Electrothermal Transformations within Graphene-Based Aerogels through High-Temperature Flash Joule Heating

Flash Joule heating of highly porous graphene oxide (GO) aerogel monoliths to ultrahigh temperatures is exploited as a low carbon footprint technology to engineer functional aerogel materials. Aerogel Joule heating to up to 3000 K is demonstrated for the first time, with fast heating kinetics (∼300 K·min–1), enabling rapid and energy-efficient flash heating treatments. The wide applicability of ultrahigh-temperature flash Joule heating is exploited in a range of material fabrication challenges. Ultrahigh-temperature Joule heating is used for rapid graphitic annealing of hydrothermal GO aerogels at fast time scales (30–300 s) and substantially reduced energy costs. Flash aerogel heating to ultrahigh temperatures is exploited for the in situ synthesis of ultrafine nanoparticles (Pt, Cu, and MoO2) embedded within the hybrid aerogel structure. The shockwave heating approach enables high through-volume uniformity of the formed nanoparticles, while nanoparticle size can be readily tuned through controlling Joule-heating durations between 1 and 10 s. As such, the ultrahigh-temperature Joule-heating approach introduced here has important implications for a wide variety of applications for graphene-based aerogels, including 3D thermoelectric materials, extreme temperature sensors, and aerogel catalysts in flow (electro)chemistry.

Brunauer-Emmett-Teller (BET) surface area measurements were conducted on a Micromeritics TriStar 3000 instrument.The samples (100 mg) were degassed in nitrogen atmosphere at 110 o C for 3 h before analysis.The nitrogen adsorption/desorption isotherms were measured at 77 K.The pore diameter distribution of the samples was determined from the desorption isotherm using the Barrett-Joyner-Halenda (BJH) method.The as-synthesized GO aerogel was carried out via Joule-heating, using the setup depicted above (Figure S2).For ultrahigh-temperature Joule-heating, aerogels were Joule-heated at a constant heating current of 10.1 A under nitrogen atmosphere for 30 s (Figure 3a).During this treatment, the emission of intense red-coloured black body radiation.Thermocouple measurements at the aerogel surface confirm that temperatures of at least 1200 °C is reached.

