Special issue research article Received: 15 October 2014

Revised: 18 December 2014

Accepted: 19 January 2015

Published online in Wiley Online Library

(wileyonlinelibrary.com) DOI 10.1002/mrc.4222

Differentiation of enantiomers by 2D NMR spectroscopy at 1T using residual dipolar couplings Martin R. M. Koos,a Ernesto Danieli,b Federico Casanova,c Bernhard Blümichb and Burkhard Luya* Keywords: low-field NMR; RDC; differentiation of enantiomers; enantiomeric excess; chiral alignment media

The ongoing miniaturization of NMR spectrometers has opened new venues for applications in many different fields of chemistry. The availability of so-called medium-field benchtop NMR spectrometers[1] operating at proton Larmor frequencies of 20–60 MHz utilizing permanent magnets without the need of cryogenic cooling brings NMR spectroscopy directly into chemical laboratories. Immediate on the spot measurements of routine 1H spectra for monitoring syntheses become possible in research facilities as well as at production sites where desktop NMR spectroscopy will be a cost-effective, powerful tool for quality control and yield optimization. In this context, it is our aim to explore in how far challenging applications are accessible by medium-field, portable NMR spectrometers. A challenging but common task in analytical chemistry is the distinction of enantiomers and the corresponding estimation of enantiomeric excess.[2] It is most often addressed by chiral chromatography and optical measurements like optical rotatory dispersion or circular dichroism spectroscopy, but also, different approaches based on NMR spectroscopy are being used. In the latter case, it is necessary to form diastereomers from the enantiomeric compounds, which is either achieved by chemical modification with the Mosher ester[3] and related substances,[4] by the formation of weak diastereomeric complexes with lanthanide shift reagents,[5] and chiral alignment media.[6] The diastereomeric entities can then in principle be distinguished either by differences in chemical shifts or by different residual dipolar (RDC) and quadrupolar couplings of partially aligned samples.[6a,c] The differences, however, are usually quite small – even on modern high-field NMR spectrometers – and it is not clear a priori if limitation in terms of resolution, sensitivity and technical issues like field stability allows such measurements at all. In order to provide a proof of principle for enantiomeric distinction using medium-field NMR spectroscopy, we focused here on a general approach independent of the magnetic field strength and adapted it to the corresponding experimental setup on a medium-field 43 MHz Magritek/ACT prototype spectrometer:[7] Because the chemical shift resolution of the spectrometer is roughly one tenth of a ppm, differentiation based on chemical shift differences would only work in favorable cases with sufficient enantiomeric differentiation power so that we concentrated on differences in coupling constants, which do not depend on the strength of the magnetic field.[8] Furthermore, as sensitivity and

Magn. Reson. Chem. (2015)

available probe heads on medium-field NMR spectrometers do not yet allow the measurement of either deuterium quadrupolar[9] or heteronuclear dipolar couplings,[6a] we focus on the distinction via homonuclear 1H-1H RDC.[10] We have chosen the amino acid alanine as test molecule for which enantiomeric distinction has been extensively studied using high-field high-resolution NMR spectroscopy.[10a,11] The simplest experiment, a 1D 1H spectrum, is shown in Fig. 1A in black for L-alanine hosted in a 25% gelatin gel at 43 MHz. The spectrum of the unstretched state with a gelatin gel of length L0 clearly shows the splittings as a result of 3JHH couplings with the underlying coupling constant being measureable on both the Hα quartet as well as the CH3 doublet to be roughly 7 Hz. If the same sample is stretched to length L using a rubber-based stretching apparatus,[11b,12] the resulting total splitting changes because of the dipolar contribution to 3THH = 3JHH + 3DHH (Fig. 1A, blue lines). Also, the isochronous methyl protons split into a triplet of doublets because of the additional 2DHH dipolar couplings.[13] The stronger the alignment as indicated by the strain or extension factor Ξ = (L L0) / L0,[10a] the larger the dipolar contribution to the multiplets. Corresponding spectra for a 2 : 1 mixture of l-alanine and d-alanine in 25% gelatin/D2O are shown in Figure 1(B). Again, the effect of RDCs is easily visible in the 1D spectra, and a closer inspection reveals that the mixture results in multiplets that are different from the l-alanine sample. A clear distinction of enantiomers as previously reported for similar samples,[10a] however, is not possible because of the limited resolution due to relatively broad lines. We therefore took the next step by implementing 2D J-resolved

* Correspondence to: Burkhard Luy, Institut für Organische Chemie (IOC) and Institut für Biologische Grenzflächen (IBG), Karlsruher Institut für Technologie (KIT), Fritz-Haber-Weg 6, 76131 Karlsruhe, Germany. E-mail: [email protected] a Institut für Organische Chemie (IOC) and Institut für Biologische Grenzflächen (IBG), Karlsruher Institut für Technologie (KIT), Fritz-Haber-Weg 6, 76131 Karlsruhe, Germany b Institut für Technische und Makromolekulare Chemie (ITMC), Lehrstuhl für Makromolekulare Chemie, RWTH Aachen University, Worringerweg 1, 52074 Aachen, Germany c Magritek GmbH, Philipsstrasse 8, 52068 Aachen, Germany

Copyright © 2015 John Wiley & Sons, Ltd.

