Exploitation of Liquid CO2 Based Greener Process for Valorization of Citronellal-Rich Essential Oils Into Flavor Grade (−)-Menthol Using Novel Sn/Al-B-NaYZE Composites

A two-step greener catalytic process has been developed for the semi-synthesis of (−)-menthol from citronellal-rich essential oils such as Cymbopogon winterianus and Corymbia citriodora by using novel bi-acidic composites. These novel composites (Al-B-NaYZE or Sn-B-NaYZE) were prepared by impregnation of Sn-B and Al-B over NaYZE framework to increase the Lewis and Bronsted acidic sites. These composites were thoroughly characterized using TDP, FT-IR, UV, XRD, SEM, HRTEM, SA, and TGA. The composite Sn-B-NaYZE cyclized 99% of citronellal to isopulegol isomers with 99% selectivity in 45 min under a liquid CO2 medium, while composite Al-B-NaYZE is showing 95% selectivity towards isopulegol isomers. Sn-B-NaYZE displayed enhanced selectivity due to its higher Lewis and Bronsted acidic characteristics. The chiral study indicated that the Sn-B-NaYZE composite is giving the highest selectivity (94%) towards (−)-isopulegol. (+)-Citronellal or citronella essential oil was found as an ideal substrate for (−)-menthol production. It observed that the Sn-B molar ratio over base material NaY-Zeolite played a major role in catalytic activity and selectivity towards (−)-isopulegol. Further, isopulegol isomers were reduced to 98% of menthol using 1%Pd/AC at 40 psi H2 pressure in 1 h. The reaction mixture was slowly frozen to -40 °C for trouble-free isolation of the crude menthol. Further, pharmaceutical and flavor grade (−)-menthol was purified from the crude menthol through an esterification process. This (−)-menthol was found to be 100% biobased on 14C-radiocarbon-dating to authenticate as nature-identical.

The reported processes might be clear-cut for menthol conversion when using the pure (+)-citronellal as substrate. But, the major challenge arises, when this conversion is attempted in essential oils containing an enantiomeric mixture of (±)-citronellal. The selective transformation of citronellal in essential oil is a challenging task due to the presence of a large number of thermolabile terpenoids. As a result, either the catalytic active sites have been reduced or constituents other than the target compound are chemically transformed. Therefore, effective catalysts for selective conversion are of utmost importance to get the economically viable process. The present work overcomes such limitations using inexpensive composites at a lower temperature, and atmospheric pressure conditions have been facilitated for the production of menthol directly from low-value essential oils. Hence, presently the preparation of menthol from crude citronellal-rich essential oils (feedstocks) such as citronella, eucalyptus, etc. has been carried out. Since the essential oils are contained enantiomeric mixtures of (±)-citronellal, hence enatiospecific semi-synthesis of (−)-isopulegol has been studied. The present process realized reasonably low-cost composites for the conversion of citronellal to isopulegol using Al-B-NaYZE or Sn-B-NaYZE in the first step using a liquid-CO 2 solvent. Liquid CO 2 is a green and eco-friendly solvent system, and it easily attained the liquid state at below critical conditions (31.1 °C, 73.8 bar) of CO 2 [10]. Importantly, in the second step, isopulegol was reduced using a very low percentage of Pd or Ru to produce the menthol. In addition, (−)-menthol was isolated, purified, and validated through the biobased index to prove it as a natureidentical molecule. Purification of (+)-citronellal from essential oils is cost as well as labor-intensive, therefore the direct selective transformation and isolation of (−)-menthol is an attractive alternative. Besides (−)-menthol, the other compounds in the modified essential oil have produced a better olfactometry profile to use as a fine fragrance and food additive in various industries.

