This item is only available as the following downloads:
University of Florida | Journal of Undergraduate Research | Volume 15, Issue 3 | Summer 2014 1 Parahydrogen Induced Polarization Studies Using a Continuous-Flow Homogeneous Hydrogenation Reactor Daniel Schulman, Ronghui Zhou, and Dr. Russ Bowers College of Liberal Arts and Sciences, University of Florida Parahydrogen induced polarization (PHIP) is a technique for enhancing NMR signals while providing the opportunity to study hydrogenation catalysis by tr ansition metal compounds. This research focu sed on testing a continuous-flow homogeneous hydrogenation reactor and PHIP enhanced prod ucts. Signals with up to 67 times enhanc ement have been observed with Wilkinsons catalyst ([Rh(PPh3)3Cl]). After testing a variety of substrates and catalysts, many challenges associated with quantifying sig nal enhancements and potential improvements have been identified. INTRODUCTION Nuclear magnetic resonance spectroscopy (NMR) is a valuable tool for obtaining structural and dynamical information of molecules. However, the method suffers from poor sensitivity compared to most other types of spectroscopy. In recent years, s ubstantial efforts have been made to improve the sensitivity of the technique. Hyperpolarization methods such as parahydrogen induced polarization (PHIP), dynamic nuclear polarization (DNP), and spin exchange optical pumping can increase NMR signals by several orders of magnitude. NMR relies on the Zeeman effect to polarize spins in the presence of a magnetic field. The numbers of spins at each energy level, N and N follows the Boltzmann distribution, whereE is the nuclear spin Zeeman energy. The amplitude of the NMR signal is proportional to N -N Hyperpolarization methods increase the population difference, N -N effectively lowering the spin temperature down to the mK regime. (1) Parahydrogen induced polariza tion was discovered in the late 1980s[1-2]. Since its discovery, it has been used to study hydrogenation catalysis by transition metal complexes [3-8]. Recently it has shown promise in in-vivo 13C magnetic resonance imaging . The aim of the present work is to study PHIP enhanced homogeneous hydrogenation prod ucts using a continuousflow reactor. Because the reaction is carried out in high field, PHIP results in PASADENA (Parahydrogen and Synthesis Allows Dramatically Enhanced Nuclear Alignment) type anti-phase signals. This differs from reactions carried out in low field which produce ALTADENA (Adiabatic Longitudinal Transport After Dissociation Engenders Net Ali gnment) type signals . A variety of small substrates were selected for study using both Wilkinsons [Rh(PPh3)3Cl] and Crabtrees [Ir(cod)(PCy3)(Pyr)PF6] catalysts. In particular, aldehyde substrates were chosen for further studies. Previous ALTEDENA experiments using Wilkinsons catalyst and acrolein showed polarization transfer to the hydrogen proton on the adjacent aldehy de group. These experiments, which utilized the shaking of a pressured NMR tube followed by quick insertion into the magnet, yielded highly varied results. To address th is problem, we constructed an in-situ continuous-flow reactor. We demonstrated that higher hyperpolarization repr oducibility and quantification can be obtained with this experimental setup. Principals of Parahydrogen Induced Polarization Parahydrogen. The PHIP effect relies on the slow interconversion of the two spin isomers of molecular hydrogen. Hydrogen exists in both the triplet orthohydrogen (o-H2) and singlet parahydrogen (p-H2) spin states (eq1) . (2) At temperatures much greater than the characteristic temperature of rotation, T >> r ( r=85.4K), the ortho : para ratio approaches 3 : 1. At T < r with the use of a paramagnetic catalyst such as activated charcoal, the mixture can become p-H2 enriched. At liquid nitrogen temperatures (77K), the p-H2 mole fraction reaches 51%. The equilibrium ratio for a given temperature can be derived from the rigid rotor partition function as shown by equation (3) .
