+ All Categories
Home > Documents > UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the...

UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the...

Date post: 09-Jul-2020
Category:
Upload: others
View: 4 times
Download: 0 times
Share this document with a friend
19
UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl) UvA-DARE (Digital Academic Repository) Entropy driven separations in nanoporous materials Torres Knoop, A. Link to publication Citation for published version (APA): Torres Knoop, A. (2016). Entropy driven separations in nanoporous materials. General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons). Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible. Download date: 10 Aug 2020
Transcript
Page 1: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

UvA-DARE is a service provided by the library of the University of Amsterdam (httpdareuvanl)

UvA-DARE (Digital Academic Repository)

Entropy driven separations in nanoporous materials

Torres Knoop A

Link to publication

Citation for published version (APA)Torres Knoop A (2016) Entropy driven separations in nanoporous materials

General rightsIt is not permitted to download or to forwarddistribute the text or part of it without the consent of the author(s) andor copyright holder(s)other than for strictly personal individual use unless the work is under an open content license (like Creative Commons)

DisclaimerComplaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests please let the Library know statingyour reasons In case of a legitimate complaint the Library will make the material inaccessible andor remove it from the website Please Askthe Library httpsubauvanlencontact or a letter to Library of the University of Amsterdam Secretariat Singel 425 1012 WP AmsterdamThe Netherlands You will be contacted as soon as possible

Download date 10 Aug 2020

CHAPTER 7

Entropic Separation of StyreneEthylbenzene Mixtures lowast

71 Introduction

Styrene is an important feedstock in the petrochemical industry The reactivity ofits vinyl group makes styrene easy to polymerize and copolymerize and thereforeit serves as raw material for the production of a great variety of materials thetwo most important being polystyrene and rubber [1] Although styrene appearsin small quantities in nature the global consumption (of the order of millions oftons per year) requires its commercial production There are two main methodsto obtain styrene dehydrogenation of ethylbenzene and co-production of styreneand propylene oxide via hydroperoxidation of ethylbenzene Direct dehydrogena-tion of ethylbenzene to styrene accounts for the majority of the production Theconventional method involves two steps the alkylation of benzene with ethyleneto produce ethylbenzene and the dehydrogenation of the ethylbenzene to producestyrene Complete conversion is not achieved in the reactor and therefore theproduct stream contains a large fraction of ethylbenzene that has to be removed

The preferred technology for the ethylbenzenestyrene separation nowadays isextractive distillation [2] and vacuum distillation [3 4] together with inhibitorslike phenylene-diamines or dinitrophenols to avoid styrene polymerization

However because of the similarity in the boiling point of styrene (418 K) andethylbenzene (409 K) this process is energetically expensive and most of the en-ergy needed for the production of styrene is used in the separation process The

lowastBased on A Torres-Knoop J Heinen R Krishna and D Dubbeldam Entropic Separationof StyreneEthylbenzene Mixtures by Exploitation of Subtle Differences in Molecular Configur-ations in Ordered Crystalline Nanoporous Adsorbents Langmuir 31(12) 2015 3771-3778

145

146 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

process is even more complicated due of the presence of side products like tolueneo-xylene and benzene

An alternative energy-efficient separation strategy involves utilizing the mo-lecular chemical and geometrical differences by means of adsorptive separationwith nanoporous materials like metal-organic frameworks and zeolites Ahmad etal [5] performed liquid chromatography separation using HKUST-1(Cu3(BTC)2)a metal-organic framework with open Cu(II) sites and 135-benzenetricarboxylate(BTC) linkers They found that styrene is preferentially adsorbed in the struc-ture because of the coordinative interaction of styrene with the Cu(II) in a π-complexation mechanism Maes et al [6 7] and Remy et al [8] reported resultson MIL-47(V) and MIL-53(Al) showing both structures are capable of separationin the liquid phase They found that in MIL-47(V) styrene selectivity is relatedto styrene capacity for packing while for MIL-53(Al) styrene selectivity is relatedto adsorption enthalpy (interaction with the carboxylate) For competitive ad-sorption in static conditions they reported separation factors of 36 and 41 forMIL-47(V) and MIL-53(Al) respectively and for an equimolar mixture in dynamicconditions (breakthrough experiments using a column filled with crystallites in anHPLC apparatus) they found separation factors of 29 and 23 They also observedthat if a more realistic mixture is taken into account (with toluene and o-xylene) inMIL-53 o-xylene and toluene are retained even longer which makes the materialgood for impurity removal Yang et al [9] conducted experiments on stationaryphase HPLC with MIL-101(Cr) a material built from a hybrid supertetrahedralbuilding unit formed by terephthalate ligands and trimeric chromium octahedralclusters Similar to HKUST-1 they reported a higher affinity towards styrene dueto the π-π interactions with the metal-organic framework walls and the unsat-urated metal sites They also reported the efficient separation of impurities likeo-xylene and toluene

Separation based on adsorption relies on either adsorption or diffusion charac-teristics At low loadings (ie the Henry regime) the selectivity is mainly drivenby enthalpic effects and favors the molecule with the strongest interaction withthe framework Selectivity is therefore strongly related to adsorbent and adsorbateproperties such as dipole moment polarizability quadrupole moment and mag-netic susceptibility At saturation conditions (industrial set-up) the selectivity isdriven by either enthalpic effects andor entropic effects like

bull commensurate freezing [10] which favors molecules which efficiently pack inintersecting-channels structures

bull size entropy [11 12] which favors the smallest molecules

bull length entropy [11 13ndash15] which favors the molecules with the shortest ef-fective length (footprint) in one-dimensional channels

bull commensurate stacking [16] which favors molecules with stacking arrange-ments that are commensurate with the dimensions of one-dimensional chan-nels

72 Methodology 147

bull face-to-face stacking [17] which favors molecules that when reoriented sig-nificantly reduce their footprint in one-dimensional channels

The various separation strategies for exploitation of molecular packing effects havebeen reviewed recently [18]

Styrene and ethylbenzene are very similar molecules the main difference beingthat styrene is a flat molecule whereas ethylbenzene is not Finding structures withselective adsorption for styrene is not easy In this work we present a screeningstudy for the separation of styrene and ethylbenzene at liquid conditions Wepropose to separate on the basis of a difference in saturation loading because it ismore cost-efficient and utilizes the pore volume most efficiently

72 Methodology

The systems were modeled using classical force fields The adsorbates were modeledwith OPLS-AA force field for organic liquids [19] In previous work [16] we haveshown that the use of these force fields is in good agreement with experimentsBecause we were interested in the selectivity of planarnon-planar molecules andnot in their conformational changes adsorbates were described as multisite rigidmolecules with properties and configurations shown in Figure 71 The parametersfor the interaction of the adsorbates (Lennard-Jones and electrostatic interactions)together with a schematic representation of the molecules showing the atom typesare presented in Table 71 Cross-interactions with other molecules and the frame-work were computed using Lorentz-Berthelot mixing rules

Boiling point 418 KFreezing point 2425 Klength 096 nmwidth 070 nmheight 034 nm

Boiling point 409 KFreezing point 178 Klength 095 nmwidth 067 nmheight 053 nm

Figure 71 Styrene (top) and ethylbenzene (bottom) configurations The figureshows the typical properties of the modeled adsorbates Distances are ldquomolecularshadow lengthsrdquo [20] from Materials Studio [21] Besides small differences in thecharges the main difference between these molecules is their height (planarity)

148 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

atomgroup q [eminus] σ [A] ε [kcal molminus1]C -0115 355 0070C1 -0115 355 0076C2 -0005 350 0066C3 -0180 350 0066C4 -0000 355 0076C5 -0115 355 0076C6 -0148 355 0076H 0115 242 0030H1 0060 242 0030H2 0074 242 0030

Table 71 OPLS-AA force field parameters for styrene and ethylbenzene [19] Thevinyl group charges () were taken from Shirley et al [22]

The frameworks were modeled as rigid with atom positions taken from crystal-lographic experimental data Most MOFs were further optimized using VASP [2324] with the cell fixed to the experimentally determined unit cell size and shape(PBE [25 26] exchange-correlation functional with dispersion corrections [27] wasused and the PAW method was applied to describe the core atoms convergencecriteria of the ionic forces was set to -1times10minus3 AeV) The metal-organic frame-works were modeled using the DREIDING force field [28] and Van der Waals para-meters not found in DREIDING were taken from the universal force field (UFF)[29] DREIDING and UFF force fields were designed to be very generic so thatbroad coverage of the periodic table including inorganic compounds metals andtransition metals could be achieved UFF was tailored for simulating moleculescontaining any combination of elements in the periodic table For the zeolites theTraPPE [30] force field was used This force field was specifically developed forzeolites

The charge-charge interactions were computed using the Ewald summation(relative precision 10minus6) Charges for the frameworks were computed by minimiz-ing the difference of the classical electrostatic potential and a quantum mechanicselectrostatic potential over many grid points using the REPEAT method [31 32]

73 Adsorption isotherms

To compute the adsorption isotherms we perform Monte Carlo simulation in thegrand-canonical ensemble (or microV T ensemble) In this ensemble the number ofadsorbates fluctuates until equilibrium conditions are reached the temperatureand chemical potential of the gas inside and outside the adsorbent are equalBecause in confined systems the fraction of successful insertions and deletions isvery low reaching equilibrium with conventional Monte Carlo methods can bevery time consuming In this study we used the Configurational Bias ContinuousFractional Monte Carlo (CBCFCMC) [33] method to enhance the success rate of

74 Results 149

insertions and deletions The method is a combination of the Configurational BiasMonte Carlo (CBMC) [34ndash36] where molecules growth is biased towards favorableconfigurations and Continuous Fractional Component Monte Carlo (CFCMC) [37]in which molecules are gradually inserted or deleted by scaling their interactionswith the surroundings We have shown in previous work [33] that the resultsobtained with this method do not differ from CBMC calculations but the efficiencyis higher

Using the dual-site Langmuir-Freundlich fits of the pure component isothermsbreakthrough calculations were carried out by solving a set of partial differentialequations for each of the species in the gas mixture [38 39] The molar loadingsof the species at any position along the packed bed and at any time were determ-ined from Ideal Adsorbed Solution Theory calculations Video animations of thebreakthrough behavior as a function of time of selected structures are provided asweb-enhanced objects online

74 Results

We perform a screening study of several zeolites and metal-organic frameworksfor the separation of styreneethylbenzene mixture focusing on saturation condi-tions Under these conditions differences in the saturation capacity of the mixturecomponents strongly dictate the separation

In systems with small pores like MRE and MTW zeolites molecules are forcedto adsorb parallel to the channels The saturation capacity is determined by theeffective length per molecule in the channel (footprint) Because of the similarityin the length of styrene and ethylbenzene the difference in saturation capacities isalmost negligible making systems with small pores unsuitable candidates for theseparation

In structures with cavities or channels much larger than styrene and ethyl-benzene molecular dimensions like IRMOF-1 and Zn-DOBDC molecules do notpresent any particular packing The observed difference in the saturation capa-cities is a consequence of the natural packing of the molecules in liquid phase(ρEb = 08665 gmL ρSt = 0909 gmL) This makes these materials also unsuit-able for the separation process

We have identified a few materials where styrene has a higher saturation capa-city than ethylbenzene In the following we describe how this difference arises fromthe previously mentioned entropic mechanisms and we highlight their applicabilityfor the separation process

Size exclusion is observed in MFI-para [40] MFI-para is a ZSM-5 zeolite whichstructure is a combination of interconnected straight and zigzag channels Thestraight channels have a diameter of 53times56 A and the zigzag channels have adiameter of 51times55 A In Figure 72b the simulated single component isothermsof styrene and ethylbenzene in MFI-para at 433 K and snapshots of styrene andethylbenzene at 1times109 Pa and 433 K are presented At low loadings molecules

150 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

preferentially adsorb in the straight channels the difference in loadings arise froma stronger interaction of styrene with MFI-para At saturation conditions styrenecan obtain almost twice the loading of ethylbenzene because of a size exclusioneffect in the zig-zag channels in which ethylbenzene does not fit due to its heightWhen an equimolar styreneethylbenzene mixture is considered the difference inloadings at saturation conditions is even larger (Figure 72c)

a

0

02

04

06

08

1

12

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

b

0

02

04

06

08

1

12

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

c

Figure 72 (a) Snapshots of ethylbenzene (top) and styrene (bottom) at 1times109

Pa and 433 K (b) Single component isotherms of styrene (red) and ethylbenzene(blue) in MFI-para at 433 K (lines are dual-site Langmuir-Freundlich fits of the purecomponents points are the pure component isotherms from CBCFCMC simulations)(c) Mixture component isotherms for an equimolar mixture in MFI-para at 433 K

Face-to-face stacking occurs in MAZ [41] and AFI [42] zeolites MAZ and AFIare 1D-channel zeolites with dimensions that allow a molecular reorientation ofethylbenzene and styrene

In Figure 73 we present the simulation results for the single component iso-therms of ethylbenzene and styrene in AFI zeolite at 433 K At low loadingsmolecules are mostly adsorbed flat on the walls (parallel to the channels axis) ad-sorption is dictated by enthalpy effects which favors ethylbenzene As the loadingincreases the molecules undergo a molecular reorientation from a configurationparallel to the channel axis to a ldquopilerdquo configuration with the phenyl groups fa-cing each other This new configuration (face-to-face stacking) reduces styrenersquosfootprint more than ethylbenzenersquos footprint (Figure 76a) because of the non-planarity of ethylbenzene allowing styrene to obtain higher saturation loadings

For MAZ zeolite a similar behavior is observed however because MAZ zeolitehas smaller channels than AFI the angle at which molecules can reorient has asmaller effect on the reduction of the moleculersquos footprint in the channels (Figure76b) When an equimolar mixture is considered both AFI and MAZ zeolites arestyrene selective at saturation conditions as shown in Figures 74a and 74b wherethe simulated mixture component isotherms at 433 K are presented

74 Results 151

0

02

04

06

08

1

12

14

100

101

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

Pure component isotherms CBCFCMC AFI 433K

ethylbenzenestyrene

Figure 73 Single component isotherms of styrene and ethylbenzene in AFI at 433 KInset snapshots of styrene (top) and ethylbenzene (bottom) at 1times103 Pa and 1times109

Pa At low loadings both molecules are adsorbed with the phenyl group parallel to thechannel axis At higher pressures styrene can arrange in a tilted face-to-face stackingconfiguration which reduces its footprint and allows for a higher saturation capacityThe difference in saturation capacities ensures the selectivity of AFI towards styreneat saturation conditions in the mixture

0

01

02

03

04

05

06

07

08

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

01

02

03

04

05

06

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

Figure 74 Mixture component isotherms for and equimolar mixture of styrene andethylbenzene at 433 K in (a) AFI zeolite and (b) MAZ zeolite

Face-to-face stacking is also observed in DON zeolite DON [43] is a structurewith 1D-channels slightly larger than AFI and MAZ zeolites This increase indimensions favors ethylbenzene to undergo a molecular reorientation into a face-to-face stacking configuration but also because the diameter of the channels arelarger than the length of styrene it induces styrene to adopt a commensurate stack-ing configuration where the stacking of two molecules with their phenyl groupsfacing each other is commensurate with the channel dimensions (Figures 75a)This is also observed in MIL-53 [44] a metal-organic framework with lozenge-shaped rhombohedric channels of approximately 085 nm as show in Figure 75b

152 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

a

b

Figure 75 (a) Snapshots of styrene (left) and ethylbenzene (right) in DON zeolite at1e9 Pa and 433 K(b) Snapshots of styrene (left) and ethylbenzene (right) in MIL-53at 1e9 Pa and 433 K Color code carbon (cyan) hydrogen (white)

where snapshots of styrene and ethylbenzene at 1times109 Pa and 433 K are presen-ted Because of styrene and ethylbenzene dimensions styrene with commensuratestacking can obtain higher saturation capacities than ethylbenzene with face-to-face stacking as shown schematically in Figure 76c This is also seen in DONand MIL-53 single component isotherms (Figure 77a) Mixture isotherms (Fig-ures 77b) and breakthrough curves (Figure 77c) further confirm that DON andMIL-53 are styrene selective structures

Commensurate stacking for both styrene and ethylbenzene is observed in MIL-47 [45] and MAF-X8 [46] MIL-47 is a metal-organic framework with lozenge-shaped rhombohedric channels which size (slightly larger than MIL-53) allows forethylbenzene to also have commensurate stacking In Figure 78 we present thesimulated single component adsorption isotherms of ethylbenzene and styrene inMIL-47 at 433 K and snapshots of styrene and ethylbenzene at 1times106 Pa and1times109 Pa At 1times109 Pa both molecules have commensurate stacking but in orderfor ethylbenzene to have commensurate stacking the phenyl groups have to beslightly shifted due to the non-planar ethyl group This shift affects the amount ofethylbenzene molecules that can be adsorbed (as compared to styrene) and there-fore causes a difference in the saturation capacity of styrene and ethylbenzeneInterestingly commensurate stacking occurs at lower loading for styrene than forethylbenzene We can see in Figure 78 that at 1times106 Pa styrene already presentscommensurate stacking while ethylbenzene seems to have more of a face-to-facestacking In a mixture this will favor styrene adsorption even more In Figure79 simulated mixture adsorption isotherms for an equimolar mixture in MIL-47at 433 K are presented MIL-47 is a styrene selective structure Our results are ingood qualitative agreement with Maes et al [6] Breakthrough simulations furtherconfirm that MIL-47 is a styrene selective structure at saturation conditions InMAF-x8 a metal-organic framework with square channels of approximately 1nm

74 Results 153

BEFORE REORIENTATION

AFTER REORIENTATION

096 nm 095 nm

034 nm 053 nm

a b

095 nm

053 nm x2

c

Figure 76 Schematic representation of different entropic effects occurring in theseparation of styrene and ethylbenzene in nanoporous materials (a) The reductionof the moleculersquos footprint in the channels because of a reorientation into a face-to-face stacking configuration (b) Effect of the channel size on the reorientationand therefore moleculersquos footprint reduction (c) Comparison of the channel lengthneeded for two molecules of styrene to have commensurate stacking vs two moleculesof ethylbenzene to have face-to-face stacking

0

05

1

15

2

25

3

35

10-1

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

0

05

1

15

2

25

3

100

101

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

0

05

1

15

2

0 05 1 15 2 25 3 35 4 45

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

0

02

04

06

08

1

12

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

a

0

01

02

03

04

05

06

07

08

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

0

05

1

15

2

0 05 1 15 2 25 3

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

c

Figure 77 Simulated isotherms and breakthrough curves of styrene and ethylben-zene in MIL-53 (top) and DON (bottom) at 433 K (a) Pure component isotherms(lines are dual-site Langmuir-Freundlich fits of the pure components points are thepure component isotherms from CBCFCMC simulations) (b) Mixture componentisotherms for an equimolar mixture The IAST results are in good agreement with themixture isotherms (c) Simulated step-type breakthrough at 1e6 Pa total fugacity

154 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

0

1

2

3

4

5

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

Figure 78 Single component adsorption isotherms for styrene and ethylbenzeneat 433 K in MIL-47 Inset styrene (top) and ethylbenzene (bottom) snapshots at1times106 Pa and 1times109 Pa Styrene has commensurate stacking at lower pressuresthan ethylbenzene Because of the out of plane ethyl group ethylbenzene moleculeshave to be slightly shifted in order to have commensurate stacking and therefore lessethylbenzene molecules can be adsorbed in MIL-47 channels at saturation conditions

0

05

1

15

2

25

3

35

4

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

02

04

06

08

1

12

14

16

18

0 1 2 3 4 5

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

b

Figure 79 Styreneethylbenzene separation using MIL-47 at 433 K (a) equimolarmixture isotherms and Ideal Adsorption Solution Theory (IAST) prediction based onpure component isotherms (b) simulated step breakthrough at 1times106 Pa total fu-gacity The IAST prediction is in excellent agreement with the mixture simulationsThe mixture and breakthrough simulations show a high styrene selectivity and loadingin the mixture

it is easier for ethylbenzene to have commensurate stacking than in MIL-47 Thesingle component isotherms of both molecules behave very similar The topologyof the structure seems to induce a shift between parallel styrene molecules and al-lows for the ethyl group of ethylbenzene to stick in the channel ldquopocketsrdquo (Figure710) This might be the reason for the smaller difference in saturation capacitiesof styrene and ethylbenzene compared to MIL-47

74 Results 155

a b

0

05

1

15

2

25

3

35

4

10-6

10-4

10-2

100

102

104

106

108

Abso

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

c

Figure 710 Snapshots of (a) styrene and (b) ethylbenzene in MAF-X8 at 1e9 Paand 433 K Both styrene and ethylbenzene have commensurate stacking Color codecarbon (cyan) hydrogen (white) (c) Simulated single component isotherms of styreneand ethylbenzene in MAF-X8 at 433 K

Combinationcompetition of mechanisms is observed in structures with a more com-plex topology An example of this is MOF-CJ3 metal-organic framework [47] Thewide segment of the channels are big enough to allow for both styrene and ethyl-benzene to form two parallel rows (commensurate stacking) however the shapeof the channels forces ethylbenzene molecules to adopt a configuration where theethyl group is pointing to the channel ldquopocketsrdquo Styrene has more freedom andmolecules can arrange in a way that an extra styrene can be adsorbed in theprotracted segments in a configuration perpendicular to the channel (face-to-facestacking) (Figure 711) The interplay between different mechanisms rarely makesthe separation better Even when all the mechanisms favor a specific molecule thecompetition between them can induce enough disorder to destroy the selectivityone could achieve with a ldquopurerdquo mechanism

Figure 711 Snapshots of styrene (top) and ethylbenzene (bottom) at 1times109 Paand 433 K in MOF-CJ3

156 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

75 Discussion

There are two important factors to consider when using adsorption for separa-tion processes at industrial conditions namely selectivity and capacity A highselectivity ensures that less cycles are needed to achieve a high degree of purity inthe separation but a high capacity implies that the regeneration time is longerFor a binary mixture the adsorption selectivity is defined as

Sads =q1q2

f1f2(71)

and the capacity is defined as the styrene loading in the adsorbed phase of a binarymixture

Capacity = q1 (72)

In Figure 712 the relationship between these two properties for different struc-tures is presented Structures with the same separation mechanism are plottedwith the same color An ideal structure for the separation would be located at theright top corner

1

10

0 1 2 3 4 5 6

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

MAF-x8

MOF-CJ3

CoBDP IRMOF-1

JUC-77

CuBTC

UiO-66

Figure 712 Selectivity as a function of styrene loading (capacity) in a binary mix-ture at 433 K and 1times106 Pa total fugacity The structures are divided in differentcolors depending on the selectivity mechanism observed Color code Size exclusion(red) face-to-face stacking (purple) commensurate stacking (blue) commensurate-stackingface-to-face stacking (green) mixed(orange) The dotted red line corres-ponds to the ratio styreneethylbenzene at liquid conditions There is a naturaltrade-off between selectivity and loading finding structures in the top right corneris not feasible The black dashed line (guide to the eye) denotes the inverse relation-ship between selectivity and capacity MIL-47 is a styrene selective material with ahigh capacity therefore a good candidate for the styreneethylbenzene separation

75 Discussion 157

In structures where the separation is driven by size-exclusion (MFI-para) theselectivity is high because there is an adsorption site available only for styrene(zig-zag channels) but the capacity is rather low since for size-exclusion to occuramong similar molecules there has to be a very tight fit between the moleculesand the adsorbent usually associated with small pore systems

In structures where the separation is driven by face-to-face stacking the se-lectivity relies on the increase of the moleculersquos footprints difference consequenceof a reorientation and piling Ideally only one of the mixture components shouldbe able to reorient but in the case of styrene and ethylbenzene the similarity intheir length and width makes no significant difference in the pore size needed forthe reorientation The selectivity relies thus in how favorable the reorientationand piling are

In MAZ and AFI both molecules can reorient but the reorientation is restrictedby the pore size forcing a tilting in the face-to-face stacking configuration that isunfavorable for ethylbenzene to form a pile but it is favorable for styrene Thisallows styrene to obtain higher saturation loadings and ensures MAZ and AFIstyrene selectivity at saturation conditions

However face-to-face stacking only occurs at high loadings In Figure 712the selectivity is presented at 1times106 Pa At this fugacity MAZ is not yet styreneselective and AFI selectivity is almost negligible The effect of face-to-face stackingin styrene selectivity can only be observed at higher loadings as shown in Figure713 where the selectivity as a function of styrene loading for higher fugacities ispresented

