ACS Publications. Most Trusted. Most Cited. Most Read
My Activity
CONTENT TYPES
RETURN TO ISSUEPREVResearch ArticleNEXT

Speciation of ZnMe2, ZnMeCl, and ZnCl2 in Tetrahydrofuran (THF), and Its Influence on Mechanism Calculations of Catalytic Processes

  • Juan del Pozo
    Juan del Pozo
    IU CINQUIMA/Química Inorgánica, Facultad de Ciencias, Universidad de Valladolid, 47011 Valladolid, Spain
  • María Pérez-Iglesias
    María Pérez-Iglesias
    IU CINQUIMA/Química Inorgánica, Facultad de Ciencias, Universidad de Valladolid, 47011 Valladolid, Spain
  • Rosana Álvarez
    Rosana Álvarez
    Departamento de Química Orgánica, Facultad de Química (CINBIO), Universidade de Vigo, Campus As Lagoas-Marcosende, 36310 Vigo, Spain
  • Agustí Lledós*
    Agustí Lledós
    Departament de Química, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Spain
    *E-mail: [email protected] (A. Lledós).
  • Juan A. Casares*
    Juan A. Casares
    IU CINQUIMA/Química Inorgánica, Facultad de Ciencias, Universidad de Valladolid, 47011 Valladolid, Spain
    *E-mail: [email protected] (J. Casares).
  • , and 
  • Pablo Espinet*
    Pablo Espinet
    IU CINQUIMA/Química Inorgánica, Facultad de Ciencias, Universidad de Valladolid, 47011 Valladolid, Spain
    *E-mail: [email protected] (P. Espinet).
Cite this: ACS Catal. 2017, 7, 5, 3575–3583
Publication Date (Web):April 11, 2017
https://doi.org/10.1021/acscatal.6b03636

Copyright © 2017 American Chemical Society. This publication is licensed under these Terms of Use.

  • Open Access

Article Views

4550

Altmetric

-

Citations

LEARN ABOUT THESE METRICS
PDF (2 MB)
Supporting Info (1)»

Abstract

This paper examines the speciation of ZnMe2, ZnMeCl, and ZnCl2 in tetrahydrofuran (THF) solution, and provides experimental (infrared (IR) and calorimetric experiments) and density functional theory (DFT) data to conclude that the species in THF solution are, beyond a doubt, ZnMe2(THF)2, ZnMeCl(THF)2, and ZnCl2(THF)2. Consequently, coordinated THF should be used in DFT mechanistic calculations (e.g., Negishi reactions in THF). Naked or single THF coordinated molecules are virtually nonexistent in THF solutions. A cluster-continuum model for the solvent is recommended to obtain a more-correct representation. The presence or absence of THF produces a marked effect on the thermodynamics of the Negishi catalysis and a lesser effect on the kinetics of the transmetalation.

Introduction

ARTICLE SECTIONS
Jump To

Many transition-metal-catalyzed reactions are carried out in noncoordinating solvents; however, more frequently, the solvents used have some—from modest to high—coordinating ability. This is seldom a problem for experimental kinetic studies of reaction mechanisms: the data measured in the experiment simply impose their own physical value, regardless of the prejudices of the experimentalist about the possible implication of the solvent to produce these values. In a theoretical calculation, however, it is the decision of the theoretician to keep open the opportunity for the solvent to participate in the mechanism, or deny it by not including explicit solvent molecules as potentially coordinating ligands. (1,2) This choice influences the entire evolution of the computational study.
Modeling organic solvents as a polarizable continuum medium characterized by its dielectric constant (implicit solvent model) is mandatory in density functional theory (DFT) calculations of organometallic reactions; (3) however, this correction cannot spare the need of explicit solvent molecules as potential ligands that can interact coordinate to acid metal centers, with the corresponding enthalpic and entropic consequences. A recent example has shown that the direct Ar–benzyl coupling of ZnPh2 with benzyl halides occurs very fast and efficiently in benzene, but becomes very slow and inefficient if tetrahydrofuran (THF) is added to the benzene solution. (4) This experimental observation was interpreted considering that THF coordination to ZnPh2 is stronger than the interaction of the ZnPh2 with the benzyl halide reagent. Since the latter is needed for the direct coupling, the addition of THF quenches the coupling. This behavior was also found by theoretical calculations, only when THF molecules were explicitly included. The case could have never been explained with neglect of the solvent molecules.
Calculations not adding solvent molecules as coordinated ligands often create in silico species that cannot be conceived to exist in the presence of a coordinating solvent (the reader will excuse us for not mentioning specific examples here). For the case of good coordinating solvents (e.g., acetonitrile (MeCN), dimethylformamide (DMF), hexamethylphosphoramide (HMPA), pyridine (py), dimethyl sulfoxide (DMSO), etc.), a chemically judicious prior guess in favor of their participation can be made. This adds complexity to the mechanism and, very importantly, brings about the problem of how to address entropy. (1,5,6) Moreover, in cases of small enthalpy values, the accompanying entropic contribution is very significant, and overestimation of the entropy might introduce high percentages of error in ΔG, perhaps leading to mistaken interpretations. (7) Thus, in these cases, the decision to use (or not use) explicit participation of solvent molecules can be controversial, and disregarding the possible participation of poorly coordinating solvents is a reasonable temptation. Papers with both practices can be found. Skipping references to other authors, in some of our papers on the mechanism of the Negishi reaction, we have used ZnMe2(THF)2; in others, we have used just ZnMe2, depending on the preference of the theoreticians contributing to the paper. (8) In both circumstances, our calculations have satisfactorily reproduced the experimental ΔG data available. For this reason, we are still in doubt regarding whether ZnMe2(THF)2 and not just ZnMe2 must be considered. Since the Negishi reaction is gaining importance, particularly for difficult coupling reactions involving C(sp3), and since THF is the usual solvent for these reactions (but also for many others!), we have decided to investigate experimentally the speciation of ZnMe2, ZnClMe, and ZnCl2 in THF.
What is known today about ZnMe2 and ZnMe2(THF)2? Back in 1859, Sir Edward Frankland reported the synthesis of ZnMe2. (9,10) He noticed that its separation from the solvent (OEt2) was almost impossible. ZnMe2 is a liquid at room temperature, which was structurally characterized as a symmetric linear molecule (sp hybridization for Zn) in the gas phase in 2003, (11) and in the solid state (dC–Zn = 1.927(6) Å) in 2011. (12) Quantum mechanical calculations indicate considerable polarization of the two identical Zn–Me bonds toward the C atoms. (13) Surprisingly, there are no recent experimental studies on the formation of complexes of ZnMe2 with ethers (often used as reaction solvents), but remarkable studies by Thiele in 1962 showed that cyclic ethers O(CH2)n (n = 2–5; this includes the case of THF) form 1:2 complexes with ZnMe2. (14) Although this study lacks modern structural characterization, the stoichiometry ZnMe2(THF)2 was ascertained by full elemental analysis. Determination of its apparent molecular weight by cryoscopy, in benzene or in cyclohexane, afforded values lower or much lower than expected for the 1:2 adduct, meaning that ZnMe2(THF)2 undergoes partial THF dissociation in these solvents (somewhat higher in benzene than in cyclohexane). (14,15) Another indication that the binding of THF to ZnMe2 is not strong was the observation of a large dissociation of ZnMe2(THF)2 in its vapor. (14) To the best of our knowledge, a structural characterization of ZnMe2(THF)2 in the solid state is still lacking. The available X-ray structure of [Zn(C6F5)2(THF)2] displays a highly distorted tetrahedral geometry, (16) with two angles (C–Zn–C = 132.1° and O–Zn–O = 92.4°) very different from the 109.47° in an ideal tetrahedron. (17) The related [ZnCl2(THF)2] also displays a distorted tetrahedral geometry with Cl–Zn–Cl = 127.24° and O–Zn–O = 97.52°, and the two slightly differently arranged THF molecules produce different Zn–Cl (2.176 and 2.184 Å) and Zn–O (2.009 and 2.016 Å) bonds. (18) ZnEtCl exists in the solid state as a coordination polymer with three-coordinated Zn. (19) As for the Schlenk-type equilibria between ZnR2 and ZnX2 (R = Me, Et; X = Cl, I), it is well established by infrared (IR) spectroscopy that they are strongly shifted toward the formation of ZnRX in THF and in OEt2. (20)
Notwithstanding the reported facility of ZnMe2(THF)2 to dissociate THF in other circumstances, (14) it seems reasonable to us that, in solutions of ZnMe2 in THF, the coordinating nature of THF and the enormous concentration of THF in itself, acting as a solvent, (21) should largely push any associative equilibrium of THF with ZnMe2 toward the tetracoordinated compound. Here, we are searching for objective data, both from experiment and from calculations, on the speciation of Zn derivatives in THF solution.

Results and Discussion

ARTICLE SECTIONS
Jump To

Nuclear Magnetic Resonance (NMR) Studies

Variable-temperature nuclear magnetic resonance (NMR) is the technique that usually allows for precise experimental studies in solution, but our experiments on ZnMe2/THF mixtures failed to provide useful information. The 1H or 13C spectra of a 1:2 molar ZnMe2:THF mixture at −60 °C showed only one Me signal. Additions of THF did not produce any significant drift. This is compatible with THF dissociation being negligible in THF or, more likely, with NMR not being appropriate for this problem.

Infrared (IR) Spectroscopy Studies

Solutions made by mixing THF and the corresponding Zn compound (ZnMe2, ZnMeCl, or ZnCl2) under nitrogen, were examined by infrared (IR) spectroscopy. Neat THF (no Zn present) displays two C–O intra-ring stretching bands at ∼1067 cm–1asym) and ∼908 cm–1sym). (22) Since, in the Zn complexes, these vibrational modes also involve participation of the O–Zn bond, the two absorptions are useful to evaluate the THF/Zn interactions. Figure 1a shows the IR spectra in the range of 800–1200 cm–1 for the different solutions of ZnMe2 in THF, mixing the two reagents in 1:1, 1:2, and 1:3 ratios in order to allow for the formation of the potential complexes ZnMe2(THF), ZnMe2(THF)2, or higher coordination numbers. The wavelengths corresponding to the main species observed in Figure 1 are listed in Table 1.

