Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2022 Nov 23;94(48):16667–16674. doi: 10.1021/acs.analchem.2c02905

Rapid Quantification of Pharmaceuticals via 1H Solid-State NMR Spectroscopy

Y T Angel Wong †, Ruud L E G Aspers †, Marketta Uusi-Penttilä ‡, Arno P M Kentgens †,*
PMCID: PMC9730298  PMID: 36417314

Abstract

graphic file with name ac2c02905_0008.jpg

The physicochemical properties of active pharmaceutical ingredients (APIs) can depend on their solid-state forms. Therefore, characterization of API forms is crucial for upholding the performance of pharmaceutical products. Solid-state nuclear magnetic resonance (SSNMR) spectroscopy is a powerful technique for API quantification due to its selectivity. However, quantitative SSNMR experiments can be time consuming, sometimes requiring days to perform. Sensitivity can be considerably improved using 1H SSNMR spectroscopy. Nonetheless, quantification via 1H can be a challenging task due to low spectral resolution. Here, we offer a novel 1H SSNMR method for rapid API quantification, termed CRAMPS–MAR. The technique is based on combined rotation and multiple-pulse spectroscopy (CRAMPS) and mixture analysis using references (MAR). CRAMPS–MAR can provide high 1H spectral resolution with standard equipment, and data analysis can be accomplished with ease, even for structurally complex APIs. Using several API species as model systems, we show that CRAMPS–MAR can provide a lower quantitation limit than standard approaches such as fast MAS with peak integration. Furthermore, CRAMPS–MAR was found to be robust for cases that are inapproachable by conventional ultra-fast (i.e., 100 kHz) MAS methods even when state-of-the-art SSNMR equipment was employed. Our results demonstrate CRAMPS–MAR as an alternative quantification technique that can generate new opportunities for analytical research.

Introduction

Active pharmaceutical ingredients (APIs) can exist as multiple solid-state forms such as polymorphs,1,2 salts,2,3 and hydrates.4 Since different forms can exhibit different physicochemical properties, drug development and manufacturing requires careful control over the desired form.1−4 For instance, APIs are often produced as salts to increase solubility and dissolution rates.2,3 Undesired conversion to the neutral form can compromise the drug performance. APIs can also display extensive polymorphism, and the identity of the polymorph can alter the solubility, dissolution rate, and bioavailability of the drug.1,2 Given the complex interplay between the solid-state forms and API performance, identification and quantification of API forms is crucial for pharmaceutical research.

API characterization can be a challenging task. The structural similarities between solid-state forms and the heterogeneous composition of a pharmaceutical product can obscure analysis.5−9 A promising method for API quantification is solid-state nuclear magnetic resonance (SSNMR) spectroscopy.10−12 SSNMR spectroscopy is a non-destructive technique that can be applied on crystalline and amorphous materials.13−20 Moreover, SSNMR spectroscopy can provide high selectivity. Structurally similar API forms can be distinguished via SSNMR spectroscopy.5−8,21−28 In some cases, SSNMR spectroscopy can outperform powder X-ray diffraction.5−7 For example, the isomorphous solvate hydrates of finasteride were found to give comparable powder X-ray diffractograms but clearly distinct 13C SSNMR spectra.5 Similar observations were made for the polymorphs of neotame.6 Furthermore, by judiciously choosing which nuclei to examine, API signals can be selectively detected. This has been demonstrated using nuclei such as 35Cl and 19F, where spectra of APIs were obtained without interferences from excipients.8,16,29−31 In contrast to many analytical techniques, SSNMR spectroscopy is inherently quantitative since the peak area is directly proportional to the number of spins. Thus, for some SSNMR experiments, quantification can be performed without calibration.18,19,21,22,31−33

13C SSNMR spectroscopy is commonly employed for API quantification since carbons are ubiquitous in APIs.15,20,22,25,33,34 Nonetheless, 13C SSNMR experiments can be time consuming due to the low natural abundance of 13C, and spectra with sufficient signal-to-noise ratio (SNR) can require days to aquire.21,35 While 13C isotopic enrichment can decrease experimental time,26 isotopic labeling can be expensive and labor intensive. Furthermore, isotopic labeling is incompatible with quality checks during manufacturing as pharmaceutical end products are not isotopically enriched. An alternative approach to reduce experimental time is 1H SSNMR spectroscopy since 1H has a significantly higher natural abundance and gyromagnetic ratio than 13C.21,35 For example, Hirsh et al. showed that 1H quantification can decrease the experimental time by one to three orders of magnitude compared to conventional 13C SSNMR approaches.35 However, acquiring 1H SSNMR data adequate for quantification can be a challenging task as well. Strong 1H–1H dipolar coupling interactions are often present in solids, resulting in broad linewidths and a severe decline in spectral resolution. A common method to boost spectral resolution is fast or ultra-fast magic-angle spinning (MAS). In this method, the sample is spun at speeds greater than 40 kHz, with 120 kHz being the maximum frequency currently commercially available.36,37 Fast or ultra-fast MAS has been applied on various APIs to increase 1H spectral resolution,14,21,30,35,38−42 with some using spinning speeds above 100 kHz.38,39,41,42 The boost in spectral resolution can facilitate quantitative studies. For instance, Li et al. employed an external magnetic field strength (Bo) of 9.40 T (νo(1H) = 400 MHz) and showed that 1H spectral resolution of pioglitazone hydrochloride (PioHCl) salts, an anti-diabetes drug, can be drastically improved at high spinning speeds .21 At 60 kHz, the 1H SSNMR spectra contain adequate resolution for quantification. Nonetheless, fast and/or ultra-fast MAS requires specialized probes that are costly. Another method to enhance 1H resolution is via combined rotation and multiple-pulse spectroscopy (CRAMPS),43−46 such as decoupling using mind-boggling optimization (DUMBO).47,48 These experiments rely on both sample spinning and radiofrequency (RF) pulses to reduce 1H linewidths. The combination allows for a slower spinning speed (e.g., ca. 12 kHz);43 thus, these experiments can be performed with readily available standard equipment. Furthermore, larger-diameter rotors can be employed at slower spinning speeds. This enables a higher volume of sample to be analyzed, which can provide a potentially better SNR. For instance, the larger rotors available for a spinning speed of 12 kHz (e.g., 3.2 mm rotor diameter) can boost the SNR by an approximate factor of 3 compared to the smaller rotors available for 60 kHz MAS (e.g., 1.2, or 1.3 mm rotor diameter). Additionally, CRAMPS can provide better resolution. Several works have shown that for spinning speeds up to 80 kHz, CRAMPS can yield higher resolution than MAS alone.49−54 Furthermore, Leskes et al. demonstrated that CRAMPS at slower spinning speeds can still give narrower signals than fast MAS.54 For glycine, CRAMPS spectra recorded at 35 to 55 kHz all provided higher resolution than fast MAS experiments at 65 kHz.

Besides challenges with data acquisition, 1H SSNMR data analysis can also be difficult. Several approaches can be used to quantitatively analyze SSNMR data, such as relaxation,15,18 chemometric,17,22,55 peak-integration,21,35 and reference-based methods.19,33 To the best of our knowledge, previous quantitative 1H SSNMR studies of APIs have only been performed with peak-integration-based methods.21,35 However, the main disadvantage of this approach is it requires well-resolved signals. Even when advanced 1H SSNMR techniques such as ultra-fast MAS and CRAMPS are used, the resulting spectra can still have inadequate resolution. The spectral congestion can render data analysis a challenging, if not impossible, task. Recently, Stueber and Dance showed that quantitative 13C and 19F SSNMR data can be processed via mixture analysis using references (MAR).33 In MAR, the mixture spectrum is fitted as a linear combination of the pure component spectra. The corresponding results provide insights into the mixture composition. MAR can offer the same level of accuracy as conventional peak-integration-based methods but does not require baseline resolution for an effortless analysis. Thus, MAR can be beneficial for quantitative 1H SSNMR studies, where the corresponding spectra are often highly congested.

