Flash Joule heating for synthesis, upcycling and remediation

Introduction

Industrial processes responsible for chemicals and raw materials production frequently demand large quantities of heat for tasks such as steam generation, melting, refining, separation and drying¹. Currently, much of this industrial heating is powered by coal or natural gas combustion² and, in theory, could be replaced by electricity-based heating methods. Therefore, industrial electrification³ — the transition from fossil fuel-based technologies to electricity-based technologies — is instrumental in decarbonizing society, as the industrial sector contributes ~26% of global carbon emissions⁴. However, industrial electrification is still in early stages and requires technological innovation⁵.

Electric heating methods encompass a range of technologies such as resistance heating⁶ (Box 1), electric arc heating⁷, induction heating⁸ and dielectric heating⁹. These methods differ in energy efficiency, temperature range and heating rate, for instance (Supplementary Table 1). Direct electric Joule resistive heating¹⁰ is emerging as a promising heating technique because of its very high energy efficiency (100% electricity-to-heat conversion efficiency), as it involves no intermediate heat transfer. Flash Joule heating (FJH) is a type of direct resistance heating that occurs when a high-power, short-duration electrical pulse from a power source passes through a material, rapidly generating intense heat directly within the chosen material without the need for a heating medium to transfer the heat¹¹. The high efficiency, short duration, solvent-free nature and minimal heat loss of FJH processes are improvements compared with other processes such as radiative heating or wet methods¹².

Following its introduction in 2020 for converting carbon sources into high-quality graphene¹¹, FJH has been scaled, and its applications have expanded. FJH has been used in various materials synthesis and thermal treatment processes (Fig. 1A), including in bottom-up synthesis of graphene^11,13, carbon nanotubes¹⁴, carbides¹⁵, borides¹⁶, nitrides¹⁷, metallic glasses^18,19, transition-metal dichalcogenides and p-block metal chalcogenides²⁰, as well as doped and functionalized variants of these compounds²¹. Waste materials can be used as feedstocks for materials synthesis, providing a substantial reduction in energy intensity and emissions for many FJH processes relative to using virgin materials^12,22. FJH has also been applied in resource recovery and waste decontamination, including in perfluoroalkyl and polyfluoroalkyl substances (PFAS) destruction^23,24, soil remediation²⁵, waste plastic upcycling^14,26, and extraction of rare earth metals²⁷.

Flash Joule heating for synthesis, upcycling and remediation — **Fig. 1: FJH principle, equipment and scale-up.**

In this Review, we describe FJH, covering its fundamental principles, equipment design, applications, industrial implementation, sustainability and technoeconomic considerations (Fig. 1A). We begin by discussing the basic principles and theories underlying FJH, as well as the equipment and reactor designs. Subsequently, we highlight its potential applications in fields such as synthesis of carbon materials and metastable inorganic materials, resource recovery and waste upcycling, and environmental remediation. Sustainability and technoeconomic considerations are further discussed. Finally, we outline the remaining challenges that must be addressed to make FJH a mainstay of manufacturing and environmental stewardship.

Box 1 Resistance heating

Resistance heating, also known as resistive heating or Joule heating, has its roots in the 1840s when James P. Joule discovered that heat could be generated by applying electricity to a resistor²⁸. Following Joule’s law, heat (Q) produced is proportional to the square of the current (I) multiplied by the electrical resistance (R) of the resistor and the heating time (t),

$$Q={I}^{2}Rt$$

(4)

Joule heating boasts a theoretical energy conversion efficiency (η) of 100%, meaning that all electric energy is converted into heat (see the table, which compares general resistive Joule heating with flash Joule heating (FJH)). This efficiency contrasts with the fuel utilization efficiency of furnaces, where chemical energy is transformed into heat through combustion. Furnace efficiency varies from 50% to 100% depending on the system design, but is often at the lower end of this range¹⁷⁹.

Resistive Joule heating is used in indirect heating resistance furnaces²⁹, which are widely used across industry, such as in metallurgy²⁹. In these furnaces, current passing through a material generates heat, which is then transferred to the medium through conduction, convection and/or radiation²⁹. Enclosures are typically used to isolate the heating process from the external environment¹⁸⁰. However, indirect resistance heating tends to have low energy efficiencies due to heat loss during the transfer process.

Unlike indirect Joule heating, direct Joule heating¹⁰ involves no intermediate heat transfer and therefore has very high energy efficiency. Direct heating for materials synthesis is used in the carbothermal shock process^105,106. This process involves the loading of metal feedstocks on a carbon substrate, followed by introduction of an electric pulse for rapid heating, resulting in nanoparticle synthesis. The feedstocks are in direct contact with the carbon paper, and heated by mainly conduction. The carbothermal shock method has been widely applied in the production of nanomaterials^{105,114,116,125}, ceramic processing^38,181, thermochemical synthesis¹³⁰ and plastic depolymerization¹⁴⁹. FJH is another type of direct Joule heating and uses high temperatures and a fast duration (see the table). In addition to the direct contact heating, FJH can use the materials themselves as the resistive medium. In this case, heat transfer is not required, such that the efficiency can be even higher than the carbothermal shock process. Many other terms are used in the literature, such as high-temperature shock, high-temperature sintering, high-temperature electrothermal process, flash carbothermic reduction, flash upcycling, flash sintering and rapid Joule heating; some of these processes can have indirect characteristics or variations of carbothermal shock and FJH.

Feature	Combustion furnace^a	Resistance furnace	Carbothermal shock	Flash Joule heating
Energy source	Combustion of fuel	Continuous electric current	Short, high-power electric pulse	Short, high-power electric pulse
Electricity supply	Not required	d.c. or a.c. power	d.c. or a.c. power	d.c. or a.c. power, capacitor
Principle	Exothermic chemical reaction, heat transfer	Joule’s law of heating, heat transfer	Joule’s law of heating, heat transfer	Joule’s law of heating, with or without heat transfer
Maximum energy efficiency	Moderate, up to 80–90% (ref. ¹⁸²)	High, up to 95–100%	High, up to 95–100%	High, up to 95–100%
Temperature range	Depending on the fuel types, typically <2,000 °C (ref. ¹⁸³)	Up to 3,000 °C	Up to 3,000 °C	Up to 3,500 °C
Heating rate	Slow, typically <50 °C min⁻¹	Slow, typically <30 °C min⁻¹ for a resistance furnace¹⁸⁴	Extremely high, up to 10⁵°C s⁻¹	Extremely high, 10²–10⁵ °C s⁻¹
Cooling rate	Slow, typically <10 °C min⁻¹	Slow, typically <10 °C min⁻¹ for a resistance furnace	Extremely high, up to 10⁴°C s⁻¹	Extremely high, 10²–10⁴ °C s⁻¹
Processing duration	Long, typically hours	Long, typically minutes to hours or days	Short, typically milliseconds to minutes	Short, typically milliseconds to minutes
Heating manner	Indirect heating	Indirect heating, mainly by convection and radiation	Direct heating by thermal conduction heating	Direct heating by Joule resistance heating or thermal conduction heating
Material suitability	Wide applicability	Conductive and semiconductive materials	Materials loaded onto a carbon substrate	Materials with appropriate resistance (typically ~0.5–10 Ω, although larger ranges can be used depending on the power supply methods) or materials in direct contact with a resistance heat source

^aCombustion furnaces do not use Joule heating and are discussed here as a comparison.

Principles, equipment and scalability

In this section, the principles behind FJH are introduced, the apparatus and systems are demonstrated, and various reactor designs targeting different applications are discussed. Next, the ongoing scaling up of the FJH process toward industry-scale implementation is briefly discussed.

Fundamentals

Joule heating is a process by which an electrical current that passes through a resistive medium heats this medium by Joule’s law, according to which the heating power is equal to the electrical potential difference times the current flowing through the medium²⁸. Here, we define flash Joule heating as a direct heating process with a high-power, short-duration electrical pulse generated by a power source that is applied directly to a resistive material, causing extremely rapid heating of the target material to a wide temperature range, followed by rapid cooling. The target material can itself be the resistive material, or the target material can be in direct contact with the resistive material. The current density can exceed 10 A mm⁻² and, when used in conjunction with a suitable electrically resistive material, leads to ultrahigh temperatures that can even exceed 3,500 °C (ref. ¹¹). The current pulse width can be as short as milliseconds to seconds, resulting in ultrafast heating rates (typically 10² to 10⁵ °C s⁻¹; refs. ^11,15) and ultrafast cooling rates (typically 10² to 10⁴ °C s⁻¹, refs. ^11,15), owing to rapid thermal dissipation.

