Formation of metallacarboxylic acids through Hieber base reaction. A density functional theory study
- 228 Downloads
Using density functional theory (B97-D/ECP2/PCM//RI-BP86/ECP1 level), we have studied the effects of ligand variation on OH− uptake by transition-metal carbonyls (Hieber base reaction), i.e., LnM(CO) + OH− → [LnM(CO2H)]−, M = Fe, Ru, Os, L = CO, PMe3, PF3, py, bipy, Cl, H. The viability of this step depends notably on the nature of the co-ligands, and a large span of driving forces is predicted, ranging from ΔG = −144 kJ/mol to +122 kJ/mol. Based on evaluation of atomic charges from natural population analysis, it is the ability of the co-ligands to delocalize the additional negative charge (through their π-acidity) that is the key factor affecting the driving force for OH− uptake. Implications for the design of new catalysts for water gas shift reaction are discussed.
KeywordsHomogeneous catalysis Water gas shift reaction Hieber base reaction Density functional theory
Development of homogeneous transition metal catalysts for H2 production from methanol [11, 12] has attracted much attention in the past few decades. Morton and Cole-Hamilton developed a ruthenium catalyst [Ru(H)2(X2)(PPh3)3] (X = N, H) for partial dehydrogenation of alcohols (including methanol) with notable turnover frequencies . Aldehydes and ketones were the main products [formaldehyde in case of methanol, Scheme 1, step (i)]. During the conversion of ethanol, significant amounts of methane and a carbonyl complex, [RuH2(CO)(PPh3)3], were produced through decarbonylation [Scheme 1, step (ii)]. However, since no CO2 was noticed, apparently these Ru complexes are not active as WGSR catalysts .
Significant research has been undertaken to reveal the mechanism of WGSRs catalyzed by homogeneous transition metal complexes, particularly the metal carbonyls of Fe, Ru, and Os [18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]. Recently, Guo et al. have studied the WGSR mechanism catalyzed by hexacarbonyl complexes of Mo and W . In all of these reactions, which are conducted under basic conditions, OH− is the nucleophile and its uptake to form a transient metallacarboxylic acid is considered as the initial step (in the grey box in Scheme 2). Such attack of OH− on carbonyl ligands is well known as Hieber base reaction .
In all studies of metal carbonyl catalyzed WGSR, this OH− uptake appeared to be highly exothermic and essentially barrierless. In contrast, in our study of WGSR in the Morton and Cole-Hamilton system , this step is predicted to be highly endergonic. The driving force for formation of the metallacarboxylic acid depends notably on the co-ligands that are present, in particular on the number of CO ligands. While large negative enthalpies and free energies are computed for the OH− uptake of Ru(CO)5 [28, 29] ΔG = 127.7 kJ/mol and 81.6 kJ/mol are predicted for [RuH2(CO)(PPh3)3] and [RuH2(CO)2(PPh3)2], respectively. It therefore appears that suitable ligand design, by varying the steric or electronic properties of the ligands, could make the process of OH− uptake feasible. In this work, we now report DFT-computed driving forces for OH− uptake in a number of metal-carbonyl complexes of Ru, Fe, and Os. Along with CO, we have made the choice of trimethylphosphine (comparable to triphenylphosphine), trifluorophosphine (a strong π-acceptor ligand, comparable to CO) , pyridine, and bipyridine ligands. For a perfect catalytic system, the OH− entry into the cycle should be facile and should not produce a very low-lying intermediate on the reaction profile that would eventually deactivate the catalytic system. This work can lead to the rational design of better catalysts for WGSR and, eventually, towards the complete decomposition of alcohols by dehydrogenation, decarbonylation, and the finally WGSR, which could facilitate entry into a hydrogen-based economy.
Results and discussion
Ligand effects on the initial uptake of OH− to form the metallacarboxylic acid
Ru(CO)(PF3)4 follows the expected trend in terms of free energy for the OH− uptake and has the largest predicted affinity for OH− of all complexes studied here (−143.7 kJ/mol, Fig. 2). It should be noted that such a large driving force for OH− uptake does not necessarily make this complex a good target for a WGSR catalyst, because a correspondingly higher energy needs to be invested to close the cycle and re-form the initial catalyst.
The results for the analogous Os complexes are very similar to those for the Ru species just discussed, with individual driving forces for OH− uptake within typically 10 kJ/mol of each other (ca. 20 kJ/mol for the bipy complex, compare Tables S1 and S3 in the SI). On going from Ru to Fe congeners, the changes in this driving force become somewhat more variable (up to ca. 30 kJ/mol, compare Tables S1 and S2 in the SI), but overall the same trends are obtained irrespective of the group 8 metal.
