Acylation of trans -2-substituted cyclohexanols: the impact of substituent variation on the pyridine-induced reversal of diastereoselectivity

Numerous methods for the stereoselective synthesis of chiral compounds exist in the literature, which use chiral templates or catalysts. Only in a limited number of cases have achiral catalysts been shown to significantly influence the stereochemical outcomes of reactions. Previous studies in our laboratories have revealed the achiral acyl-transfer catalyst pyridine to alter the stereochemical outcome of the reaction of racemic trans -2-substituted cyclohexanols with racemic 2-chloropropionyl chloride and cause a reversal of diastereoselectivity. The current paper presents the application of the reaction scheme to a wider number of substrates and reveals the importance of the heteroatom in the trans -2-substituent.


Introduction
2][3] Chiral substrates influence the stereochemical outcomes of their own transformations. 4In addition, stereoselectivity can be impacted by a number of factors, such as solvent polarity or temperature.6][7] In organometallic catalysis, chiral ligands can be utilized to control enantio-or diastereoselectivity.In some cases, even achiral additives have been shown to influence the stereochemical course of reactions, [8][9][10] and π-π stacking can play an important role as well. 11,12Our lab has previously shown that pyridine and derivatives thereof are able to catalyze the acylation of several racemic trans-2-substituted cyclohexanols, carrying e.g.tolylsulfanyl-or tolyloxy-substituents, with (±)-2-chloro-propionyl chloride while at the same time reversing the diastereoselectivity of the reaction. 13A linear correlation between the amount of pyridine used and the diastereoselective outcome of the reaction was observed. 14The purpose of this study was to further investigate this relationship and to probe the substrate scope for the reversal of diastereoselectivity with pyridine.Here, we provide a rationale for the previously observed catalyst-loading effect and show that pyridine is able to reverse the diastereoselectivity for the reaction of a variety of trans-2substituted cyclohexanols with (±)-2-chloropropionyl chloride ((±)-14) and (±)-2-chloro-2phenylacetyl chloride ((±)-15) in the presence of an auxiliary base, as depicted in Scheme 1. Scheme 1. Reaction scheme for the acylation of racemic trans-2-substituted cyclohexanols with racemic acyl chlorides in the (i) absence or (ii) presence of pyridine (0.1 eq) and ProtonSponge® (1.0 eq) as auxiliary base.R 1 = variable (see Scheme 2), R 2 = Me ((±)-14), R 2 = Ph ((±)-15.

