Thioureas have long been studied as useful synthons and bioactive materials [1], and their aminomethylation reactions have received attention [2]. For example, the reaction of the parent thiourea, H 2 N(C=S)NH 2 , with paraformaldehyde and piperidine has been shown to lead to both mono-and disubstituted materials, as illustrated in reaction (1) (Scheme 1) [2a].

The well-known Mannich reagent N,N,N',N'tetramethyldiaminomethane (Me 2 NCH 2 NMe 2 ) has also been shown to react with thiourea to form a similar bis-Me 2 NCH 2 -substituted product that is a useful reagent for further reaction with primary amines to form 1,3,5-triazinane-2-thiones [2]. However, this reagent needs the presence of metal catalysts, e.g. Cu

N,N-Dimethyl-1-trimethylsiloxymethanamine Me 3 • SiOCH 2 NMe 2 was first synthesized in 1981 by the Mironov's group [3] by the reaction of Me 2 NSiMe 3 with formaldehyde. The authors demonstrated its reactions with halosilanes R 3 SiX to form Me 2 NCH 2 X, with excess N,N-dimethyl-1-trimethylsilylmethan-amine to form diamines [reaction (3a)], and with hydroaminosilane BuNHSiMe 3 to form hexahydro-1,3,5-tributyl-1,3,5-triazine [reaction (3b)] [ We earlier reported the utility of N,N-dimethyl-1triethylsiloxymethanamine Et 3 SiOCH 2 NMe 2 (1) (an O-silylated hemiaminal) as a versatile and very efficient aminomethylating reagent that requires neither catalysts nor high temperatures [5][6][7]. Thus, compound 1 readily reacts with a range of EH, E = O, S, N, materials to form the corresponding E-CH 2 NMe 2 products in high yield. For the corresponding E = P chemistry, catalysts are required for efficient chemistry.

We now report the utility of 1 for aminomethylation of thioureas 2, 3, and 4, that results in new previously unknown materials and surprising chemistry (Scheme 3).

The reactions of compound 1 were studied in dichloromethane without any catalytic reagent and were fast and completed in a few hours at room temperature as determined by 13 C and 1 H NMR monitoring. The reactions of both N,N-dimethylthiourea (2) and 1,3-dihydroimidazole-2-thione (4) were unremarkable and led initially to monosubstitution in both cases to form compounds 5 and 6, respectively, reactions (4) and (5) (Scheme 4). The reaction of imidazole 4 continued in the presence of excess 1 to form disubstituted product 7 (both 6 and 7 are new compounds). Contrary to the chemistry of 4, we never observed a second substitution in the case of thiourea 2, regardless of the stoichiometry of the reaction. Presumably, the restricted geometry of the NH bonds in 4 accounts for this extra reactivity, although precisely how is outside the scope of our present study.

The single crystal structure of 1,3-bis(N,Ndimethylaminomethyl]-1,3-dihydro-2H-imidazole-2thione (7) is presented in Fig. 1. The molecule has a C 2 symmetry through the C=S bond, and the bond lengths and angles are as expected. The crystal structure is stabilized by hydrogen bonds (HBs) between N 2 and H 4B (2.850 Å) forming R 2 2 (6) motifs, and generating chains linked by this motif (Fig. 2). The 3-D architecture is generated by the motif C 1 2 (12) formed by the bidentate interaction of S 1 with two symmetrically equivalent H 4A (2.876 Å).

The reaction between 1 and N,N-diphenylthiourea (3) was more interesting and produced new chemistry as opposed to simply new products. By 1 H and 13 C NMR monitoring we could observe the rapid formation of a mono-aminomethylation product, PhNH(C=S)PhNCH 2 NMe 2 (8). However, all attempts to isolate this material resulted in the recovery of two major products, the asymmetrical thiourea PhNH(C=S)• NMe 2 (9) [8], and hexahydro-1,3,5-triphenyl-1,3,5triazine (10) (Scheme 5). This chemistry represents a new molecular rearrangement for such a structural unit.

If the initial reaction between compounds 1 and 3 is monitored after the reaction to form 8 has gone to completion, it can be seen that 8 is slowly converted to 9 and 10, even at room temperature. However, if the Scheme 3. initially formed solution of 8 is heated to 60°C, the transformation is complete within 40 min.

