Steam explosion is one of the most studied pretreatments since its development for commercial scale application (Masonite technology) and has been applied success­fully on a variety of lignocellulosic materials (Ballesteros et al. 2004; Taherzadeh and Karimi 2007). This process combines thermal (high temperature), mechanical (sudden vaporization of the water) and chemical (hydrolysis of hemicelluloses) alteration of biomass. During steam explosion the biomass is exposed to saturated steam at high pressure for a period of time (seconds to several minutes) after which it is suddenly depressurised. The water penetrates into the lignocellulose structure and hydrolyses acetyl residues from the hemicelluloses, which in turn promotes partial hydrolysis of the hemicelluloses. The sudden depressurization induces the mechanical rupturing of fibres and redistribution of lignin. As a result, lignocellulose is more accessible to enzymes and the solubilised hemicelluloses can be easily recovered by filtration. However, solubilised lignin and sugars can be further degraded into compounds that can be inhibitory to further bioprocessing.

The concentration of degradation products, and therefore the extent of toxicity, depends on the source of biomass, the severity of the pretreatment (often measured as the combination of temperature and residence time) and properties of the specific enzymes/microorganisms involved in the subsequent biochemical conversion.

Despite its feedstock versatility, steam explosion is not very effective on soft­wood materials owing to its lower acetyl content. The addition of an acid catalyst (H2SO4, SO2) has been recommended as an option to improve the performance of steam explosion on softwoods and other woody materials, which will result in reduction of pretreatment severity and maximise sugar yields (Ewanick et al. 2007; Garcfa-Aparicio et al. 2011). Similarly, the application of CO2 with organic acids has been suggested to reduce the severity of pretreatment, in turn reducing hemicellulose degradation into inhibitory compounds (Gfrio et al. 2010).

Liquid hot water is a hydrothermal treatment that employs compressed hot water (pressurised and above saturation point) at high temperatures to disrupt the lignocellulose structure. Compared to steam explosion, this treatment is very effective for hemicellulose solubilisation, leading to reduced inhibitor formation when the reaction pH is kept between 4 and 7. However, the concentration of hemicellulose-derived sugars is reduced due to the higher water input.

Ammonia fibre explosion (AFEX) combines the use of liquid anhydrous ammonia at high temperature (60-100 °C) and pressure. After a variable period of heating, sudden decompression provokes the expansion of ammonia gas altering lignocel — lulose structure with limited inhibitor formation. Contrary to other thermochemical pretreatments, AFEX treated material is a solid with a similar carbohydrate compo­sition to the raw material but is more digestible by cellulases and hemicellulases. Moreover, the residual ammonia after recycling can reduce nutritional requirements in the following fermentation step. Although it has been suggested that AFEX alters lignin, reducing its ability to bind enzymes (Kumar and Wyman 2009), this pretreatment is not very effective when applied to woody biomass (Kumar et al. 2009).

In ammonia recycle percolation (ARP) the biomass is subjected to aqueous ammonia (5-15 %, w) in a flow-through system (approximately 5 ml/min) at high temperatures (normally 170 °C). ARP solubilises hemicelluloses and lignin providing cellulose-enriched residues with high digestibility. ARP has shown to be efficient in increasing digestibility of hardwoods, waste paper and softwood pulp mill sludges (Gfrio et al. 2010). Sugar degradation is minimal during ARP, but the solubilised lignin can be toxic for microorganisms.

Wet oxidation involves the exposure of biomass to oxygen or air at high temperatures (170-200 °C) and pressures (10-12 bars) for short periods of time (10-15 min). It solubilizes hemicelluloses (mainly in polymeric form) and lignin (Hendricks and Zeeman 2008). The phenolic compounds are further degraded to carboxylic acids but furans formation is lower in comparison with steam explosion or liquid hot water.

The choice of pretreatment depends on the type of feedstock and the desired biofuel output. The majority of pretreatments generate a material known as slurry which consists of a solid fraction enriched in cellulose (and lignin, depending on the pretreatment) designated as the water insoluble fraction (WIS) and a liquid fraction or prehydrolysate containing the sugars solubilised during the pretreatment (mainly hemicellulose-derived sugars). Depending on the severity of the pretreatment, the sugars can be further degraded into furans that, coupled with the solubilised lignin and acetic acid released from the hemicelluloses, impact negatively on biochemical transformation. For this reason, most studies separate slurry into the separate liquid and solid fractions to optimise the conversion of each fraction into biofuels. Current research focuses not only on the development of detoxification processes but also on the cultivation of robust microorganisms that are able to effectively convert hexoses and pentoses in slurry into biofuels. This aspect seems to have less impact for biogas production given the higher tolerance of methanogenic bacteria to inhibitors (Hendriks and Zeeman 2008).