Despite the benefits of MAR, analyzing CRAMPS spectra via MAR can be non-trivial. Since MAR employs pure component spectra to fit the mixture spectrum,33 the pure component spectra must resemble the constituents of the mixture spectrum for an accurate fit. For CRAMPS, spectral features such as resolution, lineshapes, and absolute peak positions are dictated by chemical shift scaling factors.56,57 To successfully apply MAR on CRAMPS spectra, any discrepancies between the scaling factors of the mixture and pure component spectra must be accounted for during data processing. However, standard CRAMPS data processing methods, such as obtaining the scaling factor via an external reference,43 assume identical scaling factors between datasets. These methods can therefore overlook differences between the scaling factors of the mixture and pure component spectra. As such, the standard methods are inadequate for MAR, and a different data processing approach is required.

Here, we present an SSNMR approach for quantifying solid-state forms of APIs, termed CRAMPS–MAR. In this method, data acquisition is performed with 1H CRAMPS, and analysis is accomplished with MAR. CRAMPS alleviates the need for specialized equipment without sacrificing spectral resolution, while MAR enables an effortless analysis of complex spectra. Thus, the combination, CRAMPS–MAR, allows for a rapid, accurate, and straightforward API quantification procedure. Our results demonstrate that CRAMPS can provide higher 1H spectral resolution than ultra-fast MAS, even when the ultra-fast MAS experiments are performed at a significantly higher Bo. Moreover, we have developed CRAMPS data processing procedures to compensate for variations between the scaling factors of the pure component and mixture spectra, allowing for an accurate MAR fit. Using mixtures containing PioHCl and pioglitazone (Pio), we demonstrate that CRAMPS–MAR outperforms previously reported results from fast MAS with peak integration.21 To show that our method is applicable for structurally complex systems, we studied a polymorphic blend of a pharmaceutically relevant steroid,58 Org OD 14, also known as 7αMNa or tibolone. Although the structural complexity of steroids typically leads to severely congested SSNMR spectra, CRAMPS–MAR was able to provide accurate quantification results. Analogous quantification was found to be impractical with peak-integration-based methods even when the data was obtained using state-of-the-art ultra-fast MAS equipment.

Experimental Section

Sample Preparation

Pio (99% purity) was purchased from Doug Discovery. PioHCl (> 98% purity) was obtained from Tokyo Chemical Industry (TCI). Both Pio and PioHCl were used as purchased. Form I and form II polymorphs of Org OD 14 (Org-I and Org-II, respectively) were provided by Aspen Oss B.V. The chemical structures of Pio, PioHCl, and Org OD 14 are shown in Figure S1.

For the Pio/PioHCl systems, seven samples ranging from 5 to 97 weight percent (wt %) Pio were prepared. For the Org-I/-II sample, a bicomponent mixture containing 69.9 weight percent Org-I was made. Additional sample preparation details are given in the Supporting Information.

SSNMR Experiments

For samples containing Pio and/or PioHCl, all experiments were performed using a Varian VNMRS spectrometer at a Bo of 9.40 T (νo(1H) = 400 MHz), a Varian 3.2 mm T3 HXY probe, and large-volume rotors. The samples were restricted to the center third of the rotors. The exact sample weights inside the rotors were determined using an analytical balance and are ca. 30 mg. The 1H CRAMPS experiments were performed at 11.1 kHz MAS using a super-cycled59 windowed DUMBO (wDUMBO) pulse sequence with DUMBO-1 coefficients47 and a five-point phase ramp.60 The experimental temperature was regulated to 21 °C to avoid variations in Boltzmann distributions between the pure component and mixture spectra. The sample spinning speed, probe tuning frequency, and transmitter offset frequency (TOF) were optimized for resolution (see the Supporting Information for details). Moreover, CRAMPS spectra can suffer from artifacts such as rotary resonance frequency (RRF) lines. The positions of RRF lines are determined by MAS and CRAMPS cycle frequencies and are independent of the TOF.61,62 Since RRF lines do not contain any weight percent information, overlaps between sample signals and RRF lines were minimized by carefully adjusting the TOF without decreasing spectral resolution.43 The absolute chemical shift scaling factors were determined using the NH3 and CH2 signals of α-glycine at 8.0 and 2.6 ppm, respectively.63 The 1H chemical shifts were referenced using α-glycine (δiso(CH2) = 2.6 ppm).63

For the Org OD 14 samples, the 1H CRAMPS experiments were performed using the same setup and/or conditions as the Pio and PioHCl samples. Ultra-fast MAS 1H SSNMR experiments were performed on a Varian VNMRS spectrometer at a Bo of 19.96 T (νo(1H) = 850 MHz) with a Bruker 0.7 mm ultra-fast MAS probe. The spectra were acquired with a one-pulse experiment at 100 kHz MAS. The 1H chemical shifts were referenced using α-glycine (δiso(NH3) = 8.0 ppm)63 as a secondary reference.

The ultra-fast MAS spectra were fully processed using ssNake.64 For the CRAMPS spectra, ssNake64 was employed for truncation, zero-filling, Fourier transform, and phasing. The pure component and mixture spectra were first phased independently to absorption mode. The spectra were then overlaid for a visual inspection. If the spectra were out of phase with each other, the zeroth- and/or first-order phasing was adjusted until the spectra are in-phase. MAR processing was performed using an in-house MATLAB65 script. Additional experimental and data processing details and the CRAMPS–MAR processing MATLAB scripts are provided in the Supporting Information.

Results and Discussion

Description of CRAMPS–MAR

In the MAR method, the spectrum for a mixture with N components is represented as a weighted linear sum of the pure component spectra (p1, p2, ..., pN).33 The mixture spectrum, m, can therefore be expressed as

graphic file with name ac2c02905_m001.jpg

where Im,i denotes the spectral intensity of m at frequency i, Ix,i denotes the spectral intensity of px at frequency i, cx is the weighting coefficient for px, and ei represents the residual error at frequency i (e.g., noise). By fitting Im with Ix (x = 1 to N) via least-squares, cx, and therefore the mixture composition, can be determined.

Figure S2 illustrates the complete workflow of CRAMPS–MAR. To obtain accurate quantification results, the appearance of the pure component spectra must represent the components in the mixture spectrum. Therefore, the spectra must be in-phase with each other. Significant deviation in phasing will give inaccurate CRAMPS–MAR results (see the Supporting Information for a detailed discussion). Moreover, for 1H CRAMPS experiments, spectral features such as resolution, lineshape, and peak positions are dictated by chemical shift scaling factors, which are sensitive to experimental settings.56,57 These settings include the sample position relative to the RF coil, probe tuning frequency, decoupling RF field strength, and TOF.56,57 We found that the spectral resolution and lineshapes of the pure and mixture spectra are comparable if the sample positions are restricted to the center third of the rotors, and the spectra are recorded using the same probe tuning frequency, RF field strength, and TOF. However, an offset in peak positions was still observed due to small differences in the scaling factors (Figure S7).43,56,57 Thus, solely regulating the experimental settings is insufficient to standardize the scaling factors between different spectra. We have developed additional data processing procedures to account for the variations in scaling factors. These procedures involve scaling and aligning the pure component spectra to the mixture spectrum. The scaling factor must also be applied on the y-axis for the peak areas to remain the same, allowing for quantitative analysis. An in-depth description on the scaling procedure is given in the Supporting Information.

