Jiancheng
Lin
a,
Mohamed
Kilani
a,
Mahroo
Baharfar
a,
Ren
Wang
a and
Guangzhao
Mao
*ab
aSchool of Chemical Engineering, University of New South Wales (UNSW Sydney), Sydney, New South Wales 2052, Australia. E-mail: guangzhao.mao@unsw.edu.au
bSchool of Engineering, Institute for Materials and Processes, The University of Edinburgh, Robert Stevenson Road, Edinburgh, EH9 3FB, UK
First published on 7th October 2024
Electrodeposition is used at the industrial scale to make coatings, membranes, and composites. With better understanding of the nanoscale phenomena associated with the early stage of the process, electrodeposition has potential to be adopted by manufacturers of energy storage devices, advanced electrode materials, fuel cells, carbon dioxide capturing technologies, and advanced sensing electronics. The ability to conduct precise electrochemical measurements using cyclic voltammetry, chronoamperometry, and chronopotentiometry in addition to control of precursor composition and concentration makes electrocrystallization an attractive method to investigate nucleation and early-stage crystal growth. In this article, we review recent findings of nucleation and crystal growth behaviors at the nanoscale, paying close attention to those that deviate from the classical theories in various electrodeposition systems. The review affirms electrodeposition as a valuable method both for gaining new insights into nucleation and crystallization on surfaces and as a low-cost scalable technology for the manufacturing of advanced materials and devices.
Technique | Advantages | Disadvantages | Material type | Dimension | Ref. |
---|---|---|---|---|---|
Electrodeposition | Cost effective, simple process set-up, relatively easy to scale up, ambient condition process, wide variety of deposits by choosing the right precursors and conditions, one-step synthesis and assembly is possible | Requires conductive surface for growth, which can limit device configuration, can have low crystallinity and requires post treatment, unapplicable to insulating materials | Metal | 0D, 3D | 15–17 |
Metal compound, e.g., oxides, hydroxides, dichalcogenides, nitride | 0D, 1D, 2D, 3D | 10 and 18–21 | |||
Metal–organic framework | 3D | 22 and 23 | |||
Charge-transfer complex | 1D | 24 and 25 | |||
Polymer | 2D | 26 | |||
Chemical vapor deposition (CVD), e.g., metal organic, plasma enhanced, atomic layer deposition | Controllable thickness for 2D structures, scalable to some extent, high quality and conformity 2D crystal | High vacuum and temperature needed for the process, high capital and operational cost, suitable substrate needed for growth of high-quality crystals, gaseous precursor can be dangerous to handle, slow deposition rate, requires specialized trained personnel for operation | Metal compounds, e.g., oxide, dichalcogenides, carbide | 1D, 2D | 27–30 |
Metal–organic framework | 2D | 31 and 32 | |||
III–V compound | 1D, 2D | 33–35 | |||
Polymer | 2D | 36 | |||
Physical vapor deposition (PVD), e.g., e-beam, sputtering, pulsed laser, molecular beam epitaxy | Wafer-scale deposition is possible, may require lower operating temperature than CVD | High vacuum and temperature needed for the process, high capital and operational cost, slow deposition rate, requires specialized trained personnel for operation, specific substrate is necessary for highly crystalline product, poorer crystallinity or smaller grain size as compared to CVD | Metal compounds, e.g., oxides, dichalcogenides | 2D | 37 and 38 |
Solution-based chemical methods, e.g., hydrothermal, solvothermal, sol–gel, hot-injection, solvent evaporation | Some solution-based process does not require specific specialized equipment, compatible with conventional chemical reactor, can achieve morphological change by simply altering reaction conditions | Process time can be very long, up to several hours for the case of hydrothermal method, precise control over reaction rate can be challenging, often involve two-step: synthesis and assembly – difficult to scale | Metal compounds, e.g., halides, oxides | 0D, 2D | 39–41 |
Charge-transfer complex | 1D, 2D | 42 |
Fig. 1 outlines in situ analytical techniques for advancing the microscopic understanding of the electrocrystallization processes and its impact on real-world applications, which will be described in detail in the section on “Experimental Techniques for Electrochemical Nucleation and Growth”. The microscopic picture of nucleation and crystal growth, at the nanoscale or even atomic scale, continues to evolve, facilitated by advances in experimental and modeling tools. Due to the growing importance of electrodeposition in cutting-edge technologies such as battery technology, fuel cells, electrocatalysis, sensors, wearable electronics, and environmental remediation, this review fills a gap in current literature by putting together latest findings that have contributed significantly to our understanding of electrochemical nucleation and growth down to the single-nanoparticle level. We include electrodeposition of nonmetals including charge-transfer complexes (CTCs), metal organic frameworks (MOFs), mineral salts, and metal oxides, which are not typically discussed with metal electrodeposition. We benefit from earlier reviews and perspectives on aspects of electrochemical nucleation and crystal growth.11–14
Our literature survey shows that Li electrodeposition and dendrite formation have continuously been of interest to researchers because of the relevance in mitigating battery degradation. Others delve into the challenge in controlling electrodeposition for catalysis, sensing, wearable electronics, and environmental applications. A search on the Web of Science using keywords including “electroplating”, “electrodeposition”, “electrochemical deposition”, “electrocrystallization”, and “electrochemical crystallization” indicates a sustained interest in this field (Fig. 2). Despite the increasing importance of electrodeposition in cutting-edge technologies such as battery technology, fuel cells, electrocatalysis, sensors, wearable electronics, and environmental remediation, to our best knowledge there are no recent reviews on nanoscale understanding of electrodeposition. This review therefore fills this significant gap by putting together latest findings that have contributed to our understanding of electrochemical nucleation and growth down to the single-entity level. In-depth understanding of the electrodeposition process will enable precise control of nanostructure formation and enable wide adoption of electrodeposition by manufacturers of energy storage devices, advanced electrode materials, fuel cells, carbon dioxide capturing technologies, and advanced sensing devices. A better understanding of nucleation and initial crystal growth will lead to the realization of desirable and reproducible functional properties in electrodeposited thin films, coatings, devices, and systems.
(1) |
Fig. 3 Schematic representation of classical vs. multi-step nucleation mechanisms for (a) the formation of crystals from supersaturated solutions and (b) their corresponding energy barriers: a single barrier in classical nucleation (solid blue curve), a multi-step barrier model for nonclassical nucleation (dashed green curve) and associated prenucleation clusters. Redrawn from ref. 43. |
Now specific to electrocrystallization, ΔGv varies with η according to the following equation:11
ΔGv = zF|η|/Vm, | (2) |
(3) |
The number of atoms in the critical nucleus, Ncritical, is related to rcritical and proportional to γ3/|η|3:
(4) |
Here, NA is the Avogadro's number.
In the classical nucleation theory, the nucleation rate, J (number of nuclei formed per area per second), varies with η as follows.
