Tetrakis(trimethylphosphine)tungsten(II) trimethylphospinate hydride

Tetrakis(trimethylphosphine)tungsten(II) trimethylphospinate hydride (W(PMe3)4(η2-CH2PMe2)H) is an air-sensitive organotungsten complex with tungsten in the oxidation state of +2. It is an electron-rich tungsten center is and, thus, prone to oxidation. This bright-yellow complex has been used as a starting retron for some challenging chemistry, such as C-C bond activation, tungsten-chalcogenide multiple bonding, tungsten-tetrel multiple bonding, and desulfurization.

Synthesis
W(PMe3)4(η2-CH2PMe2)H was first synthesized in 1983 by reacting tungsten hexachloride with trimethylphosphine and sodium under a nitrogen atmosphere. The complex was also a very minor product synthesized as a part of a reaction aimed at generating cyclopentadienyl- and PMe3-containing tungsten complexes by co-condensing tungsten atoms, PMe3, and cyclopentene at –196°C. The same procedure, sans cyclopentene, also yields W(PMe3)4(η2-CH2PMe2)H. Alternatively, W(PMe3)4(η2-CH2PMe2)H can also be synthesized by condensing PMe3 into an ampule with Na(K) alloy and adding WCl6. WCl6 with excess PMe3 with H2 as an oxidizing agent also produces W(PMe3)4(η2-CH2PMe2)H in a 3:1 mixture with W(PMe3)5H2.

Properties
Unlike Mo, the preparation of a homoleptic, octahedral trimethylphosphine tungsten complex, hexakis(trimethylphosphine)tungsten (W(PMe3)6), is extremely difficult due to the equilibrium below. W(PMe3)4(η2-CH2PMe2)H is thermodynamically favored (ΔGrxn = –1.73 kcal mol-1).

The co-condensation method produces only W(PMe3)4(η2-CH2PMe2)H, and the Na(K) alloy method produces a mixture of W(PMe3)4(η2-CH2PMe2)H and W(PMe3)6 only under vast excess of PMe3.

W(PMe3)5, being an electron-rich, 16 electron, d6 complex, is unstable. It has been proposed as an unstable intermediate between W(PMe3)4(η2-CH2PMe2)H and W(PMe3)6. Parkin and coworkers proposed a stepwise mechanism for the equilibrium where the rate-determining step is PMe3 dissociation. The concerted mechanism was disproven by the low kinetic-isotope effect data. The reaction at 30°C between W(PMe3)6 and W(PMe3)5 has ΔG‡ = 23.2 kcal mol-1, and the reaction at 30°C between W(PMe3)4(η2-CH2PMe2)H and W(PMe3)5 has ΔG‡ = 24.5 kcal mol-1.

Selective deuteration the alkylidene derivative (vide supra) leads to a statistical distribution of deuterium throughout the hydrogen sites of that phosphine as the alkyl ligand can change through the W(PMe3)4(PMe2CH2D) intermediate. The non-η2 ligands can also undergo facile fluxional rearrangement on the NMR timescale.

Diatomic molecule activation
W(PMe3)4(η2-CH2PMe2)H activates H2 to generate W(PMe3)5H2 in PMe3 solvent. W(PMe3)5H2 can activate another H2 molecule in light petroleum upon the dissociation of PMe3, giving W(PMe3)4H4. W(PMe3)4(η2-CH2PMe2)H can also be reacted with HD to generate W(PMe3)5HD or W(PMe3)4(η2-HD) in PMe3 solvent.

W(PMe3)4(η2-CH2PMe2)H binds N2 in a nitrogen atmosphere to generate W(PMe3)5(N2). In 2 atmospheres of CO, W(PMe3)4(η2-CH2PMe2)H dissociates 2 equivalents of PMe3 and adds 3 equivalents of CO to generate fac-W(PMe3)3(CO)3.

Carbon dioxide
W(PMe3)4(η2-CH2PMe2)H reacts with 3 atmosphere of 1:1 CO2/H2 gas mix to produce W(PMe3)4(κ2-O2CO)H2 and a bimetallacyclic compound.