Most conventional solid-state synthesis techniques, including flux growth and chemical vapour deposition, involve the use of a resistive furnace²⁹. Joule heating is used in heating elements embedded within the furnace, which act as the resistive medium through which current passes. These heating elements pass heat through the walls and gas inside the furnace through thermal conduction, convection and radiation, heating the chemical reactants within the furnace²⁹. The slow heat-transfer rate limits the heating rate to a maximum of ~1 °C s⁻¹, and these furnaces have low thermal efficiency as the target reactants compose only a small minority of the total heat capacity of the volume being heated²⁹.

In contrast to furnace-based heating methods, FJH uses the materials themselves as the resistive medium, allowing for ~100% sample heating efficiency at heating rates up to 10⁵ °C s⁻¹ (ref. ¹¹). This fast heating rate and the subsequent fast cooling rate enable kinetically controlled non-equilibrium products to be formed¹⁵. The high temperature of FJH (in excess of 3,500 °C) allows for rapid reaction to be completed on a millisecond to second timescale^11,15. Features including temperature range, heating and cooling rates, duration and heating manner distinguish FJH from conventional Joule heating (Box 1).

FJH is an electrothermal process rather than a solely thermal process. Therefore, in addition to temperature increase, FJH features the passage of electric current through the reactants, which necessitates an electrical potential difference through the sample rather than just along a sample surface. The presence of an electric current passing through a material affects the chemical products¹³ and can lower reaction activation energies by as much as 50% (ref. ³⁰). The passage of current through the material also promotes crystalline alignment of the flash-Joule-heated product along the axis of current flow, and this effect is further enhanced through rapid pulse-width modulation³¹. With the same current and voltage input, Joule heating performance depends on various materials properties, including thermal conductivity, a defined resistivity range and heat capacity, so the temperature is related to the materials.

Electric systems and hardware

FJH can be performed with any system that applies a sufficient voltage across the reactant medium. The methods were initially developed using capacitor-based systems¹¹ (Fig. 1B). The energy delivered is determined by the total capacitance of the capacitors as well as the voltage to which these capacitors are charged prior to discharging, as expressed in equation (1):

$$E=frac{1}{2}C{V}^{2}$$

(1)

where E is the energy input, C is the capacitance and V is the voltage.

The advantages of a capacitor-based system include the precisely controllable timescale of power delivery through associated electronics (Fig. 1C), the controllable predetermined energy value, and the ability of many models of capacitors to be charged to several hundred volts. Consequently, capacitor-based systems can achieve high power output, up to 1 MW (ref. ²³), but are limited in output duration (usually subsecond timescales)¹¹. Engineering techniques, such as pulse-width modulation by a variable frequency drive (Fig. 1B), can increase the capacitor discharge duration. Capacitor-based FJH systems can effectively be scaled up to kilogram production using rapid cycling of smaller batches, with automated sample loading and unloading to make the system continuous and high throughput. In these systems, the rate-determining step is often the repeated recharge time of the capacitor bank after each use^23,32.

Non-capacitor systems have the benefit of allowing continuous FJH without a necessary recharging step, offering higher energy delivery — despite reduced power output— owing to the long FJH timescales permitted (Fig. 1B). These non-capacitor, continuous systems exist in alternating-current (a.c.) and direct-current (d.c.) varieties (Fig. 1B). Alternating-current flash Joule heaters use a.c. electricity directly from the laboratory or manufacturing site, and they do not rectify the current before delivering it through the sample. Such systems are commonly used for pretreatment to reduce the electrical resistivity of a sample before subsequent d.c. flashing^33,34. Direct-current systems feature a device that rectifies the a.c. input from the facility to d.c. output to the sample and are the most common type of continuous flash Joule heater²⁴. These systems have programmable power supplies with which the desired voltage and current can be chosen before the reaction, and they can be programmed to adjust during FJH. In this way, continuous systems offer superior energy controllability and output relative to capacitor-based systems, while having inferior power output. Indeed, a capacitor-based FJH reaction can be completed in milliseconds (Fig. 1D), whereas an a.c.-based or d.c.-based FJH reaction can be sustained much longer (Fig. 1E).

Reactor design

Most FJH reactions involve filling the reactant feedstock into an FJH vessel, consisting of an insulative tube capped at the two ends by FJH electrodes made of either brass or graphite (Fig. 1Fa). Although other materials such as polytetrafluoroethylene are occasionally used³⁵, the tube is typically fused quartz in laboratory settings, as its transparency allows for convenient observation and temperature measurement of the reaction through infrared and spectral methods. Fused quartz also has a high melting point (~1,650 °C), but this is rarely reached as minimal heat is transferred to the tube during the rapid FJH reaction. Additionally, the low thermal conductivity and low thermal expansion coefficient of fused quartz allow it to withstand cracking from thermal shock caused by rapid heating and cooling of FJH reactions. For large-scale applications, other refractory, cost-effective materials such as concretes, ceramics or brick might be used.

Before FJH, the feedstock is compressed between the electrodes to improve electrical contact and reduce the sample resistance¹¹. Vessels can either be sealed to contain the gaseous products of the flash reaction or instead have holes in the electrodes to allow pressure relief through reaction outgassing (Fig. 1Fc). Most FJH reaction vessels are deliberately not sealed to allow reaction outgassing and prevent high pressure buildup within the reaction vessel. This outgassing is often advantageous in removing volatile reactant impurities, resulting in higher product purity^26,33. Furthermore, outgassing prevents oxygen infiltration, protecting the reactants and products from burning at the high temperatures reached. The volatiles can also be collected for analysis of the reaction process. In some cases, the volatiles are the desired products³⁶.

The electrical current exhibited during an FJH reaction is a function of the bulk resistance of the sample and the output voltage of the FJH system, both of which determine the current value, as described by Ohm’s law:

$$frac{V}{R}=I$$

(2)

where I is the current, V is the voltage and R is the sample resistance. If the sample resistance is too high, the current value could be too low, and thus the power delivered will not be sufficient, as described by Joule’s law:

$$P=mathrm{IV}={I}^{2}R$$

(3)

where P is the power. By contrast, if the sample resistance is too low, then a substantial portion of the heating can occur in the components of the FJH system rather than in the sample. Hence, the sample resistance should be in a suitable range for delivering the most electric energy. The real-time temperature profile during the FJH reaction is usually recorded using an infrared thermometer or a spectrometer when the temperature exceeds 3,000 °C (refs. ^11,15).

As resistivity is an intrinsic property of a material, the requirement of sample resistance could restrict the applicable feedstock. To address this limitation, several techniques have been developed for non-conductive feedstocks. First, a low-power pretreatment heating step enhances the conductivity of carbon feedstocks through carbonization, which is especially useful when plastic or biomass are used as the feedstocks³⁷. Second, the addition of conductive additives, typically carbon materials, into the feedstocks³³ permits the synchronous heating of the non-conductive material. In addition to carbon, other materials with appropriate resistance may be used as the substrates or additives for FJH, but this is underexplored in the field. The conductive additive can remain in the product in some cases, or it can be separated based on the property differences between the additive and the product, such as through sieving if their sizes are different. Third, the use of a resistive heating element enables the rapid heating of the non-conductive materials in thermal contact with it (Fig. 1Fb) or on it enabled by thermal conduction^20,30,38 (Fig. 1Fd). A preferred sample resistance for capacitor-based FJH systems is 0.5 to 10 Ω (ref. ¹¹); when an a.c.-based or d.c.-based FJH system is used, a wide resistance range (0.05–100 Ω) can work, depending on the power supply¹².