Flash Joule-heating of GO-derived aerogels
Over the pre-conditioning duration, a clear decrease in heating voltage is observed within the first 5 s.This sudden drop in aerogel resistance is attributed to GO deoxygenation and impurity removal (i.e.thermal degradation of ascorbic acid residues remaining from the hydrothermal aerogel synthesis) at the high temperatures reached.Over the remaining 25 s of the preconditioning, a further, small but continuous decrease in heating voltage is observed, indicating a slight improvement in graphiticity over this duration.
The resulting, rGO 30s aerogels are extremely stable under the high-temperature conditions explored in this work.To demonstrate this point, a rGO 30s aerogel was repeatedly Joule-heated to high temperatures (Figure S3b).Specifically, the same high aerogel temperature (T surf = 1000 °C) was accurately reached over 10 Joule-heating cycles, using the same current and voltage input in each cycle.This reliable high-temperature cycling behaviour confirms aerogel stability during high-temperature flash Joule-heating treatments, such as the thermo-chemical nanoparticle synthesis described in the main text.At the same time, the preconditioned aerogels can be heated very stably to 800°C and 900°C in inert atmosphere for long durations (>30 min, Figure S3c).This high-temperature stability is important for aerogel applications beyond flash heating, e.g. in high-temperature chemistry, temperature-swing regeneration of aerogel catalysts and sorbents and aerogel-based high-temperature sensors.Table S1.Physico-chemcial characterisation of GO aerogel and rGO 30s aerogel, including electrical conductivity (), specific surface area (SSA), meso-pore volume (V Meso-porosity ), and micro-pore volume (V micro ).
Name  V Meso-porosity (cm 3 •g -1 ) V micro (cm  In terms of textural properties, N 2 adsorption/desorption measurements (Figure S4c, Table S1) indicate that aerogel pre-conditioning gives rise to a substantial increase in aerogel surface area and porosity.Specifically, the pre-conditioned aerogel (rGO 30s ) exhibits a 9 times larger specific surface area (634 m 2 g -1 ) compared to the parent GO aerogel (68 m 2 g -1 ).Similarly, substantially increases are observed in aerogel porosity upon pre-conditioning, with aerogel mesopore volume increasing from 0.34 cm 3 g -1 to 1.93 cm 3 g -1 while aerogel micropore volume increases by more than 10-fold.These substantial changes further confirm that brief hightemperature Joule-heating conditioning is highly effective in removing organic impurities remaining from the aerogel synthesis and in exposing existing micro-and meso-pores within the aerogel network.The pre-conditioned aerogels also exhibit markedly changed wettability.
Water droplet contact angle measurements of the parent GO aerogels indicate highly hydrophilic character (contact angle 0°), while the pre-conditioned aerogels exhibit clear hydrophobic character (contact angle 143°), confirming removal of polar oxygen surface groups and synthetic impurities after relatively short (30 s) high-temperature Joule-heating.High-temperature Joule-heating also substantially changes the electronic aerogel properties.
This change is reflected by a substantial increase in electronic conductivity from 1.6 Scm -1 (for the partially reduced as-synthesised GO aerogel) to 81.5 Scm -1 (after 30 s high-temperature Joule heating).Joule-heating enables a straightforward route to measure the temperature dependence of the electrical conductivity across a wide temperature range.The Arrhenius thermal activation model can therefore be readily applied to estimate the band gap of the aerogels before and after Joule-heating: 1 Equation ( 1) where  a is a pre-exponential factor; E a is the activation energy; and k B is the Boltzmann's constant.To estimate the band gaps, the logarithm of the electrical aerogel conductivity is plotted against the inverse temperature.Taking the activation energy E a of the Arrhenius fit as estimate for the bandgap, the band gap substantially decreases from 0.09 eV (GO aerogel) to 0.01 eV (rGO 30s aerogel) upon 30 s Joule-heating annealing, in line with the changes in graphitic crystallinity, discussed in the main text.It is also worth noting that the fit is considerably better for the ultrahigh-temperature Joule-heated sample, highlighting again the structural stability of the aerogels after the high-temperature conditions explored in this study.As outlined in the main text and above, short-duration, high-power Joule-heating treatments of GO HT aerogels (120W power input) result in substantial GO graphitisation and annealing, as confirmed by XRD, Raman and TEM (main-text Figure 1).The evolution of graphitic quality is further reflected in substantially improved combustion resistance after Joule-annealing as probed by TGA (Figure S6a).For the parent GO aerogels, the differential thermogram (DTG) exhibits two distinct combustion events at 250 °C and 576 °C.The event at 250 °C indicates impurity removal and GO deoxygenation, while the event at around 580 °C is likely due to combustion of the (partially reduced) GO.For the rGO 30s sample, the lower-temperature event is no longer observed, confirming complete deoxygenation and impurity removal after only 30 s Joule-heating.In addition, the DTG of rGO 30s aerogel (Figure S6a) also suggests substantial annealing of the graphene basal planes as indicated by the reduced amplitude of the 585 °C event (partially-reduced GO) and the appearance of a new combustion event at considerably higher temperatures (705 °C, likely combustion of highly graphitised rGO).This effect is even more pronounced for the rGO 300s aerogel where only the high-temperature combustion event is observed (Figure S6a).The shift to higher combustion resistance upon Joule-heating is further evidenced through TGA-FTIR analysis, where the evolution of CO 2 combustion product (characteristic CO 2 IR peaks at 2350 cm -1 , Figure S6b is observed at considerably higher temperatures for the rGO 30s and rGO 300s samples, compared to the GO HT parent material.

Joule-heating annealing of GO aerogels
These TGA findings are fully in line with the Raman and XRD results and further highlight the utility of high temperature Joule heating for effective enhancements of rGO aerogel quality, both in terms of graphitic crystallinity as well as aerogel purity.The as-synthesised GO aerogels exhibit an internal microstructure with heterogeneous, cellular macroposity, originating from the templating by ice-crystals during the aerogel fabrication process (Figures S7a-d).SEM imaging indicates that Joule-heating annealing does not substantially alter the general macropore structure, however average macropore size slightly decreases for the 30s (Figures S7e-h) and 300s (Figures S7i-l) Joule-annealed samples.
However, the pore walls seem more crumbled and porous in themselves.As evidenced by Raman and XPS data (main text, Figure 1), this is unlikely due to local combustion and is more likely due to local deformations of the multi-layer graphene stacks at the ultra-high temperatures applied.Higher magnification SEM images of the macropore walls show smooth and relatively thin rGO sheets for all samples.1.