(5) |
(6) |
Here, R is the gas constant and T is the temperature. An alternative model to the classical nucleation theory that describes electrochemical nucleation is the atomistic model developed by Milchev et al.44,45 Similarly, the atomistic model describes the nucleation process being atom-by-atom addition and predicts a critical number of atoms for a thermodynamically stable nucleus to form. The difference is that the atomistic model takes into account of the energy associated with the interaction between individual atoms and their interactions with the electrode surface, instead of the interfacial energy of the nucleus.46–48
If electrodeposition is limited by diffusion, nucleation can be defined as instantaneous or progressive. If the number of nuclei remains the same (N = N0) from the start of the potential step in potentiostatic electrodeposition, then it is instantaneous nucleation. If N increases with time, t, then it is progressive nucleation. The Scharifker–Hills model has been widely used for analyzing the current–time curves in potentiostatic electrodeposition of metals:49–54
(7) |
Taking into account of the electrochemical reaction (charge transfer) step at the solution/electrode interface, Fletcher derived the following equation for the current–time behavior in electrodeposition of hemispherical metal nuclei of radius r (the Fletcher model):58
(8) |
(9) |
(10) |
Here, iS is the current for a single nucleus formation, N0 is the number of isolated nuclei formed instantaneously, ρm is the molar density, I0 is the exchange current density determined by the Tafel plot,59α is the charge-transfer coefficient, τ is the nucleation induction time, and the rate constant of the deposition reaction, kf, is given by eqn (10). Building on these works, models were developed to better describe the current–time transient during the early stage of electrodeposition in the mixed reaction-diffusion control regime.60,61 Regardless, these models enabled the determination of J and Ncritical based on the current–time transient.
Crystal growth according to the classical theories progresses by incremental addition of atoms (or molecules), resulting in the replication of unit cells in the growing crystal, with no accompanying structural changes. Crystal growth is described by the Frank–van der Merwe 1D growth, 2D monolayer growth, or Volmer–Weber 3D island growth mechanisms.62 In general, strong adatom and surface interactions favor 2D growth while weak interactions favor 3D island growth. Under equilibrium conditions, the shape (morphology) of the single crystal is dictated by the minimization of total Gibbs surface free energy, as prescribed by the Wulff theorem.63 With increasing chemical potential driving force (e.g., higher η), the surface integration of atoms changes from the spiral 1D growth mechanism to 2D growth mechanism where atoms attach to kinks and steps on the bulk crystal surface and the rough growth mechanism associated with spherulitic, fractal, and dendritic patterns. Fractal (dendritic) growth, particularly relevant for Li batteries, can be described by diffusion-limited aggregation assuming irreversible adatom sticking upon initial contact with the crystal surface.64,65
The key steps in 2D electrodeposition are: (1) bulk diffusion, which involves atoms and molecules moving from solution to the electrode surface as adatoms; (2) charge transfer, which involves the transfer of charge at the electrode surface; (3) surface diffusion, which involves atoms or molecules moving along the electrode surface with the surface diffusion rate determined from logarithmic density of 3D islands vs. 1/T plot; and (4) transfer from adatom to step or kink position, which is defined by the transition of atoms or molecules from their initial adatom position to either a step or kink position on the crystalline surface.11 Moreover, the Winand diagram has been used as an empirical guide to predict crystal morphology based on i (normalized by the diffusion-limiting i) and the inhibition intensity.66 Inhibition refers to any factor that slows down surface reactions, such as adsorbed species requiring displacement by primary metal cations at the electrode surface. According to the Winand diagram, the combination of low i and slow surface reaction results in higher degree of order in the electrodeposited thin films.
For the 3D growth of a single particle, the crystal growth rate of a spherical particle of radius r can be expressed by eqn (11)–(13) depending on the nature of the rate-limiting step.67 If diffusion is the rate-limiting step, the crystal growth rate can be derived from Fick's first law:
(11) |
(12) |
(13) |
In electrocrystallization, the growth rate varies with η. In the case of a reaction-limited system, the growth rate of a hemispherical nucleus on an electrode surface is described as following:68
(14) |
Crystal growth through particle aggregation and coalescence has also been observed.114 In zeolite crystallization, a wide range of crystal growth pathways have been identified including monomer-by-monomer, oligomers, gel-like islands, amorphous nanoparticles, colloidal assembly, and oriented attachment.87 We want to point out that electrocrystallization models to explain nonclassical behaviors observed in electrodeposition use the same concepts of nanocluster surface diffusion and aggregation, orientated attachment, and recrystallization13,115–120 as those in zeolite crystallization. The concept of a nanocluster building block was utilized in the formulation of the generalized electrochemical aggregative growth mechanism for metal electrodeposition on low-energy surfaces (Fig. 4).13 The nanoclusters with self-limiting growth are similar in concept to the PNCs found in many nonelectrochemical crystallization systems.108–113 According to the aggregative growth model, nanocluster surface diffusion is explained by adatom dissolution and re-adsorption process, i.e., the higher the mobile adatom concentration on the nanocluster, the higher the surface diffusion coefficient of the nanocluster. The surface diffusion coefficient increases with increasing η. The rates of nanocluster coalescence and recrystallization also increase with increasing η. A major deviation from the classical picture is in the interpretation of the current–time transients in the induction period of metal electrodeposition. The induction time in the chronoamperometric data according to the electrochemical aggregative growth mechanism is related to aggregate–nucleation events rather than to the standard nucleation process. The crystal growth pathways are η dependent. This aggregative growth phenomenon during the early stages of electrodeposition highlights the risk that values of J and Ncritical calculated using the classical model may deviate significantly from the experimental results through in situ surface analysis.12
Fig. 4 Schematic diagram showing the different crystallization stages of the generalized electrochemical aggregative growth mechanism. Dots represent the nongrowing nanoclusters and blue circles around the aggregates represent their nucleation exclusion zones. Black stripes within a particle represent defects. Reprinted with permission from ref. 13. Copyright 2013 American Chemical Society. |
Variations to the classical Volmer–Weber 3D island growth model based on nanocluster aggregation have been proposed by the Unwin group.117 The nanocluster aggregation rate constant, kAGG, can be extracted from the experimental chronoamperometric curves. The nanocluster nucleation rate, kN, and the nucleus dissolution rate, kD, can be expressed as following:
(15) |
(16) |
(17) |
Here, N is the number of isolated nanoclusters, NAGG is the number of aggregated nanoclusters, QR is total charge passed in an i–t transient, and QR is the charge passed per nanocluster. kAGG varies with η.
Other nonclassical crystal growth mechanisms include oriented attachment121–126 and Ostwald ripening growth.127,128 The oriented growth model assumes that nanoclusters with short-range orders can directly attach to each other or to larger crystals. For example, CaSO4·2H2O undergoes a three-stage crystal growth process:126 (1) homogeneous precipitation of nanocrystalline hemihydrate bassanite below its predicted solubility, (2) self-assembly of bassanite into elongated aggregates co-oriented along their c axis, and (3) transformation into dihydrate gypsum. This oriented attachment has recently been observed in the electrocrystallization of the charge-transfer complex (CTC) cobalt tetracyanoquinodimethane complex (Co-TCNQ).25 For the Ostwald ripening growth, the Lifshitz–Slyozov–Wagner (LSW) theory is used to describe the crystal growth kinetics of a spherical particle with radius r a follows.129,130
(18) |
Our literature review of electrodeposition identifies many instances that deviate from the classical behaviors, which will be discussed in detail in the last part of the paper. While theoretical treatments of nucleation and crystal growth in electrodeposition remain limited, we foresee future development to be significantly aided by advances in nanomaterials characterization tools with increasing spatiotemporal resolutions. New method development for the characterization of electrochemical nucleation and early-stage crystal growth will be highlighted in the next section.
