Electrochemical Insights into Copper Electrodeposition on Nd2Fe14B Grains: A Proof-of- Concept Study

Journal of The Electrochemical Society OPEN ACCESS Electrochemical Insights into Copper Electrodeposition on Nd2Fe14B Grains: A Proof-ofConcept Study To cite this article: Črt Saksida et al 2025 J. Electrochem. Soc. 172 022505 View the article online for updates and enhancements. You may also like Improvement in coercivity, thermal stability, and corrosion resistance of sintered Nd–Fe–B magnets with Dy80Ga20 intergranular addition Beibei Zhou, , Xiangbin Li et al. - Pre-orientation sintering: an alternative process to prepare high-performance anisotropic Nd–Fe–B HDDR magnets Zexuan Wang, Lizhong Zhao, Anjian Pan et al. - Effect of reduction-diffusion time on microstructure and properties of Nd-Fe-B nanoparticles prepared by low-energy chemical method Wenli Pei, Xiaohan Zhang, Qinyu Meng et al. - This content was downloaded from IP address 193.2.4.235 on 05/05/2025 at 12:53 Electrochemical Insights into Copper Electrodeposition on Nd 2 Fe 14 B Grains: A Proof-of-Concept Study Črt Saksida, 1,z Mihaela Rebernik, 1,2 TomažTomše, 1 László Péter, 3 Carlo Burkhardt, 4 Laurence Schieren, 4 Zoran Samardžija, 1 and Kristina Žužek 1,2 1 Jožef Stefan Institute, Department for Nanostructured Materials, SI-1000 Ljubljana, Slovenia 2 Jožef Stefan International Postgraduation School, SI-1000 Ljubljana, Slovenia 3 Institute for Solid State Physics and Optics, HUN-REN Wigner Research Centre for Physics, Budapest, 1525, Hungary 4 Institute for Precious and Technology Metals, Pforzheim University, Pforzheim 75175, Germany This study explores Cu electrodeposition from a near-neutral sulphate bath onto Nd-Fe-B bulk and powder electrodes. The former served for the preliminary electrochemical tests, while the latter was used for Cu coating of the corrosion-sensitive powdery raw material. Cyclic voltammetry established the potential intervals for Cu deposition (at least −0.1 V and below) and the Nd-Fe-B oxidation (above −0.5 V). Cu electrodepositions were performed on both electrodes in potentiostatic mode for 30 s. Scanning electron microscopy/energy-dispersive spectroscopy showed that Cu deposited at high overpotentials (−1.05 and −0.5 V) had a dendritic structure mainly due to mass transport limitations. A chronoamperometric study on Nd 2 Fe 14 B powder electrodes at −0.25 V resulted in a positive current, indicating the Nd-Fe-B oxidation dominance. At −0.5 V, the current remained negative, but showed diffusion limitations. The latter was improved by using ultrasonic agitation, which resulted in a higher total negative charge and more uniform Cu deposits on Nd 2 Fe 14 B grains. Cu-coated Nd₂Fe₁₄B grains showed a mass magnetization decrease from 137 to 127 emu g −1 , corresponding to a ∼9% Cu mass increase determined via gravimetry. The study demonstrates successful Cu electrochemical deposition with no magnetization loss beyond the paramagnetic Cu phase, paving the way for grain-boundary engineering of novel Nd-Fe-B magnets. © 2025 The Author(s). Published on behalf of The Electrochemical Society by IOP Publishing Limited. This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 License (CC BY, https://creativecommons.org/ licenses/by/4.0/), which permits unrestricted reuse of the work in any medium, provided the original work is properly cited. [DOI: 10.1149/1945-7111/adb339] Manuscript submitted October 29, 2024; revised manuscript received January 10, 2025. Published February 19, 2025. Supplementary material for this article is available online The transition to a low-carbon and energy-efficient society by 2050, as envisioned by the EU’s Green Deal, requires resourceefficient solutions to achieve an 80% reduction in greenhouse-gas emissions. Highly efficient electric motors, essential for e-mobility and eco-friendly power generation, rely on rare-Earth transitionmetal permanent magnets (PMs) like neodymium–iron–boron (Nd-Fe-B). 1 Nd-Fe-B PMs play a crucial role in technological advancements and energy conservation. These magnets are widely used in applications; however, they still encounter challenges related to low coercivity at elevated temperatures, compromising their thermal stability —especially in e-mobility and high-efficiency motors. 1,2 To date, the underlying mechanism behind their low coercivity, relative to their high anisotropy field, remains unresolved. 3 In addition, given the criticality of rare-Earth elements, as highlighted by the EU’s list of essential raw materials, 4 the focus must shift towards developing magnets that maximize resource efficiency. Studies of the coercivity of Nd-Fe-B magnets have considered a demagnetization mechanism based on the nucleation of the reverse domains. 5 However, because of the experimental discoveries that have pointed to the ferromagnetic nature of the Nd-rich “subphase” or “grain-boundary phase,”a mechanism of domain-wall pinning has been proposed. 2 The latter emphasizes the significance of controlling the structural and magnetic properties at the interfaces between the main matrix Nd 2 Fe 14 B phase and the surrounding subphase(s) to enhance the coercivity. 6–8 Trace amounts of copper (around 0.1 atomic percent) positively influence the coercivity. Several annealing processes have successfully enhanced the coercivity, highlighting the importance of copper near the interface. 9–12 In annealed Nd-Fe-B materials, local-composition analyses revealed the formation of a thin neodymium-rich (Nd-rich) amorphous layer along the grain boundaries. Additionally, Cu segregates at these interfaces, enveloping the Nd₂Fe₁₄B grains with a Cu-enriched layer. 8 Optimal coercivity thus depends on the amount of Cu and how it segregates, factors which are influenced by the copper concentration and the applied thermal treatment. 13 Based on these findings, it is likely that the presence of Cu near the interface directly improves the coercivity. A first-principles study investigated a Cu-doped Nd 2 Fe 14 B-NdO x system to explore Cu’s role in enhancing the coercivity in annealed Nd-Fe-B magnets. It was shown that Cu replaces the Fe at the interface, leading to a 40% improvement in the magnetic anisotropy of the Nd atoms. 