Estimation of Joule-Heating Temperatures for Monolithic Aerogels
To determine aerogel temperature as function of electrical power input, rGO 30s aerogels were Joule-heated in inert atmosphere across a power input range of 0.5 -120 W. The Joule heating temperature of three-dimensional, uninsulated aerogel monoliths will be higher at the core compared to the aerogel surface due to heat conduction and thermal losses at the aerogel surface.Therefore, for aerogels, both core temperatures and surface temperatures need to be considered.To this end, aerogel temperatures were estimated via a previously published model, based on simplified one-dimensional heat conduction, that links aerogel surface temperature (T surf ) to aerogel core temperature (T core ) at a fixed electrical power input (q) via the aerogel's thermal conductivity (k). 2 Equation ( 2) Due to technical limitations of the temperature measurement instruments (thermocouples, high-temperature thermal cameras), aerogel Joule-heating temperatures were estimated via slightly different approaches within the medium-temperature and high-temperature regimes, as outlined in the following sections.

Medium-Temperature Regime: Aerogel Core Temperature  1200 o C
In the medium temperature regime (rt -1200 o C), regular thermocouples can be readily used to measure both T core and T surf .Specifically, a thermocouple was inserted into the centre of the rGO 30s aerogel, while a second thermocouple was brought into contact with the aerogel surface (distance, r, between core and surface position ca 5 mm, Figure S11).T core and T surf were then recorded at different power inputs between 0.5W to 60 W (Figure S12a).For each power input step, the measured T core and T surf, were also used to estimate aerogel thermal conductivity k via Eq (1). Figure S12b and Figure S12c show the obtained thermal conductivity as function of Joule heating surface temperature and as function of Joule-heating core temperature, respectively.Both relationships show a typical Umklapp scattering profile with a clear inflection point, where an initial increase of k with temperature switches over to a clear decrease of k with temperature, typical for graphitic materials.Thermocouples could only be reliably used for temperature measurements in the aerogel interior up to 1200 o C. Therefore, for power inputs > 60 W, aerogel core temperatures could no longer be measured directly, but needed to be estimated through equation ( 2) from measurements of the aerogel surface temperature.Surface temperatures could be directly measured even at very high temperatures (Figure S13), either via thermocouples (T surf <1200 o C) or thermal camera (T surf > 1200 o C).
However, to apply equation ( 2), knowledge of the thermal conductivity value  at different surface temperatures is required.The temperature-dependence of the aerogels thermal conductivity therefore needs to be estimated.To this end, the decreasing branch of the -T surf functional relationship in Figure S14a was extended to higher temperatures, using a previously established power-law fitting (PLF) method: 1 Equation ( 3) As known reference points in this equation (k KnownT , T Known ), the values at the inflection point of the measured k-T surf relationship (Figure S14a) where used, specifically k KnownT =1.3 Wm - 1 K -1 at T Known =865K.Equation ( 4)

𝑛
Where k HighTemp is the thermal conductivity at a measured high surface temperature T HighTemp .
The best fit of Equation ( 4) to the measured T surf values (for all T surf > 865 K) is obtained for an exponent of n=-1.31.However, the PLF model requires exponents to be integer.Therefore, fits against n=-1 and n=-2 where also carried out, with the best integer exponent fit obtained for n=-1.Taking this exponent into account, thermal conductivities can be estimated for all high surface temperatures measured.
Equation ( 5) The high-temperature k values estimated in this way were then used as input parameters in Equation ( 5) in order to estimate core temperatures from the measured surface temperatures (Figure S14b).The corresponding results are depicted in Figure 2c of the main text.

