Research Article
Hybrid polyacrylamide chiral stationaryphases for HPLC prepared by surface-initiated photopolymerization
Two hybrid polyacrylamide chiral stationary phases (CSPs) for HPLC have been synthe-
sized by a new surface-initiated photo-induced radical polymerization approach of enan-
tiopure N,N0-diacryloyl derivatives of (1R,2R)-diaminocyclohexane (CSP1) and (1R,2R)-
diphenylethylenediamine (CSP2). This system is based on the activation of mesoporous
silica microparticles by chemically bonded trichloroacetyl groups and dimanganese
decacarbonyl as catalyst. UV irradiation was performed using a lab-made quartz photo-
chemical reactor, ad hoc designed for the photo-induced polymerization process on the
surface of microparticles. The two phases were evaluated and compared as chromato-
graphic supports for the enantioselective HPLC of model chiral compounds. Their
physico-chemical properties and chromatographic performances were also evaluated in
comparison with those exhibited by the homologue CSPs obtained by the grafting-fromthermal-induced process (CSP3 and CSP4). The new photopolymerization approach
yielded higher grafting density than the thermal-induced one, especially in the case of the
less reactive monomer (the diacryloyl derivative of (1R,2R)-diphenylethylenediamine),
good chromatographic efficiency and a broad application field under normal phase and
polar organic mode conditions.
Keywords: Chiral polymers / Chiral stationary phases / HPLC / Photopoly-merizationDOI 10.1002/jssc.201000355
1 Introduction
Investigation on chiral stationary phases (CSPs) based on
totally synthetic, optically active polymers linked to a
chromatographic matrix is continuously evolving. Since
their introduction in 1980, polyacrylamide and polymeth-
acrylamide CSPs were used for the separation of a wide
range of chiral compounds, including benzodiazepines,
barbiturates, and hydantoins [1, 2]. Diverse strategies are
indeed available for the grafting of polymers onto the
surface of ultra-fine particles for non-chromatographic
applications [3], based on surface-initiated polymerization
(also called grafting-from or g-from approach) from initiators
bound to surfaces. Uniform and well permeable polymeric
layers are available only if the degree of polymerization and
the grafting density are carefully controlled. Such polymeric
materials are therefore highly desirable for the preparation
of improved polymeric CSPs. We recently reported [4–6] a
new hybrid organic/inorganic CSP for HPLC synthesized by
the g-from radical polymerization of an enantiopure
diacryloyl derivative of trans-1,2-diaminocyclohexane in the
presence of mesoporous, azo-activated silica micro-particles.
This was the first example of application of the g-fromapproach to the synthesis of a CSP for HPLC applications.
However, the thermal grafting process could not be surface-
confined only, since concomitant solution polymerization
may lead to ungrafted polymer chains [3]. Thus, we moved
to study another way to induce polymerization of the above
monomer, and focused our attention on photo-induced
polymerization, which has been obtaining much attention
in recent years because of its numerous industrial applica-
tions [7–9]. The concept of photopolymerization is that the
initiators generate free radicals upon light irradiation, and
the resulting radical starts the polymerization process.
Recently, several photoinitiators in free radical promoted
cationic polymerization have successfully been used [10]. In
particular, several visible light-absorbing systems able to
generate oxidable radicals were reported. For instance,
radicals formed by the irradiation of systems containing a
xanthene dye and an aromatic amine, were oxidized by a
Alessia Ciogli1
Ilaria D’Acquarica1
Francesco Gasparrini1
Carmela Molinaro1
Romina Rompietti1
Patrizia Simone1
Claudio Villani1
Giovanni Zappia2
1Dipartimento di Chimica eTecnologie del Farmaco,Sapienza Universita di Roma,Roma, Italy
2Istituto di Chimica Farmaceutica,Universita di Urbino ‘‘Carlo Bo’’,Urbino, Italy
Received May 18, 2010Revised July 22, 2010Accepted July 22, 2010
Abbreviations: 3-APSG, 3-aminopropyl silica gel; CSP, chiralstationary phase; DACH, (1R,2R)-diaminocyclohexane;
DPEDA, (1R,2R)-diphenylethylenediamine; DRIFT, diffusereflectance infrared Fourier transform; NP, normal phase;
POM, polar organic mode
Correspondence: Professor Francesco Gasparrini, Dipartimentodi Chimica e Tecnologie del Farmaco, Sapienza Universita diRoma, P.le A. Moro 5, 00185 Roma, ItalyE-mail: [email protected]: 1390649912780
& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
J. Sep. Sci. 2010, 33, 3022–30323022
diphenyliodonium salt [11]. Similarly, the dimanganese
decacarbonyl Mn2(CO)10-organic halides combination is an
efficient co-initiator for visible light cationic polymerization,
when used in conjunction with onium salts [12, 13].
Polymerization of vinyl monomers, such as methyl meth-
acrylate, styrene, and glycidyl methacrylate, was successfully
initiated by the system consisting of molibdenum hexa-
carbonyl Mo(CO)6 and surface-linked trichloroacetyl groups
as well [13, 14]. A possible mechanism of photoinitiation
involves a preliminary interaction between the metal
carbonyl and the trichloroacetyl functionalities, yielding
initiating alkyl radicals, that later react with the monomer
[15, 16].
In this work we describe for the first time the pre-
paration of two hybrid polyacrylamide CSPs by the
surface-initiated photo-induced radical polymerization of
enantiopure N,N0-diacryloyl derivatives of (1R,2R)-diamino-
cyclohexane (DACH) and (1R,2R)-diphenylethylene-
diamine (DPEDA), initiated by the system consisting
of trichloroacetyl groups on mesoporous silica particles
and dimanganese decacarbonyl Mn2(CO)10 under UV
irradiation.
