USF Libraries
USF Digital Collections

Catalytic oxidation of methane using single crystal silicon carbide

MISSING IMAGE

Material Information

Title:
Catalytic oxidation of methane using single crystal silicon carbide
Physical Description:
Book
Language:
English
Creator:
Gopalkrishna, Akshoy
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Silicon carbide   ( lcsh )
Methane -- Oxidation   ( lcsh )
methane oxidation
silicon carbide
Dissertations, Academic -- Chemical Engineering -- Masters -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: SiC is a hard man-made material and has emerged as an excellent material for a wide range of applications which are exposed to extreme conditions such as high temperatures and harsh chemical environments. These applications range from SiC being used as an abrasive, to a refractory material, to a semiconductor material for high power and high frequency electronic devices. The properties of the material for each application is different, with the semiconductor grade material for electronic devices being the most refined. SiC, with its excellent thermal properties and high resistance to harsh chemical environments, lends itself to being an ideal support for catalyst systems. Various characterisation & analysis techniques such as Scanning Electron Microscopy (SEM), Energy Dispersive Spectroscopy (EDS), X-ray Diffraction (XRD), Fourier Transform Infrared Spectroscopy (FTIR) and Gas Chromatography (GC) are used in this thesis to investigate the suitability of single crystal SiC for high temperature catalytic systems. Low temperature oxidation of methane was used to investigate the catalytic activity of: - Porous and standard 4H-SiC with and without Pd - Porous and Standard 6H-SiC with and without Pd. - Nanocrystalline Beta-SiC powder with and without Pd. Part of the samples were impregnated with Pd using Palladium Nitrate (Pd (NO3)2) which is a common precursor for Pd. Activation treatments which were investigated were oxidation and reduction. Oxidation was generally better in activating the catalyst, as was expected, since the PdO phase is known to be more active in oxidising methane. A mixed set of Pd and PdO were observed by SEM and EDS which were the main characterisation techniques used to analyze the structure of the catalysts before and after the reaction. The Beta-SiC showed by far the best activity which could be attributed to the micro-crystalline powder format in which it was used, where as all other catalysts studied here were derived from crushed wafer pieces. Type II porous 4H-SiC was another of the samples which registered impressive results, vis-à-vis catalytic activity.
Thesis:
Thesis (M.Ch.E.)--University of South Florida, 2003.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Akshoy Gopalkrishna.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 70 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 52443436
notis - AJL4048
usfldc doi - E14-SFE0000105
usfldc handle - e14.105
System ID:
SFS0024801:00001


This item is only available as the following downloads:


Full Text

PAGE 1

CATALYTIC OXIDATION OF METHANE USING SINGLE CRYSTAL SILICON CARBIDE by AKSHOY GOPALKRISHNA A thesis submitted in partial fulfillment of the requirements for the degree of Master of Chemical Engineering Department of C h emical Engineering College of Engineering University of South Florida Co Major Professor: John T. Wolan Ph.D. Co Major Professor: Stephen E. Saddow Ph.D. Member: Scott W. Campbell, Ph .D. Date of Approval: April 7 2003 Keywords: Pd and PdO catalysts, Oxidation of streams with lean methane concentration, Semiconductor catalyst supports, SiC catalyst support Single crystal SiC. Copyright 2003 Akshoy Gopalkrishna

PAGE 2

ACKNOWLEDGEMENTS This work was supported by a Defense University Research Initiative on Nanotechnology (DURINT) program administered by the Office of Naval Research under Grant N00014 01 1 0715 (P rogram M onitor : C. Wood). Partial funding was pro vided by the National Science Foundation Grant ECS 0225697 (Program D irector : Dr. J. Momoh). In addition we would like to thank Dr. R. Koshla of the National Science Foundation for his encouragement and interest in this work. I would also like to thank my committee members, Dr. John T. Wolan, Dr. Stephen E. Saddow and Dr. Scott W Campbell for their support and guidance throughout the project.

PAGE 3

i TABLE OF CONTENTS LI ST OF TABLES iv LIST OF FIGURES v ABSTRACT vii CHAPTER 1: INTRODUCTION 1 1.1 Introduction to Catalysis 1 1.2 Silicon Carbide Material Prop e rties 2 1.3 Silicon Carbide in Catalysis 4 1.4 Thesis Outline 6 CHAPTER 2: HETEROGENEOUS CATALYSIS AND METHANE OXIDATION A LITERATURE REVIEW 9 2.1 Catalysis Involving Metals 10 2.1.1 Initiating Reactions 10 2.1.2 Stabilising the Reaction Intermediates 10 2.1.3 Holding the React ants in the Right Configuration 11 2.2 Cycles of Catalysis 11 2.2.1 Adsorption 11 2.2.2 Surface Reactions 12 2.2.2.1 Activation Barr ier Considerations 12 2.2.2.2 Mechanisms of Surface Reactions 14

PAGE 4

ii 2.2.3 Desorption 15 2.3 Sites for Adsorption 16 2.4 Mechanisms of Processes in Methane Oxidation A Review 17 2.4.1 Methane Oxidatio n Test Reaction 18 2.4.2 Activati on Treatment Reduction of PdO 19 2.4.3 Reaction after Reduction of PdO 20 2.4.4 Oxidation of Deposited Pd 21 2.4.5 Reactio n over Oxidised Pd 23 CHAPTER 3: CATALYTIC REACTOR DEVELOPMENT 25 3.1 Experimental Test Bed 25 3.2 Ca talyst Preparation 27 3.3 Test Rea ction Conditions 28 3.4 Instrumentation 29 3.5 Sample Calculation for Conversion 30 CHAPTER 4: CATALYTIC REACTOR EXPERIMENTS 33 4.1 Substrates Used 33 4.2 Beta SiC vs. Polycrystalline 3C SiC an XRD Comparison 34 4.3 S EM Analyses Before the Reaction 35 4.4 Reaction Data 36 4.4.1 Under Reduction Regime 36 4.4.2 Under Oxidation Regime 39 4.5 Reaction Data For Beta SiC 42 4.6 Reaction Data for Type II Porous 4H SiC 42 4.7 Reaction Data for Standard and Porous 4H SiC 43

PAGE 5

iii 4.8 Reaction Data of Porous and Standard 6H SiC 44 CHAPTER 5: SUMMARY AND FUTURE WORK. 4 6 5.1 Summary 46 5 .2 Future Work 47 REFERENCES 49 APPENDI CES 52 Appendix : A 53 A ppendix : B 60

PAGE 6

LIST OF TABLES Table 3.1 Details of flow controllers connected for the respective gases 26 Table 3.2 FTIR peak analysis 30 Table 3.3 Analysis of FTIR data 31 Table 4.1 Loading of Pd (in wt %) on the substrate from EDS 35 Table B.1 Components used in the catalytic reactor test bed 60 iv

PAGE 7

LIST OF FIGURES Figure 1.1 Stacking sequence in 4H, 3C and 6H polytypes of SiC 3 Figure 2.1 Lennard-Jones model of (a) pure molecular adsorption, (b) activated dissociative adsorption and (c) unactivated dissociative 13 Figure 2.2 Mechanisms of surface reactions 15 Figure 2.3 Sites for adsorption 16 Figure 2.4 Coverage of adsorption sites 17 Figure 2.5 Optimized structures for critical points of the O H bond formation reaction: (a) CH 3 + H on Pd fragment; b) intermediate activation; c) transition state; d) adsorbed CH 3 + OH (Bond distances are given in ) 20 Figure 2.6 Optimized structures of; a) collinearly adsorbed methane, b) bridging adsorption, and c) the dissociated product (Bond distances given in ) 24 Figure 3.1 Block diagram of the experimental test bed 26 Figure 3.2 Snapshot of the catalytic reactor test bed 27 Figure 3.3 Sample spectrum of type II 4H-SiC/Pd under reaction conditions at 300 C 30 Figure 3.4 A typical background spectrum 31 Figure 3.5 Absorbance spectrum obtained from the ratio of the sample spectrum and the background for a Type II 4H-SiC/Pd under Reaction conditions at 300 C 32 Figure 4.1 XRD spectra of substrate 34 Figure 4.2 Reaction results from the GC 36 Figure 4.3 Conversions of various substrates under a reduction regime 37 v

PAGE 8

Figure 4.4 Reaction results of porous sic and beta SiC impregnated with Pd after reduction (activation treatment) 38 Figure 4.5 Reaction results after oxidation (activation treatment) 40 Figure 4.6 Reaction results from porous SiC substrates with pd after oxidation (activation treatment) 41 Figure 4.7 Reaction results of beta SiC 42 Figure 4.8 Reaction results of type II porous 4H-SiC 43 Figure 4.9 Reaction results of standard and porous 4H-SiC 44 Figure 4.10 Reactions results of 6H-SiC 45 Figure A.1 SEM and EDS images of 6H-SiC 53 Figure A.2 SEM and EDS images of 6H-PSC 54 Figure A.3 SEM and EDS images of type II 4H-PSC 55 Figure A.4 SEM and EDS images of 4H-PSC 56 Figure A.5 SEM and EDS images of 4H-SiC 58 vi

