Energy 280 (2023) 128185

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Valorisation of bakery waste via the bioethanol pathway

M. Nikolaou, C. Stavraki, І. Bousoulas, D. Malamis, M. Loizidou, S. Mai **, E.M. Barampouti *
National Technical University of Athens, School of Chemical Engineering, Unit of Environmental Science Technology, 9 Iroon Polytechniou Str., Zographou Campus, GR-
15780, Athens, Greece

A R T I C L E I N F O A B S T R A C T

Handling Editor: Petar Sabev Varbanov Bakery waste is a stream that is generated in enormous quantities globally and thus poses a significant envi­
ronmental problem. Furthermore, the increasing use of fossil fuels necessitates the development of alternate
Keywords: energy sources. In an effort to meet the EU circularity goals, in this study, a holistic method to address these
Bread waste issues is proposed; producing bioethanol from bakery waste. Bakery waste that was collected from local bakeries
Ethanol
and cafeterias was utilized as feedstock. Two fermentation modes were studied in laboratory scale by applying
Hydrolysis
factorial design; separate hydrolysis and fermentation and simultaneous saccharification and fermentation. For
Fermentation
1
SHF the separate hydrolysis and fermentation trials, the ethanol yield reached was almost 100% for 20 μL g−starch
SSF enzyme loading and 20% solid loading at 35 ◦ C, whereas the highest ethanol yield for simultaneous sacchari­
1
fication and fermentation was 95% for 20 μL g−starch enzyme loading and 20% solid loading. Even though a small
difference in the yield was observed between the two fermentation modes, simultaneous saccharification and
fermentation is beneficial in terms of technoeconomics. Furthermore, it was established that the valorisation of
bakery waste for the production of bioethanol is technically viable, even at the pilot scale, as 100 g L− 1 of ethanol
after 31 h in a 200 L bioreactor under the optimum conditions was observed. Nevertheless, to evaluate the
process feasibility, other techno-economical factors of the entire value chain must also be taken into account.
These include fluctuations in the bakery waste composition, collection of bread residues as well as recovery and
purification of ethanol.

1. Introduction Despite the 14% average annual increase of the Net Zero Scenario,
stronger measures are more than necessary to be enforced between 2021
In recent years rapid industrialization has led to excessive con­ and 2030 [4]. In the near future, biofuels will substitute gasoline and
sumption of fossil fuels resulting in climate change that is considered as perhaps other fossil fuels as well [5].
one of the most acute problems that the planet has to face [1]. Nowa­ One of the most auspicious renewable fuels that could potentially
days, the attention is focused on the rising concern over environmental substitute fossil-based transportation fuels as an additive to gasoline is
pollution and natural resources degradation, therefore it is considered bioethanol. Due to its greater oxygen concentration, gasoline hydro­
necessary that the primary source of energy shifts from fossil fuels to carbons oxidize more effectively, resulting in lower CO2 emissions.
clean renewable energy [2]. Many countries have implemented programs in order to decrease oil
Bioenergy is a term used to describe biofuels made from a variety of imports, enhance agricultural economies and ameliorate air quality [6].
biological sources rich in organic matter such as municipal solid waste, Bioethanol mostly derives from feedstocks that contain either sucrose or
energy crops, food waste, industrial effluents, agricultural residues, starch as well as from lignocellulosic biomass through microbial
wood and forest residues, marine algae, animal manure, cereal grains fermentation [7]. Depending on the raw material, the production pro­
etc. Conversion of these potential raw materials into biofuels would be cess complexity might vary, because of the different structural charac­
an attractive option for developing alternative energy sources and teristics or chemical composition. The total world ethanol production in
thereby decreasing greenhouse gas emissions [3]. Between 2010 and 2021 was 27,310 million gallons [8], of which 60% approximately
2019, worldwide biofuel usage grew at a rate of 5% per year on average. originates from starchy materials [9].

