pubs.acs.org/Langmuir © 2010 American Chemical Society
Adsorption of Histidine and Histidine-Containing Peptides on Au(111) Vitaliy Feyer,*,† Oksana Plekan,†,^ Nataliya Tsud,‡ Vladimı´ r Chab,§ Vladimı´ r Matolı´ n,‡ and Kevin C. Prince† †
Sincrotrone Trieste S.C.p.A., in Area Science Park, Strada Statale 14, km 163.5, I-34012 Basovizza, Trieste, Italy, ‡Charles University, Faculty of Mathematics and Physics, Department of Surface and Plasma Science, V Holesovi ck ach 2, 18000 Prague 8, Czech Republic, and §Institute of Physics, Academy of Sciences of the Czech Republic, Cukrovarnick a 10, 16253 Prague 6, Czech Republic. ^ Present address: Aarhus University, Department of Physics and Astronomy, Ny Munkegade 120, 8000 Aarhus C, Denmark Received December 12, 2009. Revised Manuscript Received March 8, 2010
The adsorption of histidine (His) and three His-derived peptides on Au(111) has been studied by soft X-ray photoelectron spectroscopy (XPS) and near-edge X-ray absorption fine structure spectroscopy (NEXAFS) at the nitrogen and oxygen K edges. The peptides were glycyl-histidine (Gly-His), glycyl-histidine-glycine (Gly-His-Gly), and glycyl-glycyl-histidine (Gly-Gly-His) and were adsorbed at saturated coverage on the Au(111) surface from aqueous solution. Coverages of 1 and 0.5 monolayers (ML) of His were adsorbed by evaporation in vacuum and compared with 1 ML deposited from solution. There were no significant chemical differences between the monolayers deposited in vacuum or from solution. The Au 4f core level shift indicates that a chemisorption rather than a physisorption bond is formed. In both deposited phases, His bonds to the gold surface in anionic form via the imino nitrogen atom of the imidazole ring and the oxygen atoms of the carboxylate group. N and O K-edge NEXAFS indicate that the ring and carboxylate triangle of adsorbed His are tilted at ∼35 and ∼27, respectively, with respect to the Au(111) surface. The peptides bond to the gold surface in a mode similar to the single His molecule, via the imino and carboxylate groups, while the peptide group is at a steep angle to the surface. However, the peptides adsorb with a higher atomic density, consistent with the peptide groups being above the surface. There are also differences between Gly-His-Gly and Gly-Gly-His, implying that the sequence within the peptide has a significant influence on the bonding geometry.
1. Introduction Interfaces between metals and biologically active molecules are important topics in biocatalysis, biocompatibility, and biosensors. There are many such organic molecules and amino acids; their oligomers (peptides) and polymers (proteins) represent an important class. There are many possible applications, and one example is the electrochemical detection of metal ions, demonstrated by Yang et al.,1 who made a sensor with sub-ppt detection limits by attachment of the tripeptide Gly-Gly-His to a gold electrode. In this device, His is of prime importance as it forms a complex with the metal ion. Histidine is of general biochemical interest, not only in proteins but also as a precursor of histamine, a biogenic amine involved in local immune responses. Because of the presence of the imidazole side chain, His and its peptides may also have important applications as corrosion inhibitors.2 Histidine consists of three functional groups, amino, imidazole (IM), and carboxylic acid, and they are potential binding sites to surfaces. In addition, its peptides contain peptide groups (see Figure 1) which may also play a significant role in molecular bonding with a surface. A variety of experimental surface science techniques such as reflection-absorption infrared spectroscopy (RAIRS), photoelectron diffraction, XPS, and NEXAFS, together with theoretical approaches such as density functional theory (DFT), have been used to study the adsorption of amino acids on gold and *Corresponding author: Tel þ39 0403758287; fax þ39 0403758565; e-mail
[email protected]. (1) Yang, W.; Jaramillo, D.; Gooding, J. J.; Hibbert, D. B.; Zhang, R.; Willett, G. D.; Fisher, K. J. Chem. Commun. 2001, 1982. (2) Xue, G.; Dong, J.; Sun, Y. Langmuir 1994, 10, 1477.
8606 DOI: 10.1021/la904684e
copper surfaces.2-13 Gold is the least reactive of the noble metals,14 while copper is an interesting metal to compare with gold due to its higher reactivity.3,8,9 Many complex molecules, such as peptides, have been deposited on well-characterized metal surfaces from the liquid phase,15-19 while some di- and tripeptides have been successfully evaporated (3) Liedberg, B.; Carlsson, C.; Lundstr€om, I. J. Colloid Interface Sci. 1987, 120, 64. (4) Hasselstr€om, J.; Karis, O.; Weinelt, M.; Wassdahl, N.; Nilsson, A.; Nyberg, M.; Pettersson, L. G. M.; Samant, M. G.; St€ohr, J. Surf. Sci. 1998, 407, 221. (5) Nyberg, M.; Hasselstr€om, J.; Karis, O.; Wassdahl, N.; Weinelt, M.; Nilsson, A.; Pettersson, L. G. M. J. Chem. Phys. 2000, 112, 5420. (6) Dodero, G.; De Michieli, L.; Cavalleri, O.; Rolandi, R.; Oliveri, L.; Dacca, A.; Parodi, R. Colloids Surf., A 2000, 175, 121. (7) Barlow, S. M.; Raval, R. Surf. Sci. Rep. 2003, 50, 201. (8) Marti, M. E.; Methivier, Ch.; Dubot, P.; Pradier, C. M. J. Phys. Chem. B 2003, 107, 10785. (9) Marti, M. E.; Quash, A.; Methivier, Ch.; Dubot, P.; Pradier, C. M. Colloids Surf., A 2004, 249, 85. (10) Zubavichus, Y.; Zharnikov, M.; Yang, Y.; Fuchs, O.; Heske, C.; Umbach, E.; Tzvetkov, G.; Netzer, F. P.; Grunze, M. J. Phys. Chem. B 2005, 109, 884. (11) Jones, J.; Jones, L. B.; Thibault-Starzyk, F.; Seddon, E. A.; Raval, R.; Jenkins, S. J.; Held, G. Surf. Sci. 2006, 600, 1924. (12) Iori, F.; Corni, S.; Di Felice, R. J. Phys. Chem. C 2008, 112, 13540. (13) Feyer, V.; Plekan, O.; Skala, T.; Chab, V.; Matolı´ n, V.; Prince, K. C. J. Phys. Chem. B 2008, 112, 13655. (14) Hammer, B.; Nørskov, J. K. Nature 1995, 376, 238. (15) Baas, T.; Gamble, L.; Hauch, K. D.; Castner, D. G.; Sasaki, T. Langmuir 2002, 18, 4898. (16) Chow, E.; Wong, E. L. S.; B€ocking, T.; Nguyen, Q. T.; Hibbert, D. B.; Gooding, J. J. Sens. Actuators, B 2005, 111-112, 540. (17) Cho, Y.; Ivanisevic, A. J. Phys. Chem. B 2005, 109, 6225. (18) Monti, S.; Carravetta, V.; Battocchio, C.; Iucci, G.; Polzonetti, G. Langmuir 2008, 24, 3205. (19) Polzonetti, G.; Battocchio, C.; Dettin, M.; Gambaretto, R.; Di Bello, C.; Carravetta, V.; Monti, S.; Iucci, G. Mater. Sci. Eng., C 2008, 28, 309.
