Donor-Activated Lithiation and Sodiation of Trifluoromethylbenzene

Article pubs.acs.org/Organometallics

Donor-Activated Lithiation and Sodiation of Trifluoromethylbenzene: Structural, Spectroscopic, and Theoretical Insights Jennifer A. Garden,† David R. Armstrong,† William Clegg,‡ Joaquin García-Alvarez,§ Eva Hevia,*,† Alan R. Kennedy,† Robert E. Mulvey,*,† Stuart D. Robertson,† and Luca Russo‡ †

WestCHEM, Department of Pure and Applied Chemistry, University of Strathclyde, Glasgow G1 1XL, Scotland, U.K. School of Chemistry, Newcastle University, Newcastle upon Tyne, NE1 7RU, U.K. § Laboratorio de Compuestos Organometálicos y Catálisis (Unidad Asociada al CSIC), Departmento de Química Orgánica e Inorgánica, Instituto Universitario de Química Organometálica “Enrique Moles”, Facultad de Química, Universidad de Oviedo, E-33071 Oviedo, Spain ‡

S Supporting Information *

ABSTRACT: Aiming to shed new light on the stability and constitution of the organometallic intermediates involved in direct ortho-metalation processes, using trifluoromethylbenzene (1) as a case study, this paper investigates the deprotonation of 1 using group 1 alkyl bases tBuLi and nBuNa in the presence of the Lewis donors TMEDA (N,N,N′,N′-tetramethylethylenediamine), THF, and PMDETA (N,N,N′,N″,N″-pentamethyldiethylenetriamine). A systematic and comprehensive study combining structural, spectroscopic, and theoretical studies reveals that these donors strongly influence the final outcome of the reactions, not only by activating the alkali-metal bases and facilitating deprotonation of 1 but also by tuning the regioselectivity of the reaction. Thus, while using tBuLi/TMEDA, ortho-metalation of 1 is preferred, switching to THF gives a complex mixture of products with the meta-regioisomer being the major species crystallizing from hexane solution. This donor effect is significantly reduced when nBuNa is employed, as ortho-regioselectivity is observed almost exclusively using THF, TMEDA, or PMDETA. DFT calculations computing the relative energies of the ortho-, meta-, and pararegioisomers obtained from these metalating systems have also been carried out. Reinforcing the experimental findings, these theoretical studies show that although in all cases the product of ortho-metalation is the most thermodynamically preferred, the energy difference between the three possible modeled regioisomers is much larger for the Na systems than for the Li ones. The structures of key reaction intermediates [(TMEDA)·Li(C6H4-CF3)]2 (2), [(TMEDA)·Na(C6H4-CF3)]2 (3), and [(PMDETA)· Na(C6H4-CF3)]2 (4) have been elucidated by X-ray crystallographic studies. All compounds exhibit a similar dimeric arrangement with a four-atom core constituting a {MCMC} ring. Interestingly for Na derivatives 3 and 4 unusual Na···F dative interactions are found, which appear to contribute to the overall stability of these compounds, therefore favoring ortho-metalation of 1, as the meta or para structures do not contain these additional interactions.

■

INTRODUCTION Directed ortho-metalation (DoM), in particular directed ortholithiation, is well established as one of the most powerful synthetic preludes to achieving the direct functionalization of aromatic molecules.1−3 Pioneering studies by Wittig4 and Gilman5 on lithiation of anisole constituted a milestone in the development of this synthetic methodology. Since then, the lithiation of this benchmark molecule in DoM chemistry has been much studied using NMR spectroscopy,6 semiempirical calculations,6,7 kinetic isotope effects,8 and X-ray crystallography.9 Acting as a directing metalating group (DMG), the electron-rich MeO group provides a Lewis basic coordination point for an incoming organometallic reagent, bringing it into close proximity to an adjacent C−H unit, enabling orthometalation, though this mechanism still remains somewhat controversial with this particular substrate.10 This phenomenon © XXXX American Chemical Society

is often referred to as a complex-induced proximity effect (CIPE). 11 Another important factor in achieving this regioselectivity is the inductive effect of the MeO substituent, which enhances the acidity of the ortho-H atoms. For substrates bearing electron-withdrawing groups (e.g., as in haloaromatic substrates),12−14 this can play a major role in directing the metalation. Surprisingly, despite the extensive employment of organolithium reagents within DoM chemistry, there are only a few known examples of structurally characterized ortho-lithiated intermediates.15,16 This paper deals with the aromatic substrate trifluoromethylbenzene, 1. Frequently utilized as a solvent in synthesis,17 1 is also a common intermediate in the synthesis of Received: August 1, 2013

A

dx.doi.org/10.1021/om4007664 | Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

Scheme 1. General Synthesis of Metalated Aromatic Products 2−5

pharmaceuticals and agrochemicals.18 Its CF3 substituent would be expected to promote DoM mainly on account of its strong electron-withdrawing capability. Indeed, ortho-metalation of 1 has been observed using nBuLi in refluxing ether solution, with subsequent quenching by carbon dioxide giving an unexpected mixture of ortho- and meta-carboxylic acids in a combined 48% yield and in a ratio of 5:1, respectively, along with negligible quantities of the para-isomer.19 Intermediate in size between an i Pr and a tBu group, it is thought that the steric bulk of the CF3 group disfavors ortho-metalation.20−22 Overcoming this, through applying the bimetallic “superbase” nBuLi·KOtBu (commonly written as LIC-KOR) in THF solvent at −78 °C, 23−30 to the metalation of 1, Schlosser observed regioselective ortho-deprotonation.31 However, no metal-active intermediates of these reactions prior to electrophilic interception were isolated or characterized within this work, and so the exact nature of the organometallic species involved is not yet known. We recently showed that the direct zincation of 1 at room temperature using the zincate [(TMEDA)Na(μ-TMP)(μ-tBu)Zn(tBu)] (TMP is 2,2,6,6-tetramethylpiperidide) results in a mixture of ortho-, meta-, and para-zincated ring products in a respective ratio of 20:11:1.32 Also studies by Uchiyama of the alumination of 1 through reaction with aluminate [LiAl(TMP)(iBu)3] at 0 °C revealed the meta-product to be the predominant regioisomer (20% isolated yield), along with 10% ortho- and 10% para-metalation products.33 Intrigued by these precedents and building on our previous metalation studies of 1 using bimetallic bases, here we provide a systematic study of the deprotonation of 1 using the alkyl metalating reagents tBuLi and nBuNa. By combining spectroscopic and structural characterization of the organometallic intermediates with theoretical calculations, new insights have been gained into the constitution and stability of the organometallic species involved and the effect that the use of donors such as THF, TMEDA (N,N,N′,N′-tetramethylethylenediamine), and PMDETA (N,N,N′,N″,N″-pentamethyldiethylenetriamine) can play in tuning the reaction regioselectivity.