1

10

1 10

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

CoBDP

CuBTC

IRMOF-1

JUC-77

MAF-X8

MIL-47

MIL-53

MOF-CJ3

AFI

DON

MAZ

MFI-para

UiO-66

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

JUC-77

MAF-x8

IRMOF-1

MOFCJ3

CoBDP

CuBTC

UiO-66

MIL-47

MIL-53

JUC-77

CuBTC

CoBDP

MOF-CJ3 IRMOF-1

MAZDON

MFI-para

AFI

UiO-66

MAF-X8

1e6 Pa1e7 Pa1e9 Pa

Figure 713 Selectivity (qstqeb) in an equimolar mixture at 3 different fugacitiesThe dashed red line corresponds to the liquid phase ratio For all the cases styreneloading increases with the fugacity (shift to the right of the plot) The selectivity alsoincreases in most of the cases with the fugacity (shift upwards) except for MFI-paraJUC-77 MIL-47 and Cu-BTC

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 2: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

CHAPTER 7

Entropic Separation of StyreneEthylbenzene Mixtures lowast

71 Introduction

Styrene is an important feedstock in the petrochemical industry The reactivity ofits vinyl group makes styrene easy to polymerize and copolymerize and thereforeit serves as raw material for the production of a great variety of materials thetwo most important being polystyrene and rubber [1] Although styrene appearsin small quantities in nature the global consumption (of the order of millions oftons per year) requires its commercial production There are two main methodsto obtain styrene dehydrogenation of ethylbenzene and co-production of styreneand propylene oxide via hydroperoxidation of ethylbenzene Direct dehydrogena-tion of ethylbenzene to styrene accounts for the majority of the production Theconventional method involves two steps the alkylation of benzene with ethyleneto produce ethylbenzene and the dehydrogenation of the ethylbenzene to producestyrene Complete conversion is not achieved in the reactor and therefore theproduct stream contains a large fraction of ethylbenzene that has to be removed

The preferred technology for the ethylbenzenestyrene separation nowadays isextractive distillation [2] and vacuum distillation [3 4] together with inhibitorslike phenylene-diamines or dinitrophenols to avoid styrene polymerization

However because of the similarity in the boiling point of styrene (418 K) andethylbenzene (409 K) this process is energetically expensive and most of the en-ergy needed for the production of styrene is used in the separation process The

lowastBased on A Torres-Knoop J Heinen R Krishna and D Dubbeldam Entropic Separationof StyreneEthylbenzene Mixtures by Exploitation of Subtle Differences in Molecular Configur-ations in Ordered Crystalline Nanoporous Adsorbents Langmuir 31(12) 2015 3771-3778

145

146 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

process is even more complicated due of the presence of side products like tolueneo-xylene and benzene

An alternative energy-efficient separation strategy involves utilizing the mo-lecular chemical and geometrical differences by means of adsorptive separationwith nanoporous materials like metal-organic frameworks and zeolites Ahmad etal [5] performed liquid chromatography separation using HKUST-1(Cu3(BTC)2)a metal-organic framework with open Cu(II) sites and 135-benzenetricarboxylate(BTC) linkers They found that styrene is preferentially adsorbed in the struc-ture because of the coordinative interaction of styrene with the Cu(II) in a π-complexation mechanism Maes et al [6 7] and Remy et al [8] reported resultson MIL-47(V) and MIL-53(Al) showing both structures are capable of separationin the liquid phase They found that in MIL-47(V) styrene selectivity is relatedto styrene capacity for packing while for MIL-53(Al) styrene selectivity is relatedto adsorption enthalpy (interaction with the carboxylate) For competitive ad-sorption in static conditions they reported separation factors of 36 and 41 forMIL-47(V) and MIL-53(Al) respectively and for an equimolar mixture in dynamicconditions (breakthrough experiments using a column filled with crystallites in anHPLC apparatus) they found separation factors of 29 and 23 They also observedthat if a more realistic mixture is taken into account (with toluene and o-xylene) inMIL-53 o-xylene and toluene are retained even longer which makes the materialgood for impurity removal Yang et al [9] conducted experiments on stationaryphase HPLC with MIL-101(Cr) a material built from a hybrid supertetrahedralbuilding unit formed by terephthalate ligands and trimeric chromium octahedralclusters Similar to HKUST-1 they reported a higher affinity towards styrene dueto the π-π interactions with the metal-organic framework walls and the unsat-urated metal sites They also reported the efficient separation of impurities likeo-xylene and toluene

Separation based on adsorption relies on either adsorption or diffusion charac-teristics At low loadings (ie the Henry regime) the selectivity is mainly drivenby enthalpic effects and favors the molecule with the strongest interaction withthe framework Selectivity is therefore strongly related to adsorbent and adsorbateproperties such as dipole moment polarizability quadrupole moment and mag-netic susceptibility At saturation conditions (industrial set-up) the selectivity isdriven by either enthalpic effects andor entropic effects like

bull commensurate freezing [10] which favors molecules which efficiently pack inintersecting-channels structures

bull size entropy [11 12] which favors the smallest molecules

bull length entropy [11 13ndash15] which favors the molecules with the shortest ef-fective length (footprint) in one-dimensional channels

bull commensurate stacking [16] which favors molecules with stacking arrange-ments that are commensurate with the dimensions of one-dimensional chan-nels

72 Methodology 147

bull face-to-face stacking [17] which favors molecules that when reoriented sig-nificantly reduce their footprint in one-dimensional channels

The various separation strategies for exploitation of molecular packing effects havebeen reviewed recently [18]

Styrene and ethylbenzene are very similar molecules the main difference beingthat styrene is a flat molecule whereas ethylbenzene is not Finding structures withselective adsorption for styrene is not easy In this work we present a screeningstudy for the separation of styrene and ethylbenzene at liquid conditions Wepropose to separate on the basis of a difference in saturation loading because it ismore cost-efficient and utilizes the pore volume most efficiently

72 Methodology

The systems were modeled using classical force fields The adsorbates were modeledwith OPLS-AA force field for organic liquids [19] In previous work [16] we haveshown that the use of these force fields is in good agreement with experimentsBecause we were interested in the selectivity of planarnon-planar molecules andnot in their conformational changes adsorbates were described as multisite rigidmolecules with properties and configurations shown in Figure 71 The parametersfor the interaction of the adsorbates (Lennard-Jones and electrostatic interactions)together with a schematic representation of the molecules showing the atom typesare presented in Table 71 Cross-interactions with other molecules and the frame-work were computed using Lorentz-Berthelot mixing rules

Boiling point 418 KFreezing point 2425 Klength 096 nmwidth 070 nmheight 034 nm

Boiling point 409 KFreezing point 178 Klength 095 nmwidth 067 nmheight 053 nm

Figure 71 Styrene (top) and ethylbenzene (bottom) configurations The figureshows the typical properties of the modeled adsorbates Distances are ldquomolecularshadow lengthsrdquo [20] from Materials Studio [21] Besides small differences in thecharges the main difference between these molecules is their height (planarity)

148 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

atomgroup q [eminus] σ [A] ε [kcal molminus1]C -0115 355 0070C1 -0115 355 0076C2 -0005 350 0066C3 -0180 350 0066C4 -0000 355 0076C5 -0115 355 0076C6 -0148 355 0076H 0115 242 0030H1 0060 242 0030H2 0074 242 0030

Table 71 OPLS-AA force field parameters for styrene and ethylbenzene [19] Thevinyl group charges () were taken from Shirley et al [22]

The frameworks were modeled as rigid with atom positions taken from crystal-lographic experimental data Most MOFs were further optimized using VASP [2324] with the cell fixed to the experimentally determined unit cell size and shape(PBE [25 26] exchange-correlation functional with dispersion corrections [27] wasused and the PAW method was applied to describe the core atoms convergencecriteria of the ionic forces was set to -1times10minus3 AeV) The metal-organic frame-works were modeled using the DREIDING force field [28] and Van der Waals para-meters not found in DREIDING were taken from the universal force field (UFF)[29] DREIDING and UFF force fields were designed to be very generic so thatbroad coverage of the periodic table including inorganic compounds metals andtransition metals could be achieved UFF was tailored for simulating moleculescontaining any combination of elements in the periodic table For the zeolites theTraPPE [30] force field was used This force field was specifically developed forzeolites

The charge-charge interactions were computed using the Ewald summation(relative precision 10minus6) Charges for the frameworks were computed by minimiz-ing the difference of the classical electrostatic potential and a quantum mechanicselectrostatic potential over many grid points using the REPEAT method [31 32]

73 Adsorption isotherms

To compute the adsorption isotherms we perform Monte Carlo simulation in thegrand-canonical ensemble (or microV T ensemble) In this ensemble the number ofadsorbates fluctuates until equilibrium conditions are reached the temperatureand chemical potential of the gas inside and outside the adsorbent are equalBecause in confined systems the fraction of successful insertions and deletions isvery low reaching equilibrium with conventional Monte Carlo methods can bevery time consuming In this study we used the Configurational Bias ContinuousFractional Monte Carlo (CBCFCMC) [33] method to enhance the success rate of

74 Results 149

insertions and deletions The method is a combination of the Configurational BiasMonte Carlo (CBMC) [34ndash36] where molecules growth is biased towards favorableconfigurations and Continuous Fractional Component Monte Carlo (CFCMC) [37]in which molecules are gradually inserted or deleted by scaling their interactionswith the surroundings We have shown in previous work [33] that the resultsobtained with this method do not differ from CBMC calculations but the efficiencyis higher

Using the dual-site Langmuir-Freundlich fits of the pure component isothermsbreakthrough calculations were carried out by solving a set of partial differentialequations for each of the species in the gas mixture [38 39] The molar loadingsof the species at any position along the packed bed and at any time were determ-ined from Ideal Adsorbed Solution Theory calculations Video animations of thebreakthrough behavior as a function of time of selected structures are provided asweb-enhanced objects online

74 Results

We perform a screening study of several zeolites and metal-organic frameworksfor the separation of styreneethylbenzene mixture focusing on saturation condi-tions Under these conditions differences in the saturation capacity of the mixturecomponents strongly dictate the separation

In systems with small pores like MRE and MTW zeolites molecules are forcedto adsorb parallel to the channels The saturation capacity is determined by theeffective length per molecule in the channel (footprint) Because of the similarityin the length of styrene and ethylbenzene the difference in saturation capacities isalmost negligible making systems with small pores unsuitable candidates for theseparation

In structures with cavities or channels much larger than styrene and ethyl-benzene molecular dimensions like IRMOF-1 and Zn-DOBDC molecules do notpresent any particular packing The observed difference in the saturation capa-cities is a consequence of the natural packing of the molecules in liquid phase(ρEb = 08665 gmL ρSt = 0909 gmL) This makes these materials also unsuit-able for the separation process

We have identified a few materials where styrene has a higher saturation capa-city than ethylbenzene In the following we describe how this difference arises fromthe previously mentioned entropic mechanisms and we highlight their applicabilityfor the separation process

Size exclusion is observed in MFI-para [40] MFI-para is a ZSM-5 zeolite whichstructure is a combination of interconnected straight and zigzag channels Thestraight channels have a diameter of 53times56 A and the zigzag channels have adiameter of 51times55 A In Figure 72b the simulated single component isothermsof styrene and ethylbenzene in MFI-para at 433 K and snapshots of styrene andethylbenzene at 1times109 Pa and 433 K are presented At low loadings molecules

150 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

preferentially adsorb in the straight channels the difference in loadings arise froma stronger interaction of styrene with MFI-para At saturation conditions styrenecan obtain almost twice the loading of ethylbenzene because of a size exclusioneffect in the zig-zag channels in which ethylbenzene does not fit due to its heightWhen an equimolar styreneethylbenzene mixture is considered the difference inloadings at saturation conditions is even larger (Figure 72c)

a

0

02

04

06

08

1

12

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

b

0

02

04

06

08

1

12

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

c

Figure 72 (a) Snapshots of ethylbenzene (top) and styrene (bottom) at 1times109

Pa and 433 K (b) Single component isotherms of styrene (red) and ethylbenzene(blue) in MFI-para at 433 K (lines are dual-site Langmuir-Freundlich fits of the purecomponents points are the pure component isotherms from CBCFCMC simulations)(c) Mixture component isotherms for an equimolar mixture in MFI-para at 433 K

Face-to-face stacking occurs in MAZ [41] and AFI [42] zeolites MAZ and AFIare 1D-channel zeolites with dimensions that allow a molecular reorientation ofethylbenzene and styrene

In Figure 73 we present the simulation results for the single component iso-therms of ethylbenzene and styrene in AFI zeolite at 433 K At low loadingsmolecules are mostly adsorbed flat on the walls (parallel to the channels axis) ad-sorption is dictated by enthalpy effects which favors ethylbenzene As the loadingincreases the molecules undergo a molecular reorientation from a configurationparallel to the channel axis to a ldquopilerdquo configuration with the phenyl groups fa-cing each other This new configuration (face-to-face stacking) reduces styrenersquosfootprint more than ethylbenzenersquos footprint (Figure 76a) because of the non-planarity of ethylbenzene allowing styrene to obtain higher saturation loadings

For MAZ zeolite a similar behavior is observed however because MAZ zeolitehas smaller channels than AFI the angle at which molecules can reorient has asmaller effect on the reduction of the moleculersquos footprint in the channels (Figure76b) When an equimolar mixture is considered both AFI and MAZ zeolites arestyrene selective at saturation conditions as shown in Figures 74a and 74b wherethe simulated mixture component isotherms at 433 K are presented

74 Results 151

0

02

04

06

08

1

12

14

100

101

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

Pure component isotherms CBCFCMC AFI 433K

ethylbenzenestyrene

Figure 73 Single component isotherms of styrene and ethylbenzene in AFI at 433 KInset snapshots of styrene (top) and ethylbenzene (bottom) at 1times103 Pa and 1times109

Pa At low loadings both molecules are adsorbed with the phenyl group parallel to thechannel axis At higher pressures styrene can arrange in a tilted face-to-face stackingconfiguration which reduces its footprint and allows for a higher saturation capacityThe difference in saturation capacities ensures the selectivity of AFI towards styreneat saturation conditions in the mixture

0

01

02

03

04

05

06

07

08

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

01

02

03

04

05

06

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

Figure 74 Mixture component isotherms for and equimolar mixture of styrene andethylbenzene at 433 K in (a) AFI zeolite and (b) MAZ zeolite

Face-to-face stacking is also observed in DON zeolite DON [43] is a structurewith 1D-channels slightly larger than AFI and MAZ zeolites This increase indimensions favors ethylbenzene to undergo a molecular reorientation into a face-to-face stacking configuration but also because the diameter of the channels arelarger than the length of styrene it induces styrene to adopt a commensurate stack-ing configuration where the stacking of two molecules with their phenyl groupsfacing each other is commensurate with the channel dimensions (Figures 75a)This is also observed in MIL-53 [44] a metal-organic framework with lozenge-shaped rhombohedric channels of approximately 085 nm as show in Figure 75b

152 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

a

b

Figure 75 (a) Snapshots of styrene (left) and ethylbenzene (right) in DON zeolite at1e9 Pa and 433 K(b) Snapshots of styrene (left) and ethylbenzene (right) in MIL-53at 1e9 Pa and 433 K Color code carbon (cyan) hydrogen (white)

where snapshots of styrene and ethylbenzene at 1times109 Pa and 433 K are presen-ted Because of styrene and ethylbenzene dimensions styrene with commensuratestacking can obtain higher saturation capacities than ethylbenzene with face-to-face stacking as shown schematically in Figure 76c This is also seen in DONand MIL-53 single component isotherms (Figure 77a) Mixture isotherms (Fig-ures 77b) and breakthrough curves (Figure 77c) further confirm that DON andMIL-53 are styrene selective structures

Commensurate stacking for both styrene and ethylbenzene is observed in MIL-47 [45] and MAF-X8 [46] MIL-47 is a metal-organic framework with lozenge-shaped rhombohedric channels which size (slightly larger than MIL-53) allows forethylbenzene to also have commensurate stacking In Figure 78 we present thesimulated single component adsorption isotherms of ethylbenzene and styrene inMIL-47 at 433 K and snapshots of styrene and ethylbenzene at 1times106 Pa and1times109 Pa At 1times109 Pa both molecules have commensurate stacking but in orderfor ethylbenzene to have commensurate stacking the phenyl groups have to beslightly shifted due to the non-planar ethyl group This shift affects the amount ofethylbenzene molecules that can be adsorbed (as compared to styrene) and there-fore causes a difference in the saturation capacity of styrene and ethylbenzeneInterestingly commensurate stacking occurs at lower loading for styrene than forethylbenzene We can see in Figure 78 that at 1times106 Pa styrene already presentscommensurate stacking while ethylbenzene seems to have more of a face-to-facestacking In a mixture this will favor styrene adsorption even more In Figure79 simulated mixture adsorption isotherms for an equimolar mixture in MIL-47at 433 K are presented MIL-47 is a styrene selective structure Our results are ingood qualitative agreement with Maes et al [6] Breakthrough simulations furtherconfirm that MIL-47 is a styrene selective structure at saturation conditions InMAF-x8 a metal-organic framework with square channels of approximately 1nm

74 Results 153

BEFORE REORIENTATION

AFTER REORIENTATION

096 nm 095 nm

034 nm 053 nm

a b

095 nm

053 nm x2

c

Figure 76 Schematic representation of different entropic effects occurring in theseparation of styrene and ethylbenzene in nanoporous materials (a) The reductionof the moleculersquos footprint in the channels because of a reorientation into a face-to-face stacking configuration (b) Effect of the channel size on the reorientationand therefore moleculersquos footprint reduction (c) Comparison of the channel lengthneeded for two molecules of styrene to have commensurate stacking vs two moleculesof ethylbenzene to have face-to-face stacking

0

05

1

15

2

25

3

35

10-1

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

0

05

1

15

2

25

3

100

101

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

0

05

1

15

2

0 05 1 15 2 25 3 35 4 45

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

0

02

04

06

08

1

12

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

a

0

01

02

03

04

05

06

07

08

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

0

05

1

15

2

0 05 1 15 2 25 3

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

c

Figure 77 Simulated isotherms and breakthrough curves of styrene and ethylben-zene in MIL-53 (top) and DON (bottom) at 433 K (a) Pure component isotherms(lines are dual-site Langmuir-Freundlich fits of the pure components points are thepure component isotherms from CBCFCMC simulations) (b) Mixture componentisotherms for an equimolar mixture The IAST results are in good agreement with themixture isotherms (c) Simulated step-type breakthrough at 1e6 Pa total fugacity

154 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

0

1

2

3

4

5

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

Figure 78 Single component adsorption isotherms for styrene and ethylbenzeneat 433 K in MIL-47 Inset styrene (top) and ethylbenzene (bottom) snapshots at1times106 Pa and 1times109 Pa Styrene has commensurate stacking at lower pressuresthan ethylbenzene Because of the out of plane ethyl group ethylbenzene moleculeshave to be slightly shifted in order to have commensurate stacking and therefore lessethylbenzene molecules can be adsorbed in MIL-47 channels at saturation conditions

0

05

1

15

2

25

3

35

4

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

02

04

06

08

1

12

14

16

18

0 1 2 3 4 5

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

b

Figure 79 Styreneethylbenzene separation using MIL-47 at 433 K (a) equimolarmixture isotherms and Ideal Adsorption Solution Theory (IAST) prediction based onpure component isotherms (b) simulated step breakthrough at 1times106 Pa total fu-gacity The IAST prediction is in excellent agreement with the mixture simulationsThe mixture and breakthrough simulations show a high styrene selectivity and loadingin the mixture

it is easier for ethylbenzene to have commensurate stacking than in MIL-47 Thesingle component isotherms of both molecules behave very similar The topologyof the structure seems to induce a shift between parallel styrene molecules and al-lows for the ethyl group of ethylbenzene to stick in the channel ldquopocketsrdquo (Figure710) This might be the reason for the smaller difference in saturation capacitiesof styrene and ethylbenzene compared to MIL-47

74 Results 155

a b

0

05

1

15

2

25

3

35

4

10-6

10-4

10-2

100

102

104

106

108

Abso

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

c

Figure 710 Snapshots of (a) styrene and (b) ethylbenzene in MAF-X8 at 1e9 Paand 433 K Both styrene and ethylbenzene have commensurate stacking Color codecarbon (cyan) hydrogen (white) (c) Simulated single component isotherms of styreneand ethylbenzene in MAF-X8 at 433 K

Combinationcompetition of mechanisms is observed in structures with a more com-plex topology An example of this is MOF-CJ3 metal-organic framework [47] Thewide segment of the channels are big enough to allow for both styrene and ethyl-benzene to form two parallel rows (commensurate stacking) however the shapeof the channels forces ethylbenzene molecules to adopt a configuration where theethyl group is pointing to the channel ldquopocketsrdquo Styrene has more freedom andmolecules can arrange in a way that an extra styrene can be adsorbed in theprotracted segments in a configuration perpendicular to the channel (face-to-facestacking) (Figure 711) The interplay between different mechanisms rarely makesthe separation better Even when all the mechanisms favor a specific molecule thecompetition between them can induce enough disorder to destroy the selectivityone could achieve with a ldquopurerdquo mechanism

Figure 711 Snapshots of styrene (top) and ethylbenzene (bottom) at 1times109 Paand 433 K in MOF-CJ3

156 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

75 Discussion

There are two important factors to consider when using adsorption for separa-tion processes at industrial conditions namely selectivity and capacity A highselectivity ensures that less cycles are needed to achieve a high degree of purity inthe separation but a high capacity implies that the regeneration time is longerFor a binary mixture the adsorption selectivity is defined as

Sads =q1q2

f1f2(71)

and the capacity is defined as the styrene loading in the adsorbed phase of a binarymixture

Capacity = q1 (72)

In Figure 712 the relationship between these two properties for different struc-tures is presented Structures with the same separation mechanism are plottedwith the same color An ideal structure for the separation would be located at theright top corner

1

10

0 1 2 3 4 5 6

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

MAF-x8

MOF-CJ3

CoBDP IRMOF-1

JUC-77

CuBTC

UiO-66

Figure 712 Selectivity as a function of styrene loading (capacity) in a binary mix-ture at 433 K and 1times106 Pa total fugacity The structures are divided in differentcolors depending on the selectivity mechanism observed Color code Size exclusion(red) face-to-face stacking (purple) commensurate stacking (blue) commensurate-stackingface-to-face stacking (green) mixed(orange) The dotted red line corres-ponds to the ratio styreneethylbenzene at liquid conditions There is a naturaltrade-off between selectivity and loading finding structures in the top right corneris not feasible The black dashed line (guide to the eye) denotes the inverse relation-ship between selectivity and capacity MIL-47 is a styrene selective material with ahigh capacity therefore a good candidate for the styreneethylbenzene separation

75 Discussion 157

In structures where the separation is driven by size-exclusion (MFI-para) theselectivity is high because there is an adsorption site available only for styrene(zig-zag channels) but the capacity is rather low since for size-exclusion to occuramong similar molecules there has to be a very tight fit between the moleculesand the adsorbent usually associated with small pore systems

In structures where the separation is driven by face-to-face stacking the se-lectivity relies on the increase of the moleculersquos footprints difference consequenceof a reorientation and piling Ideally only one of the mixture components shouldbe able to reorient but in the case of styrene and ethylbenzene the similarity intheir length and width makes no significant difference in the pore size needed forthe reorientation The selectivity relies thus in how favorable the reorientationand piling are

In MAZ and AFI both molecules can reorient but the reorientation is restrictedby the pore size forcing a tilting in the face-to-face stacking configuration that isunfavorable for ethylbenzene to form a pile but it is favorable for styrene Thisallows styrene to obtain higher saturation loadings and ensures MAZ and AFIstyrene selectivity at saturation conditions

However face-to-face stacking only occurs at high loadings In Figure 712the selectivity is presented at 1times106 Pa At this fugacity MAZ is not yet styreneselective and AFI selectivity is almost negligible The effect of face-to-face stackingin styrene selectivity can only be observed at higher loadings as shown in Figure713 where the selectivity as a function of styrene loading for higher fugacities ispresented

1

10

1 10

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

CoBDP

CuBTC

IRMOF-1

JUC-77

MAF-X8

MIL-47

MIL-53

MOF-CJ3

AFI

DON

MAZ

MFI-para

UiO-66

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

JUC-77

MAF-x8

IRMOF-1

MOFCJ3

CoBDP

CuBTC

UiO-66

MIL-47

MIL-53

JUC-77

CuBTC

CoBDP

MOF-CJ3 IRMOF-1

MAZDON

MFI-para

AFI

UiO-66

MAF-X8

1e6 Pa1e7 Pa1e9 Pa

Figure 713 Selectivity (qstqeb) in an equimolar mixture at 3 different fugacitiesThe dashed red line corresponds to the liquid phase ratio For all the cases styreneloading increases with the fugacity (shift to the right of the plot) The selectivity alsoincreases in most of the cases with the fugacity (shift upwards) except for MFI-paraJUC-77 MIL-47 and Cu-BTC

