Citation
Studies of Hydrogenation Catalysis Using Hyperpolarized NMR

Material Information

Title:
Studies of Hydrogenation Catalysis Using Hyperpolarized NMR
Creator:
Zhou, Ronghui
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (1 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Chemistry
Committee Chair:
BOWERS,CLIFFORD RUSSELL
Committee Co-Chair:
MCELWEE-WHITE,LISA ANN
Committee Members:
POLFER,NICOLAS CAMILLE
MURRAY,LESLIE JUSTIN
HAGELIN,HELENA AE
Graduation Date:
5/3/2014

Subjects

Subjects / Keywords:
Catalysts ( jstor )
Flow velocity ( jstor )
Gas flow ( jstor )
Hydrogen ( jstor )
Hydrogenation ( jstor )
Magnetic fields ( jstor )
Propane ( jstor )
Propylene ( jstor )
Signals ( jstor )
Temperature dependence ( jstor )
Chemistry -- Dissertations, Academic -- UF
altadena -- catalysis -- heterogeneous -- hydrogenation -- hyperpolarization -- isomerization -- nanoparticle -- nmr -- parahydrogen -- pasadena -- phip -- simulation -- stereoselectivity
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Chemistry thesis, Ph.D.

Notes

Abstract:
Surface processes involved in heterogeneous catalysis can be effectively studied by in-situ/operando techniques. For example, in heterogeneous hydrogenation, molecular hydrogen dissociates into atomic hydrogen upon chemisorption on metal surfaces followed by migration, spillover and even dissolution into the lattice, leading to random addition. Pairwise addition, where the two hydrogen atoms come from a single molecule and retain their nuclear spin correlations, is not typically considered as a distinct process. Nevertheless, surface reaction activities and pathways have been shown to be affected by the presence of reactants, intermediates, products and by-products such as carbonaceous deposits. Other factors affecting heterogeneous catalysis include the nature of the catalyst, such as metal type, particle size, shape, morphology, dispersion, as well as the properties of the supports, especially those with strong metal-support interactions (SMSI). Such factors are also important in determining the favorability of pairwise addition. While it would be extremely difficult to distinguish between these two pathways using conventional methods such as gas chromatography, hyperpolarized NMR based on Parahydrogen Induced Polarization (PHIP) provides a unique tool to study pairwise hydrogen addition, since only this pathway leads to dramatically enhanced (105) NMR signals. In this study, a variety of catalysts were prepared with transition metals (Pt, Ir, etc.) supported on different oxides (Al2O3, TiO2, etc.), through precipitation of metal salts onto supports from an aqueous solution by controlling solution pH followed by aging and calcination. Active surface area, dispersion and particle size of the catalysts were characterized by CO chemisorption measurements. For PHIP experiments, a home-built micro-reactor with heating unit was installed on the top of the magnet. The precise delivery of gas mixture (propylene, p-H2 and carrier gas) was facilitated with Mass Flow Controllers (MFCs). The resultant hyperpolarized propane flew down to the NMR probe for detection. Kinetic parameters, such as reaction temperature, gas composition and flow rate, were varied to optimize PHIP enhancement. Kinetic studies about reaction orders and activation energies were made possible with the dramatic signal enhancement. In addition, by comparing resultant NMR spectra from reactions with n-H2 or p-H2, contributions from pairwise addition was able to be derived. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2014.
Local:
Adviser: BOWERS,CLIFFORD RUSSELL.
Local:
Co-adviser: MCELWEE-WHITE,LISA ANN.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2016-05-31
Statement of Responsibility:
by Ronghui Zhou.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Embargo Date:
5/31/2016
Resource Identifier:
907295093 ( OCLC )
Classification:
LD1780 2014 ( lcc )

Downloads

This item has the following downloads:


Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ETOZ95H8K_F1B6AZ INGEST_TIME 2014-10-03T21:50:31Z PACKAGE UFE0046487_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

1 STUDIES OF HYDROGENATION CATALYSIS USING HYPERPOLARIZED NMR By RONGHUI ZHOU A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 201 4

PAGE 2

2 201 4 Ronghui Zhou

PAGE 3

3 To my beloved parents and family

PAGE 4

4 ACKNOWLEDGMENTS This work will not be possible without all helps received over the past four years. Especially for heterogeneous catalysis, I am very grateful to the collaboration with Prof. Helena Hagelin under the support from the Petroleum Research Foundation of the American Chemi cal Society (ACS PRF #52258 ND5). Dr. Luke Neal designed the reactor and Dr. Wei Cheng helped to initiate the LabVIEW programming, prepared all of the supported nanoparticles as heterogeneous catalysts, and provided insights, discussions and suggestions ab out steady state reaction kinetics. This collaboration is highly appreciated. I appreciate Prof. Ilya Kuprov and Prof. Mal colm H Levitt at the University of Southampton (UK) for helps with density matrix simulations using Spinach/Matlab and SpinDynamica/Ma thematica, and Prof. Konstantin Ivanov at International Tomography Center (Russia) for help with polarization transfer simulations in Matlab. I also appreciate my other supervisory committee members, Prof. Leslie Murray, Prof. Nicolas Polfer and Prof. Lisa McElwee White for their supports and helpful discussions. I would like to express my deepest gratitude to my advisor, Prof. Clifford R. Bowers for introducing me into this exciting research area and teach me from ABC of Swagelok to Quantum Mechanics and density matrix simulations of NMR spectra. His knowledge, insight, enthusiasm and dedication to scientific research have been a constant inspiration for me. I would also like to thank group members for collaborations, helpful discussions and all kinds of s upports, including Dr. Muslim Dvoyashkin, Ryan Wood, Chris Akel, Hrishi Bhase, Turgut Sonmez, Ryan Wolf, John Tokarski, Chris Reeg, Celeste Kennard, Daniel Schulman, Christianne Barry, Steven Geller, Navid Mirnazari, Lauren McCarthy, and Kristopher Schock. Especially, I highly appreciate

PAGE 5

5 helps from Ryan Wood, including but not limited to teaching me the basics of NMR theory data processing and programming, operation s of the spectrometer, and working with me side by side during the early stage of my researc h. Daniel, Christianne Steven Navid, and Ryan Wolf have been working with me over different periods of time, and contributed to some of the results included in this d i ssertation Special thanks go to Jay Horton, the Cryogenic Services Facility (Greg Labbe and John Graham), Machine Shop (Mark Link, Bill Malphurs, Ed Storch and John Vanleer) and Electronics Shop (Pete Axson and Dan Ekdahl) in the Department of Physics at University of Florida for supports on the des ign and construction of parahydrogen converters, gas handling system, and gas phase continuous flow reactor system. Finally, I would like to thank my parents, parents in law, my wife and daughter for their unconditional love and supports. Especially, than ks to my parents and parents in law who, leaving behind big famil ies in China, came to US A to take care of our daughter, my wife and I were able to spend more time on research and study I also appreciate my sister and brother in law for taking care of our parents in China. My wife has been with me for more than ten years and stayed with me through out the hard times and spent lots of time educating our daughter. I cannot thank her enough.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ .......... 10 LIST OF FIGURE S ................................ ................................ ................................ ........ 11 LIST OF SCHEMES ................................ ................................ ................................ ...... 16 LIST OF ABBREVIATIONS ................................ ................................ ........................... 17 ABSTRACT ................................ ................................ ................................ ................... 19 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 21 1.1 Hy perpolarization: Improvement of NMR Sensitivity ................................ ......... 21 1.1.1 Pushing Thermal Equilibrium ................................ ................................ ... 21 1.1.2 Polarization Transfer from Electron or Photon to Nucleus ....................... 22 1.1.2.1 Dynamic Nuclear Polarization (DNP) ................................ ............. 22 1.1.2.2 Optical Nuclear Polarization: Optically Pu mped NMR .................... 22 1.1.2. 3 Spin Exchange Optical Pumping (SEOP) ................................ ...... 22 1.1.3 Polarization Transfer from Nucleus to Nucleus ................................ ....... 22 1.1.4 Parahydrogen Induced Polarization (PHIP) ................................ ............. 23 1.2 Basics of PHIP ................................ ................................ ................................ .. 23 1.2.1 Nuclear Spin Isomers: Ortho and Para Hydrogen ................................ .. 23 1.2.2 PHIP: PASADENA versus ALTADENA ................................ ................... 25 1.2.3 Signal Amplification by Reversible Exchange (SABRE) .......................... 27 1.2.4 Basic Requirements for Observing PHIP ................................ ................. 27 1. 3 PHIP in Heterogeneous Catalysis ................................ ................................ ..... 27 1.3.1 Kinetics and Mechanisms of Reactions on Metals ................................ .. 29 1.3.1.1 The Horiuti Polanyi Mechanism ................................ ..................... 29 1.3.1.2 The Active Sites on Surfaces ................................ ......................... 30 1.3.1.3 The MARI Approximation ................................ ............................... 31 1.3.1.4 The Proximity Effect ................................ ................................ ....... 31 1.3.1.5 The Time Scale in Catalysis ................................ ........................... 31 1.3.2 The Arrhenius Equation and Activation Energy ................................ ....... 31 1.3.3 Rate Equation and the Reaction Orders ................................ .................. 32 1. 4 Applications of PHIP ................................ ................................ ......................... 33 1.4. 1 Producing Hyperpolarized Tracers for Biomedical Applications .............. 33 1.4.2 Elucidating the Kinetics and Mechanisms of Reactions on Su rfaces ...... 33 1.4.3 Operando Spectroscopy and Catalysis Imaging ................................ ...... 34

PAGE 7

7 2 INSTRUMENTATION AND EXPERIMENTAL METHODS ................................ ..... 36 2.1 Gas Handling System and Heterogeneous Catalytic Reactor ........................... 36 2.2 Parahydrogen Enrichment and Ortho to Para hydro gen Converter ................. 40 2.2.1 Preparation of 50% Enriched p H 2 ................................ ........................... 40 2.2.2 Preparation of 99% Enriched p H 2 ................................ ........................... 41 2.2.3 Determining the Ortho /Para Hydrogen Ratio by NMR ............................ 41 2.3 Preparation of Supported Metal as Heterogeneous Catalysts .......................... 42 2.3.1 Preparation of Supported Platinum (Pt) Catalysts ................................ ... 42 2.3.2 Preparation of Supported Iridium (Ir) Catalysts ................................ ....... 42 2.4 Characterization of Heterogeneous Catalysts ................................ ................... 43 2.4.1 Determine the Actual Metal Loading with ICP AES ................................ 43 2.4.2 Temperature Programmed Reduction Measurements ............................. 43 2.4.3 Metal Dispersion and Particle Size: the Chemisorption Measurement .... 44 2.4.4 Transmission Electro n Microscopy (TEM) Measurement ........................ 45 2.5 NMR Experiments ................................ ................................ ............................. 46 2.5.1 Reactor Setup ................................ ................................ ......................... 46 2.5.2 St eady State Optimization ................................ ................................ ....... 46 2.5.3 PHIP Procedure ................................ ................................ ...................... 47 2.6 The Effect of Flow Rate: Relaxation and Magnetization Buildup ...................... 48 2.6.1 Static Spectra of Propylene, Propane and the Mixture ............................ 48 2.6.2 Magnetic Field Profile ................................ ................................ .............. 49 2.6.3 Magnetization Buildup: Effect of Gas Flow Rate on Signal Inten sity ....... 50 2.6.4 Para H 2 Enrichment: Effect of Gas Flow Rate ................................ ......... 52 2.7 Kinetic Data Interpretation of Propylene Hydrogenation ................................ ... 53 2.7.1 Data Treatment for Reactions with 50% Enriched p H 2 ........................... 53 2.7.2 Propylene to Propane Conversion ................................ ........................... 53 2.7.3 Signal Enhancement Factor and Pairwise Selectivity .............................. 54 2.7.3 .1 Signal Enhancement Factor ................................ ........................... 54 2.7.3 .2 Pairwise Selectivity ................................ ................................ ........ 57 3 OXIDE SUPPORTED PLATINUM CATALYSTS: KINETIC STUDIES OF PROPYNE HYDROGENATION ................................ ................................ .............. 61 3.1 Background ................................ ................................ ................................ ....... 61 3.2 Experimental Methods ................................ ................................ ...................... 63 3.3 Results and Discussion ................................ ................................ ..................... 65 3.3.1 Temperature Dependent Studies of Propyne Partial Reduction .............. 65 3.3. 1.1 Hydrogen Partial Pressure Dependent Reduction ......................... 68 3.3. 1.2 Total Flow Rate Dependent Re duction: CAT023a ......................... 70 3.3.2 Reaction Orders with Respect to p H 2 ................................ ..................... 71 3.3. 2.1 CAT023a (1.0 Wt% Pt/Tio 2 50 mg) ................................ ............... 71 3.3. 2.2 Kinetic Isotope Effect: Reactions with D 2 ................................ ....... 73 3.3.3 Conversion and the Amount of Catalyst ................................ .................. 78 3.3.4 Stereoselectivity and Surface Isomerization ................................ ............ 78 3.3.5 Polarization Transfer ................................ ................................ ............... 87 3.4 Conclusion ................................ ................................ ................................ ........ 89

PAGE 8

8 4 TITANIA SUPPORTED IRIDIUM CATALYSTS: KINETIC STUDIES OF PROPYLENE HYDROGENATION ................................ ................................ ......... 92 4.1 Background ................................ ................................ ................................ ....... 92 4.2 Experimental Methods ................................ ................................ ...................... 93 4.3 Results and Discussion ................................ ................................ ..................... 93 4.3.1 CO Chemisorption ................................ ................................ ................... 94 4.3.2 Catalyst Durability ................................ ................................ .................... 94 4.3.3 Kinetic Studies ................................ ................................ ......................... 95 4.3. 3.1 The Arrhenius Equation and Activation Energy .............................. 96 4.3. 3.2 The Effect of Flow Rate and Relaxation in the PHIP Studies ....... 100 4.3. 3.3 Reaction Orders with Respect to H 2 and Propylene ..................... 104 4.3. 3.4 The Effect of Flow R ate on Reaction Orders ................................ 109 4.3. 3.5 Probing Propylene Desorption Process on Surfaces ................... 111 4.4 Conclusion ................................ ................................ ................................ ...... 112 5 OXIDE SUPPORTED PLATINUM/IRIDIUM CATALYSTS: THE ROLE OF SUPPORT AND PARTICLE S IZE IN HETEROGENEOUS CATALYSIS AND PHIP PERFORMANCE ................................ ................................ ......................... 114 5.1 Background ................................ ................................ ................................ ..... 114 5.1.1 The Effect of Particle Size and Support on PHIP Performance ............. 114 5.1.2 The Effect of Oxide Support on Pairwise Selectivity .............................. 115 5.2 Experimental Methods ................................ ................................ .................... 115 5.3 Results and Discussion ................................ ................................ ................... 116 5.3.1 Pt Catalysts: the Effect of Support, Loading and Quantity ..................... 116 5.3. 1.1 The Arrhenius Equation and Activation Energy ............................ 119 5.3. 1.2 The Effect of Catalyst Amoun t ................................ ..................... 124 5.3. 1.3 The Effect of Gas Composition ................................ .................... 126 5.3.2 Ir Catalyst: the Effect of Support and Loading ................................ ....... 128 5.3.3 Pt & Ir Catalyst: the Comparison ................................ ........................... 129 5.4 Conclusion ................................ ................................ ................................ ...... 130 6 EXTENDING THE LIFETIME OF PARAHYDROGEN INDUCED POLARIZATION THROUGH POLARIZATION TRANSFER ................................ 132 6.1 Background ................................ ................................ ................................ ..... 132 6.1.1 Longitudinal (Spin Lattice) Relaxation T ime Constant T 1 ...................... 132 6.1.2 T 1 Measurement: the Inversion Recovery Method ................................ 132 6.1.3 Hyperpolarization Transfer towards Spins with Extended Lifetime ........ 133 6.1.4 Rationale ................................ ................................ ............................... 134 6.2 Experimental Methods ................................ ................................ .................... 135 6.2.1 NMR Spectrometer and Materials ................................ ......................... 135 6.2.2 PHIP Procedure: the ALTADENA Protocol ................................ ............ 135 6.3 Results and Discussion ................................ ................................ ................... 136 6.3.1 Acetal ization: Equilibration of Propanal and Acetal in Methanol ............ 136 6.3.2 Solvent and Chemical Environment Dependent T 1 Measurements ....... 137

PAGE 9

9 6.3.3 Hyperpolarization Transfer towards Spins with Extended Lifetime ........ 139 6.3.4 Magnetic Field Profile for ALTADENA Experiments .............................. 140 6.3.5 Hyperpolarization Transfer Mechanisms ................................ ............... 142 6.3. 5.1 Nuclear Overhauser Effect (NOE) and 1D NOESY ...................... 142 6.3. 5.2 Magnetization Transfer through J Coupling ................................ 143 6.4 Conclusion ................................ ................................ ................................ ...... 143 7 EXTENDING THE LIFETIME OF PARAHYDROGEN INDUCED POLARIZATION THROUGH SINGLET STATE ................................ .................... 145 7.1 Background ................................ ................................ ................................ ..... 145 7.2 Experimental Methods ................................ ................................ .................... 146 7.3 Results and Discussion ................................ ................................ ................... 146 7.4 Conclusion ................................ ................................ ................................ ...... 146 8 CONCLUSION S ................................ ................................ ................................ ... 149 9 FUTURE STUDIES ................................ ................................ ............................... 152 APPENDIX: DENSITY MATRIX SIMULATION OF ETHYLENE 13 C PARAHYDROGENATION ................................ ................................ .................... 154 LIST OF REFERENCES ................................ ................................ ............................. 155 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 165

PAGE 10

10 LIST OF TABLES Table page 2 1 Magnetic field profile for the ALTADENA reactor ................................ ................ 58 2 2 Normalized integral values of propylene and propane at various flow rates: the relaxation effect. ................................ ................................ ........................... 59 2 3 Transport time (t tr ), effective relaxation time constants (T eff ), relaxation correction factors, and enhancement correction factor C EF for propylene and p ropane. ................................ ................................ ................................ ............. 60 3 1 Platinum particles supported on TiO 2 as catalysts. Courtesy of Dr. Wei Cheng. ................................ ................................ ................................ ................ 91 3 2 Surface isomerization and percentage of trans (anti ) addition product with various total gas flow rate and temperature. ................................ ....................... 91 3 3 Surface isomerization and percentage of trans (anti ) addition product with various hydrogen partial pressure and temperature. ................................ .......... 91 4 1 Iridium nanoparticle supported on TiO 2 as the catalyst. Courtesy of Dr. Wei Cheng. ................................ ................................ ................................ .............. 113 4 2 Effect of total flow rate on the activation energy. ................................ .............. 113 4 3 Reaction orders at various temperatures and total flow rates. .......................... 113 5 1 Platinum nanoparticles supported on oxides as catalysts. Courtesy of Dr. Wei Cheng. ................................ ................................ ................................ .............. 131 5 2 Iridium nanoparticles supported on oxides as catalysts. Courtesy of Dr. Wei Cheng. ................................ ................................ ................................ .............. 131

PAGE 11

11 LIST OF FIGURES Figure page 1 1 PASADENA state state correlation diagram ................................ ...................... 26 1 2 ALTADENA state state correlation diagram.. ................................ ..................... 26 1 3 The Horiuti Polanyi mechanism on palladium. ................................ .................... 29 2 1 Gas handling system for ortho to para hydrogen conversi on. ........................... 36 2 2 Ortho to para hydrogen converter working at liquid N 2 temperature. ................ 37 2 3 Ortho to para hydrogen converter working at liquid helium temperature. .......... 37 2 4 Integrated PHIP gas phase heterogeneous catalytic reactor system under LabVIEW control. ................................ ................................ ................................ 38 2 5 Gas phase heterogeneous catalytic reactor mounted on top of the superconducting magnet. ................................ ................................ ................... 39 2 6 Schematics of ALTADENA reactor (reaction outside the detection field). .......... 39 2 7 Schematics of PASADENA reactor (reaction inside the detection/high field). .... 40 2 8 Overlapped spectra of 50% enriched p H 2 and n H 2 ................................ .......... 41 2 9 Steady state optimization. ................................ ................................ .................. 46 2 10 Static 1 H NMR spectra of propylene and propane obtained in a 10 mm NMR tube at 9.4 T ( 1 H NMR 400 MHz).. ................................ ................................ ...... 48 2 11 Magnetic field profile for ALTADENA reactor based on Figure 2 6. .................... 49 2 12 Flow rate dependence of magnetization buildup for propylene and propane ..... 50 2 13 Estimate the effective relaxation constants (T eff ) for propylene and propane. .... 51 2 14 Flow rate dependence of hydrogen signal and ortho to para H 2 conversion.. ... 53 3 1 Spectra for temperature dependent propyne reduction under ALTADENA condition using CAT009 (0.5 wt% Pt/TiO 2 10 mg). ................................ ............ 66 3 2 Temperature dependent propyne reduction under ALTADENA condition using CAT009 (0.5 wt% Pt/TiO 2 10 mg).. ................................ .......................... 67 3 3 Spectra for temperature dependent propyne partial reduction under ALTADENA condition using CAT023a (1.0 wt% Pt/TiO 2 50 mg) ....................... 68

PAGE 12

12 3 4 Temperature dependent propyne reduction under ALTADENA condition using CAT023a (50 mg). ................................ ................................ .................... 69 3 5 Spectra for temperature and total flow rate dependent propyne reduction under ALTADENA condition using CAT023a (50 mg). ................................ ....... 70 3 6 Temperature and total flow rate dependent propyne reduction using CAT023a (50 mg). ................................ ................................ .............................. 71 3 7 dependent (reaction order) propyne reduction under ALTADENA condition using CAT023a (50 mg). ................................ ................................ .................... 72 3 8 D 2 as the delivery gas: hydrogen partial pressure and temperature dependent (reaction order) propyne reduction under ALTADENA condition using CAT023a (50 mg).. ................................ ................................ ................... 73 3 9 Comparison of conversion ratio and propane/propylene ratio at different temperatures (350 o C/200 o C) in the propyne reduction under ALTADENA condition usi ng CAT023a (50 mg). ................................ ................................ ..... 7 4 3 10 Comparison of ALTADENA spectra obtained with N 2 or D 2 at different temperatures in the propyne partia l reduction under ALTADENA condition using CAT023a (50 mg), (H 2 /propyne = 6/1). ................................ ..................... 75 3 11 Comparison of ALTADENA spectra obtained with N 2 or D 2 at different temperatures in the propyne partial reduction under ALTADENA condition using CAT023a (50 mg), (H 2 /propyne = 12/1).. ................................ .................. 76 3 12 Comparison of ALTADENA spectra obtained with N 2 or D 2 at different temperatures in the propyne partial reduction under ALTADENA condition using CAT023a (50 mg), (H 2 /propyne = 18/1). ................................ ................... 77 3 13 1 H NMR spectrum of thermally polarized static propylene at 1 atm. ................... 79 3 14 Six spin system density matrix simulation of syn and anti addition of p H 2 ..... 80 3 15 Spectra with a linear combination of syn and anti addition. ............................... 81 3 16 The percentage of trans polarization over the sum of trans and cis polarization versus the percentage of anti addition over the sum of anti and syn addition. ................................ ................................ ................................ ....... 82 3 17 Comparison of propylene CH 2 hyperpolarization obtained with different H 2 partial pressure at different temperatures in the propyne reduction under ALTADENA condition using CAT023a (50 mg). ................................ ................. 83 3 18 Comparison of percentage of trans polarization and anti addition at various hydrogen partial pressure and temperature.. ................................ ...................... 84

PAGE 13

13 3 19 Comparison of propylene CH 2 hyperpolarization obtained at different temperatures in the propyne reduction under ALTADENA condition using CAT023a (1.0 wt% Pt/TiO 2 50 mg). Gas flow rate (total 400 mL/min) ............... 85 3 20 Comparison of propylene CH 2 hyperpolarization obtained at different temperatures in the propyne reduction under ALTADENA condition using CAT023a (1.0 wt% Pt/TiO 2 50 mg). Gas flow rate (total 600 mL/mi n) ............... 86 3 21 Comparison of percentage of trans polarization and anti addition at various total gas flow rate and reaction temperatur e ................................ ...................... 87 3 22 Comparison of propylene CH & CH 3 hyperpolarization obtained with different H 2 partial pressure at different temperatures in the propyne reduction under ALTADENA condition using CAT023a (50 mg). ................................ ................. 88 4 1 Catalyst durability test. A&C) Remaining propylene in the product stream. B&D) Produced propane in the product stream. ................................ ................. 95 4 2 Representative (50 350 o C) spectra resulting from reactions with n H 2 and p H 2 for the temperature dependent studies of propylene hydrogenation .............. 96 4 3 3D spectra for temperature dependent catalytic performance under ALTADENA condition ................................ ................................ ........................ 96 4 4 Co mposition of the product stream at various temperatures. ............................. 97 4 5 Temperature dependent studies of propylene hydrogenation under ALTADENA condition ................................ ................................ ......................... 98 4 6 Propylene to propane conversion, enhancement factor and pairwise selectivity. ................................ ................................ ................................ ........... 99 4 7 Temperature dependent studies of propylene hydrogenation at a variety of total flow rates under ALTADENA condition ................................ .................... 100 4 8 The Arrhenius plots of the temperature dependent (50 200 o C) studies of propylene hydrogenation at a variety of total flow rates under ALTADENA condition ................................ ................................ ................................ .......... 102 4 9 Temperature and total flow rate dependent propylene to propane conversion, signal enhancement factor and pairwise select ivity. ................................ ......... 103 4 10 Total flow rate dependent activation energies for different reaction pathways. 103 4 11 Reaction orders with respect to H 2 partial pressure at various temperatures and propylene concentrations under ALTADENA condition ............................ 105

