Chemistry of Several Sterically Bulky Molecules with P=P, P=C, and C≡P Bond

Several sterically protected, low-coordinate organophosphorus compounds with P=P, P=C, and C≡P bond are described in this study. Molecules such as diphosphenes, phosphaalkenes, 1-phosphaallenes, 1,3-diphosphaallenes, 3,4-diphosphinidenecyclobutenes, and phosphaalkynes are stabilized with an extremely bulky 2,4,6-tri-t-butylphenyl (Mes*) group. The synthesis, structures, physical, and chemical properties of these molecules are discussed, together with some successful applications in catalytic organic reactions.


Introduction
Low-coordinate organophosphorus compounds with coordination numbers of 1 or 2 were once believed to only exist as unstable species. However, by introducing a sterically bulky group such as 2,4,6-tri-t-butylphenyl into a molecule, various kinds of unusual phosphorus compounds have been isolated as kinetically stable species since 1981. In this review, the characteristics of several compounds carrying P=P, P=C, and P≡C bondsincluding the preparation, structural, physical, theoretical and chemical aspects, as well as some catalytic applications to organic reactions-are described.

Steric Protection for Stabilization of Unstable Compounds
As shown in Scheme 1, Okazaki and Inamoto [1] found that 2,4,6-tri-t-butylaniline (1) is oxidized with 2 moles of perbenzoic acid in dichloromethane to yield the corresponding nitrosobenzene (2), which is stable in air at room temperature and can be purified by column chromatography with alumina. In contrast to normal nitroso compounds such as unsubstituted nitrosobenzene, compound 2 is emerald green and does not dimerize to 1,3,2,4-dioxadiazetidine (3) either in solution or the solid state due to the steric hindrance. In general, the steric protection technique seems promising for the stabilization of unstable compounds.

Introduction
Low-coordinate organophosphorus compounds with coordination numbers of 1 or 2 were once believed to only exist as unstable species. However, by introducing a sterically bulky group such as 2,4,6-tri-t-butylphenyl into a molecule, various kinds of unusual phosphorus compounds have been isolated as kinetically stable species since 1981. In this review, the characteristics of several compounds carrying P=P, P=C, and P≡C bondsincluding the preparation, structural, physical, theoretical and chemical aspects, as well as some catalytic applications to organic reactions-are described.

