Chemical Modulation of Contacts Control Energy Level Alignment in

Subscriber access provided by Washington University | Libraries

Article

Protein Electronics: Chemical Modulation of Contacts Control Energy Level Alignment in Gold-Azurin-Gold Junctions. Jerry A Fereiro, Gilad Porat, Tatyana Bendikov, Israel Pecht, Mordechai Sheves, and David Cahen J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b07742 • Publication Date (Web): 20 Sep 2018 Downloaded from http://pubs.acs.org on September 20, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Protein Electronics: Chemical Modulation of Contacts Control Energy Level Alignment in Gold-Azurin-Gold Junctions Jerry A Fereiro 1, Gilad Porat 2, Tatyana Bendikov 3, Israel Pecht 4*, Mordechai Sheves 2*, David Cahen 1* Affiliations: 1

Department of Materials and Interfaces, Weizmann Institute of Science, Rehovot, Israel. Department of Organic Chemistry, Weizmann Institute of Science, Rehovot, Israel. 3 Department of Chemical Research Support Unit, Weizmann Institute of Science, Rehovot, Israel 4 Department of Immunology, Weizmann Institute of Science, Rehovot, Israel. 2

Corresponding authors; email: [email protected], [email protected], [email protected].

1 ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 22

Abstract: Making biomolecular electronics a reality will require control over charge transport across biomolecules. Here we show that chemical modulation of the coupling between one of the electronic contacts and the biomolecules in a solid-state junction allows controlling electron transport (ETp) across the junction. Employing the protein Azurin (Az), we achieve such modulation as follows: Az is covalently bound by Au-S bonding to a lithographically prepared Au electrode (Au-Az). Au nanowires (AuNW) onto which linker molecules, with free carboxylic group, are bound via Au-S bonds, serve as top electrode. Current-voltage plots of AuNW-linkerCOOH//Az-Au junctions show step-like features, due to resonant tunneling through discrete Az energy levels, as shown earlier. Forming an amide bond between the free carboxylic group of the AuNW-bound linker and Az, yields AuNW-linkerCO-NH-Az-Au junctions. This Az-linker bond switches the ETp mechanism from resonant to off-resonant tunneling. By varying the extent of this amide bonding, the current-voltage dependence can be controlled between these two mechanisms, thus providing a platform for altering and controlling the ETp mechanism purely by chemical modification in a two-terminal device, i.e., without a gate electrode. Using results from conductance, including the energy barrier and electrode-molecule coupling parameters extracted from current-voltage fitting and Normalized Differential Conductance analysis, and from Inelastic-Electron-Tunneling- and Photoelectronspectroscopies, we determine the Az frontier orbital energies, with respect to the Au Fermi level, for four junction configurations, differing only in electrode-protein coupling. Our approach and findings open the way to both, qualitative and quantitative control of biomolecular electronic junctions.


Page 3 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction: Measuring and controlling electron transport, ETp, through molecules sandwiched between two electrodes is a major goal in molecular electronics. As electronic coupling within such molecules and with the electrodes strongly influences the ETp across such junctions

1, 2, 3

,

understanding how the coupling strength affects the ETp mechanism is of key importance in designing molecular junctions

4, 5

. One way to optimize the molecule-electrode coupling

strength is the choice of proper electrode binding groups 6, so that they can interact with both electrodes and provide efficient electronic coupling between the molecule(s) and the electrodes 7, 8, 9

. The molecule-electrode coupling strength can be reduced by introducing an ultrathin

insulating layer on the substrate, with Al2O3 as a popular choice 10, 11. Danilov et. al showed that adding a methylene unit at the end of a linearly conjugated molecule changes the charge transport mechanism from coherent tunnelling to coulomb blockading by the molecules 3. When proteins are used as charge carriers, both the coupling strength orientation relative to the electrode

12, 13

2

and protein

play a major role in ETp. Earlier we showed that by

inserting a linker molecule between the electrode and the protein Azurin(Az), the currentvoltage (I-V) / conductance-voltage (G-V) characteristics of Au-protein monolayer-Au junctions can be switched from resonant to off-resonant tunneling

14

. Here we report that by

modulating the chemical nature of protein-linker interactions we can modulate the alignment of frontier energy levels in and out of resonance, as reflected in the electron transport across the Az monolayer. A Monolayer of Mercapto-Propionic Acid (MPA), covalently bound to the Au nanowire (AuNW) via its thiol, with a free carboxylic group at its other end was used as linker/spacer (see Fig. 1B). To vary the coupling strength between linker and protein we covalently bind the linker to free amino residues on the protein surface, using the EDC/NHS crosslinking reaction

15

(see

Fig. 1F). We found that such covalent linker-protein (amide) bond increases the electrodeprotein coupling strength, so that it now becomes comparable to a junction without a spacer (i.e., bare gold AuNW). We also examined an intermediate situation, where only part of the MPA monolayer generated a covalent amide bond with the Az molecules. The resulting G-V plots indicate that the formation of the protein-linker amide bond affects the alignment of the frontier orbital energy levels. These plots also indicate that the energy separation between the 3 ACS Paragon Plus Environment


