13C Satellite-Free 1H NMR Spectra - ACS Publications - American


Oct 30, 2017 - Figure 1a shows the 1H spectrum of a sample of the proton pump inhibitor ... The new method, shown in Figure 2, is an extension of the...
3 downloads 0 Views 466KB Size


Subscriber access provided by READING UNIV

Letter

13C satellite-free 1H NMR spectra Pinelopi Moutzouri, Peter Kiraly, Andrew Richard Phillips, Steven R. Coombes, Mathias Nilsson, and Gareth A. Morris Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b03787 • Publication Date (Web): 30 Oct 2017 Downloaded from http://pubs.acs.org on October 31, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

13

C satellite-free 1H NMR spectra

Pinelopi Moutzouri,† Peter Kiraly, † Andrew R. Phillips,‡ Steven R. Coombes,§ Mathias Nilsson† and Gareth A. Morris*† †

School of Chemistry, University of Manchester, Oxford Road, Manchester M13 9PL, UK.



Pharmaceutical Sciences, AstraZeneca, Silk Road Business Park, Macclesfield, SK10 2NA, UK.

§

Pharmaceutical Technology and Development, AstraZeneca, Silk Road Business Park, Macclesfield, SK10 2NA, UK.

ABSTRACT: A new NMR experiment (Destruction of Interfering Satellites by Perfect Echo Low-pass filtration, DISPEL) is introduced that facilitates the analysis of low-level components in high dynamic range mixtures by suppressing one-bond 13 C satellite signals in 1H spectra. Since the natural abundance of 13C is around 1.1%, these satellites appear at 0.55% of the intensity of a parent peak, mimicking and often masking impurity signals. The new experiment suppresses one-bond 13C satellite signals, with high efficiency, at negligible cost in signal-to-noise ratio, and over a wide range of one-bond coupling constants, without the need for broadband 13C decoupling.

The identification and quantification of minor components in mixtures pose challenges in many areas of chemistry. In pharmaceutical manufacturing, for example, it is a requirement that all impurities above 0.1 % of a main active pharmaceutical ingredient should be identified and quantified.1 The presence of isotopomers containing 13C (natural abundance 1.1 %) is a major complication, as it gives rise to multiple 13C satellite signals at around 0.54 % of the parent peak intensity. 19F NMR is a powerful emerging tool in the analysis of low level impurities, giving highly sensitive and well-resolved spectra2-6. Recent experiments have made the use of 19F NMR even more attractive by allowing uniform quantitative excitation of its wide chemical shift range7,8, and suppressing 13C satellite signals9. However, the great majority of experiments use 1 H NMR, where the approach used in the ODYSSEUS experiment9 is not directly applicable. The much narrower chemical shift range and extensive multiplet structure greatly increase the potential for confusion and/or overlap between signals of low-level species and 13C satellites. Moreover the extensive coupling complicates the design of spectral editing pulse sequences. However the much smaller secondary isotope effects on the chemical shift mean that long-range 1H-13C couplings cause few if any problems. A new technique is proposed that gives excellent suppression of one-bond 13C satellite signals, over a wide range of one-bond coupling constants, at a very small cost in sensitivity. A variety of different methods have been used to circumvent problems of overlap between 13C satellites and signals of interest, including changing the solvent, concentration, temperature or pH of the sample to move the signals relative to one another,10,11 and using broadband 13 C decoupling to collapse the heteronuclear couplings12-17. The first class of methods is tedious and time-consuming,

relying on trial and error. Broadband 13C decoupling is more attractive, and in some cases works well, but its performance is very dependent on the sample, solvent and instrumentation used. As shown in Figure S10 of the Supplementary Information, even with the most recent instrumentation and a sample that is only moderately lossy, the high radiofrequency power required causes sample heating which broadens, distorts and shifts signals. The sidebands introduced by in the spectrum can also complicate spectral analysis. As shown in Figure S10 and S11, even with bilevel adiabatic decoupling18 careful optimisation is required to balance the conflicting demands of minimizing heating and avoiding decoupling sidebands. A subtler problem is that the secondary isotope effect on the proton chemical shift means that the decoupled signals from 13C isotopomers have slightly different (typically 1 – 2 ppb) chemical shifts from those of 12C isotopomers which slightly broadens the bases of the decoupled resonances. The new method described here avoids all these problems by editing the 1H spectrum to remove 13C satellite signals. Figure 1a shows the 1H spectrum of a sample of the proton pump inhibitor omeprazole (1, Scheme 1), used to treat excess stomach acid, spiked with a small amount (0.15 %) of its precursor omeprazole sulfide (2). The signals of 2, of which the 1H spectrum can be seen in Figure 1c, are approximately 4 times less intense compared to the 13 C satellites of omeprazole. The 13C satellites of 1, being dominant over the weaker impurity signals, complicate analysis; for example as seen in Figure 1f a signal of 2 close to 2.3 ppm is almost degenerate with a satellite signal of 1. However in the 1H spectrum of Figure 1b, obtained with the new one-bond 13C satellite suppression method, the absence of the 13C satellites makes it straightforward to

