Spontaneous Crystallization of Chiral Minerals

In order to prepare chiral crystals from achiral compounds, no special crystallization technique is required. Ordinary methods can be employed, such as cooling a hot saturated solution of a compound or evaporating the solvent to obtain crystals.

Determining if a crystal is chiral can be achieved by measuring the optical rotation in solution using a polarimeter (Koshima 2004).4

In 1990, Dilip Kondepudi, a former member of Ilya Prigogine's Brussels school of thermodynamics, surprised professional circles with a simple crystallization experiment on chiral symmetry breaking. At Wake Forest University in North Carolina, he and his team studied the crystallization of sodium chlorate NaClO3. The crystals of sodium chlorate belong to the cubic space group P2i3 and are -as in the case of quartz - optically active although the molecular compounds of the crystal are not chiral. Kondepudi et al. prepared aqueous solutions of sodium chlorate by dissolving 100 g NaClO3 in 120 mL of distilled water. After complete dissolution of the solute at 50° C, samples of 25 mL were transferred to several petri dishes and left for crystals to grow. After 4 to 5 days, the crystals bigger than 1 mm in size were separated and their optical rotation was measured by a pair of polari-zers.5 The crystals appeared blue when the polarizers were crossed and turn red on rotating the analyser clockwise or anticlockwise, depending on the crystal's hand-edness. For a 1 mm crystal, the change in colour from blue to red occurs at a rotation of about 7°. After precipitation of a total of 1000 crystals, 525 NaClO3 crystals were identified to be of /-configuration whereas 475 crystals were of ^-configuration. As expected, crystals were found in almost equal numbers in repeated experiments. Nothing new. The frequency distribution for crystal enantiomeric excesses was said to be monomodal, as demonstrated more than 100 years ago (Kipping and Pope 1898) and as illustrated in Fig. 4.2.

When the solution was simply stirred during crystallization with a magnetic stirrer, however, a dramatic effect was observed: as before, /- and d-crystals were separated and counted. Almost all of the NaClO3 crystals (99.7%) had the same handedness, either /- or d-! In a total of 32 different crystallization experiments, 18 set-ups were predominantly of crystals in /-configuration whereas 14 were predominantly of d-configuration. The racemic state became unstable. Kondepudi et al. (1990) obtained a bimodal frequency distribution of the crystal's enantiomeric excesses as given in Fig. 4.3. A total of 11 829 crystals was counted. The direction of the obtained enantiomeric excess was unpredictable and random. In this context, the system showed stochastic behaviour. The observed phenomenon occurred without the need for careful control of temperature or other experimental parameters.

4 If a crystal is composed of achiral compounds, its chirality disappears in solution. To circumvent this problem, light emitted from TFT-displays can be used in order to distinguish between the chirality of enantiomorphous crystals. This light is linearly polarized. Enantiomorphous crystals can thus be deposited directly on the screen of a notebook and regarded with a polarizing glass. The required angle of the glass to transmit the light gives information on the crystal's optical rotation and chirality.

5 For determination of the NaClO3 crystal chirality, linear polarizing laminated films from Edmund Optics can be used. One laminar was mounted on a light table (for instance, as used for photography) and the petri dish containing the NaClO3 crystals was placed on top of it. Light shines from the bottom through the diffuser then through the polarizing film (called the polarizer) and finally through the sample. Now holding and rotating a second laminar (called the analyser) between the petri dish and one's eyes makes the optical activity of an individual crystal clearly visible (Personal Communications with Thomas Buhse and Dilip Kondepudi in December 2005).

Fig. 4.2 Quantity of l-NaClO3 crystals in percent (left) and monomodal frequency distribution for crystal enantiomeric excesses (right) after precipitation in non-stirred solutions (see Kondepudi et al. 1990; Kondepudi and Asakura 2001)

This kind of breaking of the chiral symmetry in the form of enantioselective crystallization was not induced by the presence of optically active compounds, external seeds, or other external asymmetric influences.6 Also impurities in the reactants could be excluded for the observed results.

Presenting these intriguing results at his typically overcrowded lectures at international conferences, Dilip Kondepudi was often asked whether the direction of

Fig. 4.3 Quantity of l-NaClO3 crystals in percent (left) and bimodal frequency distribution for crystal enantiomeric excesses (right) after precipitation in stirred solutions (see Kondepudi et al. 1990; Kondepudi and Asakura 2001)

6 In order to distinguish between enantioselective and enantiospecific, Hellwich and Siebert (2006) clarified that enantioselectivity means that one enantiomer out of several possible enantiomers is preferentially produced in a chemical reaction. A reaction is termed enantiospecific, if chiral starting materials are converted into chiral products. If the configuration of the starting material is known, then the product's configuration of an enantiospecific reaction can be predicted.

