Alcohol Oxidation Assisted by Molybdenum Hydrazonato Catalysts Employing Hydroperoxide Oxidants

: Molybdenum(VI) catalysts were obtained from methanol or acetonitrile by the reaction of [MoO 2 (C 5 H 7 O 2 ) 2 ] and isonicotinoyl- or nicotinoyl-based aroylhydrazones. Reactions in methanol resulted in the formation of the mononuclear complexes [MoO 2 (L 1–4 )(MeOH)] ( 1a – 4a ), while the ones in acetonitrile provided polynuclear complexes [MoO 2 (L 1–4 )] n ( 1 – 4 ) . Crystals of polynuclear compound, [MoO 2 (L 3 )] n · H 2 O ( 3 · H 2 O ), suitable for X-ray diffraction analysis were obtained by the solvothermal procedure at 110 ◦ C. Complexes were characterized by infrared spectroscopy (IR-ATR), nuclear magnetic resonance (NMR), elemental analysis (EA), and thermogravimetric analysis (TGA). The prepared catalysts were tested in alcohol oxidation reactions. Carveol, cyclohexanol, and butan-2-ol were investigated substrates. Because the alcohol oxidations are very challenging due to various possible pathways, the idea was to test different oxidants, H 2 O 2 , TBHP in water and decane, to optimize the researched catalytic system. presentation of the mononuclear [MoO 2 (L 1–4 )(MeOH)] complexes coordinated with hydrazonato ligands bearing hydroxyl group R , at position 3 or 4.


Introduction
Catalytic oxidations of alcohols to corresponding aldehydes and ketones are of great importance and interest [1,2]. Since the usual oxidation processes imply the use of KMnO 4 , CrO 3 , and halogenated solvents, which are often environmentally harmful, there is a definite need for developing more eco-friendlier and economically affordable catalytic procedures [3]. Hydrogen peroxide, H 2 O 2 , is often recommended oxidizing agent in that regard. Another option is tert-butyl hydroperoxide (THBP) due to its solubility and stability in organic solvents [4]. The by-product of the reaction, tert-butanol can be easily separated by distillation, converted into methyl tert-butyl ether, and employed as a gasoline additive.
Typically, carveol is available as a mixture of cisand transisomers [5]. Trans-carveol is an expensive ingredient of Valencia orange essential oil. The successful preparative method involved α-pinene oxide and zeolite catalyst [6]. The desired product, carvone takes place in the production of pharmaceuticals, fragrances, and flavours. It can be extracted from essential spearmint oils, but the great demand requires new chemical pathways for its production. For instance, it can be produced by the catalytic oxidation of limonene obtained from the orange peels. Literature reports carveol oxidation with hydrogen peroxide, [M 4 (H 2 O) 2 (PW 9 O 34 ) 2 ] n− (M = Co II , Mn II , Fe III , Co 4 (PW 9 ) 2 , Mn 4 (PW 9 ) 2 , and Fe 4 (PW 9 ) 2 , respectively) as catalysts [7]. The conversion was almost complete, but the selectivity towards carvone was around 50%. Further, carveol oxidation with THBP over the phthalocyanine complex FePcCl 16 immobilized on the mesoporous silica SBA-15, provided 75% of conversion and 40% selectivity towards carvone [8].
On the other hand, the majority of cyclohexanol and its mixtures are used as starting material in the synthesis of caprolactan and adipic acid, intermediates in the production of nylon. Not so far ago, direct production of adipic acid from cyclohexanone with oxygen or hydrogen peroxide was investigated [9,10]. Likewise, supported phosphotungstatic acid on silica-coated MgAl 2 O 4 nanoparticles with hydrogen peroxide showed great recovery possibilities for cyclohexanol oxidation [11], while selectivity towards the desired ketone with ruthenium pyridine-imine based complexes with N-methylmorpholine-Noxide (NMO) was in the range 82-97% [12]. NMO showed great potential for selective oxidation of alcohols under mild conditions. The supreme problems of cyclohexanol oxidation rise within the steric effect of cyclohexyl group and competing reactions as aromatization to phenol, dehydration to cyclohexene, and condensation of cyclohexanone to cyclohexenyl cyclohexanone.
In the present research, the main focus was on the carveol, cyclohexanol, and butan-2ol oxidation reactions, with the assistance of molybdenum complexes containing aroylhydrazonato ligands (Scheme 1). Similar catalytic systems proved to be efficient catalysts for the cyclooctene epoxidation reactions and the aim was to extend the research to alcohol oxidations. Since the alcohol oxidations are very challenging due to different possible pathways leading to a wide class of by-products, the idea was to test different oxidants to optimize the catalytic system. For that reason, H 2 O 2 , TBHP in water, and decane were chosen as oxidizing agents.
Whereas the compound 1•MeCN partially loses MeCN already upon standing at room temperature, crystals of 3•H2O are more stable and could be handled with less precaution. On the other hand, compounds 1a-4a with coordinated solvent molecules are even more stable. According to the thermal analysis, they lose the coordinated MeOH in the range 178-210 °C (1a), 206-224 °C (2a), 120-175 °C (3a) and 214-247 °C (4a). Following decomposition, the final residue is MoO3.
As expected, the shortest bonds in polymer 3•H2O are Mo=O1 and Mo=O2 (1.724(2) and 1.694(2) Å, respectively). Selected bond lengths and angles are given in Table S1   Whereas the compound 1·MeCN partially loses MeCN already upon standing at room temperature, crystals of 3·H 2 O are more stable and could be handled with less precaution. On the other hand, compounds 1a-4a with coordinated solvent molecules are even more stable. According to the thermal analysis, they lose the coordinated MeOH in the range 178-210 • C (1a), 206-224 • C (2a), 120-175 • C (3a) and 214-247 • C (4a). Following decomposition, the final residue is MoO 3 .

