Skip to main content
JACS Au logoLink to JACS Au
. 2025 Nov 1;5(11):5656–5664. doi: 10.1021/jacsau.5c01159

Crystallographic, Electronic Structure, and Computational Studies of PHOX–Ni Aryne Complexes: Origins of Regioselectivity in Metal-Bound Aryne Synthesis and Difunctionalization

Alexander Umanzor , Nicholas A Garcia , Kevin P Quirion , Alex Lovstedt , Peng Liu ‡,*, Courtney C Roberts †,*
PMCID: PMC12648280  PMID: 41311958

Abstract

Late transition metal aryne complexes are stable, isolable counterparts to free aryne intermediates. However, their utility has largely been limited since the Aryne Distortion Model (ADM) cannot be applied to substituted aryne complex reactivity, leading to nonselective reactions. Our group recently reported the first regioselective synthesis and difunctionalization of a CyPHOX–Ni o-methoxybenzyne complex. However, to increase the utility of these complexes in synthesis, their electronic structure, reactivity, and the impact of aryne substituents on selectivity must be understood. Herein, we report the first comprehensive experimental electronic structure study of aryne complexes, which has been carried out via UV/vis spectroscopy and cyclic voltammetry (CV) with an array of o-substituted arynes. CyPHOX–Ni aryne complexes exhibit a metal-to-ligand charge transfer (MLCT), and this transition as well as their oxidation potentials trend with Hammett parameters for the aryne substituents. To gain further insight into the origins of regioselectivity in CyPHOX–Ni aryne complex formation and difunctionalization, a combination of single-crystal X-ray crystallographic and density functional theory (DFT) structural studies were carried out. Our findings lead us to propose a Metal Aryne Reactivity/Selectivity (MAR/S) Model, which shows that CyPHOX–Ni aryne binding selectivity is governed by a combination of sterics and aryne distortion, whereas selectivity in functionalizations is directed by the phosphine trans influence.

Keywords: nickel, PHOX ligands, arynes, regioselectivity, selectivity models


graphic file with name au5c01159_0011.jpg


graphic file with name au5c01159_0009.jpg

Introduction

Arynes are reactive intermediates consisting of an aromatic ring with a triple bond that have been used in the synthesis of over 75 natural products. , However, one limitation of unsymmetrically substituted free arynes is their substrate-controlled regioselectivity in reactions. The utility of o-substituted free aryne intermediates in synthesis has been expanded by the Aryne Distortion Model, developed by Garg and Houk (Figure a). In this pioneering model, it was shown that inductively electron-withdrawing substituents in o-substituted benzyne intermediates distort the internal aryne bond angles such that the meta-position is distorted toward linearity. This site has a larger internal angle and is the preferred site of nucleophilic functionalization. Greater internal angle differences are directly correlated with greater regioselectivities, allowing for more selective difunctionalizations (Figure a).

1.

1

(a) Aryne Distortion Model for free arynes. (b) This work: Spectroscopic, crystallographic, and computational structural studies lead us to propose the Metal Aryne Reactivity/Selectivity (MAR/S) Model.

Late transition metal aryne complexes are isolable counterparts to free arynes which have been shown to exhibit umpolung reactivity as nucleophiles. , They are invoked as intermediates in catalytic reactions and are capable of a wide array of reactivity. This makes them potentially useful synthons for rapidly accessing a variety of arene substitution patterns. However, unsymmetrically substituted aryne complexes often generate equimolar mixtures of regioisomeric products. For example, Hosoya and co-workers reported a difunctionalization of a (Et3P)2Ni o-methoxybenzyne complex, which gave a product ratio of 51:49. This is in stark contrast to the corresponding free aryne, which exclusively generates one regioisomer. Therefore, for aryne complexes to gain utility in synthesis, a platform must be developed to establish and predict their selectivity in reactions. The fact that they can be isolated, characterized, and reacted in a controlled stoichiometric manner allows for ease of study. Once this model is understood, selectivity can be further improved and implemented in Ni catalysis.

Thus, we sought to establish a system by which late transition metal aryne complexes could be synthesized and functionalized selectively. Our group recently reported the regioselective synthesis and difunctionalization of a CyPHOX–Ni o-methoxybenzyne, which showed that the concept of ligand control could be applied to influence o-substituted aryne complex selectivity. We now wanted to probe the electronic structure of an isostructural series of CyPHOX–Ni aryne complexes with varying aryne substituents to better understand the impact of substituent effects in combination with ligand effects on selectivity and reactivity. This was accomplished by studying these complexes via UV/vis spectroscopy, cyclic voltammetry, natural population analysis (NPA) calculations, and their correlations to linear free energy relationships. In this report, we propose the Metal Aryne Reactivity/Selectivity (MAR/S) Model, which has been developed using a combination of spectroscopy, crystallography, density functional theory (DFT), and reactivity studies (Figure b).

Results and Discussion

Synthesis and Crystallographic Studies of CyPHOX–Ni Aryne Complexes

Similarly to our previous studies, the synthesis of the aryne complexes 4-R was accomplished via a three-step route involving (i) oxidative addition of the aryne precursor, (ii) CyPHOX ligand exchange, and (iii) transmetalation using NaO t Bu as an activator to form the aryne. All novel PPh3 σ-aryls 2-R were isolable and characterized crystallographically via single-crystal X-ray diffraction (XRD). CyPHOX σ-aryls 3-CN, 3-Me, 3-Cl and 3-F were found to be isolable after steps (i) and (ii) and were characterized via XRD. As observed with the o-methoxy oxidative addition complex 3-OMe, the unsubstituted σ-aryl (3-H) was not isolable and thus was subjected to an in situ ligand exchange followed by transmetalation to form 4-H.

