Mass Spectrometry of 1 , 2 , 5-Oxadiazole N-Oxide Derivatives . Use of Deuterated Analogues in Fragmentation Pattern Studies

Reportamos neste trabalho o estudo sistemático de fragmentação dos derivados de N-óxidos de 1,2,5-oxadiazóis por espectroscopia de massa, usando análogos marcados com deutério para identificar algumas fragmentações críticas. Foi confirmada a perda neutra de CH 2 O a partir do N 2 -óxido de 3hidroximetil-4-fenil-1,2,5-oxidiazol, usando o análogo mono-deuterado. A perda de OH, a partir do oxigênio do N-óxido, por um rearranjo β-H e δ-H, foi claramente verificada a partir do N 2 -óxido de 3-(4-metilpiperazina-1-metil)-4-fenil-1,2,5-oxidiazol, usando-se o analogo tetra-deuterado adequado. O isômero N-óxido e análogos desoxigenados foram também usados para confirmar a participação do fragmento óxido no processo de defragmentação.


Introduction
As part of an ongoing research program on the chemistry and biological characterization of N-oxide containing molecules, a number of 1,2,5-oxadiazole N-oxide, benzo[1,2-c]1,2,5-oxadiazole N-oxide, quinoxaline N,N'dioxide, and 1,2,4-triazine N-oxide derivatives were synthesized and evaluated against different biological targets. 1 In the course of our synthetic chemical approach we developed 1,2,5-oxadiazole N-oxide derivatives as potential herbicides, 2 bioreductive compounds, [3][4][5][6] and antitrypanosomal drugs, 7,8 using previously described synthetic methods. 9During the structural elucidation of these derivatives, we were interested in knowing about the presence and the exact location of the N-oxide functionality.Simple and readily available spectroscopic techniques (e.g.IR and 1 H NMR) are not able to unambiguously characterize the presence of an N-oxide group.For example, the characteristic IR absorptions of =N + -O -(1300-1200 cm -1 and 970-950 cm -1 ) are not considered a conclusive proof because these bands can overlap with the fingerprint bands. 10Similarly, 1 H NMR spectroscopy can not afford important information about the =N + -O -system, because these products are heterocycles devoid of protons.However, more recently 2D-HETCOR experiments (sequences of HMQC for one-bond correlation and sequences of HMBC for long distance/carbon correlation) provide us very important data about the presence and position of the =N + -O -moiety.
In this context, the initial aim of the present study was to characterize the fragmentation pattern in mass spectrometry of the 1,2,5-oxadiazole N-oxide derivatives as an alternative structural determination technique.On the other hand, this work describes the elucidation of the fragmentation patterns of these compounds through the use of deuterated derivatives, 11 N-oxide isomer and deoxygenated (without N-oxide) analogues. 12

