Access to Benzo- and Naphtho-Azaphospholes via C−H Bond Activation of Aryl-Substituted Isonitriles

Graphical Abstract

From activation to migration. Activation of a C−H bond of phenyl- and naphthyl-substituted isonitriles with heterocyclopentane-1,3-diylene represents a new method that allows the formation of benzo- and naphtho-azaphospholes. For this method, the absence of substituents in ortho-position of phenyl isonitriles is obligatory, but various substituents in other positions are tolerated. This aromatic substitution reaction can be strongly influenced by light and stoichiometry of the isonitrile used.

Abstract

Differently substituted phenyl isonitriles (with C−H bonds in ortho-position) and naphthyl isonitriles were reacted with the cyclic biradical [⋅P(μ-N-Ter)₂P⋅] (1). Insertion of the isonitrile formed a cyclic five-membered biradical [⋅P(NTer)₂C(R)P⋅] (2R, R=phenyl, naphthyl) in the first step, followed by C−H activation at the aryl substituent, resulting in novel azaphospholes (5R), which could be isolated and fully characterized. The formation of the azaphospholes can be prevented by the addition of a second equivalent of isonitrile, which causes the blocking of the radical centers in 2R by adduct formation (3R). Quantum mechanical calculations showed that a significant increase in the aromaticity of the benzo- and naphtho-azaphospholes is one of the driving forces for the activation process leading to the formation of thermodynamically favored azaphospholes. Targeted activation of C−H bonds using biradical systems represents a new synthetic approach to generate benzo- and naphtho-azaphospholes.

Introduction

Molecule activation has become a growing area of research in synthetic chemistry and can be achieved by a number of different approaches. In the last two decades, concepts have been developed for this purpose in main group molecular chemistry, such as the use of Frustrated Lewis Pairs (FLPs) or biradicals, to name but two, besides the classical use of transition metal-based catalysts.^1-14 With regard to the activation of molecules, C−H activation also plays a major role as a widely used method for the introduction of special building blocks. It is therefore not surprising that a number of different and unusual examples are known.^15-21 Recently, so-called open-shell singlet biradicals have been increasingly used for small molecule activation because they have interesting electronic properties and their reactivity can be designed with activation in mind.^{8-10, 22-25} Depending on the strength of the interaction between the two radical centres, biradicals (diradicals) are also called biradicaloids.^{8, 26} In some cases, if a bond can be reversibly formed between the two radical centres, for example under irradiation with light or by thermal regime, these species are also referred to as molecular switches, which are becoming increasingly popular.^27-33

Currently, the use of molecular switches as photo actuators or in catalytic applications is in the foreground.^34-36 Recently, we have succeeded in using the four-membered cyclic biradical [⋅P(μ-N-Ter)₂P⋅] (1) (Scheme 1, Ter=2,6-dimesityl-phenyl) to activate molecules.^{8, 14, 37-41} For example, isonitriles interact with one of the P−N bonds of 1 to form a new five-membered biradical 2R.⁴² This in turn represents a molecular switch, i. e., on irradiation with red light a trans-annular P−P bond is formed, resulting in a housane-like species 4R.^43-45 However, this species can thermally re-open the P−P bond and the biradical 2R is regenerated. Furthermore, in addition to housane formation, the biradical 2R can add another isonitrile equivalent, leading to the formation of a hetera-bicyclo[2.1.1]hexane (3R).^{42, 44}

Scheme 1

Synthesis of the hetero-cyclopentanediyls 2 and utilisation as a molecular switch to generate 4 (R’=tBu, Me₂N, H). Adduct products 3 can be formed in a reversible equilibrium reaction^{45, 46} and in the case of aryl substituted isonitriles (bearing C−H bonds in ortho-position) biradical 2 can undergo C−H activation leading to the formation of azaphospholes 5.

Here we show that the biradical 2R is also capable of reacting with isonitriles under selective C−H activation to form novel benzo- and naphtho-azaphospholes in a straightforward, high-yielding procedure.⁴⁷ Finally, azaphospholes (with a condensed benzene ring also called phospha-indoles), for which we have accidentally discovered a new synthesis, represent a class of compounds that has been extensively studied in terms of their preparation and further use.^47-55 For example, the classical routes to prepare these compounds start with o-amino-phenyl phosphanes and involve the use of a cyclizing agent such as acid chlorides or esters (Scheme 2).^{48, 52, 55} Other routes use variously substituted phenyl isonitriles as starting materials, and reaction with elemental lithium can yield the corresponding aza-phospholes (Scheme 2).^{50, 54}

Scheme 2

Selection of known synthetic approaches to obtain azaphospholes.^{48, 50, 52, 54, 55}

Results and Discussion

As previously shown, a variety of isonitriles are able to insert into a P−N bond of biradical 1 yielding the corresponding five-membered hetero-cyclopentanediyls (2R, for example R=2,6-dimethylphenyl=Dmp, Scheme 1 and Scheme 3).^{42, 45, 46} To investigate the various equilibrium reactions and housane formation (4R) under irradiation (Scheme 1), depending on the phenyl substituent of the isonitrile, both electron-withdrawing and electron-donating functional groups were used in para-position of the phenyl substituent (R′=tBu, NMe₂). As depicted in Scheme 3, a set of seven differently aryl-substituted isonitriles were used in the reaction with biradical 1. Since we used only slightly modified synthesis routes to generate the isonitriles, they are not discussed here. All isonitrile syntheses together with the full set of analytical data can be found in the Supporting Information file. Furthermore, since we happened to observe a completely different reaction pathway depending on the substitution in 2- and 6-position at the aryl, we split the following discussion into two parts: (i) We start with aryl substituents having a methyl group in 2- and 6-position and (ii) then show in the second part that in case of a C−H substitution in 2- and 6-position a C−H activation happens which can even be controlled with light. This C−H activation on the aromatic leads to the formation of an azaphosphole (5R).

