ABSTRACT

GAS PHASE OXIDATION OF DIMETHYL SULFIDE BY DIOXIDE BASED CATALYSTS

by Sachin Kumar

In this study, a low temperature catalytic oxidation process was investigated for the oxidation of dimethyl sulfide using titania-based catalysts. TiO2 catalysts doped with vanadia were made using a wet incipient method and a flame synthesis method. The catalysts were characterized using XRD, Raman and BET surface area analysis to study the TiO2 phase transition as functions of calcination temperature and V/Ti mass ratio. A flow reactor was used to investigate the performance of the catalysts, and the exit gases were analyzed using gas chromatography. It was found that low concentrations of vanadia (V/Ti mass ratio ≤ 2%) inhibited phase transformation and sintering, which resulted in more activity per unit mass of the catalysts, and the catalysts having a V/Ti mass ratio of 2% were able to degrade dimethyl sulfide most efficiently.

GAS PHASE OXIDATION OF DIMETHYL SULFIDE BY TITANIUM DIOXIDE BASED CATALYSTS

A Thesis

Submitted to the

Faculty of Miami University

In partial fulfillment of

the requirements for the degree of

Master of Science

Department of Science and Engineering

by

Sachin Kumar

Miami University

Oxford, OHIO

2004

Advisor: Dr. Catherine Almquist

Reader: Prof. Michael H. Waller

Reader Dr. Martin D. Sikora

TABLE OF CONTENTS

Page

1.0 Introduction 1

1.1 Thermal 3

2.0 Research Goals and Objectives 5

2.1 Experimental Design 5

3.0 Experimental Methods 7

3.1 Catalyst Preparation 7

3.1.1 V2O5/TiO2 Catalysts 7

3.1.1.1 Wet Incipient method 7

3.1.1.2 Flame Synthesis Method 9

3.2 Catalyst Characterization 11

3.2.1 X-ray Diffraction Analysis 11

3.2.2 Raman Spectroscopy 15

3.2.3 BET Surface Area 15

3.3 Catalyst Performance 18

3.3.1 Thermal Catalysis 18

4.0 Results and Discussion 22

4.1 Thermal Catalysis with V2O5/TiO2 Catalysts 22

4.1.1 Catalysts Characterization 22

4.1.2 Catalyst Performance 29

4.1.3 Catalyst Performance of Flame Synthesis Catalysts 38

5.0 Summary and Recommendations for Future Work 40

ii 6.0 References 42

Appendices

Appendix A Article Published in Catalysis Today Journal

Other Appendices Available upon request

iii LIST OF TABLES

Page

Table 1. Experimental Design 6

Table 2. Catalysts selected for performance testing 29

iv LIST OF FIGURES

Page

Figure 1. Catalysts prepared using Wet Incipient Method 8

Figure 2. Flame-Synthesis Method 9

Figure 3. Bragg’s Law [6] 12

Figure 4. The THETA:THETA Goniometer [4] 13

Figure 5. Diffraction Pattern plot obtained at the different 2θ angles 14 Figure 6. SA 3100 COULTER Surface Area Analyzer [10] 17

Figure 7. Scheme of the Thermocatalytic Experimental Design 18

Figure 8. Lindberg/Blue Tube Furnace 19

Figure 9. HP 6890 Gas-Chromatograph 19

Figure 10. Diffusion of DMS in Air 20

Figure 11. Variation in Peak Area with DMS concentration 21

Figure 12. XRD spectra for pure Degussa P25 TiO2 as received 25

and after calcinations at 400, 500, 6000C in air

Figure 13. Fraction TiO2 in V2O5/TiO2 catalysts as functions of 26

V/Ti mass ratio and calcination temperatures

Figure 14. Particle size for the anatase fraction at different calcination temperatures 27

Figure 15. BET surface area reduction as functions of calcinations temperature 28

and V/Ti mass ratio

Figure 16. Raman spectra for catalysts calcined in air at 400 0C for 24 h 30

at varying V/Ti mass ratios

v Figure 17. Effect of V/Ti Ratio on DMS Degradation for Catalysts Calcined at 4000C 32

Figure 18. Effect of V/Ti Ratio on DMS Degradation for Catalysts Calcined at 5000C 32

Figure 19. Effect of V/Ti Ratio on DMS Degradation for Catalysts Calcined at 5500C 33

Figure 20. Effect of V/Ti Ratio on DMS Degradation for Catalysts Calcined at 6000C 33

Figure 21. Effect of Calcination Temperature on DMS Degradation at V/Ti=0% 36

Figure 22. Effect of Calcination Temperature on DMS Degradation at V/Ti=1% 36

Figure 23. Effect of Calcination Temperature on DMS Degradation at V/Ti=2% 37

Figure 24. Effect of Calcination Temperature on DMS Degradation at V/Ti=5% 37

Figure 25. Formation of V2O5 crystallites on the surface of the catalysts 38

Figure 26. DMS destruction for catalysts with different V/Ti ratios prepared 40

using Flame synthesis method

vi NOMENCLATURE

λ - Wavelength

θ - Angle

B - Anatase peak width at half

height height

D - Diameter

eV - Electron Volt

f - Fraction of TiO2

IA - Intensity of anatase reflection

IR - Intensity of reflection

kV - Kilo Volt

ms - milli Second

NH4VO3 - Metavanadate

TCD - Thermal Conductivity

Detector Detector

TiO2 - Titanium Dioxide

TRS - Total Reduced

VOC - Volatile Organic Compounds

V/Ti - Vanadia to Titania ratio

UV - Ultra Violet

XRD - X-Ray Diffraction

vii ACKNOWLEDGEMENT

I acknowledge the following people for their help and support.

Ž Dr. Catherine Almquist Ž Prof. Michael H Waller Ž Dr. Martin D Sikora Ž Rodney J Kolb Ž Students, Faculty and Staff of Paper Science

Ž XRD Spectral Analysis Ž John P Morton (Geology) Ž Dr. John Rakovan (Geology)

Ž RAMAN Analysis Ž Dr. Andre J Sommer (Chemistry & Biochemistry)

Ž BET Surface Area Analysis Ž Dr. James Allan Cox (Chemistry & Biochemistry) Ž Ms. Diep Vu Ca (Chemistry & Biochemistry)

Ž Collaborative Work Ž Dr. John L. Graham (UDRI, Dayton) Ž Dr. Sukh Sidhu (UDRI, Dayton)

viii 1.0 INTRODUCTION

Volatile organic compounds (VOC’s) along with and sulfur

oxides are the most important polluting gases emitted by manufacturing industries. The

predominant VOC’s emitted by the paper industry are , total reduced sulfur

(TRS) gases and chlorinated gases. VOC’s are compounds that contain and

some may contain , nitrogen and other elements, but specifically exclude ,

monoxide, , carbonic acid, and metallic carbides and carbonate

salts. VOC’s are responsible for urban smog and the reduction of air quality. Therefore

the United States Environmental Protection Agency (US EPA) has imposed regulations

for these emissions. Because of the current and future regulations concerning gaseous

emissions, there is an increasing interest in cost effective VOC abatement technologies. A

key challenge to destroy VOC’s is that they are often produced in small concentrations in

large volumes of air, and this makes the available control technologies expensive on the

basis of cost per mass of pollutant destroyed [1].

The VOC’s are commonly degraded in vent streams using controlled high-

temperature incineration, fume incineration and regenerative or recuperative thermal

oxidation. These processes, while effective, require very high temperatures, usually on

the order of 800ºC to 1,000ºC. Maintaining the necessary harsh conditions requires the

consumption of large amounts of fossil fuels (eg. methane) and to the formation of

undesirable end products such as carbon dioxide, oxides of nitrogen (NOX) and even hazardous organic compounds. Carbon dioxide is a green house gas, and the increase in carbon dioxide levels in the atmostphere over the past 100 years is regarded as a cause for the increase in the temperature of the ’s atmosphere. In addition, when the

1 concentration of the compounds to be removed is very low (traces), incineration is not

economically efficient. Thus for many applications, development of new oxidation

technologies would to energy savings and reduced air emissions.

An alternative approach to incineration is to utilize catalytic oxidation. Through

the proper selection of the catalyst media, high efficiency can be obtained at much lower

temperatures; conditions in the range of 200ºC -300ºC are typical. At these relatively

mild conditions equipment is easier to design and maintain, and capacity is higher.

Furthermore, catalysts can offer high selectivity for desirable end products and will not

produce NOX under typical operating conditions.

A significant challenge of utilizing catalysts for environmental compliance is the high cost of the precious metal catalysts used and their susceptibility to physical and chemical damage. Consequently, there is interest in developing alternative catalysts that are less expensive and more robust than the materials presently in wide use. One strategy to improve the activity of alternative catalysts is to manufacture them as nanostructured materials either in the form of or with nanoscale crystallites.

This thesis outlines a study where a low-temperature method of VOC abatement was investigated: thermal catalysis.

