Galbe and Wallberg Biotechnol Biofuels

https://doi.org/10.1186/s13068-019-1634-1
(2019) 12:294
Biotechnology for Biofuels

REVIEW Open Access

Pretreatment for biorefineries: a review

of common methods for efficient utilisation
of lignocellulosic materials
Mats Galbe* and Ola Wallberg

Abstract
The implementation of biorefineries based on lignocellulosic materials as an alternative to fossil-based refineries calls
for efficient methods for fractionation and recovery of the products. The focus for the biorefinery concept for utilisa-
tion of biomass has shifted, from design of more or less energy-driven biorefineries, to much more versatile facilities
where chemicals and energy carriers can be produced. The sugar-based biorefinery platform requires pretreatment
of lignocellulosic materials, which can be very recalcitrant, to improve further processing through enzymatic hydroly-
sis, and for other downstream unit operations. This review summarises the development in the field of pretreatment
(and to some extent, of fractionation) of various lignocellulosic materials. The number of publications indicates that
biomass pretreatment plays a very important role for the biorefinery concept to be realised in full scale. The traditional
pretreatment methods, for example, steam pretreatment (explosion), organosolv and hydrothermal treatment are
covered in the review. In addition, the rapidly increasing interest for chemical treatment employing ionic liquids and
deep-eutectic solvents are discussed and reviewed. It can be concluded that the huge variation of lignocellulosic
materials makes it difficult to find a general process design for a biorefinery. Therefore, it is difficult to define “the best
pretreatment” method. In the end, this depends on the proposed application, and any recommendation of a suitable
pretreatment method must be based on a thorough techno-economic evaluation.
Keywords: Review, Biomass, Pretreatment, Lignocellulosic materials, Biorefinery, Fractionation

Background production to be 1 × 1011 tons [3]. About 5%, or 225 EJ/

Lignocellulosic biomass has been suggested as an impor- year, is equivalent to almost half of the global energy
tant part of a sustainable society for a few decades. It is demand at present. This estimated demand (225 EJ/year)
viewed as an alternative to fossil carbon sources, such as is in line with results from other studies. Berndes et al. [4]
oil, natural gas and coal. An example is the global biofuel gathered data from 17 studies in a review and concluded
production, which in 2016 was estimated to be 137 billion the contribution from biomass to be between 100 EJ/year
L (3.3 EJ) [1]. Another field of application is for produc- and above 400 EJ/year in 2050. The main reasons for the
tion of polymers, fertilisers, etc., but the transportation quite large difference are two major uncertainties: avail-
sector is the main consumer of energy-related products. ability of land and yield in crop production. In the same
The annual global production of biomass is estimated publication, the assessed plantation capacity is estimated
to be about 4500 EJ of solar energy that is captured per to yield a sustainable availability between 50 and 240 EJ/
year [2]. Another estimation presents the global annual year. Thus, it is very difficult to make accurate estima-
tions of the actual amount available. Although global
bio-based chemical and polymer production is estimated
*Correspondence: Mats.Galbe@chemeng.lth.se
Department of Chemical Engineering, Lund University, P.O. Box 124, 221 to be around 50 million tonnes, the historic low price
00 Lund, Sweden of fossil feedstock together with optimised production

© The Author(s) 2019. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material
in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material
is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the
permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat​iveco​
mmons​.org/licen​ses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creat​iveco​mmons​.org/publi​cdoma​in/
zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 2 of 26

processes have restricted commercial production of bio- Some important considerations have been suggested
based products [5]. Some processes for production of for the biorefinery concept to become a path forward
chemicals have already reached the required technical towards a less fossil-dependent society. The development
maturity to be produced through biological routes. For of biorefineries is a vital key for integration of food, feed,
example, most lactic acid production takes place through chemicals, fuels and energy production in the future.
a biological route, while the classical chemical route is Combinations of physical and biotechnological processes
less important [6]. for production of proteins, but also for platform chemi-
cals such as lactic acid will be of importance in the future
[9, 10]. Biomass can mitigate, to some extent, the high
The biorefinery concept and full‑scale operation atmospheric levels of carbon dioxide by replacing fossil
A biorefinery can be defined as the renewable equivalent fuels; in addition, in many countries around the world,
of a petroleum refinery, the main difference being the raw the concept may be important to secure domestic energy
material. In the biorefinery, biomass can be converted carriers and the supply of chemicals. Another aspect
into a wide range of chemicals and energy carriers, and it is the possibility for erection of new industries in rural
can also contribute to the development of circular econ- areas, where it is likely that many of the large biorefin-
omy; this concept is based on the model that lignocellu- eries will be located [11]. Most lignocellulosic feedstock
losic materials, which were used to generate bio-based is generally much more complex and recalcitrant than
products can be recovered (to a certain degree), and be the currently used starchy materials in, e.g. the ethanol
recovered and recycled [7]. The International Energy industry, which poses a challenge [12]. Nevertheless, a
Agency Bioenergy Task 42 defines biorefining as “the number of demo-scale plants have been put in opera-
sustainable processing of biomass into a spectrum of tion during the last 10–15 years. Several companies
marketable bio-based products (chemicals, materials) have evaluated ethanol production through transforma-
and bioenergy (biofuels, power, heat)” [8]. However, the tion of cellulose to second-generation sugar products
different types of raw material constitute a tremendous for conversion to ethanol and other chemicals [5]. An
challenge when a large-scale production facility is con- easily accessible source (for a quick overview of various
sidered. In Fig. 1, a schematic overview of a potential facilities worldwide for the production of advanced liquid
biorefinery is shown. and gaseous biofuels intended for transport purposes) is

Lignocellulosic crops Lignocellulosic waste

Pretreatment

Hydrolysis

Lignin C6 sugars C5 sugars

Pyrolysis Gasificaon Separaon Fermentaon Chemical reacon

Chemical Polymers Bio-

Food and
Biodiesel CH 4 H2 Ethanol Building and resins materials
Feed
blocks

Fig. 1 A schematic representation of a biorefinery for production of energy carriers and chemicals
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 3 of 26

maintained by the International Energy Agency (IEA Bio- Table 1 A selection of challenges for successful
energy, Task 39) [13], which shows that there is a great implementation of biorefineries
interest in energy carriers from biomass. Some examples Problem Challenge
are POET-DSM, who operates a demo-scale plant at TRL
8, in Emmetsburg, Iowa, USA with a capacity of 75,000 t/ Scale-up to industrial scale Requires significant capital investment
Requires strong financial incentive
year of ethanol. In Finland, the company North European Investors find too low return on investment
Oil Trade Oy (formerly ST1) is operating a demo plant at Future situation unclear Laws and regulations not clear
TRL 6-7 in Kajaani for production of cellulosic ethanol Construction and design Delays in erection of plant
at a capacity of about 8000 t/year for demonstration of Testing of equipment
their conversion technology called “Cellulonix”. Similar Biomass availability All-year round supply of suitable materials
projects can be found in Brazil (Gran Bio, Bioflex 1), as Possibility to run on more than one material
well as in China (Henan 1), the NREL Integrated Biore- Logistics and supply Storage and transportation must be reliable
finery Research Facility (IBRF) in Golden (Co, USA), and Data for process design Transfer of smaller-scale data to industrial
scale
in several other locations all over the world. However,
Maturity of a process Handling at high pressures and feeding, e.g.
many earlier projects are now idle or changing focus. An for 2G plants of ethanol causes produc-
example is a demo plant operated by DuPont in Vonore, tion stop
TN, US, which is currently idle as DuPont is refocusing
on enzyme production. The data gathered from demo-
scale plants are extremely important, since they have
resulted in large sets of engineering design information, the important factors are: (1) biomass availability, which
which are necessary for the full-scale design of future includes all-year round supply of lignocellulosic biomass
biorefineries. However, since most of the recorded data at competitive cost, and access to biomass of similar kind
are proprietary information, it is not likely to be acces- to be able to run a biorefinery within a narrow range of
sible to the general engineering or science community. operating conditions. (2) All logistics and a fully working
In addition, many pulp mills can be considered, to some supply chain must be available, e.g. transports of low-
extent, to be operated as biorefineries. Today, not only density materials (straw, bagasse, etc.), as well as storage
pulp is a valuable product, but also tall oil from the kraft facilities for continuous operation to be possible. Feed-
process, which can be transformed to a diesel-like liquid stock handling and transportation are the most promi-
fuel [14], soil conditioners from organic materials [15], nent costs associated with the biorefinery concept. In the
fibres for textile production [16], lignin [17], vanillin, case of biofuel production, as much as 40–60% of the full-
microfibrillar cellulose [18], and other valuable products, scale production costs can be attributed to feedstock [20,
which in the future may be adding to the revenue from 21]. Tao et al. performed process and techno-economic
pulp production. analysis of six biomass pretreatment processes [22]. The
A number of challenges have been identified for imple- overall ethanol production, total capital investment and
mentation of successful future biorefineries, which are minimum ethanol-selling price were compared for the
summarised in Table 1. Sanford et al. [19] argues that selected methods. The result indicates that there was
scaling up a successful small-scale operation to a large- no large difference in economic performance except if a
scale biorefinery is the first step towards a significant process was operated in a way, which resulted in low glu-
capital investment. Therefore, a very strong economic cose and ethanol yields. In addition to the factors previ-
incentive must be present to justify the investment. In ously discussed, other important challenges to overcome
addition, investors also recognise the low return on include process mechanisation; transfer of small-scale
invested capital and an unstable future situation, as the laboratory data to engineering design data; and the lack
laws regarding biofuels and biochemicals are not yet of technical maturity in 2G lignocellulosic technology,
long-term stable, which means that the required financial such as feeding issues with high-solids suspension to a
ground is not yet fully prepared for large investments. high-pressure reactor.
They also conclude that many biotechnology companies,
who have been running a successful early-stage process, Platform chemicals
have suffered severe problems due to the challenges asso- A number of chemicals—platform chemicals—are more
ciated with scale-up and effects from delays in construc- important in the biorefinery concept than other products
tion, testing and operation. In some cases, capital costs or intermediate compounds. The reason for their impor-
have been cut by retrofitting existing plants. A number tance is that they can to be used as building blocks for
of bottlenecks and possible solutions for commercialisa- production of high-value chemicals. These often consti-
tion have been identified by Chandel et al. [5]. Some of tute chemicals having multi-functional groups that can
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 4 of 26

be converted to other groups of useful molecules. An expenditure costs (CAPEX) for the processing equip-
important feature of a platform chemical is its applica- ment. For less developed processes, the available data
bility for production of a multitude of compounds, such are more scarce and unreliable. The comparison between
as various forms of plastics, binders, fibres, energy carri- processes at different maturity would likely result in a
ers, etc. In a joint report from Pacific Northwest National skewed view on the performance of the different pro-
Laboratory (PNNL), National Renewable Energy Labo- cesses and would provide little general information. As
ratory (NREL) and the Office of Energy Efficiency & with LCA analysis, a techno-economic analysis there-
Renewable Energy (EERE), [23] 12 sugar-based building fore need to be performed on a case-by-case basis for a
blocks are targeted as the most important. These include certain production process integrated with the pretreat-
1,4-diacids (succinic, fumaric and malic), 2,5-furan dicar- ment process. Several estimations on techno-economic
boxylic acid, 3-hydroxy propionic acid, aspartic acid, analysis and LCA have been published over the years.
glucaric acid, glutamic acid, itaconic acid, levulinic acid, Aden and Foust [24], as well as Humbird et al. [25] per-
3-hydroxybutyrolactone, glycerol, sorbitol, and xylitol/ formed a techno-economic analysis on a process based
arabinitol. A couple of these are considered to be of more on dilute-acid pretreatment of corn stover followed by
interest in terms of their commercial usefulness: succinic, enzymatic hydrolysis and fermentation. In the studies,
fumaric, and malic acid, since they today form the basis it is concluded that techno-economic analysis plays an
for many high-value replacement products. important role in process development for targeting of
technical as well as economic hurdles, which must be
Techno‑economic estimations and life‑cycle analysis (LCA) studied for production to be successful. In another study,
for biorefineries Barta et al. [26] studied the process economics of com-
Pretreatment is one of the key unit operations in a biore- bined ethanol, biogas, heat and electricity production
finery. Thus, it is critical to evaluate the pretreatment from industrial hemp. One of the results in the study is
step in conjunction with all related equipment and pro- that it is important to maximise the recovery of poten-
cess solutions in a proposed biorefinery. Life-cycle analy- tial energy carriers in the raw material, since feedstock is
sis and techno-economic evaluations are important tools the largest cost contributor in the process. Similar studies
for finding suitable and sustainable production methods. have been presented by Dias et al. [27] on improvement
It is not possible to evaluate stand-alone pretreatment of distillation, cogeneration systems and heat integration
systems, since they are always part of a larger processing based on sugar cane bio-ethanol production, while Joels-
concept for producing fuels and chemicals. These pro- son et al. [28] and Börjesson et al. [29] studied potential
duction processes can, and in most cases do, result in a biorefineries, using agricultural and forestry residues,
pallet of products. Some of the pretreatment processes in terms of techno-economic factors, system integra-
are specialised towards certain products, while oth- tion, LCA and feedstock supply. The study aimed at find-
ers are more generic for the sugar platform. Regardless ing suitable locations for a full-scale refinery in Sweden
of which, since the pretreatment step is integrated with where various factors were evaluated, such as integration
other processing steps it is difficult to predict the data for with pulp mills, export of heat to district heating systems,
a life-cycle analysis based on a stand-alone pretreatment etc., Budzinski et al. [30] utilised a hybrid LCA multi-
process only. The primary energy for the pretreatment objective optimisation model to evaluate two constraints.
may be supplied by a downstream processing step at a The first was based on maximising profit, while the sec-
higher temperature level, and the pretreatment process ond was based on minimising global impact on climate
may supply secondary heat the rest of the process. Both change. In the study, ethanol production was found to be
of these alternatives would make the pretreatment pro- more cost-effective than ethylene production; however,
cess a net zero user of energy, which in an LCA analysis in terms of reducing climate impacts ethylene produc-
of a pretreatment system is one of the major parts, since tion was the better alternative. It must be stressed that
one of the major LCA “costs” for a biorefinery is the pri- techno-economic evaluations, as well as LCA cannot be
mary energy supply. If a by-product from the biorefinery performed without reliable experimental data. In addi-
is utilised, or if an external energy supply is used, this will tion, a rather detailed process design together with infor-
strongly influence the LCA performance. It is therefore mation on supply, logistics, economic details, etc., must
difficult to make an LCA comparison between different be available.
pretreatment processes without taking into account the
surrounding processes and the integration with these. Biomass
However, in terms of techno-economic analysis of pre- “Biomass” is a description for a very heterogeneous
treatment, it can be performed to some extent for some group of materials, which sometimes involves also
of the more mature processes in terms of capital and microorganisms. In this review, the term is used for
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 5 of 26

various plant-based material, such as agricultural, forest origin. It is not likely that a biorefinery will be capable of
and herbaceous matter. Despite great efforts invested processing all sorts of lignocellulosic materials.
worldwide to improve cellulose digestibility, cellulosic
biofuels and chemicals have yet to be cost-competitive The purpose of biomass pretreatment
with their starch-based counterparts. The challenge is The purpose of the biomass pretreatment step has some-
primarily due to their high production costs associ- what shifted during the last decade(s); previously, the
ated with pretreatment and enzymatic hydrolysis (and main interest was to use lignocellulosic materials for
the raw material itself ). The variety of plant species mainly bioethanol production. The interest in the other
is immense, which has a direct impact on the process main compounds, lignin and hemicelluloses, was lim-
concepts for their utilisation. The availability of various ited. Today, it is of great importance to find ways to max-
types of lignocellulosic materials varies from country to imise the overall yield of the valuable compounds that
country, and continent to continent. In certain regions, make up lignocellulosic materials. Pretreatment methods
forests are abundant, while agricultural species are that enable efficient recovery of carbohydrates as well as
more common in other regions. Lignocellulosic mate- lignin are desired; however, this all depends on the situ-
rials differ from one species to another. However, the ation and the final product. The energy requirements in
main constituents are basically the same, although the the production process must be met under any circum-
contents of individual carbohydrates, aromatics and stance, either by internal or external integration of high-
other compounds vary: about 50–60% are carbohy- energy streams, such as in a mill producing pulp, where
drates, i.e. cellulose and hemicelluloses, 20–30% lignin, the excess lignin is the main process-energy supplier. It
while the rest consist of extractives, fatty acids, ash, has been estimated that—in an optimised mill—between
etc. [31]. Cellulose is made up of cellobiose units, while 20 and 30% of the lignin is available for other purposes
hemicelluloses finds their structure from a mixture of than for internal energy requirements [34].
hexose and pentose sugars, in combination with vari- Pretreatment is a step that is included as one of the first
ous organic acids [32]. Hemicelluloses are more hydro- steps in the process, to alleviate access to the raw mate-
philic than cellulose and easier to hydrolyse. Lignin is rial. It is difficult to define “the best” pretreatment for
the major non-carbohydrate constituent in lignocellu- all situations and raw materials; however, it is vital that
losic materials comprising a very complex structure of some important features of the pretreatment method
aromatic compounds. It is associated with both cellu- are fulfilled, such as a high recovery of the individual
lose and hemicelluloses and is an important reason for polymers and other compounds in the lignocellulosic
the high strength of many lignocellulosic materials. In material. In addition, the formation of toxic or inhib-
general, agricultural crops and hardwood contain more iting compounds must be low to decrease the risk for
pentose sugars than does softwood [33]. The predomi- negative effects in, e.g. the enzymatic hydrolysis and fer-
nant sugar in most softwood species is mannose, while mentation steps, if they are part of the process. It is well
in hardwood and agricultural species, pentose sugars known that too severe conditions during pretreatment
are in majority [33]. will cause greater degradation of hemicellulosic sugars,
One of the major challenges that the biorefinery con- which can cause formation of highly toxic compounds,
cept faces to become successful is to find suitable raw such as furfural, HMF and organic acids. In addition, a
materials. It is likely that second-grade or waste materi- plethora of other compounds may be generated; however,
als will be the main raw material supply in a biorefinery. furfural and HMF are often used as proxies for the gen-
This includes straw, bagasse, tree roots, branches, forest eral content of inhibitory compounds [35]. Preferably, the
thinnings, etc. However, a large part of the published energy requirement must be as low as possible. It is also
research that deals with woody materials are often based an advantage if energy integration between the pretreat-
on wood chips of high quality. This is in direct compe- ment step and other parts of the production facility can
tition with, e.g. the interests of pulp producers. On the be implemented, such as utilisation of low-grade steam
other hand, the residues from agricultural operations are for distillation purposes. It has also been established that
in many cases available for conversion to other valuable the economic success of a biorefinery is heavily depend-
products. However, to maintain soil quality, such as the ent on the solid content of the pretreated materials. If too
level of nutrients, not all straw can be removed from the dilute solutions are produced the energy costs for puri-
field; an alternative is to return processed residues to the fication may be prohibitively high, which can cause an
field. Other types of waste materials include municipal otherwise well-functioning pretreatment method to be
solid waste, waste textile, waste-construction materials, discarded [36].
etc. Thus, it calls for very robust and versatile production The classical way to discuss pretreatment methods has
methods to be able to handle raw materials of different been to divide them into chemical, physical and mixtures
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 6 of 26