Scheme 3

Isonitriles used in the reaction with biradical 1.

Reaction with phenyl isonitriles featuring methyl groups in 2- and 6-position

Synthesis of 2Mtp

We started this project with the reaction of 1 with 2,6-dimethyl-4-tert-butyl-phenylisonitrile, Mtp-NC, in benzene at ambient temperatures yielding dark blue crystals of 2Mtp (Scheme 4) in good yields (82 %). Blue crystals of 2Mtp are thermally stable up to 155 °C but are extremely oxygen and moisture sensitive. In the ³¹P{¹H} NMR spectrum, two doublet resonances are observed at 222.7 and 258.7 ppm with ²J(³¹P,³¹P)=136 Hz (Figure 2, cf. 222.3 and 258.4 with ²J(³¹P,³¹P)=141 Hz for 2Dmp).⁴² Single-crystal X-ray structure analysis of 2Mtp revealed a monoclinic space group P2₁/n with Z=4. As depicted in Figure 1, both terphenyl-substituents form a pocket, in which the planar 5-membered heterocycle sits. Both P−N bond lengths are in the typical range of a highly polarized bond featuring only a small amount of double bond character (d(P1−N1)=1.725(2) and d(P2−N1)=1.654(2), cf. Σr_cov(P−N)=1.82 Å and Σr_cov(P=N)=1.62 Å).⁵⁶

Scheme 4

Synthesis of 2Mtp and generation of housane species 4Mtp upon irradiation with red light.

Figure 1

Molecular structures of 2Mtp and 3pTB in the single crystal (123 K). Colour code: grey=carbon; white=hydrogen; blue=nitrogen; orange=phosphorus. Disorder of the tBu group omitted for clarity. Selected bond lengths (Å) and angles (°): 2Mtp: P1−N1 1.725(2), P2−N1 1.654(2), P1−P2 2.9430(8); C1-N3 1.286(3); N1−P2−C1 92.83(9); N1−P1−N2 94.37(9); N1−P2−P1−C1 178.8(2).3pTB: P1−N1 1.748(1), P2−N1 1.752(1), P1−P2 2.6374(6), C1-N3 1.277(2), P1−N1−P2 97.80(7), P1−N2−C1 112.2(1); N1−P1−N2−C1 43.6(1).

Photoisomerization of 2Mtp

With blue biradical 2Mtp in hand, we investigated the photoisomerization process affording the colourless housane-type [2.1.0] bicyclic isomer 4Mtp. NMR spectra under irradiation were recorded at 101.27 MHz, using our previously published setup, which was adopted from a setup published by the Gschwind group using a fibre-coupled laser diode.^{45, 57} The ³¹P{¹H} NMR spectra of the blue solution were first recorded in the dark showing only signals of the biradical species 2Mtp. Continuous irradiation with a red laser diode (638 nm, 500 mA) for approximately two minutes, led to full conversion to the colourless housane-type species 4Mtp (two doublets at −63.5 and −126.5 ppm) as depicted in Figure 2.

Figure 2

³¹P{¹H} NMR spectrum of the biradical 2Mtp recorded at 25 °C in the dark (bottom) and the housane species 4Mtp under continuous irradiation (top).

To investigate the conversion of 4Mtp back to the biradical species 2Mtp, temperature and time variable ³¹P{¹H} NMR experiments were carried out. Once the laser diode was turned off, the thermal reverse reaction was traced in situ by recording ³¹P{¹H} NMR spectra (Figures S23–S26). According to these experiments, the thermal reverse reaction was found to be a first-order reaction with a half-life of about 7 min at ambient temperature (k=1.42(2)×10⁻³ s⁻¹), leading to quantitative recovery of 2Mtp (Figure S22, cf. 2Dmp/4Dmp: T_1/2=7 min, k=1.73(3)×10⁻³ s⁻¹).⁴⁵ The activation barrier was determined utilising Eyring theory by re-determination of the rate constants at several temperatures, giving a Gibbs free energy of activation Δ^≠G⁻=87(4) kJ mol⁻¹ (Figure S27, cf. 2Dmp/4Dmp: 88(4) kJ mol⁻¹).⁴⁵ This means that the introduction of a tBu group has no measurable effect on the equilibrium between biradical 2R and housane 4R (R=Mtp, Dmp).