14 Introducing Cu around the interface between the main matrix Nd 2 Fe 14 B phase grains and the subphases is expected to prevent the domain walls from penetrating the main-phase grains, thus effectively contributing to an increase in the coercivity. Theoretically, the Nd 2 Fe 14 B phase contains 26.7 wt% of rare Earths including Nd, Pr, Ce Tb and Dy and others, 15,16 where the Nd contributes to a majority to around ∼23 wt%. 17 The ∼72 wt% of transition metals are mainly presented by iron (Fe), although small amounts of cobalt (Co) are up to 1 wt%, are usually added to enhance the Curie temperature. 17 For simplicity, the term Nd₂Fe₁₄B is used as a general formula to represent the hard-magnetic tetragonal phase. However, commercial Nd-Fe-B sintered magnets have compositions that are much richer in rare Earth, up to 32 wt%, to ensure the liquid phase’s formation for sintering 18 (the latter being Nd-rich). However, this Nd-rich phase is prone to oxidation, jeopardizing the Nd-Fe-B magnet’s long-term stability and requiring additional costs for coating. 18,19 In this study we present an electrochemical analysis and propose a new approach to creating a copper-based grainboundary phase in Nd-Fe-B magnets. Since Cu is usually added to the Nd-Fe-B-type alloy in trace amounts (around 0.1 atomic percent), this novel phase is designed to be more resource efficient and corrosion resistant than the current Nd-rich grain-boundary phase. Additionally, considering the studies mentioned above, incorporating a Cu-based grain-boundary phase could provide controlled magnetization reversal and thereby improve the coercivity. For this purpose, Nd-Fe-B magnets were treated with the hydrogen-assisted process of magnetic scrap (HPMS) that was developed at the University of Birmingham, and involves placing z E-mail: [email protected] Journal of The Electrochemical Society, 2025 172 022505 end-of-life Nd-Fe-B permanent magnets in a reactor evacuated and filled with hydrogen at controlled pressures and ambient temperature. The Nd-rich grain boundaries react with hydrogen to form NdH ∼2.7 , while the Nd₂Fe₁₄B matrix phase accommodates dissolved (interstitial) hydrogen atoms, causing a crystal lattice volume expansion up to 5%. This leads to embrittlement and separation of the Nd₂Fe₁₄B grains and the Nd-rich phase into coarse, friable aggregates. Once hydrogen consumption stabilizes, the material is partially degassed and removed, yielding powders ranging from fine particles (<10 μm) to aggregates several mm in size consisting of the Nd₂Fe₁₄B grains and the Nd-rich phase, which require further milling to break down the agglomerates. 20–22 Those were later stripped chemically for our study to remove the Nd-rich grainboundary phase. 17,23 The process resulted in a Nd₂Fe₁₄B main matrix phase on which the Cu was deposited. Electrochemical deposition is a method of choice in our work due to its experimental and economic feasibility for coating powder particles with metals and alloys. However, several aspects need to be considered when depositing Cu on Nd-Fe-B. Neodymium is a lanthanide; it possesses an exceptionally low standard redox potential of −2.3 V vs SHE (Nd 3+ /Nd). This characteristic makes it highly sensitive to common environmental oxidants like O 2 and H 2 O, necessitating its storage and handling in an argon atmosphere. The Fe 2+ /Fe system also has a negative standard redox potential of −0.44 V vs SHE. The electrolytic bath for Cu plating also contains another oxidizing agent, Cu 2 ⁺. It is possible, though undesirable, for an electron exchange to occur between the Nd-Fe-B and the copper ions, which have a positive standard potential of +0.34 V vs SHE (concerning the Cu 2+ /Cu redox pair). This could result in the electroless deposition of copper. The oxidation, i.e., the corrosion behaviour of Nd-Fe-B sintered magnets, was studied in. 24,25 The Ndrich phase in the Nd-Fe-B system is the most susceptible to oxidation due to its high neodymium content. This makes it more reactive and prone to get oxidized compared to other phases in the system. 26 The latter was exploited for the selective leaching of the Nd-rich phase and the separation of the matrix Nd 2 Fe 14 B main phase, either electrochemically 27 or chemically, 22,23 and was used as a precursor material for our study. The literature on the electrodeposition of metals onto Nd-Fe-B powders is scarce, and mainly conducted on nanocrystalline systems that are more corrosion resistant. The literature reports the successful electrodeposition of Fe, Cu, Zn, Ni, and Sn onto melt-spun nanocrystalline Nd-Fe-B powders for further processing to bonded 28 or hot-pressed magnets. 28 However, they were all carried out in a galvanostatic mode without assessing the complete electrochemical behaviour. Generally, Cu electrodeposition on an industrial scale is based on established electrolytic baths, including either complexed alkaline cyanide and pyrophosphate systems or acid sulfate and fluoborate simple ion systems. 29 Among these, the acid sulfate bath is the most widely used in industry due to its ability to produce bright, smooth deposits, its low cost, and its effectiveness at high current densities. It typically uses a high concentration of H₂SO₄(1.8 M) to increase the conductivity and remove the oxide layers on the electrode surface. As Nd-Fe-B powder is notoriously susceptible to oxidation (i.e., dissolution or surface oxidation in acids), Na₂SO₄was employed in our study as a substitute for sulfuric acid to provide a near-neutral environment. Copper electrodeposition from solutions of divalent non-complexed Cu 2+ ions is known to take place via the formation of the Cu + intermediate, its formation from the Cu 2+ ion being the rate-determining step of the entire process. 30 It is also known that chloride ions can tune the facet distribution of the deposited Cu metal. 31 Chloride ions facilitate the formation of the cuprous intermediate from the cupric ions. 32 Even if the Cl − concentration applied falls into the mM range, no mass transport limitation was found by Gabrielli et al. in relation to the impact of the chloride ions. Rather, Cl − ions seem to reside at the Cu surface permanently as adsorbed species, hence forming a monoatomic electron transfer catalyst layer. 