Acids
HBF4 reacts with W(PMe3)4(η2-CH2PMe2)H in ether to add hydroxyl ligands to the tungsten and form [W(PMe3)4(OH)2H2][BF4]2. It can react with Na(K) alloy to form W(PMe3)4H4. It can also be further reacted with KH in tetrahydrofuran (THF) to replace the hydroxyl ligands with fluoro ligands, producing W(PMe3)4F2H2. Upon crystallization, W(PMe3)4F2H2 becomes the dodecahedral complex [W(PMe3)4F(H2O)H2]F. Inter-ligand hydrogen bonding exists between the fluoro and aqua ligands. Reacting this complex with water reverts it to W(PMe3)4F2H2. Reacting W(PMe3)4(η2-CH2PMe2)H with HF or HPF6 also forms the [W(PMe3)4F(H2O)H2]+ salt with the corresponding anion, F- and PF6-, respectively.

W(PMe3)4Cl2H2 can be generated with the reaction of W(PMe3)4(η2-CH2PMe2)H with HCl. This chlorine analogue can be converted to the W(PMe3)4Br2H2 and W(PMe3)4I2H2 with excess LiBr or NaI, respectively, in benzene.

Trifluoroacetic acid reacts with W(PMe3)4(η2-CH2PMe2)H to form [W(PMe3)4(κ2-O2CCF3)H][CF3CO2].

W(PMe3)4(η2-CH2PMe2)H can react with two equivalents of benzoic acid or pivalic acid, such that each equivalent binds differently. W(PMe3)4(η2-CH2PMe2)H first reacts to generate W(PMe3)4(κ2-O2CR)H. The addition of the second equivalent generates W(PMe3)3(κ2-O2CR)(κ1-O2CR)H2. The two carboxylate ligands are not distinguishable on the NMR timescale at room temperature despite their different bonding behavior due to fluxional rearrangement of their bonding behavior. The rearrangement can be sufficiently slowed for NMR analysis at –75°C. This complex can be used to trap water; the oxygen lone forms a L-type σ-bond with the metal while the hydrogens form hydrogen bonds with the carboxylate groups. Both experimental and computational work show that the oxygen atom has roughly sp3 hybridization. Furthermore, the contributing atoms to hydrogen binding (O-H - - - O) are not colinear, contrary to some prior misconceptions of metal-aqua complexes. The addition of water is reversible, and its removal is inducible with KH or molecular sieves.

π-systems
In 1-2 atmospheres of ethylene at room temperature, W(PMe3)4(η2-CH2PMe2)H reacts to form trans-W(PMe3)4(η2-C2H4)2.

Upon subjecting W(PMe3)4(η2-CH2PMe2)H to 2 atmospheres of ethylene at 60°C in the presence of light petroleum for a week, W(PMe3)2(η2-C4H6)2 is produced. W(PMe3)4(η2-CH2PMe2)H will ligate to buta-1,3-diene when the latter is in vast excess and in the presence of light petroleum at 50°C to make the same product as ethylene. W(PMe3)2(η2-C4H6)2 produces yellow crystals.

Much like with ethylene, propylene (2 atm) also forms C-C bonds upon reaction with W(PMe3)4(η2-CH2PMe2)H and light petroleum at 70°C. The resultant product is W(PMe3)3[η-CH2=C(Me)CH=C(cis-Me)H]H2.

W(PMe3)4(η2-CH2PMe2)H, upon reaction with cyclopentadiene in light petroleum for five days, binds cyclopentadiene and dissociates two PMe3 ligands to generate W(η5-C5H5)(PMe3)3H, W(PMe3)4H4, W(PMe3)3H6, and trace W(η5-C5H5)2H2. The crystals of this mixture are yellow and air-sensitive. In the reaction with quinoxaline (QoxH,HH) and its derivatives 6-methylquinoxaline (QoxMe,HH) and 6,7-dimethylquinoxaline (QoxMe,MeH), W(PMe3)4(η2-CH2PMe2)H forms [κ2-C2-C6RR'H2(NC)2]W(PMe3)4, (η4-C2N2-QoxR,R'H)W(PMe3)3H2 (vide infra), and W(PMe3)4H2 (R,R'=H, Me), wherein the first listed product is generated from C-C bond cleavage to form two W=C=B bond motifs. The latter two products are hypothesized to be formed from H2 generated from the C-C bond cleavage.

