Quantifying Disproportionation in Pharmaceutical Formulations with

Jul 17, 2018 - Search; Citation; Subject .... of an ionic salt of an API to a neutral form with different properties, through a process known as dispr...
0 downloads 0 Views 924KB Size
Subscriber access provided by MT ROYAL COLLEGE

Article

Quantifying Disproportionation in Pharmaceutical Formulations with Cl Solid-State NMR 35

David A. Hirsh, Yongchao Su, Haichen Nie, Wei Xu, Dirk Stueber, Narayan Variankaval, and Robert W. Schurko Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/acs.molpharmaceut.8b00470 • Publication Date (Web): 17 Jul 2018 Downloaded from http://pubs.acs.org on July 18, 2018

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 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 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.

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 34 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

Molecular Pharmaceutics

Quantifying Disproportionation in Pharmaceutical Formulations with 35Cl Solid-State NMR David A. Hirsh,1 Yongchao Su,2 Haichen Nie,2 Wei Xu,2 Dirk Stueber,2 Narayan Variankaval,2 and Robert W. Schurko1,*

1

Department of Chemistry & Biochemistry, University of Windsor, Windsor, ON,

N9B 3P4, Canada 2

Merck Research Laboratories, Merck & Co., Inc., Kenilworth, NJ, 07033, United

States * Author to whom correspondence should be addressed: R.W.S. Phone: 519-253-3000 x3548 E-mail: [email protected]

ACS Paragon Plus Environment

Molecular Pharmaceutics 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 34

Abstract Reliable methods for the characterization of drug substances are critical for evaluating their stability and bioavailability, especially in dosage formulations under varying storage conditions and usage. Such methods must provide information on the molecular identities and structures of drug substances and any potential by-products of the formulation process, as well as providing a means of quantifying the relative amounts of these substances. For example, active pharmaceutical ingredients (APIs) are often formulated as ionic salts to improve the pharmaceutical properties of dosage forms; however, exposure of such formulations to elevated temperature and/or humidity can trigger the conversion of an ionic salt of an API to a neutral form with different properties, through a process known as disproportionation. It is particularly challenging to identify changes of pharmaceutical components in solid dosage formulations, which are complex heterogeneous mixtures of the active pharmaceutical ingredient (API) and excipient components (e.g., binders, disintegrants, and lubricants). In this study, we illustrate that ultra-wideline (UW)

35

Cl solid-state NMR (SSNMR) can be used to characterize the

disproportionation reaction of pioglitazone HCl (PiogHCl) in mixtures with metallic stearate excipients.

35

Cl SSNMR can quantitatively detect the amount of PiogHCl in mixed samples

within ±1 wt-% and measure the degree of PiogHCl disproportionation in formulation samples stressed at high relative humidity and temperature. Unlike other methods used for characterizing disproportionation, our experiments directly probe the Cl− anions in both the intact salt and disproportionation products, revealing all of the chlorine-containing products in the solid-state chemical reaction without interfering signals from the formulation excipients. Keywords:

quantification,

solid-state

NMR,

dosage

formulations,

processing-induced phase transitions, chlorine-35, hydrochloride salts

2

ACS Paragon Plus Environment

disproportionation,

Page 3 of 34 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

Molecular Pharmaceutics

1. Introduction Active pharmaceutical ingredients (APIs) in dosage formulations are typically manufactured as salts, in which a weakly basic or acidic form of the API is charge balanced by a counterion. Salts are advantageous because they typically have high bioavailability and aqueous solubility, properties that are desirable for formulations.1–4 However, a pharmaceutical product containing an API salt can be problematic. For example, under certain circumstances, an API salt can undergo a proton exchange process and convert to a neutral form, which often has distinct physicochemical properties, including chemical and physical stability, processability and dissolution profile. This process is known as disproportionation.5–10 Many factors have been shown to influence the extent and rate of salt disproportionation in drug products, including the: (i) storage conditions, (ii) physicochemical properties of the API salt (e.g., pKa or intrinsic solubility), and (iii) properties of excipients (e.g., acidity or presence of a proton accepting group).5–14 In particular, disproportionation is known to be solution-mediated in many cases, and even small amounts of moisture can initiate the reaction (e.g., adsorption from the atmosphere).4 Exposure to moisture can occur during the formulation development, manufacturing, storage, and shipping processes, which makes disproportionation a major concern for the quality and viability of the final drug product. Therefore, the development and application of methods for identifying and quantifying the by-products of disproportionation, along with mechanistic studies of these reactions, are critical in the pharmaceutical sciences. Several methods have been applied successfully to quantitatively detect disproportionation in model formulations, including near infrared (NIR) and Raman spectroscopies, as well as X-ray and synchrotron diffraction methods.5,10,15–17 However, these techniques may not be applicable to all formulations; for instance, signal overlap from the

3

ACS Paragon Plus Environment

Molecular Pharmaceutics 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

formulation components can complicate the identification and quantification of signals from the API salt or disproportionation by-products. Diffraction techniques rely on long-range order, which may not be useful in samples where the free base or salt are amorphous, or in cases where interfering signals from the excipient matrix obscure those from the API.18 Therefore, alternative methods of detecting disproportionation are desirable. Solid-state nuclear magnetic resonance (SSNMR) spectroscopy is a useful method for characterizing drugs in bulk and dosage formulations.19–22 One of its primary advantages is an ability to detect signals from crystalline, amorphous, and even aqueous phases, which is important for studying solvent-mediated processes like disproportionation. SSNMR is also inherently quantitative, since the amount of signal is directly proportional to the number of NMR-active nuclei in the sample. Unlike some other quantification methods, quantitative NMR (qNMR) experiments do not generally require extensive instrument calibration or elaborate sample preparation,23–27 though some optimization of experimental conditions is often necessary (vide infra). Quantitative SSNMR studies have been used previously to measure relative amounts of crystalline or amorphous pharmaceutical components,28–32 and mixtures of crystalline pharmaceutical phases (e.g., different solid API forms);26,33–38 however, despite the potential advantages, there are relatively few examples of the use of SSNMR to study disproportionation.10,39,40 35

Cl SSNMR is a powerful technique for identifying and quantifying chlorine-containing

drug substances, and well-suited for the study of disproportionation in dosage formulations containing HCl salts. Such experiments directly probe the Cl− anions, a key structural component that is not probed by other methods, in order to detect local structural differences resulting from the disproportionation process in either bulk or dosage formulations. Furthermore, 35Cl SSNMR

4

ACS Paragon Plus Environment

Page 4 of 34

Page 5 of 34 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

Molecular Pharmaceutics

spectra of Cl− ions in dosage formulations are not influenced by the excipients, which rarely contain Cl– anions. Some formulation components that may contain covalently bound Cl atoms (e.g., C-Cl bonds) do not result in powder patterns that interfere with the relatively narrow and well-defined patterns of the Cl− anions.18 Chlorine-35 is a quadrupolar nucleus (I = 3/2), and its SSNMR spectra are affected by a combination of anisotropic quadrupolar and chemical shift interactions.41,42 The former has been shown to be particularly sensitive to the hydrogen-bonding environment of the Cl− anion,18,43–47 yielding a unique spectral fingerprint in each case. Primarily due to the magnitude of the quadrupolar interaction, central transition (CT, mI = +1/2 ↔ −1/2) 35

