Journal of Environmental Management 236 (2019) 100–107

Contents lists available at ScienceDirect

Journal of Environmental Management

journal homepage: www.elsevier.com/locate/jenvman

Research article

Variability of food waste chemical composition: Impact of thermal pre- T

treatment on lignocellulosic matrix and anaerobic biodegradability
Pamela Pagliacciaa,b, Agata Gallipolia, Andrea Gianicoa, Fausto Gironia,b, Daniele Montecchioa,
Carlo Pastorec, Luigi di Bitontoc, Camilla M. Bragugliaa,∗
a
Water Research Institute (IRSA), National Research Council (CNR), Strada Provinciale 35d, km 0.7, 00010, Montelibretti, Rome, Italy
b
Dipartimento di Ingegneria Chimica, Università degli Studi di Roma “La Sapienza”, Via Eudossiana 18, 00184, Rome, Italy
c
Water Research Institute (IRSA), National Research Council (CNR), Via F. de Blasio 5, 70132, Bari, Italy

A R T I C LE I N FO A B S T R A C T

Keywords: A comprehensive sustainable Food Waste (FW) management is globally needed in order to reduce the en-
Food waste vironmental pollution and the financial costs due to FW disposal; anaerobic digestion is considered as one of the
Lignocellulosic matrix best environmental-friendly alternatives to this aim. A deep investigation of the chemical composition of dif-
Thermal pre-treatment ferent Food waste types (cooked kitchen waste (CKW), fruit and vegetable scraps (FVS) and organic fraction of
Methane potential
municipal solid waste (OFMSW)) is here reported, in order to evaluate their relevant substance-specific prop-
Fermentability
Hydrogen
erties and their impact on anaerobic biodegradability by means of a sophisticated automatic batch test system.
Suitability for a mild thermal pre-treatment (T = 134 °C and p = 3.2 bar) to enhance the biological degradation
of hardly accessible compounds was investigated. The pre-treatment affected significantly the carbohydrates
solubilisation, and was able in reducing part of the lignocellulosic matrix. Moreover, in mesophilic conditions,
the high solubilized sugars content favoured the initial recovery of hydrogen (not consumed by hydro-
genotrophic methanogenesis), allowing to newly assess the extent of prompt fermentability. Pre-treatment en-
hanced hydrogen yields of FVS and OFMSW, with gains up to +50%, while the successive methane production,
occurring in the same reactor, resulted affected by the lack of the soluble part of carbohydrates, “subtracted” for
H2 production. Only in thermophilic conditions, when no hydrogen in the biogas was detected, pre-treatment of
OFMSW significantly increased methane yield (from 0.343 to 0.389 L CH4 g−1 VSfed). A thermal pre-treatment
seems the recommended solution in order to reduce part of the recalcitrant lignocellulosic matrix of food waste,
to improve energy recovery and to eliminate the extra cost needed for pasteurization.

1. Introduction economy. However, AD is very sensitive to disturbances, such as

changes in feed materials and temperature. This is due to the fact that
Food waste (FW) is one of the largest source of solid waste, more AD is a highly dynamic and multi-step process resulting from a con-
than paper or plastic. In particular, one-third of the food produced in tinuous interaction between substrate, microorganisms/enzymes and
the world and allocated for human consumption actually never reached environment creating a delicate equilibrium of the whole process,
the consumers table (FAO, 2015). The definition of organic fraction of which is mainly dependent on substrate characteristics (Fisgativa et al.,
municipal solid waste (OFMSW) varies regionally and nationally. In the 2016, 2017; Shen et al., 2013; Zhang et al., 2014). It is already assessed
European Union, OFMSW is considered a mixture of wastes from parks, that FW has a high potential for methane production (Braguglia et al.,
gardens and FW (Al Seadi et al., 2013), not as “fresh” as the source 2018; Jiang et al., 2010; Lee et al., 2008), because the high organic
selected FW, because of the time elapsed between the production and fraction together with the high water content makes it a perfect can-
the collection (Alibardi and Cossu, 2015). Anaerobic digestion (AD) is a didate for AD (Heo et al., 2004; Zhang et al., 2007). Physical and
proven reliable and technologically simple process for converting or- chemical characteristics largely influence the process stability and lead
ganic “leftovers” in biogas thus allowing energy recovery as well as to considerably different results in biogas production and quality. For
meeting the landfill disposal standards for biodegradable organic ma- complex matrices such as FW, it is impossible to standardize the AD
terials, towards the closure of the loop as requested by the circular process: a case-by-case approach is required in order to calculate yields

∗
Corresponding author.
E-mail address: [email protected] (C.M. Braguglia).

https://doi.org/10.1016/j.jenvman.2019.01.084
Received 6 April 2018; Received in revised form 18 January 2019; Accepted 23 January 2019
Available online 02 February 2019
0301-4797/ © 2019 Elsevier Ltd. All rights reserved.
P. Pagliaccia et al. Journal of Environmental Management 236 (2019) 100–107

