Control of Growth Front Evolution by Bi Additives during ZnAu

Jan 8, 2018 - Finally, the effects of Bi in suppressing bubble formation are also seen in this data (Figure 1D, Figure S4), consistent with previous s...
0 downloads 6 Views 2MB Size
Subscriber access provided by READING UNIV

Communication

Control of Growth Front Evolution by Bi additives during ZnAu Electrodeposition Jeung Hun Park, Nicholas M. Schneider, Daniel A. Steingart, Hariklia Deligianni, Suneel Kodambaka, and Frances M Ross Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.7b04640 • Publication Date (Web): 08 Jan 2018 Downloaded from http://pubs.acs.org on January 8, 2018

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.

Nano 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 12 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

Nano Letters

Control of Growth Front Evolution by Bi additives during ZnAu Electrodeposition Jeung Hun Park,†,‡,§ Nicholas M. Schneider,∥ Daniel A. Steingart,‡,* Hariklia Deligianni,† Suneel Kodambaka,§,* and Frances M. Ross†,* †

IBM T. J. Watson Research Center, 1101 Kitchawan Road, Yorktown Heights, New York 10598, United States Department of Mechanical and Aerospace Engineering, and Andlinger Center for Energy and the Environment, Princeton University, 86 Olden Street, Princeton, New Jersey 08544, United States § Department of Materials Science and Engineering, University of California Los Angeles, 410 Westwood Plaza, Los Angeles, California 90095, United States ∥Department of Mechanical Engineering and Applied Mechanics, University of Pennsylvania, 220 South 33rd Street, Philadelphia, Pennsylvania 19147, United States ‡

ABSTRACT: The performance of many electrochemical energy storage systems can be compromised by the formation of metal dendrites during charging. Additives in the electrolyte represent a useful strategy to mitigate dendrite formation, but understanding the mechanisms involved requires knowledge of the nanoscale effects of additives during electrochemical deposition. Here we quantify the effects of an inorganic additive on the morphology of an evolving electrochemical growth front, using liquid cell electron microscopy to provide the necessary spatial and temporal resolution. We examine deposition of ZnAu on Au in the presence of Bi additive, and show that low concentrations of Bi delay but do not prevent the formation of growth front instabilities. We describe a model in which Bi segregates at the growth front and promotes the surface diffusion and relaxation of Zn, allowing better coverage of the initial Au electrode surface. A more precise knowledge of the mechanism of inorganic additive effects may help in designing electrolyte chemistry for battery and other applications where morphology control is essential.

KEYWORDS: dendrites, additives, energy storage, electrochemistry, liquid cell electron microscopy, in situ electron microscopy

