Kelly Swinney

Detection in capillary electrophoresis.

A review of the four major, on‐line, capillary electrophoresis (CE) detection modalities is presented. It is shown that each detection method, fluorescence, absorbance (conventional and nonconventional), electrochemical and refractive index, have distinct advantages and limitations when applied to analysis in a CE format. Various aspects of CE detection are considered and a perspective regarding the applicability of the technique is provided. It is shown that because of widely varying detection limits (ranging from single molecule to 10—5 M) and detection scheme complexity, the particular application should dictate the selection of detection methodology in CE.

Quantification and evaluation of Joule heating in on‐chip capillary electrophoresis

We present the use of a novel, picoliter volume interferometer to measure, for the first time, the extent of Joule heating in chip‐scale capillary electrophoresis (CE). The simple optical configuration for the on‐chip interferometric backscatter detector (OCIBD) consists of an unfocused laser, an unaltered silica chip with a half‐cylinder channel and a photodetector. Using OCIBD for millidegree‐level noninvasive thermometry, temperature changes associated with Joule heating (≥2.81°C above ambient) in on‐chip CE have been observed in 90 νm wide and 40 νm deep separation channels. The temporal response of Joule heating in isotropically etched channels was exponential, with it taking an excess of 2.7 s to reach equilibrium. Buffer viscosity changes have also been derived from empirical on‐chip thermometry data, allowing for the determination of diffusion coefficients for solutes when separated in heated buffers. In addition, OCIBD has allowed the reduction in separation efficiency to be estimated in the absence of laminar flow and due to increased molecular diffusion and lower buffer viscosity. A 7% reduction in separation efficiency was determined for a high current drawing buffer such as Tris‐boric acid under an applied field of just 400 V/cm. Results indicate that heating effects in on‐chip CE have been underestimated and there is a need to readdress the theoretical model.

A Review of CE Detection Methodologies

Presented here is a comprehensive description of the four major detection modalities (absorbance, fluorescence, electrochemical, and refractive index) as well as a brief discussion of the less used detection techniques for CE. The focus of this manuscript is limited to direct, online detection techniques; therefore, detection via mass spectroscopy has not been included. The most sensitive on-line detection scheme for CE is laser-induced fluorescence detection, which is capable of single molecule detection but most often requires chemical derivatization. For CE applications needing universal detection, refractive index, conductivity detection and to some degree UV-Vis systems using the low UV (<210 nm) are suitable. In general, the detection scheme of choice is application specific and dictated by the level of sensitivity necessary (Table 1). While great strides have been made toward improved detection methodologies, if CE is to become a routine, robust and widely accepted analysis method, improvements in detection technology are still needed.

Ultrasmall volume refractive index detection using microinterferometry

A microinterferometric backscatter detector (MIBD) has been developed to perform subnanoliter volume refractive index measurements using a simple, folded optical train based on the interaction of a laser beam and a fused silica capillary tube. Positional changes of the interference pattern extrema (fringes) allow for the determination of Δn at the 10−7 level, corresponding to 5.3 pmole or 0.48 ng of solute, when thermal noise is controlled at 8×10−3 °C. MIBD is relatively path-length insensitive for capillaries ranging in inner diameter from 75 to 775 μm, allowing a large range of detection volumes, from 350 pL to 40 nL, to be produced. A theoretical model of the microinterferometric backscatter detector has also been developed and evaluated and has been found to be in agreement with experimental data. This model indicates increased sensitivity of the instrument as the wavelength of the probe beam and the wall thickness of the capillary tube are reduced.

Universal detection in capillary electrophoresis with a micro-interferometric backscatter detector

An optically simple, inexpensive, micro-volume refractive index detector was applied to capillary electrophoresis (CE), allowing universal solute detection at the sub-picogram level. The micro-interferometric backscatter detector (MIBD) employs direct, side illumination of an unmodified capillary by an He–Ne laser, producing a 360° fan of scattered light that contains a set of high contrast interference fringes. These light and dark spots are viewed on a flat plane in the direct backscatter configuration. A slit–photodetector assembly accomplishes signal interrogation of the time-dependent fringe shifts, produced or imparted by refractive index (RI) changes. Using an unfocused laser beam to prove the unmodified separation capillary produces a detector volume of 4.7 × 10–9 L. The separation and quantification of a mixture of organic dyes and simple sugars demonstrate the system’s utility. Submicromolar concentration detection limits of 0.46, 1.1 and 0.72 µM for Bromothymol Blue, Thymol Blue and Bromocresol Green, respectively, are achievable with CE-MIBD in the simplest configuration. The 3ς RI concentration detection limits are 2.5 times superior to those obtained by UV/VIS detection performed under the same conditions. Several carbohydrates (maltose, lactose and D-ribose) are separable and detectable at the ppm level, using no active thermal stabilization. Further demonstrating the utility of MIBD for universal detection with CE.

Noninvasive picoliter volume thermometry based on backscatter interferometry

Using the on‐chip refractive index (RI) detector based on backscatter interferometry, sensitive, small volume, noninvasive thermometry can be performed. The current optical configuration for the on‐chip interferometric backscatter detector (OCIBD) is quite simple and consists of an unfocused laser, an unaltered chip with a hemispherical channel and a photodetector. Alignment is straightforward with the only requirement being that the beam fully fills the channel. The interaction of an unfocused laser beam with the uncoated etched channel with a curvature within the silica plate (chip) produces fringes whose positional changes scale with respect to the refractive index (RI), n, of the fluid in the channel. Due to the inherently high value of dn/dT for most fluids and the high sensitivity of OCIBD to RI changes, the measurement of small temperature variations in sub‐nanoliter volumes is possible. Performing OCIBD with a 75 νm diameter laser beam on a silica chip that contains an etched channel with a 40 νm radius facilitates noninvasive thermometry on a N‐(2‐hydroxyethyl)piperazine‐ (2‐ethanesulfonic acid) (HEPES) solution in a 188×10–12L probe volume with a temperature resolution of 9.9×10–4 °C, at the 99% confidence level.