Here we review microscopic and X-ray methods that have been used to capture nucleation and crystal growth events at the single-entity level, typically at or smaller than nanoscale, most relevant to electrodeposition. An excellent review on in situ kinetic observations of crystal nucleation and growth has been provided by Li and Deepak.131 We keep this part brief by focusing on aspects of the methods that are most relevant for the monitoring of the electrochemical process. Table 2 summarizes the various techniques, principles, and applications in electrodeposition.
Technique | Advantages | Disadvantages | Applications in electrodeposition | Ref. |
---|---|---|---|---|
TEM/STEM | High resolution within atomic to micro range, possibility of structural, morphological, and phase characterization | Challenges in electron beam focusing across liquid electrolyte, the effect of electron beam on structures and electrolyte, limited probing area, electrochemical data might be impacted by using a thin layer of electrolyte | Observation of lithiation/delithiation in Si terminals, characterization of crystal nucleation and growth during electrocrystallization | 144–154 |
SEPM | Real-time observation of morphological, topographical, and electrochemical features | Lower resolution compared to TEM and STEM, the possible effect of mechanical forces on structures, probe effect | Versatile technique for observation of in situ electrocrystallization, monitoring electrocatalytic reactions in real time, imaging electrochemical activity of nanostructures, monitoring electrode surface dynamics | 155–157 |
XRD/SAXS | Non-destructive, large probing area, provides crystallographic information | Poor signal-to-noise ratio in case of limited amount of sample, insensitivity to intermediate amorphous structures | Monitoring catalytic interfaces during catalysis, real-time observation of crystallization (non-classical nucleation in particular) | 81 and 158 |
An essential component for EC-LCTEM is a liquid cell that can accommodate both the electrochemical reaction on the working electrode and TEM imaging. Radisic et al. in 2006 reported one of the earliest LCTEM studies of electrocrystallization of Cu with simultaneous chronoamperometry and TEM data capture.146 A 1–2 μm thick electrochemical cell was constructed to enclose the electrolytic solution between two Si3N4 windows, one of which contained a micropatterned Au working electrode. The Au ultramicroelectrode was connected to the leads that were connected to an electrochemical workstation to perform electrodeposition and electroanalytical measurements. In 2014, Unocic et al. introduced a microfluidic electrochemical liquid cell design for STEM imaging, diffraction, and spectroscopy.150 The microfluidic electrochemical liquid cell has a three-electrode configuration with glassy carbon (GC) and Pt microband electrodes. The electrochemical cell design was further improved by using a thin nanofluidic channel TEM liquid sample holder with two Si3N4 windows and a metal-coated microchip serving as the working electrode.144 Multiple ultramicroelectrodes can be built into the same liquid cell holder so that multiple experiments can be performed on the same microchip, and the ultramicroelectrode geometry can be varied to control the assembly of nanostructures.148 To mitigate systematic errors in electrochemical measurements inside a TEM liquid cell Fahrenkrug et al. provided design criteria for TEM liquid cells used in conjunction with electrochemical measurements.144
LCTEM and STEM studies of electrodeposition will be described in detail in the following section. Here, we highlight a few recent examples of their use in capturing atomic-level evidence of non-classical nature in nonelectrochemical crystallization systems. The purpose of including these examples is to forecast the potential use of LCTEM and STEM for electrocrystallization studies.
Low-dose LCTEM was used to monitor the transition of dispersed Au nanoprisms to an Au superlattice at the single-particle level.82 Combining real-time particle tracking with Monte Carlo simulations, a nonclassical nucleation pathway involving a dense, amorphous intermediate was identified. In another study, aberration-corrected TEM captured a nonclassical nucleation process in which Au crystal nucleates via reversible structural fluctuations between disordered and crystalline states at atomic spatial resolution and millisecond temporal resolution.80 This study cleverly used the energy of the electron beam to initiate Au recrystallization from preformed Au nanoribbons to enable real-time monitoring of Au crystallization at the single-atom resolution.
LCTEM has played a critical role in confirming the existence of PNCs in crystallization of inorganic compounds such as NaCl93,159 and CaCO3.109 Several precursors, such as amorphous calcium carbonate (ACC), prenucleation crystals, and embryos, have been captured by LCTEM in CaCO3 nucleation.160 The study found that among 150 prenucleation crystals, only one exceeded Ncritical (∼104 molecules). The lifetime of the prenucleation crystals was independent of their size, which implies the existence of several different prenucleation polymorphs. An important recent work on nucleation and early-stage crystallization was reported by Nakamuro et al. utilizing in situ LCTEM to monitor NaCl crystallization confined to a vibrating carbon nanotube cone.93 In the narrow apex of the carbon nanotube, a transient NaCl nanocluster was repeatedly found to fluctuate between amorphous and semi-ordered states with a sudden appearance of a crystalline nucleus (Fig. 5). LCTEM images revealed a two-step nucleation mechanism that involves multiple nonproductive semi-ordered nanoclusters before the final crystalline nanocluster. After reaching stable nucleation, classical homoepitaxial crystal growth was found to take place stochastically in the vibrating carbon nanotube in NaCl crystallization.
Fig. 5 Nucleation and growth of a NaCl nanocrystal in a vibrating carbon nanotube. (a) Schematic energy diagram of NaCl crystallization according to the classical nucleation theory. (b) LCTEM image of a NaCl nanocrystal. Scale bar = 1 nm. (c) Schematic drawing of the vibrating conical carbon nanotube. (d) Schematic diagram of fluctuating PNCs prior to stable nucleation confirming a two-step nucleation mechanism and subsequent classical homoepitaxial crystal growth. Reprinted with permission from ref. 93. Copyright, 2021 The Authors under CC-BY-NC-ND license, published by American Chemical Society. |
SECM utilizes ultramicroelectrodes as scanning probes for electrochemical characterization of surfaces.161–163 First reported in 1989 by Bard et al.161 the scanning ultramicroelectrode probe moves above a substrate surface in an electrolytic solution enclosed in an electrochemical liquid cell. The electrodeposition process can be monitored in the collection mode by the probe either held at a constant potential or operated during CV scans. Electrodeposition can also be monitored by the tip current change in the feedback mode. The conductive probe allows imaging of contact currents and electrochemical currents in addition to topographical imaging. The spatial resolution of SECM is limited by the size and geometry of the ultramicroelectrode scanning probe.
