Anal. Chem. 2001, 73, 1286-1291
Interfacing a Polymer-Based Micromachined Device to a Nanoelectrospray Ionization Fourier Transform Ion Cyclotron Resonance Mass Spectrometer Zhaojing Meng, Shize Qi, Steven A. Soper,* and Patrick A. Limbach*
Department of Chemistry, Louisiana State University, 232 Choppin Hall, Baton Rouge, Louisiana 70803
Here we report the design, fabrication, and operation of a polymer-based microchip device interfaced to a nanoelectrospray ionization source and a Fourier transform ion cyclotron resonance mass spectrometer. The poly(methyl methacrylate) micromachined device was fabricated using X-ray lithography to produce a network of channels with high aspect ratios. Fabrication of high aspect ratio channels allows for zero dead volume interfaces between the microchip platform and the nanoelectrospray capillary interface. The performance of this device was evaluated with standard peptide and protein samples. High-quality mass spectral data from peptide and proteins (and mixtures thereof) were obtained without any interfering chemical noise from the polymer or the developers and plasticizers used in the fabrication process. Sample crosscontamination is not a problem using this polymer-based microchip device as demonstrated by the sequential analysis of several proteins. The nanoelectrospray source was operated at flow rates from 20 to 100 nL/min using pressure-driven flow, and uninterrupted operation for several hours is demonstrated without any noticeable signal degradation. The ability to fabricate multiple devices using injection molding or hot-embossing techniques of polymers provides a lower cost alternative to silicabased devices currently utilized with mass spectrometry. Microfabricated devices offer a number of important benefits for chemical analysis including reduced analysis times, reduced waste streams, reduced sample consumption, reduced reagent costs, ease of automation, and the potential to be used in highthroughput applications. A demonstrated strength of these devices is the potential for integrating several chemical processes, including sample preparation, chemical reactions, separation, and detection, directly onto a single device. This lab-on-a-chip concept1 has shown great potential in many fields including capillary electrophoresis,2-4 on-chip reactions,5,6 polymerase chain reaction * To whom correspondence should be addressed: (phone) (225) 388-3361; (fax) (225) 388-3458; (e-mail)
[email protected]; (e-mail)
[email protected]. (1) Figeys, D.; Pinto, D. Anal. Chem. 2000, 72, 330A-335A. (2) Zhang, B.; Liu, H.; Karger, B. L.; Foret, F. Anal. Chem 1999, 71, 32583264. (3) Li, J.; Thibault, P.; Bings, N. H.; Skinner, C. D.; Wang, C.; Colyer, C.; Harrison, J. Anal. Chem. 1999, 71, 3036-3045.
1286 Analytical Chemistry, Vol. 73, No. 6, March 15, 2001
(PCR) analysis,7,8 immunoassays,9 DNA separations,10 and hybridizations.11 To date, the most common detection method for microchips is laser-induced fluorescence (LIF). LIF is extremely compatible with the first generation of microfabricated devices, which used either glass or quartz substrates, due to the low level of autofluorescence these materials generate. Recently, mass spectrometry has been shown to be a compatible detection method for such devices.2,3,5,12-20 Mass spectrometry offers a number of advantages as a method of detection due to its high sensitivity and selectivity. Electrospray ionization (ESI) has been the method of choice for interfacing microfabricated devices with mass spectrometry. ESI is an ideal ionization source for microfabricated devices because this ionization method is compatible with solution-phase sample delivery and because ESI allows for the generation of gasphase ions from biological samples. In particular, nanoelectrospray ionization is especially compatible for use with microfabricated devices because of its ability to attain high-sensitivity analysis on extremely limited sample volumes and because the flow rate for (4) Jacobson, S. C.; Koutny, L. B.; Hergenroder, R.; Moore, A. W.; Ramsey, J. M. Anal. Chem. 1994, 66, 3472-3476. (5) Xue, Q.; Dunayevskiy, Y. M.; Foret, F.; Karger, B. L. Rapid Commun. Mass Spectrom. 