Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A study of the catalytic reaction of olefins with methane Liu, Qingdong 1998

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1998-271889.pdf [ 6.02MB ]
Metadata
JSON: 831-1.0058632.json
JSON-LD: 831-1.0058632-ld.json
RDF/XML (Pretty): 831-1.0058632-rdf.xml
RDF/JSON: 831-1.0058632-rdf.json
Turtle: 831-1.0058632-turtle.txt
N-Triples: 831-1.0058632-rdf-ntriples.txt
Original Record: 831-1.0058632-source.json
Full Text
831-1.0058632-fulltext.txt
Citation
831-1.0058632.ris

Full Text

A STUDY OF THE CATALYTIC REACTION OF OLEFINS WITH METHANE by Qingdong Liu B. A. Sc., Tianjin University, Tianjin, 1982 M. A. Sc., Tianjin University, Tianjin, 1984 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PfflLOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Chemical and Bio-Resource Engineering) We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F B R T T I S H C O L U M B I A March 1998 © Qingdong L iu , 1998 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver, Canada DE-6 (2/88) dedicated ta my wife JtiActa ABSTRACT The catalytic reaction of CH4 with C3H6, referred to as CH4/C3H6 homologation or coupling, was investigated in an attempt to selectively produce C4 hydrocarbons. The CH4/C3H6 coupling reaction was conducted over various Ni catalysts at temperatures in the range 300 °C to 375 °C and 101 kPa. The effects of process variables, catalyst promoters and different supports on the CH4/C3H6 coupling reaction were examined. In addition, the role of the carbonaceous species deposited on the catalyst surface during catalyst reduction in CH4 and during the CH4/C3H6 coupling reaction, was investigated. In preliminary experiments, Ni catalysts supported on AI2O3 were modified by K and P, and used for the CH4/C3H6 coupling reaction. The calcined catalysts were reduced in CH4 at 600 °C for 1 hour before reaction. Compared with Ni/Al 20 3 catalyst, the Ni/K/Al203 and Ni/P/Al203 catalysts showed higher C3H6 conversion and C4 selectivity. The rmximum selectivity of 10.0 mole % to the desired C4 product was achieved over the Ni/K/Al203 catalyst at 350 °C and 101 kPa with a feed gas composition of 90 mol % CH4/IO mol % C3H6. The C3H6 conversion decreased significantly with time-on-stream. During catalyst reduction in CH4, decomposition of CH4 occurred and a large amount of carbonaceous species deposited on the catalyst surface. Subsequently, during the CH4/C3H6 coupling reaction, additional carbonaceous species were deposited on the catalyst surface. The presence of two types of carbonaceous species were identified using temperature programmed surface reaction (TPSR) in H 2 . The first type of carbonaceous species was relatively active, and reacted with H 2 at low temperature, in the range 192 °C to 237 °C. A second type of i i carbonaceous species was relatively inactive, and reacted with H 2 at high temperature, in the range 547 °C to 667 °C. The amount of the low temperature carbonaceous species present on the catalyst surface after the CH4/C3H6 coupling reaction, was shown to correlate with the C4 yield. The significance of the treatment of catalyst in CH4 before reaction was shown to be less important than what was claimed in previous studies. The CH4 reduced catalyst surface had higher C3H6 conversion than H 2 reduced catalysts but the C4 selectivity was unchanged. A model of the carbonaceous species deposition process on the catalyst surface was proposed. The effect of supports on the CH4/C3H6 coupling reaction was also investigated. Different reaction paths were proposed. The major reactions over non-zeolite supported Ni catalysts were C3H6 decomposition and C3H5 hydrogenation with a low activity for homologation. The C3H5 conversion was high and C3H6 metathesis and olefin dimerization were the dominant reactions over the Na-Y supported Ni catalyst. The high activity of the Ni/Na-Y catalyst was a consequence of the acidic property of the Na-Y support. After the acidic property of the Ni/Na-Y catalyst was neutralized, the catalyst activity decreased dramatically and evidence for the C3H6 homologation was apparent. Results of the present study have shown that C 4 selectivity, particularly 1-butene, was formed by the homologation reaction: CH X + C3H6 -> I-C4H8, where CH X is a surface carbon species generated either from catalyst reduction in C H 4 or from C 3 H 6 decomposition reactions that occur during reaction. Most importantly, the present study has shown that gas phase CH4 does not homologate C3H6 directly at the conditions studied. The maximum C 4 selectivity of 10 %, obtained using Ni/K/Al 203 catalyst operated at 350 °C and lOlkPa, was also very low which together with the fact that direct CH4/C3H6 homologation did not occur, point to the need for a i i i multi-step reaction if this approach is to be utilized for direct CH4 upgrading. iv T A B L E OF C O N T E N T S Page Abstract ii Table of Contents v List of Tables viii List of Figures ix Acknowledgment vx C H A P T E R 1. I N T R O D U C T I O N 1 1.1 Background 1 1.2 Objectives of this Study 5 C H A P T E R 2. L I T E R A T U R E R E V I E W 7 2.1 Direct CH4 Conversion 7 2.2 Homologation of C 3H6 with Cft, 10 2.3 Chemisorption and Carbonaceous Species on Metals 16 2.3.1 Chemisorption of C H X (x=l,2 and 3) 16 2.3.2 Chemisorption of Olefins 17 2.3.3 Carbonaceous Species on Metals 18 2.4 Related Chemical Reactions 19 2.4.1 C 3H6 Homologation with C K , 20 2.4.2 Fischer-Tropsch Synthesis 20 2.4.3 C 3H6 Decomposition and the Formation of Carbonaceous Species 22 2.4.4 Metathesis 23 2.4.5 Dimerization and Cracking 24 2.4.6 Hydrogenation of Olefins 25 2.4.7 Hydrogenolysis 26 v 2.5 Summary of Literature Review 27 C H A P T E R 3. E X P E R I M E N T A L P R O C E D U R E S 29 3.1 The Experimental Apparatus 29 3.2 Catalyst Preparation 31 3.2.1 Preparation of Ni /K/ALOs (Ni 7 wt%, K 1 wt%, A 1 2 0 3 92 wt%) by Incipient Wetness Impregnation - 32 3.2.2 Preparation of Ni/Na-Yby Ion Exchange 33 3.3 Catalyst Characterization 34 3.3.1 Surface Area 34 3.3.2 X-ray Diffraction (XRD) 36 3.3.3 Temperature Programmed Reduction (TPR) 36 3.3.4 Temperature Programmed Desorption (TPD) 37 3.4 The Catalytic Reaction and TPSR 38 C H A P T E R 4. P R E L I M I N A R Y S T U D Y OF T H E CHVCsHe C O U P L I N G R E A C T I O N O V E R MODIF IED N I / A L 2 0 3 C A T A L Y S T S 42 4.1 Modified N i / A l 2 0 3 Catalysts 42 4.2 CiU/CjiU Coupling Reaction over Modified N i Catalysts '• 46 C H A P T E R 5. C A R B O N A C E O U S DEPOSITS O N NI C A T A L Y S T S 52 5.1 Study of Surface Carbon Using TPSR 52 5.2 The Effect of Carbonaceous Species Generated during CH4 Reduction 59 5.3 Summary 70 C H A P T E R 6. E F F E C T OF P R O C E S S V A R I A B L E S O N C H 4 / C 3 H 6 COUPL ING71 6.1 The Effect of Reduction Time on the Coupling Reaction 71 6.2 The Effect of Reaction Temperature on the Coupling Reaction 74 6.3 The Effects of Residence Time on the Coupling Reaction 80 6.4 Summary 91 vi C H A P T E R 7. E F F E C T OF DIFFERENT SUPPORTS O N CHVCsHe C O U P L I N G O V E R NI C A T A L Y S T S 92 7.1 The Study of Silica Supported N i Catalysts 92 7.1.1 Activity Tests of Silica Supported N i Catalysts 93 7.1.2 The Effect of K on the CHVCsHs Coupling Reaction 98 7.1.3 Summary 101 7.2 The Study of SAPO-5 Supported N i catalysts 101 7.3 The Study of Na-Y Supported N i Catalysts 110 7.3.1 Preliminary Tests 111 7.3.2 The Effects of Residence Time on the Coupling Reaction 117 7.3.3 The Acidity Changes of Ni /Na-Y Catalysts on Coupling Reaction 121 7.4 Effects of Supports on the C3H<; / CH* Coupling Reaction 126 C H A P T E R 8. C O N C L U S I O N S A N D F U T U R E W O R K 138 8.1 Summary of Results from the Present Study 138 8.3 The Contribution of this Work 140 8.4 Future Work 141 R E F E R E N C E S 143 Appendix 1. Repeatability of Experimental Data 149 Appendix 2. Experimental Data 151 Appendix 3. Carbon Balance Data 178 vii LIST OF T A B L E S Pa2e Table 2-1. Capital and operating cost comparison, gas-to-liquid processes 8 Table 2-2. The performances of different catalysts in the FT synthesis [45] 22 Table 3-1. Calculation procedure of reaction results 41 Table 4-1. Properties of P and K promoted 7 wt % Ni on AI2O3 catalysts 42 Table 4-2. Results of C3H5 coupling with CH4 at 350 °C over Ni/Al 20 3 catalysts after 60 minutes time-on-stream. Feed = 90 mol % CH4/10 mol % CsFfc; Flowrate = 20 cm3 (20 °C and 101 kPa ) /min. Catalyst mass = 1.1 g 47 Table 5-1. Carbon content of catalyst after reduction in CH4 at 600 °C for 54 Table 5-2. Carbon content of catalyst after reduction in CH4 at 600 °C for 1 hour and reaction in 90 mol % CH4/IO mol % CsHe at 350 °C for 1 hour measured by TPSR in H 2 57 Table 5-3. Catalyst surface carbon measured by temperature programmed surface reaction (TPSR) experiments after reduction in CH4 at 600 °C for 1 hour compared to surface Ni measured by temperature programmed desorption (TPD). 64 Table 5-4. Catalyst surface carbon measured by temperature programmed surface reaction (TPSR) experiments after reduction in CH4 at 600 °C for 1 hour and reaction in CH4/C3H6 at 350 °C for 1 hour compared to surface Ni measured by temperature programmed desorption (TPD). 64 Table 7-1 Ni dispersion on Ni/Si02 or Ni/K/Si02 catalysts measured by TPD in H 2 94 Table 7-2 Properties of Ni/SAPO-5 and Ni/Na-Y measured by TPR in H 2 [87] 103 Table 7-3. C3 and C4 product mole % in total products after 60 minutes time-on-stream in the CHVCsHg coupling reaction over Ni/SAPO-5 catalyst at 350 °C and 101 kPa. 10 % C3H6 / 90 % CH4 feed gas 109 vii i LIST OF F IGURES Page Figure 1-1. Uses of CH4 in industrial processes [2] Figure 2-1. The proposed mechanism for the homologation of C3H5 with CEL» using Ni-supported catalyst. In CHX, x = 0, 1, 2 or 3. 15 Figure 2-2 Olefin adsorption on transition metals 18 Figure 3-1. Experimental Apparatus for CH4/C3H6 Coupling 30 Figure 4-1. Exit gases measured by MS during TPR of Ni/P/Al203 ( 7/3/90 wt % ) catalyst in CH4 at a flowrate 20cm3/min. and a temperature range 50 °C to 900 °C at a ramp rate 30 °C/min. 44 Figure 4-2. X-ray diffraction spectrum of Ni/K/Al203 catalyst before and after reduction in CFLj46 Figure 4-3. C3H6 conversion over Ni/K /A1203 (1% K) and Ni/P/Al203 (1% P). Reaction at 350 °C and 101 kPa. Feed gas = 90 mol % CH4/ 10 mol % C3H6. 49 Figure 4-4. C 4 selectivity over Ni/K /A1203 (1% K) and Ni/P/Al203 (1% P). Reaction at 350 °C and 101 kPa. Feed gas = 90 mol % CH4/10 mol % C3H6. 50 Figure 5-1. TPSR profile of A1203 supported catalysts following 1 hour reduction in CH4 at 600 °C ' 53 Figure 5-2. TPSR profile of A1203 supported catalysts following 1 hour reduction in CH4 at 600 °C showing low temperature peaks 55 Figure 5-3. TPSR profile of A1203 supported catalysts following 1 hour reduction in CH4 at 600 °C and reaction in CH, / C3H6 at 350°C for 1 hour 57 Figure 5-4. The relationship between C 4 selectivity and the moles of CH4 released during TPSR at low temperature peak range after 1 hour reaction at 350 °C 58 Figure 5-5. The relationship between C3H6 conversion and the moles of CH4 released during TPSR at low temperature peak range after 1 hour reaction at 350 °C 58 Figure 5-6. Comparison of C3H6 conversion (Figure A) and C4 selectivity (Figure B) over Ni/K7A1203 catalysts reduced in H 2 and CH4, Reaction at 350 °C, lOlkPa and 10 % IX C3H6 / 90 % CH4 feed gas 60 Figure 5-7. Comparison of C 3 (Figure A) and C 2 selectivity (Figure B) over Ni/K/Al 20 3 catalysts reduced in H 2 and CH4. Reaction at 350 °C, lOlkPa and 10 % C3H6 / 90 % CH4 feed gas. 66 Figure 5-8. Comparison of n-C4Hio in total C 4 over Ni/K/Al 20 3 catalysts reduced in H 2 and CH4. Reaction at 350 °C, lOlkPa and 10 % C3He / 90 % CH4 feed gas 68 Figure 5-9. Comparison of 1-butene (Figure A) and isobutylene (Figure B) in total C4's over Ni/K/Al 20 3 catalysts reduced in H 2 and CH4. Reaction at 350 °C, lOlkPa and 10 % C3H6 / 90 % CH4 feed gas • _ _ 69 Figure 6-1. The effect of reduction time in CH4 on the C3H6 conversion over Ni/K/Al 20 3. Reaction at 350 °C, 101 kPa and 10 % C3H<s / 90 % CH, feed gas 72 Figure 6-2. The effects of reduction time in CH4 on the C 4 selectivity over Ni/K/Al 20 3. Reaction at 350 °C, 101 kPa and 10 % CaHe / 90 % CH, feed gas 73 Figure 6-3. The effects of reaction temperature on the C3H6 conversion over Ni/K/Al 20 3. Reaction at 101 kPa and 10 % C3He / 90 % CH4 feed gas 75 Figure 6-4. The effects of reaction temperature on the C 4 selectivity over Ni/K/Al 20 3. Reaction at 101 kPa and 10 % C3Hs / 90 % CH, feed gas 75 Figure 6-5. The effect of reaction temperature on the C 3 (Figure A) and C 2 (Figure B) selectrvities over Ni/K/Al203. Reaction at 101 kPa and 10 % C3H<; / 90 % CH, feed gas 76 Figure 6-6. The changes of the ratio of C4 production and C3H6 conversion with time-on-stream77 Figure 6-7. The estimation of activation energy for the C3H6 conversion over Ni/K/Al 20 3. Reaction at 101 kPa and 10 % C3H6 / 90 % CH, feed gas 79 Figure 6-8. The effect of residence time (x) on the ln(l-x) (x is C3H6 conversion) over Ni/K/Al 20 3 in 30 minutes time-on-stream. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CH, feed gas. 81 Figure 6-9. The effects of residence time on the selectivity - 5 minutes time-on-stream- over Ni/K/Al 20 3 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CH, feed gas 81 Figure 6-10. The effects of residence time on selectivity - time-on-stream 30 minutes (Figure A) and 130 minutes (Figure B) - over Ni/K/Al 20 3 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CH4 feed gas 83 Figure 6-11. The changes of paraffin and olefin distribution in total C4 with time-on-stream over Ni/K/Al 20 3 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3He / 90 % CH4 feed gas 85 Figure 6-12. The effects of residence time on the paraffin and olefin in the total C4's - 30 minutes time-on-stream - over Ni/K7A1203. Reaction at 350 °C, 101 kPa and 10 % C3He / 90 % CH4 feed gas 86 Figure 6-13. The effects of residence time on the paraffin in the total C4's - 5 minutes time-on-stream in (Figure A) and 105 minutes time-on-stream (Figure B) - over Ni/K/Al 20 3 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3He / 90 % CH4 feed gas 87 Figure 6-14. The effects of residence time on the olefin in the total C4's -in stream 30 minutes (Figure A) and 105 minutes (Figure B) - over Ni/K/Al 20 3 catalysts at 350 °C. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CH4 feed gas 88 Figure 6-15. The selectivity changes with time-on-stream over Ni/K/Al 20 3 catalysts. Reaction at 350 °C, 101 kPa and 10 % CsHg / 90 % He feed gas 90 Figure 7-1. C3He conversion over Ni/Si02 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CH4 feed gas. 94 Figure 7-2. C3H<; conversion versus Ni sites in mmole/g-catalyst on the catalyst surface over Ni/Si02 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3He / 90 % CH4 feed gas... 94 Figure 7-3. C 4 selectivity over Ni/Si02 catalysts, Reaction at 350 °C, 101 kPa and 10 % CsFV 90%CH4 feed gas. 95 Figure 7-4. C 3 selectivity (Figure A) and C 2 selectivity (Figure B) over Ni/Si02 catalyst Reaction at 350 °C, 101 kPa and 10 % C3He / 90 % CH4 feed gas. 97 Figure 7-5. Comparison of C3H6 conversion over silica supported Ni catalysts, Reaction at 350 XI °C, 101 kPa and 10 % C3Hs / 90 % C H 4 feed gas. 99 Figure 7-6. Comparison of C4 selectivity over silica supported Ni catalysts, Reaction at 350 °C, 101 kPa and 10 % CaHs / 90 % C H 4 feed gas. 99 Figure 7-7. Comparison of C3 selectivity over silica supported Ni catalysts, Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CH4 feed gas. 100 Figure 7-8. Comparison of C 2 selectivity over silica supported Ni catalysts, Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % C H 4 feed gas. 100 Figure 7-9. Conversion of C3H$ over Ni/SAPO-5 catalysts prepared by ion exchange and impregnation methods. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CH, feed gas. 104 Figure 7-10. C4 selectivity over Ni/SAPO-5 catalysts prepared by ion exchange and impregnation methods. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CH, feed gas 104 Figure 7-11. C 3 selectivity (Figure A) and C 2 selectivity (Figure B) over Ni/SAPO-5 catalysts prepared by ion exchange and impregnation methods. Reaction at 350 °C, 101 kPa and 10 % CjHg / 90 % CH, feed gas 105 Figure 7-12. Cs+ selectivity over Ni/SAPO-5 catalysts prepared by ion exchange methods. Reaction at 350 °C, 101 kPa and 10 % CsHg / 90 % CFL, feed gas 106 Figure 7-13. The distribution of C4 paraffins over Ni/SAPO-5 catalysts prepared by ion exchange methods. Reaction at 350 °C, 101 kPa and 10 % C3H<; / 90 % CFL, feed gas 108 Figure 7-14. The distribution of C4 olefins over Ni/SAPO-5 catalysts prepared by ion exchange methods. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CH, feed gas 109 Figure 7-15. C3Hs conversion over Ni/Na-Y catalysts. Reaction at 350 °C,101 kPa and 10 % C3H6 / 90 % CH, feed gas. 111 Figure 7-16. Cs+,C4, C3 and C 2 selectivities over Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CH, feed gas. 112 Figure 7-17. C4 paraffins changes over Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CH, feed gas. 113 Figure 7-18. C 4 olefins changes over Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % xii C3H6 / 90 % C H 4 feed gas. 113 Figure 7-19. The conversion of C2H4 overNi/Na-Y catalyst. Reaction at 250 °C and 101 kPa and 10 % C2H4 / 90 % CH, feed gas. 114 Figure 7-20. The selectivities of the products over Ni/Na-Y catalyst, Reaction at 250 °C and 101 kPa and 10 % C2H4 / 90 % C H 4 feed gas. 115 Figure 7-21. The C 4 paraffins in total C 4 over Ni/Na-Y catalyst, Reaction at 250 °C and 101 kPa and 10 % C2H4 / 90 % CH, feed gas. 115 Figure 7-22. The C 4 olefins in total C 4 over Ni/Na-Y catalyst, Reaction at 250 °C and 101 kPa and 10 % C2H4 / 90 % CH, feed gas. 116 Figure 7-23. The effects of residence time changes on the C3H6 conversion over Ni/Na-Y catalyst at 30 and 105 minutes time-on-stream. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CH4 feed gas. 117 Figure 7-24. The effects of residence time changes on the C 4 selectivity over Ni/Na-Y catalyst at 30 and 105 minutes time-on-stream. Reaction at 350 °C, 101 kPa and 10 % CsHe / 90 % CFLt feed gas. 118 Figure 7-25. The effects of residence time changes on the C 2 , C3 and C 5 selectivity over Ni/Na-Y catalyst at 30 minutes time-on-stream. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CH4 feed gas. 119 Figure 7-26. The effects of residence time changes on the C 4 paraffins over Ni/Na-Y catalyst at 30 minutes time-on-stream. Reaction at 350 °C, 101 kPa and 10 % C3H6 /90 % CIT4 feed gas. 120 Figure 7-27. The effects of residence time changes on the C 4 olefins over Ni/Na-Y catalyst at 30 minutes time-on-stream Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CFLt feed gas. 121 Figure 7-28. The effects of a addition of a base on Ni/Na-Y catalyst on the C3H6 conversion. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CFLt feed gas 123 Figure 7-29. The effects of a addition of a base on Ni/Na-Y catalyst on C 4 selectivity. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CHt feed gas 123 Figure 7-30. The effects of a addition of a base on Ni/Na-Y catalyst on C 3 selectivity. Reaction at xi i i 350 °C, 101 kPa and 10 % C3H5 / 90 % CH4 feed gas 124 Figure 7-31. The effects of addition of a base on Ni/Na-Y catalyst on C2 selectivity. Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CFL, feed gas. 124 Figure 7-32. The effects of a base addition into Ni/Na-Y on the C 4 distribution at 30 minutes time-on-stream. Reaction at 350 °C, 101 kPa and 10 % C3FL5 190 % CFL, feed gasl25 Figure 7-33. C3FL5 conversion over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CFL, feed gas 127 Figure 7-34. C 4 selectivity % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % CsHg / 90 % CFL, feed gas 127 Figure 7-35. C3 selectivity % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % C£k 190 % CHjfeed gas 128 Figure 7-36. C2 selectivity % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % C3Hs / 90 % CFL, feed gas 128 Figure 7-37. Cs+ selectivity % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % CaHs / 90 % CH4 feed gas 129 Figure 7-38. Yield of C 4 % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CFL, feed gas 130 Figure 7-39. iso-C^io (Figure A) and n-CJlw (Figure B) % in total C4's over Ni/K/Al 20 3 and Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CFL, feed gas 133 Figure 7-40. 1-butene (Figure A) and isobutylene (Figure B) % in total C4's over Ni/K/Al 20 3 and Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CFL, feed gas 134 Figure 7-41. C-2-C4H8 (Figure A) and t -2 -C^ (Figure B) % in total C4's over Ni/K/Al 20 3 and Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % C3He / 90 % CFL, feed gas 135 Figure 7-42. The ratio of C 4 paraffin to C 4 olefin of over Ni/Na-Y (Figure A) and over Ni/K/Al 20 3 (Figure B). Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CFL, feed gas 136 XIV A c k n o w l e d g m e n t I would like to express my sincere gratitude to my supervisor Dr. K J. Smith for his guidance, responsible supervision and encouragement over the entire course of this work, especially for his excellent ideas and patient correction during the writing of my thesis and his financial support during my graduate study. I would like to give my appreciation to Dr. C. J. Lim and Dr. E. Ogryzlo for their suggestions and assistance during committee meetings and for reviewing my thesis. I would like to thank all members in Dr. Smith's research group for helpful discussions and suggestions. I also would like to thank all staff of the Workshop and Stores in the Department of Chemical and Bio-Resource Engineering for their help and assistance in my research. Finally, I would like to express my thanks to my family, Lihua Zou and Richard Liu, for their thorough understanding and wholehearted support. XV Chapter 1 Introduction 1 . 1 Background Natural gas is an abundant resource with world reserves of over 100 x 1012 m3 [1]. Although significantly smaller in magnitude than those of some other nations, Canada's reserves of natural gas are significant, totaling 2.4 x 1012m3. Most natural gas is used as a fuel for power generation and heating and about 7 % of the world production is consumed in chemical industries. Although current availability of natural gas exceeds demand, the resource is not renewable, nor is the large majority of reserves strategically placed. The use of natural gas as chemical feed stock increases its value compared with its use as a fuel. For use as a chemical feedstock it must be converted into more appropriate products, since its main component, CFLt, is chemically inert. There are several processes available for CFLt upgrading. The most important commercial process is natural gas steam reforming used in the production of synthesis gas (CO and H2). The produced synthesis gas is widely used in the chemical industry, for example in the production of CH 3OH. Since the desired product (CH3OH) is obtained via the intermediate production of synthesis gas, this approach to CFLt conversion is referred to as an indirect process. Other direct processes, for example, halogenation of CFLt to produce halohydrocarbons, reaction with ammonia to produce hydrocyanic acid and pyrolysis to produce carbon black, have very limited scales because there are limited applications for their products. Figure 1-1 presents a summary of commercial natural gas conversion processes. Since the CFLt stream reforming process is expensive and can account for more than 60 % l of the total cost of gasoline production from the Fischer-Tropsch process [3], other processes for CH, conversion have been under study since the beginning of the twentieth century to convert CFL, directly to useful chemicals or to a product that can be easily transported. Methanol Fischer-Tropsch, Carbonylation Hydrocarbons & Alcohols Hydrogen + N2 A m m i onia Methane + HCLHF Thermal, Electric Arc Halohydrocarbons + NH, Acetylene Pyrolysis Hydrocyanic Acid + Sulfur Carbon Black Carbon Disulfide Figure 1-1. Uses of CFL, in industrial processes [2] The direct thermal conversion of CH, to higher hydrocarbons requires temperatures above 1000 °C and carbon formation is a serious problem [4,5,6]. Since the thermal conversion is endothermic, it consumes considerable quantities of energy and the cost of the process is high. If the temperature for the CH, conversion reaction is lowered, the reaction is not thermodynamically favorable. 2 2 CH4 -> C2H4 + 2 H 2 AG 0 = +9.5 kcal/mole at 727 °C To lower the CH4 conversion temperature and make the conversion thermodynamically favorable, other approaches have been examined. One promising approach, oxidative coupling of CH4, has been paid much attention in recent years since the publication of Keller and Bhasin's work [7-15]. The main products of the reaction are C 2 (C2H4 and C2H6), CO and C 0 2 With oxygen or air as co-reactant, this catalytic reaction can occur at temperatures from 600 °C to 800 °C using catalysts such as alkali metals on supports such as BeO, MgO, CaO, ZnO and some rare earth oxides. The highest C 2 yield % (2 x moles of C 2 in the products/moles of CH4 feed xlOO) reported is between 20 % and 30 % with or without diluent gas present [14]. Yields greater than 30 % have not been achieved due to the fact that the hydrocarbon products are rapidly converted to C0 2 at the reaction conditions. The high yield of C0 2 is the main reason why the CH4 oxidative coupling process has not been commercialized. To avoid the formation of C0 2 , 0 2 should be eliminated from the feed gas. A new approach to upgrade CH4 was proposed recently, based on this idea in which a light olefin (C3H6) was homologated (coupled) with CH4 to produce higher hydrocarbons. 