Different hydrogen-bonded chains in the crystal structures of three alkyl N-[(E)-1-(2-benzylidene-1-methylhydrazinyl)-3-hydroxy-1-oxopropan-2-yl]carbamates

Three closely related methylated hydrazine carbamates show different hydrogen-bonding patterns, although they all result in chains.

In general, the tertiary butyl compounds form simple methylated products as described here, whereas the benzyl compounds lead to cyclized oxazolidin-2-one products (Noguiera et al., 2013). However, compound (I) described herein has not cyclized. As described below, compound (III) has undergone an unexpected racemization during the methylation step. The acidity of the -hydrogen atom in serine ISSN 2056-9890 derivatives has been variously reported (e.g., Blaskovich & Lajoie, 1993;Kovacs et al., 1984), and apparently can result in racemization in the presence of even a very weak base such as the carbonate ion. Similar racemizations have been observed in the cyclized oxazolidin-2-one products (Noguiera et al., 2015).

Structural commentary
The molecular structure of (I) is shown in Fig. 1, which confirms that methylation has occurred at N2 but no cyclization to an oxazolidin-2-one has occurred (Noguiera et al., 2015). Compound (I) crystallizes in a chiral space group but its absolute structure was indeterminate in the present experiment and C10 was assumed to have an S configuration to match the corresponding atom in the l-serine starting material. The atoms of the C14 benzene ring show notably larger displacement ellipsoids than the rest of the molecule, but attempts to model this as disorder did not lead to a significant improvement in fit. Atom N2 is statistically planar (bondangle sum = 360 ), which implies sp 2 hybridization for this atom. The C9-N2 bond length of 1.358 (6)Å is typical of an amide and the N1-N2 bond length of 1.374 (5) is shorter than the reference value of 1.40 Å for a nominal N(sp 2 )-N(sp 2 ) single bond. This suggests at least some electronic conjugation over the almost planar C7/N1/N2/C9/O1 grouping (r.m.s. deviation = 0.010 Å ): the C1 benzene ring is twisted by 6.1 (2) with respect to these atoms. The C7-N1-N2-C8 torsion angle of À1.9 (6) shows that the carbon atoms are almost eclipsed with respect to the N-N bond whereas the C9-C10-C11-O2 torsion angle of À50.9 (5) indicates a gauche conformation about the C10-C11 bond. The C9-C10-N3-H3A torsion angle is 38 and the separation between H2A (bonded to O2) and H3A is 2.5 Å .
The molecular structure of (II) can be seen in Fig. 2: again the methylation of N2 has occurred as expected. Because the absolute structure was indeterminate, the configuration of C10 (S) was assumed to be the same as that of the corresponding atom in the l-serine starting material. In terms of the C7/N1/ N2/C9/O1 grouping in (II), the C9-N2 and N1-N2 bond lengths are 1.385 (6) and 1.388 (5) Å , respectively, which are both notably longer than the corresponding bonds in (I), and the r.m.s. deviation from planarity of 0.049 Å for these five atoms is also larger than the corresponding value for (I). The dihedral angle between C7/N1/N2/C9/O1 and the C1-benzene ring in (II) is 10.5 (3) . The C7-N1-N2-C8 torsion angle is 1.2 (7) and the C9-C10-C11-O2 torsion angle is À47.4 (6) , which are similar to the equivalent data for (I). The C9-C10-N3-H3 torsion angle in (II) is 30 and the separation between H2A and H3 is 2.7 Å . These values are evidently sufficiently different from the corresponding data for (I) to lead to a different hydrogen-bonding pattern in the crystal (see below). The molecular structure of (I) showing 50% displacement ellipsoids.

Figure 2
The molecular structure of (II) showing 50% displacement ellipsoids.

Supramolecular features
In the extended structure of (I), the molecules are linked by short O2-H2AÁ Á ÁO4 i (i = 1 + x, y, z) and much longer N3-H3AÁ Á ÁO4 i hydrogen bonds (Table 1, Fig. 4) to the same acceptor oxygen atom, generating [100] chains, with adjacent molecules related by simple translation in the a-axis direction. An unusual R 1 2 (7) loop arises from these hydrogen bonds; alternately, this could be described as combined C(7) O-HÁ Á ÁO and C(4) N-HÁ Á ÁO chains. A pair of weak C-HÁ Á Á interactions are also observed but there is no aromaticstacking (shortest centroid-centroid separation > 4.7 Å ).
The extended structure of (II) also features [100] chains ( Fig. 5) with adjacent molecules related by translation, but in this case the molecules are only linked by C(7) O2-H2AÁ Á ÁO4 i (i = 1 + x, y, z) hydrogen bonds (Table 2) with almost the same local geometry as seen in (I). The N3-H3 grouping in (II) is twisted far enough away from O4 i to not form an intermolecular hydrogen bond (H3Á Á ÁO4 i = 3.2 Å ), but instead forms an intramolecular link to O1. A very long intermolecular C-HÁ Á ÁN interaction is observed but there is Fragment of a [100] hydrogen-bonded chain in the crystal of (I). Symmetry code: (i) 1 + x, y, z. All C-bound H atoms are omitted for clarity. Table 1 Hydrogen-bond geometry (Å , ) for (I).