Due to the scaling, the data points of each spectrum will occur at different frequencies. Consequently, the spectra must be binned. Excessive binning will degrade the apparent spectral resolution; thus, minimal binning should be conducted. A visual comparison between the pre- and post-binned spectra can be employed as a gauge. If the apparent spectral resolution of the post-binned spectrum is comparable to the pre-binned spectrum, then the binning is adequate. Practically, spectral resolution will be preserved upon binning if an additional zero-fill is performed during FID processing, and the Fourier transformed data is binned to at most two data points per bin. A detailed discussion on binning, and how it influences fit results and analysis time, is provided in the Supporting Information.

To obtain the mixture composition with MAR, the spectra must be normalized to the sample weight and number of scans (assuming constant temperature and Bo). A least-square fitting of Im using Ix (x = 1 to N) will then provide the mixture composition in weight percent. A weight of 0 should be assigned to the RRF lines in the least-square fit region as they do not contain any weight percent information.

Quantification of Pio/PioHCl Mixtures

The 1H wDUMBO spectra of pure Pio and PioHCl are given in Figure 1. For both samples, the highest spectral resolution is observed in the region of ca. 11 to 17 ppm. This region contains 1H signals corresponding to nitrogen-bound protons, i.e., N–H and N·HCl.21 In the aromatic and aliphatic regions (ca. 0 to 9 ppm), the spectra are more crowded. Nonetheless, distinct spectral features can still be observed, and the resolution is notably higher than spectra recorded at the same Bo (9.40 T) using fast MAS at 60 kHz (figure 3 of ref (21)). The increase in resolution is the most apparent in the 6 to 10 ppm region of Pio and the 0 to 10 ppm region of PioHCl. A similar behavior has been observed for glycine, where wDUMBO experiments at moderate spinning speeds (35 to 55 kHz) provided higher spectral resolution than MAS experiments at 65 kHz.54 The resolution of MAS spectra can be further increased via ultra-fast MAS (i.e., spinning speed ≥100 kHz).66,67 Nonetheless, the 1H linewidths do not scale linearly with spinning speeds (see figure 5 of ref (66)). For example, the 1H linewidths of l-histadine·HCl·H2O were only reduced by a maximum of ca. 37% when the spinning speed was increased from 60 to 120 kHz, which is the current limit of commercial equipment.67 As such, we expect the wDUMBO results to still compare favorably to those obtained using ultra-fast MAS. This was indeed observed for Org OD 14, where the wDUMBO spectra displayed a higher resolution than those acquired at a spinning speed of 100 kHz (vide infra).

Figure 1.

Figure 1

Open in a new tab

The 1H wDUMBO SSNMR spectra of pure (a) PioHCl and (b) Pio, and the corresponding chemical structures. Asterisks (*) denote RRF lines.

We have quantified seven bicomponent samples containing 5 to 97 wt % Pio using CRAMPS–MAR. The exact sample compositions and the CRAMPS–MAR results are given in Table S8. Figure 2 shows the results for Pio 70%. The CRAMPS–MAR fit agrees well with the experimental spectrum, and the residue spectrum does not resemble sample signals. Moreover, if the pure component spectra were properly scaled and aligned to the mixture spectrum, the difference spectrum obtained from m – c1p1 should mimic p2. Figure 2 shows that the spectrum of m – cPIOpPIO resembles the pure PioHCl spectrum while the spectrum of m – cPIOHClpPIOHCl looks like the pure Pio spectrum. Thus, the pure component spectra were adequately scaled and aligned. Comparable fit quality was found for all other samples, and the respective spectra are provided in Figure S12.

Figure 2.

Figure 2

Open in a new tab

(a) The experimental 1H wDUMBO spectrum (black trace) and the corresponding CRAMPS–MAR fit (red dotted trace) for Pio 70%. The individual fit components (Pio: purple trace, PioHCl: green trace, residue error: yellow trace) are also provided. (b) The difference spectrum (black trace) obtained by subtracting the CRAMPS–MAR-predicted Pio contribution from the experimental spectrum, and the pure PioHCl spectrum (green trace). (c) The difference spectrum (black trace) acquired by removing the CRAMPS–MAR-predicted PioHCl contribution from the experimental spectrum, and the pure Pio spectrum (purple trace). Asterisks (*) denote RRF lines, and crosses (+) denote artifacts from applying 0 weight on the RRF lines during fitting.

To further evaluate the CRAMPS–MAR results, the weight percent values from CRAMPS–MAR are plotted in Figure 3 against the sample weight values. A good agreement is found, with a coefficient of determination (R2) value greater than 0.999. Moreover, the CRAMPS–MAR values are within ±1 wt % of the sample weight values (Table S8). This is reflected by the correlation plot (Figure 3), which displays a slope of near unity (0.998) and an intercept of ca. 0. The accuracy of our CRAMPS–MAR results compare favorably with those reported for MAR-based 13C and 19F SSNMR studies (slope = 0.99, intercept = 0.24 to 0.30).33

Figure 3.

Figure 3

Open in a new tab

Correlation between the Pio weight percent (Pio wt %) determined by sample weight and CRAMPS–MAR. The line of best fit is also provided (gray dotted line). Errors bars are not shown but are provided in Table S8.

Even though previous studies have shown that 13C and/or 19F SSNMR spectroscopy can provide accurate MAR results, the experiments are time consuming since each spectrum can require hours to record.33 On the other hand, 1H SSNMR experiments can be significantly faster due to the high gyromagnetic ratio and natural abundance of 1H. This is demonstrated using Pio 10% (Figure S13). We found that even at this low weight percent value, mixture spectra suitable for accurate quantification can be acquired in 6 min. Moreover, CRAMPS–MAR analysis can be accomplished within seconds (see the Supporting Information). It should be noted that the experimental time was not limited by the SNR but by the minimal number of scans required for complete phase cycling. Furthermore, we used a recycle delay that is five times longer than the longitudinal relaxation time (T1) values. The experimental time could be further decreased if the truMAR method, which uses a recycle delay of 1.2 times the T1 values, is employed.33 As shown by Stueber and Dance using 13C SSNMR spectroscopy, a ca. 75% reduction in experimental time can be obtained via truMAR.33

Pio/PioHCl bicomponent mixtures have been previously quantified using fast MAS experiments with a spinning speed of 60 kHz. Two-dimensional correlation experiments were used to assign the 11 to 17 ppm signals. The corresponding peaks in the one-dimensional spectra were integrated and normalized based on the signal identities. Using this method, samples containing ca. 10 to 90 wt % Pio were quantified.21 Here, we accurately quantified mixtures with ca. 5 and 97 wt % Pio without signal assignments, allowing for a less labor- and time-intensive analysis procedure. Furthermore, our CRAMPS–MAR results (5 to 97 wt % Pio) exceed the limits demonstrated by fast MAS with signal integration (10 to 90 wt % Pio). The success of our methods can be attributed to two main factors. First, CRAMPS experiments can offer superior spectral resolution compared to fast MAS experiments. For MAR, high resolution provides more spectral features for the least-square fit, resulting in a more sensitive quantification. Second, MAR has higher tolerance for peak overlap than the signal integration method. In the signal integration approach, accurate quantification requires well-resolved peaks. Thus, for Pio/PioHCl, only the ca. 11 to 17 ppm region is usable (Figure 1). However, at lower weight percent, the peaks in this region can become undetectable, making quantification by signal integration unfeasible. This is seen in the Pio 5% spectrum (Figure 4), where the Pio signal at ca. 14 ppm is no longer observed. On the other hand, the MAR method is based on “fingerprinting” and can therefore be applied on highly congested regions. Hence, the 0 to 10 ppm region can be analyzed. Figure 4 shows that the CRAMPS–MAR method provides an excellent fit. The difference spectrum of m – cpioHClppioHCl looks similar to pure Pio in the 0 to 10 ppm range, verifying that this spectral region contains useful quantitative data. Thus, by utilizing valuable spectral regions that are inaccessible to the signal integration approach, the CRAMPS–MAR method outperforms the signal integration method.