Published on Web 03/25/2010
Langmuir 2010, 26(11), 8606–8613
Feyer et al.
Article
may change as a function of coverage due to intermolecular interaction.25
2. Experiment 2.1. Experimental Methods. The experiments were per-
Figure 1. Schematic structure of (a) His, (b) Gly-His, (c) Gly-HisGly, and (d) Gly-Gly-His.
under UHV conditions.20-22 There have also been many theoretical calculations, for example [ref 23 and references therein]. It is well-known that biomolecules are thermally sensitive, so liquid phase deposition is a “soft” method, and compared with evaporation, it is a step closer to possible “real world” applications. One of the questions addressed in this paper is whether there are differences in the interfacial bonding of adlayers obtained by evaporation and deposition. Depending on the pH, amino acids and peptides may interact with the surface as anions, cations, or neutrals. Vallee et al.24 showed that from acidic solution glutathione (Glu-Cys-Gly) adsorbs on the Au(111) surface in its cationic form. At moderate pH, the molecules were found to be mostly zwitterionic, while from basic solution the dominant species observed was anionic. Recently, we have investigated the electronic structure and adsorption geometry of His on Cu(110) surfaces using XPS and NEXAFS13 and provided additional information to previously published but controversial RAIRS studies.8,9 The nature of the bonding of His to gold is also controversial. High-resolution XPS has indicated that in the monolayer film His bonds to polycrystalline, preferentially (111) oriented, gold as histidinate anions, forming strong ionic-covalent bonds with the gold substrate via the amino group, imino nitrogen atom in the IM ring, and the carboxylate (COO-) group.10 A study of His on Au(111) by RAIRS concluded that histidine interacts with the surface only via the two oxygen atoms of the COO- group, and the imidazole ring is almost normal to and far from the surface.9 Using the same method, Liedberg et al.3 found that in a few monolayers prepared by adsorption of His from solution on a gold film the histidine remained zwitterionic and adsorbed solely via one oxygen atom of the carboxylate group. Given this controversy, it appears that the adsorption geometry of His on Au(111) surfaces is not fully understood. The present work focuses on the adsorption of His evaporated in UHV as well as His and His-containing peptides deposited from aqueous solutions onto Au(111) and on the characterization of electronic structure and geometry by a combination of XPS and NEXAFS. We have prepared and analyzed the surfaces at saturation monolayer coverage, and also at submonolayer coverage for some samples, because bonding (20) Barlow, S. M.; Haq, S.; Raval, R. Langmuir 2001, 17, 3292. (21) Vallee, A.; Humblot, V.; Methivier, C.; Pradier, C. M. Surf. Sci. 2008, 602, 2256. (22) Vallee, A.; Humblot, V.; Methivier, C.; Pradier, C. M. J. Phys. Chem. C 2009, 113, 9336. (23) Heinz, H.; Farmer, B. L.; Pandey, R. B.; Slocik, J. M.; Patnaik, S. S.; Pachter, R.; Naik, R. R. J. Am. Chem. Soc. 2009, 131, 9704. (24) Vallee, A.; Humblot, V.; Methivier, C.; Pradier, C. M. Surf. Interface Anal. 2008, 40, 395.
