Wavelength Modulated Back-Scatter Interferometry ... - ACS Publications


Wavelength Modulated Back-Scatter Interferometry...

0 downloads 45 Views 1MB Size

Subscriber access provided by UNIV OF DURHAM

Article

Wavelength Modulated Back-Scatter Interferometry for Universal, On-Column Refractive Index Detection in Picoliter Volumes Robert C Dunn Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b00771 • Publication Date (Web): 15 May 2018 Downloaded from http://pubs.acs.org on May 16, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Submitted, Analytical Chemistry

Wavelength Modulated Back-Scatter Interferometry for Universal, OnColumn Refractive Index Detection in Picoliter Volumes Robert C. Dunn* Ralph N. Adams Institute for Bioanalytical Chemistry University of Kansas 2030 Becker Drive Lawrence, KS 66047 Abstract Wavelength modulated back scatter interferometry (M-BSI) is shown to improve the detection metrics for refractive index (RI) sensing in micro-separations. In M-BSI, the output of a tunable diode laser is focused into the detection zone of a separation channel as the excitation wavelength is rapidly modulated. This spatially modulates the observed interference pattern, which is measured in the back-scattered direction. Using a split photodiode detector aligned on one fringe of the of interference pattern, phasesensitive detection is used to monitor RI changes as analytes are separated. Using sucrose standards, we report a detection limit of 700 µg/L in a 75 µm i.d. capillary at the 3σ level, corresponding to a detection volume of 90 pL. To validate the approach for electrophoretic separations, Na+ and Li+ were separated and detected with M-BSI and indirect-UV absorbance on the same capillary. A 4 mg/L NaCl and LiCl mixture leads to comparable separation efficiencies in the two detection schemes, with better signal-tonoise in the M-BSI detection, but less baseline stability. The latter arises in part from Joule heating, which influences RI measurements through the thermo-optic properties of the run buffer. To reduce this effect, a 25 µm i.d. capillary combined with active temperature control was used to detect the separation of sucrose, glucose, and lactose with M-BSI. The lack of suitable UV chromophores makes these analytes challenging to detect directly in ultra-small volumes. Using a 55 mM NaOH run buffer, M-BSI is shown to detect the separation of a mixture of 174 mg/L sucrose, 97 mg/L glucose, and 172 mg/L lactose in a 15 pL detection volume. The universal on-column detection in ultrasmall volumes adds new capabilities for micro-analysis platforms, while potentially reducing the footprint and costs of these systems.

*email: [email protected]

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Introduction The miniaturization of analytical separation platforms has dramatically reduced the time and sample required for analysis.1-4 Capillary electrophoresis (CE) and microchip CE systems (MCE) enable the separation of complex mixtures in seconds to minutes using nanoliter to picoliter sample volumes. These highly efficient separation platforms are widely applied in fields from the life sciences to environmental monitoring.14 While the miniaturized separation channels makes these techniques highly efficient, it also places increased demands on the detection system.5-8 Developing methods with improved sensing metrics that are compatible with ultrasmall probe volumes, therefore, is an active area of research.9-12 Mass spectrometry, electrochemical detection, and optical methods have all been adapted for detection in micro-separations.5-8 Optical methods are particularly convenient since most can be implemented in a standoff, on-column detection configuration that is easily integrated with CE and MCE separations. These methods are generally fast, inexpensive, non-destructive, and readily miniaturized for field deployable applications. Finally, the significant progress in light-emitting and laser diodes along with advances in detector technology continues to reduce the footprint and costs of these devices.13-14 While fluorescence detection has the most favorable detection limits, most analytes are not intrinsically fluorescent and thus require additional labeling steps. Ultraviolet (UV) absorbance provides a more general approach, but does not scale favorably as separations are miniaturized. Moreover, important analytes such as carbohydrates, lipids, small ions, and most amino acids lack suitable UV active chromophores, thus necessitating labeling or indirect detection schemes that further reduce performance and flexibility. Raman detection has also been coupled with microseparations and brings aboard the significant advantage of structural specificity. Conventional Raman detection is limited to millimolar concentrations, which has led to the development of methods based on surface enhanced Raman scattering (SERS). These methods can significantly improve detection limits, but lend themselves to postcolumn detection geometries.15-17 For on-column SERS detection, in situ growth of the SERS substrate or addition of silver colloids directly into the run buffer have been demonstrated.18-19 For the latter, fouling of the detection window degrades performance and limits the number of runs. Refractive index (RI) sensing is a universal approach for on-column analyte detection that does not require a chromophore. The RI quantifies the speed of light traveling through a given substance and is related to the electronic polarizability of the species. As such, all analytes affect the RI through this general mechanism and can, in theory, be detected in a separation as long as their RI differs from the background run buffer. Traditionally, deflection-based measurements have been used for RI detection, but these become problematic as detection volumes decrease. For small volume RI detection, surface plasmon resonance (SPR) and whispering gallery mode (WGM) sensing have been coupled with separations.12, 20-22 However, because of the metalized surface required in SPR and the physical size of WGM resonators, these methods are implemented in a post-column detection configuration for electrophoretic separations. For on-column RI detection in ultra-small volumes, interferometry has become the most popular approach.

