A convenient and eco-friendly cerium(III) chloride-catalysed synthesis of methoxime derivatives of aromatic aldehydes and ketones

The use of CeCl3·7H2O as an efficient and eco-friendly promoter for the convenient synthesis of methoximes derived from aromatic aldehydes and ketones, is reported. The transformations entail the use of equimolar amounts of MeONH2·HCl and NaOAc in EtOH at 50°C, and no special precautions are needed with regard to the presence of oxygen. The scope and limitations of the transformation were studied and a reaction mechanism was proposed.

This functional group is also widespread among approved crop-protecting agents, as exemplified by the synthetic strobilurins trifloxastrobin and fluoxastrobin, for conferring them valuable selectivity in their biocidal properties [10].
Furthermore, oxime ethers are frequently present in the patent literature, being broadly used as intermediates in the chemical industry and in synthetic organic chemistry [11], especially for electrocyclization [12,13] and more recently for   ortho-functionalization reactions of inactive ArC-H bonds [14,15], being also useful as aldehyde and ketone protecting groups [16] or as chromatographic derivatization agents [17]. In addition, oxime ethers have been employed as precursors of other functional groups, such as alkoxy-amides/lactams [18], nitriles [19] and amidines [20]. The methoximes are a special group among the oxime ethers. Conventional methods towards their synthesis [21] mostly rely on the reaction between methoxylamine and aromatic or aliphatic aldehydes or ketones [22]. (Hydro)alcoholic media, pyridine [22,23] and, occasionally, chlorinated compounds are employed as solvents [24]. Refluxing conditions are generally required for improved yields.
Some protocols include the use of molecular sieves, Na 2 SO 4 or MgSO 4 as water scavengers, to drive the reaction to a more rapid completion and/or to avoid heating [25,26]. Usual yields exceed 70%. The methoximation reagent is available as MeONH 2 ·HCl, which is seldom employed alone; usually, a base (pyridine, Et 3 N, NaOAc, NaHCO 3 , Na 2 CO 3 or K 2 CO 3 ) is added [27,28], which enables the use of milder reaction conditions (less time, lower temperature).

