Three-component diazaphospholium synthesis enables P(V)-to-P(III) conversion and ylide-mediated CO₂ and CS₂ activation

Abstract

Diazaphospholium triflate salts 3[OTf]₂ were synthesized through a three-component (1 + 2 + 2)-cycloaddition of the imidazoliumyl-substituted phosphenium surrogate [L_C-P(OTf)Ph]⁺ cation (L_C = 4,5-dimethyl-1,3-diisopropylimidazolium) with nitriles and N-benzylideneaniline. These P(V) heterocycles display pronounced base-dependent reactivity. Treatment with 4-dimethylaminopyridine (DMAP) triggers a 1,2-phenyl migration and a formal P(V)-to-P(III) conversion, affording cationic DMAP adducts that can be further transformed into azaphospholes bearing P–O, P–C, and P–N bonds. Key intermediates and products were isolated and fully characterized, and the proposed mechanism is supported by DFT calculations. In contrast, reaction of diazaphospholium triflate 3[OTf]₂ with 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), followed by deprotonation, furnishes an ylide featuring a directional, strong N⋯P interaction. This species exhibits not only classical Wittig-type reactivity but also unusual CO₂ and CS₂ activation, leading to oxazinedione- or thiazinedithione-type products alongside the corresponding phosphole chalcogenides.

---

Introduction

Phosphorus-containing heterocycles have garnered significant attention owing to their widespread applications in pharmaceutical agents,^1,2 materials,^3,4 and organic synthesis.^5,6 Among the reported synthetic methodologies for these compounds, condensation and cycloaddition reactions are the two most commonly employed protocols. The former typically employs simple or pre-designed phosphorus building blocks to construct more complex molecules and requires careful consideration of functional group tolerance during the process. In contrast, cycloaddition reactions form cyclic products through efficient processes and have therefore attracted considerable attention.⁷ A representative example is the (3 + 2)-cycloaddition reaction of phosphaalkynes with azides, which provides efficient access to 1,2,3,4-triazaphospholes,^8–11 albeit being constrained to highly reactive phosphaalkynes.

Phosphenium ions [R₂P]⁺ are dicoordinate phosphorus species featuring six-valence electrons, rendering them isoelectronic with neutral carbenes. Considerable efforts have been devoted to stabilizing these highly reactive species through electronic and steric effects imposed by substituents (R = e.g. amino, alkyl, aryl, cyclopentadienyl, ferrocenyl).^12–37 Owing to their high electrophilicity, they have long served as versatile synthons for constructing phosphorus-containing frameworks. They readily engage in insertion and addition reactions (e.g. (1,4)- and (2,4)-additions) with unsaturated substrates to form diverse phosphorus-containing compounds.^12,17,38,39

In 2002, Bertrand and co-workers reported a (3 + 2)-dipolar cycloaddition of phosphenium cation I with cyanamides to give adducts of type II, which react with a second equivalent of cyanamide, affording cationic five-membered heterocycles III (Fig. 1a).⁴⁰

Fig. 1 (a) Dipolar (3 + 2) cycloaddition of a phosphenium–cyanamide acid–base adduct II with a second equivalent of cyanamide, affording a cationic five-membered heterocycle III; R = Me, ⁱPr. (b) Two-component (3 + 2)-cycloaddition of an imidazoliumyl-substituted triflatophosphane IV with alkynes and nitriles to access azaphospholium salts V and VI; R₁ = H, R₂ = aromatic groups, or R₁ = R₂ = aromatic groups; R₃ = alkyl or aromatic groups. (c) This work: three-component (1 + 2 + 2)-cycloaddition of 1[OTf] with nitriles and a N-benzylideneaniline to access diazaphospholium salts 3[OTf]₂; R = aromatic groups. Stereocenters are marked by asterisk [*].

Recently, our group reported the imidazoliumyl-substituted phosphenium surrogate IV, which activates dipolarophiles such as alkynes and nitriles to access a range of azaphospholium salts V and VI via a (3 + 2)-cycloaddition route (Fig. 1b).^41,42 Inspired by these results, we envisioned that the phosphenium surrogate 1[OTf] could promote a (1 + 2 + 2)-cycloaddition, thus granting access to azaphospholium architectures (Fig. 1c). Moreover, the imidazoliumyl group should enable post-modification via L_C–P bond activation, e.g. through processes in which the L_C-substituent acts as a leaving group liberating the free NHC upon reduction or substitution reactions with a Grignard reagent.⁴³

Despite their rich chemistry in these two-component reactions, the utility of phosphenium cations in multicomponent cycloaddition reactions remains largely unexplored. Herein, we report a three-component reaction of the imidazoliumyl-substituted phosphane 1[OTf] with nitriles and an N-benzylideneaniline to afford a series of five-membered diazaphospholium triflate salts 3[OTf]₂. The L_C–P bond in these products can be selectively engaged by the choice of bases, resulting in divergent downstream reactivity. Treatment of 3a–c[OTf]₂ with 4-dimethylaminopyridine (DMAP) promotes either 1,2-phenyl migration or ring opening, leading in both cases to a rare formal P(V)-to-P(III) conversion, as supported by DFT calculations and the full characterization of intermediates and products. Further functionalization of the resulting adduct 5a[OTf] provides access to a range of azaphospholes. In contrast, treatment of 3a[OTf]₂ with 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) affords a phosphorus ylide that exhibits Wittig-type reactivity. This ylide further reacts with aldehydes, CO₂, and CS₂, most likely via a distinctive bimolecular pathway, to give a series of DBU-derived products.

Results and discussion

Synthesis

The triflato-phosphane 1[OTf] was prepared analogously to literature procedures via chloride abstraction from the chlorophosphane precursor using AgOTf.^12,44 It displays a downfield-shifted resonance in the ³¹P NMR spectrum [δ(³¹P) = 113.5 ppm], within the range reported for structurally characterized triflato-phosphanes.^41,42 To probe the electrophilic behavior of the P atom in 1[OTf], we screened its reaction with various nitriles. Treating 1[OTf] with benzonitrile or 4-methoxybenzonitrile at 80 °C for 16 hours resulted in no observable reaction. In contrast, use of the more nucleophilic 4-(dimethylamino)benzonitrile afforded the cycloaddition product 2[OTf]₂ in near quantitative yield (96%). Suitable crystals for single-crystal X-ray diffraction (scXRD) analysis confirmed the molecular structure (Fig. 2a). This result is reminiscent of the previously reported trapping of electrophilic phosphorus cations by strongly donating nitriles to furnish cationic five-membered heterocycles. Although no discrete 1⁺–nitrile adduct was isolated, the formation of 2[OTf]₂ and its structural metrics are consistent with initial nitrile coordination to cation 1⁺ as the entry point to cycloaddition.