of either TMEDA or PMDETA. Reactions were performed at −78 °C (Scheme 1), and transferral of the resultant solutions to the freezer (at −28 °C) allowed isolation of the crystalline solids [(TMEDA)·Li(C6H4-CF3)]2 (2), [(TMEDA)·Na(C6H4CF3)]2 (3), [(PMDETA)·Na(C6H4-CF3)]2 (4), and [(THF)2· Li(C6H4-CF3)]2 (5) after 1 day, in decent crystalline yields of 67−86%. Compounds were characterized in solution using multinuclear (1H, 7Li{1H}, 13C{1H}, and 19F{1H}) NMR spectroscopy and the structures of 2−4 in the solid state elucidated by X-ray crystallographic studies. Unfortunately, severe disorder in the crystallographic structure of 5 made it impossible to refine the molecular structure satisfactorily; hence no structural data can be presented. Solid-State Molecular Structures. Although solid-state structures of polar organometallic compounds can sometimes substantially differ from the structures found in solution, crystallographic analysis offers structural information that can disclose a depth of insight into a reaction pathway.34 Accordingly, metalated intermediate complexes (with respect to the normal synthetic protocol of quenching these complexes with electrophiles) 2−4 were synthesized in a crystalline form suitable for the aforementioned X-ray crystallographic determinations. Exhibiting a dimeric arrangement with the general formula [(donor)·M(C6H4-CF3)]2, (M = Li for 2; M = Na for 3 and 4), the molecular structure of each compound contains a four-membered core constituting a [MCMC] ring, with the metal corners engaged in dative bonding with Lewis donors, namely, one molecule of chelating TMEDA in 2 and 3 and PMDETA in 4. In each case, this [MCMC] ring is strictly (2 and 3) or essentially planar (4) [sum of endocyclic angles 360° for 2, 360° for 3, and 357° for 4]. This cyclic, dimeric motif is widespread in alkali-metal chemistry, with known examples including alkali-metal alkyl,35,36 alkali-metal aryl,37 alkali-metal amide,38−41 and alkali-metal alkoxide42−45 complexes. Despite this, a surprisingly limited number of metalated DoM intermediates have been isolated and structurally characterized. A rare example comes from the ortho-sodiation of N,Ndimethylaniline by nBuNa, which led to the isolation and structural characterization of dimeric [{(TMEDA)·Na(o-C6H4NMe2)}2].46 Lithiated intermediate 2 contains essentially identical distorted tetrahedral Li centers, each coordinated by two ortho-metalated trifluoromethylbenzene units and a chelating TMEDA (Figure 1). A crystallographic 2-fold axis in space group C2/c runs through both Li1 and Li2 and the centers of the C−C bonds of the TMEDA ligands, and so the asymmetric unit contains half of the dimeric molecule (Z′ = 0.5). Equivalent within experimental error, the two distinct C− Li bond lengths in 2 [Li1−C1, 2.252(2) Å; Li2−C1, 2.251(2)

■

RESULTS AND DISCUSSION Alkali Metalation of Trifluoromethylbenzene, 1. We started our studies by preparing the metalated products of the reactions of tBuLi or nBuNa with 1 in the presence of the Lewis base THF, TMEDA, or PMDETA. Following the addition of either TMEDA or THF to the 1:1 stoichiometric reaction of t BuLi with 1, a pale yellow solution was formed. The reaction of freshly prepared nBuNa with 1 produced a white suspension, which immediately changed color to brown upon introduction B


Organometallics

Article

Figure 1. Molecular structure of 2 with displacement ellipsoids at the 50% probability level and hydrogen atoms omitted for clarity. Symmetry transformation used to generate equivalent atoms (primed): 1 − x, y, 0.5 − z. Selected bond lengths (Å) and angles (deg): Li1−N1, 2.238(3); Li1−C1, 2.252(2); Li2−N2, 2.244(3); Li2− C1, 2.251(2); N1−Li1−N1′, 82.54(12); N1−Li1−C1, 112.52(5); N1−Li1−C1′, 119.99(5); N1′−Li1−C1′, 112.52(5); C1−Li1−C1′, 108.05(15); N2−Li2−N2′, 82.10(12); N2−Li2−C1, 98.26(5); N2− Li2−C1′, 137.89(6); N2′−Li2-C1, 137.89(6); Li1−C1−Li2, 71.93(10).

Figure 2. Molecular structure of 3 with displacement ellipsoids at the 50% probability level and hydrogen atoms omitted for clarity. Symmetry transformation used to generate equivalent atoms labeled ′: x, −y, z; ″: 1 − x, y, −z; ‴: 1 − x, −y, −z. Selected bond lengths (Å) and angles (deg): Na1−F1, 2.6415(7); Na1−N1, 2.5234(10); Na1− C1, 2.5580(12); F1−C7, 1.3610(12); F2−C7, 1.3412(18); Na1−C1− Na1′, 76.00(4); F1−Na1−F1″, 158.18(4); F1−Na1−N1, 102.39(3); F1−Na1−N1″, 95.29(3); F1−Na1−C1, 68.38(4); F1−Na1−C1′, 97.78(4); N1−Na1−N1″ 71.77(4); N1−Na1−C1, 161.63(3); N1− Na1−C1′, 92.78(3); C1−Na1−C1′, 104.00(4).

Å] are comparable to those in the related lithium-unsubstituted phenyl complexes [{Li(C6H5)}2]∞ (average Li−C bond length = 2.282 Å)47 and [(TMEDA)·Li(C6H5)]2 (average Li−C bond length = 2.243 Å).48 It should be noted that no significant Li···F contacts are observed within 2, with the shortest contact distance being Li2···F1 [3.2920(3) Å], which is essentially equivalent to the sum of the van der Waals radii for Li and F (3.29 Å).49,50 The related aryl lithium, [(THF)3·Li(C6HF4)], has a nonbonding Li···F interatomic distance of 3.448(4) Å.51 In contrast, shorter Li···F bond lengths were observed for [2,4,6-(CF3)3C6H2Li·Et2O]2 (2.252 Å)52 and [(THF)·{LiF(NtBu)SitBu2}2]2 (1.85 and 2.09 Å).53 Within 2, there is a dihedral angle of 77.31(6)° between the aryl ring plane and the Li1C1Li2C1′ ring plane. Presenting a rare example of an organometallic compound involving Na−F interactions, the crystal structure of 3 contains two essentially identical though crystallographically independent centrosymmetric molecules, although for brevity only the geometry of one (Figure 2) need be discussed. Each dimer lies about a 2/m center with the 2-fold rotation axis running through the metal sites and the centers of the C−C bonds of the TMEDA and the mirror plane coincident with the plane of the aromatic group. The two symmetry-equivalent Na centers of each dimer of 3 are bridged via the ortho-C of deprotonated 1. Two additional dative Na−F interactions add further stabilization. The distorted octahedral geometry of Na [cis angles ranging from 68.38(4)° to 104.00(4)° and trans angles ranging from 158.18(4)° to 161.63(3)°] is completed by a bidentate TMEDA ligand. Exemplifying the ability of Na to raise its coordination number higher than Li, each Na center is hexacoordinate, while in comparison the Li centers within analogous 2 are only tetracoordinate. The increase in coordination number and size of the alkali metal in 3 when compared to 2 translates into a more acute bite angle of the chelating TMEDA ligand [N1−Na1−N1″, 71.77(4)° in 3 vs 82.32° average for 2].

Upon moving from Li to its larger congener Na, the CF3 substituent of 1 is able to approach the metal center more closely, facilitating Na−F interactions that are further assisted by the higher coordination number achievable with Na. While Li···F interactions are not observed within 2, the Na1−F1 bond length of 2.6415(7) Å (Table 1) lies significantly below the sum of the van der Waals radii for Na and F (3.74 Å),49,50 although it is greater than the sum of the relevant ionic radii (2.31 ppm).54 This interatomic distance lies within the scope of Na− F bond lengths for other crystallographically characterized compounds,32 ranging from 2.175(3) Å for [Cr3(dpa)4F2Na(H2O)BF4]BF4·CH3OH55 at the short end of the scale56 (where dpa = 2,2′-dipyridylamine) to 3.000(2) Å for [Na(pHC6F4N(CH2)2NEt2)2Na(TMEDA)].57 The Na1−F1−C7 bond angle of 109.69(8)° also suggests that one lone pair of electrons of a tetrahedral, sp3-hybridized F atom datively coordinates to Na. Experimentally substituting bidentate TMEDA by tridentate PMDETA led to the isolation of 4 (Figure 3), where now only one CF3 substituent interacts with the Na center. Aside from this interaction, which is 0.033 Å longer than those in 3, the distorted octahedral Na center bonds to three N atoms of PMDETA and to one ortho-C of two aryl anions. The Na−C bonds are inequivalent [Na1−C10, 2.7458(15) Å; Na1−C20, 2.6946(15) Å]. Forming a short interaction, the Na1−F2 bond length is 2.6750(9) Å, while the Na1−F3 separation distance is elongated at 3.2326(12) Å (Table 1). Dimeric 4 crystallizes in the monoclinic space group P21/m with half a molecule in the asymmetric unit. The addition of the asymmetric PMDETA ligand results in loss of the 2-fold rotation symmetry seen in 2 and 3 with intramolecular symmetry now restricted to the mirror plane that is coincident with the aromatic ring planes, as shown in Figure 3. Although severe disorder in the X-ray crystallographic analysis of the THF-solvated lithium aryl 5 prevented complete elucidation of the molecular structure, 1H NMR spectroscopic C


Organometallics

Article

Table 1. Comparison of Selected Bond Lengths and Bond Angles in Complexes 2−4 bond lengths and angles compound [(TMEDA)·Li(C6H4-CF3)]2 (2) [(TMEDA)·Na(C6H4-CF3)]2 (3) [(PMDETA)·Na(C6H4-CF3)]2 (4)

M−F (Å)

M−C (Å)

C−M−C (deg)

M−C−M (deg)

M−F−C (deg)

3.2920(3) 3.4447(3) 2.6415(7) 2.6750(9) 3.2326(12)

2.252(2) 2.251(2) 2.5580(12) 2.7458(15) 2.6946(15)

108.05(15) 108.09(14) 104.00(4) 102.20(4)

71.93(10)

101.583(3) 90.593(3) 109.69(8) 113.55(5) 107.37(6)

76.00(4) 75.44(5) 77.13(5)

analysis of the crystalline material divulged that of the three potential regioisomers, the meta-metalated derivative of 1 was the prevailing product [where the relative ratio of ortho:meta:para isomers is 1:10:3 (vide inf ra)]. Solution NMR Spectroscopic Studies. Analysis of metalated intermediates 2−5 via multinuclear (1H, 7Li{1H}, 13 C{1H}, and 19F{1H}) NMR spectroscopy in C6D6 solvent confirmed that in all cases a single C−H deprotonation of 1 has taken place (Table 2). For isolated crystals of 2−4, only resonances corresponding to the ortho-isomer were observed, reflecting the solid-state structures (vide supra). The salient feature for each compound was the observation of four multiplets in the aromatic part of the spectrum, which are shifted significantly downfield when compared with the resonances of precursor 1. For example, the chemical shift for the meta-H (labeled as Hd and positioned adjacent to the site of metalation) shifts significantly downfield, from 6.90 ppm in nonmetalated 1, to 8.32 ppm for 2, 8.43 ppm for 3, and 8.48 ppm for 4 (Table 2). Comparison of the 1 H NMR spectroscopic resonances attributed to TMEDA in 2 with that of noncoordinated TMEDA suggests that in C6D6 solution TMEDA remains coordinated to the electropositive metal center, mirroring the solid-state structure. This is manifested in a significant upfield shift for the corresponding NCH3 and NCH2 resonances, from 2.12 and 2.36 ppm, respectively, for free TMEDA to 1.71 and 1.64 ppm in 2. A similar trend is observed for 3, 4, and 5, with analogous upfield deviations in the resonances associated with TMEDA in 3, PMDETA in 4, and THF in 5 suggesting that these Lewis donors also remain coordinated, as would be expected in noncoordinating C6D6 solvent. Within 13C NMR spectra, varying the metal from Li in

Figure 3. Molecular structure of 4 with displacement ellipsoids at the 50% probability level and hydrogen atoms omitted for clarity. Symmetry transformation used to generate equivalent atoms (primed): x, 0.5 − y, z. Selected bond lengths (Å) and angles (deg): Na1−F2, 2.6750(9); Na1−N1, 2.6606(11); Na1−N2, 2.6824(12); Na1−N3, 2.6655(12); Na1−C10, 2.7458(15); Na1− C20, 2.6946(15); Na1−C20−Na1′, 77.13(5); Na1−C10−Na1′, 75.44(5); F2−Na1−N1, 148.42(3); F2−Na1−N2, 82.73(3); F2− Na1−N3, 82.17(3); F2−Na1−C20, 85.41(5); F2−Na1−C10, 65.38(4); N1−Na1−N2, 70.17(3); N1−Na1−N3, 102.01(4); N1− Na1−C20, 125.78(5); N1−Na1−C10, 99.25(4); N2−Na1−N3, 68.90(3); N2−Na1−C20, 154.84(5); N2−Na1−C10, 92.88(3); N3−Na1−C20, 87.58(4); N3−Na1−C10, 144.85(5); C20−Na1− C10, 102.20(4); Na1−F2−C16, 113.55(5).

Table 2. Aromatic Chemical Shifts in 1H NMR (400.03 MHz) Spectra of Products 2−5 in C6D6 Solution (where M = Li or Na)

compound CF3C6H5 (1) [(TMEDA)·Li(C6H4-CF3)]2 (2) [(TMEDA)·Na(C6H4-CF3)]2 (3) [(PMDETA)·Na(C6H4-CF3)]2 (4) [(THF)2·Li(o-C6H4-CF3)]2 (5-ortho) [(THF)2·Li(m-C6H4-CF3)]2 (5-meta) [(THF)2·Li(p-C6H4-CF3)]2 (5-para)

δ(Ha) 7.32 7.73 7.76 7.78 7.74 7.56 7.66

δ(Hb)

(d) (d) (d) (d) (d) (d) (d)

6.90 7.21 7.16 7.19 7.21 7.31 8.32 D

(t) (t) (t) (t) (t) (t) (d)

δ(Hc) 6.97 7.35 7.32 7.35 7.38 8.40

(t) (m) (t) (t) (t) (d)

δ(Hd) 8.32 8.43 8.48 8.31 8.63

(d) (d) (d) (d) (s)