thereof, denoted physico-chemical. This is a somewhat detrimental effect on enzymatic hydrolysis and fermen-
arbitrary classification, which largely becomes more and tation. A seemingly successful pretreatment, in terms of
more out-dated since the main aim of a successful pre- hydrolysis efficacy and yield, may be far too toxic for a
treatment is now not only to achieve a cellulose fraction fermenting organism to cope with, which can result in a
that can be easily hydrolysed at a high yield, but also to stuck fermentation, such as after a too severe acid-cata-
take care of other constituents. Various forms of “pre- lysed pretreatment [38]. On the other hand, if the pur-
treatment” methods have been used for a long period pose is to produce, e.g. polymers, gels or binders, other
of time. Already in the 1940s, the Scholler process was evaluation criteria must be employed. These can include
a means to hydrolyse wood for production of ethanol mechanical strength and swelling properties of the poly-
for fuel and chemicals. However, the Scholler process is mers [39, 40]. Regarding lignin, its reactivity can be an
strictly speaking not a pretreatment method, but rather appropriate indicator [41].
an acid hydrolysis using 0.5% ­H2SO4 at around 130° for Since several platform chemicals have their origin in
45 min to hydrolyse cellulose [37]. The definition of the monomeric sugars, (e.g. glucose), enzymatic hydrolysis
unit operation pretreatment is rather vague. When does of the pretreated material is probably the most impor-
acid-catalysed steam pretreatment of a lignocellulosic tant indicator. This can be a very time-consuming pro-
material become acid hydrolysis instead? Can soda cook- cedure, since enzymatic hydrolysis requires a rather
ing of spruce be considered as pretreatment? In general, a long residence time, often between 24 and 96 h. An
pretreatment step is combined with some kind of another approach to reduce the total time and the need for many
treatment, such as enzymatic hydrolysis; however, several reactors is described by Studer et al., who developed
of the rather new methods are largely efficient fractiona- a high-throughput reaction system [42, 43], where a 96
tion methods, which makes it difficult to put a label that well-plate unit was employed for both pretreatment and
clearly defines “pretreatment”. enzymatic hydrolysis. In the system, a 1% solids suspen-
The research area concerning biomass pretreatment is sion of a genotype of Populus trichocarpa was tested with
very active. A database search using the keywords “bio- an excess of cellulolytic enzymes to speed up the assess-
mass AND pretreatment” in Scopus for the period 1973– ment procedure. Costa et al. [44, 45] report in studies
2019 results in about 9600 publications, with a suddenly concerning sugar cane hybrids, a correlation between
increasing number of publications starting around 2010. microscopic characteristics and chemical composition
At this time, the number went from about 300 to pres- regarding recalcitrance of the sugar cane. Rapid, simple,
ently more than 1000 publications per year having the and efficient screening protocols will be more and more
aforementioned keywords. One of the reasons for this important when selection of modified plants for biorefin-
may be a more efficient classification system, where the ery purposes is based on certain properties, such as easily
authors’ select proper keywords, but it also shows the accessible carbohydrates for further processing. In addi-
continuous high interest for fractionation of various lig- tion, the rapidly growing interest for lignin also shows the
nocellulosic materials. It is thus an overwhelming task requirement for procedures to validate the quality and
to cover every aspect of biomass pretreatment in all its reactivity of the fractionated lignin.
forms. Therefore, this review is a selection of currently It has to be stressed that although rapid screening
“hot” methods, as well as a historical perspective of methods are of great value, the final testing must be car-
methods that still attracts much attention. ried out at higher solid content than 1–2%. The impact
on the overall process costs is very much influenced by
How can different pretreatment methods be assessed? the achieved concentrations of sugar or other products.
One of the challenges in selecting a pretreatment method A low concentration adds to the separation and recov-
is to evaluate the effect of different methods. It is in gen- ery costs such that they may become far too high. It is
eral not enough to calculate recovery and yields of the especially important to perform high-solids enzymatic
main compounds in the lignocellulosic raw material. hydrolysis since the conversion of the lignocellulosic
These figures are very important, especially when consid- materials may become much smaller than expected from
ering the economics of a process. However, they do not low-solids experiments [46].
fully tell if a pretreatment method is suitable for a cer-
tain post-processing step. For instance, if the purpose of Modification of lignocellulosic species
pretreatment is to produce ethanol, then several indica- Modification of the plant materials on a genetic level
tors can be used to assess the pretreated material. These is starting to increase in importance. An example is a
indicators include estimation of the levels of inhibitor study by Holwerda et al. [47], who compared transgenic
or toxic compounds, such as aldehydes, organic acid, switchgrass plant lines with their non-transgenic con-
ketones, phenolic compounds, etc., which may have a trols. The results show that it is possible to achieve high
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 7 of 26

total-carbohydrate solubilisation, which is dependent where an organic solvent, an ionic liquid and similar
on a number of factors, such as feedstock modification, dissolving agents are utilised for pretreatment (unless a
feedstock choice, biocatalyst choice and augmentation catalyst is added that affects the pH). The pH has a very
of biological attack (i.e. pretreatment). The latter was large influence on the outcome of a pretreatment pro-
performed using either ball milling or CELF (Cosolvent- cess. These effects can be described roughly according to
Enhanced Lignocellulosic Fractionation, a method that the following:
involves treatment with aqueous tetrahydrofuran and
dilute acid at elevated temperatures). Shen et al. [48] • Low pH, where the desired result is hydrolysis of the
modified genetically switchgrass to reduce lignin content, hemicelluloses part to monomeric sugars, without
resulting in less formation of phenolic degradation prod- formation of inhibiting or toxic compounds, while at
ucts, while maintaining carbohydrate levels. In another the same time keeping the cellulose polymer intact.
study, Mansfield et al. [12] subjected poplar trees with • Neutral or close to neutral pH, which result in partial
altered lignin content to steam explosion or organosolv hydrolysis of the hemicelluloses caused by organic
pretreatment (using aqueous ethanol and sulphuric acid acids in the lignocellulosic materials (autohydrolysis).
as catalyst). Their conclusion was that the low-lignin Since the conditions are not severe enough, most of
poplar trees showed a 15% improvement in the conver- the hemicelluloses will stay in oligomeric or poly-
sion efficiency resulting in near-complete hydrolysis of meric form. Lignocellulosic feedstocks that are low
the poplar carbohydrates. It has to be mentioned that in organic acids, such as softwood, will not result in
enzymatic hydrolysis was performed at 2% dry-matter high solubilisation of hemicelluloses.
content, which as discussed earlier, is a non-realistic • High pH, which can result in dissolution of the lignin
dry-matter concentration in full-scale operation. It is, fraction (such as in pulping of wood, or when organic
however, a common method to assess various pretreat- solvents are employed) while some or most of the
ment methods. Wilkerson et al. [49] carried out a study hemicelluloses still is found in its solid state.
on poplar where they introduced ester linkages in the
lignin structure to improve cell wall digestibility after Acidic conditions result in hemicellulose hydrolysis,
a mild alkaline pretreatment. The concept of tailoring which also commonly results in formation of degradation
plants to use cell-wall biosynthesis to modify the natu- products, such as furfural, HMF, levulinic acid, etc. Car-
rally occurring variety may be an option to grow plants valheiro et al. [54] have summarised several pretreatment
that are more easily deconstructed for biorefinery pur- methods where pH is the main cause for the outcome of
poses, according to Wilkerson et al. Another study on the procedure. It is obvious how the distribution of the
genetically modified poplar was carried out by Biswal main polymers in a lignocellulosic material changes as
et al., who showed that it was possible to reduce recalci- the pH moves from low to high values. At low pH, almost
trance, have more easily extractable cell walls as well as a all hemicelluloses are removed from the solid material; in
more rapid growth in the selected poplar species (Popu- contrast, at high pH, lignin is dissolved and cellulose and
lus deltoides) [50]. In addition, the xylan and the pectin hemicelluloses constitute the solid fraction (Fig. 2).
contents were reduced. The classification can also be based on the expected
fractionation result. Many pretreatment methods today
Pretreatment and fractionation of lignocellulosic materials are based on solvents (or regarding their action, solvent-
The traditional manner of discussing pretreatment meth- like chemicals), yielding results, which to some extent
ods is often based on a rather loose classification. In gen- resembles the outcome from high-pH pretreatment
eral, the pretreatment methods are regarded as belonging methods, i.e. a solid phase containing relatively higher
to one of the following categories [51–53]: physical (e.g. parts of hemicelluloses and cellulose, while most of the
milling, grinding, irradiation, sonication); chemical (e.g. lignin is dissolved. It is also in this research area where a
alkali, acid, oxidising agents, organic solvents, ionic liq- multitude of combinations of chemicals is under intense
uids and deep-eutectic solvents); physico-chemical scrutiny, both regarding their suitability for pretreatment
methods (e.g. steam pretreatment w/wo catalyst, wet-oxi- and fractionation of lignocellulosic materials, but also
dation and hydrothermolysis), or biological methods. In regarding concerns for health and process safety. How-
reality, many variations may be made up of combinations ever, the somewhat older established processes are also
of two or more methods. under constant development, often in combination with
Another classification can be made according to the pH studies on lignocellulosic waste materials from all kinds
of the selected pretreatment method, if pretreatment is of agricultural, woody or herbaceous plants. Much of
performed under acidic, neutral or alkaline conditions. these residues are currently of no or low value, but still
This classification is, of course, not valid for methods represent a considerable amount of biomass that could be
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 8 of 26

a b
Hemicellulose
Hemicellulose
Lignin
Lignin

Cellulose

Cellulose

Fig. 2 The effect of pH on pretreatment of lignocellulosic materials: a low pH; b high pH

used for production of value-added chemicals and fuels for production of high-quality journal paper. Depend-
without too complicated process designs. A comment ing on the severity of the pretreatment, the extent to
about their low value is in place here: as soon as a waste which the solid material dissolves can vary to a high
can be utilised, the material cost is likely to increase, degree. If a high acid concentration is employed at high
which is one of the reasons why it is difficult to make eco- temperatures, not only hemicelluloses but also cel-
nomic estimations for a full-scale production facility. lulose converts to oligo- or mono-saccharides. Under
In Table 2, a selection of commonly utilised pretreat- even more severe acidic conditions the carbohydrates
ment methods is summarised. As already pointed out, degrade to other compounds, e.g. furfural, HMF, lev-
the vast number of pretreatment-related publications ulinic acid, etc. Typically, temperatures ranging from
makes it difficult to discuss in detail every pretreatment 140 to 200 °C and a residence time of minutes to hours
method. This review summarises frequently applied pre- are common. Alkaline conditions often have the great-
treatment methods, which can be applied in a biorefinery est effect on agricultural residues or herbaceous crops,
facility. as these materials in general contains less lignin [55, 56].
An excellent techno-economic analysis of six pretreat-
Pretreatment methods ment technologies (ammonia fibre expansion, dilute
Acid and alkaline methods acid, lime, liquid hot water, soaking in aqueous ammo-
In this section, acid and alkaline methods based on the nia and ­SO2-impregnated STEX) was published 2011, in
hydrolytic action at low or high pH are discussed. At the CAFI project (Consortium for Applied Fundamen-
low pH, the hemicelluloses fraction hydrolyses, while tals and Innovation) [22], where switchgrass was used
the cellulose and lignin fractions in general are less for evaluation of the processes. Interestingly, the result
affected. At higher pH, it is instead lignin that is solubi- shows that the differences between the technologies were
lised, which forms the basis for many pulping processes limited. However, switchgrass is not very recalcitrant in

Table 2 A selection of commonly utilised pretreatment methods

Method Active agent Mode of action

Dilute-acid pretreatment H2SO4, ­H3PO4 and other strong acids Hydrolysis of hemicelluloses
Alkali pretreatment NaOH, lime, ­Na2CO3 and similar alkaline compounds Extraction of lignin
Steam pretreatment/steam explosion High-temperature steam; catalyst may be added Hydrolysis of hemicelluloses, fibre separation
Ionic liquids Large organic cation and a small inorganic anion Fractionation of polymers
Deep-eutectic solvents Mixtures of Lewis and Brønsted acids and bases Fractionation of polymers
Organosolv Organic solvents, e.g. ethanol, butanol. Catalyst can be Extraction of lignin
added
Milling/grinding Particle size reduction Surface increase and improved access
Biological treatment Degradation of the material Brown-rot degrades hemicelluloses and cellulose
White-rot break down lignin
Soft-rot breaks down cellulose
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 9 of 26

contrast to softwood, which may have exhibited larger were conducted for 4 h to secure steady-state conditions.
variation of the estimated costs. The pH of the slurry was then adjusted to 5.2, after which
Dahunsi [57] studied the effect of acid or alkaline enzymatic hydrolysis was carried out. Pretreatment using
hydrogen peroxide pretreatment on pineapple peel prior 0.35 wt% acid, 162 °C and 0.65 wt% acid at 152 °C yielded
to anaerobic digestion. In the study, based on response almost the same result in terms of sugar recovery. Over-
surface methodology, they varied the temperature all, about 65% of the potentially available carbohydrates
between 75 and 115 °C, the residence time between 6 and could be recovered. It was also pointed out that a higher
46 min, the H­ 2SO4 concentration was 0–2% (v/v). A simi- recovery can be achieved by adding more enzymes; how-
lar procedure was employed using alkaline H ­ 2O2. The ever, there is as trade-off between the efficacy of enzy-
main results from the study were that alkali-enhanced matic hydrolysis and enzyme dosage since enzymes are
pretreatment yielded a lignin reduction and a biogas pro- expensive.
duction, which was 67% higher than the corresponding The SPORL method (sulphite pretreatment to over-
acid treatment. Another observation was that the resi- come recalcitrance of lignocellulosic) is a technology
dence time in the biogas reactor was shorter for the alka- utilising sulphite in the pretreatment step. This pretreat-
line-treated pineapple peel. In another study by Harun ment technology allows lignocellulosic materials to react
and Goek [58], NaOH was used to carry out pretreat- with a solution of a sulphite salt (with e.g. Na, Mg, or
ment on rice straw. The straw was pretreated at 55 °C for Ca) at a temperature of 160–180 °C, and a pH of 2–4
1 and 3 h, respectively. A concentration of 2–12% (w/v) for about 30 min [61]. After the sulphite cooking, a disk
of NaOH and a ratio of rice straw to NaOH solution of milling operation takes place to generate a fibrous mate-
1:20 (w/v) was applied to all samples. One of the results rial for subsequent saccharification (and fermentation).
was that an increase of the glucan content by 85.6% (rela- When spruce wood was subjected to SPORL treatment,
tive to untreated straw) was achieved using 12% NaOH the overall yield after the saccharification step was found
for 1 h. This condition also resulted in the highest delig- to be 87.9% for glucose and xylose. During the pro-
nification of the straw, 79.6%. On the other hand, straw cess, hemicelluloses were dissolved, and about 32% of
treated for 3 h with 2% NaOH was found to give the high- the lignin fraction was dissolved and sulphonated. The
est total carbohydrate content, 79.2%. SPORL process was compared with a dilute-acid pre-
An example of the current biorefinery trend, where val- treatment performed at the same temperature and resi-
uable lignocellulosic constituents are utilised for produc- dence time, which resulted in about 57% recovery of the
tion of value-added compounds, is the integrated process sugars.
for coproduction of fermentable sugars and lignin from Another technology that has been studied in pilot-scale
poplar pellets that was evaluated by Chu et al. [59]. The is the ammonia fibre-expansion method (AFEX). Sarks
process was based on alkali-sulphite pretreatment for et al. [62] investigated in two 450 L packed-bed reac-
production of lignin adsorbents and sugars. In the pro- tors pretreatment of corn stover followed by enzymatic
cess, ­Na2SO3 was used to soak poplar wood pellets prior hydrolysis and fermentation. The water content in the
to a two-stage cook, the first at 130 °C for 30 min, while corn stover was adjusted to 25% before being heated to
the second was carried out at 210° for 5 min. After pre- 80°. Then, biomass was treated with gaseous ammonia at
treatment the liquor and the solid residue was separated. a ratio of ammonia to biomass of 1:1 for 30 min. One of
After washing, enzymatic hydrolysis on the solid resi- the interesting results is the increase in yield relative to
due was performed. The liquid fraction was processed that of laboratory-scale experiments (10 L). The glucose
further using nitric acid to yield an adsorbent prod- and the xylose release after 48H increased by 19 and 15%,
uct. Over 75% of the original carbohydrate content was respectively, in the pilot-scale experiments. In addition,
released and potentially could be used for fermentation. ethanol production showed a 15% increase compared to
The adsorbent was tested using lead ions, which resulted laboratory-scale trials. The reasons for the higher conver-
in adsorption capacities of 156 and 68 mg g−1, for lignin sion is not fully clear. However, improvements were seen
from pre-hydrolysate and enzymatic hydrolysis residue, already earlier when the transition from a stirred batch
respectively. The probable reason for the heavy-metal ion reactor to the laboratory-scale equipment was made. The
adsorption capacity was suggested to be incorporation of improved conversion has a direct effect on the minimum
sulphur-containing groups in the lignocellulosic material. ethanol-selling price, which was estimated to be about
Dilute-acid pretreatment of rice straw in pilot-scale was 19% lower.
carried out by Kapoor et al. [60]. A continuous pilot-scale Biomass particle size is reported to have an effect on
reactor was utilised to pretreat rice straw at a feeding rate the distribution of the constituents of the biomass. In a
of 10 kg h−1 at temperatures ranging from 150 to 166 °C study on switchgrass, Bridgeman et al. [63] found that
with a fixed residence time of 10 min. All experiments smaller particles have a significantly larger content of
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 10 of 26