2 CH4 + 0 2 -> C2H4 + 2 H 2 0 AG 0 = -36.4 kcal/mole at 727 °C 2 CH4 + 1/2 0 2 -> C2H6 + H 2 0 AG 0 = -14.5 kcal/mole at 727 °C CH4 + C3H6 —> C4H8 + H 2 The advantage of the homologation (coupling) of an olefin with CH4 over the oxidative coupling reaction is that there is no oxygen present in the system and hence no C O 2 is produced. Also, the homologation occurs at much lower temperature (< 350 °C) than the oxidative coupling reaction. Since natural gas also contains light olefins and paraffins, the source of the light olefins for CHVolefin coupling may be the natural gas itself or olefins may be generated from the light paraffins (C2H6 and C3FL3) by oxidative dehydrogenation. The olefins produced in the proposed homologation reaction, after oligomerization, have high octane numbers and could be used in gasoline to replace aromatics which are important components of gasoline. Changes to environmental laws require the replacement of aromatics in gasoline. Ovalles et al. [16] reported the catalytic homologation of C3H6 with CH4 to form C4 hydrocarbon products (hydrocarbons with 4 carbon atoms) on Ni catalysts that were reduced in CH4 before reaction. The C4 selectrvities were 81.4 % for a Ni/Si02 catalyst and 6.2 % for a Ni/Al203 catalyst after 1 hour time-on-stream at a reaction temperature 350 °C. The authors claimed that in comparing the two catalysts (Ni/Al203 and Ni/SiC^), the better performance of Ni/SiC>2 in terms of C4 selectivity, was due to a higher Ni dispersion on the Si02 supported catalyst than on the A1203 supported catalyst. The Ni-CHX generated by the reaction of CH4 with Ni on the catalyst was responsible for the C4 formation. However, no detailed information on catalyst deactivation, hydrocarbon product distribution at different reaction conditions and the role of surface carbon was reported. In addition, the role of the initial catalyst reduction in CH4 rather than in H 2 was not investigated. Further study of this interesting process is warranted and the present study focuses on this topic. During the course of the present study, metal cations supported on zeolites were shown to 4 have the ability to act as strong hydride acceptors. Wang et al. [17] proposed that CH4 can undergo hydride abstraction on the metal cation of a suitable zeolite to produce a methyl carbenium ion. The methylation of benzene by CH4 over zeolite has been reported by He et al. [18]. The toluene and xylenes were formed in the reaction of CHj/benzene over Cu/ZSM-5, Cu/beta and H-beta zeolite catalysts at 400 °C and 5.5 MPa pressure. In this report an isotopic tracer test with 13CH4 showed that the carbon of the methyl substituents added to the aromatic nucleus was predominantly 1 3 C. A direct CH4 conversion to C2H4 and benzene over metal/zeolite has also been reported recently [17]. About 8 % CH4 conversion with 100 % selectivity to benzene was achieved over Mo/HZSM-5 and Zn/HZSM-5 catalysts. Direct methylation of naphthalene with CH, over metal (Pb, Cu, Ni and Si) substituted aluminophosphate molecular sieves has also been reported [19] and the mechanism of this reaction may also involve hydride abstraction. All these reports show that the zeolite supported metal catalysts have the ability to activate CH, at moderate reaction conditions. For this reason, the zeolite supported metal catalysts were included in the latter part of the present study. In the present study not only were the CHVolefin coupling and related reactions investigated, but the effect of different supports and promoters was also determined, in an attempt to increase the understanding of the CFL/olefin coupling reaction. The light olefin used in the olefin/CFL, coupling reaction was primarily C3H6. The desired products were C 4 hydrocarbons synthesized from the reaction between CH, and C3H6, not C3FL5 alone. 1 . 2 Objectives of this Study Since homologation (coupling) of a light olefin with CH, is a relatively new approach to 5 CH, upgrading in the absence of O2, few reports on this topic have been published in the literature. Based primarily on the report by Ovalles et al. [16], several aspects of the homologation of light olefins with CFL, need to be explored further, including: 1. The role of carbon deposition before and after reaction and its relationship to catalyst activity and product selectivity. 2. The significance of reactions other than CH4/C3H6 coupling occurring during the process and a detailed analysis of the product distribution. 3. The effect of different supports and promoters on the CH4/C3FL5 coupling reaction. In view of above, the objectives of the present study are: i. to investigate the CH4/C3H6 coupling reaction over A1203, Si02 and zeolite supported Ni catalysts, to determine product distribution as a function of different reaction conditions such as temperature, residence time and catalyst reduction conditions. ii. to investigate the role of surface carbon deposited on the catalyst in the CH4/C3FL5 coupling reaction and to correlate the results to the catalyst activity. Ultimately, the goal of the present study is to convert as much CFL, as possible to higher hydrocarbons. 6 Chapter 2 Literature Review The primary constituent of natural gas is CH4 and Canada possesses significant reserves of natural gas. How to use this nonrenewable resource more efficiently is of great concern to both industry and research. Many reports have been published on CH4 upgrading. Currently, two large-scale approaches for converting CH4 to gasoline or chemicals are used commercially. One is the methanol-to-gasoline (MTG) route and the other is the Fischer-Tropsch (FT) route. Both start with CH4 steam reforming, which is a high temperature, endothermic and costly process. An approach to convert CH4 to value added chemicals or gasoline without reforming CH4 to synthesis gas, would be very attractive. In this review the focus is on direct CH4 conversion to value added products that can be easily used as a feed stock in the chemical industry. 2.1 Direct CH4 Conversion A review of the different catalytic routes for direct CH4 upgrading was reported by Fox [15]. Three CH4 conversion processes were evaluated: CH, oxidative coupling, CH, partial oxidation and CH, oxyhydrochlorination. The CH, oxidative coupling process, operated at 800 °C and 379 kPa, had 25 % CH, conversion and 78 % selectivity to hydrocarbons. The CH, partial oxidation process produced CH 3OH when operated at 400 °C to 500 °C and 6060 kPa with 5.5 % conversion (per pass) and 80 % selectivity. The CH, oxyhydrochlorination process was operated at 242 °C and 1586 kPa with 25 % conversion. The products were primarily CH3C1 7 and CH2CI2 and selectivity to C O 2 was about 3 %. In this review the capital and operating costs of these three processes were compared with MTG and FT processes. Table 2-1 shows the comparison. Table 2-1. Capital and operating cost comparison, gas-to-Uquid processes (2304 M3/day), in minions of US dollars (1988) [15] Oxidative Partial Oxyhydro- Fixed-Bed Fluid-Bed FT Coupling Oxidation chlorination MTG MTG Process Investment cost 683.5 844.0 596.1 682.6 621.4 646.8 Contingency 136.7 168.8 119.2 0 31.1 64.7 Investment cost + Contingency 820.2 1012.8 715.3 682.6 652.5 711.5 Operating cost/year 74^ 5 75^ 0 74/7 57A 5TI 52.8 When contingency costs are added, none of the three direct CFL, conversion processes is competitive with the indirect MTG or FT processes. Clearly, significant improvement in conversion and selectivity is required for the direct CFL conversion process. Different approaches to CFL upgrading were also discussed in a review by Poirier et al. [7]. In one approach, thermal conversion of CFL, occurs by consecutive radical reactions in the gas phase. The hydrocarbon products are mainly C2H2, C2H4 and CeFL- The reaction temperatures are higher than 1000 °C. The main drawback of this approach is its energy consumption, which makes it expensive. 8 The conversion of CH4 with other co-reactants has also been widely studied. The co-reactants can be either a halogen or O2. In the case of a halogen as co-reactant, the products such as CH 3X have few applications and have to be converted to CH 3OH or higher hydrocarbons. Major disadvantages associated with CH4 halogenation are coke formation, equipment corrosion and the high cost of HX recovery. With 0 2 as co-reactant, CH4 can be partially oxidized to form CH3OH and CH 2 0 at very low CH4 conversion. The formation of CO and C 0 2 increases as the conversion of CH, increases. To increase the hydrocarbon selectivity and extend the catalyst life, it is necessary to decrease the reaction temperature of CH, oxidative coupling. Recently a review on low-temperature CH, coupling was reported by Guczi et al. [21]. This review focused on the activation of C-H bonds of CH, and C-C bond formation between CH X fragments. In the chemisorption of atomic H and C, fragments (CH, CH 2 and CH3) on the Ni (111) surface, the relative bond strengths are CH > CH 2 > H > CH 3 , with corresponding maximum adsorption energies of 3.1, 2.9, 2.7 and 1.7 electron volts (eV). The following CH, chemisorption can occur at temperatures from 30 °C to 300 °C on Ni/Si02 catalysts: CH, + 7Ni -> Ni3C + 4Ni-H The chemisorption of CH, on several other metals including Ru, Pd, Pt, Fe, Co and Cu is also possible. A two step nonoxidative CH, coupling approach was also discussed by Guczi et al. [21]. The reactions were conducted over a Co/Si02 catalyst. The first step in the reaction is the dissociation of CH, at relatively high temperature. The subsequent hydrogenation of the adsorbed surface carbonaceous species at low temperature produced higher hydrocarbons and CH,. The 9 reactions can be represented stoichiometrically as follows: 2CH4 + 6 C 0 ->2Co3C + 4H 2 at360 °C 2Co3C + 3H2 -> C 2 H5 + 6 C 0 at 100 °C Since CH4 chemisorption is favored at high temperature and hydrogenation is favored at low temperature, the two steps of the reaction must be conducted separately. Since olefin homologation with CFL, is a relatively new approach to CFL, upgrading, none of these reviews of direct CFL, conversion discussed the CFVolefin coupling approach. The present study was concerned primarily with homologation of C3FL5 with CFL, and the topics relevant to this study are reviewed in the following sections. 2.2 Homologation of C3H6 with CH4 Although CFL, oxidative coupling has been widely investigated [7-15], only a few studies have been reported on the homologation of olefins with CFL,. Homologation is defined as a reaction in which the final product has one more carbon atom than the starting reactant. In the present study, homologation reactions of the type: CH, + CH2=CH2 -> CH3CH=CH2 + H 2 CH, + CH3CH=CH2 -> CH 3CH 2CH=CH 2 + H 2 are of primary interest. Under certain conditions, the olefin homologue may be hydrogenated according to the reaction: 10 CH 3CH 2CH=CH 2 + H 2 -> CH 3 CH 2 CH 2 CH 3 Loffler et al. [22] investigated the reaction of C3He with CH, using 7.5 % Ni on Si02 as catalyst at 330 °C and 10 atm. Before the reaction, the catalyst was treated in CH, at high temperature. A CH4/C3H6 gas mixture which contained 6 % C3H6, was passed over the catalyst and 53 % d + C 2 \ 26 % C3Hg, 8.5 % C4H10 and 13 % C 5 + C 6 were obtained in the final products. They claimed that CH, did not react with olefins. The CH, apparently formed active CHX-Ni species during reduction of the catalyst and these combined with the olefin. Information on the catalyst reduction conditions and catalyst deactivation were not given. They also reported the homologation of benzene and cyclopentene with CH, to produce some toluene and methylcyclopentane on the Si02 supported nickel catalyst. Subsequently, Ovalles et al. [16] reported the homologation of C3H6 with CH, on Ni catalysts and compared the surface characteristics of the catalysts with their activity. Before reaction, the calcined catalysts were treated in CH, at 600 °C for 8 hours. It was claimed that the treatment of the catalysts in CH, was critical and resulted in reduction of the NiO, present on the calcined catalyst, to Ni. The results also showed that the CH4/C3H6 coupling reaction with CH, treated catalyst gave much higher C 4 selectivity than H 2 treated catalyst. Comparing Si02 supported and A1203 supported Ni catalysts, the Si02 supported Ni catalyst showed higher C 4 selectivity than the Ni/Al 20 3 catalyst. The best result obtained at 4.7 % C3H6 conversion was 81.4 % C 4 selectivity at 350 °C and the reaction was conducted with a 9 % C3H<; in CH, gas mixture on a 7 wt % Ni/Si02 catalyst. By comparison, the C3H6 conversion and the C 4 selectivity were 9.2 % and 6.2 %, respectively with * Note: C* means all hydrocarbons with x C atoms in the hydrocarbon molecule 11 the A1203 supported Ni catalyst. Results from X-ray Photoelectron Spectroscopy (XPS) studies showed that there were more carbonaceous species present on the SiC>2 supported catalyst surface than on the AI2O3 supported catalyst. The dispersion of Ni on the SiC>2 support was higher than on the AI2O3 support which resulted in more carbonaceous species on the Ni/Si02 catalyst than on the Ni/Al 20 3 catalyst. As a result, the C 4 selectivity was higher with the Ni/Si02 catalyst than with the Ni/Al203 catalyst. Ovalles et al.[16] claimed that CH4 was incorporated into C 4 products during the reaction. There was no distribution of the C4 products reported in this study. It can be seen from the above two studies that the treatment of catalysts in CH4 is very important in CH4/C3H6 coupling. The treatment in CFL not only reduces the NiO to Ni, but also provides carbonaceous species on the surface of the catalysts. These carbonaceous species couple with olefins to form higher olefins. Koerts and Van Santen [23] used a different technique to study the reaction sequence of olefin homologation with CH4. First, CFL, was adsorbed at 450 °C on a previously reduced Ru or Co catalyst supported on SiC>2. After cooling the catalyst in He, light olefin (C2FL or C3FL5) was co-adsorbed on the pretreated catalyst at 50 °C. Subsequently the system was exposed to H 2 at the same temperature. The product gases were analyzed by a gas chromatograph (GC) and the results showed that incorporation of pre-adsorbed CH4 into the olefin produced higher olefins, and more of the olefins were adsorbed on a CFL treated catalyst than otherwise. Tests were conducted for homologation of C3FL5, C2H4 and C 2 H 2 with 1 3 C labeled CH, over Si02 supported Ru and Co catalysts. Results showed that CH, was incorporated in the homologation products. Most recently Vasant et al. [24] reported a study on CH, conversion to higher 12 hydrocarbons. The approach was to couple CFL, to aromatics with the addition of light olefins, such as C2H4, C3FL5 and C ^ , at 400 °C to 600 °C over H-GaAlMFI zeolite catalysts (H-gaUoaluminosilicate). The catalysts showed no activity for CFL, conversion at < 600 °C in the absence of olefins. However, the addition of olefins to the feed increased the CFL, conversion from zero to 45 %, depending on the additives, the olefin/CH, molar ratio in the feed and the temperature. Higher CH, conversion occurred at higher olefin/CH, ratio and higher temperature. The selectivity of aromatics was higher than 90 % in all cases. According to the authors, the following steps were involved in the reaction: C n H 2 n + H + - — • (CnH2n+l) H - CH3 H — CH3 -Ga-O- -Ga — O -+ H - - - - C H 3 + (CnH2n+l) + : : *" CnH2n+2 + CH3 -Ga — O -C H 3 + — • C H 2 + H + 2CH2 — • C2H4 At the reaction conditions, the paraffins underwent fast dehydrogenation to regenerate olefins and the C2H4 and higher olefins were oligomerized and then dehydrocyclized to aromatics over the bifunctional zeolite catalysts: 13 C n H 2 n + 2 H % G a - O x i d e ^ C n H 2 n + H 2 C 2 to C 4 olefins -t£*. C 6 to C 1 0 olefins H%Ga-Q)dde> aromatics Since the CH4 conversion reached 45 % in this study, it is a most promising approach for the upgrading of CH4. Studies on the mechanism of the catalytic homologation of olefin with CH4 have also been reported [16,23,25]. The typically proposed mechanism [16] for the homologation on supported Ni catalyst involves: 1. The Ni on the surface reacts with CFL, to form Ni-CH X species, where x = 0, 1, 2 or 3. 2. Olefins adsorb on the activated catalyst. 3. Homologation takes place between Ni-CH X and adsorbed olefins to release higher hydrocarbons and regenerate Ni for the next cycle. Figure 2-1 shows the proposed homologation mechanism [16]. Van Santen et al. [23] suggested a similar mechanism However, in the latter case the reactions were conducted at two temperatures to improve reaction thermodynamics. Since CH4 adsorption requires high temperature and the coupling reaction is favored at low temperature, CH, was first adsorbed on the Ni site at high temperature and then the Ni-CHX reacted with C3H6 at lower temperature to form C4 and release Ni for the next cycle. There is some controversy between the studies reported by Loftier et al. [22] and Ovalles et al. [16] concerning whether CH, was incorporated into the C4 product during the coupling 14 reaction or not. Loftier et al. [22] claimed that CH4 did not react with C3H6 in the reaction and that the added CH X came from the initial pretreatment of the catalyst. Ovalles et al.[16] claimed that CH4 was incorporated during the reaction. Figure 2-1. The proposed mechanism for the homologation of C3FL5 with CFL, using Ni-supported catalyst. In CHX, x = 0, 1, 2 or 3. Other studies, including the homologation of C3FL5 with CFL, in the presence of oxygen [26,27] and the homologation of olefin using solid or liquid superacid catalysts [28,29], have also been reported. Sodesawa et al. [19] reported that the oxidative methylation of C3FL5 with CFL, over 3 wt % of Na20/La203 catalyst gave a maximum 10.1% C4 yield with 8.7% CFL, conversion during a 2 hour time-on-stream Commonly used catalysts are SbF5/HS03, HF-TaF5, TaF 5-AlF 3 andHF-SbF5 Since the olefin homologation with CFL, is a relatively new approach to CFL, upgrading, there is a lack of information in the literature on the catalyst deactivation during reaction, the product distribution and the possible reaction paths by which different products are formed. The 15 most important issue that remains unresolved is whether the incorporation of CH4 into C3FL5 to form C4 occurs during the reaction. More research is needed to fully understand this new approach to CH4 upgrading. 2.3 Chemisorption and Carbonaceous Species on Metals Heterogeneous catalytic reactions involve reactions among at least one species adsorbed on the surface of the catalyst. A better understanding of the mechanism of chemisorption on the catalyst may lead to the development of improved catalysts. 2.3.1 Chemisorption of C H X (x= 1,2 and 3) As discussed in the Section 2.2, higher hydrocarbons are formed through the coupling of an olefin with CHX, both adsorbed on the metal sites of the catalyst. Therefore, the formation of CH X species is an important step toward the formation of C 4in the CH4/C3H6 coupling reaction. CH4 is adsorbed dissociatively on metal catalysts. The C-H bond in CH4 is stronger (C-H bond energy = 104 kcal/mol) than in other hydrocarbons such as ethane (C-H bond energy = 98 kcal/mol). Dissociative adsorption is difficult and only occurs at high temperature on metals such as Ni, Fe, Co and Ru [30-35]. The dissociative adsorption may be represented as follows: CH4 + (5-x)M -» M-CH X + (4-x)(M-H) (x=l, 2 and 3) CH X has also been identified as an important surface species present in the FT synthesis. 16 Hydrogenation of CO in FT synthesis occurs on metals such as Ni, Fe and Co [31-33] by the following mechanism: CO + 2M -> M-C + M-0 H 2 + 2M -» 2M-H M-C + xM-H -» M-CFfx + xM 2M-CHX -> M-CH x.iCH 3 + M or M-CH X + M-H —> CFLj + 2M (x=l, 2 and 3) The CH X species may be hydrogenated to CFL, or react with another M-CH X to form M-CH x.iCH 3 in another catalytic reaction cycle. 2.3.2 Chemisorption of Olefins Since C3FL5 is one of the reactants in CH4/C3FL3 coupling , its chemisorption is a major concern. Generally, olefins can be adsorbed on many metals and metal oxides and the chemisorption usually follows the sequence: H H I I R - C H = C H - R + 2 M « R - C - C - R I I M M 17 Olefins are readily adsorbed on transition metals [36]. Olefins and transition metals form bonds due to the fact that electrons from olefins are transferred to d orbitals of the metals via the O" bond, and back transferred to olefins through the K bond as indicated in Figure 2-2 [37]. The bond strength can be measured by measuring the heat of adsorption. For C2FL4 adsorption on metals, the order of decreasing heats of adsorption is: Ta>W>Cr>Fe>Ni>Ru>Pt Figure 2-2 Olefin adsorption on transition metals 2.3.3 Carbonaceous Species on Metals Chemisorbed carbon plays an important role as an intermediate in many catalytic reactions including the production of hydrocarbons in FT Synthesis. The deactivation of catalysts is also 18 caused by carbon deposits (coking). The difference between the active intermediate and inactive coking is that the active carbon is in the form of carbide and the inactive form is graphitic [38-40]. By using Auger Spectroscopy, the carbidic and graphitic carbon can be identified. The carbidic carbon consists of nearly non-interacting carbon atoms while the graphitic carbon resembles the carbon in graphite. The relationship between them can be expressed as follows [41]: M-C-> M + C Note: carbon in M - C is carbidic carbon and C is graphitic carbon To avoid the formation of graphitic carbon from carbidic carbon, two methods can be effective [41]: (a) preventing the formation of a metastable metal carbide. Formation of a metastable metal carbide can be avoided by alloying of the active metal with a metal not forming a carbide of significant stability. (b) preventing the formation of relatively large carbided metal particles. This will call for relatively small metal particles strongly adhering to the support. 2.4 Related Chemical Reactions From the limited number of literature publications and the results of the experiments discussed later, it can be concluded that the coupling reaction is accompanied by several other reactions such as the hydrogenation of olefins to paraffins, the hydrogenolysis of C3 to methane and C2 and the dehydrogenation of methane on the surface of metals to form C H X and C. 19 Furthermore, the coupling mechanism is related to the FT synthesis. It is necessary to review the mechanisms of these reactions and catalysts, since they are relevant to the present study. 2.4.1 C3Iis Homologation with CH4 The homologation of C3H5 with CH4 is the target reaction of the present study. The homologation of C3H5 with CH4 is thought to occur via adsorbed carbonaceous species and C3H6 according to reactions of the type [42]: CH2 CHCH2CH3 II + I »- CH2=CHCH2CH3 + 2M M M Since C3H6 is readily adsorbed on a metal surface, the adsorption of the CH X (x=l,2 or 3) carbonaceous species is the key step in the reaction. According to the homologation mechanism described by Eloy and Leconte [42], the product of the reaction is 1-butene. In the present work, the CH2=M species may arise from CH, decomposition that occurs during the reduction of the calcined catalyst [16,22] or from CH4 and C3FL5 decornposition that may occur during the coupling reaction. 2.4.2 Fischer-Tropsch Synthesis The Fischer-Tropsch (FT) synthesis involves CH X formation and chain growth while the CH4/C3FL5 coupling reaction also deals with CH X formation and chain growth. The similarity of the FT reaction can provide some clue in the selection of catalysts for the CH4/C3H5 reaction. Since Fischer and Tropsch reported their work using alkalized Fe catalyst to produce 20 higher hydrocarbons from CO and H 2 at pressures from 1 to 50 atm, much research has been conducted on this topic. Several different mechanisms have been proposed [31,43,44], most of which involve CH X species adsorbed on the catalyst surface. One of the most common mechanisms is: Initiation: a. C O + M — - M-CO -tM*. M-C + M-0 b. H 2 + M M-H2 or 2M + H 2 —* - 2M-H c. M-C + M-H2 —>- M-CH 2 and M-C+M-H —>- M-CH M-CH 3 and M-0 + M-H 2 — 2 M + H 2 0 or 2 M-H + M-0 - » • H 2 0 + 3 M Propagation d. M-CH 2 + M-CH 2 — * • M-CHCH 3 M-CHCH 2R + M-CH 2 — - M-CHCH 2 CH 2 R Branching and Termination R R e. M-CH-R + M-CH3 - * • M-C-CH3 »Ji>. M-CH2-C-CH3 H H f. M-CH-CH2-R RCH=CH2 or RCH2CH3 The catalysts used to carry out the FT synthesis include Fe, Co, Ni and Ru. Typical performances of different catalysts are shown in Table 2-2 [45]. 21 It is well known that the alkali metals, especially K, are key promoters of Fe and Ni catalysts. Alkali addition results in an increased selectivity toward higher hydrocarbons in FT synthesis, improvement in the resistance to poisoning and a decreased rate of hydrogenolysis. The alkali effects are directly related to the electron-donor character of the promoter. Studies have shown that the metal-adsorbate bond is strengthened making dissociative adsorption more probable for the case of CO and N2 molecules [46]. K is usually added as the carbonate. The chloride, bromide and sulfate salts result in catalysts with low activity. Table 2-2. The performances of different catalysts in the FT synthesis [45] Catalyst Fe Co Ni Temp. (°C ) 371 231 205 Pressure (aim.) 35 7.8 1 H 2 /CO (mol.) 1 3 3 C\ selectivity % 36 53 81 C2-C5 selectivity % 64 47 19 2.4.3 C3H5 Decomposition and the Formation of Carbonaceous Species C3II3 decomposition can occur on Group VIE metal catalysts according to the following mechanism which explains both the formation of C2H4 and carbon deposition that occurs on the catalyst surface [47]. The H 2 produced by this mechanism may subsequently result in the hydrogenation of C2H4, C3FL3 and C4 olefins. 22 H2C—CHCFL I I H 2C=CHCH 3 + 2M ^ M M H 2 C - C H C H 3 H 2 C C H C H 3 I I ^ I I M M M + M H 2 C CHCH3 I I ^ H 2 + C + H 2 C=CH 2 +2M M + M 2 2 2 Since this reaction consumes C3H6 and also causes catalyst deactivation by carbon deposition, it is an undesired side reaction of the CH4/C3H6 coupling process. 2.4.4 Metathesis Metathesis is another side reaction that may occur in the homologation of C3FL5 with CFL,. Metathesis is catalyzed by transition metals and involves the interchange of alkylidene units between olefins. R,CH=CHR2 RiCH CHR 2 R,CH=CHR2 RiCH CHR 2 C3FL5 may undergo metathesis (disproportionation) to form ethylene and butenes accordingly: CH 2=CHCH 3 C H 2 CHCH 3 + II + II CH 2=CHCH 3 C H 2 CHCH 3 23 Note that metathesis of C3FL5 produces an equal number of moles of C2H4 and 2-C4H3. From the point of view of CH4/C3FL5 homologation, C3FL5 metathesis is an undesired reaction. 