Separation of Alkyne Enantiomers by Chiral Column HPLC Analysis of Their Cobalt-Complexes

Separation of the enantiomers of new chiral alkynes in strategic syntheses and bioorthogonal studies is always problematic. The chiral column high-performance liquid chromatography (HPLC) method in general could not be directly used to resolve such substrates, since the differentiation of the alkyne segment with the other alkane/alkene segment is not significant in the stationary phase, and the alkyne group is not a good UV chromophore. Usually, a pre-column derivatization reaction with a tedious workup procedure is needed. Making use of easily-prepared stable alkyne-cobalt-complexes, we developed a simple and general method by analyzing the in situ generated cobalt-complex of chiral alkynes using chiral column HPLC. This new method is especially suitable for the alkynes without chromophores and other derivable groups.


Introduction
Alkyne is one of the fundamental groups in organic chemistry, which exists widely in natural products and unnatural functionalized molecules [1]. The development of modern transition-metal-catalyzed reactions, such as cross coupling [2], Pauson-Khand reaction [3], click chemistry [4], etc., has caused alkynes to play a more and more important role in strategic syntheses and bioorthogonal design [1].
Enantiomeric purity is vital for chiral alkynes to function. The enantiomeric excess (ee) of chiral alkynes with unknown optical rotation could be determined generally in two ways: (1) using another chiral reagent to derivate diastereomers and then measure the diasteromeric ratio; (2) using enantioselective chromatography to directly measure the enantiomeric excess. Recently, in the course of our total synthesis of natural products, we prepared the known alkyne 1a (Figure 1), and found that two methods were used in the literatures to determine its ee [5][6][7][8][9][10][11][12][13]. Most literature used Mosher's methodology [14] which undertook an esterification with expensive Mosher's chiral reagent, and determined the enantiomeric purity by NMR analysis of the resultant diasteromeric mixture [5][6][7][8][9][10][11][12]. One study used high-performance liquid chromatography (HPLC) with chiral column to analyze the 3,5-dinitrobenzoate derivative of 1a [13]. Since both methods need a derivatization reaction with tedious workup procedure, we initially attempted a direct analysis of 1a with chiral column HPLC It was noted that most of the above HPLC experiments were carried out with the same chiral column using the most general 2-PrOH/n-hexane eluting-solvent system. Except for changing the 2-PrOH/n-hexane ratio to make the retention time between 10-20 min, no more optimization was done for HPLC conditions. The simple setup of the HPLC conditions implied the easiness and generality of our Co-complex analytic method on much wider alkyne substrates.
Although the above HPLC analysis was all carried out with purified Co-complexes, since the Aware of the insufficiency of the present ee determination methodology for chiral alkynes such as 1a, we decided to develop and report herein a simple and general method by analyzing the in situ generated cobalt-complex of chiral alkynes using chiral column HPLC. This new method is especially suitable for the alkynes without chromophores and other derivable groups.