1

ACS Paragon Plus Environment

Nano Letters 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

Zn-based flow batteries promise an environmentally friendly and low cost solution for grid-scale energy storage.1 However, the formation of metal dendrites during charging is one factor reducing the performance of these and other secondary batteries.1,2 Strategies to suppress these growth instabilities include electrolyte flow3 and modulation of the charging current4 or potential,5 but the simplest approach may be the use of additives in the electrolyte.6 Several organic additives are known to inhibit dendrite formation to a greater or lesser extent during Zn deposition.1,7,8,9 However, extensive potential cycling results in consumption of the organic additives and reduced anode capacity and durability that still limit performance.1,9 Inorganic additives can also influence electrodeposition kinetics so as to suppress dendrite formation,1,10 and can be more stable than organic additives9,11 over time. Inorganic additives are therefore widely employed to suppress H2 generation, reduce dendrite formation, and improve discharge and cycling reversibility.1,10,12 However, the mechanism of their interactions is not well documented,10 partly because it is difficult to measure additive effects on the spatially and temporally developing growth front with sufficient resolution. A better understanding of additive effects on growth front morphology is a key goal that may help in developing new strategies for additive design to optimize cycling performance. In order to quantify the evolution of the electrochemical growth front with spatial and temporal resolution, we have therefore used liquid cell transmission electron microscopy (TEM) to examine additive effects on a Zn-based system. In this technique, an electrochemical process takes place within a thin electrolyte layer held within a sample designed to allow observation in the TEM.13 The evolving morphology can be tracked with simultaneous recording of electrochemical parameters. Liquid cell TEM studies have explored several aspects of battery operation including the formation of the solid-electrolyte interphase14 and growth front morphology during potential cycling.15,16,17,18 Dendrite formation and its suppression with organic additives has been studied ex situ for alkaline Zn-based systems.19 However, the role of inorganic additives in acidic Zn-based systems has not been addressed in liquid cell TEM. We thus use the technique to explore the mechanism by which a commonly used additive, Bi,1,10,12,20,21 modifies growth front evolution in a Zn-based electrolyte. We quantify the role of the additive by tracking the morphological evolution of the material, a ZnAu alloy, that deposits onto an Au electrode from acidified zinc sulfate. We use the results to discuss a model for the effect of Bi, in which Bi both smooths the surface and inhibits the deposition of Zn on Zn, favoring deposition on the electrode between existing growth nuclei. Images of the growth front are shown in Figures 1 and 2, and Movies M1 and M2, with the simultaneously measured electrochemical parameters shown in Figure S1. The darker region at the bottom of each image is the Au electrode while the brighter region is filled with electrolyte. These data were obtained in a liquid cell equipped with parallel electrodes composed of electron-transparent, 30 nm-thick polycrystalline Au (Figure S2). One electrode serves as the working electrode (WE) and another as a combined counter/reference electrode (C/RE). The electrolyte was 0.5 M H2SO4 + 0.1 M ZnSO4 (pH = 0.4) with or without 0.25 mM Bi2O3. Deposition was carried out under controlled current without deaerating or flowing the electrolyte. To extend the lifetime of the liquid cell by minimizing dissolution of the CE22 and generation of hydrogen bubbles,23 the deposition current was applied in 10 s pulses at typically -20 nA, followed by a 10 s rest time at zero current.4,12,24 The deposition conditions and reproducibility are discussed further in the Supporting Information. Images were recorded under bright field conditions using a 300keV TEM (FEI model CM30). Low-dose imaging conditions (electron flux ~50 e- / Å2·s) were used to minimize electron beam effects in the experiments while still providing sufficient resolution to track the growth front. Images were recorded at 30 frames per second. Furthermore, after every 5 cycles (total time = 100 s) the current was paused, and the entire electrode edge was recorded as a montage of TEM images, a procedure taking several minutes. This imaging approach allowed the growth front progression to be recorded with reasonable time resolution in one location, but also over a large area every 100 s to evaluate the effect of electrolyte thickness on growth rate and morphology. In the present experiments, the electrolyte thickness varies from 250 nm to ~1.5 µm along the length of the electrode. The minimum thickness is determined by the spacer layer thickness used in the liquid cell, while the variation in electrolyte thickness due to deflection of the windows25,26 was estimated by elastic beam theory,27 2