Cl SSNMR spectra typically have broad powder patterns on the order of hundreds of kHz or

larger, so-called ultra-wideline (UW) spectra;48 the breadths of these patterns necessitate the use of specialized pulse sequences that afford uniform excitation and increased S/N.49–51 The CarrPurcell-Meiboom-Gill (CPMG) sequence52,53 is commonly used for the acquisition of spectra featuring inhomogeneously broadened patterns; signal enhancement is achieved by repeated refocusing and detection of the spin polarization for as long as the transverse relaxation (T2) permits (vide infra).54,55 Broadband excitation of UW patterns can be achieved using wideband uniform-rate smooth truncation (WURST) pulses,56,57 which employ phase- and amplitudemodulated shapes to create an effective frequency-sweep of the transmitter across the pattern. WURST pulses can be utilized in the framework of a CPMG-type pulse sequence known as WURST-CPMG, for the combined benefits of signal enhancement and uniform excitation.58,59 Quantification using SSNMR can be complicated by several factors, including: temperature, sample filling factor and positioning, and the presence of spectral artifacts. These concerns, and methods to reduce their impact, have been discussed previously.25,30,34,60 The intensities of peaks or patterns arising from magnetically distinct sites, regardless of whether

5

ACS Paragon Plus Environment

Molecular Pharmaceutics 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

they are in the same molecule, same sample, or a mixture of constituents, can be influenced by distinct relaxation rates and mechanisms, which can further complicate quantification. In order to circumvent this problem, it is often necessary to calibrate experimental conditions and peak intensities against known standards.25,27 There are a number of concerns associated with quantitative comparison of powder patterns acquired with UW NMR methods, most of which are unexplored to date. First, the observed signal intensities in spectra acquired using CPMG methods are strongly related to T2 relaxation. The primary influence on T2(35Cl) relaxation in HCl APIs is the 1H-35Cl heteronuclear dipolar coupling mechanism; in most cases, the intensity of the 35Cl NMR signal can be improved dramatically with the use of high-power 1H decoupling (producing an effective T2(35Cl), called T2eff, where T2eff > T2). However, decoupling efficiency (and the resulting T2eff) can vary for different chlorine sites in a sample. A second concern relates to quantifying signals from quadrupolar nuclei (I > ½) due to differences in their nutation behaviour, which depend on the magnitude of the quadrupolar interactions for each nuclear environment.61,62 This problem can be avoided with the use of CT selective pulses (i.e., τselective = τnonselective×(I + 1/2)−1 ) or pulses with low-RF powers.63,64 Fortunately, nutation rates are not a concern when WURST pulses are utilized, since they typically employ low RF powers (vide infra).65 Despite these additional considerations, useful quantitative information can be obtained from UW NMR spectra by comparing the integrated signal intensity from a sample in a mixture to that of an external reference (e.g., the pure material). Similar techniques have been used in 1H13

C CP/MAS NMR studies, where a variety of relaxation rates and processes can complicate

quantification.26,30 Therefore, comparison of a 35Cl SSNMR spectrum of a sample in which disproportionation has occurred to that of the pure salt can be used to measure the amount of

6

ACS Paragon Plus Environment

Page 6 of 34

Page 7 of 34 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

Molecular Pharmaceutics

disproportionation (i.e., %-conversion of the salt) and detect Cl-containing by-products. Recently, the HCl salt form of pioglitazone (PiogHCl), a thiazolidinedione used for treating type-II diabetes, has been a popular test case for various proof-of-concept methods to detect disproportionation.10,15,16,66,67 PiogHCl is known to disproportionate in mixtures with salt stearates such as magnesium stearate (MgSt) and sodium stearate (NaSt), according to the proposed reaction scheme shown in Scheme 1.8,10,15 Such stearates are common lubricants in a variety of formulations, and are typically present in small quantities (e.g., 2-3 wt-% of the final dosage form). Herein, we describe the use of 35Cl SSNMR to quantify an HCl salt of pioglitazone and study disproportionation reactions in formulations containing stearate salts. Qualitatively, comparisons of 35Cl NMR spectra of the sample mixtures before and after treatment at high temperature and relative humidity (RH) can reveal the identity of disproportionation products involving the Cl− anions, which are not adequately characterized using other methods. We propose a simple quantitative method for measuring the amount of HCl salt in a dosage formulation, and demonstrate its utility by measuring the amount of disproportionation in stressed PiogHCl samples containing different stearate salts. 2. Methods 2.1. Chemicals The salt and free-base forms of pioglitazone (PiogHCl and PiogFB, respectively) were ordered from Tokyo Chemical Industry (TCI) Co, Ltd. (Portland, OR). Magnesium stearate (MgSt) was purchased from Mallinckrodt (St. Louis, MO). Sodium stearate (NaSt) was provided by Acme-Hardesty Co. (Blue Bell, PA). Chemical structures of the materials used in this study

7

ACS Paragon Plus Environment

Molecular Pharmaceutics 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

are shown in Figure S1. Samples were stored at 5 °C with desiccant to avoid water intake and degradation while not in use. 2.2. Sample Preparation Quantification standards were prepared by mixing various amounts of PiogHCl with the free base form to produce samples ranging from 3% to 100% HCl salt (wt/wt). These samples are referred to in the text using the wt-% PiogHCl present in the sample (e.g., PiogHCl 3%, PiogHCl 100%). Sample mixtures of PiogHCl with MgSt or NaSt were prepared with 90% API wt/wt. As discussed previously,66 this weight ratio is representative of a typical tablet formulation. Samples were weighed and prepared under inert atmosphere to minimize unintended water uptake. The sample codes and masses are listed in Table 1. Samples to be used as controls (i.e., API/stearate mixtures not exposed to high temperature or humidity, NaSt_C and MgSt_C) and the quantification standards were mixed for 5 minutes with a vortex mixer before the samples were packed. Stressed samples (NaSt_X and MgSt_X, respectively) were prepared using binary mixtures of PiogHCl and MgSt or NaSt placed into petri dishes, which were subjected to 40°C/75% RH in a stability chamber for 5 days in an open dish condition. Prior work on similar materials has demonstrated that the binary mixtures can be highly hygroscopic.68 As such, two rehydrated samples (NaSt_XH, and MgSt_XH, respectively), were prepared by placing approximately 300 mg of the NaSt_X or MgSt_X samples in a 75% RH environment at 22 °C for 4 hours. Preliminary 1H and 35Cl NMR experiments (as well as previous work on this sample)68 demonstrated that NaSt_X did not absorb significant amounts of H2O upon rehydration; therefore, the NaSt_XH sample was not examined further. All samples were ground with a mortar and pestle before being packed and weighed under an inert atmosphere. For magic-angle spinning (MAS) experiments, samples were packed