2.4.5 Dimerization and Cracking Dimerization and cracking of C3FL5 are also undesired reactions that may occur in the homologation of C3FL5 with CFL,. These reactions are catalyzed by acid catalysts. In both dimerization and cracking processes, carbenium ions are intermediates in reactions catalyzed by acidic surfaces that lead to the formation and breaking of carbon-carbon (C-C) bonds. The dimerization of olefins catalyzed by acids is illustrated by the following [48]: H 2C=CHCH 3 + HX ===== CH 3 —CHCH 3 + X" CH 3 —CHCH 3 + H 2C=CHCH 3 CH 3 —CHCH 3 CH 2 CHCH 3 The dimerization of olefins can continue until coke is formed on the catalyst surface. Cracking is essentially the reverse of dimerization, occurring at the bond located P to the carbon atom bearing the positive charge. Cracking of a linear secondary carbenium ion results in the formation of a primary carbenium ion: R C H 2 C H C H 2 — C H 2 C H 2 R ' ^ RCH2CH=CH2 + CH 2CH 2R' The primary ion can undergo a rapid hydrogen shift to give a more stable secondary carbenium ion: 24 C H 2 C H 2 R CH 3 CHR This continued pattern of cracking of the straight chain at the P position leads to the formation of C3FL5 in high yield; C2H4 is not formed by this mechanism 2.4.6 Hydrogenation of Olefins Catalytic hydrogenation is one of the most important reactions in synthetic organic chemistry. Olefin hydrogenation is easily carried out over a variety of catalysts. With active catalysts the reaction is likely to be diffusion limited, since hydrogen can be consumed faster than it can be supplied to the catalyst surface. The mechanisms of olefin hydrogenation are complex and many details are still unknown. One of the mechanisms that provides a useful way of accounting for all aspects of olefin hydrogenation was proposed by Horiuti and Polanyi [49] as follows: 2 M + H 2 o M - H + M - H H H b - CH 2 - C H = C H - CH 2 - +2 M <=> - C H 2 - C - C - C H 2 -I I M M H H H I I I c - C H 2 - C - C - C H 2 - + H » - C H 2 - C - C H 2 - C H 2 - + 2 M I I I I M M M M H I d - C H 2 - C — C H 2 - C H - + H o - C H 2 - C H 2 - C H 2 - C H 2 - +2 M I I M M 25 The order of decreasing activities of catalysts used in olefin hydrogenation is [44]: Pt = Pd > Ni > Fe + Co > Cu Since Ni, used in the present study, is a good hydrogenation catalyst, the undesired hydrogenation of C3H5 may be expected to occur in the CH4/C3H6 coupling process, where H 2 is generated as a consequence of C3H6 decomposition. However, some compounds can modify, inhibit or poison the hydrogenation activity [32]: a) . Carbon monoxide is a well known activity moderator, which decreases the olefin hydrogenation activity of Pd and Pt catalysts . b) . Nitrogen compounds are also activity moderators but are mainly used for selective hydrogenation. c) . Sulfur compounds are poisons for olefin hydrogenation. d) . As, Hg, Pb and Si are all poisons for the hydrogenation of olefins. 2.4.7 Hydrogenolysis Catalytic hydrogenolysis is the cleavage of organic molecules into fragments by hydrogen in the presence of a catalyst [32]. In the present study the desired reaction is coupling of lower hydrocarbons to form higher hydrocarbons, therefore hydrogenolysis is undesired. Mechanisms of hydrogenolysis of 1-butene were proposed as follows [50]: Mechanism of hydrogenolysis: 26 1 CH2=CHCH2CH3 + M-H M-CH 2CH 2CH 2CH 3 +M M = CH 2 + M-CH 2CH 2CH 3 +H CH 4 + C 3 H 6 + 2 M H 2 CH2=CHCH2CH3 + M-H CH 3 -C- CH 2 CH 3 +M M M=CHCH3 + M-CH 2CH 3 - H — 2 C 2 H 4 +2 M From the mechanism above it can be seen that hydrogenation of the olefin precedes the cleavage of the C-C bond. The catalysts usually used in hydrogenolysis are Ru, Pt, Ni, Co and Fe. The addition of K to Fe results in a sharp decrease of the rate of ethane hydrogenolysis. This is also true for the K-Ni catalysts. The rate of benzene hydrogenation is decreased by a factor of 3.6 when K is added to Ni catalyst. The dissociative adsorption of ethylene on Ni is also depressed by the addition of K [51]. Hydrogenolysis reactions require high temperature and strong bonding of reactants to the catalyst and are therefore difficult to accomplish when compared with hydrogenation reactions. 2.5 Summary of Literature Review One of the possible processes for the conversion of CH, to more useful chemicals is CFL/olefin coupling. Up to now very few studies related to this approach have been reported in the literature and some studies have reported [16,22] opposite conclusions with respect to the involvement of CH, in the formation of C 4 products from CFL, and C3H6. In addition, it has been 27 claimed that the treatment of catalyst in CHj before reaction is a necessary step in the CH4/C3H6 coupling reaction. The lack of information on catalyst deactivation with time-on-stream, the role of catalyst treatment in CH4 before reaction, the product distribution and how different products are formed during CH4/C3H6 coupling, points to the necessity for further investigation. Since Ni catalysts are the most studied catalyst for the CH4/C3H6 coupling reaction, in CH4 activation studies [16,22,23,25,30,34,35,52], and has also been studied in FT process [31,43,44,53] which has a similarity with the CHV C3H5 coupling process in CH X formation, Ni was chosen as the catalyst in the present study. According to Ovalles et al.[16], Ni must be highly dispersed to increase C4 selectivity and Ni dispersions are strongly dependent on the support. Hence the effect of support on CtLJ C3FL5 coupling was an important issue in the present study. Zeolite supported metal catalysts have been reported to have the ability to activate CH4 by forming methyl carbenium ions through hydride abstraction [17,18]. Hence zeolites were also included as supports in the present study. To increase C4 selectivity, the reduction of side reactions was important. For the reduction of hydrogenation and hydrogenolysis of olefins, catalyst promoters such as K may be effective [51]. Although catalyst deactivation during CH4/C3H6 coupling reaction has not been reported in the literature, catalyst deactivation can be anticipated because the adsorbed species such as M-CH X and M-C can be converted to inactive carbon. 28 Chapter 3 Experimental Procedures The most important criterion for a successful catalyst is its performance in the desired catalytic reaction, usually measured in terms of catalyst activity and product selectivity. Catalyst characterization is necessary to develop an understanding of the mechanisms of the catalytic reaction and hence improve the catalyst by modification of its formulation. The experimental procedures followed in the present study emphasize these aspects and will be discussed in the following sections: • The experimental apparatus • Catalyst preparation • Catalyst characterization. • The catalytic reaction and Temperature Programmed Surface Reaction (TPSR). 3.1 The Experimental Apparatus The experimental apparatus used to carry out the catalytic reaction is shown in Figure 3-1. The catalyst, of size between 14 and 24 mesh (average diameter 0.7 mm to 1.2mm) was loaded into the fixed-bed micro-reactor (R). The reactor consisted of a quartz tube of 50 cm length and 0.7 cm diameter. The reactor and catalyst dimensions were chosen to minimize heat and mass transfer effects [54]. Using Mears criterion [73], the measured reaction rate was shown to be less than 15 % of the external mass transfer rate. Similarly the Thiele modulus [73] was estimated at < 29 0.01, indicating no internal mass transfer limiting effect. The reactor was placed in a tube furnace (Lindberg/Blue M Model 55031, Watertown WI) and the reaction temperature was controlled with a programmable temperature controller (Omega CN-2010, Cole-Palmer, LL) (TC) connected to the reactor heater. The feed gas flowrate was controlled by a mass flow controller (Brooks/Model 5878, Hatfield, PA) (MF). The product gases passed through an automatic gas sampling valve before venting through a bubble flow meter (B), the latter used to measure the reactor exit gas flowrate. The product gases were analyzed using a sampling valve and a gas chromatograph (GC) (Shimadzu GC-14A, Mandel Scientific, Toronto, ON). A Porapak Q column (6m x 2.5 mm) with a suitable temperature program and a flame ionization detector (FID) were used to separate and quantify the components of the exit gases. M F J TC VENT B ( O O P o ) a • G C • • 0 PRINTER :1 C2 C3 C4 Figure 3-1. Experimental Apparatus for CH4/C3H6 Coupling 30 The temperature-programmed-surface-reaction (TPSR) was conducted to determine the catalyst surface carbon reactivity. The apparatus described for the activity measurement was also used for TPSR testing. Pure H 2 was passed over the catalyst after treatment in CH4 or after reaction of CH4/C3H6 coupling, and the furnace temperature was increased linearly at 60 °C/min, following the procedure of McCarty and Wise [55]. The produced gases passed directly to the detectors of the gas chromatograph (GC), in which the produced gases were monitored continuously using a thermal conductivity detector (TCD) and a flame ionization detector (FID), connected in series. Catalyst characterization was achieved using a number of techniques. Temperature-programmed-reduction (TPR) was used to determine the reduction temperature of the catalyst metal oxide precursor in CFL. and temperature-programmed-desorption (TPD) was used to determine the catalyst metal dispersion. The apparatus shown in Figure 3-1 was also used for the TPR and TPD testing. In addition, a mass spectrometer (Spectramass DAQ 100/OXM) (MS) was used to continuously analyze the composition of the exit gas. Catalyst surface area measurements were made using the BET method [56] and X-ray diffraction (XRD) was used for bulk phase analysis. 3.2 Catalyst Preparation Since C3FL5 homologation with CH4 is conducted over a catalyst, it was necessary to prepare and characterize the catalysts for use in the study. The most widely used techniques for laboratory preparation of supported metal catalysts are impregnation (incipient wetness 31 impregnation), co-precipitation and ion exchange methods. Two methods were used in the present study for the catalyst preparation: incipient wetness impregnation and ion exchange. To illustrate the procedures for catalyst preparation, a detailed description of the Ni/K/Al203 and Ni/Na-Y catalyst preparation procedure is given. The preparation of the Ni/Si02 catalysts used the same procedure as the Ni/K/Al 20 3 catalysts. The preparation of the Ni/SAPO-5 catalysts used the same procedure as the Ni/K7A1203 catalysts when prepared by incipient wetness impregnation and used the same procedure as the Ni/Na-Y catalysts when prepared by ion exchange . 3.2.1 Preparation of Ni/K/Al 20 3 (Ni 7 wt%, K 1 wt%, A1203 92 wt%) by Incipient Wetness Impregnation For the preparation of Ni/K/Al 20 3 (Ni 7 wt%, K 1 wt%, A1203 92 wt%), the y-Al 20 3 (F-200-Al2O3, Aldrich Chem. Company) support, after crushing to 14-24 mesh size (particle diameter from 1.18 mm to 0.71 mm), was placed in a furnace at 500 °C for lh to remove water and other adsorbed gases. The Al 20 3had a pore volume of about 0.5 cm3/gram A solution of Ni(N03)2 (Ni(N03)2.6H20, certified A C S , Fisher Chemical), containing sufficient Ni to give 7 wt % on the catalyst, was added dropwise to the A1203 while stirring with a glass rod. The wet catalyst was left over night at room temperature and then placed in a furnace at 120 °C for 3 hours to dry, followed by calcination at 450 °C for 1 hour. K was added to the calcined Ni/Al 20 3 by addition of the appropriate volume of KC1 (KCL Fisher Chemical, certified A C S ) solution. The K doped catalyst was dried and calcined as before. After the catalyst cooled to room temperature it was stored in a sealed bottle and labeled, ready 32 for use. For the preparation of Ni/P/Al203 catalysts with 7 wt % Ni and 1, 3 or 7 wt % P, the calcined Ni/Al203 precursor was prepared as before. A solution of H 3 P0 4 (85 % o-phosphoric acid, Fisher Scientific, A.C.S) was added to the Ni/Al 20 3 dropwise with stirring to form 1, 3 or 7 wt % P loading in the finished Ni/P/Al203 catalysts. The wet catalyst was left over night at room temperature and then placed in a furnace at 120 °C for 3 hours to dry, followed by calcination at 450 °C for 1 hour. For the preparation of Ni/Si02 with 1, 7 and 14 wt % Ni and Ni/K/Si02 with 7 wt % Ni and 1 wt % K, a solution of Ni(N03)2 (Ni(N03)2.6H20, certified A.C.S, Fisher Chemical) and Si02 (Silica grade 951, Aldrich Chem Company) were used and the same preparation procedure as for the Ni/Al 20 3 and Ni/K7A1203 catalysts was followed. 3.2.2 Preparation of Ni/Na-Yby Ion Exchange For the preparation of Ni/Na-Yby ion exchange, the Na-Y zeolite (Aldrich) was placed in a beaker containing 0.2 N Ni(N03)2 solution. The solution was stirred with a magnetic rod at room temperature for 18 hours. The procedure was repeated with a fresh 0.2 N Ni(N03)2 solution. The catalyst was then washed with deionized water and the wet catalyst was left over night at room temperature. The following day, the catalyst was placed in a furnace to dry at 120 °C for 3 hours and calcined at 500 °C for 6 hours. After calcination, the catalyst was mixed with 5 % bentonite (purified grade, Fisher Chemical) as binder and deionized water, and the mix was shaped into pellets with an extruder. The wet catalyst pellets were left overnight at room temperature and then calcined at 450 °C for 1 hour. The pellets were cooled to room temperature 33 and broken into 14 to 24 mesh particles (catalyst diameter from 1.2 mm to 0.7 mm). The finished catalyst was placed into a bottle and labeled, ready for use. 3.3 Catalyst Characterization There are many techniques available for catalyst characterization. In the present study some of the most common techniques including BET (for catalyst surface area and pore volume measurement), TPR (temperature programmed reduction for determination of metal oxide reduction temperature), TPD (temperature programmed desorption for the measurement of metal dispersion on the catalyst) and XRD (X-ray diffraction for measurement of catalyst bulk properties) were used to characterize the catalysts. 3.3.1 Surface Area In a heterogeneous catalytic process, most reactions occur on the surface of the catalyst. Hence, it is necessary to know the surface area available for catalysis per unit mass of catalyst. The BET method, developed by Brunauer, Emmett and Teller [56], is the most common method used in catalytic studies to measure surface areas. Descriptions and evaluations are given by Emmett [57,58]. The mathematical expression of the BET equation is as follows: P = 1 (C- l )P V(P0-P)~VmC+ VmCP0 C = e(q'~qi)/RT 34 Where V = volume of gas adsorbed at pressure P (m3) V m = volume of gas adsorbed in monolayer, same units as V (m 3) P 0 = saturation pressure of adsorbate gas at the experimental temperature (Pa) C = a constant related exponentially to the heats of adsorption and liquefaction of the gas qi = heat of adsorption on the first layer (J) = heat of liquefaction of adsorbed gas on all other layers (J) R = the gas constant (JK^mof 1 ) T = temperature (K) Many adsorption data show very good agreement with the BET equation over values of the relative pressure P/Po between 0.05 and 0.3. With microporous substances such as zeolites, the linear region on a BET plot occurs at much lower value of P/Po, typically around 0.01 or less. After a sample is heated under vacuum to remove moisture and gas, it is cooled to liquid N 2 temperature, in the case of using N 2 as adsorbate. A known quantity of N 2 is admitted to a constant volume sample holder and allowed to equilibrate. From the equilibrate pressure and PVT relationship, the amount of N 2 adsorbed is calculated. The procedure is repeated, yielding a series of values of the volume adsorbed corresponding to a set of increasing values of the equilibrium pressure. Using the BET equation, a graph of P/V(P0-P) versus P/P0 can be drawn and V m can be determined from the slope and intercept. Once V m is known the surface area can be calculated. When the constant C is large enough, e.g., greater than 50, as it is with N 2 adsorption, the intercept [l/(VmC)] is usually small relative to the slope[(C-l)/VmC)]. Hence, a straight line can be drawn connecting the origin and one point obtained at a P/Po value of about 0.2 to 0.3 to obtain the slope of the BET equation. This method, requiring only one datum point, is called a single point surface area measurement. 35 In the present study the examination of catalyst surface area was performed by N 2 adsorption at - 196 °C using a Flowsorb 2300 (Micromeritics, Norcross, GA). Before the measurement, the catalyst was heated at 120 °C to remove any moisture and then was degassed under vacuum to remove any adsorbed gases for 0.5 to 24 hours depending on the sample. A single point BET measurement was used to obtain catalyst surface area. 3.3.2 X-ray Diffraction (XRD) XRD can be used to obtain information on the structure and composition of crystalline materials [59,60,61]. The minimum limit of detection is about 5 percent for compounds and 1 percent for elements. Common compounds can be identified from standard reference patterns. In the present study, X-ray diffraction [Diffraktometer, D5000, Siemens. Power setting: 40kV and 30 mA. Radiation source: CuKa (X=1.5406 )] was used to measure the NiO reducibility by comparing the X-ray diffraction spectra of a calcined Ni/K/Al203 and a CH, reduced Ni/K7A1203 catalyst with standard reference patterns. 3.3.3 Temp erature Programmed Reduction (TPR) TPR is a method for detenrnning the reducibility of a catalyst and many TPR studies of catalysts have been reported [62-66]. In TPR, the temperature increases linearly with time in the presence of a reducing gas. The unreduced metal oxide, placed in the reactor, undergoes reduction at specific temperatures that are characteristic of the metal oxide and this is measured by a change in the reactor exit gas composition. Hence TPR measures the reduction profile of the catalyst precursor. In the present study, CH, was used as reducing agent. The calcined catalyst 36 was loaded into the reactor and CH4 was passed through the reactor at 20 cm3/rnin. while the temperature was increased linearly at 30 °C/min. from 50 °C to 900 °C using the TC (Figure 3-1). When reduction of metal oxide occurred, the gas composition change was followed using a TCD detector of the GC whereas the MS quantified the change in gas composition. A 386 DX computer (PC) was used to record the data from the GC and MS and the data were tabulated and plotted to determine the metal oxide reduction temperature. 3.3.4 Temperature Programmed Desorption (TPD) The TPD technique is a very useful procedure for estimating the dispersion of metals of supported catalysts. It is very important to disperse the active metals on the support as much as possible during the preparation of supported catalysts. The TPD process includes chemisorption, purge of physically adsorbed gas species and chemidesorption at a programmed temperature. The commonly used probe gases are O 2 , H 2 , CO and NO. Many reports using TPD for studies of catalyst reduction are available [53,62,67-69]. In the present study, H 2 was used as probe gas. In the TPD measurement, the catalyst containing metal oxide was loaded in the reactor and pure H 2 was passed through the reactor at 20 cm3/min. Then, the catalyst was reduced in H 2 at 450 °C for 16 hours. After reduction, the catalyst was purged with He at 450 °C for 1 hour to ensure no H 2 remained on the catalyst. Pure H 2 was then passed through the catalyst at 110 °C for 1 hour to adsorb H 2 on the metal. The reactor was cooled to room temperature and purged in He at room temperature to remove physically adsorbed H 2 . He then flowed through the reactor as the temperature increased from 50 °C to 450 °C at a rate of 20 °C/min. The TCD of the GC recorded the signal for H 2 from which 37 the total amount of H 2 released from the catalyst could be calculated. The Ni dispersion was calculated using the following formula: Ni dispersion = (mmole H 2 x 2 /mmole Ni) x 100 % assuming that one Ni atom adsorbs one H atom In the present study the catalyst activity was measured on catalysts that were reduced in CH4. Hence, metal dispersion measured for the H 2 reduced catalysts may not reflect the actual Ni dispersion of the CH4 reduced catalysts. CH4 reduced catalysts could not be used to measure metal dispersion by H 2 TPD since the surface carbon formed during catalyst reduction in CH4 could react with H 2 and therefore the measured amount of adsorbed H 2 would be in error. An additional error may arise from the fact that the catalyst reduction in H 2 was conducted at 450 °C whereas the catalyst reduction in CH, was conducted at 600 °C. Higher reduction temperature may cause metal sintering, especially when the metal dispersion is high. Since Ni dispersions on the Ni/Al 20 3 catalyst and promoted Ni/Al203 catalysts, discussed in Chapter 4, were low (from 3.6 % to 7.0 %) i.e. the particles were larger, the trends in metal dispersion measured using H 2 reduced catalysts should reflect the trends of the metal dispersion of the catalysts reduced in CH,. 3.4 The Catalytic Reaction and TPSR The emphasis of the present study was on determining the activity of different catalysts for the CH4/C3H6 coupling reaction. The apparatus used for these measurements was shown in Figure 3-1. Typically 1.1 gram of catalyst was placed in the reactor. Before reaction, the catalyst was 38 reduced in CH4 at a flow rate of 20 cm3/min. as the temperature was programmed to increase from 50 °C to 600 °C over 30 minutes and was then held at 600 °C for 1 hour. The temperature of the reactor was subsequently decreased to the desired temperature and the reactants (90 mol % CH4 and 10 mol % C3H6) with a flowrate of 20 cm3/min were fed to the reactor. The produced gases were sampled using an automatic sampling valve and analyzed by GC. A bubble flow meter was connected to the exit gas line, with which the flow rate of exit gases was measured. After the GC was calibrated, the percentage of different gas components was calculated. The final reaction results were calculated as in Table 3-1. The definition of product selectivity, which follows that used by previous investigators, Ovalles etc. [16], was primarily aimed at monitoring the following reactions: Homologation: C3H5 + CH, -» C4H8 Hydrogenation: C3H5 + H 2 —> C3H8 Hydrogenolysis: C3H5 + H 2 -> C 2 H, + CH, However, it became apparent during the course of the study that propylene decomposition also occurred, which may be represented stoichiometrically as: CaHe -» 3C + 3H2 The extent of this reaction could not be accurately monitored because the exact stoichiometry of all the reactions occurring in the system was not known. However, as will be discussed in later chapters, although decomposition of propylene was an important step in the overall process, the trends in selecthities as defined in Table 3.1 were not affected by the carbon deposition. This was 39 due to the relative low level of the carbon deposition. The carbon balance data of some tests were included in the Appendix 3. The temperature programmed surface reaction (TPSR) was used to examine the reactivity of the surface carbon deposited during CH4 treatment (reduction) and after reaction with CH4/C3H6. The procedure was as follows: After reduction in CH4 at 600 °C for 1 hour, or after reduction in CH, and 1 hour reaction with CH4/C3H5 at 350 °C, the reactor was cooled to 50 °C. H 2 was subsequently passed through the reactor as the temperature was increased from 50 °C to 900 °C at a rate of 60°C /min. The exit gas was passed to the TCD detector of the GC. The reaction between H 2 and surface carbon occurred as the temperature increased. The MS was used to identify the produced gas and in the present study CH, was the exclusive product. A computer was used to manipulate the signal from the GC and the total CH, peak area was calculated. A calibration was done to obtain the ratio of the CH, peak area to the amount of CH, produced during TPSR. Hence, the amount of carbon on the surface of the catalysts was calculated. 40 Table 3-1. Calculation procedure of reaction results Term Calculation procedure C3H6 conversion % C3FL5 conversion % = [ (moles of C3H6 in feed - moles of C3PL5 in products) / (moles of C3H6 in feed)] x % Product selectivity % C x selectivity % = (moles of C x in products / moles of all products) x % *1 C 4 distribution % C4i distribution % = (moles of C 4; in products /moles of all C 4 in products ) x 100 % *2 Carbon balance % Carbon balance % = {[X(moles of C x x X xl2) in products ] / [I(moles of C x x X x 12) in feed] }x % *3 H 2 concentration % H 2 concentration % = [1- X(CX concentration in products)] x % *4 *Note: 1. C x: hydrocarbons with x C atoms excluding CFL, and C3FL3 2. C 4j: C 4 olefin and paraffin isomers 3. X: No. of C atoms in a molecule 4. X=1, 2,3,4,... 41 Chapter 4 Preliminary Study of the CH4/C3BL5 Coupling Reaction over Modified Ni/Al203 Catalysts The purpose of the present research was to upgrade CH4 through the catalytic coupling of CH4 with olefins. The research involved the selection and preparation of catalysts, activation of the catalysts and the measurement of their activity and selectivity. The main body of the research was aimed at obtaining the maximum yield of desired products and understanding the catalytic reaction mechanisms. 4.1 Modified Ni/Al203Catalysts For the preliminary study of the CH4/C3FL5 coupling reaction, a series of modified Ni catalysts supported on Y-AI2O3 (surface area = 297 m2/g) were prepared by the incipient wetness impregnation method with the nominal compositions and properties shown in Table 0-1: Table 0-1. Properties of P and K promoted 7 wt % Ni on A1203 catalysts Surface Area of Catalysts Reduction PorK Calcined Reduced in After CH4/C3H6 Temperature Ni Content CFLt Coupling Reaction from TPR Dispersion Catalyst (wt %) (m2/g) (m2/g) (m2/g) (°C) (%) Ni - 211 156 162 695 3.8 Ni/P 1 192 160 171 650 7.0 Ni/P 3 115 111 115 632 5.8 Ni/P 7 23 21 62 582 3.6 Ni/K 1 189 153 170 664 3.9 Note: The maximum deviation of N i dispersion was 17 % and the standard deviation was ±6 %. 