Search

The preparation and evaluation of packed columns for CEC with porous and non-porous stationary phases with particle diameters between 1.5 μm and 5 μm is described. They are compared with respect of achievable efficiency, magnitude of EOF, dependence on buffer pH and concentration as well as on organic modifier content. No general correlation between chromatographically measured surface silanol concentration and magnitude of EOF could be found. Problems in reproducibility of retention in CEC may arise from in-column heat generation and overloading problems. The strong dependence of column efficiency on volume overloading is demonstrated. The use of peak area measurements for quantitative analysis is recommended. The application of CEC to the separation of herbicides and DNPH derivatives is demonstrated.

A general method for noncompetitive immunoassay of small analytes using affinity probe capillary electrophoresis (APCE) is demonstrated using digoxin as a model analyte. A uniform immunoreagent was prepared from a single-chain antibody (scFv) gene specific for digoxin. Site-directed mutagenesis introduced a unique cysteine residue for uniform labeling with a thiol-reactive fluorochrome. After expression in E. coli, the scFv was purified by immobilized metal affinity chromatography (IMAC) using an added C-terminal 6-histidine sequence. The protein was renatured and labeled while immobilized on the IMAC resin. After 0.02-microm filtration to remove microaggregates, the resulting reagent was highly uniform and stable at -12 degrees C for at least 1 year. Three formats of APCE using the scFv reagent were explored. A "mix-and-inject" assay optimized for low detection limits demonstrated analysis of 10 pM digoxin in aqueous standard solutions in 10 min. A rapid mix-and-inject format in a short capillary allowed detection of 1 nM digoxin in 1 min. Digoxin samples in serum and urine were injected directly after 10-fold dilution. In combination with solid-phase extraction, 400 fM digoxin was detected in 1 mL of serum. Including solid-phase extraction, reproducibility was within 2.5%, and the linear range was 3 orders of magnitude. The strategy adopted in this paper should be of general use in the low-level analysis of small analytes.