Preparation and in Situ Characterization of Surfaces Using Soft

Anal. Chem. 2005, 77, 3452-3460

Preparation and in Situ Characterization of Surfaces Using Soft Landing in a Fourier Transform Ion Cyclotron Resonance Mass Spectrometer Jormarie Alvarez and R. Graham Cooks

Department of Chemistry, Purdue Univeristy, West Lafayette, Indiana 47907 S. E. Barlow, Daniel J. Gaspar, Jean H. Futrell, and Julia Laskin*

Fundamental Science Directorate and Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, Richland, Washington 99352

Mass-selected peptide ions produced by electrospray ionization were deposited onto fluorinated self-assembled monolayer surfaces (FSAM) surfaces by soft landing using a Fourier transform ion cyclotron resonance mass spectrometer (FT-ICR MS) specially designed for studying interactions of large ions with surfaces. Analysis of the modified surface was performed in situ by combining 2-keV Cs+ secondary ion mass spectrometry with FT-ICR detection of the sputtered ions (FT-ICR-SIMS). Regardless of the initial charge state of the precursor ion, the SIMS mass spectra included singly protonated peptide ion, peptide fragment ions, and peaks characteristic of the surface in all cases. In some experiments, multiply protonated peptide ions and [M + Au]+ ions were also observed upon SIMS analysis of modified surfaces. For comparison with the in situ analysis of the modified surfaces, ex situ analysis of some of the modified surfaces was performed by 25-keV Ga+ time-of-flight-secondary ion mass spectrometry (TOF-SIMS). The ex situ analysis demonstrated that a significant number of soft-landed peptide ions remain charged on the surface even when exposed to air for several hours after deposition. Charge retention of soft-landed ions dramatically increases the ion yields obtained during SIMS analysis and enables very sensitive detection of deposited material at less than 1% of monolayer coverage. Accumulation of charged species on the surface undergoes saturation due to coulomb repulsion between charges at close to 30% coverage. We estimated that close to 1 ng of peptide could be deposited on the spot area of 4 mm2 of the FSAM surface without reaching saturation. Interaction of ions with surfaces is an area of active fundamental research in surface science relevant to a broad range of other scientific disciplines such as materials science, mass spectrometry, imaging, and spectroscopy. Various techniques for surface preparation and modification utilizing bombardment by an energetic ion beam (typically >1 keV) in a vacuum environment * To whom correspondence whould be addressed. E-mail: [email protected].

3452 Analytical Chemistry, Vol. 77, No. 11, June 1, 2005

had been developed and commercialized by the early 1980s. These techniques utilize atomic ions that lose their charge and get incorporated into the outer layers of the surface. Alternatively, high-energy ion beams are used for physical modification of surfaces during or prior to film deposition.1 Low-energy (1-100 eV) hyperthermal ion beams have also been used for surface modification. Various physical and chemical processes occurring during interaction of both small and large hyperthermal ions with different surfaces have been extensively reviewed.2-5 This regime of collision energies is dominated by ion scattering and charge loss at surfaces. For example, less than 1% of typical projectile ions survive collisions with metal surfaces. However, neutralization is dramatically reduced when ions interact with organic surfaces such as self-assembled monolayers (SAMs) of thiols on metal substrates or with liquid polymer films.3,6,7 A unique process that occurs during low-energy collisions of polyatomic ions of some organic surfaces is soft landing (SL).8 SL is defined as the intact capture in the condensed phase (surfaces of solids or liquids) of mass-selected polyatomic ions. The term is applied to two distinct processes, one in which the molecular structure remains intact after the collision with a surface and the more demanding process in which the molecular structure remains intact and the ion preserves its charge. This study is concerned with the second phenomenon. Charge retention has been unambiguously proven for small closed-shell ions colliding at fluorinated SAM surfaces (FSAM).8-10 Soft landing of proteins (1) Handbook of Deposition Technologies for Films and Coatings: Science, Technology and Applications, 2nd ed.; Bunshah, R. F., Ed.; William Andrew Publishing/Noyes Publications: Park Ridge, NJ, 1994. (2) Cooks, R. G.; Ast, T.; Mabud, A. Int. J. Mass Spectrom. Ion Processes 1990, 100, 209-265. (3) Dongre, A. R.; Somogyi, AÄ .; Wysocki, V. H. J. Mass Spectrom. 1996, 31, 339-350. (4) Grill, V.; Shen, J.; Evans, C.; Cooks, R. G. Rev. Sci. Instrum. 2001, 72, 31493179. (5) Jacobs, D. C. Annu. Rev. Phys. Chem. 2002, 53, 379-407. (6) Morris, M. R.; Riederer, D. E. J.; Winger, B. E.; Cooks, R. G.; Ast, T.; Chidsey, C. E. D. Int. J. Mass Spectrom. Ion Processes 1992, 122, 181. (7) Cooks, R. G.; Ast, T.; Pradeep, T.; Wysocki, V. Acc. Chem. Res. 1994, 27, 316-323. (8) Miller, S. A.; Luo, H.; Pachuta, S. J.; Cooks, R. G. Science 1997, 275, 1447. (9) Luo, H.; Miller, S. A.; Cooks, R. G.; Pachuta, S. J. Int. J. Mass Spectrom. 1998, 174, 193. 10.1021/ac0481349 CCC: $30.25