GO aerogel Joule-heating at shorter timescales
To probe structural evolution in GO aerogels at timescales shorter than 30s, we have produced new hydrothermal GO aerogel samples and Joule-heated these for 10s and 30s at the same conditions as outlined in the manuscript (10 A, ~120W power input).It should be noted that these additional experiments have been carried out with a different batch of commercial GOso characterisation results for the GO aerogel and rGO 30s presented here differ slightly from the values presented in the manuscript's main text.Considering batch-to-batch variations in GO production, the values are however surprisingly similar and fully confirm and support the trends discussed in the main text, highlighting the robustness and broad applicability of the flash Joule-heating methodology.
In order to explore the structural evolution at shorter time scales, we characterised the GO, rGO 10s and rGO 30s aerogel samples via XPS and Raman spectroscopy.XPS elemental composition data (Figure S22, Table S3) very clearly show that after 10s Joule heating GO deoxygenation is almost complete, with a dramatic decrease of oxygen concentration from around 25at% oxygen in the parent GO aerogel to 1.5at% oxygen in the rGO 10s aerogel.Further Joule heating to 30s results in a small degree of additional deoxygenation (0.5 at% oxygen in rGO30s aerogel), but the data clearly indicate that 10s high temperature Joule heating are fully sufficient to induce GO de-oxygenation.
Raman spectroscopy data (Figure S23, Table S4) show that graphitisation of the lattice is also occurring on fast time scales.After 10s Joule heating, the I D /I G ratio reduces significantly from 1.46 to 0.57.This is accompanied by a narrowing in band widths, also indicating improvements in graphiticity, even at this short time scale.Longer Joule-heating for 30s allows for additional improvements in graphitcity.This is in line with the observations discussed in the main text for 300s-heated sample, that showed steady improvements in graphiticity with longer heating durations.

Structural Stability of GO aerogels upon high-temperature Joule heating
1][32][33][34] We believe that the contrast to the literature studies lies within the synthesis of the GO aerogel which was carried out under chemically reducing conditions, likely changing GO structure.Specifically, hydrothermal GO aerogel synthesis was carried out in the presence of ascorbic acid, following a well-known approach from the aerogel synthesis literature. 35corbic acid is a mild reducing agent, and has been used to aid the formation of robust GO gel networks under the mild hydrothermal synthesis temperatures used in this work.Chemical reduction by ascorbic acid likely removes some of the more labile oxygen functional groups from the GO structure, minimising violent release of gaseous products and exothermic heat upon aerogel heating in our samples (as shown for example by the DSC data in Figure S6).
This chemical reduction of our GO is in clear contrast to the pristine, oxygen-rich GO, utilised in the heating studies cited above.
XRD patterns of the untreated GO and the ascorbic-acid-treated GO (i.e. the GO present in the aerogels) showed very clear differences, mainly in the (002) GO peak position and broadness (Figures S24a-24b).The untreated GO sample, showed a relatively defined XRD peak at 2theta = 11.43°(corresponding to a relatively large d-spacing of 0.77 nm), indicating the presence of a large amount of oxygen functional groups.3] This is in contrast, to the ascorbic-acid treated GO (i.e. the GO present in the aerogels) used in our work that exhibited a substantial shift to smaller d-spacings (peak maximum corresponds to d = 0.36 nm), clearly evidencing chemical reduction, likely leading to partial GO de-oxygenation.
Raman spectroscopy seems to confirmed the XRD findings, showing a clear increase in I D /I G ratio (Figure 24c, Table S5) compared to untreated GO, suggesting some degree of sp 2 domain increase, likely associated with deoxygenation.However, the presence of ascorbic acid residues in aerogel makes Raman data interpretation uncertain.Low levels of ascorbic residues have been left in the GO aerogels on purpose in this work.(Even after four solvent exchanges around 10wt% ascorbic acid remains within the final GO aerogel structure as indicated by TGA, see also Figure S6a).Some literature studies have suggested that pyrolysis of small amounts of organics embedded in the GO network can leave behind carbonaceous residues (<1 wt%) upon heat treatment that can function as additional (covalent) crosslinking points. 36The potential crosslinking through ascorbic acid pyrolysis is extremely challenging to evidence experimentally, due to the difficulties detecting low of carbonaceous crosslinking residues against the carbon matrix of the GO network.However, we do find that GO aerogels containing small amounts of ascorbic acid prior to heating show better structural robustness after heat treatments.This is mentioned here as potential secondary effect contributing to the stability of our aerogels under Joule-heating and to explain why we chose not to remove all ascorbic acid from the GO aerogels.
It is worth noting that literature has also shown that GO deoxygenation does occur under hydrothermal conditions, [37][38] however this effect is less pronounced at acidic conditions.The increased rGO stacking and small d(002) spacings, evidenced by XRD, do however suggest that some de-oxygenation (potentially of the more labile oxygen groups) has occurred in our materials.The large oxygen contents and large d-spacings of the pristine GO used the literature studies also suggest the presence of water in the GO interlayer spaces.Such chemisorbed interlayer water could also significantly contribute to the disruptive structural changes in GO upon heating. 31Finally, it should be noted that our GO aerogels are highly porous (void volume fractions of up to 95%) and possess many relatively large, spacious pores (> 5 m) into which gaseous compounds easily can expand into upon heating, likely also substantially minimising disruptive structural changes.This is again in contrast to previous heating studies which have been carried out on much denser GO material forms, such as GO papers [30][31][32][33] and GO clumps/films. 34 summary, partial chemical reduction of GO through ascorbic acid during hydrothermal synthesis is indicated by XRD.This is in clear contrast to pristine GO previously investigated in the literature under medium-to high-temperature conditions.In our materials instead, chemical reduction of GO has potentially led to the removal of more reactive groups from the GO structure, resulting in materials considerably less prone to violent exothermic release of gaseous compounds upon fast heating.Secondary effects, such as additional crosslinking due to organic residue pyrolysis and the extremely high porosity of our aerogel samples might further mitigating against structural disintegration during high-temperature GO aerogel heating.Table S5.Raman data for and untreated, pristine GO and for GO, hydrothermally treated with ascorbic acid (i.e.equivalent to the GO aerogel synthesis conditions).