2 Materials and methods
2.1 Chemicals and reagents
Spherical Daisogel SP-300-5P (5 mm particle size, 115 m2/g
specific surface area) silica gel was purchased from Daiso
(Osaka, Japan). Diisopropylethylamine, DACH, acryloyl
chloride, (3-aminopropyl)triethoxysilane, phosphorous
pentachloride, ammonium acetate, and dry toluene were
purchased from Fluka (Sigma-Aldrich, Buchs, Switzerland);
dimanganese decacarbonyl (Mn2(CO)10), dry tetrahydro-
furan, 2,2,2-trichloroacetyl isocyanate, DPEDA, and 4,40-
azobis-4-cyanopentanoic acid were purchased from Aldrich
(Sigma-Aldrich).
1-Methoxy-2-methyl-1-(trimethylsilyloxy)-1-propene was
purchased from Lancaster (Clariant group, UK). HPLC-
grade solvents were purchased from Merck (Darmstadt,
Germany). Chloroform (analytical grade) was dried by
filtration through an open glass column filled with neutral
alumina under inert atmosphere and then degassed with
helium. Chiral solutes (compounds 1�12) were available
from previous studies.
2.2 Apparatus
Diffuse reflectance infrared Fourier transform (DRIFT) and
transmission IR (potassium bromide pellets or liquid
paraffin dispersion) spectra were recorded on a Jasco 430
Fourier transform (FT) IR spectrometer (Jasco Europe,
Cremella, Italy) at a resolution of 4 cm�1. NMR spectra were
recorded on a Bruker Avance 400 spectrometer. Melting
points were determined on a Buchi B-545 instrument
(Flawil, Switzerland). Optical rotation values were obtained
on a Jasco P1030 polarimeter.
A lab-made quartz photochemical reactor, equipped
with a refrigerating chamber, an inert gas inlet, a mechanic
stirring and a high-pressure mercury (Hg) vapor lamp,
125 W (Helios Italquartz srl, Milan, Italy) was used for the
synthesis of CSP1 and CSP2.
Analytical liquid chromatography was performed on a
Waters chromatograph equipped with a Rheodyne model
7725i 20 mL loop injector, a 1525 binary HPLC pump and a
2487 dual wavelength absorbance detector (Waters, Milford,
MA, USA). Chromatographic data were collected and
processed using Empower software.
2.3 Preparation of N-(2-acryloylamino-(1R,2R)-cyclo-
hexyl)-acrylamide (R,R-DACH-ACR)
N-(2-Acryloylamino-(1R,2R)-cyclohexyl)-acrylamide or (R,R)-
DACH-ACR was prepared starting from DACH and acryloyl
chloride as previously described [5]. Elemental analysis: found
% C, 64.68; % H, 8.21; % N, 12.46; calculated for C12H18N2O2
% C, 64.84; % H, 8.16; % N, 12.60. Mp: 2331C. ½a�20D 5 185.4
(c 5 1.0; DMSO). 1H-NMR (d6-DMSO) d (ppm): 1.20–1.30
(m, 4H), 1.60–1.70 (m, 2H), 1.85–1.95 (m, 2H), 3.60–3.70
(m, 2H), 5.52 (dd, J 5 9.90 Hz, 2.44 Hz, 2H), 6.02
(dd, J 5 17.09 Hz, 2.44 Hz, 2H), 6.15 (dd, J 5 17.09 Hz,
9.90 Hz, 2H), 7.85 (d, 2H). 13C-NMR (d6-DMSO) d (ppm):
23.93, 31.30, 52.24, 124.74, 130.28, 166.14. FT-IR (KBr): 3284,
3075, 3033, 1656, 1625, 1550, 1410 cm�1.
2.4 Preparation of N-[2-acryloylamino-(1R,2R)-di-
phenylethyl]-acrylamide (R,R-DPEDA-ACR)
To a solution of DPEDA (1.0 g, 4.7 mmol) in 17.5 mL of dry
toluene was added diisopropylethylamine (1.6 mL,
9.4 mmol). To the cooled (01C, ice bath) solution was added
dropwise over a period of 1 h, with magnetic stirring under
an argon atmosphere, a solution of acryloyl chloride
(0.8 mL; 9.9 mmol) in 25 mL of dry toluene. The reaction
mixture was kept at 01C for 1 h, with magnetic stirring
under an argon atmosphere. The whitish precipitate was
collected by filtration, washed with toluene and hexane and
dried at reduced pressure (0.1 mbar, 251C) to yield 1.1 g of a
crude solid (73% yield), which was dissolved in 2-propanol
(20 mL) and precipitated by hexane (100 mL). The white
solid was filtered, washed with hexane and dried at reduced
pressure (0.1 mbar, 401C) to give 1.0 g of title compound
(66% overall yield). TLC: Merck plates Si-60-F254 eluent
CH2Cl2/MeOH 97:3 v/v, Rf 5 0.46. Elemental analysis:
found % C, 75.02; % H, 6.32; % N, 7.69; calculated for
C20H20N2O2 % C, 74.98; % H, 6.29; % N, 8.74. Mp: 2501C.
½a�25D 5 117.7 (c 5 1.0; DMSO). 1H-NMR (d6-DMSO) d
(ppm): 5.26 (dd, 2H), 5.54 (dd, J 5 9.95 Hz, 2.25 Hz, 2H),
6.08 (dd, J 5 17.05 Hz, 2.29 Hz, 2H), 6.31 (dd, J 5 17.05 Hz,
9.95 Hz, 2H), 7.20 (m, 10H), 8.70 (d, 2H). FT-IR (KBr):
J. Sep. Sci. 2010, 33, 3022–3032 Liquid Chromatography 3023
& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
3305, 3063, 3033, 1655, 1628, 1604, 1537, 1496, 1407 cm�1.13C-NMR (d6-DMSO) d (ppm): 57.60, 126.00, 127.35, 127.87,
128.35, 132.10, 140.96, 164.62. FT-IR (KBr): 3305, 3063,
3033, 1655, 1628, 1604, 1537, 1496, 1407 cm�1.
