Syntheses of Molybdenum Adamantylimido and t-Butylimido Alkylidene Chloride Complexes Using HCI and Diphenylmethylphosphine

Abstract

Reactions between Mo(N-t-Bu)₂(CH₂-t-Bu)₂ or Mo(NAdamantyl)₂(CH₂CMe₂Ph)₂ and 3 equiv of HCl in the presence of 1 equiv of PPh₂Me yield Mo(NR)(CHR’)(PPh₂Me)Cl₂ complexes, from which Mo(NR)(CHR’)(PPh₂Me)(OAr)Cl complexes (OAr = a 2,6-terphenoxide) can be prepared. The Mo(NR)(CHR’)(PPh₂Me)(OAr)Cl complexes were evaluated as cross-metathesis catalysts between cyclooctene and Z-1,2-dichloroethylene. The efficiencies of the test reaction for complexes in which OAr = OTPP, OHMT, OHIPT, or OHTBT (where OTPP is 2,3,5,6-tetraphenylphenoxide, OHMT is hexamethylterphenoxide, OHIPT is hexaisopropylterphenoxide, and OHTBT is hexa-t-butylterphenoxide) maximize when OAr is OHMT or OHIPT. Mo(N-t-Bu)(CH-t-Bu)(PPh₂Me)Cl₂ is essentially inactive for the reaction between cyclooctene and Z-1,2-dichloroethylene. X-ray structural studies were carried out on Mo(NAd)(CHCMe₂Ph)(PPh₂Me)Cl₂, Mo(N-t-Bu)(CH-t-Bu)(PPh₂Me)(OHMT)Cl, Mo(NAd)(CHCMe₂Ph)(Cl)(OHTBT)(PMe₃), and [Mo(NAd)(CHCMe₂Ph)(PMe₃)(Cl)]₂(μ-O), the product of the reaction between Mo(NAd)(CHCMe₂Ph)(Cl)(OHTBT)(PMe₃) and 0.5 equiv of water.

Graphical Abstract

INTRODUCTION

A key step in the synthesis of molybdenum and tungsten imido alkylidene complexes traditionally has been the addition of 3 equiv of triflic acid to a M(NR)₂(CH₂R’)₂ complex (R = aryl, t-butyl, or 1-adamantyl; R’ = t-Bu or CMe₂Ph) in the presence of dimethoxyethane to give RNH₃OTf, CH₃R’, and a M(NR)(CHR’)(OTf)₂(DME) complex.¹ The triflate ligands can be replaced with a variety of sterically demanding alkoxide (or aryloxide; X) or pyrrolide (Y) ligands to give 14e M(NR)(CHR’)(X)₂, M(NR)(CHR’)(Y)₂, or M(NR)(CHR’)(X)(Y) complexes. Monoaryloxide pyrrolide M(NR)(CHR’)(X)(Y) complexes have proven useful for E-selective or Z-selective reactions,² including stereoselective ROMP reactions,³ when the aryloxide is a sterically demanding 2,6-terphenoxide such as 2,6-dimesitylphenoxide (OHMT = hexamethylterphenoxide). In the process of exploring cross-metathesis reactions in which one of the olefins is Z-ClCH=CHCl,⁴ we discovered that M(NR)(CHR’)(OAr)(L)Cl complexes are especially reactive if the donor ligand (L = a 2e donor such as a pyridine or a nitrile) is labile enough to expose the 14e M(NR)(CHR’)(OAr)Cl core. Although we devised methods of preparing some M(NR)(CHR’)(OAr)(L)Cl complexes,⁵ the monochloride complexes would be more directly accessible from M(NR)(CHR’)(L)Cl₂ or M(NR)(CHR’)(L)₂Cl₂ complexes simply through substitution of one chloride with a sterically demanding aryloxide. They should be useful catalysts if L readily dissociates from M(NR)(CHR’)(OAr)(L)Cl to a significant degree.

M(NR)(CHR’)(L)Cl₂ or M(NR)(CHR’)(L)₂Cl₂ complexes (M = Mo or W) are relatively rare. The first M(NR)(CHR’)(L)₂Cl₂ complexes were prepared through addition of 3 equiv of HCl (as pyridinium chloride) to M(NR)₂(CH₂R’)₂, where NR is N-2,6-(mesityl)₂C₆H₃ (NAr*).⁶ Molybdenum and tungsten complexes that contain a N-2,6-(2,4,6-i-Pr₃C₆H₂)₂C₆H₃ (NAr**) ligand were prepared in a similar manner.⁷ The steric demand of the NAr** ligand even allowed a W(NAr**)(CHCMe₂Ph)Cl₂ complex to be prepared, the only isolable monomeric 14e dichloride complex of this general type. W(NR)(CHR’)Cl₂(pyridine)₂ complexes (R = adamantylimido or t-butylimido) were reported soon thereafter,⁸ as were similar dichlorides in reactions between W(N-t-Bu)₂(CH₂-2-OMeC₆H₄)₂ and pyridine hydrochloride.⁹ Unfortunately, in most cases pyridine is bound too strongly to the metal in the resulting M(NR)(CHR’)(OAr)(L)Cl complexes for them to be highly active in olefin metathesis reactions. In some cases the pyridine is labile enough to be removed from solution through addition of B(C₆F₅)₃ to yield (py)B(C₆F₅)₃, in which case the 14e complexes can be generated in situ.