Steric Protection for Stabilization of Unstable Compounds
As shown in Scheme 1, Okazaki and Inamoto [1] found that 2,4,6-tri-t-butylaniline (1) is oxidized with 2 moles of perbenzoic acid in dichloromethane to yield the corresponding nitrosobenzene (2), which is stable in air at room temperature and can be purified by column chromatography with alumina. In contrast to normal nitroso compounds such as unsubstituted nitrosobenzene, compound 2 is emerald green and does not dimerize to 1,3,2,4-dioxadiazetidine (3) either in solution or the solid state due to the steric hindrance. In general, the steric protection technique seems promising for the stabilization of unstable compounds.
In an attempt to introduce a phosphorus moiety to the phenyl ring, phosphorus trichloride was allowed to react with 1,3,5-tri-t-butylbezene (4) in the presence of aluminum chloride. As reported by Cook [2], in place of the expected 2,4,6-tri-t-butylphenylphosphinic chloride (5), after partial hydrolysis, t-butyl-3,5-di-t-butylphenylphosphinic chloride (6) was obtained, indicating that one of the t-butyl groups migrated from the aromatic ring to the phosphorus atom under Friedel-Crafts reaction conditions, as shown in Scheme 2 [3]. It turned out that t-butyl is prone to move around on the ring to rearrange the original positions under the strong acidic conditions, probably due to a labile t-butyl cation, e.g., 7 and 8. Therefore, another strategy was needed to introduce a phosphorus functional group on a bulky benzene nucleus to utilize 2,4,6-tri-tbutylphenyl as a sterically demanding group (hereafter abbreviated as Mes*).
As shown in Scheme 5, dichloride 15 is allowed to react with magnesium in THF to yield a stable diphosphene 17 as an orange red crystalline material with an mp of 175-176 °C [12]. Some selected physical and X-ray data are listed in Table 1. It is noteworthy tha (E)-bis(2,4,6-tri-t-butylphenyl)diphosphene (17), which turned out to be a true "phos phobeneze," has a short P-P bond distance (2.034(2) Å) and shows a low 31 P NMR chem ical shift (δP 492.4 ppm).   (2), P-C 1.862(2) Å; P-C-C 102.8(1), C-P-P-C 172.2(1)° [12] According to the crystallographic analysis, the bond distance between the two phos phorus atoms is about 10% shorter than the regular P-P single bond, and the dihedra angle ∠C-P-P-C reveals a planar molecular system with (E)-configuration.
However, in our initial paper [12], we misreported the 31 P NMR chemical shift of 17 as δP -59.00 ppm and immediately corrected the value for 17 to 492.4 ppm (C6D6) because it is one of the most important physical characteristics [13]. In the meantime, Lappert [15 critcized our initially reported chemical shift but Cowley [16] argued our structure of di phosphene itself as diphosphane Mes*P(H)-P(H)Mes* ( 31 P NMR δP of -64.4 ppm for dl and δP of -65.0 ppm for meso) [17], despite our unambiguously determined X-ray analysis [12] On the other hand, it turned out that the coupling constant between two phosphorus nu clei, 1 JPP, is large and has been confirmed by NMR measurement for unsymmetrical di phosphene 18 prepared from the dehydrochlorination reaction of the primary phosphane 16 and RPCl2 with a base such as DBU (1,8-diazabicyclo [5.4.0]undec-7-ene) [18] (Scheme 5, second row). Some selected coupling constants and chemical shifts are listed in Table 2 [19].  As shown in Scheme 5, dichloride 15 is allowed to react with magnesium in THF to yield a stable diphosphene 17 as an orange red crystalline material with an mp of 175-176 °C [12]. Some selected physical and X-ray data are listed in Table 1. It is noteworthy that (E)-bis(2,4,6-tri-t-butylphenyl)diphosphene (17), which turned out to be a true "phosphobeneze," has a short P-P bond distance (2.034(2) Å) and shows a low 31 P NMR chemical shift (δP 492.4 ppm).   (2), P-C 1.862(2) Å; P-C-C 102.8(1), C-P-P-C 172.2(1)° [12] According to the crystallographic analysis, the bond distance between the two phosphorus atoms is about 10% shorter than the regular P-P single bond, and the dihedra angle ∠C-P-P-C reveals a planar molecular system with (E)-configuration.
However, in our initial paper [12], we misreported the 31 P NMR chemical shift of 17 as δP -59.00 ppm and immediately corrected the value for 17 to 492.4 ppm (C6D6) because it is one of the most important physical characteristics [13]. In the meantime, Lappert [15] critcized our initially reported chemical shift but Cowley [16] argued our structure of diphosphene itself as diphosphane Mes*P(H)-P(H)Mes* ( 31 P NMR δP of -64.4 ppm for dl and δP of -65.0 ppm for meso) [17], despite our unambiguously determined X-ray analysis [12] On the other hand, it turned out that the coupling constant between two phosphorus nuclei, 1 JPP, is large and has been confirmed by NMR measurement for unsymmetrical diphosphene 18 prepared from the dehydrochlorination reaction of the primary phosphane 16 and RPCl2 with a base such as DBU (1,8-diazabicyclo[5.4.0]undec-7-ene) [18] (Scheme 5, second row). Some selected coupling constants and chemical shifts are listed in Table 2 [19].   (2), P-C 1.862(2) Å; P-C-C 102.8(1), C-P-P-C 172.2(1) • [12] According to the crystallographic analysis, the bond distance between the two phosphorus atoms is about 10% shorter than the regular P-P single bond, and the dihedral angle ∠C-P-P-C reveals a planar molecular system with (E)-configuration.
However, in our initial paper [12], we misreported the 31 P NMR chemical shift of 17 as δ P −59.00 ppm and immediately corrected the value for 17 to 492.4 ppm (C 6 D 6 ) because it is one of the most important physical characteristics [13]. In the meantime, Lappert [15] critcized our initially reported chemical shift but Cowley [16] argued our structure of diphosphene itself as diphosphane Mes*P(H)-P(H)Mes* ( 31 P NMR δ P of −64.4 ppm for dl and δ P of −65.0 ppm for meso) [17], despite our unambiguously determined Xray analysis [12]. On the other hand, it turned out that the coupling constant between two phosphorus nuclei, 1 J PP , is large and has been confirmed by NMR measurement for unsymmetrical diphosphene 18 prepared from the dehydrochlorination reaction of the primary phosphane 16 and RPCl 2 with a base such as DBU (1,8-diazabicyclo[5.4.0]undec-7ene) [18] (Scheme 5, second row). Some selected coupling constants and chemical shifts are listed in Table 2 [19].  Mes = 2,4,6-Me3C6H2, Cp* = pentamethylcyclopentadienyl. 31 P NMR data for (Z)-diphosphenes 21 (δP 368 ppm) and 22 (δP 394 ppm) show a higher chemical shift than (E)-diphosphenes. As shown in Scheme 6, the formation of (Z)bis(2,4,6-tri-t-butylphenyl)diphosphene (21) is observed during the temperature-and wavelength-depending photolysis of (E)-diphosphene 17. The irradiation of 17 with a Hg lamp through a Pyrex filter at -40 °C [26], or argon-laser irradiation (514.5 nm) at −78 °C [27] yields an E/Z equilibrium mixture that returns to E-form upon warming, showing that (Z)-diphosphene 21 is thermally unstable due to the steric congestion caused by the two adjacent Mes* groups. Caminade reported the first order rate constant of E/Z isomerization reaction as ΔG ≠ 273 = 20.35 kcal/mol based on 31 P-NMR studies. When the photolysis is conducted without a Pyrex filter, regardless of the temperature between −78 °C and 0 °C [26], intramolecular insertion to a neighboring methyl group occurs and yields benzophosphaindane 24, most likely suggesting the generation of phosphinidene intermediate 23 during the photolysis. Scheme 6. Temperature-and wavelength-dependent photolysis of diphosphene 17: (a) Hg lamp irradiation through a Pyrex filter or Ar-laser irradiation at low temperature; (b) Hg lamp without a Pyrex filter.  Mes = 2,4,6-Me3C6H2, Cp* = pentamethylcyclopentadienyl. 31 P NMR data for (Z)-diphosphenes 21 (δP 368 ppm) and 22 (δP 394 ppm) show a higher chemical shift than (E)-diphosphenes. As shown in Scheme 6, the formation of (Z)bis(2,4,6-tri-t-butylphenyl)diphosphene (21) is observed during the temperature-and wavelength-depending photolysis of (E)-diphosphene 17. The irradiation of 17 with a Hg lamp through a Pyrex filter at -40 °C [26], or argon-laser irradiation (514.5 nm) at −78 °C [27] yields an E/Z equilibrium mixture that returns to E-form upon warming, showing that (Z)-diphosphene 21 is thermally unstable due to the steric congestion caused by the two adjacent Mes* groups. Caminade reported the first order rate constant of E/Z isomerization reaction as ΔG ≠ 273 = 20.35 kcal/mol based on 31 P-NMR studies. When the photolysis is conducted without a Pyrex filter, regardless of the temperature between −78 °C and 0 °C [26], intramolecular insertion to a neighboring methyl group occurs and yields benzophosphaindane 24, most likely suggesting the generation of phosphinidene intermediate 23 during the photolysis. Scheme 6. Temperature-and wavelength-dependent photolysis of diphosphene 17: (a) Hg lamp irradiation through a Pyrex filter or Ar-laser irradiation at low temperature; (b) Hg lamp without a Pyrex filter.
Z 394 -- [28] Mes = 2,4,6-Me 3 C 6 H 2 , Cp* = pentamethylcyclopentadienyl. 31 P NMR data for (Z)-diphosphenes 21 (δ P 368 ppm) and 22 (δ P 394 ppm) show a higher chemical shift than (E)-diphosphenes. As shown in Scheme 6, the formation of (Z)-bis(2,4,6-tri-t-butylphenyl)diphosphene (21) is observed during the temperature-and wavelength-depending photolysis of (E)-diphosphene 17. The irradiation of 17 with a Hg lamp through a Pyrex filter at -40 • C [26], or argon-laser irradiation (514.5 nm) at −78 • C [27] yields an E/Z equilibrium mixture that returns to E-form upon warming, showing that (Z)-diphosphene 21 is thermally unstable due to the steric congestion caused by the two adjacent Mes* groups. Caminade reported the first order rate constant of E/Z isomerization reaction as ∆G = 273 = 20.35 kcal/mol based on 31 P-NMR studies. When the photolysis is conducted without a Pyrex filter, regardless of the temperature between −78 • C and 0 • C [26], intramolecular insertion to a neighboring methyl group occurs and yields benzophosphaindane 24, most likely suggesting the generation of phosphinidene intermediate 23 during the photolysis.  Mes = 2,4,6-Me3C6H2, Cp* = pentamethylcyclopentadienyl. 31 P NMR data for (Z)-diphosphenes 21 (δP 368 ppm) and 22 (δP 394 ppm) show a higher chemical shift than (E)-diphosphenes. As shown in Scheme 6, the formation of (Z)bis(2,4,6-tri-t-butylphenyl)diphosphene (21) is observed during the temperature-and wavelength-depending photolysis of (E)-diphosphene 17. The irradiation of 17 with a Hg lamp through a Pyrex filter at -40 °C [26], or argon-laser irradiation (514.5 nm) at −78 °C [27] yields an E/Z equilibrium mixture that returns to E-form upon warming, showing that (Z)-diphosphene 21 is thermally unstable due to the steric congestion caused by the two adjacent Mes* groups. Caminade reported the first order rate constant of E/Z isomerization reaction as ΔG ≠ 273 = 20.35 kcal/mol based on 31 P-NMR studies. When the photolysis is conducted without a Pyrex filter, regardless of the temperature between −78 °C and 0 °C [26], intramolecular insertion to a neighboring methyl group occurs and yields benzophosphaindane 24, most likely suggesting the generation of phosphinidene intermediate 23 during the photolysis. In the course of the synthetic route from 15 to 17 (Scheme 5, first row), a "free" 23 does not seem to be generated since Protasiewicz failed to obtain 17 under controlled reaction conditions such as when employing extremely dried THF with activated magnesium metal [29], instead yielding 24.
Scheme 7 shows that (Z)-diphosphene 22 is prepared via photolytic ring closing metathesis (RCM) on bis(E-diphosphene) 25 [28] with a Xe-lamp in benzene for 30 min at room temperature, though the yield of 22 is not satisfactory.
Molecules 2022, 27, x FOR PEER REVIEW 5 of 21 In the course of the synthetic route from 15 to 17 (Scheme 5, first row), a "free" 23 does not seem to be generated since Protasiewicz failed to obtain 17 under controlled reaction conditions such as when employing extremely dried THF with activated magnesium metal [29], instead yielding 24.
Scheme 7 shows that (Z)-diphosphene 22 is prepared via photolytic ring closing metathesis (RCM) on bis(E-diphosphene) 25 [28] with a Xe-lamp in benzene for 30 min at room temperature, though the yield of 22 is not satisfactory. It should be noted that in the year of 1981, when a stable diphosphene 17 was isolated for the first time, West reported the successful isolation of a stable disilene (Mes2Si=SiMes2; Mes = 2,4,6-Me3C6H2) protected with four mesityl groups [30]. The molecular structures of these two compounds are simple but unusual and uncommon based on the traditional knowledge of organic, inorganic, and/or physical chemistry [31][32][33][34][35][36][37]. Since then, however, various kinds of sterically protected stable low-coordinate phosphorus compounds have been successfully prepared and characterized. Many sophisticated or simple protective groups other than Mes* with stronger or weaker steric ability, with or without electronic effect, or for special or common purposes have been developed, and some early examples have already been reviewed [18,38,39].