Scale-up and industrial implementation

FJH is theoretically scalable^15,22,39. Experimentally, FJH has been conducted at various scales. In its early demonstration for flash graphene synthesis¹¹, which used a capacitor-based system with a total capacitance of 0.06 F and a maximum voltage of 200 V, the mass of product ranged from 0.03 g to 1 g per batch. For a capacitor-based system, the voltage and/or capacitance of the capacitors can be increased to increase the mass per batch: 10 g per batch of flash graphene was prepared using an upgraded FJH system with a capacitance of 0.624 F and a maximum voltage of 480 V (ref. ²³) (Fig. 1G). Furthermore, a lab-scale automated FJH system²³ has been demonstrated to achieve the synthesis of flash graphene using metallurgical coke as the feedstock at a production rate of 730 g h⁻¹ (ref. ²³). Continuous FJH models have been introduced for flash graphene synthesis⁴⁰, including a continuous automatic device for flash graphene synthesis from biomass with a production rate of 21.6 g h⁻¹ (ref. ³⁷) (Fig. 1H).

Integrating continuous d.c. or a.c. into the FJH process offers better energy deliverability than a capacitor-based FJH system and enables scale-up, provided there is a sufficient instant power supply. For example, mass scales of 100 g per batch have been achieved using an a.c. FJH heater²⁵; a US$530 continuous FJH reactor was used to achieve 5 kg h⁻¹ flash graphene synthesis from metallurgical coke at kilogram-scale batch size⁴¹. Although lab-scale equipment is usually designed for manual operation, mechanical operation was integrated for continuous feedstock loading, FJH and product unloading for reported scale-up prototypes (Fig. 1h). Commercial scaling-up of the FJH process for flash graphene synthesis and metal recycling is ongoing. Broadly, we expect further scale-up of the FJH process for graphene production as well as other uses in the near future, given the automation equipment development, engineering advances and investment in this field.

Owing to the high-power input and high intensity of the reaction during the FJH process, there is a risk of electric shock, explosion, burns or eye damage if the electricity supply system and reactor are not properly designed, managed and used with appropriate precautions (Supplementary Note 1). Safety is paramount to the equipment design and operation, and while a 10-g scale per batch has been realized on the laboratory scale, further scaling requires due precautions. In particular, FJH users should gradually increase the energy input rather than begin reactions at maximum input, especially for these reactions with intense outgassing²³.

Production of graphene and related carbon materials

FJH was originally developed for graphene production and is widely used in flash graphene synthesis^11,14,37. Related carbon materials are also synthesized via FJH, as the ultrahigh temperature makes it especially useful for carbon materials production: almost all carbon resources, including carbon black¹¹, hard carbon⁴², coal^43,44, coke²³, plastic³³, rubber⁴⁵, biomass⁴⁶, food¹¹, CO₂-derived amorphous carbon⁴⁷, pyrogenic carbon⁴⁸, and municipal solid waste⁴⁹, are graphitized below 3,000 °C. Moreover, the temperatures reached during FJH can cure material defects and volatilize out impurities, so that the quality of the carbon material produced is usually good to excellent⁵⁰. By tuning the feedstock and optimizing FJH parameters, several carbon morphologies, including graphene, graphite, carbon nanotubes, carbon nanofibres, and their doped or mixed counterparts, have been synthesized (Fig. 2a and Supplementary Table 2).

**Fig. 2: FJH for graphene and related carbon materials synthesis.**

Flash graphene production and characterization

Flash graphene is made when an amorphous carbon feedstock, commonly from low-value or waste feedstocks such as coal, coke, plastics or biomass, is heated to 2,000–3,000 °C to induce conversion into turbostratic flash graphene (Fig. 2b)¹¹. In contrast to ordered graphene^51,52, turbostratic graphene layers are randomly angularly displaced with respect to each other, resulting in increased interlayer spacing and lowering the van der Waals interactions between the layers. This disordered stacking arises from the kinetic trapping of the graphene layers due to the rapid heating and cooling rates^11,30.

Turbostratic graphene can have an arbitrary number of layers that remain optically and electronically decoupled. Hence, it has electronic properties more similar to those of monolayer graphene than multilayer graphene, owing to that interlayer decoupling^{11,53,54,55,56}. For example, both monolayer and turbostratic graphene are zero-bandgap semiconductors and can support massless Dirac fermions, in contrast to multilayer graphene. These similarities in electronic properties result in similar Raman spectra of monolayer and turbostratic graphene (Fig. 2c), which are distinct from the spectra of multilayer graphene or graphite¹¹ (Supplementary Table 3).

Flash graphene is characterized chiefly by Raman spectroscopy (Fig. 2c), which provides information on the crystallinity, defect density and turbostratic character. Specifically, the D, G, 2D (also known as G′)⁵⁷ and sometimes TS₁, TS₂ (turbostratic indicators) and M (AB-ordered stacking indicator) bands are examined. The 2D peak height is correlated with graphene crystallinity, and the I_2D/G peak ratio is examined as one metric to determine conversion from amorphous carbon to graphene. Conventionally, an I_2D/G of at least 0.3 is necessary for graphene conversion⁵⁸. Turbostratic flash graphene typically exhibits a 2D band full-width half-maximum (FWHM) of ~15–30 cm⁻¹ and a single Lorentzian peak shape¹¹, comparable to that of single-layer graphene (~24 cm⁻¹, ref. ⁵⁷). Multilayer graphene exhibits an asymmetric 2D band composed of multiple Lorentzian peaks, and the 2D band FWHM of stacked graphene increases with the stacking number such that four-layer graphene has a FWHM of ~40 cm⁻¹ (ref. ⁵⁷) and stacks even above 10 have a FWHM of up to 100 cm⁻¹ (ref. ⁵⁶). The best flash graphene samples show an undetectable D band and a high I_2D/G, demonstrating its low defect and high crystallinity (Fig. 2c, left). The TS₁ and TS₂ peaks are low-intensity and observed at ~1,880 cm⁻¹ and ~2,030 cm⁻¹, respectively. The low-intensity M peak occurs at 1,760 cm⁻¹. When the two TS peaks and the M peak are simultaneously present in a sample, turbostratic and ordered stacking are present³⁰. Carbon-13-enriched flash graphene shows an expected redshift in all of its Raman peaks due to the increased mass of the atoms, thereby decreasing their vibrational frequencies relative to those of conventional ¹²C-flash graphene⁵⁹.

The structure of flash graphene can also be characterized using X-ray diffraction analysis¹¹ and transmission electron microscopy (TEM, Fig. 2b). Conventional ordered graphene, as obtained by graphite exfoliation, is a mixture composed of mostly Bernal (AB-stacked) graphene with a minority of rhombohedral (ABC-stacked) layers⁶⁰. Each type has (100) and (101) X-ray diffraction peaks between ~41° and 46°; thus, most graphene samples show four small peaks in this region. In turbostratic graphene, the peaks associated with this region exhibit broadening compared to those of ordered graphene and thus overlap with each other³⁰. In addition, layers of turbostratic flash graphene exhibit mixed moiré patterns under TEM, giving additional confirmation of the turbostratic stacking^11,61 (Fig. 2b).

The physical and spectroscopic characteristics of flash graphene vary with reaction parameters used in the FJH process, including power, energy⁶¹ and the feedstocks (Supplementary Table 4). Graphene produced from low-energy input has a high interlayer spacing of ~0.341–0.343 nm, low crystallinity and high defect density, and exhibits wrinkles under TEM⁶¹. This rich morphology feature is distinct from graphene produced by chemical vapour deposition, which is often pristine⁶² and ultraflat^63,64. Graphene produced from moderate-energy FJH input has high crystallinity, a more sheetlike structure and lower defect density. With higher reaction energy, through either more intense flashes or longer duration, the turbostratic graphene is converted to ordered graphene (Bernal and rhombohedral) and graphite, with a lower defect density and interlayer spacing of 0.334 to 0.337 nm (refs. ^30,65).