Methanol
W(PMe3)4(η2-CH2PMe2)H, upon addition of methanol in an ethylene atmosphere, can form W(PMe3)4(CO)H2.

W(PMe3)4(η2-CH2PMe2)H, upon MeOH ligation in an η2-fashion, dissociates PMe3 and forms W(PMe3)4(η2-CH2O)H2. This complex undergoes many similar reaction pathways as its precursor retron.

Silanes
The reaction of W(PMe3)4(η2-CH2PMe2)H with SiH4 with light petroleum leads to the dissociation of one PMe3 ligand and activation of two Si-H bonds of separate SiH4 molecules to yield W(PMe3)4(SiH3)2H2.

Unlike the reaction with silane, W(PMe3)4(η2-CH2PMe2)H reacts with SiPhH3 to form three products: W(PMe3)3(SiPhH2)(SiH2SiPh2H)H4, W(PMe3)4(SiH2SiPh2H)H3, and [W(PMe3)2(SiHPh2)H2](μ-Si,P-SiHPhCH2PMe2)(μ‑SiH2)[W(PMe3)3H2]. W(PMe3)3(SiPhH2)(SiH2SiPh2H)H4 can be converted to W(PMe3)4(SiH2SiPh2H)H3 upon addition of PMe3.

W(PMe3)4(η2-CH2PMe2)H reacts with SiPh2H2 to form the σ-silane complex W(PMe3)3(SiPh2H2)H4.

Tungsten-tetrel multiple bonding
W(PMe3)4(η2-CH2PMe2)H, in pentane and at –20°C, reacts with Ge(C6H3-2,6-Trip2)Cl (Trip=C6H2-2,4,6-iPr3, iPr=CH(CH3)2) to dissociate PMe3 and generate trans-[Cl(H)(PMe3)3W{=Ge(C6H3-2,6-Trip2)(CH2PMe2)}]. This green, air-sensitive complex can heated at 50°C with toluene or left in ambient conditions with either toluene or pentane to yield the Ge≡C bond-containing complex, trans-[Cl(PMe3)4W≡Ge-C6H3-2,6-Trip2]. This brown, air-sensitive complex can also be directly generated from W(PMe3)4(η2-CH2PMe2)H by heating with toluene and Ge(C6H3-2,6-Trip2)Cl at 50°C. trans-[Cl(PMe3)4W≡Ge-C6H3-2,6-Trip2] is, in turn, also a retron for further chemistry by substitution of the labile chloride ligand. Upon addition of lithium iodide in ether, chloride is substituted for iodide, forming red-brown trans-[I(PMe3)4W≡Ge-C6H3-2,6-Trip2]. With lithium dimethylamine in THF, the chloride is substituted for a hydride, generating red-brown, air-sensitive trans-[H(PMe3)4W≡Ge-C6H3-2,6-Trip2]. With potassium thiocynate in THF, chloride is substituted for thiocynate, forming dark brown trans-[(NCS)(PMe3)4W≡Ge-C6H3-2,6-Trip2]. W(PMe3)4(η2-CH2PMe2)H with 0.5 equivalent of {Pb(Trip)Br2}2 and in toluene at 50°C produces (PMe3)4BrW{≡Pb(C6H3-2,6-Trip2)}. Upon addition of lithium dimethylamine in THF, Br(PMe3)4W{≡Pb(C6H3-2,6-Trip2)} converts to brown, air-sensitive H(PMe3)4W{≡Pb(C6H3-2,6-Trip2)}. Alternatively, W(PMe3)4(η2-CH2PMe2)H, with 0.5 equivalent of {Pb(Trip)NMe2}2 (produced from the reaction of {Pb(Trip)Br2}2 with lithium dimethylamine) in toluene and at 80°C, also produces H(PMe3)4W{≡Pb(C6H3-2,6-Trip2)}.