Catalysts Preparation and Characterization
In [MoO 2 (L 2 )] n ·H 2 O (3·H 2 O) the hydrazone coordinates the molybdenum atom of the cis-MoO 2 2+ core tridentately through the phenolic and isonicotinoyl oxygens and azomethine-nitrogen ( Figure 1a). The isonicotinoyl nitrogen atom of an adjacent complex unit occupies the remaining sixth coordination site thus enabling the formation of a one-dimensional polymer ( Figure 1b). The distance Mo-N3 (2.427 Å) is the largest bond length within the molybdenum coordination sphere. According to the Cambridge Structural Database [17], only five Mo-coordination polymers with isonicotinoyl-based ligands were crystallized and structurally characterized so far [18][19][20][21][22]. The Mo-N bond in these structures has bond length in the range 2.426(6)-2.549(4) pm. The shortest one is in the polymer with 5-iodo-2-(olato)benzylidene)-(pyridine-4)carbohydrazonato) (Refcode GATGOA) [19] and the longest once in 4-(diethylamino)-2-(olato)benzylidene)-(pyridine-4)carbohydrazonato) ligand (Refcode GELXUS) [17]. The asymmetric and symmetric bands for {MoO2} core vibrations appear around 935-925 cm −1 in the IR spectra, and these bands tend to overlap. The interaction Mo=Ot•••Mo in 1-4 is excluded due to the absence of broadband at ~800 cm −1 in their spectra ( Figure S1). Instead, a strong band belonging to ν(O=Mo-N) around 900 cm −1 appears and is used to indicate the polymer formation. For the mononuclear complexes 1a-4a, this band is replaced with a new one of differing intensity due to ν(O=Mo-OMeOH) stretching ( Figure S2). Furthermore, a new band at ~1020 cm −1 due to the C-OMeOH vibrations appears in the spectra. However, for polymer, no significant band is observed in that region. The band at ~1340 cm −1 , assigned to the C-O group of the hydrazone moiety, and bands at ~1600 cm −1 and 1250 cm −1 , belonging to C=Nimine and C-Oph groups, respectively, indicate coordination of the ligand to the {MoO2} 2+ unit through the ONO atoms of these three functional groups.
The asymmetric and symmetric bands for {MoO 2 } core vibrations appear around 935-925 cm −1 in the IR spectra, and these bands tend to overlap. The interaction Mo=O t ···Mo in 1-4 is excluded due to the absence of broadband at~800 cm −1 in their spectra ( Figure S1). Instead, a strong band belonging to ν(O=Mo-N) around 900 cm −1 appears and is used to indicate the polymer formation. For the mononuclear complexes 1a-4a, this band is replaced with a new one of differing intensity due to ν(O=Mo-O MeOH ) stretching ( Figure S2). Furthermore, a new band at~1020 cm −1 due to the C-O MeOH vibrations appears in the spectra. However, for polymer, no significant band is observed in that region. The band at 1340 cm −1 , assigned to the C-O group of the hydrazone moiety, and bands at~1600 cm −1 and 1250 cm −1 , belonging to C=N imine and C-O ph groups, respectively, indicate coordination of the ligand to the {MoO 2 } 2+ unit through the ONO atoms of these three functional groups. The coordination via ONO donor atoms is also maintained in the solution as confirmed by the NMR analysis (Tables S3 and S4, Figures S1 and S2, Supplementary Materials). The singlets belonging to the NH (=N-NH-(C=O)-) and OH-2' protons are absent in the 1 H NMR spectra of the complexes, indicating ligand tautomerization (to =N-N=(C-OH)-) and coordination through the deprotonated oxygen atoms. Significant deshielding of carbons adjacent to donor ONO atoms is observed, being larger for carbons at positions 1 and 4 (up to 8.01 ppm and 5.66 ppm, respectively) than for the carbon at position 2 (up to 2.04 ppm), Table 1, Scheme 1. Signals arising from free MeOH (ca. one equivalent), seen in the spectra of 1a-4a, are suggesting MeOH co-ligand substitution with dmso-d 6 .
The reactions were carried out at 80 • C in a stirred solution of the substrate, catalyst, and oxidant in acetonitrile for 5 h. All the complexes were insoluble in acetonitrile at room temperature (orange slurry) but dissolved after 150 min in the carveol reaction mixture or after 20 min in the cyclohexanol reaction mixture at 80 • C (orange mixture turn yellow until the end of the reaction). The conversion of substrates was calculated according to an internal standard, acetophenone.

