Anal. Chem. 2009, 81, 7256–7261
Spray Performance of Microfluidic Glass Devices with Integrated Pulled Nanoelectrospray Emitters Peter Hoffmann, Markus Eschner, Stefanie Fritzsche, and Detlev Belder* University of Leipzig, Institute of Analytical Chemistry, Johannisallee 29, 04103 Leipzig, Germany The performance of microfluidic glass devices with a monolithically integrated nanospray tip have been evaluated. The nanospray tip is generated directly on the edge of a microfluidic glass chip by a pulling step followed by HF-etching for directed formation of a sharp tip with defined emitter area. For a fair judgment of the MSdetection sensitivity, we compared the detection performance with commercial nanospray needles. For that purpose the effect of the emitter opening on the sensitivity was studied in detail for different microfluidic chips as well as for commercial nanospray needles. A comparison of the chip-nanospray device with commercial nanospray needles revealed that a comparable spray performance is obtained at similar emitter diameters. A stable electrospray could be generated at such tapered tips without any need for hydrodynamic or electroosmotic pumping. The nanospray chips were successfully applied for coupling microchip electrophoresis and mass spectrometry. For improved performance, the separation channel of the microdevice was flushed with hydroxypropylmethylcellulose (HPMC) in order to reduce analyte-wall interactions and electroosmotic flow. The miniaturization of chemical processes from classical laboratories to fully integrated micro total analysis systems (µ-TAS)1 is an extremely fast growing field of research.2 This emerging new technology features numerous advantages like the accelerated speed of reactions and analysis, little sample and solvent consumption, and the potential for system integration. Sensitive and selective detection methods are crucial and one of the most challenging areas in the field of lab-on-a-chip technology. The most common detection technique in microfluidics is fluorescence,3 due to its unsurpassed sensitivity and the ease of practical implementation in common microscopic setups. Common fluorescence detection however suffers from crucial labeling steps, * To whom correspondence should be addressed. E-mail: belder@ uni-leipzig.de. Phone: +49-341-97-36221. Fax: +49-341 97-36229. (1) Manz, A.; Verpoorte, E.; Effenhauser, C. S.; Burggraf, N.; Raymond, D. E.; Widmer, H. M. Fresenius J. Anal. Chem. 1994, 348, 567–571. (2) West, J.; Becker, M.; Tombrink, S.; Manz, A. Anal. Chem. 2008, 80, 4403– 4419. (3) Go ¨tz, S.; Karst, U. Anal. Bioanal. Chem. 2007, 387, 183–192. (4) Schulze, P.; Belder, D. Anal. Bioanal. Chem. 2009, 393, 515–525. (5) Schulze, P.; Ludwig, M.; Kohler, F.; Belder, D. Anal. Chem. 2005, 77, 1325–1329.
