Waste Biomass Valor (2013) 4:87–93 DOI 10.1007/s12649-013-9204-4
Catalytic Hydrothermal Degradation of Carbon Reinforced Plastic Wastes for Carbon Fibre and Chemical Feedstock Recovery Jude A. Onwudili • Eyup Yildirir • Paul T. Williams
Received: 14 September 2012 / Accepted: 4 February 2013 / Published online: 27 February 2013 Ó Springer Science+Business Media Dordrecht 2013
Abstract Recovery of carbon fibre and chemical feedstock via catalytic hydrothermal degradation of waste carbon fibre reinforced plastic (CFRP) sample was investigated in a stainless steel batch reactor between 400 and 420 °C and pressures of 20 and 25 MPa, respectively. Sodium hydroxide and potassium hydroxide were used as catalysts/additives. Using supercritical water alone, a maximum of 54.5 wt% of resin was removed from the CFRP at 420 °C, but with high recovery of phenol in the liquid residual. The presence of NaOH or KOH alone in water led to up to 81 wt% resin removal, even at short reaction times. Extracts from the liquid residual contained phenol and aniline as the major components; thus representing a potential for monomer recovery. For instance, the use of KOH alone gave phenol yield of 377 mg/(g resin) and aniline yield of 112 mg/(g resin). In addition, the presence of the alkalis led to the recovery of carbon fibres with very good mechanical properties. Keywords Reinforced plastic wastes Carbon fibre Hydrothermal Feedstock recovery
Introduction Fibre reinforced plastics (FRP) industry is valued at 775 million US dollars for the North American and European market which produced about 685,000 tonnes in 2002 . Fibre reinforced composites materials find use in the
J. A. Onwudili (&) E. Yildirir P. T. Williams Energy Research Institute, School of Process Environmental and Materials Engineering, The University of Leeds, Leeds LS2 9JT, UK e-mail: [email protected]
aerospace and automotive industries, sport tools, electronics and recently in the wind energy sector [2, 3]. The production of reinforcing fibre materials has grown steadily over the last two decades due to a variety of new applications. FRPs contain between 5 and 65 wt% reinforcing materials such as carbon fibres, glass fibres, cellulose fibres and aramid fibres but the market is dominated by carbon fibre which has 40 % share of the market, followed by the glass fibre, with 31 % market share . In the United Kingdom, the annual production rate of CFRP is about 2,130 tonnes, of which the aerospace and defence industries consume 36 %, the wind energy sector 33 %; the automotive, marine and other industrial uses 31 % . The fate of carbon fibre reinforced plastics (CFRP) or composites in end-of life applications is an interesting topic of research as these materials are often classified as nonrecyclables . By their nature and composition, CFRPs are non-biodegradable and the waste is considered hazardous in the USA. In Europe, the EU Directive on End-of Life Vehicles (Directive 2000/53/EC places the responsibility of disposal of old vehicles on manufacturers. In addition, only 15 % by weight of car can be disposed of in landfill, while the remaining 85 wt% must be reused, recycled or treated for energy recovery with effect from 2006. By 2015, the proportion of a car allowed for landfill disposal will reduce further to 5 wt%. The current rise in the application of CFRPs will lead to increased generation and disposal of CFRP wastes in the next few years as aircrafts and other CRFP-associated equipment and utilities reach their end-of-life. A variety of techniques have been applied for the recycling of carbon fibre reinforced plastic composites, including thermal degradation, mechanical treatment and chemical depolymerization . In general, there is a growing trend within the research community towards the