32 Therefore, we applied NaCl in a low concentration to make the reaction as unhindered as possible. Here, we provide a complete study of the electrochemical reactions during the electrodeposition of Cu on Nd₂Fe₁₄B main matrix phase, single crystalline powders. The main objectives of the study are to promote Cu 2+ reduction, to reduce the rate of Nd-Fe-B oxidation, and to obtain a uniform morphology for the Cu-based deposit to set the guidelines for a possible novel, Cu-based, grainboundary phase in conventional Nd-Fe-B permanent magnets. Experimental Nd 2 Fe 14 B grains (i.e., Nd-Fe-B powder).—End-of-life, windturbine magnets served as the starting material. Before undergoing the HPMS treatment, the magnets were thermally demagnetized and sandblasted to remove their coatings. These magnets were converted into a hydrided powder using the HPMS process according to. 19,20 The conditions used were 3 bar of hydrogen, the pressure of which was maintained until it stabilized. The resulting powder was coarse but friable. To reduce the particle size, the powder was subjected to ball milling. For each batch, 120 grams of powder were milled with 50 steel balls (10 mm in diameter) at a frequency of 35 Hz, with three milling cycles of 10 min each, separated by a 10 min cooling period. The powder was further degassed (heating rate =2°C min −1 , max temperature =450 °C, holding time =20 min, under vacuum) and stored in a glove box under argon. This powder was then used as the base material for further experiments of selectively leaching the Nd-rich phase, performed in 0.5 mol l −1 citric acid solution for 15 min. After the set time, the acid solution was strained off, and the powder was rinsed three times with deionized water, strained with magnetic decanting, and rinsed 3 times with water and once with ethanol. The ICP-OES of the average values from the 4 batches are listed in (Table S1 in the Supplementary Information (SI)), showing the powder consists of 23.00 wt% of Nd, 0.29 wt% Pr, 4.20 wt% Dy, 1.10 wt% B, 71.90 wt% Fe, 0.48 wt% Al, 0.06 wt% Cu, 0.02 wt% Ga, 0.01 wt% Sn and 0.01 wt% Si. The leaching experiments were set to achieve the stoichiometric composition of the Nd 2 Fe 14 B with the ratio between the wt% of Fe (72.30 wt%) and total wt% of the rare Earth content (26.70 wt%) to be 2.71. The leaching procedure resulted in a Fe and total rare Earth content ratio of 2.59, closely approximating the target stoichiometry, confirming the presence of Nd₂Fe₁₄B phase grains. The resulting powder was examined with SEM/EDS, which confirmed that the powder consists of Nd 2 Fe 14 B grains with an average particle size of ∼10 μm as shown in (Fig. S1 in the SI), hereafter denoted as Nd-Fe-B powder. The electrochemical cell.—The electrochemical cell was a 50 ml beaker equipped with three electrodes: a circular Pt-mesh was used as a counter electrode and Ag/AgCl from HANNA instruments as the reference electrode. The working electrode was varied for the different experiments. All the potentials in this work are referenced vs Ag/AgCl. Working electrodes.—Cu-plate.—A thin copper plate was polished with sandpaper, washed, and dried. The plate was then used in experiments to investigate its interaction with the electrolyte and Cucontaining electrolyte solution. Bulk Nd-Fe-B electrode.—A special electrode was constructed to study the electrochemical reactions on the surface of the Nd-Fe-B. An image of such electrode can be found in Fig. S2b in the SI. The Nd-Fe-B powder was pressed into lumps and attached to a copper tape. This was immobilized with epofix and then polished. This resulted in a polished Nd-Fe-B grains’cross-section and ensured electrical wiring while keeping the magnet and copper tape sealed in plastic—separated from the solution. This electrode was used for all the voltammetric experiments. It was stored in the argon atmosphere and was polished before each experiment with 5 μm sandpaper. Journal of The Electrochemical Society, 2025 172 022505 Nd-Fe-B loose powder electrode.—The working electrode for the electrodeposition consisted of a copper tape to which 20–30 mg of Nd-Fe-B powder was attached using an external permanent magnet (see Figs. 1and S2a in the SI). During the deposition, the external magnet was rotated, which caused the magnetic powder to move correspondingly, and ensured that most of the grains were exposed to the solution. A video of this rotation is presented in the SI. Ultrasonic agitation was applied to aid the mass transport using a 30-kHz ultrasonic bath. During the brief experiment duration (30 s of deposition and up to 5 min of handling), the sample was exposed to air and water. After the Cu deposition, the powder was washed thoroughly with deionized water and isopropanol, vacuum dried, and stored in the glove box. For coating the powder used for measuring the magnetic properties the experiment was scaled up to approximately 1 g of powder per deposition (also using a larger Cu-tape and a stronger external magnet). Since the electrode potential is the parameter providing the driving force of a process and is indicating the stability regime of the phases produced, the experiments performed with the same cathode potential but with dissimilar loadings can rightfully be assumed to be qualitatively alike. Electrolytic baths.—The solution used in this investigation was a “neutralized”sulfate bath, with Na₂SO₄employed as a substitute for H₂SO₄. The aqueous electrolytic bath included CuSO₄·5H₂O (CARLO ERBA), anhydrous Na₂SO₄(Sigma Aldrich and Thermo Scientific), NaCl (Sigma Aldrich 99.99%) and FeSO₄·7H₂O (Sigma Aldrich). All the reagents were of analytical grade and used without further purification. Table Ishows the ion concentrations in the different baths used in the following experiments. Cell notation.—To describe the system being investigated in a particular experiment, the following cell notation is used: Working electrode | Solution composition | Counter electrode. For example, Nd-Fe-B-bulk|Cu 2 ⁺,Na + ,SO 4 2− ,Cl − ,H 2 O|Pt means a bulk Nd-Fe-B electrode immersed in the copper-containing solution, with a platinum mesh serving as the counter electrode. Experimental conditions.