Article

 

Cyclic C-F-Metal Containig Species. Solid State and Computational Examples

 

Maribel Arroyo Carranza¹ and Hugo Torrens Miquel²*

 

¹ Centro de Química del Instituto de Ciencias de la Benemérita Universidad Autónoma de Puebla. Edificio 103G del Complejo de Ciencias, Ciudad Universitaria, San Manuel, 72571, Puebla, Pue., Mexico. slmarroyo@hotmail.com

² División de Estudios de Posgrado, Facultad de Química, Universidad Nacional Autónoma de México, Ciudad. Universitaria, 04510 México D.F., México. torrens@unam.mx

 

Received March 20, 2012
Accepted April 16, 2013

 

Abstract

This article examines the results of research published between 2000 and 2012, which found experimental evidence of cyclic species containing the fragment carbon-fluorine-transition metal. 25 Examples (numbered throughout this paper) with 20 different metals and a wide variety of chemical environments have been gathered. A similar analysis brings together almost 40 transitional states and intermediates resulting from theoretical work studying the transition metal mediated carbon-fluorine bond activation. Among other aspects, this review is useful to recognize the specific characteristics which determine whether the C-F activation process is carried out through a concerted process, a mechanism assisted by auxiliary ligands or by free radicals.

Key words: Carbon, fluorine, bond activation, structure, theory.

 

Resumen

En este artículo se examinan los resultados de investigación publicados entre 2000 y 2012, en los que se han encontrado evidencias experimentales de especies cíclicas conteniendo al fragmento carbono-flúor-metal de transición. Se han reunido 25 ejemplos (numerados a lo largo de este trabajo) con 20 distintos metales y en una gran variedad de ambientes químicos. Un análisis similar reúne a casi 40 estados transicionales e intermediarios resultantes de trabajos teóricos enfocados al estudio de la activación del enlace carbono-flúor a través de metales de transición. Esta recopilación es de utilidad, entre otros aspectos, para reconocer las características específicas que determinan si el proceso de activación C-F se lleva a cabo mediante un proceso concertado, un mecanismo asistido por ligantes auxiliares o por radicales libres.

Palabras clave: Carbono, flúor, activación de enlace, estructura, teoría.

 

Introduction

The activation of the carbon-fluorine bond remains an area of great interest to those pursuing the replacement or removal of fluoride as well as to those who seek its incorporation into a given molecular structure. It is a process appealing to both, bond breakers and bond makers. The presence of cyclic species containing a carbon-fluorine-metal fragment has been repeatedly invoked as part of these processes either from a theoretical standpoint, as a result of kinetic studies or because the elusive carbon-fluorine-metal bonds have been observed in the solid state [1-8]. In this paper we gather some of these examples in each of these facets, although not in all cases the described metal containing cycles are directly involved in reactions promoting the activation of carbon-fluorine bonds.

Reviews focused on carbon-fluorine bond activation have been published [9] and a special issue of Organometallics has been dedicated to Fluorine in Organometallic Chemistry [10].

The compounds included in this review are grouped by metals and are described following the periodic table.

Scandium and Yttrium

In 2005, Bouwkamp et al. [11] studied the ionic metallocene complexes [Cp*₂M](BPh₄) (Cp* = C₅Me₅) of the trivalent 3d metals Sc, Ti, and V. They found that for Sc, a complex of twelve electrons, the anion interacts weakly with the metal center through one of the phenyl groups. With Sc, these Lewis acidic species react with fluorobenzene and 1,2-difluorobenzene to yield [Cp*₂Sc(κFFC₆H₅)₂](BPh₄) and [Cp*₂Sc(κ²F-1,2-F₂C₆H₄)](BPh₄) (1), the first examples of κF-fluorobenzene and κ²F-1,2-difluorobenzene adducts of transition metals. With the perfluorinated anion (B(C₆F₅)₄)^-, Sc forms the [Cp*₂Sc(κ²F-C₆F₅)B(C₆F₅)₃] (2) contact ion pair as shown in Figure 1. The nature of the metal-fluoroarene interaction was studied by density functional theory calculations and by comparison with the corresponding tetrahydrofuran adducts concluding that it is predominantly electrostatic.

Two years later, in 2007, Li et al. [12] synthesized the first cationic aminobenzyl-metal complex with a (B(C₆F₅)₄)^-adduct as shown in Figure 2, [(C₅Me₄SiMe₃)Sc(CH₂C₆H₄NMe₂-o)(κ²F-C₆F₅)B(C₆F₅)₃] (3). An X-ray diffraction study revealed that compound 3 is a contact ion pair, in which the (B(C₆F₅)₄)^-anion is coordinated to the metal center in a κ²F-fashion with two adjacent (ortho and meta) F atoms (Fig. 3). Because of the higher electron deficiency of the cationic metal center in 3, the bond distances of the Sc-C₅Me₄SiMe₃ bonds (av. 2.434(3)Å) are much shorter than those in its neutral precursor (av. 2.539(2)Å), and so are the Sc-benzyl (2.195(3)Å, precursor: 2.288(2)Å) and Sc-amino (2.300(3)Å, precursor: 2.470(2)Å) bond distances.

In a C₆D₅Cl solution, the four C₆F₅ groups in 3 showed one set of ¹⁹F NMR signals, suggesting that it is a highly fluxional molecule, possibly due to the rapid dissociation and re-coordination of the (B(C₆F₅)₄)^-unit or a solvent molecule. This fluxionality was not fixed even at -45 °C, as found in a variable-temperature ¹⁹F NMR study.

Lara-Sanchez et al. [13] studied the reaction of [Y(N(SiMe₃)₂)₃] with the fluorinated ligand (2-C₆F₅N=CH)(6-^tBu)C₆H₃OH (HL) depicted in Figure 4 and Figure 5. This reaction affords [Y(N(SiMe₃)₂)(L)₂] (4) and [Y(L)₃] (5). Both complexes are seven-coordinate in the solid state due to Y-F co-ordination to the C₆F₅ substituents.