Preparation of Composites
Paste of NaYZE was made in distilled water and then the acidified molar solution of Al(NO 3 ) 3 or Sn(SO 4 ) 2 was impregnated at 30 °C with constant stirring (800 rpm) for 2 h. Thereafter, the materials were thoroughly washed with distilled water and kept for drying at 120 °C. The composites were powdered again and calcined at 550 °C to obtain Al-NaYZE or Sn-NaYZE. Then, aqueous BA was added drop-wise to the above mixture and stirred for 2 h at 30 °C to obtain the composites by the co-precipitation method. These composites were placed in an oven at 110 °C for complete evaporation of water. Then, composites were thoroughly washed in water followed by drying at 110 °C. Finally, these composites were calcined for 5 h at 550 °C to obtain Al-B-NaYZE or Sn-B-NaYZE. Further, YZE (Y-H-Zeolite) was produced from NaYZE by ion exchange with 2 M NH 4 Cl solutions followed by washing, drying, and calcination at 550 °C. Then a similar methodology was followed for the preparation of Al-B-YZE or Sn-B-YZE composites. The molar ratio of each active component is given in Supplementary Table S1.
Reduced metal catalysts were prepared as follows: 100 mg of 10 mmol of Pd(NO 3 ) 2 or Ru(NO)(NO 3 ) 3 were dissolved in 30 mL of 1 N HCl, and then an aqueous solution of NaHCO 3 was used to maintain the pH 7. Thereafter, 15-20 mL of NaBH 4 (0.05 mmol) and 15-20 mL of glucose (0.3 mmol) were added in stirring conditions for 30 min to obtain a colloidal solution. AC was added in appropriate amounts of Pd or Ru colloidal solution to achieve 0.5-2%Pd/AC or Ru/AC catalyst. Then, catalysts were thoroughly washed in water followed by drying at 110 °C. The dried catalyst was placed in a furnace at 220 °C for 4 h before use.

Characterization of Catalysts
Catalysts were characterized using UV-Vis, FT-IR, XRD, SA (surface area), NH 3 -TPD, SEM-EDX, TEM, HRTEM-EDX, TGA, XPS, True density, and ICP-OES as per the protocol given in Supplementary data. The Lewis and Bonsted acid sites in the composites are determined by the pyridine (Py)-TGA adsorption-desorption method (Supplementary data).

Reaction Conditions
The citronellal cyclization was carried out using any one of the solvent such as cyclohexane, CHCl 3 , CH 2 Cl 2 in a ratio of citronellal to solvent (5:1 to 8:1) in an RB flask connected to the condenser. Further, a ratio of composites to citronellal (1:15 to 1:18) was added to this reaction mixture. The mixture was allowed to agitate at a rate of 400 rpm at 50-75 °C for 15 min to 2 h. Finally, the process optimization was carried out using Al-B-NaYZE and Sn-B-NaYZE composites (Table S2, and S3).
The above cyclization of citronellal was carried out in a laboratory specially designed stainless steel liquid CO 2 apparatus. The experimental system for the liquid-CO 2 extraction unit has been illustrated in Fig. S1. Citronellal (1 g) was taken in a designed beaker along with 70 mg of composites (Al-B-NaYZE or Sn-B-NaYZE). The reaction was carried out at 21 °C (approx) with CO 2 pressure of 68-70 bar. The chilled water was circulated in the cooling fingure of the apparatus at 15 °C and the bottom of the apparatus was maintained at 34 °C. The process of evaporation and condensation were repeated throughout the process for 15-45 min. Then the CO 2 was removed to obtain isopulegol isomers. Further, it was dissolved in ethanol and then filtered to obtain isopulegol solution. In the further scale-up study, citronellal (10 g) with 0.6 g of the composite was taken for the cyclization reaction under the same reaction conditions for 45 min. For direct conversion, 20 g of essential oil (Citronella or Eucalyptus) was taken with a ratio of citronellal to composite (10:0.7) for 45 min.
The different reduction catalysts are prepared as per Fig. S2, and it is clear that Pd/AC or Ru/AC is suitable for the reduction of isopulegol to menthol. The reacted mixture in the above (first) step was reduced to menthol under H 2 pressure (10-40 psi) using prepared Pd/AC or Ru/AC catalyst in a ratio of 1:10 to 1:13 (catalyst: reacted mixture) in ethanol for 10 to 60 min (Table S3, and S4). In a similar methodology, the above-reacted oil mixture was reduced using Pd/AC. After completion of the step-two reaction, the ethanol was removed in a rotary evaporator at 50 °C under 150 mbar pressure, and then the modified oil was kept in a refrigerator at -40 °C for 24 h to isolate the crude menthol crystal. For enrichment of (−)-menthol, the crude menthol isomers were derivatized to menthyl acetate using acetic anhydride or acetyl chloride (Supplementary data). After completion of the reaction, (±)-menthyl acetate was crystallized at 35 °C to isolate (−)-menthyl acetate, whereas other diastereomers remained in the liquid phase. The hydrolysis of (−)-menthyl acetate yielded (−)-menthol. Further, (−)-menthol was crystallized at -40 °C to obtain the bold crystals. The citronellal conversion is determined by using the Eqs. (1)(2)(3).
w h e r e Ccal o = initial concentration, a n d Ccal t = concentration at timet Product selectivity (Si) is calculated as The reaction rate (R) is calculated by