Reaction with phenyl isonitriles featuring hydrogen atoms in 2- and 6-position

In further experiments, we used aryl-substituted isonitriles with either electron-withdrawing or electron-donating groups attached to the aryl system to study the electronic influence on housane formation (see above). Since the synthesis of the corresponding isonitriles is easier (see Supporting Information) when no methyl groups are present in the 2- and 6-positions of the aryl substituent, we chose the corresponding aryl-substituted isonitriles. For reasons of comparability we started the investigations by using p-tert-butylphenyl isonitrile (pTB-NC, Scheme 3) to see what influence the two missing methyl groups in the 2- and 6-position (compared to Mtp-NC) would have. This led to the discovery of a new reaction channel.

Synthesis of azaphospholes (5R)

The reactions of biradical 1 with pTB-NC were performed exactly as described above for the reaction with Mtp-NC, but with completely different results: Upon treatment of biradical 1 with pTB-NC, in fact a variety of products (1⋅CNtBu, 3pTB, 4pTB and 5Benz_tBu; cf. Scheme 1 and Scheme 5) could be observed in solution according to ³¹P NMR studies, depending on the stoichiometry and irradiation. In particular, the formation of 5Benz_tBu was somewhat surprising to us, as it represents a new reaction channel, i. e., C−H bond activation, which was facilitated by the missing ortho-Me groups of the organic aryl-substituted isonitrile. Compounds 3pTB (Figure 1) and 5Benz_tBu (Figure 3, top) could be isolated and fully characterized, depending on the reaction conditions.

Scheme 5

Synthesis of 5Benz_tBu when an exact 1 : 1 ratio is used.

Figure 3

Molecular structure of different azaphospholes in the single crystal. Selected bond lengths (Å) and angles (°): 5Benz_tBu: N1−P1 1.539(2), N2−P1 1.683(2), P2−C1 1.721(2), N3-C1 1.364(3), N2−C1 1.438(3); N1−P1−N2 103.0(1), C1−P2−C2 88.4(1). 5Benz_NMe₂: N1−P1 1.539(6); N2−P1 1.710(7), P2−C1 1.725(5); N3-C1 1.363(6); N2−C1 1.427(8); N1−P1−N2 101.7(3), C1−P2−C2 88.4(2). 5_2Naph: N1−P1 1.550(2); N2−P1 1.699(2), P2−C1 1.722(2); N3-C1 1.360(3), N2−C1 1.444(3); N1−P1−N2 101.4(1), C1−P2−C2 88.4(1). 5_1Naph: N1−P1 1.542(2); N2−P1 1.683(2), P2−C1 1.717(3); N3-C1 1.368(4), N2−C1 1.435(3); N1−P1−N2 103.6(1), C1−P2−C2 88.3(1).

To further investigate the formation of the different (intermediate) products, the reaction of 1 with one and two equivalents of pTB-NC was carried out in the dark as well as under irradiation with red laser light (see below). After addition of one equivalent of pTB-NC to a solution of biradical 1 in benzene in the dark, a change of colour from red to brown occurred within 15 min (Scheme 5). The reaction can be easily followed by ³¹P NMR spectroscopy and after about 48 h, a nearly complete conversion to a new compound was observed, which could be isolated in 84 % yield (Table S3). After recrystallisation from cyclopentane yellow single crystals were obtained. Single crystal X-ray structural analysis clearly showed the presence of the benzo-azaphosphole 5Benz_tBu (Figure 3, top). The ³¹P{¹H} NMR spectrum of 5Benz_tBu shows two broad singlets with chemical shifts of 69.7 and 272.9 ppm. During the reaction, the formal di-addition product 3pTB with a chemical shift of 172.7 and 209.7 ppm (two broad doublets and a coupling constant of ²J(³¹P, ³¹P)=34 Hz) could be observed by in situ ³¹P NMR spectroscopy, which, however, slowly disappeared at the expense of 5Benz_tBu (Figure 4). Only when the isonitrile was completely consumed, 3pTB slowly converted to the final product 5Benz_tBu under release of one equivalent of pTB-NC, which then reacted with the remaining, unreacted 1. In contrast, in case of the reaction of 1 with Mtp-NC, only the biradical 2Mtp was observed and isolated, but not 3Mtp (see above). For a targeted synthesis of the intermediate 3pTB, biradical 1 was treated with two equivalents of pTB-NC, resulting in the formation of a yellow solution within 15 min from which colourless crystals were obtained overnight (η=80 %). Single crystal structure elucidation revealed unequivocally the presence of the desired product, the diadduct 3pTB (Scheme 1, Figure 1 bottom). Apparently, in the presence of an excess of isonitrile, the formation of 5Benz_tBu is suppressed, as the formation of 3pTB is much faster, in agreement with ³¹P NMR studies.

Figure 4

³¹P{¹H} NMR spectra of the conversion of 3pTB with one equivalent of biradical 1 to form 5Benz_tBu as stable product in C₆D₆ at t=0 h (blue), t=24 h (black) and t=48 h (red).