© 2005 American Chemical Society Published on Web 05/03/2005

with retention of structure has been observed at FSAM surfaces, while retention of configuration and biological activity (but not charge) has been observed for protein landing at liquid surfaces.11,12 SL can be utilized for a specific modification of surfaces using a beam of mass-selected ions of any size and composition. Furthermore, retention of biological activity by soft-landed proteins opens opportunities for controlled preparation of protein arrays using SL.11 Whereas protein arrays are best suited to investigating protein-protein and protein-ligand interactions, peptide arrays can be used to precisely characterize molecular recognition events at the amino acid level.13 Peptide chips have recently generated widespread interest because many enzymatic processes, including kinase and protease activity, can be readily studied using peptides as model substrates.14 Existing technologies for the production of peptide arrays have been reviewed.13 These are based on a variety of synthetic strategies. The major disadvantage of most of the existing approaches is the lack of specificity and the relatively large quantities of purified material required for the preparation of peptide chips. SL of mass-selected ions on surfaces provides a means for highly specific deposition of peptides in surfaces using only a small fraction of material utilized in standard synthetic approaches. In this work, we present a novel experimental approach to fundamental studies of factors affecting SL of peptide ions on surfaces. Surfaces are modified by a beam of mass-selected ions of varying kinetic energy in a Fourier transform ion cyclotron resonance mass spectrometer (FT-ICR MS). The final kinetic energy of the precursor ions is changed electrostatically in a strong magnetic field, which ensures that ion beam trajectories are unperturbed by the deceleration optics and are independent of collision energy.15 This is important for quantitative studies of the effect of the initial kinetic energy of the incoming ions in SL. Modified surfaces were probed in situ by combining 2-keV Cs+ secondary ion mass spectrometry with FT-ICR detection of the sputtered ions (FT-ICR-SIMS). For comparison, ex situ analysis of some of the modified surfaces was performed using 25-keV Ga+ time-of-flight-secondary ion mass spectrometry (TOF-SIMS). Several groups have used SIMS inside the ICR cell as a means of generating ions and introducing them into the mass spectrometer.16-20 However, utilization of in-cell ionization techniques has been greatly reduced with the advent of soft ionization (10) Shen, J.; Yim, Y. H.; Feng, B.; Grill, V.; Evans, C.; Cooks, R. G. Int. J. Mass Spectrom. 1999, 183, 423-435. (11) Ouyang, Z.; Takats, Z.; Blake, T. A.; Gologan, B.; Guymon, A. J.; Wiseman, J. M.; Oliver, J. C.; Davisson, V. J.; Cooks, R. G. Science 2003, 301, 1351. (12) Gologan, B.; Taka´ts, Z.; Alvarez, J.; Wiseman, J. M.; Talaty, N.; Ouyang, Z.; Cooks, R. G. J. Am. Soc. Mass Spectrom. 2004, 15, 1874. (13) Reimer, U.; Reineke, U.; Schneider-Mergener, J. Curr. Opin. Biotechnol. 2002, 12, 315. (14) Houseman, B. T.; Huh, J. H.; Kron, S. J.; Mrksich, M. Nat. Biotechnol. 2002, 20, 270. (15) Laskin, J.; Denisov, E. V.; Shukla, A. K.; Barlow, S. E.; Futrell, J. H. Anal. Chem. 2002, 74, 3255. (16) Castro, M. E.; Russell, D. H. Anal. Chem. 1984, 56, 578-581. (17) Castro, M. E.; Russell, D. H. Anal. Chem. 1985, 57, 2290-2293. (18) Castro, M. E.; Mallis, L. M.; Russell, D. H. J. Am. Chem. Soc. 1985, 107, 5652-5657. (19) Amster, I. J.; Loo, J. A.; Furlong, J. J. P.; McLafferty, F. W. Anal. Chem. 1987, 59, 313-317. (20) Hill, N. C.; Limbach, P. A.; Shomo, R. E.; Marshall, A. G. Rev. Sci. Instrum. 1991, 62, 2612-2617.