With this scope of electronically varied arynes, the aryne complexes were observed in the solution state by 31P­{1H} NMR spectroscopy in regioisomeric ratios ranging from 67:33 to 88:12 (Figure b). In order to explain the relationship between our aryne regioisomers, 31P­{1H}–31P­{1H} NOESY experiments were employed. As was found with 4-OMe, experiments show that the substituted arynes 4-R all interconvert on the 2.5 s mixing time scale. This is indicative of slow interconversion between regioisomers. We hypothesized that a structural feature could correlate with the observed binding selectivities. Using the 4-R solid-state structures (Figure c), the most significant feature exhibited between the crystal structures was the variable C1–C2 aryne bond lengths. It should be noted that the substituent will impact the electronics of both termini of the aryne through both resonance and inductive effects through the conjugated system. The arynes surveyed cover a wide range of C1–C2 bond lengths. 4-Me (1.333(16) Å) is comparable to other late transition metal benzyne complexes reported in the literature (dcpe–Ni 1.332(6) Å, (Cy3P)2–Pd 1.324­(8) Å). , However, the aryne bond length of 4-F is significantly longer, at 1.365(9) Å. This approaches the double bond length of benzene (1.396 Å). The minor regioisomers were not observed universally across the substituted complexes 4-R in the solid state; they manifested as rotational disorder for 4-Me, 4-OMe, and 4-Cl, and were not observed for 4-CN and 4-F, the CyPHOX–Ni aryne complexes with the longest C1–C2 bond lengths. The Dewar-Chatt-Duncanson model has long been used to describe the bonding in metal π complexes. , Crystallographic bond length comparisons can provide insight into the degree of π backdonation to the aryne π* orbitals. Our series of CyPHOX–Ni aryne complexes 4-R exhibit a wide range of aryne C1–C2 bond lengths (1.329(6) to 1.365(9) Å). Since 4-CN and 4-F are thus closer to the metallacycloarene end of the spectrum, this increased degree of π backdonation to the aryne may induce a higher degree of rigidity that hinders aryne rotation. Examining other features of the crystal structures, the Ni–C2 distance is longer than the Ni–C1 distance across the series. This confirms that the trans influence of the P donor is stronger than that of the N donor. Consequently, C2 will likely be the more reactive site. Despite these observations, we were unable to find any trend that correlated with the aryne complex regioisomeric ratios. Analysis of the internal aryne angles (θC1 and θC2) also did not correlate with the regioisomeric ratios observed in the solution state despite this angle difference (Δθ) being a computationally derived predictor in the Garg/Houk Aryne Distortion Model. This is likely due to crystal packing effects that distort bond lengths and angles. ,

2.

2

(a) Representative 3-step synthetic route toward CyPHOX–Ni aryne complexes. (b) Scope of CyPHOX–Ni aryne complexes in this study. (c) Single crystal X-ray crystal structures and select structural features of CyPHOX–Ni aryne complexes in this study. Note: For crystal structures with Z′ > 1 and/or disorder, the reported bond lengths and angles are the average of those in the asymmetric unit. aref .

Electronic Structure Studies

Recently, our group reported that 5-membered Ni N-heteroaryne complexes can react as both nucleophiles and electrophiles. Therefore, it is important to understand the electronic structure of complexes 4-R in order to inform reactivity studies. Because electronic structure studies in the solid state can be limited due to crystal packing effects, as well as the fact that only a single regioisomer was observed in the solid state for multiple complexes in our scope (R = CN, F), these factors posed limitations on the ability to find trends between structural and electronic features. Thus, we turned to solution-state studies, which can provide added insight into complex systems. In our previous studies with a CyPHOX–Ni o-methoxybenzyne (4-OMe), a proposed weak metal-to-ligand charge transfer (MLCT) band was observed in the UV/vis spectrum. This MLCT band was also observed in all of our complexes 4-R. We then proceeded to study the aryne substituent effects on the MLCT (Figure a). It was found that the MLCT transition across CyPHOX–Ni arynes lie in the cyan-green region of the visible light spectrum (466 nm < λmax < 509 nm). It was found that the λmax for the MLCT is strongly correlated to the σm Hammett parameter (Figure b, R 2 = 0.99). Thus, the amount of electron density at the meta-position (C1) has the greatest effect on the energy required for this transition. Regarding the intensity of the transition, 4-CN (5950 M–1·cm–1 ) exhibited the strongest MLCT, while 4-OMe had the weakest extinction coefficient (ε = 457 M–1·cm–1). Interestingly, the extinction coefficient for the MLCT was found to be correlated with the σp+ Hammett parameter (Figure c, R 2 = 0.99). From this trend, it appears that substituents that withdraw electron density through both resonance and inductive effects from the ortho-position (C2) correlate to a more intense charge transfer. Given the degree of correlation with Hammett parameters for the respective aryne substituents, we hypothesize that the MLCT involves the transfer of an electron from the Ni center to a π* orbital on the aryne. The σp+ parameter accounts for stabilization of a cation at the benzylic position. Since this proposed electron transfer would leave the “benzylic” Ni center cationic and form an intermittent NiI species, this parameter most accurately describes the substituent effects on this transition. Additionally, due to the absence of d–d transitions across aryne complexes 4-R, these spectral features preliminarily suggest a Ni(0) oxidation state. However, further studies (vide infra) were needed in order to support this hypothesis.