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 5

identify those signals that do not originate from omeprazole. Here in addition to the signals of 2, identifiable by comparison with the omeprazole sulfide spectrum of Figure 1c, a small number of very weak signals, marked with asterisks, of an unknown impurity can be seen. Less obvious but evident in Figure 1d and 1g, the spectral region around 6.8 ppm shows that the 13C satellite signal of omeprazole at 6.76 ppm completely masks one of the omeprazole sulfide signals. The new method, shown in Figure 2, is an extension of the 1JCH low-pass filter,19-23 used for example in the HMBC20 and ODYSSEUS9 experiments, in which a 90° 13C pulse is used to

Scheme 1. Omeprazole (1), its precursor omeprazole sulfide (2), quinine (3), and cinchonidine (4). convert 1H coherence that is antiphase with respect to 13C into unobservable multiplet quantum coherence. The presence of 1H-1H scalar coupling means that this basic low-pass filter causes the phases of multiplet components to diverge, distorting and complicating phase-sensitive spectra. This J modulation can be refocused (for short times) by using a perfect echo24-26 instead of a simple spin echo.27 The perfect echo uses an orthogonal 90° pulse between two spin echoes to reverse the sense of J modulation. Provided the duration 2 τ of each echo is short compared to the inverse of the couplings JHH (2 τ JHH << 1), both the chemical shift and the homonuclear scalar coupling are refocused. The need to use a perfect echo to refocus the effects of JHH makes it possible to use a four-stage low-pass J filter, with four different delays and four 90° 13C pulses, to cover a wide range in 1JCH at no extra cost in time.19-23 At each stage a 90° 13C pulse converts all 1H coherences that are antiphase with respect to 13C into multiple quantum coherence that is subsequently dephased. The result is that the 1H coherence of 13C isotopomers is repeatedly depleted; if the delays are appropriately optimised the result is efficient suppression of ipso-13C isotopomer signals over a very wide range of one-bond coupling constants 1JCH. As shown in Figure S3, when the τ1, τ2, τ3, and τ4 delays are set to the optimized values 3.2, 1.1, 3.95 and 1.56 ms respectively, satellite signals are suppressed, typically by two orders of magnitude, for 1JCH from 120 Hz to 360 Hz, covering the full range of couplings likely to be encountered in practice.

1

13

Figure 1. 500 MHz (a) H spectrum, and (b) one-bond C 1 satellite-suppressed H spectrum acquired with the pulse sequence of Figure 2 for a mixture containing omeprazole (1) spiked with 0.15 % of its precursor omeprazole sulfide (2). (c) 1 H spectrum of omeprazole sulfide. Spectra (d) - (f), (g) - (i) and (j) – (l) are vertical expansions of spectra (a), (b) and (c) 13 respectively. In spectrum (i), the residual C satellites are at approximately 0.01 % of the parent omeprazole signal. The 13 solid and dashed arrows indicate the one-bond C satellites of omeprazole, and the signals of omeprazole sulfide, respectively, while the asterisks indicate very weak unidentified signals.

The z-filter28 at the end of the pulse sequence rejects out of phase coherences, suppressing any zero quantum term which would introduce unwanted anti-phase dispersive components in the spectrum. As shown in Figure S6, it also helps with the elimination of any homonuclear JHH modulation that was not perfectly refocused by the quadrature 90° pulse of the perfect echo. Additionally, it dephases any residual anti-phase heteronuclear term. Compared with the conventional pulse-acquire experiment, the 1H spectrum acquired with the pulse sequence of Figure 2 causes negligible sample heating and has the same resolution. There is a very small sensitivity penalty

ACS Paragon Plus Environment

Page 3 of 5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

because of the effects of pulse imperfections and of T2 relaxation during the 4τ duration of the perfect echo.