Figure 5
Fragment of a [100] hydrogen-bonded chain in the crystal of (II). Symmetry code: (i) 1 + x, y, z. All C-bound H atoms are omitted for clarity.
nostacking in (II), as the shortest centroid-centroid separation is greater than 5.3 Å . The packing in the centrosymmetric structure of (III) leads to [010] chains ( Fig. 6) with adjacent molecules related by the 2 1 screw axis, so that the C1-benzene ring is 'flipped' from one side of the chain to the other in adjacent molecules. As noted above, the hydroxyl group is disordered over two orientations. The hydrogen bond from the major orientation of O2A-H2A is still a bond to O4 i (Table 3), where i = 1 À x, y À 1 2 , 1 2 À z. The minor disorder component (O2B-H2B) forms an O-HÁ Á ÁO hydrogen bond in the opposite chain direction to O1 ii (ii = 1 À x, y + 1 2 , 1 2 À z): O1 also accepts an intramolecular N-HÁ Á ÁO hydrogen bond, as seen in (II). Once again, no aromaticstacking is observed in the crystal of (III), as the minimum centroid-centroid separation is greater than 4.6 Å .

Database survey
There are no -OCONHCH(CH 2 OH)CON(CH 3 )N CHfragments reported in Version 5.36 of the Cambridge Structural Database (Groom & Allen, 2014) but there are 14 unmethylated -OCONHCH(CH 2 OH)CONHN CHgroupings with different substituents at each end of the fragment, all of which have been reported by us in the last few years (Howie et al., 2011 and references therein). All of these materials crystallize in chiral space groups.

Synthesis and crystallization
Potassium carbonate (1.76 mmol) was added to a solution of the appropriate (E)-(S)-ROCONHCH(CH 2 OH)-CONHN CH-benzene compound (Noguiera et al., 2013) in acetone (10 ml) and the reaction mixture was vigorously stirred at room temperature for 5 minutes, before adding methyl iodide (1.80 mmol). The reaction mixture was stirred at 323 K for 24-48 h and the solvent removed by rotary

Figure 6
Fragment of an [010] hydrogen-bonded chain in the crystal of (III). Both disorder components of the OH group are shown. Symmetry codes: (i) 1 À x, y À 1 2 , 1 2 À z; (ii) 1 À x, y + 1 2 , 1 evaporation. The residue was subjected to column chromatography on silica gel, using a chloroform:methanol (100 ! 95%) gradient. The colourless crystals used in the structure determinations were recrystallized from ethanol solution at room temperature. For further details and spectroscopic data, see: Noguiera et al. (2013).

Refinement
Crystal data, data collection and structure refinement details are summarized in Table 4. The crystal of (I) gave a poor diffraction pattern and indexing initially established a large triclinic unit cell [a = 9.512 (12) ]. An atomic model could be developed in space group P1 with Z = 6, but a PLATON (Spek, 2009) symmetry check indicated that the smaller monoclinic cell reported above was more appropriate and the unit cell transformed by the matrix (À 1 3 À 1 3 0 / À 2 3 1 3 0 / 0 0 À1). It is notable that the aromatic rings of the benzyl groups of all six molecules in the triclinic supercell showed a high degree of thermal motion. The transformation to monoclinic symmetry resulted in a rather low data completion percentage of 92%, but we consider that the refinement is satisfactory and a good geometrical precision results. For each structure, the O-and N-bound H atoms were located in difference maps, repositioned in idealized locations and refined as riding atoms [H1N was freely refined in structure (III)]. The C-bound H atoms were placed geometrically (C-H = 0.95-1.00 Å ) and refined as riding atoms. The constraint U iso (H) = 1.2U eq (carrier) or 1.5U eq (methyl carrier) was applied in all cases. The H atoms of the hydroxyl groups were allowed to rotate about their C-O bond (SHELXL HFIX 83 instruction with O-H = 0.84 Å and C-O-H = 109.5 ) to best fit the electron density. The methyl groups were allowed to rotate, but not to tip, to best fit the electron density (AFIX 137 instruction).

Computing details
For all compounds, data collection: CrystalClear (Rigaku, 2012); cell refinement: CrystalClear (Rigaku, 2012); data reduction: CrystalClear (Rigaku, 2012); program(s) used to solve structure: SHELXS97 (Sheldrick, 2008); program(s) used to refine structure: SHELXL97 (Sheldrick, 2008); molecular graphics: ORTEP-3 for Windows (Farrugia, 2012); software used to prepare material for publication: publCIF (Westrip, 2010).   (13) Special details Geometry. All e.s.d.'s (except the e.s.d. in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell e.s.d.'s are taken into account individually in the estimation of e.s.d.'s in distances, angles and torsion angles; correlations between e.s.d.'s in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell e.s.d.'s is used for estimating e.s.d.'s involving l.s. planes. 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 R-factors(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.

Hydrogen-bond geometry (Å, º)
Cg2 is the centroid of the C14-C19 ring.   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 R-factors(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.

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