Organometallics

Article

monitoring of a solution of crystalline 5 in C6D6 over 5 h revealed that the 1:10:3 ratio of ortho:meta:para regioisomers persisted. However, dissolution of 5 in deuterated THF solution induced a remarkably different outcome. Thus, as illustrated in Figure 4a, the aromatic resonances in the 1H NMR spectrum of 5 in this Lewis base donor solvent are considerably broad, and after 20 min in solution they show the almost complete conversion of both 5-meta and 5-para to 5ortho (Figure 4c). It is worthy of mention that within the context of DoM chemistry Wheatley et al. have found that when isolated crystals of the ortho-lithiated product of 2-ethylN,N-diisopropyl-1-benzamide are dissolved in deuterated THF, a rearrangement takes place in solution, to give the product of lateral lithiation, which in this case is most thermodynamically favored.16 DFT calculations support the proposal that this rearrangement occurs through an externally solvated, monomeric intermediate. In contrast, no changes were observed in the NMR spectra of crystalline 2−4 in either C6D6 or d8-THF solution. For completeness, we next monitored the regioselectivity of the lithiation of 1 over a period of 24 h through 1H NMR spectroscopic analysis of the in situ iodine quench products, to probe the possibility of a thermodynamically driven rearrangement. The results of this study (Table 4) demonstrate that the

2 to Na in congeneric 3 translates as a downfield shift of the C−M resonance, from 185.9 ppm to 192.6 ppm (Table 3), a difference of 6.6 ppm. Table 3. Metal−Carbon Chemical Shifts in 13C{1H} NMR (100.59 MHz) Spectra of Products 2−5 in C6D6 Solution compound

δ(C−M) ppm

[(TMEDA)·Li(C6H4-CF3)]2 (2) [(TMEDA)·Na(C6H4-CF3)]2 (3) [(PMDETA)·Na(C6H4-CF3)]2 (4) [(THF)2·Li(o-C6H4-CF3)]2 (5-ortho) [(THF)2·Li(m-C6H4-CF3)]2 (5-meta) [(THF)2·Li(p-C6H4-CF3)]2 (5-para)

185.9 192.6 194.4 185.9 179.5 178.7

It should be noted that, although the NMR spectra of isolated crystals of compounds 2−4 displayed the resonances of a single product resulting from ortho-deprotonation of 1, NMR spectroscopic analysis of the in situ reaction mixtures prior to crystallization showed that, in the case of the reaction of tBuLi/ TMEDA, small but significant amounts of meta and para lithiation products are present in solution (11% and 7%, respectively, Table S1). Contrastingly, switching to nBuNa as a metalating reagent in the reaction solutions producing 3 and 4 led to the almost exclusive formation of ortho-metalated product.58 Surprisingly given the results for 2−4, NMR spectroscopic analysis of isolated crystals of 5 showed that in C6D6 solution the major product is the meta-isomer (Figure 4d), followed by

Table 4. Effect of Reaction Time on the Yield and Ratio of Iodo-trifluoromethylbenzene Regioisomers

iodinated product yield (%)a entry 1 2 3

metal reagent t

BuLi t BuLi t BuLi

donor

reaction time (h)

ortho

THF THF THF

10 min 100 min 24 h

8 32 30

meta para 8 31 26

4 17 13

total 20 80 69

a

Conditions: CF3Ph (2 mmol), tBuLi (2 mmol), THF (4 mmol), hexane (10 mL). Yields were determined through addition of an internal standard (10 mol % hexamethylbenzene) to the iodinated product.

relative ratio of ortho-, meta-, and para-iodo-trifluoromethylbenzene isomers remains largely consistent over time, though the collective yield after only 10 min is considerably smaller than those after 100 min or 24 h. These findings show that the presence of excess Lewis basic THF solvent is required for the thermodynamically driven rearrangement to the ortho-product. It should be noted that the yields of the ortho-, meta-, and paraiodo-trifluoromethylbenzene (Table 4) give good agreement with those of the in situ-metalated aryl product (Table S1) obtained prior to iodine quenching. DFT Computational Studies. To attempt to quantify the energetics involved in the formation of these metalated species, the aforementioned solid-state and solution studies were augmented by a complementary gas phase theoretical study investigating the reaction of 1 (model I) with tBuLi/TMEDA, n BuNa/TMEDA, tBuLi/THF, or nBuNa/THF. Calculations were performed to probe the relative energies of the regioisomers of compounds 2 (model II), 3 (model III), 5

Figure 4. Aromatic region of 1H NMR spectra for deuterated THF solutions (a−c) and deuterated benzene solutions (d−f) of isolated crystals of 5 after (a) 0 min, (b) 80 min, (c) 5 h, (d) 0 min, (e) 80 min, and (f) 5 h.

para and ortho with a relative ratio of 10:3:1. Contrastingly, when the reaction was repeated without removal of the crystalline solid, the same analysis of the crude reaction mixture after stirring for 24 h revealed that, overall, the ortho-metalated product was the marginally preferred regioisomer (with an ortho:meta:para ratio of 3:2:1, Table S1). Given that the solid isolated is predominantly 5-meta, this suggests that the metaproduct is the least soluble and preferentially crystallizes from solution. Within the 13C{1H} NMR spectra, the attenuating influence of the electron-withdrawing CF3 substituent upon moving from 5-ortho, to 5-meta, to 5-para significantly impacts the chemical shift, with an upfield sequential progression from 185.9, to 179.5, to 178.7 ppm, respectively (Table 3). 1H NMR E


Organometallics

Article

Figure 5. Relative energies of DFT-modeled compounds II, III, V, and VI (picturing the energetically favored ortho- isomers).

kcal mol−1) than between those of the sodium analogues IIIortho, III-meta, and III-para (spanning 19.15 kcal mol−1), which supports the superior selectivity toward the ortho-isomer observed experimentally for the reaction of nBuNa/TMEDA with 1. This trend is mirrored to a less significant extent by models V and VI (where the energy difference of V-ortho, Vmeta, and V-para spans 11.25 kcal mol−1 compared with 15.44 kcal mol−1 for VI-ortho, VI-meta, and VI-para). Comparing geometrical parameters of model II with those of compound 2 shows a significant overestimation in the theoretical model for Li···F interactions [with lengths of 2.897 and 3.123 Å vs experimental values of 3.2920(3) and 3.4447(3) Å]. A similar but more modest effect is observed for Na···F contacts in III and 3 [2.519 vs 2.6415(7) Å]. This general trend predicts a similar overestimation for model V, where the Li···F interatomic distances [2.995 and 3.001 Å] are too long to be significant. Collectively, these DFT calculations reinforce the experimental observations within this investigation. However, the predominance of the meta-lithiated species (5-meta) in the solid state is surprising, as calculations reveal V-ortho to be the most

(model V), and [(THF)2·Na(C6H4-CF3)]2 (model VI), first utilizing the B3LYP59 functionals and the 6-311G (d,p)60 basis set (see Supporting Information for full details). The resultant optimized geometries were subjected to a frequency analysis, and the total energy computed by the DFT calculation was adjusted by inclusion of the zero-point energy contribution. These compounds were modeled as dimeric structures, to match the arrangements observed in the experimental structural studies. A comparison of the geometrical parameters calculated for II and III are in good agreement with those observed experimentally from X-ray determination of the solid-state structures (see Supporting Information). In each case the theoretical model resulting from ortho-deprotonation of 1 was determined to be the energetically preferred product. In both cases, the difference between ortho and meta is much larger (+7.48 and 15.47 kcal mol−1, respectively) than that between meta and para (+3.10 and 3.68 kcal mol−1, respectively), as shown in Figure 5. Comparison of the DFT calculations for II and III shows that there is a smaller energy difference between the regioisomers II-ortho, II-meta, and II-para (spanning 10.58 F


