Homogeneous hydrogenation of aqueous hydrogen carbonate to formate under exceedingly mild conditionsmdash;a novel possibility of carbon dioxide activationdagger;

摘要

Homogeneous hydrogenation of aqueous hydrogen carbonate to formate under exceedingly mild conditionsmdash;a novel possibility of carbon dioxide activationdagger; Ferenc Jooacute;,*ab Gaacute;bor Laurenczy*c, Levente Naacute;dasdibc and Jaacute;nos Elekb a Institute of Physical Chemistry and b Research Group of Homogeneous Catalysis of the Hungarian Academy of Sciences, L. Kossuth University, H-4010 Debrecen, PO Box 7, Hungary. E-mail: jooferenc@tigris.klte.hu c Universitegrave; de Lausanne, Institut de Chimie Minegrave;rale et Analytique, CH-1015 Lausanne-Dorigny, Switzerland Received (in Cambridge, UK) 24th March 1999, Accepted 26th April 1999 Water soluble ruthenium(ii)ndash; and rhodium(i)ndash;phosphine complexes catalyze the hydrogenation of aqueous HCO32 to HCO22 under mild conditions with turnover frequencies up to 262 TO h21.Reduction of carbon dioxide into useful starting materials of organic synthesis is an aim of paramount importance both for economic and environmental reasons.1 It is intriguing that while in the dark reactions of photosynthesis carbon dioxide reacts in an aqueous medium under very mild conditions, efficient synthetic systems usually require elevated temperatures and pressures to achieve reasonable reaction rates.In homogeneous systems, both in the liquid and in the supercritical state, several ruthenium complexes, such as RuCl2(PMe3)4 1 and RuCl2(dppe)2 2 showed impressive catalytic activity in the reaction of CO2, H2 and HNMe2 yielding dimethyl formamide. 2,3 Leitner et al.4 investigated the hydrogenation of CO2 in aqueous solutions with RhCl(tppts)3 3 as catalyst tppts = tris(3-sulfonatophenyl)phosphine which gave formate with initial reaction rates of 7260 h21 at 81 deg;C and 1365 h21 at 23 deg;C in the presence of HNMe2 under 40 atm total pressure (CO2/H2 = 1). It is of interest that under identical conditions RuCl2(tppts)3 4 showed4 a TOF of only 6 h21 (23 deg;C), and no formic acid was detected in aqueous solutions with 3 without an amine. Water soluble phosphine complexes of platinum metals were introduced into the homogeneous catalysis scene by Jooacute; and Beck5 who also noted the interaction of CO2 with HRuCl(tppms)3 5 (tppms = 3-sulfonatophenyldiphenylphosphine) in aqueous solution.6 Recently it was established that formation of the water soluble hydrides of ruthenium(ii), such as 5 H2Ru(tppms)4 6 and HRuCl(tppms)22 7 from RuCl2(tppms)22, 8, tppms and H2, and their distribution in aqueous solution is strongly pH-dependent which leads to opposite selectivities in acidic and basic solutions during the biphasic hydrogenation of unsaturated aldehydes.7 This prompted us to reinvestigate the hydrogenation of carbon dioxide and its derivatives in aqueous solutions catalyzed by water soluble platinum group metals phosphine complexes as a function of pH.In addition to RuCl2(tppms)22, RhCl(tppms)3 9 HRu(ac)(tppms)3 10, trans-IrCl(CO)(tppms)2 11, PdCl2(tppms)2 12, and complexes of 1,3,7-triaza-7-phosphaadamantane (pta), such as RuCl2(pta)4 13 and RhCl(pta)3 14 were also studied.8 