Stationary phase computer modeling

Analytical shape computation techniques were applied for the detection of cavities and the calculation of molecular surface properties of isolated cavity features and other ordered formations within these resultant alkyl stationary-phase simulation models [227]. Deep cavities (8-10 A wide) within the alkyl chains were identified for Cig polymeric models representing shape selective stationary phases (Figure 5.23). Similar-structure cavities with significant alkyl-chain ordered regions (>11 A) were isolated from two independent Cig models (differing in temperature,... 

Pyrolysis-Gas Chromatography-Mass Spectrometry. In the experiments, about 2 mg of sample was pyrolyzed at 900°C in flowing helium using a Chemical Data System (CDS) Platinum Coil Pyrolysis Probe controlled by a CDS Model 122 Pyroprobe in normal mode. Products were separated on a 12 meter fused capillary column with a cross-linked poly (dimethylsilicone) stationary phase. The GC column was temperature programmed from -50 to 300°C. Individual compounds were identified with a Hewlett Packard (HP) Model 5995C low resolution quadruple GC/MS System. Data acquisition and reduction were performed on the HP 100 E-series computer running revision E RTE-6/VM software. 

Figure 12.4 Mastering column distribution design. On the left-hand side are shown computational fluid dynamic modeling results, and on the right-hand side are displayed pictures of the stationary phase cross section after an experiment with a dye (flow goes from the bottom to the top) (a) without a distributor, and (b) with a correctly designed distributor.

A computational procedure for the modeling of chromatographic separation of racemic Co(acac)3 into enantiomers on a dinitrobis(arginine)cobalt(III) complex as a chiral selector was described. Predicted elution order calculated from the differences in total energy of interaction for A and A selectands were found to be in agreement with the experimental results. The predictive power of the method and its possible practical applications in designing efficient chiral stationary phases was demonstrated . ... 

Automation allows batch chromatography to be run as a continuous process. Multiple injections using a separate pump and fraction collection provide an opportunity for continuous unattended operation. In iso-cratic separations, sample injection is often made before previously injected product elutes from the column, thus reducing cycle time and solvent consumption. Continuous and automated processes are always used with smaller columns and lower amounts of expensive enantioselective stationary phases. One of the future goals for modern PHPLC optimization would be the creation of software that would allow computer simulation modeling of nonlinear effects in preparative chromatography. 

Most RP-HPLC separations are done in the iso-cratic mode (i.e., where the composition of the mobile phase is held constant during the analysis). This approach is suitable when the sample consists of analytes having similar properties or where their hydrophobic-ities encompass a small or moderate range. Under these conditions, all solutes in the sample will be eluted over a reasonable time span (i.e., not too short to prevent resolution of individual analytes and not too long to result in an inconvenient analysis period). Therefore, proper selection of the mobile-phase composition is essential in the development of any re-versed-phase separation method. Fortunately, due to the decades of long practice of RP-HPLC, there exists in the literature and from commercial sources, a wealth of information on suitable mobile-phase compositions for particular types of sample, especially for the Cig stationary phase. In addition, the retention of solutes on hydrophobic phases has been modeled mathematically and there exist computer programs for assisting in the optimization of mobile-phase composition in the solution of various separation problems. 

Prediction of Elution Profiles (Linear Equilibrium). For the case of local linear equilibrium (infinite rate of mass transfer), Lapidus and Amundson (25) derived equations for computing concentration distributions in a packed column. With concentrations at the inlet of the column, and initial conditions throughout the column known, concentration profiles at a specific distance from the column inlet can be computed. The derivation was based on a semi-infinite column, which differs mathematically from a finite column, in that effects of the mobile phase leaving the stationary phase are not modeled. Nonetheless, the solution obtained is useful for giving a qualitative picture of important parameters in column performance. The equation is ... 

Another report on ion-pair chromatography, where computed molecular descriptors from molecular modeling were nicely correlated with experimental separation factors, was published by Karlsson, Luthman, Pettersson and Hacksell [81]. They examined factors responsible for separation of aminotetralins on achiral stationary phases in the presence of the chiral additive N-benzyloxycarbonylglycyl-L-proline (L-ZGP), a protected peptide derivative. 

These computational studies are comparable to those described in the section covering Type I CSPs. Experimentally the only difference between these separations and those above is that here the selectors are not stationary phases but rather are co-additives that form the diastereomeric complexes. Because no computational studies on type IV CSPs exist, molecular modeling of inorganic coordination complexes directed towards rationalizing enantioselective binding and chiral recognition presents itself as a ripe area for exploration. 

Type V CSPs are protein phases. Because of the well established chemo- and stereospecificity of enzymes, a large number of experimentalists have adapted proteins in one form or another as stationary phases for chiral separations. The intermolecular forces responsible for analyte binding to these biopolymers are the same as for most other CSPs but the size and complexity of proteins makes them difficult to study computationally. One would think that with approximately 400 entries in the Brookhaven Protein Databank to select from, separation scientists would have used one of these proteins as a chiral selector and then use those atomic coordinates to carry out molecular modeling studies. Only one example has appeared in the literature where information from the PDB has been used to serve as a beginning point for molecular modeling of a protein CSP. In all other examples the CSP is viewed as having an unknown structure and Quantitative Structure-Enantioselective Retention Relationships (QSERRs) have been carried out. 

Window diagram Window diagrams, developed by Laub and Purnell for optimizing the composition of mixed stationary phases in gas chromatography, can be used for optimizing mobile phase composition in LC. From two initial experiments (if a linear relationship is assumed between log k and mobile phase composition) or more (in the case of a quadratic relationship), the retention models are calculated for all solutes, and a response function (selectivity between every possible pair of solutes) is calculated and plotted versus the mobile phase composition. Areas or windows in which all solutes are separated can be located graphically. No particular effort of computation is required in such a procedure. 

After selecting a stationary phase and the mobile-phase components, several isocratic experiments are required to build a retention model. A computer-assisted multivariate procedure is often used to find the best combination of the working parameters.Separation selectivity is often monitored by maximum resolvable components in the shortest time at separation of multicomponent samples. 

---