Characterization of essential oils and reaction mixtures
The percentage composition was determined by PerkinElmer GC connected to a flame ionization detector (FID). WCOT column coated with diphenyl dimethyl siloxane with a dimension of 25 m × 0.25 mm × 0.25 μm (HP-5) was used for this analysis. H 2 was used as carrier gas with a flow rate of 1.2 mL min −1 under initial column pressure of 42 kPa. Compounds were separated with clear baseline distinction following a linear temperature program of 60 °C-200 °C at 2 °C min −1 and finally held at 200 °C for 30 min. The peak normalization method was used for the calculation of the relative area percentage of the compounds. Again, Perki-nElmer GC/MS was used for compound confirmation under similar GC-FID conditions. MS conditions were as follows the detector voltage (1.5 V), ionization voltage (70 eV), peak width (2 s), mass range (50-400 amu), etc. Individual peaks were confirmed by comparing the mass spectra of compounds with the libraries available on the computer (NIST-1, NIST-2, Adams). For confirmation of the isomers, the linear relative retention indices (Kovats indices) of the compounds were calculated from standard n-alkanes (C8-C22) chromatograms [4,11]. The chiral-GC-FID analysis was carried out in β-Dex column (Rt™-bDEXse, ResTek corporation) under similar reaction conditions. Further, 13 C-NMR spectrum, menthol (10 mg) was dissolved in CDCl 3 and recorded on DPX-500 Bruker system at 125 MHz at 26 °C with trimethylsilane (TMS) as the internal standard.

Biobased evaluation
Biobased evaluation of the samples is calculated by quantification of 14 C content, which is reported in the percentage of modern carbon (pMC) values. Accelerator Mass Spectrometry (AMS) was used for the quantification of biobased Change in the concentration of reactants (mol∕l) Time taken (s) carbon content in each sample [12]. Samples were combusted in automated graphitization equipment (AGE), and converted into graphite powder for enabling AMS measurements to quantify the 14 C/ 12 C ratio. An ion accelerator (500 kV) based AMS system was used for the quantification of 14 C/ 12 C present in each of the samples [13]. The percentage of biobased "C" is calculated as per Eq. (4) The reference value for the sample collected in the particular year was obtained as per ASTM D6866-16 [14].

Characterization of spent catalysts
The reaction medium after filtration was used for the estimation of B, Al, Si, Sn, and Na. The used composites were thoroughly washed with acetone and water, and calcined at 550 °C for 5 h. There was not much difference in SA and porosity of the re-used composites. A leaching test of the filtrate was performed by removing the composites from the reaction mixture after 45 min. Similarly, 1%Pd/AC was thoroughly washed with warm water (60 °C) and calcined at 220 °C for 4 h before using the next batch of reactions.