—The experiments were always performed at 30 °C ± 2 °C. The platinum mesh was soaked in nitric acid, washed and torched in a butane flame for 30 s, each time before use. A Gamry Reference 600 potentiostat/galvanostat equipped with PHE200 software was used for all the measurements and depositions. Cyclic voltametric (CV) studies investigated the potential region from −1.6 V to +0.7 V. The scan rate was 50 mV s −1 in all the experiments. The scan was initiated immediately after soaking the working electrode in solution without a stabilization time to keep the oxidation at the lowest level possible. To ensure consistency, a CV study for the Nd-Fe-B-bulk|Cu 2 ⁺,Na + ,SO 4 2− ,Cl − ,H 2 O|Pt system was additionally performed in a de-aerated solution. As there was no significant difference in the voltammogram shape, we concluded that dissolved O 2 does not participate in the reactions to an appreciable extent. For the Cu potentiostatic electrodeposition, experiments were conducted on both bulk and loose powder Nd-Fe-B electrodes. The deposition potential was −0.25 V, −0.5 V or −1.05 V vs Ag/AgCl, with a deposition time of 30 s in all experiments. Electrogravimetry.—To analyse the copper content, the coated powder was weighed, dissolved in nitric acid and the solution was placed in an electrolytic cell with a magnetic mixer, consisting of a platinum wire counter, Ag/AgCl reference and platinum-mesh working electrode. Copper was deposited galvanostatically, at a current I=−600 mA until the appearance of a potential jump and the occurrence of hydrogen bubbles. A bright copper deposit was obtained, and its mass was determined from the weight difference of the mesh. SEM/EDS and VSM evaluation.—For studying the powder morphology using SEM, the powder particles were mounted on conductive adhesive carbon tape, sputtered with an 8-nm layer of carbon, and examined without further modification. To analyse the particle cross-sections, the powder was embedded into an epoxy resin, allowed to cure overnight, and then ground using SiC papers with grain sizes down to 5 μm, followed by final polishing with diamond paste of 3 μm and 0.25 μm. For the SEM analysis of the copper deposits on the bulk Nd-Fe-B electrode, the entire electrode was encapsulated in epoxy and cut perpendicular to the deposition surface. After removing the copper tape and the magnet, the sample was polished using the same procedure as for the particle crosssection analysis. Microstructural characterization of the particle morphology and/or polished particles was performed using a fieldemission-gun scanning electron microscope (FEGSEM) Verios G4 HP (Thermo Fisher Scientific Company, 5350 NE Dawson Creek Drive, Hillsboro, Oregon 97124, USA) equipped with an energydispersive X-ray spectroscopy system (EDS) AZtec with 65 mm 2 silicon-drift detector (SDD) UltimMax (Oxford Instruments NanoAnalysis, Halifax Road, High Wycombe HP12 3SE, UK). Samples were analysed at accelerating voltages of 18 kV and 7 kV Figure 1. Electrochemical cell with Pt mesh as counter electrode (CE), Ag/AgCl reference electrode (RE) and copper tape with Nd-Fe-B powder as working electrode (WE). Table I. Electrolytic bath composition. [Cu 2+ ]/M [SO 4 2− ]/M [Cl − ]/mM [Fe 2+ ]/M Copper containing solution 0.25 2.05 1.41 0 Copper-free solution 0 2.05 1.41 0 Iron(II)-containing solution 0 2.05 1.41 0.25 Journal of The Electrochemical Society, 2025 172 022505 and a beam current of 0.2 nA. Electron micrographs were recorded using secondary-electron imaging (SE) to observe the particle topography/morphology and using backscattered electron (BSE) imaging to enhance the material contrast, i.e., compositional atomic number Z-contrast and phase composition. EDS spectra for the compositional analyses were acquired to accumulate a large number of X-ray counts of the order of magnitude 10 6 counts per spectrum, thus ensuring reliable elemental identification and quantification with high accuracy. Quantitative compositional EDS analysis was done using an AZtec internal standards database with the XPPPhiRoZ quantitative algorithm, which is included in the software package. In the analysed EDS spectra we detected only the elements present in the Cu-coated Nd-Fe-B system, and no other impurities were found. Compositional analysis.—For precise quantitative analysis, ICPOES measurements were performed. The ground powders were digested in aqua regia (HNO₃:HCl, 1:3) using a “Mars 6”microwave digester (CEM GmbH) and analyzed with an “iCAP 7000 Series” ICP spectrometer (Thermo Fisher Scientific Inc.). Magnetic characterization.—The magnetic measurements of the powdered Nd-Fe-B samples coated with Cu at −0.5 V for 30 s were performed using a vibrating-sample magnetometer (VSM LakeShore 8600) with magnetic fields in the range ±1000 kA m −1 . Results and Discussion Cyclic voltammety study.—ACVoftheCuplateimmersedin copper-containing solution denoted as [Cu-plate|Cu 2 ⁺,Na + ,SO 4 2− ,Cl − , H 2 O|Pt system] was taken and is shown in Fig. 2a. The potential was swept from 0 V to −1.6 V, then to +0.5 V and finished at 0 V with a scan rate of 50 mV s −1 . In Fig. 2a, an increasing reduction current was measured below −0.1 V, and reached the first peak (C Cu 2+ )at−0.36 V. We attribute it to the process of Cu 2+ reduction to elementary copper. Similar to other studies with suflate-based baths with minor Cl – concentration, 33 the peak immediately after the start of the Cu deposition (here, in the −0.2 to −0.5 V interval) is relatively large compared to the further diffusion-limited current density. Given the low concentration of Cl – relative to Cu 2+ , the intermediate formation of [CuCl 2 ] – does not appear as a distinct step before the two-electron reduction of Cu 2+ ions, unlike in more concentrated solutions of chloride salts. 34 In the further section of the cathodic sweep, Cu deposition is controlled by the mass transport. The current rise (C H2O ) around −1.4 V is due to hydrogen evolution, and the appearance of bubbles is visually observed. The potential of water decomposition occurs at a very negative potential because Cu is not a good electrocatalyst (as compared to, e.g., platinum), and the solution is not acidic. 