Revealing the Nature of MoleculeâElectrode Contact in Tunneling

Subscriber access provided by TUFTS UNIV

Letter

Revealing the Nature of Molecule-Electrode Contact in Tunneling Junctions using Raw Data Heat-Maps Jacob Sporrer, Jiahao Chen, Zhengjia Wang, and Martin M. Thuo J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.5b02300 • Publication Date (Web): 30 Nov 2015 Downloaded from http://pubs.acs.org on December 1, 2015

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 free 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 accessible to all readers and 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.

The Journal of Physical Chemistry Letters 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 23

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

The Journal of Physical Chemistry Letters

Revealing the Nature of Molecule-Electrode Contact in Tunneling Junctions using Raw Data Heat-Maps Jacob Sporrer, 1§ Jiahao Chen,1,2§ Zhengjia Wang,1 Martin M. Thuo1,2,3* 1

Department of Materials Science and Engineering, Iowa State University, 2220 Hoover Hall, Ames, IA 50011 USA

2

Micro-electronic research center, Iowa State University, 133 Applied Sciences Complex I, 1925 Scholl Road, Ames, IA 50011 USA 3

Biopolymer and Bio-composites Research Team, Center for Bioplastics and Bio-composites, Iowa State University, 1041 Food Sciences Building, Ames, IA 50011 USA

§

These authors contribute equally to this work.

AUTHOR INFORMATION Corresponding Author *Email: [email protected]. Phone: +1-515-294-8581

ACS Paragon Plus Environment

1


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 23

ABSTRACT Mechanistic understanding of charge transport through molecular tunnel junctions requires reproducible and statistically relevant data sets. This challenge has been overcome by development of large area junctions, especially those based on liquid-metal physi-sorbed top-electrodes, such as eutectic gallium-indium. A challenge with these junctions, however, is an inability to diagnose the quality of contact between the top-electrode and the SAMs. Since tunneling currents are dependent on the distance between the two electrodes, we demonstrate that by analyzing all raw unfitted data derived from a measurement using heat-maps, one can deduce the quality of contact and other minor bias-dependent fluctuations in the charge transport behavior. We demonstrate that the use of 3D plots would be challenging to interpret but adoption of heat maps clearly capture details on junction quality irrespective of the total size of the dataset or molecules used. We propose representation of raw data, rather than reliance on statistics, as proof of quality junctions. TOC GRAPHICS

KEYWORDS: Molecular electronics, data analysis, charge tunneling, junction quality, large area junction.


2

Page 3 of 23

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


The promise of molecular electronics rests on the ability to engineer molecular wave-function through organic synthesis. For some time, however, the main challenge was a need for platforms to reproducibly and reliably study unimolecular systems. The liquid top-electrode platforms, especially with EGaIn (eutectic gallium indium alloy) pioneered by Whitesides (Figure 1a-b), have offered promise in their ability to rapidly collect large volumes of stable, reproducible and statistically significant data.1-5 As the technique develops, improvements in characterization and understanding of different components is vital, for example, the nature or characteristics of the top-electrode upon contact is still an on-going study.6-12 Alongside this, challenges in processing large volumes of data and validating the obtained data will enable design of tools and molecules for mechanistic studies needed for the translation of the obtained information into working devices.13 Background: There are two main classes of molecular junctions, viz; i) small area junctions, and, ii) Large area junctions. Small area junctions, usually single molecule junctions, are mainly based on the break junction technique - which include mechanically controllable break junctions (MCBJs) and electro-migration break junctions (EBJs).8, 14 Despite the fact that break junction techniques can produce large data sets,15 challenges in the fabrication process, difficulty in identifying the contact configuration, low-yield of working junctions, and, limited compatibility of molecules to electrodes (types of bifunctional molecules) make the derived data suffer from either large fluctuation (i.e. junction-to-junction variation) or small-volume of high quality information rich data.15-16 To overcome the fluctuations, conductance histograms have been adopted in most of the statistical data analysis.17-18 The use of heat-maps has also been reported, for example by Smaali and co-workers and Tao and co-workers, although this is limited to a few studies and no reports exist on their use with large area junctions.19-20


