Electronic Structure Analysis and Reactivity of the Bimetallic Bis-Titanocene Vinylcarboxylate Complex, [(Cp₂Ti)₂(O₂C₃TMS₂)]

Abstract

Multimetallic redox cooperativity features heavily in both bioinorganic and synthetic reactions. Here, the electronic structure of the bimetallic Ti/Ti complex 11, [(Cp₂Ti)₂(O₂C₃TMS₂)] has been revisited with EPR, confirming a predominantly Ti^III/Ti^III electronic structure. Reactions of 11 with 2,6-dimethylphenyl isocyanide (CNXyl), TMSCl, MeI, and BnCl were explored, revealing differential redox chemistry of the bimetallic core. In reactions with CNXyl and TMSCl, the metallacyclic Ti^III center remained unperturbed, with reactions taking place at the pendent κ²(O,O)-titanocene fragment, while reaction with MeI resulted in remote oxidation of the metallacyclic Ti center, indicative of a cooperative redox process. All structures were studied via X-ray diffraction and EPR spectroscopic analysis, and their electronic structures are discussed in the context of the covalent bond classification (CBC) electron counting method.

Graphical Abstract:

Introduction

The insertion of CO₂ into the Ti–C bonds of titanocene alkene, alkyne, and benzyne complexes has been well studied over the last 50 years. For example, in 1971 Vol’pin reported CO₂ insertion into transiently-generated benzyne complexes (IM1) from Cp₂TiPh₂ 1 (Cp = C₅H₅; Ph = C₆H₅), yielding a titanabenzofuranone 2 (Figure 1, top),[1] while Bercaw later reported CO₂ insertion into Cp*₂Ti(η²-C₂H₄) 3 (Cp* = C₅Me₅) to form titanalactone 4 (Figure 1, middle).[2] Later, a number of additional CO₂ insertion products with titanocene alkyne adducts Cp^R₂Ti(R′C=CR′) 5 ( Cp^R = C₅H₅, C₅Me₅; R’ = Ph, TMS), were reported wherein the final products—either binuclear carbonate complex 6, titanafuranone 7-9, or the bridged binuclear σ-alkenylcarboxylate 10 and 11—are determined by the structure of the metallocene (Figure 1, bottom).

Figure 1.

Examples of CO₂ insertion into Ti–C bonds of titanocene complexes.[1–6]

Complex 11 is interesting from an electron-counting perspective: one could reasonably draw two potential major resonance contributors: a Ti^III/Ti^III contributor (Figure 2, A), or a Ti^IV/Ti^II contributor (Figure 2, B).[3] At first pass, each configurational extreme has some merit according to Green’s Covalent Bond Classification (CBC) method:[7, 8] each Ti in A has an ML₅X₃ configuration, which is somewhat uncommon (found in approximately 9% of reported Ti compounds); while B has one Ti in the very common ML₄X₄ configuration (49%) along with one Ti in the uncommon ML₆X₂ configuration (4%).[9]

Figure 2.

Limiting resonance structures of binuclear 11.

Magnetic susceptibility of 11 revealed a magnetic moment of 1.6 μ_B per Ti atom, close to the spin-only value expected for d¹ Ti^III and consistent with configuration A where the two centers are ferromagnetically coupled; however, reaction of 11 with O₂ resulted in exclusive reaction with the carboxylate-bound Ti center, yielding oxidized titanafuranone 13 along with oxidation products of “Cp₂Ti^II” (Figure 3, top).[3] The mechanism of this oxidation reaction was rationalized through the participation of resonance form B. This type of bi- or multimetallic redox cooperativity is common in many bioinorganic systems, [10, 11] and is an emerging tool for reaction control in synthetic reactions,[12, 13] as well. In an effort to better understand the electronic structure of this interesting compound and the communication between the two Ti centers, we herein report the reactivity of 11 toward neutral donor CNXyl (Xyl = 2,6,-(CH₃)₂C₆H₃) and electrophiles TMSCl, MeI, and BnCl (Figure 3, bottom), along with characterization of the resultant products through X-ray diffraction and EPR studies. These reactions reveal substrate-dependent oxidation of the bimetallic Ti₂ core, demonstrating that the effect of electronic coupling between the two metal centers on subsequent reactivity is variable. In memory of Malcolm Green, we explicitly discuss each structure using the CBC method, showing the importance of simple electron counting techniques in building chemical intuition about organometallic structures.