* Corresponding author.
** Corresponding author.
E-mail addresses: mai@central.ntua.gr (S. Mai), belli@central.ntua.gr (E.M. Barampouti).

https://doi.org/10.1016/j.energy.2023.128185
Received 15 November 2022; Received in revised form 19 May 2023; Accepted 18 June 2023
Available online 20 June 2023
0360-5442/© 2023 Elsevier Ltd. All rights reserved.
M. Nikolaou et al. Energy 280 (2023) 128185

On the other hand, bakery products and waste consist of starchy pilot plant of 200 L capacity. The course of the study is illustrated in the
matter and represent a rich source of easily extractable fermentable following figure (Fig. 1).
sugars, that makes them as one of the most auspicious food waste
sources for ethanol production [10]. In order to meet the consumers’ 2.1. Raw material
demands in freshness, bakery products are manufactured in excess and
as a result a large part of them is left unsold due to staling. Therefore, Bakery waste, consisting mainly of white bread, was obtained from
bakery products have become one of the most wasted food in the local bakeries and was delivered to UEST (Unit of Environmental Sci­
developed world and pose a very significant issue in the majority of the ence and Technology), School of Chemical Engineering of the National
European countries. In fact, the production of bread is almost 100 Technical University of Athens. The obtained loafs were roughly cut into
million tons per year globally and it has been approximated that, 4–7 cm dices and then they were dried and ground by a GAIA GC-100
globally and annually, 10% of these bakery products are wasted [11]. rotary drum dryer at 60◦ C to reach a constant moisture, approxi­
Some of these wastes may be utilized as a partial flour replacement, or mately 3.3%. To comminute further and achieve an easy manageable
used as animal feed, but its usage for animal or/and human feeding raw material, a laboratory blender was used. The produced homoge­
might be perilous to consumer’s health because of mycotoxins growth. neous powder was kept in air-tight bags at room temperature until its
These issues lead bakery wastes to landfills or an interesting alternative processing.
would be that they are used as a source of energy [12]. Indeed, in
Scandinavian countries, most of the wasted bread is valorised as a 2.2. Bioethanol production
feedstock for fuel ethanol production. Specifically in Sweden, animal
feed and bioethanol are produced from dry bread. The bioethanol For bioethanol production two different processes were applied to
generated is utilized as fuel for automobiles in Denmark, England and bakery waste: Separate Hydrolysis Fermentation (SHF), Simultaneous
Germany whereas the carbon dioxide generated during ethanol Saccharification and Fermentation (SSF). At first, all experiments were
fermentation is led to the beverage industrial sector as carbonic acid conducted on lab scale in 100 mL autoclavable bottles, using a water
[13]. This indicates that valorisation of this food waste is sustainable bath shaker (Unitronic-Orbital, PSelecta). The most effective conditions
and contributes importantly to circular economy. were then applied on a 200 L pilot plant in effort of upscaling the
Even though, bread waste has been recognized as a very promising process.
waste stream for bioethanol production, very few studies have experi­
mentally demonstrated it, to the authors’ knowledge. Pietrzak and 2.2.1. Separate Hydrolysis fermentation (SHF)
Kawa-Rygielska [12] proved in lab scale that wheat-rye bread could Starch hydrolysis was performed enzymatically by adding Spirizyme
produce over 350 g ethanol per kg of bread using granular starch hy­ Excel XHS, a glucoamylase blend, formulated by alpha-amylase and
drolyzing enzyme. Similar were the experimental results of Ebrahimi cellulase and provided by the Danish Novozymes company. There is an
et al. [10] that applied at lab scale separate hydrolysis and fermentation important impact of several variables in enzymatic hydrolysis of starch,
process on flat-wheat bread producing 350 g bioethanol/kg bread. such as the type of substrate and reaction conditions, thus the operating
The aim of this study was to investigate the potential of producing
bioethanol from bakery waste by testing two distinct experimental
procedures: Separate Hydrolysis and Fermentation (SHF) and Simulta­
neous Saccharification and Fermentation (SSF). The main distinction
between the two procedures is that saccharification in SHF is conducted
at higher temperature (50–60 ◦ C), than in SSF in which saccharification
and fermentation occurs at the same time, at the temperature at which
yeast is able to ferment (35 ◦ C), which means lower energy demands.
Apart from bioprocess optimization, one of the main goals of the present
study was the upscaling of the process, progressing technology readiness
level (TRL) and reducing the risks of discrepancies for full scale appli­
cations and modeling. In light of the experimental results, an analysis of
the possibilities, technical advances as well as challenges in bioethanol
production from bakery residues is provided.

2. Materials and methods

Within this study, a bulk amount of bakery waste was collected from
local bakeries and homogenized (dried, sieved), in order to ensure a
stable feedstock for all the experimental trials. Following, a factorial
experimental design was used as an optimization tool for bioethanol
production at lab scale. The main advantage of the factorial design is
that it allows the determination of the effect of several factors and even
interactions between them with the least number of trials and the
highest degree of accuracy. The experimental trials are conducted in
accordance with the factorial design. The derived data are statistically
processed in order to reveal the operational conditions that could lead to
the highest product yields.
Two experimental procedures were applied at lab scale. The first one
involved saccharification of the carbohydrates and then fermentation of
sugars (Separation Hydrolysis and Fermentation – SHF). In the second
experimental procedure, simultaneously saccharification of sugars and
fermentation to ethanol was carried out (Simultaneous Saccharification
and Fermentation – SSF). The optimum conditions were upscaled in a Fig. 1. Stages of the experimental work.