2 1.1 THERMAL CATALYSIS

Thermal catalysis is an alternative process to incineration for degrading VOC’s at

lower temperatures. Oxidation catalysts consisting of V2O5 deposited on TiO2 have drawn considerable attention due to their successful application in the selective catalytic reduction of NOX reactions. They have also been demonstrated as effective catalysts for

partial oxidation of organic compounds. The goal of this work is to assess whether this

catalyst system could be used to destroy dimethyl sulfide (DMS), one of the key -

causing pollutants by the pulp and paper industry.

From a scientific viewpoint, this catalytic system is an interesting example of the

strong interaction between the support (TiO2) and the active phase (V2O5). In particular, the spreading of (V), over the TiO2 support leads to a modification of the

chemical-physical peculiarities of TiO2 and to an enhancement of its catalytic properties

[2]. V/TiO2 (anatase) catalysts are generally obtained by depositing V species on

commercial anatase by impregnation techniques. The overall V content that can be

introduced is largely dependent on the surface area of the anatase. Segregated vanadia

(V2O5) crystallites are evidenced, once a nominal vanadia monolayer covering the anatase surface is exceeded. Other methods, including sol-gel chemistry, co-precipitation, and laser induced pyrolysis methods, have opened the possibility of generating small anatase particles containing highly accessible V species, resulting in apparently high reaction rates per gram of catalyst [2].

The TiO2 support of this catalyst system exists in three crystal structures - , anatase and rutile. Catalytic studies suggest that the anatase support results in a more active and selective catalysis than rutile [2]. Unfortunately the anatase form of TiO2

3 is metastable and transforms into the thermodynamically stable rutile form upon

calcination. It has been found that small anatase particles (specific surface areas typically

greater than 100 m2/g) transform into rutile more easily than larger particles [2]. Some controversies about the role of the TiO2 crystalline phase in determining the characteristic

of the active phase are found in the literature [2]. However it is generally agreed that

TiO2 anatase gives rise to superior active catalysts as compared to TiO2 rutile for thermal catalytic processes [2].

4 2.0 RESEARCH GOALS AND OBJECTIVES

The overall goal of this study is to demonstrate the feasibility of nanostructured

V2O5/TiO2 catalysts for low temperature oxidation of dimethyl sulfide (DMS), a malodorous pollutant emitted by the pulp and paper industry. The specific research goals are:

Thermal Catalysis

• To investigate the thermal stability of V2O5/TiO2 catalysts as a function

of V/Ti mass ratio

• To investigate the activity of the catalysts for the oxidation of DMS as a

function of the V/Ti mass loading and calcination temperature.

2.1 EXPERIMENTAL DESIGN

• Thermocatalytic experimental system

In experimental system, the destruction of DMS was studied because it is among the key odor-causing VOC’s emitted by the paper industry.

Table 1 outlines the experimental design used for this research:

5 Table 1. Experimental Design

Thermal Catalysis

Objectives Variables Constants Tools

Thermal stability • V/Ti mass ratio (0-10%) Catalyst synthesis Catalyst of catalyst • Calcination temperature • TiO2 support (Degussa Characterization

(400-6000C) P25) • BET (surface area)

• V precursor (NH4VO3) • XRD (crystal phase)

• Synthesis method (Wet • Raman (V species)

Incipient)

Catalytic • Catalyst (see above) • Mass of catalyst (50mg) • Tube furnace with

Activity • Reaction temperature • Concentration of DMS temperature

(~4000ppm) controller

• Flow rate (5ml/min) • Gas chromatograph

with TCD for DMS

analysis

Photocatalysis

Objectives Variables Constants Tools

Catalytic activity • Calcination temperature • Degussa P25 • Photoreactor

of TiO2 on beads • Borosilicate glass beads • Gas chromatograph

(2 mm OD) with TCD for DMS

analysis

6 3.0 EXPERIMENTAL METHODS

3.1 CATALYST PREPARATION

3.1.1 V2O5/TiO2CATALYSTS

3.1.1.1 WET INCIPIENT METHOD

The catalysts used for thermal catalysis (V2O5/TiO2) in this study were prepared

by the wet incipient method. Briefly, 60 mL deionized was added to 5 grams of

2 Degussa P25 TiO2 (surface area 50 m /g, anatase fraction ~ 80%, and anatase crystal size

~ 20nm) and stirred using a magnetic stirrer. A measured amount of ammonium

metavanadate (Fisher Scientific) was added to the slurry to achieve V/Ti mass ratios

ranging from 0% to 10%. The slurry was heated to ~ 700C and stirred constantly to evaporate the water until the slurry was a thick paste. This paste was dried in an oven at

0 100 C overnight. Once dry, the powder (V2O5/TiO2) was crushed using a mortar and pestle and separated into 1 gram aliquots. The aliquots were calcined at 400, 500, 550 and

6000C in air for 24 hours. After calcination, the powders were crushed again using a

mortar and pestle and transferred to a vial for storage and subsequent characterization.

From the visual inspection of the prepared catalysts, as shown in Figure 1 it was observed that as the V/Ti mass ratio increased from 0 to 10%, the catalysts changed color from to faint yellow to dark yellow and finally to brown.

7 Pure TiO2 0% V/Ti 0.1% V/Ti

0.5% V/Ti 1% V/Ti 2% V/Ti

5% V/Ti 10% V/Ti Pure V2O5

Figure 1. Catalysts prepared using Wet Incipient Method

8 3.1.1.2 FLAME SYNTHESIS METHOD

In the present work, an alternative method of catalysts preparation was also

investigated to verify the feasibility of a flame synthesis method for V/TiO2 catalysts.

This was performed to obtain preliminary data to assess the feasibility of performing this method of catalyst synthesis with the test system available in our laboratory.

In this method, the organometallic precursor, titanium tetra-isopropoxide (TTIP) and vanadium oxytriethoxide (VOTE), in an impinger were mixed and introduced as a mixed vapor to a flame as shown in Figure 2. The resulting V2O5/TiO2 catalyst, formed in the flame, was collected thermophoretically on a water-cooled plate. No calcination of catalysts was necessary. All catalysts prepared by this method were anatase TiO2.

The catalysts prepared using flame synthesis method were prepared by using

TTIP and vanadium oxytriethoxide in volumetric ratios of 0%, 2% and 5%. Based upon

the observations during sample collection and the color of the samples, the V/Ti mass

ratios in the catalysts were greater than 10%, and the samples themselves were not

uniform. No other analysis was done to confirm V/Ti mass ratio in these catalysts.

Out Cooling Water In

Oxygen Methane Mass Flow Mass Flow Controller Controller Clean Air

Mass Flow Collision Controller NBuebubblizleerr

Figure 2. Flame-Synthesis Method

9 The vapor pressures of TTIP and VOTE, are not equal and so it can be stated that the concentration of liquid mixture of the two chemicals and the resulting V/Ti catalyst changed with time. This can very well support our observation about the color change observed in the prepared catalysts. It was much lighter in color initially, but became very dark with time.

One of the most important observations regarding this feasibility study for the flame synthesis method is the catalyst collection system. The experiment schematic as shown in Figure 1 was kept inside a properly vented hood. During the synthesis, which went on for approximately 8 hours, for each catalyst, a lot of catalyst was lost to the hood due to the air flow inside the hood, and the catalyst particles were seen on surfaces inside the hood. The amount of catalyst collected was very small, after running the experiment for nearly 8 hours only 50 mg of catalyst was collected.

Due to the poor collection efficiency of the system it was not possible to do a proper analysis of the collected catalysts using ICP, DCP or AA, which are analytical techniques that could be used for determining the V/Ti mass ratios in the prepared catalysts. For carrying out the performance testing of the catalysts in thermal catalysis, 50 mg of the catalyst was used and so only 1 run could be done to test the performance of the catalyst.

Based on the observations and results of this work, recommendations were made to improve the feasibility of the method flame synthesis in our laboratory.

10 3.2 CATALYST CHARACTERIZATION

3.2.1 X-RAY DIFFRACTION ANALYSIS

The V2O5/TiO2 catalysts were characterized using X-Ray Diffraction

(XRD) to study the effect of V/Ti mass ratio on the phase conversion of TiO2 from anatase to rutile, the latter being less desirable for catalysis [2].

The X-ray diffraction pattern of material is like a fingerprint of the material. The powder diffraction method is ideally suitable for characterization and identification of polycrystalline phases of materials. When X-rays interact with a crystalline material, a diffraction pattern is achieved. Every crystalline material gives a pattern, and the same material always gives the same pattern each analysis in a mixture of materials. The mechanical assembly consists of a sample holder, detector arm and associated gearing, called a goniometer [4]. It works on the basis of Bragg’s law. It has been shown in

Figure 3.

Bragg’s law states that for X-rays scattered from a crystalline solid [5], n λ = 2d Sin θ (1)

Where, θ = angle of incidence of X-ray beam, degrees d = the distance between atomic layers in a crystal, Å

λ = the wavelength of the incident X-ray beam, Å n = an integer

11

Figure 3: Bragg’s Law [6] In the goniometer the distance between the X-ray focal spot and the sample is the same as

the distance between the sample and the detector. In a θ:θ goniometer, the sample is stationary in a horizontal position. The X-ray tube and the detector both move simultaneously over the angular range THETA. This is shown is Figure 4.