PAGE 14

14 4 12 Propylene to propane conversion and pairwise selectivity with respect to H 2 partial pressure at various temperatures and propylene concentratio ns under ALTADENA condition ................................ ................................ ...................... 106 4 13 Reaction orders with respect to propylene partial pressure at various temperatures under ALTADENA condition ................................ ...................... 107 4 14 Propylene to propane conversion and pairwise selectivity with respect to propylene partial pressure at various temperatures under ALTADENA condition ................................ ................................ ................................ .......... 108 4 15 Reaction orders with respect to A) H 2 and B) propylene partial pressure at 300 o C under ALTADENA condition .. ................................ ............................... 109 4 16 Propylene to propane conversion, enhancement factor and pairwise selectivity with respect to H 2 and propylene partial pressure at 300 o C. ........... 110 4 17 Propylene desorption process. ................................ ................................ ......... 111 5 1 Catalytic and PHIP performance comparison for Pt catalysts at 150 o C under ALTADENA condition ................................ ................................ ...................... 117 5 2 Catalytic and PHIP performance comparison for Pt catalysts at 150 o C under ALTADENA condition ................................ ................................ ...................... 117 5 3 Conversion, enhancement factor and pairwise selectivity for particles with different sizes at various temperatures under ALTADENA condition ............... 118 5 4 Temperature dependence and activation energy for t emperature dependent studies of propylene hydrogenation under ALTADENA condition ..................... 119 5 5 Conversion rate, enhancement factor and pairwise selectivity for t emperature dependent studies of propylene hydrogenation. ................................ ............... 120 5 6 Temperature dependence and activation energy for t emperature dependent studies of propylene hydrogenation under ALTADENA condition .................... 121 5 7 Conversion rate, enhancement factor and pairwise selectivity for t emperature dependent studies of propylene hydrogenation. ................................ ............... 122 5 8 Temperature dependence and activation energy for t emperature dependent studies of propylene hydrogenation under ALTADENA condition ..................... 123 5 9 Conversion rate, enhancement factor and pairwise selectivity for t emperature dependent studies of propylene hydrogenation. ................................ ............... 123 5 10 Temperature dependence and activation energy of propylene hydrogenation under ALTADENA condition . ................................ ................................ ........... 124

PAGE 15

15 5 11 Conversion rate, enhancement factor and pairwise selectivity for t emperature dependent studies of propylene hydrogenation. ................................ ............... 125 5 12 Comparing the effect of catalyst quantity on the c onversion rate, enhancement factor and pairwise selectivity for t empera ture dependent studies of propylene hydrogenation under ALTADENA condition .................... 126 5 13 Reaction orders with respect to h ydrogen partial pressure of propylene hydrogenation at 150 o C under ALTADENA condition using CAT009 (0.5PtTi450). ................................ ................................ ................................ ..... 127 5 14 Convers ion rate, enhancement factor and pairwise selectivity of propylene hydrogenation at 150 o C on CAT009 (0.5PtTi450). ................................ .......... 127 5 15 Representative overlapped spectra for propylene hydrogenation at 150 o C under ALTADENA condition using Ir/TiO 2 with n H 2 and p H 2 ......................... 128 5 16 Temperature dependent propylene hydrogenation under ALTADENA condition on supported Ir catalysts ................................ ................................ ... 129 6 1 The inversion recovery pulse sequence. ................................ .......................... 132 6 2 Propanal and acetal in equilibrium. ................................ ................................ ... 137 6 3 Comparison of T 1 values of propanal and acetal. ................................ ............. 138 6 4 T 1 of propanal in different solvents. T 1 and chemical shift values of propanal in different solvents are different. ................................ ................................ ...... 138 6 5 Hyperpolarized aldehyde through spontaneous polarization transfer from para ....... 139 6 6 Magnetic field variation profile over the distance. ................................ ............. 141 6 7 Magnetic field variation profile over the time. ................................ ................... 141 6 8 1D NOESY of propanal & acetal mixture. ................................ ......................... 142 6 9 Other unsaturated aldehydes. ................................ ................................ ... 144 7 1 Ethylene 13 C parahydrogenation under PASADENA condition. ....................... 147 7 2 Chemical shifts and J coupling constants in ethane 13 C ................................ ... 147 7 3 Comparison of experimental spectra and de nsity matrix simulation ................. 148

PAGE 16

16 LIST OF SCHEMES Scheme page 3 1 Propyne hydrogenation over Pt/TiO 2 ................................ ................................ 63 4 1 Hydrogenation of propylene over Ir/TiO 2 at elevated temperatures. ................... 94 6 1 Para hydrogenation of unsaturated aldehyde. ................................ .......... 135 6 2 Equilibration of propanal and acetal. ................................ ................................ 136 6 3 Para ................................ 140 7 1 Hydrogenation of ethylene 13 C over Pt/TiO 2 ................................ .................... 146

PAGE 17

17 LIST OF ABBREVIATIONS 1D one dimensional ACF activated carbon fiber ALTADENA adiabatic longitudinal transport engenders net alignment BET Brunauer Emmett Teller CSA chemical shift anisotropy DD dipole dipole (interaction) ICP AES inductively coupled plasma atomic emission spectroscopy Ir Iridium MagLab National High Magnetic Field Laboratory MFC mass flow controller min minute MARI M ost Abundant Reaction Intermediate MOFs metal organic frameworks NMR n uclear magnetic resonance NOE nuclear Overhauser effect NOESY nuclear Overhauser enhancement spectroscopy PASADENA para hydrogen and synthesis allow dramatic enhancement of nuclear alignment Pd palladium PHIP P arahydrogen I nduced P olarization Pt platinum REU R esearch E xperiences for U ndergraduates S ABRE signal amplification by reversible exchange sccm standard cubic centimeters per minute SEOP spin exchange optical pumping

PAGE 18

18 SMSI strong metal support interaction TEM transmission electron microscopy TOF t urnover frequency TPR temperature programmed reduction XPS X ray photoelectron spectroscopy XRD X ray diffraction

PAGE 19

19 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy STUDIES OF HYDROGENATION CATALYSIS USING HYPERPOLARIZED NMR By R onghui Z hou May 201 4 Chair: Clifford R. Bowers Major: Chemistry Surface processes involved in heterogeneous catalysis can be effectively studied by in situ /operando techniques. For example, in heterogeneous hydrogenation, molecular hydrogen dissociates into atomic hydrogen upon chemisorption on metal surfaces followed by migration, spillover and even dissolution into the lattice l eading to random addition. Pair wise addition, where the two hydrogen atoms come from a single molecule and retain their nuclear spin correlation is not typically considered as a distinct proc ess. Nevertheless, surface reaction activit ies and pathway s have been shown to be affected by the presence of reactants, intermediates, products and by products such as carbon aceous deposits. Other factors affecting heterogeneous catalysis include the natu re of the catalyst, such as metal type, particle size, s hape, morphology, dispersion as well as the properties of the supports, especially those with strong metal su pport interactions (SMSI). Such factors are also important in determining the favorability of pairwise addition. While it would be extremely difficult to distinguish between these two pathways using conventional methods such as gas chromatography, hyperpolarized NMR based on Parahydrogen Induced Polarization (PHIP) provides a

PAGE 20

20 unique tool to stu dy pairwise hydrogen addition, since only this pathway leads to dramatically enhanced ( up to 10 5 ) NMR signals. In this study, a variety of catalysts were prepared with transition metals (Pt, Ir, etc.) supported on different oxides (Al 2 O 3 TiO 2 etc.), thr ough precipitation of metal salts onto supports from an aqueous solution by controlling solution pH followed by aging and calcination. For PHIP experiments, a home built micro reactor with heating unit wa s installed on the top of the magnet. The precise de livery of gas mixture (propylene p H 2 and carrier gas ) wa s facilitated with Mass Flow Controller s (MFCs) T he resultant hyperpolarized propane flew down to the NMR probe for detection. Kinetic parameters, such as react ion temperature, gas composition and flow rate were varied to optimize PHIP enhancement. Kinetic studies about reaction orders and activation energies were made possible with the dramatic signal enhancement. In addition, b y comparing resultant NMR spectra from reactions with n H 2 or p H 2 co ntributions from pairwise addition was able to be derived.

PAGE 21

21 CHAPTER 1 INTRODUCTION 1.1 Hyperpolarization : Improvement o f NMR Sensitivity Nuclear magnetic resonance spectroscopy (NMR) is a valuable tool for obtaining structural and dynamical information of molecules. However th is method suffers from poor sensitivity compared to m ost other types of spectroscopy. This situation is even worse in gas phase where the nuclear spin density is rather sparse, roughly three orders of magnitude lower than in liquids. Th e intrinsic low sensitivity due to the small population difference between nuclear spin states determined by the Boltzmann distribution, can be solved to some extent, by va rious hyperpolarization methods. 1.1.1 Pushing Thermal Equilibrium NMR signal i s proportional to polarization which is defined as the difference over the sum of the populations in the case of spin 1 /2 nuclei such as 1 H shown in E quation ( 1 1 ). At room temperature and 11.7 Tesla ( 1 H: 500 MHz), this population difference is around 0.00 008. Th us NMR is a rather insensitive technique. ( 1 1 ) One way to increase the polarization is to increase the magnetic field or decrease the temperature as seen in the following E quation ( 1 2 ). ( 1 2 )

PAGE 22

22 1.1.2 Polarization Transfer f rom Electron o r Photo n t o Nucleus 1.1.2.1 Dynamic Nuclear Polarization (DNP) Nuclear spin polarization can be enhanced by transferring the polarization from that of electrons known as the nuclear Overhauser effect. 1 2 T his is the basis of dynamic nuclear polarization (DNP) and widely used in lots of applications, including solid state NMR experiments of biomolecules. It relies on the fact that electron has a much large r population difference between the two spin states based on the Boltzma nn statistics 1.1.2.2 Optical N uclear P olarization: Optical ly Pump ed NMR In this method, the polarization of circularly polarized light can be transferred to electron or nuclear spins through specially designed pulse sequences, leading to hyperpolarizat ion. Four to five order s of magnitude of sensitivity enhancement has been achieved by optically pumped NMR ( OPNMR ) and this method is commonly employed in the studies of semiconductor band structures 3 1.1.2. 3 Spin Exchange Optical Pumping (SEOP) Nob l e gases such as Xenon 129 Helium 3 and Kr 89, can be hyperpolarized relatively easily with spin exchange optical pumping method. 4 8 This is achieved by polarization transfer from circularly polarized light to noble gas via alkali metal electrons. 9 Hyperpolarized Xenon is widely used in diffusio n studies and magnetic resonance imaging. 1.1.3 Polarization Transfer f rom Nucleus t o Nucleus The n uclear Overhauser e ffect (NOE) is a common phenomenon in resonance where the spin polarization of one nucleus is transferred to another via cross relaxation.

PAGE 23

23 be enhanced by the microwave irradiation of the conduction electrons in certain metals. 1 NOE effect is widely used in NMR for characterizing organic chemical structures and even the three dimensional struct ures of biological macromolecules in solutions by determin ing the sp atia l correlations between nuclei. Transferring polarization from high (such as 1 H) to low (such as 13 C) nucleus in solid s is used routinely to improve the sensitivity and well cross polarization 10 1.1. 4 Parahydrogen Induced Polarization (P HIP ) In the late 1980s, a new concept was proposed 11 and shortly demonstrated 12 by Bowers and Weitekamp that dramatic signal enhancement c ould be achieved by converting the singlet nuclear spin order of parahydrog en into hyperpolarization, through catalytic hydrogenation reactions. This method initially observed but misinterpreted, 13 is called P arahydrogen I nduced P olarization ( PHIP ) 14 20 Up to 5 orders of magnitude is achievable and it is possible to observe transient i ntermediates and distinguish reaction pathways. Thus it soon became a powerful mechanistic tool in homogeneous catalysis. 17 21 24 Besides, the observation of PHIP in heterogeneous catalysis 25 allows a direct visualization of mechanisms on metal surfaces and production of metal free hyperpolarized fl uids for novel biomedical applications. 26 1 .2 Basics of PHIP 1.2.1 Nuclear Spin Isomers: Ortho a nd Para Hydrogen The existence of spin isomers of molecular hydrogen was first predicted by Werner Heisenberg b ased on the Born Oppenheimer Approximation and the Pauli Exclusion Principle There are two isomers, namely ortho hydrogen and para hydrogen (Equation 1 3) According to the Pauli Exclusion Principle the total wavefunction of

PAGE 24

24 fermions (a class of partic le s to which protons belong) is anti symmetric with respect to permutation of the coordinates of the two nuclei. The overall wavefunction of molecular hydrogen consists of electronic, translational, vibrational, rotational and nuclear contributions For th e ground state, electronic, translational, and vibrational components are all symmetric, thus singlet para hydrogen ( p H 2 ) is associated with even rotational quantum states and triplet ortho hydrogen ( o H 2 ) is with odd states. ( 1 3 ) The ratio between ortho and para hydrogen depends on the rotational partition functions ( Equation 1 4 ). 14 Thus p H 2 can be enriched to 50% at liquid n itrogen temperature (77 K) and 99% at 20 K. Transitions between singlet and triplet states are magnetic dipole forbidden without catalysts (FeO(OH), activated carbon, etc.), allowing for extended storage of the p H 2 enric hed gas. This is also the reason why p H 2 is NMR silent and the p H 2 concentration can be measured using the signal intensity which is directly proportional to the o H 2 concentration. (1 4)

PAGE 25

25 1 .2.2 PHIP: PASADENA v ersus ALTADENA Parahydrogen induced polarization utilizes the ability of the singlet state nuclear spin order of p H 2 to be converted into observable nuclear spin magnetization by the addition of the p H 2 molecule to an unsaturated substrate. There are two variants of P HIP based on the procedure: 1) PASADENA (Para hydrogen And Synthesis Allow Dramatic Enhancement of Nuclear Alignment) ( Figure 1 1 ) 12 and 2) ALTADENA (Adiabatic Longitudinal Transport Engenders Net Alignment) ( Figure 1 2 ) 27 In the case of PASADENA, the reaction and NMR detection are both carried out in high magnetic fields. D ramatically enhanced anti phase multiplets are observed which can be rationalized based on sudden approximation The transition probability from an initial state to a final state can be calculated from the square of the projection of the initial state to the final state. In ALTADENA case, fields and detection is in high fields following an adiabatic transport to the magnet. This relies on both sudden approximation for the product formation and then adiabatic transport to high fields. There is no further change in the internal Hamiltonian when the sample is transferred to the magnet until the detection by NMR. From Figure 1 1 it is clear that the use of ortho enric hed hydrogen would also lead to hyperpolarization, though with opposite phases to the para hydrogen derived signals and much less degree of hyperpolarization.

PAGE 26

26 Figure 1 1 PASADENA state state correlation diagram A ) Left and right show the singlet a nd triplet states of p H 2 and o H 2 respectively, and the middle is the spin states of the adduct as a weakly coupl ed two proton (AX) system. The population distribution of the adduct is found by projecting the initial state onto the final states, under th e sudden approximation B ) Conventional NMR spectrum of the adduct under thermal equilibrium. C ) PASADENA spectrum, exhibiting antiphase doublets. Figure 1 2 ALTADENA state state correlation diagram. A ) Left shows the singlet of p H 2 middle is the adduct state in low magnetic fields, and right shows the final state after adiabatic transport to the magnet. B ) Conventional NMR spectrum of the adduct under thermal equilibrium. C ) ALTADENA spectrum, exhibiting net alignment.

PAGE 27

27 1.2. 3 Signal Amplification b y Reversible Exchange (SABRE) The requirement of an unsaturated precursor for PHIP drove the development of a new hyperpolarization technique without the incorporation of p H 2 molecules into the substrate This was termed as SABRE (S ignal Amplification by Reversible Exchange). 28 30 This polarization transfer from para hydrogen t o the substrate molecule is accomplished through the J coupling when both of them bind to transition metal complexes at low magnetic field. SABRE greatly extended the scope of p arahydrogen i nduced p olarization (PHIP) to molecules which are not readily acc essible by hydrogenations. 1.2. 4 Basic Requirement s f or Observing P HIP To observe PHIP effects, all of the following conditions have to be fulfilled: 1. Pairwise addition: the two protons from p H 2 must be transferred to the substrate in such a way that their spin order is preserved. 2. Symmetry break: the hydrogen nuclei must be deposited into magnetically inequivalent sites while retaining a spin coupling between them. 3. Weak coupling: the chemical s hift difference of the two coupled spins (in the detection magnetic field) has to be much greater than the J coupling constant. 4. Rela xation limit: the reaction time scale must be shorter than the lifetime of hyperpolarization, approximately the spin lattice relaxation time constant (T 1 ). 1 3 PHIP in H eterogeneous C atalysis Interest in PHIP has been increasing ever since its first discovery It has increased further when the ability to obtain rapid assays and images of metabolism in living tissues was demons trated using PHIP hyperpolarized molecular probes. 31 32 To extend such studies to humans, it is critical to have catalyst free hyperpolarized molecular tra cers, which is possible in HET ( heterogeneous ) PHIP. 33 34 Besides the

PAGE 28

28 easi ness of separation of hyperpolarized fluids from the catalyst, it is also easy to control the whole hydrogenation process in heterogeneous catalysis T he dramatic signal enhancement enable s mechanistic studies on heterogeneous catalysis itself and in situ/ operando studies of reactors Based on the dissociative Horiuti Polanyi mechanism m olecular hydrogen readily dissociates into atomic hydrogen, followed by migration, spillover and even dissolution into the metal lattice. 35 This creates a pool of surface hydrogen atoms available for reactions 36 but also causes the loss of the spin order of p H 2 In such cases, random addition dominates over pairwise addition. However the dramatic signal enhancement due to PHIP ma de a small (< 3 %) 37 pairwise contribution detectable in heterogeneous catalysis with immobilized Rh(I) complexes (RhCl(PPh) 3 ), 15 38 39 Iridium complexes, 40 Au(III) Schiff base complexes immobilized within a metal organic framework, 41 and supported Pt/Pd metal nanoparticles. 42 The surface reaction activity and pathway have been shown to be affected by the presence of reactants, intermediates, products and by products such as carbon deposits 43 whose presence parti tions the vast metal surface into isolated islands and lead s to localization of catalytic sites. This can hinder the ready migration of H atoms and increase the chance of pairwise addition of two H atoms coming from the same hydrogen molecule. Other factor s affecting heterogeneous catalysis include the nature of the catalyst, such as metal type, particle size, shape, morphology, dispersion and loading, as well as the properties of the supports, especially those with strong metal su pport interactions ( SMSI ). 37 44 47

PAGE 29

29 To further expand the scope of PHIP, bulk metals, such as Pt black, and metal oxides, such as PtO 2 CaO, Cr 2 O 3 ZrO 2 CeO 2 were reported to produce PHIP recently, attributed to the reduced mobility of surface H atoms related with the strength of adsorbate support interaction. 48 Even more interestingly, metal free compounds ansa aminoborane tweezers were found to exhibit PHIP effect recently 49 1. 3 1 Kinetics a nd Mechanisms o f Reactions o n Metals Kinetic stu dy is typically conducted to evaluate the microscopic impacts of macroscopic parameters such as relative ratios between reagents, their concentrations, pressures, temperatures, and flow rates in a continuous flow reactor. 50 52 1. 3 1 1 The Horiuti Polany i M echanism Figure 1 3 The Horiuti Polanyi mechanism on palladium. Pd atoms are blue, carbon atoms are black, hydrogen atoms are green, and deuterium atoms are red. Step 1. Alkene adsorbs and hydrogen dissociates into atomic hydrogen on the metal surface. Step 2. One hydrogen atom migrates and adds reversibly bond between the metal and Carbon or a alkyl moiety. Step 3. A second hydrogen atom adds irreversibly to form the product. 55 Reprinted with permiss ion from J. Chem. Educ. 2013 90 (5), pp 613 619 Copyright (2013) American Chemical Society.

PAGE 30

30 Horiuti Polanyi mechanism is the p revailing mechanism in heterogeneous hydrogenation catalysis at typical reaction temperatures, 53 consisting of three steps (Figure 1 3 ): (1) alkene adsorbs and hydrogen dissociates into atomic hydrogen on the metal surface, (2) one hydrogen atom migrates and adds reversibly to alkene with the formation of a bond (or a al kyl moiety) between the metal and c arbon 54 and finally (3) in a reductive elimination step a second hydrogen atom adds irreversibly to form the product. 55 1. 3 1 2 The Active Sites o n Surfaces There are two kinds of adsorption: molec ular and dissociative adsorption. Generally, it is assumed tha t there is a random distribution of non interacting adsorbed species on the surface where all sites are equal and all geometric sites are active This is the so called mean field approximation, 56 58 and can only be fulfilled at low coverage or high tempe rature. While at high coverage and low temperature, the interactions, being attractive or repulsive, between adsorbed species can have significant impact s on the kinetics. In contrast, Taylor proposed that only certain sites or centers on surfaces are cat alyzing the reactions. 59 Those active sites, composing of an ensemble of atoms, might be sparsely and randomly distributed on the surface and prone to poisoning by just a few molecules or species. Obviously, flat surfac e atoms or terrace sites are more highly coordinated than those located at corners, edges, surface steps or kinks. Three distinct adsorption sites as atop, bridge, and hollow co exist on (111) face of a face centered cubic (fcc) metal such as platinum or i ridium. 60 Accordingly, reactions can also be categorized into two groups: structure sensitive (demanding) and structure insensitive (facile) reactions. 61 65 Catalyst

PAGE 31

31 optimization is of critical impo rtance for st ructure sensitive processes, where local atomic structure gives rise to the variations of adsorbate bond strengths and reaction kinetics. 62 66 Single site catalysts where either a single ato m (or ion), or a very small cluster of atoms serves as the reaction center, were reported and expected to be highly selective due to its uniformity across the surface. 67 1. 3 1 3 T he MARI A pproximation In a lot of cases, one of the reagents or intermediates adsorbs more strong ly th an the other participants and to a good extent dominates the surface. This intermediate is called the Most Abundant Reaction Intermediate ( MARI ) 58 1. 3 1 4 T he Proximity Effect The d electrons of metals are normally actively involved in chemisorption, weakening chemical bonds and making them much easier to break. It is well accepted as the proximity effect that bonds only in close proximity to metal surfaces can break. 58 1. 3 1 5 The Time Scale i n Catalysis A complete catalytic cycle is more than just bonds breaking and reforming. Often, surface processes such as diffusion and transport are much slower and limit the reaction kinetics. Typically, microscopic chemical bond breaking and formation happen in the s cale of picoseconds while the macroscopic diffusion through metal catalysts and porous supports takes between seconds to minutes. 58 The time scale of PHIP reactions should lie within the lifetime ( spin lattice relaxation time T 1 ) of hyperpolarized signal s 1 3 2 T he Arrhenius Equation a nd Activation Energy A common phenomen on is that the rate of catalytic r eactions goes up when temperature increases. A potential energy barrier has to be overcome as, for instance, it costs energy to break chemical bonds. This relationship is governed by the Arrhenius

PAGE 32

32 E quation ( 1 5 ) which can be re written in the logarithmic scale ( 1 6 ) At certain conditions diffusion phenomena start to impose influences on reactions and limit the kinetics. ( 1 5 ) ( 1 6 ) In a more practical situation, for low temperatures t he rate increases with increasing temperature and t he apparent activation energy is positive. However, a t a certain temperature the diffusion and transport of reactants start to limit the kinetics as the reaction rate becomes so high. A concentration gradient between the bulk gas and the s urface starts to develop, resulting in a slower rate increase. A t even high er temperatures the residence time of adsorbed species decreases and there is a lack of surface bound re a gents. T he rate thus decreases with temperature resulting in a negative ap parent activation energy 58 1 3 3 Rate Equation a nd t he Reaction Orders For the propylene hydrogenatio n reaction: ( 1 7 ) The rate of reaction can be expressed in the following equation: ( 1 8 )

PAGE 33

33 Here r is the reaction rate, and k is the rate constant and assumed to be temperature independent For an easier manipulation, this equation is re arranged to the logarithmic scale as follows: ( 1 9 ) The reaction order is the corresponding slope of the linear fitting when the natural log of reaction rate is plotted against the natural log of the concentration of that species. It represents the extent of the effect of the concentration of a species on the rate, and also provides information about which species affects t he rate more significantly 1 4 Applications of PHIP 1 4 1 Producing Hyperpolarized Tracers f or Biomedical Applications The use of heterogeneous catalysts for generating PHIP provides a number of significant advantages over the homogeneous processes, inc luding the possibility to produce hyperpolarized gases, better control over the hydrogenation process, and the easi ness of separating hyperpolarized fluids from the catalyst. The latter advantage is of paramount importance in light of the recent tendency t oward s utilization of hyperpolarized substances in in vivo spectroscopic and imaging applications of NMR. 68 70 The production of hyperpolarized propane 71 and its utilization in gas phase NMR imaging were initially demonstrate hydrogenation. 34 72 73 1 4 2 Elucidating t he Kinetics a nd Mechanisms o f Reactions o n Surfaces In addition, the dramatic signal enhancement associated with PHIP makes it a sensitive tool for in situ studies of the kinetics and mechanisms of operating catalytic reactors. Short lived and trace amount of reaction intermediates are now possible to be

PAGE 34

34 identified, which provides insight into the whole process. Activation energies and reaction orders associ ated with pairwise addition are valuable information towards understanding the fundamentals of different surface reactions. PHIP has been applied successfully to the kinetic studies in heterogeneous hydrogenations 74 76 and reaction orders with respect to either n or p H 2 were reported under PASADENA protocol (reaction in high magnetic fields) over Pt/Al 2 O 3 75 1 4 3 Operando Spectroscopy a nd Catalysis Imaging Solid state NMR spectroscopy has been used in the in situ/operando studies of reaction kinetics and catalyst deactivation in heterogeneous catalysis. 77 78 Y et the sophisticated experimental setup related with in situ MAS (Magic Angle Spinning) NMR, the difficult y to mimic industrially relevant high pressure and high temperature processes, and the requirement of large signal average due to poor sensitivity lim it the scope of potential applications. On the other hand, t he existence of microscopic fluid flow, as well as mass and heat gradients within a heterogeneous reactor bed complicates the reactor design and catalyst optimizations. Although computational mode lling is the routine method to approach the coupling between heat transfer, fluid dynamics and surface reaction kinetics, a direct visualization of the flow distribution, local density profiles and temperature maps is critical to validate such models. 79 The significant signal enhancement of PHIP can be extended to magnetic resonance imaging (MRI) of operating catalytic reactors. Bouchard and co workers were able to image the hyperpolarized propane in the void space of the reactor, 33 and the gas velocity map and density of active catalyst (silica gel in a packed bed micro reactor. 34 They even mapped the gas temperatures for metal

PAGE 35

35 nanoparticles catalyzed pro pylene reduction, taking advantage of the inverse relationship between NMR linewidths and the temperature caused by motional averaging in a weak magnetic field gradient. 80 This motional averaging NMR method based on PHIP is much more sensitive than the conventional NMR thermometry method fo r instance, based on the temperature sensitive characteristics of 27 Al signal of alumina, 81 and should find wide applications in the operando thermodynamic studies of heterogeneous catalysis.