3


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 4 of 23

Large area junctions (predominantly self-assembled monolayer (SAM) based), are formed by introducing a top electrode in contact with a uniform assembly of many molecules (SAM) fabricated on a bottom electrode (predominantly Au or Ag). Large area junctions, therefore, measure the average properties across many molecules to rapidly accumulate reproducible and statistically relevant data. The presence of SAM defects (Figure 1c), however, results in variations between junctions fabricated on the same substrate, primarily due to the nature of contact (hence the differentiation of surface defects to thin-area and thick-area defects2, 21).22 Defects in organic molecular junctions include; grain-boundaries, step edges, adventitious impurities, and, molecular back-folding of long-chain molecules, can have different effect on the data from measurements. 4, 22 Inspecting and analyzing the whole range of data, however, can distinguish measurements made on defective areas from those made on fairly uniform junction areas. A large amount of data is, therefore, required to statistically minimize the effects of defects and overcome the resulting variance. The hanging mercury (Hg) drop electrodes (HMDE) technique, due to its relative simplicity and compatibility with a large selection of molecules, is a good tool for molecular junction studies.8, 23-24

Use of Hg is, however limited by major drawbacks that include; i) Hg is toxic, ii) Hg is

volatile, iii) Junctions are complicated by Hg electro-migration, and, iv) amalgamation with entire surface upon failure.9 The EGaIn technique, as an alternative to Hg, has been widely adopted for its stability, high yield and reproducibility.9, 21, 25 EGaIn is a non-Newtonian liquid metal at room temperature and can be shaped into a conical probe that can be used as the top electrode, forming a Metal-SAM//Ga2O3-EGaIn junction (Figure 1a-b). 4, 6, 26-27 The current EGaIn technique also has its own challenges, for example, the lack of an ability to make uniformly configured top electrodes and the difficulty in defining the influence of


4

Page 5 of 23

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


morphology of the tip on measurements.6 Degree of force applied on the top electrode to establish contact (which can alter transport mechanism28), or the method to prepare the top electrode, induces variant amount of stress on the SAM resulting in discrepancies in distribution of data between users and or labs. An example is the magnitude of the widely-observed odd-even effect in SAMs and in charge transport across the alkanethiolate junctions.1, 4, 25-26, 29-33 Disregarding the fact that a defective/rough surface will increase the chance to diminish the alignment of molecules and break the structure of SAMs, the origin of the recent discrepancy in the magnitude of the observed odd-even effect in charge transport has not been established. Thuo et al.4 and Jiang et al.1 reported the odd-even oscillation of tunneling current across AgTSSCn//GaOx/EGaIn junction (AgTS is the template stripped Ag surface and SCn is the nAlkanethiolate SAMs) using soft contact, while Baghbanzadeh et. al.29 showed an absence of the effect when they altered the method used to prepare the top-electrode. A closer look at the three data sets shows that where the odd-even effect has diminished, the variance in data point is small (σlog = 0.3), and reproducible not only across different molecules but across different substrates, suggesting that the data is precise. In the case where the odd-even effect was observed, the standard deviations varied across different molecules with Thuo et. al., using the conical tip electrode, reported the largest variance (σlog = 0.2-1.1) while Jiang et. al., with microfluidicsbased top electrode reported slightly lower range (σlog = 0.12-0.57). The variations in the latter two studies indicates less precise measurements. We hypothesized that the nature of contact between different electrodes might be the origin of the observed difference and that analysis of the raw data set can reveal the nature of the contact based on the distribution of the outliers and the skew of the distributions in counts. We inferred that with the different types of interfaces,


5


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 6 of 23

judgement of the nature of non-optimal contacts, which is either hard or soft, would be manifested in either a domination by high or low value outliers respectively.

Figure 1: Schematics of molecular tunneling junction and diagrams of analysis of tunneling currents thought the junctions. a) Schematics of performing current measurement using EGaIn tip. b) The formation of MSAMs//Ga2O3/EGaIn junction and interface 1 and 2. c) A schematic representation of defects in metal substrate and SAMs. d-f) Common approaches to represent charge-transport data derived from large area junctions, with each being limited to two variables at a time. g) a 3D plot of I-V data from EGaIn-based junctions illustrating that a simple histogram plot is not very informative or easy to interpret.