Here we report that 8ndash;10 and 13 catalyze the hydrogenation of CO2 in aqueous solutions under mild conditions without the need of an amine additive.When solutions of 8, 9 and 13 were agitated under a CO2ndash;H2 atmosphere (up to 80 atm total pressure) at 24 deg;C, small amounts of formic acid were detected by 1H/13C NMR with TOFs not exceeding 1.5 h21 (Table 1).However, when the solvent was changed for 0.2ndash;1.0 M NaHCO3 a substantial increase in the catalytic activity was observed and TOFs reached 262 h21 with 9 at 24 deg;C. Further increase of the pH decreased the rate of hydrogenation (entry 7). These data represent a significant improvement over previous results in that the turnover frequencies for Ru-based catalysts are among the highest obtained so far in aqueous solutions, moreover, the reductions do not require addition of an organic co-solvent or an amine or alcohol to stabilize the product.In contrast, no formate production took place with 11, 12 and 14 under the conditions of entry 13. Instead, 12 yielded a dark brown precipitate while the originally yellow solutions of 11 and 14 became colorless; these reactions were not investigated in further detail. We note, however, that Pruchnik et al.9 observed catalysis of the reverse water gas shift reaction in aqueous solution by Rh2(ac)4- (H2O)2 + pta (Rh : pta = 1 : 3); no sign of CO formation in our system with 14 was detected.Most of the experiments were done on the hydrogenation of aqueous NaHCO3 solutions; the results are summarized in Table 1. No carbon containing products other than formate were dagger; Dedicated to Prof. Mihaacute;ly T. Beck on the occasion of his 70th birthday. Table 1 Reduction of CO2 in aqueous systems at various pH catalyzed by water soluble Ru- and Rh-phosphine complexesDagger; Entry Catalyst P/Ma Base/solvent p(CO2)/atm p(H2)/atm T/deg;C t/h TON TOF/h21 average 1 8 5 H2O 20 60 24 14.5 21.6 1.49 2 9 6 H2O 20 60 24 14.5 1.6 0.11 3 13 4 H2O 20 60 24 13.5 3.2 0.24 4 8 4 1 M NaHCO3 mdash; 60 54 209 289 1.4 (47)b 5 13 4 1 M NaHCO3 mdash; 60 54 456 358 0.78 (4.5)b 6 13 4 1 M NaHCO3 mdash; 60 103 22.5 358 16 (49)b 7 13 4 0.2M NaHCO3+0.8 M Na2CO3 mdash; 60 54 286 75 0.26 8 10 6 1 M NaHCO3 mdash; 10 50 18 284 16 9 8 5 200 mg CaCO3/2 ml H2Oc 20 60 24 14 372d 26.6 10 9 6 200 mg CaCO3/2 ml H2Oc 20 60 24 14 262 18.7 11 13 4 200 mg CaCO3/2 ml H2Oc 20 60 24 14 35 2.5 12 8 2 0.2 M NaHCO3 mdash; 10 50 4 0.04 0.01 13 8 4 0.2 M NaHCO3 mdash; 10 50 4 60 15 14 8 6 0.2 M NaHCO3 mdash; 10 50 4 60 15 15 8 4 0.2 M NaHCO3 + 0.2 M KI mdash; 10 50 4 55 14 16 9 7 1 M NaHCO3 5 35 24 2 524 262 a Total phosphorus to metal ratio; M = 2.0ndash;2.5 mM.b Initial turnover frequencies in parentheses. c Suspensions. d 0.93 M HCO2 ndash; final solution. Chem. Commun., 1999, 971ndash;972 971detected. With any of 8ndash;10 and 13 the reactions started with no appreciable induction period (Fig. 1). Increase of the temperature from 54 to 103 deg;C caused an 11-fold increase of the initial rate of formate production with 13 (entries 5, 6). Despite the similarity10 of pta to PMe3, 13 proved considerably less active than 8. However, the anionic ligand of the catalyst (precursor) had no significant influence: either 8 in the