"Understanding the Synthesis of Turbostratic/Flash Graphene via Joule Heating" is an open-access review published in the MDPI journal Materials in June 2025 (received 19 May, accepted 16 June). Its authors are Faisal Mahmood, Christian Fabrice Magoua Mbeugang, Furqan Asghar, Xing Xie, Dan Lin, Dongjing Liu and corresponding author Bin Li, with funding from the National Natural Science Foundation of China, the State Key Laboratory of Coal Combustion and Anhui Agricultural University. The paper reports no new experiments; its data statement says no new data were created or analysed. Its value to this dossier is that it gathers the case for how Turbostratic Graphene forms during Flash Joule Heating (FJH), and it states plainly where that case is still unproven.
What the review sets out to do
The authors say plenty of reviews already survey graphene production in general. They narrow the focus to one route, Joule heating with turbostratic ("flash") graphene as the product, and cover four things: the history of the Joule effect, how laboratory set-ups work, how turbostratic graphene is identified, and what is known about why it forms. A comparison table places flash graphene among other methods. It rates the method as producing bulk powder with poor control over layer number, moderate scalability, low-to-moderate cost and moderate-to-high quality. Chemical vapour deposition, by contrast, is rated excellent for layer control but high in cost. A later table in the review rates flash graphene's scalability differently (see below).
From Joule's law to the flash
The historical thread is brief. It starts with James P. Joule reporting to the Royal Society of London in 1851, and the heat law Q = I²Rt (see Joule Heating). The review then moves through everyday resistive heating (rice cookers, irons, welding) to materials work, citing ceramics sintering, powder metallurgy and nanoparticle fabrication (compare Flash Sintering (Field-Assisted Sintering)). It credits the method with avoiding furnaces, solvents and process gases, and with residence times under 10 s. These efficiency advantages are asserted by citing earlier studies. The review itself gives no energy-consumption comparison.
Two ways to wire the sample
The review distinguishes two electrode arrangements:
- Plate-like. The feedstock sits between two strips or fabrics of carbon. Current heats the strips, and they pass heat to the sample by radiation. The examples given are activated-carbon work at 900 °C (Zhang et al.) and a KOH activation study (Liu et al.). In the KOH study, fast heating and cooling is said to give finer, more uniform pores than a furnace, where molten KOH droplets have time to merge into larger ones.
- Rod-like. The feedstock is the conductor. Carbon black is added if needed to keep its resistance between 1 and 1000 Ω, and a capacitor bank discharges through it in milliseconds, with relays timing the pulse. The review reports temperatures of about 3000 °C and bright flashes from black-body radiation. It cites Luong et al.'s 2020 Nature paper for a yield of 80 to 90% and purity of almost 99% without post-treatment. It also cites Eddy et al. (2024) for the claim that the electric field lowers the activation energy and helps align layers.
How turbostratic graphene is told apart
Most of the technical detail concerns characterisation:
- Raman spectroscopy. Turbostratic layers interact only weakly, so the 2D peak stays narrow and Lorentzian whatever the layer count. The usual layer-counting rules therefore do not apply. The review cites a 2D/G ratio as high as 17 in Luong et al., against a commonly assumed maximum of 4 for single-layer CVD graphene. Weak TS₁ (1880 cm⁻¹) and TS₂ (2030 cm⁻¹) peaks should be present. The M peak at 1750 cm⁻¹, linked to ordered stacking, should be absent, and the review says it was absent in Luong et al.'s spectra.
- X-ray diffraction. A (002) peak at 26.3° corresponds to about 3.45 Å spacing, wider than graphite's 3.36 Å, along with a broadened, asymmetric (002) peak and a weak (100) peak.
- Electron microscopy. TEM images show moiré striations where layers are rotationally mismatched. Alongside them are small, bent graphitic particles, labelled wrinkled graphene, of 3 to 8 layers.
