Crystal structure and photoreactive behaviour of N,N-diisopropyl(p-phenylphenyl)glyoxylamide

The photoreactive behaviour of the title compound, C20H23NO2, was investigated. Solid-state photoreaction did not occur in the solid-state but it occurred in acetonitrile solution.


Chemical context
The solid-state photochemistry of N,N-dialkyl--oxoamides has been studied in relation to penicillin chemistry (Aoyama et al., 1979). The amides undergo Norrish type II cyclization giving -lactams (Aoyama et al., 1978). The achiral molecule N,N-diisopropylphenylglyoxylamide 1a crystallizes in the chiral space group P2 1 2 1 2 1 and is transformed to the optically active -lactam derivative 2a upon UV light irradiation ( Fig. 1; Toda et al., 1987;Sekine et al., 1989). N,N-Diisopropyl(mchloro or m-methyl or o-methylphenyl)glyoxylamides 1b and 1c also form chiral crystals, and photoirradiation in the solid state gives optically active -lactam derivatives 2b and 2c, respectively (Toda & Miyamoto, 1993;Hashizume et al., 1995Hashizume et al., , 1996Hashizume et al., , 1998. However, N,N-diisopropyl(p-chloro or o-chloro or p-methylphenyl)glyoxylamide 1b and 1c do not form chiral crystals, and their photoirradiation in the solid state gives racemic -lactam derivatives 2b and 2c, respectively. Therefore, we synthesized the novel title compound 1d having a phenyl group and investigated whether optically active -lactam derivative 2d could be obtained by photoreaction. It was found that 1d formed a chiral crystal in the chiral space group P2 1 2 1 2 1 , but photoreaction did not proceed in the solid  state. However, photoreaction of 1d in acetonitrile solution proceeded to give racemic 3-(p-phenylphenyl)-3-hydroxy-Nisopropyl-4,4-dimethylazetidin-2-one 2d in 26% yield. In this study, although 1d formed a chiral crystal, the reason why the photoreaction product of 1d in the solid state was not obtained was clarified by single-crystal X-ray structural analysis, UV spectroscopy and time-dependent density functional theory (TDDFT) calculations. Table 1 summarizes intra and intermolecular hydrogen bonds observed in the title compound. The phenyl rings in the biphenyl group are coplanar with the carbonyl group (C7 O1). The torsion angles C2-C1-C7-O1 and C3-C4-C15-C16 are 7.8 (3) and À0.4 (2) , respectively, and the torsion angles O1-C7-C8-O2 and C7-C8-N1-C9 are 97.1 (2) and À3.9 (2) , respectively (Fig. 2). The corresponding torsion angles in 1a are 88.0 (4) and À5.1 (4) . In order for the Norrish-Yang reaction to take place, the reacting atoms in the molecular structure must be in close proximity. The Yang cyclization of -oxoamides to -lactams starts with abstraction of the -hydrogen (with respect to the benzylic carbonyl) by the benzylic carbonyl oxygen in the excited state. In the title compound, there are two -hydrogen atoms (H5 on C9 and H12 on C12). The distances between the carbonyl oxygen atom O1 and the respective -hydrogen atoms H5 and H12 are 2.65 and 5.01 Å . The former interatomic distance is within the ideal value of up to about 2.7 Å , at which photoreaction can proceed in the crystal (Konieczny et al., 2018). Moreover, the distance between the reacting C7 and C9 carbon atoms is 2.840 (2) Å , which is in the range of ideal values of up to about 3.2 Å . The corresponding distances are 2.78 (4) and 2.871 (4) Å in 1a. As shown in Fig. 3, the geometries of the oxoamide moiety of 1d and 1a are almost the same. Despite satisfying the geometry and distance requirements for the photoreaction, the corresponding -lactam was not detected in the solid-state reaction. From the UV spectrum of 1d, it is considered that the biphenyl group of 1d absorbs ultraviolet light preventing the solid-state reaction (Fig. 4). In other words, the photocyclization reaction does not proceed in the solid state for at least 300 h because the irradiated UV light is absorbed by the -* transition of the biphenyl group.

Figure 2
The molecular structure of 1d. Displacement ellipsoids for non-H atoms are drawn at the 50% probability level.

Figure 3
Overlay of molecules 1a (in red) and 1d (in blue).
etries of 1a and 1d were obtained from XRD data. Hydrogen atoms were optimized at the B3LYP/6-311G(d,p) level (Becke, 1993). The UV-vis spectra of 1a and 1d were calculated by the time-dependent density functional theory [TDDFT, B3LYP/6-311G(d,p)] method. In the calculated UV-vis spectra, there were two weak peaks at 254 and 362 nm for 1a, and there was an intense and broad peak at 310 nm for 1d. The calculated spectra were similar to the experimental spectra ( Fig. 5). For 1a, the peak at 254 nm corresponds to the -* transition of the Ph group, while that at the longer wavelength of 362 nm is due to n-* transitions of the carbonyl groups. For 1d, the adsorption peak at 376 nm was assigned to n-* transitions of carbonyl groups. A very weak absorption peak was observed around 370 nm in the experimental spectrum. A mercury lamp has an intense emission at 365 nm, such that the photoreaction for 1a proceeds rapidly in the solid state. In contrast, the large and broad absorption prevents the solid-state photoreaction for 1d. Since the molecules can move freely in solution, light irradiation for 60 h was uniformly performed, and it seemed that the reaction proceeded slightly. It has been reported that an oxoamide derivative having a naphthyl group slows down the photoreaction (Natarajan et al., 2005). The relationship between photoreactivity and irradiation wavelength is under investigation.

Synthesis and crystallization
The title compound was prepared according to a reported method (Toda et al., 1987;Sekine et al., 1989), i.e., chlorination of 2-oxo-2-(4-phenylphenyl)acetic acid with thionyl chloride followed by reaction with N,N-diisopropylamine. Thus, to an ice-cooled solution of N,N-diisopropylamine (16 mL, 0.11 mol) in dry diethyl ether (45 mL) was added a solution of 4-phenylbenzoylformyl chloride (13.8 g, 0.0564 mol) in dry diethyl ether (45 mL), and the reaction mixture was stirred for 10 h at room temperature. After filtration of N,N-diisopropylammonium chloride, the filtrate was washed with dilute HCl and aqueous NaHCO 3 and dried over MgSO 4 . The crude product was purified by silica gel column chromatography (toluene:ethyl acetate = 9:1) and recrystallized from toluene to give 1d as colorless prisms (

Refinement
Crystal data, data collection and structure refinement details are summarized in Table 2. All H atoms were positioned in geometrically calculated positions (C-H = 0.95-0.98 Å ) and refined using a riding model with U iso (H) = 1.2U eq (C) and 1.5Ueq(C-methyl). The Flack parameter x is 0.1 (4) as shown in   and photoreactive behaviour of N,N-diisopropyl(p-phenyl-

Special details
Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes.