CTAB-Influenced Electrochemical Dissolution of ... - ACS Publications

Article pubs.acs.org/Langmuir

CTAB-Influenced Electrochemical Dissolution of Silver Dendrites Colm O’Regan,†,‡,§ Xi Zhu,⊥ Jun Zhong,†,‡ Utkarsh Anand,†,‡,§,∥ Jingyu Lu,†,‡,§,∥ Haibin Su,*,⊥ and Utkur Mirsaidov*,†,‡,§,∥ †

Centre for BioImaging Sciences, Department of Biological Sciences, National University of Singapore, 14 Science Drive 4, Singapore 117543 ‡ Department of Physics, National University of Singapore, 2 Science Drive 3, Singapore 117551 § Centre for Advanced 2D Materials, National University of Singapore, 6 Science Drive 2, Singapore 117546 ∥ NanoCore, National University of Singapore, 4 Engineering Drive 3, Singapore 117576 ⊥ Division of Materials Science, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798 S Supporting Information *

ABSTRACT: Dendrite formation on the electrodes of a rechargeable battery during the charge−discharge cycle limits its capacity and application due to short-circuits and potential ignition. However, understanding of the underlying dendrite growth and dissolution mechanisms is limited. Here, the electrochemical growth and dissolution of silver dendrites on platinum electrodes immersed in an aqueous silver nitrate (AgNO3) electrolyte solution was investigated using in situ liquid-cell transmission electron microscopy (TEM). The dissolution of Ag dendrites in an AgNO3 solution with added cetyltrimethylammonium bromide (CTAB) surfactant was compared to the dissolution of Ag dendrites in a pure aqueous AgNO3 solution. Significantly, when CTAB was added, dendrite dissolution proceeded in a step-by-step manner, resulting in nanoparticle formation and transient microgrowth stages due to Ostwald ripening. This resulted in complete dissolution of dendrites and “cleaning” of the cell of any silver metal. This is critical for practical battery applications because “dead” lithium is known to cause short circuits and high-discharge rates. In contrast to this, in a pure aqueous AgNO3 solution, without surfactant, dendrites dissolved incompletely back into solution, leaving behind minute traces of disconnected silver particles. Finally, a mechanism for the CTAB-influenced dissolution of silver dendrites was proposed based on electrical field dependent binding energy of CTA+ to silver.

■

INTRODUCTION

immersed in an AgNO3 electrolyte solution. CTAB surfactant was added to an aqueous solution of AgNO3 and the “dissolution” of dendrites formed was compared to dendrite dissolution in a pure aqueous AgNO3 electrolyte. The use of CTAB is a common method for the formation and modulation of silver (Ag) nanostructure shape.19−21 Previously, surfactants have also been used to control the electrochemical growth of Ag dendrites,22 but their application toward the understanding of growth or dissolution mechanisms is still missing. In our case, we show that dendrites completely dissolve in solution when CTAB is added. The dissolution starts at the dendrite tip and proceeds down to the electrode surface, completely “cleaning” the solution of silver metal. This prevents the formation of any “dead” metal in the electrochemical liquid cell, critical for avoiding short circuits and high-discharge rates in battery systems.3,23 Without CTAB, the dissolution does not start at the tip, and in some cases the silver does not fully

Metallic dendrite growth on negative electrodes during the charge−discharge cycle of a rechargeable battery is a major obstacle to the commercialization of next-generation electrochemical storage devices.1−4 These dendritic structures are responsible for the short-circuiting of batteries in portable electronics5−7 and their formation can lead to cell ignition or thermal runaway.8,9 Thus, dendrite formation in electrochemical cells presents an immediate safety concern regarding next-generation battery development.10 Compounding this problem is the continued miniaturization of device structures,11,12 which necessitates the development of methods that enable advanced operando investigations of nanoscale reactions inside an electrochemical cell.13−15 Despite the advancement of such techniques, there is still a significant knowledge gap with regards to dendrite formation and dissolution on electrodes. The probing of these processes and the underlying mechanisms is crucial in order to obtain a better understanding of how to prevent electrochemical dendrite growth.16 Here, we report the use of in situ liquid-cell transmission electron microscopy (TEM)17,18 to investigate the electrochemical dissolution of silver dendrites on platinum electrodes © 2016 American Chemical Society