2.5 Preparation of 3-aminopropyl silica gel (3-APSG)
Spherical Daisogel SP-300-5P silica (10.0 g) was dried at
reduced pressure (0.1 mbar) at 1501C for 1 h. Dried silica
was placed in a 500 mL three-necked round bottom flask
equipped with a Dean Stark trap, reflux condenser and inert
gas inlet, and 240 mL of toluene were added. The slurry was
heated to reflux temperature with mechanical stirring under
an argon atmosphere, and 25 mL of distillate were collected
over a 1 h period.
After cooling to room temperature, 5.0 mL of
(3-aminopropyl)triethoxysilane (21.5 mmol) was added at
once and the slurry was heated to reflux temperature for 4 h,
with mechanical stirring under an argon atmosphere. After
cooling to room temperature, modified silica gel (3-APSG)
was collected by filtration, washed with 200 mL portions of
toluene, methanol, dichloromethane, and dried at reduced
pressure (0.1 mbar, T 5 601C) up to a constant weight
(weight increment: 4.0%). Elemental analysis: % C, 1.47; %
H, 0.46; % N, 0.44, corresponding to 325 mmol/g of starting
silica (2.82 mmol/m2) or 314 mmol/g of final matrix (based
on nitrogen). FT-IR (DRIFT): 3644, 2978, 2940, 1874, 1614,
1099, 948, 809 cm�1.
2.6 Activation of 3-aminopropyl silica gel with 2,2,2-
trichloroacetyl isocyanate (3-APSG-COCCl3)
3-Aminopropyl silica gel (3-APSG) obtained as described in
Section 2.5 (3.0 g) was dispersed in 10 mL of tetrahydrofur-
an, with mechanical stirring and under an argon atmo-
sphere. To the slurry was added at once 2,2,2-trichloroacetyl
isocyanate (1.0 g; 5.3 mmol), the stirring was continued at
room temperature for 1 h, and then the slurry was heated to
701C for 2 h. Modified silica gel (3-APSG-COCCl3) was then
isolated by filtration, washed with 50 mL portions of
tetrahydrofuran, methanol, acetone, dichloromethane, and
dried at reduced pressure (0.1 mbar, T 5 251C) up to a
constant weight (weight increment: 3.9%). Elemental
analysis: % C, 2.60; % H, 0.46; % N, 0.83, corresponding
to 324 mmol g�1 of starting silica (2.82 mmol m�2) or
296 mmol g�1 of final matrix (based on nitrogen). FT-IR
(DRIFT): 2947, 2885, 1732, 1707, 1548, 1479 cm�1.
2.7 General preparation of photo-induced polyacryl-
amide stationary phases (CSP1 and CSP2)
Into a lab-made quartz photochemical reactor methanol
(220 mL) was first placed and degassed by helium, and then
heated to 401C, under mechanical stirring (Fig. 1). The
diacryloyl monomer (1R,2R)-DACH-ACR (0.87 g; 3.9 mmol)
or (1R,2R)-DPEDA-ACR (0.87 g; 2.7 mmol) was charged into
the reactor under continuous stirring until complete
dissolution, followed by dimanganese decacarbonyl (0.1 g).
3-APSG-COCCl3 (2.9 g) was then added, under mechanical
stirring, and the slurry was degassed by helium. The reactor
was purged with argon for 5 min and then irradiated with a
125 W high-pressure Hg lamp with continuous stirring at
401C for 3 h. Modified silica gels (CSP1 and CSP2) were
isolated by filtration, washed with 100 mL portions of
methanol, tetrahydrofuran, dichloromethane, and dried at
reduced pressure (0.1 mbar, T 5 601C).
CSP1 (CSP-hn-poly-(R,R)-DACH-ACR): weight incre-
ment 22%. Elemental analysis: % C, 15.12; % H, 2.19; % N,
2.95, corresponding to 1082 mmol/g of starting silica
(9.41 mmol/m2) or 810 mmol/g of final matrix (based on
carbon). FT-IR (DRIFT): 2941, 2863, 1713, 1653, 1543,
1453 cm�1.
CSP2 (CSP-hn-poly-(R,R)-DPEDA-ACR): weight incre-
ment 20%. Elemental analysis: % C 17.25; % H, 1.99; % N,
2.33, corresponding to 799 mmol/g of starting silica
(6.95 mmol/m2) or 591 mmol/g of final matrix (based on
carbon). FT-IR (DRIFT): 3072, 3032, 2941, 1725, 1660, 1538,
1457 cm�1.
2.8 Activation of 3-APSG with 4,40-azobis-4-cyano-
pentanoic acid dichloride (3-APSG-AZO)
3-APSG was activated with 4,40-azobis-4-cyanopentanoic
acid dichloride as previously described [5] to yield 3-APSG-
AZO silica gel.
2.9 General preparation of thermal-induced poly-
acrylamide stationary phases (CSP3 and CSP4)
To a heated (601C) solution of the diacryloyl monomer
(1R,2R)-DPEDA-ACR in anhydrous, degassed chloroform
(0.45 g; 1.4 mmol in 40 mL) was added 3-APSG-AZO (3.0 g)
with mechanical stirring and under an argon atmosphere.
The slurry was heated and kept at reflux temperature for 6 h,
with continuous stirring. After cooling to room tempera-
ture, modified silica (CSP4) was isolated by filtration,
washed with 100 mL portions of methanol, acetone,
dichloromethane and dried at reduced pressure (0.1 mbar,
T 5 601C).