Characterization of composites
The surface morphology of the composites (Sn-B-NaYZE, Al-B-NaYZE, Sn-B-YZE, and Al-B-YZE) were examined by SEM and TEM. The SEM micrograph revealed that all particles persist in a well-shaped structure, with sizes ranging from 0.54 to 0.78 µm (Fig. S3). Also, the SEM images of 1%Pd/AC and 1%Ru/AC have shown the proper incorporation of Pd and Ru on AC surface (Fig. S3c,f). Moreover, the elemental distribution of composites is determined through SEM-EDAX analysis (  [15,16]. The lack of any hump and soft horizontal line in the background indicated the complete crystallinity in composites without any amorphous species (SiO 2 ) [17,18]. In Sn-B-NaYZE has been shown 2θ corresponds to 26.4° and 33.9°, these are matched with lattice planes of 110 and 101 of SnO 2 , respectively [19]. The resemblance of XRD patterns of Sn-B-NaYZE and Sn-B-YZE have indicated the successful incorporation of Sn into the NaYZE and YZE framework, and it is supported by decreasing the molar ratio of SiO 2 /Al 2 O 3 in Sn-B-NaYZE and Sn-B-YZE (Table 2). Al/Sn impregnated ZE had reduced crystallinity in comparison to the normal one, which was reflected by decreasing the intensity of diffraction peaks. Sn-B-NaYZE was less crystalline as compared to Al-B-NaYZE and Na-YZE. The incorporation of Sn is led to enhance the Na content ( Table 2). The conventional YZE framework was composed of SiO 4 and AlO 4 species, which were arranged  in tetrahedral symmetry. Due to the different oxidation states of Al 3+ and Si 4+ , which impacted the overall negative charge to the framework, thereby need monovalent ion (Na + ) for maintaining the charge balance. For the Sn 2+ composite, the framework has shifted the burden to increase the negative charge, hence additional monovalent Na + was needed for maintaining the charge neutrality to facilitate better incorporation of Sn 2+ [17]. The intensity of characteristic peak indicated the extent, as well as the concentration of crystallinity in composites could be expressed according to the equation described by Pal et al. [20]. The absence of any extra peak at 28° indicated no more free BA in the composites [21]. Similarly, 1%Pd/AC has shown the characteristics 2θ peaks at 40.5°, 47.0°, and 68.5°. The data for NH 3 -TPD of Al-B-Na-YZE, Sn-B-Na-YZE, Al-B-YZE, and Sn-B-YZE depicted two peaks: one narrow peak at 150 °C and another broad peak at 500 °C (Fig. 3a). The low-temperature peak was derived from the weakly adsorbed NH 3 molecules via H-bonding. Whereas, the high-temperature peak was ascribed to strong Bronsted acid sites. As shown in Fig. 3a, the total peak area corresponding to Al-B-YZE, Sn-B-YZE, Al-B-Na-YZE, and for Sn-B-NaYZE and Sn-B-YZE, respectively. From the above results, Sn-B-NaYZE is a strong acidic composite as compared to Sn-B-YZE. The lower percentage of B loading slightly impacted the Si/B ratio, which helped to shift the orientation towards the tetrahedral coordination. On the other hand, higher B loading decreased the Si/B ratio, hence facilitating the trigonal forms of coordination [23]. Therefore a definite percentage of B incorporation has been shown to enhance the Prins reaction, due to its proper balance between the trigonal and tetrahedral coordination forms. Trigonal B represented the Lewis acidic character, whereas B exhibited the Bronsted acidic behavior in dynamic equilibrium to favor the extraordinary activity of these composites [24].
The incorporation of Sn into the ZE (zeolite) framework is verified by examining the bands formed by charge transfer from ligand to the metal as analyzed from the UV-Vis spectra (Fig. 3b). Sn-B-NaYZE and Sn-B-YZE exhibited a strong UV-Vis characteristic absorption band centered at 219 nm, which was featured in the presence of tetracoordinate Sn 4+ in the ZE framework. The intensity of this band was proportional to the loading of tetrahedrally coordinated Sn 4+ ions [25]. The band at 250 nm was attributed to the polymeric SnO 2 species. Therefore, the stronger band intensity of Sn-B-NaYZE in comparison to that of the Sn-B-YZE suggested the greater formation of Sn 4+ species during its incorporation procedure [25]. The absorption spectra of pure SnO 2 demonstrated an extensive band at ~ 240-400 nm, which was allocated to the polymeric Sn-O-Sn category species. The absorption spectra of Sn-B-NaYZE and Sn-B-YZE exhibited a blue shift relative to the pure SnO 2 . This blue shift could be attributed to small SnO 2 species and is largely dependent on the size-quantization of the SnO 2 particles in the nanometer regime. SnO 2 species were smaller in size and impacted a stronger blue shift [19]. The blue shift is more significant for Sn-B-NaYZE, which indicated the presence of smaller SnO 2 particles (Fig. 3b). The absorption spectrum of Sn-B-NaYZE has contained a strong band at ~ 205 nm with very weak shoulder bands in between ~ 290-420 nm, it indicated that most of the Sn was inserted into the tetrahedral framework, and very few of them were present in the extra framework. However, SnO 2 phases were not detected in Sn-B-YZE. Furthermore, the spectra of Al-B-NaYZE and Al-B-YZE were quite similar to YZE spectra, which means no such changes were observed after Al impregnation on the ZE framework.
