Electronic structure of dipeptides in the gas-phase and as an adsorbed monolayer

Cunlan Guo ab, Soumyajit Sarkar a, Sivan Refaely-Abramson a, David A. Egger a, Tatyana Bendikov c, Keiichirou Yonezawa d, Yosuke Suda d, Takuma Yamaguchi e, Israel Pecht f, Satoshi Kera deg, Nobuo Ueno d, Mordechai Sheves b, Leeor Kronik a and David Cahen *a
aDepartment of Materials and Interfaces, Weizmann Institute of Science, Rehovot, 76100, Israel. E-mail: david.cahen@weizmann.ac.il
bDepartment of Organic Chemistry, Weizmann Institute of Science, Rehovot, 76100, Israel
cDepartment of Chemical Research Support, Weizmann Institute of Science, Rehovot, 76100, Israel
dGraduate School of Advanced Integration Science, Chiba University, Chiba 263-8522, Japan
eSOKENDAI (The Graduate University for Advanced Studies), Hayama, Kanagawa 240-0193, Japan
fDepartment of Immunology, Weizmann Institute of Science, Rehovot, 76100, Israel
gInstitute for Molecular Science, Myodaiji, Okazaki, 444-8585, Japan

Received 30th November 2017 , Accepted 10th February 2018

First published on 12th February 2018


Peptide-based molecular electronic devices are promising due to the large diversity and unique electronic properties of biomolecules. These electronic properties can change considerably with peptide structure, allowing diverse design possibilities. In this work, we explore the effect of the side-chain of the peptide on its electronic properties, by using both experimental and computational tools to detect the electronic energy levels of two model peptides. The peptides include 2Ala and 2Trp as well as their 3-mercaptopropionic acid linker which is used to form monolayers on an Au surface. Specifically, we compare experimental ultraviolet photoemission spectroscopy measurements with density functional theory based computational results. By analyzing differences in frontier energy levels and molecular orbitals between peptides in gas-phase and in a monolayer on gold, we find that the electronic properties of the peptide side-chain are maintained during binding of the peptide to the gold substrate. This indicates that the energy barrier for the peptide electron transport can be tuned by the amino acid compositions, which suggests a route for structural design of peptide-based electronic devices.


Introduction

Biomolecules such as proteins and peptides are potential candidates for building diverse functional molecular electronic devices.1 Unlike commonly used alkyl-like and aromatic molecules, peptides exhibit special structures, owing to amide bonds and a variety of side chains, from completely saturated to aromatic moieties. Furthermore, peptide bonds introduce a dipole along the peptide backbone, with varying magnitude and direction, depending on the peptide structure. The structure of the amide backbone and flexibility in side-chain selection allow for tunability of the peptide electronic structure and electrode–molecule coupling, making them attractive for possible multifunctional solid-state electronics.2–13

To understand peptide junctions, we need to consider the mechanism of electron transport (ETp) via the peptides that bridge between the contacting electrodes. The electronic structure of the peptide at the electrode interface is key to understanding ETp in and across proteins, especially those involved in electron transfer (ET) in biology, in energy conversion, and in signaling.11,14–16 Specifically, the electronic structure of a peptide monolayer in contact with an electrode can be directly related to the energy barrier and electrode–molecule coupling in the resulting junctions.17

Ultraviolet photoelectron spectroscopy (UPS), contact potential difference measurement by Kelvin probe, and Kelvin probe force microscopy have been used extensively to study the electronic structure of molecular thin films or monolayers on a solid surface. Among other quantities of interest, these experiments can provide information on the ionization potential (IP) of the molecular layer and the work function (WF) of the molecularly modified electrode.18,19 Clearly, monolayer energetics are strongly influenced by molecule-electrode interactions and monolayer dipole moment, which are also related to the WF of the electrode and the energy alignment/coupling between electrode and monolayer.20–24 Here, in order to obtain insights into peptide functionality for future applications in biomolecular electronics, we compare the electronic structure of two dipeptides – 2Ala and 2Trp (see Fig. 1) – which differ greatly in terms of their side chains. Ala has a methyl group, while Trp has an aromatic indole ring. We study the two dipeptides using UPS in both the gas phase and on a monolayer bound to Au via a thiol linker. This allows us to identify intrinsic similarities and differences in the electronic structure of the dipeptides, in a systematic way that is devoid of additional complications brought about by extrinsic solvent effects. Combining the experimental data with results from first-principles density functional theory (DFT) calculations, we analyze the electronic structure of both dipeptides in the gas phase and compare the changes in structure and energy levels between the gas phase and Au-adsorbed monolayer. The results show that the S–Au bonding, achieved via an MPA (3-mercaptopropionic acid) linker group, mostly preserves the gas-phase electronic structure of the dipeptides, resulting primarily in a rigid shift of the energy levels. The peptide–substrate interaction can further affect the frontier orbitals to a certain extent, via orbital hybridization, depending on the nature of peptides. This finding provides insights into parameters that control coupling between peptide molecules and a metallic substrate that serves as an electrical contact, with implications for the design of peptide-based electronic devices.


image file: c7cp08043c-f1.tif
Fig. 1 Structures of the dipeptides used in the study: (A) 2Ala, (B) 2Trp, (C) MPA-2Ala, (D) MPA-2Trp. MPA = mercapto-propionic acid, the linker used for binding to Au via an Au–S bond.