Chemical Reactivity of Diphosphenes
Diphosphenes demonstrate a variety of chemical reactions including photolysis, oxidation, reduction, hydrolysis, sulfurization, carbene-addition, and coordination to transition-metals [31,32]. In this section, selected examples are described in more detail.
Diphosphene 17 is allowed to react with sulfur in triethylamine at room temperature overnight to yield a stable diphosphene mono-sulfide 26, and the desulfurization to the starting diphosphene 17 is carried out with hexamethylphosphorous triamide in benzene at room temperature for 2 h. Sulfide 26 is converted to thiadiphosphirane 27 upon the irradiation of light with a medium-pressure Hg lamp at 0 °C for 5 min or heating in toluene for 30 min at 95 °C (Scheme 8) [40,41]. Unsymmetrical diphosphene 18B is allowed to react with (THF)Cr(CO)5 to yield an end-on complex 28 as a result of coordination at the less hindered site. However, complex 28 is converted to isomer 29 upon the irradiation of light with a medium-pressure Hg lamp in hexane at 0 °C for 5 min due to the E/Z isomerization around the P=P bond. The Z-ligand on chromium in 29 is thermally stable, probably because the steric environment reduces the enthalpy change between 28 and 29, in contrast to that of free ligand 18B (Scheme 9) [42]. It should be noted that in the year of 1981, when a stable diphosphene 17 was isolated for the first time, West reported the successful isolation of a stable disilene (Mes 2 Si = SiMes 2 ; Mes = 2,4,6-Me 3 C 6 H 2 ) protected with four mesityl groups [30]. The molecular structures of these two compounds are simple but unusual and uncommon based on the traditional knowledge of organic, inorganic, and/or physical chemistry [31][32][33][34][35][36][37]. Since then, however, various kinds of sterically protected stable low-coordinate phosphorus compounds have been successfully prepared and characterized. Many sophisticated or simple protective groups other than Mes* with stronger or weaker steric ability, with or without electronic effect, or for special or common purposes have been developed, and some early examples have already been reviewed [18,38,39].