Received: January 6, 2016 Revised: March 25, 2016 Published: March 27, 2016 3601

DOI: 10.1021/acs.langmuir.6b00037 Langmuir 2016, 32, 3601−3607

Article

Langmuir

catalyzing of the proton reduction at the electrode surface made it a suitable material for our in situ studies. A DC current−voltage profile was applied and the dynamics of the interface between the electrode and liquid electrolyte were monitored in real time inside a JEOL 2010FEG transmission electron microscope at a frame rate of 10 frames s−1 and 1024 × 1024 pixels with an Orius SC200 camera (Gatan Inc. Pleasanton, CA). Dendrite growth and dissolution from this AgNO3 solution in our electrochemical cell is shown in Figure 2A. This shows an in situ TEM image series of Ag dendrites growing and dissolving on the working electrode when an electrical bias was applied (Supporting Information, Video 1). The image series of dendrite growth dynamics in a thin film were acquired after the liquid receded due to bubble formation26 from electron beam exposure. For all observations reported, the potential difference applied between the working and counter electrodes was between +2 V and −2 V. The working electrode is visible in the bottom right corner of all frames as a reference (Figure 2A). Dendrite growth usually proceeded over several voltage cycles (Figures S2 and S3), where dendrites grew with a negative bias and dissolved with a positive bias. Electron diffraction analysis was performed on dendrites in both CTAB and aqueous AgNO3 solutions, and are indexed to the face-centered-cubic silver lattice (JCPDS file No. 04-0783) (Figure S4). The area of two branches of the dendrite shown in Figure 2A, during growth and dissolution (t = 0 s to t = 97.1 s), is given in Figure 2B. The initial phase of growth of the dendrite in Figure 2 is fast, from t = 5.5 s to t = 12 s, as seen from the projected area curves, after which growth slows down and the area of the dendrite levels off due to a depletion of Ag+ ions surrounding the electrode. During the growth, most of the reduced Ag0 in solution was deposited on the electrode and contributed toward dendrite growth. Hence, as time proceeded, less and less Ag+ ions were available in solution for reduction. When the bias was switched from −2 to +2 V at t = 28 s, the dendrites dissolved back into solution over a time period of approximately 60 s (panels: t = 66.9 s, t = 77.2 s, t = 97.1 s, Figure 2A and the brown shaded region of Figure 2B). Importantly, the dissolution started at the electrode surface and proceeded upward toward the dendrite tip, as observed by Schneider et al.27 This disconnected the dendrite from the electrode surface, leaving “dead” silver in solution. This can be seen in the last panel of Figure 2A (t = 97.1 s), where even after 90 s, pieces of Ag dendrite are still present in solution. Note that the effect of the electron beam on dendrite dissolution was also tested and was found to be negligible (Figure S5). We avoided in situ imaging at high magnification to keep the electron flux rate low (≤1 e Å−2 s−1) and to minimize the beam induced effects (Figure S6). To investigate the effect of surfactants on Ag dendrite dissolution, the aforementioned experiment was also performed in a solution containing 1 mM AgNO3 and 0.5 mM CTAB surfactant. The dissolution of dendrites with a positive bias of +2 V in a 1 mM AgNO3 solution with and without CTAB surfactant is shown in Figure 3A and B, respectively. Figure 3A represents a time frame of 0 to 13.2 s, and Figure 3B represents a time frame of 0 to 15.4 s. The CTAB concentration was kept to 0.5 mM, as CTAB exhibits a critical micelle concentration of approximately 1 mM.28 As shown in Figure 3A (Video 2), dendrite dissolution proceeded in a stepwise manner. This process involved the dendrite completely dissolving in solution, starting at the tip, with nanoparticle formation observed. In

dissolve. Instead, the dendrite disconnects from the electrode surface, leaving silver metal in solution.

■

RESULTS AND DISCUSSION The electrochemical cell setup used for this study is illustrated in Figure 1. Our electrochemical liquid cell was assembled in a

Figure 1. Schematic of the liquid cell setup used during the in situ TEM study. The top and bottom SiNx membranes sealed the AgNO3 solution in place. The solution was injected into the cell using a syringe and flow station. The bottom membrane contained three electrodes, spaced 20 μm apart. The cell, after assembly, was then placed in a flow-holder and inserted into the transmission electron microscope. The bottom panel shows the magnified electrode region of the liquid cell.

Hummingbird Scientific flow holder using two ultrathin (50 nm) electron transparent SiNx membranes (Supporting Information, Figure S1). The silver nitrate (AgNO3) electrolyte was injected into the cell at 2 μL min−1 using a Hummingbird Scientific flow-through pump station. Ag was selected as the metal of choice because it has sufficient mass−thickness contrast for imaging inside liquids within a transmission electron microscope due to its large atomic number (Z = 47). More importantly, it also has a positive standard thermodynamic potential (+0.80 V) with respect to the standard hydrogen electrode (SHE),24 so the reduction reaction at the electrode surface proceeds spontaneously, as given by the reactions below: Ag + + e− → Ag

