10 zo


Figure 5 The HPLC trace for Xenon light degraded (1)

In order to identify the nature of the major degradation product or products of compound (1) and compound (2) form II, samples of both solids were exposed to xenon light for long periods (several days) to facilitate degradation of over 50%. Examination of the products of this procedure showed significant amounts of material insoluble in acetonitrile. The insoluble material was subsequently isolated for further analysis and was found to be soluble in dimethyl sulphoxide (DMSO) and dimethyl formamide (DMF). Using these solvents in the mobile phase, Gel Permeation Chromatography (GPC) was carried out, the results of which are given in Figures 6 and 7.

Figure 6 GPC traces for (1) using UV detection. Column: Styragel HR-1 300x7.8 mm, Sample:80/20THF/DMSO, Eluent: THF (40 deg. C) Detection:UV at 280nm (l=blank, 2=unstressed (1), 3/4 xenon light stressed material, 5=extracted insolubles from xenon light stressed material).

The GPC traces clearly show significant peaks due to high molecular weight material not present in unirradiated samples. These data suggest that the principal products of light degradation of compound (1) and compound (2) form II may be polymeric in nature. The appearance of the GPC traces suggest the polymeric material is itself a complex mixture.

The insoluble material was analysed using spectroscopic techniques. The proton nuclear magnetic resonance spectrum showed broad signals consistent with polymeric materials. Infrared spectroscopy indicated a loss in the acrylate double bond. Analysis of the insoluble material was attempted using mass spectrometry. A number of ionisation techniques were used, including electron impact, chemical ionisation, electrospray, fast atom bombardment and matrix assisted laser desorption. In all cases no mass spectrometric evidence for high molecular weight material was found. The reason for this may be due to the insoluble and involatile nature of the polymeric materials which may well posses a degree of cross linking making it unamenable to these techniques.

Figure 7 GPC traces for (1) using refractive index detection. Column: Stryagel HR-1 300 x 7.8 mm, Sample: DMF, Eluent: DMF, Detection: RI (l=blank, 2=unstressed material, 3-5= xenon light stressed material).

1 2 4 6 8 exposure time (hours)

HPLC assay %

Area of TLC baseline peak

Figure 8 The correlation between loss of HPLC assay and intensity of baseline spot on TLC plate as measured using a densitometer for light degraded (2) form II.

1 2 4 6 8 exposure time (hours)

HPLC assay %

Area of TLC baseline peak

Figure 8 The correlation between loss of HPLC assay and intensity of baseline spot on TLC plate as measured using a densitometer for light degraded (2) form II.

The GPC traces of lighted stressed (1) were broad in nature and quantitation of the polymeric species was not possible. The final part of the analytical work was to develop a simple method to indicate the presence of the insoluble materials at low levels and to act as a semi-quantitative or limit test for the polymeric d├ęgradants. For this work a Thin Layer

Chromatography (TLC) method was used which exploited the insoluble nature of the degradants. The exposed samples were first dissolved in dimethlyl formamide and applied to a TLC plate and the solvent evaporated. The plate was then developed carefully whereby only the insoluble material was left on the baseline. The insoluble material may be visualized in UV light and may be made to fluoresce. A densitometer scanner was used to record the spot and compare densities of samples irradiated for different lengths of time. The results of this experiment are summarised for (2) form II in Figure 8. In this way a simple method for detecting the insoluble material was developed and routinely used.

3.2 The Influence of Solid State Structure on Light Stability

The most striking feature of Table 2 is the difference in light stability between the two polymorphic forms of (2). Form I degrades almost exclusively to the cyclobutane dimer (4) in the presence of xenon light. Indeed, 24 hours exposure under regular laboratory fluorescent tubes has also been shown to produce around 4% of (4). In the case of form II, however, the light instability is akin to that of (1), where polymeric products are formed.

Clearly the arrangement of atoms in the solid state structure of the two forms of (2) has a profound effect on resultant light stability. In order to examine this effect the x-ray crystal structures of the two forms were measured and examination of the arrangement of the molecules in the crystal lattice permitted a plausible explanation of the observed effects.

Views of the crystal structures are given in Figures 9 (form I) and 10 (form II). In Figure 9 the acrylate double bonds in neighboring molecules in the unit cell are well aligned to undergo a light catalysed cycloaddition reaction in the solid state to (4). In Figure 10, however, the double bonds are too remote for such a reaction to take place in the solid state. Therefore the absorption of light energy in form II promotes another chemical process which initiates principally the formation of polymers in far lower amounts than the formation of (4) in the case of form I.

Figure 9 Crystal structure of (2) form I.
Figure 10 Crystal structure of (2) form II.
0 0

Post a comment