A technique related to SECM is ECAFM, which offers improved spatial resolution to SECM by combining liquid-cell AFM (LCAFM) with a sharp but nonconductive probe and a three-electrode electrochemical cell. ECAFM and ECSTM, connectively known as electrochemical scanning probe microscopy (ECSPM), enable in situ measurements of electrodeposition on conductive sample surfaces.156 An earlier ECSTM investigation focused on the electrodeposition and dissolution of Ag at the HOPG working electrode.164 The electrical circuit was made of a four-electrode cell in a bipotentiostat mode so that the electrochemical current at a constant potential could be monitored simultaneously with the tunneling current between the STM tip and the sample/working electrode surface. ECSTM captured a single Ag nanoparticle whose morphology is consistent with the 3D island crystal growth mechanism. More recently, ECAFM has been used to simultaneously characterize the size and electrochemical properties of individual Pt nanoparticles on Si though an electroless plating process.157 While STM offers atomic resolution it also has more stringent requirements such as sample conductivity required to generate the tunneling current. In contrast, ECAFM is more versatile, only requiring the sample to be connected to an external potentiostat. Development of commercial high-speed ECSPM will further enable measurements of fast nucleation and crystal growth phenomena at the electrode surface.165 Like all scanning probe techniques effects of the probe need to be considered in interpreting electrodeposition data.
SICM utilizes a nano-sized electrolyte-filled glass pipette as a scanning probe.166–168 In the original nonmodulated mode, ion current is generated by applying a constant potential bias between an electrode inserted into the electrolyte in the pipette and the reference electrode in the bath electrolyte outside the pipette. A topographical image is obtained while the probe scans over the surface through a feedback mechanism similar to that of AFM. More advanced modes have been developed including modulated, pulse, hopping, and hybrid modes.167 SECCM improves on the SICM technique by employing a mobile meniscus containing the electrolytic solution in contact with a working electrode surface. Combined with SEM, TEM, AFM, Raman spectroscopy, and electron backscatter diffraction (EBSD), precise knowledge of the size of the electrochemical cell formed between the nanopipette meniscus and the surface can be obtained.169–173 The Unwin group developed the nonclassical nucleation–aggregative growth–detachment model118 based on the data from their SECCM studies and will be described in more details below.
X-ray diffraction (XRD) is an essential tool for crystal structure determination. For nucleation and early-stage crystal growth in electrodeposition, surface X-ray diffraction (SXRD) or grazing-incidence XRD combined with synchrotron can resolve atomic layer structures of a crystal as well as species near the electrolyte/electrode interface on the solution side such as the Stern layer.158 X-rays scattered at a sharp surface form crystal truncation rods whose intensities can be modelled to yield dynamic information on surface structures such as chemical bond length, adlayer ordering, and surface restructuring when operated in situ. In an in situ SXRD study of Pt electrochemistry, an electrochemical cell was constructed with electrolyte droplet in a capillary in contact with a working electrode to limit the electroactive area of study (improving the quality of CV),174 reminiscent of SECCM. Other advanced XRD methods include energy dispersive SXRD, high-energy SXRD, transmission SXRD, and coherent surface X-ray scattering.158
In situ X-ray can also be combined with spectroscopies such as Raman spectroscopy and molecular dynamics (MD) simulations to further determine the chemical structure of deposited species during nucleation and crystal growth.159,175
Next, we delve into individual studies that shed new lights on electrochemical nucleation and early-stage crystal growth of diverse materials starting with the most studied class of materials – metals.
Fig. 6 STEM images of Au nanoclusters and dynamic interactions with Au nanoparticles during electrodeposition. (a–c) Au nanocluster and nanoparticle interactions during Au electrodeposition in three different areas at growth times of 5, 10, and 30 ms. The electrodeposition potential is −0.5 V vs. SCE. Au nanoclusters disintegrate to provide atoms to a neighboring crystalline nanoparticle. The nanoparticle becomes disordered first and then ordered due to recrystallization. Scale bar = 3 nm. Reproduced from ref. 120 with permission from the American Chemical Society. |
High-speed ECAFM has been used to study Cu electrodeposition.177,178 Nucleation and growth of multilayer Cu islands were observed within seconds after applying the potential step, which grew rapidly. The shape evolution was analyzed quantitatively by determining the island diameter vs. growth time. Here, a change in the growth law was attributed to the transition from hemispherical to planar diffusion. In addition, the experiments revealed a strong increase in the nucleation density with increasing η. Unlike the cases for Au and Ag, these results for Cu electrodeposition do not deviate from classical considerations under a combined reaction and mass transport control.
Jacobse et al. provided insights on the surface roughening process of Pt(111) when subjected to repeated oxidation–reduction cycles (ORCs), by linking, for the first-time, electrochemical measurements to structural information obtained from in situ high-resolution ECSTM.180 Two growth regimes were identified: (1) the nucleation and early growth regime where nano-islands nucleate and grow laterally; and (2) the late growth regime where the nano-islands coalesce and the growth becomes predominately in the height direction. Interestingly, no correlation between the roughness of the surface and the electrochemical signal was observed during the earlier regime while a linear correlation was observed during the later regime. The reason for the nano-islands not contributing to the electrochemical signal during the nucleation and early-growth regime remains unclear. Nonetheless, this approach enables a quantitative correlation between electrochemical characterization and STM data in Pt electrodeposition.
A novel investigation was conducted by the Dick group, which combined water-in-oil emulsion with the ultramicroelectrode technique to allow observations of the electrodeposition of a single Pt nanoparticle on a 10 μm Pt ultramicroelectrode.140 Nanodroplets of an aqueous/glycerol solution containing PtCl62− precursor ions were dispersed in tetrabutylammonium perchlorate (TBAP) and 1,2-dichloroethane (DCE) by ultrasonication. Upon nanodroplet collision with the electrode, the precursor ion was reduced to Pt metal in a four-electron process (Fig. 8A). The small droplet volume ensures rapid precursor ion depletion and permits the observation of reaction-controlled electrodeposition of individual Pt nanoparticles. The current–time traces at low η exhibit a parabolic rising edge, proportional to t2, consistent with electrokinetically controlled crystal growth. The current–time traces at high η exhibit an edge that is proportional to t1/2, indicative of diffusion-controlled crystal growth (Fig. 8B and C). This is the first time the crystal growth rate constant was accurately determined for Pt electrodeposition to be 0.003 cm s−1 from analysis of numerous current–time curves of individual Pt nanoparticle formation events. This method holds potential for measuring kinetic and mass-transfer rates of nanoscale electrodeposition of a broad range of materials and systems.
Fig. 8 (A) Schematic representation of Pt nanoparticle electrodeposition on a Pt ultramicroelectrode by the single nanodroplet method. (Bottom) experimental and simulated amperometric response of (B) electrokinetic-controlled electrodeposition at low η (0.3 V vs. Ag/AgCl), showing a rising edge proportional to t2 and (C) diffusion-controlled electrodeposition at high η (0.0 V vs. Ag/AgCl), showing a rising edge proportional to t1/2. The nanodroplet consists of 50 mM chloroplatinate in 1:1 water/glycerol suspended in 0.1 M TBAP + DCE. Reproduced from ref. 140 with permission from the American Chemical Society. |
The Bard group studied electrodeposition of isolated Pt atoms and nanoclusters (9-atom cluster) on a Bi ultramicroelectrode and characterized the deposited nanostructure by the electrochemical hydrogen evolution reaction (HER).181,182 By combining ultralow concentration of H2PtCl6 in water and the ultramicroelectrode technique a single Pt atom deposition was detected by CV via HER electrocatalytic amplification. A minimum diffusion-limiting current of 55 pA in the steady-state voltammograms points to the presence of a single Pt atom on Bi. This result has important implications for emerging single-atom catalysis. Individual nanoclusters with radii less than 1 nm and Pt nanoparticles of 1–10 nm in size were also studied for their electrocatalytic activity. The HER kinetics were found to increase with Pt nanoparticle radius until reaching a plateau at ∼4 nm.