Figure 3.

Reaction of 11 with O₂[3] and summary of reactions presented herein.

Results and Discussion

In this study, we sought to further examine the electronic structure of bimetallic 11 through EPR analysis and with additional reactivity studies. Compound 11, although first reported by Shur, was not examined using EPR. The X-band EPR spectrum of 11 (and all subsequent compounds) was collected as a toluene glass at 4 K (Figure 4, top; See Supporting Information for full spectrum). The spectrum of 11 displays the features consistent with a dinuclear S = 1 Ti^III/Ti^III complex: the six features typical of ΔM_s = 1 transitions and a weak half-field signal for the formally forbidden ΔM_s = 2 transition. This EPR data is consistent with earlier magnetic susceptibility measurements[3] (1.6 μ_B per Ti atom, close to the spin-only value for d¹ Ti^III). The unassigned peaks in the spectrum may be due to solvent coordination equilibria and monomeric impurities that are consistent with other reports of Ti^III/Ti^III EPR.[14–17]

Figure 4.

Top: Experimental X-band EPR spectrum of 11 as a toluene glass collected at 4 K. The six features typical of ΔM_S = 1 transitions for triplet system are observed (denoted by *). Bottom: Ti^…Ti distances of 11 obtained from EPR analysis and XRD data.

Using the obtained g-tensors and zero-field splitting parameter D, the Ti^…Ti separation (i.e. R) was calculated according to Eq. 1, where the value of D was taken as an approximation for the dipole—dipole zero-field parameter D_d.[18] The obtained g-tensors and zero-field splitting parameters for compound 11 were g₁ = 1.991, g₂ = 1.990, g₃ = 1.978 and D = 0.0256 cm⁻¹, E=0.00567 cm⁻¹, respectively. Using Eq. 1, the Ti^…Ti distance in 11 was calculated to be 4.06 Å. For comparison, the Ti^…Ti distance in 11 was measured to be 4.36 Å by crystallography (Figure 4, bottom). Similar differences between EPR- and crystallographically-derived Ti^…Ti distances have been reported.[15, 18–22]

$$D_d=\frac{-{\beta }^2}{3R^3}[2g_z^2+\frac{g_x^2+g_y^2}{2}]$$ Eq. 1

Adduct Formation with Isocyanide.

Next, the reaction of 11 with CNXyl (Xyl = 2,6,-(CH₃)₂C₆H₃), TMSCl, MeI, and BnCl were explored. When 11 was treated with CNXyl, the green solution immediately turned red, indicative of isocyanide coordination and the formation of 14 (Figure 5, top). Interestingly, analogous reaction attempts with the more electron-rich cyclohexyl isocyanide (CNCy) resulted in no reaction. Crystallization of 14 from a concentrated pentane solution at −35 °C resulted in a 49% isolated yield of 14. Much like 11, 14 could potentially exist as one of two limiting configurations: Ti^III/Ti^III (Figure 5, middle), or a Ti^IV/Ti^II (Figure 5, bottom). One might expect that the mixed-valent Ti^IV/Ti^II (ML₆X₂, 4%[9]) configuration may play a role, given that the π-acceptor nature of the isocyanide ligand could stabilize a low-valent Ti^II species. However, XRD and EPR analysis of 14 (see below) indicate that the most probable electronic structure is similar to that of 11, wherein each Ti is in an ML₅X₃ Ti^III d¹ configuration (Figure 5, middle).

Figure 5.

Top: Reaction of 11 with CNXyl results in formation of 14. Middle: Electron counting diagram for 14 assuming a Ti^III/Ti^III configuration. Bond lengths determined by XRD (see Figure 6 and Table 1). Bottom: Electron counting diagram for 14 assuming an alternate Ti^IV/Ti^II configuration.

