Polymorphic Pharmaceuticals and Fine Chemicals - Essay Example

Academia-research.com order 113216 Topic: An approach fine chemicals. Approximate word count: @ 2340 Total number of pages: 11 (eleven) General Introduction (to the key words used). Chemically all substances exist in three states: i. Solid, ii. Liquid, and/or iii. Gaseous states. Some substances in solid state have a distinct regular array of their constituent particles (atoms, ions or molecules). Such an arrangement facilitates the formation of constant angles between their faces. The faces are found to have distinct edges. Obviously, a definite relationship is observed between the faces or edges and/to the arrangement of the constituent particles. Those substances, in solid state, that conform to the above description are referred to as crystals and the process of forming crystals is termed as crystallization. Crystals usually are formed from a solution saturated with the solute. Upon breaking, large crystals form small crystals. The most commonly noticed forces in crystals are the weak Van der waals forces. Consequently, the melting point of crystals is never very high (1, 2). Structures of many substances, including crystals, vary with temperature. This ability of various substances to exist in more than one form is known as Polymorphism. Allotropy is a synonym for polymorphism and is usually used in the context of elements. Dimorphism is the word reserved for a substance that exists in two forms. Yet another definition exits for polymorphism. It also refers to the multiple crystals that might form owing to improper solvents used during the process of crystallization. Structures of almost all substances consist of bonds, which can be intermolecular or intramolecular. In these, occasionally, one finds that hydrogen (H) bound to a strongly electronegative element (X) acquires a positive charge owing to the bond polarization by the electronegative element (represented as X- H+). Such a polarity charged hydrogen is available for interaction directly with the electronegative elements of adjacent molecules, and the resultant intermolecular bond is referred to as Hydrogen Bond (1,2) represented as three dots: X- -H+ . . . X- -H+ It is this hydrogen bond that accounts for the unusually high boiling points of some liquids, viz., Water (H2O), Hydrogen fluoride (HF), etc. Hydrogen bond is also involved in dimer formation as in carboxylic acid, and is the bond responsible for the stability observed in nucleic acids. When hydrogen bonding is present in crystals, it significantly affects the crystal molecular geometry (2). Thus, hydrogen bond profoundly influences the physical and chemical properties of various substances. Continuing the discussion on similar lines, in a covalent bond, the electrons between atoms or groups with different electronegativities tend to be polarized towards the more electronegative constituent. In such situations, a partial charge can be attributed to the constituents owing to the partial ionic nature of the bond. The ability of an atom in a molecule to attract electrons towards itself is termed electronegativity (EN). The ionic character of the bond can be used as a measure of the magnitude of this effect (the partial charge/EN). When the effect is small, the bond is referred to as a polar bond and treated using dipole moments (DM). Covalent bonds are expected to have a DM of zero, provided the electrons are shared equally by the two atoms (1, 2). In a way, DM is a quantitative measure of polarity, with Debye as units. If bond angles are known, DM is estimated by vector addition of individual bond moments. Possession of a dipole moment permits direct interaction with electric fields or with the electric component of radiation (1). With measurements comes mathematics. Whenever a process/object/concept is characterized in terms of mathematics, relatively simple manipulation of variables can be achieved fundamentally to determine how the process, object or concept behaves in different situations. Such an exercise is traditionally referred to as mathematical modelling. And when a person applies a collection of techniques primarily to model or mimic molecular behaviour, s/he is carrying out molecular modelling. Based on the experiences gained either by modelling, observation or scientific reasoning, a person can foretell, or the effort commonly referred to as prediction. In this process of prediction, a person can use results obtained by simulation (3). In any solution, there is a solvent in which the solute is dissolved. Consequently, a range of electrostatic interactions is observed between the molecules present inside the solution. Evaluation of these interactions represents a major challenge in macromolecular simulations. Particularly noteworthy in these evaluations is the polarization of the surrounding solvent in response to the charge distribution on the macromolecules. Error creeps in if these polarizations are inadequately represented. Sometimes reorienting DM on the water molecules, if any, influences the effective electrostatic interactions between the atomic charges of the macromolecule. This effect is referred to as 'solvent screening' effect. Solvent screening effect must be accurately modelled because it is known to be very strong, and is found to affect the forces between macromolecular charges by up to a factor equal to the dielectric constant of water at room temperature. Several techniques for measuring the same exist (4). Because crystallization is a process that involves solutions, usage of improper solvents results in unwanted polymorphic forms. This is where accurate estimation of the various parameters and effects mentioned above comes into picture. Additionally, information, thus generated is also useful in 'separation science', a subject that revolves around separation of different macromolecules in solution (5). To aid