Nahil Al-Zuhaikaa,
Simon Severinab,
Madeleine D. Schmucklera,
Alan J. Lougha and
Douglas W. Stephan*a
aDepartment of Chemistry, University of Toronto, 80 St. George St, Toronto, ON M5S3H6, Canada. E-mail: dstephan@chem.utoronto.ca
bInstitute of Inorganic and Analytical Chemistry, University of Münster, Muenster, Germany
First published on 18th August 2025
The diphenylphosphirenium salts of the form [Ph2PC(R)C(H)][AlCl4] are readily generated and select examples react with secondary phosphines, R2PH, to give dissymmetric bidentate phosphonium salts, [Ph2PC(R)C(H)P(R′)2H][AlCl4]. While these reactions work well for sterically encumbered combinations of the phosphirenium cations and secondary phosphines (R = tBu, Cy, Mes), less encumbered combinations provide a mixture of products arising from alkyne displacement. As expected the protonated bis-phosphine salts are easily deprotonated, demonstrating easy access to a rare class of dissymmetric bidentate phosphine ligands.
In a recent effort targeting new routes to dissymmetric bidentate phosphines, we21,22 recognized that the phosphino-phosphenium cation (PPC) originally described by Burford and coworkers23–27 and others28,29 can act as frustrated Lewis pairs (FLPs) and provide access to both phosphorus Lewis acidic and basic sites. As such, they undergo addition to alkynes to afford cis-substituted olefins to give cations of the form [cis-R3PCHC(R′′)PR′2]+.21 Suitable modification of the precursors afforded [cis-R2ClPCHC(R′′)PR′2]+ which were readily reduced by nBu3P to give the dissymmetric bidentate phosphine in good yields (Scheme 1).21
While this new protocol provides access to a rare set of bidentate phosphines, we sought to further improve the generality and facility of the synthetic protocol. We recognized that one of the limitations of the above method is the need for reduction of the intermediate phosphorus(V) halide salt (Scheme 1). We envisioned deprotonation as a simpler, cheaper process. Moreover, this avenue offers increased variability as a broad selection of secondary phosphines are commercially available. However, we also recognized that the reactivity of PPCs prepared using a secondary phosphine as the donor are readily deprotonated affording the neutral diphosphine R2PPR′2. Our previous mechanistic study revealed that phosphino-phosphination was initiated by alkyne displacement of the phosphine donor in the PPC generating a three membered phosphirenium cationic ring,21 which undergoes subsequent ring opening by the chlorophosphines. Herein, we explore the reactivity of phosphirenium cations targeting ring-opening by secondary phosphines as a route to protonated dissymmetric bidentate phosphine cations, species which are readily deprotonated to give olefin-linked dissymmetric bidentate phosphines.
To initiate this study, the phosphirenium salt of the form [R2P(R′CCH)][AlCl4] were prepared according to literature procedures30–34 via a 1:
1
:
1 reaction of the chlorodiphenylphosphine, terminal alkyne and AlCl3 (Table 1). Initially using tBuCCH, the species [Ph2PC(tBu)C(H)][AlCl4] 1 was generated as evidenced by the 31P NMR signal at 104.6 ppm, characteristic of the phosphirenium cation.
Cmpd | R | T (h) | Solvent | 31P shift |
---|---|---|---|---|
1 | tBu | 0.3 | CH2Cl2 | −104.6 |
2 | Me3Si | 1 | CH2Cl2 | −122.0 |
3 | Ph | 94 | o-DFB | −105.5 |
4 | p-tol | 0.3 | CH2Cl2 | −107.3 |
5 | p-(HCC)C6H4 | 24 | o-DFB | −105.9 |
6 | p-IC6H4 | 48 | CH2Cl2 | −106.2 |
7 | p-BrC6H4 | 0.3 | CH2Cl2 | −105.8 |
8 | C4H9 | 95 | o-DFB | −106.9 |
9 | c-C3H5 | 0.5 | CH2Cl2 | −105.8 |
10 | c-C5H9 | 2 | C6H5Me | −107.2 |
11 | Cy | 1 | CH2Cl2 | −107.2 |
In a similar fashion, this protocol was used to generate a series of related phosphirenium cations [Ph2PC(R)C(H)][AlCl4] (R = SiMe3 2, Ph 3, p-tol 4, p-(HCC)C6H4 5, p-IC6H4 6, p-BrC6H4 7, C4H9 8, C3H5 9, C5H9 10, C6H11 11). While these species were not isolated all reactions showed a clean singlet in the 31P NMR spectrum in the range of 104–122 ppm, typical of phosphirenium cations.