and composition of the biogas produced (Curry and Pillay, 2012). cafeteria serves approximately 300 researchers per day producing
Taking this into account, an accurate characterization (especially in 400 kg FW per week. For this study, food waste underwent an accurate
terms of macromolecular compounds analysis of the input material as manual screening in order to maximize the amount of pasta and rice,
protein, carbohydrate, lipid, and lignocellulosic matrix seems to be an and fruit and vegetable wastes. CKW, in fact, consisted of mixed cooked
essential step in FW AD (Alibardi and Cossu, 2016; Elbeshbishy and food as pasta, vegetable soup, rice, and bread, while FVS was mainly
Nakhla, 2012; Fisgativa et al., 2016; Fisgativa et al., 2017). Since the composed by fruit and vegetable peelings. Successively, CKW and FVS
biodegradability of a substrate depends on its properties and is affected were firstly manually chopped and then shredded by a food processor,
by many factors, at present poorly understood (Raposo et al., 2011), a before being stored at (−20)°C. OFMSW was typical source sorted
detailed characterization is also important to determine whether a pre- waste collected from the municipality of Treviso (North-East Italy)
treatment is needed or not, and which pre-treatment is optimal in every deriving from domestic source. It was mechanically sorted (plastic, iron
single case (Cabbai et al., 2013). Starch containing wastes are usually and ferrous materials were removed and screened) and shredded using
easily hydrolyzed into its carbohydrate components, while cellulose, a blade hammer, 15 mm cut size. The collected material contained 15%
hemicellulose and lignin containing wastes need further pre-treatment, of “garden waste” and 10% of 'Paper and card' of the total waste
because the lignin matrix, acting as physical barrier, often render it sample.
unavailable to microbial breakdown (Zheng et al., 2014). Lig-
nocellulose is commonly present in four major sources, namely forest 2.2. Autoclave pre-treatment
residues, agricultural residues, energy crops and, last but not least, also
municipal solid waste (OFMSW or food waste) (Romero-Cedillo et al., Before the pre-treatment test, the waste sample was thawed and
2017). The role of pre-treatments is to specifically modify these struc- mixed with tap water at different weight ratios, depending on the solids
tural properties in order to make the substrate more accessible to mi- concentration. The addition of water was essential to ease the final
crobial degradation (Loow et al., 2016). Pre-treatment has been a major grinding phase and to maximize the effectiveness of pre-treatment.
argument of research during the past 30 years and the AD improvement After the grinding phase, one portion of the waste was thermally
in terms of increased hydrogen/methane yields or solids reduction are pre-treated in a bench scale autoclave Laboklav 25b, with total capacity
well established (Braguglia et al., 2018). Anyway, there is a lack of of 25 L and able to work at Tmax of 134 °C and pmax 3.2 bar. Retention
common and standardised protocols for the evaluation of the pre- time was set at 20 min on 300 mL of sample. The applied specific
treatment efficiency (Komilis and Ham, 2003). In fact, there is the thermal energy resulted in 503 kJ kg−1FW.
tendency of correlating solubilisation and enhancement in substrate The effectiveness of the autoclave pre-treatment was expressed by
biodegradability, while there is no correlation at all, as widely docu- means of the degree of parameters changed availability (for COD,
mented in literature (Ariunbaatar et al., 2014; Carlsson et al., 2012; Jin proteins, carbohydrates in soluble phase while for lignin, cellulose,
et al., 2016; Karthikeyan et al., 2017; Liu et al., 2012; Pagliaccia et al., hemicellulose and starch in particulate phase), according to equation
2016; Strong et al., 2010; Tampio et al., 2014). Nevertheless, thermal (1):
pre-treatments may result in a reduced biodegradability due to the
formation of complex and refractory organic compounds, such as X after pre − treatment − Xbefore pre − treatment
X change (%) = •100
Maillard compounds (Ariunbaatar et al., 2014; Kondusamy and Xbefore pre − treatment (1)
Kalamdhad, 2014; Liu et al., 2012; Menon et al., 2016; Tampio et al.,
2014), generated at different temperatures depending on the type of the In the case of negative value, the absolute value was taken as the
substrate (Ariunbaatar et al., 2014). Lignocellulosic substrates, as also grade of transformation of the initial compound in other forms.
food waste, contain an adequate amount of sugar to be considered
suitable for bio-hydrogen production, but they primarily need to be 2.3. Anaerobic batch digestion tests-experimental design
deprived of the rigid structure, before sugars can be saccharified
(Lalaurette et al., 2009): pre-treatment seems to be a necessary step for Batch anaerobic digestion tests were conducted using the previously
biohydrogen production for this kind of substrate. described different substrates to evaluate the effect of a variable che-
So far, few recent studies report detailed and exhaustive FW char- mical composition as well as the influence of the pre-treatment. The
acterization based on a source base (Cabbai et al., 2013; Edwiges et al., Biomethane Potential (BMP) tests for determining the anaerobic bio-
2018; Fisgativa et al., 2016, 2017) but no detailed investigation of pre- degradability in terms of methane (and hydrogen, if feasible) yield of
treatment effect on chemical composition (in particular lignocellulosic CKW, FVS, and OFMSW were carried out by using the Automatic
matrix) and related AD process performances was reported. Methane Potential Test System (AMPTS) II (Bioprocess, Sweden).
In this context, the present paper aims to evaluate the composition AMPTS consisted of 15 parallel batch reactors of 500 mL capacity,
variability among different types of FW with an in-depth analysis of mechanically stirred, heated and filled up to 300 mL of working vo-
pre-treatment effects on the different properties (lignocellulosic frac- lume. The CO2-fixing unit vials were filled with 3N NaOH, thus al-
tion in particular). The relationship between physico-chemical prop- lowing the absorption of carbon dioxide and sulfuric acid. Methane
erties and AD performances was assessed in terms of parameters change content was measured online using automated data logging system,
and accurate anaerobic batch tests by means of a sophisticated auto- while hydrogen and biogas composition was analysed by gas chroma-
matic batch test system. The whole set of experimental data was sta- tography (Fig. 1).
tistically analysed by ANOVA to identify the possible correlations and Each reactor bottle was partially filled with inoculum and substrate,
the statistical differences between variables and parameters. according to a substrate/inoculum (S/I) ratio of 0.6 on a VS basis ac-
cording to the optimized results of our preliminary assessment
2. Materials and methods (Braguglia et al., 2018; Pagliaccia et al., 2016). Temperature regimen
was mesophilic (37 °C) for FVS and OFMSW digestion tests, and ther-
2.1. Food waste origin mophilic (55 °C) for OFMSW and CKW tests.
Sludge used as inoculum for mesophilic tests was extracted from a
Three different FW samples were investigated in this study: cooked mesophilic anaerobic digester while thermophilic inoculum was batch
kitchen waste (CKW), fruit and vegetable scraps (FVS) and organic acclimated before starting the experiment, without any pre-treatment.
fraction of municipal solid waste (OFMSW). Each sample was collected Both inocula were degassed for several days until no significant me-
in a single acquisition. CKW and FVS were collected from the cafeteria thane production was observed, in order to deplete the residual bio-
of the research area “Roma 1” of the National Research Council. The degradable organic matter (Angelidaki et al., 2009).

101
P. Pagliaccia et al. Journal of Environmental Management 236 (2019) 100–107

In order to evaluate a “prompt fermentability” value from the so-

luble sugars fraction of each feedstock, together with the classical
biodegradability value, the BHPth and the BMPth calculated for each
type of feedstock were compared to the experimental ones (respectively
BHPexp and BMPexp) and following the ensuing equations (4) and (5) to
BMPexp
Biodegradability = •100
BMPth (4)