PAGE 9

Catalytic Oxidation of Methane using Single Crystal Silicon Carbide Akshoy Gopalkrishna ABSTRACT SiC is a hard man-made material and has emerged as an excellent material for a wide range of applications which are exposed to extreme conditions such as high temperatures and harsh chemical environments. These applications range from SiC being used as an abrasive, to a refractory material, to a semiconductor material for high power and high frequency electronic devices. The properties of the material for each application is different, with the semiconductor grade material for electronic devices being the most refined. SiC, with its excellent thermal properties and high resistance to harsh chemical environments, lends itself to being an ideal support for catalyst systems. Various characterisation & analysis techniques such as Scanning Electron Microscopy (SEM), Energy Dispersive Spectroscopy (EDS), X-ray Diffraction (XRD), Fourier Transform Infrared Spectroscopy (FTIR) and Gas Chromatography (GC) are used in this thesis to investigate the suitability of single crystal SiC for high temperature catalytic systems. Low temperature oxidation of methane was used to investigate the catalytic activity of: Porous and standard 4H-SiC with and without Pd Porous and Standard 6H-SiC with and without Pd. Nanocrystalline Beta-SiC powder with and without Pd. vii

PAGE 10

Part of the samples were impregnated with Pd using Palladium Nitrate (Pd (NO 3 ) 2 ) which is a common precursor for Pd. Activation treatments which were investigated were oxidation and reduction. Oxidation was generally better in activating the catalyst, as was expected, since the PdO phase is known to be more active in oxidising methane. A mixed set of Pd and PdO were observed by SEM and EDS which were the main characterisation techniques used to analyze the structure of the catalysts before and after the reaction. The Beta-SiC showed by far the best activity which could be attributed to the micro-crystalline powder format in which it was used, where as all other catalysts studied here were derived from crushed wafer pieces. Type II porous 4H-SiC was another of the samples which registered impressive results, vis--vis catalytic activity. viii

PAGE 11

1 CHAPTER 1: INTRODUCTION 1.1 Introduction to Catalysis Catalysis as a phenomenon has been under investigation since Berzelius described it in 1836 [1] However to date the phenomenon is not quantitatively well understood; there are many current hypotheses on various aspects of catalysis that are generally accepted. Through the decades, many assumptions and theories have been rejected as improvements in analytic al characterization and computer aided simulation technologies usher in newer and more refined views of the older hypotheses. This thesis attempts to apply some of the currently accepted hypotheses on various aspects of catalysis in literature in order to explore the possibilities of using silicon carbide (SiC) along with promoters as an effective catalytic system. Catalysts by general definition are materials that increase the rate of the desired reaction by lowering the activation energy required for the reaction to take place without themselves being used up in the process. Catalysis has been widely used in the industry for varying purposes from increasing selectivity or yield of the desired product to increasing the rate of reaction. Catalysts can be b roadly classified either as homogenous catalysts or as heterogeneous catalysts. Of the two, the latter is more significant and favoured industrially. Homogenous catalysts generally dissolve in the reactant solution and enhance the rate of

PAGE 12

2 reaction by actin g uniformly throughout the reactant mixture. Due to the path of action taken by homogeneous catalysts, they have to be either separated from the product stream in order to be recovered or lost with the product, thus increasing the cost of the process. Hete rogeneous catalysts on the other hand do not dissolve into the reacting solution; instead they facilitate the reaction on its surface by providing an advantageous electronic and/or structural environment for the desired reaction to occur. The catalytic sy stem under investigation in this thesis belongs to the heterogeneous class of catalysts. SiC was the first man made abrasive substance. Invented by Dr. Edward Goodrich Acheson more than 100 years ago [2] ; it was given the trade name Carborundum and was hard enough to cut glass. An extremely rare, naturally occurring mineral form of SiC is found in meteorites and is called Moissanite, named after the scientist who first cla ssified SiC as a crystal, Henri Moissan. Researchers have mainly considered catalyst systems with polycrystalline forms of SiC as opposed to single crystal [3 9] In this thesis we will investigate the catalytic activity of crushed single crystal SiC. 1.2 Silicon Carbide Material Properties SiC has a tetragonal structure with a C/Si atom at the centre connected to 4 Si/C atoms. The distance between the C atom and the nei ghbouring Si atoms is found to be the same for all polytypes [10] and is approximately 3.08 .. The exact physical properties of SiC depend on the crystal struct ure realized. SiC has been observed to form over 200 polytypes or families of crystals [11] Polytypes in SiC differ in the arrangement of the layers of Si and C atoms. The polytypes are named according to the periodicity of these layers, for example one of the most common polytypes is called 6H, and this means a hexagonal type lattice with an arrangement of 6 different Si + C layers

PAGE 13

3 needed before the pattern repeats itself. Some of the most common structures used are 6H, 4H and 3C where the, C stands for cubic lattice structure. Figure 1.1 Stacking s equence in 4H, 3C and 6H p olytypes of SiC [12] SiC has excellent materia l properties, which is a primary motivation for using this material for applications exposed to extreme conditions i.e. high melting point (Sublimes at temperatures greater than 1800 C), exhibits excellent thermal conductivity (above 3 W/cm K at roo m temperature) and low thermal expansion, thus displays good thermal shock resistance. In addition, demonstrates high hardness, corrosion resistance and stiffness SiC also possesses interesting electrical properties due to its semiconductor characterist ics: The high bond strength in SiC leads to a wide energy bandgap ( 4H SiC: 3.26 eV; 6H SiC: 3.03 eV) allowing it to operate at extremely high temperatures without suffering from intrinsic conduction effects and also allows it to emit and detect short wave length light. It has a high breakdown electric field { 4H SiC: 2.2 x 10 6 V/cm; 6H SiC: 2.4

PAGE 14

4 x 10 6 V/cm (for 1000 V operation)} which is an important parameter for high voltage, high power devices. It has a high, saturated electron drift velocity { 4H SiC: 2. 0 x 10 7 cm/sec; 6H SiC: 2.0 x 10 7 cm/sec (at E = 2 x 10 5 V/cm)} enabling operation at high frequencies. As the properties of SiC began to be investigated, one of the initial results was the first light emitting diode (LED) made as early as 1907 [13] although far more efficient material systems have been fabricated for making LEDs instead of SiC. Currently the semiconductor industry has been in the process of utilising these robust material properties of SiC for high temperature, high frequency and high power applications. However no commercial use of single crystal SiC has been reported for catalytic applications. 1.3 Silicon Carbide in Catalysis To date catalysis, researchers [4 6, 8, 9, 14] have mainly explored the vi ability of using ceramic grade polycrystalline SiC as a catalyst/catalytic support, work that has been on going since the mid nineteen sixties [15, 16] Newer and more improved meth ods of manufacturing ceramic grade SiC have been found and quite a few of them have been patented [17, 18] The two main conclusions researchers have established regarding the attractiveness of SiC as a catalytic support for high temperature combustion reaction s are: SiC has a thermal conductivity close to that of metals such as Ag or Cu [19] thereby reducing thermal shocks which can lead to sintering of the support an d the active phase. It has been found to be physically and chemically stable in inert gases up to temperatures of 1650 o C [8] The chemical stability allows easy recovery of the

PAGE 15

5 active phase and inhibits the reaction of the support with the active phase or other chemical reagents. Lianos, et al [20] have tested the catalytic activity of Pd on a single crystal 6H SiC(0001) for 1,3 butadiene hydrogenation reaction. Pd is the most common catalyst for this reacti on. Pd was deposited on a single crystal 6H SiC by an atomic deposition technique under ultra high vacuum (UHV) conditions. The catalytic activity of the Pd/SiC system was found to be higher than that of Pd surface atoms on pure Pd (111) and Pd(110) cryst als. This suggests a synergistic affect between the support (SiC), and transition metal (Pd) of known catalytic activity. There are many possible reasons for this observed increased activity including: Some amount of catalytic activity shown by the suppor t. Optimum surface structure of the transition metal after depositing onto the support. Extended surface area provided by the support. Electronic interactions between Pd atoms and the support to lower the activation energy. Dispersion of the transitional m etal on the support. Better heat dissipation over the extended surface area of the support. Methivier, et al [19] have conducted tests with methane oxidation using Pd on polycrystalline SiC and found that different activation methods generally lead to different sizes of active Pd particles on the support. The sizes of active Pd particles were determined by H 2 thermal desorption and Transmission Electron Microscopy (TEM). Activities of the relatively smaller particles (d 5 nm) were much greater than the larger particles (d 25 nm). However, the larger particles showed a much improved resistance to deactivation at

PAGE 16

6 temperatures greater than 800 C over the smaller parti cles. Clearly the optimum catalytic geometry is application specific. 1.4 Thesis Outline This thesis research involved the investigation of catalytic activity of Pd supported on single crystal substrates of porous and non porous 4H, 6H and polycrystalli ne 3C forms of SiC. In addition, a newly developed Type II porous 4H material developed by M. Mynbaeva was also explored. Investigations featured Different combinations of material systems. Different activation techniques of the Pd catalysts via: o Reducti on. o Oxidation. Characterisation of catalysts using techniques such as : o Secondary Electron Microscopy (SEM) o Energy Dispersive Spectroscopy (EDS) Exhaust gas analysis techniques using: o Fourier Transform Infrared Spectroscopy. o Gas Chromatography. Now that th e motivation for the research has been presented, in the second chapter more details of the current accepted hypotheses of catalysis are presented. A strong

PAGE 17

7 correlation between catalytic activity & surface structure exists and this is discussed in detail, vis vis adsorption mechanisms, role of activation barriers and mechanisms of catalytic action. Current accepted models of catalytic kinetics are also discussed in this chapter. The third chapter discusses current hypotheses found in literature for the m echanisms of methane oxidation on supported Pd and PdO catalysts including the mechanisms for their activation treatments, namely oxidation and reduction. Chapter four discusses the features of the catalytic reactor test bed built for analysing catalytic m aterial systems. The fifth chapter summarizes the results and concludes with ideas for future research work that could be adopted.