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 3: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

146 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

process is even more complicated due of the presence of side products like tolueneo-xylene and benzene

An alternative energy-efficient separation strategy involves utilizing the mo-lecular chemical and geometrical differences by means of adsorptive separationwith nanoporous materials like metal-organic frameworks and zeolites Ahmad etal [5] performed liquid chromatography separation using HKUST-1(Cu3(BTC)2)a metal-organic framework with open Cu(II) sites and 135-benzenetricarboxylate(BTC) linkers They found that styrene is preferentially adsorbed in the struc-ture because of the coordinative interaction of styrene with the Cu(II) in a π-complexation mechanism Maes et al [6 7] and Remy et al [8] reported resultson MIL-47(V) and MIL-53(Al) showing both structures are capable of separationin the liquid phase They found that in MIL-47(V) styrene selectivity is relatedto styrene capacity for packing while for MIL-53(Al) styrene selectivity is relatedto adsorption enthalpy (interaction with the carboxylate) For competitive ad-sorption in static conditions they reported separation factors of 36 and 41 forMIL-47(V) and MIL-53(Al) respectively and for an equimolar mixture in dynamicconditions (breakthrough experiments using a column filled with crystallites in anHPLC apparatus) they found separation factors of 29 and 23 They also observedthat if a more realistic mixture is taken into account (with toluene and o-xylene) inMIL-53 o-xylene and toluene are retained even longer which makes the materialgood for impurity removal Yang et al [9] conducted experiments on stationaryphase HPLC with MIL-101(Cr) a material built from a hybrid supertetrahedralbuilding unit formed by terephthalate ligands and trimeric chromium octahedralclusters Similar to HKUST-1 they reported a higher affinity towards styrene dueto the π-π interactions with the metal-organic framework walls and the unsat-urated metal sites They also reported the efficient separation of impurities likeo-xylene and toluene

Separation based on adsorption relies on either adsorption or diffusion charac-teristics At low loadings (ie the Henry regime) the selectivity is mainly drivenby enthalpic effects and favors the molecule with the strongest interaction withthe framework Selectivity is therefore strongly related to adsorbent and adsorbateproperties such as dipole moment polarizability quadrupole moment and mag-netic susceptibility At saturation conditions (industrial set-up) the selectivity isdriven by either enthalpic effects andor entropic effects like

bull commensurate freezing [10] which favors molecules which efficiently pack inintersecting-channels structures

bull size entropy [11 12] which favors the smallest molecules

bull length entropy [11 13ndash15] which favors the molecules with the shortest ef-fective length (footprint) in one-dimensional channels

bull commensurate stacking [16] which favors molecules with stacking arrange-ments that are commensurate with the dimensions of one-dimensional chan-nels

72 Methodology 147

bull face-to-face stacking [17] which favors molecules that when reoriented sig-nificantly reduce their footprint in one-dimensional channels

The various separation strategies for exploitation of molecular packing effects havebeen reviewed recently [18]

Styrene and ethylbenzene are very similar molecules the main difference beingthat styrene is a flat molecule whereas ethylbenzene is not Finding structures withselective adsorption for styrene is not easy In this work we present a screeningstudy for the separation of styrene and ethylbenzene at liquid conditions Wepropose to separate on the basis of a difference in saturation loading because it ismore cost-efficient and utilizes the pore volume most efficiently

72 Methodology

The systems were modeled using classical force fields The adsorbates were modeledwith OPLS-AA force field for organic liquids [19] In previous work [16] we haveshown that the use of these force fields is in good agreement with experimentsBecause we were interested in the selectivity of planarnon-planar molecules andnot in their conformational changes adsorbates were described as multisite rigidmolecules with properties and configurations shown in Figure 71 The parametersfor the interaction of the adsorbates (Lennard-Jones and electrostatic interactions)together with a schematic representation of the molecules showing the atom typesare presented in Table 71 Cross-interactions with other molecules and the frame-work were computed using Lorentz-Berthelot mixing rules

Boiling point 418 KFreezing point 2425 Klength 096 nmwidth 070 nmheight 034 nm

Boiling point 409 KFreezing point 178 Klength 095 nmwidth 067 nmheight 053 nm

Figure 71 Styrene (top) and ethylbenzene (bottom) configurations The figureshows the typical properties of the modeled adsorbates Distances are ldquomolecularshadow lengthsrdquo [20] from Materials Studio [21] Besides small differences in thecharges the main difference between these molecules is their height (planarity)

148 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

atomgroup q [eminus] σ [A] ε [kcal molminus1]C -0115 355 0070C1 -0115 355 0076C2 -0005 350 0066C3 -0180 350 0066C4 -0000 355 0076C5 -0115 355 0076C6 -0148 355 0076H 0115 242 0030H1 0060 242 0030H2 0074 242 0030

Table 71 OPLS-AA force field parameters for styrene and ethylbenzene [19] Thevinyl group charges () were taken from Shirley et al [22]

The frameworks were modeled as rigid with atom positions taken from crystal-lographic experimental data Most MOFs were further optimized using VASP [2324] with the cell fixed to the experimentally determined unit cell size and shape(PBE [25 26] exchange-correlation functional with dispersion corrections [27] wasused and the PAW method was applied to describe the core atoms convergencecriteria of the ionic forces was set to -1times10minus3 AeV) The metal-organic frame-works were modeled using the DREIDING force field [28] and Van der Waals para-meters not found in DREIDING were taken from the universal force field (UFF)[29] DREIDING and UFF force fields were designed to be very generic so thatbroad coverage of the periodic table including inorganic compounds metals andtransition metals could be achieved UFF was tailored for simulating moleculescontaining any combination of elements in the periodic table For the zeolites theTraPPE [30] force field was used This force field was specifically developed forzeolites

The charge-charge interactions were computed using the Ewald summation(relative precision 10minus6) Charges for the frameworks were computed by minimiz-ing the difference of the classical electrostatic potential and a quantum mechanicselectrostatic potential over many grid points using the REPEAT method [31 32]

73 Adsorption isotherms

To compute the adsorption isotherms we perform Monte Carlo simulation in thegrand-canonical ensemble (or microV T ensemble) In this ensemble the number ofadsorbates fluctuates until equilibrium conditions are reached the temperatureand chemical potential of the gas inside and outside the adsorbent are equalBecause in confined systems the fraction of successful insertions and deletions isvery low reaching equilibrium with conventional Monte Carlo methods can bevery time consuming In this study we used the Configurational Bias ContinuousFractional Monte Carlo (CBCFCMC) [33] method to enhance the success rate of

74 Results 149

insertions and deletions The method is a combination of the Configurational BiasMonte Carlo (CBMC) [34ndash36] where molecules growth is biased towards favorableconfigurations and Continuous Fractional Component Monte Carlo (CFCMC) [37]in which molecules are gradually inserted or deleted by scaling their interactionswith the surroundings We have shown in previous work [33] that the resultsobtained with this method do not differ from CBMC calculations but the efficiencyis higher

Using the dual-site Langmuir-Freundlich fits of the pure component isothermsbreakthrough calculations were carried out by solving a set of partial differentialequations for each of the species in the gas mixture [38 39] The molar loadingsof the species at any position along the packed bed and at any time were determ-ined from Ideal Adsorbed Solution Theory calculations Video animations of thebreakthrough behavior as a function of time of selected structures are provided asweb-enhanced objects online

74 Results

We perform a screening study of several zeolites and metal-organic frameworksfor the separation of styreneethylbenzene mixture focusing on saturation condi-tions Under these conditions differences in the saturation capacity of the mixturecomponents strongly dictate the separation

In systems with small pores like MRE and MTW zeolites molecules are forcedto adsorb parallel to the channels The saturation capacity is determined by theeffective length per molecule in the channel (footprint) Because of the similarityin the length of styrene and ethylbenzene the difference in saturation capacities isalmost negligible making systems with small pores unsuitable candidates for theseparation

In structures with cavities or channels much larger than styrene and ethyl-benzene molecular dimensions like IRMOF-1 and Zn-DOBDC molecules do notpresent any particular packing The observed difference in the saturation capa-cities is a consequence of the natural packing of the molecules in liquid phase(ρEb = 08665 gmL ρSt = 0909 gmL) This makes these materials also unsuit-able for the separation process

We have identified a few materials where styrene has a higher saturation capa-city than ethylbenzene In the following we describe how this difference arises fromthe previously mentioned entropic mechanisms and we highlight their applicabilityfor the separation process

Size exclusion is observed in MFI-para [40] MFI-para is a ZSM-5 zeolite whichstructure is a combination of interconnected straight and zigzag channels Thestraight channels have a diameter of 53times56 A and the zigzag channels have adiameter of 51times55 A In Figure 72b the simulated single component isothermsof styrene and ethylbenzene in MFI-para at 433 K and snapshots of styrene andethylbenzene at 1times109 Pa and 433 K are presented At low loadings molecules

150 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

preferentially adsorb in the straight channels the difference in loadings arise froma stronger interaction of styrene with MFI-para At saturation conditions styrenecan obtain almost twice the loading of ethylbenzene because of a size exclusioneffect in the zig-zag channels in which ethylbenzene does not fit due to its heightWhen an equimolar styreneethylbenzene mixture is considered the difference inloadings at saturation conditions is even larger (Figure 72c)

a

0

02

04

06

08

1

12

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

b

0

02

04

06

08

1

12

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

c

Figure 72 (a) Snapshots of ethylbenzene (top) and styrene (bottom) at 1times109

Pa and 433 K (b) Single component isotherms of styrene (red) and ethylbenzene(blue) in MFI-para at 433 K (lines are dual-site Langmuir-Freundlich fits of the purecomponents points are the pure component isotherms from CBCFCMC simulations)(c) Mixture component isotherms for an equimolar mixture in MFI-para at 433 K

Face-to-face stacking occurs in MAZ [41] and AFI [42] zeolites MAZ and AFIare 1D-channel zeolites with dimensions that allow a molecular reorientation ofethylbenzene and styrene

In Figure 73 we present the simulation results for the single component iso-therms of ethylbenzene and styrene in AFI zeolite at 433 K At low loadingsmolecules are mostly adsorbed flat on the walls (parallel to the channels axis) ad-sorption is dictated by enthalpy effects which favors ethylbenzene As the loadingincreases the molecules undergo a molecular reorientation from a configurationparallel to the channel axis to a ldquopilerdquo configuration with the phenyl groups fa-cing each other This new configuration (face-to-face stacking) reduces styrenersquosfootprint more than ethylbenzenersquos footprint (Figure 76a) because of the non-planarity of ethylbenzene allowing styrene to obtain higher saturation loadings

For MAZ zeolite a similar behavior is observed however because MAZ zeolitehas smaller channels than AFI the angle at which molecules can reorient has asmaller effect on the reduction of the moleculersquos footprint in the channels (Figure76b) When an equimolar mixture is considered both AFI and MAZ zeolites arestyrene selective at saturation conditions as shown in Figures 74a and 74b wherethe simulated mixture component isotherms at 433 K are presented

74 Results 151

0

02

04

06

08

1

12

14

100

101

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

Pure component isotherms CBCFCMC AFI 433K

ethylbenzenestyrene

Figure 73 Single component isotherms of styrene and ethylbenzene in AFI at 433 KInset snapshots of styrene (top) and ethylbenzene (bottom) at 1times103 Pa and 1times109

Pa At low loadings both molecules are adsorbed with the phenyl group parallel to thechannel axis At higher pressures styrene can arrange in a tilted face-to-face stackingconfiguration which reduces its footprint and allows for a higher saturation capacityThe difference in saturation capacities ensures the selectivity of AFI towards styreneat saturation conditions in the mixture

0

01

02

03

04

05

06

07

08

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

01

02

03

04

05

06

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

Figure 74 Mixture component isotherms for and equimolar mixture of styrene andethylbenzene at 433 K in (a) AFI zeolite and (b) MAZ zeolite

Face-to-face stacking is also observed in DON zeolite DON [43] is a structurewith 1D-channels slightly larger than AFI and MAZ zeolites This increase indimensions favors ethylbenzene to undergo a molecular reorientation into a face-to-face stacking configuration but also because the diameter of the channels arelarger than the length of styrene it induces styrene to adopt a commensurate stack-ing configuration where the stacking of two molecules with their phenyl groupsfacing each other is commensurate with the channel dimensions (Figures 75a)This is also observed in MIL-53 [44] a metal-organic framework with lozenge-shaped rhombohedric channels of approximately 085 nm as show in Figure 75b

152 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

a

b

Figure 75 (a) Snapshots of styrene (left) and ethylbenzene (right) in DON zeolite at1e9 Pa and 433 K(b) Snapshots of styrene (left) and ethylbenzene (right) in MIL-53at 1e9 Pa and 433 K Color code carbon (cyan) hydrogen (white)

where snapshots of styrene and ethylbenzene at 1times109 Pa and 433 K are presen-ted Because of styrene and ethylbenzene dimensions styrene with commensuratestacking can obtain higher saturation capacities than ethylbenzene with face-to-face stacking as shown schematically in Figure 76c This is also seen in DONand MIL-53 single component isotherms (Figure 77a) Mixture isotherms (Fig-ures 77b) and breakthrough curves (Figure 77c) further confirm that DON andMIL-53 are styrene selective structures

Commensurate stacking for both styrene and ethylbenzene is observed in MIL-47 [45] and MAF-X8 [46] MIL-47 is a metal-organic framework with lozenge-shaped rhombohedric channels which size (slightly larger than MIL-53) allows forethylbenzene to also have commensurate stacking In Figure 78 we present thesimulated single component adsorption isotherms of ethylbenzene and styrene inMIL-47 at 433 K and snapshots of styrene and ethylbenzene at 1times106 Pa and1times109 Pa At 1times109 Pa both molecules have commensurate stacking but in orderfor ethylbenzene to have commensurate stacking the phenyl groups have to beslightly shifted due to the non-planar ethyl group This shift affects the amount ofethylbenzene molecules that can be adsorbed (as compared to styrene) and there-fore causes a difference in the saturation capacity of styrene and ethylbenzeneInterestingly commensurate stacking occurs at lower loading for styrene than forethylbenzene We can see in Figure 78 that at 1times106 Pa styrene already presentscommensurate stacking while ethylbenzene seems to have more of a face-to-facestacking In a mixture this will favor styrene adsorption even more In Figure79 simulated mixture adsorption isotherms for an equimolar mixture in MIL-47at 433 K are presented MIL-47 is a styrene selective structure Our results are ingood qualitative agreement with Maes et al [6] Breakthrough simulations furtherconfirm that MIL-47 is a styrene selective structure at saturation conditions InMAF-x8 a metal-organic framework with square channels of approximately 1nm

74 Results 153

BEFORE REORIENTATION

AFTER REORIENTATION

096 nm 095 nm

034 nm 053 nm

a b

095 nm

053 nm x2

c

Figure 76 Schematic representation of different entropic effects occurring in theseparation of styrene and ethylbenzene in nanoporous materials (a) The reductionof the moleculersquos footprint in the channels because of a reorientation into a face-to-face stacking configuration (b) Effect of the channel size on the reorientationand therefore moleculersquos footprint reduction (c) Comparison of the channel lengthneeded for two molecules of styrene to have commensurate stacking vs two moleculesof ethylbenzene to have face-to-face stacking

0

05

1

15

2

25

3

35

10-1

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

0

05

1

15

2

25

3

100

101

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

0

05

1

15

2

0 05 1 15 2 25 3 35 4 45

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

0

02

04

06

08

1

12

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

a

0

01

02

03

04

05

06

07

08

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

0

05

1

15

2

0 05 1 15 2 25 3

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

c

Figure 77 Simulated isotherms and breakthrough curves of styrene and ethylben-zene in MIL-53 (top) and DON (bottom) at 433 K (a) Pure component isotherms(lines are dual-site Langmuir-Freundlich fits of the pure components points are thepure component isotherms from CBCFCMC simulations) (b) Mixture componentisotherms for an equimolar mixture The IAST results are in good agreement with themixture isotherms (c) Simulated step-type breakthrough at 1e6 Pa total fugacity

154 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

0

1

2

3

4

5

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

Figure 78 Single component adsorption isotherms for styrene and ethylbenzeneat 433 K in MIL-47 Inset styrene (top) and ethylbenzene (bottom) snapshots at1times106 Pa and 1times109 Pa Styrene has commensurate stacking at lower pressuresthan ethylbenzene Because of the out of plane ethyl group ethylbenzene moleculeshave to be slightly shifted in order to have commensurate stacking and therefore lessethylbenzene molecules can be adsorbed in MIL-47 channels at saturation conditions

0

05

1

15

2

25

3

35

4

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

02

04

06

08

1

12

14

16

18

0 1 2 3 4 5

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

b

Figure 79 Styreneethylbenzene separation using MIL-47 at 433 K (a) equimolarmixture isotherms and Ideal Adsorption Solution Theory (IAST) prediction based onpure component isotherms (b) simulated step breakthrough at 1times106 Pa total fu-gacity The IAST prediction is in excellent agreement with the mixture simulationsThe mixture and breakthrough simulations show a high styrene selectivity and loadingin the mixture

it is easier for ethylbenzene to have commensurate stacking than in MIL-47 Thesingle component isotherms of both molecules behave very similar The topologyof the structure seems to induce a shift between parallel styrene molecules and al-lows for the ethyl group of ethylbenzene to stick in the channel ldquopocketsrdquo (Figure710) This might be the reason for the smaller difference in saturation capacitiesof styrene and ethylbenzene compared to MIL-47

74 Results 155

a b

0

05

1

15

2

25

3

35

4

10-6

10-4

10-2

100

102

104

106

108

Abso

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

c

Figure 710 Snapshots of (a) styrene and (b) ethylbenzene in MAF-X8 at 1e9 Paand 433 K Both styrene and ethylbenzene have commensurate stacking Color codecarbon (cyan) hydrogen (white) (c) Simulated single component isotherms of styreneand ethylbenzene in MAF-X8 at 433 K

Combinationcompetition of mechanisms is observed in structures with a more com-plex topology An example of this is MOF-CJ3 metal-organic framework [47] Thewide segment of the channels are big enough to allow for both styrene and ethyl-benzene to form two parallel rows (commensurate stacking) however the shapeof the channels forces ethylbenzene molecules to adopt a configuration where theethyl group is pointing to the channel ldquopocketsrdquo Styrene has more freedom andmolecules can arrange in a way that an extra styrene can be adsorbed in theprotracted segments in a configuration perpendicular to the channel (face-to-facestacking) (Figure 711) The interplay between different mechanisms rarely makesthe separation better Even when all the mechanisms favor a specific molecule thecompetition between them can induce enough disorder to destroy the selectivityone could achieve with a ldquopurerdquo mechanism

Figure 711 Snapshots of styrene (top) and ethylbenzene (bottom) at 1times109 Paand 433 K in MOF-CJ3

156 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

75 Discussion

There are two important factors to consider when using adsorption for separa-tion processes at industrial conditions namely selectivity and capacity A highselectivity ensures that less cycles are needed to achieve a high degree of purity inthe separation but a high capacity implies that the regeneration time is longerFor a binary mixture the adsorption selectivity is defined as

Sads =q1q2

f1f2(71)

and the capacity is defined as the styrene loading in the adsorbed phase of a binarymixture

Capacity = q1 (72)

In Figure 712 the relationship between these two properties for different struc-tures is presented Structures with the same separation mechanism are plottedwith the same color An ideal structure for the separation would be located at theright top corner

1

10

0 1 2 3 4 5 6

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

MAF-x8

MOF-CJ3

CoBDP IRMOF-1

JUC-77

CuBTC

UiO-66

Figure 712 Selectivity as a function of styrene loading (capacity) in a binary mix-ture at 433 K and 1times106 Pa total fugacity The structures are divided in differentcolors depending on the selectivity mechanism observed Color code Size exclusion(red) face-to-face stacking (purple) commensurate stacking (blue) commensurate-stackingface-to-face stacking (green) mixed(orange) The dotted red line corres-ponds to the ratio styreneethylbenzene at liquid conditions There is a naturaltrade-off between selectivity and loading finding structures in the top right corneris not feasible The black dashed line (guide to the eye) denotes the inverse relation-ship between selectivity and capacity MIL-47 is a styrene selective material with ahigh capacity therefore a good candidate for the styreneethylbenzene separation

75 Discussion 157

In structures where the separation is driven by size-exclusion (MFI-para) theselectivity is high because there is an adsorption site available only for styrene(zig-zag channels) but the capacity is rather low since for size-exclusion to occuramong similar molecules there has to be a very tight fit between the moleculesand the adsorbent usually associated with small pore systems

In structures where the separation is driven by face-to-face stacking the se-lectivity relies on the increase of the moleculersquos footprints difference consequenceof a reorientation and piling Ideally only one of the mixture components shouldbe able to reorient but in the case of styrene and ethylbenzene the similarity intheir length and width makes no significant difference in the pore size needed forthe reorientation The selectivity relies thus in how favorable the reorientationand piling are

In MAZ and AFI both molecules can reorient but the reorientation is restrictedby the pore size forcing a tilting in the face-to-face stacking configuration that isunfavorable for ethylbenzene to form a pile but it is favorable for styrene Thisallows styrene to obtain higher saturation loadings and ensures MAZ and AFIstyrene selectivity at saturation conditions

However face-to-face stacking only occurs at high loadings In Figure 712the selectivity is presented at 1times106 Pa At this fugacity MAZ is not yet styreneselective and AFI selectivity is almost negligible The effect of face-to-face stackingin styrene selectivity can only be observed at higher loadings as shown in Figure713 where the selectivity as a function of styrene loading for higher fugacities ispresented

1

10

1 10

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

CoBDP

CuBTC

IRMOF-1

JUC-77

MAF-X8

MIL-47

MIL-53

MOF-CJ3

AFI

DON

MAZ

MFI-para

UiO-66

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

JUC-77

MAF-x8

IRMOF-1

MOFCJ3

CoBDP

CuBTC

UiO-66

MIL-47

MIL-53

JUC-77

CuBTC

CoBDP

MOF-CJ3 IRMOF-1

MAZDON

MFI-para

AFI

UiO-66

MAF-X8

1e6 Pa1e7 Pa1e9 Pa

Figure 713 Selectivity (qstqeb) in an equimolar mixture at 3 different fugacitiesThe dashed red line corresponds to the liquid phase ratio For all the cases styreneloading increases with the fugacity (shift to the right of the plot) The selectivity alsoincreases in most of the cases with the fugacity (shift upwards) except for MFI-paraJUC-77 MIL-47 and Cu-BTC

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 4: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

72 Methodology 147

bull face-to-face stacking [17] which favors molecules that when reoriented sig-nificantly reduce their footprint in one-dimensional channels

The various separation strategies for exploitation of molecular packing effects havebeen reviewed recently [18]

Styrene and ethylbenzene are very similar molecules the main difference beingthat styrene is a flat molecule whereas ethylbenzene is not Finding structures withselective adsorption for styrene is not easy In this work we present a screeningstudy for the separation of styrene and ethylbenzene at liquid conditions Wepropose to separate on the basis of a difference in saturation loading because it ismore cost-efficient and utilizes the pore volume most efficiently

72 Methodology

The systems were modeled using classical force fields The adsorbates were modeledwith OPLS-AA force field for organic liquids [19] In previous work [16] we haveshown that the use of these force fields is in good agreement with experimentsBecause we were interested in the selectivity of planarnon-planar molecules andnot in their conformational changes adsorbates were described as multisite rigidmolecules with properties and configurations shown in Figure 71 The parametersfor the interaction of the adsorbates (Lennard-Jones and electrostatic interactions)together with a schematic representation of the molecules showing the atom typesare presented in Table 71 Cross-interactions with other molecules and the frame-work were computed using Lorentz-Berthelot mixing rules

Boiling point 418 KFreezing point 2425 Klength 096 nmwidth 070 nmheight 034 nm

Boiling point 409 KFreezing point 178 Klength 095 nmwidth 067 nmheight 053 nm

Figure 71 Styrene (top) and ethylbenzene (bottom) configurations The figureshows the typical properties of the modeled adsorbates Distances are ldquomolecularshadow lengthsrdquo [20] from Materials Studio [21] Besides small differences in thecharges the main difference between these molecules is their height (planarity)

148 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

atomgroup q [eminus] σ [A] ε [kcal molminus1]C -0115 355 0070C1 -0115 355 0076C2 -0005 350 0066C3 -0180 350 0066C4 -0000 355 0076C5 -0115 355 0076C6 -0148 355 0076H 0115 242 0030H1 0060 242 0030H2 0074 242 0030

