Enhanced Reactivity of N‐heterocyclic Halophosphines and Phosphenium Cations: The Activation of Carbon─Heteroatom and Boron─Heteroatom Bonds

Abstract

N‐heterocyclic phosphines (NHPs), and especially their ionic derivatives, phosphenium cations, have gained significant attention not only as isoelectronic species to N‐heterocyclic carbenes (NHCs) but mainly, as highly Lewis acidic, active organocatalysts. A novel type of saturated NHPs has been introduced, with the nitrogen atoms substituted by phosphoramidothioates–{[κ²‐O,O‐(OCH₂C(Me)₂CH₂O)P(= S)]₂‐μ‐[NCH₂CH₂N]}PX, where X = F, Cl, Br, I, H, Ph, and their derivatives. Thanks to an increased positive charge at the P(III) atom, these compounds are prone to activate C─O, B─H, B─C, or B─F bonds and open cyclic ethers through the formation of intermediate species–phosphenium cations. The hydrido derivative can effectively catalyze the transfer hydrogenation of azobenzene by ammonia borane when irradiated by UV light at room temperature. The self‐ionization of the iodido derivative promoted by the coordination of less sterically demanding N‐heterocyclic carbene enables the isolation of a phosphenium cation.

Novel class of saturated NHPs with phosphoramidothioate substituents exhibit enhanced reactivity in photoinitiated transfer hydrogenation reactions and activate C─O, B─H, B─C, and B─F bonds via phosphenium cation formation.

1. Introduction

N‐heterocyclic phosphines (Scheme 1, NHPs) constitute a broad class of phosphorus‐containing compounds, where the phosphorus atom is chelated by two amido substituents. These species are, on the one hand, exceptionally stable thanks to both kinetic and thermodynamic stabilization; on the other hand, they also exhibit specific properties and reactivity.^{[ 1 ]} There are two particular ways of influencing the properties of the phosphorus atom: the charge distribution within the heterocyclic ring and also the nature of the exocyclic P─X (X = halogen, hydrogen, or organic group) bond. This is manifested by hundreds of reported NHPs containing common aromatic or aliphatic substituents at nitrogen atoms.^{[ 1 ]} In contrast, the number of the species having other heteroatoms directly connected to nitrogen is limited to thirty up to date.^{[ 2} , ³ , ⁴ , ^{5 ]} Another possibility is to design compounds with π‐electron‐conjugated heterocycles. The most pronounced representatives of compounds with unsaturated rings are 1,3,2‐diazaphospholenes (DAPs), which have a wide range of applications in organic synthesis^{[ 6 ]} or as ligands in organometallic chemistry^{[ 7} , ⁸ , ^{9 ]} and also offer great potential in homogenous catalysis.^{[ 10} , ¹¹ , ¹² , ¹³ , ^{14 ]} Unlike DAPs, saturated NHPs lack the ability of π‐electron conjugation, as a result of which their exocyclic P─X bond is more polarized. They have also found applications in various fields of chemistry, but the number of papers published on them is significantly lower when compared to DAPs.^{[ 15} , ¹⁶ , ^{17 ]} The heterolytic cleavage of the phosphorus─halogen bond in N‐heterocyclic halophosphines leads to the formation of phosphenium cations, isoelectronic analogues of N‐heterocyclic carbenes (NHCs).^{[ 18 ]} Phosphenium cations have found applications as ligands for transition metals,^{[ 19} , ^{20 ]} in enantioselective imine reduction,^{[ 21 ]} or in the preparation of light‐emitting materials.^{[ 22 ]} The carbobiscarbene‐stabilized phosphenium cations and dications can efficiently activate E─H bonds (E═B, C, Si).^{[ 23 ]}

Scheme 1.

Examples of the molecular structures of various NHPs. NBO (QTAIM) charges on phosphorus atoms in e, DAPs: R =  ^t Bu, Mes, Cy, 2,6‐(diisopropyl)phenyl.^{[ 24} , ^{25 ]} Saturated: R =  ^t Bu, Me, Ph, 4‐OMe‐C₆H₄.^{[ 26} , ²⁷ , ²⁸ , ^{29 ]} More information on the charge distribution in the heterocycle is presented in Table S8.