42 Coke formation on catalysts is common in catalytic reactions of hydrocarbon synthesis and cracking. The reduction of coke is very important to maintain catalyst activity during reaction. Furthermore, it was suggested by Ovalles et al. [16] that to achieve high C 4 selectivity in CFL/CsFLs coupling, high Ni dispersion was necessary. Hence, the metal dispersion is also important in the present study. The addition of K to Ni/ catalyst catalysts is known to reduce coke formation [70,71]. The addition of P to Ni/Al 20 3 is known to decrease the formation of inactive NiAl204 during catalyst calcination [72] resulting in an increase in the NiO available for reduction and consequently, increased Ni dispersion. Therefore, K (KC1) and P (85 % o-phosphoric acid) were chosen as promoters of the Ni catalysts for these prehminary experiments. K is usually added as K 2 C0 3 , but in the present study K 2 C 0 3 gave lower activity than KC1. Hence K was added as KC1 for all the results presented here. The different results obtained by using the KC1 compared to K 2 C 0 3 as additive might be due to changes in catalyst acidity caused by the addition of KC1 or K 2 C0 3 . Calcination of the K 2 C 0 3 containing catalyst will decompose K 2 C 0 3 to K 2 0 and C0 2 . The basic properties of K 2 0 would neutralize Al203's acidity. Calcination of the KC1 added catalyst would also yield K 2 0 but CI would remain on the A1203 increasing the acidity. According to Ovalles et al. [16], the calcined catalyst (containing NiO) must be reduced in CH, to obtain an active CH,-C3H6 coupling catalyst. To determine the required temperature for reduction in CH,, temperature programmed reduction (TPR) in CH, was carried out. Figure 4-1 shows the gas components present in the effluent from the TPR apparatus for a Ni/P/Al203 catalyst. The presence of H 2 0 in the exit gas confirmed the reduction of NiO with CH,. However, the presence of large amounts of H 2 in the gas phase suggested that during reduction a significant amount of CH, decomposition, carbon deposition and H 2 production also occurred. For these 43 experiments the amount of C H 4 decomposed, the carbon deposition and H2 production were not quantified. In Chapter 5, the carbon deposition is quantified, from which the CH4 decomposition and H 2 production can be calculated. The catalyst reduction temperature determined by TPR, the BET surface areas of the calcined catalyst, the CH4 reduced catalyst and the catalyst after reaction, as well as the Ni dispersion measured by H 2 chemidesorption on the H 2 reduced catalysts, are summarized in Table 4-1. Figure 4-1. Exit gases measured by MS during TPR of Ni/P/Al203 ( 7/3/90 wt % ) catalyst in CH4 at a flowrate 20cm3/min. and a temperature range 50 °C to 900 °C at a ramp rate 30 °C/min. 44 The surface areas of the calcined catalysts decreased with increased concentration of P added to the Ni/Al203. The surface areas of the CH4 reduced and used Ni catalysts also decreased with the addition of P to Ni/Al203. Clearly the P2Os reduced the catalyst surface area presumably by a pore blocking mechanism. The addition of P also decreased the temperature at which catalyst reduction in CH4 occurred. Since P is known to reduce the formation of NiAl 20 4 [72] and NiAl 20 4 is more difficult to reduce than NiO, it follows that the reduction temperature would decrease. The addition of 1 wt % P to the Ni/Al203 catalyst increased the Ni dispersion compared to the Ni/Al203 catalyst. As the P concentration increased above 1 wt %, the Ni dispersion decreased. Apparently the addition of P decreased the formation of NiAl 20 4, resulting in higher reduction, but as P concentration increased, the support surface area decreased preventing a further increase in Ni dispersion. The effect of 1 wt % K addition to Ni/Al203 was similar to the effect of 1 wt % P addition to Ni/Al203. In both cases the surface area and TPR peak temperature decreased. However, the Ni dispersion did not change significantly with the addition of K Based on the TPR results shown in Figure 4-1 and Table 4-1, a reduction temperature of 600 °C for 1 hour was chosen as the standard procedure for reduction of the catalysts in CH4. The reduction of NiO to Ni at these conditions was confirmed by X-ray diffraction as shown in Figure 4-2 for the Ni/K/Al 20 3 catalyst. 45 Figure A for Calcined N i / K / A l 2 0 3 Figure B for Reduced Ni /K/Al 2 0 j Figure 4-2. X-ray diffraction spectrum of Ni/K/Al 20 3 catalyst before and after reduction in CH4 4.2 CH4/C3H6 Coupling Reaction over Modified Ni Catalysts A preliminary set of activity tests of the Ni catalysts was conducted in a 10% C3H6 and 90% CH4 feed gas mixture. The reaction was carried out at 350 °C and at atmospheric pressure after the calcined catalyst had been reduced in CFL, at 600 °C for 1 hour. The results of these activity tests are shown in the Table 4-2. Although the desired product was C 4 hydrocarbons, other components were present in the exit gases including C 2 to C5 hydrocarbons. The presence of a range of products implied that reactions other than desired the CH4-C3H6 coupling reaction occurred in the reactor. Note that in these preliminary experiments, the product distribution among the components having the same carbon number, was not measured. 46 Table 4-2. Results of C3H6 coupling with CH4 at 350 °C over Ni/Al 20 3 catalysts after 60 minutes time-on-stream Feed = 90 mol % CFL / 10 mol % C 3FL; Flowrate = 20 cm3 (20 °C and 101 kPa) /min. Catalyst mass = 1.1 g C3FL C 2 c 3 c 4 c 4 PorK Conversion Selectivity Selectivity Selectivity Yield Catalysts (wt %) (mol %) (mol %) (mol %) (mol %) (mol %) Ni - 5.0 66 34 0.0 0.0 Ni/P 1 9.9 54 40 6.0 0.6 Ni/P 3 8.4 38 55 7.0 0.6 Ni/P 7 8.6 27 71 2.0 0.2 Ni/K 1 45 30 59 10 4.5 Note: 1. In the case of Ni/K/Al 20 3, product selectivities consisted of C 2 , C 3 C 4 and C 5 + . 2. The standard deviation for the results can be referred to Appendix 1. Also, the CFL. conversion was not included in the data of Table 4-2 since the net CFL conversion was too low to be measured accurately. Although some CFL. may have been consumed in the coupling reaction, CH4 could also be produced by hydrogenolysis or decomposition of C3FL. The addition of 1 wt % and 3 wt % P to the Ni/Al 20 3 catalyst increased the C 4 selectivity and C3H<5 conversion compared to the unpromoted Ni/Al 20 3 catalyst. The increase in C3H5 conversion and C 4 selectivity can be attributed to the higher Ni dispersion [16]. However, with the addition of 7 wt% P, the C 4 selectivity decreased, most likely a consequence of the decrease in Ni dispersion [16]. The C 4 products were most likely generated by either C 3 H 6 homologation with a CH2=M 47 surface species or by C2H4 dimerization. Metathesis of C3FL5 produces one mole of C2H4 and one mole of C4H8 per mole of C3FL5 converted. Since there was much more C2 than C 4 in the products, metathesis of C3FL5 could only have occurred to a small extent. Hence, it is most likely that the C2's were mainly generated from the C3H5 decomposition reaction. The formation of C3H3 was likely the result of C3H5 hydrogenation. The H-M necessary for C3H5 hydrogenation could be generated from either C3H6 decomposition or from the carbonaceous species formed during the reduction. The CH X carbonaceous species are known to release H-M during the reaction according to the following sequence [21]: C H 3 - M + M C H 2 = M + H - M C H 2 = M + M C H — M + H - M C H — M + M C = M + H - M The addition of K to the Ni catalysts increased the C3HS conversion and C 4 selectivity, but in this case the increase was not due to metal dispersion since both the Ni/Al 20 3 and the Ni/K/Al203had equivalent metal dispersions. The electron donation from K to Ni was likely more important, increasing the metal-carbon bond strength [46]. The selectivity shift toward heavier hydrocarbons is assumed to be the consequence of a concentration increase of carbon-containing species on the metal [73]. Figure 4-3 and Figure 4-4 shows the C3H6 conversion and C 4 selectivity change with time-on-stream over the Ni/K7A1203 and Ni/P/Al203 catalysts. The effect of catalyst deactivation has not been reported in previous studies. That the catalyst activity declined with time-on-stream was likely the result of the decomposition of C3H6. The produced carbonaceous species deposited on the catalyst surface and deactivated the catalysts during the reaction. Another possible reason for 48 the catalyst deactivation is that both the active CH X carbonaceous species and H adsorbed on the catalyst surface, that were formed during the catalyst reduction in CH4, were depleted by reaction with C3H6. If the amount of new active species generated from C H 4 or C3FL5 dissociation during the reaction was insignificant, the C3FL5 conversion would decrease as shown in Figure 4-3. The C 4 selectivity over the Ni/K7A1203 catalyst was higher than over the Ni/P/Al203 catalyst. The C 4 selectivity over the Ni/P/Al203 catalyst decreased with time-on-stream while the C 4 selectivity over the Ni/K/Al203 catalyst did not change significantly with time-on-stream In conclusion, the C3FI6 conversion and C 4 selectivity increased with addition of K and P to the Ni catalysts. The Ni/K/Al203 gave the highest C3FL5 conversion and C 4 selectivity of all the catalysts examined. Among the Ni/P/Al203 catalysts, Ni/P/Al2C>3 with 1% P was better than Ni/P/Al203 with 7% P due to the fact that the former had a higher Ni dispersion and surface area u -I 1 1 1 1 1 1 1 0 2 0 4 0 6 0 80 100 1 2 0 1 4 0 T i m e - o n - s t r e a m , min • N i / K / A l 2 0 3 • N i / P / A l 2 0 3 Figure 4-3. C3H6 conversion over Ni/K /A1203 (1% K) and Ni/P/Al 20 3 (1% P). Reaction at 350 °C and 101 kPa. Feed gas = 90 mol % CFL,/10 mol % C3Hs. 49 5 - \ 0 20 40 60 80 100 120 140 T i m e - o n - s t r e a m , min • N i / K / A l 2 0 3 • N i / P / A l 2 0 3 Figure 4-4. C 4 selectivity over Ni/K /A1203 (1% K) and Ni/P/Al203 (1% P). Reaction at 350 °C and 101 kPa. Feed gas = 90 mol % CFL»/10 mol % C3H6. than the latter. Since the Ni/K/Al 20 3 had similar Ni dispersion as the unpromoted catalyst but gave much higher C3Hs conversion and higher C4 selectivity, it is concluded that a chemical interaction between the Ni and K played more important role than metal dispersion in determining catalyst performance. The decline in catalyst activity with time-on-stream is thought to be related to carbon deposition on the catalyst surface. Catalyst reduction was carried out in CFL» before reaction, and carbon was deposited on the catalyst surface during this process. The reactivity and type of carbon on the catalyst surface and its role in the subsequent reaction needs to be examined. In addition, Table 4-2 shows that reactions other than the desired CH4/C3H6 coupling reaction occurred and the product distribution changed with the type of catalyst. It is important to investigate the product distribution at different reaction conditions and to examine the different 50 types of reactions occurring on the catalyst. In the following chapters, experiments designed to address these questions are described and discussed. 51 Chapter 5 Carbonaceous Deposits on Ni Catalysts During the reduction of the calcined catalysts in CBU, some carbonaceous species were deposited on the surface of the catalysts. During the reaction with C3H5 and CH4, carbonaceous species from the reactants were also formed on the surface of catalysts. Temperature programmed surface reaction (TPSR) in H 2 was conducted to investigate the relationship between the reactivity of the carbonaceous species on the surface of the catalysts and the activity of the catalysts, Hence, the reactivity of the carbonaceous species formed in the reduction process as well as in the reaction process was determined. In addition, the CH4/C3HS coupling reaction was conducted over a H 2 reduced Ni catalyst and a comparison was made between the reaction results from the H 2 reduced Ni catalyst and the CH4 reduced Ni catalyst. In this chapter, the results of these experiments are presented and discussed. 5.1 Study of Surface Carbon Using TPSR In TPSR, carbonaceous species deposited on the catalyst surface are reacted with H 2 as the reaction temperature is increased linearly. The product gas released during the TPSR was exclusively CH4 for all the Ni catalysts examined herein. 52 4e+6 0 100 200 300 400 500 600 700 800 900 Temperature °C Figure 5-l.TPSR profile of A1203 supported catalysts following 1 hour reduction in CFL, at 600 °C Figure 5-1 shows the TPSR profiles for the Ni/Al 20 3, Ni/K/Al 20 3 and Ni/P/Al203 catalysts after 1 hour reduction in CFL, at 600 °C. It can been seen that two types of carbon deposit resulted from the CFL, reduction, each identified by their reactivity to H 2 . One group of peaks, which were relatively small, occurred in the low temperature range 190 °C to 300 °C. The second group of peaks, which were very large, appeared in the high temperature range 540 °C to 670 °C. The high temperature peaks represent less reactive carbonaceous species, perhaps graphitic [55,70,75], that would not be hydrogenated at the temperature of the C3H<5 coupling reaction 53 (350 °C). The lower temperature peaks, which may be carbidic [55,70,75], were generated at 192 °C to 237 °C. These carbon species would likely be reactive during the coupling reaction since H is present on the surface ( as a consequence of CH4 decomposition during reduction) and the reaction temperature is 350 °C. Therefore, the low temperature peaks of the TPSR profile were examined in more detail. Table 5-1 shows the amount of carbon on the different catalyst surfaces at different temperatures obtained by integration of the TPSR profiles. In each case, the amount of surface carbon reactive at high temperature was much larger than that reactive at low temperature. Table 5-1. Carbon content of catalyst after reduction in CH4 at 600 °C for one hour measured by TPSR in H 2 Peak Temperature and Carbon Amount Catalyst PorK T P Amount T P Amount (wt %) <°C) (mmol/gcat) <°C) (mmol/gcat) Ni/Al 20 3 0 192 0.014 547 0.637 Ni/K/Al 20 3 1 217 0.011 637 0.914 Ni/P/Al203 1 237 0.008 667 0.575 (a) The amount of carbon calculated by integration of TPSR profile for each peak identified by the peak temperature T p . 54 Figure 5-2 shows the TPSR low temperature peak profiles of AI2O3 supported catalysts after 1 hour reduction in CFLt at 600 °C. The addition of P or K to the Ni/Al 20 3 catalysts decreased the amount of the low temperature carbonaceous species. The data show that the amount of reactive carbonaceous species on the catalyst after reduction in CFLt increased in the order Ni > Ni/K > Ni/P. However, as shown in Table 4-2, both C3H6 conversion and C4 selectivity increased upon addition of K or P to the Ni/Al203 catalyst. Therefore, there is no relationship between the catalyst performance and the amount of reactive carbon on the catalyst surface after reduction. In a second set of experiments, TPSR was performed after 1 hour reduction in CFLt at 600 °C and 1 hour reaction with CH4/C3H6 at 350 °C. Table 5-2 shows the peak temperature (Tp) and the amount of surface carbon that reacted with H 2 during the TPSR. If the peak temperatures in Table 5-2 are compared with the peak temperatures in Table 5-1, a shift to higher peak temperature (from T p = 192 - 237 °C to T p = 292 - 342 °C ) is observed. One possible reason for this shift is the release of H from the carbonaceous species to form less reactive species: C H 2 CHCH 2 CH 3 C II + I 1 + CH 3 CH 2 CH 3 + M M M M 55 1e+5 Figure 5-2. TPSR profile of AI2O3 supported catalysts following 1 hour reduction in CH4 at 600 °C showing low temperature peaks According to Koerts and Van Santen [75] the adsorption heat of CH X-M increases with decreasing H content i.e. CH X is more strongly adsorbed to the catalyst surface and therefore requires higher temperature to be hydrogenated. The shift to higher peak temperature may also be due to the build-up of carbonaceous species and formation of C-C species that would be more graphitic and less reactive than the C=M species [75]. 56 Table 5-2. Carbon content of catalyst after reduction in CF£» at 600 °C for 1 hour and reaction in 90 mol % CH4/IO mol % C3H<; at 350 °C for 1 hour measured by TPSR in H 2 Peak Temperature and Carbon Amount Catalyst PorK T P Amount T P Amount (wt %) (°C) (mmol/gcat) (°C) (mmol/gcat) Ni/Al 20 3 0 292 0.015 612 1.43 Ni/K/Al 20 3 1 317 0.078 602 1.74 Ni/P/Al203 1 342 0.042 617 1.21 (a) The amount of carbon calculated by integration of TPSR profile for each peak identified by the peak temperature T p . Figure 5-3 shows the low temperature TPSR profiles of Ni catalysts after reduction in CH, and reaction with CH4-C3FI6. The addition of P or K to Ni/Al 20 3 catalysts increased the amount of low-temperature carbon compared to the unpromoted Ni/Al 20 3 catalyst, as quantified in Table 5-2. 57 Figure 5-3. TPSR profile of AI2O3 supported catalysts following 1 hour reduction in CFL, at 600 °C and reaction in CFL,/ CsHs at 350°C for 1 hour Figure 5-4 and Figure 5-5 show the relationship between the amount of low temperature carbon determined by TPSR after reaction, and the C 4 selectivity and the C3FL5 conversion. Clearly a correlation between the CFL, released during the TPSR at low temperature and the catalyst performance is apparent. Both C3FL5 conversion and C 4 selectivity increased with increased amount of low temperature carbonaceous species on the catalyst surface. 58 10 Ni/K/ALO. o :> Ni/P/Al,0, CO 2 0.00 0.02 0.04 0.06 0.08 0.10 mmole C H 4 Figure 5-4. The relationship between C4 selectivity and the moles of CFL, released during TPSR at low temperature peak range after 1 hour reaction at 350 °C Ni/K/AljO, • (D 5, 20 o 15 o O Ni/AI.O. 0.04 0.06 mmole C H . Figure 5-5. The relationship between C3FL5 conversion and the moles of CH4 released during TPSR at low temperature peak range after 1 hour reaction at 350 °C 59 The question that arises from these results is whether the carbonaceous species produced from CH4 reduction was involved in the CH4/C3H6 coupling reaction or whether only carbonaceous species generated during reaction reacted with C3Hs to form C4? In an attempt to answer this question, the following experiments were conducted. 5.2 The Effect of Carbonaceous Species Generated during CH4 Reduction To examine the role of the carbonaceous species generated during the CFL, reduction step, the Ni/K/Al203 catalyst was reduced with H 2 instead of CH4 prior to the CH4/C3H6 activity measurement. Differences between the CBL» reduced catalyst and the H 2 reduced catalyst in the subsequent CH4/C3H6 reaction must be caused by the carbonaceous species that existed on the catalyst surface before the reaction. Figure 5-6 A shows that the initial conversion of C3H5 was higher over the CH4 reduced catalyst than the H 2 reduced catalyst. The C3FL5 conversion over C H 4 reduced catalyst decreased monotonically, whereas the C3FL5 conversion over the H 2 reduced catalyst initially increased with time-on-stream and then reached a maximum value. Initially the carbonaceous species and H generated on the catalyst surface by reduction in CH4 may react with C3FL5, which contributes to the high C3FLs conversion initially. In support of this assertion, note that on the H 2 reduced catalyst once carbonaceous species built up on surface, the C3FL5 conversion increased. The C3FL5 conversion decrease with time-on-stream over the CFL, reduced catalyst was likely caused by the continuous deposition of carbonaceous species and consumption of H on the surface during the catalyst activity measurement. 60 Figure 5-6. Comparison of C3FL5 conversion (Figure A) and C4 selectivity (Figure B) over Ni/K/Al 20 3 catalysts reduced in H 2 and CFL,, Reaction at 350 °C, lOlkPa and 10 % C3FL5 / 90 % CFL feed gas 61 The change in C3FL5 conversion over the H 2 reduced catalyst with time-on-stream can be divided into three periods. In the initial period, carbonaceous and H species build up on the metal sites of the catalyst surface as a consequence of C3FL3 decomposition reactions. These species react with C3FL3, resulting in an increase in C3H6 conversion. In the second period, some of the deposited carbonaceous species migrate from the metal sites onto the support [86]. Since the species on the metal sites were reactive and metal sites were still exposed to the reactants, the C3FL5 conversion did not change significantly. In the third period, more carbonaceous species deposit, the carbonaceous species on the metal sites age, lose H and become less reactive. Hence C3H5 conversion decreases. A model of this process of carbonaceous species deposition is illustrated as follows: Metal Sites Support Phase 1 Bond formed between metal and CH X or H CH X CH X H Phase 2 62 Carbidic Carbon CH X CH X CH X C H c c Phase 3 Graphitic Carbon CC CCC CCC' C-C C-C C-C C C-C C C C-C CH X cc i c c Phase 4 In phase 1, which exists in the case of the H 2 reduced catalyst before reaction, there was no carbonaceous species on either the reduced metal sites or the support. In phase 2, the initial period of reaction for the H 2 reduced catalyst, some carbonaceous species deposit on the metal sites and these carbonaceous species are reactive (carbidic carbon). In phase 3, the metal site is saturated with carbonaceous species and some of the carbonaceous species migrate to the support. Phase 3 likely also exists in the case of the CH4 reduced catalyst before reaction. In phase 4, with continued deposition of carbonaceous species, some of the carbonaceous species migrate to the support and some react with each other to form graphitic carbon and remain on the metal site deactivating the metaL as was observed in the case of the CH4 reduced catalyst and the H 2 reduced catalyst after a period of reaction. 63 The proposed model can be confirmed by the comparison of the results calculated from the Ni dispersion data, measured by temperature programmed desorption, and the results of the TPSR experiments. Table 5-3 and Table 5-4 show the comparisons. Table 5-3. Catalyst surface carbon measured by temperature programmed surface reaction (TPSR) experiments after reduction in CH4 at 600 °C for 1 hour compared to surface Ni measured by temperature programmed desorption (TPD). TPSR Low TPSR High N i available on Ratio of low temperature temperature catalyst surface temperature Ratio of total carbon carbon from TPD test carbon to N i carbon to N i Catalyst P o r K Amount Amount Amount (wt %) (mmol/gcat) (mmol/gcat) (mmol/gcat) (mmol/mmol) (mmol/mmol) N i / A l 2 0 3 0 0.014 0.64 0.045 0.31 14.5 N i / K / A l 2 0 3 1 0.011 0.91 0.047 0.23 19.6 N i / P / A l 2 0 3 1 0.008 0.58 0.083 0.09 7.08 Table 5-4. Catalyst surface carbon measured by temperature programmed surface reaction (TPSR) experiments after reduction in CH4 at 600 °C for 1 hour and reaction in CH4/C3H6 at 350 °C for 1 hour compared to surface Ni measured by temperature programmed desorption (TPD). TPSR Low TPSR High N i available on Ratio of low temperature temperature catalyst surface temperature Ratio of total carbon carbon from TPD test carbon to N i carbon to N i Catalyst P o r K Amount Amount Amount (wt %) (mmol/gcat) (mmol/gcat) (mmol/gcat) (mmol/mmol) (mmol/mmol) N i / A l 2 0 3 0 0.015 1.43 0.045 0.33 32.1 N i / K / A l 2 0 3 1 0.078 1.74 0.047 1.73 38.7 Ni/P/A12Q3 1 0.042 1.21 0.083 0.50 15.1 64 The data of Table 5-3 and Table 5-4 lead to the following important conclusions: 1. The amount of carbon deposited on the catalyst during the CH4 reduction exceeded the amount of surface Ni available on the catalyst. 2. Deposition of carbonaceous species continued during the CH4/C3H6 reaction at 350 °C. 3. Since carbon decomposition continued well beyond a monolayer coverage of the surface Ni, migration of the carbon deposit from the Ni to the support must have occurred. All these conclusions are consistent with the proposed carbon deposition model. By comparing the data in Table 5-3 and 5-4, an estimate of the carbon deposition that occurs during the reaction step can be made. For example, for the Ni/K/Al203 catalyst, the amount of carbon that is added during 1 hour of reaction is about 0.9 mmol/g cat. If one assumes that all the carbon is generated from C3FL5 decomposition, then this amount of carbon corresponds to about a 6 % C3FI6 conversion to surface carbon, in agreement with the previous assertion that the level of C3FL5 conversion to carbon species is low and does not significantly affect the selectivities reported according to the formula given in Table 3.1. Figure 5-6 B shows that the C 4 selectivities changed little with time-on-stream and were about the same over both the H 2 reduced catalyst and the CFL, reduced catalyst. Hence, although the CFL reduced catalyst had carbonaceous species deposited during reduction whereas the H 2 reduced catalyst did not, the C 4 selectivities were similar. This fact should be taken as evidence that the Ni-CHX generated during catalyst reduction in CFL, was not involved in C 4 formation at 65 these experimental conditions (high C3FL5 conversion). As shown in Figure 5-7, initially the C3FL3 selectivity over the CFL, reduced catalyst was higher than over the H 2 reduced catalyst, however, the C3 selectivity over the CFL, reduced catalyst decreased with time-on-stream As discussed previously, carbonaceous species with stoichiometry CH X (x = 1, 2 or 3) as well as H were generated during the reduction in CFL,. That the C 3 selectivity was high at the beginning of the reaction over the CFL, reduced catalyst was due to C3FL5 hydrogenation by the H available on the catalyst surface. There was also F£2 formed from C3FL5 decomposition during reaction. The C3FL, conversion on the CFL, reduced catalyst decreased with time-on-stream due to a consumption of the H deposited during CFL, reduction. Since there was no H on the catalyst at the beginning of the reaction over the H 2 reduced catalyst, H was generated from C3FL5 decomposition only. Hence, since the catalyst activity for the C3FL5 decomposition reaction did not change significantly, C3 and C 2 selectivities would not change much either. These results suggest that the carbonaceous deposit generated during CFL, reduction at high temperature (600 °C) was not the same as that on the surface generated during reaction at low temperature (350 °C) in terms of its reactivity with C3FL5. A similar conclusion was drawn from the TPSR studies in Chapter 4. Finally, it is worth noting that, in contrast to the claims of Ovalles et al. [16] and Loftier et al. [22], CFL reduction did not have a strong effect on C 4 selectivity on the catalysts of the present study. However, the present results were obtained at relatively high C3H6 conversions. In Chapter 6 it is shown that some of the carbonaceous species generated during CH, reduction do homologate C3FL5, but this reaction is only observed at low C3FL5 conversions. 66 Figure 5-7. Comparison of C3 (Figure A) and C2 selectivity (Figure B) over Ni/K7A1203 catalysts reduced in H 2 and CH4. Reaction at 350 °C, lOlkPa and 10 % C3FL5 / 90 % CFL, feed gas. 67 Figure 5-8 and Figure 5-9 compare the change in the distribution of C 4 products with time-on-stream over the H 2 reduced and CFLt reduced catalysts. In both cases the products included 1-butene, 2-C4H8, iso-butylene and n-C4Hi0. There was no iso-C4Hi0 detectable in the products. The change in distribution of C^s over the H 2 reduced catalyst was much less than that observed over the CH4 reduced catalyst. The differences in C4 distribution between the H 2 reduced catalyst and the C H 4 reduced catalyst are also related to the carbonaceous species and H on the catalyst surface. Initially, there was no carbon or H on the surface of the H 2 reduced catalyst. Hence hydrogenation of C4 olefins was a minor reaction and the n-C4H10 selectivity (Figure 5-8) did not show a sharp decrease over the H 2 reduced catalyst. In contrast, the n-C4Hi0 selectivity decreased rapidly over the CFLt reduced catalyst with time-on-steam Since there were carbonaceous species and H on the surface of the CH4 reduced catalyst before reaction, the initial n-C4Hi0 selectivity was high, caused by C4 olefin hydrogenation. With more carbonaceous species deposited on the CFLt reduced catalyst with time-on-steam, the catalyst lost its activity and less H was formed by C3FL5 decomposition. Therefore the n-C4Hi0 selectivity decreased. 