A plausible mechanism associated with this transformation, involving an intramolecular base substitution at the thiocarbonyl group by the strongly basic Me 2 N functionality, is outlined in Scheme 6.

Attempted isolation of triamine 8 using many solvent systems to recrystallize the solid gave a product that spectroscopically appeared to be pure. Sometimes from such attempts, in addition to 9 and 10, we isolated small amounts of a new aminomethanol, PhHN(C=S)NPhCH 2 OH (11), and could obtain a crystal structure of this unusual material (Fig. 3).

In the solid state, compound 11 forms an intramolecular HB with the sulfur atom as acceptor (2.629 Å), while forming chains through the O 1 -H•••N 1 HBs (2.209 Å). At the same time, the C-N and C=S bonds show no significant differentiation in comparison to 7, suggesting that the electronic factors are similar. The presence of the phenyl group reduces the conjugation of the nitrogen lone pair with the C-S system, as shown by a comparison of the structure in focus with that of the related compound [CH 3 HN(C=S)• NHCH 2 OH] (12) [9]. The C-S bond distance of 1.728(8) Å in 12 is considerably reduced compared to 1.6870 (11) and 1.678(3) Å for 7 and 11, respectively. Furthermore, in the case of 12 the intra-molecular HB is absent, indicating that it is probably the result of steric interactions due to the phenyl rings in compound 11.

We surmised that 11 was formed by the hydrolysis of 8 by residual water in our dried solvent systems and thus attempted a treatment of 8 with water. This experiment resulted in no identifiable products.

In another attempt to derivatize 8 we treated it with MeI hoping to obtain a quaternary ammonium salt. This reaction led only to the isolation of methyl N,N'-diphenylcarbamimidothioate (13). After several attempts to get suitable crystals, an acceptable structure was solved in terms of R 1 (Table 1), although with a high R 1 descriptor (Fig. 4). The structure data collected at 100 K showed 4 symmetrically independent mole- cules in an asymmetric unit cell for a total of 16 molecules in the unit cell.

Examination of the N 1 -C 1 [1.366( 9) Å] and N 2 -C 1 [1.288(9) Å] distances in compound 13 clearly reveals distinctions between the sp 3 -C-N and sp 2 -C-N bond lengths, whereas the C phenyl -N bond lengths are equivalent: 1.413(8) and 1.425(9) Å, respectively. These observations apply to all the 4 conformers in the asymmetric unit. With respect to the packing, the NH hydrogen forms HBs with the sp 2 -C-N nitrogen of a neighboring molecule, generating a chain of the type

It is of interest to compare the neutral molecule 13 with its HI salt 14, reported in [10]. The first striking difference is in the conformation adopted by the charged compound (Fig. 4), where the two phenyl groups appear to form a π-stacking motif, although the phenyl rings are divergent rather than parallel. Overall, the chemistry delineated in this article shows that 1 is an efficient, low energy, Me 2 NCH 2transfer agent to thioureas and requires no catalyst. Additionally, it is clear that a significant new chemistry is to be obtained by further investigation of such simple materials. Scheme 7 illustrates the overall chemistry of compound 8. Scheme 5. Chemical transformations in the reaction between 1 and PhNH(C=S)NHPh.

THF was distilled under nitrogen from benzophenone ketyl prior to use. Hexane, benzene, and toluene were dried over sodium metal and distilled before use. 1,3-Dimethylthiourea, 1,3-diphenylthiourea, and imidazole-2-thiol were purchased from Sigma-Aldrich. Et 3 SiOCH 2 NMe 2 was synthesized by the reported method [11]. The NMR spectra were recorded on 300 MHz Bruker spectrometer in CDCl 3 . The SC-XRD analysis of 7 was performed using a Bruker Venture Duo diffractometer with a micro-source and a Photon 200 detector, and the data for 11 were collected on a Bruker APEX CCD diffractometer. Both structures were solved using APEX3 crystallography software suite [12].

