Protective Layers Based on Carbon Paint To Yield High-Quality Large-Area Molecular Junctions with Low Contact Resistance

A major obstacle for transforming large-area molecular junctions into a viable technology is the deposition of a top, metallic contact over the self-assembled monolayer (SAM) without chemically damaging the molecules and preventing an interface-limited charge transport. Often a thin conducting layer is softly deposited over the SAM to protect it during the deposition of the metal electrode which requires conditions under which organic molecules are not stable. We report a new protective layer based on carbon paint which is highly conductive and has metallic-like behavior. Junctions made of SAMs of n-alkanethiolates supported by Au were characterized with both dc and ac techniques, revealing that carbon paint protective layers provide a solution to three well-known challenges in molecular junctions: series resistance of the leads, poor interface conductance, and low effective contact area related to the roughness of the interfaces. Transport is constant with coherent tunneling down to 10 K, indicating the carbon paint does not add spurious thermally activated components. The junctions have both high reproducibility and good stability against bias stressing. Finally, normalized differential conductance analysis of the tunneling characteristics of the junctions as a function of molecular length reveals that the scaling voltage changes with molecular length, indicating a significant voltage drop on the molecules rather than on the molecule–electrode interface. There is a clear inverse dependence of the scaling voltage on length, which we deduced has a tunneling barrier height of close to 2 eV. The paper establishes the reliability of carbon paint protective layers and provides a procedure for discriminating genuine molecular effects from interfacial contributions.


S5
Electrical measurements. We followed previously reported procedures for the data collection and analysis of the J(V) curves. 3 We recorded the I-V curves and performed current retention tests using the home-written code in LabView (Version 2010) and a Keithley 6430 sub-femtoamp remote source meter.
Normalized differential conductance (NDC) analysis. We used MatLab code and applying the NDC function (Eq. 5, main text) over parabolic approximation (Eq. 4, main text) to extract tunneling parameters. The J-V curves were re-casted in NDC plots after smoothing of the data using the spline-smoothing function of MatLab (factor = 0.9999) to reduce the noise. Impedance spectroscopy. We performed impedance measurements as described previously. 4 Briefly, we measured impedance spectroscopy using a Solartron impedance analyzer (Solartron 1260A with 1296A dielectric interface) in reference mode with a standard 10 pF capacitor as the external reference. All the impedance measurements were Section S1. To measure the thickness of the CP film, we prepared a patterned CP films on a glass substrate spin-coating CP on a glass substrate that was partially covered with Scotch tape. We removed the Scotch tape after spin-coating. The inset of Figure S2 shows a photograph of the glass surface with only one side coated with CP. Figure S2 shows the height profile (measured with a profilometer) of the edge of the CP film from which we determined the thickness of the film of ~1.0 µm. We determined the average size of the graphite flakes from SEM images obtained from diluted layers of the CP using different lot numbers to understand the batch-to-batch consistency of the CP product. Figure S3 shows the histograms of the graphite flake size distribution with Gaussian fits to these histograms for 3 different lot numbers. The Gaussian fit yields the Gaussian mean, along with the Gaussian standard deviation, of the size of the graphite flakes: 0.18 ± 0.02 µm for lot number 1220501 (Fig. S3A), 0.19 ± 0.01 µm for lot number 1230420 ( Fig. S3B), and 0.20 ± 0.02 µm for lot number 1239429 (Fig. S3C). Thus, we confirm that the CP film is highly reproducible and does not depend on the lot number. Figure S4 shows all the fabrication steps of the template-stripped carbon paint (CP TS ). We prepared a SAM of SC12 on Au TS (Fig. S4A) using a previously reported method. 1 The SAMs were formed by immersion of the Au TS substrate into the corresponding ethanoic solution of SC12 (Fig. S4B). The CP was deposited by spincoating at 6000 rpm ( Fig. S4C) followed by applied scotch tape support on the CP film ( Fig. S4D). After applied scotch tape support, we applied small pressure on the tape using S7 cotton buds to make sure tape support properly stick to the CP film surface. Finally, the CP film with tape support striped from the Au TS /SAM surface and then the CP TS film used for AFM, and XPS characterization. To study the batch-to-bacth consistency of the CP TS film, we measured topography of the CP surfaces with AFM for CP TS films prepared from the three different batches (lot numbers, 1220501, 1230420, and 1239429); these results are summarized in Figure S5. Figure S6 shows the AFM images that were used to determine the area available for contact following the same procedure reported in ref 5 . Briefly, we estimated area available for the contact from digital analysis of a high resolution AFM image of a Au TS and template stripped CP film, the available contact is the number of pixel within 0.2 nm from the top-most average plane of the digital image divided by the total number of pixels using Nanoscope analysis software.   Section S4. Statistical data analysis. We recorded statistically large number of data (see Table S1) for the Au TS -SCn//CP//Au junctions. We followed the procedure for statistical analysis of the junction data as reported before. 9 We plotted the value of log10|J| at given bias voltage ( Figure S11) in histograms and fitted Gaussians to these histograms to obtain the Gaussian log-standard deviation (σlog G) and the Gaussian log-mean of the value of J, <log10|J|>G, for each applied voltage; this was repeated for the various lengths of alkanthiolate SAMs. These data were used to construct the <log10|J|>G vs. V curves shown in the main text.      where the molecule is highly distorted by the huge electric field. Scenario C may partly apply for highly rough top-electrode that cannot produce an intimate contact.

