Samantha M.
Gibson
,
Rachel M.
Lanigan
,
Laure
Benhamou
,
Abil E.
Aliev
and
Tom D.
Sheppard
*
Department of Chemistry, University College London, 20 Gordon St, London, WC1H 0AJ, UK. E-mail: tom.sheppard@ucl.ac.uk; Tel: +44 (0)2076 792467
First published on 29th July 2015
In this paper we describe the use of a chiral aldehyde derived from lactate esters for determining the enantiopurity of primary amines, via the formation of diastereomeric imines. The method was shown to be suitable for reproducibly determining the enantiopurity of a diverse set of chiral amines. Both enantiomers of the aldehyde can be prepared in two steps from commercially available materials.
For chiral amines, however, derivatisation methods are somewhat more difficult. If amides are prepared using a chiral acid such as 1, then the resulting compounds are frequently difficult to analyse via NMR due to the presence of rotational isomers caused by slow rotation around the C–N bond.1,2 An alternative strategy is to use a chiral aldehyde to generate an imine.3–5 This strategy is potentially useful as imine formation is often rapid. However, the aldehyde used must be fairly hindered in order to give the E-imine selectively as a single geometrical isomer, and it must also be stable towards racemisation under the conditions used for imine formation. There have been a few reports in this area, including the use of myrtenal 22,3 as well as o-formylbenzene boronic acid 3 in combination with enantiomerically enriched BINOL.4 Myrtenal 2, has not been widely employed, with the original reports simply demonstrating that the diastereomeric imines derived from racemic amines could easily be distinguished by NMR. This is, however, very different from the measurement of the enantiopurity of an amine derived from a chemical reaction in which small quantities of the minor diastereoisomer must typically be quantified as accurately as possible. The enantiomer of 2 is also not readily available. Aldehyde 3, in combination with chiral bis-phenols, has been used to measure the enantiopurity of a range of amines synthesised by various methods.5 In our hands it has proved suitable for the measurement of the enantiopurity of simple branched amines (e.g. α-alkylbenzylamines), but we were not able to apply it to determine the enantiopurity of a series of α-aminoamides prepared in our amidation chemistry.7
In this paper, we describe the use of chiral aldehyde 4 as a new method for the determination of the enantiopurity of chiral amines. We demonstrate its application to a diverse set of commercially available chiral amines including samples covering a wide range of enantiopurities as well as samples of α-aminoamides obtained directly from amidation reactions.
Scheme 1 Synthesis of both enantiomers of aldehyde 4. aThe enantiopurity was determined by imine formation with L-phenylalanine methyl ester.† |
To determine the suitability of 4 for the derivatisation of chiral amines we examined the imine formation reaction between racemic amino amide 7 and (S)-4 in a range of deuterated NMR solvents (Scheme 2 and Table 1). Pleasingly, the corresponding diastereomeric mixture of imines 8a/8b was formed rapidly at room temperature in a variety of different solvents.
Solvent | Peak separation (ppm) |
---|---|
CDCl3 | 0.11 |
DMSO-d6 | 0.09 |
MeOH-d4 | 0.21 |
C6D6 | 0.13 |
Unfortunately, the 1H NMR did not offer sufficient resolution between the two diastereoisomers to enable accurate quantification. However, the 13C NMR clearly showed two separate peaks for the imine carbons around 170 ppm which were readily resolved in all of the solvents used (Table 1). Fortunately, using modern NMR machines it is relatively straightforward to obtain accurate quantification from 13C NMR, provided the carbon signals under scrutiny have similar relaxation times. Typically, a 15-minute 13C NMR experiment was sufficient to obtain spectra with a strong enough signal to noise ratio, enabling efficient analysis of multiple samples in a short period of time.
With samples of both enantiomers of aldehyde 4 in hand, we explored the derivatisation of a selection of commercially available chiral amines 9–20 (Table 2). In the majority of cases, excellent resolution of the diastereomers was observed, enabling the enantiopurity9 of the chiral amines to be determined. There was good agreement between the measurements obtained with the two enantiomeric aldehydes, and the two values were averaged to give a more accurate measurement. The method could be used to derivatise a selection of benzylamines (10–12, 16, 19) as well as amines containing only aliphatic substituents at the adjacent carbon atom (9, 13–15). Bicyclic amines such as 17, 18 and 20 could also be analysed, although in the latter case the two imine carbon signals were very close together so the ratio of the two diastereoisomers could only be quantified by peak height and not by integration.