Results and Discussion
In our initial studies we focused in greater detail on the concentration-dependence of the observed diastereomeric ratio (dr) on pyridine.While equimolar amounts of pyridine or dimethylaminopyridine (DMAP) led to the reversal of dr compared to the uncatalyzed reaction (Table 1, Entries 1+3), use of catalytic amounts gave dr's closer to the uncatalyzed reaction (Table 1, Entry 2).Additionally, our studies had previously revealed that non-nucleophilic bulky bases such as trimethylamine gave approximately 1:1 ratio of diastereomers for the product (Table 1, Entry 4).However, when using DMAP in catalytic quantities (0.1 equivalents) with one equivalent of trimethylamine, the overall yield was improved and the observed dr was comparable to the case of equimolar amounts of DMAP (Table 1, Entry 5).This effect of an auxiliary base was also seen when using a basic ion exchange resin (Amberlite IRA-400 OH; Table 1, Entry 6).
Table 1.Removal of catalyst-load dependence through the addition of auxiliary base, using 0.1 mmol amounts of (±)-1 and (±)-14 in 1 mL CH 2 Cl 2 in the presence of 0. The reason for this improvement in diastereoselectivity and yield through the addition of an auxiliary base is possibly the more effective neutralization of the hydrochloric acid which is released as a side-product of the reaction.When pyridine (or DMAP) is used in catalytic amounts, it is protonated by HCl over the course of the reaction and thus becomes unavailable for catalysis.This reverts the reaction back to the non-catalyzed reaction pathway which yields opposite diastereoselectivity.The effect is an overall reduction of the dr.The addition of an auxiliary base causes its protonation in the place of pyridine, leaving the latter to remain available to catalyze the acylation reaction.The effect is an overall improvement of the dr.Next, the reaction conditions were optimized to give highest yield.It was noted from previous experiments that despite full consumption of (±)-2-chloropropionyl chloride only relatively low yields of the desired ester were obtained.Thus, screening of the alcohol to acyl chloride ratio revealed the 2:1 ratio thereof to give optimal ester formation for use in further reactions (Table 2).
As shown in Table 3, a small number of potential auxiliary bases were screened to find optimal conditions for the reaction of racemic trans-2-substituted cyclohexanols with racemic acyl chlorides.Based on the best combination of yield and dr, ProtonSponge® (1,8bis(dimethylamino)-naphthalene was chosen as a homogeneous auxiliary base to be employed in the reaction.
Table 2. Screening of alcohol (±)-1 to acyl chloride (±)-14 ratio for optimal yield at a 0.1 mmol scale in 1 mL CH 2 Cl 2 in the presence of 0.1 mmol pyridine for 24 h at rt. ( a Determined by 1 H NMR.) Entry ROH (eq.) AcylCl (eq.) Consumption a (Acyl Cl) Ester Yield a With these improved catalytic reaction conditions in hand, we set out to investigate the source of the reversal of diastereoselectivity in the reaction.For this purpose, a small library of racemic trans-2-substituted cyclohexanols, carrying a variety of OR, SR and CH 2 R motifs (where R=aryl, alkyl), was generated for substrate screening.The alcohols were allowed to react with two commercially available racemic acyl chlorides, (±)-2-chloropropionyl chloride (±)-14 and (±)-2-chloro-2-phenylacetyl chloride (±)-15.Compounds (±)-1-7 and (±)-10 were synthesized via basic epoxide opening of cyclohexene oxide in ethanol, while the carba-analog of (±)-3 ((±)-12) was obtained from benzyl magnesium bromide and cyclohexene oxide in THF.The epoxide opening with electron-deficient nitrophenol was unsuccessful even under basic conditions.Compound (±)-8 with a cyclohexyloxy substituent was obtained from the reaction of cyclohexene oxide and cyclohexanol under basic conditions.The OtBu-substitutent ((±)-9) was installed with catalytic amounts of Cu(BF 4 ) 2 as activator for the epoxide opening of cyclohexane oxide with tertbutanol.Compound (±)-11 was synthesized from 1 through hydrogen peroxide oxidation in acetic acid.A second carba-analog ((±)-13) was commercially available.
The collection of racemic trans-2-substituted cyclohexanols was then reacted with (±)-2chloropropionyl chloride in dichloromethane for 24 h at room temperature either (i) without or (ii) with the addition of pyridine and ProtonSponge® based on the established optimal reaction conditions elaborated above.Yield and dr were determined from the crude reaction mixture by 1 H NMR analysis and the diastereomers were assigned arbitrarily as A and B, with A being the diastereomer giving the more deshielded signal of the proton signal used for the determination of the ratio of diastereomers (compare Experimental section).The results of the substrate screening are summarized in Table 4.