8

ACS Paragon Plus Environment

Page 8 of 34

Page 9 of 34 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

Molecular Pharmaceutics

into 4 mm o.d. zirconia rotors sealed with air-tight Teflon® screws. For experiments conducted under static sample conditions, samples were packed into glass tubes with one open end, which was then sealed using Teflon® tape and vacuum grease. 2.3. Solid-State NMR (SSNMR) Experiments All solid-state NMR (SSNMR) experiments were conducted on a Bruker Avance III HD console with an Oxford B0 = 9.4 T wide-bore magnet (ν0(1H) = 399.73 MHz, ν0(13C) = 100.53 MHz, ν0(23Na) = 105.74 MHz, ν0(35Cl) = 39.16 MHz). 1H and 35Cl experiments used a Varian/Chemagnetics 5mm HX static probe with a horizontal coil alignment. All other experiments used a Varian/Chemagnetics 4 mm HX MAS probe. 35

Cl chemical shifts were referenced with respect to NaCl(s) (δiso = 0.0 ppm). 1H and 13C

chemical shifts were referenced to tetramethylsilane (TMS, δiso = 0.0 ppm) using adamantane (δiso = 1.85 and 38.57 ppm, respectively) as a secondary reference. 23Na chemical shifts were referenced with respect to a 1.0 M NaCl(aq) solution (δiso = 0.0 ppm). A full list of experimental parameters used for the SSNMR experiments is given in the Supporting Information (Tables S1-S4). 35Cl{1H} spectra were acquired using the WURSTCPMG sequence.58,59 1H and 23Na{1H} direct-excitation MAS experiments were conducted using a rotor-synchronized Hahn echo pulse sequence of the form (π/2)x − τ1 − (π)y − τ2 − acq. 1H-13C cross-polarization (CP) experiments used a ramped-amplitude spin lock pulse on the 1H channel.69–71 High power 1H decoupling was applied in all 35Cl, 23Na, and 13C experiments, using either continuous wave (for 35Cl) or swept field two-pulse phase-modulation (sw-TPPM)72,73 (for 23

Na and 13C) decoupling sequences. All experiments were conducted at 20 °C with temperature

control provided by a Varian VT stack and Bruker BCU II 80/60 chiller.

9

ACS Paragon Plus Environment

Molecular Pharmaceutics 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

2.4. Powder X-ray Diffraction (PXRD) Experiments PXRD patterns for all of the samples were collected using a Bruker DISCOVER X-ray diffractometer with a Cu-Kα (λ = 1.54056 Å) radiation source and Bruker AXS HI-STAR area detector. Samples were packed into 0.9 mm o.d. glass capillary tubes and analyzed for 30 minutes with the detector centered at 2θ = 18º. Diffraction patterns were processed using the CrystalDiffract software package. 3. Results and Discussion 3.1. Pioglitazone HCl The 35Cl{1H} WURST-CPMG NMR spectrum of pioglitazone HCl (PiogHCl) acquired under static sample conditions is shown in Figure 1. The spectrum has a powder pattern spanning roughly 150 kHz with a shape typical of the second-order quadrupolar interaction. This pattern can be simulated with NMR tensor parameters as follows: δiso = 100(20) ppm, CQ = 4.8(1) MHz, and ηQ = 0.63(5). The shape of the pattern, and associated tensor parameters, are unique to the Cl− environment in PiogHCl, and therefore serve as a spectral fingerprint of the API.18,43,44 The data can be processed by either performing a Fourier transform (FT) of the echo train itself (producing a spectrum with a series of spikelets), or by coadding the echoes to form a single echo before the FT (producing a spectrum with a continuous line that traces the edge of the powder pattern, so-called echo coaddition).74 As seen in Figure 1, both processing methods yield patterns with the same shape. 3.2. PiogHCl:NaSt mixture 3.2.1.

35

Cl NMR Spectra

The 35Cl{1H} WURST-CPMG NMR spectra of the NaSt-containing samples are shown in Figure 2. The spectrum of the control mixture (NaSt_C, Figure 2a) has the same shape as the 10

ACS Paragon Plus Environment

Page 10 of 34

Page 11 of 34 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

Molecular Pharmaceutics

spectrum of pure PiogHCl, indicating that the physical mixing does not change the local structure of the Cl− ion. The spectrum of the stressed sample, NaSt_X (i.e., stored at 40 °C/75% RH for 5 days), has the same breadth and all of the discontinuities present in the spectrum of pure PiogHCl; however, this spectrum also has a distinct feature that results from the disproportionation reaction: a sharp peak at roughly 0 ppm (Figure 2b, see Figure S2 for a deconvolution). As shown in the deconvolution, the pattern arising from pure PiogHCl accounts for most of the integrated intensity, indicating that most of the PiogHCl does not disproportionate when the sample is stressed at elevated temperature and humidity (N.B., no quantitative assessment is made at this point, vide infra). The signal from disproportionation products can be isolated by examining the difference between the spectra, i.e., by subtracting the spectrum of the control sample from that of the stressed sample, as shown in Figure 2c. This signal is largely negative, because the amount of PiogHCl decreases in the disproportionation process. Decreasing the scale of the spectrum of NaSt_C to account for the decreased amount of PiogHCl before taking the difference produces a spectrum with entirely positive signal corresponding to the disproportionation product (Figure 2d). The result is a sharp feature, which indicates a site with a small quadrupolar interaction, such as an ionic chloride in an environment of high spherical (Platonic) symmetry.45,75,76 The narrow breadth of this pattern and its shift (0 ppm) are consistent with NaCl(s), which has previously been proposed as a by-product of the disproportionation reaction.10,66 No other peaks are observed in the 35Cl NMR spectrum, which suggests that the Cl– ions that separate from the API salt exclusively react with Na+ ions in the stearate to form NaCl, which is also confirmed by 23Na NMR and PXRD, vide infra. The magnitudes of the quadrupolar coupling constants, CQ, (and concomitantly their quadrupolar frequencies, νQ = CQ/[4I(2I – 1)2π]) are very different for Cl– ions in the API salt

11

ACS Paragon Plus Environment

Molecular Pharmaceutics 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

and the disproportionation product (νQ(35Cl)API salt >> νQ(35Cl)disp), giving distinct 35Cl nutation behaviour for the CT patterns corresponding to each environment.61,62 Such differences in nutation frequencies can be exploited by choosing an appropriate pulse sequence and applied rf amplitude, ν1. For example, using a Hahn echo sequence with non-selective excitation pulses (i.e., where νQ(35Cl)disp 0.999); using either measure of spikelet intensity, the measured salt contents are within ±1 wt-% of the expected value. The root mean square deviation (RMSD) and standard deviations (σ) of the measurements (Table 2) are larger when using ISI (primarily at low wt-% PiogHCl), due to low S/N of these spectra. Thus, MSI seems to be preferable for quantification of these patterns (especially for low salt contents), and is the method we used for all further quantification experiments (vide infra). 3.5. Quantification of Disproportionation Given the success of our procedure for quantifying PiogHCl in standard samples using 35