Amine | Δ(δC)/ppm | Enantiomeric ratio9 | ||
---|---|---|---|---|
(solvent) | (S)-4 | (R)-4 | Average | |
a The ratio of imines was determined using peak height rather than integration. | ||||
0.070 (d4-MeOH) | 98:2 | 99:1 | 98.5:1.5 | |
98:2 | 99:1 | 98.5:1.5 | ||
0.093 (d4-MeOH) | 98:2 | 99:1 | 98.5:1.5 | |
98:2 | 99:1 | 98.5:1.5 | ||
0.139 (d4-MeOH) | 97:3 | 98:2 | 97.5:2.5 | |
98:2 | 99:1 | 98.5:1.5 | ||
0.163 (d4-MeOH) | 99:1 | 98:2 | 98.5:1.5 | |
0.264 (CDCl3) | 99:1 | 98:2 | 98.5:1.5 | |
0.792 (CDCl3) | 99:1 | 99:1 | 99:1 | |
0.164 (d4-MeOH) | 99:1 | 99:1 | 99:1 | |
0.186 (d4-MeOH) | 98:2 | 98:2 | 98:2 | |
0.116 (d4-MeOH) | 97:3 | 96:4 | 96.5:3.5 | |
0.173 (CDCl3) | 95:5 | 96:4 | 95.5:4.5 | |
0.200 (d4-MeOH) | 51:49 | 50:50 | 50.5:49.5 | |
0.036 (C6D6) | 50:50 | 50:50 | 50:50a |
A key test of any enantiomeric determination method is the ability to accurately and reproducibly measure the enantiomeric purity of a set of unknown samples. To this end, we challenged the method with a set of ‘blind’ tests (Table 3). Three ‘unknown’ scalemic samples of chiral amine were prepared by each of three researchers by weighing out appropriate quantities of commercially available amines 9–11. The three samples were then analysed ‘blind’ by the other two researchers using both enantiomers of aldehyde 4. The measured er in each case was calculated by taking the average of the values obtained with (S)-4 and (R)-4. As can be seen from the results, extremely consistent measurements were obtained by the three researchers, and in all cases these were in very close agreement with the actual enantiopurity of the sample based on the original weighings. There is a small divergence in the measured enantiopurities at high er (entries 4 and 7), but this is reasonably expected as the errors in measuring the small integral for the minor isomer in these cases is likely to be relatively large.
Entry | Amine | Prepared by | Actual era (R:S) | Measured erb (R:S) |
---|---|---|---|---|
a Calculated using weights of amines used and the measured er values for amines 9–11 from Table 2. b Average of values obtained with (S)-4 and (R)-4, see supporting information for original data and the corresponding NMR spectra. | ||||
1 | 11 | RML | 69:31 | 70:30 (LB) |
69:31 (SMG) | ||||
2 | 10 | RML | 34:66 | 34:66 (LB) |
35:65 (SMG) | ||||
3 | 9 | RML | 77:23 | 77:23 (LB) |
78:22 (SMG) | ||||
4 | 11 | LB | 6:94 | 8:92 (RML) |
9:91 (SMG) | ||||
5 | 10 | LB | 86:14 | 87:13 (RML) |
85:15 (SMG) | ||||
6 | 9 | LB | 24:76 | 25:75 (RML) |
25:75 (SMG) | ||||
7 | 11 | SMG | 89:11 | 88:12 (RML) |
87:13 (LB) | ||||
8 | 10 | SMG | 53:47 | 53:47 (RML) |
53:47 (LB) | ||||
9 | 9 | SMG | 16:84 | 17:83 (RML) |
18:82 (LB) |
Pleasingly, an amine (±)-21 bearing an adjacent quaternary centre reacted with aldehyde 4 to give two easily resolved diastereomeric imines (Scheme 3). However, the formation of enamine 23 (as a single geometrical isomer) also took place, and significant amounts of 23 were observed before all of the starting amine 21 had been consumed. The formation of imine 22 was relatively slow between the enantioenriched amine (S)-21 and (S)-4. As can be seen in Fig. 2, the proportion of enamine present in the sample is already ∼35% by the time the starting amine is completely consumed (∼1.5 hours). Eventually, after a total reaction time of three hours the diastereomeric imine 22b is observed and its concentration gradually increases over time demonstrating that in this particular case racemisation of imine 22avia the enamine 23 takes place. In the corresponding reaction between (S)-21 and (R)-4, a similar reaction profile is observed, although the initial imine formation did not quite go to completion in this case. This nevertheless enables us to conclude that the original amine (S)-21 is of high enantiopurity (∼99:1 er) as no minor imine 22b is observed in the reaction of (S)-21 and (S)-4 until consumption of the amine 21 is complete. This demonstrates that either: (i) no detectable imine 22b is formed (there is no significant quantity of (R)-21 present); or (ii) imine 22b is formed but is rapidly converted into the enamine 23. The latter is unlikely to be the case as no significant quantity of the minor imine (in this case ent-22a) is observed when (S)-21 is reacted with (R)-4.
Fig. 2 Imine/enamine formation from amine 21 over time: (a) reaction of (S)-21 and (S)-4; (b) reaction of (S)-21 and (R)-4. |
In the case of all other amines analysed (Table 2), imine formation was observed to be much more rapid than enamine formation, enabling accurate determination of the enantiopurity in each case. However, enamine formation was seen to take place in some cases if samples of the imine were stored for prolonged periods of time. It is therefore recommended that NMR analysis should be carried out immediately after sample preparation in order to obtain accurate results.
We next went on to examine some compound samples obtained from our amidation chemistry. Amides 25a–25c were prepared via direct borate-mediated amidation of the corresponding Boc-amino acids 24a–24c (Scheme 4).7 Acidic deprotection of the Boc group gave the corresponding amines as their trifluoroacetate salts which were converted to the free amine by filtration through a basic ion exchange resin. The enantiopurity of the amines 26 was then measured using both enantiomers of aldehyde 4, and the enantiopurity of the Boc-protected compounds 25 was measured by chiral HPLC. In all three cases, good agreement was seen between the two methods, demonstrating that the amidation reactions and subsequent deprotection occur with very little loss in enantiopurity under these conditions.
Scheme 4 Determination of the enantiopurity of amidation products using aldehydes 4 and chiral HPLC. |
Footnote |
† Electronic supplementary information (ESI) available: Experimental procedures, spectroscopic data and 1H and 13C NMR spectra. See DOI: 10.1039/c5ob01398d |
This journal is © The Royal Society of Chemistry 2015 |