6

Page 7 of 23

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


The EGaIn junctions generate a large volume of data, hence Gaussian fits (that assume a stochastic process) and related plots (Figure 1d-f) have been deployed as reliable platforms for reporting charge transport data.1-5,

13, 26, 34-35

Although these data representations allow for valuable

comparisons, for example through comparison of Gaussian fits (J-V) and associated Gaussian mean, standard deviations and medians,1-2, 4-5, 13, 26 the eliminate outliers (through the fit or by the user) and often override skewness, key parameters in establishing bias in the measurements. To take advantage of this information, J-V data needs to be represented in a single plot that includes total counts at each bias. Since current methods of representing data (Figure 1d-f) are insufficient, while a simple 3D plot is difficult to decipher (Figure 1g), deploying heat-maps - a planar version of the 3D data, we not only capture the average J-V trace (Figure 1e), but also capture the skewness and outliers (that could be due to nature of contact or defects in the SAM) in a single figure. The ability to display the raw data also eliminates any bias associated with the quality of fit, or the fitting algorithms employed. Discussion: We demonstrate the diagnostic/quality-control capability of this simple approach to data analysis and representation, a Matlab based Large Area Junction Analysis (LAJA) method, by analyzing data collected from junctions of n-alkanethiolate SAMs (AuTS-SCn//Ga2O3/EGaIn) with; i) different levels of experimental quality (molecule quality, hence, quality of the SAM, different expertise, different experimental conditions), ii) a molecule that is known to back-fold (SC19) upon self-assembly,4 and, iii) soft or forced contacts (that is, we barely bring the topelectrode into contact with the SAM or push the top-electrode beyond what is normally considered a good physical contact). First, we intentionally fabricated junctions under different experimental conditions using a medium chain length even-carbon molecule (SC14), that is, we intentionally prepared the SAMs


7


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 8 of 23

under experimental conditions, many of which are known to affect the SAM quality (did not purify the thiol, data was corrected at different humidity and temperature conditions, and no vibration isolation), and, data was collected by users with different levels of expertise. We represented the data in both 2D and 3D formats to illustrate how low quality data (Figure 2). The obtained data is at least bimodal, that is, there are two dominant distributions with ‘shadow’ peaks on lower and higher end of each distribution, suggesting a skewness. We hypothesized that this skewness is an indication of the origin of the adjacent major distribution, with the higher log|J| values being from thin-area defects and, similarly, the lower log|J| values being from thick-area defects. The conventional method using Gaussian fits at a single bias fails to capture the extent of the noise in the data, especially when the data is fitted to a single Gaussian. To interpret the data, a 3D histogram of the distribution of current density with voltage (Figure 2a) was generated from LAJA, showing the discrete multimodal J-V distribution, a simple 3D histogram is challenging to decipher. In contrast, the corresponding heat-map (Figure 2b), generated from an aerial top or bottom) view of that ‘skyscraper’ histogram plot is clearer and not only gives the data distribution but also captures any variations across all the biases. For data that are this noisy, a J-V trace generated from a single Gaussian fit is also equally effective at showing that the data is not scientifically sound (Figure 2d). The poor quality is only revealed upon plotting data across the span of the bias ranges, but the J-V trace fails when a shorter bias range is use. It is common in the area of large-area junctions to consider data at a narrower bias range, e.g. ±0.1V,8 but even with a poor data set this can be misleading. For example, in the current data set data within ±0.2V gives a well-defined J-V trace erroneously suggesting that this data set is reliable. Using the standard deviation (slog) from a single Gaussian fit at ±0.2V J-V trace, also erroneously suggests that the data has a narrow distribution (Figure 2cii). These limitations, therefore, highlight the limitations


8

Page 9 of 23

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


to adopting distributions and/or data at a single bias as a representation of the whole and, we infer, that a method that represents all data points without fits or pre-selection is desired.