DANIEL SCHULMAN, RONGHUI ZHOU, PROF. RUSS BOWERS University of Florida | Journal of Undergraduate Research | Volume 15, Issue 3 | Summer 2014 2 Transitions between these states are magnetic dipole forbidden allowing for extended storage of the p-H2 enriched gas. This forbidden transition causes p-H2 to be NMR silent. Hence the p-H2 concentration can be measured using the signal intensity which is directly proportional to the o-H2 concentration. PASADENA Effect. Parahydrogen induced polarization utilizes the ability of the singlet state nuclear spin order of parahydrogen to be converted in to observable nuclear spin magnetization by the addition of the p-H2 molecule to a unsaturated substrate. The effect, when carried out in high field, was given the acronym PASADENA (Parahydrogen and Synthesis Allows Dramatically Enhanced Nuclear Alignment). Since, the effect has been observed with a variety of homogeneous catal ysts[3-8], heterogeneous catalysts[12-14] and colloidal catalysts. Certain requirements must be met in order to observe the PHIP effect . Hydrogen atoms must be transferred pairwise to the substrate and maintain their spin pairing. The atoms must also occupy coupled, magnetically inequivalent sites on the product. In order to observe the hyperpolarization, the reaction must proceed faster than the spin-lattice relaxation time constant (T1). The maximum PASADENA signal amplitude is observed using a 45 pulse. Figure 2a shows the correlation diagram for the nuclear spin states resulting from both p-H2 and o-H2 addition. Figure 2b shows the energy level transitions and resulting 1H NMR spectra of a PASADENA enhanced system. The theoretical enhancement factor is defined as the ratio of the hyperpolarized and thermal equilibrium signal amplitudes of the 12 transition. This depends on the p-H2 enrichment, magnetic field strength, temperature, and pulse angle. Under ideal conditions, PASASENA signal intensities resulting from 100% p-H2 approach those of a fully polarized nuclear spin system. Catalyst Mechanisms. Since the pairwise addition of the p-H2 is crucial for the PHIP effect, the catalytic mechanism is central. Wilkinso ns catalyst, RhCl(PPh3)3, was the first to exhibit the PHIP effect in the hydrogenation of acrylonitrile. Figure 3 shows the widely accepted hydride route mechanism for W ilkinsons catalyst in which the key intermediate is structure II . First, the phosphine ligand dissociates allowing hydrogen coordination followed by the olefin addition. The pathway can be simplified to the following two step process. I II (4) P Despite Wilkinsons catalys ts high efficiency, the iridium analog is entirely inactive. Hydrogen irreversibly binds producing a stable [ IrClH2(PPh3)3] adduct state. Cationic iridium catalysts such as Ir(cod)(PCy3)(Pyr)PF6, Crabtrees catalyst, have su rpassed Wilkinsons catalyst with rates over 100 times greater and they possess the ability to hydrogenate tri-subs tituted alkenes. The key to Crabtrees reactivity is the use of high polarity, noncoordinating solvents such as CH2Cl2. Cationic rhodium complexes have also have been shown to act as efficient catalysts. However, they behave much differently than cationic iridium complexes . METHODS Parahydrogen production A parahydrogen converter was constructed from a coiled copper tube packed with activat ed charcoal. Prior to use, the converter was reactivated and bound contaminant gases were removed. The reactor was purged with dry N2 for approximately 24 hours, followed by evacuation for 12 hours by rough pumping. The final hour of pumping was completed using a high vacuum turbo pump. At this point, the reactor was submerged in a dewar of liquid nitrogen and allowed to cool for 2 hours. A H2 flow rate of approximately 1ml/min was used to allow adequate reaction time between the H2 gas and the catalyst. The first 30 minutes of p-H2 product was discarded. Over the course of 4 hours, 3 bar of p-H2 was produced and stored in a 5L aluminum container. Over the course of 2 weeks, the p-H2 was periodically checked using control reactions to ensure there was no significant para to ortho H2 backconversion. Sample Preparation Each experiment used 1.5mg of catalyst which was Figure 1 : Percent of p-H2 as a function of T .