In the literature, there are approximately fifty structurally characterized phosphenium cations of unsaturated NHPs^{[ 24} , ^{25 ]} and fewer than twelve for saturated ones.^{[ 26} , ²⁷ , ²⁸ , ²⁹ , ³⁰ , ^{31 ]} Our hypothesis is based on the presumption that a heteroatom, which will be in a high oxidation state as a substituent of the ring, can deflate electron density and thus increase the charge on the phosphorus(III) atom enormously. To the best of our knowledge, no such examples of phosphenium cations with substituents other than carbon at nitrogen atoms have been reported. We have chosen the P(V)‐based substitution of the nitrogen atoms, which is known as a very robust connection.

The present work introduces a new type of NHP based on the doubly deprotonated N,N′‐bis(5,5‐di­methyl‐2‐thioxo‐1,3,2‐dioxaphos­phinan‐2‐yl)ethane‐1,2‐diamine,^{[ 32 ]} abbreviated as SNOOP, which can be regarded as a prototype NHP species with a flexible ligand backbone and a suggested increase of reactivity.

One of the main aims of this work was to prepare and explore the reactivity of a phosphenium cation (SNOOP‐P⁺) derived from species containing the SNOOP²⁻ ligand. Although the parent diamine has been prepared and commercialized as an extremely cheap flame retardant,^{[ 33 ]} its potential in coordination chemistry has remained unexplored. This protoligand has a number of advantages: easy and cost‐effective preparation, remarkable stability in the air, and structural flexibility with both soft‐ and hard‐donor structural conformations. It is important to note that ligands closely related in structure are already used for the preparation of metal (Mg, Ca, Sr, and Ba) complexes, which are then utilized in ring‐opening polymerisation processes.^{[ 34} , ^{35 ]} In contrast, only a few examples of unsaturated NHPs with a similar structure are known.^{[ 2} , ³ , ^{36 ]}

2. Results and Discussion

Primary DFT calculations have shown the charge distribution in the hypothetical heterocycle of the SNOOP‐P⁺ (QTAIM, ^{[ 37 ]} NBO^{[ 38 ]} – Scheme 1, Figure S95, Table S8). The values obtained were compared with structurally characterized N‐heterocyclic phosphenium cations. The P(III) in SNOOP‐P⁺ carries the highest positive charge in the series, connected with an increase in reactivity as well as its possible applications.

A series of halide derivatives, starting materials for further reactivity studies, denoted as SNOOP‐PX (X = Cl (2), Br(4)), have been synthesized from 1 with PX₃ in the presence of a base (Scheme 2). SNOOP‐PI (5) has been prepared by the reaction of trimethylsilyl iodide and 2. SNOOP‐PF (3) is the product of a metathetical exchange from the chloride 2 by NaF. These compounds also serve as a benchmark for evaluation of P─X‐bond polarity.

Scheme 2.

A general scheme of the reactivity studied; (NHC( ⁱ Pr) = 1,3‐diisopropyl‐1H‐imidazol‐3‐ium‐2‐ide), R  = P(= S)((μ‐OCH₂)₂C(CH₃)₂) fragment.

We suggested using the ³¹P NMR parameters not only for compounds’ characterization but also for the investigation of the electronic situation in the molecules. The ³¹P{¹H} NMR chemical shifts for P(III) (125.7 (P─F), 138.6 (P─Cl), 154.3 (P─Br), and 169.9 (P─I) ppm) are comparable with already reported compounds of similar composition (111.2–195.7 ppm).^{[ 39} , ^{40 ]} Especially the iodo‐derivative 5 seems to be the only NHP or similar species with a covalent P‐I, as limited literature examples tend to be ion pairs both in solution and solid state,^{[ 40 ]} which is another proof of a higher electrophilicity of the P center in SNOOP derivatives. At the same time, the chemical shift for P(V) varies slightly between 64.2 (P─F) and 58.3 (P─I) ppm, which reflects an electronic communication between phosphorus atoms. The scalar ³¹P‐³¹P couplings could also be a measure of these connections, but values for 2–5 are very close to each other (∼80 Hz). These values later served to compare significant differences in bond character for the more covalently bound hydride and the cation (see below – 70.5 and 95.8 Hz).