BHPexp
Prompt sugars fermentability = •soluble sugars fraction·100
BHPth
Fig. 1. Automatic Methane Potential Test System (AMPTS) device composed of (5)
sample incubation unit (A), CO2-absorption traps tray (B), gas volume mea-
suring device (C), gas-sampling unit (d) and data acquisition system (e). Statistical analysis was performed by using descriptive statistics
data from the physical and chemical characterization, as the BHP and
BMP values. Analysis of variance (ANOVA) was used to identify the
The average characteristics of the degassed inocula were: a pH of existence of significant differences between the treatment means, and
8.5 ± 0.2, total solids (TS) content of 19.0 ± 2 g L−1, volatile solids the Tukey test (p < 0.05) was performed in order to group the data.
(VS) content of 9.8 ± 1 g L−1 and total COD of 11.2 ± 2 g L−1.
Reactors were flushed with N2 gas before sealing in order to create
2.5. Analytical methods
an anaerobic environment in all batch headspaces. Blank tests were
performed for each set of BMP test so that the methane production
Total and volatile solids were determined according to standard
attributable to the residual organic matter contained in the inoculum
methods (APHA, 1998). The pH was detected by a portable pH meter
could be accounted for. In order to follow the changes in the FW
Eutech Instruments pH 700. Soluble (sCOD) and total COD, measured in
composition throughout the digestion, each test was carried out in four
duplicates, were determined by means of COD Cell Test of Spectroquant
reactors and then at regular time intervals (2, 6 and 14 days) one re-
Merck (EPA method 410.4). Total nitrogen was determined photo-
actor was stopped and analysis were performed. The last reactor was
metrically after transformation of organic and inorganic nitrogen
run for over 30 days until no significant methane production was ob-
compounds into nitrate according to Koroleff's method by treatment
served over 1-week period. All tests were performed in duplicates.
with an oxidizing agent in a thermoreactor (T = 120 °C for 1 h) and
after reaction with 2,6-dimethylphenol (DMP) to form 4-nitro-2,6-di-
2.4. Theoretical methane potential (BMPth)
methylphenol in a solution acidified with sulfuric and phosphoric acid.
Nitrogen (total) cell tests by Spectroquant Merck (digestion analogous
The theoretical methane potential is widely used to represent the
to EN ISO 11905-1 and determination of nitrate analogous to DIN
methane production of a specific organic substrate. In this study, thanks
38405-9) were used. Ammonium nitrogen (NH4+-N) was determined
to an accurate and complete chemical characterization of the different
according to APHA (1998). To analyse proteins and carbohydrates,
feedstocks, the BMPth was calculated and used as touchstone to eval-
samples aliquots were filtered through glass filters (1.2 μm; GF/C
uate the anaerobic biodegradability of CKW, FVS and OFMSW. The
Whatman); protein content was determined by bicinchoninic acid
individual components of cellulose and hemicellulose are degradable in
protein assay (Pierce, Rockford, USA) using standard solution of bovine
anaerobic conditions; only when these components are tightly bound by
serum albumin, which was modified from Lowry method (Lowry et al.,
lignin (i.e. as lignocellulosic matrix) do they become inaccessible for
1951), while carbohydrates determination was based on a modified
microorganisms and thereby non-degradable. The theoretical CH4
DuBois method (Braguglia et al., 2012; DuBois et al., 1956). Total
production of each major component in the VS fraction was calculated
proteins content was estimated by multiplying the organic nitrogen
using the stoichiometric equation (Symons and Buswell, 1933); BMPth
content by a factor 6.25.
(LCH4 g−1VSfed) and anaerobic biodegradability were calculated using
NREL method for “Determination of Structural Carbohydrates and
following equation (2):
Lignin in Biomass” was partially adapted (di Bitonto et al., 2016) and
L STP L L applied on the different lyophilised food waste samples in order to
BMPth = 0.415 • + 0.496 • STP + 1.014 • STP + 0.727
g carbohydrates g proteins glipids evaluate the content and the composition of simple and structural
L STP complex sugars. 500 mg of lyophilised solids were suspended into
• 100 mL of 4%wt of sulfuric acid and stirred at room temperature for 1 h
glignin (2)
(Suspension A). 1 mL of this suspension was filtered, opportunely di-
−1
where BMPth is expressed in LCH4 g VS and C57H104O6, C5H7O2N, luted and analysed for determination of free sugars (glucose, fructose
C6H10O5 and C10H13O3 are the empirical formulas for lipid, protein, and saccharose were mainly found). Then, Suspension A was refluxed
carbohydrate and lignin, respectively (Triolo et al., 2011). for 1 h. The resultant filtered suspension was analysed, by allowing
For the first time, a BHPth (Biohydrogen Potential) was calculated hemicelluloses, starches and pectinic sugars to be determined. Whereas,
referring the observed hydrogen production to the catabolic fermenta- the solid filtered from “Suspension A” was washed with over 100 mL of
tion of the soluble sugars via the acetate pathway (disregarding the MilliQ water and dried at 378 K for 24 h. Resultant dried solids were
anabolism), reported in equation (3). finally weighted. Then, 100 mg of this residue was suspended and kept
under agitation at 303 K for 60 min in 3 mL of 72% sulfuric acid and
C6 H12 O6 + 2H2 O→ 2CH3 COOH + 2CO2 + 4H2 (3)
with 84 mL of MilliQ water, the acid solution was transferred into a
This hypothesis was corroborated by the absence of methane pro- 250 mL glass balloon and kept under reflux for 1 h. The suspension was
duction during the first 48 h of the test (when H2 production stopped) cooled, filtered by crucible, and analysed for sugars determinations. On
and the decrease of sugars concentration in the reactor after the same the other side, filtered solids were abundantly washed with MilliQ
period, suggesting sugars fermentation and no hydrogenotrophic me- water and dried at 378 K for 24 h and weighted. Insoluble lignin was
thanogenesis. The maximum theoretical stoichiometric production for calculated by the difference between this weight and the respective
equation (2) corresponds to 4 mol of H2 per 1 mol of glucose consumed, ashes obtained at 823 K for 3 h. Sugars were determined by using a
that is, in terms of COD, a theoretical hydrogen yield of 0.498 LH2 GS50 chromatography system (Dionex-Thermo Fisher Scientific, Sun-
g−1glucose. nyvale, CA, USA) equipped with an autosampler (AS50) and a pulsed

102
P. Pagliaccia et al. Journal of Environmental Management 236 (2019) 100–107

Table 1
Characterization of untreated and thermal pre-treated wastes.
CKW FVS OFMSW