7256
Analytical Chemistry, Vol. 81, No. 17, September 1, 2009
which can be overcome by the application of native fluorescence detection4 utilizing deep UV laser excitation5,6 or two-photon excitation.7 One of the most appealing detection techniques in microfluidics is mass spectrometry as it allows unequivocal identification of analytes. The first approaches for coupling microchips and mass spectrometry were introduced independently by Ramsey and Karger in 1997.8,9 In these reports electrospray was generated at the planar edge of a microfluidic chip. The striking simplicity of this approach was however accompanied by several disadvantages. The large emitter area caused an undesired dead volume in terms of a big liquid droplet which resulted in insufficient spray stability and added a significant dead volume impeding high-efficiency separations. An obvious method to improve the spray performance is the use of external emitters especially nanoelectrospray (nanoES) tips. According to the principle of nanoelectrospray,10,11 the emitter diameter is proportional to the size of solvent droplets and the ionization efficiency is indirectly proportional to the solvent droplet size. The use of external emitters12,13 reduces the emitter area significantly and is a feasible way for coupling chip-electrophoresis and mass spectrometry especially in academic laboratories. The addition of such capillary emitters relies however on tricky assembling procedures, which cannot exclude dead volumes in a reliable manner. The multitude of methods for chip-MS interfacing was discussed in five recent reviews.14-18 Monolithic integration of a confined emitter is the most appealing method because band broadening that is associated with dead volumes can be eliminated. With the use of common microfabrication techniques such as injection molding, laser ablation, or UV-lithography, this approach can be elegantly realized in polymeric materials.19-23 Microfluidic polymeric MS chips with emitter tips formed by laser (6) Schulze, P.; Ludwig, M.; Belder, D. Electrophoresis 2008, 29, 4894–4899. (7) Schulze, P.; Schu ¨ ttpelz, M.; Sauer, M.; Belder, D. Lab Chip 2007, 7, 1841– 1844. (8) Ramsey, R. S.; Ramsey, J. M. Anal. Chem. 1997, 69, 1174–1178. (9) Xue, Q.; Foret, F.; Dunayevskiy, Y. M.; Zavracky, P. M.; McGruer, N. E.; Karger, B. L. Anal. Chem. 1997, 69, 426–430. (10) Wilm, M. S.; Mann, M. Int. J. Mass Spectrom. 1994, 136, 167–180. (11) Wilm, M. S.; Mann, M. Anal. Chem. 1996, 68, 1–8. (12) Bings, N. H.; Wang, C.; Skinner, C. D.; Colyer, C. L.; Thibault, P.; Harrison, D. J. Anal. Chem. 1999, 71, 3292–3296. (13) Zhang, B.; Foret, F.; Karger, B. L. Anal. Chem. 2000, 72, 1015–1022. (14) Zamfir, A. D.; Bindila, L.; Lion, N.; Allen, M.; Girault, H. H.; Peter-Katalinic, J. Electrophoresis 2005, 26, 3650–3673. (15) Sung, W. C.; Makamba, H.; Chen, S. H. Electrophoresis 2005, 26, 1783– 1791. (16) Foret, F.; Kusy, P. Electrophoresis 2006, 27, 4877–4887. (17) Lazar, I. M.; Grym, J.; Foret, F. Mass Spectrom. Rev. 2006, 25, 573–594. (18) Koster, S.; Verpoorte, E. Lab Chip 2007, 7, 1394–1412. 10.1021/ac9015038 CCC: $40.75 2009 American Chemical Society Published on Web 07/29/2009
ablation are commercially available for chip chromatography and mass spectrometry coupling.24 Monolithic integration of emitter tips in glass is more challenging, but glass has several advantages as chip material such as its chemical inertness, the mechanical stability, and the optical transparency. Several approaches aiming at a reduction of the emitter size in microfluidic glass chips have been reported.25 Yue et al.26 studied mechanical processes to obtain a sharpened emitter tip by cutting and sanding. They found that channel centering is challenging in the mechanical process they used and proposed polishing of a microfluidic device in only two dimensions. An approach where the channel is centered just in the corner of a thin microfluidic chip has been presented recently by Mellors et al.27 This approach can also be regarded as a sort of “tip” confined in two dimensions. They found ESI-MS sensitivity levels comparable to commercial pulled capillary nanospray emitters tapered to 5 µm at the tip. These results were obtained at flow rates of ∼40 nL/min delivered either by a syringe pump or by an integrated electroosmotic pump. Recently, we introduced a microfluidic glass chip with a monolithically integrated pulled emitter for MS-coupling.28 This nanospray emitter which sprays directly from the chip was manufactured by a two step procedure, namely, computer numerical control (CNC)-milling to generate a cone followed by pulling the heated pin. With the use of this