Fig. 2 Molecular structures of 2²⁺ in 2[OTf]₂ (a), anti-3a²⁺ in anti-3a[OTf]₂ (b), 5a⁺ in 5a[OTf]·0.5 CH₂Cl₂ (c), 6²⁺ in 6[OTf]₂·0.666 CH₂Cl₂ (d) and anti-4⁺ in anti-4[OTf] (e); hydrogen atoms, anions, and solvent molecules are omitted for clarity; thermal ellipsoids are displayed at 50% probability (100 K); selected bond lengths (in Å) and angles (in °): 2[OTf]₂, P1–C1 1.819(3), P1–C2 1.793(3), P1–N1 1.615(3); anti-3a[OTf]₂, P1–C1 1.8160(14), P1–C2 1.7682(15), P1–N1 1.6312(12); 5a[OTf], P1–C1 1.9338(12), P1–N1 1.6862(10), P1–N2 1.7840(9); 6[OTf]₂, P1–C1 1.835(4), P1–C2 1.830(4), P1–N1 1.702(3); anti-4[OTf], P1–C1 1.942(3), P1–C2 1.831(3), C1–N2 1.521(3), P1–N1 1.711(2).

Motivated by this, we investigated a strategy to engage less nucleophilic nitriles by introducing a second, complementary two-atom component. N-Benzylideneaniline was found to be a suitable substrate for this three-component conversion and the synthetic route is outlined in Scheme 1. Treatment of stoichiometric amounts of benzonitrile, N-benzylideneaniline, and 1[OTf] in DCM at 80 °C for 12 h in a pressure-sealed tube afforded the desired product 3a[OTf]₂. The product was recrystallized from CH₂Cl₂/Et₂O and isolated in 84% yield on a 10 gram scale. The ³¹P NMR spectrum of 3a[OTf]₂ revealed two resonances at δ(³¹P) = 55.0 ppm [syn-(S_P, R_C), 93%] and δ(³¹P) = 60.2 ppm [anti-(S_P, S_C), 7%], corresponding to two diastereomeric configurations and in agreement with a larger characteristic ²J_HP coupling constant of 11 Hz observed in the ¹H NMR spectrum for syn-3a[OTf]₂.⁴⁵ These assignments were unambiguously confirmed by scXRD analysis of triflate salts of both diastereomers (see the SI, Fig. S106). The two diastereomers of cations 3a–c²⁺ arise from the stereogenic carbon atom in the N-benzylideneaniline and the phosphorus atom upon cycloaddition.⁴⁶

Scheme 1 Synthesis of diazaphospholium salts 2[OTf]₂ and 3a–c[OTf]₂. (i) 2 equiv. 4-(dimethylamino)benzonitrile, DCM, r.t., 30 min, 96%. (ii) 1 equiv. 1[OTf], 1 equiv. benzonitrile derivative, and 1 equiv. N-benzylideneaniline, DCM, 80 °C (pressure sealed tube), 12 h, 84% (for 3a[OTf]₂), 78% (for 3b[OTf]₂) and 73% (for 3c[OTf]₂). Stereocenters are marked by asterisk [*].

This reaction tolerates both electron-donating and electron-withdrawing substituents, such as 4-methoxy- and 4-nitrobenzonitrile, affording 3b,c[OTf]₂ in good yields (78 and 73%). The molecular structures of 3a,b[OTf]₂ were verified by scXRD analysis (see anti-3a[OTf]₂, Fig. 2b and anti-3b[OTf]₂ in the SI, Fig. S21). The observed P–C bond lengths in 3a²⁺ [P1–C1 1.8160(14) Å, P1–C2 1.7682(15) Å] fall within the typical single bond range for azaphospholium salts. The P1–N1 single bonds in azaphospholium cations 3a,b²⁺ [3a²⁺, 1.6312(12) Å; 3b²⁺, 1.625(2) Å] are comparable to that of 2²⁺ [1.615(3) Å], further indicating that the cycloaddition involves interaction of the electrophilic phosphorus with the cyano group.

Phosphonium salts containing an α-C–H bond readily undergo deprotonation and have been exploited in Wittig reactions.^47–49 Given their structural similarity to such phosphonium salts, these azaphospholium salts are anticipated to exhibit related reactivity. At the same time, the L_C–P bond is base-sensitive, which potentially enables its functionalization. Accordingly, the deprotonation of these cyclic azaphospholium salts 3a–c[OTf]₂ was evaluated. However, treatment with ionic bases, such as potassium bis(trimethylsilyl)amide (KHMDS) or potassium tert-butoxide (KO^tBu) led to unselective conversions in all cases. In contrast, reactions with neutral N-bases such as DMAP or DBU yielded well-defined and isolable species.

Reactions with nitrogen base DMAP

Addition of DMAP to a suspension of 3a[OTf]₂ in C₆H₅F or THF gave a yellow solution within 2 hours at room temperature. ³¹P NMR spectroscopic analysis of aliquots removed from the reaction mixture revealed two distinct resonances at δ(³¹P) = 98.0 and 112.3 ppm, and the ¹H NMR spectrum confirmed the formation of [Lc–H][OTf] (see the SI, section S4.1).

Repeated ³¹P NMR spectroscopic analysis after another 16 hours at room temperature revealed the complete consumption of both resonances in favor of a new resonance at δ(³¹P) = 141.4 ppm, which was unambiguously identified as 5a[OTf] by scXRD (Fig. 2c). The P–N bond length in 5a⁺ (P1–N2 1.7840(9) Å) is consistent with that of the known N–P(III) acid–base adduct (e.g. DMAP–PPh₂: 1.789(1) Å) (Scheme 2).⁵⁰

Scheme 2 Reactions of 3a–c[OTf]₂ with DMAP. (i) +DMAP, −[L_C–H][OTf], C₆H₅F or THF, r.t., 2 h, 89% (for 4[OTf]). (ii) Rearrangement to 5a,b[OTf] (r.t., 12 h), 95% (for 5a[OTf]), 59% (for 5b[OTf]) or to 5c[OTf] (r.t., 7 days; or dissolution in DCM, overnight), 89%. (iii) Competing C–P bond cleavage observed predominantly for 3b[OTf]₂, affording 6[OTf]₂, 38%. Stereocenters are marked by asterisk [*].

In an effort to rationalize this reactivity and identify the observed intermediates, conversions of 3b[OTf]₂ and 3c[OTf]₂ with DMAP were conducted under identical conditions. The reaction of 3b[OTf]₂ afforded two main products that were isolated, fully characterized, and identified as 5b[OTf] (see the SI, Fig. S40) and 6[OTf]₂ (Fig. 2d) by scXRD, respectively. The product 6[OTf]₂ is consistent with C–P bond scission following nucleophilic addition of DMAP; its connectivity was established by multinuclear NMR spectroscopy. In contrast, treatment of 3c[OTf]₂ led to a brown suspension after 3 days, and ³¹P NMR analysis of the mixture by dissolution in DCM revealed resonances at δ(³¹P) = 100.6 and 110.8 ppm, consistent with those observed for 3a[OTf]₂ under analogous conditions. Recrystallization of the crude solid from DCM/Et₂O allowed identification of diastereomeric 4[OTf] (Fig. 2e), featuring two stereogenic centers. Prolonged reaction times (>1 week) or dissolution of 4[OTf] in DCM overnight resulted in exclusive formation of 5c[OTf]. Taken together, these results are consistent with DMAP attack at the internal α-carbon to give a pyridinium-type intermediate (4⁺). From this branch point, two competing pathways are operative: (i) departure of the L_C group to either furnish isolable 4[OTf] or trigger 1,2-phenyl migration to give 5[OTf], or (ii) C–P bond scission to give the ring-opened product 6[OTf]₂ (predominantly for R = OMe). Notably, the nitro-substituted intermediate 4[OTf] is sufficiently stabilized to be isolated.