2
M. Nikolaou et al. Energy 280 (2023) 128185

conditions and dosage of the enzyme needs investigation [14]. Table 2

The experimental trials were performed out under a range of oper­ Levels of controlling parameters in SSF procedure.
ational conditions as defined by three independent parameters (hydro­ Factors Levels
lysis temperature, enzyme loading, solid loading), while hydrolysis
Low (− ) High (+) Centre (0)
time, fermentation time, fermentation temperature and quantity of yeast
were constant. More specifically, as presented in Table 1, experiments
1
Enzyme loading (μL g−starch ) 20 60 40
1 Solid loading (% w /w) 10 20 15
with 10, 15 and 20% w/w solid loading, 20, 40 and 60 μL⋅ g−starch dosage
of enzyme and 35, 50 and 65 ◦ C of hydrolysis temperature, were
examined. After 1 h of enzymatic hydrolysis, the samples were cooled 2.4. Factorial design
down, in order to add 2% w/w of yeast Saccharomyces cerevisiae for
ethanolic fermentation at 35◦ C for 48 h. The aim of this paper was to investigate bioethanol production, in
terms of ethanol yield, scrutinizing the influence of inputs on the process
2.2.2. Simultaneous Saccharification Fermentation (SSF) outputs. This approach may be achieved by designing factorial experi­
Simultaneous Saccharification Fermentation technique (SSF) was ments [24]. In this study two factorial experiments were performed. In
examined, as it is considered high yielding and more cost – effective the SHF procedure, the examined factors were the hydrolysis tempera­
process than SHF [15]. Experiments were performed under two inde­ ture, the enzyme loading and the solid loading, while in SSF procedure
pendent variables (enzyme loading and solid loading). Enzymes load­ were the enzyme and solid loading, according to Tables 1 and 2.
1
ings of 20, 40 and 60 μL⋅g−starch was combined with solid loading of 10, 15 Mean values and standard deviations were calculated, in order to
and 20% w/w (Table 2). The SSF process was conducted at 35 ◦ C for 48 assess random errors with 95% statistical significance. Moreover,
h using 2% w/w of yeast Saccharomyces cerevisiae. Cochran criterion was applied to validate the homogeneity of fluctua­
tions. Furthermore, mathematical models were developed, showing the
2.3. Chemical analyses and calculations impact and the significance of the chosen factors to the optimization
parameters. In SHF procedure, saccharification yield and ethanol yield
The characterization of feedstocks and residues was performed in were the optimization parameters, while in SSF procedure ethanol yield
accordance with the NREL laboratory analytical protocol. After every was examined as the sole optimization parameter. Checking of devel­
process, all samples were centrifugated for 9 min at 3200 rpm (IEC/ oped mathematical model adequacy was achieved by Fisher criterion
Centra CL2) and filtered, in order to proceed for further chemical [19].
analysis. In the solid fraction, total solids, moisture, volatile solids,
water soluble solids, ash [16], lignin (acid soluble lignin and acid 3. Results and discussion
insoluble residue), hemicellulose, cellulose [17] and starch [18] were
measured. Starch was estimated by Total Starch Assay Kit, using AOAC 3.1. Chemical composition
Method 996.11 (K-TSTA-100A) [19]. In the liquid phase, glucose was
determined with a commercial kit (Biosis S.A., Greece), using the Table 3 summarizes the compositional analysis of the bakery waste
Glucose oxidase-peroxidase method (GOD/PAP), ethanol concentration used in this study on a dry basis.
was calculated by Ethanol Assay Kit, using AOAC Method 2019.08 Bakery waste consists mainly of starchy matter, which is the main
(K-ETOH 05/21) [20], total reducing sugars by the 3,5-dinitrosalicylic hydrolysis substrate for sugars extraction. In general, the composition of
acid method [21], total organic carbon (TOC) and total nitrogen (TN) bakery waste varies according to many parameters, such as the pro­
were estimated by standard methods, using SHIMADZU TOC-VCHS and duction process and the consumers preferences. According to literature,
TNM – 1 [22]. The activity of Spirizyme XHS was measured equal to there is a wide range of starch percentage values from 47.1 to 83.0% [3,
2337 U/mL [23]. 25,26]. In our study, starch concentration fell within the range of
Two yields were quantified in order to assess the efficiency of hy­ literature but the fluctuations in starch content were low. Additionally,
drolysis and fermentation. In SHF process, saccharification yield was the oil content of the feedstock studied was relatively low, possibly due
expressed as follows: to the fact that mainly bread waste was included in the feedstock. Ac­
cording to literature, the oil content may reach over 20% in bakery
Produced Glucose (g)
SG = ⋅100% (2.1) waste implying that food species such as pies or cookies are included [3,
Theoretical Glucose (g)
25,26]. Therefore, it is crucial to characterize the delivered bakery
Theoretical glucose represents the maximal glucose production if all waste, in order to determine its chemical composition, which also in­
carbohydrates were fully converted, plus free sugars. dicates its biofuel potential [27].
Ethanol yield, which is the optimization parameter of the research
was also quantified by the following equation: 3.2. Lab scale
Produced Ethanol (g)
YEtOH = ⋅100% (2.2) 3.2.1. SHF trials
Theoretical Ethanol (g)
The results of the factorial experiment for the SHF trials are pre­
Theoretical ethanol also represents the maximum ethanol production sented in Table 4 and Table 5. Firstly, glucose concentration was
by the total conversion of carbohydrates plus free sugars.
Table 3
Composition of bakery waste.
Composition (% d.b.) Experimental value
Table 1
Levels of controlling parameters in SHF procedure. Volatile Solids (VS) 96.7 ± 0.1
Oils 2.9 ± 0.1
Factors Levels Water Solubles (WS) 18.1 ± 0.1
Low (− ) High (+) Centre (0) Cellulose (CELL) 7.1 ± 1.4
Hemicellulose (HCELL) 1.2 ± 1.0
Hydrolysis temperature (◦ C) 35 65 50 Starch (ST) 62.5 ± 1.9
1
Enzyme loading (μL g−starch ) 20 60 40 Acid Soluble Lignin (ASL) 0.5 ± 0.0
Solid loading (% w /w) 10 20 15 Acid Insoluble Residue (AIL) 7.1 ± 3.1

