Flame-made Particles for Sensors, Catalysis, and Energy Storage Applications

Abstract

Flame spray pyrolysis of precursor-solvent combinations with high enthalpy density allows the design of functional nanoscale materials. Within the last two decades, flame spray pyrolysis was utilized to produce more than 500 metal oxide particulate materials for R&D and commercial applications. In this short review, the particle formation mechanism is described based on the micro-explosions observed in single droplet experiments for various precursor-solvent combinations. While layer fabrication is a key to successful industrial applications toward gas sensors, catalysis, and energy storage, the state-of-the-art technology of innovative in situ thermophoretic particle production and deposition technology is described. In addition, noble metal stabilized oxide matrices with tight chemical contact catalyze surface reactions for enhanced catalytic performance. The metal-support interaction that is vital for redox catalytic performance for various surface reactions is presented.

Figure 1

(A) Particle formation routes during flame aerosol synthesis. During particle formation, the gas-to-particle route is realized for the combustion of metalorganic precursor–solvent with high enthalpy density, while the droplet-to-particle route is observed for nonvolatile and/or endothermic combustion of the precursors. (B) The formation of single and/or multicomponent particles via gas-to-particle mechanism. Adapted with permission from refs ( and 5). Copyright 2010 Royal Society of Chemistry and 2020 American Chemical Society.

Figure 2

TEM images of the particle morphology obtained using precursor–solvent combinations with different combustion enthalpy densities: (A–D and F–I) TEM/SEM images of hollow and large Bi₂O₃ particles (100–500 nm). The precursor–solvent combination such as Bi(NO₃)₃ + HNO₃ + C₂H₅OH with low combustion enthalpy density produces hollow particles. The particles (E and J) obtained from spray pyrolysis of Ce(ac)₃ + pure acetic acid were as large as 300 nm distributed over smaller sized particles in the range of 10 nm. The inhomogeneities of the particles were due to the low combustion enthalpy of acetic acid. However, when combustible precursor solvent combinations such as bismuth neodecanoate dissolved in xylene were used, crystalline and ultrafine particles were obtained (O and T). The data reflect the importance of precursor chemistry for high quality particles. Loosely interconnected networks of (K) as prepared nanocrystalline of Ce_0.5Zr_0.5O₂ followed by (P) calcination. The crystalline and fine particles resulted from the combustible organic precursors and solvents such as hydrated Cerium(III) acetate hydrate and zirconium tetraacetylacetonate dissolved in lauric–acetic acid. The images of FSP synthesized ZnO nanoparticles (L) and ZnO quantum dots (Q). Very fine WO₃ particles obtained from W(CO)₆ dissolved in solvent with high combustion enthalpy such as tetrahydrofuran (M, N, R, and S). Adapted with permission from refs (−14). Copyright Material Research Society (2002), American Institute of Physics (2002), American Ceramic Society (2002, 2005), Royal Society of Chemistry (2003, 2016), American Chemical Society (2010).

Figure 3

Proposed mechanistic particle formation routes during single droplet combustion of tin-2ethylhexanoate dissolved in xylene (left) and the occurrence of microexplosions from the surface of the droplets. Adapted with permission from ref (44). Copyright 2020 Elsevier.

Figure 4

Role-to-role layer transfer process of thermophoretically collected particles (a) gas phase synthesis of TiO₂ nanoparticles and subsequent layer transfer via lamination from the collecting unit to the substrate. (b) Scanning electron microscopic images of the fabricated layer (I) before (II) after role-to-role lamination. (c) Compaction simulation of the thermophoretic TiO₂ nanoparticle layer. The φ, p, and h denote the porosity, pressure and film height, respectively. Adapted with permission from refs ( and 53). Copyright Elsevier (2013) and Royal Chemical Society (2019).

Figure 5

(A) Scanning electron microscopy image of the cross section of the directly deposited double layer sensing substrates where an Al₂O₃/Pd filter was coated on top of a SnO₂ layer. Sensing signals observed at 200 and 400 °C in dry air at (B) 230 ppm of CH₄ and at (C) 10 ppm of CO and (D) 10 ppm EtOH. Adapted with permission from refs ( and 56). Copyright 2007 Elsevier and Materials Research Society.

Figure 6

Thermophoretically deposited In_1.9Sn _0.1O₃ and In₄Sn₃O₁₂ layers for formaldehyde sensing. (A) The flame picture showing gas phase particle production. (B) Scanning electron microscopy image of the directly deposited sensor layer. (C) FIB cross-section of the sensing layer demonstrating the layer thickness (∼30 μm) and porosity. (D) and (E) The crystallographic projections of hexagonal In₄Sn₃O₁₂ and cubic In_1.9Sn _0.1O₃, respectively. (F) Phase compositions of In–Sn–O ternary materials with different In/Sn ratios and (G) sensor signals to 73 ppb of formaldehyde in 50% RH. The highest sensor signals are observed for 43% of Sn due to metastable In₄Sn₃O₁₂. Adapted with permission from ref (16). Copyright 2012 Elsevier.

Figure 7

(A) LTO/C synthesis and electrode fabrication via role-to-role layer transfer and doctor blading. (a) Flame spray-produced LTO and carbon aerosol streams mixed prior to deposition on a glass-fiber filter. Gas phase thermophoretic particle production technique enables homogeneous mixing of the particles and heterojunction formation in the aerosol stream when multiple flames are used (Table 1). (b) The glass-fiber unit with LTO/C particles placed upside down on a copper current collector. The LTO/C particle layer is transferred to the copper foil using a role-to-role laminator. (c) The LTO/C powder from the collection unit in (a) is removed and fabricated into a paste using NMP and PVDF. The paste is spread on a copper foil, dried, and calendared for the electrode fabrication. Long-term cycling and rate test capability of the pouch cells derived from three different LTO obtained from precursors such as lithium nitrate, lithium tert-butoxide, and lithium acetylacetonate. (B) dis-/charge curves in the 1st, 2nd, and 450th cycles, and (C) long-term cycling performance, (D) LTO flex-TFB preparation via FSP-lamination (a) mounting the Cu-substrates on the flame spray holder and covering the substrates using mask (b) thermophoretic deposition of LTO on the substrates (c) role-to-role lamination for the layer compression (d) cell assembly after sputtering electrolyte (LiPON) and Li electrode, (E) photography of the custom-built bending apparatus for mechanical bending in static condition of the battery cells, (F) ADCs of LTO flex-TFB cells at 2 μA (1C) in prebent, bent, and postbent conditions. Adapted with permission from refs (, , and 79). Copyright (2017) American Chemical Society and Elsevier (2018).

Figure 8

(A) The experimental double-flame setting for the production of the Pd/SiO₂–Al₂O₃ catalysts. (B–D) Transmission electron microscopy (STEM, low and high resolution TEM modes) of 5% Pd/SiO₂–Al₂O₃ (Si/Al = 30/70) catalysts. The bright white spots in STEM images are crystalline Pd particles. (E) ¹H MAS NMR and (F) ²⁷Al MQMAS NMR spectrum of 5% Pd/SiO₂–Al₂O₃ (Si/Al = 30/70). Data show the density of surface Brønsted acid sites on 5% Pd/SA could be varied using different SiO₂/Al₂O₃ ratios. (G) Chemoselective hydrogenation mechanism on double flame-made FSP Pd/SA catalysts. Adapted with permission from ref (9). Copyright 2013 Elsevier.