35 When the potential scan is reversed, a higher current is measured, because H 2 bubbles provide some agitation, thereby aiding the mass transport and enabling more copper ions to be reduced. As the bubble effect vanished by returning the electrode potential to a value where hydrogen is no longer evolved (E > ∼−1.2 V), the Cu-deposition current returns to the value seen in the cathodic-going scan just prior to the onset of water decomposition. At potentials above zero volts, the increasing anodic current (A Cu )is measured for the whole investigated potential range up to 0.5 V. We attribute it to the oxidation of the deposit and the electrode material (both made of metallic copper). The positive-going and the negativegoing arms nearly coincide. There is no difference in the reactant supply, because a solid copper tape is being oxidized. To further investigate this oxidation, a copper-free solution was prepared, and 2 CV cycles were taken and are presented in Fig. 2b. Figure 2. CV of the Cu-plate in copper-containing solution (a) and copperfree solution (b). Figure 3. CVs of the Nd-Fe-B-bulk electrode (a) in copper-free (light grey) and copper-containing solution (black), started at 0 V, and (b) in iron(II)- containing (dark grey), copper-free (light grey) and copper-containing solution (black), started at −0.6 V. Journal of The Electrochemical Society, 2025 172 022505 The potential was swept from 0 V to −1.6 V, then to +0.7 V and finished at −1.6 V. When the potential is first swept in the negative direction (black solid line), starting at 0 V, no reduction current is measured except for hydrogen evolution (C H2O ) at potentials lower than −1.2 V. Then, at potentials above zero volt, oxidation (A Cu ) occurs, but we cannot describe it with a single chemical reaction. In this region the previously transparent solution takes on a pale-blue colour indicating the formation of disolved Cu 2+ and also the metal electrode begins to change its colour. It becomes darker (black) but also contains orange-brown colours suggesting the formation of solid copper oxides/hydroxides. Then, when the potential is again swept in the negative direction (grey dashed line) two reduction peaks occur. The first one (C Cu 2+ )at−0.3 V indicates the reduction of hydrated, dissolved Cu 2+ ions to elementary copper and is found at the same potential as in a copper-containing solution (Fig. 2). We attribute the peak (C CuO )at−1.05 V to the reduction of the adherent solid layer composed of couprous/coupric oxides or hydroxides since it does not appear if the copper is not previously oxidized. This is in good agreement with, 36 who studied copper redox behaviour in NaOH solution and attributed the peak at approximately the same potential to the reduction of copper(I) oxide and copper(II) oxide to metalic Cu. Publications dealing with the electrochemical oxidation of Cu in neutral or alkaline media containing chloride ions report the formation of various oxides, hydroxides and layers containing chloride salts, 37,38 sometimes in a varying ratio as a function of ripening time. 39 It is agreed that the intense electrochemical oxidation of Cu might lead to a slowly dissolving CuCl layer that can decelerate the electrochemical processes. Figure 3a shows two CVs of the bulk Nd-Fe-B electrode scanned in both copper-containing solution (black) and copper-free solution (light grey). The potential was swept first in the negative direction from 0 V to −1.6 V, then in the positive direction till +0.7 V and ended at 0 V. In Fig. 3b three CVs of the bulk Nd-Fe-B electrode scanned in copper-containing, copper-free and iron(II)-containing solutions are shown. The potential was swept first in the negative direction from −0.6 V to −1.6 V, then in the positive direction till +0.7 V and ended at −0.6 V. Here, only the first cathodic part of the scan is presented, but full CVs can be found in Fig. S3 in the SI. As the potential in Fig. 3a is swept down from 0 V, a high positive current is measured in both solutions, resulting from the oxidation of the Nd-Fe-B electrode. From previous CVs (see Fig. 2), we know that Cu 2 ⁺ions begin to be reduced at potentials as high as −0.1 V in the Cu-plate|Cu 2 ⁺,Na + ,SO 4 2− ,Cl − ,H 2 O|Pt system. This implies that they undergo reduction at all potentials lower than −0.1 V. In copper-containing solution, at potentials above −0.5 V, a positive current is measured that is approximately 0.01 A lower than in copper-free solution. This occurs because the reduction of copper ions, which contributes a negative current, takes place simultaneously with the oxidation of the Nd-Fe-B electrode. Below −0.6 V, Nd-Fe-B oxidation is suppressed, and the current in copper-free solution drops to zero. In the copper-containing solution, a plateau of negative current is reached, as the Cu 2+ reduction is controlled by diffusion. At −1.05 V a cathodic peak (C Fe 2+ ) appears in both solutions, which we attribute to the reduction of Fe 2+ ions, formed during the oxidation of the Nd-Fe-B electrode at potentials higher than −0.5 V. Although the peak occurs at the same potential as in the second cycle (Fig. 2b) of Cu-plate|Na + ,SO 4 2− ,Cl − ,H 2 O|Pt system, it represents a different reaction which was studied more in detail in Fig. 3b. Here, as the potential is swept down from −0.6 V, in copper-free solution no cathodic current is measured until −1.2 V, because Fe 2+ ions cannot be present if the electrode is not previously oxidized. When Fe 2+ are introduced to the solution (iron(II)- containing solution, dark grey line), an intense C Fe 2+ peak appears, thus assigning it to reduction of Fe 2+ to elementary iron. The potential of this peak is in accordance with the thermodynamics, as it is below the Fe 2+ /Fe equilibrium potential (−0.65 V vs Ag/AgCl 40 ). In copper-containing solution the overall current is approximately 0.01 A greater than in the copper-free solution because of a constant, diffusion limited Cu 2+ reduction. Appearance of C Fe 2+ indicates the presence of Fe 2+ , even though the potential below −0.6 V (too negative for Nd-Fe-B electrochemical oxidation) has been kept from the start. This means that Fe 2+ is formed when Nd-Fe-B is immersed into a copper-containing solution as a result of electroless Cu 2+ deposition (Fe +Cu 2+ →Fe 2+ +Cu). Nd 3+ is also likely to be released during this spontaneous reaction, but do not result in a cathodic peak, because their reduction potential is much too negative. When the potential in Fig. 3a becomes more negative than −1.2 V, an increasing cathodic current is measured due to hydrogen evolution (C H2O ). In the positive scan an anodic current is observed at potentials higher than −0.5 V, again indicating the oxidation of Nd-Fe-B. In a copper-free solution, the oxidation current (A Nd-Fe-B ) continues to increase until +0.4 V, where it reaches a peak and starts decreasing, indicating some degree of passivation. In copper-containing solution, this peak (A Nd-Fe-B ) is already reached at +0.1 V, possibly due to deposited Cu covering the Nd-Fe-B grains and thus preventing the Nd-Fe-B from further oxidation, but after +0.2 V a rise in current is again observed due to oxidation of the deposited copper (A Cu ) which is found at a similar potential as in Fig. 2. As the previously deposited copper is anodically stripped off, the free surface of the Nd-Fe-B grains is getting oxidized, which additionally increases the current during A Cu . Cu electrodeposition study.—To obtain a metallic copper deposit with the maximum suppression of Nd-Fe-B oxidation and prevention of spontaneous electron exchange between Nd-Fe-B and Cu 2+ ,an extremely negative potential of −1.05 V, corresponding to the reduction peak in Fig. 3, was chosen for the first electrodeposition. The deposition was performed potentiostatically for 30 s from a copper-containing solution on the bulk Nd-Fe-B electrode (current transient presented in Fig. S4 in the SI). The SEM BSE image of this deposit in cross-section is shown in Fig. 4a. The brighter grains are the bulk Nd-Fe-B electrode and the slightly darker tree-like structures on the left-hand side of both images are the copper deposit. As seen from the SEM image, the deposited Cu exhibits a nonuniform dendritic structure. The latter can be explained by the large overpotential, diffusion-limited growth, and the onset of the hydrogen evolution, as evident from the CV presented in Fig. 3.A more positive potential of −0.5 V for 30 s was later used for the Cu potentiostatic deposition. The corresponding microstructure is presented in Fig. 4b, which reveals a more homogenous Cu deposit with only an onset of the dendritic Cu growth. Based on those results, the Cu deposition was performed at −0.5 V from a copper-containing solution for 30 s; however, now using a loose Nd-Fe-B powder electrode. Figure 4c presents a SEM SE and Fig. 4d BSE image of Nd-Fe-B powder particles after performing the Cu deposition. The number of coated particles was high, but the deposit, despite enhanced powder movements and reduced diffusion-limited growth, exhibited dendritic morphology. A cross-section of the Cu-coated Nd-Fe-B powder particle is presented in Fig. 4. Qualitative EDS analysis of the deposit/coating detected Cu as the main element in the deposit (i.e., in the interfacial zone) with a low oxygen content. EDS quantitative analysis of these tiny submicrometre-sized phases was not accurate enough because of their small size. To make the morphology of the Cu coating more homogenous, the potential was set to an even less negative value. Figure 4e shows the SEM SE images of the surface of the Nd-Fe-B powder particles coated with Cu and Fig. 4f shows the SEM BSE the cross-section of the polished particles, of a deposition at −0.25 V for 30 s on a loose Nd-Fe-B powder electrode. Under these conditions, approximately 50% of the particles are coated (see Fig. S5 in the SI), but the copper coatings exhibited substantial variations in deposit thickness. While some particles were covered with thin deposits, with thickness of ∼300 nm (Fig. 4f, the right-hand grain), others were considerably Journal of The Electrochemical Society, 2025 172 022505 thicker, reaching up to 2.5 μm (Fig. 4f, the left-hand grain). A quantitative EDS analysis performed on the thicker coating revealed a composition rich in Cu (i.e., 94.6% Cu, 0.7% O, 3.8% Fe, 1.0% Nd in wt%). Additionally, it has been proven with a low-voltage EDS setup at 7 kV (see Fig. S6), that only Cu and O are present in the coating, which means that the Nd and Fe signals originate from neighbouring Nd-Fe-B particles. This confirms that the deposit is metallic copper with a negligible degree of oxidation. In order to investigate and optimize the electrodeposition of Cu onto loose Nd-Fe-B powders, the temporal evolution of current profiles for different Cu deposition conditions, at E=−0.25 V and E=−0.5 V, with and without ultrasonic agitation, are presented in Fig. 5for a time interval of 30 s. Let us first consider the depositions without ultrasonic agitation, which are presented by the black lines. For both investigated potentials, the currents are negative at the beginning of the reaction. However, when the potential was set to E =−0.25 V (Fig. 5a), the current shifted to positive values after 8 s of deposition, indicating a prevailing oxidation over reduction, as already expected from the cyclic voltammogram in Fig. 3. When the applied potential was set to E=−0.5 V (Fig. 5b), the current was seen to approach zero with time; however, it remained negative for the whole deposition time interval of 30 s. When the bulk Nd-Fe-B electrode is oxidized in a copper-free solution at E=−0.25 V (see Fig. S7) the current initially increases during the first 4 s and then gradually decreases for the rest of the deposition process. Since the oxidation rate is not expected to change significantly during this time (except for the impact in the reduced free surface area due to the Cu deposition), the increasingly positive (or less negative) current observed in all the experiments shown in Fig. 5likely reflects a decrease in the reduction current, meaning a slower deposition rate. This pattern is typical for processes with insufficient mass transport. It is important to emphasize that these experiments were conducted using