E0 = + 0.80V vs SHE

H 2 → 2H+ + 2e−

E0 = 0.000V vs SHE

Ag + + H 2 → 2H+ + Ag E0 = + 0.80V vs SHE

Consequently, the electroplating of Ag is relatively consistent, when compared to other metals such as copper (+0.34 V) and lead (−0.13 V).24 AgNO3 was used specifically as the electrolyte because it dissolves readily in water, meaning facile electrolyte preparation. Platinum is a common electrode material used for electrochemical experiments because it does not corrode or dissolve in solution. Additionally, excellent reproducibility of the electrode potential25 and the effective 3602


Article

Langmuir

Figure 2. (A) TEM image series showing one cycle of Ag growth and dissolution on the working electrode upon the application of an electrical bias, as indicated by the label on the electrode. Importantly, parts of the dendrite, as well as Ag particulates are still visible at t = 97.1 s, approximately 60 s after dissolution started. (B) 3D schematic of the dendrite in Figure 2A at t = 9 s, along with the area curves for the two branches labeled 1 and 2. The red curve represents branch 1, and the yellow curve represents branch 2. The blue shaded region represents a negative bias (−2 V) applied to the electrode shown in (A), resulting in dendrite growth, while the brown shaded region represents a positive bias (+2 V) resulting in dissolution.

Figure 3. Comparison of dendrite dissolution in an aqueous AgNO3 + CTAB versus AgNO3 solution. (A) TEM image series showing dendrite dissolution in 1 mM AgNO3 + 0.5 mM CTAB aqueous solution. Particle formation and microgrowth was observed during stage-wise dissolution. The dendrite dissolved from its tip toward the root. No leftover Ag metal was observed in solution after the dendrite dissolved. (B) TEM image series showing dendrite dissolution in 1 mM AgNO3 aqueous solution in the absence of CTAB surfactant. Dendrite dissolution started at its root. Remaining Ag nanoparticulates are observed in solution.

pure aqueous AgNO3, the dendrites did not dissolve completely into solution, with small trace amounts of silver particulates remaining behind, as seen in Figure 3B (Video 3). For Figure 3A (also see Figure S7 and Video 2), at t = 2 s, the dendrites began to dissolve, starting at the tip. Dissolution initially occurred from t = 2 s to t = 3.8 s, after which

approximately one-third of the dendrite has dissolved. This was followed by nanoparticle formation, and microgrowth of new branches at t = 3.8 s at the tip of the dissolving dendrite due to Ostwald ripening - smaller particles giving way to larger particles, which eventually dissolved back into solution. At t = 5.5 s, dendrite dissolution began again and proceeded to t = 7.2 3603


Article

Langmuir

more Ag is exposed to the solution and oxidized to form Ag+ ions. The change in relative CTA+ coverage of silver was plotted in the inset of Figure 4 at T = 300 K. This plot shows the CTA+ coverage change as a function of the electric field strength. The reduction in CTA+ coverage over the dendrite surface will be greater as the binding energy decreases. This plot suggests that for the electric field strength of 2.53 × 108 V m−1 (estimated from a Debye length of 7.9 nm; see Methods and Materials), an almost 100% reduction in CTA+ coverage would occur. This dissolution process of dendrites in an AgNO3 solution with added CTAB surfactant is shown schematically in Figure 5. We propose that the dissolution process of dendrites in CTAB solution is due to the detachment of CTA+ ions from the Ag. Step 1 of the process began when the bias is switched to positve, which modulates the electric field and influences the binding energy. The higher the field strength, the lower the binding energy of CTA+ to the Ag surface, as already seen in Figure 4. The electric field will be stronger at the sharp points of the dendrite tips. Therefore, CTA+ ions will desorb first in the vicinity of these regions to expose silver directly to the solution in step 2. Step 3 in Figure 5 is the oxidation of silver at the dendrite branch tips followed by dissolution of these exposed regions of dendrite. Finally, step 4 is Ostwald ripening leading to microgrowth at the dendrite tip and particle formation during the dissolution process. This dissolution stage with the four above-mentioned steps is then repeated until the dendrite is fully dissolved back into solution. Figure 3A (Video 2) shows three dissolution stages, which are highlighted in the bottom left section of Figure 5. The first major dissolution stage occurred at approximately 4.0 s, the second at 8.4 s, and the final stage at approximately 13.0 s. Ostwald ripening of CTAB-capped nanoparticles has been observed previously,32,33 as well as Ostwald ripening of nanoparticles stabilized by other molecules.34 Jang et al.33 reported the Ostwald ripening of CTAB-capped Au nanoparticles which originated from the combination of redox reactions between H2O2 and AuNP’s under weakly acidic conditions. In particular, they highlighted the role of Br− ions in the process.33 In our case of electrochemical growth, H2O2 was not required to initiate the oxidation of CTA+-covered Ag dendrites, as this was accomplished when the electrical bias was switched to positive. Particle formation was initiated by a reduction in coverage of the dendrite surface by CTA+ due to the positive bias. The surfactant stabilized the surface of the Ag dendrites initially, which explains why they did not dissolve back into solution from root to tip,32 as observed with the pure aqueous AgNO3 solution. However, the gradual detachment of CTA+ due to the strong electric field allowed oxidation of Ag due to the positive bias and subsequent Ostwald ripening to form larger particles in solution. We propose that Ostwald ripening occurred due to the simultaneuous oxidation of the Ag dendrites due to the positive bias and reduction caused by Br− in solution, leading to microgrowth and particle formation. This was likely due to the formation of AgBrx− complexes in solution, as seen with Au by Jang et al.33 Also, this Ostwald ripening process resulted in the complete dissolving of silver dendrites in CTAB solution, “cleaning” the cell of any remaining metal. This formation of particles in solution during the dendrite dissolution process is an interesting occurrence, considering the catalytic capability of Ag and other metals. Overall, formation of electrode-bound nanoparticles in battery systems may prove