Phosphine adducts of imido alkylidene complexes are relatively well-known.¹ Therefore, we explored the possibility of using phosphonium chlorides as the means of delivering HCl and chloride to generate imido alkylidene dichloride complexes as phosphine adducts. We discovered that addition of 2 equiv of PhMe₂PHCl to Mo(NC₆F₅)₂(CH₂CMe₂Ph)₂ led to formation of Mo(NC₆F₅)(CHCMe₂Ph)(PhMe₂P)₂Cl₂ complexes in good yield.⁵ In that paper, we reported that “treatment of Mo(N-t-Bu)₂(CH₂-t-Bu)₂ with phosphonium halides generated a complex mixture of alkylidene-containing compounds along with other unidentified products.” Upon optimization of conditions and choice of phosphine, we have now found that reactions between Mo(N-t-Bu)₂(CH₂-t-Bu)₂ or Mo(NAdamantyl)₂(CH₂CMe₂Ph)₂ and 3 equiv of HCl in the presence of 1 equiv of PPh₂Me produce Mo(NR)(CHR’)(PPh₂Me)Cl₂ complexes, from which Mo(NR)(CHR’)(PPh₂Me)(OAr)Cl complexes can be prepared. In this paper we report the syntheses of Mo(NR)(CHR’)(PPh₂Me)Cl₂ and Mo(NR)(CHR’)(PPh₂Me)(OAr)Cl complexes, along with a side-by-side comparison of all catalysts for a ring-opening cross-metathesis between cyclooctene and Z-ClCH=CHCl.

RESULTS

The molybdenum imido alkylidene dichloride complexes, Mo(NAd)(CHCMe₂Ph)(PPh₂Me)Cl₂ (1a(PPh₂Me)) and Mo(N-t-Bu)(CH-t-Bu)(PPh₂Me)Cl₂ (1b(PPh₂Me)) are obtained in approximately 50% yield from the corresponding Mo(NR)₂(CH₂R’)₂ complexes upon addition of 1 equiv of PPh₂Me, followed by 3 equiv of HCl in diethyl ether, or alternatively, 1 equiv of PPh₂Me•HCl, followed by 2 equiv of HCl in ether (eq 1). Three equivalents of HCl are required to complete the reaction because the t-butylamine or adamantylamine that is formed is readily protonated to give the ammonium salt, which is not a strong enough acid to allow the reaction to be completed. The reactions typically are carried out at −78 °C initially, followed by warming the reaction mixtures slowly to 22 °C, although reactions that start at ~−30 °C have been equally successful. Exploratory reactions in the presence of 1 equiv of PPh₃ or PPhMe₂ so far have not led to acceptable yields of complexes analogous to 1a(PPh₂Me) and 1b(PPh₂Me).

All NMR data are consistent with 1a(PPh₂Me) and 1b(PPh₂Me) being monophosphine complexes that do not contain a plane of symmetry. Two alkylidene doublet resonances are found in the proton NMR spectra of each. The major alkylidene α proton resonances found near 13 ppm (in CD₂Cl₂) have ¹J_CH values (123 Hz for 1a(PPh₂Me) and 118 Hz for 1b(PPh₂Me)) that are consistent with the alkylidene being in the syn orientation. In the minor isomer (5–10 % of the mixture) the value for ¹J_CH (148 Hz for 1a(PPh₂Me)) is characteristic of an anti isomer. The doublet resonance for the syn alkylidene α carbon in 1a(PPh₂Me) is found at 309.8 ppm (²J_CP = 18.9 Hz; CD₂Cl₂) and in 1b(PPh₂Me) is found at 314.5 ppm (²J_CP = 18.7 Hz; CD₂Cl₂). Recrystallization of mixtures of syn and anti isomers leads to samples that are essentially pure syn isomers, which suggests that syn and anti isomers do not interconvert readily in solution, most likely because alkylidenes can interconvert only as 14e phosphine-free species^10,11 and PPh₂Me is bound too strongly in 1a(PPh₂Me) and 1b(PPh₂Me).

An X-ray structural study of 1a(PPh₂Me) (Figure 1) shows the overall structure to be approximately halfway between a trigonal bipyramid and a square pyramid (τ¹¹ = 0.54) with the alkylidene in the syn orientation. Relevant bond angles at the metal are N1–Mo–Cl1 = 132.38(7)°, Cl1–Mo–C1 = 120.92(8)°, C1–Mo–N1 = 106.24(10)°, Cl2–Mo–P1 = 164.85(2)°, C1–Mo–P1 = 89.15(8)°, and C1–Mo–Cl2 = 96.40(7)°. The Mo=C1 (1.887 (2) Å) and Mo=N1 (1.717 (2) Å) distances and Mo=C–C (147.39(19)°) and Mo=N–C (163.15(17)°) angles are normal for five-coordinate complexes of this general type.

Figure 1.

Structure of Mo(NAd)(CHCMe₂Ph)(PPh₂Me)Cl₂ (1a(PPh₂Me)). Hydrogen atoms, except on C1, have been omitted for clarity.