Macpherson, Unwin, and coworkers pioneered the use of nanopipettes for the study of metal electrodeposition. They used the microcapillary electrochemical cell to study Pd nanoparticle electrodeposition on single walled carbon nanotubes (SWNTs).186 The nanostructures were characterized by current–time tracing and ex situ AFM and field-emission SEM (FESEM) imaging. At low deposition η values (10–50 mV), nucleation takes place preferentially at defect sites on the SWNTs. This feature can potentially be used to identify defect sites on the SWNTs and their chemical nature. At high η, Pd growth takes place on non-defected regions of the SWNT, referred to as random nucleation. The nanoparticle density was found to vary little during electrodeposition on SWNTs, which indicates that Pd nanoparticle nucleation occurs at sub-ms timescale. The experimentally determined nanoparticle size distributions by AFM and FESEM imaging were found to match the calculated sizes from the Scharifker–Hills model (eqn (7)) fit to the current–time traces.
A more recent work by the Unwin group studied Pd electrodeposition directly on a carbon-coated TEM grid from a nanodroplet (avoiding the materials transfer step in a typical TEM experiment).117 A double potential step chronoamperometry method was used to jump the applied potential back and forth from a base potential where no reaction occurs. TEM analysis shows a majority of deposited Pd nanoparticles to be 1–2 nm in diameter. Moreover the results are consistent with the nucleation and aggregative growth model for metal electrodeposition (Fig. 3).13 The first step involves the reversible nucleation and dissolution of discrete nanoparticles. The second step involves the attainment of a critical size after which point nanoparticles aggregate to grow larger. They reported quantitative measurements of aggregate rate constants at 3 different potentials:
kagg (0.05 V) = (7.89 ± 0.01) × 104 s−1 |
kagg (0.035 V) = (9.57 ± 0.03) × 104 s−1 |
kagg (0.025 V) = (8.11 ± 0.02) × 104 s−1 |
The solvent effect on Pd electrodeposition was studied by the McPherson group by changing the mole fractions of water and acetonitrile (MeCN).187Ex situ FESEM and STEM imaging of electrodeposited Pd was carried out on BDD electrodes. From the shift in the potential necessary to initiate Pd deposition, it was determined that the process is kinetically more favorable in the presence of water. Water molecules bind to the acetate ligands and free up the Pd cation for solvation by MeCN and electrodeposition. For equal volumes of water and MeCN, electrodeposited Pd was in the forms of single atoms, amorphous atom clusters, monocrystalline nanoparticles, defected nanoparticles, and more complex nanostructures with the latter three comprising 96% of all morphologies. The defected and elongated nanostructures were attributed to nearby nanoparticle coalescence or possibly to Ostwald ripening. But no evidence of aggregation prior to coalesce was observed. In MeCN rich solution, precursors to Pd nucleation including Pd atoms and amorphous atom clusters were observed similar to those observed in Au electrodeposition.120 The solvation effect of metal precursor ions on electrodeposition warrants more attention in order to fully understand and control electrocrystallization.
Instead of coating the electrode with thin films to promote epitaxial growth of Zn, Shen et al. explored the use of co-solvents such as cyclic tetramethylene sulfone to promote uniform distribution of adsorbed Zn2+ precursor ions.192 Cyclic tetramethylene sulfone molecules preferentially adsorb on the Zn (0001) face compared to water and form oriented dipole arrays, where the low electron density and large steric hindrance slows down Zn2+ adsorption on the electrode surface. This effectively reduces the reaction rate and was shown to improve Zn battery cycle stability. This study provides another example of using co-solvents to control electrodeposition kinetics.
Li et al. applied in situ ECTEM to study early-stage Zn electrodeposition on Au working electrodes.193 The applied potential vs. time traces in galvanostatic electrodeposition captured a nucleation η spike at the beginning followed by a η plateau corresponding to continued growth of Zn dendrites. The ECTEM results are consistent with the classical picture for Zn electrodeposition in a two-stage process, nucleation followed by crystal growth. In stage one, the metastable Zn nanoclusters fluctuate between deposition and dissolution. In stage two, η decreases to a plateau value with stable Zn nucleus growth.
In another recent work, ECAFM captured the initial Zn deposition on polycrystalline Cu electrodes.194 It was determined that preexisting CuO particles on the Cu surface act as nucleation sites for Zn electrodeposition. To better follow and capture the onset of Zn deposition with successively applied small changes in potential, AFM images were recorded at an increased scan rate. The deposited Zn was found to dissolve after reducing the anodic potential with a morphological change from the triangular growth to a rounded shape and was redeposited again after increasing the anodic potential. The results show a reversible Zn deposition process with no significant deviation from classical crystal growth models.
SICM has been used to study lithiation on Sn/Cu films.200 Prior to the lithiation, surface roughness determined from the topographical image of the as-deposited Sn/Cu film matched well with the ion current image, indicating that variations in the ion current were due to the thickness variations in the thin film. Following the lithiation, local areas on the sample displayed an increase in topography attributed to the decomposition of the Li-containing electrolyte at the Sn/Cu film surface. An increase in current at a tall topographical feature and a drop in current around the peripheral regions was attributed to the SEI formation, which obstructs the flow of ions to the underlying Sn/Cu film while inducing a higher rate of lithiation at the protruding area. Tall features in the topographical map sometimes showed decreased ion current, which was attributed to the SEI film blocking the current flow in the Sn electrode.
Ex situ SEM was used to monitor the nucleation of Li metal on Cu substrates by the Cui group and the data were described by the classical nucleation theory.201 The Li nucleus size was found to be proportional to the 1/η (eqn (3)) and the number density of nuclei was proportional to η3. By applying the new knowledge, the researchers achieved improved Li deposit uniformity and particle density, desirable attributes for high-performance Li metal batteries. The authors concluded that more potentiostatic experiments are necessary for understanding Li nucleation mechanism.