ACS Paragon Plus Environment

Page 2 of 17

Page 3 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

RI detection using interferometry has been integrated with both capillary and microchip separations using various configurations.23-25 A coherent light source passed through a capillary or microchip channel will generate an interference pattern as light refracts and reflects from the various interfaces. A particularly convenient geometry for micro-separations is retroreflected26 or back-scatter interferometry (BSI)25, 27, where the interference pattern is imaged from the same side as the excitation source. Optical pathlength changes from analytes migrating through the detection zone, shifts the BSI fringe pattern, which is quantified through imaging the interference pattern or intensity measurements taken on a single fringe.10, 27 This approach scales favorably as dimensions are reduced and is readily integrated with miniaturized platforms.26-27 Unlike UV or indirect UV methods, BSI is readily integrated with microchip formats for universal detection. The utility of BSI detection in microchannels also extends beyond separations. Label-free BSI detection, for example, has been used to quantify proteinbinding interactions, probe membrane-ligand associations, and quantify the hydrodynamic radii of species in solution, although some have questioned the former.2833

The many advantages of RI detection, however, are tempered by modest sensitivities in ultra-small volumes. To help address this issue, we introduce a new approach for BSI that extends the detection limits and opens new applications in microfluidic detection. The approach scans the excitation wavelength to shift the BSI fringe pattern, providing a mechanism for modulating the back-scattered signal. Wavelength modulated back-scatter inteferometry (M-BSI) exploits the significant advantages inherent in phase-sensitive detection to improve RI detection in microseparations. The output of a tunable diode laser is modulated at 250 Hz and focused into the detection zone of a separation capillary. The central fringe in the back-scattered interference pattern is detected on a split photodiode and demodulated with a lock-in amplifier to detect shifts arising from sample RI changes. Using sucrose standards passed through a 75 µm i.d. capillary, we report a detection limit of 700 µg/L in a 90 pL volume. This is over a factor of 50 improvement in detection limits over similar BSI studies conducted using larger detection volumes.30 To validate the approach for electrophoretic separations, Na+ and Li+ are separated and detected with both M-BSI and indirect UV (i-UV) absorbance on the same 75 µm i.d. CE capillary. Finally, M-BSI is used to detect the separation of sucrose, glucose and lactose in a 25 µm i.d. capillary, with corresponding detection volume of 15 pL. These species are difficult to detect directly with other optical methods, illustrating the potential of M-BSI for universal detection in ultra-small volumes. Materials and Methods Chemicals: Reagent grade sucrose, glucose, lactose monohydrate, sodium chloride, lithium chloride, sodium hydroxide and imidazole were purchased from Fisher Scientific (Fair Lawn, New Jersey) and used without further purification. Stock solutions of sucrose, glucose, lactose monohydrate, sodium chloride, and lithium chloride were prepared at 1 g/L in ultra-pure water (Milli-Q Academic A10, Burlington, MA) and diluted to the appropriate concentrations. Run buffers of 5 mM imidazole and 55 mM NaOH were prepared in ultra-pure water and stored at 4oC. All solutions were equilibrated to room temperature and passed through a 0.22 µm filter (Fisher Scientific) prior to use.

