Accepted Manuscript

An investigation on the structure and catalytic activity of cryptomelane-typemanganese oxide materials prepared by different synthesis routes

Catherine Almquist, Mark Krekeler, Lulu Jiang

PII: S1385-8947(14)00542-7DOI: http://dx.doi.org/10.1016/j.cej.2014.04.102Reference: CEJ 12081

To appear in: Chemical Engineering Journal

Received Date: 20 January 2014Revised Date: 23 April 2014Accepted Date: 26 April 2014

Please cite this article as: C. Almquist, M. Krekeler, L. Jiang, An investigation on the structure and catalytic activityof cryptomelane-type manganese oxide materials prepared by different synthesis routes, Chemical EngineeringJournal (2014), doi: http://dx.doi.org/10.1016/j.cej.2014.04.102

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customerswe are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, andreview of the resulting proof before it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

An investigation on the structure and catalytic activity of cryptomelane-type

manganese oxide materials prepared by different synthesis routes

By

Catherine Almquist1*, Mark Krekeler2, Lulu Jiang1

1Chemical and Paper Engineering Department, Miami University, Oxford, Ohio 45056

2Geology Department, Miami University, Hamilton, Ohio 45011

*Corresponding Author, almquic@miamioh.edu; 513-529-0767

Abstract

Cryptomelane is an octahedral molecular sieve with a 2x2 tunnel structure (OMS-2)

and a stoichiometric formula of approximately KMn8O16. Cryptomelane OMS-2 is

gaining increasing attention in a range of applications, including catalysis. However,

the impacts of synthesis methods on catalytic performance are still not well

understood. Cryptomelane OMS-2 was prepared in this study by three methods: 1) a

sol-gel method, 2) a reflux method, and 3) a milling method. The physical and

chemical characteristics of these materials were investigated, and the catalytic

performances of these materials were compared for the gas phase oxidation of ethanol.

A novel aspect of this study is that cryptomelane was prepared by the milling method

using post-consumer alkaline battery waste, and this novel cryptomelane material was

compared with others prepared from commercially-available precursors and

published laboratory methods.

Cryptomelane prepared from battery waste via the milling method has a low surface

area (4 m2/g) and a lower catalytic activity than cryptomelane samples prepared in

this study with purchased chemicals. However, all cryptomelane samples used in this

study, including the samples prepared from battery waste, could oxidize ethanol

vapors at inlet concentrations ≤ 500 ppm by greater than 95% at 200 C. The catalytic

activities of all cryptomelane samples studied were stable for up to 8 days at a

reaction temperature of 175 C at an inlet ethanol concentration of 400 ppm. The

implication of this study is that post-consumer alkaline batteries may be used as a raw

material for volatile organic carbon oxidation catalysts.

1. Introduction

Cryptomelane is a manganese oxide octahedral molecular sieve (OMS) with a 2x2

tunnel structure and a stoichiometric formula of approximately KMn8O16. It has been

investigated as a useful and active catalyst in several types of reactions in both liquid

and gas phases [1]. For example, cryptomelane was studied as a liquid-phase catalyst

for selective oxidation reactions [2-10] and as a photocatalyst for degradation of

contaminants in water [11]. It has been studied as a gas phase oxidation catalyst for

the oxidation of carbon monoxide (CO) [12-17] as well as for the oxidation of

volatile organic carbon (VOC) compounds, such as formaldehyde [18-23], ethyl

acetate [24-29], ethanol [30-35] and other oxygenates [36,37], and aromatics [38-41],

among others.

The activity of the catalyst has been reported to be a result of the mixed valence state

of the manganese (Mn) and the mobility of the lattice oxygen in the cryptomelane

surface structure [1,3-5,29,34,35, 37]. In a review by Suib (2008), it is stated that

mixed valence is important in electron transfer, energy transfer, and redox catalysis

and that oxygen vacancies are related to potential catalytic activity [1]. Therefore,

the average oxidation state (AOS) of cryptomelane is important in characterizing the

redox properties of cryptomelane, as some researchers have correlated AOS or redox

properties with catalytic activity [1,19,31,33,34, 37,38 ]. Other properties that impact

catalytic activity have been found to include surface area [3,19,21] acidity [6, 15],

hydrophobicity [15,38,40], and morphology [16,19-22,33,38], among others.

The kinetic model most often proposed for the catalytic oxidation of organics on

cryptomelane is the Mars-Van Krevelen mechanism [4,12,37,38]. There are two key

steps in this mechanism: 1) lattice oxygen at the surface of the catalyst reacts with

adsorbed organic compounds, which reduces manganese in cryptomelane from Mn4+

to Mn2+ and creates an oxygen vacancy in the catalyst, and 2) gas-phase oxygen

replenishes the oxygen vacancy and oxidizes the reduced catalyst. According to a

discussion by Makwana, et al. [4], when the catalysis system is at steady-state, the

rate at which the adsorbed organic compounds are oxidized by lattice oxygen equals

the rate at which gas-phase oxygen replenishes the resulting oxygen vacancy. Under

conditions of steady-state and excess oxygen in the system, the initial rate at which

the organic compounds are oxidized over cryptomelane takes the form [4] depicted in

Equation 1:

][][][Rcb

Ra

dt

Rd

+

=− (1)

where [R] represents the concentration of organic substrate and a,b, and c are model parameters.

Researchers have doped cryptomelane in an effort to enhance its catalytic activity

[12,14,23,29,37,38]. Enhanced catalytic activity as a result of adding a dopant to

cryptomelane has been attributed to enhanced oxygen mobility, changes in average

oxidation state, and changes in the reducibility of Mn at the surface of cryptomelane

[13,37]. However, the dopant does not necessarily change the morphology of the

cryptomelane. For example, Hernandez, et al (2010) doped cryptomelane with

copper (Cu), cobalt (Co), nickel (Ni), and zinc (Zn) and found that the dopants do not

have any noticeable effect on the structural or textural properties of cryptomelane

[13]. However, the activity for CO oxidation was enhanced with Cu and Co, whereas

very little change in activity was observed with Ni and Zn dopants [13].

This study focuses on the preparation, characterization and evaluation of

cryptomelane and its activity for the catalytic oxidation of ethanol. Ethanol was

selected because of its importance as an additive to gasoline and as a renewable fuel.

In fact, the final EPA Renewable Fuel Standard targets 6 million gallons of cellulosic

biofuels for 2013 [42], of which ethanol is a key component. The manufacture,

distribution, and use of ethanol will generate ethanol emissions, which must be

treated in pollution control operations. Catalytic oxidation utilizing volatile organic

compound (VOC) catalysts is one method of controlling such emissions, and in this

study, cryptomelane prepared by different synthesis methods is investigated as a

VOC catalyst.

The motivation for this research was to assess whether cryptomelane prepared from

post-consumer alkaline battery waste could be used as a VOC oxidation catalyst.

This is the novel aspect of this study; the precursor to one of the cryptomelane

materials prepared here is post-consumer alkaline battery waste, which contains zinc,

manganese dioxide, and potassium hydroxide. It is estimated that nearly three billion

of these batteries are purchased and thrown away each year, amounting to nearly

100,000 tons of battery waste [43]. This waste typically ends up in landfills, where

the metals in the batteries have potential to leach into the environment. Therefore,

finding a use for this waste stream would have both economic and environmental

benefits to battery manufacturers and users.

The recovered Mn-oxide and potassium hydroxide components from the batteries

contains zinc, and so the resulting cryptomelane prepared from the battery waste also

has zinc in it. What impacts will the zinc have on the catalytic activity of the

cryptomelane?

The four key objectives of the study are to 1) assess the impact of synthesis method

on the chemical and physical properties of cryptomelane and its catalytic activity for

ethanol oxidation; 2) assess the impact of calcination temperature on the chemical

and physical properties of cryptomelane and its catalytic activity, 3) assess the impact

of zinc on cryptomelane’s physical and chemical properties and on its catalytic

activity for the gas-phase oxidation of ethanol, since alkaline battery waste contains

zinc, and 4) compare the catalytic activity of cryptomelane made from battery waste

with those synthesized with purchased chemicals. These objectives are supported by

Prieto, et al (2003) who state that small changes in synthesis parameters can lead to

materials with different catalytic properties [44].

2. Experimental Methods

2.1 Cryptomelane synthesis

The cryptomelane was prepared in this study using three methods: 1) a sol-gel

method, 2) a reflux method, and 3) a milling method. Each of these methods is

described in more detail below.

2.1.1 Sol-gel Method. The sol-gel method used here was described by Ching and

Roark [45]. Briefly, a 100 mM solution of potassium permanganate (KMnO4) was

prepared. Fumaric acid (C4H4O4) was added to the solution at a molar ratio of 3:1

KMnO4/C4H4O. After adding the two chemicals into the deionized water, the mixture

forms a brownish flocculent gel. The gel is then filtered, washed, dried, and calcined

in air for 2 hours. Finally the black cryptomelane is washed again with acidic water

and then deionized water.

The sol-gel catalysts were used to investigate the role of calcination temperature on

the properties and catalytic activity of cryptomelane. Aliquots of the prepared

crytomelane were calcined in air for 2 hours at 350 C, 400 C, 500 C, and 600 C.

