Injection Metal-Assisted Catalytic Etching (MACE) of Si Powder: Discovery of Low-Load MACE and Pore Distribution Tunability Using Ag, Au, Pd, Pt and Cu Catalysts

Abstract

Injection of H2O2 to control both the rate and extent of etching allowed us to discover a new regime of MACE with extremely small quantities of deposited metal as a catalyst: low-load MACE (LL-MACE).1 The structure of particles subjected to LL-MACE is completely different compared to conventional MACE. Si nanowires and mesoporous micro- or nano-particles are produced with high yield, low cost and controlled properties suitable for applications in e.g. lithium-ion batteries, drug delivery, as well as biomedical imaging and contrast enhancement. Different metal catalysts behave in distinct manners than can be traced to differences in nucleation and reaction kinetics.2 Not only Ag, Au, Pd and Pt but also inexpensive Cu can be used to perform LL-MACE. Previously, controlled reactant injection and acetic acid were used to transform stain etching into the improved process known as regenerative electroless etching (ReEtching).3 Here injection and acetic acid are exploited to enhance MACE and extend it to the low-load regime. Ag, Au, Pd, Pt and Cu can be galvanically deposited onto Si powder out of solutions of HF(aq). Nucleation of deposited metal nanoparticles is made more uniform by dispersing the Si powder in acetic acid and injecting the dissolved metal ions at a controlled rate. The use of a syringe pump to deliver not only metal ions during deposition but also the oxidant (H2O2) during MACE is essential for increased product uniformity and yield. Temperatures of ~0–50 °C and grades of Si (i.e. electronics grade Si with various doping levels or metallurgical grade Si) produced significantly different pore size distributions. Doping levels and the associated changes in band bending and space charge layer width4 are linked to changes in the pore size distribution. As shown in Fig. 1(A), high loading of Ag resulted in deposition of dendrites and Ag particles. These broke into ~50–100 nm Ag nanoparticles that etched in a correlated fashion to generate parallel etch track pores. The uniform size distribution of the Ag nanoparticles was generated dynamically during etching. This conventional high-load MACE (HL-MACE) did not produce Si nanowires (Si NWs) directly. However, etch track pore walls were cleaved readily by ultrasonic agitation to form Si NWs. Etching far removed from the metal nanoparticles (remote etching) also occurred for highly doped Si. As shown in Fig. 1(B), reducing the loading of Ag by a factor of 100 to 1000 compared to HL-MACE created 10–30 nm nanoparticles that etched in an uncorrelated fashion and generated randomly directed etch track pores. Significant etching far removed from the Ag catalyst also occurred. The combination of localized and remote etching resulted in a bimodal distribution of mesoporosity peaking at ~4 nm from remote etching and 10–30 nm from localized etching. The bimodal distribution was confirmed by the BJH method of nitrogen adsorption porosimetry. The pore size distribution could be controlled by changing the process temperature, the grade of Si and the metal composition. Whereas Ag, Au, Pd and Pt performed well in both the HL and LL limits, Cu generated porosification only in the LL limit. Funded by grants #314552 and #314412 from Academy of Finland and the National Science Foundation award #1825331. The microscopy studies were performed using the facilities in the UConn/Thermo Fisher Scientific Center for Advanced Microscopy and Materials Analysis (CAMMA). 1 K. Tamarov, J. D. Swanson, B. A. Unger, K. W. Kolasinski, A. T. Ernst, M. Aindow, V.-P. Lehto, and J. Riikonen, ACS Appl. Mater. Interfaces 12, 4787 (2020). 2 K. W. Kolasinski, W. B. Barclay, Y. Sun, and M. Aindow, Electrochim. Acta 158, 219 (2015). 3 K. W. Kolasinski, N. J. Gimbar, H. Yu, M. Aindow, E. Mäkilä, and J. Salonen, Angew. Chem., Int. Ed. Engl. 56, 624 (2017). 4 E. Torralba, S. Le Gall, R. Lachaume, V. Magnin, J. Harari, M. Halbwax, J.-P. Vilcot, C. Cachet-Vivier, and S. Bastide, ACS Appl. Mater. Interfaces 8, 31375 (2016). Figure 1