PAGE 18

9 CHAPTER 2: HETEROGENEOUS CATALYSIS AND METHANE OXIDATION A LITERATURE REVIEW The most common example of a hetero geneous catalyst is the catalytic converter on the exhaust pipe of automobiles. Heterogeneous catalysts are preferred over their homogeneous counterparts in most industrial processes despite the fact that homogeneous catalysts increase the rate of the reac tion, to a greater extent than their heterogeneous counterparts. But the phase difference in heterogeneous catalysis makes the recovery of catalyst easier, than in homogeneous catalysis where the catalyst is lost with the products as they dissolve in the r eactant matrix. Heterogeneous catalysis is a purely surface phenomena, hence the nature of the surface plays an important role as will be discussed later in this chapter. Most heterogeneous catalysts are solids and include: Supported metals. Transition m etal oxides and sulphides. Solid acids and bases. Immobilized enzymes and other polymer bound species. The scope of this thesis will be confined to the activity shown by supported transition metal catalysts. Supports tend to spread the active promoter unif ormly, provide a larger surface area for catalytic activity and act as efficient means for energy transfer.

PAGE 19

10 Work has been done using various transition metals as active promoters along with polycrystalline ceramic grade SiC as a support. We intend to use single crystal electronic grade SiC as a support. Crystalline structure plays a major role in catalysis as will be shown later in this chapter. Investigations of porous as well non porous samples of different crystal structures of SiC, namely 4H SiC and 6H SiC, are presented. 2.1 Catalysis Involving Metals The main pathways of catalytic action via metals are Initiating reactions Stabilising intermediates Holding the reactants in the right configuration 2.1.1 Initiating Reactions Most gas phase reaction s follow the initiation propagation mechanism, where the formation of a free radical is essential to trigger the reaction. The ability of metals to dissociate the molecules rapidly allows a fast initiation reaction thus increasing the overall rate of react ion. Gases like oxygen and hydrogen will readily dissociate to form radicals so as to trigger the reaction. 2.1.2 Stabilising the Reaction Intermediates The rate of the reaction is proportional to the concentration of the intermediates in a reaction. Thus if one increases the concentration of the available intermediates in some way one can enhance the rate effectively. Metals generally have a good cache of electrons in the bulk of the material system; these electrons are confined within the metal itself. But on the surface the story is different;

PAGE 20

11 the free electrons can escape the metal and go into the gas phase. Thus any radical in the vicinity of the surface of the metal can be stabilised by these free electrons. This leads to an increase in concentration of the radicals or the so called stabilised radicals on the surface of the metal catalyst, thus increasing the rate of the reaction. The d bands play a major role in this process, by reducing the activation barrier for the reaction thus favouring the form ation of bonds. 2.1.3 Holding the Reactants in the Right Configuration Reactants have to come together to react. This is sometimes very well facilitated by the catalyst which provides the reactants, active sites for the reactants to get adsorbed to and thus facilitates the transformation to the product state. 2.2 Cycles of Catalysis Catalysis in general can be shown to follow a cycle: Adsorption of one or all of the reactants. Reaction amongst the adsorbed species and or other unadsorbed reactant sp ecies. Desorption of the product species 2.2.1 Adsorption Adsorption is solely a surface phenomenon. Masel in his book [1] defines the main types of adsorptio n as: Physisorption where the adsorbate and the adsorbent do not undergo a significant change in their electronic structures.

PAGE 21

12 Chemisorption where the adsorbate and the adsorbent undergo a significant change in their electronic structures. According to Masel [1] a molecule is first physisorbed and then converted into a chemisorbed state, hence adsorption is primarily discussed via the chemisorption process. Ch emisorption of gases on surfaces can be either dissociative or molecular depending specific conditions of temperature, pressure so as to overcome the activation barrier for the reaction as will be discussed later in this chapter. In dissociative adsorptio n the incoming molecule is first adsorbed and then undergoes dissociation depending on the magnitude of the activation barrier. Molecular chemisorption involves adsorption of a molecule without the dissociation. The rate of dissociation is dependant on the crystal structure of the adsorbent. Having said that, there has been no quantitative method to show this but a number of qualitative methods/experiments have shown that the nature of the crystal face, in terms of crystal orientation, the density of steps terraces and kinks play a major role. Adsorption may or may not change the structure of the adsorbent depending on the conditions of temperature, pressure and nature of the reacting species, namely the adsorbent and the adsorbate. Each crystal face has i ts own set of parameters vis vis reactivity, active sites etc. 2.2.2 Surface Reaction 2.2.2.1 Activation Barrier Considerations Activation barriers are crucial for any reaction to occur, but more so for catalysis. The whole idea of catalysis is based on lowering the activation barriers so that the reactants can be transformed into the products.

PAGE 22

13 Consider a simple model put forward by Lennard Jones to show the influence of the activation barrier over the nature of chemisorption that occurs. Consider an interaction betw een a flat surface and a molecule say B 2 Assuming the interaction is a function of only a single dimension i.e. the distance of the molecule from the surface. The presence of a barrier influencing the nature of desorption can be quantitatively explained b y the figure below. Figure 2.1 Lennard Jones model of (a) pure molecular adsorption, (b) activated dissociative adsorption and (c) unactivated dissociative adsorption [1] In figure 2.1 (a) the molecular state has a lower energy state than the dissociative state, hence molecular dissociation will dominate. In figure 2.1 (b) the dissociated state has a lower energy than its counterpart. But there is a fi nite activation barrier which the molecule has to overcome, which in turn requires some energy. Hence in this case the molecular chemisorption will be dominant but dissociative chemisorption is also possible provided some energy is provided to the incoming molecule. This is the regime of operation for most catalytic reactions. In figure 2.1 (c) the energy state of the dissociated state is far lower than the molecular state and the activation energy barrier is below the energy of the molecule, hence the in coming molecule will always dissociate.

PAGE 23

14 For every reaction there is a certain barrier which has to be overcome by the reactants in order to be transformed into products. In other words specific amount of energy has to be supplied to the reactants so that t he reaction can proceed. Masel in his book [1] has outlined the main causes for the activation barrier. They are: For a reaction to happen bonds have to be brok en, thus requiring them to stretch and distort, which in turn requires energy. For molecules to react, they have to come close to each other. To do so they have to overcome Pauli repulsions and other steric effects. Thus requiring some form of activation e nergy. In a few cases quantum effects prevent a transition from the reactants to the products, hence requiring a certain amount of energy to overcome those effects. In certain cases reactants need to be in their excited state for the reaction to occur, whi ch can be provided by the activation energy/barrier. In catalysis the first two causes are the main reasons for the activation barrier as compared to the other reasons which are for specific cases. 2.2.2.2 Mechanisms of Surface Reactions The adsorbed speci es has unique set of properties. It may be strongly or weakly bonded with the surface based on the electronic interactions between the surface and the adsorbate. Most catalytic surface reactions are said to follow either the three commonly accepted mechani sms of surface reactions, or some combination of these: Langmuir Hinshelwood mechanism Rideal Eley mechanism, and

PAGE 24

15 Pre cursor mechanism. Consider a reaction where A + B AB In the Langmuir Hinshelwood approach as shown in figure 2.2 (a) below both reactan ts are adsorbed, they react to form A B adsorbed complex on the surface which then desorbs to form the product AB. Under the Rideal Eley mechanism one of the reactants say A is adsorbed. The adsorbed species reacts with the other reactant B to give the A B adsorbed complex which then desorbs. Fi gure 2.2 Mechanisms of surface r eactions [1] .In the precursor mechanism A is adsorbed. B collides with the surface to form a mobile precursor state which rebounds on the surface until it collides with A to form the A B adsorbed complex, which later desorbs. 2.2.3 Desorption Masel [21] discerns 4 main types of desorption processes in metal catalysts: Simple molecular desorption Recombinative desorption Displacement desorption, and scissions

PAGE 25

16 In simple molecular desorption, the adsorbate complex formed from the surface reaction detaches itself from the surface to give the final product. In most of the cases the molecule leaves without significant rearrangement. In recombinative des orption the adsorbed species react to form a stable molecular species which can leave the surface. This process is dominant in supported metal catalysts In displacement desorption a molecule from the solution that the surface is exposed to, comes in and displaces a ligand from the adsorbed species. This process i s more dominant in metal clusters. scissions are dominant in transition metal clusters wherein the desorbing species leaves a Hydrogen atom at the site. 2.3 Sites for Adsorption It has also been seen that an adsorbate can be held onto a series of dis tinct sites on Figure 2.3 Sites for a dsorption [1] The surface, some of which are shown in the figure below, namely Directly above a surface atom call ed linear or on top site. Between two adjacent surface atoms bridge bound site Above a threefold hollow triple coordinated site.