In situ ECTEM and STEM have been used by several groups to study Li electrodeposition.202–205 Mehdi et al. applied HAADF-STEM to study Li deposition and dissolution at the interface between the Pt working ultramicroelectrode and the electrolyte solution during 3 charge and discharge cycles of the operando battery cell built into the STEM.205 The primary finding is the existence of hysteresis in Li electrodeposition. The Unocic group applied in situ STEM to directly visualize the early stages of SEI growth and Li electrodeposition on a 30–50 nm thick GC working electrode.202,203 The STEM images show that Li nucleates at the edge of the GC electrode and grows perpendicular to the electrode surface. Li deposits do not have the classical fractal shape exhibited by other metals during dendritic growth. Rather Li deposits tend to be clusters of globular objects. It was shown that Li grows in bursts – after the initial growth to a critical diameter of 400–500 nm a second nucleation burst follows. The globular nucleus shape was also captured by Kushima et al. in an in situ ECTEM study of Li nucleation and early growth on patterned Au electrodes.204 The TEM images show distinct stages in Li electrodeposition: (1) spherical nuclei emerge at the surface of the Au electrode with their size increasing with t1/2 indicating diffusion-limited growth; (2) the growth of Li whiskers at the SEI pushes the initial deposits away from the electrode surface suggesting preferential Li deposition at the Li/Au interface; (3) thickening of the SEI decreases Li metal crystal growth rate; and (4) kink formation on the whisker divides it into two segments under accumulated stress due to the SEI (SEI fracture). A crystal growth model for Li electrodeposition was derived by Thirumalraj et al. by considering both 3D diffusion-controlled instantaneous nucleation and electrolyte decomposition due to the SEI fracture.206 The equation allows the extraction of important kinetic parameters such as diffusion coefficient, number of nucleation sites, and rate constant of the electrolyte decomposition due to SEI fracture during Li electrodeposition by fitting the equation to the experimental current–transient curves.
In a final example of Li electrodeposition, an operando reflection interference microscope (RIM) was applied for real-time imaging of the entire Li nucleation dynamics at the single nanoparticle level (Fig. 9).207 Li electrodeposition starts with progressive nucleation and then changes into instantaneous nucleation. The RIM images show inhomogeneity of the surface electrochemical environment, which impacts Li nucleation and growth on the Cu electrode. The electrode surface heterogeneity causes local variation in η, which varies with the nucleus size (Fig. 9A). In the first 75 s of deposition, nuclei emerge randomly on the sample surface (Fig. 9B). With further deposition from 75 s to 140 s, the surface electrochemical heterogeneity is reduced to promote more uniform growth of the Li thin film (Fig. 9C and D).
Fig. 9 Localized overpotential mapping by RIM. (A) Proof-of-concept curve fitting correlates η with the nucleus size. (B) The nuclei formation time map. (C) The overall η map at the deposition time of 75 s. (D) The overall η map at the deposition time of 140 s. “NaN” represents the region without nuclei. Reproduced from ref. 207 with permission. Copyright, 2021, under CC-BY license, published by Science Advances. |
Electron beam-induced precipitation of ZnO from solution mimics the electrodeposition of ZnO in terms of the crystal growth pathway.211 ZnO particles after nucleation were found to aggregate and grow anisotropically without coalescence according to the in situ TEM investigation.
Tay et al. developed an in situ three-electrode cell capable of in situ monitoring of the nanostructure during electrodeposition with a transmission X-ray microscope.212 A electrochemical liquid cell was designed for simultaneous electrodeposition and synchrotron X-ray imaging. The cell has a bulk volume of 30 ml, larger than the typical TEM liquid cells, and tapers to a width of 2 mm at the bottom where the X-ray beam can pass through two Kapton windows. One window was coated with 10 nm Au as the working electrode. Pt was used as the counter electrode and Ag/AgCl microelectrode was used as the reference electrode. X-ray absorption images were recorded at intervals of 2.2 s during ZnO electrodeposition. Nanorods were found at 5 mM Zn(NO3)2 and nanoplates were produced at 50 mM Zn(NO3)2 at an applied potential of −0.97 V or −0.75 V. The nucleation phase was indicated by an initial current density peak at 25 s for nanorod formation and 90 s for nanoplate formation. Imaging blurring in the X-ray absorption near edge structure (XANES) images suggests the existence of precipitated ZnO particles near the electrode surface in the solution phase (Fig. 10). The presence of swirling, smeared-out particles is prominent in Fig. 9a–d during the first 200 s of electrodeposition and remains visible in the latter time period of 332–336 s (Fig. 9n–p). This X-ray technique provides information on the particle structure in the solution phase, which cannot be easily obtained by in situ ECTEM or SEPM. This information would be useful to obtain a complete picture of the dynamic nature in electrodeposition, for example, confirming the detachment step in nucleation–aggregative growth–detachment,118 a growth pathway likely to occur in many electrodeposition systems with weak deposit/substrate interfaces. The spatial resolution of the XANES technique is at the micrometer scale, which presents a limitation of this method for the study of early-stage crystallization.
Fig. 10 (a–p) The sequence of absorption images that captured every 2.2 s, is organized into 4 time periods: ∼200 s (a–d), 270 s (e–h), 315 s (i–l), and 330 s (m–p). Scale bar = 5 μm. (q) Current density vs. deposition time curve with vertical lines corresponding to the 4 time periods under concentration of Zn(NO3)2 is 50 mM. The deposition potential is −0.75 V. (r) A schematic representation focuses on the area near the working electrode, showing regions (colored green) where ZnO precipitation occurs due to the supersaturation of ZnO. Reproduced from ref. 212 with permission from Royal Society of Chemistry. |
Hauwiller et al. applied real-time EC-LCTEM to study the nucleation and dendritic growth of iron oxide.214 Nucleation was initiated by electron beam illumination, a process resembling electrodeposition. The nanoclusters grew to 4–6 nm in diameter and then changed into nano-dendrites due to supersaturation-induced growth instability. Ex situ energy dispersive X-ray spectroscopy (EDS) elemental maps revealed that the as-grown nanostructure was amorphous, and it transformed into crystalline Fe2O3/Fe3O4 under prolonged electron beam irradiation. This work investigated further the dendritic growth mechanism. The growth was diffusion limited. The dendritic tip curvature was found to be in a linear relationship with the growth rate as predicted by the theory.215 The tip splitting can also be explained by the established analytical model. The growing tip was found to impact the growth of a neighboring tip in close proximity due to precursor depletion. This paper established a strong correlation between fractal growth of iron oxide nano-dendrites and the morphological evolution predicted by the classical crystal growth theories.
Composite coatings of metal oxides can also be made by electrodeposition in direct current, pulsed direct current, pulsed reverse current, potentiostatic, and pulsed potentiostatic modes to improve coating performance.216,217 But the literature on metal oxide composite electrodeposition is largely on the empirical side without mechanistic understanding of the complex process. This area warrants further fundamental research.
More recently a nanoelectrode study was conducted by Blount et al. for the understanding of the nucleation and growth of ionic crystals with CaCO3 as a model system.219 The experiment was carried out on Pt nanoelectrodes in an aqueous solution of NaHCO3 and Ca2+. Pt nanoelectrodes with radii less than 100 nm were fabricated by a bench-top method originally developed by the White group.143 CaCO3 precipitation was initiated by electroreduction of water to create a higher local pH near the electrode surface. The excess OH− converted HCO3− into CO32− and produced a local supersaturation sufficiently high for CaCO3 nucleation. A sudden cathodic current drop in the voltammogram signaled the nucleation and growth of nonconductive CaCO3 on the electrode surface, which blocks the electroactive surface area of the electrode. It is interesting to note that the authors attributed the noisy residual current to the dynamic attachment and detachment of the nucleated CaCO3 crystals, which may share the same origin as Ag electrodeposition on HOPG of the nucleation–aggregative growth–detachment mechanism.118 The characteristic peak current was used to quantify the supersaturation ratio required for nucleation based on the classical nucleation theory. The supersaturation ratio, 220–420, was used to calculate the nucleation energy for CaCO3, 12–14 kJ mol−1.