Cl SSNMR, we applied the same method to measure the amount of PiogHCl in the stearate-

containing sample mixtures. In order to avoid interfering signal from the disproportionation products, we analyzed spikelets between −40 and −65 kHz, which are outside of the range of signals from NaCl or MgCl2 (see the difference spectra, Figures 2c and 4d, respectively). The amounts of PiogHCl measured are listed in Table 3. As discussed above, the Cl− anions involved in disproportionation react with the cation of the stearate to produce either NaCl or MgCl2 (and its hydrates); the limiting reagent in the disproportionation reaction appears to be the stearate. With a 9:1 wt/wt ratio of API to stearate, the maximum stoichiometric limits of PiogHCl conversion are 14.77% and 14.25% in mixtures with MgSt and NaSt, respectively. The former value is within 0.5 %-conversion of what we have determined using 35Cl SSNMR, supporting our measured value. MgSt seems to cause a substantially larger loss of PiogHCl than NaSt (vide infra), in agreement with the previous observations.10,66 Interestingly, the values measured previously with other techniques10,16,66 range between 20-30 %-conversion for the MgSt mixture, which may indicate the presence of Cl− ions that are not bound to stearate cations (e.g., hexaaqua-coordinated Cl− species or evolution of HCl gas79–81), but we have not observed

17

ACS Paragon Plus Environment

Molecular Pharmaceutics 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

these species in our work. We note that 35Cl SSNMR can directly quantify the loss of the API HCl salt, providing an accurate measure of disproportionation; while the production of NaCl, MgCl2, and other by-products can also be quantified, this may not provide a precise measure of disproportionation. The amount of disproportionation measured in the NaSt-containing sample, 7(2) %-conversion, is roughly half what is predicted based on the stoichiometry of the mixture, indicating a partial conversion of the stearate. This result is supported by the 23Na NMR data (Figure 3b), which reveal that some of the NaSt remains in the NaSt_X sample after it has been stressed (i.e., some of the NaSt does not react with the API). As was previously noted,66 the decreased %-conversion of NaSt_X relative to MgSt_X is counterintuitive; NaSt has a higher surface area, alkalinity, and stronger hygroscopicity than MgSt, all of which are factors that are thought to increase disproportionation. However, as verified in the current work, the disproportionation in MgSt_X produces MgCl2, whose strong hygroscopicity is a driving force for the reaction. Such conclusions require further study, for which NMR should be helpful, given its ability to study both solid and solution products of disproportionation. 4. Conclusions In this work, we demonstrate the use of 35Cl SSNMR for qualitative and quantitative assessment of the disproportionation reaction of PiogHCl in model formulations with metallic stearates. Our SSNMR spectra provide conclusive proof of the reactions between the chloride anions in the API salt and the metal cations in the stearates to produce solid salts (i.e., NaCl or MgCl2; no hexaaqua-coordinated Cl− species are observed). With judicious choice of pulse sequences, we can selectively excite signal from the disproportionation products, and filter signal from the intact API salt. Such spectra provide rapid detection of disproportionation (i) with a

18

ACS Paragon Plus Environment

Page 18 of 34

Page 19 of 34 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

Molecular Pharmaceutics

much greater efficiency than 13C NMR and (ii) without interfering signal from excipient molecules; therefore, the acquisition of such spectra holds much promise for the high-throughput screening of formulations. We also present a novel method for quantification of ultra-wideline (UW) NMR spectra. We have illustrated how, with careful sample preparation and experimental setup, 35Cl SSNMR experiments can be used to accurately determine the amount of PiogHCl present in standardized binary mixtures containing the free base form (within 1 wt-% of the expected value, and at concentrations as low as 3 wt-%). To the best of our knowledge, this is the first demonstration of quantification using UW NMR spectra of any nucleus. This method has been used to quantitatively measure disproportionation in PiogHCl mixtures with metal stearates. By using spikelet spectra obtained from a Fourier transform of the entire CPMG echo train, our method can be used to selectively quantify peaks from the API salt without interference from the disproportionation products. Our results suggest that sample mixtures with NaSt and MgSt that have been stressed at 75% RH and 40 °C for 5 days produce disproportionation (%-conversion of PiogHCl) of roughly 7(2)% and 15(2)%, respectively. Quantification of API phase conversions in pharmaceutical development is critical to understand the risk of physical instability. Our results will aid in the mechanistic understanding of disproportionation, facilitating the risk assessment of API stabilities in the development of salt-containing pharmaceutical dosage formulations. Furthermore, with the novel approach for quantification of UW NMR spectra, we hope to encourage the use of quadrupolar nuclei and UW spectra in quantification studies of pharmaceuticals, as well as the wide variety of other important materials containing such nuclei.

19

ACS Paragon Plus Environment

Molecular Pharmaceutics 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

Acknowledgements R.W.S thanks the Natural Sciences and Engineering Research Council of Canada (NSERC) for funding this research (NSERC in the form of a Discovery Grant). R.W.S. is also grateful for a 50th Anniversary Golden Jubilee Chair from the University of Windsor. We are also grateful for the funding of the Laboratories for Solid-State Characterization at the University of Windsor from the Canadian Foundation for Innovation, the Ontario Innovation Trust, and the University of Windsor. The authors would like to thank Dr. John Christopher (MRLs, Merck) for fruitful scientific discussions.

Supporting Information The following additional information is detailed in the supporting information: additional experimental parameters for NMR experiments, SSNMR spectra, PXRD data, and detailed analysis of the processing of the 35Cl SSNMR spectra used for quantification.

20

ACS Paragon Plus Environment

Page 20 of 34

Page 21 of 34 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

Molecular Pharmaceutics

References (1) (2) (3) (4) (5) (6) (7)

(8) (9) (10)

(11)

(12)

(13)

(14) (15)

(16)

(17)

(18)

Berge, S. M.; Bighley, L. D.; Monkhouse, D. C. Pharmaceutical salts. J. Pharm. Sci. 1977, 66, 1–19. Elder, D. P.; Holm, R.; Lopez de Diego, H. Use of pharmaceutical salts and cocrystals to address the issue of poor solubility. Int. J. Pharm. 2013, 453, 88–100. Nelson, E. Comparative dissolution rates of weak acids and their sodium salts. J. Am. Pharm. Assoc. Am. Pharm. Assoc. (Baltim). 1958, 47, 297–299. Serajuddin, A. T. M. Salt formation to improve drug solubility. Adv. Drug Deliv. Rev. 2007, 59, 603–616. Guerrieri, P.; Taylor, L. S. Role of salt and excipient properties on disproportionation in the solid-state. Pharm. Res. 2009, 26, 2015–2026. Thakral, N. K.; Kelly, R. C. Salt disproportionation: A material science perspective. Int. J. Pharm. 2017, 520, 228–240. Thakral, N. K.; Behme, R. J.; Aburub, A.; Peterson, J. A.; Woods, T. A.; Diseroad, B. A.; Suryanarayanan, R.; Stephenson, G. A. Salt Disproportionation in the Solid State: Role of Solubility and Counterion Volatility. Mol. Pharm. 2016, 13, 4141–4151. Nie, H.; Byrn, S. R.; Zhou, Q. Stability of pharmaceutical salts in solid oral dosage forms. Drug Dev. Ind. Pharm. 2017, 43, 1215–1228. Stephenson, G. A.; Aburub, A.; Woods, T. A. Physical Stability of Salts of Weak Bases in the Solid-State. J. Pharm. Sci. 2011, 100, 1607–1617. John, C. T.; Xu, W.; Lupton, L. K.; Harmon, P. A. Formulating Weakly Basic HCl Salts: Relative Ability of Common Excipients to Induce Disproportionation and the Unique Deleterious Effects of Magnesium Stearate. Pharm. Res. 2013, 30, 1628–1641. Hsieh, Y.-L.; Taylor, L. S. Salt Stability - Effect of Particle Size, Relative Humidity, Temperature and Composition on Salt to Free Base Conversion. Pharm. Res. 2015, 32, 549–561. Serajuddin, A. T. M.; Rosoff, M. pH-Solubility Profile of Papaverine Hydrochloride and Its Relationship to the Dissolution Rate of Sustained-Release Pellets. J. Pharm. Sci. 1984, 73, 1203–1208. Serajuddin, A. T. M.; Jarowski, C. I. Effect of Diffusion Layer pH and Solubility on the Dissolution Rate of Pharmaceutical Acids and Their Sodium Salts II: Salicylic Acid, Theophylline, and Benzoic Acid. J. Pharm. Sci. 1985, 74, 148–154. Hsieh, Y.-L.; Merritt, J. M.; Yu, W.; Taylor, L. S. Salt Stability – The Effect of pHmax on Salt to Free Base Conversion. Pharm. Res. 2015, 32, 3110–3118. 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. Koranne, S.; Govindarajan, R.; Suryanarayanan, R. Investigation of Spatial Heterogeneity of Salt Disproportionation in Tablets by Synchrotron X-ray Diffractometry. Mol. Pharm. 2017, 14, 1133–1144. Skrdla, P. J.; Zhang, D. Disproportionation of a crystalline citrate salt of a developmental pharmaceutical compound: characterization of the kinetics using pH monitoring and online Raman spectroscopy plus quantitation of the crystalline free base form in binary physical mixtures. J. Pharm. Biomed. Anal. 2014, 90, 186–191. Namespetra, A. M.; Hirsh, D. A.; Hildebrand, M. P.; Sandre, A. R.; Hamaed, H.; Rawson,