CSP3 (CSP-D-poly-(R,R)-DACH-ACR) was prepared as
previously described [5]. Elemental analysis: % C, 10.90; %
H, 1.76; % N, 2.30, corresponding to 590 mmol/g of starting
silica (5.14 mmol/m2) or 488 mmol/g of final matrix (based
on carbon). FT-IR (DRIFT): 3078, 2941, 2863, 2237, 1713,
1653, 1543, 1453 cm�1.
CSP4 (CSP-D-poly-(R,R)-DPEDA-ACR): weight incre-
ment: 5.6%. Elemental analysis: % C, 7.99; % H, 1.13; % N,
1.39, corresponding to 229 mmol/g of starting silica
(1.99 mmol/m2) or 196 mmol/g of final matrix (based on
J. Sep. Sci. 2010, 33, 3022–30323024 A. Ciogli et al.
& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
carbon). FT-IR (DRIFT): 3065, 3034, 2941, 2244, 1654, 1532,
1455 cm�1.
2.10 Columns packing and geometry
Stainless steel columns (250 mm� 4.6 mm id for CSP3;
250 mm� 4.0 mm id for CSP1, CSP2, and CSP4) supplied
by Alltech (IL, USA) were packed with CSP1�CSP4 at
9000 psi by using a previously described slurry packing
procedure [17]. The column dead times (t0) were determined
from the elution times of an unretained marker (1,3,5-tri-
tert-butylbenzene), using as eluent a mixture made up of
97% dichloromethane (amylene stabilized) and 3% metha-
nol, at T 5 251C, a flow-rate of 1.0 mL/min, and UV
detection at 254 nm. To maintain the same linear velocity,
flow-rate for CSP3 was adjusted to 1.3 mL/min.
2.11 Chromatographic conditions
HPLC separations were carried out under normal phase
(NP) and polar organic mode (POM) conditions on
CSP1�CSP4. In the first case, the mobile phase was made
up of 97% dichloromethane stabilized with ethanol
(�0.25%) and 3% methanol (v/v) for compounds 1�5. For
compound 6, a mobile phase consisting of 10% ethanol in
hexane was used. In the second case (POM mode), a
mixture of acetonitrile–methanol 70:30 v/v plus 20 mM
ammonium acetate was used as mobile phase for
compounds 7�9. For compounds 10�12, a mobile phase
made up of acetonitrile–methanol 85:15 v/v plus 20 mM
ammonium acetate was instead used. The flow-rate was set
to 1.0 mL/min for CSP1, CSP2, and CSP4, and adjusted to
1.3 mL/min for CSP3, to maintain the same linear velocity.
All the columns were thermostated at 251C. Samples were
dissolved in mobile phase and aliquots of 10–20 mL were
injected. Chromatograms were recorded by monitoring the
UV trace at 254 nm.
3 Results and discussion
3.1 Choice of a new synthetic strategy to hybrid
polyacrylamide CSPs
We recently reported a new hybrid organic/inorganic CSP
for HPLC synthesized by the g-from radical polymerization
of an enantiopure diacryloyl derivative of trans-1,2-diamino-
cyclohexane in the presence of mesoporous, azo-activated
silica micro-particles [4–6]. Hybrid CSPs based on synthetic
polymers as chiral selectors linked to inorganic supports
have gained attention and are proposed as synthetic
counterparts of the CSPs based on derivatized polysaccha-
rides. The expected advantages are increased mechanical,
thermal, and chemical stability accompanied by large
chromatographic efficiencies [18–22].
With this experience in hand, we moved to study
another way to induced polymerization of the above
monomer, and focused our attention on photo-induced
polymerization, which has been obtaining much attention
in the recent years because of its numerous industrial
Figure 1. Pictures of empty photo-chemical reactor (left) and filled withthe slurry of silica gel microparticlesbefore the photo-induced polymeri-zation (right).
J. Sep. Sci. 2010, 33, 3022–3032 Liquid Chromatography 3025
& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
applications [7–9]. We thus prepared two hybrid poly-
acrylamide stationary phases, CSP1 and CSP2, by a new
photo-induced polymerization approach based on the acti-
vation of mesoporous silica microparticles by chemically
bonded trichloroacetyl groups and using dimanganese
decacarbonyl Mn2(CO)10 as catalyst. The chiral polymeriz-
able monomers were enantiopure N,N0-diacryloyl derivatives
of trans-DACH (DACH-ACR) and DPEDA (DPEDA-ACR),
for CSP1 and CSP2, respectively.
3.2 Photochemical lab-made reactor
The surface-initiated photochemical grafting process was
carried out in a lab-made reactor tailored to address some
particular requirements of an heterogeneous reaction. A
photoreactor system was in fact sought in which all the
silica particles could be homogeneously exposed to the
irradiating region of the UV lamp and where mechanical
damage and gravitational segregation of the silica particles
could be avoided. On the basis of such design requirements
we developed the photoreactor shown in Fig. 1. We used a
quartz jacketed photoreactor system equipped with a high-
pressure mercury (Hg) vapor lamp and a propeller blade
stirrer for efficient, uniform motion of the reaction mixture.
The reactor is also equipped with gas inlet and sampling
ports, to provide easy degassing and O2 depletion of the
reaction medium by He sparging, to keep an inert atmo-
sphere (argon) during UV irradiation, and to allow for
sample withdrawing for reaction progress monitoring.