Langmuir 2010, 26(11), 8606–8613
formed at the Materials Science Beamline at the Elettra synchrotron light source in Trieste.26 The beamline is equipped with a plane mirror monochromator providing synchrotron light in the energy range of 40-800 eV, a Specs Phoibos 150 hemispherical electron energy analyzer, low-energy electron diffraction (LEED) optics, and dual-anode X-ray source. During the experiments the base pressure in the main chamber was in the low 10-10 mbar range. The Au 4f core level spectra were recorded at 120 eV of photon energy in normal emission geometry (incidence/emission angles of 60/0), and the total resolution (analyzer þ beamline) was 0.15 eV. The C 1s and N 1s XPS spectra were collected in the same geometry, and the photon energy and total resolution were 500 and 0.45 eV, respectively. The O 1s core level spectra were measured with the same Phoibos analyzer using Mg KR radiation as the excitation source and the total energy resolution was 0.85 eV. This source was also used to measure C 1s and N 1s spectra during preparation for the synchrotron radiation experiments. The binding energy (BE) was calibrated by measuring the Fermi edge. The NEXAFS spectra were taken at the N and O K edges using the nitrogen and oxygen KVV Auger yield, at normal (NI, 90) and grazing (GI, 10) incidence of the photon beam with respect to the surface. The energy resolution for the N and O Kedge spectra was estimated to be 0.35 and 0.8 eV, respectively. The polarization of light from the beamline has not been measured but is believed to be between 80% and 90% linear, as the source is a bending magnet. The raw NEXAFS data were normalized to the intensity of the photon beam, measured by means of a hightransmission gold mesh and divided by corresponding spectra of the clean sample, recorded under identical conditions. 2.2. Sample Preparation. The sample was a gold disk of 10 mm diameter and 2 mm thickness with (111) surface orientation supplied by MaTeck. It was cleaned in situ using standard procedures: cycles of Arþ sputtering (kinetic energy 1.0 keV), followed by annealing at 773-873 K. The surface order and cleanliness were monitored by LEED and XPS. Contaminants (such as C, N, and O) were below the detection limits. The compounds, His, Gly-His, Gly-His-Gly, and Gly-Gly-His, with the highest commercially available purities, were obtained from Sigma-Aldrich and used without further purification. For samples evaporated in UHV, His was deposited on the surface in a separate preparation chamber, with base pressure 5 10-9 mbar (mainly water), using a homemade Knudsen cell type evaporator. Before the deposition, the His powder was degassed in vacuum at ∼373 K, then heated to ∼420 K, and dosed onto the surface. The deposition rate was ∼1.0 monolayer (ML) in 600 s, corresponding to a local pressure of the order of 5 10-8 mbar, and the integrated intensities (areas of peaks) of C 1s, N 1s, and O 1s spectra increased linearly with time. The single ML, vacuum-deposited sample was prepared by adsorbing a multilayer of His on the surface and then flashing to 400 K to desorb the weakly adsorbed His species. This temperature is a plateau in the desorption sequence, in the sense that the coverage obtained is the same over the temperature range 375-425 K. This coverage was then defined as 1.0 ML and characterized by the integrated intensities (areas of peaks) of C 1s, N 1s, and O 1s measured with Mg KR excitation. The submonolayer coverage of His was formed by reducing the deposition time, and the coverage was determined by comparing the XPS signals with those of the 1 ML sample. (25) Feyer, V.; Plekan, O.; Prince, K. C.; Sutara, F.; Skala, T.; Chab, V.; Matolı´ n, V.; Stenuit, G.; Umari, P. Phys. Rev. B 2009, 79, 155432. (26) Vasina, R.; Koları´ k, V.; Dolezel, P.; Mynar, M.; Vondracek, M.; Chab, V.; Slezak, J.; Comicioli, C.; Prince, K. C. Nucl. Instrum. Methods Phys. Res. A 2001, 467-468, 561.
DOI: 10.1021/la904684e
8607
Article The samples with adsorption from the liquid phase were prepared from saturated aqueous solutions, made from distilled water and the compound. The pH of the solutions was 8 (His), 7 (His-Gly) 6 (Gly-His-Gly), and 7 (Gly-Gly-His) with an error of (0.5. The solution was prepared and all deposition steps were performed under a nitrogen atmosphere in a glovebag connected to a fast entry lock of the chamber. The Au(111) crystal, after preparation in UHV, was withdrawn from the chamber via the fast entry lock. A drop of the prepared solution was placed on the clean gold surface for 5 min, then rinsed with distilled water, and dried under nitrogen gas flow, after which the gold crystal was transferred back into vacuum. Gly-His-Gly adsorbed on the surface under these deposition conditions only at saturated coverage, and no multilayers were observed. However, for all other molecules adsorbed from the liquid phase, multilayers formed. For these molecules a saturated coverage (close to the ML regime) was obtained using the following procedure. First we adsorbed multilayers on the surface and then flashed from 325 to 500 K in steps of 25 K; at 400-425 K the weakly adsorbed species desorbed. We assumed that the measured XPS signals (areas of peaks) of C, N, and O 1s after flashing to 400-425 K correspond to the saturated coverage. Then the saturated coverage was prepared by deposition of a multilayer, followed by rinsing several times with distilled water, and the core level signals were similar to those found after heating to 400-425 K. This coverage was further studied using synchrotron radiation. Under some circumstances, carbon contamination was observed and was easily identifiable in the C 1s spectra. Subsequent heating to 450 K of the single ML His samples, deposited both in vacuum and from the liquid phase, and Hiscontaining peptides did not change the areas of the C, N, and O 1s XPS signals (measured with Mg KR radiation), while annealing to higher temperatures led to partial decomposition and desorption of molecules from the surface, which was indicated by a decrease of intensity and substantial changes in the XPS spectra. Checks for radiation damage were done by monitoring the valence band and XPS spectra, and no spectral changes were observed after 1 h. Thus, His and His-derived peptide films are reasonably stable under our experimental conditions.
3. Results and Discussion 3.1. Core Level XPS Results. The Au 4f7/2 core level spectra of the surface before and after adsorption of 1 ML of His in UHV and Gly-His-Gly from the liquid phase are shown in Figure 2. For the other compounds the spectra show similar behavior. The two components in the spectra of the clean surface are due to the bulk (83.98 eV) and surface (83.69 eV) states. After adsorption of one ML of histidine or histidine-derived peptide on Au(111), the surface state of the clean surface shifts and/or decreases in intensity, which indicates most surface Au atoms are affected by the adsorption of histidine-containing molecules. This energy shift suggests fairly strong adsorbate-substrate interaction (chemisorption rather than physisorption). The observed energy shift is much smaller than that in the spectra of Zubavichus et al.10 measured at a photon energy of 610 eV. Because of the higher kinetic energy of the photoelectrons in that work (526 eV), their spectra should be less surface sensitive than the present spectra (36 eV kinetic energy). The difference in surface core level shift may be due to the different substrates, polycrystalline Au-covered silicon wafers with preferential (111) orientation in ref 10 and a Au(111) monocrystal in the present work. The C, N, and O 1s core level spectra of 0.5 and 1 ML coverage of histidine adsorbed in UHV are shown in Figure 3. Both C 1s core level spectra show two strong features A and B centered at 288.1 and 285.85 eV with intensity ratio A:B = 1:5. The peak A is assigned to the core level of the carboxylate group carbon, and the energy is similar to that found in other adsorbed amino 8608 DOI: 10.1021/la904684e
Feyer et al.