To corroborate our experimental findings, we performed DFT calculations at the BP86-D4/def2-TZVP level of theory (COSMO = C₆H₅F) and propose the mechanism outlined in Fig. 3, which involves four transition states (TS) and three intermediates (INT) that are consistent with experimental observations of transient species by in situ NMR spectroscopy.

Fig. 3 Reaction profile (ΔG° in kcal mol⁻¹) for the transformation process, at the BP86-D4/def2-TZVP level of theory (COSMO = C₆H₅F). The reactants were simplified as 3a′–c′. Specifically, the phenyl group attached to the nitrogen atom of N-benzylideneaniline was replaced with a methyl group, as this group is not involved in the mechanistic steps. DMAP was replaced by 4-aminopyridine (AP), and the L_C substituent was simplified as 1,3-dimethylimidazolium (denoted as ). Y = H, [3a]²⁺; OMe, [3b]²⁺; NO₂, [3c]²⁺.

For computational efficiency, all structures were optimized on simplified models (see caption of Fig. 3). The mechanism proceeds through several key steps. The initial step involves the abstraction of an acidic proton adjacent to the phosphorus atom by 4-aminopyridine (AP), leading to the formation of the first intermediate (INT1). Subsequently, INT1 converts to the second intermediate (INT2) via elimination of . In this process, AP acts as an acid–base catalyst, facilitating this elimination. The next step is a nucleophilic attack, in which a second AP molecule acts as a nucleophile, attacking the deprotonated carbon atom of the C=P⁺(Ph)N group (INT2). This forms the intermediates 4a′–c′, depending on the Y-substituent. Finally, a 1,2-phenyl shift yields the final products 5a′–c′ in a concerted manner, as further discussed below.

The reaction profile provides the following energetic insights (Gibbs energies in kcal mol⁻¹). The first transition step, proton abstraction, presents a low activation barrier, ranging from 8.2 kcal mol⁻¹ (for Y = NO₂) to 10.1 kcal mol⁻¹ (for Y = H). It is endergonic for all substituents, with Y = OMe being the least stable (4.6 kcal mol⁻¹). Notably, for Y = OMe, the reverse reaction barrier is very small (4.8 kcal mol⁻¹). Therefore, in this case, the concentration of INT1 is very low compared to the starting material, which rationalizes an increased susceptibility toward competing nucleophilic pathways under these conditions (including C–P bond scission), although the ring-opening pathway itself was not explicitly modelled computationally. Further support for this interpretation is the lower positive charge at the H-atom in 3b′ compared to 3a′ and 3c′, suggesting lower acidity and allowing the competitive nucleophilic attack (Table 1).

Table 1 Charges in [e] for the H atom α to the P-atom in 3a′–c′ using different methods

Method	3a′	3b′	3c′
Mulliken	0.1764	0.1755	0.1780
NPA	0.2543	0.2497	0.2566
MK	0.1339	0.1281	0.1422
Löwdin	0.1650	0.1635	0.1660

---

The subsequent step, elimination of , is identified as the turnover-limiting event within the computed free-energy profiles, leading to the second intermediate. This step is exergonic and irreversible, with step barriers of 13–14 kcal mol⁻¹. The overall barrier energies are higher for Y = H and OMe (18.0 and 17.4 kcal mol⁻¹, respectively) compared to Y = NO₂ (14.3 kcal mol⁻¹). The nucleophilic attack by AP exhibits a very low activation barrier (1–2 kcal mol⁻¹), yielding compounds 4a′–c′. This step is also exergonic, with the nitro derivative (Y = NO₂) showing significant additional stabilization, which aligns well with the experimental observation that compound 4[OTf] is the only isolable intermediate. The final 1,2-phenyl shift is exergonic. Notably, while this step can display a comparatively large intrinsic barrier relative to its preceding minimum, the preceding intermediate is substantially stabilized such that the corresponding transition state lies only slightly above the starting materials on the overall profile.

The optimized geometries of TS1–TS4 for Y = H were used as representative models for the series and are shown in the SI, together with the Cartesian coordinates for all computed structures. Standard bond-forming and bond-breaking distances are observed for TS1 and TS3. TS4, which corresponds to the 1,2-phenyl shift, represents a concerted but asynchronous process, indicating that the intermediate and final product are connected by a single transition state. For all three systems studied, the AP ring initiates the rearrangement first, followed by the subsequent migration of the phenyl ring. In the transition state, the phenyl ring is almost equidistant from the carbon and phosphorus atoms, while the AP ring is considerably closer to the phosphorus atom.

Functionalization of 5[OTf]

As reported previously by our group, the P(III) reagent, (DMAP)₃P[OTf]₃ has served as a versatile electrophilic P(III) synthon for constructing a variety of P-functionalized compounds.⁵¹ Likewise, 5a[OTf] proved amenable to further functionalization by acting as an electrophilic azaphosphole-type P(III) fragment transfer reagent. Accordingly, reaction of 5a[OTf] with PhMgBr, PhOH (in the presence of Et₃N), and KNPh₂ furnished the P–C-, P–O-, and P–N-functionalized azaphosphole derivatives 7, 8, and 9, respectively (Scheme 3), which were isolated as analytically pure solids and fully characterized (see the SI for details; including scXRD for 7–9).

Scheme 3 Functionalization reactions of 5a[OTf]; (i) for 7, +PhMgBr, −DMAP, −MgBrOTf, r.t., 2 h, 84%; for 8, +PhOH, +Et₃N, −DMAP, −Et₃N–HOTf, 60 °C, 2 h, 69%; for 9, +KNPh₂, −KOTf, −DMAP, r.t., 2 h, 64%.

Compared to the parent azaphosphole framework, the ³¹P NMR resonances of these products are shifted upfield (δ(³¹P) = 98.1 ppm (7), 108.4 ppm (8), and 147.5 ppm (9)), consistent with substitution at the tricoordinate phosphorus center.⁵² Overall, this methodology provides straightforward access to structurally elaborate cyclic phosphorus compounds that are difficult to obtain by conventional approaches. The coordination and catalytic applications of these azaphospholes are under investigation in our group.