12

Figure 4: The THETA:THETA Goniometer [4]

X-rays have wavelengths on the order of a few angstroms, a typical distance between atoms in crystalline materials. X-ray diffraction occurs as X-rays interact with a regular structure of the catalyst whose distance is about the same as the wavelength. That means

X-rays can be diffracted from materials which, by definition, are crystalline and have regularly repeating atomic structures.

The spectra (instrument response (counts/s) vs 2theta) that are obtained from

XRD represent the diffracted X-rays, and they are characteristic of the crystal structure of the catalyst. The predominant peak for anatase TiO2 is at a 2θ of 25.5, and that for rutile

is 27.5.

For a given sample, the ratio (IR/IA) of the intensities of the strongest rutile

reflection (IR) to the intensity of the strongest anatase reflection (IA) is calculated. This

ratio is used in the formula [7] below to calculate the fraction (f) of TiO2 crystal structure that is anatase. f = 1/(1+1.26(IR/IA)) (2)

13 The anatase crystal size of the catalysts were calculated using the Scherrer

equation and the XRD spectras. The mean size (D) of the anatase crystallite in the

catalyst is determined using the following equation [8]:

D = k λ / B1/2 cos (θ) (3) Where,

k , is the constant ranging between 0.7 to 1.71

λ, is the wavelength of X-Ray [8-9], Å

θ, is the angle measured from the diffraction patterns. Note that the XRD gives the values

as intensity versus 2θ, degrees

B1/2, is measured from the diffraction pattern figure obtained at the different 2θ angles

(degrees), as it can be approximated as width of the line breadth at half the intensity,

which is labeled as B1/2, as shown in Figure 5.

B1/2 Intensity

24 25 2 Theta 26

Figure 5. Diffraction Pattern plot obtained at the different 2θ angles

14 3.2.2 RAMAN SPECTROSCOPY

Raman spectrometry is based on inelastic light scattering by . In the analytical technique, a laser is used to illuminate the process sample. The molecules in the sample vibrate or rotate at characteristic frequencies. The light scattered back from these molecules is shifted in frequency by amounts specific to each type of . The scattered light is collected from the process using a window probe or an insertion probe.

The light is then analyzed in a spectrographic detector and computer to determine the concentration of the components in the process. The effect is both qualitative and quantitative. The advantage of Raman spectrometry is that certain molecules can be measured with higher sensitivity and greater ease than with other analytical methods because of their characteristic stretching frequency.

In our system of catalysts, the low concentration of V2O5 could not be detected with XRD but they were observed qualitatively using Raman.

3.2.3 BET SURFACE AREA

The BET (Brunauer, Emmett and Teller) calculation was first introduced in 1938, and it is the most commonly used calculation method for the characterization of specific surface area. This calculation is based upon gas adsorption onto the catalyst, according to the multilayer adsorption theory.

The multilayer adsorption theory is based upon the assumption that the first layer of molecules adsorbed on the surface involves adsorbate-adsorbent energies, and subsequent layers of molecules adsorbed involve the energies of vaporization

(condensation) of the adsorbate-adsorbate interaction.

15 The BET equation [10] shown below produces a straight line plot between

volume adsorbed (VA) vs Relative pressure (PS/PO), the linear form of which is most

often represented as:

PS/VA(P0-PS) = [(C-1)/VMC] PS/P0 + 1/VMC (4)

Where, VM = volume of monolayer

VA = volume adsorbed (cc/g) (STP)

PS = sample pressure (Pa)

P0 = saturation pressure (Pa)

C = constant related to enthalpy of adsorption

Gas sorption is the most widely used and accurate method for the measurement

for total surface area within the approximate range of 10 to 200 m2/g. In this method gas molecules of known size are condensed on sample surface. The quantity of gas condensed and the resultant sample pressure are used for the calculation of specific surface area [10].

In this study, the specific surface area of each catalyst was measured using nitrogen adsorption at 77 K using a Beckman-Coulter SA3100 surface area analyzer shown below in Figure 6. Prior to analyses, the catalysts were degassed at 120ºC with a helium purge for 60 minutes. The adsorption process in the Coulter SA 3100 is measured volumetrically with a static fully equilibrated procedure. The isotherm volume data is calculated by subtracting the freespace of the sample tube, which is that volume of the sample tube not occupied by the sample, from the total volume of gas dosed to the sample. An incremental data set is formed. Each data point is processed by calculating

16 the volume adsorbed and measuring the sample pressure which is then divided by the saturation vapor pressure.

Figure 6: SA 3100 COULTER Surface Area Analyzer [10]

17 3.3 CATALYST PERFORMANCE

3.3.1 THERMAL CATALYSIS

The performance of the catalysts in a thermal catalytic process was assessed for

the destruction of DMS.

In our test system DMS was passed through a tube furnace, and the exit gases

were analyzed using a gas chromatograph. The schematic of thermal process is shown in

Figure 7.

In this system 5 ml/min air, controlled by a mass flow controller, was mixed with

DMS vapor, generated using a diffusion cell. This mixture was passed through ¼” OD alumina reaction tube where 0.05g of catalyst was secured between 2 plugs of quartz wool. The exit of the reaction tube was connected to a gas chromatograph (GC) (HP

6890), shown in Figure 9 that has a thermal conductivity detector (TCD). Samples of the reactor effluent were introduced to the GC by way of gas sampling valves. A tube furnace

(Figure 8) maintained the catalysts at the desired reaction temperatures of 100, 150, 200,

225, 250, 275 and 300 degrees Celsius.

Mass Flow Tube Furnace Controller

Diffusion GC Cell Air

Figure 7. Scheme of the Thermocatalytic Experimental Design

18

Figure 8: Lindberg/Blue Tube Furnace

Figure 9: HP 6890 Gas-Chromatograph

19 DMS vapors were generated using a diffusion cell. The diffusion cell shown in

Figure 10, consisted a glass vial connected to the influent air line by a 1/8” I.D. tube.

DMS diffusion from the liquid reservoir of DMS takes place through the 1/8” I.D. tube to

the air line, to generate a stable vapor concentration to the reactor. The rate of diffusion,

NA, of DMS was calculated using the equation [11],

NA = DAB P/RT (z2-z1) ln (P-pA2) / (P-pA1) (5)

2 Where, DAB is diffusivity of DMS (cm /s), obtained from CRC handbook for chemical

engineers [15].

NA is rate of diffusion, moles/sec

P is total pressure (Pa)

pA2 and pA1 are partial pressure of DMS (Pa)

pB2 and pB1 are partial pressure of air (Pa)

Air pA1

NA

pA1 DMS (A)

Figure 10: Diffusion of DMS in Air

The linear response to DMS concentration was verified by plotting the calculated

Peak area vs DMS concentration, shown in Figure 11.

20 1000 900 800 y = 0.2064x - 62.64 700 R2 = 0.9985

ea 600 r

A 500 ak

e 400 P 300 200 100 0 1000 1500 2000 2500 3000 3500 4000 4500 DMS conc. (ppm)

Figure 11. Variation in Peak Area (unitless) with DMS concentration

In addition, the system was checked for any possible leakage. The temperature of

DMS in the diffusion cell was kept constant by keeping it in a water bath at room temperature.

21 4.0 RESULTS AND DISCUSSION

4.1 THERMAL CATALYSIS WITH V2O5/TIO2 CATALYSTS

4.1.1 CATALYSTS CHARACTERIZATION

The prepared catalysts were characterized to investigate the effects of V/Ti mass

ratio and calcination temperature on the structure and properties of the catalysts.

Specifically the following changes in catalyst properties were investigated:

- Phase transformation from anatase to rutile

- Change in surface area

- Change in anatase crystal size

- Presence of V species on the surface of catalyst

It is desirable for catalyst activity to keep the anatase fraction relatively

high [12]. The fraction of anatase phase in the prepared catalysts were determined from

XRD spectra using the ratios of the predominant anatase and rutile peak intensities as

described in the previous section.

As shown in Figure 12, the diffraction patterns for Degussa P25 as received and

after calcination show that the nature of TiO2 changed from anatase to rutile at

calcination temperatures between 5000 C and 6000 C. This is apparent by comparing the predominant anatase and rutile peaks at 2θ values of 25.50 and 270, respectively. The

transformation of anatase to rutile is expected to deteriorate catalytic performance [12].

22 A R R A R A R

600 C Y IT 550 C NS INTE D 500 C RMALIZE

NO 400 C

As Received

20 25 30 35 40 45 50 55 2 THETA

Figure 12. XRD spectra for pure Degussa P25 TiO2 as received and after calcinations at 400, 500, 6000C in air, A = Anatase R = Rutile

Figure 13 shows the anatase fraction in the V2O5/TiO2 catalysts as functions of

both the V/Ti mass ratio and calcination temperature. The V/Ti mass ratio at calcination

temperatures up to 500 0C had little effect on the anatase fraction present in the catalysts.