NMR spectroscopy under irradiation

Finally, to get further insight into the rather complex equilibrium chemistry, we added an excess of isonitrile to 1 at very low temperatures under irradiation with red light (for details see Supporting Information Table S3, Figure 5). Interestingly, at −40 °C in the dark, immediately after addition of R-NC to 1, we observed the very labile intermediate 1⋅CNR (at 301.6 and 305.5 ppm, R=p-tert-butyl-phenyl, Scheme 1). With the onset of irradiation (638 nm, 500 mA) accompanied by slow warming, the concentration of 1⋅CNR decreased rapidly while the concentration of 3pTB increased. At −10 °C, only traces of 1⋅CNR were observed, while the housane 4pTB appeared simultaneously (−66.5 and −127.3 ppm, cf. −63.4 and −129.2 ppm for 4Dmp).⁴⁵ The formation of the housane 4pTB is a clear indication of the in situ formation of the five-membered cyclic biradical 2pTB.^{42, 45} Between −10 and +25 °C, only 3pTB and 4pTB were now present, but neither biradical 2pTB nor azaphosphole 5Benz_tBu, which is the thermodynamically favoured product (see below). When the irradiation was switched off, 4pTB disappeared immediately and only 3pTB was observed. From these NMR experiments we have learned that the formation of 5Benz_tBu is prevented either by adduct formation (3pTB) in the presence of an excess of isonitrile or by housane formation upon irradiation.

Figure 5

³¹P{¹H} NMR spectra of reaction of 1 with an excess of pTB-NC from −40 °C to 25 °C (spectra recorded under irradiation of the sample are indicated by asterisks).

Synthesis of Naphtho- and further Benzo-Azaphospholes

After understanding the rather complex reaction that produces 5Benz_tBu, we performed ³¹P NMR experiments to show whether other substitutions on the phenyl ring also lead to the formation of the azaphospholes (Scheme 1 and Scheme 3). For R′=Me₂N and H, the formation of the corresponding benzo-azaphosphole (phosphaindoles, 5Benz_NMe₂ and 5Benz_H) was also observed (Table S3) and for 5Benz_NMe₂ we succeeded in isolating single crystals (Figure 3).

To broaden the scope, naphthyl-substituted isonitriles were also studied to see if they show the same C−H activation chemistry. Therefore, biradical 1 was first treated with 2-naphthyl isonitrile in benzene at a 1 : 1 ratio in the dark (Scheme 6). After 24 h of stirring and recrystallisation from a saturated benzene solution, yellow crystals could be isolated (yield 87 %), which were investigated by single-crystal structural elucidation that revealed the presence of 5_2Naph as desired (Figure 3). Like 5Benz_tBu, 5_2Naph can be synthesized in large quantities, has a high melting point (195 °C, cf. 5Benz_tBu 202 °C), is readily soluble in benzene and stable for long periods in a sealed glass tube. As expected, two broad ³¹P signals were detected at δ=69.8 (CPC) and 273.6 ppm (NPN, Table S3). Interestingly, the reaction of biradical 1 with 2-naphtylisonitrile selectively led only to the one product; no other isomers were observed. A very similar situation is found when 1-naphthyl isonitrile is reacted with biradical 1. When using 1-naphthyl isonitrile for the reaction with 1 (1 : 1 ratio, Scheme 6), full conversion leads to the exclusive formation of 5_1Naph, which can be isolated in 80 % yield as yellow single crystals (Figure 3) with a melting point of 228 °C. In the ³¹P NMR spectrum two signals at 76.3 and 275.2 ppm were observed (for comparison see Table S3). Again, only one isomer was observed, indicating a highly selective reaction as found for 5_2Naph (see above).

Scheme 6

Synthesis of 5_1Naph and 5_2Naph.

Structure elucidation

The molecular structures of 2Mtp, 3pTB, 5Benz_R’ (R’=tBu, NMe₂) and 5Naph in the crystal are depicted in Figure 1 and Figure 3. When the isonitrile is inserted into the 4-membered ring of biradical 1, a planar 5-membered ring is formed as exemplified by 2Mtp. When a second equivalent of isonitrile is added, as for example in 3pTB the bridging occurs along the P1−P2 axis of the 5-membered ring by formal adduct formation (angles N1−P1−N2−C1 43.6(1)°).^{8, 26, 42, 44, 45} The P−N distances are all in the range of strongly polarized single bonds (also see discussion below). Probably the most striking structural motif of the species of type 5 (the azaphosphole ring) is that it is nearly planar, as are the condensed rings (maximum deviation from planarity <6°). The exocyclic N(Ter)PN-Ter unit is almost perpendicular to the phosphole ring system, creating a pocket formed by the bulky terphenyl substituents, into which the heterocycle fits very well. The N1−P1 bonds (1.539–1.550 Å; Σr_cov(P−N)=1.82 Å and Σr_cov(P=N)=1.62 Å)⁵⁶ are rather short and are in the range of a double bond, whereas the P1−N2 distances (1.683–1.709 Å) are significantly longer and in the range of a polarized P−N single bond. The bond lengths within the azaphosphole ring are all in the range of partial double bonds (e. g., C1−P2: 1.721–1.724; P2−C2 1.780–1.781; cf. Σr_cov(C=P)=1.69 Å),⁵⁶ as expected for aromatics (see below).