Results and Discussion
The spectra for selected derivatives 1-13 (Figure 1) revealed different fragmentation patterns, resulting from different side chains of the 1,2,5-oxadiazole heterocycle at C-3.The most relevant mass spectral data for derivatives 1-13 are presented in Table 1 and Figure 2 rationalizes important fragmentation pathways.
The molecular ion was detected in all cases.However, the abundances were very low for derivatives 4-6, which contain a residue proven to undergo retro-Diels-Alderfragmentation (morpholine, thiomorpholine, and methylpiperazine). 13The same was observed for derivative 12, that bears an aliphatic side chain on the semicarbazone moiety.The [M-16u] +• ion, corresponding to an oxygen loss, was observed in all cases as a relatively small peak.However, the peak corresponding to m/z [M-17u] + , became increasingly important for derivatives 4-7 and 11 as result of an OH • loss.This kind of fragmentation has been previously described for similar moieties, e.g., a NO 2 group losses OH • in o-nitrotoluenes and o-nitroanilines, 13 or for other adequately substituted N-oxide heterocycles. 14,15In these compounds, two different rearrangement processes could explain this radical loss, β-H and/or δ-H rearrangement (Figure 3).The high abundance of this fragment ion for derivatives 4-7 (non-aromatic cyclic amine derivatives) compared to fragment ion abundance's for derivatives 8-10 (phenylamino derivatives) made us to think that a δ-H rearrangement was the main process in this kind of structures.With a conventional EI/MS equipment it is not possible to study this fact and others, e.g., to determine whether the [M-30u] + fragment ion in derivative 1 corresponds to a NO loss (as for the other derivatives) or a CH 2 O loss from the hydroxymethyl substituent in 3-position of the 1,2,5oxadiazole heterocycle is due to an initial γ-H transference, which produces the stable neutral-product formaldehyde (Figure 4).
In order to explain these critical fragmentations in mass spectrometry, using a EI/MS equipment, we centered our The results are the averages for three independent experiments.c The "-" denotes that the fragment ion was not observed.efforts on the synthesis of deuterium analogues of some selected furoxan derivatives. 16,17Initially, we tried to prepare di-deuterium analogue, at the benzylic position, of derivative 6.The synthesis of the di-deuterium chloride 3 was attempted by trying to exchange the "acidic" benzylic-protons using a biphasic system NaOD-D 2 O/ CDCl 3 at different temperatures (room temperature to reflux for 24 h) (as shown in Scheme 1).Unfortunately, 1 H-NMR monitoring of the reaction mixture showed no significant exchange under these conditions.When CD 3 O -Na + / CD 3 OD was used, exchange took place, but the ether 14 was obtained as the result of the substitution by the powerful nucleophile deuterated methoxide (Scheme 1).These synthetic problems led us to undertake the preparation of the mono-deuterium analogue of derivative 6, at the benzylic position, with 3-d 1 as starting material (Scheme 1).Using the aldehyde 2, the mono-deuterium alcohol 1-d 1 was produced with more than 95% of  deuterium incorporation verified by 1 5).The product 1-d 1 was then transformed in a good yield into the corresponding chloride 3-d 1 using thionyl chloride (Scheme 1).
Finally, the amine derivative 6-d 1 was obtained by the reaction between chloride 3-d 1 and N-methylpiperazine (Scheme 2).The molecular structure of derivative 6-d 1 has been determined by X-ray diffraction methods (Figure 6). 18,19In addition, we prepared 6-d 4 , the tetra-deuterium analogue in the 2,6-position of piperazine ring, through reaction between the chloride 3 and the heterocyclic amine  Scheme 1. 17 (Scheme 2).This amine was deuterated via the nitrosamine 15, prepared following the Ravindran et al. procedure, 20 which was converted into the tetra-deuterium analogue 16 following the Keefer-Fodor methodology. 21o transform nitrosamine 16 into amine 17 we tried a procedure by Kano et al. (reduction with NaBH 4 :TiCl 4 (2:1) in diglyme) with bad results. 22The reduction process did not occur and probably, a complex between the methylpiperazine nitrosamine and TiCl 4 was obtained.The use of H 2 in Raney-Nickel at room temperature and atmospheric pressure led to compound 17 in an adequate yield. 23The deuterium incorporation in compound 16 was more than 95% (by 1 H-NMR analysis).
To know how the N-oxide group position affect on the furoxans' mass spectrometry behavior, we prepared the Noxide-positional isomer 3-i.This compound was obtained via the alcohol 1-i, which was prepared following the Gasco et al. methodology (Scheme 3). 10 To study the fragmentation patron of the molecule "N-oxide-free", we prepared the deoxygenated analogues 1-deoxy and 7-deoxy using Zn in NH 4 Cl solution as the reduction reagent (Scheme 3). 6,24These products were clearly confirmed through HETCOR experiments (HMQC and HMBC).
Mass spectrometry was carried out on all the analogues developed.The most characteristic peaks in the mass spectrum (EI/MS) for derivatives 1-d 1 , 1-deoxy, 3-d 1 , 3-i, 6-d 1 , 6-d 4 and 7-deoxy, together with those for parent compounds 1, 3, 6 and 7, are presented in Table 2.The results clearly indicate that the fragmentation process of compound 1-d 1 does not occur exclusively through a NO • loss, the [M-31u] +• ion (not present in the parent compound 1) probably arise from a CDHO loss.On the other hand, comparing the relative abundance of the M +• and [M-30u] +• ions in compounds 1-deoxy and 7-deoxy, 9.9% and 1.6% respectively, we could notice that the [M-30u] +• ion was more abundant in the first one, this fact is in accordance with the neutral CH 2 O loss fragmentation process in derivative 1.
The mass spectrum of positional isomer 3-i did not show the [M-17u] +• and [M-31u] +• ions, this fact could be indicative that the parent compound 3 losses OH .and HNO as a β-H participation.Deuterium labeling of compound 6 indicated that the [M-17u] +• ion of this derivative was the result of β-H and δ-H rearrangements (see Table 2).While derivative 6 showed an [M-17u] +• ion abundance of 50.4%, derivatives mono-and tetra-deuterated (6-d 1 and 6-d 4 ), that could present the β-H and the δ-H rearrangement phenomena, showed the abundances of the corresponding  Table 2. Abundance of the critical fragment ions in the corresponding 1,2,5-oxadiazole N-oxide analogues EI mass spectrum a MS experiments were performed using a Shimadzu MS QP 1100 EX equipment, with EI at 70 eV, with direct insertion probe, the ion source temperature 250 ºC and the mass range was 40-500 amu.b The results are the averages for three independent experiments.c The "-" denotes that the fragment ion was not observed.d At 20 eV, ion source temperature 150 ºC.These conditions were used in order to increase the abundance of the peak of the molecular ion, other conditions and other products were not studied.