PAGE 36

36 CHAPTER 2 INSTRUMENTATION AND EXPERIMENTAL METHODS 2 .1 G as H andling S ystem and H eterogeneous C ataly tic R eactor To start with, the design and construction of a gas handling system enabling ortho to para hydrogen conversion and continuous flow reaction is necessary. T he whole system incorporates all desired functions and allows easy manipulations: 1) Pressures can be monitored everywh ere with pressure gauges and a relief valve ensure s safety (Figure 2 1) 2) The o / p H 2 conversion system can be purged with nitrogen to remove any absorbed gas or moisture, and evacuated with a high vacuum turbo pump. 3) p H 2 can be enriched to 50% with a converter working at 77 K in liquid nitrogen (Figure 2 2) or to 99% with a converter working at 20 K in liquid helium (Figure 2 3) 4) Catalysts can be activated in situ in the NMR tube (PASADENA) or in a U shaped glass tube reactor (ALTADENA) outside the magnet at elevated temperatures by flowing through N 2 gas or reducing agent (H 2 ) (Figure 2 4) 5) Gaseous reaction mixtures can be prepared and charged to the reactor under precise control of three Mass Flow Controllers (MFCs ) (Figure 2 4) 6) The gas flow rate can be easily adjusted with a LabVIEW program connected to MFCs. 7) Reaction temperature can be well maintained up to 400 o C for steady state reaction conditions. 8) The exhaust gas is vented out of the building directl y. Figure 2 1. Gas handling system for ortho to para hydrogen conversion

PAGE 37

37 Figure 2 2 Ortho to para hydrogen converter working at liquid N 2 temperature Figure 2 3. Ortho to para hydrog en converter working at liquid helium temperature In collaboration with Prof. Helena Hagelin flow gas phase reactor was constructed ( Figure 2 4). The precise delivery of gas es (N 2 n / p H 2 and propylene /propyne is facilitated with three mass flow controllers (MFCs) T he reactor ( Figure 2 5) is made of a U shaped glass tube and placed inside a tube furnace

PAGE 38

38 for steady state reaction conditions at variable temperatures (up to 400 o C). Schematics of ALTADENA and PASADENA reactors are shown in Figure 2 6 and 2 7 respectively. The whole sy s tem is under easy control from LabVIEW program which allows easy adjustment of flow rates and temperatures, as well as NMR acquisition triggering and feedback. It is also capable of produc ing a stream of hyperpolarized products (such as hyper polarized propane) for further applications in diffusion and MRI Figure 2 4. Integrated PHIP gas phase heterogeneous catalytic reactor system under LabVIEW control.

PAGE 39

39 Figure 2 5 Gas phase heterogeneous catalytic reactor mounted on top of the superconducting magnet Figure 2 6 Schematics of ALTADENA reactor (reaction outside the detection field).

PAGE 40

40 Figure 2 7 Schematics of PASADENA reactor (reaction inside the detection /high field). 2 .2 Parahydrogen E nrichment and O rtho to P ara hydrogen C onverter The converter (Figure 2 2) is a helical coil of copper tubing which is filled with a spin catalyst currently, activated carbon. This converter is designed to work in liquid N 2 producing 50% enriched p H 2 (ortho /para =1:1). 2 .2.1 Preparation o f 50% Enriched p H 2 For an easy operation, p H 2 is enriched to 50% by passing n H 2 into the converter working at liquid N 2 temperature ( Figure 2 2 ). A typical procedure is as follows: (1) the converter is purged with N 2 overnight to remov e any adsorbed air or moisture, (2) it is then evacuated under rough pump for several hours followed by turbo pump for another hour, (3) it is pressurized with n H 2 to around 2 atm and immersed into liquid N 2 while connected to the n H 2 source (4) wait for 2 hours to ensure the whole converter is maintained at 77 K, (5) finally 50% enriched p H 2 is produced and ready for

PAGE 41

41 experiments. Pressure is monitored on both sides of the converter and a relief valve is attached as well to ensure a safe op eration ( Figure 2 1 ). 2 .2. 2 Preparation o f 99 % Enriched p H 2 The other converter for 99% enrichment ( Figure 2 3 ) is designed to work inside a liquid helium dewar. The heli cal coil filled with a spin catalyst is held within a vacuum space filled with low pressure helium as the heat exchange gas. 20 K is maintained by the helium gas together with a heating wire and silicon diode as the temperature sensor. Highly enriched p H 2 is expected to furth er increase the sensitivity of PHIP 2 .2. 3 Determining t he Ortho /Para Hydrogen Ratio b y NMR The NMR determination of para enrichment is based upon the fact that singlet p H 2 is NMR silent but o H 2 is not. As n H 2 is composed of 75% o H 2 enrichment to 50% p H 2 (and 50% o H 2 ) would result in a decrease in the NMR signal by one third (Figure 2 8 ) Comparison of peak integrals of both spectra indicated a 50% enrichment of p H 2 All p H 2 used in this dissertation was actually 50% enriched p H 2 Spectra obtained for reactions using 50% enriched p H 2 were subtracted with that obtained using n H 2 and regarded as p H 2 results throughout this dissertation. Figure 2 8 Overlapped spectra of 50% enriched p H 2 and n H 2 (after NMR probe background subtraction) Orange: 50% enriched p H 2 blue: n H 2 Comparison of peak integrals indicate d 50% enrichment.

PAGE 42

42 2 3 Preparation of S upported M etal as H eterogeneous C atalysts All supported metal catalysts were prepared and characterized by Dr. Wei Cheng ( and some Iridium catalysts together with Steven Geller ) in Prof. Helena Hagelin laboratory in the Department of Chemical Engineering at the University of Florida. 82 83 Iridium or Platinum nanoparticles supported on oxide s were prepared by precipitation of corresponding precursors onto the supports from an aqueous solu tion controlling solution pH using sodium hydroxide. 2 3 1 Preparation o f Supported Platinum ( Pt ) Catalysts 1.98 g of TiO 2 support (Alfa Aesar) was dispersed in 100 m L of deionized water under constant stirring. H 2 PtCl 6 6H 2 O (0.053 g) was dissolved in 5 m L deionized H 2 O and was poured into the support/water mixture. Platinum hydroxide was deposited onto the support by dropwise addition of a 2.5 mM NaOH solution until the pH value of the catalyst dispersion mixture reached 11. The catalyst dispersion mixtur e was then aged overnight under continuous stirring, and titrated with a diluted acetic acid solution to pH 7. Then, the mixture was filtered, re dispersed in deionized water, and filtered a second time after stirring overnight. The aging step was necessary to prepare a reproducible and active catalyst, and washing was necessary to remove residual sodium ions and any chloride contaminant as these can accumulate on the surface and reduce activity. The p re catalyst was dried at 105 C overnight, and then calci ned at 350 C for 3 hours. 2 3 2 Preparation o f Supported Iridium ( Ir ) Catalysts 1.98 g of TiO 2 support was dispersed in 100 m L of deionized water under constant stirring. H 2 IrCl 6 6H 2 O (0.056 g) was dissolved in 5 m L deionized H 2 O and poured into the suppo rt/water mixture dropwise. Ir nanoparticle was deposited onto the

PAGE 43

43 support by titrating the above solution with a 2.5 mM NaOH solution until the pH of the dispersion mixture reaches 11 under vigorous stirring at room temperature. The mixture was then aged o vernight under continuous stirring before it was filtered, re dispersed in deionized water. It was then filtered a second time. The re dispersed catalyst was filtered and dried at 105 C overnight, and then calcined at 450 C for 3 hours. 2 4 Characterizat ion of H eterogeneous C atalysts 2 4 1 Determine t he Actual Metal Loading w ith ICP AES The metal loading used in this dissertation were reported as the weight percentage in the initial preparation step. For a more accurate measurement, t he actual metal loading can be characterized by inductively coupled plasma atomic emission spectroscopy (ICP AES) on a Perkin Elmer Optima 3200 RL In short, catalyst should be fused with sodium peroxide at 500 C for 1 hour, followed by dissolution in water a nd neutralization with hydrochloric acid. The concentration of metal species in fusion solution will be analyzed by ICP AES measurement, and this information can be used for calculating the actual metal loading in the catalyst. 2 4 2 Temperature P rogramme d R eduction M easurements Temperature programmed reduction provides useful information on the temperature needed for the complete reduction of a catalyst. This was performed using a Quantachrome ChemBET 3000 instrument. For the measurement, 200 mg of the ca talyst was loaded in a quartz tube reactor, and secured with a plug of quartz wool at the gas flow outlet end. The temperature in the catalyst bed was monitored by an Omega K type thermocouple. Reduction condition was optimized at a heating rate of 10 o C p er minute up to a temperature of 500 o C A gas flow of 5% H 2 /N 2 mixture at a total

PAGE 44

44 flow rate of 70 sccm was used for reduction until no further hydrogen adsorption on the catalyst was observed. 2 4 3 Metal Dispersion a nd Particle Size: t he C hemisorption Measurement Carbon monoxide (CO) c hemisorption is a powerful technique for measuring the surface area of supported metal catalyst s The number of surface metal atoms can be calculated based on the uptake of CO molecules, provided the stoichiometry of CO to metal atoms on the surface is known Further information about the metal dispersion and particle size can be derived. In this study, chemisorption measurements w ere performed on a ChemBET 3000 workstation ( Quantachrome Instruments ) The fresh catalysts w ere first oxidized at 5% oxygen in helium at T C ( T equals 350 for Ir and 170 for Pt respectively) for 0.5 h and outgassed in helium, before the catalyst was reduced in 5% hydrogen in nitrogen at T C for 0.5 h. The oxidation and reduction cycle provide t he catalyst yield a well defined reduced surface on the catalyst. The catalyst was outgassed in helium at T C for another 15 minutes in order to remove all physically absorbed hydrogen on the catalyst surface before cooling down to room temperature. The activation temperature T was chosen for performing hydrogen reduction based on the information obtained from temperature programmed reduction (TPR) measurement, which indicate d that most of the metal species was reduced at T C. The catalyst was then subje cted to CO adsorption measurements to determine the metal surface area. By titrating a known amount of adsorbate gas (CO) into a stream of inert makeup gas (helium), the titrated CO gas stream would flow through the sample surface, and the probing gas CO w ould adsorb onto the exposed metal surface until it is saturated

PAGE 45

45 with CO. The amount of CO gas in the makeup gas helium stream not adsorbed on the metal surface wa s measured by a thermal conductivity detector (TCD). By subtracting known amount of unabsorbed CO from TCD signal from total amount of titrated CO, the absorbed CO gas can be quantitatively calculated Estimates of the metal particle sizes were made from these CO adsorption measurements (Equation 2 1) 83 (2 1) Here av is the average particle size, S av is the average stoichiometry: CO/Ir= 1 ; 84 85 k is the shape factor, and a value of 5 was chosen for Ir in this study, which related to a cube with one side attached to the support and five sides opposed to the environment; V m is the molar volume; N a is the Avogadro's number; m is the metal density; V g is the volume of gas adsorbed; C m is the surface density of metal atoms. The value 1.16 10 15 atoms/cm 2 was used in this study for Ir Similarly, the iridium metal surface area (S Ir ) can be calculated according to the following E quation 2 2 : ( 2 2) 2 4 4 Transmission Electron Microscopy ( TEM ) Measurement Thin film of catalyst solutions (dispersed in isopropanol) was drop cast ed onto a 400 mesh ultrathin carbon film su pported copper grids (TED PELLA I nc. USA). TEM micrographs were obtained in a microscope JEOL Model JEM 2010F System in the Major Analytical Instrumentation Center (MAIC) at UF

PAGE 46

46 2 5 NMR Experiments 2 5 .1 Reactor Setup The continuous flow gas phase heterogeneous cataly tic reactor system was employed for these studies. This system provid es a well controlled reaction system, highly reproducible thanks to the precise control of gas delivery using m ass flow controllers (MFCs). The position of the thermocouple inside the reactor and also by the outlet of the catalyst bed provides highly accurate temperature control. While for the heterogeneous PASADENA case, 75 the temperature was obtained based on the th ermocouple in the probe outside the NMR tube, and wa s not a direct measurement. As the reaction i s so exothermic, this approach is much more reliable. 2 5 2 Steady State Optimization Figure 2 9 Steady state optimization. Different amounts of time wer e required to achieve steady state conditions. Reactor was set to 75 o C under a N 2 flow of 200 mL/min before switching to reaction gas mixture s A significant amount of N 2 was necessary to dissipate heat.

PAGE 47

47 As steady state condition is critical to kinetic studies, and hydrogenation reactions were normally very exothermic, it is important to optimize the reaction conditions. In this set of experiments, the reactor was set at 75 o C under a N 2 flow of 200 mL/min be fore switching to reaction gas mixtures of different ratios among N 2 /H 2 /propylene. Results (Figure 2 9) indicated that reactor temperature started to increase upon switching, and different amounts of time were required to get stabilized. Without N 2 dilutio n, the reactor temperature could not get back to 75 o C even after 30 min. Thus it is important to introduce N 2 as the heat dissipater. Reaction conditions were optimized based on the total gas flow rate accordingly. 2 5 3 PHIP Procedure All NMR experiments were carried out on a Bruker Avance 400 MHz (B 0 = 9.4 Tesla) NMR spectrometer. Certain amount of the catalyst was placed on the bottom of the reactor. Prior to experiments, the catalyst was activated under a stream of H 2 (100 mL/min ) for 30 minutes at 350 C (for Ir catalysts) or 170 C (for Pt catalysts) followed by purging under a stream of N 2 ( 4 00 mL/min ) for 15 minutes at the same temperature and then set to designed experimental temperature under the same N 2 stream. The precise delivery of reaction gas mixture of N 2 (as carrier gas), H 2 (either n or p ) D 2 and propylene /propyne was facilitated with mass flow controllers (Alicat Scientific, Inc.). p H 2 was produced in situ by passing n H 2 charcoal and immersed in a liquid N 2 (77 K) Dewar. The gas mixture fl ew through the reactor maintained at elevated temperatures (50 350 C) and then into the NMR tube positioned inside the magnet for detectio n. The reaction was exothermic and temperature arose upon switching the gas from N 2 to reaction mixture. To reduce the heat generation and time taken to reach steady state condition, a considerable amount

PAGE 48

48 of N 2 wa s introduced into the reaction mixture as t he carrier gas and heat dissipater in most cases It should be noted that this also significantly reduced the possibility of achieving high signal enhancement. Upon reaching steady state reaction conditions, NM R spectra were acquired with multiple scans us ing a 90 degree pulse for ALTADENA conditions and a 45 degree pulse for PASADENA conditions 2 6 The E ffect of F low R ate : R elaxation and M agnetization B uildup 2 6 .1 Static Spectra o f Propylene, Propane a nd t he Mixture Figure 2 10 Static 1 H NMR spectra of propylene and propane obtained in a 10 mm NMR tube at 9.4 T ( 1 H NMR 400 MHz). A) Propane. B) Propylene and propane mixture. C) Propylene. Figure 2 10 shows the spectra for static propylene, propane and their mixture, all at around 1 atm. All peaks were well separated from each other and assign ed to specific proton accordingly. Special attention should be paid to the CH 3 group of

PAGE 49

49 propylene (with a chemical shift of 1.72 ppm) and the CH 2 group of propane (with a chemical shift of 1.44 ppm), as these two w ere used for conversion and enhancement factor calculations in this dissertation For ALTADENA experiments the CH 2 and CH 3 groups of propane should be expected to show net alignment with CH 2 pointing up and CH 3 pointing down. 2 6 .2 Magnetic Field Profile The magnetic field profile for this ALTADENA reactor (Figure 2 6) was measured with a Lakeshore Gauss meter Zero point is the point of product formation in the catalytic reactor on top of the magnet. The volume of g as flow path was calculated according t o the si z e of the tube connecting the reactor and the NMR tube. A sample time scale magnetic field profile (assuming a gas flow rate of 400 mL/min) as well as the distance scale profile was presented in Figure 2 1 1 and Table 2 1 Dramatic magnetic field ch ange was observed only in the last part of the flow path. Figure 2 1 1 Magnetic field profile for ALTADENA reactor based on Figure 2 6. A gas flow rate of 400 mL/min was assumed for. Magnetic field was measured with a Lakeshore Gauss meter. Red: time scale. Blue: distance scale.

PAGE 50

50 2 6 3 Magnetization Buildup: Effect o f Gas Flow Rate o n Signal Intensity As the magnetization of propylene and propane w as essentially zero outside t he magnet, it takes time to get polarized during the process of fl owing from the gas handling system to the detection region. Typically, to reach a maximum magnetization or thermal equilibrium at the detection field, it takes 3 5 times of the spin lattice relaxation time constant T 1 It is worth mentioning that, as relax ation time is magnetic field dependent, the effective relaxation time constant T eff here does not necessarily agree with the T 1 value of static gas measured by conventional inversion recovery or saturation recovery methods. As the flow rate increases, each individual gas molecule has less time to build up the magnetization and this will significantly reduce the NMR signal. To evaluate the relaxation process, or to what extent the flow rate affects the signal intensity, a series of experiments were repeated with pure propylene and propane respectively (Figure 2 1 2 ). Signals were normalized to the static condition. Figure 2 1 2 Flow rate dependence of magnetization buildup for propylene and propane (based on the integral values of propylene CH 3 and propane CH 2 groups). Integral values were normalized to that of the static spectrum of propylene and propane respectively.

PAGE 51

51 Results indicated that flow rate d id affect the signal substantially Signal reduce d exponentially with increasing flow rate an d at very high flow rate of 700 mL/min both gases only obtained around 10% magnetization. Figure 2 1 3 Estimate the effective relaxation constants (T eff ) for propylene and propane. A) Propylene CH 3 : 2.24 s. B) Propane CH 2 : 1.94 s. The schematics of the reactor setup used for reactions outside the magnetic field, or the ALTADENA condition, is shown in Figure 2 6. The distance and volume gas molecules travel from zero field t o detection field (9.4 Tesla) was calculated as Table 2 1 Based on this information, relaxation process is presented in the time scale (Figure 2 1 3 & Table 2 2 ). This is very similar to a saturation recovery relaxation time constant ( T 1 ) measurement an d observed magnetization follows the following equations de pending on the flow rate (2 3) (2 4)

PAGE 52

52 (2 5) Here is the observed thermally polarized magnetization, is the relaxation correction factor for thermally polarized signal, t tr is the transport time for molecules to travel from zero field to detection field, T eff is the effective relaxation time constant V is the traveling volume, and f is the gas flow rate F rom the fitting equation, effective relaxation time constants T eff were deduced (Table 2 3 ). T he CH 3 group of propylene and CH 2 group of propane have T eff values of 2.24 s and 1. 94 s respectively (Figure 2 1 3 ) This flow rate dependence affects the pai rwise selectivity calculation. Slower flow rate is better for thermally polarized signals, as they have more time to build up the magnetization. On the other hand, as relaxation r educe s the hyperpolarized signal, higher flow rate is better for reactions wi th p H 2 All kinetic d ata presented in this dissertation were corrected for relaxation effect based on information obtained here This relaxation effect was reported in the literature and corrected by a charcoal insert before the catalyst in PASADENA conditions. 86 But the method presented here is not feasible in ALTADENA conditions as charcoal insert will also relax the hyperpolarized signal O ur method is applicable to both PASADENA and ALTADENA conditions. 2 6 4 Para H 2 Enrichment: Effect o f Gas Flow Rate As in Figure 2 1 4 the relaxation time of H 2 is so short that gas flow rate d id n o t h ave a significant effect on the thermal intensity up to 1 L/min. But ortho to para H 2 conversion wa s affected with flow rates high er than 500 mL/min For low er flow rates, this difference in enrichment wa s negligible.

PAGE 53

53 Figure 2 1 4 Flow rate dependence of hydrogen signal and ortho to para H 2 conversion. A) Flow rate dependent peak integral of H 2 signal for n H 2 (blue triangle) and p H 2 (red square) B) Flow rate dependent enrichment (percentage of p H 2 ). 2 7 Kinetic D ata I nterpretation of P ropylene H ydrogenation 2 7 .1 Data Treatment f or Reactions w ith 50% Enriched p H 2 In this dissertation, 50% enriched p H 2 was used and spectra obtained for hydrogenation reactions were subtracted with those spectra obtained using n H 2 and treated as the pure p H 2 data. 2 7 2 Propylene t o Propane Conversion For reactions with n H 2 the conversion value is defined as the fr action of reactant propylene converted into product propane, and can be estimated by comparing the concentration of propane and propylene (by integrating the CH 3 group in propylene and CH 2 group in propane ) in the product gas stream. 75 As discussed in section 2.6.2, thermally polarized signals were normally underestimated at high flow rate and proper correction should be applied. According to

PAGE 54

54 E quation (2 3), the relation between observed signal and thermal equilibrium signal for prop ylene and propane are as follows: ( 2 6 ) ( 2 7 ) Here is the thermal equilibrium signal of propane, is the observed signal of propane, is the relaxation correction factor for propane, is the thermal equilibrium signal of propylene, is the observed signal of propylene, and is the relaxation correction factor for propylene. CH 3 group of the starting mater ial propylene and CH 2 group of the product propane were used for calculating conversion as follows: ( 2 8 ) Here is the conversion, is the integral value of propane CH 2 group, and is the integral value of propylene CH 3 group. 2 7 3 S ignal Enhancement Factor a nd Pairwise Selectivity 2 7 3 .1 Signal Enhancement Factor The experimental enhancement factor ( ) was calculated as the ratio of the peak integral of CH 2 group of hyperpolarized propane over that of thermally polarized propane

PAGE 55

55 ( 2 9 ) Here is the integral value of CH 2 group of propane obtained in reactions using p H 2 is the integral value of CH 2 group of propane obtained in reactions using n H 2 As discussed in section 2.6.2, relaxation effect plays a significant role at high gas flow rate. Proper treatment of the observed signal is necessary to minimize errors. In contrast to the fact that thermally polarized magnetization takes time to build up when molecules flow from zero field to high field (equation 2 3), h yperpolarized magnetization relaxes exponentially after being produced, according to the following equation s : (2 10 ) (2 11 ) (2 12) Here is the observed hyperpolarized magnetization, is the initial hyperpolariz ed magnetization, is the relaxation correction factor for hyperpolarized signal, t tr is the transport time for hyperpolarized signal to reach the detection field, and T eff is the effective relaxation time constant. Thus, the experimental enhancement factor after relaxation correction could be calculated as follows :

PAGE 56

56 (2 1 3 ) Let: (2 1 4 ) So: (2 1 5 ) Here is the correction coefficient for enhancement factor. In this dissertation, propane CH 2 group was used for enhancement factor calculation s (Table 2 2). The theoretical enhancement factor was calculated using the following adapted equation for the ALTADENA prot oc ol 14 75 as ALTADENA protocol has 2 times the theoretical enhancem ent factor of PASADENA protoc o l: ( 2 1 6 ) Here is the theoretical enhancement factor, k is the Boltzmann constant T is the detection temperature (297.13 K), is the fraction of p H 2 in the mixture of hydrogen spin isomers (50%), and is the number of equivalent protons ( 2 for the CH 2 group of propane )

PAGE 57

57 The maximum signal enhancement that can be produced in a PASADENA 240f, where f is the uncompensated fraction of p H 2 in the hydrogen mixture used (f=1/3 for 1:1 o:p ratio). 14 In our experiment s B 0 is 9.4 Tesla, is 0.5, is 2. Taking into account that only one H at om of the two in the CH 2 group is derived from p H 2 and that experiments were performed at 9.4 T (9.4/4.7=2), the maximum possible ratio of the integrals reduces to approximately 2*31240/(3*1*2*2)*(298.15/300)=2586. 2 7 3 2 Pairwise Selectivity Pairwise hydrogen addition represents the minor surface process compared to the so called random addition where one substrate harvests two hydrogen atoms originating from different hydrogen molecules. Pairwise selectivity is calculated by dividing the experimen tal signal enhancement over that of theoretical value. ( 2 1 7 ) In ALTADENA experiments, it takes time for the hyperpolarized product to diffuse out of the catalyst and be transported into the detection region. S pin lattice relaxation starts once the product is formed, and is dramatically accelerated by collisions with the porous surfaces. This significantly reduces e xperimental enhancement factor and data obtained here represent the lower limit

PAGE 58

58 Table 2 1. Magnetic field profile f or the ALTADENA reactor based on Figure 2 6. A gas flow rate of 400 mL/min was assumed for time scale calculation. Magnetic field was measured with a Lakeshore Gauss meter. Distance Volume Time Magnetic field (cm) (m) (m L ) (s) (mT) (T) 0 0.00 0.00 0.00 4.0 0.0040 5 0.05 0.27 0.04 3.0 0.0030 10 0.10 0.54 0.08 2.5 0.0025 15 0.15 0.80 0.12 2.0 0.0020 20 0.20 1.07 0.16 1.5 0.0015 40 0.40 1.47 0.22 3.0 0.0030 45 0.45 1.57 0.23 3.5 0.0035 50 0.50 1.67 0.25 4.5 0.0045 55 0.55 1.76 0.26 5.7 0.0057 60 0.60 1.86 0.28 7.7 0.0077 65 0.65 1.96 0.29 10.3 0.0103 70 0.70 2.06 0.31 13.6 0.0136 75 0.75 2.16 0.32 19.4 0.0194 80 0.80 2.26 0.34 28.0 0.0280 85 0.85 2.36 0.35 41.4 0.0414 90 0.90 2.46 0.37 63.2 0.0632 95 0.95 2.56 0.38 98.4 0.0984 100 1.00 2.65 0.40 173.3 0.1733 105 1.05 2.75 0.41 308.0 0.3080 110 1.10 2.85 0.43 596.0 0.5960 115 1.15 2.95 0.44 1240.0 1.24 120 1.20 3.05 0.46 2820.0 2.82 125 1.25 3.15 0.47 5270.0 5.27 130 1.30 3.25 0.49 7670.0 7.67 135 1.35 3.35 0.50 8610.0 8.61 140 1.40 3.45 0.52 8780.0 8.78

