Week 5: Representing and Interpreting Data in Single-Molecule Electronics

Rohan V -

Hi all, it’s Rohan! Last week, I explained how data is commonly modeled in single-molecule experiments. This week, I’ll explain how researchers try and interpret the vast amount of data that comes out of these experiments. As a refresher, because of the small scale of individual molecules, thousands of individual ‘experiments’ need to be run on any given molecule to ensure consistency of results. Any individual experiment might be affected by a slight variation in ambient temperature, a change in machine voltage, etc; this inconsistency is counteracted by performing the experiment many times.

Due to the size of the experimental output, it’s inefficient for researchers to look at thousands of graphs of how the junction conductance evolves as it is stretched apart. To more quickly and accurately analyze their data, researchers often represent trace data in a two-dimensional (2-D) histogram. These histograms aggregate the distance-conductance data from all of the thousands of traces in the experimental dataset, grouping them together into a single graph.

Above is a 2-D histogram generated from the experiment we ran this past week! Here, a deeper shade of blue indicates that more traces pass through that specific point in distance-conductance space. Recalling how a general trace from this molecule looks, we can see that the 2-D histogram broadly represents an initial exponential conductance decay, followed by a ‘molecular plateau’ where the molecule is bound to gold electrodes.

However, the 2-D histogram is much broader than the individual trace. This phenomenon arises because traces collected from the same molecule can vary greatly in behavior, as explained above. The idea of representing single-molecule data with 2-D histograms has become a staple feature of molecular electronics research. However, the way that researchers extract information about molecular conductance from these outputs is still in development. Oftentimes, the ‘conductance’ of a molecule is defined to be the conductance value that the most traces pass through (represented as the darkest shade on the 2-D histogram above). This method of extracting conductance, while effective for a general understanding of molecular behavior, fails to account for the variable conductance behavior of many traces. In fact, in the single trace depicted below, conductance varies by a significant amount even within the molecular plateau.

My project involves developing a more accurate and efficient way to model single-molecule conductance. Over the last few weeks, I’ve spent a significant amount of my time working with MATLAB code, using existing documentation from my lab as well as programs developed by other researchers, to achieve this goal. In the coming weeks, I’ll dive into more specifics: explaining existing algorithms to model single-molecule conductance, their limitations, and the approach I’ve been using to create an improved program that can be widely applied to many molecules.

More Posts

Comments:

All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.

    camille_bennett
    Hi Rohan, looks like fascinating work. You mentioned using a 2-D histogram to make sense of the data. Can you explain what a histogram is and why it’s helpful in organizing the data from experiments?
    rohan_va
    Hi Ms. Bennett! A histogram is a graph which displays the number of times a variable hits a specific value. Oftentimes, histograms are made for only one variable. In that case, the range of the values the variable can take on is on the x-axis, and the number of times the variable hits a specific value is on the y-axis. Histograms are helpful when researchers want to see the distribution of data, in other words, if they're interested in how many data points equal a certain value. In single-molecule electronics, the main variable of interest is the conductance; so researchers histogram conductance data to see what the most frequently hit conductance is. In the case of single-molecule experiments, we have two variables of interest: distance as well as conductance. So, we make a 2-D histogram, where we use both axes for variables (distance and conductance), but add a third metric - color scale - to indicate how many of our data points are that specific distance-conductance value pair. As a summary, histograms depict distributions of data, and they are useful for single-molecule researchers to see what conductance and distance values are most frequently 'hit' as we pull apart the single-molecule junction. I hope this helped!

Leave a Reply

Your email address will not be published. Required fields are marked *