Results and discussion
At the outset of the study, based on previous experience [40] and on the literature regarding nonpromoted methoximations, the development of an optimal protocol was sought, with acetophenone (1a) as the chosen model starting material, in EtOH as solvent. Trial and error experiments, which included heating from room temperature to reflux, revealed that incubating the reaction at 50°C provided the most convenient conversions in short times and under mild conditions. On the other hand, progressive reduction of the excess of methoxylamine (range 2.5-1.1 equiv.) exposed that 1.5 equiv. of each, MeONH 2 ·HCl and the base, were required for optimal results. Lowering the amount of MeONH 2 ·HCl to 1.2 equiv. or below afforded either unfinished transformations or long reaction times. Hence, warming the reaction at 50°C in the presence of 1.5 equiv. each methoxylamine hydrochloride and the base were set as the initial conditions for further improvements.
The transformations were monitored by GC-MS, employing anisole as internal standard, with the results detailed in table 1. This enabled us to observe that, in the absence of a promoter, the reaction afforded only 21% yield of the expected methoxime 2a after 40 min (entry 1) and required 12 h to reach completion. It was also evident that Ba(OAc) 2 ·H 2 O was ineffective under these conditions (entry 2), with results analogous to the non-promoted process.      Copper salts performed only slightly better; however, regardless of their different solubility and oxidation state, the outcome was similar. At 40 min CuI afforded only 30% yield of the product 2a (entry 3), whereas Cu(OAc) 2 ·H 2 O gave the methoxime in 31% yield (entry 4), both being found to be not viable to favour this transformation. Interestingly, an undisclosed copper salt supported on silica gel has been recently proposed as promoter for the preparation of oximes derived from aromatic aldehydes, in reactions taking 2-3 h to reach completion [53].
The use of La(NO 3 ) 3 .6H 2 O gave similar results (entry 5), while the tested iron salts furnished approximately 10% increment in the yields (entries 6 and 7); an analogous behaviour was observed in the reactions promoted by Mn(OAc) 3 ·2H 2 O and CdCl 2 ·xH 2 O, which resulted in 44% and 47% of 2a, respectively, at the 40 min check time (entries 8 and 9).
Interestingly, despite the importance of the counterions in Lewis acid-mediated carbonyl activation on the outcome of the thus triggered transformations [54], the performances of the reactions promoted by MnSO         the eco-friendly CeCl 3 ·7H 2 O is inexpensive, commercially available and easier to handle, it was selected for further optimization of the model reaction.
The nature of the reaction solvent was also optimized (table 2, entries 1-11), with the use of CeCl 3 ·7H 2 O as promoter in the presence of NaOAc. It was observed that the transformation did not proceed at all in PhMe, CHCl 3 , MeCN and dioxane (entries 1-4), whereas it afforded a meagre 30% yield of 2a in THF, after 40 min (entry 5).
Oppositely, alcoholic solvents (MeOH, EtOH, i PrOH, t BuOH) proved to be suitable media to achieve moderate-to-excellent yields of the product (64-89% after 20 min, entries 6-9). Probably, this is owing to the better solubility of all the reactants in alcoholic solvents and to their potential interaction with the promoter. Interestingly, it was shown that CeCl 3 forms a dimeric adduct with MeOH, [Ce 2 Cl 6 (MeOH) 8 ] that persists in solution and the corresponding ethanol adduct can also be prepared [58,59].
Hence, EtOH emerged from these experiments as the most advantageous solvent alternative (entry 9). It was also found that the reactions can also be carried out without special protection against oxygen. However, the absolute grade solvent proved to be more efficient than its mixtures (up to 20% v/v) with water (entries 9-11). In these cases, it was observed that the presence of H 2 O did not hinder the transformation, but it seemed to slightly lower the reaction rate.
On the other hand, the aptitude of mild bases (NaHCO 3 , K 2 CO 3 , K 2 HPO 4 and Et 3 N) other than NaOAc, to free the methoxime base was also evaluated (entries 12-15) in EtOH. However, despite their excellent performances, especially in the case of K 2 HPO 4 (99% at 30 min and 100% at 40 min, entry 14), none of them surpassed that of NaOAc.
In addition, when 0.15 M solutions of MeONH 2 ·HCl in EtOH (2.5 ml) were treated with NaOAc (0.90, 1.0 and 1.1 equiv.) and diluted with water (5.0 ml), they exhibited essentially the same pH values (4.83, 4.89 and 4.91, respectively), confirming the robustness of the method. Under these conditions Ce(III) is stable in solution and it has been shown that oximation reactions are slow and their rate has a maximum between pH 4 and 5 [60]. Accordingly, NaOAc was selected as the added base for further experiments.
Finally, the load of the promoter was analysed, in the range 2-8 mol% (entries 9, [16][17][18][19], observing that the product yield was quantitative at the 30 min checkpoint with loads of at least 5 mol% (entries 9, 18 and 19). However, while it was found that the reaction was complete in 40 min at a 4 mol% promoter level, it was also concluded that loads above 5 mol% did not produce any substantial improvement at the 20 min control time. Therefore, the latter level was chosen as the optimum.
The resulting protocol (MeONH 2 ·HCl and NaOAc (1.5 equiv. each) and 5 mol% CeCl 3 .7H 2 O in EtOH at 50°C) proved to be mild and respectful towards sensitive compounds, while remaining a highly discriminating condition against the non-catalysed process.
Next, the scope of the optimized methodology was explored, employing various aromatic ketones and aldehydes with different substituents and substitution patterns (table 3). In general, very goodto-excellent yields were obtained at 50°C with both, aromatic ketones (entries 1-12) and aldehydes (entries 13-18), usually taking place in short reaction times. Being more reactive, the best results were achieved with the latter ones, in which cases the transformations were also completed in comparatively shorter times.
Further analysis of the results revealed that no significant effects were detected owing to the presence of either electron withdrawing (entries 2, 6, 8, 17 and 18) or electron donating (entries 3-5, 7, 9 and 10) groups attached to different positions of the aromatic ring. However, the nitro derivatives of entries 6 and 8 reacted at a lower rate, taking longer times to reach completion.
The reaction conditions also proved to be compatible with ortho-substituents (entries 3, 8-10 and 14-18), without substantial loss of performance, except that compounds exhibiting bulky orthosubstituents (entries 8, 9 and 12) and ketones displaying ortho hydroxy groups (entries 9 and 10) were methoximated in good yields, at the expense of rather longer reaction periods.
In the case of 2-hydroxyketones, some starting material was recovered at the end of the reaction period. Presumably, this may be a result of the presence of a hydrogen bond between the phenol and the carbonyl moieties. The transformation was also viable in the presence of free phenols and free amines in different positions (entries 5, 6, 9, 10, 14 and 16) and took place with aromatic ketones other than acetophenones (entries 10-12).
The conjugate addition of hydroxylamine derivatives to α,β-unsaturated carbonyls is a serious side reaction in certain systems, leading sometimes to undesired products [49,50,66]. Lewis acids have been found to promote carbonyl activation, favouring this process [51,67,68]. Fortunately, however, no products arising from conjugate addition were observed in the experiments of entries 20 and 21, suggesting that the reaction conditions are mild enough to prevent this reaction, and that carbonyl methoximation is faster than the Michael addition. The methoxime product, being less reactive, is less likely to undergo a conjugate addition.
In addition, it was observed that the Ce(III)-promoted reaction was also successful with hydroxylamine. When acetophenone was used as substrate, 95% yield of the expected oxime 2w was obtained after 25 min (entry 23); in comparison, the non-catalysed process took over 140 min to reach completion under the same conditions.
The structures of the different products were assessed by their melting points, as well as by IR and NMR ( 1 H and 13 C) spectroscopy, being all in full agreement with their proposed structures and with the corresponding literature data. Not unexpectedly, in many cases they were obtained as mixtures of anti/syn (E/Z) isomeric compounds that could not be separated chromatographically.
The major products were assigned as the anti-isomers on the basis of comparative analysis of their 1 H NMR spectral data and the known tendency of acetophenone oximes to preferentially adopt the anti-configuration [69]. Furthermore, most of the methoximes are known compounds and were chosen with the purpose of comparing the performance of the proposed cerium(III)-promoted transformation with previous results (cf. electronic supplementary material).