Flash graphene has been demonstrated in various applications, especially those that require bulk amounts of graphene. For example, flash graphene is used as an additive in cement composites that increases compressive strength by ~25% with just 0.05 wt% flash graphene addition¹¹. It is also used in epoxy composites⁶⁶, concrete aggregate substitutes⁶⁷, lubricant additives⁶⁸, lithium-ion batteries⁶⁹ and conductive inks⁷⁰. In addition to the preparation of pristine graphene, FJH is also adopted for an in situ graphene coating on other materials to enhance their performances. For example, graphene-coated Cu particles show enhanced oxidation resistance⁷¹, and graphene-coated lithium iron phosphate has improved rate performance in batteries⁷².

Modified and functionalized flash graphene

FJH carbon feedstocks can produce modified graphene products and different carbon nanostructures. For instance, heteroatom-doped flash graphene (an effective material in battery electrodes and supercapacitors⁷³) is produced when a carbon feedstock is heated in the presence of a heteroatom-containing compound²¹ (Fig. 2d). The doped graphene exhibits increased defect densities and an increased interlayer spacing relative to non-doped graphene²¹. Heteroatom doping ratios have been achieved above 10 at% (ref. ²¹).

When the feedstock is composed of a high proportion of non-carbon elements, such as boron or nitrogen, 2D turbostratic boron–carbon–nitrogen ternary compounds can be synthesized⁷⁴. When a sealed flashing vessel is used, and the carbon feedstock is mixed with a fluorinated polymer, new carbon morphologies can be formed, including fluorinated nanodiamond, fluorinated graphene and fluorinated concentric carbon³⁵. These different structures have a time evolution in which the nanodiamonds form first at ~10 ms, and concentric carbon forms last at ~1 s. Using solid-state relays with millisecond-scale controllability, the reaction can be stopped in 1-ms increments along any point of the evolution. The heteroatom functionalization strategies of FJH can be extended to other materials such as 2D transition-metal carbides⁷⁵, underscoring the versatility of FJH.

High-surface-area graphene is desirable for applications including electrocatalysis, battery anodes and sorption media. The flash graphene usually has a lower surface area than the feedstock, owing to thermal-induced aggregation of pores⁷⁶. Typically, feedstocks with high surface area, such as carbon black, form higher-surface-area flash graphene¹¹ than do low-surface-area feedstocks such as coke²³. Porous carbon and flash graphene synthesized from bituminous coal exhibit effective adsorptive properties^77,78.

High-surface-area flash graphene can be produced from engineered high-surface-area precursors. For example, FJH of graphene oxide can produce highly defective and, thus, high-surface-area graphene⁷⁹, and ultrafine metal nanoparticles can be decorated on a reduced graphene oxide aerogel when metal precursors are preloaded on it⁸⁰. Similarly, when hollow mesoporous carbon spheres are used as feedstock, high-surface-area graphene hollow spheres can be produced with a surface area of up to 670 m² g⁻¹ (ref. ⁸¹). Another strategy uses an etchant during FJH to increase the surface area. In the presence of Ca(OAc)₂, FJH of plastics yields holey and wrinkled flash graphene with a surface area of up to 874 m² g⁻¹ and pore volume up to 0.32 cm³g⁻¹ (ref. ⁷⁶). In another case, KCl/K₂CO₃ salts have been used for the molten-salt synthesis of porous carbon from anthracite by FJH, achieving a surface area of 1,338 m² g⁻¹ and pore volume up to 9.95 cm³g⁻¹ (ref. ⁸²). As a comparison, activated carbon typically possesses a surface area of ~500–1,500 m² g⁻¹ and a pore volume of 0.3–0.8 cm³g⁻¹ (ref. ⁸³).

One-dimensional carbon nanostructure synthesis

One-dimensional carbon materials, including multiwalled carbon nanotubes (CNTs) and bamboo-like carbon fibres, are formed when a carbon feedstock is treated by FJH in the presence of a CNT growth catalyst, such as iron or nickel¹⁴ (Fig. 2e). The product formed from these reactions can be tuned by the FJH reaction temperature. A temperature of ~1,000 °C produces CNT without graphene; increasing temperature to ~2,000 °C leads to the formation of graphene–CNT hybrid structures; at temperatures beyond this, the ratio of CNT to flash graphene decreases such that virtually no CNTs remain by ~3,000 °C (ref. ¹⁴). Yield of CNT or hybrid morphologies, rather than 2D graphene morphologies, is estimated to be up to 90%, as observed by SEM imaging. Elemental purity of carbon can reach 98% even when waste plastics are used as feedstock¹⁴. Commercial multiwalled CNTs could range from 90% to 99% in carbon purity^84,85.

Further, FJH treatment can convert some of the CNTs into graphene. The conversion proportion is affected by the reaction duration and temperature. These CNT–graphene composites are effective reinforcing additives in epoxy composites⁸⁶. Similarly, in the presence of a heteroatom-containing compound, heteroatom-doped CNT has been synthesized by FJH^32,87. When ammonia borane is used as the feedstock, boron nitride nanotubes can be produced⁸⁸. Other carbon forms with different graphitization degrees and morphologies can be produced by FJH, such as hard carbon^89,90 and graphitic carbon cages⁹¹.

Inorganic materials production

FJH is a versatile tool for phase engineering^15,35,92 as it has a broadly tunable energy input capable of reaching temperatures >3,000 °C. FJH-synthesized materials with rationally engineered morphologies and electronic structures have unique properties that position them, for example, as promising high-performance catalysts in renewable energy devices. This section discusses the synthesis of inorganic materials, including metastable materials, solid-state materials and nanocatalysts.

Synthesis of metastable materials

Metastable materials can be produced through rapid heating or cooling during FJH, as the system does not have enough time to reach equilibrium⁹³. For instance, turbostratic flash graphene is a metastable phase as opposed to the more stable Bernal-stacked structures⁹⁴. FJH is emerging as a tool for engineering other metastability, as well, such as structural dislocation and defects^95,96.

The temperature tunability of FJH allows access to many metastable phases by making them thermodynamically preferable. A notable example is the phase conversion of transition-metal dichalcogenides⁹², where bulk conversion of molybdenum dichalcogenides (MoS₂) and tungsten dichalcogenides (WS₂) from 2H phases to 1T phases was achieved in milliseconds, reaching a phase ratio of up to 76%. The 1T phases are metastable with higher free energy than the 2H phases. FJH induces structural changes in the 2H phase, particularly the formation of sulfide vacancies, which reverses the thermodynamic preference⁹². A similar approach is used for the phase-controlled synthesis of transition-metal carbide nanocrystals¹⁵. The ultrahigh-temperature capability of FJH enables the carbothermic reduction of metal oxides, leading to the synthesis of metal carbides within 1 s (Fig. 3a, left). By controlling pulse voltages, phase-pure molybdenum carbides, including stable β-Mo₂C and metastable α-MoC1−x and η-MoC_1−x, can be selectively synthesized (Fig. 3a, right). Carbon vacancies introduced during the FJH process are the structural factor for the phase transition of carbides¹⁵.

**Fig. 3: FJH for phase-controlled synthesis of functional inorganic materials.**

Beyond vacancies, FJH can precisely tune the morphology and size of inorganic nanomaterials, providing another route for phase engineering, such as in the phase transformation synthesis of high-surface-area corundum nanoparticles (Fig. 3b, left)³⁹. Aluminium oxide (Al₂O₃) has an unusual surface-area-dependent formation energy: α-Al₂O₃ is the thermodynamically stable phase of coarse crystals, whereas γ-Al₂O₃ has a lower surface area, causing nanocrystalline Al₂O₃ to usually crystallize in the γ-phase⁹⁷. The pulsed direct current input in the FJH process creates resistive hotspots at the interfaces between γ-Al₂O₃ nanoparticles (Fig. 3b, left), leading to controlled coarsening and an accompanying phase conversion from γ-phase to δ′-phase and then to α-phase (Fig. 3b, right).

The ultrafast cooling rate of the FJH process can kinetically retain the metastable phase at room temperature. This process is demonstrated by the kinetically controlled synthesis of metallic glass nanoparticles¹⁸, as metallic glass can be obtained by rapid quenching⁹⁸. Metal precursors loaded on a carbon substrate are subjected to FJH, rapidly raising the temperature; the resulting alloy melts and cools at an ultrafast rate (10⁴ °C s⁻¹), vitrifying into glassy nanoparticles. Palladium-based and platinum-based metallic glass nanoparticles have been produced with this technique¹⁸.