As these phosphirenium cations proved to be readily generated, we next assessed the ability of a secondary phosphine to effect ring opening of these phosphacyclic cations. To this end, we first generated a CH2Cl2 solution of 1 and added an equivalent of tBu2PH. Stirring the mixture 20 minutes followed by subsequent workup afforded a reddish-brown sticky solid 12 in 85% yield. This species shows a 31P{1H} NMR spectrum consisting of two doublets at 28.6 and −1.1 ppm with a coupling constant of 86 Hz. The corresponding 31P NMR spectrum revealed that the downfield signal also showed P–H coupling of 474 Hz consistent with a one-bond P–H coupling constant. Thus, the upfield resonance was attributed to a PPh2-fragment, while the downfield signal was assigned to the protonated tBu2PH-fragment. These data together with 1H NMR and MS data supported the formulation of 12 as [Ph2PC(tBu)C(H)(HPtBu2)][AlCl4] (Scheme 2). On standing slow vaporization of CH2Cl2 from the reaction mixture provided crystals of 12 suitable for an XRD study. The structural data confirmed the formulation of the salt 12 (Fig. 1) in which cation and anion are well separated. In the cation, the phosphorus centers adopt a cis-disposition with P–C bond distances to the linking olefinic fragment found to be 1.847(6) Å and 1.793(6) Å. The shorter of the two is associated with the cationic tBu2PH fragment. Despite the phosphonium center, the tBu2PH fragment affords the shorter P–C bond, consistent with the presence of the electron donating alkyl substituents and the less hindered carbon center. The olefin linker showed the typical C–C distance of 1.336(9) Å.
![]() | ||
Fig. 1 POV-ray depiction of the crystallographic structure of the cation of 12. Hydrogen atoms except for the PH unit are omitted for clarity. C: black, P: orange, H: white. |
In a similar fashion following this protocol, reaction of 1 with Cy2PH, afforded a sticky solid in 73% yield. This product 13 showed 31P{1H} NMR resonances at 9.5 and −2.8 ppm with a P–P coupling constant of 57 Hz. The corresponding 1H coupled 31P NMR data revealed the downfield signal was also coupled to proton with a coupling constant of 479 Hz. These data as well as mass spectral data affirmed the formulation of 13 as [Ph2PC(tBu)C(H)(HPCy2)][AlCl4] (Scheme 2).
In the case of the analog using the phosphirenium salt 2, reaction with tBu2PH was performed on a 2 g scale following a similar procedure. The ultimate yield of the [Ph2PC(Me3Si)C(H)(HPtBu2)][AlCl4] 14 was 91% yield (Scheme 2). This species showed the two expected resonances in the 31P{1H} NMR spectrum at 20.9 and −3.39 ppm with a 3JPP of 81 Hz, while the 31P NMR spectrum revealed a 1JPH of 466 Hz for the upfield resonance, again consistent with the formation of 14.
We next examined the reactions of related aryl-substituted phosphirenium cations. Thus, 4 was independently reacted with tBu2PH or Cy2PH. These reactions proceeded similar to those above affording reddish-brown and yellowish-orange products 15 and 16. These products were further purified by reprecipitation from C6H5Cl and isolated in 89% and 62% yields respectively. Compound 15 exhibited 31P resonances at 26.6 and −4.5 ppm with 3JPP of 74 Hz and the downfield signal exhibiting a 1JPH of 461 Hz. The corresponding data for 16 showed 31P signals at 9.6 and −2.6 ppm with 3JPP of 70 Hz with the downfield signals also exhibiting 475 Hz, respectively. These data again affirmed the formulation of these products as [Ph2PC(p-tol)C(H)(HPR2)][AlCl4] (R = tBu 15, Cy 16) (Scheme 2).
Switching the phosphirenium cation to 7, the corresponding derivative [Ph2PC(p-BrC6H4)C(H)(HPtBu2)][AlCl4] 17 was formed in 76% yield (Scheme 2) as evidenced by the 31P resonances at 27.0 and −4.0 ppm with 3JPP and 1JPH of 77 and 450 Hz, respectively. While this formulation was also supported by 1H and MS data (see SI) it was also supported by a crystallographic study. This unambiguously confirmed the formation of 17 where two phosphorus atoms on an olefinic linker are cis-disposed (Fig. 2). Further the PPh2 fragment resides on the substituted carbon of the olefinic link, and the PtBu2 fragment is protonated. The resulting P–C bond distances to the linking olefin were found to be 1.862(2) Å and 1.780(2) Å, with the shorter P–C bond arising from tBu2P-fragment. The remainder of the metric parameters are unexceptional.
![]() | ||
Fig. 2 POV-ray depiction of the crystallographic structure of 17. Hydrogen atoms except for the PH unit are omitted for clarity. C: black, P: orange, Cl: green, Al: white, Br: red. |
In a similar fashion reaction of 7 with Mes2PH gave an orange-yellow solid in 91% yield. Two 31P resonances at −2.3 and −31.6 ppm showed a 3JPP of 100 Hz and the upfield signals showed a 1JPH of 497 Hz, consistent with the formulation of 18 as [Ph2PC(p-BrC6H4)C(H)(HPMes2)][AlCl4] (Scheme 2) which was also supported by 1H and MS data (see SI).