Monoaryloxide monochloride complexes were prepared in reactions between 1a(PPh₂Me) or 1b(PPh₂Me) and 1 equiv of LiOAr (OAr = OTPP, OHMT, OHIPT, OHTBT) in benzene, where OTPP is 2,3,5,6-tetraphenylphenoxide, OHMT is hexamethylterphenoxide (2,6-dimesitylphenoxide), OHIPT is hexaisopropylterphenoxide, and OHTBT is hexa-t-butylterphenoxide.¹² We chose OTPP, OHMT, OHIPT, and OHTBT because of the dramatic degree to which their steric properties vary in the order OTPP<OHMT<OHIPT<OHTBT. Compound 5a(PMe₃) is the first alkylidene complex to be prepared that contains the OHTBT ligand.¹²

Proton NMR spectra of complexes 2a(PPh₂Me) and 2b(PPh₂Me) reveal largely a doublet resonance for a syn alkylidene isomer, and under some conditions a small doublet for an anti alkylidene isomer, as found for 1a(PPh₂Me) and 1b(PPh₂Me). However, a few percent of each sample (~5%, depending upon the concentration) are the phosphine-free 14e complexes, 2a and 2b, respectively, as syn isomers, according to ¹H NMR spectra. Free phosphine also can be observed in the ³¹P NMR spectra (see the SI). The amount of free phosphine increases for 3a(PPh₂Me) and 3b(PPh₂Me), and again for 4b(PPh₂Me). As expected, addition of PPh₂Me shifts the equilibrium toward the PPh₂Me adduct (as a syn isomer), while addition of 1 equiv of B(C₆F₅)₃ results in scavenging of the PPh₂Me to produce spectra consistent with formation of the phosphine-free species. It is clear that the phosphine-free complexes are extremely soluble, even in pentane, and none has been isolated in crystalline form. Mo(NR)(CHR’)(OAr)Cl complexes also appear to be exceedingly sensitive to water (vide infra).

An X-ray crystallographic study of 3b(PPh₂Me) reveals that the crystals contain two molecules that are enantiomers,¹³ with equal occupancy over the entire unit cell (see the SI for details). The disorder could be resolved and both structures of 3b(PPh₂Me) solved. They are best described as approximately square pyramids (τ = 0.24 and 0.36) with the apical alkylidene in a syn conformation and the PPh₂Me bound trans to the chloride. Solvents are commonly found in the crystal lattices of the compounds reported here (sometimes partial occupation; see the SI), a complication that has compromised the precision of elemental analyses in some cases.

The structure of the isomer of 3b(PPh₂Me) with τ = 0.24 is shown in Figure 2. Relevant bond angles around the metal are Cl1–Mo1–P1 = 167.39(4)°, N1–Mo1–O1 = 153.12(9)°, C1–Mo1–N1 = 101.32(11)°, C1–Mo1–O1 = 105.51(9)°, C1–Mo1–Cl1 = 100.33(8)° and C1–Mo1–P1 = 89.01(8)°. The Mo=C1 (1.921(2) Å), Mo=N1 (1.722(2) Å), Mo–Cl1 (2.4457(7) Å), Mo–P1 (2.5587(13) Å), and Mo–O1 (1.9895(14) Å) distances and Mo=C1–C2 (142.4(2)°), Mo=N1–C6 (160.15 (19)°), and Mo–O1–C10 (146.90(12)°) angles are typical of other five-coordinate complexes of this general type.⁵

Figure 2.

Structure of one of the molecules of Mo(N-t-Bu)(CH-t-Bu)(PPh₂Me)(OHMT)Cl (3b(PPh₂Me)); τ = 0.24. Hydrogen atoms (except on C1), disordered atoms, and lattice solvent toluene have been omitted for clarity.

The reaction of 1a(PPh₂Me) with KOHTBT (OHTBT = hexa-t-butylterphenoxide) gives essentially phosphine-free 4-coordinate 5a, according to NMR studies; it is too soluble, even in pentane, to be crystallized (so far). However, a PMe₃ complex, 5a(PMe₃), could be isolated in crystalline form in 47% yield upon addition of 5 equiv of PMe₃ to 5a in dichloromethane. Compound 5a(PMe₃) is an off-white powder in the solid state, but its solutions are yellow because PMe₃ dissociates to a significant degree to give yellow 5a, as can be shown upon scavenging PMe₃ in solutions of 5a(PMe₃) with B(C₆F₅)₃. The proton NMR spectrum of 5a(PMe₃) in Figure 3 (top) shows that the ratio of 5a(PMe₃) to 5a is 73:27 (20 mg of 5a(PMe₃) in 0.6 mL of C₆D₆, 31 mM initial concentration); therefore K_eq is 3.1 mM in C₆D₆ at 20 °C. The rate of PMe₃ exchange in 5a(PMe₃) is such that the alkylidene proton is essentially decoupled from phosphorus. Addition of 2 equiv of PMe₃ to the solution of 5a(PMe₃) results in a mixture that is ~97% 5a(PMe₃) (Figure 3 bottom). The small doublet alkylidene resonance at 13.55 ppm in the spectra in Figure 3 is the result of a reaction between 5a(PMe₃) and traces of water (vide infra).

Figure 3.

Alkylidene resonances in proton NMR spectra of Mo(NAd)(CHCMe₂Ph)(Cl)(OHTBT)(PMe₃) 5a(PMe₃) in C₆D₆ at 20 °C; *5a(PMe₃); **5a.

An X-ray study of 5a(PMe₃) (Figure 4) shows its structure to be closer to a square pyramid (τ = 0.27) with the alkylidene in the syn orientation, analogous to that of 3b(PPh₂Me) (Figure 2). Relevant bond angles at the metal are Cl1–Mo–P1 = 168.122 (13)°, N1–Mo–O1 = 151.86(5)°, C1–Mo–O1 = 106.44(5)°, C1–Mo–N1 = 101.36(6)°, C1–Mo1–Cl1 = 97.67(4)°, and C1–Mo1–P1 = 94.21(4)°. The Mo=C1 (1.8973(14) Å), Mo=N1 (1.7345(12) Å), Mo–Cl1 (2.4480(5) Å), and Mo–P (2.5168(5) Å) distances are similar to what they are in 3b(PPh₂Me). The Mo-O1 (2.0243(10) Å) distance is slightly longer than what it is in 3b(PPh₂Me), presumably as a consequence of the extreme steric bulk of the OHTBT ligand. The steric bulk of the OHTBT ligand must also be the origin of the relatively large Mo–O1–C21 (159.86(9)°) angle. For comparison, in Mo(N-t-Bu)(CH-t-Bu)Cl(OHIPT)(3-Bromopyridine) the Mo–O–C_aryl angle is 151.4(3)°.⁴ The Mo=C-C (142.48(11)°) and Mo=N–C (166.59(11)°) angles within 5a(PMe₃) are consistent with complexes of this type.