Figure 1

Figure 1. (a) IR spectra of mixtures of ZnMe2 and THF in ratios of 1:1 (black), 1:2 (red), and 1:3 (blue); (b) IR spectra of mixtures of ZnMe2 (blue), ZnMeCl (red) or ZnCl2, and THF. The bands of uncoordinated THF (1067 and 908 cm–1) are observed in all of the spectra.

Table 1. IR Wavenumbers for C–O Stretching of THF
entrycompoundν(C–O)sym; ν(C–O)asym (cm–1)Δν(C–O)asym; Δν(C–O)sym (cm–1)a
1neat THF1067; 908 
2ZnMe2 + 1 equiv THF1048;b 888b19; 20b
3ZnMe2(THF)21051; 89016; 18
4ZnClMe(THF)21033; 87434; 34
5ZnCl2(THF)21028; 86639; 42
a

Δν(C–O) = ν(C–O)neat – ν(C–O)coord.

b

Deceptive!!

In the 1:1 mixture of THF and ZnMe2, the 1067 cm–1 ν(C–O)asym band of THF shifts down to 1048 cm–1. An analogous effect (shift from 908 to 888 cm–1) is found for the symmetric stretching band (Figure 1a and Table 1, entry 1), suggesting weakening of the C–O bond upon THF coordination.
The addition of further THF is done first to reach the stoichiometric proportion ZnMe2:THF = 1:2, and then to provide a large excess of THF (1:3 ratio), giving rise to an apparent shift of the minimum of the band assigned to coordinated THF, from 1048 cm–1 to 1051 cm–1 and then to 1055 cm–1. A principal component analysis of the spectra (Figures S2 and S3 in the Supporting Information (SI)) shows that this shift is a deceptive appearance that is due to the overlap of ν(C–O)asym for ZnMe2(THF)2 (1048 cm–1) and free THF (1067 cm–1) in different proportions. The same deceptive effect is observed for ν(C–O)sym. Interestingly, in the ZnMe2:THF = 1:2 mixture, the ν(C–O)asym band shows a shoulder at 1067 cm–1, revealing that ZnMe2(THF)2 undergoes some dissociation. (23)
In conclusion, all the IR spectra observed detect only ZnMe2(THF)2 and free THF, regardless of the ZnMe:THF ratio. Unobservable ZnMe2 must exist for THF/Zn ratios lower than 2:1. The existence of small amounts (below the limits of observation) of ZnMe2(THF) in equilibrium cannot be fully disproved; however, overall, the coordination of a second THF molecule to an initial species ZnMe2(THF) is clearly preferred to the coordination of a first THF molecule to ZnMe2. In other words, as suggested in Scheme 1, the existence of three-coordinated ZnMe2(THF) seems very unfavorable versus its rearrangement to a mixture of ZnMe2 and ZnMe2(THF)2.

Scheme 1

Scheme 1. THF Coordination Equilibria, As Suggested by IR Observations
The changes in ν(C–O) upon coordination to ZnMe2 are small (Table 1, (Δν(C–O) = 20 cm–1), suggesting that the O–Zn dative bond in ZnMe2(THF)2 is weak. The O–Zn bonds are expected to be stronger for more acidic ZnXY fragments. In effect, the differences Δν(C–O) rise to 34/34 cm–1 and 39/42 cm–1, respectively, for the symmetric/asymmetric vibrations in ZnClMe(THF)2 and ZnCl2(THF)2 (Figure 1b). (24)
IR evidence is fairly conclusive, supporting that, in THF, the three Zn derivatives ZnMe2, ZnMeCl, and ZnCl2 are only observed as their ZnXY(THF)2 complexes. However, dissociation equilibria with very small ΔG0 values (e.g., 2 kcal mol–1) could make the concentration of other species, such as ZnXY(THF), unobservable. Furthermore, IR observation cannot detect uncoordinated ZnXY molecules, such as ZnMe2. Consequently, below, we examine further evidence via other methods.

Calorimetric Studies

The heat of reaction of liquid ZnMe2 with THF was measured in a microcalorimeter by adding THF to neat ZnMe2. First, two moles of THF per mole of ZnMe2 were added to reach the stoichiometric ratio of ZnMe2:THF = 1:2. Then, an excess of THF was injected. The first addition of THF produced ΔH = −6.3 kcal mol–1. The second injection produced only a modest additional heat release (ΔH = −1.6 kcal mol–1). These results support that coordination had not been fully completed after the addition of stoichiometric THF and required an excess of solvent in order to further shift the coordination equilibrium toward ZnMe2(THF)2. This tendency of ZnMe2(THF)2 to establish some dissociation equilibrium in THF is in agreement with the IR observations, and it means that part of the heat released in the second addition comes from fulfilling the tetracoordination of ZnMe2 and is not due to the dilution enthalpy of ZnMe2(THF)2 (l) in THF (l) (so-called excess enthalpy). The overall heat released after the addition of an excess of THF (−7.9 ± 0.2 kcal mol–1) accounts for the enthalpy of coordination of THF to ZnMe2 to form ZnMe2(THF)2 plus the excess enthalpy, but the two contributions cannot be measured independently. Excess enthalpy values reported in the literature suggest that this contribution should be small, compared with the bond energies. Reported values of excess enthalpy, corresponding to the dilution of nonprotic solvents (that we take as models of ZnMe2) in THF, span from about +0.19 kcal mol–1 for n-hexane (maximum value for a molar fraction of χ = 0.5), (25) to +0.026 kcal mol–1 for acetonitrile (maximum value for χ = 0.35). (26) Also reported, the heat of dilution of MgEt2(THF)2 from 0.2 M to 0.043 M in THF is −0.2 kcal mol–1. (27) Taking these data as reference, a reasonable error estimation if ignoring the heat of dilution in our measured figures may be approximately ±0.2 kcal mol–1, leading to ΔHsolv0 = −7.9 ± 0.4 kcal mol–1 for the bonding enthalpy of two THF molecules to ZnMe2 (∼4 kcal mol–1 for one bond). This is a remarkably small value for an overall coordination enthalpy involving two ligands. Comparing with related structures, it is only about one-third of the calculated bonding enthalpy of two molecules of THF to MgMe2. (28)
Experimental calorimetric data could not be obtained under similar conditions for ZnMeCl or for ZnCl2, (29) but the THF coordination energy is expected to be stronger for more-electronegative substituents on Zn (ZnMe2 < ZnMeCl < ZnCl2). This is supported by the calculations discussed below.

DFT Studies: Structural Results

DFT optimization of the structures of [Zn(C6F5)2(THF)2] were made with different functionals and compared with its X-ray determined structure, (16) The best agreement was found with the M06 functional (see the SI), which was selected to calculate the structures of the Zn compounds and their possible complexes with one or two THF molecules (Figure 2). Table 2 lists informative computed distances and angles of these structures.

Figure 2

Figure 2. Representation of the DFT computed structures. [Legend: red, O; yellow, Zn; gray, C; and green, Cl.]

Table 2. DFT Computed Distances and Angles of Zn Derivativesa
  Angles (deg)Distances (Å)
entrycompoundO–Zn–OC–Zn–XZn–OO–CZn–CZn–Cl
1ZnMe2 (1) 180  1.932 
        
2ZnMeCl (2) 179.9  1.9202.175
        
3THF (3)   1.426, 1.426  
        
4ZnMe2(THF) (4) 174.22.2081.437, 1.4371.944, 1.944 
        
5ZnMe2(THF)2 (5)82.7161.22.2111.429, 1.4321.967, 1.967 
    2.2111.429, 1.432  
        
6ZnMeCl(THF) (6) 166.92.0921.443, 1.4441.9322.190
        
7ZnMeCl(THF)2 (7)89.8147.32.0521.439, 1.4371.9652.258
    2.0521.442, 1.437  
        
8Zn(C6F5)2(THF)2 (8)92.4(1)132.1(2)2.093 1.999, 2.012 
    2.113   
        
9ZnCl2(THF)2 (9)97.5(2)127.2(1)2.009, 2.016  2.176, 2.184
a

Data for 8 and 9 have been taken from their X-ray structures.

The set of structures and their geometrical parameters show remarkable features (more marked for ZnMe2). Since the donor–acceptor Zn–THF bond distances are long, compared to the sum of covalent radii for Zn and O, these bonds must be weak, because of smaller orbital overlap, and have a relatively significant electrostatic contribution. There are precedents for this analysis. In fact, the highly distorted tetrahedral structure calculated for ZnMe2(THF)2, with small O–Zn–O and large Me–Zn–Me angles, finds similar structural features in isoelectronic Mg(alkyl)2(THF)2 compounds. (30) For these Mg complexes, an analysis of the chemical bonding with generalized valence bond (GVB) wave functions has concluded that the THF molecules and the Mg atom are held together only by classical electrostatic forces. (31) Although somewhat higher covalent participation can be expected for Zn, with a smaller covalent radius than Mg (rcov(Zn) = 1.22 Å, rcov(Mg) = 1.41 Å), (32) there are no steric hindrance reasons to be argued for the long Zn–O calculated distances (up to 2.211 Å; ∼1.90 Å is expected for a covalent single bond). (33) We find it plausible that the p-orbitals in ZnMe2(THF)2 are less prone to participate in covalent interactions, especially in donor–acceptor interactions, because of notable nuclear shielding on Zn by the strongly donor Me substituents, which makes the p-orbitals high in energy and, consequently, bad acceptors. Increasing the covalence for the Zn–O bonds can be expected when more-electronegative substituents replace Me groups, increasing the electrophilicity of Zn. This is reflected in a shortening of the Zn–O bond lengths (from 2.211 Å for entry 5 in Table 2 to 2.009 Å for entry 9 in Table 2). In the same sense, the C–Zn–X bond closes (from 161° for entry 5 in Table 2 to 127.2° for entry 9 in Table 2) and the O–Zn–O angle opens, less steeply (from 82.7° for entry 5 in Table 2 to 97.2° for entry 9 in Table 2).
ZnMe2 (1) and ZnMeCl (2) have perfectly linear calculated structures, as expected for sp orbitals for Zn. The changes upon coordination are discussed first for ZnMe2. Upon coordination of one THF molecule, the linear bonding arrangement in 1 undergoes a moderate deformation of the Me–Zn–Me angle, giving a T-shaped structure for the putative ZnMe2(THF) (4), not detected experimentally. Coordination of a second THF molecule forms tetracoordinated 5, with a very large Me–Zn–X angle and a very small O–Zn–O angle, relative to identical 109.47° angles in a regular tetrahedron (Table 2). Assuming that the Zn–O interactions are significantly electrostatic, as in Mg(alkyl)2(THF)2, the THF ligands should be placed where the nuclear charge is less shielded (that is, in the plane bisecting the Me–Zn–Me bond). It is there where the THF ligands are found in 4 and 5. The O–Zn–O angle, close to 90° in 5, is due to the directional covalent donor–acceptor contribution to this bond, which will make more efficient overlapping in the direction of the empty Zn orbitals. In simple terms, we could say that, for the Me–Zn–Me bonds, the Zn atom is contributing two approximately spz orbitals (with somewhat less s and more p in the contribution, in order to bend the angle from 180°) and, approximately, the empty pxpy orbitals (with less p and some s contribution) determining the O–Zn–O angle.
The trend for Zn to produce angles getting closer to tetrahedral (getting closer to sp3 hybridization) when the electronegativity of the substituents increases reflects the stabilization and increasing covalent participation of the p orbitals. The progressive shortening of the O–Zn distances (2.21 Å in ZnMe2(THF)2, 2.052 Å in ZnMeCl(THF)2, 2.009 Å in ZnCl2(THF)2) supports an increasingly covalent (donor–acceptor) participation in the O–Zn bond. The variation of O–Zn distances is in agreement with the observed trend of vibrational frequencies in the IR region.
The possible existence of five-coordinated species with three-coordinated THF molecules was investigated. (34) All attempts to optimize ZnMe2(THF)3, starting with three solvent molecules coordinated to zinc always ended producing ZnMe2(THF)2 and one THF molecule expelled from the metal. This suggests that such an adduct with three solvent molecules is not stable.