PARAHYDROGEN INDUCED POLARIZATION USING A CONTINUOUS-FLOW HOMOGENEOUS HYDROGENATION REACTOR University of Florida | Journal of Undergraduate Research | Volume 15, Issue 3 | Summer 2014 3 weighed and transferred into a 10mm NMR tube. A volume of 1.5ml of solvent a nd 0.1ml of substrate were added to the tube. The tube was immediately capped and vigorously shaken to promoting dissolving of the catalyst. Once dissolved, the capillary tube cap was screwed on. To remove oxygen from the tube, the tube was purged with nitrogen for approximately 15s at a rate of 20ml/ minute. Experimental Procedures The sample tube containing the reaction mixture (catalyst, substrate and solvent) was inserted into the Bruker Avance 400MHz spectrome ter. The spectrometer was tuned, shimmed and a reference spectrum was acquired. All spectra were acquired with a single scan and a pulse angle of 45. There was also a delay of at least 3 minutes between acquisitions to ensure full spin-lattice relaxation. PHIP was initiated by closing the nitrogen valve and opening the p-H2 valve. To react the substrate, the p-H2 enriched hydrogen was bubbled at a rate of 40ml/ minute for 60 seconds. This flow rate was selected as it allowed adequate bubbling but was not too vigorous to cause significant solvent evaporation. A flow time of 60 seconds was selected because it allowed the residual nitrogen in the line to be displaced by the p-H2 and allowed maximum enhancements to be observed in all solvents, catalysts, and substrates. After bubbling, a 5s delay was allowed prior to collection of the free induction decay. A fully-relaxed thermal equilibrium spectrum was acquired after approximately 5 minutes. In some experiments, the polarization had not fully rela xed after 5 minutes, and in Figure 4 Lower: PHIP enhanced spectrum of acrolein to propanal after 60s of 50% p-H2 bubbling. 1.5ml C6D6 0.1ml acrolein, 1.5mg Wilkinsons catalyst. Upper: t hermal equilibrium spectrum after hydrogenation. Ha Hb Figure 3. Hydride mechanism of hydrogenation for Wilkinsons catalyst .
DANIEL SCHULMAN, RONGHUI ZHOU, PROF. RUSS BOWERS University of Florida | Journal of Undergraduate Research | Volume 15, Issue 3 | Summer 2014 4 this case, another spectrum was acquired after 5 additional minutes. Only the first hyperpolarized spectrum was used when calculating enhancement values. Subsequent experiments on the same sample were used to verify results and to ensure that longer bubbling times did not affect hyperpolarization levels. After each experiment, the capillary tube was rinsed in acetone and the line was purged with N2 to flush the p-H2 mixture. This was done to avoid any premature hydrogenation of the subseque nt samples. The NMR tube was then rinsed 3 times with acetone and dried in an oven. Data Analysis Enhancements of each product peak were calculated by integrating both the hyper polarized and relaxed product signals. The enhancement factor is defined as the ratio of the hyperpolarized signal intensity to the relaxed signal intensity. A reactant peak was chosen to be the integral reference for both the hyperpolarized and relaxed spectra. Due to the PASADENA peak pattern, the hyperpolarized intensity was taken as the sum of the absolute integrals of the absorption and emission peaks. RESULTS Wilkinsons catalyst yielded the highest enhancements under the given conditions. As shown in table 1, the largest observed enhancement was 67. This compound was highly effective in catalyzing hydrogenation reactions of unsaturated aldehydes with low steric hinderances, benzylic alkenes and terminal alkynes to alkanes. Crabtrees catalyst was remarkably efficient at hydrogenating styrene and 1,5-hexadiene; however it did not produce PHIP enhancement. Some reactions occurred at a very slow rate, thus not yielding a visible product. In those cases, a hyperpolarized signal was visible but an enha ncement factor could not be calculated without the accurate integration of the relaxed peaks. Wilkinsons Catalyst The hydrogenation rates of substrates of Wilkinsons catalyst are consistent with published observations. It is assumed that product yield, as shown in table 1, is directly rated to the reaction rate. Although substrates with electron withdrawing substituents are known to be highly reactive , steric hindrances can significantly reduce reactivities . This explains the slower hydrogenation of cinnamaldehyde and lack of reactivity for 2methylpentenal compared to arolein. Steric hinderence also explains the slower reactivity of 2-hexyne compared to 1hexyne. In the literature it is reported that styrene hydrogenation is approximately 12 times faster than 1,5hexadiene, confirming 1,5-hexadienes slow reaction rate. Due to peak overlap, the percent conversion for 1hexyne could not be accura tely determined. In the Table 1: Enhancement Factors and Product Yield % for Various Substrates, Catalyst and Solvent Combinations. Substrate Catalyst Solvent* HA Enhancement factor HB Enhancement factor Product Yield % Acrolein Wilkinson B 67 24 1.2 Acrolein Crabtree A 25 15 0.60 Styrene Wilkinson B 18 10 0.22 Styrene Crabtree B Large Product, no polarization 34 Cinnamaldehyde Wilkinson B 6.5 2.3 0.55 Cinnamaldehyde Crabtree A No reaction 2 methylpentenal Crabtree A No reaction 1,5 hexadiene Wilkinson B Small enhancement** 1,5 hexadiene Crabtree A Large product, no polarization 12 1 hexyne Wilkinson B Small alkene enhancement, large alkane product 1 hexyne Crabtree A Significant enhancement** 2 hexyne Wilkinson B Small enhancement**, no alkane product 2 hexyne Crabtree A 13*** 2.4 *B: Benzene, A: Acetone ***Cant differentiate HA and HB due to similar chemical shift ** Not enough product to calculate enhancement