Figure 2. Au 4f7/2 core level spectra of the clean surface and after (a) adsorption of His in UHV and (b) adsorption of Gly-His-Gly from the liquid phase. Photon energy 120 eV.
acids.4,10,11,13 This peak has a fwhm ∼1.1 eV, and as mentioned above, the resolution was 0.45 eV, so subtraction in quadrature gives estimated intrinsic widths of ∼1.0 eV, related to the unresolved vibrational structure, natural lifetimes of the respective core-hole states, and solid-state effects. The expected spread in the BE for other carbon atoms is about ∼1.7 eV,27,28 but due to the unresolved vibrational structure, these carbon atoms form the unresolved feature B. A weak shoulder on the higher energy side of peak B is possibly due to carbon atoms with C-N bonds. The energy separation from the center of the two peaks A and B in the C 1s spectra is about 2.25 eV, and it is significantly smaller than that in the gas phase spectra of amino acids,29,30 where the differences between the carboxylic (COOH) and other carbon core levels were about 3-4 eV. This is further evidence that the adsorbed species contains a carboxylate (COO-) rather than the neutral carboxylic (COOH) group. This in turn implies that His interacts with the surface in its anionic and/or zwitterionic form and excludes the cationic and neutral forms which contain carboxylic groups. Both features A and B shift ∼0.3 eV to lower BE at 0.5 ML compared with 1 ML of His. We attribute this shift to final state screening, which is more efficient in thinner films.10 The peak B in the submonolayer is slightly broader (fwhm = 1.55 eV) compared with the ML (fwhm = 1.45 eV), which may indicate some disorder or heterogeneity in the bonding sites of histidine at submonolayer coverage (see Figure 3). The C 1s core level spectra of His, Gly-His, Gly-His-Gly, and Gly-Gly-His adsorbed from the liquid phase on Au(111) are shown in Figure 4a, and two prominent features labeled A and B as well as the weak shoulder C are present. The spectrum of His adsorbed from aqueous solution (see Figure 4a) is very similar to that of 1 ML adsorbed by UHV evaporation (see Figure 3a), so the features A and B are attributed to the carbon atoms as (27) Apen, E.; Hitchcock, A. P.; Gland, J. L. J. Phys. Chem. 1993, 97, 6859. (28) Kaznacheyev, K.; Osanna, A.; Jacobsen, C.; Plashkevych, O.; Vahtras, O.; Agren, H.; Carravetta, V.; Hitchcock, A. P. J. Phys. Chem. A 2002, 106, 3153. (29) Plekan, O.; Feyer, V.; Richter, R.; Coreno, M.; de Simone, M.; Prince, K. C.; Carravetta, V. J. Phys. Chem. A 2007, 111, 10998; 2008, 112, 7806. (30) Zhang, W.; Carravetta, V.; Plekan, O.; Feyer, V.; Richter, R.; Coreno, M.; Prince, K. C. J. Chem. Phys. 2009, 131, 035103.
Langmuir 2010, 26(11), 8606–8613
Feyer et al.
Article
Figure 3. C 1s (a), N 1s (b), and O 1s (c) core level spectra of histidine adsorbed by evaporation in UHV onto Au(111).
discussed above. Also, the ratio A:B = 1:4.5 is close to the observed intensity ratio in the 1 ML spectrum. The dipeptide GlyHis contains two more carbon atoms than His, one of which is located in the peptide CONH moiety while the other is bonded to the amino group (see Figure 1b). The ratio of A:B in the C 1s spectrum is 1:3, which suggests that the carbon atom of the peptide group contributes to the A feature and the other additional carbon atom in Gly-His to the peak B. The C 1s spectra of Gly-His-Gly and Gly-Gly-His further support the proposed assignment as the intensity of the feature A increases with respect to the low binding energy peak B (A:B = 1:2.5) due to the higher number of carbon atoms in peptide moieties in tripeptides. This assignment is also in agreement with that proposed by Chatterjee et al.,31 who reported photoelectron spectra of Gly and its peptides in powder form. We relate the weak feature C in the C 1s spectra of all molecules adsorbed from aqueous solution to carbon containing impurities, and the BE of this peak suggests that they are hydrocarbons. The intensity of feature C is about 1-3% of the total. It is important to note that this feature was stronger when lower than saturated coverage was deposited on the surface. However at saturated coverage, the His-containing molecules prevent the adsorption of impurities and form an almost uniform layer. The N 1s spectra of histidine adsorbed at different coverages in UHV are shown in Figure 3b. The spectrum of the multilayer (not shown) resembles the multilayer spectrum of histidine measured on polycrystalline gold and Cu(110).10,13 There are two strong features A and B in the N 1s spectra of the 1 and 0.5 ML samples of His adsorbed in UHV (see Figure 3b). The intensity ratios A:B (31) Chatterjee, A.; Zhao, L.; Zhang, L.; Pradhan, D.; Zhou, X.; Leung, K. T. J. Chem. Phys. 2008, 129, 105104.