Computational studies

To better understand the results of the NMR experiments discussed above, we calculated both thermodynamic and kinetic data at the DLPNO-CCSD(T)/def2-TZVP//PBE−D3/def2-TZVP level of theory.^58-63 As illustrated in Figure 6, all additions of isonitrile to 1 are exergonic. Irrespective of the amount of isonitrile used, the formation of 2pTB is thermodynamically less favorable than the (partial) formation of the adduct 3pTB. For example, with one equivalent of isonitrile (n=1), the formation of 0.5 equiv. of 3pTB and 0.5 equiv. of unreacted 1 is favored by about 20.4 kJ mol⁻¹. This explains why we could not observe 2pTB even when using only one equivalent of isonitrile.

Figure 6

Gibbs free energy of the reaction of n RNCwith 1 with R=(p-tBu)C₆H₄ at the DLPNO-CCSD(T)/def2-TZVP//PBE−D3/def2-TZVP level of theory (c°=1 mol/L).

Consequently, using two equivalents of isonitrile leads to quantitative formation of 3pTB. The final product 5Benz_tBu is the thermodynamically most stable product, in agreement with experimental observation. To understand its relatively slow formation, the reaction pathway from the biradical 2pTB to the azaphosphole 5Benz_tBu was investigated, since we assume that the biradical should be the active species for C−H activation (Figure 7). The activation process begins with an electrophilic attack of a phosphorus atom at the ortho-carbon atom of the aromatic substituent. This leads to the transient intermediate Int1, with an activation barrier of 107 kJ mol⁻¹. The proton migration from the aromatic ring to the phosphorus atom is almost barrier-free (TS2). Further proton migration steps follow along the reaction pathway (TS3→Int3→TS4) until the final product 5Benz_tBu is formed in an overall exergonic step. The relatively high barrier of TS4 for the proton migration^{64, 65} that occurs here can probably be overcome through tunnelling as indicated by the transition temperature T_c=33.6 °C (according to an imaginary frequency of 1339 i cm⁻¹ at the barrier top).^66-68

Figure 7

Mechanism for the reaction from 2pTB to 5Benz_tBu at the DLPNO-CCSD(T)/def2-TZVP//PBE−D3/def2-TZVP level of theory.

Since we assumed that one of the driving forces for the formation of azaphosphole 5 is its larger aromaticity as opposed to the five-membered biradical 2, NICS(1)_zz values (Nucleus-Independent Chemical Shifts) and magnetically induced ring current susceptibilities, which represent common parameters to describe aromaticity, were calculated for the model compounds 2H and 5H (Figure 8, Table 1).^70-77 The net induced ring current (13.2/12.2 nAT⁻¹ for 6- and 5-membered rings, resp.) and NICS(1)_zz values (−30.2/−28.3 ppm) of 5H are in the range of other related aromatic ring systems, such as indole and azadi-phosphaindane-1,3-diyl, indicating that compound 5H possesses significant aromatic character. In contrast, values for 2H (which was previously investigated by us) differ notably and are significantly smaller (3.5 nAT⁻¹, −7.2 ppm), implying that 2H is best regarded as a non-aromatic compound.^{69, 70} In Figure 8 the current density susceptibility of the model system 5H is depicted as streamline representation. The diatropic π ring current is clearly visible above (and below) the ring system of the azaphosphole. Similar representations of the model systems listed in Table 1 may be found in the Supporting Information (Figure S35).

Figure 8

Streamline plot of the current density susceptibility in a plane placed 1 a.u. above the molecular plane of model system 5H. The diatropic π ring currents circle around the system counter clockwise.

Table 1. Net induced currents and NICS(1)_zz values of model systems 2H, 5H and further reference molecules.^{69, 70} Except for 2H the values are assigned to the five-membered(⑤) and six-membered ring system (⑥). For further details see Supporting Information.

	indole	C6H4P2NH	2H	5H
Net induced current [nAT^−1]	13.1 (⑥)	11.2 (⑥)	3.5	13.2 (⑥)
	12.1 (⑤)	13.5 (⑤)		12.2 (⑤)
NICS(1)zz [ppm]	−30.6 (⑥)	−24.9 (⑥)	−7.2	−30.2 (⑥)
	−30.3 (⑤)	−31.1 (⑤)		−28.3 (⑤)

Conclusion

A new route for the synthesis of azaphospholes starting from four-membered biradicals of the type [⋅P(μ-N-Ter)₂P⋅] (1) and aryl-substituted isonitriles was developed. It was shown that two different reaction channels exist in the reaction of biradical 1 with aryl-substituted isonitriles, depending on the substitution on the aryl. When methyl groups are present in the 2- and 6-positions on the aryl, an insertion into the 4-ring occurs in the first step, resulting in the formation of a five-membered planar biradical [⋅P(NTer)₂C(R)P⋅] (2), and a bridging addition in the second step, which leads to a cage compound, a hetero-bicyclo[2.1.1]hexane. When the 2- and 6-positions are not blocked by a methyl group, C−H activation occurs, which is highly selective. Again, in the first step of the reaction, an insertion of the isonitrile into a P−N bond of the biradical occurs, leading to the formation of a five-membered biradical 2 as found for the methyl-substituted aryls. However, now the ortho-C−H bond of the aryl substituent of the isonitrile can be activated by the five-membered biradical in an electrophilic substitution reaction, resulting in a condensed azaphosphole ring after further proton migration. This electrophilic attack is relatively slow due to high activation barriers, but is highly exergonic. However, when the biradical is blocked, for example by adduct formation with excess of isonitrile or by irradiation with red light triggering subsequent housane formation, the formation of an azaphosphole is suppressed. Hence, the aromatic substitution reaction can be controlled both by stoichiometry and light. Further investigations into the use of other isonitriles as building blocks for the synthesis of azaphospholes are part of ongoing research.