Page 4 of 22

frontier energy levels and the electrode’s Fermi level increases with the coupling i.e., covalent linker - protein bond formation pushes the frontier energy levels further away from the Au Fermi level, in accordance with a tight binding picture, so that the levels are not anymore in resonance. Inelastic Electron Tunnelling (IETS) Spectra obtained from the four different junction configurations show clear differences in terms of line shapes and intensities. The IETS results, along with the key parameters (barrier height (εo) and the coupling strength (Г)) extracted from fitting I-V curves to a cubic polynomial using Taylor expansion and mapping it to single level Landauer model, show that decreasing the electrode-protein coupling strength enhances the vibrational inelastic tunneling intensity electrode-protein coupling strength

16, 17, 18

19, 20, 21

. The correlation of IETS intensity/shape with

along with results from Normalized Differential

Conductance analysis and Photoelectron Spectroscopy measurements then allows to determine the position of Az frontier orbital energies, close to, and with respect to Au Fermi level, for configurations differing only in electrode-protein coupling. Thus, these results demonstrate how the energy level alignment in a two-terminal protein junction can be modulated by chemical modification, without using a gate electrode. Junction fabrication: Protein monolayer formation and junction fabrication were carried out following previously published procedures 14, 22. Briefly, µm-sized Au electrode pairs were fabricated on a Si wafer by photolithography. The disulphide bridge of Az opens upon contact with the Au surface and one or both SH-groups of the cysteines then form Au-S bonds to it 23, 24. This leads to the formation of controlled, oriented, protein monolayers, sufficiently robust to carry out solid-state ETp measurements 5 down to cryogenic temperatures (~ 10K) 22. Characterization of the protein monolayers was done by ellipsometry and AFM (thickness), AFM (topography), UV-Vis (see SI, section 8) and PM-IRRAS spectroscopies (see SI, section 3). A “soft” nondestructive contact was made on top of the protein monolayer, using the “suspended-wire” technique

25, 26

, where individual Au nanowires (AuNWs) are electrostatically trapped between

pairs of lithographically prepared Au micro-electrodes

14

, forming junction between the Az

monolayer on the Au micro-electrode and the electrostatically trapped single AuNW (see schematics in Fig. 3). 4 ACS Paragon Plus Environment

Page 5 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1: Schemes of the Au nanowire (AuNW) chemical modifications A) AuNW; B) Au NW surface modified with MPA (see text); C) and D) show the Au NW surface after treatment with different concentrations of EDC/NHS (see text). E) Az monolayers attached covalently to the Au substrate via thiolate bonds with surface exposed amino residues F) Covalent amide bond formed between the linker’s activated carboxylic group and a free amino group (e.g. of a lysyl residue) exposed on the Az surface. N.B. sketches and their parts are NOT to scale! Figure 1 A, B, C and D schematically illustrates the different reactions which lead to the chemical modification of the AuNW, before being electrostatically trapped between the Au micro-electrodes. For details of the preparation of the AuNW and their characterization, see SI section-1. The AuNW were immersed in a 10mM aqueous solution of MPA forming selfassembly of a monolayer via binding its -SH terminal to the Au surface yielding S-Au bonds (Fig. 1B). To characterize the structure and the quality of the monolayers on AuNW, similar monolayers were prepared on a 2 cm2 Au substrates. Fig. 2A shows the PM-IRRAS spectra of the MPA monolayer on the Au surface, with the main peak at 1730 cm-1 of the (C=O) stretching vibration of free carboxylic (-COOH) group. 5 ACS Paragon Plus Environment


Page 6 of 22

The free carboxylic group of the MPA monolayer was then reacted with 1-ethyl-3(3dimethylaminopropyl) carbodiimide hydrochloride (EDC), followed by reaction with Nhydroxysulfo-succinimide (sulpho-NHS) (see Fig. 1D)

15

. This process was followed by PM-

IRRAS, showing that after these reactions the main IR peak which was previously observed at 1730 cm-1 (-COOH groups) was replaced by a peak at 1750 cm-1, (due to the presence of carbonyl stretching vibration of O=C-ONR group), along with two other satellite peaks corresponding to the carbonyl stretching vibrations of the succinimide moiety at 1780 and 1815 cm-1 15. In order to produce an intermediate situation, where only a part of MPA molecules in the monolayers form covalent amide bond with Az, a lower concentration of EDC/NHS was used to chemically activate the free -COOH group of the MPA monolayer (see Fig. 1C). Formation of this intermediate situation, is supported by the PM-IRRAS spectrum (see SI Fig. S5), showing the presence of both free -COOH and EDC/NHS activated carboxylic groups on the surface. The next step was reacting the latter succinimide moiety of the monolayer with surface exposed amino residues (e.g. lysines and N-terminal amino group) of Az to form an amide bond

27

(see Fig. 1F). The PM-IRRAS results of this product (see Fig. 2C) show

formation of the amide I and amide II peaks at 1653 and 1544 cm-1 respectively 28, consistent with the presence of Az on the surface. The likely candidates for this reaction with Az are surface-exposed Lys residues that can face the AuNW. Which specific one this are depends on the orientation of Az with respect to the AuNW. A slight change in Az orientation between the NW can change the point of forming lysine-protein amide bond(s) (see SI, section-2). Az is tens of times larger than the crosslinker arms diagrammed in the reaction (see Fig. 1), thereby leaving some unreacted fraction of EDC/NHS groups on the AuNW surface as indicated by the PM-IRRAS results shown in Fig. 2C.