Tungsten-chalcogenide multiple bonding
W(PMe3)4(η2-CH2PMe2)H reacts with an equivalent of H2Se to generate the green, mono-selenido complex W(PMe3)4SeH2. Unique to this intermediate is the ability to generate mixed-terminal chalcogenide complexes, as the addition of H2S and elimination of H2 generates W(PMe3)4(Se)S. Alternatively, upon addition of another equivalent of H2Se and elimination of H2, green trans-W(PMe3)4Se2 is formed. Upon reaction with either formaldehyde or benzaldehyde, trans-W(PMe3)4Se2 reversibly becomes blue-green cisoid W(PMe3)2Se2(η2-OCHR) (R=H, Ph). The PMe3 ligands of this complex are noted to be weakly ligating, evidenced by the easy of ligand substitution. The reverse reaction can be performed by adding PMe3. This η2 ligand (which Rabinovich and Parkin argue to be a WVI metallaoxirane in some instances but calls a WIV aldehyde in others) is also weakly bound. trans-W(PMe3)4Se2 can also react with tBuNC (tBu=C(CH3)3), forming trans, trans, trans-W(PMe3)2(CNtBu)2Se2. Two equivalents of H2S's hydrogens are eliminated upon reaction with W(PMe3)4(η2-CH2PMe2)H in pentane to give yield the yellow thiol-ligated complex W(PMe3)4(SH2)H2. The complex then undergoes reduction elimination in the presence of a hydrogen trap, producing two equivalents of H2 (in the form of W(PMe3)4H3(OC6H5)) (vide supra) and purple trans-W(PMe3)4S2. Two equivalents of PMe3 are reversibly lost upon reaction with RCHO (R=H, CH3, C6H5, para-C6H4CH3, para-C6H4OCH3) to yield red-purple, 16 electron, cisoid W(PMe3)2S2(η2-OCHR). Alternatively, W(PMe3)2S2(η2-OCHR) can be reacted with tBuNC to generate trans, trans, trans-W(PMe3)2(CNtBu)2S2. A more general reaction of isocynanide ligand substitution was demonstrated for W(PMe3)2S2(η2-OCHR), to which RNC (R=iPr, tBu, ortho-C6H11) was added to form trans, trans, trans-W(PMe3)2(CNR)2S2. Unlike with sulfur, the formation of W=Te bonds with W(PMe3)4(η2-CH2PMe2)H requires the use of elemental tellurium, wherein it is hypothesized the PMe3 acts as a catalyst to form PMe3Te and deliver Te to W(PMe3)4(η2-CH2PMe2)H. The resultant compound is red-brown trans-W(PMe3)4Te2, the first of its kind with a terminal telluride ligand. Like trans-W(PMe3)4S2, trans-W(PMe3)4Te2 reacts reversibly with aldehydes (e.g., formaldehyde and benzaldehyde) to form cisoid, red-brown, diamagnetic W(PMe3)2Te2(η2-OCHR) (R=H, Ph). Both trans-W(PMe3)4Te2 and W(PMe3)2Te2(η2-OCHR) can react with tBuNC to form dark green-brown W(PMe3)(η2-Te2)(tBuNC)4, wherein the planar tBuNC ligands show significant π-backbonding as evidenced by the sp2-like geometry at N.

Hydrodesulfurization
The ability for W(PMe3)4(η2-CH2PMe2)H to desulfurize thiophenes was demonstrated with thiophene, benzothiophene, and dibenzothiophene. The pre-hydrogenation complexes result form C-S bond cleavage. In the case of thiopene, one C-S bond is cleaved and two PMe3 ligands dissociate to generate (η5-C4H5S)W(PMe3)2(η2-CH2PMe2). Hydrogenation changes η5-C4H5S to η1-C4H9S, generating W(PMe3)4(C4H9S)H3. Heating the complex at 100°C dissociates but-1-ene. Benzothiophene similarly experiences one C-S bond cleavage, leading to two PMe3 ligand dissociations to generate both (κ1,η2-CH2CHC6H4S)W(PMe3)3(η2-CH2PMe2) and (κ1,η2-CH2CC6H4S)W(PMe3)4. Upon hydrogenation, both products give W(PMe3)4(SC6H4CH2CH3)H3. Further heating at 100°C dissociates ethylbenzene. The desulfurization of dibenzothiophene proceeds through the ditungsten complex, [(κ2-C12H8)W(PMe3)](μ-S)(μ-CH2PMe2)(μ-PMe2)[W(PMe3)3]. Hydrogenation at 60°C liberates biphenyl.

