1. Introduction
Industry is being driven evermore towards the use of renewable, environmentally benign feedstocks as the push to further sustainable development increases in pace, in line with the seventh principle of green chemistry [
1]. This is largely motivated by a requirement to rapidly decarbonise and reduce environmental impact to create a sustainable future in mitigating climate change. One such method for achieving these goals may lie in the use of abundant secondary biomass, namely lignocellulosic residues from biorefineries as a feedstock, which are a potentially rich source of chemical precursors, such as phenol and furfural.
Phenol is a value-added chemical component used in the polymer, pharmaceutical and dye industries as a feedstock. Key examples of valuable phenol derivates may generally be described as alkyl-phenols used as monomers for plastics production and more specifically compounds such as hydroquinone and salicylic acid as pharmaceutical reagents.
The state-of-the-art processing route to produce phenol is via the cumene process, at times licensed as the Hock process, during which cumene is oxidised in several stages and then cleaved to form phenol and acetone [
2]. Hence, the natural motivation in this case is to search for alternative production methods of the value-adding chemicals to further sustainable development of these industries. Lignocellulosic lignin derived from woody residues is a prime candidate, as any carbon emissions related to the process could offset by the carbon sequestration properties of the forest sources. The structure of lignin is that of a phenylpropanoid type polymer [
3], of which it is constituents are
p-hydroxyphenyl (H), guaiacyl (G) and syringyl (S) units. Lignin can be largely isolated from woody components into commercially derived products such as Etek (hydrolytic type) lignin from Sekab’s bioethanol production process [
4]. Uniquely, due to the treatment process, the lignin itself is high in derived sugars and related products [
5], which makes it suitable for pyrolysis to generate bio-oil, while makes it less attractive for production of pure lignin due to the structural modifications and rearrange on the surface of the lignocellulose when compared to organosolv methods [
6]. Therefore, the composition of the raw feedstock will have a strong impact on the composition of lignin and especially the option to separate the extractives and other lignocellulosic compounds from lignin and all that will impact the gas and bio-oil composition [
7].
Pyrolysis is an important step of most thermochemical processes that aims to densify biomass, simultaneously reducing its high O content. The conditions applied largely influence the solid, liquid and gas yields of the resulting bio-product [
7]. Since pyrolysis is not able to completely remove O, it is typically followed by hydrodeoxygenation (HDO), which minimise O content in presence of H
2 and catalysts [
8]. Growing research is focusing on hydropyrolysis, also known as H
2-aided pyrolysis, with the aim to maximise organics recovery in the bio-oil fraction and minimise coking in the HDO step [
9]. This is mostly due to the success of the ex-situ IH2
® process that Shell Catalysts and Technologies licensed from GTI and is testing at demonstration scale (5 t/day) in India [
10]. For example, the ex-situ hydropyrolysis of beech wood was studied in 26 bar hydrogen in a fluid bed reactor operated at 450 °C with several different catalysts, followed by HDO in fixed bed (370–400 °C) using a sulphided commercial NiMo/Al
2O
3 catalyst. Their results show that the best performing catalyst (HYCP) gave a condensable organic yield of 25 wt% daf corresponding to the highest obtained energy recovery of 58% and reduced coke on the spent catalyst to ~20 wt% [
9].
Pyrolysis of biomass and its components has been well studied, paying particular attention to application of catalysts by Direct Deoxygenation (DDO), in order to selectively upgrade the pyrolysis vapour products into bio-oils of desired compositions and reduce O content to minimise H
2 consumption in the following HDO step [
11,
12,
13,
14]. Other proposed reactions pathways include alkylation (AL), demethoxylation (DMO), oxidation (OX), demethylation (DME) and hydrogenation (HYD) among others [
15,
16,
17,
18].
The pyrolysis of biomasses of varying structures showed that catalytic cellulose pyrolysis produced 50–70% oxygenates (including all oxygenated aromatics except for phenols classified separately), whilst lignin pyrolysis produced 60–80% phenolics showing a clear advantage for using lignin-rich biomass for phenolics [
12]. The pyrolysis of Etek type lignin was compared to that of lignins derived from different hydrolysis processes resulting in the lowest char yield (~22%), due to its intrinsic high content in holocellulose, suggesting it is a good feedstock for pyrolysis [
3].
It is well known that DDO by catalytic pyrolysis in presence of zeolites is particularly attractive due to the selective production of aromatics but has the drawback of resulting in severe deactivation by coking and excessive carbon losses in gas phase. Metal oxides are also widely investigated with the purpose to increase the yield of desired products by first combining small oxygenates into big ones within transportation fuel range due to their superior ability to ketonisation and aldol condensation and then undergoing HDO to remove oxygen [
19]. For example, CaO can effectively reduce acids, anhydrosugars and phenols, increasing the formation of hydrocarbons and cyclopentanones, while metal oxides such as ZrO
2 and Zr
0.5Ce
0.5O
2 can efficiently promote the conversion of light oxygenates to big molecules through ketonisation and/or aldol-condensation [
19]. The use of ZrO
2-TiO
2 for the catalytic upgrading of pyrolysis vapours concerning a lignin model compound, guaiacol, resulted in the increase in phenol yields (21.4%) when using guaiacol as a feedstock [
13]. The impact on pyrolysis of cellulose and lignin of the addition of alkali metals such as Na to ZrO
2 was also studied indicating that the resulting Na/ZrO
2 promoted decomposition of cellulose favouring hydrogen production derived from cracking of pyrolysis derivatives, while, in the case of lignin, it resulted in the largest combined yield of monomeric phenolics (17.5 wt%) and alkylphenols (6 wt%) (compared to Ce, NiCe and MgCe addition), which was linked to the mild basicity of Na/ZrO
2 [
17,
18]. Addition of Ce to ZrO
2 instead resulted in increase in bio-oil yield [
18]. Furthermore, the oxidative depolymerisation of prot and alkali lignin in the presence of cobalt impregnated ZrO
2 catalysts at 140 °C selectively produced (67 area%) guaiacol monomers [
15]. These studies suggest that the mixture of basic metal oxides and ZrO
2 is a good support for lignin depolymerisation.
The literature shows that catalysts for biomass pyrolysis still suffer from coking, so that it is essential to develop coke resistant materials with the ability to retain organics in the bio-oil fraction. Since previous work showed that Na/ZrO
2 is a good catalyst for recovering phenols in bio-oil and ceria is resistant to coking, for this study, a combination of metal oxides of sodium, cerium and zirconium has been selected due to the potential synergies in enhancing the bio-oil yield and phenolics recovery and reducing coke formation in the catalytic pyrolysis of Etek lignin. To do so, an ex-situ bench scale configuration (where the biomass and the upgrading catalyst are not in contact) was selected for the catalytic pyrolysis, as it is considered a fast and reliable way to evaluate catalysts’ performance [
19]. The catalysts studied were then characterised using XRD, and the bio-oil analysed using GC-MS and FTIR. Cat_A with equimolar composition resulted in a remarkable reduction of coke to 4 wt% and a propensity to recover phenols from Etek lignin.