use of solvolysis to remove the resin materials in CRFP waste and recover the carbon fibre for re-use in suitable applications. Recently, Morin et al.  carried out an extensive review of the application of solvolysis as a chemical recycling technology for the treatment of waste CFRPs mainly for carbon fibre recovery. Supercritical solvents generally applied for solvolysis of CFRPs include water, methanol, ethanol, propan-1-ol, acetone, ethylene glycol. Among these solvent-based treatment methods, hydrothermal processes are being considered as benign options since they involve the use of water as solvent [5, 8– 10]. Under hydrothermal conditions, water attains special physico-chemical properties similar to common organic solvent. For instance, supercritical water is able to solubilize many organic compounds and offers a single-phase reaction medium with very high reaction rates . Therefore, in the application of hydrothermal processes for the recycling of CRFP, the focus would be on the ability of hydrothermal fluid to remove the resin via depolymerization or degradation and release the carbon fibre. Hydrothermal processing has shown great potential for the depolymerization and degradation of thermosetting plastics. Goto  reported the depolymerization of poly(ethylene) terephthalate (PET) in subcritical and supercritical water to yield the monomers terephthalic acid (TPA) and ethylene glycol (EG). The author found that nearly 100 % yield of TPA was achieved at 400 °C and 40 MPa after 30 min of reaction. Sugeta et al.  found that virtually no reaction occurred when a CFRP sample cured with phenolic resin was treated at a temperature of 380 °C for 5 min. However, the authors observed an improvement in the removal of resin when the reaction was carried out in the presence of alkali and an alcohol combined with water, leading to the recovery of monomers of the phenolic resin. Pin˜ero-Hernanz et al.  carried out catalytic hydrothermal depolymerization of CFRP in the presence of KOH and hydrogen peroxide and achieved 90 wt% resin removal from CRFP waste at 400 °C, 27 MPa, and the recovered carbon fibre retained between 90 and 98 % of its mechanical properties. However, Bai et al.  showed that the presence of molecular oxygen could lead to severe oxidation of the carbon fibre and a reduction of its mechanical strength. The use of alkaline hydrothermal conditions for recycling CFRP waste appears to be evolving into a viable chemical recycling process. However, recent literature is mostly limited to the removal of resins in order to recover carbon fibres. There is therefore potential to further develop this process for both carbon fibre and chemicals’ recovery in order to enhance the viability of the process. In this work, catalytic hydrothermal degradation of a CFRP sample has been carried out at temperatures of 400 and 420 °C and pressures of 20 and 25 MPa, respectively. There is a lack of detailed report in literature about the fate of the reaction products beside carbon fibre from chemical
Waste Biomass Valor (2013) 4:87–93
recycling of CFRPs. This present work therefore includes the detailed analyses of the reaction products, including the liquid residuals, with the aim of investigating the conditions suitable for the recovery of carbon fibre and chemicals from waste CFRP. This would be useful in the understanding of the hydrothermal processing of CFRP wastes.
Experimental Materials Waste CFRP sample used was obtained from an aerospace company in the UK and cut into strips of approximately 1–3 cm before use. The elemental (CHSN-O) composition of the CFRP was as follows; 84.5 % carbon, 2.4 % hydrogen, 6.6 % nitrogen, 0.6 % sulphur and 5.6 % oxygen by difference. In addition, proximate analyses showed 33 % volatile matter, 66.6 % fixed carbon and 0.4 % ash. TGA results showed that the CFRP sample was composed of approximately 61.5 % carbon fibre and 38.5 % resin. Sodium hydroxide (pellets) and potassium hydroxide (pellets) were purchased from Sigma-Aldrich, UK. Dichloromethane for the extraction of organic compounds in the liquid effluent was also obtained from SigmaAldrich, UK. Details of the 500 mL stainless steel reactor obtained from Parr Instruments Co. USA, have been provided in a previous publication . Procedure Approximately 5.0 g of the CFRP strips was loaded into the reactor along with 60 mL of distilled water. When alkalis were used, 1.0 g of NaOH or KOH was dissolved in the water, corresponding to 0.42 M NaOH and 0.30 M KOH aqueous concentrations. The reactor was then sealed and purged with nitrogen gas for 5 min and heated to the designated temperature and corresponding pressure. The heating rate was kept constant throughout the experiments and the reactor was quickly withdrawn from the heater as soon as the designated conditions were reached. The reactor was then quickly cooled to room temperature with compressed air and a fan. On cooling, the cold-gas pressure (between 2 and 6 bar) and temperature were noted before gas sampling. Analyses of Reaction Products Gaseous effluent was sampled for offline analysis by means of two Varian 3,380 gas chromatographs for permanent gases and hydrocarbons, respectively . Results of the gas analyses were obtained in mole percent from the gas