Carveol Oxidation
In order to find efficient and eco-friendlier catalytic system reactions were conducted with low Mo loading n(Mo):n(substrate):n(oxidant) = 1:400:800, where oxidant is H 2 O 2 or THBP (solution in water or decane).
In the case of H 2 O 2 , the conversion of carveol is high (84-91%) for all tested molybdenum complexes, following the order 4a > 3a > 1 > 2 = 1a > 2a > 4 > 3 ( Table 2). All catalysts show similar values of selectivity towards carvone (41-44%). TOF 20 min for the complexes obtained from 2,3-dihydroxybenzaldehyde (1, 2, 1a, 2a) showed higher values in comparison to the complexes obtained from 2,4-dihydroxybenzaldehyde (3, 4, 3a, 4a) implying faster activation time and conversion to the pentacoordinate active species [MoO 2 L]. TON values for all the complexes, except 3 and 1a, were up to 300. Since the oxidation system employing H 2 O 2 and the complexes obtained from the ligands H 2 L 1 and H 2 L 2 showed better results in terms of tested catalytic parameters, further investigation was continued with the complexes 1, 2, 1a, and 2a. Kinetic profiles of polynuclear and mononuclear complexes were presented in Figure 2.  On the other hand, for the tested molybdenum catalysts, 1, 2, 1a, and 2a, conversion of carveol with aqueous THBP was in the range 56 to 66%, with low selectivity towards carvone (10-19%). The kinetic profile was presented in Figure 3. Interestingly, complex 2a with aqueous THBP showed the lowest carveol conversion (56%), but the highest selectivity towards carvone (19%) compared to other catalysts. It seems that a slower reaction rate seems to favour carvone formation. Additionally, the catalytic activity of the complex 2a was tested in the presence of THBP solution in decane, and remarkable carveol conversion is achieved (99%) after 20 min of the reaction. It is not surprising that TOF20 min achieved a value of 1001. On the other hand, for the tested molybdenum catalysts, 1, 2, 1a, and 2a, conversion of carveol with aqueous THBP was in the range 56 to 66%, with low selectivity towards carvone (10-19%). The kinetic profile was presented in Figure 3. Interestingly, complex 2a with aqueous THBP showed the lowest carveol conversion (56%), but the highest selectivity towards carvone (19%) compared to other catalysts. It seems that a slower reaction rate seems to favour carvone formation. Additionally, the catalytic activity of the complex 2a was tested in the presence of THBP solution in decane, and remarkable carveol conversion is achieved (99%) after 20 min of the reaction. It is not surprising that TOF 20 min achieved a value of 1001. carvone (10-19%). The kinetic profile was presented in Figure 3. Interestingly, complex 2a with aqueous THBP showed the lowest carveol conversion (56%), but the highest selectivity towards carvone (19%) compared to other catalysts. It seems that a slower reaction rate seems to favour carvone formation. Additionally, the catalytic activity of the complex 2a was tested in the presence of THBP solution in decane, and remarkable carveol conversion is achieved (99%) after 20 min of the reaction. It is not surprising that TOF20 min achieved a value of 1001.  However, due to low selectivity towards carvone (4%) ( Table 2), further testing of other complexes was not performed under these conditions.
Since carveol contains two double bonds that may be targeted by the oxidant, relatively high conversion of carveol, accompanied by low selectivity towards carvone, can be expected. According to the literature, carveol epoxidation is the most likely to occur. which further explains low carvone yield (Figure 4) [23][24][25]. However, due to low selectivity towards carvone (4%) ( Table 2), further testing of other complexes was not performed under these conditions.
Since carveol contains two double bonds that may be targeted by the oxidant, relatively high conversion of carveol, accompanied by low selectivity towards carvone, can be expected. According to the literature, carveol epoxidation is the most likely to occur. which further explains low carvone yield (Figure 4) [23][24][25]. The obtained results imply that the substituent on the ligand (isonicotinic vs. nicotinic acid hydrazide) and type of complex (polynuclear vs. mononuclear) do not influence a lot on carveol oxidation. On the other hand, the effect of the used oxidant on the selectivity towards carvone should be discussed. During the reaction, a different consumption of cis-and trans-carveol was observed in the GC chromatogram for each of the oxidants. Having the latest in mind, different carvone selectivity can be explained by reaction stereoselectivity towards cis-or trans-carveol. By 1 H-NMR analysis [26] of the obtained reaction mixtures, it was seen that in the presence of H2O2, the preferred substrate is ciscarveol, while with aqueous THBP, is trans-carveol ( Figure 5). The obtained results imply that the substituent on the ligand (isonicotinic vs. nicotinic acid hydrazide) and type of complex (polynuclear vs. mononuclear) do not influence a lot on carveol oxidation. On the other hand, the effect of the used oxidant on the selectivity towards carvone should be discussed. During the reaction, a different consumption of cisand transcarveol was observed in the GC chromatogram for each of the oxidants. Having the latest in mind, different carvone selectivity can be explained by reaction stereoselectivity towards cisor transcarveol. By 1 H-NMR analysis [26] of the obtained reaction mixtures, it was seen that in the presence of H 2 O 2 , the preferred substrate is cis-carveol, while with aqueous THBP, is trans-carveol ( Figure 5). a lot on carveol oxidation. On the other hand, the effect of the used oxidant on the selectivity towards carvone should be discussed. During the reaction, a different consumption of cis-and trans-carveol was observed in the GC chromatogram for each of the oxidants. Having the latest in mind, different carvone selectivity can be explained by reaction stereoselectivity towards cis-or trans-carveol. By 1 H-NMR analysis [26] of the obtained reaction mixtures, it was seen that in the presence of H2O2, the preferred substrate is ciscarveol, while with aqueous THBP, is trans-carveol ( Figure 5).  In addition, a high peak, belonging to the unknown compound A, could be detected in the GC chromatogram. In the reaction mixture with H 2 O 2 , carvone was the main product and unknown compound A is a by-product. However, when using TBHP as an oxidant, unknown compound A was the main product. Further discussion in that regard is available in the Supplementary Materials.