3.3 Outcome of the polymerization and character-
ization of the new polyacrylamide CSPs
The whole synthetic process (Fig. 2) starts with the
derivatization of 3-APSG with trichloroacetyl isocyanate to
generate the activated silica with surface-linked R-NH-CO-
NH-COCCl3 moieties (R 5 propyl, 3-APSG-COCCl3) which,
under irradiation with a mercury (Hg) lamp in the presence
of Mn2(CO)10, initiates the co-polymerization of the chiral
diacrylamide [(R,R)-DACH-ACR or (R,R)-DPDEA-ACR]
directly from the silica surface, yielding the final CSP1
and CSP2, respectively, containing the covalently grafted
chiral polymers [5]. A plausible mechanism of photoinitia-
tion involves a preliminary decomposition of Mn2(CO)10
into Mn(CO)5 species that react with trichloroacetyl groups,
yielding carbon-centered, surface-confined radicals [12, 13].
The latter, in the presence of chiral vinyl monomers, initiate
the polymerization yielding the final surface-grafted chiral
polymeric stationary phase [15, 16]. Grafting reactions were
monitored by DRIFT spectroscopy (Fig. 3), and the final
CSPs were fully characterized by elemental analysis (see
Table 1) and DRIFT. Activation of 3-APSG by trichloroacetyl
isocyanate is accompanied by some changes in the DRIFT
spectra (Fig. 3, top), which clearly indicate the presence of
additional absorption bands, diagnostic of the presence of
two different carbonyl groups on the silica surface (1732,
1707, 1548, and 1479 cm�1), in agreement with the
proposed structure containing the 1-trichloroacetyl-3-propyl-
urea fragment.
In the DRIFT spectrum of CSP1 (Fig. 3, middle),
intense signals arising from the chiral polymer of DACH-
ACR are clearly visible in the amide stretching region (1713,
1653, 1543 cm�1) and in the aliphatic stretching and bend-
ing regions (2941, 2863, 1453 cm�1). Similar stretching and
bending vibration bands are evident in the DRIFT spectrum
of CSP2 (Fig. 3, bottom), as they are diagnostic of the
presence of the chiral polymer of DPEDA-ACR (2941, 1725,
1660, 1538, 1457 cm�1); additional bands above 3000 cm�1
are symptomatic of phenyl groups (3072 and 3032 cm�1,
aromatic C–H stretching).
For comparative purposes, we prepared two homologue
CSPs (CSP3 and CSP4) by the thermal-induced approach
described in [5] from DACH-ACR and DPEDA-ACR,
respectively. Surface coverage characterization of the poly-
acrylamide CSP1–CSP4 and of aminopropyl silica gel
precusors are collected in Table 1. Carbon content data on
CSP1 show an average surface density of monomer units of
NH2NH2 NH2
NH
NH
CCl3
CHCl2 CHCl2 CHCl2
silica
B
NH
NH
O
O
NH
NH
O
O
or
O
O
NH
NH
CCl3
O
O
NH
NH
CCl3O
ONH
NH
O
O
NH
NH
O
O
NH
NH
O
O
hν, cat.
A
Figure 2. Synthetic pathway to polyacrylamide CSPs by surface-initiated photopolymerization. (A) 2,2,2-trichloroacetyl isocya-nate, tetrahydrofuran, r.t., 1 h-701C, 2 h; (B) Mn2(CO)10, metha-nol, irradiation with a 125 W high-pressure Hg lamp, 401C, 3 h.
J. Sep. Sci. 2010, 33, 3022–30323026 A. Ciogli et al.
& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
9.41 mmol/m2, corresponding to 810 mmol of monomer
units per gram of matrix. These values are larger than those
obtained on the thermal triggered CSP3, prepared from the
same chiral monomer and azo-activated silica (5.14 mmol/
m2, corresponding to 488 mmol/g of matrix).
The surface coverage of CSP2 based on carbon
content is 6.95 mmol/m2, corresponding to 591 mmol/g of
matrix. Again, these values are higher when compared to
those found on the homologue thermal triggered CSP4
(1.99 mmol/m2, corresponding to 196 mmol/g of matrix).
Further inspection of elemental analysis data yields the
average number of chiral monomer units of the new CSPs.
Thus, considering the number of mol of monomer units per
square meter of matrix (last column in Table 1), it is clearly
seen that all the aminogroups are converted to trichlor-
oacetyl derivatives, and that the averaged number of
monomer units in CSP1 and CSP2 is about 3.3 and 2.5,
respectively.
These findings clearly indicate that the photo-induced
grafting reaction yields higher polymer loading and grafting
efficiency in a shorter time, compared to the thermal
process, and no radical fragments escape from the silica
surface. Moreover, activation of the aminopropyl silica is
much easier in this case, as no by-products are generated.
Notably, the photo-induced process turned out to be the
method of choice for the grafting polymerization of DPEDA-
ACR, which showed low reactivity under thermal conditions
(6.95 versus 1.99 mmol/m2). Moreover, the solvent change
from chlorofom (the solvent used in the thermal process)
to methanol (the low wavelength UV cut-off solvent
used in the photo-triggered process) has a negligible effect
on the outcome of the DPEDA-ACR thermal grafting
polymerization.
With regard to the temperature of photochemical reac-
tion, we chose 401C to be sure that both chiral monomers
would be soluble and yield a suitable final concentration in
the reaction solvent.
3.4 Enantioresolution capabilities of the new poly-
acrylamide CSPs
In an effort to better understand the relationship between
structural features of the analytes and their retention and
enantioselectivity, a preliminary investigation was carried
out by screening a set of chiral compounds on the analytical
columns packed with the new CSPs. For comparison
purposes, retention and enantioselectivities were also
checked, under the same experimental conditions, on the
CSPs prepared by the thermal process.
The structures of the investigated analytes are shown in
Fig. 4. They can be divided into two sets, according to their
relative polarity and the elution mode chosen to analyze
them. Compounds 1�6 have no easily ionizable fragments
and can be eluted under NP conditions, using a single polar
mobile phase for 1�5 (dichloromethane/methanol 97:3 v/v)
and a less polar eluent for 6 (hexane/ethanol 90:10 v/v).