68 16 0 20 40 60 80 100 120 140 T i m e - o n - s t r e a m , m i n s • H 2 R e d u c e d Cata lyst • C H „ R e d u c e d Cata lys t Figure 5-8. Comparison of n-C4Hi0 in total C 4 over Ni/K/Al 20 3 catalysts reduced in H 2 and CH4 Reaction at 350 °C, lOlkPa and 10 % C3Fl6 / 90 % CFLt feed gas As shown in Figure 5-9, 1-butene increased and n-C+Hio decreased significantly with time-on-stream over the CFL, reduced catalyst. As shown in Section 6.3, the hydrogenation of 1-butene produced n-C4Hi0. The CFLt reduced catalyst activity decreased with time-on-stream as no more H was released from the previously deposited carbonaceous species. Hence hydrogenation decreased which resulted in an increase in 1-butene and decrease in n-C4Hi0 selectivity with time-on-stream It should be noted that the expected product of C3H6 homologation is 1-butene. 69 Figure B "1 i i i i i i I 0 20 40 60 80 100 120 140 Time-on-stream, mins • H 2 Reduced Catalyst • CH 4 Reduced Catalyst Figure 5-9. Comparison of 1-butene (Figure A) and isobutylene (Figure B) in total C4's over Ni/K/Al 20 3 catalysts reduced in H 2 and CFL,. Reaction at 350 °C, lOlkPa and 10 % C3FL / 90 % CFL feed gas 70 5.3 Summary In this chapter the carbon deposited on the catalyst during reduction and reaction was investigated. Two kinds of carbon deposit were identified: one reactive to H 2 at low temperature (190 °C to 350 °C) and another reactive to H 2 at high temperature (545 °C to 670 °C). As the amount of the low temperature carbon generated during reaction increased, the C3H5 conversion and C4 selectivity increased. The role of the carbonaceous species generated during the reduction in CH4 was examined by comparing product distributions obtained over CH4 reduced and H 2 reduced catalyst. The catalyst activity can be increased by the H and the surface carbon formed during catalyst reduction in CH4 and the catalyst activity can also be decreased by the surface carbon depending on the type and amount of surface carbon on the catalyst surface. Based on the experimental data, a carbon deposition model was proposed. By comparison of the results of the CH4/C3FL3 coupling reaction (with high C3H6 conversion) over the CH4 reduced and the H 2 reduced catalysts, it may be concluded that the carbonaceous species generated during the CH4 reduction was not involved in C4 formation. However, as discussed in following chapters, this conclusion is valid only at the relatively high C3H6 conversion conditions examined here. 71 Chapter 6 Effect of Process Variables on CH4/C3H6 Coupling The results presented in Chapter 4 showed that the C4 selectivity from CH4/C3H6 reaction was low (< 10 %). However, the results from Chapter 5 suggest that the selectivity can be increased by increasing the amount of the low temperature carbonaceous species on the catalyst surface. In an attempt to increase C 4 selectivity and to better understand the product distribution and reaction mechanism of the CH4/C3H5 reaction, the effect of different reaction conditions was investigated. The variables studied included: - reduction time - reaction temperature - residence time from which the different reaction pathways for the CH4/C3H6 coupling and other side reactions were identified. 6.1 The Effect of Reduction Time on the Coupling Reaction The carbonaceous species on the catalyst formed during reduction in CH4 may increase the observed catalyst activity by reacting with C3H6 , or may decrease the activity of the catalyst by covering Ni sites. Since the reduction time will likely affect the type and amount of carbonaceous species on the catalyst surface [76], a number of tests were conducted to determine the optimum catalyst reduction time in CH4. 72 Figure 6-1 and Figure 6-2 show the effects of reduction time on the subsequent conversion of C3H6 and resulting C 4 selectivity. The data clearly show that if the catalyst was not reduced, there was no C3H5 conversion. Furthermore, the catalyst activity decreased with increased length of reduction time. This was most likely caused by an increase in the carbonaceous deposit with increased reduction. Hence, coverage of the Ni sites increased and the activity decreased. It was also possible that reactive carbonaceous species were converted into less reactive carbonaceous species with low H content, when exposed to 600°C for longer periods [76]. 100 O ID O O 40 60 80 100 120 Time-on-stream, mins • No Reduction • 15 mins Reduction A 60 mins Reduction • 90 mins Reduction 140 Figure 6-1. The effect of reduction time in CH, on the C3FL5 conversion over Ni/K7A1203. Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CH4 feed gas 73 40 60 80 100 Time-on-stream, mins 140 No Reduction 60 mins Reduction 15 mins Reduction 90 mins Reduction Figure 6-2. The effects of reduction time in CFL, on the C4 selectivity over Ni/K/Al203. Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CFL, feed gas The C 4 selectivity was almost the same for all reduction times, as shown in Figure 6-2. This result indicates that the reduction time only changed the amount of Ni-CF£X or Ni on the catalyst, which in turn affected the C3FL5 conversion but not the C4 selectivity. Since the catalyst reduced for 60 minutes gave stable C4 selectivity, the results reported in the following sections were obtained over catalysts reduced for 60 minutes. 74 6.2 The Effect of Reaction Temperature on the Coupling Reaction The effects of reaction temperature on the conversion and product distribution were investigated in the range 300 °C to 375 °C. The choice of the temperature range used here was based on the fact that there was no reaction at below 300 °C (275 °C) and all C3H6 was converted to CFL,, F£2 and C at above 375 °C ( 400 °C). Figure 6-3, Figure 6-4 and Figure 6-5 show the conversion of C3FL5 and the product selectivities from the reaction at different temperatures. The data show that high temperature resulted in high conversion of C3FL5. When the reaction temperature was > 350 °C, the C4 selectivity decreased. When the reaction proceeded at a temperature lower than 325 °C, the conversion was less than 30 %. Considering the data obtained at 375 °C, the conversion was almost 100 % at the beginning of the experiment and decreased rapidly with time-on-stream Furthermore, the C 3 selectivity decreased and the C4 selectivity increased with time-on-stream At this high temperature, C3FL5 was completely converted, generating more H 2 and carbon on the catalyst than at low temperature. As time proceeded, the extent of the hydrogenation reaction decreased (C3 selectivity decreased) as did olefin decomposition and F£ generation (C3FL5 conversion) because of the cumulative carbon deposition and deactivation of the catalyst. Hence, the C4 selectivity increased. At 300°C, the C 4 selectivity rapidly decreased from 9 % to 3 % within 60 minutes while the C3H<; conversion decreased from 11 % to 5 %. At these low C3FL3 conversions, the C 4 product was likely formed from the combination of C3FL5 with carbonaceous species (CHX-M) generated during catalyst reduction [16]. Since the carbonaceous species formed during the reduction could 75 Figure 6-3. The effects of reaction temperature on the C3H6 conversion over Ni/K7A1203, Reaction at 101 kPa and 10 % C3FL5 / 90 % CFL, feed gas Figure 6-4. The effects of reaction temperature on the C4 selectivity over Ni/K7A1203. Reaction at 101 kPa and 10 % C3H5 / 90 % CFL, feed gas 76 80 0s > t5 0) 0) CO m o 75 70 £ 65 60 55 50 45 Figure A i 20 40 60 80 100 120 140 20 40 60 80 100 120 Time-on-stream, mins 140 • 300 °C 350 °C 3 2 5 ° C 375 °C Figure 6-5. The effect of reaction temperature on the C3 (Figure A) and C 2 (Figure B) selectivities over Ni/K/Al 20 3. Reaction at 101 kPa and 10 % C3FL5 / 90 % CH4 feed gas 77 not be regenerated at low C3FL5 conversion, the C4 selectivity decreased with time-on-stream. In this way, the carbonaceous species (CHX-M) generated during catalyst reduction in CH4 were involved in the C4 formation. This conclusion is in apparent contradiction with the conclusion drawn in Section 5.2. A similar conclusion is made later in this Chapter, i.e., the carbonaceous species generated during CH4 reduction were involved in the formation of C4 products. This apparent contradiction will be discussed further in a later section of this Chapter. T ime-on-s t ream Figure 6-6. The changes of the ratio of C4 production and C3H6 conversion with time-on-stream over Ni/K/Al 20 3. Reaction at 101 kPa and 10 % C3IL5 / 90 % CFL, feed gas The effect of temperature on C4 selectivity discussed above, has not considered the possible influence of C3FL5 decomposition to carbon species on the selectivity trends. An alternative definition for C4 selectivity is to consider the ratio of moles of C 4 produced to the moles of C3FL3 converted. If C3FL5 conversion to carbonaceous deposits is significant this 78 definition of C4 selectivity will be very different from that reported in the present work and defined in Table 3.1. However, Figure 6-6 shows the same trends in the C4 produced/CsFLs consumed ratio as observed for C4 selectivity in Figure 6.4. Hence the effect of C3FL5 conversion to carbonaceous species on the selectivity as defined in Table 3.1, must be small. Indeed, in the present work the difference between the moles of C3FL5 converted and the total moles of C 2 , C 3 plus C4 produced reflects the carbon balance. For the range of conditions studied here, the carbon balance was basically >90 % (see Appendix 3). Hence the trends in C4 selectivity based on the definition given in Table 3.1 will not be affected to any significant degree by the decomposition of C3KL5 for the range of operating conditions examined. The C3 selectivity decreased with time-on-stream for all reaction temperatures studied. The route to C3H8 is by hydrogenation of C3IL5. The required H 2 for hydrogenation may be provided through the decomposition of C3FL5 and from carbonaceous species formed during reduction (the carbonaceous species release H-M according to reaction CH X -M + M —> CH x.i-M + H-M). Since for all temperatures C3FL5 conversion decreased with time-on-stream, the H produced from C3H6 decomposition would also decrease with time-on-stream. Hence the rate of hydrogenation and consequent C3H8 selectivity decreased with time-on-stream The effect of temperature on the C3H6 conversion was also described by the conventional Arrhenius equation: Ea k = k0e~RT The rate constant k was calculated assuming an ideal plug-flow reactor (PFR) and first order reaction kinetics: 79 dC dt A _ = JfeC\ A plot of Ink versus 1/T shown in was linear from which Ea and ko were estimated as follows: Ea = 200.4 kJ/mol and ko = 1.13X1019 min"1 \ • 1.50 1.55 • 1.60 1 .65 1.70 1 .75 1.80 1/T (K')/1000 Figure 6-7. The estimation of activation energy for the C3FL5 conversion over Ni/K/Al203. Reaction at 101 kPa and 10 % C3H5 / 90 % CFL. feed gas Activation energies for some of the reactions of interest in the present work have not been reported in the literature. However, the activation energy for C3FL5 hydrogenation over Rh/Zeolite catalyst at 130°C -170 °C was reported as 101.4 kJ/mol [77]. The activation energy 80 for C3Hg hydrogenolysis over Ni/Si02 catalyst at 181°C - 210 °C and over Ni/Ti02 catalyst at 181°C - 283 °C [78] was about 155 kJ/mol. However no value for C3H5 decomposition is available. Since the C3H6 conversion reaction involved C3H5-CH4 coupling, C3H5 decomposition, C3H6 hydrogenation and C3H; hydrogenolysis, the activation energy obtained here represented an overall reaction activation energy. The magnitude of the estimated activation energy for C3H<5 conversion of the present work is high, as one might expect for C-C bond breakage as the dominant reactions, yielding CH4, C2H4, C2H6 and carbonaceous deposits on the catalyst surface. 6.3 The Effects of Residence Time on the Coupling Reaction The effect of residence time (x = W/F where W is the catalyst weight in the reactor and F is the feed gas flowrate ) on the product distribution, especially the C4 product distribution, was examined in an attempt to identify the reactions leading to C4's. In these tests, the weight of catalyst used in the reactor was changed while the flow rate of the reactants was kept constant. The conversion of C3FL5 declined with the time-on-stream as shown before. Hence the comparisons presented in this section were made at the same time-on-stream. If one assumes that the C3H6 conversion (x) follows first order kinetics, then a plot of ln(l-x) with residence time (x) would be linear. Figure 6-8 shows a linear relationship between ln(l-x) and x indicating that the estimation of activation energy for the C3H6 conversion, based on first order reaction kinetics, was correct. 81 -1 1 1 1 1 1 1 0.02 0.03 0.04 0.05 0.06 0.07 O.OS Residence Time X , g .min/ml Figure 6-8. The effect of residence time (x) on the ln(l-x) (x is C3H5 conversion) over M/K/AI2O3 in 30 minutes time-on-stream. Reaction at 350 °C, 101 kPa and 10 % CJth / 90 % CH4 feed gas. Figure 6-9. The effects of residence time on the selectivity - 5 minutes time-on-stream- over Ni/K/Al203 catalysts. Reaction at 350 °C, 101 kPa and 10 % CjHs / 90 % CH4 feed gas 82 To minimize the effect of catalyst deactivation, product selectrvities after 5 minutes time-on-stream are shown in Figure 6-9. The C4 selectivity decreased with increased residence time, which suggests that C 4 was a primary product from the C3H6/CHX coupling, at least in the initial stage of the reaction. When the selectrvities were measured after longer time-on-stream, the effect of residence time changed. Unlike the C4 selectivity in Figure 6-9 which decreased with residence time, at longer time-on-stream the C 4 selectrvities increased with residence time in Figure 6-10, suggesting the C 4 was not formed by a primary reaction. The change of C 4 formation from a primary reaction with time-on-stream suggests that the C 4 products were initially formed by homologation between C3FL5 and Ni-CHX species, the latter formed during catalyst reduction in CH4. As the reaction proceeded, the amount of the Ni-CHX species from catalyst reduction decreased and the amount of Ni-CHX from C3FL5 decomposition increased. Consequently, the C 4 products were only formed by a primary reaction initially. From the above argument, it may be concluded that CH4 participated in CFL/C3FL5 coupling reaction only as Ni-CHX species formed during the catalyst reduction in CFL,. The C4 products formed later in the reaction were from the reaction of C3FL5 and the Ni-CHX formed by C3FL5 decomposition or possibly by dimerization of C2 fragments generated from C3H6 decomposition. 83 0.02 0.03 0.04 0.05 0.06 0.07 0.08 Residence Time, W/F, g.min/ml C 4 Selectivity • C 3 Selectivity A c 2 Selectivity Figure 6-10. The effects of residence time on selectivity - time-on-stream 30 minutes (Figure A) and 130 minutes (Figure B) - over Ni/K/Al 20 3 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CFL, feed gas 84 Figure 6-10 shows the selectivities as a function of residence time after 30 minutes and 130 minutes time-on-stream The trends in C3 and C2 selectivities with residence time shown in Figure 6-10 (30 and 130 mins time-on-stream) were the same as the trends of C3 and C2 selectivities shown in Figure 6-9 (5mins time-on-stream). These trends suggest that the C3 product was formed via a secondary reaction and the C 2 products were formed via a primary reaction. The conclusion that CFU participated in CFL/C3H6 coupling reaction only as Ni-CHX species formed during the reduction and the conclusion drawn in Section 6.2, contradict the conclusion in Section 5.2. In Section 5.2, it was concluded that the carbonaceous species generated during CFL, reduction were not involved in the formation of C4 products. This conclusion was based on C 4 selectivities measured on H 2 reduced and CFL reduced Ni/K7A12C>3 catalysts. In both cases, the C3FL5 conversion was above 30 % and the comparisons were based on time-on-stream > 5 minutes. Hence, the initial C3H5 reaction with CHX-Ni generated in the reduction step was likely not observed. The results from the present study suggest therefore, that initially CH X species generated during catalyst reduction react with C3FL5. However, the concentration of these species is low, they are rapidly consumed and subsequently CH X species generated from C3FL5 decomposition are primarily involved in the C3H6/CH4 coupling. The C 4 components formed in the reaction included isomers of paraffins and olefins. Before the effects of residence time on C 4 distribution are presented, the changes in paraffin and olefin distribution with time-on-stream are presented in Figure 6-11. Most of the C4's were olefins and the olefin selectivity increased with time-on-stream The paraffins consisted of a small part (<10%) of the total C4's and decreased with the time-on-stream As the catalyst activity 85 decreased, the olefins were not hydrogenated. 0 20 40 60 80 100 120 140 T i m e - o n - s t r e a m , mins • Paraffin • Olefin Figure 6-11. The changes of paraffin and olefin distribution in total C 4 with time-on-stream over Ni/K/Al203 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CH, feed gas The effects of residence time on the paraffin and olefin distribution in total C4's are shown in Figure 6-12. The paraffin content increased and the olefin content decreased with residence time. This trend shows that the paraffins result from a secondary reaction and are a product of olefin hydrogenation. The change in C4 distribution with the residence time at different times-on-stream are shown in Figure 6-13 and Figure 6-14. Figure 6-13 shows that n-C+Hio , which accounted for most of the C4 paraffins, increased with residence time. ISO-C4H10 was formed after 5 minutes time-on-stream and longer residence time. There was no iso-C4Hi0 after 105 minutes time-on-stream 86 0.02 0.04 0.06 0.08 Residence Time, W/F, g.mins/ml • Paraffin • Olefin Figure 6-12. The effects of residence time on the paraffin and olefin in the total C4's - 30 minutes time-on-stream - over Ni/K/Al203. Reaction at 350 °C, 101 kPa and 10 % CsFfe / 90 % CFL, feed gas Figure 6-14 shows the change in C4 olefin distribution with residence time. The data show a decreasing trend in 1-butene and an increasing trend in other C 4 olefins with increasing residence time, after 30 minutes and 105 minutes time-on-streams. The increase in n-C4Hi0 in Figure 6-13 and the decrease in 1-butene in Figure 6-14 with increased residence time were consistent with the statement that n-C4Hi0 was the product of 1-butene hydrogenation and that 1-butene was a primary product. Furthermore, the data of both Figure 6-13 (A and B) and Figure 6-14 (A and B), show that with increased time-on-stream the n-C4Hio selectivity decreased andl-butene increased, implying that as the catalyst deactivated, it lost its ability to hydrogenate 1-butene to n-C4H10. 87 50 Figure 6-13. The effects of residence time on the paraffin in the total C/s - 5 minutes time-on-stream in (Figure A) and 105 minutes time-on-stream (Figure B) - over Ni/K/Al 20 3 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3He / 90 % CFLt feed gas 88 45 in ~* O o c o is at c <u 40 35 30 I 25 20 15 10 0.02 50 -t 45 -at <S 40 -o h- 35 -c c 30 -o 1 JQ •c 25 -to b c 20 -"5 o 15 -10 -0.04 Figure A 0.06 0.08 0.02 0.04 0.06 Residence Time, W/F, g.min/ml 0.08 c-2-C,H 8 t-2-C 4 H„ • 1-butene • isobutylene Figure 6-14. The effects of residence time on the olefin in the total CVs -in stream 30 minutes (Figure A) and 105 minutes (Figure B) - over Ni/K/Al 20 3 catalysts at 350 °C. Reaction at 350 °C, 101 kPa and 10 % CsHe / 90 % CFL, feed gas 89 The increase in C4 olefins other than 1-butene with the residence time could be explained by the secondary isomerization of 1-butene [79]: I-C4FL3 —^  2-C4FL3 I-C4H8 -> iso-C4Hg Alternatively, at longer residence time the C4 may result from dimerization of C 2 fragments, generated from C3FL5 decomposition. Such a reaction sequence would also be consistent with secondary formation of C4's. In the literature, mechanisms of olefin homologation have been proposed by Rodriguez et al. [80] and Leconte et al. [81], which suggested that the product of C3FL5 homologation with M=CH2 is 1-butene and the product of C2H4 dimerization is 2-butene. Since the 1-butene had a higher percentage in the total C 4 products at low residence time, it may be concluded that C3FL5 coupling with M=CH2 was the primary reaction over Ni/K7A1203 catalysts. However, the M=CH2 species involved in the C3FL5 coupling reaction may come from three sources: from CH, decomposition during catalyst reduction, from CH, decomposition during reaction and from C3He decomposition during reaction. To clarify involvement of CH, in the C 4 formation during reaction, a test using He/C3H6 instead CH4/C3H6 was conducted at the same reaction conditions. Figure 6-15 shows the selectivities of the reaction using He/C3H6 as the feed gas. Compared to the C 4 selectivity of CFL, reduced catalyst reacted in CH4/C3FL5 shown in Figure 5-6 B, the C 4 selectivity in Figure 6-15 was almost identical. Given these results, two conclusion can be drawn: 90 1. There was no significant consumption of CH4 during the CH4/C3H6. However CH4 was involved in the reaction through CH X species produced during the CH4 reduction step. 2. The initial C4 formation occurred as a result of reaction between C3H5 and CH X species generated during the CH4 reduction step. Subsequently, C4 formation was a consequence of decomposition and recombination reactions of C3H6 alone. 0 20 40 60 80 100 120 140 T i m e - o n - s t r e a m , min . • C 2 Selectivity • C , Selectivity A C 4 selectivity Figure 6-15. The selectivity changes with time-on-stream over Ni/K/Al203 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % He feed gas These conclusions are consistent with the claim reported by Loftier et al. [22], but contradicts the claims of Ovalles et al. [16]. All the experiment data of the CH4/C3FL5 coupling reactions over Ni/K/Al203 catalysts at 91 different time-on-stream are shown in the Appendix 2. 6.4 Summary The examination of the effects of the reduction time, reaction temperature and residence time on the coupling reaction showed that: (a) If there was no reduction, the catalyst was not active and if the reduction time was too long, too much carbon deposited on the catalyst surface, deactivating the catalyst, (b) When reaction temperature was high, the C3H6 conversion was high. Since the C4 selectivity was highest and the C3H6 conversion was the second highest at 350 °C, the reactions were conducted at 350 °C. C3IL5 conversion followed first order reaction kinetics over the range of conditions studied, (c) The effect of residence time on the coupling reaction was examined in detail. After very short time-on-stream, the C 4 selectivity decreased with residence time, which suggested that the C 4 was a primary product. After longer time-on-stream, the C 4 selectivity increased with residence time, which suggested that C 4 products were not formed by a primary reaction. Additional tests were conducted with He/CsILs as feed gas to determine the involvement of CH4 in the C 4 formation. The results from this test combined with other results from this Chapter and Chapter 5 showed that the C 4 products were formed mainly from the decomposition and recombination reactions of C3H6. The CH X formed during catalyst reduction in CH* was only involved in the initial C 4 formation. After longer time-on-stream and high C3H6 conversion, CH X species generated by C3H6 decomposition dominated the reaction. 92 Chapter 7 Effect of Different Supports on C H J / C S H G Coupling over Ni Catalysts In the literature review, a number of supports were used to support Ni catalysts in the CH4/C3H6 coupling and other related reactions. Since these supports were reported to have the potential to enhance the CH4/C3H6 coupling reaction [16-19,22,82], some of them were chosen for investigation in the present study. In the following sections the CH4/C3H6 reaction conducted with Ni catalysts supported on SiC>2, SAPO-5 and Na-Y zeolite at different reaction conditions is described. 7.1 The Study of Silica Supported Ni Catalysts Unlike AI2O3, SiC>2 does not react easily with Ni to form Ni-SiC>2 compounds. Therefore more Ni is available on the catalyst surface and higher Ni dispersion is possible. Ovalles et al. claimed [16] that high Ni dispersion on the catalyst surface was desired for high C4 selectivity in the CH4/C3H6 coupling reaction. The decrease of Ni/K/Al203 catalyst activity with time-on-stream was mainly caused by the carbon deposition on the catalyst surface. The reduction of carbon deposition was important to keep the catalyst active. It is known that alumina has acidic properties while silica is a very weak acid [83]. One of the purposes of using less acidic silica supported Ni catalysts was to reduce the coking and increase the catalyst activity. The effects of K on the Ni/Si02 were also examined. 93 7.1.1 Activity Tests of Silica Supported Ni Catalysts Three Ni/Si02 catalysts with different Ni loadings and a 7 wt % Ni/Si02 catalyst doped with lwt % K, were used in this part of the study. The Ni dispersions of the Si02 supported catalysts are shown in Table. 7-1. As expected, the lowest Ni loading gave the highest Ni dispersion. However, the K doped catalyst had a significantly reduced dispersion. Note, that the number of Ni sites/g-cat on the 14 wt % Ni/Si02 catalyst was highest, even though this catalyst had the lowest dispersion. All catalysts reported in this section were reduced for 1 hour in CH4 before reaction at 350 °C. Figure 7-1 shows the C3H6 conversions over the Ni/Si02 catalysts and Figure 7-2 shows the relationship between the number of Ni sites per gram of catalyst and the C3FL5 conversion. It is obvious that more Ni sites on the catalyst resulted in a higher conversion of C3H6. It may suggest, however, as shown in Figure 7-1 that if the number of Ni sites is too low, Ni would be rapidly covered by deposited carbon and lose its activity. Table. 