The schemes in Fig. 3 show the four different Au-(Az) - Au junction configurations used in this study, obtained from electrostatically trapped AuNWs modified with different linkers. In all cases, the bottom electrode was covalently bound to Az. For clarity of notation, the following names are given to the different employed junction configurations: (I) Au-Az-AuNW (II) Au-Az//MPA-AuNW; III is represented as Au-Az-X-AuNW [ -X- =(CH2)2CO-NH and //MPA] for the intermediate situation where only part of MPA is activated with EDC/NHS to 6 ACS Paragon Plus Environment

Page 7 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


form the amide bonds leaving behind inactivated MPA molecules on the Au surface; (IV) AuAz-(CH2)2CO-NH-AuNW, for the situation where the MPA has been fully activated with EDC/NHS. In the short-hand notation that we use for the junctions, we distinguish between the strong and weak interactions: “-” denotes strong covalent and “//” denotes weak electrostatic/mechanical bond. Here on, for simplicity, different junction configurations will be denoted as I, II, III and IV.

Figure 2: PM-IRRAS spectra in the 1100 and 1900 cm-1 range measured on A) Monolayer of MPA bound to the Au surface, main peak at 1730 cm-1 of the (C=O) stretching vibration of free carboxylic (-COOH) group B) MPA monolayer on Au surface following reaction with EDC/NHS, main peak at 1750 cm-1 (due to the presence of carbonyl stretching vibration of O=C-ONR group), along with two other satellite peaks at 1780 and 1815 cm-1 , corresponding to the carbonyl stretching vibrations of the succinimide moiety C) EDC/NHS activated MPA monolayer on Au surface following the reaction with Az forming the amide I and amide II peaks at 1653 and 1544 cm-1 respectively; the absorbance values are given in arbitrary units. Current-Voltage and Conductance measurements: Analysing the shape of the I-V and G-V plots of the four different configurations at cryogenic temperatures (10K) provides an indication of the ETp mechanism via those junctions. 7 ACS Paragon Plus Environment


Page 8 of 22

In all these configurations, Az is covalently bound to the bottom electrode via cysteine thiolate residues. Therefore, we can ascribe the differences observed in I-V and G-V relations between the different junction configurations, mostly to the altered coupling strength of Az to the top AuNW electrode. For configurations I and IV, the I-V curves are linear, whereas for II and III show non-linear behaviour (see Fig. 3). This difference in the shape of I-V curves (linear to non-linear), is consistent with a transition from strong to weak electronic coupling for the given configurations

1, 29

. I-V plots for configuration II exhibits a step-like behaviour, whereas for

configuration III a sigmoidal I-V shape was observed. Step-like features observed in the I-V plots was previously reported for single molecule ETp measurements 30, 31 and by us for protein monolayers 14.

Az monolayers attached covalently to the Au substrate via thiolate bonds consist of a sequence of Au-binding peptides on the freely exposed surface, providing strong coupling interaction with the AuNW at the other end. Although the nature of the AuNW-Az interaction is ill-defined (being only a physical contact), the linear I-V (see Fig. 3A, red curve) plot observed for configuration I (i.e., using bare Au electrodes) indicate strong electronic coupling of Az with both electrodes

29

. The symmetry observed in the I-V plot at different bias polarity (Fig.

3A) provides further indication that the coupling strength to the bottom Au electrode (Au-S) is comparable to that of the opposite end (Az-AuNW contact) 32, 33. The G-V plot (Fig. 3A, black curve), has small kinks (Fig. 3A-black line), consistent with the opening of inelastic conducting channels at voltages corresponding to energies of vibrational modes

34, 22

. The observed dip at

near zero bias is attributed to the large number of low energy- vibrations 22, 35.


Page 9 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3: A) (top left), Current – voltage, I-V (red) and conductance - voltage, dI/dV – V (black) plots of the Az junction between -0.5 and +0.5 V for configuration I; B) (Top right) I-V (green) and conductance - voltage, dI/dV – V (black) plots for configuration II. The small red circle represents the shoulder peaks observed at the tail of the main resonance peaks; C) (Bottom left) I-V (blue) and conductance - voltage, dI/dV – V (black) plots for configuration III; D) (bottom right) I-V (violet) and conductance - voltage, dI/dV – V (black) plots for configuration IV. Yellow circle shows the similar (bias) voltages at which the conductance peak was observed for configuration II and III. All data are for experiments at ~10 K. I-V measurements on configuration II junctions, AuNW with MPA monolayer, exhibit step-like behaviour, typical for resonant tunneling (see Fig. 3B, green line). As shown previously 14, the observed conductance peak (see Fig. 3B, black line) can be interpreted as the result of the alignment of one of the frontier orbital energy levels with the work function of the



Page 10 of 22

MPA-modified Au NW (~ 4.75 eV vs. vacuum 36, for UPS data see Fig. 5D ). Such alignment of the frontier orbital energy levels with the modified work function makes the orbital energy levels accessible to resonant transport

14

. In our recent report, we have already shown that the

resonance peaks observed in Fig. 3B (black line) do not form a progression, whereas the shoulder peaks observed at the tails of the main resonance peaks in Fig. 3B (marked in red circles) do indeed fit the progression pattern of vibronic states 14, 29. These peaks of ~ 50 mV or 400 cm-1 away from the main one (see Fig. 3B) fits well with the Cu-S stretching vibrations of Az. These satellite peaks were thus interpreted as the vibronic signature of the Az Cu(II) coordination sphere 37, 38 providing direct evidence for the involvement of metal ion in resonant tunneling transport via the protein molecule.

