‘5'I47II‘II ”,5. III ,..II' 1 wfiw 77'7" 7".77 . 7 I II II III I'" “if:
, '1: )7 "I‘ 777 ”"177. ":7'",-'7 “I: '777i‘77757t77777777‘77771777777777 "H7,“ 7'I77'I7I 7:77:77RP' .nII “fl EMA“

’3 I§I.3I:I.:'II "772?" 'II777 7I7 7777M II‘J'I'“ 7": 77 ' 7I77I 7777771777537: 7"”I’I7III'II31' 7&7?77"'5:J '7'!
7‘71: ‘77. 37:19:. ' :7 I717 7‘-,"I‘I-7I.’ I, I I7'7j 777 7' 77 77 77 77:;77739‘ 77 ',;II"'!h{ ' 7
777: : l 2&17'7I'IIVLI17 I. }77"77' 777 'I77 77III' 7I77 'II' .I'II'I i'§::"7'7:7 77777777 {7:177:73‘777377

I I I 'l' I, 1",, I

I . . .2; . I' II‘IzIEI‘IP ',1,|'7u737'7;':"7l:"I'I7‘:|I|:7':I77II7'II1"IH
‘II I "7; 77777:; li""'I‘$ 77|7"‘ ("7 I!" ’ 'I:'777'7\ '

i " I: I. ‘I'ImIL, ,. ,II.I|I:..‘I 77:717.,h77’ "HEW!" 77;E7L7EI7,;:IIE&iEI:.;77H 'Isz' i' in,,L,I
.5!’ III, 1.7'.'." 'I'77"'"I'IT"7"'... ..I’I-'I:.HII’-'II'I r" “‘II‘II '7
II

 

7

,1

     

   
  

A
V

x
'
~_

.
'z'
gt.’ a
.
2.-
..‘..l-
--
'25..
_/
-
-_-'--.-'
...
.4:
_.;.
“I“:
-
”..—
_..‘
b.-

7777;77:717
“'0‘!" V", I: , 'I
7'I' ' ":7’ {'7'7 “"’ I"I"I '-' I:5 "III" II 77:37." ":1 ‘7 II'W' I‘I'L'77EEL'EW 3177777
7| 71!;775or 7"777777777 .hl,: '_"I ' I0 ' ..
77'II 'HIU7'7 7717:." '7577'1'7I “{77'I 7:" I7‘"' -.-.I; IIII'I‘I.III:I"ItTI-4II7 7'” .
- 7'31: IIIU' 'tIIL-‘IIIIII'I'I‘IIi'II .H' 71"I7III7I‘ ,jI.. "LEI I'I7'II 7 7E.
I ; v ‘ , . . ' hl ,
1 .' 'Wm'""'77" '“'7I7"7'I77'II'-'I7‘7‘I’IIII.:II In: :7 IIII"7"77 Mi; "'I|II IIIII II‘II'II17 :79. fig}- :sz "77$
1 II~LIII III :Jffi'i" 1,,7I7IISI II,.,I.IIIIII I;;;fI.I 71‘ 'If‘. “1,5? "7"II!""77‘7 13,}, ”I: ,4! .I I
‘

 

'LIII’.

 

:7 I ,7‘ 7“] 5 ‘ ‘7 l7 J E‘ '. .‘ I H . '
' '“uJ :‘E’I: 321:7 '7 I77 |:;| {77‘1' ' 77‘7“ ' .7'7177 7'1" .| Ir‘:777'"1'j7~,' 7:777" "77 77:7;7' I77 9 I '7' 71'1" ‘ 1’37 F‘fi‘flflb 377;)“: 77.
h, 7777777777'1'15777hl7’7 .,7,',, .’ , '7..lll::'l|;‘, :II.I‘.I ;-,:II:.II,:,I:;I IN , ._|1,,.~;|;II:;I .I, I:I+.,.I,;f4;v,|,I,;fl,. I. :~:-.I‘.,. ‘,I,$7’.:‘-,.L' l,v
" 7 "3,, IL=’II‘~‘=I‘I:-I'EII "' I'I'I’ II'"? ‘ I'II .i‘I'I" (”II'I‘II ‘IIII III'I ‘I' 'III. IIII‘II I§;’I.5.I4I."¥II. .' I.
77 77“; :0 I7 . I777: “I"III. . ,5‘I . ‘ ", ::~.-,,?x'i'§ "-YIIII"
[ {'1' .7:77 577735.; :1": 7,757“- '777 77'7“”; 'I7'~ '77777'7' 7777 777777777 7 7 ' ' '7' 7I'7I 17: '7' ~I'77"'7"'.II7 ’7' 7':""‘7' .'-II ENS; 75v, 7,77 777673747337#5775777?'71‘77 M '77
‘ ‘-‘; ‘ ' I. ‘ ' II I . I III 1" u" o I '- I' I
E IN 7E, 2.]. 775,. 77:7 7‘7 ,7 .I777u7l77777 (I7 7:7; 777747,“ 77 7‘: 777.7 "7) t . _.‘ I' '33: s 77.71; '7‘, . 77:7"( “-7 77:377f7:..77"‘3:-1 7
«adufu .-;'_'7I7H ',I.,}:. ”‘1 _ I ; 777E77777 ‘1‘. 77: H .II .. | I g I" ,0“ I In Q7\".t..:7;‘vli4.7‘l:b’ 72577'7'7I f." v ', .I_,.
,. IE7: I . I 33411;. ;, 77.,::2; I” . . ',-':'.7. 'I,| {77 l I 11,”. my": I II M, .. ;.,I ,, 1‘” |..r".,;§;v'l,:-;I .51! ., :hu
‘. I ,L, ‘ ‘7, 01"" t I :.I:;, l : | “1;. '1. fl "7:777 ‘ 9;; 9;, :.I; .| 7.:
. I .‘ '.'I' ‘l7 " '7 | ..‘I . I' 37:. I
7 .7031 '-I '07 II 'l;. . H1: ..' :' ;'II.II,.II|’I it}. ;I, I
u |. l ,Zz'r;...;!;.;. ;;,- 0-,:1.’ 1 1" ’1: ..’ ,I,, I I, I 3.") : e 2: ..': ,'|.; ,: .;_ I I ;
';I;,II,*,I, -.-";' ,I,-, g“,- 5, ,:II;;II ”:III ,,;.I,. .;- Vial-75' ,_'..I, «2-: MN.” ,,.I:; I ,,.
"1'7: :I-“I' I. ..«JI‘V' II. "I ‘r'f'IIIII:"' ..
. .‘t, ‘1: L7) \ .I 79.79" h . 77:24:; 2:377:7777:3'7:77 “3-350”. ,7!“ 1;»: I ”2:, . t't.|3 "h, ‘7'”.7: . ,
I .'~; “"‘II .I"‘.' .7. .. i-J‘ I'.' I15} ;_. '
..‘ 7'7 v. I Lit-7V '7’t777‘ 'a;?,;;' ;-." . ',;|M I . MAIN“ ;a _I
. ,. I. ,I l I O, l . w.
I I I II c I, 5 'I .. ~ a II .1.
.I - 7 . I I7_-7I‘ - - - I; I I I , I. I I-
':-7I-: I727 7 7 7 7 0 \ § 77 I 7' 7 - I! ’ - -I
- :‘:I7"'-I~- I I" "7 '. ‘ - I . I ' ‘ -Ii" .II' -
7 7 77 777' 7I7777 777177.777“: 777:3- ’77 7 7 7 s7 a . 7 7 7.777 I 777 7 I. 7 -
., . I . '72,. . ‘.,7; ‘7 _ I ' | 7 7 ' . '2‘ ,. I . 'f'," I 7: -
' ‘ ‘I | ‘ I .- I - "" u ' ' I' ,-
I7 7 ' I "31 ' I' ':" I 7 7 '7 L
I» ;-2".I . I t'?- ' " " ' ' 7 7’-
,7' .7,:,'.. .7“I: 7;. I 7 , I I. I. . _, I' 77 . {‘34: 77 ’
..‘ ’ 1.7.77 ..' ; \ 7" C . - . - , I o ..‘ . {II‘ 7,, I :I 7'7 ‘ ,, ; ‘5'
'77« 7‘7" 77757 7 *I ' 7 '7 L 7 ’ 7 I 77.; 7 “'7
I 711- .‘I 7‘I 7.1.7 I II ' ' ' 7~ ,7 I '- .- 7 ." . “.. 0
. 'I 'I . I '- I’ I. :- --' -
f; . II II ' . II . :a I
I17. 777 ' 7 7 :07 7 7' 77 7 7 : ' '- ' I 77' 77" l'
,. ',"I7 .I;1I , 'I_ ,. .g 1 . ' - "If" 7. "' .‘
7,“I 7. I7 7,.I-c , 7 . .‘ ' 7.7'77l .7 I 7‘8; I: n, ' I" o
'73. ‘73:: Io;‘|. , I , :u I ' c , ., :il.7:7 r ..‘. . ‘I,:.:f| : 'I . II
0' I - 3 ‘
J" ' . I: ', g - ;, ,I,7 I I 717.7
. I I: II I II'» ,I' .
I' | E ,|-NII t ' 1,. - . .,;v ".5, ..’
l I ' I I 7. , I ' ‘I I L
I-I'I I: I. . . I I III ,,, , I_I,I‘ I'I:,I,,
7 1‘7 7 :7I ‘7Il' ‘.u'. I . ' " I ' 77 ; 777.'Z,I"71.7':;l77"73.'
'1' , l7 I'II,;7 II| 7I:|7. /' |' , l I. I." 7 .‘ I 117',77l7..' ~77" .,;\ 7
E 5:073;,|Ii|,_,: 7,||,;.,I I ‘I: : , 7 ’ ' 7 ,‘7,| I: ,;' 7 I. :I\I.:'_'7 I77 fit, 7177‘77'i77'l7777:7w'f7712L‘77;'07::
. II' II I‘ - 'I I I'I'IIrI I‘I II‘-I'- II ._ III .I. ‘ II. . 77"; ‘I'II " -II- . 'I
h 97 II,T;',I: 7.! ‘ ;I,, | I, II'7| “I. I‘l u, .I' I. I' 'I, ;.,";,'., II | . . .7'777;--.-
I |§:I|I;|I|I|.' 'l,’,' I. III' IIII' 1,1It, ' I..I I,'.I' ,'I' '|~I;II‘I I}: ,.; 7 """d, I'M "’ 4477’ III I‘ “'1‘,-,
g I a". ‘ I I. I ,' ".7 ,' ")1 ., I . ' ,. I,7.,I ”777177;. In."77 , ‘
I ”n .‘I I, I_I'I ‘I I,:|I{7'11I:l7:£t:7|h7'1‘ |I
I I.IIII ,.:,.III I I7|§7n III II.I‘.II IIIII ”I ,,.I ..; ':'E: 'I I.IIIIII I
: :I;|:7t7" '|7,.‘”' I‘7 .477I'7 ..|I ': 'H'H !"::,",'|||1 I'- h 77':|'7 |I||\II| I,”,IIIt|I7 MIN, |,‘P'I I:I'IIII| 77"" 7711!,9,II1,|II:1;I-
77I"7 7"I"I:77 'I7'I' |II’I-I 'I.|‘ ' _: II" l. :'I H: 7'!“ I,I " 7:7"; '5‘”: :IIII' ""‘I ”if. III: II'”
‘17":7':'HII'I".I‘;."":" I"I‘I 'FI |"I‘|" I,II||| ,HIIHIII I""I|7I"" if 777L '3'9:
,, 7II , . I
III; III||:; 1"!“ III: | . 7|: {III N W, .2'55‘I ' I " ’ ':",‘II III] E "7 ' ""I" 43:1,?" EtizIfI
'27s f'HE I: ff:
,7. §

   
   

  

.II, ,7.” I,.,, .:.IIIIII:,I: ”5|
:9in ,fl, ,‘7, "III 77'7"" 7., I ' ,7.'7I.II|I,,,I.7.II. :: III'E‘I |
I I II I I I I
'7'77777 ":I77'. II'II

Eh!

I

‘1 E; *3:

I:I||LII’I 1“:
I771

2.4%

        

  

“1‘7
P71

I

773777!

     
 

I
If.

Y

 

["71 I...II.I.

 

II}

V; - . i" '7IL,‘,1 I77g'III;I1I:. I ':7IuI.II ‘ M ‘ I
A»: :':f777 7""“I ,, ”in“, .I .,, .,,II,. ,II,I;77 I‘IIII'..II IIII: IIIIIJ,II,,L,,, ‘III'.2'(.I§?:,I~III

III

fun“.
:II
III, F,

 

$.11..—
u
.3:
ca.
3253.”?—
” "..‘
QC...
Iaxrm'g‘
. w - HI "'
., .43....— .-E
m... ‘_
A; a; ..‘:
“run-A '.
.....7.."‘"‘
.- ?‘A—-oq
’11,,“
.121
:33“?

 

”hf—‘wr—VFVT.‘ «gm u: :s .. $19.25.??? *Mam‘w,uprwnmmo

    
 
 

:' 77":
3:71:77

     

   
     

 

I

.
m AXE. «, .u—am

‘
.
-.. . . ..
“-30:1-
' W -‘

 

7:4."
SEE 1‘3

I I I 5
1:43. 23?“! IIFIIIE:
3' 7 37.77:: 74:514.?
"I' .- I' I ’IiIIfiI"

1 ‘I :15. I

‘I I...

llllllHllllllllIllUH“lllllllllllllllllllllllllllllllllllll

31293 00904 9416

This is to certify that the

thesis entitled

Preparation and Characterization of
Vanadium Oxide/Aluminum Antimonate and
Supported Antimony Oxide Catalysts

presented by

Thomas John Curtis

has been accepted towards fulfillment
of the requirements for

M. S . degree in Chemis try

 

 

    

ajor professor

 

0-7639 MS U is an Affirmative Action/Equal Opponunity Institution

 

 

 

 

» LIBRARY
Michigan State
University J

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or baton date due.

DATE DUE DATE DUE DATE DUE H

__J
:1;
~ ‘1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Tl—T

J
MSU Is An Affirmative Action/Equal Opportunity Institution
cmmma-m

 

 

 

 

PREPARATION AND CHARACTERIZATION OF
VANADIUM OXIDE/ALUMINUM ANTIMONATE AND
SUPPORTED ANTIMONY OXIDE CATALYSTS

By

Thomas John Curtis

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Chemistry

1993

ABSTRACT

PREPARATION AND CHARACTERIZATION OF
VANADIUM OXIDE/ALUMINUM ANTIMONATE AND
SUPPORTED ANTIMONY OXIDE CATALYSTS

By

Thomas John Curtis

A variety of bulk and surface characterization techniques have been used to
examine the structure and reactivity of novel V/AleO4, Sb/Al203, and Sb/Si02 catalysts.
This work will establish the active phase(s) of these catalysts.

V/AleO4 has been charaterized using XPS, XRD, IR, solid-state NMR, and EPR.
XPS V 2pm binding energies for the catalysts indicate that V is in the +5 oxidation state.
For V205 loading 5 20 wt.%, XPS data suggest that the V is highly dispersed over the
Ale04 carrier. For loading 2 25 wt.%, XPS indicates that V is poorly dispersed. These
results are consistent with XRD and IR data obtained. Propylene oxidation data shows
that vanadium promoted sample shows a greater selectivity to acrolein (90%) for 5 wt.%
vanadium than to that of the pure support (59%). The increase in selectivity of propylene
to acrolein is due to V=O bond.

Sb/A1203 and Sb/Si02 catalysts have been characterized using XPS and XRD.
XPS Sb 3d3,2 binding energies indicate that the Sb is in the +5 oxidation state. XPS data
indicates that Sb is highly dispersed over the alumina and silica supports. CO oxidation
activity is highest for catalysts with active phase loadings of 10-20 wt.% Sb205 on both
A1203 and SiOz. CO oxidation activity over Ale04 is greater than that measured for the
Sb/Al203 catalysts.

To my parents, Thomas and June, my sister, Denise,
and my nephews, Nicholas and Alexander

ACKNOWLEDGEMENTS

I would like to thank Dr. Jeffrey Ledford for his guidance and support during the
course of this work and for all the knowledge I obtained in our conversations. Without
his support this project would never have been accomplished. I would like to thank my
parents, my sister and my nephews for their support and understanding during the past few
years. I would also like to thank the Ledford group members, Jeff R., Paul, Kathy, Greg,
Mike, Mark and especially Ed and Per with their help with ChemDraw and Designer. 1
would like to thank the rest of the graduate students, especially Tim, Scott and Doug for
their friendship during this time. Finally, I would like to thank all people who have

provided assistance on this project.

iv

TABLE OF CONTENTS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Page
List of Tables - -- _ - -- -- - - .... _ _ vii
List of Figures viii
Chapter 1: Introduction to Selective Oxidation Catalysis..............-..- . -- 1
1.1. Overview of Selective Oxidation - - ........... - - - 1
1.2. Overview of Methane Conversion _ - - - - - - - -_ 8
1.3. Overview of CO Oxidation- ...... - - - 14
1.4. Focus of Research _ - - - -- _ - ..... 16
1.4.1. Selective Oxidation.--- - -- - 16
1.4.2. Methane and CO Oxidation _________ - 17
References -_ -- - ..... -- ...... 19
Chapter 2: Preparation and Characterization of Mixed Metal Oxide Catalysts ............... 22
2.1. Preparation of Catalysts - ......... - .... _ 22
2. 2. Characterization of Heterogenous Catalysts ...... - -- . 26
2. 2 1. Bulk Characterization -- - ........... _ . . ..... 26
2. 2. 2. X-ray Photoelectron Spectroscopy (XPS) - 28
2. 2. 3. Catalytic Activity--- - -_ - ..................... 34
References - -- -- - -- 38
Chapter 3: Vanadium Oxide/Aluminum Antimonate------- - 39
3.1. Introduction - -- ...... 39
3.2. Experimental - _ - - - - - - 40
3.3. Results - - _- - -- - - _ 45
3.4. Discussion - _- 59
3.5. Conclusions - - .......... - --.....64
References _ - -- - - ..... 65
Chapter 4: Antimony Oxide Supported Catalysts -- - -- _- - --_-67
4.1. Introduction - ........... 67
4.2. Experimental -- -_ ....... - - 69

 

 

 

 

 

 

 

 

 

 

4.3. Results 73
'4.4. Discussion _ - 83
4.5. Conclusions ..... -- - ............. . --86
References - _ ..... _ ...... 86
Chapter 5: Future Work- -- -- - - ................ - 89
5.1. Characterization __ _ - - _ -- 89
5.2. Vanadium Oxide/Aluminum Antimonate Catalysts- - - ..... _ - 90
5.3. Antimony Oxide/Supported Catalysts ......... - 93
5.4. Activation Energy ................. 94
References 95

 

Table

Table 3.1.

