Electrochemical Transmission Electron Microscopy (EC-TEM) of Capacitance-Induced Electrodeposition and Coarsening on Graphene

Heterostructures that combine metal nanoparticles and two-dimensional (2D) materials fabricated by electrochemical deposition offer the opportunity to enhance performance and develop unique material properties for technological applications ranging from catalysis [1,2] to nanoelectronic devices [3].Continued improvements in the precision and application of electrochemical fabrication require a deep understanding of the interplay of kinetic and thermodynamic phenomena that operate during the process.Here, we explore the deposition of copper (Cu) metal on a graphene electrode using liquid cell transmission electron microscopy (TEM), using the temporal and spatial resolution of this technique to explore the nucleation and growth of nanocrystals.The technique reveals an unexpected transient phenomenon under certain applied voltage conditions, and we discuss possible mechanisms.We transferred graphene to be used as a working electrode onto liquid cell chips patterned with multiple platinum (Pt) electrodes, using Pt as reference and counter electrodes.We then performed electrochemical deposition of Cu on graphene by controlling the deposition potential in an acidified copper sulfate solution.In situ movies reveal that under applied potential, the formation of Cu nanocrystals follows the expected electrodeposition model of progressive nucleation and diffusion-limited growth: the density of nanocrystals increases, and their size at the end of the voltage pulse decreases with an increase in the applied overpotential [4].Typical examples are shown in Fig. 1 for potentials in the range of -40 to -100 mV.When the applied potential is low, particles immediately dissolve as the potential is turned off.However, we observe that when a higher potential is applied during the deposition, the growth of Cu nanocrystals continues after the voltage goes off.The total amount of Cu deposited can greatly exceed the amount deposited in the 10s voltage pulse.This mode of deposition of Cu on graphene is shown in Fig. 2, with images extracted from a movie recorded during and after pulse voltammetry at -260 mV for 10 s in a liquid cell filled with 0.1 M CuSO 4 + 0.1 M H 2 SO 4 electrolyte.In the first row of Fig. 2 Cu islands nucleate and grow on the graphene electrode during the 'V on' stage, consistent with results from Fig. 1.However, the later movie frames in Fig. 2 show continued growth of Cu islands even after the voltage is turned off at t=10 s, followed by eventual disappearance of all islands after tens of seconds.Tracking the size of the individual particles shows that smaller particles disappear earlier while the largest particles grow to quite large sizes before themselves eventually disappearing.The decrease in the number of islands with time is particularly apparent at higher applied potentials and is shown in Fig. 3.The behavior is consistent with Ostwald ripening, and the size threshold that separates crystals that grow or shrink appears consistent with Ostwald ripening models [5].The eventual dissolution of Cu nanocrystals in all experiments would not ordinarily be expected at zero volts applied potential.However, the comparison of cyclic voltammetry curves among different configurations of graphene electrodes, where 1, 2 or all 3 of the Pt electrodes are covered with separate graphene flakes, suggests that there is an intrinsic potential between the graphene and Pt electrodes [6].We conclude that this built-in potential is high enough to dissolve the deposited Cu even when the applied potential is 0 mV.This would be in effect when the potential applied for growth is small (-40 mV, -60 mV, -80 mV, and -100 mV) and the nanocrystals dissolve immediately after the applied potential is turned off, as well as when the applied potential is large (-180 mV, -220 mV, -240 mV and -260 mV), when continued growth takes place, but nanocrystals eventually dissolve, requiring up to 2 minutes as shown in Figure 3.We consider several mechanisms that may be responsible for the unexpected transient deposition.One possible explanation for the additional growth could be that Cu is indeed deposited during the potential pulse but does not form into nanocrystals until later.This could arise if Cu is intercalated in the graphene in a uniform manner during the applied potential, with coalescence into islands occurring later [7].However, the measured charge that flows during the potential pulse is not strongly dependent on voltage, in particular not varying enough to account for the large deposited volume at high potentials.A second possible explanation may be attributed to the capacitance of graphene.Graphene has high capacitance based on electrochemical double-layer capacitance (EDLC) measurements, and it has been shown that the electrochemical interfacial capacitance increases for larger (negative) applied potential [8].The plots in Fig. 3 show that the islands grown at negatively higher potential last longer, consistent with additional growth of Cu islands during the pulse-off stage being due to charge accumulation from the capacitance of the graphene when the applied voltage is on.A final mechanism for the effect could be that the application of low cathodic potential may cause the reduction of various oxygen functional groups on the graphene surface, resulting in an increase of -OH groups on the surface [9].These charged species could sustain additional growth by causing a reduction of Cu ions in the solution.Further experiments will be required to distinguish between these possibilities.