Figure 1. Schematic view of the FT-ICR mass spectrometer used for SL experiments.

techniques such as electrospray (ESI)21 and matrix-assisted laser desorption/ionization22 and the subsequent development of approaches for introducing externally generated ions into the ICR cell.23 In addition, reduced ion yields of high-mass ions in SIMS, combined with the smaller dynamic range of the FT-ICR MS as compared to TOF-SIMS, present serious limitations for the utility of FT-ICR SIMS for analysis of large molecules. The approach adopted in this study is quite different. We utilize FT-ICR SIMS for analysis of preformed ions deposited on the surface by SL, which provides a unique opportunity for in situ characterization of modified surfaces and eliminates possible ion loss between the ion deposition and the analysis step. EXPERIMENTAL SECTION Soft Landing of Peptides Using an ESI-FT-ICR Mass Spectrometer. SL experiments were performed in a custom-built 6-T FT-ICR mass spectrometer (Pacific Northwest National Laboratory, Richland, WA) described elsewhere (Figure 1).15 A syringe pump (Cole Parmer, Vernon Hills, IL) was used for direct infusion of the sample at a flow rate of 30 µL/h. Protonated peptides are formed in an external ESI source. Efficient transmission of ions into the vacuum system is achieved using an electrodynamic ion funnel. After the ion funnel, the ions undergo collisional relaxation in a collisional quadrupole (CQ) followed by mass selection using a commercial Extrel (Pittsburgh, PA) quadrupole mass filter (resolving quadrupole, RQ). The massselected ions are transmitted through a third quadrupole (accumulation quadrupole, AQ) operated in the rf-only mode and pass into an electrostatic ion guide that consists of a series of five tube lenses that enable precise positioning and shaping of the ion beam. An electrostatic quadrupole bender located after the second tube lens turns the ion beam by 90° to avoid contamination of the surface by neutrals. After exiting the last tube lens, mass-selected ions are transported into the ICR cell through a long flight tube. Ions are decelerated inside the strong magnetic field by two deceleration plates located at the entrance of the ICR cell. SL occurs when selected ions collide at various energies with a (21) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Science 1989, 246, 64. (22) Hillenkamp, F.; Karas, M.; Beavis, R. C.; Chait, B. T. Anal. Chem. 1991, 63, A1193. (23) Marshall, A. G. Int. J. Mass Spectrom. 2000, 200, 331 and references therein.

Analytical Chemistry, Vol. 77, No. 11, June 1, 2005

3453

Figure 2. Kinetic energy distributions of ions produced in the ESI source.