Waste Biomass Valor (2013) 4:87–93
chromatographs and converted to actual mass of individual gas yields using the ideal gas equation. The reactor contents, including liquid and solid residues were emptied into a holding beaker and separated by vacuum filtration, washing several times with distilled water. The solid residue was dried to a constant weight in an oven at 105 °C to determine its weight. Percentage resin removal was calculated on solid residue dry-weight basis by the formula; Weight of CFRP feed Weight of Solid residue 100 Weight of Resin in CFRP feed In a previous paper  the resin removal was calculated after post-oxidation of the solid residue, which gave slightly higher values. The quality of the solid residue produced from the experiment was monitored using scanning electron microscopy (SEM) to determine the surface roughness of the recovered material. A LEO 1530 Gemini FEGSEM with Oxford instruments INCA 350 EDX system by Carl Zeiss was used to analyse the morphology, diameter and the visual signs of residual resin. The solid residues were prepared by cutting a small piece of the sample and placed on carbon discs and coated with palladium powder. A portion of the aqueous residuals was then taken and analyzed for total organic carbon (TOC) and inorganic carbon (IC). These were used to compute the carbon balance results. Portions of the liquid residual were extracted with dichloromethane (DCM) for GC/MS/MS analyses. The extraction and analyses of phenol and aniline in water are often difficult due to their polar nature and high water solubility under certain pH ranges . In this study, pH adjustments of the aqueous residuals with freshly prepared 10 % HCl and 10 % KOH solutions, respectively were used to separate these two compounds during solvent extraction with DCM. In this procedure, a 20 mL portion of the aqueous residual was taken for liquid-liquid extraction after vigorous shaking of the sample. To this portion, 3 g of sodium chloride was dissolved in order to increase the ionic strength of the aqueous medium and facilitate the partitioning of the organic compounds present into the organic solvent (DCM). Thereafter the pH of this portion was firstly raised to 12 with 10 % KOH to extract free aniline; after which 10 % HCl was used to bring the pH of the same sample to 2 for phenol extraction. The 20 mL portions taken were extracted three times with 20 mL of DCM each time to ensure quantitative extraction and the extract combined and thoroughly mixed. The combined volume of each of the extracts of *60 mL was noted before GC analysis. Furthermore, 100, 1,000 and 2,000 ppm standard solutions of phenol and aniline in water were, respectively prepared for recovery analyses. These solutions were extracted with DCM under conditions identical to the ones used for the experimental samples.
The DCM extract were then analyzed on a GC/MS/MS instrument using external standard method for phenol and aniline. The GC/MS/MS system consisted of a Varian 3800-GC coupled to a Varian Saturn 2200 ion trap MS/MS equipment. The column used was a 30 m 9 0.25 mm inner diameter Varian VF-5 ms (DB-5 equivalent) while the carrier gas was helium, at a constant flow rate of 1 mL min-1. The GC injector was held at 290 °C. The oven temperature programmed was as follows; 40 °C held for 2 min and ramped to 280 °C at a rate of 5 °C min-1 and then held at 280 °C for 10 min. The transfer line temperature was 280 °C, manifold was at 120 °C and the trap temperature was held at 200 °C. Spectral searches on the installed NIST 2008 Library  were used to qualitatively identify the major ‘unknown’ compounds in the oil products. The recovery analyses which were done in triplicates gave average phenol recovery of 86 % with a standard deviation of 5.2; while average aniline recovery was 91 % with a standard deviation of 4.5. These values were used to adjust the yields of phenol and aniline in the extracted aqueous residuals.
