Electrochemical 3D Printing of Copper

Additive manufacturing (AM), also referred attractively as 3D printing, is a well-established process of creating complex 3D geometries through a layer-by-layer building approach. 3D printing has received considerable attention in automotive, aerospace, and biomedical sectors. Direct metal laser sintering (DMLS) is the most commonly used method for 3D printing of metals which involves selective laser sintering of layers of metal powders. However, the high capital cost, defects in the manufactured components and inability to work with multiple materials, limit 3D printing of metals by DMLS and warrant the development of novel non-laser based 3D printing techniques.

Electrochemical additive manufacturing (ECAM) (electrochemical 3D printing) involves deposition of thin and highly adherent layers of metal with a layer-by-layer approach onto a conducting surface through reduction of metal ions from the electrolyte. It combines the basic principles of electrodeposition and 3D printing. Two different approaches, viz., localized electrochemical deposition (LCD) and meniscus confined electrode (MCE), have been used for electrochemical 3D printing of metals. These methods, however, require the use of expensive piezo-based movement stages and nanopippettes/ultrafine electrodes. The slow rate deposition rate of metals (Cu: from ≈100 nm/s to ≈0.18 μm/s), inability to develop components of complex geometries, porosity and roughness are some of the major limitations. Moreover, majority of the structures fabricated using these methods are simple wire-based architectures. Methods for 3D electrochemical printing of metals with improved deposition rates and capable of printing complex structures are warranted. In this perspective, researchers at Dyson School of Design Engineering and Department of Earth Science and Engineering, Imperial College London, UK, have reported a novel strategy for ECAM using a MCE approach.

The electrochemical 3D printer assembly comprised of a plastic syringe and nozzle (diameter 400 μm) with a porous sponge filled with 1 M CuSO₄ was mounted on a carriage and its movement in the x-, y- and z- directions was precisely controlled using computer controlled stepper motors. Two copper rods suspended in the CuSO₄ electrolyte served as the counter and reference electrodes while a copper plate served as the working electrode. The print head was moved to contact the copper plate and retract back by a small amount. The sufficient back pressure provided by the porous sponge to the hydraulic head enabled the formation of a stable meniscus (localized electrolyte). When a positive potential was applied between the working and counter electrodes, deposition of Cu occurs on the working electrode through reduction of Cu²⁺ ions with a simultaneous replenishment of Cu²⁺ ions from the counter electrode. To create 3D printed structures, the print head was moved in x-, y- and z-directions with varying traverse speed as directed by the 3D model. The various components of the electrochemical 3D printer assembly is schematically represented in Figs. 1 (a), 1(b) and 1(c). Optical image of the 3D printed Cu with the shape of the letters “I”, “C”, and “L” using 1 M CuSO₄ at 4 V at a print head speed of 0.4 mm/s is shown in Fig. 1(d).

Fig. 1 Schematic illustration of the electrochemical 3D printer assembly: (a) Print head set-up; (b) electrode arrangement; (c) print nozzle and sponge in the tip, highlighting how they act during the deposition of Cu; and (d) optical images of the printed Cu structures featuring the letters “I”, “C”, and “L” printed using 1 M CuSO₄ at 4 V at a print head speed of 0.4 mm/s.

Electrochemical 3D printing of Cu dots and lines (lateral print head velocity: 0.4 mm/s) were made using 1 M CuSO₄ as the electrolyte at varying voltages from 1 to 6 V for 1 h. The Cu dots printed at 1 V exhibited a dense structure with a high degree of concentricity. In spite of a faster rate of growth along with a dense structure, those printed at 2 V exhibited a convex shape due to preferential deposition at the center. Formation of Cu dendrites becomes apparent at 3V due to the limitations imposed by the mass transport and the a continuous increase in porosity is observed with a further increase in deposition potential (Fig. 2(a)). Hence, achieving dimensionally accurate structures would be difficult using electrochemical printing of Cu dots at potentials beyond 2 V. Irrespective of the deposition potentials in the rage of 1-6 V, electrochemically printed Cu lines fails to exhibit any Cu dendrites since the relative position of the print head aids mass transport of Cu²⁺ ions or might have assisted mechanical removal of the dendrites (Fig. 2(b)). The thickness of the Cu lines is increased from 3 μm at 1 V to 15 μm at 4 V, beyond which it decreased due to the decrease in deposition efficiency. For a given lateral speed of 0.4 mm/s for 3600 s, Cu lines can be printed for 144 passes over a distance of 10 mm.