Organometallics

Article

energetically stable product, and indeed 5-ortho is the major product observed experimentally in the in situ reaction mixture. The predominance of 5-meta and 5-para in the solid state may originate with the relatively large dimerization energy gain associated with the formation of these species (−11.23 kcal mol−1 for V-meta compared to −9.32 kcal mol-1 for V-para and −5.18 kcal mol−1 for V-ortho). DFT modeling of V in the theoretical monomeric state shows the ortho-product to be more thermodynamically favorable than the meta- and paraproducts (by +9.37 and +9.76 kcal mol−1, respectively), which can be attributed in part to the presence of Li···F interactions. Although V-ortho remains the energetically preferred product upon moving to the more realistic dimeric model, the balance swings in favor of V-meta (+6.62 kcal mol−1). The relatively low dimerization energy associated with V-ortho is likely to be influenced by the disruption of the Li···F interaction present in the monomeric model as it is replaced by a secondary Li−C bond. NMR Spectroscopic Analysis: Electrophilic Quenching Studies with Iodine. After running the metalation reactions for 24 h at ambient temperature, in situ iodine quenching of reaction mixtures of compounds 2−5 was performed, followed by analysis via multinuclear (1H, 13C{1H}, and 19F{1H}) NMR spectroscopy in deuterated chloroform solvent. This showed the successful monometalation of 1 that led to the formation of iodo-trifluoromethylbenzene,32 with further two-dimensional (1H−1H COSY) NMR experiments confirming the presence of three regioisomers (ortho, meta, and para). As iodotrifluoromethylbenzene is light-sensitive,58 the reaction was protected from light throughout the quenching process. This collection of results (Table 5) discloses that the careful choice

It is a fact well exploited by synthetic chemists that the introduction of Lewis donors to polar organometallic reagents can greatly boost their reactivity.28 Thus in this study we found that no metalation of 1 occurs when neat tBuLi was administered in hexane solution (Table 5, entry 1). Exemplifying the essential requirement for the donor solvent, use of tBuLi in tandem with THF gave ortho-, meta-, and parametalated 1 in a relative ratio of 30:26:13 (entry 2, Table 5). Simply switching to the more basic TMEDA as the donor solvent improved the regioselectivity to a ratio of 65:10:6 (entry 3, Table 5), diminishing the quantities of meta- and para-isomers and promoting that of the thermodynamically favored ortho-product (Scheme 2). It is worth noting that the metalation of 1 using tBuLi and PMDETA was also investigated; however spectroscopic analysis of the quenched reaction mixture revealed that a complex mixture of products was obtained, eight different product signals being observed in the 19F{1H} NMR spectrum. Reflecting the absence of a reactivity-enhancing Lewis basic solvent, a poor yield of 8% of ortho-iodo-trifluoromethylbenzene was obtained upon treatment of 1 with nBuNa alone (Table 5, entry 4). Turning from Li to Na revealed that the formation of the ortho-products became much more favorable, in line with increasing donor strength (entries 5−7, Table 5); notably only 1% of meta- and para-iodo-trifluoromethylbenzene were observed for PMDETA (entry 7), and these regioisomers were absent completely when Lewis donor THF or TMEDA was used, though this was counterbalanced by diminished yields. This general trend of increasing favorability of the orthometalation product upon moving from Li to Na is likely to be influenced by the increased atomic radius and increased coordination requirements of Na facilitating Na−F interactions, thus driving the regioselectivity toward ortho-metalation.

Table 5. Effect of Varying the Metal Reagent and Donor Solvent on the Ratio of Iodo-trifluoromethylbenzene Regioisomers

■

CONCLUSIONS Using a systematic approach combining X-ray crystallographic and NMR spectroscopic studies with theoretical calculations, the metalation of trifluoromethylbenzene (1) by group 1 bases t BuLi and nBuNa in the presence of several Lewis bases, including TMEDA and THF, two of the most commonly used in organic synthesis, has been investigated. These studies have revealed not only that the use of these Lewis bases as additives is crucial for the metal−hydrogen exchange reactions to occur but also that they can play a major role in tuning the regioselectivity of the overall process. This is particularly noticeable for lithiation reactions, where using chelating diamine TMEDA yielded compound [(TMEDA)·Li(o-C6H4CF3)]2 (2-ortho), where the metalation has occurred exclusively at the ortho-position of 1. Contrastingly when monodentate THF is employed, compound [(THF)2·Li(m-C6H4-CF3)]2 (5meta) was the isolated crystalline product. NMR spectroscopic analysis of 5 revealed a mixture of the three possible regioisomers for the deprotonation of 1, showing a preference for the meta- and paraproducts, with the ortho-isomer being the minor species present in solution. Highlighting the complex behavior of these organometallic intermediates in solution, these studies also revealed that the observed ratio of the three regioisomers can be greatly influenced by the solvent employed. Thus, while in C6D6 solutions this ratio remains constant over time, when donor solvent d 8-THF is used, complete equilibration of 5-meta and 5-para to 5-ortho is observed after 20 min at room temperature. DFT studies predict 5-ortho to be

iodinated product yield (%)a entry 1 2 3 4 5 6 7

metal alkyl t

BuLi BuLi t BuLi n BuNa n BuNa n BuNa n BuNa t

donor

ortho

meta

para

total

none THF TMEDA none THF TMEDA PMDETA

0 30 65 8 43 72 80

0 26 10 0 0 0 1

0 13 6 0 0 0 1

0 69 81 8 43 72 82

a

Yields were determined through addition of an internal standard (10 mol % hexamethylbenzene) to the iodinated product.

of metal reagent and Lewis basic solvent enables us to improve the regioselectivity of metalation. The best yield (82%) with high selectivity (80% ortho-directed) was obtained via nBuNa/ PMDETA. In general, the yields of the iodinated products for entries 2/3 and 6/7 (Table 5) exhibit little deviation from the metalated yields (see Supporting Information, Table S1). “Even if suspected to produce sometimes a “placebo effect”,10 N,N,N′,N′-tetramethylethylenediamine (TMEDA) is often added to n-butyllithium as a reactivity enhancing complexand.”3 G


Organometallics

Article

Scheme 2. Regioselective Metalation of 1 Using tBuLi Activated by THF or TMEDA

the most thermodynamically preferred regiosomer. This suggests that the CIPE, that is, the complexation of an organometallic reagent to the functional group of the substrate prior to proton transfer, plays little or no role in determining the regioselectivity when CF3 is present as the DMG. This differs significantly from the case with other, stronger DMGs.11 Using nBuNa as a metalating reagent also has important consequences for the regioselectivity of the reaction, showing that independent of the donor employed (THF, TMEDA, or PMDETA), metalation occurs almost exclusively at the orthoposition of 1. Supporting this superior chemoselectivity, DFT studies comparing the relative stability of the three possible regioisomers for deprotonation of 1 by tBuLi/TMEDA or n BuNa/TMEDA showed that although in both cases the orthoproducts are the most energetically preferred, the calculated energy differences between the ortho-, meta-, and pararegioisomers for the nBuNa/TMEDA reactions are significantly larger than those observed for the tBuLi/TMEDA system. Finally, new structural insights on the constitution of the organometallic intermediates involved in these reactions have shed some light on the regioselectivities observed. Thus, sodium derivatives [(TMEDA)·Na(C6H4-CF3)]2 (3) and [(PMDETA)·Na(C6H4-CF3)]2 (4) contain stabilizing dative Na···F interactions, which appear to be key in the stabilization of the ortho-metalated products.