s where more particle formation and microgrowth was observed at the dendrite tip. At t = 7.2 s, two-thirds of the dendrite had dissolved into solution. Dissolution began once again at t = 8.5 s, and proceeded until the entire dendrite had dissolved. This dissolution starts at the dendrite tip and proceeds toward the electrode surface (Figure S7). Plots showing the change in area as a function of time for dendrites in both Figure 3A and B are given in Figure S8. Significantly, this dissolution process from dendrite tip to electrode surface may benefit battery technology due to the prevention of “dead” lithium−metal which has disconnected from the electrode, and can cause short circuits and high discharge rates.3,23 Therefore, dissolution starting at the tip and moving down toward the electrode, which is observed in CTAB solution, means the entire dendrite is removed. The formation of “dead” silver during the dendrite dissolution in an aqueous AgNO3 solution (in the absence of CTAB surfactant) is also shown in Figure S9. In order to resolve the observed mechanism of dendrite dissolution in a AgNO3 solution with added CTAB surfactant, the binding energies of CTA+ to a Ag (111) surface were computed by first-principles method, and plotted against the electric field strength (see Methods and Materials), in Figure 4.

Figure 4. Binding energy of CTA+ to the Ag dendrite surface plotted as a function of the electric field strength. Inset shows a plot of the change in relative coverage versus the electric field strength, |E ⃗|. This relative change in CTA+ coverage of the Ag surface at T = 300 K is C /C0 = exp(−ΔE b /kBT ). Here, given by ΔE b = E b(|E ⃗|) − E b(|E ⃗| = 0). The higher the electric field strength (at sharp edges and dendrite tips), the less the CTA+ coverage, and the more Ag is exposed to the solution and oxidized.

This plot shows how the binding energy of CTA+ decreases as the field strength increases. When a positive external electrical bias is applied, the binding energy between CTA+ and the silver dendrite surface decreases as the electric field strength increases. This reduction in binding energy leads to the desorption of CTA+ from the Ag dendrite surfaces, exposing the Ag to the solution and causing it to oxidize to Ag+ ions. Due to the concentration of electrical lines near sharp edges,29 the regions with high curvature in the dendrite, such as the tips of the dendrite branches, exhibit the highest electric field strength.30,31 The higher the electric field strength (at sharp edges and dendrite tips), the less the CTA+ coverage and the 3604