Compounds 7�12 are N-protected amino acids with differ-
ent fragments on the amino group and structural variations
on the side chains, and all of them have a free carboxyl
group that required polar organic solvents containing
20 mM ammonium acetate for their HPLC elution.
Inspection of CSPs and analytes structures suggests
that CSP-solute H-bonding is the dominant interaction
governing retention and, to some extent, enantioselectivity
4000 20000
20
40
60
80
100
120
% T
20
40
60
80
100
120
% T
0
20
40
60
80
100
% T
wavenumber / cm-1
3600 3200 2800 2400 1600 1200 800 400
4000 20003600 3200 2800 2400 1600 1200 800 400
4000 20003600 3200 2800 2400 1600 1200 800 400
Figure 3. FT-IR (DRIFT) spectra of 3-APSG-COCCl3 silica gel,CSP1 (CSP-poly-(R,R)-DACH-ACR), and CSP2 (CSP-poly-(R,R)-DPEDA-ACR) (from top to bottom).
J. Sep. Sci. 2010, 33, 3022–3032 Liquid Chromatography 3027
& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
in the examined systems. Amide C=O groups of our CSPs
may accept H-bonds from alcoholic, phenolic, amidic, or
carboxylic groups of the analytes, whereas amide NH groups
on the CSPs can act as H-bond donors towards H-bond
acceptor sites of the analytes. Dipole–dipole interactions
and, for CSP2 and CSP4, p–p interactions with aromatic
substrates are also expected to modulate both retention and
enantioselectivity.
Chromatographic data collected on the four CSPs 1�4
using NP elution mode for compounds 1�6 are listed in
Table 2. Inspection of retention (k1 and k2) and enantio-
selectivity (a) data disclosed marked differences between the
chromatographic behavior of photo- and thermal-induced
poly-DACH-ACR CSPs. Retention of both the enantiomers
is always larger on CSP1 (with the exception of the second
eluted enantiomer of compound 3), as a result of the higher
polymer loading obtained in the photo-initiated process (see
Table 1). An additional aspecific contibution to retention on
CSP1 can in principle arise from the polar acylurea frag-
ments present on the surface of the photo-initiated CSP.
Enantioselectivity is in favor of CSP3 for compounds 2�4and 6: for instance, compound 2 is resolved with a5 4.36 on
CSP3 and a5 2.61 on CSP1, and the difference in enan-
tioselectivity in favor of CSP3 is due to the decreased
Table 1. Surface coverage characterization of the polyacrylamide CSP1�CSP4 and of aminopropyl silica gel precusors
Support Chiral selector C (%) H (%) N (%) mmol/gc) mmol/m2
3-APSGa) � 1.47 0.46 0.44 325 (314) 2.82
3-APSG-COCCl3a) � 2.60 0.46 0.83 324 (296) 2.82
CSP1 (hn-induced)b) poly-(R,R)-DACH-ACR 15.12 2.19 2.95 1082 (810) 9.41
CSP3 (D-induced)b) poly-(R,R)-DACH-ACR 10.90 1.76 2.30 590 (488) 5.14
CSP2 (hn-induced)b) poly-(R,R)-DPEDA-ACR 17.25 1.99 2.33 799 (591) 6.95
CSP4 (D-induced)b) poly-(R,R)-DPEDA-ACR 7.99 1.13 1.39 229 (196) 1.99
a) Calculation was based on nitrogen content.
b) Calculation was based on carbon content.
c) mmol of monomer units per gram of starting silica (in parentheses are reported the same data relative to final matrix).
Figure 4. Chemical structures of chiral compounds.
J. Sep. Sci. 2010, 33, 3022–30323028 A. Ciogli et al.
& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
retention of both the enantiomers, compared to CSP1.
Compound 5, on the other hand, is better resolved on the
photo-induced CSP1 (a5 3.19) than on the thermal-induced
CSP3 (a5 2.07). A similar picture is observed for retention
data collected on the photo- and thermal-induced poly-
DPEDA-ACR CSPs, with the analytes showing larger affi-
nities for the photo-induced CSP2 compared to the corre-
sponding thermal-induced CSP4. Enantioselectivity in these
cases is in favor of CSP4 for compounds 1�3, whereas
compounds 5 and 6 are better resolved on CSP2, and
compound 4 shows no enantioseparation at all on the two
CSPs (a5 1.00). Figures 5 and 6 show the chromatographic
resolutions of compounds 6 and 4, respectively, on the four
different CSPs, under the same experimental conditions. In
most cases, narrow and symmetric peaks are observed for
the resolved enantiomeric analytes, indicating a fast and
efficient solute mass transfer between mobile phase and
polymeric stationary phases.
Taken together, these data clearly indicate that the two
CSPs prepared by photo-initiated polymerization have
increased retention under NP elution for compounds 1�6,
compared to the CSPs prepared by thermal-initiated poly-
merization. The higher polymer loading and the polar
acylurea moiety from which the polymer grows in CSP1 and
CSP3 may account for the observed differences in retention.