3.

3

(a) UV/vis spectra of CyPHOX–Ni arynes 4-R in this study. Spectra were collected on 0.3 mM samples under Ar in THF at 298 K. (b) Trend between the maximum absorbance wavelength λmax for MLCT and the σm Hammett parameter for the respective aryne substituent in 4-R. (c) Trend between the extinction coefficient ε for MLCT and the σp+ Hammett parameter for the respective aryne substituent in complexes 4-R.

Intrigued by the photophysical properties of 4-R, we then moved on to study their redox properties. Like the previously studied 4-OMe, the voltammograms of all the arynes exhibited a single, irreversible oxidation event (E pa) and a single, quasi-reversible reduction (E pc) in Figure a. The difference in oxidation potentials across the full 4-R series was 470 mV and the difference in the reduction potentials was 160 mV. These differences in redox potentials highlight that the identity of the R group imposes significant effects on the electrochemical properties of complexes 4-R. A summary of all oxidation and reduction potentials for 4-R is found in Figure b. In contrast to the MLCT features, the oxidation and reduction potentials of these complexes were found to have disparate correlations to Hammett parameters. The σm parameter shows good correlation with the oxidation potential, suggesting that inductively electron-withdrawing groups lead to harsher oxidation potentials (Figure c, R 2 = 0.82). Conversely, the reduction potential exhibits a poor trend with both σp (SI Figure S124, R 2 = 0.43) and σp+ (SI Figure S125, R 2 = 0.19).

4.

4

(a) Representative voltammogram displaying full electrochemical window events at 100 mV/s for 4-R (4-H shown). CVs collected in 0.1 M [nPr4N]­[BAr4 F] electrolyte solutions in THF with 3 mM analyte under Ar. (b) Tabulated redox potentials for aryne complexes 4-R. aref . (c) Hammett parameter trend between oxidation potential E pa and σm. (d) Overlay of reduction event E pc. for compounds 4-R at 100 mV/s.

Considering the distorted trigonal planar geometry, diamagnetic character, and MLCT exhibited by complexes 4-R, they are hypothesized to exist in the Ni0 formal oxidation state. Therefore, the presence of a reduction in the voltammograms warranted further investigation as it is likely not metal-based. When 4-Me was compared to the CV of free CyPHOX ligand and 3-Me, the latter only exhibited oxidation events related to the ligand and 2 reduction events presumed to correspond to NiI/II and Ni0/I redox events (SI Figure S122). Thus, this reduction is unique to the aryne complexes 4-R.

Therefore, we sought to chemically reduce our CyPHOX–Ni benzyne complex (4-H) to gain further insight into the identity of this reduced species, and whether the reduction is aryne- or CyPHOX-based. The unsubstituted complex 4-H was chosen as a model system. Single-electron reduction of 4-H with KC8 at −78 °C in THF with 18-crown-6 as an encapsulant cleanly produced complex 5-H. A single new resonance in the 31P­{1H} NMR spectrum was observed at 69.0 ppm, which is downfield of the neutral species (38.3 ppm) (SI Figure S64). The aromatic region of the 1H NMR did not exhibit appreciable changes in the chemical shifts of the resonances. However, broadening of the resonances was observed and the chemical shift of the methylene protons on the oxazoline ring shifted significantly from 3.62 to 2.94 ppm. Crystals suitable for X-ray crystallographic analysis were grown from a saturated THF solution with vapor diffusion of pentane. The crystal structure of this complex showed a monoanionic CyPHOX–Ni benzyne complex (5-H) with K­(18-crown-6)­(THF)2 as the counterion (Figure a). The most dramatic structural change is seen in the imine CN bond length (1.284(3) Å to 1.331(2) Å), suggesting partial reduction of this bond (Figure b). As expected, the radical character appears to be delocalized throughout the conjugated π system of the CyPHOX backbone, manifesting in alternating bond lengths of the aryl ring (SI Figure S96). The aryne CC bond only slightly lengthened (1.345(4) Å vs 1.349(2) Å), and there is an unexpected but substantial bond elongation of the C2–C3 bond (1.354(3) Å to 1.387(2) Å). The DFT-calculated spin densities show that the radical character is indeed delocalized throughout the CyPHOX ligand (Figure c). This manifests in the UV/vis spectrum of 5-H, which shows that it is capable of similar transitions to the parent benzyne, but the former exhibits a more intense charge transfer at λmax = 257 nm (ε = 10,140 M–1·cm–1). This represents, to our knowledge, the first reduced metal aryne complex. Considering that the MLCT is still present after reduction of the benzyne complex (SI Figure S105), this supports our hypothesis that this charge transfer involves transfer of an electron from the Ni center to the aryne. Interestingly, EDA-NOCV computational analyses carried out by Mondal and co-workers suggest an alternative bonding scenario in d10 metal aryne complexes to closed-shell σ-donation/π-backdonation described by the Dewar-Chatt-Duncanson model. Their studies show that a mix of σ-donation and an e sharing π-bond between positively charged ligand–metal and negatively charged aryne fragments is another plausible bonding scenario that is at times energetically favored over purely dative bonding. Their studies suggest that enhancing the ability of the ligand–metal fragment to undergo electron transfer to the aryne fragment may be beneficial toward reactivity and stability. Thus, we postulate that this visible light-enabled MLCT observed for aryne complexes 4-R may be helpful in stabilizing these complexes.

5.

5

(a) Chemical reduction of 4-H with KC8. (b) X-ray crystal structure of reduced CyPHOX–Ni benzyne complex 5-H. K+[18-crown-6]­(THF)2 counterion omitted for clarity. (c) Calculated spin density of 5-H shows that the radical character is localized on CyPHOX.