inorganic matter than larger particles. In contrast, the countries. The authors point out the importance of test-
larger particles had a relatively higher content of carbo- ing the pretreated straw regarding the health effects on
hydrates, but were lower in nitrogen content. The influ- the animals and, in addition, if the milk and meat are
ence of particle size on AFEX-pretreated rice straw was safe for human consumption. Bals et al. [65] carried out
studied by Harun et al. [64] who pretreated milled (2 and a study on the concentration of acetamide (an ammo-
5 mm) and cut (2 and 5 cm) particles. In addition, two niation by-product) in AFEX-treated crop residues and
levels of severity were employed for each particle size. in milk and meat from cattle and buffalo that had been
The less severe experiments were performed at 100 °C for fed with the material. When the AFEX-treated mate-
30 min at a straw: ammonia ratio of 2:, while another set rial was introduced, the levels of acetamide in the blood
of experiments were conducted using more higher tem- increased. It is possible that the levels will decrease over
perature (140 °C), longer residence time (50 min), and a time due to an adaptation by the animal; however, con-
ratio of 1:1 of ammonia to straw. Enzymatic hydrolysis of tinued research is necessary to assess the risk, if any, for
the AFEX-pretreated fractions showed an impact of par- ruminants and humans.
ticle size, in that larger particles resulted in lower sugar
conversion for the less severe conditions. In contrast, the Steam explosion (pretreatment) and extrusion methods
more severe conditions yielded a higher conversion with Steam explosion (or steam pretreatment) is one of the
increasing particle size. The largest, cut particles signifi- most widely investigated methods and has been so for a
cantly demonstrated a higher conversion than did smaller long time. In Table 3, a selection of results from steam
particles at the highest severity. pretreatment using typically investigated materials is
The AFEX technology has also been employed for presented. STEX has been used for many purposes on a
production of livestock feed. The technology was used large range of lignocellulosic materials, and in combina-
to upgrade wheat straw in an attempt to reach fodder- tion with other pretreatment procedures. This includes
level quality. The true IVOMD (in vitro organic mat- studies on liquid and gaseous fuels production, e.g.
ter digestibility), which is a measure for fodder quality, ethanol and butanol, or biomethane through anaerobic
improved on average by 193 g kg−1 to reach an average digestion, respectively. The variation in lignocellulosic
of 844 g kg−1, which can make a large difference in many materials treated by STEX is very large. They range from

Table 3 Selected steam pretreatment results using various lignocellulosic materials

Biomass Catalyst/procedure T, °C t, min Main product(s) Yield(s) Refs.

Alpine hay No catalyst 160–220 5–15 Glucose, biogas Glucose: 90% [67]
Biogas: 469 L­ N
Poplar wood Mechanical refining + STEX 210 5 Carbohydrates Carbohydrates: 81% [68]
Neutral sulphonation Sulphonated lignin
Hybrid poplar 4.5% (w/w) S­ O2 195 4.5 Ethanol prod. for evaluation 15% improvement in etha- [12]
of modified lignin content nol for low-lignin content
breeds
Sugar cane bagasse Pre-soaking in reverse osmo- 185–215 10–15 16 g/100 g DM (whole Max combined carbohy- [69]
sis water slurry) drate yield: 65%
Corn cob 0.5% ­H2SO4 (12 h) 180 10 Lignin 57.3% (purity 99%) [70]
Wheat straw 1% acetic acid 190 10 Ethanol 0.32 g/g (of glucose and [71]
xylose)
Agave bagasse No catalyst 142–179 2.8–22 Biogas BM: 0.290 ­LN ­gCOD −1 [72]
Rice straw Moisture content 0–70% 160–205 1–10 Biogas Increase in production rates [73]
up 2.4 time untreated
straw
Switchgrass Hydrated with water to 30% 170–200 5–15 Glucose After EH: 88.3% [74]
DM
Sesame seed Soaked in water (12 h) 212 (2 MPa) 10 s Lipid extraction 45% (treated) [75]
180 (1 MPa) 30 s 38% (untreated)
Barley straw No catalyst 180 30 Ethanol 50 g L−1 [76]
(99% cellulose recovery; 82%
hemicelluloses after STEX)
Wheat straw/corn stover Soaking 0.2% ­H2SO4 190 5 Ethanol > 50 g L−1 [77]
(mixed) Overall yield 74–78%
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 11 of 26

food-industry waste to forest and agricultural residues, to temperatures between 115 and 160 °C for 10 min. The
but also include wood chips from high-quality wood. It is study aimed to study the effect of the process conditions
obvious that STEX still is of high interest for biorefinery on hydrolysis, for kraft cooking of spruce chips, and for
applications, based on the large number of publications, alkali extraction experiments. Their conclusion was that
both past and present. mild STEX makes the wood structure more accessible for
The term “steam explosion” (STEX) is originally based enzymes and for pulping. Thus, the implementation of
on the notion that steam, under high pressure inside the mild STEX prior to further treatment can be beneficial
cell structure, is rapidly released and expanded when as a first step in biorefinery concepts, where further pro-
the reactor vessel is de-pressurised. This causes disrup- cessing by enzymes are planned. In addition, traditional
tion of the cell structure similar to that of an explosive pulping can be improved by such a treatment, since the
action. The term “steam explosion” is somewhat mislead- distribution profile of cooking chemicals inside the chips
ing since an explosion in the physical sense does not take appear to be more even. In the study, one of the conclu-
place. However, since there is a loud bang when the mate- sions was that at lower temperatures, the mechanical
rial is suddenly de-pressurised, and the material can be effect of STEX was of importance.
brought open to show individual fibres, the term is still The size of the chips or straw pieces is likely to have an
widely used. The background is the Madison process for effect on the overall enzymatic sugar release. Standard
production of fibreboard, which was described already pulp chips are rather well defined, about 30 to 50 mm in
in 1930 [66]. In the process, wood chips are subjected to the fibre direction and 5–10 mm in the cross-fibre direc-
steam at around 265 °C for a short time, and then pressed tion. Further size reduction can consume large amounts
to a fibreboard material. of energy; thus, it is preferable to avoid additional size-
It is not completely clear if the explosion itself is nec- reducing steps. DeMartini et al. [81] steam-pretreated
essary to yield an easily hydrolysable material. Brownell aspen (Populus tremuloides) using a standardised wood-
et al. [78] investigated the mechanical effect of the pres- chip size at 180 °C for 4–18 min. The resulting pretreated
sure drop by comparing two sets of experiment: one with materials were subjected to enzymatic hydrolysis for a
a rapid pressure release, and another where 80% of the residence time ranging from 24 to 168 h. They experi-
steam was vented off prior to complete de-pressurisation. ments were evaluated by MRI, SEM and staining tech-
The result was that there was no significant difference in niques. The study shows that it is possible to have a
the yields of glucose whether the release was from full uniform pretreatment also for wood chips of industrial
steam pressure, or from a reduced pressure. size. Their conclusion was that the rapid decompression
Muzamal et al. [79] evaluated three effects of STEX achieved by steam explosion might alter accessibility at
pretreatment on Norway spruce wood in specially lower temperature conditions.
designed STEX experiments at temperatures ranging Vidal et al. [82] reviewed the effect of particle size for
from 165 to 195 °C: various pretreatment methods. Particle size may have a
direct effect on enzymatic hydrolysis. The particle size is,
• The influence of STEX time and pressure, however, less important than the removal of, e.g. hemi-
• The effect of pressure-release rate (slow or rapid), celluloses and lignin. By employing physical size reduc-
• The effect caused by shearing between the chips, and tion, conversion yields of about 50% can be achieved.
the material impact with the vessel walls. Thermochemical pretreatment methods can handle par-
ticle sizes to a varying degree. The maximal size range
Their study showed that STEX pretreatment clearly that a certain pretreatment method can handle is stated
increases the porosity of the wood fibres. The explosion to be decreasing as follows: STEX > LHW > dilute acid
step alone does not disintegrate the chips. The impact and alkali. The feedstock also appear to influence the out-
of highly softened wood chips grinding other chips, and come of EH; thus, herbaceous or grassy biomass exhib-
effect from the impact against the vessel walls were more its a lower maximal size (< 3 mm) than woody biomass.
important in this regard. STEX enlarges pores in wood Another aspect is how the lignin structure is affected by
through chemical changes during steam treatment, by the STEX process. Wang et al. [83] used several analytical
opening the wood structure with the expansion of vapour methods such as SEC, FTIR and NMR(HSQC) to study
inside the tracheid. In summary, all three steps were the effect of STEX on poplar wood, which was treated at
considered to have a synergistic effect that increases the 209 °C for 7 min. Samples from native poplar wood, as
effectiveness of STEX pretreatment. well as from the STEX-treated solid and liquid fractions
Similar studies were performed by Jedvert et al. [80] were evaluated. The results show that STEX pretreat-
who utilised mild STEX for production of pulp from Nor- ment reduces the amounts of β-O-4′, β-β, and spirodi-
way spruce (Picea abies). The spruce wood was subjected enone structure, and increases the syringyl/guiacyl ratio
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 12 of 26

from 1.14 (native poplar) to 1.70 (solid) and 1.86 (liquid), to the continuous steam-pretreatment method discussed
respectively. In addition, SEC results showed that de- and earlier. The motion and the mixing of the lignocellulosic
re-polymerisation are the main reasons for the increase materials improves heat and mass transfer in the reactor.
in lignin average molecular weight in the solid fraction, However, the extrusion process results in a much more
while it decreases in the liquid fraction. An important intense abrasion of the material.
conclusion is that although the lignin structure changed, Steam explosion has been used not only as a prepara-
the backbone structure was not modified to a large tive step prior to enzymatic hydrolysis for further pro-
degree. cessing for the sugar-platform concept, but also for
Thus, although STEX has been studied for many years, enhancing biogas production [67]. In a study performed
the conclusion from these studies shows that the over- by Bauer et al., hay was steam pretreated at temperatures
all effect of STEX is still not clear. However, the higher ranging from 160 to 220 °C in intervals of 15 °C. The
the temperature, the more important the chemical effect residence times were 5, 10 or 15 min. Subsequently, the
becomes, i.e. the reactions caused by acids and other steam-pretreated materials were enzymatically hydro-
degradation reactions inside the lignocellulosic mate- lysed. In comparison with untreated hay, the methane
rial. The approach to use continuous steam-pretreatment yield increased by about 16%.
equipment where the lignocellulosic material is trans-
ported through the reactor is also a means to increase Hydrothermal methods
the reaction rate in comparison with non-stirred batch Hydrothermolysis is a method that uses liquid water
equipment. In a continuous reactor, the mixing proper- under pressure at high temperatures. The method is
ties are improved largely. Mixing makes it possible to also known as hydrothermal treatment, autohydrolysis,
have less severe conditions during the pretreatment step, pressure-cooking in water, etc. It has mostly been used
which is beneficial since non-desired side reactions, such in studies regarding hardwood lignocellulosic materials
as furfural formation, is promoted by higher tempera- [87–89] and annual plants [90, 91]. The mechanism of
tures. Cornejo et al. [84] used thermochemical pretreat- hydrothermolysis is believed to be working by auto-ion-
ment (steam pretreatment) equipment to pretreat poplar, isation of water, causing acetyl groups in hemicelluloses
wheat straw and pine employing sulphuric acid as a cat- to form acetic acid. The dissociated acetic acid catalyses,
alyst (1–3% for wheat straw; 0–4¤ for pine and poplar). e.g. carbohydrate degradation and dissolution of hemi-
The selected temperatures were between 164 and 192 °C celluloses. Hydrothermolysis has also been evaluated for
for poplar and pine chips, and ranging from 173 to 187 °C extraction of lignin and carbohydrates from pine. Ståhl
for wheat straw. The residence times were between 5 and et al. [31] treated pine chips at temperatures ranging from
15 min. The purpose was to solubilise the hemicelluloses 200–240 °C having a liquid-to-wood ratio of 40:1. Within
fraction for transformation to furfural, and for produc- 10 min, approximately 35% of the lignin was solubilised
tion of glucose. Very high conversions of cellulose to during hydrothermolysis at 240 °C. This amount of lignin
glucose were achieved after enzymatic hydrolysis: poplar corresponds to about 97 kg per ton of dry wood. If the
biomass resulted in the highest yield (40 g/100 g DM), reaction was extended, increased condensation of lignin
while pine only yielded 25 g/100 g DM) over the enzy- took place. Two fractions of lignin were recovered in the
matic hydrolysis step. The results are in line with previ- liquid phase, which were found to be reactive; in addi-
ous results where pine has been shown to be difficult to tion, the recovered lignin was sulphur-free, which can
hydrolyse efficiently [85]. be an advantage for use in other applications. The hemi-
Del Carmen Fong Lopez et al. [86] studied a continuous celluloses fraction was found to be completely liquefied
process combining an alkaline pretreatment and neutrali- already at 200 °C, where also about 12.5% of the cellulose
sation in a twin-screw extruder. In the study, dehydrated fraction was dissolved. Ståhl et al. also propose a kinetic
sweet corn co-products (SCC) were exposed to NaOH in model for degradation of xylan, mannan and glucan. The
concentrations between 3.9 and 8.0% w/w and tempera- model has a number of constants [18] that were fit to the
tures ranging from 52 to 168 °C. The accessibility of the suggested rate equations. Lignin in the solid residue can
solid extrudate was assessed by enzymatic hydrolysis at be calculated from analytical expressions.
5% DM. At optimal conditions, a resulting cellulose con- The hydrothermolysis technology has also been used
version of about 70% was achieved. The authors state for fractionation of agricultural materials such as rape-
several phenomena that may be involved in the extru- seed flour into, e.g. amino acids, carbohydrates, fatty
sion process. These include structural biomass modi- acids and organic acids [92]. The treatment was carried
fication. In addition, the sugar yield could be improved out with subcritical water at temperatures between 180
by a higher temperature, while decreasing the amount of and 280 °C for residence times up to 60 min. The high-
catalyst. To some extent, extrusion of biomass is similar est yield of amino acids was obtained at 200 °C and a
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 13 of 26