Figure 4.

Figure 4

Open in a new tab

(a) The 1H wDUMBO spectrum (black trace) of Pio 5% with the corresponding CRAMPS–MAR fit (red dotted trace). (b) The difference spectrum (black trace) generated by subtracting the CRAMPS–MAR-predicted PioHCl contribution from the experimental spectrum. The pure Pio spectrum (purple trace) is also provided for reference. Asterisks (*) denote RRF lines, and crosses (+) denote artifacts from applying 0 weight on the RRF lines during fitting.

Quantification of Org OD 14

To the best of our knowledge, 1H SSNMR-based quantification of steroids has yet to be attempted, likely due to the structural complexity of these compounds. To further investigate the limits of CRAMPS–MAR, we have quantified a steroid, Org OD 14, using a sample (Org-I 70%) that contains 69.9 wt % of polymorph Org-I and 30.1 wt % of polymorph Org-II. Previous studies have shown that Org-I/-II mixtures can be quantified via natural abundance 13C SSNMR spectroscopy.58 However, this can be time consuming. For example, in our experience, it can take up to a week to acquire quantifiable 13C SSNMR data for Org-I/-II samples with 1 wt % Org-II, even when sensitivity enhancement techniques such as 1H →13C cross-polarization are employed.

Figure 5 shows the 1H wDUMBO spectra of Org-I and Org-II. Based on the number of crystallographically inequivalent hydrogens in the Org-I and Org-II crystal structures, 56 and 28 1H SSNMR signals, respectively, can be expected.68,69 However, only ca. six and five distinct spectral features were observed in our Org-I and Org-II spectra, respectively. The spectral congestion is attributed to the complex chemical structure of Org OD 14 and the presence of strong 1H–1H dipolar interactions. For comparison, we have also acquired 1H SSNMR spectra of Org-I and Org-II using an ultra-fast MAS speed of 100 kHz and an ultra-high Bo of 19.96 T (Figure 5). Remarkably, the 1H wDUMBO spectra display higher resolution even though they were acquired with a substantially slower spinning speed (ca. 11 kHz) and at ca. half the Bo value (9.40 T). This is the most notable in the ca. 2 to 4 ppm region of Org-I, where the wDUMBO spectra present more detailed spectral features. The increase in resolution is also reflected by the full width at half maximum (FWHM) value of the well-resolved Org-I signal at 4.5 ppm. An FWHM of 350 Hz (0.41 ppm) was observed at 100 kHz spinning, while an FWHM of 140 Hz (0.35 ppm, rescaled linewidth) was found using wDUMBO. Further increasing the spinning speed to the currently commercially available limit of 120 kHz will have negligible influence as 1H linewidths do not scale linearly with spinning speed.66 To boost the spectral resolution in a meaningful manner, the spinning speed needs to be increased substantially. Nonetheless, even when such technology is available, the rotors will be impractically small, potentially resulting in sample preparation and SNR issues.

Figure 5.

Figure 5

Open in a new tab

The chemical structure of Org OD 14, and the 1H SSNMR spectra of (a) Org-I and (b) Org-II recorded using wDUMBO (top) and ultra-fast MAS (bottom) experiments. Asterisk (*) denotes impurity.

The pure component spectra (Figure 5) have similar appearances as Org-I and Org-II are polymorphs. Furthermore, the resolution of these spectra indicates that absolute quantification is inaccessible by peak integration. This is demonstrated by the 1H wDUMBO spectrum of Org-I 70% shown in Figure 6a. None of the Org OD 14 1H signals are well resolved except for the Org-I signal at ca. 4.5 ppm. The lack of distinct peaks for Org-II renders absolute quantification using peak integration impractical. Hirsh et al. showed that peak-integration-based quantification can be accomplished using a single peak.35 However, a calibration plot is required, leading to an increase in sample preparation and experimental time.

Figure 6.

Figure 6

Open in a new tab

(a) The 1H wDUMBO spectrum (black trace) and the CRAMPS–MAR fit (dotted red trace) for Org-I 70%. The individual fit components are also provided (Org-I: purple trace, Org-II: green trace, residue error: yellow trace). (b) The difference spectrum (black trace) calculated by subtracting the CRAMPS–MAR-predicted Org-II contribution from the experimental spectrum. The pure Org-I spectrum is also provided for reference (purple trace). (c) The difference spectrum (black trace) obtained by deducting the CRAMPS–MAR-predicted Org-I contribution from the experimental spectrum. The pure Org-II is also given (green trace).

Using CRAMPS–MAR, we successfully quantified the Org-I 70% sample with ease (Figure 6a). The mixture and reference spectra were each acquired within 8 min, allowing for rapid quantification. Figure 6a shows a good agreement between the CRAMPS–MAR fit and the experimental spectrum. The fit results were found to be highly accurate, arriving within ±1 wt % of the sample weight values (Table 1). No obvious sample signals were observed in the residue spectrum. Moreover, the corresponding difference spectra (m – cOrg-IpOrg-I and m – cOrg-IIpOrg-II) mimic the pure component spectra, indicating that the pure component spectra were properly scaled and aligned (Figure 6b,c).

Table 1. The Weight Percent of Org-I in the Org-I/-II Mixture Determined by Sample Weight and CRAMPS–MARa.

method Org-I
sample weightb 69.9 ± 0.4
CRAMPS–MARc 70.3 ± 2.3
Open in a new tab
a

All errors are expressed as 95% confidence intervals.

b

Errors derived from the uncertainty associated with the analytical balance.

c

Errors were propagated from the least-square fitting errors.

To estimate the quantification limit of CRAMPS–MAR for Org-I/-II mixtures, we have simulated a series of mixture spectra using the pure component spectra (Figure S14). According to our Pio/PioHCl results, CRAMPS–MAR can accurately quantify mixtures that display spectral features for the minor component. For Org-I and Org-II, the corresponding signals can still be seen when the mixture contains 10 and 85 wt % of Org-I, respectively. Thus, we estimate the quantification limit to be ca. 10 wt % for Org-I and ca. 15 wt % for Org-II. The experimental limit is likely slightly lower than our estimate since, as shown by our Pio 5% results, accurate quantification can be accomplished even when the minor component signals are not clearly observed in the mixture spectra. Our Org OD 14 results thus demonstrate the broad applicability of CRAMPS–MAR in quantifying solid-state structures of pharmaceuticals. In particular, accurate and calibration-free quantification can be accomplished even when analogous analyses cannot be realized using ultra-fast MAS with peak integration.

Spectral resolution plays a crucial role in a successful CRAMPS–MAR analysis. However, the 1H spectra of structurally complex APIs can be highly congested. As seen with Org-I/-II, the corresponding mixture spectrum can exhibit very little spectral features for the minor component. Moreover, spectral crowding will intensify as more components (> 2) are added into the mixture. Thus, we expect the performance of CRAMPS–MAR to decline as the analytes’ structures and/or the mixture becomes more complex. Nonetheless, the resolution of CRAMPS is continuously being improved.49,51,53,70,71 With the recent progress, we believe that CRAMPS–MAR can still advance, allowing complex, multi-component pharmaceutical mixtures to be quantified with ease.

