(684f) Can the Solvation Manner Leads to a Nucleation Diversity for Polymorphic System? the Case of D-Mannitol | AIChE

(684f) Can the Solvation Manner Leads to a Nucleation Diversity for Polymorphic System? the Case of D-Mannitol


Liu, S. - Presenter, National Engineering Research Center of Industry Crystallization Technology
Gong, J., Tianjin University

Can the
solvation manner leads to a nucleation diversity for polymorphic
system? The case of D-mannitol

There has been much
interest in the role of solution chemistry plays in determining the polymorphic
nucleation process, especially for the field requires a precise control and
even design of the final crystalline product, such as pharmaceutical and fine chemical
industries. In this study we focused on a quite
challenging subject: D-mannitol (D-man). An interesting
solvent-dependent nucleation behavior was found by us: in pure methanol the alpha
form can be easily harvested regardless of the supersaturation and temperature,
in water the alpha can also be sporadically observed but the solution mainly crystallized
beta form, whereas for mixtures of water and methanol, interestingly, the pure delta
form was exclusively yielded instead of either alpha or beta. We wondered the
reason for this solvent dependent crystallization outcomes, especially the
unexpected appearance of delta form, however, the nearly identical molecular
conformations and extensive hydrogen bonds throughout the crystal structures making D-man a system
difficult to resolve [1], as shown in Fig. 1.

points are mainly concerned in this study : 1. exploring is there any
structural feature could differentiate three anhydrous polymorphs of D-man
properly; 2. based on such difference, whether a ¡°link¡± could be detected
between the nature of intermolecular interactions present in solution and the
crystallization outcomes, especially the unexpected appearance of delta form.

Though a Hirshfield
Surface analysis [2], we found two important characteristics can be used to
differentiate three polymorphs, namely the dimerization effect and terminal
packing intensity, as illustrated in Fig.2 and TABLE 1.

In light of above
comparison, the nature of interactions in different D-man solutions was
investigated in terms of the radial distribution function (RDF) calculation. The
solvation effect on D-man¡¯s body parts (-O3H3) and terminal parts (-O1H1) were
both investigated. The RDF results together with snapshots for the molecular
distribution within different solutions are shown below:

From above results, no
link was found between the solvation effect on D-man¡¯s body part and the dimerization
effect in three forms, indicating that the dimerization strength cannot be a
probe for the nucleation behavior explanation. While a good correlation was
observed for terminal packing characteristics: in pure methanol and water, the
solvation intensities at the end part are 3 and 2.7, respectively, while in the
mix solvent, much higher solvating power on terminals was observed, with the
terminal-water in 4 for g(r) and terminal-methanol in 3. A double-layer solvent
shell with water molecules in the inside track while methanol in outsides was
formed, considering the higher probability of methanol-water assemblies than
the homologous solvent aggregates (as illustrated in Fig. 5(c)), we have reason
to believe that those outside methanol may be bonded to the inside water shell
to some degree. In the following process of solute molecule assembly, this
distribution will serve as a shield for solute aggregation, create limited
space for solute-solute interactions, thus facilitate a head-to-tail connection
due to the relatively small volume of both ends, and meanwhile bring about a lower
local packing compactness in final structure, consequently leads to the
appearance of delta form. On the other hand, the light solvation degree in
other two solutions can provide favorable condition for more solute molecules
to occupy the area adjacent to terminals. As lots of solutes accumulate around
the ends, one exposed hydroxyl group at the molecular terminal may be likely to
bond with others in more orientations, contribute a lower selectivity for
terminal groups, which could be supported by the separate peaks of C1 and C6
(two terminal carbon atoms) obtained from solid state NMR characterization, as
shown in Fig. 6. The broader space in solution together with the optional
bonding orientation finally stabilized an approximate herringbone linkage
network and a dense packing motif exist in alpha and beta.

  Thus we have found that the packing
characteristics in the lattice of D-man should be a direct reflection of the
solvation state of D-man in solution, the solvating power of different solvents
on the both ends of D-man can provide distinct selectivity for terminal ¨COH
groups to interact with each other. To verify the assumption, same experiments
were carried out in the acetone-water system. The results showed that in
water-acetone solvent, the delta form was not crystallized exclusively, instead
a mixture of alpha and delta was usually yield. This phenomenon to a certain
extent supports our conclusion, as acetone possesses lower HB donor and
acceptor abilities than methanol, cannot construct a firm double-solvent-shell
as observed in methanol-water, as a result a local environment being suitable
for both alpha and delta was created.  

This study, for the first
time, investigated the solvent-dependent nucleation behavior of D-man which is one
of the most important excipients in pharmaceutical and food industry [3, 4],
and revealed the significant roles played by solvation manners in determining
polymorphic outcomes of D-man. In this regard, the use of local packing
compactness as a probe might constitute a potential way could be extended to
other systems for polymorphic nucleation research.


[1] Carespacheco, M. G.,
Vacamedina, G., Calvet, R., Espitalier, F., Letourneau, J. J., & Rouilly,
International Journal of Pharmaceutics, 475 (2014) 69.

[2] M.A. Spackman, J.J. McKinnon,
CrystEngComm. 4 (2002) 378.

[3] Wagner, C.M., Pein, M.,
Breitkreutz, Pow. Tech. 270 (2015) 470.

[4] Rowe, R.C., Sheskey, P.J.S.,
Owen, S.C., 2006. Handbook of pharmaceutical excipients, 449.