Reactions with nitrogen base DBU

The aforementioned results suggest that the basicity and nucleophilicity of the base significantly influence the activation pathway. To further probe this effect, other bases were examined in reactions with diazaphospholium compounds 3a–c[OTf]₂. The reaction with 1,5-diazabicyclo[4.3.0]non-5-ene (DBN) afforded non-separable mixtures in all cases (Fig. S112 and S113). In contrast, treatment of 3a[OTf]₂ with one equivalent of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) in THF at room temperature gave a yellow suspension within 2 hours. Analysis of the supernatant by ¹H NMR spectroscopy confirmed the formation of [L_C–H][OTf], whereas a solution of the precipitate (10[OTf]) in CD₂Cl₂ exhibited two resonances at δ(³¹P) = 23.4 and 18.0 ppm (6 : 1 integral ratio) in the ³¹P NMR spectrum, assigned to syn- and anti-diastereomers of 10[OTf] (Scheme 4, I). Single crystals suitable for scXRD analysis were obtained by diffusion of Et₂O into a saturated DCM solution, and the molecular structure of syn-10⁺ is shown in Fig. 4a. Contrary to expectation, functionalization of DBU with the azaphosphole fragment occurs via C–P over N–P bond formation, which has been previously observed in phosphorylation reactions of DBU.^53,54 Notably, the P1–N1 atom distance (2.306 Å) is longer than the single-bond length (P–N 1.82 Å),⁵⁵ yet shorter than the sum of the van der Waals radii (3.35 Å),⁵⁶ indicating a possible intramolecular interaction. The P1–C1 bond length (1.836 Å) aligns with a typical P–C single bond.

Scheme 4 (I) (i) +DBU, −[L_C–H][OTf], THF, r.t., 2–12 h, 78%. (ii) +KHMDS, −HMDS, −KOTf, THF, r.t., 2 h, 86%. (iii) THF, r.t., 2–12 h or 60 °C, 2 h. Paraformaldehyde, 13[OTf], 74%; 4-MeOPhCHO, E-14[OTf], 64%; 4-MePhCHO, E-15[OTf], 76%; 4-NO₂PhCHO, Z-16[OTf], 60%; E-cinnamaldehyde, Z,E-17[OTf], 78%. (iv) 2 bar CO₂, −196 °C to r.t., C₆D₆, 10 min or 1 bar CO₂, r.t., THF, 10–30 min, 71%; (v) CS₂, THF, r.t., 10 min, 52%. (II) Photographic representation of CO₂ activation by 11 in THF under a CO₂ atmosphere (1 atm) at room temperature. The THF solution of 11 before exposure to CO₂ (a), after exposure to CO₂ for 10 min (b), and after exposure to CO₂ for 20 min (c). Stereocenters are marked by asterisk [*].

Fig. 4 Molecular structures of syn-10⁺ in syn-10[OTf] (a) and 11 (b); hydrogen atoms and other solvent molecules are omitted for clarity; thermal ellipsoids are displayed at 50% probability (100 K); selected bond lengths (in Å) and angles (in °): 10[OTf]: P1–C1 1.836(3), C1–C2 1.490(5), C2–N1 1.298(5), P1–C1–C2 98.8(2), C1–C2–N1 110.5(3); 11: P1–C1 1.7488(12), C1–C2 1.3987(17), C2–N1 1.3652(17), P1–N1 1.9670(11), C1–P1–N1 72.39(5), P1–C1–C2 94.47(8), C1–C2–N1 105.58(10), C2–N1–P1 86.41(8).

Reactions of 3b[OTf]₂ and 3c[OTf]₂ with DBU gave complex mixtures. These observations are well supported by the computed electronic properties of the systems. The decreased α-proton acidity in 3b redirects the pathway toward unselective nucleophilic side-reactions, whereas the strong stabilization of intermediates induced by the nitro group in 3c increases the effective kinetic barriers for the subsequent steps, leading in both cases to the observed complex mixtures.

The mechanism for the transformation of 3a′ to 10′ was investigated using DFT calculations, and the proposed pathway is illustrated in Fig. 5. The mechanism begins with a proton abstraction similar to that described above for DMAP. Specifically, the nitrogen atom of DBU abstracts the proton adjacent to the positively charged phosphorus atom via TS1. This step exhibits a low activation energy and is exergonic by 7.8 kcal mol⁻¹. Subsequently, L_C⁺ abstracts a proton from the seven membered ring, which is activated and sufficiently acidic due to the adjacent iminium group. This proton transfer constitutes the rate determining step of the reaction, possessing a global activation barrier of 23.3 kcal mol⁻¹. This process generates an enamine intermediate along with the formation of . Finally, the nucleophilic enamine attacks the phosphorus atom (TS3) with a concurrent proton transfer, yielding the final product 10. This final step has a low activation energy, with TS3 located only 3.0 kcal mol⁻¹ above the starting materials. Moreover, the formation of 10 is highly exergonic by −10.1 kcal mol⁻¹, providing strong thermodynamic support for its formation.

Fig. 5 Reaction profile (ΔG° in kcal mol⁻¹) for the transformation process, at the BP86-D4/def2-TZVP level of theory. The reactant was simplified as 3a′. Specifically, the phenyl group attached to the nitrogen atom of N-benzylideneaniline was replaced with a methyl group, as this group is not involved in the mechanistic steps.

Small molecules activation

Deprotonation of the phosphonium salt 10[OTf] with KHMDS readily formed the corresponding ylide 11 in THF solution at room temperature, as evidenced by an upfield-shifted resonance at δ(³¹P) = −23 ppm in the ³¹P NMR and the disappearance of the C1–H proton resonance (δ(¹H) = 4.14 ppm) in 10[OTf]. Compound 11 gradually decomposes within several days under an inert atmosphere, but was isolated in high yield (86%) by extraction with Et₂O and fully characterized. Orange-colored crystals suitable for scXRD analysis were obtained by vapor diffusion of n-pentane into a saturated Et₂O solution and verified the assignment (Fig. 4b). Relative to the parent structure, a prominent structural change is the formation of a strained four-membered ring, as evidenced by the remarkably shortened P1⋯N1 distance of 1.9670(11) Å. This distance lies between that of azaphosphetidines [2.170(3) Å] and their saturated analogues [1.7985(5) Å],^57,58 and is longer than that of reported amidophosphoranes [1.842(7) Å],⁵⁹ consistent with a strongly constrained interaction between nitrogen (N1) and phosphorus (P1) rather than a fully localized P–N single bond. The P1–C1 bond length is significantly shortened to 1.7488(12) Å, approaching the range observed for phosphonium-ylides.⁶⁰ In addition, the shortened C1–C2 bond [10[OTf], 1.490(5) Å; 11, C1–C2 1.3987(17) Å] and the elongated C2–N1 bond [10[OTf], 1.298(5) Å; 11, 1.3652(17) Å] indicate electron density transfer from N toward the central P-center. Thus, the strong P/N interaction results in a rare strain-enforced bonding environment at the P atom.