7-1 Ni dispersion on Ni/Si02 or Ni/K7Si02 catalysts measured by TPD in H 2 Ni Loading wt% 1 7 7 14 K Loading wt% 1 Ni Dispersion % 16\0 1L6 2^ 5 10.5 Figure 7-3 and Figure 7-4 show the product selectivities. Note that the highest C4 selectivity was obtained after 5 minutes time-on-stream over the 1 wt % Ni/Si02 catalyst but for this catalyst the C4 selectivity decreased sharply with time-on-stream while the C3FL, conversion was lower than 2 %. This result was similar to the results obtained over the Ni/K/Al203 catalyst at 94 100 0 I * T * T * " — - l • r « r 0 20 40 60 80 100 120 140 Tlme-on-stream, mins • 1 wt% NI and 99 wt% S i 0 2 • 7 wl% NI and 93 wt% S i 0 2 * 14 wt% Ni and 86 wt% S i Q 2 Figure 7-1. CsHe conversion over Ni/Si02 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CH4 feed gas. Figure 7-2. C3H6 conversion versus Ni sites in mmole/g-catalyst on the catalyst surface over Ni/Si02 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CFL, feed gas. 95 0 20 40 60 80 100 120 140 T i m e - o n - s t r e a m , m i n s • 1 wt% Ni and 99% S i O , • 7 wt% NI and 93 wt% S i 0 2 14 wt% NI and 86 wt% SI0 2 Figure 7-3. C 4 selectivity over Ni/Si02 catalysts, Reaction at 350 °C, 101 kPa and 10 % C3H5/ 90 % C H 4 feed gas. 300 °C reported in Section 6.2. Hence, the initial high C 4 selectivity is in agreement with the previous proposal that C3FL5 reacted with the CH X species formed during the CH4 reduction to form C4 products. Since the reaction over the 1 wt % Ni/Si02 catalyst did not produce enough active carbonaceous species to sustain the C3FL5 and CH, coupling reaction, the C 4 selectivity decreased rapidly. The initial C4 selectivity over the 1 wt % Ni/Si02 catalyst is taken as additional evidence that CH, participated in the CH4/C3H5 coupling reaction via the CH, catalyst pre-treatment step. Similar to the Ni/K7A1203 catalysts, the C 2 and C3 selectrvities were higher than the C 4 selectivity over the Ni/Si02 catalysts. Figure 7-4 A shows the C3 selectivity and Figure 7-4 B shows the C 2 selectivity over the three Ni/Si02 catalysts. Since there were only few data available 96 on the C 2 and C 3 selectivities over the 1 wt % Ni/Si02 catalyst, the comparison was restricted to the 7 wt % Ni/Si02 and the 14 wt % Ni/Si02 catalysts. Since the C 4 selectivities over the 7 wt % Ni/Si02 and the 14 wt % Ni/Si02 catalysts were similar, the percentage C3EL5 converted to C 2 plus C3 hydrocarbons must also be similar over the two catalysts. As discussed in Chapter 6, the formation of C3 product was due to the hydrogenation of C3FL5, in which the required H 2 was produced from C3EL5 decomposition. C3H6 hydrogenolysis, which yields C2FJLt, C 2 H6 and C H 4 , would also occur, but likely at a lower rate. Hence, as expected, the C3H8 selectivity was greater than the C 2 selectivity for the 14 wt % Ni/Si02 catalyst. However, for the 7 wt % Ni/Si02 catalyst the C 2 selectivity was higher than the C3H8 selectivity, suggesting that the decomposition of C3EL5 occurred according to reactions of the type: C3H6 -» C 2 H, + C + H 2 (1) rather than C 3H5 -> 3C + 3H2 (2) The difference in C3H5 conversion, C 2 and C3 selectivities between the 7 wt % Ni/Si02 and 14 wt % Ni/Si02 is a the result of a greater number of Ni sites on the catalyst surface in the latter case. Hence, more C3EL5 decomposition occurred according to reaction (2) rather (1). 97 0 20 40 60 80 100 120 140 Time-on-stream, mins • 1 wt%Niand99wt%Si02 • 7 wt% Ni and 93 wt% Si0 2 A 14 wt% Ni and 86 wt% Si0 2 Figure 7-4. C3 selectivity (Figure A) and C 2 selectivity (Figure B) over Ni/Si02 catalyst Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CFL, feed gas. 98 7.1.2 The Effect of K on the CH4/C3H6 Coupling Reaction To investigate the promotion effect of K, the reaction of C3H6 with CFI, over Ni/K/Si02 (7 wt % Niand lwt % K) was conducted. Figure 7-5 to Figure 7-8 show the effects of K addition on the C3FI6 conversion and product selectivity compared with the unpromoted 7 wt % Ni/Si02 catalysts. Figure 7-5 shows that the C3H6 conversion was higher over Ni/K/Si02 catalyst than over Ni/Si02 catalyst. Figure 7-7, Figure 7-8 and Figure 7-8 show the C 4, C 3 and C 2 selectivities over Ni/K7Si02 and Ni/Si02 catalyst. The C 4 selectivity decreased with the addition of K to Ni/Si02. The increased C 3 selectivity with the addition of K is a result of the increase in catalyst activity reflected by higher C3FL5 conversion compared to that over the Ni/Si02 catalyst. With the AI2O3 supported catalyst, K addition increased both conversion and C 4 selectivity and this was ascribed to an electronic effect of the K The oleffin adsorption was increased on Ni site when electrons were transfered from K to Ni. The data obtained over the Si02 catalyst showed reduced C 4 selectivity upon K addition, suggesting that the K acts differently, depending on the support. The different behavior is likely due to the acid/base properties of the supports, with K addition neutralizing any acid sites of the Al203. The importance of support acidity is discussed further in section 7.2 and 7.3. 99 Figure 7-5. Comparison of C3FL5 conversion over silica supported Ni catalysts, Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CH4 feed gas. Figure 7-6. Comparison of C 4 selectivity over silica supported Ni catalysts, Reaction at 350 °C, 101 kPa and 10 % C3IL5 / 90 % CFL, feed gas. 100 Figure 7-7. Comparison of C3 selectivity over silica supported Ni catalysts, Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CH4 feed gas. Figure 7-8. Comparison of C 2 selectivity over silica supported Ni catalysts, Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CFL, feed gas. 101 7.1.3 Summary In this section the reaction of CH4/C3FL5 coupling was conducted over Ni/Si02 catalysts with different Ni loadings. High loading of Ni resulted in high C3H5 conversion. The changes in C 2 and C 3 selectivity with Ni loading were discussed in detail. The effects of K addition to Ni/Si02 catalyst were examined. Like the addition of K to Ni/Al2C>3 catalyst, the addition of K to Ni/Si02 catalyst increased the catalyst activity. However, unlike the Ni/K/Al 20 3 catalyst, C4 selectivity was not increased by addition of K 7.2 The Study of SAPO-5 Supported Ni catalysts According to recent literature, the direct methylation of naphthalene with CH4 has been successfully conducted over zeolite supported metal catalysts, based on the idea that hydride abstraction occurred on the metal site of zeolite supported metal catalysts [17]. The methylation of benzene with CH, was also reported over zeolite supported catalysts [18]. In the present study, two types of zeolite (Na-Y and SAPO-5) were used as Ni catalyst supports for the reaction of C3H6 with CH4. The SAPO-5 was prepared according to the procedure described in US Patent 4,440,871 with the composition (Sio.74Alo.i8Po.o8)02 [87]. The characterization of the Ni/Na-Y and Ni/SAPO-5 catalysts were reported previously using TPR in H 2 and the results are summerized in Table 7-2 [87]. Before the catalyst was used in the coupling reaction, a blank experiment with SAPO-5 was carried out to see if the support was active for the reaction. The results from the experiment proved that at the same reaction conditions as used previously for the Ni/Al 20 3 and Ni/Si02 102 catalysts, the SAPO-5 was inactive. Table 7-2 Properties of Ni/SAPO-5 and Ni/Na-Y measured by TPR in H 2 [87] Catalyst Ni content Exchange H 2 consumption NiO reduction |1 mol/g cat % u. mol/g cat % Ni/SAPO-5 Exchanged Ni/SAPO-5 Impregnated Ni/Na-Y 26 438 579 81 3 0 444 81 0 100 14 28 w (a) assuming Ni 2 4 " + H 2 - » N i + 2H* (b) assuming N i 2 * + 1/2 H 2 —» N i + + H * The C3FL5 conversion and product selectivities over the Ni/SAPO-5 catalysts prepared by impregnation and ion-exchange methods are shown in Figure 7-9 to Figure 7-12. Figure 7-9 shows that the activities of both catalysts decreased with the time-on-stream. It is obvious that the ion-exchanged catalyst was significantly better than the impregnated catalyst in terms of both the C3FL5 conversion and C 4 selectivity. The higher C3H6 conversion over the ion-exchanged catalyst compared with the impregnated catalyst could be the result of highly dispersed Ni generated by the ion exchange preparation method. 103 Figure 7-10 shows the C4 selectivities over the ion-exchanged and the impregnated Ni/SAPO-5 catalyst. The C4 selectivities were much higher, especially over ion-exchanged catalyst, than over Ni/K/Al 20 3 and Ni/Si02 catalysts. Figure 7-11 A and B shows the C3 and C 2 selectivities. There were also differences in the C 2 and C3 selectivities between reactions catalyzed by the ion-exchanged catalyst and the impregnated catalyst, of which the latter was similar to the Ni/K7A1203 and Ni/Si02 catalysts. Also, there was 25 mol % Cs+ hydrocarbons in the products over the ion-exchanged catalyst shown in Figure 7-12. These differences in product selectivities suggest that the reaction paths over the ion-exchanged Ni/SAPO-5 catalyst may be different from that over Ni/K/Al203 or Ni/Si02 catalysts. As discussed in Section 2.4, C 4 products may be formed via different reaction pathways, including CIVC3FI6 coupling, C3FL5 metathesis and C2FL» dimerization. As concluded in Chapter 5 and Chapter 6, over the Ni/K/Al203 catalysts the C4 products were formed during the initial stage of the CHJC3H6 coupling reaction; thereafter the C4's were generated via C3FL5 decomposition to form CH X and CFL/C3H6 coupling to form C 4 . C3FL5 metathesis and C2FL, dimerization to produce C 4 were not major reactions over the Al203 supported Ni catalysts. In the case of using the Ni/SAPO-5 catalyst prepared by ion exchange, the C 2 and C3 selectivities, shown in Figure 7-11, were much lower than that obtained over the AUO3 supported Ni catalysts. The low C 2 and C3 selectivities mean that most of the C3FL5 consumed was converted into C 4 and C5 products. If C3FL5 decomposition was the major reaction, the C3 selectivity would be high due to C3FL5 hydrogenation. 104 0 20 40 60 80 100 120 140 T i m e - o n - s t r e a m , m i n s • Impregnation • Ion Exchange Figure 7-9. Conversion of C3FL5 over Ni/SAPO-5 catalysts prepared by ion exchange and impregnation methods. Reaction at 350 °C, 101 kPa and 10 % C3H<; / 90 % CH4 feed gas. 1 o 120 20 40 60 80 100 T i m e - o n - s t r e a m , m i n s • Impregnation • Ion Exchange 140 Figure 7-10. C 4 selectivity over Ni/SAPO-5 catalysts prepared by ion exchange and impregnation methods. Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CFL, feed gas 105 • Impregnation • ion Exchange Figure 7-11. C 3 selectivity (Figure A) and C2 selectivity (Figure B) over Ni/SAPO-5 catalysts prepared by ion exchange and impregnation methods. Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CFLt feed gas 106 50 > w + ID o 45 40 H 35 30 25 20 H 15 10 5 0 20 40 60 80 100 R e a c t i o n T i m e , m i n s 120 140 Figure 7-12. Cs+ selectivity over Ni/SAPO-5 catalysts prepared by ion exchange methods. Reaction at 350 °C, 101 kPa and 10 % CsFfc / 90 % CFL, feed gas The low C 3 selectivity obtained with the ion exchanged catalyst may suggest that C3H6 decomposition was not a major reaction. Furthermore, if C3FLs metathesis was a major reaction, then C 2 selectivity would be as high as C4 selectivity because of the following reaction stoichiometry: 2 C3FLs -> C4FL3 + C2FI4 If the CHX/C3H6 coupling and the C3H<; metathesis were the major reactions, then the low C2 and C 3 selectivities would not be possible. Hence C2H4 dimerization must occur to lower the C 2 selectivity and it may be concluded that the most probable reactions that dominated over the ion exchanged Ni/SAPO-5 catalyst were C3FL5 metathesis and C2H4 dimerization. 107 The C4 component distribution over ion-exchanged Ni/SAPO-5 catalyst with time-on-stream is shown in Figure 7-13 and Figure 7-14. Figure 7-13 shows the paraffin distribution with the time-on-stream There was much more 1SO-C4H10 than n-C4Hi0 in the products and total paraffins was about 30 mol % of total C 4 products. The iso-C4Hi0 decreased and the n-C4Hi0 increased marginally with the time-on-stream Figure 7-14 shows the C 4 olefin distribution with the time-on-stream, which accounts for about 70 mol % of total C 4 products. The possible reaction pathways can be discussed in view of the C 4 component distribution data. When a mole of C3FL5 decomposes, one to three moles of paraffin may be produced via hydrogenation using the H 2 generated from the decomposition. However, if one mole of C3H6 is consumed through the metathesis reaction, no H 2 is formed. Since it was concluded that the C 4 products could not be formed by the reaction of feed CH, with C3H5, the CH X needed for the CHX/C3FL5 coupling was formed by the C3H6 decomposition, which also produced H 2 available for olefin hydrogenation to form paraffin. Therefore, the product paraffin/olefin ratio can be taken indirectly as a indicator of the relative rate of the CHX/C3H6 coupling to the C3H5 metathesis reaction. Since the C 2 selectivity was low (about 3 %) and the C 5 + components were not determined, only the C3 and C 4 product distribution are discussed. Table 7-3 shows the mole % of the paraffin and olefin of C 3 and C 4 products in total products. Since the olefin % was higher than paraffin %, it can be assumed that both the C3H5 metathesis and the CHX/C3H6 coupling occurred and the C3FL3 metathesis reaction rate was higher than the CHx/C3Hs coupling reaction rate. Given that the C 5 + products were formed from C3H6 dimerization and from C 2 H,, C3H6 and C ^ recombination, a error from excluding C 5 + components in the calculation should not invalidate the result. 108 Table 7-3. C3 and C4 product mole % in total products after 60 minutes time-on-stream in the CH4/C3FI6 coupling reaction over Ni/SAPO-5 catalyst at 350 °C and 101 kPa. 10 % CsHg / 90 % CH4 feed gas CsHg iso-C4Hio n-C4H10 I-C4H8 iso-C4H8 C-2-C4H8 t-2-C4H8 Total % % % % % % % % Paraffin <M) 18^ 0 06 27.6 Olefin 4.5 20.0 6.5 9.5 40.5 Note: each C 4 component mole % is calculated by C 4 selectivity x C 4 distribution % in total C 4 . (0 o o I-c c e (0 Q. O 10 40 60 80 100 120 140 160 180 Reaction Time, mins » i s o - C . H . „ • n-C,,H,„ Figure 7-13. The distribution of C 4 paraffins over Ni/SAPO-5 catalysts prepared by ion exchange methods. Reaction at 350 °C, 101 kPa and 10 % CsHe / 90 % CFL, feed gas 109 50 0 20 40 60 80 100 120 140 160 180 Reaction Time, mins * 1-butene • isobutylene A c - 2 - C 4 H B • t -2-C 4 H 8 Figure 7-14. The distribution of C4 olefins over Ni/SAPO-5 catalysts prepared by ion exchange methods. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CH4 feed gas The isobutylene increased with the time-on-stream while the other C4 isomers decreased little. It is known that the olefin isomerization is catalyzed by zeolite catalyst [84]. The 2-C4FL3 formed from the metathesis of C3IL3 could be isomerized into isobutylene over the Ni/SAPO-5 catalyst. The decrease in iso-C4H10 and increase in isobutylene with time-on-stream suggests that the iso-C4Hio was the product of isobutylene hydrogenation. With the catalyst activity decreasing, the catalytic hydrogenation ability decreased and less isobutylene was hydrogenated. That the 1-butene percentage was low showed that C3FL5 coupling with Ni-CHX was not a major reaction over the M-SAP05 catalyst. 110 On the other hand, the impregnated Ni/SAPO-5 catalyst gave higher C 2 and C 3 selectivities than the ion exchanged Ni/SAPO-5 catalyst and higher C4 selectivity than over the Ni/K/Al203 and Ni/Si02 catalysts. This fact shows that the performance of the Ni catalyst was affected by different supports and also affected by the preparation technique. Summarizing the results from this section, tests of CFL/C3FL5 coupling over Ni/SAPO-5 catalysts were conducted. High C4 selectivity (about 60%) was obtained over ion-exchange Ni/SAPO-5 catalyst. However the data suggest that the high C4 selectivity was mainly the result of the metathesis of C 3 H; and the dimerization of C2Ht which was formed by the C3FL5 decomposition and C3H5 metathesis. That the C3FL5 conversion was higher over the ion-exchange Ni/SAPO-5 catalyst than over the impregnated catalyst may be the result of higher Ni dispersion over the ion-exchange Ni/SAPO-5 catalyst than over the impregnated catalyst. The low 1-butene selectivity was indicative of very low extent of the CHx/C3Fi<5 coupling reaction. 7.3 The Study of Na-Y Supported Ni Catalysts The results from tests using Na-Y zeolite supported Ni catalysts are reported in this section. The reactions were conducted with a 90 mol % CH4 and 10 mol % C3H5 feed gas at 350 °C and atmospheric pressure, as before. The results are reported in the following three subsections: preliminary test results that show the effect of changes with time-on-stream, results on the effect of residence time and tests on the effects of acidity changes to the catalyst. For the tests on the effect of the residence time, the flow rate of the reactants was kept constant while the weight of the catalysts used in the reactor was varied. The effects of acidity changes of Ni/Na-Y catalyst on the products were determined by the addition of a base to the Ni/Na-Y catalyst. i l l 7.3.1 Preliminary Tests Results from the reaction of CFL, and C3FL5 over Na-Y supported Ni catalysts are shown in Figure 7-15 and Figure 7-16. The C3FL5 conversion was very high initially but decreased with time-on-stream The C4 selectivity decreased and the C 3 selectivity increased with time-on-stream The selectivity of C2 did not change significantly with the time-on-stream The Cs+ selectivity was about 20 % and increased with time-on-stream As discussed in Section 7-2, the zeolite supported Ni may have promoted metathesis, dimerization and isomerization reactions. The metathesis of C3FL5 produced C4 and C2, the dimerization of C2 produced C4, the dimerization of C3FL3 and recombination of C 2 , C 3 and C4 produced Cs+ components. The rapid decline in C3FL5 conversion suggested that coking, catalyzed by the acidic zeolite, was significant [85]. As the catalyst lost its activity with time-on-stream, the C4 selectivity decreased. Figure 7-15. C3FL5 conversion over Ni/Na-Y catalysts. Reaction at 350 °C,101 kPa and 10 % C3H5 / 90 % CFL, feed gas. 112 70 u -1 1 1 1 1 1 1 1 0 20 40 60 80 100 120 140 T i m e - o n - s t r e a m , m i n s • C 6 Selectivity % • C„ Selectivity % A c3 Selectivity % • C 2 Selectivity % Figure 7-16. C 5 + , C 4 , C 3 and C 2 selectivities over Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CFL, feed gas. The changes in the C4 component distribution with time-on-stream are shown in Figures 7-17 and Figure 7-18. The iso-C4H10, which accounted for 55 % to 90 % of the total C4's, decreased with the time-on-stream The n-C4Hi0, which was less than 10 % of the total C4's, did not change significantly with the time-on-stream Since all C 4 paraffins were the products of C 4 olefin hydrogenation, the decrease in C4 paraffins with time-on-stream is indicative of less H 2 consumption. All C4 olefins (1-butene, isobutylene, c-2-C4H8 and t-2-C4H8) increased with time-on-stream Since the amount of F£2 was limited by the reaction of C3H6 decomposition, the decrease in C3Ff6 conversion with time-on-stream should be responsible for the decrease in olefin hydrogenation. The high percentage formation of iso-GJTio may come from the hydrogenation of isobutylene after dimerization of C2FL, formed via metathesis of C3FLs. When the hydrogenation of 113 Figure 7-17. C 4 paraffins changes over Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CFL, feed gas. 16 -14 -12 -(0 I T o I To 10 - / • "o 1- I II c 8 -/ // # / // m 6 - / // P c o w // / O 4 -2 - ^ ^ ^ ^ ^ Q i i i i i i 0 20 40 60 80 100 120 140 T i m e - o n - s t r e a m , m i n s • 1-butene • isobutylene • t - 2 - C 4 H a A c - 2 - C 4 H 8 Figure 7-18. C 4 olefins changes over Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % C3F£6 / 90 % CFL, feed gas. 114 isobutylene and dimerization of C2H4 decreased, caused by the catalyst activity decreasing with the time-on-stream, the C 4 olefins increased. To prove the assumption that iso-C4Hi0 comes from the dimerization of C2H4, a test was conducted with CH4/C2H4 feed gas instead of CFVC3FL5 over Ni/Na-Y catalyst. The results of the reaction are shown in Figure 7-19 to Figure 7-22. Since the C2FI4 was more reactive than C3FL5, the conversion was high even though the reaction was conducted at 250 °C and 101 kPa. The conversion declined with time-on-stream as shown in Figure 7-19. Figure 7-20 shows the selectivities of all products. The highest selectivity was for C4's, which confirms that the dimerization of C2FL, was the main reaction over Ni/Na-Y catalysts. 100 c o 'S2 01 > c o O 0) c <D •5. & 20 40 60 80 100 T i m e - o n - s t r e a m , m i n s 120 140 Figure 7-19. The conversion of C2FL, over Ni/Na-Y catalyst. Reaction at 250 °C and 101 kPa and 10 % C2H4 / 90 % CFL, feed gas. 115 — I 1 1 1 1 — 20 40 60 80 100 T i m e - o n - s t r e a m , m i n s 120 C , Selectivity C 4 Selectivity • C 3 Selectivity • C 5 * Selectivity 140 Figure 7-20. The selectivities of the products over Ni/Na-Y catalyst, Reaction at 250 °C and 101 kPa and 10 % C2FI4 / 90 % CFL, feed gas. o* s a c M C 1 40 60 80 100 T i m e - o n - s t r e a m , min 140 Figure 7-21. The C 4 paraffins in total C 4 over Ni/Na-Y catalyst, Reaction at 250 °C and 101 kPa and 10 % C2H4 / 90 % CFL, feed gas. 116 Figure 7-22. The C 4 olefins in total C 4 over Ni/Na-Y catalyst, Reaction at 250 °C and 101 kPa and 10 % C2H4 / 90 % CFLt feed gas. C2H4 + C2H4 —> C4Hg The Cs* selectivity was also high showing that the olefin condensation catalyzed by acid was the major reaction. The C3 and C2 selectrvities were similar to the reaction with the feed of C3H6/CH1. Figure 7-21 and Figure 7-22 show the C 4 distribution in total C 4 products. The high percentage of iso-OF^ in the total C 4 rather than 1-butene further confirmed that C 4 products in the CH4/C3FL3 coupling reaction come from C2FLt dimerization followed by isomerization. 117 7.3.2 The Effects of Residence Time on the Coupling Reaction Figure 7-23 shows the change in C3FL5 conversion with residence time after 30 and 105 minutes time-on-stream for the Ni-NaY catalyst. The catalyst activity declined with the time-on-stream for all reactions. Note that when the residence time (W/F) was greater than 0.04 at 30 minutes time-on-stream, the C3H6 conversion approached 100 % and therefore data beyond this point (W/F = 0.04) will not be discussed. R e s i d e n c e t i m e , W / F , g . m i n / m l 30 mins Time-on-stream • 10S mins Time-on-stream Figure 7-23. The effects of residence time changes on the C3EL conversion over Ni/Na-Y catalyst at 30 and 105 rninutes time-on-stream. Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CH4 feed gas. Figure 7-24 and Figure 7-25 show selectivity as a function of residence time. Both the C 3 and C4 selectivity increased with the residence time whereas the C 2 selectivity decreased slightly with residence time. Similar to the reaction over ion exchanged Ni/SAPO-5 catalyst, the Cs+ selectivity over Ni/Na-Y catalyst was much higher than that over Ni/K/Al 20 3 catalyst. These 118 trends in selectivity are quite different from those observed over the K/Ni/Al203 catalyst pointing to the occurrence of different reaction pathways. As discussed earlier, over the Ni-zeolite catalysts, the C4 hydrocarbons were formed via C3H6 metathesis and the C2H4 dimerization. Because of these reactions the C4 selectivity was much higher over Ni/zeolite catalyst than over the Ni/K7A1203 or Ni/Si02 catalysts. Most importantly, with increasing residence time the C4 selectivity increased over the Ni-NaY catalyst pointing to C4 as a secondary product. so 1 45 -40 -35 -m * * > ectivity 30 -25 -Sel 20 -0 15 -10 -5 -0 - I I I I I I I 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 Residence Time, W/F, g.min/ml • 30 mins Time-on-stream • 105 mins Time-on-stream Figure 7-24. The effects of residence time changes on the C 4 selectivity over Ni/Na-Y catalyst at 30 and 105 minutes time-on-stream. Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CFL feed gas. 119 50 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 R e s i d e n c e T i m e , W / F , g . m i n / m l • C 2 selectivity • c 3 selectivity * C 6 * selectivity Figure 7-25. The effects of residence time changes on the C2, C3 and C 5 selectivity over Ni/Na-Y catalyst at 30 minutes time-on-stream Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CFL, feed gas. The hydrogenation of C3FL5, which was responsible for the C3 selectivity, could occur only when there was H 2 formed via C3FL5 decomposition. With the increase of residence time, more C3FL3 was decomposed and more FL2 was available, which increased the C3 selectivity. The C2FL, dimerization increased with the residence time and therefore, the C 2 selectivity decreased with residence time. The formation of Cs+ products could occur from primary reactions (C3FL5 dimerization) and secondary reactions (C2FL, and C3FLs coupling, C4 and C5 coupling with CFL,). The different C4 components formed in the reaction over Ni/Na-Y also changed with the residence time. Figure 7-26 shows the C4 paraffin distribution and Figure 7-27 the C4 olefin distribution with the residence time at 30 minute time-on-stream In Figure 7-26 both iso-C4Hi0 120 and n-C4Hio increased with the residence time, and iso-C^o increased more than n-CiHio. The sharp increase in iso-GiHio with residence time confirmed the statement that the formation of the iso-C+Hio was the result of a secondary reaction. Figure 7-27 shows that all C4 olefins decreased with residence time. The decrease in C4 olefins with residence time indicates that they were primary products, unlike the trends with K/Ni/Al 20 3 catalysts where only 1-butene was identified as a primary product. 100 o "5 o f— c CL 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 R e s i d e n c e T i m e , W / F , g . m i n / m l i s o - C , H 1 0 n - C 4 H 1 0 Figure 7-26. The effects of residence time changes on the C 4 paraffins over Ni/Na-Y catalyst at 30 minutes time-on-stream. Reaction at 350 °C, 101 kPa and 10 % C3H5 /90 % CFL feed gas. 121 30 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 Residence Time, W/F, g.min/ml • 1-butene • isobutylene • t -2-C 4 H 8 • c -2 -C 4 H 8 Figure 7-27. The effects of residence time changes on the C 4 olefins over Ni/Na-Y catalyst at 30 rninutes time-on-stream. Reaction at 350 °C, 101 kPa and 10 % CjH* / 90 % CFL, feed gas. 7.3.3 The Acidity Changes of Ni/Na-Y Catalysts on Coupling Reaction The acidic property of Na-Y may be responsible for the C3FL5 conversion over Ni/Na-Y being much higher than the C3FL5 conversion over alumina, silica or SAPO-5 supported Ni catalysts. To confirm this statement a base, in the form of K 2 0 was added to the Ni/Na-Y catalyst. The results of the reaction over base doped Ni/Na-Y are shown in Figure 7-28 to Figure 7-31 with the comparison of the results from the reaction over undoped Ni/Na-Y. The addition of the base caused a sharp decline of the C3FL5 conversion, a fast decrease in C 4 selectivity and almost complete elimination of C5 selectivity, which was zero except for 3.2 % 122 after 5 minutes time-on-stream. The C3 and C2 selectrvities increased over the base doped Ni/Na-Y catalyst. All of these observations meant that once the acidic property was removed, the catalyst activity decreased and the product distribution changed significantly. The once dominant reactions such as C3H6 metathesis and C2H4 dimerization catalyzed by acid catalyst decreased sharply. As a result, the C 4 and C5 selectrvities were low. The C3FL5 decomposition and C3Ff6 coupling with CH4 were the major reactions resulting in the high C 3 and C 2 selectrvities compared with that over Ni/Na-Y catalyst. As far as the C4 distribution in total C4 hydrocarbons was concerned, the difference of performance by the addition of a base on the Ni/Na-Y was more apparent and is shown in Figure 7-32. The majority of C4's from the reaction over K20-Ni/Na-Y was C 4 olefin while C 4 paraffin formed the majority of the total C 4 from the reaction over Ni/Na-Y. On the K 2 C0 3 -Ni/Na-Y catalyst, the C3H3 metathesis occurred, which yielded 2-butene, and the isomerization catalyzed by acid was low reflected by the low isobutylene formation. Furthermore, the presence of 1-butene is indicative of CFL/C3FL5 coupling on this catalyst. 