3
M. Nikolaou et al. Energy 280 (2023) 128185

Table 4 low organic content and may be easily used as dilution water for the SHF
Characteristics of liquid phase after 1 h of hydrolysis of SHF trials. trials. The nitrogen content is also low (0.905 ± 0.350 g L− 1 ), thus it
No Conditions Liquid phase Yield shall not induce any issues during saccharification or fermentation or
after hydrolysis even during disposal in wastewater treatment plant. Indeed, it may serve
Hydrolysis Enzyme Solid Glucose SG (%) as a nutrient medium in an anaerobic system.
Temperature loading loading concentration In regard to the optimization part of this analysis, the response of two
(◦ C) 1
(μL g−starch ) (%) (g L− 1 ) parameters was examined: saccharification yield (SG ) and ethanol yield
1 35 20 10 38.5 ±0.3 48.1 ± (YEtOH ). The following equations were generated, in order to indicate the
1.2 impact of the statistically important operational conditions to the
2 35 20 20 105.8 ±0.4 58.9 ± maximum saccharification and ethanol yield.
2.1
3 65 20 10 65.4 ±0.1 81.3 ± SG = 68.9 + 15 x2 (2.3)
2.2
4 65 20 20 137.7 ±0.9 76.5 ± SG = 18.9 + THydrolysis (2.4)
1.1
5 35 60 10 33.3 ±0.6 41.4 ±
1.7 YEtOH = 91.0 (2.5)
6 35 60 20 120.9 ±1.1 67.2 ±
0.5 Eq. (2.3) refers to coded values, while eq. (2.4) refers to physical values,
7 65 60 10 69.8 ±0.8 86.7 ± where x2 and THydrolysis are the coded and physical values of hydrolysis
0.7
temperature. Therefore, hydrolysis temperature is the statistically
8 65 60 20 163.5 ±2.1 91.0 ±
0.5 important parameter for hydrolysis, whereas all the other parameters
9 50 40 15 113.8 ±1.3 89.6 ± and the interactions among them can be statistically neglected. Τhe
1.2 results presented above are related to 1 h of hydrolysis, which means
that hydrolysis time is also a factor that could be investigated further.
The fact that the enzymatic hydrolysis temperature is a statistical
measured after 1 h of hydrolysis under given conditions, in order to
important factor was anticipated, given that the experimental conditions
assess saccharification efficiency. The optimum results were obtained at
studied were not just the optimum; that is the conditions suggested by
65 ◦ C, 60 μL g− 1 of initial starch and 20% solid loading resulting for 91%
the enzyme’s manufacturer. On the other hand, ethanol yields do not
saccharification yield. After the fermentation step, complete assimila­
seem to be influenced by the control parameters within the boundaries
tion of glucose took place, providing final ethanol concentration of
76 g L− 1 which corresponds to ethanol yield of 87.6%. Nevertheless, it is
clear from Table 5, that even higher ethanol yields were achieved by Table 6