Fig. 2 SEM images of (a) single Cu dot; and (b) Cu lines (lateral print head speed: 0.4 mm/s) electrochemically printed using 1 M CuSO₄ at 3-6 V for 1 h

The electrochemical 3D printed Cu dots and Cu lines exhibit a higher hardness than cold worked cast Cu. The Vickers hardness of Cu dots varies from 184 to 196 MPa, Cu lines ranging from 211 to 228 MPa when compared to the hardness of cold worked cast Cu, which varies from 50 to 176 MPa; the smaller the grain size, the higher is the hardness. The electrical conductivity of Cu lines is ranging between 1.31×10⁶ and 6.86×10⁶ S/m, which agrees well with that of nanocrystalline Cu (5.41×10⁶ S/m) but relatively lower than coarse grained Cu.

Electrochemical 3D printing of Cu builds a foundation for future work to achieve high speed deposition as well as printing of multi-metal functional structures. The z-height resolution opens up new avenues for the fabrication of functional electronics such as sensors.

T.S.N. Sankara Narayanan

For more information, the reader may kindly refer: Xiaolong Chen et al., A Low Cost Desktop Electrochemical Metal 3D Printer, Adv. Mater. Technol. 2017, 1700148, DOI: 10.1002/admt.201700148

Achieving Improved Stability of Li Anode through Dendrite Free Li Deposition using Ag Nanoparticles

Lithium metal batteries due to their high energy densities assume significance in portable electronics and electric vehicles. Nevertheless, dendrite formation during deposition, instability of the Li metal interface, and huge volume change are the major limitations that need to be solved for their effective utilization. The uncontrolled dendrite-growth of Li is considered responsible for the low reversibility, short cycle life, internal short circuiting, and safety hazards. Researchers at Department of Materials Science and Engineering, University of Maryland at College Park, USA have demonstrated a rapid Joule heating method to anchor Ag nanoparticles (Ag NPs) on carbon nanofibers (CNFs) to guide seeded nucleation and growth of Li to obtain smooth Li metal anode without dendrite growth.

CNFs prepared by electrospinning served as the host material. They were soaked in silver acetate and rapidly heated by Joule heating setup (Fig. 1(a)). When heated above the melting point of Ag (962 K) for only 0.1 s, the molten Ag gets self-assembled as Ag NPs. The high temperature promotes a strong bonding between Ag NPs and CNFs. The defects in the CNFs constrain the migration of Ag NPs. Rapid quenching of the CNFs seeded with Ag NPs below the melting point of Ag prevents agglomeration of Ag NPs. Fortunately, the CNFs could withstand the  thermal shock and preserved its graphitic structure.

Fig. 1 (a) Schematic of the Joule heating method for coating Ag NPs on CNFs (inset: morphology of CNFs prepared by electrospinning); (b) Digital image of the Joule heating set up. The sample was connected to Cu electrodes and heated by a current pulse in Ar-filled glove box.

The morphologies of Ag NPs on CNFs obtained by Joule heating for 0.05, 0.1, 0.5, and 4 s (Fig. 2) indicate that they are homogenous with an average size of 29–57 nm. The size of Ag NPs show a strong dependence on the thermal shock time; the shorter the time, the lesser the particle size (Fig. 2).

Fig. 2 SEM images of Ag NPs deposited on CNFs by Joule heating method  for (a) 0.05 s; (b) 0.5 s; and (c) 4 s.