DFT Studies: Thermodynamic Results

Inclusion of solvent molecules in the coordination sphere of a metal is a particular case of association–dissociation process in a condensed phase. The accurate description of such processes by conventional DFT methodology is fraught with difficulties, among them, those related with entropy changes in solution, (35) which can be circumvented by ab initio molecular dynamics (AIMD) simulations, in which the solute is placed in a box containing hundreds of solvent molecules (not only the two coordinating molecules). (36,37) However, AIMD calculations are extremely time-intensive. (38) For this reason, two simpler approximations to address solvent coordination effects were applied here to calculate the ΔG0 of the THF-bonded zinc species in solution.
The simplest approximation, which is frequently used in published works, introduces explicitly only two THF molecules per Zn, in consonance with our experimental results (Scheme 2a). (39) The solvent molecules that will eventually get involved in coordination appear as individual noninteracting independent particles, and the total number of particles in the equilibrium decreases when THF molecules coordinate to Zn, from n + 1 to n, and then to n – 1. This is a major unfavorable entropic contribution to the Gibbs energy of coordination. This model may be correct for reactions in gas phase, but it overestimates the entropy contribution for reactions in condensed phase, where the solvent molecules are interacting with each other, and are not free molecules.

Scheme 2

Scheme 2. Coordination Equilibria of THF Plus ZnMe2: (a) Considering the Solvent as Individual Molecules and (b) Considering the Solvent as a Cluster of n Molecules (n = 2–4)
In our case, because of the small magnitude of the enthalpic contributions measured by calorimetry for ZnMe2(THF)2, a very likely overestimation of entropy might be particularly misleading for the calculated ΔG0 values. One way to adjust this oversimplified description and attenuate the entropic contribution of THF is to consider the uncoordinated solvent molecules not as free individuals, but as being involved in intermolecular dipolar links with other solvent molecules. This new perspective can be modeled considering a certain number (n) of solvent molecules as constituting a cluster: (THF)2, (THF)3, (THF)4, etc., from which the coordinating molecules will proceed. This means that, for n ≥ 3, the total number of particles remains constant (two particles) throughout the equilibria of formation of the two ZnMe2 complexes (4 and 5) with one (K1) or, respectively, two (K3) THF molecules, as shown in Scheme 2, eliminating a major contribution to entropy changes upon coordination. This cluster-continuum model has been successfully applied to the calculation of pKa values, (40) and it has been extended to the calculation of Gibbs energy changes in solution. (41)
The calculated Gibbs energy values (ΔG0) for the equilibria in Schemes 2a (only for the case of ZnMe2 and two individual THF molecules) and 2b (for ZnMe2 with 2–5 THF clusters; for ZnClMe and ZnCl2 with 4 THF cluster molecules) are gathered in Table 3.
Table 3. DFT Calculated ΔG0 Values (at 0 °C) for the Equilibria in Scheme 2
  ΔG0 (kcal mol–1)
entryn (model)for K1for K2for K3
1 (ZnMe2)n = 1 + 1 (2a)0.1–6.1–6.0
2 (ZnMe2)n = 2 (2b)–3.9–6.1–10.0
3 (ZnMe2)n = 3 (2b)–4.6–10.1–14.7
4 (ZnMe2)n = 4 (2b)–3.7–10.8–14.5
5 (ZnMe2)n = 5 (2b)–3.2–9.8–13.0
6 (ZnMeCl)n = 4 (2b)–8.3–15.7–24.0
7 (ZnCl2)n = 4 (2b)–17.3–23.4–40.7
Regardless of the approximation chosen and the size of the cluster, all of the entries in Table 3 share the characteristic that the coordination of the second THF molecule is more exergonic than the first coordination. This is particularly marked for ZnMe2. The differences between these values for K1 and K2 are so high that the concentration of ZnMe2(THF) in equilibrium with ZnMe2(THF)2 should be negligible. This is consistent with our nonobservation of the three-coordinated species.
Usually, the incorporation of a second covalently coordinated ligand to a metal center, in reactions that proceed via the substitution of weak ligands for stronger ligands, is less favorable than incorporation of the first one, because after the first coordination, the metal center exhibits less acceptor nature toward the second one. The opposite observation here indicates that the higher stabilization of the molecule in the incorporation of the second ligand is not due to only the bond energy of the new Zn–THF bond that is formed. As a matter of fact, the two Zn–O bonds are slightly longer (recall Table 2) and, hence, are presumably weaker in ZnMe2(THF)2 than in ZnMe2(THF), and the same applies to the two Zn–Me bonds. Structural changes can induce other effects that can be decisive to the overall stability of a molecule. We suggest that part of the molecular stabilization of ZnMe2(THF)2 is due to the structural change produced upon coordination of the second THF molecule, closing the Me–Zn–Me angle (this angle varies only very little upon coordination of the first THF molecule). Extrapolating from the well-known trans-influence in square-planar complexes, where highly donor ligands try to avoid mutually trans positions, the destabilizing contribution associated with two powerful bond dipolar moments oriented opposite to each other in ZnMe2 (180°) is lowered progressively when the angle between these dipoles closes, as in ZnMe2(THF)2 (161°). (42) Overlooking these structural effects on the ΔG0 values upon coordination might lead to the deceptive conclusion that the Zn–O bond energies in ZnMe2(THF)2 are higher than for the same bond in ZnMe2(THF). As indicated previously, the bond distances in Table 2 support exactly the opposite.
The calculations with only one coordinating THF (Table 3, entry 1) suggest that the formation of ZnMe2(THF) (K1) should be slightly unfavorable (positive ΔG0). Anyone trusting a prior exploration with this calculation would be misled to decide not to include coordinating THF. All the data in our study prove that it is compulsory to introduce two coordinating THF molecules for correct calculations. The misleading result of entry 1 in Table 3 for K1 is a warning of the danger of trusting partial results.
Although it is not possible to separate enthalpic and entropic contributions in calculations with a continuum model, (43) some observations can be made by combining the calculated ΔG0 results with the experimentally determined ΔHsolv0. Looking at the most relevant data in Table 3 (those in column K3 in Table 3) and taking into account the experimental enthalpy value (ΔHsolv0 = −7.9 kcal mol–1 for the coordination of two THF), the ΔG0 values show that
(i)

moving from method 2a (entry 1) to method 2b (entry 2) produces an immediate change from unfavorable entropy contribution when the solvent is considered as independent molecules (TΔS0 = −1.9, since ΔG0 = −6.0 is less negative than ΔHsolv0), to favorable entropic contribution (TΔS0 = +2.1, since ΔG0 = −10.0 is more negative than ΔHsolv0) when the solvent is handled as a cluster; and

(ii)

this effect improves when the equilibrium in Scheme 2 keeps the number of particles during the coordination reaction constant and equal to 2 (for n > 2, entries 3–5), and becomes approximately steady in the range of n = 3–5.

We suggest cluster (THF)4 as a less expensive but sufficiently good representation of the solvent. Clearly, all the calculations predict that the coordination of two THF molecules to ZnMe2 is significantly exergonic, in agreement with the THF coordination equilibria proposed in the experiment (Scheme 1). ΔG0 calculations for ZnMeCl or ZnCl2 (entries 6 and 7 in Table 3) show, as expected, more exergonic processes than ZnMe2.

DFT Studies: Relevance to the Transmetalation Step in the Negishi Coupling Using ZnMe2

These ΔG0 values have direct translation to the thermodynamics of the transmetalation steps of Me to a metal (often corresponding to Me for Cl exchanges), producing much larger differences in the transmetalation equilibria when coordinating THF is considered. Using the values in Table 3, we note that, in the (THF)4 cluster model, releasing the two THF molecules to obtain the corresponding naked ZnX2 compound would cost 14.5 kcal mol–1 from ZnMe2(THF)2 (entry 4), 24.0 kcal mol–1 from ZnMeCl(THF)2 (entry 6), and 40.7 kcal mol–1 from ZnCl2(THF)2 (entry 7). Consequently, the contribution of the presence of coordinating THF to the formation of ZnMeCl(THF)2 from ZnMe2(THF)2, calculated from the data in Table 3, is 9.5 kcal mol–1. (44) In other words, in a Negishi coupling, this transmetalation will be favored by 9.5 kcal mol–1 in THF, compared to a noncoordinating solvent. Similarly, if the Zn reagent was ZnMeCl and the transmetalation product was ZnCl2, the thermodynamic benefit for transmetalation in THF will be, compared to a noncoordinating solvent, 16.7 kcal mol–1. Thus, the Zn organometallics are more powerful transmetalating reagents in THF, and calculations omitting explicit THF give a largely inexact image of these transmetalation equilibria. The influence of the coordinating solvent on the values of the activation energy of the transmetalation, and, consequently, on the kinetic behavior with or without THF, is not as easy to analyze.
In order to examine this question in more detail, we have carried out a comparative DFT study of the transmetalation profiles on the model complex PdMeCl(PMe3)2 + ZnMe2, using or excluding coordinating THF (see Figure 3). (45,46)

Figure 3

Figure 3. Gibbs energy profile of the Negishi transmetalation step (kcal mol–1): (top) without explicit THF and (bottom) with two coordinating THF molecules (M06 optimizations in THF described by the SMD continuum model).