ACS Paragon Plus Environment

Page 2 of 12

Page 3 of 12 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

Nano Letters

assuming 1 atm. pressure difference between the cell interior and the microscope vacuum and a Young’s modulus of 300 GPa for silicon nitride. Compositional analysis of the deposited layer was carried out post-growth by X-ray energy dispersive spectroscopy (XEDS) in a scanning electron microscope (SEM) after flowing DI water through the cell then separating and drying the chips (see Supporting Information). When current first flows we expect the initial deposition to take place over the entire electrode surface. Once the electrolyte directly above the electrode is depleted of ions, this “surface growth” should transition to “lateral growth”, or a growth front that extends outwards from the edge of the electrode. The initial surface growth is visible in the data in Figure 1 if the image contrast is stretched to emphasize details on the electrode, as is shown in Figure S3. However, we analyze the lateral growth regime for a more quantitative measure of additive effects on growth morphology. The electrode edge initially appears fairly smooth. Ramified structures, which we denote generally as dendrites, form in the plain solution within the first 5 cycles and become longer with additional cycles (Figure 1AD). Similar to observations of Pb,15 Cu,28 and Ag29 growth, each dendrite grows approximately along the direction of the electric field (see Figure S2 for detailed geometry of the liquid cell). The presence of Bi causes clear changes to the growth front evolution (Figure 1F-I). Growth remains smooth during the first 5 cycles, with an irregular growth front beginning to appear after 10 cycles. Without Bi, the dendrite structures are more fern-like while Bi produces more nodular and rounded dendrite tips. A second effect of Bi is in the suppression of hydrogen. 20 cycles without Bi typically forms large bubbles (a bubble is present in Figure 1D, accounting for the improved contrast in that image), but such bubbles were not observed in experiments in which Bi was present. We quantify these observations by marking the growth front and superimposing the line profiles (Figure 1E and J). A distinctive difference between the two experiments is that while many points along the interface stop growing in the plain solution, resulting in deep valleys in the growth front, in the Bi-containing solution the growth front continues to progress at most locations, and is thus smoother overall. Based on the electrochemical data in Figure S1, the overall deposition rate in the two solutions is similar. Bi appears to have its greatest effects at early times, promoting deposition on the electrode surface. To quantify the effects of electrolyte thickness on morphology, in Figure 1E and J we divide the electrode edge into three regions of interest (ROI) labeled A-C, at which the corresponding liquid thickness are approximately 250 nm, 500 nm, and 1 µm. Note that these liquid thicknesses are larger than is often used in liquid cell TEM experiments. Thick liquid layers were chosen to provide a better model for bulk processes, even though the quality of the images obtained through such thick layers is reduced.30 In each ROI we measure the RMS roughness and the distance from the initial to the most advanced position of the interface (Figure 1K and L). In the plain solution, the overall lateral growth rate depends on electrolyte thickness: the dendrite length is approximately proportional to electrolyte thickness at each time. Bi additives reduce both average dendrite length and this variation in growth rate, supporting the idea that Bi ions promote more homogenous deposition. Bi appears to suppress electrolyte thickness effects for at least the first 20 cycles of deposition. After this time, the morphology at the growth front becomes dendritic and the dendrite length starts to show a stronger dependence on liquid thickness (see Figure S5, with data after 25 cycles and t = 500 s); in other words, the Bi effects disappear after some time. A dependence of overall growth rate on electrolyte thickness in the plain solution is not surprising because variations in electrolyte thickness will change the supply rate of ions. The lack of such an effect in the early stage growth with Bi may reflect a larger role of surface diffusion causing more uniform growth along the electrode. Finally, the effects of Bi in suppressing bubble formation are also seen in this data (Figure 1D, S4), consistent with previous suggestions that Bi suppresses hydrogen bubble generation.31 The growth data at higher spatial and temporal resolution are shown in Figure 2 and Movie M1 and M2. The initial planar growth is highlighted in Figure S4. The lateral growth regime is characterized in terms of growth front roughness in Figure 2D and S4C (obtained using a fully automated imaging processing algorithm32). In the plain solution, the growth front roughens quickly, increasing with each cycle, and develops needle-shaped features. In the 3