presence of NaI or 10 gave approximately the same rate as the unmodified 8 (entries 8, 13, 15). Both with 8 and 9 some excess of tppms proved beneficial for activity and catalyst stabilitymdash;in fact 8 was almost completely inactive without added tppms; however, its catalytic activity was no further increased when the total tppms/Ru ratio exceeded 4 (entries 12ndash;14).As expected, increase of the hydrogen pressure increased the rate of reduction with all catalysts. Conversely, while CO2 in the gas phase was detrimental with 8 and 13, it was essential for high reaction rates with 9 (although a much slower reaction still proceeded in its absence).The substrate bicarbonate can be produced in situ in the reaction of carbonates and CO2. When a 2 ml aqueous suspension of 200 mg CaCO3 was pressurized with CO2 (20 atm) and H2 (60 atm) and shaken in an NMR tube overnight at 24 deg;C with 8, calcium formate was produced with TOF = 26.6 h21, the total TON reaching 372. Similar experiments with 9 and 13 yielded formate with TOF 18.7 and 2.5 h21, respectively (see also Table 1).The presence of CaCO3 in these systems is highly beneficial, since under the same conditions CO2 alone is reduced with TOF 1.49 (8), 0.11 (9), and 0.24 h21 (13); the rate increase with 9 is more than 160-fold! It is premature to speculate on detailed mechanisms of the hydrogenation of bicarbonate in these systems. It is known7 that in aqueous KCl solutions, at the pH (8.3) corresponding to that of 1 M of NaHCO3 solutions used in the present experiments, 8 + tppms + H2 rapidly yields 6 in which the total P/Ru ratio is 4, found optimal for reductions here (entry 13).However, pHstatic titrations (similar to those described in ref. 7) of 8 under H2 in NaHCO3 solutions containing an excess of tppms showed formation of only 0.4ndash;0.6 mol H+ for 1 mol Ru during the early phase of the reaction (instead of 2 mol required for 6). This suggests a monohydride, possibly HRu(HCO3)(tppms)4 as a more likely intermediate. Both of these catalysts would account for the observed independence of the rate from the starting composition of the catalyst precursor.Assuming a constant catalyst composition in the pH range from 8.3 (1 M NaHCO3) to 10.8 (0.2 M NaHCO3 + 0.8 M Na2CO3) the lower reactivity of CO3 2ndash; (entry 7) is consistent with its known inferior tendency of oxygen exchange with H2O compared to HCO3 ndash;.11 So far homogeneous catalytic hydrogenation of HCO3 ndash; in aqueous solutions has not been reported in the literature. Stalder et al.12 used various heterogeneous Pd catalysts with a maximum TOF of 35 h21 (Pd/C, 25 deg;C).Kudo et al.13 studied in detail the reduction of CO2 (40 atm) catalyzed by PdCl2 in aqueous KOH at forcing conditions (106 atm H2, 240 deg;C). In our opinion, based on the CO3 22/HCO32/H2O equilibrium, all similar reactions using aqueous solutions or mixture of amines or carbonates under CO2 pressure, in fact, may have utilized HCO32 as the real substrate in the catalytic cycle. This could also explain the beneficial effect of a small amount of H2O on CO2 reductions often observed4b,14 in organic solvent systems. This work was supported by the Hungarian National Research Foundation (OTKA