In analogy to conventional Wittig ylides, 11 exhibited canonical Wittig-type reactivity toward carbonyl electrophiles.^47,61–63 Reaction with paraformaldehyde and various substituted benzaldehydes afforded the corresponding DBU-derived products, isolated as their protonated triflate salts after work-up in good yields (Fig. 6a, Z,E-17[OTf]; for 14–17[OTf], see the SI for details). The phosphole oxide by-product 12 was isolated and characterized as a mixture of two diastereomers. Their diastereomeric configurations are differentiated by the relative orientation of the α-proton and the oxygen atom on the five-membered ring, as confirmed by scXRD analysis (see the SI, Fig. S94).

Fig. 6 Molecular structures of Z,E-17⁺ in Z,E-17[OTf] (a), 18 (b), and 20 (c). Hydrogen atoms, anions, and solvent molecules are omitted for clarity; thermal ellipsoids are displayed at 50% probability (100 K).

Beyond its canonical Wittig-type behavior, compound 11 exhibited distinctive reactivity toward CO₂ (Scheme 4). In an initial attempt, a C₆D₆ solution of 11 in a J. Young NMR tube was frozen in liquid nitrogen, evacuated, and charged with 2 bar CO₂. During thawing, an immediate color change from orange to pale yellow was observed, and the ³¹P NMR spectrum confirmed the formation of 12. Alongside 12, the main product of this conversion was identified as oxazinedione derivative 18 by scXRD (Fig. 6b) and was isolated as an air- and moisture-stable white solid. Remarkably, this transformation also proceeded efficiently at room temperature under 1 atm CO₂ (Scheme 4, II).

Similarly, CS₂ can also be activated in this manner. The reaction can proceed rapidly even at −78 °C, affording the phosphole sulfide 19 and the thiazinedithione derivative 20, a heavier chalcogenide analogue of 18 featuring a C₂S₃ fragment (Fig. 5c). Both species are air-stable and can be conveniently separated by extraction with toluene (52% for 20).

Phosphorus ylides have shown potential for CO₂ transformation by forming zwitterionic intermediates with CO₂, which can undergo further reactions (including hydrolysis) to furnish carboxylic acid derivatives.^64,65 Interestingly, in our case, the endocyclic N atom participates in the formation of oxazinedione derivative 18, consistent with a concerted bimolecular pathway rather than simple CO₂ adduct formation. Bimolecular CO₂ activation has only been rarely reported, typically in metal-mediated systems,⁶⁶ whereas more recent studies have mainly described the formation of CO₂ adducts.^{59,64,67–69}

In contrast, compound 11 operates via a rare activation mode, which was further investigated using DFT analysis. Fig. 7a illustrates the proposed mechanism for the transformation of ylide 11 with two equivalents of CO₂ (an analogous mechanism can be formulated for CS₂) (see detailed mechanism in Fig. S118). For computational economy, we replaced the phenyl groups by methyl groups in compound 11 (denoted as 11′). The process starts with the opening of the four-membered ring via a conrotatory reaction as has been studied previously in the literature.⁷⁰ This transient intermediate is immediately trapped by the CO₂ via a (4 + 2) cycloaddition reaction, forming a six-membered ring. The nucleophilic C-atom of the enamine-like unit attacks a second molecule of CO₂, forming a zwitterionic intermediate.

Fig. 7 (a) Gibbs energies for the conversion of 11′ with CO₂ and CS₂ to yield products 12′/18 and 19′/20. (b) BP86-D4/def2-TZVP geometries of the transition states for the (4 + 2) cycloaddition of 11′ with CO₂. (c) The resulting intermediates (INT) for CO₂. Distances in Å. A truncated model for 11 was used for calculation.

Then the anhydride functional group is formed upon the attack of the carboxylate oxygen atom to the “phosphaneyl-carbamate” fragment, eliminating compound 12′ and regenerating the enamine group to furnish 18. The mechanism was analyzed from a thermodynamic point of view and by computing the transition state for the conrotatory opening of the four-membered ring in order to explain the activation process. The results are gathered in Fig. 7b, evidencing that the first step is basically barrierless (0.6 kcal mol⁻¹) for CO₂ and proceeds with a low barrier (5.6 kcal mol⁻¹) for CS₂, which is consistent with the experimentally observed fast reaction and its occurrence at low temperature. This first mechanistic step is more exergonic for CO₂ than for CS₂ (−8.0 and −4.9 kcal mol⁻¹, respectively), further supporting that this initial step is significantly more favored for CO₂.⁷¹ The following addition of the second molecule of CO₂/CS₂ and elimination of the phosphole (sulfur/oxygen) chalcogenides 12′/19′ is also exergonic for both compounds, but in this case is more exergonic for CS₂ (−24.3 kcal mol⁻¹) than for CO₂ (−9.0 kcal mol⁻¹).

Conclusions

In summary, we have established a modular synthetic platform that converts readily accessible precursors into diverse cyclic phosphorus-containing frameworks. A practical three-component (1 + 2 + 2)-cycloaddition provides straightforward access to a series of diazaphospholium triflate salts. The imidazoliumyl (L_C) substituent serves as a programmable handle that directs subsequent, base-mediated diversification of these P(V) heterocycles. Treatment with DMAP triggers an unusual 1,2-phenyl migration, effecting a formal P(V)-to-P(III) conversion and furnishing cationic P(III)–DMAP adducts that can be further transformed into P-functionalized azaphospholes bearing P–C, P–O, and P–N bonds through reactions with appropriate nucleophiles. Together, these findings reveal a rare, operationally simple strategy for converting azaphospholium P(V) salts into isolable P(III) synthons. In contrast, treatment with DBU affords a DBU-derived phosphonium salt that can be deprotonated to an ylide. This Wittig-type species retains canonical olefination reactivity towards aldehydes while also enabling an atypical bimolecular activation of CO₂ and CS₂, leading to oxazinedione-type or thiazinedithione products alongside the corresponding phosphole chalcogenides. Ongoing work in our group aims to expand the (1 + 2 + 2)-cycloaddition to other dipolar two-atom components and nitrile classes, and to explore the coordination chemistry and catalytic utility of these new azaphosphole derivatives.

Author contributions

L. H., K. S., and J. J. W. conceptualized the study; L. H. conducted the experiments and optimized the syntheses, isolations, and purifications; R. G. and A. F. were responsible for mechanistic studies; J. F., P. R., and J. J. W. were responsible for X-ray data collection and refinement; K. S. and J. J. W. conceived, oversaw, and directed the project; L. H. wrote the initial draft of the paper. All authors contributed to data analysis, manuscript review and editing, and discussion.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting the findings of this study are available in the supplementary information (SI). Supplementary information: CIF files, NMR spectra, and computational details. See DOI: https://doi.org/10.1039/d6qi01242f.