Rapid sintering and solid-state synthesis

Solid-state synthesis, also known as the ceramic method, is a reliable and versatile method for materials production⁹⁹ and typically yields thermodynamically stable phases⁹⁹. FJH is useful in solid-state synthesis owing to its ultrahigh-temperature capability and rapid heating rates, which secure the thermodynamic spontaneity of many reactions and enable fast diffusion and reaction kinetics. However, most feedstocks are not conductive enough for FJH, necessitating unique designs to deliver electricity and heat to the feedstock.

Graphite sheets are used as the heating element for ceramic sintering (Fig. 3c, top). Resistance heating is already widely used in flash sintering¹⁰⁰ and spark plasma sintering¹⁰¹. Joule-heating-based sintering techniques can be faster, achieve higher temperatures and require less expensive apparatus³⁸. In a typical high-temperature sintering set-up, the pressed pellet of a ceramic precursor powder is sandwiched between two woven graphite sheets, which rapidly heat the pellet by conduction and radiation (Fig. 3c, bottom)³⁸. The sintering can be completed in seconds, making it especially promising for solid-state electrolytes to prevent loss of volatile elements such as lithium^38,102. Rapid Joule-heating-based sintering has also been used to construct interfaces between solid electrolytes and cathodes, overcoming large interfacial resistances¹⁰³. In addition, Joule-heating-based sintering has been previously explored in structural ceramic sintering, such as alumina ceramics from corundum nanoparticles³⁹; however, the densification is relatively poor because of the limited sintering time. Joule-heating-based sintering would be useful for sintering ceramics that have volatile element components but less stringent requirements for densification.

Flash-within-flash (FWF) Joule heating has also been developed for inorganic synthesis^20,104 (Fig. 3d). The FWF process involves two quartz vessels: an outer vessel filled with conductive additives and an inner vessel loaded with feedstocks for the solid-state synthesis (Fig. 3d, left). The current applied to the outer vessel generates intense heat, which is then transferred through an inner vessel to the inner feedstock by conduction and radiation to drive chemical reactions. Thus, FWF is an indirect heating method. The FWF can be used multiple times or in an anion-exchange mode (Fig. 3d, right). FWF is a versatile, efficient and scalable method for the production of phase-selective, single-crystal bulk powder materials at the gram scale, as demonstrated by the synthesis of ten transition-metal dichalcogenides, three group XIV dichalcogenides and nine non-transition-metal dichalcogenide materials²⁰. In the FWF configuration, metal precursors are not in contact with the carbon in the outer tube, so metal carbide formation is mitigated. These two designs — woven graphite sheet heating elements and outer–inner tube FWF configurations — enable the use of non-conductive materials, or even materials that are too conductive, greatly expanding the versatility of the FJH process. These methods rely primarily on thermal conductive heating, but rapid heating rates and efficiencies differentiate them from traditional furnace heating methods.

Carbon-supported inorganic nanocatalysts

Inorganic nanomaterials can be synthesized by FJH, typically using carbon materials such as carbon black¹⁸ and carbon nanofibre¹⁰⁵ as conductive additives and substrates, as their electrical resistance makes them suitable for Joule heating. After loading precursors onto the carbon support, the current passing through the carbon instantly generates heat. The heat leads to the decomposition, reaction and fusion of the precursors, and then solidification to form nanoparticles during the cooling stage¹⁰⁵. Various inorganic nanomaterials have been produced using this method, including single metals^{96,106,107,108,109}, multimetallic alloys^110,111, metallic heterostructures^112,113, high-entropy alloys^105,114,115, single atoms¹¹⁶, and metallic compounds¹¹⁷ such as sulfides¹¹⁸, oxides^119,120, selenides¹²¹ and carbides^117,118,122.

The precise control of reaction time down to milliseconds during FH allows for the formation of dispersed, ultra-small particles by preventing their agglomeration¹⁰⁵. The carbon supports are vital for the dispersion and stability of nanoparticles by providing numerous nucleation and binding sites^123,124. Moreover, as FJH is a dry process, the surface of the as-produced nanomaterials remains clean, in contrast to materials synthesized by wet chemistry, which are often contaminated by surfactants and capping agents¹⁸. Their cleanliness and structure make the FJH-produced materials suitable for use as catalysts in a variety of applications, including thermal catalysis^114,115, electrocatalysis^{95,107,111,118,125}, environmental catalysis¹⁰⁸ and photocatalysis^126,127.

FJH is primarily used for solid-state synthesis, but with proper reactor design it can be integrated into wet chemistry and gas reactions. A wet-interfacial Joule heating approach has been proposed to synthesize various nanomaterials from solution feedstocks^128,129. By instantaneous evaporation of solvents that are on the carbon heater, synthesis is completed in <1 s (ref. ¹²⁸). For thermochemical synthesis, a non-equilibrium synthesis that uses programmable electric current to rapidly heat and quench gas reactions has been demonstrated¹³⁰, achieving methane pyrolysis with high selectivity to C₂ products.

Resource recovery and waste upcycling

Thermal treatment is commonly used in resource recovery and waste upcycling¹³¹. However, increased energy efficiency and reduced costs and emissions are needed to ensure that the value of the recovered products can offset the process costs. FJH is therefore being explored for the recovery of metals from waste streams such as electronic waste (e-waste), industrial wastes and spent batteries^{22,27,36,132,133}, the upcycling of inorganic wastes^22,134, and the conversion of carbonaceous waste (such as plastics and rubbers) into graphitic materials^12,33.

Metal recovery and inorganic waste upcycling

Metals are often the most valuable and recoverable components of waste¹³⁵. Their recovery involves altering their chemical forms, speciation, and distribution to enable their separation from wastes based on property differences. FJH, with its high-temperature capability and cost-effectiveness, makes these conversions feasible and economic³⁶.

Critical metals recovery from e-waste and industrial wastes

FJH in metal recovery was originally demonstrated in urban mining, through the gram-scale recycling of precious metals from e-wastes — specifically, waste printed circuit boards³⁶. The difference between the vapour pressures of metals and those of the substrate materials (carbon, ceramics and glass) allows the separation of metals, which is called evaporative separation. Precious metals in e-wastes are heated and evaporated by ultrahigh-temperature FJH; then the metal vapours are transported and condensed in a cold trap (Fig. 4a). With the assistance of chlorination (converting the metal into its chloride by reacting with chlorinating agents), recovery yields of >80% were achieved for rhodium, palladium, silver and gold within 1 second (ref. ³⁶).

**Fig. 4: Metal recovery and waste upcycling by FJH.**

Integration of FJH into an electrothermal chlorination or electrothermal carbochlorination process is performed for selective separation of critical metals from waste feedstocks¹³⁶. The electrothermal chlorination process leverages the differences in the free energy of formation of metal chlorides as well as the kinetic selectivity due to the ultrafast heating and cooling capability of FJH. This process has been demonstrated in the recovery of gallium, indium and tantalum from e-wastes with purities >95% and yields >88% (ref. ¹³⁶).

FJH has also been used to recover rare earth metals from coal fly ash²⁷, a by-product of coal combustion. In this process, FJH thermally converts the hard-to-dissolve rare earth phosphates into rare earth oxides and metals with high solubility (Fig. 4b). The FJH activation increased the recovery yield of rare earth metals roughly twofold compared with directly acid leaching the raw materials. This method is also applicable to other wastes like e-wastes and bauxite residues (red mud)²⁷.

Recycling and regeneration of spent batteries

Battery cathodes can be recycled through hydrometallurgy, which typically involves acid leaching of metals. However, the transition metals in active cathode materials with high valence states lead to low leaching efficiency¹³⁷. FJH has been applied to make this process more efficient by heating the black mass, which is the combined anode and cathode waste routinely used in the recycling industry¹³² (Fig. 4c). This process led to the thermal decomposition of the compact solid–electrolyte interphase (SEI) and the carbothermic reduction of the spent cathode materials (LiNi_xMn_yCo_1−x−yO₂, LiCoO₂, LiNi_xCo_yAl_1−x−yO₂ and LiFePO₄) into their lower oxidation state or metallic form. With this FJH activation, the metal recovery yield was improved from <35% to >98% (Fig. 4c).