Probing the generality of this reactivity further, the corresponding reaction of 7 with Ph(tBu)PH was found to give a complex mixture of products including unreacted phosphirenium cation, the expected product [Ph2PC(p-BrC6H4)C(H)(HP(tBu)(Ph)][AlCl4], [Ph2P(PHPh(tBu))]+, [Ph(tBu)P(PHPh2)]+ and [H2P(tBu)(Ph)]+ (Scheme 3) as evidenced by the 31P NMR signals. The complexity of this mixture precluded the isolation of individual products, but did provide some mechanistic insights.
![]() | ||
Scheme 3 Divergent reaction pathways for less hindered combinations of phosphirenium cations and secondary phosphines. |
The product mixture is consistent with two reaction pathways. Reaction of the phosphirenium cation with Ph(tBu)PH can proceed via ring-opening to protonated bisphosphine salt, alternatively alkyne displacement affords phosphino-phosphenium cations which in the presence of secondary phosphine also generates [H2P(tBu)(Ph)]+. We note that no evidence of this latter pathway was seen in the previous examples where either bulky substituents were incorporated into the secondary phosphine or the alkyne. These observations can be rationalized on the basis of steric demands associated with the nucleophilic attack of the phosphirenium cation as well as the basicity of the alkyne used to generate the phosphirenium cation. While sterically demanding secondary phosphines lead to attack at the unsubstituted carbon of the phosphirenium cation, sterically less encumbered phosphines can access the σ*-orbital on phosphorus, prompting alkyne displacement and the formation of phosphino-phosphenium cations.
To further probe this notion, the phosphirenium salt 11 was reacted with tBu2PH affording the dissymmetrically substituted [Ph2PC(Cy)C(H)(HPtBu2)][AlCl4] 19 (Scheme 2) as evidenced by the 31P NMR signals at 28.5 and −6.5 ppm with a P–P coupling constant of 93 Hz and the P–H coupling of the downfield signal of 474 Hz. In contrast, reaction of 11 with the less sterically encumbered Cy2PH generated a mixture of products including [Ph2PPCy2H]+, [Ph2PHPCy2]+ and [Cy2PH2]+ (see SI) as evidenced by 31P NMR spectroscopy (Scheme 3).
Collectively, these data affirm that less hindered combinations of phosphirenium cations and secondary phosphines react via divergent pathways to afford a mixture of products. On the other hand, more sterically demanding systems are shown to readily afford protonated, dissymmetrically substituted, olefin-linked phosphino-phosphonium cations 12–19. To confirm that such cations can be deprotonated in a trivial fashion, compound 18 was treated with an equivalent of 1,8-bis(dimethylamino)-naphthalene (proton sponge) for 30 minutes. Following solvent removal under vacuum, the product 20 (Scheme 2) was obtained in a 90% yield. The 31P NMR spectrum of 20 shows resonances at −7.6 and −35.5 ppm with a coupling constant of 148 Hz consistent with the formulation of 20 as the bidentate phosphine, Ph2PC(p-BrC6H4)C(H)(PMes2). This formulation was also confirmed via a crystallographic study (Fig. 3). The cis-bisphosphine and dissymmetric nature of 20 revealed P–C distances to the linking olefinic fragment of 1.841(4) Å and 1.824(4) Å for the PMes2 and PPh2 fragments respectively, with a CC bond length of 1.349(6) Å.
![]() | ||
Fig. 3 POV-ray depiction of the crystallographic structure of 20. Hydrogen atoms are omitted for clarity. C: black, P: orange, Br: red. |
Herein, we demonstrated that the salts [Ph2PC(R)CH][AlCl4] are readily generated using a variety of terminal alkynes. The reactions of such species with secondary phosphines showed that less encumbered combinations give rise to mixtures of ring-opened and alkyne displacement products, while bulkier combinations afford exclusively the ring-opening products, [Ph2PC(R)C(H)PR′2H][AlCl4]. Facile deprotonation affording a dissymmetric bidentate phosphine ligand was demonstrated confirming that this new synthesis route provides ready access to a class of compounds that are generally rare and otherwise challenging to prepare. We are continuing to examine the reactions of phosphirenium cations as well as new strategies to desirable phosphorus-based compounds.
NSERC of Canada and Syensqo are thanked for financial support. Drs E. Conrad, J. Zheng, and Z. Yan of Syensqo are thanked for helpful discussions.
NA, SS and MDS performed the experimental work, AL performed one of the X-ray structural studies while DWS wrote the manuscript. All authors edited the draft.
CCDC 2444558 (17), 2444559 (12) and 2444560 (20) contain the supplementary crystallographic data for this paper.35–37
This journal is © The Royal Society of Chemistry 2025 |