surface positioned at the rear trapping plate of the ICR cell. This surface is introduced through a vacuum interlock assembly. During SL experiments, static dc potentials are applied to the front trapping plate, the ring electrode, and the rear trapping plate of the ICR cell. The collision energy of the ions was adjusted by changing the dc offset applied to the rear trapping plate of the ICR cell. Because the final kinetic energy of the precursor ions is changed in the strong magnetic field, ion trajectories are unperturbed by the deceleration optics. This ensures that the shape of the beam and the ion flux do not depend on the kinetic energy of the ionssa unique feature of this experimental configuration. Ion kinetic energy per unit charge is determined by the potential difference between the ion source and the surface. The initial kinetic energy of ions was measured using the retarding potential method by systematically varying the potential on the rear trapping plate of the ICR cell and monitoring the ion current on the surface. In this experiment, the surface was pulled back from the ICR cell by ∼100 mm. The resulting signal was differentiated to give the distribution of kinetic energies shown in Figure 2. The distribution peaks around 14 eV and has a full width at half-maximum of ∼5 eV. A small peak at lower kinetic energies is the result of ion collisions with background gas in the extraction region of the CQ. The SL spot size was determined to be ∼4 mm2 by TOF-SIMS imaging of the surface after SL, using a Ga+ beam with a cross section of 0.01 mm2. SL experiments utilized ion currents in the range of 1-30 pA. Peptides were purchased from Sigma and used as received. Samples were dissolved in a 70:30 (v/v) methanol/ water solution with 1% acetic acid to a concentration of 0.1 mg/ mL. Mass-to-charge ratios are reported using the thomson unit (1 Th ) 1 Da/unit charge).24 In Situ Analysis of Modified Surfaces Using FT-ICR-SIMS. The modified surface was subjected to in situ Cs+ ion desorption analysis within 5 min of completion of the ion SL experiment. The last of the five tube lenses (lens 5) of the FT-ICR electrostatic ion guide is mounted in a custom-built moving stage, which is used to interchange it with a cesium ion gun for SIMS analysis of the surface. Primary Cs+ ions are generated with a cesium ion gun, transported into the ICR cell through a long flight tube, and collided with the surface positioned at the rear trapping plate of the ICR cell. Static SIMS conditions with a total ion flux of ∼1.5 (24) Cooks, R. G.; Rockwood, A. L. Rapid Commun. Mass Spectrom. 1991, 5, 93.

3454 Analytical Chemistry, Vol. 77, No. 11, June 1, 2005

Figure 3. Schematic view of voltages applied to the ICR cell in SIMS experiments.

× 1010 ions/cm2 (current 50 nA, duration 100 µs, spot size 0.05 cm2, 25 shots) were achieved when the following potentials were applied to various focusing elements: Cs+ gun floating voltage, +2000 V; Cs+ ion gun shield, +1909 V; flight tube, -650 V; and deceleration plates, 0 V. Data acquisition was accomplished with a MIDAS data station.25 The Cs+ ion beam was pulsed by alternating the potential applied to the flight tube between -650 and +2300 V, the high value being used to block the Cs+ beam from reaching the surface. SIMS spectra were averaged over 5-25 shots. Figure 3 summarizes the event sequence employed in SIMS experiments. The ring electrode of the ICR cell was kept at 0 V throughout the experiment. The potentials on the trapping plates of the ICR cell were optimized for the best SIMS signal. During the Cs+ pulse, the front trapping plate of the ICR cell was kept at +7 V, while the surface and the rear trapping plate were held at +2 V. The same potentials were applied to the trapping plates during the recoil event, when the Cs+ gun was switched off. This choice of trapping potentials ensures that ions sputtered off the surface are slightly accelerated toward the center of the cell and reflected back into the cell by the potential difference between the surface and the front trapping plate of the cell. This procedure improves collection of ions over a broad range of masses and kinetic energies. The optimum Cs+ pulse width was determined experimentally to be 100 µs. After the recoil event, the sputtered ions were captured by raising the potentials on the front and rear trapping plates of the ICR cell to 30 V to ensure efficient trapping of ions with relatively high kinetic energies. It should be noted that such high trapping voltages are 5-10 times higher than commonly used trapping voltages. However, our cylindrical ICR cell was specifically designed to eliminate the fourth-order term in the electrostatic trapping field and is a nearly perfect quadratic field trapping cell. It has been demonstrated that such a quadratic trap can be operated with much higher trapping voltages than conventional ICR cells without impairing cell performance.26 The sensitivity of the measurement was further enhanced by adding two excitation eventssto expel Cs+ (133 Th) and Au+ (197 Th) ions from the ICR cellsprior to the final excitation/detection event. The time elapse between the trapping event and the excitation/detection event was ∼0.4 s. (25) Senko, M. W.; Canterbury, J. D.; Guan, S.; Marshall, A. G. Rapid Commun. Mass Spectrom. 1996, 10, 1839. (26) Barlow, S. E.; Tinkle, M. Rev. Sci. Instrum. 2002, 73, 4185.