Enantioselectivity has no clear cut dependence on the
analyte structure and on the polymerization conditions:
here, the interplay of several factors including polymer
loading, polymer structure, conformation, and mobility,
aspecific solute–CSP interactions established which of the
Table 2. Chromatographic data obtained for the enantioresolution of compounds 1–6 (Fig. 4) by polyacrylamide CSPs under NP
conditions
Normal phase (NP) poly-DACH-ACR-CSPs poly-DPEDA-ACR-CSPs
hn-induced CSP1 D-induced CSP3 hn-induced CSP2 D-induced CSP4
Compound Eluent k1 k2 a2,1 k1 k2 a2,1 k1 k2 a2,1 k1 k2 a2,1
1 CH2Cl2/MeOH 97:3 0.10 0.19 1.90 0.09 0.16 1.78 0.19 0.48 2.53 0.04 0.14 3.50
2 CH2Cl2/MeOH 97:3 3.41 8.89 2.61 1.25 5.45 4.36 1.98 2.72 1.37 0.61 0.95 1.56
3 CH2Cl2/MeOH 97:3 7.99 9.54 1.19 6.03 9.81 1.63 8.27 15.02 1.82 3.77 7.36 1.95
4 CH2Cl2/MeOH 97:3 3.55 4.88 1.37 2.00 3.49 1.75 0.35 0.35 1.00 0.15 0.15 1.00
5 CH2Cl2/MeOH 97:3 0.21 0.67 3.19 0.15 0.31 2.07 0.33 0.59 1.79 0.28 0.33 1.18
6 Hexane/EtOH 90:10 11.59 11.59 1.00 4.85 4.98 1.03 6.35 12.14 1.91 3.80 4.56 1.20
0
Minutes
0
Minutes
0
Minutes
0
Minutes
Δ Δ
5 10 15 20 25 10 20 30 40 50
5 10 15 20 255 10 15 20 25
Figure 5. NP HPLC of compound 6 onhn-induced – (top right) and D-induced – (bottom right) CSPs-poly-(R,R)-DACH-ACR, together with hn-induced – (top left) and D-induced –(bottom left) CSPs-poly-(R,R)-DPEDA-ACR. Eluent: n-hexane/etha-nol, 90:10 v/v; flow-rate: 1.0 mL forCSP1, CSP2, CSP4 and 1.3 mL/min forCSP3; T 5 251C; detection: UV at254 nm.
J. Sep. Sci. 2010, 33, 3022–3032 Liquid Chromatography 3029
& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
two phases performs better in terms of enantiomeric
recognition capability.
Chromatographic data collected on the four CSPs 1�4
using POM for compounds 7�12 are listed in Table 3.
Compounds 7�12 all have a free carboxyl group whose
presence required the addition of ammonium acetate
in the mobile phase to have short analysis times and
acceptable peak shapes (see Fig. 7 for compound 7).
Retention of the analytes under these conditions is
mainly controlled by the ionizable carboxyl fragment, and
retention differences between photo- and thermal activated
CSPs are less pronounced compared to NP conditions. In
most cases, retention is higher on the CSPs prepared by
thermal-initiated polymerization. The differences in enan-
tiorecognition abilities between photo- and thermal-induced
poly-DACH CSPs are smaller compared to the NP elution
mode, and here are always in favor of CSP3, except for
compound 6. Comparison of enantioselectivity data of
photo- and thermal-induced poly-DPEDA CSPs shows, on
the other hand, that the photo-induced CSP2 has larger
values and that CSP4 fails to resolve five out of six
compounds.
Examination of the enantioselective properties of the two
CSPs prepared by photo-initiated polymerization suggests
that their selectivities are influenced by a combination of
different factors, thus offering rather complementary appli-
0
Minutes
0
Minutes
0
Minutes
Minutes
Δ Δ
5 10 15 20 251 2 3 4 5
0 1 2 3 4 5 5 10 15 20 25
Figure 6. NP HPLC of compound 4 onhn-induced – (top right) and D-induced – (bottom right) CSPs-poly-(R,R)-DACH-ACR, together with hn-induced – (top left) and D-induced –(bottom left) CSPs-poly-(R,R)-DPEDA-ACR. Eluent: dichloro-methane/methanol, 97:3 (v/v); flow-rate: 1.0 mL for CSP1, CSP2, CSP4and 1.3 mL/min for CSP3; T 5 251C;detection: UV at 254 nm.
Table 3. Chromatographic data obtained for the enantioresolution of compounds 7–12 (Fig. 4) by polyacrylamide CSPs under POM
conditions.
Polar organic mode (POM) poly-DACH-ACR-CSPs poly-DPEDA-ACR-CSPs
hn-induced CSP1 D-induced CSP3 hn-induced CSP2 D-induced CSP4
Compound Eluent k1 k2 a2,1 k1 k2 a2,1 k1 k2 a2,1 k1 k2 a2,1
7 ACN/MeOH 70:30 1 20 mM AcONH4 0.67 0.82 1.22 0.73 0.97 1.33 0.53 0.68 1.28 0.73 0.73 1.00
8 ACN/MeOH 70:30 1 20 mM AcONH4 0.63 0.78 1.24 0.72 0.90 1.25 0.63 0.63 1.00 0.87 0.87 1.00
9 ACN/MeOH 70:30 1 20 mM AcONH4 0.51 0.71 1.40 0.56 0.90 1.61 0.46 0.46 1.00 0.51 0.51 1.00
10 ACN/MeOH 85:15 1 20 mM AcONH4 1.90 2.17 1.14 1.26 1.53 1.22 1.08 1.20 1.11 1.30 1.30 1.00
11 ACN/MeOH 85:15 1 20 mM AcONH4 1.89 2.47 1.31 1.25 1.66 1.33 1.20 1.33 1.11 1.28 1.28 1.00
12 ACN/MeOH 85:15 1 20 mM AcONH4 2.43 2.74 1.13 1.71 2.10 1.23 1.50 1.98 1.32 1.53 1.64 1.07
J. Sep. Sci. 2010, 33, 3022–30323030 A. Ciogli et al.
& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
cation fields. This is highligthed by the comparative plots of
Figs. 5 and 6, where enantioselectivity towards compound 6is observed only on poly-(R,R)-DPEDA-ACR CSP (see Fig. 5),
whereas compound 4 was resolved only by poly-(R,R)-DACH-
ACR CSP (see Fig. 6). Additionally, the two polymeric
selectors may have different enantioselectivities under the
same experimental condition, as shown in Fig. 7: a non-
racemic mixture ad hoc prepared by mixing the two enan-
tiopure enantiomers of Fmoc-Phe (compound 7) was resolved
on both CSP1 and CSP2, but an elution order inversion took
place by changing the selector moiety.