Hayes et al. [14] isolated [Y(κF,κC-C₆F₅)(C₆F₅)(THF)(ArNC₆H₄CHNAr)] (6), a six-coordinated Y compound containing a 4 membered cyclometallated ring, taking advantage from the fact that the precursor [YPh₂((THF)(ArNC₆H₄CHNAr)], above -20 °C, undergo a C₆F₅ transfer process, ultimately yielding the bis-pentafluorophenyl derivative 6 and unidentified boron containing products. Compound 6 was isolated in low yield and identified by X-ray crystallography and features a distinct Y-F interaction involving one of the ortho F atoms in the solid state, as shown in Figure 6 and Figure 7.

While exploring the coordination chemistry of the fluorinated diamino-dialkoxy tetradentate ligands [OC(CF₃)₂CH₂N(Me)(C H₂)_nN(Me)CH₂C(CF₃)₂O]^2-(n = 2, [ON²NO]^2-; n = 3, [ON³NO]^2-) onto group 3 and group 4 metals, Lavanant and collaborators [15] found an efficient route to the yttrium complex [Y(ON²NO)(CH₂SiMe₃)(THF)] (7). X-ray crystallographic analysis reveals that this compound adopts in the solid state a seven-coordinate distorted structure, due to Y-F coordination to one of the four CF₃substituents forming a 5-membered ring (Y-F 2.806(2) Å) as shown in Figure 8. In contrast, NMR studies show that it adopts a C2-symmetric structure in solution with no Y-F interaction.

Titanium, Zirconium and Hafnium

As mentioned above, Bouwkamp et al. [11] studied the ionic metallocene complexes [Cp*₂M](BPh₄) (Cp* = C₅Me₅) of the trivalent 3d metals Sc, Ti, and V. They found that for M = Ti and V, the cations are naked. Crystallographic studies shown that they each contain one strongly distorted Cp* ligand, with two agostic C-H-Ti interactions involving the Me groups of Cp*. For Ti, these Lewis acidic species react with fluorobenzene and 1,2-difluorobenzene to yield [Cp*₂Ti(κF-FC₆H₅)](BPh₄) and [Cp*₂Ti(κ²F-1,2F₂C₆H₄)](BPh₄) (8), the first examples of κF-fluorobenzene and κ²F-1,2-difluorobenzene adducts of transition metals. With the perfluorinated anion (B(C₆F₅)₄)^-, Ti form a [Cp*₂Ti(κ²F-C₆F₅)B(C₆F₅)₃] (9) contact ion pair (Fig. 9). The nature of the metal-fluoroarene interaction was studied by density functional theory calculations and by comparison with the corresponding tetrahydrofuran adducts and was found to be predominantly electrostatic for all metals studied.

Andino et al. [16] have studied the isomeric alkylidene complexes syn- and anti-[(PNP)Ti(=C^tBu(C₆F₅))(F)] and [(PNP)Ti(= C^tBu(C₇F₇))(F)], generated from C-F bond addition of hexafluorobenzene (C₆F₆) and octafluorotoluene (C₇F₈) by the alkylidyne ligand of the transient [(PNP)Ti≡C^tBu], (PNP^-= N(2-P(CHMe₂)₂-4-methylphenyl)₂).

Two mechanistic scenarios for the C-F bond activation of C₆F₆ are considered: 1,2-C-F addition of C₆F₆ to the fragment Ti≡C^tBu to form (PNP)Ti≡C^tBu(σ-C₆F₆) and [2+2]-cycloaddition of C₆F₆ to form the metallacyclobutene intermediate which undergo a β-fluoride elimination to yield 1-syn. Upon formation of the alkylidenes, the kinetic and thermodynamic alkylidene product is the syn isomer, which gradually isomerizes to the corresponding anti isomer to ultimately establish an equilibrium mixture (65/35) if the solution is heated in benzene to 105 °C for 1 h.

Single crystal X-Ray crystallographic data obtained for the cis and trans isomers are in good agreement with the computed DFT-optimized models shown in Figure 10 and Figure 11.

Wondimagegn and collaborators [17] have combined quantum mechanical and molecular mechanical models, to explore possible transformation routes for ion pair systems used as catalysts in olefin polymerization. The catalyst systems include [(NPR₃)₂TiMe]⁺(MeB(C₆F₅)₃)^-, [(Cp)(NCR₂)TiMe]⁺(MeB(C₆F₅)₃)^-, and [(Cp)(SiMe₂NR)TiMe]⁺(MeB(C₆F₅)₃)^- (Fig. 12). The possible thermal decomposition routes for the catalyst systems include fluorine transfer from (MeB(C₆F₅)₃)^-to the metal center, aryl transfer (C₆F₅) to the growing chain, transfer of B(C₆F₅)₃ from (MeB(C₆F₅)₃)^-to the ancillary ligand, exchange of methyl on butyl from the ancillary ligand with aryl in (MeB(C₆F₅)₃)^-, transfer of hydrogen from the Cp* ring to the growing chain, and transfer of hydrogen from the methyl group on tertiary butyl to the growing chain. The activation barriers fall in the range 15.5-68.7 kcal/mol. The transfer of fluorine from the counter ion to the metal center is the most facile deactivation pathway.

As part of their study of hetero bimetallic compounds, Gade and coworkers [18] determined the crystal structures of [HC{SiMe₂N(2,3,4-F₃C₆H₂)}₃Zr(μ-Cl)₂Li(OEt₂)₂] (10) and [HC{SiMe₂N(2-FC₆H₄)}₃Zr(μ-Cl)₂Li(OEt₂)₂] (11) in order to compare the role which the peripheral fluorine atoms play in both compounds. As in the monofluorophenyl analogue, the coordination at the central zirconium atom at 10 may be described, see Figure 13, as highly distorted octahedral with three amido-nitrogen atoms in facial sites, two bridging chloro ligands and with the sixth site being occupied by the ortho-fluorine atom of one of the phenyl groups. Considerable bonding character is indicated by the extent to which the phenyl ring is forced over to allow this chelate ring formation, giving a N-C-C intra-chelate angle of 114.6(4)° compared with 125.3(4)° for a non F bonded N-C-C angle.

