Biomass Waste Carbonization in Piranha Solution: A Route to Hypergolic Carbons?

Abstract

In the present work we report for the first time the carbonization of biomass waste, such as stale bread and spent coffee, in piranha solution (H₂SO₄-H₂O₂) at ambient conditions. Carbonization is fast and exothermic, resulting in the formation of carbon nanosheets at decent yields of 25–35%, depending on the starting material. The structure and morphology of the nanosheets were verified by X-ray diffraction, Raman, X-ray photoelectron and microscopy techniques. Interestingly, the obtained carbon spontaneously ignites upon contact with fuming nitric acid HNO₃ at ambient conditions, thus offering a rare example of hypergolicity involving carbon as the solid fuel (i.e., hypergolic carbon). Based on the relatively large interlayer spacing of the as-produced carbons, a simple structural model is proposed for the observed hypergolicity, wherein HNO₃ molecules fit in the gallery space of carbon, thus exposing its basal plane and defect sites to a spontaneous reaction with the strong oxidizing agent. This finding may pave the way towards new type hypergolic propellants based on carbon, the latter exclusively obtained by the carbonization of biomass waste in piranha solution.

1. Introduction

Piranha solution is a powerful oxidizing agent that is typically made of concentrated sulfuric acid (H₂SO₄) and 30% hydrogen peroxide (H₂O₂) at a 3:1 ratio. The solution tends to attack organic matter, particularly carbohydrates, through the dehydration of the organic molecules into carbon and the subsequent etching of the as-produced carbon into CO₂ by in situ formed oxygen radicals. The overall result is combustion of the organic molecules into CO₂ and H₂O at room temperature without flame (e.g., a sort of cold combustion). As is common in every combustion process, an excess of oxygen and hence, an excess of piranha solution will lead to the complete combustion and etching of the organic matter, whereas an oxygen shortage and hence, a shortage of piranha solution will lead to the incomplete combustion and formation of char-like carbons. The advantages of the piranha solution treatment are that it takes place at ambient conditions without pyrolysis, and it is fast and exothermic, liberating energy that can be converted into useful work, and lastly, the released greenhouse CO₂ gas can be collected at ambient conditions without escaping into open air. Disadvantages include high toxicity and corrosiveness, safety issues related to use and disposal, as well as loss of strength over time; the latter implies that freshly prepared solutions must be used. Therefore, careful handling is necessary before, during and after use in a particular process.

Despite these disadvantages, due to its strong oxidizing power, piranha solution finds numerous practical uses in the microelectronics industry for cleaning/modifying the surface of silicon wafers or for etching copper from printed circuit boards and in the chemistry laboratory for cleaning glassware, as well as in the hydrophilization of glass by creating surface silanol Si-OH groups that are easily post-functionalized, e.g., through silanization by organofunctional alkoxysilanes. Moreover, this mixture, which also goes under the name of piranha etch, serves in carbon materials synthesis either in producing advanced carbon nanomaterials, such as graphene and photoluminescent carbon dots [1,2], or in treating the surface of carbon materials (graphene oxide nanosheets, nanotubes, fibers) to induce the formation of oxidized derivatives with tailored properties suitable for applications in cell labeling, energy and composites [3,4,5,6].

Stale bread and spent coffee are common biomass food wastes produced in large quantities. Characteristically, nearly 900,000 tonnes of bread and 6,000,000 tonnes of spent coffee end up annually in landfills worldwide before their full potential is exploited. This motivates on-going research interest for utilizing this waste as a resource for biogas/biofuel production or for the extraction of useful chemicals [7,8,9,10]. Bread and spent coffee wastes have also been employed as starting material for the development of advanced functional carbon materials with interesting properties, often produced via conventional pyrolytic or hydrothermal carbonization methods [11,12,13,14,15,16]. However, to the best of our knowledge, no report on bread and coffee biomass waste carbonization in piranha solution has been published so far. Interestingly, bread and coffee contain a significant fraction of carbohydrates, proteins and fats [17,18] that are all prone to carbonization by piranha solution. Therefore, this procedure merits further attention as it could provide an alternative processing method to pyrolysis or hydrothermal treatment, yielding carbons with unique properties not achievable by conventional carbonization methods.