in such selections of solvents, International Conference on Harmonization (ICH) has come up with three different classes of solvents, named as classes 1, 2 and 3 (6). Mirmehrabi and Rohani's paper [henceforth referred to as 'the authors' and 'the paper'] lists the properties of two of these classes (7). Furthermore, the authors have used partial charge calculation as a tool for estimating the solvent impact. As case studies, the authors used Ranitidine Hydrochloride (RAN-HCl) and Stearic acid (StA). Before we proceed to discuss the specifics, it is imperative that we also talk about the background information and the lacunae in the same. Background. As mentioned in the introduction, polymorphism appears to be an inherent property of crystals. Consequently, when appropriate care is not taken during crystal formation, there's every possibility of ending up with an unwanted crystal structure. This has significance in biological systems wherein a wrong/unwanted structure of a therapeutic agent might have adverse consequences on patients. Additionally, such wrong structures might have no effect or might not perform the intended function, defeating the purpose of the entire exercise. Also, to achieve the correct structures avoiding polymorphic forms, usage of appropriate solvents, as stated earlier, becomes vital. While a number of effects, viz., solvent screening effect, etc., influence the selection of appropriate solvent, the International conference on Harmonization (ICH) has laid certain guidelines for usage of appropriate solvent taking into account the safety of the solvents when used in humans (6). The Lacunae. A major problem that plagues chemistry, in general, and separation science or pharmaceutical sciences, in particular, lies in the fact that partial charges, i.e., DM, for a number of substances is yet to be determined. This means that the chemical properties of a number of solvents are yet to be documented. Put simply, owing to the lack of proper information on solvents, pharmaceutical companies/pharmacists perhaps end up with substandard polymorphic forms of the biologically active compounds. Hence, characterization of various pharmacologically active solvents becomes vital. 'The Paper'. To fulfil these lacunae, the authors attempted to characterize/study the solvent impact on crystallization of various substances. Such an effort is essential given the fact that solvents do influence formation of crystal polymorphs. Towards this end, the authors suggest a method based on group theory. The results thus obtained are correlated and used for predicting hydrogen-bonding ability of the solute and/or the solvent molecules. Eventually, the predictions are compared with the correlations obtained using quantum mechanics. As case studies, the authors used RAN-HCl and StA. Whereas RAN-HCl blocks acid production in stomachs, coating of tablets is done frequently using StA. In other words, the authors used partial charge calculations as tools for predicting solvent impact. Needless-to-say, and loosely put, in the long run, the findings, predictions, and conclusions of this study aid in contributing to the production of safer pharmaceuticals. The Details. In every hydrogen bond, there's a donor (HBD) and an acceptor (HBA). It has been previously documented that the hydrogen bond formation is a direct function of the dipole-dipole interactions of HBD and HBA. In such a situation, the bond energy is found to increase significantly, if the donor molecule were to be positively charged. However, it was found that an increase in solvent polarity tends to disrupt the inter and intra-molecular hydrogen bonds of the solute (8). This means that increasing the solvent polarity, in part, makes the solute come out of the solution. In other words, the solute crystallizes. Hence, development of correlations to predict hydrogen bonding ability aids in and has immense potential in i. Separation science, ii. Crystallizations, and iii. Consequently, pharmaceuticals. Precisely, this is what the authors did - proposing a group theory for partial charge calculation, and using these partial charges to estimate hydrogen bonding ability. This they followed it up with correlations and predictions with regard to solvent behaviour. A number of factors influence the hydrogen bonding abilities of molecules/atoms. Some of these include: i. Atomic charges of HBD or HBA atoms, ii. Molecular orbital charges. Till recently, quantum mechanics were employed to calculate the atom and molecular orbital charges. Additionally, computer software tools are available for the calculation of the same (9). However, the authors state that using traditional methods makes for a difficult, cumbersome, and computationally expensive effort. In contrast, the authors claim, although debatable, their method is easy and requires minimal effort (7). Here again, there is a hitch. How does one calculate the partial charges Conventionally, calculation of partial charges involved EN. But then, how does one quantify EN Although, since 1930s it is known that EN of an atom is related to its electron affinity or its ionization potential, the exact relationship is still unknown (10). Given this situation, Pauling developed an electronegativity scale by assuming that EN of an element is proportional to its bond strength in various compounds (11). Nevertheless, a limitation exists. An initial EN must be postulated. Despite this, the EN scale developed by Pauling continues to be a gold standard for many researchers, and for reasons that will be elaborated here at a later stage. When partial charges were calculated as explained above, the authors observed same partial charge for all similar atoms in various parts of the molecule. This, the authors state that, is a consequence of not considering the position of atoms in the molecule in addition to the bond strengths, in the earlier methods. Hence, the