M. R. M. Koos et al. experiments,[14] which are well known to inherently compensate static magnetic field inhomogeneities in the indirectly acquired dimension by formation of Hahn echoes. Indeed, corresponding spectra show significantly improved line widths in the J-dimension of 1.8 Hz compared to approximately 4.5 Hz in the conventional 1D. Figure 2 shows the directly processed 2D spectra with the typical tilts of the multiplets together with the coupling contributions from the isotropic and partially aligned

1

Figure 1. 1D H spectra of (A) L-alanine and (B) a mixture of D-alanine and L-alanine each dispersed in 25% gelatin/D2O. Several spectra for the two samples with different degrees of stretching corresponding to extension factors Ξ are shown. The line width of the Hmethyl peaks is approximately 4.5 Hz in the unstretched sample. With increasing alignment, multiplets 2 3 change because of DHH and DHH residual dipolar couplings. Thirty-two scans were accumulated every 3 s, acquiring 8192 points in 1.64 s each.

Figure 2. J-resolved spectra of the methyl groups of (A, C) L-alanine and (B, D) D + L-alanine (2 : 1) in gelatin/D2O measured using the unstretched, isotropic samples (Ξ = 0) (A, B) and the stretched samples (Ξ ≈ 1) (C, D). The splittings caused by the different couplings are identified. This figure is available in colour online at wileyonlinelibrary.com/journal/mrc

wileyonlinelibrary.com/journal/mrc

Figure 3. Sum projection of (A) Hα and (B) Hmethyl multiplets of stretched L-alanine (Ξ ≈ 1) from 2D J-resolved spectra, overlaid with the simulated multiplet of L-alanine (blue). The spectrum was integrated over a range of 5.5 Hz in the directly detected direction symmetrically around the center of the peak. The line intensities of the Hmethyl multiplet roughly follow the expected 1 : 1 : 2 : 2 : 1 : 1 multiplet of the weak coupling limit, slightly distorted by second-order artifacts.

Figure 4. Sum projection of the (A) Hα and (B) Hmethyl multiplets of the stretched mixture of D + L-alanine (2 : 1) from 2D J-resolved spectra overlaid with the simulated peaks (blue). The projection was integrated over a range of 5.5 Hz in the directly detected direction symmetrically around the top of the peak. The individual multiplets of D-alanine (red) and L-alanine (green) that constitute the spectrum are displayed below. The D-alanine Hmethyl multiplet intensities (B, red) approach the ratio 1 : 2 : (1 : 1) : 2 : 1. From the intensities of the outer and central peaks, the fraction of D-alanine can be estimated to be 1.9 while comparison of all peaks to a simplified 1 : 3 : 6 : 4 : 6 : 3 : 1 multiplet model indicates an error of approximately 15%. This figure is available in colour online at wileyonlinelibrary.com/journal/mrc

Copyright © 2015 John Wiley & Sons, Ltd.

Magn. Reson. Chem. (2015)