Table 3.2.

Table 3.3

Table 4.1.

Table 4.2.

List of Tables

 

 

 

 

Page
BET surface area (mZ/g) of V205/AleO4
catalysts. ........... 46
Particle size V205/AISbO4 calculated from XRD
data. _ -- -- - . ................ -..--51
Conversion and selectivity of V205/A1Sb04 for propylene oxidation at
200 °C and 100 cc/minute flow rate ...... - 58
BET surface area (mZ/g) of SbZOS/A1203 and SiOz.----- - - - . . --.-.-74
CO conversion for Sb205/A1203 and SbZOS/SiOZ at 450°C
and 15 cc/minute flow rate. - - .......... 82

 

vii

List of Figures
Figure . Page

Figure 1.1. Reaction mechanism for selective oxidation and

 

 

 

 

 

 

 

 

 

 

 

ammoxidation of propylene.--- - ................ ........ - -3
Figure 1.2. Mechanism of selective ammoxidation (lefi) and oxidation

(right) over bismuth molybdate.

(FromJ. Catal, 87, 373, 1984).-- - ....... - - ........ -5
Figure 1.3. Reaction mechanism for selective oxidation and

ammoxidation of propane.

(From Ind Eng. Chem. Res., 31, 107, 1992). - ........... - . . 7
Figure 1.4. Reaction mechanism for methane conversion.

(FromJ. Catal., 112, 168, 1988) ................................................................ 10
Figure 1.5. Reaction mechanism for non-catalytic coupling

ofmethane. (FromJ. Carat, 121, 122, 1990). - _- - - - --12
Figure 1.6. Random mtile structure. (From Acta Chem. Scanda. A,

29(9), 804, 1975). - - . ................. 18
Figure 2.1. Various types of active oxide-support interaction.

(From Preparation of Catalysts II, 439, 1979). — - 23
Figure 2.2. Reaction of metal alkoxide with support. (From .1. Carat, 101, 1,

1986). - - - - -- - - - --25
Figure 2.3 XPS process..- - -------- - - - - ---29
Figure 2.4. Model of the catalyst particle. (From .1. Phys. Chem,

83, 1612, 1979). - . - ------32
Figure 2.5. Protocol for study of heterogeneous catalysts by reactive

gas treatment combined with XPS.

(From Anal. Chem, 58(12), 1184 A, 1986).. ----- - - -- 35
Figure 2.6. Reaction chamber. - -------- - ........ 36
Figure 2.7. Catalytic reactor ------- . - - ..... - - -37

 

viii

Figure 3.1.
Figure 3.2.
Figure 3.3.
Figure 3.4.
Figure 3.5
Figure 3.6.
Figure 3.7.

Figure 3.8.
Figure 4.1.

Figure 4.2.

Figure 4.3.

Figure 4.4.

Figure 4.5.

XRD of V205/AleO4 a) V25, b) vzo, and c) vo

 

 

 

 

 

 

 

 

 

 

catalysts. ......... 47
27Al MAS NMR of a) V25 and b) V0

catalysts. 48
Semi-quantitative XRD for physical mixtures of

V205 and AISDO4. - ------ 50
XPS spectra of V25 catalyst a) 1n propylene/Olee, b) 02/I-Ie,

c) spray coated, and d) as powder--- - - 53
XPS intensity ratio of V 2p3,2/Al 2p as a function of

V/Al atomic ratio ....................................................................................... 54
IR data of a) V205, b) V25, c) V20,

d) V15, e) V10, and 1) V5 catalysts ............................................................ 55
Vanadium Wide-line NMR of a) V25 and b) V5

catalysts- . - 56
EPR V“ spectra measured for reacted a) V25

and b) V5 catalysts. Spectra were collected at room

temperature. -- - ...-.57

XRD Of SD205/N203 a) SD40“, D) SD35AI, C) SD30AI,

d) Sb25Al, e) Sb20Al, t) SblSAl, g) SblOAl,

h) SbSAI, and i) SbOAl catalysts.----- 76
Semi-quantitative XRD for physical mixtures of Sb205 and

A1203. 77

XRD of Sb205/Si02 a) Sb408i, b) Sb3SSi, c) Sb3OSi,

d) SbZSSi, e) SbZOSi, f) SblSSi, g) SblOSi,

h) SbSSi, and i) SbOSi catalysts. --------- 78

Semi-quantitative XRD for physical mixtures of Sb205 and

Si02.- ------------------- 79

XPS intensity ratio of Sb 3d3,2/Al 2p as a fimction of Sb/AI

atomic ratio. - 80

 

ix

Figure 4.6.

Figure 5.1.

Figure 5.2.

XPS intensity ratio of Sb 3d3,2/Si 2p as a function of Sb/Si

 

 

atomic ratio. 81
In situ IR cell. - - - - - - - 91
Gas handling system to in situ IR cell. -- - -- - . .............. 92

 

Chapter 1

Introduction to. Selective Oxidation Catalysis

1.1. Overview of Selective Oxidation Reactions

Selective oxidation of hydrocarbons is an important industrial process that uses
transition metal oxide catalysts to convert less valuable hydrocarbons into more valuable
compounds (e.g., acrolein and acrylonitrile). The most common catalysts used for these
processes are based on metal oxides of vanadium and molybdenum. A common feature of
these reactions is that the desired products are often not the most thermodynamically
favorable ones. For reactions carried out in either air or oxygen, the most
thermodynamically favored products are water and carbon dioxide. The desired products
are alcohols, aldehydes, ketones, acids, anhydrides, or alkenes and dienes [1]. Thus, for
each hydrocarbon, numerous products of various extents of oxidation are formed before
carbon oxides are produced. Hence, it is a challenge to understand the factors that
account for the selectivity and activity of a catalyst and to develop a practical catalyst. In
selective oxidation reactions, the selectivity is usually dictated by the ability of the metal
oxide to catalyze the formation of C-0 bonds while minimizing the breaking of C-C
bonds.

Selective oxidation reactions can be classified into two types: dehydrogenation and
dehydrogenation followed by oxygen or ammonia insertion. In dehydrogenation reactions,
aliphatic molecules are converted into alkenes by breaking C-H bonds and forming C=C
bonds. Often oxygen is used as an oxidant to yield water as a byproduct. The oxygen
additionally provides the thermodynamic driving force for this process. Other oxidants
(e.g., 1, Br, and N20) can be used, but oxygen allows the reaction to be conducted at

lower temperatures [1].

2

In dehydrogenation and oxygen insertion processes, C-H bonds are broken and
C-0 bonds are formed. Oxygen is used as an oxidant for the formation of oxygenates and
in the formation of water in the dehydrogenation step. Exceptions to this general rule are
the oxidation of ethylene to ethylene oxide in which no C-H bonds are broken and
ammoxidation reactions (such as propylene to acrylonitrile) in which C-N bonds are
formed. Oxidation of benzene to maleic anhydride is an example of a reaction that
involves the breaking of C-C bonds and the insertion of oxygen.

Catalysts used for selective oxidation and ammoxidation reactions contain several
types of sites: olefin chemisorption sites, (Jr-hydrogen abstraction functions, oxygen or NH
insertion sites, and redox couples associated with dioxygen dissociation. The olefin
chemisorption site and the oxygen or NH insertion sites are often associated with elements
in their highest oxidation state (e.g., Mo+6 or Sb+5). The (Jr-hydrogen abstraction site
consists of metalloids with free electron pairs (e.g., Bi+3 or Sb”) that give radical
character to the oxygen bond at the site. The presence of radical oxygen facilitates the
abstraction of a-hydrogen as a radical from the chemisorbed olefin (an important step in
the formation of the n-allyl intermediate). Oxygen or NH insertion fi'om maximum
valency metal sites into the n-allyl species and forms the o-allyl surface intermediate.
During this process, reduction of the metal (e.g., Mo+6 to Mo+4 in molybdate catalysts)
occurs followed by desorption of the product. A redox couple with a reduction potential
greater than that of the oxygen or NH insertion site may also be present to facilitate
reoxidation of these elements back to their original active state. The redox element also
serves as a site for 02 reduction [2]. Figure 1.1 shows the general reaction mechanism for

the selective oxidation of propylene.

CH CH=CH
> ACTIVE\ Km 02
SITE 13!le

nM[] REOXIDATION
H ' 0 SITE
:MQ/MZ/m f
“[1
NH
1 3 CHZ=CHCHO + H20
C112=CHCN 4.31120

Figure 1.1. Reaction mechanism for selective oxidation and ammoxidation of propylene.

4
Among the most commercially important examples of selective oxidation are the

oxidation of propylene to acrolein:
CH3CH=CH2 +02-—>CH2 =CHCHO+H20 (l)
and ammoxidation of propylene to acrylonitrile:
CH3CH=CH2 +NH3+%_02—-)CH2 =CHCN+3H20 (2)

which are used for fiirther reactions with hydrocarbons to produce valuable compounds.

In 1948, Adams used cuprous oxide as the catalyst to transform propylene into
acrolein, with a yield of approximately 50% [3]. In 1959, Idol improved the yield of
propylene oxidation to acrolein by using a bismuth-molybdate catalyst [4,5]. The bismuth-
molybdate catalyst was also found to convert a reactant mixture of propylene, air, and
ammonia to acrylonitrile [6]. Commercial vapor-phase oxidation and ammoxidation was
developed by Standard Oil of Ohio (801110). In 1965, Adams [7] found that the bismuth—
molybdate catalyst could produce butadiene from butene. Also, in 1965, a more selective
uranium-antimony catalyst was introduced for the same set of reactions. In order to
eliminate the radiation concerns associated with the use of uranium, an iron-antimony
catalyst was developed. This catalyst is still used in Japan [8-11].

In 1970, SOHIO introduced the first multicomponent catalysts [12,13]. These
catalysts were composed of a variety of elements including Bi and Mo. Based on research
by various workers, Burrington et a1. [14] proposed detailed mechanisms for both
oxidation and ammoxidation over bismuth molybdate. The mechanisms for oxidation and

ammoxidation are shown in Figure 1.2.

Figure 1.2. Mechanism of selective ammoxidation (left) and oxidation (right) over
bismuth molybdate. (FromJ. Catal, 87, 373, 1984).

6

In these mechanisms, the surface active site is a Bi-Mo pair site that is composed
of a Bi-O group responsible for allyic H abstraction and a molybdenum dioxo group for
nitrogen insertion. It is believed that propylene absorption occurs at the molybdenum
cation. The oxidation reaction proceeds by dissociative adsorption of propylene to
produce a rt-allyl species. It is not clear, at present, whether dissociation of propylene
occurs upon adsorption or propylene is first adsorbed molecularly as a 1: complex and
allylic H abstraction follows. The next step is the formation of C-0 bonds and a second
hydrogen abstraction in the form of a (1,4) shift to produce adsorbed acrolein and Mo-
OH. Finally, the product desorbs and the catalyst is reoxidized.

In ammoxidation, a surface molybdenum diimido species is formed by the reaction
of dioxo groups with ammonia. The activation of the propylene occurs in the same
manner as the oxidation cycle. The reaction proceeds by dissociative adsorption of
propylene, followed by formation of a C-N bond and two additional hydrogen abstractions
producing acrylonitrile. The reduced surface site is reoxidized and then reconverted to the
diimido species.

Because of the reduced cost and the greater availability of alkane feedstock,
significant effort has been devoted to understanding the reaction mechanism for the
oxidation and ammoxidation of propane. At present, Cantani et a1. [15] suggest that the
reaction mechanism occurs as shown in Figure 1.3. In this scheme, the propane is
converted to the main product of propylene, which then further reacts to form acrylonitrile
(ACN), acetonitrile (AcCN), hydrogen cyanide (HCN), ethane and ethylene (C2), and
carbon oxides (COX). They suggest that the formation of acrylonitrile passes through the
intermediate formation of propylene and the production of acrylonitrile is limited by the

slow rate of propylene formation.

 

 

 

 

 

 

ACN
v
Propane > Prop lene ’ cox
NC /
9 HCN
> C2

 

 

Figure 1.3. Reaction mechanism for selective oxidation and ammoxidation of propane.
(From Ind. Eng. Chem. Res., 31, 107, 1992).

1.2. Overview of Methane Conversion

Recently, there has been . increased interest in the oxidative coupling of
hydrocarbons primarily using metal oxide, rare earth oxide, alkaline earth oxide and alkali
metal-doped oxide catalysts. The oxidative coupling of methane to form ethylene and
ethane has received considerable attention due to the abundance and low cost of methane.
Interest in oxidative coupling has also increased due to the possibility of conversion to
either gasoline, distillate or other products. The conversion over metal oxides is thought

to occur as follows:

xCH4+(x—1)MO-—>CXH2x+2 +(x—1)H20+(x-1)M (3)

In 1969, Union Carbide investigated the feasibility of using methane coupling for
ethylene manufacturing [16]. They performed extensive studies on different metal oxides
for this process and found that most of the metal oxides of group IIIA, IVA, and VA
exhibit both high activity and selectivity for methane coupling. In 1979, Mitchell and
Waghome of Exxon [17] achieved 40-50% conversion of methane to mainly ethane and
ethylene over a catalyst consisting of a group VIII noble metal, a group VIB metal oxide,
and a group IIA metal. Between 1983 and 1986, Baems et al. [18] studied several lead
oxide supported on ‘y-alumina, silica, and alumina/silica catalysts. Their work showed low
acidity supports are preferred for high selectivity to Cz's. From 1984-1987, Arco [19]
screened a large number of metal oxides supported on silica. They reported that
manganese, indium, germanium, antimony, tin, bismuth, and lead oxides achieved
approximately 15% conversion and give 10-50% selectivity to higher hydrocarbons. They
also found that more acidic catalysts were less selective to C2 formation. Further, they

reported that addition of alkali metal leads to an increase in Cz's selectivity [20,21].

9

Both reducible and irreducible oxides have been used for methane coupling.
Bhasin and Keller [22] found that oxides of Sn, Pb, Sb, Bi, T1, Cd, and Mn were the most
active. They performed the reaction without the addition of 02 in the gas stream. They
concluded that the common characteristic of the active metals was that they could cycle
between at least two oxidation states. Thus, these reactions can be considered as oxide
reduction.

A number of irreducible metal oxides (e.g., MgO, Li/MgO, Na/CaO, Sr/La203,
and Li/T 102) have been studied as methane coupling catalysts [23-26]. In these systems,
oxygen in the gas stream is required to generate and maintain activity, but the oxidation
state of the metal oxide does not change. These catalysts also have an advantage over the
reducible catalyst because they are easier to control.

The oxidative coupling reaction mechanism and the nature of catalytic coupling
site over Li/MgO has been extensively studied by Lunsford et al. [27]. Oxidative coupling
is proposed to proceed via the abstraction of hydrogen by lattice oxygen from the metal
oxide followed by gas-phase coupling of methyl radicals. The methyl radicals can then
combine to form ethane or react with oxygen to form products as shown in Figure 1.4.
The pathway leading to complete oxidation is more complex. High selectivity to C02 is
observed when the reaction is conducted in the presence of 02. This suggests a significant
gas-phase contribution to COX production, possibly by methylperoxy radicals or adsorbed
oxygen. In addition, a surface assisted route evident during the initial phases of the
reaction in the absence of 02 can contribute to COx production [28].

In addition to methane coupling, ethane, and propane coupling have received
considerable attention. Recently, Otsuka et al. [29] showed that at temperatures less than
650 K, Na202 partially oxides methane, ethane, and propane to ethane, butane, and
hexane, respectively. Kinetic studies by Otsuka suggests that the activation of methane is

initiated by diatomic oxygen such as 02", 02'2 or 02 which are adsorbed on the

10

02 (g) 1‘
N CH302 (g/a)
MOx MOx - 1 \

Figure 1.4. Reaction mechanism for methane conversion. (From .1. Carat, 112, 168,
1988)

11
surface [30,31]. The proposed mechanism for these reactions is shown in Figure 1.5.

In addition to the coupling of aliphatic hydrocarbons, there is interest in the
coupling of aromatic hydrocarbons. such as methylbenzenes. The catalytic oxidation of
methylbenzenes by metal oxides to the corresponding aldehyde and acid are well known
industrial processes [32]. In a recent patent [33], anthraquinone was reported to be the
major product of the oxidation of methylbenzene. King [34] recently found that under
anaerobic conditions methyl-aryl coupling reactions occurred to produce
methyldiphenylmethanes from the methyl-substituted benzenes. The corresponding
aldehydes and acids were minor products. From this work, King concluded that methyl-
aryl coupling is more favorable using more basic and ionic metal oxides formed from
elements on the left side of the periodic table. A possible explanation for this finding is
that a methylbenzene molecule may be polarized on the oxide and a proton from the
methyl group abstracted to form OH‘ on the surface. If the multivalent metal oxidant
(MN) is easily reducible, the benzyl anion may transfer two electrons to form a benzyl
cation. The overall reaction involves reduction of the metal oxide and formation of water

fi'om OH' and H". The reaction is believed to occur by the following mechanism:

+
Ar-CH3+Mn+ ——>Ar-CH;+M(““) (4)
+
Ar-CH;+M“+ ——+Ar-CH;+M("") +11+ (5)
Ar-CH;+Ar—CH3—->Ar-CH2~Ar-CH3+H+ (6)

For easily reducible metal oxides, the reaction occurs through the methylene group on the
methylbenzenes. The first step of methylbenzene oxidation on the oxide is believed to
involve the abstraction of (Jr-hydrogen to form water. The oxide at the surface may be

reduced to its metallic form which has very weak interaction with the adsorbed organic

12

 

 

in3 H
CH‘ + N3202 NaO 0N8
C2H6 + N8202 k2 N80 0N8
can, H
CSHB 4’ N8202 A N80 0N8
CH3

k
2 N80 —4—> N8202 + C2H6

 

 

T211.
2 N80 Na202 + C4H10
T3117
II

It
20 Na—7—u. Na20 + H20

Figure 1.5. Reaction mechanism for non-catalytic coupling of methane. (From J. Catal.,
121, 122, 1990).

13

species: A proposed mechanism on PbO is as follows:

ZAI-CH3+PbO——)2Ar-CH2'+PD+H20 (7)
ZN-CHZ'fiN-CHz-CHz‘N (8)

The formation of bibenzyl occurs through methyl-methyl coupling.