Direct recycling (regeneration of battery materials without destroying their crystal structures) has gained attention due to potential reduced environmental impacts and economic costs relative to destructive recycling such as hydrometallurgy¹³⁸. High-temperature calcination that gasifies organics for graphite anode regeneration is technically straightforward¹³⁹, but as graphite is less valuable than cathode materials, furnace-based extended calcination is often not economically viable¹³⁹. FJH has been applied to decompose the impurities and regenerate the entire graphite anode in 1 second, which significantly reduces the energy consumption and carbon emission compared with high-temperature calcination recycling¹⁴⁰ (Fig. 4d). The recycled anode preserves the graphite structure while being coated with an SEI-derived carbon shell, contributing to high battery performance¹³³.

Another FJH-based process involves converting the loose SEI layer that is coated onto the degraded graphite into a compact and mostly inorganic mass that encloses active lithium, leading to >100% initial Coulombic efficiency, superior to commercial graphite¹⁴¹. A similar strategy was applied to regenerate spent cathode carbon blocks of aluminium electrolytic cell¹⁴². In addition to anode regeneration, FJH can achieve the direct recycling of cathodes, such as the relithiation of LiCoO₂ and repair of its crystal structure¹⁴³. FJH was also combined with magnetic separation and solid-state relithiation to restore fresh cathodes from waste cathodes, with battery metal recovery yields of ~98% (ref. ¹⁴⁴).

Upcycling inorganic wastes into value-added materials

Recycling inorganic wastes that are less valuable than critical metals, such as glass and silicon, is often less profitable. As a result, these waste streams are usually directly landfilled instead of recycled. However, similar to its application in the synthesis of inorganic materials, FJH can be used to convert inorganic waste to value-added inorganic materials. As low-value materials often constitute a major part of inorganic wastes, their upcycling by FJH represents a promising path toward their secondary utilization.

For example, FJH has been used to upcycle glass-fibre-reinforced plastics into silicon carbide²², a high-performance reinforcement and semiconducting material. Waste glass-fibre-reinforced plastics were mixed with waste carbon-fibre-reinforced plastics as conductive additives (Fig. 4e). FJH rapidly converted the glass fibre to SiC by carbothermic reduction (Fig. 4e). The phase of SiC can be controlled by the FJH process, similar to the phase-controlled synthesis of molybdenum carbides¹⁵. The SiC powder obtained was further used as an anode for lithium-ion batteries²². In another example, the rapid Joule heating process converts photovoltaic silicon waste into silicon nanowire electrodes for lithium-ion batteries¹³⁴.

Carbonaceous waste upcycling

FJH production of graphene was initially optimized with amorphous carbon and coal¹¹, but many other low-value carbonaceous waste streams³³ have been upcycled into graphene via FJH. Upcycling carbonaceous waste into high-value products such as graphene can economically incentivize the responsible disposal and recovery of resources¹⁴⁵. The value of the product required to offset process costs is dependent on the demonstrated application, and several promising pathways are discussed here.

Upcycling of waste plastics and rubbers into flash graphene

FJH can be used to upcycle waste plastics³³ and other petrochemical wastes, such as rubber⁴⁵, pyrolysis ash¹⁴⁶ and asphaltenes³⁴. For example, mixed plastic wastes have been converted to graphene via FJH in a process that does not require sorting or washing (which involves high process costs)³³. Graphene derived from waste plastics has been tested in applications of lubricants⁶⁸, automotive composite applications¹², 3D printing and corrosion resistance³⁴, and electromagnetic field absorption¹⁴⁷. Graphene derived from coal has been used as a total substitute for concrete aggregates, producing concrete that is 25% lighter and has superior mechanical properties⁶⁷.

Because of the resistive heating nature of FJH, samples must be conductive enough to support the rapid discharge of current required to achieve high temperatures during the process. As waste plastics are not conductive enough, conductive additives are needed. Typically, 10–20 wt% of conductive additives, such as carbon black, tyre-carbon black, charcoal or coal, can be added to reduce the resistance to <10 Ω; however, this can be a costly additive when processing waste streams. A two-step FJH strategy has been developed to use as little as 2–3 wt% of carbon black in the conversion of ground waste plastics into graphene¹² (Fig. 5a). In this two-step method, a longer-duration (10-s), lower-current discharge is conducted to carbonize the carbonaceous feedstock, increasing the conductivity of the material and resulting in volatile outgassing. After this low-current FJH, where some volatile mass is lost dependent on feedstock, the sample resistance is <10 Ω, allowing high-current FJH to occur, as shown in the graph of current versus time in Fig. 5a. Raman spectroscopy shows defective graphitic structure after low-current FJH, which is then annealed into high-quality turbostratic graphene after high-current FJH (Fig. 5b). The carbon black additives are simultaneously converted to graphene, so their separation is not required¹².

**Fig. 5: Upcycling of waste plastics and biomass into flash graphene and clean hydrogen.**

The longer heating duration of the two-step FJH process allows alternative products to be synthesized. For example, the incorporation of <1 wt% salts such as FeCl₃ catalysed the formation of 1D carbon nanotubes and nanofibres with controllable diameters. By parameter tuning, hybrid morphologies of graphene domains embedded with 1D nanofibres were also produced, which resulted in superior composite mechanical properties when tested in epoxy composites¹⁴. The incorporation of calcium salts can also function as blowing agents and proppants to increase and maintain the specific surface area during the FJH process. Through the evolution of gases, porosity and wrinkles can be included in the material, resulting in holey, wrinkled flash graphene⁷⁶.

Upcycling of waste plastics into hydrogen and chemicals

Carbonization occurs during the low-current FJH of plastics, forming hydrocarbon volatiles including substantial proportions of H₂ gas (Fig. 5c). High yields of H₂ were observed regardless of plastic type (Fig. 5d), with lower initial sample resistances resulting in more complete hydrocarbon breakdown to form higher yields of H₂ and flash graphene. Hydrogen efficiencies as high as 93% were observed, and no CO₂ is produced during the FJH of polyolefins. Compared with traditional pyrolysis²⁶, the higher heating rate of FJH results in a substantially different product distribution (90 vol% H₂). FJH is, therefore, a catalyst-free method to produce clean H₂ from mixed waste plastics, with the costs potentially offset by the co-production of high-value graphene.

Depolymerization is another widely adopted strategy for waste plastic upcycling¹⁴⁸. An electrified spatiotemporal heating process based on the rapid Joule heating technique has been tested in the pyrolysis of commodity plastics¹⁴⁹. This catalyst-free, far-from-equilibrium thermochemical depolymerization method promotes depolymerization while suppressing unwanted side reactions, leading to high-yield monomer recovery: 36% for polypropylene and 43% for poly(ethylene terephthalate). H-ZSM-5 was later introduced as a catalyst to efficiently deconstruct polyolefin plastic wastes into light olefins C₂–C₄ (ref. ¹⁵⁰). In that demonstration, the pulsed current input was critical for producing a narrow distribution of gaseous products.

Upcycling of biomass into flash graphene

Although thermal conversion of biomass to biochar has been widely adopted¹⁵¹, the conversion of biomass to high-value carbon materials such as graphene represents a value-added upcycling route. Biomass such as sawdust and straw¹⁵², lignin¹⁴⁷, waste food¹¹, hair⁴⁶ and even mixed municipal waste⁴⁹ can be upcycled into graphene by FJH. No catalyst is required. However, biomass is not conductive, so a conductive additive such as carbon black is required. The biomass usually contains a high oxygen content, such that massive pyrolytic volatiles are released during the FJH process, which accounts for 60–70% of carbon emissions¹⁵². A two-step process has been developed to allocate energy efficiently³⁷ and address these emissions. Initially, pyrolysis is used to release biomass pyrolytic volatiles; subsequent FJH reaction is carried out to optimize the flash graphene structure (Fig. 5e). An integrated automatic FJH system has also been built (Fig. 5f), enabling the continuous production of biomass-derived graphene at the productivity of 21.6 g h⁻¹ (ref. ³⁷).