Results and Discussion
To overcome the above-mentioned two weak points of alkynes, our idea was to make use of its metal complex. It is well known that Co 2 (CO) 8 can easily react with alkynes to lose two carbon monoxides and form stable Co 2 (CO) 6 -alkyne complexes. Figure 1a showed this transformation from alkyne 1a to its complex 2a. By this simple complexation, (1) the alkyne-chain became significantly bulky, which was very different in size from the other side-chain; (2) the Co-complexation substructure was a good chromophore, making the UV absorption of 2a more than 100 times stronger than 1a (Figure 1b,c), and a wide range of detection wavelength (200-400 nm) could be selected in the HPLC experiment. As a result, comparison with the unsuccessful result of 1a (Figure 1d), 2a was easily resolved by HPLC using a CHIRALPAK-IB column with 0.4:99.6 2-PrOH/n-hexane eluting-solvent system, 1 mL·min −1 flow rate, and 350 nm detection wavelength at 25 • C ( Figure 1e). Under the same conditions, an enantioenriched 2a was also tested (Figure 1f), and the detected enantiomeric ratio was consistent with that determined by the modified Mosher's methodology [11].
After confirmation of our idea on compounds 1a and 2a, we next prepared a series of Co 2 (CO) 6 -alkyne complexes 2b-k (Table 1) from none-chromophore alkynes, and checked their resolution by HPLC chiral column. As a result, all the enantiomeric pairs of Co-complexes were successfully baseline-separated. For alcohol substrates, the size of another side-chain (butyl, allyl, ethyl, methyl) or the distance between the C≡C triple bond and the hydroxyl group did not show any notable effect for the enantiomeric separation (2a-e). Even a tertiary alkynol-complex 2f could be well resolved, although a CHIRALPAK-IA column was used in this case. Internal alkyne-complexes 2g and 2h also gave good separation. Replacing the hydroxyl group with halogen atoms might change the interaction between the substrates and the stationary phase of the column. Fortunately, the chloride 2i and the fluoride 2j were both resolved, although lower polar eluting-solvent systems were needed for the low polarity of these substrates. Notably, the tetrahydropyran (THP)-protected alkynol-complex 2k could also been enantiomerically separated. Actually, compounds such as 1i-k, are not good substrates for classical pre-column derivatization methods. Aware of the insufficiency of the present ee determination methodology for chiral alkynes such as 1a, we decided to develop and report herein a simple and general method by analyzing the in situ generated cobalt-complex of chiral alkynes using chiral column HPLC. This new method is especially suitable for the alkynes without chromophores and other derivable groups.

Results and Discussion
To overcome the above-mentioned two weak points of alkynes, our idea was to make use of its metal complex. It is well known that Co2(CO)8 can easily react with alkynes to lose two carbon monoxides and form stable Co2(CO)6-alkyne complexes. Figure 1a showed this transformation from alkyne 1a to its complex 2a. By this simple complexation, (1) the alkyne-chain became significantly bulky, which was very different in size from the other side-chain; (2) the Co-complexation substructure was a good chromophore, making the UV absorption of 2a more than 100 times stronger than 1a (Figure 1b,c), and a wide range of detection wavelength (200-400 nm) could be selected in the HPLC experiment. As a result, comparison with the unsuccessful result of 1a (Figure 1d), 2a was easily resolved by HPLC using a CHIRALPAK-IB column with 0.4:99.6 2-PrOH/n-hexane eluting-solvent system, 1 mL·min −1 flow rate, and 350 nm detection wavelength at 25 °C ( Figure 1e). Under the same conditions, an enantioenriched 2a was also tested (Figure 1f), and the detected enantiomeric ratio was consistent with that determined by the modified Mosher's methodology [11].
After confirmation of our idea on compounds 1a and 2a, we next prepared a series of Co2(CO)6-alkyne complexes 2b-k (Table 1) from none-chromophore alkynes, and checked their resolution by HPLC chiral column. As a result, all the enantiomeric pairs of Co-complexes were successfully baseline-separated. For alcohol substrates, the size of another side-chain (butyl, allyl, ethyl, methyl) or the distance between the C≡C triple bond and the hydroxyl group did not show any notable effect for the enantiomeric separation (2a-e). Even a tertiary alkynol-complex 2f could be well resolved, although a CHIRALPAK-IA column was used in this case. Internal alkyne-complexes 2g and 2h also gave good separation. Replacing the hydroxyl group with halogen atoms might change the interaction between the substrates and the stationary phase of the column. Fortunately, the chloride 2i and the fluoride 2j were both resolved, although lower polar eluting-solvent systems were needed for the low polarity of these substrates. Notably, the tetrahydropyran (THP)-protected alkynol-complex 2k could also been enantiomerically separated. Actually, compounds such as 1i-k, are not good substrates for classical pre-column derivatization methods. Aware of the insufficiency of the present ee determination methodology for chiral alkynes such as 1a, we decided to develop and report herein a simple and general method by analyzing the in situ generated cobalt-complex of chiral alkynes using chiral column HPLC. This new method is especially suitable for the alkynes without chromophores and other derivable groups.