PARAHYDROGEN INDUCED POLARIZATION USING A CONTINUOUS-FLOW HOMOGENEOUS HYDROGENATION REACTOR University of Florida | Journal of Undergraduate Research | Volume 15, Issue 3 | Summer 2014 5 literature it is stated that terminal alkynes should react approximately 20% slower than terminal alkenes . Crabtrees Catalyst The reactivity of Crabtrees catalyst was much lower than expected based on literature reports. This, however, could be attributed to improper solvent choice. Crabtrees catalyst requires a polar non coordinating solvent such as CH2Cl2. Figure 5 shows the hydrogenation catalytic cycle where [IrH2(cod)L2]+ is a key precursor. As shown by reaction (4), acetone can form a semi stable [IrH2(acetone)2L2]+ complex which prevents the formation of the required [IrH2(cod)L2]+ complex. (5) Despite the inappropriate r eaction conditions, the rate of hydrogenation for styrene and 1,5-hexadiene still surpassed that of Wilkinsons catalyst. Futures studies should be conducted under more favor able reaction conditions allowing for a hydrogenation rate fast enough to yield a measurable product for all substrates. While both catalysts hydrogenate d most substrates fairly slowly, Crabtrees catalyst was able to hydrogenate both styrene and 1,5-hexadiene extremely quickly. Crabtrees catalyst hydrogenated styrene with a 34% yield while Wilkinsons catalyst only achieved a 1.2% yield. These reactions also resulted in no PHIP enhancement. It would appear that Crabtrees catalyst is not maintaining the spin pairing during the addition step. It is not currently know whether this is due to a different mechanism or an effect of this rapid reaction rate. PHIP Enhancement Quantifying PHIP signal enhancements using the described method has many co mplications. The aims of these experiments were, first, to see large reproducible PHIP signals, and, second, to quantitatively calculate the enhancement factors. Howe ver, these two goals have differing requirements. In order to observe the largest PHIP signals, the reaction must proceed at a high rate therefore producing large amounts of product. In order to achieve this, the solution must be saturated with p-H2. This saturation is not instantly achieved. It is assumed that saturation is achieved when maximum PHIP signals are observed. With this assumption, hydrogen saturation occurs after approximately 15-30 seconds of bubbling. In contrast, accurate quantitative measurements require short reaction times and therefore short bubbling times. In this case, the reaction time is significantly longer than the hyperpolarization lifetime, which is determined by the spin lattice relaxation time (T1). Consequently, the observed hyperpolarized signal represents products formed during within a few T1s before de tection. This results in artificially low enhancement factors. Another issue that compounds this error is the continued reaction after the first data acquisition. After the hyperpolarized spectrum is acquired, the mixture continues to react. Therefore, the relaxed product measured is much larger than the amount of product at the time of the first acquisition. This also deflates the enhancement. In order to study the correlation between the enhancement factor and the re action rate, the enhancement factor was plotted against the percent yield in figure 6. It is assumed that the percent yield is directly related to the reaction rate. Since each enhanced signal is due only to the single added hydrogen while the relaxed signal is due to all the equivalent hydrogens, enha ncement factors will vary based on the number of equivalent hydrogens on the substrate. To normalize the data, the enhancement factor was multiplied by the number of equivalent hydrogens. For a CH3 group, the enhancement was multiplied by a factor of 3. The enhancement factor for both the Ha and Hb were then averaged. Acrolein exhibits a linear trend in the plot of % yield vs. normalized enhancement factor. The normalized enhancement factor proporti onally increases with percent yiel. This could be explai ned by the continued reaction after the hyperpolarized spectrum is acquired. After the bubbling is discontinued, ther e is a finite amount of hydrogen gas that can still react For a fast reaction, this amount is negligible compared to the amount that had already reacted. Therefore, for fast reactions, this continued reaction is less pronounced. For slower Figure 5. Proposed catalytic cycle for hydrogenation using Crabtrees catalyst .