XPS spectra of composites are presented in Fig. S7. The XPS spectra illustrated two principal signals that related to  Sn 3d 5/2 and Sn 3d 3/2 photoelectrons with binding energies of 487.0 and 495.5 eV, respectively (Fig. 3c), which indicated the existence of Sn 4+ particles on the surface [24]. The count Sn-B-NaYZE was higher as compared to the Sn-YZE. The consequences implied that the Sn-B-NaYZE acquired more framework of Sn particles on the surface, which complied with the results of UV-Vis and SEM-EDX ( Table 2) (Fig. S7c).
FT-IR spectra of the composites are given in Fig. 4a,b. The band at 1150 and 1017 cm −1 were assigned to asymmetric stretching vibration Al-O-Si, while 791 and 450 cm −1 were allocated to symmetric stretching and bending vibration of Al-O-Si. Sn-impregnated composites have been shown the characteristic peaks at 592, 915, and 1627 cm −1 . Al-impregnated composites have been given a characteristic peak at 1110 cm −1 . The peak observed at 460 cm −1 indicated the Si-O bending vibration, and at 580 cm −1 was originated due to the double six-membered ring structure of the ZE [26]. All the composites have displayed a distortion between the 910-950 cm −1 , which is allocated to asymmetrical stretching of Si-O-M(M: Al or Sn). Sn-B-YZE and Al-B-YZE composites were shown bands at 3434, 3227, and 2260 cm −1 , which were arisen due to stretching vibration of SiO-H, BO-H, and B-OH, respectively. The B-OH bending vibration was assigned at 1187 cm −1 [21]. Whereas, these bands are completely absent in Al-B-NaYZE and Sn-B-NaYZE composites. The stretching and bending vibration of Si-O-B (borosilicate linkage) are observed at 927 and 649 cm −1 , respectively (Fig. S4a,b). In addition, the composites incorporated B in the crystal lattices in the form of trigonal as well as tetrahedral valency states [23]. Thus, the tetrahedral B was assigned the peak at 920-930 cm −1 , whereas trigonal B was at 1395 cm −1 .
Nitrogen adsorption-desorption was carried out to determine the SA and pore volume. The isotherms are investigated at a relative pressure of P/Po ranging from 0 to 0.9 (Fig. 5a,b). The SA is calculated using Eq. (5), the SA, pore volume, and pore radius are presented in Table 3. where SA: Surface area m 2 /g, Vm: Volume of monolayer m 3 , Am: Area occupied by one molecular of nitrogen in The N 2 adsorption reveals that Al/Sn-B-Na-YZE belongs to type-1 isotherm and Al/Sn-B-YZE composite belongs to type-IV isotherm (Fig. 5a,b). Isotherms of NaYZE and Al/ Sn-B-NaYZE exhibited microporous adsorption behavior as missing hysteresis loops within the framework, and the pattern of pore size distribution indicated the lack of mesoporous behavior, which thereby confirmed the microporous nature. While, the isotherms of YZE and Al/ Sn-B-YZE indicated the presence of some mesopores, due to the presence of a hysteresis loop in the framework. The isotherm was raised at a low-pressure range with an increase in relative pressure (P/Po), and on further increasing the relative pressure the N 2 adsorption was increased, which confirmed the presence of mesopores in the composites [21]. The small change in SA and micropore volume after the adsorption of Sn and Al indicated the successful incorporation of Sn and Al on ZE (NaYZE and YZE) ( Table 3). Whereas, the significant decrease in SA and micropore volume was recorded when B was incorporated into the Sn-NaYZE framework which might be formed complex oxides , Langmuir SA (28.4, 6.9 m 2 g −1 ), total pore volume (0.022, 0.006 cm 3 g −1 ) with improved pore diameter (503.4, 583.8 Å), respectively. On the other hand, 1%Pd or 1%Ru reduced catalysts are prepared using inexpensive AC as a major percentage, which gave improved SA and porosity for enhancing the activity (Table 3).
Composites have been shown thermo-stability behavior up to 1000 °C (Fig. S8). The total mass losses of composites such as Sn-B-NaYZE, Al-B-NaYZE, Sn-B-YZE, and Al-B-YZE were 3.5%, 5.2%, 4.1%, and 6.7%, respectively in the entire temperature range, which indicated prepared catalysts were thermally more stable as compared to the respective base materials. The difference in the stability of the base material catalyst and the metal-incorporated base material catalyst is due to hindrance in the diffusion of components by the incorporated metals, also hindering the phase transition kinetics. It was validated that Na was firmly associated with the composites to give extra stability and helped to generate the acidic sites. The phase transition kinetics is shown by the TGA evaluation of composites before and after Py adsorption (Fig. 6). The initial mass loss after Py adsorption was attributed to desorption from weakly acid sites, and the mass loss at a higher temperature (> 400 °C) was attributed to desorption at strongly acidic sites up to 700 °C, which was well correlated the findings of NH 3 -TPD result.
Metal loadings of the composites are determined by ICP-OES and SEM-EDX ( Table 2, Table S5). The composites contained the rational percentage of Al, Sn, Na, and B, which formed the complex architecture. Therefore, the reduction of SA and pore volume may be due to the partial pore blockage of ZE with these complex oxides [25]. It is justified that the drastic reduction (> 50 times) of micropore volume and also the SA in Sn-B-NaYZE and Al-B-NaYZE (Table 3). But, the pore diameter (503.4-583.8 A o ) of the above composites was enhanced more than two times.