ACS Paragon Plus Environment

Nano Letters 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

presence of Bi, the morphology is more globular with some degree of faceting and does not become very rough over the first 100 s, forming a compact deposit (i.e. no deep valleys). A final feature of this data is the apparently disconnected features that form above the electrochemical growth front. In the plain solution, these are needles; in the presence of Bi ions they are rounder islands. These appear more frequently in areas that are under continuous electron beam irradiation, as is required during video recording, but can be seen in areas with only occasional irradiation. Metal ions in solution are known to be reduced by beam-induced radiolysis species.13,33 Separate experiments34 using various electrolyte chemistries (such as 0.1 M HCl) and higher electron flux (>250 e- / Å2·s) suggest that Au3+ ions released from the CE during current flow (a process discussed below) can reach concentrations high enough to allow beam-induced nucleation of nanoparticles. The electrolytes used here can be irradiated for tens of minutes without nanoparticles appearing35, but nanoparticle formation during electrochemical deposition is consistent with release of Au3+ ions from the CE followed by chemical reaction with radiolytic or other species present in the electrolyte.33,34,35 The composition of the deposited material is not discernable in the TEM images but was measured post-growth by XEDS. Zn (and Bi, in the case of Bi-containing electrolyte) are detectable along the edge and over the electrode surface (Figure 3), presumably from the initial deposition. The subsequent lateral growth is an alloy of Zn and Au that is rich in Au, with 11 at% Zn in the plain solution and 23 at% Zn in the Bi-containing solution (and Bi concentration ~0.03 at% in the deposit). The XEDS penetration depth is in the micrometers range so this composition reflects the bulk of the deposit and not just surface segregation. The deposited Au originates from the CE, as discussed above in the context of nanoparticle formation. The presence of Bi appears to correlate with higher Zn composition in the deposit. The electrochemical, compositional and morphological data combined allow us to discuss the sequence of events during deposition. In the plain electrolyte, Zn and Au are co-deposited because Au is released from the CE during current flow. Although we do not have accurate reference potential measurements in these two-terminal experiments, Figure 2C shows that the voltage becomes more negative over the 10 s current pulse. Initially the imposed current drives a relatively low voltage. The first reaction at the CE is then the reduction of oxygen at approximately -0.3V, while at the WE, Zn2+ is converted to zero-valent Zn. However, the voltage does not return to its initial value even after the 10 s rest time, and continues to become more negative until a steady state is reached after several cycles (Figure S1B). As Zn2+ is depleted at the Au WE surface, the potential shifts more negative until it reaches a value that supports other reactions at the CE and WE that can supply the desired current. These reactions are Au dissolution at the CE and Au+Zn deposition at the WE. In strong acids, Au dissolution begins with oxidation to form a soluble Au(III) species or film of Au(OH)3, but Au(I) also forms at more negative potentials.36,37 The dissolution of Au is initiated above -1.35 V (Ref. 38) and occurs until -1.55 V (vs. Au plated Pt reference) in 0.5 M H2SO4 (Ref. 38 ). For the electrolyte used here, the theoretical redox potential for Au3+ is E○ = -1.51 V (Ref. 39). These reported values are consistent with our observations. The released Au ions diffuse under the electric field gradient and are co-deposited at the WE along with Zn that diffuses from more distant regions of the electrolyte. The deposit composition depends on the relative supply rates of Au and Zn. Under these conditions, the growth front eventually becomes dendritic because in electrochemical models a dendritic morphology is associated with growth that is limited by diffusion of the depositing ions. Phasefield simulations for plain Zn show that ion concentration and electrical potential at the tips of asperities are higher than those at other positions, leading to a concentration gradient and heterogeneity of local current densities at the tips40 that promote dendrite growth.40,41 The presence of Bi changes this deposition in several ways. At early times (t ~100 s), potentials are generally more negative in the presence of Bi for the same current (Figure 2C). More negative potentials in the presence of additives are attributed to the absorption of additives on the electrode and partial blockage of active sites which decreases the nucleation rate and increases overpotential, and inhibits electro-crystallization onto the adsorbed additive layer.42 Bi 4