1997, 11, 1253-1256. (6) Jacobson, S. C.; Hergenroder, R.; Moore, A. W.; Ramsey, J. M. Anal. Chem. 1994, 66, 4127-4132. (7) Khandurina, J.; McKnight, T. E.; Jacobson, S. C.; Waters, L. C.; Foote, R. S.; Ramsey, J. M. Anal. Chem. 2000, 72, 2995-3000. (8) Waters, L. C.; Jacobson, S. C.; Kroutchinina, N.; Khandurina, J.; Foote, R. S.; Ramsey, J. M. Anal. Chem. 1998, 70, 158-162. (9) Chiem, N.; Harrison, D. J. Anal. Chem. 1997, 69, 373-378. (10) Chen, Y.-H.; Chen, S.-H. Electrophoresis 2000, 21, 165-170. (11) Fan, Z. H.; Mangru, S.; Granzow, R.; Heaney, P.; Ho, W.; Dong, Q.; Kumar, R. Anal. Chem. 1999, 71, 4851-4859. (12) Bings, N. H.; Wang, C.; Skinner, C. D.; Colyer, C. L.; Thibault, P.; Harrison, D. J. Anal. Chem. 1999, 71, 3292-3296. (13) Chan, J. H.; Timperman, A. T.; Qin, D.; Aebersold, R. Anal. Chem. 1999, 71, 4437-4444. (14) Lazar, I. M.; Ramsey, R. S.; Sundberg, S.; Ramsey, J. M. Anal. Chem. 1999, 71, 3627-3631. (15) Pinto, D. M.; Ning, Y.; Figeys, D. Electrophoresis 2000, 21, 181-190. (16) Ramsey, R. S.; Ramsey, J. M. Anal. Chem. 1997, 69, 1174-1178. (17) Xue, Q.; Foret, F.; Dunayevskiy, Y. M.; Zavracky, P. M.; McGruer, N. E.; Karger, B. L. Anal. Chem. 1997, 69, 426-430. (18) Zhang, B.; Foret, F.; Karger, B. L. Anal. Chem. 2000, 72, 1015-1022. (19) Figeys, D.; Gygi, S. P.; McKinnon, G.; Aebersold, R. Anal. Chem. 1998, 70, 3728-3734. (20) Xu, N.; Lin, Y.; Hofstadler, S. A.; Matson, D.; Call, C. J.; Smith, R. D. Anal. Chem. 1998, 70, 3553-3556. 10.1021/ac000984a CCC: $20.00
© 2001 American Chemical Society Published on Web 02/15/2001
Figure 1. (a) PMMA microchip nanoelectrospray ionization platform developed in this work. (b) Schematic diagram of PMMA microchip. Two channels were blocked using epoxy to generate a single-channel device. (c) Optical micrograph of capillary interface region of PMMA microchip.
nanoelectrospray is similar to flow rates achievable with microfabricated devices.21 Although silica-based microfabricated devices have a number of demonstrated advantages, polymer-based microfabricated devices are receiving increased attention.10,13,22-24 The design and fabrication process for polymer-based devices is quicker, easier, and less expensive as compared to silica-based devices. Polymerbased devices offer the potential for mass production using replication procedures such as casting, embossing, or injection molding. Although polymer-based devices have been utilized with LIF detection,10,23,25 very little research has been performed to determine whether such devices would be compatible with mass spectrometry as the detection method. Aebersold and co-workers have shown that a poly(dimethylsiloxane) (PDMS) microfabricated device could be coupled via nanoelectrospray ionization to a quadrupole ion trap mass spectrometer for protein analysis.13 Smith and co-workers recently described a polycarbonate-based microchip20 with a machined electrospray emitter fabricated onchip coupled to a quadrupole ion trap mass spectrometer.26 Here, we have chosen to fabricate micromachined devices using poly(methyl methacrylate) (PMMA). PMMA is well-suited for use as a substrate for microfabricated devices because of its high dielectric constant, thermal conductivity comparable to silica, (21) Wilm, M.; Mann, M. Anal. Chem. 1996, 68, 1-8. (22) Martynova, L.; Locascio, L. E.; Gaitan, M.; Kramer, G. W.; Christensen, R. G.; MacCrehan, W. A. Anal. Chem. 1997, 69, 4783-4789. (23) Duffy, D. C.; McDonald, J. C.; Schueller, O. J. A.; Whitesides, G. M. Anal. Chem. 1998, 70, 4974-4984. (24) Soper, S. A.; Ford, S. M.; Qi, S.; McCarley, R. L.; Kelly, K.; Murphy, M. C. Anal. Chem. 2000, 72, 642A-651A. (25) Soper, S. A.; Ford, S. M.; Xu, Y.; Qi, S.; McWhorter, S.; Lassiter, S.; Patterson, D.; Bruch, R. C. J. Chromatogr., A 1999, 853, 107-120. (26) Wen, J.; Lin, Y.; Xiang, F.; Matson, D. W.; Udseth, H. R.; Smith, R. D. Electrophoresis 2000, 21, 191-197.