Chemical Reactivity of Diphosphenes
Diphosphenes demonstrate a variety of chemical reactions including photolysis, oxidation, reduction, hydrolysis, sulfurization, carbene-addition, and coordination to transitionmetals [31,32]. In this section, selected examples are described in more detail.
Diphosphene 17 is allowed to react with sulfur in triethylamine at room temperature overnight to yield a stable diphosphene mono-sulfide 26, and the desulfurization to the starting diphosphene 17 is carried out with hexamethylphosphorous triamide in benzene at room temperature for 2 h. Sulfide 26 is converted to thiadiphosphirane 27 upon the irradiation of light with a medium-pressure Hg lamp at 0 • C for 5 min or heating in toluene for 30 min at 95 • C (Scheme 8) [40,41].
Molecules 2022, 27, x FOR PEER REVIEW 5 of 21 In the course of the synthetic route from 15 to 17 (Scheme 5, first row), a "free" 23 does not seem to be generated since Protasiewicz failed to obtain 17 under controlled reaction conditions such as when employing extremely dried THF with activated magnesium metal [29], instead yielding 24.
Scheme 7 shows that (Z)-diphosphene 22 is prepared via photolytic ring closing metathesis (RCM) on bis(E-diphosphene) 25 [28] with a Xe-lamp in benzene for 30 min at room temperature, though the yield of 22 is not satisfactory. It should be noted that in the year of 1981, when a stable diphosphene 17 was isolated for the first time, West reported the successful isolation of a stable disilene (Mes2Si=SiMes2; Mes = 2,4,6-Me3C6H2) protected with four mesityl groups [30]. The molecular structures of these two compounds are simple but unusual and uncommon based on the traditional knowledge of organic, inorganic, and/or physical chemistry [31][32][33][34][35][36][37]. Since then, however, various kinds of sterically protected stable low-coordinate phosphorus compounds have been successfully prepared and characterized. Many sophisticated or simple protective groups other than Mes* with stronger or weaker steric ability, with or without electronic effect, or for special or common purposes have been developed, and some early examples have already been reviewed [18,38,39].

Chemical Reactivity of Diphosphenes
Diphosphenes demonstrate a variety of chemical reactions including photolysis, oxidation, reduction, hydrolysis, sulfurization, carbene-addition, and coordination to transition-metals [31,32]. In this section, selected examples are described in more detail.
Diphosphene 17 is allowed to react with sulfur in triethylamine at room temperature overnight to yield a stable diphosphene mono-sulfide 26, and the desulfurization to the starting diphosphene 17 is carried out with hexamethylphosphorous triamide in benzene at room temperature for 2 h. Sulfide 26 is converted to thiadiphosphirane 27 upon the irradiation of light with a medium-pressure Hg lamp at 0 °C for 5 min or heating in toluene for 30 min at 95 °C (Scheme 8) [40,41]. Unsymmetrical diphosphene 18B is allowed to react with (THF)Cr(CO)5 to yield an end-on complex 28 as a result of coordination at the less hindered site. However, complex 28 is converted to isomer 29 upon the irradiation of light with a medium-pressure Hg lamp in hexane at 0 °C for 5 min due to the E/Z isomerization around the P=P bond. The Z-ligand on chromium in 29 is thermally stable, probably because the steric environment reduces the enthalpy change between 28 and 29, in contrast to that of free ligand 18B (Scheme 9) [42]. Unsymmetrical diphosphene 18B is allowed to react with (THF)Cr(CO) 5 to yield an end-on complex 28 as a result of coordination at the less hindered site. However, complex 28 is converted to isomer 29 upon the irradiation of light with a medium-pressure Hg lamp in hexane at 0 • C for 5 min due to the E/Z isomerization around the P=P bond. The Z-ligand on chromium in 29 is thermally stable, probably because the steric environment reduces the enthalpy change between 28 and 29, in contrast to that of free ligand 18B (Scheme 9) [42].

Stable Phosphaalkenes and the Related Compounds
In 1966, Märkl reported the interesting preparation of a stable phosphorus-containing heterocyclic compound from 2,4,6-triphenylpyrylium tetrafluoroborate (30) [43]. The obtained triphenylphosphabenzene (or phosphinine) (31) contains a P=C bond in its canonical form, but the stability is estimated to be due to delocalization or aromaticity (Scheme 10).
Scheme 12. Preparation of phosphaalkene 33 and the Peterson reaction.
Interestingly, phosphaethene 33 in the (E)-form is isomerized in benzene upon the irradiation of light with a 100 W medium pressure Hg lamp at 0 °C for 6 h to yield an equilibrium mixture (3:7) with the corresponding (Z)-phosphaethene 34 [45], as shown in Scheme 13. After separation with silica-gel column chromatography, both isomers were Later in 1978, Bickelhaupt reported a thermally stable phosphaalkene 32 with a localized P=C double bond that is sterically protected with mesityl group [44], as shown in Scheme 11. This successful isolation of 32 triggered our interest in the field of sterically protected, low-coordinate phosphorus compounds.

Stable Phosphaalkenes and the Related Compounds
In 1966, Märkl reported the interesting preparation of a stable phosphorus-containing heterocyclic compound from 2,4,6-triphenylpyrylium tetrafluoroborate (30) [43]. The obtained triphenylphosphabenzene (or phosphinine) (31) contains a P=C bond in its canonical form, but the stability is estimated to be due to delocalization or aromaticity (Scheme 10).
Scheme 12. Preparation of phosphaalkene 33 and the Peterson reaction.

Stable Phosphaalkenes and the Related Compounds
In 1966, Märkl reported the interesting preparation of a stable phosphorus-containing heterocyclic compound from 2,4,6-triphenylpyrylium tetrafluoroborate (30) [43]. The obtained triphenylphosphabenzene (or phosphinine) (31) contains a P=C bond in its canonical form, but the stability is estimated to be due to delocalization or aromaticity (Scheme 10).
Scheme 12. Preparation of phosphaalkene 33 and the Peterson reaction.
Interestingly, phosphaethene 33 in the (E)-form is isomerized in benzene upon the irradiation of light with a 100 W medium pressure Hg lamp at 0 °C for 6 h to yield an equilibrium mixture (3:7) with the corresponding (Z)-phosphaethene 34 [45], as shown in Scheme 13. After separation with silica-gel column chromatography, both isomers were Scheme 12. Preparation of phosphaalkene 33 and the Peterson reaction.
Interestingly, phosphaethene 33 in the (E)-form is isomerized in benzene upon the irradiation of light with a 100 W medium pressure Hg lamp at 0 • C for 6 h to yield an equilibrium mixture (3:7) with the corresponding (Z)-phosphaethene 34 [45], as shown in Scheme 13. After separation with silica-gel column chromatography, both isomers were analyzed by X-ray crystallography [46], and it was found that separated (Z)-phosphaethene 34 is thermally stable and does not isomerize to 33 even it is heated at 100 • C in toluene for 24 h. analyzed by X-ray crystallography [46], and it was found that separated (Z)-phosphaethene 34 is thermally stable and does not isomerize to 33 even it is heated at 100 °C in toluene for 24 h.