Also, single-crystal X-ray structure analyses of [HC{SiMe₂N(2-FC₆H₄)}₃Zr(μ-CS₂)Fe(CO)₂Cp] (12) and [HC{SiMe₂N(2FC₆H₄)}₃Zr(μ-SCNPh)Fe(CO)₂Cp] (13) were carried out. It is clear that the overall coordination of the zirconium atom in both 12 and 13 bears a striking resemblance to that found in the dichloride 10 and 11.

As in the dichlorides, the tripodal ligands in the insertion products 12 and 13 both effectively occupy four coordination sites of the zirconium atom, with three facial nitrogen donors and an interaction from the ortho-fluoro atom of a peripheral phenyl ring at the fourth site. As in 11 the coordinating fluorophenyl ring is apparently being forced over to facilitate chelation [N-C-C 125.1(8)° and 124(1)° for the non F bonded rings, and N-C-C 114.8(8)° and 116(1)° for the F bonded rings in 12 and 13, respectively]. In both dichlorides, the zirconium-nitrogen bond within the five membered chelate ring is the longest and that trans to the coordinating fluorine atom is the shortest. The shortening of the Zr-N(1) bond in each of these compounds is consistent with enhanced π donation from the amide donor trans to the highly electronegative fluorine atom. The persistence of the fluoro-zirconium interaction, from its presence in the starting material 10 to that in the heterodinuclear insertion products 12 and 13, demonstrates the importance of the "active" ligand periphery in these compounds.

Giesbrecht and coworkers [19] have studied the compound {K(η⁶-C₇H₈)₂}{ZrCl₂(N(C₆F₅)₂)₃} (14) incorporating decafluorodiphenylamido ligands. The ¹⁹F NMR spectrum of 14 reveals pentafluorophenyl resonances in a 2:1 ratio, consistent with a trigonal bipyramidal structure being maintained in solution. The crystal structure of 14 however, reveals a pseudo octahedral structure, with the sixth-coordination site being completed by a weak Zr-F interaction with a pentafluorophenyl group of an amido ligand as show in Figure 14.

Among the outstanding results from the group of Arndt et al. [20], there are a series of compounds containing cyclometallated rings resulting from F-Zr interactions. Besides suggested intermediate as [rac-(ebthi)Zr(η²-Me₃SiC₂SiMe₃)] the crystal structure has been resolved for [(ebthi-(2-BH(C₆F₅)₂)Zr(κF-C₆F₅)] (15) (ebthi = 1,2-ethylene-1,1'-bis(η⁵-tetrahydroindenyl)), Figure 15 and Figure 16.

rac-(ebthi)Zr(η²-Me₃SiC₂SiMe₃)] undergoes electrophilic substitution at one five-membered ring of the ebthi ligand upon reaction with B(C₆F₅)₃, forming the corresponding alkenyl complex 15. In this complex, the boranate group is substituted at the 3-position of the five-membered ring, and one of its ortho fluorine atoms is coordinated to the zirconium center. The σ-bound alkenyl group at the zirconium center participates in an additional agostic interaction.

Studying the bonding and bending in Zirconium(IV) and Hafnium(IV) hydrazides, Herrmann and collaborators [21] follow the reactions of the dichloro complexes [M(N₂^TBSN_py)Cl₂] (M = Zr, Hf) with LiNHNPh₂ to afford mixtures of the two diastereomeric chlorohydrazido complexes [M(N₂^TBSN_py)(NHNPh₂)Cl] and also the reaction of [M(N₂^TBSN_py)(NNPh₂)(py)] with one molar equivalent of B(C₆F₅)₃ to give [Zr(N₂^TBSN_py)(Ph₂NN{B(C₆F₅)₂(κF-C₆F₅)})] (16) and [Hf(N₂^TBSN_py)(Ph₂NN{B(C₆F₅)₂(κFC₆F₅)})] (17). In these products, B(C₆F₅)₃ becomes attached to the Nαatom of the side-on bound hydrazinediide and there is an additional interaction of an ortho-F atom of a C₆F₅ ring with the metal centre (Fig. 17 and Fig. 18).

In a work directed to evaluate the role of ortho halides in 1-hexene polymerization, Schrock et al. [22] determined by X-ray studies the structure of a series of compounds including the fluorinated, [Hf{(2,6-F₂C₆H₃NCH₂)₂C(CH₃)(2-C₅H₄N)}(ⁱBu)₂] (18), with two fluorides weakly bonded to the metal as shown in Figure 19.

Activation of these type of complexes with (Ph₃C)(B(C₆F₅)₄) in bromobenzene led to the formation of [Hf{(2,6X₂C₆H₃NCH₂)₂C(CH₃)(2-C₅H₄N)}(ⁱBu)](B(C₆F₅)₄) species that are also active for the polymerization of 1-hexene. The rate of consumption of 1-hexene followed a first-order dependence on 1-hexene (and hafnium), although the rates were substantially slower compared to those for the known catalyst with mesityl substituents on the amido nitrogens and slower when X = F than when X = Cl. The ease of preparation of the cations also followed the order aryl = mesityl > 2,6-Cl₂C₆H₃ > 2,6-F₂C₆H₃. Finally, the quality of the polymerization, in terms of its living characteristics, deteriorated markedly when the aryl was 2,6-F₂C₆H₃. In conclusion the introduction of ortho chlorides or fluorides decreases the rate of polymerization and also encourages βhydride elimination and formation of shorter polymer chains, therefore compromising the living characteristics of the polymerization.

Vanadium

Giesbrecht et al. [19] have studied the interaction of VCl₃ with 3 equivalents of KN(C₆F₅)₂. This reaction generates the 'metallate' complex K[VCl(N(C₆F₅)₂)₃] (19); the solid state structure of 19 reveals a distorted trigonal bipyramid with the coordination sphere being completed by a weak V-F interaction with an ortho-fluorine of one of the decafluorodiphenylamido ligands as shown in Figure 20.