To further elucidate the electronic structure of complex 5–H, X–Band EPR studies were conducted in fluid THF to elucidate the nature of the radical at 298 K. An anisotropic signal was observed with g x,y,z values of 1.9954, 2.065, and 2.067, respectively (SI Figure S126). Resolved significant hyperfine coupling to the imine nitrogen (A iso = (14N = 1, n = 1) = 159.7 MHz) and phosphine (A iso = (31P = 1, n = 1) = 167.8 MHz) was observed. Additional hyperfine coupling was observed to the four distinct, inequivalent protons on the ligand backbone (A iso = (1H, n = 1) = 34.9, 52.6, 54.2, and 70.6 MHz). No superhyperfine coupling was observed with the most abundant 58Ni nucleus (I = 3/2, 68.08%). These findings are consistent with a radical delocalized across the CyPHOX ligand backbone as observed in the computed spin–density plot (vide supra). Subjecting 4–H to the same conditions resulted in no observable signal at 298 K, further supporting the closed–shell electronic structure of the proposed Ni(0) center.

Origins of Binding Selectivity in CyPHOX–Ni Benzyne Complexes

Given that the solid-state structures did not provide significant insight into the observed regioisomeric ratios of complexes 4-R, DFT calculations were employed to study the structural features of the metal aryne complexes and their impact on regioisomeric ratios. Whereas the solid-state studies were limited due to crystal packing effects, computational studies allow for modeling of species in the solution state, which may better represent the system. Thus, to elucidate the origins of CyPHOX ligand-induced selectivity as well as aryne substituent effects, the geometries of CyPHOX–Ni benzyne and substituted aryne regioisomers 4-R were optimized at the B3LYP-D3/def2-SVP level of theory. The optimized geometries for the major (4-F-a) and minor (4-F-b) regioisomers of aryne complex 4-F are shown in Figure a. Single point energies were calculated with ωB97X-D/def2-TZVP/SMD­(THF). Gibbs free energy differences (SI Figure S132) and quasi-harmonic corrected enthalpy differences (ΔH qh, Figure b) between regioisomers calculated at this level of theory were found to correlate well with the experimentally observed regioisomeric ratios, which suggests that this is a valid level of theory for modeling these complexes. ,

6.

6

(a) Optimized geometries of regioisomers of aryne 4-F and their internal aryne angle and energy differences. (b) Correlation between experimentally observed regioisomeric ratios of aryne complexes and calculated quasi-harmonic corrected enthalpy difference between regioisomers (ΔH qh). (c) Correlation between calculated ΔΔθ (internal aryne angle differences) vs experimentally observed regioisomeric ratios for CyPHOX–Ni aryne complexes.

In order to decouple the effects of the CyPHOX ligand and the aryne substituent on binding selectivity, benzyne complex 4-H was used as a benchmark. It was found to exhibit an appreciable CyPHOX ligand-induced distortion present in the internal aryne angles (θC1 and θC2) leading to an angle difference Δθ = θC1 – θC2 = 1.9°, where θC2 is the smaller internal angle (that which is trans to phosphine). The major regioisomers of the substituted aryne complexes followed this same trend but exhibited varying degrees of additional aryne distortion–while major regioisomer 4-F-a features the greatest angle difference of Δθmajor = 4.3°, 4-Me-a only has a 1.8° difference (SI Figure S127). Interestingly, the opposite is true for the calculated geometries of all minor aryne regioisomers 4-R-b in this study. That is, placing the substituent on the phosphine side of the CyPHOX ligand results in the internal aryne angle θC1 becoming the smaller internal angle (Figure a). In order to include both regioisomers in our model, we determined the difference in angle distortion ΔΔθ between each pair of substituted aryne regioisomers, where ΔΔθ = Δθmajor – Δθminor. Gratifyingly, similar to the Aryne Distortion Model, this ΔΔθ parameter was found to have a strong correlation with the experimentally observed regioisomeric ratios (R 2 = 0.92, Figure c). It appears that for the major regioisomers, the ligand and aryne distortion are “matched,” to yield the lowest energy conformation. On the other hand, in the case of the minor regioisomer, the ligand and aryne distortion are “mismatched,” and the aryne substituent erodes the CyPHOX ligand-induced distortion to make θC1 the smaller internal angle (SI Figure S127). Furthermore, the greater the minor regioisomer geometry (4-R-b) deviates from the major (4-R-a), the higher the energy penalty, leading to higher regioisomeric ratios. Our model suggests that the CyPHOX ligand-induced aryne distortion is a governing effect that can be tuned to further improve selectivities and is necessary for Ni aryne complexes to be synthesized and functionalized selectively.

In studying the electronic structure of CyPHOX–Ni aryne complexes, HOMO/LUMO and natural population analysis (NPA) charge analysis was performed in order to gain further insight into reactivity. These calculations show that the HOMO across all substituted aryne complexes 4-R is aryne-based (SI Figure S130). The LUMO is imine-based (SI Figure S132). An appreciable NPA charge difference (Δq NPA = |q C2q C1|) was observed for all aryne complexes across the aryne carbons. As a charge-controlled model had once been postulated for free arynes, , we evaluated this possibility against our distortion model. The ΔqNPA values for the major regioisomers 4-R-a also showed a strong trend with our observed selectivities (SI Figure S131). Thus, distortion-interaction analysis was needed in order to determine whether binding selectivity in this CyPHOX–Ni aryne system is governed by distortion or atomic charge control.