residence time of 60 min (124 g per kg rapeseed flour). cleavage. In addition, some of the hemicelluloses are dis-
However, carbohydrates were obtained at different con- solved and degraded to smaller compounds [105].
ditions: xylose and glucose reached maximum concentra- Ethanol–water mixtures are still one of the most com-
tions of 51 g kg−1 (220 °C, 20 min) and 51 g kg−1 (260°, mon ways to perform organosolv pretreatment. A catalyst
10 min), respectively. A similar study was performed on is often added, which enhances the effect of the solvent.
pectin from citrus and apple pectin, with the purpose of Bouxin et al. [95] examined the effect of heated acidi-
producing uronic acids [93]. Uronic acids are valuable fied aqueous ethanol treatment on Sitka spruce sawdust
chemicals that are used in several application, e.g. in the [95]. Sulphuric acid having concentrations between 0.75
cosmetic industry. The highest yield was 79.7 g kg−1 at a and 1.25% (w/w) was used as catalyst. The range of pre-
temperature of 150 °C and a residence time of 40 min. treatment temperatures was 150 °C to 180 °C followed by
A different application than most other hydrother- enzymatic hydrolysis. The solid organosolv residue was
molysis investigations was studied by Chu et al. [94]. subjected to a saccharification step. A decrease in lignin
They explored the direct conversion of a sediment from content from 35 to 22% did not significantly have an
the conventional kraft-pulping process (black liquor acid impact on the enzymatic hydrolysis. The highest sacchar-
sediment) into phenolics. Depolymerisation of black liq- ification yield was obtained at a residence time of 60 min
uor lignin is a means to produce intermediates that can at a temperature of 180 °C, with an ethanol concentration
be processed further to yield high-value chemicals. In the of 60% and 1% ­H2SO4. Under these conditions, a sacchar-
study, the effects of temperature and residence time were ification yield of 86% was reached. A reduction of pen-
investigated. The temperature was varied between 260– tose degradation to furfural was also noticed. A large part
340 °C, and the maximum residence time was 120 min. of the hemicelluloses sugars was also converted to ethyl
The yields of four products—oil phase, aqueous phase, glycosides, which is a valuable product as an intermediate
char product and gas phase—were evaluated. Their con- in the sustainable production of value-added chemicals.
clusion was that direct hydrothermal conversion of the Matsakas et al. [106] used a modified steam-explosion
black liquor is a promising method. The selectivity for reactor for operation with ethanol instead of steam. They
phenolic compounds, especially catechol was high. pretreated spruce at a temperature of 200 °C, varying the
residence time between 15 and 60 min, the ethanol con-
tent between 52 and 65% (v/v), and addition of sulphuric
Organic solvents/organosolv/green solvents acid (0–1% w/w). The pretreated solid material was pro-
Pretreatment aided by an organic solvent is known as cessed using high-gravity SSF at 22% solid content. The
organosolv. One of the important features of organosolv highest resulting ethanol concentration was 61.7 g L−1,
methods is the ability to cause biomass dissolution. The with a total yield of 68.6% of the theoretical maximum.
organosolv pretreatment method employs organic sol- This is in the same range as reported by Hoyer [107] who
vents of widely different kinds, such as ethanol [95–97], reached 65 g ­L−1 and a yield of 72.1% employing STEX.
acetone [97], organic acids (e.g. formic and acetic acid) Recovery of the solvent is of high importance to enable
[96], ethylene glycol [98]. More recently, liquids such as a cost-efficient process, where make-up chemicals should
γ-valerolactone [99, 100], methyl isobutyl ketone (MIBK) be kept at a minimum. Lê et al. [108] processed Eucalyp-
[101], tetrahydrofuran (THF) mixed with water [102] and tus globulus wood chips using GVL/water fractionation
2-methyltetrahydrofuran [103] have been investigated. with different liquid-to-solids ratios ranging from 2 L
An excellent review by Zhang et al. [96] on numerous ­kg−1 to 10 L ­kg−1 at a reaction temperature of 180 °C and
solvents that have been experimentally tested for orga- a residence time of 150 min. The GVL content in the liq-
nosolv, discusses in detail pretreatment using organo- uor was 50%. Lignin was recovered from the spent cook-
solv. In the review, the pros and cons of many solvents ing liquor by the addition of water, and the precipitated
are reviewed in terms of physical properties (e.g. boiling lignin could be collected by spinning in a centrifuge. The
points, solvent polarity, their rating as a green solvent, recovery of GVL was performed utilising several separa-
etc.); in addition, a thorough survey of the effect of the tion methods, e.g. distillation at reduced pressure, lignin
organic solvent on saccharification using cellulolytic precipitation by water addition, and liquid ­CO2 extrac-
enzymes is presented. Organosolv does—as does alkali— tion. The combination of precipitation and distillation at
remove lignin in rather large amounts. Alkali mainly reduced pressure resulted in 90% recovery of GVL and
works by disrupting the bonds between lignin and hemi- the formation of a sticky residue, a mixture of lignin and
celluloses to cleave off ether and ester bonds between GVL. Distillation at reduced pressure is a rather com-
carbohydrate and lignin. It also breaks C–C bonds in plex unit operation, which increases the risk for process
lignin by hydrolysis [104]. Organosolv acts by cleaving stability should leaks occur in the distillation unit. How-
β-aryl-ether bonds either by acidolysis and (or) homolytic ever, it is frequently used in the petroleum refineries for
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 14 of 26

the heavy oil fractions. Using liquid C

­ O2, 87% of the GVL aqueous mixture to an estimated optimum of 92–96 wt%
could be recovered in the extract, and the rest was found GVL. If water is added lignin precipitates, which opens a
in the raffinate. Six recovery schemes were proposed of way for separation and recovery of both lignin and GVL
which two were considered to become feasible given fur- [111].
ther optimisation. It was also pointed out that the pro- A combination of two technologies, organosolv and
cess for water recycling of the washing water must be dilute-acid pretreatment was performed by Chin et al.
investigated. [98] who used ethylene glycol as a first step in pretreat-
GVL was utilised in a study by Alonso et al. [99]. The ing empty fruit bunch at 85° for 45 min with 50% ethylene
solvent has shown high applicability for fractionation of glycol in the presence of 3% NaOH. The resulting solid
lignocellulosic biomass. By adding an acid to serve as a fraction was then subjected to a two-stage dilute-acid
catalyst, a mixture of GVL, water and acid can dissolve hydrolysis, where stage 1 was performed with 36–90
the hemicelluloses and the lignin parts, resulting in a wt% ­H2SO4 at 80 °C for 45 min. Stage 2 employed dilute
high-purity cellulose residue. The cellulose can be pro- acid (3–15%) at a temperature of 100 °C for 60 min. The
cessed further to produce dissolving pulp for textile-fibre highest release of sugars was about 90%, while the for-
production. A potential advantage of GVL is the manu- mation of degradation products was rather low. The best
facturing process of the compound itself. GVL is pro- conditions was found to be an acid concentration of 45%
duced from levulinic acid, which is a degradation product in stage 1 at 65 °C for 30 min, followed by a second step
from hexoses; thus, through hydrolysis of part of the using 12% acid at 100 °C for 120 min.
available cellulose into glucose, hydroxymethylfurfural A selection of various biomass materials, which have
can be dehydrated and yield formic acid and levulinic been pretreated utilising different organosolv-like pre-
acid, which is the precursor for GVL [109]. From a safety treatment methods, is presented in Table 4. The number
perspective, GVL has physical properties that makes of organic solvents, which have been utilised in the orga-
it suitable for storage and transportation, e.g. a low nosolv process, is very large; therefore, only some of the
vapour pressure (3.5 kPa at 80°) [110]. The solubility of typical solvents, as well as some more recently investi-
lignin increases with increasing GVL concentration in an gated solvents are included.

Table 4 A selection of organosolv methods utilised for pretreatment of lignocellulosic materials

Biomass Solvent Glucan Xylan Lignin Comment Refs.
recovery removal removal
(%) (%) (%)

Hybrid Pennisetum Water 91.1 36.8 11.6 [97]

Acetone 90.4 47.8 32.8
Ethanol 93.4 53.5 37.6
THFA 92.3 46.8 46.8
GVL 90.9 43.4 50.3
Coir (coconut fibre) 1,4-Butanediol/acidic IL 87.5–90 77–93 75–88 Combination of organosolv and ionic liquids [112]
Eucalyptus Ethanol 1. Glucan content (solids): 75.6% [113]
1. Bark Oxalic acid 2. 50%
2. Wood Water
1. SCB Butanol 81 87 Glucose yield (%) after EH: [114]
2. Tall fescue 80 87 1. 77
3. Sugar beet 65 85 2. 74
4. Eucalyptus 80 80 3. 69
5. Beech 81 72 4. 65
6. Japanese cedar 79 12 5. 65
6. 1
Eucalyptus Ethanol 74–93 15–70 25–52 [115]
Bark
Rice husks Ethanol 88–90 86.8 77.5 [116]
Empty fruit bunch (palm tree) Ethylene glycol (3% NaOH) 90.6 10–54 67.2 Starting material was decomposed fruit bunch [117]
Corn stover Methanol/NaOH 97.5 16.5 37.3 Enzyme hydrolysis: [118]
Glucan: 97.2%
Xylan: 80.3%
Bamboo Formic acid 98 22 83 [119]
SCB Autohydrolysis + glycerol 80–90 55–68 48–84 [120]
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 15 of 26

Ionic liquids Table 5 A selection of cations commonly used for ionic

Efficient pretreatment and fractionation methods for liquids
production of value-added products heavily relies on the Acronym Cation Refs.
availability of efficient and not too expensive processes. It
is also imperative that the products can be recovered at a [Hmim] 1-Methylimidazolium [132]
high yield using common separation and recovery equip- [C4-py] 1-Butyl-pyridinium [133, 134]
ment. The potential candidates for commercialisation [Bmim] 1-Butyl-3-methylimiazolium [127, 132]
must be safe from health and environmental perspec- [Emim] 1-Ethyl-3-methylimidazolium [127, 134]
tives. A particular interesting group of chemicals is the [Mmim] 1,3-Dimethylimidazolium [135]
ionic liquids. Ionic liquids are salts that consist of an inor- [Pdmim] 1-Propyl-2,3-dimethylimidazolium [136]
ganic anion and an organic cation [121–123]. The cations [Hexpy] 1-Hexylpyridinium [134]
are typically composed of organic cores, e.g. imidazo- [Bmpyr] 1-Butyl-3-methyl pyridinium [137]
lium, phosphonium, pyrrolidinium, cholinium and many [Hmpyr] 1-Hexyl-3methyl pyridinium [137]
more [123]. The combination of a large organic cation, [Amim] 1-Allyl-3-methylimdazolium [127]
in which the positive part is either sterically hindered or [Cnpyr] 1-Alkylpyridinium [138]
shielded, and a smaller inorganic cation, causes the salts [mDBN] 5-Methyl-1,5-diazabicyclo[4.3.0]-non-5-enium [139]
to prefer the liquid state in many systems. ILs are usually [DBNH] 1,5-Diazabicyclo[4.3.0] non-5-enium [139]
divided into two classes, i.e. protic and aprotic ILs [124].
The seed for a virtual explosion in the number of ionic
liquids was the discovery of the 1-ethyl-3-methylimida- individual lignocellulosic components, such as cellulose,
zolium cation [EtMeim] in 1982 [125, 126]. A drawback can be recovered by addition of an anti-solvent, e.g. water
at the time was the requirement to avoid moisture enter- [143, 144], acetone [145] or ethanol [146] by, for example
ing the IL system. The development has been rapid since precipitation. The ILs are usually quite costly (c.f. DESs
and more and more combinations of salts are utilised in the next section) to produce, and the viscosity is high,
for very different purposes, such as man-made cellulosic which makes pumping and mixing quite energy consum-
fibres [127] and for extraction of lignin from biomass ing. Another issue is their sensitivity to moisture and the
[128]. The combination of cations and anions making recovery cost, which can be very high. Zhou et al. have
up ionic liquids are almost endless [127]. The possibility evaluated, in a very comprehensive review, a number of
to make combinations of various cations and anions is recovery options, such as distillation, extraction, adsorp-
an advantage, since the properties of the IL may change tion, membrane separation (including pervaporation and
depending on the selected combination, which opens up membrane distillation), etc. [147]. Distillation has often
for fine-tuning of the pretreatment process. Table 5 pre- been suggested as the main unit operation to be used
sents a summary of various combinations of ionic liquids. for recovery of the ILs. However, the energy cost asso-
The acronyms are not always the same in different pub- ciated with distillation is rather high, which may impair
lications. As mentioned earlier, the number of potential the overall feasibility of a large-scale process [147, 148].
ILs are very large; therefore, only some commonly uti- Lynam et al. [146] employed direct-contact membrane
lised ILs are presented. The ILs have been referred to by distillation (DCMD) at low temperatures and ambient
a number of notations: Room-temperature ionic liquid, pressure (in contrast to pervaporation methods, where
non-aqueous ionic liquid, molten salt, liquid organic salt, lower pressures are used), which was successfully used
and fused salt have all been used to describe salts in the to separate water from ­ [C2mim][O2CH] and ­ [C2mim]
liquid phase [129]. They are non-volatile, non-flammable [OAc]. The resulting IL–water mixture must be separated
and have high chemical and thermal stability. Room-tem- prior to IL recycling because water inhibits IL–biomass
perature ionic liquids (RTILs) are of special interest since pretreatment.
they can be used at low temperatures (below 100 °C) The utilisation of ILs has attracted interest regard-
[129–131]. ing their potential effect on environment and health
Ionic liquids have a potentially wide field of applications consequences. Gathergood et al. [149] investigated the
for biomass processing. A number of areas have been biodegradability of commonly used ILs. They designed
identified, such as fractionation, conversion into chemi- and evaluated ILs containing ester or amide groups in
cals, dissolution and hydrolysis. Some hurdles are yet to the alkyl side chain. They found that ILs with an ester
overcome to make ILs entering in large-scale processes in a sidechain were generally liquids at room tempera-
[140, 141]. The separation and recovery of the IL from a ture. The [Bmim]-derived ILs were found to have poor
mixture of carbohydrates, lignin and proteins can result or negligible biodegradability. However, if an ester
in challenges for large-scale biorefineries [135, 142]. The was incorporated in a side chain, the biodegradability
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 16 of 26

increased significantly. The non-toxic character and Lignin is of major interest in current biorefinery-related
their bio-degradability has been studied and several research for production of for example fuels and chemi-
publications provide great amounts of information cals [159]. Therefore, pretreatment methods that can
regarding their bio-degradability, eco-toxicity, as well extract lignin from lignocellulosic materials in its native
as cytotoxicity [122, 130, 137, 150, 151]. ILs may have form are of particular interest. However, native lignin
an effect on all levels of life; therefore, ILs have been differs from extracted lignin using available commercial
studied in various life forms, including bacteria, fungi, methods. The solubility of lignin that has been extracted
plants, animals, etc. In addition, since ILs are very from, e.g. cooking of wood for production of pulp, and
water soluble, they may have a large environmental used for solubility tests in various ILs may not say any-
impact should accidents occur [122, 150, 151]. It has thing about the ILs capacity to dissolve lignin from a
been shown that the choice of cation is of great impor- real biomass substrate [144, 160–162]. Similar studies
tance. The biodegradability of ILs having the cation were performed by Achinivu et al. [163] on three differ-
cholinium (sometimes classified as belonging to the ent protic IL cations [Mim], [Pyrr] and [Py] using acetate
DES family, which is discussed in the next section) is as anion. They also included xylan and microcellulose
usually high [122, 152]. in the study. All protic ILs were able to dissolve lignin
ILs have been used to pretreat various lignocellulosic in large quantities, while cellulose was insoluble. Xylan,
materials. Alayoubi et al. [121] used [Emim][OAc] to pre- on the other hand, has a varying solubility in the tested
treat three cellulose-containing materials: cotton, spruce ILs. However, the results cannot be directly transferred
and oak sawdust. In addition, they performed enzymatic to their ability to extract lignin from a lignocellulosic bio-
hydrolysis of the untreated and pretreated materials. In mass. The amount of extracted lignin after pretreatment
short, 2% (w/w) substrate was added to 10 ml [Emim] at 90 °C for 24 h employing [Pyrr][OAc] was greater than
[OAc] and incubated the suspension at 45 °C for 40 min. 70% of the original content in corn stover.
After incubation, the pretreated substrate was precipi- The number of publications that presents results from
tated by addition of water (2:1 v/v water-to-IL). The solid IL pretreatment of various lignocellulosic materials is
fraction was enzymatically hydrolysed and fermented to rapidly growing. In Table 6, a selection is presented of ILs
ethanol. The glucose yields after pretreatment was 70% that have been utilised for fractionation of different types
for cotton, 60% for oak sawdust, while spruce sawdust of lignocellulosic materials.
resulted in a yield of only 50%. The ethanol yields were
in all cases around 50%. This has to be considered a low Deep‑eutectic solvents
yield if compared with commonly reported ethanol yields Deep eutectic solvents (DESs) are a class of eutectic mix-
from other studies performed at higher temperatures tures of Lewis or Brønsted acids and bases, which can be
[153, 154]. Although the IL pretreatment is followed by made up from a variety of anionic and cationic species.
careful rinsing of the solid materials, residual ILs may It is considered to be a new class of ILs, since they have
have a toxic or inhibiting effect on enzymes and/or fer- many properties and characteristics in common [167].
menting organisms [137, 155, 156]. Sitepu et al. [157] They are generally classified according to four classes,
scanned one hundred and sixty-eight strains of wild yeast type I, II, II and IV and can be described by a general
(including the Saccharomyces genus) for their tolerance formula Cat+X−zY, where Cat+ can be an ammonium, a
to [Emim][OAc]. Based on growth in media containing phosphonium or a sulphonium cation. X is a Lewis base,
[Emim][OAc], tolerance levels between 1 and 5% of the which can interact with either a Brønsted or a Lewis acid
IL were observed for more than 80 strains. This indicates [167], and z is the number of Y molecules that interacts
that residual IL in the solid material after pretreatment with the selected anion. DESs can be produced relatively
is not a critical issue from this point of view. In a study easy by mixing the compounds together at moderate
by Auxenfans et al. [158] on simultaneous pretreatment temperatures. They usually have low volatility and high
and enzymatic saccharification of lignocellulosic and thermal stability [168]. The term deep-eutectic solvent
cellulosic substrates concerning spruce, utilisation of comes from the difference in the expected freezing point
­[C2mim][MeO(H)PO2 led to significant glucose yields up if two DES-forming chemicals are mixed; when the two
to a concentration of 30% IL (v/v) in the saccharification compounds are mixed at a certain ratio, to eventually
step, while [­ C2mim][OAc] was the better IL for oak saw- hit the eutectic point, the freezing point becomes much
dust. However, beyond an IL concentration of 10% (v/v) lower than the corresponding freezing points for the
the yields of glucose slowly decreased until the enzymatic pure components. The DES components are reported to
hydrolysis stopped at 50% IL or higher. Auxenfans et al. be less expensive to produce than conventional ILs. One
also found that β-glucosidase activity was sensitive to estimation states that they are about 20% cheaper than
­[C2mim][OAc] [158]. ILs, ranging from $20 kg−1 for choline chloride and for
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 17 of 26

Table 6 A selection of fractionation methods for of various lignocellulosic materials utilising ionic liquids
Biomass IL t (h) T (°C) Main product(s) Yield(s) Refs.

