Dynamic optical property changes: implications for reflectance feedback control of photocoagulation.

During laser treatment, coagulation affects the optical properties of the tissue. In particular, the formation of a white lesion significantly increases the scattering coefficient. This change in the optical properties in turn affects the laser light distribution in the tissue. The white lesion formed during photocoagulation of the retina has a dynamic effect upon reflection and fluence rate. This problem has been simulated on a model medium consisting of a thin absorbing layer covered with a 1 cm thick layer of albumin. The albumin layer is subdivided into coagulated (white) and uncoagulated (clear) layers. The optical properties of each layer have been determined and these values have been used to model light distribution in the medium. One-dimensional adding-doubling and three-dimensional Monte Carlo methods have provided light distributions in the medium for varying thicknesses of the coagulated albumin. Computed fluence reaching the absorbing layer decreased in the presence of a 275 microns or thicker coagulated layer. The coagulated layer attenuates light because it is highly scattering; however, this scattering also leads to a sub-surface peak in fluence rate at a level higher than the incident fluence. The latter effect outweighed the former for coagulated layer thicknesses less than 275 microns. Computed reflectance of argon laser light from a semi-infinite coagulated region initially increased linearly as a function of thickness. As the coagulation thickness increased beyond 4-5 optical depths, the reflectance approached a constant value, R infinity, at 9 optical depths (2 mm). Experimentally measured total reflectance is shown to be an inadequate indicator of the thickness of a lesion (finite coagulated volume); however, central reflectance from a lesion measured with a CCD camera confirmed the computed trends. These results provide a theoretical foundation for control of lesion thickness using reflectance images.     

Improved chiral stationary phase for ß-blocker enantioseparations

The previously described -Burke 1 chiral stationary phase (CSP) was designed for the chromatographic separation of the enantiomers of ß-blockers. Difficulties with the reproducibility of the free radical addition reaction, used in the attachment of the chiral selector to the chromatographic support, have required the development of an alternative silane immobilization process (-Burke 2 CSP). While the enantioselectivity afforded by this new CSP is generally equivalent to that of the original CSP, the -Burke 2 CSP demonstrates longer analyte retention, necessitating the use of mobile phases of greater eluotropic strength. The increased retention of the new CSP presumably results from a greater surface density of functional selectors, an interpretation which is supported by the observation that the preparative capacity of the -Burke 2 CSP is greater than that of the original. Some of the factors influencing the retention and separation of a group of 23 ß-blockers on the -Burke 2 CSP are discussed.     

New chiral crown ether stationary phase for the liquid chromatographic resolution of alpha-amino acid enantiomers

A new chiral stationary phase (CSP) for the liquid chromatographic separation of enantiomers was prepared by bonding a novel enantiopure (diphenyl-substituted 1,1'-binaphthyl) crown ether to 5 microm silica gel. The resulting CSP was applied to the separation of the enantiomers of various natural and unnatural alpha-amino acids. All alpha-amino acids tested were resolved very well on the new CSP, with the exception of proline, which does not contain a primary amino group. The resolution of alpha-amino acid enantiomers on this new CSP was found to be dependent on the type and amounts of organic and acidic modifiers, and on column temperature.     

Efficient synthesis of NK(1) receptor antagonist aprepitant using a crystallization-induced diastereoselective transformation

An efficient stereoselective synthesis of the orally active NK(1) receptor antagonist Aprepitant is described. A direct condensation of N-benzyl ethanolamine with glyoxylic acid yielded a 2-hydroxy-1,4-oxazin-3-one which was activated as the corresponding trifluoroacetate. A Lewis acid mediated coupling with enantiopure (R)-1-(3,5-bis(trifluoromethyl)phenyl)ethan-1-ol afforded a 1:1 mixture of acetal diastereomers which was converted into a single isomer via a novel crystallization-induced asymmetric transformation. The resulting 1,4-oxazin-3-one was converted via a unique and highly stereoselective one-pot process to the desired alpha-(fluorophenyl)morpholine derivative. Interesting and unexpected [1,2]-Wittig and [1,3]-sigmatropic rearrangements were identified during the optimization of these key steps. In the final step, a triazolinone side chain was appended to the morpholine core. The targeted clinical candidate was thus obtained in 55% overall yield over the longest linear sequence.