Article

Langmuir

Figure 5. Schematic showing the mechanism of Ag dendrite dissolution in AgNO3 + CTAB aqueous solution. Step 1 involves the switching of the bias to positive, which influences the binding energy of CTA+ to the Ag surface. The electric field was strongest at the tips and sharp points of the dendrite, which results in CTA+ detaching from these areas first, as seen in step 2. Step 3 is the oxidation of this exposed Ag to Ag+ ions in solution, resulting in dendrite dissolution beginning at the dendrite’s upper regions. This is followed by step 4, the formation of larger particles and microgrowth at the dendrite tips due to Ostwald ripening. This four-step stage of the dissolution process repeats from tip to electrode surface until the entire dendrite is dissolved. The TEM image on the bottom left represents the dendrite after the first dissolution stage has finished. 10 mL of water. This was then diluted down to 0.5 mM by adding 0.05 to 0.95 mL of 1 mM AgNO3 solution. Experimental Methods. Prior to loading, the top and bottom chips of the cell were oxygen plasma treated to render their membrane surfaces hydrophilic. This was done with a plasma coater (K100X Glow Discharge System, Quorum Technologies, Lewes, East Sussex, U.K.) discharging at 10 mA for 30 s with a base pressure of 2 × 10−1 mbar. The holder tubing system was flushed slowly with ultrapure water at a rate of 200 μL min−1 for approximately 1 h prior to loading the AgNO3 solution to ensure all possible contaminants were flushed from the system. The top and bottom chips were then secured in place within the liquid flow-through holder (Hummingbird Scientific, Lacey, WA), and the integrity of the membranes tested using a pumping station. The solution was loaded at a rate of 2 μL min−1 from a 1 mL syringe, until approximately 100 μL had passed through the flowholder. The low rate of 2 μL min−1 was used to ensure the integrity of the ultrathin SiNx membranes remained intact during loading. Finally, the integrity of the cell membranes was tested again using the pumping station (Hummingbird Scientific, Lacey, WA). We used a JEOL 2010FEG transmission electron microscope operating at 200 kV for in situ imaging with electron flux rates of less than 1 e Å−2 s−1. Imaging and recording was performed using an ORIUS SC200 CCD camera (Gatan Inc. Pleasanton, CA), at a rate of 10 frames s−1 and 1024 × 1024 pixels. When irradiated with the electron beam, the liquid in the cell retracted, leaving behind an ultrathin layer of electrolyte solution on the membrane, which was ideal for imaging dendrite formation. Finally, electrical data was recorded using an Agilent B2901A Precision Source/Measure Unit and a Keithley 2450 SourceMeter that were switched between +2 and −2 V. DC voltage profiles were used for our experiments. Image Processing. We implemented our image processing algorithm in the Miniconda python37 distribution using the numpy,38 opencv,39 scikit-image,40 cython,41 mahotas,42 and matplotlib43 libraries. The sequence of the raw image frames were inverted so that dendrites have higher intensity values than the background. In order to get rid of the illumination gradient the inverted images were subjected to a Gaussian blur σ = 1 pixels, followed by a gray scale tophat transform with a disc shaped structuring element of radius 20

advantageous, as silver nanoparticles have been shown to enhance the performance of Li−air batteries by catalyzing the formation of Li−O bonds.35 Other metals have shown similar behavior and this suggests Ag particles may hold promise in future battery technologies.36

■

CONCLUSIONS In summary, this study revealed, through in situ liquid-cell TEM, that CTAB surfactant influenced the electrochemical dendrite dissolution in an AgNO3 solution. When CTAB was added, Ag dendrites dissolved through the process of nanoparticle formation and microgrowth due to Ostwald ripening. The dendrites proceeded to dissolve from tip to root (i.e., toward the electrode surface) in steps, leading to complete dendrite dissolution and the absence of remaining “dead” silver. In contrast, without CTAB, the dendrites did not dissolve completely into solution. Instead, they dissolved from root to tip and thus disconnected from the electrode surface at the root, leaving behind small traces of silver metal. This study suggests possible control of electrochemical dendrite growth and dissolution in battery systems using additives. Silver nanoparticle formation is an important study for Li battery systems due to their electrocatalytic behavior. Hence, future work may highlight specific additives that can be utilized to suppress the growth or enhance the dissolution of different dendrites.

■

METHODS AND MATERIALS

Materials. A 10 mM AgNO3 electrolyte solution was prepared by dissolving 0.017 g of AgNO3 powder (Cat# 204390, Sigma-Aldrich Co., St Louis, MO) in 10 mL of water. This was then diluted down to 1 mM by adding 0.1 to 0.9 mL of ultra pure H2O (Cat# 320072, Sigma-Aldrich Co., St Louis, MO). A 10 mM CTAB solution was prepared by dissolving 0.036 g of CTAB surfactant (Cat# H9151, Sigma-Aldrich Co., St Louis, MO) in 3605


Langmuir

■

pixels. The background subtracted images were again subject to another Gaussian blurring with σ = 2 pixels to smoothen out the high frequency noise features. Then, the Otsu44 threshold algorithm was used to segment out the dendrites, where all the pixels above a certain intensity value were marked as 1, and 0 otherwise. Note that, at the end of this step, most of the regions with dendrites were marked as 1. The dendrites had a very poor contrast, and the above method failed for some branches in the field of view. In order to get rid of this fluctuation in segmentation, the observation that dendrites keep growing for a certain duration of time was used, implying that the area of dendrites can only increase. Boolean “OR” operations were performed on the frames where dendrites were growing, wherein the segmentation result of n+1th frame was replaced by the Boolean OR result between nth frame and n+1th frame. A similar operation between the nth and n−1th frames was carried out when the dendrites were dissolving. Further refinement of the segmented images was done based on the area of the segmented regions. A connected region which does not fall within a certain permissible limit of area was classified as 0, and henceforth considered to be a part of the background. Binding Energy Calculations. Here, we chose the Ag (111) surface to model the dendrite facets since this surface has the lowest surface energy and so is the most stable. The surface structure consisted of 384 Ag atoms, with one CTAB molecule adsorbed on it. First principle density functional theory was implemented in SIESTA code,45 within the local density approximation.46 A cutoff energy of 4082 eV and a double-ζ polarized basis45 were chosen for all elements, to investigate the energetic nature of this structure. The computed binding energy (Eb) was defined as