approach, very fine tips with tapered microchannels could be fabricated. These tips were found to be comparable to commercially available nanospray needles. They exhibit emitter dimensions of only a few micrometers which allow stable spray performance without additional pumping. A difficulty in this method was the controlled opening of the drawn tips. This was achieved mechanically just as with conventional nanospray needles11,29 by breaking or cutting the tip. With this rather random procedure it was difficult to generate reproducible tip openings. As the emitter area has a strong impact on the spray performance it would be desirable to generate a defined emitter area and geometry in a directed and reproducible manner. This is especially important for a fair comparison of the spray performance with commercial nanospray emitters and with literature data from other approaches. In this contribution we present an improved method for an exact centering of the channel during CNC-milling and for a well controlled opening of the fused pulled tips. This allows the (19) Svedberg, M.; Pettersson, A.; Nilsson, S.; Bergquist, J.; Nyholm, L.; Nikolajeff, F.; Markides, K. Anal. Chem. 2003, 75, 3934–3940. (20) Rossier, J. S.; Youhnovski, N.; Lion, N.; Damoc, E.; Becker, S.; Reymond, F.; Girault, H. H.; Przybylski, M. Angew. Chem., Int. Ed. 2003, 42, 53–58. (21) Schilling, M.; Nigge, W.; Rudzinski, A.; Nezer, A.; Hergenro¨der, R. Lab Chip 2004, 4, 220–224. (22) Tuomikoski, S.; Sikanen, T.; Ketola, R. A.; Kostiainen, R.; Kotiaho, T.; Franssila, S. Electrophoresis 2005, 26, 4691–4702. (23) Sikanen, T.; Tuomikoski, S.; Ketola, R. A.; Kostiainen, R.; Franssila, S.; Tapio Kotiaho, T. J. Mass Spectrom. 2008, 43, 726–735. (24) Yin, H. F.; Killeen, K. J. Sep. Sci. 2007, 30, 1427–1434. (25) Petersen, D.; Varjo, S.; Geschke, O.; Riekkola, M.-L.; Kutter, J. P. Proceedings of the Micro-TAS 2002 Symposium; Baba, Y., Shoji, S., van den Berg, A., Eds.; Nara, Japan, November 3-7, 2002; pp 691-693. (26) Yue, G. E.; Roper, M. G.; Jeffery, E. D.; Easley, C. J.; Balchunas, C.; Landers, J. P.; Ferrance, J. P. Lab Chip 2005, 5, 619–627. (27) Mellors, J. S.; Gorbounov, V.; Ramsey, R. S.; Ramsey, J. M. Anal. Chem. 2008, 80, 6881–6887. (28) Hoffmann, P.; Ha¨usig, U.; Schulze, P.; Belder, D. Angew. Chem. Int.Ed. 2007, 46, 4913–4916. (29) Schmidt, A.; Karas, M.; Du ¨ lcks, T. J. Am. Soc. Mass Spectrom. 2003, 14, 492–500.
generation of nanospray tips with various well-defined emitter areas, and furthermore this etching step sharpens the tips significantly. In order to allow a fair judgment of the obtainable detection sensitivity, we compared the performance of these emitter chips with common nanospray needles with similar emitter dimensions. EXPERIMENTAL SECTION Deionized water from Millipore (Billerica) and methanol, acetic acid, and hydroxypropylmethylcellulose (HPMC) from Merck (Darmstadt, Germany) were used. Ephedrine, leucine-enkephalin, ubiquitin, spermine, arginine, and nicotinamide were obtained from Sigma-Aldrich (Steinheim, Germany). Emitter Fabrication. Common chips for electrophoresis (Micronit/The Netherlands) in simple cross layout were used for emitter fabrication. The injection channel has a length of 10 mm while the dimensions of the microfluidic channel are 50 µm × 20 µm (width × depth). The basic procedure of realizing a microfluidic glass chip with a fully integrated pulled nanoelectrospray emitter achieving flow rates of about 25 nL/min without pumping has been described earlier.26 A more straightforward centering of the channel is now achieved by camera-supported calibration of a CNC-milling machine. For this purpose a mounted chargecoupled device (CCD) camera (WAT-704R from Watec/Japan) was installed in the position of a milling head. After exchange of the camera mounting by the milling head, a 0.3 mm thick and 1.2 mm long cone with a centered channel was generated. This microcone was then drawn to a sharp tip using a platinum heating coil as described earlier. The closed channel was then opened by HF etching. For this purpose the drawn tip was immersed vertically into a 25% aqueous hydrofluoric acid solution (Merck, Darmstadt, Germany), which was covered with silicone oil to prevent evaporation of HF.30 Surface tension pulls the etchant up to the hydrophilic glass wall, and a concave meniscus is formed. A faint nitrogen flow applied to the channel indicates the end point of the etching process (after about 20 min) by bubble formation and, moreover, avoids etching of the channel interior. Subsequently the chip was rinsed with water. With the use of this approach, microfluidic glass chips with monolithically integrated