Herein, we present for the first time the carbonization of stale bread and spent coffee wastes in piranha solution, resulting in the formation of carbon nanosheets. The nanosheets exhibit relatively high interlayer spacing (d₀₀₂ > 4 Å), a result of released CO₂ hot gas pushing the carbon layers apart from each other. This allows HNO₃ molecules to intercalate carbon, thus reaching its basal space and defect sites to trigger the hypergolic ignition of the solid at ambient conditions. The observed carbon hypergolicity is a rare phenomenon in the literature and appears to be quite unique to the present piranha solution treatment. In this respect, biomass waste carbonization in piranha solution is expected to pave the way towards the future advancement of a larger series of chemically ignitable carbons as part of a new generation of hypergolic propellants.

2. Materials and Methods

All syntheses were conducted in a fume hood. Concentrated H₂SO₄ 98% and H₂O₂ 30% were supplied by Merck KGaA, Darmstadt, Germany. Piranha solution is highly corrosive and should be handled with great caution inside a fume hood.

In the case of bread, a cut slice of stale bread weighing a total of 30 g was immersed in 100 mL piranha solution (3:1, 75 mL H₂SO₄ 98%–25 mL H₂O₂ 30%) as shown in the set-up of Figure 1. The fast (within 5–10 min) and exothermic reaction that took place led to pieces of homogeneously carbonized bread. The pieces were washed with de-ionized water until neutral pH, with acetone until no coloration of the solvent was observed and with dimethylformamide (DMF, Sigma–Aldrich, St. Louis, MO, USA) until no coloration of that solvent was observed. DMF is a highly polar solvent capable of washing away organic species that strongly adhere to the carbon surface. After washing with DMF, the sample was once again washed with water and acetone in order to remove the polar solvent, air-dried at 100 °C and crushed into a fine black powder with a yield of 25% (specific surface area from N₂ BET was 10–15 m² g^–1). The sample is thereafter denoted as BRPS (from BRead-Piranha Solution).

In the case of spent coffee, 40 g of dried spent coffee grains (from brewed coffee, brand name JACOBS DECAF) were slowly added in 100 mL piranha solution (3:1, 75 mL H₂SO₄ 98%–25 mL H₂O₂ 30%) as shown in Figure 2. Similarly as above, the fast and exothermic reaction that took place led to a homogeneous carbonization of the spent coffee grains. The carbonized solid was rinsed with de-ionized water until neutral pH, with acetone until no coloration of the solvent was observed and with N-methyl-2-pyrrolidone (NMP, Sigma–Aldrich, St. Louis, MO, USA) until no coloration of that solvent was observed. NMP is another highly polar solvent capable of washing away organic species that strongly adhere to the carbon surface. After washing with NMP, the sample was once again washed with water and acetone in order to remove the polar solvent, air-dried at 100 °C and crushed into a fine black powder with a yield of 35% (the specific surface area from N₂ BET was 10–15 m² g^–1). The sample is thereafter denoted as SCPS (from Spent Coffee-Piranha Solution).

At this point it should be mentioned that washing is an inevitable part of this method in order to obtain products with a well-defined structure and composition. Typically, nearly half a liter of a solvent is used in order to purify the carbon product. Taking into consideration that this waste processing also creates solvent waste, the sustainability and viability of the method should be taken seriously into consideration in terms of large-scale production.