Molybdenum and Tungsten

Liu et al. [23] have found that monofluorinated benzenes undergo facile oxidative addition with the [W(HB(C₃H₃N₂)₃)(NO) (PMe₃)] system, forming seven-coordinate aryl fluoride complexes. DFT calculations reveal two viable reaction pathways, one passing through a κF complex and one passing through an η²-arene intermediate. From the κF complex, a modest 3 kcal/mol barrier is estimated to stand in the way of C-F activation, offering an explanation for the uncommonly rapid insertion of the metal. In contrast to reports of other second- or third-row transition metal complexes, C-H addition is not observed with monofluorinated benzenes, though this reaction is thought to be kinetically accessible. When the number of fluorines increases to four, C-H insertion becomes competitive, and for hexafluorobenzene or fluoronaphthalene, η²-coordination dominates (Fig. 21).

Beltran and coworkers [24] have studied the catalytic hydrodefluorination (HDF) of pentafluoropyridine in the presence of arylsilanes which is catalyzed by the tungsten(IV) and molybdenum(IV) cluster hydrides of formula [M₃S₄H₃(dmpe)₃]⁺, M = Mo or W (dmpe = 1,2-(bis)dimethylphosphinoethane). The reaction proceeds regioselectively at the 4-position under microwave radiation to yield the 2,3,5,6-tetrafluoropyridine.

A mechanism for the HDF reaction has been proposed that explains these results based on DFT calculations. The mechanism involves partial decoordination of the diphosphine ligand that generates an empty position in the metal coordination sphere. This position together with its neighbor M-H site are used to activate the C-F bond of the pentafluoropyridine through a M-H/C-F σ-bond metathesis mechanism involving a four-center transition state to give 2,3,5,6-tetrafluoropyridine. Subsequent coordination of the dangling diphosphine affords the para-substituted product and the fluoride cluster [M₃S₄F₃(dmpe)₃]⁺, M = Mo or W, as seen in the following Figure 22.

Ruthenium and Osmium

Huang and coworkers [25] have found that the reactions of both [RuH(Ph)(CO)(P^tBu₂Me)₂] and [OsH(Ph)(CO)(P^tBu₂Me)₂] with vinyl fluoride readily form M-F bonds; however, Ru eliminates benzene,while Os eliminates ethylene. In contrast, [Ru(H)₂(CO)(P^tBu₂Me)₂] and [Os(H)₂(CO)(1-butene)(P^tBu₂Me)₂] both react with vinyl fluoride to give ethylene and [MHF(CO)(P^tBu₂Me)₂]. Ethylene production from both dihydrides is attributed to β-F migration to M from a four membered MCH₂CH₂F transient, Figure 23, while the unique behavior of [RuH(Ph)(CO)(P^tBu₂Me)₂] (giving the fluorine containig product [Ru(η¹-vinyl)F(CO)(P^tBu₂Me)₂] and benzene) is attributed to the difficulty of achieving ruthenium(IV), and the ability of the strongly π-acidic vinyl fluoride to rapidly trigger reductive elimination of benzene. The products of reaction of [RuH(Ar)(CO)(P^tBu₂Me)₂] with vinyl fluoride are redirected more towards ethylene formation when Ar carries fluorine substituents. The reaction products of [OsH(R)(CO)(P^tBu₂Me)₂] with vinyl fluoride revert to R-H elimination when R is methyl. Finally, the more π-acidic H₂C=CF₂ triggers very rapid CH₄ elimination from [OsH(CH₃)(CO)(P^tBu₂Me)₂]; cleavage of the second C-F bond yields the vinylidene [OsF₂(CCH₂)(CO)(P^tBu₂Me)₂].

Ritter et al. [26] have described the synthesis, structure, and the activity of new ruthenium-based olefin metathesis catalysts featuring fluorinated N-heterocyclic carbene ligands [RuCl₂(carbene)(ligand)] (20).

The introduction of ortho halogen atoms profoundly alters the catalytic metathesis performance. Structural investigation suggested that an uncommon fluorine-ruthenium interaction, seen on Figure 24, is responsible for the significant rate enhancement. This is the first example of such an interaction resulting in increased catalytic activity.

Cobalt, rhodium and iridium

Giesbrecht et al. [19] found that CoI₂ reacts with 2 equivalents of. NaN(C₆F₅)₂ in the presence of pyridine to produce the expected product [Co{N(C₆F₅)₂}{N(C₆F₅)(κF-C₆F₅)}(py)₂] (21); X-ray crystallography reveals a five-coordinate species in the solid state which is additionally stabilized by a weak Co-F interaction, Figure 25.

Dugan and coworkers [27] have studied the overall binuclear oxidative addition of fluorinated arenes to the cobalt(I) complex [L^tBuCo] (L^tBu = 2,2,6,6-tetramethyl-3,5-bis(2,6-diisopropylphenylimido)-hept-4-yl) to give cobalt(II) complexes [L^t-BuCo(μ-F)]₂ and [L^tBuCo(Ar)], in a 1:2 molar ratio. The C-F activation reaction has a first-order rate dependence on both cobalt and fluorobenzene concentrations. The rate is increased by meta-fluoride substituents, and is slowed by ortho-fluoride substituents, suggesting electronic and steric influences on the transition state, respectively. The authors find that data are most consistent with a mechanism beginning with rate-limiting oxidative addition of the aryl fluoride to cobalt(I), followed by rapid reduction of the cobalt(III) aryl fluoride intermediate by a second molecule of [L^tBuCo] (Fig. 26).

