Flash Graphene Synthesis in Bulk Quantity

Overview

I built a flash Joule heating system to investigate graphene synthesis from carbon black. Brief electrical pulses rapidly heat and cool the feedstock, yielding approximately 100 mg of material per batch. Raman measurements show graphitic carbon with possible turbostratic character; confirming the layer structure remains a next step.

Experimental Setup

Photograph of the custom-built flash graphene synthesis apparatus — Flash graphene synthesis setup.

Additional photograph of the flash graphene synthesis apparatus — Another view of the experimental apparatus.

A custom AC–DC converter charges the capacitor bank, and an Arduino controls the discharge switch. Carbon black sits between copper electrodes in a quartz tube, where rapid heating to approximately 3,000 K is followed by rapid cooling.

Capacitor bank: 70 mF
Charging voltage: Up to 300 V DC
Switch on-time: ~30 ms

Experiment Video

Production & Raman Characterization

Six spectra show strong D and G bands and a weaker, broad 2D band, indicating disordered or edge-rich graphitic carbon. The 2D profile is compatible with turbostratic material but does not establish its stacking or layer count.

The spectra below use local background subtraction and normalization to the fitted G-peak height, without smoothing.

Six Raman spectra after local background subtraction and G-peak normalization, showing D near 1335 and G near 1575 inverse centimeters. — **D and G bands.** Fitted D/G peak-height ratio: 1.06–1.15.

Six background-subtracted 2D bands centered near 2662 to 2667 inverse centimeters, with fitted widths of 99 to 121 inverse centimeters and heights around one-quarter to one-third of G. — **2D band.** Fitted 2D/G peak-height ratio: 0.25–0.34.

Peak measurements

Approximate fitted parameters for the six spectra
Spectrum	D center (cm⁻¹)	G center (cm⁻¹)	2D center (cm⁻¹)	2D FWHM (cm⁻¹)	I_D/I_G	I_2D/I_G
dark-1	1333	1576	2662	101	1.14	0.25
dark-2	1333	1575	2662	99	1.15	0.25
dark-3	1335	1573	2665	108	1.06	0.34
dark-4	1335	1573	2665	106	1.06	0.34
dark-5	1337	1576	2666	121	1.07	0.27
dark-6	1337	1576	2667	118	1.07	0.27

FWHM: full width at half maximum. Ratios use fitted peak heights, not integrated areas; values depend on the background and fitting models.

Analysis method and limitations

Unsmoothed data were fitted with Lorentzian peaks and a linear background: two peaks over 1100–1800 cm⁻¹ for D/G, and one over 2450–2850 cm⁻¹ for 2D. Only the local background was subtracted.

A single-peak 2D model was favored in five of six spectra; the extra component in dark-6 did not establish a resolved peak. Alternative fits gave 2D widths of 95–127 cm⁻¹, a sensitivity range rather than a confidence interval.

D′ was not fitted separately. Excitation wavelength and instrumental resolution were unavailable, and stacking-sensitive bands were unresolved. No layer count, defect dimensions, or turbostratic fraction can be inferred.

Next steps are higher-quality Raman measurements, microscopy and diffraction to verify layer stacking, and increased production capacity.