Figure 2. DISPEL (Destruction of Interfering Satellites by Perfect Echo Low-pass filtration) pulse sequence for the ac13 1 quisition of one-bond C satellite-suppressed H spectra. Narrow and wide rectangles represent 90° and 180° hard RF pulses respectively. The trapezoid represents a 180° smoothed chirp pulse applied simultaneously with a matched field gradient pulse (G3). The grey rectangles G1 and G2 indicate field gradient pulses for coherence transfer pathway selection, while G4 represents a homospoil pulsed field gradient. The delay τ is set to 4.05 ms, to accommodate the numerically optimized delays τ1, τ2, τ3, and τ4 (3.2, 1.1, 3.95 and 1.56 ms respectively) that allow suppression of couplings 1 from 120 Hz to 360 Hz. The phase of the second 90° H pulse is in quadrature to that of the first. A minimum of 2 scans is needed; the phase cycle is detailed in the Supporting Information.

With the delays optimised as above, the overall time over which T2 causes signal attenuation is about 16 ms ; the z-filter adds a small additional T1 weighting. Where necessary, a shorter version of the pulse sequence can be used (Figure S4), with a total T2 weighting of 8.4 ms. As shown in Figure S5, it offers at least 50-fold suppression for a range of 1JCH couplings between 120 Hz and 157 Hz, which is adequate for many purposes. The quantification by NMR of mixture components below 1% is challenging, not least because of the presence of 13 C satellites. The DISPEL experiment facilitates quantification of dilute components by suppressing the interfering satellite signals, at the price of a small additional uncertainty, typically of the same order as the contribution of the spectral noise, due to differences in relaxation times. As is usual in quantitation by NMR, the lost accuracy can if necessary be retrieved either by correction for relaxation, or by calibration by standard addition. 1

In H spectra the proposed approach generally does away with any need for broadband 13C decoupling. The one-bond 13C satellites are suppressed, while any longrange 13C satellites are typically hidden under the multiplet structure of the parent 12C signals and do not interfere with the detection and quantification of low level signals. However in cases where signals in the 1H spectrum are very narrow, whether because of the absence of multiplet structure or because the spectrum is measured in pure shift mode29,30, the satellite filtration may be combined with low power adiabatic 13C decoupling to collapse

the remaining, long range, 1H-13C couplings. Only low power is needed, minimising sample heating, because nJCH << 1JCH.

1

1

Figure 3. 500 MHz (a) H spectrum, and (b) DISPEL H spec1 trum of a 30 mM sample of quinine in DMSO-d6 (3). (c) H spectrum of cinchonidine for comparison. The asterisks indicate signals from other impurities, presumably the dihydro analogues.

Figure 3 shows a second illustration of the DISPEL experiment. This time a sample of the natural product quinine (3), which is extracted from the bark of cinchona tree was used, as purchased from Acros Organics (99%) without further purification. An area of the 1H spectrum of quinine can be seen in Figure 3a. After the filtration of the 1 JCΗ couplings it was very straightforward to spot the presence of at least an impurity, which later was identified as cinchonidine (4), another alkaloid extracted from the same tree. The 13C satellites of quinine, are of comparable intensity to the impurity signals, while in more than one case they overlap with the signals of 4. The absence of 13C satellites in the spectrum of Figure 3b greatly assists identification of the cinchonidine signals. The new method introduced here is straightforward to implement and offers the possibility of acquiring clean 1H spectra, free from 13C satellites. This will greatly facilitate both the analysis of low level impurities in the presence of strong signals and the measurement of 1H spectra in protiated, as opposed to deuteriated, solvents, while retaining the high resolution and sensitivity of the conventional 1D 1H experiment.

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website. Additional figures as noted in text, pulse sequences and phase cycling, sample details, experiment details, four-stage DISPEL experiment theoretical analysis, two-stage DISPEL experiment theoretical analysis, comparison of filter perfor13 mance. and comparison with broadband C decoupling (PDF).

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

All raw experimental data and pulse sequence codes can be downloaded from DOI: http://dx.doi.org/10.17632/29y4xynz34.2.

AUTHOR INFORMATION Corresponding Author * E-mail: [email protected]

Author Contributions Ms Moutzouri performed the experiments; the pulse sequences were designed and analysed by Ms Moutzouri, Dr Kiraly, Dr Nilsson and Prof Morris; the problem of omeprazole was suggested by Dr Coombes and Dr Phillips; the work was supervised by Dr Coombes, Dr Phillips, Dr Nilsson and Prof Morris. All authors contributed to the writing of the manuscript and supplementary information.

Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT This work was supported by AstraZeneca and by the Engineering and Physical Sciences Research Council (EP/N033949/1). The application to protiated solvents was kindly brought to our attention by Dr PWA Howe, who has independently developed a closely-related experiment.