authors recommend calculations of the partial charges using their proposed group theory. Based on this, various correlations were developed for oxygen and nitrogen considering each of these two molecules/atoms as HBD and HBA. The results and conclusions of these calculations, correlations, in particular, and the procedure, in general, were applied to RAN-HCl and stearic acid. The authors conclude the paper by saying that hydrogen bonding is one of the most important parameters in solvent selection for polymorphic isolation of various compounds. Additionally, they also confirm the findings of Kamlet et al., (8). Drawbacks. Though significant, this paper raises a few questions. Foremost among them being, 'when authors themselves have clearly stated that using Pauling's EN values they found that there was less error in correlations; of what advantage is the authors proposed method, as the same conclusions can be arrived at using the methods currently in vogue and developed by earlier researchers' Secondly, if the method proposed by the authors also relies on EN, which it does as per their own suggestion, the limitations of earlier methods remain, and hence contribution of this paper reduces to the various values mentioned in the tables and case studies. [Additional reasons why Pauling's EN scale can be considered more useful than the rest]. Additionally, a number of markers mentioned in all the tables listed in the paper were missing. Constant reference in the text of the paper to the compounds identified by the missing markers confuses the reader. It is like referring to something that was never there. Generating a model, be it mathematical or molecular, and forecasting basing on those models, makes the reader lose interest in the publication and makes the paper an 'also ran' in this vast info-junkyard of scientific literature. This particular point becomes important when we consider the fact that publications in reputed scientific journals are read by a number of researchers working in various fields and need not necessarily be only by the people of 'the authors' research field. In this situation, absence of a list of abbreviations also adds to the confusion. Perhaps the journal personnel and publishers should also take some responsibility for the same. Needless-to-say adequate care in the preparation of the manuscript was missing making 'the paper' appear, at certain places, as 'cut and paste' from earlier published work of 'the authors'. Despite these, the paper is definitely useful for chemists, biochemical engineers and pharmacists for the values of various compounds listed in the tables, the case studies and a comparative review of methods of different calculations. Of course, the paper makes for a good reference and might be cited by different authors in future. 0 - 00 - 0 - Endnotes. For representing dissociation or ionization constants in 'the paper', the standard symbol K was used. 'r' is the correlation coefficient and can have any value ranging from zero (no association; viz., the size of tomato fruits and number of seeds) to +1 (complete positive correlation; perhaps, age and body length of insects like locusts) or -1 (absolute negative dependence; like atmospheric oxygen pressure and rate of spiracle opening in insects). The square of the correlation coefficient, r2, indicates for what proportion of the initial variable the association with the other variable can accounted for. Coefficient of multiple determinations, R2, says what proportion of the variation of the dependent variable is explained by a set of independent variables (12, 13). Cited References/Sources (in the order they appear in the text): 1. John Daintith (Ed.). 1983. Cosmo Key facts - Dictionary of Chemistry. Cosmo Publications, New Delhi, India. 2. Jerome L. Rosenberg. June 1983. Schaum's Outline of Theory and problems of college chemistry, 6th Ed. Asian Student Edition. Schaum's outline series, McGraw-Hill International Book Company, Singapore. 3. http://www.google.co.in/ accessed on-line on 17-Feb-2006. 4. Sonja M. Schwarzl et al. 2003. How well does charge reparametrisation account for solvent screening in molecular mechanics calculations The example of myosin. In silico Biology; 3, 0016. Accessed on-line at http://www.bioinfo.de/isb/2003/03/0016/main.html on 17-Feb-2006. 5. http://www.answers.com/separation%20science accessed on 17-Feb-2006. 6. ICH Topic Q3C. Residual solvent. http://www.emea.eu.int/pdfs/human/ich/028395en.pdf as mentioned in Reference/source 7, and ICH guidelines, accessed on-line at http://www.ich.org/cache/compo/276-254-1.html on 18-Feb-2006 7. Mirmehrabi M, Rohani S. July 2005. An approach to solvent screening for crystallization of polymorphic pharmaceuticals and fine chemicals. Journal of Pharmaceutical Sciences; vol. 94 No.7: 1560 - 1576. 8. Kamlet MJ, Dickson C. 1982. Linear solvation energy relationship. 16. Dipole/Dipole contribution to formation constants of some "hydrogen bonded complexes". J Org Chem; 47: 4971 - 4975, as mentioned in Reference/source 7. 9. Gancia E et al. 2001. Theoretical hydrogen bonding parameters for drug design. J Mol Graph Mod; 19: 349 - 362. 10. Abraham MH et al. 1989. Hydrogen bonding. Part 9. Solute proton donor and proton acceptor scales for use in drug design. J Chem Soc Perkin Trans II; 1355 - 1375, as mentioned in Reference/source 7. 11. Pauling L. 1967. The Chemical Bond. New York: Cornell University Press, and Pauling L. 1989. The nature of chemical bond. IV. The energy of single bonds and the relative electronegativity of atoms. J Am Chem Soc 54: 3570 - 3582, as mentioned in Reference/source 7. 12. Abramson JH. 1984. Survey Methods in community medicine - an introduction to epidemiological and evaluative studies, 3rd Ed. Churchill Livingstone, New York, USA. 13. Taylor DJ et al. 2002. Biological Science, 3rd Ed. (Soper R, Ed.). Cambridge University Press, Chennai, India. Read More