more preferable conditions (lower temperature and lower solid loading). – Degradation of solid fraction during SHF trials.
In experiment 2, the ethanol yield reached was almost 100%. Although No Conditions Degradation
different hydrolysis conditions led to lower saccharification yields after Enzyme Hydrolysis Solid Solid Starch
1 h of hydrolysis, it seems that ethanol yields are in all cases quite high. loading Temperature (◦ C) loading (% (%) (%)
This fact demonstrates that saccharification is also achieved at 35 ◦ C, 1
(μL g−starch ) )
even though the optimum operational temperature for Spirizyme Excel 1 20 35 10 78.5 ± 99.6 ±
XHS is 65 ◦ C, and even at lower enzyme dosages (see Table 6). 0.2 0.2
Similar ethanol yields were also observed by Demirci et al. who 2 20 35 20 77.3 ± 98.8 ±
optimized enzymatic hydrolysis and the maximum starch conversion to 0.5 0.3
3 20 65 10 76.6 ± 98.7 ±
glucose by waste bread was 86% [25]. Furthermore, Narisetty et al.
0.7 0.4
proved that enzymatic hydrolysis is preferable, achieving 95.9% 4 20 65 20 74.6 ± 100.0 ±
saccharification yield contrarily to acid hydrolysis (73.5% saccharifi­ 0.1 0.6
cation yield) [27]. Although bakery waste in not a thoroughly studied 5 60 35 10 78.8 ± 98.2 ±
0.9 0.1
feedstock in literature, it is obvious that the results obtained from the
6 60 35 20 77.0 ± 99.4 ±
literature as well as from the present study are very promising. 0.4 0.4
From the concentrations of TOC, it can be estimated that ethanol in 7 60 65 10 76.4 ± 97.9 ±
all cases is 92.3 ±3.8% of TOC. Τhe remaining rate may correspond to 0.3 0.5
sugars not consumed by the yeast or other metabolic products generated 8 60 65 20 75.1 ± 99.1 ±
0.9 0.8
during fermentation [28]. This indicates that almost all organic fraction
9 40 50 15 77.2 ± 99.0 ±
of bakery waste that is released in the fermentation broth is ethanol that 0.2 0.3
can be recovered via distillation. Hence, the liquid residue has a very

Table 5
Characteristics of liquid phase after fermentation of SHF trials.
No Conditions Liquid phase after fermentation Yield
1
Enzyme loading (μL g−starch ) Hydrolysis Temperature (◦ C) Solid loading (%) Ethanol concentration (g L− 1 ) TOC (g L− 1 ) TN (g L− 1 ) YEtOH (%)

1 20 35 10 37.5 ±1.2 22.2 ± 0.8 0.6 ±0.1 91.8 ±1.9

2 20 35 20 92.0 ±0.9 51.1 ±0.6 1.5 ±0.2 100.0 ±0.6
3 20 65 10 35.1 ±0.5 20.0 ±0.3 1.4 ±0.2 85.7 ±2.1
4 20 65 20 76.4 ±0.6 42.4 ±0.2 0.6 ±0.1 82.7 ±1.7
5 60 35 10 40.2 ±0.3 23.5 ±0.1 0.8 ±0.1 97.9 ±2.2
6 60 35 20 88.4 ±1.5 50.5 ±1.1 1.1 ±0.3 95.6 ±3.1
7 60 65 10 37.5 ±0.4 22.1 ±0.2 0.5 ±0.0 91.7 ±4.1
8 60 65 20 76.0 ±1.2 40.2 ±0.9 0.8 ±0.1 82.7 ±1.4
9 40 50 15 50.1 ±0.8 26.8 ±0.5 0.8 ±0.2 77.0 ±0.5