Figure 4.

Structure of Mo(NAd)(CHCMe₂Ph)(Cl)(OHTBT)(PMe₃) 5a(PMe₃); τ = 0.27. Hydrogen atoms (except on C1) and lattice solvent dichloromethane have been omitted for clarity.

In an attempt to remove PMe₃ from 5a(PMe₃) in the solid state to give 5a, a sample of 50 mg of 5a(PMe₃) in a Pyrex tube was exposed to a vacuum (25 mTorr) for 24 h. We were surprised to find that 5a(PMe₃) decomposed under these conditions to give HTBTOH and a single alkylidene complex that is relatively insoluble in pentane and that contains PMe₃, but not OHTBT. HTBTOH can be separated from the decomposition product (6a(PMe₃)) through addition of a small quantity of pentane to the mixture of 6a(PMe₃) and HTBTOH. A ¹H NMR spectrum of 6a(PMe₃) shows that the syn-alkylidene proton resonance is a doublet at 13.55 ppm (³J_PH = 6.4 Hz; ¹J_CH = 127 Hz). Because 6a(PMe₃) also could be obtained in 67% yield in a reaction between 5a(PMe₃) and 0.5 equiv of water in ether, we suspect that 6a(PMe₃) is formed from a selective hydrolysis of the OHTBT ligand in 5a(PMe₃), even in vacuo in the particular experiment described above.

An X-ray study of 6a(PMe₃) (Figure 5) showed it to be [Mo(NAd)(CHCMe₂Ph)(PMe₃)(Cl)]₂(μ-O). The overall structure of each Mo center is close to a square pyramid (τ = 0.21 at Mo1 and τ = 0.16 at Mo2) with each alkylidene in the syn orientation. Relevant bond angles at the metal are Mo1–O–Mo2 = 120.37(7)°, N1–Mo1–O1 = 149.41(7)°, Cl1–Mo1–P1 = 161.730(19)°, N2–Mo2–O1 = 152.02 (7)°, Cl2–Mo2–P2 = 161.34(2)°. The Mo1–O1 (1.9423(13) Å), Mo2–O1 (1.9774(13) Å), Mo1=C101 (1.907(2) Å), Mo2=C201 (1.907(2) Å), Mo1=N1 (1.7507(17) Å), Mo2=N2 (1.7492(17) Å), Mo1–Cl1 (2.4871(6) Å), Mo2–Cl2 (2.4870(6) Å), Mo1–P1 (2.4980(6) Å), Mo2–P2 (2.4894(7) Å) distances and Mo1=C–C (141.95(6)°), Mo2=C–C (143.61(16)°), Mo1=N–C (158.81(14)°), and Mo2=N–C (160.82(15)°). In view of the steric crowding within 6a(PMe₃), we were surprised to find that the Mo1–O1–Mo2 angle is only 120°.

Figure 5.

Structure of [Mo(NAd)(CHCMe₂Ph)(PMe₃)(Cl)]₂(μ-O) (6a(PMe₃)).

We evaluated the efficacy of 1b, 2b, 3b, 4b, 5a, and 6a, and their phosphine adducts), for the ring-opening cross-metathesis of cyclooctene and Z-dichloroethylene (equation 3), as has been described in an earlier paper for closely related monomeric complexes.⁵ The results are shown in Table 1.

Table 1.

Results (A/B/C Percentages) of Reactions between Cyclooctene (A) and Z-1,2-Dichloroethylene (1.25 equiv) in C₆D₆ to Give Polycyclooctene (B) and/or the Cross-metathesis Product (C)

initiator	20 min	1 h	6 h	% Z,Z of C
1b(PPh2Me)	100/0/0	100/0/0	100/0/0	n/a
2b(PPh2Me)	100/0/0	92/8/0	45/55/0
3b(PPh2Me)	3/51/46	0/9/91	0/1/99	86.4
4b(PPh2Me)	4/32/64	0/24/76	0/4/96	100
5a(PMe3)	100/0/0	100/0/0	100/0/0	n/a
6a(PMe3)	100/0/0	100/0/0	100/0/0	n/a
1ba	100/0/0	100/0/0	100/0/0	n/a
2ba	0/2/98	0/1/99		89.6
3ba	0/3/97	0/3/97		99.3
4ba	0/0/100			99.7
5aa	100/0/0	100/0/0	100/0/0	n/a
6aa	78/22/0	61/39/0	5/95/0	n/a

^aGenerated in situ through addition of 1 equiv of B(C₆F₅)₃ to the phosphine adduct.