Experimental and computational details

Chemicals and materials

Dipeptides with/without mercaptopropionic acid (MPA) linker at the N-terminus were purchased from GL Biochem Ltd (Shanghai, China) with purity >95% (HPLC). A small amount of TFA (trifluoroacetic acid) is contained to keep the dipeptides in neutral condition. For preparing MPA-dipeptide monolayers, MPA-dipeptides were dissolved in DMF with a concentration of 0.25 mM.

Peptide monolayer preparation on Au surface

Au substrates were cleaned by sonication in water and ethanol for 5 min, followed by UV/ozone treatment for 10 min. The Au substrates were further treated by hot ethanol to remove the oxidized Au. After that, the Au substrates were incubated in peptide solution for 48 hours, cleaned with DMF and ethanol and dried by nitrogen.

Polarization modulation-infrared reflection–absorption spectroscopy (PM-IRRAS)

A Nicolet 6700 FTIR spectrometer coupled with a PEM-90 photoelastic modulator (Hinds Instruments, Hillsboro, OR) was used to collect PM-IRRAS data for the peptide monolayers on Au with incident angle 80° and resolution of 2 cm−1.

Ellipsometry measurements

A Woollam M-2000 V multiple-wavelength ellipsometer at an incidence angle of 70° was used to measure the ellipsometry of the sample. The Cauchy model was used to estimate the thickness of the peptide layers. The parameters in the Cauchy model for “n” and “k” are An = 1.45; Bn = 0.01; Cn = 0.

Electrochemistry

Cyclic voltammetry was obtained with VersaSTAT 3 Potentiostat Galvanostat (Princeton Applied Research). Au substrates with MPA-dipeptide monolayers were used as a working electrode in a three-electrode cell configuration. All potentials are referred to a Hg/Hg2Cl2 (saturated KCl) electrode. The electrochemical measurements were performed in 0.5 M KOH solution after 15 min of nitrogen purging. The number of molecules involved in Au–S bond breaking was calculated from the area of the reduction peak near −1.14 V (vs. Hg/Hg2Cl2), which was calculated from the onset of the peak in order to decrease (but not necessarily completely eliminate) the charge contribution from the double layer. With the estimated Au electrode area, the mean molecular area of one MPA-dipeptide was evaluated,25 which implicitly assumes a similar interaction of the two peptides with the Au substrate.

Contact potential difference (CPD)

CPD measurements were performed in a N2-filled glove box at room temperature. The CPD values were transformed to WF by using freshly peeled highly oriented pyrolytic graphite (HOPG) as a reference, which has a known WF of 4.6 eV.26

UPS measurement in gas phase

Gas-phase UPS measurements were performed in a custom-built UPS system with a VG-CLAM4 analyzer, a He I (21.22 eV) radiation source, and a high-temperature gas photoionization cell (gas cell) with a miniature container of an organic powder sample. The powders of dipeptides were sublimated at 430 K and 545 K for 2Ala and 2Trp, respectively, for photoionization. The energy resolutions were ∼60 meV for 2Ala and ∼100 meV for 2Trp which are determined by full-width at half maximum for the 2P3/2 (15.76 eV) peak of Ar. This peak was also used for calibrating the energy scale, i.e., the energy scale is from the vacuum level, which is determined from simultaneous measurement of this Ar 2P3/2 peak (±1 meV).

UPS measurement for monolayers on Au surface

A Kratos AXIS ULTRA system with a hemispherical analyzer for photo-emitted electron detection was used to collect UPS data of peptide monolayers. UPS was measured with a helium discharge lamp, using He I (21.22 eV) radiation lines. The total energy resolution was better than 100 meV, as determined from the Fermi edge of an Au reference sample. A logarithmic intensity scale was used to determine the HOMO level at low binding energies by extrapolating the signal “edge” to the background signal level. The results were not significantly different from those obtained using a linear intensity scale.

