A FILTER-TRUST-REGION METHOD FOR LC^1 UNCONSTRAINED OPTIMIZATION AND ITS GLOBAL CONVERGENCE

In this paper we present a filter-trust-region algorithm for solving LC1 unconstrained optimization problems which uses the second Dini upper directional derivative. We establish the global convergence of the algorithm under reasonable assumptions.     

A non-Marcus model for electrostatic fluctuations in long range electron transfer

We propose a new model for the elementary act of electron transfer between two species in solution. The central idea is that the solution in the immediate vicinity of each species may be represented by an equivalent circuit consisting of a Debye circuit shunted by a resistor. Based on this insight, we derive a new formula for the one-dimensional potential energy profile of a coupled donor–acceptor pair at finite (but large) separation d, along a charge-fluctuation reaction co-ordinate, at fixed radii of the transition states. The corresponding reorganisation energy of the reaction is also derived, and it is found to differ from that in the Marcus theory. In particular, the new model predicts that the reorganisation energy is independent of the static dielectric constant of the solution, whereas the old model predicts a strong dependence. The difference is traced to the fact that the Marcus theory omits consideration of the work required to form the charge fluctuations and focuses instead on the work required to localise the charge fluctuations. In general, the equivalent circuit approach permits many of the difficult-to-derive equations of non-equilibrium polarisation theory to be written down by inspection.     

Investigation of particle-based and monolithic columns for cation exchange protein displacement chromatography using poly(diallyl-dimethylammonium chloride) as displacer

The overall topic of the investigation was the separation of basic proteins by cation exchange displacement chromatography. For this purpose two principal column morphologies were compared for the separation of ribonuclease A and alpha-chymotrypsinogen, two proteins found in the bovine pancreas. These were a column packed with porous particles (Macro-Prep S, 10 microm, 1000 A) and a monolithic column (UNO S1). Both columns are strong cation exchangers, carrying -SO3(-)-groups linked to a hydrophilic polymer support. Poly(diallyl-dimethylammonium chloride) (PDADMAC), a linear cationic polyelectrolyte composed of 100-200 quaternary pyrrolidinium rings, was used as displacer. The steric mass action (SMA) model and, in particular, the operating regime and dynamic affinity plots were used to aid method development. To date the SMA model has been applied primarily to simulate non-linear displacement chromatography of proteins using low molar mass displacers. Here, the model is applied to polyelectrolytes with a molar mass below 20000 g mol(-1), which corresponds to a degree of polymerization below 125 and an average contour length of less than 60 nm. The columns were characterized in terms of the adsorption isotherms (affinity, capacity) of the investigated proteins and the displacer.     

Validation study of a method for assaying DE-310, a macromolecular carrier conjugate containing an anti-tumor camptothecin derivative, and the free drug in tumor tissue by high performance liquid chromatography/atmospheric pressure chemical ionization tandem mass spectrometry

DE-310 is a macromolecular carrier conjugate containing an anti-tumor camptothecin derivative, DX-8951, which is conjugated to a water-soluble polymer via a peptide spacer. Assay methods have been developed for the determination of a polymer-bonded DX-8951 conjugate, DX-8951, and Glycyl-DX-8951 concentrations in murine Meth A tumor tissue. Free DX-8951 and Glycyl-DX-8951 were extracted from tumor tissue homogenates by protein precipitation and analyzed by LC/MS/MS (method I). Conjugated DX-8951 was isolated by solid-phase extraction after digestion with a thermolysin. The productive phenylalanyl-glycyl-DX-8951 was analyzed by LC/MS/MS (method II). The lower limits of quantitation of DX-8951, Glycyl-DX-8951, and conjugated DX-8951 were 1.36, 1.34 and 73.7 ng/g (as DX-8951 equivalent). These two methods showed satisfactory sensitivity, precision and accuracy. To study the pharmacokinetics of DE-310, it would be of great help to assay the polymer-bonded DX-8951 and its released drugs in tumor tissue.     

Development of an Ion-pair HPLC method for the determination of 2,4- and 2,6-toluendiamine in human plasma

Summary An ion-pair HPLC method is presented for the determination in plasma of 2,4- and 2,6-toluendiamine (TDA), known carcinogens. The chromatographic conditions consisted in isocratic elution on a reversed phase C₁₈ column with 5 mM octanesulfonic acid in methanol-water, 45∶55, as mobile phase. UV detection was performed at 235 nm. Samples were analyzed after a simple single step liquid-liquid extraction and the method was validated by measurement of precision (interassay and intraassay), sensitivity, specificity, linearity, and recovery. The detection limit for both TDAs in plasma was 20 ng mL⁻¹. The correlation coefficients based on the intrassay calibration curve were 0.998 and 0.997 for 2,6- and 2,4-TDA, respectively. The intraassay accuracy, expressed in terms of recovery, was found to be up to 89.91% and 97.05% for 2,6- and 2,4-TDA, respectively.