C-H bond activation
The highly electron-rich nature of W(PMe3)4(η2-CH2PMe2)H allows it to activate C-H bonds. Moreover, the reaction with phenol/phenol-derivatives exhibit highly selective activation of bonds with specific hybridizations. If there is a methyl group in either of the sites ortho to the hydroxyl group, W(PMe3)4(η2-CH2PMe2)H will activate the sp3 site, forming a five-membered oxometallacycle. Complexes of this type include W(PMe3)4(κ2-OC6H2(CH2)RR')H2 (R, R'=H, CH3) If there is no methyl group in either ortho-site, W(PMe3)4(η2-CH2PMe2)H will activate the ortho sp2 site, forming a four-membered oxometallacycle. Complexes of this type include W(PMe3)4(κ2-OC6H3R)H2 (R=H, Et, iPr, tBu, Ph; Et=CH2CH3; Ph=C6H5). The lack of formation of the dogmatically stable six-membered ring may be due to the 18 electron nature of the tungsten complex not needing extra electron density donation from O→M π-bonding that is maximized with a six-membered cycle. The proposed mechanism for the formation of the oxometallacycle was based on PhOD, wherein the first step is the attack of the W-C bond, forming the W-O and a C-D bond in W(PMe3)4(PMe2CH2D)(OPh)H. The deuterated PMe3 can dissociate and substitute the non-deuterated PMe3's, leading to a statistical distribution of deuterium in the PMe3 ligands. Finally, the C-H bond is activated, creating a hydride and the oxometallacycle.

These complexes can undergo reductive elimination to form the intermediate W(PMe3)4(κ2-OC6H2(CH3)RR')H (R, R'=H, CH3) or W(PMe3)4(κ2-OC6H4R)H (R=H, Et, iPr, tBu; Et=CH2CH3) before reacting with either H2 or D2, forming, through hydrogenation, W(PMe3)4(κ2-OC6H2(CH3)RR')H3 (R, R'=H, CH3) or W(PMe3)4(κ2-OC6H4R)H3 (R=H, Et, iPr, tBu; Et=CH2CH3) or, through deuteration, W(PMe3)4(κ2-OC6H2(CH3)RR')HD2 (R, R'=H, CH3) or W(PMe3)4(κ2-OC6H4R)HD2 (R=H, Et, iPr, tBu; Et=CH2CH3). The equilibria are modulated by the aryl substituents. HD can dissociate through reductive elimination, leading to reformation of the κ2 complexes. These equilibria lead to the statistical distribution of deuterium throughout the hydride sites and sp2 ortho-sites at room temperature and ortho-methyl substituents upon heating at 55°C for 24 hours.

2,2′-methylenebis(4,6-dimethylphenol) (CH2(ArMe2OH)2) ligates to W(PMe3)4(η2-CH2PMe2)H, forming [κ2-O,C-CH2(ArMe2OH){(C6H2Me)(CH2)O}]W(PMe3)4H2. 2,2′-methylenebis(4,6-dimethylphenol) binds in the same fashion as other ortho-methylated phenols until the resultant complex is heated at 60°C. This generates both [κ2,η2-CH2(ArMe2O)2]W(PMe3)3H2 and [κ3-CH(ArMe2O)2]W(PMe3)3H3. The former complex forms an agostic bond from a hydrogen atom to the tungsten atom, whereas the latter complex has a hydride. Both structures exist in equilibrium — favoring the agostic-bond containing complex — in deuterated toluene, whereas only the latter exists in the solid-state. To imitate oxygen-rich surfaces (e.g., silica) for catalysis, W(PMe3)4(η2-CH2PMe2)H was reacted with a calixarene, para-tert-butyl-calix[4]arene. This reaction forms {[para-tert-butyl-Calix(OH)2(O)2]W(PMe3)3H2} which is in equilibrium with {[para-tert-butyl-Calix(OH)2(-H)(O)2]W(PMe3)3H3}, where (-H) denotes the loss of a hydrogen atom. Both are seen in deuterated benzene solution, whereas only the latter is seen in the solid-state, unlike the biphenol complex. Based on 2D-NMR, the calixarene complex exchanges rapidly between these two binding motifs. Furthermore, the W(PMe3)3H2 fragment can migrate across the hydroxyl rim of the calixarene. This exchange leads to selective exchange within the endo-hydride sites, as evidenced by the addition of deuterium. Upon addition of (PhC)2, the tungsten binds to all calixarene phenolates simultaneously, generating para-tert-butyl-Calix(O)4]W=C(Ph)Ar (Ar=PhCC(Ph)CH2Ph).