Results and Discussion
To overcome the above-mentioned two weak points of alkynes, our idea was to make use of its metal complex. It is well known that Co2(CO)8 can easily react with alkynes to lose two carbon monoxides and form stable Co2(CO)6-alkyne complexes. Figure 1a showed this transformation from alkyne 1a to its complex 2a. By this simple complexation, (1) the alkyne-chain became significantly bulky, which was very different in size from the other side-chain; (2) the Co-complexation substructure was a good chromophore, making the UV absorption of 2a more than 100 times stronger than 1a (Figure 1b,c), and a wide range of detection wavelength (200-400 nm) could be selected in the HPLC experiment. As a result, comparison with the unsuccessful result of 1a (Figure 1d), 2a was easily resolved by HPLC using a CHIRALPAK-IB column with 0.4:99.6 2-PrOH/n-hexane eluting-solvent system, 1 mL·min −1 flow rate, and 350 nm detection wavelength at 25 °C ( Figure 1e). Under the same conditions, an enantioenriched 2a was also tested (Figure 1f), and the detected enantiomeric ratio was consistent with that determined by the modified Mosher's methodology [11].
After confirmation of our idea on compounds 1a and 2a, we next prepared a series of Co2(CO)6-alkyne complexes 2b-k (Table 1) from none-chromophore alkynes, and checked their resolution by HPLC chiral column. As a result, all the enantiomeric pairs of Co-complexes were successfully baseline-separated. For alcohol substrates, the size of another side-chain (butyl, allyl, ethyl, methyl) or the distance between the C≡C triple bond and the hydroxyl group did not show any notable effect for the enantiomeric separation (2a-e). Even a tertiary alkynol-complex 2f could be well resolved, although a CHIRALPAK-IA column was used in this case. Internal alkyne-complexes 2g and 2h also gave good separation. Replacing the hydroxyl group with halogen atoms might change the interaction between the substrates and the stationary phase of the column. Fortunately, the chloride 2i and the fluoride 2j were both resolved, although lower polar eluting-solvent systems were needed for the low polarity of these substrates. Notably, the tetrahydropyran (THP)-protected alkynol-complex 2k could also been enantiomerically separated. Actually, compounds such as 1i-k, are not good substrates for classical pre-column derivatization methods. Aware of the insufficiency of the present ee determination methodology for chiral alkynes such as 1a, we decided to develop and report herein a simple and general method by analyzing the in situ generated cobalt-complex of chiral alkynes using chiral column HPLC. This new method is especially suitable for the alkynes without chromophores and other derivable groups.

Results and Discussion
To overcome the above-mentioned two weak points of alkynes, our idea was to make use of its metal complex. It is well known that Co2(CO)8 can easily react with alkynes to lose two carbon monoxides and form stable Co2(CO)6-alkyne complexes. Figure 1a showed this transformation from alkyne 1a to its complex 2a. By this simple complexation, (1) the alkyne-chain became significantly bulky, which was very different in size from the other side-chain; (2) the Co-complexation substructure was a good chromophore, making the UV absorption of 2a more than 100 times stronger than 1a (Figure 1b,c), and a wide range of detection wavelength (200-400 nm) could be selected in the HPLC experiment. As a result, comparison with the unsuccessful result of 1a (Figure 1d), 2a was easily resolved by HPLC using a CHIRALPAK-IB column with 0.4:99.6 2-PrOH/n-hexane eluting-solvent system, 1 mL·min −1 flow rate, and 350 nm detection wavelength at 25 °C ( Figure 1e). Under the same conditions, an enantioenriched 2a was also tested (Figure 1f), and the detected enantiomeric ratio was consistent with that determined by the modified Mosher's methodology [11].
After confirmation of our idea on compounds 1a and 2a, we next prepared a series of Co2(CO)6-alkyne complexes 2b-k (Table 1) from none-chromophore alkynes, and checked their resolution by HPLC chiral column. As a result, all the enantiomeric pairs of Co-complexes were successfully baseline-separated. For alcohol substrates, the size of another side-chain (butyl, allyl, ethyl, methyl) or the distance between the C≡C triple bond and the hydroxyl group did not show any notable effect for the enantiomeric separation (2a-e). Even a tertiary alkynol-complex 2f could be well resolved, although a CHIRALPAK-IA column was used in this case. Internal alkyne-complexes 2g and 2h also gave good separation. Replacing the hydroxyl group with halogen atoms might change the interaction between the substrates and the stationary phase of the column. Fortunately, the chloride 2i and the fluoride 2j were both resolved, although lower polar eluting-solvent systems were needed for the low polarity of these substrates. Notably, the tetrahydropyran (THP)-protected alkynol-complex 2k could also been enantiomerically separated. Actually, compounds such as 1i-k, are not good substrates for classical pre-column derivatization methods.  As a result, 1a showed very low UV-absorption, and it was difficult to characterize under normal HPLC concentration using a UV detector. Aware of the insufficiency of the present ee determination methodology for chiral alkynes such as 1a, we decided to develop and report herein a simple and general method by analyzing the in situ generated cobalt-complex of chiral alkynes using chiral column HPLC. This new method is especially suitable for the alkynes without chromophores and other derivable groups.