In addition to oxidative coupling of methane to higher hydrocarbons, the deep
oxidation of CH4 to C02 has potential for large scale application. Deep oxidation of
methane was widely studied in the 1960's and 1970's when it was hoped that less costly
alternatives to Pt/A1203 total combustion catalysts could be found [35]. The use of
catalysts for total oxidation of hydrocarbons offers several advantages over more
conventional destruction methods such as thermal incineration. For example, the
formation of carbon and nitrogen oxides can be completely prevented by using appropriate
catalyst formulations. The deep oxidation of methane is described by the following

reaction:
CH4 +202 -—)C02 +2H20 (9)

In addition to Pt/Ale3, other catalysts such as Pd, Rh, and first row transition metal
oxides have been used for this reaction [3 6].

Recently, Lacombe et al. [37] have studied the total oxidation of methane pathway
in oxidative coupling of methane over lanthanum oxide catalysts. They found that total
oxidation into C02 proceeds mainly through surface reactions by slow steps which
combine with the initial slow step of methane activation.

In addition to methane coupling and total oxidation, partial oxidation of methane

to CO and hydrogen has received considerable attention. Among the most important

14
commercial methods of synthesis gas manufacturing is the steam reforming of natural gas

and light hydrocarbons. For methane, the reaction is as follows [3 8]:

CH4+H20-—-—)CO+3H2 (10)

Typical catalyst used are CaO and/or K20 promoted Ni/or-alumina, MgO or MgA1204.
Prettre et al. [39] were among the first to study synthesis gas formation by catalytic
conversion of CH4/02 mixtures with 10 wt.% nickel/refractory oxide catalyst. They
found that the compositions of the final CH4/CO/C02/Hz reaction mixture agreed with

thermodynamic predictions.

1.3. Overview of CO Oxidation

In recent years, there has been increased interest in the environment and the effect
of technology on the environment. Air pollution is generated from emission sources such
as electric power generation, refuse burning, industrial, and domestic fuel burning,
industrial processes and transportation. CO is the most abundant air pollutant in the lower
atmosphere.

CO oxidation was first studied by Langmuir [40] in 1922 over group VIII metals.

In general, the reaction occurs as follows:

2c0+o2 ——)co2 (11)

Even though the overall reaction is very simple, the reaction kinetics are, in fact, rather
complicated. For example, when performed on Pd, the rate dependencies of the reaction

vary from negative first order to positive first order for each reactant [41,42]. In addition,

15
surfacedifl’usion of adsorbed CO [43,44] and metal surface restructuring under reaction
conditions [45] also complicate the interpretation of kinetics data. CO oxidation has been
extensively studied on single-crystal. Pd and supported Pd crystallites. Evidence suggests
a surface reaction between an adsorbed CO molecule and an oxygen atom occurs.

Most commercially employed catalysts for CO oxidation utilize noble metals. This
is because they have high intrinsic activity for oxidation, are not greatly deactivated by
sulfur in fuel at temperatures < 500°C, and have good thermal stability. However, at
temperatures between SOC-900°C, the pure metal sinters rapidly. Furthermore, it has been
found that the noble metal can disperse as oxides on supports at temperatures below the
decomposition temperature of the oxides [51].

Although noble metal catalysts are more developed, base metal oxide catalysts are
of importance because of the natural abundance and lower cost. Frontier research in this
area focuses on the development of contaminant resistant and thermally stable base metal
oxide catalysts for catalytic combustion [52].

In 1923, Jones and Taylor reported that copper is an active catalyst for the
reaction between CO and 02 [47]. Ertl [48], Harbraken ct al. [49], and Arlow and
Woodrufi‘ [50] have performed studies on Cu single-crystals and found that high oxygen
coverage retards the reaction rate and that a surface reaction between an adsorbed CO
molecule and an oxygen atom occurs. However, there is no agreement as to structural
sensitivity of the reaction. Since oxygen adsorbs more strongly than CO on Cu, high
oxygen coverage usually exists and reactions rates are lower. Thus, copper oxide requires
higher temperatures to obtain similar activities per unit surface area in comparison with
noble metal catalysts. Prokopowicz et al. [53] have studied CO oxidation over CuO/SiOz
and observed first-order dependency on C0 (characteristic of base metals) and zero-order

on 02 above 4% 02. Huang et al. [54] have reported a strong dependence of

16
pretreatment on the activity of CO oxidation on CuO/y-AIZO3 catalysts and concluded the
reaction is structure sensitive.

In general, high oxidation activity requires metal ions that can have more than one
valence state and can participate in redox reactions. Among the most promising oxides
are C0304, N10, Cr203, and CuO because they are refractory oxides and thus maintain
their activity at high temperature. Mixtures of these oxides often exhibit greater stability
and activity than the single oxide. One such mixture that has been shown to have activity
near that of noble metal catalysts is the copper-chromium catalyst. In this system, it has
been suggested that c0pper oxide is the active component and chromium oxide is a

promoter.

1.4. Focus of Research

1.4.1. Selective Oxidation

The discovery of selective oxidation and ammoxidation of alkenes has been the
subject of much commercial interest. Recent research effort has focused on the
development of selective oxidation and ammoxidation catalysts for the activation of less
expensive feedstocks, such as propane and methane. The catalysts commonly used are
prepared by mixing a hydrosol or gel of the carrier with a sluny of a solution containing
the active phase(s) and the promoters such as W and/or Mo. This procedure can produce
very complicated catalyst structures which makes comprehensive characterization difficult.
Thus, the effect of surface structure on the properties of the catalyst and the active sites
are not well known. In order to understand the influence of the catalyst structure on

hydrocarbon activation, a simple catalytic system must be examined.

 

17

Several authors have reported that vanadium antimonate catalysts are promising
systems for the ammoxidation or propylene [55] and propane [56] to acrylonitrile. Centi
et al. [57] have investigated the. structure and activity for selective oxidation and
ammoxidation of propane for VSbOx mixed oxides supported on SiOZ, A1203, and T102
and promoted with W or Mo. Their work suggests that VSbOx phase plus szOs is not
the active phase for ammoxidation. They proposed that the random rutile phase of
Ale04 (Figure 1.6) modified with vanadium is the active phase for the formation of
propylene and vanadium or other redox elements having M=O bonds (W or M0) is needed
to form acrolein or acrylonitrile. Although VSbOx/A1203 is not as complex as the
commercial catalysts, further understanding of hydrocarbon activation would be obtained
by studying the interaction of the pure phases believed to be responsible for selective
oxidation and ammoxidation.

The objective of this work is to study mixed oxide systems which model specific
active phase-support phase interactions that have been proposed to play a role in selective
oxidation. The work will focus on V oxide supported on aluminum antimonate. The
nature of the V-AleO4 interaction as well as the influence of V dispersion and reactivity

on the catalytic prOperties of the mixed oxide will be determined.

1.4.2. Methane and CO Oxidation

The surface catalyzed coupling of methane has had limited success mainly because
direct coupling of methane is a thermodynamically forbidden process at the reaction
conditions of interest. However, if an oxidizing agent is used in the gas stream, either
partial oxidation to methanol or oxidative coupling can take place. A number of
researchers [58-62] have shown that it is possible to convert methane to higher

hydrocarbons by reaction with metal oxides, thus coupling the methane through the

 

18

 
 

 

    

  

 

 

Figure 1.6. Random rutile structure. (From Acta Chem. Scanda. A, 29(9), 804, 1975).

19
abstraction of lattice oxygen and the formation of water.

Antimony oxide has been used for methane coupling because the melting point of
the compound lies in the low melting region of the periodic table and has been reported to
give high C2 selectivity. In addition, Sb204 is a well known allylic oxidation catalyst
promoter in propylene oxidation [63].

In addition to methane conversion, antimony oxide has also been used as a
promoter for the oxidation of CO [64]. Ali-Zade et al. [65] found that the activity of Cu,
Co, Cr, and Mn oxides supported on y—alumina is improved when Sb oxide is added.
Antimony oxides have also been used in conjunction with oxides of Be, Mg, Zn, Al, Si and
rare earth metals [66] and Sn-Mn-Pb oxides [67] for treatment of exhaust gases.

The goal of this research is to investigate the surface structure and catalytic
properties of antimony catalysts supported on A1203 and SiOz. This work will focus on
catalyst preparation and characterization and will establish correlation between activity
and structural information obtained from bulk and surface spectroscopies. CO oxidation
will be used as a probe reaction to obtain information about the nature of active site(s) for

the reaction.

References

1. H.H. Kung, "Transition Metal Oxides: Surface Chemistry and Catalysis," Elsevier,
Amsterdam, 1989.

2. JD. Burrtington, C.T. Hartisch, and R.K. Grasselli, J. Catal., 81, 489. 1983.

3. G.W. Heam and ML. Adams, US. Patent 2,451,485, 1948.

4. JD. Idol Jr., US. Patent 2,904,580, 1959.

5. F. Veacth, J.L. Callahan, EC. Milberger, and R.W. Foreman, "Proceedings, 2nd
Int. Congr. Catal,” 1, 2647, Technip, Paris, 1960.

6. F. Veacth, J .L. Callahan, and EC Milberger, Chem. Eng. Process, 5b, 65, 1960.

7. CR. Adams, ”Proceedings, 3rd Int. Congr. Catal.," 1, 240, Amsterdam, 1965.

8. IL. Callahan and B. Sertisser, US. Patent 3,198,750, 1965.

9. G.K. Boreskov, S.A. Ven'yaminov, A. Dzis'ko, D.V. Tarasova, V.M. Dindoin,

N.W. Sazanova, J.P. Olen'kova, and L.M. Kefeli, Kinet. Karat, 10, 1109, 1969.

 

10.

ll.
12.
13.
14.
15.
16.

17.
18.
19.

20

21.
22.
23.
24.

25.
26.
27.
28.
29.

30.
31.
32.

33.
34.
35.

36.
37.
38.
39.
40.
41.

42.
43.

20
T. Yoshino, S. Saito, J. Ishikara, T. Sasaki, and K. Sofugawa, Japanese Patent
7,102,802, 1971.
T. Yoshino, S. Saito, and B. Sobukawa, Japanese Patent 7,103,438, 1971.
RK. Grasselli, G.H. Heights, and IL Callahan, US. Patent 3,414,631, 1968.
R.K. Grasselli, G.H. Heights, and HF. Hardman, US. Patent 3,642,930, 1972.
JD. Buntington, C.T. Hartisch, and R.K. Grasselli, J. Catal., 87, 363, 1984.
R Cantani, G. Centi, and F. Trifiro, Ind Eng. Chem. Res., 31, 107, 1992.
M.M. Bhasin, "Methane Conversion," Elsevier Science Publishers, Amsterdam, p.
343, 1988.
H.L. Mitchell III and RH. Waghorse, US. Patent 4,172,810, 1979.
W. Bytyn and M. Baems, Appl. Catal., 28, 199, 1986.
J.A. Sokfranko, J.J. Leonard, and CA. Jones, J. Catal., 103, 302, 1987.
CA. Jones, 1]. Leonard, and J.A. Sokfi'anko, J. Catal., 103, 311, 1987.
J.A. Sokfranko, 11 Leonard, and CA. Jones, Energy and Fuels, 1, 12, 1987.
GE. Keller and M.M Bhasin, J. Carat, 73, 9, 1982.
K.D. Campbell and J.H. Lunsford, J. Phys. Chem, 92, 5792, 1988.
Y. Chen, H.T. Tohver, J. Narayan, and M.M. Abraham, J. Phys. Chem, 16(12),
5535, 1977.
C-H. Chen, J-X. Wang, and J.H. Lunsford, J. Catal., 111, 302, 1988.
J.M. DeBoy and RF. Hicks, J. Catal, 113, 517, 1988.
CHH Lin, K.D. Campbell, and J.H. Lunsford, J. Phys. Chem, 90, 534, 1986.
M.-Y. Lo, S.I(. Agarwal, and G. Marcelin, J. Catal., 112, 168, 1988
K. Otsuka, Y. Murakami, Y. Wada, A.A. Said, and A. Morikawa, J. Catal., 121,
122, 1990.
K. Ostuka and K. Jinno, Inorg. Chim Acta, 121, 237, 1986.
K. Ostuka, A.A Said, K. Jinno, and T. Komatsu, Chem. Lett, 77, 1987.
Kirk-Othmer, ”Encyclopedia of Chemical Technology,” 3rd. ed., 3, p. 738, Wiley,
New York, 1978. '1‘
K. Kojo, Japanese Patent 1,883,121,238, 1983. .
S.T. King, J. Catal, 131, 215, 1991.
Catalysts for the Control of Automotive Pollutants: Advances in Chemistry Series
143; JD. Mc Evory, ed.; American Chemical Society: Washington, DC, 1974.
EH. Foules, J.H. Labinger, J.L. Beauchamp, and B. Fulty, J. Phys. Chem, 95,
7393, 1991
S. Lacombe, J.G. Sanchez, M.P. Delidore, H. Mozzanega, J.M. Tatibouet, and C.
Mirodatos, Catal. Today, 13, 273, 1992.
D. Dissanayake, M.P. Rosynek, K.C.C. Kharas, and J.H. Lunsford, J. Catal., 132,
117, 1991.
M. Prettre, C. Eichner, and M. Perrin, Trans. Faraday Soc, 43, 335, 1946.
I. Langmuir, Trans. Faraday Soc., 17, 672, 1922.
T. Engel and G. Ertl, J. Phys. Chem, 69, 1267, 1978.
PI. Berlowitz, HF. Peden, and D.W. Goodman, J. Phys. Chem, 92, 521, 1988.
W.H. Weinberg, C.M. Comrie, and RM. Lambert, J. Catal, 41, 493, 1976.
B. Mantolin and E. Gillet, Surf Sci., 166, L115, 1986.

11mm; ..1 h"! ht,”

 

 

1w-

45.
47.
48.
49.

50.
51.

52.
53.

54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
65.
66.

67.

21
I. Keverekidis, L.D. Schmidt, and R.Aris, Surf Sci, 137, 151, 1984.
HA. Jones and H.S.Taylor, J. Phys. Chem, 27, 623, 1923.
G. Ertl, Surf Sci., 7, 309, 1967.
F.P.H.M. Habraken, E. P. Keifer, and GA. Bootsma, Surf Sci, 83, 45, 1979.
F.P.H.M. Habraken, C.M.A.M. Mestera, and GA. Bootsma, Surf Sci., 97, 264,
1980.
IS Arlow and DP. Woodruff, Surf Sci, 180, 89, 1987.
IR Prasad, L.A. Kennedy, and AB Ruckenstein, Cara]. Rev. Sci. Eng., 26,
1,1984.
B.L. Yang, S.F. Chan, W.S. Chang, and Y2. Chen, J. Catal., 130, 52,1991.
RA Prokopowicz, P.L. Silveston, R.R. Hudgins, and DE. Irish, React. Kinet.
Cami. Lett., 37, 63, 1988.
T-J. Huang,T-C. Yu, and S-H. Chang, App]. Carat, 52, 157, 1989.
F]. Berry and ME. Brett, J. Catal., 88, 232,1984.
G. Centi, D. Pesheva, and F. Trifiro, Appl. Caral., 33, 343, 1982.
G. Centi, RK. Grasselli, E. Patane, and F. Trifiro,"New Developments in Selective
Oxidation," Stud. Surf Sci. Catal., 55, 515, 1990.
GE. Keller and M.M Bhasin, J. Catal., 73, 9, 1982.
IT. Ali Emesh and Y. Amenaniya, J. Phys. Chem, 90, 4785, 1986.
W. Hinsen, W. Bytyn, and M. Baems, "Proceedings 8th Intr. Congr. Catal., Berlin,
1984,” 3, p.581, Dechema, FrankfiJn-am-Maion, 1984.

J. Driscoll, W. Martin, J.X. Wang, and J.H. Lunsford, J. Amer. Chem. Soc., 107,
58, 1985.

T. Ito and J .H. Lunsford, Naturedondon), 314, 721, 1985.

V. Fattore, Z.A. Fuhnnan, G. Manara, and BJ. Notari, J. Catal., 37, 215, 1975.
IL Mikhailova, I.S. Sazonova, N.P. Keier, and TV. Belodtseva, Kinet. Karat,
9(3), 565, 1968.

FM. Ali-Zade, N.M. Mardanova, PM. Sorgin, RM. Talshinskii, A.A.
Medzhidov, and RG. Rizaev, U.S.S.R. Patent 49,052,169, 1991.

S. Blumrich, R. Brand, B. Engler, W. Honnen, and E. Koberstein, European
Patatent 410,440, 1991.

Y. Ohno and K. Morishita, Japanese Patent 49,052,169, 1974.

 

Chapter 2

Preparation and Characterization of Mixed Metal Oxide Catalysts

2.1. Preparation of Catalysts

Recent advances in the preparation of catalysts illustrate the importance of carrier,
impregnation technique, calcination temperature, and activation treatment on the structure
and properties of heterogeneous catalysts [1-3]. A number of different types of
interactions between an active metal oxide and the support are possible (Figure 2.1).
Among them are the weak interaction between the supported oxide and the carrier,
aggregates of deposited oxide electronically interacting with the carrier, metal oxide
monolayers, covalently or ionically bound aggregates, solid solution of the active phase in
the support, and new surface compounds. The objective of this research is to employ
sophisticated synthetic procedures to control catalyst structure, thoroughly characterize
the surface and bulk of the catalyst, and correlate catalyst structure with catalytic
properties.

The term impregnation generally refers to a procedure whereby a certain volume
of solution containing a compound of the active element is contacted with a high surface
area support. The most common methods of impregnation include wet impregnation,
incipient wetness, grafting, chemical vapor deposition, and co-precipitation.

Wet impregnation usually involves contacting a porous carrier with a solution of
the active element to produce a suspension. The suspension is dried and the catalyst is
activated by converting the dried phase to its active form through various physical and
chemical changes (e.g., oxidation or reduction). Incipient wetness involves impregnation

of the carrier with a solution whose volume corresponds to the total pore volume of the

22

 

23

 

 

 

 

    

 

 

 

 

 

' 2 cl .....;9. r 5
, 379—19: i—ououo‘
' , , . ° ‘, it 'a .1. o A m
y I / ///;'// / .‘l // /,/_ ./I ,’ [0 'A O ‘ 0 A o l‘
1 /1 1 1,1 1 1 1 1 1' ’ 1/1/ 1. 1 1'" 1‘ 1’ 1 1 1 A J
3 Weak iorces 13 Junction c Monolayer

 

 

 

 

 

 

 

r ' .
I .
J i A 4 A 1 A A‘ ' A A 4 J

 

a Chemically bound e Soho sotution t New compound
aggregate 1n support with support

 

Figure 2.1. Various types of active oxide-support interaction. (From Preparation of
Catalysts II, 439, 1979).