PAGE 26

17 In a condition where the coverage of the adsorbate is low, random site filling is dominant where the sites fill up randoml y as in case (a) in the figure below. Fig ure 2.4 Coverage of adsorption s ites [21] Under moderate coverages (case b) there is a presence of a weak order and under a strong coverage ( case c) islands are formed. In some cases the adsorption of some species can occur only in certain fixed sites (case d) called random incommensurate adsorption. 2.4 Mechanisms of Processes in Methane Oxidation A Review Researchers have not delved deepl y into the possibility of using single crystal SiC as a viable catalyst support apart from a handful of researchers like Lianos L. [20] Berthet A. [22] and Ledoux M.J. [3] Research interest in the topic waned in the 1990s until the latter half of the decade.

PAGE 27

18 2.4.1 Methane Oxidation Test R eactio n A test reaction was chosen to evaluate the below mentioned catalytic systems. Methane oxidation seemed an ideal candidate for this purpose for the following reasons: Methane is the least reactive of all the hydrocarbons and therefore the most difficult to oxidize. Complete oxidation of methane without any catalyst was found to occur at 950 C in the test bed which is discussed in detail later in this chapter. Hence the reaction is an ideal candidate to lower the temperature of the reaction with the help of a catalyst. The methane molecule can trap energy twenty one times more than a CO 2 molecule, thus making it a potent Greenhouse gas Low temperature methane oxidation can significantly reduce NOx emissions which are formed as by products from high temp erature combustions. Sources of methane are mainly from production and transportation of coal, natural gas, and oil. Methane emissions also result from the decomposition of organic wastes in municipal solid waste landfills, and the raising of livestock. Po ssible applications based on this reaction could be low temperature Catalytic converters, Gas sensors, Exhaust scrubbers, H reforming, The ease in modeling the C1 chemistry is also an additional bonus in choosing this test reaction system. Reactions invo lved: CH 4 + O 2 CO + 2H 2 ?H r (STP) = 35.7 KJ/mol (partial oxidation) CH 4 + 2O 2 CO 2 + 2H 2 O ?H r (STP) = 890 KJ/mol (complete oxidation)

PAGE 28

19 Both reactions are exothermic! Methane oxidation has been well res earched by many researchers but some of its aspects have not been fully understood. Most of the papers reviewed on CH 4 oxidation have either used Alumina or Zirconia as their choice of support, with the latter gaining the favour of most researchers. Many p ossible mechanisms have been reported in literature for the various processes we have used to test the samples, namely, Activation treatment. Reduction. Reaction on reduced Pd. Activation treatment Oxidation. Reaction on PdO. 2.4.2 Activation Treatmen t Reduction of PdO Although a lean mixture of CH 4 along with an inert like N 2 can be used as a reducing mixture, Ciuparu et al in their review of CH 4 oxidation using Pd Based catalysts [23] observed that using H 2 at a high enough temperature reduced the uptake of O on th e cooling cycle. Su et al. [24] found t hat reduction with H 2 as an activation treatment for a Pd/ZrO 2 at around 533 K, occurs so quickly that a layer of metallic Pd forms in a shell wise manner. We used a 50/50 : H 2 /N 2 mixture at 1073K for 2 hours as our reducing agent. Pd metal is formed on the surface screening the PdO layer. As the temperatures are increased so does the rate of reduction. The rate limiting step was found to be the removal of oxygen from the surface. Ciuparu et al [23] found that a t low coverages, the PdO phase is known to spread over the s upport, while the metallic Pd phase forms larger, more segregated crystallites. This was evident in our samples as well.

PAGE 29

20 According to them the encapsulation of metallic clusters into the oxide phase is likely because of their higher surface tension compar ed to the oxide phase, consistent with the metal core growing in the PdO shell model. 2.4.3 Reaction after Reduction of PdO According to Ciuparu et al. [23] dissociative adsorption of O 2 is much faster than that of CH 4 CH 4 dissociatively adsorbs onto metallic Pd as an a dsorbed hydrogen species and an adsorbed methyl species. The Langmuir Hinshelwood model of adsorption can be used to approximate the dissociative adsorption of CH 4 Competitive adsorption patterns between O 2 and CH 4 make the adsorption of CH 4 the rate con trolling step for methane oxidation over metallic Pd. At higher temperatures the coverage extent of surface O has been found to decrease, thus allowing higher reactivity at higher temperatures. Ciuparu et al [23] in their review have noted that researchers have observed th at at low coverages the PdO phase is well spread out, where as the Pd phase segregates to form larger crystallites. This was also observed in our samples as well where clusters of Pd (size range: 1 5 microns) were detected by EDS (SEM images of the chunk s are available in the Appendix A). They also observed that CH 4 oxidation over metallic Pd is structure sensitive as researchers have noticed that activity increased as the size of Pd increased. Figure 2.5 Optimized structures for critical points of t he O H bond formation reaction: (a) CH 3 + H on Pd fragment; b) intermediate activation; c) transition state; d) adsorbed CH 3 + OH ( Bond distances are given in ) [25]

PAGE 30

21 Broclawik et al [25] suggest a mechanism for the reaction which follows the following path as shown in the figure above a) Dissociative adsorption of CH 4 molecule at a bridge position, i.e., between two Pd atoms. b) An intermediate activated state. c) A transition state in which both adsorbed methyl and hydrogen species are weakly bound allowing the H species to move around the Pd Pd dimer easily or to bind itself to a neighbouring surface O atom as well; and d) the final product is a surface hydroxyl. They found that the direct O H interaction is apparently switched on This activation costs 27 kcal/mol and the energy gain after forming a strong OH bond is 15 kcal / mol The methyl species attains the position of the initially adsorbed methane and the next C H bond becomes activated, thus the entire cycle may be repeated The abstraction of next hydrogen atoms may be even easier than the first one and the adsorbed OH group may thus react to form water molecules 2.4.4 Oxidation of D eposited Pd Due to the disparity in the Pd Pd distance in Pd (0.275 nm) and PdO (0.305 nm) the crystal structure has to undergo significant reconstructions during oxidation of Pd. The same holds true for reduction of PdO as well. According to Ciuparu et al [23 ] the (001) and (110) bulk terminated PdO surfaces expose only a single type of O species as no distinction can be made between surface, adsorbed or lattice O atoms. The O atoms are bridge bound to two Pd atoms. Analysing the results from various res earchers Ciuparu et al [23] in their review of Pd based catalysts for methane oxidation, put forth a three step mechanism for the oxidation of Pd (111). They are: Rapid adsorption that stops at 0.25 ML of atomic oxygen

PAGE 31

22 Oxygen penetration into the surface accompanied by isl and growth leading to the formation of two phases that are intermediate between Pd and PdO. and Formation of bulk PdO. 0.25 ML saturation coverage of O is formed on the surface of the Pd atom at temperatures below and up to room temperatures and pressures on the order of 10 5 torr. They argue that oxygen growth on the surface has been found to take place via a nucleation mechanism due to the following reasons: Researchers have observed that the rate of oxidation increases with the temperature until a thres hold value is reached after which it decreases considerably, which is a characteristic of the nucleation mechanism. TEM studies have shown that reoxidised Pd particles undergo considerable surface roughening which is another trait of the nucleation mechani sm. The chemisorbed O atoms from the first monolayer tend to penetrate the surface forming Pd O clusters along with other Pd atoms on the surface. These clusters grow to form islands until the whole surface is covered under favourable conditions. The PdO t hus formed say, Carstens et al [26] is more amorphous than that formed under reactions conditions at the same temperature. They have examined the catalyst Pd/ZrO 2 for methane combustion under reaction conditions and have found that a thin crystalline PdO film formed over met allic Pd of approximately 7 monolayers thick is required for achieving high activity. They also suggest that PdO along with small amounts of Pd increases the activity of the catalyst as CH 4 dissociates readily over metallic Pd rather than PdO.