In order to determine the phenomenon that controls binding selectivity, distortion-interaction analysis was carried out on the aryne complexes 4-R. , The interaction energy differences (ΔΔE interaction) between regioisomers are much smaller than the distortion energy differences between regioisomers (ΔΔE distortion) (SI Figure S128). These results indicate that the regioselectivity is mainly controlled by distortion energies. Further decomposition of the distortion energies indicates that both the CyPHOX–Ni and aryne fragments are slightly more distorted in the minor regioisomers (SI Figure S129). This supports that the steric repulsions between the bulkier phosphine arm of the C1-symmetric ligand and aryne substituent are among the factors that contribute to regioinduction (SI Figure S127).

Benchmarking the Translation of Selectivity from Aryne Binding to Difunctionalizations

Our group previously reported that upon difunctionalization of 4-OMe, the arene product ratio (r.r. 88:12) was slightly enhanced compared to the observed aryne regioisomeric ratio (r.r. 81:19). In order to apply our knowledge of selectivity to reactivity, we hypothesized that the product ratios of difunctionalizations would continue to be in agreement with the ratios of the aryne complexes across the series 4-R. Therefore, we subjected the substituted arynes 4-R to methyl triflate (MeOTf) followed by deuterated trifluoroacetic acid (TFA-d), to give the corresponding difunctionalized products (Figure a). 4-Me gave the corresponding products in 79:21 r.r., once again showing good retention in selectivity from aryne to arene product. Excitingly, the difunctionalizations of 4-Cl and 4-F were found to produce a single regioisomer. 4-CN was found to not be compatible with this sequence, as MeOTf is known to methylate nitriles and methyl iodide was not a strong enough methylating source, which agrees with the CV data which shows that 4-CN exhibits the harshest oxidation potential. These difunctionalizations proceeded with low yields, primarily due to the rather inert nature of the methylated intermediates.

7.

7

(a) Difunctionalization of CyPHOX–Ni aryne complexes 4-Me, 4-Cl, and 4-F via methylation/deuteration shows retention in selectivity from aryne complexes to arene products. Yields calculated via 2H NMR using CD2Cl2 as an internal standard. Calculated methylation transition states originating from (b) the major regioisomer 4-F-a. and (c) the minor regioisomer 4-F-b. Transition states were calculated at the ωB97X-D/def2-TZVP/SMD­(THF)//B3LYP-D3/def2-SVP level of theory.

Having validated that our unsymmetrical ligand-controlled approach could be leveraged to not only synthesize aryne complexes regioselectively but maintain selectivity in difunctionalizations, we wanted to further investigate the relative contributions of phosphine trans influence- and aryne-induced distortion on selectivity in these difunctionalizations. Thus, transition state calculations were carried out at the ωB97X-D/def2-TZVP/SMD­(THF)//B3LYP-D3/def2-SVP level of theory for the methylations of both regioisomers of aryne complexes 4-Me and 4-F at each terminus of the aryne (SI Figures S134–137). The methylation with the major regioisomer 4-F-a prefers the ortho-methylation at the C2 position, trans to the phosphine ( o -TS-4-F-a). This process proceeds via a direct SN2-type reaction between the aryl carbon and the MeOTf electrophile to furnish a methylated σ-aryl intermediate, which is consistent with previous computational studies on methylation of (dcpe)­Ni–benzynes by Hosoya. By contrast, meta-methylation of 4-F-a at the C1 position, trans to oxazoline, is less favorablethe same direct SN2-type transition state could not be located, whereas the alternative MeOTf oxidative addition process has a 4.6 kcal/mol higher barrier in energy (SI Figure S136). Similar calculations with the minor regioisomer 4-F-b show that the meta-methylation of C2, which is now positioned trans to the phosphine (Figure c), has a 4.2 kcal/mol lower barrier than the ortho-methylation trans to oxazoline (SI Figure S137). Our computational studies suggest the phosphine trans influence governs the regioselectivity in aryne complex difunctionalizations, with the preferred methylation occurring trans to the phosphine ligand in the reactions with both regiosiomers of the aryne.

Given the isomerization of the aryne complexes 4-R observed via 31P­{1H}–31P­{1H} NOESY as well as the rotational disorder between several regioisomer pairs observed in crystallo, the transition state for this isomerization process was calculated for 4-F, and was found to have a 31.6 kcal/mol energy barrier (SI Figure S138). This high kinetic barrier leading to slow isomerization as well as the modest yields in these model difunctionalizations are likely contributors for the selectivities in the difunctionalizations being in good agreement with the aryne ratios, but not identical.

Development of the Metal Aryne Reactivity/Selectivity (MAR/S) Model

With the combined experimental, spectroscopic, and computational insights gathered from this study, we propose the Metal Aryne Reactivity/Selectivity (MAR/S) Model. For unsymmetrically substituted CyPHOX–Ni benzynes, the binding selectivity is controlled by an interplay of ligand and aryne substituent-induced distortion. These two effects are matched in the major regioisomer to minimize ring strain, compounding on C2 to result in both a longer Ni–C2 bond and a smaller internal angle θC2, respectively (Figure a). This orientation also minimizes the steric interactions between the aryne substituent and the ligand. Ultimately, it is selective for ortho-methylation at the C2 position, leading to formation of major product 6-R-a, as supported by our transition state analysis for the methylations (vide supra). In the minor regioisomer, the ligand and aryne substituent-induced distortion are mismatched, acting on C2 to produce a longer Ni–C2 bond and a smaller internal angle θC1, respectively (Figure b). This binding orientation is generally higher in energy due to a combination of higher ring strain as well as steric interactions between the aryne substituent and the bulky cyclohexyl groups of the phosphine donor. However, it demonstrates that the trans influence is the predominant factor in maintaining selectivity from aryne complex to arene product, as it is selective for meta-methylation to produce the minor product 6-R-b.