The metallacarboxylic acid arising from Hieber base reaction of [Ru(CO)3Cl3]− has been implicated as a key reactive intermediate in a complex variety of reactions . Indeed, despite forming a dianion from two monoions, OH− uptake of [Ru(CO)3Cl3]− affording [Ru(CO)2(CO2H)Cl3]2− is highly exergonic, with a free energy of −119.0 kJ/mol. This large driving force is fully consistent with the fact that this complex is a reactive intermediate that can be formed through Hieber base reaction . Experimentally, [Ru(CO)2(CO2H)Cl3]2− appears to lose a chloride ion consistent with our calculations as at our level, as this process is computed to be slightly exergonic, by −5.9 kJ/mol.
Natural population analysis
What is the origin of the huge variation in driving forces for OH− uptake in these complexes? Hypothesizing that a key factor should be delocalization of the additional negative charge brought into the complex, we used natural population analysis (NPA)  to evaluate the extent of charge transfer from OH− upon attack on the carbonyl ligand. To this end, we simply calculated the natural charge on the OH− fragments in the ruthenacarboxylic acid products, assessing how it changes from the value in free OH−, where it is −1. A substantial reduction from this absolute value is found in the complexes, indicating that most of the charge is actually delocalized into the complex, but there is still a notable variation of this charge, between −0.32 and −0.19 (see Table S4 in the SI).
The PMe3 ligand has σ-donating abilities, which pushes electron density to the metal center, which increases the amount of backbonding interaction between the filled metal d-orbital and the empty π*- orbital of the carbon atom of CO. The overall affect makes it difficult for the OH− fragment to delocalize electron density over the metal complex. As we replace more CO ligands with PMe3 ligands, the natural charge at the OH− fragment decreases. In [Ru(CO)2(CO2H)(PMe3)2]−, presence of a CO ligand at the axial position trans to the –CO2H− fragment increases its distance from the metal center, making it less available for OH− fragment to accommodate the charge density. Here the trans influence dominates the electronic nature of the ligands and a small discrepancy in the natural charges of the OH− fragment occurs when we move from [Ru(CO)2(CO2H)(PMe3)2]− to [Ru(CO)(CO2H)(PMe3)3]−, similar is the case with py and bipy ligands. For the PF3 ligands, on replacing each with the CO ligands, the natural charge at the OH− fragment gradually increases, as expected (Table S4).
In summary, using an appropriate DFT level, we have computed the driving forces for formation of metallacarboxylic acids from group 8 carbonyl complexes through uptake of OH−. This reaction (Eq. 2), known as Hieber base reaction, is the first step of water–gas shift reaction (WGSR) that can be catalyzed by transition metal complexes under basic conditions. According to our findings, the driving force for this step is surprisingly sensitive to the nature of the co-ligands at the metal, and can range from ΔG = −144 kJ/mol to +122 kJ/mol [for R = F and Me, respectively, in Ru(CO)(PR3)4]. Far from being innocent spectator ligands, these co-ligands actively take part in OH− uptake through delocalization of the negative charge, as apparent in the computed atomic charges from natural population analysis. Fe and Ru pentacarbonyls are prototypical WGSR catalysts; it is remarkable how replacement of CO ligands with electron-rich phosphines (which are ubiquitous in modern transition metal chemistry) can impede the first step of this WGSR catalytic cycle. In that case, use of phosphines with electron-withdrawing substituents (where we have used PF3 as extreme example) or aromatic N-donor ligands can increase the driving force for Hieber base reaction. Compared to these ligand effects, the nature of the metal (Fe, Ru, or Os) or its oxidation state [e.g., Ru(0) vs. Ru(II)] seems to be of lesser importance for OH− uptake.
We are convinced that the tunability of the driving force for Hieber base reaction through appropriate choice of co-ligands can inform on the rational design of new WGSR catalysts. As this quantity, a simple reaction (free) energy, can be readily computed with modern DFT tools, large libraries of ligands can be screened computationally, opening up new avenues for applications of molecular modeling in homogeneous catalysis.
In order to identify the most stable isomers and conformers of each reactant and product, an exhaustive screening of the possible stereoisomers was undertaken. Only the results for the most stable forms are reported. The conformation of the carboxylic acid group was uniformly taken as that where the hydrogen of the OH− fragment is pointing towards the metal center. We investigated the stability of these metalla-acids against those in which the hydrogen of the OH− fragments points away from the metal center, particularly in Fe complexes, the former complexes are appeared to be more stable (see Table S5).
We thank EaStCHEM and the School of Chemistry for support. Computations were carried out on a local Opteron PC cluster maintained by Dr. H. Früchtl.