Compounds 2–5 have also been characterized by sc‐XRD techniques (Figures S72–S75). The P─X bond distances are: 1.614(1) Å for P─F, 2.1374(7) Å for P─Cl, 2.3125(5) Å for P─Br, and 2.497(2) Å for P─I. The values for P─F and P─Cl derivatives correspond to P─X bonds for other NHP structures with purely covalent bonds found in the Cambridge Structural Database. (1.654(1) Å for P─F^{[ 39 ]} and 2.109(2) Å^{[ 4 ]} for P─Cl) The extremely long distances for the reported P─Cl/Br/I bonds (2.759(2) Å for P─Cl,^{[ 39 ]} 2.432(3)–2.947(1) Å for P─Br,^{[ 40} , ^{41 ]} and 3.426(1) Å for P─I^{[ 40 ]}) have been reported in cases of DAPs due to intra‐ and intermolecular factors,^{[ 40 ]} which are not taking place in SNOOP derivatives with predominantly covalent P─X connections. There is also a significant structural difference in the N1‐C1‐C2‐N2 torsion angle (13.08–31.05°), which can serve as a measure of heterocycle distortion within the series of 2–5. A detailed analysis of geometrical parameters is presented in Table S3. Based on these NMR and sc‐XRD findings, one is not convinced about the difference of SNOOP‐based compounds from already reported species.

The chloride 2 is easily hydrolyzed and partially oxidized when its powder is left on the air. This is more pronounced for the samples of bromide (4) and iodide (5). On the contrary, the sample of fluoride (3) is air‐stable. Surprisingly, bromide (4) and iodide (5) give the THF‐ring‐opening products SNOOP‐PO(CH₂)₄Br (7) and SNOOP‐PO(CH₂)₄I (8) (Figures 1 and S77) when dissolved under mild conditions. The complete conversion of iodide takes 10–15 minutes, whereas 4 reacts slowly at room temperature, but full conversion is achieved under reflux within several hours. Similar, but slightly slower, reactivity was observed in 2‐methyltetrahydrofuran (Me‐THF). To the best of our knowledge, the only compound of such an electronegative element that is capable of opening the THF ring in moderate yield (at 50 °C) is the phosphorus in the NHP‐copper complex, where the assistance of the copper is necessary.^{[ 42 ]} As a consequence of the halide for alkoxide substitution, the chemical shifts of P(III) in 4 and 5 change from 154.3 and 169.9 ppm to 117.5 and 116.3 ppm, respectively.

Figure 1.

The molecular structure of 7. Thermal ellipsoids are at the 40% probability level. Selected bond distances (Å) and angles (°): P1–O5 1.651(3), P1–N1 1.716(3), P1–N2 1.711(3), N1–C1 1.483(5), N2–C2 1.466(5), C1–C2 1.471(6); N1–P1–N2 88.9(2), O5–P1–N1 101.9(2).

We propose the mechanism for this reaction (Figure S89), where the first step is the dissociation of the P‐I bond in solution. This leads to the compensation of the positive charge on P(III) by the complexation of oxygen from the THF molecule connected to the polarization of the O‐C bond within the cycle (INT‐1). In the next step, the I⁻ nucleophile attacks the carbon atom and is responsible for the opening of the cycle (TS‐1, Figure S90).

In order to understand the nature of the P─X/H bonds, various topological parameters have been analyzed within the QTAIM framework (Table S11).^{[ 43} , ^{44 ]} It is clear that within the series, P─X bonds, from P─F to P─I, progressively lose their covalent character and become ionic and thus more reactive. Strongly covalent character of the P─H bond has been calculated for the corresponding phosphine SNOOP‐PH (6). Hence, it can be assumed that 6 species could be robust enough for further reactivity and catalysis. Despite the predictions, we have found that the preparation and isolation of 6 is rather challenging because common hydride sources such as LiAlH₄ and LiEt₃BH react with 2 in the sense of an unwanted side reaction to 1. The most successful way to obtain 6 is the reaction of 2 with ⁱ Bu₂AlH in THF (Scheme 1). The crystallization of 6 from various solvents (THF, toluene, benzene) typically resulted in the isolation of diphosphine 10. Although the UV irradiation (quartz tube, 254 nm) of the THF solution of 6 is not necessary for the described reactivity, as demonstrated on samples stored in the dark, it significantly accelerates the process. Crystalline 6 was finally obtained from a concentrated THF solution immediately after the reaction at −60 °C (Figure S76). One of the possibilities to stabilize compound 6 by introducing a BH³ group was used upon immediate formation of 11 (Figure S80 – crystal structure), which is favored in THF solution by 7.5 kcal/mol (Figure S95). Compound 11 was also prepared by an unprecedented reaction of 2 or 5 with an excess of BH₃·SMe₂ in Me‐THF at 80 °C for 3 hours (in the case of 5, it requires 30 minutes only). It is worth mentioning that no opening of the Me‐THF cycle is observed when 5 is reacted, which indicates that it prefers the reaction with borane over Me‐THF coordination and ionization. At the same time, if the reaction mixture of 11 is irradiated by UV light (254 nm) for a couple of hours, the ³¹P NMR spectra show a signal pattern corresponding to 10 (Table S2, Figures S64–S66). A possible mechanism of the thermally conducted process proposed by DFT calculations is shown in Figure 2.