On the other hand, the piranha solution acts through the following cascade of reactions towards carbonization as shown in Scheme 1 (https://en.wikipedia.org/wiki/Piranha_solution, last accessed on 15 January 2021):

X-ray diffraction (XRD) was conducted on background-free Si wafers using Cu Kα radiation (λ = 1.54 Å) with a Bruker Advance D8 diffractometer (Bruker, Billerica, MA, USA). Raman spectra were recorded with a RM 1000 Renishaw micro-Raman system using a laser excitation line at 532 nm (Nd:YAG) in the range of 100–4000 cm⁻¹. A laser power of 10 mW was used with a 2 μm focus spot. X-ray photoelectron spectroscopy (XPS) was performed with a SSX-100 (Surface Science Instruments) spectrometer, equipped with a monochromatic Al Kα X-ray source (hν = 1486.6 eV) and operating at a pressure of 1 × 10⁻⁹ mbar. The samples were pressed onto silver plates; a flood gun was used to neutralize charging effects due to the insulating nature of the samples. The photoelectron take-off angle was 37° with respect to the surface normal. The analyzed spot size on the sample was 1000 μm in diameter. The experimental resolution was set to 1.67 eV for overview spectra and to 1.26 eV for the detailed scans of the various core level regions. Binding energies are reported ±0.1 eV and referenced to the C1s signal at a binding energy (BE) of 284.6 eV. All XPS spectra were analyzed using the least squares curve-fitting program Winspec (developed at LISE, University of Namur, Namur, Belgium). Deconvolution of the spectra included a Shirley background subtraction and fitting with a minimum number of peaks consistent with the chemical structure of the sample, taking into account the experimental resolution. The profile of the peaks was taken as a convolution of Gaussian and Lorentzian functions. At least 3 spots on the sample were analyzed, in order to verify the homogeneity of the sample. The uncertainty in the peak intensity determination was 2% for all core levels reported. Scanning electron microscopy (SEM) images were obtained using a JSM-6510 LV SEM Microscope (JEOL Ltd., Tokyo, Japan) equipped with an X-Act EDS-detector by Oxford Instruments (Abingdon, Oxfordshire, UK, an acceleration voltage of 20 kV was applied). The samples were studied using transmission electron microscopy TEM (JEOL 2010 TEM equipped with a LaB₆ type emission gun operating at 160 kV, JEOL Ltd., Tokyo, Japan). Scanning transmission electron microscopy high-angle annular dark-field (HAADF) imaging analyses for the EDS mapping of elemental distributions on the products were performed with a FEI Titan HR-TEM microscope operating at 80 kV. Atomic force microscopy (AFM) images were collected in tapping mode with a Bruker Multimode 3D Nanoscope (Ted Pella Inc., Redding, CA, USA) using a microfabricated silicon cantilever type TAP-300G, with a tip radius of <10 nm and a force constant of ca. 20–75 N m⁻¹. The Si wafers (P/Bor, single-side polished, Si–Mat, Kaufering, Germany) used in AFM imaging were cleaned before use for 20 min in an ultrasonic bath (160 W) with water, acetone (≥99.5% Sigma–Aldrich, St. Louis, MO, USA) and ethanol (≥99.5% Sigma–Aldrich, St. Louis, MO, USA).

3. Results

3.1. Structural Characterization and Morphology

The powder XRD patterns of BRPS and SCPS (Figure 3, top) show broad (002) reflections with corresponding interlayer spacings of d₀₀₂ = 4.3 Å and d₀₀₂ = 4.4 Å, respectively. Such spacings are larger than that of crystalline graphite (d₀₀₂ = 3.4 Å [19]) and due to the generation of hot CO₂ gas that pushes the carbon layers apart from each other. The broadness of the reflections along with the relatively high d₀₀₂ values pinpoints the formation of amorphous carbons adopting a turbostratic ordering of incompletely crystallized layers [20,21]. In contrast, highly crystalline graphite displays sharper reflections with smaller d₀₀₂ values [19]. It is also worth noting that the d₀₀₂ value of SCPS is higher than that of pyrolyzed spent coffee with d₀₀₂ = 3.5 Å [16]. Based on the Scherrer equation, the L_c crystallite height of the particles along the c direction was estimated at about 23 Å or 2.3 nm for both samples, values which correspond to stacks of 6–7 carbon layers per particle [21]. On the other hand, Raman spectroscopy (Figure 3, bottom) shows the characteristic for carbon D and G bands at 1356 and 1585 cm^–1, respectively, with an intensity ratio I_D/I_G = 0.7 for both samples (i.e., >0.2 typical for crystalline graphite [19]), indicating amorphous carbon [20,22,23]. Based on the I_D/I_G ratio and its close relation to the L_α size of the sp² nanodomains within a layer [22], L_α was estimated to be 6.3 nm. Moreover, the broad band centered near 2900 cm^–1 can either be attributed to N incorporation into the carbon lattice (e.g., nitrogenous carbon) [24,25,26] or/and to a set of broad and superimposed D + D”, 2D, D + G and 2D’ pre-graphite bands [19], due to incomplete carbon crystallization.