2.1.2 Reflux Method. The reflux method used here was described by Frias, et al [46]

and by Valente, et al [47]. 11 g of manganese acetate was added to 40 mL of

deionized water. 5 mL of glacial acetic acid was added to the solution. This solution

was brought to a boil in a fluxing apparatus with vigorous stirring. In a separate

beaker, 6.5 g of KMnO4 was added to 150 mL of deionized water. This KMnO4

solution was then added slowly to the refluxing solution and allowed to react for 24

hours under refluxing conditions. After cooling, the black precipitate was filtered,

washed with deionized water, and dried. These samples were then calcined at 450 C

for 24 hours.

2.1.3 Milling Method. The milling method, or solid state method, used here was

described by Ding, et al. [48], Li, et al. [49], and Valente, et al. [47]. In brief,

KMnO4 and Mn acetate were placed in a mortar in a molar ratio of 2:3. The materials

were ground for several minutes until the mixture was a gray color and uniform

texture. The mixtures were then placed in a glass vial and heated to 80 C for 24

hours. The samples were then washed with deionized water, dried, and calcined in

air at 450 C for 24 hours. According to Li, et al., the milling method is simple and

easy [49]. However, the amount of water and the length of grinding time could

influence the morphology of the resulting powders [49].

The milling method was also used to prepare cryptomelane from post-consumer

alkaline battery waste as a precursor. Duracell AA cell sized alkaline batteries with

varying controlled discharge levels provided by Duracell were digitally photographed

and catalogued, recording the expiration date, manufacturing country, weight, and

residual voltage. Using a Dremel tool equipped with a diamond tipped or embedded

rotary cutting blade, the batteries were cut perpendicular to the length at the positive

end. The batteries were then cut parallel to the length. The last cut is perpendicular to

the length at the negative end of the battery. The cathode was gently pulled from the

battery, and the metal casing was peeled off, often using a hand tool for leverage.

The black, outer manganese cathode is then separated and removed from the battery,

by means of pulling or gently chipping or chiseling, and it was then placed in a vial.

The zinc anode that was wrapped in the paper lining was removed and placed in to a

separate vial. The spent cathode in the vials was then powdered, using an agate

mortar and pestle. It was crushed into as fine a powder as possible often requiring 10

minutes, to as much as 60 minutes for each battery, depending on how indurated the

material is. Material was crushed in an up-and-down motion and not ground in a

rotary motion so as to preserve as much crystallinity of poorly crystalline phases as

possible. The powder was then placed into ceramic alumina crucibles in preparation

for the heat modification. Each vial of material was placed into a separate crucible,

to avoid contamination between samples. The loaded alumina crucibles were then

placed in a programmable muffle furnace heated to 550°C for 30 minutes in ambient

atmosphere. The accuracy of the thermocouple is accepted to be ± 3 ˚C and the

temperature fluctuation is typically none (approximately 95% of the time) or may be

as much as 3 ˚C. After the samples have been heated, the crucibles are taken out of

the furnace and allowed to cool for an hour before being placed in new, labeled vials.

The reflux and milling methods of synthesis were used to investigate the effect of

zinc on the catalytic activity of cryptomelane. Zinc was added to the cryptomelane

during synthesis by adding zinc acetate to the reaction mixtures in a Zn/Mn mass

ratio of 0.05 Zn/Mn. The cryptomelane prepared from battery waste has a similar

amount of Zn because it is contained in the post-consumer battery waste.

2.2 Cryptomelane Characterization

Cryptomelane was characterized for BET surface area and pore size distribution by

nitrogen adsorption, thermal stability by thermal gravimetric analyses (TGA), crystal

structure and crystallinity by X-ray diffraction (XRD), atomic composition by energy

dispersion spectroscopy (EDS), and morphology by scanning electron microscopy

(SEM). Each is described in turn below.

2.2.1 BET Surface area and pore size distribution. The BET surface area and pore

size distribution for each cryptomelane sample was measured by nitrogen adsorption

at 77 K using a Micromeritics Tri-Star BET surface area analyzer. Prior to analysis,

the samples were degassed for 30 minutes at 150 C with a helium purge.

2.2.2 Thermal Gravimetric Analyses (TGA). Thermal gravimetric analyses (TGA)

were conducted on all cryptomelane samples used in this study with a TA

Instruments Q500 TGA. The temperature was varied from ambient temperature to

900 C using a 10 C/min ramp. Nitrogen gas was used to purge the furnace at 60

mL/min. A platinum pan was used as the sample holder.

2.2.3 X-ray Diffraction (XRD). Powder X-ray diffraction data was collected using

a Scintag X-1 powder diffractometer using monchromated Cu radiation (1.5418Å)

equipped with a Peltier detector, operating at 40 kV and 35 mA. Cryptomelane

samples were scanned from 5◦ to 65◦ 2θ at 0.02◦ steps using a count time of 7.5s per

step. To analyze the phases present within the minerals, powder diffraction files

(PDFs) were used as analogues. The PDFs used were cryptomelane (PDF #042-1348)

and hausmanite (PDF #024-0734). The samples were prepared for XRD analysis by

gently grinding the material with an agate mortar and pestle and using pack mounts.

2.2.4 Scanning Electron Microscope and Energy Dispersive Spectrometry

(SEM/EDS). The scanning electron microscope that was used to collect data is a

variable pressure Zeiss Supra 35VP FEG field emission scanning electron microscope

(FESEM). Images were acquired in variable pressure mode using N2 as the

compensating gas. The instrument is equipped with an EDAX Genesis 2000 energy

dispersive spectrometry (EDS) detector, which has a detection limit of approximately

0.1 wt% for most elements. To prepare the samples for SEM, small chunks of the

rock were carefully broken off and applied to a carbon sticky tab placed on an

aluminum tab using clean forceps.

2.2.5 Average oxidation state (AOS). The average oxidation state (AOS) was

determined in two parts: 1) the determination of total Mn in a sample of cryptomelane

using Inductively Coupled Plasma Optical Emission Spectrometry (Agilent 720ES

axial-viewing ICP-OES), and 2) the determination of oxidation state using reaction

and titration methods described by Tali [50]. Approximately 20 mg of cryptomelane

was dissolved in 5 mL concentrated hydrochloric acid (HCl) under heat

(approximately 80 C). During this step, argon gas was bubbled through the HCl and

Mn slurry at 20 ccpm. The chlorine gas (Cl2) resulting from the reduction of Mn3+

and Mn4+ to Mn2+ and oxidation of Cl- to Cl0 was captured in 1N potassium iodide

(KI) solution, also under heat (~80 C). The reaction between HCl and Mn was

allowed to continue until the solution of Mn became clear. At the end of the reaction

step, the Mn/HCl solution was diluted to 100 mL with deionized water, and these

diluted solutions were analyzed for total Mn using ICP-OES. The KI solutions

turned burnt orange to brown in color as Cl2 reacted with iodide (I-) to form I2. With

excess I- in solution, the iodine reacted to form I3-. The I3

- was finally titrated with

0.1N sodium thiosulfate solution until the solution turned clear. The AOS test system

used for this analysis is shown schematically in Figure 1a.The chemical reactions that

were used in these analyses are shown below, where Equations 2 and 3 are those that

reduce Mn3+ and Mn4+ to Mn2+ and generate Cl2, Equation 4 describes the reaction

between Cl2 and I- to form I2, and Equation 5 describes the titration of I2 back to I-

with thiosulfate solution:

)(21

)()()( 223

gClaqMnaqClsMn +→++−+ (2)

)()()(2)( 224 gClaqMnaqClsMn +→+

+−+ (3)

)()(2)(2)( 22 aqIaqClaqIgCl +→+−− (4)

)()(2)(2)( 264

2322 aqOSaqIaqOSaqI

−−−

+→+ (5)

The average oxidation state, then, was determined according to Equations 6 and 7,

making the assumption that Mn is present in the cryptomelane sample in Mn3+ and

Mn4+ oxidation states:

totalMn

VCa −= 2 (6)

AOS = 3α + 4(1- α) (7)

Where α = fraction of total Mn that is Mn3+

1-α = fraction of total Mn that is Mn4+

V = volume of sodium thiosulfate solution used in titration (L)

C = concentration of sodium thiosulfate solution (0.1 M)

Mntotal = moles of Mn in the sample (based upon ICP-OES)

2.3 Gas-phase catalytic oxidation of ethanol

The cryptomelane samples were investigated as catalysts for the oxidation of ethanol.

The test system used for the catalytic oxidation tests is shown in Figure 1b. Ethanol

vapor was generated using a diffusion cell at concentrations that ranged from 400

ppm to 2000 ppm. The ethanol vapor in air was directed at 100 mL/min through a

stainless steel tube (0.635 cm OD; 0.124 cm wall thickness) reactor containing 50 mg

cryptomelane (ground to powder with morter and pestle to particle size < 1 mm) in

each run (gas-hourly space velocity (GHSV) = 120,000/hr). The tube furnace used in

this study was a Lindberg Blue M tube furnace. The reactor effluent was collected at

least one hour after the reaction temperature was attained and analyzed via gas

chromatography (Agilent 6890) with flame ionization detection (GC/FID).