Results and Discussion
To overcome the above-mentioned two weak points of alkynes, our idea was to make use of its metal complex. It is well known that Co2(CO)8 can easily react with alkynes to lose two carbon monoxides and form stable Co2(CO)6-alkyne complexes. Figure 1a showed this transformation from alkyne 1a to its complex 2a. By this simple complexation, (1) the alkyne-chain became significantly bulky, which was very different in size from the other side-chain; (2) the Co-complexation substructure was a good chromophore, making the UV absorption of 2a more than 100 times stronger than 1a (Figure 1b,c), and a wide range of detection wavelength (200-400 nm) could be selected in the HPLC experiment. As a result, comparison with the unsuccessful result of 1a (Figure 1d), 2a was easily resolved by HPLC using a CHIRALPAK-IB column with 0.4:99.6 2-PrOH/n-hexane eluting-solvent system, 1 mL·min −1 flow rate, and 350 nm detection wavelength at 25 °C ( Figure 1e). Under the same conditions, an enantioenriched 2a was also tested (Figure 1f), and the detected enantiomeric ratio was consistent with that determined by the modified Mosher's methodology [11].
After confirmation of our idea on compounds 1a and 2a, we next prepared a series of Co2(CO)6-alkyne complexes 2b-k (Table 1) from none-chromophore alkynes, and checked their resolution by HPLC chiral column. As a result, all the enantiomeric pairs of Co-complexes were successfully baseline-separated. For alcohol substrates, the size of another side-chain (butyl, allyl, ethyl, methyl) or the distance between the C≡C triple bond and the hydroxyl group did not show any notable effect for the enantiomeric separation (2a-e). Even a tertiary alkynol-complex 2f could be well resolved, although a CHIRALPAK-IA column was used in this case. Internal alkyne-complexes 2g and 2h also gave good separation. Replacing the hydroxyl group with halogen atoms might change the interaction between the substrates and the stationary phase of the column. Fortunately, the chloride 2i and the fluoride 2j were both resolved, although lower polar eluting-solvent systems were needed for the low polarity of these substrates. Notably, the tetrahydropyran (THP)-protected alkynol-complex 2k could also been enantiomerically separated. Actually, compounds such as 1i-k, are not good substrates for classical pre-column derivatization methods.  As a result, 1a showed very low UV-absorption, and it was difficult to characterize under normal HPLC concentration using a UV detector. Aware of the insufficiency of the present ee determination methodology for chiral alkynes such as 1a, we decided to develop and report herein a simple and general method by analyzing the in situ generated cobalt-complex of chiral alkynes using chiral column HPLC. This new method is especially suitable for the alkynes without chromophores and other derivable groups.