Carbonyl-ene Rearrangement of Citronellal
The carbonyl-ene rearrangement of citronellal is studied in Al-B-NaYZE, Sn-B-NaYZE, Al-B-YZE, and Sn-B-YZE   (Fig. S9) Among these, Al-B-NaYZE and Sn-B-NaYZE are shown better activity (Table S2). The order followed by the catalytic activity of prepared composites is Sn-B-NaYZE > Al-B-NaYZE > Sn-B-YZE > Al-B-YZE. Further, the elemental ratios of both these composites have been optimized for getting higher citronellal cyclization. The ratios of Sn-B-NaYZE (0.02:0.05:0.5) and Al-B-NaYZE (0.01:0.05:0.45) are optimum for performing a better activity (Fig. 3). This transformation has been studied in CHCl 3 , CH 2 Cl 2 , and cyclohexane. It has been observed that cyclohexane gave 96% conversion of citronellal to about 98% of isopulegol isomers in 2 h at 70 °C (Fig. S10). The kinetics of transformation from citronellal to isopulegol has been optimized to obtain maximum conversion in 2 h (Fig.  S11). For a greener approach, the above chemical kinetics was studied over a liquid-CO 2 medium for 60 min. There is a 99% conversion of citronellal to about 99% of selectivity towards isopulegol isomers attained in 45 min (Fig. 7a,b). Sn-B-NaYZE have been shown promising results with almost enantioselectivity to (−)-isopulegol ( Table 4). The enantiospecific analysis was revealed (−)-isopulegol (94%), (+)-neoisopulegol (3%), (+)-neoisoisopulegol (2%), and (+)-isoisopulegol (1%) from (+)-citronellal. Thus, the present composites have lower SA, but improved pore diameter and strong acidic character, which facilitated the exceptional conversion rate and selectivity to isopulegol. From the HRTEM image, it was observed that the Sn preferred spherical layer type of arrangements in Sn-B-NaYZE with lattice plane distance between 0.25 and 0.35 nm to facilitate the carbonyl-ene rearrangement. The reaction probably took place on Lewis and Bronsted acidic sites of the composites leading to a protonated citronellal, which is further transformed to more stable carbocation via intermolecular rearrangement followed by deprotonation to isopulegol (Fig.  S9). Further, the experimental figures obtained against logarithmic of reactant concentration vs time revealed that the reaction is followed by first-order kinetics (Fig. 8). Therefore, the reaction was dependent on the initial concentration of citronellal, and the decrease of concentration was measured along with the progress of the reaction. As displayed, both the composites are associated with trigonal and tetragonal geometry, and there is a high possibility for the effective and selective transformation of carbonyl-ene rearrangement to isopulegol (Fig. S12). The present catalytic synthesis was equally effective for the conversion of either (+)-citronellal or its racemic mixture. Similarly, the developed composites were not only selective and efficient for the conversion of pure citronellal, but they were also equally effective in citronellal-rich essential oils. Table 6, shows the comparison of catalyst activity in the cyclization of citronellal to isopulegol with the reported literature. Meso-microporous-ZSM-5 transformed 85.1% of citronellal to 82% of isopulegol in 3 h at atmospheric pressure [26]. The transformation of citronellal (90%) to isopulegol (97%) was reported using 25%WO 3 /Al 2 O 3 /SiO 2 catalyst in 24 h [27]. The acidic site of 269-293 μmol g −1 was responsible for this carbonyl-ene rearrangement. S-Zr/ MMT-15 (Zr sulfated MMT-15) and Ru(Bpy)-Saponite were effective to produce 84.56% and 89.22% of isopulegol under microwave irradiation [28,29]. Zr content and sulfation increase the acidity (1.15 mmol g −1 ) of the catalyst, which further facilitated the isopulegol conversion [28]. In addition, the solvent (isopropanol) facilitated H-transfer under microwave irradiation and provided the stability to the intermediate complex [29]. 1%Fe-Beta produced 94% of isopulegol in 24 h. Fe (1%) modified Beta had weak and strong acidic sites of 388 and 215 μmol g −1 , respectively to facilitate the isopulegol conversion [30]. SiO 2 -Al 2 O 3 was the effective catalyst to produce 97.7% of isopuegol under 5 bar pressure [31]. While the study on the influence of solvent found that the semipolar and non-poar solvents were most effective for isopulegol conversion. The selectivity concerning solvent is as follows the order of chloroform > toluene > cyclohexane. The citronellal was partially adsorbed over SiO 2 -Al 2 O 3 and leading to high-density acidic sites in chloroform as solvent. In addition, this catalyst was associated with strong Lewis acidic sites and weak Bronsted acidic sites with a ratio of 0.79:0.21 were facilitated the carbonyl-ene rearrangement [31]. But all these discussed processes [26][27][28][29][30][31] are not mentioned the percentage of (−)-isopulegol, which is a key enantiomer for (−)-menthol synthesis.