Untreated Pre-treated Untreated Pre-treated Untreated Pre-treated

pH 4.6 ± 0.1 4.6 ± 0.1 4.7 ± 0.1 4.7 ± 0.1 4.7 ± 0.1 4.7 ± 0.1
TS (g kg−1
ww) 236 ± 8 244 ± 10 54 ± 4 59 ± 6 334 ± 18 354 ± 15
VS/TS (%) 95 ± 1 96 ± 1 83 ± 1 83 ± 1 89 ± 1 91 ± 1
Soluble COD (g kg−1
ww) 60 ± 4 105 ± 8 21 ± 1 23 ± 1 88 ± 3 127 ± 4
Total COD (g kg−1
ww) 325 ± 18 325 ± 14 55 ± 3 58 ± 3 350 ± 25 395 ± 35
COD/VS 1.4 ± 0.1 1.4 ± 0.1 1.2 ± 0.1 1.2 ± 0.1 1.2 ± 0.2 1.2 ± 0.2
Total Nitrogen (%TS) 2.9 ± 0.1 2.9 ± 0.1 2.4 ± 0.1 2.4 ± 0.1 1.3 ± 0.1 1.3 ± 0.1
Total Proteins (%COD) 13.3 ± 0.5 14.4 ± 0.4 12.6 ± 1.0 13.6 ± 1.1 7.1 ± 0.7 5.2 ± 1.5
Total Carbohydrates (%COD) 79.8 ± 2.7 74.7 ± 2.7 50.1 ± 0.4 49.8 ± 0.4 47.3 ± 0.7 53.2 ± 1.0
Soluble 7.4 ± 0.5 14.0 ± 1.1 15.2 ± 0.3 16.7 ± 0.3 15.2 ± 0.3 25.5 ± 0.7
Hemicellulose 5.5 ± 0.1 4.8 ± 0.1 15.8 ± 0.2 16.0 ± 0.2 5.3 ± 0.3 4.3 ± 0.3
Cellulose 0.7 ± 0.1 0.7 ± 0.1 6.0 ± 0.1 5.8 ± 0.1 2.4 ± 0.3 1.9 ± 0.3
Starch 66.2 ± 2.6 55.2 ± 2.5 13.1 ± 0.2 11.3 ± 0.2 24.4 ± 0.5 21.5 ± 0.5
Lignin (%COD) n.q.a n.q.a 15.6 ± 1.1 15.5 ± 1.1 20.8 ± 0.5 20.0 ± 0.5
Lipids (%COD) 6.0 ± 0.5 7.0 ± 0.5 10.0 ± 1.5 10.0 ± 1.0 10.0 ± 1.0 11.0 ± 1.0
Not mapped COD (VFA and alcohols, %COD) 0.9 ± 0.1 3.9 ± 1.0 11.7 ± 2 11.1 ± 2.0 14.8 ± 2.0 10.6 ± 1.5
Lignocellulose matrix (%COD) 6.2 ± 0.3 5.5 ± 0.2 37.4 ± 0.2 37.3 ± 0.2 28.5 ± 0.3 26.2 ± 0.3
BHPth (LH2 g−1VSfed) 0.037 ± 0.001 0.070 ± 0.002 0.082 ± 0.002 0.086 ± 0.002 0.080 ± 0.002 0.144 ± 0.005
BMPth (LCH4 g−1VSfed) 0.441 ± 0.020 0.441 ± 0.020 0.462 ± 0.030 0.459 ± 0.030 0.434 ± 0.030 0.430 ± 0.030

a
n.q.: not quantifiable.

amperometric detector (ED50, with a gold electrode). The above- carbohydrates, proteins and lipids (Table 1). However, the low biode-
mentioned microfiltered solutions were injected (25 μL loop) and ana- gradability of the lignocellulose matrix in AD reactors is due principally
lysed with a Carbopack PA10 Analytical column (250 mm, 4 mm; to lignin being non-degradable in anaerobic conditions, while cellulose
Dionex) using a flow of 1.0 mL min−1 of an aqueous gradient of KOH and hemicellulose are tightly packed in the lignin and are therefore
generated by an EG40 Eluent generator (KOH concentration pro- responsible for drastically reduced biodegradability (Triolo et al.,
file:18 mM for 30 min, 200 mM for 10 min and again 18 mM for 2011). In order to evaluate the possible effect of pre-treatment methods
15 min). Glucose, Fructose, Arabinose, Mannose, Xylose and Galactose aiming at breaking up the lignocellulose structure, knowledge of the
were used as standard for qualitative and quantitative determinations. quantitative distribution of lignocellulose fibres in the feedstock is
The biogas composition was measured using a PerkinElmer Auto mandatory. An organic component balance was calculated as the sum of
System Gas Chromatograph equipped with a thermal conductivity de- cellulose, hemicellulose, starch, lignin, protein and lipids in terms of
tector (TCD). Methane and hydrogen productions were referred to 35 °C COD; the sum covered 85–99% of the total COD, suggesting that the
and 1 atm. analysed components constituted the majority of the organic matter
present in each feedstock.
3. Results and discussion In particular, CKW was mainly composed of carbohydrates. In
particular, starch (more than 65% of the total COD) was considerably
3.1. Characterization of the investigated food wastes higher than in the other FW (13% for FVS and 24% for OFMSW).
Hemicellulose and cellulose were mostly abundant in FVS (22% of total
The characterization of different compounds (carbohydrates, pro- COD), while lignin, the bio-recalcitrant fraction, was in the range
teins, lipids and fibres), in terms of COD, and the quantification of the 16–21% for both FVS and OFMSW, and not present is CKW. Lignin, in
theoretical potential to produce hydrogen (BHPth) and methane fact, typically comes from the vegetable–fruit part of the food waste and
(BMPth) are summarized in Table 1, for untreated and pre-treated CKW, mainly from the paper/cardboard fraction of the OFMSW (Komilis and
FVS and OFMSW. Ham, 2003). The lignin content of food and simulated food waste re-
Solids concentration of the investigated wastes (CKW, FVS and ported in literature ranges from 0.9 to 12%TS (Pecorini et al., 2016).
OFMSW) ranged between 50 and 334 g kg−1 ww with high organic sub- The high lipid and “not mapped COD” concentrations of OFMSW were
stance content (VS/TS = 83–95%), revealing significant water content probably due, respectively, to meat, fish, oil and cheese residues wasted
and high transformation potential of these substrates in AD process. in household kitchens, and to pre-degradation phenomena starting
FVS are low in TS and VS if compared with CKW and OFMSW, as fruit during the storage producing intermediates (Fisgativa et al., 2017).
and vegetable were mostly made up of water, as reported also in the Moreover, the composition of OFMSW is generally heavily dependent
literature (Alibardi and Cossu, 2015; Edwiges et al., 2018). On the other on the generating source and the collection time in a specific munici-
hand, if the ratio between soluble and total COD was taken into ac- pality or area (Alibardi and Cossu, 2015). The pH of the three in-
count, FVS presented the highest value, namely 38%, due to a more vestigated feedstocks was 4.7 ± 0.1, according to the statistical ana-
efficient grinding phase (the higher water content in the FVS might lysis reported by Fisgativa et al. (2016), where the pH of food waste was
have favoured the conveyance of organic matter into solution). Ni- 5.1 ± 0.7 whatever the FW categories. In the case of fruit and vege-
trogen content was lower (1.3% TS) in the case of OFMSW while ranged table wastes in Lebanon, however, the pH was 5.1 (Ghanimeh et al.,
between 2.4 and 2.9% for the cafeteria food waste (CKW and FVS). 2018), while French municipal biowaste presented a pH of 4.7 ± 0.2
Similar food waste investigated elsewhere were characterized by TS (Fisgativa et al., 2017). The low pH combined with the high carbohy-
ranging between 10 and 40%, with VS/TS between 87 and 97.1%, and drates contents of these feedstocks may induce high production of VFA
nitrogen content ranging between 2.1 and 5.2% of TS (Alibardi and and hydrogen (Braguglia et al., 2018; Pagliaccia et al., 2016).
Cossu, 2015; Zhang et al., 2007; Kim et al., 2006; Lim et al., 2008; The accurate chemical characterization of the feedstocks permitted
Mata-Alvarez et al., 1992). to calculate also the theoretical hydrogen and methane potentials
As regards chemical composition, food waste contain various sub- (BHPth and BMPth) for each feedstock. In detail, the BHPth was calcu-
stances belonging to the three main groups of biodegradable organics: lated from the available soluble carbohydrates transformed