■

Figure 6. Labeling scheme for metalated 1.

temperature. Subsequently, TMEDA (2 mmol, 0.30 mL) was added to the mixture at −78 °C, and the reaction mixture was stirred for 30 min at room temperature to give a white precipitate. Briefly heating the suspension with a heat gun afforded a colorless solution, from which colorless crystals of 2 were deposited upon cooling this solution to −28 °C (yield 0.41 g, 76%). 1 H NMR (400.13 MHz, C6D6, 300 K): δ 1.64 and 1.71 [s, 32H, CH2 and CH3-TMEDA respectively], 7.21 [t, 2H 3JHH = 7.7 Hz, HbC6H4], 7.35 [m, 2H, Hc-C6H4], 7.73 [d, 2H, 3JHH = 7.62 Hz, HaC6H4], 8.32 ppm [d, 2H, 3JHH = 6.4 Hz, Hd-C6H4]. 13C{H NMR (100.63 MHz, d6-benzene, 300 K): δ 45.3 [CH3 of TMEDA], 56.7 [CH2 of TMEDA], 122.7 [q, 3JCF = 3.0 Hz, C(Ha)-C6H4], 124.0 [s, C(Hb)-C6H4], 127.7 [s, C(Hc)-C6H4], 130.6 [br m, CF3], 141.8 [q, 2 JCF = 23.6 Hz, i-C6H4], 142.5 [s, C(Hd)-C6H4], 185.9 ppm [br s, C(Li)-C6H4]. 7Li{1H} NMR (155.50 MHz, d6-benzene, 300 K, reference LiCl in D2O at 0.00 ppm): δ 2.0 ppm. 19F{1H} NMR (376.40 MHz, d6-benzene, 300 K): δ −60.4 ppm. Preparation of [(TMEDA)·Na(C6H4-CF3)]2, 3. Mr = 568.60 g. Under Ar atmosphere, CF3Ph (2 mmol, 0.25 mL) was added to a suspension of nBuNa (2 mmol, 0.16 g) in hexane (10 mL) at −78 °C. The reaction mixture was subsequently stirred for 30 min at room temperature. TMEDA (2 mmol, 0.30 mL) was then added to the mixture at −78 °C, and the mixture was stirred for 30 min at room temperature, which produced a white solid. Briefly heating the suspension with a heat gun produced a brown solution, from which crystalline 3 was afforded at −28 °C (yield 0.49g, 86%). 1H NMR (400.13 MHz, d6-benzene, 300 K): δ 1.61 [s, 32H, CH2 and CH3TMEDA], 7.16 [t, 2H, 3JHH = 7.5 Hz, Hb-C6H4], 7.32 [t, 2H, 3JHH = 6.7 Hz, Hc-C6H4], 7.76 [2H, d, 3JHH = 7.8 Hz, Ha-C6H4], 8.43 ppm [d, 2H, 3JHH = 6.5 Hz, Hd-C6H4]. 13C{H} NMR (100.63 MHz, d6benzene, 300 K): δ 44.8 [CH3-TMEDA], 56.7 [CH2-TMEDA], 121.8 [q, 3JCF = 3.7 Hz, C(Ha)-C6H4], 121.9 [s, C(Hb)-C6H4], 123.4 [s, C(Hc)-C6H4], 130.2 [br m, CF3], 142.4 [q, 2JCF = 23.8 Hz, i-C6H4],

EXPERIMENTAL SECTION

All reactions were performed under a protective argon atmosphere using standard Schlenk techniques. Hexane and THF were obtained from Aldrich and dried by heating to reflux over sodium metal and distilled under nitrogen prior to use. nBuNa was prepared according to a literature method.61 TMEDA, PMDETA, and trifluoromethylbenzene were obtained from Aldrich, distilled from CaH2, and stored over 4 Å molecular sieves. Other chemicals were obtained from Aldrich or Alfa Aesar and were used as received. NMR spectra were recorded on a Bruker AV3 or AV400 MHz spectrometer. Correlations between protons and carbon atoms were obtained by using COSY and HSQC NMR spectroscopic methods. Due to the acute air sensitivity of these compounds, satisfactory C, H, N analyses could not be obtained. Preparation of [(TMEDA)·Li(C6H4-CF3)]2, 2. Mr = 536.50 g. CF3Ph (2 mmol, 0.25 mL) was added to a solution of tBuLi (2 mmol, 1.18 mL of a 1.7 M solution in hexane) in hexane (10 mL) at −78 °C, and the reaction mixture was stirred for 30 min at ambient H


Organometallics

Article

× 15 mL). Magnesium sulfate was used to dry the combined organic layers. After filtration, the solvent was removed in vacuo, and the crude residue was spiked with 10 mol % hexamethylbenzene [0.0324 g, 0.2 mmol for 2 mmol scale reactions]. 1H NMR spectroscopic analysis was performed, and the relative yields of ortho-, meta-, and para-iodotrifluoromethylbenzene were determined by relative integration. Crystal Structure Determination. Single-crystal data were measured at low temperature with Oxford Diffraction or Agilent instruments using monochromated Mo Kα radiation (λ = 0.71073 Å). The structures were refined to convergence against F2 using all independent reflections and by full-matrix least-squares using the SHELXL-97 program.62 Selected parameters are given in the Supporting Information, and full details are given in the deposited cif files. CCDC reference numbers 952115−952117 contain the supplementary crystallographic data for this paper. These data can be obtained free of charge from the Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.