The composition of BRPS and SCPS was determined by XPS. The XPS surveys of both samples (Figure 4a,c) show intense peaks at BEs of 284.0 and 530.0 eV, which are attributed to the C1s and O1s core level lines, respectively [27]. For BRPS (Figure 4a) other peaks with significantly lower intensity are identified at BEs of 400.0, 168.0, 51.0 and 31.0 eV; they correspond to the N1s, S2p, Mg2p and Na2p core level lines and are typical for the composition of bread [28]. For SCPS (Figure 4c), the XPS survey, in addition to the main elements C and O, showed the signature of N and S at BEs of 400.0 and 168.0 eV, respectively, in agreement with the typical elemental composition of coffee waste [29]. The C1s spectrum for BRPS (Figure 4b) needs six components to achieve a good fit; this implies that, at a BE 282.9 eV, C-Me (Me: Mg, Na) bonds contribute with 10% to the total carbon intensity; at 284.5 eV, C-C bonds are responsible for 18.2%; at 286.1 eV, C-S bonds are responsible for 21.8%; at 287.6 eV, C-O/C-N bonds are responsible for 34.4%; C=O at 289.6 eV for 11.1% and O-C=O at 291.6 eV contribute with 4.4% [30,31]. On the other hand, the deconvolution of the C1s spectrum for SCPS necessitates four components (Figure 4d); the signal at a BE of 284.6 eV is attributed to C-C bonds and amounts to 56.7% of the total C1s intensity; the signal at 285.8 eV stems from C-S bonds (relative intensity 30.6%); the one at 287.4 eV is assigned to C-O/C-N bonds (7.8%), and, at 289.3 eV, one finds the contribution of C=O bonds accounting for 4.9% of the carbon signal [32,33].

The morphology of the BRPS and SCPS samples was assessed by SEM and TEM electron microscopies, showing the presence of plate-like nanometer-sized flakes with multilayer structure (Figure 5). From the height profiles of the AFM examination, the thickness of the flakes was estimated to be circa 3 nm (Figure 6), i.e., close to the L_c crystallite height calculated from X-ray diffraction (2.3 nm).

On the other hand, chemical mapping of BRPS and SCPS (Figure 7), carried out by energy-dispersive X-ray spectroscopy conducted with the HR-TEM, revealed a homogeneous C, O, N and S elements distribution throughout the nanometer-sized flakes, thus confirming their uniform composition.

These results suggest that the effects of a piranha solution treatment should also be considered for other types of waste discarded in large quantities, such as apples, potatoes, cigarette butts, medicine waste (aspirin, paracetamol) and cardboard. We started with cardboard, which produces sizeable amounts of waste in spite of recycling. For the carbonization process, 30 g of cardboard were first cut into small pieces and then added to 70 mL of piranha solution (3:1 H₂SO₄/H₂O₂). The obtained carbon was washed similarly as the BRPS and SCPS samples, except for using dimethyl sulfoxide (DMSO, Merck KGaA, Darmstadt, Germany) as a highly polar solvent. The carbon yield was 15%. Once more, XRD showed a broad (002) reflection, which translates into an interlayer spacing of 4.3 Å; in addition the signature peaks of cellulosic residues from cardboard were detected (Figure 8, top). On the other hand, Raman spectroscopy showed the characteristic D and G bands of carbon (Figure 8, bottom). Note that some cardboard residues in the sample could be seen by naked eye and hence could be separated by hand.

Similar trials with other waste products gave carbon yields that are too low for an economically viable exploitation, namely 0.5% for apples, 0.1% for potatoes, 5% for cigarette butts and 0.1% for medicine waste (aspirin, paracetamol).