MoS2 is the one of the most widely explored 2D TMDCs for electrodeposition. Many synthesis methods have been proposed, including the use of different working electrode material, aqueous vs. non-aqueous syste.223–225 Recently, the Kees de Groot group demonstrated the electrodeposition of thin MoS2 layers on a graphene working electrode in a nonaqueous environment, using a three-electrode system.18 The group modified an existing technique and achieved highly anisotropic lateral growth of MoS2 from TiN microelectrode over an insulating surface. In this work, the flat planar graphene working electrode was replaced with TiN sandwiched between two SiO2 layers.226 However, in both cases, the as-deposited film is amorphous and required annealing to achieve crystallinity. Our literature survey found a lack of mechanistic studies of the electrodeposition process, which could be a key to achieve high crystallinity 2D film via electrodeposition.
ECAFM was used to monitor the in situ anodic Cu dissolution and deposition of Cu-MOF HKUST-1 (composed of Cu2+ cations and 1,3,5-benzenetricarboxylic acid (BTC) ligands).233 The real-time imaging together with in situ EC Raman spectroscopy confirmed a previous hypothesis of a 4-step process in Cu-MOF nucleation and crystal growth.234 Firstly, a critical Cu2+ concentration is necessary to initiate Cu-MOF nucleation. Nucleation occurs at defects on the roughened electrode surface. Then the nuclei grow through the classical 3D island mechanism. After the third intergrowth step, the Cu-MOF crystals detach from the electrode surface due to prolonged metal substrate corrosion. No Cu2O or amorphous nucleation intermediates103,235 were observed in this study.
A precursor ion layer was found to promote the nucleation and growth Co-MOF crystals on carbon fiber cloth in cathodic electrodeposition.236 The self-assembled Co2+ ions at the negative potential on the electrode induce the nucleation and growth of Co-MOF in 3D rodlike crystal bundles. The nucleation was found to start from defects on the Co coating. The nucleated crystals grow through the 3D island growth mechanism to several microns and then merge into a continuous layer.
Ward has defined the basic steps in CTC electrocrystallization:242
(1) Electrochemical reaction: The process begins with the electrochemical reduction or oxidation of redox-active precursors at the electrode surface.
(2) Formation of ion pairs and clusters: Following the electrochemical reaction, the generated species interact with counterions present in the solution to form ion pairs. These ion pairs can further aggregate into larger clusters, setting the stage for nucleation.
(3) Nucleation: Nucleation occurs when the clusters reach a critical size and configuration, allowing for the stable formation of a new crystalline phase. This critical size is influenced by the local supersaturation and electrochemical conditions such as the applied current density and potential.
(4) Crystal growth: The crystal grows by further deposition of ion pairs onto the existing nuclei. Growth tends to occur preferentially along specific crystallographic directions depending on the molecular and ionic arrangements, driven by the local electrochemical and intermolecular interactions, such as π–π stacking.
(5) Morphological development: The morphology of the crystals and rate of growth can be significantly affected by the electrochemical conditions, such as the current density, electrode potential, and the presence of impurities, leading to different crystal shapes and sizes.
TTF-Br nanocrystals with width of 30–600 nm and aspect ratios of 20 or higher were electrochemically synthesized on Pt nanoparticle seeds by Favier et al.243 The Pt nanoparticles imposed a nanoconfinement effect on the TTF-Br crystal morphology in which the width of the nucleated TTF-Br crystals varied linearly with the Pt nanoparticle diameter to which TTF-Br is attached. A paper by Mas-Torrent and Hadley demonstrated the electrochemical growth of 1–2 μm long and 200 nm wide TTF-Br microcrystals on microfabricated Au electrodes.244 Ren et al. made significant contributions on the understanding on controlled CTC micro/nanowire electrodeposition on carbon nanotubes.245 They synthesized 12 CTC types of CTC micro/nanowires with combinations of the donors and acceptors shown in Fig. 11. Compared to CTC grown on bulk Au electrodes, CTCs grown on carbon nanotube nanoelectrodes are longer in length and smaller in diameter indicating a nanoconfinement effect imposed by the carbon nanotubes. The CTC morphology was shown to be dependent on the intrinsic structure and properties of the CTCs. CTC crystals based on TCNQ are smaller in diameter than CTC crystals containing TTF, TMTSF, and ET.
Beyond TTF- and TCNQ-based molecular complexes, electrocrystallization of linear chain complexes was studied by Wysocka et al. by electrooxidation of [IrCl2(CO)2]− in dichloromethane containing different tetra(alkyl)ammonium cations (TAA+) on Au working electrodes.246 The experimental chronoamperometric curves were fitted with eqn (7) to determine the nucleation and growth kinetics. Values of the exponent n and the constant K were found to depend on the type of nucleation (instantaneous or progressive) and growth (1D, 2D, or 3D) mechanisms.49–51 The linear double logarithmic fit with a slope close to 1 indicated progressive nucleation followed by 1D growth of needlelike microcrystals. Additional kinetic parameters can be extracted using the following equation:
I = nFAL2kt, | (19) |
Our group has been exploring seed-mediated electrocrystallization as a solution to nanosensor manufacturing by directly synthesizing CTC nanowire sensors in a controlled manner on sensor substrates. This stems from an original discovery made by Mao et al. that nanoparticles of monolayer-protected CdSe and Au are effective seeds for the nucleation of 1D nanorods in solution crystallization via solvent evaporation.247,248 The small nanoparticle seed size (or radius of curvature) was hypothesized to be responsible for the confined nucleation and growth of the molecular crystals.248,249 Subsequently, this hypothesis has been validated in different crystallization systems using compounds including aliphatic carboxylic acids, TTF-Br, and Krogmann's salts.247–252 Electrochemical methods have been used for TTF-Br and Krogmann's salt crystallization on Au nanoparticle seeds and microelectrode patterns.
We further examined the nucleation and growth mechanisms of TTF-Br in nanoconfinement using Pt microdisk electrodes.253 The current increases linearly with time due to the increasing electroactive area upon conductive TTF-Br microwire growth. The linear growth of the CTC microwires on the Pt microdisk electrode enables precise control of the microwire length for device fabrication. In addition, this work demonstrated the use of ultramicroelectrodes in controlling the aspect ratio, orientation, and number of CTC microwire crystals, all are desirable features for scaling up chemiresistive gas sensor technology based on CTC electrodeposition on patterned micro/nanoelectrodes. Currently there are no other methods that enable the control of all these features except with predefined templates.