CCDC 2543129–2543151 contain the supplementary crystallographic data for this paper.^72a–w

Additional data can be obtained from the corresponding author upon reasonable request.

Acknowledgements

This work was supported by the German Research Foundation (DFG; WE 4621/6–1, WE 4621/6-2). L. H. was supported by the China Scholarship Council (CSC No. 202106360039). A. F. and R. M. G. thank MICIU/AEI (Spain) for financial support, project PID2023-148453NB-I00, co-funded by FEDER funds. We also thank Technische Universität Dresden (TUD) for support.

References

1. X. Chen, D. J. Kopecky, J. Mihalic, S. Jeffries, X. Min, J. Heath, J. Deignan, S. Lai, Z. Fu, C. Guimaraes, S. Shen, S. Li, S. Johnstone, S. Thibault, H. Xu, M. Cardozo, W. Shen, N. Walker, F. Kayser and Z. Wang, Structure-guided design, synthesis, and evaluation of guanine-derived inhibitors of the eIF4E mRNA-cap interaction, J. Med. Chem., 2012, 55, 3837–3851.
2. F.-R. Alexandre, A. Amador, S. Bot, C. Caillet, T. Convard, J. Jakubik, C. Musiu, B. Poddesu, L. Vargiu, M. Liuzzi, A. Roland, M. Seifer, D. Standring, R. Storer and C. B. Dousson, Synthesis and biological evaluation of aryl-phospho-indole as novel HIV-1 non-nucleoside reverse transcriptase inhibitors, J. Med. Chem., 2011, 54, 392–395.
3. F. Mathey, The organic chemistry of phospholes, Chem. Rev., 1988, 88, 429–453.
4. M. P. Duffy, W. Delaunay, P.-A. Bouit and M. Hissler, π-Conjugated phospholes and their incorporation into devices: components with a great deal of potential, Chem. Soc. Rev., 2016, 45, 5296–5310.
5. W. Tang and X. Zhang, New chiral phosphorus ligands for enantioselective hydrogenation, Chem. Rev., 2003, 103, 3029–3070.
6. H. Fernández-Pérez, P. Etayo, A. Panossian and A. Vidal-Ferran, Phosphine-phosphinite and phosphine-phosphite ligands: preparation and applications in asymmetric catalysis, Chem. Rev., 2011, 111, 2119–2176.
7. Cycloaddition reactions in organic synthesis, ed. S. Kobayashi and K. A. Jorgensen, Wiley-VCH, Weinheim, Germany, 2002.
8. L. Dettling, M. Papke, J. A. W. Sklorz, D. Buzsáki, Z. Kelemen, M. Weber, L. Nyulászi and C. Müller, A new access to diazaphospholes via cycloaddition-cycloreversion reactions on triazaphospholes, Chem. Commun., 2022, 58, 7745–7748.
9. S. Weidner, J. Renner, U. Bergsträßer, M. Regitz and H. Heydt, Organophosphorus Compounds, Part 168;1 1,3-Dipolar Cycloaddition Reactions of 1,3,5-Triphosphinines with Nitrile Oxides, Synthesis, 2004, 241–248.
10. J. A. W. Sklorz, S. Hoof, N. Rades, N. de Rycke, L. Könczöl, D. Szieberth, M. Weber, J. Wiecko, L. Nyulászi, M. Hissler and C. Müller, Pyridyl-functionalised 3H-1,2,3,4-triazaphospholes: synthesis, coordination chemistry and photophysical properties of low-coordinate phosphorus compounds, Chem. – Eur. J., 2015, 21, 11096–11109.
11. S. L. Choong, A. Nafady, A. Stasch, A. M. Bond and C. Jones, The facile assembly of bis-, tris- and poly(triazaphosphole) systems using “click” chemistry, Dalton Trans., 2013, 42, 7775–7780.
12. A. H. Cowley, R. A. Kemp, J. G. Lasch, N. C. Norman and C. A. Stewart, Reaction of phosphenium ions with 1,3-dienes: a rapid synthesis of phosphorus-containing five-membered rings, J. Am. Chem. Soc., 1983, 105, 7444–7445.
13. R. J. Boyd, N. Burford and C. L. B. Macdonald, Ab Initio Studies of the Contrasting Butadiene Cheletropic and Diels–Alder Cycloaddition Reactivities Observed for “Carbenic” Phosphorus (Phosphenium) and Arsenic (Arsenium) Cations, Organometallics, 1998, 17, 4014–4029.
14. M. B. Abrams, B. L. Scott and R. T. Baker, Sterically Tunable Phosphenium Cations: Synthesis and Characterization of Bis(arylamino)phosphenium Ions, Phosphinophosphenium Adducts, and the First Well-Defined Rhodium Phosphenium Complexes, Organometallics, 2000, 19, 4944–4956.
15. N. Burford, P. J. Ragogna, R. McDonald and M. J. Ferguson, Phosphine Coordination Complexes of the Diphenylphosphenium Cation: A Versatile Synthetic Methodology for P–P Bond Formation, J. Am. Chem. Soc., 2003, 125, 14404–14410.
16. C. A. Caputo, J. T. Price, M. C. Jennings, R. McDonald and N. D. Jones, N-heterocyclic phosphenium cations: syntheses and cycloaddition reactions, Dalton Trans., 2008, 3461–3469.
17. A. H. Cowley and R. A. Kemp, Synthesis and reaction chemistry of stable two-coordinate phosphorus cations (phosphenium ions), Chem. Rev., 1985, 85, 367–382.
18. N. Đorđević, R. Ganguly, M. Petković and D. Vidović, E-H (E = B, Si, C) Bond Activation by Tuning Structural and Electronic Properties of Phosphenium Cations, Inorg. Chem., 2017, 56, 14671–14681.
19. N. Đorđević, M. Q. Y. Tay, S. Muthaiah, R. Ganguly, D. Dimić and D. Vidović, C-F bond activation by transient phosphenium dications, Inorg. Chem., 2015, 54, 4180–4182.
20. B. D. Ellis, P. J. Ragogna and C. L. B. Macdonald, Computational insights into the acceptor chemistry of phosphenium cations, Inorg. Chem., 2004, 43, 7857–7867.
21. C. Hering, A. Schulz and A. Villinger, On the synthesis and reactivity of highly labile pseudohalogen phosphenium ions, Inorg. Chem., 2013, 52, 5214–5225.
22. A. Jayaraman, S. Nilewar, T. V. Jacob and B. T. Sterenberg, Sequential Electrophilic Substitution Reactions of Tungsten-Coordinated Phosphenium Ions and Phosphine Triflates, ACS Omega, 2017, 2, 7849–7861.
23. Á. Kozma, T. Deden, J. Carreras, C. Wille, J. Petuškova, J. Rust and M. Alcarazo, Coordination Chemistry of Cyclopropenylidene-Stabilized Phosphenium Cations: Synthesis and Reactivity of Pd and Pt Complexes, Chem. – Eur. J., 2014, 20, 2208–2214.
24. L. L. Liu, D. A. Ruiz, F. Dahcheh and G. Bertrand, Isolation of a Lewis base stabilized parent phosphenium (PH₂⁺) and related species, Chem. Commun., 2015, 51, 12732–12735.
