Anal. Chem. 2001, 73, 4607-4616
Evaluations of the Stability of Sheathless Electrospray Ionization Mass Spectrometry Emitters Using Electrochemical Techniques Stefan Nilsson,† Malin Svedberg,‡ Jean Pettersson,† Fredrik Bjo 1 refors,§ Karin Markides,† and ,† Leif Nyholm*
Department of Analytical Chemistry, Uppsala University, Box 531, SE-751 21 Uppsala, Sweden, Department of Materials Science, Uppsala University, Box 534, SE-751 21 Uppsala, Sweden, and Division of Applied Physics, Department of Physics and Measurement Technology, Linko¨ping University, SE-581 83 Linko¨ping, Sweden
The processes that cause the failure of sheathless electrospray ionization (ESI) emitters, based on different kinds of gold coatings on fused-silica capillaries, are described and explained. The methods chosen for this study include electrochemical methods, ICPMS analysis of the electrolytes used, SEM studies, and electrospray experiments. Generally, the failure occurs by loss of the conductive coating. It is shown that emitters with sputtercoated gold lose their coatings because of mechanical stress caused by the gas evolution accompanying water oxidation or reduction. Emitters with gold coatings on top of adhesion layers of chromium and nickel alloy withstand this mechanical stress and have excellent durability when operating as cathodes. When operating as anodes, the adhesion layer is electrochemically dissolved through the gold film, and the gold film then flakes off. It is shown that the conductive coating behaves as a cathode even in the positive electrospray mode when the magnitude of a superimposed reductive electrophoretic current exceeds that of the oxidative electrospray current. Fairy-dust coatings developed in our laboratory (see Barnidge, D. R.; et al. Anal. Chem. 1999, 71, 4115-4118,) by gluing gold dust onto the emitter, are unaffected by the mechanical stress due to gas evolution. When oxidized, the fairydust coatings show an increased surface roughness and decreased conductivities due to the formation of gold oxide. The resistance of this oxide layer is however negligible in comparison with that of the gas phase in ESI. Furthermore, since no flaking and only negligible electrochemical etching of gold was found, practically unlimited emitter lifetimes may be achieved with fairy-dust coatings. Sheathless electrospray ionization (ESI) is known to provide excellent sensitivities in conjunction with a mass spectrometer (MS).2 The sheathless interface offers compatibility with flow rates of nanoliters per minute and introduces no postseparation band * Corresponding author: (e-mail)
[email protected]; (fax) +46 (0)18 471 3692. † Department of Analytical Chemistry, Uppsala University. ‡ Department of Materials Science, Uppsala University. § Linko ¨ping University. 10.1021/ac010300e CCC: $20.00 Published on Web 09/06/2001
© 2001 American Chemical Society
broadening when the separation device and the emitter are unified. It is therefore often chosen when interfacing low flow rate separation techniques such as capillary electrophoresis (CE), capillary electrochromatography (CEC), and nanoliquid chromatography, as well as low flow rate infusion, to MS. However, sheathless emitters prepared by the application of gold or silver conductive coating on the tip of a sharpened fused-silica capillary generally suffer from poor stability.1-19 This requires frequent replacement of the coating or the entire capillary making routine applications of sheathless ESI-MS difficult. A common way of obtaining ESI-MS conductive fused-silica tips involves the sputtering or evaporative deposition of gold5,10,11,14,15 or, less commonly, silver.3 To improve the mechanical stability of the film, an adhesion layer containing chromium,2,19 chromium and nickel alloy,12 or various organic compounds such as silanes6 (1) Barnidge, D. R.; Nilsson, S.; Markides, K. E. Anal. Chem. 1999, 71, 