ACS Paragon Plus Environment

Page 4 of 12

Page 5 of 12 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

Nano Letters

also smooths the surface, both in the electrodeposited material and in the isolated nanocrystals. However, the Bi effects appear to diminish with time. After ~300 s the potential is similar for the two solutions (-1.56 ± 0.05 V and 1.58 ± 0.05 V, respectively) and the growth front has become dendritic in both cases. However, deposition is more uniform along the length of the electrode and there are fewer valleys with no growth. Bi thus does not suppress roughening but appears to delay its onset. The result is a more compact and void-free deposit. This morphological change also agrees with a change in local current distribution near the tips due to additive adsorption, which increases the surface polarization and creates a more uniform current distribution.6 The details of the activity of Bi as an additive is generally understood in terms of its adsorption onto the electrode surface during electrodeposition.12,20,21,43 One can expect Bi ions to be reduced to solid Bi before Zn ions are reduced, ° ° due to its larger standard reduction potential (‫ܧ‬஻௜ = +0.317 V; ‫ܧ‬௓௡ = -0.761).44,45,46 Similar to Bi deposition in 21,43,47 alkaline solution, cyclic voltammetry shows that Bi is expected to be deposited onto Au at less negative potentials than are required for Zn deposition onto Au (Fig. S1). In the experiments shown here, as the potential becomes more negative during each deposition pulse, the Bi is therefore deposited onto Au before Zn onto Au. Surface energy considerations48,49 suggest that while the current is off, this Bi should migrate towards the electrode surface. As the asperities develop during diffusion-limited deposition, preferential accumulation of Bi ions at the locations with higher potentials (the tips) would lead to the formation of a screening layer (and hence increase in the voltage required for deposition), similar to the action of Cs or Rb additives on Li.41 The screening layer can favor Zn or Au deposition adjacent to the protrusion, rather than at the tip, and thus slow the development of the asperity. This mechanism is consistent with the growth mode observed, in which growth continues over most or all of the front rather than only at a few locations. This model differs from our understanding of levelling in other systems such as Cu with polyethylene glycol additive, which acts as a leveler because it is an ohmic block. The action of Bi instead appears more like the direct plating additives used to assist Cu deposition on Ru.50 A separate effect of Bi may also be an enhanced surface diffusion at the growth front. In particular, to account for the smoother surfaces of the disconnected nanocrystals in the Bi-containing electrolyte, we note that surface diffusion is a factor that will suppress the development of surface irregularities.51,52 This smoothing effect appears to persist for long times (~300 s) even after the initial electrode surface is covered. This is consistent with studies in alkaline43 electrolytes that also suggest that Bi at ppm levels enhances mass transport to form smoother surfaces. It is interesting to consider that the morphology of beam-induced nanocrystals, which is formed in the absence of the electrochemical driving force, may provide information that complements measurements of the morphology of the electrochemical growth front. In summary, spatially and temporally resolved images of growth front propagation allow us to evaluate the effects of an inorganic additive on growth morphology. We find that the presence of Bi has several effects on the deposition process. Bi appears to facilitate more homogenous deposition over the electrode surface, even in the presence of changes in electrolyte thickness and hence ion supply. It delays but does not prevent the growth of ramified features at the ZnAu growth front. It appears to enhance the surface diffusivity of the growth species, producing a morphology (both at the electrochemical growth front and at beam-induced, disconnected crystals) that is smoother at the local scale compared to Bi-free deposition. Bi also appears to suppress hydrogen bubble formation. We have discussed the effect of Bi in terms of a model in which Bi deposits on the growth front prior to Zn deposition, inhibiting continued addition of Zn onto asperities, enhancing surface diffusion and favoring Zn deposition onto Au. We suggest that to optimize the formation of a smooth and compact deposition morphology, it may be advantageous to combine Bi ions at ppm levels with pulse plating techniques as were used here. Indeed, pulse plating of Zn in acidic electrolytes is already known to have beneficial effects on mass transport, electrode kinetics and the nucleation of growth centers.53,54 The use of liquid cell observations to develop a model for this particular additive may be a useful strategy in understanding general inorganic additive effects on the control of unstable, dendritic growth, with applications in improving secondary battery performance.

5

ACS Paragon Plus Environment

Nano Letters 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

■ ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publication website at DOI: 10.1021/acs.nanolett. Materials and methods, discussion of experimental parameters for electrochemical deposition of ZnAu dendrites in the continuous flow cell, and details of TEM movies (PDF). Bright field TEM movies of growth interface evolution in additive-free Zn solution (Movie M1, AVI) and Bi additive containing Zn solution (Movie M2, AVI). ■ AUTHOR INFORMATION Corresponding Authors * E-mail: [email protected] * E-mail: [email protected] * E-mail: [email protected] Author contributions FMR and JHP conceived the project and designed experiments. FMR and SK supervised the project. JHP performed all experiments and analyzed the data. DAS and LD provided expertise in electrochemistry and the growth model. NMS provided the computational algorithm for growth front analysis. All authors discussed the results and contributed to the manuscript. Notes The authors declare no conflict of interest. ■ ACKNOWLEDGMENTS We gratefully acknowledge funding supports from the National Science Foundation (NSF-GOALI: DMR-1310639) and the BP Carbon Mitigation Initiative, M. C. Reuter for assistance with experimental details, and C. A. Orme and D. Yu for helpful discussions.