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Experimental Apparatus: An in-house assembled CE using a Spellman CZE 1000R (Hauppauge, NY) high voltage source and custom designed pressure injection timing electronics was used for all studies. A Thermo Separations Products UV1000 (Fremont, CA) absorbance detector was added for the indirect detection of Na+ and Li+ in an imidazole run buffer. Both 25 µm and 75 µm i.d. (363 µm o.d.) fused silica separation capillary was purchased from Polymicro Technologies (Lisle, IL) and used at the indicated lengths. The capillaries were conditioned by flushing 100 mM NaOH through the capillary for 15 min followed sequentially by deionized water and run buffer. Following each separation, run buffer was flushed through the capillary for at least 5 min. After no more than five separations, the capillary was re-conditioned with 100 mM NaOH. For the separation of Na+ and Li+, the M-BSI detection zone of the capillary was mounted flush on a 4 cm x 4 cm x 2 cm copper block for passive temperature stabilization. For the separation of carbohydrates, active temperature control was implemented. The detection zone of the capillary was affixed to a 4 cm x 4 cm thermoelectric Peltier cooler mounted on the copper block heat sink. A thermistor was attached to the capillary mount with its output sent to a temperature controller (Stanford Research Systems LDC500, Sunnyvale, CA) driving the thermoelectric element. A PID feedback loop implemented by the temperature controller held the capillary mount at 18.000 ± 0.004 oC during the carbohydrate separations. For M-BSI detection, the output of a fiber coupled tunable diode laser centered near 635 nm (New Focus Velocity TLB 6700-LN, Irvine, CA) is passed through a lens (7 cm focal length), a 45o beam splitter (Chroma Technology, Bellows Falls, VT), and into the detection zone of the separation capillary. The retroreflected interference pattern from the capillary is directed out of the excitation path with the beamsplitter and sent to a split photodiode detector (EG&G Optoelectronics, Quebec, Canada). The excitation optics and photodiode detector are mounted on separated xyz manipulators (Linetool, Allentown, PA) for proper alignment. A selected fringe from the interference pattern is aligned on the split photodiode and the outputs are differentially amplified (Stanford Research Systems SR560, Sunnyvale, CA). The amplified output is sent to a dualphase lock-in amplifier (Stanford Research Systems SR830, Sunnyvale, CA), whose sinusoidal reference signal is used to sweep the wavelength of the tunable diode laser. For all experiments reported here, the laser modulation frequency and amplitude were kept constant at 250 Hz and 1.1 Vrms, respectively. The latter corresponds to sweeping the laser wavelength 0.1 nm. The R output from the lock-in amplifier (magnitude of the vector sum of the in-phase and quadrature signal components) is sent to a computer and recorded using a program written in LabView (National Instruments, Austin, TX). Results and Discussion Figure 1A schematically shows how BSI is traditionally coupled with microseparations.34 Using CE as an example, an unfocussed laser source is directed into the detection zone of the separation capillary. As the light traverses the capillary, each interface leads to reflection and refraction of the incident beam. Since a coherent sourced is used, this produces a scattered interference pattern around the capillary that can be modeled with wave-based and ray-tracing models.32, 35-37 Figure 1A, for example, shows the interference pattern measured in the back-scattered direction, where the detector is positioned on the same side as the excitation source.