Cyclohexanol Oxidation
At first, reactions were conducted with low Mo loading, (n(Mo):n(substrate):n(oxidant) = 1:400:800, where oxidant is H 2 O 2 or THBP (in water or in decane), protocol A in the Experimental part. Since cyclohexanol conversion was extremely low in H 2 O 2 , <10% for all the catalysts, the testing with other oxidants was continued only with the complexes 2a. The cyclohexanol conversion is extremely low, 11% for 2 and 13% for 2a, but selectivity towards cyclohexanol is moderate, 64% for 2 and 59% for 2a.
All relevant catalytic data are presented in Table 3. A catalytic system with complex 2a and THBP solution in the decane provides the highest cyclohexanol conversion (28%) and cyclohexanone yield (16%), while the selectivity towards cyclohexanone is the same as with H 2 O 2 (59%). The use of TBHP in water did not result in better 2a activity, and selectivity towards cyclohexanone was even diminished. However, the use of TBHP in decane provided better substrate conversion, 28%, while the cyclohexanone selectivity remained the same as with H 2 O 2 . Due to the very low results, higher catalyst loading, 1%, and lower content of oxidant was investigated (protocol B in the Experimental part), with H 2 O 2 as an oxidant. All molybdenum(VI) complexes were tested within the following reaction conditions, n(Mo):n(substrate):n(H 2 O 2 ) = 1:100:300. The results are summarized in Table 4. Slightly higher conversion for all the tested catalysts is achieved after 5 h (15-21%), followed by the higher values of selectivity towards cyclohexanone, the lowest being 57% for complex 3 and the highest one 71% for complex 2a.