We found that product C is formed readily when 3b and 4b, in which the PPh₂Me is dissociated to the greatest extent, are the initiators. Compound C is essentially 100% Z,Z when 4b(PPh₂Me) is the initiator and 86.4% Z,Z when 3b(PPh₂Me) is the initiator, as a consequence of the greater steric demand of the OHIPT versus the OHMT terphenoxide ligand.⁵ Some poly(cyclooctene) is formed when 2b(PPh₂Me) is the initiator, but the cross-metathesis reaction is comparatively sluggish unless B(C₆F₅)₃ is added, in which case the activity is high, but the stereoselectivity is only ~90%. Compound 1b(PPh₂Me) is inactive for the cross-metathesis reaction, most likely because PPh₂Me is too tightly bound to the metal. 1b(PPh₂Me), cyclooctene, and ClCH=CHCl are still present after 10 h and no initial metathesis product (t-BuCH=CHCl) is formed. We were surprised to find that 5a(PMe₃) is inactive, even though PMe₃ is dissociated to a significant degree from Mo in 5a(PMe₃) (Figure 3, top) and even when B(C₆F₅)₃ has been added to 5a(PMe₃) to give 5a. No C is formed using 6a(PMe₃) or 6a.

DISCUSSION

It has long been known through experiments,^1,14 and has now been confirmed through extensive calculations,¹⁵ that the 14e core of a high oxidation state Mo or W complex must be accessible in order to be an efficient olefin metathesis catalyst, and the metallacyclobutane intermediate must be a trigonal bipyramid with the MC₃ ring in two equatorial positions.¹⁵ As we have shown here and elsewhere,^4,5 monoaryloxide monochloride complexes are unusually reactive and capable of cross-metathesizing electron-poor Z-ClCH=CHCl with an “ordinary” olefin such as cyclooctene. Fortunately, the adamantylimido and t-butylimido alkylidene complexes now can be prepared relatively directly as diphenylmethylphosphine adducts using HCl to trigger formation of the alkylidene through protonation and removal of one imido ligand and to provide the chloride anion, all in one event. Phosphine is required in order to isolate the dichloride or monochloride complexes, but in the right circumstances the phosphine is labile enough for the compound to initiate and sustain a metathesis reaction that involves Z-ClCH=CHCl. The degree of dissociation of PPh₂Me depends dramatically upon the size of the aryloxide (OAr) ligand in the order OTPP < OHMT < OHIPT < OHTBT, but the reactivity of the resulting 14e complex is likely to follow the opposite trend (OTPP > OHMT > OHIPT > OHTBT). These opposing trends result in 3b(PPh₂Me) and 4b(PPh₂Me) being the “best compromise” for the reaction shown in equation 3. Sterically less demanding and more strongly bound phosphines (PMe₃) or other ligands (e.g., pyridine) are more likely to hinder or block reaction with an olefin through formation of a 16e or 18e complex. More reactive, smaller olefins (e.g., ethylene) may be able to compete successfully for a 14e complex, whereas Z-ClCH=CHCl, or even cyclooctene, cannot.

We have to ascribe the low activity of phosphine-free 5a entirely to the fact that OHTBT is simply too sterically demanding for the reaction shown in equation 3, or even polymerization of cyclooctene, under the conditions employed. Low reactivity has been observed, largely for steric reasons, for other 14e imido alkylidene complexes. One example is Mo(NAd)(CHCMe₂Ph)[OSi(t-Bu)₃]₂, which does not react readily with ethylene at 5 atm and 120 °C.¹⁶ The low reactivity of phosphine-free 6a also can be ascribed to the steric demand of one of the Mo entities with respect to reaction of substrate at the other Mo center.

Detailed studies that document the consequence of reaction of water with high oxidation state Mo or W alkylidene complexes¹⁷ (or alkylidyne complexes¹⁸) are rare. What we have found here in the case of 6a(PMe₃) is that the alkylidene ligand is not necessarily the first ligand to be protonated upon addition of water. The protonation process is likely to consist of water first binding to the metal followed by a proton being transferred (intramolecularly or intermolecularly) to the aryloxide oxygen. It also cannot be assumed that, if an alkylidene ligand survives partial hydrolysis, the resulting complex will be a more active catalyst than the original alkylidene complex, especially if a μ-oxo dimer or larger cluster is formed, as is found for 6a(PMe₃). In fact, a lower reactivity may be more likely in view of the propensity of an oxo ligand to bridge between metals in many circumstances.

CONCLUSIONS

We have shown that adamantylimido and t-butylimido alkylidene dichloride complexes can be prepared in the presence of diphenylmethylphosphine and 3 equiv of HCl. These serve as starting materials for the synthesis of monoaryloxide chloride complexes in which the aryloxide is a terphenoxide (OTPP, OHMT, OHIPT, or OHTBT). A 16e monoaryloxide monochloride complex is active in metathesis reactions when the terphenoxide is large enough to labilize the phosphine, but small enough to allow reactions at the 14e metal center to proceed. The metathesis reaction is most stereoselective when the terphenoxide is OHIPT.

Experimental Section

General Procedures.