Results and Discussion
To overcome the above-mentioned two weak points of alkynes, our idea was to make use of its metal complex. It is well known that Co2(CO)8 can easily react with alkynes to lose two carbon monoxides and form stable Co2(CO)6-alkyne complexes. Figure 1a showed this transformation from alkyne 1a to its complex 2a. By this simple complexation, (1) the alkyne-chain became significantly bulky, which was very different in size from the other side-chain; (2) the Co-complexation substructure was a good chromophore, making the UV absorption of 2a more than 100 times stronger than 1a (Figure 1b (Figure 1f), and the detected enantiomeric ratio was consistent with that determined by the modified Mosher's methodology [11].
After confirmation of our idea on compounds 1a and 2a, we next prepared a series of Co2(CO)6-alkyne complexes 2b-k (Table 1) from none-chromophore alkynes, and checked their resolution by HPLC chiral column. As a result, all the enantiomeric pairs of Co-complexes were successfully baseline-separated. For alcohol substrates, the size of another side-chain (butyl, allyl, ethyl, methyl) or the distance between the C≡C triple bond and the hydroxyl group did not show any notable effect for the enantiomeric separation (2a-e). Even a tertiary alkynol-complex 2f could be well resolved, although a CHIRALPAK-IA column was used in this case. Internal alkyne-complexes 2g and 2h also gave good separation. Replacing the hydroxyl group with halogen atoms might change the interaction between the substrates and the stationary phase of the column. Fortunately, the chloride 2i and the fluoride 2j were both resolved, although lower polar eluting-solvent systems were needed for the low polarity of these substrates. Notably, the tetrahydropyran (THP)-protected alkynol-complex 2k could also been enantiomerically separated. Actually, compounds such as 1i-k, are not good substrates for classical pre-column derivatization methods.  As a result, 1a showed very low UV-absorption, and it was difficult to characterize under normal HPLC concentration using a UV detector. Aware of the insufficiency of the present ee determination methodology for chiral alkynes such as 1a, we decided to develop and report herein a simple and general method by analyzing the in situ generated cobalt-complex of chiral alkynes using chiral column HPLC. This new method is especially suitable for the alkynes without chromophores and other derivable groups.

Results and Discussion
To overcome the above-mentioned two weak points of alkynes, our idea was to make use of its metal complex. It is well known that Co2(CO)8 can easily react with alkynes to lose two carbon monoxides and form stable Co2(CO)6-alkyne complexes. Figure 1a showed this transformation from alkyne 1a to its complex 2a. By this simple complexation, (1) the alkyne-chain became significantly bulky, which was very different in size from the other side-chain; (2) the Co-complexation substructure was a good chromophore, making the UV absorption of 2a more than 100 times stronger than 1a (Figure 1b,c), and a wide range of detection wavelength (200-400 nm) could be selected in the HPLC experiment. As a result, comparison with the unsuccessful result of 1a (Figure 1d), 2a was easily resolved by HPLC using a CHIRALPAK-IB column with 0.4:99.6 2-PrOH/n-hexane eluting-solvent system, 1 mL·min −1 flow rate, and 350 nm detection wavelength at 25 °C ( Figure 1e). Under the same conditions, an enantioenriched 2a was also tested (Figure 1f), and the detected enantiomeric ratio was consistent with that determined by the modified Mosher's methodology [11].
After confirmation of our idea on compounds 1a and 2a, we next prepared a series of Co2(CO)6-alkyne complexes 2b-k (Table 1) from none-chromophore alkynes, and checked their resolution by HPLC chiral column. As a result, all the enantiomeric pairs of Co-complexes were successfully baseline-separated. For alcohol substrates, the size of another side-chain (butyl, allyl, ethyl, methyl) or the distance between the C≡C triple bond and the hydroxyl group did not show any notable effect for the enantiomeric separation (2a-e). Even a tertiary alkynol-complex 2f could be well resolved, although a CHIRALPAK-IA column was used in this case. Internal alkyne-complexes 2g and 2h also gave good separation. Replacing the hydroxyl group with halogen atoms might change the interaction between the substrates and the stationary phase of the column. Fortunately, the chloride 2i and the fluoride 2j were both resolved, although lower polar eluting-solvent systems were needed for the low polarity of these substrates. Notably, the tetrahydropyran (THP)-protected alkynol-complex 2k could also been enantiomerically separated. Actually, compounds such as 1i-k, are not good substrates for classical pre-column derivatization methods.  2 4 6 8 10 12 14 16 18 20 22 24 26        It was noted that most of the above HPLC experiments were carried out with the same chiral column using the most general 2-PrOH/n-hexane eluting-solvent system. Except for changing the 2-PrOH/n-hexane ratio to make the retention time between 10-20 min, no more optimization was done for HPLC conditions. The simple setup of the HPLC conditions implied the easiness and generality of our Co-complex analytic method on much wider alkyne substrates.
Although the above HPLC analysis was all carried out with purified Co-complexes, since the CHIRALPAK immobilized-polysaccharide type columns can tolerate a wide range of solvents, we next attempted to directly measure the reaction system of 1a, Co 2 (CO) 8 , and CH 2 Cl 2 (Figure 1a). To our delight, by just diluting a small amount of the reaction mixture with n-hexane and then injection into HPLC, the in situ generated 2a could be monitored without any problems (Figure 2). Since the manipulation was rather simple without the tedious workup procedure like those of Mosher's or other carbonyl chloride methods [5][6][7][8][9][10][11][12][13], this result implied a promising application of our Co-complexation method in monitoring the asymmetric alkyne-synthesis reactions. we next attempted to directly measure the reaction system of 1a, Co2(CO)8, and CH2Cl2 (Figure 1a). To our delight, by just diluting a small amount of the reaction mixture with n-hexane and then injection into HPLC, the in situ generated 2a could be monitored without any problems (Figure 2). Since the manipulation was rather simple without the tedious workup procedure like those of Mosher's or other carbonyl chloride methods [5][6][7][8][9][10][11][12][13], this result implied a promising application of our Co-complexation method in monitoring the asymmetric alkyne-synthesis reactions.