Results and Discussion Resin Removal Effect of Different Reaction Media and Reaction Temperature Table 1 shows the percent products distribution in relation to reaction temperature and hydrothermal reaction media after the treatment of CRFP waste in this present study. Three sets of experiments were carried out involving different reaction media including water alone and two aqueous solutions containing 1.0 g (0.42 M) NaOH and 1.0 g (0.30 M) KOH each. Each experiment was conducted at 400 and 420 °C, respectively and for zero min reaction times at the designated temperatures. Due to the use of water as reaction medium, it was difficult to evaluate the exact amount of liquid products; however since both the solid and gas products were accurately determined, the liquid products were obtained by difference. Sodium hydroxide gave the lowest yields of solid residue, which corresponded to the highest yields of liquid products at the two reaction temperatures in this study. Apparently, more degradation of the resin materials was achieved in the presence of the alkaline additives than with supercritical water alone. Since the CFRP was the only source of carbon, accounting for the carbon distribution would give the best measure of the degradation of the CFRP. Table 2 presents
Waste Biomass Valor (2013) 4:87–93
the carbon balance from these experiments and in all, carbon balance closures were above 93 %. As expected, major proportions of the carbon in the feed were obtained as solid residue. Increase in temperature would weaken the bonds between carbon fibre and resin and this would clearly increase the ability of supercritical water to attack the already weakened bonds between the carbon fibre and the resin material. This could be explained by increased hydrogen bonding between water and the hydroxyl, phenolic, carboxyl and carbonyl groups present within the cross-liked polymer [18, 19]. Hence, there were slight improvements in the resin removal efficiency with increasing reaction temperature from 400 to 420 °C, with corresponding reduction in the amount of solid residue. In addition, the introduction of NaOH and KOH changed the product distribution as well as improved the resin removal efficiency as the temperature increased. These additives have the ability to increase the hydroxide ions in supercritical water and therefore increase the hydrolysis of the resin as well as cleaving the bonds between the resin and the carbon fibre . About 54.5 wt% of resin removal was obtained in supercritical water alone. Both sodium hydroxide and potassium hydroxide improved the degradation of the resin materials by up to 30 % compared to experiments with water alone. Sodium hydroxide appeared to be more effective than potassium hydroxide in degrading the resin into liquid products, comparing the yields of solid residue in Table 1. NaOH and KOH have been extensively used in the hydrothermal conversion of organic materials, such as biomass, into useful products including chemicals such as acetic acid [20, 21]. These two alkalis have also been used
for the production of hydrogen gas during alkaline hydrothermal gasification of biomass [14, 22–24]. The mechanism of reaction between these alkalis and biomass appears to involve the cleavage of polar bonds via nucleophilic attack , leading to the production of simple molecules for gasification. A similar mechanism may be responsible for their high activity in the depolymerization of resins. In addition, Pin˜ero-Hernanz et al.  have shown that KOH can be used to recover carbon fibre from CRFP in supercritical water due to the ability of the alkali to degrade the resin. These authors achieved up to 95 wt% resin removal in the presence of KOH but with 10 % loss in the mechanical properties of the recovered carbon fibre. In this present work, resin removal in the presence of the alkalis ranged from 81.8 to 85.2 % and KOH gave higher resin removal than NaOH in almost all cases. Interestingly, the resin removal ability of either NaOH or KOH did not differ significantly with reaction time. Chemicals Recovery The GC/MS/MS chromatogram presented in Fig. 1 shows the presence of a huge peak corresponding to the co-elution
Table 1 Products distribution from carbon fibre reinforced plastic waste under different hydrothermal media (reaction time = 0 min) Reaction media
Water ? NaOH
Water ? KOH
Table 2 Carbon distribution in relation to reaction media and temperature during hydrothermal processing of carbon fibre reinforced plastic waste (reaction time = 0 min)
Fig. 1 Typical GC/MS/MS chromatogram with major components of the organic extract from hydrothermal processing of CFRP waste (coelution of phenol and aniline)
Solid residue (g)
Resin removal (%)
Waste Biomass Valor (2013) 4:87–93
Fig. 2 Effect of reaction media and temperature on the yields of phenols and aniline during hydrothermal processing of CRFP
Table 3 Mechanical properties of virgin and recovered fibre at 420 °C in specified hydrothermal media Recovered carbon fibre/0.3 M KOH % Retention
Recovered carbon fibre/0.42 M NaOH