The nucleation and growth of Li seeded by Ag NPs on CNFs is schematically represented in Fig. 3(a). Due to the zero nucleation overpotential, selective nucleation of Li occurs on AgNP/CNFs (Fig. 3(c)). Plating of Li is proceeded by alloying of Li with Ag NPs. The strong anchoring of Ag NPs on CNFs guides the formation of a smooth Li coating. During the growth stage, Li from AgNP/CNFs gradually fills the voids between the CNFs, resulting in the formation of an even Li metal anode without dendrite growth (Fig. 3(d)). The ability of the Ag NPs strongly bound on to CNFs to retain itself on the surface of the anode even after stripping of Li (Fig. 3(e)), could repeatedly guide the seeded nucleation of Li. Figs. 3 (f) and 3(g) show the inability of bare CNFs to promote uniform deposition of Li metal, due to the poor wettability of CNFs with Li, thus justifying the beneficial role of Ag NPs.

Fig. 3 (a) Schematic of Li nucleation and growth seeded by Ag NPs on CNFs; (b-g) SEM images: (b) pristine AgNP/CNFs without Li deposition; (c) initial Li nucleation on AgNP/CNFs; (d) Li deposited on CNFs guided by Ag NPs at 1 mA h/cm² of; (e) AgNP/CNFs after the first plating/stripping cycle: (f) bare CNFs without Ag nanoseeds; and (g) Li deposited on bare CNFs

The cycling performance of Li metal anodes using AgNP/CNFs as host (size of Ag NPs ≈40 nm) indicate an exceptional cycling stability at 0.5 mA/cm² for 500 h without short-circuiting with a high Coulombic efficiency of ≈98%. In contrast, self-nucleation of Li resulting in the formation of pillars and dendrites of Li metal dramatically decrease cycling stability of bare CNFs to 100 h. The discharge/charge profiles of Li anode seeded by AgNP/CNFs show a low overpotential (≈25 mV) with a negligible nucleation overpotential at 0.5 mA/cm², which is likely to promote a controlled growth of Li. In contrast, plating or stripping of Li on bare CNFs is accompanied with an initial nucleation overpotential.

The CNFs host modified by Ag NPs effectively regulates the deposition of Li, thus enabling the formation of a smooth Li anode without dendrites. The Li metal anodes developed using AgNP/CNFs exhibits a low voltage overpotential, an exceptional cycling stability and avoids problems due to short-circuiting.

T.S.N. Sankara Narayanan.

For more information, the reader may kindly refer: Chunpeng Yang et al., Ultrafine Silver Nanoparticles for Seeded Lithium Deposition toward Stable Lithium Metal Anode, Adv. Mater. 2017, 1702714, DOI: 10.1002/adma.201702714

Restraining Dendrite Growth in Li Anodes through Vertically Aligned Microchannels Developed on Copper Collectors

Lithium-ion batteries (LIBs) have received considerable attention in portable electronics. Uneven plating/stripping and uncontrolled dendrite growth of Li that could induce internal short circuits and explosion of the battery are the major limitations. Researchers at CAS Key Laboratory of Molecular Nanostructure and Nanotechnology,  Chinese Academy of Sciences (CAS), School of Chemistry and Chemical Engineering University of Chinese Academy of Sciences (CAS) and Beijing Institute of Nanoenergy and Nanosystems, China have developed a novel porous Cu current collector with vertically aligned microchannels (VAMCs), which regulates the current density distributions to prevents dendrite growth of Li metal anodes.

Vertically aligned Cu micro-channels with varying pore radius, pore depth and pore spacing were fabricated using a laser micro-processing system (Fig. 1(a)). The sample of porous Cu with a pore radius of 5 μm, a pore depth of 50 μm, and a pore spacing of 12 μm is designated as porous Cu-5-50-12. The porous Cu current collector with VAMCs due to its large specific surface area and low local current density successfully prevents the dendrite growth of Li. The large surface area of VAMCs enables uniform deposition of Li not only on their surface but also in the microchannels.