The shape of the G-V plot (see Fig. 3C, black curve) for configuration III shows features similar to those observed for configuration I and II. For example, the dip in conductance 34 at low bias is seen also with configuration I, whereas the conductance peak (see Fig. 3C, marked in yellow circles) observed at higher bias is seen also with configuration II. Comparison of the conductance peak observed for configurations II and III shows similar (bias) voltages at which the peaks were observed (see Fig. 3B & 3C, marked in yellow circles). Unlike the strongly coupled (I and IV), beyond ~ ±0.1V, the I-V plots of configuration III break its linearity. At this voltage the energy is getting close to resonance, we do not observe as clear conductance peaks as in configuration II, possible due to less defined crosslinking reaction. To substantiate this interpretation of the G-V data for configuration III, we carried out a Normalized Differential Conductance (NDC) analysis

39

of the I-V curve for configuration III

(Fig. 3C, blue curve). NDC is a generic tool that is applicable for any molecular junction and for tunneling charge transport in general. Different I-V relations (e.g., Ohmic, generic tunneling) are represented differently on NDC plots 39. The sharp NDC rise observed near ±0.1V (see Fig. S9B in the SI) resembles resonance peaks of a single-level Landauer model, though the attenuated NDC response at higher biases (|| > ±0.1), indicates the competition between different transport mechanism 39. The I-V and G-V plots obtained for configuration (IV) are similar to those obtained for configuration I (even though IV has peptide bonds linking the Az to the MPA). The similarity 10 ACS Paragon Plus Environment

Page 11 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


of the results indicates that off-resonant tunneling dominates the transport mechanism for configuration IV. It has been shown previously that electronic inductive effects can lead to shifts in molecular energy levels in the system with respect to the electrodes’ Fermi levels (EF), in molecular monolayers 16 and single molecule junctions 1. The observed shift in energy levels is the result of electronic equilibration of the system upon its perturbation by e.g. chemical bond formation between the electrode and the inserted linker molecule. We therefore postulate that the covalent bond formed between the linker and the protein leads to strong electronic inductive effects shifting the energy levels further away from the Au Fermi level resulting in off-resonant tunneling. This notion is also supported by the higher barrier ( ) values obtained from the I-V fitting analysis (see Table- 1) for the different configurations. The stability and the reproducibility of all the junctions studied were tested using different samples prepared on different days with several junctions tested within each sample (see SI Fig. S8). When configuration I and IV are employed, ~ 90% of the junctions measured shows similar results. The resonance peak position observed for configuration II varies slightly (~0.1 V) between different junctions, and such peak-like structures were seen in ~ 70% of the junctions measured whereas for configuration III similar results were obtained for ~55% of the measured junctions.

A detailed ab initio computation of the electronic structure of the Au-Az-AuNW junctions, is needed to provide insights into coupling values via these junctions. Although these calculations can provide good predictive power regarding the position of energy level, detailed ab initio calculation still remains a challenging task. Therefore the (first-principles) justification of the value of those parameters is, regrettably, well beyond the scope of the current study. Hence, in the present study we have used an alternative method to extract key parameters (see Table-1), i.e., by fitting the I-V curves (in Fig. 3) to a cubic polynomial using Taylor expansion and then mapping it to a single level Landauer model

39

(for more details see SI section 5).

Section 4 in SI gives a detailed theoretical explanation of the I-V and G-V relationships, observed for different junction configurations. Briefly, tunneling current due to transport facilitated by a single energy level according to the Landauer model is given by: I = 4NG Γ

ε +

− ∙ − 2

(1)



Page 12 of 22

where ‘e’ is the electron charge, ‘’ measures the asymmetry of the bias partition (found to be negligible in this case). ‘ ’ is the barrier height (in eV) and ‘N’ is the number of molecules in the junction, which is chosen and is a constant and ‘" ’ is the quantum conductance. Fig. S5 in the SI shows the I-V curve fitting analysis carried out on different junction configurations. The extracted parameters from the I-V curve fitting such as the position of the frontier energy levels ‘ ’ and its Lorentzian broadening, ‘Γ# ’ (Γ# = $Γ%∙ Γ& ) (in eV) are given in Table-1 (also see SI, section 5). For all configurations studied here the protein is bound to the Au substrate by cysteines C3 and/or C26; therefore, we assume that coupling to the bottom electrode Γ% is

similar in all cases and that the difference observed in Γ# that we find, for the different junctions

originates mainly from the difference in coupling to the top electrode (Γ& ). To understand the +

trend in the coupling with different configurations we define relative coupling Γ'() = * ,,. 0 . + ,,/

This value provides the ratio between the coupling to the top electrode of configuration i (where i = configuration I, II, III or IV) and configuration I. Table-1 shows that the value of Γ'() varies

in order: configuration I > IV > III > II. The extracted parameters values ( 123 Γ# ) support our previous interpretation that forming a linker-protein covalent bond increases the coupling strength and pushes the frontier energy levels further apart (see Table-I), so that the one that was closest in energy to the electrode’s Fermi level is now farther removed, resulting in offresonant tunneling.