low cost, and ease of microfabrication.27 Our objective here was to interface a single-channel PMMA microchip with nanoelectrospray ionization mass spectrometry. We were interested in evaluating the polymer compatibility with typical electrospray solvent systems,24 and we were interested in whether the microchip would yield sufficient ion current for use with an FTICR mass spectrometer. We have evaluated the suitability of this device for biological applications using standard peptide and protein samples. We have found that this polymer-based device is ideally suited for use with nanoelectrospray ionization mass spectrometry, exhibiting no background due to the polymer, no sample memory effects, minimal incompatibilities with organic solvent solutions necessary for protein analysis, and stable ionization current during extended operation. High-quality mass spectral data can be obtained with minimal sample consumption. EXPERIMENTAL SECTION Chemicals. Angiotensin I, cytochrome c, and ubiquitin were obtained from Sigma (St. Louis, MO). Acetic acid and methanol were obtained from Fisher Scientific (Pittsburgh, PA). All reagents were used as received. All sample solutions were made to the desired concentrations in a 50% methanol/water solution with 1% acetic acid. Microfabrication of PMMA Device. The layout of the PMMA microchip used in this work is shown in Figure 1. Epoxy glue was used to block all microchannels except one to obtain a singlechannel device. The use of X-ray lithography for fabrication allows for zero dead volume interfaces (Figure 1c) to be designed to couple to the two transfer capillaries, one of which is used for sample introduction and the other of which is used to interface the device to the nanoelectrospray source. (27) Ford, S. M.; Davies, J.; Kar, B.; Qi, S. D.; McWhorter, S.; Soper, S. A.; Malek, C. K. J. Biomech. Eng. 1999, 121, 13-21.
Analytical Chemistry, Vol. 73, No. 6, March 15, 2001
1287
The single channel is 6 cm in length, 20 µm wide, and 40 µm deep with two transfer capillary interfaces that are 6 mm in length and 60 µm in width. The microchip was fabricated as described previously.27 Briefly, to fabricate this device, the microchip layout was designed and transferred to an optical mask using a pattern generator. Next, a 10-µm-thick photoresist layer was applied on a Kapton film (Fralock, Canoga Park, CA) by spin coating. After exposure of the photoresist to UV light and subsequent development, a 10-µm layer of gold, which is an X-ray absorber, was electroplated to the exposed area to produce an X-ray mask. The X-ray mask was situated over a 5 cm × 6 cm PMMA wafer (Goodfellow Co., Berwyn, PA) and the assembly placed in an X-ray beam line at the LSU Center for Advanced Microstructure and Devices (CAMD) (mas ) 4 keV, ∆ ) 4 keV (fwhw), incident power on sample: 1 W cm-1 100 mA-1). Microstructure channels were formed by developing the X-ray-exposed PMMA in the standard GG developer (15% water, 60% butoxyethanol, 20% terahydrooxazine, and 5% aminoethanol). A cover piece of PMMA was spin-coated with a 2-µm-thick layer of poly(butyl methacrylateco-methyl methacrylate) (PBMA-co-PMMA). The coated side of the cover was thermally bonded to the microchip substrate at 120 °C for 1 h. A 20-µm-i.d., 90-µm-o.d. fused-silica capillary (Polymicro Technologies, Phoenix, AZ) was chemically etched using HF to provide a tight fit within the microchannel capillary interface and inserted ∼6 mm into both ends of the microchannel and fixed in place with epoxy. Nanoelectrospray-PMMA Microchip Interface. The nanoelectrospray interface utilized in this work is similar to that previously described by Hannis and Muddiman.28 The fused-silica transfer capillary extended out of the microchip ∼20 cm. A 0.0625in.-o.d., 0.005-in.-i.d. stainless steel capillary was connected to the free end of the capillary via a 0.01 in. i.d. × 1/16 in. o.d. Teflon tube (Upchurch Scientific, Belfont, PA). One end of the Teflon tube was machined to slide over the stainless steel capillary, and the other end was attached onto the fused-silica capillary with a stainless steel ferule. The nanoelectrospray tip was in-house pulled on a bunsen burner from a 20-µm-i.d., 90-µm-o.d. fused-silica capillary (Polymicro). The nanoelectrospray tip was attached to the free end of the stainless steel capillary by utilizing Teflon tubing in the same manner as described above. The electrospray voltage was applied at the stainless steel capillary and was controlled by an Analytica (Analytica of Branford, Branford, CT) ESI power supply. The nanoelectrospray source was mounted on a Newport x-y-z translational stage (Newport, Irvine, CA), which allowed for the alignment of the nanoelectrospray tip with the mass spectrometer interface. During typical operation, the nanospray tip is positioned ∼1 mm from the mass spectrometer interface. Mass Spectrometry. Mass spectra were obtained from an IonSpec HighRes ESI Fourier transform mass spectrometer (Irvine, CA) equipped with a 4.7-T superconducting magnet. The original Analytica ESI transfer optics were modified to contain a heated metal capillary in place of the standard glass capillary as previously described.29 The temperature of the heated metal capillary was held constant at 200 °C. Typical operating conditions (28) Hannis, J. C.; Muddiman, D. C. Rapid Commun. Mass Spectrom. 