25. V. Nesterov, D. Reiter, P. Bag, P. Frisch, R. Holzner, A. Porzelt and S. Inoue, NHCs in Main Group Chemistry, Chem. Rev., 2018, 118, 9678–9842.
26. M. Olaru, D. Duvinage, E. Lork, S. Mebs and J. Beckmann, Transient Phosphenium and Arsenium Ions versus Stable Stibenium and Bismuthenium Ions, Chem. – Eur. J., 2019, 25, 14758–14761.
27. M. Olaru, A. Mischin, L. A. Malaspina, S. Mebs and J. Beckmann, The Bis(ferrocenyl)phosphenium Ion Revisited, Angew. Chem., Int. Ed., 2020, 59, 1581–1584.
28. J. Petuškova, H. Bruns and M. Alcarazo, Cyclopropenylylidene-Stabilized Diaryl and Dialkyl Phosphenium Cations: Applications in Homogeneous Gold Catalysis, Angew. Chem., Int. Ed., 2011, 50, 3799–3802.
29. M. A. Samsonov, J. Vrána, P. Švec, V. Němec, E. Procházková, J. Cvačka and A. Růžička, Enhanced Reactivity of N-heterocyclic Halophosphines and Phosphenium Cations: The Activation of Carbon–Heteroatom and Boron–Heteroatom Bonds, Chem. – Eur. J., 2025, 31, 2–8.
30. J. M. Slattery and S. Hussein, How Lewis acidic is your cation? Putting phosphenium ions on the fluoride ion affinity scale, Dalton Trans., 2012, 41, 1808–1815.
31. C. K. SooHoo and S. G. Baxter, Phosphenium ions as dienophiles, J. Am. Chem. Soc., 1983, 105, 7443–7444.
32. S. A. Weissman and S. G. Baxter, Evidence For The Rearrangement Of P-Chloro-Phosphirenium Ions To P-Vinyl-Phosphenium Ions, Tetrahedron Lett., 1990, 31, 819–822.
33. C. Stoian, N. Schmidt, T. J. Kuczmera, P. Puylaert, E. Lork, B. J. Nachtsheim, E. Hupf and J. Beckmann, Oxidative addition of diaryldichalcogenides to the diferrocenylphosphenium ion: synthesis, structure and organocatalytic activity, Chem. Commun., 2025, 61, 8256–8259.
34. M. Q. Y. Tay, G. Ilić, U. Werner-Zwanziger, Y. Lu, R. Ganguly, L. Ricard, G. Frison, D. Carmichael and D. Vidović, Preparation, Structural Analysis, and Reactivity Studies of Phosphenium Dications, Organometallics, 2016, 35, 439–449.
35. R. Yadav, A. Sharma, B. Das, C. Majumder, A. Das, S. Sen and S. Kundu, Air and Water Stable Bicyclic (Alkyl)(Amino)Carbene Stabilized Phosphenium Cation: Reactivity and Selective Fluoride Ion Affinity, Chem. – Eur. J., 2024, 30, e202401730.
36. Y. Unoh, K. Hirano and M. Miura, Metal-Free Electrophilic Phosphination/Cyclization of Alkynes, J. Am. Chem. Soc., 2017, 139, 6106–6109.
37. A. Jayaraman and B. T. Sterenberg, Phosphorus–Carbon Bond Forming Reactions of Diphenylphosphenium and Diphenylphosphine Triflate Complexes of Tungsten, Organometallics, 2016, 35, 2367–2377.
38. D. Gasperini, S. E. Neale, M. F. Mahon, S. A. Macgregor and R. L. Webster, Phosphirenium Ions as Masked Phosphenium Catalysts: Mechanistic Evaluation and Application in Synthesis, ACS Catal., 2021, 11, 5452–5462.
39. L. You, D. Roth and L. Greb, Structural constraint at a P-P bond: phosphinophosphination of alkenes, alkynes, and carbonyls by a concerted mechanism, Chem. Sci., 2025, 16, 1716–1721.
40. A. Dumitrescu, H. Gornitzka, W. W. Schoeller, D. Bourissou and G. Bertrand, A Crystalline Phosphenium Salt Featuring the Electron-Withdrawing 2,6-Bis(trifluoromethyl)phenyl Group, Eur. J. Inorg. Chem., 2002, 2002, 1953–1956.
41. J. Fidelius, K. Schwedtmann, S. Schellhammer, J. Haberstroh, S. Schulz, R. Huang, M. C. Klotzsche, A. Bauzá, A. Frontera, S. Reineke and J. J. Weigand, Convenient access to π-conjugated 1,3-azaphospholes from alkynes via [3 + 2]-cycloaddition and reductive aromatization, Chem, 2024, 10, 644–659.
42. J. Fidelius, K. Schwedtmann, S. Schellhammer, R. Huang, F. Hennersdorf, M. Fink, J. Haberstroh, A. Bauzá, A. Frontera, S. Reineke and J. J. Weigand, 1,3-Dipolar cycloaddition reactions of triflatophosphanes to afford functionalized azaphospholium salts and azaphospholes, Inorg. Chem. Front., 2025, 12, 3324–3333.
43. P. Royla, K. Schwedtmann, Z. Han, J. Fidelius, D. P. Gates, R. M. Gomila, A. Frontera and J. J. Weigand, Cationic Phosphinidene as a Versatile P₁ Building Block: LC-P+ Transfer from Phosphonio-Phosphanides [LC-P-PR₃]⁺ and Subsequent LC Replacement Reactions (LC = N-Heterocyclic Carbene), J. Am. Chem. Soc., 2023, 145, 10364–10375.
44. J. J. Weigand, K.-O. Feldmann and F. D. Henne, Carbene-stabilized phosphorus(III)-centered cations [LPX₂]⁺ and [L₂PX]²⁺ (L = NHC; X = Cl, CN, N₃), J. Am. Chem. Soc., 2010, 132, 16321–16323.
45. R. M. Tipker, J. A. Muldoon, D. H. Pham, B. R. Varga, R. P. Hughes, D. S. Glueck, G. J. Balaich and A. L. Rheingold, Configurational Lability at Tetrahedral Phosphorus: syn/anti -Isomerization of a P-Stereogenic Phosphiranium Cation by Intramolecular Epimerization at Phosphorus, Angew. Chem., Int. Ed., 2022, 61, e202110753.
46. S. Fankel, H. Helten, G. von Frantzius, G. Schnakenburg, J. Daniels, V. Chu, C. Müller and R. Streubel, Novel access to azaphosphiridine complexes and first applications using Brønsted acid-induced ring expansion reactions, Dalton Trans., 2010, 39, 3472–3481.
47. G. Wittig, From diyls to ylides to my idyll, Science, 1980, 210, 600–604.
48. D. Lednicer, Preparation of 1,6-diarylhexatrienes by a modified Wittig reaction, J. Org. Chem., 1971, 36, 3473–3474.
49. Y. Masuda, D. Ikeshita, K. Higashida, M. Yoshida, N. Ishida, M. Murakami and M. Sawamura, Photocatalytic 1,2-Phosphorus-Migrative [3 + 2] Cycloaddition of Tri(t-butyl)phosphine with Terminal Alkynes, J. Am. Chem. Soc., 2023, 145, 19060–19066.
50. N. Burford, P. Losier, A. D. Phillips, P. J. Ragogna and T. S. Cameron, Nitrogen ligands on phosphorus(III) Lewis acceptors: A versatile new synthetic approach to unusual N-P structural arrangements, Inorg. Chem., 2003, 42, 1087–1091.