At calcination temperatures of 550 0C and 600 0C, the presence of low concentrations of

V (V/Ti mass ratio ≤ 2%) inhibited the phase transformation of anatase to rutile TiO2, whereas at V/Ti mass ratios of 5% and 10%, the V appeared to promote the anatase to rutile phase transformation. This is suggested by the maxima anatase fraction at a V/Ti mass ratio of 2% in the 550 and 600 0C calcination temperature plots (Figure 13). These results support the conclusion presented by Balikdjian et al. [13] that a threshold amount

23 of surface V species is required to decrease the temperature at which phase transformation occurs.

0.9 0 0.8 400 C 5000C 0.7

0.6 ase at

n 0.5 0 0.4 550 C A

0.3 Fract 0.2

0.1 6000C 0 024681012 V/Ti mass ratio (%)

Figure 13. Fraction anatase TiO2 in V2O5/TiO2 catalysts as functions of V/Ti mass

ratio and calcination temperatures

Figure 14 shows the calculated crystal size of TiO2 at different calcinations temperatures. Two trends are observed in Figure 14:

1. Crystal size increases as the calcination temperature increases

2. Crystal size increases as V/Ti mass ratio increases, especially at 6000 C

24 48 10% 5%

43 ) 38

(nm 2% ze i 1%

s 33

le 0.5% ic t r 28 0.1% Pa 0% 23

18 400 450 500 550 600 Calcination Temperature (0C)

Figure 14. Particle size for the anatase fraction at different calcination temperatures

From the BET surface area analysis shown in Figure 15, we observe the reduction in BET surface area as functions of calcination temperature and V/Ti mass ratio. As shown, there is little effect of V/Ti mass ratios up to 2%, suggesting that the catalysts with low amount of V species on the surface sinter to the same extent as catalysts without V. However, at V/Ti mass ratios of 5% and 10%, the BET surface area reduction, or degree of sintering, increases with increasing V/Ti mass ratio. Besselmann et el. [14] observed that the V species on the surface of V2O5/TiO2 catalysts when present

in low concentrations are predominantly monomeric and polymeric species in strong

interaction with the TiO2 support. At high V concentrations, the V species on the surface

of the catalysts are predominantly polymeric and crystalline vanadia species. Considering

the BET surface area, XRD and Raman data in our catalysts, and the observations by

25 Besselmann et al. [14], monomeric and polymeric species in strong interaction with the

TiO2 may indeed inhibit phase transformation, whereas crystalline and polymeric vanadia at the surface of the catalysts (present in our catalysts with V/Ti mass ratios of 5% and

10%) promotes sintering and anatase to rutile phase transformation.

60 V/Ti = 0% V/Ti = 0.1% 50 V/Ti = 0.5% V/Ti = 1%

) g / 40 V/Ti = 2% 2 ^ V/Ti = 5% m

( V/Ti = 10% ea

r 30

ce A a

rf 20 u S 10

0 400 450 500 550 600 650 Calcination Temperature (C)

Figure 15. BET surface area reduction as functions of calcinations temperature and V/Ti mass ratio

For this research work we used Raman Spectroscopy as a tool that can be used to detect V species on TiO2. Raman allows us to characterize V groups, which are not detected by XRD at low concentrations. Figure 16 shows a summary of Raman spectra

0 of V2O5/TiO2 catalysts calcined at 400 C. V peaks were observed in catalysts with V/Ti mass ratios of 5% and 10% at approximately 995 cm-1, which are indicative of terminal

-1 V=O groups of bulk V2O5 [14]. No other peaks between 900 and 1100 cm were

26 observed. Because these groups are observed only when the V/Ti mass ratio is 5% and

10%, and because the enhanced degree of sintering and phase transformation is also observed when the V/Ti mass ratio is 5% and 10%, this supports our conclusion and that of Besselmann et al. [14] that crystalline vanadia groups on the surface of the TiO2 are responsible for the enhanced degree of sintering and phase transformation.

27

8

7

6

5

4

3 Instrument Response 2

1

0 700 750 800 850 900 950 1000 1050 1100 Frequency (cm^-1)

8 V/Ti = 0.1 7 V/Ti = 0.05 6 V/Ti = 0.02 5 V/Ti = 0.01

4 V/Ti = 0.005 3 Instrument Response TiO2

2

NH4VO3 1 V2O5 0 100 200 300 400 500 600 700 800 900 1000 1100 Frequency (cm^-1)

Figure 16. Raman spectra for catalysts calcined in air at 400 0C for 24 h at varying

V/Ti mass ratios

28 4.1.2 CATALYST PERFORMANCE

Preliminary catalyst performance testing was performed on all the catalysts using

the method described in thermal catalysis section. Based upon the preliminary testing,

only selected catalysts were subsequently used in performance testing as shown in

Table 2.

Table 2. Catalysts selected for performance testing

Calcination temperature (0C) V/Ti V/Ti V/Ti V/Ti

400 0% 1% 2% 5%

500 0% 1% 2% 5%

550 0% 2% 5%

600 0% 2% 5%

The performance results for catalysts with V/Ti mass ratios of 0%, 2% and 5%

were selected for presentation because the performance of catalysts having V/Ti mass

ratios between 0% and 2% were very close and differences in them were difficult to

distinguish. The performance results for the catalysts with V/Ti mass ratio of 10% were

not presented because in preliminary experiments, they did not perform as well as

catalysts having V/Ti mass ratios less than or equal to 5%.

The results for the destruction of DMS over the V2O5/TiO2 catalysts calcined at

400ºC, 5000C, 5500C and 600ºC are shown in Figures 17, 18, 19, 20 respectively.

29 120

100 ) (%

g 80 n i n i 0%_400C a 60 m 1%_400C Re

S 40 2%_400C

DM 5%_400C 20 Empty Tube Quartz Wool 0 100 150 200 250 300 Reaction Temperature (0C)

Figure 17. DMS Degradation vs Reaction temperature for Various V/Ti mass ratios

at calcination temperature of 4000 C

120

100 ) % ( 80

60 0%_500C 1%_500C 40 2%_500C 5%_500C DMS Remaining 20 Empty Tube Quartz Wool 0 100 150 200 250 300 Reaction Temperature (0C)

Figure 18. DMS Degradation vs Reaction temperature for Various V/Ti mass ratios

at calcination temperature of 5000 C

30

120

) 100 % ( 80

60 0%_550C 40 2%_550C 5%_550C DMS Remaining 20 Empty Tube 0 Quartz Wool 100 150 200 250 300 Reaction Temperature (0C)

Figure 19. DMS Degradation vs Reaction temperature for Various V/Ti mass ratios

at calcination temperature of 5500 C

120

100 ) % 80 ing ( in 60 ma

e 0%_600C

S R 40 2%_600C M

D 5%_600C 20 Empty Tube

0 Quartz Wool 100 150 200 250 300 Reaction Temperature (0C)

Figure 20. DMS Degradation vs Reaction temperature for Various V/Ti mass ratios

at calcination temperature of 6000 C

31 Figure 17, shows the effect of V/Ti mass ratios of the catalysts on the DMS

degradation for the catalysts calcined at 4000C. A few experimental runs were conducted

keeping the same controls but not using catalysts. In one control the alumina tube was

kept empty and in the other control the alumina tube was plugged using only quartz wool.

The results obtained using these control conditions were then compared to the results

obtained when catalysts were used. Results show that with catalysts there was ~ 20%

destruction of at 2000C, but more than 90% destruction of DMS at 2500C. It is observed,

as shown in Figure 17, that for some catalysts, there was an initial increase in the DMS

concentration between 1000C and 2000C. This can be explained by the fact that at temperatures less than 2000C there was some amount of DMS adsorbed on the catalysts,

and as the temperature increases DMS desorbs. The TCD detects this desorbed DMS, and

so we observe an increase in the peak area.

Figure 18 shows the catalytic performance of catalysts calcined at 5000C at varying V/Ti mass ratio. It was observed that greater than 90% DMS destruction occured for catalysts having 1% and 2% V/Ti mass ratio at 2500C. However, the performance of catalysts with V/Ti mass ratio of 5% degraded at the higher calcination temperature, and only 70% destruction was observed at 2500C.

In comparing Figures 17 and 18, the data also show that the calcining temperature between 400ºC and 500ºC had very little effect on catalyst performance except only the V/Ti mass ratio of 5% catalyst. Similar trends (fraction of DMS remaining vs reaction temperature) were observed for all the catalysts at calcining temperatures, 400ºC and 500ºC. The reason for the different catalytic activities of the catalysts with V/Ti mass ratios of 1% and 5%, when calcined at 500ºC appears to be due

32 to their difference in surface area. At 400ºC calcining temperature, the difference in

surface area between the catalysts with V/Ti mass ratios of 1% and 5% was small (48

m2/g vs 45 m2/g); this difference was much more prominent when the catalysts were calcined at 500ºC (38 m2/g vs 30 m2/g).

Figures 19 and 20 shows that the catalyst activity for the catalysts calcined at

5500C and 6000C. On comparing the fraction of DMS remaining in the system for the

catalysts having V/Ti mass ratios of 0%, 2% and 5% we observe that 2% V/Ti mass ratio

catalyst appears to be the most active catalyst and it is degrading the DMS most

efficiently among these three catalysts.