- 4.Zhang YHP (2011) Hydrogen production from carbohydrates: a mini-review. ACS Symp Ser 1067 (Sustainable Production of Fuels, Chemicals, and Fibers from Forest Biomass):203-216. https://doi.org/10.1021/bk-2011-1067.ch008
- 9.Veziroglu TN, Editor (1976) Conference Proceedings of the 1st World Hydrogen Energy conference, Vol. 1: March 1-3, 1976 Miami Beach, Fla. vol Copyright (C) 2018 American Chemical Society (ACS). All Rights Reserved. Univ. Miami Press,Google Scholar
- 10.Ishida Y, Kumabe K, Hata K, Tanifuji K, Hasegawa T, Kitagawa K, Isu N, Funahashi Y, Asai T (2009) Selective hydrogen generation from real biomass through hydrothermal reaction at relatively low temperatures. Biomass Bioenergy 33:8–13. https://doi.org/10.1016/j.biombioe.2008.04.004 CrossRefGoogle Scholar
- 13.Morton D, Cole-Hamilton DJ (1988) Molecular hydrogen complexes in catalysis: highly efficient hydrogen production from alcoholic substrates catalyzed by ruthenium complexes. J Chem Soc, Chem Commun:1154-1156. https://doi.org/10.1039/c39880001154
- 14.Morton D, Cole-Hamilton DJ, Utuk ID, Paneque-Sosa M, Lopez-Poveda M (1989) Hydrogen production from ethanol catalyzed by group 8 metal complexes. J Chem Soc, Dalton Trans, pp 489–495Google Scholar
- 16.Sieffert N, Reocreux R, Lorusso P, Cole-Hamilton DJ, Bühl M (2014) On the importance of decarbonylation as a side-reaction in the ruthenium-catalysed dehydrogenation of alcohols: a combined experimental and density functional study. Chem. Eur. J. 20:4141–4155. https://doi.org/10.1002/chem.201303722 CrossRefPubMedGoogle Scholar
- 17.Lorusso P, Ahmad S, Schmid K, Cole-Hamilton DJ, Sieffert N, Bühl M Manuscript in preparationGoogle Scholar
- 30.Schaper L-A, Herrmann WA, Kuehn FE Water-Gas Shift Reaction. In, 2018. Wiley-VCH Verlag GmbH & Co. KGaA, pp 1689-1698. https://doi.org/10.1002/9783527651733.ch38
- 32.Hill AF (2000) "simple" carbonyls of ruthenium: new avenues from the Hieber base reaction. Angew. Chem. Int. Ed. 39:130–133. https://doi.org/10.1002/(SICI)1521-3773(20000103)39:1<130::AID-ANIE130>3.0.CO;2-6 CrossRefGoogle Scholar
- 34.Ru(CO)3(PMe3)2,is a model for Ru(CO)3(PPh3)2,, also known as Roper’s catalyst: Cavit BE, Grundy KR, Roper WR (1972) Dicarbonyltris(triphenylphosphine)ruthenium and -osmium. Ethylene complex of ruthenium and a dioxygen complex of osmium. J Chem Soc, Chem Commun:60-61. https://doi.org/10.1039/c3972000060b
- 35.e.g. [Ru(phen)2(CO)(CO2H)]+ (phen = 1,10-phenanthroline): Gibson DH, Ding Y, Andino JG, Mashuta MS, Richardson JF (1998) Synthesis, Characterization, and Reactions of Ruthenium Phenanthroline Complexes Bearing C1 Ligands: Formyl, Metallocarboxylate, and CO2-Bridged Complexes. Organometallics 17:5178-5183. https://doi.org/10.1021/OM9806242
- 36.e.g. [Ru(bipy)2(CO)(CO2H)]+ (bipy = 2,2′=bipyridine): Toyohara K, Nagao H, Adachi T, Yoshida T, Tanaka K (1996) Crystal structure of [Ru(bpy)2(CO)(η1-C(O)OH)]+ (bpy = 2,2′-bipyridine) as a key intermediate in CO2/CO conversion. Chem Lett :27-28. https://doi.org/10.1246/cl.1996.27
- 37.In terms of reaction enthalpies (i.e. without the temperature-dependent entropy part) the driving forces are ΔH = 36.7 kJ/mol and 33.0 kJ/mol for PMe3 and PPh3, respectively. These values are thus rather similar for both phosphines, indicating little steric hindrance with the bulkier PPh3Google Scholar
- 41.Becke AD (1988) Density-functional exchange-energy approximation with correct asymptotic behavior. Phys Rev A: Gen Phys 38:3098–3100. https://doi.org/10.1103/PhysRevA.38.3098
- 42.Perdew JP (1986) Density-functional approximation for the correlation energy of the inhomogeneous electron gas. Phys. Rev. B Condens. Matter 33:8822–8824. https://doi.org/10.1103/PhysRevB.33.8822
- 45.Dong W, Yan M, Zhang M, Liu Z, Li Y (2005) A computational and experimental investigation of the interaction between the template molecule and the functional monomer used in the molecularly imprinted polymer. Anal. Chim. Acta 542:186–192. https://doi.org/10.1016/j.aca.2005.03.032 CrossRefGoogle Scholar
- 46.Frisch MJ et al (2009) Gaussian 09. Revision D.01 edn. Gaussian, Inc., PittsburghGoogle Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.