PAGE 32

23 Crystalline PdO has been found to be stable in temperatures under 1046 K. Researchers have found that crystalline PdO is more active than metallic Pd at lower temperatures. the activation energies of methane oxidation on crystalline PdO is lesser than metallic Pd. [23] 2.4.5 Reacti o n over Oxidised Pd The mechanism for oxidation of methane over PdO is not entirely understood and many hypotheses have been put forward but the most recent published article on the subject was from Ciuparu and Pfefferle [27] They propose that methane oxidation on PdO follows a redox mechanism where the CH 4 molecule first dissociates on a Pd O lattice site as shown in the figure 2.6 suggested by Broclawik [28] The dissociation products are an adsorbed methyl and surface hydroxyl species. This mechanism suggests that any surface dimer site can participate in methane ac tivation and the gas phase methane molecule is directly adsorbed on the surface. The methyl radical undergoes further H abstraction to produce more surface hydroxls and CO 2 CO 2 is desorbed rapidly creating a vacant site which can be occupied either by O f rom the gas side or from the support based on the partial pressure of O over the catalyst and the O mobility in the support. Finally when the hydroxyl recombine with another adsorbed H species and desorb, further vacant sites are created. The hydroxyl deso rption was found to be the rate controlling step for this mechanism. They also suggest that due the slow recombination of the surface hydroxyl and desorption of water prevents the reoxidation of the site. Ciuparu and Pfefferle in [29] propose 2 regimes of methane oxidation. The first regime occurs at lower temperatures where recombination of the surface hydroxyl is the rate

PAGE 33

24 controlling step and the second regime is at higher temper atures where methane activation is the rate controlling step. Figure 2.6 Optimized structures of; a) collinearly adsorbed methane, b) bridging adsorption, and c) the dissociated product (B ond distances given in ) [28] (a) (b) (c)

PAGE 34

25 CHAPTER 3: CATALYTIC REACTOR DEVELOPMENT The catalytic reactor test bed shown in figure 3.1, was setup which consisted of a bank of mass flow controllers connected to the tubular quartz reactor, which in turn is connected to the Fourier Transform Infrared Spectrometer (FTIR) for analysing the exhaust gases. 3.1 Experimental Test Bed A gas delivery system was designed to setup a 2 % (vol) methane stream with the help of four mass flow controllers (MFC) bought from MKS Instruments Inc.; of type 1179A; of which two were rated for a flow of 20 sccm, another for 50 sccm and the final MFC was rated for 100 sccm. The MFCs were calibrated by the manufacturer for nitrogen flow and the appropriate gas correction factor was applied via a MKS 247D four channel flow controller readout, which was used to control the MFCs. The correction factors were made prior to running the various gases through the MFCs. The gases from the respe ctive MFCs were mixed in a static mixer of length 6 inches. A vent/purge line was also added to this delivery system for safety purposes. The 2% CH 4 mixture which emerges out of the static mixer is then passed through a final flow controller before enteri ng the reactor.

PAGE 35

26 Table 3.1 Details of the flow controllers con nected for the respective gases Figure 3.1 Block diagram of the catalytic reactor test b ed Figure 3.1 shows a block diagram of the test bed. The reactor is a quartz tube with a constriction in the centre so as to create a packed bed o f 1 cm in length where the catalyst material is loaded. The ends of the tube are connected with Ultra Torr tees to provide sample ports for inlet and outlet streams for analysis using with the Gas Chromatograph (GC) or the FTIR. The reactor is placed in a Lindberg/Blue M tube furnace which is capable of heating till 1100 0 C. The gases, on leaving the reactor go through a gas cell accessory of the FTIR for analysis from where they are properly vented. Gases Flow (sccm) Correction factor Nitrogen 50 1 Oxygen 20 198 Methane 20 144 Final mixture 100 99 FC Throttle Valve To Scrubber To Scrubber MFCs Pump GC/FTIR Port GC/FTIR Port Static Mixer Reagents Flow Meter Bubbler P gauge Heterogeneous Reactor CH 4 H 2 N 2 O 2

PAGE 36

27 Figure 3.2 Snapshot of the catalytic reactor t est b e d A Bio Rad Excalibur series FTIR was used for analysis of the exhaust gases from the reactor. It is fitted with a liquid nitrogen cooled, Mercury Cadmium Telluride (MCT) detector which can analyze effectively for wavenumbers between 4000 & 600 cm 1 a range that is relevant for most of the gases used and produced in the reactor. The gas cell provides an infra red beam path length of 2.1 metres through multiple reflections from KBr coated optical mirrors. The gas cell is kept under a constant vacuum with the help of a roughing pump as well as to reduce the ambient CO 2 and water vapour contaminations in the background infrared (IR) spectrum. A photograph of the FTIR system is shown in figure 3.2 3.2 Catalyst Preparation The catalyst compound t o be evaluated in the catalytic test bed was prepared in the following manner: Reaction Tube Loade d Catalyst

PAGE 37

28 n Dodecanol, being a non polar solvent, was used as a solvent to dissolve the Pd(NO 3 ) 2 salt. Crushed pieces of wafer/powder are then introduced into this n dodecanol/Pd(NO 3 ) 2 m atrix. The alcohol is then slowly evaporated over a period of approximately 2 hours. The residue is then calcined under atmospheric conditions at 500 0 C for 2 hours. As the calcined catalyst attains room temperature it is weighed and loaded into the reacto r. (Activation step) Once loaded into the reactor it is oxidised under a 50/50:N 2 /O 2 mixture at a flow rate of 10 sccm through the reactor at about 800 0 C for about 2 hours. Secondary Electron Microscopy (SEM) and Energy Dispersive Spectroscopy (EDS) were carried out on the samples before and after the reaction. 3.3 Test Reaction Conditions The Nitrogen flow through the reactor was 88 volume %; oxygen was 10 volume % and methane was at 2 volume %. The total flow through the reactor was 10 sccm. The product s of the reaction were analysed using the FTIR spectra at every 50 0 C increments ranging from room temperature till about 950 0 C, or until all the methane was spent, whichever came first. From the FTIR spectra a conversion plot was created based on the con centrations of CH 4 /CO 2 over the entire temperature range, to compare the abilities of the various catalyst systems tested. Before describing how the data was analysed, the analytical instrumentation is presented.

PAGE 38

29 3.4 Instrumentation The FTIR used for ana lysing the exhaust gases from the reactor was a Digilab (formerly Bio Rad) Excalibur FTS 3000 which was fitted with a liquid N 2 cooled Mercury Cadmium Telluride (MCT) detector which has a range of 4000 600 cm 1 wavenumbers. respectively. A gas cell ac cessory, bought from Pike Technologies Inc., was fitted into the sample compartment chamber of the FTIR. The gas cell is permanently aligned with a infrared beam path length of 2.4 m It has a borosilicate glass body, with KBr coated windows and an approxi mate volume of 0.1 liters. The mirrors of the gas cell accessory were manually aligned with the infrared beam of the FTIR by adjusting the mirrors on the gas cell manually. At the start of every exhaust gas analysis run the MCT detector was cooled by filli ng the liquid N 2 reservoir which, once filled, lasts for 8 10 hours of continuous operation. There was a constant N 2 purge applied to the sample compartment chamber and the chamber, which housed the interferometer. Once the detector was cooled the gas ce ll was purged with N 2 which flowed at 10 sccm through the reactor for 40 minutes to purge the gas cell of residual CO 2 and H 2 O contamination from the previous run. The voltage response from the detector was checked and was always adjusted between 5 and 7 volts as suggested by the manufacturer. Changing the aperture diameter or the sensitivity through the program supplied by Digilab to control the instrument did this adjustment. A reference scan was then taken with N 2 flowing at 10 sccm and also at the reac tion configuration. The total flow at anytime through the reactor was kept constant at 10 sccm. The FTIR was run at the following scan settings: Number of scans: 16 Scan resolution: 0.1 cm 1 An inventory of the components used in the experimental test bed are given in Appendix B

PAGE 39

30 3.5 Sample Calculation for Conversion The sample scans were ratioed with their appropriate backgrounds collected at the beginning of the sample run and the conversion of CH 4 into products were calculated as is explained F igur e 3.3 Sample s pect rum of type II 4H SiC/Pd under reaction c onditions at 300 C in this section. F igure 3.3 is a sample spectrum at 300 0 C for the type II porous 4H SiC/Pd sample with 10 sccm flow of the reactant mixture. Figure 3.4 is a background spectr um collected at room temperature with 10sccm N 2 flow through the gas cell. The peaks were identified as follows: Table 3.2 FTIR p eak a nalysis Species CH 4 CO 2 CO 2 CO Lower Limit 2650 2230 620 2060 Upper Limit 3200 2400 720 2230 Sample Spectrum at 300 C

PAGE 40

31 Figure 3.4 A typ ical background spectrum An absorbance s pectrum is obtained by the ratio of the sample spectrum and the b ackground s pectrum The area under absorbance spectrum is a measure of the concentration of the species, at that time. The number within the brack ets in figure 3.5 are the values for the area under the curve, which is obtained from the Digilab (WINIR PRO) software. The area under the curve for all temperatures is tabulated for methane and carbon dioxide species as follows: Table 3.3 Analysis of FT IR d ata Temperature ( C) CO 2 (Area under the curve) CH 4 (Area under the curve) Conversion 950 53.359 1.871 0.905923 300 3.732 19.807 0.063361 50 19.888 0 Background spec trum

PAGE 41

32 o Figure 3.5 Absorbance s pectrum o btained from the ra tio of the sample spectrum and the background for a type II 4H SiC/Pd under reaction c onditions at 300 C The conversion for any temp is calculated as follows: Inlet concentration of methane at room temp (area under the curves are noted) = 19.888 Final concen tration of methane at 950 C = 1.871 Hence final conversion is: (19.888 1.871) / 19.888 = 0.9059 Hence fraction of methane spent is 0.9059 Therefore, If the area under the peak for CO 2 @ 950 C which is 53.359 corresponds to conversion of 0.9059 of metha ne Then the CO 2 area of 3.732 at 300 C corresponds to a methane conversion of : 3.732 x 0.9059/53.359 = 0.063361 CO 2 CH 4 Absorbance Spectrum at 300 C H 2 O

PAGE 42

33 CHAPTER 4: CATALYTIC REACTOR EXPERIMENTS 4.1 Substrates Used Samples of porous and non porous 4H and 6H SiC wafers (obtained from the DURINT Program) and also 3C Beta SiC powder (obtained from Marketech Intl.) were investigated as substrates for a catalyst system. Pd was chosen as the active phase in the catalyst on the basis of literature for the methane oxidation reaction system. Th e substrates tried were 4H SiC o porous o non porous/standard 4H SiC impregnated with Pd o porous o non porous/standard 6H SiC o porous o non porous/standard 6H SiC impregnated with Pd o porous o non porous/standard