A 20 mL round-bottom flask was charged with 0.125 g (1.25 mmol) of imidazole-2-thiol and 0.48 g (0.25 mmol) of 1 in 10 mL of dichloromethane. The solution was stirred overnight at room temperature. Volatiles were then removed under vacuum to leave a brown solid, repeated recrystallization of which from a mixture of hexane and dichloromethane gave a 0.7 : 1.0 mixture of 1-(dimethylamino)imidazole-2-thiol (6) and 1,3-bis(dimethylamino)imidazole-2-thiol (7). Repeating the reaction of 1 with imidazole-2-thio in 1 : 2 molar ratio in dichloromethane gave a 1 : 2 mixture of compounds 6 and 7, but using a larger excess of 1 (4-fold) gave compound 7 as a single product (yield 85%).

Reaction of Et 3 SiOCH 2 NMe 2 with 1,3-diphenylthiourea. A 20 mL round-bottom flask was charged with 0.165 g (0.88 mmol) of Et 3 SiOCH 2 NMe 2 and 0.2 g (0.88 mmol) of 1,3-diphenylthiourea in 10 mL of dichloromethane. The solution was stirred overnight at room temperature, after which the solvent and triethylsilanol were removed under vacuum to obtain 1-[(dimethylamino)methyl]-1,2-diphenylthiourea Ph• NH(C=S)NPhCH 2 NMe 2 (8), yield 80%. This new material is thermally labile and undergoes slow transformation even at room temperature.

The reaction between 1 and 1,3-diphenyl thiourea in 2 : 1 molar ratio, too, resulted exclusive formation of 8; unexpectedly, no disubstituted product was obtained. Due to the limited thermal stability of 8, we could not perform its CHN analysis. 1

A Pyrex NMR tube was charged with 0.183 g (0.64 mmol) of compound 8 in 0.3 mL of CDCl 3 . The tube was sealed under vacuum and immersed in an oil bath maintained at 70°C. After 5 h, the 1 H, 29 Si and 13 C NMR monitoring showed a quantitative transformation of 8 to 1,1-dimethyl-3-phenylthiourea PhNH(C=S)NMe 2 (9) [8] and hexahydro-1,3,5-triphenyltriazine (PhNCH 2 ) 3 (10).

A 20 mL round-bottom flask was charged with a solution of 0.24 g (1.27 mmol) of 1 and 0.29 g (1.27 mmol) of 1,3-diphenylthiourea in 10 mL of dichloromethane. The solution was stirred overnight at room temperature. Excess of MeI (1 mL) was then added with a syringe to the reaction mixture. After 15 min of stirring, a white solid started to precipitate, and the reaction mixture was further stirred for 2 h. The solvent and triethylsilanol were removed under vacuum. The white solid was treated twice with 15 mL of hexane, the extract was filtered, and the solvent was removed to obtain methyl N,N-diphenylcarbamimidothioate PhN=C(SMe)NHPh (13), yield 62% yield, mp 106°C (107°C [13]).

Reaction of Et 3 SiOCH 2 NMe 2 with 1,3-dimethylurea. In a typical experiment, a 10 mL round-bottom flask was charged with 0.22 g (1.16 mmol) of Et 3 SiOCH 2 NMe 2 and 0.10 g (0.1.16 mmol) of 1,3-dimethyurea in 5 mL of dichloromethane. The reaction mixture was stirred overnight at room temperature. The solvent and triethylsilanol were removed under vacuum overnight to isolate 1-[(dimethylamino)methyl]-1,2-dimethylurea MeNH(C=O)NMeCH 2 NMe 2 as a colorless oil, yield 80% The material is thermally stable and does not undergo significant decomposition even on heating at 45°C for 2 days. 1 H NMR spectrum, δ, ppm: 2.13 s (6H, CH 3 ), 2.69 d (3H, CH 3 , J 4.8 Hz), 2.87 s (3H, CH 3 ), 3.51 s (2H, CH 2 ), 6.31 br.s (1H, NH). 13 C NMR spectrum, δ, ppm: 26.8 (Me), 35.6 (Me), 41.5 (N-Me 2 ), 73.9 (CH 2 ), 160.6 (C=O).

A similar reaction between Et 3 SiOCH 2 NMe 2 and 1,3-diphenylurea in different molar ratios in dichloromethane failed to yield any dimethylaminosubstituted product.