ACS Paragon Plus Environment

Page 4 of 17

Page 5 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

The observed interference pattern is influenced by the RI of the fluid filling the bore of the capillary, forming the basis of RI sensing in BSI.38 Changes in the lumenal RI will alter the optical path length through the channel, shifting the interference pattern as shown in Fig. 1A. For separations, this leads to a reversible shift in the BSI pattern as analyte bands traverse the detection zone and transiently change the RI from the BGE level.25-26 The BSI geometry is particularly convenient for micro-separations, since the excitation and collection optics are positioned on the same side of the device. The detection optics are also easily miniaturized and the method is straightforward to implement and align for on-column detection. This, along with the universal nature of RI detection, makes BSI a compelling approach for integration with micro-separation platforms. Here, the BSI method is modified to improve detection metrics and streamline data collection and analysis. As indicated in Fig. 1A, changes in the excitation wavelength will also shift the interference pattern in BSI. This provides a straightforward route for improving RI detection metrics using wavelength modulation coupled with phase-sensitive detection. Modulated-backscatter interferometry (M-BSI), is shown schematically in Fig. 1B. The excitation wavelength from a tunable diode laser, centered at 635 nm, is modulated using a sinusoidal waveform. The frequency of the waveform determines how fast the laser is swept and the amplitude determines the wavelength range. As shown in Fig. 1B, the sine wave is generated by the reference signal of a lock-in amplifier and sent to the laser controller. For all experiments reported here, the frequency was set at 250 Hz with an amplitude of 1.1 Vrms, corresponding to a 0.1 nm sweep of the laser wavelength. In addition to wavelength modulation, the M-BSI approach shown in Fig. 1B differs slightly from traditional BSI in that the excitation is focused into the detection zone and the scattered light is detected in a retroreflected geometry.26 The back-scattered radiation from the tubular capillary produces a symmetric interference pattern, with a central spot flanked by a progression of smaller spots.26, 38 The central feature is the most sensitive to lumenal RI changes since the light forming this spot travels a pathlength equal to twice the inner diameter of the capillary.32 A segmented photodiode is aligned on the central spot with the outputs sent to a differential amplifier. The amplified output is sent to the same lock-in amplifier providing the reference sine wave driving the laser modulation. For all measurements, the R output from the lock-in amplifier was recorded. To characterize the RI response, sucrose solutions of increasing concentration were driven under pressure through a 75 µm i.d. capillary (363 µm o.d.). Figure 2 plots the M-BSI signal with the introduction of 23 mg/L (67.5 µM), 46 mg/L (135.0 µM), 65 mg/L (189.0 µM), 162 mg/L (472.5 µM), and 323 mg/L (945.1 µM) sucrose solutions. Each signal level change corresponds to the RI difference (∆RI) between pure water (RI = 1.331932) and the sucrose solution at 21.5oC.39-40 This difference ranges from 3.3x106 refractive index units (RIU) at 23 mg/L to 4.6x10-5 RIU at 323 mg/L. After the introduction of each sucrose standard, ultra-pure water was injected into the capillary to ensure the signal returned to the baseline level as shown in Fig. 2. The inset shows the calibration curve constructed from the measured level changes. At the particular settings used (modulation amplitude, lock-in settings, amplifier gain, etc.), a plot of RI versus lock-in output (volts) yields a sensitivity of 3x10-5 RIU/V. The standard deviation (σ) in the baseline signal is 1.1 mV, yielding a limit of

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

-7

detection (LOD) at the 3σ level of 1x10 RIU or 700 µg/L sucrose (2 µM) in a 90 pL detection volume. By way of comparison, recent BSI measurements using the conventional imaging approach reported a detection limit of 35 mg/L for glucose in a 100 µm i.d. capillary.30 While this is not a direct comparison, the 50-fold improvement here in a smaller detection volume does suggest that the M-BSI approach enhances detection capabilities. To explore detection capabilities under electrophoretic separation conditions, mixtures of Na+ and Li+ were detected with M-BSI. For these experiments, indirect UV (iUV) detection was implemented on the same capillary. The polyimide coating of a 75.5 cm total length separation capillary (75 µm i.d., 363 µm o.d.) was removed at 51 cm and 62 cm from the inlet to create windows for the i-UV and M-BSI detectors, respectively. Since Na+ and Li+ are UV silent, a 5 mM imidazole background electrolyte (BGE) at pH 5 was used to enable i-UV detection at 214 nm, as previously reported.41 Mixtures of NaCl and LiCl were pressure injected into the separation capillary at 6 psi for 0.1 s. Figure 3 shows the i-UV and M-BSI electropherograms for the separation of 1.33 g/L NaCl (22.7 mM) and LiCl (31.4 mM) at an applied potential of 24 kV. The Na+ and Li+ peaks are clearly resolved and have similar profiles in the two electropherograms. The migration times in the M-BSI electropherogram are slightly delayed due to the detectors placement downstream from the UV detector. This also accounts for the slightly higher efficiency in the separation parameters tabulated in Fig. 3. The Na+ band migrates first followed by Li+, with band shapes indicative of electrodispersion.42 As seen in Fig. 3, the baseline stability for i-UV detection is better than that of M-BSI. In general, RI detection is susceptible to environmental influences that can alter the RI and lead to low frequency drifts in the signal. For electrophoretic separations, temperature gradients arising from Joule heating are the most problematic. Most liquids have thermo-optic coefficients, ⁄ , on the order of 10-4 RIU/oC and for water it is 0.9x10-4 RIU/oC at 635 nm.23, 39 Small temperature fluctuations, therefore, influence the RI value which remains a major challenge for RI detection. These effects can be minimized using small-bore capillaries combined with active temperature control, which is discussed later. For these particular measurements, however, only passive temperature control using a copper block heat sink to support the separation capillary was used. Figure 4 compares electropherograms measured at lower ion concentrations. A mixture of 4 mg/L NaCl (68.4 µM) and 4 mg/L LiCl (94.4 µM) was injected onto the capillary and separated at a lower applied potential of 12 kV to reduce Joule heating. With the lower concentration, electrodispersion is minimized and the peaks appear symmetric in both electropherograms.42 This, along with the longer migration times, result in significant improvements in separation efficiency as tabulated in the Fig. 3. Comparing the electropherograms in Fig. 4 reveals better signal-to-noise with M-BSI detection, while background stability remains better in the i-UV electropherogram. The enhanced sensitivity of M-BSI and the universal nature of RI detection, provides a new tool for CE and MCE separations of analytes traditionally difficult to detect using other optical approaches. Carbohydrates, for example, are challenging to detect directly in micro-separations since they lack a UV active chromophore.43-45 Nuetral carbohydrates, moreover, are very weak acids which require strongly alkaline run buffers for electrophoretic separation. Sucrose, glucose, and lactose, for example, have pKas of 12.62, 12.28, and 11.98, respectively, at 25oC. The importance of