41154118. (2) Ramsey, R. S.; McLuckey, S. A. J. Microcolumn Sep. 1995, 7, 461. (3) Smith, R. D.; Olivares, J. A.; Nguyen, N. T.; Udseth, H. R. Anal. Chem. 1988, 60, 436-441. (4) Wahl, J. H.; Gale, D. C.; Hofstadler, S. A.; Udseth, H. R.; Smith, R. D. Presented at 41st ASMS conference, San Francisco, CA, 1993; Paper ThP 226. (5) Valaskovic, G. A.; Kelleher, N. L.; Little, D. P.; Aaserud, D. J.; McLafferty, F. W. Anal. Chem. 1995, 67, 3802-3805. (6) Krieger, M. S.; Cook, K. D.; Ramsey, R. S. Anal. Chem. 1995, 67, 385389. (7) Wahl, J. H.; Gale, D. C.; Smith, R. D. J. Chromatogr., A 1994, 659, 217222. (8) Wu, J.-T.; Qian, M. G.; Li, M. X.; Liu, L.; Lubman, D. M. Anal. Chem. 1996, 68, 3388-3396. (9) Li, M. X.; Liu, L.; Wu, J.-T.; Lubman, D. M. Anal. Chem. 1997, 69, 24512456. (10) Valaskovic, G. A.; McLafferty, F. W. J. Am. Soc. Mass Spectrom. 1996, 7, 1270-1272. (11) Wilm, M.; Mann, M. Anal. Chem. 1996, 68, 1-8. (12) Kelly, J. F.; Ramaley, L.; Thibault, P. Anal. Chem. 1997, 69, 51-60. (13) Bateman, K. P.; White, R. L.; Thibault, P. Rapid Commun. Mass Spectrom. 1997, 11, 307-315. (14) McComb, M. E.; Krutchinsky, A. N.; Ens, W.; Standing, K. G.; Perrault, H. J. Chromatogr., A 1998, 800, 1-11. (15) Hulthe, G.; Petersson, M. A.; Fogelqvist, E. Anal. Chem. 1999, 71, 29152921. (16) Herring, C. J.; Qin, J. Rapid Commun. Mass Spectrom. 1999, 13, 1-7. (17) Barroso, M. B.; de Jong, A. P. J. Am. Soc. Mass. Spectrom. 1999, 10, 12711278. (18) Bateman, K. P. J. Am. Soc. Mass Spectrom. 1999, 10, 309-317. (19) Barnidge, D. R.; Nilsson, S.; Markides, K. E.; Rapp, H.; Hjort, K. Rapid Commun. Mass Spectrom. 1999, 13, 994-1002.
Analytical Chemistry, Vol. 73, No. 19, October 1, 2001 4607
has been used. Electroplating has also been tried, either as reinforcement of sputtered gold coatings12,13,17,18 or in the form of electroless deposition of silver.8,9 The use of silver or gold conducting epoxy coatings has also been reported.4,7,16 The latter approach is related to the recently proposed fairy-dust technique1 in which gold particles are secured on a fused-silica capillary tip with the help of a polymer, such as polyimide or silicone.20 The lifetimes of the different conductive coatings reported in the literature, except fairy dust, generally range from less than 1 h to more than 100 h.5,6,10,11,13-15,17,18 The fairy-dust coating is an interesting exception. Fairy-dust emitters have been shown1 to last for 2000 h of continuous spraying without visible signs of wear or malfunction. This exceptional stability of the fairy-dust emitters compared to that for thin-layer coated emitters sparked an interest in evaluating the reasons for the differences in performance. Previous studies have clearly demonstrated that electrochemical reactions take place at the electrospray emitter during ESI operation.21-23 An oxidation takes place in the positive electrospray mode while a reduction occurs in the negative electrospray mode. If a sheathless setup is used with a gold emitter, the reactions mainly involve the oxidation and reduction of water, respectively. These electrochemical reactions occurring at the metal-liquid interface may be destructive for two reasons. First, the oxidation or reduction of water causes the formation of oxygen or hydrogen bubbles on the surface that induce mechanical stress upon the coating.24 Second, the conductive coating may itself be oxidized and thereby be stripped from the surface of the ESI emitter. An oxidation of the ESI emitter has been shown to occur for emitters made of stainless steel, zinc, and silver in positive ESI-MS.21 For gold emitters, the oxidation of the emitter has not been reported to be a problem in ESI-MS although it is well known that electrochemical oxidation gives rise to the formation of oxide on gold electrodes.25-27 It is also known that oxidation of gold electrodes in chloride solutions gives rise to loss of gold by the formation of soluble gold chloride