Table 1. Shows the parameters obtained by fitting the I-V curves shown in Fig.3 for different configurations with single-Landauer cubic polynomial using Taylor expansion. Assuming = 0.5 and N=100. Relative Γ'() is calculated using configuration I as a standard. * The values were extracted by fitting the I-V curve in a low bias range (-0.1 to 0.1V). Comparison of IETS Results:


Page 13 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4 shows a plot of the second derivative of the current-voltage (d2I/dV2) vs. V; namely, the IETS results for configurations I, II, III and IV. IETS provides insight into some important spectroscopic and structural data, particularly, information regarding the strength of the vibronic coupling itself

22, 21, 19

. IETS measurements were carried out by recording the first

and the second derivatives as harmonic signals, using lock-in techniques with a DC bias added to an AC signal. Once a stable junction has been established, we were able to sweep the bias voltage several times without significant change of the IETS, allowing repeated IETS measurements on the same junction. For the different configurations studied here we observed clear differences in the IETS data in both their line shapes and intensities (see Fig. 4). The IETS spectra for configuration I and IV exhibits peaks, indicating a far off-resonant tunneling as the dominant transport mechanism

19

. Although there are no selection rules in IETS as there are in IR and Raman

spectroscopy, certain selection preferences have been established

40

. Molecular vibrations with

net dipole moments perpendicular to the interface of the tunneling junction have larger peak intensities than vibrations with net dipole moments parallel to the interface

40

.The IETS

spectrum obtained for configuration I and IV, is dominated by the C-H stretching peak observed around 3000 cm-1, and the amide I and amide II bands observed at 1640/1520 cm-1 respectively (see Fig. 4A & D).



Page 14 of 22

Figure 4: IETS, d2I/dV2 – V, for different junction configurations; significant modifications in spectral intensity and line shape are observed. All data are for experiments at 10 K. Unlike results obtained with configuration I and IV (see Fig. 4A and D), d2I/dV2) vs. V for configuration II (see Fig. 4B) does not show any clear peak features; rather we observe derivative-like line shapes. The amplitude of the IETS features has increased more than hundred-fold compared to that observed using configuration I. Theoretical and experimental results have shown that the difference in energy between the Fermi levels of the electrodes and the frontier molecular orbitals is one of the key parameter that determines the intensity and line shape of the IETS spectrum

34, 21, 41

. The enhancement in intensity and the modification of the

line shape observed using configuration II is due to the formation of new hybrid energy levels, lying closer to electrode’s Fermi level than to the frontier energy levels of the protein itself. These hybrid energy levels are formed by the interaction of Cu(II) (coordination shell) with the 14 ACS Paragon Plus Environment

Page 15 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


linker (MPA) molecule 14. Previously we have shown that the Cu(II) ion is required in order to get these derivative-like IETS features, from which we deduced that Cu(II) ion is involved in the formation of these hybrid energy levels 14. The large change in intensity and spectral line shape observed for configuration II, when the frontier energy level is closer to electrode’s Fermi level (on-resonance) can be explained using low-order perturbative model reported by Baratoff and Persson model

42

42, 43

. Considering this

, the IETS spectrum for configuration II, where resonant tunneling was observed,

depends on the position of the frontier energy levels (56 ) , on the width (7) of the Cu(II) induced hybrid energy level and on its coupling to the molecular vibrations, 85 (see Fig. 6B). The electron-phonon coupling term 85 describes the shift of 56 under the influence of the excitation of a particular vibrational mode.

Baratoff and Persson found that the relative change in the conductance upon reaching the threshold for excitation of a molecular vibration |eV| = Ω is proportional to: Γ (59: + Ω − 56 ) − 85 2 Γ Γ (56 − 59 : ) + (59: + Ω − 56 ) + 2 2

(2)

where 59: = 5: − , with 5: being the Fermi level of the unbiased electrodes, ‘e’ is the charge

of an electron, ‘V’ is the applied bias, and Ω is the energy of the vibrational mode. The

numerical modelling results of Galperin et al.

35

also show the evolution of an IETS feature

from a peak (off-resonant) to a peak like-derivative (resonant) as the position of the frontier energy level moves closer to the Fermi level. For configurations II and III, depending on the position of E? and the width (Γ) of the Cu(II) induced hybrid energy level, the amplitude of the IETS features can vary, and the line shape can be a peak- like derivative.

IETS profiles (intensities and the peak shape) have previously been used as an indirect probe to determine the alignment of frontier energy levels with respect to the metal E@ 21, 41. The larger the enhancement in amplitude the closer the frontier energy levels to EF 21. Therefore, for configurations II and III, where the IETS intensities are larger, the frontier energy levels are energetically closer to the EF than those obtained using configuration I and IV (far off15 ACS Paragon Plus Environment


Page 16 of 22

resonance), in agreement with the theoretical calculations. For the off-resonant cases (e.g., configurations I and IV), the tail of the frontier level enhances the probability of vibrational excitations observed as peaks in the IETS (d2I/dV2 -V) spectrum

35

. Whereas, for the on-

resonant case (configuration II), the transported electrons occupy a transient vibronic state observed as peaks in the G-V spectra. This leads to an even stronger vibrational signal as observed for configuration II.

UPS data: UPS measurements were used to correlate molecular junction conductance with energy level alignment

16, 17, 44

. Work function values, determined from UPS measurements are

sensitive to a local vacuum level shift induced by the new chemical composition of the (electrode) surface

17

. The UPS spectrum can be used to determine the work function of the

sample using the high binding energy cut-off (HBEC) where WF= 21.22 - HBEC (21.21eV is the energy of the He I radiation source), as shown in Fig. 5.