General Methods
All reactions were carried out under an atmosphere of Ar unless otherwise indicated. 1 H-, and 13 C-NMR spectra were acquired on Mercury-300 (Agilent, Santa Clara, CA, USA), AVANCE III-400 (Bruker, Billerica, MA, USA), WNMR-I-500 (Zhongke Niujin Co., Ltd., Wuhan, China), or VNMRS-600 spectrometers (Agilent). Chemical shifts are indicated in parts per million (ppm) downfield from tetramethylsilane (TMS, δ = 0.00) with residual undeuterated solvent peaks as internal reference for 1 H-NMR and deuterated solvent peaks shifts for 13 C-NMR. Multiplicities are reported as s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), br (broad) or combinations of those. For NMR analysis of the Co 2 (CO) 6 -alkyne complexes, the sample solution in CDCl 3 should pass through a disposable syringe filter (Nylon 66, 0.22 µm, 13 mm) immediately before the NMR experiment, to remove the small amount of paramagnetic material. Mass spectra (MS) are electron ionization (EI) or electrospray ionization (ESI). EI-MS data were measured on GCT mass spectrometer (Micromass, Manchester, UK). ESI-MS data were measured on Thermo-Fisher Accela liquid chromatography system coupled with Exactive Plus Orbitrap mass spectrometer (Thermo-Fisher, Bremen, Germany). Reagents and compounds 1c-f, 1k were purchased from commercial suppliers and used as received. Compound 1a was prepared according to the literature procedure [8]. Compounds 1b, 1g-j, and 2a-k were synthesized as following. See Supplementary Materials for spectra and chromatograms of the prepared products.

Conclusions
We have developed a simple and general method to separate the enantiomers of chiral alkynes using their in situ generated cobalt-complex by chiral column HPLC. HPLC analysis of enantiomeric purity of metal complexes with carbon-metal bonds is not usual [16]. Our method is particularly useful for the alkynes without chromophores and other derivable groups. Since the decomplexation of Co-alkyne complexes is well known [17], our method is also promisingly suitable for the preparative resolution of racemic alkynes.  13 C-NMR spectra of 1b, and 1g-j; (4) HPLC chart of 1a; (5) 1 H-, 13 C-NMR spectra and HPLC charts of 2a-k; (6) HPLC monitor of the reaction of 1a and Co 2 (CO) 8 .