144.0 [s, C(Hd)-C6H4], 192.5 ppm [q, 3JCF = 13.7 Hz, C(Na)-C6H4]. 19 1 F{ H} NMR (376.40 MHz, d6-benzene, 300 K): δ −61.5 ppm. Preparation of [(PMDETA)·Na(C6H4-CF3)]2, 4. Mr = 682.79 g. CF3Ph (2 mmol, 0.25 mL) was added to a suspension of nBuNa (2 mmol, 0.16 g) in hexane (10 mL) at −78 °C, and the reaction mixture was stirred for 30 min. PMDETA (2 mmol, 0.42 mL) was subsequently added to the reaction mixture at −78 °C, which was then stirred at ambient temperature for a further 30 min. The resultant brown solution was transferred to the freezer, whereupon a crop of colorless crystals was deposited overnight (0.46 g, 67% yield). 1H NMR (400.13 MHz, d6-benzene, 300 K): δ 1.59 [s, 6H, N-CH3 PMDETA], 1.86 [s, 16H, N-CH2 PMDETA], 1.91 [s, 24H, N(CH3)2 PMDETA], 7.19 [br s, 2H, Hb-C6H4], 7.35 [br s, 2H, Hc-C6H4], 7.78 [br s, 2H, Ha-C6H4] 8.48 ppm [br s, 2H, Hd-C6H4]. 13C{H} NMR (100.63 MHz, d6-benzene, 300 K): δ 42.6 [N-CH3 PMDETA], 45.3 [N(CH3)2 PMDETA], 54.7 [N-CH2 PMDETA], 57.0 [N-CH2 PMDETA], 122.0 [C(Ha)-C6H4], 122.8 [C(Hb)-C6H4], 127.1 [C(Hc)-C6H4], 142.1 [q, 2JCF = 23.9 Hz, i-C6H4], 144.0 [C(Hd)-C6H4], 194.4 ppm [q, 3JCF = 15.0 Hz, C(Na)-C6H4]. 19F{1H} NMR (376.40 MHz, d6-benzene, 300 K): δ −60.7 ppm. Preparation of [(THF)2·Li(C6H4-CF3)]2, 5. Mr = 592.51 g. Under Ar atmosphere, CF3Ph (2 mmol, 0.25 mL) was added to a solution of t BuLi (2 mmol, 1.18 mL of a 1.7 M hexane solution) in hexane (10 mL) at −78 °C. The reaction mixture was subsequently stirred for 30 min at ambient temperature. Following this THF (4 mmol, 0.32 mL) was added to the reaction mixture at −78 °C, producing a yellow solid. After stirring for 5 min at ambient temperature, a primrose solution was produced. Refrigeration of this solution at −28 °C afforded a crop of colorless crystals after 18 h (yield 0.41g, 69% yield). The NMR of the crystals reveals a mixture of ortho-, meta-, and para-products in a respective ratio of 1:10:3. As labile THF is removed under vacuum upon isolation of crystalline 5, ideal NMR integration values could not be obtained. 1H NMR (400.13 MHz, d6-benzene, 300 K): δ 1.20 [m, 9H, β-CH2 THF], 3.22 ppm [m, 9H, α-CH2 THF]. 5-ortho: δ 7.21 [t, 3JHH = 7.5 Hz, 1H, Hb-C6H4], 7.38 [t, 3JHH = 6.8 Hz, 1H, Hc-C6H4], 7.74 [d, 3JHH = 8.0 Hz, 1H, Ha-C6H4], 8.31 ppm [d, 1H, Hd-C6H4]. 5-meta: δ 7.31 [t, 3JHH = 7.2 Hz, 1H, Hb-C6H4], 7.56 [d, 3JHH = 7.7 Hz, 1H, Ha-C6H4], 8.40 [d, 3JHH = 6.1 Hz, 1H, HcC6H4], 8.63 [s, 1H, Hd-C6H4] ppm. 5-para: δ 7.66 [d, 3JHH = 7.05 Hz, 2H, Ha-C6H4], 8.32 [d, 2H, Hb-C6H4] ppm. 13C{H} NMR (100.63 MHz, d6-benzene, 300 K): δ 25.2 [β-CH2-THF], 68.1 ppm [α-CH2THF]. 5-ortho: δ 122.8 [q, 3JCF = 3.6 Hz, C(Ha)-C6H4], 125.4 [s, C(Hb)-C6H4], 128.7 [s, C(Hc)-C6H4], 142.7 [s, C(Hd)-C6H4], 185.9 [br, C(Li)-C6H4]. Signals for CF3 and i-C6H4 were not detected. 5meta: δ 122.1 [br, C(Ha)-C6H4], 125.7 [s, C(Hb)-C6H4], 138.1 [br, C(Hd)-C6H4] 142.4 [q, 2JCF = 26.0 Hz, i-C6H4], 146.1 [s, C(Hc)C6H4], 179.5 [s, C(Li)-C6H4]. Signal for CF3 not detected. 5-para: δ 121.6 [s, C(Ha)-C6H4], 142.5 [s, C(Hb)-C6H4], 178.7 [br, C(Li)C6H4]. Signals for CF3 and i-C6H4 were not detected. 7Li{1H} NMR (d6-benzene, 155.50 MHz, 300 K, reference LiCl in D2O at 0.00 ppm): δ 2.0 ppm. 19F{1H} NMR (376.40 MHz, d6-benzene, 300 K): 5-ortho: δ −61.1 ppm (s), 5-meta: δ −61.4 ppm (s), 5-para: δ −61.6 ppm (s). Iodine Quench Studies: General Procedure. To a solution of tBuLi (2 mmol, 1.18 mL of a 1.7 M hexane solution for compounds 2 and 5) or a suspension of nBuNa (2 mmol, 0.16 g for compounds 3 and 4) in hexane (10 mL) was injected aromatic substrate 1 via syringe at −78 °C. The cold bath was then removed, and the reaction mixture was allowed to stir for 90 min at ambient temperature, prior to the addition of a donor solvent (TMEDA, 2 mmol, 0.30 mL for compounds 2 and 3; PMDETA, 2 mmol, 0.42 mL for 4; THF, 4 mmol, 0.32 mL for 5) at −78 °C. Subsequently, the reaction mixtures were stirred for 24 h at ambient temperature prior to treatment with a freshly prepared solution of iodine (4 mmol, 2 mL of a 2 M THF solution) at −78 °C. Following the addition of iodine, the reaction was stirred at ambient temperature for 1 h. As iodo-trifluoromethylbenzene is lightsensitive,58 the reaction was protected from light throughout the quenching process. A 10 mL amount of a saturated solution of NH4Cl was added along with the addition of saturated Na2S2O3 until bleaching occurred (20 mL). The organic layer was separated from the aqueous layer and the aqueous layer was washed with diethyl ether (4

■

ASSOCIATED CONTENT

S Supporting Information *

NMR spectra and computational details, X-ray data in crystallographic file (CIF) format for compounds 2, 3, and 4. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was generously supported by George Fraser (scholarship award to J.A.G.), the Royal Society/Wolfson Foundation (research merit award to R.E.M.), the Royal Society and the European Research Council (awards to E.H.), Royal Society of Edinburgh (BP Trust Fellowship to S.D.R.), and the MICINN and the European Social Fund (Ramón y Cajal to J.G.-A.)

■

REFERENCES

(1) Snieckus, V. Chem. Rev. 1990, 90, 879−933. (2) Clayden, J. Organolithiums: Selectivity for Synthesis, 1 ed.; Elsevier Science Ltd.: Oxford, UK, 2002; Vol. 23. (3) Organometallics in Synthesis; Schlosser, M., Ed.; John Wiley and Sons, Inc.: Hoboken, NJ, 2013. (4) Wittig, G.; Fuhrman, G. Chem. Ber. 1940, 73, 1197−1218. (5) Gilman, H.; Bebb, R. L. J. Am. Chem. Soc. 1939, 61, 109−112. (6) Bauer, W.; Schleyer, P. v. R. J. Am. Chem. Soc. 1989, 111, 7191− 7198. (7) Saá, J. M.; Deyá, P. M.; Suñer, G. A.; Frontera, A. J. Am. Chem. Soc. 1992, 114, 9093−9100. (8) Stratakis, M. J. Org. Chem. 1997, 62, 3024−3025. (9) Harder, S.; Boersma, J.; Brandsma, L.; van Mier, G. P. M.; Kanters, J. A. J. Organomet. Chem. 1989, 364, 1−15. (10) Collum, D. B. Acc. Chem. Res. 1992, 25, 448−454. (11) Whisler, M. C.; MacNeil, S.; Snieckus, V.; Beak, P. Angew. Chem., Int. Ed. 2004, 43, 2206−2225. (12) Wittig, G. Naturwissenschaften 1942, 30, 696−703. (13) Huisgen, R.; Rist, H. Naturwissenschaften 1954, 41, 358−359. (14) Wittig, G. Angew. Chem., Int. Ed. Engl. 1965, 4, 731−737. (15) Clayden, J.; Davies, R. P.; Hendy, M. A.; Snaith, R.; Wheatley, A. E. H. Angew. Chem., Int. Ed. 2001, 40, 1238−1240. I