3.2. Carbon Hypergolicity

Hypergolic fuels are mainly used as propellants in rocket engines and simply consist of a fuel and a strong oxidizer that ignite spontaneously upon contact at ambient conditions without external stimuli. Fuels include hydrazine derivatives, heterocyclics, long-chain aliphatics, N-alkyl diamines, aromatics, sugars, ionic liquids, metal-organic frameworks and metal powders (Al, Mg), whereas strong oxidizers include fuming nitric acid HNO₃, N₂O₄, NH₄ClO₄, KClO₃, concentrated H₂O₂ and Na₂O₂ to name a few [34,35,36,37,38,39,40,41,42,43,44,45,46]. Some criteria for hypergolic fuels are high energy value (25–50 kJ·g⁻¹ [41]), high density (e.g., preferably solid fuels that occupy less space), inexpensiveness, low toxicity and fast ignition upon contact with a strong oxidizer. More and more often, nanocarbons such as fullerenes, nanotubes, graphene/graphene oxide or nanodiamonds are added as fillers in the composition of hypergolic fuels in order to boost and sustain a smooth ignition [47]. Nevertheless, none of these nanocarbons are inherently hypergolic upon contact with a strong oxidizer at ambient conditions. In fact, the phenomenon of carbon hypergolicity is quite rare in literature, and the very few cases documented comprise: activated charcoal reported to ignite upon contact with Na₂O₂ in the presence of a small amount of water (see Bretherick’s Handbook of Reactive Chemical Hazards); fullerenes shown to ignite upon mechanochemical mixing with Na₂O₂ [48]; fine carbon dust, which can explode in air under certain circumstances [49]; coal found to ignite and burn upon contact with F₂ gas at ambient conditions [50]; and carbon nanotubes, for which flash ignition was shown [51]. Another example not to be forgotten is “pyrophoric carbon” [52,53], a self-ignitable form of carbon obtained from the mild pyrolysis of cut herbage at 250–300 °C. In all these cases, carbon ignition takes place spontaneously at ambient conditions without external stimuli. Although carbon materials have high energy value (ca. 30 kJ·g⁻¹), low cost and low toxicity, their direct application as solid propellants in rocket engines has remained largely unexplored.

In the current study, the carbons obtained from the piranha solution treatment of stale bread and spent coffee were shown to be inherently hypergolic when brought in contact with fuming nitric acid HNO₃ (Figure 9). To this aim, 2 g of BRPS or SCPS were placed in an open glass vial and 3 mL or 5 mL fuming HNO₃ (100% Sigma–Aldrich, St. Louis, MO, USA) were added respectively. The ignition is sensitive to the amount of HNO₃: excess of nitric acid may lead to the wetting of the particles, thus inhibiting ignition (boiling point of pure HNO₃: 83 °C), while the addition of too small amounts cannot sustain ignition. After a delay time of circa 1 min, both carbon solids ignited spontaneously at ambient conditions giving bursts of yellow flames.

By assuming that the average size of HNO₃ molecule is 3.5 Å (i.e., close to the diameter of NO₃⁻ ion) and knowing that the interlayer spacing of the carbon nanosheets is 4.3 Å, we conclude that the HNO₃ molecules match well the interlayer spacing of both carbons. Generally, nitric acid is commonly used to intercalate graphite [54]. In the present case, the larger interlayer space allows for a higher amount of HNO₃ to be included between the layers. Hence, as shown in Figure 10, the strong oxidizer may easily access the basal spacing and defect sites of the carbon lattice, thus triggering ignition according to the following exothermic (ΔH_rxn^o < 0) and spontaneous (ΔG_rxn^o < 0) reaction at 25 °C (T = 298 K) (ΔH_rxn^o = −1626 kJ, ΔS_rxn^o = +1.22 kJ·K⁻¹, ΔG_rxn^o = ΔH_rxn^o − T∙ΔS_rxn^o = −1990 kJ):