complexes.28-30 The influence of such dissolution of the gold coatings on their performance in ESI has apparently not been reported. To our knowledge, only positive electrospray ionization has previously been used to evaluate the stability of sheathless ESI emitters. The aim of the present study is to investigate the electrochemical processes involved in the wear of gold conductive coatings used in ESI-MS. Gold-coated fused-silica ESI emitters were therefore used as working electrodes in an electrochemical cell and the behavior of these coatings, when used both as anodes and cathodes, was investigated using cyclic voltammetry (CV), chronoamperometry (CA), chronopotentiometry (CP), ac imped(20) Nilsson, S.; Markides, K. E. Rapid Commun. Mass Spectrom. 2000, 14, 6-11. (21) Blades, A. T.; Ikonomou, M. G.; Kebarle, P. Anal. Chem. 1991, 63, 21092114. (22) Van Berkel, G. J.; Zhou, F.; Aronson, J. T. Int. J. Mass Spectrom. Ion Processes 1997, 162, 55-67. (23) Van Berkel, G. J. J. Am. Soc. Mass Spectrom. 2000, 11, 951-960. (24) Anderson, J. L.; Winograd, N. In Laboratory Techniques in Electroanalytical Chemistry, 2nd ed.; Kissinger, P. T., Heineman, W. R., Eds.; Marcel Dekker: New York, 1996; Chapter 11. (25) Conway, B. E. Prog. Surf. Sci. 1995, 49, 331-452. (26) Lebreton, C.; Wang, Z. Z. Surf. Sci. 1997, 382, 193-200. (27) Joudkazis, K.; Joudkazyte, J.; Jasulaitiene, V.; Lukinskas, A.; Sebeka, B. Electrochem. Commun. 2000, 2, 503-507. (28) Lebreton, C.; Wang, Z. Z. Appl. Phys. A 1998, 66, S777-S782. (29) Ye, S.; Ishibashi, C.; Uosaki, K. Langmuir 1999, 15, 807-812. (30) Turyan, I.; Matsue, T.; Mandler, D. Anal. Chem. 2000, 72, 3431-3435.
4608
Analytical Chemistry, Vol. 73, No. 19, October 1, 2001
ance measurements, and electrospray durability studies. In these experiments, the electrochemical conditions under which a loss of the electrical contact to the coatings occurred were studied. In addition, scanning electron microscopy (SEM) was used to inspect some of the emitters that had experienced electrochemical or electrospray tests. Furthermore, the concentrations of the coating components in the solutions used for the electrochemical experiments were determined by inductively coupled plasma - mass spectrometry (ICPMS). It is shown that the results of the electrochemical experiments are in good agreement with previously reported findings as well as results of ESI experiments carried out with continuous infusion and positive CE (i.e., with a positive voltage applied to the injection end of the capillary). The present results clearly demonstrate the advantages of using fast electrochemical experiments as a complement to ESI-MS runs in fundamental studies of electrochemical reactions in ESI aiming, for example, at the development of improved sheathless ESI conducting coatings. EXPERIMENTAL SECTION Materials. The experiments were carried out using fused-silica capillaries, 25 µm i.d. × 360 µm o.d., purchased from Polymicro Technologies (Tucson, AZ). Deionized water was obtained from a Milli-Q+ system (Millipore Corp., Marlborough, MA). Suprapur nitric acid and pro analysi hydrochloric acid were obtained from Merck (Darmstadt, Germany). The hydrochloric acid was further cleaned by distillation below the boiling point (sub-boiled) in a distillation apparatus (Heraeus, Germany) prior to use. Gold and the 2-µm gold particles used as coating materials as well as the solid gold wire, 1 mm in diameter (all 99.95% purity), were obtained from Goodfellow (Cambridge, U.K.) while the NiCr alloy (60/40 w/w%, 99.9% purity) was purchased from Cerac Inc. (Milwaukee, WI). Alltech (Deerfield, IL) supplied the polyimide sealing resin. Shaping of Fused-Silica Capillaries. The shaping procedure employed was identical to the procedure described by Barnidge et al.19 The capillary was loaded into a pin vise held in a handheld drill through a piece of Teflon tubing. While the pin vise holding