9


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 10 of 23

Figure 2: Discrete data from unhealthy measurement of tunneling junctions. a) Histogram diagram of J-V in terms of counts. b) Corresponding heat map diagram of tunneling current vs applied voltage. c) The Gaussian fitting curves at each voltage step. d) The J-V curve by plotting Gaussian mean current density value with applied voltage


10

Page 11 of 23

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


Understanding the nature of contact in EGaIn-based junctions: The convention in understanding conical tip based junctions is that a physical contact is established when the reflection of the shaped liquid metal merges with its own reflection on a metal surface. Electrical contact is confirmed by running a single J-V trace and the stability of the junction ascertained by running at least 20 traces. The success of the visual method of establishing contact between a hundreds of microns wide electrode and often a nanometer thick monolayer is surprising, but also a potential major source of variances between users or between different laboratories. The challenge in learning what a reliable contact is has led to development of other methods of introducing the top-electrode that is independent of the user.2 Despite this development the conical tip is rapid, easy to use and does not require any pre-fabrication or advanced fabrication techniques hence the continued interest in the method.6, 27, 29, 34 Based on the dependence of the current density on distance, thick- and thinarea defects can be observed from the direction of the skew in the data and from the associated outliers, in which case the accumulation of outliers on one side of the dominant distribution would indicate either thin- or thick-area defects. In cases where the substrate and SAMs are of high quality, the thin- and thick-area defects would be associated with ‘hard’ and ‘soft’ contacts respectively. We define hard contacts as junctions formed when there is significant pressure applied on the top-electrode leading to deformation of the SAM (we infer the shear modulus of EGaIn, ≈0.5 N/m,36-37 is larger than that of a liquid-like SAM). Figure 3a shows data derived from a hard contact in which there are significant shorts and a skew of the data towards high values of J. A simple 3D plot (Figure 3d) or the conventional J-V trace (Figure 3g) do not give valuable information on the quality of the junction. Most current density data collected from a soft contact (Log |J| (A/cm2)|) is almost 3 orders of magnitude lower than those from a hard contact. A generated Gaussian mean J-V curve from hard contact shows a reversed configuration than


11


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 12 of 23

observed in most dielectric devices (Figure 3g). The abnormal current density peak at -0.1 V is caused by calculating the Gaussian mean with the inclusion of the asymmetrically distributed and abnormally large shorting current in the device. Figure 3b shows the analogous case when the electrode barely merges with its reflection on the surface (soft contact). A skew towards lower currents, and a large number of low-current outliers, is observed. For a control experiment, we collected data from C19 thiol, a molecule that is known to back-fold4 an therefore a source of thin area defects. As expected, the data was bimodal with a large distribution at higher currents. Although the fully extended chains dominated the current, a small but significant amount of current was observed through the thin-area regions. Unlike in the hard contacts, we did not observe a large number of shorts junctions (log|J| ≥ 2). The ‘skyscraper’ 3D histograms (Figure 3e-f) and J-V curves (Figure 3h-i), as expected, failed to illustrate the quality of the junctions but the heat-maps clearly showed the variances and as such can be used to inform the quality of the experiments.

Figure 3: Typical defective junctions diagnosed by heat map, 3D histogram and Gaussian value curves. a-c) The heatmaps of current density across molecular junctions with hard contact, soft contact and back-folding molecules, respectively. d-f) The corresponding 3D histogram associated with to the heat-maps. g-i) The plots of current density, calculated from Gaussian average, against voltage.


12

Page 13 of 23

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


Besides revealing the quality of the contacts through capturing thin-area and thick-area defects, the deployment of heat-maps as a means to represent charge transport data enables visualization of a large volume of data, especially from a homologous series. Figure 4a-b show the tunneling current though a SC10 SAM junction in heat-map format and the corresponding 3D histogram. The heat-map clearly shows median log|J| at each bias, and where a purely stochastic process occurs, the mean and median overlap or are very close.13 A similar inference can be draw by looking at the Gaussian fits to log|J| at each bias (Figure 4c). For single molecule measurements, not only is the nature of the contact revealed, but any voltage dependent fluctuations will also be revealed – in this data set, there are no statistically significant voltage dependent fluctuations.