PAGE 43

34 Beta SiC powder Type II 4H PSC 4.2 Beta SiC vs. Po lycrystalline 3C SiC an XRD Comparison Most of the SiC samples were obtained from the DURINT ONR program. The beta SiC powder was obtained from Marketech International. An XRD comparison was conducted between a standard 3C SiC wafer obtained from the O NR program and the beta SiC. The beta SiC was of a lower crystalline quality than the 3C wafer as is evident from the graph below, when one compares the FWHM values for both. Beta-SiC Powder 30 35 40 45 50 55 60 65 70 75 80 2 Theta (111) FWHM : 0.171 (311) FWHM : 0.873 (220) FWHM : 0.177 (200) (222) Particle size =< 30nm by TEM Surface area >= 109 sq.m/gm 3C-SiC Wafer Sample holder (111) FWHM : 0.821 (200) FWHM : 0.199 (220) FWHM : 0.818 (311) FWHM : 0.182 Figure 4. 1 XRD spectra of substrate

PAGE 44

35 4.3 SEM Analyses Before the Reaction Th e SEM images before the reaction did not show too many differences between the samples. (See Appendix A for images). On most of the SiC samples, Pd tended to aggregate as a bunch of spheres or lobes the size of which were not greater than 0.1 A network of these lobes & bunches were formed. Pd was also deposited as whole chunks measuring upto 10 in size in some cases. Pd was well dispersed over the supports. The 4H polytype showed a relatively greater affinity towards Pd than the rest of the samples. Th is is fairly evident from the figures in Table 4.1 below. Table 4.1 Loading of Pd (in wt %) on the s ubstrate from EDS The figures in the table were obtained from EDAX EDS. They were averaged from the readings obtained at 10 different spots on each of the samples. The figures for Beta SiC powder could not be ascertained because the sample was lost at the time of unloading the sample after running the reactio n. SEM and EDS were primarily used to characterize the catalyst. XPS and XRD were also tried out but the results from these analyses turned out to be vague because: The design of the XPS system we had access to did not facilitate analysis of powders. The amount of catalyst sample involved was too small to conduct a decent XRD analysis. Sample 6H STD/Pd 6H PSC/Pd 4H STD/Pd 4H PSC/Pd Type II 4H PSC /Pd Wt% 1.85 1.89 4.30 3.41 1.78

PAGE 45

36 4.4 Reaction Data 4.4.1 Under Reduction Regime The most common activation treatment for most metal/supported metal catalysts is a reduction process. Hence we used reduction using a 50/50 : N 2 /H 2 stream as the activation step. The initial reaction results were obtained using the Gas Chromatograph (GC) using a 6 silica gel column. 0 0.2 0.4 0.6 0.8 1 350 450 550 650 750 Temperature (deg C) Conversion/gm of Catalyst 4H-SiC/Pd 4H-SiC 6H-SiC (30sccm) 6H-SiC/Ni (After Reduction) Polycrystalline SiC Quartz tube (30sccm) Figure 4.2 Reaction results from the GC Of particular interest was the activity shown by bare po rous substrates i.e. 4H PSC and 6H PSC 600 C and 700 C mark respectively, which indicated some limited amount of catalytic action albeit a small amount unlike other supports which were pretty much inert to the reaction. The porous 4H SiC (4H PSC) coated with Pd showed some activity around 450 C, which had been observed in other catalyst systems.

PAGE 46

37 The set of reactions from here on were monitored by the Bio Rad FTIR which are relatively more accurate. Under the same reduction regime as the activation ste p the tests were carried on to give the following results as shown by the graph below: Under Reduction Regime 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 250 350 450 550 650 750 850 950 Temperature (deg C) Conversion (%) Beta SiC Powder 6H-PSC 4H-STD Type II 4HPSC Light-off Temperature Figure 4. 3 Conversions of various substrates under a reduction regime. Most samples follow a similar route wherein they have a range of temperature in which there i s some minute activity i.e. conversion is less than 10% and then the gas reaction takes over after 875 C. The beta SiC powder starts to show some conversion (less than 10%) from 500 C onwards till 900 C when the gas reaction takes over at that temperatu re. The type II porous 4H SiC follows a similar line but the start of activity is delayed till 700 C. The standard/non porous 4H SiC shows its share of minute activity from around 750 C or so. Under Reduction Regime

PAGE 47

38 This shifted our thinking from the viewpoint that SiC in its elf could be a catalyst on its own to that SiC may facilitate catalysis in very controlled amounts unlike other inert supports. After testing the substrates, the reaction data for the loaded substrates under the reduction regime is shown in the graph below Most samples in this set follow similar suit, but the presence of Pd marks the difference between the graph below and the previous graph. For the initial range of temperatures till around 650 C, the conversion is below 10% after which the catalytic acti on takes over. Under Reduction Regime with Pd R 2 = 0.9924 R 2 = 0.9988 R 2 = 0.9994 0 0.15 0.3 0.45 0.6 0.75 0.9 250 350 450 550 650 750 850 950 Temperature (deg C) Conversion Beta SiC/Pd Type II 4H-PSC/Pd (~1.7 wt%) 6H-PSC/Pd (~1.8 wt%) 4H-PSC/Pd (~3.4 wt%) Light-off Temperature Figure 4.4 Reaction results of porous SiC and beta SiC impregnated with Pd after reduction (activation treatment) Beta SiC coated with Pd showed a light off at about 350 C after which the activity steadily increases unlike the other sam ples. It does not have a region of very low activity (conversion < 10%) which is interesting. Under reduction regime with Pd

PAGE 48

39 Type II porous 4H SiC which has a Pd loading of about 1.7 % by weight shows the distinctive region of low activity from 300 C onwards 600 C after which the cata lytic action takes over the reaction more definitively. The porous 6H SiC has a Pd loading of around 1.8 % by weight, follows a similar route as the previous sample. The region between 350 and 750 C is the region of low activity where conversions are les s than 10%. After 750 C the catalytic action takes over the reaction. The porous 4H SiC has the highest Pd loading of about 3.4 % by weight. This sample does not show the region with consistently low activity. Instead activity starts from 600 C and from thereon itself the catalytic action takes over. 4.4.2 Under Oxidation Regime Instead of reducing we thought of oxidising the catalyst sample to see the changes that it may bring to the catalytic action. We found in literature [30] that PdO is a much more potent catalyst than Pd itself. Hayes et al say that palladium oxide (PdO) is readily formed when Pd supported on alumina is heated in an oxygen environment above 600 C. Hence we switched the activation step from redu ction to an oxidation step with a 50/50 : N 2 /O 2 mixture. The results were much better across all samples in terms of the catalytic action: Under the oxidation regime beta SiC powder coated with Pd had a much better result where it lit off around 200 C and showed the sustained catalytic activity till the end. This sample however did sinter forming aggregates & plugging the reactor tube. Pd Na

PAGE 49

40 The standard 4H SiC sample which had a Pd loading of about 4.3 % by weight starts to show catalytic action from around 225 0 C. Under Oxidation Regime R 2 = 0.9954 R 2 = 0.9845 R 2 = 0.9978 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 100 200 300 400 500 600 700 800 900 Temperature (deg C) Conversion 4H-STD Type II 4H-PSC Beta-SiC/Pd 4H-STD/Pd (~4.3 wt%) 6H-STD/Pd (~1.8 wt%) Light-off Temperature Figure 4.5 Reaction results after oxidation (activation treatment) The standard 6H SiC sample with a Pd loading of around 1.8 % by weight comparatively has a higher temperature from where it shows activity i.e. around 475 C. Under oxidation regime

PAGE 50

41 Under Oxidation Regime R 2 = 0.9766 R 2 = 0.9921 R 2 = 0.9625 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 100 200 300 400 500 600 700 800 900 Temperature (deg C) Conversion Type II 4H-PSC/Pd (~1.7 wt%) 4H-PSC/Pd (~3.4 wt%) 6H-PSC/Pd (~1.8 wt%) Light-off Temperature Figure 4.6 Rea ction results from porous SiC substrates with Pd after oxidation (activation t reatment) The porous type II 4H SiC sample with a Pd loading of about 1.7 % by weight lights off the earliest amongst all the samples i.e. 175 C. But it rises gradually till 6 00 C and only then takes off. The porous 4H SiC with a relatively Pd loading of about 3.4 % by weight follows the route taken by the Type II. Catalytic activity rises gradually from 225 0 C onwards till around 650 0 C after which the reaction really takes o ff. The porous 6H SiC with a Pd loading of around 1.8 % by weight shows some minute activity (conversion < 10%) from 375 C till about 750 C after which the reaction really starts. Under oxidation regime