Page 17 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5: A Close-up of the HBEC region for Au surface, modified with different molecular structures, showing that electron withdrawing groups shifts the apparent work function to higher energies. B shows the measured work function and the changes in work function (∆WF) obtained using different linkers. C, D, and E show the schematics of the vacuum level shift, induced by binding respective linkers to Au surface. The Au surface with a measured work function of 5.05 eV (in the UHV of the UPS spectrometer) was used as the reference (see Fig. 5A). The modification of Au surface with MPA monolayers caused only a small shift (∆ϕ= -0.32 eV) (see Fig. 5A, green line) in HBEC, whereas the succinimide moiety of EDC/NHS causes a significant shift (∆ϕ ~ -0.8 eV, red line in Fig. 5A). Surface bound molecules with electron withdrawing groups such as -COOH or succinimide shift the apparent work function towards higher energies. The table in Fig. 5B shows the work functions of the modified Au surfaces, along with the change ∆WF with respect to unmodified Au surface. Here we stress that the measured work function for the Au-MPA sample (4.73 eV) is very close to the redox-potential of the Cu(II) centre in aqueous solution, translated to the solid state scale vs. vacuum, (~ 4.75 eV) 36. For the case of the cyclic pyridinyl ring of EDC/NHS the measured work function is 4.25 eV, which is 0.5 eV away from the redox potential of Cu(II) centre. The succinimide forms a strong covalent bond with the free lysine groups on the protein surface. The strong coupling can further lead to charge re-distribution, causing a further shift in work function. For configuration (III), where only partial peptide bond formation has occurred, the measured work function is 4.45eV, i.e.~ 0.3eV away from the redox potential of the Cu(II) centre. The UPS results for configuration III provides additional support for our previous claim that the formation of a covalent bond between the linker and the protein pushes the frontier energy level away from resonance.



Page 18 of 22

Figure 6: A shows schematic illustration of a solid-state protein junction for configuration I & IV, indicating off-resonant tunneling transport at an applied bias, V. The strong coupling is expressed by the large broadening of the energy levels (Lorentzian shown in yellow). The black arrow indicates elastic tunneling whereas, the blue and the brown arrow shows inelastic tunneling process occuring at different energies, represented by a molecular potential energy curve with two different excited vibrational energy levels at the bottom of the figure. B The onresonant tunneling case in the presence of the linker at an applied bias, V (configuration II). Resonant transport becomes possible if one of these levels (horizontal blue bars) falls within the bias window, V. Here the Cu(II) localized electronic states falls within the applied bias window at both low (indicated by black arrow) and high bias (indicated by red arrow). The arrows also indicate the resonant electronic excitation due to vibronic interaction at the Cu(II) localized electronic states. C Scheme of electron transport at low applied bias, indicating off-resonant tunneling. The black arrow indicates elastic tunneling whereas, the blue arrow shows inelastic tunneling process, represented by a molecular potential energy curve at the bottom of the figure. With increase in bias, one of the level falls within the bias window V, and resonant transport become possible (configuration III), indicated by red arrow. Fig. 6 illustrates the proposed transport mechanism for the different solid-state junction configurations used in this study. Based on the G-V and I-V fitting analysis along with the IETS and UPS results we conclude that when the proteins are strongly coupled to both the electrodes (in configuration I and IV), off-resonant tunneling dominates (see Fig. 6A) the charge transport. However, upon lowering the coupling strength at one end (configuration II) by introducing a linker/spacer resonant tunneling becomes dominant (see Fig. 6B) mechanism. Fig. 6C shows the proposed scheme of electron transport for configuration III, namely via the intermediate situation, where we can still observe resonant tunnelling, though at higher bias than in the configuration of the non-bound MPA linker.

Conclusions: Systematic study of the G-V, I-V patterns along with IETS measurements were carried out at low temperatures in order to investigate the charge transport mechanism via different AzAu-Az junction configurations. In all studied configurations Az was covalently bound to the Au surface on one side and chemical modifications were carried out on the linker to the other end. The G-V and IETS profiles obtained from the different configuration strongly suggest that by chemically modulating the coupling strength to the electrodes, one can switch the tunneling mechanism ‘IN’ and ‘OUT’ of resonance. The shape of the G-V plots, intensity of the IETS spectra and the barrier height ( ) values, obtained by fitting the I-V curves provide the 18 ACS Paragon Plus Environment

Page 19 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


position of the frontier energy levels with respect to the electrode Fermi level, indicating that the energy levels are moved away upon increased coupling. This shows that the energy-level alignment in Az based junctions can be regulated by chemically modifying interaction between linker and the protein. The results presented here therefore provide an attractive method for controlling the charge transport characteristics across protein-based two terminal devices without the requirement of a gate electrode.

