Desolvation Processes in Channel Solvates of Niclosamide

The antiparasitic drug niclosamide (NCL) is notable for its ability to crystallize in multiple 1:1 channel solvate forms, none of which are isostructural. Here, using a combination of time-resolved synchrotron powder X-ray diffraction and thermogravimetry, the process-induced desolvation mechanisms of methanol and acetonitrile solvates are investigated. Structural changes in both solvates follow a complicated molecular-level trajectory characterized by a sudden shift in lattice parameters several degrees below the temperature where the desolvated phase first appears. Model fitting of kinetic data obtained under isothermal heating conditions suggests that the desolvation is rate-limited by the nucleation of the solvent-free product. The desolvation pathways identified in these systems stand in contrast to previous investigations of the NCL channel hydrate, where water loss by diffusion initially yields an anhydrous isomorph that converts to the thermodynamic polymorph at significantly higher temperatures. Taking the view that each solvate lattice is a unique “pre-organized” precursor, a comparison of the pathways from different starting topologies to the same final product provides the opportunity to reevaluate assumptions of how various factors (e.g., solvent binding strength, density) influence solid-state desolvation processes.


Supporting Information Desolvation Processes in Channel Solvates of Niclosamide
Jen E. Mann, Renee Gao, Shae S. London and Jennifer A. Swift* Georgetown University, Department of Chemistry, Washington, DC 20057-1227

Table of Contents Page
Figure S1.Optical micrographs of SMeOH, SMeOH and HA.Each form crystalizes as needles but with slightly different aspect ratios.
Table S1.Solid state reaction models and integral expressions used for kinetic analyses.

Figure S2.
Solvent channels in HA, SMeOH, and SACN viewed with the Mercury Solvate Analyzer tool using a probe radius of 1.0 Å and a grid spacing of 0.3 Å.Each solvate is viewed along and normal to the channel axis.020), ( 120), and (1-1-2) diffraction lines of SACN with temperature.From 25 to 62 °C, the peaks decrease by 18.7%, 24.5%, and 18.6%, respectively.Table S3.Correlation coefficients associated with different solid state reaction models for SACN (ground) isothermal TGA data (40, 45, and 50 °C).Reaction models with R 2 > 0.99 are red and with R 2 ≥ 0.999 are red and bold.Table S2.Correlation coefficients associated with different solid state reaction models for SMeOH (ground) isothermal TGA data (40, 45, and 50 °C).Reaction models with R 2 > 0.99 are red and with R 2 ≥ 0.999 are red and bold.Table S3.Correlation coefficients associated with different solid state reaction models for SACN (ground) isothermal TGA data (40, 45, and 50 °C).Reaction models with R 2 > 0.99 are red and with R 2 ≥ 0.999 are red and bold.

Figure S3 .
Figure S3.DSC curves of heat flow vs. temperature for HA, SMeOH and SACN that were (A, B, C) hand-ground with a mortar and pestle or (D, E, F) unground.All samples were heated at 5 ℃/min to 250 ℃ in aluminum pans with unsealed lids.

Figure S4 .
Figure S4.TGA of heat flow vs. temperature for (A) SMeOH and (B) SACN.Samples were heated at 5 ℃/min in open pans.

Figure S5 .
Figure S5.Experimental sPXRD of SMeOH at the start of the dehydration experiment compared against the simulated PXRD from the single crystal structure.

Figure S7 .
Figure S7.Experimental sPXRD of the SMeOH dehydration product at 71 ℃ and 115 ℃ compared against the simulated PXRD of the F1 single crystal structure.

Figure S9 .
Figure S9.Ea values determined from model-based and model-free kinetic analyses of SMeOH TGA isothermal desolvation at 40, 45 and 50°C.(A) The three nucleation models with the highest R 2 values (A2, A3 and B1) yielded similar Ea values.(B) Representative Arrhenius plot of the A2 model.Time-dependent Ea values calculated from model-free (C) Friedman and (D) Standard methods show a decreasing Ea as a function of reaction progress.

Figure S10 .
Figure S10.Experimental sPXRD of SACN at the start of the dehydration experiment compared against the simulated PXRD from the single crystal structure.

Figure S12 .
Figure S12.Contour plots of SACN heated to and held isothermal at (A) 40 ℃ and (B) 45 ℃.Samples were loaded into the capillary as a wet paste (40 and 45 ℃) and were able to completely transition to phase pure Form 1 after approximately 31 and 20 minutes, respectively.

Figure S13 .
Figure S13.Ea values determined from model-based and model-free kinetic analyses of SACN TGA isothermal desolvation at 40, 45 and 50°C.(A) The three nucleation models with the highest R 2 values (A2, A3 and B1) yielded similar Ea values.(B) Representative Arrhenius plot of the A2 model.Time-dependent Ea values calculated from model-free (C) Friedman and (D) standard analysis methods.

Figure S15 .
Figure S15.Thermal expansion in HA and H* over the temperature range up to 150 °C.

Figure S1 .
Figure S1.Optical micrographs of SMeOH, SMeOH and HA.Each form crystalizes as needles but with slightly different aspect ratios.

Figure S2 .
Figure S2.Solvent channels in (left) HA, (middle) SMeOH, and (right) SACN viewed with the Mercury Solvate Analyzer tool using a probe radius of 1.0 Å and a grid spacing of 0.3 Å.Each solvate is viewed along and normal to the channel axis.

Figure S3 .
Figure S3.DSC curves of heat flow vs. temperature for HA, SMeOH and SACN that were (A, B, C) hand-ground with a mortar and pestle or (D, E, F) unground.All samples were heated at 5 ℃/min to 250 ℃ in aluminum pans with unsealed lids.

Figure S4 .
Figure S4.TGA of heat flow vs. temperature for (A) SMeOH and (B) SACN.Samples were heated at 5 ℃/min in open pans.

Figure S5 .
Figure S5.Experimental sPXRD of SMeOH at the start of the dehydration experiment compared against the simulated PXRD from the single crystal structure.

Figure S9 .Figure S10 .
Figure S9.Ea values determined from model-based and model-free kinetic analyses of SMeOH TGA isothermal desolvation at 40, 45 and 50°C.(A) The three nucleation models with the highest R 2 values (A2, A3 and B1) yielded similar Ea values.(B) Representative Arrhenius plot of the A2 model.Time-dependent Ea values calculated from model-free (C) Friedman and (D) Standard methods show a decreasing Ea as a function of reaction progress.

Figure S12 .
Figure S12.Contour plots of SACN heated to and held isothermal at (A) 40 ℃ and (B) 45 ℃.Samples were loaded into the capillary as a wet paste (40 and 45 ℃) and were able to completely transition to phase pure Form 1 after approximately 31 and 20 minutes, respectively.

Figure S13 .
Figure S13.Ea values determined from model-based and model-free kinetic analyses of SACN TGA isothermal desolvation at 40, 45 and 50°C.(A) The three nucleation models with the highest R 2 values (A2, A3 and B1) yielded similar Ea values.(B) Representative Arrhenius plot of the A2 model.Time-dependent Ea values calculated from model-free (C) Friedman and (D) standard analysis methods.

Figure S15 .
Figure S15.Thermal expansion in HA and H* over the temperature range up to 150 °C.

Table S2 .
Correlation coefficients associated with different solid state reaction models for SMeOH (ground) isothermal TGA data (40, 45, and 50 °C).Reaction models with R 2 > 0.99 are red and with R 2 ≥ 0.999 are red and bold.

Table S1 .
Solid state reaction models and integral expressions used for kinetic analyses.