Modeling the effect of ligands and solvation on hydrolysis variants in the Pd(II)-Catalyzed hydroxycarbonylation of pentenoic acids
Graphical abstract
We studied computationally how the number of H2O molecules affects the rate determining hydrolysis step (concerted or stepwise) of the Pd(II)-catalyzed hydroxycarbonylation of pentenoic acid isomers.
Introduction
Transition-metal-mediated alkene transformations are well-known to play a key role in synthesizing industrially important products. Reactions involving the carbonylation of alkenes to generate acyl complexes are noteworthy examples [1,2]. The hydroxycarbonylation, which belongs to this important class of reactions, is one of the most common homogeneously catalyzed reactions in the chemical industry [[3], [4], [5], [6]]. Recently, Meurs and co-workers reported [7,8] the hydroxycarbonylation of pentenoic acids (PEAs) catalyzed by a Pd(II) catalyst bearing the bulky ligand L2 = 1,2-bis[(di-R)phosphinomethyl]benzene, DTBPX, with R = tert-butyl. This catalyst achieves a selectivity of over 95% toward the linear product, adipic acid (ADA), in comparison to the branched regio-isomers methyl glutaric acid (MGA), ethyl succinic acid (ESA), and propyl malonic acid (PMA) [[7], [8], [9], [10], [11], [12]], Scheme 1, Scheme 2.
For Pd(II) catalysts bearing less bulky ligands L2 of the same overall composition, but with the substituents R = methyl (L2 = DMPX) or phenyl (L2 = DPPX), the branched products PMA (75%) and ESA (90%) become more favorable [12]. The proposed mechanism [12] consists of three major steps: (i) substrate isomerization, (ii) CO insertion, and (iii) hydrolysis, Scheme 1. Hydrolysis, the rate determining step, was studied computationally using a (H2O)3 cluster for the concerted pathway, which is analogous to the approach for methanolysis, with three methanol molecules, (CH3OH)3 [11]. Furthermore, in that study [12] a stepwise pathway was proposed, comprising a hydration step followed by product release, Scheme 2; see also Scheme S1 in the Supplementary Data (SD) of the present work.
DFT modeling [12] showed that the active catalyst [L2PdII–H]+ isomerizes the PEAs to their equilibrium mixture of Pd(II)-alkyl intermediates, from which carbonylation leads to a stable Pd(II)-acyl complex, Scheme 1. Subsequent hydrolysis of the latter complex results in the formation of linear or branched products. For the crucial hydrolysis step, the calculated barriers decreased significantly when the number of participating H2O molecules increased from 1 to 3, (H2O)m, m = 1–3, with m = 3 being the optimum number of water molecules [12]. The free energy barriers for the hydrolysis step essentially do not change for n = 4 where the added H2O molecule acts as a spectator or a solvent [12]. Using a (H2O)3 cluster was determined to favor the stepwise mechanism over the analogous concerted variant of the hydrolysis of the Pd-acyl complexes [12]. For the ligand DTBPX, the upper free energy barrier of the stepwise hydrolysis is by 42 kJ mol−1 higher (109–67 kJ mol−1) than the barrier of the concerted hydrolysis step leading to ADA, Scheme 1. The hydrolysis step exhibited a preference for the product ADA for all ligands, at variance with the experimentally observed selectivity for the branched products PMA and ESA for less bulky DMPX, and DPPX ligands [12].
The goal of the present computational work is to evaluate the preferential hydrolysis mechanism, concerted vs stepwise with the number of water molecules (H2O)m, m = 1, 2, followed by a comparison with the recently reported hydrolysis using a (H2O)3 cluster [12]. For the calculated rate-limiting hydrolysis barriers (concerted or stepwise), we analyzed the contribution of various energy terms – corrections due to larger basis set, inclusion of dispersion interactions, and solvation – to understand how each of these terms is affecting the decision between branched vs linear products.
Section snippets
Computational details
All computational work, geometry optimizations and single-point calculations, were carried out with the software package Gaussian 09 [13], invoking the B3LYP functional [14,15]. For Pd we used the small-core, quasi-relativistic Stuttgart-Dresden effective core potential, with its associated valence basis set [16]. For C, H, and O atoms we employed the all-electron 6-31G(d,p) basis set [[17], [18], [19]] for optimizations, thermal corrections and solvent contributions. The reported Gibbs free
Hydrolysis of the Pd-acyl complex
General mechanistic details of the hydrolysis process of 1XY2, i.e., in the presence of one H2O molecule, are shown in Scheme 2, while Schemes S2 and S3 of the SD provide details of the corresponding process with two or three H2O molecules close to the metal center. Starting from 1XY2 the stepwise hydrolysis pathway consists of two steps, Scheme 2: (i) the hydration of the carbonyl moiety in the encounter complex 1XY2, leading to the formation of a diol intermediate 1XY3 via 1XY2-3, and (ii)
Summary
We examined how the number of water molecules, (H2O)m, m = 1–3, affects the rate-determining hydrolysis process of the Pd-catalyzed hydroxycarbonylation of pentenoic acids using DTBPX, DMPX, and DPPX as ligands. We explored the concerted and the stepwise mechanism variants, and studied the product selectivity of these systems towards the linear product, adipic acid (ADA) vs. the branched products methyl glutaric acid (MGA), ethyl succinic acid (ESA), and propyl malonic acid (PMA).
- –
Preferred
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
The authors thank Torstein Fjermestad for helpful suggestions. This work was supported by grant no.1527700033 of the A∗STAR Science and Engineering Research Council as well as by generous allotments of computational resources at the A∗STAR Computational Resource Center (ACRC) and the National Super Computing Center Singapore (NSCC).
References (39)
- et al.
J. Mol. Catal.
(1994) - et al.
Organometallics
(2017) - et al.
Organometallics
(2015) - et al.
Chem. Rev.
(2012) - et al.
ChemSusChem
(2013) - et al.
J. Org. Chem.
(1992) - et al.
Angew. Chem. Int. Ed.
(2000) - et al.
Adv. Synth. Catal.
(2009) - et al.
ChemistrySelect
(2016) - et al.
J. Am. Chem. Soc.
(2012)
J. Am. Chem. Soc.
ACS Catal.
Gaussian 09, Revision A.1
Phys. Rev. B
J. Chem. Phys.
J. Chem. Theor. Comput.
J. Chem. Phys.
J. Chem. Phys.
Theor. Chim. Acta
Cited by (0)
- 1
Both authors contributed equally to this work.