DANIEL SCHULMAN, RONGHUI ZHOU, PROF. RUSS BOWERS University of Florida | Journal of Undergraduate Research | Volume 15, Issue 3 | Summer 2014 6 reactions, this hydrogen can be significant compared to small amount that reacted during the bubbling time. The continued reaction can significantly increase this product. The result is a drastic decrease in the calculated enhancement factor. Kinetics studies conducted using normal hydrogen and determination of T1 using inve rsion recovery can be used to correct the enhancement factor. By plotting product formation as a function of bubbling time, the amount of product formed during a single T1 time can be calculated. It is expected that after an initial mixing time, this relationship should be linear. Since the hyperpolarized signal essentially only results from the product formed during this T1 time, the corrected product concentration should be substituted in for the thermal equilibrium concentration in the enhancement factor equation. CONCLUSION Significant PHIP enhancements were observed using the continuous flow reactor with both Wilkinsons catalyst and Crabtrees Catalyst. The calculated enhancement factors, however, are artificially low due to a reaction time which is much longer than T1, and the continued reaction after the hyperpolarized spectrum acquisition. For future work, kinetics studies with normal hydrogen and determination of T1 can correct this for error. It is expected that these corrected enhancement factors will be similar to other published results. The absence of a PHIP signal in by styrene and 1,5hexadiene using Crabtrees catalyst needs to be further studied. It is unclear what is responsible for the loss of spin pairing. Compared to other reactions, these reactions produced significantly higher yields. By improving reaction conditions of Crabtrees catalyst, future reactions should also proceed at significantly higher rates. These experiments will provide insight as to whether the loss of spin pairing is related to the reaction rate or due to the specific substrate. ACKNOWLEDGMENTS This project was supported by the University Scholars Program at the University of Florida and the American Chemical Society Petroleum Research Fund, Project #52258-ND5. I would like to thank my advisor, Prof. Russ Bowers, and Ronghui Zhou for their guidance and support. REFERENCES  Bowers, C. R., & Weitekamp, D. P. (1986). Transformation of Symmetrization Order to Nucl ear-Spin Magnetization by Chemical Reaction and Nuclear Magnetic Resonance. Physical Review Letters 57 (21), 2645-2648. doi: 10.1103/PhysRevLett.57.2645  Bowers, C. R., & Weitekamp, D. P. (1987). Parahydrogen and synthesis allow dramatically enhanced nuclear alignment. Journal of the American Chemical Society 109 (18), 5541-5542. doi: 10.1021/ja00252a049  Giernoth, R., Huebler, P., & Ba rgon, J. (1998). Intermediate Product-Catalyst Complexes in the Homogeneous Hydrogenation of Styrene Derivatives with Parahydrogen and Cationic RhI Catalysts. Angewandte Chemie International Edition 37 (18), 2473-2475. doi: 10.1002/(SICI)15213773(19981002)37:183.0.CO;2-J  Harthun, A., Barkemeyer, J., Selke, R., & Bargon, J. (1996). ChemInform Abstract: The Use of in situ NMR Spectroscopy and Parahydrogen as Tools for Chiral Synthesis. ChemInform 27 (4), No-No. doi: 10.1002/chin.199604059  Aime, S., Gobetto, R., Reineri, F., & Canet, D. (2003). Hyperpolarization transfer from parahydrogen to deuterium via carbon-13. The Journal of Chemical Physics 119 (17). doi: 10.1063/1.1615522  Korchak, S. E., Ivanov, K. L., Yurkovskaya, A. V., & Vieth, H. (2009). Para-hydrogen induced polarization in multi-spin systems studied at variable magnetic field. Physical Chemistry Chemical Physics doi: 10.1039/b914188j  Matthes, J., Pery, T., Grndemann, S., Buntkowsky, G., SaboEtienne, S., Chaudret, B., & Limbach, H. (2004). Bridging the Gap between Homogeneous and Heterogeneous Catalysis: Ortho/Para HCon version, Hydrogen Isotope Scrambling, and Hydrogenation of Olefins by Ir(CO)Cl(PPh). Journal of the American Chemical Society 126 (27), 8366-8367. doi: 10.1021/ja0475961  Adams, R. W., Aguilar, J. A., Atkinson, K. D., Cowley, M. J., Elliott, P. P., Duckett, S. B., ... Williamson, D. C. (2009). Reversible Interactions with para-Hydrogen Enhance NMR Sensitivity by Polarization Transfer. Science 323 (5922), 17081711. doi: 10.1126/science.1168877  Gutmann, T., Ratajczyk, T., Dill enberger, S., Xu, Y., Grnberg, A., Breitzke, H., ... Buntkowsky, G. (2011). New investigations of technical rhodium and iridium cat alysts in homogeneous phase employing para-hydrogen induced polarization. Solid State Nuclear Magnetic Resonance 40 (2), 88-90. doi: 10.1016/j.ssnmr.2011.08.002  Goldman, M., Johannesson, H., Axelsson, O., & Karlsson, M. (2005). Hyperpolarization of C through order transfer from parahydrogen: A new contrast agent for MRI. Magnetic 0 20 40 60 80 100 120 012Enhancement factor (normalized)% Yield% Conversion vs. Normalized Enhancement Acrolein Styrene Cinnamaldehyde Figure 6. Plot of the normalized enhancement factor vs. the percent yield of the reaction.
