(W1130-02-09) Evaluating Analytical Methods for Moisture Determination in Sulfobutylether β-Cyclodextrin and Characterization of Sorption Behavior by DVS

Purpose: Water content is a critical attribute defining true excipient mass of sulfobutylether β-cyclodextrin (SBECD) solids. We compared four analytical techniques: dynamic vapor sorption (DVS), thermogravimetric analysis (TGA), and coulometric and volumetric Karl Fischer titration (C-KF, V-KF), to identify a practical, precise method for routine SBECD moisture assessment.

Methods: Twelve SBECD lots were obtained, including four test lots for method screening and five evaluation lots for follow-up analysis. Moisture content was quantified using dynamic vapor sorption (DVS), thermogravimetric analysis (TGA), coulometric Karl Fischer titration (C-KF), and volumetric Karl Fischer titration (V-KF), each performed in quintuplicate (n = 5). ANOVA was used to assess statistical differences between methods and lots (α = 0.05). DVS was also used to generate sorption–desorption isotherms (0–60 % RH, 10 % steps) for selected SBECD and non-derivatized β-cyclodextrin lots.

Results: TGA, C-KF, and V-KF produced statistically equivalent results (p = 0.559) (Fig. 1). DVS consistently reported lower moisture values than other techniques, likely due to its limited temperature range and dynamics associated with reaching equilibrium states, often complicated by hysteresis. Using the three preferred methods, the moisture content differed significantly among lots evaluated using these methods (p = 0.022). Lot MS-02 had the highest moisture level, likely due to differences in packaging and storage over time. TGA showed the best balance of sample size, run-time, and precision (RSD = 7.87 %) and was applied to five additional lots. (Fig. 2) One lot, V-01, appeared to have moisture levels similar to MS-02. The average water level across all lots by TGA is 6.5±2.0%, or eliminating high and low lots, 6.4±1.3%, with a median of all lots being at 6.1% water content. While DVS did not provide consistent values for moisture content, it can be used to understand the sensitivity of the material to relative humidity and how important that may be in determination of the values of water content. DVS isotherms for lots of V‑02 and V‑04 showed type‑II/III behavior with reversible, near‑linear uptake with minimal hysteresis up to 60 % RH (Fig. 3). Typical humidity ranges in a lab are from 30-50%, winter to summer, which will result in water contents ranging from 4-6% versus 10-11%, respectively, if not carefully packaged and stored. (Fig.3)

Conclusion: TGA, C-KF, and V-KF are interchangeable for SBECD moisture quantification, but TGA offers superior practicality and acceptable precision for routine use. DVS remains valuable for sorption studies, but underestimates absolute moisture and is highly dependent on instrument parameters used in evaluation. These findings guide method selection when determining and controlling water content in SBECD and presumably related drug-CD systems.

References: Ohtake S, Shalaev E. J Pharm Sci. 2013;102(4):1139–1154.
Stella VJ, Rajewski RA. Int J Pharm. 2020;583:119396.
Pereva S, et al. Beilstein J Org Chem. 2019;15:1592–1600.

Average water content of four SBECD test lots measured by DVS, TGA, C-KF, and V-KF (n = 5). Error bars show standard deviation.

Average water content and residuals for five SBECD validation lots measured by TGA (n = 5). Error bars show standard deviation; residuals indicate deviation from lot means.

Dynamic Vapor Sorption isotherms for SBECD lots V‑04 (A) and V‑02 (B). Left panels: full adsorption–desorption cycle (0–60 % RH). Right panels: adsorption only, extended to 90 % RH.