Differentiation of enantiomers by 1T bench-top 2D NMR spectroscopy L-alanine and D + L-alanine samples. While especially the methyl group shows a well-recognizable multiplet structure for pure L-alanine, the overlap of signals in the mixture leads to a complex pattern that is best addressed by modeling the spectra as shown in Figs. 3 and 4: After tilting the multiplets, 1D slices along the J-dimension were extracted and compared with multiplets simulated using self-written Matlab code until satisfying agreement was achieved. For the L-alanine sample, dipolar contributions of 2 DHH = 5.65 Hz and 3DHH = 3.05 Hz could be identified, and for the enantiomeric mixture, corresponding couplings are 2 DHH = 2.55 Hz and 3DHH = 2.00 Hz for D-alanine and 2 DHH = 7.35 Hz and 3DHH = 3.97 Hz for L-alanine. All peaks but the outermost peaks of L-alanine and the central peak of the D-alanine multiplet consist of contributions from both molecules at the same ratio. This unfortunate combination of dipolar couplings leaves only three peaks for measuring the relative ratio D-alanine versus L-alanine. Using the intensities of a sum projection (Fig. 4), the ratio can be estimated to 1.9, corresponding to an enantiomeric excess of 0.31, which is very close to the weighed-in ratio of 2.0 (ee = 0.33). Comparison of the multiplet intensity to a simple 1 : 3 : 6 : 4 : 6 : 3 : 1 multiplet indicates an estimated error of up to 15% in the measurement of the enantiomeric ratio. We conclude from the experimental results that the distinction of enantiomers is readily possible using low-field 43 MHz 1H NMR spectroscopy and partial alignment in a chiral alignment medium. In contrast to chemical shift differences, for which resolution is limited by the homogeneity of the static magnetic field, differences in coupling constants are directly measured using J-resolved spectroscopy with a resolution that only depends on natural line width and overall spectrometer stability. Especially, the latter aspect has been a serious issue with benchtop medium-field spectrometers so far, but spectrometer design during the last years improved considerably, and state-of-the-art spectrometers available today appear to fulfill the requirements for the acquisition of well-resolved 2D J-spectra. The other issue of low-field NMR spectroscopy next to resolution is sensitivity. The setup used for the presented experiments allows a signal-to-noise ratio of 30.000 : 1 for H2O in a single scan. This easily allows the acquisition of spectra from samples with concentrations in the 100 mM range that are readily available for most small molecules. As long as spin systems are fairly small, as in the case of alanine, the reduction of peak intensity as a result of the multiplet splitting can be tolerated. For larger spin systems, this decrease in intensity might become an issue using a simple, non-selective J-resolved pulse sequence, but a number of more sophisticated, selective experiments existn[10b,15] that have been designed to eliminate most homonuclear couplings. In the near future, medium-field NMR spectroscopy is expected to benefit from the use of advanced techniques for sensitivity enhancement such as the combination with hyperpolarization techniques.[16] Schemes like para-hydrogen NMR[17] that transfer spin order with chemical selectivity are of particular interest because they boost sensitivity and simplify the spectra at the same time. The distinction of enantiomers could be further enhanced by the applicability of heteronuclear experiments like BIRD-decoupled variants of an HSQC[11c] or the direct acquisition of deuterium spectra.[18] First, benchtop spectrometers are now equipped with 13 19 C, F or 31P coils in addition to 1H acquisition, and the foreseeable applications are manifold. Considering everyday applications, the presented approach could be used, for example, in quality control applications involving enantiomeric compounds. Corresponding NMR spectra are able to

Magn. Reson. Chem. (2015)

address both chemical and enantiomeric purity of small molecules at the same time. The non-routine sample preparation could be simplified by directed chromatography or electrophoresis into the gel-based alignment medium, and data analysis would be straight forward as soon as an initial protocol is established for the molecule of interest. In contrast to conventional high-field NMR, low-field benchtop NMR spectrometers can be placed directly at the production site ensuring an effective work flow. Finally, medium-field NMR spectroscopy provides a general alternative to costly high-resolution NMR spectroscopy in chemical research, which will lead to a change of how NMR is applied in laboratories. The possibility to differentiate small molecule enantiomers certainly adds to the attractiveness of benchtop NMR spectrometers.

Experimental section All experiments were performed on an early prototype magnet provided by ACT/Magritek working at 42.8 MHz. The desktop NMR magnet was thermally stabilized, and its homogeneity improved by electrical shim coils providing up to second-order corrections. Two 25% gelatin/D2O gel samples were prepared containing either 167 mmol/l L-alanine or a mixture of 111 mmol/l D-alanine and 56 mmol/l L-alanine with a trace amount of NaN3 added for conservation. Both samples were inserted into a stretching apparatus as described in ref. [12b]. 1D pulse-acquire spectra utilizing a 5.5 μs excitation pulse, recording 8192 data points in 1638.4 ms acquisition time with a repetition time of 3 s, and 32 scans were acquired of both samples with varying alignment strengths. Acquired FIDs were zero filled to 32 768 points without use of a window function. 2D J-resolved spectra were acquired of both samples with alignment strength corresponding to extension factors of 0 (no stretching) and approx. 1 (corresponding to a gel of twice the unstretched length). 4096 × 48 data points were recorded with 819.2 ms acquisition time, with a repetition time of 1.5 s and eight scans per increment. The spectral width in the indirect dimension was 50 Hz. Individual scans were aligned, using the highest peak of the spectrum as a reference, to compensate for remaining B0 instability. Spectra were zero filled to 8192 × 384 points without use of a window function and tilted by 45° using standard procedures for J-resolved spectroscopy. To preserve information, no symmetrization was performed. Simulations were performed using MATLAB R2012b and selfwritten code. The full homonuclear secular Hamiltonians for chemical shift, J-coupling and D-coupling and analytical pulses were included. Acknowledgements We greatly acknowledge funding by the Deutsche Forschungsgemeinschaft (DFG Gerätezentrum Pro2NMR, DFG LU 835/4-2 and DFG FOR 934) and the HGF (programme BioInterfaces). We thank Dr. Juan Perlo for helpful discussions regarding technical issues with the experimental setup. M.R.M.K. thanks the Fonds der Chemischen Industrie e.V. for his PhD fellowship.

