Experimental

2.3.1Materials and Buffers

All materials were used as shipped without further modification. 2-

morpholinoethansulfonic acid (MES), 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES), 1,4-piperazinediethanesulfonic acid (PIPES), 3-(cyclohexylamino)-1- propanesulfonic acid (CAPS), 2-(cyclohexylamino)-ethanesulfonic acid (CHES), 3-(N- morpholino)-propanesulfonic acid (MOPS), L-Arginine, L-Histidine, propionic acid, hydroxybenzylamine, and borax were all obtained from Sigma (St. Louis, MO). Sodium hydroxide solution (NaOH), hydrochloric acid (HCl), sodium chloride (NaCl), and lithium chloride (LiCl) were all obtained from Fisher Scientific (Fair Lawn, NJ).

Potassium chloride (KCl) was obtained from Mallinckrodt (Paris, NY). All solutions were prepared using deionized water filtered through a Barnstead Nanopure Filtration System (Boston, MA). Only buffer solutions involved in pH studies were adjusted using either sodium hydroxide or hydrochloric acid; otherwise buffer solutions were at their native pH. Borate solutions were prepared from borax. Normal capillary electrophoresis buffer solutions were 20 mM MES/20 mM Histidine at pH 6.1. All other buffer solutions were prepared at a concentration giving a similar background conductivity value as the normal solution, which was determined experimentally. All solutions were filtered using a 0.2 µm nylon membrane filter from Grace Davidson (Deerfield, IL) and vacuum degassed using a Cole Parmer 8891 Ultrasonic Bath (Vernon Hills, IL).

2.3.2Capillary Electrophoresis System

Capillary electrophoresis was performed using an untreated fused silica capillary (Polymicro Technologies, Phoenix, AZ) 68 cm in length. The outer diameter (o.d.) of all capillaries used was 360 µm. The inner diameter (i.d.) for all studies, with the exception of the capillary inner diameter study, was 50 µm. Electrophoresis was driven by a reversible 30 kV power supply (Spellman High Voltage Electronics, Inc., Plainview, NY). The power supply current was limited using an in-house built current delimiting circuit with a resistance value of 111 MΩ. The electrophoresis current was monitored by measuring the voltage drop across a 220 kΩ resistor at the outlet end of the

electrophoresis system, which was connected to ground. Injections were performed electrokinetically at 2 kV for 4.5 s, unless otherwise stated.

2.3.3Contactless Conductivity Detection System

The entire experimental setup is shown in Figure 2-2. A DS335 digital function generator (Stanford Research Systems, Sunnyvale, CA) was used to supply the AC excitation signal. Typically, the function generator was set to output 20 Vp-p at 100 kHz; however, these values were changed in studies where the effect of the excitation signal was investigated. The excitation signal was applied to the conductivity cell via a metal electrode. Electrodes were made from 23 gauge, 304 stainless steel hypo tubing (Small Parts, Miami Lakes, FL) with the center drilled out to give an inner diameter of 390 µm. Electrodes 5 mm in length were used unless otherwise specified.

As described previously, the electrodes were soldered onto a printed circuit board (ExpressPCB, Redwood City, CA), which defines the dimensions of the conductivity cell.4 Conditioning of the electrode using soap, water, and 6 M Hydrochloric acid is necessary to ensure sufficient adhesion of the electrode to the PCB. Placement of the electrodes was accomplished by feeding a capillary through each electrode, with a 1 mm spacer in between. Once the electrodes were properly soldered to the PCB, the capillary was removed and the 1 mm spacer taken out, giving a 1 mm air gap between the

electrodes. The excitation signal was fed into the PCB using a crimp terminal and pin header (DigiKey Electronics, Thief River Falls, MN).

Once the signal progressed through the detection cell, an OPA 602 operational amplifier (Texas Instruments, Dallas, TX) was used in a current to voltage configuration for normal runs. This op amp was chosen for its low noise characteristics (0.6 fA / √Hz). A 1 MΩ feedback resistor (Multicomp, Chicago, IL) was used to provide an

was chosen due to its temperature stability (50 ppm / °C) and precision (1%). The op amp was used under gain peaking conditions where the signal peaked at 400 kHz before experiencing any rolling off, as shown previously.4 Power was supplied to the op amp by connecting four 6 V batteries in series. The series was grounded between the second and third batteries to give ± 12 V at the output ends. Power and ground were connected to the PCB via a ten pin header (Newark Electronics, Chicago, IL). To remove any high

frequency oscillations in the feedback loop of the operational amplifier, 1000 pF capacitors (Newark Electronics, Chicago, IL) were used between the power supply and conductivity electrodes. The entire detection cell was placed in a grounded metal box to provide additional shielding from outside noise sources.

An SR810 digital lock-in amplifier (Stanford Research Systems, Sunnyvale, CA) was used for signal isolation and amplification in all cases. The external reference setting was used and was supplied by a split signal from the DS335 function generator. The signal from the detection cell was input into the lock-in amplifier through the ten pin header previously discussed. Due to the frequency input limits of the lock-in amplifier, the 100 kHz signal was chosen (limit of 102 kHz). The lock-in was set with a 100 ms time constant and a 24 dB/octave slope, giving a bandwidth of 0.78 Hz. The sensitivity was chosen such that the maximum signal could be obtained. Line and 2x line filters were used to decrease any contribution to noise from external sources.

The output of the lock-in amplifier was digitized using a PCI-MIO-16XE-50 DAQ card (National Instruments, Austin, TX) and collected on a personal computer using custom software written in LabVIEW (National Instruments, Austin, TX). Data

analysis was performed using Igor Pro and included median filter baseline subtraction prior to analysis.

2.4Experimental Results and Discussion