Section 6. Impedance data analysis
The impedance data were validated (i.e., to establish the junctions were in thermodynamic equilibrium and stable during the measurements, and that higher harmonics were not important) with Kramers−Kronig (KK) transformations (see Section S7 and 8, Fig. S18 and 19) and Table S2 lists the values of χ 2 KK and the values of χ 2 fit which are similar to χ 2 KK. Figure S20 shows the Nyquist plots for the three different junctions. The phase spectrum shows one peak ( Figure 6B main text) and the Nyquist plot shows the presence of one semicircle ( Figure S20), which indicates the presence of one capacitance in the junctions.
The CPE contains an additional constant (ne) in the exponent which is used to account for non-ideal capacitive behavior due to defects in the electrodes (induced by, S26 e.g., surface roughness): ne = 1 represents an ideal capacitor, but here we used slightly smaller values in the range of ne = 0.98 -0.99 to obtain good fits (Section 7). The parallel plate equation is given by where ε0 is the permittivity of the free space, εSAM is the dielectric constant of the nalkanethiolate SAM, and Ageo is the geometrical area of the junctions. The intercept of the fit with Eq. S1 (Fig. S21) and the y-axis yields the stray capacitance of 6.19 µF/cm 2 , which close to the stray capacitance estimated from the equation S1 (see for details Section S9). The AlOx layer is very thin and we have shown before that Au directly deposited on this thin layer results in pinholes, 2 but the impedance results indicate that the CP does not penetrate the AlOx layer. We have confirmed this conclusion by measuring the J(V) curve of a Au TS //AlOx//CP//Au junction across which we could not measure a current within the detection limit of our electrometer (see Section S10, Fig.   S22). Thus, the SAM inside the micropore dominates the charge transport process across the junctions. From the capacitance of the SAM, CSAM, we extracted a relative dielectric constant of 3.1 ± 0.4 which is similar to previously reported values for n-alkanthiolate SAMs in other types of junctions. [12][13][14] and the results confirm that stray capacitances are insignificant (see Section S10 for measurements on Au TS //AlOx//CP/Au junctions).
Section S7. Kramer-Kronig (KK) analysis. Figure S18 shows the KK-plots of the Au TS -SCn//CP//Au junctions. No trends are visible, therefore we conclude that the data are linear and that the junctions did not change during the measurements. The measured impedance data are of reasonable quality (Table S2) with acceptable signal-to-noise S27 ratios (although noise levels were in general 2-3% in the low-frequency regime and increased to roughly 15% in the high-frequency regime). S28 Section S8. Residual plots. Figure S19 shows the residual plots for nonlinear least square fitting of the impedance data to the equivalent circuit shown in Figure 6A as described in the main text. The values of χ 2 Fit are summarized in Table S2. The χ 2 Fit values are close to the χ 2 KK values.   where, d, ε0, and εr, are the contact area of Au TS //AlOx//CP/Au junction, the distance between two electrodes of 10 nm, ε0 is the permittivity of the free space, and εr the relative dielectric constant of insulator (9 for AlOx), respectively. 1 This estimation yields Cstray = 0.79 μF/cm 2 . This estimation confirms that Cstray is not important in our devices.
Section S10. To confirm that the leakage current across the 10 nm AlOx is not important for our large area junctions, we deposited 10 nm AlOx on a patterned Au TS substrate S32 using ALD and then spin-coated a ~1.0 µm thick CP film on the Au TS /AlOx substrate.
Next, we deposited 100 nm Au with a dimension of 50 × 1000 µm through a shadow mask after which the excess the CP was removed with oxygen plasma etching. Figure   S21 shows the J(V) curve for the Au TS //AlO×//CP//Au junction, we did not observe a significant current across the junction, which confirms that the leakage current across the 10 nm AlOx is insignificant. Figure S22. The I(V) characteristics of a Au TS //10 nm AlO× //CP/Au junction.