Conductivity Detection

The conductivity detector, in a direct contact mode, first appeared in the literature as a means of detection for use with capillary isotachophoresis (CITP) systems in the 1970s. Development of this detector was necessitated by the need for a method that had high spatial resolving power, unlike thermometric detection techniques commonly used in CITP. Everaerts and colleagues sandwiched a 50 µm sheet of Terylene polymer film, with a 100 µm thick layer of a metallic foil on each side, between two pieces of Kel-F, in a butt-connected manner to ensure little fluidic leakage. The plastic Kel-F pieces were then fitted into a brass capillary holder. The center of the Terylene piece and the Kel-F pieces were then drilled out to match the inner diameter of the capillary used, to provide a uniform channel. In this configuration, the metal foil deposited on each side of the

Terylene served as a sensing electrode, while the width of the Terylene served to define the detection gap between the electrodes. In the initial proof of concept experiments, both

alternating current (AC) and direct current (DC) modes were tested. However, it was found that in the DC mode, polarization of the electrode surface made definition of zone boundaries impossible. It was also noted that drops in the potential gradient occurred due to gas formation at the surface of the electrodes, further increasing the noise in the system. In the AC mode, it was observed that gas formation on the surface of the electrodes did not occur to a significant degree, despite having potential gradients of 50 mV along the length of the electrodes. It was noted, however, that absorption of hydrogen into the metallic layer decreased the sensitivity of the detection method over time.25

One method for circumventing the problems associated with direct contact measurements is to use the conductivity detector in a contactless mode. This was first developed for CITP in the 1980s as a means of detection that did not suffer from the irreproducibility and electrode fouling that commonly occurs with direct contact measurements. Gas and co-workers used a four electrode setup for detection, wherein a high frequency signal is applied to two electrodes and detected by the remaining two electrodes at the output end of a capacitive cell. The electrodes were made of 200 µm copper enameled wire placed in an equiplanar arrangement around the capillary. A high frequency signal was applied to each excitation electrode, with the signals being 180° out of phase. An operational amplifier set in differential mode, using the two detection electrodes, serves as an amplifier prior to rectification and filtering. While it was

determined that the baseline stability and reproducibility of the detection was much better than that seen with a direct conductivity detector, it was also observed that the spatial resolution of contactless conductivity detection was poorer than that seen in the contact mode.26

Conductivity detection was first applied to capillary zone electrophoresis systems in an on-column manner in the late 1980s, by Huang, et al. This initial attempt at on-column detection was performed in the contact mode by placing 25 µm (o.d.) platinum wires into 40 µm (i.d.) holes drilled into the wall of a fused silica capillary directly opposite each other. The holes were placed such that the potential difference between them due to the electrophoretic separation field was minimized, lessening any electrochemical effects from the high separation field. The wires, or electrodes, were held in place using epoxy and the entire cell was sealed in a Plexiglas jacket. Using this setup, the detection volume was determined to be 30 pL, with detection of Li+ exceeding three orders of magnitude with good linearity.27 This setup was also used to quantify Li+ in human serum and provide a means for detection of low molecular weight carboxylic acids.28-30 However, the construction of the detector made this technique very difficult to implement, and as a result, it never found widespread use.

In 1998, two research groups independently developed a contactless conductivity detector for use with capillary electrophoresis. da Silva and do Lago developed a detection cell by painting two 2 mm silver electrodes onto the outside of a fused silica capillary with a gap of 1 mm. A high-frequency signal was then applied to one electrode and detected at a subsequent electrode, where information about the solution impedance was collected. Separation of simple inorganic cations was possible, with a limit of detection of 1.5 µM for Li+.31 Zemann and co-workers developed an identical painted electrode system; however, they also used 15-50 mm long electrodes made from syringe cannulas in a similar detection scheme. As all of these electrodes were cylindrical in nature and placed axially rather than radially across the capillary, decreased sensitivity

from shorter path lengths was not of concern. One added advantage of using cylindrical electrodes is the ease of placement. The electrodes can be placed anywhere along the length of the capillary and the placement can be changed quickly. The detection of inorganic cations and anions was demonstrated with this detector and shown to have limits of detection on the order of 200 ppb for sodium and chloride over a dynamic range that was more than three orders of magnitude.32

Since the introduction of the contactless conductivity system by both groups, there have been numerous papers in the literature dealing with the development and

optimization of the detection electronics,33-35 electrode configuration,36-41 and cell geometry.42, 43 There have also been papers dealing with the overall fundamental aspects of the detection method,43-46 including capillary inner diameter,47, 48 excitation frequency and voltage,33, 49-51 and even high voltage excitation.52, 53 As was discussed earlier, conductivity detection is a concentration sensitive detector, and as a result, there have also been a number of accounts where detection has occurred on miniaturized

platforms,54-66 as well as in conjunction with various other separation techniques such as anion exchange,67 _{reverse-phase ion pair microcapillary electrochromatography,}68 _size- exclusion electrokinetic chromatography (SEKC),69 and high performance liquid chromatography (HPLC)70 to name just a few. Work presented in this dissertation will include further development and optimization of experimental parameters, including a thorough investigation of sources of noise and the effect of choice of buffer system on the overall noise.