Article

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.6b00037. Schematic of electrochemical liquid cell, SAED analysis, detailed image series showing dendrite tip dissolution, image series showing “dead” metal left behind in solution in pure aqueous AgNO3 solution, and further image series of dendrite growth in pure aqueous AgNO3 solution (PDF) Growth and dissolution of silver dendrite on a platinum electrode immersed in a pure aqueous 1 mM silver nitrate solution; video shows 1 cycle of growth and dissolution at −2 V and +2 V bias voltage, respectively (AVI) Dissolution of silver dendrite from a platinum electrode immersed in a 1 mM silver nitrate and 0.5 mM CTAB solution; video highlights the step-by-step dissolving of silver dendrites upon the application of +2 V bias to the electrode, along with particle formation due to Ostwald ripening; all silver metal is “cleaned” from solution after complete dissolution (AVI) Dissolution of silver dendrite on a platinum electrode immersed in a pure aqueous 1 mM silver nitrate solution; video highlights the disconnection of the dendrites from the electrode and the incomplete dissolving of the dendrite into solution, upon the application of +2 V bias; no Ostwald ripening is observed and traces of Ag can be seen after dissolving (AVI)

E b = ECTA+/Ag(111) − EAg(111) − ECTA+ where ECTA+, EAg(111), and ECTA+/Ag(111) are the computed energies of CTA+, the Ag(111) surface, and CTA+ adsorbed on the Ag(111) surface, respectively. The computed binding energy was 2.23 eV, which is in good agreement with previous results.21 To account for the change of the binding energy due to the external electric field, we calculated the binding energy with the presence of the electric field using SIESTA code. The electric field dependent binding strength was plotted in Figure 4. Since the variation of the binding energy lead to a change in the coverage of CTA+ on the Ag dendrites, the relative change in coverage of CTA+, C/C0 on Ag dendrites is given by47

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (H.S., binding energy calculation). *E-mail: [email protected] (U.M., experiments). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Singapore National Research Foundation’s Competitive Research Program funding (NRFCRP9-2011-04).

C = exp(−ΔE b /kBT ) C0

■

Here, k B is the Boltzmann constant, while C = C(|E ⃗|) and C0 = C0(|E ⃗| = 0) are the CTA+ coverage of the Ag surface in the presence of an electric field of strength |E ⃗| and in the absence of an electric field (|E ⃗| = 0 ), respectively. We computed the change in the binding energy, ΔEb = Eb(|E ⃗|) − Eb(|E ⃗| = 0), at a temperature of T = 300 K using SIESTA codes. The relative coverage of silver surface by CTA+ ions was plotted in the inset of Figure 4. The Debye length (δ) was calculated using47

REFERENCES

(1) White, E. R.; Singer, S. B.; Augustyn, V.; Hubbard, W. A.; Mecklenburg, M.; Dunn, B.; Regan, B. C. In situ transmission electron microscopy of lead dendrites and lead ions in aqueous solution. ACS Nano 2012, 6 (7), 6308−6317. (2) Aurbach, D.; Zinigrad, E.; Cohen, Y.; Teller, H. A short review of failure mechanisms of lithium metal and lithiated graphite anodes in liquid electrolyte solutions. Solid State Ionics 2002, 148 (3), 405−416. (3) Tarascon, J.-M.; Armand, M. Issues and challenges facing rechargeable lithium batteries. Nature 2001, 414 (6861), 359−367. (4) Ross, F. Electrochemical nucleation, growth and dendrite formation in liquid cell TEM. Microsc. Microanal. 2010, 16 (S2), 326−327. (5) Li, H.; Wang, Z.; Chen, L.; Huang, X. Research on advanced materials for Li-ion batteries. Adv. Mater. 2009, 21 (45), 4593. (6) Palacin, M. R. Recent advances in rechargeable battery materials: a chemist’s perspective. Chem. Soc. Rev. 2009, 38 (9), 2565−2575. (7) Quartarone, E.; Mustarelli, P. Electrolytes for solid-state lithium rechargeable batteries: recent advances and perspectives. Chem. Soc. Rev. 2011, 40 (5), 2525−2540.