ACS Paragon Plus Environment

Page 6 of 17

Page 7 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

carbohydrates in fields from the medical sciences to food and drink analysis, however, has motivated the development of various chemical modifications that aid in both the separation and detection of these analytes.3, 45-46 These approaches, however, add time and complexity to the analysis.45 Figure 5A shows an electropherogram for the direct detection of 1.74 g/L sucrose (5.1 mM), 0.97 g/L glucose (5.4 mM), and 1.72 g/L lactose (4.8 mM) using M-BSI. Following previous work, a 55 mM NaOH BGE was used to separate the three carbohydrates, which migrate in order of decreasing pKa with sucrose migrating first, followed by glucose and then lactose.44, 47 The high ionic strength of the BGE leads to increased Joule heating and fluctuations in the M-BSI signal. To reduce these effects, a 25 µm i.d. (363 µm o.d.) capillary was used in these separations. The total length of the capillary was 89 cm with a length to detector of 76 cm. To further minimize temperature effects, the detection zone of the capillary was mounted on a thermoelectric cooler that was held at 18.000 ± 0.004 oC. The sample was pressure injected onto the capillary at 2.5 psi for 10 s and the separation carried at a potential of 21 kV. Three peaks are clearly resolved in Fig. 5A following the neutral peak at 705 s. Migration times and efficiencies are tabulated in the inset. With the smaller lumen of the capillary (25 µm i.d.), the detection volume is reduced to 15 pL. Figure 5B shows an electropherogram for a mixture at lower concentrations of 174 mg/L sucrose (510 µM), 97 mg/L glucose (540 µM), and 172 mg/L lactose (480 µM). This mixture was separated at a potential of 24 kV using the same injection parameters. These initial measurements show that M-BSI can significantly improve RI detection in ultra-small volumes. While CE is used here to demonstrate the approach, the universal sensing, ease of on-column detection, and potential for miniaturization should prove especially useful for microfluidic chip-based analysis platforms. Previous studies integrating BSI with microfluidic platforms have already demonstrated the utility of this approach.28-29 As with all interferometry measurements, environmental influences from mechanical instabilities and temperature fluctuations can increase noise and drift in the signal. This can be especially challenging for electrophoretic separations, where Joule heating can affect the measured signal through the thermo-optic coefficient of the BGE. This contributes to the baseline instability seen in Figs. 4 and 5 that currently limit the detection metrics. With better thermal management and enhanced mechanical stability, however, we believe that detection into the nanomolar regime should be possible using the M-BSI approach. Summary M-BSI is introduced to improve detection for universal RI measurements in ultrasmall volumes. Using sucrose standards, we demonstrate detection limits of 700 µg/L at the 3σ level in a 75 µm i.d. capillary. This corresponds to measuring a 1x10-7 RIU change in a 90 pL detection volume. The electrophoretic separation of Na+ and Li+ is compared using both M-BSI and i-UV detection on the same capillary, validating the approach for micro-separations. Finally, the direct on-column detection of carbohydrates separated in a 25 µm i.d. capillary illustrates the potential for detecting species normally difficult to measure optically. The separation of sucrose, glucose, and lactose in a 55 mM NaOH BGE are detected directly at the 500 µM level in a 15 pL detection volume. Extending M-BSI detection for microfluidic platforms is straightforward and currently underway. For those applications, the ease of M-BSI alignment, stand-off on-column

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

detection, flexibility towards channel size and geometry, and potential for miniaturization should enable new applications for universal detection in ultra-small volumes. Acknowledgements We gratefully acknowledge support from the University of Kansas.