Figure 4: Single-molecule and multi-molecule analysis of tunneling current through junctions using versatile 3Dviewing tool. a-c) analysis of tunneling current across decanethiolate (C10) molecular junction, including heat-map, 3D histogram and Gaussian curves. d-f) analysis of all the tunneling current data obtained for molecular junctions from nanonethiolates to hexadecanethiolates (C9-C16), including heat-map, histogram and Gaussian curves.


13


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 14 of 23

In the case of a many molecules study, data from all the molecules can be presented in a single heat-map to inform homogeneity in the data set and/or pin-point variations that might bias any statistical comparisons between molecules. Figure 4d-f summarizes a multi-molecule data set containing 147,000 data points, a statistically relevant volume. Figure 4d, shows all the raw data (C9-C16) represented in a heat map, allowing for identification of regions with low or high data points – hence likely to dominate volume-dependent comparative parameters, for example calculation of mean. Unlike the commonly used stacked histograms at a single bias (Figure 4e), the heat-maps allows for comparison across all biases while also revealing differences that could be attributed to differences in data volume. In the heat-map, a well distributed data should having equal distribution of data across the biases would result in a Gaussian distribution of the current densities allowing for the data set to be treated as one. This type of distribution can be revealed at every bias using a heat-map (Figure 4d). Conclusion: We demonstrate that the adoption of a simple data presentation can reveal otherwise concealed information in charge transport measurements. By converting an otherwise uninformative ‘skyscraper’ histogram into a heat-map, thin- and thick-area defects in molecular junctions can be revealed allowing for the nature of contact between the SAM and top-electrode to be evaluated. To support that soft or hard junctions have been formed, asymmetry in the distribution of the current density –skewness toward low (soft contact) or high (hard contact) values of current densities further support the inference. For good contacts, the distribution in current densities is symmetric. Besides unveiling the SAM-electrode contact, other thick area defects like molecular back-folding have also been demonstrate. These heat-maps can be used with single or multi-molecule data sets. Experimental


14

Page 15 of 23

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


The experimental data was gathered using the EGaIn technique, as previously reported.4, 6, 26, 29 The plots were created using an in-house developed Matlab program, large area junction analysis (LAJA). Data was sorted into a hierarchal folder setup. A dataset contains molecules that contain samples that contain junctions that contain traces, enabling LAJA to analyze the data at different experimental levels by separating out unique samples, junctions, or molecules. The program allows users to specify relevant variables like voltage ranges, plots to produce, and current density ranges. LAJA can generate all of the data analysis figures shown in this work as well as a 3d scatter plot of the histogram, a table of the statistical analysis, and a 3d scatter plot of every individual trace divided up by molecule. The program uses a window-based system where each command, specification, and file is accessed sequentially by its own window, taking the user through the analysis. The MATLAB® code is available at; www.mse.iastate.edu/thuo/LAJA. An introductory video tutorial is also available online at the same website. All SAMs and charge transport measurements were performed as previously reported with modifications described in the text of this publication. ASSOCIATED CONTENT ACKNOWLEDGMENT This work was supported by Iowa State University through startup funds and through a Black and Veatch faculty fellowship to MT. MT and JC acknowledge support from Engineering Research Institute through a Catron Fellowship. Supporting Information. A video illustrating the real-time transformation of the ‘skyscraper’ histogram into a heat-map is included while the matlab code is freely available as described in the experimental section. This material is available free of charge via the Internet at http://pubs.acs.org.