4 Concluding remarks
The surface-initiated photopolymerization of chiral, enan-
tiopure diacryloyl derivatives of 1,2-diamines on mesopor-
ous silica yielded hybrid polyacrylamide CSPs that have
been characterized by elemental analysis and DRIFT
spectroscopy. The new photopolymerization process is
highly efficient in terms of polymer loading and reaction
time, and leads to very stable hybrid polymeric organic–
inorganic CSPs. The photopolymerization process is
succesfull also for those monomers (DPEDA-ACR) that
are less reactive in the thermal polymerization. HPLC
columns packed with the above CSPs show good enantios-
electivities towards a broad range of compounds under NP
and POM elution conditions. Enantioselectivity is accom-
panied by high chromatographic efficiency and permeabil-
ity, as a result of a polymerization process that generates a
uniform polymer layer on the silica surface.
We thank Sapienza Universita di Roma, Italy (Funds forselected research topics 2008�2010) and Istituto Pasteur-Fondazione Cenci Bolognetti, Roma, Italy, for financial support.
The authors have declared no conflict of interest.
5 References
[1] Blaschke, G., Angew. Chem. Int. Ed. 1980, 19, 13–24.
[2] Blaschke, G., J. Liq. Chromatogr. 1986, 9, 341–368.
[3] Edmondson, S., Osborne, V. L., Huck, W. T. S., Chem.Soc. Rev. 2004, 33, 14–22.
[4] Gasparrini, F., Misiti, D., Villani, C., PCT Int. Appl. (2003),WO 2003079002 A2 20030925.
[5] Gasparrini, F., Misiti, D., Rompietti, R., Villani, C.,J. Chromatogr. A 2005, 1064, 25–38.
[6] Cavazzini, A., Dondi, F., Marmai, S., Minghini, E., Massi,A., Villani, C., Rompietti, R., Gasparrini, F., Anal. Chem.2005, 77, 3113–3122.
[7] Fouassier, J. P., Photoinitiation, Photopolymerization,and Photocuring: Fundamentals and Applications,Hanser Publishers, Munich Vienna, New York, 1995.
[8] Mishra, M. K., Yagci, Y., Handbook of Radical VinylPolymerization, Marcel Dekker, New York, 1998,pp. 167�172.
[9] Wu, G. Q., Nie, J., J. Photochem. Photobiol. A 2006, 183,154–158.
[10] Dursun, C., Degirmenci, M., Yagci, Y., Jockusch, S.,Turro, N. J., Polymer 2003, 44, 7389–7396.
0
Minutes
0
Minutes
Δ
Minutes
Minutes
Δ
2 4 6 8 10 0 2 4 6 8 10
0 2 4 6 8 102 4 6 8 10
Figure 7. POM HPLC of compound 7
(non-racemic mixture ad hocprepared by mixing the two enantio-pure enantiomers of Fmoc-Phe) onhn-induced – (top right) and D-induced – (bottom right) CSPs-poly-(R,R)-DACH-ACR, together with hn-induced – (top left) and D-induced –(bottom left) CSPs-poly-(R,R)-DPEDA-ACR. Eluent: acetonitrile/methanol, 70:30 v/v 1 20 mM ammo-nium acetate; flow-rate: 1.0 mL forCSP1, CSP2, CSP4 and 1.3 mL/min forCSP3; T 5 251C; detection: UV at254 nm.
J. Sep. Sci. 2010, 33, 3022–3032 Liquid Chromatography 3031
& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
[11] Bi, Y., Neckers, D. C., Macromolecules 1994, 27,3683–3693.
[12] Yagci, Y., Hepuzer, Y., Macromolecules 1999, 32,6367–6370.
[13] Shirai, Y., Shirai, K., Tsubokawa, N., J. Polym. Sci. PartA 2001, 39, 2157–2163.
[14] Shirai, Y., Tsubokawa, N., React. Funct. Polym. 1997, 32,153–160.
[15] Bamford, C. H., in: Jenkins, A. D., Ledwith, A. (Eds.),Reactivity, Mechanism and Structure in PolymerChemistry, Wiley, New York 1974, p. 52.
[16] Bamford, C. H., in: Allen, N. S., Rabek, J. F., (Eds.), NewTrends in the Polymer Photochemistry, Elsevier AppliedScience, New York 1995, pp. 129�145.
[17] D’Acquarica, I., Gasparrini, F., Giannoli, B., Badaloni, E.,Galletti, B., Giorgi, F., Tinti, M. O., Vigevani, A.,J. Chromatogr. A 2004, 1061, 167–173.
[18] Barnhart, W. W., Gahm, K. H., Hua, Z., Goetzinger, W.,J. Chromatogr. B 2008, 875, 217–229.
[19] Lee, K.-P., Choi, S.-H., Kim, S.-Y., Kim, T.-H., Ryoo, J. J.,Ohta, K., Jin, J.-Y., Takeuchi, T., Fujimoto, C., J. Chro-matogr. A 2003, 987, 111–118.
[20] Allenmark, S. G., Andersson, S., Moller, P., Sanchez, D.,Chirality 1995, 7, 248–256.
[21] Andersson, S., Allenmark, S. G., Moeller, P., Persson,B., Sanchez, D., J. Chromatogr. A 1996, 741, 23–31.
[22] Lammerhofer, M., J. Chromatogr. A, 2010, 1217, 814–856.
J. Sep. Sci. 2010, 33, 3022–30323032 A. Ciogli et al.
& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com