Attenuated Total Internal Reflectance Infrared ... - ACS Publications

Sample separations using a programmed series of pressure, voltage, and again ...... Cao, W.; Liu, J.; Yang, X.; Wang, E. Electrophoresis 2002, 23, 368...
0 downloads 0 Views 125KB Size
Anal. Chem. 2004, 76, 3826-3832

Attenuated Total Internal Reflectance Infrared Microspectroscopy as a Detection Technique for Capillary Electrophoresis Brian M. Patterson, Neil D. Danielson, and Andre´ J. Sommer*

Molecular Microspectroscopy Laboratory, Department of Chemistry and Biochemistry, Miami University, Oxford, Ohio 45056

Attenuated total internal reflection (ATR) infrared (IR) spectroscopy has experienced tremendous growth over the past several years, in part because of its versatility and the ease with which samples are analyzed. Several portable versions of ATR infrared spectrometers are commercially available and find use in applications such as forensics and hazardous materials identification, specifically for homeland security.1,2 When incorporated into a microscope configuration, that is, ATR infrared microspectroscopy, the method has several advantages over traditional IR techniques.3-7 These advantages include a small, well-defined optical path length and a magnification factor that is equal to the

refractive index of the internal reflection element (IRE). From a microscopic standpoint, these benefits translate into an increased spatial and volumetric resolution.7 One final advantage of ATR infrared microscopy is its ability to provide direct molecular information of structurally related compounds. As a result, it has high selectivity when compared to other detection techniques commonly employed in chromatography. Online flow cells using ATR have been developed for liquid chromatography (LC) using the multibounce CIRCLE and ultramicro CIRCLE cells; however, they are not applicable to capillary LC or CE. A 25-µm capillary, 94 cm in length, has a volume of 460 nL; however, the ultramicro CIRCLE cell has a volume of 1.75 µL.8 Unacceptable band broadening in the detector cell would result. Only recently, several techniques both offline and online have been developed to hyphenate capillary electrophoresis (CE) to Fourier transform infrared (FT-IR). Offline techniques9-11 using solvent elimination involve deposition of the eluent by spraying onto an IR-inactive salt substrate as it exits the capillary. The salt substrate may be IR-transparent, in which case, spectra are collected in the transmission mode.10 It is also possible to deposit eluent onto the IRE of the Harrick SplitPea.7,11 Eluent deposition characteristics have been extensively studied to remove electrolytes and minimize spot area while maximizing eluent separation. Spectra of 0.5 ng of poly(ethylene glycol)7 and 155 ng of sodium benzoate11 have been collected using this technique. Some disadvantages include band broadening due to the spraying; extra time due to the post separation measurement; and most importantly, a complex setup. Online techniques, typically using specialty transmission flow cells made out of CaF2 salt plates,12,13 avoid the IR absorption of silica. The flow cell channel has the dimensions of 150 µm wide and 2 mm long with a 15-µm IR path length. The mass detection limit for guanosine was shown to be 800 pg (3 S/N). The volume of the IR flow cell is 4.5 nL, which is approximately one-half the injection volume of 10 nL. According to the electropherogram, with a flow rate of 98 nL/min, the sample

* To whom correspondence should be sent. E-mail: [email protected]. (1) Rein, A. J. Microsc. Today 2003, 11, 16-17. (2) Norman, M. L.; Gagnon, A. M.; Reffner, J. A.; Schiering, D. W.; Allen, J. D. Proceedings of SPIE, Photonics East, October 27-30, 2003, Providence, RI, in press. (3) Reffner, J. A.; Alexay, C. C.; Hornlein, R. W. SPIE 1991, 1575, 301-302. (4) Harrick, N. J.; Milosevic, M.; Berets, S. L. Appl. Spectrosc. 1991, 45 (6), 944-948. (5) Nakano, T.; Kawata, S. Scanning 1994, 16, 368-371. (6) Lewis, L.; Sommer, A. J. Appl. Spectrosc. 1999, 53, 375-380. (7) Sommer, A. J.; Hardgrove, M. Vib. Spectrosc. 2000, 24, 93-100.

(8) McKittrick, P. T.; Danielson, N. D.; Katon, J. E. Microchem. J. 1991, 44, 105-116. (9) He, L.-T.; de Haseth, J. A. AIP Conf. Proc. 1998, 430 (Fourier Transform Spectroscopy), pp 407-410. (10) Jarman, J. L.; Todebush, R. A.; de Haseth, J. A. J. Chromatogr. 2002, 976, 19-26. (11) Jarman, J. L.; Seerley, S. I.; Todebush, R. A.; deHaseth, J. A. Appl. Spectrosc. 2003, 57, 1078-1086. (12) Kolhed, M.; Hinsmann, P.; Svasek, P.; Frank, J.; Karlberg, B.; Lendl, B. Anal. Chem. 2002, 74, 3843-3848. (13) Kolhed, M.; Lendl, B.; Karlberg, B. Analyst 2003, 128, 2-6.