24
carrier [8]. The loading of the active element is controlled by varying the concentration of
the impregnating solution.

Chemical vapor deposition. (CVD) involves the use of volatile inorganic or
organometallic compounds to prepare the catalyst. This method has been found to
produce well-dispersed active components on a variety of carriers [4—7].

The active and supporting oxides or their precursors may be co-precipitated from a
solution containing compounds of each element [8]. This usually produces an intimate
mixing of the active phase and the support. From this close proximity with the support,
the active phase is dispersed throughout the bulk as well as the surface of the catalyst.
Other less involved methods of impregnation include heating a physical mixture of the
oxides to induce spreading of the active phase over the supporting oxide [9].

Recently, grafting reactions have been used to obtain monolayers of various active
phases [10]. This preparation method involves a chemical reaction between the active
element (often in the form of an alkoxide) and the hydroxyl groups of the support. One of
the alkoxide ligands reacts with a surface hydroxyl of the support to produce a support-
active phase bond and the corresponding alcohol. During calcination of the impregnated
oxide, alkenes are formed from the destruction of the surface phase. The typical reaction
used to produce vanadium oxide monolayers is shown in Figure 2.2.

The grafting method usually involves wet impregnation of the support with a
nonaqeous solution of the active element which is then filtered and the impregnated
catalyst (filtrate) is washed. This impregnation procedure has been modified slightly for

this work by employing incipient wetness impregnation of the appropriate solution.

25

 

 

i—C4H9—0 O—i—C4H9
1.1—OH .1—c,H9_o_v=o _. Al—O—Y=O . i- C4H90H
i—C4H9—0 O—i—C4H9
Q-i-C4H9 OH
Al-O-Y=0 AT Al—O—\:7=0 . ZC4H3
O-i—C4H9 OH
o—i—cmg
Al-O—V=O on
/ o-i—c.,H9 /Al—O—Y=0
0 AT 0 o +4C4H8+H20
9'i-C4H9 \Al-o—ir=o
“TO—Yd) 0H
o-i—c,H9

Figure 2.2. Reaction of metal alkoxide with support. (From J. Catal., 101, 1, 1986).

 

26

2.2. Characterization of Heterogeneous Catalysts
2.2.1. Bulk Characterization

In order to understand the effect of preparation method, composition and
pretreatment on the structure of heterogeneous catalysts, a variety of bulk characterization
methods should be used. The bulk characterization methods that provide important
information in this study are the Brunauer-Emmett-Teller (BET) method for measuring
surface area, X-ray diffraction (XRD), nuclear magnetic resonance (NMR), infrared
spectroscopy (IR), and electron paramagnetic resonance (EPR).

Surface Area. Catalyst surface areas can be determined using the Brunauer-
Emmett-Teller (BET) method. In this method, the Langmuir adsorption isotherm is
extended to multilayer adsorption. The first layer is as assumed in the Langmuir
adsorption isotherm. For subsequent layers, the rate of adsorption is assumed to be
proportional to the fi'action of the lower layers which are vacant. The summation over an

infinite number of adsorbed layers is defined by the following equation [11]:

P _ 1 +(C-1)
v(1>o —P) vmc vaPo

 

 

(1)

where C is a constant related to the heat of adsorption and liquefaction of the gas, Vm is
the monolayer capacity (cm3 STP), V is the volume of the adsorbatc at STP (cm3/mole
STP), P is the pressure of the adsorbatc, and P0 is the total pressure.

X-ray Difi'raction. X-ray diffraction (XRD) can be used to identify crystalline
phases with particle sizes greater than or equal to 3.0 nm. The particle size can be

calculated from line broadening measurements using the Scherrer equation [12]:

 

K}.

E =
BcosO

 

(2)

where d is the mean dimension of the particle, K is a constant (particle shape factor) taken
as 0.9 (for cubic particles), [3 is the full width-half maximum of the diffraction line, 7» is the
X-ray wavelength, and 9 is the angle between the atomic plane of both the incident and
reflected beams. In addition, semi-quantitative XRD analysis can be performed in order to
determine the weight percent of crystalline phases present in a catalyst series. In this
procedure, a physical mixture of the phases are ground in order to produce a homogenous
mixture. The areas of the peaks of interest are measured and used to generate a
calibration plot. Comparison of intensity ratios measured for the catalyst and the physical
mixtures can be used to determine the amount of crystalline phases present in the catalysts.
This method requires the particle sizes of the active phase to be similar in the catalyst
series and the physical mixture of components.

Nuclear Magnetic Resonance. Solid-state NMR is able to provide qualitative
information about chemically distinct sites, clarify physical states (coordination) [13], and
also provide quantitative information [14]. Since only the local environment of the atom is
probed, NMR is well suited for the structural analysis of systems that do not have long
range order. The type of interactions involved in solid-state NMR are direct dipole-dipole
coupling between nuclei, interaction of the nuclei with electrons in the environment
creating chemical shifts, and quadrupolar interactions with electric field gradients.

Infrared Spectroscopy. Infrared spectroscopy (IR) has been useful for obtaining
information about the structure of catalysts and species adsorbed on the catalyst surface.
IR has wide applicability in the study of dispersed metal or oxide catalysts. In addition,

spectra can be readily be acquired at high temperature and pressure. Thus one may study

 

28
catalysts under realistic reaction conditions and subsequently obtain information about the
nature of active sites in the catalysts.

Electron Paramagnetic Resonance. Electron paramagnetic resonance (EPR) has
been extensively used to study paramagnetic species of interest in catalysis such as
supported metal ions, surface defects, and adsorbed molecules and ions. The extent of
information obtainable from EPR data varies from a simple confirmation that an unknown
paramagnetic species is present to a detailed description of the bonding and orientation of
a surface complex. Factors such as spin-spin interactions, crystal field interaction and
relaxation time have a significant effect on the spectrum [15]. The high sensitivity of EPR

can be useful for the study of low concentrations of active phase species.

2.2.2. X-ray Photoelectron Spectroscopy (XPS)

A variety of surface analytical methods have been used to study heterogeneous
catalysts. X-ray photoelectron spectroscopy (XPS) is one of the most powerful surface
sensitive techniques for the study of electronic structure of filled levels of surface atoms

and adsorbates. The process occurring in XPS can be illustrated as follows [16]:

 

 

 

 

2 p /i i/ /i i/ /i i
2 s /i i/
h‘D
1 s 1
XPS
Figure 2.3. XPS process
A+hv——>A++e' (3)

E(kinetic) = E(photon) - E(binding) (4)

The kinetic energy of the photoemitted electron is essentially the difference between the
energy of the incident photon and the binding energy of the electron. Due to a variety of

processes in solids, equation 4 is more appropriately written as follows [17]:

EKE =hv—Eb+r-:,+E, (5)

 

30
where E1. is the intra-atomic relaxation energy and E, is the extra-atomic relaxation energy
associated with the solid environment. These relaxation shifts result when a core hole is
created by photoionization. When this occurs, other electrons relax to lower energy states
and partially screen the hole and thus make more energy available to the outgoing
photoelectron.

XPS can be used to obtain qualitative and quantitative information about surface
species. Even though XPS is dominated by atomic rather than solid-state effects, the local
electronic environment in the solid can influence the observed peak position and line
shapes through both initial and final state effects. Initial state effects refer to shifts in the
original binding energy due to changes in the electronic environment of the atom. Final
state effects are shifts associated with intra- and extra-atomic relaxations [17].

The signal intensity, 15, of XPS can be related to the concentration by the following

expression:

11 = 1071.1)(81 ”(81) (6)

where 19 is the X-ray flux, m is the concentration, oi is the photoelectron cross section
[18], D(ei) is the efficiency of the analyzer, and M6,) is the mean free path of the
photoelectron.

Inelastic mean free path or escape depth (3.) of low energy electrons in the solid is
a key parameter in describing the surface sensitivity of XPS. To obtain the mean free path

the following expression is used [19]:

8

ME) = [a(ln e + b)]

(7)

31
where M8) is the mean free path of the excited electron with energy 8 and a and b depend
on the electron concentration of the solid.

In order to obtain quantitative information about catalysts, a number of models
have been proposed to relate active phase/support peak intensity ratios to catalyst
structure. It has been shown by Defosse et al. [20] that one can calculate the theoretical
intensity ratio (Ip°/Is°) expected for a supported phase (p) atomically dispersed on a
carrier (8). Kerkhof and Moulijn [21] extended the Defosse model and derived
expressions based on a model catalyst that consists of sheets of support with cubic
particles of active phases deposited between the support layers (Figure 2.4).

For species dispersed as monolayers, the relationship is given by the following

equation:

 

1° -13
p (.2) D(ep)op13,(1+e 2) (3)

17’- - s b D(e,)o,2(1-e'52)

s monolayer

where ID is the intensity of the electrons from element p in the supported phase, Is is the
intensity of the electrons from element 3 in the carrier, GP and as are the photoelectron
cross sections of the respective elements, D(e)p,s are the detector efficiencies, ens are the
kinetic energies of the electrons, and [51.2 is the thickness (t) per mean escape depth of the

photoelectrons from the support (its) or from the promoter (hp). For crystallites, the

equation can be simplified to:

. \\\\\\\
2\\\\\\\\\\

S

I

 

33

[11] = 1.0") 1L. (9)
I 1° d
s cm] J 3:.

 

 

where (II’/196W is the experimentally determined intensity ratio, (Ip°/Is°) is the theoretical
monolayer intensity ratio calculated using equation 9, d is the length of the edge of the
cubic crystallites of the deposited phase, and XP is the mean escape depth of the
photoelectron in the promoter [22].

The ability of XPS to determine active phase particle size independent of chemical
state (metal or metal oxide) offers distinct advantages over more conventional methods of
catalyst characterization. For example, chemisorption may be used to determine the
average particle size of supported metals; however, the application of this technique to the
study of supported metal oxides has been limited to a few select systems. Using XRD, it
is possible to determine crystalline phases only if the particle size is 2 3.0 nm and the
phase is present in sufficient quantities. An important limitation of XRD is that it cannot
be used to study highly dispersed or amorphous phases. Electron microscopy may not be
useful for the characterization of mixed oxides due to the limited contrast between the
support phase and the canier.

An important capability of surface spectroscopy is the ability to observe changes in
the surface structure. Surface reactions provide useful information for understanding
changes in the surface structure that result during reactions (Figure 2.5). Species which
strongly interact with the support tend to undergo fewer chemical and structural changes
during reactive gas treatment in comparison to weakly interacting phases. By choosing

reactive gases that are involved in oxidation or ammoxidation reactions, it is possible to

34
study the reactivity of catalyst phases under conditions similar to those in actual reaction
conditions. In situ surface characterization of catalysts reacted with reducing molecules
such as H2, NH3, CO and hydrocarbons will be performed using a preparation chamber

attached directly to the surface science instrument (Figure 2.6).
2.2.3. Catalytic Activity

The catalytic activity experiments reported in this work were performed in a
differential reactor as shown in Figure 2.7. As shown in the figure, the gases are passed
through a zeolite bed to remove water. The flow of the gas is controlled by mass flow
controllers. The gas mixture can either be routed to the reactor or can bypass the reactor
for direct analysis of the gas stream prior to reaction. By correlating activity with the

catalyst structure it will be possible to identifyI active sites of the catalysts.

35

 

 

 

 

 

 

 

 

 

II' from
, /M(A)
_"=_.
M(A) reacts I.
to form NP 1.13
E M(A)
E "(3) CO I“ from M(B)
we) M(B) reacts -
to form M° all)
<— Binding energy C)

 

 

 

Figure 2.5. Protocol for study of heterogeneous catalysts by reactive gas treatment
combined with XPS. (From Anal. Chem, 58(12), 1184 A, 1986).

36

 

Figure 2.6. Reaction chamber.

 

 

 

 

 

 

 

 

h
”t

37

FURNACE

 

 

 

 

La
59

 

 

3-WAY
VALVE

 

 

 

 

 

 

 

 

“it t

 

 

 

 

 

 

arLJL

 

z-zaoura

 

Figure 2.7. Catalytic reactor.

w-Massnowcomaorm
m-GASOROMAM
air-mm
m-mromnonneracroa
rep-Wm

 

. fATr _.. tmtfl‘lfi'

 

38

References

1. F. Roozeboorn, D.D. Cordingley, and P]. Gellings, J. Catal., 68, 464, 1981.

2 LE. Wachs, R.Y. Saleh, S.S.' Chan, and CC. Chersich, App]. Catal., 15, 339,
1985.

3. PL. Villa, ”Catalyst Deactivation,” B.Delmon and GP. Fromont ed., Elsevier,
Amsterdam, p. 103, 1980.

4. J.C.W. Chien, J. Amer. Chem. Soc, 93, 4675, 1971.

5. H. Praliaud and M.V. Mathieu, J. Chim. Phys, 73, 689, 1976.

6. V.A. Khalif, E.L. Aptekar’, O.V. Krylov, and G. Ohlmann, Kinet. Catal. (USSR),
18,867,1977.

7. F. Roozeboorn, T. Fransen, P. Mars, and PI. Gellings, Z. Anorg. Allg. Chem, 25,
449, 1979.

8. E.T.C. Vogt, A.J. van Dillen, J .W. Geus, and F .J.J .G. Janssen, Catal. Today, 2,
569, 1988.

9. S. Shan and D.Hoenicke, Chem-lng.-Tech., 61, 321, 1989.

10. A. Baiker and A. Wokaun, Naturwiss, 76, 168, 1989.

11. ON. Satterfield, "Heterogenous Catalysis in Practice," New York, McGraw-Hill,
1980.

12. HP. Klug and LE. Alexander, "X-ray Diffraction Procedures for Polycrystalline
and Amorphous Materials," New York, Wiley-Interscience Publications, 1974.

13. C. Dybowski, Chemetech., March, 186, 1985.

14. BC. Gerstein, Anal. Chem, 55(7), 782A, 1983.

15. W.N. Delgass, G.L. Haller, R. Kellerrnan, and J .H. Lunsford, "Spectroscopy in
Heterogeneous Catalysis,” Academic Press, 1979.

16. D.M. Hercules and SH. Hercules, J. Chem. Ed, 61(6), 483, 1984.

17. DP. Woodruff and TA. Delchar,:Modem Techniques of Surface Science," F
Cambridge, Cambridge University Press, 1986.

18. J. H. Scofield, J. Electron Spectrosc. Relat. Phenom, 8, 129, 1979.

19. D. R. Penn, J. Electron Spectrosc. Relat. Phenom, 9, 29, 1976.

20. C. Defosse, P. Canesson, P. G. Rouxlet, and B. Delmon, J. Catal., 51, 269, 1979.

21. F. P. J. M. Kerkhof and J. A. Moulijn, J. Phys. Chem, 83, 1612, 1979.

22. M. Houalla, in "Preparation of Catalysts III", Elsevier, Amsterdam, 273, 1983.

 

Chapter 3

Vanadium Oxide/Aluminum Antimonate

3.1. Introduction

Considerable research effort has been focused on the development of selective
oxidation and ammoxidation catalysts for the activation of less expensive feedstocks such
as propane, propylene and methane. Several researchers have reported that vanadium
antimonate catalysts are promising systems for ammoxidation of propylene [1] and
propane [2] to acrylonitrile. Typical industrial catalysts are prepared by mixing a hydrosol
or gel of the carrier with a shiny or solution containing the active phase(s) and the
promoters, such as W and/or Mo. This procedure can produce very complicated catalyst
structures that make comprehensive characterization difficult. Thus, the effect of surface
structure on the properties of the catalyst and active site(s) are not well known. In order
to understand the influence of the catalyst structure on alkane or alkene activation, a
simple catalytic system must be examined.

Centi et al. [3] have investigated the structure and selective oxidation and
ammoxidation of propane activity for VSbOx mixed oxides supported on SiOz, A1203,
and Ti02 and promoted with W or Mo. Their work suggests that VSbOx phase plus
Sb205 is not the active phase for ammoxidation. Instead, they proposed that AleO4
rutile phase modified with V is the active phase for the formation of propylene and V or
other redox elements having M=O bonds (W or Mo) are needed to form acrolein or
acrylonitrile. Although VSbOx/A1203 is not as complex as the commercial catalysts,
further understanding of propane and propylene activation would be obtained by studying
the interaction of the pure phases believed to be responsible for selective oxidation and

ammoxidation.
39

. ‘:'l.'1"a‘

 

40

Recently, Volta et a1. [4,5] have examined the structure and propane oxidation
activity of vanadium oxide on AINbO4. AleO4 was chosen because of the structural
similarity between AleO4 and T102. These researchers showed that it is possible to
control the structure and reactivity of vanadia monolayers by modifying the local structure
of the oxide support. They also found that the Ale04 disorganized structure is present
and has good selectivity for the oxidative dehydrogenation of propane into propylene.

This work will correlate surface structure with propylene oxidation activity for V
oxide catalysts supported on AleO4. The focus will be on defining the nature of the V-
AleO4 interaction and influence of V dispersion and reactivity on the catalytic properties
of the mixed oxides. In order to determine structure and reactivity of the species in the
mixed oxide catalysts, X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS),
infrared spectroscopy (IR), solid-state nuclear magnetic resonance (NMR), and electron
paramagnetic resonance (EPR) will be used. Structural information will be correlated with
propylene oxidation activity measurements in order to establish the nature of propylene

activation site(s) in V/AleO4 mixed oxide catalysts.

3.2. Experimental

Catalyst Preparation. AleO4 was prepared by coprecipitation of a 1:1 atomic
ratio of Al(NO3)3-9H20 (J.T. Baker, A.C.S. grade) dissolved in deionized water and
SbC15 (Aldrich, 99%). SbC15 was added to a vigorously stirred solution of A1(NO3)3 over
a 30 minute period. NH40H (J .T. Baker, A.C.S. grade) was added until the solution pH
reached 6-7. The neutralized solution was allowed to stir for 1 hour. The precipitate
formed was filtered and washed with deionized water, dried in air at 120°C for 12 hours,
ground to produce a homogeneous mixture, and calcined in air at 750°C for 12 hours.
The resulting AleO4 support had a pore volume of 0.1 ml/g and a surface area of 155

m2/g. Prior to impregnation, the support was calcined at 500°C for 12 hours.

41

.V/AleO4 catalysts were prepared by pore volume impregnation of Ale04 using
vanadium triisopropoxide oxide (Alfa, 95-99%). The catalysts were dried under nitrogen
for 48 hours at 25°C, dried in air for. 12 hours at 120°C and calcined in air at 500°C for 12
hours. The vanadium content varied from O to 25 wt % V205 (V/Al atomic ratio of 0 to
0.78). Catalysts will be designated V# where # is the weight percent vanadium as V205.