Langmuir 2010, 26(11), 8606–8613
in the 1 and 0.5 ML spectra are equal to 3:1 and 2:1, respectively. For the 0.5 ML spectrum, we assign peak B to the imino nitrogen atom, while the two amino nitrogen atoms (of the amino group and the IM ring) contribute to the A peak. The peak separation is 2.05 eV, which is significantly larger than the separation observed for the two nitrogen atoms in imidazole in the solid and gas phases, 1.3 eV,27 or in our multilayers (1.55 eV). The amino peak shifts by 0.1 eV between the multilayer and monolayer spectra, probably due to more efficient screening by the metal, while the imino peak shifts by 0.6 eV. We conclude that the imino peak is chemically shifted due to formation of a chemical bond with the surface. Previously, Xue et al. reported a chemical shift of 0.7 eV to lower energy of the imino feature in N 1s spectra due to the chemisorption of imidazole on silver.32 Our assignments are also supported by studies of other amino acids adsorbed on Au(111), Ag(111), Cu(110), and Pt(111) surfaces4,6,11,33,34 and imidazole in the solid state.27 Since the intensity ratio A:B is 3:1 in the 1 ML instead of 2:1 in the 0.5 ML spectrum, the bonding must also be different. We attribute the difference to species that are not chemisorbed via the imino N but only weakly bound to the surface, giving rise to intensity at lower binding energy than peak A, as in solid-state imidazole (binding energy difference of 1.3 eV). These species may be stabilized by intermolecular hydrogen bonds, which would further shift the imino N 1s toward the amino N 1s binding energy.25 At higher coverage, there is stronger intermolecular interaction, and it appears that this is comparable in energy to the surface-molecule interaction, so that for some (32) Xue, G.; Dai, Q.; Jiang, S. J. Am. Chem. Soc. 1988, 110, 2393. (33) L€ofgren, P.; Krozer, A.; Lausmaa, J.; Kasemo, B. Surf. Sci. 1997, 370, 277. (34) Schiffrin, A.; Riemann, A.; Ausw€arter, W.; Pennec, Y.; Weber-Bargioni, A.; Cvetko, D.; Cossaro, A.; Morgante, A.; Barth, J. V. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 5279.
DOI: 10.1021/la904684e
8609
Article
Feyer et al. Table 1 Au 4f intensity intensity attenuamolecule C 1s N 1s tion His Gly-His Gly-His-Gly Gly-Gly-His
Figure 4. C 1s (a), N 1s (b), and O 1s (c) core level spectra of saturated coverage (1.0 ML) of His, Gly-His, Gly-His-Gly, and Gly-Gly-His adsorbed from the liquid phase on Au(111).
molecules there is no longer a strong surface-His interaction via the imino nitrogen. On the basis of the intensity ratios, the fraction of histidine molecules coordinated to the surface via the imino nitrogen atom is estimated to be about 75% and 100% in the 1 and 0.5 monolayer samples, respectively. No changes in the intensity ratios or binding energies were observed in the spectrum of the 0.5 ML histidine film after annealing to 400 K (spectrum not shown). On the basis of the comparison of the 1 and 0.5 ML spectra, and the lack of an effect on the N 1s core level spectra due to annealing, we suggest that histidine bonds to the surface via the imino nitrogen atom, so that the imidazole ring is oriented to allow interaction with the surface. This is in agreement with Zubavichus et al.10 but in contradiction with previous RAIRS studies.3,9 In addition, Zubavichus et al. proposed that the amino group also contributes to the histidine-substrate bonding.10 We would expect a chemical shift of the N 1s peak associated with the amino nitrogen, but this was not observed in our data. Therefore, we conclude that probably the nitrogen atom of the amino group is not strongly bonded to the Au(111) surface. The N 1s spectra of His and its peptides adsorbed from the liquid phase on Au(111) are shown in Figure 4b. The N 1s spectrum of His is similar to the spectrum of His adsorbed in UHV and shows two prominent features A and B (A:B = 2.1:1) centered at 400.55 and 398.9 eV, respectively. There is a slight shift of the spectrum and filling in of the minimum between the features, which may be due to some disorder in the adlayer. We assign the features as in the case of the N 1s spectrum of His 8610 DOI: 10.1021/la904684e
1.00 1.96 2.03 2.62
1.00 2.13 2.11 2.71
0.83 0.48 0.38 0.31
effective thickness (A˚)
cube root of molecular volume (A˚)
0.9 3.6 4.0 5.7
5.7 6.3 6.4 6.7
deposited in UHV. The peak A is slightly broader compared with feature B and is due to the two amino nitrogen atoms, which may have an unresolved chemical shift. The spectrum of Gly-His (Figure 4b) is a single unresolved and asymmetric peak, and to interpret it, we have fitted it with three peaks A, B, and C, with the same energy (to within 0.05 eV) and fixed widths for A and B, whose values are those found in the His spectrum. An additional peak C is found at ∼399.8 eV, and we assign it to the nitrogen atom in the peptide CONH moiety in Gly-His. The fit yields an intensity ratio for the features A, B, and C of 1.0, 0.5, and 0.5 ( 0.03. This is in good agreement with the expected stoichiometric ratio, supporting our interpretation of the spectrum. A similar procedure was used to analyze the N 1s spectra of the tripeptides. The intensity ratios for the peaks A, C, and B are 1.0, 0.97, and 0.49. Again, this agrees with the expected stoichiometric ratio, for two amino, two peptide, and one imino nitrogen core level, and is consistent with our other data. The O 1s core level spectra of His adsorbed on the Au(111) surface in UHV are shown in Figure 3c. The spectral shapes of the 1.0 and 0.5 ML coverage histidine samples are similar, and only a small shift of ∼0.3 eV toward lower BE is observed in the submonolayer regime. As for the C and N 1s core level data, this is probably a final state shift reflecting more efficient screening by the metallic substrate. The O 1s spectra show a single peak with fwhm of about 1.6 eV. This implies that adsorbed His molecules have a deprotonated carboxylic (COOH) group, and this is reflected in the single O 1s peak of carboxylate (COO-) rather than two peaks for oxo and hydroxy forms of oxygen separated by 1.7-1.8 eV.29,30 This is also in agreement with the C 1s spectra discussed above. The asymmetric shape we relate to unresolved vibrational structure rather than to presence of the species with a protonated COOH group. The single feature in the O 1s spectra of the 1 and 0.5 ML samples indicates that the two oxygen atoms are in very similar chemical states, as no chemical splitting is observed. This is in agreement with RAIRS data of Marti et al.9 but inconsistent with the results of Liedberg et al.3 where coordination of the carboxylate group of His on gold by only one oxygen atom has been proposed. The data of Zubavichus et al.10 showed one or two peaks in the O 1s spectra, depending on sample preparation; the one peak structure was assigned by the authors to equivalent oxygen atoms of uncoordinated or symmetrically coordinated oxygen atoms, while the two peaks are attributed to bonding of His to gold via one oxygen atom. The O 1s core level spectra of His and its peptides adsorbed on Au(111) from aqueous solution are shown in Figure 4c and resemble the spectrum of His adsorbed in UHV, with slightly larger fwhm of 1.75 eV (see Figure 4c). This suggests that adsorbed peptides also have deprotonated carboxylic groups (see discussion above), and the oxygen atoms of the peptide moieties have similar binding energies, as no chemical splitting is observed. The photoemission intensities for carbon and nitrogen 1s have been quantified in order to estimate the relative surface areas occupied by the molecules (Table 1). Amino acids and peptides are flexible and can lie flat on the surface or bend to protrude from the surface, thus accommodating a larger number of molecules Langmuir 2010, 26(11), 8606–8613
Feyer et al.