Some Reactions of Phosphaalkenes
The Peterson method can be applied to prepare 3-phenyl-1-(2,4,6-tri-t-butylphenyl)-1-phosphaallene (36), as shown in Scheme 15 [47], indicating that the extended P=C system (phosphacumulenes) is also stable once sterically protected. On the other hand, when a silylphosphide is allowed to react with half of an equivalent of carbon dioxide, 1,3-bis(2,4,6-tri-t-butylphenyl)-1,3-diphosphaallene (37) is formed [48,58] via the phospha-Peterson reaction, as shown in Scheme 16 [47]. It is interesting to note that diphosphaallene 37 can be alternatively prepared from diphosphene 17 and dichlorocarbene followed by methyllithium with the Doering-Moore-Skattebøl-type reaction (Scheme 17) [20,59], probably via a bracketed carbenoid intermediate. Carbene can be generated in aqueous medium with the Makosza method [60], in the presence of benzyltriethylammonium chloride in a mixture of aqueous 50% NaOH, trichloromethane, and hexane. This one-carbon homologation method [61] can also be applied to prepare phosphaallene (36), as shown in Scheme 18 with a combination of 33 and dichlorocarbene [59]. It is interesting to note that diphosphaallene 37 can be alternatively prepared from diphosphene 17 and dichlorocarbene followed by methyllithium with the Doering-Moore-Skattebøl-type reaction (Scheme 17) [20,59], probably via a bracketed carbenoid intermediate. Carbene can be generated in aqueous medium with the Makosza method [60], in the presence of benzyltriethylammonium chloride in a mixture of aqueous 50% NaOH, trichloromethane, and hexane.

Some Reactions of Phosphaalkenes
The Peterson method can be applied to prepare 3-phenyl-1-(2,4,6-tri-t-butylphenyl)-1-phosphaallene (36), as shown in Scheme 15 [47], indicating that the extended P=C system (phosphacumulenes) is also stable once sterically protected. On the other hand, when a silylphosphide is allowed to react with half of an equivalent of carbon dioxide, 1,3-bis(2,4,6-tri-t-butylphenyl)-1,3-diphosphaallene (37) is formed [48,58] via the phospha-Peterson reaction, as shown in Scheme 16 [47]. It is interesting to note that diphosphaallene 37 can be alternatively prepared from diphosphene 17 and dichlorocarbene followed by methyllithium with the Doering-Moore-Skattebøl-type reaction (Scheme 17) [20,59], probably via a bracketed carbenoid intermediate. Carbene can be generated in aqueous medium with the Makosza method [60], in the presence of benzyltriethylammonium chloride in a mixture of aqueous 50% NaOH, trichloromethane, and hexane. This one-carbon homologation method [61] can also be applied to prepare phosphaallene (36), as shown in Scheme 18 with a combination of 33 and dichlorocarbene [59]. Phosphacumulenes 36 and 37 have axial dissymmetry according to their X-ray structures [62][63][64], so, under achiral circumstances, the reaction products are expected to be a 1:1 mixture of enantiomers, though in Schemes 15-18 for 36 and 37, only the (S)-isomers were displayed. Actually both are mixtures of enantiomers and could be separated with high-performance liquid chromatography (HPLC). Each separated enantiomer is racemized on exposure to light due to the rotation around the axis of the molecules [47,65], but they were found to be thermally stable in the dark at room temperature, even at 50 • C for 15 h for 36.
As shown in Scheme 19 (third and fourth rows), the Doering-Moore-Skattebøl method can be applied to the P=C=C and P=C=P systems (41 and 37) to yield 1-phospha-1,2,3butatriene 38 [49] and 1,4-diphospha-1,2,3-butatrienes (39 and 40, ca 4:1) [51], respectively, which implies that two carbon atoms are inserted into P=C and P=P in a stepwise manner. Additionally, Märkl described the preparation of extended cumulenes (38-40) from 2,4,6-trit-butylphenylphosphonous dichloride (15) and allenyllithium compounds as nucleophiles (Scheme 19, first and second rows) [50,52]. Phosphacumulenes 36 and 37 have axial dissymmetry according to their X-ray structures [62][63][64], so, under achiral circumstances, the reaction products are expected to be a 1:1 mixture of enantiomers, though in Schemes 15-18 for 36 and 37, only the (S)-isomers were displayed. Actually both are mixtures of enantiomers and could be separated with high-performance liquid chromatography (HPLC). Each separated enantiomer is racemized on exposure to light due to the rotation around the axis of the molecules [47,65], but they were found to be thermally stable in the dark at room temperature, even at 50 °C for 15 h for 36.

Stable Phosphaalkynes
Phosphaalkynes have the lowest coordination number of 1, and once sterically protected, this type of compound can be isolated. Using tris(trimethylsilyl)phosphane (44) and acid chlorides, Becker [55] reported the t-butyl derivative 45 in 1981, as shown in Scheme 21. With a similar method, Märkl [56] described the 2,4,6-tri-t-butylphenyl derivative 46 in 1986.  Table 3 also lists P=C-containing compounds such as 42 [53] and 43 [54], whose chemical shifts have higher values, which could be rationalized if their canonical forms are considered as depicted in Scheme 20, where the anion is located on the phosphorus atom. Phosphacumulenes 36 and 37 have axial dissymmetry according to their X-ray structures [62][63][64], so, under achiral circumstances, the reaction products are expected to be a 1:1 mixture of enantiomers, though in Schemes 15-18 for 36 and 37, only the (S)-isomers were displayed. Actually both are mixtures of enantiomers and could be separated with high-performance liquid chromatography (HPLC). Each separated enantiomer is racemized on exposure to light due to the rotation around the axis of the molecules [47,65], but they were found to be thermally stable in the dark at room temperature, even at 50 °C for 15 h for 36.
Cyclic phosphide anion 61 serves as a good nucleophile to yield various kinds of biradical species in addition to methyl derivative 62. Anion 61 reacts with several arynes or benzynes to yield phenyl-substituted biradicals [80]. In place of the t-butyllithium shown in Scheme 25, a series of carbon nucleophiles (including t-butyl iodide and even nitrogen nucleophiles such as LDA (lithium diisopropylamide)) can be used [77]. It should be noted that nucleophiles initially attack the phosphorus atom to form 61, in contrast to the corresponding nitriles.