The review makes one methodological point that matters for the whole dossier. Raman is a local measurement and samples vary from region to region, so it warns that showing a single spectrum "can be misleading". It recommends averaged spectra or scatter plots of peak ratios instead.
The proposed mechanism and what simulation could not show
The account the review gives runs as follows. Current follows the path of least resistance, so particles along that path graphitise first. Poorly conducting defects keep heating until they too anneal. Bonds break at extreme temperature and carbon re-forms as sp² sheets. Cooling is so fast that the sheets are kinetically trapped before they can settle into ordered Bernal (ABA) stacking. Graphitised particles, or added carbon black, act as seeds for the growing layers.
The review then marks the limits of this account. Molecular-dynamics simulations in Luong et al. tracked rising sp² content at 3000 K and 3500 K. They agreed with experiment on how temperature relates to graphene formation and defect healing. But they could not reproduce discrete turbostratic sheets. Computing limits kept the runs under 5 ns, and the review states that the simulated annealing time was only about 0.002% of the real annealing time. Separately, it gives the experimental heating as temperatures above 3000 K for almost 5 ms. These figures do not match: 5 ns is about 0.0001% of 5 ms, not 0.002%, and the review does not explain the difference. The simulations also left out resistance, charge fluctuations and the current itself, and they modelled pure amorphous carbon rather than feedstocks such as biomass with impurities and volatiles. The kinetic-trapping idea rests instead on experiment. Lengthening the flash to 5 s gave more ordered stacking in Raman analysis.
A second simulation is cited from Guan et al. (2025), with 10 lignin molecules and one carbon-black particle. Species distributions levelled off near 5 ns, with C₆–C₂₀ species at 10% and C₁–C₅ and C₂₀₊ up to 40%. Ring counts first fell and then rose. The review takes this as evidence that the carbon-black particle templated graphene growth.
What the document does not establish
- No independent data. Every yield, purity and property figure comes from cited papers, and the turbostratic-specific figures come mainly from a single 2020 paper, Luong et al. The review's background reference on large-scale flash synthesis is co-authored by Luong and James M. Tour.
- Internal inconsistencies. One passage gives heating and cooling rates of "tens to hundreds of degrees Celsius per second", which is hard to reconcile with reaching about 3000 °C within milliseconds. The interlayer spacing appears as 3.45 Å, ~3.44 Å and 0.34 nm in different passages. A statement attributed to Stanford et al. carries reference number 48, which in the reference list is a review of graphene-cement composites. Stanford et al. is listed as reference 34. The scalability rating also shifts: the first comparison table rates flash graphene's scalability as moderate, while the later property table describes it as low in cost and highly scalable from waste or biomass.
- Unsupported table values. In the property table, flash graphene's thermal conductivity of up to about 2000 W m⁻¹·K⁻¹ is not discussed or sourced in the text beyond the column's single citation.
- Commercialisation. The review says Universal Materials Inc. in Canada has commercialised Joule heating, citing an earlier review, and gives no further detail.
Proposed next steps
The authors suggest continuous rotary or belt-fed reactors to replace batch flashing. They also propose making composites in situ from metal–organic frameworks or metal oxides, doping graphene with heteroatoms from biomass, and using machine-learning models that link feedstock traits to product quality. Other suggestions are real-time purification, printable graphene inks, and systematic studies of mixed feedstocks. All of these are offered as recommendations, not results.
Related pages
Flash Joule Heating (FJH) · Turbostratic Graphene · Joule Heating · James P. Joule · Flash Sintering (Field-Assisted Sintering) · Flash Joule Heating Claims vs. Evidence: The Dossier's Figures and the Absence of Independent Verification
References
- Mahmood F., Magoua Mbeugang C.F., Asghar F., Xie X., Lin D., Liu D., Li B. "Understanding the Synthesis of Turbostratic/Flash Graphene via Joule Heating." Materials, 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC12195536/
Source notes & attribution
- https://pmc.ncbi.nlm.nih.gov/articles/PMC12195536/