Scheme 2. Generated library of racemic trans-2-substituted cyclohexanols.
Little change in the dr in either (i) the absence or (ii) presence of pyridine and ProtonSponge® was observed when the aromatic moiety on the trans-2-substituent carried metaor para-substituents (Table 4, Entries 1-6).The larger naphthyloxy-substituent in (±)-4 also did not lead to higher dr in (ii), which indicated that extended π-π-stacking or other electronic effects on the aromatic moiety aside from the phenyl ring were not significant.
However, (±)-7 and (±)-11 showed a strong increase in dr in the catalyzed reaction.This may be a result of an increased bulk in the substrate close to the reaction center or a significant change in conformation of the substituent relative to the cyclohexanol moiety.Surprisingly, it was observed that alkyloxy-substituted compounds (±)-8 and (±)-9 gave higher absolute dr in the uncatalyzed reaction, contrary to the trend observed in aryloxysubstituted compounds.Although arguments related to electronics or sterics of these substituents and potential interactions with (±)-14 could be made, the exact reason for this observation remains unknown and further exploration is needed to elucidate the cause.Compound (±)-10 showed no reversal of dr between (i) and (ii).This is most likely the result of the pyridine moiety on the trans-2-substituent acting as an intramolecular acyl-transfer catalyst in the absence of pyridine.The carba-analogs (±)-12 and (±)-13 both did not show a reversal of diastereoselectivity upon the addition of pyridine and auxiliary base.This appears to be direct consequence of the lack of a heteroatom on the trans-2-substituent.The heteroatom may significantly influence the transition state energies in the reaction with the acyl transfer catalyst relative to the reaction without it or even actively participate in the mechanism, as previously proposed. 13A heteroatom (oxygen or sulfur, in the cases above) may thus be essential for the reversal of dr.A small selection of racemic trans-2-substituted cyclohexanols was also subjected to (±)-2chloro-2-phenylacetyl chloride (±)-15 in the same reaction scheme as above.The results are shown in Table 5.The initial expectation that a bulkier acyl chloride would yield higher dr was not borne out by the data.Although reversal of diastereoselectivity was seen generally, the dr was lower in the majority of cases with (±)-15 than with (±)-14.The reason for this may be that the replacement of a methyl group with a phenyl group in the γ-position on the acyl chloride leads to steric crowding in the transition state, meaning that stereodifferentiation is more successful with less sterically demanding acyl substrate.Curiously, no reversal of diastereoselectivity was observed for (±)-8 (Table 5, Entries 7 and 8).The reason for this lack of reversal is not immediately apparent and warrants further study.
In order to further investigate the reason for the reversal of dr with the addition of pyridine, the transition state structures of the ester formation were computed at the B3LYP/6-31G* level of theory for (±)-1, (±)-2 and (±)-13 with (±)-14 and the acyl-pyridinium intermediate of (±)-14, respectively, under basic conditions (alcoholate).However, the reversal of diastereoselectivities due to the introduction of pyridine was not borne out quantitatively by the computed relative transition state energies for the different diastereomeric transition states.(For full details, transition state energies and structures, see Supplemental Material.)Modeling of the reaction at a higher level of theory and potentially with consideration of the solvent may give a more accurate representation of the experimentally observed situation, especially given the fact that the observed dr's would correspond to relatively small differences in the transition state energies.Instead, a qualitative interpretation of the computational results revealed some interesting observations that could help explain the experimentally observed reversal of dr. Figure 1 depicts a characteristic selection of the structures of the lowest transition states found for the respective reactions of (±)-2 and (±)-13 with their LUMO shown.The structures and geometric coordinates for all diastereomeric transition states can be found in the Supplemental Materials.Notably, the lowest transition states had the trans-2-substituted cyclohexanol configured with both substituents in the axial position for (±)-2.
The LUMO was found to be localized almost entirely in the pyridinium ring for the catalyzed reaction in all cases.In the catalyzed reaction, there appears to be an interaction of either the heteroatom or the aromatic moiety on the trans-2-substituent with the pyridinium ring, on which the LUMO is localized, causing the trans-2-substituent to be in close vicinity.This is not the case for (±)-13, where both heteroatom and aromatic moiety are absent.As a result, no clear preference in conformation of the transition state was observed.This interaction is also notably absent in the uncatalyzed reaction and no such interaction could be inferred between the heteroatom or the aromatic moiety and the chlorine leaving group for (±)-2.These computations suggest that the presence of a heteroatom and/or aromatic moiety on the trans-2-substituent of cyclohexanol causes it to interact favorably with the pyridinum species in the transition state.This overall change in conformation from the uncatalyzed to the catalyzed reaction for (±)-2 may be the reason for the observed reversal in diastereoselectivity.