4
M. Nikolaou et al. Energy 280 (2023) 128185

examined, as all terms are not statistically significant and the predicted Table 8
value of ethanol yield is consequently equal to a constant value. This Degradation of solid fraction and ethanol yields of SSF trials.
implies that the range selected was the optimum, given the absolute No Conditions Degradation Yield
values of ethanol yield. Therefore, the center of the experimental design
Enzyme loading Solid Solid (%) Starch (%) YEtOH (%)
1
(50 ◦ C, Spirizyme 40 μL g−starch and 15% solid loading) could be consid­ 1
(μL g−starch ) loading (%)
ered the optimum conditions for achieving the maximum ethanol yield
1 20 10 79.7 ± 0.3 99.5 ± 0.2 88.7 ± 4.3
under SHF conditions. 2 20 20 75.9 ± 0.4 99.2 ± 0.2 95.0 ± 3.9
3 60 10 80.5 ± 0.4 99.5 ± 0.0 91.8 ± 8.7
3.2.2. SSF trials 4 60 20 75.7 ± 0.8 99.3 ± 0.3 93.6 ± 9.2
Aiming to investigate the effect of enzyme and solid loading, under 5 40 15 77.6 ± 0.4 99.4 ± 0.2 84.6 ± 5.6
SSF conditions, the liquid phase of fermented residues was character­
ized, and the results are summarized in Table 7. The maximum ethanol The results obtained in the present paper demonstrated that SSF
concentration (87.5 g L− 1 ), was observed with 20 μLspirizyme g−starch
1
and process could lead to comparably higher efficiencies (92.3%) than SHF
20% solid loading, which are quite favorable conditions, since the (91%) which is advantageous considering that SSF is less time
enzyme loading is low and the solids loading is high implying lower consuming, less energy demanding, and more cost efficient than SHF.
reactor volumes. The almost zero glucose concentration at the end of the Moreover, the overall conversion efficiencies indicate the potential of
experiments demonstrates glucose’s complete assimilation from bakery waste as a feedstock for large scale bioethanol production.
Saccharomyces cerevisiae. Hence, these conditions (SSF, Spirizyme 20 μL g−starch1
and 20% solid
Table 8 presents the degradation of the solid fraction along with the loading) were applied in the first scale-up step in the 200 L bioreactor.
ethanol yields resulting from the experiments. As expected, starch was
converted to glucose, which, in turn, was fully consumed achieving 99%
starch degradation. Thus, enzyme dosages seemed to be sufficient. As far 3.3. Pilot scale
as ethanol efficiencies are concerned, ethanol yields proved to be high in
all cases. The optimum results were obtained in experiment 2 where the In view of the results obtained, an experiment was conducted on pilot
high solid loading and the low enzyme loading were applied, resulting in scale (14 kg bakery waste) under the optimum SSF conditions (20% solid
95% ethanol yield. 1
loading, enzyme loading 20 μL g−starch , 2% w/w Saccharomyces cerevisiae)
According to Pietrzak and Kawa – Rygielska [12], the fermentation at 35 C for 48 h. The experiment was conducted in a 200 L bioreactor,
◦

of untreated bread waste resulted in 80% ethanol yield. Under SSF which is 2000 times bigger than lab scale bottles. Аt regular time in­
conditions, higher ethanol efficiencies (96% ethanol yield) have been tervals a sample was collected to measure glucose and ethanol in order
reported in literature by Pietrzak and Kawa – Rygielska [29], by use of to study the kinetics of the reaction and to determine the maximum
granular starch hydrolyzing enzyme (GSHE). In terms of microorgan­ ethanol concentration. A glucose peak in the first few hours is observed.
isms, Saccharomyces cerevisiae has been the most widely utilized yeast The glucose is completely consumed by the yeast which is responsible
for alcoholic fermentation. However, Kawa – Rygielska et al. [30] sup­ for ethanol production in the following hours. The highest ethanol
port that 82% ethanol yield was also achieved by other fungi, such us concentration observed was 100 g L− 1 after 31 h (Fig. 1), which con­
Aspergillus oryzae and Neurospora intermedia. firms the lab scale results. As shown in Fig. 1 there is a drop in ethanol
As in SHF, it was estimated that bioethanol in all cases was 91.5 ± concentration after 31 h. This is attributed to yeast stress. Due to the
8.1% of TOC, indicating once more that almost all organic fraction of high ethanol concentration and almost full assimilation of glucose and
bakery waste that is released in the fermentation broth is ethanol that nutrients, yeast starts consuming ethanol. The accumulation of alcohol
can be recovered. Thus, the liquid residue may be easily utilized as is toxic, thereby killing the yeast cells [31]. The concentration achieved
dilution water for the SSF trials as well. The nitrogen content is also low at pilot scale was even higher than that of lab scale (100 versus 87.5 g
(0.900 ± 0.274 g L− 1 ). L− 1). This fact could be due to the better stirring of the mixture (me­
The results showed that similar ethanol yields were obtained in each chanical stirring versus magnetic stirrer). Additionally, it is obvious
experiment, indicating the statistically insignificant effect of solid from the Figure for the pilot experiment that at 48 h the ethanol con­
loading and enzyme quantity on ethanol yield within the examined centration had decreased after the maximum peak at 31 h. From Fig. 2,
boundaries. As far as the process optimization is concerned and in line the rate of bioethanol production can be calculated equal to 2.94 g L− 1
with the results of ethanol yield from the factorial experiment presented h− 1 and there is also a lag phase of about 1 h that can be attributed to the
above, Eq. (2.5) was generated. time needed for the amylolytic enzyme to act. This bioethanol produc­
tion rate is similar to those reported by Moodley and Kana [32] for
YEtOH = 92.3 (2.6)
bioethanol produced from sugarcane leaf waste (2.44 and 2.85 g L− 1
Similarly as in SHF, this implies that the range of operational pa­ h− 1) but higher than those reported by Srimachai et al. [33] who studied
rameters selected was the optimum, given the absolute values of ethanol ethanol production from oil palm frond juice (0.24 g L− 1 h− 1). However,
yield. Therefore, the center of the experimental design (Spirizyme 40 in this case the lag phase was much shorter (0.12 h).
1
μL g−starch and 15% solid loading) could be considered the optimum con­ Similarly, according to Sudiyani et al. the yield of ethanol production
ditions for achieving the maximum ethanol yield under SSF conditions. by food waste, not bakery waste, in a 300 L reactor reached 69% under
SSF conditions [34]. Yan et al. also claimed that very high bioethanol