15


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 16 of 23

AUTHOR INFORMATION Corresponding Author *Email: [email protected] Notes The authors declare no competing financial interests. REFERENCES 1. Jiang, L.; Sangeeth, C. S. S.; Nijhuis, C. A. The Origin of the Odd-Even Effect in the Tunneling Rates across EGaIn Junctions with Self-Assembled Monolayers (SAMs) of nAlkanethiolates. J. Am. Chem. Soc. 2015, 137, 10659-10667. 2. Jiang, L.; Sangeeth, C. S. S.; Wan, A.; Vilan, A.; Nijhuis, C. A. Defect Scaling with Contact Area in EGaIn-Based Junctions: Impact on Quality, Joule Heating, and Apparent Injection Current. J. Phys. Chem. C 2015, 119, 960-969. 3. Reus, W. F.; Thuo, M. M.; Shapiro, N. D.; Nijhuis, C. A.; Whitesides, G. M. The SAM, not the Electrodes, Dominates Charge Transport in Metal-Monolayer//Ga2O3/Gallium-Indium Eutectic Junctions. ACS Nano 2012, 6, 4806-4822. 4. Thuo, M. M.; Reus, W. F.; Nijhuis, C. A.; Barber, J. R.; Kim, C.; Schulz, M. D.; Whitesides, G. M. Odd-Even Effects in Charge Transport across Self-Assembled Monolayers. J. Am. Chem. Soc. 2011, 133, 2962-2975. 5. Yuan, L.; Breuer, R.; Jiang, L.; Schmittel, M.; Nijhuis, C. A. A Molecular Diode with a Statistically Robust Rectification Ratio of Three Orders of Magnitude. Nano Lett. 2015, 15, 5506-5512. 6. Barber, J. R.; Yoon, H. J.; Bowers, C. M.; Thuo, M. M.; Breiten, B.; Gooding, D. M.; Whitesides, G. M. Influence of Environment on the Measurement of Rates of Charge Transport across AgTS/SAM//Ga2O3/EGaIn Junctions. Chem. Mater. 2014, 26, 3938-3947. 7. Chiechi, R. C.; Weiss, E. A.; Dickey, M. D.; Whitesides, G. M. Eutectic Gallium-Indium (EGaIn): a Moldable Liquid Metal for Electrical Characterization of Self-Assembled Monolayers. Angew. Chem., Int. Ed. 2008, 47, 142-144. 8. Hylke B. Akkerman, B. d. B. Electrical Conduction through Single Molecules and SelfAssembled Monolayers. Journal of Physics: Condensed Matter 2008, 20, 013001. 9. Jeong, H.; Kim, D.; Kim, P.; Cho, M. R.; Hwang, W.-T.; Jang, Y.; Cho, K.; Min, M.; Xiang, D.; Park, Y. D.; Jeong, H.; Lee, T. A new Approach for High-Yield Metal-MoleculeMetal Junctions by Direct Metal Transfer Method. Nanotechnology 2015, 26, 25601. 10. Park, S.; Wang, G.; Cho, B.; Kim, Y.; Song, S.; Ji, Y.; Yoon, M.-H.; Lee, T. Flexible Molecular-Scale Electronic Devices. Nat. Nanotechnol. 2012, 7, 438-442. 11. Wang, G.; Kim, T.-W.; Lee, T. Electrical Transport Characteristics through Molecular Layers. Journal of Materials Chemistry 2011, 21, 18117-18136. 12. Wang, G.; Kim, Y.; Choe, M.; Kim, T.-W.; Lee, T. A New Approach for Molecular Electronic Junctions with a Multilayer Graphene Electrode. Adv. Mater. 2011, 23, 755-760.