T023997 and F 023159) and by the Office Feacute;deacute;ral de lrsquo;Education et de la Science, Suisse (OFES C98.0011 to G.L.).L. N. is grateful for an OFES fellowship, 1998/99. The loan of RuCl3.xH2O by Johnson- Matthey PLC is gratefully acknowledged. This research is part of the collaboration within the COST Action D10 /0001 Working Group.Notes and references Dagger; Non-SI units: p/atm = 101.325 kPa; T/K = T/deg;C + 273.15. Ligands tppms, pta and their Rh, Ru and Ir complexes were prepared as described in refs 8ndash;10. Hydrogenation experiments were carried out in well stirred, heavy walled glass tubes with 10 atm H2 or in high pressure sapphire NMR tubes with 60 atm H2 (pressures at room temperature) intensively shaken (300 min21) on top of a laboratory shaker. Formate concentration was determined by HPLC or the reaction mixture was analyzed in situ by 1H and/ or 13C NMR spectroscopy (Bruker AC 200, AM 360, DRX 400) using 13Cenriched (99) NaHCO3. 1 Most recent reviews: D. Walther, M. Ruben and S. Rau, Coord. Chem. Rev., 1999, 182, 67; X. Yin and J. R. Moss, Coord. Chem. Rev., 1999, 181, 27. 2 P. G. Jessop, T. Ikariya and R. Noyori, Nature, 1994, 368, 231. 3 O. Krouml;cher, R. A Kouml;ppel and A. Baiker, Chem. Commun., 1997, 453. 4 (a) W. Leitner, Angew. Chem., 1995, 107, 2391; Angew.Chem., Int. Ed. Engl., 1995, 34, 2207; (b) W. Leitner, E. Dinjus and F. Gassner, in Aqueous-Phase Organometallic Catalysis, ed. B. Cornils and W. A. Herrmann, Wiley-VCH, Weinheim, 1998, p. 486. 5 F. Jooacute; and M. T. Beck, Magy. Keacute;m. Foly., 1973, 79, 189; F. Jooacute;, Proc. 15th ICCC (Moscow, USSR, 1973), p. 556. 6 F. Jooacute; and M. T. Beck, React. Kinet. Catal. Lett., 1975, 2, 257. 7 F. Jooacute;, J. Kovaacute;cs, A. Cs. Beacute;nyei and Aacute;. Kathoacute;, Catal. Today, 1998, 42, 441. 8 (8, 10): Z. Toacute;th, F. Jooacute; and M. T. Beck, Inorg. Chim. Acta., 1980, 42, 153; (9, 11): F. Jooacute;, J. Kovaacute;cs, Aacute;. Kathoacute;, A. C. Beacute;nyei, T. Decuir and D. J. Darensbourg, Inorg. Synth., 1998, 32, 1; (12): K.-C. Tin, N.-B. Wang, R.-X. Li, Y.-Z. Li and X.-J. Li, J. Mol. Catal. A: Chem., 1998, 137, 113; (13): D. J. Darensbourg, F. Jooacute;, M. Kannisto and Aacute;. Kathoacute;, Organometallics, 1992, 11, 1990; (14): F. Jooacute;, L. Naacute;dasdi, A. Cs. Beacute;nyei and D. J. Darensbourg, J. Organomet. Chem., 1996, 512, 45. 9 F. P. Pruchnik, P. Smolenski and I. Raksa, Pol. J. Chem., 1995, 69, 5. 10 J. R. Delerno, L. M. Trefonas, M. Y. Darensbourg and R. J. Majeste, Inorg. Chem., 1976, 15, 816. 11 H. Gamsjauml;ger and R. K. Murmann, in Adv. Inorg. Bioinorg. Mech., 1983, 2, 343. 12 C. J. Stalder, S. Chao, D. P. Summers and M. S. Wrighton, J. Am. Chem. Soc., 1983, 105, 6318. 13 K. Kudo, N. Sugita and Y. Takezaki, Nippon Kagaku Kaishi, 1997, 302; Chem. Abstr., 1977, 86, 178068s. 14 Y. Inoue, H. Izumida, Y. Sasaki and H. Hashimoto, Chem. Lett., 1976, 863. Communication 9/02368B Fig. 1 Time course of formate production in reaction of HCO32 with H2 catalyzed by RuCl2(pta)4 (a), (b), (d) and by 1/2RuCl2(tppms)22 + 2 tppms (c) at 50 (c), 54 (a), 90 (b) and 103 deg;C (d) followed by 1H (a), ndash;xndash; or 13C NMR (all other cases). 972 Chem. Commun., 1999, 971ndash;972