Experimental Section

All manipulations were carried out under oxygen- and moisture free conditions under argon atmosphere using standard Schlenk or glovebox techniques. The reported reaction products are mostly sensitive toward oxygen and moisture and need to be handled carefully to prevent decomposition reactions. All starting materials were produced using literature procedures as stated in the Supporting Information. The reactants and solvents from commercial sources were dried and purified. If not stated otherwise, experiments and crystallisation attempts were carried out at ambient temperature (25 °C). The removal of solvents in vacuo was carried out at 1×10⁻³ mbar and at 25 °C if not stated otherwise. Further information on experimental procedures, data acquisition and processing, purification of starting materials and solvents, and on computational investigations as well as a full set of analytical data for each compound and crystallographic information can be found in the Supporting Information.

Synthesis of [⋅P(NTer)₂C(Mtp)P⋅] (2Mtp): In an argon filled dry box one equivalent of [TerNP]₂ (1) (0.22 g, 0.31 mmol) and one equivalent of MtpNC (0.057 g, 0.31 mmol) were combined in a vial and benzene (5 mL) was added. An immediate change of colour from orange to dark blue was observed. After stirring the solution for 15 min the solvent was left to evaporate directly in the dry box over a period of 2–3 days. Yield: 0.23 g (0.25 mmol, 82 %). Mp: 155 °C (decomposition). EA: calc. (found) in %: C 81.03 (80.06), H 7.47 (7.08), N 4.65 (4.56). ³¹P{¹H} NMR (298 K, C₆D₆, 202.46 MHz): δ=222.7 (d, 1 P, ²J(³¹P,³¹P)=136 Hz, NPC); 258.7 (d, 1 P, ²J(³¹P,³¹P)=136 Hz, NPN). ¹H NMR (298 K, C₆D₆, 500.13 MHz): δ=1.28 (s, 9H, tBu-CH₃); 1.72 (s, 6H, Mes o-CH₃); 1.74 (s, 6H, Mes o-CH₃); 1.95 (s, 12H, Mes o-CH₃); 2.27 (s, 6H, CH₃); 2.29 (s, 6H, CH₃); 2.30 (s, 6H, CH₃); 6.72–7.10 (m, 16H, Ph-CH). ¹³C{¹H} NMR (THF-d₈, 298 K, 125.77 MHz): δ=19.3 (s, CH₃); 19.4 (s, CH₃); 21.3 (s, CH₃); 21.4 (s, CH₃); 21.6 (s, CH₃); 21.6 (s, CH₃); 21.7 (s, CH₃); 21.8 (s, CH₃); 32.2 (s, C-(CH₃)₃); 34.5 (s, C-(CH₃)₃); 125.6 (s, PhCH); 128.3 (s, PhCH); 128.4 (s, PhCH); 129.2 (s, PhCH); 129.3 (s, PhC); 129.5 (s, PhCH); 130.3 (s, PhCH); 132.1 (s, PhCH); 132.9 (s, PhCH); 136.4 (s, PhC); 137.0 (s, PhC); 137.4 (s, PhC); 137.6 (s, PhC); 137.7 (s, PhC); 138.0 (s, PhC); 138.2 (s, PhC); 139.8 (s, PhC); 141.0 (d, PhC, J=4 Hz); 143.7 (s, PhC); 149.2 (s, PhC); 149.2 (s, PhC).

Synthesis of TerN(CN(H)PC₆H₄tBu)PNTer (5Benz_tBu): [TerNP]₂ (1) (0.184 g, 0.257 mmol) was dissolved in benzene (5 mL). Afterwards a solution of 4-tert-butylphenyl isocyanide in benzene (1.85 mL, 0.139 mol/L, 0.257 mmol) was added at ambient temperature, whereupon an immediate change of colour from red to brown (over a period of 15 min) occurred. After stirring for two hours and storage of the reaction mixture for 48 h (at 25 °C) the solvent was removed in vacuo (1×10⁻³ mbar, 60 °C, water bath) yielding the corresponding azaphosphole Yield: 0.189 g (0.216 mmol, 84 %). Mp: 202.4 °C. EA: calc. (found) in %: C 80.88 (80.33), H 7.25 (7.58), N 4.80 (4.02). ³¹P{¹H} NMR (C₆D₆, 298 K, 121.5 MHz): δ=69.7 (br. s, 1P, CPC); 272.9 (br. s, 1P, NPN). ¹H NMR (C₆D₆, 300 K, 250 MHz): δ=1.38 (s, 9H, tBu-CH₃); 1.76 (s, 12H, o-Ter-CH₃); 1.93 (br. s, 12H, o-Ter-CH₃); 2.20 (s, 6H, p-Ter-CH₃); 2.32 (s, 6H, p-Ter-CH₃); 6.60–7.03 (m, 17H, Ph-H); 7.31 (s, 1H, NH). ¹³C{¹H} NMR (C₆D₆, 298 K, 62.9 MHz): δ=20.9 (s, Ter-CH₃); 21.6 (s, Ter-CH₃); 21.8 (s, Ter-CH₃); 22.1 (br. s, Ter-CH₃); 32.3 (s, tBu-CH₃); 35.0 (s, tBu-C); 112.9 (s, PhCH); 122.7 (s, PhCH); 123.2 (s, PhCH); 124.9 (s, PhCH); 128.7 (s, PhCH); 129.0 (s, PhCH); 131.6 (s, PhCH); 131.9 (s, PhCH); 132.0 (s, PhCH); 136.4 (s, PhC); 137.0 (s, PhC); 137.6 (s, PhC); 137.8 (m, PhC); 138.4 (s, PhC); 139.0 (s, PhC); 139.1 (s, PhC); 142.0 (s, PhC); 142.2 (s, PhC); 142.7 (s, PhC); 143.5(s, PhC); 143.9 (s, PhC).