For VAMCs with a fixed pore radius of 5 μm, for a pore spacing of 10 μm the current efficiency is lower at locations away from the channels (Fig. 1(b)). When the pore spacing is decreased to 6 and 2 μm, the current efficiency at these locations is increased but the extent of increase is much lower than those experienced in the mouth of the channels (Figs. 1(c) and 1(d)). Since the current density within the microchannels is much larger than that on the upper surface of the porous Cu, preferential nucleation of Li occurs inside the mouth of channels (Fig. 1(e)).

Fig. 1 (a) Schematic of the porous Cu current collectors; (b–d) current density distribution on the surface of porous Cu collectors obtained from COSMOL simulation: (b) Cu-5-50-20; (c) Cu-5-50-16; (d) Cu-5-50-12; and (e) schematic diagram depicting preferential deposition of Li inside the mouth of channels.

For VAMCs with a fixed pore radius of 5 μm, an increase in pore depth increases the current density around the entire mouth of the channel. The current density distribution gradient is inevitable, which helps to accommodate most of the Li inside the channels. Hence, systems having a highest current efficiency in the pores and lowest current efficiency at the locations away from the pores is expected to effectively suppress the Li dendrite formation.

The morphology of Li deposits formed using 1 M lithium bis(trifluoromethane-sulfonyl)imide in 1:1 1,3-dioxolane/1,2-dimethoxyethane as the electrolyte containing 1 wt% LiNO₃ on porous Cu with different pore radii as well as those formed on planar Cu at 3 mA h/cm² is compared (Figs. 2(a)–2(e)). Li deposits on porous Cu-5-50-12 indicate enrichment of Li in the VAMCs (Fig. 2(a)) while an increase in pore radius decreased the ability of VAMCs to restrict Li deposition within the microchannels (Figs 2(b)-2(d)). Formation of Li deposits with a spherical shape could not be observed on planar Cu (Fig. 2e). The voltage profiles of Li deposition on porous and planar Cu current collectors (Fig. 2(f)) indicate a lower overpotential of ≈144 mV for porous 5-50-12 whereas for planar Cu, the overpotential is 280 mV under similar conditions. The lower overpotential values obtained for porous Cu points out a decrease in local current density due to the larger specific surface area of the VAMCs.

Fig. 2 Morphology of Li deposits formed on porous and planar Cu current collectors: (a–d) SEM images of Li deposits formed on the porous Cu with varying pore radii: (a) 5 μm; (b) 7.5 μm; (c) 10 μm; (d) 15 μm; (e) SEM image of Li deposits form on the planar Cu; and (f) Voltage profiles of Li deposition on Cu current collectors.

Galvanostatic cycling measurements performed using symmetrical cells indicate that porous Cu-7.5-50-17 possesses a low voltage hysteresis of ≈20 mV and improved cycling stability even after 300 h. In contrast, under similar conditions, planar Cu exhibits a gradual rise in voltage hysteresis after 50 h.  The cycling stability and CE of porous Cu collectors are also ascertained by cells assembled using commercial lithium foil as the counter electrodes. For porous Cu-5-50-12, the CE of the cell remains stable for 200 cycles with an average CE of 98.5%. In contrast, for planar copper, the CE of the cell reaches 98.2% after 13 cycles and exhibits a rapid decline to 68.1% after 81 cycles, with an average CE of 94%.

Full cell galvanostatic cycling of Li/LFP cells performed using planar Cu as well as porous Cu-5-50-12 anodes indicate that cells with porous Cu anode exhibits good cycling behavior, delivering a capacity of 134 mA h/g (Fig. 3(a) with a capacity retention of ≈90% after 100 cycles (Fig. 3(b). The cell with planar Cu anode exhibits a larger polarization effect of 177 mV (Fig. 3(a)) and a poor capacity retention of 80% after 100 cycles (Fig. 3(b)).