a moving electrode, but the agitation provided by the external magnet proved inadequate to totally eliminate the mass-transport limitations. Consequently, the additional agitation of the electrolyte solution is required, which in this case involves the use of ultrasonication. In Fig. 5, chronoamperometric data for two new experiments conducted in a copper-containing solution at E =−0.25 V (a) and E =−0.5 V (b) under ultrasonic agitation are presented with grey lines. Ultrasonic agitation during the Cu deposition notably impacted the chronoamperometric behaviour. Without agitation, the current becomes more positive (or less negative) over time, indicating a decrease in reduction current. When ultrasonic agitation is applied, this trend persists, but the rate at which the current becomes more positive is less pronounced. Also, the initial current is more negative if ultrasonic agitation is applied. The US agitation was found to increase the total negative charge passed within the 30 s deposition period, changing from +0.61 C to −0.42 C at −0.25 V and from −2.21 C to −4.10 C at −0.5 V, as given in the right bottom of both graphs presented in Fig. 5. Since agitation has no effect on the thermodynamics of deposition, the observed increase in the cathodic current must be due to the improved mass transport. In Fig. 4g we present an SEM SE image of the morphology and in Fig. 4h a BSE image of the cross-section of the particles after copper deposition from a copper-containing solution at −0.5 V with ultrasonic agitation. A comparison of SEM images in Figs. 4g–4h with Figs. 4c–4d, where the Cu deposition was performed without the US agitation, confirms that Cu coatings on Nd-Fe-B powders Figure 5. Chronoamperometric data for Cu deposition at E =−0.25 V for 30 s (a) without agitation (black line) and with US agitation (grey line). In the bottom right the total negative charge passed within the 30 s deposition period is presented for both cases (a). Current–time curves for Cu deposition at E =−0.5 V for 30 s (b) without agitation (black line) and with agitation (grey line). In the bottom right the total negative charge passed within the 30 s deposition period is presented for both cases (b). Figure 4. SEM images of Cu deposits on Nd-Fe-B electrodes after 30 s of deposition under various applied potential conditions: (a) BSE image, bulk Nd-Fe-B electrode, −1.05 V; (b) BSE image, bulk Nd-Fe-B electrode, −0.5 V; (c) SE image, powdered Nd-Fe-B electrode (Nd₂Fe₁₄B grains), −0.5 V; (d) BSE image, Nd₂Fe₁₄B grains in cross-section, −0.5 V; (e) SE image, powdered Nd-Fe-B electrode (Nd₂Fe₁₄B grains), −0.25 V; (f) BSE image, Nd₂Fe₁₄B grains in cross-section, −0.25 V; (g) SE image, powdered Nd-Fe-B electrode (Nd₂Fe₁₄B grains), −0.5 V ultrasonic agitation; (h) BSE image, Nd₂Fe₁₄B grains in cross-section, −0.5 V, ultrasonic agitation. Journal of The Electrochemical Society, 2025 172 022505 with ultrasonic agitation exhibit a significantly less dendritic surface, suggesting a deposition not, or only partly, limited by diffusion (mixed activation and mass-transport control). 41 At −0.25 V, ultrasonication supplies an additional −1.03 C, and −1.88 C at −0.5 V, as calculated by subtracting the current integrals in Fig. 5, showing that the reduction rate is increased. Ultrasonication, therefore, positively affects the mass transport, which leads to an increase in the average cathodic current and enhances the morphology and the number of coated particles during the 30 s of deposition. Analysis of the magnetic properties.—Based on our experimental findings, the optimal conditions for Cu deposition were achieved using the potentiostatic mode at −0.5 V for 30 s by applying the US agitation. Under these conditions, we successfully produced 5 grams of Cu-coated Nd-Fe-B powder. The magnetic properties of the coated material were subsequently measured with the VSM. The magnetic hysteresis loops for both the Cu-coated NdFe-B powder and the original powder samples are shown in Fig. 6a, indicating mass magnetization values and coercivity. Both samples were magnetically soft with a coercivity of 0.24 kA m −1 for the Cucoated Nd-Fe-B powder and 0.34 kA m −1 for the original Nd-Fe-B powder. The mass magnetization saturates at ≈127 emu g −1 for the Cu-coated Nd-Fe-B sample and at 137 emu g −1 for the uncoated NdFe-B sample, which is in accordance with our previous findings on a similar system. 23 This reduction in mass magnetization is attributed to the dilution with a non-ferromagnetic phase, i.e., the deposited copper, which does not contribute to the magnetic moment but increases the overall sample mass. This trend is obvious from a comparison of the normalized hysteresis loops. By multiplying the curve of the original uncoated Nd-Fe-B particles by a factor of 0.92, a perfect overlap with the hysteresis loop of the Cu-coated Nd-Fe-B powder is found (Fig. 6b). The copper content, determined by electrogravimetry, was 9 mass % (see experimental section), which aligns well with the VSM normalization factor, indicating that the sample contains 8 mass % Cu. Conclusions By employing Na 2 SO 4 for the supporting electrolyte instead of the commonly used H 2 SO 4 , we were able to electrodeposit copper on easily oxidizable substrates based on Nd 2 Fe 14 B grains. A CV study indicated that part of the potential region for Cu 2 ⁺reduction overlaps with the region for Nd-Fe-B oxidation (above −0.5 V), resulting in a spontaneous reaction between Nd-Fe-B and Cu 2+ . As a result Cu electroless deposition occurs and Fe 2+ ions are released into solution leading to loss of ferromagnetic material. During the potentiostatic deposition of Cu on bulk Nd-Fe-B-based electrodes at such highly negative potentials (−1.05 V), the copper deposits exhibit a nonuniform dendritic structure due to diffusion-limited growth and the onset of hydrogen evolution. Less-negative potentials, such as −0.5 V, still resulted in a negative current throughout the whole deposition experiment indicating prevailing Cu reduction over the Nd-Fe-B oxidation. The achieved Cu deposits were more uniform both on bulk and powder Nd-Fe-B-based electrodes; however, they still had minor dendritic growth and the