PAGE 59

59 Table 2 2 Normalized integral values of propylene and propane at various flow rates: the relaxation effect Flow rate Time Normalized integral (m L /min) (s) Propylene Propane CH CH 2 CH 3 CH 2 CH 3 0 NA 1.00 1.00 1.00 1.00 1.00 50 4.14 0.67 0.65 0.78 0.83 0.75 100 2.07 0.49 0.47 0.59 0.67 0.59 150 1.38 0.40 0.39 0.50 0.53 0.51 200 1.04 0.30 0.28 0.39 0.46 0.38 250 0.83 0.23 0.22 0.32 0.34 0.31 300 0.69 0.19 0.18 0.26 0.30 0.25 350 0.59 0.16 0.16 0.23 0.25 0.22 400 0.52 0.15 0.14 0.21 0.22 0.18 450 0.46 0.13 0.13 0.19 0.20 0.17 500 0.41 0.12 0.11 0.17 0.19 0.15 550 0.38 0.10 0.10 0.15 0.17 0.15 600 0.35 0.09 0.09 0.13 0.17 0.15 650 0.32 0.09 0.08 0.13 0.16 0.12 700 0.30 0.08 0.08 0.13 0.13 0.12

PAGE 60

60 Table 2 3 T ransport time (t tr ), effective relaxation time constants (T eff ) relaxation correction factors (C TP for thermally polarized signal and C HP for hyperpolarized signal) and enhancement correction factor C EF for propylene and propane In this dissertatio n, C EF of propane CH 2 is used for enhancement factor calculation s Flow rate Transport Propylene Propane m L /min t tr (s) CH CH 2 CH 3 CH 2 CH 3 T eff (s) NA NA 3.22 3.39 2.24 1.94 2.34 C TP 300 0.69 0.19 0.18 0.27 0.30 0.26 400 0.52 0.15 0.14 0.21 0.23 0.20 500 0.41 0.12 0.11 0.17 0.19 0.16 600 0.35 0.10 0.10 0.14 0.16 0.14 700 0.30 0.09 0.08 0.12 0.14 0.12 C HP 300 0.69 0.81 0.82 0.73 0.70 0.74 400 0.52 0.85 0.86 0.79 0.77 0.80 500 0.41 0.88 0.89 0.83 0.81 0.84 600 0.35 0.90 0.90 0.86 0.84 0.86 700 0.30 0.91 0.92 0.88 0.86 0.88 C EF 300 0.69 4.18 4.43 2.77 2.34 2.92 400 0.52 5.74 6.06 3.85 3.27 4.04 500 0.41 7.29 7.70 4.93 4.20 5.17 600 0.35 8.84 9.33 6.01 5.14 6.29 700 0.30 10.40 10.97 7.09 6.07 7.42

PAGE 61

61 CHAPTER 3 OXIDE SUPPORTED PLATINUM CATALYSTS: KINETIC STUDIES OF PROPYNE HYDROGENATION 3 .1 Background S elective partial reduction of hydrocarbons with multiple unsaturation, i.e. alkyne and dienes is of particular petrochemical industrial interests 87 88 as t he presence of trace alkyne may poison the catalyst for alkene polymerization or selective oxidation processes. 89 Catalyst and reaction condition o ptimizations are promising ways, and insights towards the surface kinetics and mechanisms will be helpful. 90 The differences in the surface affinity and adsorption/desorption thermodynamics of alkyne and alkene make it possible for the partial reduction product a lkene being replaced by incoming alkyne in the reaction stream. Alkyne has a higher adsorption enthalpy than alkene, and dominates surface coverage unless at a significant ly low concentration. 90 TPD suggested a value of 160 kJ mol 1 for propyne adsorption on Pt(111), 91 and a value of 91 kJ mol 1 for propylene. 92 On the other hand, the stereoselectivity of alkyne reduction provides insights into the reaction mechanisms. In c requirement of pairwise pathway provide s a judgment of the syn or anti addition. 44 74 76 93 In propyne hydrogenation the initial addition of two hydrogen atoms lead s to the partial reduction product propylene, and further reduction lead s to propane. Especially in the partial reduction product propylene, the two hydrogen atoms at the end of the double bond have distinct chemical shifts. In this case, PHIP can provide direct information if only one of them or both come from p H 2 Similar to the case using 94 a stereoselective syn addition of H 2 was observed using 95 and reduction with

PAGE 62

62 immobilized Au(III) complex led to the hyperpolarization of the atom only trans to the methyl group as well 41 In contrast when the reaction was carried with supported Pt as the catalyst, both hydrogen atoms exhibited hyperpolarization, which was attributed to the isomerization process on surfaces. 96 Interestingly, in the gas phase reaction using 95 three distinct stages were observed Initially at 70 o C, a syn addition was observed for a short period of time. The catalyst then became inactive and no product was detected. An elevation of temperature to 110 o C re activated the system at the cost of losing stereoselectivity. Both trans and cis prop yl ene configurations were hyperpolarized, probably due to the partial reduction of the metal complex at high temperature. Propyne reduction over supported palladium catalysts were proposed to have two kinds of active sites. 97 Type I represents the sites where full reduction to propane happens and type II is the selective reduction part. Type II is formed fr om the carbonaceous deposits at the early stage of the catalytic reaction. 98 100 The hydrogen dissolved in the lattice is reactive but unselective upon reaching the surface, and carbonaceous deposits serve as the obstacle, thus increasing the s electivity. 99 Palladium is so far the b est catalyst for selective alkyne hydrogenations. 90 Pd nanoparticles in a supported ionic liquid phase Pd 0 IL1/C and Pd 0 IL2/C were reported to be se lective in acetylene 101 an d propyne 44 partial reduction. This high selectivity towards partial reduction was als o demonstrated in PHIP studies using two different ionic liquids supported on activated carbon fiber (ACF) materials and containing Pd nanoparticles. 44 High selectivity was also reported using supported gold catalysts for the reduction of butadiene, butyne and acetylene. 102 105 In particular, a 100% selectivity in

PAGE 63

63 the temperature rang e between 313 and 523 K was reported for the acetylene to ethylene hydrogenation over Au/Al 2 O 3 105 In this chapter, propyne reduction (Scheme 3 1) was studied using Pt/TiO 2 as the catalyst. Conditions were varied to favor either partial or full reduction product. The dependence of selectivity on temperature and hydrogen partial pressure were explored In another case, introducing D 2 into the reaction system would suppress the production of hyperpolarized propylene as D 2 compete s with p H 2 Further reaction with p H 2 still lead s to hyperpolariz ation. Detailed studies about temperature dependence and reaction orders will be presented Scheme 3 1. Propyne hydrogenation over Pt/TiO 2 3 .2 Experimental Methods The catalyst s (Table 3 1) used in this chapter w ere prepared and characterized by Dr. Wei Cheng according the method described in C hapter 2 Catalysts were characterized mainly by chemisorption measurements. PHIP experiments were conducted according to the method described in C hapter 2 T he total gas flow rate wa s 400 mL/min and for the temperature dependent studies, the gas composition wa s N 2 /H 2 /Propyne = 140/240/20 mL/min (35/60/5%)

PAGE 64

64 As discussed in C hapter 2 all data need to be corrected for relaxation effect Based on E quation (2 11), the initial hyperpolarized signal can be calculated. As different protons have different effective relaxation time constants (Table 2 2) data should be corrected accordingly. (3 1) (3 2) (3 3) (3 4) (3 5) As the total flow rate wa s 400 mL/min for all experiments done in this chapter, wa s 0. 52 s (Table 2 2) Thus ratios between different protons can be calculated accordingly. (3 6) (3 7)

PAGE 65

65 (3 8) (3 9) Here is the ratio between CH 3 and CH 2 in hyperpolarized propane, is the ratio between CH 3 and CH in hyperpolarized propylene, is the ratio between CH 2 and CH in hyperpolarized propylene, is the ratio between hyperpolarized propa ne and hyperpolarized propylene, and calculated b y comparing propane CH 2 and propylene CH groups. 3 .3 Results and Discussion Results indicated that p ropyne wa s less reactive than propylene and almost no product was hardly detectable in the reaction with n H 2 similar to the case using immobilized Iridium complexes as the catalyst 89 But thanks to the dramatic signal enhancement, pairwise contribution wa s easily recognizable. 3 .3.1 Temperature Dependent Studies o f Propyne Partial Reduction Propyne wa s partially reduced to propylene and fully reduced to propane. The ratio between propylene and propyne depend s on the re action temperature. From Figure 3 1 (reactions with CAT009 0.5 wt% Pt/TiO 2 ) and Figure 3 3 (reactions with CAT023a 1.0 wt% Pt/TiO 2 ) propylene wa s more favored than propane at high tempera ture and propane production decreas e d as temperature went up. This might be explain ed by the surface desorption process of propylene. At high temperature,

PAGE 66

66 propylene h a d high kinetic energy and wa s more prone to desorb ing from the surface. Decreased surfac e propylene lead to lower propane production. Figure 3 1. Spectra for t emperature dependent propyne reduction under ALTADENA condition using CAT009 ( 0.5 wt% Pt/TiO 2 10 mg). A ) Reactions with n H 2 B) Reactions with 50% p H 2 after subtraction with n H 2 spectra. Gas composition: N 2 /H 2 /propyne = 140/240/20 mL/min (35/60/5%). Figure 3 2B showed that the activation energy for propyne to propylene reduction is 20. 9 2.6 kJ/mol. The ratio between propylene and propane in the product stream represents the sel ectivity towards partial reduction. Clearly for both catalysts, propane production decreased with increasing temperature. For CAT009 (Figure 3 2D), at 150 o C, propane/propylene was around 1. 3 while at 350 o C, it dropped to around 0.3. The decreasing showe d a linear behavior. Similar trend was observed with CAT023a (Figure 3 4D), but the ratio was around 5 .5 at 150 o C, and around 1 at 350 o C. Thus, CAT023a seemed to be less favorable for partial selectivity The change of full to partial reduction at increased temperature could be attributed to the decrease of surface residence time of propylene with increasing temperature and the stronger binding energy of propyne. Thus as CAT023a seemed to produce more propane than CAT009, the surface o f CAT023a wa s more favorable for

PAGE 67

67 propylene adsorption or retention due to diffusion limit or more active metal sites. This difference m ight also be due to the more abundance of sites favoring pairwise selectivity on the surface of CAT023a. Figure 3 2 Temperature dependent propyne reduction under ALTADENA condition using CAT009 ( 0.5 wt% Pt/TiO 2 10 mg ) Gas composition: N 2 /H 2 /propyne = 140/240/20 mL/min (35/60/5%). A ) P eak integral value versus temperature ( o C ). B ) Arrhenius plot for propyne to propylen e reduction C ) Peak integral ratio between CH 3 /CH groups in propylene and CH 3 /CH 2 in propane D ) Peak integral ratio between propane CH 2 and propylene CH

PAGE 68

68 Figure 3 3 Spectra for t emperature dependent propyne partial reduction under ALTADENA condition using CAT0 23a ( 1.0 wt% Pt/TiO 2 5 0 mg) Gas composition: N 2 /H 2 /propyne = 140/240/20 mL/min (35/60/5%). A ) Reactions with n H 2 B) Reactions with p H 2 Thus it is possible to optimize the selectivity towards partial reduction by changing the temperature or optimizing the catalyst. 3 .3. 1. 1 Hydrogen Partial Pressure Dependent Reduction For CAT023a, three different hydrogen partial pressures were used to study the influence of H 2 concentration on the reduction selectiv ity ( Figure 3 4 ). In all three cases, propane production decreased and propylene production increased with increasing temperature. Set 2 (H 2 /propyne = 12/1) produced most propane (Figure 3 4 A) but set 1 (H 2 /propyne = 6/1) had the most propylene (Figure 3 4 B). While it make sense to favor partial reduction with least H 2 the optimum condition for propane production may be due to the blockage of active sites by hydrogen when hydrogen is in large excess and predominant on the surface ( H 2 /propyne = 18/1 ) decreasin g the interaction s of propyne and propylene with surfaces. This was confirmed by the reduced production of propylene as well.

PAGE 69

69 Activation energies were calculated based on the Arrhenius plot of temperature dependent studies (Figure 3 4 ) for three increasing hydrogen partial pressures: 14.4 1.7 18. 8 2.7 and 15. 2 2. 3 kJ/mol respectively. Figure 3 4 Temperature dependent propyne reduction under ALTADENA condition using CAT023a (50 mg). A ) P ropane CH 2 peak integral value versus temperature ( o C) B ) Arrhenius plot for propylene CH peak integral C ) Peak integral ratio between groups in propane ( C 1) and propylene ( C 2) D ) Peak integral ratio between propane and propylene. Red square: Set 1 (H 2 /Propyne = 6/1) Blue triangle: Set 2 (H 2 /Propyne = 12/1) Cyan diamond: Set 3 (H 2 /Propyne = 18/1) The ratios between the CH 3 and CH 2 groups of propane were around 1, as expected, but the ratios of propylene CH 3 to CH were between 0.5 and 0.6 for all three cases. Although it was not expected, as no hydrogen in C H 3 group came from p H 2 this

PAGE 70

70 polarization gain might be attributed to the polarization transfer from other coupled spins. This is be further explained in section 3.3.5 As shown in Figure 3 4 D, the more hydrogen, the more favorable for full reduction. And all decreased with increasing temperature, consistent with previous observations. 3 .3. 1. 2 Total Flow Rate Dependent Reduction : CAT023a For CAT023a two different flow rates ( 400 mL/min and 600 mL/min ), with the same gas composition ( H 2 /Propyne = 12/1), wer e employed to investigate the effect of gas flow rate on catalytic performance Figure 3 5 Spectra for t emperature and total flow rate dependent propyne reduction under ALTADENA condition using CAT023a (50 mg). Gas composition: N 2 /H 2 /propyne = 140/240/20 mL/min (35/60/5%). A ) T otal flow rate 400 mL/min. B) T otal flow rate 600 mL/min The temperature dependence was more or less consistent for the two different flow rates (Figure 3 5 & 3 6). Figure 3 6B demonstrated that at higher flow rate react ion wa s more sensitive to temperature, and ha d higher apparent activation energy for propylene production. Figure 3 6 D Shows at 100 o C, propane production wa s small compared to propylene, but 150 o C gave the best propane/propylene ratio, which decrease d wi th increasing temperature.

PAGE 71

71 Figure 3 6 Temperature and total flow rate dependent propyne reduction using CAT023a (50 mg). A ) propane CH 2 peak integral value versus temperature ( o C) B ) Arrhenius plot for propylene CH peak integral. C ) Peak integral ratio between groups in propane ( C 1) and propylene ( C 2) D ) Peak integral ratio between propane and propylene. Red square: total flow rate 400 mL/min. Blue triangle: total flow rate 600 mL/min Gas composition: N 2 /H 2 /propyne = 35/60/5% 3 .3.2 Reaction Or ders w ith Respect t o p H 2 3 .3. 2. 1 CAT023a ( 1.0 Wt% Pt /Tio 2 5 0 m g) The dependence of PHIP on hydrogen partial pressure was studied at two different temperatures (200 and 350 o C). The total flow rate was 400 mL/min, with a flow rate of 20 mL/min for propyne.

PAGE 72

72 Figure 3 7 dependent (reaction order) propyne reduction under ALTADENA condition using CAT0 23a ( 5 0 mg). A ) Natural log of propane CH 2 peak integral value versus natural log of hydrogen partial pressure B ) Natural log of propylene CH peak integral value versus natural log of hydrogen partial pressure C ) Peak integral ratio between groups in propane ( C 1) and propylene ( C 2). D ) Peak integral ratio between propane and propylene. T otal flow rate: 400 mL/min Red square: reactions at 350 o C. Blue triangle: reactions at 200 o C. The reaction orders with respect to H 2 (Figure 3 7 A) wa s around 1 for propane production at both temperatures (200 and 350 o C). B ut for propylene (Figure 3 7B the reaction orders were 0.7 2 0.02 (with respect to n H 2 ) and 0.5 4 0.03 (with respect to p H 2 ) at the beginning for 200 and 350 o C respectively then the signal started to decrease with increasing H 2 partial pressure, lead ing to more full reduction to propa ne. This indicated that full reduction was more sensitive to the hydrogen partial pressure than partial reduction. Again, t he ratio of CH 3 and CH 2 groups in propane wa s around 1

PAGE 73

73 for all cases and that of CH 3 and CH groups in propane wa s around 0.5. L ow tempera ture (200 o C) favor ed more full reduction propane production. 3 .3. 2. 2 Kinetic Isotope Effect : Reactions w ith D 2 Here D 2 was used as the third gas, replac ing N 2 (Figure 3 8 ). R eaction orders were 0. 60 0.07 (with respect to n H 2 ) and 0. 80 0.04 (with respect to p H 2 ) Figure 3 8 D 2 as the delivery gas: h ydrogen partial pressure and temperature dependent (reaction order) propyne reduction under ALTADENA condition using CAT0 23a ( 5 0 mg). A ) Natural log of propane CH 2 peak integral value ver sus natural log of hydrogen partial pressure B ) Natural log of propylene CH peak integral value versus natural log of hydrogen partial pressure C ) Peak integral ratio between groups in propane ( C 1) and propylene ( C 2). D ) Peak integral ratio between propane and propylene. T otal flow rate: 400 mL/min Red square: reactions at 350 o C. Blue triangle: reactions at 200 o C.

PAGE 74

74 Reaction s with D 2 would not lead to hyperpolarization, but further reduction with p H 2 w ould result in hy perpolarized propane. Results show ed that more hyperpolarized propane was observed than hyperpolarized propylene in this case. Catalytic performance s in reactions with D 2 or N 2 as the third gas w ere compared and are presented in Figure 3 9 In both cases, high temperature (350 o C) favor ed partial reduction to propylene, as the ratio between propane and propylene was less compared to reactions at 200 o C The use of D 2 should reduce the production of hyperpolarized propylene due to competitive re actions of D 2 over p H 2 But for hyperpolarized propane, it did not matter either D 2 or H 2 was added during the first step partial reduction, a consequent step with p H 2 would in any case lead to hyperpolarization. Figure 3 9 Comparison of conversion ratio and propane/propylene ratio at different temperatures (350 o C /200 o C ) in the propyne reduction under ALTADENA condition using CAT023a (50 mg). A) The ratio of propane over propylene at 350 o C /200 o C with N 2 as the third gas. B) The ratio of propane over propylene at 350 o C /200 o C with D 2 as the third gas. Total flow rate: 400 mL/min R ed squ a re: 350 o C B lue triangle: 200 o C.

PAGE 75

75 Figure 3 10 (H 2 /Propyne = 6/1), Figure 3 11 (N 2 /H 2 /Propyne = 12/1) and Figure 3 12 (N 2 /H 2 /Propyne = 18/1) showed the comparison of using N 2 and D 2 at different H 2 /propyne ratios. Figure 3 1 0 Comparison of ALTADENA spectra obtained with N 2 or D 2 at different temperatures in the propyne partial reduction under ALTADENA condition using CAT023a (50 mg) (H 2 /propyne = 6/1) A) N 2 /H 2 /Propyne = 260/120/20 mL/min. A 1 ) 350 o C. A 2 ) 200 o C. B) D 2 /H 2 /Propyne = 260/120/20 mL/min. B 1 ) 350 o C. B 2 ) 200 o C.

PAGE 76

76 Figure 3 1 1 Comparison of ALTADENA spectra obtained with N 2 or D 2 at different temperatures in the propyne partial reduction under ALTADENA condition using CAT023a (50 mg) (H 2 /propyne = 12 /1) A) N 2 /H 2 /Propyne = 140/240/20 mL/min. A 1 ) 350 o C. A 2 ) 200 o C. B) D 2 /H 2 /Propyne = 140/240/20 mL/min. B 1 ) 350 o C. B 2 ) 200 o C.

PAGE 77

77 Figure 3 1 2 Comparison of ALTADENA spectra obtained with N 2 or D 2 at different temperatures in the propyne partial reduction under ALTADENA condition using CAT023a (50 mg) (H 2 /propyne = 18 /1) A) N 2 /H 2 /Propyne = 20/360/20 mL/min. A 1 ) 350 o C. A 2 ) 200 o C. B) D 2 /H 2 /Propyne = 20/360/20 mL/min. B 1 ) 350 o C. B 2 ) 200 o C. For those reactions with N 2 reduction product propylene and low temperature favored full reduction product propane. For those with D 2 compared to N 2 cases, as D 2 competed with p H 2 leading to non hyperpolarization. Propane hyperpolarization showed some interesting results. At high temperature, D 2

PAGE 78

78 cases had stronger hyperpolarized propa ne signal than those obtained with N 2 While in low temperature case, the result was opposite. Those with D 2 had smaller signal than that of N 2 As hyperpolarized propylene production was significantly reduced but hyperpolarized propane product was not aff ected too much, kinetic isotope effect was not so significant for the first step reduction than that of second step. D 2 competed with H 2 for first step reduction, leading to reduced hyperpolarization for propylene. But H 2 was much more favorable than D 2 fo r the second step reduction, as results for N 2 or D 2 were comparable, just as if no D 2 was involved much. Future experiments may be conducted to compare the results of u sing D 2 /H 2 /Propyne and n H 2 /H 2 /Propyne. Reduction with either D 2 or n H 2 will not lead to hyperpolarization, but further reduction with p H 2 will result in hyperpolarized propane. By comparing the conversion of propyne and resulting hyperpolarized propylene and propane, a better understanding of the kinetic isotope effect may be achieved On the other hand, introduction of D 2 to propylene reduction should also give a direct insight about the kinetic isotope effect. 3 .3.3 Conversion a nd t he Amount o f Catalyst Comparing the reaction orders for CAT009 (10 mg) and CAT023a (50 mg) at 350 o C, CAT023 gave more propane as the product while CAT009 lead to more propylene. It seemed that the mo re catalyst was used the more full reduction was obtained But as the reactivity wa s different for different catalysts, f urther experiments should be done using the same catalyst (such as CAT023a ) but d ifferent quantities 3 .3.4 Stereoselectivity a nd Surface Isomerization Figure 3 1 3 is the 1 H NMR spectrum of static propylene at 1 atm.

PAGE 79

79 In the PHIP studies of propyne reduction on Pt/TiO 2 the two protons in propylene CH 2 group behaved differently. The proton cis to CH group (H3) was always hyperpolarized together with the CH group, in agreement with syn addition of p H 2 While the proton trans to CH group (H2) showed some extent of hyperpolarizat ion, depending on the reaction conditions such as temperature and p H 2 partial pressure. This wa s interesting as it was reported that both protons showed hyperpolarization over supported Pt, which was attributed to the catalytic isomerization of prop yl ene on surfaces and reduction of metal surfaces 96 Figure 3 1 3 1 H NMR spectrum of thermally polarized static propylene at 1 atm. A) Full spectrum. B) CH 2 group. Chemical shift (ppm): 5.85 (m, 1H, CH) 5.03 (d, 1H, trans to CH), 4.91 (d, cis to CH), 1.72 CH 3 (m, 3H, CH 3 ). The spectra (Figure 3 14) resulting from syn or anti addition of p H 2 were obtained using six spin density matrix simulations using the SpinDynamica package developed by Prof. Malcolm Levitt, University of Southampton, UK. Results indicated that polarization is shared between the two protons of the propylene CH 2 group. In Fig. 3 14A, it is apparent that the trans proton is hyperpolarized following syn addition due to

PAGE 80

80 coherence t ransfer through the J coupling network at low field under ALTADENA reaction conditions. Similarly, the Fig. 3 14 B shows that the cis proton is hyperpolarized following anti addition. Figure 3 1 4 Six spin system d ensity matrix simulation of syn and an ti addition of p H 2 A) Syn addition: p H 2 added to H1 & H3 positions in Figure 3 13. B) Anti addition: p H 2 added to H1 & H2 positions in Figure 3 13. C) Enlargement of syn addition on CH 2 D) Enlargement of anti addition on CH 2 Simulation was carried o ut in SpinDynamica package developed by Prof. Malcolm Levitt, University of Southampton, UK.

PAGE 81

81 In a surface hydrogenation reaction, both syn and anti additions may occur Spectra resulting from a linear combination of these two processes were simulated and representative spectra with certain percentage of anti addition are shown in Figure 3 15). Figure 3 1 5 Simulated spectra resulting from a linear combination of sy n and anti addition. A) 100% Syn. B) 20% Anti. C) 40% Anti. D) 60% Anti. E) 80% Anti. F) 100% Anti. Pure s yn and anti spectra l simulations are shown in Fig. 3 14.

PAGE 82

82 A calibration curve correlating the percentage of trans polarization and the percentage o f anti addition is constructed from the simulated spectra (Figure 3 16). Trans polarization is proportional to the integral ( ) of the trans proton peak (proton number 2 in Figure 3 13) Cis polarization is proportional to the peak integral of proton number 3 ( ) in Figure 3 13. Thus, (3 10) With this calibration curve, the percentage of anti addition can be deduced (Table 3 2, Table 3 3) fro m the measured percentage of trans polarization of the experimental ALTADENA spectrum Figure 3 1 6 The percentage of trans polarization over the sum of trans and cis polarization versus the percentage of anti addition over the sum of anti and syn addition. Figure 3 1 7 indicates that at l ow temperature (200 o C) both protons showed similar hyperpolarization, meaning more surface isomerization. At high temperature

PAGE 83

83 ( 35 0 o C), the apparent polarization of the trans H was significantly less than that of cis H Furthermore, fraction of syn addition increased with increasing hydrogen pressure (from top to bottom in Figure 3 1 7 ) Figure 3 1 7 Comparison of propylene CH 2 hyperpolarization obtained with different H 2 partial pressure at different temperatures in the propyne reduction under ALTADENA condition using CAT023a (50 mg). N 2 /H 2 /Propyne : A) 260 / 12 0/20 mL/min ; 350 o C. B) 260 / 12 0/20 mL/min ; 200 o C. C) 140 / 24 0/20 mL/min ; 350 o C. D) 140 / 24 0/20 mL/min ; 200 o C. E) 20 / 36 0/20 mL/min ; 350 o C. F ) 20 / 360/20 mL/min ; 200 o C.