6

ACS Paragon Plus Environment

Page 6 of 12

Page 7 of 12 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

Nano Letters

FIGURES

Figure 1. In situ progression of the ZnAu growth front. (A-J) Time series of bright field TEM montages. (A-E) Electrolyte 0.1 M ZnSO4; (F-J) Electrolyte 0.1 M ZnSO4 + 0.25 mM Bi2O3. Images were obtained after each 5 deposition/relaxation cycles at the times indicated. Each cycle consisted of current -20 nA for 10 s and 0 nA for 10 s. The electron dose rate was ~50 e- / Å2·s. The images are ~ 10k × 1.2k pixels and cover a field of view 25 µm × 5 µm. Image D shows better contrast due to a hydrogen bubble that formed at 289 s; another formed at 469 s. (Image D is therefore after 289 s rather than 300 s.) We did not observe bubbles in the Bi containing Zn solution over the entire 500 s experiment. After 10 successive depositions the total charge passed is ~ 4.5 × 10-4 C/cm2 which should deposit ~ 31 nm on the electrode (at two e- per deposited Zn). The lateral dendrite lengths were typically 400 nm at this time. (E, J) The superimposed outlines of the growth fronts in the two solutions. Black, red, green, and blue lines show the growth front at t = 0, 100, 200, and 300 s, respectively. (K, L) Average length of maximum height of ZnAu dendrites in the two solutions. Three regions of interest, A, B, and C are marked with black square (■), red circle (●), and blue triangle (▲) symbols. Note that the disconnected islands are only loosely attached to the windows, moving and rotating toward the WE during growth.

7

ACS Paragon Plus Environment

Nano Letters 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. Images extracted from movie M1 and M2 recorded in (A) 0.1 M ZnSO4 and (B) 0.25 mM Bi2O3 + 0.1 M ZnSO4 during cycling -20 nA for 10 s followed by 0 nA for 10 s. t is the total time elapsed. (C) The applied current and measured voltage during the first 5 cycles. (D) Root-mean square surface roughness of the two growth movies M1 and M2 vs. time. The images were analyzed with a partial field of view (1.5 µm × 1 µm) after the correction of image drift and rotation. The stepwise increase in roughness for each current pulse is visible. With Bi, roughness does not increase until 200 s and then increases in a stepwise manner.

8

ACS Paragon Plus Environment

Page 8 of 12

Page 9 of 12 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

Nano Letters

Figure 3. Post-growth SEM characterization of the deposited material after 25 cycles with boxes showing the area used for chemical analysis. (A) Plain solution; (B) Bi-containing solution. (C, D) Magnified SEM images and elemental XEDS maps from the boxes in (A) and (B), used to measure the deposit composition. It is not possible to determine whether O was incorporated during or post-growth.