Figure 4. FT-ICR SIMS spectra of the FSAM surface (a) in the positive mode and (b) in the negative mode.

Ex Situ Analysis of Modified Surfaces Using a TOF-SIMS Mass Spectrometer. Ex situ analysis of the surfaces modified in the FT-ICR was performed using a PHI TRIFT TOF-SIMS mass spectrometer (Physical Electronics, Eden Prairie, MN) available at the Pacific Northwest National Laboratory. High-resolution 25keV Ga+ TOF-SIMS analysis was performed using a beam current of 60 pA for 120 s at a frequency of 10 kHz and a primary pulse width of 1 ns. The experimental conditions were set for static SIMS analysis at an ion flux of ∼4.5 × 109 ions/cm2 and a spot size of 10-4 cm2. To confirm that static SIMS conditions were in fact used, a modified surface was bombarded for a period of 30 min. The signal of the peptide was monitored during this time and only a 30% decrease in the signal was seen during this period, which is far longer than the 2 min used to acquire data on the modified surfaces. Self-Assembled Monolayer Surfaces. Fluorinated self-assembled monolayer surfaces were used as targets for SL experiments. The surfaces were prepared following literature procedures.27 In this study, CF3(CF2)9(CH2)2SH was used to form the self-assembled monolayer by exposure of the gold surface (International Wafer Service, Portola Valley, CA) to an ethanol solution of the thiol for 24-36 h. The surface was removed from the SAM solution and ultrasonically washed in ethanol for 5 min to remove extra layers. Sequentially the surface was dried under nitrogen gas before being introduced into the instrument. RESULTS AND DISCUSSION Surface Characterization Using FT-ICR SIMS. Positive and negative SIMS spectra of the FSAM surface recorded by Cs+ bombardment are shown in Figure 4. In the positive mode, the SIMS spectrum shows strong peaks corresponding to gold cluster ions (Au2+ and Au3+) and gold/sulfur cluster ions (Au2SH+ and Au3S+). The presence of these latter ions is evidence of the strength of the Au-S bond at the metal-thiol interface. A large variety of gold-containing ions such as AuCF2+ and Au2F+ are also observed upon Cs+ desorption ionization of the FSAM surface (27) Chidsey, C.; Liu, G.; Rowntree, P.; Scoles, G. J. Chem. Phys. 1989, 91, 4421.

but with low relative intensities. This agrees with previous static SIMS studies of fluorinated SAM surfaces performed in a quadrupole-based instrument with Xe+ at a primary ion beam energy of 7 keV.28 The negative ion static SIMS mass spectrum also shows Au/fluorine-containing fragments, including prominent ions such as AuF2-, Au2F-, and Au2SF-; however, no molecular ions are present in the spectrum in contrast to the case for alkanethiol (M ) CnHn+1SH) surfaces on gold.28 In agreement with the literature,28 SIMS spectra of SAM surfaces of alkanethiol (M ) CnHn+1SH) on gold show a rich variety of Au-molecular cluster ions such as Au[M - H]2-, AuM-, AuSM-, and Au2[M - H]- in addition to gold-sulfur cluster ions such as AuS-, AuS2H-, and Au2S-. A number of experimental parameters have been optimized to improve the collection and detection of high-mass ions. The time delay between switching off the Cs+ pulse and gated trapping of sputtered ionssthe recoil timesdetermines the range of m/z values that reside close to the center of the cell at the time when the potentials on the trapping plates are increased for trapping. Figure 5a shows the dependence of the SIMS signal for ions of different masses on the recoil time. At short delays, SIMS spectra are dominated by low-mass ions represented by CF3+ and C2F5+ in the figure. The abundance of low-mass ions rapidly decreases at longer recoil times while the intensity of higher mass ions maximizes at longer recoil times of ∼50 µs. The optimal recoil time for ions of a range of m/z values is shown in Figure 5b. In this study, SIMS analysis of surfaces was typically carried out using a 50-80 µs recoil time. It follows that ions with m/z