DFT and electronic structure calculations

Gas-phase DFT calculations were performed using the Q-Chem package, version 4.1,27 with the structural configuration of the peptides optimized using the PBE0 hybrid functional. For the electronic-structure calculations we used the optimally-tuned range-separated hybrid (OT-RSH) functional28 approach. Specifically, we used an RSH functional containing 80% short-range PBE exchange, 20% short-range Fock exchange, 100% long-range Fock exchange and full PBE correlation. The range-separation parameter was optimally tuned so as to obey the IP theorem for the peptides in the neutral form. The basis-sets used were the correlation-consistent polarized valence triple zeta (cc-pVTZ)29 for both electronic structure and geometry optimization. DFT-based ab initio molecular dynamics (MD) calculations were performed using the Born–Oppenheimer potential energy surface as implemented in Q-Chem. Each MD run consisted of 2000 steps with a time-step of 0.484 femtoseconds at a temperature of 430 K (2Ala) and 545 K (2Trp), and calculated the electronic structure at every tenth time-step. DFT calculations for “floating” and Au-adsorbed monolayers were performed within the plane wave-based Vienna Ab initio Simulation Package (VASP),30,31 using projector-augmented wave (PAW) potentials to treat the core electrons. The structure was optimized using the PBE functional and the density of states (DOS) was calculated using the HSE32,33 functional. A k-point mesh of 6 × 6 × 1 was considered for the Brillouin zone integration. A dipole-correction was applied in the direction perpendicular to the surface-normal in all band-structure calculations.

Results and discussion

We use short dipeptides because these can be brought into the gas phase without decomposition and can be readily analyzed by first principles electronic structure calculations. We choose to compare alanine (Ala) and tryptophan (Trp) as these natural amino acids have the largest IP difference.34,35 The chemical structures of 2Ala and 2Trp without and with a thiol binding group are shown in Fig. 1.

We first compare experimental UPS data of the gas-phase dipeptides (Fig. 2) with simulated spectra obtained from density functional theory (DFT) based on the optimally-tuned range-separated hybrid (OT-RSH) approach28 (see Methods section). In recent years, this method was used for accurate simulations of UPS data for a large variety of organic molecules.36–39 Specifically, the HOMO levels obtained from this method can be compared directly with the IP from the UPS.


image file: c7cp08043c-f2.tif
Fig. 2 Comparison of experimental and simulated gas-phase UPS data for 2Ala (A, left) and 2Trp (B, right): (a) gas-phase UPS data, at 430 K for 2Ala and at 545 K for 2Trp; (b) electronic structure calculation at 0 K; (c) electronic structure calculation based on molecular dynamics at the experimental temperatures. The vacuum level is defined as 0 eV. All computational data have been rigidly shifted to the left by 0.1 eV (2Ala) and 0.4 eV (2Trp) for the best agreement with experiment (see text for a detailed discussion). Vertical bars in the simulated spectra represent energy levels (in (c) for different configurations), with the solid curves obtained from Gaussian-based broadening of the levels.

Yuan et al.40 have shown that single amino acids can have several conformations close to the minimum total energy, depending on the relative orientation of the standard backbone dihedral angles, ψ and φ. Being close to amino acids in terms of their short backbone structure, it is reasonable for dipeptides to possess several conformations as well. This is particularly so for 2Ala, whereas in 2Trp the much larger indole-ring side-chains limit the possibility of multiple conformations at energies close to the minimum total energy. Therefore, we theoretically optimized several 2Ala conformations with different dihedral angles. Our calculations showed that different conformations can be found within an energy range of 40 meV from the minimum energy, as shown in Fig. S1 of the ESI. Here, in absence of detailed information as to which of these is the most relevant one in the experiment, we consider the conformation that resulted in an eigenvalue spectrum in best agreement with the experimental UPS data (see Fig. S2 in the ESI for details).

The OT-RSH computed IP at 0 K for 2Ala is 9.2 eV, somewhat smaller than the measured IP of 9.6 eV at 430 K (Fig. 2A). Naturally, flexible molecules possess many different conformations at temperatures above 0 K. To rule out that this difference is primarily due to thermal broadening, we performed additional DFT-based molecular dynamics at an experimentally relevant temperature and averaged over the IP of the different configurations. The averaged IP of the resulting structures is 9.25 eV, i.e., close to the calculated value at 0 K.

The gas phase UPS of 2Trp was measured at 545 K (Fig. 2B) (compared to 2Ala, higher temperature is required for 2Trp vaporization). Similar to 2Ala, the measured IP for 2Trp (7.8 eV) is higher than the computed IP for the gas phase at 0 K (7.3 eV). An average over a molecular dynamics simulation at 545 K yielded a value of 7.2 eV, also close to the calculated IP at 0 K. We note that OT-RSH calculations are designed to yield HOMO values that are predictive for measured IP values and are found to exhibit a mean absolute error of only 0.2 eV across a benchmark set of 148 different molecules.41 Here, the underestimation of ∼0.4–0.5 eV by the OT-RSH approach is somewhat larger, but still considerably smaller than what is obtained with standard DFT functionals.28 The remaining shifts can be due to the diverse molecular geometries in the gas-phase, which are known to produce a variety of dipole values.40 Given this difference, the simulated spectra shown in Fig. 2 have been slightly shifted (by 0.1 eV and 0.4 eV for 2Ala and 2Trp, respectively) to obtain the best visual agreement between theory and experiment. This agreement clearly indicates that the overall electronic structure is well-captured by the OT-RSH calculation, which allows for its use in interpretation of the experimental data. Thus, we can examine the molecular orbitals corresponding to the 0 K calculations (Fig. 3). The frontier orbitals of 2Ala are distributed mostly over a large part of the backbone; the HOMO and HOMO−1 exhibit an energy difference of ∼0.3 eV. For 2Trp, the frontier orbitals are largely distributed on one or the other of the two indole rings with a difference of ∼0.3 eV from HOMO to HOMO−1.