Table 7
Characteristics of liquid phase of fermentation residues from SSF trials.
No Conditions Liquid phase

Enzyme loading (μL 1

g−starch ) Solid loading (%) Ethanol concentration (g L− 1 ) Glucose concentration (g L− 1 ) TOC (g L− 1 ) TN (g L− 1 )

1 20 10 36.3 ± 1.8 0.1 ± 0.1 22.7 ± 0.2 0.6 ± 0.0

2 20 20 87.5 ± 3.5 1.6 ± 2.1 45.9 ± 1.4 1.0 ± 0.1
3 60 10 37.5 ± 8.5 0.1 ± 0.0 23.4 ± 0.3 0.9 ± 0.0
4 60 20 86.0 ± 3.5 1.6 ± 1.9 49.2 ± 0.6 1.3 ± 0.2
5 40 15 55.0 ± 3.5 0.0 ± 0.0 25.3 ± 11.0 0.7 ± 0.3

5
M. Nikolaou et al. Energy 280 (2023) 128185

Fig. 2. The profile of glucose consumption and ethanol production of bakery waste (14 kg dry matter) on pilot scale under SSF conditions (20 % solid loading,
enzyme loading 20 μL g − 1 of initial starch at 35 ◦ C for 48 h).

production using pilot scale (90–94% ethanol yield) facilities were an additional added value. The EU’s promotion of the zero-waste idea is
achieved by the utilization of sugars obtained from enzymatic sacchar­ also further encouraged.
ification of kitchen waste [35].
After 48 h, several samples were taken to characterize the solid Credit author statement
fraction of the experiment. 99% degradation of starch and 75% degra­
dation of solid was achieved, which is comparable with the results ob­ M. Nikolaou: Investigation, C. Stavraki: Investigation, Data curation,
tained during the lab scale trials. Writing - original draft, І. Bousoulas: Investigation, D. Malamis: Project
Thus, if the total amount of bakery waste produced globally was administration, М. Loizidou: Funding acquisition, Resources, S. Mai:
valorised, around 5 million m3 of bioethanol could be produced. This Conceptualization, Methodology, Supervision, Visualization, Writing -
seems as a huge amount of ethanol but of course, the full valorisation of review & editing, Supervision, E.M. Barampouti: Conceptualization,
bakery waste is not feasible, given that the waste is spread over a vast Methodology, Supervision, Visualization, Writing - review & editing,
geographic area. Nevertheless, smaller valorisation units able to collect Supervision.
smaller volumes of bread waste may still be viable. The economics of the
treatment train should also be examined, prior to determining the Declaration of competing interest
viability of the process.
The authors declare that they have no known competing financial
4. Conclusions and perspectives interests or personal relationships that could have appeared to influence
the work reported in this paper.
In this paper, the use of two fermentation modes, SHF and SSF, for
the production of bioethanol from bakery waste were initially studied at Data availability
laboratory scale. Even at temperatures below the optimal for the
amylolytic enzyme activity, such as 35 ◦ C, the enzyme’s performance Data will be made available on request.
was shown to be effective. SSF seems to be a more viable fermentation
mode for producing bioethanol from bread waste. The optimum condi­ Acknowledgements
tions were further upscaled at pilot scale (200 L capacity), achieving
almost 95% ethanol yield and ethanol concentrations up to 100 g L− 1. This work was supported by the LIFE CIRCforBIO (LIFE Ref. No:
The results of this study may thereby encourage the production of LIFE18 CCM/GR/001180) project.
ethanol from bakery waste going beyond the state-of-the-art. Such high
ethanol yields accompanied by high ethanol concentration are very References