123 0 20 40 60 80 100 120 T i m e - o n - s t r e a m , m i n s Ni /Na-Y • N i / K 2 C 0 3 / N a - Y 140 Figure 7-28. The effects of a addition of a base on Ni/Na-Y catalyst on the C3FL5 conversion. Reaction at 350 °C, 101 kPa and 10 % C3EL5 / 90 % CH4 feed gas Figure 7-29. The effects of a addition of a base on Ni/Na-Y catalyst on C4 selectivity. Reaction at 350 °C, 101 kPa and 10 % CjHe / 90 % CH4 feed gas 124 Figure 7-30. The effects of a addition of a base on Ni/Na-Y catalyst on C 3 selectivity. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CFLt feed gas Figure 7-31. The effects of addition of a base on Ni/Na-Y catalyst on C 2 selectivity. Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CFLt feed gas. 125 100 iso-C4H10 1-butene n-C4H10 isobutylene C-2-C4H8 t-2-C4H8 HH K2C03-Ni/Na-Y i l l Ni/Na-Y Figure 7-32. The effects of a base addition into Ni/Na-Y on the C4 distribution at 30 minutes time-on-stream. Reaction at 350 °C, 101 kPa and 10 % C3FL; / 90 % CH4 feed gas In summary, the acidic site on the Ni/Na-Y catalyst played a major role in the catalytic reaction of CFL/C3FL3 and Ni/Na-Y catalyzed a series of reactions that led to undesired products. The dimerization of C2H4 was the key reaction in which C4 was formed. On the other hand when the acidic site was neutralized by the addition of a base, the reaction gave more C4 olefins in which the linear olefins (1-butene and 2-butene) were significant. As discussed previously, the 2-butene was a result of C3FL5 metathesis and the 1-butene a result of the reaction: CFf3CH=CH2 + M-CFf2-» CFf3CH2CH=CH2 + M Since the C4 selectivity over K20-Ni/Na-Y catalyst decreased to zero in a short time-on-stream, it can be said that the CFL. mostly came from the surface carbonaceous species formed during the reduction process. 126 In this section CH4/C3H6 coupling reactions were conducted over the Ni/Na-Y catalysts under different reaction conditions. Both the C3H5 conversion and C4 selectivity increased with the residence time. The major reactions were C3FL5 metathesis and C2H4 dimerization. The detailed product distribution was discussed. With the addition of a base to the catalyst, the catalyst activity decreased sharply. It was concluded that the acidity of the Na-Y played a important role in the CH4/C3H5 coupling reactions over Ni/Na-Y catalysts and these reactions did not involve CFL nor the CH X surface species generated during reduction . 7.4 Effects of Supports on the C3H6 / CH4 Coupling Reaction So far the activities of Ni catalysts with four different supports have been tested in the coupling reaction of C3IL3 with CH4. To compare the effects of different supports on the Ni catalyst performance, four catalysts were chosen which were the best in each support group in terms of C 4 yield. The catalysts were Ni/K/Al 20 3 (7 wt % Ni, 1 wt % K and 92 wt % A1203), Ni/SiC-2 (7 wt % Ni and 93 wt % Si02), Ni/SAPO-5 and Ni/Na-Y (both prepared by ion exchange). Figure 7-33 to Figure 7-37 show the C3FL5 conversions and product selectrvities over these four catalysts. Since Cs+ products were not formed over Ni/Si02 catalyst, Ni/Si02 catalyst was not included in Figure 7-37. 127 2 0 4 0 6 0 80 T i m e - o n - s t r e a m , 1 — 1 0 0 m i n s 1 2 0 1 4 0 N i / S A P O - 5 • N i / S i O , • N i /Na-Y Figure 7-33. C3H3 conversion over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % CsHe / 90 % CFL feed gas 2 0 4 0 6 0 80 1 0 0 T i m e - o n - s t r e a m , m i n s 120 1 4 0 N l / K / A l 2 0 3 N i / S A P O - 5 N i / S i 0 2 Ni /Na-Y Figure 7-34. C 4 selectivity % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CH4 feed gas 128 80 > 70 60 50 40 0) a> W 30 CO o 20 H 10 0 —I 1 1 1 1 — 20 40 60 80 100 T i m e - o n - s t r e a m , m i n s 120 140 N i / S A P O - 5 N i / S i 0 2 Ni /Na-Y Figure 7-35. C3 selectivity % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CFL feed gas 80 70 -60 -50 -4-* ts 40 -<D to 30 -o N 20 -10 -20 40 60 80 100 T i m e - o n - s t r e a m , m i n s 120 140 N i / S A P O - 5 • N i / S i 0 2 • Ni /Na-Y Figure 7-36. C 2 selectivity % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % Cl^feed gas 129 20 40 60 80 100 Time-on-st ream, mins A Ni/SAPO-5 • 120 140 Ni/Na-Y Figure 7-37. Cs+ selectivity % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % CjHe / 90 % CH4 feed gas The C3H6 conversion shown in Figure 7-34 decreased with time-on-stream over all Ni catalysts with different supports and the catalysts activity for the CjHe conversion had the order: Ni/Na-Y > Ni/K/Al 20 3 > Ni/SAPO-5 > Ni/Si02 It should be noted that the high C3FL5 conversion over Ni/Na-Y was caused by undesired reactions (C3FL5 metathesis and C2H4 dimerization). Figure 7-34 shows that the C 4 selectivities were almost the same over Ni/Al 20 3 and Ni/Si02 catalysts. The C 4 selectivities were much higher over zeolite supported Ni catalysts than non-zeolite supported Ni catalysts, in which Ni/SAPO-5 gave higher C4 selectivities than Ni/Na-Y catalyst. The C 4 selectivities were in the following order: Ni/SAPO-5 > Ni/Na-Y > Ni/K/Al 20 3 = Ni/Si02 130 The difference in C4 selectivity between the zeolite and non zeolite supported Ni catalysts was caused by the different reactions occurring on the catalyst. For the zeolite supported Ni catalysts, C3H6 metathesis followed by the C2FI4 dimerization were the main reactions, which gave high C4 selectivity. There were also significant Cs+ products from the reactions over zeolite supported Ni catalysts due to dimerization. For non zeolite supported Ni catalysts, the C3FL5 decomposition was the major reaction, therefore, the C2 and C3 selectivity were high as shown in Figure 7-35 and Figure 7-36. Figure 7-38 shows a comparison of the C 4 yields from the reactions over Ni catalysts supported with different supports. The C4 yields were in the following order: Ni/Na-Y > Ni/SAPO-5 > Ni/K/Al 20 3 > Ni/Si02 160 T i m e - o n - s t r e a m , mins I I N l / K / A l 2 0 3 mmS NI/SIOj N I / S A P O - 5 W!2m N l /Na-Y Figure 7-38. Yield of C 4 % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CFL, feed gas 131 It is clear that the zeolite type supports (SAPO-5 and Na-Y) and the non-zeolite type supports (AI2O3 and Si02) played much different role in the reactions, which was reflected in the product distribution. The main difference in the product distribution was that the zeolite supported catalysts gave high percentage of C4 and Cs+ components that resulted from the metathesis of C3H6, oligomerization, condensation and isomerization. The dimerization of C2H4 resulted in the high C 4 and low C2 selectrvities and oligomerization and condensation of C 2 and C 3 olefins formed the C 5 + components with 15 to 30 % selectivity for both SAPO-5 and Na-Y supported Ni catalysts. On the other hand, the A1 20 3 and Si02 supported Ni catalysts gave low C4 selectivity (about 10 %) and high C2 and C3 selectivity. Since the acidity of the A1 2 0 3 and Si0 2 supported Ni catalysts was relatively low, the major reactions were catalyzed by the Ni rather than by the acid catalyst. The CH4 and C3II5 coupling, the C3H6 decomposition and C3H6 hydrogenation reactions were responsible for the product distribution. It could be concluded that the reaction mechanisms were very different between with the zeolite supported Ni catalysts and the non-zeolite supported Ni catalysts. There were also differences between the zeolite supported Ni catalysts in the product distributions. The C 4 selectivity was lower and C3 selectivity was higher over Ni/Na-Y catalyst than over the Ni/SAPO-5 catalyst. This was the result of a higher rate of C3H6 decomposition followed by the hydrogenation of C3H5, over the Ni/Na-Y catalyst than over the Ni/SAPO-5 catalyst. There were also differences between the Ni/K/Al203 catalyst and the Ni/Si02 catalyst in the selectrvities. In spite of these differences in the results over the zeolite supported Ni catalysts and in the results over the non-zeolite supported Ni catalysts, two series of reactions can be identified as discussed above. 132 To compare further the differences of the two kind of supports, the two catalysts were chosen to represent the zeolite supported Ni catalyst group (Ni/Na-Y) and non-zeolite supported Ni catalyst group (Ni/K/Al203). Figure 7-39 to Figure 7-41 show the comparison of different C4 components over Ni/K7A1203 and Ni/Na-Y catalysts. The main C 4 products over the Ni/Na-Y catalyst were the paraffins in which iso-C4Hio took much more percentage than n-C4Hi0. In contrast the reaction over the Ni/K/Al203 catalyst gave more olefins than paraffins. Figure 7-42 shows the C 4 paraffin to olefin ratio over Ni/K/Al203 and Ni/Na-Y catalysts. Although the ratio of C 4 paraffin to C 4 olefin was much higher over the Ni/Na-Y catalysts than over the Ni/K/Al2C>3 catalysts, it was the same that the ratio of C 4 paraffin to C 4 olefin over both Ni/K/Al203 and Ni/Na-Y catalysts decreased with the time-on-stream This means that more C 4 olefin and less C 4 paraffin were formed as the catalyst activity decreased with the time-on-stream, which once again confirmed the reaction sequence: C 4 olefin => C 4 paraffin Since hydrogenation of olefin is catalyzed by metals, the sharp decrease of the paraffin/olefin ratio over Ni/Na-Y catalyst meant the rapid deactivation of Ni catalyst. To summarize the effects of the supports on the coupling reaction with CFL,, the zeolite supported Ni catalysts can catalyze more deep and complex reactions, while the Al203 and Si02 supported Ni catalysts involved relative simple reactions. If only the C 4 selectivity was taken as the criteria, the performances of the catalysts were as follows: Ni/SAPO-5 > Ni/Na-Y » Ni/K/Al 20 3 = Ni/Si02 133 100 80 « 60 o x~ o 6 40 20 Figure A (0 o I-c X O i c 20 40 60 80 100 Time-on-stream, mins • Ni /K/Al 2 0 3 • Ni/Na-Y 120 140 140 Figure 7-39. iso-C^io (Figure A) and n-C^io (Figure B) % in total C4's over Ni/K7A1203 and Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CFL feed gas 134 40 35 in 30 O otal 25 H c c* 20 4) C 4> 15 isobu 10 I 40 60 80 100 Time-on-stream, mins 140 Ni/K/Al203 Ni/Na-Y Figure 7-40. 1-butene (Figure A) and isobutylene (Figure B) % in total C4's over Ni/K/Al 20 3 and Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CFL, feed gas 135 40 in O i 35 -30 -25 .£ 20 SO o O 10 5 0 Figure A 20 40 60 i 80 100 120 140 40 60 80 100 Time-on-stream, mins 140 Ni /K/Al 2 0 3 Ni/Na-Y Figure 7-41. 02-C4H8 (Figure A) and t-2-C4Hg (Figure B) % in total C4's over Ni/K/Al 20 3 and Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CFL feed gas 136 0 20 40 60 80 100 120 140 Time-on-stream, mins Figure 7-42. The ratio of C4 paraffin to C4 olefin of over Ni/Na-Y (Figure A) and over Ni/K/Al 20 3 (Figure B). Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CFL feed gas But when the C3FL5-CFL, coupling reaction was considered, the C4 products were formed mainly from C3FL itself over zeolite supported catalysts. In this sense, the catalysts, which promote the simple reactions, should be easier to be improved to reduce the side reactions and focus to the C3H6-CH4 coupling reaction. In this section 7.4, the results of CH4/C3FL5 coupling reaction over Ni catalysts on different supports were compared. The zeolite supported Ni catalysts gave higher C4 yields than the non-zeolite catalysts. The reactions that occurred over different catalysts were discussed. It can be concluded that the Ni catalysts supported by the zeolite (SAPO-5 and Na-Y) catalyzed mainly the reactions of metathesis of C3Fl6, dimerization and condensation, and the Ni catalysts supported by 137 the non-zeolite (Si02 and AI2O3) catalyzed mainly the reactions of CH4/C3H6 coupling, C3FL5 deconnjosition and C3H6 hydrogenation. 138 Chapter 8 8. Conclusions and Future work In the present study, the CH4/C3H5 coupling reaction over supported Ni catalysts was investigated as a potential new approach to CFL upgrading. The study emphasized the effects of process variables, catalyst supports and promoters on the CH4/C3H6 coupling reaction and resulting product distribution. In the following sections, a summary and conclusions from the present study are given followed by proposals for future work. 8.1 Summary of Results from the Present Study A preliminary series of CFL/C3FL coupling experiments over Ni/Al203 catalysts with and without the addition of P or K, showed that the addition of P and K addition increased the catalyst activity and C4 selectivity. The maximum C4 selectivity achieved over the Ni/K/Al203 catalyst was 10 mol % at 350 °C and 101 kPa with a feed gas composition of 90 mol % CFL and 10 mol % C3H5. The catalyst activity decreased with time-on-stream as a result of carbon deposition on the catalyst surface. TPSR was used to examine these surface carbon species. Two types of surface carbon species were identified; active carbon which reacted with H 2 at low temperature (192 °C to 237°C) and inactive carbon which reacted with H 2 at high temperature (547 °C to 667°C). A correlation was established between the amount of active carbon species generated during the coupling reaction and the catalyst performance (C3H5 conversion and C 4 selectivity). A comparison of the results from CH4/C3H5 coupling over H 2 reduced catalyst and CFL reduced 139 catalyst, led to a carbon deposition model being proposed to explain the changes in catalyst activity. To better understand the CH4/C3H6 coupling reaction, the effect of different reaction conditions, namely catalyst reduction temperature, reaction temperature and reactant residence time were investigated. The C4 product distribution with time-on-stream and residence time were also examined in detail. Evidence from these data showed that the C3FL5 coupling reaction occurred via the reaction of CH X with C3H6 to form 1-butene. The CH X was formed either from catalyst reduction in CH4, or from C3FL5 decomposition during the CH4/C3FI6 coupling reaction. C3FL5 hydrogenation also occurred which resulted in high C3 selectivity. An examination of the effect of catalyst supports was carried out using Ni/Si02, Ni/SAPO-5 and Ni/Na-Y catalysts for the CH4/C3FL3 coupling reaction. In some cases the Ni loading and the methods of catalyst preparation were also varied. The high C4 selectivity over zeolite supported Ni catalysts was a consequence of C3FL5 metathesis and olefin dimerization. 8.2 Conclusions From the results of the CH4/C3FL5 coupling reaction over supported Ni catalysts, the following conclusions can be drawn: 1. The CH4/C3FL5 coupling reaction occurred when CH X reacted with CsHeto form 1-butene. CFL, was only involved in C4 formation as CH X species, generated during the catalyst reduction process. Louring the CH4/C3FL5 coupling reaction, CFL, did not react with C3H6 to form C 4 2. The catalyst surface carbon species consist of two types in terms of activity. The amount 140 of active carbon species, generated during the CH4/C3I-L5 coupling reaction, can be correlated with the catalyst performance (C3FL5 conversion and C4 selectivity). 3. The major reactions in the CH4/C3H6 coupling reaction over non-zeolite supported Ni catalyst were decomposition of C3H5, hydrogenation of C3H6 and CHJCjHe coupling. The major reactions over zeolite supported Ni catalysts were metathesis of C3IL3 and dimerization of olefins. 4. Since the maximum C4 selectivity of 10 mol % obtained using Ni/K7A1203 catalysts and since CH4 did not react with C3H3 directly, it is concluded that a multi-step reaction will be needed if this approach is to be utilized for direct CFL upgrading. 8.3 The Contribution of this Work The contribution of the present study consisted of three parts: 1. The desired CH4/C3H6 coupling reaction has been shown to occur at the initial stage of the reaction from CH X and C3H5. The CH X species were formed during the catalyst reduction in CH4. During reaction, CH4 was not involved in the CH4/C3II5 coupling reaction. 2. A correlation between the type and amount of surface carbon and the conversion and C4 selectivity of the CH4/C3FL3 coupling reaction over Ni catalysts has been established. The fiinction of the surface carbon in the CH4/C3H6 coupling reaction was discussed in detail. 3. The investigation of CH4/C3FL5 coupling reaction over Ni catalysts with different 141 supports and promoters, over a range of reaction conditions, identified the different reactions that occurred at the conditions of the present study. The desired CH4/C3H6 coupling reaction was shown to occur at a much lower level than these undesired reactions. All the above three parts are unknown in the literature. The significance of the present work is that it demonstrates that that CFL, can be activated at high temperature and the resulting CH X surface species can be reacted with C3H6 at low temperature. The main contribution of this study is that its points to the potential of a cycling process in which reactions occur at different temperatures; a high temperature to activate CH4 and lower temperature to couple the CH X species with C3FL5 to form higher hydrocarbons. 8.4 Future Work The desired CH4/C3FL5 coupling reaction occurred at the initial stage of the reaction following catalyst reduction in CH4. To improve the CFL/C3FL5 coupling process, three stage process may increase the C4 selectivity and decrease the C3H6 consumption. In the first stage, the catalyst is reduced in CH4 and the surface carbon is formed. In the second stage, the C3FL5 is inputted into the reactor and the CFL/C3FL5 coupling occurs. In the third stage, the used catalyst is reactivated, which means the inactive carbon is removed from the catalyst surface and the catalyst is reduced in CFL, again. Research on this three stage cycling process should include the determination of the optimum amount of surface carbon on the catalysts, the determination of the duration of the coupling stage and the determination of the condition to reactivate the used catalyst. 142 References 1. IGT World Reserves Survey, Institute of Gas Technology, Chicago, IL, (1989). 2. St-Just, J., Basset, J. M., Bousquet, J. and Martin, G. A., La Recherche 21, (222), 730-738 (1990) 3. Leprince, P., La Recherche 19, (199), 664-665 (1998) 4. Bossard, A. R. and Back, M. H., Can. J. Chem,. 68, 1401-1407, (1990) 5. Anderson, J. R, Appl. CataL, 47, 177-196, (1998) 6. Le Bee, R_, Marquaire, P. M. and Come, G. M., Ind. Eng. Chem Res., 29, 39-544, (1990) 7. Poirier, M.G., Sanger, A.R. and Smith, K.J., Can. J. Chem Eng., 69, 1027-1035,(1991) 8. Keller, G. E., and Bhasin, M. M., J. Catal., 73, 9-19, (1982) 9. Chan, T. and Smith, K. J, Appl. Catal., 60, 13-31, (1990) 10. Geerts, J. W. M. H., Chen, Q., Kasteren, J. M. N. van and Wiele, K van der, Catal. Today, 6,519-526,(1990) 11. Sofranko, J. A , Leonard, J. J. and Jones, C. A., J. Catal., 103, 302-310, (1987) 12. Smith, K J., Painter, T. M. and Galuszka, J., Catal. Letters, 11, 301-308, (1991 ) 13. Choudhary, V. R and Rane, V. H., J. Catal., 130, 411-422, (1991) 14. Zanthoff, H., Zhang, Z., Grzybek, T., Lehmann, L. and Baerns, M. Catal. Today 13, 469-480, (1992) 15. Fox HL J. M., Catal. Rev.-Sci. Eng., 35(2), 169-212, (1993) 16. Ovalles, C , Leon, V., Reyes, S. and Rosa, R, J. Catal., 129, 368-382, (1991) 17. Wang, L., Tao, L., Xie, T., Xu, G , Huang, J., and Xu, Y., Catal. Lett., 21, 35-41 (1991) 143 18. He, S. J. X. and M. A. Long, Energy & Fuels, 9, 616-619, (1995) 19. He, S. J. X., Long, M. A , Attalla, M. I., Wilson, M. A, Energy and Fuels, 6(4), 498-503, (1992) 20. Fox HI, J. M., T. P. Chen and B. D. Degen, Chem Eng. Prog., 42-50, (1990) 21. GuczL L., R. A. Van Santen and K V. Sarma, Catal. Rev.-Sci. Eng., 38(2), 249-296, (1996) 22. Loffler, I. D., Maier, W. F., Andrade, G. J., Thies, I. and Schleyer, P. von K , J. Chem. Soc, Chem. Commun., 1177-1179, (1984) 23. Koerts, T. and Santen, van R. A , J. Chem. Soc., Chem Commun., 345-346, (1992) 24. Vasant, R. C , Kinage, A. K and Choudhary, T. V., Science, 275, 1286-1288, (1997) 25. Tanaka, K-i., Yaegashi, I. and Aomura, K , J. Chem Soc, Chem Commu., 938-940, (1982) 26. Sodesawa, T., Matsubara, M., Satoshi, S. and Nozaki, F., Chem Letters, 1513-1514, (1987) 27. Khcheyan, K E., Revenko, O. M. and Shatalova, A. N., Proc. Xlth World Petrol. Congr., 4, (1983) 28. Olah, G. A , Felberg, J. D and Lamtnertsma, K , J. Am Chem. Soc, 105, 6529-6531, (1983) 29. Siskin, M., J. Am. Chem Soc, 98, 5413-5415, (1976) 30. Moffat, J.B., "Theoretical Aspects of Heterogeneous Catalysis", Van Nostrand Reinhold, New York, (1989) 31. Anderson, J. R. and Boudart, M., "Catalysis", 1, Springer-Verlag Berlin, Heidelberg, New York, (1980) 32. Delmon, B and Froment G. F., "Catalyst Deactivation", Elsevier Scientific Pubhsbing Company (1987) 144 33. He, Hong, Nakamura, J. and Tanaka, K-L, Catalysis Letters, 16, 407-412, (1992) 34. Anderson, A. B., Malonet, J. J. and Yu, J., J. Catal., 112, 392-400, (1988) 35. Anderson, A. B. and Yu, J., J. Catal., 119, 135-145, (1989) 36. Kiselev, V. F. and Krylov, O. V., "Adsorption and Catalysis on Transition Metals and their Oxides", Springer Verlag (1991) 37. Jolly, W. L., "Modern Inorganic Chemistry", McGram-Hill Book Company, (1984) 38. Jacobsen, K W. andNorskov, J. K , Surface Science, 166, 539-553, (1986) 39. BonzeL H. P. and Krebs, H. J., Surface Science, 91, 499-513, (1980) 40. Coad, J.P. and Riviere, J. C , Surface Science, 25, 609-624, (1971) 41. Vogt, E. T. C , Dillen, A. J. van and Geus, J. W., "Catalyst Deactivation", Elsevier Scientific Pubhshing Company, (1987) 42. Eloy, R., M. Leconte, J. Am. Chem Soc. 110, 275-278 (1988) 43. Rofer-DePoorter, C. K , Chem Rev., 81, 447-474, (1981) 44. Anderson, R. B., "Heterogeneous Catalysis", McMaster University 3rd edition, (1980) 45. Anderson, R. B, 'Tischer-Tropsch Synthesis", Academic Press, (1981) 46. Imelik, B., "Metal-Support and Metal-Additive Effects in Catalysis", Elsevier Scientific Pubhshing Company, (1982) 47. Tanaka, K. "Olefin Metathesis and Polymerization Catalysis", 271-301, Kluwer Academic Publishers, (1990) 48. Pines, H., "The Chemistry of Catalytic Hydrocarbon Conversions", Academic Press (1981) 49. Rylander, P. N., "Hydrogenation Methods", McGraw-Hill Book Company (1985) 50. Pines, H , "The Chemistry of Catalytic Hydrocarbon Conversion", Academic Press, (1981) 145 51. Imelik, B., "Metal-Support and Metal-Additive Effects in Catalysis", Elsevier Scientific Publishing Company, (1982) 52. Benndorf C , Nieber, B. and Krueger, B., Surface Science, 177, L907-914, (1986) 53. Fierro, J. G. andBanda, J. G., Catal. Rev.-Sci. Eng., 28(2-3), 265-333, (1986) 54. Bradley, S. A., Gattuso, M. J. and Bertolacini, R. J., "Characterization and Catalyst Development", Chapter 11., American Chemical Society, (1989) 55. McCarty, J.G. and Wise, F£., J. Catal. 57 406-416 (1979) 56. Brunauer, S., Emmett, P. H. and Teller, E., J. Am Chem Soc, 60, 309-312, (1938) 57. Emmett, P. H., Adv. Catal., 1, 65-68, (1948) 58. Emmett, P. H., "Catalysis", vol. 1, Reinhold , New York, (1954) 59. Satterfield, C. N., "Heterogeneous Catalysis in Industrial Practice", 2nd. ed., Mcgraw -HilL Inc., New York, (1991) 60. Busca, G., Cavani, R, Centi, G. and Trifiro, R, J. Catal., 99, 400-414, (1986) 61. Guo, I.,Yu, T-T., Wanke, S. E. and Ward, J. W. (Editor), "Catalysis", Springer Verlag, (1988) 62. Sachdev, A. and Schwank, J., Proc. Int. Cogr. Catal. 9th, 3, 1275-1282, (1988) 63. Burch, R. J., Catal., 71, 348-359, (1981) 64. Sexton, B. A, Hughes, A. E. and Foger, K , J. Catal., 88, 466-477, (1984) 65. Clause, O., Bonneviot, L., Che, M. and Dexpert, H , J. Catal., 130, 21-28, (1991) 66. Kadkhodayan, A. and Brenner, A , J. Catal., 117, 311-321, (1989) 67. Forni, L., Toscano, M. and PolleseL P., J. Catal., 130, 392-402, (1991) 68. Forni, L., Toscano, M. and PolleseL P., J. Catal., 130, 403-410, (1991) 146 69. Muhler, M., Nielsen, L. P., Toernqvist, E., Clausen, B. S. and Topsoe, H., Catal Letters, 14, 241-249, (1992) 70. Chen, I. and Chen, F-L., Ing. Eng. Chem. Res., 29, 550-554 (1990) 71. Martin, G. A. "Metal-Support and Metal-Additive Effects in Catalysis", Elsevier Scientific Publishing Company, (1982) 72. Lewis, J. M and Kydd, R. A , J. Catal. 132, 465-471 (1991) 73. Fogler, H.S., "Elements of Chemical Reaction Engineering", Second Edition, Prentice HalL 1992 74. Liu, Q. and Smith, J. K., Can. J. Chem. Eng., 73, 337-344, (1995) 75. Koerts, T and Van Santen, R.A., "Catal. Letters, 6, 49-58, (1990) 76. Vogt, E. T. C , Van Dillen, A. J. and Geus, J. W., "Catalyst Deactivation", Elsevier Scientific PubUshing Company (1987) 77. Arai, H. and Tominaga, H., J. Catal. 