Table 71 OPLS-AA force field parameters for styrene and ethylbenzene [19] Thevinyl group charges () were taken from Shirley et al [22]

The frameworks were modeled as rigid with atom positions taken from crystal-lographic experimental data Most MOFs were further optimized using VASP [2324] with the cell fixed to the experimentally determined unit cell size and shape(PBE [25 26] exchange-correlation functional with dispersion corrections [27] wasused and the PAW method was applied to describe the core atoms convergencecriteria of the ionic forces was set to -1times10minus3 AeV) The metal-organic frame-works were modeled using the DREIDING force field [28] and Van der Waals para-meters not found in DREIDING were taken from the universal force field (UFF)[29] DREIDING and UFF force fields were designed to be very generic so thatbroad coverage of the periodic table including inorganic compounds metals andtransition metals could be achieved UFF was tailored for simulating moleculescontaining any combination of elements in the periodic table For the zeolites theTraPPE [30] force field was used This force field was specifically developed forzeolites

The charge-charge interactions were computed using the Ewald summation(relative precision 10minus6) Charges for the frameworks were computed by minimiz-ing the difference of the classical electrostatic potential and a quantum mechanicselectrostatic potential over many grid points using the REPEAT method [31 32]

73 Adsorption isotherms

To compute the adsorption isotherms we perform Monte Carlo simulation in thegrand-canonical ensemble (or microV T ensemble) In this ensemble the number ofadsorbates fluctuates until equilibrium conditions are reached the temperatureand chemical potential of the gas inside and outside the adsorbent are equalBecause in confined systems the fraction of successful insertions and deletions isvery low reaching equilibrium with conventional Monte Carlo methods can bevery time consuming In this study we used the Configurational Bias ContinuousFractional Monte Carlo (CBCFCMC) [33] method to enhance the success rate of

74 Results 149

insertions and deletions The method is a combination of the Configurational BiasMonte Carlo (CBMC) [34ndash36] where molecules growth is biased towards favorableconfigurations and Continuous Fractional Component Monte Carlo (CFCMC) [37]in which molecules are gradually inserted or deleted by scaling their interactionswith the surroundings We have shown in previous work [33] that the resultsobtained with this method do not differ from CBMC calculations but the efficiencyis higher

Using the dual-site Langmuir-Freundlich fits of the pure component isothermsbreakthrough calculations were carried out by solving a set of partial differentialequations for each of the species in the gas mixture [38 39] The molar loadingsof the species at any position along the packed bed and at any time were determ-ined from Ideal Adsorbed Solution Theory calculations Video animations of thebreakthrough behavior as a function of time of selected structures are provided asweb-enhanced objects online

74 Results

We perform a screening study of several zeolites and metal-organic frameworksfor the separation of styreneethylbenzene mixture focusing on saturation condi-tions Under these conditions differences in the saturation capacity of the mixturecomponents strongly dictate the separation

In systems with small pores like MRE and MTW zeolites molecules are forcedto adsorb parallel to the channels The saturation capacity is determined by theeffective length per molecule in the channel (footprint) Because of the similarityin the length of styrene and ethylbenzene the difference in saturation capacities isalmost negligible making systems with small pores unsuitable candidates for theseparation

In structures with cavities or channels much larger than styrene and ethyl-benzene molecular dimensions like IRMOF-1 and Zn-DOBDC molecules do notpresent any particular packing The observed difference in the saturation capa-cities is a consequence of the natural packing of the molecules in liquid phase(ρEb = 08665 gmL ρSt = 0909 gmL) This makes these materials also unsuit-able for the separation process

We have identified a few materials where styrene has a higher saturation capa-city than ethylbenzene In the following we describe how this difference arises fromthe previously mentioned entropic mechanisms and we highlight their applicabilityfor the separation process

Size exclusion is observed in MFI-para [40] MFI-para is a ZSM-5 zeolite whichstructure is a combination of interconnected straight and zigzag channels Thestraight channels have a diameter of 53times56 A and the zigzag channels have adiameter of 51times55 A In Figure 72b the simulated single component isothermsof styrene and ethylbenzene in MFI-para at 433 K and snapshots of styrene andethylbenzene at 1times109 Pa and 433 K are presented At low loadings molecules

150 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

preferentially adsorb in the straight channels the difference in loadings arise froma stronger interaction of styrene with MFI-para At saturation conditions styrenecan obtain almost twice the loading of ethylbenzene because of a size exclusioneffect in the zig-zag channels in which ethylbenzene does not fit due to its heightWhen an equimolar styreneethylbenzene mixture is considered the difference inloadings at saturation conditions is even larger (Figure 72c)

a

0

02

04

06

08

1

12

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

b

0

02

04

06

08

1

12

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

c

Figure 72 (a) Snapshots of ethylbenzene (top) and styrene (bottom) at 1times109

Pa and 433 K (b) Single component isotherms of styrene (red) and ethylbenzene(blue) in MFI-para at 433 K (lines are dual-site Langmuir-Freundlich fits of the purecomponents points are the pure component isotherms from CBCFCMC simulations)(c) Mixture component isotherms for an equimolar mixture in MFI-para at 433 K

Face-to-face stacking occurs in MAZ [41] and AFI [42] zeolites MAZ and AFIare 1D-channel zeolites with dimensions that allow a molecular reorientation ofethylbenzene and styrene

In Figure 73 we present the simulation results for the single component iso-therms of ethylbenzene and styrene in AFI zeolite at 433 K At low loadingsmolecules are mostly adsorbed flat on the walls (parallel to the channels axis) ad-sorption is dictated by enthalpy effects which favors ethylbenzene As the loadingincreases the molecules undergo a molecular reorientation from a configurationparallel to the channel axis to a ldquopilerdquo configuration with the phenyl groups fa-cing each other This new configuration (face-to-face stacking) reduces styrenersquosfootprint more than ethylbenzenersquos footprint (Figure 76a) because of the non-planarity of ethylbenzene allowing styrene to obtain higher saturation loadings

For MAZ zeolite a similar behavior is observed however because MAZ zeolitehas smaller channels than AFI the angle at which molecules can reorient has asmaller effect on the reduction of the moleculersquos footprint in the channels (Figure76b) When an equimolar mixture is considered both AFI and MAZ zeolites arestyrene selective at saturation conditions as shown in Figures 74a and 74b wherethe simulated mixture component isotherms at 433 K are presented

74 Results 151

0

02

04

06

08

1

12

14

100

101

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

Pure component isotherms CBCFCMC AFI 433K

ethylbenzenestyrene

Figure 73 Single component isotherms of styrene and ethylbenzene in AFI at 433 KInset snapshots of styrene (top) and ethylbenzene (bottom) at 1times103 Pa and 1times109

Pa At low loadings both molecules are adsorbed with the phenyl group parallel to thechannel axis At higher pressures styrene can arrange in a tilted face-to-face stackingconfiguration which reduces its footprint and allows for a higher saturation capacityThe difference in saturation capacities ensures the selectivity of AFI towards styreneat saturation conditions in the mixture

0

01

02

03

04

05

06

07

08

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

01

02

03

04

05

06

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

Figure 74 Mixture component isotherms for and equimolar mixture of styrene andethylbenzene at 433 K in (a) AFI zeolite and (b) MAZ zeolite

Face-to-face stacking is also observed in DON zeolite DON [43] is a structurewith 1D-channels slightly larger than AFI and MAZ zeolites This increase indimensions favors ethylbenzene to undergo a molecular reorientation into a face-to-face stacking configuration but also because the diameter of the channels arelarger than the length of styrene it induces styrene to adopt a commensurate stack-ing configuration where the stacking of two molecules with their phenyl groupsfacing each other is commensurate with the channel dimensions (Figures 75a)This is also observed in MIL-53 [44] a metal-organic framework with lozenge-shaped rhombohedric channels of approximately 085 nm as show in Figure 75b

152 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

a

b

Figure 75 (a) Snapshots of styrene (left) and ethylbenzene (right) in DON zeolite at1e9 Pa and 433 K(b) Snapshots of styrene (left) and ethylbenzene (right) in MIL-53at 1e9 Pa and 433 K Color code carbon (cyan) hydrogen (white)

where snapshots of styrene and ethylbenzene at 1times109 Pa and 433 K are presen-ted Because of styrene and ethylbenzene dimensions styrene with commensuratestacking can obtain higher saturation capacities than ethylbenzene with face-to-face stacking as shown schematically in Figure 76c This is also seen in DONand MIL-53 single component isotherms (Figure 77a) Mixture isotherms (Fig-ures 77b) and breakthrough curves (Figure 77c) further confirm that DON andMIL-53 are styrene selective structures

Commensurate stacking for both styrene and ethylbenzene is observed in MIL-47 [45] and MAF-X8 [46] MIL-47 is a metal-organic framework with lozenge-shaped rhombohedric channels which size (slightly larger than MIL-53) allows forethylbenzene to also have commensurate stacking In Figure 78 we present thesimulated single component adsorption isotherms of ethylbenzene and styrene inMIL-47 at 433 K and snapshots of styrene and ethylbenzene at 1times106 Pa and1times109 Pa At 1times109 Pa both molecules have commensurate stacking but in orderfor ethylbenzene to have commensurate stacking the phenyl groups have to beslightly shifted due to the non-planar ethyl group This shift affects the amount ofethylbenzene molecules that can be adsorbed (as compared to styrene) and there-fore causes a difference in the saturation capacity of styrene and ethylbenzeneInterestingly commensurate stacking occurs at lower loading for styrene than forethylbenzene We can see in Figure 78 that at 1times106 Pa styrene already presentscommensurate stacking while ethylbenzene seems to have more of a face-to-facestacking In a mixture this will favor styrene adsorption even more In Figure79 simulated mixture adsorption isotherms for an equimolar mixture in MIL-47at 433 K are presented MIL-47 is a styrene selective structure Our results are ingood qualitative agreement with Maes et al [6] Breakthrough simulations furtherconfirm that MIL-47 is a styrene selective structure at saturation conditions InMAF-x8 a metal-organic framework with square channels of approximately 1nm

74 Results 153

BEFORE REORIENTATION

AFTER REORIENTATION

096 nm 095 nm

034 nm 053 nm

a b

095 nm

053 nm x2

c

Figure 76 Schematic representation of different entropic effects occurring in theseparation of styrene and ethylbenzene in nanoporous materials (a) The reductionof the moleculersquos footprint in the channels because of a reorientation into a face-to-face stacking configuration (b) Effect of the channel size on the reorientationand therefore moleculersquos footprint reduction (c) Comparison of the channel lengthneeded for two molecules of styrene to have commensurate stacking vs two moleculesof ethylbenzene to have face-to-face stacking

0

05

1

15

2

25

3

35

10-1

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

0

05

1

15

2

25

3

100

101

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

0

05

1

15

2

0 05 1 15 2 25 3 35 4 45

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

0

02

04

06

08

1

12

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

a

0

01

02

03

04

05

06

07

08

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

0

05

1

15

2

0 05 1 15 2 25 3

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

c

Figure 77 Simulated isotherms and breakthrough curves of styrene and ethylben-zene in MIL-53 (top) and DON (bottom) at 433 K (a) Pure component isotherms(lines are dual-site Langmuir-Freundlich fits of the pure components points are thepure component isotherms from CBCFCMC simulations) (b) Mixture componentisotherms for an equimolar mixture The IAST results are in good agreement with themixture isotherms (c) Simulated step-type breakthrough at 1e6 Pa total fugacity

154 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

0

1

2

3

4

5

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

Figure 78 Single component adsorption isotherms for styrene and ethylbenzeneat 433 K in MIL-47 Inset styrene (top) and ethylbenzene (bottom) snapshots at1times106 Pa and 1times109 Pa Styrene has commensurate stacking at lower pressuresthan ethylbenzene Because of the out of plane ethyl group ethylbenzene moleculeshave to be slightly shifted in order to have commensurate stacking and therefore lessethylbenzene molecules can be adsorbed in MIL-47 channels at saturation conditions

0

05

1

15

2

25

3

35

4

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

02

04

06

08

1

12

14

16

18

0 1 2 3 4 5

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

b

Figure 79 Styreneethylbenzene separation using MIL-47 at 433 K (a) equimolarmixture isotherms and Ideal Adsorption Solution Theory (IAST) prediction based onpure component isotherms (b) simulated step breakthrough at 1times106 Pa total fu-gacity The IAST prediction is in excellent agreement with the mixture simulationsThe mixture and breakthrough simulations show a high styrene selectivity and loadingin the mixture

it is easier for ethylbenzene to have commensurate stacking than in MIL-47 Thesingle component isotherms of both molecules behave very similar The topologyof the structure seems to induce a shift between parallel styrene molecules and al-lows for the ethyl group of ethylbenzene to stick in the channel ldquopocketsrdquo (Figure710) This might be the reason for the smaller difference in saturation capacitiesof styrene and ethylbenzene compared to MIL-47

74 Results 155

a b

0

05

1

15

2

25

3

35

4

10-6

10-4

10-2

100

102

104

106

108

Abso

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

c

Figure 710 Snapshots of (a) styrene and (b) ethylbenzene in MAF-X8 at 1e9 Paand 433 K Both styrene and ethylbenzene have commensurate stacking Color codecarbon (cyan) hydrogen (white) (c) Simulated single component isotherms of styreneand ethylbenzene in MAF-X8 at 433 K

Combinationcompetition of mechanisms is observed in structures with a more com-plex topology An example of this is MOF-CJ3 metal-organic framework [47] Thewide segment of the channels are big enough to allow for both styrene and ethyl-benzene to form two parallel rows (commensurate stacking) however the shapeof the channels forces ethylbenzene molecules to adopt a configuration where theethyl group is pointing to the channel ldquopocketsrdquo Styrene has more freedom andmolecules can arrange in a way that an extra styrene can be adsorbed in theprotracted segments in a configuration perpendicular to the channel (face-to-facestacking) (Figure 711) The interplay between different mechanisms rarely makesthe separation better Even when all the mechanisms favor a specific molecule thecompetition between them can induce enough disorder to destroy the selectivityone could achieve with a ldquopurerdquo mechanism

Figure 711 Snapshots of styrene (top) and ethylbenzene (bottom) at 1times109 Paand 433 K in MOF-CJ3

156 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

75 Discussion

There are two important factors to consider when using adsorption for separa-tion processes at industrial conditions namely selectivity and capacity A highselectivity ensures that less cycles are needed to achieve a high degree of purity inthe separation but a high capacity implies that the regeneration time is longerFor a binary mixture the adsorption selectivity is defined as

Sads =q1q2

f1f2(71)

and the capacity is defined as the styrene loading in the adsorbed phase of a binarymixture

Capacity = q1 (72)

In Figure 712 the relationship between these two properties for different struc-tures is presented Structures with the same separation mechanism are plottedwith the same color An ideal structure for the separation would be located at theright top corner

1

10

0 1 2 3 4 5 6

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

MAF-x8

MOF-CJ3

CoBDP IRMOF-1

JUC-77

CuBTC

UiO-66

Figure 712 Selectivity as a function of styrene loading (capacity) in a binary mix-ture at 433 K and 1times106 Pa total fugacity The structures are divided in differentcolors depending on the selectivity mechanism observed Color code Size exclusion(red) face-to-face stacking (purple) commensurate stacking (blue) commensurate-stackingface-to-face stacking (green) mixed(orange) The dotted red line corres-ponds to the ratio styreneethylbenzene at liquid conditions There is a naturaltrade-off between selectivity and loading finding structures in the top right corneris not feasible The black dashed line (guide to the eye) denotes the inverse relation-ship between selectivity and capacity MIL-47 is a styrene selective material with ahigh capacity therefore a good candidate for the styreneethylbenzene separation

75 Discussion 157

In structures where the separation is driven by size-exclusion (MFI-para) theselectivity is high because there is an adsorption site available only for styrene(zig-zag channels) but the capacity is rather low since for size-exclusion to occuramong similar molecules there has to be a very tight fit between the moleculesand the adsorbent usually associated with small pore systems

In structures where the separation is driven by face-to-face stacking the se-lectivity relies on the increase of the moleculersquos footprints difference consequenceof a reorientation and piling Ideally only one of the mixture components shouldbe able to reorient but in the case of styrene and ethylbenzene the similarity intheir length and width makes no significant difference in the pore size needed forthe reorientation The selectivity relies thus in how favorable the reorientationand piling are

In MAZ and AFI both molecules can reorient but the reorientation is restrictedby the pore size forcing a tilting in the face-to-face stacking configuration that isunfavorable for ethylbenzene to form a pile but it is favorable for styrene Thisallows styrene to obtain higher saturation loadings and ensures MAZ and AFIstyrene selectivity at saturation conditions

However face-to-face stacking only occurs at high loadings In Figure 712the selectivity is presented at 1times106 Pa At this fugacity MAZ is not yet styreneselective and AFI selectivity is almost negligible The effect of face-to-face stackingin styrene selectivity can only be observed at higher loadings as shown in Figure713 where the selectivity as a function of styrene loading for higher fugacities ispresented

1

10

1 10

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

CoBDP

CuBTC

IRMOF-1

JUC-77

MAF-X8

MIL-47

MIL-53

MOF-CJ3

AFI

DON

MAZ

MFI-para

UiO-66

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

JUC-77

MAF-x8

IRMOF-1

MOFCJ3

CoBDP

CuBTC

UiO-66

MIL-47

MIL-53

JUC-77

CuBTC

CoBDP

MOF-CJ3 IRMOF-1

MAZDON

MFI-para

AFI

UiO-66

MAF-X8

1e6 Pa1e7 Pa1e9 Pa

Figure 713 Selectivity (qstqeb) in an equimolar mixture at 3 different fugacitiesThe dashed red line corresponds to the liquid phase ratio For all the cases styreneloading increases with the fugacity (shift to the right of the plot) The selectivity alsoincreases in most of the cases with the fugacity (shift upwards) except for MFI-paraJUC-77 MIL-47 and Cu-BTC

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 5: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

148 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

atomgroup q [eminus] σ [A] ε [kcal molminus1]C -0115 355 0070C1 -0115 355 0076C2 -0005 350 0066C3 -0180 350 0066C4 -0000 355 0076C5 -0115 355 0076C6 -0148 355 0076H 0115 242 0030H1 0060 242 0030H2 0074 242 0030

Table 71 OPLS-AA force field parameters for styrene and ethylbenzene [19] Thevinyl group charges () were taken from Shirley et al [22]

The frameworks were modeled as rigid with atom positions taken from crystal-lographic experimental data Most MOFs were further optimized using VASP [2324] with the cell fixed to the experimentally determined unit cell size and shape(PBE [25 26] exchange-correlation functional with dispersion corrections [27] wasused and the PAW method was applied to describe the core atoms convergencecriteria of the ionic forces was set to -1times10minus3 AeV) The metal-organic frame-works were modeled using the DREIDING force field [28] and Van der Waals para-meters not found in DREIDING were taken from the universal force field (UFF)[29] DREIDING and UFF force fields were designed to be very generic so thatbroad coverage of the periodic table including inorganic compounds metals andtransition metals could be achieved UFF was tailored for simulating moleculescontaining any combination of elements in the periodic table For the zeolites theTraPPE [30] force field was used This force field was specifically developed forzeolites

The charge-charge interactions were computed using the Ewald summation(relative precision 10minus6) Charges for the frameworks were computed by minimiz-ing the difference of the classical electrostatic potential and a quantum mechanicselectrostatic potential over many grid points using the REPEAT method [31 32]

73 Adsorption isotherms

To compute the adsorption isotherms we perform Monte Carlo simulation in thegrand-canonical ensemble (or microV T ensemble) In this ensemble the number ofadsorbates fluctuates until equilibrium conditions are reached the temperatureand chemical potential of the gas inside and outside the adsorbent are equalBecause in confined systems the fraction of successful insertions and deletions isvery low reaching equilibrium with conventional Monte Carlo methods can bevery time consuming In this study we used the Configurational Bias ContinuousFractional Monte Carlo (CBCFCMC) [33] method to enhance the success rate of

74 Results 149

insertions and deletions The method is a combination of the Configurational BiasMonte Carlo (CBMC) [34ndash36] where molecules growth is biased towards favorableconfigurations and Continuous Fractional Component Monte Carlo (CFCMC) [37]in which molecules are gradually inserted or deleted by scaling their interactionswith the surroundings We have shown in previous work [33] that the resultsobtained with this method do not differ from CBMC calculations but the efficiencyis higher

Using the dual-site Langmuir-Freundlich fits of the pure component isothermsbreakthrough calculations were carried out by solving a set of partial differentialequations for each of the species in the gas mixture [38 39] The molar loadingsof the species at any position along the packed bed and at any time were determ-ined from Ideal Adsorbed Solution Theory calculations Video animations of thebreakthrough behavior as a function of time of selected structures are provided asweb-enhanced objects online

74 Results

We perform a screening study of several zeolites and metal-organic frameworksfor the separation of styreneethylbenzene mixture focusing on saturation condi-tions Under these conditions differences in the saturation capacity of the mixturecomponents strongly dictate the separation

In systems with small pores like MRE and MTW zeolites molecules are forcedto adsorb parallel to the channels The saturation capacity is determined by theeffective length per molecule in the channel (footprint) Because of the similarityin the length of styrene and ethylbenzene the difference in saturation capacities isalmost negligible making systems with small pores unsuitable candidates for theseparation

In structures with cavities or channels much larger than styrene and ethyl-benzene molecular dimensions like IRMOF-1 and Zn-DOBDC molecules do notpresent any particular packing The observed difference in the saturation capa-cities is a consequence of the natural packing of the molecules in liquid phase(ρEb = 08665 gmL ρSt = 0909 gmL) This makes these materials also unsuit-able for the separation process

We have identified a few materials where styrene has a higher saturation capa-city than ethylbenzene In the following we describe how this difference arises fromthe previously mentioned entropic mechanisms and we highlight their applicabilityfor the separation process

Size exclusion is observed in MFI-para [40] MFI-para is a ZSM-5 zeolite whichstructure is a combination of interconnected straight and zigzag channels Thestraight channels have a diameter of 53times56 A and the zigzag channels have adiameter of 51times55 A In Figure 72b the simulated single component isothermsof styrene and ethylbenzene in MFI-para at 433 K and snapshots of styrene andethylbenzene at 1times109 Pa and 433 K are presented At low loadings molecules

150 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

preferentially adsorb in the straight channels the difference in loadings arise froma stronger interaction of styrene with MFI-para At saturation conditions styrenecan obtain almost twice the loading of ethylbenzene because of a size exclusioneffect in the zig-zag channels in which ethylbenzene does not fit due to its heightWhen an equimolar styreneethylbenzene mixture is considered the difference inloadings at saturation conditions is even larger (Figure 72c)

a

0

02

04

06

08

1

12

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

b

0

02

04

06

08

1

12

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

c

Figure 72 (a) Snapshots of ethylbenzene (top) and styrene (bottom) at 1times109

Pa and 433 K (b) Single component isotherms of styrene (red) and ethylbenzene(blue) in MFI-para at 433 K (lines are dual-site Langmuir-Freundlich fits of the purecomponents points are the pure component isotherms from CBCFCMC simulations)(c) Mixture component isotherms for an equimolar mixture in MFI-para at 433 K

Face-to-face stacking occurs in MAZ [41] and AFI [42] zeolites MAZ and AFIare 1D-channel zeolites with dimensions that allow a molecular reorientation ofethylbenzene and styrene

In Figure 73 we present the simulation results for the single component iso-therms of ethylbenzene and styrene in AFI zeolite at 433 K At low loadingsmolecules are mostly adsorbed flat on the walls (parallel to the channels axis) ad-sorption is dictated by enthalpy effects which favors ethylbenzene As the loadingincreases the molecules undergo a molecular reorientation from a configurationparallel to the channel axis to a ldquopilerdquo configuration with the phenyl groups fa-cing each other This new configuration (face-to-face stacking) reduces styrenersquosfootprint more than ethylbenzenersquos footprint (Figure 76a) because of the non-planarity of ethylbenzene allowing styrene to obtain higher saturation loadings

For MAZ zeolite a similar behavior is observed however because MAZ zeolitehas smaller channels than AFI the angle at which molecules can reorient has asmaller effect on the reduction of the moleculersquos footprint in the channels (Figure76b) When an equimolar mixture is considered both AFI and MAZ zeolites arestyrene selective at saturation conditions as shown in Figures 74a and 74b wherethe simulated mixture component isotherms at 433 K are presented