PAGE 84

84 Figure 3 1 8 Variation of percentage of trans polarization and percentage of anti addition with hydrogen partial pressure and temperature. Propyne partial pressure was constant as 0.05 atm. Total gas flow rate: 400 mL/min. A) Trans polarization percentage (Eq uation 3 10 ) B) Anti addition percentage (from calibration curve, Fig. 3 16) Red: 350 o C. Blue: 200 o C. The extent of anti addition or surface isomerization is presented in Table 3 2 and Figure 3 1 8 Lower temperature is mor e favorable for anti addition and thus surface isomerization. At high temperature, the residence time of a substrate molecule on the metal surface decrease s and the probability of surface isomerization is reduced. In addition, increased hydrogen partial pressure also increased the extent of surface isomerization. Interestingly, the results show that anti addition constituted around 60 % of the whole product. At present, there is no clear explanation for this result. The stereoselectivity studies were ext ended to several other temperatures and flow rates. Figure 3 19 and Figure 3 20 compare the isomerization at various temperature with a total gas flow rate of 400 mL/min and 600 mL/min respectively.

PAGE 85

85 Figure 3 1 9 Comparison of propylene CH 2 hyperpolarization obtained at different temperatures in the propyne reduction under ALTADENA condition using CAT0 23a ( 1.0 wt% Pt/TiO 2 5 0 mg) Gas flow rate (total 400 mL/min): N 2 /H 2 /propyne = 140/240/20 mL/min (35/60/5%). A ) 200 o C. B) 250 o C. C) 300 o C. D) 350 o C. E) 375 o C.

PAGE 86

86 Figure 3 20 Comparison of propylene CH 2 hyperpolarization obtained at different temperatures in the propyne reduction under ALTADENA condition using CAT0 23a ( 1.0 wt% Pt/TiO 2 5 0 mg) Gas flow rate (total 600 mL/min): N 2 /H 2 /propyne = 210/360/30 mL/min (35/60/5%). A ) 200 o C. B) 250 o C. C) 300 o C. D) 350 o C. E) 375 o C.

PAGE 87

87 For both cases, at low temperature, syn and anti additions have comparable hyperpolarization, but as temperature increased, syn addition hyperpolarization b uilt up fast than anti product. The extent of anti addition or surface isomerization was presented here in Figure 3 21 Comparing data at 200 and 375 o C, almost only half extent of surface isomerization was observed at high temperature than low temperatur e in both flow rate cases. High flow rate was less affected as shown below. At low temperature, f low rate did not affect the extent of isomerization. Figure 3 21 Comparison of percentage of trans polarization and anti addition at various total gas flow rate and reaction temperature Gas ratio: N 2 /H 2 /propyne = 35/60/5%. A) Trans polarization percentage. B) Anti addition percentage. Total flow rate: 400 mL/min (blue); 600 mL/min (red). 3 .3. 5 Polarization Transfer Interestingly f or the partial reduct ion o f propyne to propylene, the CH 3 group of propylene (chemical shift at around 1.72 ppm) wa s also hyperpolarized to a great extent (Figure 3 22 )

PAGE 88

88 Figure 3 22 Comparison of propylene CH & CH 3 hyperpolarization obtained with different H 2 partial pressure at different temperatures in the propyne reduction under ALTADENA condition using CAT023a (50 mg). A) N 2 /H 2 /Propyne = 260 / 12 0/20 mL/min, 350 o C. B) N 2 /H 2 /Propyne = 260 / 12 0/20 mL/min, 200 o C. C) N 2 /H 2 /Propyne = 140 / 24 0/20 mL/min, 350 o C. D ) N 2 /H 2 /Propyne = 140 / 24 0/20 mL/min, 200 o C. E) N 2 /H 2 /Propyne = 20 / 36 0/20 mL/min, 350 o C. D) N 2 /H 2 /Propyne = 20 / 360/20 mL/min, 200 o C.

PAGE 89

89 This hyperpolarization was not normally expected. As shown S cheme 3 1, CH 3 group (H4) was not involved in the reaction and no hydrogen atom came from p H 2 The gain of its hyperpolarization wa s probably due to the spontaneous polari zation sharing and transfer among coupled spins via isotropic mixing at low magnetic field where the reaction was taking place. Reactions in PASADENA conditions did not report this polarization share to the neighboring group. The peak at around 6 ppm (left most) corresponds to CH group of propylene, whose H was originating from p H 2 Thus it is interesting to compare the extent of hyperpolarization between these two peaks. The ratio between propylene CH 3 and CH wa s about 0.5, while the ratio between CH 2 and CH 3 groups of the full reduction product propane wa s around 1 for both catalysts (Figure 3 1 & 3 3). T his polarization transfer extend ed the scope of potential applications especially if, in some cases, different protons could be involved in different reactions or interactions with surroundings. 3 .4 Conclusion In this chapter, the selectivit y of propyne partial reduction over Pt/TiO 2 was studied using parahydorgen induced polarization. It was found to be dependent on the catalyst quantity, reaction temperature and hydrogen partial pressure. High er temperature wa s more favorable for selectivit y towards partial reduction. The observation of polarization transfer to couple spins and the hyperpolarization of trans H surface isomerization, as well as the kinetic isotope effect were among the most interesting parts of this work. Especially surface isomerization, thus the polarization pattern, was found to be related with reaction temperature and hydrogen concentration as well. The introduction of D 2 into the system, replacing N 2 shed some lights on the detail of surface processes and the relative kinetic isotope effect. Further

PAGE 90

90 studies should be focused on catalyst optimizations towards better control of pairwise and stereochemistry selectivit y

PAGE 91

91 Table 3 1. Platinum particles supported on TiO 2 as catalysts. Courtesy of Dr. Wei Cheng Catalog Catalyst Metal Support Loading Calcination No. Name (w/w%) T ( ) CAT009 0.5PtTi450 Pt TiO 2 0.5 450 CAT023a 1.0PtTi350 Pt TiO 2 1 350 Table 3 2. Surface isomerization and percentage of trans (anti ) addition product with various total gas flow rate and temperature Propyne partial pressure: 0.05 atm. P H2 Reaction temperature (atm) 350 o C 200 o C Trans Cis Trans/Cis Trans % Anti % Trans Cis Trans/Cis Trans % Anti % 0.30 9.16 30.87 0.30 23 35 3.95 10.09 0.39 28 39 0.60 9.36 26.84 0.35 26 37 6.28 11.53 0.54 35 46 0.90 11.87 19.81 0.60 37 49 8.89 10.43 0.85 46 60 Table 3 3. Surface isomerization and percentage of trans (anti ) addition product with various hydrogen partial pressure and temperature Hydrogen partial pressure: 0.60 atm. Propyne partial pressure: 0.05 atm. T ( o C) Total gas flow rate 400 mL/min 600 mL/min Trans Cis Trans/Cis Trans % Anti % Trans Cis Trans/Cis Trans % Anti % 200 3.10 4.11 0.75 43 56 10.27 14.54 0.71 41 53 250 3.62 5.29 0.68 41 53 10.33 18.00 0.57 36 47 300 2.43 7.94 0.31 23 35 12.02 24.39 0.49 33 44 350 1.05 6.19 0.17 14 27 22.41 78.60 0.29 22 34 375 1.31 8.31 0.16 14 27 23.68 89.93 0.26 21 33

PAGE 92

92 CHAPTER 4 TITANIA SUPPORTED IRIDIUM CATALYST S : KINETIC STUDIES OF PROPYLENE HYDROGENATION 4 .1 Background Hydrogenation of olefinic materials on heterogeneous metallic catalysts is of co nsiderable industrial interest. Its kinetic behavior depends significantly on the properties of the catalyst itself, including the metal type, particle size, support and even preparation conditions. Propylene represents the simplest asymmetric olefin and its hydrogenation has been employed as a model reaction extensively in parahydrogen induced polarization (PHIP) based on heterogeneous catalysis. Hyperpolarization was observed in a variety of catalytic systems, including immobilized inorganic complexes and supported metals. Catalyst particle size, support and reaction conditions were varied to gain insights into the factors governing pairwise selectivity. Kinetic st udies employing the high signal enhancement of PHIP was reported recently with Pt/Al 2 O 3 in both PASADENA and ALTADENA experiments. 42 96 Ir complexes have been involved in homogeneous hydrogenation reactions for long time 106 108 109 It has been involved in the development of PHIP in the form of [(COD)Ir( PCy 3 )(Py)] + PF 6 (COD = 1,4 cyclooctadiene, PCy 3 = tricyclohexylphosphine, Py = pyridine), and catalyst Ir(CO)Cl(PPh 3 ) 2 Here we prepared Iridium nanoparticles supported on titanium oxide as the heterogeneous catalyst. The activation energy for propylene hydrogenation over pumice supported Ir was reported to be 62.7 k J /mol. 110 Employing the hydrogenation

PAGE 93

93 of propylene as the model reaction, the dependence of PHIP performance on the reaction temperature and relative ratio of reactants were studied in details. The apparent activation energy for the overall reaction and the reaction orders with respect to H 2 and propylene were determined with either n H 2 or 50% enriched p H 2 This represents the first kinetic study of propylene hydrogenation o ver supported i ridium catalysts. 4 .2 Experimental Methods The catalyst used in this chapter was prepared and characterized by Dr. Wei C heng according to the procedure described in C hapter 2 PHIP experiments were conducted based on the procedure described in C hapter 2 For most of the experiments, a total flow rate of 300 mL/min was employed, except for the flow rate dependent studies. All NMR spectra were acquired with 128 scans and a 90 o pulse under steady state conditions. Kinetic data were corrected for relaxation effect based on E quations (2 3), (2 10) (2 13), and Table 2 2 4 .3 Results and Discussion ALTADENA experiments for the hydrogenation of propylene (Scheme 4 1) were carried out and hyperpolarization (net alignment) was achieved in most cases. Activation en ergies of the reaction were derived from temperature dependent studies, and reaction orders were also determined by fixing the total flow rate but changing only one component at a time. Two different surface pathways, namely random and pairwise additions, were distinguished by reacting with n H 2 or 50% enriched p H 2 respectively.

PAGE 94

94 Scheme 4 1. Hydrogenation of propylene over Ir/TiO 2 at elevated temperatures 4 .3. 1 CO Chemisorption Estimation of the p article size of the prepared Ir catalyst (Table 4 1) could be calculated based on CO adsorption measurements with certain assumptions 83 The metal loading is reported here as the weight percentage of the reagents, but t he actual metal loading could be determined by ICP AES 4 .3. 2 Catalyst Durability One of the criteria for being a good catalyst is the robustness. Catalytic durability test of the catalyst prepared for this work w as carried out under continuou s flow condition over 8 hours. Propylene and propane were monitored by NMR R esults (Figure 4 1 A&B ) showed that no significant change of the catalytic activity was observed over the whole experiments for both reactions at 50 or 150 o C. More propane was obtained at 150 o C than 50 o C, as the catalyst is more reactive at higher temperature. While at higher temperature (Figure 4 1C&D ), reactions were less stable and more fluctuations were observed. Reactivity decreased slightly over time for reactions at 350 o C, but still not too bad. At very high temperature (350 o C), reactivity decreased and surface decomposition to methane further reduced propane production.

PAGE 95

95 Figure 4 1 Catalyst durability test. A &C ) Remaining propylene in the product stream. B &D ) Produced propane in the product stream. Gas flow rate (total 300 mL/min ): N 2 / p H 2 /Propylene = 120/150/30 mL/min Red square: reactions at 150 o C B lue triangle: reactions at 50 o C. Green dot: reactions at 250 o C Black star: reactions at 350 o C 4 .3. 3 Kinetic Studies It was suggested that n H 2 and p H 2 have different kinetic behavior and in this part, the temperature dependence and reaction orders were studie d with either n H 2 or 50% enriched p H 2 under otherwise identical conditions.

PAGE 96

96 4 .3. 3 .1 The Arrhenius Equation a nd Activation Energy Figure 4 2 Representative (50 350 o C) spectra resulting from reactions with n H 2 and p H 2 for the t emperature dependent studies of propylene hydrogenation ( p H 2 spectra obtained by subtracting n H 2 spectra from 50% p H 2 ) under ALTADENA condition Gas flow rate (total 300 mL/min ): N 2 / p H 2 /Propylene = 120/150/30 mL/min. A ). n H 2 spectra. B) p H 2 spectra. Figure 4 3 3D spectra for temperature dependent catalytic performance under ALTADENA condition Gas flow r ate (total 300 mL/min ): N 2 /H 2 /Propylene = 120/150/30 mL/min A) Spectra obtained for reactions with n H 2 B) Spectra obtained for reactions with p H 2

PAGE 97

97 As described in Chapter 1, reactions normally depend on the temperature, thus it is important to quantify their relationship by running propylene hydrogenation reactions at a series of temperatures and the apparent activation energy associated with the complete catalytic cycle can be derived. Figure 4 4 Composition of the product stream at various temperatures. Red square: starting material propylene. Blue triangle: product propane. Green circle: by product. Figure 4 2 and Figure 4 3 (3D) show ed representative spectra obtained with either n or p H 2 The starting material propylene and the product propane were detected based on the 1 H NMR spectra. At high temperature (>250 o C), another unknown by product appeared and this was proposed as CH 4 based on the chemical shift value of 0.2 ppm. As shown in Figure 4 4 of the comparison of integral values, propane production started to decrease after 250 o C but propylene conversion seemed

PAGE 98

98 to be stabilized. This could be explained by the emergence of the by product which increased with increasing temperature and quickly dominated over the product formation. At high temperature, the diffusion and transport processes of adsorbed species on surfaces started to limit the reaction kinetics and a drop of reactivity was observed Figure 4 5 Temperature dependent studies of propylene hydrogenation under ALTADENA condition Gas flow rate (total 300 mL/min): N 2 / p H 2 /Propylene = 120/150/30 mL/min. Red square: reactions with p H 2 Blue triangle: reactions with n H 2 A) Catalytic performance in the temperature range of 50 350 o C. B) The Arrhenius plot over the temperature range of 50 250 o C. To take a closer look at the dependence of propane formation on temperature, two distinct curves were observed for reactions with n or p H 2 (Figure 4 5). While the plot on the left showed the complete data for the temperature range of 50 350 o C, the right figure was plotted based on the Arrhenius equation for the low temperature region. A linear relationship was observed for temperatures between 50 and 250 o C. According to the Arrhenius equation, the slope of the curve is E a so activation energies can be calculated for pairwise (16.8 0. 5 kJ/mol) and random (9.8 0.8 kJ/mol) pathways.

PAGE 99

99 This result demonstrated that the two pathway ha ve different reaction barriers and pairwise addition was more energy demanding This difference might be related with the location of the reaction sites on the surface. As was suggested in the literature, pairwise addition tends to occur at sites with low coordination numbers, such as (100) facet, edges and corners, where less interactions are possible between the adsorbed sp ecies and the surface. Figure 4 6 Propylene to propane conversion, enhancement factor and pairwise selectivity. Gas flow rate (total 300 mL/min): N 2 / p H 2 /Propylene = 120/150/30 mL/min. A) Propylene to propane conversion. B) Enhancement factor and pair wise selectivity. As shown in Figure 4 6, p ropane production increased with temperature up to 250 o C, and decreased at high temperature region. Although H 2 had 5 times higher partial pressure than propylene, the maximum conversion was still less than 40 %. The use of limited quantity of catalyst, the introduction of N 2 as the delivery gas and high flow rate may account for this behavior. Nevertheless, these parameters were optimized for the best control of the steady state reaction condition under all tempe ratures, as the reaction was rather exothermic and efficient heat dissipation was critical for maintaining

PAGE 100

100 steady state conditions. Pairwise selectivity proved to be a minor process, as it never reached 1 %, but increas ed with increasing temperature 4 .3. 3 .2 The Effect o f Flow Rate a nd Relaxation i n t he PHIP Studies Figure 4 7 Temperature dependent studies of propylene hydrogenation at a variety of total flow rates under ALTADENA condition Total gas flow rates (mL/min): red square 400, blue triangle 5 00, green dot 600, pink star 700. A) Reactions with n H 2 B) Reactions with p H 2 As i t t a k es time for the hyperpolarized product to reach the detect ion region, hyperpolarization can be partially lost during this process. On the other hand, magnetization buildup or relaxation, for thermally polarized molecules also affect s the NMR signal, leading to less intensity as the flow rate increase s These factors could be corrected based on the relaxation parameters obtained in Chapter 2. But flow rate might also affect the reactivity itself. Here 4 different total flow rates were studied : 400, 500, 600, 700 mL/min The ratio of gas composition thus the partial pressure of each component, wa s the same for all cases, N 2 /H 2 /propylene = 40/50/10 %

PAGE 101

101 Similar temperat ure dependence was obse rved for all flow rates (Figure 4 7), including the one mentioned in previous section (300 mL/min ). Propane production started to decrease at 250 o C, consistent with the production of by product. Figure 4 7 A showed that propane signal s obtained from reactions with n H 2 were similar for these four different flow rates, after relaxation correction based on the magnetization buildup as described in section 2 6 .2. This mean t that flow rate did not affect the catalytic reaction too mu ch within this range. Contrary to that, for hyperpolarized signals, higher flow rate meant less hyperpolarization lost, leading to the observation of stronger signal But data correction for the relaxation effect using the method developed in C hapter 2 pro ved to be effective again. A s shown on the ri ght of the figure, results were quite consistent for different flow rates at low temperature range. But at higher temperature, higher flow rate had higher hyperpolarized signal. This behavior discrepancy between thermally polarized signal and hyperpolarized signal might be attributed to the reduced contact time of hyperpolarized signal with the metal surfaces minimizing the relax ation effect from interactions with metal surfaces Another possible reason could be the increase of pairwise selectivity at higher flow rate. With speeded replacement by the feeding stock, surface bound propylene or reaction intermediates had less chance to receive two hydrogen atoms originating from different molecules. Plots based on the Arrhenius equation of these experiments provided the chance to calculate the activation energies (Figure 4 8). Again, reactions with p H 2 seemed to require more energy to overcome the barriers than their counterparts with n H 2 Another general trend wa s that activation energy increased with increasing total flow rate for both cases. A t high flow rate, r eagents had less contact time with surfaces, reducing the

PAGE 102

102 chance of bond weakening effects of the metal reagent interactions. As a consequence, higher re action barrier might be experienced to break bonds Figure 4 8 The Arrhenius plots of the temperature dependent (50 200 o C) studies of propylene hydrogenation at a variety of total flow rates under ALTADENA condition Total gas flow rates (mL/min): red square 400, blue triangle 500, green dot 600, pink start 700. A) Reactions with n H 2 B) Reactions with p H 2 Propylene to propane conversion, PHIP signal enhancement factor and pairwise selectivity were presented in F igure 4 9. L ower propylene to propane conversion at high flow rate might be due to the less contact time, thus less chance of reacting with H 2 of propylene on the surface. On the other hand, enhancement factor and pairwise selectivity increased with incre asing flow rate, partially due to the reduced relaxation effect caused by metal substrate interaction as mentioned above

PAGE 103

103 Figure 4 9 Temperature and total flow rate dependent propylene to propane conversion, signal enhancement factor and pairwise selectivity. Total gas flow rates (mL/min): red square 400, blue triangle 500, green dot 600, pink start 700. A) Propylene to propane conversion. B) Enhancement factor and pairwise selectivity. Figure 4 10 Total flow rate dependent activation energies for different reaction pathways. Red: pairwise pathway. Blue: random pathway. Error bar represents standard error from linear fitting of the Arrhenius e quation.

PAGE 104

104 The relationship between activation energy and flow rate c an be found in Figure 4 10 and Table 4 3. Activation energy increased with increasing gas flow rate, and pairwise reaction mechanism was more energy demanding than random addition. 4 .3. 3 .3 Reaction Orders w ith Respect t o H 2 a nd Propylene As mentioned previously, even 5 times excess of H 2 did not make 100% conversion of propylene to propane. To optimize the ratio between H 2 and propylene, as well as to gain further information about the kinetics of surface reactions, experiments were performed at three different temperatures (150, 25 0 and 350 o C) to vary the partial pressure of H 2 or propylene while keeping the other at a fixed value Results indicated that reactions with n H 2 and p H 2 showed distinct behaviors, further confirming that pairwise and random additions took place at diff erent sites on surface s To determine the reaction orders with respect to hydrogen, propylene partial pressure was fixed by means of keeping its flow rate constant. The total flow rate was kept always as 300 mL/min and propylene was set as 30 mL/min in on e set of studies and 150 mL/min in another. By changing H 2 pressure, kinetic data were obtained for three different temperatures and two different propylene partial pressures (Figure 4 11) For the case of propylene partial pressure as 0.1 atm (Figure 4 11A, C & E) reactions with p H 2 or pairwise pathway, had higher reaction order s than that with n H 2 at all three temperatures Both reaction orders also increased with increasing temperature. As higher reaction order means stronger dependence on its partial pressure, increasing p H 2 concentration and reaction temperature would lead to better hyperpolarization. And similar reaction orders were observed with respect to n H 2 and p H 2

PAGE 105

105 In contrast, for the case of propy lene partial pressure as 0.5 atm (Figure 4 11B, D & F), reactions with p H 2 or pairwise pathway, had lower reaction orders than that with n H 2 at all three temperatures. Figure 4 11 Reaction orders with respect to H 2 partial pressure at various temperatures and propylene concentrations under ALTADENA condition Propane CH 2 integral values were used. Red square: reactions with p H 2 Blue triangle: reactions with n H 2 A) 150 o C, P propylene =0.1 atm. B) 150 o C, P propylen e =0.5 atm. C) 250 o C, P propylene =0.1 atm. D) 250 o C, P propylene =0.5 atm. E) 350 o C, P propylene =0.1 atm. F) 350 o C, P propylene =0.5 atm.

PAGE 106

106 This difference even increased with increasing temperature. As propylene was in large excess, reaction was more sensitive to hydrogen partial pressure and thus higher reaction orders, as compared to the above case. Conversion rate, enhancement factor and pairwise selectivity were presented in Figure 4 12. Figure 4 12 Propylene to propane conversion and pairwise selectivity with respect to H 2 partial pressure at various temperatures and propylene concentrations under ALTADENA condition Red square: reactions with p H 2 Blue triangle: reactions with n H 2 A) 150 o C, P propylene =0.1 atm. B) 150 o C, P propylene =0.5 atm C) 250 o C, P propylene =0.1 atm. D) 250 o C, P propylene =0.5 atm. E) 350 o C, P propylene =0.1 atm. F) 350 o C, P propylene =0.5 atm.

PAGE 107

107 The conversion rate increased with increasing hydrogen partial pressure for all cases, including different propylene partial pressures and different reaction At propylene partial pressure of 0.1 atm, pairwise selectivity increased with increasing hydrogen partial pressure, while it was opposite at propylene partial pressure of 0.5 atm. When the surface was surrounded with more propylene, random addition was more favored and the access of p H 2 to lots of sites favoring pairwise selectivity might be blocked by propylene. But the relative p airwise selectivity was higher when propylene partial pressure was 0. 5 atm than 0. 1 atm in all three different temperatures. Thus to get high signal enhancem ent, high propylene concentration (50%) and low hydrogen concentration i s more desirable. On the other hand, reaction orders with respect to propylene were also studied at 250 and 350 o C, and plotted in Figure 4 13. In this case, hydrogen partial pressure was kept as 0.5 atm by keeping its flow rate as 150 mL/min while the total flow rate was 300 mL/min Figure 4 14 showed the conversion and pairwise selectivity data. Figure 4 13 Reaction orders with respect to propylene partial pressure at various te mperatures under ALTADENA condition Propane CH 2 peak integral was used. Red square: reactions with p H 2 Blue triangle: with n H 2 A) 250 o C, P H2 =0.5 atm. B) 350 o C, P H2 =0.5 atm.

PAGE 108

108 Reaction orders with respect to propylene were smaller than that of hydrogen. At 250 o C (Figure 4 13A) pairwise addition had a reaction order of 0.8, while random addition had 0.5 Pairwise pathway had stronger dependence on propylene concentration. While on the other hand, at the reaction temperature of 350 o C (Figure 4 13B), this was reversed. Pairwise addition had a reaction order of 0.4, while random addition had 0.7. At high temperature, random addition was more sensitive to the propylene concentratio n. Figure 4 14 Propylene to propane conversion and pairwise selectivity with respect to propylene partial pressure at various temperatures under ALTADENA condition Red square: reactions with p H 2 Blue triangle: with n H 2 A) 250 o C, P H2 =0.5 atm. B) 350 o C, P H2 =0.5 atm. Conversion (Figure 4 14) decreased with increasing propylene at both temperature, as hydrogen became less excess. At high temperature, conversion of propylene to propane was greatly affected by the emergence of by product. Pairwise selectivity increased with increasing propylene at 250 o C, but decreased at 350 o C. It seemed that a t high temperature, random addition was enhanced to a higher degree than pairwise addition.