Article
Figure 5. N K-edge (a) and O K-edge (b) NEXAFS spectra of 1.0 ML of histidine adsorbed in UHV on Au(111) measured at GI (10) and NI.
per unit area; this conformational adaptability is at the basis of protein folding. In addition, these compounds are expected to interact via van der Waals forces or hydrogen bonding, and the distance between, and packing of, the molecules can change, thus changing the average area per unit adsorbate. We do know that all adsorbates cover the surface sufficiently well to prevent adsorption of contaminants. The spectra were all taken in the same geometry and with the same electron energy analyzer parameters. The data were normalized to the photon flux, and the peak areas were calculated. The C and N 1s intensities have been further normalized to a value of 1 for the His intensities in Table 1. In addition, the attenuation of the substrate Au 4f intensity is included in the table; this is the intensity of the adsorbate-covered surface spectrum divided by that of the clean surface. The C and N 1s intensities for Gly-His are twice that of His, indicating a much higher atomic density. If Gly-His occupied the same area as His, we would expect an increase in intensity of 1.33 due to the larger numbers of C and N atoms. The greater increase in intensity indicates that the packing density has increased, and this could happen in two ways. First, the distance between His may be larger than between Gly-His due to differences in hydrogen bonding. Second, the conformation of the molecule may change. Our NEXAFS data indicate that the IM ring and carboxylate group are close to parallel to the surface (see below), and these two groups may be further apart, as in Gly-His-Gly (Figure 1c), or closer together, as in Gly-Gly-His (Figure 1d). Furthermore, the peptide and amino groups in Gly-His can twist up and away from the surface, which is also consistent with our N K-edge NEXAFS. The C and N 1s intensities of Gly-His-Gly are the same within error as those of Gly-His, although it contains 25% more carbon and nitrogen. This indicates that the area occupied by each molecule is 25% larger to maintain the same number of C and N atoms per unit area, which is reasonable if the molecule is at least partly spread out on the surface. Gly-Gly-His is an isomer of Gly-His-Gly but has significantly higher intensity, and so Gly-Gly-His occupies a much smaller surface area than its isomer. Again, this could be due to hydrogen bonding causing tighter packing or to conformational changes. In order to obtain a semiquantitative estimate of the coverage, we use the parametrized inelastic mean free path of Seah and Dench35 for organic materials: λm ¼ 49=Ek 2 þ 0:11Ek 1=2 mg=m - 2 (35) Seah, M. P.; Dench, W. A. Surf. Interface Anal. 1979, 1, 2.
Langmuir 2010, 26(11), 8606–8613
These values were converted to distances by dividing by the densities of each compound,36 which varies from 1.42 (His) to 1.68 (Gly-Gly-His), to yield inelastic mean free paths for His of 11.3, 8.1, and 4.1 A˚ for the C 1s, N 1s, and Au 4f photoelectrons, respectively. Since the densities are rather similar, there are only small variations in mean free paths. Using the values for Au 4f, we obtain the effective thicknesses shown in the Table 1. The effective thickness is an index of how thick a layer is within a continuum model of the overlayer, so the number is essentially qualitative as we do not have enough structural information at the atomic level to quantify more precisely. The carbon and oxygen intensities vary by a factor of about 2.7 while the thickness varies from 0.9 to 5.7 A˚, giving a discrepancy of about a factor 2 in the two numbers. However, the data show the same trend: the di- and tripeptides cover the surface more densely, and the peptide groups are located above the surface, as the nitrogen intensity apparently increases faster than the carbon intensity, even though the nitrogen photoelectrons have a shorter inelastic mean free path. It is interesting that in the case of the tripeptides the sequence of the peptide influences the area of the surface occupied and therefore the adsorption geometry. The C and N 1s intensity as well as the attenuation of Au 4f consistently supports this conclusion. The cube roots of the volumes of a single molecule of each material are also given in Table 1 to give an index of the molecular dimensions. For Gly-Gly-His, the effective thickness approaches the size of a molecule in the form of a cube, and so it is appropriate to consider whether adsorption is occurring as a bilayer20 or in a single layer. In the event of bilayer adsorption, we would expect that the second layer is either zwitterionic or neutral. If it were neutral, we would expect two O 1s peaks due to the ionization of the COOH group, and these peaks should be separated by about 1.7 eV. However, we see a single peak. If the second layer were zwitterionic, as in the suggested structure of the peptide trialanine,20 then we would expect a N 1s peak due to ionization of the protonated amino group, at ∼402.5 eV,31 but no such feature was observed. We therefore conclude that it is unlikely that bilayer formation is occurring. 3.2. NEXAFS Spectra of Histidine Layers. We have analyzed the geometry of the adsorbed species using NEXAFS spectroscopy. The N and O K-edge NEXAFS spectra of 1.0 ML of His adsorbed in UHV, and His and His-derived peptides deposited from aqueous solution on the Au(111) surface, were measured as a function of the incidence angle of the photons, and the spectra are shown in Figures 5 and 6. The N K-edge spectra show four prominent features marked A, B, C, and D. The spectra of the 0.5 ML sample deposited in UHV (not shown) had a similar (36) http://www.chemspider.com.