Figure 2.

The DFT‐estimated Gibbs free‐energy profile (kcal/mol) for the reaction mechanism of 2 with BH₃·SMe₂, R  = P(S)((μ‐OCH₂)₂C(CH₃)₂).

First, BH₃ is coordinated to the lone electron pair of the phosphorus atom via TS‐1, forming INT‐1. This is followed by the process of intramolecular exchange, in which the chlorine atom is replaced by hydrogen. One of the P─N bonds is dissociated, and the BH₂ group is inserted into a six‐membered heterocycle (INT‐2). Subsequently, the cycle is contracted to five‐membered again, but the phosphorus coordinates BH₂Cl (INT‐3). Notably, INT‐3 is not stable, probably due to steric repulsion involving the chlorine atom. Consequently, BH₂Cl is de‐coordinated by means of THF, making space for the new BH₃ group, which, in turn, leads to the formation of the final product 11. In the THF solution, there is an alternative pathway from INT‐4 (6); when sufficient energy is transferred through UV irradiation, 6 becomes capable of forming the diphosphine 10 through the transition state TS‐6. An alternative mechanism involving phosphenium cation coordination to BH₃ is not taken into account, as the reaction was experimentally driven in THF and no solvent opening product was detected. The P─P bond dissociation energy in 10 is 35.8 kcal/mol, indicating that the radical formation is highly unfavorable.

In order to compare the efficiency of our compounds promoting catalytically driven reactions, N═N double‐bond hydrogenation was selected. The same azobenzene catalytic hydrogenation was also reported via the redox P^III/P^V cycling.^{[ 45 ]} Specifically, the model reaction selected was the NHP‐catalyzed reduction (transfer hydrogenation) of azobenzene by the NH₃•BH₃ complex, already employed for several unsaturated and one saturated NHPs.^{[ 46 ]}

Considering the lower thermal robustness of 6 and its particular instability toward the irradiation by UV light, we tried to turn these disadvantages into a benefit and use 6 in such a process. First, we modified the reported noncatalyzed experiments with negligible conversion (azobenzene and 4 mol. equiv. of NH₃•BH₃; CHCl₃, 80 °C, 2 days) to a process that is more suitable for 6 (unstable in chlorinated solvents). The conversion of that noncatalyzed/irradiated reaction, monitored by ¹H NMR, increased to 15 mol% in THF at 50 °C after 42 hours (Tables 1 and S11, Run 1, Figure S103).

Table 1.

Optimization of the reaction conditions between azobenzene and ammoniaborane catalyzed by 6.^{[ a ]}

Run	6 [molar equiv.]	Solvent / T [°C] / UV Irradiation [300 nm]	Azobenzene Conversion [mol. %]	Molar ratio 6′ and 1,2‐diphenyl hydrazine
1	none	THF / 50 / OFF	15	0 / 15
2	none	THF / RT / ON	43	0 / 43
3	none	MeCN / RT / ON	32	0 / 32
4	1	THF / RT / ON	100	[ b ]
5	1	THF / 50 / OFF	56	51 / 5
6	0.5	THF / RT / ON	100	[ b ]
7	0.05	THF / RT / ON	69	0 / 69
8	0.05	MeCN / RT / ON	31	5 / 26

^[a] Reaction conditions: azobenzene (0.30 mmol), NH₃•BH₃ (1.20 mmol), solvent (1.5 mL), 42 hours for each run.