Conclusions

In this work, we proposed a novel SSNMR approach, CRAMPS–MAR, for rapid quantification of APIs. CRAMPS can provide high-resolution 1H SSNMR data with standard equipment, while MAR allows complex data to be effortlessly analyzed. By combining the two techniques, a rapid, straightforward, and sensitive quantification can be realized. First, quantification can be accomplished within minutes, which is significantly faster than the common 13C SSNMR approach. Second, CRAMPS–MAR data analysis is simple to conduct. Unlike the traditional peak-integration method, accurate CRAMPS–MAR analysis does not entail peak assignment. Third, the “fingerprinting” nature of CRAMPS–MAR allows for a more relaxed prerequisite on spectral resolution, thereby enabling a lower quantification limit.

By utilizing CRAMPS–MAR, we successfully quantified mixtures of Pio/PioHCl and Org-I/-II. Our results surpass those obtained using fast and/or ultra-fast MAS with peak integration. For Pio/PioHCl, CRAMPS–MAR provided a lower quantification limit by utilizing crowded spectral regions that are inaccessible for fast MAS with peak integration. Consequently, we were able to quantify mixtures with 5 wt % of Pio and 3 wt % of PioHCl, respectively. The corresponding results were extremely accurate, landing within ±1 wt % of the sample weight values. To demonstrate the robustness of CRAMPS–MAR, we used a polymorphic mixture of a steroid (Org OD 14). We found absolute quantification to be unfeasible by ultra-fast MAS with peak integration. This is attributed to the spectral complexity, which was unalleviated even when state-of-the-art equipment were employed. Remarkably, higher resolution was observed in our wDUMBO spectra, and CRAMPS–MAR accurately determined the composition of the Org-I/-II mixture. Thus, our results demonstrate the suitability of CRAMPS–MAR in previously unachievable quantification analyses. Although our study was conducted on API mixtures, the methodology described should be extendable to other complex mixtures. Further advances in the resolution of CRAMPS spectra will certainly lower the limitations and increase the applicability of CRAMPS–MAR. As such, we believe that CRAMPS–MAR will evolve into an indispensable tool in analytical sciences.

Acknowledgments

The authors thank Ernst R. H. van Eck and Hans Jansen (MRRC) for their technical expertise. We thank Dr. Edwin Kellenbach (Aspen Oss B.V.) for fruitful discussions. Support of the Dutch Research Council (NWO)-funded Netherlands’ Magnetic Resonance Research School (NMARRS, project number 022.005.029) and the NWO-funded solid-state NMR facility for advanced materials science, which is part of the uNMR-NL ROADMAP facilities (project number 184.035.002), are gratefully acknowledged. Y.T.A.W. also acknowledges the Natural Sciences and Engineering Research Council of Canada (NSERC) for a graduate scholarship.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.analchem.2c02905.

  • Pio, PioHCl, and Org OD 14 chemical structures; additional sample preparation details; additional experimental and data processing details; CRAMPS–MAR procedure schematic; CRAMPS–MAR phasing discussion; CRAMPS–MAR scaling procedure discussion; CRAMPS–MAR binning discussion; Pio/PioHCl sample compositions and the CRAMPS–MAR results; CRAMPS–MAR fits for Pio 97 to 5% samples; CRAMPS–MAR fits for Pio 10% with 6 min and 1 h acquisition time; simulated Org-I/-II mixture spectra; in-house MATLAB script used for CRAMPS–MAR (PDF)

The authors declare no competing financial interest.

Supplementary Material

ac2c02905_si_001.pdf (2MB, pdf)