DOI: 10.1021/la904684e
8611
Article
Feyer et al.
Figure 6. N K-edge (a) and O K-edge (b) NEXAFS spectra of 1.0 ML of His, Gly-His, Gly-His-Gly, and Gly-Gly-His on Au(111), deposited from aqueous solution. Spectra measured at GI (10) and NI.
shape with a stronger A peak. If we normalize the spectrum for 0.5 ML to the intensity of the B feature in the 1 ML spectrum, and fit with two Lorentzian functions, the intensity of feature A is about 20-30% stronger compared with the ML spectrum. This is consistent with the observed intensity ratio difference in the N 1s XPS spectra. The published N K-edge NEXAFS spectra of polycrystalline powder films showed two prominent peaks in the π* region centered at ∼400.0 and ∼401.5 eV, which correspond to the N(imino) 1s f π* and N(amino) 1s f π* transitions of zwitterionic histidine.37,38 Imidazole in the solid state also showed a double structure due to the same transitions at 400.4 and 401.8 eV, with similar relative intensities for the two resonances.27 Following the assignments given there, the two features A and B at 399.9 and 401.35 eV in the N K-edge NEXAFS spectra of His and its peptides are attributed to transitions of the two IM nitrogen atoms, N(imino) 1s f π* and N(amino) 1s f π*. The two broad features C and D centered at 406.5 and 412.5 eV are attributed to transitions of 1s electrons of all nitrogen atoms to σ* (N-C) resonances. These assignments are supported by previously reported data on adsorbed histidine and other amino acids and imidazole in the solid phase.4,11,13,27 The N K-edge NEXAFS spectra show strong angular dependence: the π* (A and B) resonances are stronger than the σ* resonances (C and D) at GI, while at NI the π*/σ* intensity ratios reverse (Figures 5a and 6a), indicating that the IM ring of His and the His-peptides is preferentially oriented. The angular dependence of the π* resonance intensity varies as I cos2 θ,39 where θ is the angle between the E-vector and direction of the orbital. Using this equation, we conclude that the tilt angle of the IM ring is ∼32-35 with respect to the Au(111) surface. The peaks due to N(imino) 1s f π* and N(amino) 1s f π* transitions in the spectra of polycrystalline histidine powder films as well as imidazole in the gas and solid states showed similar relative intensities.27,37,38 In contrast, in the present N K-edge NEXAFS spectra of histidine and its peptides, these peaks show a different intensity ratio (see (37) Zubavichus, Y.; Shaporenko, A.; Grunze, M.; Zharnikov, M. J. Phys. Chem. A 2005, 109, 6998. (38) Leinweber, P.; Kruse, J.; Walley, F. L.; Gillespie, A.; Eckhardt, K.-U.; Blyth, R. I. R.; Regier, T. J. Synchrotron Radiat. 2007, 14, 500. (39) St€ohr, J. NEXAFS Spectroscopy; Springer: Berlin, 1992.
8612 DOI: 10.1021/la904684e
Figures 5a and 6a). We attribute this to the chemical interaction of the imino nitrogen atom with the gold surface, which changes the pole strength of the corresponding 1s f π* transition. This is consistent with our model in which the IM ring of these molecules strongly interacts with the gold surface via the imino nitrogen. The intensity ratio of peaks A:B in the spectrum of His have a constant ratio of 1:2.5 at GI and NI, as expected because they are both due to 1s f π* transitions of the two nitrogen atoms of the IM ring and so have the same angular dependence. However, these features in the spectra of His-containing peptides show a different intensity ratio at GI (see Figure 6a); the ratio of A:B in the His spectrum is 1:2.5, while in Gly-His it is equal to 1:2.75. In the case of the tripeptides, the intensity of feature B further increases (A:B = 1:3). In contrast at NI the intensity ratios A:B for all molecules adsorbed from aqueous solution are similar. This behavior indicates a contribution of the peptide 1s f π*CN transitions to the peak B. The assignment is in agreement with previous electron energy loss studies and NEXAFS measurement of Gly-Gly in the solid state.40-42 The dipole moment of the π*CN orbital is oriented perpendicular to the peptide plane, and the stronger intensity of feature B in the NEXAFS spectra of the Hiscontaining peptides at GI indicates that the peptide group is oriented essentially perpendicular to the Au(111) surface. The features C and D in the N K-edge NEXAFS spectra of His of all phases are attributed to N 1s f C-N σ* resonances for both nitrogen atom in the IM ring and amino nitrogen atom of amino group (see Figures 5a and 6a).4,11,27 The C and D peaks are stronger in the spectra measured at NI and weaker at GI; therefore, we suggest that the N-C bonds are oriented close to parallel to the Au(111) surface. This is in agreement with the RAIRS results of Liedberg et al.3 The O K-edge spectra of His adsorbed by evaporation in UHV and from the liquid phase are shown in Figures 5b and 6b. Three features are observed: a sharp peak at 532.5 eV (E) and broader (40) Gordon, M. L.; Cooper, Gly.; Morin, C.; Araki, T.; Turci, C. C.; Kaznatcheev, K.; Hitchcock, A. P. J. Phys. Chem. A 2003, 107, 6144. (41) Cooper, Gly.; Gordon, M.; Tulumello, D.; Turci, C.; Kaznatcheev, K.; Hitchcock, A. P. J. Electron Spectrosc. Relat. Phenom. 2004, 137-140, 795. (42) Zubavichus, Y.; Zharnikov, M.; Schaporenko, A.; Grunze, M. J. Electron Spectrosc. Relat. Phenom. 2004, 134, 25.