covering frequency of the powder Nd-Fe-B-based electrode was low. The application of ultrasonic agitation enhances the mass transport, leading to improved copper-deposit uniformity on the Nd-Fe-B powders and an increased Cu deposition rate, as evidenced by measuring a more reductive current. Choosing more positive potentials like −0.25 V would improve the Cu deposit’s morphology but cause a significant Nd-Fe-B oxidation, as evidenced by a significant anodic current. This study demonstrates the successful electrochemical deposition of Cu onto Nd₂Fe₁₄B grains, with no magnetization reduction aside from the introduction of the paramagnetic Cu phase. These findings lay the groundwork for future research into sintering and microstructural optimization to explore Cu’s role as a novel grain boundary phase in enhancing coercivity and magnetic performance. Investigating a system of Nd₂Fe₁₄B/Cu, combined with densification and sintering approaches, will be essential to fully understand copper’s impact on key magnetic properties such as remanent magnetization and coercivity in bulk Nd-Fe-B magnets. Copper electrodeposits thus represent a critical first step in advancing grainboundary engineering, introducing new phases that enhance corrosion resistance and promote cost and resource efficiency. Acknowledgments This research was funded by the Slovenian Research and Innovation Agency (research core funding no. P2–0084) and the European Union’s Horizon Europe Research and Innovation Program REESILIENCE (grant agreement no. 101058598) and GREENE (grant agreement no. 101129888). All figures created with BioRender.com. ORCID Črt Saksida https://orcid.org/0009-0002-4648-1578 László Péter https://orcid.org/0000-0001-5604-0982 Kristina Žužek https://orcid.org/0000-0003-2652-1966 References 1. B. Podmiljšak, B. Saje, P. Jenuš, T. Tomše, S. Kobe, K. Žužek, and S. Šturm, Materials,17, 848 (2024). 2. J. W. Heim II and R. L. Vander Wal, Minerals,13, 1274 (2023). 3. J. Li, H. Sepehri-Amin, T. Sasaki, T. Ohkubo, and K. Hono, Sci. Technol. Adv. Mater.,22, 386 (2021). 4. M. Grohol and C. Veeh, European Commission: Directorate-General for Internal Market, Industry, Entrepreneurship and SMEs, Study on the critical raw materials for the EU 2023, Publications Office of the European Union, 2023, (2024), https:// data.europa.eu/doi/10.2873/725585. 5. O. Gutfleisch, J. Phys. D,33, R157 (2000). Figure 6. As-measured magnetic hysteresis loops of Nd-Fe-B powder before and Cu after deposition at −0.5 V for 30 s with US agitation (a) and the normalized hysteresis loop of Cu-coated Nd-Fe-B powder at −0.5 V for 30 s with US agitation overlapping the loop of uncoated Nd-Fe-B powder (b). Journal of The Electrochemical Society, 2025 172 022505 6. T. Schrefl, J. Fidler, and H. Kronmüller, Physical Review B,49, 6100 (1994). 7. K. Hono and H. Sepehri-Amin, Scr. Mater.,67, 530 (2012). 8. H. Sepehri-Amin, T. Ohkubo, T. Shima, and K. Hono, Acta Mater.,60, 819 (2012). 9. J. Fidler and J. Bernardi, J. Appl. Phys.,70, 6456 (1991). 10. S. Pandian, V. Chandrasekaran, M. Garimella, K. Iyer, and K. Rao, J. Appl. Phys., 92, 6082 (2002). 11. A. Yan, X. Song, Z. Chen, and X. Wang, J. Magn. Magn. Mater.,185, 369 (1998). 12. W. Li, T. Ohkubo, and K. Hono, Acta Mater.,57, 1337 (2009). 13. H. J. Engelmann, A. S. Kim, and G. Thomas, Scr. Mater.,36, 55 (1997). 14. Y. Tatetsu, S. Tsuneyuki, and Y. Gohda, Phys. Rev. Appl.,6, 064029 (2016). 15. J. Dai, Z. Yang, and Q. Liu, Materials,13, 4360 (2020). 16. S. Pandian, V. Chandrasekaran, M. Garimella, K. Iyer, and K. Rao, J. Alloys Compd.,364, 295 (2004). 17. A. Mishra, S. Khoshsima, T. Tomše, B. Podmiljšak, S. Šturm, C. Burkhardt, and K. Žužek, Materials,16, 6565 (2023). 18. W. Li, T. Ohkubo, K. Hono, and M. Sagawa, J. Magn. Magn. Mater.,321, 1100 (2009). 19. L. Schultz, A. M. El-Aziz, G. Barkleit, and K. Mummert, Mater. Sci. Eng. A,267, 307 (1999). 20. A. Walton, Y. Han, N. A. Rowson, J. D. Speight, V. S. J. Mann, R. S. Sheridan, A. Bradshaw, I. R. Harris, and A. J. Williams, J. Clean. Prod.,104, 236 (2015). 21. A. Walton, A. J. Williams, J. D. Speight, and I. R. Harris, U.S. Patent, 8,734,714 B2 (2014). 22. C. Burkhardt*, S. van Nielen, M. Awais, F. Bartolozzi, J. Blomgren, P. Ortiz, M. B. Xicotencatl, M. Degri, S. Nayebossadri, and A. Walton, J. Magn. Magn. Mater., 588, 171475 (2023). 23. S. Khoshsima, J. Vidmar, Z. Samardžija, T. Tomše, M. Kušter, A. Mishra, S. Šturm, and K. Žužek, Materials,16, 5181 (2023). 24. H. Bala, S. Szymura, and J. J. Wysłocki, J. Mater. Sci.,25, 571 (1990). 25. G. W. Warren, G. Gao, and Q. Li, J. Appl. Phys.,70, 6609 (1991). 26. P. Venkatesan, T. Hoogerstraete, T. Hennebel, K. Binnemans, J. Sietsma, and Y. Yang, Green Chem.,20, 1065 (2018). 27. X. Xu, S. Šturm, Z. Samardžija, J. Vidmar, J. Ščančar, and K. Rožman, ChemSusChem,12, 4754 (2019). 28. J. Zheng, X. Cheng, H. Chen, L. Qiao, M. Lin, L. Jiang, and S. Che, J. Adhes. Sci. Technol.,30, 275 (2016). 29. M. S. Paunovic and M. Schlesinger, Modern Electroplating 5th ed. ed.736 (New York), (Wiley, Hoboken, NJ) (2010). 30. G. G. Láng, M. Ujvári, and G. Horányi, J. Electroanal. Chem.,522, 179 (2002). 31. Q. Wu and D. Barkey, J. Electrochem. Soc.,147, 1038 (2000). 32. C. Gabrielli, P. Moçotéguy, H. Perrot, and R. Wiart, J. Electroanal. Chem.,572, 367 (2004). 33. W. Shao, G. Pattanaik, and G. Zangari, J. Electrochem. Soc.,154, D201 (2007). 34. , Low, , Ponce, de León, , and Walsh, Transactions of the IMF,93, 74 (2015). 35. P. Venkatesan, Z. H. Sun, J. Sietsma, and Y. Yang, Proceedings of the 1st European Rare Earth Resources Conference163 Milos, Greece (2014). 36. J. B. Raoof, R. Ojani, A. Kiani, and S. Rashid-Nadimi, Int. J. Hydrogen Energy,35, 452 (2010). 37. S. M. Abd El Haleem and B. G. Ateya, J. Electroanal. Chem. Interfacial Electrochem.,117, 309 (1981). 38. J. Kunze, V. Maurice, L. H. Klein, H-H. Strehblow, and P. Marcus, Electrochim. Acta,48, 1157 (2003). 39. S. Kologo, M. Eyraud, L. Bonou, F. Vacandio, and Y. Massiani, Electrochim. Acta, 52, 3105 (2007). 40. S. Sathiyanarayanan, M. Sahre, and W. Kautek, Corros. Sci.,41, 1899 (1999). 41. C. Lupó and D. Schlettwein, J. Electrochem. Soc.,166, D3182 (2018). Journal of The Electrochemical Society, 2025 172 022505