Rodolfo Pinal, Ph.D. Associate Professor of Industrial and Physical Pharmacy Director, Dane O. Kildsig Center for Pharmaceutical Processing Research Phone: (765) 496-6247 Fax: (765) 494-6545 E-mail: rpinal@purdue.edu Full directory listing

Specialization: Drug solubility and solubilization. Strategies for enhancing the bioavailability of poorly soluble drugs.

Education

Research: Drug solubility and solubilization. Strategies for enhancing the bioavailability of poorly soluble drugs.

Antiplasticization

The interactions of water with polymers such as microcrystalline cellulose and starch present an effect that has received little attention. Water is a known plasticizer of these materials, having the effect of softening them. However, low water levels of the solvent have the exact opposite effect: antiplasticization. This means that small amounts of water actually harden the polymer before higher moisture levels begin to soften it. This phenomenon is actually widespread, it also occurs with other polymer-solvent combinations, such as those used in packaging. The adjacent figure shows how there is a range in the water content for microcrystalline cellulose (MCC), where its compacts are stronger than those having either higher or lower levels of water.

We have found a similar effect on the permeability of water. There is a water level where the ability of water to permeate through the MCC compact is minimum, slower than in either dryer or wetter matrices. The practical implications of this phenomenon can be significant in cases where the presence of water is a concern. While being more difficult and more expensive, water removal to complete dryness may not be as effective as an optimally selected low moisture level.

Is the Young's modulus the only property where this type of effect is observed?

Not by a long shot (Tablets & Capsules 2007, 5, 22-33). In fact, this phenomenon is quite common and amazingly easy to reproduce in the laboratory, hence in practice, with different properties.

For example, the figure below shows the permeability of water vapor in the polymer Eudragit E-100 as a function of relative humidity (RH). With sorbed water acting as a plasticizer, the common expectation is that permeability ought to increase as the amount of water present increases. This is indeed the case, with the notable exception of the first portion of the curve. When water is present at very low levels, it depresses permeability before greater concentrations begin to increase it.

Antiplasticization is a rather general phenomenon with important implications to pharmaceutical products. You can readily see it in the mechanical, transport and structural relaxation properties of pharmaceutical polymers. Even though it is not widely recognized in the pharmaceutical field, this is a remarkably persistent phenomenon, whose impact goes beyond the physical properties of pharmaceutical materials and formulations. Antiplasticization is observable in the drug release performance of some formulations (Colloids Surf. A, DOI:10.1016/j.colsurfa.2008.05.047).

Solubility

Dissolution limited bioavailability is a very common problem in pharmaceutical development. The well known adage "like dissolves like" is, I am sorry to say, not true. Well, it is true, but only for liquids. If the solute is a solid, then the solid properties of the solute play an often dominant role. Many drugs are hydrophobic, which makes them poorly soluble in water. But when we have a nightmarish solubility problem, one where the hydrophobic solute does not dissolve in water, but it does not dissolve in hydrophobic liquids either, then it is best to look at the solid instead of at the solvent.

The phenanthrene-anthracene example is very nice because it makes it very easy to show an important point. The figure to the left shows the "magnitude" of the problem. It is a log scale, so that the solubility in benzene is about a million times higher than in water for each of the two compounds. This means that by solvent manipulation we can increase the solubility of these compounds a million fold. If instead of phenanthrene and anthracene we had say drug A and drug B, could we solubilize drug B to the same level of drug A? The answer is no.

Anthracene will be 25 times less soluble than phenanthrene in every solvent we care to try. The black portions of the bars are a property of the solid and are completely unaffected by the choice of solvent. The white part of the bars in the figure are labeled "Lyophobicity," that is to say, how much the solute "dislikes" the solvent. If the solvent is water, then lyophobicity=hydrophobicity, but if the solvent is benzene for example, we see that the dislike of antrhacene and phenanthrene for this solvent is, as expected, not there. Anthracene and phenanthene both like benzene equally and very much, this is what is meant by "like dissolves like." The reason anthracene is 25 times less soluble is not that it dislikes benzene; it is that it has a strong crystal that makes it difficult to dislodge the molecules from it.

This is an extremely common problem during drug candidate selection, where we may have two or more drugs with remarkably similar structure but very different solubilities. There is so much we can do by solvent manipulation if we want to improve the bioavailability of a drug that is very poorly soluble and therefore dissolves very slowly. We need to add energy to that solid so that its crystal structure gets disrupted.

Structural Effects on Solubility

What makes the solubility of two isomers like anthracene and phenanthrene so different?

Even though they are chemically very similar, the two isomers have different symmetry. That gives them the ability to arrange into crystalline structures of different strength. It is something like having two jigsaw puzzles made with the same cardboard but different shape of cuttings, so that one pattern is more difficult to tear apart than the other. The different size of the black portions of the bars in the figure above is not the result of different intermolecular interactions in the crystals of anthracene or phenanthrene. It is the result of the different type of physical montage each can form. The more symmetrical a molecule, the stronger the crystal it can form. The reason is that it is easier for a symmetrical molecule to orient and align into a crystalline lattice than for a non symmetrical one. This effect is easiest seen with benzene. If I take a molecule of benzene and rotate 60° clockwise, the result will look identical to the original position. I can do the 60° rotation six times with the same result. Then I can flip the benzene molecule so that the front is back and the back is front, and again, I have six possibilities to make the picture look identical. So benzene has a symmetry number (σ) of 12, because there are 12 indistinguishable ways I can rotate the same molecule. So, a benzene molecule has 12 different ways of occupying the same "spot" in a benzene crystal. Benzene by the way, has a strong crystal (high melting point) if we compare it with toluene for example. Anthracene and phenanthrene have symmetry numbers of 4 and 2, respectively. So anthracene can form a stronger crystal. My view (Org. Biomol. Chem. 2004, 2, 2692-2699) is that in the ideal case, the solubilities of isomers of different symmetry are related to each other by a relationship of the following form:

where T is the experimental temperature, S denotes solubility, Tm is the melting point, and the subscripts 1 and 2 denote compound 1 and 2, respectively, and compound 2 has the higher symmetry (σ₂ > σ₁).