Butan-2-ol Oxidation
Considering that in our previous investigations, mononuclear complexes usually showed slightly better activity, catalysts 1a-4a were tested for oxidation of butan-2-ol. Knowing the fact that MeCN provides better activity and selectivity towards the desired product, MeCN was added to both catalytic systems, containing H 2 O 2 and aqueous TBHP. As seen from the results compiled in Table 5, MeCN did not have any tremendous effect on the tested reaction. In general, the selectivity towards butan-2-one is slightly better when MeCN is added and the catalyst is faster transferred into active species (concluded from TOF 20 min values). However, TON values remain similar, no matter the addition of the solvent. Furthermore, the nature of the used oxidant does not have a dramatic effect on the catalytic process. Table 5.
Results  In comparison to the similar reported investigations with molybdenum Schiff base catalysts [13][14][15], the catalytic process presented within this research, provide great potential. For carveol oxidation, the system employing H 2 O 2 provided to be the best one, justifying and following the principles of green processes. Cyclohexanol and butan-2-ol oxidation reactions, after 5 h, resulted with very good ketone yields, while the conversion parameter demands further optimization.

General Procedure for the Oxidation of Secondary Alcohols
Alcohol, 0.1 mL acetophenone (internal standard), and 2.5 mL acetonitrile (if added, details hereunder) were stirred together. Mo(VI) (pre)catalyst was added to the mixture. The mixture was stirred and heated to 80 • C, and the oxidant was added. The reaction was monitored by withdrawing small aliquots of the reaction mixture at a definite time interval (0, 20, 50, 90, 150, and 300 min), and analysed quantitatively by gas chromatography. The same as protocol A and with 2.5 mL of MeCN.

Physical Methods
The C, H, and N mass contents were provided by the Ruder Bošković Institute, Zagreb. Thermogravimetric analysis was performed on a Mettler TG 50 thermobalance using Al 2 O 3 crucibles in an oxygen atmosphere with a flow rate of 100 cm 3 min −1 and heating rates of 10 K min −1 . IR-ATR spectra were recorded at room temperature using a Perkin Elmer Spectrum Two FTIR Spectrometer using the Attenuated Total Reflectance technique (ATR). NMR spectra were obtained on a 400 MHz Bruker Avance III HD spectrometer in dmso-d6.
Single-crystal X-ray diffraction data for 3·H2O was collected on an XtaLAB Synergy-S diffractometer with CuK α (λ = 1.54184 Å) radiation at 170 K. Data reduction was performed using the CrysAlis software package [29,30]. The solution, refinement, and analysis of the structures were done using the programs integrated with the WinGX [31] and OLEX2 [32] systems. All structures were solved and refined with the SHELX programme suite [33]. Structural refinement was performed on F2 using all data. All hydrogen atoms were placed at calculated positions and treated as riding on their parent atoms. Geometrical calculations were done using PLATON [34]. Drawings of the structures were prepared using PLATON and MERCURY programs [35]. The catalytic reactions were followed by gas chromatography on an Agilent 6890A chromatograph equipped with an FID detector and a DB5-MS capillary column (30 m × 0.32 mm × 0.25 mm). The GC parameters were quantified using authentic samples of the reactants and products. The conversion of carveol, cyclohexanol, and butan-2-ol and the formation of cyclooctene oxide were calculated from calibration curves (r 2 = 0.999, 0.997, 0.999, respectively) relative to acetophenone as an internal standard.

Conclusions
Molybdenum(VI) complexes, mononuclear [MoO 2 (L)(MeOH)] and polynuclear [MoO 2 (L)] n ones, were tested as catalysts for the oxidation of secondary alcohols, carveol, cyclohexanol, and butan-2-ol, in the presence of different oxidants, H 2 O 2 and THBP (in water or in decane). Carveol oxidation provided very good conversions. However, carvone yield was better with H 2 O 2 and can be compared to the previously reported results. Catalytic investigation for cyclohexanol was less successful and require further optimization of reaction conditions. Although the reaction was relatively slow, selectivity towards cyclohexanone gives great potential to the tested catalytic system. In the end, butan-2-ol oxidation, with the assistance of mononuclear complexes used as catalysts, provided interesting results in terms of solvent (MeCN) addition to the investigated arrangement. For the research under these conditions, additional solvent did not show a positive effect and consequently was not needed, justifying green chemistry principles. Since the complexes with different ligand substituents showed similar results, the general conclusion was that the ligand influence on the catalysis is not significant, as well as the type of the used complex (mononuclear or polynuclear).

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/10 .3390/catal11080881/s1, CCDC no: 2091707 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/structures.