74 Results 151

0

02

04

06

08

1

12

14

100

101

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

Pure component isotherms CBCFCMC AFI 433K

ethylbenzenestyrene

Figure 73 Single component isotherms of styrene and ethylbenzene in AFI at 433 KInset snapshots of styrene (top) and ethylbenzene (bottom) at 1times103 Pa and 1times109

Pa At low loadings both molecules are adsorbed with the phenyl group parallel to thechannel axis At higher pressures styrene can arrange in a tilted face-to-face stackingconfiguration which reduces its footprint and allows for a higher saturation capacityThe difference in saturation capacities ensures the selectivity of AFI towards styreneat saturation conditions in the mixture

0

01

02

03

04

05

06

07

08

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

01

02

03

04

05

06

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

Figure 74 Mixture component isotherms for and equimolar mixture of styrene andethylbenzene at 433 K in (a) AFI zeolite and (b) MAZ zeolite

Face-to-face stacking is also observed in DON zeolite DON [43] is a structurewith 1D-channels slightly larger than AFI and MAZ zeolites This increase indimensions favors ethylbenzene to undergo a molecular reorientation into a face-to-face stacking configuration but also because the diameter of the channels arelarger than the length of styrene it induces styrene to adopt a commensurate stack-ing configuration where the stacking of two molecules with their phenyl groupsfacing each other is commensurate with the channel dimensions (Figures 75a)This is also observed in MIL-53 [44] a metal-organic framework with lozenge-shaped rhombohedric channels of approximately 085 nm as show in Figure 75b

152 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

a

b

Figure 75 (a) Snapshots of styrene (left) and ethylbenzene (right) in DON zeolite at1e9 Pa and 433 K(b) Snapshots of styrene (left) and ethylbenzene (right) in MIL-53at 1e9 Pa and 433 K Color code carbon (cyan) hydrogen (white)

where snapshots of styrene and ethylbenzene at 1times109 Pa and 433 K are presen-ted Because of styrene and ethylbenzene dimensions styrene with commensuratestacking can obtain higher saturation capacities than ethylbenzene with face-to-face stacking as shown schematically in Figure 76c This is also seen in DONand MIL-53 single component isotherms (Figure 77a) Mixture isotherms (Fig-ures 77b) and breakthrough curves (Figure 77c) further confirm that DON andMIL-53 are styrene selective structures

Commensurate stacking for both styrene and ethylbenzene is observed in MIL-47 [45] and MAF-X8 [46] MIL-47 is a metal-organic framework with lozenge-shaped rhombohedric channels which size (slightly larger than MIL-53) allows forethylbenzene to also have commensurate stacking In Figure 78 we present thesimulated single component adsorption isotherms of ethylbenzene and styrene inMIL-47 at 433 K and snapshots of styrene and ethylbenzene at 1times106 Pa and1times109 Pa At 1times109 Pa both molecules have commensurate stacking but in orderfor ethylbenzene to have commensurate stacking the phenyl groups have to beslightly shifted due to the non-planar ethyl group This shift affects the amount ofethylbenzene molecules that can be adsorbed (as compared to styrene) and there-fore causes a difference in the saturation capacity of styrene and ethylbenzeneInterestingly commensurate stacking occurs at lower loading for styrene than forethylbenzene We can see in Figure 78 that at 1times106 Pa styrene already presentscommensurate stacking while ethylbenzene seems to have more of a face-to-facestacking In a mixture this will favor styrene adsorption even more In Figure79 simulated mixture adsorption isotherms for an equimolar mixture in MIL-47at 433 K are presented MIL-47 is a styrene selective structure Our results are ingood qualitative agreement with Maes et al [6] Breakthrough simulations furtherconfirm that MIL-47 is a styrene selective structure at saturation conditions InMAF-x8 a metal-organic framework with square channels of approximately 1nm

74 Results 153

BEFORE REORIENTATION

AFTER REORIENTATION

096 nm 095 nm

034 nm 053 nm

a b

095 nm

053 nm x2

c

Figure 76 Schematic representation of different entropic effects occurring in theseparation of styrene and ethylbenzene in nanoporous materials (a) The reductionof the moleculersquos footprint in the channels because of a reorientation into a face-to-face stacking configuration (b) Effect of the channel size on the reorientationand therefore moleculersquos footprint reduction (c) Comparison of the channel lengthneeded for two molecules of styrene to have commensurate stacking vs two moleculesof ethylbenzene to have face-to-face stacking

0

05

1

15

2

25

3

35

10-1

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

0

05

1

15

2

25

3

100

101

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

0

05

1

15

2

0 05 1 15 2 25 3 35 4 45

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

0

02

04

06

08

1

12

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

a

0

01

02

03

04

05

06

07

08

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

0

05

1

15

2

0 05 1 15 2 25 3

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

c

Figure 77 Simulated isotherms and breakthrough curves of styrene and ethylben-zene in MIL-53 (top) and DON (bottom) at 433 K (a) Pure component isotherms(lines are dual-site Langmuir-Freundlich fits of the pure components points are thepure component isotherms from CBCFCMC simulations) (b) Mixture componentisotherms for an equimolar mixture The IAST results are in good agreement with themixture isotherms (c) Simulated step-type breakthrough at 1e6 Pa total fugacity

154 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

0

1

2

3

4

5

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

Figure 78 Single component adsorption isotherms for styrene and ethylbenzeneat 433 K in MIL-47 Inset styrene (top) and ethylbenzene (bottom) snapshots at1times106 Pa and 1times109 Pa Styrene has commensurate stacking at lower pressuresthan ethylbenzene Because of the out of plane ethyl group ethylbenzene moleculeshave to be slightly shifted in order to have commensurate stacking and therefore lessethylbenzene molecules can be adsorbed in MIL-47 channels at saturation conditions

0

05

1

15

2

25

3

35

4

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

02

04

06

08

1

12

14

16

18

0 1 2 3 4 5

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

b

Figure 79 Styreneethylbenzene separation using MIL-47 at 433 K (a) equimolarmixture isotherms and Ideal Adsorption Solution Theory (IAST) prediction based onpure component isotherms (b) simulated step breakthrough at 1times106 Pa total fu-gacity The IAST prediction is in excellent agreement with the mixture simulationsThe mixture and breakthrough simulations show a high styrene selectivity and loadingin the mixture

it is easier for ethylbenzene to have commensurate stacking than in MIL-47 Thesingle component isotherms of both molecules behave very similar The topologyof the structure seems to induce a shift between parallel styrene molecules and al-lows for the ethyl group of ethylbenzene to stick in the channel ldquopocketsrdquo (Figure710) This might be the reason for the smaller difference in saturation capacitiesof styrene and ethylbenzene compared to MIL-47

74 Results 155

a b

0

05

1

15

2

25

3

35

4

10-6

10-4

10-2

100

102

104

106

108

Abso

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

c

Figure 710 Snapshots of (a) styrene and (b) ethylbenzene in MAF-X8 at 1e9 Paand 433 K Both styrene and ethylbenzene have commensurate stacking Color codecarbon (cyan) hydrogen (white) (c) Simulated single component isotherms of styreneand ethylbenzene in MAF-X8 at 433 K

Combinationcompetition of mechanisms is observed in structures with a more com-plex topology An example of this is MOF-CJ3 metal-organic framework [47] Thewide segment of the channels are big enough to allow for both styrene and ethyl-benzene to form two parallel rows (commensurate stacking) however the shapeof the channels forces ethylbenzene molecules to adopt a configuration where theethyl group is pointing to the channel ldquopocketsrdquo Styrene has more freedom andmolecules can arrange in a way that an extra styrene can be adsorbed in theprotracted segments in a configuration perpendicular to the channel (face-to-facestacking) (Figure 711) The interplay between different mechanisms rarely makesthe separation better Even when all the mechanisms favor a specific molecule thecompetition between them can induce enough disorder to destroy the selectivityone could achieve with a ldquopurerdquo mechanism

Figure 711 Snapshots of styrene (top) and ethylbenzene (bottom) at 1times109 Paand 433 K in MOF-CJ3

156 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

75 Discussion

There are two important factors to consider when using adsorption for separa-tion processes at industrial conditions namely selectivity and capacity A highselectivity ensures that less cycles are needed to achieve a high degree of purity inthe separation but a high capacity implies that the regeneration time is longerFor a binary mixture the adsorption selectivity is defined as

Sads =q1q2

f1f2(71)

and the capacity is defined as the styrene loading in the adsorbed phase of a binarymixture

Capacity = q1 (72)

In Figure 712 the relationship between these two properties for different struc-tures is presented Structures with the same separation mechanism are plottedwith the same color An ideal structure for the separation would be located at theright top corner

1

10

0 1 2 3 4 5 6

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

MAF-x8

MOF-CJ3

CoBDP IRMOF-1

JUC-77

CuBTC

UiO-66

Figure 712 Selectivity as a function of styrene loading (capacity) in a binary mix-ture at 433 K and 1times106 Pa total fugacity The structures are divided in differentcolors depending on the selectivity mechanism observed Color code Size exclusion(red) face-to-face stacking (purple) commensurate stacking (blue) commensurate-stackingface-to-face stacking (green) mixed(orange) The dotted red line corres-ponds to the ratio styreneethylbenzene at liquid conditions There is a naturaltrade-off between selectivity and loading finding structures in the top right corneris not feasible The black dashed line (guide to the eye) denotes the inverse relation-ship between selectivity and capacity MIL-47 is a styrene selective material with ahigh capacity therefore a good candidate for the styreneethylbenzene separation

75 Discussion 157

In structures where the separation is driven by size-exclusion (MFI-para) theselectivity is high because there is an adsorption site available only for styrene(zig-zag channels) but the capacity is rather low since for size-exclusion to occuramong similar molecules there has to be a very tight fit between the moleculesand the adsorbent usually associated with small pore systems

In structures where the separation is driven by face-to-face stacking the se-lectivity relies on the increase of the moleculersquos footprints difference consequenceof a reorientation and piling Ideally only one of the mixture components shouldbe able to reorient but in the case of styrene and ethylbenzene the similarity intheir length and width makes no significant difference in the pore size needed forthe reorientation The selectivity relies thus in how favorable the reorientationand piling are

In MAZ and AFI both molecules can reorient but the reorientation is restrictedby the pore size forcing a tilting in the face-to-face stacking configuration that isunfavorable for ethylbenzene to form a pile but it is favorable for styrene Thisallows styrene to obtain higher saturation loadings and ensures MAZ and AFIstyrene selectivity at saturation conditions

However face-to-face stacking only occurs at high loadings In Figure 712the selectivity is presented at 1times106 Pa At this fugacity MAZ is not yet styreneselective and AFI selectivity is almost negligible The effect of face-to-face stackingin styrene selectivity can only be observed at higher loadings as shown in Figure713 where the selectivity as a function of styrene loading for higher fugacities ispresented

1

10

1 10

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

CoBDP

CuBTC

IRMOF-1

JUC-77

MAF-X8

MIL-47

MIL-53

MOF-CJ3

AFI

DON

MAZ

MFI-para

UiO-66

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

JUC-77

MAF-x8

IRMOF-1

MOFCJ3

CoBDP

CuBTC

UiO-66

MIL-47

MIL-53

JUC-77

CuBTC

CoBDP

MOF-CJ3 IRMOF-1

MAZDON

MFI-para

AFI

UiO-66

MAF-X8

1e6 Pa1e7 Pa1e9 Pa

Figure 713 Selectivity (qstqeb) in an equimolar mixture at 3 different fugacitiesThe dashed red line corresponds to the liquid phase ratio For all the cases styreneloading increases with the fugacity (shift to the right of the plot) The selectivity alsoincreases in most of the cases with the fugacity (shift upwards) except for MFI-paraJUC-77 MIL-47 and Cu-BTC

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 6: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

74 Results 149

insertions and deletions The method is a combination of the Configurational BiasMonte Carlo (CBMC) [34ndash36] where molecules growth is biased towards favorableconfigurations and Continuous Fractional Component Monte Carlo (CFCMC) [37]in which molecules are gradually inserted or deleted by scaling their interactionswith the surroundings We have shown in previous work [33] that the resultsobtained with this method do not differ from CBMC calculations but the efficiencyis higher

Using the dual-site Langmuir-Freundlich fits of the pure component isothermsbreakthrough calculations were carried out by solving a set of partial differentialequations for each of the species in the gas mixture [38 39] The molar loadingsof the species at any position along the packed bed and at any time were determ-ined from Ideal Adsorbed Solution Theory calculations Video animations of thebreakthrough behavior as a function of time of selected structures are provided asweb-enhanced objects online

74 Results

We perform a screening study of several zeolites and metal-organic frameworksfor the separation of styreneethylbenzene mixture focusing on saturation condi-tions Under these conditions differences in the saturation capacity of the mixturecomponents strongly dictate the separation

In systems with small pores like MRE and MTW zeolites molecules are forcedto adsorb parallel to the channels The saturation capacity is determined by theeffective length per molecule in the channel (footprint) Because of the similarityin the length of styrene and ethylbenzene the difference in saturation capacities isalmost negligible making systems with small pores unsuitable candidates for theseparation

In structures with cavities or channels much larger than styrene and ethyl-benzene molecular dimensions like IRMOF-1 and Zn-DOBDC molecules do notpresent any particular packing The observed difference in the saturation capa-cities is a consequence of the natural packing of the molecules in liquid phase(ρEb = 08665 gmL ρSt = 0909 gmL) This makes these materials also unsuit-able for the separation process

We have identified a few materials where styrene has a higher saturation capa-city than ethylbenzene In the following we describe how this difference arises fromthe previously mentioned entropic mechanisms and we highlight their applicabilityfor the separation process

Size exclusion is observed in MFI-para [40] MFI-para is a ZSM-5 zeolite whichstructure is a combination of interconnected straight and zigzag channels Thestraight channels have a diameter of 53times56 A and the zigzag channels have adiameter of 51times55 A In Figure 72b the simulated single component isothermsof styrene and ethylbenzene in MFI-para at 433 K and snapshots of styrene andethylbenzene at 1times109 Pa and 433 K are presented At low loadings molecules

150 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

preferentially adsorb in the straight channels the difference in loadings arise froma stronger interaction of styrene with MFI-para At saturation conditions styrenecan obtain almost twice the loading of ethylbenzene because of a size exclusioneffect in the zig-zag channels in which ethylbenzene does not fit due to its heightWhen an equimolar styreneethylbenzene mixture is considered the difference inloadings at saturation conditions is even larger (Figure 72c)

a

0

02

04

06

08

1

12

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

b

0

02

04

06

08

1

12

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

c

Figure 72 (a) Snapshots of ethylbenzene (top) and styrene (bottom) at 1times109

Pa and 433 K (b) Single component isotherms of styrene (red) and ethylbenzene(blue) in MFI-para at 433 K (lines are dual-site Langmuir-Freundlich fits of the purecomponents points are the pure component isotherms from CBCFCMC simulations)(c) Mixture component isotherms for an equimolar mixture in MFI-para at 433 K

Face-to-face stacking occurs in MAZ [41] and AFI [42] zeolites MAZ and AFIare 1D-channel zeolites with dimensions that allow a molecular reorientation ofethylbenzene and styrene

In Figure 73 we present the simulation results for the single component iso-therms of ethylbenzene and styrene in AFI zeolite at 433 K At low loadingsmolecules are mostly adsorbed flat on the walls (parallel to the channels axis) ad-sorption is dictated by enthalpy effects which favors ethylbenzene As the loadingincreases the molecules undergo a molecular reorientation from a configurationparallel to the channel axis to a ldquopilerdquo configuration with the phenyl groups fa-cing each other This new configuration (face-to-face stacking) reduces styrenersquosfootprint more than ethylbenzenersquos footprint (Figure 76a) because of the non-planarity of ethylbenzene allowing styrene to obtain higher saturation loadings

For MAZ zeolite a similar behavior is observed however because MAZ zeolitehas smaller channels than AFI the angle at which molecules can reorient has asmaller effect on the reduction of the moleculersquos footprint in the channels (Figure76b) When an equimolar mixture is considered both AFI and MAZ zeolites arestyrene selective at saturation conditions as shown in Figures 74a and 74b wherethe simulated mixture component isotherms at 433 K are presented

74 Results 151

0

02

04

06

08

1

12

14

100

101

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

Pure component isotherms CBCFCMC AFI 433K

ethylbenzenestyrene

Figure 73 Single component isotherms of styrene and ethylbenzene in AFI at 433 KInset snapshots of styrene (top) and ethylbenzene (bottom) at 1times103 Pa and 1times109

Pa At low loadings both molecules are adsorbed with the phenyl group parallel to thechannel axis At higher pressures styrene can arrange in a tilted face-to-face stackingconfiguration which reduces its footprint and allows for a higher saturation capacityThe difference in saturation capacities ensures the selectivity of AFI towards styreneat saturation conditions in the mixture

0

01

02

03

04

05

06

07

08

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

01

02

03

04

05

06

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

Figure 74 Mixture component isotherms for and equimolar mixture of styrene andethylbenzene at 433 K in (a) AFI zeolite and (b) MAZ zeolite

Face-to-face stacking is also observed in DON zeolite DON [43] is a structurewith 1D-channels slightly larger than AFI and MAZ zeolites This increase indimensions favors ethylbenzene to undergo a molecular reorientation into a face-to-face stacking configuration but also because the diameter of the channels arelarger than the length of styrene it induces styrene to adopt a commensurate stack-ing configuration where the stacking of two molecules with their phenyl groupsfacing each other is commensurate with the channel dimensions (Figures 75a)This is also observed in MIL-53 [44] a metal-organic framework with lozenge-shaped rhombohedric channels of approximately 085 nm as show in Figure 75b

152 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

a

b

Figure 75 (a) Snapshots of styrene (left) and ethylbenzene (right) in DON zeolite at1e9 Pa and 433 K(b) Snapshots of styrene (left) and ethylbenzene (right) in MIL-53at 1e9 Pa and 433 K Color code carbon (cyan) hydrogen (white)

where snapshots of styrene and ethylbenzene at 1times109 Pa and 433 K are presen-ted Because of styrene and ethylbenzene dimensions styrene with commensuratestacking can obtain higher saturation capacities than ethylbenzene with face-to-face stacking as shown schematically in Figure 76c This is also seen in DONand MIL-53 single component isotherms (Figure 77a) Mixture isotherms (Fig-ures 77b) and breakthrough curves (Figure 77c) further confirm that DON andMIL-53 are styrene selective structures

Commensurate stacking for both styrene and ethylbenzene is observed in MIL-47 [45] and MAF-X8 [46] MIL-47 is a metal-organic framework with lozenge-shaped rhombohedric channels which size (slightly larger than MIL-53) allows forethylbenzene to also have commensurate stacking In Figure 78 we present thesimulated single component adsorption isotherms of ethylbenzene and styrene inMIL-47 at 433 K and snapshots of styrene and ethylbenzene at 1times106 Pa and1times109 Pa At 1times109 Pa both molecules have commensurate stacking but in orderfor ethylbenzene to have commensurate stacking the phenyl groups have to beslightly shifted due to the non-planar ethyl group This shift affects the amount ofethylbenzene molecules that can be adsorbed (as compared to styrene) and there-fore causes a difference in the saturation capacity of styrene and ethylbenzeneInterestingly commensurate stacking occurs at lower loading for styrene than forethylbenzene We can see in Figure 78 that at 1times106 Pa styrene already presentscommensurate stacking while ethylbenzene seems to have more of a face-to-facestacking In a mixture this will favor styrene adsorption even more In Figure79 simulated mixture adsorption isotherms for an equimolar mixture in MIL-47at 433 K are presented MIL-47 is a styrene selective structure Our results are ingood qualitative agreement with Maes et al [6] Breakthrough simulations furtherconfirm that MIL-47 is a styrene selective structure at saturation conditions InMAF-x8 a metal-organic framework with square channels of approximately 1nm

74 Results 153

BEFORE REORIENTATION

AFTER REORIENTATION

096 nm 095 nm

034 nm 053 nm

a b

095 nm

053 nm x2

c

Figure 76 Schematic representation of different entropic effects occurring in theseparation of styrene and ethylbenzene in nanoporous materials (a) The reductionof the moleculersquos footprint in the channels because of a reorientation into a face-to-face stacking configuration (b) Effect of the channel size on the reorientationand therefore moleculersquos footprint reduction (c) Comparison of the channel lengthneeded for two molecules of styrene to have commensurate stacking vs two moleculesof ethylbenzene to have face-to-face stacking

0

05

1

15

2

25

3

35

10-1

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

0

05

1

15

2

25

3

100

101

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

0

05

1

15

2

0 05 1 15 2 25 3 35 4 45

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

0

02

04

06

08

1

12

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

a

0

01

02

03

04

05

06

07

08

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

0

05

1

15

2

0 05 1 15 2 25 3

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

c

Figure 77 Simulated isotherms and breakthrough curves of styrene and ethylben-zene in MIL-53 (top) and DON (bottom) at 433 K (a) Pure component isotherms(lines are dual-site Langmuir-Freundlich fits of the pure components points are thepure component isotherms from CBCFCMC simulations) (b) Mixture componentisotherms for an equimolar mixture The IAST results are in good agreement with themixture isotherms (c) Simulated step-type breakthrough at 1e6 Pa total fugacity

154 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

0

1

2

3

4

5

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

Figure 78 Single component adsorption isotherms for styrene and ethylbenzeneat 433 K in MIL-47 Inset styrene (top) and ethylbenzene (bottom) snapshots at1times106 Pa and 1times109 Pa Styrene has commensurate stacking at lower pressuresthan ethylbenzene Because of the out of plane ethyl group ethylbenzene moleculeshave to be slightly shifted in order to have commensurate stacking and therefore lessethylbenzene molecules can be adsorbed in MIL-47 channels at saturation conditions

0

05

1

15

2

25

3

35

4

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

02

04

06

08

1

12

14

16

18

0 1 2 3 4 5

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

b

Figure 79 Styreneethylbenzene separation using MIL-47 at 433 K (a) equimolarmixture isotherms and Ideal Adsorption Solution Theory (IAST) prediction based onpure component isotherms (b) simulated step breakthrough at 1times106 Pa total fu-gacity The IAST prediction is in excellent agreement with the mixture simulationsThe mixture and breakthrough simulations show a high styrene selectivity and loadingin the mixture

it is easier for ethylbenzene to have commensurate stacking than in MIL-47 Thesingle component isotherms of both molecules behave very similar The topologyof the structure seems to induce a shift between parallel styrene molecules and al-lows for the ethyl group of ethylbenzene to stick in the channel ldquopocketsrdquo (Figure710) This might be the reason for the smaller difference in saturation capacitiesof styrene and ethylbenzene compared to MIL-47

74 Results 155

a b

0

05

1

15

2

25

3

35

4

10-6

10-4

10-2

100

102

104

106

108

Abso

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

c

Figure 710 Snapshots of (a) styrene and (b) ethylbenzene in MAF-X8 at 1e9 Paand 433 K Both styrene and ethylbenzene have commensurate stacking Color codecarbon (cyan) hydrogen (white) (c) Simulated single component isotherms of styreneand ethylbenzene in MAF-X8 at 433 K

Combinationcompetition of mechanisms is observed in structures with a more com-plex topology An example of this is MOF-CJ3 metal-organic framework [47] Thewide segment of the channels are big enough to allow for both styrene and ethyl-benzene to form two parallel rows (commensurate stacking) however the shapeof the channels forces ethylbenzene molecules to adopt a configuration where theethyl group is pointing to the channel ldquopocketsrdquo Styrene has more freedom andmolecules can arrange in a way that an extra styrene can be adsorbed in theprotracted segments in a configuration perpendicular to the channel (face-to-facestacking) (Figure 711) The interplay between different mechanisms rarely makesthe separation better Even when all the mechanisms favor a specific molecule thecompetition between them can induce enough disorder to destroy the selectivityone could achieve with a ldquopurerdquo mechanism

Figure 711 Snapshots of styrene (top) and ethylbenzene (bottom) at 1times109 Paand 433 K in MOF-CJ3

156 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

75 Discussion

There are two important factors to consider when using adsorption for separa-tion processes at industrial conditions namely selectivity and capacity A highselectivity ensures that less cycles are needed to achieve a high degree of purity inthe separation but a high capacity implies that the regeneration time is longerFor a binary mixture the adsorption selectivity is defined as

Sads =q1q2

f1f2(71)

and the capacity is defined as the styrene loading in the adsorbed phase of a binarymixture

Capacity = q1 (72)

In Figure 712 the relationship between these two properties for different struc-tures is presented Structures with the same separation mechanism are plottedwith the same color An ideal structure for the separation would be located at theright top corner