The two calculated ΔG values for the model studied are very similar (ΔGTHF = 13.5 kcal mol–1; ΔGnaked = 11.5 kcal mol–1) with a difference of only 1.9 kcal mol–1). Such a small difference suggests that the two methods could match equally well (or equally poorly) an experimental value for the activation energy (for instance, 12 kcal mol–1) if accepted approximations were applied. Things are very different, with regard to the ΔG0 values. Both profiles predict an unfavorable (endergonic) transmetalation, (47) but ΔG0 is very much affected by THF coordination (ΔGTHF0 – ΔGnaked0 = −9.5 kcal mol–1), as deduced from the data in Table 3 (entries 4 and 6). If ΔGTHF0 matches an experimental value reasonably well, ΔGnaked0 will largely fail. Moreover, since the naked profile predicts that the formation of the dimethyl palladium complex would cost 18.7 kcal mol–1, E becomes higher in energy than the transmetalation transition state C. In these circumstances, the transmetalation activation energy (associated with C) would become kinetically irrelevant if the process had to continue to produce the coupling product. However, in the profile with coordinated THF, the transition state H, being higher than J, would still be relevant and potentially rate-determining (depending on the activation barriers of other steps in the cycle).
In experimental studies, the ΔGexper values can be determined with good precision, usually by nuclear magnetic resonance (NMR), measuring the reaction rates of a transmetalation process isolated from the catalytic cycle. In contrast, quantification of ΔG0 requires the observation and integration of the products at both sides of the equilibrium, which is only occasionally accessible. (48) For this reason, often, the only experimental datum determined is ΔGexper. (8) Consequently, a satisfactory experiment–calculation match only requires a good reproduction of ΔGexper, which is equally achieved in both profiles of Figure 3. Possible mismatches in other points of a more-complex profile, or in ΔG0 in the simple profile of Figure 3, can remain hidden or be overlooked when ΔGexper values are not available. (49)
Figure 3 illustrates well that both calculations can afford reasonably good values of ΔG but, for a reaction in THF of the most frequent case of Negishi catalysis, which involves alkyl/Cl exchange, the one without THF will strongly fail in ΔG0. However, it is convenient to note that other exchanges in Negishi catalysis have less marked electronegativity differences of the groups exchanged, which will lead to lower ΔGTHF0 – ΔGnaked0 calculated differences for them. As a matter of fact, in our own previous studies without coordinated THF, the results for ΔG0 were not warning us of the problem studied here, because they never involved the primary Me/Cl exchange, (50) but only Me/Me exchanges in ZnMe2-catalyzed Pd isomerization processes (ΔG0 = 0), (8c) or Me/Ph exchanges in the study of so-called “secondary transmetalations” (ΔG0 moderately affected by THF coordination). (8d) Finally, it is also clear that solvents more strongly coordinating than THF will produce larger ΔGsolvent0 – ΔGnaked0 differences.

Conclusions

ARTICLE SECTIONS
Jump To

All the experimental and calculated data support, beyond any doubt, that, in THF solution, the species ZnMe2, ZnMeCl, and ZnCl2 are coordinated with two THF molecules. The bonding forces for this coordination have a high electrostatic content for ZnMe2(THF)2 and become increasingly covalent for more electronegative substituents on Zn, which increase the electrophilicity of the Zn center. The structural effect is that the high distortion from an ideal tetrahedron, found in ZnMe2(THF)2 moves toward less-distorted structures and smaller Y–Zn–X angles in the same sense. Particularly for ZnMe2, the structural change upon closing the Me–Zn–Me angle is probably a significant contribution to the stabilization of the complex, whereas, for species with more electronegative Y and/or Z substituents, the formation of two extra and stronger Zn–THF bonds is more relevant.
Any theoretical study in THF should consider the species ZnMe2(THF)2, ZnMeCl(THF)2, and ZnCl2(THF)2 as the ones participating in reactivity. Calculations in the literature for reactions in THF (or other coordinating solvents) without these two coordinating molecules may be considered useful models for reactions carried out in noncoordinating solvents (obviously needing a rectification of the implicit solvent model correction), but do not represent what happens in a coordinating solvent. Structural proposals featuring intermolecular or intramolecular interactions to Zn should be examined with caution. They are probably inconsistent with the speciation in a real solution, where stronger Zn–THF interactions will be formed instead. (51)
Studies carried out with two explicit molecules of THF to coordinate Zn are, in principle, fully correct. The use of a cluster model for the solvent is recommended as a nonarbitrary procedure to improve the estimated ΔG and ΔG0 values.
In the model case studied here, the lack of two molecules of coordinated THF produces only a moderate difference in ΔG, but the difference in ΔG0 for Me/Cl transmetalation is very large, showing that the information on the transmetalation equilibrium in THF is very unrealistic if obtained in calculations without coordinated THF. It can be more acceptable for exchanges of groups of similar electronegativity.
A practical consequence of this study is that the transmetalation equilibria ZnMe2/ZnMeCl are much more favorable in a coordinating solvent than in a noncoordinating one. Moreover, the ΔG0 difference in a coordinating solvent is even larger for the transmetalation pair ZnMeCl/ZnCl2. Combining ZnMeCl and a coordinating solvent should provide the most favorable thermodynamics for Negishi reactions.
Although the observations in this paper concern direct studies of Negishi catalysis, whether theoretical or experimental, similar solvent dependence can be expected for other catalysis using organometallics of other electropositive metals as nucleophiles. The structural similarity of zinc and magnesium derivatives, already noted in the text, immediately connects these conclusions to the Kumada coupling.
Finally, this study is a good occasion to share our thoughts about the best way to progress in the use of experimental + computational chemistry. Although calculations are not real life, they provide invaluable content and knowledge that is otherwise inaccessible. Progress in calculations needs experimental reference values as the touchstone of developments in theoretical methods. It is a hard task to get these experimental data. The accuracy of the experimental values is dependent on the precision and correctness of the experiments, not on whether they match the calculations better or worse. The accuracy of the computations is dependent on the ability to describe many molecular factors mathematically, and improved methods must be in continuous development. Common sense must be exercised to get the best of both worlds with a reasonable effort.
For experimental works that include calculations, less-exact calculations can provide sufficient mechanistic information. Often, the very valuable structural and mechanistic information that calculations provide will not change with the choice of functional or with the size of model used. In fact, methodological changes such as the choice of functional can produce large changes in the calculated thermodynamic parameters without significant changes in the mechanistic and structural information. (52) Experimental papers should not be charged, without a good reason, with excessive demands to perform heavy adjustment of their accompanying DFT calculations. They cannot be punished for having experimentally determined the value that is pursued with an improved calculation.
On the other hand, for theoretical papers aimed at developing methodology, tuning a more perfect match of calculation to an experimental value is the natural way to check their quality. However, papers containing only calculations with mechanistic proposals should, sometimes, be required to provide some experimental support, or at least be asked for bibliographic support of plausibility in real conditions of their mechanistic proposals. Theoretical projects should be aware, when setting the theoretical contour conditions of the calculation, that the absence of just one species present in the flask but not in the computer can make the results unrealistic. This can be, in our opinion, more important than pursuing higher accuracy.

Supporting Information

ARTICLE SECTIONS
Jump To

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acscatal.6b03636.

  • Experimental methods and data, computational methods and data, and associated references (PDF)

Terms & Conditions

Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.

Author Information

ARTICLE SECTIONS
Jump To

  • Corresponding Authors
  • Authors
    • Juan del Pozo - IU CINQUIMA/Química Inorgánica, Facultad de Ciencias, Universidad de Valladolid, 47011 Valladolid, SpainPresent Address: Presently at Merkert Chemistry Center, Boston College, Chestnut Hill, MA 02467, USA
    • María Pérez-Iglesias - IU CINQUIMA/Química Inorgánica, Facultad de Ciencias, Universidad de Valladolid, 47011 Valladolid, Spain
    • Rosana Álvarez - Departamento de Química Orgánica, Facultad de Química (CINBIO), Universidade de Vigo, Campus As Lagoas-Marcosende, 36310 Vigo, Spain
  • Funding

    The following entities are acknowledged for providing financial support: Spanish Ministerio de Economía, Industria y Competitividad (MINECO); Spanish Ministerio de Educación (MINEDUC); and Spanish Consejería de Educación, Junta de Castilla y León (JCyL).

  • Notes
    The authors declare no competing financial interest.

Acknowledgments

ARTICLE SECTIONS
Jump To

Financial support is gratefully acknowledged to the Spanish MINECO (Project Nos. CTQ2013-48406-P, CTQ2014-54071-P, CTQ2014-52796-P, CTQ 2016-80913-P) and the Junta de Castilla y León (Project No. VA256U13). J.d.P. thanks the Ministerio de Educación for an FPU grant and the Alfonso Martín Escudero Foundation for a postdoctoral fellowship. M.P.-I. thanks the Junta de Castilla y León for a contract as predoctoral researcher.

References

ARTICLE SECTIONS
Jump To

This article references 52 other publications.

  1. 1
    Plata, R. E.; Singleton, D. A. J. Am. Chem. Soc. 2015, 137, 38113826,  DOI: 10.1021/ja5111392
  2. 2
    Winter, A. A. Nat. Chem. 2015, 7, 473475,  DOI: 10.1038/nchem.2267
  3. 3
    Harvey, J. N. Faraday Discuss. 2010, 145, 487505,  DOI: 10.1039/B907340J
  4. 4
    Dunsford, J. J.; Clark, E. R.; Ingleson, M. J. Angew. Chem., Int. Ed. 2015, 54, 56885692,  DOI: 10.1002/anie.201411403
  5. 5

    In an oversimplified view, solvent coordination seems enthalpically favorable but entropically unfavorable, but this relationship is not that simple.