On comparing the catalysts activity based on the calcination at 5500C, we observe that around 80%, 20% and 50% DMS was still there up to 2500C for 0%, 2% and 5%

V/Ti mass ratios catalysts respectively. For catalysts having calcination temperature of

6000C. The DMS fractions remaining at 2500C were 76%, 19% and 66%. These values were 64%, 3% and 8% for catalysts having a calcination temperature of 4000C and 47%,

4% and 25% for catalysts having a calcination temperature of 5000C. This suggests that the catalytic activity is decreasing at the calcination temperatures greater than 5000C.

This can be explained using the XRD plots which indicates the conversion of anatase

0 phase of TiO2 to inactive rutile phase at calcination temperatures of more than 500 C.

33 120

100

80 ) (% g in in

a 60 m e R S M D 40 400C 500C

20 550C 600C

0 100 150 200 250 300

Reaction Temperature (0C)

Figure 21. DMS Degradation vs Reaction temperature for various Calcination

temperatures at V/Ti=0%

120

100 ) (%

g 80 n i n i

a 60 m

Re 40 S DM 20 400C 500C 0 100 150 200 250 300 Reaction Temperature (0C)

Figure 22. DMS Degradation vs Reaction temperature for various Calcination

temperatures at V/Ti=1%

34 120

100 )

% 80 ( g in in a 60 m e

S R 400C M

D 40 500C 20 550C 600C 0 100 150 200 250 300

Reaction Temperature (0C)

Figure 23. DMS Degradation vs Reaction temperature for various Calcination

temperatures at V/Ti=2%

120

100

80 ng (%) i n i

a 60 m Re S 40 400C DM 500C

20 550C 600C 0 100 150 200 250 300

Reaction Temperature (0C)

Figure 24. DMS Degradation vs Reaction temperature for various Calcination

temperatures at V/Ti=5%

Figures 21, 22, 23 and 24 shows the catalytic activity for catalysts having V/Ti mass ratios of 0%, 1%, 2% and 5% respectively, as functions of calcination temperature.

35 Figures 22 and 23 reveal that performance of catalyst with, 1% and 2% V are not

strongly affected by calcination temperature up to 5000C.

But Figure 24 for 5% V2O5/TiO2 clearly shows the negative effect of the increase in calcination temperature over the catalytic activity of the catalysts. The DMS fraction remaining at 2500C increases as the calcination temperature of the catalysts increases.

Possible reasons include the following.

1. At higher calcination temperature the anatase phase of TiO2 changes into

catalytically less active rutile phase.

2. The higher calcination temperature for the catalyst lowers the specific surface

area of the catalysts and so the activity per unit surface area decreases.

3. Formation of V2O5 crystallites on the surface of the catalysts is evident by the

Raman spectra as shown in Figure 25

36

Figure 25. Formation of V2O5 crystallites on the surface of the catalysts

37 4.1.3 CATALYST PERFORMANCE OF FLAME SYNTHESIS CATALYSTS

Catalytic performance testing of the catalysts prepared using flame synthesis method was done using the same experimental methods described for the wet incipient catalysts.

The fractions of DMS remaining at different reaction temperatures were plotted and are shown in Figure 26.

Flame synthesis method did not produced sufficient catalyst to conduct the characterization studies on the catalysts prepared. In order to distinguish the catalysts, it had been labeled as 0%, low and high vanandium catalysts, where 0% catalyst means there is no vanadium in the catalyst, low vanadium catalyst were made using 2% volumetric ratio of vanadium oxytriethoxide to titanium tetra-isopropoxide in impinger and high vanadium catalyst were made using 5% volumetric ratio of vanadium oxytriethoxide to titanium tetra-isopropoxide in impinger.

The catalysts without vanadium were white, catalyst with low vanadium were pale yellow and the catalyst with high vanadium were brown.

38 120

100 ) % ( 80 ining ma e

r 60 S M D

ion 40 t

c 0% a

Fr Low Vanadium 20 High Vanadium

0 100 150 200 250 300 Reaction Temperature (0C)

Figure 26. DMS destruction vs Reaction temperature for catalysts with different

V/Ti ratios prepared using flame synthesis method

From Figure 26, it can be seen that among the three catalysts having V/Ti ratios of 0%, low vanadium and high vanadium, the catalysts having low vanadium destroyed

100% of DMS at 2500 C, whereas the other two catalysts do not show total destruction at

the same temperature.

The catalyst having no vanadia doping was least active among the three catalysts,

which supports our wet incipient catalyst data. This result also suggests that putting

vanadia in excess hinders the catalytic activity of the catalyst. Since this was only a

feasibility study to investigate the catalyst synthesis method,only three catalysts were

made. The results of these three catalysts show a trend similar to that observed with wet

incipient catalysts. However, there is not enough data in this study to defensibly compare

the two synthesis methods.

39 5.0 SUMMARY AND RECOMMENDATIONS FOR FUTURE WORK

The feasibility of using nanostructured V2O5/TiO2 catalysts for the low temperature oxidation of DMS has been shown in this work. It was found that V2O5/TiO2 catalysts can be used to achieve the oxidation of DMS at moderate temperatures

(<3000C). Thermal stability of the catalysts were also investigated as a function of V/Ti

mass ratio, and it was found that the catalysts having up to 2% V/Ti mass ratio were most

stable against the phase reversal process during calcinations. This is supported by the

BET surface area analysis results, where it was seen that when the V/Ti mass ratio was

increased to 5% and 10% a greater degree of sintering was observed and the particle size

was also increased. The thermal stability of the catalysts can be summarized as there is

little effect of V/Ti mass ratios up to 2%, and the catalysts with low amount of V species

on ht e surface sinter to the same extent as the catalysts without V. However, at greater

V/Ti mass ratios (5% and 10%), the catalysts becomes thermally less stable.

Catalysts having V/Ti mass ratio of 2% was found to be catalytically most active.

The experimental data suggests that crystalline vanadia species on the surface of the catalysts are responsible for higher catalytic activity for the oxidation of DMS. Calcining temperatures between 400 and 5000 C had only a small effect on the catalyst activity.

Future work with this system should include a detailed analysis of the reaction

products and kinetics. This will help in identifying the types of compounds for which this

method of destruction is effective.

Based on the results obtained for the flame synthesis method it is recommended to

conduct future studies with the use of proper analytical techniques like Atomic

Adsorption Spectroscopy (AA), X-ray Fluorescence Spectroscopy (XRF), Direct Current

40 Plasma Spectroscopy (DCP) and Induced Coupled Plasma Spectroscopy (ICP) for the determination of the V/Ti mass ratios in the collected catalysts.

There is need for more intensive catalysis study using flame synthesis method.

More variations in the vanadia loading should be done in future. This will help in determining the optimum vanadia loading level for maximum activity of the catalysts.

Incorporation of a better catalyst synthesis and collection system is recommended in order to obtain more quantity of uniform catalysts in less time.

There is need for more intensive catalysis study using flame synthesis method.

More variations in the vanadia loading should be done in future. This will help in determining the optimum vanadia loading level for maximum activity of the catalysts.

Incorporation of a better catalyst synthesis and collection system is recommended in order to obtain a greater quantity of uniform catalysts in less time.

41 6.0 REFERENCES

[1] P.Avila, A.Bahamonde, J.Blanco, B.Sanchez, A.I.Cordona, M.Romero “Gas

phase photo-assisted mineralization of volatile organic compounds by monolithic

titania catalysts”: Applied Catalysis B: Environmental 17 (1998) 75-88

[2] J.P.Balikdjian, A.Davidson, S.Launay, H.Eckert and M.Che “Sintering and phase

transformation of V-loaded anatase materials containing bulk and surface V

species”: Journal of Physical Chemistry B (2000), 104, 8931-8939

[3] Catherine Almquist: Doctoral Dissertation, “The synthesis and characterization of

titanium dioxide photocatalysts and their performance in selected applications of

air and water remediation and organic synthesis”: University of Cincinnati, June

2001

[4] Basics of X-ray Diffraction, by Scintag Inc., www.scintag.com

[5] www.bmsc.washington.edu/people/merritt/bc530/bragg/

[6] www.matter.org.uk/diffraction/geometry/geometry_of_diffraction_braggs-

law_2.htm

[7] Robert A.Spurr and Howard Myers, Anal. Chem. 29 (1957) 760–762

[8] H.P.Klug and L.E.Alexander, X-Ray Diffraction procedures, John Wiley and

Sons Inc., New York, Chapter 9 (1954)

[9] P.Scherrer.Gottinger Nachrichten, (1918) 2, 98

[10] Product Manual: COULTER Series surface area and pore size analyzer, Coulter

Corporation, Miami, Florida

[11] Christie J.Geankoplis, Transport Process and Unit Operations, 2nd Edition, Allyn

and Bacon Inc., Principles of mass transfer, chapter 6, 378-379

42 [12] C.B.Rodella, P.A.P.Nascente, R.W.A.Franco, C.J.Magon, V.R.Mastelaro, and

A.O.Florentino “Surface characterization of V2O5/TiO2 catalytic system”: Phys.