PAGE 109

109 The reaction order data were summarized in Table 4 2. As higher reaction orders means stronger dependence of reaction rate on species concentration, adjusting hydrogen and propylene partial pressure seemed to be a suitable way to improv e product formation, or the degree of hyperpolarization. 4 .3. 3 .4 The E ffect o f Flow Rate o n Reaction Orders In previous section, all experiments were carried out with a total flow rate of 300 mL/min As discussed befo re, the flow rate would affect the signal enhancement and kinetic behavior. To further investigate that, anot her set of experiments were carried out at 300 o C and with a total flow rate of 700 mL/min The gas mixture was composed of N 2 H 2 and propylene. In one case, the catalytic dependence on hydrogen partial pressure was studied by fixing propylene partial pre ssure as 0.5 atm with a flow rate of 350 mL/min In another case, H 2 partial pressure was fixed and the dependence on propylene partial pressure was investigated. Thus, reaction orders with respect to either hydrogen or propylene were obtained (Figure 4 1 5 ). Figure 4 15 Reaction orders with respect to A) H 2 and B) propylene partial pressure at 300 o C under ALTADENA condition Total flow rate: 700 mL/min. For A, propylene flow rate=350 mL/min, partial pressure = 0.5 atm. For B, H 2 flow rate=350 mL/min, partial pressure = 0.5 atm

PAGE 110

110 In this case, random addition had a slightly higher reaction order than pairwise addition with respect to hydrogen partial pressure, but both were around 1. Quite differently, this relationship was reverse for the reaction orders with respect to propylene, with pairwise addition as 0.6 and random addition as 0.4 (Table 4 3). Figure 4 16 Propylene to propane conversion, enhancement factor and pairwise selectivity with respect to H 2 and propylene partial pres sure at 300 o C. Total flow rate: 700 mL/min. A) Conversion rate with various H 2 partial pressure. B) Enhancement factor and pairwise selectivity with various H 2 partial pressure. C) Conversion rate with various propylene partial pressure. D) Enhancement factor and pairwise selectivity with various propylene partial pressure. For A & B, propylene flow rate=350 mL/min, partial pressure = 0.5 atm. For C & D, H 2 flow rate=350 mL/min, partial pressure = 0.5 atm Figure 4 1 6 showed the propylene to propane conversion rate, PHIP enhancement factor as well as pairwise selectivity. As for fixed propylene partial

PAGE 111

111 pressure, increasing hydrogen pressure increased the overall conversion rate but decreased the pairwise selectivity (Figure 4 16 A&B). Contrary to that, with fixed hydrogen partial pressure, the conversion rate decreased with increasing propylene concentration but the pairwise selectivity increased (Figure 4 16 C&D ) The increase of hydrogen partial pressure increased the chance of propylene to collide and reac t with hydrogen to form propane at the cost of reduced pairwise selectivity But when hydrogen pressure was fixed, the more propylene, the less chance of prop ylene to collide with H 2 rather than another propylene molecule. Nevertheless, it seemed that more propylene were able to reach the sites favoring pairwise addition, leading to increas ed pairwise selectivity. 4 .3. 3 5 Probing Propylene Desorption Process o n Surfaces Figure 4 17 Propylene desorption process. Propylene was suggested to bind strongly to metal atoms through several different mechanisms. The adsorption process was studied by PHIP at 350 o C The catalyst was first saturated with propylene by flow ing propylene gas through it, followed by N 2 purge to remove the propylene and introduction of p H 2 and then NMR detection 1.5 2.0 2.5 3.0 3.5 4.0 0 5 10 15 Integral (CH 2 ) (A.U.) N 2 purge time (min) Propylene desorption process

PAGE 112

112 N 2 purge time was varied and the result i s presented in Figure 4 1 7. In all cases, hype rpolarization was observed. The propylene desorption process seemed to behave exponentially. Further experiments should be conducted at various temperature s to investigate the adsorption activation energies. 4 .4 Conclusion To the best of our knowledge, thi s work represent s the first observation of PHIP on supported Iridium particles. Taking advantage of our well designed continuous flow heterogeneous catalytic reactor setup, steady state reaction conditions were achieved over a wide range of temperature s T hus it was possible to study the temperature dependence of catalytic performance. The activation energies were determined based on the Arrhenius equation. Surface processes such as adsorption, diffusion and desorption are also known to affect the kinetic behavior of reactions. To investigate that, a series of different total flow rates were employed to study the temperature dependence. The reaction order s with respect to H 2 and propylene were investigated by changing the partial pressure of one component b ut keeping the other constant. Results indicated that p airwise and non pairwise hydrogen addition showed distinct behaviors. Moreover, it was found that the NMR signal enhancement depend ed on the total gas flow rate, reaction temperature and partial press ure of p H 2 or propylene These are valuable information towards understanding the surface processes and improving the efficiency of producing hyperpolarized species for potential applications.

PAGE 113

113 Table 4 1 Iridium nanoparticle supported on TiO 2 as the catalyst Courtesy of Dr. Wei Cheng Catalog Catalyst Metal Support Loading Calcination No. Name (w/w%) T ( ) CAT018 1.0IrTi350 Ir TiO 2 1 350 Table 4 2 Effect of total flow rate on the activation energy Error bar represents standard error from linear fitting of the Arrhenius equation. Total flow rate Activation Energy (mL/min) (kJ/mol) Random Pairwise 300 9.8 0.8 16.8 0.5 400 14 .0 0.9 20.4 2.0 500 16.8 1.1 26.5 2.8 600 15.2 0.5 24.5 1.2 700 16.5 0.8 27.2 1.9 Average 14.5 23.1 Table 4 3 Reaction orders at various temperatures and total flow rates Error bar represents standard error from linear fitting of the Arrhenius equation. Temperature Reaction Orders ( ) H 2 (propylene: 0.1) H 2 (propylene: 0.5) Propylene Pairwise Random Pairwise Random Pairwise Random 150 0.88 0.06 0.79 0.04 0.97 0.02 0.97 0.05 NA NA 250 1.08 0.11 1.07 0.08 1.18 0.05 1.66 0.04 0.82 0.01 0.48 0.05 350 1.72 0.06 1.5 0 0.06 2.04 0.06 2.77 0.13 0.36 0.02 0.67 0.12

PAGE 114

114 CHAPTER 5 OXIDE SUPPORTED PLATINUM / IRIDIUM CATALYSTS: THE ROLE OF SUPPORT AND PARTICLE SIZE IN HETEROGENEOUS CATALYSIS AND PHIP PERFORMANCE 5 .1 Background The co existence of several types of active sites which may have distinct catalytic behaviors is well recognized for supported metal catalysts. Exact reproducibility is very less likely even for catalysts prepared with the same procedure. Even slight and unintended deviation in the preparatio n or activation process can lead to noticeable variation in particle shape, size, morphology and thus catalytic activity. 65 86 111 112 One important aspect of research in heterogeneous catalysis is the relationship of catalyst structure with reactivity. It is even more important for PHIP applications, as p airwise addition pathway contributes to only a small portion of surface processes, a small improvement on surface structure may lead to huge difference in signal enhancement. The general catalytic activities in propylene hydrogenation with different metal types are as follows : Rh > Ir> Ru > Pt > Pd > Ni > Fe > Co > Os. 110 5 .1. 1 The Effect o f Particle Size a nd Support o n PHIP Performance Supporte d m etal particles tend to expose thermodynamically stable facets, such as (111) plane in fcc structures, on surfaces Thus the composition of different facets on surfaces varies. More thermodynamically stable facets contribute more in big particles. In addition, different atom ic arrangements and coordination numbers associated with various facets lead to non uniform adsorption properties and catalytic activities 113 PHIP studies with supported Pt and Pd nanoparticles suggested th at small particles with more edges and corners and less coordination numbers tend ed to

PAGE 115

115 facilitate the pairwise addition pathway on surfaces thus better hyperpolarization The presence of surface adsorbed species such as carbonaceous deposit s and reactan ts may effectively isolate surfaces into small static or dynamic domains where hydrogen diffusion is limited and pairwise pathway is en hanced 5 .1. 2 The Effect o f O xide Support o n Pairwise Selectivity Metal tends to minimize their surface potential energy by aggregating to large particles, but the use of supports in heterogeneous catalysis can prevent the sintering of active sites Other functions of supports include providing high surface area anchorag e sites, as well as thermal and mechanical stabilities. On the other hand, the presence of supports reduces the availability of active metal sites, and limit s the diffusion and transport of reactants to surface active sites. Heat transfer and dissipation m ay also be a problem. Thus it is important to choose proper supports to improve the catalytic performance. Another interesting phenomen on is s o called the s trong m etal s upport i nteraction (SMSI) due to the charge transfer in the metal support interface and particle morphology change Comparison of PHIP effects with different supports using p latinum or p alladium as the catalysts were reported in the literature. TiO 2 was found be the best support for PHIP activ ity which was attributed to the presence of SMSI leading to highest signal enhancement 37 5 .2 Experimental Methods In this chapter, the effects of particle size and support on PHIP performance were evaluated by running identical reactions with a series of catalysts (Pt or Ir) of different supports (TiO 2 or Al 2 O 3 ).

PAGE 116

116 The catalysts used in this chapter were prepared and characterized by Dr. Wei Cheng according the method described in C hapter 2 Calcination at different temperatures lead to different CO chemisorption values Catalysts were characterized mainly by chemisorption measurements. PHIP experiments were conducted based on the procedure described in C hapter 2 The total gas flow rate used here was 400 mL/min (N 2 /H 2 /propylene = 280/100/20 mL/min = 70/25/5%). This was selected for best performance a nd practical temperature control for steady state conditions. Pt c atalysts were activated at 170 o C and Ir catalysts at 350 o C before experiments. Spectra were accumulated with 1 6 scans with a 90 degree pulse and a recycle delay of 2 s. Kinetic data were corrected for relaxation effect based on equations (2 3), (2 10) and (2 13), and Table 2 2. 5 .3 Results and Discussion 5 .3.1 Pt Catalysts: t he Effect o f Support, Loading and Quantity A series of Pt catalysts were prepared (Table 5 1 & 5 2), and their catal ytic and PHIP performance were compared at various temperatures. Among the Pt catalysts, only CAT005 had Al 2 O 3 as the support, and all others had TiO 2 as the support. Spectra obtained with either n H 2 or p H 2 were presented in Figure 5 1 and Figure 5 2. A ll catalysts showed decent hyperpolarization and thus PHIP activity. Enhancement factor and pairwise selectivity (Figure 5 3) increased with increasing temperature for all catalysts. Different pairwise selectivity was observed for different catalysts, but it never reached 2% in this case.

PAGE 117

117 Figure 5 1. Catalytic and PHIP performance comparison for Pt catalyst s at 150 o C under ALTADENA condition Gas flow rate (total 400 mL/min): N 2 /H 2 /Propylene = 280/100/20 mL/min. A) n H 2 spectra. B) p H 2 spectra. Catalyst quantity: 10 mg. Pink: CAT019, 1.0PtTi350. Green: CAT023a, 1.0PtTi350. Blue: CAT017, 1.0PtTi350. Red: CAT011, 1.0PtTi450. Catalyst details see Table 5 1. Figure 5 2. Catalytic and PHIP performance comparison for Pt catalysts at 150 o C under ALTADENA condition Gas flow rate (total 400 mL/min): N 2 /H 2 /Propylene = 280/100/20 mL/min. A) n H 2 spectra. B) p H 2 spectra. Catalyst quantity: 10 mg. Black: CAT025a, 0.5PtTi350. Cyan: CAT009, 0.5PtTi450. Pink: CAT015, 0.75PtTi450. Green: CAT026b 0.75PtTi450. Blue: CAT005, 0.5PtAl350. Red: CAT030b, 2.5PtTi450. Catalyst details see Table 5 1.

PAGE 118

118 Figure 5 3. Conversion, enhancement factor and pairwise selectivity for particles with different sizes at various temperatures under ALTADENA condition A) Conversion for Pt catalysts (I) B) Conversion for Pt catalysts (I) C) Enhancement factor and pairwise selectivity for Pt catalysts (II) D) Enhancement factor and pairwise selectivity for Pt catalysts (II) Catalyst quantity: 10 mg.

PAGE 119

119 5 .3. 1 1 The Arrhenius Equation a nd Activation Energy The activation energy for propylene hydrogenation over Pt supported on alumina, silica or titania was reported to range from 23.8 to 58.5 kJ /mol, depending on the particle size, shape as well as support 114 118 For instance, Pt/Al 2 O 3 with 80% cubic shape of 8 10 nm size gave 40.6 2.1 k J /mol. 114 Results presented here indicated that catalyst with Al 2 O 3 as the support was less reactive and more energy demanding, with much higher activation energies than those with TiO 2 as the support CAT017: 1 .0 wt % Pt/TiO 2 It s temperature dependence was studied between 50 and 300 o C, with a total gas flow rate of 400 mL/min (N 2 /H 2 /Propylene = 280/100/20 mL/min ) for the best control of steady state conditions. Figure 5 4 Temperature dependence and activation energy for t emperature dependent studies of propylene hydrogenation under ALTADENA condition Gas flow rate (total 400 mL/min): N 2 /H 2 /Propylene = 280/100/20 mL/min. A) Signal integral versus temperature. B). The Arrhenius plot. Catalyst: CAT017 ( 1.0PtTi350 ), 1 0 mg. Error bar represents standard deviation of three measureme nts

PAGE 120

120 PHIP hyperpolarization was observed for all temperatures and increased with increasing temperature with a maximum at 300 o C. Figure 5 4 showed the change of integral values of the CH 2 group of propane. No evidence of reaching a diffusion/transport l imit was observed upto 300 o C and the reaction followed very well the Arrhenius behavior, under the control of intrinsic kinetics. Random and pairwise pathways showed distinct performance and a ctivation energies were derived based on the slope of the linear fitting i n (B) : 7.0 0.6 kJ/mol (ra ndom) and 1 4 9 0.8 kJ/mol (pairwise). Propylene to propane conversion PHIP enhancement factor and pairwise selectivity were shown in Figure 5 6 Both increased with increasing temperature up to 300 o C. Reaction a t 350 o C resulted in decreas ing conver sion due to the diffusion limit, but still increas ing pairwise selectivity, although to a less extent. Figure 5 5 Conversion rate, enhancement factor and pairwise selectivity for t emperature dependent studies of propylene hydrogenation. Gas flow rate (total 400 mL/min): N 2 /H 2 /Propylene = 280/100/20 mL/min. A) Propylene to propane conversion rate. B) Enhancement factor and pairwise selectivity. Catalyst: CAT017 ( 1.0PtTi350 ), 1 0 mg. E rror bar represents standard deviation of three measurements

PAGE 121

121 CAT005: 0.5 wt % Pt/Al 2 O 3 This wa s the only catalyst with Al 2 O 3 as the support tested in this section. Its temperature dependent catalytic and PHIP performance was studied follow ing the same pro cedure as described above. No PHIP hyperpolarization was observed at 50 o C and also negligible at 100 o C and relatively poor hyperpolarization was observed at higher temperatures compared to the catalyst above with TiO 2 as the support Figure 5 6 showed the change of integral values of the CH 2 group of propane. Diffusion star ted to take part in the kinetic control at temperatures higher than 2 50 o C. Activation energies (50 225 o C) were derived based on the slope of the linear fitting in (B): 12.5 0.8 kJ/mol (random) and 35.7 1. 2 kJ/mol (pairwise). Figure 5 6 Temperature dependence and activation energy for t emperature dependent studies of propylene hydrogenation under ALTADENA condition Gas flow rate (total 400 mL/min): N 2 /H 2 /Propylene = 280/100/20 mL/min. A) Signal integral versus temperature. B). The Arrhenius plot. Catalyst: CAT005 ( 0.5PtTi350 ), 1 0 mg.

PAGE 122

122 Propylene to propane conversion enhancement factor and pairwise selectivity were shown in Figure 5 7 Conversion rate inc reased with increasing temperature, while enhancement factor and pairwise selectivity started to decrease at 250 o C Figure 5 7 Conversion rate, enhancement factor and pairwise selectivity for t emperature dependent studies of propylene hydrogenation. Gas flow rate (total 400 mL/min): N 2 /H 2 /Propylene = 280/100/20 mL/min. A) Propylene to propane conversion rate. B) Enhancement factor and pairwise selectivity. Catalyst: CAT005 ( 0.5PtTi350 ), 1 0 mg. CAT023a: 1 .0 wt % Pt/TiO 2 The temperature dependent catalytic and PHIP performance of CAT023a was studied follow ing the same procedure as described above. PHIP hyperpolarization was observed for all temperatures and increased with increasing temperature with a maximum at 300 o C. Figure 5 8 showed the chan ge of integral values of the CH 2 group of propane. Diffusion sta r ted to take part in at temperature higher than 250 o C Random and pairwise pathways showed distinct performance and activation energies were derived based on the slope of the linear fitting i n (B): 11.2 0. 9 kJ/mol (random) and 20 .7 1. 5 kJ/mol (pairwise).

PAGE 123

123 Figure 5 8 Temperature dependence and activation energy for t emperature dependent studies of propylene hydrogenation under ALTADENA condition Gas flow rate (total 400 mL/min): N 2 /H 2 /Propylene = 280/100/20 mL/min. A) Signal integral versus temperature. B) The Arrhenius plot. CAT023a ( 1.0PtTi350 ), 1 0 mg. Figure 5 9 Conversion rate, enhancement factor and pairwise selectivity for t emperature dependent studies of propylene hydroge nation. Gas flow rate (total 400 mL/min): N 2 /H 2 /Propylene = 280/100/20 mL/min. A) Propylene to propane conversion rate. B) Enhancement factor and pairwise selectivity. Catalyst: CAT023a ( 1.0PtTi350 ), 1 0 mg.

PAGE 124

124 Propylene to propane conversion, enhancement fac tor and pairwise selectivity were shown in Figure 5 9. Conversion rate started to decrease at temperature higher than 250 o C, but enhancement factor and pairwise selectivity increased with increasing temperature even up to 350 o C. 5 .3. 1 2 The Effect o f Catalyst Amount Here, 20 mg of CAT017 was used and NMR spectra were acquired with 128 scans (rec eiver gain: 8192) Hyperpolarization was observed at all temperatures. Figure 5 1 0 showed the change of integral values of the CH 2 group of propane. Both pairw ise and random addition products increased with increasing temperature up to 300 o C, and the activation energies were derived: 5.4 0. 4 kJ/mol (random addition) and 25. 2 1. 4 kJ/mol (pairwise addition). Figure 5 10 Temperature dependence and activation energy of propylene hydrogenation under ALTADENA condition Gas flow rate (total 400 mL/min): N 2 /H 2 /Propylene = 280/100/20 mL/min. A) Signal integral versus temperature. B) The Arrhenius plot. Catalyst: CAT017 ( 1.0PtTi 350 ), 2 0 mg.

PAGE 125

125 Figure 5 1 1 showed the conversion enhancement and pairwise selectivity. Except for 350 o C, all values followed the same pattern as increasing with increasing temperature up to 300 o C. Figure 5 11 Conversion rate, enhancement factor and pairwise selectivity for t emperature dependent studies of propylene hydrogenation. Gas flow rate (total 400 mL/min): N 2 /H 2 /Propylene = 280/100/20 mL/min. A) Propylene to propane conversion rate. B) Enhancement factor and pairwise selectivity. Catalyst: CA T017 ( 1.0PtTi350 ), 2 0 mg. The effect of catalyst amount on catalytic and PHIP performance was compared in Figure 5 1 2 The more catalyst, the higher conversion rate as demonstrated in (A). But for enhancement factor and pairwise selectivity, too much catalyst did not seem to be helpful. Data obtained with 10 and 20 mg were comparable in terms of pairwise selectivity But for the case with 50 mg, enhancement factor started to decrease at 150 o C It seemed that the increase of catalyst amount increased t he chance of random addition more significantly than pairwise addition.

PAGE 126

126 Figure 5 12 Comparing the effect of catalyst quantity on the c onversion rate, enhancement factor and pairwise selectivity for t emperature dependent studies of propylene hydrogen ation under ALTADENA condition Gas flow rate (total 400 mL/min): N 2 /H 2 /Propylene = 280/100/20 mL/min. A) Propylene to propane conversion rate. B) Enhancement factor and pairwise selectivity. Catalyst: CAT017 ( 1.0PtTi350 ) Red square: 10 mg. Blue triangle: 20 mg. Green star: 50 mg. 5 .3. 1 3 The Effect o f G as Composition The hydrogen partial pressure dependent studies of catalytic and PHIP performance of CAT009 (50 mg) were studied at 150 o C with a fixed total flow rate of 400 mL/min and propylene 20 mL/min As propylene partial pressure was so low (0.05 atm), reaction rate reached maximum quickly, limited by the availability of surface bounded propylene. For random addition, an initial reaction order of 0.82 0.06 with respect to hydrogen was observed (Figu re 5 1 3 ). With increasing hydrogen partial pressure, the conversion of propylene to propane increased, but it seemed that random addition dominated the whole reaction more leading to decrease of PHIP enhancement factor and pairwise selectivity (Figure 5 1 4 ).

PAGE 127

127 Figure 5 13 Reaction orders with respect to h ydrogen partial pressure of propylene hydrogenation at 150 o C under ALTADENA condition using CAT009 ( 0.5 PtTi 4 50 ). Gas flow rate (total 400 mL/min): N 2 /H 2 /Propylene = 280/100/20 mL/min. A) n H 2 spectra. B) p H 2 spectra. Catalyst quantity: 50 mg. Figure 5 14 Conversion rate, enhancement factor and pairwise selectivity of propylene hydrogenation at 150 o C on CAT009 ( 0.5 PtTi 4 50 ). Gas flow rate (total 400 mL/min): N 2 /H 2 /Propylene = 280/100/20 mL/min. A) Propylene to propane conversion rate. B) Enhancement factor and pairwise selectivity. Catalyst quantity: 50 mg.

PAGE 128

128 5 .3. 2 Ir Catalyst: t he Effect o f Support a nd Loading Iridium catalysts (Table 5 2) were compared for catalytic performance and PHIP enhancement. A representative set of spectra obtained by reacting with n or p H 2 was plotted as in Figure 5 1 5 Hyperpolarization (net alignment) according to ALTADENA protocol was observed. Figure 5 15. Representative overlapped spectra for propylene hydrogenation at 150 o C under ALTADENA condition using Ir/TiO 2 with n H 2 and p H 2 Blue: n H 2 Red: 50% p H 2 (spectrum subtracted using n H 2 Catalyst (CAT043b 1.0IrAl450 ): 1% Iridium supported on Al 2 O 3 calcination temperature 450 o C. All catalysts except CAT043 a were tes ted in propylene hydrogenation. These catalysts showed great catalytic performance but CAT04 4b and CAT04 4c did not show PHIP activity or hyperpolarization These catalysts were compared for their catalytic and PHIP performance (Figure 5 16).All catalysts had similar temperature dependence, pairwise selectivity increased with increasing temperature up to around 200 o C, and then decreased with the emergence of by product methane.

PAGE 129

129 Figure 5 16. Temperature dependent propylene hydrogenation under ALTADENA condition on supported Ir catalysts A) Reaction with n H 2 B) Reaction with p H 2 C) Propylene to propane conversion. D) Enhancement factor and pairwise sele ctivity. Gas flow rate: 400 mL/min (N 2 /H 2 /propylene = 280/100/20 mL/min) Red square: CAT020, 0.5IrTi350 Blue triangle: CAT045b, 0.5IrAl450 Green dot: CAT043b, 1.0IrAl450 Pink star: CAT022, 1.0IrAl350 Cyan diamond: CAT021, 1.0IrTi350 5 .3. 3 Pt & Ir Catalyst: t he Comparison Results indicated that Ir catalysts showed higher conversion than Pt catalysts b ut in terms of pairwise selectivity, Ir catalysts had much smaller pairwise selectivity.

PAGE 130

130 5 .4 Conclusion A series of supported p l atinum and i ridium ca talysts were prepared and their catalytic and PHIP kinetics were compared. The amount of catalyst seemed to affect the conversion and pairwise selectivity in an opposite way. PHIP optimization is possible with optimum conditions of temperature, hydrogen p artial pressure and catalyst amount etc. Generally, Pt showed better pairwise selectivity than Ir catalysts And TiO 2 is a better support than Al 2 O 3 for both Pt and Ir catalysts, in agreement with literature reports. Work towards preparing more catalysts with a larger variety of particle sizes and supports, such as SiO 2 ZrO 2 is in progress.

PAGE 131

131 Table 5 1. Platinum nano particles supported on oxides as catalysts Courtesy of Dr. Wei Cheng Catalog Catalyst Metal Support Loading Calcination No. Name (w/w%) T ( ) CAT025a 0.5PtTi350 Pt TiO 2 0.5 350 CAT009 0.5PtTi450 Pt TiO 2 0.5 450 CAT015 0.75PtTi450 Pt TiO 2 0.75 450 CAT026b 0.75PtTi450 Pt TiO 2 0.75 450 CAT017 1.0PtTi350 Pt TiO 2 1 350 CAT019 1.0PtTi350 Pt TiO 2 1 350 CAT023a 1.0PtTi350 Pt TiO 2 1 350 CAT011 1.0PtTi450 Pt TiO 2 1 450 CAT030b 2.5PtTi450 Pt TiO 2 2.5 450 CAT005 0.5PtAl350 Pt n Al 2 O 3 0.5 350 Table 5 2. Iridium nanoparticles supported on oxides as catalysts. Courtesy of Dr. Wei Cheng Catalog Catalyst Metal Support Loading Calcination PHIP No. Name (w/w%) T ( ) Active CAT020 0.5IrTi350 Ir TiO 2 0.5 350 YES CAT021 1.0IrTi350 Ir TiO 2 1 350 YES CAT045b 0.5IrAl450 Ir n Al 2 O 3 0.5 450 YES CAT022a 1.0IrAl350 Ir n Al 2 O 3 1 350 YES CAT043b 1.0IrAl450 Ir n Al 2 O 3 1 450 YES CAT044b 2.0IrAl450 Ir n Al 2 O 3 2 450 NO CAT044c 2.0IrAl650 Ir n Al 2 O 3 2 650 NO

PAGE 132

132 CHAPTER 6 EXTENDING THE LIFETIME OF PARAHYDROGEN INDUCED POLARIZATION THROUGH POLARIZATION TRANSFER 6 .1 Background 6 .1.1 Longitudinal (Spin Lattice) Relaxation Time Constant T 1 The lifetime of hyperpolarization defines the time window for potential applications. I t is mainly limited by the spin lattice relaxation time constant (T 1 ) and is on the order of a few tens of seconds for protons in favorable cases. 119 This relaxation is caused largely by transient magnetic field fluctuations produced by inter a nd intramolecular spin interactions with spectral density close to the Larmor frequency. The relaxation constant depends on the molecular structure and details of the motion (rotational motion of the same molecule and translational motion of the other mole cules due to collisions), as well as the applied magnetic field. A variety of relaxation mechanisms, such as dipole dipole interaction, spin rotation interaction and chemical shift anisotropy ( CSA ) allow nuclear spins to exchange energies with their surrou ndings, leading to the thermal equilibration of spin populations. 120 121 Hyperpolarization transfer to spins with longer T 1 is of particular interest for extending the lifetime. 6 .1. 2 T 1 Measurement: t he Inversion Recovery Method The relaxation time constant T 1 can be determined experimentally with either the saturation recovery or inversion recovery techniques Figure 6 1. The inversion recovery pulse sequence.