In a remarkable paper, Erhardt and Macgregor [28] used density functional theory calculations to model the reaction of C₆F₆ with [IrMe(PEt₃)₃], this reaction proceeds with both C-F and P-C bond activation to yield trans-[Ir(C₆F₅)(PEt₃)₂(PEt₂F)], C₂H₄, and CH₄. Using a model species, trans-[IrMe(PH₃)₂(PH₂Et)], a low-energy mechanism involving nucleophilic attack of the electron-rich Ir metal center at C₆F₆ with displacement of fluoride has been identified. A novel feature of this process is the capture of fluoride by a phosphine ligand to generate a metallophosphorane intermediate [Ir(C₆F₅)(Me)(PH₃)₂(PH₂EtF)]. These events occur in a single step via a 4-centered transition state, in a process that has been termed "phosphine-assisted C-F activation". Alternative mechanisms based on C-F activation via concerted oxidative addition or electron-transfer processes proved less favorable. From the metallophosphorane intermediate the formation of the final products can be accounted for by facile ethyl group transfer from phosphorus to iridium followed by β-H elimination of ethene and reductive elimination of methane. The interpretation of phosphine-assisted C-F activation in terms of nucleophilic attack is supported by the reduced activation barriers computed with the more electron-rich model reactant trans-[IrMe(PMe₃)₂(PMe₂Et)] and the higher barriers found with lesser fluorinated arenes. Reactivity patterns for a range of fluoroarenes indicate the dominance of the presence of ortho-F substituents in promoting phosphine-assisted C-F activation, and an analysis of the charge distribution and transition state geometries indicates that this process is controlled by the strength of the Ir-aryl bond that is being formed (Fig. 27).

In a fundamental work, Choi and collaborators [29] have found the addition of C(sp³)-F bonds to a pincer compound [(PCP)Ir(NBE)] (NBE = norbornene) via the initial, reversible cleavage of a C-H bond. One example is shown on Figure 28.

The results of density functional theory calculations, offer additional support for the proposed mechanism above. On the basis of the approximate reaction rates, the experimental free energy barrier is about 23 kcal/mol. Direct addition of the C-F bond of CH₃F (i.e., addition via a three-centered transition state Ir-F-CH₃) is calculated to have a prohibitively high barrier (∆G^‡ = 31.1 kcal/mol relative to the fragment (PCP)Ir or 37.5 kcal/mol relative to [(PCP)Ir(NBE)]). In contrast, the pathway depicted on Figure 28 is calculated to have a substantially lower barrier (∆G^‡ = 16.5 kcal/mol relative to the fragment (PCP)Ir or 22.9 kcal/mol relative to [(PCP)Ir(NBE)]) fully consistent with the observed rate, attributable to the transition state of an α-fluorine-migration rate-determining step (three-centered transition state Ir-F-CH₂). Density Functional Theory calculations were performed using the Gaussian09 collection of computer programs. Emploing the M06 model. For Ir, the Hay-Wadt relativistic effective and the LANL2TZ basis set were applied. Vibrational frequencies were employed to determine zero-point energy corrections.

Nickel, Palladium and Platinum

In an impressive work, Schaub and collaborators [30] have described the reaction of [Ni₂(ⁱPr₂Im)₄(COD)] or [Ni(ⁱPr₂Im)₂(η²-C₂H₄)] (ⁱPr₂Im = 1,3-di(isopropyl)imidazol-2-ylidene), with different fluorinated arenes. These reactions occur with a high chemo- and regioselectivity. In the case of polyfluorinated aromatics of the type C₆F₅X such as hexafluorobenzene, octafluorotoluene, trime thyl(pentafluorophenyl)silane, or decafluorobiphenyl, the C-F activation regioselectively takes place at the C-F bond in the para position to the X group to afford the complexes [Ni(ⁱPr₂Im)₂(X)(C₆F₅)]. trans-[Ni(ⁱPr₂Im)₂(F)(4-(SiMe₃)C₆F₄)] was structurally characterized by X-ray diffraction.

The reaction of [Ni₂(ⁱPr₂Im)₄(COD)] with partially fluorinated aromatic substrates C₆H_xF_y leads to the products of a C-F activation, for example trans-[Ni(ⁱPr₂Im)₂(F)(2-C₆FH₄)]. The reaction of [Ni₂(ⁱPr₂Im)₄(COD)] with octafluoronaphthalene yields exclusively trans-[Ni(ⁱPr₂Im)₂(F)(1,3,4,5,6,7,8-C₁₀F₇)], the product of an insertion into the C-F bond in the 2-position, whereas for the reaction of [Ni(ⁱPr₂Im)₂(η²-C₂H₄)] with octafluoronaphthalene two isomers are formed.

The reaction of [Ni₂(ⁱPr₂Im)₄(COD)] or [Ni(ⁱPr₂Im)₂(η²-C₂H₄)] with pentafluoropyridine at low temperatures affords trans[Ni(ⁱPr₂Im)₂(F)(4-C₅NF₄)] as the sole product, whereas the reaction of [Ni(ⁱPr₂Im)₂(η²-C₂H₄)] performed at room temperature leads to the generation of trans-[Ni(ⁱPr₂Im)₂(F)(4-C₅NF₄)] and trans-[Ni(ⁱPr₂Im)₂(F)(2-C₅NF₄)].

The detection of intermediates as well as kinetic studies gives some insight into the mechanistic details for the activation of an aromatic carbon-fluorine bond at the {Ni(ⁱPr₂Im)₂} complex fragment.

The intermediates of the reaction of [Ni(ⁱPr₂Im)₂(η²-C₂H₄)] with hexafluorobenzene and octafluoronaphthalene, [Ni(ⁱPr₂Im)₂(η²C₆F₆)] and [Ni(ⁱPr₂Im)₂(η²-C₁₀F₈)], have been detected in solution. They convert into the C-F activation products.

Furthermore, density functional theory calculations on the reaction of [Ni₂(ⁱPr₂Im)₄(COD)] with hexafluorobenzene, octafluoronaphthalene, octafluorotoluene, 1,2,4-trifluorobenzene, and 1,2,3-trifluorobenzene are presented. Several calculated C-F-Ni cyclic species are described on this work (Fig. 29).