the capillary was rotated in the hand-held drill, pressure was applied to the capillary tip onto a piece of waterproof silicon carbide paper No. 800 (Struers, Denmark) mounted on a piece of circular plastic that was rotating in a drill press. The tips were subjected to shaping for 1 min and were then ready for the appliance of the conductive coatings. Application of Fairy Dust to Capillary Tips. Shaped fusedsilica capillaries were coated with 2-µm gold particles according to the procedure described by Barnidge et al.1 Briefly, a thin layer of polyimide sealing resin was applied onto ∼3 cm of the sharpened capillary end. Gold particles were dusted upon the polyimide with a coating density of ∼1 mg of gold/cm of capillary. The polyimide was then cured in a gas chromatographic oven using a temperature program from 50 to 250 °C (32 °C/min) terminated by a 15-min hold time at 250 °C. Application of NiCr Alloy and Au to Capillary Tips. This coating procedure was identical to the procedure described by Barnidge et al.19 except for the fact that a nickel-chromium alloy was used instead of pure chromium. After cleaning in NH3/H2O2/ H2O (400 mL of 25%/400 mL of 31%/2 L) at 60 °C, the capillaries were mounted in a vacuum evaporator (Edwards FL 400, BOC
Edwards) and heated with an IR lamp to 200 °C during evacuation to 1 × 10-6 mbar. To maintain a water-free surface, the heating was applied until the first layer of metal had been deposited. The NiCr alloy was resistively heated and deposited on the capillaries followed by the deposition of gold in the same manner. The thickness of the deposited layer was approximately 1500 Å for the NiCr alloy and 1500 Å for Au. It should be noted that since the NiCr alloy was evaporated onto the capillary tips, the composition of the deposited coating was not identical to the composition in the alloy used and also changed during the course of the deposition. The alloy was used instead of chromium due to its assumed better corrosion resistance compared to pure chromium. Emitters with only NiCr alloy deposited on them were also manufactured using the procedure described above. In this case, the IR lamp was not used. An additional cleaning in HCl/ H2O2/H2O (333 mL of 37%/333 mL of 31%/2 L) at 60 °C was also performed before the coating step. The thickness of the NiCr alloy film was ∼760 Å. Sputter Coating of Capillary Tips. Gold was sputtered on one side of the capillary tips in a E5000M SEM coater (Bio-Rad, Hercules, CA) using a plasma current of 10 mA, a voltage of 1.2 kV, and an argon pressure of 0.1 mbar for 3 min. The thickness of the obtained coating was ∼350 Å. Electrochemical Instrumentation. Cyclic voltammetric, chronoamperometric, and chronopotentiometric experiments were carried out using a PAR 273 potentiostat (EG&G Princeton Applied Research, Princeton, NJ). The working electrode was either a piece of a solid gold wire or a fused-silica capillary tip with a conductive coating. Four different types of gold-coated capillary tips were studied: (i) gold evaporated on fused-silica capillary tips having an adhesion layer of NiCr alloy, (ii) gold sputtered on fusedsilica capillary tips, (iii) gold sputtered on fused-silica capillary tips lacking a polyimide coating, and (iv) fused-silica capillary tips with fairy-dust coatings. For comparison, fused-silica capillary tips with coatings of only NiCr were also used in some experiments. All emitters, apart from type iii, had a polyimide layer underneath the metal coatings except for in the shaped part of the tip where the metal coating was deposited directly on the fused-silica surface. A glassy carbon rod with a diameter of 3 mm was used as the counter electrode, and the reference electrode was an Ag/AgCl electrode positioned in a bridge containing the electrolyte. Before each experiment, a cyclic voltammogram was recorded in order to verify that a good contact to the emitter was