The structure of 14 is presented in Figure 6, and relevant bond lengths are presented in Figure 5 (middle) and Table 1. The structure of 14 contains features similar to its precursor 11, where the local structure of the κ²(O,C) titanacycle of Ti1 is preserved. However, the Ti2 titanocene that was formerly κ²(O,O) in 11 is now κ¹-O coordinated to the vinylcarboxylate, wherein the new CNXyl L donor has displaced the L donor from the carboxylate. The Ti1–O1 (2.0824(9) Å) and Ti2–O2 (2.098(1) Å) bond distances, as well as the C1–O1 (1.268(2) Å) and C1-O2 (1.277(2) Å) are similar, indicating resonance throughout the bridging carboxylate moiety. Overall, the bond distances in 14 remain similar to the distances in 11 with the exception of Ti1–O1, which shows significant contraction from 2.178(4) to 2.0824(9) Å. This contraction could be expected as O1 is no longer participating in bonding to both Ti centers.

Figure 6.

Displacement ellipsoid plot of 14 with ellipsoids drawn at the 50% probability level. Hydrogen atoms and disorder of one Cp ring were omitted for clarity. Selected bond distances (Å) are found in Table 1.

Table 1.

Summary of XRD bond distances (Å) of compounds 11, 14, 15, and 16.

	11	14	15	16
First Ti Center
Ti1–cent1	2.075	2.073	2.065	2.049
Ti1–cent2	2.091	2.092	2.070	2.050
Ti1–O1	2.178(4)	2.0824(9)	2.1289(9)	1.960(1)
Ti1–C3	2.256(6)	2.242(1)	2.248(1)	2.184(2)
Second Ti Center
Ti2–cent3	2.049	2.053	-	2.071
Ti2–cent4	2.064	2.052	-	2.071
Ti2-O1	2.257(4)	-	-	-
Ti2–O2	2.094(4)	2.098(1)	-	2.138(1)
Ti2-CNR	-	2.168(1)	-	-
Ti2-I	-	-	-	2.9103(8)
Carboxylate
C1–O1	1.310(7)	1.268(2)	1.246(2)	1.293(2)
C1–O2	1.266(6)	1.277(2)	1.326(2)	1.250(2)
C1–C2	1.488(8)	1.500(2)	1.480(1)	1.504(2)

The C–N bond lengths of previously reported Ti-CNXyl complexes with formal oxidation states ranging from +2 to +4 are summarized in Table 2. As expected from changes in π-backbonding contributions, Ti^IV-CNXyl complexes have longer Ti–C bonds and shorter C–N bonds whereas Ti^II complexes have shorter Ti–C bonds and longer C–N bonds. The C–N and Ti–CNXyl bond lengths in 14, 1.159(2) Å and 2.168(1) Å, respectively, are similar to a previously reported Ti^III complex Cp*₂Ti(CNXyl)(C₂TMS),[23] which has a C–N bond distance of 1.162(3) Å and Ti–CNXyl distance of 2.130(2) Å. This suggests that the Ti2 in 14 has a formal +3 oxidation state. The ν(CN) frequencies of these Ti complexes are summarized in Table 2. Compound 14 shows a strong ν(CN) stretch at 2038 cm⁻¹. This frequency is significantly lower than that of the Ti^III Cp*₂Ti(CNXyl)(C₂TMS)[23] (2087 cm⁻¹) and even lower than a reported Ti^II complex Cp*₂Ti(CNXyl)(H₂C₃NXyl)[24] (2066 cm⁻¹), indicating that electronic communication between the two Ti centers complicates concrete formal oxidation state assignments. The Ti–Cp(centroid) distances could suggest that these are Ti^III metal centers, 2.052-2.092 Å, however, Ti^IV–Cp(centroid) literature values[25–28] vary upwards to 2.060 Å which render these distances too close to reliably assign the metal oxidation state using solely XRD metrical parameters.

Table 2.

Bond length ranges of Ti complexes containing CNXyl and ν(CN) stretching frequencies.