Stat. Sol. (1) (2001) 187, No. 1, 161-169

[13] J.P. Balikdjian, A. Davidson, S. Launay, H. Eckert, M. Che,J. Phys. Chem. B 104

(2000) 8931–8939

[14] S. Besselmann, E. Loffler, M. Muhler, J. Molec, Catal. A Chem. 162 (2000) 401–

411

[15] CRC Handbook for Chemical Engineers

43 Catalysis Today 88 (2003) 73–82

An investigation of nanostructured vanadia/titania catalysts for the oxidation of monochlorobenzene John L. Graham a, Catherine Bothe Almquist b,∗, Sachin Kumar b, Sukh Sidhu a a Environmental Science and Engineering, University of Dayton, Research Institute, Dayton, OH 45469, USA b Department of Paper Science and Engineering, Miami University, Oxford, OH 45056, USA

Abstract

In this work, the effects of V/Ti mass ratios in nanostructured V2O5/TiO2 catalysts on catalyst characteristics and catalyst activity were investigated. The destruction of monochlorobenzene was used as a measure of catalyst activity. The V2O5/TiO2 catalysts, which have been traditionally used as de-NOx catalysts and selective oxidation catalysts (e.g. in the production of phthalic anhydride), were made by the wet incipient method. This synthesis method provides nanostructured V2O5/TiO2 catalysts with the vanadium species on the surface of the TiO2. The catalysts were characterized using X-ray diffraction (XRD), BET surface area, and Raman spectroscopy. The performance of the catalysts for the destruction of organic pollutants was assessed in a differential tube flow reactor for gas-phase thermal oxidation reactions. It was found that the presence of crystalline V2O5 on the surface of TiO2 (V/Ti mass ratio 0.05 and 0.1 ) is necessary for the oxidation of monochlorobenzene at temperatures <300 ◦C. © 2003 Elsevier B.V. All rights reserved.

Keywords: Vanadia/titania catalysts; Oxidation; Monochlorobenzene; Nanostructured

1. Introduction or even hazardous organic compounds. An alternative approach is to utilize catalytic oxidation. Through The oxidation of organic vapors is an important the proper selection of the catalyst media high effi- component of environmental compliance. This is of- ciency can be obtained at much lower temperatures, ◦ ten achieved through some form of thermal process conditions in the range of 200–300 C are typical. At such as controlled high-temperature incineration, these relatively mild conditions equipment is easier to fume incineration and regenerative or recuperative design and maintain, and capacity is higher. Further- thermal oxidation. These processes, while effective, more, catalysts can offer high selectivity for desirable require very high-temperatures, usually on the order end products and will not produce oxides of nitrogen of 800–1000 ◦C. Maintaining the necessary harsh con- under typical operating conditions. A significant chal- ditions may require the consumption of large amounts lenge of utilizing catalysts for environmental compli- of fossil fuels and may also lead to the formation of ance is the high cost of the precious metal catalysts undesirable end products such as oxides of nitrogen used and their susceptibility to physical and chemical damage. Consequently, there is interest in developing alternative catalysts that are less expensive and more ∗ Corresponding author. Tel.: +1-513-529-2203; robust than the materials presently in wide use. One fax: +1-513-529-2201. strategy to improve the activity of alternative catalysts E-mail address: [email protected] (C.B. Almquist). is to manufacture them as nanostructured materials

0920-5861/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.cattod.2003.08.008 74 J.L. Graham et al. / Catalysis Today 88 (2003) 73–82 either in the form of nanoparticles or with nanoscale support than rutile TiO2 [8]. The optimum vanadium crystallites. content has been reported to be as low as 4% [9] and Vanadia/titania (V2O5/TiO2)-based catalysts are as high as 30% [10], depending upon the catalyst syn- commonly used in selective catalytic reduction (SCR) thesis method and the reaction conditions. The ther- because of their high catalytic activity and ther- mal stability of the V2O5/TiO2 catalysts is dependent mal stability [1]. V2O5/TiO2-based catalysts also upon the catalysts synthesis method and how the se- exhibit resistance against SO2 poisoning, which is lective catalytic reduction (V) species is incorporated generally present in exhaust gases from stationary into or onto the TiO2. This has implications on the combustion sources [2]. V2O5/TiO2 based catalysts TiO2 crystal phase, and ultimately, the stability of the have also been recently used to control emissions of catalyst over time. chlorinated organics including chlorinated dioxins, The goals of this study were to investigate the ef- furans and [3–5]. The results of experi- fect of V/Ti mass ratio on TiO2 phase transformation ments with chlorinated volatile organic compounds from anatase to rutile, surface area reduction (degree (VOCs) such as chloride, dichloromethane, of sintering), and catalyst activity for the oxidation of ethyl chloride, dichloroethane, tetrachloroethene, monochlorobenzene (MCBz). MCBz was selected be- tetrachloromethane, and chlorobenzenes have been cause it is emitted from many combustion-based man- reported [3–6]. In these studies, the effective tem- ufacturing processes, cement kilns and incinerators. perature for the destruction of the chlorinated VOCs To accomplish our experimental goals, V2O5/TiO2 ◦ was found to be above 300 C. Only the TiO2-based catalysts were synthesized with a range of V/Ti mass V2O5/WO3 catalysts showed effective destruction of ratios from 0 to 0.1, and for each V/Ti ratio, the cata- tetrachloroethene at temperatures less than 300 ◦C lysts were calcined in air at temperatures from 400 to (230 ◦C). It becomes evident from results of recent 600 ◦C. The catalysts were characterized using X-ray studies that V2O5/TiO2-based catalysts with increased diffraction (XRD), BET surface area, and Raman oxidation potential due to higher vanadium content spectrometry. can be optimized to control both NOx and organic pollutant emissions [6]. However, a further increase in catalyst efficiency is needed for V2O5/TiO2-based 2. Experimental catalysts before they can meet emissions stan- dards for VOCs. Also there is a need to eliminate 2.1. Catalyst synthesis formation of higher chlorinated products, which have been observed in previous pollutant oxidation The catalysts used in this study were prepared by studies [7]. the wet incipient method. Briefly, 60 ml deionized wa- One approach to improving the efficiency of cat- ter was added to 5 g of Degussa P25 TiO2 (surface alytic materials is preparing them as nanoscale parti- area 50 m2/g, anatase fraction 80%, and anatase crys- cles (mean diameter <100 nm). It has been shown that tal size 20 nm [11]) and stirred using a magnetic stir- nanoparticles are often far more reactive than materials rer. A measured amount of ammonium metavanadate in the bulk phase. Hence, it is possible that nanopar- (Fisher Scientific) was added to the slurry to achieve ticle catalysts of V2O5/TiO2 will be able to reduce V/Ti mass ratios ranging from 0 to 0.1. The slurry ◦ NOx and oxidize organic pollutants with increased ef- was heated to approximately 70 C and stirred con- ficiencies so as to comply with existing and future en- stantly to evaporate the water until the slurry was a vironmental emission standards. There is also need to thick paste. This paste was dried in an oven at 103 ◦C address inconsistencies in the literature regarding the overnight. Once dry, the powder (V2O5/TiO2)was effects of catalyst properties and synthesis methods on crushed using a mortar and pestle and separated into V2O5/TiO2 performance. For example, there are con- 1 gm aliquots. The aliquots were calcined at 400, 500, ◦ troversies in the literature about the role of the TiO2 550 or 600 C in air for 24 h. After calcination, the crystalline phase in determining the characteristics of powders were crushed again using a mortar and pestle, the active phase of V2O5/TiO2 catalysts [8], although and transferred to a vial for storage until subsequent it is generally stated that anatase TiO2 is a more active characterization. J.L. Graham et al. / Catalysis Today 88 (2003) 73–82 75

2.2. Catalyst characterization diameter of two micrometers at the sample. Power at the sample did not exceed 6 mW. The same objective 2.2.1. X-ray diffraction was employed to collect the scattered radiation. Spec- −1 The TiO2 crystal phase of each catalyst was charac- tra were collected at 4 cm resolution and represent terized using a Scintag X-ray Powder Diffractometer the average of five individual scans. The integration (XRD) with a Cu K␣ source from 20◦ < 2θ<60◦ time for each spectral element was 30 s. V species at a scan rate of 1◦/min. The XRD was equipped with peaks that were not observed using XRD were ob- voltage and current stabilizers and a computer with served in samples with V/Ti mass ratios ≥ 0.05. necessary software to record the diffraction patterns. It was operated at a voltage of 40 kV with a filament 2.3. Catalyst performance current of 35 mA. The XRD data were used to study the effect of calcination temperature and V/Ti mass The overall activity of each catalyst for the oxi- ratio on the phase conversion of TiO2 from anatase to dation of monochlorobenzene (Aldrich, 99.99%) was rutile, the latter being less desirable for catalysis [8]. determined using a flow reactor system called the ther- mal stability analyzer (TSA). Fig. 1 shows the TSA 2.2.2. BET surface area as configured for this study. As this figure shows, the The surface area of each catalyst was mea- reactor assembly consists of a sample inlet system, a sured using nitrogen adsorption at 77 K using a quartz tubular flow reactor, and a total hydrocarbon Beckman-Coulter SA3100 surface area analyzer. Prior analyzer. The sample inlet system includes a syringe ◦ to analyses, the catalysts were degassed at 120 C pump to meter the sample into the inlet stream and a with a helium purge for 60 min. mixing tee. The reactor is a quartz tube in which 20 mg of catalyst is captured between a pair of quartz wool 2.2.3. Raman spectroscopy plugs and positioned at the mid-point of a single zone Raman spectra were collected using a Renishaw tube furnace. The total hydrocarbon analyzer consists 2000 confocal Raman microprobe. Samples were of a helium make-up inlet, a mixing tee, and a hydro- excited with a HeNe (632 nm) laser. This source was gen flame ionization detector (HFID). Note that the focused onto the sample using a 50× (0.85 N.A.) HFID is a very sensitive detector that responds linearly objective, which produced an approximate beam to organic vapors and does not respond to inorganic