  6. 6
    (a) Cooper, J.; Ziegler, T. Inorg. Chem. 2002, 41, 66146622,  DOI: 10.1021/ic020294k
    (b) Dub, P. A.; Poli, R. J. Mol. Catal. A: Chem. 2010, 324, 8996,  DOI: 10.1016/j.molcata.2010.03.003
  7. 7

    A recent study of the difficulties associated with calculations has made very clear how difficult it is to decide which functional is reasonable for a given calculation or how to address entropy contributions. See ref (1).

  8. 8
    (a) Fuentes, B.; García-Melchor, M.; Lledós, A.; Maseras, F.; Casares, J. A.; Ujaque, G.; Espinet, P. Chem. - Eur. J. 2010, 16, 85968599,  DOI: 10.1002/chem.201001332
    (b) García-Melchor, M.; Fuentes, B.; Lledós, A.; Casares, J. A.; Ujaque, G.; Espinet, P. J. Am. Chem. Soc. 2011, 133, 1351913526,  DOI: 10.1021/ja204256x
    (c) delPozo, J.; Gioria, E.; Casares, J. A.; Álvarez, R.; Espinet, P. Organometallics 2015, 34, 31203128,  DOI: 10.1021/acs.organomet.5b00329
    (d) del Pozo, J.; Salas, G.; Álvarez, R.; Casares, J. A.; Espinet, P. Organometallics 2016, 35, 36043611,  DOI: 10.1021/acs.organomet.6b00660
  9. 9
    von Frankland, E. Ann. Chem. Pharm. 1849, 71, 171213,  DOI: 10.1002/jlac.18490710205
  10. 10
    (a) Seyferth, D. Organometallics 2001, 20, 29402955,  DOI: 10.1021/om010439f
    (b) Seyferth, D. Organometallics 2004, 23, 11721172,  DOI: 10.1021/om0499557
  11. 11
    Haaland, A.; Green, J. C.; McGrady, G. S.; Downs, A. J.; Gullo, E.; Lyall, M. J.; Timberlake, J.; Tutukin, A. V.; Volden, H. W.; Østby, K.-A. Dalton Trans. 2003, 43564366,  DOI: 10.1039/B306840B
  12. 12
    Bacsa, J.; Hanke, F.; Hindley, S.; Odedra, R.; Darling, G. R.; Jones, A. C.; Steiner, A. Angew. Chem., Int. Ed. 2011, 50, 1168511687,  DOI: 10.1002/anie.201105099
  13. 13
    Antes, I.; Frenking, G. Organometallics 1995, 14, 42634268,  DOI: 10.1021/om00009a032
  14. 14
    Thiele, K. H. Z. Anorg. Allg. Chem. 1962, 319, 183195,  DOI: 10.1002/zaac.19623190309
  15. 15
    Bochmann has reported the formation in toluene of Zn(C6F5)2·(toluene) from ZnMe2 and B(C6F5)3. It is thought that the toluene is coordinated to Zn. These interactions could explain the dissociation of ZnMe2(THF)2 in benzene observed by Thiele. See:Walker, D. A.; Woodman, T. J.; Hughes, D. L.; Bochmann, M. Organometallics 2001, 20, 37723776,  DOI: 10.1021/om0103557
  16. 16
    Weidenbruch, M.; Herrndorf, M.; Schäfer, A.; Pohl, S.; Saak, W. J. Organomet. Chem. 1989, 361, 139145,  DOI: 10.1016/0022-328X(89)85378-1
  17. 17

    The high electrophilicity of Zn(C6F5)2 is shown by the isolation of Zn(C6F5)2·(toluene) and Zn(C6F5)2·(C6Me6), obtained from toluene solutions, and is believed to have η2-coordinated arene.

  18. 18
    Dashti, A.; Niediek, K.; Werner, B.; Neumiiller, B. Z. Anorg. Allg. Chem. 1997, 623, 394402,  DOI: 10.1002/zaac.19976230163
  19. 19
    Guerrero, A.; Hughes, D. L.; Bochmann, M. Organometallics 2006, 25, 15251527,  DOI: 10.1021/om051043x
  20. 20
    Evans, D. F.; Wharf, I. J. Chem. Soc. A 1968, 783787,  DOI: 10.1039/j19680000783
  21. 21

    The activity concept cannot be rigorously applied under these conditions, but the simple numerical calculation for the number of moles in 1 L of THF would afford a concentration of 12.33 M.

  22. 22

    The labels “asym” and “sym” are assigned to describe the symmetry within the THF ring. Strictly speaking, the vibrational modes involve also the O–Zn bond in the case of the Zn complexes.

  23. 23

    Note that, in the spectrum with a lower concentration of THF (the black trace in Figure 1a), the bands corresponding to some noncoordinated THF are still observable, as a shoulder for the asymmetric stretching, and as a clear band at 908 cm–1 for the ν(C–O)sym. Thus the composition of this mixture contains some noncoordinated THF, tetrahedral ZnMe2(THF)2 and, necessarily, the remaining Zn as non-observable linear ZnMe2. The same applies to the dissociation observed for ZnMe2(THF)2 (the red trace in Figure 1a).

  24. 24

    The concentrations of the three samples recorded in the infrared (IR) spectroscopy analysis are very different, because of the low solubility of ZnCl2 in THF. As a result, the intensity of the noncoordinated THF band is also different for each sample.

  25. 25
    Guillén, M. D.; Gutiérrez Losa, C. J. Chem. Thermodyn. 1978, 10, 567576,  DOI: 10.1016/0021-9614(78)90045-9
  26. 26
    Letcher, T. M.; Domanska, U. J. Chem. Thermodyn. 1994, 26, 7584,  DOI: 10.1006/jcht.1994.1022
  27. 27
    Smith, M. B.; Becker, W. E. Tetrahedron 1967, 23 (11), 42154227,  DOI: 10.1016/S0040-4020(01)88819-0
  28. 28
    Tammiku-Taul, J.; Burk, P.; Tuulmets, A. J. Phys. Chem. A 2004, 108, 133139,  DOI: 10.1021/jp035653r
  29. 29

    Nonsolvated ZnMeCl is not available. ZnMeCl is usually produced in solution from ZnMe2 and ZnCl2, which already bear coordinated solvent. Mixing solutions of these compounds in a different solvent (e.g., toluene) is also unpractical because of the existence of THF dissociation equilibria, as reported in ref (14)), and because of the insolubility of ZnCl2 in noncoordinating solvents.

  30. 30
    Fischer, R.; Suxdorf, R.; Görls, H.; Westerhausen, M. Organometallics 2012, 31, 75797585,  DOI: 10.1021/om300880f
  31. 31
    Henriques, A. M.; Barbosa, A. G. H. J. Phys. Chem. A 2011, 115, 1225912270,  DOI: 10.1021/jp202762p
  32. 32
    Cordero, B.; Gómez, V.; Platero-Prats, A. E.; Revés, M.; Echeverría, J.; Cremades, E.; Barragán, F.; Álvarez, S. Dalton Trans. 2008, 28322838,  DOI: 10.1039/b801115j
  33. 33
    For instance, the two Zn–O distances in the tetrahedral cations of [Zn(OMe)(L)]·[Zn(OH)(L)]·2(BPh4) (L = cis,cis-1,3,5-tris[(E,E)-3-(2-furyl)acrylideneamino]cyclohexane) are 1.896 and 1.879 Å:Cronin, L.; Walton, P. H. Chem. Commun. 2003, 15721573,  DOI: 10.1039/B302895J
  34. 34
    ZnCl2(dioxane)2 shows a terminal and a bridging THF on a pentacoordinated Zn with bpt coordination:Boardman, A.; Small, R. W. H.; Worrall, I. J. Acta Crystallogr., Sect. C: Cryst. Struct. Commun. 1983, 39, 10051007,  DOI: 10.1107/S0108270183007179
  35. 35
    Fey, N.; Ridgway, B. M.; Jover, J.; McMullin, C. L.; Harvey, J. N. Dalton Trans. 2011, 40, 1118411191,  DOI: 10.1039/c1dt10909j
  36. 36
    (a) Vidossich, P.; Lledós, A.; Ujaque, G. Acc. Chem. Res. 2016, 49, 12711278,  DOI: 10.1021/acs.accounts.6b00054
    (b) Vidossich, P.; Ujaque, G.; Lledós, A. Chem. Commun. 2014, 50, 661663,  DOI: 10.1039/C3CC47404F
  37. 37
    Peltzer, R. M.; Eisenstein, O.; Nova, A.; Cascella, M. J. Phys. Chem. B , in press (DOI: 2017 DOI: 10.1021/acs.jpcb.7b02716 ).
  38. 38
    Literally millions of CPU hours.
  39. 39

    The two approximations shown in Scheme 2 include the correction for THF as bulk solvent with a polarizable continuum medium with dielectric constant ε = 7.4257 (SMD model), but differ in the number and state of the explicit THF molecules that participate in the equilibrium.

  40. 40
    (a) Pliego, J. R., Jr.; Riveros, J. M. J. Phys. Chem. A 2002, 106, 74347439,  DOI: 10.1021/jp025928n
    (b) Kelly, C. P.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. A 2006, 110, 24932499,  DOI: 10.1021/jp055336f
    (c) Marenich, A. V.; Ding, W.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. Lett. 2012, 3, 14371442,  DOI: 10.1021/jz300416r
  41. 41
    (a) Bryantsev, V. S.; Diallo, M. S.; Goddard, W. A., III. J. Phys. Chem. B 2008, 112, 97099719,  DOI: 10.1021/jp802665d
    (b) Ho, J.; Ertem, M. Z. J. Phys. Chem. B 2016, 120, 13191329,  DOI: 10.1021/acs.jpcb.6b00164
  42. 42

    We are aware that, in transition metals, the trans effect is reasoned, taking into account that trans ligands are sharing common d-orbitals, whereas in Zn, the 3d orbital is full, highly stabilized, and does not participate in bonding. However, the ultimate destabilizing reason of these trans phenomena is that electron densities prefer to avoid, as much as they can, high electronic repulsions when sharing a common space and this preference holds also as discussed for ZnMe2.