21

ACS Paragon Plus Environment

Molecular Pharmaceutics 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

(19) (20) (21) (22) (23) (24) (25) (26)

(27) (28)

(29)

(30)

(31)

(32)

(33)

(34) (35)

J. M.; Schurko, R. W. 35 Cl solid-state NMR spectroscopy of HCl pharmaceuticals and their polymorphs in bulk and dosage forms. CrystEngComm 2016, 18, 6213–6232. Tishmack, P. A.; Bugay, D. E.; Byrn, S. R. Solid-State Nuclear Magnetic Resonance Spectroscopy-Pharmaceutical Applications. J. Pharm. Sci. 2003, 92, 441–474. Geppi, M.; Mollica, G.; Borsacchi, S.; Veracini, C. A. Solid‐State NMR Studies of Pharmaceutical Systems. Appl. Spectrosc. Rev. 2008, 43, 202–302. Vogt, F. G. Characterization of Pharmaceutical Compounds by Solid-state NMR. eMagRes 2015, 4, 255–268. Harris, R. K. NMR studies of organic polymorphs & solvates. Analyst 2006, 131, 351–373. Holzgrabe, U.; Deubner, R.; Schollmayer, C.; Waibel, B. Quantitative NMR spectroscopy—Applications in drug analysis. J. Pharm. Biomed. Anal. 2005, 38, 806–812. Stephenson, G. A.; Forbes, R. A.; Reutzel-Edens, S. M. Characterization of the solid state: Quantitative issues. Adv. Drug Deliv. Rev. 2001, 48, 67–90. Harris, R. K. Quantitative aspects of high-resolution solid-state nuclear magnetic resonance spectroscopy. Analyst 1985, 110, 649–655. 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. Ziarelli, F.; Caldarelli, S. Solid-state NMR as an analytical tool: Quantitative aspects. Solid State Nucl. Magn. Reson. 2006, 29, 214–218. Offerdahl, T. J.; Salsbury, J. S.; Dong, Z.; Grant, D. J. W.; Schroeder, S. A.; Prakash, I.; Gorman, E. M.; Barich, D. H.; Munson, E. J. Quantitation of crystalline and amorphous forms of anhydrous neotame using 13C CPMAS NMR spectroscopy. J. Pharm. Sci. 2005, 94, 2591–2605. Liu, J.; Nagapudi, K.; Kiang, Y.-H. H.; Martinez, E.; Jona, J. Quantification of compaction-induced crystallinity reduction of a pharmaceutical solid using 19 F solidstate NMR and powder X-ray diffraction. Drug Dev. Ind. Pharm. 2009, 35, 969–975. Lefort, R.; De Gusseme, A.; Willart, J.-F.; Danède, F.; Descamps, M. Solid state NMR and DSC methods for quantifying the amorphous content in solid dosage forms: an application to ball-milling of trehalose. Int. J. Pharm. 2004, 280, 209–219. 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. Correa-Soto, C.; Trasi, N. S.; Schmitt, P. D.; Su, Y.; Liu, Z.; Miller, E.; Variankaval, N.; Marsac, P. J.; Simpson, G. J.; Taylor, L. S. Second harmonic generation microscopy as a tool for the early detection of crystallization in spray dried dispersions. J. Pharm. Biomed. Anal. 2017, 146, 86–95. Suryanarayanan, R.; Wiedmann, T. S. Quantitation of the Relative Amounts of Anhydrous Carbamazepine (C15H12N2O) and Carbamazepine Dihydrate (C15H12N2O · 2H2O) in a Mixture by Solid-State Nuclear Magnetic Resonance (NMR). Pharm. Res. 1990, 7, 184– 187. Gao, P. Determination of the composition of delavirdine mesylate polymorph and pseudopolymorph mixtures using 13C CP/MAS NMR. Pharm. Res. 1996, 13, 1095–1104. Tozuka, Y.; Ito, A.; Seki, H.; Oguchi, T.; Yamamoto, K. Characterization and

22

ACS Paragon Plus Environment

Page 22 of 34

Page 23 of 34 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

Molecular Pharmaceutics

(36) (37)

(38)

(39)

(40)

(41) (42) (43)

(44)

(45) (46)

(47)

(48) (49)

(50) (51) (52)