ACS Paragon Plus Environment

Page 8 of 17

Page 9 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

References 1. Harstad, R. K.; Johnson, A. C.; Weisenberger, M. M.; Bowser, M. T., Capillary Electrophoresis. Anal Chem 2016, 88 (1), 299-319. 2. Patabadige, D. E. W.; Jia, S.; Sibbitts, J.; Sadeghi, J.; Sellens, K.; Culbertson, C. T., Micro Total Analysis Systems: Fundamental Advances and Applications. Anal Chem 2016, 88 (1), 320-338. 3. Nuchtavorn, N.; Suntornsuk, W.; Lunte, S. M.; Suntornsuk, L., Recent applications of microchip electrophoresis to biomedical analysis. J Pharm Biomed Anal 2015, 113, 72-96. 4. Ouimet, C. M.; D'amico, C. I.; Kennedy, R. T., Advances in capillary electrophoresis and the implications for drug discovery. Expert Opin Drug Dis 2017, 12 (2), 213-224. 5. Swinney, K.; Bornhop, D. J., Detection in capillary electrophoresis. Electrophoresis 2000, 21 (7), 1239-1250. 6. Wu, J.; Gu, M., Microfluidic sensing: state of the art fabrication and detection techniques. J Biomed Opt 2011, 16 (8), 080901. 7. Gencoglu, A.; Minerick, A. R., Electrochemical detection techniques in micro- and nanofluidic devices. Microfluid Nanofluid 2014, 17 (5), 781-807. 8. Mogensen, K. B.; Klank, H.; Kutter, J. P., Recent developments in detection for microfluidic systems. Electrophoresis 2004, 25 (21-22), 3498-3512. 9. Son, S.; Grover, W. H.; Burg, T. P.; Manalis, S. R., Suspended microchannel resonators for ultralow volume universal detection. Anal Chem 2008, 80 (12), 47574760. 10. Wang, Z. L.; Bornhop, D. J., Dual-capillary backscatter interferometry for highsensitivity nanoliter-volume refractive index detection with density gradient compensation. Anal Chem 2005, 77 (24), 7872-7877. 11. Powell, P. R.; Woods, L. A.; Ewing, A. G., Characterization of etched electrochemical detection for electrophoresis in micron inner diameter capillaries. J Sep Sci 2005, 28 (18), 2540-2545. 12. Kim, D. C.; Dunn, R. C., Integrating Whispering Gallery Mode Refractive Index Sensing with Capillary Electrophoresis Separations Using Phase Sensitive Detection. Anal Chem 2016, 88 (2), 1426-1433. 13. Macka, M.; Piasecki, T.; Dasgupta, P. K., Light-Emitting Diodes for Analytical Chemistry. Annu Rev Anal Chem 2014, 7, 183-207. 14. Wu, J. G.; Zheng, G. A.; Lee, L. M., Optical imaging techniques in microfluidics and their applications. Lab Chip 2012, 12 (19), 3566-3575. 15. Prikryl, J.; Kleparnik, K.; Foret, F., Photodeposited silver nanoparticles for oncolumn surface-enhanced Raman spectrometry detection in capillary electrophoresis. J Chromatogr A 2012, 1226, 43-47. 16. Negri, P.; Schultz, Z. D., Online SERS detection of the 20 proteinogenic L-amino acids separated by capillary zone electrophoresis. Analyst 2014, 139 (22), 5989-5998. 17. Negri, P.; Flaherty, R. J.; Dada, O. O.; Schultz, Z. D., Ultrasensitive online SERS detection of structural isomers separated by capillary zone electrophoresis. Chem Commun 2014, 50 (21), 2707-2710. 18. Nirode, W. F.; Devault, G. L.; Sepaniak, M. J., On-column surface-enhanced Raman spectroscopy detection in capillary electrophoresis using running buffers containing silver colloidal solutions. Anal Chem 2000, 72 (8), 1866-1871. 19. Leopold, N.; Lendl, B., On-column silver substrate synthesis and surfaceenhanced Raman detection in capillary electrophoresis. Anal Bioanal Chem 2010, 396 (6), 2341-2348.