1998, 12, 443-448. (29) Chowdhury, S. K.; Katta, V.; Chait, B. T. Rapid Commun. Mass Spectrom. 1990, 4, 81-87.
1288
Analytical Chemistry, Vol. 73, No. 6, March 15, 2001
for the ESI source were as follows: ESI source voltage 1-3 kV; skimmer voltage 10 V; capillary exit voltage 60 V; hexapole offset voltage 6.5 V; shutter voltage -75 V. Samples were pressure injected onto the PMMA microchip, and pressure-driven flow was used to acquire all mass spectra shown in this work. The flow rate was estimated to be about 20100 nL/min and was determined by measuring the time necessary for the sample to traverse the microchip device. The flow rate was adjusted by regulating the backing pressure. Ions were transferred to a 2-in. cylindrical Penning ion trap (trapping voltage 1.000 V) through a hexapole ion guide (operated at 932.5 kHz) by opening the mechanical shutter separating the ESI source from the ion transfer optics for 10 ms/scan. To efficiently trap the injected ions, a brief pulse (∼4 ms) of argon was used. The trapped ions were subjected to broadband frequency sweep excitation and detection. The time domain ICR signal was subjected to baseline correction followed by Hanning apodization and one zero-fill before Fourier transformation and magnitude-mode calculation. Spectra were obtained by averaging five scans if not specifically mentioned otherwise. RESULTS AND DISCUSSION As an initial evaluation of the feasibility of a PMMA microchip for use with mass spectrometry, we examined the following operational aspects of the device: background due to chemical noise, solvent compatibility, sample cross-contamination, source stability, sensitivity, and mass spectrometer efficiency. The first three aspects are related to the characteristics of the microchip device, and the last three aspects are related to the interface between the microchip device and the mass spectrometer. The operational behavior of the device was investigated using standard peptide and protein analytes analyzed in positive ion mode. PMMA Microchip Background. One initial concern regarding the suitability of a polymer-based microchip device is the potential for increased chemical noise arising from the polymer material or developers and plasticizers used in the processing of the polymer-based device. Chan et al. reported that microchips fabricated from PDMS exhibit an increase in chemical noise as compared to microchips fabricated from glass or silica-based substrates.13 In addition, mass spectrometry is a more universal detection method as compared to optical methods previously utilized with these devices,10,23,25 and mass spectrometry may be more sensitive to the chemicals utilized in the fabrication of the device. Figure 2a shows the mass spectrum obtained from initial analysis of a 10 µM solution of angiotensin I. While the mass spectrum appears to demonstrate that a functional interface between the PMMA microchip and the nanoelectrospray source was established, no ions representative of angiotensin I are detected. The mass spectrum is characterized by a “peak at every mass” which is characteristic of a chemical noise contamination problem. To remove the residual chemical interferences, the microchip was washed with 50% MeOH/H2O solution overnight. Figure 2b shows the mass spectrum obtained from the same angiotensin I sample solution after washing overnight. As can be seen in Figure 2b, the spectrum no longer shows residual chemical noise after the solvent wash and ions representative of the sample are readily detected. No further washing of the PMMA microchip was found
Figure 2. (a) Initial mass spectrum obtained from PMMA microchip device. The mass spectrum is characterized by a substantial chemical noise background most likely due to residual developer present in the microchannels. (b) Electrospray mass spectrum obtained with 10 µM angiotensin I in 50:50 MeOH/H2O containing 1% acetic acid after the PMMA microchip has been washed with 50:50 MeOH/H2O. The simple solvent wash completely eliminates the residual chemical contamination arising from the microchip fabrication step.