Scheme 25. Two types of stable singlet biradicals (62 and 64).
Compound 62 is a deep blue solid with an mp of 158-160 °C and thermally stable even in a solution in air. The 31 P NMR spectrum shows an ABq at δP 55.9 (t-BuP) and -11.3 (MeP) ppm with 2 JPP 362.8 Hz, and the 13 C NMR spectrum of the ring C shows a dd at δC 111.3 (ring C) with 1 JCP 10.7 and 3.3 Hz. An X-ray study indicated no apparently direct bonding between either P-P (2.43 Å) or C-C (2.50 Å) in the ring with two almost perfectly planar carbon atoms. No signal was observed in EPR measurements, indicating that the compound is a singlet biradical species. The chemical reactivity of singlet biradicals is of high interest, including hydrolysis with water [77], oxidation with molecular oxygen [77], TEMPO (2,2,6,6-tetramethyl-1-piperidinoxy) [77], and ammoniumyl antimonate [81] Scheme 24. Structures of phosphanylidene carbenes and phosphaalkynes. Concerning the energy difference between phosphanylidenecarbene (linear 56 or bent 57) and phosphaethyne 46, we calculated total energy with an ab initio method and found that 46 has total energy of 87.3 kcal/mol less than that of 56 or 57 [70]. Additionally, in 1986, Nguyen theoretically computed 83.9 kcal/mol as the difference between 58 and 60 in a model system [74], and Lehmann [75] calculated the difference between 59 and 60 as 84.1 kcal/mol in 1985. Each computed result indicates that phosphaalkynes are far more stable than the corresponding phosphanylidenecarbene isomers (Scheme 24).
Concerning the energy difference between phosphanylidenecarbene (linear 56 or bent 57) and phosphaethyne 46, we calculated total energy with an ab initio method and found that 46 has total energy of 87.3 kcal/mol less than that of 56 or 57 [70]. Additionally, in 1986, Nguyen theoretically computed 83.9 kcal/mol as the difference between 58 and 60 in a model system [74], and Lehmann [75] calculated the difference between 59 and 60 as 84.1 kcal/mol in 1985. Each computed result indicates that phosphaalkynes are far more stable than the corresponding phosphanylidenecarbene isomers (Scheme 24).
Cyclic phosphide anion 61 serves as a good nucleophile to yield various kinds of biradical species in addition to methyl derivative 62. Anion 61 reacts with several arynes or benzynes to yield phenyl-substituted biradicals [80]. In place of the t-butyllithium shown in Scheme 25, a series of carbon nucleophiles (including t-butyl iodide and even nitrogen nucleophiles such as LDA (lithium diisopropylamide)) can be used [77]. It should be noted that nucleophiles initially attack the phosphorus atom to form 61, in contrast to the corresponding nitriles.