Standard Materials. V02 (99.9%) and V203 (99%) were obtained fiom Aldrich.
V205 was prepared by calcining NH4VO3 (Johnson Matthey, 99.99%) in air at 450°C for
4 hours. Sb205 was prepared by the addition of SbC15 to deionized water under vigorous
stirring. The pH of the solution was adjusted to 6-7. The precipitate formed was filtered
and washed with deionized water, dried in air at 120°C for 12 hours and calcined at 500°C
for 12 hours. XRD patterns of all standard compounds matched the appropriate ASTM
powder diffraction file.

BET Surface Area. Surface area measurements were detemtined using a
QuantaChrome Quantasorb Jr. Sorption System. Approximately 0.10 grams of the
catalyst was outgased between 165°-170°C for 12 hours prior to absorption
measurements. Measurements were made using relative pressures of N2 to He of 0.05,
0.08, and 0.15 (surface area of N2 = 0.162 nm2) at 77 K. Data were processed using a
Macintosh computer.

X-ray Drflraction. X-ray powder diffraction patterns were obtained using a Rigaku
XRD diffractometer which employs Cu Kor radiation (1.541838 A). The x-ray was
operated at 45 kilovolts and 100 milliamps. The patterns were scanned between 10°-75°
(2 theta) at a scan rate of 0.1 °/min. with DS and SS = 0.5. Powdered samples were
mounted on glass slides by pressing the powder in an indentation on one side of the slide.
Particle sizes were determined from line broadening measurements using the Scherrer

equation [6]:

(1)

 

where d is the mean dimension of the particle, K is a constant (particle shape factor) taken
as 0.9 (for cubic particles), 6 is the hill width-half maximum of the V205 <001>
difi‘raction line, 3. is the X-ray wavelength, and 0 is the angle between the atomic plane of
both the incident and reflected beams. Semi-quantitative XRD has been performed using
physical mixtures of V205 and Ale04 (0.2 wt %, 0.5 wt %, and 1.0 wt % V205). The
areas of the V205 <001> and Ale04 <110> peaks were measured and used to generate a
calibration curve that can be used to determine the amount of crystalline V205 present in
the V/AleO4 catalysts. Peak areas were obtained using Googly software [7] and a linear
background was assumed over the peak base.

X-ray Photoelectron Spectroscopy. X-ray photoelectron spectroscopy (XPS) data
were obtained using a Perkin-Elmer spectrometer that is equipped with a Mg (1253.6
eV)/Al (1486.6 eV) dual anode and a 10-360 hemispherical analyzer with an omnifocus
small spot lens. Samples were mounted as powders on double-sided sticky tape or spray
coated on glass slides. Spray coated samples were prepared by air brushing a 10%
suspension of the catalyst (20% deionized water and 80% acetone) onto a glass slide
heated at 60°C. XPS binding energies of the catalyst were referenced to the Al 2p (74.5
eV) peak. The binding energies for standard compounds were referenced to the C Is
(284.6 eV). Peak areas and binding energies were obtained using Googly software. XPS
binding energies were measured with a precision of d: 0.2 eV, or better.

In situ XPS analyses of spray coated V/AleO4 were performed using a reactor
attached directly to the spectrometer. The samples were initially analyzed, then reacted

first with 100 cc/min mixture of 8.0% 02 (99.98%) and 92% He (99.9%) for 1 hour at

 

43

350°C and then with 100 cc/min 1.7% propylene (99.9%), 8.0% 02 and 90.3% He for 10
minutes at 200°C.

Quantitative )G’S. A number of models have been proposed to relate active
phase/support XPS peak intensity ratios to catalyst structure. It has been shown by
Defosse et a1. [8] that one can calculate the theoretical intensity ratio (If/15°) expected for
a supported phase (p) atomically dispersed on a carrier (5). Kerkhof and Moulijn [9]
extended the Defosse model and derived expressions based on model catalysts that consist
of sheets of support with cubic particles of active phase deposited between the support
layers. The photoelectron cross-sections (0’) and mean escape depths (k) of the
photoelectrons used in the calculations are taken from Scofield [10] and Penn [11],
respectively. For species dispersed as monolayers, the relationship is given by the

following equation:

 

1?, =(E) D(ep)ch,(1+e'l32) (2)

I? b D(es)052(1-e-52)

S
monolayer

where Ip° is the intensity of the electrons from element p in the supported phase, 15° is the
intensity of electrons from element 5 in the carrier, GP and as are photoelectron cross
sections of the respective elements, D(ep’s) are detector efficiencies, 8w are the kinetic
energies of the electrons, and 61,2 is the thickness (t) per the mean escape depth of the
photoelectron fi'om the support (3.5) or the promoter (hp). This model predicts a linear
increase in the intensity ratio of the supported phase (p) to the carrier (5) as the (p/s)
atomic ratio is increased.

Infrared Spectroscopy. Infiared spectra were obtained using a Mattson 3020

FTIR spectrometer with a DTGS detector. The catalyst was suspended in a 80% acetone

44

20% water mixture and sprayed on a 13x2 mm KBr disk heated at 60°C. Spectra were
obtained with 2 cm'1 resolution and both background and sample had 32 scans performed.
Data acquisition were performed using a FIRST software package. Data manipulation
were performed using Googly software. A linear background was subtracted to produce
Figure 3.6.

Nuclear Magnetic Resonance. Solid-state nuclear magnetic resonance
experiments were performed using a Varian VXRS 400 spectrometer. The powdered
catalysts were packed into a zirconium oxide roater. The samples were spun in a Bruker
MAS probe at 4 kHz. The frequency used for aluminum catalysts was 104.230 MHz.

Wide-line solid-state NMR experiments were also performed. The amplifier is
fi'om American Microwave Technology and the probe is a standard Varian wide-line
probe. The powder samples were placed in a quartz tube and then placed in the wide-line
probe. The frequency used for the vanadium catalysts was 105.152 MHz. A SUN
workstation was used to operate the spectrometer and process the data.

Electron Paramagnetic Resonance. Electron paramagnetic resonance data for
V/AleO4 catalysts were obtained using Bruker ER 300D spectrometer. A Hewlett
Packard 5245L electronic counter with a 5255A frequency converter plug in and a Bruker
ERO35M gaussmeter were used to measure microwave frequency and magnetic field,
respectively. The cavity used was a rectangular TEroz- The samples were run as powders
after reaction in 1.7% propylene, 8.0% 02 and 90.3% He.

Propylene Oxidation Activity Measurements. Catalytic activity experiments were
carried out in a differential type flow reactor at low conversion (<10%) with
Oz/propylene/He (8.0/l.7/90.3%) at a temperature of 250°C. The catalysts were
pretreated at 350°C for 3 hours with a 8.0% 02/92% He mixture. For both pretreatment
and reactions, the flow rate was 100 cc/minute. Approximately 0.2 grams were used for
each reaction. Reactions reached steady-state after 11 hours. The gas flow rates were

held constant with Brooks 5850 and Porter 201 mass flow controllers. The reactor

45

temperature was controlled with an Omega CN 1200 controller. Product gases were
analyzed with a Varian 3700 gas chromatograph equipped with a TCD (for C0, C02) and
FID (for hydrocarbon, acrolein, and acetaldehyde). The column used for permanent gas
separation was a 5 ft. 60/80 mesh Carboxen (Supelco) column. For hydrocarbon and
acrolein separation a 5 it 80/ 100 Porapak QS column was used. The chromatograph was
interfaced to a Hewlett-Packard 3394A integrator. The column temperature was
programmed from 55°C to 190°C with analysis time of 30 minutes to allow complete
separation of all components. Selectivities are reported for catalysts operated at similar

conversion.
3.3. Results

Ale04. The BET surface area of the Ale04 support is given in Table 3.1. The
XRD powder pattern of the support (Figure 3.1) matches results reported in the literature
[12]. The Ale04 particle size determined using x-ray diffraction line broadening
calculations (equation 1) is 2.7 nm. No other peaks due to antimony oxides or aluminum
oxides were observed. 2"Al MAS solid-state NMR of the aluminum antimonate support
has been performed and peaks at 1, 34, and -31 ppm in the NMR spectrum were observed
(Figure 3.2). The FTIR spectrum of the support was also obtained and no peaks were
observed between approximately 920 cm'1 and 1050 cm'1 (the region of interest of the
vanadium oxide). The XPS Sb 3d3,2 binding energy measured for the support was found
to be 540.5 eV. The XPS binding energy for Sb205 was 540.5 eV. The XPS Sb3d3,2/Al
2p intensity ratio indicates that Sb/Al atomic ratio in the surface of the Ale04 support is
1.0.

Vanadium/Aluminum Antimonate. Variation in the BET surface area of the
V/AleO4 catalysts as a function of vanadium content is shown in Table 3.1. The surface

area decreases from 140 to 108 m2/g as the WA] atomic ratio increases from 0.12 to 0.78

46

Table 3.1. BET surface area (mZ/g) of V205/N5b04 catalysts.

 

 

Atomic Ratio V/Al Weight % V205 Surface Area (ngLI
0 0 155
0.12 5 140
0.26 10 1.32
0.41 15 123
0.58 20 1 11
0.78 25 108

 

 

 

 

 

 

47

 

 

 

 

 

1 1 m4 1 I
<110>_
AleO4
<101>
<001>
a
b
c
l l 1 I
10.0 20.0 30.0 40.0

DEGREE (2 THETA)

Figure 3.1. XRD of V20,/AISb04 a) V25, b) V20, and c) V0 catalysts.

48

 

 

 

 

ITTTrf‘YVITYTY‘IWY'YIYVY rrfifrrrr

. r . h
300 200 100 0 -100 -200 -300

1’1”“

Figure 3.2. 27Al MAS NMR of a) V25 and b) V0 catalysts.

49

(5 wt.% to 25 wt.% V205). However, corrections made for active phase loading indicate
that the surface area of the support does not change.

For V/AleO4 catalysts with V/Al atomic ratios of < 0.78, XRD patterns showed
peaks characteristic only of the Ale04 carrier (Figure 3.1). The diffraction pattern
measured for the V25 catalysts showed peaks characteristic of Ale04 and crystalline
V205. For the V25 catalyst, XRD line broadening calculations (equation 1) indicated that
the particle size of the V205 crystallites was 29 nm (Table 3.2). Figure 3.3 shows the
variation of the V205 <001>/AISb04 <110> intensity ratios measured for 0.2, 0.5 and 1.0
wt.% V205/AleO4 physical mixtures. The value measured for the V25 catalyst with is
also given. The results indicate that the amount of V205 crystallite in the catalyst is less
than 0.1 wt.% V205.

XPS V 2p3,2 binding energies measured for the catalysts, run as powders, were
found to be 517.5 a: 0.2 eV. The XPS V 2pm binding energy measured for pure V205,
V02 and V203 were 517.5 eV, 517.3 eV and 516.5 eV, respectively. The binding
energies of the Sb 3d3,2 for the series of catalysts were found to be 540.4 :1: 0.2 eV. The
binding energies for both vanadium and antimony are independent of vanadium loading.

XPS V 2pm binding energies measured for the spray coated and propylene/OZ/He
reacted catalysts were 517.3 :h 0.2 eV. For 02/He reacted catalysts the V 2pm binding
energies were 517.0 i 0.3 eV. The Sb 3d3,2 for all the samples under each condition was
540.3 :1: 0.2 eV. Variation in the position of the vanadium peak as a function of reaction
condition measured for the V25 catalyst is shown in Figure 3.4.

Figure 3.5 shows the variation in the V 2p3,2/Al 2p XPS intensity ratio measured
for powder, spray coated, oxidized, and reacted catalysts as a fimction of WA] atomic
ratio. There is a linear increase in intensity ratios for the spray coated and reacted samples
up to a V/Al atomic ratio of 0.58. Between an V/Al atomic ratio of 0.58 and 0.78, there is
a sharp decrease in the WA] intensity ratio measured for the spray coated and reacted

samples. For the powdered samples, there is a linear increase in the WA] intensity ratio as

50

 

 

 

 

 

 

0.08 --
I
" Physical Mixtures ,
/
— "' " Best Fit /
/
0.06 " t 25 wt.% /
‘ V/AleO4 /
/
7p /
_o_ /
V /
v /
3 ) /
3 /
S. 0 04 ‘* /
A /
g /
n /
c; /
> / '
-‘-" /
/
/
/
0.02 -- /
/
/
/
/
/
/
A a
/
o ./ 1 1 1 t 1
0 0.2 0.4 0.6 0.8 1
Weight Percent V205

Figure 3.3. Semi-quantitative XRD for physical mixtures of V205 and Ale04.

51

Table 3.2. Particle size for V205/AleO4 calculated from XRD data.

 

 

 

 

 

XRD Particle size (nm)
V/Al Atomic Ratio XRD(nm)
0.12 J
0.26 _‘
0.41 —"
0.58 J
‘ 0.78 29

 

' - No V205 peaks observed. Thus, no V205 particle size can be calculated.

52

as the V/Al atomic ratio increases. For catalysts with V/Al atomic ratios < 0.58 and V25
catalyst the intensity ratios for the powder, spray coated, and reacted samples are similar,
within experimental error. The V20 V/Al intensity ratios measured for spray coated and
reacted catalysts were identical, within experimental error. However, the WA] intensity
ratio measured for the powder catalyst was significantly lower than those obtained for the
spray coated and reacted catalysts.

IR. Figure 3.6 shows the IR spectra measured between approximately 920 cm'1
and 1050 cm"1 for the catalysts. The spectrum measured for the V205 is shown for
comparison. The spectra contains a broad band centered at 985 cm"1 for all loadings. As
the V/Al atomic ratio increases from 0.26 to 0.41, another band appears at 995 cm‘l. A
band at 1022 cm"1 (which is due to crystalline V205 [13]) is observed in the spectrum
measured for the V25 catalyst.

Nil/IR. The 27Al MAS NMR spectra measured for the aluminum antimonate
support and the vanadium catalyst series show one main peak at approximately 1.0 ppm
and additional peaks at 34 ppm and -31 ppm (Figure 3.2). Figure 3.7 shows the wide-line
solid-state NMR spectra obtained for the V5 and V25 catalysts measured at a frequency
of 105.152 MHz. The vanadium peak was found to shift by 40 ppm from the 5 wt.% to
the 25 wt.% catalysts.

EPR EPR spectra were measured for the catalysts calcined at 500°C and after
exposure to reaction conditions. The calcined catalyst showed no signal in the EPR
spectrum. However after reaction, the catalyst showed a very broad signal due to V4+
with a g value of 1.98 (Figure 3.8). Hyperfine structure was observed in the spectrum
measured for the V5 catalyst. However because of the concentration of the vanadium,
detailed information could not be obtained. Dilution of the catalysts with support did not
improve the spectra obtained.

Propylene Oxidation. Table 3.3 shows the average percent conversion and

selectivity to acrolein, acetaldehyde, C0, and C02 measured for the V/Ale04 catalysts.

 

 

 

l l l 1
520.0 515.0 515.0 514.0
BINDING ENERGY (eV)

 

Figure 3.4. XPS spectra of V25 catalyst a) in propylene/OZ/I-Ie, b) 02/He, c) spray
coated, and d) powder.

54

   
  
  
    
  

 

6 .1.-
' Spray Coated
e
S '1'- . Ozme
A PropyleneJ02/He
Powder

 

 

 

    

 

   

 

CE}- 4 .1 e
3
555 z
a: . I
N
Z. 3 ‘

2 ..

l

0 1 1 1 1 1 1 1 a

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

V/Al Atomic Ratio

Figure 3.5. XPS intensity ratio of V 2pm/A1 2p as a function of WA] atomic ratio.

55

 

__ 1‘)

 

 

 

—
_

l 1
1040.0 1WD 960.0 920.0

 

FREQUENCY (cm 4 )

Figure 3.6. IR data of a) V205, b) V25, c) V20, d) V15, e) V10, and 1) V5 catalysts.

 

 

MT 1 Y Y I rY Y f1 1W
200 0 ~100

Figure 3.7. Vanadium Wide-line NMR of a) V25 and b) V5 catalysts.

-200

 

 

 

l I I 1
2000.0 3000.0 4000.0 5000.0

 

FIELD (GAUSS)

Figure 3.8. EPR V4+ spectra measured for reacted a) V25 and b) V5 catalysts. Spectra
were collected at room temperature.

58

 

 

 

Table 3.3 Conversion and selectivity of V205/AleO4 for propylene oxidation at
200°C and 100 cc/minute flow rate.
V/Al Percent Selectivity Selectivity C0 C02
Atomic Conversion to to

Ratio Acetaldehyde 3 Acrolein b

0 0.5 37.1 59.5 3.0 <1.0
0.12 1.4 9.6 90.4 <0.1 0.1
0.26 2.2 13.6 86.2 <0.1 <0.1
0.41 3.3 17.3 82.3 <0.2 0.1
0.58 3.3 19.0 80.5 <0.2 <0.1
0.78 4.1 21.3 78.0 <0.5 <0.1
0.12 3.1 7.0 89.6 2.1 1.3
0.78 3.6 21.4 78.1 <0.4 <0.1

 

 

 

 

 

 

 

1' - Error bars of d: 20%.
b - Error bars of i 10%.

 

59

The production of carbon oxides was the highest for the pure AleO4. As the amount of
vanadium oxide increases, the percent conversion increases. The addition of 5 wt.% (V5
catalyst) vanadium oxide increases the selectivity to acrolein and decreases the amount of
carbon oxides formed. For higher vanadium oxide loadings, the selectivity to acrolein
decreases. When the conversion of the extremes of the catalysts were similar, the
selectivity to acrolein for the V5 catalyst was 89.6% and for the V25 catalyst was 78.0%.
The amount of C0 produced by the V5 catalyst increased from < 0.2% to 2.1% as the

conversion increased.

3.4. Discussion

Aluminum Antimonate. BET, XRD, and NMR results indicate that base hydrolysis
of aluminum nitrate and antimony pentachloride followed by calcination in air at 750°C
results in a pure, high surface area (155 m2/g) AleO4 support. The XPS Sb 3d3,2
binding energy measured for AleO4 carrier is consistent with the formation of Sb5+
expected for this compound. In addition, the Sb/Al atomic ratio determined fi’om XPS
intensity ratios (1.0) suggests that no residual aluminum or antimony phases exist in the
surface of the carrier. Volta et al. [4,5] have reported that coprecipitation followed by
calcination at 750°C leads to the formation of crystalline AleO4.