The catalytic oxidation of ethanol oxidation results in two predominant reaction

products: acetaldehyde and carbon dioxide [27,28,32,33]. In this study, the carbon

dioxide in the effluent was collected by a precipitation method in KOH/EtOH

solution in selected experimental trials. The carbon dioxide reacts with KOH to form

potassium bicarbonate, which is insoluble in ethanol. The potassium bicarbonate was

subsequently filtered, dried, and weighed to gravimetrically determine the amount of

CO2 captured as KHCO3. After each experimental trial in which carbon dioxide was

captured, a carbon balance was conducted to assess losses of carbon. In each

experiment for which CO2 was measured, the moles of carbon accounted for in

ethanol, acetaldehyde, and CO2 in the effluent was within 90% of the total moles of

carbon fed to the reactor. Trace compounds in the effluent, which may include

ethylene, carbon monoxide, and acetic acid, were not observed with the analytical

methods used in this study.

3. Results and Discussion

3.1 Surface area, pore volumes, and pore size distributions.

Heterogeneous catalysis is a surface phenomena, and so surface area, pore volumes,

and pore size distributions in the cryptomelane samples all have implications to the

number of and accessibility of substrates to active sites on the surface of the catalyst.

All of the cryptomelane samples have been analyzed for BET surface area, and the

results are shown in Table 1. The surface area of cryptomelane decreases

significantly with calcination temperature among the sol-gel samples. This is

expected due to sintering of aggregates and particle growth with temperature. The

presence of Zn (up to 3 wt% ) in the reflux and milled cryptomelane samples

decreased the surface area by 23% and 28%, respectively. The Zn likely blocked

pores in the samples, which decreased the surface area accessible to the nitrogen gas

during the analyses. The surface area of the battery-waste derived cryptomelane is

very low compared to the other lab-made cryptomelane samples. Reasons for the low

surface area for the battery-derived cryptomelane may include the presence of sulfur

and zinc on the surface of the cryptomelane, which after calcination, sinters and

possibly blocks pores and surface area on the battery-waste-derived cryptomelane.

The pore volumes are also provided in Table 1. Although micropores (< 20 Å),

mesopores (20Å - 500Å), and macropores (> 500 Å) exist in the cryptomelane

samples, the use of nitrogen as the adsorbate to measure surface area limits the

micropore analyses [1,15]. Nitrogen access to the micropores is limited due to the

presence of the potassium ion in the narrow (< 7Å) channels [1,15]. Therefore, the

pore volumes reported in Table 1 were measured between 17Å and 3000Å. For the

cryptomelane samples without Zn, the pore volumes vary inversely with bulk density

of the materials, as shown in Figure 2. The reflux sample has the highest pore

volume of all samples prepared for this study at 0.47 ±0.12 cm3/g. Approximately 80

percent of the pore volume in the reflux sample is contained in macropores, which

accounts for its relatively low bulk density (0.37 g/cm3). The pore volumes for all

other cryptomelane samples used in this study are predominantly contained in the

mesopores (20Å - 500Å). Zn lowers the pore volume in the reflux and milled

samples by greater than 60 percent and 20 percent, respectively. This is likely due to

Zn blocking both mesopores and macropores that form from agglomeration and

sintering of cryptomelane particles during synthesis and calcination. Calcination of

the sol-gel samples also decreased pore volumes in samples calcined from 350 C to

500 C. However, the sol-gel sample calcined at 600 C had a similar pore volume to

that calcined at 350 C. Upon further inspection of the raw data, as calcination

temperature of the sol-gel samples increases, the mesoporosity deceases while the

macroporosity increases. Sintering of the aggregates decreases the mesoporosity,

which is observed in the sol-gel samples as calcination temperature increases from

350 C to 600 C. The increased macroporosity observed after calcination of the sol-

gel sample at 600 C may be attributed to the coalescing of particles, or particle

coarsening, which would form larger particles with larger pores between them

compared to samples calcined at lower temperatures.

The nitrogen adsorption and desorption curves for representative cryptomelane

samples are shown in Figures 3a-d. The hysteresis in the adsorption/desorption

curves indicates mesoporosity in the cryptomelane samples. Insets in Figures 3a-d

show the mesopore size distributions as calculated by BJH adsorption dV/dD pore

volumes. The reflux cryptomelane sample has a very narrow mesopore size

distribution centered at 27 Å. However, as stated above, most of the pore volume in

the reflux sample is contained in macropores, which are not shown in the insets. In

contrast, the milled cryptomelane samples (those derived from purchased chemicals

(Figure 3b) and from battery waste (Figure 3d)) have much broader mesopore size

distributions, centered at approximately 200 Å. The milled samples prepared with

purchased chemicals have no observable pore volume in the macropore region,

whereas the battery waste-derived cryptomelane has nearly 50 percent of the pore

volume contained in macropores. The sol-gel samples have mesopore size

distributions (Figure 3c) that are skewed, but the maximum in the mesopore size

distribution is near 100 Å. The mesopores in these samples are likely the result of

cryptomelane particles aggregating and sintering during calcination.

3.2 Average oxidation state (AOS)

The AOS of the cryptomelane samples ranged from 3.4 to 3.8, as shown in Table 1.

These values are supported by those reported in other studies for cryptomelane [5, 6,

16]. The AOS of the samples used in this study did not reveal any trends with

calcination temperature or presence of Zn. Chen, et al. [18] has concluded that AOS

in the narrow range of 3.5 to 3.9 does not obviously influence catalytic performance.

A comparison of the AOS values obtained for cryptomelane (Table 1) with the

catalytic reaction rates of ethanol provided in Table 2 show that there is no

observable trend in catalytic performance with AOS in this study, either.

3.3 Thermal stability using TGA

The TGA curves (% weight change vs temperature) for the cryptomelane samples

used in this study are shown in Figures 4a-c. According to the published literature [2,

3, 8, 15], weight losses at temperatures < 200 C are due to losses of physically

adsorbed water, between 200 C and 500 C are due to losses of chemisorbed water and

lattice oxygen in the cryptomelane, between 500 C and 700 C are due to losses of

oxygen due to changes in the crystal structure from cryptomelane to bixbyite, and

losses at temperatures higher than 700 C are due to losses of oxygen from the catalyst

due to a change in crystal structure to hausmannite. A comparison of weight losses

over specified temperature ranges from cryptomelane samples used in this study is

shown in Table 3. The reflux and milled samples, including the milled sample from

battery waste, behave similarly between 500 C and 900 C, with ~4% weight loss

starting at 500 C and another ~3% weight loss starting at 700 C. These weight losses

correlate to changes in crystal structure from cryptomelane to bixbyite to

hausmannite, starting at 500 C and 700 C, respectively, as described in the literature

[2,3]. The sol-gel samples lose much more water at temperatures up to 200 C than

the other cryptomelane samples (4%-7% of initial mass compared to <2% for the

reflux and milled samples). In addition, there is no marked change in mass at 500 C

with the sol-gel samples like there is in the reflux and milled samples. Rather, there

is a relatively steady change in mass (2.5% to 4% of initial mass) between 200 C and

700 C. At 700 C, the sol-gel samples have a marked weight loss of 4%-5%,

indicating a loss of oxygen and change in structure, similarly to that observed in the

reflux and milled cryptomelane samples at 700 C. The weight losses for all

cryptomelane samples used in this study at temperatures between 200 C and 900 C

total approximately 8%.

The apparent weight gain observed in the TGA curve for the sol-gel sample calcined

at 350 C is likely an artifact. Such artifacts can occur due to slight fluctuations in the

nitrogen purge gas. Other possible reasons for the apparent weight gain include

adsorption or reactions with the nitrogen or impurities in the purge gas.

3.4 Crystal structure and crystallinity

Figures 5 and 6 show XRD diffraction patterns of cryptomelane samples used in this

study. Comparisons of the patterns in Figures 5 and 6 with those of a standard

cryptomelane diffraction pattern (JCPDS file 29-1020, not shown) indicate that all of

our samples have the cryptomelane crystal structure. Figure 5 summarizes the XRD

diffraction patterns of the sol-gel samples calcined at temperatures from 350 C to 600

C. As calcination temperature increases, the XRD peaks sharpen up to a calcination

temperature of 500 C, indicating crystal growth. Using the Scherrer equation, the

crystal sizes of the cryptomelane samples were calculated using the XRD peaks at the

[211] plane, and they ranged from approximately 20 nm to 30 nm. The results are

summarized in Table 1. The crystal sizes obtained in this study are supported by

Schurz, et al [3], who also calculated crystal sizes using the [211] plane between 10

nm and 30 nm for cryptomelane samples. The crystal size increases with calcination

temperature among the sol-gel samples calcined at temperatures up to 500 C. There

is no clear effect of Zn on crystal size for reflux and milled samples. However, the

milled samples, in general, have smaller crystal sizes after calcination than reflux or

sol-gel cryptomelane samples. This is apparent in both the milled samples prepared

with purchased precursors and those prepared with alkaline battery waste. This result

is due to the lower extent of contact between the precursor chemicals prior to

calcination. According to Portchault, et al [51], tunnel-based manganese oxide

materials crystallization occurs through two steps: 1) a disordered or layered

manganese oxide precursor is formed, and 2) the transformation of that precursor

structure into 1-dimensional tunnel-based manganese oxide upon aging. The tunnel

structures are generally stabilized by water and by the cation (K+) [51]. Whereas

sol-gel and reflux methods are solution-phase methods of synthesis, the solid state or

milling method is dry. The dry method may limit the extent of contact and re-

structuring of the precursors prior to calcination, resulting in a smaller crystal size for

the milled samples.