In 2007, Yandulov and Tran [31], reported a computational analysis on C-F reductive elimination from aryl palladium(II) fluorides. They explored N-heterocyclic carbenes and phosphines as auxiliary ligands. Three-coordinate T-shaped geometry of [PdL(Ar)F] (L = NHC, PR₃) was shown to offer kinetics and thermodynamics of Ar-F elimination largely compatible with synthetic applications, whereas coordination of strong fourth ligands to Pd or association of hydrogen bond donors with F each caused pronounced stabilization of palladium(II) reactant and increased activation barrier beyond the practical range. Potential C-F reductive elimination from [Pd(PMe₃)(Ph)F] was computed. Yandulov's calculations predicted that in the related tricoordinate NHC complex [Pd(Me₂NHC)(Ph)F] reductive elimination of Ar-F is more facile for electron-withdrawing substituents, similar to related carbon-heteroatom reductive elimination reactions and nucleophilic aromatic substitution reactions as shown in Figure 30 and Figure 31.

Decreasing donor ability of L promotes elimination kinetics via increasing driving force and para-substituents on Ar exert a sizable SNAr-type TS effect.

Synthesis and characterization of the novel [Pd(C₆H₄-4-NO₂)ArL(µ-F)]₂ (L = P(o-Tolyl)₃; P(t-Bu)₃) revealed stability of the fluoride-bridged dimer forms of the requisite [PdL(Ar)(F)] as the key remaining obstacle to Ar-F reductive elimination in practice.

In 2010, the first systematic mechanistic study of C-F reductive elimination reactions from aryl palladium(IV) fluoride complexes was reported by Furuya et al. [32], with particular focus on the C-F reductive elimination from the palladium(IV) compound [PdF(LN)(LN₂(SO₂)(NO₂))] pyridyl-sulfonamide stabilized cationic Pd(IV) fluoride shown below. The proposed mechanism for C-F reductive elimination is based on activation parameters, rate dependence on the polarity of the reaction medium, Hammett analysis, and DFT calculations.

The mechanistic study confirmed that C-F reductive elimination proceeds efficiently from aryl palladium(IV) fluoride complexes, supported by pyridyl-sulfonamide ancillary ligands. It was proposed that the pyridyl-sulfonamide ligand plays a crucial role for facile and efficient C-F bond formation. The ability of the pyridyl-sulfonamide ligand to function as a bidentate-tridentatebidentate coordinating ligand during oxidation and reductive elimination, combined with the appropriate electronic requirements of the sulfonamide to position the aryl substituent trans to the sulfonamide ligand in the palladium(II) complex and the fluoride ligand trans to the sulfonamide ligand in the palladium(IV) complex, may be the reasons for facile C-F bond formation (Fig. 32).

Nova et al. [33] have examined the computed mechanisms for C-F bond activation at the 4-position of pentafluoropyridine by the model zero-valent bis-phosphine complex, [Pt(PH₃)(PH₂Me)]. Three distinct pathways leading to square-planar platinum(II) products are persistent. Direct oxidative addition leads to cis-[Pt(F)(4-C₅NF₄)(PH₃)(PH₂Me)] via a conventional 3-center transition state. This process competes with two different phosphine-assisted mechanisms in which C-F activation involves fluorine transfer to a phosphorus center via novel 4-center transition states. Figure 33. The more accessible of the two phosphine-assisted processes involves concerted transfer of an alkyl group from phosphorus to the metal to give a platinum(alkyl)(fluorophosphine) compound, trans-[Pt(Me)(4-C₅NF₃)(PH₃)(PH₂F)], analogues of which have been observed experimentally. The second phosphine assisted pathway sees fluorine transfer to one of the phosphine ligands with formation of a metastable metallophosphorane intermediate from which either alkyl or fluorine transfer to the metal is possible. Both Pt-fluoride and Pt(alkyl)(fluorophosphine) products are therefore accessible via this route. The calculation of Nova et al., Figure 33 and Figure 34, highlight the central role of metallophosphorane species, either as intermediates or transition states, in aromatic C-F bond activation. In addition, the similar computed barriers for all three processes suggest that Pt-fluoride species should be accessible. This is confirmed experimentally by the reaction of [Pt(PR₃)₂] species (R = isopropyl (ⁱPr), cyclohexyl (Cy), and cyclopentyl (Cyp)) with 2,3,5-trifluoro-4-(triflu oromethyl)pyridine to give cis-[Pt(F){2-C₅NHF₂(CF₃)}(PR₃)₂]. These species subsequently convert to the trans isomers, either thermally or photochemically.

Moxey et al. [34] studied the reactions of a series of fluorinated bis(aryl)platinum(II) complexes, cis-[Pt(Ar_F)₂(1,5-C₆H₁₀)] (Ar_F = p-C₆HF₄, p-C₆(OMe)F₄ or C₆H₂F₃-2,4,6) with the four-membered gallium(I) or indium(I) heterocycles, [E{[N(Ar)]₂CN(C₆H₁₁)₂}] (E = Ga or In, Ar = C₆H₃ ⁱPr₂-2,6) under various stoichiometries. These yielded either the 2:1 complexes, cis-[Pt(Ar_F)₂{Ga{[N(A r)]₂CN(C₆H₁₁)₂}}₂], trans-[Pt(C₆H₂F₃-2,4,6)₂-{Ga{[N(Ar)]₂CN(C₆H₁₁)₂}}₂] and trans-[Pt(Ar_F)₂{In{[N(Ar)]₂CN(C₆H₁₁)₂}}₂], or the 3:1 complexes, trans-[Pt(Ar_F)₂{In{[N(Ar)]₂CN(C₆H₁₁)₂}}₃] (Ar_F = p-C₆HF₄ (22) or p-C₆(OMe)F₄ (23)), all of which were crystallographically characterized. The differing outcomes of these reactions are explained in terms of the lesser nucleophilicity and greater electrophilicity of the indium heterocycle relative to its gallium counterpart (Fig. 35). As show on Figure 36, in the solid state, the 3:1 indium complexes exhibit strong intramolecular In---F interactions which are indicative of their heterocyclic ligands displaying a "Lewis amphoteric" behaviour.