Bleached birch kraft pulp [Emim][OAc] 3 60 Pulp 2.37a; 2.40b [139]

[Emim][DMP] 1.28a; 1.44b
[Emim][Cl] 1.99a; 1.35b
[mDBN][DMP] 3.56a; 2.82b
[DBN][OAc] 6.56a; 1.18b
[DBN][EtCOOH] 6.94a; 4.35b
Switchgrass [C2mim][OAc] Glucose 90–95 [164]
[FurEt2NH][H2PO4]c Xylose 70–75
[VanEt2NH][H2PO4]
[p-AnisEt2NH][H2PO4]
Oak sawdust [C2mim][OAc] 40 min 110 Glucose 67–79 [153]
Spruce sawdust 66–73
Cotton fibre
Southern yellow pine [C2mim][OAc] Holocellulose/lignin 59/31 [145]
a) Triticale 1. [Emim][OAc] 0.5–24 70–150 Lignin For a) & 1: [161]
b) Wheat straw 2. [Bmim][Cl] Glucose Lignin: 52.7%
c) Flax shives 3. DMEAF Glucose: >95
4. DMEAA
5. DMEAG
6. DMEAS
Pine 1. ­[HBim][HSO4] 0.5–8 120–170 Lignin For 3: [165]
2. ­[ TEA][HSO4] Glucose Lignin: 70
3. ­[DMBA[HSO4] Glucose: 75%
Rice straw 1. ­[C2mim][Cl] 1 110 Lignin For 3: [148]
2. ­[C2mim][Cl/water] Glucose Lignin: 93.7
3. ­[C2mim][Cl/K2CO3] Glucose: 92.1
Cotton-based waste textiles [Amim][Cl] 0.5–150 90–130 Bacterial cellulose 10.8 g L−1 of nano- [143]
cellulose fibres
Wheat straw [Emim][DEP] 10–120 min 25–150 Reducing sugars (RS) ­ M−1
RS: 54.8 g g D [166]
[@130°, 30 min)
a
Residual xylan
b
Dissolved cellulose
c
The ILs utilised in [164] were synthesised from aromatic aldehydes derived from the major by-products of biofuel production from lignocellulosic materials: furfural,
vanillin and p-anisaldehyde

urea $35 kg−1 [168]. However, Socha et al. suggest a price DESs have been used for various kinds of biomass frac-
estimate for the IL ­[C2mim][OAc] of approximately $17– tionation tests. Chen et al. [172] evaluated six ternary
25 kg−1 [164]. Thus, the production costs are somewhat DESs for fractionation of switchgrass where the hydro-
uncertain. The potential applications for DESs show that gen-bond acceptor was either choline chloride (ChCl)
they are very versatile, and they have been reported to be or guanidine hydrochloride (GH). The hydrogen-bond-
used in very different situations, e.g. in the areas of elec- donators were chosen to be one of ethylene glycol (EG),
trochemistry, pharmaceuticals, fossil fuels, food and feed propylene glycol (PG), glycerine (GLY) or p-toluenesul-
production and for lignocellulosic biomass fractionation fonic acid (PTSA). The combination GH-EG-PTSA was
[169]. the most efficient, resulting in a removal of 79% xylan and
DESs can only be recognised as green solvents if they 82% lignin in 6 min at 120 °C with 10 wt% solid loading. If
can fulfil health and safety regulations, but also sustain- the solids loading increased to 35 wt%, a removal of more
ability criteria [170] just as is the case for ILs. Most DESs than 60% in 30 min was possible using GH-EG-PTSA and
are based on the hydrogen-bond acceptor choline chlo- ChCl-EG-PTSA. The resulting cellulose-rich fraction was
ride. While choline chloride is a common chemical, fre- subjected to fed-batch hydrolysis to a final solids loading
quently used in animal feed, it is believed to be safe from of 20%, which produced 128 g ­L−1 glucose.
a health perspective. It also supports important biologi- The hardwood willow was used in a study by Li et al.
cal functions in human beings [171]. In addition, several [173] using ChCL as HBA and one of lactic acid, glycerol
publications show that cholinium-based liquids can allow or urea as HBD. The combination of ChCl and lactic acid
protein structures and enzyme functions to be main- at a 1:10 molar ratio had the highest lignin-extraction
tained or even enhanced. efficacy when treated at 120 °C for 12 h. The lignin yield
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 18 of 26

was 91.8%; if the reaction time was prolonged to 42 h, The potential of DESs to be used for fractionation pur-
the increase in lignin extraction was only small reach- poses in a biorefinery is likely to increase, as the inter-
ing 94.8%. They also noticed a small decrease in cellu- est for these types of compounds is becoming larger and
lose content indicating that degradation was initiated. larger. In Table 7, a selection of DESs, which have been
The extracted lignin fraction had a purity of about 95%, utilised for fractionation of various biomass materials, is
which is very high; in addition, the ash content was about presented. It is obvious that not only woody materials are
0.5%. FTIR, 13C-NMR and 31P-NMR showed that syrin- of interest, but also various kinds of agricultural and fruit
gyl and guaiacyl units were the main constituents of the waste have been evaluated.
extracted lignin.
Another lignocellulosic biomass, loblolly pine, was Biotechnical methods
evaluated by Lynam et al. [174]. Loblolly pine (also known Biological pretreatment can be carried by applying fungi,
as Southern yellow pine) is considered as a very recalci- which breaks down lignin (white-rot fungi), cellulose
trant material. Three HBDs (formic acid, lactic acid and (soft-rot fungi), or hemicelluloses and cellulose (brown-
acetic acid) and three HBAs (choline chloride, betaine rot fungi). It has not been common to utilise any of these
and proline) were utilised in various combinations and fungi for pretreatment purposes for, e.g. ethanol produc-
ratios. Loblolly pine was added to the DESs at a ratio of tion. The rate of degradation is slow, which makes them
1:10. The fractionation was performed at a temperature impractical for industrial use. Another disadvantage is
of 155 °C for 2 h. The washed cellulose-rich fractions loss of material, which potentially could have been uti-
were assessed by enzymatic hydrolysis. An initial evalua- lised for other purposes. Nevertheless, biological pre-
tion of the lignin-solubilising capacity was performed on treatment has been used to improve methane production
alkali lignin, xylan from beech wood and medium fibrous from lignocellulosic biomass [182]. Akyol et al. point out
cellulose. The DES consisting of formic acid and choline that some components in lignocellulosic biomass are
chloride at a ratio of 2:1 exhibited the highest lignin solu- difficult to degrade employing anaerobic digestion for
bility and the lowest cellulose and xylan solubility. This biogas production. By including an aerobic step prior to
DES was also selected for fractionation of loblolly pine. the anaerobic digestion step, where Trametes versicolor
The resulting glucose yield after enzymatic hydrolysis (a white-rot fungi) was applied on the lignocellulosic
for 72 h was about 70% of the cellulose content in the biomass, the methane yield improved by 10–18%. The
untreated material. The lignin-extraction yield was not optimum residence time for the aerobic pretreatment
explicitly presented. step was found to be 6 days. The impact of an additional
One of the proposed benefits of utilising DESs is the step will have effects on the overall costs, which need to
suggestion that the extracted lignin stays in a more reac- be evaluated through a techno-economic analysis. Biogas
tive form than, e.g. the lignin from STEX pretreatment. production is the result from anaerobic digestion where
Tan et al. [41] applied acidic DESs on oil-palm empty a consortium of microorganisms is involved. Ali et al.
fruit bunch (EFB) for lignin extraction. Nine organic [183] proposed a biological pretreatment of oak sawdust
acids were selected as HBDs while choline chloride was by means of a microbial consortium prior to the biogas
the HBA choice. The molar ratios of the various combi- production step. The bacterial pretreatment caused a
nations of the DESs were ranging from 2:1 to 1:15 (HBA: significant reduction of cellulose, hemicelluloses and
HBD), while the ratio of EFB to DES was 1:10 by weight. lignin content (compared with the untreated sawdust) of
The suspensions were allowed to react for 8 h at 120 °C. 35.8, 37.1 and 46.2%, respectively, after 5 days pretreat-
The results show that a higher lignin yield was achieved ment. The biological pretreatment enhanced biometh-
for monocarboxylic-based DESs, than for the corre- ane formation from untreated oak sawdust by 92% after
sponding di- or tricarboxylic acids. The same relationship 40 days of anaerobic digestion. T. versicolor has also
was also found by Hou et al. [175] for DES extraction of been employed for production of biosurfactants, which
rice straw. Tan et al. suggests that the increase of possible belong to a product family of high value from a biore-
hydrogen-bonding sites in di- or tri-carboxylic acids may finery. Lourenço et al. [184] utilised olive-mill waste in
restrict the mobility of solvent molecules, which weakens a solid-state fermentation system. The biosurfactant was
the solvent–lignin interaction thus decreasing the lignin- able to reduce the surface tension by up to 34.5 mN ­m−1.
extraction efficacy. The DES having the highest lignin- Surfactants have a large application area, such as deter-
extraction capability was shown to be choline chloride: gents, cosmetics, pulp and paper and many other areas.
lactic acid, which at a molar ratio of 1:2 extracted more Ceriporiopsis subvermispora is another fungus that have
than 60% of the lignin. This lignin also exhibited a reac- been studied by Vasco-Correa et al. for pretreatment of
tivity on par with technical commercial lignin, based on lignocellulosic biomass [185]. The investigated the effect
the phenolic hydroxyl content in the extract. of fungal pretreatment on four lignocellulosic materials:
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 19 of 26

Table 7 A selection of DESs utilised for biomass fractionation

Biomass DES t (h) T (°C) Main product(s) Yield(s) Refs.

Lettuce leaves ChCl:glycerol 3–16 80–150 Bio-butanol Glucose: 94.9%; xylose: 75% @150°, 16 h [176]
Corn stover 1. ChCl:formic acid 0.5–3 90–130 Bio-butanol For 1: [177]
2. ChCl:urea Glucose: 99% (17 g L−1)
3. ChCl:glycerol Butanol: 5.6 g L−1 (0.17 g g−1 sugar)
4. ChCl:acetic acid
5. ChCL:oxalic acid
6. ChCl:malonic acid
7. ChCl:citric acid
Willow 1. ChCl:lactic acid 6–42 90–120 Lignin For 1: [173]
2. ChCl:glycerol Purity: 94.5%
3. ChCl:urea Yield: 91.8%
Switchgrass 1. GH-PG-PTSAa 0.1 Fractionation: Highest removal for 2: [172]
2. GH-EG-PTSAa Cellulose Cellulose: 0.70
3. GH-GLY-PTSAa Xylan Xylan: 79.4
4. ChCl-PG-PTSAa Lignin Lignin: 82.1
5. ChCl-EG-PTSAa @6 min, 120 °C
6. ChCl-GLY-PTSAa
7. GH-Eg-PTSAb
8. ChClEG-PTSAb
9. GH-EG-PTSAc
10. ChCl-EG-PTSAc
Corncob 1. ChCl:glycerol 24 90 Lignin For 1: [178]
2. ChCl:lactic acid Glucose Lignin: 71.3%
3. ChCl:glycolic acid Glucose: 96.4%
4. ChCl:levulinic acid (after EH)
5. ChCl:malonic acid
6. CHCl:glutaric acid
7. ChCl:oxalic acid
8. ChCl:ethylene glycol
9. ChCl:glycerol
Potato peels 1. ChCl:glycerol 3 60–150 Lignin For 1: [179]
2. ChCl:ethylene glycol Glucose Removal of lignin: 33%
Glucose yield: 0.80 g/g glucan
@150 °C
Apple residues 1. ChCl:glycerol 3 60–150 Lignin For 1: [179]
2. ChCl:ethylene glycol Glucose Removal of lignin: 62%
Glucose yield: 0.95 g/g glucan
@150 °C
Rice straw 1. ChCl:malic acid 0.5–12 60–121 Lignin For 4: [180]
2. ChCl:citric acid Glucose Lignin removal: 57.2%
3. ChCl:tartaric acid Ethanol 10. Glucose yield: 87.1%
4. ChCl:lactic acid 10. Ethanol yield: 89.5%
5. ChCl:oxalic acid
6. ChCl:malonic acid
7. ChCl:ethylene glycol
8. ChCl:1,2 propane diol
9: ChCl:urea
10. ChCl:glycerol
Spruce saw dust 1. ChCl:boric acid 24 80 Glucose < 20% after EH [181]
2. ChCl:glycerol
3. Betaine:glycerol
a
10% solids
b
30% solids
c
35% solids

corn stover, miscanthus, pine and white ash. The pre- was successful for the first generation of fungi on non-
treatment procedure was carried out for 14 days of incu- sterile material. However, for subsequent generations
bation time. The feedstock was treated either sterilised or only material that had been sterilised showed signs of
non-sterilised and the results changed with the sterility. degradation. Regarding corn stover, fungal pretreatment
Fungal pretreatment of pine, white ash and miscanthus of non-sterilised material was inefficient. The highest
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 20 of 26

glucose yield after enzymatic hydrolysis was found to on iron salts in combination that in combination with
be about 37% for first-generation, non-sterile pretreated hydrogen peroxide can oxidise organic contaminants,
white ash wood. such as inhibitors and toxic substances that can be detri-
In addition to fungi, bacteria exists that have the capa- mental to microorganisms. In the study, birch wood was
bility to degrade various lignocellulosic components. Guo first steam pretreated at 210 °C for 10 min. Subsequently,
et al. tested various genera from bacteria for laccase pro- the STEX material was subjected to ultrasonication at a
duction and for their hydrolytic capacity on miscanthus pH of 4 for 2 h, with or without addition of hydrogen per-
[186]. The strains included, for instance, Bacillus, Pseu- oxide and ­FeCl3. The results suggest that the combination
domonas, Exiguobacterium and Aeromonas. Miscanthus of a harsh ultrasonication and Fenton-like treatment have
was pretreated for 96 h by adding the substrate to the a negative impact on the biogas production rate, while a
culture medium containing the microorganism. At the milder treatment caused increased production rate. This
end of the pretreatment, the solid material was washed resulted in a shorter residence time to reach close to the
and subjected to enzymatic hydrolysis for 72 h. The result maximum BMP, which has a large impact on reactor size
was a lignin removal after pretreatment of about 30–60%. (or biogas capacity), which is the major capital cost in a
In addition, the increase in glucose release (compared biogas plant.
with untreated Miscanthus) was 1.3- to 2.2-fold higher. The combination of thermal pretreatment and ultra-
A maximum of 87% cellulose digestibility after enzymatic sonication was also investigated by John et al. [189] for
hydrolysis was recorded. hydrolysis of sweet lime peel. First, thermal pretreatment
A biological method, which is not by itself a pretreat- was performed with addition of sulphuric acid (0.25%
ment method, is the consolidated bioprocessing. It v/v) with a solids concentration of 17%. In an autoclave
still requires some prior pretreatment to be effective. at 121° for 1 h. The solid material after the thermal treat-
However, it has some interesting properties, such that ment was then subjected to ultrasound assisted dilute
it combines enzyme production, enzymatic hydroly- acid hydrolysis employing a design of experiments
sis and fermentation in one vessel. The most commonly method to generate the experimental conditions. The
employed microorganism are strains of Clostridium ther- design variables were acid concentration, peel concen-
mocellum, which has been found to be suitable for, e.g. tration, sonication time, temperature and amplitude. The
ethanol production. In a study by Kothari et al. [187], five highest reducing sugar yield was about 0.2 g/g pretreated
different pure model cellulose substrates were experi- peel, which is equivalent to about 60% yield based on the
mentally evaluated for the effect of enzymatic diges- cellulose content after pretreatment.
tion by C. thermocellum in comparison with fungal Chen et al. [190] compared four different treat-
enzyme mixtures. In the study, no pretreatment prior ment methods on white birch chips to evaluate the
to degradation took place. The results show that diges- effect of thermochemical pretreatment in terms of its
tion using C. thermocellum was more or less affected composition. They also assessed thermochemical and
by structural properties of the substrates, while fungal disk-refining treatment for production of fermentable
enzymes yielded a glucan conversion that diminished in carbohydrates. In addition, enzymatic hydrolysis was
the following order: milled filter paper > Avicel > Sigma- performed to evaluate the yield of sugars after the pre-
cell > α-cellulose > cotton linter. The study showed that treatment steps. The evaluation was carried out in pilot-
conversion was rather unaffected by cellulose micro- scale, where the starting material was 100 kg of birch
accessibility in contrast to the selected fungal enzyme chips. Sodium hydroxide (5% w/v dry biomass) was
blend. It needs to be mentioned that the study was car- added during the thermochemical step, which was car-
ried out at very low solid content, between 0.5 and 5%. ried out at 140 °C for 30 min. Addition of NaOH during
In addition, at 5% solid content, the resulting glucan the thermal pretreatment was found to reduce the energy
conversion dropped from above 90% to about 20–30%. consumption in the disk-refining step by up to 86%.
However, it is necessary to evaluate real lignocellulosic Particle size reduction can be carried out by a process
materials to make proper conclusions. that is common in the pulp industry, mechanical refining,
which yields a pulp that is processed further to pulp qual-
Grinding/ultrasound/other mechanical methods ities such as newspaper and paperboard. In an attempt
In Northern Europe there is a great interest in waste liq- to increase the recovery of carbohydrates and reduce the
uid streams from pulp and paper mills. Lamb et al. [188] enzyme addition, Chandra et al. [68] utilised mechani-
used ultrasonication, a Fenton-like reaction, or com- cally refined pulp for subsequent steam pretreatment of
binations of the two on steam-pretreated birch wood poplar pulp. The reactivity of lignin decreased during
to investigate if a positive impact on biomethane pro- steam pretreatment while the accessibility to the cellulose
duction could be achieved. A Fenton reaction is based fraction was improved. Compared with refiner pulp, the
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 21 of 26