Figure 6a compares the XRD diffraction patterns for cryptomelane prepared by

different methods. All samples in Figure 6a show similar characteristic peaks for

cryptomelane. Figures 6b and 6c show XRD diffraction patterns that compare

cryptomelane with and without Zn. The XRD patterns suggest that Zn at

concentrations up to 3 wt% has little to no apparent effect on the crystal structure or

crystallinity of cryptomelane. This is supported by Hernandes, et al [13] who also

found no observable change in structure or texture properties of cryptomelane after

doping with transition metals.

3.5 Morphology

SEM images of all cryptomelane samples prepared in this study were acquired and

are shown in Figures 7a-d and 8a-d. Figures 7a-d show SEM images of sol-gel

cryptomelane calcined at different temperatures. The fiber lengths significantly

increase with calcination temperature. The sample calcined at 600 ℃ appeared to be

relatively non-uniform, and features that were characteristic of those observed after

calcination at 350 C, 400 C, and 500 C were also observed in the 600 C sample. A

time-temperature profile during calcination that was not consistent with those for

other samples may be the reason for the non-uniformity in the 600 C sample.

Figures 8a-d show SEM micrographs of cryptomelane prepared by the reflux method,

milling method, sol-gel method, and cryptomelane prepared with battery waste.

Cryptomelane prepared by the sol-gel and reflux methods have significantly longer

fibers than those samples prepared by the milling method. Valente, et al [47] also

compared the cryptomelane samples prepared using solid state or the milling method

with the reflux method. They, too, observed larger and more organized fibers in the

reflux-prepared materials compared with the milling-method materials [47].

3.6 Effect of zinc in cryptomelane

Figures 6b, 6c, and 9 show the effect of Zn in the cryptomelane samples. Figures 6b

and 6c show XRD diffraction patterns that indicate no observable effect of Zn on the

crystal structure of cryptomelane. Figure 9 compares SEM micrographs of reflux and

milled cryptomelane samples with and without the addition of Zn. The comparisons

in Figure 9 show little effect of Zn on the morphology of the samples. This is

supported by Peiteado, et al [52] who found that Zn does not diffuse into manganese

oxides to a detectable extent until calcination at 700 C. In addition, Hernandez, et al

[13], also observed no significant changes in the structural or textural properties of

cryptomelane after doping with various transition metal cations, including Zn.

3.7 Catalytic Activity

A comparison of the ethanol conversions with reaction temperature for cryptomelane

samples prepared by the sol-gel (400 C), reflux, and milling method, including that

prepared with battery waste is shown in Figure 10. Table 2 summarizes the observed

reaction rates per mass of catalyst and per surface area of catalyst, and it lists the

calculated activation energies for the catalytic oxidation of ethanol. All catalysts

studied have catalytic activity for the oxidation of ethanol. Greater than 95%

conversion of ethanol at inlet concentrations ≤ 500 ppm can be achieved with the

cryptomelane catalysts used in this study at a reaction temperature of 200 C.

Considering the conversion of ethanol as a measure of catalytic activity, the catalytic

activity of cryptomelane prepared by all three methods (sol-gel, reflux, and milling

method) with purchased precursors are comparable, approximately 30 umoles/min/g

catalyst at 175 C and 400 ppm inlet ethanol concentration; the rate of ethanol vapor

oxidation over battery-waste-derived cryptomelane under the same conditions

averaged slightly lower.

The rate of ethanol oxidation over cryptomelane increases with inlet ethanol

concentration, as shown in Figure 11. According to the rate equation developed for

Mars Van Krevelen kinetics at steady-state under excess oxygen (Equation 1)[4]:

][

][][

Rcb

Ra

dt

Rd

+

=− (1)

the rate of ethanol oxidation should increase with ethanol concentration at low

ethanol concentrations or as long as parameter b >> c[R]. As [R] increases, the rate of

oxidation should approach zero order (when parameter b << c[R]). For the battery-

waste-derived cryptomelane, which has very low surface area compared to the other

cryptomelane samples used in this study, the rate of reaction increases with ethanol

inlet concentration but much more flatly than the increase in reaction rate observed

for the other cryptomelane catalysts in this study. One reason for the higher activity

of the purchased chemical-derived cryptomelane is their higher specific surface area.

The surface areas of the reflux and milled cryptomelane samples (~70 m2/g),

prepared with purchased precursors were more than an order of magnitude greater

than the surface area of the battery-waste-derived cryptomelane (~4 m2/g). The

flatter increase in rate of ethanol oxidation with inlet ethanol concentration over

battery-waste-derived cryptomelane suggests that the reaction is catalyst active site-

limited compared to the other cryptomelane samples. This is supported by Schurz, et

al [3], who found that reactivity of cryptomelane depends almost linearly with

specific surface area for the aerobic partial oxidation of benzyl alcohol. Similar

trends between surface area and catalyst activity were also observed by others for the

partial oxidation of benzyl alcohol [5,8].

However, surface area alone cannot be used to explain the differences in catalytic

activity. In this study, a comparison of the catalytic activity of sol-gel samples

(Figures 12a,b) shows that surface area, alone, cannot be used to predict activity. In

Figure 12a, the sol-gel samples were compared with an inlet ethanol concentration of

500 ppm. All samples were able to completely oxidize ethanol vapor at 200 C under

the conditions used in this study. However, in Figure 12b, the inlet ethanol

concentration was doubled to 1000 ppm. In Figure 12b, it is clear to see that the sol-

gel sample calcined at 400 C had the highest activity of the sol-gel samples.

Although the cryptomelane sample calcined at 350 C had higher surface area, it did

not perform as well as the cryptomelane sample calcined at 400 C. This is supported

by Tian, et al. [20], who found two cryptomelane-type manganese oxide materials

having similar specific surface areas have very different catalytic activities. Tian, et

al [20] suggest that morphology and accessibility to the pores influence the catalytic

activity more so than specific surface area. In this study, SEM micrographs of the sol-

gel cryptomelane calcined at 350 C contained evidence of a carbon layer on the

surface of the catalyst (not shown), suggesting that the precursors used for

synthesizing the cryptomelane may not have been completely removed during

calcination. The presence of this carbon layer or deposit may have affected the

apparent catalytic activity of this catalyst.

The catalyst activity of cryptomelane is related to the structure, oxygen mobility, and

the number of and accessibility to surface sites on the catalyst [24, 33, 34,35]. The

accessibility to surface area for the ethanol oxidation reaction will be dependent upon

the relative pore sizes of the cryptomelane samples. In a review by Spivey [53], most

industrial gas-phase reactions are controlled by mass transport within the catalyst

particle or by heat transport between the particle and the flowing gas. Therefore,

mass transport resistances can, in part, explain the differences in apparent catalytic

activity. With the cryptomelane samples used in this study having different pore size

distributions, differences in accessibility to the active surface sites and pore structure

may impact the observed differences in catalytic activity.

Figures 13a and b show the effect of zinc on the catalytic activity of cryptomelane.

There is no observable effect of the zinc on catalytic activity for both the reflux and

milled cryptomelane samples. This is supported by Hernandez, et al [13], who also

showed very little impact of zinc on cryptomelane’s catalytic activity.

Figure 14 shows that the predominant reaction products observed during this study

are acetaldehyde and carbon dioxide. In this study, acetaldehyde was observed at

reaction temperatures between 150 C and 200 C. Peluso, et al [32] also observed

acetaldehyde when ethanol was reacted over manganese oxide-based catalysts.

Santos, et al [35] observed acetaldehyde as a partial oxidation product over

cryptomelane at reaction temperatures between 100 C and 210 C. In this study, a

carbon balance of greater than 90% was obtained during reactions with high (>2000

ppm) initial ethanol concentrations. The high initial ethanol concentrations resulted

in carbon dioxide concentrations in the reactor effluent that facilitated its capture and

quantification. Defensible and reproducible quantification of CO2 was not achieved

at initial ethanol concentrations less than 2000 ppm due to the small concentrations of

CO2 in the reactor effluent. The calculated (open triangles) and measured (solid

triangles) carbon dioxide shown in Figure 14 were similar. Trace reaction products,

which may include carbon monoxide, ethylene, acetic acid, and acetone were not

observed in this study.