  43. 43

    The continuum model only provides Gibbs energy of solvation, not enthalpies or entropies. Despite the frequent mistake of taking the energy in solution to be equivalent to enthalpy, comparison of this parameter with experimental ΔH values cannot be done. See ref (3), ref (6b) and:

    (a) Braga, A. A. C.; Ujaque, G.; Maseras, F. Organometallics 2006, 25, 36473658,  DOI: 10.1021/om060380i
  44. 44

    Since the Pd complex does not change by the presence of THF, the Gibbs energy differences come only from the Zn molecules considered.

  45. 45

    As a matter of fact, this exact reaction has never been measured experimentally, but there are closely related calculations available in the papers in ref (8) and in the following:

    (a) Ribagnac, P.; Blug, M.; Villa-Uribe, J.; Le Goff, X.-F.; Gosmini, C.; Mézailles, N. Chem.—Eur. J. 2011, 17, 1438914393,  DOI: 10.1002/chem.201102369
    (b) Nicolas, E.; Ohleier, A.; D’Accriscio, F.; Pécharman, A.-F.; Demange, M.; Ribagnac, P.; Ballester, J.; Gosmini, C.; Mézailles, N. Chem.—Eur. J. 2015, 21, 76907694,  DOI: 10.1002/chem.201500192
  46. 46

    Since the purpose of the study is to compare the results with or without explicit solvent molecules, the ΔG or ΔG0 values shown in Figure 3 do not contain any procedure to modify them by using commonly accepted approximations.

  47. 47

    It is very common that PdR2L2 + [M]-Cl will be less stable than PdRClL2 + [M]-R. In catalysis, the subsequent irreversible C–C coupling will pull forward the transmetalation, in the counterthermodynamic sense.

  48. 48

    For instance, for an equilibrium of A + B = C + D, a value of ΔG0 = 2 kcal mol–1 is associated with an ∼100:1 concentration ratio of the components at the two sides, which practically precludes NMR observation of one of the sides of that uneven equilibrium.

  49. 49

    Equilibria can alternatively be studied by calorimetric methods but these afford ΔH0, not ΔG0.

  50. 50

    These were made with two explicit molecules of coordinated THF. See refs (8a) and (8b).

  51. 51
    (a) For instance, the very recently reported and X-ray characterized (phen)Ar2Pt–Zn(C6F5)2 adduct, containing a dative Pt–Zn bond can be formed in toluene and crystallized from that solution. See: Liberman-Martin, A. L.; Levine, D. S.; Ziegler, M. S.; Bergman, R. G.; Tilley, T. D. Chem. Commun. 2016, 52, 70397042. DOI: 10.1039/C6CC02433E
    (b) More recently, the X-ray characterized (bhq)2Pd-Zn(C6F5)2, containing a dative Pd–Zn bond, has been captured and crystallized from CH2Cl2. The very labile Pd–Zn bond undergoes chemical exchange at −100 °C. See: Oeschger, R. J.; Chen, P. Organometallics in press (DOI: 2017 DOI: 10.1021/acs.organomet.7b00113 ).
    (c) Similar intermediates [Pd]–ZnMe2, with a much less electrophilic Zn center, found in silico in our calculations without explicit THF (refs (8c) and (8d)), are plausible molecules in cyclohexane and perhaps in toluene or CH2Cl2, but not in THF, where coordination of THF would split the extremely weak Pd–ZnMe2 dative bond.
  52. 52
    For an interesting example (a very simple ligand coordination reaction in Rh) of enormous differences in ΔG, depending on the functional used, see:Sieffert, M.; Bühl, M. Inorg. Chem. 2009, 48, 46224624,  DOI: 10.1021/ic900347e

Cited By

ARTICLE SECTIONS
Jump To

This article is cited by 29 publications.

  1. Estefanía Gioria, Juan del Pozo, Agustí Lledós, Pablo Espinet. Understanding the Use of Phosphine-(EWO) Ligands in Negishi Cross-Coupling: Experimental and Density Functional Theory Mechanistic Study. Organometallics 2021, 40 (14) , 2272-2282. https://doi.org/10.1021/acs.organomet.1c00001
  2. Baojian Xiong, Ting Wang, Haotian Sun, Yue Li, Søren Kramer, Gui-Juan Cheng, Zhong Lian. Nickel-Catalyzed Cross-Electrophile Coupling Reactions for the Synthesis of gem-Difluorovinyl Arenes. ACS Catalysis 2020, 10 (22) , 13616-13623. https://doi.org/10.1021/acscatal.0c03993
  3. Fedor M. Miloserdov, Nasir A. Rajabi, John P. Lowe, Mary F. Mahon, Stuart A. Macgregor, Michael K. Whittlesey. Zn-Promoted C–H Reductive Elimination and H2 Activation via a Dual Unsaturated Heterobimetallic Ru–Zn Intermediate. Journal of the American Chemical Society 2020, 142 (13) , 6340-6349. https://doi.org/10.1021/jacs.0c01062
  4. Andrea Darù, Xile Hu, Jeremy N. Harvey. Iron-Catalyzed Reductive Coupling of Alkyl Iodides with Alkynes To Yield cis-Olefins: Mechanistic Insights from Computation. ACS Omega 2020, 5 (3) , 1586-1594. https://doi.org/10.1021/acsomega.9b03578
  5. Mikhail V. Polynski, Valentine P. Ananikov. Modeling Key Pathways Proposed for the Formation and Evolution of “Cocktail”-Type Systems in Pd-Catalyzed Reactions Involving ArX Reagents. ACS Catalysis 2019, 9 (5) , 3991-4005. https://doi.org/10.1021/acscatal.9b00207
  6. Daniel H. Yi, Chang-Yong Nam, Gregory Doerk, Charles T. Black, Robert B. Grubbs. Infiltration Synthesis of Diverse Metal Oxide Nanostructures from Epoxidized Diene–Styrene Block Copolymer Templates. ACS Applied Polymer Materials 2019, 1 (4) , 672-683. https://doi.org/10.1021/acsapm.8b00138
  7. Konstantinos D. Vogiatzis, Mikhail V. Polynski, Justin K. Kirkland, Jacob Townsend, Ali Hashemi, Chong Liu, Evgeny A. Pidko. Computational Approach to Molecular Catalysis by 3d Transition Metals: Challenges and Opportunities. Chemical Reviews 2019, 119 (4) , 2453-2523. https://doi.org/10.1021/acs.chemrev.8b00361
  8. Maria Besora, Pietro Vidossich, Agustí Lledós, Gregori Ujaque, and Feliu Maseras . Calculation of Reaction Free Energies in Solution: A Comparison of Current Approaches. The Journal of Physical Chemistry A 2018, 122 (5) , 1392-1399. https://doi.org/10.1021/acs.jpca.7b11580
  9. Michael Busch, Matthew D. Wodrich, Clémence Corminboeuf. A Generalized Picture of C–C Cross-Coupling. ACS Catalysis 2017, 7 (9) , 5643-5653. https://doi.org/10.1021/acscatal.7b01415
  10. Ashok Mandi, Nandagopal Bar, Debrupa Biswas, Arindam Ray, Keya Ghosh, Dhrubajyoti Mondal, Gourab Kanti Das, Pranesh Chowdhury. Rhodamine 6G salicylaldehyde hydrazone-Zn complex: Synthesis, characterization, photochromism, understanding, and application. Journal of Luminescence 2024, 271 , 120611. https://doi.org/10.1016/j.jlumin.2024.120611
  11. Manting Mu, Marconi N. Peñas-Defrutos, Max García-Melchor. Challenges in the computational modelling of bimetallic C–H activation processes. 2024https://doi.org/10.1016/bs.adomc.2024.03.001
  12. Jordan Rio, Lionel Perrin, Pierre‐Adrien Payard. Structure–Reactivity Relationship of Organozinc and Organozincate Reagents: Key Elements towards Molecular Understanding. European Journal of Organic Chemistry 2022, 2022 (44) https://doi.org/10.1002/ejoc.202200906
  13. Aude Salamé, Jordan Rio, Ilaria Ciofini, Lionel Perrin, Laurence Grimaud, Pierre-Adrien Payard. Copper-Catalyzed Homocoupling of Boronic Acids: A Focus on B-to-Cu and Cu-to-Cu Transmetalations. Molecules 2022, 27 (21) , 7517. https://doi.org/10.3390/molecules27217517
  14. Yaju Wang, Zhiming Zhou, Chao Wang, Leihong Zhao, Qineng Xia. Hydrogenolysis of glycerol over TiO2-supported Pt-WOx catalysts: Effects of the TiO2 crystal phase and WOx loading. Frontiers in Chemistry 2022, 10 https://doi.org/10.3389/fchem.2022.1004925
  15. Gantulga Norjmaa, Gregori Ujaque, Agustí Lledós. Beyond Continuum Solvent Models in Computational Homogeneous Catalysis. Topics in Catalysis 2022, 65 (1-4) , 118-140. https://doi.org/10.1007/s11244-021-01520-2
  16. Desiré Carrasco, Juan A. Casares. Experimental study of the [ZnCl2(THF)2] catalyzed cis/trans-isomerization of [Pd(C6Cl2F3)Me(PPh3)2] and of the transmetalation of trans-[PdCl(C6Cl2F3)(PPh3)2] with [ZnMeCl(THF)2]. Inorganica Chimica Acta 2021, 517 , 120206. https://doi.org/10.1016/j.ica.2020.120206
  17. Amir H. Hoveyda, Yuebiao Zhou, Ying Shi, M. Kevin Brown, Hao Wu, Sebastian Torker. Sulfonate N‐Heterocyclic Carbene–Copper Complexes: Uniquely Effective Catalysts for Enantioselective Synthesis of C−C, C−B, C−H, and C−Si Bonds. Angewandte Chemie International Edition 2020, 59 (48) , 21304-21359. https://doi.org/10.1002/anie.202003755
  18. Amir H. Hoveyda, Yuebiao Zhou, Ying Shi, M. Kevin Brown, Hao Wu, Sebastian Torker. Sulfonate N‐Heterocyclic Carbene–Copper Complexes: Uniquely Effective Catalysts for Enantioselective Synthesis of C−C, C−B, C−H, and C−Si Bonds. Angewandte Chemie 2020, 132 (48) , 21488-21543. https://doi.org/10.1002/ange.202003755
  19. Odile Eisenstein, Gregori Ujaque, Agustí Lledós. What Makes a Good (Computed) Energy Profile?. 2020, 1-38. https://doi.org/10.1007/3418_2020_57
  20. Eric D. Slack, Raina Seupel, Donald H. Aue, Gerhard Bringmann, Bruce H. Lipshutz. Atroposelective Total Synthesis of the Fourfold ortho ‐Substituted Naphthyltetrahydroisoquinoline Biaryl O , N ‐Dimethylhamatine. Chemistry – A European Journal 2019, 25 (62) , 14237-14245. https://doi.org/10.1002/chem.201903832
  21. Matthew A. Lynn, John R. Miecznikowski, Jerry P. Jasinski, Manpreet Kaur, Brandon Q. Mercado, Eric Reinheimer, Emilse M. Almanza, Rami M. Kharbouch, Michael R. Smith, Samantha E. Zygmont, Nicole F. Flaherty, Amber C. Smith. Copper(I) SNS pincer complexes: Impact of ligand design and solvent coordination on conformer interconversion from spectroscopic and computational studies. Inorganica Chimica Acta 2019, 495 , 118996. https://doi.org/10.1016/j.ica.2019.118996
  22. Mikhail V. Polynski, Evgeny A. Pidko. Intermetallic species in the Negishi coupling and their involvement in inhibition pathways. Catalysis Science & Technology 2019, 9 (17) , 4561-4572. https://doi.org/10.1039/C9CY00752K
  23. James Sherwood, James H. Clark, Ian J. S. Fairlamb, John M. Slattery. Solvent effects in palladium catalysed cross-coupling reactions. Green Chemistry 2019, 21 (9) , 2164-2213. https://doi.org/10.1039/C9GC00617F
  24. Hugo Lingua, Nejib Dwadnia, Didier Siri, Michèle P. Bertrand, Laurence Feray. Reactivity of benzylidene and alkylidenemalonates in radical addition mediated with dialkylzincs – An intriguing story. Tetrahedron 2018, 74 (52) , 7507-7515. https://doi.org/10.1016/j.tet.2018.11.029
  25. Yang Liu, Alessandro Cerveri, Assunta De Nisi, Magda Monari, Olalla Nieto Faza, Carlos Silva Lopez, Marco Bandini. Nickel catalyzed regio- and stereoselective arylation and methylation of allenamides via coupling reactions. An experimental and computational study. Organic Chemistry Frontiers 2018, 5 (22) , 3231-3239. https://doi.org/10.1039/C8QO00729B
  26. Tomaz Neves-Garcia, Andrea Vélez, Jesús M. Martínez-Ilarduya, Pablo Espinet. Highly enantioselective addition of dimethylzinc to fluorinated alkyl ketones, and the mechanism behind it. Chemical Communications 2018, 54 (83) , 11809-11812. https://doi.org/10.1039/C8CC06358C
  27. Luka Rejc, Vanessa Gómez-Vallejo, Jesús Alcázar, Nerea Alonso, José Ignacio Andrés, Ana Arrieta, Fernando P. Cossío, Jordi Llop. Negishi coupling reactions with [ 11 C]CH 3 I: a versatile method for efficient 11 C–C bond formation. Chemical Communications 2018, 54 (35) , 4398-4401. https://doi.org/10.1039/C8CC01540F
  28. Amalia I. Poblador Bahamonde, Stéphanie Halbert. Computational Study of the Cu‐Free Allylic Alkylation Mechanism with Grignard Reagents: Role of the NHC Ligand. European Journal of Organic Chemistry 2017, 2017 (39) , 5935-5941. https://doi.org/10.1002/ejoc.201701010
  29. Jesús Campos, Ainara Nova, Eugene L. Kolychev, Simon Aldridge. A Combined Experimental/Computational Study of the Mechanism of a Palladium‐Catalyzed Bora‐Negishi Reaction. Chemistry – A European Journal 2017, 23 (51) , 12655-12667. https://doi.org/10.1002/chem.201702703
  • Abstract