Organometallics

Article

(49) Pauling, L. The Nature of the Chemical Bond, 3rd ed. ed.; Cornell University Press: Ithaca, NY, 1960. (50) Bondi, A. J. Phys. Chem. 1964, 68, 441−451. (51) Kottke, T.; Sung, K.; Lagow, R. J. Angew. Chem., Int. Ed. Engl. 1995, 34, 1517−1519. (52) Stalke, D.; Whitmire, K. H. J. Chem. Soc., Chem. Commun. 1990, 833−834. (53) Stalke, D.; Keweloh, N.; Klingebiel, U.; Noltemeyer, M.; Sheldrick, G. M. Z. Naturforsch., B: Chem. Sci. 1987, 42, 1237−1244. (54) Pauling, L. General Chemistry: An Introduction to Descriptive Chemistry and Modern Chemical Theory; W.H. Freeman: San Francisco, 1947. (55) Li, H.; Lee, G.-H.; Peng, S.-M. J. Mol. Struct. 2004, 707, 179− 186. (56) Vedavathi, B. M.; Vijayan, K. Acta Crystallogr. B 1977, 33, 946− 948. (57) Deacon, G. B.; Forsyth, C. M.; Junk, P. C.; Wang, J. Chem. Eur. J. 2009, 15, 3082−3092. (58) For the nBuNa/TMEDA mixture, compound 3-ortho is formed exclusively in solution, whereas for nBuNa/PMDETA, 1% meta- and 1% para-metalated product are observed. (59) Kohn, W.; Becke, A. D.; Parr, R. G. J. Phys. Chem. 1996, 100, 12974−12980. (60) McLean, A. D.; Chandler, G. S. J. Chem. Phys. 1980, 72, 5639− 5648. (61) Schade, C.; Bauer, W.; Schleyer, P. v. R. J. Organomet. Chem. 1985, 295, C25−C28. (62) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112−122.

(16) Armstrong, D. R.; Boss, S. R.; Clayden, J.; Haigh, R.; Kirmani, B. A.; Linton, D. J.; Schooler, P.; Wheatley, A. E. H. Angew. Chem., Int. Ed. 2004, 43, 2135−2138. (17) Ogawa, A.; Curran, D. P. J. Org. Chem. 1997, 62, 450−451. (18) Banks, R. E. Organofluorine Chemicals and their Industrial Applications; Ellis Horwood LTD: Chichester, 1979. (19) Roberts, J. D.; Curtin, D. Y. J. Am. Chem. Soc. 1946, 68, 1658− 1660. (20) Charton, M. J. Org. Chem. 1976, 41, 2217−2220. (21) de Riggi, I.; Virgili, A.; de Moragas, M.; Jaime, C. J. Org. Chem. 1995, 60, 27−31. (22) Wolf, C.; König, W. A.; Roussel, C. Liebigs Ann. 1995, 781−786. (23) Morton, A. A.; Patterson, G. H.; Donovan, J. J.; Little, E. L. J. Am. Chem. Soc. 1946, 68, 93−96. (24) Morton, A. A.; Holden, M. E. T. J. Am. Chem. Soc. 1947, 69, 1675−1681. (25) Morton, A. A.; Marsh, F. D.; Coombs, R. D.; Lyons, A. L.; Penner, S. E.; Ramsden, H. E.; Baker, V. B.; Little, E. L.; Letsinger, R. L. J. Am. Chem. Soc. 1950, 72, 3785−3792. (26) Lochmann, L.; Pospíšil, J.; Lím, D. Tetrahedron Lett. 1966, 7, 257−262. (27) Schlosser, M. J. Organomet. Chem. 1967, 8, 9−16. (28) Hsieh, H. L.; Wofford, C. F. J. Polym. Sci., Part A 1969, 7, 449− 460. (29) Schlosser, M. Pure Appl. Chem. 1988, 60, 1627−1634. (30) Schlosser, M.; Choi, J. H.; Takagishi, S. Tetrahedron 1990, 46, 5633−5648. (31) Schlosser, M.; Katsoulos, G.; Takagishi, S. Synlett 1990, 747− 748. (32) Lying in the midst of this range, the related sodium zincate [(TMEDA)Na(TMP)(C6H4-CF3)Zn(tBu)] possesses a Na···F bond length of 2.435 Å, as modeled through DFT calculations for the orthozincated isomer. Armstrong, D. R.; Blair, V. L.; Clegg, W.; Dale, S. H.; García-Á lvarez, J.; Honeyman, G. W.; Hevia, E.; Mulvey, R. E.; Russo, L. J. Am. Chem. Soc. 2010, 132, 9480−9487. (33) These moderate yields were achieved using an excess of the aluminate base (2.2 equiv of base per 1 equiv of substrate). Uchiyama, M.; Naka, H.; Matsumoto, Y.; Ohwada, T. J. Am. Chem. Soc. 2004, 126, 10526−10527. (34) Granitzka, M.; Pöppler, A.-C.; Schwarze, E. K.; Stern, D.; Schulz, T.; John, M.; Herbst-Irmer, R.; Pandey, S. K.; Stalke, D. J. Am. Chem. Soc. 2012, 134, 1344−1351. (35) Setzer, W. N.; Schleyer, P. v. R. Adv. Organomet. Chem. 1985, 24, 353−451. (36) Nichols, M. A.; Williard, P. G. J. Am. Chem. Soc. 1993, 115, 1568−1572. (37) Bauer, W.; Müller, G.; Pi, R.; Schleyer, P. v. R. Angew. Chem., Int. Ed. Engl. 1986, 25, 1103−1104. (38) Gregory, K.; Schleyer, P. v. R.; Snaith, R. Adv. Inorg. Chem. 1991, 37, 47−142. (39) Barr, D.; Clegg, W.; Cowton, L.; Horsburgh, L.; Mackenzie, F. M.; Mulvey, R. E. J. Chem. Soc., Chem. Commun. 1995, 891−892. (40) Kennedy, A. R.; Klett, J.; O’Hara, C. T.; Mulvey, R. E.; Robertson, G. M. Eur. J. Inorg. Chem. 2009, 2009, 5029−5035. (41) Garden, J. A.; Kennedy, A. R.; Mulvey, R. E.; Robertson, S. D. Dalton Trans. 2011, 40, 11945−11954. (42) Seebach, D. Angew. Chem., Int. Ed. Engl. 1988, 27, 1624−1654. (43) Hevia, E.; Henderson, K. W.; Kennedy, A. R.; Mulvey, R. E. Organometallics 2006, 25, 1778−1785. (44) Calvo, B.; Davidson, M. G.; Garcι ́a-Vivó, D. Inorg. Chem. 2011, 50, 3589−3595. (45) Garden, J. A.; Kennedy, A. R.; Mulvey, R. E.; Robertson, S. D. Chem. Commun. 2012, 48, 5265−5267. (46) Armstrong, D. R.; Balloch, L.; Hevia, E.; Kennedy, A. R.; Mulvey, R. E.; O’Hara, C. T.; Robertson, S. D. Beilstein J. Org. Chem. 2011, 7, 1234−1248. (47) Dinnebier, R. E.; Behrens, U.; Olbrich, F. J. Am. Chem. Soc. 1998, 120, 1430−1433. (48) Thoennes, D.; Weiss, E. Chem. Ber. 1978, 111, 3157−3161. J


Donor-Activated Lithiation and Sodiation of Trifluoromethylbenzene

Recommend Documents