References

  1. Censi R.; Di Martino P. Polymorph Impact on the Bioavailability and Stability of Poorly Soluble Drugs. Molecules 2015, 20, 18759–18776. 10.3390/molecules201018759. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Williams H. D.; Trevaskis N. L.; Charman S. A.; Shanker R. M.; Charman W. N.; Pouton C. W.; Porter C. J. H. Strategies to Address Low Drug Solubility in Discovery and Development. Pharmacol. Rev. 2013, 65, 315–499. 10.1124/pr.112.005660. [DOI] [PubMed] [Google Scholar]
  3. Serajuddin A. T. M. Salt Formation to Improve Drug Solubility. Adv. Drug Delivery Rev. 2007, 59, 603–616. 10.1016/j.addr.2007.05.010. [DOI] [PubMed] [Google Scholar]
  4. Healy A. M.; Worku Z. A.; Kumar D.; Madi A. M. Pharmaceutical Solvates, Hydrates and Amorphous Forms: A Special Emphasis on Cocrystals. Adv. Drug Delivery Rev. 2017, 117, 25–46. 10.1016/j.addr.2017.03.002. [DOI] [PubMed] [Google Scholar]
  5. Othman A.; Evans J. S. O.; Evans I. R.; Harris R. K.; Hodgkinson P. Structural Study of Polymorphs and Solvates of Finasteride. J. Pharm. Sci. 2007, 96, 1380–1397. 10.1002/jps.20940. [DOI] [PubMed] [Google Scholar]
  6. Padden B. E.; Zell M. T.; Dong Z.; Schroeder S. A.; Grant D. J. W.; Munson E. J. Comparison of Solid-State 13C NMR Spectroscopy and Powder X-ray Diffraction for Analyzing Mixtures of Polymorphs of Neotame. Anal. Chem. 1999, 71, 3325–3331. 10.1021/ac9902446. [DOI] [Google Scholar]
  7. Testa C. G.; Prado L. D.; Costa R. N.; Costa M. L.; Linck Y. G.; Monti G. A.; Cuffini S. L.; Rocha H. V. A. Challenging Identification of Polymorphic Mixture: Polymorphs I, II and III in Olanzapine Raw Materials. Int. J. Pharm. 2019, 556, 125–135. 10.1016/j.ijpharm.2018.12.008. [DOI] [PubMed] [Google Scholar]
  8. Namespetra A. M.; Hirsh D. A.; Hildebrand M. P.; Sandre A. R.; Hamaed H.; Rawson J. M.; Schurko R. W. 35Cl Solid-state NMR Spectroscopy of HCl Pharmaceuticals and Their Polymorphs in Bulk and Dosage Forms. CrystEngComm 2016, 18, 6213–6232. 10.1039/C6CE01069E. [DOI] [Google Scholar]
  9. Nie H.; Liu Z.; Marks B. C.; Taylor L. S.; Byrn S. R.; Marsac P. J. Analytical Approaches to Investigate Salt Disproportionation in Tablet Matrices by Raman Spectroscopy and Raman Mapping. J. Pharm. Biomed. Anal. 2016, 118, 328–337. 10.1016/j.jpba.2015.10.024. [DOI] [PubMed] [Google Scholar]
  10. Li M.; Xu W.; Su Y. Solid-state NMR Spectroscopy in Pharmaceutical Sciences. TrAC, Trends Anal. Chem. 2020, 135, 116152. [Google Scholar]
  11. Geppi M.; Mollica G.; Borsacchi S.; Veracini C. A. Solid-State NMR Studies of Pharmaceutical Systems. Appl. Spectrosc. Rev. 2008, 43, 202–302. 10.1080/05704920801944338. [DOI] [Google Scholar]
  12. Stephenson G. A.; Forbes R. A.; Reutzel-Edens S. M. Characterization of the Solid State: Quantitative Issues. Adv. Drug Delivery Rev. 2001, 48, 67–90. 10.1016/S0169-409X(01)00099-0. [DOI] [PubMed] [Google Scholar]
  13. Lu X.; Huang C.; Li M.; Skomski D.; Xu W.; Yu L.; Byrn S. R.; Templeton A. C.; Su Y. Molecular Mechanism of Crystalline-to-Amorphous Conversion of Pharmaceutical Solids from 19F Magic Angle Spinning NMR. J. Phys. Chem. B 2020, 124, 5271–5283. 10.1021/acs.jpcb.0c02131. [DOI] [PubMed] [Google Scholar]
  14. Lu X.; Xu W.; Hanada M.; Jermain S. V.; Williams R. O. III; Su Y. Solid-State NMR Analysis of Crystalline and Amorphous Indomethacin: An Experimental Protocol for Full Resonance Assignments. J. Pharm. Biomed. Anal. 2019, 165, 47–55. 10.1016/j.jpba.2018.11.001. [DOI] [PubMed] [Google Scholar]
  15. Nasu M.; Nemoto T.; Mimura H.; Sako K. Development of Qualitative and Quantitative Analysis Methods in Pharmaceutical Application with New Selective Signal Excitation Methods for 13C Solid-state Nuclear Magnetic Resonance Using 1H T1rho Relaxation Time. J. Pharm. Sci. 2013, 102, 154–161. 10.1002/jps.23345. [DOI] [PubMed] [Google Scholar]
  16. Wenslow R. M. 19F Solid-State NMR Spectroscopic Investigation of Crystalline and Amorphous Forms of a Selective Muscarinic M3 Receptor Antagonist, in Both Bulk and Pharmaceutical Dosage Form Samples. Drug Dev. Ind. Pharm. 2002, 28, 555–561. 10.1081/DDC-120003451. [DOI] [PubMed] [Google Scholar]
  17. Brus J.; Urbanova M.; Sedenkova I.; Brusova H. New Perspectives of 19F MAS NMR in the Characterization of Amorphous Forms of Atorvastatin in Dosage Formulations. Int. J. Pharm. 2011, 409, 62–74. 10.1016/j.ijpharm.2011.02.030. [DOI] [PubMed] [Google Scholar]
  18. Liu J.; Nagapudi K.; Kiang Y.-H.; Martinez E.; Jona J. Quantification of Compaction-induced Crystallinity Reduction of a Pharmaceutical Solid Using 19F Solid-state NMR and Powder X-ray Diffraction. Drug Dev. Ind. Pharm. 2009, 35, 969–975. 10.1080/03639040902729424. [DOI] [PubMed] [Google Scholar]
  19. Huang C.; Klinzing G.; Procopio A.; Yang F.; Ren J.; Burlage R.; Zhu L.; Su Y. Understanding Compression-Induced Amorphization of Crystalline Posaconazole. Mol. Pharmaceutics 2019, 16, 825–833. 10.1021/acs.molpharmaceut.8b01122. [DOI] [PubMed] [Google Scholar]
  20. Saïdi F.; Taulelle F.; Martineau C. Quantitative 13C Solid-State NMR Spectra by Multiple-Contact Cross-polarization for Drug Delivery: From Active Principles to Excipients and Drug Carriers. J. Pharm. Sci. 2016, 105, 2397–2401. 10.1016/j.xphs.2016.05.025. [DOI] [PubMed] [Google Scholar]
  21. Li M.; Lu X.; Xu W.; Troup G. M.; McNevin M. J.; Nie H.; Su Y. Quantifying Pharmaceutical Formulations from Proton Detected Solid-State NMR under Ultrafast Magic Angle Spinning. J. Pharm. Sci. 2020, 109, 3045–3053. 10.1016/j.xphs.2020.06.026. [DOI] [PubMed] [Google Scholar]
  22. Virtanen T.; Maunu S. L. Quantitation of a Polymorphic Mixture of an Active Pharmaceutical Ingredient with Solid State 13C CPMAS NMR Spectroscopy. Int. J. Pharm. 2010, 394, 18–25. 10.1016/j.ijpharm.2010.04.017. [DOI] [PubMed] [Google Scholar]
  23. Burgess K. M. N.; Perras F. A.; Lebrun A.; Messner-Henning E.; Korobkov I.; Bryce D. L. Sodium-23 Solid-state Nuclear Magnetic Resonance of Commercial Sodium Naproxen and Its Solvates. J. Pharm. Sci. 2012, 101, 2930–2940. 10.1002/jps.23196. [DOI] [PubMed] [Google Scholar]
  24. Maruyoshi K.; Iuga D.; Watts A. E.; Hughes C. E.; Harris K. D. M.; Brown S. P. Assessing the Detection Limit of a Minority Solid-State Form of a Pharmaceutical by 1H Double-Quantum Magic-angle Spinning Nuclear Magnetic Resonance Spectroscopy. J. Pharm. Sci. 2017, 106, 3372–3377. 10.1016/j.xphs.2017.07.014. [DOI] [PubMed] [Google Scholar]