image file: c7cp08043c-f3.tif
Fig. 3 OT-RSH based electronic structure calculations for several highest occupied molecular orbitals of 2Ala (top panel) and 2Trp (bottom panel), shown along with the (unshifted) 0 K broadened, simulated UPS data. The vacuum level is defined as 0 eV.

To form a uniform peptide monolayer on Au, we introduced an MPA linker at the N-terminus of the dipeptide, allowing assembly of the peptide on Au through Au–S chemical bonding.42 To examine the effect of this linker group in the process of peptide monolayer formation, we first study the electronic levels of the MPA-dipeptide. As it is experimentally not feasible to sublime the MPA-dipeptide for gas-phase UPS measurement without dissociating the molecule, we turn to theory to find how the linker affects the molecular energy levels (Fig. 4). The calculations reveal that, as in the parent molecules, the frontier orbitals of the MPA-2Ala are distributed over a large part of the molecule (Fig. 4A) and that the frontier orbitals of MPA-Trp are distributed mainly on the indole rings (Fig. 4B), with the electron distributions in the dipeptide not dramatically changed after introducing the MPA linker. Introduction of the MPA linker to the 2Trp swaps the HOMO and HOMO−1 ordering and reduces their difference in energy from 0.4 eV to 0.1 eV. However, the frontier orbitals remain localized on the indole groups and do not reside at the linker site, in contrast to the case for MPA-2Ala, where the HOMO resides at the linker site. While the IP values for MPA-2Ala and 2Ala are quite close, the IP value of MPA-2Trp increases by 0.4 eV compared to that of 2Trp (Table 1).


image file: c7cp08043c-f4.tif
Fig. 4 OT-RSH based electronic structure calculations at 0 K for several highest occupied molecular orbitals of MPA-2Ala (top panel) and MPA-2Trp (bottom panel). The vacuum level is defined as 0 eV. The thiol group of the MPA linker is at the N-terminus of dipeptide.
Table 1 Summary of ionization potentials deduced from experiment and computationa
2Ala (gas phase) 2Ala-MPA (gas phase) 2Ala-MPA (on Au) 2Trp (gas phase) 2Trp-MPA (gas-phase) 2Trp-MPA (on Au)
a The vacuum level is defined as 0 eV. b Value estimated from difference in calculated values for this dipeptide; for details see text.
Experiment (eV) 9.6 ± 0.06 (9.5)b 6.85 7.8 ± 0.1 (8.2)b 5.55
Calculation (eV) 9.2 9.1 7.3 7.7


In order to check the interaction of the MPA-dipeptide monolayer with the substrate, we considered two different scenarios: an Au-adsorbed MPA-dipeptide monolayer and a hypothetical “floating monolayer”, which is similar to the one adsorbed on Au but lacks the Au–substrate and, therefore, the Au–S bonding.43 In both cases, we assumed a laterally ordered monolayer. Because the OT-RSH approach is not fully self-consistent for a molecule–metal interface,44 we performed this comparison, for both cases, by using the Heyd–Scuseria–Ernzerhof (HSE) functional.32,33Fig. 5 presents the projected density of states (pDOS) of the MPA-dipeptides for the “floating monolayer” and for the Au-adsorbed monolayer. After hybridization with the Au surface, the peak related to the HOMO of the MPA-2Ala changes significantly, whereas that related to the HOMO of the MPA-2Trp does not. This is consistent with the above-discussed nature of the HOMO in the two gas-phase MPA-dipeptides: hybridization with the substrate induces significant broadening only if the gas-phase MPA-dipeptide HOMO exhibits significant localization on the linker.


image file: c7cp08043c-f5.tif
Fig. 5 Density of states (DOS) for (A) MPA-2Ala and (B) MPA-2Trp peptide monolayers, projected onto the atomic orbitals of the molecular components. Blue lines indicate the projected DOS of each peptide in a “floating monolayer” and red lines show the projected DOS of a peptide in a monolayer adsorbed on an Au(111) surface. The arrows point to the HOMO orbital, which strongly hybridizes with the Au surface in the case of MPA-2Ala but barely hybridizes with the Au surface in the case of MPA-2Trp. To facilitate the comparison, all spectra have been shifted such that the HOMO level is aligned to 0 eV.