Scheme 25. Two types of stable singlet biradicals (62 and 64).
Compound 62 is a deep blue solid with an mp of 158-160 °C and thermally stable even in a solution in air. The 31 P NMR spectrum shows an ABq at δP 55.9 (t-BuP) and -11.3 (MeP) ppm with 2 JPP 362.8 Hz, and the 13 C NMR spectrum of the ring C shows a dd at δC 111.3 (ring C) with 1 JCP 10.7 and 3.3 Hz. An X-ray study indicated no apparently direct bonding between either P-P (2.43 Å) or C-C (2.50 Å) in the ring with two almost perfectly planar carbon atoms. No signal was observed in EPR measurements, indicating that the compound is a singlet biradical species. The chemical reactivity of singlet biradicals is of high interest, including hydrolysis with water [77], oxidation with molecular oxygen [77], TEMPO (2,2,6,6-tetramethyl-1-piperidinoxy) [77], and ammoniumyl antimonate [81] Scheme 25. Two types of stable singlet biradicals (62 and 64).
Cyclic phosphide anion 61 serves as a good nucleophile to yield various kinds of biradical species in addition to methyl derivative 62. Anion 61 reacts with several arynes or benzynes to yield phenyl-substituted biradicals [80]. In place of the t-butyllithium shown in Scheme 25, a series of carbon nucleophiles (including t-butyl iodide and even nitrogen nucleophiles such as LDA (lithium diisopropylamide)) can be used [77]. It should be noted that nucleophiles initially attack the phosphorus atom to form 61, in contrast to the corresponding nitriles.
Compound 62 is a deep blue solid with an mp of 158-160 • C and thermally stable even in a solution in air. The 31 P NMR spectrum shows an ABq at δ P 55.9 (t-BuP) and −11.3 (MeP) ppm with 2 J PP 362.8 Hz, and the 13 C NMR spectrum of the ring C shows a dd at δ C 111.3 (ring C) with 1 J CP 10.7 and 3.3 Hz. An X-ray study indicated no apparently direct bonding between either P-P (2.43 Å) or C-C (2.50 Å) in the ring with two almost perfectly planar carbon atoms. No signal was observed in EPR measurements, indicating that the compound is a singlet biradical species. The chemical reactivity of singlet biradicals is of high interest, including hydrolysis with water [77], oxidation with molecular oxygen [77], TEMPO (2,2,6,6tetramethyl-1-piperidinoxy) [77], and ammoniumyl antimonate [81] yielding a radical cation (66), sulfurization with elemental sulfur [82], reduction with hydride (67) [83,84], muonium addition [85], respectively. Oligomeric poly-biradical species 68 [86] can be prepared and pyrimidine derivatives such as 69 are promising for HF-capture reagents [87]. The formation of neutral radical 65, which is obtained via the partial oxidation of cyclic phosphide anion 61 with iodine, is also noteworthy. The radical is stable in air, and X-ray analyses, theoretical calculations, and EPR spectroscopic studies have indicated that 65 is a carbon-centered radical [88] (Scheme 26).
As shown in Scheme 13 for phosphaethene 33, (E,E)-3,4-bis(2,4,6-tri-t-butylphenylphosphinidene)-1,2-bis(trimethylsilyl)cyclobutene (78) can undergo E/Z photoisomerization and yield (E,Z)-isomer 79 (Scheme 28). A simple DPCB 80, which is derived from 78 and tetrabutylammonium fluoride (TBAF), also shows a similar photoisomerization for 81. It should be noted that in both cases, even after a longer irradiation time, the (Z,Z)-isomer is not formed, probably due to the severe congestion around the adjacent Mes* groups [93]. Bulky groups such as Mes* on phosphorus atoms appear to suffer from free rotation around the P-C bond and thus hinder the rotation that distinguishes syn-and anti-rotamers for DPCB 85, which is protected with two unsymmetrical substituents such as 2,4di-t-butyl-6-methylphenyl (Scheme 30). Two types of the tetracarbonyltungsten complexes of 85 were separated by column chromatography, and the syn-isomer could be analyzed by X-ray crystallography. The anti-isomer was further analyzed with a chiral HPLC column, and it was revealed that the isomer consists of two enantiomers, which was confirmed with the CD spectra [95]. As shown in Scheme 13 for phosphaethene 33, (E,E)-3,4-bis(2,4,6-tri-t-butylphenylphosphi nidene)-1,2-bis(trimethylsilyl)cyclobutene (78) can undergo E/Z photoisomerization and yield (E,Z)-isomer 79 (Scheme 28). A simple DPCB 80, which is derived from 78 and tetrabutylammonium fluoride (TBAF), also shows a similar photoisomerization for 81. It should be noted that in both cases, even after a longer irradiation time, the (Z,Z)-isomer is not formed, probably due to the severe congestion around the adjacent Mes* groups [93].
It is noteworthy that iodine initiates E/Z isomerization, as shown in Scheme 29. (E,E)-DPCB 82 was found to be isomerized to the corresponding (E,Z)-isomer 83 in the presence of iodine in THF at room temperature for 1 h [94]. Facile rotation between phosphorus and carbon in cyclobutenylium iodide 84 could operate as an intermediate during the process of isomerization.
Molecules 2022, 27, x FOR PEER REVIEW 13 of 21 As shown in Scheme 13 for phosphaethene 33, (E,E)-3,4-bis(2,4,6-tri-t-butylphenylphosphinidene)-1,2-bis(trimethylsilyl)cyclobutene (78) can undergo E/Z photoisomerization and yield (E,Z)-isomer 79 (Scheme 28). A simple DPCB 80, which is derived from 78 and tetrabutylammonium fluoride (TBAF), also shows a similar photoisomerization for 81. It should be noted that in both cases, even after a longer irradiation time, the (Z,Z)-isomer is not formed, probably due to the severe congestion around the adjacent Mes* groups [93]. Bulky groups such as Mes* on phosphorus atoms appear to suffer from free rotation around the P-C bond and thus hinder the rotation that distinguishes syn-and anti-rotamers for DPCB 85, which is protected with two unsymmetrical substituents such as 2,4di-t-butyl-6-methylphenyl (Scheme 30). Two types of the tetracarbonyltungsten complexes of 85 were separated by column chromatography, and the syn-isomer could be analyzed by X-ray crystallography. The anti-isomer was further analyzed with a chiral HPLC column, and it was revealed that the isomer consists of two enantiomers, which was confirmed with the CD spectra [95]. Bulky groups such as Mes* on phosphorus atoms appear to suffer from free rotation around the P-C bond and thus hinder the rotation that distinguishes synand anti-rotamers for DPCB 85, which is protected with two unsymmetrical substituents such as 2,4-di-tbutyl-6-methylphenyl (Scheme 30). Two types of the tetracarbonyltungsten complexes of 85 were separated by column chromatography, and the syn-isomer could be analyzed by X-ray crystallography. The anti-isomer was further analyzed with a chiral HPLC column, and it was revealed that the isomer consists of two enantiomers, which was confirmed with the CD spectra [95].
The reaction of 1,2-diphenyl-3,4-bis(2,4,6-triisopropylphenylphosphinidene)cyclobutene (95) with pentacarbonyltungsten leads to an interesting result, as depicted in Scheme 33: it not only serves as a bidentate ligand to yield 97 but also it yields 96 as a monodentate ligand. More interestingly, the ligand serves as a side-on ligand to yield 98 [104].
The reaction of 1,2-diphenyl-3,4-bis(2,4,6-triisopropylphenylphosphinidene)cyclobutene (95) with pentacarbonyltungsten leads to an interesting result, as depicted in Scheme 33: it not only serves as a bidentate ligand to yield 97 but also it yields 96 as a monodentate ligand. More interestingly, the ligand serves as a side-on ligand to yield 98 [104].
A stable (π-allyl)palladium complex such as 94 (shown in Scheme 32) [103] has shown a wide range of catalytic reactivity including in the Tsuji-Trost-type reaction that enables allylic alcohols to yield amines via the elimination of water [36,105,106] under mild conditions such as room temperature for 1 h with a 0.1 mol% DPCB-palladium catalyst.