All air- and moisture-sensitive materials were manipulated under a nitrogen atmosphere in a Vacuum Atmospheres glovebox or on a dual-manifold Schlenk line. Glassware was either oven-dried or flame-dried prior to use. Acetonitrile, benzene, methylene chloride, diethyl ether, 1,2-dimethoxyethane, and toluene were degassed, passed through activated alumina columns, and stored over 4 Å Linde-type molecular sieves. Pentane was washed with H₂SO₄, followed by water and a saturated solution of aqueous NaHCO₃, and dried over CaCl₂ pellets for at least 2 weeks prior to use in the solvent purification system. Deuterated solvents were dried over 4 Å Linde-type molecular sieves prior to use. ¹H and ³¹P{¹H} NMR spectra were acquired on a Varian Mercury 300 MHz or a Varian Inova-500 MHz NMR spectrometer. ¹³C{¹H} NMR spectra were acquired on a Varian Inova-500 MHz NMR spectrometer. 2D NMR spectra were acquired on a Varian Inova-500 MHz NMR spectrometer. Chemical shifts for ¹H, ³¹P, and ¹³C NMR spectra are reported as parts per million relative to tetramethylsilane and referenced to the residual ¹H or ¹³C resonances of the deuterated solvent (¹H δ: benzene 7.16 ppm, methylene chloride 5.32 ppm; ¹³C δ: benzene 128.06 ppm, methylene chloride 53.84 ppm). Gas Chromatography was performed on an Agilent system equipped with an HP-5 column (ID 320 μm, 0.25 μm, and length 30 m). TMSCl was purchased from Alfa Aesar and degassed by a free-pump-thaw method prior to use. B(C₆F₅)₃ was purchased from Strem and used as received. LiOAr was prepared by the addition of 1 equiv of n-butyllithium to a cold pentane or Et₂O solution of ArOH, and the solid was collected on a glass frit, washed with pentane, and dried in vacuo. PMe₃, PPhMe₂, and PPh₂Me were purchased from Strem chemicals and used as received. The hydrochloride salts of the phosphines were prepared through addition of ethereal HCl (1.0 M; Sigma Aldrich) to the phosphine in diethyl ether while stirring the mixture; the phosphine hydrochloride salt was collected on a glass frit, washed with minimal ether, and dried in vacuo. Mo(NAd)₂Cl₂(DME),¹⁹ Mo(NAd)₂(CH₂CMe₂Ph)₂,⁹ and Mo(N-t-Bu)₂Cl₂(DME)²⁰ were prepared as described in the literature. Syntheses of Mo(N-t-Bu)₂(CH₂-t-Bu)₂ from Mo(N-t-Bu)₂Cl₂ and neopentyllithium²¹ and from Mo(N-t-Bu)₂Cl₂(DME) and neopentyl Grignard²² have been reported in the literature, but no detailed procedures were provided; we provide a procedure here. Elemental analyses were performed at the CENTC Elemental Analysis Facility at the University of Rochester, New York.

Mo(N-t-Bu)₂(CH₂-t-Bu)₂.

Mo(N-t-Bu)₂Cl₂(DME) (4.01 g, 10.03 mmol) was dissolved in ~30 mL of diethyl ether in a glovebox under nitrogen. The reaction vessel was connected to a Schlenk manifold and cooled to −78 °C with a dry ice/acetone bath. Neopentylmagnesium chloride (20.1 mL, 20.09 mmol, 2 equiv., 1.0 M in THF) was syringed into the reaction vessel over a few minutes. The reaction mixture was stirred overnight to give a turbid brown-orange solution. All volatiles were removed in vacuo. In a glovebox ~30 mL of pentane was added and the mixture was filtered through Celite. All volatiles were removed from the filtrate in vacuo to yield Mo(N-t-Bu)₂(CH₂-t-Bu)₂ as a brown oil that was used without purification; crude yield 3.77 g (99%): ¹H NMR (500 MHz, C₆D₆, 22 °C) 1.76 (s, 4H), 1.41 (s, 18 H), 1.19 (s, 18 H) ppm.

Mo(NAd)(CHCMe₂Ph)(PPh₂Me)Cl₂ (1a(PPh₂Me)).

Mo(NAd)₂(CH₂CMe₂Ph)₂ (1.05 g, 1.59 mmol, 1.00 equiv) was dissolved in 30 mL of diethyl ether. The solution was stirred rapidly and solid PPh₂Me·HCl (0.375 g, 1.59 mmol, 1.00 equiv) was added. The reaction mixture was stirred for 30 min at room temperature and then chilled to – 78 °C. HCl (1.0 M in diethyl ether; 3.17 mL, 2.00 equiv) was added to the mixture with a syringe. The cooling bath was removed and the mixture was stirred overnight. The volatiles were removed from the reaction in vacuo, the crude residue was triturated with 1–2 mL of diethyl ether, and the off-white solid was collected on a glass frit. The crude product was dissolved in ~20 mL of a mixture of benzene and toluene (1:1 v/v) and the solution was filtered through Celite to remove the t-butylammonium chloride. The solvents were removed from the extract in vacuo to afford the product; yield 0.540 g (53 %). Only alkylidene NMR resonances are listed here: ¹H NMR (500 MHz, CD₂Cl₂) δ 13.10 (d, ³J_HP = 4.9 Hz, ¹J_CH = 123 Hz, 1H, Mo=CH_syn); ¹³C{¹H} NMR (126 MHz, CD₂Cl₂) δ 309.8 (d, ²J_CP = 22.4 Hz, Mo=C). Anal. Calcd for C₃₃H₄₀Cl₂MoNP (648.52 g/mol): C, 61.12%; H, 6.22%; N, 2.16%. Found: C, 61.21%; H, 6.19%; N, 2.51%.

Slow evaporation of benzene from a saturated solution of 1a(PPh₂Me) at room temperature gave crystals suitable for X-ray data collection.

Mo(N-t-Bu)(CH-t-Bu)Cl₂(PPh₂Me) (1b(PPh₂Me)).