Turning to the Au-adsorbed molecules, polarization modulation-infrared reflection–absorption spectra (PM-IRRAS, Fig. 6) show the amide peaks of both MPA-dipeptides and an additional peak of MPA-2Trp corresponding to the C[double bond, length as m-dash]C vibration in the indole ring of Trp. These peaks prove the adsorption of MPA-2Ala and MPA-2Trp on the Au substrate.45,46 The differences between the measured thickness from ellipsometry (Fig. S3, ESI) and the estimated molecular length from standard bond lengths suggest that the MPA-2Ala monolayer is well-ordered and slightly tilted from the perpendicular orientation towards Au substrate, while the MPA-2Trp monolayer is less ordered, likely due to its bulky indole ring (Table 2). The mean molecular area of an MPA-dipeptide on an Au substrate was additionally estimated by an electrochemical method25 (Table 2) and found to be larger than the mean area of n-alkanethiols on Au, but of the same order of magnitude25 and similar to the mean area of both helical polyalanine on Au by electrochemistry9 and helical hetero-peptide on Au by MD.47 This result indicates that both MPA-2Ala and MPA-2Trp are highly packed in their monolayers, even though the mean molecular area of MPA-2Trp is larger than that of MPA-2Ala, owing to the large indole ring on the Trp.


image file: c7cp08043c-f6.tif
Fig. 6 PM-IRRAS of (A) MPA-2Ala and (B) MPA-2Trp monolayers on an Au surface.
Table 2 Thickness and coveragea of an MPA-dipeptide monolayer on an Au surface
MPA-2Ala MPA-2Trp
a As measured by electrochemistry.
Thickness from ellipsometry (Å) 8.7 ± 0.12 6.7 ± 0.03
Molecular length calculated from standard bond lengths (Å) 12.6 8.9
Mean molecular areaa2 per molecule) 51.1 ± 12.6 64.1 ± 19.0


The electronic structure of the MPA-dipeptide monolayer on Au was studied by UPS. The effective WF values of Au with MPA-2Ala or MPA-2Trp monolayers (Fig. 7A) are found to be similar, in agreement with non-ultrahigh vacuum measurements by CPD (Table 3). These similar WF values indicate similar interactions with the Au substrate of both MPA-2Ala and MPA-2Trp. This result suggests that the interaction of the MPA linker with Au is the dominant part that affects the WF. The absolute IP values for the monolayers can be obtained by adding the measured position of the HOMO level to the measured work function value, given relative to the Au Fermi level (Fig. 7 and Table 1). For both monolayers, the IP of the experimentally measured MPA-dipeptide monolayer is substantially smaller than that of the measured gas-phase dipeptide, with the differences being 2.75 eV for 2Ala and 2.25 eV for 2Trp (Table 1).


image file: c7cp08043c-f7.tif
Fig. 7 UPS spectra of MPA-2Ala (red) and MPA-2Trp (blue) monolayers on Au: (A) photoemission cutoff region; (B) photoemission onset region. The binding energy is given with respect to the position of the Fermi level (EF) of Au, defined as 0 eV. The WF is calculated by the difference between the ultraviolet photon energy and the measured cutoff value. The values of the onsets are recorded from the semi-log spectra.
Table 3 Work function (from UPS and CPD), HOMO position with respect to Fermi level (from UPS), and ionization potential (IP, sum of HOMO position and work function) for an MPA-dipeptide monolayer on Au. All values are referenced to the Fermi level of the Au substrate
W F from UPS (eV) W F from CPD (eV) Onset from UPS (eV) IP (eV)
MPA-2Ala 4.55 4.65 2.30 6.85
MPA-2Trp 4.60 4.65 0.95 5.55


A priori, two main factors can lead to the observed IP value shifts – adding the MPA linker to the dipeptide and/or adsorbing the MPA-dipeptide on a Au surface to form a monolayer. As discussed above, calculations show that adding the MPA linker yields a 0.4 eV difference between MPA-2Trp and 2Trp and 0.1 eV between MPA-2Ala and 2Ala. This change is far too small to make the MPA linker a major player in the IP change, pointing to the adsorption of the peptide as a monolayer on the Au as the dominant cause for the IP shifts.

To quantify the IP decrease from gas-phase MPA-dipeptide to MPA-dipeptides monolayer on Au, we estimate the experimental IP values of the MPA-dipeptides in the gas phase, based on the calculated IP values and the 0.4–0.5 eV difference between experiment and calculation of dipeptide IPs (Table 1). A summarized illustration for the energy level changes from gas-phase dipeptide to MPA-dipeptide monolayer on Au, including the measured and estimated experimental data, is shown in the SM (Fig. S4, ESI). This procedure yields ∼9.5 eV for the IP of MPA-2Ala and ∼8.2 eV for the IP of MPA-2Trp in the gas-phase. Therefore, the estimated IP differences of MPA-dipeptide between the monolayer on Au and the gas-phase are 2.65 eV for both MPA-2Ala and MPA-2Trp.