Conclusions
After optimization of the catalytic reaction conditions, using auxiliary base to overcome the catalyst-load dependence on the dr, thirteen racemic trans-2-substituted alcohols were screened with two racemic acyl chlorides for reversal of dr.Highest dr was found with cyclohexyloxysubstituted (±)-8 in the case of the uncatalyzed reaction with (±)-15 (dr 1:10) and in the pyridinecatalyzed reaction with the tolylsulfonyl-containing compound (±)-11 with (±)-14 (dr 15:1).Stereoselectivity was generally higher with the less sterically demanding acyl chloride (±)-14as opposed to (±)-15.The heteroatom on the trans-2-substituent appears to be essential for the reversal of dr to be observed.Computational modeling of the reaction points to the importance of the heteroatom and/or aromatic moiety on the trans-2-substituent, as well.With further improvements to the dr, especially through modification close to the alcohol functionality of the cyclohexanols substrate, this method may provide valuable in the stereodivergent resolution of racemic acyl chlorides.

Experimental Section
General.Sodium borate was purchased from EMScience, Copper (II) tetrafluoroborate from Alfa Aesar, Naphthanol from Matheson Coleman & Bell and Phenol from Mallinckrodt.All other reagents were obtained from Sigma-Aldrich and used without further purifications.Solvents were distilled prior to use.Column chromatography was performed on silica gel (Sorbent Technologies, 40-75 µm) and fractions analyzed with TLC run on equivalent mobile phase and analyzed through UV or charring with H 2 SO 4 /MeOH.Melting points were determined using a Stanford Research Systems Digimelt MPA160. 1 H-NMR and 13 C-NMR spectra were acquired on a JEOL ECA-600 NMR-spectrometer (600 and 150 MHz, respectively).Structural assignments were corroborated by homonuclear and heteronuclear 2D NMR methods (COSY, HMQC, HMBC and TOCSY where necessary).Accurate mass measurements were performed on a JEOL Direct Analysis in Real Time (DART) Mass Spectrometer with AccuTOF mass analyzer (Peabody, MA, USA) with polyethyleneglycol as an internal calibrant.Samples were ionized directly from a drop of solution on the tip of a glass capillary under ambient conditions without sample preparation. 1H and 13 C NMR spectra and DART-MS spectra of all compounds are supplied in the Supplemental Material.

Compound (±)-2 [(±)-trans-2-(p-tolyloxy)cyclohexanol].
To 50 mL of a 0.5 M solution of Na in absoluted ethanol was added cyclohexene oxide (5.0 mL, 49.4 mmol) and p-cresol (5.36 g, 49.6 mmol) with stirring and the reaction was heated to 80 ºC with stirring for 24 hours.The yellow solution was then cooled, quenched with 10 mL H 2 O and neutralized using conc.HCl acid.The solution was then diluted with 50 mL of CH 2 Cl 2 and transferred to a separatory funnel.The organic layer was separated and the aqueous layer washed with CH 2 Cl 2 (20 mL).The combined organic layers (yellowish liquid) were washed with H 2 O (30 mL) and sat.NaCl solution (30 mL) consecutively and then dried over Na 2 SO 4 .The drying agent was filtered off and the solvent evaporated to give 10.02 g of a tan solid.The crude product was recrystallized from hexane and combined with a second crop of crystals from the filtrate to yield 7.24 g (35.1 mmol, 71%) of white needle-like crystals.mp 84-87 ºC. 1  ].To 10 mL of a roughly 0.4 M solution of Na in ethanol was added cyclohexene oxide (1.0 mL, 9.9 mmol) and phenol (0.93 g, 10.2 mmol) with stirring and the reaction mixture was heated to gentle reflux with continued stirring for 24 hours.The solution was then cooled and quenched with 10 mL water.Then, it was neutralized using conc.HCl acid.The yellow liquid was then diluted with 10 mL of CH 2 Cl 2 and transferred to a separatory funnel.The organic layer was separated and the aqueous layer washed with CH 2 Cl 2 (2 x 10 mL).The combined organic layers (yellow liquid) were washed with 15 mL sat.NaCl solution and dried over Na 2 SO 4 .The drying agent was filtered off and the solvent evaporated to give an off-white to yellow solid.The crude product was recrystallized from hexane to yield 0.89 g (4.6 mmol, 47%) of fine white crystals.mp 83-84 ºC. 1