9

ACS Paragon Plus Environment

Nano Letters 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

References and Notes (1) Li, Y.; Dai, H. Chem. Soc. Rev. 2014, 43, 5257–5275. (2) Wang, Z.-L.; Xu, D.; Xu, J.-J.; Zhang, X.-B. Chem. Soc. Rev. 2014, 43, 7746–7786. (3) Naybour, R. D. J. Electrochem. Soc. 1969, 116, 520–524. (4) Arouete, S.; Blurton, K. F.; Oswin, H. G. J. Electrochem. Soc. 1969, 116, 166–169. (5) Chin, D.-T.; Sethi, R.; McBreen, J. J. Electrochem. Soc. 1982, 129, 2677–2685. (6) Banik, S. J.; Akolkar, R. Electrochim. Acta 2015, 179, 475–481. (7) Lan, C.; Lee, C.; Chin, T. Electrochim. Acta 2007, 52, 5407–5416. (8) Wang, J. M.; Zhang, L.; Zhang, C.; Zhang, J. Q. J. Power Sources 2001, 102, 139–143. (9) Bass, K.; Mitchell, P. J.; Wilcox, G. D.; Smith, J. J. Power Sources 1991, 35, 333–351. (10) Shin, J. W.; You, J.-M.; Lee, J. Z.; Kumar, R.; Yin, L.; Wang, J.; Meng, Y. S. Phys. Chem. Chem. Phys. 2016, 18, 26376–26382. (11) Banik, S. J.; Akolkar, R. J. Electrochem. Soc. 2013, 160, D519–D523. (12) McBreen, J.; Gannon, E. J. Electrochem. Soc. 1983, 130, 1980–1982. (13) Ross, F. M. Science 2015, 350, aaa9886. (14) Zeng, Z.; Liang, W. I.; Liao, H. G.; Xin, H. L.; Chu, Y. H.; Zheng, H. Nano Lett. 2014, 14, 1745–1750. (15) White, E. R.; Singer, S. B.; Augustyn, V.; Hubbard, W. A.; Mecklenburg, M.; Dunn; B.; Regan, B. C. ACS Nano 2012, 6, 6308–6317. (16) Mehdi, B. L.; Stevens; A.; Qian, J.; Park, C.; Xu, W.; Henderson, W. A.; Zhang, J.-G.; Mueller, K. T.; Browning, N. D. Sci. Rep. 2016, 6, 34267. (17) Rong, G.; Zhang, X.; Zhao, W.; Qiu, Y.; Liu, M.; Ye, F.; Xu, Y.; Chen, J.; Hou, Y.; Li, W.; Duan, W.; Zhang, Y. Adv. Mater. 2017, 29, 1606187. (18)Yang, C.; Han, J.; Liu, P.; Hou, C.; Huang, G.; Fujita, T.; Hirata, A.; Chen, M. Adv. Mater. 2017, 1702752. (19) Sun, K. E. K.; Hoang, T. K. A.; Doan, T. N. L.; Yu, Y.; Zhu, X.; Tian, Y.; Chen, P. ACS Appl. Mat. Interfaces 2017, 9, 9681–9687. (20) McBreen, J.; Gannon, E. J. Electrochem. Soc. 1983, 130, C303. (21) McBreen, J.; Gannon, E. J. Power Sources 1985, 15, 169–177. (22) Chandrasekar, M. S.; Pushpavanam, M. Electrochim. Acta 2008, 53, 3313–3322. (23) Grogan, J. M.; Schneider, N. M.; Ross, F. M.; Bau, H. H. Nano Lett. 2014, 14, 359–364. (24) This current should give an average deposition rate over the surface of the active electrode of ~0.31 nm/s (assuming 2 electrons per Zn2+ ion). (25) Holtz, M. E.; Yu, Y.; Gao, J.; Abruña, H. D.; Muller, D. A. Microsc. Microanal. 2013, 19, 1027–1035. (26) Ring, E. A.; Peckys, D. B.; Dukes, M. J.; Baudoin, J. P.; de Jonge, N. J. Microsc. 2011, 243, 273–283. (27) T. Kubiak. Static and dynamic buckling of thin-walled plate structures. Springer International Publishing, 2013; pp. 27–47. DOI: 10.1007/978-3-319-00654-3_2. (28) Zhang, X.; Wang, G.; Liu, X.; Wu, H.; Fang, B. Cryst. Growth Des. 2008, 8, 1430–1434. (29) Fang, J.; You, H.; Kong, P.; Yi, Y.; Song, X.; Ding, B. Cryst. Growth Des. 2007, 7, 864–867. (30) de Jonge, N.; Ross, F. M. Nat. Nano 2011 6, 695–704. (31) Yano, M.; Fujitani, S.; Nishio; K.; Akai, Y.; Kurimura, M. Appl. Energ. 1998, 74, 129–134. (32) Schneider, N. M.; Park, J. H.; Norton, M. M.; Ross, F. M.; Bau, H. H. Adv. Struct. Chem. Imaging 2016, 2, 2. (33) Park, J. H.; Schneider, N. M.; Grogan, J. M.; Reuter, M. C.; Bau, H. H.; Kodambaka, S.; Ross, F. M. Nano Lett. 2015, 15, 5314–5320. (34) Park, J. H.; Steingart, D. A.; Kodambaka, S.; Ross, F. M. Sci. Adv. 2017, 3, e1700234. (35) Park, J. H.; Reuter, M. C.; Kodambaka, S.; Ross, F. M. Microsc. Microanal. 2014, 20 (S3), 1598–1599. (36) Frankenthal, R. P.; Thompson, D. E. J. Electrochem. Soc. 1976, 123, 799–804. (37) Cadle, S. H.; Bruckenstein, S. Anal. Chem. 1974, 46, 16–20. (38) Shrestha, B. R.; Nishikata, A.; Tsuru, T. J. Electroanal. Chem. 2012, 665, 33–37. (39) In the case that the voltage is sufficiently negative, the following electrochemical reactions are expected: 2Au + 3H2O → Au2O2 + 6H+ + 6e-, E○ = -1.457 + 0.0591× log[H+] or Au+ + 3H2O → Au(OH)2 + 3H + 3e-, E○ = -1.511 + 0.0591× log[H+]. Therefore, the electrochemical dissolution of Au and formation of Au3+ can be described as Au → Au3+ + 3e-, E○ = -1.498 + 0.0197 × log[Au3+]. (40) Wang, K.; Pei, P.; Ma, Z.; Chen, H.; Xu, H.; Chen, D.; Wang, X. J. Mat. Chem. A 2015, 3, 22648–22655. (41) Ding, F.; Xu, W.; Graff, G. L.; Zhang, J.; Sushko, M. L.; Chen, X.; Shao, Y.; Engelhard, M. H.; Nie, Z.; Xiao, J.; Liu, X.; Sushko, P. V.; Lu, J.; Zhang, J.-G. J. Am. Chem. Soc. 2013, 135, 4450–4456. (42) Esfahani, M.; Zhang, J.; Durandet, Y.; Wang, J. J. Electrochem. Soc. 2016, 163, D476–D484. 10