Furnace w/ Catalyst Reactor HFID Quartz Wool Sampling Port

Inlet Air Catalyst Sample Pump Mass Flow Helium Make-up Controller Controller

Fig. 1. Schematic of the experimental test system with differential catalytic flow reactor (the thermal stability analyzer) for measuring catalyst activity. 76 J.L. Graham et al. / Catalysis Today 88 (2003) 73–82 gases. Consequently, the output of this detector is to make sure the bed was in the form of a dense packed linearly proportional to the concentration of organic layer securely held in place between the two quartz species in the reactor effluent. As configured for wool tufts. The reactors were also inspected after be- this study the method detection limit was 0.03 ppmv, ing removed from the furnace to make sure settling which is equivalent to a fraction remaining of 0.01%. did not occur during the tests. The finished bed mea- A stock sample of MCBz vapor was prepared by sured approximately 2 mm long by 4 mm in diameter. injecting 20 ␮l of liquid MCBz into a clean glass sam- A photograph of the powdered catalyst secured in the ple tube filled with 450 ml of dry air drawn from the reactor is shown in Fig. 2. The prepared reactor tube same compressed gas cylinder used to supply the reac- was then mounted in the tube furnace and the gas con- tor system and allowed to evaporate. From this stock nections made. Dry air was then supplied to the reactor sample 10 ml was drawn from the sample tube using a with mean flow rate of 64.8 ␮l/s and the reactor was gas-tight syringe. A fresh stock sample was prepared heated to the temperature at which the catalyst was for each test. calcined. The sample was then fed into the reactor in- For each analysis a reactor tube was assembled with let at a constant rate of 1.68 ␮l/s giving a mean initial approximately 20 mg of catalyst. The catalyst bed was concentration of 265 ppmv. The combined flow rate configured using a tight-packed powder design devel- of the inlet and sample gases gave a mean residence oped previously for measuring the activity of small time at 450 ◦C of 0.124 s, or a mean space velocity of fixed beds of fine powdered catalysts [12]. Briefly, a 29,000 h−1. Once the total hydrocarbon analyzer indi- tuft of quartz wool is positioned at the midpoint of the cated the reactor was at steady-state (i.e. the HFID out- reactor tube. The inside face of the tuft was tamped and put and system temperatures were stable for 10 min) shaped into a densified flat face. With the reactor tube the reactor was cooled at approximately 2 ◦C/min. held vertically the measured catalyst powder was then The raw data from the total hydrocarbon analyzer placed on this prepared face using a long stem glass was in the form of the HFID signal and tempera- pipette. An ultrasonic shaker was then applied to the ture recorded at 1 s intervals. The HFID signal was reactor tube to settle the powder into a compact layer converted to fraction remaining by first subtracting lying on the face of the tuft. A matching quartz wool the baseline signal recorded prior to feeding the sam- tuft was then inserted into the reactor tube and tamped ple to the reactor, then normalizing by the average in place. The catalyst bed was then inspected visually steady-state signal recorded at the end of the run.

Fig. 2. Photograph of the fine powder catalyst secured in quartz reactor tube. Dimensions of catalyst plug are 2 mm × 4 mm in diameter. J.L. Graham et al. / Catalysis Today 88 (2003) 73–82 77

3. Results and discussion rutile TiO2, whereas at V/Ti mass ratios of 0.05 and 0.1, the V appeared to promote the anatase to rutile 3.1. Catalyst characterization phase transformation. This is suggested by the max- ima anatase fraction at a V/Ti mass ratio of 0.02 in the ◦ The anatase TiO2 fraction in each prepared catalyst 550 and 600 C calcination temperature plots (Fig. 4). was determined using XRD spectra. The phase conver- These results support the conclusion presented by Ba- sion as a function of calcination temperature for V-free likdjian et al. [14] that a minimal amount of surface TiO2 is shown in Fig. 3. Without the addition of V, the V species is required to decrease the temperature at crystal phase of Degussa TiO2 converts from anatase which phase transformation occurs. to rutile at temperatures between 500 and 600 ◦C. Fig. 5 shows the reduction in BET surface area as The anatase fraction of TiO2 for each sample pre- functions of calcination temperature and V/Ti mass pared by the wet-incipient method was calculated us- ratio. As shown, there is little effect of V/Ti mass ra- ing XRD spectra and the equation by Spurr and Myers tio on BET surface area reduction at V/Ti mass ra- [13]: tios up to 0.02, suggesting that the catalysts with low amounts of V species on the surface sinter to the same 1 f = extent as catalysts without V. However, at V/Ti mass + . (I /I ) 1 1 265 R A ratios of 0.05 and 0.1, the BET surface area reduc- where f is the fraction anatase, IR is the intensity of tion, or degree of sintering, increases with increasing predominant rutile peak (2θ = 27.4), and IA is the V/Ti mass ratio. Besselmann et al. [15] observed that intensity of predominant anatase peak (2θ = 25.3). the V species on the surface of V2O5/TiO2 catalysts Fig. 4 shows the anatase fraction in the V2O5/TiO2 when present in low concentrations are predominantly catalysts as functions of both the V/Ti mass ratio and monomeric and polymeric species in strong interaction calcination temperature. The V/Ti mass ratio at calci- with the TiO2 support. At high V concentrations, the nation temperatures up to 500 ◦C had little effect on V species on the surface of the catalysts are predomi- the anatase fraction present in the catalysts. At cal- nantly polymeric and crystalline vanadia species. Con- cination temperatures of 550 and 600 ◦C, the pres- sidering the BET surface area, XRD, and Raman (next ence of low concentrations of V (V/Ti mass ratio ≤ section) data for our catalysts, and the observations 0.02) inhibited the phase transformation of anatase to by Besselmann et al. [15], monomeric and polymeric

ARRA A R 600 C

550 C

500 C

400 C NORMALIZED INTENSITY

As Received

20 25 30 35 40 45 50 2 THETA

Fig. 3. X-ray Diffraction spectra for Degussa P25 as received and after calcination in air for 24 h at temperatures of 400, 500, 550, and ◦ 600 C. A: anataseTiO2; R: rutile TiO2. 78 J.L. Graham et al. / Catalysis Today 88 (2003) 73–82

1

0.9

0.8

0.7

0.6

0.5

0.4 Fraction Anatase 0.3 400 C 500 C 0.2 550 C 0.1 600 C

0 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 V/Ti Mass Ratio

Fig. 4. Fraction anatase TiO2 in V2O5/TiO2 catalysts as functions of V/Ti mass ratio and calcination temperature.

60 V/Ti = 0 V/Ti = 0.001 50 V/Ti = 0.005 V/Ti = 0.01 40 V/Ti = 0.02 V/Ti = 0.05 30 V/Ti = 0.10

20 Surface Area (m^2/g)

10

0 400 450 500 550 600 650 Calcination Temperature (C)

Fig. 5. BET surface area reduction as functions of calcination temperature and V/Ti mass ratio.

species in strong interaction with the TiO2 may in- Fig. 6 shows a summary of Raman spectra of ◦ deed inhibit phase transformation, whereas crystalline V2O5/TiO2 catalysts calcined at 400 C. V peaks and polymeric vanadia at the surface of the catalysts were observed in catalysts with V/Ti mass ratios of (present in our catalysts with V/Ti mass ratios of 0.05 0.05 and 0.1 at approximately 995 cm−1, which are and 0.1) promote sintering and anatase to rutile phase indicative of terminal V=O groups of bulk V2O5 [15]. transformation. No other peaks between 900 and 1100 cm−1 were J.L. Graham et al. / Catalysis Today 88 (2003) 73–82 79

8

7

6

5

4

3 Instrument Response 2

1

0 700 750 800 850 900 950 1000 1050 1100 Frequency (cm^-1)

8 V/Ti = 0.1 7 V/Ti = 0.05 6 V/Ti = 0.02 5 V/Ti = 0.01

4 V/Ti = 0.005 3 Instrument Response TiO2

2

NH4VO3 1 V2O5 0 100 200 300 400 500 600 700 800 900 1000 1100 Frequency (cm^-1)