103
P. Pagliaccia et al. Journal of Environmental Management 236 (2019) 100–107

theoretically into H2, while the BMPth is calculated considering the hemicellulose trapped in lignin cross-links (Tampio et al., 2014). Lignin
contribution of each component, except the carbohydrate fraction fraction, when present, underwent a minor conversion (below 5%), due
transformed during the first hours into H2 along equation (3). to the mild temperature and low pressure of the pre-treatment (Lawther
The lowest BHPth of 0.037 LH2 g−1VSfed was estimated for CKW, et al. (1996).
rich of starch while poor in soluble sugars, whereas FVS and OFMSW The impact of the different feedstock composition on the pre-
presented similar high BHPs, in the range of 0.080–0.082 LH2 g−1VSfed. treatment effects is showed in Fig. 2. FVS, characterized by the highest
The theoretical BMPth calculated for the three feedstocks (CKW, FVS fraction of lignocellulosic material presented the lowest organics solu-
and OFMSW) resulted not statistically different and was in the range bilisation (+10% of COD); while for CKW the soluble COD increase was
0.434–0.462 ± 0.030 LCH4 g−1VSfed, in line with literature values over 70%. In the case of OFMSW, thermal pre-treatment was highly
(Bayard et al., 2017). In particular, fruit and vegetable waste collected effective in reducing all the three polymers linked together in the lig-
over one year in Brazil presented an average theoretical BMP of nocellulosic matrix, probably due to the low hemicellulose content. The
0.475 ± 41 L CH4 g−1 VSfed (Edwiges et al., 2018), not statistically different composition of the lignocellulosic matrix could be the reason
different from 0.462 ± 0.030 L CH4 g−1 VSfed obtained here for FVS. for this result. While FVS lignocellulosic matrix was composed by more
than 40% of hemicellulose, OFMSW matrix presented a significantly
3.2. Impact of thermal pre-treatment on FW organic fractions lower content of hemicellulose (less than 20%) affecting positively the
accessibility of the matrix and consequently the effect of the pre-
Pre-treatment had a significant impact on the distribution of the treatment. Hemicellulose degradation depends, in fact, on its chemical
biochemical components in soluble and particulate form. Complex nature and relationship with lignin within the matrix (Jördening and
carbohydrates (starch, cellulose and hemicellulose) and lignin percen- Winter, 2005).
tage were determined also in the pre-treated substrate to determine the
efficiency of thermal pre-treatment on solubilisation of recalcitrant fi- 3.3. Food waste fermentability and biodegradability assessment
bres. Autoclave pre-treatment was found to be effective in the solubi-
lisation of organic matter for all the substrates analysed. Solubilisation The main factors affecting anaerobic batch tests are linked to the
affected mostly the carbohydrate fraction, whose soluble concentration food and inoculum characteristics (source, activity), the operational
almost doubled for CKW and OFMSW. Soluble COD and proteins had conditions and the substrate to inoculum ratio (S/I) (Raposo et al.,
nonetheless relevant increases, although the values were variable ac- 2011). In this study, in order to assess the intrinsic biodegradability of
cording to the substrate. Pecorini et al. (2016) also reported similar the different feedstocks, inoculum characteristics and S/I were main-
results in their study: soluble COD, carbohydrates and proteins were tained constant, while temperature regimen was changed.
found in higher concentrations in the pre-treated OFMSW sample. Major BMP results are shown in Table 2.
Carbohydrates underwent a 55–60% increase compared to almost In mesophilic conditions, FVS and OFMSW produced only hydrogen
100% reached in the present study for the most similar substrate. Dif- in the first hours of digestion (Fig. 3), with statistically comparable
ference can be explained in the lower pressure (2 bar) adopted. Menon specific production of 0.027 ± 0.001 LH2 g−1VSfed. Significant pre-
et al. (2016) registered the highest solubilisation degree of organic treatment effect was observed on the hydrogen conversion rates that
substrate with their autoclave treatment on canteen FW. Even higher increased up to +50% for OFMSW, due principally to the enhanced
results were achieved by Liu et al. (2012) for kitchen waste and vege- sugar solubilisation that favoured hydrogen generation. As reported
table/fruit residues, but pre-treatment temperature was sensibly up- also in Ding et al. (2017) in the case of two-stage digestion, the hy-
raised (175 °C for 60 min). drogen yield of the thermal pre-treated FW gradually increased and
Lignocellulosic and starch fraction were found to be lower in the peaked at 43.0 mL g−1VS, which was 23.9% higher than that of the
pre-treated substrate, according to the corresponding increase in the untreated FW (34.7 mL g−1VS). The variation trend in hydrogen yield
soluble carbohydrates (Fig. 2). with pre-treatment temperature was in correspondence with the
Starch content is by far the easiest to process and was found to be changes in soluble carbohydrate content in FW.
the one with the higher grade of transformation (14% as average). If The prompt fermentability of the available soluble sugars accounted
exposed to temperatures rising up to 140 °C, starch swells, dextrinizes for 5.2–5.4% for the untreated feedstocks of FVS and OFMSW, but in-
and hydrolyses to reducing sugar, lowering its content in the garbage creased up to 8% for the pre-treated OFMSW evidencing the role of
(Ren et al., 2006). The lignocellulose matrix is primarily made up of thermal pre-treatment in fermentation enhancement.
three biopolymers, i.e. cellulose, hemicellulose, and lignin, which are It is worth noting that the acidification phase occurring during
strongly intermeshed and chemically bonded together by non-covalent hydrogen production was reversible. The declining pH recovered,
forces and by covalent cross-linkages (Perez et al., 2002). Results con- without carbonate addition, during the digestion period assuring suc-
firm that autoclaving is partly able to release cellulose and cessive methane production for untreated and pre-treated substrates.
Nevertheless, methane yields were affected by the consumption of the
soluble part of the carbohydrate fraction, converted to hydrogen (see
equation (3)). In fact, the methane yield of the untreated feedstock was
even higher. The operating temperature affected the initial production
of hydrogen, as demonstrated by the different results obtained during
batch tests carried out with same OFMSW in mesophilic and thermo-
philic conditions.
In fact, in thermophilic conditions no H2 production from OFMSW
and CKW could be spotted, probably due to the enhanced kinetics fa-
vouring the immediate conversion into methane. Regarding OFMSW,
CH4 production occurred instantly and just after 4 days around 95% of
the total CH4 was produced, while for the thermal pre-treated OFMSW
time was reduced to 3.5 days. For CKW, due to the lowest content of
soluble bioavailable sugars (7% of total COD), CH4 conversion rate was
Fig. 2. Pre-treatment effect on lignocellulosic (cellulose, hemicellulose and slower and most of the production occurred after 6 days. Moreover, no
lignin) and soluble fractions (carbohydrates, proteins) of CKW, FVS and effect of thermal pre-treatment on kinetics and biodegradability was
OFMSW. observed.