nanoelectrospray emitters can be fabricated easily. The emitter diameter depends on the immersion time into the HF solution after the channel is reopened. Nanoelectrospray needles from Proxeon (Odense, Denmark) were etched accordingly to get similar tip diameters. Emitter Evaluation. All MS experiments were performed with a single quadrupole mass spectrometer LC-MS 2010 EV from Shimadzu (Duisburg, Germany) in which the standard ESI interface was replaced with a custom-built interface. Electrospray voltages of 1.0-1.4 kV were applied directly onto the microchannel of the glass chip connected via the buffer inlet. The emitter potential is defined by the voltage drop during the spray process which made a conductive emitter obsolete. The MS settings were as follows: The heat block temperature was 200 °C, and the temperature of the curved desolvation line was 250 °C. No nebulizing gas was used. Ephedrine and leucine-enkephalin were recorded in the scan mode with a mass range of 100-1000 Da, (30) Turner, D. R. Etch Procedure for Optical Fibers. U.S. Patent 4,469,554, September 4, 1984.
Analytical Chemistry, Vol. 81, No. 17, September 1, 2009
7257
Figure 1. Schematic drawing and light microscopic image of a microfluidic glass chip with an integrated nanoelectrospray emitter with an emitter opening of 3 µm.
while ubiquitin was recorded in the profile mode with a mass range of 100 - 2000 Da. Both the nanoelectrospray needles and the glass chips were mounted on a three-axis translation stage and positioned at a distance of 0.5 mm in front of the MS inlet. The position was controlled by an SZ 51 stereomicroscope from Olympus (Hamburg, Germany), equipped with a calibration grid. Generally the chip was filled by support of a vacuum pump. Safety Information. HF solutions are corrosive and extremely hazardous. HF readily penetrates skin and destructs deeper tissue layers. Therefore, HF solutions should be handled solely in a ventilated hood and appropriate protective clothing is necessary. RESULTS AND DISCUSSION Fabrication of the nanoES emitter chips is performed by CNCmilling and tip pulling, and previously the glass chip was broken to reopen the microchannel, which is a rather random process. Therefore it is difficult to produce defined emitter geometries in a reproducible manner. This problem can be overcome when the emitter is opened by a controlled HF etching process as described
in the Experimental Section. With this approach it is possible to manufacture tips with well-defined opening diameters in a reproducible manner. Furthermore this etching step sharpens the tip by decreasing the wall thickness, which significantly reduces the effective emitter area. A bright-field microscopic view of such an etched tip is shown in Figure 1. In Figure 2 scanning electron microscopy images of the previously used broken (part A) and improved etched (part B) tips are shown. Both emitters are characterized by an inner diameter (i.d.) of 10 µm, but as a liquid will spread over the surface, the effective emitter area is significantly smaller in the etched version. As shown in Figure 2, both emitters deliver a steady nanoelectrospray as confirmed by the stabilities of the total ion current. The HF etching step reduces the emitter area approximately by a factor of 20, which leads to a significantly higher total ion current (TIC) and molecule ion current (MIC). In this experiment, the signal intensity of the ephedrine molecule peak is 7-fold higher with the etched compared to the broken tip. The exemplary chosen etched emitter in Figure 2B exhibits some roughness on the external wall due to the etching process. A microscopic investigation of emitters which varied in surface appearances and a comparison in mass spectrometry indicated that this did not induce liquid spreading and did accordingly not notably alter the spray stability or sensitivity. As the overall sensitivity and the limit of detection is strongly dependent on various instrumental parameters, especially the utilized mass spectrometer, a comparison with other chip-emitters from literature data is difficult. A reasonable measure for the spray performance is the comparison with common nanospray needles
Figure 2. Comparison of total ion current (TIC) stability and signal intensity using mechanically opened (A) or etched tips (B). Sample: 1 µg/mL ephedrine in MeOH/H2O 50:50 (v/v). 7258
Analytical Chemistry, Vol. 81, No. 17, September 1, 2009
Figure 3. Influence of emitter opening diameter on signal intensity (molecule ion current) obtained with nanospray chips or with etched nanospray needles. Conditions: 1 µg/mL (6.1 µM) ephedrine in MeOH/ H2O 50:50 (v/v).