  25. Tinmanee R.; Larsen S. C.; Morris K. R.; Kirsch L. E. Quantification of Gabapentin Polymorphs in Gabapentin/excipient Mixtures Using Solid State 13C NMR Spectroscopy and X-ray Powder Diffraction. J. Pharm. Biomed. Anal. 2017, 146, 29–36. 10.1016/j.jpba.2017.07.048. [DOI] [PubMed] [Google Scholar]
  26. Booy K.-J.; Wiegerinck P.; Vader J.; Kaspersen F.; Lambregts D.; Vromans H.; Kellenbach E. The Use of 13C Labeling to Enhance the Sensitivity of 13C Solid-state CPMAS NMR to Study Polymorphism in Low Dose Solid Formulations. J. Pharm. Sci. 2005, 94, 458–463. 10.1002/jps.20236. [DOI] [PubMed] [Google Scholar]
  27. Griffin J. M.; Martin D. R.; Brown S. P. Distinguishing Anhydrous and Hydrous Forms of an Active Pharmaceutical Ingredient in a Tablet Formulation Using Solid-State NMR Spectroscopy. Angew. Chem. Int. Ed. 2007, 46, 8036–8038. 10.1002/anie.200702582. [DOI] [PubMed] [Google Scholar]
  28. Veinberg S. L.; Johnston K. E.; Jaroszewicz M. J.; Kispal B. M.; Mireault C. R.; Kobayashi T.; Pruski M.; Schurko R. W. Natural Abundance 14N and 15N Solid-state NMR of Pharmaceuticals and Their Polymorphs. Phys. Chem. Chem. Phys. 2016, 18, 17713–17730. 10.1039/C6CP02855A. [DOI] [PubMed] [Google Scholar]
  29. Hirsh D. A.; Rossini A. J.; Emsley L.; Schurko R. W. 35Cl Dynamic Nuclear Polarization Solid-state NMR of Active Pharmaceutical Ingredients. Phys. Chem. Chem. Phys. 2016, 18, 25893–25904. 10.1039/C6CP04353D. [DOI] [PubMed] [Google Scholar]
  30. Lu X.; Skomski D.; Thompson K. C.; McNevin M. J.; Xu W.; Su Y. Three-Dimensional NMR Spectroscopy of Fluorinated Pharmaceutical Solids under Ultrafast Magic Angle Spinning. Anal. Chem. 2019, 91, 6217–6224. 10.1021/acs.analchem.9b00884. [DOI] [PubMed] [Google Scholar]
  31. Holmes S. T.; Hook J. M.; Schurko R. W. Nutraceuticals in Bulk and Dosage Forms: Analysis by 35Cl and 14N Solid-State NMR and DFT Calculations. Mol. Pharmaceutics 2022, 19, 440–455. 10.1021/acs.molpharmaceut.1c00708. [DOI] [PubMed] [Google Scholar]
  32. Hirsh D. A.; Su Y.; Nie H.; Xu W.; Stueber D.; Variankaval N.; Schurko R. W. Quantifying Disproportionation in Pharmaceutical Formulations with 35Cl Solid-State NMR. Mol. Pharmaceutics 2018, 15, 4038–4048. 10.1021/acs.molpharmaceut.8b00470. [DOI] [PubMed] [Google Scholar]
  33. Stueber D.; Dance Z. E. X. Component Quantification in Solids with the Mixture Analysis Using References Method. Anal. Chem. 2020, 92, 11095–11102. 10.1021/acs.analchem.0c01045. [DOI] [PubMed] [Google Scholar]
  34. Harris R. K.; Hodgkinson P.; Larsson T.; Muruganantham A. Quantification of Bambuterol Hydrochloride in a Formulated Product Using Solid-state NMR. J. Pharm. Biomed. Anal. 2005, 38, 858–864. 10.1016/j.jpba.2005.01.052. [DOI] [PubMed] [Google Scholar]
  35. Hirsh D. A.; Wijesekara A. V.; Carnahan S. L.; Hung I.; Lubach J. W.; Nagapudi K.; Rossini A. J. Rapid Characterization of Formulated Pharmaceuticals Using Fast MAS 1H Solid-State NMR Spectroscopy. Mol. Pharmaceutics 2019, 16, 3121–3132. 10.1021/acs.molpharmaceut.9b00343. [DOI] [PubMed] [Google Scholar]
  36. Dudek M. K.; Kaźmierski S.; Potrzebowski M. J. Fast and Very Fast MAS Solid State NMR Studies of Pharmaceuticals. Annu. Rep. NMR Spectrosc. 2021, 103, 97–189. 10.1016/bs.arnmr.2021.02.002. [DOI] [Google Scholar]
  37. Nishiyama Y. Fast Magic-angle Sample Spinning Solid-state NMR at 60–100kHz for Natural Abundance Samples. Solid State Nucl. Magn. Reson. 2016, 78, 24–36. 10.1016/j.ssnmr.2016.06.002. [DOI] [PubMed] [Google Scholar]
  38. Lu X.; Tsutsumi Y.; Huang C.; Xu W.; Byrn S. R.; Templeton A. C.; Buevich A. V.; Amoureux J.-P.; Su Y. Molecular Packing of Pharmaceuticals Analyzed with Paramagnetic Relaxation Enhancement and Ultrafast Magic Angle Pinning NMR. Phys. Chem. Chem. Phys. 2020, 22, 13160–13170. 10.1039/D0CP02049D. [DOI] [PubMed] [Google Scholar]
  39. Struppe J.; Quinn C. M.; Sarkar S.; Gronenborn A. M.; Polenova T. Ultrafast 1H MAS NMR Crystallography for Natural Abundance Pharmaceutical Compounds. Mol. Pharmaceutics 2020, 17, 674–682. 10.1021/acs.molpharmaceut.9b01157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Baias M.; Widdifield C. M.; Dumez J.-N.; Thompson H. P. G.; Cooper T. G.; Salager E.; Bassil S.; Stein R. S.; Lesage A.; Day G. M.; et al. Powder Crystallography of Pharmaceutical Materials by Combined Crystal Structure Prediction and Solid-state 1H NMR Spectroscopy. Phys. Chem. Chem. Phys. 2013, 15, 8069–8080. 10.1039/c3cp41095a. [DOI] [PubMed] [Google Scholar]
  41. Li M.; Meng F.; Tsutsumi Y.; Amoureux J.-P.; Xu W.; Lu X.; Zhang F.; Su Y. Understanding Molecular Interactions in Rafoxanide–Povidone Amorphous Solid Dispersions from Ultrafast Magic Angle Spinning NMR. Mol. Pharmaceutics 2020, 17, 2196–2207. 10.1021/acs.molpharmaceut.0c00317. [DOI] [PubMed] [Google Scholar]
  42. Mathew R.; Uchman K. A.; Gkoura L.; Pickard C. J.; Baias M. Identifying Aspirin Polymorphs from Combined DFT-based Crystal Structure Prediction and Solid-state NMR. Magn. Reson. Chem. 2020, 58, 1018–1025. 10.1002/mrc.4987. [DOI] [PubMed] [Google Scholar]
  43. Paruzzo F. M.; Emsley L. High-resolution 1H NMR of Powdered Solids by Homonuclear Dipolar Decoupling. J. Magn. Reson. 2019, 309, 106598. 10.1016/j.jmr.2019.106598. [DOI] [PubMed] [Google Scholar]
  44. Madhu P. K. High-resolution Solid-state NMR Spectroscopy of Protons with Homonuclear Dipolar Decoupling Schemes under Magic-angle Spinning. Solid State Nucl. Magn. Reson. 2009, 35, 2–11. 10.1016/j.ssnmr.2008.11.001. [DOI] [PubMed] [Google Scholar]
  45. Brown S. P. Probing Proton–proton Proximities in the Solid State. Prog. Nucl. Magn. Reson. Spectrosc. 2007, 50, 199–251. 10.1016/j.pnmrs.2006.10.002. [DOI] [Google Scholar]
  46. Coelho C.; Rocha J.; Madhu P. K.; Mafra L. Practical Aspects of Lee–Goldburg Based CRAMPS Techniques for High-resolution 1H NMR Spectroscopy in Solids: Implementation and Applications. J. Magn. Reson. 2008, 194, 264–282. 10.1016/j.jmr.2008.07.019. [DOI] [PubMed] [Google Scholar]
  47. Sakellariou D.; Lesage A.; Hodgkinson P.; Emsley L. Homonuclear Dipolar Decoupling in Solid-state NMR Using Continuous Phase Modulation. Chem. Phys. Lett. 2000, 319, 253–260. 10.1016/S0009-2614(00)00127-5. [DOI] [Google Scholar]
  48. Lesage A.; Sakellariou D.; Hediger S.; Eléna B.; Charmont P.; Steuernagel S.; Emsley L. Experimental Aspects of Proton NMR Spectroscopy in Solids Using Phase-modulated Homonuclear Dipolar Decoupling. J. Magn. Reson. 2003, 163, 105–113. 10.1016/S1090-7807(03)00104-6. [DOI] [PubMed] [Google Scholar]