steam-pretreated material performed better during enzy- decreased with an increasing disk gap. The suggested
matic hydrolysis. The addition of a pre-processing step process solution is especially attractive in an already
where poplar chips were allowed to soak in a solution of existing pulp mill. In a study performed by Huang et al.
either sodium sulphite, or sodium sulphite and sodium [192] dilute-alkali assisted ball milling of bagasse and
carbonate, at 60 °C for 16 h prior to steam pretreatment, pennisetum was employed, having NaOH concentrations
caused lignin sulphonation as well as cellulose hydrolysis ranging from 0.25% to 4.0%, followed by mild hydrother-
to become enhanced. In addition, most of the carbohy- mal pretreatment at 80 or 100 °C. The highest reducing
drate fraction retained in the solid fraction, which is in sugar yield from bagasse (40.75%) was achieved using
contrast to acid-catalysed steam pretreatment where the 4% NaOH at 100 °C for 40 min for bagasse, while 55.74%
hemicelluloses fraction solubilise. yield was obtained for pennisetum after treatment at
Physical pretreatment does not only refer to grind- 80° for 60 min. The definition of the obtained yields are
ing or milling operations. A study by Falls et al. [191] somewhat unclear. A similar study was presented by Yu
explores the effect of “shock treatment”. The method is et al. [193], who compared phosphoric acid pretreatment
based on the effect that a sudden shock wave causes on a with intense pulverisation of corn stover, which rendered
material, if treated further using, e.g. enzymatic hydroly- the material more susceptible to enzymatic hydrolysis.
sis. In the study, Falls et al. tested several lignocellulosic Other combinations of pretreatment methods include
materials, such as bagasse, corn stover, poplar wood, sor- a study by Lee et al. [194], employing ionic liquid frac-
ghum and switchgrass. One of the goals was to compare tionation, followed by a solid acid saccharification and
ball milling with shock treatment in terms of the glu- enzymatic hydrolysis; mild acid pretreatment of euca-
can enzymatic digestibility of oxidative-lime pretreated lyptus, birch or wheat straw followed by an organosolv
(OLP) substrates. For a hydrolysis time of 24 h, shock step utilising ethanol–water mixtures [195]; sequential
treatment was more efficient than ball milling for all fractionation of hardwood by a combination of STEX
substrates. With increasing residence time, ball milling and hydrotropic treatment [196]. Enzymatic hydrolysis of
was performing better than shock treatment. Compared pine from auto-hydrolysis at temperatures ranging from
with only OLP substrates an increase in digestibility was 150 to 200 °C, was used by Rigual et al. in an ionic-liquid
found, resulting in higher glucose formation. pretreatment step [197], which resulted in a digestibility
of about 79% of available glucan content.
Combination of pretreatment methods
The combination of pretreatment methods opens up
Combinations of one or more pretreatment methods to for better fractionation in many cases, if a selective frac-
improve the pretreatment process may be a possibility tion can be employed, such as pre-extraction of hemi-
to find process designs that will be suitable for enhanced celluloses by acid treatment, which could be used for
fractionation of the raw material. This could be, for high-value products such as barrier films or hydrogels,
instance, to yield process streams, which are optimised followed by an organosolv step for lignin and cellulose
for hemicelluloses, while other streams are optimised for recovery [198]. Thus, there are many suggestions for
other compounds. However, implementation of several, improving the utilisation of the raw materials by combin-
different pretreatment methods comes with an additional ing various pretreatment methods. However, it is critical
cost, if the methods are dissimilar. Therefore, it would to find combinations that match each other to avoid addi-
not be advisable to apply widely different pretreatment tional complications in terms of process design and costs.
methods. Nevertheless, pretreatment is commonly pre-
ceded by a size-reduction step, which can be regarded
as mechanical pretreatment, if the size reduction is Conclusions
thorough. The reverse operational procedure is also a The increasing interest in the utilisation of lignocellu-
possibility. Chen et al. [190] used hydrothermal treat- losic materials from agricultural, forest and other plants
ment at 140 °C for 30 min of white birch chips that had and residues is shown in the number of publications in
been impregnated with a solution of 5 wt% NaOH. The the field. The biorefinery concept is advancing from
collected material was further treated in a disk refiner an interesting idea to a promising alternative for many
at various gap sizes ranging from 0.15 to 1.00 mm. The fossil-based products. Already today, several produc-
refined material was enzymatically hydrolysed for 72 h. tion facilities can be regarded as a kind of biorefinery,
The addition of NaOH improved the combined pretreat- e.g. pulp mills, where more than one product is resulting
ment process. Compared with a disk-refining step only, from lignocellulosic materials. However, recovery of even
the process improved total-reducing sugar yield 6.4 to more constituents and production of other types of com-
42.2% based on the available cellulose, while the specific pounds calls for fractionation methods that may be dif-
energy consumption was reduced by 62%. The sugar yield ferent from traditional pulping methods.
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 22 of 26

The huge variation in lignocellulosic materials makes it Authors’ contributions

MG wrote the draft of the manuscript and finalised the text. OW edited and
difficult to find a general process design for all raw mate- commented the text. Both authors read and approved the final manuscript.
rials. The recalcitrance of softwood is much higher than
that of most agricultural or herbaceous crops and resi- Funding
Not applicable.
dues. Therefore, it is difficult to define “the best pretreat-
ment method”. In the end, this depends on the proposed Availability of data and materials
application, and any recommendation must be based on Not applicable.
a thorough techno-economic evaluation, where data has Ethics approval and consent to participate
been collected from a scale of at least pilot size. Selec- Not applicable.
tion of a suitable pretreatment method largely depends
Consent for publication
on the final application. Nevertheless, steam pretreat- Not applicable.
ment in various forms has shown to be attractive in the
first demo plants for 2G ethanol production; however, for Competing interests
The authors declare that they have no competing interests.
more extensive fractionation of lignocellulosic materials,
it may be advantageous to employ pretreatment methods Received: 22 October 2019 Accepted: 11 December 2019
that act by solubilisation.
The impressive number of suggested pretreatment
methods and publications clearly shows that the research
field is very much alive. The great interest in the more References
recent pretreatment methods, where ionic liquids and 1. IEA. Tracking clean energy progress 2017. Report. International Energy
Agency; 2017 June 6, 2017.
deep-eutectic solvents are investigated, contributes to 2. Ladanai S, Vinterbäck J. Global potential of sustainable biomass for
the publication list; however, the large number of studies energy. Uppsala Sweden: Swedish University for Agricultural Sciences
regarding steam explosion, organosolv, etc. indicates that Dept. of Energy and Technology; 2009. Report No.: ISSN 1654-9406.
3. Menon V, Rao M. Recent trends in valorization of lignocellulose to bio-
they still have a very important role to play when plan- fuel. In: Microorganisms in sustainable agriculture and biotechnology.
ning for future full-scale biorefineries. An indication can 2012. p. 381–409.
be the number of publications in the last 5 years, where 4. Berndes G, Hoogwijk M, Van Den Broek R. The contribution of biomass
in the future global energy supply: a review of 17 studies. Biomass
about 1500 papers with keywords containing steam pre- Bioenergy. 2003;25(1):1–28.
treatment (explosion) or organosolv were published. A 5. Chandel AK, Garlapati VK, Singh AK, Antunes FAF, da Silva SS. The path
similar search on ionic liquids or deep-eutectic solvents forward for lignocellulose biorefineries: bottlenecks, solutions, and
perspective on commercialization. Bioresour Technol. 2018;264:370–81.
results in a steady increase from almost none in 2010 to 6. Endres HJ. Bioplastics. Advances in biochemical engineering/biotech-
about 200 papers a year today. It is likely that the latter nology. 2019. p. 427–68.
will continue to increase with time. 7. Capolupo L, Faraco V. Green methods of lignocellulose pretreat-
ment for biorefinery development. Appl Microbiol Biotechnol.
2016;100(22):9451–67.
8. Bio-based chemicals. Value added products from biorefineries. https​://
Abbreviations www.ieabi​oener​gy.com/wp-conte​nt/uploa​ds/2013/10/Task-42-Bioba​
LN: standard litre (dry gas volume at 0 °C and 1013 mbar); VS: volatile solids sed-Chemi​cals-value​-added​-produ​cts-from-biore​finer​ies.pdf.
g kg−1; COD: chemical oxygen demand; TRL: technical readiness level; SEM: 9. Kamm B, Schönicke P, Kamm M. Biorefining of green biomass—techni-
scanning electron microscopy; MRI: magnetic resonance imaging; NMR: cal and energetic considerations. Clean Soil Air Water. 2009;37(1):27–30.
nuclear magnetic resonance; STEX: steam explosion (steam pretreatment); 10. Bello S, Ríos C, Feijoo G, Moreira MT. Comparative evaluation of ligno-
LHW: liquid hot-water treatment; BMP: biochemical methane potential; EH: cellulosic biorefinery scenarios under a life-cycle assessment approach.
enzymatic hydrolysis; SEC: size-exclusion chromatography; FTIR: Fourier- Biofuels Bioprod Biorefin. 2018;12(6):1047–64.
transform infrared spectroscopy; NMR (HSQC): heteronuclear single quantum 11. Cherubini F. The biorefinery concept: using biomass instead of
coherence nuclear magnetic resonance; RTILs: room-temperature ionic liquids; oil for producing energy and chemicals. Energy Convers Manage.
GVL: γ-valerolactone; THFA: tetrahydrofuran; IL: ionic liquid; DES: deep-eutectic 2010;51(7):1412–21.
solvent; ChCl: choline chloride; GH: guanidine hydrochloride; [Bmim]: 1-butyl- 12. Mansfield SD, Kang KY, Chapple C. Designed for deconstruction—
3-methylimidazolium; [HBim][HSO4]: 1-butylimidazolium hydrogen sulfate]; poplar trees altered in cell wall lignification improve the efficacy of
[DMBA][HSO4]: N,N-dimethylbutylammonium hydrogen sulfate; [TEA][HSO4]: bioethanol production. New Phytol. 2012;194(1):91–101.
triethylammonium hydrogen sulfate; [EMim]: 1-ethyl-3-methylimidazolium; 13. IEA Bioenergy T. Database on facilities for the production of advanced
[DEP]: diethyl phosphate; [C2mim][MeO(H)PO2]: 1-ethyl-methylimidazolium liquid and gaseous biofuels for transport 2019. https​://demop​lants​
methylphosphonate; [AMIM]Cl: 1-allyl-3-methylimidazolium chloride; [Py]: .bioen​ergy2​020.eu/. Accessed 25 Sept 2019.
pyridinium; [Mim]: 1-methylimidazolium; [Pyrr]: pyrrolidinium; [OAc]: acetate; 14. SunPine. From forest to green diesel and rosin 2019. https​://www.sunpi​
EG: ethylene glycol; PG: propylene glycol; GLY: glycerin; PTSA: p-toluenesul- ne.se/en/tekno​logi/. Accessed 26 Sept 2019.
fonic acid; HBA: hydrogen-bond acceptor; HBD: hydrogen-bond donator; 15. Birla A, Domsjö. 2019. http://www.domsj​o.adity​abirl​a.com/en/sidor​/
DMEAF: N,N-dimethylethanolammonium formate; DMEAA: N,N-dimethyleth- More-produ​cts.aspx. Accessed 26 Sept 2019.
anolammonium acetate; DMEAG: N,N-dimethylethanolammonium glycolate; 16. Cell S. Dissolving pulp. https​://www.sodra​.com/en/pulp/produ​cts/disso​
DMEAS: N,N-dimethylethanolammonium succinate; SCB: sugar cane bagasse. lving​-pulp/. Accessed 30 Sept 2019.
17. Ragauskas AJ, Beckham GT, Biddy MJ, Chandra R, Chen F, Davis MF, et al.
Acknowledgements Lignin valorization: improving lignin processing in the biorefinery. Sci-
Not applicable. ence. 2014;344(6185):1246843.
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 23 of 26

18. A/S B. Borregaard A/S. https​://borre​gaard​.com/Produ​cts-Solut​ions. 42. Studer MH, DeMartini JD, Brethauer S, McKenzie HL, Wyman CE. Engi-
Accessed 29 Sept 2019. neering of a high-throughput screening system to identify cellulosic
19. Sanford K, Chotani G, Danielson N, Zahn JA. Scaling up of renewable biomass, pretreatments, and enzyme formulations that enhance
chemicals. Curr Opin Biotechnol. 2016;38:112–22. sugar release. Biotechnol Bioeng. 2010;105(2):231–8.
20. Humbird D, Davis R, Tao YH, Kinchin C, Hsu DD, Aden A, et al. Process 43. Lindedam J, Bruun S, Jørgensen H, Decker SR, Turner GB, DeMartini
design and economics for biochemical conversion of lignocellulosic JD, et al. Evaluation of high throughput screening methods in pick-
biomass to ethanol. 2011. ing up differences between cultivars of lignocellulosic biomass for
21. Joelsson E, Erdei B, Galbe M, Wallberg O. Techno-economic evaluation ethanol production. Biomass Bioenergy. 2014;66:261–7.
of integrated first- and second-generation ethanol production from 44. Costa THF, Masarin F, Bonifácio TO, Milagres AMF, Ferraz A. The
grain and straw. Biotechnol Biofuels. 2016;9(1):1. enzymatic recalcitrance of internodes of sugar cane hybrids with
22. Tao L, Aden A, Elander RT, Pallapolu VR, Lee YY, Garlock RJ, et al. Process contrasting lignin contents. Ind Crops Prod. 2013;51:202–11.
and technoeconomic analysis of leading pretreatment technologies 45. Costa THF, Vega-Sánchez ME, Milagres AMF, Scheller HV, Ferraz A.
for lignocellulosic ethanol production using switchgrass. Bioresour Tissue-specific distribution of hemicelluloses in six different sugar-
Technol. 2011;102(24):11105–14. cane hybrids as related to cell wall recalcitrance. Biotechnol Biofuels.
23. Werpy T, Petersen G. Top value added chemicals from biomass. 2004. 2016;9(1):99.
24. Aden A, Foust T. Technoeconomic analysis of the dilute sulfuric acid 46. Weiss ND, Felby C, Thygesen LG. Enzymatic hydrolysis is limited by
and enzymatic hydrolysis process for the conversion of corn stover to biomass-water interactions at high-solids: improved performance
ethanol. Cellulose. 2009;16(4):535–45. through substrate modifications. Biotechnol Biofuels. 2019;12(1):3.
25. Humbird D, Davis R, Tao L, Kinchin C, Hsu D, Aden A, et al. Process 47. Holwerda EK, Worthen RS, Kothari N, Lasky RC, Davison BH, Fu C, et al.
design and economics for biochemical conversion of lignocellulosic Multiple levers for overcoming the recalcitrance of lignocellulosic
biomass to ethanol: dilute-acid pretreatment and enzymatic hydrolysis biomass. Biotechnol Biofuels. 2019;12(1):15.
of corn stover. United States; 2011. 48. Shen H, Poovaiah CR, Ziebell A, Tschaplinski TJ, Pattathil S, Gjersing
26. Barta Z, Kreuger E, Björnsson L. Effects of steam pretreatment and E, et al. Enhanced characteristics of genetically modified switchgrass
co-production with ethanol on the energy efficiency and process (Panicum virgatum L.) for high biofuel production. Biotechnol Biofu-
economics of combined biogas, heat and electricity production from els. 2013;6(1):71.
industrial hemp. Biotechnol Biofuels. 2013;6(1):56. 49. Wilkerson CG, Mansfield SD, Lu F, Withers S, Park JY, Karlen SD, et al.
27. Dias MOS, Modesto M, Ensinas AV, Nebra SA, Filho RM, Rossell CEV. Monolignol ferulate transferase introduces chemically labile linkages
Improving bioethanol production from sugarcane: evaluation of into the lignin backbone. Science. 2014;344(6179):90–3.
distillation, thermal integration and cogeneration systems. Energy. 50. Biswal AK, Hao Z, Pattathil S, Yang X, Winkeler K, Collins C, et al.
2011;36(6):3691–703. Downregulation of GAUT12 in Populus deltoides by RNA silencing
28. Joelsson E, Wallberg O, Börjesson P. Integration potential, resource results in reduced recalcitrance, increased growth and reduced
efficiency and cost of forest-fuel-based biorefineries. Comput Chem xylan and pectin in a woody biofuel feedstock. Biotechnol Biofuels.
Eng. 2015;82:240–58. 2015;8(1):41.
29. Börjesson P, Ahlgren S, Barta Z, Björnsson L, Ekman A, Erlandsson P, et al. 51. Galbe M, Zacchi G. Pretreatment of lignocellulosic materials for efficient
Sustainable performance of lignocellulose-based ethanol and biogas bioethanol production. Biofuels. In: Advances in biochemical engineer-
co-produced in innovative biorefinery systems. 2013. ing/biotechnology. 2007. p. 41–65.
30. Budzinski M, Cavalett O, Nitzsche R, Strømman AH. Assessment of ligno- 52. Kumar P, Barrett DM, Delwiche MJ, Stroeve P. Methods for pretreatment
cellulosic biorefineries in Germany using a hybrid LCA multi-objective of lignocellulosic biomass for efficient hydrolysis and biofuel produc-
optimization model. J Ind Ecol. 2019;23(5):1172–85. tion. Ind Eng Chem Res. 2009;48(8):3713–29.
31. Ståhl M, Nieminen K, Sixta H. Hydrothermolysis of pine wood. Biomass 53. Alvira P, Tomás-Pejó E, Ballesteros M, Negro MJ. Pretreatment technolo-
Bioenergy. 2018;109:100–13. gies for an efficient bioethanol production process based on enzymatic
32. Sjöström E. Wood chemistry. In: Sjöström E, editor. Wood chemistry. 2nd hydrolysis: a review. Bioresour Technol. 2010;101(13):4851–61.
ed. San Diego: Academic Press; 1993. p. iv. 54. Carvalheiro F, Duarte LC, Gírio FM. Hemicellulose biorefineries: a review
33. Hon DNSS, N. Wood and Cellulosic Chemistry. Second Edition, Revised on biomass pretreatments. J Sci Ind Res. 2008;67(11):849–64.
and Expanded Edited by David N.-S. Hon (Clemson University) and 55. Kim S, Holtzapple MT. Lime pretreatment and enzymatic hydrolysis of
Nubuo Shiraishi (Kyoto University). Marcel Dekker: New York and corn stover. Bioresource Technology. 2005; 96(18 SPEC. ISS.):1994–2006.
Basel. 2001. vii + 914 pp. $250.00. ISBN 0-8247-0024-4. J Am Chem Soc. 56. Chang VS, Kaar WE, Burr B, Holtzapple MT. Simultaneous sac-
2001;123(36):8880–1. charification and fermentation of lime-treated biomass. Biotech Lett.
34. MISTRA. The Eco-cyclic pulp mill. The foundation for strategic environ- 2001;23(16):1327–33.
mental research. 2003. Report No.: KAM report A100. 57. Dahunsi SO. Liquefaction of pineapple peel: pretreatment and process
35. Jönsson LJ, Alriksson B, Nilvebrant NO. Bioconversion of lignocellulose: optimization. Energy. 2019;185:1017–31.
inhibitors and detoxification. Biotechnol Biofuels. 2013;6(1):16. 58. Harun S, Geok SK. Effect of sodium hydroxide pretreatment on rice
36. Macrelli S, Galbe M, Wallberg O. Effects of production and market fac- straw composition. Indian J Sci Technol. 2016;9(21):95245.
tors on ethanol profitability for an integrated first and second genera- 59. Chu Q, Song K, Hu J, Bu Q, Zhang X, Chen X. Integrated process for
tion ethanol plant using the whole sugarcane as feedstock. Biotechnol the coproduction of fermentable sugars and lignin adsorbents from
Biofuels. 2014;7(1):26. hardwood. Bioresour Technol. 2019;289:121659.
37. Faith WL. Development of the Scholler process in the United States. Ind 60. Kapoor M, Soam S, Agrawal R, Gupta RP, Tuli DK, Kumar R. Pilot scale
Eng Chem. 1945;37(1):9–11. dilute acid pretreatment of rice straw and fermentable sugar recovery
38. Tengborg C, Stenberg K, Galbe M, Zacchi G, Larsson S, Palmqvist E, at high solid loadings. Bioresour Technol. 2017;224:688–93.
et al. Comparison of S­ O2 and H­ 2SO4 impregnation of softwood prior to 61. Shuai L, Yang Q, Zhu JY, Lu FC, Weimer PJ, Ralph J, et al. Comparative
steam pretreatment on ethanol production. Appl Biochem Biotechnol. study of SPORL and dilute-acid pretreatments of spruce for cellulosic
1998;70–72:3–15. ethanol production. Bioresour Technol. 2010;101(9):3106–14.
39. Al-Rudainy B, Galbe M, Hernandez MA, Jannasch P, Wallberg O. Impact 62. Sarks C, Bals BD, Wynn J, Teymouri F, Schwegmann S, Sanders K, et al.
of lignin content on the properties of hemicellulose hydrogels. Poly- Scaling up and benchmarking of ethanol production from pelletized
mers. 2018;11(1):35. pilot scale AFEX treated corn stover using Zymomonas mobilis 8b. Biofu-
40. Qi XM, Chen GG, Gong XD, Fu GQ, Niu YS, Bian J, et al. Enhanced els. 2016;7(3):253–62.
mechanical performance of biocompatible hemicelluloses-based 63. Bridgeman TG, Darvell LI, Jones JM, Williams PT, Fahmi R, Bridgwater AV,
hydrogel via chain extension. Sci Rep. 2016;6:33603. et al. Influence of particle size on the analytical and chemical properties
41. Tan YT, Ngoh GC, Chua ASM. Effect of functional groups in acid of two energy crops. Fuel. 2007;86(1–2):60–72.
constituent of deep eutectic solvent for extraction of reactive lignin.
Bioresour Technol. 2019;281:359–66.
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 24 of 26