8.

8

MAR/S Model for the (a) major and (b) minor regioisomers of CyPHOX–Ni benzyne complexes.

Conclusion

In summary, we have carried out the first comprehensive experimental electronic structure study across a series of CyPHOX–Ni aryne complexes. While X-ray crystallographic studies showed important trends regarding π-backdonation and the trans influence of the phosphine, no structural feature was found to correlate to the aryne complex regioisomeric ratios. CyPHOX–Ni aryne complexes exhibit intense MLCT absorbances, the properties for which are well-described by Hammett parameters (ε by σp+ and λmax by σm, respectively). In terms of their redox properties, they are capable of an irreversible one-electron oxidation and a quasi-reversible, one-electron, ligand-centered reduction event. Trends in oxidation potentials are well-described by Hammett parameters (σm for E pa). We have probed the reduction experimentally, and an anionic benzyne complex has been synthesized, characterized, and studied computationally. This finding leads us to propose a Ni0 oxidation state for CyPHOX–Ni benzyne complexes.

In addition to the electronic structure studies, DFT and NMR spectroscopic studies have allowed us to develop the MAR/S Model. Selectivity in aryne formation is a result of a combination of sterics and CyPHOX ligand and aryne substituent-induced aryne distortion. Similarly to the Aryne Distortion Model, calculated geometries show that greater ΔΔθ values correlate with higher CyPHOX–Ni aryne complex regioisomeric ratios. Furthermore, due to the slow interconversion of the substituted arynes and the phosphine trans influence, difunctionalizations proceed with excellent retention of selectivity from aryne to arene products. However, unlike the Aryne Distortion Model, the relative internal angles do not dictate the site selectivity in functionalizations. Transition state analysis performed on the methylations at each aryne terminus for arynes 4-Me and 4-F support that electrophilic addition consistently occurs at the position trans to phosphine for both aryne regioisomers. This governing phenomenon essentially leads to regiospecific difunctionalizations with respect to the aryne complex regiochemistry. Interestingly, the ligand-induced aryne distortion contribution was found to be very consistent across the unsymmetrically substituted aryne complexes surveyed, suggesting that ligand-induced distortion can be tuned via ligand design to further improve selectivities. CyPHOX proved to be a helpful first-generation unsymmetrical ancillary ligand for stabilizing Ni aryne complexes for these electronic structure studies and uncovering the origins of selectivity in their synthesis and difunctionalization. Future work will focus on ligand modifications to expand the reactivity of Ni aryne complexes based on this conceptual framework, improving selectivities, and applications to catalytic reactivity.

Supplementary Material

au5c01159_si_001.pdf (23.9MB, pdf)

Acknowledgments

We thank Dr. Thomas Smith for help with NMR experiments and Victor Young for help with crystallography. We thank the Gladfelter lab for access to UV–vis instrumentation. We thank Rana Abdu for help with EPR. Instrumentation for the UMN Chemistry NMR facility was supported from a grant through the National Institutes of Health (NIH) (S10OD011952). Mass spectrometry analysis was performed at the UMN Department of Chemistry Mass Spectrometry Laboratory (MSL), supported by OVPR, CSE, and the Department of Chemistry at UMN, as well as the NSF (CHE-1336940). Computational studies are supported by the NIH (R35 GM128779) with supercomputer resources provided by the University of Pittsburgh Center for Research Computing and Data and the Advanced Cyberinfrastructure Coordination Ecosystem: Services & Support (ACCESS) program, supported by NSF award numbers OAC-2117681, OAC-1928147, and OAC-1928224.

Glossary

Abbreviations

Cy

cyclohexyl

PHOX

phosphinooxazoline

COD

cyclooctadiene

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacsau.5c01159.

  • Experimental details and characterization of all new compounds (NMR, UV–vis, CV, etc.) (PDF)

The manuscript was written through contributions of all authors.

A.U., N.A.G., and C.C.R. acknowledge the NIH R35GM146957 for funding. K.P.Q. and P.L. acknowledge the NIH R35 GM128779 for funding. C.C.R. is supported by an Amgen Young Investigator Award, 3 M Alumni Professorship, a McKnight Land-Grant Professorship, Dreyfus Teacher Scholar Award, and Sloan Fellowship. A.U. is supported by an NSF Graduate Research Fellowship (#2237827).

The authors declare no competing financial interest.