1

10

0 1 2 3 4 5 6

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

MAF-x8

MOF-CJ3

CoBDP IRMOF-1

JUC-77

CuBTC

UiO-66

Figure 712 Selectivity as a function of styrene loading (capacity) in a binary mix-ture at 433 K and 1times106 Pa total fugacity The structures are divided in differentcolors depending on the selectivity mechanism observed Color code Size exclusion(red) face-to-face stacking (purple) commensurate stacking (blue) commensurate-stackingface-to-face stacking (green) mixed(orange) The dotted red line corres-ponds to the ratio styreneethylbenzene at liquid conditions There is a naturaltrade-off between selectivity and loading finding structures in the top right corneris not feasible The black dashed line (guide to the eye) denotes the inverse relation-ship between selectivity and capacity MIL-47 is a styrene selective material with ahigh capacity therefore a good candidate for the styreneethylbenzene separation

75 Discussion 157

In structures where the separation is driven by size-exclusion (MFI-para) theselectivity is high because there is an adsorption site available only for styrene(zig-zag channels) but the capacity is rather low since for size-exclusion to occuramong similar molecules there has to be a very tight fit between the moleculesand the adsorbent usually associated with small pore systems

In structures where the separation is driven by face-to-face stacking the se-lectivity relies on the increase of the moleculersquos footprints difference consequenceof a reorientation and piling Ideally only one of the mixture components shouldbe able to reorient but in the case of styrene and ethylbenzene the similarity intheir length and width makes no significant difference in the pore size needed forthe reorientation The selectivity relies thus in how favorable the reorientationand piling are

In MAZ and AFI both molecules can reorient but the reorientation is restrictedby the pore size forcing a tilting in the face-to-face stacking configuration that isunfavorable for ethylbenzene to form a pile but it is favorable for styrene Thisallows styrene to obtain higher saturation loadings and ensures MAZ and AFIstyrene selectivity at saturation conditions

However face-to-face stacking only occurs at high loadings In Figure 712the selectivity is presented at 1times106 Pa At this fugacity MAZ is not yet styreneselective and AFI selectivity is almost negligible The effect of face-to-face stackingin styrene selectivity can only be observed at higher loadings as shown in Figure713 where the selectivity as a function of styrene loading for higher fugacities ispresented

1

10

1 10

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

CoBDP

CuBTC

IRMOF-1

JUC-77

MAF-X8

MIL-47

MIL-53

MOF-CJ3

AFI

DON

MAZ

MFI-para

UiO-66

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

JUC-77

MAF-x8

IRMOF-1

MOFCJ3

CoBDP

CuBTC

UiO-66

MIL-47

MIL-53

JUC-77

CuBTC

CoBDP

MOF-CJ3 IRMOF-1

MAZDON

MFI-para

AFI

UiO-66

MAF-X8

1e6 Pa1e7 Pa1e9 Pa

Figure 713 Selectivity (qstqeb) in an equimolar mixture at 3 different fugacitiesThe dashed red line corresponds to the liquid phase ratio For all the cases styreneloading increases with the fugacity (shift to the right of the plot) The selectivity alsoincreases in most of the cases with the fugacity (shift upwards) except for MFI-paraJUC-77 MIL-47 and Cu-BTC

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 7: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

150 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

preferentially adsorb in the straight channels the difference in loadings arise froma stronger interaction of styrene with MFI-para At saturation conditions styrenecan obtain almost twice the loading of ethylbenzene because of a size exclusioneffect in the zig-zag channels in which ethylbenzene does not fit due to its heightWhen an equimolar styreneethylbenzene mixture is considered the difference inloadings at saturation conditions is even larger (Figure 72c)

a

0

02

04

06

08

1

12

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

b

0

02

04

06

08

1

12

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

c

Figure 72 (a) Snapshots of ethylbenzene (top) and styrene (bottom) at 1times109

Pa and 433 K (b) Single component isotherms of styrene (red) and ethylbenzene(blue) in MFI-para at 433 K (lines are dual-site Langmuir-Freundlich fits of the purecomponents points are the pure component isotherms from CBCFCMC simulations)(c) Mixture component isotherms for an equimolar mixture in MFI-para at 433 K

Face-to-face stacking occurs in MAZ [41] and AFI [42] zeolites MAZ and AFIare 1D-channel zeolites with dimensions that allow a molecular reorientation ofethylbenzene and styrene

In Figure 73 we present the simulation results for the single component iso-therms of ethylbenzene and styrene in AFI zeolite at 433 K At low loadingsmolecules are mostly adsorbed flat on the walls (parallel to the channels axis) ad-sorption is dictated by enthalpy effects which favors ethylbenzene As the loadingincreases the molecules undergo a molecular reorientation from a configurationparallel to the channel axis to a ldquopilerdquo configuration with the phenyl groups fa-cing each other This new configuration (face-to-face stacking) reduces styrenersquosfootprint more than ethylbenzenersquos footprint (Figure 76a) because of the non-planarity of ethylbenzene allowing styrene to obtain higher saturation loadings

For MAZ zeolite a similar behavior is observed however because MAZ zeolitehas smaller channels than AFI the angle at which molecules can reorient has asmaller effect on the reduction of the moleculersquos footprint in the channels (Figure76b) When an equimolar mixture is considered both AFI and MAZ zeolites arestyrene selective at saturation conditions as shown in Figures 74a and 74b wherethe simulated mixture component isotherms at 433 K are presented

74 Results 151

0

02

04

06

08

1

12

14

100

101

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

Pure component isotherms CBCFCMC AFI 433K

ethylbenzenestyrene

Figure 73 Single component isotherms of styrene and ethylbenzene in AFI at 433 KInset snapshots of styrene (top) and ethylbenzene (bottom) at 1times103 Pa and 1times109

Pa At low loadings both molecules are adsorbed with the phenyl group parallel to thechannel axis At higher pressures styrene can arrange in a tilted face-to-face stackingconfiguration which reduces its footprint and allows for a higher saturation capacityThe difference in saturation capacities ensures the selectivity of AFI towards styreneat saturation conditions in the mixture

0

01

02

03

04

05

06

07

08

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

01

02

03

04

05

06

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

Figure 74 Mixture component isotherms for and equimolar mixture of styrene andethylbenzene at 433 K in (a) AFI zeolite and (b) MAZ zeolite

Face-to-face stacking is also observed in DON zeolite DON [43] is a structurewith 1D-channels slightly larger than AFI and MAZ zeolites This increase indimensions favors ethylbenzene to undergo a molecular reorientation into a face-to-face stacking configuration but also because the diameter of the channels arelarger than the length of styrene it induces styrene to adopt a commensurate stack-ing configuration where the stacking of two molecules with their phenyl groupsfacing each other is commensurate with the channel dimensions (Figures 75a)This is also observed in MIL-53 [44] a metal-organic framework with lozenge-shaped rhombohedric channels of approximately 085 nm as show in Figure 75b

152 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

a

b

Figure 75 (a) Snapshots of styrene (left) and ethylbenzene (right) in DON zeolite at1e9 Pa and 433 K(b) Snapshots of styrene (left) and ethylbenzene (right) in MIL-53at 1e9 Pa and 433 K Color code carbon (cyan) hydrogen (white)

where snapshots of styrene and ethylbenzene at 1times109 Pa and 433 K are presen-ted Because of styrene and ethylbenzene dimensions styrene with commensuratestacking can obtain higher saturation capacities than ethylbenzene with face-to-face stacking as shown schematically in Figure 76c This is also seen in DONand MIL-53 single component isotherms (Figure 77a) Mixture isotherms (Fig-ures 77b) and breakthrough curves (Figure 77c) further confirm that DON andMIL-53 are styrene selective structures

Commensurate stacking for both styrene and ethylbenzene is observed in MIL-47 [45] and MAF-X8 [46] MIL-47 is a metal-organic framework with lozenge-shaped rhombohedric channels which size (slightly larger than MIL-53) allows forethylbenzene to also have commensurate stacking In Figure 78 we present thesimulated single component adsorption isotherms of ethylbenzene and styrene inMIL-47 at 433 K and snapshots of styrene and ethylbenzene at 1times106 Pa and1times109 Pa At 1times109 Pa both molecules have commensurate stacking but in orderfor ethylbenzene to have commensurate stacking the phenyl groups have to beslightly shifted due to the non-planar ethyl group This shift affects the amount ofethylbenzene molecules that can be adsorbed (as compared to styrene) and there-fore causes a difference in the saturation capacity of styrene and ethylbenzeneInterestingly commensurate stacking occurs at lower loading for styrene than forethylbenzene We can see in Figure 78 that at 1times106 Pa styrene already presentscommensurate stacking while ethylbenzene seems to have more of a face-to-facestacking In a mixture this will favor styrene adsorption even more In Figure79 simulated mixture adsorption isotherms for an equimolar mixture in MIL-47at 433 K are presented MIL-47 is a styrene selective structure Our results are ingood qualitative agreement with Maes et al [6] Breakthrough simulations furtherconfirm that MIL-47 is a styrene selective structure at saturation conditions InMAF-x8 a metal-organic framework with square channels of approximately 1nm

74 Results 153

BEFORE REORIENTATION

AFTER REORIENTATION

096 nm 095 nm

034 nm 053 nm

a b

095 nm

053 nm x2

c

Figure 76 Schematic representation of different entropic effects occurring in theseparation of styrene and ethylbenzene in nanoporous materials (a) The reductionof the moleculersquos footprint in the channels because of a reorientation into a face-to-face stacking configuration (b) Effect of the channel size on the reorientationand therefore moleculersquos footprint reduction (c) Comparison of the channel lengthneeded for two molecules of styrene to have commensurate stacking vs two moleculesof ethylbenzene to have face-to-face stacking

0

05

1

15

2

25

3

35

10-1

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

0

05

1

15

2

25

3

100

101

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

0

05

1

15

2

0 05 1 15 2 25 3 35 4 45

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

0

02

04

06

08

1

12

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

a

0

01

02

03

04

05

06

07

08

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

0

05

1

15

2

0 05 1 15 2 25 3

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

c

Figure 77 Simulated isotherms and breakthrough curves of styrene and ethylben-zene in MIL-53 (top) and DON (bottom) at 433 K (a) Pure component isotherms(lines are dual-site Langmuir-Freundlich fits of the pure components points are thepure component isotherms from CBCFCMC simulations) (b) Mixture componentisotherms for an equimolar mixture The IAST results are in good agreement with themixture isotherms (c) Simulated step-type breakthrough at 1e6 Pa total fugacity

154 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

0

1

2

3

4

5

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

Figure 78 Single component adsorption isotherms for styrene and ethylbenzeneat 433 K in MIL-47 Inset styrene (top) and ethylbenzene (bottom) snapshots at1times106 Pa and 1times109 Pa Styrene has commensurate stacking at lower pressuresthan ethylbenzene Because of the out of plane ethyl group ethylbenzene moleculeshave to be slightly shifted in order to have commensurate stacking and therefore lessethylbenzene molecules can be adsorbed in MIL-47 channels at saturation conditions

0

05

1

15

2

25

3

35

4

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

02

04

06

08

1

12

14

16

18

0 1 2 3 4 5

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

b

Figure 79 Styreneethylbenzene separation using MIL-47 at 433 K (a) equimolarmixture isotherms and Ideal Adsorption Solution Theory (IAST) prediction based onpure component isotherms (b) simulated step breakthrough at 1times106 Pa total fu-gacity The IAST prediction is in excellent agreement with the mixture simulationsThe mixture and breakthrough simulations show a high styrene selectivity and loadingin the mixture

it is easier for ethylbenzene to have commensurate stacking than in MIL-47 Thesingle component isotherms of both molecules behave very similar The topologyof the structure seems to induce a shift between parallel styrene molecules and al-lows for the ethyl group of ethylbenzene to stick in the channel ldquopocketsrdquo (Figure710) This might be the reason for the smaller difference in saturation capacitiesof styrene and ethylbenzene compared to MIL-47

74 Results 155

a b

0

05

1

15

2

25

3

35

4

10-6

10-4

10-2

100

102

104

106

108

Abso

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

c

Figure 710 Snapshots of (a) styrene and (b) ethylbenzene in MAF-X8 at 1e9 Paand 433 K Both styrene and ethylbenzene have commensurate stacking Color codecarbon (cyan) hydrogen (white) (c) Simulated single component isotherms of styreneand ethylbenzene in MAF-X8 at 433 K

Combinationcompetition of mechanisms is observed in structures with a more com-plex topology An example of this is MOF-CJ3 metal-organic framework [47] Thewide segment of the channels are big enough to allow for both styrene and ethyl-benzene to form two parallel rows (commensurate stacking) however the shapeof the channels forces ethylbenzene molecules to adopt a configuration where theethyl group is pointing to the channel ldquopocketsrdquo Styrene has more freedom andmolecules can arrange in a way that an extra styrene can be adsorbed in theprotracted segments in a configuration perpendicular to the channel (face-to-facestacking) (Figure 711) The interplay between different mechanisms rarely makesthe separation better Even when all the mechanisms favor a specific molecule thecompetition between them can induce enough disorder to destroy the selectivityone could achieve with a ldquopurerdquo mechanism

Figure 711 Snapshots of styrene (top) and ethylbenzene (bottom) at 1times109 Paand 433 K in MOF-CJ3

156 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

75 Discussion

There are two important factors to consider when using adsorption for separa-tion processes at industrial conditions namely selectivity and capacity A highselectivity ensures that less cycles are needed to achieve a high degree of purity inthe separation but a high capacity implies that the regeneration time is longerFor a binary mixture the adsorption selectivity is defined as

Sads =q1q2

f1f2(71)

and the capacity is defined as the styrene loading in the adsorbed phase of a binarymixture

Capacity = q1 (72)

In Figure 712 the relationship between these two properties for different struc-tures is presented Structures with the same separation mechanism are plottedwith the same color An ideal structure for the separation would be located at theright top corner

1

10

0 1 2 3 4 5 6

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

MAF-x8

MOF-CJ3

CoBDP IRMOF-1

JUC-77

CuBTC

UiO-66

Figure 712 Selectivity as a function of styrene loading (capacity) in a binary mix-ture at 433 K and 1times106 Pa total fugacity The structures are divided in differentcolors depending on the selectivity mechanism observed Color code Size exclusion(red) face-to-face stacking (purple) commensurate stacking (blue) commensurate-stackingface-to-face stacking (green) mixed(orange) The dotted red line corres-ponds to the ratio styreneethylbenzene at liquid conditions There is a naturaltrade-off between selectivity and loading finding structures in the top right corneris not feasible The black dashed line (guide to the eye) denotes the inverse relation-ship between selectivity and capacity MIL-47 is a styrene selective material with ahigh capacity therefore a good candidate for the styreneethylbenzene separation

75 Discussion 157

In structures where the separation is driven by size-exclusion (MFI-para) theselectivity is high because there is an adsorption site available only for styrene(zig-zag channels) but the capacity is rather low since for size-exclusion to occuramong similar molecules there has to be a very tight fit between the moleculesand the adsorbent usually associated with small pore systems

In structures where the separation is driven by face-to-face stacking the se-lectivity relies on the increase of the moleculersquos footprints difference consequenceof a reorientation and piling Ideally only one of the mixture components shouldbe able to reorient but in the case of styrene and ethylbenzene the similarity intheir length and width makes no significant difference in the pore size needed forthe reorientation The selectivity relies thus in how favorable the reorientationand piling are

In MAZ and AFI both molecules can reorient but the reorientation is restrictedby the pore size forcing a tilting in the face-to-face stacking configuration that isunfavorable for ethylbenzene to form a pile but it is favorable for styrene Thisallows styrene to obtain higher saturation loadings and ensures MAZ and AFIstyrene selectivity at saturation conditions

However face-to-face stacking only occurs at high loadings In Figure 712the selectivity is presented at 1times106 Pa At this fugacity MAZ is not yet styreneselective and AFI selectivity is almost negligible The effect of face-to-face stackingin styrene selectivity can only be observed at higher loadings as shown in Figure713 where the selectivity as a function of styrene loading for higher fugacities ispresented

1

10

1 10

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

CoBDP

CuBTC

IRMOF-1

JUC-77

MAF-X8

MIL-47

MIL-53

MOF-CJ3

AFI

DON

MAZ

MFI-para

UiO-66

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

JUC-77

MAF-x8

IRMOF-1

MOFCJ3

CoBDP

CuBTC

UiO-66

MIL-47

MIL-53

JUC-77

CuBTC

CoBDP

MOF-CJ3 IRMOF-1

MAZDON

MFI-para

AFI

UiO-66

MAF-X8

1e6 Pa1e7 Pa1e9 Pa

Figure 713 Selectivity (qstqeb) in an equimolar mixture at 3 different fugacitiesThe dashed red line corresponds to the liquid phase ratio For all the cases styreneloading increases with the fugacity (shift to the right of the plot) The selectivity alsoincreases in most of the cases with the fugacity (shift upwards) except for MFI-paraJUC-77 MIL-47 and Cu-BTC

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 8: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

74 Results 151

0

02

04

06

08

1

12

14

100

101

102

103

104

105

106

107

108

109

Ab

so

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

Pure component isotherms CBCFCMC AFI 433K

ethylbenzenestyrene

Figure 73 Single component isotherms of styrene and ethylbenzene in AFI at 433 KInset snapshots of styrene (top) and ethylbenzene (bottom) at 1times103 Pa and 1times109

Pa At low loadings both molecules are adsorbed with the phenyl group parallel to thechannel axis At higher pressures styrene can arrange in a tilted face-to-face stackingconfiguration which reduces its footprint and allows for a higher saturation capacityThe difference in saturation capacities ensures the selectivity of AFI towards styreneat saturation conditions in the mixture

0

01

02

03

04

05

06

07

08

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

01

02

03

04

05

06

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

Figure 74 Mixture component isotherms for and equimolar mixture of styrene andethylbenzene at 433 K in (a) AFI zeolite and (b) MAZ zeolite

Face-to-face stacking is also observed in DON zeolite DON [43] is a structurewith 1D-channels slightly larger than AFI and MAZ zeolites This increase indimensions favors ethylbenzene to undergo a molecular reorientation into a face-to-face stacking configuration but also because the diameter of the channels arelarger than the length of styrene it induces styrene to adopt a commensurate stack-ing configuration where the stacking of two molecules with their phenyl groupsfacing each other is commensurate with the channel dimensions (Figures 75a)This is also observed in MIL-53 [44] a metal-organic framework with lozenge-shaped rhombohedric channels of approximately 085 nm as show in Figure 75b

152 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

a

b

Figure 75 (a) Snapshots of styrene (left) and ethylbenzene (right) in DON zeolite at1e9 Pa and 433 K(b) Snapshots of styrene (left) and ethylbenzene (right) in MIL-53at 1e9 Pa and 433 K Color code carbon (cyan) hydrogen (white)

where snapshots of styrene and ethylbenzene at 1times109 Pa and 433 K are presen-ted Because of styrene and ethylbenzene dimensions styrene with commensuratestacking can obtain higher saturation capacities than ethylbenzene with face-to-face stacking as shown schematically in Figure 76c This is also seen in DONand MIL-53 single component isotherms (Figure 77a) Mixture isotherms (Fig-ures 77b) and breakthrough curves (Figure 77c) further confirm that DON andMIL-53 are styrene selective structures

Commensurate stacking for both styrene and ethylbenzene is observed in MIL-47 [45] and MAF-X8 [46] MIL-47 is a metal-organic framework with lozenge-shaped rhombohedric channels which size (slightly larger than MIL-53) allows forethylbenzene to also have commensurate stacking In Figure 78 we present thesimulated single component adsorption isotherms of ethylbenzene and styrene inMIL-47 at 433 K and snapshots of styrene and ethylbenzene at 1times106 Pa and1times109 Pa At 1times109 Pa both molecules have commensurate stacking but in orderfor ethylbenzene to have commensurate stacking the phenyl groups have to beslightly shifted due to the non-planar ethyl group This shift affects the amount ofethylbenzene molecules that can be adsorbed (as compared to styrene) and there-fore causes a difference in the saturation capacity of styrene and ethylbenzeneInterestingly commensurate stacking occurs at lower loading for styrene than forethylbenzene We can see in Figure 78 that at 1times106 Pa styrene already presentscommensurate stacking while ethylbenzene seems to have more of a face-to-facestacking In a mixture this will favor styrene adsorption even more In Figure79 simulated mixture adsorption isotherms for an equimolar mixture in MIL-47at 433 K are presented MIL-47 is a styrene selective structure Our results are ingood qualitative agreement with Maes et al [6] Breakthrough simulations furtherconfirm that MIL-47 is a styrene selective structure at saturation conditions InMAF-x8 a metal-organic framework with square channels of approximately 1nm

74 Results 153

BEFORE REORIENTATION

AFTER REORIENTATION

096 nm 095 nm

034 nm 053 nm

a b

095 nm

053 nm x2

c

Figure 76 Schematic representation of different entropic effects occurring in theseparation of styrene and ethylbenzene in nanoporous materials (a) The reductionof the moleculersquos footprint in the channels because of a reorientation into a face-to-face stacking configuration (b) Effect of the channel size on the reorientationand therefore moleculersquos footprint reduction (c) Comparison of the channel lengthneeded for two molecules of styrene to have commensurate stacking vs two moleculesof ethylbenzene to have face-to-face stacking

0

05

1

15

2

25

3

35

10-1

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

0

05

1

15

2

25

3

100

101

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

0

05

1

15

2

0 05 1 15 2 25 3 35 4 45

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

0

02

04

06

08

1

12

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

a

0

01

02

03

04

05

06

07

08

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

0

05

1

15

2

0 05 1 15 2 25 3

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

c

Figure 77 Simulated isotherms and breakthrough curves of styrene and ethylben-zene in MIL-53 (top) and DON (bottom) at 433 K (a) Pure component isotherms(lines are dual-site Langmuir-Freundlich fits of the pure components points are thepure component isotherms from CBCFCMC simulations) (b) Mixture componentisotherms for an equimolar mixture The IAST results are in good agreement with themixture isotherms (c) Simulated step-type breakthrough at 1e6 Pa total fugacity

154 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

0

1

2

3

4

5

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

Figure 78 Single component adsorption isotherms for styrene and ethylbenzeneat 433 K in MIL-47 Inset styrene (top) and ethylbenzene (bottom) snapshots at1times106 Pa and 1times109 Pa Styrene has commensurate stacking at lower pressuresthan ethylbenzene Because of the out of plane ethyl group ethylbenzene moleculeshave to be slightly shifted in order to have commensurate stacking and therefore lessethylbenzene molecules can be adsorbed in MIL-47 channels at saturation conditions

0

05

1

15

2

25

3

35

4

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

02

04

06

08

1

12

14

16

18

0 1 2 3 4 5

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

b

Figure 79 Styreneethylbenzene separation using MIL-47 at 433 K (a) equimolarmixture isotherms and Ideal Adsorption Solution Theory (IAST) prediction based onpure component isotherms (b) simulated step breakthrough at 1times106 Pa total fu-gacity The IAST prediction is in excellent agreement with the mixture simulationsThe mixture and breakthrough simulations show a high styrene selectivity and loadingin the mixture

it is easier for ethylbenzene to have commensurate stacking than in MIL-47 Thesingle component isotherms of both molecules behave very similar The topologyof the structure seems to induce a shift between parallel styrene molecules and al-lows for the ethyl group of ethylbenzene to stick in the channel ldquopocketsrdquo (Figure710) This might be the reason for the smaller difference in saturation capacitiesof styrene and ethylbenzene compared to MIL-47

74 Results 155

a b

0

05

1

15

2

25

3

35

4

10-6

10-4

10-2

100

102

104

106

108

Abso

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

c

Figure 710 Snapshots of (a) styrene and (b) ethylbenzene in MAF-X8 at 1e9 Paand 433 K Both styrene and ethylbenzene have commensurate stacking Color codecarbon (cyan) hydrogen (white) (c) Simulated single component isotherms of styreneand ethylbenzene in MAF-X8 at 433 K

Combinationcompetition of mechanisms is observed in structures with a more com-plex topology An example of this is MOF-CJ3 metal-organic framework [47] Thewide segment of the channels are big enough to allow for both styrene and ethyl-benzene to form two parallel rows (commensurate stacking) however the shapeof the channels forces ethylbenzene molecules to adopt a configuration where theethyl group is pointing to the channel ldquopocketsrdquo Styrene has more freedom andmolecules can arrange in a way that an extra styrene can be adsorbed in theprotracted segments in a configuration perpendicular to the channel (face-to-facestacking) (Figure 711) The interplay between different mechanisms rarely makesthe separation better Even when all the mechanisms favor a specific molecule thecompetition between them can induce enough disorder to destroy the selectivityone could achieve with a ldquopurerdquo mechanism