Associated content: Supporting Information Information related to structure of Az molecule, materials and methods section, characterization data (PM-IRRAS, UV-Vis and AFM), detailed explanation of Landauer single-level model, I-V curve fitting analysis, Normalized Differential Conductance (NDC) analysis, reproducibility test and XPS analysis of Az on Au surface (including Figures S1-S12). Acknowledgments: We thank Prof. Juan Carlos Cuevas (Universidad Autónoma de Madrid, Spain), Dr. Ayelet Vilan and Mr. Ben Kayser (Weizmann Inst.) for fruitful discussions. JF is grateful to the Azrieli Foundation for the award of an Azrieli Fellowship. DC and MS thank the Israel Science Foundation, the Minerva Foundation (Munich), the Benoziyo Endowment Fund for the Advancement of Science and J & R Center for Scientific Research and the Kimmelman center for Biomolecular Structure and Assembly for partial support. M.S. holds the Katzir-Makineni Chair in Chemistry. References: 1. Frisenda, R.; van der Zant, H. S. J., Transition from strong to weak electronic coupling in a single-molecule junction. Physical Review Letters 2016, 117 (12). 2. Amdursky, N.; Ferber, D.; Bortolotti, C. A.; Dolgikh, D. A.; Chertkova, R. V.; Pecht, I.; Sheves, M.; Cahen, D., Solid-state electron transport via cytochrome C depends on electronic coupling to electrodes and across the protein. Proceedings of the National Academy of Sciences of the United States of America 2014, 111 (15), 5556-5561. 3. Danilov, A.; Kubatkin, S.; Kafanov, S.; Hedegard, P.; Stuhr-Hansen, N.; Moth-Poulsen, K.; Bjornholm, T., Electronic transport in single molecule junctions: Control of the molecule-electrode coupling through intramolecular tunneling barriers. Nano Letters 2008, 8 (1), 1-5. 4. Alessandrini, A.; Facci, P., Electron transfer in nanobiodevices. European Polymer Journal 2016, 83, 450-466. 5. Baldacchini, C.; Bizzarri, A. R.; Cannistraro, S., Electron transfer, conduction and biorecognition properties of the redox metalloprotein Azurin assembled onto inorganic substrates. European Polymer Journal 2016, 83, 407-427.



Page 20 of 22

6. Kaliginedi, V.; Rudnev, A. V.; Moreno-Garcia, P.; Baghernejad, M.; Huang, C. C.; Hong, W. J.; Wandlowski, T., Promising anchoring groups for single-molecule conductance measurements. Physical Chemistry Chemical Physics 2014, 16 (43), 23529-23539. 7. Xue, Y. Q.; Ratner, M. A., Microscopic study of electrical transport through individual molecules with metallic contacts. I. Band lineup, voltage drop, and high-field transport. Physical Review B 2003, 68 (11). 8. Xue, Y. Q.; Ratner, M. A., End group effect on electrical transport through individual molecules: A microscopic study. Physical Review B 2004, 69 (8). 9. Chen, F.; Li, X. L.; Hihath, J.; Huang, Z. F.; Tao, N. J., Effect of anchoring groups on singlemolecule conductance: Comparative study of thiol-, amine-, and carboxylic-acid-terminated molecules. Journal of the American Chemical Society 2006, 128 (49), 15874-15881. 10. Qiu, X. H.; Nazin, G. V.; Ho, W., Vibrationally resolved fluorescence excited with submolecular precision. Science 2003, 299 (5606), 542-546. 11. Wu, S. W.; Nazin, G. V.; Chen, X.; Qiu, X. H.; Ho, W., Control of relative tunneling rates in single molecule bipolar electron transport. Physical Review Letters 2004, 93 (23). 12. Venkat, A. S.; Corni, S.; Di Felice, R., Electronic coupling between azurin and gold at different protein/substrate orientations. Small 2007, 3 (8), 1431-1437. 13. Ocampo, O. E. C.; Gordiichuk, P.; Catarci, S.; Gautier, D. A.; Herrmann, A.; Chiechi, R. C., Mechanism of orientation-dependent asymmetric charge transport in tunneling junctions comprising photosystem I. Journal of the American Chemical Society 2015, 137 (26), 8419-8427. 14. Fereiro, J. A.; Yu, X.; Pecht, I.; Sheves, M.; Cuevas, J. C.; Cahen, D., Tunneling explains efficient electron transport via protein junctions. Proceedings of the National Academy of Sciences 2018, 115 (20), E4577-E4583. 15. Soeriyadi, A. H.; Gupta, B.; Reece, P. J.; Gooding, J. J., Optimising the enzyme response of a porous silicon photonic crystal via the modular design of enzyme sensitive polymers. Polymer Chemistry 2014, 5 (7), 2333-2341. 16. Sayed, S. Y.; Fereiro, J. A.; Yan, H. J.; McCreery, R. L.; Bergren, A. J., Charge transport in molecular electronic junctions: Compression of the molecular tunnel barrier in the strong coupling regime. Proceedings of the National Academy of Sciences of the United States of America 2012, 109 (29), 11498-11503. 17. Kim, B.; Choi, S. H.; Zhu, X. Y.; Frisbie, C. D., Molecular tunnel junctions based on pi-conjugated oligoacene thiols and dithiols between Ag, Au, and Pt contacts: Effect of surface linking group and metal work function. Journal of the American Chemical Society 2011, 133 (49), 19864-19877. 18. Cahen, D.; Kahn, A., Electron energetics at surfaces and interfaces: Concepts and experiments. Advanced Materials 2003, 15 (4), 271-277. 19. Kim, Y.; Song, H., Investigation of molecular junctions with inelastic electron tunneling spectroscopy. Applied Spectroscopy Reviews 2016, 51 (7-9), 603-620. 20. Reed, M. A., Inelastic electron tunneling spectroscopy. Materials Today 2008, 11 (11), 46-50. 21. Song, H.; Kim, Y.; Jang, Y. H.; Jeong, H.; Reed, M. A.; Lee, T., Observation of molecular orbital gating. Nature 2009, 462 (7276), 1039-1043. 22. Yu, X.; Lovrincic, R.; Sepunaru, L.; Li, W.; Vilan, A.; Pecht, I.; Sheves, M.; Cahen, D., Insights into solid-state electron transport through proteins from inelastic tunneling spectroscopy: The case of azurin. ACS Nano 2015, 9 (10), 9955-9963. 23. Friis, E. P.; Andersen, J. E. T.; Kharkats, Y. I.; Kuznetsov, A. M.; Nichols, R. J.; Zhang, J. D.; Ulstrup, J., An approach to long-range electron transfer mechanisms in metalloproteins: In situ scanning tunneling microscopy with submolecular resolution. Proceedings of the National Academy of Sciences of the United States of America 1999, 96 (4), 1379-1384.