Page 4 of 5

(13) Fu, R.; Bodenhausen, G., Chem. Phys. Lett., 1995, 245, 415420. (14) Fujiwara, T.; Anai, T.; Kurihara, N.; Nagayama, K., J. Magn. Reson., 1993, 104, 103-105. (15) Kupče, E.; Freeman, R., Chem. Phys. Lett., 1996, 250, 523527. (16) Shaka, A.J.; Barker, P.B.; Freeman, R., J. Magn. Reson., 1985, 64, 547-552. (17) Simeral, L.S., Appl. Spectrosc., 1995, 49, 400-402. (18) Kupče, E.; Freeman, R.; Wider, G.; Wüthrich, K., J. Magn. Reson., 1996, 122, 81-84. (19) Bax, A.; Griffey, R.H.; Hawkins, B.L., J. Magn. Reson., 1983, 55, 301-315. (20) Bax, A.; Summers, M.F., J. Am. Chem. Soc., 1986, 108, 2093-2094. (21) Kogler, H.; Sørensen, O.W.; Bodenhausen, G.; Ernst, R.R., J. Magn. Reson., 1983, 55, 157-163. (22) Müller, L., J. Am. Chem. Soc., 1979, 101, 4481-4484. (23) Nyberg, N.T.; Sørensen, O.W., Magn. Reson. Chem., 2006, 44, 451-4. (24) Aguilar, J.A.; Nilsson, M.; Bodenhausen, G.; Morris, G.A., Chem. Commun., 2012, 48, 811-813. (25) Takegoshi, K.; Ogura, K.; Hikichi, K., J. Magn. Reson., 1989, 84, 611-615. (26) van Zijl, P.C.M.; Moonen, C.T.W.; von Kienlin, M., J. Magn. Reson., 1990, 89, 28-40. (27) Hahn, E.L., Phys. Rev., 1950, 80, 580-594. (28) Thrippleton, M.J.; Keeler, J., Angew. Chem. Int. Ed., 2003, 42, 3938-3941. (29) Adams, R.W., Pure Shift NMR Spectroscopy. In eMagRes, John Wiley & Sons, Ltd: 2007. (30) Castañar, L.; Parella, T., Magn. Reson. Chem., 2015, 53, 399-426.

REFERENCES (1) I.C.H., Impurities in new drug substances Q3A (R2) International Conference on Harmonisation, IFPMA, Geneva, Switzerland, 2006. (2) Mistry, N.; Ismail, I.M.; Duncan Farrant, R.; Liu, M.; Nicholson, J.K.; Lindon, J.C., J. Pharm. Biomed. Anal., 1999, 19, 511-517. (3) He, W.; Du, F.; Wu, Y.; Wang, Y.; Liu, X.; Liu, H.; Zhao, X., J. Fluorine Chem., 2006, 127, 809-815. (4) Lindon, J.C.; Wilson, I.D., 19F NMR Spectroscopy: Applications in Pharmaceutical Studies. In eMagRes, John Wiley & Sons, Ltd: 2007. (5) Trefi, S.; Gilard, V.; Malet-Martino, M.; Martino, R., J. Pharm. Biomed. Anal., 2007, 44, 743-754. (6) Trefi, S.; Gilard, V.; Balayssac, S.; Malet-Martino, M.; Martino, R., J. Pharm. Biomed. Anal., 2008, 46, 707-722. (7) Power, J.E.; Foroozandeh, M.; Adams, R.W.; Nilsson, M.; Coombes, S.R.; Phillips, A.R.; Morris, G.A., Chem. Commun., 2016, 52, 2916-2919. (8) Power, J.E.; Foroozandeh, M.; Moutzouri, P.; Adams, R.W.; Nilsson, M.; Coombes, S.R.; Phillips, A.R.; Morris, G.A., Chem. Commun., 2016, 52, 6892-6894. (9) Moutzouri, P.; Kiraly, P.; Phillips, A.R.; Coombes, S.R.; Nilsson, M.; Morris, G.A., Chem. Commun., 2017, 53, 123-125. (10) Holzgrabe, U.; Branch, S.K., Magn. Reson. Chem., 1994, 32, 192-196. (11) Michaleas, S.; Antoniadou-Vyza, E., J. Pharm. Biomed. Anal., 2006, 42, 405-410. (12) Freeman, R.; Kupče, E., NMR Biomed., 1997, 10, 372-380.

ACS Paragon Plus Environment

Page 5 of 5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Authors are required to submit a graphic entry for the Table of Contents (TOC) that, in conjunction with the manuscript title, should give the reader a representative idea of one of the following: A key structure, reaction, equation, concept, or theorem, etc., that is discussed in the manuscript. Consult the journal’s Instructions for Authors for TOC graphic specifications.

Insert Table of Contents artwork here

ACS Paragon Plus Environment