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

20. Scholten, K.; Collin, W. R.; Fan, X.; Zellers, E. T., Nanoparticle-coated microoptofluidic ring resonator as a detector for microscale gas chromatographic vapor analysis. Nanoscale 2015, 7 (20), 9282-9289. 21. Whelan, R. J.; Zare, R. N., Surface plasmon resonance detection for capillary electrophoresis separations. Anal Chem 2003, 75 (6), 1542-1547. 22. Wade, J. H.; Bailey, R. C., Refractive Index-Based Detection of Gradient Elution Liquid Chromatography using Chip-Integrated Microring Resonator Arrays. Anal Chem 2014, 86 (1), 913-919. 23. Bruno, A. E.; Krattiger, B.; Maystre, F.; Widmer, H. M., On-Column Laser-Based Refractive-Index Detector for Capillary Electrophoresis. Anal Chem 1991, 63 (23), 26892697. 24. Yang, X. H.; Zhou, M. H.; Li, S.; Liu, Z. H.; Yang, J.; Zhang, Y.; Yuan, T. T.; Qi, X. X.; Li, H. Y.; Yuan, L. B., On-line dynamic detection in the electrophoretic separation by tapered optical fiber interferometer. Sensor Actuat B-Chem 2017, 242, 667-672. 25. Wang, Z. L.; Swinney, K.; Bornhop, D. J., Attomole sensitivity for unlabeled proteins and polypeptides with on-chip capillary electrophoresis and universal detection by interferometric backscatter. Electrophoresis 2003, 24 (5), 865-873. 26. Deng, Y. Z.; Li, B. C., On-column refractive-index detection based on retroreflected beam interference for capillary electrophoresis. Appl Optics 1998, 37 (6), 998-1005. 27. Swinney, K.; Markov, D.; Bornhop, D. J., Ultrasmall volume refractive index detection using microinterferometry. Rev Sci Instrum 2000, 71 (7), 2684-2692. 28. Baksh, M. M.; Kussrow, A. K.; Mileni, M.; Finn, M. G.; Bornhop, D. J., Label-free quantification of membrane-ligand interactions using backscattering interferometry. Nat Biotechnol 2011, 29 (4), 357-360. 29. Bornhop, D. J.; Latham, J. C.; Kussrow, A.; Markov, D. A.; Jones, R. D.; Sorensen, H. S., Free-solution, label-free molecular interactions studied by backscattering interferometry. Science 2007, 317 (5845), 1732-1736. 30. Saetear, P.; Chamieh, J.; Kammer, M. N.; Manuel, T. J.; Biron, J. P.; Bornhop, D. J.; Cottet, H., Taylor Dispersion Analysis of Polysaccharides Using Backscattering Interferometry. Anal Chem 2017, 89 (12), 6710-6718. 31. Jepsen, S. T.; Jorgensen, T. M.; Zong, W.; Trydal, T.; Kristensen, S. R.; Sorensen, H. S., Evaluation of back scatter interferometry, a method for detecting protein binding in solution. Analyst 2015, 140 (3), 895-901. 32. Jorgensen, T. M.; Jepsen, S. T.; Sorensen, H. S.; di Gennaro, A. K.; Kristensen, S. R., Back scattering interferometry revisited - A theoretical and experimental investigation. Sensor Actuat B-Chem 2015, 220, 1328-1337. 33. Baksh, M. M.; Finn, M. G., An experimental check of backscattering interferometry. Sensor Actuat B-Chem 2017, 243, 977-981. 34. Swinney, K.; Bornhop, D. J., Quantification and evaluation of Joule heating in onchip capillary electrophoresis. Electrophoresis 2002, 23 (4), 613-620. 35. Swinney, K.; Markov, D.; Bornhop, D. J., Chip-scale universal detection based on backscatter interferometry. Anal Chem 2000, 72 (13), 2690-2695. 36. Xu, Q. W.; Tian, W. J.; You, Z. H.; Xiao, J. H., Multiple beam interference model for measuring parameters of a capillary. Appl Optics 2015, 54 (22), 6948-6954. 37. Tarigan, H. J.; Neill, P.; Kenmore, C. K.; Bornhop, D. J., Capillary-scale refractive index detection by interferometric backscatter. Anal Chem 1996, 68 (10), 1762-1770. 38. Bornhop, D. J., Microvolume Index of Refraction Determinations by Interferometric Backscatter. Appl Optics 1995, 34 (18), 3234-3239.