Environmental remediation

Environmental remediation processes, such as thermal treatment, thermal desorption and high-temperature vitrification, can have high energy usage and by-products. Improving energy efficiency, reducing by-product emissions and using renewable energy are key research areas toward sustainability in environmental remediation¹⁵³. Owing to its versatility, energy efficiency, widely tunable temperature range (up to 3,000 °C) and lack of secondary waste, FJH has been applied in environmental remediation, including as a thermal process for the decontamination of hazardous wastes and remediation of soil, and in material synthesis for pollution degradation, as discussed in the following sections.

Decontamination of hazardous wastes

Thermal treatment is widely used in hazardous waste decontamination and environmental remediation^153,154. FJH can achieve a wide temperature range, which is sufficient to degrade organic pollutants and remove heavy metals by evaporation. Heavy metals (including Cd, As, Pb, Ni and Co) have high vapour pressure (Fig. 6a), enabling their evaporative removal with efficiencies of 70–90% at temperatures just below 3,000 °C. For example, during the urban mining of precious metals by FJH, toxic heavy metals, such as Hg, As, Cd, Pb and Cr, are removed by evaporation and then captured by condensation³⁶. The heavy metal content in the residual waste is reduced to within safe limits after two to three FJH pulses under 120 V FJH. Heavy metals in coal fly ash¹⁵⁵ have also been removed during treatment with FJH (Fig. 6a), allowing the purified coal fly ash to be used as a low-carbon cementitious material with reduced heavy metal leaching¹⁵⁵. FJH has also been used to recycle hazardous zinc-enriched hyperaccumulators into graphitic materials while removing 98.6% of the zinc¹⁵⁶.

FJH has been shown to be an effective technique for the electrothermal mineralization of PFAS¹⁵⁷. A common strategy to remove PFAS in water is adsorption, using sorbent-like granular activated carbon¹⁵⁸. Traditional thermal methods for treatment of this PFAS-contaminated carbon, such as incineration, are energy intensive and can release hazardous gaseous fluorocarbons; they therefore require secondary flame scrubbers to mitigate the gaseous release¹⁵⁹. In the presence of sodium or calcium hydroxides, PFAS sorbed on granular activated carbon are converted into non-toxic fluoride salts by FJH (Fig. 6b). High fluorine conversion efficiencies of >96% are achieved within 1 s, with undetectable gaseous fluorocarbons, and it is much faster than traditional incineration¹⁵⁷.

Soil remediation

Thermal desorption¹⁵³, such as in thermal conduction heating and electrical resistive heating, is a practical method for soil remediation. Traditional thermal desorption technologies usually operate at temperatures <500 °C, so they are only suitable for the remediation of volatile or semivolatile contaminants¹⁶⁰. FJH can be applied to soil remediation to remove multiple pollutants simultaneously²⁵: heavy metals are removed by evaporation, while persistent organic pollutants such as polycyclic aromatic hydrocarbons are removed by carburization (Fig. 6c).

Compared with conventional thermal techniques, the electrothermal remediation processes are much faster, within seconds rather than days or even months, and they have higher degradation capability for pollutants²⁵. For example, the removal efficiencies of tested heavy metals such as Cd, Hg and Pb were >80% in a single FJH pulse²⁵; the degradation degrees of tested polycyclic aromatic hydrocarbons including pyrene, fluorene and benz[a]anthracene were >93% in three FJH pulses²⁵. The high-temperature electrothermal process of FJH can potentially be integrated with vapour extraction pipes for in situ remediation²⁵. However, in contrast to traditional electrical resistance heating techniques that rely on the thermal conductivity and slow heat diffusion of the soil itself, in the FJH process, premixing conductive carbon (such as biochar) soil particles is required to provide sufficient conductivity for the Joule heating. The conductive additives can be recycled by mechanical sieving based on the size differences between the soil and the carbon additives and then reused to minimize the materials consumption of the remediation process²⁵. Currently, FJH for soil remediation has been demonstrated on a laboratory scale; further field tests are required to assess the soil properties.

FJH has also been used for the remediation of PFAS in soil at the kilogram scale²⁴ (Fig. 6d). With biochar as a conductive additive, FJH is used to ramp the soil temperature to >1,000 °C in seconds. PFAS then reacts with in situ calcium compounds to form calcium fluorides and other alkaline earth fluorides, which are non-toxic and the natural mineralized form of fluoride in the environment. High PFAS removal efficiencies of >99% have been achieved with this method without the use of any external reagents. The soil retains most of its physicochemical properties after the FJH remediation process^25,79. This method is a proof-of-concept work at the bench scale, and further scaling up is required to assess its applicability and economic viability at larger scales.

Functional materials for pollutant degradation and removal

Environmental catalysis for pollutant degradation can be synthesized via FJH without producing secondary waste during their synthesis. For example, iron-based material is one of the most important catalysts for the advanced oxidation process to degrade organic pollutants in wastewater¹⁶¹. With soft carbon as the conductive additive, FJH was used to transform a low-grade, inexpensive FeS mineral into nanoscale Fe(0)/FeS heterostructures embedded in thin graphene layers¹⁰⁸. With Fe(NO)₃ as a precursor, Fe single atoms and nanoparticles were also produced¹⁰⁹. These materials catalysed an advanced oxidation process for chloramphenicol degradation^108,109. For example, the Fe-based materials show a chloramphenicol removal efficiency of >94% via catalytic activation of sodium peroxydisulfate¹⁰⁸. FJH was also applied to synthesize carbon-coated Mg(0)/MgO, which is used in activating peroxydisulfate to degrade sulfamethoxazole, an antibiotic present in wastewater¹⁶².

In addition to acting as catalysts for oxidation processes, low-valent or zero-valent metals are promising for reductive remediation¹⁶³. During FJH synthesis, a carbon layer is coated on the Fe(0) nanoparticle, which prevents its surface oxidation¹⁶⁴. This Fe–C composite demonstrates superior reductive remediation of multiple pollutants in wastewater, such as water-soluble PFAS, Cr(VI) and Sb(V). By further incorporating Mg with the Fe, a reductive Fe–Mg bimetallic nanocomposite can be synthesized using FJH¹⁶⁵. This composite leads to the efficient removal of heavy metals from wastewater, including Cr(VI), Sb(V), Ni(II) and Cu(II), with a 2–3 times increase in efficiency compared with monometallic composites.

FJH can also produce the materials for the removal of volatile organic compounds, such as carbon-supported Ag–Co₃O₄ bimetallic catalyst for the catalytic oxidation removal of formaldehyde¹⁶⁶. In another example, Joule heating serves as the thermal source to drive the catalytic oxidation of volatile organic compounds using Pt/CeO₂ as the catalysts. This Joule-heating-based catalytic system exhibits an ultralow input power¹⁶⁷, 87% lower than that of a conventional heating furnace.

Sustainability and technoeconomic considerations

Quantifying resource use and waste emissions throughout the production, use, or disposal aspects of the FJH allows for direct comparison of different methods. Life-cycle assessment (LCA) and technoeconomic analysis (TEA) have been used to understand the impacts of FJH. These assessments indicate that the high efficiency, short duration, solvent-free nature and minimal heat loss of FJH processes result in substantial reductions in emissions and energy consumption compared with radiative heating or wet-chemistry methods. However, the maximum scales of the processes differ, so there is some uncertainty. This section examines the sustainability of graphene production and waste management using FJH and notes the limitations of these approaches.

Graphene and inorganic materials production

LCA typically includes raw material and resource extraction, which can have high environmental burdens. These impacts can be minimized or even eliminated when using waste feedstocks for flash graphene and other material synthesis by FJH (Fig. 7a), as it requires no graphite mining or solvents in processing. Using waste materials, such as pyrolysis ash¹⁴⁶, waste plastics^12,14, waste printed circuit boards³⁶, e-wastes¹³⁶, retired wind turbine blades²², coal fly ash¹⁵⁵ and spent rechargeable batteries¹³², as feedstocks has provided a substantial reduction in energy intensity and emissions for many FJH processes.