The significant increase in the HOMO energy (closer to the electrode Fermi level as well as to the vacuum level) upon adsorption onto the metal is consistent with the physical phenomenon of surface-induced gap renormalization.44,48–51 The reason for this renormalization is as follows. When the molecule is close to the surface, electrons in the metal respond to and screen single-particle excitations in the molecule, i.e., carriers in the metal polarize when a hole/electron is added to the molecule, thereby screening the Coulomb interactions in the molecule. As a result, when the molecule approaches the surface, the HOMO and LUMO energies move closer to the Fermi level. This renormalization is therefore often referred to as an “image-charge effect” (there is an additional, similar effect due to the embedding of the molecule in the monolayer). While polarization-induced renormalization is an important factor in the observed changes, it likely cannot account for the full increase of the HOMO energy. Indeed, the HOMO energy level alignment, the IP of a chemisorbed monolayer, and hence the transport properties, are expected to be influenced also by other factors, including orbital hybridization at the molecule–metal interface, intrinsic molecular dipoles, and inter-molecular hybridization and/or depolarization, as discussed, e.g. in ref. 21 and 24. Interestingly, all these factors together lead to the same IP shift during the formation of the two MPA-dipeptide monolayers from gas-phase MPA-dipeptides, with the electronic properties of the peptide side-chains preserved. It suggests that the energy barrier in a peptide junction can be determined by the side-chain, even in the absence of strong secondary structure effects. Thus, the nature of the peptide side-chain, not only the peptide-electrode coupling, can significantly affect the electron transport in peptide-based solid-state devices. This is consistent with our previous reports on peptide electron transport.10,17

Conclusions

In summary, by combining UPS experiments, in the gas phase and on monolayers on Au, with DFT simulations we analyzed the energy level change of two dipeptides as a result of forming a monolayer on Au via an MPA linker added to the peptide. The (gas-phase) calculations show that adding an MPA linker to the dipeptide can slightly shift the HOMO energy. The frontier orbitals of MPA-2Ala are still distributed over a significant part of the molecular backbone, while those of MPA-2Trp remain localized on one of the indole rings. In the process of surface adsorption and monolayer formation, the energy levels shift due to several effects, notably metal-induced renormalization, and possibly dipole–dipole interactions and orbital hybridization. The frontier orbitals of the two dipeptides exhibit partially different interactions with the metal, as the electronic states of 2Ala hybridize with the Au substrate, while the more localized 2Trp orbitals are barely altered. Nevertheless, our analysis indicates that both dipeptides undergo similar IP shifts from gas phase to a monolayer adsorbed on Au. Taken together, the findings strongly suggest that at least part of the gas-phase peptide side-chain electronic properties is maintained, which offers the possibility of fine-tuning electron transmission through peptide monolayers. This behavior may be used for designing specific functionalities in future biomolecular electronic devices. In addition, our results demonstrate that the amino acid side-chains may play an important role in modulating electron transport across proteins.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

DC and LK thank the Israel Science Foundation through its Centers of Excellence program, DC thanks the Nancy and Stephen Grand Center for Sensors and security. DC and MS thank the Israel Science Foundation and the Minerva Foundation (Munich). MS thanks the Kimmelman Center for Biomolecular Structure and Assembly, and the Jonathan Beare and Estate of George Hecht for partial support. LK acknowledges support from the European Research Council. SK acknowledges support from JSPS KAKENHI (no. 26248062, 23360005). SRA acknowledges an Adams Fellowship of the Israel Academy of Sciences and Humanities. SS acknowledges the Koshland Foundation and support from a McDonald-Leapman grant. DAE acknowledges support by the Koshland Foundation and the Austrian Science Fund (FWF): J3608-N20. DC thanks Chiba University for a visiting professorship. This research was made possible in part by the historic generosity of the Harold Perlman family. This joint work was partly supported by the Global-COE Program of MEXT (G03), Advanced School for Organic Electronics, operated at Chiba University. MS holds the Katzir-Makineni Professorial Chair in Chemistry; DC held the Rowland and Sylvia Schaefer Professorial Chair in Energy Research.