to be necessary to ensure that background chemical noise would not interfere with mass spectral analysis. This chemical noise most likely resulted from the developer used to remove exposed PMMA following X-ray exposure. The GG developer consists of several organic solvents and water (15% water, 60% butoxyethanol, 20% terahydrooxazine, and 5% aminoethanol), and the microchannels are saturated with GG for several minutes with sonication at elevated temperatures during fabrication. As demonstrated here, the residual developer is easily removed by washing the device prior to use and no further background contamination from the fabrication steps are seen during the operation of the device. Devices fabricated using injection molding or hot-embossing techniques should be less likely to be contaminated by the developer, thus reducing initial contamination concerns. Solvent Compatibility. For PMMA-based microchip devices to be useful for coupling to ESI-MS, these microchips should be compatible with the solvent systems typically used for biopolymer analysis. As mentioned earlier, the PMMA device was frequently flushed with 50% aqueous methanol overnight and experiments were performed the next day using 50% aqueous methanol with 1% acetic acid. During extended operation (∼4 months), no noticeable effects from the use of this solvent system has been observed. Microscopic inspection indicated no noticeable polymer swelling due to the use of this solvent system. In addition, solvents can be retained in the microchip overnight with no adverse effect on the operation of the device. Although it is known that PMMA is susceptible to particular solvents,24 our results suggest that the
Figure 3. Electrospray mass spectra of (a) 10 µM cytochrome c in 50:50 MeOH/H2O containing 1% acetic acid and (b) 10 µM ubiquitin in 50:50 MeOH/H2O containing 1% acetic acid. The mass spectrum in (b) was obtained immediately after the mass spectrum in (a), demonstrating that sample cross-contamination in the microchip is minimal.