16

Page 17 of 23

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


13. Reus, W. F.; Nijhuis, C. A.; Barber, J. R.; Thuo, M. M.; Tricard, S.; Whitesides, G. M. Statistical Tools for Analyzing Measurements of Charge Transport. J. Phys. Chem. C 2012, 116, 6714-6733. 14. Xiang, D.; Jeong, H.; Lee, T.; Mayer, D. Molecular Electronics: Mechanically Controllable Break Junctions for Molecular Electronics Adv. Mater. 2013, 25, 4818. 15. Ratner, M. A Brief History of Molecular Electronics. Nat. Nanotechnol. 2013, 8, 378381. 16. Tsutsui, M.; Taniguchi, M. Single Molecule Electronics and Devices. Sensors 2012, 12, 7259-7298. 17. Kiguchi, M.; Nakashima, S.; Tada, T.; Watanabe, S.; Tsuda, S.; Tsuji, Y.; Terao, J. Single-Molecule Conductance of π-Conjugated Rotaxane: New Method for Measuring Stipulated Electric Conductance of π-Conjugated Molecular Wire Using STM Break Junction. Small 2012, 8, 726-730. 18. Reuter, M. G.; Hersam, M. C.; Seideman, T.; Ratner, M. A. Signatures of Cooperative Effects and Transport Mechanisms in Conductance Histograms. Nano Lett. 2012, 12, 2243-2248. 19. Guo, S.; Hihath, J.; Díez-Pérez, I.; Tao, N. Measurement and Statistical Analysis of Single-Molecule Current–Voltage Characteristics, Transition Voltage Spectroscopy, and Tunneling Barrier Height. J. Am. Chem. Soc. 2011, 133, 19189-19197. 20. Smaali, K.; Clément, N.; Patriarche, G.; Vuillaume, D. Conductance Statistics from a Large Array of Sub-10 nm Molecular Junctions. ACS Nano 2012, 6, 4639-4647. 21. Nijhuis, C. A.; Reus, W. F.; Whitesides, G. M. Molecular Rectification in Metal-SAMMetal Oxide-Metal Junctions. J. Am. Chem. Soc. 2009, 131, 17814-17827. 22. Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. SelfAssembled Monolayers of Thiolates on Metals as a Form of Nanotechnology. Chem. Rev. 2005, 105, 1103-1169. 23. Wimbush, K. S. Supramolecular Tunneling Junctions. Ph.D Thesis, University of Twente, The Netherlands, 2012. 24. Mann, B.; Kuhn, H. Tunneling through Fatty Acid Salt Monolayers. J. Appl. Phys. 1971, 42, 4398-405. 25. Nerngchamnong, N.; Yuan, L.; Qi, D.-C.; Li, J.; Thompson, D.; Nijhuis, C. A. Role of Van der Waals Forces in Performance of Molecular Diodes. Nat. Nanotechnol. 2013, 8, 113-118. 26. Thuo, M. M.; Reus, W. F.; Simeone, F. C.; Kim, C.; Schulz, M. D.; Yoon, H. J.; Whitesides, G. M. Replacing -CH2CH2- with -CONH- does not Significantly Change Rates of Charge Transport through AgTS-SAM//Ga2O3/EGaIn Junctions. J. Am. Chem. Soc. 2012, 134, 10876-10884. 27. Cademartiri, L.; Thuo, M. M.; Nijhuis, C. A.; Reus, W. F.; Tricard, S.; Barber, J. R.; Sodhi, R. N. S.; Brodersen, P.; Kim, C.; Chiechi, R. C.; Whitesides, G. M. Electrical Resistance of AgTS-S(CH2)n-1CH3//Ga2O3/EGaIn Tunneling Junctions. J. Phys. Chem. C 2012, 116, 1084810860. 28. Wang, G.; Jo, G.; Kim, Y.; Lee, T. Effect of Molecular Tilt Configuration on Molecular Electronic Conduction. AIP Conference Proceedings 2011, 1399 (Pt. B, Physics of Semiconductors), 913-914. 29. Baghbanzadeh, M.; Simeone, F. C.; Bowers, C. M.; Liao, K.-C.; Thuo, M.; Baghbanzadeh, M.; Miller, M. S.; Carmichael, T. B.; Whitesides, G. M. Odd-Even Effects in Charge Transport across n-Alkanethiolate-Based SAMs. J. Am. Chem. Soc. 2014, 136, 1691916925.