5 C _(s) + 4 HNO_{3 (l)} → 2 N_{2 (g)} + 5 CO_{2 (g)} + 2 H₂O _(g)

The above enthalpy change value compares quite well or even better to that of other hypergolic pairs used in rocket propellants, such as the N₂H₄-H₂O₂ system, for which holds:

N₂H_{4 (l)} + 2 H₂O_{2 (l)} → N_{2 (g)} + 4 H₂O _(g) ΔH = −780 kJ

Based on the mechanism proposed in Figure 10, washing of the carbon solids with highly polar solvents (NMP, DMF or DMSO) is critical in order to remove any organic species that strongly adhere to the carbon surface, thus blocking the interlayer space. Worth noting, crystalline graphite with considerably fewer defect sites and d₀₀₂ = 3.4 Å (i.e., less than the nominal size of HNO₃, thus limiting influx of the latter molecules into the gallery space of carbon) gave no ignition with fuming nitric acid at any ratio. The same holds true also for pyrolyzed spent coffee discussed earlier with d₀₀₂ = 3.5 Å. Hence, the ignition property seems to be unique to the current piranha solution treatment. Furthermore, the proposed structural model might also explaining the observed ignition delay, e.g., HNO₃ molecules need some time to diffuse into the gallery space of carbon. Although an ignition delay is undesirable in hypergolic propellants, higher surface area carbon nanosheets (porous or/and with smaller lateral dimensions) could perhaps trigger a faster ignition. Ignition with fuming nitric acid was also found for the cardboard-derived carbon (Figure 11). In this case, thermal camera imaging of the ignition process showed a maximum reached temperature of circa 300 °C.

It is interesting to note that the relatively large interlayer spacing of the derived carbons also allows for the intercalation of ignitable flat aromatic molecules (thickness circa 3.5 Å) between the nanosheets, thus further boosting carbon hypergolicity when in contact with fuming HNO₃. To demonstrate this, we chose aniline (C₆H₅NH₂), which is well known to react hypergolically with fuming nitric acid HNO₃ [55,56,57]. A total of 1.5 g of the cardboard-derived carbon was mixed with 1.5 g aniline (99%, Sigma–Aldrich, St. Louis, MO, USA) in a sealed glass vial, and the mixture was placed at 100 °C for a day in an oven in order to force the aniline molecules into the gallery space of carbon. Aniline intercalation was evidenced by XRD (Figure 12), wherein a small increase in interlayer spacing (d₀₀₂ = 4.7 Å) and a sharpening of the (002) reflection due to better lamellar ordering were observed [58,59].

The intercalated solid (2.5 g) was placed in a flask, and 2.5 mL fuming HNO₃ was slowly added. Instantly, a bright ignition took place irrespectively of the added amount of acid (Figure 13). Therefore, aniline intercalation might eventually hold the key to overcome ignition delays and achieve faster and brighter ignitions at ambient conditions.

4. Conclusions

With its pros and cons, biomass waste carbonization in piranha solution appears to be a versatile new method for obtaining carbon materials with unusual properties not displayed by pyrolytically or hydrothermally derived carbons from biomass. As examples, we have mainly shown here the piranha solution-mediated carbonization of stale bread and spent coffee wastes, both discarded in landfills in large quantities without further processing. In both cases, carbonization proceeded fast and exothermically leading to carbon nanosheets at decent yields (25–35%); the carbon nanosheets are amorphous, displaying turbostratic arrangement of incompletely crystallized layers, have a thickness of circa 3 nm and contain O, N and S structural heteroatoms from the parent waste. Because of their relatively large interlayer spacing (4.3 Å), the obtained carbons react hypergolically upon contact with fuming nitric acid at ambient conditions. Since carbon hypergolicity is a rare phenomenon in literature, the present findings may stimulate further research in developing hypergolic carbons through the piranha solution treatment of biomass waste. Preliminary results with aniline intercalation indicate that the structural engineering of the derived carbons (surface area, porosity, lateral size and intercalation with hypergolically active molecules) may boost faster and brighter ignition, which is highly desired for rocket propellants. Such energetic carbons could be used as cheaper and less toxic alternative solid fuels in hypergolic propellant compositions.