Diphosphenes

In 1877, Köhler and Michaelis claimed what would have been the first isolated diphosphene (PhP=PPh),^[2] a publication that inspired little. However, the heavier pnictogens were known to form oligomers in oxidation state I, and by 1958, chemists had begun to reconsider the structure of Köhler and Michaelis' product.^{[3][original research?]} During the subsequent decade (the 1960s), molecular weight determination^[4] and X-ray crystallographic analysis^[5] proved that this "diphosphene" only had P-P single bonds and was in fact primarily a four-membered ring of the form (PPh)₄. Nevertheless, the contemporary discoveries of the first diphosphorus ylide and first phosphaalkene suggested that compounds with multiply-bonded phosphorus could be made.^[6]

The modern diphosphene field properly begins with field Yoshifuji et al's isolation of a more sterically-hindered diphosphene in 1981.^[6] That compound's P-P bond distance is 2.034 Å, which is much shorter than the average bond length in (C₆H₅P)₅ (2.217 Å) and (C₆H₅P)₆ (2.237 Å) and indicates double-bond character.^[7]

Following Maasaka Yoshifuji and his coworkers' 1981 isolation of bis(2,4,6-tri-tert-butylphenyl)diphosphene,^[7] most disphosphene syntheses dehalogenate a bulkyl alkyldichlorophosphine with an active metal.^[8] Such a synthesis works for arylphosphenes,^[7] trisalkylsilylphosphines,^[8] or N-heterocyclic boro-phosphines.^[9]

In 2019, Stephan and co-workers at the University of Toronto reported the first examples of di-vinyl-substituted diphosphenes via a ring opening/dimerization process from kinetically unstable 2H-phosphirenes. However, the conjugation caused the compounds to exhibit reactivity closer to a phosphinidene.^[10]

Cyclic voltammetry and UV/Vis spectra illustrate that boryl-substituted diphosphenes have lower LUMO level and larger HOMO-LUMO gap than aryl-substituted diphosphenes.^[9]

X-ray analysis indicates certain important bond lengths and angles of the first diphosphene, bis(2,4,6-tri-tert-butylphenyl)diphosphene: P-P = 2.034 (2) Å; P-C = 1.826 (2) Å; ${\displaystyle \angle }$ P-P-C = 102.8 (1)^o; ${\displaystyle \angle }$ C-P-P-C = 172.2 (1)^o.^[7] Compared with the bond length of a P-P single bond in H₂PPH₂ (2.238 Å),^[11] the P-P bond distance is much shorter, which reveals double bond character. The trans orientation is the thermodynamically preferred isomer.^[12]

Diphosphene compounds usually exhibit a symmetry-allowed (${\displaystyle \pi \rightarrow \pi ^{*}}$ ) (intense) and symmetry-forbidden electronic transitions (${\displaystyle n\rightarrow \pi ^{*}}$ ) (weak).^[13] Raman spectroscopy observes significant enhancement of P=P stretch in the resonance with allowed electron ${\displaystyle \pi \rightarrow \pi ^{*}}$  transition than with the forbidden ${\displaystyle n\rightarrow \pi ^{*}}$  transition due to different geometries of excited states and enhancement mechanism.^[14] Also the observed strong Raman shifts for (CH(SiMe
₃)
₂)
₂P
₂and (CH(SiMe₃)₂P=PC(SiMe₃)₂) suggest stronger dipnictenes feature^[which?] of diphosphene compared with P-P single bond.^{[15][failed verification]}

Diphosphenes react very similarly to olefins. Hydride reagents such as lithium aluminum hydride can reduce diphosphene to give stable diphosphanes:^[16] $${\displaystyle {\ce {ArP=PAr\ +\ AlH\rightarrow ArHP-PHAr}}}$$ 

Visible radiation induces cis-trans isomerization,^[12] although further irradiation can excite the molecule to a triplet diradical state. In triplet trans-HPPH, the P-P bond length is predicted to be 2.291 Å. It is not only longer than the P-P double bond in ground state trans-bis(2,4,6-tri-tert-butylphenyl)diphosphene, but also longer than that of P-P single bond in H₂PPH₂. Calculation of the dihedral angle of trans-HPPH suggests that it is almost 90 degree, which means the formation of ${\displaystyle \pi }$  and ${\displaystyle \pi ^{*}}$  P-P bonds is forbidden and σ bond is enhanced.^[17]

The compounds form a wide variety of transition metal alkene complexes (see § Coordination to transition metals), as well as the traditional complexation to the phosphorus lone pair, or to any aryl moieties present.

Diphosphene is inert to ground-state oxygen but cycloadds to triplet oxygen or ozone to give highly unstable phosphorus-oxygen rings that tend to attack the phosphorus' organyl substituents.^[18][19] The reaction with ozone is much more rapid and indicates a 2:1 (ozone:diphosphene) stoichiometry.^[19]

Carbenes add across the double bond, to give diphosphiranes, which further rearrange to 1,3-diphospha-allenes in strong bases.^[20] Unlike with olefins, the ylides traditionally called persistent carbenes instead tend to cleave the central bond, forming two phosphaalkene/phosphinidene compounds.^[21]

Diphosphenes can bind to transition metal either in a η¹ mode by donating a lone pair on phosphorus, or in a η² behavior via a ${\displaystyle \pi }$  interaction. If the bulky groups are aryl- groups, arene-coordinated products of η⁶-type coordination are also possible.

In 1983, Philip P. Power synthesized a transition-metal complex containing P=P double bond (trans–{[Fe(CO)
₄][PCH(SiMe
₃)
₂]2}) via a simple one-step procedure.^[22] They mixed Na₂[Fe(CO)₄] and dichlorobis(trimethylsilyl)methylphosphine and got dark red-brown crystals, which was the first complex that contained an unbridged P-P double bond. Each phosphorus exhibited terminal coordination nature and the P-P distance was essentially unchanged. Later in 1983, A. H. Cowley reported ArP=PArFe(CO)₅ (with Ar=2,4,6-tri-tert-butylphenyl) by treating diphosephene with Fe₂(CO)₉ or Na₂Fe(CO)₄.^[23] In this synthesis procedure, there was only one terminal P-coordination and P-P double bond had Z configuration. Apart from iron, other similar transition metal complexes by reacting diphosphenes with transition metal carbonyls of nickel, tungsten, and chromium were discovered and they all exhibited Z configuration. M. Yoshifuji proved E/Z isomerization can take place under lighting, probably via migration of the metal moiety from one side to the other.

Apart from the very bulky substituents, a η²-coordination of diphosphene to a metal is also possible to stabilize the P-P double bond. In 1982, K. R. Dixon et al. synthesized platinum and palladium complexes (M(PhP=PPh)L₂) (with M=Pt or Pd and L=(PPh3)₂ or Ph₂P[CH₂]₂PPh₂), which contained side-on coordination.^[24] Different from η¹ coordination complex, where P-P still kept the double bond nature, P-P distance in side-on coordination complexes (2.121Å in Pd(PhP=PPh)PPh₃CH₂CH₂PPh₃) was significantly longer than that in non-coordinated bis(2,4,6-tri-tert-butylphenyl)diphosphene.

If there are aryl- groups on phosphorus, transition-metal can not only bind to the phosphorus directly, but also form arene-coordinated products of η⁶-type coordination. Refluxing diphosphene in 1,4-dioxane with the excess of Cr(CO)₆ can generate mono and bis arene tricarbonylchromium(0) complexes.^[25]

• Diazene
• Double bond rule