Equations 8 and 9 show the stoichiometry of gas-phase ethanol oxidation to

acetaldehyde and carbon dioxide, which is supported by several published works

[32,33,35]:

OHOHCOOHHC 242252 2

1+→+ (8)

OHCOOOHC 22242 222

5+→+ (9)

According to Santos, et al [35], ethanol is adsorbed onto the catalyst surface as

ethoxides, which are converted to acetates as temperature increases. The ethoxy

species were also reported by Li, et al on Mn and cerium-based catalysts [54] and by

Peluso, et al over cryptomelane catalysts [30]. The interactions of these adsorbed

intermediates with the catalyst surface favor hydrogen abstraction to form

acetaldehyde, or cleavage of the C-C bonds for complete oxidation. The types of

surface sites that contribute to the activity of the catalyst include lattice oxygen sites

(O2-) and hydroxyl sites (OH-) [35].

The steps in the Mars Van Krevelen mechanism are depicted in Equations 10- 13,

leading to an observed reaction as depicted in Equation 8. Ethanol is adsorbed as an

ethoxide on the surface of cryptomelane. The lattice oxygen oxidizes the adsorbed

ethanol, while Mn4+ is reduced to Mn2+. The hydrogens abstracted from ethanol are

combined with the lattice oxygen and are liberated from the surface as water, leaving

an oxygen vacancy (*) in the catalyst. Gaseous oxygen replenishes the vacancy while

oxidizing the reduced catalyst from Mn2+ to Mn4+.

CH3CH2OH (g) + Mn4+

� CH3C(O)H (g) + Mn2+

+ 2 H+ (10)

2H+ + O2

2- � H2O(g) + ½ O2 (g) + * (11)

O2 (g) + * + 2e-� O2

2- (12)

Mn2+ � Mn4+ + 2e- (13)

Combining Equations 10-13 gives the overall reaction that is observed for the initial

oxidation of ethanol to acetaldehyde, depicted in Equation 8.

The activation energies for ethanol on cryptomelane catalysts used in this study range

from approximately 59 kJ/mole to 67 kJ/mole (Table 2). These values are within the

range of reported activation energies for the catalytic oxidation of ethanol, which

range from approximately 25 kJ/mole to greater than 130 kJ/mole for a variety of

catalysts [55-61]. The activation energies reported in Table 2 were obtained by

calculating the observed reaction rate constant for each sample as a function of

reaction temperature. For this analysis, it was assumed that the oxidation of ethanol

was apparently first order with respect to ethanol concentration. This is supported by

Ismagilov, et al [60], who found that the oxidation of ethanol followed a power law,

with the power for ethanol in the conversion of ethanol to acetaldehyde being close to

unity. In addition, it is supported by Equation 1 [4] when the concentration of

ethanol is relatively low. Arrhenius’s equation (Equation 10, linearized form in

Equation 11) was used to graphically determine the activation energy for each

catalyst. Figure 15 shows an example of one such plot of ln (k) vs (1/T) for the

cryptomelane sample prepared by the reflux method. The slope (-E/R) was used to

calculate E, the activation energy.

RT

E

Aek−

= (10)

−=

TR

EAk

1)ln()ln( (11)

Where k = reaction rate constant (s-1)

A = pre-exponential factor (s-1)

E = activation energy (kJ/mole)

R = gas constant (0.008314 kJ/mole/K)

T = absolute temperature (K)

Figure 16 shows that the cryptomelane samples prepared for this study are stable for

the oxidation of ethanol for greater than 8 days. Although little to no change in

reactivity was noted in the catalytic experiments, SEM micrographs of the samples

taken before and after use in the reaction for 8 days (not shown) suggest that carbon

deposits formed on the cryptomelane samples. The stability of Mn-based catalysts

was also observed by Peluso, et al [30] and by Li, et al [54] for the catalytic oxidation

of ethanol. Chen, et al [14] observed a decay in catalyst activity with time when

investigating CO oxidation over silver (Ag)-doped cryptomelane at low temperatures

(<100 C). They attributed the deactivation to an accumulation of carbonates on the

surface of the catalyst. Santos, et al [24] investigated the stability of cryptomelane

catalysts for the total oxidation of ethyl acetate. They observed slight deactivation of

the catalyst over the first 20 hours of reaction, but the catalyst was stable afterwards

through 90 hours’ time on stream. The slight deactivation, however, was not

attributed to coking [24]. Lahousse, et al [25] also found the cryptomelane catalysts

to be time-stable over several days. However, variations in VOC conversions were

observed, depending upon the conditions of the reaction. These variations have been

attributable to surface coverage of reactants and products, including that of water [24].

Santos, et al [39] investigated the stability of cryptomelane during the oxidation of

toluene vapor. They concluded that deactivation at temperatures lower than 270 C

can be attributable to adsorbed reactants and products on the catalyst surface.

Deactivation does not occur, however, at higher temperatures [39].

Conclusions

The conclusions of this study are summarized below:

• All cryptomelane samples synthesized and used in this study have catalytic

activity for the gas-phase oxidation of ethanol, including the cryptomelane

sample prepared from alkaline battery waste. Nearly 100% conversion of

ethanol vapors were observed at 200 C when the inlet ethanol concentration

was < 500 ppm.

• The synthesis method and calcination temperature impact the physical and

chemical properties of cryptomelane catalysts.

o All cryptomelane samples used in this study have XRD patterns

characteristic of cryptomelane.

o Synthesis method has a significant impact on morphology. Solution

methods of synthesis result in fiber-like morphologies of cryptomelane.

Dry methods of synthesis (milling) form much shorter fiber

morphology.

o Calcination increases fiber length, decreases surface area, and

increases the pore volume in macropores.

o Pore size distributions are significantly different with synthesis

method. The reflux method results in predominantly macropore

volume with a narrow distribution of mesopores centered at

approximately 30 Å. The milled samples contain pore volume in the

mesopore range with a broad pore-size distribution centered at

approximately 200 Å. Sol-gel samples have pore sizes predominantly

in the mesopore region with a skewed pore size distribution that has a

maximum at approximately 100 Å.

o TGA curves for the reflux and milled samples show that marked

weight losses occur at 500 C and 700 C, which correspond to

structural changes in the catalyst. TGA curves for the sol-gel samples

do not have the marked weight loss at 500 C, however, the weight loss

is gradual between 200 C and 700 C. The weight losses between 200

C and 900 C for all cryptomelane samples are comparable at

approximately 8%.

• Zn on cryptomelane decreased surface area and pore volume. However, there

was no apparent effect of Zn on the crystal structure and catalytic activity of

cryptomelane.

• The reaction of ethanol vapor at inlet concentrations > 500 ppm on battery-

waste-derived cryptomelane appeared to be limited by it available surface area.

The rate of ethanol oxidation increased with increasing inlet ethanol

concentration, but that increase was relatively flat with the battery-derived

cryptomelane compared to those with higher specific surface areas.

• The cryptomelane catalysts were stable for the oxidation of ethanol for 8 days

at a reaction temperature of 175 C. Although no apparent degradation in

catalyst activity was observed, carbon deposition was qualitatively observed

on the catalysts following the 8-day reaction period in SEM micrographs.

• Battery waste can be used as a pre-cursor to VOC oxidation catalysts. The Zn

impurity in the battery waste should not impact the catalyst activity for the

gas-phase oxidation of ethanol. Future efforts should be made to develop a

method to increase the surface area of the battery-waste derived catalysts.

Acknowledgements

The authors would like to acknowledge Dr. John Morton of Miami’s Geology

Department, who assisted us on this manuscript with XRD and ICP-OES analyses. In

addition, the authors would like to acknowledge Dr. Richard Edelman of Miami’s

Molecular Microscopy Lab for his time and assistance on SEM images and EDS

analyses. The authors would also like to acknowledge Mr. Jared Minges and Mr.

Ryan Flannery for efforts made in the lab toward the progress of this project.

References

[1] Suib, S.L. (2008). Structure, porosity, and redox in porous mangaenes oxide

octahedral layer and molecular sieve materials. Journal of Materials Chemistry,

18, pp. 1623-1631.

[2] Jothiramalingam, R, Viswanathan, B, and Caradarajan, TK (2006). Synthesis,

characterization and catalytic oxidation activity of zirconium-doped K-OMS-2

type manganese oxide materials. Journal of Molecular Catalysis A: Chemical,

252, pp. 49-55.

[3] Schurz, F., Bauchert, J.M., Merker, T., Schleid, T., Hasse, H., and Glaser, R.

(2009). Octahedral molecular sieves of the type K-OMS-2 with different particle

sizes and morphologies: Impact on the catalytic properties in the aerobic partial

oxidation of benzyl alcohol. Applied Catalysis A: General, 355, pp. 42-49.

[4] Makwana, V.D., Son, Y-C., Howell, A.R., and Suib, S.L. (2002). The role of

lattice oxygen in selective benzyl alcohol oxidation using OMS-2 catalyst: A

kinetic and isotope-labeling study. Journal of Catalysis, 210, pp. 46-52.

[5] Makwana, V.D., Garces, L.J., Liu, J., Cai, J., Son, Y-C., Suib, S.L. (2003).

Selective oxidation of alcohols using octahedral molecular sieves: influence of

synthesis method and property-activity relations. Catalysis Today, 85, pp. 225-

233.

[6] Kumar, R., Sithambaram, S., and Suib, S.L., (2009). Cyclohexane oxidation

catalyzed by manganese oxide octahedral molecular sieves-Effect of acidity of the

catalyst. Journal of Catalysis, 262, pp. 304-313.