PAGE 51

42 4.5 Reaction Data For Beta SiC Beta SiC/Pd R 2 = 0.9954 R 2 = 0.9917 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 100 200 300 400 500 600 700 800 900 Temperature (deg C) Conversion Beta-SiC/Pd Under Oxidation Beta SiC/Pd Under Reduction Light-off Temperature Figure 4.7 Reaction r esults of b eta S iC The switch from reduction to oxidation as the activation step affected beta SiC the most. It was by far the best sample in terms of activity shown. Having said that, the sample was the only one in which sintering effects could clearly be seen in the re actor. A possible reason for this sample doing so well as it did might be because of it fine powder format whereas the other samples were in the form of small wafer pieces. Unfortunately the Pd loading on this sample could not be found out as the sample wa s lost in the reactor while being removed. 4.6 Reaction Data for Type II Porous 4H SiC Type II porous 4H SiC was by far one of the better samples showing activity from 150 C onwards. Oxidation definitely seems to work a lot better as an activation treat ment Beta S iC/Pd

PAGE 52

43 when considering Pd. There was no evidence of sintering as well for this sample. Type II 4H-PSC R 2 = 0.9785 0 0.15 0.3 0.45 0.6 0.75 0.9 0 100 200 300 400 500 600 700 800 900 Temperature (deg C) Conversion Without any catalyst Type II 4H-PSC Under Reduction Type II 4H PSC/Pd Under Reduction (~1.7 wt%) Type II 4H-PSC/Pd Under Oxidation (~1.7 wt%) Type II 4H-PSC Under Oxidation Type II 4H-PSC/Pd Under Oxidation (~1.7 wt%) Light-off Temperature Figure 4.8 Reaction r esults of t ype II p orous 4H SiC 4.7 Reaction Data for Standard and Porous 4H SiC Both the 4H samples, porous and standard showed a greater aff inity towards getting coated with Pd than any of the samples as shown in Figure 4.10. The loading for the porous 4H sample was around 3.8 % by weight and that of the standard sample was about 4.3 % by weight. They do show a lot of promise as materials for catalytic supports. Type II 4H PSC

PAGE 53

44 4H-STD & 4H-PSC R 2 = 0.9912 R 2 = 0.9921 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 100 200 300 400 500 600 700 800 900 Temperature (deg C) Conversion 4H-STD/Pd Under Reduction (~4.3 wt%) 4H-STD/Pd Under Oxidation (~4.3 wt%) 4H-PSC/Pd Under Oxidation (~3.4 wt%) Light-off Temperature Figure 4.9 Reaction results of standard and p orous 4H SiC 4.8 Reaction Data of Porous and Standard 6H SiC The porous 6H SiC sample was one of the worst in the set that was tested. The standard 6H showed much more reliable catalytic activity from 500 C than the Porous sample which showed similar activity from 750 C onwards. There did not seem to be much of a change even when the activation step was altered. 4H STD and PSC

PAGE 54

45 6H-SiC R 2 = 0.9834 0 0.15 0.3 0.45 0.6 0.75 0.9 0 100 200 300 400 500 600 700 800 900 Temperature (deg C) Conversion 6H-PSC/Pd Under Reduction (~1.8 wt%) 6H-PSC Under Reduction 6H-PSC/Pd Under Oxidation (~1.8 wt%) 6H-STD/Pd Under Oxidation (~1.8 wt%) Poly. (6H-PSC/Pd Under Oxidation (~1.8 wt%)) Light-off Temperature Figure 4.10 Reactions results of 6H SiC 6H SiC

PAGE 55

46 CHAPTER 5: SUMM ARY AND FUTURE WORK 5.1 Summary In this study porous and non porous samples of semiconductor grade single crystal SiC and nanocrystalline b SiC powder were investigated as substrates in catalytic systems. Methane oxidation was chosen as the test reaction for which Pd and PdO supported on alumina or zirconia are the current catalytic systems of choice. Two forms of activation pre treatments were explored, viz., reduction and oxidation. The oxidation pre treatment proved to be far more effective in activat ing the catalyst system as compared to reduction. This was well supported by researchers that PdO along with a Pd phase together is more effective for methane oxidation. As EDS showed for the samples used in this work, Pd segregated and the PdO phase spre ad out well around the Pd phase and also over the support. Nanocrystalline beta SiC powder obtained from Marketech international initially showed the most promise vis vis the lower light off temperature and the sustained high level of conversion after l ight off. These results could be attributed to the fact that the beta SiC/Pd was the only sample in a powder form thus increasing the surface area of contact with the reactant gases. The other samples were derived from crushed wafer pieces. However, the b SiC/Pd system did sinter easily and did show considerable amount of CO as compared to the other wafer derived samples. Therefore, this material can prove to be the most expensive and environmentally hazardous. The 4H SiC polytype sample shows

PAGE 56

47 tremendous pr omise, the type II porous in particular. This may be due to an efficient handling of the adsorbed OH species which is the rate controlling step for methane oxidation over PdO based catalysts. Researchers have ascribed to the fact that, supports in catalys t. systems which showed greater ability to move oxygen showed better results. The standard 4H polytype also showed results which were almost at par with its porous counterpart. The 6H polytype porous and standard showed less activity than the 4H samples this may be due to the lower electron mobility of 6H SiC vs. 4H SiC; further investigation is needed. Small amounts of CO were observed as one of the reaction products. For most samples CO was observed at temperatures between 700 900 C. The research wor k conducted on these catalyst systems, being of preliminary nature, more work needs to be done for commercial applications, in particular, for MEMS, micro reactors and portable fuel cells to be produced based on this material system. 5.2 Future Work Thes e were some of the thoughts or directions in which further research work could be conducted: Most of the catalyst systems used were crushed wafer pieces except for Beta SiC. To provide level playing field catalyst systems need to be finely powdered before reaction takes place. Characterisation data from either/or XPS & XRD needs to be collected and investigated. This would definitely aid us in finding out the dominant surface phase at each step of the process. XRD would give us the nature of the active phas e and its crystalline quality along with the crystal face. We could not use either of these techniques as the XPS system we have is not really

PAGE 57

48 suited to analysing powders and the amount of catalyst material was less than required for a proper XRD analysis Catalyst aging studies need to be carried out along with temperature programmed desorption and adsorption studies. The life of the catalyst is an important aspect which can increase or decrease the cost of the process. Different deposition/impregnation techniques to deposit the Pd/PdO phase onto the SiC support need to be looked into as well. Electron beam implantation could be one of the possible alternatives for a better uniform deposition. Since SiC is a semiconductor, it could be used as a catalyst in itself by biasing it, to promote the exchange of electrons with the adsorbed species, thus aiding in catalysis so as to open a vista for a plethora of microelectronic devices. SiC could be well suited as part of a photocatalytic system, where on expos ure to UV radiation, activated oxygen is formed, which could trigger most gas phase reactions at lower temperatures thus also decomposing organic systems. Although Pd has emerged as the transition metal of choice for catalytic oxidation of methane, combin ations of different transition elements could be tried out to enhance the quality of the result, vis vis the selectivity. Ni and Rh could be used specifically for partial oxidation Different stoichiometric ratios of reactants could also be experimented to see if there are any effects, positive or negative from it, as CH 4 in O 2 lean mixtures can act as a reducing agent.

PAGE 58

49 REFERENCES 1. Masel, R.I., Principles of Adorption and Reaction on Solid Surfaces. 1996: Wiley Interscience. 2. Ceramics., S. G.A., History of Silicon Carbide. 3. Ledoux, M.J. and C. Pham Huu, Silicon carbide a novel catalyst support for heterogeneous catalysis. Cattech, 2001. 5(4): p. 226 246. 4. Lednor, P.W., Synthesis, stability and catalytic properties of high surface area silicon oxynitride and silicon carbide. Catalysis Today, 1992. 15: p. 243 261. 5. M. B. Kizling, P.S., S. Andersson, A. Frestad., Characterisation and catalytic activity of SiC powder as catalyst support in exhaust catalysts. Applied Catalysi s B Environmental, 1992. 1: p. 149 168. 6. M. Benaissa, C.P. H., J. Werckmann, C. Crouzet, M.J. Ledoux, A sub nanometer structural study of Pt Rh catalysts on Ce doped SiC. Catalysis Today, 1995. 23: p. 283 298. 7. Cuong Pham Huu, N.K., Gaby Ehret, Marc J. Ledoux, The First Preparation of Silicon Carbide Nanotubes by Shape Memory Synthesis and Their Catalytic Potential. Journal of Catalysis, 2001. 200(2): p. 400 410. 8. T. Okutani, Y.N., M. Suzuki, H. Nagai., Development of stable supports consisting of SiC Si composite for high temperature combustion catalysts. Catalysis Today, 1995. 26: p. 247 254. 9. S. Maruko, T.N., M. Onodera, Multistage catalytic combustion systems and high temperature combustion systems using SiC. Catalysis Today, 1995. 26: p. 309 317 10. Knippenberg, W.F., Philips Research Reports, 1963. 18(No. 3): p. 161 274. 11. Davis, R.F. in Proceedings of the International Conference in SiC and Related Materials 93, Washington DC, USA, Inst. Phys. Conf. Ser. 1994. 12. Kordina, O., SiC as a Power Device Material in Dept. of Physics and Measurement Technology, Biology and Chemistry Lingkoping University. 13. Round, H.J., Electrical World, 1907. 19: p. 309.