⎛ ε εk T ⎞1/2 δ = ⎜ 20 B ⎟ ⎝ e NA 2C ⎠ where C is the ion concentration of CTAB and AgNO3 (0.5 and 1 mM respectively), e = −1.6 × 10−19 C is the charge of an electron, ε = 78 is the relative permittivity of water, ε0 = 8.854 × 10−12 C2 J−1 m−1 is the permittivity of free space, NA = 6.022 × 1023 mol−1 is Avogadro’s constant, and T = 300 K. δ was calculated to be 7.9 nm, which led to an estimated electrical field strength of |E ⃗| = V /δ ∼ 2.5 × 108 V m−1 for a potential of 2 V. 3606


Article

Langmuir (8) Ostrower, J.; Pasztor, A. Microscopic ″Dendrites ″a Focus in Boeing Dreamliner Probe. Wall Street Journal [Online], Feb 11, 2013, http://www.wsj.com (accessed Jan 20, 2016). (9) Yoshino, A. The Birth of the Lithium-Ion Battery. Angew. Chem., Int. Ed. 2012, 51 (24), 5798−5800. (10) Kim, H.; Jeong, G.; Kim, Y.-U.; Kim, J.-H.; Park, C.-M.; Sohn, H.-J. Metallic anodes for next generation secondary batteries. Chem. Soc. Rev. 2013, 42 (23), 9011−9034. (11) Armstrong, M. J.; O’Dwyer, C.; Macklin, W. J.; Holmes, J. D. Evaluating the performance of nanostructured materials as lithium-ion battery electrodes. Nano Res. 2014, 7 (1), 1−62. (12) Song, J.; Yang, X.; Zeng, S.-S.; Cai, M.-Z.; Zhang, L.-T.; Dong, Q.-F.; Zheng, M.-S.; Wu, S.-T.; Wu, Q.-H. Solid-state microscale lithium batteries prepared with microfabrication processes. J. Micromech. Microeng. 2009, 19 (4), 045004. (13) Whittingham, M. S. Lithium batteries and cathode materials. Chem. Rev. 2004, 104 (10), 4271−4302. (14) Sacci, R. L.; Black, J. M.; Balke, N.; Dudney, N. J.; More, K. L.; Unocic, R. R. Nanoscale Imaging of Fundamental Li Battery Chemistry: Solid-Electrolyte Interphase Formation and Preferential Growth of Lithium Metal Nanoclusters. Nano Lett. 2015, 15 (3), 2011−2018. (15) Abellan, P.; Mehdi, B. L.; Parent, L. R.; Gu, M.; Park, C.; Xu, W.; Zhang, Y.; Arslan, I.; Zhang, J.-G.; Wang, C.-M.; et al. Probing the degradation mechanisms in electrolyte solutions for Li-ion batteries by in situ transmission electron microscopy. Nano Lett. 2014, 14 (3), 1293−1299. (16) Li, W.; Zheng, H.; Chu, G.; Luo, F.; Zheng, J.; Xiao, D.; Li, X.; Gu, L.; Li, H.; Wei, X.; et al. Effect of electrochemical dissolution and deposition order on lithium dendrite formation: a top view investigation. Faraday Discuss. 2014, 176, 109−124. (17) Sutter, E. A.; Sutter, P. W. Determination of Redox Reaction Rates and Orders by In Situ Liquid Cell Electron Microscopy of Pd and Au Solution Growth. J. Am. Chem. Soc. 2014, 136 (48), 16865− 16870. (18) Woehl, T. J.; Park, C.; Evans, J. E.; Arslan, I.; Ristenpart, W. D.; Browning, N. D. Direct observation of aggregative nanoparticle growth: Kinetic modeling of the size distribution and growth rate. Nano Lett. 2014, 14 (1), 373−378. (19) Zheng, X.; Zhu, L.; Wang, X.; Yan, A.; Xie, Y. A simple mixed surfactant route for the preparation of noble metal dendrites. J. Cryst. Growth 2004, 260 (1), 255−262. (20) Rao, G. P. C.; Tharmaraj, V.; Yang, J. Surfactant-assisted electroless deposition of silver nanoparticles on Ge crystal for ultrasensitive detection by surface-enhanced infrared absorption spectroscopy. RSC Adv. 2015, 5 (26), 20390−20395. (21) Zeng, Q.; Jiang, X.; Yu, A.; Lu, G. M. Growth mechanisms of silver nanoparticles: a molecular dynamics study. Nanotechnology 2007, 18 (3), 035708. (22) Hong, X.; Wang, G.-z.; Wang, Y.; Zhu, W.; Shen, X.-s. Controllable electrochemical synthesis of silver dendritic nanostructures and their SERS properties. Chin. J. Chem. Phys. 2010, 23 (5), 596−602. (23) Guo, Z.; Zhu, J.; Feng, J.; Du, S. Direct in situ observation and explanation of lithium dendrite of commercial graphite electrodes. RSC Adv. 2015, 5 (85), 69514−69521. (24) Gileadi, E. Physical electrochemistry: fundamentals, techniques and applications; Wiley VCH: Weinheim, Germany, 2011. (25) Sawyer, D. T.; Sobkowiak, A.; Roberts, J. L. Electrochemistry for chemists; Wiley: New York, 1995. (26) Grogan, J. M.; Schneider, N. M.; Ross, F. M.; Bau, H. H. Bubble and pattern formation in liquid induced by an electron beam. Nano Lett. 2014, 14 (1), 359−364. (27) Schneider, N.; Park, J.; Grogan, J.; Kodambaka, S.; Steingart, D.; Ross, F.; Bau, H. In-Situ Electron Microscopy of Electrochemical Deposition, Dendrite Growth, and Etching. J. Heat Transfer 2014, 136 (8), 080910. (28) Neugebauer, J. M. Detergents: an overview. Methods Enzymol. 1990, 182, 239−253.