acidified water/methanol solution used for peptide and protein analysis in these experiments is compatible with this device. Future investigations will examine the range of solvent systems that is compatible with this microchip device. Sample Cross-Contamination. Two issues were investigated with respect to sample cross-contamination. The first issue was the effect on the mass spectral data from leaving a particular sample in the microchip device for extended times. Previously, Chan et al. noted that PDMS-based devices exposed to sample solutions for extended periods of time exhibited poor performance.13 To investigate this issue, a 10 µM solution of angiotensin I was injected into the PMMA microchip and left to sit overnight. Mass spectral data were then obtained on the following day. While the first mass spectrum obtained the next day exhibited a lower signal-to-noise ratio than usual, no residual chemical noise was detected in any of the subsequent mass spectra and the signalto-noise ratio was similar to that obtained during the previous day’s operation (data not shown). In addition, the angiotensin I sample solution could be removed from the PMMA chip after washing with 50% MeOH/H2O and no signal representative of angiotensin I was detected. Thus, the PMMA device does not appear to exhibit memory effects even when exposed to a particular sample for extended periods of time. To investigate the second concern related to sample crosscontamination occurring between different samples analyzed on the same microchip, 10 µM solutions of angiotensin I, cytochrome c, and ubiquitin were introduced and analyzed sequentially without any solvent wash in between. Figure 3 shows representative mass Analytical Chemistry, Vol. 73, No. 6, March 15, 2001
1289
spectra obtained under these conditions. No sample crosscontamination is detected among any of the two subsequent samples injected (i.e., cytochrome c and ubiquitin) investigated, suggesting that this polymer-based microchip device does not exhibit sample memory effects even upon sequential injection of multiple samples. In addition, the ability to fabricate capillary interfaces reduces the dead volume between the nanospray interface and the microchip device, thereby reducing potential sample cross-contamination in the interface region. Fabrication of low dead volume interfaces, as opposed to postassembly construction,12 is easier to implement, is more precise, and reduces potential damage to the assembled microchip device. While previous investigators have shown that polymer-based devices demonstrate protein adsorption through modifications in the electroosmotic flow of the device or degradation in separation efficiency when used for electrophoresis,30 our results indicate that the proteins utilized in these experiments are either irreversibly adsorbed to the wall of the device such that no leaching of protein from the wall during operation is detectable by the mass spectrometer or that no protein adsorption occurs. Although preliminary water contact angle measurements of PMMA soaked in various protein solutions made in our laboratory indicate that adsorption does occur,31 no significant loss of signal intensity on the first run of a device, which would be indicative of strong protein adsorption, was noted. In any event, protein absorption may pose a problem in selected applications and the use of different polymer materials for the microdevice could alleviate this potential problem. Nanoelectrospray Source Stability. The stability of the PMMA microchip ESI assembly was also investigated. One advantage of nanoelectrospray is its low detection limits, which arise due to the low flow rates necessary for the operation of this source.21 We were interested in determining whether the PMMA chip would be compatible with extended analysis times or whether the device would have a limited experimental time frame after which the sensitivity of the assembly would decrease. To test the assembly stability, a 10 µM solution of cytochrome c was injected into the microchip device and analyzed for several hours. Panels a and b of Figure 4 respectively show mass spectra obtained at the beginning and end (240 min later) of the analysis. As can be seen by comparing the two mass spectra, minimal loss in performance of this assembly was noted. Thus, uninterrupted operation of several hours, which is substantially longer than the operation demonstrated from PDMS fabricated devices,13 is possible using a PMMA-fabricated device. We have noticed that a +98 Da adduct ion increases in abundance as the length of time that the sample is in contact with the ESI solvent and microchip device increases. Although it is possible that this adduct ion arises from the PMMA device (the mass of methyl methacrylate is 100 Da), we have observed this adduct when cytochrome c is analyzed without use of a microfabricated device. In addition, no adducts were observed during the extended analysis of other peptide and protein analytes. Thus, it appears that this adduct is associated with cytochrome c and is not due to the PMMA device. Mass Spectral Performance. A brief investigation into the sensitivity of the PMMA microchip was performed. Panels a and (30) Locasio, L. E.; Perso, C. E.; Lee, C. S. J. Chromatogr. 1999, 857, 275-284. (31) Hazel, D.; Soper, S. A., unpublished results.
1290 Analytical Chemistry, Vol. 73, No. 6, March 15, 2001
Figure 4. Electrospray mass spectra of 10 µM cytochrome c in 50: 50 MeOH/H2O containing 1% acetic acid. (a) Mass spectrum obtained at the beginning of the experimental run. (b) Mass spectrum obtained about 4 h later. No degradation of the signal is seen under extended operation of the PMMA microchip.
