Tetramethyl 1,1,2-triphenyl-2H-1λ5-phosphole-2,3,4,5-tetracarboxylate

The title compound, C30H27O8P (1), was formed as one of two products {(1) and (2) [Krawczyk et al. (2010 ▶). Acta Cryst. E66 (cv2753)]} in the reaction of dimethyl acetylenedicarboxylate with triphenylphosphine. The molecule of (1) consists of a five-membered ring, in which the P atom is incorporated. One of the phenyl groups of the triphenylphosphine migrated to a vicinal C atom during the reaction. The five-membered ring of (1) is corrugated [r.m.s. deviation = 0.0719 (8) Å], whereas that in compound (2) is planar, the r.m.s. deviation being only 0.009 (2) Å.

The title compound, C 30 H 27 O 8 P (1), was formed as one of two products {(1) and (2) [Krawczyk et al. (2010). Acta Cryst. E66 (cv2753)]} in the reaction of dimethyl acetylenedicarboxylate with triphenylphosphine. The molecule of (1) consists of a five-membered ring, in which the P atom is incorporated. One of the phenyl groups of the triphenylphosphine migrated to a vicinal C atom during the reaction. The five-membered ring of (1) is corrugated [r.m.s. deviation = 0.0719 (8) Å ], whereas that in compound (2) is planar, the r.m.s. deviation being only 0.009 (2) Å .

Comment
Several derivatives of dimethylenesuccinic anhydride (fulgides) have been the subject of intensive studies due to their photochromic properties (Hadjoudis & Mavridis, 2004). As early as in 1893 Stobbe (Stobbe, 1893) discovered a very effective synthetic procedure leading to E,E-diarylfulgides. The double Stobbe condensation, after some minor modifications (Gordaliza et al., 1996) still remains the method of choice in the construction of fulgide-type compounds (Datta et al., 2001).
However, considering some disadvantages of this procedure, e.g. a need for the use of strong bases, which may cause resinification of some aldehydes, there have been continuous efforts towards development of alternative approaches. Especially the Wittig reaction (Maercker, 1965) between dialkyl bis[triphenylphosphoranylidene]succinates and the appropriate benzaldehydes appeared to be of particular value. The ylide component of the Wittig reaction seemed to be easily accessible by the condensation between triphenylphosphine and dialkyl acetylenedicarboxylate. However, this reaction, performed in diethyl ether, gave tetramethyl 1,1,2-triphenyl-2H-1λ 5 -phosphole-2,3,4,5-tetracarboxylate (1) and another adduct -trimethyl-3-methoxy-4-oxo-5-triphenylphosphoranylidenecyclopent-1-ene-1,2,3-tricarboxylate in 42% and 21% yield respectively (Shaw et al., 1967). We found, that when dry toluene was used as a solvent, and the reaction was performed at -78°C, (1) was formed in 63% yield, and the other adduct in 28% yield. In the present comunication we report on the crystal structure of compound (1). This structure was already proposed in 1971 (Waite et al., 1971) on the basis of spectral data. The crystal structure of the other adduct could also be determined via single-crystal diffraction (Krawczyk et al., 2010).
In compound (1) (Fig. 1) two acetyl groups at C2 and C4, respectively, are almost co-planar with the five-membered ring with the dihedral angle of 11.6 (1) and 8.47 (9)°, respectively, whereas the two remaining acetyl groups at C1 and C3 are strongly rotated from the ring plane (the dihedral angles of 67.(1) and 80.51 (8)°, respectively). The phenyl rings bonded to the phosphorous atoms in (1) have similar conformations to that observed at room temperature for the parent triphenylphosphine in both polymorphic structures (Spek, 1987;Thomas & Hamor, 1993) assuring the lowest repulsion of the neighboring fragments.

Refinement
H atoms were placed in calcluated positions and were included in the refinement with U iso (H) = 1.2U eq (C) [1.5 in the case of methyl groups H atoms]. Isotropic displacement parameters for hydrogen atoms bonded to either oxygen or nitrogen atoms were refined independently.  (1) showing the atomic labelling and 30% probability displacement ellipsoids.
Tetramethyl 1,1,2-triphenyl-2H-1λ 5 -phosphole-2,3,4,5-tetracarboxylate Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > σ(F 2 ) is used only for calculating Rfactors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.