achieved and that the quality of the gold coating was good. Data from the PAR 273 potentiostat were collected using the PAR Electrochemistry Software 4.0 (EG&G Princeton Applied Research). Further evaluation was done on a Macintosh Power PC computer using IgorPro (Wavemetrics, Lake Oswego, OR) and Microsoft Excel 98. The impedance data were obtained with an Autolab PGSTAT 20 (EcoChemie, Utrecht, The Netherlands), equipped with a FRA2 frequency response analyzer module. All experiments were performed in the three-electrode mode using an Ag/AgCl reference electrode and a platinum wire as the counter electrode. The working electrode was either a piece of gold-coated glass or a fused-silica capillary coated with a fairy-dust coating. The goldcoated glass pieces were manufactured by vacuum deposition of a 2000-Å gold film on top of a 10-Å adhesion layer of chromium. These gold surfaces were pressed against a Viton O-ring, sur-
rounding a hole in the wall of a Teflon-based electrochemical cell. The area of the electrode exposed to the solution was determined to 0.20 cm2. The fairy-dust capillaries were dipped 10 mm into a 30-mL solution of 100 mM HNO3 or HCl when used as working electrodes. The impedance data were recorded at 21 frequencies (logarithmic distribution) in the range 2 kHz to 220 Hz, at 0 V versus Ag/AgCl. The interfacial capacitance was obtained by fitting the real and imaginary impedance to an equivalent RC circuit, where R denotes the resistance in series with the total capacitance C. The capacitance was evaluated using the constant phase element procedure.31 The data evaluation was made using the FRA 2.3 software (EcoChemie). Off-Line ESI. The evaluation of the long-term stability of the emitters was made using a benchtop off-line ESI source. In these experiments, the emitters were inserted into a stainless steel capillary from which the emitter extended 10 mm. A flat (4 cm high and 6 cm wide) piece of steel was placed 1 cm from the emitter as a counter electrode. For positive electrospray, a potential of 2.5-3 kV was applied to the stainless steel capillary by a Spellman SL10 high-voltage supply (Hauppauge, NY) while the counter electrode was grounded through a 1-MΩ resistor. For negative electrospray, a reverse polarity setup was used. The voltage drop across the resistor was measured by a laboratorybuilt high-resistance voltmeter, yielding an ESI current sensitivity of 1 mV/nA. The measured current ranged between 75 and 150 nA in all experiments. For infusion studies, an LC pump (Jasco Inc., Easton, MD) was used to continuously infuse buffer solution at a flow rate of 1000 nL/min through the emitter via a split-flow connection. For CE-ESI studies, a second high-voltage supply (Bertan, Hicksville, NY) was employed to apply 30 kV to a 200mL reservoir filled with buffer. To avoid interruptions of the electroosmotic flow (EOF), the buffer level was raised 10 cm above the level of the emitter. Measurements of the ESI current and electrospray voltage and visual inspections with a binocular microscope were all used to evaluate the emitter performance. Some of the emitters were also further studied by SEM. ICPMS Instrumentation. Some of the electrolytes from the electrochemical experiments were sampled for further analysis with ICPMS. In these cases, hydrochloric acid was added to a concentration of 1 M immediately after the sampling to prevent losses of dissolved gold.32 The inductively coupled plasma timeof-flight mass spectrometer (ICP-TOF-MS) used was a Renaissance (Leco, St Joseph, MI). Power, torch position, nebulizer flow, coolant flow, and auxiliary flow were optimized with a 50 ppb indium solution for maximum signal-to-noise ratio. Second-order polynomials with R2 values of more than 0.9997 were obtained for six calibration solutions ranging from 0 to 100 ppb of Au, Ni, and Cr. The confidence intervals for the blank runs (Student’s t-test, p
