US12054391B2, "Flash Joule heating synthesis method and compositions thereof", is the core United States patent behind "flash graphene", the product most associated with Flash Joule Heating (FJH). It is assigned to William Marsh Rice University (Rice University), and its listed inventors are James Mitchell Tour (James M. Tour), Duy X. Luong, Wilbur Carter Kittrell and Weiyin Chen. The patent claims priority from 5 September 2018. It was filed on 23 August 2019 as application US17/272,895, published as US20210206642A1 on 8 July 2021, and granted on 6 August 2024. Google Patents lists it as active with an adjusted expiry of 7 April 2040, and flags that the patent family "has litigation". Google notes that its status, date and assignee entries are assumptions, not legal conclusions.
What the patent describes as the invention
The core idea, as the patent describes it, is simple: make graphene by sending a voltage pulse through a conductive carbon source that is "substantially not graphene". The description then sets out wide operating ranges for possible embodiments:
- pulse length from 1 microsecond to 5 seconds, with a narrower range of 100–500 ms;
- 2 to 100 repeated pulses, with a narrower range of 2 to 10;
- current density from 0.01 to 10000 A/cm², and field strength from 10 to 4000 V/cm;
- feedstock conductivity above 10⁻⁵ S/cm, preferably above 10⁻³ S/cm.
Feedstocks that conduct poorly can be brought up to that level by mixing in carbon black, metal powder or flash graphene from an earlier run. The list of possible feedstocks is very broad. It runs from anthracite, calcined petroleum coke and biochar to plastics, rubber, food waste, wood, proteins, feces and animal carcasses. The described product yields range from "at least 70%" graphene up to 100%. The patent also describes a bulk material weighing at least 1 gram in which most of the graphene, and in some versions at least 90 wt%, is turbostratic.
The patent also reaches well beyond synthesis. It describes composites containing 0.001 to 10 wt% turbostratic graphene in concrete, plastics, paints, metals, asphalt and other materials. In these composites, tensile, compressive or yield strength "can be" at least 10% higher than without the graphene. Other proposed uses include drilling-fluid additives, dry lubricants, scale and corrosion inhibitors, gas and nerve-agent sensors, and optoelectronic devices. The patent also proposes applying the same pulsing to other two-dimensional materials: 1T′- and 2H-MoS₂, hexagonal boron nitride and black phosphorus. The patent argues that very short pulses can capture metastable phases that are hard to make in other ways.
How the apparatus is described
The laboratory system in the description is small. A quartz tube holds the compressed carbon powder between electrodes, which may be copper, graphite or another conductive refractory material. The tube sits in a chamber at atmospheric pressure or under a mild vacuum to let gases escape. The power comes from a bank of ten 450 V, 6 mF electrolytic capacitors, which the patent says can handle batches of up to 0.5 g. A second bank of ten 400 V, 18 mF capacitors extends this to 1.0 g. A mechanical relay switches the discharge, and a 24 mH inductor keeps the current from spiking and causing arcs at the relay. A microcontroller board controls the sequence. The sample holder was built from a small commercial vise and laser-cut wooden parts.
According to the patent, the discharge runs at about 400 V and reaches up to 1000 A in under 100 ms. It heats the carbon above 3000 K, and the inventors write that the process is fast enough to prevent AB-ordered layering. The document includes practical safety warnings. Many thousands of joules released in milliseconds can make relays or capacitors explode, so it recommends DC-rated circuit breakers in place of ordinary switches.
The patent also proposes two scale-up designs. One is a continuous belt of quartz or ceramic cups. A powder feed fills each cup, one geared electrode compresses and preheats the sample, and a second delivers the flash. Gas jets blast cooling gas onto the electrodes and clean the cups; after a cup is depressed, a separate stream of gas from a nozzle clears the finished flash graphene into a collection bin. The inventors say they believe this could reach at least 60 pulses per second. Beyond a statement that Pyrex cups have been tested and work well, the patent reports nothing about the belt being built. The other design draws on mains three-phase power: zero-crossing relays and six rectifying diodes shape it into a DC pulse. Pulse energy is adjusted by the number of half-cycles let through, in steps of 1/360 second (2.78 ms). No throughput data are given for either design.
The figures the inventors report
The description gives figures for its own experiments. From high-carbon sources such as carbon black, calcined coke or anthracite, the patent reports yields of 80 to 90% with carbon purity above 99%. It puts electrical energy at about 7.2 kJ per gram. In the Raman spectra, the D band is weak or missing, and the I2D/G ratio exceeds 10 and sometimes reaches 17. On that basis the inventors assert that flash graphene is "among the lowest defect graphene ever reported".
Temperatures were measured by fitting black-body emission at 600–1100 nm, and the patent identifies 3000 K as the threshold for low-defect product. Carbon-black flash graphene had a BET surface area of about 295 m²/g. Batches made in tubes of different sizes weighed 0.003 g, 0.1 g and 1 g. According to the patent, used coffee grounds and the plastic samples were mixed with 5 wt% carbon black, other non-plastic feedstocks with 5 to 10 wt%, and biochar needed no additive. It adds that 2 to 5 wt% flash graphene from a previous run can replace the carbon black. The mixed-plastic blend was HDPE 40%, PETE 40%, PP 10% and PVC 10%.
As evidence for turbostratic stacking, the patent points to two Raman features it labels TS1 (near 1880 cm⁻¹) and TS2 (near 2030 cm⁻¹). It also cites a single narrow 2D peak whose Lorentzian fit gives R² of 0.999. The patent distinguishes its product from an earlier detonation-synthesis patent (Sorensen '857), saying that patent's 2D band of 43–63 cm⁻¹ is too broad for turbostratic graphene. It also describes Tour's earlier laser-induced graphene as a surface-only method that yielded less than 1 gram per day.
What the document does and does not establish
A granted patent shows that the filing was examined and granted. It does not show that the process performs as described (see Patent as Evidence). The reported yields, purities, energies, Raman ratios and BET surface area all come from the inventors' own laboratory. The patent does not say which spectrum produced the I2D/G of 17. The described ranges are much wider than anything the description reports testing: pulses of microseconds to seconds, a 10⁶-fold span of current density, and feedstocks such as animal carcasses. The ≥10% composite improvements are written as permissive language ("can be"), not as measured results. The automated belt and three-phase systems are designs, and no output figures are given for them. The patent's molecular-dynamics runs lasted up to 5×10⁻⁹ s and address graphitization and defect healing. The 2025 Materials review notes that simulations of this length have not reproduced turbostratic stacking itself (How Flash Graphene Forms: A 2025 Review of Joule-Heating Synthesis, Characterisation and the Limits of Its Mechanism (Mahmood et al., Materials)).
Place in the dossier
As described by its inventors, this patent sets out the early figures of the flash graphene programme: gram batches from capacitor discharge, about 7.2 kJ/g, and 80–90% yield. Later figures in Flash Joule Heating Claims vs. Evidence: The Dossier's Figures and the Absence of Independent Verification can be checked against these baselines. Background on the stacking question is in Turbostratic Graphene.
Source notes & attribution
- https://patents.google.com/patent/US12054391