  49. Nishiyama Y.; Lu X.; Trébosc J.; Lafon O.; Gan Z.; Madhu P. K.; Amoureux J.-P. Practical choice of 1H–1H Decoupling Schemes in Through-bond 1H–{X} HMQC Experiments at Ultra-fast MAS. J. Magn. Reson. 2012, 214, 151–158. 10.1016/j.jmr.2011.10.014. [DOI] [PubMed] [Google Scholar]
  50. Gan Z.; Madhu P. K.; Amoureux J.-P.; Trébosc J.; Lafon O. A Tunable Homonuclear Dipolar Decoupling Scheme for High-Resolution Proton NMR of Solids from Slow to Fast Magic-angle Spinning. Chem. Phys. Lett. 2011, 503, 167–170. 10.1016/j.cplett.2010.12.070. [DOI] [Google Scholar]
  51. Salager E.; Dumez J.-N.; Stein R. S.; Steuernagel S.; Lesage A.; Elena-Herrmann B.; Emsley L. Homonuclear Dipolar Decoupling with Very Large Scaling Factors for High-resolution Ultrafast Magic Angle Spinning 1H Solid-state NMR Spectroscopy. Chem. Phys. Lett. 2010, 498, 214–220. 10.1016/j.cplett.2010.08.038. [DOI] [Google Scholar]
  52. Salager E.; Stein R. S.; Steuernagel S.; Lesage A.; Elena B.; Emsley L. Enhanced Sensitivity in High-resolution 1H Solid-state NMR Spectroscopy with DUMBO Dipolar Decoupling Under Ultra-fast MAS. Chem. Phys. Lett. 2009, 469, 336–341. 10.1016/j.cplett.2008.12.073. [DOI] [Google Scholar]
  53. Tognetti J.; Franks W. T.; Lewandowski J. R.; Brown S. P. Optimisation of 1H PMLG Homonuclear Decoupling at 60 kHz MAS to Enable 15N– 1H Through-bond Heteronuclear Correlation Solid-state NMR Spectroscopy. Phys. Chem. Chem. Phys. 2022, 24, 20258–20273. 10.1039/D2CP01041K. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Leskes M.; Steuernagel S.; Schneider D.; Madhu P. K.; Vega S. Homonuclear Dipolar Decoupling at Magic-angle Spinning Frequencies Up to 65kHz in Solid-state Nuclear Magnetic Resonance. Chem. Phys. Lett. 2008, 466, 95–99. 10.1016/j.cplett.2008.10.029. [DOI] [Google Scholar]
  55. Zumbulyadis N.; Antalek B.; Windig W.; Scaringe R. P.; Lanzafame A. M.; Blanton T.; Helber M. Elucidation of Polymorph Mixtures Using Solid-State 13C CP/MAS NMR Spectroscopy and Direct Exponential Curve Resolution Algorithm. J. Am. Chem. Soc. 1999, 121, 11554–11557. 10.1021/ja9929786. [DOI] [Google Scholar]
  56. Brouwer D. H.; Horvath M. Minimizing the Effects of RF Inhomogeneity and Phase Transients Allows Resolution of Two Peaks in the 1H CRAMPS NMR Spectrum of Adamantane. Solid State Nucl. Magn. Reson. 2015, 71, 30–40. 10.1016/j.ssnmr.2015.10.005. [DOI] [PubMed] [Google Scholar]
  57. Vega A. J. Controlling the Effects of Pulse Transients and RF Inhomogeneity in Phase-modulated Multiple-pulse Sequences for Homonuclear Decoupling in Solid-state Proton NMR. J. Magn. Reson. 2004, 170, 22–41. 10.1016/j.jmr.2004.05.017. [DOI] [PubMed] [Google Scholar]
  58. Sas G. A. J. M. T.; Van Doornum E. M. Pharmaceutical Composition which Contains a Pharmaceutically Suitable Carrier and the Compound Having the Structure (7α,17α)-17-hydroxy-7-methyl-19-nor-17-pregn-5(10)-en-20-yn-3-one. United States 2006, 5037817. [Google Scholar]
  59. Paul S.; Thakur R. S.; Goswami M.; Sauerwein A. C.; Mamone S.; Concistrè M.; Förster H.; Levitt M. H.; Madhu P. K. Supercycled Homonuclear Dipolar Decoupling Sequences in Solid-state NMR. J. Magn. Reson. 2009, 197, 14–19. 10.1016/j.jmr.2008.11.011. [DOI] [PubMed] [Google Scholar]
  60. Brauckmann J. O.High Resolution μMAS of Nanoliter Volumes & Structural Studies of Solid Materials. PhD Thesis, Radboud University, 2019. [Google Scholar]
  61. Vega S.; Olejniczak E. T.; Griffin R. G. Rotor Frequency Lines in the Nuclear Magnetic Resonance Spectra of Rotating Solids. The Journal of Chemical Physics 1984, 80, 4832–4840. 10.1063/1.446504. [DOI] [Google Scholar]
  62. Vinogradov E.; Madhu P. K.; Vega S. Phase Modulated Lee–Goldburg Magic Angle Spinning Proton Nuclear Magnetic Resonance Experiments in the Solid State: A Bimodal Floquet Theoretical Treatment. J. Chem. Phys. 2001, 115, 8983–9000. 10.1063/1.1408287. [DOI] [Google Scholar]
  63. Stievano L.; Tielens F.; Lopes I.; Folliet N.; Gervais C.; Costa D.; Lambert J.-F. Density Functional Theory Modeling and Calculation of NMR Parameters: An ab Initio Study of the Polymorphs of Bulk Glycine. Cryst. Growth Des. 2010, 10, 3657–3667. 10.1021/cg100525h. [DOI] [Google Scholar]
  64. van Meerten S. G. J.; Franssen W. M. J.; Kentgens A. P. M. ssNake: A Cross-Platform Open-source NMR Data Processing and Fitting Application. J. Magn. Reson. 2019, 301, 56–66. 10.1016/j.jmr.2019.02.006. [DOI] [PubMed] [Google Scholar]
  65. MATLAB R2019b; The MathWorks, Inc.: Natick, Massachusetts, USA. [Google Scholar]
  66. Marchand T. L.; Schubeis T.; Bonaccorsi M.; Paluch P.; Lalli D.; Pell A. J.; Andreas L. B.; Jaudzems K.; Stanek J.; Pintacuda G. 1H-Detected Biomolecular NMR under Fast Magic-Angle Spinning. Chem. Rev. 2022, 122, 9943–10018. 10.1021/acs.chemrev.1c00918. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Pandey M. K.; Zhang R.; Hashi K.; Ohki S.; Nishijima G.; Matsumoto S.; Noguchi T.; Deguchi K.; Goto A.; Shimizu T.; Maeda H.; Takahashi M.; Yanagisawa Y.; Yamazaki T.; Iguchi S.; Tanaka R.; Nemoto T.; Miyamoto T.; Suematsu H.; Saito K.; Miki T.; Ramamoorthy A.; Nishiyama Y. 1020MHz Single-channel Proton Fast Magic Angle Spinning Solid-state NMR Spectroscopy. J. Magn. Reson. 2015, 261, 1–5. 10.1016/j.jmr.2015.10.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Declercq J.-P.; van Meerssche M.; Zeelen F. J. Conformational Analysis of 3-oxo 5(10)-unsaturated Steroids. Single-crystal X-ray Structure Analysis of 17-hydroxy-7α-methyl-19-nor-17α-pregn-5(10)-en-20-yn-3-one (Org OD 14). Recl. Trav. Chim. Pays-Bas 1984, 103, 145–147. [Google Scholar]
  69. Schouten A.; Kanters J. A. Structure of the Triclinic Modification of 17β-hydroxy-19-nor-7α-methyl-17α-pregn-5(10)-en-20-yn-3-one (ORG OD14). Acta Crystallogr., Sect. C 1991, 47, 1754–1756. 10.1107/S0108270190014196. [DOI] [Google Scholar]
  70. Aebischer K.; Wili N.; Tošner Z.; Ernst M. Using Nutation-frequency-selective Pulses to Reduce Radio-frequency Field Inhomogeneity in Solid-state NMR. Magnetic Resonance 2020, 1, 187–195. 10.5194/mr-1-187-2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Halse M. E.; Schlagnitweit J.; Emsley L. High-Resolution 1H Solid-State NMR Spectroscopy Using Windowed LG4 Homonuclear Dipolar Decoupling. Isr. J. Chem. 2014, 54, 136–146. 10.1002/ijch.201300101. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ac2c02905_si_001.pdf (2MB, pdf)

Articles from Analytical Chemistry are provided here courtesy of American Chemical Society

RESOURCES