Quantitation of Clarithromycin Polymorphs by Powder X-Ray Diffractometry and SolidState NMR Spectroscopy. Chem. Pharm. Bull. (Tokyo). 2002, 50, 1128–1130. Vickery, R. D.; Nemeth, G. A.; Maurin, M. B. Solid-state carbon NMR characterization of the polymorphs of roxifiban. J. Pharm. Biomed. Anal. 2002, 30, 125–129. Stueber, D.; Jehle, S. Quantitative Component Analysis of Solid Mixtures by Analyzing Time Domain 1 H and 19 F T 1 Saturation Recovery Curves (qSRC). J. Pharm. Sci. 2017, 106, 1828–1838. Tinmanee, R.; Larsen, S. C.; Morris, K. R.; Kirsch, L. E. Quantification of gabapentin polymorphs in gabapentin/excipient mixtures using solid state 13 C NMR spectroscopy and X-ray powder diffraction. J. Pharm. Biomed. Anal. 2017, 146, 29–36. Rohrs, B. R.; Thamann, T. J.; Gao, P.; Stelzer, D. J.; Bergren, M. S.; Chao, R. S. Tablet dissolution affected by a moisture mediated solid-state interaction between drug and disintegrant. Pharm. Res. 1999, 16, 1850–1856. Merritt, J. M.; Viswanath, S. K.; Stephenson, G. A. Implementing Quality by Design in Pharmaceutical Salt Selection: A Modeling Approach to Understanding Disproportionation. Pharm. Res. 2013, 30, 203–217. Ashbrook, S. E.; Duer, M. J. Structural information from quadrupolar nuclei in solid state NMR. Concepts Magn. Reson. Part A 2006, 28A, 183–248. Bryce, D. L.; Sward, G. D. Solid-state NMR spectroscopy of the quadrupolar halogens: chlorine-35/37, bromine-79/81, and iodine-127. Magn. Reson. Chem. 2006, 44, 409–450. Hamaed, H.; Pawlowski, J. M.; Cooper, B. F. T.; Fu, R.; Eichhorn, S. H.; Schurko, R. W. Application of Solid-State 35 Cl NMR to the Structural Characterization of Hydrochloride Pharmaceuticals and their Polymorphs. J. Am. Chem. Soc. 2008, 130, 11056–11065. Hildebrand, M.; Hamaed, H.; Namespetra, A. M.; Donohue, J. M.; Fu, R.; Hung, I.; Gan, Z.; Schurko, R. W. 35 Cl solid-state NMR of HCl salts of active pharmaceutical ingredients: structural prediction, spectral fingerprinting and polymorph recognition. CrystEngComm 2014, 16, 7334. Widdifield, C. M.; Chapman, R. P.; Bryce, D. L. In Annual Reports on NMR Spectroscopy; Webb, G. A., Ed.; Elsevier, 2009; Vol. 66, pp 195–326. Bryce, D. L.; Gee, M.; Wasylishen, R. E. High-Field Chlorine NMR Spectroscopy of Solid Organic Hydrochloride Salts: A Sensitive Probe of Hydrogen Bonding Environment. J. Phys. Chem. A 2001, 105, 10413–10421. Bryce, D. L.; Bultz, E. B. Alkaline Earth Chloride Hydrates: Chlorine Quadrupolar and Chemical Shift Tensors by Solid-State NMR Spectroscopy and Plane Wave Pseudopotential Calculations. Chem. Eur. J. 2007, 13, 4786–4796. Schurko, R. W. Ultra-Wideline Solid-State NMR Spectroscopy. Acc. Chem. Res. 2013, 46, 1985–1995. Perras, F. A.; Viger-Gravel, J.; Burgess, K. M. N.; Bryce, D. L. Signal enhancement in solid-state NMR of quadrupolar nuclei. Solid State Nucl. Magn. Reson. 2013, 51–52, 1– 15. Ashbrook, S. E.; Sneddon, S. New Methods and Applications in Solid-State NMR Spectroscopy of Quadrupolar Nuclei. J. Am. Chem. Soc. 2014, 136, 1–60. Schurko, R. W. In Encyclopedia of Magnetic Resonance; John Wiley & Sons, Ltd: Chichester, UK, 2011; pp 77–93. Carr, H. Y.; Purcell, E. M. Effects of Diffusion on Free Precession in Nuclear Magnetic Resonance Experiments. Phys. Rev. 1954, 94, 630–638.

23

ACS Paragon Plus Environment

Molecular Pharmaceutics 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

(53) (54)

(55) (56) (57) (58) (59)

(60)

(61)

(62)

(63) (64)

(65) (66)

(67)

(68)

(69)

(70)

Meiboom, S.; Gill, D. Modified Spin‐Echo Method for Measuring Nuclear Relaxation Times. Rev. Sci. Instrum. 1958, 29, 688–691. Larsen, F. H.; Farnan, I. 29Si and 17O (Q)CPMG-MAS solid-state NMR experiments as an optimum approach for half-integer nuclei having long T1 relaxation times. Chem. Phys. Lett. 2002, 357, 403–408. Hung, I.; Gan, Z. On the practical aspects of recording wideline QCPMG NMR spectra. J. Magn. Reson. 2010, 204, 256–265. Kupce, E.; Freeman, R. Adiabatic Pulses for Wideband Inversion and Broadband Decoupling. J. Magn. Reson. 1995, 115, 273–276. Bhattacharyya, R.; Frydman, L. Quadrupolar nuclear magnetic resonance spectroscopy in solids using frequency-swept echoing pulses. J. Chem. Phys. 2007, 127, 194503. O’Dell, L. A.; Schurko, R. W. QCPMG using adiabatic pulses for faster acquisition of ultra-wideline NMR spectra. Chem. Phys. Lett. 2008, 464, 97–102. O’Dell, L. A.; Rossini, A. J.; Schurko, R. W. Acquisition of ultra-wideline NMR spectra from quadrupolar nuclei by frequency stepped WURST-QCPMG. Chem. Phys. Lett. 2009, 468, 330–335. Ziarelli, F.; Viel, S.; Sanchez, S.; Cross, D.; Caldarelli, S. Precision and sensitivity optimization of quantitative measurements in solid state NMR. J. Magn. Reson. 2007, 188, 260–266. Samoson, A.; Lippmaa, E. Excitation phenomena and line intensities in high-resolution NMR powder spectra of half-integer quadrupolar nuclei. Phys. Rev. B 1983, 28, 6567– 6570. Kentgens, A. P. M.; Lemmens, J. J. M.; Geurts, F. M. M.; Veeman, W. S. Twodimensional solid-state nutation NMR of half-integer quadrupolar nuclei. J. Magn. Reson. 1987, 71, 62–74. Vega, A. J. In Encyclopedia of Magnetic Resonance; Grant, D. M., Harris, R. K., Eds.; John Wiley & Sons, Ltd: Chichester, 1996; Vol. 4, pp 3869–3888. Man, P. P.; Klinowski, J.; Trokiner, A.; Zanni, H.; Papon, P. Selective and non-selective NMR excitation of quadrupolar nuclei in the solid state. Chem. Phys. Lett. 1988, 151, 143–150. O’Dell, L. A. The WURST kind of pulses in solid-state NMR. Solid State Nucl. Magn. Reson. 2013, 55–56, 28–41. Nie, H.; Xu, W.; Ren, J.; Taylor, L. S.; Marsac, P. J.; John, C. T.; Byrn, S. R. Impact of Metallic Stearates on Disproportionation of Hydrochloride Salts of Weak Bases in SolidState Formulations. Mol. Pharm. 2016, 13, 3541–3552. Nie, H.; Xu, W.; Taylor, L. S.; Marsac, P. J.; Byrn, S. R. Crystalline solid dispersion-a strategy to slowdown salt disproportionation in solid state formulations during storage and wet granulation. Int. J. Pharm. 2017, 517, 203–215. Nie, H.; Su, Y.; Zhang, M.; Song, Y.; Leone, A.; Taylor, L. S.; Marsac, P. J.; Li, T.; Byrn, S. R. Solid-State Spectroscopic Investigation of Molecular Interactions between Clofazimine and Hypromellose Phthalate in Amorphous Solid Dispersions. Mol. Pharm. 2016, 13, 3964–3975. Peersen, O.; Wu, X.; Smith, S. Enhancement of CP-MAS Signals by Variable-Amplitude Cross Polarization. Compensation for Inhomogeneous B1 Fields. J. Magn. Reson. Ser. A 1994, 106, 127–131. Peersen, O. B.; Wu, X. L.; Kustanovich, I.; Smith, S. O. Variable-Amplitude Cross-