PAGE 133

133 The basic pulse sequence of the inversion recovery method (Figure 6 1) consists of the followi ng steps 1) A recovery delay (d1) ensures the signal relaxed to thermal equilibrium 2) A 180 pulse (p1) inverts the magnetization, from the positive z axis to the negative z axis 3) D uring the d2 delay relaxation along the z axis takes place and m agnetization returns to wards the original equilibrium z magnetization Here, d 2 is varied so that time dependent magnetization can be sampled. (4) A final 90 pulse (pw) creates transverse magnetization and a cquisition is performed as usual The time dependence of the magnetization parallel (longitudinal) to the a pplied field can be written as: ( 6 1 ) This has th e general solution: ( 6 2 ) In the case of inversion recovery : ( 6 3 ) Thus the magnetization over time can be expressed as follows: ( 6 4 ) 6 .1. 3 Hyperpolarization Transfer t owards Spins w ith Extended Lifetime The hyperpolarization from PHIP is not limited to the two protons originating from p H 2 It has been demonstrated that hyperpolarization can be transferred to other spins (both hetero and homo nucleus) originally present on the substrate (i.e., involved in a J coupling network), 122 127 leading to sharing of hyperpolarization among coupled spins In the ALTADENA case, all homonuclear spins are essentially chemically equivalent in spectrum line width. 128

PAGE 134

134 The s haring a nd redistribution of hyperpolarization have been proposed to occur through either coherent pathways among scalar coupled spins over chemical bonds, 129 131 or incoherent processes such as through space cross relaxation as exemplified by the n uclear Overhauser ef fect (NOE) 1 T he final polarization distribution can be controlled to some extent, by appropriate choice of parameters such as the magnetic field for sample preparation mixing time, field variation profile and radio frequency pulse flip angles. 128 130 131 Hyperpolarization transfer towards coupled homonuclei has been demonstrated experimentally within 2 bonds with significant enhancement 128 and more than three weekly coupled bonds but with fairly detectable enhancement 131 6 .1. 4 Rationale The direct dipole dipole (DD) interaction is the main relaxation mechanism for protons in liquids, as constant molecular tumbling creates fluctuating magnetic fields. While this mechanism operates most effectively be tween directly bonded nuclei, its resonating and moving nuclei, the distance between them and the rate of motions. In this chapter the carbon carbon double bond of unsatu rated aldehydes was para hydrogenated. The carbonyl proton normally has a longer relaxation time constant, due to weaker DD interactions. This is because the DD interaction decreases with increasing distance between dipoles. The feasibility of hyperpolarization transfer towards the carbonyl proton (H a in Scheme 6 1 ) wa s considered.

PAGE 135

135 Scheme 6 1 P ara h ydrogenation of unsaturat ed aldehyde. R 1 R 2 R 3 are substituting groups; H 1 H 2 come from p H 2 and H a is the polarization transfer target 6 .2 Experimental Methods 6 .2.1 NMR Spectrometer a nd Materials For this chapter, all experiments were carried out in UF Chemistry NMR Facility with an Agilent Inova 500 MHz spectrometer equipped with a 5 mm liquids probe. Chemicals were purchased from Sigma Aldrich and used as obtained. Deuterated solvents were purchased from Cambridge Isotope Laboratories (CIL). 5 mm thin wall precision quick p ressure valve NMR sample tube (catalog no.: 528 QPV 8) were purchased from Wilmad LabGlass. 6 .2. 2 PHIP Procedure: t he ALTADENA Protocol catalyst (3.7 mg, 5 mM) was dissolved in 0.6 m L of the solvent [ C 6 D 6 /CD 3 OD (1/1) ] and the solution was transferred to the NMR tube and sealed. This solution was degassed three times by the freeze pump thaw procedure (freezing with dry ice/acetone at 78 o C, and pumping to 10 2 mtorr) before and after introducing 0.05 m L (1.77 M) of the substrate acrolein. A 1 H NMR spectrum was acquired. Then the sample was charged with around 3 bar of 50% enriched p H 2 and the sample was kept frozen in the dry ice/acetone cooling system while b eing transported to the NMR lab. There, the NMR

PAGE 136

136 tube was thawed in the air during around 1 min and another 1 H NMR spectrum was acquired again. After that, the tube was ejected out of the magnet, shaken by hands rigorously for 10 30 seconds nearby the ultra shielded magnet (magnetic field: 1 2.5 mT) and inserted back into the magnet over the period of 6 s A third 1 H NMR spectrum was acquired. A last spectrum was acquired after a few minutes for relaxed signal. All spectra were acquired with a single scan us ing a 90 o pulse for maximum signal detection 6 .3 Results and Discussion 6 .3. 1 Acetalization : Equilibration o f Propanal a nd Acetal i n Methanol Due to its active chemical nature, propanal is in equilibrium with its dimethoxyacetal form once getting in contact with deuterated methanol in the solvent mixture (Scheme 6 2). The ratio between the two species can be derived by comparing the peak integra l of the CH 3 group in each molecule (Figure 6 2). In the case of a mixed solvent of C 6 D 6 /CD 3 OD (1:1) 75% of propanal was converted to its acetal form. Scheme 6 2 Equilibration of propanal and acetal. Solvent: C 6 D 6 /CD 3 OD (1:1), J is the scalar cou pling constant; ratio determined by comparing the integral value of corresponding CH 3 group in 1 H NMR spectrum ( Figure 6 2 )

PAGE 137

137 Figure 6 2. Propa nal and acetal in equilibrium. 1 H NMR (C 6 D 6 /CD 3 OD (1:1), 500 MHz, ppm) propanal : 9.49 ( t 1H, CH), J =1 Hz; 2.13 ( m 2H, CH 2 ), J =7.5, 1 Hz; 0.89 ( t 3H, CH 3 ), J =7.5 Hz; acetal : 4.43 ( t 1H, CH), J =5.5 Hz; 1.64 ( m 2H, CH 2 ), J =7.5, 5.5 Hz; 0.94 ( t 3H, CH 3 ), J =7.5 Hz. Data acquired together with Daniel Schulman. 6 .3. 2 Solvent a nd Chemical Environment Dependent T 1 Measurements Using the inversion recovery technique, T 1 values of propanal were measured in different solvent s with and without metal catalysts. The removal of trace paramagnetic oxygen by standard freeze pump thaw procedure along with its DD interaction, significantly increase d the T 1 values of all protons. T 1 also varie d in different chemical environments ( Figure 6 3 )

PAGE 138

138 Figure 6 3. Comparison of T 1 values of propanal and acetal. Solvent: C 6 D 6 /CD 3 OD (1:1); CH has a longer T 1 than CH 2 and CH 3 in all cases. Data acquired together with Daniel Schulman. Figure 6 4. T 1 of propanal in different solvents. T 1 and chemical shift values of propanal in different solvents are different. Spectra were acquired with an Agilent Inova 500 MHz spe ctrometer. Green: C 6 D 6 /CD 3 COCD 3 (1:1) ; 9.56 ( t 1H, CH), J =1.5 Hz; 2.17 ( m 2H, CH 2 ), J =7.5, 1.5 Hz; 0.9 ( t 3H, CH 3 ), J =7.5 Hz. Purple: CD 3 COCD 3 ; 9.74 ( t 1H, CH), J =1.5 Hz; 2.45 ( m 2H, CH 2 ), J =7.5, 1.5 Hz; 1.03 ( t 3H, CH 3 ), J =7.5 Hz. Red: C 6 D 6 /CD 3 OD (1:1); 9.49 ( t 1H, CH), J =1.5 Hz; 2.13 ( m 2H, CH 2 ), J =7.5, 1.5 Hz; 0.89 ( t 3H, CH 3 ), J =7.5 Hz. Data acquired together with Daniel Schulman. 0 50 CH CH2 CH3 10.1 2.8 3.1 40.8 17.2 15.9 T 1 (S) Structure dependence of T 1 Acetal Propanal 0 20 40 60 80 100 CH CH2 CH3 95.4 29.8 25.1 74 30.4 25.9 40.8 17.2 15.9 T 1 (S) Solvent dependence of T 1 Benzene/Acetone (1:1) Acetone Benzene/MeOD (1:1)

PAGE 139

139 The T 1 and chemical shift values were solvent dependent as well ( Figure 6 4 ). In a deuterated benzene/aceton e mixture, the T 1 values of all protons, especially the aldehyde CH, were longer than the corresponding values in acetone only or benzene/methanol solvents. Further efforts are needed to understand this difference. 6 .3. 3 Hyperpolarization Transfer t owards Spins w ith Extended Lifetime Figure 6 5. Hyperpolarized aldehyde through spontaneous polarization transfer from para Spectrum acquired after the signal relaxed back to therm al equilibrium. B) Spectrum acquired after shaking the tube for 10 s outside of the magnet followed by insertion into the magnet within 6 s for detection. C) Spectrum acquired before shaking the tube. Propanal aldehyde ( 9.47 ppm, CH) is hyperpolarized th rough spontaneous polarization transfer from protons originating from para hydrogen ( 2.10 ppm, CH 2 ; 0.89 ppm, CH 3 ). Spectra acquired with an Agilent Inova 500 MHz spectrometer equipped with a 5 mm liquids probe together with Daniel Schulman.

PAGE 140

140 Polarizat ion transfer as well as PHIP enhancement was demonstrated with the homogeneous hydrogenation of acrolein to propanal under the ALTADENA protocol ( Scheme 6 3 ) Interestingly significant signal enhancement of up to 100 times was observed for the aldehyde H ( Figure 6 5 ), which is coupled with para hydrogenation sites with only a small coupling constant, 1.0 Hz. As t h e carbonyl proton of the product propanal has a significantly longer T 1 than the methyl and methylene groups where p H 2 was added (Figure 6 3) Its hyperpolarization also persist ed longer than those of methylene and methyl groups as the carbonyl CH group remained hyperpolarized after the polarization of the CH 2 and CH 3 groups had decayed ( Figure 6 5 ) Thus the polarization transfer towards the CH group extend ed the lifetime of para hydrogen induced hyperpolarization Scheme 6 3 P ara h 6 .3. 4 Magnetic Field Profile f or ALTADENA Experiments To understand the details about the polarization transfer, the magnetic field profile was measured with a Lakeshore Gauss meter. The change of the magnetic field over the distance from the top of the magnet ( Figure 6 6 ) and over time during the insertion process ( Figure 6 7 ) were plotted A linear increase wa s assumed for the dropping speed of the NMR sample tube.

PAGE 141

141 Figure 6 6. Magnetic field variation profile over the distance. Zero distance point: the top of the magnet; magnetic fi elds measured with a Lakeshore Gaussmeter 421 together with Daniel Schulman. Figure 6 7. Magnetic field variation profile over the time. Zero time point: sample tube was placed on the top for insertion after vigorous shaking. Data acquired together w ith Daniel Schulman. 0 2 4 6 8 10 12 14 0 20 40 60 80 100 Magnetic field (Tesla) Distance from the top of the magnet (cm) Field variation profile over distance 0 2 4 6 8 10 12 14 0 2 4 6 8 Magnetic field (Tesla) Time (s) Field variation profile over time

PAGE 142

142 6 .3. 5 Hyperpolarization Transfer Mechanisms 6 3 5 .1 Nuclear Overhauser Effect ( NOE ) a nd 1D NOESY The NOE effect depends on the dipolar mechanism acting as the dominant relaxation pathway. The major factors are molecular tumbling frequencies and internuclear distances. The intensity of the NOE is proportional to r 6 where r is the distance between the two spins. The s elective 1D NOESY spectrum ( Figure 6 8 ) of the propanal sample indicate d no observable NOE effect between the aldehyde H peak and neighboring protons in this high field (11.7 Tesla 1 H NMR 500 MHz ). Figure 6 8. 1D NOESY of propanal & acetal mixture. catalyst in C 6 D 6 /CD 3 OD mixed solvent, degassed. A) 1D NOESY. B) Normal spectrum. The spectrum was acquired with an Agilent Inova 500 MHz did not sh ow observable NOE effect. Data acquired together with Daniel Schulman.

PAGE 143

143 6 3 5 2 Magnetization Transfer t hrough J Coupling Magnetization can be shared among scalar coupled spins at low fields through isotropic mixing. This is similar to the principle of SABR E, where hyperpolarization is achieved without hydrogenation but through coupling to the same metal center. Although it is hard to visualize experimentally, density matrix simulation would provide an alternative approach. 6 .4 Conclusion Hyperpolarization transfer towards the propanal aldehyde proton (weakly coupled with neighboring methylene CH 2 J =1.0 Hz) was observed with over 1 0 0 times enhancement. This hyperpolarization persist ed for an extended period of time, partially due to its ex treme ly long spin lattice relaxation time constant T 1. In the presence of deuterated methanol, the resulting propanal was in equilibrium with acetal. The fact that a cetal CH peak was also observed to be hyperpolarized to a lesser extent in some cases may s uggest that the time scale of propanal acetal conversion is longer than the lifetime of hyperpolarization. Further investigation with various substrates ( Figure 6 9 ) and solvents may be conducted to search for higher transfer extents and longer lifetimes. As nuclear Overhauser effect is inversely proportional to the magnetic field, it is also possible that cross relaxation occurs in low magnetic fields where the sample is shaken. Theoretical approaches using numerical density matrix simulations may be emplo yed to investigate the polarization transfer mechanisms. These nuclear spin states with extended lifetime have many potential applications, including stud ies of slow molecular motion and molecular transport, the

PAGE 144

144 storage and transport of hyperpolarized nuclear spin order, and as a probe for local molecular geometry and dynamics. 119 F igure 6 9. Other unsaturated aldehydes.

PAGE 145

145 CHAPTER 7 EXTENDING THE LIFETIME OF PARAHYDROGEN INDUCED POLARIZATION THROUGH SINGLET STATE 7 .1 Background The coupling of two spin 1/2 particles leads to a composite system with total spin 0, which is known as a singlet state, and total spin 1, con sisting of three triplet states 132 Zeeman interaction splits the triplet into three energy levels, where spin lattice relaxation (T 1 ) takes place. T 1 typically defines the time frame of hyperpolarization. On the other hand, singlet triplet interconversion is normally dipole forbidde n and slow and to a much extent sensitive to the surroundings Its time constant, denoted as T s defines the lifetime of singlet order in a much bigger time scale than T 1 This opens the possibility of generating long lived nuclear spin order by storing it in the singlet state. An eight fold lifetime extension was reported for the singlet state of 15 N labeled nitrous oxide in solution, which exhibited a 25 min decay time constant. 133 Nuclear singlet state represents a promising way to extend the hyperpolarization lifetime, and has been applied to the stu dies of chemical exchange in slow dynamic processes, 134 protein folding/unfolding, 135 and measuring slow molecular diffusion coefficients 136 139 It is also applicable to storing and transporting of hyperpolarized nuclear spin orders generated by parahydr ogen induced polarization (PHIP) 128 140 148 In the reaction of acetyle ne with p H 2 over Pt/TiO 2 catalyst, i t was demonstrated that polarization from PHIP could be stored in ethylene as a singlet state for minutes and then released by a further reaction with electrophilic arenesulfenyl chloride to break the symmetry and exhibit the hyperpolarization 149

PAGE 146

146 7 .2 Experimental Methods 10 mg of 1.0 wt % Pt/TiO 2 (CAT023a, Table 3 1) was placed in the bottom of a 10 mm NMR probe heated to 115 o C. The catalyst was activated under H 2 flow and used in the reaction (Scheme 7 1). Norma l ethylene was used to establish steady state reaction conditions before a brief burst of ethylene 13 C was introduced to the system and detected with a single scan of 45 o pulse Scheme 7 1 Hydrogenation of ethylene 13 C over Pt/TiO 2 7 .3 Results and Discussion PASADENA pattern signal was observed as shown in Figure 7 1 Red line represents the spectrum obtained using 50% p H 2 and the blue one from n H 2 experiments. Clearly, anti phase pattern was observed with significan t enhancement. 7 4 Conclusion The significance of this research lies on the fact that the coupling between 13 C and 1 H can be removed under proper radio frequency pulse sequence, thus making the parahydrogenation product ethane symmetric and the spin order locked in singlet state. This hyperpolarization can be retrieved in a later time upon turning off decoupling pulse and thus symmetry break. Studies towards this end is in progress.

PAGE 147

147 Figure 7 1. Ethylene 13 C parahydrogenation under PASADENA condition Catalyst: 10 mg of 1.0 Pt/TiO 2 (CAT023a). Reaction Temperature: 115 o C. Gas flow rate: H 2 /Ethylene 13 C = 400/20 mL/min. NMR parameters: 45 degree pulse (8 us, 6 db), 8 scans, d1 = 2 s, receiver gain = 8 k, line broadening = 15 Hz. A) Overlap of spectra obtained with p H 2 and n H 2 B) Spectrum obtained with p H 2 C) Spectrum with n H 2 Figure 7 2. Chemical shifts and J coupling constants in ethane 13 0.782 ppm. J H1 C13 = 125 Hz, J H2 C13 = 4 .67 Hz, J H1 H2 = 8 Hz.

PAGE 148

148 Figure 7 3 Comparison of experimental spectra and density matrix simulation results. Line broadening: 15 Hz. A) Spectra obtained with p H 2 B) Spectra obtained with n H 2 Blue: spectra obtained under PASADENA condition. Red: spectra obtained with density matrix simulation using SpinDynamica /Mathematica package (see Appendix A )

PAGE 149

149 CHAPTER 8 CONCLUSIONS Para Hydrogen Induced Polarization (PHIP) has been serving as a crucia l tool in the exploration of catalytic reaction mechanisms and intermediates. Its application towards generating hyperpolarized molecular probes for in vivo imaging further boosts this research area. Thus, high efficient production of metal free molecular probes with long lived hyperpolarized nuclear spins is a primary focus. The easy separation of hyperpolarized product from transition metals makes heterogen eous PHIP an area of particular interest for biomedical applications D issociative chemisorption followed by rapid diffusion and tumbling are normally expected for hydrogen molecules on surfaces, which leads to a loss of the high spin order of singlet p H 2 Thus, systematic studies with a great variety of particle size, shape, morphology, and support are very meaningful in search of optimal conditions for pairwise selectivity on surfaces. In C hapter 2 we first assembled a gas handling system and continuou s flow reactor. Para hydrogen was enriched to 50% by passing normal hydrogen through a converter filled with activated charcoal and immersed in liquid nitrogen. Gas phase heterogeneous catalysis was well controlled under steady state condition s with mass flow controllers and LabVIEW program. A variety of catalysts (Pt and Ir) with different particle sizes and supports (Al 2 O 3 and TiO 2 ) were prepared and characterized by Dr. Wei Cheng in Prof. Hagelin according to literature methods As i t takes three to five times relaxation time constant T 1 to build up the thermal equilibrium when molecules transport from zero field to high magnetic field, partial polarization should be expected at high gas flow rat e. The dependence of signal on flow

PAGE 150

150 rat e was investigated and equations for correcting the data were developed. Effective relaxation time constant s and relaxation correction factors for both thermally polarized and hyperpolarized magnetizations were calculated for both propylene and propane. In C hapter 3, selectivity for p ropyne hydrogenation was studied using Pt/TiO 2 as the catalyst. It was found that selectivity was affected by reaction conditions, such as reaction temperature, hydrogen partial pressure and gas flow rate. At high temperature, partial reduction product propylene desorbs more rapidly and had reduc ed contact time with catalysts for further reduc tion to propane. Other interesting observations included surface isomerization, polarization transfer and the kinetic isotope effect. In C hapter 4, Ir supported on TiO 2 was prepared and detailed kinetic studies were carried out. Temperature dependent studies based on the Arrhenius equation gave the activation energies and results indicated that pairwise pathway was more energy demanding and had a higher activation energy than random addition mechanism. It was also demonstrated that hyperpolarization could be enhanced with optimal p H 2 partial pressure and reaction orders with respect to hydrogen or propylene partial pressures were presented. More studies on propylene hydrogenation reactions over Pt or Ir catalysts supported on TiO 2 or Al 2 O 3 were conducted in C hapter 5. For Pt catalysts, TiO 2 was a much better support than Al 2 O 3 potentially due to the strong metal support interaction. Generall y speaking, Pt catalysts showed better pairwise selectivity than Ir counterparts. As the lifetime of hyperpolarization defines the time frame of potential applications, research towards extending the lifetime is of great importance. Polarization transfer to other nuclei, both homo and hetero is one of the promising ways. In

PAGE 151

151 C hapter 6 polarization transfer to wards coupled spins with longer lifetime was investigate d Spin lattice relaxation time constant T 1 was measured using the inversion recovery method and result s indicated that protons had different T 1 value s depending on the locat ion in the molecule and their surroundings. In the hydrogenation of acrolein to propanal, aldehyde proton was observed hyperpolarized and longer lived. Though normally not expected, as this proton d id not originate from p H 2 its hyperpolarization might be gained through J coupling with other spins. As singlet state has even longer lifetime beyond T 1 its incorporation into PHIP could greatly enhance the capacity of potential application s and progress towards this end was initiated in C hapter 7. Here we hydrogenated single C 13 label ed ethylene and converted it into ethane 13 C. The presence of 13 C, together with the discrepancy in the coupling between this 13 C and the two protons originating from p H 2 broke the symmetry and released the hyperpolarization But this coupling can be removed to ensure a singlet state, a s 13 C 1 H decoupling is routinely used in conventional 13 C NMR Hyperpolarization could be thus stored in singlet state and released later upon introducing back this scalar coupling. The ultima te goal of this research is to generat e long lived and hyperpolarized nuclear spin states without metal contamination. This may be resolved through optimizing PHIP based on heterogeneous catalysis. Catalyst optimization towards clarifyi ng the factors governing heterogeneous pairwise selectivity is still underway.

PAGE 152

152 CHAPTER 9 FUTURE STUDIES As particle size is affecting the catalytic performance, further efforts towards more defined particles with narrow size distribution will be investigate d On the other hand, catalytic activity is also significantl y dependent on the morphologies 150 P articles with different structure, sizes and shapes, such as nanocubes, nanobars, nanorods and nanoprisms (and thus the ratio of exposed (111) and (110) surfaces) 111 113 151 155 are possible to be prepare d according to literature methods 156 159 It has been suggested that different active centers co exist on supported Pt catalysts and pairwise mechanism tends to happen on the less dense (such as (100) or (110)) crystal faces where carbon deposition is more facile as compared to the close packed P t (111) surface. 37 Secondly, the electronic structure, or the center of the d band with respect to the Fermi level was suggested to play a significant role in the catalytic performance 158 159 It is possible to fine tune this property by mixing two or more metals novel metal embedded metal organic frameworks (MOFs) 160 164 and atomically precise Au clusters 165 166 are also showing promising hydrogenation activities. The use of MOFs in heterogeneous catalysis has numerous unique and outstanding adva ntages. In addition to extraordinarily high surface areas, their crystal structures, compositions and nanoporosity properties can be well controlled in a systematic way to enhance catalytic activities and selectivity. In the meantime the atomic level stra tegy of Au clusters with unique atom packing structures and electronic properties is emerging as a new design of nanocatalysts to overcome the poly dispersity of conventional nanoparticle catalysts. 166 PHIP oriented mechanistic investigations using a systematic variety of atomically precise noble metal catalysts

PAGE 153

153 supported on or embedded in MOFs may open up new oppor tunities of achieving high pair wise selectivity. 58 for the optimum catalyst, the energy of interaction between metal surfaces and p H 2 molecules should be strong enough for activation but not too strong to permit the retention of singlet nuclear spin orders and thus pairwise selectivity.

PAGE 154

154 APPENDIX A DENSITY MATRIX SIMULATION OF ETHYLENE 13 C PARAHYDROGENATION This simulation wa s carried out using SpinDynamica/Mathematica package version 2.8.2, developed by Prof Malcolm Levitt. ( http://www.spindynamica.soton.ac.uk/ ) Needs["NMR`SpinDynamics`"]; SetSpinSystem[{{"A", 1/2}, {"B", 1/2}, {"M", 1/2}}]; Az = opI["A", "z"]; Ax = opI["A", "x"]; Ay = opI["A", "y"]; Bz = opI["B", "z"]; Bx = opI["B", "x"]; By = opI["B", "y"]; Mz = opI["M", "z"]; Mx = opI["M", "x"]; My = opI["M", "y"]; A = opI["A"]; B = opI["B"]; M = opI["M"]; Bz.Mz; H J coupling *) 4.67); (* H 13 C 2J coupling *) (* H 13 C coupling *) opI["A"].opI["B"]; spectrum = Re@FT@Signal1D[ {0, 1, 1/1024}, EvolutionBackground > H0, InitialDensityOperator Preparation > {{None, t}, EnsembleAverage > {t Range[0, 10 10^ 2, 1 10^ 2]}]; SimulationFolder = "D: \ \ "; SetDirectory[SimulationFolder]; Expor t["C1_3spin_pH2.xlsx", spectrum "XLSX"]; ListPlot[spectrum, Joined > True, Frame > True, PlotRange > All]

PAGE 155

155 LIST OF REFERENCES (1) Overhauser, A. W. Physical Review 1953 92 411. (2) Carver, T. R.; Slichter, C. P. Physical Review 1953 92 212. (3) Tycko, R.; Reimer, J. A. The Journal of Physical Chemistry 1996 100 13240. (4) Goodson, B. M. J. Magn. Reson. 2002 155 157. (5) Brunner, E.; Haake, M.; Kaiser, L.; Pines, A.; Reimer, J. A. J. Magn. Reson. 1999 138 155. (6) Ruppert, K.; Mata, J. F.; Brookeman, J. R.; Hagspiel, K. D.; Mugler, J. P. Magnetic Resonance in Medicine 2004 51 676. (7) Mair, R. W.; Hrovat, M. I.; Patz, S.; Rosen, M. S.; Ruset, I. C.; Topulos, G. P.; Tsai, L. L.; Butler, J. P.; Hersman, F. W.; Walsw orth, R. L. Magnetic Resonance in Medicine 2005 53 745. (8) Pavlovskaya, G. E.; Cleveland, Z. I.; Stupic, K. F.; Basaraba, R. J.; Meersmann, T. Proc. Natl. Acad. Sci. U. S. A. 2005 102 18275. (9) Happer, W. Reviews of Modern Physics 1972 44 169. (10) Hartmann, S. R.; Hahn, E. L. Physical Review 1962 128 2042. (11) Bowers, C. R.; Weitekamp, D. P. Phys. Rev. Lett. 1986 57 2645. (12) Bowers, C. R.; Weitekamp, D. P. J. Am. Chem. Soc. 1987 109 5541. (13) Seidler, P. F.; Bryndza, H. E.; Frommer, J. E.; Stuhl, L. S.; Bergman, R. G. Organometallics 1983 2 1701. (14) Bowers, C. R. In Encyclopedia of Magnetic Resonance ; John Wiley & Sons, Ltd: 2002. (15) Green, R. A.; Adams, R. W.; Duckett, S. B.; Mewis, R. E .; Williamson, D. C.; Green, G. G. R. Prog. Nucl. Magn. Reson. Spectrosc. 2012 67 1. (16) Duckett, S. B.; Mewis, R. E. Acc. Chem. Res. 2012 45 1247. (17) Duckett, S. B.; Sleigh, C. J. Prog. Nucl. Magn. Reson. Spectrosc. 1999 34 71. (18) Natterer, J.; Bargon, J. Prog. Nucl. Magn. Reson. Spectrosc. 1997 31 293. (19) Gloggler, S.; Colell, J.; Appelt, S. J. Magn. Reson. 2013 235 130. (20) Duckett, S. B.; Mewis, R. E. Top. Curr. Chem. 2013 338 75.