Alkylidene generation
Upon the addition of bromobenzene, iodobenzene, or para-bromotoluene, W(PMe3)4(η2-CH2PMe2)H undergoes hydride abstraction to form the cation [W(PMe3)4(η2-CHPMe2)H]+ with the corresponding halide anion. The metallocycle was determined to be an alkylidene based on the W-C bond length of the crystal structure. η2-CHPMe2 is an LX2 ligand, where the Fenske-Hall molecular orbitals indicate the phosphorus atom as the L part and the carbon atom as the X2 part. LiAlH4 can be used to do the reverse reaction from [W(PMe3)4(η2-CHPMe2)H]+ to W(PMe3)4(η2-CH2PMe2)H.

C-C bond activation mechanism
The novel activation of the aromatic C-C bond in QoxH by W(PMe3)4(η2-CH2PMe2)H under relatively mundane conditions inspired mechanistic theorizations. In their original publication, Sattler and Parkin suggested a mechanism in QoxH first acts as an L-type ligand from the N lone pair. The Qox ligand then changes its bonding behavior, with the bonding atoms shifting counterclockwise per Qox's numbering scheme. Upon reaching η2-C2 binding, the complex undergoes reductive elimination of its two hydrides to form H2. Finally, the complex cleaves its C-C bond to form the two W=C bonds.

Miscione and coworkers — using the B3LYP functional with energy-adjusted pseudopotential and DZVP basis sets — provided the first computational study of the proposed mechanism, wherein they provided a few pathways, building on Sattler and Parkin's work. The first pathway suggests that the hydride moves towards the tucked-in alkyl ligand to form W(PMe3)5 before QoxH binds. Upon the loss of a PMe3 ligand, Qox can then bond in an η2-N,C fashion, forming a hydride which subsequently moves to be trans to Qox. In the second pathway, PMe3 occurs first, followed by QoxH's ligation. Then, the agostic interaction is transformed into a standard PMe3 L-type ligand to join the first pathway in following the original proposed mechanism. The third pathway diverges from the first pathway at W(PMe3)5, wherein Qox instead interacts at the 2-H site before either bonding in a κ1-C fashion or losing a PMe3 to interact with both the 2-H and 3-H sites. Both intermediates then form (along with the loss of PMe3 in the former complex) a κ1-C complex with a 3-H interaction, before rejoining the original mechanism at the η2-C2 complex. Of these paths, path 2 is the least favored due to the ~30-40 kcal/mol energy barrier in breaking the agnostic interaction. Paths 1 and 3 are reported to be of roughly equal thermodynamic favorability with energy barriers mostly around 10-20 kcal/mol, until the maximum of the energy surface, the three-membered ring-containing η2-C2 intermediate (33.7 kcal/mol higher than W(PMe3)4(η2-CH2PMe2)H). Miscione and coworker's results substantiate Sattler and Parkin's hypothesis that the ring strain in the η2-C2 complex facilitates the C-C bond cleavage. They also report the reaction as being slightly net endergonic by 3.3 kcal/mol.