75, 188-190, (1982) 78. Burke, P. A. and Ko, E. I., J. Catal., 116, 230-233, (1989) 79. Gates, B. C , Katzer, J. R. and Schuit, G. C. A., "Chemistry of Catalytic Processes", Chemical Engineering Series, McGraw-Hill Book Company, (1979) 80. Rodriguez, E., Leconte, M., Basset, J. M. and Tanaka, K , J. Catal., 119, 230-237, (1989) 81. Leconte, M., Theolier, A. and Basset, J. M., J. of Molecular Catalysis, 28, 217-231, (1985) 82. Ishihara, T., Sumi, H., Takita, Y., Chem Letter, 8, 1499-1502 (1994) 83. Tanabe, K, Misono, M., Ono, Y. and Hattori, H., "New Solid Acids and Bases", Elsevier Scientific Pubhshing Company, (1989) 84. Imelik, B., Naccacha, C , Taarit, Y. B., Verdrine, J.C., Coudurier, G. and Prahaud. H., 147 "Catalysis by Zeolites" Elsevier Scientific Pubhshing Company. (1980) 85. Solinas, V., Monad, R., Rombi, E. and Forni, L., " Coke Deposition on Y-Zeolites in Methylnaphthalene Isomerisation", Elsevier Scientific PubUshing Company (1980) 86. Zadeh, S. M., Smith, K, J., "Kinetics of CFLt Decomposition on Supported Co Catalysts", in press J. Catal., in 1998 87. Smith, K J., "Preliminary Study of Methane Coupling on Metal Containing Zeohtes in the Absence of Oxygen", Final Report, Natural Resources Canada, Contract No. 23440-4-1404/01-SQ, Nov. 1995 148 Appendix 1 R e p e a t a b i l i t y o f e x p e r i m e n t a l d a t a To measure the repeatability of the experimental data, two tests at different time but using the same catalyst and at the same reaction condition were conducted. Table A-1 below shows the comparison of the results. Table A- l . Reaction results of two tests over the CH4 reduced Ni/K/Al 20 3 catalysts (7 wt % N i , 1 wt % K and 92 wt % Al 2 03). Reaction at 350 °C, lOlkPa and 10 % C3H5 / 90 % CFLt feed gas. Time C3FL5 Conversion C4 Selectivity C 3 Selectivity C 2 Selectivity (min.) (%) (%) (%) (%) 5 33.7 26.4 9.3 9.9 58.3 68.5 31.3 21.2 30 41.9 42.4 9.3 9.3 57.9 57.7 31.7 31.9 55 43.9 42.6 9.8 9.8 58.3 57.5 30.9 31.4 80 43.4 41.8 9.8 9.7 58.5 57.5 30.7 31.2 105 40.5 39.6 9.7 9.8 57.7 57.5 31.5 31.7 130 38.8 36.6 9.7 9.9 57.4 57.1 31.9 32.2 Test No. 1 2 1 2 1 2 1 2 To calculate the repeatability of the experimental data, the following equation were used: 1=1 STD = ^ fsT^ Note: S2x is the variance of a set of scores and STD is the standard deviation 149 Table A-2. Calculation of the Variation (S2x) and Standard Deviation (STD) of the results shown in Table A-1 Time C3H6 Conversion C4 Selectivity C3 Selectivity C2 Selectivity (min.) S2x STD S2x STD S2x STD S2x STD 5 13.3 3.65 0.12 0.34 26.0 5.10 25.5 5.05 30 0.06 0.25 0.00 0.02 0.01 0.10 0.01 0.10 55 0.42 0.65 0.00 0.05 0.16 0.40 0.06 0.25 80 0.64 0.80 0.00 0.03 0.25 0.50 0.06 0.25 105 0.64 0.80 0.00 0.03 0.25 0.50 0.06 0.25 130 1.21 1.10 0.01 0.12 0.02 0.15 0.02 0.15 Test No. 1 2 1 2 1 2 1 2 Table A-3 Calculation of the Variation and Standard Variation from the results in Table A-2 C3H6 Conversion C4 Selectivity C3 Selectivity C 2 Selectivity S(STD)/n 1.45 S(STD)/n 0.11 S(STD)/n 1.35 S(STD)/n 1.21 Note: S(STD)/n = X(STD)/n and is the standard deviation of the two data groups The results shown in Table A-3 prove that the experimental data are repeatable in the present study. Appendix 2 E x p e r i m e n t a l D a t a Figure 4-3 and Figure 4-4. C3H5 conversion and selectivity over Ni/K/Al 20 3 (1 wt % K) and Ni/P/Al203 (1 wt % P). Reaction at 350 °C and 101 kPa. Feed gas = 90 mol % CFL / 10 mol % C3H6. Catalyst mass = 1.1 g. flow rate = 20 ml/min. Over Ni/K/Al 20 3 catalyst Time-on-stream C 3 H 6 conversion C 4 selectivity C 3 selectivity C 2 selectivity C 5 selectivity Minutes % % % % % 5 77.9 8.6 64.1 26.4 1.0 30 53.0 10.4 59.4 29.0 1.2 55 44.8 10.1 59.0 29.8 1.1 80 38.6 9.9 58.3 30.9 1.0 105 31.0 8.5 57.3 32.9 0.3 130 27.0 8.8 54.9 35.2 1.1 Over Ni/P/Al 20 3 catalyst Time-on-stream C 3 H 6 conversion C 4 selectivity C 3 selectivity C 2 selectivity C 5 selectivity Minutes % % % % % 5 23.1 9.3 63.3 27.4 0 30 16.4 8.0 61.5 30.5 0 55 10.0 8.0 41.5 50.4 0 80 9.8 6.3 39.8 53.9 0 105 9.0 6.2 40.5 53.3 0 130 8.5 6.0 38.5 54.9 0 151 Figure 5-6. Comparison of C3FL5 conversion and C4 selectivity over N1/K/AI2O3 catalysts reduced in H 2 (A) and CFL, (B), Reaction at 350 °C, lOlkPa and 10 % C3FL5 / 90 % CFL, feed gas. Catalyst mass = 1.1 g, flow rate = 20 ml/min. A B A B Time-on-stream C 3 H 6 conversion C 3 H 6 conversion Time-on-stream C 4 selectivity C 4 selectivity Minutes % % Minutes % % 5 33.7 77.9 5 9.3 8.6 30 41.9 51.1 30 9.3 10.4 55 43.9 44.8 55 9.8 10.1 80 43.4 38.6 80 9.8 9.9 105 40.5 31.0 105 9.7 8.7 130 38.8 25.9 130 9.7 8.8 Figure 5-7. Comparison of C 3 and C 2 selectivity over Ni/K/Al203 catalysts reduced in H 2 (A) and CFLt(B). Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CFL, feed gas. Catalyst mass = 1.1 g, flow rate = 20 ml/min. A B A B Time-on-stream C 3 selectivity C 3 selectivity Time-on-stream C 2 selectivity C 2 selectivity Minutes % % Minutes % % 30 57.9 59.4 30 31.8 29.0 55 58.3 59.0 55 30.9 29.8 80 58.5 58.3 80 30.7 30.9 105 57.7 57.3 105 31.5 32.9 130 57.4 54.9 130 31.9 35.2 152 Figure 5-8 and Figure 5-9. Comparison of C4 distribution in total C4's over Ni/K7A1203 catalysts reduced in H 2 (A) and CIL,(B). Reaction at 350 °C, 101 k Pa and 10 % C 3 H6 / 90 % CFL, feed gas. Catalyst mass = 1.1 g, flow rate = 20 ml/min. A B A B A B Time-on-stream n-C4Hio n-C 4 H 1 0 1-butene 1-butene isobutylene isobutylene Minutes % % % % % % 5 3.4 11.1 25.0 13.9 19.3 27.2 30 2.9 5.0 24.4 21.9 22.2 27.0 55 2.9 3.4 23.3 29.5 22.3 28.4 80 3.0 2.7 24.0 31.5 22.0 27.4 105 2.2 1.8 24.4 34.5 22.2 25.5 130 2.4 0 25.3 38.1 21.7 23.8 A B A B Time-on-stream c-2-C 4H 8 c-2-C 4H 8 t-2-C 4H 8 t-2-C 4H 8 Minutes % % % % 5 22.7 19.4 29.5 28.3 30 22.2 21.9 28.9 28.9 55 22.3 26.1 29.1 12.5 80 22.0 26.0 29.0 12.3 105 22.2 25.5 28.9 12.7 130 22.9 23.8 27.7 14.3 153 Figure 6-1. The effect of reduction time in CFL. on the C3FL5 conversion over Ni/K/Al203. Reaction at 350 °C, 101 kPa and 10 % C3FL3 / 90 % CFLj feed gas. Catalyst mass = 1.1 g, flow rate = 20 ml/min. Time-on-stream Minutes C 3 H 6 conversion % with different catalyst reduction time (minutes) 0 min. reduction 15 min. reduction 60 min. reduction 90 min. reduction (%) (%) (%) (%) 5 0 69.4 77.9 61.5 30 0 54.2 51.1 37.1 55 0 48.6 44.8 31.6 80 0 45.0 38.6 26.9 105 0 42.9 31.0 23.8 130 0 38.1 25.9 20.9 Figure 6-2. The effect of reduction time in CFL. on the C4 selectivity over NI/K/AI2O3. Reaction at 350 °C, 101 kPa and 10 % C3FL3 / 90 % CFL, feed gas. Catalyst mass = 1.1 g, flow rate = 20 ml/min Time-on-stream Minutes C 4 selectivity % with different catalyst reduction time (minutes) 0 min. reduction 15 min. reduction 60 min. reduction 90 min. reduction (%) (%) (%) (%) 5 0 8.7 8.6 8.6 30 0 9.8 10.1 10.4 55 0 9.9 10.2 10.1 80 0 9.9 10.1 9.9 105 0 9.7 10.1 8.5 130 0 9.6 9.8 8.8 154 Figure 6-3. The effects of reaction temperature on the C3FL5 conversion over Ni/K7A1203. Reaction at 101 kPa and 10 % C3H5 / 90 % CH4 feed gas. Catalyst mass = 1.1 g, flow rate = 20 ml/min Time-on-stream Minutes C 3 H 6 conversion % vwth different reaction temperature ( °C) Reaction at 300 °C Reaction at 325 °C Reaction at 350 °C Reaction at 375 °C (%) (%) (%) (%) 5 12.2 26.7 78.7 100 30 6.1 16.6 59.0 97.1 55 5.1 15.0 54.2 91.4 80 4.8 15.6 48.0 76.9 105 2.7 14.4 42.0 56.1 130 - 14.7 36 39.1 Figure 6-4. The effects of reaction temperature on the C4 selectivity over Ni/K/Al203. Reaction at 101 kPa and 10 % C3H6 / 90 % CH4 feed gas. Catalyst mass = 1.1 g, flow rate = 20 ml/min. Time-on-stream Minutes C 4 selectivity % vwth different reaction temperature ( °C) Reaction at 300 °C Reaction at 325 °C Reaction at 350 °C Reaction at 375 °C (%) (%) (%) (%) 5 9.3 9.6 8.6 3.0 30 4.8 8.0 10.1 3.4 55 3.1 8.1 10.2 4.9 80 3.1 8.6 10.1 7.3 105 3.0 8.5 10.1 8.0 130 - 8.3 9.8 8.4 155 Figure 6-5. The effects of reaction temperature on the C 2 and C 3 selectivity over Ni/K/Al 20 3. Reaction at 101 kPa and 10 % C3F£o / 90 % CFL, feed gas. Catalyst mass = 1.1 g, flow rate = 20 ml/min. | C 3 selectivity % with different reaction temperature ( °C) Time-on-stream Reaction at 300 °C Reaction at 325 °C Reaction at 350 °C Reaction at 375 °C Minutes (%) (%) (%) (%) 5 64.4 63.5 64.5 70.0 30 63.9 58.2 61.1 67.5 55 62.1 57.5 61.4 65.5 80 61.8 56.3 60.9 61.7 105 58.5 55.5 59.6 57.7 130 - 55.6 58.0 55.1 C 2 selectivity % with different reaction temperature ( °C) Time-on-stream Reaction at 300 °C Reaction at 325 °C Reaction at 350 °C Reaction at 375 °C Minutes (%) (%) (%) (%) 5 26.0 26.3 26.2 28.5 30 32.9 33.5 27.8 29.1 55 34.8 33.8 28.0 29.6 80 35.1 34.7 28.8 31.1 105 36.0 35.8 30.0 33.8 130 - 35.8 32.0 37.2 156 Figure 6-8. The effects of residence time on the selectivity - 5 minutes time-on-stream- over Ni/K/Al 20 3 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3FL3 / 90 % CFL feed gas. Residence time (W/F, g.min/ml) C 4 selectivity (%) C 3 selectivity (%) C 2 selectivity (%) 0.0275 10.3 59.4 30.0 0.0375 10.2 59.4 29.5 0.055 8.6 64.1 26.4 0.065 8.5 65.4 25.4 0.075 4.0 69.6 24.0 Figure 6-9. The effects of residence time on the selectivity - time-on-stream 30 minutes (Table A) and 130 minutes (Table B) - over Ni/K7A1203 catalysts. Reaction at 350 °C, lOlkPa and 10 % C3H5 / 90 % CFL feed gas. Table A Residence time C 4 selectivity C 3 selectivity C 2 selectivity (W/F, g.min/ml) (%) (%) (%) 0.0275 9.0 53.7 37.1 0.0375 10.5 56.2 32.3 0.055 10.4 59.4 29.0 0.065 9.9 61.9 27.2 0.075 9.3 63.6 26.1 fable B Residence time C 4 selectivity C 3 selectivity C 2 selectivity (W/F, g.min/ml) (%) (%) (%) 0.0275 3.9 54.4 41.7 0.0375 5.7 51.6 42.7 0.055 8.8 54.9 35.2 0.065 10.1 60.3 28.9 0.075 9.9 59.8 29.3 157 Figure 6-10. The changes of paraffin and olefin distribution in total C4 with time-on-stream over Ni/K/Al 20 3 catalysts. Reaction at 350 °C, lOlkPa and 10 % C3H6 / 90 % CH4 feed gas. Catalyst mass = 1.1 g, flow rate - 20 ml/min. Time-on-stream minutes Paraffins in total C 4 (%) Olefins in total C 4 (%) 5 11.1 88.9 30 3.9 96.1 55 3.4 96.6 80 2.7 97.3 105 1.8 98.2 130 0 100 Figure 6-11. The effects of residence time on the paraffin and olefin in the total C4's - 30 minutes time-on-stream - over Ni/K/Al203. Reaction at 350 °C, 101 kPa and 10 % QFfe / 90 % CFL, feed gas. Residence time Paraffins in total C 4 Olefins in total C 4 (W/F, g.min/ml) ( % ) ( % ) 0.0275 0 100 0.0375 2.2 97.8 0.055 3.9 96.1 0.065 6.9 93.1 158 Figure 6-12. The effects of residence time on the paraffin in the total C4's - 5 minutes time-on-stream in (A) and 105 minutes time-on-stream (B) - over Ni/K7A1203 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CFL, feed gas. Iso-C 4H 1 0 in total C 4 n-C 4 H 1 0 in total C 4 lso-C 4 H 1 0 in total C 4 n-C 4 H 1 0 in total C 4 Residence time A A B B (W/F, g.min/ml) (%) (%) (%) (%) 0.0275 0 5!6 6 0~ 0.0375 0 4.6 0 0.3 0.055 0.8 11.1 0 1.2 0.065 1.6 17.0 0 3.3 0.075 8.5 41.2 0 4.0 Figure 6-13. The effects of residence time on the olefin in the total C4's -in stream 30 minutes (Table A ) and 105 minutes (Table B ) - over Ni/K/Al 20 3 catalysts at 350 °C. Reaction at 350 °C, lOlkPa and 10 % C3FL5 / 90 % CFL, feed gas. Table A Residence time lso-C 4 H 1 0 in total C 4 n-C 4 H 1 0 in total C 4 lso-C 4 H 1 0 in total C 4 n-C 4 H 1 0 in total C 4 (W/F, g.min/ml) (%) (%) (%) (%) 0.0275 34 9 209 18!6 25!6 0.0375 24.4 22.2 22.2 28.9 0.055 21.9 23.4 21.9 28.9 0.065 17.0 23.9 21.6 30.7 0.075 16.1 25.6 21.4 29.8 Table B Residence time lso-C4Hio in total C 4 n-C 4 H 1 0 in total C 4 lso-C 4 H 1 0 in total C 4 n-C 4 H 1 0 in total C 4 (W/F, g.min/ml) (%) (%) (%) (%) 0.0275 38.1 19.1 19.1 23.8 0.0375 32.1 21.4 21.4 25.0 0.055 28.0 22.0 25.5 26.0 0.065 24.0 22.9 22.9 30.7 0.075 22.0 23.6 22.1 28.2 159 Figure 6-14. The selectivity changes with time-on-stream over Ni/K/Al 20 3 catalysts. Reaction at 350 °C, lOlkPa and 10 % C3H6 / 90 % He feed gas, flow rate = 20 ml/min. Time-on-stream Minutes C4 selectivity % C3 selectivity % C2 selectivity % 5 24.3 67.4 8.3 30 28.4 60.8 10.0 55 28.7 60.0 10.4 80 29.3 59.4 10.4 105 29.8 59.0 10.1 Figure 7-1. C3He conversion over Ni/Si02 catalysts Reaction at 350 °C, 101 kPa and 10 % C3He / 90 % CFL feed gas, flow rate = 20 ml/min. C 3 H 6 conversion with different Ni wt % in Ni/Si0 2 catalysts Time-on-stream 1 wt % Ni 7 wt % Ni 14 wt % Ni Minutes Conversion ( % ) Conversion ( % ) Conversion ( % ) 5 1.7 28.4 76.8 25 1.4 9.7 47.7 45 1.3 8.6 45.5 65 0.0 7.4 42.0 85 0.0 6.8 38.3 100 0.0 6.8 37.2 125 0.0 6.4 34.8 160 Figure 7-3. C 4 selectivity over Ni/Si02 catalysts, Reaction at 350 °C,101 kPa and 10 % C 3 F V 9 O % CH4 feed gas, flow rate = 20 ml/rnin. C 4 selectivity with different Ni wt % in Ni/Si0 2 catalysts Time-on-stream 1 wt % Ni 7 wt % Ni 14 wt % Ni Minutes C 4 selectivity ( % ) C 4 selectivity ( % ) C 4 selectivity ( % ) 5 53.3 9.9 8.1 25 0.0 10.2 8.7 45 0.0 10.0 8.9 65 0.0 9.7 7.0 85 0.0 9.9 6.7 105 0.0 8.7 6.5 125 0.0 8.2 6.8 Figure 7-4. C 3 selectivity (Table A) and C 2 selectivity (Table B) over Ni/Si02 catalyst. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CKL feed gas, flow rate - 20 ml/min. Table A C 3 selectivity with different Ni wt % in Ni/Si0 2 catalysts Time-on-stream 1 wt % Ni 7 wt % Ni 14 wt % Ni Minutes C 3 selectivity ( % ) C 3 selectivity ( % ) C 3 selectivity ( % ) 5 33.0 65.9 73.3 25 55.0 50.0 73.1 45 46.1 44.0 73.0 65 0 37.3 74.0 85 0 36.2 74.0 105 0 35.5 73.6 125 0 35.3 72.8 161 Table B C 2 selectivity with different Ni wt % in Ni/Si0 2 catalysts Time-on-stream 1 wt % Ni 7 wt % Ni 14 wt % Ni Minutes C 2 selectivity (% ) C 2 selectivity (% ) C 2 selectivity ( % ) 5 13.7 24.2 18.7 25 45.0 45.0 18.3 45 54.0 51.0 18.1 65 0 53.0 19.0 85 0 53.9 19.9 105 0 55.8 19.9 125 0 56.4 20.4 Figure 7-5 and Figure 7-6. Comparison of C3H6 conversion and C4 selectivity over Ni/Si02 and Ni/K/Si02 catalysts. Reaction at 350 °C, 101 kPa and 10 % C3Hs / 90 % CH4 feed gas, flow rate = 20 ml/min. Time-on-stream Minutes C3H6 Ni/Si02 Conversion (%) Ni/K/SiC-2 C 4 selectivity ( % ) Ni/Si0 2 Ni/K/SiOz 5 28.4 24.4 9.9 7.0 25 12.0 13.5 10.2 4.0 45 8.6 11.2 10.0 4.0 65 7.4 9.9 9.7 5.8 85 6.8 9.3 9.9 5.6 105 6.8 9.0 8.7 3.9 162 Figure 7-7 and Figure 7-8 Comparison of C3 selectivity over silica supported Ni catalysts. Reaction at 350 °C, 101 kPa and 10 % CsFk / 90 % CFL, feed gas, flow rate = 20 ml/min. Time-on-stream C 3 selectivity ( % ) C 2 selectivity ( % ) Minutes Ni/Si02 Ni/K/Si02 Ni/Si0 2 Ni/K/SiOz 5 65.9 70.9 24.2 22.1 25 50.0 69.4 51.5 26.7 45 44.0 64.0 54.3 26.3 65 37.3 53.0 53.0 44.9 85 36.2 49.1 53.9 45.3 105 35.5 47.0 55.8 46.5 Figure 7-9 and Figure 7-10. Conversion of C3FL5 and C 4 selectivity over Ni/SAPO-5 catalysts prepared by ion exchange and impregnation. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CFL, feed gas, flow rate = 20 ml/min. Time-on-stream C 3 H 6 conversion ( % ) C 4 selectivity ( % ) Minutes impregnation ion exchange impregnation ion exchange 5 17.3 50.6 24.0 66.1 25 2.8 37.0 25.0 61.0 45 1.5 33.2 25.0 56.5 65 1.9 25.6 24.0 56.6 85 1.5 19.2 22.0 56.6 105 1.4 16.3 20.1 56.4 163 Figure 7-11 and Figure 7-12, Cs+ selectivity, C3 selectivity and C2 selectivity over Ni/SAPO-5 catalysts prepared by ion exchange and impregnation. Reaction at 350 °C,101 kPa and 10 % C3H5 / 90 % CFL feed gas, flow rate = 20 ml/min. Time-on-stream C3 selectivity ( % ) C2 selectivity ( % ) C5+ selectivity ( % ) Minutes impregnation ion exchange impregnation ion exchange ion exchange 5 59.6 11.4 28.4 2.5 20.1 25 50.0 30.0 35 7.6 2.0 26.0 45 43.0 30.8 65 40.0 9.0 31.0 1.8 28.0 85 41.0 33.3 95 10.0 2.0 28.0 105 40.0 36.5 125 40.0 11.3 37.2 3.0 29.0 Figure 7-13 and Figure 7-14. The distribution of C4 paraffins and olefins in total C4's over Ni/SAPO-5 catalysts prepared by ion exchange. Reaction at 350 °C,101 kPa and 10 % C3H6 / 90 % CFt, feed gas, flow rate = 20 ml/min. Time-on-stream minutes iso-C4H10 % n-C4Hio % 1-butene % isobutylene % c-2-C4H8 % t-2-C4H8 % 5.0 38.2 0.8 8.2 23.8 11.8 17.2 35.0 34.5 1.1 7.6 29.3 11.2 16.3 65.0 29.4 1.1 7.6 35.0 10.8 16.0 95.0 26.6 1.3 7.6 38.2 10.6 15.8 125.0 25.0 1.6 7.4 39.7 10.6 15.7 155.0 23.5 1.8 7.4 41.1 10.7 15.6 164 Figure 7-15 and Figure 7-16. C3FL5 conversion and selectivities over Ni/Na-Y catalysts. Reaction at 350 °C,101 kPa and 10 % CsHe / 90 % CFI4 feed gas, flow rate = 20 ml/min. Time-on-stream Conversion C 5 + selectivity C 4 selectivity C 3 selectivity C 2 selectivity Minutes (%) (%) (%) (%) (%) 5 98.4 13.7 47.9 33.2 5.2 30 98.1 17.3 35.2 42.1 5.4 55 96.6 20.1 30.7 43.9 5.3 80 92.7 21.2 28.1 45.3 5.5 105 80.9 21.8 25.4 46.8 6.0 130 53.7 22.5 23.2 47.8 6.5 Figure 7-17 and Figure 7-18. C4 paraffin and olefin changes over Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % CsHe / 90 % CFL, feed gas, flow rate - 20 ml/min. Time-on-stream iso-C4Hio n-C4Hio 1-butene isobutylene c-2-C 4H 8 t-2-C 4H 8 minutes % % % % % % 5 91.0 7.0 0.3 0.9 0.3 0.5 30 89.9 8.4 0.2 0.7 0.3 0.5 55 88.8 8.7 0.3 1.0 0.5 0.7 80 86.0 9.0 0.7 1.7 1.0 1.5 105 77.2 9.0 2.2 3.8 3.2 4.7 130 54.5 7.6 6.2 9.7 8.9 13.1 165 Figure 7-19 and Figure 7-20. C2H4 conversion and selectrvities over Ni/Na-Y catalyst. Reaction at 250 °C and 101 kPa and 10 % C2H4 / 90 % CH4 feed gas, flow rate = 20 rm/min. Time-on-stream Conversion C 2 selectivity C 3 selectivity C 4 selectivity C 5 + selectivity Minutes (%) (%) (%) (%) (%) 5 90.6 21.6 2.3 59.9 16.2 30 83.3 17.4 1.6 58.7 22.3 55 74.1 16.9 3.1 55.3 24.7 80 64.1 17.1 3.3 54.8 24.9 105 55.8 17.2 3.8 55.1 24.0 130 42.8 14.5 10.0 49.7 25.8 Figure 7-21 and Figure 7-22. C4 paraffins and olefins in total C4 over Ni/Na-Y catalyst, Reaction at 250 °C and 101 kPa and 10 % C2H4 / 90 % CFL, feed gas, flow rate = 20 rnl/rnin. Time-on-stream iso-C 4 H 1 0 n-C 4 H 1 0 1-butene isobutylene c-2-C 4H 8 t-2-C 4H 8 minutes % % % % % % 5 89.9 5.7 0.6 1.0 1.1 1.7 30 78.5 14.6 1.1 0.6 2.0 3.3 55 70.4 17.8 1.7 0.7 3.5 5.8 80 62.2 17.8 2.8 0.9 5.9 9.7 105 53.7 17.9 4.3 1.1 8.5 14.5 130 39.7 15.1 6.7 1.5 13.7 23.3 166 Figure 7-23 and Figure 7-24. Effect of residence time changes on C3FL5 conversion and C 4 selectivity over Ni/Na-Y catalyst at 30 (A) and 105 (B) minutes time-on-stream Reaction at 350 °C, 101 kPa and 10 % C3FL5 I 90 % CFL, feed gas, flow rate = 20 ml/min. C 3 H 6 conversion C 4 selectivity Residence time A B A B (W/F, g.min/ml) % % % % 0.00625 24.5 6.0 24.0 12.7 0.0125 47.9 9.4 29.0 17.0 0.0275 88.0 23.7 32.2 21.5 0.0413 94.0 48.6 34.5 24.6 0.055 98.1 80.9 35.2 25.4 0.0688 98.6 94.5 36.8 29.1 Figure 7-25. Effects of residence time changes on the C2, C 3 and C5 selectivity over Ni/Na-Y catalyst at 30 minutes time-on-stream Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CFL, feed gas, flow rate = 20 ml/min. Residence time C 2 selectivity C 3 selectivity C 4 selectivity (W/F, g.min/ml) (%) (%) (%) 6.25E-03 40.8 6.2 13.8 0.0125 39.5 5.8 22.0 0.0275 41.6 5.3 21.4 0.0413 42.0 5.0 19.9 0.055 42.1 4.8 17.3 0.0688 43.8 4.2 13.0 167 Figure 7-26 and Figure 7-27. Effects of residence time changes on the C4 paraffins and olefins over Ni/Na-Y catalyst at 30 minutes time-on-stream Reaction at 350 °C, 101 kPa and 10 % C3Ho / 90 % CFL, feed gas. Residence time iso-C4Hio n-C4Hio 1-butene isobutylene c-2-C 4H 8 t-2-C 4H 8 (W/F, g.min/ml) % % % % % % 0.00625 21.9 4.6 11.0 22.9 16.1 26.5 0.0125 52.8 5.6 8.0 11.4 8.8 12.2 0.0275 86.3 7.7 0.9 2.1 1.5 1.8 0.04125 89.7 8.1 0.4 0.9 0.4 0.7 0.055 89.9 8.7 0.4 0.7 0.3 0.5 0.06875 89.8 8.8 0.3 0.5 0.2 0.3 Figure 7-28 and Figure 7-29. Effects of a addition of a base on Ni/Na-Y catalyst on the C3Ff6 conversion and C 4 selectivity. Reaction at 350 °C, 101 kPa and 10 % C3H<; / 90 % CFL, feed gas, flow rate = 20 ml/min. Time-on-stream C 3 H 6 conversion ( % ) C 4 selectivity ( % ) Minutes without a base with a base without a base with a base 5 98.4 42.5 47.9 11.2 30 98.1 7.6 35.2 5.6 55 96.6 5.1 30.7 0 80 92.7 4.2 28.1 0 105 80.9 1.5 25.4 0 130 53.7 0 23.2 0 168 Figure 7-30 and Figure 7-31. Effects of a addition of a base on Ni/Na-Y catalyst on C3 and C 2 selectrvities. Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CFL feed gas, flow rate = 20 ml/min. Time-on-stream C 3 selectivity ( % ) C 2 selectivity ( % ) Minutes without a base with a base without a base with a base 5 33.2 66.7 5.2 19.0 30 42.1 71.0 5.4 20.0 55 43.9 73.0 5.3 23.0 80 45.3 74.0 5.5 26.0 105 46.8 - 6.0 -130 47.8 - 6.5 -Figure 7-32. Effect of a base addition into Ni/Na-Y on the C 4 distribution at 30 minutes time-on-stream. Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CFL, feed gas, flow rate = 20 ml/min. C 4 name Ni/Na-Y with a base % in total C 4 's Ni/Na-Y without a base % in total C 4 's iso-C4H-io 8.5 91.0 1-butene 13.2 0.3 n-C4Hio 4.4 7.0 isobutylene 27.2 0.9 c-2-C 4H 8 18.9 0.3 t-2-C 4H 8 27.8 0.5 169 Figure 7-33. C2FL» conversion and the product selectivities over Ni/K/Na-Y catalyst, Reaction at 250 °C, 101 kPa and 10 % C2H4 / 90 % CFL feed gas, flow rate = 20 ml/min. Time-on-stream C2H4 conversion C4 selectivity C3 selectivity C2 selectivity Minutes % % % % 5 &2 871 &8 6VJ 30 1.0 70.0 7.6 23.0 55 0.1 68.0 3.8 26.6 Figure 7-34. C3FL, conversion over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % C3FL / 90 % CFL feed gas, flow rate = 20 ml/min. Time-on-stream Minutes Ni/K/Al203 C3H6 conversion ( % ) Ni/SiOz Ni/SAPO-5 Ni/Na-Y 5 77.9 28.4 50.6 98.4 30 55.0 9.7 33.2 98.1 55 44.8 8.6 25.5 96.6 80 38.6 7.4 19.2 92.7 105 31.0 6.8 16.3 80.9 130 25.9 6.8 14.5 53.7 170 Figure 7-35. C 4 selectivity % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CH4 feed gas, flow rate = 20 rrd/rnin. Time-on-stream C 4 selectivity ( % ) Minutes N i / K / A l 2 0 3 Ni/SiOz Ni/SAPO-5 Ni/Na-Y 5 8.6 9.9 66.1 47.9 30 10.4 10.2 62.3 35.2 55 10.1 10.0 58.0 30.7 80 9.9 9.7 56.6 28.1 105 8.5 9.9 57.0 25.4 130 8.8 8.7 57.0 23.2 Figure 7-36. C3 selectivity % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % C3H6 / 90 % CFL, feed gas, flow rate = 20 ml/min. Time-on-stream C 3 selectivity (% ) Minutes N i / K / A l 2 0 3 Ni/SiOz Ni/SAPO-5 Ni /Na-Y 5 64.0 65.9 11.4 33.2 30 59.4 38.4 7.6 42.1 55 59.0 35.8 9.0 43.9 80 58.2 36.0 11.0 45.3 105 57.9 36.2 11.3 46.8 130 54.9 35.5 11.6 47.8 171 Figure 7-37. C 2 selectivity % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % C3F£6 / 90 % CFL. feed gas, flow rate = 20 ml/min. Time-on-stream C 2 selectivity ( % ) Minutes N i / K / A l 2 0 3 Ni/SiQz Ni/SAPO-5 Ni/Na-Y 5 26.4 24.2 2.5 5.2 30 29.0 51.5 2.0 5.4 55 29.8 54.3 1.8 5.3 80 30.9 53.0 2.2 5.5 105 32.9 53.9 2.3 6.0 130 35.2 55.8 2.4 5.6 Figure 7-38. C 5 + selectivity % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % C3H„ / 90 % CFL, feed gas, flow rate = 20 ml/min. Time-on-stream C 5 + selectivity ( % ) Minutes N i / K / A l 2 0 3 Ni/SiOz Ni/SAPO-5 Ni/Na-Y 5 1.0 0 20.1 13.7 30 1.2 0 28.1 17.3 55 1.1 0 31.0 20.1 80 1.0 0 31.0 21.2 105 1.0 0 30.0 21.8 130 1.1 0 28.6 22.5 172 Figure 7-39. Yield of C4 % over Ni catalysts supported by different supports. Reaction at 350 °C, 101 kPa and 10 % C3H<; / 90 % CH4 feed gas, flow rate = 20 ml/min. Time-on-stream C 4 yield ( % ) Minutes N i / K / A l 2 0 3 Ni/SiOz Ni/SAPO-5 Ni/Na-Y 5 6.7 2.8 33.4 47.3 30 5.3 1.0 20.7 34.7 55 4.5 0.9 14.4 30.2 80 3.8 0.7 10.9 27.3 105 2.6 0.7 9.4 24.0 130 2.3 0.6 8.3 20.3 Figure 7-40. iso-C4H10 and n-C^o % in total C4's over Ni/K/Al 20 3 and Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % C3FLs / 90 % CFL feed gas, flow rate = 20 ml/min. N i / K / A l 2 0 3 Ni/Na-Y N i / K / A l 2 0 3 Ni/Na-Y Time-on-stream iso-C4H-io iso-C4rHio n-C4Hio n - C 4 H 1 0 Minutes % % % % 30 0.0 89.9 5.0 8.4 55 0.0 88.8 3.4 8.7 80 0.0 86.0 2.7 9.0 105 0.0 77.2 1.8 9.0 130 0.0 54.5 0.0 7.6 173 Figure 7-41. 1-butene and isobutylene % in total C4's over Ni/K/Al203 and Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CFL, feed gas, flow rate = 20 ml/min Ni/K/Al 20 3 Ni/Na-Y Ni/K/Al 20 3 Ni/Na-Y Time-on-stream 1-butene isobutylene 1-butene isobutylene Minutes % % % % 5 13.9 0.3 27.2 0.9 30 21.9 0.2 27.5 0.7 55 27.0 0.3 28.4 1.0 80 31.5 0.7 27.4 1.7 105 34.5 2.2 25.5 3.8 130 38.1 6.2 23.8 9.7 Figure 7-42. c-2-GJfg and t-2-C4H8 % in total C4's over Ni/K/Al 20 3 and Ni/Na-Y catalysts. Reaction at 350 °C, 101 kPa and 10 % C3FL5 / 90 % CFL, feed gas, flow rate = 20 ml/min. Ni/K/Al 20 3 Ni/Na-Y Ni/K/Al 20 3 Ni/Na-Y Time-on-stream c-2-C 4H 8 t-2-C 4H 8 c-2-C 4H 8 t-2-C 4H 8 Minutes % % % % 5 19.4 0.3 28.3 0.5 30 21.9 0.3 21.0 0.5 55 26.1 0.5 14.0 0.7 80 26.0 1.0 12.3 1.5 105.0 25.5 3.2 12.7 4.7 130 23.8 8.9 14.3 13.1 174 Figure 7-43. The ratio of C 4 paraffin to C 4 olefin of over Ni/Na-Y and over Ni/K/Al203. Reaction at 350 °C, 101 kPa and 10 % C3H5 / 90 % CFL feed gas, flow rate = 20 ml/min. Time-on-stream Minutes N i / K / A l 2 0 3 C 4 / C 4 = Ni/Na-Y C 4 / C 4 = 5 0.125 49.1 30 0.041 55.7 55 0.035 38.5 80 0.028 19.0 105 0.018 6.2 130 0.000 1.6 The changes in C3FL; conversion and selectrvities with time-on-stream over Ni/K/Al203 catalyst. Reaction at 350 °C and 101 kPa and 10 % C3H5 / 90 % CFL feed gas, flow rate = 20 ml/min. Catalyst mass = 0.55 g Time-on-stream Conversion C 2 selectivity C 3 selectivity C 4 selectivity C 5 + selectivity Minutes ( % ) ( % ) ( % ) ( % ) ( % ) 5 51.5 30.0 59.4 9.87 0.75 30 26.9 37.1 53.7 9.02 0.20 55 22.0 38.2 52.9 8.66 0.18 80 21.1 39.3 52.2 8.27 0.20 105 17.7 37.4 55.5 7.08 0.00 130 14.8 41.7 54.4 3.86 0.00 175 The changes in C3FL5 conversion and selectivities with time-on-stream over Ni/K/Al203 catalyst. Reaction at 350 °C and 101 kPa and 10 % C3H<; / 90 % CH4 feed gas, flow rate = 20 ml/min. Catalyst mass = 0.75 g Time-on-stream Conversion C 2 selectivity C 3 selectivity C 4 selectivity C 5 + selectivity Minutes ( % ) ( % ) ( % ) ( % ) ( % ) 5 50.6 29.5 59.4 10.2 0.90 30 38.3 32.3 56.2 10.5 0.97 55 33.9 33.0 56.5 10.1 0.34 80 26.1 35.5 55.2 8.95 0.38 105 20.5 38.2 53.3 8.51 0.00 130 15.7 42.7 51.6 5.74 0.00 The changes in C3H6 conversion and selectivities with time-on-stream over Ni/K/Al203 catalyst. Reaction at 350 °C and 101 kPa and 10 % C3H6 / 90 % CH4 feed gas, flow rate = 20 ml/min. Catalyst mass = 1.1 g Time-on-stream Conversion C 2 selectivity C 3 selectivity C 4 selectivity C 5 + selectivity Minutes ( % ) ( % ) ( % ) ( % ) ( % ) 5 77.9 26.4 64.1 8.56 1.02 30 51.1 29.0 59.4 10.4 1.24 55 44.8 29.8 59.0 10.1 1.05 80 38.6 30.9 58.3 9.88 1.01 105 31.0 32.9 57.3 8.49 0.34 130 25.9 35.2 54.9 8.77 1.12 176 The changes in C3IL5 conversion and selectrvities with time-on-stream over Ni/K7A1203 catalyst. Reaction at 350 °C and 101 kPa and 10 % C3H6 / 90 % CFL feed gas, flow rate = 20 rril/rnin. Catalyst mass = 1.3 g Time-on-stream Conversion C 2 selectivity Minutes ( % ) ( % ) 5 79.7 25.4 30 64.0 27.2 55 62.5 27.3 80 57.4 27.5 105 51.9 27.9 130 46.2 28.9 C 3 selectivity C 4 selectivity C 5 + selectivity ( % ) ( % ) ( % ) 6JT4 8~!47 0 7 6 61.9 9.97 0.89 62.0 9.95 0.81 61.7 9.98 0.76 61.3 10.0 0.79 60.3 10.1 0.77 The changes in C3H6 conversion and selectrvities with time-on-stream over Ni/K7A1203 catalyst. Reaction at 350 °C and 101 kPa and 10 % C3H5 / 90 % CH4 feed gas, flow rate = 20 rrd/min. Catalyst mass = 1.5 g Time-on-stream Conversion C 2 selectivity C 3 selectivity C 4 selectivity Cs* selectivity Minutes ( % ) ( % ) ( % ) ( % ) ( % ) 5 98.0 26.2 69.6 3.99 0.29 30 71.8 26.1 63.6 9.29 0.99 55 65.1 26.3 62.6 10.1 0.94 80 58.2 26.8 61.8 10.3 1.00 105 52.9 27.9 60.7 10.4 1.00 130 46.3 29.3 59.8 9.87 0.98 177 Appendix 3 C a r b o n B a l a n c e D a t a Table 1. C3FL5 conversion and carbon balance over Ni/K/Al203 (1 wt % K). Reaction at 325 °C to 375 °C and 101 kPa. Feed gas = 90 mol % CFL, /10 mol % GjHg. Catalyst mass = 1.1 g. flow rate = 20 ml/min. Carbon balance (%) Temp.°C \ Time-on-stream (min.) 5 30 55 80 105 130 325 °C 97 96 96 96 96 96 350 °C 90 92 93 94 95 95 375 °C 82 83 85 89 91 C3FL5 conversion (%) Temp.°C \ Time-on-stream (min.) 5 30 55 80 105 130 325 °C 27 17 15 16 14 14 350 °C 78 51 45 39 31 26 375 °C 97 91 77 56 39 Since most of reactions were conducted at 350 °C and carbon balances were more than 90 %, the selectivities calculated in Table 3-1 can reflect real selectivities (for example, the C 4 selectivity should be equal to C 4 product / all products including coke formed in the reaction). It could be noticed that high C3FL5 conversion resulted in low carbon balance. Since the 178 C3H5 conversion over the Ni/K7A1203 catalyst was high compared to the C3H6 conversions over other Ni catalysts (the Ni/Na-Y catalyst was the only exception), it could conclude that carbon deposition did not effect the calculation of the selectrvities significantly. Table 2. Carbon balance over different catalysts. Reaction at 350 °C and 101 kPa. Feed gas = 90 mol % CFL / 10 mol % C3H6. Catalyst mass = 1.1 g. flow rate = 20 ml/min. Carbon balance (%) Catalyst \ Time-on-stream (min.) 5 30 55 80 105 130 Ni/Si02(7wt%Ni) 95 96 95 95 96 97 Ni/SAPO-5 (ion exchange) 94 95 95 98 98 98 Ni/Na-Y (ion exchange) 79 90 91 91 92 96 179 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0058632/manifest

Comment

Related Items