^[b] Can't be established precisely due to broad overlapping signals.

When the same conditions were used for the reaction where the equimolar ratio of azobenzene and 6 had been used, the azobenzene conversion reached 56% (Table 1, Run 5, Figures S110–S113). The major compound obtained was the product of NHP addition to azobenzene (6′), analogous to previously reported species.^{[ 46 ]} Surprisingly, when the comparative experiment without 6 was carried out under UV‐light irradiation in THF at room temperature, the conversion increased to 43%, whereas the same setup in MeCN, reported as the best‐performing solvent, led to only 32% (Table 1, Runs 2–3, Figures S104–S105). Complete azobenzene conversion was achieved when one or a half equivalent of 6 had been added to the reaction under the same conditions (Table 1, Runs 4 and 6, Figures S106–S109, S114–S116). In addition, NMR spectroscopy did not reveal any evidence of diphosphine (10) or the borane NHP complex (11) after those UV‐light‐irradiated reactions. Employing 6 in the catalytic amount of 5 molar % initiated reactions to 31% and 69% yields of 1,2‐diphenylhydrazine in MeCN and THF, respectively, thus outperforming the reported processes catalyzed by phosphines and saturated NHP but exhibiting slightly worse results than the processes catalyzed by unsaturated NHP, used under thermal conditions. One could assume that the use of 11 in this catalyzed reaction would be more efficient than 6, but a lack of conversion was observed. Attempts to mechanistically explain the high performance of the UV‐light‐initiated processes were made. First, the azobenzene should isomerize between E‐ and Z‐isomers at room temperature (Figure S98), while its thermal fading will occur at elevated temperatures. Transition states for thermal hydrogenation processes of Z‐isomers by both NH₃•BH₃ and 6 were calculated, showing rather high barriers of ∼40 kcal/mol (Figures S100, S101, S98), proving that more photochemical steps are involved.

The explanation of the instability of 4 and 5 in THF solution indicates a possible involvement of the phosphenium cation coordinated by a donor. The initial attempt to prepare such a species was made by reacting 2 with sodium tetraphenylborate. However, instead of the expected product [SNOOP‐P]⁺[BPh₄]⁻, a mixture of two different substances, SNOOP‐PPh (9) and (SNOOP‐P)₂ (10) (Figures S78 and S79), was obtained (∼ratio of 9:1). Phenyl transfer from BPh₄ ⁻ to both more electropositive and electronegative elements is quite common,^{[ 47} , ^{48 ]} but the formation of the P─Ph bond in a similar reaction is reported for saturated NHP (R = Me) or its copper complex only.^{[ 26} , ^{42 ]} However, in this particular reaction process, no corresponding diphosphine was observed, likely because of the coordination of BPh₃ to the phosphorus atom. The molecular structure of 9 is the only representative, in which both sulfur atoms are rotated to opposite directions with respect to their orientation in 2 (Figure S71). Diphosphine 10 is clearly identifiable in ³¹P{¹H} NMR spectra by the characteristic multiplet signal of P(III) at 102.0 ppm (Figure S41). In the crystalline form, 10 can be obtained as two polymorphic modifications with different space groups: P‐1 (THF, toluene, benzene, Figure 3) or P2₁/n (CH₂Cl₂, Figure S79), depending on the solvent used.

Figure 3.

The molecular structure of 10. Thermal ellipsoids are at the 40% probability level. Selected bond distances (Å) and angles (°): P1–P1A 2.268(2), P1–N1 1.734(3), P1–N2 1.735(3), N1–C1 1.489(5), N2–C2 1.488(5), C1–C2 1.507(6); N1–P1–N2 92.2(2), P1A–P1–N1 97.4(1).

For understanding the mechanism of the reaction of 2 with sodium tetraphenylborate, this process was investigated in a sealed NMR tube. The NMR spectra of the sample were recorded after ultrasonic activation for 5, 30, 60, and 180 minutes (at room temperature, Table S1, Figures S58–S63). NMR spectra were recorded after another two days without any activation. A comparison of chemical shift values and their integral intensities in ³¹P{¹H} NMR spectra is presented in Table S1. In the first period (5 minutes), there was a minor signal (δ = 110.5), which corresponded to the product from the reaction of the phosphenium cation with dichloromethane‐solvent molecules (see later, Table S3, Figures S67–S69), but no significant increase in its integral intensity was observed later. After 30 minutes, the spectrum showed two additional signals, corresponding to 9 (δ = 91.2 ppm) and 10 (δ = 101.8 ppm), whose integral intensities gradually increased. With insufficient activation in the ultrasonic bath, the reaction practically did not proceed.