24

ACS Paragon Plus Environment

Page 24 of 34

Page 25 of 34 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

Molecular Pharmaceutics

(71) (72) (73)

(74)

(75) (76)

(77)

(78)

(79)

(80)

(81)

Polarization MAS NMR. J. Magn. Reson. Ser. A 1993, 104, 334–339. Metz, G.; Wu, X. L.; Smith, S. O. Ramped-Amplitude Cross Polarization in Magic-AngleSpinning NMR. J. Magn. Reson. Ser. A 1994, 110, 219–227. Bennett, A. E.; Rienstra, C. M.; Auger, M.; Lakshmi, K. V.; Griffin, R. G. Heteronuclear decoupling in rotating solids. J. Chem. Phys. 1995, 103, 6951–6958. Thakur, R. S.; Kurur, N. D.; Madhu, P. K. K. An analysis of phase-modulated heteronuclear dipolar decoupling sequences in solid-state nuclear magnetic resonance. J. Magn. Reson. 2008, 193, 77–88. Lefort, R.; Wiench, J. W.; Pruski, M.; Amoureux, J.-P. Optimization of data acquisition and processing in Carr–Purcell–Meiboom–Gill multiple quantum magic angle spinning nuclear magnetic resonance. J. Chem. Phys. 2002, 116, 2493. Akitt, J. W.; McDonald, W. S. Arrangements of ligands giving low electric field gradients. J. Magn. Reson. 1984, 58, 401–412. Autschbach, J.; Zheng, S.; Schurko, R. W. Analysis of electric field gradient tensors at quadrupolar nuclei in common structural motifs. Concepts Magn. Reson. Part A 2010, 36A, 84–126. Blaakmeer, E. S. (Merijn); Antinucci, G.; Busico, V.; van Eck, E. R. H.; Kentgens, A. P. M. Solid-State NMR Investigations of MgCl 2 Catalyst Support. J. Phys. Chem. C 2016, 120, 6063–6074. Bryce, D. L.; Bultz, E. B. Alkaline Earth Chloride Hydrates: Chlorine Quadrupolar and Chemical Shift Tensors by Solid-State NMR Spectroscopy and Plane Wave Pseudopotential Calculations. Chem. Eur. J. 2007, 13, 4786–4796. Elder, D. P.; Delaney, E.; Teasdale, A.; Eyley, S.; Reif, V. D.; Jacq, K.; Facchine, K. L.; Oestrich, R. S.; Sandra, P.; David, F. The Utility of Sulfonate Salts in Drug Development. J. Pharm. Sci. 2010, 99, 2948–2961. Lin, S.-L.; Lachman, L.; Swartz, C. J.; Huebner, C. F. Preformulation Investigation I: Relation of Salt Forms and Biological Activity of an Experimental Antihypertensive. J. Pharm. Sci. 1972, 61, 1418–1422. Thakral, N. K.; Behme, R. J.; Aburub, A.; Peterson, J. A.; Woods, T. A.; Diseroad, B. A.; Suryanarayanan, R.; Stephenson, G. A. Salt Disproportionation in the Solid State: Role of Solubility and Counterion Volatility. Mol. Pharm. 2016, 13, 4141–4151.

25

ACS Paragon Plus Environment

Molecular Pharmaceutics 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

For Table of Contents Use Only

Quantifying Disproportionation in Pharmaceutical Formulations with 35Cl Solid-State NMR David A. Hirsh, Yongchao Su, Haichen Nie, Wei Xu, Dirk Stueber, Narayan Variankaval, and Robert W. Schurko

26

ACS Paragon Plus Environment

Page 26 of 34

Page 27 of 34 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

Molecular Pharmaceutics

Table 1. Sample masses used to prepare mixed samples for quantification with 35Cl SSNMR. Amount of compound added [/mg]b

PiogHCl 100% PiogHCl 90% PiogHCl 60% PiogHCl 30% PiogHCl 10% PiogHCl 5% PiogHCl 3%

100% 90.0% 60.3% 29.9% 10.1% 5.2% 3.0%

300.0 271.1 181.8 90.1 30.2 15.7 8.9

– 30.1 119.9 210.8 270 284.9 291.3

– – – – – – –

– – – – – – –

Packed Sample Weight [/mg]b,c 137.2 139.6 151.2 146.0 145.8 150.8 143.5

MgSt_C MgSt_X (Stressed at 40 °C, 75% RH) MgSt_XH (rehydrated for 4 hours at 75% RH)d

90.1%

273.0



30



191.2

90.0%

270.0



30.0



136.3

90.0%

270.0



30.0



148.8

90.1%

272.0





30.0

146.0

90.0%

270.0





30.0

132.6

Sample Name

NaSt_C NaSt_X (Stressed at 40 °C, 75% RH)

Wt-% Salta

PiogHCl PiogFB MgSt

NaSt

Wt-% PiogHCl salt calculated from the masses of the materials in the sample mixture. bAll reported masses are accurate to within 0.1 mg. cMass of the packed sample used for 35Cl SSNMR experiments. dThe MgSt_XH sample was prepared by placing 300 mg of the MgSt_X sample in a 75% RH humidity chamber for 4 hours immediately before packing the sample. a

1 ACS Paragon Plus Environment

Molecular Pharmaceutics 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 28 of 34

Table 2. Experimental and calculated amounts of PiogHCl in the standard samples. Wt-% Measurement a methodb PiogHCl

Sample

Measured wt-% PiogHClc

Standard RMSD deviation, σ [/wt-%]d [/wt-%]e

PiogHCl 90%

90.0%

MSI ISI

90.4% 90.2%

1.81% 1.95%

1.81% 1.98%

PiogHCl 60%

60.3%

MSI ISI

59.4% 59.4%

2.42% 2.80%

2.06% 2.71%

PiogHCl 30%

29.9%

MSI ISI

30.0% 30.0%

3.15% 3.98%

3.22% 4.05%

PiogHCl 10%

10.1%

MSI ISI

11.0% 11.4%

1.34% 2.51%

0.95% 2.14%

PiogHCl 5%

5.2%

MSI ISI

5.8% 5.8%

0.81% 4.25%

0.50% 0.52%

PiogHCl 3%

3.0%

MSI ISI

3.9% 3.9%

1.15% 6.22%

1.74% 1.13%

Calculated from the masses of PiogHCl and PiogFB used in the sample mixture, see Table 1. Denotes the method used to measure spikelet intensity, either: (i) maximum spikelet intensity (MSI) or integrated spikelet intensity (ISI). cMeasured values from 35Cl SSNMR experiments, reported as the average value obtained from spikelets between 39 kHz and −84 kHz, see text for details. dRoot mean square deviation, calculated using the formula: where N = number of spikelets, xˆ = the wt-% PiogHCl measured by 2 1 N RMSD = ∑ xˆ − x i sample mass, and xi = the measured wt-% PiogHCl of spikelet i. N i =1

a

b

(

)

Standard deviation, σ, calculated using the formula: where N = number of spikelets, x = the average measured wt-% N 2 1 σ= x − x PiogHCl, and xi = the measured wt-% PiogHCl of spikelet i. ∑ N −1 i =1 i

e

(

) (

)