(29) Brailsford, D. F.; Robertson, A. J. Calculation of electric field strengths at a sharp edge. Int. J. Mass Spectrom. Ion Phys. 1968, 1, 75− 85. (30) Rezinkina, M. Modeling of the dendrite shape variation with applied electric field strength in poly (ethylene). Tech. Phys. Lett. 2000, 26 (3), 196−198. (31) Kłos, J.; Sommer, J.-U. Simulations of a Grafted Dendritic Polyelectrolyte in Electric Fields. Macromolecules 2015, 48 (4), 1179− 1186. (32) Gubicza, J.; Lábár, J. L.; Quynh, L. M.; Nam, N. H.; Luong, N. H. Evolution of size and shape of gold nanoparticles during long-time aging. Mater. Chem. Phys. 2013, 138 (2), 449−453. (33) Jang, E.; Lim, E.-K.; Choi, J.; Park, J.; Huh, Y.-J.; Suh, J.-S.; Huh, Y.-M.; Haam, S. Br-assisted Ostwald ripening of Au nanoparticles under H2O2 redox. Cryst. Growth Des. 2012, 12 (1), 37−39. (34) Deivaraj, T.; Lala, N. L.; Lee, J. Y. Solvent-induced shape evolution of PVP protected spherical silver nanoparticles into triangular nanoplates and nanorods. J. Colloid Interface Sci. 2005, 289 (2), 402−409. (35) Lu, J.; Cheng, L.; Lau, K. C.; Tyo, E.; Luo, X.; Wen, J.; Miller, D.; Assary, R. S.; Wang, H.-H.; Redfern, P. Effect of the size-selective silver clusters on lithium peroxide morphology in lithium−oxygen batteries. Nat. Commun. 2014, 5, 4895. (36) Dathar, G. K. P.; Shelton, W. A.; Xu, Y. Trends in the catalytic activity of transition metals for the oxygen reduction reaction by lithium. J. Phys. Chem. Lett. 2012, 3 (7), 891−895. (37) Millman, K. J.; Aivazis, M. Python for scientists and engineers. Comput. Sci. Eng. 2011, 13 (2), 9−12. (38) Van Der Walt, S.; Colbert, S. C.; Varoquaux, G. The NumPy array: a structure for efficient numerical computation. Comput. Sci. Eng. 2011, 13 (2), 22−30. (39) Bradski, G. The opencv library. Doctor Dobbs Journal 2000, 25 (11), 120−126. (40) Van Der Walt, S.; Schönberger, J. L.; Nunez-Iglesias, J.; Boulogne, F.; Warner, J. D.; Yager, N.; Gouillart, E.; Yu, T. scikitimage: image processing in Python. PeerJ 2014, 2, e453. (41) Behnel, S.; Bradshaw, R.; Citro, C.; Dalcin, L.; Seljebotn, D. S.; Smith, K. Cython: The best of both worlds. Comput. Sci. Eng. 2011, 13 (2), 31−39. (42) Coelho, L. P. Mahotas: Open source software for scriptable computer vision. arXiv:1211.4907. arXiv.org e-Print archive. http:// arxiv.org/abs/1211.4907. (43) Hunter, J. D. Matplotlib: A 2D graphics environment. Comput. Sci. Eng. 2007, 9 (3), 90−95. (44) Otsu, N. A threshold selection method from gray-level histograms. Automatica 1975, 11 (285−296), 23−27. (45) Soler, J. M.; Artacho, E.; Gale, J. D.; García, A.; Junquera, J.; Ordejón, P.; Sánchez-Portal, D. The SIESTA method for ab initio order-N materials simulation. J. Phys.: Condens. Matter 2002, 14 (11), 2745. (46) Perdew, J. P.; Zunger, A. Self-interaction correction to densityfunctional approximations for many-electron systems. Phys. Rev. B: Condens. Matter Mater. Phys. 1981, 23 (10), 5048. (47) Israelachvili, J. N. Intermolecular and surface forces, revised third ed.; Academic Press: New York, 2011.

3607