Theoretically, the process (Figure 4 and S82) is initiated by the formation of the phosphenium cation (INT‐1), which attacks the ipso carbon of the tetraphenylborate anion. This leads to the formation of the major (kinetic) product via TS‐1 (Figure S83). However, if another molecule of INT‐1 is close enough to 9, there is a chance to stabilize the system through an alternative pathway. This pathway indicates positive charge transfer from the phosphenium cation to the phenyl ring of INT‐1 (TS‐2) and the cleavage of the B─Ph bond. The energy profile of the reaction shows that 10 is more thermodynamically favorable, but its formation strongly depends on kinetics. In order to understand precisely the process of charge redistribution on the key atoms during the approach to TS‐2 and afterwards, we conducted a comprehensive analysis, including an intrinsic reaction coordinate scan (IRC) and NBO analysis (Tables S6–S7, Figures S84–S88). The results show a decrease in the positive charge on the central phosphorus atoms and an increase in the positive charge on the carbon atom in the phenyl ring. The significance of the phenyl ring for this pathway is experimentally confirmed in reactions of 2 with NaBH₄ and NaBF₄ (Scheme 1), where no formation of 10 was detected by ³¹P NMR.

Figure 4.

The DFT‐estimated Gibbs free energy profile (kcal/mol) for the reaction mechanism of 2 with NaBPh₄. R = P(S)((μ‐OCH₂)₂C(CH₃)₂).

In the reactions of 2 with NaBH₄ and NaBF₄, B‐H/B‐F bonds are activated, which leads to the formation of 11 and 3, respectively. The reaction of a saturated NHP with sodium borohydride, resulting in a product similar to 11, has been previously reported.^{[ 49 ]} However, in the case of sodium tetrafluoroborate, to the best of our knowledge, this is the first observation of B─F bond cleavage by the NHP cation. It should be noted that the B─F σ‐bond is one of the strongest (183 kcal/mol) chemical connections.^{[ 50 ]} For instance, literature describes the breaking of the BF₃ molecule with the participation of a platinum complex as [(Cy₃P)₂Pt]^{[ 51} , ^{52 ]} or activation of anion in [AgNHC]⁺BF₄ ⁻.^{[ 53 ]} Another option is the BF₄ ⁻ to use as the fluoride source.^{[ 54 ]}

In order to investigate the bonding situation at the phosphorus atom by a standard NMR benchmark test, ^{[ 55 ]} the phosphorus selenide derivative 13 (Figure S81) was prepared by the oxidation of 8 with Se powder refluxing in MeCN. This test makes it possible to compare the nature of the ligand's influence on the central atom by comparing the scalar ¹ J(³¹P, ⁷⁷Se) coupling constants. A phenyl derivative of saturated NHP^{[ 56 ]} substituted by benzyl groups exhibits the ¹ J(³¹P, ⁷⁷Se) of 795 Hz, whereas the value of 881 Hz measured for 13 indicates its electron‐acceptor character is stronger than the carbo‐substituted NHP.

Other attempts to obtain SNOOP‐P⁺ by reactions of 2 with various reagents containing weakly coordinating anions coming from several reagents (Ag[SbF₆], GaCl₃, TMSOTf, Ag[CB₁₁H₁₂], Na[{3,5‐(CF₃)₂C₆H₃}₄B], Li[B(C₆F₅)₄], Ag[Al(OC(CF₃)₃)₄]) did not yield any desired result. Efforts to stabilize the cation with different bases (DMAP, Et₃N, pyridine, bipyridine, PMe₃, PPh₃) were not successful either. Instead of 2, compound 5, where the P─I bond is more ionic, was used. Nevertheless, theoretical calculations have shown that all the products [SNOOP‐P⁺‐Base]I⁻ are also thermodynamically unfavorable (∆G_r > 0, Table S9). Of those species for which the energy parameters were calculated, the only one with promising values was the adduct SNOOP‐P‐NHC( ⁱ Pr) (12): the charge on P(III) was 1.265 (1.842) e and ∆G_r = −16.83 kcal/mol.