	Ti Formal Ox. St.	Ti–C Bond Range (Å)	C–N Bond Range (Å)	ν(CN) (cm−1)	C.N.a	Ref
Cp*2Ti(CNXyl)(H2C3NXyl)b	2	2.099(2)	1.167(3)	2066	8	[24]
14	3	2.168(1)	1.159(2)	2038	8	this work
Cp*2Ti(CNXyl)(C2TMS)b	3	2.130(2)	1.162(3)	2087	8	[23]
Cp*Ti(CNXyl){MeC(NiPr)2}(N2CPh2)]b	4	2.232(3)	1.160(3)	2124	7	[29]
[TiCl4(CNXyl)]2	4	2.235(6)	1.147(8)	2210	6	[30]

^aCoordination number; cyclopentadienyl ligands occupy 3 coordination numbers

^bCp* = C₅Me₅

The EPR spectrum (4 K, toluene glass) of 14 is presented in Figure 7 (top). Similar to 11, the EPR spectrum of 14 displays features consistent with a dinuclear S = 1 Ti^III/Ti^III complex (the six features typical of ΔM_s = 1 transitions and a weak half-field signal for the formally forbidden ΔM_s = 2 transition). The unassigned peaks in the spectrum may be due to solvent coordination equilibria and monomeric impurities that are consistent with other reports of Ti^III/Ti^III EPR.[14–17]. Thus, the combined EPR and X-ray structural data point to the Ti^III/Ti^III electronic structure (Figure 7, bottom) as the major contributor. Analysis of the Ti^…Ti distance in 3 was carried out using g₁ = 1.983, g₂ =1.988, g₃ =1.981, and D = 0.0100 cm⁻¹, E= 0.00190 cm⁻¹. The EPR-derived Ti^…Ti separation in 3 was calculated to be 5.54 Å, which is in reasonable agreement with the crystallographically measured Ti^…Ti distance of 5.39 Å.[18, 31]

Figure 7.

Top: Experimental X-band EPR spectrum of 14 as a toluene glass collected at 4 K. The six features typical of ΔM_S = 1 transitions for triplet system are observed (denoted by *). Bottom: Ti^…Ti distances of 14 obtained from EPR analysis and XRD data.

Silylation Using TMSCl.

When 11 was treated with excess TMSCl, the reaction solution remained green and an orange precipitate was generated. After removing the orange precipitate, green crystals were grown from a concentrated pentane solution at −35 °C resulting in an 88% isolated yield of 15 (Figure 8, top). XRD and EPR analysis of 15 (see below) indicate that the product is a monometallic complex where the Ti center is in an ML₅X₃ Ti^III d¹ configuration (Figure 8, bottom). The orange precipitate was recrystallized and was later confirmed as Cp₂TiCl₂ via unit-cell analysis with XRD.

Figure 8.

Top: Reaction of 11 with TMSCl results in formation of 15. Bottom: Electron counting diagram for 15. Bond lengths determined by XRD (see Figure 9).

The structure of 15 is presented in Figure 9 with relevant bond lengths highlighted in Figure 8 (bottom) and Table 1. The metrical parameters about the Ti1 center of the monometallic complex 15 are similar to the Ti1 center in compounds 11 and 14. This suggests that the Ti1 in 15 is also Ti^III. Additionally, the C1–O1 bond length contracts from 1.310(7) to 1.246(2) Å consistent with a C=O double bond, while the C1–O2 bond length increases from 1.266(6) to 1.326(2) Å consistent with a C–O single bond. Unlike in the reaction of 11 with CNXyl, however, the reaction of 11 with TMSCl does not preserve the bimetallic structure and instead silylates the vinylcarboxylate, likely eliminating a “Cp₂TiCl” moiety, which undergoes oxidation to Cp₂TiCl₂.

Figure 9.

Displacement ellipsoid plot of 15 with ellipsoids drawn at the 50% probability level. Hydrogen atoms were omitted for clarity.

The EPR spectrum of 15 was collected at 4 K as a toluene glass (Figure 10). Unlike that of 11 and 14, the EPR spectrum of 15 is of a mononuclear Ti^III compound 15 displays an isotropic EPR spectrum with g = 1.984 and coupling constants of A_Ti = 20.0 MHz, A_{H(proximal-TMS)} = 20.7 MHz, and A_{H(distal-TMS)} = 2.3 MHz (Figure 10).

Figure 10.

Experimental X-band EPR spectrum of 15 as a toluene glass collected at 4 K; experimental spectrum, — simulated spectrum.