17


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 18 of 23

30. Chen, J.; Wang, Z.; Oyola-Reynoso, S.; Gathiaka, S. M.; Thuo, M. Limits to the Effect of Substrate Roughness or Smoothness on the Odd-Even Effect in Wetting Properties of nAlkanethiolate Monolayers. Langmuir 2015, 31, 7047-7054. 31. Newcomb, L. B.; Tevis, I. D.; Atkinson, M. B. J.; Gathiaka, S. M.; Luna, R. E.; Thuo, M. Odd-Even Effect in the Hydrophobicity of n-Alkanethiolate Self-Assembled Monolayers Depends upon the Roughness of the Substrate and the Orientation of the Terminal Moiety. Langmuir 2014, 30, 11985-11992. 32. Nurbawono, A.; Liu, S.; Nijhuis, C. A.; Zhang, C. Odd-Even Effects in Charge Transport through Self-Assembled Monolayer of Alkanethiolates. J. Phys. Chem. C 2015, 119, 5657-5662. 33. Yuan, L.; Thompson, D.; Cao, L.; Nerngchangnong, N.; Nijhuis, C. A. One Carbon Matters: The Origin and Reversal of Odd-Even Effects in Molecular Diodes with SelfAssembled Monolayers of Ferrocenyl-Alkanethiolates. J. Phys. Chem. C 2015, 119, 1791017919. 34. Simeone, F. C.; Yoon, H. J.; Thuo, M. M.; Barber, J. R.; Smith, B.; Whitesides, G. M. Defining the Value of Injection Current and Effective Electrical Contact Area for EGaIn-Based Molecular Tunneling Junctions. J. Am. Chem. Soc. 2013, 135, 18131-18144. 35. Yoon, H. J.; Shapiro, N. D.; Park, K. M.; Thuo, M. M.; Soh, S.; Whitesides, G. M. The Rate of Charge Tunneling through Self-Assembled Monolayers is Insensitive to Many Functional Group Substitutions. Angew. Chem., Int. Ed. 2012, 51, 4658-4661, S4658/1S4658/23. 36. Dickey, M. D.; Chiechi, R. C.; Larsen, R. J.; Weiss, E. A.; Weitz, D. A.; Whitesides, G. M. Eutectic Gallium-indium (EGaIn): A Liquid Metal Alloy for the Formation of Stable Structures in Microchannels at Room Temperature. Advanced Functional Materials 2008, 18, 1097-1104. 37. Larsen, R. J.; Dickey, M. D.; Whitesides, G. M.; Weitz, D. A. Viscoelastic properties of Oxide-Coated Liquid Metals. Journal of Rheology 2009, 53, 1305-1326.


18

Page 19 of 23

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


TOC graphic 217x158mm (150 x 150 DPI)



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: Schematics of molecular tunneling junction and diagrams of analysis of tunneling currents thought the junctions. a) Schematics of performing current measurement using EGaIn tip. b) The formation of MSAMs//Ga2O3/EGaIn junction and interface 1 and 2. c) A schematic representation of defects in metal substrate and SAMs. d-f) Common approaches to represent charge-transport data derived from large area junctions, with each being limited to two variables at a time. g) a 3D plot of I-V data from EGaIn-based junctions illustrating that a simple histogram plot is not very informative or easy to interpret. 403x502mm (150 x 150 DPI)


Page 20 of 23

Page 21 of 23

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 2: Discrete data from unhealthy measurement of tunneling junctions. a) Histogram diagram of J-V in terms of counts. b) Corresponding heat map diagram of tunneling current vs applied voltage. c) The Gaussian fitting curves at each voltage step. d) The J-V curve by plotting Gaussian mean current density value with applied voltage 185x331mm (150 x 150 DPI)



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: Typical defective junctions diagnosed by heat map, 3D histogram and Gaussian value curves. a-c) The heat-maps of current density across molecular junctions with hard contact, soft contact and backfolding molecules, respectively. d-f) The corresponding 3D histogram associated with to the heat-maps. g-i) The plots of current density, calculated from Gaussian average, against voltage. 374x177mm (150 x 150 DPI)


Page 22 of 23

Page 23 of 23

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: Single-molecule and multi-molecule analysis of tunneling current through junctions using versatile 3D-viewing tool. a-c) analysis of tunneling current across decanethiolate (C10) molecular junction, including heat-map, 3D histogram and Gaussian curves. d-f) analysis of all the tunneling current data obtained for molecular junctions from nanonethiolates to hexadecanethiolates (C9-C16), including heat-map, histogram and Gaussian curves. 335x256mm (150 x 150 DPI)


Revealing the Nature of MoleculeâElectrode Contact in Tunneling

Recommend Documents

Revealing the Nature of MoleculeâElectrode Contact in Tunneling