2 ACS Paragon Plus Environment

Page 29 of 34 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

Molecular Pharmaceutics

Table 3. Experimental and calculated amounts of PiogHCl in the stearate mixture samples. Measured Theoretical Measured % Conversion % Conversion of Sample wt-% PiogHCl of PiogHCl PiogHCl [/%]a b [/%] [/%]c NaSt_C NaSt_X

93(2) 84(2)

– 7(2)

– 14.25

MgSt_C MgSt_X

92(3) 77(2)

– 15(2)

– 14.77

Measured values from 35Cl SSNMR experiments, reported as the average value obtained from spikelets between −40 and −65 kHz with the standard deviation listed in parentheses, see text for details. bCalculated using the formula: amount of PiogHCl consumed 90%−measured wt-% PiogHCl % conversion = = amount of PiogHCl supplied 90%

a

(

)

Conversion of PiogHCl that would occur if all of the stearate cations reacted stoichiometrically with Cl− anions from the salt. c

a) 2 PiogHCl + (R-COO–)2Mg2+ Pioglitazone HCl

b)

Magnesium stearate (MgSt)

PiogHCl + (R-COO–)Na+

Pioglitazone HCl

75% RH, 40 °C

2 Piog + 2 (R-COOH) + MgCl2 Pioglitazone Free base

75% RH, 40 °C

Sodium stearate (NaSt)

Stearic acid

Magnesium chloride

Piog + (R-COOH) + NaCl

Pioglitazone Free base

Stearic acid

Sodium chloride

Scheme 1. Proposed disproportion reactions of PiogHCl with a) MgSt and b) NaSt.

3 ACS Paragon Plus Environment

Molecular Pharmaceutics 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

c) Simulated spectrum

b) Experimental spectrum (coadded echo train) a) Experimental spectrum (FT echo train) 200 6000

100 4000

0

–100

2000 0 –2000 35 Cl Chemical Shift

–200 [kHz] –4000

[ppm]

Figure 1. Experimental 35Cl{1H} WURST-CPMG NMR spectra of pure PiogHCl under static sample conditions, a) and b), and the corresponding analytical simulation, c). Experimental data were processed by a) applying a Fourier transform (FT) to the entire echo train (blue), or b) coadding the echoes before performing a FT (red), see details in the text.

d) Difference: b – (0.7912)a

×2.6346

c) Difference: b – a

×1.5665

b) NaSt_X a) NaSt_C 200 6000

100 4000

0

–100

2000 0 –2000 35 Cl Chemical Shift

–200 [kHz] –4000

[ppm]

Figure 2. 35Cl{1H} WURST-CPMG NMR spectra of PiogHCl samples mixed with NaSt (9:1 wt/wt), a) is the spectrum of the control sample (NaSt_C), and b) is the spectrum of the sample stressed at 40 °C/75% RH for 5 days (NaSt_X). The difference between the spectra (control – stressed) is shown in c) and d). When calculating the latter, the intensity of the spectrum of the control (a) was decreased to produce only positive signal (see text for details).

4 ACS Paragon Plus Environment

Page 30 of 34

Page 31 of 34 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

Molecular Pharmaceutics

c) NaSt

×0.10

×7

b) NaSt_X

0

−5 −10 −15 −20

a) NaSt_C 30

20

10 0 −10 −20 Na Chemical Shift (ppm)

23

−30

Figure 3. 23Na{1H} MAS (νrot = 12 kHz) NMR spectra of a) NaSt_C, b) NaSt_X, and c) pure NaSt. The inset shows the NaSt region of the spectrum of NaSt_X with increased intensity.

e) Difference: b – (0.6139)a

×2.3784

d) Difference: b – a

×1.4348

c) MgSt_XH (rehydrated) b) MgSt_X a) MgSt_C 200 6000

100 4000

0

–100

2000 0 –2000 35 Cl Chemical Shift

–200 [kHz] –4000

[ppm]

Figure 4. 35Cl{1H} WURST-CPMG NMR spectra of PiogHCl samples mixed with MgSt (9:1 wt/wt), a) the control sample (MgSt_C), b) the sample stressed at 40 °C/75% RH for 5 days (MgSt_X), and c) the stressed sample immediately after being rehydrated at 20 °C/75% RH for 4 hours (MgSt_XH). The difference between the spectra (control – stressed) is shown in d) and e). When calculating the latter, the intensity of the spectrum of the control (a) was decreased to produce only positive signal.

5 ACS Paragon Plus Environment

Molecular Pharmaceutics 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

100% PiogHCl

Page 32 of 34

90% PiogHCl

60% PiogHCl

30% PiogHCl

10% PiogHCl 5% PiogHCl

50 1000

0

−50 [kHz]

0

−1000 [ppm]

50 1000

0 0

−50 [kHz] 50 −1000 [ppm]

1000

0

−50 [kHz]

0

−1000 [ppm]

50 1000 35

0 0

−50 [kHz] 50 −1000 [ppm]

1000

0

−50 [kHz]

0

−1000 [ppm]

50 1000

0

−50 [kHz]

0

−1000 [ppm]

3% PiogHCl

50 1000

0

−50 [kHz]

0

−1000 [ppm]

Cl Chemical Shift

Figure 5. 35Cl{1H} WURST-CPMG NMR spectra of the PiogHCl/PiogFB mixtures. The data are shown after processing by either (i) applying a FT to the entire echo train (“spikelet spectrum”) or (ii) combining the echoes together before performing the FT (“echo co-addition”).

6 ACS Paragon Plus Environment

Page 33 of 34 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

Molecular Pharmaceutics

Sample Packing Center sample in NMR coil FT

Ultra-wideline (UW) 35 Cl solid-state NMR data acquisition

Frequency

Time

Spikelet selection

Maximum spikelet intensity (MSI)

or

Integrated spikelet intensity (ISI)

MSI

ISI

Scale data based on sample mass

Normalize to reference sample

Figure 6. Flow chart detailing the quantification protocol used to analyze APIs with UW 35Cl SSNMR spectra.

7 ACS Paragon Plus Environment

Molecular Pharmaceutics

a) Maximum Spikelet Intensity (MSI)

b) Integrated Spikelet Intensity (ISI)

80%

100% Experimental Ideal

60% 40% R2 = 0.9998

20% 0%

0%

20%

40%

60%

80%

wt-% PiogHCl from SSNMR

100% wt-% PiogHCl from SSNMR

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 34 of 34

100%

80%

Experimental Ideal

60% 40% R2 = 0.9997

20% 0%

0%

wt-% PiogHCl from Sample Mass

20%

40%

60%

80%

100%

wt-% PiogHCl from Sample Mass

Figure 7. Plots of PiogHCl wt-% measured experimentally using 35Cl SSNMR (y-axis) and the wt-% determined from the salt : free base mass ratio of the sample (x-axis). Data were obtained from spikelets in the spectra by measuring a) the peak intensity or b) the integrated intensity of each spikelet. The plotted values are the average measurement obtained from the spikelets in a given spectrum between 39 and − 84 kHz, and the error bars represent the root mean square deviation (RMSD). Standard deviations, σ, in the measured values (not shown) are smaller than the point markers. (See Table 2 for definitions of these two statistical values).

8 ACS Paragon Plus Environment