“CLASSIC NMR”: An In-Situ NMR Strategy for Mapping the Time-Evolution of Crystallization Processes by Combined Liquid-State and Solid-State Measurements

A new in-situ NMR strategy (termed CLASSIC NMR) for mapping the evolution of crystallization processes is reported, involving simultaneous measurement of both liquid-state and solid-state NMR spectra as a function of time. This combined strategy allows complementary information to be obtained on the evolution of both the solid and liquid phases during the crystallization process. In particular, as crystallization proceeds (monitored by solid-state NMR), the solution state becomes more dilute, leading to changes in solution-state speciation and the modes of molecular aggregation in solution, which are monitored by liquid-state NMR. The CLASSIC NMR experiment is applied here to yield new insights into the crystallization of m-aminobenzoic acid.

measurement of the T 1 ( 13 C) values for the super-saturated solutions that exist during the in-situ study of the crystallization process are not straightforward, given the meta-stable nature of these solutions and the fact that the T 1 ( 13 C) values are expected to vary as a function of time during the crystallization process as a consequence of changes in the solution-state concentration and speciation.
The temperature of the sample was estimated from a lead nitrate calibration, [3] with MAS at 12 kHz found to correspond to a temperature increase of 13 °C. This calibration was corroborated by recording 1 H NMR spectra for methanol, for which the chemical shift difference between the two 1 H resonances is temperature dependent [4] (while this calibration method was exploited previously for solid-state measurements by soaking the methanol into TTMSS, [5] neat methanol was used inside a sealed rotor in the present case).
Figure S1 Direct-excitation (liquid-state) 13 C NMR spectra recorded in our CLASSIC NMR study of crystallization of m-ABA from DMSO. The first liquid-state 13 C NMR spectrum is shown in (a) and the last liquid-state 13 C NMR spectrum (recorded 14.8 hrs after commencing the experiment) is shown in (b). From comparison of (a) and (b), it is clear that the 13 C NMR spectrum shown in (b) does not contain any contribution from the solid phase of m-ABA that was present at this stage of the crystallization experiment.

Analysis of CLASSIC NMR Data, Including Methodology for Fitting the Chemical Shift Data
In the 13 C NMR data recorded in the CLASSIC NMR experiment, there is a general drift of peaks over time. The magnitude of the drift decreases approximately exponentially with time, converging on a final peak position at lower ppm than the initial peak position. In addition, there is a secondary effect, starting after approximately 2 hrs, which increases or decreases the drift away from a single exponential. The overall drift is attributed to the fact that the magnetic shims take several hours to cool, thus altering the field generated over this period of time. The secondary effect is assigned to changes in concentration due to crystallization, as it begins at the same time that the first peaks are observed in the solid-state ( 13 C CPMAS) spectra. The contribution due to the magnetic shims is removed using a fit to a bi-exponential model.
To remove the effect of the field drift due to slow cooling of the magnetic shims during the crystallization experiment, the changes in isotropic peak positions were fitted to a bi-exponential model with the function: where the term xe -t/τ 1 models the field drift due to the shims and the term y i e -(t-t c )/τ 2,i models the concentration effect. The values of x,  1 and t c were the same for all eight peaks but the values of y i and  2,i were allowed to vary independently for each peak position (denoted i). The term t c is the time at which concentration effects begin. The result of this fit (with 19 variables) is an exponential curve given by xe -t/τ 1 (x = -0.46 ppm,  1 = 2.20 hr) which may be subtracted from the plots shown in Figure S2 to reveal the nature of the concentration effects, as shown in Figure S3

Solid-State 13 C NMR Studies of the Five Polymorphs of m-ABA
To characterize the solid-state 13 C NMR properties of the five known polymorphs of m-ABA prior to our CLASSIC NMR studies, each polymorph was prepared using the procedures reported previously. [6] The high-resolution solid-state 13 C NMR spectrum was recorded for a powder sample of each polymorph using a Chemagnetics Infinity Plus spectrometer [ 1 H Larmor frequency, 300.2 MHz (75.48 MHz for 13 C); 4 mm zirconia rotor; MAS frequency 12 kHz; ramped 1 H→ 13 C CP [1] with contact time, 2 ms; TPPM 1 H decoupling [7] ]. Different recycle delays were used for each polymorph and were at least five times T 1 ( 1 H). The value of T 1 ( 1 H) for each polymorph was determined by applying inversion recovery on the 1 H channel followed by 1 H→ 13 C CP and acquisition of the 13 C NMR spectrum.

Liquid-State 13 C NMR Studies of Saturated Solutions of m-ABA
To characterize the liquid-state 13 C NMR properties of m-ABA in different solvents, solutionstate 13 C NMR spectra were recorded at 33 °C for saturated solutions prepared by dissolution of Form III of m-ABA in water (6.5 mg/g), methanol (71.0 mg/g), DMSO (745.25 mg/g) and 1,4-dioxane (62.6 mg/g). In all cases, the data were recorded on a 500 MHz Bruker solution-state NMR spectrometer. The solution was contained in a 5 mm glass tube and was deuterium locked using D 2 O in a sealed glass insert.

In-Situ Solid-State 13 C NMR Study of Crystallization of m-ABA from Methanol
Our  Figure S4), the MAS frequency was 12 kHz. The rotor was heated to 85 °C for one hour to ensure complete dissolution. The rotor was then cooled to 33 °C and high-resolution solid-state 13 C NMR spectra were acquired repeatedly, using ramped 1 H→ 13 C CP (CP contact time, 1.5 ms) with 1 H decoupling using SPINAL-64 (nutation frequency, 91 kHz). Each spectrum involved the acquisition of 256 scans, with a recycle delay of 9 s. The total time to record each spectrum was 38.4 mins, representing the time resolution of the in-situ study. Figure S4 In-situ solid-state 13 C NMR spectra (showing only the region for the carboxylate/carboxylic acid group) recorded as a function of time during crystallization of m-ABA from methanol. The initial crystallization product is identified as Form I (black dashed line), which subsequently undergoes a polymorphic transformation to produce Form III (green dashed line). After ca. 9 hrs, the crystallization product is a monophasic sample of Form III.