Direct alkylation can be regioselectively achieved using carbanions from active methylene compounds such as ethyl acetylacetate (105) as a nucleophile in order to formally substitute the OH group of allylic alcohol 104 with an alkylation product (106) in a high yield, as shown in Scheme 37 [112], where an intermediate 107 is supposed to be functional. It is not necessary to convert alcohol 104 to a more reactive derivative (such as an ester or halide) as an allylic starting material in advance of conventional Tsuji-Trost reactions.
Scheme 37. DPCB-palladium-catalyzed Tsuji-Trost-type alkylation reaction; pd = (DPCB)Pd + . Furthermore, the reaction from a chiral alcohol 108 of 98.5 %ee yields corresponding amine 109 of the same chirality in a 92% yield, which indicates that a double inversion process is involved and surrounds intermediate 110 during the reaction course to control the stereochemistry (Scheme 38) [112].
These results indicate that DPCB ligands have a high coordination ability due to their σ-donation/π-back donation interactions, strong π-acceptor property due to their low-lying LUMO (π*), and unique coordination structure with planarity. In addition, the catalysts are highly stable toward air and reusable after reactions in most cases. The above characteristics are superior to catalysts with ligands having sp 3 phosphorus or sp 2 nitrogen atoms for catalytic organic reactions. Direct alkylation can be regioselectively achieved using carbanions from active methylene compounds such as ethyl acetylacetate (105) as a nucleophile in order to formally substitute the OH group of allylic alcohol 104 with an alkylation product (106) in a high yield, as shown in Scheme 37 [112], where an intermediate 107 is supposed to be functional. It is not necessary to convert alcohol 104 to a more reactive derivative (such as an ester or halide) as an allylic starting material in advance of conventional Tsuji-Trost reactions.
A stable (π-allyl)palladium complex such as 94 (shown in Scheme 32) [103] has shown a wide range of catalytic reactivity including in the Tsuji-Trost-type reaction that enables allylic alcohols to yield amines via the elimination of water [36,105,106] under mild conditions such as room temperature for 1 h with a 0.1 mol% DPCB-palladium catalyst.
Direct alkylation can be regioselectively achieved using carbanions from active methylene compounds such as ethyl acetylacetate (105) as a nucleophile in order to formally substitute the OH group of allylic alcohol 104 with an alkylation product (106) in a high yield, as shown in Scheme 37 [112], where an intermediate 107 is supposed to be functional. It is not necessary to convert alcohol 104 to a more reactive derivative (such as an ester or halide) as an allylic starting material in advance of conventional Tsuji-Trost reactions.
Scheme 37. DPCB-palladium-catalyzed Tsuji-Trost-type alkylation reaction; pd = (DPCB)Pd + . Furthermore, the reaction from a chiral alcohol 108 of 98.5 %ee yields corresponding amine 109 of the same chirality in a 92% yield, which indicates that a double inversion process is involved and surrounds intermediate 110 during the reaction course to control the stereochemistry (Scheme 38) [112].
These results indicate that DPCB ligands have a high coordination ability due to their σ-donation/π-back donation interactions, strong π-acceptor property due to their low-lying LUMO (π*), and unique coordination structure with planarity. In addition, the catalysts are highly stable toward air and reusable after reactions in most cases. The above characteristics are superior to catalysts with ligands having sp 3 phosphorus or sp 2 nitrogen atoms for catalytic organic reactions. Furthermore, the reaction from a chiral alcohol 108 of 98.5 %ee yields corresponding amine 109 of the same chirality in a 92% yield, which indicates that a double inversion process is involved and surrounds intermediate 110 during the reaction course to control the stereochemistry (Scheme 38) [112].
A stable (π-allyl)palladium complex such as 94 (shown in Scheme 32) [103] has shown a wide range of catalytic reactivity including in the Tsuji-Trost-type reaction that enables allylic alcohols to yield amines via the elimination of water [36,105,106] under mild conditions such as room temperature for 1 h with a 0.1 mol% DPCB-palladium catalyst.
Direct alkylation can be regioselectively achieved using carbanions from active methylene compounds such as ethyl acetylacetate (105) as a nucleophile in order to formally substitute the OH group of allylic alcohol 104 with an alkylation product (106) in a high yield, as shown in Scheme 37 [112], where an intermediate 107 is supposed to be functional. It is not necessary to convert alcohol 104 to a more reactive derivative (such as an ester or halide) as an allylic starting material in advance of conventional Tsuji-Trost reactions.
Scheme 37. DPCB-palladium-catalyzed Tsuji-Trost-type alkylation reaction; pd = (DPCB)Pd + . Furthermore, the reaction from a chiral alcohol 108 of 98.5 %ee yields corresponding amine 109 of the same chirality in a 92% yield, which indicates that a double inversion process is involved and surrounds intermediate 110 during the reaction course to control the stereochemistry (Scheme 38) [112].
These results indicate that DPCB ligands have a high coordination ability due to their σ-donation/π-back donation interactions, strong π-acceptor property due to their low-lying LUMO (π*), and unique coordination structure with planarity. In addition, the catalysts are highly stable toward air and reusable after reactions in most cases. The above characteristics are superior to catalysts with ligands having sp 3 phosphorus or sp 2 nitrogen atoms for catalytic organic reactions. These results indicate that DPCB ligands have a high coordination ability due to their σ-donation/π-back donation interactions, strong π-acceptor property due to their low-lying LUMO (π*), and unique coordination structure with planarity. In addition, the catalysts are highly stable toward air and reusable after reactions in most cases. The above characteristics are superior to catalysts with ligands having sp 3 phosphorus or sp 2 nitrogen atoms for catalytic organic reactions.
Recently, Ozawa reported using pincer ligands such as 111-113 (Scheme 39) to activate CO bonds in carbon dioxide, taking advantage of the frustrated lone pair (FLP) within the ligand [113,114]. Further fundamental and applied investigation are expected to lead to progress in research of this type of low-coordinate phosphorus ligand for ideally efficient catalytic reactions.
Recently, Ozawa reported using pincer ligands such as 111-113 (Scheme 39) to activate CO bonds in carbon dioxide, taking advantage of the frustrated lone pair (FLP) within the ligand [113,114]. Further fundamental and applied investigation are expected to lead to progress in research of this type of low-coordinate phosphorus ligand for ideally efficient catalytic reactions. Scheme 39. Some pincer ligands with sterically bulky P=C bonds (111)(112)(113).

Closing Remarks
Forty years have passed since a true "phosphobenzene" was prepared for the first time in 1981 via the steric protection methodology with an extremely bulky 2,4,6-tri-tbutylphenyl group (Mes*). It was believed, even theoretically, that molecules such as those with heavier main-group elements never existed as stable compounds, but various kinds of "unusual" phosphorus compounds have been isolated and characterized by utilizing the Mes* group. Interest in basic research on this class of compounds and synthetic applications for organic synthesis is high. This review article was based on a plenary lecture delivered by the author at the 23rd International Conference on Phosphorus Chemistry (ICPC-23) remotely held in Częstochowa, Poland, 5-9 July 2021 [115].
Funding: This research received no external funding.