Reduction of Isopulegol Isomers
For the reduction of isopulegol isomers, different metaldoped activated carbon catalysts were prepared and utilized at fixed reaction conditions to screen out the effective one (Fig. S1). Among, 1%Pd/AC and 1%Ru/AC catalysts were shown the better activity. Further, the effect of temperature  Table S4. The kinetics of reaction and distribution of menthol enantiomers with time using 1%Pd/AC and 1%Ru/AC catalysts are presented in Fig. 9. This reduction was attained by 99% conversion of isopulegols to menthols in 1 h at 60 °C and 40 psi H 2 pressure using 1%Pd/AC with 98% selectivity towards (±)-menthol. While, in the case of 1%Ru/AC catalyst the conversion percentage of isopulegol was quite comparable but the selectivity to (±)-menthol isomers was reduced. From the comparative study, it was noticed that the high metal loading and high temperature lead to enhance yield of un-desirable menthol isomers. Therefore optimized temperature, pressure, and catalyst loading not only helped in conversion efficiency but also lead the pathway for semisynthesis of enantiospecific (−)-menthol.

Composites Activity
The rearrangement of citronellal to isopulegol is the ratelimiting step (Fig. 4e). These composites were associated with both Lewis and Bronsted acid sites, which facilitated the cyclization of citronellal. The composites comprised of Al, Si, and Sn were provided Lewis acid sites and properly coped with B for the generation of Bronsted acid sites.
Composites prepared using different ratios of Al/Sn, YZE/ Na-YZE, and B have been studied for the conversion of citronellal to isopulegol (Fig. S13, Table S1). The results showed that 0.5-1.0%B and 4-6% of Al/Sn in YZE/NaYZE are effective combinations for the conversion of citronellal to isopulegol (Table S1). These composites associated with high Lewis and Bronsted acidity with improved pore diameter facilitated the composite activities towards the cyclization reaction. Metal-impregnated ZE were effective catalysts for monoterpene valorization [37,38]. Ru/H-beta-300 extrudates were utilized for citronellal conversion in a continuous flow reactor and they obtained 67-73% of (−)-menthol [37]. Maki-Arvela et al. [8] reported the isopulegol dimers (di-isopulegyl ethers) in the process of cyclization, but the present composites are very selective towards isopulegol without any unwanted dimers. Whereas in the current process, the total 98% conversion to menthol isomers is attained with 95% of enantioselectivty to (−)-menthol.