References

  1. Tadross P. M., Stoltz B. M.. A Comprehensive History of Arynes in Natural Product Total Synthesis. Chem. Rev. 2012;112(6):3550–3577. doi: 10.1021/cr200478h. [DOI] [PubMed] [Google Scholar]
  2. Takikawa H., Nishii A., Sakai T., Suzuki K.. Aryne-Based Strategy in the Total Synthesis of Naturally Occurring Polycyclic Compounds. Chem. Soc. Rev. 2018;47(21):8030–8056. doi: 10.1039/C8CS00350E. [DOI] [PubMed] [Google Scholar]
  3. Sanz R.. Recent applications of aryne chemistry to organic synthesis. a review. Org. Prep. Proced. Int. 2008;40(3):215–291. doi: 10.1080/00304940809458089. [DOI] [Google Scholar]
  4. Cheong P. H.-Y., Paton R. S., Bronner S. M., Im G.-Y. J., Garg N. K., Houk K. N.. Indolyne and Aryne Distortions and Nucleophilic Regioselectivites. J. Am. Chem. Soc. 2010;132(4):1267–1269. doi: 10.1021/ja9098643. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Im G.-Y. J., Bronner S. M., Goetz A. E., Paton R. S., Cheong P. H.-Y., Houk K. N., Garg N. K.. Indolyne Experimental and Computational Studies: Synthetic Applications and Origins of Selectivities of Nucleophilic Additions. J. Am. Chem. Soc. 2010;132(50):17933–17944. doi: 10.1021/ja1086485. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Medina J. M., Mackey J. L., Garg N. K., Houk K. N.. The Role of Aryne Distortions, Steric Effects, and Charges in Regioselectivities of Aryne Reactions. J. Am. Chem. Soc. 2014;136(44):15798–15805. doi: 10.1021/ja5099935. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Goetz A. E., Garg N. K.. Enabling the Use of Heterocyclic Arynes in Chemical Synthesis. J. Org. Chem. 2014;79(3):846–851. doi: 10.1021/jo402723e. [DOI] [PubMed] [Google Scholar]
  8. Bronner S. M., Goetz A. E., Garg N. K.. Overturning Indolyne Regioselectivities and Synthesis of Indolactam V. J. Am. Chem. Soc. 2011;133(11):3832–3835. doi: 10.1021/ja200437g. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bennett M. A., Schwemlein H. P.. Metal Complexes of Small Cycloalkynes and Arynes. Angew. Chem., Int. Ed. Engl. 1989;28(10):1296–1320. doi: 10.1002/anie.198912961. [DOI] [Google Scholar]
  10. Bennett M. A., Wenger E.. The Reactivity of Complexes of Nickel(0) and Platinum(0) Containing Benzyne and Related Small-Ring Alkynes. Chem. Ber. 1997;130(8):1029–1042. doi: 10.1002/cber.19971300803. [DOI] [Google Scholar]
  11. Dhokale R. A., Mhaske S. B.. Transition-Metal-Catalyzed Reactions Involving Arynes. Synthesis. 2018;50(01):1–16. doi: 10.1055/s-0036-1589517. [DOI] [Google Scholar]
  12. Yang Y., Huang H., Wu L., Liang Y.. Palladium-Catalyzed Annulation of Benzynes with N-Substituted-N-(2-Halophenyl)­Formamides: Synthesis of Phenanthridinones. Org. Biomol. Chem. 2014;12(29):5351–5355. doi: 10.1039/c4ob00997e. [DOI] [PubMed] [Google Scholar]
  13. Wagner C. J., Dong G.. Ring Expansion via Palladium-Catalyzed Aryne Insertion into C–C Bonds of Benzocyclobutenones. Angew. Chem. Int. Ed. 2025;64:e202500148. doi: 10.1002/anie.202500148. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Yoshida H., Honda Y., Shirakawa E., Hiyama T.. Palladium–Iminophosphine-Catalysed Carbostannylation of Arynes: Synthesis of Ortho-Substituted Arylstannanes. Chem. Commun. 2001;(18):1880–1881. doi: 10.1039/B103745P. [DOI] [PubMed] [Google Scholar]
  15. Tang C.-Y., Wu X.-Y., Sha F., Zhang F., Li H.. Pd-Catalyzed Assembly of Phenanthridines from Aryl Ketone O-Acetyloximes and Arynes through C–H Bond Activation. Tetrahedron Lett. 2014;55(5):1036–1039. doi: 10.1016/j.tetlet.2013.12.075. [DOI] [Google Scholar]
  16. Qiu Z., Xie Z.. Nickel-Catalyzed Three-Component [2 + 2+2] Cycloaddition Reaction of Arynes, Alkenes, and Alkynes. Angew. Chem., Int. Ed. 2009;48(31):5729–5732. doi: 10.1002/anie.200902006. [DOI] [PubMed] [Google Scholar]
  17. Sumida Y., Sumida T., Hashizume D., Hosoya T.. Preparation of Aryne–Nickel Complexes from Ortho-Borylaryl Triflates. Org. Lett. 2016;18(21):5600–5603. doi: 10.1021/acs.orglett.6b02831. [DOI] [PubMed] [Google Scholar]
  18. Bennett M. A.. Aryne Complexes of Zerovalent Metals of the Nickel Triad. Aust. J. Chem. 2010;63(7):1066–1075. doi: 10.1071/CH10198. [DOI] [Google Scholar]
  19. Retbøll M., Edwards A. J., Rae A. D., Willis A. C., Bennett M. A., Wenger E.. Preparation of Benzyne Complexes of Group 10 Metals by Intramolecular Suzuki Coupling of Ortho-Metalated Phenylboronic Esters: Molecular Structure of the First Benzyne-Palladium(0) Complex. J. Am. Chem. Soc. 2002;124(28):8348–8360. doi: 10.1021/ja0264091. [DOI] [PubMed] [Google Scholar]