Synthesis of [P(μ-NTer)]₂[tBuC₆H₄NC]₂ (3pTB): [TerNP]₂ (1) (0.178 g, 0.248 mmol) was dissolved in benzene (5 mL). Afterwards a solution of 4-tert-butylphenyl isocyanide in benzene (2.19 mL, 0.226 mol/L, 0.496 mmol) was added at ambient temperature, whereupon an immediate change of colour from red to brown ongoing to yellow (over a period of 15 min) occurred. After stirring for two hours the solvent was removed in vacuo (1×10⁻³ mbar, 60 °C, water bath) yielding the adduct compound. Yield: 0.205 g (0.198 mmol, 80 %). Mp: 195.7 °C. EA: calc. (found) in %: C 81.21 (80.21), H 7.40 (6.81), N 5.41 (5.45). ³¹P{¹H} NMR (C₆D₆, 298 K, 121.5 MHz): δ=172.7 (br. d, ²J(³¹P, ³¹P)=34 Hz, 1P, NPC); 209.7 (br. d, ²J(³¹P, ³¹P)=34 Hz, 1P, NPN). ¹H NMR (C₆D₆, 298 K, 300 MHz): δ=1.16 (s, 9H, tBu-H); 1.37 (s, 9H, tBu-H); 1.65–2.55 (m, 36H, Ter-CH₃); 6.50 (m, 2H, CH) ); 6.54 (m, 1H, CH); 6.67 (d, 2H, CH); 6.76 (m, 4H, CH); 6.87 (m, 3H, CH); 6.99 (s, 1H, CH); 7.05 (m, 2H, CH); 7.17 (m, 5H, CH); 7.27 (m, 2H, CH). ¹³C{¹H} NMR (C₆D₆, 298 K, 75.47 MHz): δ=21.4 (s, CH₃); 21.7 (s, CH₃); 21.9 (s, CH₃); 21.9 (s, CH₃); 22.1 (s, CH₃); 22.3 (s, CH₃); 22.5 (s, CH₃); 22.6 (s, CH₃); 31.7 (s, tBu-CH₃); 32.1 (s, tBu-CH₃); 34.7 (s, tBu-C); 34.8 (s, tBu-C); 122.1 (s, PhCH); 122.4 (s, PhCH); 123.2 (s, PhCH); 126.0 (s, PhCH); 126.2 (s, PhCH); 128.1 (s, PhCH); 128.3 (s, PhC); 128.5 (s, PhC); 128.7 (s, PhCH); 128.9 (s, PhCH); 129.2 (s, PhCH); 129.4 (s, PhCH); 129.5 (s, PhCH); 130.0 (s, PhCH); 130.2 (s, PhCH); 131.9 (s, PhCH); 136.4 (s, PhC); 136.5 (s, PhC); 136.6 (s, PhC); 137.4 (s, PhC); 137.8 (s, PhC); 137.9 (s, PhC); 138.3 (s, PhC); 138.8 (s, PhC); 141.4 (s, PhC); 142.9 (s, PhC); 145.8 (s, PhC); 149.5 (s, PhC); 150.4 (s, PhC); 150.4 (s, PhC).

