Reports: ND1050966-ND10: Sub-Diffraction Imaging of Energy Conversion Reactions on Graphene

Peng Chen, Cornell University

Developing convenient and efficient methods to study the electrocatalytic activity distribution within graphene and the related graphene oxide (GO) sheets is important for devising strategies to tailor their performances in electro-conversion reactions. We have found that their electrocatalytic activity in the reactions of our interest likely stems from the oxygen containing groups and thus moved our focus to GO sheets instead of graphene sheets. We used a fluorogenic electrocatalytic reaction to probe the activity of individual GO sheets supported on transparent conducting ITO electrodes. By imaging the fluorescence, thus the electrocatalytic activity, using fluorescence microscopy, we were able to map out the activity distribution on GO sheets. This fluorescence microscopy approach enabled us to image GO sheets and resolve their fine features, including boundaries, wrinkles and folded regions. By further manipulating their activity with prior electrochemical reduction, we could also differentiate domains within individual GO sheets. The division lines between neighboring domains were mapped out as well, and they possibly reflect the grain boundaries. These discoveries were summarized in our 2012 report. Below we summarize our progresses in the last year:

1) The detected fluorescent signal in our imaging is from the products of electrocatalytic reactions on GO sheets. Thus its first time derivative reflects the current passing through the sheet. We extracted the fluorescence-time trajectories for all regions, sub-regions and domains. The applied potential was swiping cyclically. By taking the first time derivatives of these intensity trajectories and converting time into voltage, we generated the fluorescence based cyclic voltammetry with spatial resolving power. The peak voltages were examined as well as their corresponding currents for different regions. It turned out that the wrinkled regions required the least of driving force yet still showed highest rates in both reduction and oxidation reactions.   

2) In our previous studies, we observed the correlation between sheet activity and its oxygen containing functionalities. Here we applied a constant negative voltage to GO sheets to electrochemically reduce the number of oxygen containing functionalities. This electrochemical reduction was followed by our fluorescence based cyclic voltammetry to image the activity changes. These electrochemical reductions were performed at three different voltages from -0.7 V to -0.9 V (vs. Ag/AgCl) for 1 hour. The effect of electrochemical reduction was significant. For all three voltages we observed increasing driving forces needed in both reduction and oxidation reactions. The increments could be as large as 50 mV for some sheets. The general trend from averaged GOs showed that as the reduction potential went more negative, larger increments in driving force were needed. The maximal fluorescent intensity obtained from GO reflects the highest fluorescent product molecules generated on the GO sheets, and the highest peak in its first time derivative reflects the fastest rate of product generation. These two quantities are important in terms of depicting the activity of GOs. In partially reduced GO, the maximal intensity dropped, while the maximal rate increased. Our hypothesis here is that the maximal intensity is proportional to the local area density of active sites on each sheet; while the maximal rate is proportional to both the activity of a single site and the active site density. The applied reduction potential has two-fold influences: At one hand, it reduces the area density of active sites, which causes the decrease in maximal surface fluorescent intensity; on the other hand, it restores the backbones of graphene network, which facilitates electron transfer, and thus activity, for each site.

3) To observe the changes of electrocatalytic activity more closely, we performed multistage reduction treatments of the GO sheets. By applying the reduction potential for 20 minutes each time, and measuring the maximal fluorescence intensity after each reduction, we were able to monitor the gradual change of sheet activity. The fluorescence intensity at each image pixel followed an exponential decay behavior in multistage reduction. By fitting the decay and extracting the apparent rate, we were able to reconstruct a decay map for an entire sheet.  It turned out that there exist discrete sites where the decay rates were much faster than the rest of the sheet. And by examining the spatial locations of these discrete sites, we found that electrochemical reduction preferentially started from boundaries and wrinkles.

4) We also measured the Raman spectrums at various different locations on GO to better understand the structural-activity correlation. The same pieces of sheets were later put under our optical microscope to correlate with electrocatalytic activity measurements. It is reported that the increase of oxygen functionalities causes the shift of Raman G band to lower wavenumbers. In our measurements, the boundary regions redshift the most, while wrinkles and flat sheets had almost the same Raman frequencies. These results indicate that the GO boundary region is most abundant in oxygen containing functionalities, as expected.

This PRF funded research has enabled the PI to pursue a new direction, namely, optical imaging of electrocatalysis of graphene and GO sheets. And it also funded a graduate student who was able to be trained in fluorescence microscopy, and electrochemistry/catalysis. We are at the moment trying to wrap up the study for publication.