The procedure was virtually the same as that for 1a(PPh₂Me) starting with Mo(N-t-Bu)₂(CH₂-t-Bu)₂ (2.00 g, 5.26 mmol, 1.00 equiv) in 30 mL of diethyl ether, PPh₂Me·HCl (1.24 g, 5.26 mmol, 1.00 equiv), and HCl (1.0 M in diethyl ether; 10.5 mL, 2.00 equiv); yield 1.31 g (49 %). Only alkylidene NMR resonances are listed here: ¹H NMR (500 MHz, CD₂Cl₂) δ 13.01 (d, ³J_HP = 4.9 Hz, ¹J_CH = 118 Hz, 1H, Mo=CH_syn); ¹³C{¹H} NMR (126 MHz, CD₂Cl₂) δ 314.5 (d, ²J_CP = 22.7 Hz, Mo=C). Anal. Calcd for C₂₂H₃₂Cl₂MoNP (508.34 g/mol): C, 51.98%; H, 6.35%; N, 2.76%. Found: C, 52.04%; H, 6.29%; N, 2.61%.

Mo(NAd)(CHCMe₂Ph)(Cl)(OTPP)(PPh₂Me) (2a(PPh₂Me)).

A solution of LiOTPP (0.218 g, 0.216 mmol, 1 equiv) in benzene solution was added dropwise to 1a(PPh₂Me) (0.140 g, 0.216 mmol, 1 equiv) in benzene. The reaction was stirred overnight and filtered through Celite. The solvents were removed in vacuo and the tacky yellow solid was triturated in 5 mL of pentane to afford the product as an off-white solid; yield 0.187 g (86%). Only alkylidene NMR resonances are listed here: ¹H NMR (500 MHz, CD₂Cl₂) δ 11.55 (d, ³J_HP = 4.4 Hz, 1H, Mo=CH); ¹³C{¹H} NMR (126 MHz, CD₂Cl₂) δ 311.9 (d, ²J_CP = 16.7 Hz, Mo=C). Anal. Calcd for C₆₃H₆₁ClMoNOP (1010.57 g/mol): C, 74.88%; H, 6.08%; N, 1.39%. So far elemental analyses that agree with the calculated figures has not been successful.

Mo(N-t-Bu)(CHCMe₂Ph)(Cl)(OTPP)(PPh₂Me) (2b(PPh₂Me)).

A solution of LiOTPP (0.114 g, 0.238 mmol, 1 equiv) in benzene was added dropwise to 1b(PPh₂Me) (0.121 g, 0.238 mmol, 1 equiv) in benzene. The isolation procedure was the same as that for 2a(PPh₂Me); yield 0.151 g (72.9%). Only alkylidene NMR resonances are listed here: ¹H NMR (500 MHz, CD₂Cl₂) δ 11.58 (d, ³J_HP = 4.5 Hz, 1H, Mo=CH); ¹³C{¹H} NMR (126 MHz, CD₂Cl₂) δ 316.5 (d, ²J_CP = 16.9 Hz, Mo=C). Anal. Calcd for C₅₂H₅₃ClMoNOP (870.39 g/mol): C, 71.76%; H, 6.14%; N, 1.61%. Found: C, 72.05%; H, 6.25%; N, 1.19%.

Mo(NAd)(CHCMe₂Ph)(Cl)(OHMT)(PPh₂Me) (3a(PPh₂Me)).

A solution of LiOHMT (0.063 g, 0.154 mmol, 1 equiv) was added dropwise to 1a(PPh₂Me) (0.100 g, 0.154 mmol, 1 equiv) in benzene. The isolation procedure was the same as that for 2a(PPh₂Me); yield 0.111 g (76.4%). Only alkylidene NMR resonances are listed here: ¹H NMR (500 MHz, CD₂Cl₂) δ 12.45 (d, ³J_HP = 4.6 Hz, 1H, Mo=CH); ¹³C{¹H} NMR (126 MHz, CD₂Cl₂) δ 313.6 (d, ²J_CP = 17.9 Hz, Mo=C). Anal. Calcd for C₅₇H₆₅ClMoNOP (942.54 g/mol): C, 72.64%; H, 6.95%; N, 1.49%. Found: C, 72.13%; H, 6.88%; N, 1.31%.

Mo(N-t-Bu)(CH-t-Bu)(Cl)(OHMT)(PPh₂Me) (3b(PPh₂Me)).

A solution of LiOHMT (0.067 g, 0.163 mmol, 1 equiv) in benzene solution was added dropwise to 1b(PPh₂Me) (0.083 g, 0.163 mmol, 1 equiv) in benzene. The isolation procedure was the same as that for 2a(PPh₂Me); yield 0.072 g (55%). Only alkylidene NMR resonances are listed here: ¹H NMR (500 MHz, CD₂Cl₂) δ 12.27 (d, ³J_HP = 4.7 Hz, 1H, Mo=CH); ¹³C{¹H} NMR (126 MHz, CD₂Cl₂) δ 318.5 (d, ²J_CP = 17.7 Hz, Mo=C). Anal. Calcd for C_49.5H₆₁ClMoNOP (848.42 g/mol): C, 70.07%; H, 7.25%; N, 1.65%. Found: C, 70.05%; H, 7.27%; N, 1.51%.

X-ray quality crystals were grown from a saturated toluene solution at –25 °C. The lattice contains 0.5 molecules of toluene per 3b(PPh₂Me).

Mo(NAd)(CHCMe₂Ph)(Cl)(OHIPT)(PPh₂Me) (4a(PMe₃)).