[7] Jing, HU, Keqiang, SUN, Daiping, HE, Boqing, XU (2007). Amorphous

manganese oxide for catalytic aerobic oxidation of benzyl alcohol. Chinese

Journal of Catalysis, 28(12), pp. 1025-1027.

[8] Dharmarathna, S., King'ondu, C.K., Pedrick, W., Pahalagedara, L., and Suib, S.L.,

(2012). Direct sonochemical synthesis of manganese octahedral molecular sieve

(OMS-2) nanomaterials using cosolvent systems, their characterization, and

catalytic applications. Chemistry of Materials, 24(4), pp. 705-712.

[9] Liu, J., Son, Y-C., Cai, J., Shen, Z., Suib., S.L., and Aindow, M. (2004). Size

control, metal substitution, and catalytic application of cryptomelane

nanomaterials prepared using cross-linking reagents. Chemistry of Materials,

16(2),pp. 276-285.

[10] Villegas, J.C., Garces, L.J., Gomez, S., Furand, J.P., and Suib, S.L. (2005).

Particle size control of cryptomelane nanomaterials by use of H2O2 in acidic

conditions. Chemistry of Materials, 17, pp. 1910-1918.

[11] Iyer, A., Galindo, H., Sithambaram, S., King’ondu, C., Chen, C-H., and Suib,

S.L. (2010). Nanoscale manganese oxide octahedral molecular sieves (OMS-2)

as efficient photocatalysts in 2-propanol oxidation. Applied Catalysis A: General,

375, pp. 295-302.

[12]Liu, X-S., Jin, Z-N., Lu, J-Q., Wang, X-X., and Luo, M-F. (2010). Highly active

CuO/OMS-2 catalysts for low temperature CO oxidation. Chemical Engineering

Journal, 162, pp. 151-157.

[13] Hernandez, W. Y., Centeno, M. A., Romero-Sarria, F., Ivanova, S., Montes, M.,

and Odriozola, J.A. (2010). Modified cryptomelane-type manganese dioxide

nanomaterials for preferential oxidation of CO in the presence of hydrogen.

Catalysis Today, 157, pp. 160-165.

[14] Chen, J., Li, J., Li, H., Huang, X., and Shen, W. (2008). Facile synthesis of Ag-

OMS-2 nanorods and their catalytic applications in CO oxidation. Microporous

and Mesoporous Materials, 116, pp. 586-592.

[15] Luo, J., Zhang, Q., Garcia-Martinez, J., and Suib, S.L. (2008). Adsorptive and

acidic properties, reversible lattice oxygen evolution, and catalytic mechanism of

cryptomelane-type manganese oxides as oxidation catalysts. Journal of the

American Chemical Society, 130, pp. 3198-3207.

[16] Ozacar, M., Payraz, A.S., Genuino, H.C., Kuo, C-H., Meng, Y.,and Suib, S.L.

(2013). Influence of silver on the catalytic properties of the cryptomelane and

Ag-hollandite types manganese oxides OMS-2 in the low-temperature CO

oxidation. Applied Catalysis A: General, 462-463, pp. 64-74.

[17] Gac, W. (2007). The influence of silver on the structural, redox and catalytic

properties of the cryptomelane-type manganese oxides in the low-temperature CO

oxidation reaction. Applied Catalysis B: Environmental, 75, pp. 107-117.

[18] Chen, T, Dou, H, Li, X, Tang, X, Li, J, and Hao, J (2009). Tunnel structure

effect of manganese oxides in complete oxidation of formadehyde. Microporous

and Mesoporous Materials, 122, pp. 270-274.

[19] Tian, H., He, J., Liu, L., Wang, D., Hao, Z., Ma, C. (2012). Highly active

manganese oxide catalysts for low-temperature oxidation of formaldehyde.

Microporous and Mesoporous Materials, 151, pp. 397-402.

[20] Tian, H., He, J., Zhang, X., Zhou, L., and Wang, D. (2011). Facile synthesis of

porous manganese oxide K-OMS-2 materials and their catalystic activity for

formaldehyde oxidation. Microporous and Mesoporous Materials, 138, pp. 118-

122.

[21] Tian, H., He, J., Liu, L., Wang, D. (2013). Effects of textural parameters and

noble metal loading on the catalytic activity of cryptomelane-type manganese

oxides for formaldehyde oxidation. Ceramics International, 39, pp. 315-321.

[22] Chen, H, He, J, Zhang, C, and He, H (2007). Self-assembly of novel

mesoporous manganese oxide nanostructures and their application in oxidative

decomposition of formaldehyde. Journal of Physical Chemistry, 111, pp. 18033

– 18038.

[23] Tang, X, Li, J, and Hao, J (2010). Significant enhancement of catalytic activities

of manganese oxide octahedral molecular sieve by marginal amount of doping

vanadium. Catalyst Communications, 111, pp. 871-875.

[24] Santos, VP, Pereira, MRF, Orfao, JJM, and Figueired, JL (2009). Synthesis and

characterization of manganese oxide catalysts for the total oxidation of ethyl

acetate. Topics in Catalysis, 52, pp. 470-481.

[25] Lahousse, C, Bernier, A, Grange, P, Delmon, B, Papaefthimiou, P, Ioannides, T,

and Verykios, X (1998). Evaluation of γ-MnO2 as a VOC removal catalyst:

comparison with a noble metal catalyst. Journal of Catalysis, 178, pp. 214-225.

[26] Gandhe, AR, Rebello, JS, Figueiredo, JL, and Fernandes, JB (2007).

Manganese oxide OMS-2 as an effective catalyst for total oxidation of ethyl

acetate. Applied Catalysis B: Environmental, 72, pp. 129-135.

[27] Sanz, O, Delgado, JJ, Navarro, P, Arzamendi, G, Gandia, LM, and Montes, M

(2011). VOCs combustion catalyzed by platinum supported on manganese

octahedral molecular sieves. Applied Catalysis B: Environmental, 110, pp. 231 –

237.

[28] Santos, VP, Pereira, MFR, Orfao, JJM, and Figueiredo, JL (2011). Mixture

effects during the oxidation of toluene, ethyl acetate, and ethanol over a

cryptomelane catalyst. Journal of Hazardous Materials, 185, pp. 1236 – 1240.

[29] Santos, V.P., Soares, O.S.G.P., Bakker, J.J.W., Pereira, M.F.R., Orfao, J.J.M.,

Gascon, J., Kapteijn, F., and Figueiredo, J.L. (2012). Structural and chemical

disorder of cryptomelane promoted by alkali doping: Influence on catalytic

properties. Journal of Catalysis, 293, pp. 165-174.

[30] Peluso, M.A., Pronsato, E., Sambeth, J.E., Thomas, H.J., and Busca, G. (2008).

Catalytic combustion of ethanol on pure and alumina supported K-Mn oxides: An

IR and flow reactor study . Applied Catalysis B- Environmental, 78, pp. 73-79.

[31] Peluso, M.A., Gambaro, L.A., Pronsato, E., Giazzoli, D., Thomas, H.J., and

Sambeth, J.E. (2008). Synthesis and catalytic activity of manganese dioxide (type

OMS-2) for the abatement of oxygenated VOCs. Catalysis Today, 133, pp. 487-

492.

[32] Peluso, M.A., Sambeth, J.E., and Thomas, H.J. (2003). Complete oxidation of

ethanol over MnOx. Reaction Kinetics Catalysis Letters, 80(2), pp. 241-248.

[33] Lamaita, L., Peluso, M.A., Sambeth, J.E., and Thomas, H.J. (2005). Synthesis

and characterization of manganese oxides employed in VOC abatement. Applied

Catalysis B-Environmental, 61, pp. 114-119.

[34] Bastos, SST, Orfao, JJM, Freitas, MMA, Pereira, MFR, and Figueireda, JI

(2009). Mangenese oxide catalysts synthesized by exotemplating for the total

oxidation of ethanol. Applied Catalysis B: Environmental, 93, pp. 30-37.

[35] Santos, VP, Pereira, MRF, Orfai, JJM, and Figueiredo, JI (2010). The role of

lattice oxygen on the activity of manganese oxides towards the oxidation of

volatile organic compounds. Applied Catalysis B: Environmental, 99, pp 353-363.

[36] Mccaffrey, E.F., Klissurski, D.G., and Ross, R.A. (1972). The effect of oxygen

content in a series of manganese oxide catalysts on their catalytic activity and

selectivity in the decomposition of isopropyl alcohol. Journal of Catalysis, 26, pp.

380-387.

[37] Sun, M., Yu, L., Ye, F., Diao, G., Yu, Q., Hao, Z., Zheng, Y., and Yuan, L.

(2013). Transition metal doped cryptomelane-type manganese oxide for low

temperature catalytic combustion of dimethyl ether. Chemical Engineering

Journal, 220, pp. 320-327.

[38] Genuino, HC, Dharmarathna, S, Njagi, EC, Mei, MC, and Suib, SL (2012).

Gas-phase total oxidation of benzene, toluene, ethylbenzene, and xylenes using

shape-selective manganese oxide and copper manganese oxide catalysts. Journal

of Physical Chemistry, 116, pp. 12066-12078.