stirring in the form of left-stirring versus right-stirring of the stir bar might have induced the excess of either d- or /-enantiomorphous NaClÜ3 crystals. He answered - smiling - that in stirred solution three distinct processes are of importance: primary nucleation, autocatalytic secondary nucleation, and crystal growth.7 In the above experiment, secondary nuclei were rapidly produced from a primary nucleus. The chirality of the primary nucleus thereby determines the overall chirality of the system, a process that is chirally autocatalytic. The rapid production of secondary nuclei was confirmed by videotaping the collision of a NaClO3 crystal and the stir bar. Kondepudi and colleagues assumed, however, that secondary nucleation alone is not sufficient to produce the observed homochirality. According to Kondepudi et al. (1990), this process must be accompanied by the suppression of nucleation of crystals of the opposite handedness in "a form of competition". But, what is that?

In Chap. 10, the Frank model will be introduced to explain the amplification of small enantiomeric enhancements. If we use the Frank model as a reference to better understand the formation of almost homochiral crystals by NaClO3 crystallization in stirred solutions, we should note that mutual inhibition of two enantiomers is essential for the amplification of enantiomeric enhancement. In the NaClO3 crystallization, the mutual inhibition is accomplished by the fast event of secondary nucle-ation in which the solute concentration rapidly drops below the saturation level so that no more crystals can be formed. In this context, secondary nucleation is, however, not understood in detail and Kondepudi's explanation remains debated (Buhse et al. 2000; Viedma 2005).

When the NaClO3 solution was not stirred, there was no rapid autocatalytic production of nuclei observable: all of the nuclei were produced through primary nu-cleation and their handedness was random, producing a statistically equal number of /- and d-enantiomorph NaClO3 crystals.

The observed chiral symmetry breaking also occurs in stirred solutions of sodium bromate NaBrO3. Since most, if not all, crystals can generate secondary nuclei, Kon-depudi and his collaborators expect this phenomenon to be observable in crystallization reactions of other achiral compounds that crystallize in enantiomeric forms, too (Kondepudi and Asakura 2001).

Later studies on the kinetics of the spontaneous crystallization of chiral crystals showed that small amounts of homochiral D- or L-lysine added as primary nuclei to a stirred solution of D/L-glutamic acid were capable of determining the chiral resolution of glutamic acid crystals (Buhse et al. 1999).

The symmetry breaking in spontaneous crystallization processes is not limited to the precipitation of chiral crystals from solution. The analogous effect was observed for the crystallization of a chiral polyaromatic hydrocarbon molecule from a melt. In this case, chiral 1,1'-binaphthyl served as the ideal system (Fig. 4.4): the racemization half-life measured at room temperature for a 1,1'-binaphthyl solution in benzene is relatively long and was determined to be 9 h. In the molten state above 158°C, the enantiomers interconvert rapidly with a racemization half-life of less than a second.

7 In fact, the idea of inducing an inverse enantiomeric enrichment into a chemical system by stirring to the right or to the left is not trivial as it turned out very recently and as we will discuss in Sect. 4.3.

Fig. 4.4 Chemical structures of 1,1'-binaphthyl structures that interconvert slowly in benzene solution at room temperature and rapidly in the molten state above 158°C (Kondepudi and Asakura 2001)

Fig. 4.4 Chemical structures of 1,1'-binaphthyl structures that interconvert slowly in benzene solution at room temperature and rapidly in the molten state above 158°C (Kondepudi and Asakura 2001)

The non-stirred crystallization of 1,1'-binaphthyl was observed from its melt by lowering the temperature below 158°C. This experiment resulted in the formation of chiral crystals following a monomodal frequency distribution for the enantiomeric excess with a maximum at zero. Crystallization from a melt stirred by a Teflon stir bar, however, showed - as in the case of enantiomorph NaClO3 crystals - crystal symmetry breaking since the enantiomeric excess generated in almost every crystallization is greater than 80%. The frequency distribution of the enantiomeric excess in this case was bimodal in contrast to the monomodal distribution obtained in the non-stirred experiment (Kondepudi and Asakura 2001). This phenomenon seems to be a general one since other systems such as the chiral octahedral cobalt complex cis-[Co(H2O)(OH)(en)2]2+ show equally a monomodal frequency distribution of enantiomeric excesses in non-stirred solution compared to a bimodal distribution by stirring (Asakura et al. 2000).

Such generation of highly enantioenriched solid systems from the racemic or even achiral liquid state shows us how, in nature, chiral asymmetry in crystals may spread during the processes of chemical evolution before the origin of biological life. Kondepudi et al. concluded that similar processes including autocatalysis and competition might help in understanding the possible origins of biomolecular ho-mochirality. This model is, however, not yet capable of explaining the preference of a particular configuration of one enantiomer over the other.

Was this article helpful?

0 0

Post a comment