51. M. Donath, K. Schwedtmann, T. Schneider, F. Hennersdorf, A. Bauzá, A. Frontera and J. J. Weigand, Direct conversion of white phosphorus to versatile phosphorus transfer reagents via oxidative onioation, Nat. Chem., 2022, 14, 384–391.
52. F. Mathey, The organic chemistry of phospholes, Chem. Rev., 1988, 88, 429–453.
53. A. Kers, I. Kers and J. Stawinski, The reaction of diphenyl and dialkyl phosphorochloridates with 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU). Formation of phosphonate diesters via N→C phosphorus migration, J. Chem. Soc., Perkin Trans. 2, 1999, 2071–2075.
54. M. O'Reilly, R. Pattacini and P. Braunstein, Anion effect in the diastereoselective formation of bischelated Ni(II) complexes with a novel, chiral phosphine derived from 1,8-diazabicyclo5.4.0undec-7-ene (DBU), Dalton Trans., 2009, 6092–6095.
55. P. Pyykkö and M. Atsumi, Molecular single-bond covalent radii for elements 1–118, Chem. – Eur. J., 2009, 15, 186–197.
56. M. Mantina, A. C. Chamberlin, R. Valero, C. J. Cramer and D. G. Truhlar, Consistent van der Waals radii for the whole main group, J. Phys. Chem. A, 2009, 113, 5806–5812.
57. T. Kawashima, T. Soda and R. Okazaki, Synthesis, Structure, and Thermolysis of N-Apical 1,2λ⁵-Azaphosphetidines with a Pentacoordinate P Center and the First Observation of Their N-Equatorial Pseudorotamers, Angew. Chem., Int. Ed. Engl., 1996, 35, 1096.
58. N. Kano, A. Kikuchi and T. Kawashima, The first isolable pentacoordinate 1,2 lambda 5-azaphosphetine: synthesis, X-ray crystallographic analysis, and dynamic behaviour, Chem. Commun., 2001, 2096–2097.
59. L. J. Hounjet, C. B. Caputo and D. W. Stephan, Phosphorus as a Lewis acid: CO₂ sequestration with amidophosphoranes, Angew. Chem., Int. Ed., 2012, 51, 4714–4717.
60. J. C. J. Bart, Structure of the non-stabilized phosphonium ylid methylenetriphenylphosphorane, J. Chem. Soc. B, 1969, 350–365.
61. B. E. Maryanoff and A. B. Reitz, The Wittig olefination reaction and modifications involving phosphoryl-stabilized carbanions. Stereochemistry, mechanism, and selected synthetic aspects, Chem. Rev., 1989, 863–927.
62. E. Vedejs and C. F. Marth, Mechanism of Wittig reaction: evidence against betaine intermediates, J. Am. Chem. Soc., 1990, 3905–3909.
63. P. A. Byrne and D. G. Gilheany, The modern interpretation of the Wittig reaction mechanism, Chem. Soc. Rev., 2013, 42, 6670–6696.
64. H. Zhou, G.-X. Wang, W.-Z. Zhang and X.-B. Lu, CO₂ Adducts of Phosphorus Ylides: Highly Active Organocatalysts for Carbon Dioxide Transformation, ACS Catal., 2015, 5, 6773–6779.
65. H. J. Bestmann, T. Denzel and H. Salbaum, Reaktion von phosphinalkylenen mit CO₂. Eine neue möglichkeit zur synthese von carbonsäuren, allenen und acylyliden, Tetrahedron Lett., 1974, 15, 1276–1974.
66. J. Langer, A. Hamza and I. Pápai, RuBisCO-Inspired CO₂ Activation and Transformation by an Iridium(I) Complex, Angew. Chem., Int. Ed., 2018, 57, 2455–2458.
67. D. W. Stephan, Frustrated Lewis pairs: from concept to catalysis, Acc. Chem. Res., 2015, 48, 306–316.
68. D. W. Stephan and G. Erker, Frustrated Lewis pair chemistry: development and perspectives, Angew. Chem., Int. Ed., 2015, 54, 6400–6441.
69. C.-X. Guo, K. Schwedtmann, J. Fidelius, F. Hennersdorf, A. Dickschat, A. Bauzá, A. Frontera and J. J. Weigand, Bifunctional Fluorophosphonium Triflates as Intramolecular Frustrated Lewis Pairs: Reversible CO₂ Sequestration and Binding of Carbonyls, Nitriles and Acetylenes, Chem. – Eur. J., 2021, 27, 13709–13714.
70. Z. Mucsi, I. Hermecz, B. Viskolcz, I. G. Csizmadia and G. Keglevich, The influence of exocyclic phosphorous substituents on the intrinsic stability of four-membered heterophosphetes: a theoretical study, Tetrahedron, 2008, 64, 1868–1878.
71. K. An and J. Zhu, Why Does Activation of the Weaker C=S Bond in CS₂ by P/N-Based Frustrated Lewis Pairs Require More Energy Than That of the C=O Bond in CO₂? A DFT Study, Organometallics, 2014, 33, 7141–7146.
72. (a) CCDC 2543129: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbfx; (b) CCDC 2543130: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbgy; (c) CCDC 2543131: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbhz; (d) CCDC 2543132: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbj0; (e) CCDC 2543133: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbk1; (f) CCDC 2543134: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbl2; (g) CCDC 2543135: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbm3; (h) CCDC 2543136: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbn4; (i) CCDC 2543137: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbp5; (j) CCDC 2543138: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbq6; (k) CCDC 2543139: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbr7; (l) CCDC 2543140: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbs8; (m) CCDC 2543141: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbt9; (n) CCDC 2543142: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbvb; (o) CCDC 2543143: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbwc; (p) CCDC 2543144: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbxd; (q) CCDC 2543145: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbyf; (r) CCDC 2543146: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcbzg; (s) CCDC 2543147: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcc0j; (t) CCDC 2543148: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcc1k; (u) CCDC 2543149: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcc2l; (v) CCDC 2543150: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcc3m; (w) CCDC 2543151: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2rcc4n.

---