    Figure 1

    Figure 1. (a) IR spectra of mixtures of ZnMe2 and THF in ratios of 1:1 (black), 1:2 (red), and 1:3 (blue); (b) IR spectra of mixtures of ZnMe2 (blue), ZnMeCl (red) or ZnCl2, and THF. The bands of uncoordinated THF (1067 and 908 cm–1) are observed in all of the spectra.

    Scheme 1

    Scheme 1. THF Coordination Equilibria, As Suggested by IR Observations

    Figure 2

    Figure 2. Representation of the DFT computed structures. [Legend: red, O; yellow, Zn; gray, C; and green, Cl.]

    Scheme 2

    Scheme 2. Coordination Equilibria of THF Plus ZnMe2: (a) Considering the Solvent as Individual Molecules and (b) Considering the Solvent as a Cluster of n Molecules (n = 2–4)

    Figure 3

    Figure 3. Gibbs energy profile of the Negishi transmetalation step (kcal mol–1): (top) without explicit THF and (bottom) with two coordinating THF molecules (M06 optimizations in THF described by the SMD continuum model).

  • References

    ARTICLE SECTIONS
    Jump To

    This article references 52 other publications.

    1. 1
      Plata, R. E.; Singleton, D. A. J. Am. Chem. Soc. 2015, 137, 38113826,  DOI: 10.1021/ja5111392
    2. 2
      Winter, A. A. Nat. Chem. 2015, 7, 473475,  DOI: 10.1038/nchem.2267
    3. 3
      Harvey, J. N. Faraday Discuss. 2010, 145, 487505,  DOI: 10.1039/B907340J
    4. 4
      Dunsford, J. J.; Clark, E. R.; Ingleson, M. J. Angew. Chem., Int. Ed. 2015, 54, 56885692,  DOI: 10.1002/anie.201411403
    5. 5

      In an oversimplified view, solvent coordination seems enthalpically favorable but entropically unfavorable, but this relationship is not that simple.

    6. 6
      (a) Cooper, J.; Ziegler, T. Inorg. Chem. 2002, 41, 66146622,  DOI: 10.1021/ic020294k
      (b) Dub, P. A.; Poli, R. J. Mol. Catal. A: Chem. 2010, 324, 8996,  DOI: 10.1016/j.molcata.2010.03.003
    7. 7

      A recent study of the difficulties associated with calculations has made very clear how difficult it is to decide which functional is reasonable for a given calculation or how to address entropy contributions. See ref (1).

    8. 8
      (a) Fuentes, B.; García-Melchor, M.; Lledós, A.; Maseras, F.; Casares, J. A.; Ujaque, G.; Espinet, P. Chem. - Eur. J. 2010, 16, 85968599,  DOI: 10.1002/chem.201001332
      (b) García-Melchor, M.; Fuentes, B.; Lledós, A.; Casares, J. A.; Ujaque, G.; Espinet, P. J. Am. Chem. Soc. 2011, 133, 1351913526,  DOI: 10.1021/ja204256x
      (c) delPozo, J.; Gioria, E.; Casares, J. A.; Álvarez, R.; Espinet, P. Organometallics 2015, 34, 31203128,  DOI: 10.1021/acs.organomet.5b00329
      (d) del Pozo, J.; Salas, G.; Álvarez, R.; Casares, J. A.; Espinet, P. Organometallics 2016, 35, 36043611,  DOI: 10.1021/acs.organomet.6b00660
    9. 9
      von Frankland, E. Ann. Chem. Pharm. 1849, 71, 171213,  DOI: 10.1002/jlac.18490710205
    10. 10
      (a) Seyferth, D. Organometallics 2001, 20, 29402955,  DOI: 10.1021/om010439f
      (b) Seyferth, D. Organometallics 2004, 23, 11721172,  DOI: 10.1021/om0499557
    11. 11
      Haaland, A.; Green, J. C.; McGrady, G. S.; Downs, A. J.; Gullo, E.; Lyall, M. J.; Timberlake, J.; Tutukin, A. V.; Volden, H. W.; Østby, K.-A. Dalton Trans. 2003, 43564366,  DOI: 10.1039/B306840B
    12. 12
      Bacsa, J.; Hanke, F.; Hindley, S.; Odedra, R.; Darling, G. R.; Jones, A. C.; Steiner, A. Angew. Chem., Int. Ed. 2011, 50, 1168511687,  DOI: 10.1002/anie.201105099
    13. 13
      Antes, I.; Frenking, G. Organometallics 1995, 14, 42634268,  DOI: 10.1021/om00009a032
    14. 14
      Thiele, K. H. Z. Anorg. Allg. Chem. 1962, 319, 183195,  DOI: 10.1002/zaac.19623190309
    15. 15
      Bochmann has reported the formation in toluene of Zn(C6F5)2·(toluene) from ZnMe2 and B(C6F5)3. It is thought that the toluene is coordinated to Zn. These interactions could explain the dissociation of ZnMe2(THF)2 in benzene observed by Thiele. See:Walker, D. A.; Woodman, T. J.; Hughes, D. L.; Bochmann, M. Organometallics 2001, 20, 37723776,  DOI: 10.1021/om0103557
    16. 16
      Weidenbruch, M.; Herrndorf, M.; Schäfer, A.; Pohl, S.; Saak, W. J. Organomet. Chem. 1989, 361, 139145,  DOI: 10.1016/0022-328X(89)85378-1
    17. 17

      The high electrophilicity of Zn(C6F5)2 is shown by the isolation of Zn(C6F5)2·(toluene) and Zn(C6F5)2·(C6Me6), obtained from toluene solutions, and is believed to have η2-coordinated arene.

    18. 18
      Dashti, A.; Niediek, K.; Werner, B.; Neumiiller, B. Z. Anorg. Allg. Chem. 1997, 623, 394402,  DOI: 10.1002/zaac.19976230163
    19. 19
      Guerrero, A.; Hughes, D. L.; Bochmann, M. Organometallics 2006, 25, 15251527,  DOI: 10.1021/om051043x
    20. 20
      Evans, D. F.; Wharf, I. J. Chem. Soc. A 1968, 783787,  DOI: 10.1039/j19680000783
    21. 21

      The activity concept cannot be rigorously applied under these conditions, but the simple numerical calculation for the number of moles in 1 L of THF would afford a concentration of 12.33 M.