References [1] a) B. Blümich, F. Casanova, S. Appelt. Chem. Phys. Lett. 2009, 477, 231–240; b) M. Cudaj, G. Guthausen, T. Hofe, M. Wilhelm. Macromol. Rapid Commun. 2011, 32, 665–670; c) B. Luy. Angew. Chem. Int. Ed. 2011, 50, 354–356.

Copyright © 2015 John Wiley & Sons, Ltd.

wileyonlinelibrary.com/journal/mrc

M. R. M. Koos et al. [2] [3] [4] [5] [6]

[7] [8] [9] [10] [11]

K. W. Busch, M. A. Busch, Chiral Analysis, Elsevier, Amsterdam, 2006. J. A. Dale, H. S. Mosher. J. Am. Chem. Soc. 1973, 95, 512–519. J. M. Seco, E. Quinoa, R. Riguera. Chem. Rev. 2004, 104, 17–117. T. C. Morrill, Lanthanide Shift Reagents for Stereochemical Analysis, John Wiley & Sons, 1987. a) M. Sarfati, P. Lesot, D. Merlet, J. Courtieu. Chem. Commun. 2000, 2069–2081; b) I. Canet, J. Courtieu, A. Loewenstein, A. Meddour, J. M. Pechine. J. Am. Chem. Soc. 1995, 117, 6520–6526; c) B. Luy. J. Indian Inst. Sci. 2010, 90, 119–132. E. Danieli, J. Perlo, B. Blümich, F. Casanova. Angew. Chem. Int. Ed. 2010, 49, 4133–4135. G. Kummerlöwe, B. Luy, in Annu. Rep. NMR Spectrosc, Vol. 68 (Ed: G. A. Webb), 2009, pp. 193–230. P. Lesot, D. Merlet, A. Loewenstein, J. Courtieu. Tetrahedron: Asymmetry 1998, 9, 1871–1881. a) C. Naumann, W. A. Bubb, B. E. Chapman, P. W. Kuchel. J. Am. Chem. Soc. 2007, 129, 5340–5341; b) J. Farjon, D. Merlet, P. Lesot, J. Courtieu. J. Magn. Reson. 2002, 158, 169–172. a) G. Kummerlöwe, M. U. Kiran, B. Luy. Chem. Eur. J. 2009, 15, 12192–12195; b) P. W. Kuchel, B. E. Chapman, N. Müller, W. A. Bubb, D. J. Philp, A. M. Torres. J. Magn. Reson. 2006, 180, 256–265; c) K. Kobzar, H. Kessler. Angew. Chem. Int. Ed. 2005, 44, 3145–3147; d) U. Eliav, G. Navon, B. Luy. J. Am. Chem. Soc. 2006, 128, 15956–15957.

wileyonlinelibrary.com/journal/mrc

[12] a) G. Kummerlöwe, E. F. McCord, S. F. Cheatham, S. Niss, R. W. Schnell, B. Luy. Chem. Eur. J. 2010, 16, 7087–7089; b) G. Kummerlöwe, F. Halbach, B. Laufer, B. Luy. The Open Spectrosc. J. 2008, 2, 29–33. [13] P. Tzvetkova, S. Simova, B. Luy. J. Magn. Reson. 2007, 186, 193–200. [14] a) W. P. Aue, J. Karhan, R. R. Ernst. J. Chem. Phys. 1976, 64, 4226–4227; b) G. Bodenhausen, R. Freeman, D. L. Turner. J. Chem. Phys. 1976, 65, 839–840; c) R. Freeman, H. D. W. Hill. J. Chem. Phys. 1971, 54, 301–313; d) R. L. Vold, S. O. Chan. J. Chem. Phys. 1970, 53, 449–451. [15] N. Giraud, L. Beguin, J. Courtieu, D. Merlet. Angew. Chem. Int. Ed. 2010, 49, 3481–3484. [16] H. W. Spiess. Angew. Chem. Int. Ed. 2008, 47, 639–642. [17] R. A. Green, R. W. Adams, S. B. Duckett, R. E. Mewis, D. C. Williamson, G. G. R. Green. Prog. Nucl. Magn. Reson. Spectrosc. 2012, 67, 1–48. [18] J. P. Bayle, J. Courtieu, E. Gabetty, A. Loewenstein, J. M. Pechine. New J. Chem. 1992, 16, 837–838.

Supporting information Additional supporting information may be found in the online version of this article at the publisher’s web site.

Copyright © 2015 John Wiley & Sons, Ltd.

Magn. Reson. Chem. (2015)

Differentiation of enantiomers by 2D NMR spectroscopy at 1 T using residual dipolar couplings.

Differentiating enantiomers using 2D bench-top NMR spectroscopy. Spectrometers working with permanent magnets at 1 T field strength allow the acquisit...
987KB Sizes 0 Downloads 2 Views