as a kind of gold standard. As it is well-known from conventional nanospray emitters,29,31,32 with the spray performance strongly dependent on the emitter area and the tip opening, a reasonable comparison is possible only if similar emitter areas are utilized. In order to study the relationship between emitter area and signal intensity in more detail, we have generated a series of chip emitters which differ in size of the openings from 1 to 20 µm. This could be achieved by variation of the etching time. In an analogous way, commercial nanospray needles were etched for comparison with our nanospray chip device. In a set of experiments similar to that shown in Figure 2, we evaluated the impact of the emitter diameter on the signal intensity. The obtained results for the nanospray chip and nanospray needles are compared in Figure 3. For statistical evaluation, each data point was measured 6-fold (n ) 6) and the error bars indicate the standard deviation. Similar data curves could be observed for the chip emitter and the needle with an optimum of about 8-10 µm in emitter diameter. At our experimental conditions we found an optimum in emitter area at about 10 µm. However earlier work on common nanospray needles with a more fundamental investigation of the nanospray process, including a thorough consideration of the flow rates, indicates that the sensitivity can also continue to rise with smaller emitter diameters.29 We prepared another set of emitters where we tried to meet exact emitter opening diameters of 10 µm. In these experiments we analyzed leucine-enkephalin and ubiquitin and found very similar results for the chip emitter and the nanospray needles. These results were obtained at slightly modified conditions compared to those in Figure 3. The channels were flushed with 0.05% HPMC, which covered the glass surfaces, in order to generate equally suppressed electroosmotic flow and adsorptivity behavior in both configurations. An electrical potential of 1.2-1.4 kV was applied and acetic acid was added to improve the ionization process. Optimal signal intensities of leucine-enkephalin (0.1% HAc) and ubiquitin (1% HAc) were obtained at different amounts of acetic acid. For both device configurations, very similar signal intensities were obtained in repetitive experiments with single
Figure 4. Total ion currents and obtained mass spectra for ESI-MS analysis of leucine-enkephalin or ubiquitin with the chip-emitter using an opening diameter of 10 µm (A) 10 µg/mL leucine-enkephalin (0.1% HAc) and (B) 10 µg/mL ubiquitin (1% HAc) in 50% MeOH/H2O (v/v).