Experimental
All starting materials were commercially available research-grade chemicals and used without further purification.All solvents were dried and distilled prior to use.All the reactions were carried out in a nitrogen atmosphere.The typical work-up included washing with brine and drying the organic layer with sodium sulphate.Compounds 1-3, 1-i, 1-deoxy, 7, 7-deoxy, 15 and 16 were prepared as previously described. 4,6,9,20,21Elemental analyses were obtained from vacuum-dried samples (over phosphorous pentoxide, 24 h at room temperature) and performed on a Fisons EA 1108 CHNS-O analyzer, and were within ± 0.4% of theoretical values.Infrared spectra were recorded on a Perkin Elmer 1310 apparatus, using potassium bromide tablets; the frequencies are expressed in cm -1 . 1 H-NMR spectra and HETCOR experiments were recorded on a Bruker DPX-400 (at 400 MHz and 100 MHz) instrument, with tetramethylsilane as the internal reference; the chemical shifts are reported in ppm.MS experiments were performed using the Shimadzu MS QP 1100 EX equipment, with EI at 20 or 70 eV, with direct insertion probe, the ion source was set a 150 ºC or 250 ºC and the mass range was 40-500 amu.

Figure 3 .
Figure 3. Postulated mechanism for the OH • loss in compounds 4, 5, 6 and 7 (Note: we are gratefully thank to one of the referee for suggesting the structure of the final product of the δ-H rearrangement).

Figure 4 .
Figure 4. Postulated mechanism for the NO • and CH 2 O loss in compounds 1.

Figure 6 .
Figure 6.Molecular plot of derivative 6-d 1 .The ORTEP drawing of the molecule shows the labeling of the non-H(D) atoms and their displacement ellipsoids at 30% probability level.

Table 1 .
Abundance of the most characteristic fragment ions in the corresponding 1,2,5-oxadiazole N-oxide derivatives EI mass spectrum a Analytical conditions for EI/MS: direct injection, ion source temperature 250 ºC, energy 70 eV.b