  20. Bennett M. A., Hockless D. C. R., Wenger E.. Generation of (2,3-.Eta.)-Naphthalyne-Nickel(0) Complexes and Their Reactions with Unsaturated Molecules. Organometallics. 1995;14(4):2091–2101. doi: 10.1021/om00004a071. [DOI] [Google Scholar]
  21. Umanzor A., Garcia N. A., Roberts C. C.. Ligand-Controlled Regioinduction in a PHOX-Ni Aryne Complex. ACS Org. Inorg. Au. 2024;4(1):97–101. doi: 10.1021/acsorginorgau.3c00046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kukolich S. G., McCarthy M. C., Thaddeus P.. Molecular Structure of o -Benzyne from Microwave Measurements. J. Phys. Chem. A. 2004;108(14):2645–2651. doi: 10.1021/jp031344p. [DOI] [Google Scholar]
  23. Chatt J., Duncanson L. A.. 586. Olefin Co-Ordination Compounds. Part III. Infra-Red Spectra and Structure: Attempted Preparation of Acetylene Complexes. J. Chem. Soc. 1953;(No. 0):2939–2947. doi: 10.1039/JR9530002939. [DOI] [Google Scholar]
  24. Frenking G., Fröhlich N.. The Nature of the Bonding in Transition-Metal Compounds. Chem. Rev. 2000;100(2):717–774. doi: 10.1021/cr980401l. [DOI] [PubMed] [Google Scholar]
  25. Parkin G.. Bond-Stretch Isomerism in Transition Metal Complexes: A Reevaluation of Crystallographic Data. Chem. Rev. 1993;93:887–911. doi: 10.1021/cr00019a003. [DOI] [Google Scholar]
  26. Rheingold A. L., Harper J. R.. A Comparison of the Solid-State Structures of Four Crystallographically Independent Molecules of Bis­(Pentamethylcyclopentadienyl)­Dioxodimolybdenum­(VI) Oxide, [Cp* MOO,] 2O. J. Organomet. Chem. 1991;403:335–344. doi: 10.1016/0022-328X(91)86480-E. [DOI] [Google Scholar]
  27. Humke J. N., Belli R. G., Plasek E. E., Kargbo S. S., Ansel A. Q., Roberts C. C.. Nickel Binding Enables Isolation and Reactivity of Previously Inaccessible 7-Aza-2,3-Indolynes. Science. 2024;384(6694):408–414. doi: 10.1126/science.adi1606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hansch Corwin., Leo A., Taft R. W.. A Survey of Hammett Substituent Constants and Resonance and Field Parameters. Chem. Rev. 1991;91(2):165–195. doi: 10.1021/cr00002a004. [DOI] [Google Scholar]
  29. Lin Q., Dawson G., Diao T.. Experimental Electrochemical Potentials of Nickel Complexes. Syn lett. 2021;32:1606–1620. doi: 10.1055/s-0040-1719829. [DOI] [Google Scholar]
  30. Suthar S., Mondal K. C.. Unveiling the Bonding Scenario in Metal–Aryne Complexes with EDA-NOCV Analyses. Organometallics. 2024;43(3):261–272. doi: 10.1021/acs.organomet.3c00406. [DOI] [Google Scholar]
  31. Li Y.-P., Gomes J., Sharada S. M., Bell A. T., Head-Gordon M.. Improved Force-Field Parameters for QM/MM Simulations of the Energies of Adsorption for Molecules in Zeolites and a Free Rotor Correction to the Rigid Rotor Harmonic Oscillator Model for Adsorption Enthalpies. J. Phys. Chem. C. 2015;119(4):1840–1850. doi: 10.1021/jp509921r. [DOI] [Google Scholar]
  32. Luchini G., Alegre-Requena J., Funes-Ardoiz I., Paton R.. GoodVibes: Automated Thermochemistry for Heterogeneous Computational Chemistry Data. F1000Research. 2020;9(291):291. doi: 10.12688/f1000research.22758.1. [DOI] [Google Scholar]
  33. Liu Z., Larock R. C.. Facile N-Arylation of Amines and Sulfonamides and O-Arylation of Phenols and Arenecarboxylic Acids. J. Org. Chem. 2006;71(8):3198–3209. doi: 10.1021/jo0602221. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Tadross P. M., Gilmore C. D., Bugga P., Virgil S. C., Stoltz B. M.. Regioselective Reactions of Highly Substituted Arynes. Org. Lett. 2010;12(6):1224–1227. doi: 10.1021/ol1000796. [DOI] [PubMed] [Google Scholar]
  35. Ess D. H., Houk K. N.. Distortion/Interaction Energy Control of 1,3-Dipolar Cycloaddition Reactivity. J. Am. Chem. Soc. 2007;129:10646–10647. doi: 10.1021/ja0734086. [DOI] [PubMed] [Google Scholar]
  36. Bickelhaupt F. M., Houk K. N.. Analyzing Reaction Rates with the Distortion/Interaction-Activation Strain Model. Angew. Chem. Int. Ed. 2017;56:10070–10086. doi: 10.1002/anie.201701486. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Booth B. L., Jibodu K. O., Proença M. F. J. R. P.. The Chemistry of Nitrilium Salts. Part 2. The Preparation of Nitrilium Trifluoromethanesulphonate Salts and Their Reactions with Some Oxygen and Sulphur Nucleophiles. J. Chem. Soc., Perkin Trans. 1. 1983:1067–1073. doi: 10.1039/P19830001067. [DOI] [Google Scholar]
  38. Hatakeyama M., Sakamoto Y., Ogata K., Sumida Y., Sumida T., Hosoya T., Nakamura S.. A Study on an Unusual SN2Mechanism in the Methylation of Benzyne through Nickel-Complexation. Phys. Chem. Chem. Phys. 2017;19(39):26926–26933. doi: 10.1039/C7CP04739H. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

au5c01159_si_001.pdf (23.9MB, pdf)

Articles from JACS Au are provided here courtesy of American Chemical Society

RESOURCES