Fig. 3 (a, b) Electrochemical performance of Li/LiFePO₄ cells with (a) porous Cu-5-50-12 anode; and (b) planar Cu anode; and (c) cycling performance of the Li/LFP cells

The ability of porous Cu current collector with VAMCs to control the dendrite growth of Li is mainly due to the larger specific surface area, lower local current density, restriction in lithium volume change, lower charge transfer resistance, and high Li⁺ transport properties in the cell. The porous structure served as a cage for Li, thus accommodating large amounts of Li in the pores.

T.SN. Sankara Narayanan

For more information, the reader may kindly refer: Shu-Hua Wang et al., Stable Li Metal Anodes via Regulating Lithium Plating/Stripping in Vertically Aligned Microchannels, Adv. Mater. 2017, 1703729, DOI: 10.1002/adma.201703729

Cyro-electron microscopy reveals the secret of what is limiting the life time of lithium ion batteries

In lithium-ion batteries (LIBs), during charge-discharge cycles, both Li metal and organic electrolyte become unstable. The continuous deposition and stripping of the Li metal results in a large structural change while dendrite growth worsen the situation. Decomposition of the organic electrolyte at the anode leads to the formation of a solid electrolyte interphase (SEI) layer consisting of organic and inorganic components. The changes in the SEI layer as well as the growth directions of the dendrites could alter the efficiency of the system. Since the Li containing electrode material, organic electrolyte and the SEI layer are chemically reactive and sensitive to electron-beam irradiation, it is hard to characterize them using transmission electron microscopy. These attributes poses difficulty in identification of failure mechanism of LIB.

Researchers at the Stanford University, USA, ShanghaiTech University, China, Universität Erlangen–Nürnberg, Germany, National Accelerator Laboratory, USA have developed a cryo-transfer method (Fig. 1) based on cyro-electron microscopy (cyro-EM) and demonstrated that it would be possible to obtain  atomic-resolution images of sensitive battery materials in their native state (Yuzhang Li et al., Science, 358, Issue 6362, 2017, pp. 506-510).

Fig. 1 Preserving and stabilizing Li metal by cryo-transfer method: (a) Li metal dendrites are electrochemically deposited directly onto a Cu TEM grid and then plunged into liquid N₂ after battery disassembly; and (b) The specimen is then placed onto the cryo-TEM holder while still immersed in liquid nitrogen and isolated from the environment by a closed shutter. During insertion into the TEM column, temperature is not increased > –170 °C, and the shutter prevents air exposure to the Li metal.

The cryo-TEM and cyro-SEM images of the electrodeposited Li metal dendrites (Figs. 2(a) and 2(b)) reveal that the dendrite structure is preserved during the cryo-transfer method. Time time-lapse images obtained under constant electron-beam irradiation (~50 e Å^–2 s^–1) in cryogenic conditions (Figs. 2(c), 2(d) and 2(e)) show no signs of damage in the dendrite morphology even after 10 min. The lack of reactivity of the Li metal with liquid N₂, helps the dendrites to retain their electrochemical state so that the relevant structural and chemical information could be obtained. The inferences made in this study reveal that in carbonate-based electrolyte, Li metal dendrites grow as single-crystalline nanowires along three primary growth directions: <111>, <110>, and <211> (Figs. 2(f), 2(g) and 2(h)) with 49% growth along the <111> direction, followed by 32% along <211> and 19% along <110> direction. In spite of growing as single-crystalline nanowires along a linear direction, the Li metal dendrites often change their growth directions.

Fig. 2 (a) Cryo-TEM and (b) Cryo-SEM images of Li metal dendrites depicting that the morphology is preserved by the cryo-transfer method; (c to e) time-lapse images of  Li dendrite; (f to h) growth of Li metal dendrites along: (f) <111>; (g) <110>; and          (h) <211> directions.

The methodology described in this work is likely to provide a complete understanding of the failure mechanisms in high-energy batteries.

T.S.N. Sankara Narayanan

 For more information, the reader may kindly refer: Yuzhang Li et al., Atomic structure of sensitive battery materials and interfaces revealed by cryo–electron microscopy, Science, 358, Issue 6362, pp. 506-510.