ACS Paragon Plus Environment

Page 10 of 12

Page 11 of 12 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

Nano Letters

(43) Gallaway, J. W.; Gaikwad, A. M.; Hertzberg, B.; Erdonmez, C. K.; Chen-Wiegart, Y.-C. K.; Sviridov, L. A.; Evans-Lutterodt, K.; Wang, J.; Banerjee, S.; Steingart, D. A. J. Electrochem. Soc. 2014, 161, A275–A284. (44) Bard, A. J.; Parsons, B.; Jordon J. Standard Potentials in Aqueous Solutions, Dekker: New York, 1985. (45) Milazzo, G.; Caroli, S.; Sharma, V. K. Tables of Standard Electrode Potentials, Wiley: London, 1978. (46) Swift, E. H.; Butler, E. A. Quantitative Measurements and Chemical Equilibria; Freeman: New York, 1972. (47) Wei, X.; Desai, D.; Yadav, G. G.; Turney, D. E.; Couzis, A.; Banerjee, S. Electrochim. Acta 2016, 212, 603– 613. (48) Puckrin, E.; Slavin, A. J. Phys. Rev. B 1990, 41, 4970–4975. (49) Martínez, A. J. Phys. Chem. A 2014, 118, 5894–5902. (50) Radisic, A.; Vereecken, P. Direct Copper Plating in Nanostructure Science and Technology, Copper Electrodeposition for Nanofabrication of Electronics Devices; Kondo, K.; Rohan, N. A.; Dale, P. B.; Yokoi, M., Eds.; Springer Science+Business Media: New York, 2014; Ch. 7, 131–173. (51) Aogaki, R.; Makino, T. J. Electrochem. Soc. 1984, 131, 40–46. (52) Aogaki, R.; Makino, T. J. Electrochem. Soc. 1984, 131, 46–51. (53) Li, M; Luo, S.; Qian, Y.; Zhang, W.; Jiang, L.; Shen, J. J. Electrochem. Soc. 2007, 154, D567–D571. (54) Gomes, A.; da Silva Pereira, M. I. Electrochimica Acta 2006, 51, 1342–1350.

11

ACS Paragon Plus Environment

t=0s

Page 12 of 12

t = 200 s

No additives

t=0s

RMS roughness (nm)

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

Nano Letters

30

No additives Bi additives

20 10 0

0

100 200 300 Deposition time (s)

Bi additives ACS Paragon Plus Environment

t = 200 s