Improving the signal-to-noise ratio by cross-correlation in flow injection analysis and high-performance liquid chromatography

Electrochemical and ultraviolet detection are two detection techniques used most often in flow injection analysis (FIA) and high performance liquid chromatography (HPLC). When these detectors are operated below a signal-to-noise ratio (S/N) of three, the precision of replicate determinations exceeds ten percent relative standard deviation (RSD). This dissertation demonstrates how cross correlation was used to improve the S/N in real-time at a fixed time delay (zero seconds) in both FIA and HPLC. A flow injection (FI) response below a S/N of three is cross correlated with another FI response of very high S/N. In one case, both FI responses are generated simultaneously by operating the FI systems in parallel and then cross correlating their outputs using an analog correlation circuit. In another case, a computer is used to digitize and store a high S/N FI response. After each injection, the high S/N FI signal is cross correlated with the injection response using the analog correlation circuit. The dynamic range of the amperometric detector was extended an order of magnitude after cross correlation. The lowest standard of dopamine determined after correlation was 0.12nM. The S/N at this level was 130, an improvement of 28 over the direct detector response. The detection limit at S/N = 3 was 3 $\times$ 10$\sp{-12}$M dopamine, corresponding to 63fg injected. When blank solutions were injected into a flowing stream injection peaks were observed. A novel injection technique is implemented in this work, which positions a valve downstream from the sample plug and separates the sample response from the injection peak. The use of this injection technique with cross correlation allowed detection of dopamine at the 0.84nM level at a S/N = 92, which corresponds to a detection limit for dopamine of 0.03nM. The increase in detection limit using this technique is due to increased dispersion in the flow system and increased pump noise due to the use of peristaltic pumps. Cross correlation has also been used to improve the S/N in the HPLC separation of catecholamines. Cross correlation resulted in an improvement in S/N 20 times better than that obtained using low pass filtering