After the promising reaction was carried out, a triplet was detected at 76.1 ppm (Figure S51) in the ³¹P{¹H} NMR spectrum, which corresponds to the calculated value of 74.8 ppm. XRD analysis confirmed the expected structure of 12 (Figure 5). The angle Cg–P1–C(NHC) is significantly larger (111.87°) than Cg–P–I (98.96°) in the parent iodide (Cg is the centroid of (P1, N1/1A, N2, C1/1A, C2)). The P─C(NHC) bond (1.874(4)Å) is only slightly longer than the P─Ph (1.834(3)Å) bond. The compound 12 is well soluble only in CH₂Cl₂, but it gradually begins to react with the solvent to form SNOOP‐PCCl₃ (12′ – identified by mass spectrometry, Table S3, Figures S67–S70). In the ³¹P{¹H} NMR spectrum of 12′, the chemical shift of P(III) (109.6 ppm) is similar to that observed in the first reaction step of 2 with NaBPh₄ (110.5 ppm, Table S1, Figure S59).

Figure 5.

The molecular structure of 12. Thermal ellipsoids are at the 40% probability level. The anion part I⁻ has been omitted for clarity. Selected bond distances (Å) and angles (°): P1–C7 1.874(4), P1–N1 1.735(3), N1–C1 1.467(5), C1–C1A 1.473(8); N1–P1–N1A 88.5(2), C7–P1–N1 103.7(1).

Unfortunately, a direct comparison of ¹ J(¹³C, ³¹P) = 140.0 Hz, found for 12, with reported values is impossible because there is no description available for a structurally characterized heterocycle in which the NHC would also be used to stabilize the phosphenium cation. For this reason, the comparison was performed on [L(R)PCl₂]⁺OTf⁻ (L = imidazolium‐2‐yl, R = Dipp) ¹ J(¹³C, ³¹P) = 132.7 Hz,^{[ 49 ]} confirming that the SNOOP‐P⁺ fragment exhibits one of the highest positive charges on the phosphorus atom found. It is thus quite ironic that after numerous attempts to obtain an isoelectronic analogue of carbene, it is necessary to use the carbene itself. Last but not least, the calculated value of the fluoride ion affinity,^{[ 57 ]} used as standard criteria of Lewis acidities, for SNOOP‐P⁺ (188.6 kcal/mol) overperforms not only Piers borane (B(C₆F₅)₃–111.6 kcal/mol) but most of P‐cationic species (for saturated NHP⁺ = 179 kcal/mol – Table S10).

3. Conclusion

In conclusion, this work highlights the unconventional chemical behavior of NHPs substituted at the nitrogen atoms by a P(V) moiety. This substitution results in an increase of the positive charge on the phosphorus atom in comparison with other NHPs. This leads to higher NHP reactivity and the opening of unusual reaction pathways. Although the phosphenium species derived from this NHP is robust only when stabilized by NHC, it is responsible for the activation of strong σ‐bonds such as B─F or C─O. The opening of cyclic ethers by electronegative elements such as phosphorus, except for some of its metal complexes, was elusive. Derivative 6, with the more covalent P─H bond, can efficiently catalyze transfer hydrogenation between ammonia borane and azobenzene upon UV irradiation at room temperature. Although some of these reactions were already described in the literature, the SNOOP‐based system showed its multitalented nature by activating most of these processes.

Supporting Information

The Supporting Information contains details of preparation, characterization, sc‐XRD structures (also CIF files), NMR spectra, computational details (xyz files) and reactivity of prepared compounds. The authors have cited additional references within the Supporting Information.^{[ 43} , ⁴⁴ , ⁵⁸ , ⁵⁹ , ⁶⁰ , ⁶¹ , ⁶² , ⁶³ , ⁶⁴ , ⁶⁵ , ⁶⁶ , ⁶⁷ , ⁶⁸ , ⁶⁹ , ⁷⁰ , ⁷¹ , ⁷² , ⁷³ , ^{74 ]}

Conflict of Interest

The authors declare no conflict of interest.

Supporting information

Supporting Information

Supporting Information

Supporting Information

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.