Micro-interferometric backscatter detection using a diode laser

Micro-interferometric backscatter detection (MIBD) is performed with a simple, folded optical train based on the interaction of a diode laser beam and a fused silica capillary tube allowing for refractive index (RI) determinations and detection of optically active molecules in small volumes. Side illumination of the capillary by a laser produces a 360 fan of scattered light that contains two sets of high contrast interference fringes. These light and dark spots are viewed on a flat plane in the direct backscatter configuration. Signal interrogation for polarimetry is based on quantifying the relative intensities (depth of modulation (DOM)) of adjacent high frequency (HF) interference fringes for polarimetry and relative fringe position for RI detection. Positional changes of the interference pattern extrema (fringes) allow for the determination of 1n at the 10 7 level or 5.3 pmol or 0.48 ng of solute. The MIBD-RI detection volume is just 5.0 nl. DOM changes allow for optical activity detection limits of 5.7 10 5 (mandelic acid, [a] 23 = 153, and D-glucose, [a] 25 = +52.5), and a 2 detection limit of 7.5 10 4 M (D-glucose) and 1.14 10 3 M (R-mandelic acid). The probe volume of MIBD-polarimetry was 38 nl, and within the probed volume at the limit of detection, about 28.7 pmol of mandelic acid or about 43.7 pmol of D-glucose is present. Furthermore, DOM (polarimetry signal) is unchanged when a non-optically active solute is interrogated by the MIBD-polarimeter. Finally, an optical model was derived and used to evaluate the advantages and pitfalls of using diode laser for MIBD. ©1999 Elsevier Science B.V. All rights reserved.

A chip-scale universal detector for electrophoresis based on backscattering interferometry.

An on-chip detector based on backscatter interferometry has been developed to perform sub-nanoliter volume refractive index measurements. The detection system consists of a simple, folded optical train based on the interaction of a laser beam and an etched channel in a silica (glass) plate. This etched channel is composed of two radii joined by a flat portion which define a curved surface in the shape of a half cylinder in a silica (glass) plate. The backscattered light from the channel takes on the form of a high contrast interference pattern that contains information related to the bulk properties of the fluid located within the probe volume. Positional changes of the interference pattern extrema (fringes) allow for the determination of refractive index changes at the 10(-6) level in a detection volume of 188 x 10(-12) L. Under capillary electrophoresis (CE) conditions, the injected mass detection limits for small molecules with little native absorption ranges from 530 fmol (0.18 ng) for sucrose to 720 fmol (0.43 nanograms) for raffinose. Fluorescein was also used to evaluate the technique for universal CE and under further optimized conditions can be quantified at the 150 microM level. Separation performance for the solutes tested ranged from about 2300 to 15,500 plates or 61,000 to 400,000 N m-1. The results presented here indicate there is potential for using the simple optical train of backscattering interferometry for on-chip universal solute analysis.

Capillary-scale polarimetry for flowing streams

A micro-polarimeter with a 40 nL probe volume was configured so that it is compatible with capillary-scale flowing stream analysis. The optical configuration consists of two polarizing optics, a capillary, a laser source and a photodetector which is very simple to configure with low cost components. This unique polarimeter is based upon the interaction of a linearly polarized laser beam and a capillary tube, in this case one with an inner diameter of 250 microns. Side illumination of the tube results in a 360 degrees fan of scattered light, which contains a set of interference fringes that change in response to optically active solutes. Solutes that exhibit optical activity are quantifiable and are detected by analyzing the polarization state of the backscattered light. The ability of the instrument to make extremely sensitive optical activity measurements in flowing streams is shown by the determination of (R)-mandelic acid, with a detection limit of 66 x 10(-6) M (507 x 10(-12) g), and the non-optically active control, glycerol. Additionally, the detector was configured to minimize refractive index perturbations.

D-β-Hydroxybutyrate Reaction Kinetics Studied in Nanoliter Volumes Using a Capillary Polarimeter

A laser-based capillary polarimetric detector (CPD) has been configured to allow for the analysis of optically active molecules at physiological temperatures and within a 40 nL sample volume. In the CPD, interference fringes contained within a 360° fan of scattered light result from the interaction between a polarized laser beam and a 240 μm i.d. capillary tube. Here it is shown that these fringes change in a reproducible manner with solute optical activity, facilitating a 3σ detection limit of 1.7 × 10−3 M d-β-hydroxybutyrate corresponding to 68 × 10−12 mole (7 × 10−9 gram) of solute. Further, with the use of this improved and novel polarimeter, the noninvasive study of enzyme reaction kinetics for d-β-hydroxybutyrate (50 mg/dL) and hydroxybutyrate dehydrogenase (100 μL of 50 units/L) has been evaluated for the first time in the absence of NAD. The rate constant for this reaction was found to follow first-order kinetics and determined to be 2.2 × 10−2 s−1. The prospect of directly studying the influence of enzyme cofactors on reaction kinetics is at hand, and the observations presented suggest that the application of monitoring insulin therapy for patients experiencing ketoacidosis is a tractable problem for the CPD.