    22. 22

      The labels “asym” and “sym” are assigned to describe the symmetry within the THF ring. Strictly speaking, the vibrational modes involve also the O–Zn bond in the case of the Zn complexes.

    23. 23

      Note that, in the spectrum with a lower concentration of THF (the black trace in Figure 1a), the bands corresponding to some noncoordinated THF are still observable, as a shoulder for the asymmetric stretching, and as a clear band at 908 cm–1 for the ν(C–O)sym. Thus the composition of this mixture contains some noncoordinated THF, tetrahedral ZnMe2(THF)2 and, necessarily, the remaining Zn as non-observable linear ZnMe2. The same applies to the dissociation observed for ZnMe2(THF)2 (the red trace in Figure 1a).

    24. 24

      The concentrations of the three samples recorded in the infrared (IR) spectroscopy analysis are very different, because of the low solubility of ZnCl2 in THF. As a result, the intensity of the noncoordinated THF band is also different for each sample.

    25. 25
      Guillén, M. D.; Gutiérrez Losa, C. J. Chem. Thermodyn. 1978, 10, 567576,  DOI: 10.1016/0021-9614(78)90045-9
    26. 26
      Letcher, T. M.; Domanska, U. J. Chem. Thermodyn. 1994, 26, 7584,  DOI: 10.1006/jcht.1994.1022
    27. 27
      Smith, M. B.; Becker, W. E. Tetrahedron 1967, 23 (11), 42154227,  DOI: 10.1016/S0040-4020(01)88819-0
    28. 28
      Tammiku-Taul, J.; Burk, P.; Tuulmets, A. J. Phys. Chem. A 2004, 108, 133139,  DOI: 10.1021/jp035653r
    29. 29

      Nonsolvated ZnMeCl is not available. ZnMeCl is usually produced in solution from ZnMe2 and ZnCl2, which already bear coordinated solvent. Mixing solutions of these compounds in a different solvent (e.g., toluene) is also unpractical because of the existence of THF dissociation equilibria, as reported in ref (14)), and because of the insolubility of ZnCl2 in noncoordinating solvents.

    30. 30
      Fischer, R.; Suxdorf, R.; Görls, H.; Westerhausen, M. Organometallics 2012, 31, 75797585,  DOI: 10.1021/om300880f
    31. 31
      Henriques, A. M.; Barbosa, A. G. H. J. Phys. Chem. A 2011, 115, 1225912270,  DOI: 10.1021/jp202762p
    32. 32
      Cordero, B.; Gómez, V.; Platero-Prats, A. E.; Revés, M.; Echeverría, J.; Cremades, E.; Barragán, F.; Álvarez, S. Dalton Trans. 2008, 28322838,  DOI: 10.1039/b801115j
    33. 33
      For instance, the two Zn–O distances in the tetrahedral cations of [Zn(OMe)(L)]·[Zn(OH)(L)]·2(BPh4) (L = cis,cis-1,3,5-tris[(E,E)-3-(2-furyl)acrylideneamino]cyclohexane) are 1.896 and 1.879 Å:Cronin, L.; Walton, P. H. Chem. Commun. 2003, 15721573,  DOI: 10.1039/B302895J
    34. 34
      ZnCl2(dioxane)2 shows a terminal and a bridging THF on a pentacoordinated Zn with bpt coordination:Boardman, A.; Small, R. W. H.; Worrall, I. J. Acta Crystallogr., Sect. C: Cryst. Struct. Commun. 1983, 39, 10051007,  DOI: 10.1107/S0108270183007179
    35. 35
      Fey, N.; Ridgway, B. M.; Jover, J.; McMullin, C. L.; Harvey, J. N. Dalton Trans. 2011, 40, 1118411191,  DOI: 10.1039/c1dt10909j
    36. 36
      (a) Vidossich, P.; Lledós, A.; Ujaque, G. Acc. Chem. Res. 2016, 49, 12711278,  DOI: 10.1021/acs.accounts.6b00054
      (b) Vidossich, P.; Ujaque, G.; Lledós, A. Chem. Commun. 2014, 50, 661663,  DOI: 10.1039/C3CC47404F
    37. 37
      Peltzer, R. M.; Eisenstein, O.; Nova, A.; Cascella, M. J. Phys. Chem. B , in press (DOI: 2017 DOI: 10.1021/acs.jpcb.7b02716 ).
    38. 38
      Literally millions of CPU hours.
    39. 39

      The two approximations shown in Scheme 2 include the correction for THF as bulk solvent with a polarizable continuum medium with dielectric constant ε = 7.4257 (SMD model), but differ in the number and state of the explicit THF molecules that participate in the equilibrium.

    40. 40
      (a) Pliego, J. R., Jr.; Riveros, J. M. J. Phys. Chem. A 2002, 106, 74347439,  DOI: 10.1021/jp025928n
      (b) Kelly, C. P.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. A 2006, 110, 24932499,  DOI: 10.1021/jp055336f
      (c) Marenich, A. V.; Ding, W.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. Lett. 2012, 3, 14371442,  DOI: 10.1021/jz300416r
    41. 41
      (a) Bryantsev, V. S.; Diallo, M. S.; Goddard, W. A., III. J. Phys. Chem. B 2008, 112, 97099719,  DOI: 10.1021/jp802665d
      (b) Ho, J.; Ertem, M. Z. J. Phys. Chem. B 2016, 120, 13191329,  DOI: 10.1021/acs.jpcb.6b00164
    42. 42

      We are aware that, in transition metals, the trans effect is reasoned, taking into account that trans ligands are sharing common d-orbitals, whereas in Zn, the 3d orbital is full, highly stabilized, and does not participate in bonding. However, the ultimate destabilizing reason of these trans phenomena is that electron densities prefer to avoid, as much as they can, high electronic repulsions when sharing a common space and this preference holds also as discussed for ZnMe2.

    43. 43

      The continuum model only provides Gibbs energy of solvation, not enthalpies or entropies. Despite the frequent mistake of taking the energy in solution to be equivalent to enthalpy, comparison of this parameter with experimental ΔH values cannot be done. See ref (3), ref (6b) and:

      (a) Braga, A. A. C.; Ujaque, G.; Maseras, F. Organometallics 2006, 25, 36473658,  DOI: 10.1021/om060380i
    44. 44

      Since the Pd complex does not change by the presence of THF, the Gibbs energy differences come only from the Zn molecules considered.

    45. 45

      As a matter of fact, this exact reaction has never been measured experimentally, but there are closely related calculations available in the papers in ref (8) and in the following:

      (a) Ribagnac, P.; Blug, M.; Villa-Uribe, J.; Le Goff, X.-F.; Gosmini, C.; Mézailles, N. Chem.—Eur. J. 2011, 17, 1438914393,  DOI: 10.1002/chem.201102369
      (b) Nicolas, E.; Ohleier, A.; D’Accriscio, F.; Pécharman, A.-F.; Demange, M.; Ribagnac, P.; Ballester, J.; Gosmini, C.; Mézailles, N. Chem.—Eur. J. 2015, 21, 76907694,  DOI: 10.1002/chem.201500192
    46. 46

      Since the purpose of the study is to compare the results with or without explicit solvent molecules, the ΔG or ΔG0 values shown in Figure 3 do not contain any procedure to modify them by using commonly accepted approximations.

    47. 47

      It is very common that PdR2L2 + [M]-Cl will be less stable than PdRClL2 + [M]-R. In catalysis, the subsequent irreversible C–C coupling will pull forward the transmetalation, in the counterthermodynamic sense.

    48. 48

      For instance, for an equilibrium of A + B = C + D, a value of ΔG0 = 2 kcal mol–1 is associated with an ∼100:1 concentration ratio of the components at the two sides, which practically precludes NMR observation of one of the sides of that uneven equilibrium.

    49. 49

      Equilibria can alternatively be studied by calorimetric methods but these afford ΔH0, not ΔG0.

    50. 50

      These were made with two explicit molecules of coordinated THF. See refs (8a) and (8b).

    51. 51
      (a) For instance, the very recently reported and X-ray characterized (phen)Ar2Pt–Zn(C6F5)2 adduct, containing a dative Pt–Zn bond can be formed in toluene and crystallized from that solution. See: Liberman-Martin, A. L.; Levine, D. S.; Ziegler, M. S.; Bergman, R. G.; Tilley, T. D. Chem. Commun. 2016, 52, 70397042. DOI: 10.1039/C6CC02433E
      (b) More recently, the X-ray characterized (bhq)2Pd-Zn(C6F5)2, containing a dative Pd–Zn bond, has been captured and crystallized from CH2Cl2. The very labile Pd–Zn bond undergoes chemical exchange at −100 °C. See: Oeschger, R. J.; Chen, P. Organometallics in press (DOI: 2017 DOI: 10.1021/acs.organomet.7b00113 ).
      (c) Similar intermediates [Pd]–ZnMe2, with a much less electrophilic Zn center, found in silico in our calculations without explicit THF (refs (8c) and (8d)), are plausible molecules in cyclohexane and perhaps in toluene or CH2Cl2, but not in THF, where coordination of THF would split the extremely weak Pd–ZnMe2 dative bond.
    52. 52
      For an interesting example (a very simple ligand coordination reaction in Rh) of enormous differences in ΔG, depending on the functional used, see:Sieffert, M.; Bühl, M. Inorg. Chem. 2009, 48, 46224624,  DOI: 10.1021/ic900347e
  • Supporting Information

    Supporting Information

    ARTICLE SECTIONS
    Jump To

    The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acscatal.6b03636.

    • Experimental methods and data, computational methods and data, and associated references (PDF)


    Terms & Conditions

    Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.

Pair your accounts.

Export articles to Mendeley

Get article recommendations from ACS based on references in your Mendeley library.

Pair your accounts.

Export articles to Mendeley

Get article recommendations from ACS based on references in your Mendeley library.

You’ve supercharged your research process with ACS and Mendeley!

STEP 1:
Click to create an ACS ID

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

MENDELEY PAIRING EXPIRED
Your Mendeley pairing has expired. Please reconnect