References

  1. P. Facci, Biomolecular electronics: bioelectronics and the electrical control of biological systems and reactions, William Andrew Inc., Waltham, MA, 2014, pp. 1–240 Search PubMed .
  2. L. Adler-Abramovich and E. Gazit, Chem. Soc. Rev., 2014, 43, 6881–6893 RSC .
  3. N. Amdursky, ChemPlusChem, 2015, 80, 1075–1095 CrossRef CAS .
  4. M. Eckshtain-Levi, E. Capua, S. Refaely-Abramson, S. Sarkar, Y. Gavrilov, S. P. Mathew, Y. Paltiel, Y. Levy, L. Kronik and R. Naaman, Nat. Commun., 2016, 7, 10744 CrossRef CAS PubMed .
  5. J. R. Horsley, J. X. Yu, K. E. Moore, J. G. Shapter and A. D. Abell, J. Am. Chem. Soc., 2014, 136, 12479–12488 CrossRef CAS PubMed .
  6. J. Juhaniewicz, J. Pawlowski and S. Sek, Isr. J. Chem., 2015, 55, 645–660 CrossRef CAS .
  7. M. Matmor, G. A. Lengyel, W. S. Horne and N. Ashkenasy, Phys. Chem. Chem. Phys., 2017, 19, 5709–5714 RSC .
  8. L. Scullion, T. Doneux, L. Bouffier, D. G. Fernig, S. J. Higgins, D. Bethell and R. J. Nichols, J. Phys. Chem. C, 2011, 115, 8361–8368 CAS .
  9. S. Sek, A. Tolak, A. Misicka, B. Palys and R. Bilewicz, J. Phys. Chem. B, 2005, 109, 18433–18438 CrossRef CAS PubMed .
  10. L. Sepunaru, S. Refaely-Abramson, R. Lovrincic, Y. Gavrilov, P. Agrawal, Y. Levy, L. Kronik, I. Pecht, M. Sheves and D. Cahen, J. Am. Chem. Soc., 2015, 137, 9617–9626 CrossRef CAS PubMed .
  11. A. Shah, B. Adhikari, S. Martic, A. Munir, S. Shahzad, K. Ahmad and H.-B. Kraatz, Chem. Soc. Rev., 2015, 44, 1015–1027 RSC .
  12. X. Y. Xiao, B. Q. Xu and N. J. Tao, J. Am. Chem. Soc., 2004, 126, 5370–5371 CrossRef CAS PubMed .
  13. X. Y. Xiao, B. Q. Xu and N. J. Tao, Angew. Chem., Int. Ed., 2004, 43, 6148–6152 CrossRef CAS PubMed .
  14. E. Gatto and M. Venanzi, in Peptide materials: from nanostructures to applications, ed. C. Alemán, A. Bianco and M. Venanzi, John Wiley & Sons, 2013, ch. 4, pp. 105–147 Search PubMed .
  15. H. B. Gray and J. R. Winkler, Q. Rev. Biophys., 2003, 36, 341–372 CrossRef CAS PubMed .
  16. Y. T. Long, E. Abu-Rhayem and H. B. Kraatz, Chem. – Eur. J., 2005, 11, 5186–5194 CrossRef CAS PubMed .
  17. C. L. Guo, X. Yu, S. Refaely-Abramson, L. Sepunaru, T. Bendikov, I. Pecht, L. Kronik, A. Vilan, M. Sheves and D. Cahen, Proc. Natl. Acad. Sci. U. S. A., 2016, 113, 10785–10790 CrossRef CAS PubMed .
  18. D. Cahen and A. Kahn, Adv. Mater., 2003, 15, 271–277 CrossRef CAS .
  19. N. Ueno and S. Kera, Prog. Surf. Sci., 2008, 83, 490–557 CrossRef CAS .
  20. S. Duhm, G. Heimel, I. Salzmann, H. Glowatzki, R. L. Johnson, A. Vollmer, J. P. Rabe and N. Koch, Nat. Mater., 2008, 7, 326–332 CrossRef CAS PubMed .
  21. D. A. Egger, F. Rissner, E. Zojer and G. Heimel, Adv. Mater., 2012, 24, 4403–4407 CrossRef CAS PubMed .
  22. G. Heimel, L. Romaner, E. Zojer and J.-L. Brédas, Acc. Chem. Res., 2008, 41, 721–729 CrossRef CAS PubMed .
  23. Y. L. Huang, W. Chen, F. Bussolotti, T. C. Niu, A. T. S. Wee, N. Ueno and S. Kera, Phys. Rev. B: Condens. Matter Mater. Phys., 2013, 87, 085205 CrossRef .
  24. A. Natan, L. Kronik, H. Haick and R. T. Tung, Adv. Mater., 2007, 19, 4103–4117 CrossRef CAS .
  25. C. A. Widrig, C. Chung and M. D. Porter, J. Electroanal. Chem., 1991, 310, 335–359 CrossRef CAS .