Silver

Decken and coworkers [35] examined the synthetically useful solvent-free silver(I) salt Ag[Al(pftb)₄] (pftb = OC(CF₃)₃). The salt was prepared by metathesis reaction of Li[Al(pftb)₄] with Ag[SbF₆] in liquid SO₂. The solvated complexes [Ag(OSO)][Al(pftb)₄] (24), [Ag(OSO)₂][SbF₆], and [Ag(CH₂Cl₂)₂][SbF₆] were prepared and isolated by special techniques at low temperatures and structurally characterized by single-crystal X-ray diffraction. The SO₂ complexes provide the first examples of coordination of the very weak Lewis base SO₂ to silver(I).The SO₂ molecule in [Ag(OSO)][Al(pftb)₄] is η¹-O coordinated to Ag⁺, while the SO₂ ligands in [Ag(OSO)₂][SbF₆] bridge two Ag⁺ ions in an η²-O,O' (trans, trans) manner.

Cadmium

Knapp and Mews [36] studied the ability of the sulfur-nitrogen-carbon bicycle F₃CCN₅S₃ to act as a donor towards transition metal cations. F₃CCN₅S₃ forms complexes with [M(SO₂)₂](AsF₆)₂ (M = Co, Cu, Zn, Cd) in the ratio 2:1 with the composition [M(F₃ CCN₅S₃)₂(OSO)₂(FAsF₅)₂] (M = Co, Zn), [Cu(F₃CCN₅S₃)₂(μ-F)(μ-F₂AsF₄)]₂, and [Cd(F₃CCN₅S₃)(μ-F₃CCN₅S₃)(η²-F₂AsF₄)₂]₂ (25) in liquid sulfur dioxide.

The reaction with the larger and softer cadmium(II) cation results in a dinuclear complex that contains terminal and bridging F₃CCN₅S₃ ligands. Terminal ligands coordinates through the bridging nitrogen atom of F₃CCN₅S₃. Each bridging ligand coordinates two cadmium centers through two nitrogen atoms and, in addition, a third weak bonding interaction between one fluorine atom of the trifluoromethyl substituent and a cadmium(II) center is observed as shown in Figure 37. Bidentate hexafluoroarsenate ions complete the coordination sphere around the metal center, to give a coordination number of eight.

 

Final remarks

Across the periodic table a wide variety of transition metals are capable of forming the elusive carbon-fluorine-metal molecular fragment under the appropriate conditions. Table 1 shows the more relevant structural parameters —distances and angles— as well as elements on the cyclic fragment and their number. Whether these fragments will end up activating the carbon fluorine bond either under a catalytic or stoichiometric regime, generally depends on very subtle experimental details. On the other hand, as expected, elegant and powerful theoretical calculations have shown new insights and pointed towards new targets on this area fueling the synergic experimental-theoretical research loop.

A broad picture would show that the driving forces behind the chemistry of the carbon-fluorine-metal bonds comes either from radical species, concerted oxidative addition, reductive elimination, fluorine substitution, ancillary ligand participation or a combination of these reaction mechanisms.

In general, the fluoride affinity of the highly electrophilic early transition metals tends to preclude their use in catalysis and the work described here for the metals of groups 3 to 5 show this tendency. Accordingly, the search for further developments has been directed to the use of low valent electron-rich transition metals.

As for ancillary ligand participation, several experiments result in relatively stable structures with the carbon-fluorine-metal bond originating from the ligand whereas calculations have shown that a phosphine-assisted C-F activation mechanism is possible with electron-rich precursors. The barrier associated with the phosphine-assisted process seems to be however, very similar to that for conventional oxidative addition and the balance between these processes is extremely delicate.

Several ligand based systems show promise and should be further studied as model compounds for systematic studies directed toward catalytic C-F bond activation processes.

From this work it is evident that significant progress has consistently been made in the area of metal-assisted C-F bond activation, early efforts typically employed forcing conditions and obtained low yields. C-F activation can now be accomplished under extremely mild conditions using a suitable transition-metal complex. However, considering the examples described in this paper, undoubtedly, the next challenge appears to be the activation of saturated perfluorocarbons and future efforts ought to profit from improved C-F bond cleavage processes directed toward the ultimate functionalization of the C-F bond.

 

Acknowledgement

We are grateful to DGAPA-UNAM-IN217611, to CONACYT-177498 and VIEP (ARCS-NAT-12-G) for support.

 

References

1. O'Hagan, D. Chem. Soc. Rev. 2008, 37, 308-319.         [ Links ]

2. Clot, E.; Megret, C.; Eisenstein, O; Perutz, R. N. J. Am. Chem. Soc. 2009, 131, 7817-7827.         [ Links ]

3. Furuya,T.; Klein, J.E.M.N.; Ritter T. Synth. 2010, 11, 1804-1821.         [ Links ]

4. Clot, E.; Eisenstein, O.; Jasim, MacGregor, N. S.; McGrady, J. E.; Perutz R. N. Acc. Chem. Res. 2011, 44, 333-348.         [ Links ]

5. Sun, A. D.; Love, J. A. Dalton Trans. 2010, 39, 10362-10374.         [ Links ]

6. Partyka, D. V. Chem. Rev. 2011, 111, 1529-1595.         [ Links ]

7. Crespo, M. Organometallics 2012, 31, 1216-1234.         [ Links ]

8. Nova, A.; Mas-Ballesté, R.; Lledós, A. Organometallics 2012, 31, 1245-1256.         [ Links ]

9. a) Kiplinger, J. L.; Richmond T. G. and Osterberg, C. E. Chem. Rev. 1994, 94, 373-431.         [ Links ] b) Torrens H.; Coord. Chem. Rev., 2005, 249, 19571985.         [ Links ] c) Perutz, R. N. and Braun, T.Transition Metal-mediated C-F Bond Activation. In Comprehensive Organometallic Chemistry III,         [ Links ] d) Crabtree R. H. and Mingos, M. P. ed., Elsevier, Oxford, England, 2007; pp 725-758.         [ Links ] e) Meike Reinhold, M.; McGrady, J. E.; and Perutz R. N. J. Amer. Chem. Soc. 2004, 126, 5268-5276.         [ Links ]