Synthesis of TerN(CPN(H)Naph)PNTer (5_2Naph): [TerNP]₂ (1) (0.18 g, 0.25 mmol) was dissolved in benzene (5 mL) together with 2-naphtylisocyanide (0.038 g, 0.25 mmol) whereupon an immediate change of colour from red to brown and back to red (over a period of 5 min) occurred. After stirring for 24 h the solvent was removed and the residue dried in vacuo (1×10⁻³ mbar, 60 °C, water bath) yielding the product as a yellow/orange coloured powder. Yield: 0.19 g (0.22 mmol, 87 %). Mp: 194.6 °C. EA: calc. (found) in %: C 81.45 (80.64), H 6.60 (6.37), N 4.83 (4.75). ³¹P{¹H} NMR (C₆D₆, 298 K, 121.5 MHz): δ=69.8 (br. s, 1P, CPC); 273.6 (br. s, 1P, NPN). ¹H NMR (C₆D₆, 300 K, 500 MHz): δ=1.73 (s, 12H, o-Ter-CH₃); 1.92 (br. s, 12H, o-Ter-CH₃); 2.20 (s, 6H, p-Ter-CH₃); 2.25 (s, 6H, p-Ter-CH₃); 6.68 (s, 4H, Ph-H); 6.77–7.01 (m, 10H, Ph-H); 7.32 (m, 1H, Ph-H); 7.40–7.49 (m, 2H, Ph-H); 7.57 (s, 1H, NH); 7.78 (d, 1H, ³J(¹H, ¹H)=8 Hz, Ph-H); 8.50 (d, 1H, ³J(¹H, ¹H)=8 Hz, Ph-H). ¹³C{¹H} NMR (C₆D₆, 298 K, 125.8 MHz): δ=21.0 (s, Ter-CH₃); 21.6 (s, Ter-CH₃); 21.7 (s, Ter-CH₃); 114.1 (s, PhCH); 123.2 (s, PhCH); 124.8 (s, PhCH); 125.5 (s, PhCH); 125.7 (s, PhCH); 125.8 (s, PhCH); 126.9 (s, PhCH); 127.1 (s, PhCH); 128.7 (s, PhCH); 128.9 (br. s, PhC); 129.0 (s, PhCH); 129.4 (s, PhCH); 129.5 (s, PhCH); 131.6 (s, PhCH); 132.0 (m, PhC); 136.3 (s, PhC); 136.5 (s, PhC); 136.7 (s, PhC); 136.9 (s, PhC); 137.0 (s, PhC); 137.3 (m, PhC); 137.6 (s, PhC); 138.3 (s, PhC); 138.7 (s, PhC).

Synthesis of TerN(CPN(H)Naph)PNTer (5_1Naph): [TerNP]₂ (1) (0.18 g, 0.25 mmol) was dissolved in benzene (5 mL). Afterwards a solution of 1-naphtylisocyanide in benzene (0.30 mL, 0.83 mol/L, 0.25 mmol) was added at ambient temperature, whereupon an immediate change of colour from red to brown and back to red (over a period of 5 min) occurred. After stirring for 2 h the solvent was removed and the residue dried in vacuo (1×10⁻³ mbar, 60 °C, water bath) yielding the product as a yellow/orange coloured powder. Yield: 0.17 g (0.20 mmol, 80 %). Mp: 228.1 °C. EA: calc. (found) in %: C 81.45 (81.49), H 6.60 (6.64), N 4.83 (4.93). ³¹P{¹H} NMR (C₆D₆, 298 K, 121.5 MHz): δ=76.3 (br. s, 1P, CPC); 275.2 (br. s, 1P, NPN). ¹H NMR (C₆D₆, 298 K, 300 MHz): δ=1.68 (s, 12H, o-Ter-CH₃); 1.86–2.17 (m, 12H, o-Ter-CH₃); 2.29 (m, 12H, p-Ter-CH₃); 6.69 (s, 4H, Ph-H); 6.83 (m, 8H, Ph-H); 6.93 (m, 2H, Ph-H); 7.29 (m, 2H, Naph-H); 7.38 (m, 2H, Naph-H); 7.74 (m, 1H, Naph-H); 7.87 (dd, 1H, ³J(¹H, ¹H)=9 Hz, Naph-H); 8.24 (s, 1H, NH). ¹³C{¹H} NMR (C₆D₆, 298 K, 125.8 MHz): δ=21.1 (s, Ter-CH₃); 21.6 (s, Ter-CH₃); 21.8 (s, Ter-CH₃); 120.5 (s, PhCH); 120.6 (s, PhCH); 121.0 (s, PhCH); 122.8 (s, PhCH); 123.5 (s, PhCH); 125.0 (s, PhCH); 125.6 (s, PhCH); 127.1 (s, PhCH); 127.3 (s, PhCH); 128.7 (s, PhCH); 128.7 (s, PhCH); 129.0 (s, PhCH); 129.2 (m, PhCH); 129.4 (s, PhCH); 131.6 (s, PhCH); 132.3 (m, PhC); 136.2 (s, PhC); 136.5 (s, PhC); 136.9 (s, PhC); 137.3 (m, PhC); 137.6 (s, PhC); 137.9 (s, PhC); 138.4 (s, PhC); 138.9 (m, PhC).

Deposition Number(s) 2204253 (for 2Mtp), 2157618 (for 5Benz_tBu), 2157619 (for 3pTB), 2204254 (for 5_2Naph), 2244148 (for 5_1Naph), 2169321 (for 5Benz_NMe₂) contain(s) the supplementary crystallographic data for this paper. These data are provided free of charge by the joint Cambridge Crystallographic Data Centre and Fachinformationszentrum Karlsruhe Access Structures service.

Acknowledgments

Y. P. wishes to thank the University of Rostock for financial support via the PhD Scholarship Program. This research was supported by the Deutsche Forschungsgemeinschaft (DFG, SCHU 1170/12-2). The University of Rostock and Malte Willert are acknowledged for access to the cluster computer and support with software installations. We also wish to thank Johanna Winkler for her investigations on the NMe₂ substituted compounds and Henrik Beer for his help with the Mtp derivatives. Open Access funding enabled and organized by Projekt DEAL.

Conflict of interest

The authors declare no conflict of interest.

Open Research

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.