Structure of Calcined Vanadium/A luminum Antimonate Catalysts. For V/AleO4
catalysts with V/Al atomic ratio S 0.58, the absence of vanadium oxide peaks in the XRD
pattern indicates that no crystalline vanadium oxide phases are formed in these catalysts.
However, semi-quantitative XRD analysis suggests that the V25 catalyst contains small
amounts of crystalline V205 (< 0.1 wt.% V205).

The XPS V 2pm binding energies measured for the powder and spray coated

V/Ale04 catalysts indicate that the vanadium is in the *5 oxidation state for all active

60

phase loadings. XPS Sb 3d3,2 binding energy data indicates that the antimony is in the +5
oxidation state for all vanadium loadings.

The IR spectra measured for the catalysts suggest that monomeric forms of
vanadium oxide are present in the catalysts because of the presence of a band centered at
985 cm"1 [13]. Polymeric forms of vanadium oxide are observed in catalyst with an V/Al
atomic ratio 2 0.26 due to the presence of the band centered at 995 cm'1 [13]. For the
V25, the observation of a peak at 1020 cm'1 (due to V205) [13] is consistent with XRD
results that show the presence of V205 crystallites.

Wide-line solid state NMR was used to determine the coordination of the
vanadium in the catalysts. Eckert and Wachs [14] used wide-line NMR to study the local
environments in two-dimensional vanadium(V) oxide surface layers on titania and alumina
supports. Two main surface vanadium oxide species were detected with different bonding
environments assigned to 4- and 6- coordinate V-O environments. Their results indicate
that a marked dependence of the surface vanadium oxide structure on the metal oxide
support material. For the V5 catalyst the vanadium is in a 4-coordinate environment. As
the vanadium loading is increased to 25 wt.% (V25), the vanadium becomes 6-coordinate.
This shift is as expected due to the change in the form of vanadium oxide from monomeric
at low loadings to polymeric forms at high loadings as observed from IR data.

MAS solid-state NMR on the V/Ale04 catalysts were identical to the spectrum
measured for the support. The signal between 0-2 ppm is assigned to aluminum in the
octahedral environment [15]. The peak at 34 ppm and -31 ppm has been attributed to
pentacoordinated atoms or to aluminum in a tetrahedrally distorted coordination [16].
2"Al MAS NMR spectra measured for both the support and catalysts show that there is no
transformation of structure of the antimonate after the vanadium is added and the sample
is calcined. Similar results were obtained by Volta et al. [17] for V0" on AleO4.

The method of preparation used was designed to produce a single monolayer of

V0x species on the Ale04 surface. In order to determine the theoretical monolayer

61

coverage, Roozeboom et al. [18] have defined 02.5 as the average area of supported
surface which V015 unit occupies. For V205, this value is approximately 0.105 nmz.
Thus, the monolayer capacity of the. AISb04 support is calculated to be 22.5 wt.% V205.
From the XRD pattern we see that crystallite formation is observed between 0.58 and 0.78
atomic ratio of V205. This is as expected based on the estimated monolayer capacity.

For V/AleO4 catalysts with V/Al atomic ratio 5 0.58, XPS and XRD data
indicate that the vanadium is highly dispersed. XRD results show that large particles of
V205 (~ 29 nm) are formed on catalysts with higher vanadium loadings. Figure 3.5 shows
that the V 2pm/Al 2p intensity ratio increases linearly with vanadium loading up to a
V/Al atomic ratio of 0.58. Between V/Al atomic ratio of 0.58 and 0.78 the intensity ratio
decreases (observed for the spray coated and reacted samples). This suggests that the
dispersion of the vanadium oxide decreases drastically at this level. In addition, this is the
same loading for the formation of crystallites as observed in the XRD data and that
estimated fi'om the monolayer capacity calculation. However, the powder sample of WA]
of 0.58 does not follow the trend as seen for the spray coated and reacted samples. At
present we can not determine the cause of the difference in WA] intensity ratios measured
for the spray coated, reacted, and powder V20 catalysts. Samples with WA] 2 0.58 were
prepared twice and the same trend was observed.

IR results also suggest that the vanadium is highly dispersed due to the presence of
bands characteristic of monomeric and polymeric vanadium oxide species. The band
centered at 985 cm'1 is assigned to vanadate dispersed on the support as a monolayer.
Distortion of the structure causes a shift of the V=0 stretching band fi'om 1020 to 985
cm'l [13]. The band centered at 995 cm'l is attributed to polymeric forms of vanadium
on the support. The band at 1022 cm'1 is due to crystallite forms of V205. The absence
of a band characteristic of vanadate (950 cm") suggests that no aluminum or antimony
vanadates are present [17]. MAS NMR obtained for the support and the catalysts
supports this finding. Nakagawa et a1. [13] studied monolayer species of V205/Ti02 and

62

found a shift fiom 1020 (crystallite V205) to 980 cm'1 for the V=0 stretching frequency.
Their results lead to the conclusion that at low coverage V0x is amorphous and at high
coverage both amorphous and crystalline V205 are observed. Similar results were
obtained for V205 supported on T102, A1203, Zr02, and S102 deposited either fiom
aqueous metavanadate or gaseous VOC13 precursors [19-21].

Bond et a1. [22] studied V205/T102 catalysts prepared by wet impregnation and
grafting methods. They observed that as V205 concentration was increased, the XPS V
2p3,2/T i 2pm intensity ratio increased to approximately 7 wt.%, but after two monolayers
were deposited the ratio remained approximately constant. These results suggest that
once the V205 amount exceeds that necessary for formation of the first monolayer, the
V205 is not dispersed uniformly, but crystallites of V205 cover a fraction of the
monolayer surface. Based on XPS, XRD, IR and NMR vanadium oxide is present as
monomeric species as the loading increase, polymeric forms of vanadium oxide are present
and at 25 wt.% crystalline form of V205 are present.

Structure of Reacted Vanadium/A luminum Antimonate Catalysts. The V/Al XPS
intensity ratios for the reacted samples suggest that the dispersion of the vanadium does
not change under reaction conditions. In order to determine if XPS can distinguish
between V5+ and V4+ the binding energies of standard compounds have been obtained.
Unfortunately, the binding energies of standard compounds have shown that it is difficult
to distinguish between V5+ and V“. However, as seen from Figure 3.4, slight changes in
the position of the vanadium peak for the powder, spray coated, propylene/Oz/He, and
02/He occurs, but the shifts observed are not as large as expected for a change in the
vanadium oxidation state. The V 2pm binding energies reported in the literature for
V205 vary fi'om 516.4 to 517.4 eV [23,24]. For V41“, represented by V204, the binding
energies range from 515.4 to 515.7 eV [25,26]. Based on literature values, it appears that
the vanadium in the catalyst is V5+. Unfortunately, XPS cannot unequivocally determine

if a change in vanadium oxidation state occurs due to reaction.

63

EPR spectra clearly show that some of the vanadium is reduced from V5+ to V4+
under reaction conditions. Because of the high concentration of vanadium present in this
system, it is not possible to obtain any useful information from the hyperfine splitting of
the catalyst series. With conventional EPR, noise caused by the cavity was seen and at
lower concentrations this signal was greater than that of the signal due to the vanadium.
Thus, at present, we cannot obtain coordination of the vanadium as a firnction of loading
and we cannot obtain quantitative information about the amount of reacted vanadium.

Influence of Catalyst Structure on Activity and Selectivity. The V5 catalyst shows
the greatest selectivity towards acrolein production. With the conversion extremes
similar, the selectivity toward acrolein for the monomeric form is still greater than that
with crystalline forms present. In addition, the pure aluminum antimonate support
produced higher levels of carbon oxides than the promoted support. Activity
measurements of the vanadium/aluminum antimonate catalyst suggest that monomeric
forms of vanadium oxide are more selective in the production of acrolein than polymeric
and crystalline forms of vanadium oxide. However, all forms of supported vanadium
oxide catalysts showed greater selectivity toward acrolein than the aluminum antimonate
support.

By the addition of vanadium to the surface of the support, a redox element having
a V=O bond is present. The presence of a terminal V=O site, the selectivity during the
oxidation process has been shown to be effected [27]. It is generally believed that too
weak M-O bonds results in non-discriminative C-O bond formation which leads to
combustion. Ifthe M-O bond is too strong, lattice oxygen is unavailable to participate in
the reaction [28]. In addition to M=O bonds and bond strength, geometric effects also
play a role in that the active site must contain "the right amount" of oxygen [29]. In this
system, the presence of the vanadium increases the selectivity to acrolein at all active
phase loadings. However, as crystalline phases of V205 become present, an increase in

CO" and the by-product, acetaldehyde is observed. This may be due to the reactive nature

64

of crystalline phases. In this system, as with MoO3-based mixed oxide catalysts, it is
possible that the rate of allylic oxidation is controlled by the rate of allylic-intermediate
formation [30].

Straguzzi et al. [31] studied the conversion of l-butene to butadiene on aluminum
antimonate catalyst and found that the selectivity to butadiene was 75% at conversions
less than 25% and showed activity of 1.0 mmol/mzh. Straguzzi suggests that due to the
dimculty in reducing Al“, the activity of the aluminum antimonate can be attributed to Sb
cations) However, the aluminum antimonate they prepared had a surface area of 51 m2/g
and was higher in crystallinity. The aluminum antimonate that was produced in this work
demonstrated a low selectivity to acrolein, 59.5%.

The structure of the antimonate site for selective oxidation in uranium antimonate
is composed of bridging oxygen species which connect two Sb5+ and two Sb3+ atoms. In
this, the Sb3+-0 bridge serves as the H abstracting species and gives radical character to
the oxygen bond at the site. This facilitates the H abstraction. The Sb5+-O moieties are
used as olefin chemisorption sites and sites for the insertion of oxygen [32]. In order to

determine which of these processes are occurring, firrther investigation is needed.

3.5. Conclusions

The combined use of several techniques to investigate the structure and dispersion
of V/AleO4 catalysts leads to the following conclusions:

1. XPS, XRD, IR, and NMR data indicates that Ale04 is a pure phase with no
formation of aluminum or antimony oxide species present.

2. XPS, XRD, and IR data indicate that V is highly dispersed over the Ale04
carrier. XRD and XPS results suggest the trend of crystallite formation is consistent with

estimated monolayer capacity calculations.

65

3. Based on XPS, data both Sb and V species are in the *5 oxidation state. XPS
binding energies of V and Sb do not change as a firnction of loading or reaction
conditions. .

4. IR and NMR data suggest that the vanadium oxide does not form a vanadate-
type structure with the aluminum antimonate support.

5. EPR data indicates that vanadium is reduced fiom V5+ to V“. during
propylene oxidation.

6. Activity data suggests that low loadings of vanadium oxide (V/Al S 0.26)
increase selectivity to acrolein. Pure aluminum antimonate support shows lower

selectivity to acrolein than vanadium oxide supported catalysts at all loadings.

References

fl

F. J. Berry and M. E. Brett, J. Catal., 88, 232, (1984).

2. G. Centi, D. Pesheva, and F. Trifiro, App]. Catal., 33, 343, (1987).

G. Centi, R. K. Grasselli, E. Patane, and F. Trifiro, in "New Developments in

Selective Oxidation”, Stud Surf Sci. Catal., 55, (1990).

4. PG. Pries DE Oliveira, F. Lefebvre, M. Primet, J.G. Eon, and J .C. Volta, J. Cata],
130, 293, (1991).

5. PG. Pries DE Oliveira, J .G. Eon, and 1C. Volta, J. Catal, 137, 257, (1992).

6. H. P. Klug and L. E. Alexander,”X-ray Diffraction Procedures for Polycrystalline
and Amorphous Materials", New York, N.Y., Wily-Interscience Publication, 1974.

7. A. Proctor, University of Pittsburgh.

8. C. Defosse, P. Canesson, P. G. Rouxlet, and B. Delmon, J. Catal., 51, 269, 1979.

9. F. P. J. M. Kerkhof and J. A. Moulijn, J. Phys. Chem, 83, 1612, 1979.

10. J. H. Scofield, J. Electron Spectrosc. Relat. Phenom, 8, 129, 1979.

11. D. R. Penn, J. Electron Spectrosc. Relat. Phenom, 9, 29, 1976.

12. D.V. Tarasova, V.A. Dzis'ko, I.P. Olen'kova, L.T. Tsikza, L.G. Karakchiev, and
V.V. Molodtsova, Kin. i Kata, 14(2), 481, 1973.

13. Y. Nakagawa, T. Ono, H. Miyata, and Y. Kubokawa, J. Chem. Soc, Faraday
Trans. 1, 79, 2929, 1983.

14. H. Eckert and LE. Wachs, J. Phys. Chem, 93, 6796, 1989.

15. G. Engelhardt and D. Michel, High-Resolution Solid-State WR of Silicates and
Zeolites, Wiley, New York, 1987.

16. P. Massiani, B. Chauvin, F. Fajula, F. Figueras, and C. Gueguen, Appl. Catal., 42,

105, 1988.

l”

17.

18.

19.
20.

21.
22. '
23.
24.
25.
26.

27.
28.

29.
30.

31.

32.

66

PG. Pries de Oliveria, F.Lefebvre, M. Primet, J .G. Eon, and J .C. Volta, J. Chem.
Soc, Chem. Commun, 1480, 1990.

F. Roozeboom, T. Fransen, P. Mars, and PI. Gellings, Anorg. Allg. Chem, 449,
25, 1979.

AA. Davydov, Zh. Fiz. Khim, 65,1803, 1991.

H. Miyata, K. Fujii, T.Ono, Y.Kubokawa, T.Ohno, and F. Hatayama, J. Chem.
Soc, Faraday Trans, 83(1), 675, 1987.

Y. Kera and K. I-Iirata, J. Phys. Chem, 37(11), 3973, 1969.

G. Bond and P. K6nig, J. Catal., 77, 309, 1982.

QC. Bond, J.P. Zurita, and S. Flamerz, Appl. Catal., 27, 353, 1986.

B. Blaauw, F. Leenhouts, F. van der Woude, and GA. Sawatsky, J. Phys. C., 8,
367, 1975.

J. Mendialdua, Y. Barbaux, L. Gengembre, J .-P. Bonnelle, B. Grzybowska, and M.
Gasior, Bull. Pol. Acad. Sci. Chem, 35, 213, 1987.

B. Blaauw, F. Leenhouts, F. van der Woude, and GA. Sawatsky, J. Phys. C., 8,
367, 1975.

K. Mori, Y. Yoda, N. Tamazoe ,and T. Seiyona, J. Phys. Chem, 84, 2564, 1984.
W.M.H. Sachtler, G.J.H. Dorgels, V. Farenfort, and R.J.H. Voorhoeve, Proc. 4th
Intern. Cong. Catal., 1, 355, 1970.

RK. Grasselli, Appl. Catal., 15, 127, 1985.

H.H. Voge and CR. Adams, ”Advances in Catalysis", Vol. 17, p. 151, Academic
Press, New York, 1967.

G.I. Straguazzi, K.B. Bischoff, TA. Koch, and G.C.A. Schuit, J. Catal., 104,
47,1987.

J.D. Burrington, CT. Kartisek, and R.K. Grasselli, J. Catal., 87, 363, 1984.

 

Chapter 4
Supported Antimony Oxide Catalysts

4.1. Introduction

Antinomy oxide is often employed to promote the activity of catalysts used for
selective oxidation [1-5] and for related reactions such as oxidative coupling [6]. Among
the more common reactions are the oxidation and ammoxidation of propylene to acrolein
and acrylonitrile, the conversion of isobutene to methacrolein, and the condensation of
isobutene and formaldehyde to isopropene.

Grasselli and co-workers [7] examined the relationship between structure and
catalytic activity of uranium-antimonates for the synthesis of acroylnitrile fi'om propylene,
ammonia, and air. They found that USb05 and USb3010 phases are catalytically active,
but only USb3010 is selective. The selectivity has been attributed to the orthorhombic
structure that has no adjacent U-O-U structures. In USb05, U-O-U-O-U structures
present in the catalyst are centers for waste formation.

In addition, work has been performed on iron antimonate catalysts which are used
for the conversion of butene into butadiene [8-14]. Numerous hypotheses have been
proposed to explain the reactivity of Fe-Sb-O systems based on surface geometry,
electronic properties, and bulk structure [10,11,13]. Considerable work has also been
performed on Sb oxides supported on rutile structures such as Sn02 [15-18] and Ti02
[19] for the selective oxidation of propylene to acrolein. It appears that in these systems,
the catalytic performance for the oxidation of olefins is related to an enhancement of
antimony at the surface of the rutile support. This enhancement is achieved by high

temperature calcination which results in migration of the antimony through the bulk.

67

68

Lo et al. [20] reported the oxidative coupling of methane using a potassium-
modified silica supported antimony oxide catalyst. This study found that of the Sb6013,
KSbO3, and a-Sb204 phases most predominantly formed, only a-Sb204 was selective in
methane oxidative coupling reactions and gave high selectivity to C2.

Antimony oxide has also been used as a promoter for catalysts used in CO
oxidation [21]. Ali-Zade et al. [22] found that the activity of Cu, Co, Cr, and Mn oxides
supported on y-alumina is improved when Sb oxide is added. Antimony oxides have also
been used in conjunction with oxides of Be, Mg, Zn, A1, Si and rare earth metals [23] and
Sn-Mn-Pb oxides [24] for treatment of exhaust gases. Straguzzi et al. [8,10] examined
C0 oxidation and selective oxidation of 1-butene to butadiene on iron, aluminum, cobalt,
chromium, and rhodium antimonates. They found that these reactions have the same trend
in activity, which suggests that the reactions occur on the same site and indicates that the
antimony serves two functions: 1) to limit the size of the active oxygen ensemble and 2) to
adsorb 02.

In recent years, there has been an increased interest in preparation methods to
produce highly dispersed metal oxides on supports such as Si02, A1203, and TiOz
[25,26]. In general, these procedures are directed at producing monolayers with high
thermal stability of the metal oxide coating. However, it has been found that the sorption
[27,28] and catalytic properties [29-31] of the monolayer metal oxides are often different
from those observed for the unsupported oxide phase. For example, Denofre et al. [28]
found that Nb205/Si02 irreversibly adsorbs ascorbic acid forming a surface complex.
Shirai et al. [29] reported that the activity obe205/8102 was 20 times greater than niobic
acid bulk catalysts for esterification from ethanol and acetic acid.