A novel detector for capillary electrophoresis (CE) using single-bounce attenuated total internal reflectance (ATR) Fourier transform infrared (FT-IR) microspectroscopy is presented. The terminus of the CE capillary is placed ∼1 µm from the internal reflectance crystal at the focus of an ATR infrared microscope. Using pressure driven flow injection, concentration and volume detection limits have been determined for 25- and 10-µm-i.d. silica capillaries. Upon injection of 820 pL of succinylcholine chloride in a 10-µm capillary, a concentration detection limit of ∼0.5 parts per thousand (ppt), or 410 pg, is found. The injection volume detection limit using a 108 ppt solution is 2.0 pL (216 pg). Sample separations using a programmed series of pressure, voltage, and again pressure on 25-, 50-, and 75-µm-i.d. capillaries are shown. CE separations of citrate and nitrate, as well as succinylcholine chloride with sodium salicylate using acetone as a neutral marker, are demonstrated. Several advantages of this CE-FT-IR technique include: (1) minimization of postcolumn broadening as a result of a small detector volume; (2) the ability to signal average spectra of the same aliquot, thereby improving the signal-to-noise in a stopped-flow environment; and (3) simplicity of design.

3826 Analytical Chemistry, Vol. 76, No. 13, July 1, 2004

10.1021/ac0400111 CCC: $27.50

© 2004 American Chemical Society Published on Web 05/19/2004

has a peak width of ∼30 s and a volume of ∼50 nL. For best sensitivity, the detector volume should not be more than onetenth the sample end-column volume.14 In typical CE, larger bore capillaries are used to improve spectrophotometric concentration sensitivity, with a sacrifice in efficiency and mass-injected detection limits. Alternatively, CE systems with electrochemical, conductivity, and electrochemiluminescence detection utilizing end-column15-19 monitoring have been extensively studied. The use of capillaries with diameters as small as 12.7 µm20 reduces solution heating, and the use of concentration-sensitive detectors improves mass detection limits. Spatial studies of electrodes in relation to the capillary end have been reviewed.15,16 Instead of placing the sensing electrode next to the outlet of the capillary,17 studies have shown for conductivity detection that sensitivity is improved by inserting the electrode ∼7 mm into the capillary, where a small hole was drilled in the side for the sample to exit.18 Best sensitivity is found for electrochemiluminescence detection when the distance between the capillary and detector is as high as 200 µm.19 To the best of our knowledge, end-column capillary detection for spectrophotometric methods has not been previously considered. The separation and detection of pharmaceuticals using reversedphase microbore end-column detection by ATR microscopy has been demonstrated.21 Detection limits of 0.7 parts per thousand (ppt) of succinylcholine chloride in water were also shown. Extending this approach, a novel end-column detector for CE using single bounce ATR-FT-IR is presented here. The terminus of the CE capillary is placed at the focus of an ATR infrared microscope. With this detection technique, the IR path length is completely independent of the capillary diameter. CE separations using a pressure/voltage/pressure program on 25-, 50-, and 75µm capillaries are shown. Because high analyte concentrations between 70 and 140 ppt were generally required, it was necessary to separate compounds with running electrolyte salt concentrations as high as 100 ppt NaCl (1.7 M), to provide electrostacking and good peak shape. CE separations of succinylcholine chloride with sodium salicylate, as well as citrate and nitrate, using acetone as a neutral marker are demonstrated. EXPERIMENTAL SECTION All spectra, except as noted, were collected with a Perkin-Elmer (Shelton, CT) AutoIMAGE microscope interfaced to a Spectrum 2000 FT-IR. The microscope was equipped with a liquid-nitrogencooled, wide-band 250 × 250 µm MCT detector and the standard Ge ATR microsampling accessory. The aperture at the primary image plane was set to 100 × 100 µm, which produced a 180-fL sampling volume. Data collection and processing were performed with Perkin-Elmer Spectrum TimeBase (Version 2.0) software on a Dell PC running Windows 95. No modifications were made to (14) Chervet, J. P.; Ursem, M.; Salzmann, J. P. Anal. Chem. 1996, 68, 15071512. (15) Matysik, F.-M. Anal. Chem. 2000, 72, 2581-2586. (16) Matysik, F.-M. Electroanalysis 2000, 12, 1349-1355. (17) Huang, X.; Zare, R. N.; Sloss, S.; Ewing, A. G. Anal. Chem. 1991, 63, 189192. (18) Huang, X.; Zare, R. N. Anal. Chem. 1991, 63, 2193-2196. (19) Cao, W.; Liu, J.; Yang, X.; Wang, E. Electrophoresis 2002, 23, 36833691. (20) Wallingford, R. A.; Ewing, A. G. Anal. Chem. 1988, 60, 1972-1975. (21) Patterson, B. M.; Danielson, N. D.; Sommer, A. J. Anal. Chem. 2003, 75, 1418-1424.