**Fig. 7: Life-cycle assessment and technoeconomic analysis of processes conducted by FJH.**

For example, production of flash graphene from end-of-life vehicles has been compared with two common, commercially scaled production methods including graphite exfoliation by sonication or chemical oxidation and reduction, based on a functional unit of 1 kg of graphene powder produced (Fig. 7b–d). There was a substantial reduction in the required co-feedstocks owing to the use of waste materials, and the FJH process overall resulted in >80% reductions in all burdens including greenhouse gas emissions, cumulative energy demand, and water use (Fig. 7b–d). Similarly, the FJH process for upcycling waste biomass into flash graphene results in ~90% reduction in the environmental footprint, especially carbon emissions and freshwater use, when compared with traditional production processes like electrochemical exfoliation¹⁵². FJH synthesis of other graphitic nanomaterials, such as carbon nanotubes and nanofibres, also has an 86–94% reduction in emissions and energy demand versus conventional production methods¹⁴. Similar results were observed for transition-metal dichalcogenide production compared with current production methods²⁰.

In solid-state syntheses, FJH competes primarily with chemical exfoliation¹⁶⁸, chemical vapour deposition^169,170, and electrochemical intercalation¹⁷¹ (Fig. 7e). Chemical vapour deposition and exfoliation methods require expensive feedstock costs, long reaction times and high energy usage, and low reaction yields are typical¹⁷⁰. In contrast, FJH has short reaction times and can use low-cost or waste feedstocks with minimal waste by-products, so there can be a substantial decrease in the production cost of flash graphene and related products^49,152. A TEA of preparation shows that flash graphene can be made at US$0.16 kg⁻¹, markedly lower than the ultrasonication of graphite and reduction of graphene oxide²³, which costs up to US$44 kg⁻¹. Similarly, a preliminary product cost of 2D materials synthesis shows that the FWF FJH offers large cost savings over other synthesis methods, such as autoclave and chemical vapour transport²⁰. For example, it takes ~US$600 to synthesize 100 g of MoSe₂ materials by FWF FJH, whereas this number can be up to ~US$4,500 by chemical vapour transport²⁰.

Waste upcycling and decontamination

The application of FJH for metal recovery, waste decontamination and environmental remediation has advantages relative to various widely used methods, based on LCA and TEA. For example, the recycling of battery cathode metals by FJH has reduced water consumption, energy usage, greenhouse gas emissions and costs compared to virgin mining, hydrometallurgical processing, and pyrometallurgical processing¹³² (Fig. 7f-h). Similarly, compared to synthetic graphite production and high-temperature calcination recycling¹³³, the flash recycling of graphite anode materials had substantial reductions in water and energy consumption and greenhouse gas emissions. These reductions were attributed to the rate and energy efficiency of the FJH process. The environmental impacts and costs of upcycling plastic-reinforced glass fibre by FJH have also been compared with other approaches²². The operating cost of FJH was equivalent to ~0.2% and ~3.4% of the solvolysis and incineration processes, respectively, to recycle the same amount of waste plastic-reinforced glass fibre and produce the same amount of silicon carbides.

FJH seems to be economically feasible and a viable approach to waste decontamination and environmental remediation. In an LCA of heavy metal removal from coal fly ash by FJH and its reuse in cement¹⁵⁵, there was a 30% reduction in greenhouse gas emissions compared with direct landfilling of the coal fly ash. These results indicate that the coal fly ash can be used as a lower-carbon-footprint cement. A TEA of FJH in soil remediation demonstrated that the electrothermal remediation process shows a 5–70% reduction in operating expenses relative to the different established methods, such as thermal desorption, soil washing and chemical oxidation, that are used at the industrial scale²⁵. Similar analyses have been conducted on electrothermal PFAS mineralization by FJH^24,157, showing its substantial reduction in energy consumption, greenhouse gas emissions, water consumption and operation costs when compared with existing methods such as incineration, ball milling and chemical oxidation.

Limitations of current LCA and TEA

As with all preliminary LCA and TEA, the current analyses of the FJH process have limitations. These include assumptions or omissions regarding scalability, transportation, waste feedstock availability and cost, as well as disposal of waste by-products. Widespread adaptation of sustainable technologies based solely on top-line, preliminary LCA and TEA data might result in over-estimation of environmental benefits or efficiency improvements, owing to altered consumption behaviours^172,173. This rebound effect in the circular economy is a result of projected efficiency gains of new sustainable technology being unrealized, owing to increased consumption of goods once the technology has reached large scales¹⁷⁴.

As many of the published LCAs conducted on the FJH processes are preliminary in scope and make assumptions about scale and adoption, most do not consider rebound effects. For example, a lower cost of graphene would logically result in increased consumption, which could result in an overall increase in burdens and footprint attributable to the graphene market, even with the improved efficiency of the FJH process. This would be a direct rebound effect, but indirect rebound effects could also be observed in which the increased efficiency and lowered cost might result in an increased prevalence of unrelated manufacturing processes, such as increased cathode production due to the implementation of cheaper anode recycling. Both direct and indirect rebound effects can be difficult to predict, and they also apply to TEA, not only to LCA projections¹⁷⁵. Further examples of rebound effects include the increased demand for biomass-derived FJH graphene, resulting in higher overall biomass consumption and thus greater land use burdens or increased emissions from transportation.

Scientists can help to acknowledge these possibilities by conducting thorough uncertainty analysis in LCA and TEA, as well as considering a wide array of scenarios, including process energy sources, materials transportation, and comparison to associated primary production routes that might have increased use rather than a decline in use¹⁷⁶. Furthermore, detailed LCA and TEA with a broader scope, using different assumptions, comparing different methods or analysing different scale FJH processes, can enhance the quantified understanding of cost and burden efficiency. Overall, conducting LCA and TEA during the early stages of FJH developments, optimization and scale-up can help to identify steps of high burden, as well as benchmark or suggest the sustainability in quantifiable terms.

Summary and future perspectives

FJH has emerged as an efficient electrification technology for materials production, metal recycling, waste upcycling and environmental remediation. Its hallmark features, including high energy efficiency, short reaction duration, solvent-free processing, minimal heat loss, versatile operation models and compact reactor design, provide advantages over traditional combustion-based heating or wet-chemistry methods. Despite progress in the development and application of FJH, efforts are needed to elucidate the underlying mechanisms, better harness its characteristics and advance FJH toward industrial-scale implementation.

More experimental and theoretical analyses are needed to uncover the mechanism and improve the controllability of FJH. Advanced characterization under realistic in operando reaction conditions will be necessary to clarify the reaction and conversion details¹⁷⁷. FJH is a multi-physical phenomenon that combines electric fields, thermal fields, and intense light emission. A comprehensive understanding of the multifield coupling effect is crucial for enhancing its controllability. Further, improving temperature control is necessary for better uniformity of the reaction conversions.

FJH complements existing heating techniques, such as furnace heating, microwave heating and induction heating. Its differences from these techniques offer new possibilities for researchers working in diverse disciplines applications, as demonstrated by the successful application of FJH in materials production, chemical syntheses and environmental remediation. In particular, few applications in wet chemistry and gas reactions have been reported, but these warrant further investigation as the ultrafast and highly localized heating of FJH could offer distinct kinetic and thermodynamic advantages over traditional heating methods. FJH for waste recycling and environmental remediation also requires more detailed comparative studies than established methods before deployment.

The scale-up of the FJH process and its industry-scale implementation and automation are pivotal to realizing its use in energy-efficient applications. The prototype productivity of flash graphene production had been reported at larger scales, but the FJH synthesis of inorganic materials is at the gram-scale level in the laboratory. Ideally, lessons on scalable FJH technology in graphene production can be adapted for the mass production of other materials. Scale-up is also needed for waste recycling and environmental remediation, given the quantities of solid waste generated. For instance, coal power plants currently produce >750 million tonnes of fly ash per year¹⁷⁸. For further scale-up, the electricity supply system needs design upgrades to ensure high and rapid power input⁴¹. For pilot-scale or industrial implementation of FJH, waste heat recovery and off-gas treatment systems should be integrated into FJH systems, allowing heat and gas to be recovered from the FJH process. This design requires collaboration between academia and industry to bring FJH into large-scale practice.