Recently, Benvenutti et al. [32] have reported the thermal stability and the acidic
properties of Sb205 grafted on silica. Their results suggest that when antimony atoms are
distant thermal sintering occurs formation of crystalline Sb205 is inhibited. The highly

dispersed material contains Lewis acid sites due to unsaturated coordination of the Sb5+

69

ion. The sites disappear upon thermal treatment at 500°C due to the extensive reticulation
of Sb5’+ with the silica surface.

In this chapter, the surface. structure of silica and alumina supported antimony
oxide catalysts we will correlated with their activity for C0 oxidation. The focus will be
on defining the nature of the Sb-support interaction and the influence of Sb dispersion and
reactivity on the catalytic properties of the mixed oxides. In order to determine structure
and reactivity of the species in the mixed oxide catalysts, X-ray diffraction (XRD) and X-
ray photoelectron spectroscopy (XPS) will be used. The results obtained will be correlated
with CO oxidation activity measurements in order to establish the nature of CO oxidation

site(s) in the supported Sb oxide catalysts.

4.2. Experimental

Cata]yst Preparation. The supported antimony oxide catalysts were prepared by
pore volume impregnation of y-A1203 (Cyanamid; surface area: 204 mz/g; pore volume:
0.6 ml/g) and Si02 (Davison; surface area: 310 mZ/g; pore volume: 1.15 ml/g) which had
been ground to 230 mesh. The supports were calcined in air at 500°C for 12 hours. The
Si02 was allowed to sit in a desiccator for 12 days and A1203 for 7 days prior to
impregnation. The catalysts were prepared by incipient wetness using a solution of
antimony (III) n-butoxide (Strem, 99%) in dry n-hexane (Aldrich, 99.9%). The
impregnated catalysts were dried at room temperature in a nitrogen atmosphere for 48
hours, dried in air at 120°C for 24 hours and then calcined in air at 500°C for 12 hours.
Nominal loadings of the active element range from 0-40 wt.% as Sb205 (Sb/Al atomic
ratio of 0 to 0.21 and Sb/Si atomic ratio of 0 to 0.25). Catalysts will be designated as
Sb#A1 for the Sb/A1203 series and Sb#Si for the Sb/8102 series where # is the weight

percent of antimony as Sb205.

70

Standard Materials. Sb205 was prepared by the addition of SbC15 to deionized
water under vigorous stirring. The pH of the solution was then adjusted to a pH of 6-7.
The precipitate formed was filtered and washed with deionized water, dried in air at 120°C
for 12 hours and calcined at 500°C for 12 hours. Ale04 was prepared by coprecipitation
of a 1:1 atomic ratio of AI(N03)3-9H20 (J .T. Baker, A.C.S. grade) dissolved in deionized
water and SbCl5 (Aldrich, 99%). SbC15 was added to a vigorously stirred solution of
A1(NO3)3 over a 30 minute period. NH4OH (J .T. Baker, A.C.S. grade) was added until
the solution pH reached 6-7. The neutralized solution was allowed to stir for 1 hour. The
precipitate formed was filtered and washed with deionized water; dried in air at 120°C for
12 hours, ground to produce a homogeneous mixture, and calcined at 750°C for 12 hours.
XRD patterns of the standard compounds matched the appropriate ASTM powder
dimaction file.

BET Surface area. Surface area measurements were determined using a
QuantaChrome Quantasorb Jr. Sorption System. Approximately 0.1 grams of the catalyst
was outgased between 160°-165°C for 12 hours prior to adsorption measurements. The
measurements were made using relative pressures of N2 to He of 0.05, 0.08, and 0.15
(surface area N2: 0.162 nmz) at 77 K. Data were processed using a Macintosh computer.

X-ray Diflraction. XRD patterns were obtained using a Rigaku XRD
dim'actometer which employs Cu Ka radiation (1.541838 A). The x-ray was Operated at
45 kilovolts and 100 milliamps. The patterns were scanned between 10°-75° (2 theta) for
A1203 catalysts and 10°-45° (2 theta) Sb/Si02 catalysts. The scan rate was 0.1°/min. with
DS and S5 = 0.5. Powdered samples were mounted on glass slides by pressing the
powder in an indentation on one side of the slide. Semi-quantitative XRD has been
performed by using physical mixtures of Sb205 and A1203 with 0.05 wt %, 0.1 wt %,
0.25 wt %, and 0.5 wt % Sb205. For Sb/Si02, physical mixtures of Sb205 and Si02 with
0.25 wt.%, 0.45 wt.%, and 0.7 wt.% Sb205 were prepared. The samples were ground in

order to produce a homogenous mixture. The areas of the Sb205 <400> and the A1203

71

<400> or Si02 <101> peaks were measured and used to generate a calibration curve that
can be used to determine the amount of crystalline Sb205 present in the Sb/A1203 or
SblSiOz catalysts. Peak areas were obtained using Googly software [33] and a linear
background was assumed over the peak base.

X-ray Photoelectron Spectroscopy. X-ray photoelectron spectra (XPS) were
obtained using a Perkin-Elmer spectrometer equipped with a Mg (1253.6 eV)/A1 (1486.6
eV) dual anode and a 10-360 hemispherical analyzer with an omnifocus small spot lens.
Data were collected using a PC 137 interface board and a Zeos 386SX computer.
Samples were mounted as powders on double-sided sticky tape or spray coated on glass
slides or stainless steel plates. Spray coated samples were prepared by air brushing a 10%
suspension of the catalyst (20% deionized water and 80% acetone) onto a glass slide or a
stainless steel plate heated at 60°C. XPS binding energies of the Sb/A1203 catalysts were
referenced to the A1 2p (74.5 eV) and SblSiOz catalysts to the Si 2p (103.5 eV) peak.
The binding energies for standard compounds were referenced to the C 1s (284.6 eV)
peak. XPS binding energies were measured with a precision of i 0.2 eV, or better. The
data were processed using Googly software.

In situ XPS experiments were performed on catalysts treated in a reaction chamber
attached to the spectrometer and transferred directly to the analysis chamber without
exposure to air. The antimony oxide catalysts were reacted with 100 cc/minute of 5% 02
and 95% He for 1 hour at 350°C, then reacted in 20 cc/minute of H2 at 450°C for 10
minutes and then for 10 minutes at 450°C with 20 cc/minute of 02.

Quantitative ”S. A number of models have been proposed to relate active
phase/support XPS peak intensity ratios to catalyst structure. It has been shown by
Defosse et al. [34] that one can calculate the theoretical intensity ratio (If/15°) expected
for a supported phase (p) atomically dispersed on a carrier (5). Kerkhof and Moulijn [35]
extended the Defosse model and derived expressions based on model catalysts that consist

of sheets of support with cubic particles of active phase deposited between the support

 

72

layers. . The photoelectron cross-sections (o) and mean escape depths (A) of the
photoelectrons used in the calculations are taken from Scofield [36] and Penn [37],
respectively. For species dispersed as monolayers, the relationship is given by the

following equation:

 

1° -13
p =(p) D(ep)op01(l+e 2) (2)

I; b D(es)052(1-e'32)

S

.;
monolayer
where 19° is the intensity of the electrons from element p in the supported phase, 15° is the
intensity of electrons from element s in the carrier, GP and as are photoelectron cross
sections of the respective elements, D(ep,s) are detector efficiencies, 6p,s are the kinetic
energies of the electrons, and [31,2 is the thickness (t) per the mean escape depth of the
photoelectron from the support (15) or the promoter (21,). This model predicts a linear
increase in the supported phase (p) to the carrier (5) peak intensity ratio as the (p/s) atomic
ratio is increased.

CO Oxidation. C0 oxidation experiments were carried out in a differential type
flow reactor at low conversions (<10%). Steady-state was reached between 1.5 and 2
hours for each catalyst series. The gas flow rates were held constant with Brooks 5850
and Porter 201 mass flow controllers. The reactor temperature was monitored with an
Omega CN 1200 controller. The gas products for Sb/A1203 and 8102 were analyzed with
a Varian 920 gas chromatograph equipped with a TCD. The chromatograph was
interfaced to a Hewlett-Packard 3394A integrator. The column used for permanent gas

separation was a 5 foot 60/80 mesh Carbosieve column. The catalysts were pretreated

with 5% 02 (99.98%)/95% He (99.9%) at 350°C for 1 hour prior to reaction. The

73

reactorwas Operated at 450°C with a 15 cc/minute flow of 4.8% C0/9.8% 02/85.4% He
(mixture from AGA with purity > 99.9%).

4.3. Results

BET. Variation in the BET surface area of Sb/A1203 and Sb/SiOz catalysts as a
function of antimony oxide content are shown in Table 4.1. For both series, the surface
area decreases as the loading of the antimony increases. However, corrections made for
the active phase loading indicate that the surface area of the support does not change.

X-ray Diffraction (RD). XRD patterns measured for the Sb/A1203 catalysts
showed lines characteristic of the alumina carrier (Figure 4.1). For Sb205 loadings S 15
wt.% only peaks of the alumina carrier were seen. For loadings 2 20 wt.% a new feature
is observed between 20-40° (2 theta). The feature is centered at approximately 29° (2
theta) for all loadings. As the Sb205 loading increases, the intensity of this feature
increases. At loadings of 35 and 40 wt.% Sb205, Sb205 <400> and Sb205 <111> peaks
are observed. Figure 4.2 shows the variation of the Sb205 <400>/A1203 <400> intensity
ratios measured for Sb205/A1203 physical mixtures. The value measured for the Sb40A1
catalyst is also given. The results indicate that the amount of Sb205 crystallite in the
catalyst is less than 0.2 wt.% Sb205.

XRD patterns measured for the Sb/Si02 catalysts show the broad peak centered at
22° (2 theta) characteristic of the silica carrier (Figure 4.3). For Sb205 loadings less than
10 wt.% only the peak from the silica support was observed. For Sb205 loadings > 10
wt.% a new feature between 20° and 40° (2 theta) is observed centered at approximately
29° (2 theta). As the loading of Sb205 increases, the intensity of this feature increases. At
loadings of 25, 35, and 40 wt.% Sb205, Sb205 <400> and Sb205 <111> peaks are
observed. Figure 4.4 shows the variation of the Sb205 <400>/S103 <101> intensity ratios

measured for Sb205/Si02 physical mixtures. The results indicate that the amount of

 

Table 4.1. BET surface area (mZ/g) of SbZOS/A1203 and SiOz.

74

 

 

 

 

 

Weight % Sb205 A1203 Si02
0 203.6 307.8
5 195.2 285.5
10 179.5 261.1
15 172.5 241.3
20 167.5 228.3
25 162.7 210.1
30 149.1 198.9
35 135.8 175.2
40 129.1 163.4

 

 

75

Sb205 crystallite in each catalyst is less than 0.5 wt.% Sb205.

The presence of XRD peaks due to Sb205 observed for high Sb loadings can be
attn'buted to errors in catalyst preparation. During the preparation of Sb rich catalysts
(Sb205 > 20 wt.%) it is likely that very small amounts of the antimony alkoxide migrates
out of the catalyst bed onto the walls of the crucible used to prepare the catalysts. During
calcination the alkoxide is converted to Sb205 which subsequently falls into the catalyst
bed. Unfortunately, the crystallites are not large enough to be separated by sieving. Thus,
we believe that much of the crystalline Sb205 observed in the catalyst is due to this
preparation error. Samples were prepared three times in an effort to eliminate this
problem. Results are presented for the samples that gave the lowest Sb205 <400> XRD
peak intensity.

The XPS Sb 3d3,2 binding energies measured for the powder, spray coated and
reacted Sb/A1203 and Sb/SiOz catalysts were 540.6 i 0.2 eV. The XPS Sb 3d3,2 binding
energy measured for Sb205 was 540.5 eV. The XPS Sb 3d3/2 binding energy measured
for Ale04 was 540.5 eV.

Figure 4.5 shows the variation of the Sb 3d3,2/Al 2p intensity ratio as a firnction of
the Sb/Al atomic ratio measured for the Sb/A1203 catalysts. For Sb205 loadings < 30
wt.% the intensity ratios increase linearly with Sb loading for powder, spray coated, and
reacted samples. For loadings 2 35 wt.% the increase in Sb/Al intensity ratios with Sb
loading is less pronounced. Variation of the Sb 3d3,2/Si 2p XPS intensity ratio as a
function of the Sb/Si atomic ratio is shown for the Sb/S102 catalysts in Figure 4.6. There
is a linear increase in Sb/Si intensity ratio as a fiinction of Sb loading for powder, spray
coated, and reacted samples.

CO Oxidation Activity. Table 4.2 shows the average percent conversion for
Sb/A1203 and Sb/S102 catalyst series. For the Sb/A1203 catalysts, the percent CO
conversion increases with Sb loading up to the SblSAl catalyst. The activity decreases as

the Sb loading increases from 15 to 40 wt.% Sb205. For Sb loadings 2 25 wt.% Sb205,

76

 

 

 

 

 

1 1 1 1 1
10.0 20.0 33.0 40.0 50.0
DEGREE (2 THETA)
Figure 4.1. XRD of Sb205/A1203 a) 51140111, 11) 51135111, c) 31130111, d) 31125.11, e)

Sb20Al, f) SblSAl, g) SblOAl, h) SbSAl, and i) SbOAl catalysts.

«sz0, <400>)/1(A1203 <400>)

 

77

 

 

 

 

 

0.03 T
I
' Physical Mixtures / /
"' ‘ - Best Fit /
/
0 40 wt.% //
Sb/A1203 /
/
/
0.02 ‘1- /
/
/
/
/
/
/
/
/
/
/
/
/
0.01 "i' /
d
/
/
/
/
/ I
/I
/
/
/
0 9’ 1 1 1 1 a
0 0.1 0.2 0.3 0.4 0.5
Weight Percent Sb205

Figure 4.2. Semi-quantitative XRD physical mixtures of Sb205 and A1203.

78

 

59205

 

 

 

—
—

l l
1 0.0 20.0 30.0 40.0

 

DEGREE (2 THETA)

Figure 4.3. XRD of Sb205/Si02 a) SMOSi, b) Sb3SSi, c) SbBOSi, d) SbZSSi, e) SbZOSi,
1) SblSSi, g) SblOSi, h) SbSSi, and i) SbOSi catalysts.

 

79

 

 

 

 

 

 

0.8

0.006 -
' Physical Mixtures .
/
— - - Best Fit / .
/
l 40 wt.% Sir/8102 /
' /

° 35 wt.% Sb/Sioz /
A 0.004 -- //
/\
§ / /
V r
5'
e; ) /
t: /
"" /
§ /
'1 /
2- /
a /
"' -_ /

0.002 /
/
/
/
/
/
/ a
/
/
a
/
/
0 1! 1 1 1 1
0 0.2 0.4 0.6
Weight Percent 511205

Figure 4.4. Semi-quantitative XRD physical mixtures of Sb205 and S102.

80

 

      
     

 

 

 

 

 

6 T-
5 a
' powder .
I
a
0 Hydrogen . z
A Spray Coated A
4 we
' Oxygen a
:5? ram/957.1%
3
57:1 3 -
1‘3"
M
.0
9’3
2 d
l _1
0 i i l i 41
0 0.05 0.1 0.15 0.2 0.25
Sb/Al Atomic Ratio

Figure 4.5. XPS intensity ratio of Sb 3dm/A1 2p as a function of Sb/Al atomic ratio.

81

 

 

 

 

 

 

 

4 q.-
' Powder
3 ° Hydrogen
1‘ Spray Coated
0
' Oxygen .
‘ 5% 02/95% He .
’3.
N
123,
1'5"
(‘1
.D
9
1 .m-
0 1 i 1 1 i
0 0.05 0.1 0.15 0.2 0.25

Sb/Si Atomic Ratio

Figure 4.6. XPS intensity ratio of Sb 3d3,2/Si 2p as a function of Sb/Si atomic ratio.

Table 4.2. CO conversion for Sb205/A1203 and SbZOS/Sioz at 450°C and 15 cc/minute

82

 

 

flow rate.
Weight % Percent Conversion Percent Conversion
Sb205 for Sb/AIZOf for Sb/SiOza
0 2.7 0.4
5 5.1 1.1
10 7.0 2.0
15 7.3 1.3
20 4.0 1.9
25 2.2 0.9
30 1.8 0.9
35 1.7 0.8
40 1.0 0.8
AleO4 9.6 —

 

 

 

 

3 - Error bars oft 10%.

 

83

the activity is lower than the value obtained for the support.

For the SblSiOz catalysts, the percent CO conversion increases with Sb loading up
to the SbZOSi catalyst. The activity decreases as the Sb loading increases from 20 to 40
wt.% Sb205. As shown from the table the activity of antimony oxide supported on
alumina is greater than that of the silica. However, the two series show similar trends in
the region of greatest activity (between 10 and 20 wt.%. Sb205). CO oxidation was
performed using the Ale04 catalyst. The average percent conversion at steady state is

9.6%.

4.4. Discussion

Structure of Calcined Antimony Oxide/A lumina Catalysts. For Sb/A1203 catalysts
with Sb205 loadings S 30 wt.%, absence of peaks that can be attributed to any antimony
oxide phases indicates that no crystalline antimony oxide phases are formed in these
catalysts. However, the observation of the Sb205 <400> peak for catalysts with 2 35
wt.% Sb205 indicates crystalline Sb205 is present. Semi-quantitative XRD analysis
suggests that these catalysts contain less than 0.2% crystalline Sb205. In addition, the
broad peak centered at approximately 29° (2 theta) is in the same region of the crystalline
Sb205 <400>. This may be attributed to an amorphous form of antimony oxide. The
XPS Sb 3d3,2 binding energies measured for Sb/A1203 catalyst series indicates that the
antimony is in the *5 oxidation state and does not vary as a fiinction of antimony loading.

The XPS Sb/Al intensity ratios measured for the alumina catalysts with S 30 wt.%
Sb205 show a linear increase in intensity ratio as the Sb/Al atomic ratio increases. This
suggests that the antimony phase is well dispersed over the alumina carrier up to 30 wt.%.
For higher Sb loadings, the deviation from linearity indicates that the antimony oxide is

not as highly dispersed over the alumina support. This is consistent with XRD results.

84

Further, the absence of any significant change in Sb/Al intensity ratio following reactions
suggests that the antimony oxide phase is not mobile under reaction conditions.

Structure of Calcined Antimony Oxide/Silica Catalysts. For Sb/Si02 catalysts
with S 30 wt.% Sb205, the absence of peaks that can be attributed to antimony oxide
phases indicate that no crystalline forms of antimony oxide are present. However, the
observation of Sb205 <400> peak for catalysts with 25 wt.% and 2 35 wt.% Sb205
indicates that crystalline Sb205 is present. In addition, the broad peak centered at
approximately 29° (2 theta) is in the same region of the crystalline Sb205 <400>. This
result indicates that an amorphous form of antimony oxide is produced on the support.
The XPS Sb 3d3,2 binding energies measured for Sb/Si02 catalyst series indicates that the
antimony is in the +5 oxidation state and is independent of antimony loading. Binding
energies for the series are similar to those reported previously [3 8,39].