PAGE 156

156 (21) Duckett, S. B.; Wood, N. J. Coord. Chem. Rev. 2008 252 2278. (22) Blazina, D.; Duckett, S. B.; Dunne, J. P.; Godard, C. Dalton Trans. 2004 2601. (23) Natterer, J.; Bargon, J. Prog. Nucl. Magn. Reson. Spectrosc. 1997 31 293. (24) Eisenberg, R. Acc. Chem. Res. 1991 24 110. (25) Koptyug I. V.; Kovtunov, K. V.; Burt, S. R.; Anwar, M. S.; Hilty, C.; Han, S. I.; Pines, A.; Sagdeev, R. Z. J. Am. Chem. Soc. 2007 129 5580. (26) Kovtunov, K. V.; Zhivonitko, V. V.; Skovpin, I. V.; Barskiy, D. A.; Salnikov, O. G.; Koptyug, I. V. J. Phys. Chem. C 2013 117 22887. (27) Pravica, M. G.; Weitekamp, D. P. Chem. Phys. Lett. 1988 145 255. (28) Atkinson, K. D.; Cowley, M. J.; Elliott, P. I. P.; Duckett, S. B.; Green, G. G. R.; Lopez Serrano, J.; Whitwood, A. C. J. Am. Chem. Soc. 2009 131 13362. (29) Adams, R. W.; Duckett, S. B.; Green, R. A.; Williamson, D. C.; Green, G. G. R. J. Chem. Phys. 2009 131 (30) Adams, R. W.; Aguilar, J. A.; Atkinson, K. D.; Cowley, M. J.; Elliott, P. I. P.; Duckett, S. B.; Green, G. G. R.; Khazal, I. G.; Lopez Serran o, J.; Williamson, D. C. Science 2009 323 (31) Bhattacharya, P.; Chekmenev, E. Y.; Perman, W. H.; Harris, K. C.; Lin, A. P.; Norton, V. A.; Tan, C. T.; Ross, B. D.; Weitekamp, D. P. J. Magn. Reson. 2007 186 (32) Chekmenev, E. Y.; Hoevener, J.; Norton, V. A.; Harris, K.; Batchelder, L. S.; Bhattacharya, P.; Ross, B. D.; Weitekamp, D. P. J. Am. Chem. Soc. 2008 130 (33) Bouchard, L. S.; Kovtunov, K. V.; Burt, S. R.; Anwar, M. S.; Koptyug, I. V.; Sagdeev, R. Z.; Pines, A. Angewandte Chemie International E dition 2007 46 4064. (34) Bouchard, L. S.; Burt, S. R.; Anwar, M. S.; Kovtunov, K. V.; Koptyug, I. V.; Pines, A. Science 2008 319 442. (35) Conner, W. C.; Falconer, J. L. Chem. Rev. (Washington, DC, U. S.) 1995 95 759. (36) Cremer, P. S.; Su, X.; She n, Y. R.; Somorjai, G. A. The Journal of Physical Chemistry 1996 100 16302. (37) Zhivonitko, V. V.; Kovtunov, K. V.; Beck, I. E.; Ayupov, A. B.; Bukhtiyarov, V. I.; Koptyug, I. V. J. Phys. Chem. C 2011 115 13386. (38) of, U. t. l. p.; heterogenized cat alystsSkovpin, I. V.; Zhivonitko, V. V.; Koptyug, I. V. Appl. Magn. Reson. 2011 41 393.

PAGE 157

157 (39) Gutmann, T.; Ratajczyk, T.; Xu, Y.; Breitzke, H.; Grnberg, A.; Dillenberger, S.; Bommerich, U.; Trantzschel, T.; Bernarding, J.; Buntkowsky, G. Solid State Nucl Magn. Reson. 2010 38 90. (40) Skovpin, I. V.; Zhivonitko, V. V.; Kaptein, R.; Koptyug, I. V. Appl. Magn. Reson. 2013 44 289. (41) Kovtunov, K. V.; Zhivonitko, V. V.; Corma, A.; Koptyug, I. V. Journal of Physical Chemistry Letters 2010 1 1705. (42) Kovtunov, K. V.; Beck, I. E.; Bukhtiyarov, V. I.; Koptyug, I. V. Angew Chem Int Edit 2008 47 1492. (43) Bond, G. C. Applied Catalysis a General 1997 149 3. (44) Kovtunov, K. V.; Zhivonitko, V. V.; Kiwi Minsker, L.; Koptyug, I. V. Chem. Commun. (Cambrid ge, U. K.) 2010 46 5764. (45) Mogorosi, R. P.; Fischer, N.; Claeys, M.; van Steen, E. J. Catal. 2012 289 140. (46) Liu, X. Y.; Liu, M. H.; Luo, Y. C.; Mou, C. Y.; Lin, S. D.; Cheng, H. K.; Chen, J. M.; Lee, J. F.; Lin, T. S. J. Am. Chem. Soc. 2012 134 10251. (47) Linsmeier, C.; Taglauer, E. Applied Catalysis a General 2011 391 175. (48) Kovtunov, K. V.; Barskiy, D. A.; Salnikov, O. G.; Khudorozhkov, A. K.; Bukhtiyarov, V. I.; Prosvirin, I. P.; Koptyug, I. V. Chem. Commun. (Cambridge, U. K.) 2014 50 875. (49) Zhivonitko, V. V.; Telkki, V. V.; Chernichenko, K.; Repo, T.; Leskel, M.; Sumerin, V.; Koptyug, I. V. J. Am. Chem. Soc. 2013 (50) Hattori, H. Chem. Rev. (Washington, DC, U. S.) 1995 95 537. (51) Boudart, M. Chem. Rev. (Washington, DC, U. S.) 1995 95 661. (52) Kang, H. C.; Weinberg, W. H. Chem. Rev. (Washington, DC, U. S.) 1995 95 667. (53) Horiuti, I.; Polanyi, M. Transactions of the Faraday Society 1934 30 1164. (54) Cremer, P. S.; Su, X.; Shen, Y. R.; Somorjai, G. A. J. Am. Chem. S oc. 1996 118 2942. (55) Mattson, B.; Foster, W.; Greimann, J.; Hoette, T.; Le, N.; Mirich, A.; Wankum, S.; Cabri, A.; Reichenbacher, C.; Schwanke, E. J. Chem. Educ. 2013 90 613. (56) Ceyrolles, W. J.; Viot, P.; Talbot, J. Langmuir 2002 18 1112.

PAGE 158

158 (57) Kunz, L.; Maier, L.; Tischer, S.; Deutschmann, O. In Modeling and Simulation of Heterogeneous Catalytic Reactions ; Wiley VCH Verlag GmbH & Co. KGaA: 2011, p 113. (58) Chorkendorff, I.; Niemantsverdriet, J. W. Concepts of Modern Catalysis and Kinetics ; Wile y VCH, 2007. (59) Taylor, H. S. Proceedings of the Royal Society of London. Series A 1925 108 105. (60) Vdrine, J. C. Applied Catalysis A: General (61) Boudart, M. In Adv. Catal. ; D.D. Eley, H. P., Paul, B. W., Eds.; Academic Press: 1969; Vol. Volume 2 0, p 153. (62) Somorjai, G. A.; Carrazza, J. Industrial & Engineering Chemistry Fundamentals 1986 25 63. (63) Somorjai, G. A. Catal. Lett. 1990 7 169. (64) Taghavi, M. B.; Pajonk, G. M.; Teichner, S. J. J. Colloid Interface Sci. 1979 71 451. (65) Rioux, R. M.; Song, H.; Yang, P.; Somorjai, G. A. In Metal Nanoclusters in Catalysis and Materials Science ; Corain, B., Schmid, G., Toshima, N., Eds.; Elsevier: Amsterdam, 2008, p 149. (66) Somorjai, G. A.; Marsh, A. L. Philosophical Transactions of the Ro yal Society A: Mathematical, Physical and Engineering Sciences 2005 363 879. (67) Thomas, J.; Saghi, Z.; Gai, P. Top. Catal. 2011 54 588. (68) Munnemann, K.; Kolzer, M.; Blakey, I.; Whittaker, A. K.; Thurecht, K. J. Chem. Commun. (Cambridge, U. K.) 2012 48 1583. (69) Kadlecek, S.; Vahdat, V.; Nakayama, T.; Ng, D.; Emami, K.; Rizi, R. NMR Biomed. 2011 24 933. (70) Bhattacharya, P.; Chekmenev, E. Y.; Reynolds, W. F.; Wagner, S.; Zacharias, N.; Chan, H. R.; Bunger, R.; Ross, B. D. NMR Biomed. 2011 24 1023. (71) Koptyug, I. V.; Kovtunov, K. V.; Burt, S. R.; Anwar, M. S.; Hilty, C.; Han, S. I.; Pines, A.; Sagdeev, R. Z. J. Am. Chem. Soc. 2007 129 (72) Bouchard, L. S.; Kovtunov, K. V.; Burt, S. R.; Anwar, M. S.; Koptyug, I. V.; Sagdeev, R. Z.; Pines A. Angewandte Chemie International Edition 2007 46 4064. (73) Telkki, V. V.; Zhivonitko, V. V.; Ahola, S.; Kovtunov, K. V.; Jokisaari, J.; Koptyug, I. V. Angewandte Chemie International Edition 2010 49 8363.

PAGE 159

159 (74) Kovtunov, K. V.; Beck, I. E.; Zhivoni tko, V. V.; Barskiy, D. A.; Bukhtiyarov, V. I.; Koptyug, I. V. Phys. Chem. Chem. Phys. 2012 14 11008. (75) Salnikov, O. G.; Kovtunov, K. V.; Barskiy, D. A.; Bukhtiyarov, V. I.; Kaptein, R.; Koptyug, I. V. Appl. Magn. Reson. 2013 44 279. (76) Barskiy, D. A.; Kovtunov, K. V.; Primo, A.; Corma, A.; Kaptein, R.; Koptyug, I. V. Chemcatchem 2012 4 2031. (77) Blasco, T. Chem. Soc. Rev. 2010 39 4685. (78) Zhang, W.; Xu, S.; Han, X.; Bao, X. Chem. Soc. Rev. 2012 41 192. (79) Schwiedernoch, R.; Ti scher, S.; Deutschmann, O.; Warnatz, J. Proceedings of the Combustion Institute 2002 29 1005. (80) Jarenwattananon, N. N.; Gloggler, S.; Otto, T.; Melkonian, A.; Morris, W.; Burt, S. R.; Yaghi, O. M.; Bouchard, L. S. Nature 2013 502 537. (81) Koptyug, I. V.; Khomichev, A. V.; Lysova, A. A.; Sagdeev, R. Z. J. Am. Chem. Soc. 2008 130 10452. (82) Neal, L. M.; Hagelin Weaver, H. E. Journal of Molecular Catalysis a Chemical 2008 284 141. (83) Neal, L. M.; Hernandez, D.; Hagelin Weaver, H. E. Journal of M olecular Catalysis a Chemical 2009 307 29. (84) Corro, G.; Gomez, R. React. Kinet. Catal. Lett. 1978 9 325. (85) Tauster, S. J.; Fung, S. C.; Baker, R. T. K.; Horsley, J. A. Science 1981 211 1121. (86) Kovtunov, K.; Zhivonitko, V.; Skovpin, I.; Barsk iy, D.; Koptyug, I. In Hyperpolarization Methods in NMR Spectroscopy ; Kuhn, L. T., Ed.; Springer Berlin Heidelberg: 2013; Vol. 338, p 123. (87) Arnold, H.; Dbert, F.; Gaube, J. In Handbook of Heterogeneous Catalysis ; Wiley VCH Verlag GmbH & Co. KGaA: 2008. (88) Bos, A. N. R.; Westerterp, K. R. Chemical Engineering and Processing: Process Intensification 1993 32 1. (89) Kovtunov, K. V.; Zhivonitko, V. V.; Skovpin, I. V.; Barskiy, D. A.; Koptyug, I. V. Top. Curr. Chem. 2013 338 123. (90) Molnr, .; Srkny, A.; Varga, M. J. Mol. Catal. A: Chem. 2001 173 185.

PAGE 160

160 (91) Peck, J. W.; Mahon, D. I.; Beck, D. E.; Bansenaur, B.; Koel, B. E. Surf. Sci. 1998 410 214. (92) Delbecq, F.; Sautet, P. J. Catal. 2002 211 398. (93) Boutain, M.; Duckett, S. B.; Dunne J. P.; Godard, C.; Hernandez, J. M.; Holmes, A. J.; Khazal, I. G.; Lopez Serrano, J. Dalton Trans. 2010 39 3495. (94) Osborn, J. A.; Jardine, F. H.; Young, J. F.; Wilkinson, G. Journal of the Chemical Society A: Inorganic, Physical, Theoretical 1966 1711. (95) Skovpin, I.; Zhivonitko, V.; Koptyug, I. Appl. Magn. Reson. 2011 41 393. (96) Kovtunov, K. V.; Koptyug, I. V. In Magnetic Resonance Microscopy ; Wiley VCH Verlag GmbH & Co. KGaA: 2009, p 99. (97) Kennedy, D. R.; Webb, G.; Jackson, S. D.; Lennon D. Applied Catalysis a General 2004 259 109. (98) Langmuir 1997 13 883. (99) Teschner, D.; Vass, E.; Hvecker, M.; Zafeiratos, S.; Schnrch, P.; Sauer, H.; Knop Gericke, A.; Schlgl, R.; Chamam, M.; Wootsch, A.; Canning, A. S.; Gamman, J. J.; Jackson, S. D.; McGregor, J.; Gladden, L. F. J. Catal. 2006 242 26. (100) Kennedy, D. R. ; Webb, G.; Jackson, S. D.; Lennon, D. Applied Catalysis A: General 2004 259 109. (101) Ruta, M.; Laurenczy, G.; Dyson, P. J.; Kiwi Minsker, L. The Journal of Physical Chemistry C 2008 112 17814. (102) Bond, G. C.; Thompson, D. T. Catalysis Reviews 199 9 41 319. (103) Sermon, P. A.; Bond, G. C.; Wells, P. B. Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases 1979 75 385. (104) Haruta, M. Catal. Today 1997 36 153. (105) Jia, J.; Haraki, K.; Kondo, J. N.; Domen, K.; Tamaru, K. The Journal of Physical Chemistry B 2000 104 11153. (106) Verendel, J. J.; Pmies, O.; Diguez, M.; Andersson, P. G. Chem. Rev. (Washington, DC, U. S.) 2013 (107) Church, T. L.; Andersson, P. G. Coord. Chem. Rev. 2008 252 513. (1 08) Kllstrm, K.; Munslow, I.; Andersson, P. G. Chemistry A European Journal 2006 12 3194.

PAGE 161

161 (109) Pfaltz, A.; Blankenstein, J.; Hilgraf, R.; Hrmann, E.; McIntyre, S.; Menges, F.; Schnleber, M.; Smidt, S. P.; Wstenberg, B.; Zimmermann, N. Adv. Synth. Catal. 2003 345 33. (110) Mann, R. S.; Lien, T. R. J. Catal. 1969 15 1. (111) Ortiz Soto, L.; Monnier, J.; Amiridis, M. Catal. Lett. 2006 107 13. (112) Klasovsky, F.; Claus, P. In Metal Nanoclusters in Catalysis and Materials Science ; Corain, B., Schmid, G., Toshima, N., Eds.; Elsevier: Amsterdam, 2008, p 167. (113) Ahmadi, T. S.; Wang, Z. L.; Henglein, A.; El Sayed, M. A. Chem. Mater. 1996 8 1161. (114) Yoo, J. W.; Lee, S. M.; Kim, H. T.; El Sayed, M. A. Bull. Korean Chem. Soc. 2004 25 843. (1 15) Cocco, G.; Campostrini, R.; Cabras, M. A.; Carturan, G. J. Mol. Catal. 1994 94 299. (116) Oteroschipper, P. H.; Wachter, W. A.; Butt, J. B.; Burwell, R. L.; Cohen, J. B. J. Catal. 1977 50 494. (117) Yoo, J. W.; Hathcock, D.; El Sayed, M. A. J. Phys Chem. A 2002 106 2049. (118) Yoo, J. W.; Hathcock, D. J.; El Sayed, M. A. J. Catal. 2003 214 1. (119) Levitt, M. H. In Encyclopedia of Magnetic Resonance ; John Wiley & Sons, Ltd: 2007. (120) Solomon, I. Physical Review 1955 99 559. (121) Bloembergen, N.; Purcell, E. M.; Pound, R. V. Physical Review 1948 73 679. (122) Aime, S.; Gobetto, R.; Reineri, F.; Canet, D. J. Magn. Reson. 2006 178 (123) Aime, S.; Gobetto, R.; Reineri, F.; Canet, D. J. Chem. Phys. 2003 119 8890. (124) Eisenschmid, T. C.; McDonald, J.; Eisenberg, R.; Lawler, R. G. J. Am. Chem. Soc. 1989 111 (125) Barkemeyer, J.; Haake, M.; Bargon, J. J. Am. Chem. Soc. 1995 117 2927. (126) Natterer, J.; Schedletzky, O.; Barkemeyer, J.; Bargon, J.; Glaser, S. J. J. Mag n. Reson. 1998 133 92. (127) Stephan, M.; Kohlmann, O.; Niessen, H. G.; Eichhorn, A.; Bargon, J. Magn. Reson. Chem. 2002 40 157.

PAGE 162

162 (128) Canet, D.; Bouguet Bonnet, S.; Aroulanda, C.; Reineri, F. J. Am. Chem. Soc. 2007 129 (129) Ivanov, K. L.; Miesel, K .; Yurkovskaya, A. V.; Korchak, S. E.; Kiryutin, A. S.; Vieth, H. M. Appl. Magn. Reson. 2006 30 513. (130) Ivanov, K. L.; Yurkovskaya, A. V.; Vieth, H. M. J. Chem. Phys. 2008 128 (131) Korchak, S. E.; Ivanov, K. L.; Yurkovskaya, A. V.; Vieth, H. M. Phy s. Chem. Chem. Phys. 2009 11 11146. (132) Levitt, M. H. Annual Review of Physical Chemistry, Vol 63 2012 63 89. (133) Pileio, G.; Carravetta, M.; Hughes, E.; Levitt, M. H. J. Am. Chem. Soc. 2008 130 12582. (134) Sarkar, R.; Vasos, P. R.; Bodenhausen, G. J. Am. Chem. Soc. 2006 129 328. (135) Bornet, A.; Ahuja, P.; Sarkar, R.; Fernandes, L.; Hadji, S.; Lee, S. Y.; Haririnia, A.; Fushman, D.; Bodenhausen, G.; Vasos, P. R. ChemPhysChem 2011 12 2729. (136) Cavadini S.; Dittmer, J.; Antonijevic, S.; Bodenhausen, G. J. Am. Chem. Soc. 2005 127 15744. (137) Sarkar, R.; Ahuja, P.; Vasos, P. R.; Bodenhausen, G. ChemPhysChem 2008 9 2414. (138) Ahuja, P.; Sarkar, R.; Vasos, P. R.; Bodenhausen, G. J. Am. Chem. Soc. 2009 131 7498. (139) Cavadini, S.; Vasos, P. R. Concepts in Magnetic Resonance Part A 2008 32A 68. (140) Jonischkeit, T.; Bommerich, U.; Stadler, J.; Woelk, K.; Niessen, H. G.; Bargon, J. The Journal of Chemical Physics 2006 124 (141) Aroulanda, C.; Star ovoytova, L.; Canet, D. J. Phys. Chem. A 2007 111 10615. (142) Vinogradov, E.; Grant, A. K. J. Magn. Reson. 2008 194 46. (143) Chekmenev, E. Y.; Hovener, J.; Norton, V. A.; Harris, K.; Batchelder, L. S.; Bhattacharya, P.; Ross, B. D.; Weitekamp, D. P. J. Am. Chem. Soc. 2008 130 4212. (144) Warren, W. S.; Jenista, E.; Branca, R. T.; Chen, X. Science 2009 323 1711. (145) Vasos, P. R.; Comment, A.; Sarkar, R.; Ahuja, P.; Jannin, S.; Ansermet, J. P.; Konter, J. A.; Hautle, P.; van den Brandt, B.; Bodenhausen, G. Proc. Natl. Acad. Sci. U. S. A. 2009 106 18469.

PAGE 163

163 (146) Adams, R. W.; Aguilar, J. A.; Atkinson, K. D.; Cowley, M. J.; Elliott, P. I. P.; Duckett, S. B.; Green, G. G. R.; Khazal, I. G.; Lopez Serrano, J.; Williamson, D. C. Science 2009 323 1708. (147) Bornet, A.; Jannin, S.; Bodenhausen, G. Chem. Phys. Lett. 2011 512 151. (148) Ahuja, P.; Sarkar, R.; Jannin, S.; Vasos, P. R.; Bodenhausen, G. Chem. Commun. (Cambridge, U. K.) 2010 46 8192. (149) Zhivonitko, V. V.; Kovtunov, K. V.; Chapovsky, P. L.; Koptyug, I. V. Angewandte Chemie International Edition 2013 52 13251. (150) Li, Y.; Shen, W. Chem. Soc. Rev. 2014 (151) Xiong, Y. J.; Cai, H. G.; Wiley, B. J.; Wang, J. G.; Kim, M. J.; Xia, Y. N. J. Am. Chem. Soc. 2007 129 3665. (152) Uzun, A.; Dixon, D. A.; Gates, B. C. ChemCatChem 2011 3 95. (153) Mori, K.; Yamashita, H. Phys. Chem. Chem. Phys. 2010 12 14420. (154) Mori, K.; Yamashita, H. J. Jpn. Pet. Inst. 2011 54 1. (155) Bonnemann, H.; Nagabhushana, K. S. In Metal Nanoclusters in Catalysis and Materials Science ; Corain, B., Schmid, G., Toshima, N., Eds.; Elsevier: Amsterdam, 2008, p 21. (156) Burns, S.; Jackson, S. D.; Vass, E. M. Kinetics and Catalysis 2011 52 723. (157) Lee, I.; Zaera, F. J. Catal. 2010 269 359. (158) Millstone, J. E.; Hurst, S. J.; Metraux, G. S.; Cutler, J. I.; Mirkin, C. A. Small 2009 5 6 46. (159) Niu, W. X.; Xu, G. B. Nano Today 2011 6 265. (160) Opelt, S.; Turk, S.; Dietzsch, E.; Henschel, A.; Kaskel, S.; Klemm, E. Catal. Commun. 2008 9 1286. (161) Ellena, S.; Viale, A.; Gobetto, R.; Aime, S. Magn. Reson. Chem. 2012 50 529. (162) M eilikhov, M.; Yusenko, K.; Esken, D.; Turner, S.; Van Tendeloo, G.; Fischer, R. A. Eur. J. Inorg. Chem. 2010 3701. (163) Corma, A.; Garca, H.; Llabrs i Xamena, F. X. Chem. Rev. (Washington, DC, U. S.) 2010 110 4606. (164) Yoon, M.; Srirambalaji, R.; K im, K. Chem. Rev. (Washington, DC, U. S.) 2011 112 1196.

PAGE 164

164 (165) Zhu, Y.; Qian, H. F.; Drake, B. A.; Jin, R. C. Angew Chem Int Edit 2010 49 1295. (166) Zhu, Y.; Qian, H. F.; Jin, R. C. Chemistry a European Journal 2010 16 11455.

PAGE 165

165 BIOGRAPHICAL SKETCH Ronghui Zhou was born in Linhai City of Zhejiang Province, China in 1980. He earned his Bachelor of Science in the Department of Chemistry at Zhejiang University in Hangzhou, China Science and Technology, and got his Master of Science in 2006 under the guidance of Prof. Sung Kee Chung. His focus was on bioorganic and medicinal chemistry, working on dendrimeric molecular tr ansporters for targeted cancer drug delivery. He then joined visiting scholar, working on solid phase peptide synthesis and multi resistant cancer drug delivery. After s erving as a synthetic/process organic chemist in the pharmaceutical industry in Shanghai and Taizhou, China he came to the University of Florida in 2010 to pursue a Ph.D. in physical chemistry. He spent the first year learning preparation and physical pro perties of nanomaterials under the guidance of Prof. W ei David Wei and then moved on to work with Prof. Clifford R. Bowers on the Parahydrogen Induced Polarization (PHIP) project and got his Ph.D. in the spring of 2014 He is interested in spectroscopy, so lid state NMR, hyperpolarization, catalysis, and other energy related domains. He would like to advance the science in academia, or turn h is knowledge and skills, accumulated over the years, to productivities in chemical, pharmaceutical or energy related i ndustries.