104
P. Pagliaccia et al. Journal of Environmental Management 236 (2019) 100–107

Table 2
Summary of major results for the batch digestion tests.
FVS (meso) OFMSW (meso) OFMSW (thermo) CKW (Thermo)

Untreated Pre-treated Untreated Pre-treated Untreated Pre-treated Untreated Pre-treated

−1
Hydrogen yield (LH2 g VSfed) 0.026 ± 0.001 0.031 ± 0.001 0.027 ± 0.001 0.040 ± 0.002 – – – –
Methane yield (LCH4 g−1VSfed) 0.274 ± 0.008 0.269 ± 0.006 0.315 ± 0.014 0.249 ± 0.009 0.343 ± 0.018 0.389 ± 0.020 0.425 ± 0.020 0.410 ± 0.017
Prompt fermentabilitya (%) 5.2 ± 0.2 6.2 ± 0.2 5.4 ± 0.2 8.0 ± 0.3 – –
Biodegradability (%) 59 ± 1 59 ± 1 72 ± 1 58 ± 1 75 ± 1b 84 ± 1b 96 ± 2 93 ± 2
Hydrogen gain (%) +19 +48 – –
Methane gain (%) – – +13 –

a
From the theoretical BHP calculated with only prompt soluble sugars.
b
For OFMSW (thermo) when no hydrogen was produced, the biodegradability was related to a BMPth value of 0.457 and 0.463 for untreated and pre-treated,
respectively.

availability for the consequent methane production. The enhancement

of the OFMSW biodegradability, (from 75 to 84% ought to pre-treat-
ment) was strictly related to the 9% reduction of the lignocellulosic
matrix, whereas for the pre-treated FVS feedstock the reduction was
only 1.3%, not enough to increase the biomethane potential of the pre-
treated FVS sample.
ANOVA indicated positive correlation coefficients and related sta-
tistical significance (p < 0.05) for BMP and “non lignocellulosic car-
bohydrates” (i.e. the sum of starch and soluble sugars), and significant
negative correlation for BMP and each component of the lignocellulosic
matrix, namely cellulose (0.715), lignin (0.650), and hemicellulose
(0.542).
The volatile solids removal for FVS and OFMSW was always lower
than 85% because of the consistent presence of lignin and hemicellulose
hindering the anaerobic biodegradability. On the contrary, for CKW the
experimental BMP was statistically comparable with the calculated
theoretical one, suggesting the total biodegradability of this feedstock
(see Table 1).
In Fig. 4 the sum of the prompt fermentability and the biodegrad-
ability (both normalized with respect to the total VS) was compared
with the percentage of volatile solids removed during the test, re-
presenting the direct indicator of the organic material degraded during
the process. The matching of the fermentability/biodegradability value
with the VS removal, except in the case of OFMSW attributable to an
underestimation of the biodegradability because of the high “not
mapped COD”, highlights the robustness of these data and the im-
portance of an accurate feedstock characterization and organics balance
into energy.

Fig. 3. Hydrogen cumulative specific productions for FVS and OFMSW in me-
4. Conclusions
sophilic conditions, for untreated (U) and pre-treated (P) samples.

This work provides a deep chemical characterization of different

Highest total biodegradability (96 ± 2%) and highest methane
yield (0.425 LCH4 g−1VSfed) were observed for CKW, rich of starch and
lignin-free, whereas FVS and OFMSW showed significantly lower bio-
degradability, and for OFMSW biodegradability was independent on
the temperature regimen.
The effect of the thermal pre-treatment on BMP and biodegrad-
ability resulted greatly dependent on the composition and substrate
inner characteristics, and in particular was strictly linked to the lig-
nocellulosic matrix reduction. For FVS and CKW there is no statistical
difference between untreated and pre-treated yields, highlighting the
inadequacy of the pre-treatment type, if we refer to the only methani-
zation process (H2, when produced, was always favoured by the pre-
treatment). On the contrary, for OFMSW, the difference between un-
treated and pre-treated final methane yields was statistically sig-
nificant. OFMSW showed that thermophilic conditions promoted a
positive effect over mesophilic ones, even though H2 production was
Fig. 4. Comparison of fermentability (blue bars) and biodegradability (orange
only detected at 37 °C, meaning that part of the intrinsic potential
bars) of the three different feedstocks (untreated and pre-treated) with the
(translated by BMPth) was actually used for H2 production lowering its experimental VS removal (triangles).