Virgin carbon fibre
of aniline and phenol, with molecular ion peaks at m/z 93.1 and 94 respectively. The individual yields of both phenol and aniline were obtained after their separation by liquidliquid extraction using pH adjustment and quantified on the GC/MS/MS, as described earlier. From Fig. 1 it can be seen that both phenol and aniline represented more than 90 % of the total peak areas on the chromatogram and are such were the predominant organic products in the liquid residuals obtained from the hydrothermal processing of the CFRP sample. Apart from phenol and aniline, minor degradation products detected include alkyl phenols, toluidines, quinolines and N-substituted anilines. Previous thermogravimetric analysis suggested that the resin used in this CFRP was possibly based on polybenzoxazine, of which phenol and aniline would be feedstocks . Hence, it was not surprising that these compounds were dominant
in the DCM extracts from obtained in this present study. Nahil and Williams  also reported that aniline was the major compound in the oil obtained from the pyrolysis of a similar sample of CFRP. The results of these analyses for some selected liquid residuals are shown in Fig. 2. Given that the CFRP waste contained 38.5 % resin, the nominal weight of resin in 5 g of CFRP would be approximately 1.93 g. The yields of phenol and aniline in mg/g were based on this nominal weight of resin in the CFRP. With supercritical water alone, the yield of phenol and aniline were 254 and 4.19 mg/g at 400 °C and this increased to 355 and 6.46 mg/g at 420 °C, which showed that phenol yield increased by 38 % whereas aniline increased by 54 %. It would appear that the degradation of the resin favoured the production of phenol over aniline. With the use of KOH and NaOH as alkaline hydrothermal media at 420 °C appreciable yields of both phenol and aniline were obtained. KOH gave 377 and 112 mg/g of phenol and aniline, respectively, which were the highest yields of both monomers obtained in this study. In addition, these alkalis gave the highest yields of N-methyl aniline at 38 mg/g for KOH and 22 mg/g for NaOH. Recovered Carbon Fibre Table 3 shows the critical mechanical properties of the recovered fibre under different hydrothermal reaction media at 420 °C. The use of KOH and NaOH alone gave carbon fibres with good mechanical properties, some of which are more than 90 % compared to the properties of a sample of virgin carbon fibre. The carbon fibre recovered
Waste Biomass Valor (2013) 4:87–93
Fig. 3 SEM images (magnification: 95,000) of a Virgin carbon fibre; b recovered carbon fibre with KOH at 420 °C
from the use of 0.3 M KOH gave slightly better properties than the one involving the use of 0.42 M NaOH as shown in Table 3. This could be related to the difference in the actual concentrations of the alkalis even though 1.0 g of each was used. Pin˜ero-Hernanz et al.  showed that with 0.5 M KOH in supercritical water, up to 10 % loss in tensile strength of the recovered fibre was observed compare to virgin fibre at 400 °C and 28 MPa. The results from the use of alkalis presented in this study compare very well with literature but also showed that lower concentrations of alkali could be used at slightly higher temperature of 420 °C and a reaction time of zero min. The observed changes in the mechanical properties of the recovered carbon fibre could be attributed to the increased elongation of individual fibres by up to 17 % after the alkaline hydrothermal degradation process. Fig. 3 presents the SEM pictures of a virgin fibre and the one recovered from waste CFRP at 420 °C in the presence of KOH. Clearly, no visible difference could be observed between the two carbon fibres; indeed the recovered fibres appeared cleaner. The similarity in the appearances of the carbon fibre surface may imply that the surface properties may be similar; however this would need to be experimentally determined in future work.
Conclusions Results from the catalytic hydrothermal depolymerization of waste CFRP at 400 and 420 °C. The presence of the alkalis led to the recovery of carbon fibres with very good mechanical properties comparable to literature. In addition, the alkalis gave more than 81 % resin removal but produced high yields of phenol and aniline, while supercritical water alone gave the lowest resin removal, high yields of phenol but very little yields of aniline. The liquid effluent obtained from the resin depolymerization process using 0.3 M KOH yielded phenol and aniline of up to 377 and 112 mg/g of resin content of CRFP, respectively. The mechanical property of the recovered carbon fibre such as tensile strength, Young’s modulus and breaking force decreased appreciably, while elongation increased by up to
17 % after reacting with the alkalis. Further work on the surface properties of the recovered fibre would be required to ascertain the possible applications of the recovered carbon fibres. Acknowledgments The authors would like to thank the Turkish Government for the provision of a PhD Scholarships for E. Yildirir.