Langmuir 2010, 26(11), 8606–8613
Feyer et al.
Figure 7. Adsorption model of His and His-derived peptides on Au(111).
ones at 538 eV (F) and 544 eV (G). On the basis of the NEXAFS and electron energy loss studies of amino acids in the solid state, the feature E is attributed to the O 1s f π*COO- while G and F are assigned to the σ* resonances.4,11,37,40 No sharp feature is observed that is characteristic of a hydroxyl group, at about 535 eV,41,43 which is a further proof that adsorbed His molecules have a deprotonated carboxylate (COO-) group rather than carboxylic (COOH). The O K-edge NEXAFS spectra of His adsorbed in UHV and from aqueous solution show strong angular dependence: the π* (E) resonances are stronger than the σ* resonances (E and G) at GI, while at NI the intensity ratios reverse. Using the above-mentioned equation, we conclude that the tilt angle of the O-C-O triangle of carboxylate moieties is oriented at ∼27 with respect to the Au(111) surface. This result is inconsistent with RAIRS data, where it was proposed that histidine was adsorbed with the COO- moieties almost normal to the surface.9 In the π* region of the O K-edge spectra of His-containing peptides, the resonances due to the carboxylate O 1s f π*COOand the peptide group O 1s f π*CONH transitions overlap and are observed as peak E.40-43 The intensity of the peak E in the spectrum of Gly-His at NI is higher compared with His, and it further increases in the spectra of the tripeptides (see Figure 6b). Since the π*CONH orbital has a dipole polarized perpendicular to the NCO peptide plane, this behavior suggests that the peptide group has opposite orientation with respect to the carboxylic acid group, thus reducing the angular dependence of the resonance. This interpretation is consistent with that of the nitrogen NEXAFS. 3.3. Adsorption Model. The results described above can be combined to construct a model for the adsorption of histidine and its peptides (Figure 7). The photoemission results indicate all species bond via the imino nitrogen and the carboxylic acid group to form histidinates. His is adsorbed with the IM ring and carboxylate group at shallow angles to the surface, as observed with N K-edge NEXAFS. The O K-edge NEXAFS and photoemission spectra of the peptides suffer from overlap of the carboxylate and peptide features, but the fitted results are consistent with the carboxylate having the same geometry as in His and with the peptide not forming a surface bond. (43) Feyer, V.; Plekan, O.; Richter, R.; Coreno, M.; Prince, K. C.; Carravetta, V. J. Phys. Chem. A 2009, 113, 10726.
Langmuir 2010, 26(11), 8606–8613
Article
Quantitative comparison of the C and N 1s intensities indicates that the atomic area density increases for the peptides. Since the increase is about a factor of 2, one might conclude that bilayer adsorption is occurring. However, such a model would require that the second layer molecule has the same charge state of the carboxylate group, to explain the O 1s photoemission, and the same orientation of the imidazole ring, to explain the N edge NEXAFS. This appears unlikely, and instead we attribute the increase in intensity to conformations in which the peptide groups are above the surface. The evidence favors a model in which the peptide group is roughly perpendicular to the surface, and given that the atomic density is higher for the peptides, we propose that they are above the IM and carboxylate groups, as shown schematically in Figure 7.
4. Conclusions Using high-resolution synchrotron radiation-based core spectroscopies, the electronic structure and adsorption geometry of His and His-containing peptides on Au(111) have been studied. We propose that histidine at 1.0 and 0.5 ML coverages deposited in UHV, and at saturated coverage deposited from aqueous solution, interacts with the gold surface in its anionic form via the two oxygen atoms of the carboxylate group and the imino nitrogen atom of the IM ring. The bond to the carboxylate group is stronger than to the imino group, so that at high coverage not all His molecules bond through the IM group. From the N and O K-edge NEXAFS spectra measured at different incidence angles with respect to the surface, we conclude that the tilt angles of the IM ring and COO- triangle, of the vacuum and liquid phase deposited His, are ∼35 and 27 with respect to the Au(111) surface, respectively. The imidazole rings of the His-derived peptides bond to the gold surface in a mode similar to the single His species, with the peptide plane perpendicular to the surface. The C 1s spectra of aqueous solution deposited His and Hiscontaining peptides show that these compounds bond to the surface sufficiently strongly to displace impurities and form an almost uniform layer. The bonding of His is the same for both modes of deposition, which implies that conclusions drawn from UHV evaporation are valid for systems formed by aqueous adsorption. Analysis of the C, N, and Au photoemission intensity data indicates that the packing density, and therefore the local geometry is significantly different for the two tripeptides and that the di- and tripeptides adsorb with a higher atomic density than His. The knowledge that the sequence of residues in a peptide can significantly influence the adsorption geometry will aid in designing and engineering metal-bio interfaces. Acknowledgment. We gratefully acknowledge the assistance of our colleagues at Elettra for providing good quality synchrotron light. The Materials Science Beamline is supported by the Ministry of Education of the Czech Republic under Grant LC06058. V. Chab acknowledges the Academy of Sciences of the Czech Republic for support under Grant No IAA100100905.
DOI: 10.1021/la904684e
8613