AFEX™ pretreated rice straw as source of fermentable sugars: the influ-

64. Harun S, Balan V, Takriff MS, Hassan O, Jahim J, Dale BE. Performance of 86. del Carmen Fong Lopez M, Rigal M, Rigal L, Vilarem G, Vandenbossche
V. Influence of temperature and soda concentration in a thermo-mech-
ence of particle size. Biotechnol Biofuels. 2013;6(1):40. ano-chemical pretreatment for bioethanol production from sweet corn
65. Bals B, Teymouri F, Haddad D, Julian WA, Vismeh R, Jones AD, et al. co-products. Ind Crops Prod. 2019;133:317–24.
Presence of acetamide in milk and beef from cattle consuming AFEX- 87. Walch E, Zemann A, Schinner F, Bonn G, Bobleter O. Enzymatic
treated crop residues. J Agric Food Chem. 2019;67(38):10756–63. saccharification of hemicellulose obtained from hydrothermally
66. Boehm RM. The Masonite process. Ind Eng Chem. 1930;22(5):493–7. pretreated sugar cane bagasse and beech bark. Bioresour Technol.
67. Bauer A, Lizasoain J, Theuretzbacher F, Agger JW, Rincón M, Menardo S, 1992;39(2):173–7.
et al. Steam explosion pretreatment for enhancing biogas production 88. Borrega M, Nieminen K, Sixta H. Degradation kinetics of the
of late harvested hay. Bioresour Technol. 2014;166:403–10. main carbohydrates in birch wood during hot water extraction
68. Chandra RP, Chu Q, Hu J, Zhong N, Lin M, Lee JS, et al. The influence in a batch reactor at elevated temperatures. Bioresour Technol.
of lignin on steam pretreatment and mechanical pulping of poplar to 2011;102(22):10724–32.
achieve high sugar recovery and ease of enzymatic hydrolysis. Biore- 89. Szczodrak J, Ilczuk Z, Rogalski J, Leonowicz A. Intensification of Oak saw-
sour Technol. 2016;199:135–41. dust enzymatic hydrolysis by chemical or hydrothermal pretreatment.

gens JF. Ethanol production potential from AFEX™ and steam-exploded

69. Mokomele T, Da Costa Sousa L, Balan V, Van Rensburg E, Dale BE, Gör- Biotechnol Bioeng. 1986;28(4):504–10.
90. Prado JM, Lachos-Perez D, Forster-Carneiro T, Rostagno MA. Sub- and
sugarcane residues for sugarcane biorefineries. Biotechnol Biofuels. supercritical water hydrolysis of agricultural and food industry residues
2018;11(1):127. for the production of fermentable sugars: a review. Food Bioprod
70. Zhang X, Zhu J, Sun L, Yuan Q, Cheng G, Argyropoulos DS. Extraction Process. 2016;98:95–123.
and characterization of lignin from corncob residue after acid-catalyzed 91. Manzanares P, Ballesteros I, Negro MJ, Gonzlez A, Oliva JM, Ballesteros
steam explosion pretreatment. Ind Crops Prod. 2019;133:241–9. M. Processing of extracted olive oil pomace residue by hydrothermal
71. Bondesson PM, Galbe M. Process design of SSCF for ethanol production or dilute acid pretreatment and enzymatic hydrolysis in a biorefinery
from steam-pretreated, acetic-acid-impregnated wheat straw. Biotech- context. Renew Energy. 2020;145:1235–45.
nol Biofuels. 2016;9(1):222. 92. Pińkowska H, Krzywonos M, Wolak P. Valorization of rapeseed meal by
72. Weber B, Estrada-Maya A, Sandoval-Moctezuma AC, Martínez-Cienfue- hydrothermal treatment—effect of reaction parameters on low molec-
gos IG. Anaerobic digestion of extracts from steam exploded Agave ular products distribution. Cellul Chem Technol. 2019;53(7–8):755–65.
tequilana bagasse. J Environ Manage. 2019;245:489–95. 93. Pińkowska H, Krzywonos M, Wolak P, Złocińska A. Production of uronic
73. Aski AL, Borghei A, Zenouzi A, Ashrafi N, Taherzadeh MJ. Effect of steam acids by hydrothermolysis of pectin as a model substance for plant
explosion on the structural modification of rice straw for enhanced bio- biomass waste. Green Process Syn. 2019;8(1):683–90.
degradation and biogas production. BioResources. 2019;14(1):464–85. 94. Chu J, Jiang W, Wu S, Lucia LA, Lei M. Hydrothermal-controlled conver-
74. Bonfiglio F, Cagno M, Rey F, Torres M, Böthig S, Menéndez P, et al. sion of black liquor acid sediment directly to phenolics. Energy Fuels.
Pretreatment of switchgrass by steam explosion in a semi-continuous 2017;31(2):1638–43.
pre-pilot reactor. Biomass Bioenergy. 2019;121:41–7. 95. Bouxin FP, David Jackson S, Jarvis MC. Organosolv pretreatment of
75. Yi J, Zhang Q, Li X, Wang X, Li B, Zhu W. Steam explosion technology Sitka spruce wood: conversion of hemicelluloses to ethyl glycosides.
based for oil extraction from sesame (Sesamum indicum L.) seed. J Saudi Bioresour Technol. 2014;151:441–4.
Soc Agric Sci. 2019;18(1):1–6. 96. Zhang Z, Harrison MD, Rackemann DW, Doherty WOS, O’Hara IM.
76. Álvarez C, Sáez F, González A, Ballesteros I, Oliva JM, Negro MJ. Produc- Organosolv pretreatment of plant biomass for enhanced enzymatic
tion of xylooligosaccharides and cellulosic ethanol from steam- saccharification. Green Chem. 2016;18(2):360–81.
exploded barley straw. Holzforschung. 2019;73(1):35–44. 97. Tan X, Zhang Q, Wang W, Zhuang X, Deng Y, Yuan Z. Comparison study
77. Nielsen F, Galbe M, Zacchi G, Wallberg O. The effect of mixed agricul- of organosolv pretreatment on hybrid pennisetum for enzymatic sac-
tural feedstocks on steam pretreatment, enzymatic hydrolysis, and charification and lignin isolation. Fuel. 2019;249:334–40.
cofermentation in the lignocellulose-to-ethanol process. Biomass 98. Chin DWK, Lim S, Pang YL, Lim CH, Lee KM. Two-staged acid hydrolysis
Conversion and Biorefinery. 2019. on ethylene glycol pretreated degraded oil palm empty fruit bunch for
78. Brownell HH, Yu EKC, Saddler JN. Steam-explosion pretreatment of sugar based substrate recovery. Bioresour Technol. 2019;292:121967.
wood: effect of chip size, acid, moisture content and pressure drop. 99. Alonso DM, Hakim SH, Zhou S, Won W, Hosseinaei O, Tao J, et al. Increas-
Biotechnol Bioeng. 1986;28(6):792–801. ing the revenue from lignocellulosic biomass: maximizing feedstock
79. Muzamal M, Jedvert K, Theliander H, Rasmuson A. Structural changes in utilization. Sci Adv. 2017;3(5):e1603301.
spruce wood during different steps of steam explosion pretreatment. 100. Lê HQ, Ma Y, Borrega M, Sixta H. Wood biorefinery based on
Holzforschung. 2015;69(1):61–6. γ-valerolactone/water fractionation. Green Chem. 2016;18(20):5466–76.
80. Jedvert K, Wang Y, Saltberg A, Henriksson G, Lindström ME, Theliander 101. Bozell JJ, Black SK, Myers M, Cahill D, Miller WP, Park S. Solvent frac-
H. Mild steam explosion: a way to activate wood for enzymatic treat- tionation of renewable woody feedstocks: organosolv generation of
ment, chemical pulping and biorefinery processes. Nord Pulp Pap Res J. biorefinery process streams for the production of biobased chemicals.
2012;27(5):828–35. Biomass Bioenergy. 2011;35(10):4197–208.
81. Demartini JD, Foston M, Meng X, Jung S, Kumar R, Ragauskas AJ, et al. 102. Patri AS, Mostofian B, Pu Y, Ciaffone N, Soliman M, Smith MD, et al. A
How chip size impacts steam pretreatment effectiveness for biological Multifunctional cosolvent pair reveals molecular principles of biomass
conversion of poplar wood into fermentable sugars. Biotechnol Biofu- deconstruction. J Am Chem Soc. 2019;141(32):12545–57.
els. 2015;8(1):209. 103. Geilen FMA, Engendahl B, Harwardt A, Marquardt W, Klankermayer J,
82. Vidal BC Jr, Dien BS, Ting KC, Singh V Jr. Influence of feedstock particle Leitner W. Selective and flexible transformation of biomass-derived
size on lignocellulose conversion—a review. Appl Biochem Biotechnol. platform chemicals by a multifunctional catalytic system. Angewandte
2011;164(8):1405–21. Chemie. 2010;49(32):5510–4.
83. Wang H, Liu Z, Hui L, Ma L, Zheng X, Li J, et al. Understanding the struc- 104. Qian M, Lei H, Villota E, Mateo W, Zhao Y, Huo E, et al. Optimization of
tural changes of lignin in poplar following steam explosion pretreat- delignification from Douglas fir sawdust by alkaline pretreatment with
ment. Holzforschung. 2019. sodium hydroxide and its effect on structural and chemical properties
84. Cornejo A, Alegria-Dallo I, García-Yoldi Í, Sarobe Í, Sánchez D, Otazu E, of lignin and pyrolysis products. Bioresour Technol Rep. 2019;8:100339.
et al. Pretreatment and enzymatic hydrolysis for the efficient produc- 105. Yao L, Yoo CG, Pu Y, Meng X, Muchero W, Tuskan GA, et al. Phys-
tion of glucose and furfural from wheat straw, pine and poplar chips. icochemical changes of cellulose and their influences on Populus
Bioresour Technol. 2019;288:121583. trichocarpa digestibility after different pretreatments. BioResources.
85. Kruyeniski J, Ferreira PJT, Carvalho MDG, Vallejos ME, Felissia FE, 2019;14(4):9658–76.
Area MC. Physical and chemical characteristics of pretreated slash 106. Matsakas L, Raghavendran V, Yakimenko O, Persson G, Olsson E, Rova
pine sawdust influence its enzymatic hydrolysis. Ind Crops Prod. U, et al. Lignin-first biomass fractionation using a hybrid organo-
2019;130:528–36. solv—steam explosion pretreatment technology improves the
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 25 of 26