[39] Santos, VP, Bastos, SST, Pereira, MFR, Orfao, JJM, and Figueiredo, JL (2010).

Stability of a cryptomelane catalyst in the oxidation of toluene. Catalysis Today,

154(3-4), pp. 308-311.

[40] Luo, J, Zhang, Q, Huang, A, and Suib, SL (2000). Total oxidation of volatile

organic compounds with hydrophobic cryptomelane-type octahedral molecular

sieves. Microporous and Mesoporous Materials, 35-36, pp. 209-217.

[41] Sun, H., Chen, S., Wang, P., Quan, X. (2011). Catalytic oxidation of toluene

over manganese oxide octahedral molecular sieves (OMS-2) synthesized by

different methods. Chemical Engineering Journal, 178, pp. 191-196.

[42] http://www.eia.gov/todayinenergy/detail.cfm?id=1253, accessed on August 23,

2013.

[43] Krekeler, M.P.S., Barrett, H.A., Davis, R., Burnette, C., Doran T., Ferraro, A.

Meyer, A. (2012). An investigation of mass and brand diversity in a spent battery

recycling collection with an emphasis on spent alkaline batteries: Implications for

waste management and future policy concerns. Journal of Power Sources, 203, pp.

222-226.

[44] Prieto, O., del Arco, M., Rives, V. (2003). Structural evolution upon heating of

sol-gel prepared birnessites. Thermochimica Acta , 401, pp. 95-109.

[45] Ching, S and Roark, JL (1997). Sol-gel route to the tunneled manganese oxide

cryptomelane. Chemistry of Materials, 9, pp. 750-754.

[46] Frias, D., Nousir, S., Barrio I., Montes, M., Lopez, T., Centeno, MA., and

Odriozola, J.A. (2007), Synthesis and characterization of cryptomelane- and

birnessite-type oxides: Precursor effect. Materials Characterization, 58(8-9),

pp.776-781.

[47] Valente, J.S., Frias, D., Navarro, P., Montes, M, Delgao, J.J., Fregoso-Israel, E.,

and Torres-Garcia, E. (2008). Mangaese cryptomelane-type oxides: A thermo-

kinetic and morphological study. Applied Surface Science, 254, pp. 3006-3013.

[48] Ding, Y-s., Shen, X-f, Sithambaram, S., Gomez, S., Kumar, R., Crisostomo,

VMB, Suib, S.L., and Aindow, M. (2005). Synthesis and catalytic activity of

cryptomelane-type manganese dioxide nanomaterials produced by a novel

solvent-free method. Chemistry of Materials, 17, pp. 5382-5389.

[49] Li, Q., Luo, G., Li, J., and Xia, X. (2003). Preparation of ultrafine MnO2

powders by the solid state method reaction f KMnO4 with Mn(II) salts at room

temperature. Journal of Materials Processing Technology, 137, pp. 25-29.

[50] Tali, R. (2007). Determination of average oxidation state of Mn in ScMnO3 and

CaMnO3 by using iodometric titration. Damascus University Journal for Basic

Sciences, 23(1), pp. 9-19.

[51] Portchault, D., Cassaignon, S., Baudrin, E., and Jolivet, J-P. (2007).

Morphology control of cryptomelane type MnO2 nanowires by soft chemistry,

growth mechanisms in aqueous medium. Chemistry of Materials, 19, pp. 5410-

5417.

[52] Peiteado, M., Caballero, A.C., and Makovec, D. (2007). Diffusion and reactivity

of ZnO-MnOx system. Journal of Solid State Chemistry, 180, pp. 2459-2464.

[53] Spivey, J.J. (1987). Complete catalytic oxidation of volatile organics. Industrial

& Engineering Chemistry Research, 26, pp. 2165-2180.

[54] Li, H., Qi, G, Tana, Zhang, X., Huang, X., Le, W., and Shen, W. (2011). Low-

temperature oxidation of ethanol over a Mn0.6Ce0.4O2 mixed oxide. Applied

Catlaysis B: Environmental, 103, pp. 54-61.

[55] Popa, A, Sasca, V., and Halasz, J. (2008). Catalytic properties of molecular

sieves MCM41 type doped with heteropolyacids for ethanol oxidation. Applied

Surface Science, 255, pp. 1830-1835.

[56] Jiang, B-S., Chang, R., and Lin, Y-C. (2013). Industrial and Engineering

Chemistry Research, 52, pp/ 37-42.

[57] Ionescu, N.I., Jaeger, N.I., Plath, P.J., and Hornoiu, C. (2008). Activation

energy of ignition for catalytic oxidation of ethanol in oscillatory regime. Journal

of Thermal Analysis and Calorimetry, 91(2), pp. 381-384.

[58] Campesi, M.A., Mariani, N.J., Pramparo, M.C., Barbero, B. P., Cadus, L.E.,

Martinez, O.M., and Barreto, G.F. (2011). Combustion of volatile organic

compounds on a MuCu catalyst: A kinetic study. Catalysis Today, 176, pp. 225-

228.

[59] Zhang, W., Desikan, J.A., and Oyama, S.T. (1995). Effect of support in ethanol

oxidation on molybdenum oxide. Journal of Physical Chemistry, 99, pp. 14468-

14476.

[60] Ismagilov, I.Z., Michurin, E.M., Sukhova, O.B., Tsykoza, L.T., Matus, E.V.,

Kerzhentsev, M.A., Ismagilov, Z.R., Zagoruiko, A.N., Rebrov, E.V., deCroon,

M.H.J.M., and Schouten, J.C. (2008). Oxidation of organic compounds in a

microstructured catalytic reactor. Chemical Engineering Journal, 135S, pp. S57-

S65.

[61]Boronat, M., Corma, A., Illas, F., Radilla, J., Rodenas, T., and Sabater, M.J.

(2011). Mechanism of selective alcohol oxidation to aldehydes on gold catalysts:

Influence of surface roughness on reactivity. Journal of Catalysis, 278, pp. 50-58.

Table 1. Summary of cryptomelane surface areas, pore volumes, crystal sizes, and bulk densities of catalysts used in this study.

Cryptomelane

Sample

BET

Surface

Area

(m2/g)

1Pore Volume

(cm3/g)

(% pore volume

in macropores)

AOS Bulk

density

(g/mL)

2Crystal

size

(nm)

3Elemental Composition

O/Mn

Atom

ratio

K/Mn

Atom

ratio

Zn/Mn

Atom

ratio

Zn/Mn

Mass

ratio

Reflux 62.5 ± 5.4 0.47 ± 0.12

(80 ± 6%)

3.7 ± 0.1 0.37 29 2.21 0.15 -nd- -nd-

Zn-Reflux 47.9 ± 0.7 0.17 ± 0.01

(61 ± 4%)

3.6 ± 0.1 0.37 28 2.01 0.17 0.013 0.016

Milling 73.1 ± 2.2 0.26 ± 0.01

(-nd-)

3.6 ± 0.1 1.04 23 2.10 0.15 -nd- -nd-

Zn-Milling 52.8 ± 2.1 0.21 ± 0.02

(-nd-)

3.7 ± 0.0 1.18 25 2.19 0.16 0.025 0.030

Sol-gel

-350 C

-400 C

-500 C

-600 C

91.0 ± 6.6

69.0 ± 0.4

33.5 ± 1.1

24.4 ± 0.4

0.12 ± 0.002

(1.8 ± 0.2%)

0.10 ± 0.002

(4.9 ± 0.0%)

0.084 ± 0.001

(7.4 ± 2%)

0.124 ± 0.001

(64 ± 0.3%)

3.4 ± 0.7

3.3 ± 0.1

3.6 ± 0.1

3.6 ± 0.0

1.33

1.13

1.31

1.29

20

28

31

26

2.56

1.83

2.10

2.54

0.182

0.297

0.267

0.314

-nd-

-nd-

-nd-

-nd-

-nd-

-nd-

-nd-

-nd-

Battery-

derived

4.2 ± 0.8 0.022 ± 0.002

(47 ± 2%)

3.2 ± 0.3 1.61 23 2.58 0.14 0.027 0.032

1BJH adsorption cumulative volume of pores between 17 Å and 3,000 Å. (% of pore volume contained in macropores (>500 Å)

estimated from cumulative BJH adsorption curves) 2Based upon the Scherrer equation and XRD patterns

3Based upon SEM/EDS analyses

-nd- Zn was not detected in samples.