Table 1. Spray Performance of Commercial Nanospray Needles and Chip Emitters at Equal Emitter Openings of 10 µm (Data from Molecule Ion Currents) leucine-enkephalin [M + H]+ ) 556.3 signal intensity [× 106 counts/s] relative standard deviation [%] S/N-ratio
ubiquitin [M + 12H]12+ ) 714.8
needle
chip-emitter
needle
chip-emitter
2.92
3.08
6.89
6.71
9
7
14
11
285
300
148
138
emitters (n ) 6). Results from the chip emitter including total ion current and extracted mass spectra and are shown in Figure 4. The respective molecule ion current data comparing the spray performance of both emitter types are displayed in Table 1. It is evident that the performance of the chip-emitter is very similar to the nanospray needle if both openings are at 10 µm. A benefit of nanospray emitters is the improved performance at pure aqueous electrolytes, which is especially appealing for an intended coupling of chip electrophoresis with mass spectrometry. (31) Benkestock, K.; Sundqvist, G.; Edlund, P.-E.; Roerade, J. J. Mass Spectrom. 2004, 39, 1059–1067. (32) Juraschek, R.; Dulcks, T.; Karas, M. J. Am. Soc. Mass Spectrom. 1999, 10, 300–308.
Analytical Chemistry, Vol. 81, No. 17, September 1, 2009
7259
Figure 5. Influence of organic content (MeOH) on signal response in nanoelectrospray MS experiments with 1 µg/mL ephedrine as the test compound using the chip-emitter with a tip opening of 10 µm and 1.2 kV applied potential to the buffer inlet (BI) vial. The bars indicate the stability of the total ion current.
Figure 6. Reconstructed selected ion electropherograms with mass spectra and the influence of HPMC as dynamic coating on the separation of spermine (50 µg/mL), arginine (25 µg/mL), and nicotinamide (25 µg/mL). Electrolyte: 0.1% aqueous acetic acid solution containing 25% methanol (v/v).
The influence of the water content on the electrolyte composition was studied with a chip emitter opening of 10 µm; the respective data are shown in Figure 5. With 1 µg/mL ephedrine as the test solute, the highest signal response was be obtained at about 65 vol % MeOH, and with pure aqueous system we could still get about 20% in signal response. The stability of the total ion current is as well dependent on the solvent composition as displayed with bars in Figure 5; the highest fluctuations were observed at high organic solvent contents. In these studies, aimed at a balanced comparison of a chip emitter with commercial pulled nanospray needles, we used low micromolar concentrations and obtained signal-to-noise ratios of 7260
Analytical Chemistry, Vol. 81, No. 17, September 1, 2009
100-300. For these experiments, we used a simple and economic single quadrupole mass spectrometer. Detection of lower analyte concentrations should be possible using more sensitive MS instruments. Taking into account that the chip emitter is comparable to commercial nanospray needles, similar sensitivities should be obtained for both devices when coupled to other mass spectrometer types. This will be studied in future experiments focusing on coupling the nanospray chips to TOF-MS, FT-MS, and orbitrap systems. After characterization of the spray performance, the improved chip emitters were applied to chip electrophoresis with MS detection. For this purpose, we applied the electrodeless approach
without defined electrical connection at the tip where the outlet potential is left floating.12,33,34 Spermine (50 µg/mL), arginine (25 µg/mL), and nicotinamide (25 µg/mL) were chosen as test mixture compounds, and 0.1% aqueous acetic acid solution containing 25% (v/v) methanol was employed as an electrolyte (pH 3.5). The separation channel of the used chip had a length of 68 mm with an emitter diameter of 10 µm. The applied voltages during the separation were as follows: buffer inlet vial (BI) 5 kV, sample inlet/outlet vial (SI/SO) 2.5 kV, and grounded MS entrance. The experiment was repeated 6-fold resulting in a migration time variance of the arginine peak of 6%. The chip electrophoresisMS analysis was finished within 1.2 min but without baseline resolution as shown in Figure 6A (R ) 0.72 and 0.56, respectively). In order to improve the resolution we coated the inner surface of the channels with the hydrophilic polymer hydroxypropylmethylcellulose (HPMC).35-37 For this purpose, the