A solution of LiOHIPT (0.089 g, 0.177 mmol, 1 equiv) was added dropwise to 1a(PPh₂Me) (0.115 g, 0.177 mmol, 1 equiv) in benzene. The solution was stirred for 30 h and filtered through Celite. The solvents were removed from the filtrate in vacuo and the tacky residue was extracted with pentane to give an orange solution. PMe₃ (0.100 g, 1.31 mmol, 7.4 equiv) was added to the orange pentane solution. The crystals that grew when the solution was cooled to – 25 °C were collected by filtration; yield 0.040 g (23%). Only alkylidene NMR resonances are listed here: ¹H NMR (500 MHz, C₆D₆) δ 12.74 (d, ³J_HP = 4.8 Hz, 1H, Mo=CH); ¹³C{¹H} NMR (126 MHz, C₆D₆) δ 300.9 (d, ²J_CP = 18.4 Hz, Mo=C). Anal. Calcd for C₅₉H₈₅ClMoNOP (986.72 g/mol): C, 71.82%; H, 8.68%; N, 1.42%. Found: C, 71.49%; 8.55%; N, 1.20%.

Mo(N-t-Bu)(CH-t-Bu)(Cl)(OHIPT)(PPh₂Me) (4b(PPh₂Me)).

A solution of LiOHIPT (0.139 g, 0.275 mmol, 1 equiv) in benzene solution was added drop wise to a solution of Mo(N-t-Bu)(CH-t-Bu)(PPh₂Me)Cl₂, 1b(PPh₂Me) (0.140 g, 0.275 mmol, 1 equiv) in benzene. The solution was stirred rapidly overnight. The mixture was filtered through Celite, and the volatiles were removed from the filtrate in vacuo. The tacky yellow solid was stirred in 5 mL pentane to afford 4b(PPh₂Me) as an off-white solid; yield 0.168 g (62.8%). Only alkylidene NMR resonances are listed here: ¹H NMR (500 MHz, C₆D₆) δ 12.99 (br d, 1H, 5-coordinate Mo=CH), 11.62 (s, 1H, 4-coordinate Mo=CH); ¹³C NMR (obtained from ¹H-¹³C deHSQC spectrum 500 MHz, C₆D₆) δ 292.9 (4-coordinate Mo=C). Anal. Calcd for C₅₈H₈₁ClMoNOP (970.68 g/mol): C, 71.77%; H, 8.41%; N, 1.44%. Found: C, 71.39%; H, 8.27%; N, 1.22%.

Mo(NAd)(CHCMe₂Ph)(Cl)(OHTBT)(PMe₃) (5a(PMe₃)).

KOHTBT (300 mg, 0.482 mmol, 1 equiv) was added to 20 mL of benzene containing Mo(NAd)(CHCMe₂Ph)(Cl)₂(PPh₂Me) 1a(PPh₂Me) (313 mg, 0.482 mmol, 1 equiv), and the mixture was heated at 60 °C for 18 h. The benzene was evaporated in vacuo, 5 mL of dichloromethane was added, and the mixture filtered through Celite to give a red solution. PMe₃ (0.245 mL, 184 mg, 2.41 mmol, 5 equiv) was added, and the resulting solution was kept at −20 °C overnight. During this time, pale off-white crystals formed, which were filtered off and washed with 1 mL of cold dichloromethane. The product was dried in vacuo to give Mo(NAd)(CHCMe₂Ph)(OHTBT)(Cl)(PMe₃).2CH₂Cl₂ as off-white crystals; yield 280 mg (47%). The product was stored in the glovebox refrigerator at – 20 °C in order to avoid slow decomposition. Only alkylidene NMR resonances are listed here: ¹H NMR (400 MHz, C₆D₆) δ 13.47 (br s, 1H, Mo=CH); ¹³C{¹H} NMR (151 MHz, C₆D₆) δ 311.1 (d, ²J_CP = 15.1 Hz, Mo=C). Anal. Calcd for C₆₅H₉₇ClMoNOP (1070.88 g/mol): C, 72.90%; H, 9.13%; N, 1.31%. Found: C, 73.49%; H, 9.40%; N, 1.05%.

Crystals suitable for X-ray data collection were obtained by crystallizing 5a(PMe₃) from dichloromethane in the presence of 2 equiv of PMe₃ at room temperature.

[Mo(NAd)(CHCMe₂Ph)(Cl)(PMe₃)]₂(μ-O) (6a(PMe₃)).

Mo(NAd)(CHCMe₂Ph)(OHTBT)(Cl)(PMe₃).2CH₂Cl₂ (5a(PMe₃)) (100 mg, 0.08 mmol, 1 equiv) was suspended in 2 mL of ether and the mixture was cooled to – 78 °C. A stock solution of wet ether was prepared through addition of 0.005 mL of water to 1 mL ether. A 0.145 mL (0.726 mg, 0.04 mmol, 0.5 equiv of water) of this solution was added, and the resulting suspension was stirred at room temperature overnight. The solid was filtered and washed by 1 mL of ether. All solvent was removed in vacuo to give [Mo(NAd)(CHCMe₂Ph)(Cl)(PMe₃)]₂(μ-O) (6a(PMe₃) as an off-white powder; yield 27 mg (67%). Only alkylidene NMR resonances are listed here: ¹H NMR (400 MHz, C₆D₆) δ 13.57 (d, ³J_HP = 6.3 Hz, 2H, Mo=CH); ¹³C{¹H} NMR (151 MHz, C₆D₆) δ 307.7 (d, ²J_CP = 19.8 Hz, Mo=C). Anal. Calcd for C₄₆H₇₂Cl₂Mo₂N₂OP₂ (993.86 g/mol): C, 55.59%; H, 7.30%; N, 2.82%. Found: C, 55.60%; H, 7.11%; N, 2.24%.

Crystals of 6a(PMe₃) suitable for X-ray data collection were obtained through crystallization from a mixture of pentane and toluene – 20 °C.