Cocrystal Stabilization in Aqueous Media

Introduction & Background

Development of cocrystals for use in pharmaceutical applications is a rapidly expanding field of study.  However, the use of cocrystals as a tool to increase in vivo exposure has yet to gain widespread use. Unlike salts, where dissolution leads to an ionized free form (at least initially), cocrystal dissolution leads to formation of the unionized free form that can quickly crystallize back to the less soluble free API. However, for APIs that have pKa(s) below relevant physiologic pHs, the benefit of using salts versus cocrystals may be negligible given the in vivo exposure for both is controlled by the kinetics of recrystallization of the free form of the API from the metastable, supersaturated solution either in the bulk solution or on the surface of the dissolution layer.

The following schematic shows the possible form changes that can happen in vivo for both salts and cocrystals

Thus, enhanced ‘solubility’ may not guarantee improved performance and sufficient screening may be required to find a cocrystal with the right dissolution properties to give a “spring and parachute” profile [1].

In the worst case, dissolution of the cocrystal can be impeded by the formation of a low-solubility hydrate of the API which deposits on the dissolution surface.  In this case, the dissolution profile initially spikes but drops quickly to Chydrate.

Research Motivation and Objectives

Depending on the kinetics of phase transformation, concentration enhancement and in vivo performance may be negatively impacted. The purpose of this study is to demonstrate that the presence of excess coformer in the cocrystal matrix modifies the dissolution boundary layer and inhibits the formation of the free form of the drug at the dissolution surface, thereby increasing the dissolution rate and the effective concentration of the API with respect to the crystalline free form in solution.

Preparation and Characterization of CBZ cocrystals

Carbamazepine-nicotinamide (CBZ-NCT, 1:1) and carbamazepine-succinic acid (CBZ-SUC, 2:1) cocrystals were used as model compounds.

CBZ-NCT and CBZ-SUC cocrystals were prepared according to literature methods [2,3] and their powder x-ray diffraction (XRPD) patterns were consistent with the reference patterns.

Instruments and Methods

Dissolution was done in 300 mL of 1% sodium dodecyl sulfate aqueous solution at 37°C using a Vankel dissolution apparatus.  Approximately 100 mg of the desired material was pressed into a Wood’s die at 4K pounds on a Carver press.  The dies (n=3) were rotated at 100 RPM and samples (~1 ml) were taken at various time points over the course of 1 hour.

For both CBZ-NCT + NCT and CBZ-SUC + SUC pellets, 1:1 by wt of the coformer was added.  In the case of the coground materials, both the CBZ cocrystals and the respective coformers were ground together in a mortar until uniform (~1 min).  XRPD results of preground and post grinding materials were consistent, indicating no significant loss in crystallinity.

The surface of the pellets was analyzed by FTIR (ATR) to determine the composition of the solid form.  Comparison was made to reference spectra taken of CBZ- dihydrate, the cocrystals, and mixtures of the cocrystals and their respective coformers.

Solution concentrations were determined using an Agilent 1100 series HPLC equipped with a reverse phase column (Waters, Symmetry C18, 3.5 µm, 4.6*100mm). An isocratic method with 10mM sodium dodecyl sulfate (65%) as mobile phase A and acetonitrile (35%) as mobile phase B was used at a flow rate of 0.5 mL/min. The injection volume was 10 µL and column temperature was 25°C. The peak area of carbamazepine was determined at a wavelength of 254nm. Linearity was determined from a range of 0.25 µg/mL to 100 µg/mL .

Results & Discussion

Two model cocrystals (CBZ-NCT and CBZ-SUC) which have been shown to have solubilities greater than their respective API (CBZ) [4] were studied. The intrinsic dissolution rates of pellets made with these cocrystals and pellets made with their respective excess coformer both added as-is and in intimate contact (by grinding) were compared.

The lowest solubility form of CBZ in water is the CBZ dihydrate. The kinetics of conversion can be rapid. Precipitation of the dihydrate from the cocrystal represents a significant drop in solubility (solubility ratio of the cocrystal to CBZ dihydrate is 152 for CBZ-NCT and 5.2 for CBZ-SUC) [4].  In both cases, exposure of the cocrystals to an aqueous media at sink conditions quickly leads to formation of CBZ dihydrate on the dissolution surface. The dissolution rate observed from both is consistent (see Figure 4 and 5 below) and the surfaces show nearly full conversion to the dihydrate by FTIR (see figure 7).

Attempts were made to modify the kinetics of precipitation and thereby increase the dissolution rate by adding excess coformer to the dissolution pellets.

A physical mixture of cocrystal and coformer should lower the solubility and dissolution rate due to Le Chatlier’s principle.  However in these systems it is posited that the added coformer suppresses the supersaturation of CBZ with respect to the dihydrate thereby increasing the dissolution rate.

Initial attempts were made to determine the solid form on the surface of the pellets by XRPD.  However, penetration by x-rays beyond the surface limited its utility for this system.  Due to the sensitivity to surface effects FTIR with an ATR accessory was used.

  • ˜Dissolution of CBZ-NCT and CBZ-SUC cocrystals yielded intact smooth pellets with predominately CBZ dihydrate on the surface (not shown).
  • ˜Dissolution of CBZ-NCT + NCT and CBZ-SUC + SUC physical mixtures yielded intact rough pellets.  The CBZ-NCT + NCT pellet was predominately NCT on the surface (small amount of CBZ-NCT) whereas the CBZ-SUC + SUC was mostly the cocrystal.
  • ˜Dissolution of CBZ-NCT + NCT coground showed rapid expansion of the pellets with breakage.  One pellet had mostly the cocrystal on the surface by FTIR while the other two showed partial dihydrate formation.  CBZ concentration results were not consistent and showed large variance.
  • ˜Dissolution of CBZ-SUC + SUC coground yielded intact smooth pellets. Two of the pellets were mainly the cocrystal while one was predominately succinic acid.  Again CBZ concentration results were not consistent and showed large variance.


Surface analysis by FTIR indicated that addition of excess coformer in both the CBZ-NCT and CBZ-SUC cocrystal matrices suppresses the formation of CBZ dihydrate at the dissolution surface.  Further study of the impact on the dissolution rates would be necessary to show that the CBZ concentration in solution can be rationally affected to increase in vivo exposure in biological models. Proper physical characterization and application of this technique to the formulation development of cocrystals may expand the number of useful cocrystals which may otherwise have had limited utility, and increase the likelihood of successful development for therapeutic applications.


[1] N. Jagadeesh Bubu and A. Nangia Cryst. Growth Des. 2011, 11, 2662-2679.
[2] Nehm, S.; Rodríguez-Spong, B.; Rodríguez-Hornedo, N. Cryst. Growth Des. 2006, 6, 592-600.
[3] S. Childs, N. Rodríguez-Hornedo, L.S. Reddy, A. Jayasankar, C. Maheshwari, L. McCausland, R. Shipplett, B.C. Stahly, CrystEng Comm, 10: 856-864 (2008).
[4] Good D. J. and Rodríguez-Hornedo N. Cryst. Growth Des. 2009, 9(5), 2252-2264.