105
P. Pagliaccia et al. Journal of Environmental Management 236 (2019) 100–107

food wastes and the related pre-treatment effect (as lignin, cellulose and methane co-production. Bioresour. Technol. 241, 491–499.
hemicellulose reduction) suggesting interesting and novel ideas as re- DuBois, M., Gilles, K.A., Hamilton, J.K., Rebers, P.A., Smith, F., 1956. Colorimetric
method for determination of sugars and related substances. Anal. Chem. 28,
gards sugars fermentability and successive biodegradability of food 350–356.
waste. Edwiges, T., Frare, L., Mayer, B., Lins, L., Triolo, J.M., Flotats, X., de Mendonça Costa,
Feedstock composition variability determines non-negligible varia- M.S.S., 2018. Influence of chemical composition on biochemical methane potential of
fruit and vegetable waste. Waste Manag. 71, 618–625.
tions in pre-treatment effects and digestion performances: a good Elbeshbishy, E., Nakhla, G., 2012. Batch anaerobic co-digestion of proteins and carbo-
characterization definitely helps in understanding the behaviour of the hydrates. Bioresour. Technol. 116, 170–178.
specific feedstock during the several treatment steps. Thermal pre- FAO, 2015. Food Wastage Footprint & Climate Change. http://www.fao.org/3/a-bb144e.
pdf.
treatment affected significantly the carbohydrates solubilisation, and Fisgativa, H., Tremier, A., Dabert, P., 2016. Characterizing the variability of food waste
was suitable in reducing slightly the lignocellulosic matrix. The high quality: a need for efficient valorisation through anaerobic digestion. Waste Manag.
bioavailable sugars content, together with the initial pH conditions, 50, 264–274.
Fisgativa, H., Tremier, A., Le Roux, S., Bureau, C., Dabert, P., 2017. Understanding the
favoured a fermentation phase and hydrogen recovery for FVS and
anaerobic biodegradability of food waste: relationship between the typological,
OFMSW in mesophilic conditions, and pre-treatment enhanced hy- biochemical and microbial characteristics. J. Environ. Manag. 188, 95–107.
drogen yields, with gains up to +50%. Nevertheless, in the single stage Ghanimeh, S.A., Al-Sanioura, D.N., Saikaly, P.E., El-Fadel, M., 2018. Correlation between
reactor, the successive methane production resulted affected by the lack system performance and bacterial composition under varied mixing intensity in
thermophilic anaerobic digestion of food waste. J. Environ. Manag. 206, 472–481.
of the soluble part of the carbohydrate fraction, “subtracted” for hy- Heo, N.H., Park, S.C., Kang, H., 2004. Effects of mixture ratio and hydraulic retention
drogen production, and no longer available for methanogenesis. Only in time on single-stage anaerobic co-digestion of food waste and waste activated sludge.
thermophilic conditions, for OFMSW, when no hydrogen was detected, J. Environ. Sci. Health 39 (7), 1739–1756.
Jiang, J., Du, X., Ng, S., Zhang, C., 2010. Comparison of atmospheric pressure effects on
pre-treatment implemented significantly methane yield. The enhance- the anaerobic digestion of municipal solid waste. Bioresour. Technol. 101 (16),
ment of the OFMSW methane yield (from 0.343 to 0.389 L CH4 g−1 6361–6367.
VSfed ought to pre-treatment) was strictly related to the 9% reduction of Jin, Y., Li, Y., Li, J., 2016. Influence of thermal pretreatment on physical and chemical
properties of kitchen waste and the efficiency of anaerobic digestion. J. Environ.
the lignocellulosic matrix, whereas for the pre-treated FVS feedstock Manag. 180, 291–300.
the reduction was only 1.3%, probably due to the more complex lig- Jördening, H.J., Winter, J., 2005. Environmental Biotechnology Concepts and
nocellulosic structure of vegetable wastes. In this sense, a pre-treatment Applications. Wiley –VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, pp. 463.
Karthikeyan, O.P., Trably, E., Mehariya, S., Bernet, N., Wong, J.W.C., Carrere, H., 2017.
seems a recommended solution in order to reduce the recalcitrant lig- Pretreatment of food waste for methane and hydrogen recovery: a review. Bioresour.
nocellulosic matrix by breaking the interlinkages making more bioa- Technol (in press). https://doi.org/10.1016/j.biortech.2017.09.105.
vailable the trapped compounds, increasing anaerobic biodegradability. Kim, J.K., Oh, B.R., Chun, Y.N., Kim, S.W., 2006. Effects of temperature and hydraulic
retention time on anaerobic digestion of food waste. J. Biosci. Bioeng. 102, 328–332.
Komilis, D.P., Ham, R.K., 2003. The effect of lignin and sugars to the aerobic decom-
Acknowledgements position of solid wastes. Waste Manag. 23 (5), 419–423.
Kondusamy, D., Kalamdhad, A.S., 2014. Pre-treatment and anaerobic digestion of food
This work was supported by a PRIN 2012 project titled “Advanced waste for high rate methane production – a review. J. Environ. Chem. Eng. 2,
1821–1830.
processes to convert organic wastes in innovative, sustainable and Lalaurette, E., Thammannagowda, S., Mohagheghi, A., Maness, P.C., Logan, B.E., 2009.
useful products” co-financed by the Italian Minister of University and Hydrogen production from cellulose in a two-stage process combining fermentation
Scientific Research (MIUR). Authors wish to thank the University of and electrohydrogenesis. Int. J. Hydrogen Energy 34, 6201–6210.
Lawther, J.M., Sun, R., Banks, W.B., 1996. Effect of steam treatment on the chemical
Roma La Sapienza which funded the Ph.D. work of Pagliaccia. composition of wheat straw. Holzforschung 50, 365–371.
Lee, M., Hidaka, T., Tsuno, H., 2008. Effect of temperature on performance and microbial
References diversity in hyperthermophilic digester system fed with kitchen garbage. Bioresour.
Technol. 99 (15), 6852–6860.
Lim, S.J., Kim, B.J., Jeong, C.M., Ahn, Y.H., Chang, H.N., 2008. Anaerobic organic acid
Al Seadi, T., Rutz, D., Janssen, R., Drosg, B., 2013. Biomass Resources for Biogas production of food waste in once-a-day feeding and drawing-off bioreactor.
Production. The Biogas Handbook. pp. 19–51. Bioresour. Technol. 99, 7866–7874.
Alibardi, L., Cossu, R., 2015. Composition variability of the organic fraction of municipal Liu, X., Wang, W., Gao, X., Zhou, Y., Shen, R., 2012. Effect of thermal pretreatment on the
solid waste and effects on hydrogen and methane production potentials. Waste physical and chemical properties of municipal biomass waste. Waste Manag. 32,
Manag. 36, 147–155. 249–255.
Alibardi, L., Cossu, R., 2016. Effects of carbohydrate, protein and lipid content of organic Loow, Y.-L., Wu, T.Y., Jahim, J.M., Mohammad, A.W., Teoh, W.H., 2016. Typical con-
waste on hydrogen production and fermentation products. Waste Manag. 47, 69–77. version of lignocellulosic biomass into reducing sugars using dilute acid hydrolysis
Angelidaki, I., Alves, M., Campos, L., Bolzonella, D., Borzacconi, L., Guwy, A.J., and alkaline pretreatment. Cellulose 23 (3), 1491–1520.
Kalyuzhnyi, S., Jenicek, P., Van Lier, J.B., 2009. Defining the biomethane potential Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurements
(BMP) of solid organic wastes and energy crops: a proposed protocol for batch assays. with the folin phenol reagent. J. Biol. Chem. 193, 265–275.
Water Sci. Technol. 59 (5), 927–934. Mata-Alvarez, J., Llabres, P., Cecchi, F., Pavan, P., 1992. Anaerobic digestion of the
APHA, 1998. Standard Methods for the Examination of Water and Wastewater, twentieth Barcelona central food market organic wastes: experimental study. Bioresour.
ed. American Public Health Association, Washington DC. Technol. 39, 39–48.
Ariunbaatar, J., Panico, A., Frunzo, L., Esposito, G., Lens, P.N., Pirozzi, F., 2014. Menon, A., Ren, F., Wang, J.Y., Giannis, A., 2016. Effect of pretreatment techniques on
Enhanced anaerobic digestion of food waste by thermal and ozonation pretreatment food waste solubilisation and biogas production during thermophilic batch anaerobic
methods. J. Environ. Manag. 146, 142–149. digestion. J. Mater. Cycles Waste Manag. 18, 222–230.
Bayard, R., Benbelkacem, H., Gourdon, R., Buffière, P., 2017. Characterization of selected Pagliaccia, P., Gallipoli, A., Gianico, A., Montecchio, D., Braguglia, C.M., 2016. Single
municipal solid waste components to estimate their biodegradability. J. Environ. stage anaerobic bioconversion of food waste in mono and co-digestion with olive
Manag (in press). https://doi.org/10.1016/j.jenvman.2017.04.087. husks: impact of thermal pretreatment on hydrogen and methane production. Int. J.
di Bitonto, L., Lopez, A., Mascolo, G., Mininni, G., Pastore, C., 2016. Efficient solvent-less Hydrogen Energy 41 (2), 905–915.
separation of lipids from municipal wet sewage scum and their sustainable conver- Pecorini, I., Baldi, F., Carnevale, E.A., Corti, A., 2016. Biochemical methane potential
sion into biodiesel. Renew. Energy 90, 55–61. tests of different autoclaved and microwaved lignocellulosic organic fractions of
Braguglia, C.M., Gagliano, M.C., Rossetti, S., 2012. High frequency ultrasound pretreat- municipal solid waste. Waste Manag. 56, 143–150.
ment for sludge anaerobic digestion: effect on floc structure and microbial popula- Perez, J., Mu, Dorado J., De-la-Rubia, T., Martz, J., 2002. Biodegradation and biological
tion. Bioresour. Technol. 110, 43–49. treatments of cellulose, hemicellulose and lignin: an overview. Int. Microbiol. 5,
Braguglia, C.M., Gallipoli, A., Gianico, A., Pagliaccia, P., 2018. Anaerobic bioconversion 53–63.
of food waste into energy: a critical review. Bioresour. Technol. 248 (part A), 37–56. Raposo, F., Fernandez-Cegri, V., De la Rubia, M.A., Borja, R., Béline, F., Cavinato, C.,
Cabbai, V., Ballico, M., Aneggi, E., Goi, D., 2013. BMP tests of source selected OFMSW to Demirer, G., Fernandez, B., Fernandez-Polanco, M., Frigon, J.C., 2011. Biochemical
evaluate anaerobic codigestion with sewage sludge. Waste Manag. 33 (7), methane potential (BMP) of solid organic substrates: evaluation of anaerobic biode-
1626–1632. gradability using data from an international interlaboratory study. J. Chem. Technol.
Carlsson, M., Lagerkvist, A., Morgan-Sagastume, F., 2012. The effects of substrate pre- Biotechnol. 86 (8), 1088–1098.
treatment on anaerobic digestion: a review. Waste Manag. 32, 1634–1650. Ren, L., Nie, Y., Liu, J., Jin, Y., Sun, L., 2006. Impact of hydrothermal process in the
Curry, N., Pillay, P., 2012. Biogas prediction and design of a food waste to energy system nutrient ingredients of restaurant garbage. J. Environ. Sci. (China) 18 (5),
for the urban environment. Renew. Energy 41, 200–209. 1012–1019.
Ding, L., Cheng, J., Qiao, D., Yue, L., Li, Y.-Y., Zhou, J., Cen, K., 2017. Investigating Romero-Cedillo, L., Poggi-Varaldo, H.M., Ponce-Noyola, T., Ríos-Leal, E., Ramos-
hydrothermal pretreatment of food waste for two-stage fermentative hydrogen and Valdivia, A.C., Cerda-García Rojas, C.M., Tapia-Ramírez, J., 2017. A review of the