channel was flushed with an aqueous 0.01% HPMC (v/v) solution prior to analyses. After HPMC treatment, the applied voltages were optimized for these conditions (BI 6 kV and SI/SO 3 kV). In Figure 6B, the respective MICs are plotted and showed baseline separation of spermine, arginine, and nicotinamide within 1.1 min with a relative standard deviation of 2% for arginine (n ) 6). With comparison of the analysis times obtained with and without HPMC in Figure 6, it is at first sight surprising that similar migration times are observed as one would expect increased migration times when HPMC suppresses the electroosmotic flow (EOF). However, as in the current electrodeless interface design, the emitter potential cannot be defined independently from the separation potentials and the electrical field strengths are not equal. In order to induce a stable spray at suppressed EOF conditions, the potentials had to be altered when HPMC was applied as a dynamic coating. In that case, voltages of 3 (sample inlet/outlet vial) and 6 kV (buffer inlet vial) were applied. Without HPMC coating and with use of the same voltages, no stable spray was obtained. For this reason the applied voltages were altered to 2.5 (sample inlet/outlet vial) and 5 kV (buffer inlet vial). While (33) Mazereeuw, M.; Hofte, A. J. P.; Tjaden, U. R.; van der Greef, J. Rapid Commun. Mass Spectrom. 1997, 11, 981–986. (34) Vrouwe, E. X.; Gysler, J.; Tjaden, U. R.; van der Greef, J. Rapid Commun. Mass Spectrom. 2000, 14, 1682–1688. (35) Belder, D.; Ludwig, M. Electrophoresis 2003, 24, 3595–3606. (36) Gilges, M.; Kleemiss, M. H.; Schomburg, G. Anal. Chem. 1994, 66, 2038– 2064. (37) Lindner, H.; Helliger, W.; Dirschlmayer, A.; Jaquemar, M.; Puschendorf, B. Biochem. J. 1992, 83, 467–471.
the electrodeless configuration has several benefits, such as straightforward technical implementation and stable nanospray operation, a further drawback of this approach is the difficulty of performing a pinched injection as the potential at the chip outlet cannot be exactly defined. Accordingly, the injection of a focused sample zone is impeded as monitored in video microscopy of the injection process (data not shown). This can explain that we observed, with 11 000 plates/m for signal 3 in Figure 6A (without HPMC) and 89 000 plates/m in Figure 6B (with HPMC), only moderate separation efficiencies. In order to overcome this problem, we tested several conductive coatings described in the literature, such as sputtered gold, silver paint,38 and the polymeric coating polyaniline.39 The best results where obtained with polyaniline and conductive silver paint where freshly coated tips worked satisfactory. The run to run reproducibility was however very poor, and the coating degraded just after a few runs, most probably due to arching end electrochemical processes. In a future chip device, we intend to include a makeup-flow channel, which would then allow defining the separation and the spray potentials independently. CONCLUSIONS In this approach, nanoelectrospray emitters were monolithically integrated on microfluidic glass chips enabling a reliable and sensitive coupling of microfluidic glass chips with nanospray mass spectrometry. An improved tip geometry was realized by an optimized manufacturing process including CNC-milling and HF etching. A detailed comparison of the chip-nanospray device with commercial nanospray needles revealed that a comparable spray performance is obtained at similar emitter diameters. In nanospray experiments without additional pumping, we found an optimum emitter area of about 10 µm. Such microfluidic glass chips with integrated nanospray emitters can be used for straightforward and sensitive coupling of various microfluidic techniques with mass spectrometry. As it allows MS-coupling without additional pumping for spray stabilization, this approach is especially attractive for nonpressurized separation techniques like electrophoresis. Received for review May 5, 2009. Accepted July 18, 2009. AC9015038 (38) Chen, Y. R.; Her, G. R. Rapid Commun. Mass Spectrom. 2003, 17, 4337– 441. (39) Maziarz, E. P., III; Lorenz, S. A.; White, T. P.; Wood, T. D. J. Am. Soc. Mass Spectrom. 2000, 11, 659–663.
Analytical Chemistry, Vol. 81, No. 17, September 1, 2009
7261