Compound (±)-4 [(±)-trans-2-(napthalen-2-yloxy)cyclohexanol].
In a 100 mL round-bottom flask equipped with reflux condenser was placed a previously prepared solution of NaOH in ethanol (pH 14, 18 mL) to which was added cyclohexene oxide (2.0 mL, 20.0 mmol) and 2naphthanol (2.95 g, 20.5 mmol).The sandcolored suspension was stirred and heated to about 90 ºC (oil bath temperature), which caused 2-naphthanol to dissolve and gave a clear brown solution.Reaction progress was monitored via TLC (CH 2 Cl 2 ).After consumption of starting material, solution was allowed to cool to give a light-brown to yellow solution with off-white solid.The suspension was then diluted with 5 mL water and neutralized using conc.HCl.The product was then extracted using 20 mL CH 2 Cl 2 and the aqueous layer washed twice with 10 mL CH 2 Cl 2 .The organic layers were combined and dried over Na 2 SO 4 .After filtering off the drying agent, solvent was removed to give a sandcolored solid, which was recrystallized from ethanol to yield the product as fine white needle-shaped crystals (2.48 g, 52 %).mp 136-137 ºC. 1   14).A reflux condenser was attached and the reaction heated to reflux using an oilbath.The dark brown solution was heated for 6 hours, then allowed to cool to RT.The solution was diluted with 15 mL H 2 O and then extracted using 30 mL of CH 2 Cl 2 .The aqueous layer was washed twice more using 10 mL of CH 2 Cl 2 each and the combined organic layers were dried over Na 2 SO 4 .The drying agent was filtered off and solvent removed using a rotavap to give a brown viscous oil which solidified after cooling.The crude product was recrystallized from hexane to yield offwhite to tan crystals (3.38 g, 81%).mp 76-81 ºC. 1

Compound (±)-6 [(±)-trans-2-(p-tert-butyl-phenyloxy)cyclohexanol].
In a 100 mL roundbottom flask equipped with reflux condenser was placed a previously prepared solution of Na in ethanol (0.5 M, 20 mL) to which was added cyclohexene oxide (2.0 mL, 19.8 mmol) and p-tertbutylphenol (3.21 g, 21.4 mmol).The reaction mixture was heated to about 90 ºC (oil bath temperature), giving a clear solution.Reaction progress was monitored via TLC (CH 2 Cl 2 ) and DART-HRMS.After consumption of starting material (165 min), solution was allowed to cool to give a light-yellow solution.The solution was then diluted with 10 mL water and neutralized using conc.HCl, during which a white solid precipitated out.Then, 20 mL CH 2 Cl 2 was added to extract the product, giving two opaque colorless layers and the organic layer was separated.The aqueous layer was extracted twice with 12.5 mL CH 2 Cl 2 .The organic layers were combined and dried over Na 2 SO 4 .After filtering off the drying agent, solvent was removed to give an off-white solid.The crude product was recrystallized twice from hexane to yield the product as fine white needle-shaped crystals (1.12 g, 23 %).mp 88-91 ºC. 1

Compound (±)-7 [(±)-trans-2-(2,6-dimethylphenyloxy)cyclohexanol].
To a solution of Na (0.1 g) in ethanol (10 mL) was added cyclohexene oxide (1.0 mL, 9.9 mmol) and 2,6-dimethylphenol (1.25 g, 10.2 mmol) with stirring and the reaction mixture was heated to gentle reflux for 24 hours.The solution was then cooled and quenched with 10 mL water.Then, it was neutralized using conc.HCl acid.The dark-brown liquid was then diluted with 10 mL of CH 2 Cl 2 and transferred to a separatory funnel.The organic layer was separated and the aqueous layer washed with CH 2 Cl 2 (2 x 10 mL) until a nearly clear yellow aqueous layer remained.The combined organic layers (dark-brown liquid) were washed with 15 mL sat.NaCl solution and dried over Na 2 SO 4 .The drying agent was filtered off and the solvent evaporated to give a dark-brown liquid (2.0 g).The crude product was purified via column chromatography (mob phase 90:10 hexane:ethyl acetate) to yield 1.52 g (6.9 mmol, 80%) of a clear pale-yellow liquid (R f = 0.30). 1