For Sb/SiOz catalysts, the linear increase in Sb/Si XPS intensity ratio with
increasing Sb loading suggests that antimony oxide is highly dispersed. This is consistent
with XRD results which showed a broad feature that we have attributed to dispersed
antimony oxide. In addition, no change is observed under reaction conditions. This
indicates that the reactions do not affect the dispersion of the antimony.

Structure ofReacted Antimony Oxide/A lamina and Silica Catalysts. The Sb 3d3,2
XPS binding energy for Sb205 is 5397-5405 eV [40]. The Sb 3d3,2 XPS binding energy
measured for Sb203 is 539.5 eV. Thus, since the binding energy difference between Sb3+
and Sb5+ is approximately 0.7 eV, XPS cannot readily distinguish between these two
species [41-44]. Since the difference in binding energies is small, a broad peak results
when both peaks are present. Thus, curve fitting of the Sb XPS peak has some ambiguity.
As a result, it is not possible to determine if changes are occurring as a function of
reaction condition.

Comparison of Antimony Oxide on Alumina and Silica. The concentration of the

surface OH groups/surface area of the alumina support is 1.7x10'3 mmol/m2 and for the

85

silica support the value is 9.1x10'4 mmol/m2 [45]. Thus, the ratio of hydroxyls of the
alumina support to the hydroxyls of the silica is approximately 2:1. If it is assumed that
one Sb-alkoxide reacts with one hydroxyl on the support, monolayer capacity of the
alumina is expected to be approximately 35 wt.% and of silica approximately 40 wt.%
(taking the surface area of the support into effect). XPS and XRD data for the alumina
and silica supported catalysts are in agreement with this prediction.

Influence of Catalyst Structure on C O Oxidation Activity. CO oxidation reactions
have shown that the most active catalysts contain 10 to 20 wt.% loadings of Sb205. As
the loading increases (> 25 wt.%) in both series, the activity of the catalyst decreased. In
the case of 30-40 wt.% Sb/A1203 the activity is lower than that of the support. As for
SiOz, the activity decreases to approximately 0.8% which is higher than that of the pure
support (0.4%). The drastic change in the support from that of the 40 wt.% catalyst may
be due to the activity of Sb205 itself. This is shown by the activity of the higher loadings
of the active phase. At the loadings of greatest activity, it is possible that a spillover, as
seen with Pt catalysts, is occurring between the Sb and the support. The trend of
observing a maximum then as the loading increases lends support to this explanation.
Once the surface approaches saturation with the Sb phase, the interaction sites available
between the support and the Sb decrease, which decreases the activity of the catalyst
itself. Davydov [46] used CO to probe VzOs/A1203 catalysts and observed spillover of
carbon-oxygen species from the active vanadium pentoxide component on the alumina
carrier.

CO oxidation on the aluminum antimonate support showed an increase in
conversion over the supported Sb205/A1203. This suggests that the interaction between
Sb and A1 is important in this reaction. Straguzzi et a1. [10] studied CO oxidation and
found that at 460°C the conversions were less than 10% and showed activity of 1.0
mmol/mzh. Straguzzi suggests that because of the difficulty in reducing Al“, the activity

of the aluminum antimonate is attributed to Sb cations.

 

86

4.5. Conclusions

The use of surface and bulk spectroscopic techniques and CO oxidation activity
data for Sb205/A1203 and Sb205/S102 catalysts leads to the following conclusions:

1. XRD and XPS data indicate that Sb is well dispersed over the A1203 and SiOz
carriers.

2. Semi-quantitative XRD for A1203 suggests that less than 0.2 wt.% crystallites
of Sb205 are present and less than 0.5 wt.% crystallites of Sb205 are present in the Si02
supported catalysts. We tentatively ascribe this to errors in catalyst preparation.

3. The binding energy of Sb indicates that the antimony is in the +5 oxidation state
in both catalysts and is independent of Sb loading and reaction conditions.

4. CO oxidation suggests that the highest activity occurs with active phase
loadings of 10-20 wt.% Sb205 on both A1203 and SiOz. As loading increases above this

level, the activity decreases. This is tentatively ascribed to spillover.

References

1. RK. Grasselli and JD. Burrington, Adv. Catal., 30, 133, 1981.

2. R.K. Grasselli, React. Kinet. Catal. Lett., 35, 327, 1981.

3. G. Centi and F. Trifirc‘), Catal. Rev. Sci. Eng, 28, 165, 1986.

4. 1.0 Vedrine, G. Coudurier, M. Forissier, and 1.0 Volta, Catal. Today, 1, 261,

1987.

5. F.J. Berry, Adv. Catal., 30, 97, 1981.

6. SE. Golunski and D. Jackson, App]. Catal., 48, 123, 1989.

7. RK. Grasselli and J.L. Callahan, J. Catal., 25, 273, 1972.

8. GI Straguzzi, K.B. Bischoff, TA. Koch, and 6.0 Schuit, App]. Catal., 25, 257,
1986.

9. G1. Straguzzi, K.B. Bischoff, T.A. Koch, and QC. Schuit, J. Catal., 103, 357,
1987.

10. G1. Straguazzi, K.B. Bischoff, T.A. Koch, and G.C.A. Schuit, J. Catal., 104, 47,
1987.

11. F. Sala and F. Trifirc‘), J. Catal., 34, 68,1974.
12. V. Fattore, Z.A. Furhman, G. Manara, and B. Notari, J. Catal., 37, 223, 1975.

LEE-.ufl'. Airh- ' ..‘ fi.‘ .1“ -’

 

13.

14

15.

16.

17.
18.

19.

20.
21.

22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.

41.

87

M. Burriesci, F. Garbassi, M. Petrisa, and G. Petrini, J. Chem. Soc, Faraday
Trans 1, 78, 817, 1982.

I. Aso, S. Furukawa, N. Yamazoe, and T. Seiyama, J. Catal., 64, 29, 1980.

Y. Boudeville, F. Figueras, M. Foissier, J-L. Portefaix, and J.C. Vedrine, J. Catal.,
58, 52, 1979.

J.L. Barclay, 1.11. Bethell, J.°B. Bream, D.J. Hadley, RH. Jenkins, no. Stewart,
and B. Wood, Brit. Patent 864,666.

F.J. Berry and D]. Smith, J. Catal., 88, 107, 1984.

J.C. Volta, B. Benaichouba, I. Mutin, and J.C. Vedrine, App]. Cata]., 8, 237.
1983.

GA. Zenkovets, D.V. Tarasova, T.V. Andrushkevich, G.I. Aleshina, T.A. Nikoro,
and RG. Ravilov, Kinet. Catal., 20, 380, 1979.

M.-Y. Lo, S.K. Agarwal, and G. Marcelin, J. Catal., 112, 168, 1988.

IL Mikhailova, I.S. Sazonova, N.P. Keier, and TV. Belodtseva, Kinet. Katal,
9(3), 565, 1968.

FM. Ali-Zade, NM. Mardanova, PM. Sorgin, RM. Talshinskii, AA.
Medzhidov, and RC. Rizaev, U.S.S.R. Patent 49,052,169, 1991.

S. Blumrich, R. Brand, B. Engler, W. Honnen, and E. Koberstein, European
Patatent 410,440, 1991.

Y. Ohno and K. Morishita, Japanese Patent 49,052,169, 1974.

G. Busa, Langmuir, 2, 577, 1986.

GO Bond, S. Flamerz, and R. Shukri, Faraday Discuss. Chem. Soc, 87, 65,
1989.

Y. Gushikem, C.R.M. Peixoto and LT. Kubota, New Developmemts in Ion
Exchange, ICIE, 607, 1991.

S. Denofie, Y. Gushikem, and C.U. Davanzo, Eur. J. Solid State Inorg. Chem,
28, 1295, 1991.

M. Shirai, K. Asakura, and Y. Iwasawa, J. Phys. Chem, 95, 499, 1991.

GO. Bond and K. Briickman, Faraday Discuss. Chem. Soc, 72, 235,198].

E.C. Alyea, K.F. Brown, and K.J. Fisher, J. Mol. Catal., 63, L11, 1990.

E.V. Benvenutti, Y. Gushikem, C.U. Dananzo, 8.0 de Castro, and IL. Torriani,
J. Chem. Soc. Faraday Trans, 88, 3193, 1992.

A. Proctor, University of Pittsburgh.

C. Defosse, P. Canesson, P. G. Rouxlet, and B. Delmon, J. Cata]., 51, 269, 1979.
F. P. J. M. Kerkhof and J. A. Moulijn, J. Phys. Chem, 83, 1612, 1979.

J. H. Scofield, J. Electron Spectrosc. Relat. Phenom, 8, 129, 1979.

D. R. Penn, .1. Electron Spectrosc. Relat. Phenom, 9, 29, 1976.

E.V. Benvenutti, Y. Gushikem, C.U. Dananzo, 8.0 de Castro, and IL. Torriani,
J. Chem. Soc. Faraday Trans, 88, 3193, 1992.

A. Li-Dun, J. Zhi-Cheng, and Y. Yuan-Gen, in ”Catalyst Deactivation 1987”, ed.
B. Delmon and GP. Froment, 159, 1987.

RA. Lemberanskii, F.M. Poladov, I.B. Annenkova, and AP. Shepelin, React.
Kinet. Catal. Lett., 32(1), 103, 1986.

WE. Morgan, W.J. Stec, and JR. Van Wazer, Inorg. Chem, 12, 953, 1973.

 

42.

43.

45.

46.

88

T. Birchall, J .A. Connor, and 1H. Hiller, J. Chem. Soc, Dalton Trans, 2003,
1975.

AF. Orchard and G. Thornton, J. Chem. Soc.,Dalton Trans, 123 8, 1977.

F. Garbassi, Surf Interface Anal, 2, 165, 1980.

J. Kijenski, A. Baiker, M. Glinski, P. Dollenmeier, and A. Wokaun, J. Catal., 101,
1, 1986. '

AA Davydov, Zh. Fiz. Khim., 65,1803, 1991.

 

Chapter 5

. Future Work

5.1. Characterization

Characterization of Mixed Metal Oxide Catalysts. In addition to the use of XPS
to determine the nature and distribution of the active sites in complex mixed oxides, ion
scattering spectroscopy (188) can also provide information on the active sites(s) following
various reactions. In the case of vanadium oxide and antimony oxide monolayer catalysts,
reactions of the surface phase with the reducing molecules could lead to surface migration
and subsequent multilayer formation. Monolayer-to-multilayer transitions can be difficult
to follow using XPS due to the relatively large sampling depth of the technique. Thus,
188 can be useful to follow subtle changes in the geometry of the vanadium oxide and
antimony oxide caused by surface reactions. Along with XPS, Raman spectroscopy can
also provide information on the active site(s) under various reaction conditions. Raman
spectroscopy can also be used to monitor differences in the chemical reactivity of surface
species present.

In addition to using EPR to identify V4+ in the reacted vanadium oxide on
aluminum antimonate catalysts, it would also be useful to characterize the active site(s) by
performing in situ reaction studies on both the V/AleO4 and Sb/SiOz and Sb/A1203
systems. An example of this use of EPR can be found by the double bond isomerization of
butene over y-alumina [1]. Both the isomerization reactions and the butene-deuterium
exchange reaction have been extensively studied with a view to understanding the nature
of the active sites on alumina.

NMR can also be used in much the same way as EPR to study active site(s), by

using 13C NMR studies of adsorbed molecules. For example, Stejskal et al. [2] studied

89

 

sum.- .'

90

physically adsorbed C02 on zeolites. In addition, 13C NMR has been used to study
chemisorbed molecules [3]. NMR is unique in its ability to view molecules on the time
scale of a surface reaction. Since the motion of molecules and ions on the surface is an
integral part of heterogeneous catalysis, these studies have aided in the development of a
dynamic picture of adsorption and proton mobility.

In situ IR will also provide information on the adsorption of probe molecules or
reaction gases on the active site(s) of the catalysts. The IR cell shown in Figure 5.1 will
allow the catalyst to be studied under controlled atmosphere and temperature. The gases

can be delivered to the in situ IR cell by the system shown in Figure 5.2.

5.2. Vanadium Oxide/Aluminum Antimonate Catalysts

Propane Oxidation and Ammoxidation. As mentioned earlier, there has been an
increased interest in selective oxidation and ammoxidation of inexpensive feedstocks, such
as propane and methane. The next area of investigation would be the study of propane
oxidation and ammoxidation over the V/AleO4 catalyst. In order to learn more about
the mechanism(s) for propane oxidation and ammoxidation, careful kinetics studies on the
catalyst must be performed. In order to obtain this information, it will be necessary to
study the effects of propane, oxygen and ammonia partial pressures on the rates of
propylene, acrolein, acetaldehyde, acrylonitrile, acetonitrile and carbon oxides formation.

Centi et a1. [4] reported that specific oxygen and ammonia concentrations exist
which enhance the rates of propylene, acrylonitrile and acetonitrile formation over a V-Sb-
W (1:5: 1) A1203 (70 wt.%) catalysts. The kinetic experiments proposed will determine if
there is a similar ”optimum" ratio that exists for V/Ale04. If such a ratio exists, it will
then be useful to perform IR and surface spectroscopic studies to determine the surface
phases responsible for the enhanced catalytic properties.

Influence of Excess SbOx on AleO,, supported V cata]ysts. It is well known that

91

Heater ports

/\ Vent 8
Thermocouple

Leads

1

   

  

 

 

 

Clip Thermocouple

S'd Vi Top View
I e ew Center Piece

 

Figure 5.1. In situ IR cell.

92

 

 

 

 

 

 

 

 

 

 

 

Z-EOUTE
W-MASSFLOWCON'I'ROUER
I'm-FOURIERTRANSFORMW
tic-mm

 

 

 

 

Figure 5.2. Gas handling system to in situ IR cell.

93

crystalline and monolayer phases have different effects on activity. Recently, Beny and
Brett [5-7] reported that biphasic interactions between VSbO4 and a-Sb204 are important
for the selective oxidation of propylene. However, the influence of excess SbOx phase on
the structure and activity of the catalysts is not well understood. The effects of excess
SbOx phase on the chemical state and dispersion of the supported V oxide and
subsequently on the activity and selectivity of the mixed oxide catalysts for propylene and
propane oxidation and ammoxidation should be investigated. From the different phases
produced, it will be possible to determine the effects of crystalline SbOx and highly
dispersed SbOx phases on the properties of V/AleO4 catalyst. Future experiments will

involve preparation and characterization of V/SbOx/AleO4 catalysts.

5.3. Antimony Oxide! Supported Catalysts

Sb/Supported Catalysts. There has been considerable research effort in methane
coupling and total oxidation of methane. As mentioned previously, Lo et al. [8] reported
a-Sb204 was a selective methane coupling catalyst. Dang and Ding [9] found that
Sb/Si02 catalysts that contain an Sb-SiOz surface compound (Sb loadings < 5 %) are
active for condensation reactions. However, the condensation activity decreases
significantly on the formation of a-Sb204 (Sb loadings > 5 %). Thus, the combination of
methane coupling and condensation of isobutene and formaldehyde to is0prene may be
useful reactions to probe the structure of supported Sb catalysts. It should be noted that
regardless of the ability of the test reactions to probe the structure of the mixed oxides,
valuable information about the nature of active site(s) in Sb supported catalysts will be
obtained.

By varying the partial pressures of methane and oxygen it will be possible to study
the effects of the partial pressures on the rate of carbon oxides produced. In addition, by

having a higher level of methane it may be possible to control the selectivity of methane

 

94

coupling over total oxidation of methane. In order to increase the selectivity of total
oxidation it would be necessary to have a higher partial pressure of oxygen over methane.
However, these kinetic experiments will determine if Optimum ratios exist for each
reaction (i.e. methane coupling and total oxidation). If such a ratio exists, IR and surface
spectroscopic studies will be used to determine the surface phases responsible for the
enhanced catalytic properties.

In addition, the effect of calcination temperature on the catalyst structure and
activity should be examined. Sb205 decomposes to form B—Sb204 at 730°C and Ot-Sb204
at 950°C. By performing solid-state reactions, it will be possible to determine Optimum
conditions for methane coupling and/or total oxidation of methane as a function of

antimony oxides formed.

5.4. Activation Energy

Activation Energy. It would also be useful to determine the activation energy for
oxidation reactions of the V on aluminum antimonate and the Sb on alumina and silica, to
improve our understanding of these reactions. Jonson et al. [10] have performed
activation energy studies Of V on titania. From this, it was found that at least two
different active sites are present with a active phase loading of 005% V; one for the
support, which is the least active phase and one for the vanadium species, which is the
more active phase. Also, at 2% V and above, the active part of the support surface is
covered and changes in the Arrhenius parameters can be attributed to the presence in these
catalysts of an additional V species at higher loadings. By performing such activity
measures and correlating the data with that obtained from spectroscopic techniques, it will
be possible to better define the nature Of metal oxide-support interaction and the influence

of the metal oxide dispersion and reactivity on the catalytic properties of the mixed oxides.

 

 

95

References

1. J.H. Lunsford, L.W. Zingery, and MP. Rosynek, J. Catal., 38, 179, 1975.

2. E0. Stejskal, J. Schaefer, J.M.S. Henis, and MK. Tripodi, J. Chem. Phys, 61,
2351, 1974.

3. JJ. Chang, A. Pines, J.J. Fripiat, and HA. Resing, Surface Sci, 47, 661, 1975.

4. G. Centi, R.K. Grasselli, E. Patane, and F. Trifiro,"New Developments in Selective
Oxidation,” Stud Surf Sci. Catal., 55, 515, 1990.

5. F.J. Berry, ME. Brett, and W.R. Patterson, J. Chem. Soc. Dalton, 1, 9, 1983.

6. F.J. Berry, M.E. Brett, and W.R. Patterson, J. Chem. Soc. Dalton, 1, 13, 1983.

7. F.J. Berry and ME. Brett, J. Catal, 88, 232, 1984.

8. M.-Y. Lo, S.K. Agarwal, and G. Marcelin, J. Cata]., 112, 168, 1988.

9. Z. Dang and S. Ding, Fenzi Cuihua, 1(3), 146, 1987.

10. B. Jonson, B. Rebenstrol, R. Larsson, and S.L.T. Andersson, J. Chem. Soc,

Faraday Trans. 1, 84(10), 3547, 1988.