Table 2. Observed reaction rates at 175 C and activation energies calculated for each

cryptomelane sample

Cryptomelane

Sample

Observed Reaction Rate

(umoles/min/g cat)

@

(Inlet ethanol

Concentration)

Observed Reaction Rate

(umoles/min/m2 cat)

@

(Inlet ethanol

Concentration)

Activation

Energy

(kJ/mole)

Reflux

Zn-Reflux

32.0 (400 ppm)*

54.0 (660 ppm)

57.4 (700 ppm)

0.51 (400 ppm)*

0.87 (660 ppm)

1.20 (700 ppm)

66.6

66.0

Milling

Zn-Milling

35.0 (460 ppm)*

42.8 (670 ppm)

44.6 (620 ppm)

0.48 (460 ppm)*

0.59 (670 ppm)

0.84 (620 ppm)

67.1

64.9

Sol-Gel

-350 C

-400 C

-500 C

-600 C

31.9 (410 ppm)

33.3 (430 ppm)

59.5 (980 ppm)

31.4 (450 ppm)

30.9 (540 ppm)

80.9 (1040 ppm)

31.7 (470 ppm)

28.5 (450 ppm)*

43.8 (880 ppm)

32.5 (500 ppm)

27.3 (980 ppm)

0.35 (410 ppm)

0.37 (430 ppm)

0.65 (980 ppm)

0.46 (450 ppm)

0.45 (540 ppm)

1.17 (1040 ppm)

0.95 (470 ppm)

0.85 (450 ppm)*

1.31 (880 ppm)

1.33 (500 ppm)

1.12 (980 ppm)

60.3

59.0

61.8

61.4

Battery Waste

Derived

16.5 (370 ppm)

19.8 (360 ppm)

31.3 (430 ppm)*

19.7 (620 ppm)

25.6 (1400 ppm)

47.8 (2300 ppm)

3.93 (370 ppm)

4.71 (360 ppm)

7.45 (430 ppm)*

4.64 (620 ppm)

6.10 (1400 ppm)

10.9 (2300 ppm)

67.0

*Results of Long-term tests at 175 C. Reflects average observed reaction rate over 8 days.

Table 3. Summary of weight changes in cryptomelane samples with temperature from TGA

curves.

Cryptomelane

Sample

Weight Loss (% of initial mass) Temp < 200 C 200 C ≤ T ≤

500 C

500 C ≤ T ≤

700 C

Temp >700 C Total

200 C < T <

900 C

T < 900 C

Reflux 0.9 1.0 4.5 2.9 8.4 9.3

Zn-Reflux 0.8 0.8 4.5 2.9 8.2 9.0

Milling 1.9 1.1 4.0 3.2 8.3 10.2

Zn-Milling 1.6 1.3 3.2 3.7 7.8 9.4

Sol-gel

-350 C

-400 C

-500 C

-600 C

7.1

5.6

5.6

4.2

2.4

2.5

2.0

1.1

1.5

1.4

1.1

1.3

4.4

3.4

3.7

4.2

8.3

7.3

6.8

6.8

15.4

12.9

12.4

11.0

Battery-

derived

0.2 0.2 3.4 3.5 7.1 7.3

(a)

Vapor Generation

Ethanol

Air @ 100 ccpm

Furnace Sample Collection for GC analyses

~400 ppmvEthanol in air

50 mg catalyst in¼”OD ss tube reactor

(b)

Figure 1. Schematic of test systems. (a) Test apparatus for AOS determinations. (b) Catalytic

oxidation experimental test system.

y = -0.3733x + 0.5995R² = 0.9265

0

0.1

0.2

0.3

0.4

0.5

0 0.5 1 1.5 2

Pore

Vo

lum

e (c

m3

at S

TP/g

)

Bulk Density (g/cm3)

Figure 2. Relationship between pore volume in cryptomelane and bulk density.

Figure 3. Nitrogen adsorption and desorption curves for cryptomelane prepared by (a) Reflux;

(b) Milled; (c) Sol-Gel 500 C; and (d) Battery-waste derived cryptomelane. Insert: BJH

Adsorption dV/dD pore volume.

-18%

-16%

-14%

-12%

-10%

-8%

-6%

-4%

-2%

0%

0 100 200 300 400 500 600 700 800 900 1000

% W

eig

ht

Ch

ange

Temperature ( C )

600 C

500 C400 C

350 C

-14%

-12%

-10%

-8%

-6%

-4%

-2%

0%

0 100 200 300 400 500 600 700 800 900 1000

% W

eig

ht

Ch

ange

Temperature ( C )

Battery Waste

500 C

Reflux

Milled

-12%

-10%

-8%

-6%

-4%

-2%

0%

0 100 200 300 400 500 600 700 800 900 1000

% W

eig

ht

Ch

an

ge

Temperature ( C )

Reflux

Zn Reflux

Milled

Zn Milled

Figure 4. Thermal gravimetric analyses of the cryptomelane samples used in this study. (a) Sol-

gel samples; (b) Comparison of TGA curves for Reflux, Milled, Sol-gel calcined at 500 C, and

Battery-waste-derived cryptomelane; (c) Effect of Zn on TGA curves.

(a)

(b)

(c)

10 20 30 40 50 60

Re

lati

ve In

ten

sity

Bragg Angle (2θ)

350 ͦC

400 ͦC

500 ͦC

600 ͦC

[110]

[200]

[310]

[211]

[301]

[411]

[600]

Figure 5. XRD patterns of cryptomelane prepared by the sol-gel method, showing the effect of

calcination temperature.

10 20 30 40 50 60

Rel

ativ

e In

ten

sity

Bragg Angle (2θ)

[20

0]

[31

0]

[21

1]

[30

1]

[41

1]

[60

0]

[11

0]

Milling

Reflux

Sol-Gel 500 C

Battery Waste

Sol-Gel 400 C

Figure 6. XRD patterns of cryptomelane samples used in this study. (a) Comparison of

cryptomelane samples prepared by different methods; (b) comparison of reflux cryptomelane

with and without Zn dopant; (c) comparison of milling method cryptomelane with and without

Zn dopant.

(a)

(b)

(c)

Figure 7. SEM pictures to show effect of calcination temperature on sol-gel samples 50

C;

00 C 500 C 00 C

(a) (b)

(c) (d)

200 nm 200 nm

200 nm 200 nm

(a) Reflux

(b) Milling Method

(c) Sol-Gel 500 C

( d) Battery Derived

Figure 8. SEM analyses of cryptomelane samples used in this study at magnifications of

10,000X, 50,000X, and 125,000X.

1 um 200 nm 100 nm

1 um

1 um

1 um

200 nm

200 nm

200 nm

100 nm

100 nm

100 nm

Reflux Method

Milling Method

Figure 9. SEM analyses (125,000 X) that show effect of Zn on morphology. (a) Reflux sample;

(b) Zn Reflux sample; (c) Milled Sample; (d) Zn Milled sample.

(a) (b)

(c) (d)

100 nm 100 nm

100 nm 200 nm

Figure 10. Ethanol conversion over cryptomelane prepared by different methods. Air flow =

100 ccpm; 50 mg catalyst; Inlet ethanol concentation 500 ppm.

0

30

60

90

120

0 500 1000 1500 2000 2500

Eth

ano

l Oxi

dat

ion

Rat

e (u

mo

les/

min

/g c

atal

yst)

Inlet Ethanol Concentration (ppm)

Reflux Milled Sol-Gel 350 C

Sol-Gel 400 C Sol- Gel 500 C Sol-Gel 600 C

Battery Waste

Figure 11. Observed trend for apparent reaction rates at 175 C as inlet ethanol concentration

increases.

0%

20%

40%

60%

80%

100%

25 50 75 100 125 150 175 200

% C

on

ve

rsio

n o

f E

tha

no

l

Reaction Temperature ( C)

350 C

400 C

500 C

600 C

Quartz Wool

0%

20%

40%

60%

80%

100%

25 50 75 100 125 150 175 200

% C

on

vers

ion

of

Eth

an

ol

Reaction Temperature ( C)

350 C

400 C

500 C

600 C

Figure 12. Effects of calcination temperature on the apparent catalytic activity of sol-gel

cryptomelane samples. Inlet ethanol concentration (a) 500 ppm; (b) 1000 ppm. 100 ccpm air

flow; 50 mg catalyst.

(a)

(b)

Figure 13. Effect of Zn on apparent catalytic activity of cryptomelane. 100 ccpm air flow; 50

mg catalyst; Inlet ethanol concentration 700 ppm.

0

3

6

9

12

15

18

21

24

25 50 75 100 125 150 175 200 225 250

Flo

w R

ate

in E

fflu

ent

(um

oles

/min

)

Reaction Temperature (C )

Ethanol

CO2

Acetaldehyde

Figure 14. Catalytic oxidation of ethanol over battery-waste-derived cryptomelane at 2200 ppm

ethanol at inlet; 100 ccpm air flow; 100 mg catalyst. Solid triangles are CO2 flow rates as

measured by precipitation method. Open triangles are CO2 flow rates calculated assuming no

losses of carbon in test system.

Figure 15 Arrhenius’s equ tion: Plot of ln pp rent r te onst nt vs inverse solute

temperature. Slope = -(activation energy, E / gas constant, R). This data is for cryptomelane

made by the reflux method without Zn dopant.

Figure 16. Ethanol concentration in effluent of reactor at 175 C for up to 8 days of continuous

operation.

Highlights

• Cryptomelane was prepared by three different methods: reflux, sol-gel, and milling

• Post-consumer alkaline battery waste was used to prepare a milled sample • > 95% degradation of ethanol (inlet conc. < 500 ppm) at 200 C for all samples • The activation energies for ethanol oxidation over cryptomelane was ~65

kJ/mole • The catalyst activity of cryptomelane samples was stable for up to 8 days