One-Electron Oxidation with MeI.

When a pentane solution of 11 is treated with excess MeI, the solution remains green. However, when the reaction is left to stir overnight, a blue precipitate and brown solution form. The blue precipitate can be crystallized from a THF/pentane solution at room temperature resulting in a 30% isolated yield of 16 (Figure 11, top). 16 could potentially exist as one of two limiting configurations: Ti^IV/Ti^III (Figure 11, middle), or Ti^III/Ti^IV (Figure 11, bottom). XRD and EPR analysis of 16 (see below) suggests that the major electronic structure contributor is Ti^IV/Ti^III where the two Ti centers (left to right) are in ML₄X₄ Ti^IV d⁰ and ML₅X₃ Ti^III d¹ configurations (Figure 11, middle).

Figure 11.

Top: Reaction of 11 with MeI results in formation of 16. Middle: Electron counting diagram for 16 assuming a Ti^IV/Ti^III configuration. Bond lengths determined by XRD (see Figure 12). Bottom: Electron counting diagram for 16 assuming a assuming a Ti^III/Ti^IV configuration.

Select metrical parameters of 16 are summarized in Table 1. Like in 14, the Ti2 center is no longer κ²(O,O) chelated, and an iodide ligand is coordinated to Ti2. The Ti–I bond length in 16 is 2.9103(8) Å and is significantly longer than any reported terminal Ti^III–I bond in the literature, which range from 2.6898(4) to 2.7893(6) Å.[32–36] Only one terminal Ti^II–I was reported by Ellis[32] with a distance of 3.0541(6) Å. As discussed with 14, the Ti–Cp(centroid) distances in 16 are also very similar to Ti^III–Cp(centroid) and Ti^IV–Cp(centroid) distances in the literature and cannot be reliably used to determine the oxidation state of the metal centers.[25–28]

EPR was used to help clarify the oxidation states of the two Ti centers in 16. It was found that 16 was a mixed-valent Ti^IV/Ti^III compound with an isotropic EPR spectrum similar to compound 15, which suggests significant valence localization (Figure 13). Consistent with localization of the unpaired spin on the Ti center bearing the iodide ligand, the EPR spectrum can be simulated with g = 2.001, A_Ti = 22.0 MHz, A_H = 21.7 MHz, and A_I = 35.0 MHz. Thus, 16 is the product of single-electron remote oxidation, wherein the radical reaction of Ti2 with I• results in oxidation of Ti1. Unfortunately, reaction with other alkyl halides capable of generating radicals did not new yield organometallic products. For example, when 11 is treated with BnCl an orange precipitate is deposited. NMR of the reaction mixture shows the presence of bibenzyl, Cp₂TiCl₂, and unreacted BnCl (Supporting Information).

Figure 13.

Experimental X-band EPR spectrum of 16 as a toluene glass collected at 4 K. experimental spectrum, — simulated spectrum.

Conclusions

In summary, three new reactions of the bimetallic Ti^III/Ti^III complex 11 are reported, exhibiting differential reactivity of the reduced dititanium core. In the case of simple ligand exchange with an L donor such as CNXyl, the bimetallic structure is retained, wherein XRD, EPR, and IR analysis indicate that the Ti^III/Ti^III electronic structure remains predominant. The outcome of reactions with various electrophiles are found to depend on the identity of the electrophiles. Reaction with TMSCl 11 results in a well-characterized monometallic Ti^III complex with a chelating α,β-unsaturated silyl ester moiety, which is produced via Ti–O/Si–Cl exchange. In contrast, reaction with MeI results in a radical reaction, in which Ti1 undergoes oxidation to form a mixed valent Ti^III/Ti^IV complex, consistent with earlier proposals that cooperative reactivity through electronic communication in the bimetallic core is possible. Studies into the fundamental oxidation chemistry of these complexes will hopefully motivate future work on catalytic reactions related to organometallic CO₂ coupling using Ti.

Experimental

General Considerations.