Scheme 1.
Proposed reaction mechanism for the catalytic cycle of the CeCl 3 ·7H 2 O-promoted methoximation of aromatic aldehydes and ketones.
The reaction can be assumed to take place through a stepwise process. In the first stage, the oxophilic promoter coordinates with the carbonyl moiety of 1 to afford the activated intermediate i. This intermediate is more likely than the starting carbonyl derivative to undergo nucleophilic attack by the nitrogen of the methoxylamine and furnish intermediate ii in a second step.
In the next phase, a proton transfer within this intermediate would generate intermediate iii, which is prone to suffer dehydration, with concomitant deprotonation, release of the methoxime product 2 and regeneration of the promoter. Most probably, the steps from 1 to iii are readily reversible under the mild reaction conditions, whereas once formed compound 2, it is less likely that it could revert to the starting carbonyl derivative 1 under the same conditions, owing to the relative hydrolytic stability of the oxime ethers [70].

Conclusion
In conclusion, we have developed an expeditious and efficient CeCl 3 ·7H 2 O-based catalytic method for the methoximation of aromatic aldehydes and ketones under mild conditions and demonstrated that the system is also operative on their aliphatic counterparts and for the synthesis of acetophenone oximes. This efficacious reagent reduced substantially the reaction times, the amounts of MeONH 2 ·HCl and base required, and afforded very good-to-excellent product yields.
The transformation takes place with an eco-friendly promoter and a sustainable solvent. Further, it was observed that there is no need to employ the anhydrous reagent nor an anhydrous solvent; however, for shorter reaction times, the use of CeCl 3 ·7H 2 O in absolute EtOH is preferred.
This catalytic system proved to be robust, and capable to accept a wide variety of aldehydes and ketones. It is also tolerant to electron poor and electron rich substrates, as well as those with some steric demand; in the latter case, at the expense of longer reaction times. These are promising results in the field of synthesis of oximes, which suggest they will find wide use in multistep syntheses of more complex molecules.
Data accessibility. All data used in this research are included in figures and tables. The datasets supporting this article have been uploaded as part of the electronic supplementary material.