ACS Paragon Plus Environment

Page 10 of 17

Page 11 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

39. Daimon, M.; Masumura, A., Measurement of the refractive index of distilled water from the near-infrared region to the ultraviolet region. Appl Optics 2007, 46 (18), 38113820. 40. ICUMSA Proceedings of the International Commission for Uniform Methods of Sugar Analysis: 16th session; Ankara. 1974. 41. Lee, Y. H.; Lin, T. I., Determination of Metal Cations by Capillary Electrophoresis - Effect of Background Carrier and Complexing Agents. J Chromatogr A 1994, 675 (1-2), 227-236. 42. Weinberger, R., Practical capillary electrophoresis. 2nd ed.; Academic Press: San diego, CA, 2000; p xvii, 462 p. 43. Landers, J. P., Handbook of capillary electrophoresis. 2nd ed.; CRC Press: Boca Raton, 1997; p 894 p. 44. Xu, X.; Kok, W. T.; Poppe, H., Sensitive Determination of Sugars by Capillary Zone Electrophoresis with Indirect Uv Detection under Highly Alkaline Conditions. J Chromatogr A 1995, 716 (1-2), 231-240. 45. Suzuki, S.; Honda, S., Miniaturization in carbohydrate analysis. Electrophoresis 2003, 24 (21), 3577-3582. 46. de Oliveira, M. A. L.; Porto, B. L. S.; Bastos, C. D.; Sabarense, C. M.; Vaz, F. A. S.; Neves, L. N. O.; Duarte, L. M.; Campos, N. D.; Chellini, P. R.; da Silva, P. H. F.; de Sousa, R. A.; Marques, R.; Sato, R. T.; Grazul, R. M.; Lisboa, T. P.; Mendes, T. D.; Rios, V. C., Analysis of amino acids, proteins, carbohydrates and lipids in food by capillary electromigration methods: a review. Anal Methods-Uk 2016, 8 (18), 3649-3680. 47. Colon, L. A.; Dadoo, R.; Zare, R. N., Determination of Carbohydrates by Capillary Zone Electrophoresis with Amperometric Detection at a Copper Microelectrode. Anal Chem 1993, 65 (4), 476-481.

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1: (A) Schematic of BSI where an unfocused laser beam directed onto a CE capillary generates an interference pattern detected in the back-scattered direction. Fringes in the BSI pattern shift in response to changes in the refractive index of the solution filling the capillary. (B). Schematic of M-BSI where the output of a fiber coupled tunable diode laser is focused onto the CE capillary. The reference signal from a lock-in modulates the laser wavelength, which shifts the interference pattern as indicated in panel A. A fringe in the retroreflected interference pattern is detected on a split photodiode, amplified, and sent to a lock-in amplifier to measure changes in refractive index.

ACS Paragon Plus Environment

Page 12 of 17

Page 13 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 2: M-BSI response to increasing concentrations of sucrose solutions. Sucrose standards at the indicated concentrations were passed through a 75 µm i.d. (363 µm o.d.) using pressure. Following each standard, ultra pure water was introduced and the signal returned to baseline as shown. The inset shows the calibration plot generated from the measured signal level changes.

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3: (A) Indirect UV (i-UV) and (B) M-BSI detection of the separation of 1.33 g/L NaCl and LiCl on the same 75 µm i.d. (363 µm o.d.) capillary. Comparable peak shapes and separation parameters are measured in the two detection schemes, with slightly better baseline stability observed in the i-UV electropherogram.

ACS Paragon Plus Environment

Page 14 of 17

Page 15 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 4: (A) i-UV and (B) M-BSI detection of the separation of 4.0 mg/L NaCl and LiCl on the same 75 µm i.d. (363 µm o.d.) capillary. As in Fig. 3, comparable peak shapes and separation parameters are measured in the two electropherograms. While baseline stability remains somewhat better in the i-UV electropherogram, the M-BSI measurements reveal better signal-to-noise levels.

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 5: The separation of sucrose, glucose, and lactose in samples at high (A) and low (B) concentrations, detected directly with on-column M-BSI. The migration order following the neutral peak is determined by the pKa of the carbohydrates, with the separation parameters tabulated as shown. To reduce Joule heating in the 55 mM NaOH BGE, a 25 µm i.d. (363 µm o.d.) capillary was used, leading to a detection volume of 15 pL. Separation potentials of 21 kV and 24 kV were used in (A) and (B), respectively.

ACS Paragon Plus Environment

Page 16 of 17

Page 17 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

For Table of Contents Only

ACS Paragon Plus Environment