All syntheses and manipulations described below were conducted under nitrogen with exclusion of air using glovebox, Schlenk-line, and high-vacuum techniques. Compound 11 was prepared using previously published procedures.[3] CNXyl was sublimed prior to use. BnCl, TMSCl and MeI were vacuum transferred and passed through activated alumina in the glovebox prior to use. NMR solvent C₆D₆ was dried over Na⁰/Ph₂CO and vacuum transferred before passing through activated alumina in the glovebox. Pentane, hexanes, and toluene were dried on a Pure Process Technology solvent purification system prior to use. ¹H NMR spectra were obtained on a Bruker Avance 400 MHz spectrometer at 298 K. IR data was collected on a Nicolet iS 5 FTIR Spectrometer as a toluene solution. X-band EPR spectra were recorded on a Bruker ELEXSYS Spectrometer with a cryogen-free in-cavity temperature control system at 4K. Elemental analysis was shipped to and performed by Midwest Microlab. In a glovebox, a 4 mm thin wall EPR tube was charged with a toluene (0.15 mL) solution of the sample which was then capped with a rubber septum and measured on the EPR spectrometer. SC-XRD data of 14 was collected on a Bruker-AXS Venture Photon-III using a Mo source, 15 was collected on a Rigaku SuperNova using a Cu source, and 16 was collected on a Bruker-AXS Smart Apex-II using a Mo source. Selected crystal and refinement data are available in Table 3.

Table 3.

Selected crystal data parameters and refinement for 14-16.

	14	15	16
CCDC Number	2085892	2085890	2085891
Empirical Formula	C38H47NO2Si2Ti2	C22H37O2Si3Ti	C29H38O2Si2Ti2I
Formula Weight	701.70	465.66	697.43
Temperature (K)	125(2)	100(2)	123(2)
Source	Mo	Cu	Mo
a, Å	10.0752(7)	14.63740(12)	16.031(5)
b, Å	12.2919(10)	10.01815(8)	10.047(3)
c, Å	15.4949(12)	8.66141(17)	19.447(6)
α, °	78.737(3)	90	90
β, °	79.024(3)	111.5691(10)	105.551(4)
γ, °	78.713(3)	90	90
Volume, Å3	1822.3(2)	2544.88(4)	3017.5(16)
Z	2	4	4
Crystal System	Triclinic	Monoclinic	Monoclinic
Space Group	P-1	P21/n	P21/n
dcalc, g/cm3	1.279	1.215	1.535
θ Range, °	2.303 to 30.576	3.2940 to 71.6060	1.921 to 27.606
μ, mm−1	0.537	4.310	1.656
Abs. Correction	Multi-scan	Gaussian	Multi-scan
GOF	1.037	1.048	1.051
R1 a	R1 = 0.0354	R1 = 0.0246	R1 = 0.0231
wR2b [I>2σ(I)]	wR2 = 0.0931	wR2 = 0.0642	wR2 = 0.0536

^aR₁ = Σ‖F_o|-|F_c‖/Σ|F_o|.

^bwR₂ = [Σ[w(F_o²-F_c²)²]/Σ[w(F_o²)²]^1/2.

Synthesis of 14.

In a N₂-filled glovebox, a colorless toluene (5 mL) solution of CNXyl (70 mg, 0.53 mmol) was added dropwise to a green toluene (10 mL) solution of 11 (300 mg, 0.53 mmol). The solution became immediate dark red and was left to stir overnight. Toluene was removed in vacuo from the dark red mixture, and a dark red powder was isolated yielding 14 (184 mg, 49%). Single crystal X-ray quality crystals were obtained from a concentrated pentane solution of 3 at −35 °C. IR: 2781sh, 2756w, 2693w, 2676w, 2665w, 2641sh, 2619w, 2596w, 2562m, 2556sh, 2527w, 2516w, 2502w, 2473w, 2473w, 2453sh, 2445w, 2397w, 2375m, 2346w, 2323w, 2298m, 2271m, 2242m, 2225sh, 2218m, 2197m, 2174m, 2131m, 2073sh, 2038s(ν_C≡N), 2003sh, 1977w, 1910s, 1838m, 1783w, 1755w, 1715w, 1702w, 1685w(ν_C=O), 1668w, 1647w, 1342w, 1322w, 1289m(ν_C–O), 1259m, 1231m, 1199w. 1170w, 1163w, 1143sh, 1127m, 1118w. Elemental analysis for 14 gave values consistent with half an equiv. of toluene added to the complex likely due to carry over from work-up Anal. Calcd. for C₃₈H₄₇NO₂Si₂Ti₂·0.5C₇H₈: C, 66.66; H, 6.87; N, 1.87. Found: C, 66.56; H, 6.77; N, 2.36.