PARAHYDROGEN INDUCED POLARIZATION USING A CONTINUOUS-FLOW HOMOGENEOUS HYDROGENATION REACTOR University of Florida | Journal of Undergraduate Research | Volume 15, Issue 3 | Summer 2014 7 Resonance Imaging 23 (2), 153-157. doi: 10. 1016/j.mri.2004.11.031  Bowers, C. R. (2002). Sensitivity Enhancement Utilizing Parahydrogen. In D. M. Gr ant & R. K. Harris (Eds.), Encyclopedia of nuclear magnetic resonance. (Vol. 9, Advances in NMR, pp. 750-770). New York: Wiley.  Skovpin, I. V., Zhivonitko, V. V., Kaptein, R., & Koptyug, I. V. (2013). Generating Parahydrogen-Induced Polarization Using Immobilized Iridium Complexes in the Gas-Phase Hydrogenation of CarbonCarbon Double and Triple Bonds. Applied Magnetic Resonance 44 (1-2), 289-300. doi: 10.1007/s00723-012-0419-5  Salnikov, O. G., Kovtunov, K. V ., Barskiy, D. A., Bukhtiyarov, V., Kaptein, R., & Koptyug, I. V. (2013). Kinetic Study of Propylene Hydrogenation over Pt/Al2O3 by ParahydrogenInduced Polarization. Applied Magnetic Resonance 44 (1-2), 279288. doi: 10.1007/s00723-012-0400-3  Balu, A. M., Duckett, S. B., & Luque, R. (2009). Para-hydrogen induced polarisation effects in liquid phase hydrogenations catalysed by supporte d metal nanoparticles. Dalton Transactions (26), 5074. doi: 10.1039/b906449d  Eichhorn, A., Koch, A., & Bargon, J. (2001). In situ PHIP NMR a new tool to investigate hydr ogenation mediated by colloidal catalysts. Journal of Molecular Catalysis 174 (1-2), 293295.  Duckett, S. B., & Sleigh, C. J. (1999). Applications of the parahydrogen phenomenon: A chemical perspective. Progress in Nuclear Magnetic Resonance Spectroscopy 34 (1), 71-92.  Osborn, J. A., Jardine, F. H., Young, J. F., & Wilkinson, G. (1966). The preparation and properties of tris(triphenylphosphine)halogenorhodium(I) and some reactions thereof including catalytic homogen eous hydrogenation of olefins and acetylenes and their derivatives. Journal of the Chemical Society A: Inorganic, Physical, Theoretical 1711. doi: 10.1039/j19660001711  Crabtree, R. (1979). Iridi um compounds in catalysis. Accounts of Chemical Research 12 (9), 331-337. doi: 10.1021/ar50141a005  Jardine, F. H., Osbron, J. A., & Wilkinson, G. (1967). Further studies on the homogeneous hydr ogenation of olefins using tris(triphenylphosphine)h alogenorhodium(I) catalysts. Journal of the Chemical Society A: I norganic, Physical, Theoretical 1574. doi: 10.1039/j19670001574  Candlin, J. P., & Oldham, A. R. (1968). Selective hydrogenation of unsaturated carbon-carbon bonds with rhodium complexes. Discussions of the Faraday Society 46 60. doi: 10.1039/df9684600060  Crabtree, R. H., Demou, P. C., Eden, D., Mihelcic, J. M., Parnell, C. A., Quirk, J. M., & Morris, G. E. (1982). Dihydrido olefin and solvento complexes of iridium and the mechanisms of olefin hydrogenation and alkane dehydrogenation. Journal of the American Chemical Society 104 (25), 6994-7001. doi: 10.1021/ja00389a018