PAGE 59

50 14. P. Cuong, S.M., M.J. Ledoux, M. Weibel, G. Ehret, M. Benaissa, E. Peschiera, J. Guille. Synthesis and characterisation of Platinum Rhodium supported on SiC and SiC doped with Cerium: Catalytic activity for automobile exhaust reactions. Applied Catalysis B Environmental, 1994. 4: p. 45 63. 15. Yoshihito, M., Journal of Catalysis, 1966. 5: p. 116. 16. Yoshihito, M., Journal of Catalysis, 1967. 7: p. 23. 17. Tsukada, K., Porous silicon carbide sinter and its production 1988. 18. M.J. Ledoux, J.L.G., S. Hantzer, D. Dubots., Process for the production of silicon carbide with a large specific sur face area and use for high temperature catalytic reactions 1988: France. 19. Methivier, C., et al., Pd/SiC catalysts Characterization and catalytic activity for the methane total oxidation. Journal of Catalysis, 1998. 173(2): p. 374 382. 20. Lianos, L., et al., Properties of Pd deposited on SiC(0001) single crystal surfaces. Journal of Catalysis, 1998. 177(1): p. 129 136. 21. Masel, R.I., Chemical Kinetics and Catalysis 2001: Wiley Interscience. 22. Berthet, A., et al., Comparison of Pd/(bulk SiC) catal ysts prepared by atomic beam deposition and plasma sputtering deposition: Characterization and catalytic properties. Journal of Catalysis, 2000. 190(1): p. 49 59. 23. Ciuparu, D., et al., Catalytic combustion of methane over palladium based catalysts. Catalysis Reviews Science and Engineering, 2002. 44(4): p. 593 649. 24. Su S.C., Carstens J.N., and B. A.T., A study of the dynamics of Pd oxidation and PdO reduction by H2 and CH4. Journal of Catalysis, 1998. 176(1): p. 125 136. 25. Broclawik E., et al., Density functional study on the activation of methane over Pd2, PdO and Pd2O clusters. International Journal of Quantum Chemistry, 1997. 61(4): p. 673 682. 26. Carstens J.N., Su S.C., and B. A.T., Factors affecting the catalytic activity of Pd/ZrO2 for the combustion of methane. Journal of Catalysis, 1998. 176(1): p. 136 142. 27. Ciuparu, D. and L.D. Pfefferle, Contributions of lattice oxygen to the overall balance during methane combustion over PdO based catalysts. Catalysis Today, 2002. 77(3): p. 167 179. 28. Broclawik E., et al., On the electronic structure of the palladium monoxide and the methane adsorption: Density functional calculations. Journal of Chemical Physics, 1996. 104(11): p. 4098 4104. 29. Ciuparu, D. and L. Pfefferle, Support and wat er effects on palladium based methane combustion catalysts. Applied Catalysis A: General, 2001. 209(1 2): p. 415 428.

PAGE 60

51 30. Hayes R. E. et al., The palladium catalysed oxidation of methane: reaction kinetics and the effect of diffusion barriers. Chemica l Engineering Science, 2001. 56: p. 4815 4835.

PAGE 61

52 APPENDICES

PAGE 62

53 Appendix: A Before reaction: After reaction: EDS map after reaction: SEM image Pd Si O Figure A.1 SEM and EDS images of 6 H SiC

PAGE 63

54 App endix A (Continued) Before r eaction: After reaction: EDS map after reaction: SEM image Pd Si O Figure A.2 SEM and EDS images of 6H PSC

PAGE 64

55 Appendix A (Continued) C Figure A.2 continued Before reaction: After reaction: EDS map after reaction: SEM image Pd Figure A.3 SEM and EDS images of t ype II 4H PSC

PAGE 65

56 Appendix A (Continued) Si O C Figure A.3 continued Before rea ction: After reaction: Figure A.4 SEM and EDS images of 4H PSC

PAGE 66

57 Appendix A (Continued) EDS map after reaction: SEM image Pd Si O C Figure A.4 continued

PAGE 67

58 Appendix A (Continued) Before reaction: After reaction: EDS map after reaction: SEM image Pd Si O Figure A.5 SEM and EDS images of 4H SiC

PAGE 68

59 Appendix A (Continued) C Figure A.5 continued

PAGE 69

60 A ppendix : B Table B.1 Components used in the catalytic reactor test bed COMPONENT DESCRIPTION COMPONENT NUMBER QUANTITY VENDOR Digital Pressure Gauge EW 68111 20 1 Cole Parmer 316 SS Tube Mixer E W 04669 56 1 Cole Parmer Gas Flow Meter E W 32915 15 1 C ole Parmer Mass flow controllers 1179A21CS1B 1 4 MKS I nstruments MFC Power supply/Readout Unit 247D 1 MKS I nstruments Cable Connects CB259 5 10 4 MKS I nstruments Cylinder Regulators, 2 stage E12 244B 1 Air Products & Chemicals Cylinder Regulators, 1 stage E11 215B0X 1 Air Products & Chemicals 500gm 1 Dodecanol 4154 1 Lancas t er Synthesis Inc 1/4" .035" SS pipe SS T4 S 035 20 60 ft Swagelok Tube fitting plug SS 400 P 20 Swagelok Tube cap SS 400 C 20 Swagelok Set ferrule set SS 400 50 Swagelok Ball valve SS 33 VS4 2 Swagelok Nut SS 402 1 20 Swagelok S Series fine meter SS SS4 VH 1 Swagelok 3 way ball valve SS 42 XS4 4 Swagelok

PAGE 70

61 Appendix B (Continued) Table B.1 continued COMPONENT DESCRIPTION COMPONENT NUMBER QUANTITY VENDOR 1/4" elbow SS 4 UT 9 2 Swagelok 1/4" tube fitting SS 4 UT 6 400 2 Swagelok 1/4" Ultra Torr Fitting Union SS 4 UT 6 2 Swagelok 1/4" Ultra Torr Fitting Union Elbow SS 4 UT 9 2 Swagelok 1/4" Ultra Torr Tube Fitting Union SS 4 UT 6 400 2 Swagelok Flex Tubing 321 4 X 6 B2 1 Swagelok 1/4" Ultra Torr Tube Fitting Union SS 4 UT 6 400 2 Swagelok 1/4" Ultra Torr Tee SS 4 UT 3 2 Swagelok Quartz tubing 4mm id Type 214 12ft National Scientific Co. Quartz tubing 10mm id Type 214 4ft. National Scientific Co. Palladium Nitrate 11035 2 gm Alfa Aesar Gas Cell FTIR Accessory 162 2510 1 Pike Technologies 25gauge/50mm gas tight syringe (fixed needle) 2 Sci Con Technologies Cut septa 1 Sci con Technologies Shimadzu septa 1 Sci con Technologies Removable needle gas tight syr inge 1 Sci con Technologies Removable needles 1 Sci con Technologies


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200433Ka 4500
controlfield tag 006 m d
007 cr bn
008 031007s2003 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000105
035
(OCoLC)52443436
9
AJL4048
b SE
040
FHM
c FHM
049
FHME
090
TP145
1 100
Gopalkrishna, Akshoy.
0 245
Catalytic oxidation of methane using single crystal silicon carbide
h [electronic resource] /
by Akshoy Gopalkrishna.
260
[Tampa, Fla.] :
University of South Florida,
2003.
502
Thesis (M.Ch.E.)--University of South Florida, 2003.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 70 pages.
520
ABSTRACT: SiC is a hard man-made material and has emerged as an excellent material for a wide range of applications which are exposed to extreme conditions such as high temperatures and harsh chemical environments. These applications range from SiC being used as an abrasive, to a refractory material, to a semiconductor material for high power and high frequency electronic devices. The properties of the material for each application is different, with the semiconductor grade material for electronic devices being the most refined. SiC, with its excellent thermal properties and high resistance to harsh chemical environments, lends itself to being an ideal support for catalyst systems. Various characterisation & analysis techniques such as Scanning Electron Microscopy (SEM), Energy Dispersive Spectroscopy (EDS), X-ray Diffraction (XRD), Fourier Transform Infrared Spectroscopy (FTIR) and Gas Chromatography (GC) are used in this thesis to investigate the suitability of single crystal SiC for high temperature catalytic systems. Low temperature oxidation of methane was used to investigate the catalytic activity of: Porous and standard 4H-SiC with and without Pd Porous and Standard 6H-SiC with and without Pd. Nanocrystalline Beta-SiC powder with and without Pd. Part of the samples were impregnated with Pd using Palladium Nitrate (Pd (NO3)2) which is a common precursor for Pd. Activation treatments which were investigated were oxidation and reduction. Oxidation was generally better in activating the catalyst, as was expected, since the PdO phase is known to be more active in oxidising methane. A mixed set of Pd and PdO were observed by SEM and EDS which were the main characterisation techniques used to analyze the structure of the catalysts before and after the reaction. The Beta-SiC showed by far the best activity which could be attributed to the micro-crystalline powder format in which it was used, where as all other catalysts studied here were derived from crushed wafer pieces. Type II porous 4H-SiC was another of the samples which registered impressive results, vis--vis catalytic activity.
590
Co-adviser: Saddow, Stephen E.
Co-adviser: Wolan, John T.
650
Silicon carbide.
Methane
x Oxidation.
653
methane oxidation.
silicon carbide.
690
Dissertations, Academic
z USF
Chemical Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
949
FTS
SFERS
ETD
TP145 (ONLINE)
sv 6/16/03
4 856
u http://digital.lib.usf.edu/?e14.105