  26. C. Sommerhalter, T. W. Matthes, T. Glatzel, A. Jager-Waldau and M. C. Lux-Steiner, Appl. Phys. Lett., 1999, 75, 286–288 CrossRef CAS .
  27. Y. Shao, L. F. Molnar and Y. Jung, et al. , Phys. Chem. Chem. Phys., 2006, 8, 3172–3191 RSC .
  28. L. Kronik, T. Stein, S. Refaely-Abramson and R. Baer, J. Chem. Theory Comput., 2012, 8, 1515–1531 CrossRef CAS PubMed .
  29. T. H. Dunning, J. Chem. Phys., 1989, 90, 1007–1023 CrossRef CAS .
  30. G. Kresse and J. Hafner, Phys. Rev. B: Condens. Matter Mater. Phys., 1993, 47, 558–561 CrossRef CAS .
  31. G. Kresse and D. Joubert, Phys. Rev. B: Condens. Matter Mater. Phys., 1999, 59, 1758–1775 CrossRef CAS .
  32. J. Heyd, G. E. Scuseria and M. Ernzerhof, J. Chem. Phys., 2003, 118, 8207–8215 CrossRef CAS .
  33. J. Heyd, G. E. Scuseria and M. Ernzerhof, J. Chem. Phys., 2006, 124, 219906 CrossRef .
  34. P. H. Cannington and N. S. Ham, J. Electron Spectrosc. Relat. Phenom., 1979, 15, 79–82 CrossRef CAS .
  35. E. W. Schlag, S.-Y. Sheu, D.-Y. Yang, H. L. Selzle and S. H. Lin, Angew. Chem., Int. Ed., 2007, 46, 3196–3210 CrossRef CAS PubMed .
  36. I. E. Brumboiu, G. Prokopiou, L. Kronik and B. Brena, J. Chem. Phys., 2017, 147, 044301 CrossRef PubMed .
  37. D. A. Egger, S. Weissman, S. Refaey-Abramson, S. Sharifzadeh, M. Dauth, R. Baer, S. Kuemmel, J. B. Neaton, E. Zojer and L. Kronik, J. Chem. Theory Comput., 2014, 10, 1934–1952 CrossRef CAS PubMed .
  38. D. Lüftner, S. Refaely-Abramson, M. Pachler, R. Resel, M. G. Ramsey, L. Kronik and P. Puschnig, Phys. Rev. B: Condens. Matter Mater. Phys., 2014, 90, 075204 CrossRef .
  39. S. Refaely-Abramson, S. Sharifzadeh, N. Govind, J. Autschbach, J. B. Neaton, R. Baer and L. Kronik, Phys. Rev. Lett., 2012, 109, 226405 CrossRef PubMed .
  40. Y. N. Yuan, M. J. L. Mills, P. L. A. Popelier and F. Jensen, J. Phys. Chem. A, 2014, 118, 7876–7891 CrossRef CAS PubMed .
  41. I. Tamblyn, S. Refaely-Abramson, J. B. Neaton and L. Kronik, J. Phys. Chem. Lett., 2014, 5, 2734–2741 CrossRef CAS PubMed .
  42. J. C. Love, L. A. Estroff, J. K. Kriebel, R. G. Nuzzo and G. M. Whitesides, Chem. Rev., 2005, 105, 1103–1169 CrossRef CAS PubMed .
  43. G. Heimel, L. Romaner, J. L. Brédas and E. Zojer, Phys. Rev. Lett., 2006, 96, 196806 CrossRef PubMed .
  44. Z. F. Liu, D. A. Egger, S. Refaely-Abramson, L. Kronik and J. B. Neaton, J. Chem. Phys., 2017, 146, 092326 CrossRef .
  45. A. Barth, Prog. Biophys. Mol. Biol., 2000, 74, 141–173 CrossRef CAS PubMed .
  46. E. Bellet-Amalric, D. Blaudez, B. Desbat, F. Graner, F. Gauthier and A. Renault, Biochim. Biophys. Acta, Biomembr., 2000, 1467, 131–143 CrossRef CAS .
  47. D. E. Lopez-Perez, G. Revilla-Lopez, D. Jacquemin, D. Zanuy, B. Palys, S. Sek and C. Aleman, Phys. Chem. Chem. Phys., 2012, 14, 10332–10344 RSC .
  48. C. Freysoldt, P. Rinke and M. Scheffler, Phys. Rev. Lett., 2009, 103, 056803 CrossRef PubMed .
  49. J. M. Garcia-Lastra, C. Rostgaard, A. Rubio and K. S. Thygesen, Phys. Rev. B: Condens. Matter Mater. Phys., 2009, 80, 245427 CrossRef .
  50. J. C. Inkson, J. Phys. C: Solid State Phys., 1973, 6, 1350–1362 CrossRef CAS .
  51. J. B. Neaton, M. S. Hybertsen and S. G. Louie, Phys. Rev. Lett., 2006, 97, 216405 CrossRef CAS PubMed .

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/c7cp08043c

This journal is © the Owner Societies 2018