Page 21 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


24. Chi, Q. J.; Zhang, J. D.; Friis, E. P.; Andersen, J. E. T.; Ulstrup, J., Electrochemistry of selfassembled monolayers of the blue copper protein Pseudomonas aeruginosa azurin on Au(111). Electrochemistry Communications 1999, 1 (3-4), 91-96. 25. Smith, P. A.; Nordquist, C. D.; Jackson, T. N.; Mayer, T. S.; Martin, B. R.; Mbindyo, J.; Mallouk, T. E., Electric-field assisted assembly and alignment of metallic nanowires. Applied Physics Letters 2000, 77 (9), 1399-1401. 26. Freer, E. M.; Grachev, O.; Duan, X. F.; Martin, S.; Stumbo, D. P., High-yield self-limiting singlenanowire assembly with dielectrophoresis. Nature Nanotechnology 2010, 5 (7), 525-530. 27. Friis, E. P.; Andersen, J. E. T.; Madsen, L. L.; Moller, P.; Ulstrup, J., In situ STM and AFM of the copper protein Pseudomonas aeruginosa azurin. Journal of Electroanalytical Chemistry 1997, 431 (1), 35-38. 28. The PM-IRRAS spectrum shown here is obtained by incubating Az with chemically activated EDC/NHS group. 29. Cuevas, J. C.; Scheer, E., Molecular electronics: An introduction to theory and experiment. World Scientific: 2017; Vol. 1. 30. Moth-Poulsen, K.; Bjornholm, T., Molecular electronics with single molecules in solid-state devices. Nature Nanotechnology 2009, 4 (9), 551-556. 31. Selzer, Y.; Allara, D. L., Single-molecule electrical junctions. In Annual Review of Physical Chemistry, 2006; Vol. 57, pp 593-623. 32. Kushmerick, J. G.; Whitaker, C. M.; Pollack, S. K.; Schull, T. L.; Shashidhar, R., Tuning current rectification across molecular junctions. Nanotechnology 2004, 15 (7), S489-S493. 33. Sepunaru, L.; Pecht, I.; Sheves, M.; Cahen, D., Solid-state electron transport across azurin: From a temperature-independent to a temperature-activated mechanism. Journal of the American Chemical Society 2011, 133 (8), 2421-2423. 34. Galperin, M.; Ratner, M. A.; Nitzan, A., Molecular transport junctions: vibrational effects. Journal of Physics-Condensed Matter 2007, 19 (10). 35. Galperin, M.; Ratner, M. A.; Nitzan, A., Inelastic electron tunneling spectroscopy in molecular junctions: Peaks and dips. Journal of Chemical Physics 2004, 121 (23), 11965-11979. 36. Skov, L. K.; Pascher, T.; Winkler, J. R.; Gray, H. B., Rates of intramolecular electron transfer in Ru(bpy)(2)(im)(His83)-modified azurin increase below 220 K. Journal of the American Chemical Society 1998, 120 (5), 1102-1103. 37. Andrew, C. R.; Yeom, H.; Valentine, J. S.; Karlsson, B. G.; Bonander, N.; Vanpouderoyen, G.; Canters, G. W.; Loehr, T. M.; Sandersloehr, J., Raman-spectroscopy as an indicator of Cu-S bond-length in type-1 and type-2 copper cysteinate proteins. Journal of the American Chemical Society 1994, 116 (25), 11489-11498. 38. Andrew, C. R.; Loehr, T. M.; Sandersloehr, J., Raman-spectroscopy as an indicator of Cu-S bond lengths and coordination geometries in copper-cysteinate proteins. Abstracts of Papers of the American Chemical Society 1994, 208, 362-INOR. 39. Vilan, A., Revealing tunnelling details by normalized differential conductance analysis of transport across molecular junctions. Physical Chemistry Chemical Physics 2017, 19 (40), 27166-27172. 40. Kirtley, J.; Hall, J. T., Theory of Intensities in inelastic-electron tunneling spectroscopy orientation of adsorbed molecules. Physical Review B 1980, 22 (2), 848-856. 41. Yu, L. H.; Zangmeister, C. D.; Kushmerick, J. G., Origin of discrepancies in inelastic electron tunneling spectra of molecular junctions. Physical Review Letters 2007, 98 (20). 42. Baratoff, A.; Persson, B. N. J., Theory of the local tunneling spectrum of a vibrating adsorbate Journal of Vacuum Science & Technology a-Vacuum Surfaces and Films 1988, 6 (2), 331-335. 43. Persson, B. N. J.; Baratoff, A., Inelastic electron-tunneling from a metal tip-The contribution from resonant processes. Physical Review Letters 1987, 59 (3), 339-342. 21 ACS Paragon Plus Environment


Page 22 of 22

44. Qi, Y. B.; Yaffe, O.; Tirosh, E.; Vilan, A.; Cahen, D.; Kahn, A., Filled and empty states of alkanethiol monolayer on Au (111): Fermi level asymmetry and implications for electron transport. Chemical Physics Letters 2011, 511 (4-6), 344-347.

TOC Figure:


Chemical Modulation of Contacts Control Energy Level Alignment in

Recommend Documents