106
P. Pagliaccia et al. Journal of Environmental Management 236 (2019) 100–107

potential of pretreated solids to improve gas biofuels production in the context of an Triolo, J.M., Sommer, S.G., Møller, H.B., Weisbjerg, M.R., Jiang, X.Y., 2011. A new al-
OFMSW refinery. J. Chem. Technol. Biotechnol. 92, 937–958. gorithm to characterize biodegradability of biomass during anaerobic digestion: in-
Shen, F., Yuan, H., Pang, Y., Chen, S., Zhu, B., Zou, D., Liu, Y., Ma, J., Yu, L., Li, X., 2013. fluence of lignin concentration on methane production potential. Bioresour. Technol.
Performances of anaerobic co-digestion of fruit & vegetable waste (FVW) and food 102 (20), 9395–9402.
waste (FW): single-phase vs. two-phase. Bioresour. Technol. 144, 80–85. Zhang, R.H., El-Mashad, H.M., Hartman, K., Wang, F.Y., Liu, G.Q., Choate, C., Gamble, P.,
Strong, P.J., McDonald, B., Gapes, D.J., 2010. Combined thermochemical and fermen- 2007. Characterization of food waste as feed stock for anaerobic digestion. Bioresour.
tative destruction of municipal biosolids: a comparison between thermal hydrolysis Technol. 98, 929–935.
and wet oxidative pre-treatment. Bioresour. Technol. 102 (9), 5520–5527. Zhang, C., Su, H., Baeyens, J., Tan, T., 2014. Reviewing the anaerobic digestion of food
Symons, G.E., Buswell, A.M., 1933. The methane fermentation of carbohydrates. J. Am. waste for biogas production. Renew. Sustain. Energy Rev. 38, 383–392.
Chem. Soc. 55 (5), 2018–2036. Zheng, Y., Zhao, J., Xu, F., Li, Y., 2014. Pretreatment of lignocellulosic biomass for en-
Tampio, E., Ervasti, S., Paavola, T., Heaven, S., Banks, C., Rintala, J., 2014. Anaerobic hanced biogas production. Prog. Energy Combust. Sci. 42, 35–53.
digestion of autoclaved and untreated food waste. Waste Manag. 34, 370–377.

107