Liu et al. — using the B3LYP* functional with the LANL2DZ and 6-31G(d,f) basis sets — proposed two mechanisms based on Sattler and Parkin's original proposal. Both pathways start by dissociating both equatorial PMe3 ligands in the beginning before binding QoxH and generating a κ1-N QoxH ligand. It then switches to η2-N,C-Qox with a hydride which must move to be trans to Qox. κ1-N Qox then transitions to κ1-C Qox, followed by the transformation into η2-N,C Qox. Dissociation of PMe3 follows suite. Liu ''et. al.'' 's mechanism suggests that the C-C bond is broken at this stage, with a two electron oxidation of tungsten to form a double bond to the already bound carbon and a single bond to the other. The latter carbon's C-H bond forms an agostic interaction with tungsten to account for the lost electron density. The complex then gains its second W-C bond along with a hydride ligand. At this point, the two pathways branch. In the first pathway, an axial PMe3 moves down to the equatorial plane along with loss of the W=C bonds and reformation of the C-C bond, allowing another PMe3 to associate and rejoining the original mechanism at the dihydride-containing η2-C2 Qox complex. The second pathway sees the two hydride ligands move such that they are cis to the W=C bonds before undergoing reductive elimination. PMe3 then associates, forming the final complex. Liu ''et. al.'' claims that the final step to C-C bond cleavage is the concerted, not stepwise, elimination of H2 and formation W=C bonds. Per their calculations, Sattler and Parkin's mechanism spans a range of 42.0 kcal/mol energy range, in large part due to the aforementioned concerted step. The second pathway was calculated to have energy barriers of ~10 kcal/mol in all steps post-branching, leaving the second PMe3 dissociation as the highest energy barrier in the mechanism. Liu et al. 's calculations suggest that the mechanism is exergonic, releasing a net 9.2 kcal/mol of energy.

Li and Yoshizawa — using the B3LYP* functional with the LANL2TZ(f) and 6-31G(d,f) basis sets — also proposed two mechanisms which start with ligand dissociations. Both mechanisms start with the dissociation of an equatorial PMe3 ligand, before diverging. The first pathway sees the dissociation of the second equatorial PMe3, leaving the agostic interaction and the hydride. This complex then binds to QoxH, generating a κ1-N QoxH ligand. Qox then changes its binding to the η2-N,C fashion, as well generating a hydride bond, before breaking the agostic interaction to form a PMe3 L-type interaction. Another PMe3 ligates before Qox switches to η2-C2-type bonding as well as an H2 ligand. H2 dissociation, followed by C-C bond cleavage, then leads to the final product. In the second pathway, the agostic bond is broken for a PMe3 L-type interaction after the first PMe3 dissociation. QoxH then binds in a κ1-N fashion before changing to η2-N,C with a hydride bond to tungsten and rejoining pathway 1. Li and Yoshizawa concluded that, between their pathways, pathway 1 is the most thermodynamically favorable. The reformation of PMe3 after immediately after the first PMe3 dissociation in pathway 2 has a barrier of 26.3 kcal/mol relative to W(PMe3)4(η2-CH2PMe2)H. In contrast, the energy maximum of pathway 1 is from the H2 dissociation step shared by both pathways. Overall, Li and Yoshizawa's work suggest that the C-C bond mechanism is exergonic overall, with the product being 18.5 kcal/mol lower in energy relative to W(PMe3)4(η2-CH2PMe2)H.

η4-C2N2 quinoxaline binding
The η4-C2N2-QoxH ligand is a novel binding behavior discovered from the reaction of W(PMe3)4(η2-CH2PMe2)H with QoxH. Miscione et al. and Liu et al. also investigated these mechanisms. The former group suggests that upon formation of W(PMe3)5 (vide infra), the tungsten undergoes the oxidative addition of H2, forming hydride bonds. Then, one PMe3 ligand is dissociated, allowing QoxH to bind, first in a η2-N,C fashion before switching to the final η4-C2N2 fashion via a 7.3 kcal/mol rearrangement energy barrier. The latter group suggests that one PMe3 first dissociates, followed by the oxidative addition of H2, forming an ML6 complex. One of the axial PMe3 ligands is lost, allowing QoxH to bind, forming the η4-C2N2-QoxH ligand. Both sets of calculations agree that the mechanism is net exergonic, with the product being ~20 kcal/mol lower in energy than W(PMe3)4(η2-CH2PMe2)H.