NMR Study of Natural and Synthetic Menthol
(−)-Menthol, sourced from semi-synthesized essential oils (present process) and also from the petrochemical origin (synthetic available in the market) have been displayed in the same 13 C-NMR spectra (Fig. S14). 13 Fig. S15a,b). From the DEPT study, it is found that C-2, C-5, and C-6 are secondary carbon and others are either primary or tertiary carbons (Fig. S14c). It is clear that menthol from natural (essential oil) and synthetic (petrochemical origin) (−)-menthol cannot be differentiated by chiral-GC-FID and NMR analyses.

Biobased Index
The enantiomers of menthol are differentiated by Chiral-GC-FID and 13 C-NMR, however, the 14 C-radiocarbon analysis of synthetic (petrochemicals) and natural (−)-menthols are expected to give dissimilar biobased index ( 14 C content). The bio-based index (100%) of the isolated (−)-menthol is shown a higher pMC value (> 99%), and it is very close to the menthol obtained from the M. arvensis essential oil. On the other hand, the synthetic menthol prepared by the chemical industry from petrochemicals has been shown a biobased value of 0% (Table 5). This is an interesting report on the origin authentication of (−)-menthol using 14 C-radiocarbondating to validate it as nature identical.

Re-usability of Catalyst
The reusability of the heterogenous catalyst is the most promising feature of this conversion process. The isopulegol/ menthol conversion rate for both the steps using Sn/Al-B-Na-YZE or 1%Pd/AC were stable up to four consecutive runs with only a maximum 5% loss in yield indicating the high reusability potential of the above catalysts, thereby making the overall process economical (Fig. 11). The reuse catalysts SAA, pore volume up to three consecutive runs are presented ( Table 6). Only some slight variation in the SAA of reused catalyst was observed as compared to the fresh catalyst, which signified their true heterogeneous nature of catalysts.

Conclusions
Al-B-NaYZE and Sn-B-NaYZE have displayed enhanced catalytic activity for carbonyl-ene rearrangement of citronellal to isopulegol, without the formation of any isopulegol dimers. Composites facilitated the enantiospecific semisynthesis of isopulegol due to the formation of the complex (Na-B-Si-Al/Sn). The Al, Si, and Sn were found responsible for the promotion of Lewis and Bronsted acidic sites. The advantage of Sn-B-NaYZE over the other composites was mainly due to the exceptional enhanced Lewis and Bronsted acidic properties along with d 3/2 and d 5/2 oxidation states. Further, Sn-B-NaYZE has shown very selective properties in terms of conversion of citronellal to isopulegol, which is a key rate-limiting step. The preparations of these composites are reported first time, where efficiency is maintained in five consecutive runs. Sn-B-NaYZE has displayed the  distribution of Sn particles in spherical layers with its crystal lattice distance of 0.25-0.35 nm, this is further confirmed by calculating the d-spacing values of XRD data. The organic solvents were replaced by green solvent (liquid CO 2 ) for cyclization of citronellal at ambient temperature in 45 min with better enantiospecificity. (+)-Citronellal was selectively converted to 99% of (−)-isopulegol in liquid-CO 2 (step-I), and it is further reduced by 1%Pd/AC to produce (−)-menthol of 98% purity in ethanol (step-II). It is demonstrated that the present process is effective for the selective conversion of citronellal in essential oils to menthol. Menthol was easily isolated from the modified oils by the process of slow freezing to -40 °C. Further, esterification of racemic menthol mixture to obtain (−)-menthyl acetate at 35 °C, followed by hydrolysis to acquire (−)-menthol. The isolated yield of (−)-menthol from Eucalyptus and Citronella essential oils was 30 and 20%, respectively.
Most importantly, our group has established a process to obtain flavor important (−)-menthol from low-value essential oils (fully citronellal transformation) directly using an environment-friendly and economical route. The liquid-CO 2 is a promising solvent and the conditions are maintained only by the cylinder pressure. This solvent is very much ideal for thermolabile terpenoids and gave enantiospecific (−)-menthol (> 90%) from citronellal. The present comprehensive semi-synthesis process has the potential to fill the gap of nature-identical (−)-menthol demand. This semi-synthesis (−)-menthol was proven to be nature-identical with a 100% of the biobased index as compared to menthol sourced from petrochemical with almost zero morden carbon (< 1%). The modified oils, as well as the spent oils have enhanced olfactometry profiles and better alternative feedstocks for cosmeceutical applications. The enrichment of minor terpene alcohols such as 1,8-cineole, terpinen-4-ol, α-terpineol, dihydro-citronellol, citronellol, citronellyl acetate, geranyl