Fig. 6. Raman spectra of catalysts calcined in air at 400 ◦C for 24 h at varying V/Ti mass ratios. observed. Because these groups are observed only 3.2. Catalyst performance when the V/Ti mass ratio is 0.05 and 0.1, and because the enhanced degree of sintering and phase transfor- Catalyst performance was investigated only with mation is also observed when the V/Ti mass ratio is catalysts that were calcined at 400 and 500 ◦C. The 0.05 and 0.1, this supports our previous conclusion selection of these catalysts isolated the effects of V and that of Besselmann et al. [15] that crystalline loading on the performance of the catalysts, since the vanadia groups on the surface of the TiO2 are respon- fraction anatase in these catalysts were similar and sible for the enhanced degree of sintering and phase changes in surface area were relatively small. The re- transformation. sults for the destruction of MCBz over the V2O5/TiO2 80 J.L. Graham et al. / Catalysis Today 88 (2003) 73–82

100 0.0050

80 0.053 0.010

60 0.0010 0 0.11 40 Fraction Remaining, % 20

0 100 150 200 250 300 350 400 450 500 Temperature, C

Fig. 7. MCBz degradation in the differential catalytic flow reactor for catalysts calcined in air at 400 ◦C for 24 h. catalysts calcined at 400 and 500 ◦C are shown in that MCBz is degraded to inorganic by-products. Al- Figs. 7 and 8, respectively. though we did not specifically identify the inorganic The results presented in Figs. 7 and 8 for cata- by-products of MCBz in this study, they are likely to lysts containing V/Ti ratios of 0 and 0.001 show that be CO, CO2,H2O, and HCl, since the tests were con- MCBz is stable in the presence of the catalysts at tem- ducted under oxidative conditions. peratures up to 300 ◦C. At temperatures greater than The results for the catalysts with V/Ti ratios of 300 ◦C, MCBz, or more specifically, the total hydro- 0.005 and 0.01 show bimodal behavior, which is typ- carbon content in the reactor effluent, degrades, as in- ical of TSA data for which the feed material decom- dicated by the linear decrease in HFID response with poses to form organic by-products in significant yields. temperature. The decrease in HFID response indicates The HFID signal for these two catalysts indicates that

100 0 0.0050

80

0.11 60

0.010 40 0.053 Fraction Remaining, % 20

0 100 150 200 250 300 350 400 450 500 Temperature, C

Fig. 8. MCBz degradation in the differential catalytic flow reactor for catalysts calcined in air at 500 ◦C for 24 h. J.L. Graham et al. / Catalysis Today 88 (2003) 73–82 81

MCBz is stable up to 200 ◦C. The decrease in HFID mass ratios of 0.05 and 0.1 when calcined at 500 ◦C signal for these catalysts at temperatures between 200 is due to their difference in surface area. At 400 ◦C and 300 ◦C indicates that MCBz starts to degrade to calcining temperature, the difference in surface area either organic by-products that do not desorb readily between the catalysts with V/Ti mass ratios of 0.05 from the catalyst or to a combination of organic and and 0.1 was small (48 m2/g versus 45 m2/g); this dif- inorganic by-products. The stabilized HFID signal ob- ference was much more prominent when the catalysts served between 300 ◦C and approximately 400 ◦C sug- were calcined at 500 ◦C (38 m2/g versus 30 m2/g). gests that MCBz degrades to form organic by-products Fig. 9 shows the temperature at which 10% conver- that are stable in the presence of the catalysts in this sion of MCBz is achieved (90% remaining) as a func- range of temperatures. Finally, the decrease in HFID tion of V/Ti mass ratio for catalysts calcined at 400 signal at temperatures above 400 ◦C indicates further and 500 ◦C. The results show that the overall catalyst degradation of MCBz and its by-products to inorganic activity increased with V/Ti mass ratios from 0.00 to compounds that are not detected by the HFID. 0.05. The activities of the catalysts with V/Ti mass The results for the catalysts with V/Ti mass ratios ratios of 0.05 and 0.1 were similar. of 0.05 and 0.1 show higher activity than those stud- Only those catalysts with V/Ti mass ratios of ied with lower V/Ti mass ratios. The constant decrease 0.05 and 0.1 achieved oxidation of MCBz and the in the HFID response at temperatures greater than organic by-products of MCBz oxidation at temper- 150 ◦C shows effective degradation of MCBz and the atures <250 ◦C. The TSA data show that organic organic by-products of MCBz oxidation to inorganic by-products of MCBz are formed as the V/Ti mass compounds that are not detected by the HFID. ratio increases up to 0.01, but these by-products ox- In comparing Figs. 7 and 8, the data also show that idize further when the V/Ti mass ratio is increased the calcining temperature between 400 and 500 ◦C to 0.05 and 0.1 (Figs. 7 and 8). This behavior clearly had very little effect on catalyst performance. Simi- suggests a significant increase in catalyst activity for lar trends (fraction remaining versus reaction temper- MCBz oxidation as the V/Ti mass ratio is increased ature) were observed for all the catalysts at calcining to 0.05 and 0.1. temperatures, 400 and 500 ◦C. The reason for the ap- According to Besselmann et al., crystalline and parently different activity of the catalysts with V/Ti polymeric vanadia at the surface of the catalysts,

350

Calcined at 400 C Calcined at 500 C 300 Mean Trend Line

250

200

150 Temperature For 10% C Conversion, Temperature For

100 0 0.02 0.04 0.06 0.08 0.1 0.12 V/Ti Ratio

Fig. 9. The temperature for 10% conversion as a function of V/Ti mass ratio and calcination temperature (◦C). 82 J.L. Graham et al. / Catalysis Today 88 (2003) 73–82 which were observed by Raman spectroscopy in our activity. Future work with this system should include a catalysts with a V/Ti mass ratio ≥ 0.05, are able to detailed analysis of the reaction products and kinetics. adsorb gas-phase oxygen and insert it readily into the adsorbed organic species [15]. Contrarily, the monomeric vanadyl species, assumed to be predom- References inant on our catalysts with V/Ti mass ratios < 0.05, function mainly as oxidative adsorption sites for [1] S. Morikawa, K. Takahashi, J. Mogi, S. Kurita, Bull. Chem. Soc. Jpn. 55 (1982) 2254–2257. aromatic hydrocarbons [15]. Hence, we observed [2] H. Bosch, F.J.J.G. Janssen, Catal. Today 2 (1988) 369–521. oxidation of MCBz and by-products of MCBz oxi- [3] H. Hagenmaier, K.-H. Tichaczek, H. Brunner, G. Mittelbach, ◦ dation at temperatures <250 C when crystalline and Organohalogen Comp. 3 (1990) 65–68. polymeric vanadia sites predominate on the surface. [4] Y. Ide, K. Kashiwabara, S. Okada, T. Mori, M. Hara, Stable organic by-products of MCBz oxidation were Chemosphere 32 (1996) 189–198. < ◦ [5] B. Ramachandran, H.L. Green, S. Chatterjee, Appl. Catal. B observed at temperatures 400 C when monomeric 8 (1996) 157–182. V species predominate on the catalyst surface. With [6] S. Krishnamoorthy, J.P. Baker, A. Amiridis, Catal. Today 40 trace or no V species on the surface of TiO2, oxida- (1998) 39–46. tion of MCBz and its by-products was observed at [7] R.M. Lago, M.L.H. Green, S.C. Zang, M. Odlya, Appl. Catal. temperatures >300 ◦C. B 8 (1996) 107–121. [8] C.B. Rodella, P.A.P. Nascente, R.W.A. Franco, C.J. Magon, V.R. Mastelaro, A.O. Florentino, Phys. Stat. Sol. 187 (2001) 161–169. 4. Conclusions [9] C. Hoang-Van, O. Zegaoui, P. Pichat, J. Non-Cryst. Sol. 225 (1998) 157–162. This investigation shows that V O /TiO catalysts [10] M. Schneider, M. Maciejewski, S. Tschudin, A. Wokaun, A. 2 5 2 Baiker, J. Catal. 149 (1994) 326–343. can be used to achieve the oxidation of MCBz at mod- [11] M. Nargiello, T. Hertz, in: D.F. Ollis, H. Al-Ekabi (Eds.), ◦ erate temperatures (<300 C). The data also show that Photocatalytic Purification and Treatment of Water and Air, there is a significant increase in the overall activity Elsevier, Amsterdam, 1993, p. 67. for MCBz oxidation as the V/Ti mass ratio increased [12] Z. Xu, K. Fritsky, J. Graham, B. Dellinger, Organic halogen to 0.05. In addition, our data support that crystalline compounds 45 (2000) 419. [13] R.A. Spurr, H. Myers, Anal. Chem. 29 (1957) 760–762. vanadia species on the surface of the catalyst are re- [14] J.P. Balikdjian, A. Davidson, S. Launay, H. Eckert, M. Che, sponsible for higher catalytic activity for the oxida- J. Phys. Chem. B 104 (2000) 8931–8939. tion of MCBz. Calcining temperatures between 400 [15] S. Besselmann, E. Loffler, M. Muhler, J. Molec, Catal. A and 500 ◦C had only a small effect on the catalyst Chem. 162 (2000) 401–411.