Figure S1 .
Figure S1.Schematic depiction of hydrothermal synthesis process to produce as-synthesised GO aerogels.

Figure S2 .
Figure S2.Schematic and photographs of high-temperature Joule-heating set-up

Figure S4 .
Figure S4.Materials characterisation of pre-conditioned aerogels.(a-b) Water contact angles of GO aerogel and rGO 30s aerogel.(c-d) Nitrogen adsorption-desorption isotherms and pore size distributions of the GO aerogel and rGO 30s aerogel.(e-f) SEM images of internal microstructure for the GO aerogel and rGO 30s aerogels.

Figure S5 .
Figure S5.(a) Electrical conductivity as function of Joule-heating temperature of GO aerogel and GO 30s aerogel.(b) Arrhenius model fitting of GO aerogel and rGO 30s aerogel (bandgap for GO aerogel is 0.09 eV and for rGO 30s aerogel is 0.01 eV).

Figure S8 .
Figure S8.Raman maps, depicting variations in the intensity ratio of the D peak to the G peak (I D /I G ) for the (a) GO aerogel, (b) rGO 30s aerogel and (c) rGO 300s aerogel across 50um x 50 um viewing field.Improvements in graphitisation and graphitic homogeneity upon Joule-annealing are further evidenced by Raman mapping (Figure S8).The average I D /I G ratios in the Raman maps of the GO aerogel (I D /I G =1.8), rGO 30s aerogel (I D /I G =2.1) and rGO 300s aerogel (I D /I G =0.5) aerogels follow the same trend as observed by Raman spectroscopy of the bulk powders (main text Figure 1).In line with the Tungista-Konig plot for graphitic carbon materials of fundamentally different crystallinity, the substantial improvement in graphitic crystallinity for the highly disordered GO materials after 30 s high-temperature Joule heating (T core ~ 3000 K) is indicated by an increase in I D /I G ratio (accompanied by substantial narrowing of peak width, see also Raman spectra Figure 1 main text).Further, high temperature annealing of the now relatively crystalline rGO 30s material is then indicated by a decrease in I D /I G ratio for the rGO 300s .As for the I 2D /I G Raman maps discussed in the main text, I D /I G Raman maps also indicate an overall improvement of graphitic homogeneity with increasing Joule-annealing, as indicated by the more uniform colour distributions in the maps with increasing Joule-annealing duration.

Figure S11 .
Figure S11.Schematics of estimating thermal conductivity of a rGO 30s aerogel (for Joule heating core temperature < 1200 o C) using a previously reported thermal gradient fitting method.

Figure S12 .
Figure S12.(a) Joule-heating core temperature and surface temperature of rGO 30s aerogel as function of power input.(b) Thermal conductivity of an rGO 30s aerogel versus Joule-heating surface temperature; (c) Thermal conductivity of an rGO 30s aerogel versus Joule-heating core temperature.

Figure S13 .
Figure S13.Schematics of estimating thermal conductivity and core temperature of an rGO 30s aerogel (Joule-heating core temperature > 1200 o C) using a power-law expression.

Figure S14 .
Figure S14.Thermal conductivity versus (a) Joule heating surface temperature and (b) core temperature, fitted according to Equation (3) using different n values.