Synthesis of 15.

In a N₂-filled glovebox, TMSCl (70 μL, 1.10 mmol) was added dropwise to a green pentane (10 mL) solution of 11 (300 mg, 0.53 mmol). The solution remained green and was left to stir overnight. The reaction mixture was filtered to remove orange precipitate formed. Pentane was removed in vacuo from the green filtrate and a green powder was isolated yielding 15 (218 mg, 88%). Single crystal X-ray quality crystals were obtained from a concentrated pentane solution of 15 at −35 °C. IR: 2782m, 2776w, 2762m, 2697m, 2678m, 2664m, 2620m, 2596w, 2562m, 2524m, 2514w, 2500w, 2841w, 2472w, 2443m, 2397m, 2377m, 2346m, 2322m, 2298m, 2272m, 2242m, 2218m, 2197m, 2206m, 2133sh, 2129m, 2046sh, 2028m, 2079m, 2053m, 2045m, 2023sh, 2003w, 1977w, 1908m, 1894sh, 1836w, 1781w, 1767w, 1753m, 1714m, 1700w, 1684m(ν_C=O), 1669m, 1651w, 1646w, 1556w, 1340w, 1318w, 1286s(ν_C–O), 1256s, 1242s, 1200w, 1125w. Elemental analysis for 15 gave unsatisfactory values likely due to air oxidation Anal. Calcd. for C₂₂H₃₇O₂Si₃Ti: C, 56.75; H, 8.01. Found: C, 54.78; H, 7.67.

Synthesis of 16.

In a N₂-filled glovebox, excess MeI (1 mL) was added dropwise to a green pentane (10 mL) solution of 11 (300 mg, 0.53 mmol). The solution remained green and was left to stir overnight. The resulting blue reaction mixture was filtered to obtain a blue precipitate. The blue solids were dried in vacuo isolating yielding 16 (112 mg, 30%). Single crystal X-ray quality crystals were obtained by dissolving the 50 mg blue powder in THF (0.5 mL) and layered into pentane (5 mL) at room temperature. IR: 2790m, 2711w, 2702sh, 2667w, 2656w, 2643w, 2632w, 2617w, 2564w, 2554w, 2522w, 2510w, 2486w, 2471w, 2433w, 2398w, 2393w, 2380w, 2349w, 2337w, 2322w, 2283m, 2274m, 2241w, 2228w, 2219w, 2204m, 2197m, 2174w, 2142w, 2130m, 2109sh, 2087s, 2047w, 2021w, 2002sh, 1996w, 1976m, 1903m, 1889m, 1860w, 1856w, 1834w, 1830w, 1816w, 1752s, 1715m, 1699w, 1684m, 1649s(ν_C=O), 1554sh, 1546w, 1344w, 1318w, 1302sh, 1291s(ν_C–O), 1263s, 1243s, 1201w, 1127w. Elemental analysis for 16 gave unsatisfactory values likely due to air oxidation Anal. Calcd. for C₂₉H₃₈O₂Si₂Ti₂I: C, 49.94; H, 5.49. Found: C, 47.11; H, 5.53.

Reaction of 11 and BnCl.

In a N₂-filled glovebox, BnCl (22 mg, 0.18 mmol) was added dropwise to a green pentane (10 mL) solution of 11 (100 mg, 0.18 mmol). The reaction mixture became immediately orange. Pentane was removed in vacuo yielding an orange solid. NMR and the unit cell obtained from a single crystal[27, 37] confirmed the formation of Cp₂TiCl₂ and bibenzyl[38] was identified by NMR.

Figure 12.

Displacement ellipsoid plot of 16 with ellipsoids drawn at the 50% probability level. Hydrogen atoms and a disordered Me group were omitted for clarity.