Figure 711 Snapshots of styrene (top) and ethylbenzene (bottom) at 1times109 Paand 433 K in MOF-CJ3

156 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

75 Discussion

There are two important factors to consider when using adsorption for separa-tion processes at industrial conditions namely selectivity and capacity A highselectivity ensures that less cycles are needed to achieve a high degree of purity inthe separation but a high capacity implies that the regeneration time is longerFor a binary mixture the adsorption selectivity is defined as

Sads =q1q2

f1f2(71)

and the capacity is defined as the styrene loading in the adsorbed phase of a binarymixture

Capacity = q1 (72)

In Figure 712 the relationship between these two properties for different struc-tures is presented Structures with the same separation mechanism are plottedwith the same color An ideal structure for the separation would be located at theright top corner

1

10

0 1 2 3 4 5 6

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

MAF-x8

MOF-CJ3

CoBDP IRMOF-1

JUC-77

CuBTC

UiO-66

Figure 712 Selectivity as a function of styrene loading (capacity) in a binary mix-ture at 433 K and 1times106 Pa total fugacity The structures are divided in differentcolors depending on the selectivity mechanism observed Color code Size exclusion(red) face-to-face stacking (purple) commensurate stacking (blue) commensurate-stackingface-to-face stacking (green) mixed(orange) The dotted red line corres-ponds to the ratio styreneethylbenzene at liquid conditions There is a naturaltrade-off between selectivity and loading finding structures in the top right corneris not feasible The black dashed line (guide to the eye) denotes the inverse relation-ship between selectivity and capacity MIL-47 is a styrene selective material with ahigh capacity therefore a good candidate for the styreneethylbenzene separation

75 Discussion 157

In structures where the separation is driven by size-exclusion (MFI-para) theselectivity is high because there is an adsorption site available only for styrene(zig-zag channels) but the capacity is rather low since for size-exclusion to occuramong similar molecules there has to be a very tight fit between the moleculesand the adsorbent usually associated with small pore systems

In structures where the separation is driven by face-to-face stacking the se-lectivity relies on the increase of the moleculersquos footprints difference consequenceof a reorientation and piling Ideally only one of the mixture components shouldbe able to reorient but in the case of styrene and ethylbenzene the similarity intheir length and width makes no significant difference in the pore size needed forthe reorientation The selectivity relies thus in how favorable the reorientationand piling are

In MAZ and AFI both molecules can reorient but the reorientation is restrictedby the pore size forcing a tilting in the face-to-face stacking configuration that isunfavorable for ethylbenzene to form a pile but it is favorable for styrene Thisallows styrene to obtain higher saturation loadings and ensures MAZ and AFIstyrene selectivity at saturation conditions

However face-to-face stacking only occurs at high loadings In Figure 712the selectivity is presented at 1times106 Pa At this fugacity MAZ is not yet styreneselective and AFI selectivity is almost negligible The effect of face-to-face stackingin styrene selectivity can only be observed at higher loadings as shown in Figure713 where the selectivity as a function of styrene loading for higher fugacities ispresented

1

10

1 10

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

CoBDP

CuBTC

IRMOF-1

JUC-77

MAF-X8

MIL-47

MIL-53

MOF-CJ3

AFI

DON

MAZ

MFI-para

UiO-66

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

JUC-77

MAF-x8

IRMOF-1

MOFCJ3

CoBDP

CuBTC

UiO-66

MIL-47

MIL-53

JUC-77

CuBTC

CoBDP

MOF-CJ3 IRMOF-1

MAZDON

MFI-para

AFI

UiO-66

MAF-X8

1e6 Pa1e7 Pa1e9 Pa

Figure 713 Selectivity (qstqeb) in an equimolar mixture at 3 different fugacitiesThe dashed red line corresponds to the liquid phase ratio For all the cases styreneloading increases with the fugacity (shift to the right of the plot) The selectivity alsoincreases in most of the cases with the fugacity (shift upwards) except for MFI-paraJUC-77 MIL-47 and Cu-BTC

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 9: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

152 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

a

b

Figure 75 (a) Snapshots of styrene (left) and ethylbenzene (right) in DON zeolite at1e9 Pa and 433 K(b) Snapshots of styrene (left) and ethylbenzene (right) in MIL-53at 1e9 Pa and 433 K Color code carbon (cyan) hydrogen (white)

where snapshots of styrene and ethylbenzene at 1times109 Pa and 433 K are presen-ted Because of styrene and ethylbenzene dimensions styrene with commensuratestacking can obtain higher saturation capacities than ethylbenzene with face-to-face stacking as shown schematically in Figure 76c This is also seen in DONand MIL-53 single component isotherms (Figure 77a) Mixture isotherms (Fig-ures 77b) and breakthrough curves (Figure 77c) further confirm that DON andMIL-53 are styrene selective structures

Commensurate stacking for both styrene and ethylbenzene is observed in MIL-47 [45] and MAF-X8 [46] MIL-47 is a metal-organic framework with lozenge-shaped rhombohedric channels which size (slightly larger than MIL-53) allows forethylbenzene to also have commensurate stacking In Figure 78 we present thesimulated single component adsorption isotherms of ethylbenzene and styrene inMIL-47 at 433 K and snapshots of styrene and ethylbenzene at 1times106 Pa and1times109 Pa At 1times109 Pa both molecules have commensurate stacking but in orderfor ethylbenzene to have commensurate stacking the phenyl groups have to beslightly shifted due to the non-planar ethyl group This shift affects the amount ofethylbenzene molecules that can be adsorbed (as compared to styrene) and there-fore causes a difference in the saturation capacity of styrene and ethylbenzeneInterestingly commensurate stacking occurs at lower loading for styrene than forethylbenzene We can see in Figure 78 that at 1times106 Pa styrene already presentscommensurate stacking while ethylbenzene seems to have more of a face-to-facestacking In a mixture this will favor styrene adsorption even more In Figure79 simulated mixture adsorption isotherms for an equimolar mixture in MIL-47at 433 K are presented MIL-47 is a styrene selective structure Our results are ingood qualitative agreement with Maes et al [6] Breakthrough simulations furtherconfirm that MIL-47 is a styrene selective structure at saturation conditions InMAF-x8 a metal-organic framework with square channels of approximately 1nm

74 Results 153

BEFORE REORIENTATION

AFTER REORIENTATION

096 nm 095 nm

034 nm 053 nm

a b

095 nm

053 nm x2

c

Figure 76 Schematic representation of different entropic effects occurring in theseparation of styrene and ethylbenzene in nanoporous materials (a) The reductionof the moleculersquos footprint in the channels because of a reorientation into a face-to-face stacking configuration (b) Effect of the channel size on the reorientationand therefore moleculersquos footprint reduction (c) Comparison of the channel lengthneeded for two molecules of styrene to have commensurate stacking vs two moleculesof ethylbenzene to have face-to-face stacking

0

05

1

15

2

25

3

35

10-1

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

0

05

1

15

2

25

3

100

101

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

0

05

1

15

2

0 05 1 15 2 25 3 35 4 45

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

0

02

04

06

08

1

12

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

a

0

01

02

03

04

05

06

07

08

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

0

05

1

15

2

0 05 1 15 2 25 3

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

c

Figure 77 Simulated isotherms and breakthrough curves of styrene and ethylben-zene in MIL-53 (top) and DON (bottom) at 433 K (a) Pure component isotherms(lines are dual-site Langmuir-Freundlich fits of the pure components points are thepure component isotherms from CBCFCMC simulations) (b) Mixture componentisotherms for an equimolar mixture The IAST results are in good agreement with themixture isotherms (c) Simulated step-type breakthrough at 1e6 Pa total fugacity

154 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

0

1

2

3

4

5

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

Figure 78 Single component adsorption isotherms for styrene and ethylbenzeneat 433 K in MIL-47 Inset styrene (top) and ethylbenzene (bottom) snapshots at1times106 Pa and 1times109 Pa Styrene has commensurate stacking at lower pressuresthan ethylbenzene Because of the out of plane ethyl group ethylbenzene moleculeshave to be slightly shifted in order to have commensurate stacking and therefore lessethylbenzene molecules can be adsorbed in MIL-47 channels at saturation conditions

0

05

1

15

2

25

3

35

4

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

02

04

06

08

1

12

14

16

18

0 1 2 3 4 5

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

b

Figure 79 Styreneethylbenzene separation using MIL-47 at 433 K (a) equimolarmixture isotherms and Ideal Adsorption Solution Theory (IAST) prediction based onpure component isotherms (b) simulated step breakthrough at 1times106 Pa total fu-gacity The IAST prediction is in excellent agreement with the mixture simulationsThe mixture and breakthrough simulations show a high styrene selectivity and loadingin the mixture

it is easier for ethylbenzene to have commensurate stacking than in MIL-47 Thesingle component isotherms of both molecules behave very similar The topologyof the structure seems to induce a shift between parallel styrene molecules and al-lows for the ethyl group of ethylbenzene to stick in the channel ldquopocketsrdquo (Figure710) This might be the reason for the smaller difference in saturation capacitiesof styrene and ethylbenzene compared to MIL-47

74 Results 155

a b

0

05

1

15

2

25

3

35

4

10-6

10-4

10-2

100

102

104

106

108

Abso

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

c

Figure 710 Snapshots of (a) styrene and (b) ethylbenzene in MAF-X8 at 1e9 Paand 433 K Both styrene and ethylbenzene have commensurate stacking Color codecarbon (cyan) hydrogen (white) (c) Simulated single component isotherms of styreneand ethylbenzene in MAF-X8 at 433 K

Combinationcompetition of mechanisms is observed in structures with a more com-plex topology An example of this is MOF-CJ3 metal-organic framework [47] Thewide segment of the channels are big enough to allow for both styrene and ethyl-benzene to form two parallel rows (commensurate stacking) however the shapeof the channels forces ethylbenzene molecules to adopt a configuration where theethyl group is pointing to the channel ldquopocketsrdquo Styrene has more freedom andmolecules can arrange in a way that an extra styrene can be adsorbed in theprotracted segments in a configuration perpendicular to the channel (face-to-facestacking) (Figure 711) The interplay between different mechanisms rarely makesthe separation better Even when all the mechanisms favor a specific molecule thecompetition between them can induce enough disorder to destroy the selectivityone could achieve with a ldquopurerdquo mechanism

Figure 711 Snapshots of styrene (top) and ethylbenzene (bottom) at 1times109 Paand 433 K in MOF-CJ3

156 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

75 Discussion

There are two important factors to consider when using adsorption for separa-tion processes at industrial conditions namely selectivity and capacity A highselectivity ensures that less cycles are needed to achieve a high degree of purity inthe separation but a high capacity implies that the regeneration time is longerFor a binary mixture the adsorption selectivity is defined as

Sads =q1q2

f1f2(71)

and the capacity is defined as the styrene loading in the adsorbed phase of a binarymixture

Capacity = q1 (72)

In Figure 712 the relationship between these two properties for different struc-tures is presented Structures with the same separation mechanism are plottedwith the same color An ideal structure for the separation would be located at theright top corner

1

10

0 1 2 3 4 5 6

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

MAF-x8

MOF-CJ3

CoBDP IRMOF-1

JUC-77

CuBTC

UiO-66

Figure 712 Selectivity as a function of styrene loading (capacity) in a binary mix-ture at 433 K and 1times106 Pa total fugacity The structures are divided in differentcolors depending on the selectivity mechanism observed Color code Size exclusion(red) face-to-face stacking (purple) commensurate stacking (blue) commensurate-stackingface-to-face stacking (green) mixed(orange) The dotted red line corres-ponds to the ratio styreneethylbenzene at liquid conditions There is a naturaltrade-off between selectivity and loading finding structures in the top right corneris not feasible The black dashed line (guide to the eye) denotes the inverse relation-ship between selectivity and capacity MIL-47 is a styrene selective material with ahigh capacity therefore a good candidate for the styreneethylbenzene separation

75 Discussion 157

In structures where the separation is driven by size-exclusion (MFI-para) theselectivity is high because there is an adsorption site available only for styrene(zig-zag channels) but the capacity is rather low since for size-exclusion to occuramong similar molecules there has to be a very tight fit between the moleculesand the adsorbent usually associated with small pore systems

In structures where the separation is driven by face-to-face stacking the se-lectivity relies on the increase of the moleculersquos footprints difference consequenceof a reorientation and piling Ideally only one of the mixture components shouldbe able to reorient but in the case of styrene and ethylbenzene the similarity intheir length and width makes no significant difference in the pore size needed forthe reorientation The selectivity relies thus in how favorable the reorientationand piling are

In MAZ and AFI both molecules can reorient but the reorientation is restrictedby the pore size forcing a tilting in the face-to-face stacking configuration that isunfavorable for ethylbenzene to form a pile but it is favorable for styrene Thisallows styrene to obtain higher saturation loadings and ensures MAZ and AFIstyrene selectivity at saturation conditions

However face-to-face stacking only occurs at high loadings In Figure 712the selectivity is presented at 1times106 Pa At this fugacity MAZ is not yet styreneselective and AFI selectivity is almost negligible The effect of face-to-face stackingin styrene selectivity can only be observed at higher loadings as shown in Figure713 where the selectivity as a function of styrene loading for higher fugacities ispresented

1

10

1 10

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

CoBDP

CuBTC

IRMOF-1

JUC-77

MAF-X8

MIL-47

MIL-53

MOF-CJ3

AFI

DON

MAZ

MFI-para

UiO-66

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

JUC-77

MAF-x8

IRMOF-1

MOFCJ3

CoBDP

CuBTC

UiO-66

MIL-47

MIL-53

JUC-77

CuBTC

CoBDP

MOF-CJ3 IRMOF-1

MAZDON

MFI-para

AFI

UiO-66

MAF-X8

1e6 Pa1e7 Pa1e9 Pa

Figure 713 Selectivity (qstqeb) in an equimolar mixture at 3 different fugacitiesThe dashed red line corresponds to the liquid phase ratio For all the cases styreneloading increases with the fugacity (shift to the right of the plot) The selectivity alsoincreases in most of the cases with the fugacity (shift upwards) except for MFI-paraJUC-77 MIL-47 and Cu-BTC

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 10: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

74 Results 153

BEFORE REORIENTATION

AFTER REORIENTATION

096 nm 095 nm

034 nm 053 nm

a b

095 nm

053 nm x2

c

Figure 76 Schematic representation of different entropic effects occurring in theseparation of styrene and ethylbenzene in nanoporous materials (a) The reductionof the moleculersquos footprint in the channels because of a reorientation into a face-to-face stacking configuration (b) Effect of the channel size on the reorientationand therefore moleculersquos footprint reduction (c) Comparison of the channel lengthneeded for two molecules of styrene to have commensurate stacking vs two moleculesof ethylbenzene to have face-to-face stacking

0

05

1

15

2

25

3

35

10-1

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

0

05

1

15

2

25

3

100

101

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

0

05

1

15

2

0 05 1 15 2 25 3 35 4 45

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

0

02

04

06

08

1

12

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

a

0

01

02

03

04

05

06

07

08

102

103

104

105

106

107

108

109

Com

ponent lo

adin

g q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

b

0

05

1

15

2

0 05 1 15 2 25 3

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

c

Figure 77 Simulated isotherms and breakthrough curves of styrene and ethylben-zene in MIL-53 (top) and DON (bottom) at 433 K (a) Pure component isotherms(lines are dual-site Langmuir-Freundlich fits of the pure components points are thepure component isotherms from CBCFCMC simulations) (b) Mixture componentisotherms for an equimolar mixture The IAST results are in good agreement with themixture isotherms (c) Simulated step-type breakthrough at 1e6 Pa total fugacity

154 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

0

1

2

3

4

5

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

Figure 78 Single component adsorption isotherms for styrene and ethylbenzeneat 433 K in MIL-47 Inset styrene (top) and ethylbenzene (bottom) snapshots at1times106 Pa and 1times109 Pa Styrene has commensurate stacking at lower pressuresthan ethylbenzene Because of the out of plane ethyl group ethylbenzene moleculeshave to be slightly shifted in order to have commensurate stacking and therefore lessethylbenzene molecules can be adsorbed in MIL-47 channels at saturation conditions

0

05

1

15

2

25

3

35

4

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

02

04

06

08

1

12

14

16

18

0 1 2 3 4 5

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

b

Figure 79 Styreneethylbenzene separation using MIL-47 at 433 K (a) equimolarmixture isotherms and Ideal Adsorption Solution Theory (IAST) prediction based onpure component isotherms (b) simulated step breakthrough at 1times106 Pa total fu-gacity The IAST prediction is in excellent agreement with the mixture simulationsThe mixture and breakthrough simulations show a high styrene selectivity and loadingin the mixture

it is easier for ethylbenzene to have commensurate stacking than in MIL-47 Thesingle component isotherms of both molecules behave very similar The topologyof the structure seems to induce a shift between parallel styrene molecules and al-lows for the ethyl group of ethylbenzene to stick in the channel ldquopocketsrdquo (Figure710) This might be the reason for the smaller difference in saturation capacitiesof styrene and ethylbenzene compared to MIL-47

74 Results 155

a b

0

05

1

15

2

25

3

35

4

10-6

10-4

10-2

100

102

104

106

108

Abso

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

c

Figure 710 Snapshots of (a) styrene and (b) ethylbenzene in MAF-X8 at 1e9 Paand 433 K Both styrene and ethylbenzene have commensurate stacking Color codecarbon (cyan) hydrogen (white) (c) Simulated single component isotherms of styreneand ethylbenzene in MAF-X8 at 433 K

Combinationcompetition of mechanisms is observed in structures with a more com-plex topology An example of this is MOF-CJ3 metal-organic framework [47] Thewide segment of the channels are big enough to allow for both styrene and ethyl-benzene to form two parallel rows (commensurate stacking) however the shapeof the channels forces ethylbenzene molecules to adopt a configuration where theethyl group is pointing to the channel ldquopocketsrdquo Styrene has more freedom andmolecules can arrange in a way that an extra styrene can be adsorbed in theprotracted segments in a configuration perpendicular to the channel (face-to-facestacking) (Figure 711) The interplay between different mechanisms rarely makesthe separation better Even when all the mechanisms favor a specific molecule thecompetition between them can induce enough disorder to destroy the selectivityone could achieve with a ldquopurerdquo mechanism

Figure 711 Snapshots of styrene (top) and ethylbenzene (bottom) at 1times109 Paand 433 K in MOF-CJ3

156 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

75 Discussion

There are two important factors to consider when using adsorption for separa-tion processes at industrial conditions namely selectivity and capacity A highselectivity ensures that less cycles are needed to achieve a high degree of purity inthe separation but a high capacity implies that the regeneration time is longerFor a binary mixture the adsorption selectivity is defined as

Sads =q1q2

f1f2(71)

and the capacity is defined as the styrene loading in the adsorbed phase of a binarymixture

Capacity = q1 (72)

In Figure 712 the relationship between these two properties for different struc-tures is presented Structures with the same separation mechanism are plottedwith the same color An ideal structure for the separation would be located at theright top corner

1

10

0 1 2 3 4 5 6

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

MAF-x8

MOF-CJ3

CoBDP IRMOF-1

JUC-77

CuBTC

UiO-66

Figure 712 Selectivity as a function of styrene loading (capacity) in a binary mix-ture at 433 K and 1times106 Pa total fugacity The structures are divided in differentcolors depending on the selectivity mechanism observed Color code Size exclusion(red) face-to-face stacking (purple) commensurate stacking (blue) commensurate-stackingface-to-face stacking (green) mixed(orange) The dotted red line corres-ponds to the ratio styreneethylbenzene at liquid conditions There is a naturaltrade-off between selectivity and loading finding structures in the top right corneris not feasible The black dashed line (guide to the eye) denotes the inverse relation-ship between selectivity and capacity MIL-47 is a styrene selective material with ahigh capacity therefore a good candidate for the styreneethylbenzene separation

75 Discussion 157

In structures where the separation is driven by size-exclusion (MFI-para) theselectivity is high because there is an adsorption site available only for styrene(zig-zag channels) but the capacity is rather low since for size-exclusion to occuramong similar molecules there has to be a very tight fit between the moleculesand the adsorbent usually associated with small pore systems

In structures where the separation is driven by face-to-face stacking the se-lectivity relies on the increase of the moleculersquos footprints difference consequenceof a reorientation and piling Ideally only one of the mixture components shouldbe able to reorient but in the case of styrene and ethylbenzene the similarity intheir length and width makes no significant difference in the pore size needed forthe reorientation The selectivity relies thus in how favorable the reorientationand piling are

In MAZ and AFI both molecules can reorient but the reorientation is restrictedby the pore size forcing a tilting in the face-to-face stacking configuration that isunfavorable for ethylbenzene to form a pile but it is favorable for styrene Thisallows styrene to obtain higher saturation loadings and ensures MAZ and AFIstyrene selectivity at saturation conditions

However face-to-face stacking only occurs at high loadings In Figure 712the selectivity is presented at 1times106 Pa At this fugacity MAZ is not yet styreneselective and AFI selectivity is almost negligible The effect of face-to-face stackingin styrene selectivity can only be observed at higher loadings as shown in Figure713 where the selectivity as a function of styrene loading for higher fugacities ispresented

1

10

1 10

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

CoBDP

CuBTC

IRMOF-1

JUC-77

MAF-X8

MIL-47

MIL-53

MOF-CJ3

AFI

DON

MAZ

MFI-para

UiO-66

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

JUC-77

MAF-x8

IRMOF-1

MOFCJ3

CoBDP

CuBTC

UiO-66

MIL-47

MIL-53

JUC-77

CuBTC

CoBDP

MOF-CJ3 IRMOF-1

MAZDON

MFI-para

AFI

UiO-66

MAF-X8

1e6 Pa1e7 Pa1e9 Pa

Figure 713 Selectivity (qstqeb) in an equimolar mixture at 3 different fugacitiesThe dashed red line corresponds to the liquid phase ratio For all the cases styreneloading increases with the fugacity (shift to the right of the plot) The selectivity alsoincreases in most of the cases with the fugacity (shift upwards) except for MFI-paraJUC-77 MIL-47 and Cu-BTC

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 11: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

154 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

0

1

2

3

4

5

100

101

102

103

104

105

106

107

108

109

Absolu

te loadin

g q m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

Figure 78 Single component adsorption isotherms for styrene and ethylbenzeneat 433 K in MIL-47 Inset styrene (top) and ethylbenzene (bottom) snapshots at1times106 Pa and 1times109 Pa Styrene has commensurate stacking at lower pressuresthan ethylbenzene Because of the out of plane ethyl group ethylbenzene moleculeshave to be slightly shifted in order to have commensurate stacking and therefore lessethylbenzene molecules can be adsorbed in MIL-47 channels at saturation conditions

0

05

1

15

2

25

3

35

4

100

101

102

103

104

105

106

107

108

109

Co

mp

on

en

t lo

ad

ing

q

i m

ol kg

-1

Total bulk fluid phase fugacity ft Pa

ethylbenzenestyreneIAST

a

0

02

04

06

08

1

12

14

16

18

0 1 2 3 4 5

Dim

en

sio

nle

ss c

on

ce

ntr

atio

n in

exit g

as

cic

i0

Dimensionless time τ=tuεL

ethylbenzenestyrene

b

Figure 79 Styreneethylbenzene separation using MIL-47 at 433 K (a) equimolarmixture isotherms and Ideal Adsorption Solution Theory (IAST) prediction based onpure component isotherms (b) simulated step breakthrough at 1times106 Pa total fu-gacity The IAST prediction is in excellent agreement with the mixture simulationsThe mixture and breakthrough simulations show a high styrene selectivity and loadingin the mixture

it is easier for ethylbenzene to have commensurate stacking than in MIL-47 Thesingle component isotherms of both molecules behave very similar The topologyof the structure seems to induce a shift between parallel styrene molecules and al-lows for the ethyl group of ethylbenzene to stick in the channel ldquopocketsrdquo (Figure710) This might be the reason for the smaller difference in saturation capacitiesof styrene and ethylbenzene compared to MIL-47

74 Results 155

a b

0

05

1

15

2

25

3

35

4

10-6

10-4

10-2

100

102

104

106

108

Abso

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

c

Figure 710 Snapshots of (a) styrene and (b) ethylbenzene in MAF-X8 at 1e9 Paand 433 K Both styrene and ethylbenzene have commensurate stacking Color codecarbon (cyan) hydrogen (white) (c) Simulated single component isotherms of styreneand ethylbenzene in MAF-X8 at 433 K