b of Figure 5 show representative mass spectra of 5 µM and 500 nM solutions of cytochrome c, respectively. Ions representative of cytochrome c are detected in both cases, although the signalto-noise level decreased substantially for the 500 nM solution (Figure 5b). Although no optimization was attempted to achieve the lowest possible detection limits from the assembly, these results demonstrate that submicromolar concentrations are compatible with this device. Further improvements in sensitivity should be achieved by optimizing the FTICR experimental parameters. The last issue addressed was whether the use of the PMMA microchip nanoelectrospray assembly would generate sufficient ion current that could be transferred to the FTICR MS Penning trap and analyzed at sufficiently low pressures to permit highresolution analysis of the sample solution. As a test of the performance of this assembly, a mixture of cytochrome c and ubiquitin (10 µM each) was injected into the PMMA microchip and analyzed. Figure 6a shows a representative mass spectrum of this mixture. A sufficient ion population was delivered to the ICR cell for this analysis, and the conditions of the FTICR MS system are such that isotopic resolution of each charge state is possible (which allows for the assignment of charge states directly). The inset is the spectrum arising from the transformation of the data in Figure 6a. As seen in this figure, the two components are clearly identified in the deconvoluted mass spectrum and no other components (representative of background contamination) are seen. To further illustrate the feasibility of high-resolution measurements, Figure 6b is an expansion of the 772-782 m/z region and shows the isotopically resolved peaks of the 16+ ion of cytochrome c and the 11+ ion of ubiquitin. Thus, the PMMA
Figure 5. (a) Electrospray mass spectrum of 5 µM cytochrome c in 50:50 MeOH/H2O containing 1% acetic acid. (b) Electrospray mass spectrum of 0.5 µM cytochrome c in 50:50 MeOH/H2O containing 1% acetic acid.
Figure 6. (a) Electrospray mass spectrum of a 10 µM mixture of cytochrome c and ubiquitin in 50:50 MeOH/H2O containing 1% acetic acid with transformed molecular mass spectrum of electrospray data presented in the inset. (b) Expansion of data obtained in (a) to demonstrate that isotopic resolution of the charge states for both cytochrome c and ubiquitin is obtained.
microchip nanoelectrospray assembly is capable of delivering suitable ion currents for high-resolution analysis. CONCLUSIONS Taken together, these experiments demonstrate that a polymerbased microchip device fabricated from PMMA is readily suited for use with nanoelectrospray ionization mass spectrometry. Minimal contamination problems are found using the PMMA device, and the straightforward nanoelectrospray interface allows such a device to be easily coupled to mass spectrometry-based detection systems. Future experiments will investigate the feasibility of multichannel devices operated using electroosmotic pumping for sample delivery and fabricated using injection molding or hotembossing techniques of polymers. The issue with polymer-based devices and electroosmotic pumping is the low electroosmotic flow that they can generate. For example, PMMA possesses an electroosmotic flow of 1.4 × 10-4 cm2/V‚s.27 Therefore, to generate a volume flow rate of ∼20 nL/min in a microchannel measuring 20 × 50 µm will require an electric field of 240 V/cm. This type of field is well below the breakdown voltage of PMMA and, for low-conductivity carrier buffers, will not generate excess Joule heat, which would degrade separation efficiency. An additional issue surrounding the use of FTICR MS as a detector for microfabricated devices is the lower duty cycle of this mass (32) Hofstadler, S. A.; Wahl, J. H.; Bruce, J. E.; Smith, R. D. J. Am. Chem. Soc. 1993, 115, 6983-6984. (33) Wang, Y.; Shi, S. D.-H.; Hendrickson, C. L.; Marshall, A. G. Int. J. Mass Spectrom. 2000, 198, 113-120.
spectrometer, especially if it is to be used for CE/MS. Smith and co-workers previously showed that the duty cycle can be improved sufficiently for CE/MS by use of high-speed cyropumping.32 In addition, modification of the hexapole ion guide to serve as a hexapole ion trap can also improve the instrument duty cycle.33 Although not investigated in this work, the low-cost rapid fabrication of multiple devices using injection molding or hotembossing techniques offers the potential for mass production of such devices at a greatly reduced cost. Future studies will investigate the operational characteristics of polymer-based devices fabricated using such techniques. The results presented here and the technology available for polymer-based fabrication suggests that this platform has the potential to offer a low-cost alternative to silica or glass-based microchip devices for sample analysis. ACKNOWLEDGMENT Financial support of this work was provided by the National Institute of Health HG01499 (to S.A.S.) and HG01777 (to P.A.L.). We also thank CAMD for the use of the microfabrication facilities and the synchrotron beam line for this work.
Received for review August 17, 2000. Accepted January 11, 2001. AC000984A Analytical Chemistry, Vol. 73, No. 6, March 15, 2001
1291