10. Special Issue: Fluorine in Organometallic Chemistry. Organometallics. Special issue Published February 27, 2012.         [ Links ]

11. Bouwkamp, M. W.; Budzelaar, P. H. M.; Gercama, J.; Del Hierro Morales, I.; de Wolf, J.; Meetsma, A.; Troyanov, S. I.; Teuben, J. H.; Hessen, B. J. Amer. Chem. Soc. 2005, 127, 14310-14319.         [ Links ]

12. Li, X.; Nishiura, M.; Mori, K.; Mashiko, T.; Hou, Z. Chem. Commun. 2007, 4137-4139.         [ Links ]

13. Lara-Sanchez, A.; Rodriguez A.; Hughes D.L.; Schormann M.; Bochmann M., J. Organomet. Chem. 2002, 663, 63-69.         [ Links ]

14. Hayes, P.G.; Welch G. C.; H. Emslie D. J.; Noack C. L.; Piers W. E.; Parvez M. Organometallics 2003, 22, 1577-1579.         [ Links ]

15. Lavanant, L.; Chou, T. Y.; Chi, Y.; Lehmann, C. W.; Toupet, L.; Carpentier, J. F. Organometallics 2004, 23, 5450-5458.         [ Links ]

16. Andino, J. G.; Fan, H.; Fout, A. R.; Bailey, B. C.; Baik, M-H.; Mindiola, D. J. J. Organomet. Chem. 2011, 696, 4138-4146.         [ Links ]

17. Wondimagegn, T.; Xu, Z.; Vanka, K.; Ziegler, T. Organometallics 2005, 24, 2076-2085.         [ Links ]

18. Gade, L. H.; Memmler, H.; Kauper, U.; Schneider, A.; Fabre, S.; Bezougli, I.; Galka, M. L. C.; Scowen, I. J.; McPartlin, M. Chem. Eur. J. 2000, 6, 692-708.         [ Links ]

19. Giesbrecht, G. R.; Gordon, J. C.; Clark, D. L.; Hijar, C. A.; Scott, B. L.; Watkin, J. G., Polyhedron 2003, 22,153-163.         [ Links ]

20. a) Arndt, P.; Jager-Fiedler, U.; Klahn, M.; Baumann, W.; Spannenberg, Aa; Burlakov, V. V.; Rosenthal, U. Angew. Chem. Int. Ed. 2006, 45, 4195-4198.         [ Links ] b) Arndt, P.; Baumann, W.; Spannenberg, A.; Rosenthal, U.; Burlakov, V. V.; Shur, V. B. Angew. Chem. Int. Ed. 2003, 42, 1414-1418.         [ Links ]

21. Herrmann, H.; Lloret Fillol, J.; Gehrmann, T.; Enders, M.; Wadepohl, H.; Gade, L. H. Chem. Eur. J. 2008, 14, 8131-8146.         [ Links ]

22. Schrock, R. R.; Adamchuk, J.; Ruhland, K.; Pia L; Lopez H. Organometallics 2003, 22, 5079-5091.         [ Links ]

23. Liu, W.; Welch, K.; Trindle, C. O.; Sabat, M.; Myers, W. H.; Harman, W. D. Organometallics, 2007, 26, 2589-2597.         [ Links ]

24. Beltran, T. F, Feliz, M.; Llusar, R.; Mata, J. A.; Safont, V. S. Organometallics 2011, 30, 290-297.         [ Links ]

25. Huang, D.; Renkema, K. B.; Caulton, K. G. Polyhedron 2006, 25, 459-468.         [ Links ]

26. Ritter, T.; Day, M. W.; Grubbs, R. H. J. Am. Chem. Soc. 2006, 128, 11768-11769.         [ Links ]

27. Dugan, T. R.; Goldberg, J. M.; Brennessel, W. W.;Holland, P. L. Organometallics 2012, 31, 1349-1360.         [ Links ]

28. Erhardt, S.; Macgregor, S. A. J. Am. Chem. Soc. 2008, 130, 15490-15498.         [ Links ]

29. Choi, J.; Wang, D. Y.; Kundu, S.; Choliy, Y.; Emge, T. J.; Krogh-Jespersen, K.; Goldman, A. S. Science2011, 332, 1545-1548.         [ Links ]

30. Schaub, T.; Fischer, P.; Steffen, A.; Braun, T.; Radius, U.; Mix, A. J. Am. Chem. Soc. 2008, 130, 9304-9317.         [ Links ]

31. Yandulov, D. V. Tran, N. T. J. Am. Chem. Soc., 2007, 129, 1342-1358.         [ Links ]

32. Furuya, T.; Benitez, D.; Tkatchouk, E.; Strom, A. E.; Tang, P.; Goddard, W. A. III, Ritter, T. J. Am. Chem. Soc. 2010, 132, 3793-3807.         [ Links ]

33. Nova, A.; Erhardt, S.; Jasim, N. A.; Perutz, R. N.; Macgregor, S. A.; McGrady, J. E.; Whitwood, A. C. J. Am. Chem. Soc. 2008, 130, 1549915511.         [ Links ]

34. Moxey, G. J.; Jones, C.; Stasch, A.; Junk, P. C.; Deacon, G. B.; Woodul W. D.; Drago P. R. Dalton Trans., 2009, 2630-2636.         [ Links ]

35. Decken, A.; Knapp, C.; Nikiforov, G. B.; Passmore, J.; Rautiainen, J. M.; Wang, X.; Zeng, X. Chem.-Eur.J. 2009, 15, 65046517.         [ Links ]

36. Knapp, C.; Mews, R. Eur. J. Inorg. Chem. 2005, 3536-3542.         [ Links ]