Recently we demonstrated precise control of TTF-Br electrodeposition on lithographic electrode patterns by varying the shape of the electrodes.24 Triangular shaped Au patterns with tip angles of 5°, 30°, and 90°, and thickness of 5–500 nm were made by photolithography. The electrode geometry was shown to have a strong effect on crystal growth rates. SECM imaging shows TTF-Br to nucleate randomly along the edge of the Au electrode initially. At a later time, crystals near the electrode tip grew more at the expense of the ones away from the tip, which was explained by 3D finite element simulations of the diffusion flux near a triangular tip. The simulation results and the experimental data matched each other, they collectively demonstrate that sharp electrode patterns are more effective in promoting 1D crystal growth of CTCs and their precise placement across source and drain electrodes for making chemiresistive gas sensors. Another interesting finding of this work is the existence of transient PNCs as indicated by multiple small current spikes in the induction period of the current–time transients at a low TTF concentration (Fig. 12a–b(i–iv)) on Au microdisk electrodes. The current peak height, ip, offers an estimate of the corresponding PNC radius, r:254
ip = 2πzFDcr. | (20) |
Fig. 12 (a) The current–time transients of TTF-Br electrodeposition at 300, 280, 270, and 250 mV, respectively, on the Au microdisk electrode in 0.1 mM TTF and 20 mM TBAB MeCN solution. (b) The magnified views from (a) showing individual current spikes represented by solid lines, alongside fits corresponding to zero-order kinetics represented by dashed lines. (c) The nanocluster size histograms at different applied potentials. Reproduced from ref. 24 with permission from John Wiley and Sons. |
The PNC particle size histograms based on eqn (20) are shown in Fig. 11c. Measurements of the largest PNC size provided a lower bound of rcritical to be 29 nm at 300 mV and 75–89 nm at 270–280 mV. This is among the first reports of PNCs in CTC electrodeposition.
In addition to applying electrodeposition for the understanding of nucleation, CTC electrochemistry has also been also revealed a nonclassical crystal growth mechanism in Co-TCNQ electrodeposition on a ultramicroelectrode.25 The presence of Co in the otherwise organic solid enabled us to examine the nanostructure of Co-TCNQ crystals by HR-TEM, which shows distinct crystalline domains with sizes in the range of only a few nanometers. These domains feature lattice fringes arranged along a preferred direction – the long axis of the Co-TCNQ crystal. This system this resembles closely the oriented attachment crystallization pathway observed in nonelectrochemical systems.121–126 Density function theory (DFT) calculations suggest that under high electric fields, the (100) facet of Co-TCNQ can become sufficiently energetic to be the preferential growth facet. A preferred attachment of crystallites in the long axis (π–π stacking) direction of the Co-TCNQ crystal may therefore be the result of applied potential-induced surface energy change.
We have demonstrated the gas sensing capabilities of electrodeposited TTF-Br, Co-TCNQ, and Krogmann's salt nanowires on microelectrode patterns to be effective chemiresistive gas nanosensors for ammonia, nitrogen dioxide, and other gases.25,251,255 We therefore advocate for the use of electrodeposition as a platform technology for precise deposition and assembly of conductive and semiconductive nanowires into electronics for gas sensing and potentially other optoelectronic applications.
In this review, both classical and nonclassical nucleation and crystal growth theories are discussed as they apply to electrochemical systems. The classical nucleation theory provides a knowledge foundation for defining and controlling important parameters such as the critical nucleus size, the number of atoms/molecules in the critical nucleus, and the nucleation rate as a function of the applied overpotential. The Scharifker–Hills model remains valid for the interpretation of current–time behaviors in metal electrodeposition. Classical crystal growth mechanisms include the Frank–van der Merwe 1D growth, 2D monolayer growth, and Volmer–Weber 3D island growth mechanisms have been presented. The relationship between the crystal growth rate of a hemispherical nucleus on an electrode surface as a function of the applied overpotential during electrodeposition is described. At the same time, more evidence emerges of discrete nanoclusters acting as the primary building blocks in electrochemical nucleation. The role of nanocluster building blocks in electrocrystallization is explained both by the generalized electrochemical aggregative growth model and the nucleation–aggregative growth–detachment mechanism. Electrocrystallization is made more complicated by the dynamic nature of the nanoclusters, which can undergo surface diffusion, aggregation, merging, disintegration, and detachment, and reattachment. We expect to see exciting research in further understanding the dynamic nature of electrocrystallization due to rapid advances in both experimental and computational methods with ever-improving sophistication and spatiotemporal resolutions. Research topics for further investigation include multiple nucleation sites overlapping each other, crystal growth after the nucleation stage, electrodeposition of materials beyond elemental metals and multiscale computational method to model the electrochemical driven nucleation and crystal growth processes. Recent work on charge-transfer complexes shows evidence of prenucleation clusters and oriented attachment during their electrochemical crystal growth. Further research may unveil electrocrystallization phenomena and underlying mechanisms already been observed in nonelectrochemical systems such as calcium carbonate and zeolite crystallization or completely new phenomena specific to the electrochemical process. With regard to the decades long debate of classical vs. nonclassical nucleation theory, future studies can attempt to reconcile these two theories. We should not rule out the possibility that, similar to nonelectrochemical systems, the pathway for nucleation is condition dependent.
We reviewed advanced characterization methods that are applicable for the study of electrodeposition. They include high-resolution transmission electron microscopy, various electrochemical scanning probe microscopy methods, and X-ray diffraction and scattering methods. Many of the advanced imaging methods utilize ultramicroelectrodes and patterned nanoelectrodes for simultaneous electrochemical experimentation and real-time imaging of nanoparticle formation during electrodeposition at a precise location. There have been significant advances in method development that extend the capability of the ultramicroelectrode technique significantly by integrating different technologies such as nanopipettes, nanofluidics, liquid metal nanodroplets, microemulsion, and nanoparticle electrocatalysis. This review highlighted significant advances using these new methods to capture electrodeposition structures and dynamics at an unprecedented resolution – down to the single-atom size. For the study of fast electrodeposition kinetics, further development of high-speed cameras and data acquisition specifically for electrochemical systems will be necessary. Further advancement in characterization instruments to achieve higher temporal and spatial resolution, with less stringent sample preparation requirements will no doubt offer deeper insight into the fundamental of nucleation and crystal growth in electrodeposition. It is interesting to note that stable single atoms have been observed in Au, Pt, and Pd electrodeposition on certain electrode surfaces and thus paving the way for single atom electrocatalytic and magnetic applications. Our survey of recent publications did not discover substantial new work on nucleation and crystal growth mechanisms of electrodeposited 2D crystalline materials, e.g., II–V or III–V semiconductors, transition metal chalcogenides, etc. Future studies should explore the microscopic understanding of electrochemically driven nucleation and crystal growth of 2D thin film materials to overcome the problem of poor crystallinity. This may lead to exciting new discoveries and offer opportunity for less energy intensive, simpler, and more scalable synthesis methods for device manufacturing.
The precise control over the morphology, nucleation density, and growth rate remains a major challenge for device manufacturing. This has prevented real-world applications of electrodeposition in microelectronics and optoelectronics, and hindered the advancement of technologies where controlled electrocrystallization is essential. In conclusion, there have been significant advances with many future opportunities to unveil the mysteries of electrochemical nucleation and growth of functional crystals of diverse chemical nature.
This journal is © The Royal Society of Chemistry 2024 |