References 1. Thomas, S., Pothan, L.: Natural Fibre Reinforced Polymer Composites: From Macro to Nanoscale. Old City Publishing Incorporated, Philadelphia (2008) 2. Naus, D.J., Corum, J.M., Battiste, R.L., Klett, L., Davenport, M.: Durability of carbon fibre composites. Automotive Light Weighthing Materials, Progress Report (2005) 3. Cunliffe, A.M., Jones, N., Williams, P.T.: Recycling of fibre and reinforced polymeric waste by pyrolysis: thermo-gravimetric and bench-scale investigations. J. Anal. Appl. Pyrol. 70, 315–338 (2003) 4. Job, S.: Composite recycling: summary of recent research and development. Knowledge transfer network materials. (2010). Available at: http://www.compositesuk.co.uk. Accessed Aug 25 2012 5. Bai, Y., Wang, Z., Feng, L.: Chemical recycling of carbon fibres reinforced epoxy resin composites in oxygen in supercritical water. Mater. Design 31, 999–1002 (2010) 6. Nahil, M.A., Williams, P.T.: Recycling of carbon fibre reinforced polymeric waste for the production of activated carbon fibres. J. Anal. Appl. Pyrol. 91, 67–75 (2011) 7. Morin, C., Loppinet-Serani, A., Cansell, F., Aymonier, C.: Nearand supercritical solvolysis of carbon fibre reinforced polymers (CFRPs) for recycling carbon fibres as a valuable resource: State of the art. J. Supercrit. Fluids 66, 232–240 (2012) 8. Pin˜ero-Hernanz, R., Dodds, C., Hyde, J., Garcia-Serna, J., Poliakoff, M., Lester, E., Cocero, M.J., Kingman, S., Pickering, S., Wong, K.H.: Chemical recycling of carbon fibre resin composites in subcritical water: Synergistic effect of phenol and KOH on the decomposition efficiency. Polym. Degrad. Stab. 97, 214–220 (2008) 9. Yuyan, L., Guohua, S., Linghui, M.: Recycling of carbon fibre reinforced composites using water in subcritical conditions. Mater. Sci. Eng. A 520, 179–183 (2009) 10. Suzuki, U., Tagaya, H., Asou, T., Kadokawa, J., Chiba, K.: Decomposition of pre-polymers and molding materials of phenol resin in subcritical and supercritical water under an Argon atmosphere. Ind. Eng. Chem. Res. 38(4), 391–1395 (1999) 11. Goto, M.: Super-critical water process for the chemical recycling of waste plastics. The 2nd International Symposium on Aqua
Waste Biomass Valor (2013) 4:87–93
Science, Water Resource and Low Carbon Energy, AIP Confer. Proceed. 1251, 169–172 (2010) Sugeta, T., Nagaoka, S., Otake, K., Sako, T.: Decomposition of fibre reinforced plastics using fluid at high temperature and pressure. Kobunshi Ronbunshu 58(10), 557–563 (2001) Williams, P.T., Onwudili, J.A.: Composition of products from the supercritical water gasification of glucose: a model biomass compound. Ind. Eng. Chem. Res. 44, 8739–8749 (2005) Onwudili, J.A., Williams, P.T.: Role of sodium hydroxide in the production of hydrogen gas from the hydrothermal gasification of biomass. Int. J. Hydrogen Energy 34, 5645–5656 (2009) Onwudili, J. A., Yildirir, E., Williams, P.T., Recycling of fibre reinforced plastic wastes by catalytic hydrothermal degradation. 4th International Conference on Engineering for Waste and Biomass Valorisation (2012) Santana, C.M., Ferrera, Z.S., Padro´n, M.E.T., Rodrı´guez, J.J.S.: Methodologies for the extraction of phenolic compounds from environmental samples: new approaches. Molecules 14, 298–320 (2009) Varian, Inc, NIST08 Mass Spectral Library CD-Rom Number 360601818 (2008) Goodship, V., Ogur, E.O.: Polymer processing with supercritical fluids. Shropshire: Rapra Review Reports: Rapra Technology Limited; 6–25 (2004)
93 19. Moreno-Castilla, C.: Adsorption of organic molecules from aqueous solutions on carbon materials. Carbon 42, 83–92 (2004) 20. Hsieh, Y., Du, Y., Jin, F., Zhou, Z., Enomoto, H.: Alkaline pretreatment of rice hulls for hydrothermal production of acetic acid. Chem. Eng. Res. Design 8(7), 13–18 (2009) 21. Kishida, H., Jin, F., Yan, X., Moriya, T., Enomoto, H.: Formation of lactic acid from glycolaldehyde by alkaline hydrothermal reaction. Carbohydr. Res. 341, 2619–2623 (2006) 22. Watanabe, M., Sato, T., Inomata, H., Smith, R.L., Arai, K., Kruse, A., Dinjus, E.: Chemical reactions of C-1 compounds in near-critical and supercritical water. Chem. Rev. 104, 5803–5821 (2004) 23. Yang, F., Hanna, M.A., Marx, D.B., Sun, R.: Optimization of hydrogen production from supercritical water gasification of crude glycerol—byproduct of biodiesel production, Int. J. Energy Res. (2012). DOI:10.1002/er.2969. In Press 24. Onwudili, J.A., Williams, P.T.: Reactions of different carbonaceous materials in alkaline hydrothermal media for hydrogen gas production. Green Chem 13, 2837–2843 (2011)