saccharification and fermentability of spruce biomass. Bioresour 128. Achinivu EC. Protic ionic liquids for lignin extraction—a lignin charac-
Technol. 2019;273:521–8. terization study. Int J Mol Sci. 2018;19(2):428.
107. Hoyer K. Production of ethanol from spruce at high solids concentra- 129. Hallett JP, Welton T. Room-temperature ionic liquids: solvents for syn-
tions: an experimental study on process development of simultaneous thesis and catalysis 2. Chem Rev. 2011;111(5):3508–76.
saccharification and fermentation: Department of Chemical Engineer- 130. Dias AMA, Cortez AR, Barsan MM, Santos JB, Brett CMA, De Sousa HC.
ing, Lund University; 2013. Development of greener multi-responsive chitosan biomaterials doped
108. Lê HQ, Pokki JP, Borrega M, Uusi-Kyyny P, Alopaeus V, Sixta H. Chemi- with biocompatible ammonium ionic liquids. ACS Sustain Chem Eng.
cal recovery of Î-valerolactone/water biorefinery. Ind Eng Chem Res. 2013;1(11):1480–92.
2018;57(44):15147–58. 131. Tanaka K, Ishiguro F, Chujo Y. POSS ionic liquid. J Am Chem Soc.
109. Huber GW, Iborra S, Corma A. Synthesis of transportation fuels 2010;132(50):17649–51.
from biomass: chemistry, catalysts, and engineering. Chem Rev. 132. Seddon KR. A taste of the future. Nat Mater. 2003;2(6):363–5.
2006;106(9):4044–98. 133. Xiong D, Wang H, Li Z, Wang J. Recovery of ionic liquids with aqueous
110. Horváth IT, Mehdi H, Fábos V, Boda L, Mika LT. γ-Valerolactone-a sustain- two-phase systems induced by carbon dioxide. 2012;5(11):2255–61.
able liquid for energy and carbon-based chemicals. Green Chem. 134. Choudhury AR, Winterton N, Steiner A, Cooper AI, Johnson KA.
2008;10(2):238–42. In situ crystallization of low-melting ionic liquids. J Am Chem Soc.
111. Lê HQ, Zaitseva A, Pokki JP, Ståhl M, Alopaeus V, Sixta H. Solubility of 2005;127(48):16792–3.
organosolv lignin in γ-valerolactone/water binary mixtures. Chemsu- 135. Zhang Q, Hu J, Lee DJ. Pretreatment of biomass using ionic liquids:
schem. 2016;9(20):2939–47. research updates. Renew Energy. 2017;111:77–84.
112. Cheng F, Sun J, Wang Z, Zhao X, Hu Y. Organosolv fractionation 136. Scurto AM, Aki SNVK, Brennecke JF. Carbon dioxide induced separation
and simultaneous conversion of lignocellulosic biomass in aque- of ionic liquids and water. Chem Commun. 2003;3(5):572–3.
ous 1,4-butanediol/acidic ionic-liquids solution. Ind Crops Prod. 137. Docherty KM, Kulpa CF Jr. Toxicity and antimicrobial activity of imidazo-
2019;138:111573. lium and pyridinium ionic liquids. Green Chem. 2005;7(4):185–9.
113. da Silva CMS, Vital BR, Rodrigues FDÁ, de Almeida ÊW, Carneiro 138. Brandt A, Hallett JP, Leak DJ, Murphy RJ, Welton T. The effect of the ionic
ADCO, Cândido WL. Hydrothermal and organic-chemical treatments liquid anion in the pretreatment of pine wood chips. Green Chem.
of eucalyptus biomass for industrial purposes. Bioresour Technol. 2010;12(4):672–9.
2019;289:121731. 139. Roselli A, Hummel M, Vartiainen J, Nieminen K, Sixta H. Understanding
114. Schmetz Q, Teramura H, Morita K, Oshima T, Richel A, Ogino C, et al. the role of water in the interaction of ionic liquids with wood polymers.
Versatility of a dilute acid/butanol pretreatment investigated on various Carbohyd Polym. 2017;168:121–8.
lignocellulosic biomasses to produce lignin, monosaccharides and cel- 140. Asim AM, Uroos M, Naz S, Sultan M, Griffin G, Muhammad N, et al.
lulose in distinct phases. ACS Sustain Chem Eng. 2019;7(13):11069–79. Acidic ionic liquids: promising and cost-effective solvents for process-
115. Romaní A, Larramendi A, Yáñez R, Cancela Á, Sánchez Á, Teixeira JA, ing of lignocellulosic biomass. J Mol Liq. 2019;287:110943.
et al. Valorization of Eucalyptus nitens bark by organosolv pretreat- 141. Klein-Marcuschamer D, Simmons BA, Blanch HW. Techno-economic
ment for the production of advanced biofuels. Ind Crops Prod. analysis of a lignocellulosic ethanol biorefinery with ionic liquid pre-
2019;132:327–35. treatment. Biofuels Bioprod Biorefin. 2011;5(5):562–9.
116. Kim TH, Ryu HJ, Oh KK. Improvement of organosolv fractionation 142. Feng Y, Li Q, Kang G, Ji G, Tang Y, Tu J. Aqueous two-phase/reverse
performance for rice husk through a low acid-catalyzation. Energies. micelle continuous process for recycling and simultaneous purifica-
2019;12(9):1800. tion of polar ionic liquid from enzymatic hydrolysate. J Chem Technol
117. Chin DWK, Lim S, Pang YL, Leong LK, Lim CH. Investigation of organo- Biotechnol. 2016;91(2):394–9.
solv pretreatment to natural microbial-degraded empty fruit bunch for 143. Hong F, Guo X, Zhang S, Han SF, Yang G, Jönsson LJ. Bacterial cellulose
sugar based substrate recovery. Energy Procedia. 2019;158:1065–71. production from cotton-based waste textiles: enzymatic sacchari-
118. Yuan W, Gong Z, Wang G, Zhou W, Liu Y, Wang X, et al. Alkaline organo- fication enhanced by ionic liquid pretreatment. Bioresour Technol.
solv pretreatment of corn stover for enhancing the enzymatic digest- 2012;104:503–8.
ibility. Bioresour Technol. 2018;265:464–70. 144. Brandt A, Gräsvik J, Hallett JP, Welton T. Deconstruction of lignocellu-
119. Zhang Y, Hou Q, Fu Y, Xu C, Smeds AI, Willför S, et al. One-step fractiona- losic biomass with ionic liquids. Green Chem. 2013;15(3):550–83.
tion of the main components of bamboo by formic acid-based organo- 145. Sun N, Rahman M, Qin Y, Maxim ML, Rodríguez H, Rogers RD. Complete
solv process under pressure. J Wood Chem Technol. 2018;38(3):170–82. dissolution and partial delignification of wood in the ionic liquid
120. Meighan BN, Lima DRS, Cardoso WJ, Baêta BEL, Adarme OFH, Santucci 1-ethyl-3-methylimidazolium acetate. Green Chem. 2009;11(5):646–55.
BS, et al. Two-stage fractionation of sugarcane bagasse by autohy- 146. Lynam JG, Chow GI, Coronella CJ, Hiibel SR. Ionic liquid and water
drolysis and glycerol organosolv delignification in a lignocellulosic separation by membrane distillation. Chem Eng J. 2016;288:557–61.
biorefinery concept. Ind Crops Prod. 2017;108:431–41. 147. Zhou J, Sui H, Jia Z, Yang Z, He L, Li X. Recovery and purification of ionic
121. Alayoubi R, Mehmood N, Husson E, Kouzayha A, Tabcheh M, liquids from solutions: a review. RSC Adv. 2018;8(57):32832–64.
Chaveriat L, et al. Low temperature ionic liquid pretreatment of 148. Gao J, Chen C, Wang L, Lei Y, Ji H, Liu S. Utilization of inorganic salts
lignocellulosic biomass to enhance bioethanol yield. Renew Energy. as adjuvants for ionic liquid–water pretreatment of lignocellulosic
2020;145:1808–16. biomass: enzymatic hydrolysis and ionic liquid recycle. 3 Biotech.
122. Gomes JM, Silva SS, Reis RL. Biocompatible ionic liquids: fundamental 2019;9(7):264.
behaviours and applications. Chem Soc Rev. 2019;48(15):4317–35. 149. Gathergood N, Garcia MT, Scammells PJ. Biodegradable ionic liquids:
123. Costa SPF, Azevedo AMO, Pinto PCAG, Saraiva ML. Environmental part I. Concept, preliminary targets and evaluation. Green Chem.
impact of ionic liquids: recent advances in (Eco)toxicology and (Bio) 2004;6(3):166–75.
degradability. Chemsuschem. 2017;10(11):2321–47. 150. Amde M, Liu JF, Pang L. Environmental application, fate, effects,
124. Kundu K, Chandra GK, Umapathy S, Kiefer J. Spectroscopic and com- and concerns of ionic liquids: a review. Environ Sci Technol.
putational insights into the ion-solvent interactions in hydrated aprotic 2015;49(21):12611–27.
and protic ionic liquids. Phys Chem Chem Phys. 2019;21(37):20791–804. 151. Egorova KS, Gordeev EG, Ananikov VP. Biological activity of ionic liquids
125. Wilkes JS, Zaworotko MJ. Air and water stable 1-ethyl-3-methylimidazo- and their application in pharmaceutics and medicine. Chem Rev.
lium based ionic liquids. J Chem Soc Chem Commun. 1992;13:965–7. 2017;117(10):7132–89.
126. Wilkes JS, Levisky JA, Wilson RA, Hussey CL. Dialkylimidazolium 152. Hou XD, Liu QP, Smith TJ, Li N, Zong MH. Evaluation of toxicity and
chloroaluminate melts: a new class of room-temperature ionic liquids biodegradability of cholinium amino acids ionic liquids. PLoS ONE.
for electrochemistry, spectroscopy, and synthesis. Inorg Chem. 2013;8(3):e59145.
1982;21(3):1263–4. 153. Auxenfans T, Buchoux S, Larcher D, Husson G, Husson E, Sarazin C.
127. Hummel M, Michud A, Tanttu M, Asaadi S, Ma Y, Hauru LKJ, et al. Ionic Enzymatic saccharification and structural properties of industrial wood
liquids for the production of man-made cellulosic fibers: opportunities sawdust: recycled ionic liquids pretreatments. Energy Convers Manage.
and challenges. In: Advances in polymer science. 2015. p. 133-68. 2014;88:1094–103.
Galbe and Wallberg Biotechnol Biofuels (2019) 12:294 Page 26 of 26

154. Auxenfans T, Buchoux S, Djellab K, Avondo C, Husson E, Sarazin C. Mild pre- 179. Procentese A, Raganati F, Olivieri G, Russo ME, Rehmann L, Marzocchella
treatment and enzymatic saccharification of cellulose with recycled ionic A. Deep eutectic solvents pretreatment of agro-industrial food waste.
liquids towards one-batch process. Carbohyd Polym. 2012;90(2):805–13. Biotechnol Biofuels. 2018;11(1):37.
155. Wang Y, Radosevich M, Hayes D, Labbé N. Compatible ionic liquid-cellulases 180. Kumar AK, Parikh BS, Shah E, Liu LZ, Cotta MA. Cellulosic ethanol produc-
system for hydrolysis of lignocellulosic biomass. Biotechnol Bioeng. tion from green solvent-pretreated rice straw. Biocatal Agric Biotechnol.
2011;108(5):1042–8. 2016;7:14–23.
156. Liu Z, Li L, Liu C, Xu A. Saccharification of cellulose in the ionic liquids and 181. Wahlström R, Hiltunen J, de Sirkka MP, Vuoti S, Kruus K. Comparison of three
glucose recovery. Renew Energy. 2017;106:99–102. deep eutectic solvents and 1-ethyl-3-methylimidazolium acetate in the
157. Sitepu IR, Shi S, Simmons BA, Singer SW, Boundy-Mills K, Simmons CW. Yeast pretreatment of lignocellulose: effect on enzyme stability, lignocellulose
tolerance to the ionic liquid 1-ethyl-3-methylimidazolium acetate. FEMS digestibility and one-pot hydrolysis. RSC Adv. 2016;6(72):68100–10.
Yeast Res. 2014;14(8):1286–94. 182. Akyol Ç, Ince O, Bozan M, Ozbayram EG, Ince B. Biological pretreatment with
158. Auxenfans T, Husson E, Sarazin C. Simultaneous pretreatment and enzymatic Trametes versicolor to enhance methane production from lignocellulosic
saccharification of (ligno) celluloses in aqueous-ionic liquid media: a biomass: a metagenomic approach. Ind Crops Prod. 2019;140:111659.
compromise. Biochem Eng J. 2017;117:77–86. 183. Ali SS, Al-Tohamy R, Manni A, Luz FC, Elsamahy T, Sun J. Enhanced digestion
159. Zakzeski J, Bruijnincx PCA, Jongerius AL, Weckhuysen BM. The catalytic of bio-pretreated sawdust using a novel bacterial consortium: micro-
valorization of lignin for the production of renewable chemicals. Chem bial community structure and methane-producing pathways. Fuel.
Rev. 2010;110(6):3552–99. 2019;254:115604.
160. Pu Y, Jiang N, Ragauskas AJ. Ionic liquid as a green solvent for lignin. J Wood 184. Lourenço LA, Magina MD, Tavares LBB, de Souza SMA, Román M, Altmajer
Chem Technol. 2007;27(1):23–33. D. Biosurfactant production by Trametes versicolor grown on two-
161. Fu D, Mazza G, Tamaki Y. Lignin extraction from straw by ionic liquids and phase olive mill waste in solid-state fermentation. Environ Technol.
enzymatic hydrolysis of the cellulosic residues. J Agric Food Chem. 2018;39(23):3066–76.
2010;58(5):2915–22. 185. Vasco-Correa J, Luo X, Li Y, Shah A. Comparative study of changes in compo-
162. Lee SH, Doherty TV, Linhardt RJ, Dordick JS. Ionic liquid-mediated selective sition and structure during sequential fungal pretreatment of non-sterile
extraction of lignin from wood leading to enhanced enzymatic cellulose lignocellulosic feedstocks. Ind Crops Prod. 2019;133:383–94.
hydrolysis. Biotechnol Bioeng. 2009;102(5):1368–76. 186. Guo H, Zhao Y, Chen X, Shao Q, Qin W. Pretreatment of Miscanthus with
163. Achinivu EC, Howard RM, Li G, Gracz H, Henderson WA. Lignin extraction biomass-degrading bacteria for increasing delignification and enzymatic
from biomass with protic ionic liquids. Green Chem. 2014;16(3):1114–9. hydrolysability. Microb Biotechnol. 2019;12(4):787–98.
164. Socha AM, Parthasarathi R, Shi J, Pattathil S, Whyte D, Bergeron M, et al. 187. Kothari N, Bhagia S, Zaher M, Pu Y, Mittal A, Yoo CG, et al. Cellulose hydrolysis
Efficient biomass pretreatment using ionic liquids derived from lignin and by: Clostridium thermocellum is agnostic to substrate structural proper-
hemicellulose. Proc Natl Acad Sci USA. 2014;111(35):E3587–95. ties in contrast to fungal cellulases. Green Chem. 2019;21(10):2810–22.
165. Gschwend FJV, Chambon CL, Biedka M, Brandt-Talbot A, Fennell PS, Hallett 188. Lamb JJ, Islam MH, Hjelme DR, Pollet BG, Lien KM. Effect of power ultrasound
JP. Quantitative glucose release from softwood after pretreatment with and Fenton reagents on the biomethane potential from steam-exploded
low-cost ionic liquids. Green Chem. 2019;21(3):692–703. birchwood. Ultrason Sonochem. 2019;58:104675.
166. Li Q, He YC, Xian M, Jun G, Xu X, Yang JM, et al. Improving enzymatic hydroly- 189. John I, Pola J, Appusamy A. Optimization of ultrasonic assisted saccharifica-
sis of wheat straw using ionic liquid 1-ethyl-3-methyl imidazolium diethyl tion of sweet lime peel for bioethanol production using Box–Behnken
phosphate pretreatment. Bioresour Technol. 2009;100(14):3570–5. method. Waste Biomass Valorization. 2019;10(2):441–53.
167. Smith EL, Abbott AP, Ryder KS. Deep eutectic solvents (DESs) and their appli- 190. Chen J, Adjallé K, Barnabé S, Perrier M, Paris J. Mechanical and thermal pre-
cations. Chem Rev. 2014;114(21):11060–82. treatment processes for increasing sugar production from woody biomass
168. Gorke JT, Srienc F, Kazlauskas RJ. Hydrolase-catalyzed biotransformations in via enzymatic hydrolysis. Waste Biomass Valorization. 2019;10(7):2057–65.
deep eutectic solvents. Chem Commun. 2008;10:1235–7. 191. Falls M, Madison M, Liang C, Karim MN, Sierra-Ramirez R, Holtzapple MT.
169. Satlewal A, Agrawal R, Bhagia S, Sangoro J, Ragauskas AJ. Natural deep Mechanical pretreatment of biomass—Part II: shock treatment. Biomass
eutectic solvents for lignocellulosic biomass pretreatment: recent Bioenergy. 2019;126:47–56.
developments, challenges and novel opportunities. Biotechnol Adv. 192. Huang J, Zhu Y, Liu T, Sun S, Ren J, Wu A, et al. A novel wet-mechanochemical
2018;36(8):2032–50. pretreatment for the efficient enzymatic saccharification of lignocel-
170. Mbous YP, Hayyan M, Hayyan A, Wong WF, Hashim MA, Looi CY. Applications luloses: small dosage dilute alkali assisted ball milling. Energy Convers
of deep eutectic solvents in biotechnology and bioengineering—prom- Manage. 2019;194:46–54.
ises and challenges. Biotechnol Adv. 2017;35(2):105–34. 193. Yu H, Xiao W, Han L, Huang G. Characterization of mechanical pulverization/
171. Shahriari S, Tomé LC, Araújo JMM, Rebelo LPN, Coutinho JAP, Marrucho IM, phosphoric acid pretreatment of corn stover for enzymatic hydrolysis.
et al. Aqueous biphasic systems: a benign route using cholinium-based Bioresour Technol. 2019;282:69–74.
ionic liquids. RSC Adv. 2013;3(6):1835–43. 194. Lee KM, Ngoh GC, Chua ASM, Yoon LW, Ang TN, Lee MG. Comparison study
172. Chen Z, Jacoby WA, Wan C. Ternary deep eutectic solvents for effective bio- of different ionic liquid pretreatments in maximizing total reducing sugars
mass deconstruction at high solids and low enzyme loadings. Bioresour recovery. BioResources. 2014;9(1):1552–64.
Technol. 2019;279:281–6. 195. Løhre C, Laugerud GAA, Huijgen WJJ, Barth T. Lignin-to-liquid-solvolysis (LtL)
173. Li T, Lyu G, Liu Y, Lou R, Lucia LA, Yang G, et al. Deep eutectic solvents (DESs) of organosolv extracted lignin. ACS Sustain Chem Eng. 2018;6(3):3102–12.
for the isolation of willow lignin (salix matsudana cv. zhuliu). Int J Mol Sci. 196. Olsson J, Novy V, Nielsen F, Wallberg O, Galbe M. Sequential fractionation of
2017;18(11):2266. the lignocellulosic components in hardwood based on steam explosion
174. Lynam JG, Kumar N, Wong MJ. Deep eutectic solvents’ ability to solubilize and hydrotropic extraction. Biotechnol Biofuels. 2019;12(1):1.
lignin, cellulose, and hemicellulose; thermal stability; and density. Biore- 197. Rigual V, Domínguez JC, Santos TM, Rivas S, Alonso MV, Oliet M, et al. Autohy-
sour Technol. 2017;238:684–9. drolysis and microwave ionic liquid pretreatment of Pinus radiata: imaging
175. Hou XD, Li AL, Lin KP, Wang YY, Kuang ZY, Cao SL. Insight into the structure- visualization and analysis to understand enzymatic digestibility. Ind Crops
function relationships of deep eutectic solvents during rice straw pretreat- Prod. 2019;134:328–37.
ment. Bioresour Technol. 2018;249:261–7. 198. Brosse N, Hussin MH, Rahim AA. Organosolv processes. In: Advances in
176. Procentese A, Raganati F, Olivieri G, Russo ME, Rehmann L, Marzocchella A. biochemical engineering/biotechnology 2019. p. 153–76.
Low-energy biomass pretreatment with deep eutectic solvents for bio-
butanol production. Bioresour Technol. 2017;243:464–73.
177. Xu GC, Ding JC, Han RZ, Dong JJ, Ni Y. Enhancing cellulose accessibility of Publisher’s Note
corn stover by deep eutectic solvent pretreatment for butanol fermenta- Springer Nature remains neutral with regard to jurisdictional claims in pub-
tion. Bioresour Technol. 2016;203:364–9. lished maps and institutional affiliations.
178. Zhang C-W, Xia S-Q, Ma P-S. Facile pretreatment of lignocellulosic biomass
using deep eutectic solvents. Bioresour Technol. 2016;219:1–5.