Amorphous formulations

Why do we want amorphous formulations?

When the crystal structure of the solute is so strong that the molecules simply do not get dislodged from the crystal to go into solution and be subsequently absorbed, then we need to explore the possibility of a non-crystalline formulation. We apply energy (by different possible means) so that we obtain a solid without the ordered structure that makes crystals stable. This high energy form is more soluble, but also less stable. At present, amorphous dispersions of drugs in polymers is a subject that generates intense interest. I am particularly interested in exploiting the properties of two-phase amorphous dispersions. These are dispersions where the amorphous drug and the polymer are not miscible.

Having the amorphous drug in a separate phase offers some important advantages. In a very practical sense, the stability of the formulation will be to a good extent independent of drug concentration. This way, the characterization work can focus primarily on the properties of the amorphous drug and the properties of the mixture itself, although still important, can be treated separately.

Understanding the Amorphous State

With all their potential advantages, amorphous formulations present some significant challenges. The same properties that make them so attractive for enhancing the dissolution of insoluble drugs make them also unstable. It is critical to be able to have a very good handle on the molecular mobility of amorphous substances. Molecular mobility will determine the feasibility for a compound to be developed as an amorphous formulation. It will also determine the changes in reactivity and physical stability over the shelf-life of the product.

Amorphous mixtures

There is one aspect of amorphous mixtures that I find troublesome. From Kauzmann's now classic entropy paradox, we know that entropy plays a critical role in glass formation. However, if we look at the equations available for estimating the glass transition temperature of amorphous mixtures, we see that a term to account for the effect of the entropy of mixing is conspicuously missing in every case. The typical argument is that entropy of mixing in polymer mixtures is small, so it gets ignored. But being small is not the same as being absent. The polymer argument gets extended to amorphous mixtures of low molecular weight compounds and the effect of the entropy of mixing is still absent. My position (Entropy 2008, 10, 207-223) is that the entropy of mixing must play a role on the glass transition temperature of amorphous mixtures and that to a first approximation, the effect of entropy of mixing has the form:

where T_g^" is the glass transition temperature as influenced by the entropy of mixing, and T_g is the value for the case where the entropy of mixing has no effect on glass formation, ΔS_mix^c is the configurational entropy of mixing and ΔC_p is the difference in heat capacity between the liquid and the glass.

Modeling Molecular Mobility

Time and temperature dependence of the structural relaxation time.

The stability of amorphous formulations is of particular importance given that these systems are by nature metastable. One way or another, as the product changes from the metastable to the stable form, the change involves molecular rearrangement. The structural relaxation time is a quantitative indicator of the degree of molecular mobility in amorphous systems. The relaxation time changes drastically with temperature, and this aspect has been the subject of numerous studies in the past. However, the relaxation time is also strongly dependent on time. This aspect has not received nearly the same degree of attention. Stability at constant temperature is a very important question and my group studies the time- and temperature-dependence of the structural relaxation time.

It is possible to model the relaxation time from a small set of carefully conducted heat capacity measurements. In this approach, the experimental values are fitted to the model. From the fitting the glass forming parameters of the material are "extracted". The extracted values can then be used again in the model to predict the behavior of the same material when exposed to different temperatures for different lengths of time, heating, cooling, and combinations of such steps. One important point, this model has no adjustable parameters, all parameters that are "extracted" can be measured separately. For example, Moynihan's activation enthalpy (the change in the glass transition temperature as a function of heating rate) is one of the extractable parameters. The extracted parameters in the model can be put to work to predict time consuming experiments. For example, the DSC profiles of relaxation enthalpy measurements after various annealing times can be estimated by the model.

An important conceptual aspect of the model is the smooth cross-over between the liquid and the glass. This is an important point because the Arrhenius behavior of the glass and the super-Arrhenius (VTF) behavior of the liquid give place to a kink when they meet, giving place to sharp glass transition temperature instead of to a more realistic glass transition region. In this model, the expressions gradually evolve over a glass transition region from the Arrhenius (glass) to super-Arrhenius (liquid) profile.

Interests

Solution chemistry, solubility and solubilization techniques, mixtures. Polymer-based composites as means of control of product performance (mechanical strength, dissolution, release rate and bioavailability) of bioactive compounds. Antiplasticization. Molecular relaxation in amorphous organic materials.

Representative Publications

S. P. Chamarthy and R. Pinal. Moisture induced antiplasticization of microcrystalline cellulose. Pharm. Res. (2006) submitted.

Y. Miyako, H. Tai, K. Ikeda, R. Kume and R. Pinal.* Solubility screening on a series of structurally related compounds. Cosolvent-induced changes on the activity coefficient of hydrophobic solutes. Drug Dev. Ind. Pharm. 34, 499-505 (2008).

C. Mao, R. Pinal, and K. R. Morris. A quantitative model to evaluate solubility relationship of polymorphs from their thermal properties. Pharm. Res. 22: 1149-1157 (2005).

S. P. Chamarthy and R. Pinal. Moisture-induced antiplasticization in microcrystalline cellulose compacts. Tablets & Capsules, 5, 22-33 (2007)

R. Pinal. Effect of molecular symmetry on melting temperature and solubility. Org. Biomol. Chem. 2: 2692-2699 (2004).

Curriculum Vitae

Click here for a full CV for Rodolfo Pinal. (an Adobe Acrobat file)