the instrument or computer. All spectra were collected at 8 cm-1 resolution. Single-scan backgrounds of flowing mobile phase were collected prior to each flow injection (FI) or CE run. The average time to collect a spectrum was 1.6 s. The entire optical path of the instrument was purged with boil-off from a liquid nitrogen tank. All FI and CE experiments employed a Crystal CE model 300 (Madison, WI) instrument with an autosampler. Its maximum voltage and current are 30 kV and 200 µA. Injection was made by applying pressure to the sample vial. Using Poiseuille’s Law (eq 1), the injection volume was calculated from the volume of a cylinder, where P is the applied guage pressure (mbar), t is time (s), x is the length of the sample (mm), d is the diameter of the capillary (µm), η is the viscosity of the fluid (cP), and L is the total length of the capillary (cm).22

Pt )

3200xηL d2

(1)

Solutions of sodium chloride, sodium citrate, sodium nitrate, succinylcholine chloride, sodium salicylate, and acetone were dissolved in distilled, deionized water. Samples were filtered with 0.2-µm PVDF (polyvinylidene fluoride) Gelman filters and degassed in an ultrasonicator for at least 15 min prior to use. Capillaries used for separations were flushed for at least 15 min each day with 1 M NaOH prior to use. Fused-silica capillaries of 10-, 25-, 50-, or 75-µm i.d. and 360µm o.d. were used for all experiments (Polymicro Technologies, Phoenix, AZ). The capillary was carefully cut to ensure that its terminus was perpendicular to its axis. This planarity is extremely important when bringing the terminus of the capillary into the focus of the IRE so that the evanescent wave penetrates into the central channel. Several millimeters of the capillary polyimide coating were burned off with a match to remove any frays. The end of the capillary was mounted on a metal slide and held in place with a plastic ferrule and metal nut. The end of the capillary was 2 mm above the metal nut, and held flush with the surface of the electrolyte solution in the reservoir. Because of the extremely low flow rates, the solution may evaporate faster than it is replaced, and a dry sample may show up in the IR spectrum and not be washed away. To correct this problem, a plastic cylindrical reservoir with an outer diameter of 3.8 cm, and a height of 0.76 cm was machined out of poly(vinyl chloride). The reservoir, holding more than 3.4 mL of solution, eliminated the possibility of memory effects and allowed the immersion of the Pt wire for grounding. The reservoir was press-fit onto the top of the metal nut. A schematic diagram is shown in Figure 1. The metal slide with the capillary end was mounted onto the AutoIMAGE stage and brought into focus. A visual image was made to track the terminus of the capillary with the position of the IRE. After visually focusing on a point midway between the inside edge and the outside edge of the capillary, the tip of the IRE was lowered, and a background spectrum was taken. The TimeBase software was used to monitor the spectra for changes while the capillary was raised. The capillary was raised 1 µm at a time until the silica absorption was just visible. The capillary was (22) Users Manual, Prince Autosampler; Lauerlabs B.V.: The Netherlands, 1994, p 15.

Analytical Chemistry, Vol. 76, No. 13, July 1, 2004

3827

Figure 1. Diagram of CE-ATR-FT-IR interface showing Ge ATR crystal, reservoir, capillary and platinum grounding wire. A representative analyte plug about to exit the capillary is shown. Top of the reservoir was open and could be filled with the desired electrolyte solution using a pipet. Note: not to scale.

then lowered 1 µm. In effect, this procedure put the IRE ∼1 µm from the end of the capillary. The capillary was then slowly moved on center with the IRE. It takes ∼1 h to set up and align the capillary. The reservoir was then filled with electrolyte solution to a point where capillary action with the IRE assured immersion for grounding. Only the tip of the IRE crystal was in the solution. The instrument was then wrapped in plastic bags to aid the nitrogen boil-off purge. All calibration curve data points were based on the integrated absorbance of the C-O-C stretch of succinylcholine chloride, a moderate infrared absorber, over the range of 1193-1133 cm-1. The stronger carbonyl band was not used because of its proximity to the water band. Each run was preceded by a single scan background. The five center spectra for each run were averaged for calibration calculations. Each data point is the average of three separate runs. Noise was calculated as 0.2 times the peak-to-peak noise (rms noise) over the range 2100-1900 cm-1. The detection limit was calculated by taking three times the noise of the blank and dividing by the slope of the calibration plot. Each experimental condition produced its own detection limit. FI experiments were initially completed to find mass and volume detection limits on 25- and 10-µm i.d. silica capillaries with the ATR detector using a three-stage program. The first stage was 2000 mbar pressure for 1 min to flush the system. The second stage was the sample injection. All injections were 20 s in duration. The third stage used 1500 mbar of pressure until after the analyte was detected. The concentration detection limit was measured using a constant injection volume, 10 nL for the 25-µm capillary and 0.8 nL for the 10-µm capillary, and various concentrations between 5 and 100 ppt. Volume detection limits were measured keeping the concentration constant (108 ppt) while varying the injection pressure between 100 and 1500 mbar. Injection volumes varied from 3.2 to 21 nL with the 25-µm capillary and from 0.06 to 0.8 nL for the 10-µm capillary. CE separations were carried out with a five-stage program (Table 1). Collection of the IR electropherogram always began