Combinationcompetition of mechanisms is observed in structures with a more com-plex topology An example of this is MOF-CJ3 metal-organic framework [47] Thewide segment of the channels are big enough to allow for both styrene and ethyl-benzene to form two parallel rows (commensurate stacking) however the shapeof the channels forces ethylbenzene molecules to adopt a configuration where theethyl group is pointing to the channel ldquopocketsrdquo Styrene has more freedom andmolecules can arrange in a way that an extra styrene can be adsorbed in theprotracted segments in a configuration perpendicular to the channel (face-to-facestacking) (Figure 711) The interplay between different mechanisms rarely makesthe separation better Even when all the mechanisms favor a specific molecule thecompetition between them can induce enough disorder to destroy the selectivityone could achieve with a ldquopurerdquo mechanism

Figure 711 Snapshots of styrene (top) and ethylbenzene (bottom) at 1times109 Paand 433 K in MOF-CJ3

156 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

75 Discussion

There are two important factors to consider when using adsorption for separa-tion processes at industrial conditions namely selectivity and capacity A highselectivity ensures that less cycles are needed to achieve a high degree of purity inthe separation but a high capacity implies that the regeneration time is longerFor a binary mixture the adsorption selectivity is defined as

Sads =q1q2

f1f2(71)

and the capacity is defined as the styrene loading in the adsorbed phase of a binarymixture

Capacity = q1 (72)

In Figure 712 the relationship between these two properties for different struc-tures is presented Structures with the same separation mechanism are plottedwith the same color An ideal structure for the separation would be located at theright top corner

1

10

0 1 2 3 4 5 6

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

MAF-x8

MOF-CJ3

CoBDP IRMOF-1

JUC-77

CuBTC

UiO-66

Figure 712 Selectivity as a function of styrene loading (capacity) in a binary mix-ture at 433 K and 1times106 Pa total fugacity The structures are divided in differentcolors depending on the selectivity mechanism observed Color code Size exclusion(red) face-to-face stacking (purple) commensurate stacking (blue) commensurate-stackingface-to-face stacking (green) mixed(orange) The dotted red line corres-ponds to the ratio styreneethylbenzene at liquid conditions There is a naturaltrade-off between selectivity and loading finding structures in the top right corneris not feasible The black dashed line (guide to the eye) denotes the inverse relation-ship between selectivity and capacity MIL-47 is a styrene selective material with ahigh capacity therefore a good candidate for the styreneethylbenzene separation

75 Discussion 157

In structures where the separation is driven by size-exclusion (MFI-para) theselectivity is high because there is an adsorption site available only for styrene(zig-zag channels) but the capacity is rather low since for size-exclusion to occuramong similar molecules there has to be a very tight fit between the moleculesand the adsorbent usually associated with small pore systems

In structures where the separation is driven by face-to-face stacking the se-lectivity relies on the increase of the moleculersquos footprints difference consequenceof a reorientation and piling Ideally only one of the mixture components shouldbe able to reorient but in the case of styrene and ethylbenzene the similarity intheir length and width makes no significant difference in the pore size needed forthe reorientation The selectivity relies thus in how favorable the reorientationand piling are

In MAZ and AFI both molecules can reorient but the reorientation is restrictedby the pore size forcing a tilting in the face-to-face stacking configuration that isunfavorable for ethylbenzene to form a pile but it is favorable for styrene Thisallows styrene to obtain higher saturation loadings and ensures MAZ and AFIstyrene selectivity at saturation conditions

However face-to-face stacking only occurs at high loadings In Figure 712the selectivity is presented at 1times106 Pa At this fugacity MAZ is not yet styreneselective and AFI selectivity is almost negligible The effect of face-to-face stackingin styrene selectivity can only be observed at higher loadings as shown in Figure713 where the selectivity as a function of styrene loading for higher fugacities ispresented

1

10

1 10

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

CoBDP

CuBTC

IRMOF-1

JUC-77

MAF-X8

MIL-47

MIL-53

MOF-CJ3

AFI

DON

MAZ

MFI-para

UiO-66

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

JUC-77

MAF-x8

IRMOF-1

MOFCJ3

CoBDP

CuBTC

UiO-66

MIL-47

MIL-53

JUC-77

CuBTC

CoBDP

MOF-CJ3 IRMOF-1

MAZDON

MFI-para

AFI

UiO-66

MAF-X8

1e6 Pa1e7 Pa1e9 Pa

Figure 713 Selectivity (qstqeb) in an equimolar mixture at 3 different fugacitiesThe dashed red line corresponds to the liquid phase ratio For all the cases styreneloading increases with the fugacity (shift to the right of the plot) The selectivity alsoincreases in most of the cases with the fugacity (shift upwards) except for MFI-paraJUC-77 MIL-47 and Cu-BTC

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 12: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

74 Results 155

a b

0

05

1

15

2

25

3

35

4

10-6

10-4

10-2

100

102

104

106

108

Abso

lute

lo

ad

ing

q

m

ol kg

-1

Bulk fluid phase fugacity f Pa

ethylbenzenestyrene

c

Figure 710 Snapshots of (a) styrene and (b) ethylbenzene in MAF-X8 at 1e9 Paand 433 K Both styrene and ethylbenzene have commensurate stacking Color codecarbon (cyan) hydrogen (white) (c) Simulated single component isotherms of styreneand ethylbenzene in MAF-X8 at 433 K

Combinationcompetition of mechanisms is observed in structures with a more com-plex topology An example of this is MOF-CJ3 metal-organic framework [47] Thewide segment of the channels are big enough to allow for both styrene and ethyl-benzene to form two parallel rows (commensurate stacking) however the shapeof the channels forces ethylbenzene molecules to adopt a configuration where theethyl group is pointing to the channel ldquopocketsrdquo Styrene has more freedom andmolecules can arrange in a way that an extra styrene can be adsorbed in theprotracted segments in a configuration perpendicular to the channel (face-to-facestacking) (Figure 711) The interplay between different mechanisms rarely makesthe separation better Even when all the mechanisms favor a specific molecule thecompetition between them can induce enough disorder to destroy the selectivityone could achieve with a ldquopurerdquo mechanism

Figure 711 Snapshots of styrene (top) and ethylbenzene (bottom) at 1times109 Paand 433 K in MOF-CJ3

156 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

75 Discussion

There are two important factors to consider when using adsorption for separa-tion processes at industrial conditions namely selectivity and capacity A highselectivity ensures that less cycles are needed to achieve a high degree of purity inthe separation but a high capacity implies that the regeneration time is longerFor a binary mixture the adsorption selectivity is defined as

Sads =q1q2

f1f2(71)

and the capacity is defined as the styrene loading in the adsorbed phase of a binarymixture

Capacity = q1 (72)

In Figure 712 the relationship between these two properties for different struc-tures is presented Structures with the same separation mechanism are plottedwith the same color An ideal structure for the separation would be located at theright top corner

1

10

0 1 2 3 4 5 6

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

MAF-x8

MOF-CJ3

CoBDP IRMOF-1

JUC-77

CuBTC

UiO-66

Figure 712 Selectivity as a function of styrene loading (capacity) in a binary mix-ture at 433 K and 1times106 Pa total fugacity The structures are divided in differentcolors depending on the selectivity mechanism observed Color code Size exclusion(red) face-to-face stacking (purple) commensurate stacking (blue) commensurate-stackingface-to-face stacking (green) mixed(orange) The dotted red line corres-ponds to the ratio styreneethylbenzene at liquid conditions There is a naturaltrade-off between selectivity and loading finding structures in the top right corneris not feasible The black dashed line (guide to the eye) denotes the inverse relation-ship between selectivity and capacity MIL-47 is a styrene selective material with ahigh capacity therefore a good candidate for the styreneethylbenzene separation

75 Discussion 157

In structures where the separation is driven by size-exclusion (MFI-para) theselectivity is high because there is an adsorption site available only for styrene(zig-zag channels) but the capacity is rather low since for size-exclusion to occuramong similar molecules there has to be a very tight fit between the moleculesand the adsorbent usually associated with small pore systems

In structures where the separation is driven by face-to-face stacking the se-lectivity relies on the increase of the moleculersquos footprints difference consequenceof a reorientation and piling Ideally only one of the mixture components shouldbe able to reorient but in the case of styrene and ethylbenzene the similarity intheir length and width makes no significant difference in the pore size needed forthe reorientation The selectivity relies thus in how favorable the reorientationand piling are

In MAZ and AFI both molecules can reorient but the reorientation is restrictedby the pore size forcing a tilting in the face-to-face stacking configuration that isunfavorable for ethylbenzene to form a pile but it is favorable for styrene Thisallows styrene to obtain higher saturation loadings and ensures MAZ and AFIstyrene selectivity at saturation conditions

However face-to-face stacking only occurs at high loadings In Figure 712the selectivity is presented at 1times106 Pa At this fugacity MAZ is not yet styreneselective and AFI selectivity is almost negligible The effect of face-to-face stackingin styrene selectivity can only be observed at higher loadings as shown in Figure713 where the selectivity as a function of styrene loading for higher fugacities ispresented

1

10

1 10

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

CoBDP

CuBTC

IRMOF-1

JUC-77

MAF-X8

MIL-47

MIL-53

MOF-CJ3

AFI

DON

MAZ

MFI-para

UiO-66

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

JUC-77

MAF-x8

IRMOF-1

MOFCJ3

CoBDP

CuBTC

UiO-66

MIL-47

MIL-53

JUC-77

CuBTC

CoBDP

MOF-CJ3 IRMOF-1

MAZDON

MFI-para

AFI

UiO-66

MAF-X8

1e6 Pa1e7 Pa1e9 Pa

Figure 713 Selectivity (qstqeb) in an equimolar mixture at 3 different fugacitiesThe dashed red line corresponds to the liquid phase ratio For all the cases styreneloading increases with the fugacity (shift to the right of the plot) The selectivity alsoincreases in most of the cases with the fugacity (shift upwards) except for MFI-paraJUC-77 MIL-47 and Cu-BTC

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 13: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

156 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

75 Discussion

There are two important factors to consider when using adsorption for separa-tion processes at industrial conditions namely selectivity and capacity A highselectivity ensures that less cycles are needed to achieve a high degree of purity inthe separation but a high capacity implies that the regeneration time is longerFor a binary mixture the adsorption selectivity is defined as

Sads =q1q2

f1f2(71)

and the capacity is defined as the styrene loading in the adsorbed phase of a binarymixture

Capacity = q1 (72)

In Figure 712 the relationship between these two properties for different struc-tures is presented Structures with the same separation mechanism are plottedwith the same color An ideal structure for the separation would be located at theright top corner

1

10

0 1 2 3 4 5 6

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

MAF-x8

MOF-CJ3

CoBDP IRMOF-1

JUC-77

CuBTC

UiO-66

Figure 712 Selectivity as a function of styrene loading (capacity) in a binary mix-ture at 433 K and 1times106 Pa total fugacity The structures are divided in differentcolors depending on the selectivity mechanism observed Color code Size exclusion(red) face-to-face stacking (purple) commensurate stacking (blue) commensurate-stackingface-to-face stacking (green) mixed(orange) The dotted red line corres-ponds to the ratio styreneethylbenzene at liquid conditions There is a naturaltrade-off between selectivity and loading finding structures in the top right corneris not feasible The black dashed line (guide to the eye) denotes the inverse relation-ship between selectivity and capacity MIL-47 is a styrene selective material with ahigh capacity therefore a good candidate for the styreneethylbenzene separation

75 Discussion 157

In structures where the separation is driven by size-exclusion (MFI-para) theselectivity is high because there is an adsorption site available only for styrene(zig-zag channels) but the capacity is rather low since for size-exclusion to occuramong similar molecules there has to be a very tight fit between the moleculesand the adsorbent usually associated with small pore systems

In structures where the separation is driven by face-to-face stacking the se-lectivity relies on the increase of the moleculersquos footprints difference consequenceof a reorientation and piling Ideally only one of the mixture components shouldbe able to reorient but in the case of styrene and ethylbenzene the similarity intheir length and width makes no significant difference in the pore size needed forthe reorientation The selectivity relies thus in how favorable the reorientationand piling are

In MAZ and AFI both molecules can reorient but the reorientation is restrictedby the pore size forcing a tilting in the face-to-face stacking configuration that isunfavorable for ethylbenzene to form a pile but it is favorable for styrene Thisallows styrene to obtain higher saturation loadings and ensures MAZ and AFIstyrene selectivity at saturation conditions

However face-to-face stacking only occurs at high loadings In Figure 712the selectivity is presented at 1times106 Pa At this fugacity MAZ is not yet styreneselective and AFI selectivity is almost negligible The effect of face-to-face stackingin styrene selectivity can only be observed at higher loadings as shown in Figure713 where the selectivity as a function of styrene loading for higher fugacities ispresented

1

10

1 10

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

CoBDP

CuBTC

IRMOF-1

JUC-77

MAF-X8

MIL-47

MIL-53

MOF-CJ3

AFI

DON

MAZ

MFI-para

UiO-66

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

JUC-77

MAF-x8

IRMOF-1

MOFCJ3

CoBDP

CuBTC

UiO-66

MIL-47

MIL-53

JUC-77

CuBTC

CoBDP

MOF-CJ3 IRMOF-1

MAZDON

MFI-para

AFI

UiO-66

MAF-X8

1e6 Pa1e7 Pa1e9 Pa

Figure 713 Selectivity (qstqeb) in an equimolar mixture at 3 different fugacitiesThe dashed red line corresponds to the liquid phase ratio For all the cases styreneloading increases with the fugacity (shift to the right of the plot) The selectivity alsoincreases in most of the cases with the fugacity (shift upwards) except for MFI-paraJUC-77 MIL-47 and Cu-BTC

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 14: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

75 Discussion 157

In structures where the separation is driven by size-exclusion (MFI-para) theselectivity is high because there is an adsorption site available only for styrene(zig-zag channels) but the capacity is rather low since for size-exclusion to occuramong similar molecules there has to be a very tight fit between the moleculesand the adsorbent usually associated with small pore systems

In structures where the separation is driven by face-to-face stacking the se-lectivity relies on the increase of the moleculersquos footprints difference consequenceof a reorientation and piling Ideally only one of the mixture components shouldbe able to reorient but in the case of styrene and ethylbenzene the similarity intheir length and width makes no significant difference in the pore size needed forthe reorientation The selectivity relies thus in how favorable the reorientationand piling are

In MAZ and AFI both molecules can reorient but the reorientation is restrictedby the pore size forcing a tilting in the face-to-face stacking configuration that isunfavorable for ethylbenzene to form a pile but it is favorable for styrene Thisallows styrene to obtain higher saturation loadings and ensures MAZ and AFIstyrene selectivity at saturation conditions

However face-to-face stacking only occurs at high loadings In Figure 712the selectivity is presented at 1times106 Pa At this fugacity MAZ is not yet styreneselective and AFI selectivity is almost negligible The effect of face-to-face stackingin styrene selectivity can only be observed at higher loadings as shown in Figure713 where the selectivity as a function of styrene loading for higher fugacities ispresented

1

10

1 10

Ad

so

rptio

n s

ele

ctivity

Sads

-

styrene loading mol kg-1

CoBDP

CuBTC

IRMOF-1

JUC-77

MAF-X8

MIL-47

MIL-53

MOF-CJ3

AFI

DON

MAZ

MFI-para

UiO-66

AFIMAZ

DON

MFI-para

MIL-47

MIL-53

JUC-77

MAF-x8

IRMOF-1

MOFCJ3

CoBDP

CuBTC

UiO-66

MIL-47

MIL-53

JUC-77

CuBTC

CoBDP

MOF-CJ3 IRMOF-1

MAZDON

MFI-para

AFI

UiO-66

MAF-X8

1e6 Pa1e7 Pa1e9 Pa

Figure 713 Selectivity (qstqeb) in an equimolar mixture at 3 different fugacitiesThe dashed red line corresponds to the liquid phase ratio For all the cases styreneloading increases with the fugacity (shift to the right of the plot) The selectivity alsoincreases in most of the cases with the fugacity (shift upwards) except for MFI-paraJUC-77 MIL-47 and Cu-BTC

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 15: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

158 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

Face-to-face stacking can only occur when the reoriented molecules are com-mensurate with the channel diameter This has an important restriction in thepore size and therefore in the capacity

In structures where styrene has commensurate stacking an ethylbenzene hasface-to-face stacking (as observed in MIL-53 and DON) the difference in saturationcapacities arises because the channel length needed to accommodate moleculesof styrene in commensurate stacking is smaller than the channel length neededto accommodate molecules of ethylbenzene in face-to-face stacking When anequimolar mixture is considered styrene will be favored even more because it doesnot have to undergo any reorientation to achieve higher loadings Structures withpore sizes that allow styrene to have commensurate stacking but not ethylbenzneare styrene selective and have a higher saturation capacity than structures thatpresent face-to-face stacking or size exclusion

In structures where both styrene and ethylbenzene can have commensuratestacking (MIL-47 and MAF-X8) the selectivity will depend on the efficiency inwhich the molecules can stack In Figure 714 we present schematic commen-surate stacking configurations of styrene and ethylbenzene For ethylbenzene tohave commensurate stacking the ldquominimal lengthrdquo the ldquopacking lengthrdquo or bothhave to be larger than for styrene At saturation conditions this implies that morestyrene molecules can be adsorbed than ethylbenzene ones favoring the adsorptionof styrene over ethylbenzene in a mixture Commensurate stacking enhances thedimensional differences of styrene and ethylbenzene as ldquopairsrdquo of molecules Be-cause commensurate stacking occurs in structures with larger pores high capacitycan be attained

minim

al length

packing length

Figure 714 Schematic differences of the channel dimensions needed for commensur-ate stacking Because styrene is a planar molecule commensurate stacking can occurin smaller channels Commensurate stacking is a powerful separation mechanism forplanarnon-planar molecules

Commensurate stacking is the best mechanism for the separation of ethylben-zene and styrene It offers a geometrical solution to the separation problem thatensures a high selectivity and it occurs in open pore structures ensuring a high

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 16: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

76 Conclusion 159

capacity The channel topology can facilitate or impede the selectivityIn the case of MIL-47 the almost planar walls force ethylbenzene molecules to

be shifted increasing the difference in ldquolengthrdquo per pair of molecules in the channelas compared to styrene This makes MIL-47 a highly styrene selective material andthe best candidate for the separation of styreneethylbenzene from the structureswe screened

76 Conclusion

Styrene and ethylbenzene are very similar molecules and finding structures thatcan discriminate between them is not easy At liquid conditions the success in theseparation process is strongly related with difference in saturation capacities of themixture components which in turn is strongly dictated by the underlying entropicmechanisms occurring in the nanoporous material Commensurate stacking offersthe best trade-off between saturation capacity and selectivity and is therefore avery efficient mechanism for the separation of styrene and ethylbenzene Amongthe different structures we studied MIL-47 a styrene selective structure is thebest candidate for the adsorptive separation of styreneethylbenzene mixture innanoporous materials Commensurate stacking offers a geometrical solution to theseparation of planarnon-planar molecules this enables a convenient approach todesigning materials for the separation

Acknowledgments

This material is supported by the Netherlands Research Council for Chemical Sci-ences (NWO-CW) also through a VIDI grant (David Dubbeldam) and by theStichting Nationale Computerfa- ciliteiten (National Computing Facilities Found-ation NCF) for the use of supercomputing facilities

Associated Content

This material provides (a) structural details of various materials investigated (b)pure component and mixture adsorption data (c) dual-Langmuir-Freundlich fitsparameters for unary isotherms (d) transient breakthrough simulation results forvarious materials httppubsacsorgdoiabs101021acslangmuir5b00363

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 17: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

160 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

References

[1] J R Wunsch Polystyrene Sythesis Production and Applications (RapraTechnology Ltd 2000)

[2] J C Gentry S Kumar and R Wright-Wytcherley Hydrocarb Process 9362 (2004)

[3] G A Randall Method of separating ethylbenzene from styrene by low pressuredrop distillation httpwwwgooglecompatentsUS3084108 (1963) uSPatent 3084108

[4] H M Van Tassell Separation of ethylbenzene and styrene by low pressurehigh temperature distillation httpwwwgooglecompatentsUS3398063(1968) uS Patent 3398063

[5] R Ahmad A G Wong-Foy and A J Matzger Langmuir 25 11977 (2009)

[6] M Maes L Alaerts F Vermoortele R Ameloot S Couck V Finsy J F MDenayer and D E De Vos J Am Chem Soc 132 2284 (2010)

[7] M Maes F Vermoortele L Alaerts S Couck C E A Kirschhock J F MDenayer and D E De Vos J Am Chem Soc 132 15277 (2010)

[8] T Remy L Ma M Maes D E D Vos G V Baron and J F M DenayerInd Eng Chem Res 5 14824 (2012)

[9] C-X Yang and X-P Yan Anal Chem pp 7144ndash7150 (2011)

[10] B Smit and T Maesen Nature 374 42 (1994)

[11] J Talbot AIChE J 43 2471 (1997)

[12] Z Du G Manos T J H Vlugt and B Smit AIChE J 44 1756 (1998)

[13] J M van Baten and R Krishna Microporous Mesoporous Mater 84 179(2005)

[14] R Krishna B Smit and S Calero Chem Soc Rev 31 185 (2002)

[15] R Krishna and J M van Baten Phys Chem Chem Phys 13 10593 (2011)

[16] A Torres-Knoop R Krishna and D Dubbeldam Angew Chem Int 537774 (2014)

[17] A Torres-Knoop S R G Balestra R Krishna S Calero and D Dub-beldam ChemPhysChem p accepted (2015)

[18] R Krishna Phys Chem Chem Phys 17 39 (2015)

[19] W L Jorgensen D S Maxwell and J Tirado-Rives J Am Chem Soc118 11225 (1996)

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 18: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

References 161

[20] R H Rohrbaugh and P C Jurs Anal Chim Acta 199 99 (1987)

[21] Accelrys Materials studioaccelrys ccopy2001-2007 accelrys software inc httpaccelryscomproductsmaterials-studioindexhtml

[22] S W I Siu K Pluhackova and R A Bockmann J Chem Theory Comput8 1459 (2012)

[23] G Kresse and J Hafner Phys Rev B 47 558 (1993)

[24] G Kresse and J Furthmuller Phys Rev B 54 11169 (1996)

[25] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 77 3865 (1996)

[26] J P Perdew K Burke and M Ernzerhof Phys Rev Lett 78 1396 (1997)

[27] S Grimme J Comp Chem 27 1787 (2006)

[28] S L Mayo B D Olafson and W A Goddard J Phys Chem 94 8897(1990)

[29] A K Rappe C J Casewit K S Colwell W A Goddard and W M JSkiff J Am Chem Soc 114 10024 (1992)

[30] P Bai M Tsapatsis and J I Siepmann J Phys Chem C 117 24375 (2013)

[31] C Campana B Mussard and T K Woo J Chem Theory Comput 5 2866(2009)

[32] T Watanabe T A Manz and D S Sholl J Phys Chem C 115 4824(2011)

[33] A Torres-Knoop S P Balaji T Vlugt and D Dubbeldam J Chem TheoryComput 10 942 (2014)

[34] J I Siepmann Mol Phys 70 1145 (1990)

[35] J I Siepmann and D Frenkel Mol Phys 75 59 (1992)

[36] D Frenkel G C A M Mooij and B Smit J Phys Condens Matter 43053 (1992)

[37] W Shi and E J Maginn J Chem Theory Comput 3 1451 (2007)

[38] R Krishna Microporous Mesoporous Mater 185 30 (2014)

[39] R Krishna and J R Long J Phys Chem C 115 12941 (2011)

[40] H van Koningsveld H v B F Tuinstra and J C Jansen Acta Cryst B45423 (1989)

[41] E Galli Rend Soc Ital Mineral Petrol 31 599 (1975)

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)

Page 19: UvA-DARE (Digital Academic Repository) Entropy driven ...conventional method involves two steps: the alkylation of benzene with ethylene to produce ethylbenzene and the dehydrogenation

162 Chapter 7 Entropic Separation of StyreneEthylbenzene Mixtures

[42] S Qiu W Pang H Kessler and J L Guth Zeolites 9 440 (1989)

[43] T Wessels C Baerlocher L B McCusker and E J Creyghton J AmChem Soc 121 6242 (1999)

[44] C Serre F Millange C Thouvenot M Nogues G Marsolier D Louer andG Ferey J Am Chem Soc 124 13519 (2002)

[45] K Barthelet J Marrot D Riou and G Ferey Ang Chem Int Ed 41 281(2002)

[46] C T He J Y Tian S Y Liu G Ouyang J P Zhang and X M ChenChem Sci 4 351 (2013)

[47] Z L Fang S R Zheng J B Tan S L Cai J Fan X Yan and W GZhang Journal of Chromatography A pp 132ndash8 (2013)


Recommended