promising for full scale implementation. They could also stand as a
roadmap for future research aimed at determining the upscaling pro­ [1] Zhao J, Shahbaz M, Dong X, Dong K. How does financial risk affect global CO2
emissions? The role of technological innovation. Technol Forecast Soc Change
cedure for bioethanol production. An alternative treatment train for
2021:168. https://doi.org/10.1016/j.techfore.2021.120751.
effectively processing bakery waste is presented to important stake­ [2] Saeed Meo M, Karim MZA. The role of green finance in reducing CO2 emissions: an
holders and interested parties in order to lessen the environmental empirical analysis. Borsa Istanbul Rev 2022;22:169–78. https://doi.org/10.1016/j.
impact and maximize the financial value from their production line. bir.2021.03.002.
[3] Li J, Zhao R, Xu Y, Wu X, Bean SR, Wang D. Fuel ethanol production from starchy
Thus, the viability of bioethanol produced from food waste depends on grain and other crops: an overview on feedstocks, affecting factors, and technical
an efficient waste management system. advances. Renew Energy 2022;188:223–39. https://doi.org/10.1016/j.
In conclusion, the findings of this study suggest that bakery waste renene.2022.02.038.
[4] Hodgson D, Bains P, Moorhouse. Bioenergy. Paris: International Energy Agency
may be an appropriate source of energy, fulfilling the goals of the cir­ (IEA); 2022. Available online: https://www.iea.org/reports/bioenergy. [Accessed
cular economy. To overcome the significant obstacles of upscaling and 15 November 2022].
market penetration, further technoeconomic analysis of the suggested [5] Nanda S, Rana R, Sarangi PK, Dalai AK, Kozinski JA. A broad introduction to first-,
second-, and third-generation biofuels. Recent Adv. Biofuels Bioenergy Util. 2018:
schemes must be conducted. Given that ethanol is produced from a 1–25. https://doi.org/10.1007/978-981-13-1307-3_1/TABLES/1.
waste stream, it is regarded as an advanced biofuel. Consequently, it has [6] Sánchez ÓJ, Cardona CA. Trends in biotechnological production of fuel ethanol
from different feedstocks. Bioresour Technol 2008;99:5270–95.

6
M. Nikolaou et al. Energy 280 (2023) 128185

[7] Arpia AA, Chen WH, Lam SS, Rousset P, de Luna MDG. Sustainable biofuel and [22] Baird RB, Eaton AD, Rice EW, editors. Standard methods for the examination of
bioenergy production from biomass waste residues using microwave-assisted water and wastewater. Am. Public work. Assoc. 23th ed., vol. 1469. Washington DC:
heating: a comprehensive review. Chem. Eng. J. 2021;403:126233. https://doi. APHA Press; 2017.
org/10.1016/j.cej.2020.126233. [23] Xiao Z, Storms R, Tsang A. A quantitative starch-iodine method for measuring
[8] Annual ethanol production Available online: https://ethanolrfa.org/markets-and- alpha-amylase and glucoamylase activities. Anal Biochem 2006;351:146–8.
statistics/annual-ethanol-production (accessed on 4 November 2022). https://doi.org/10.1016/j.ab.2006.01.036.
[9] Friedl A. Bioethanol from sugar and starch. Energy from Org. Mater. 2019:905–24. [24] Dweiri F, Khan SA, Khattak MNK, Saeed M, Zeyad M, Mashaly R, Hamad S.
https://doi.org/10.1007/978-1-4939-7813-7_432. Environment and sustainability approach to manage sweet bakery waste product.
[10] Ebrahimi F, Khanahmadi M, Roodpeyma S, Taherzadeh MJ. Ethanol production Sci Total Environ 2021:772. https://doi.org/10.1016/j.scitotenv.2021.145557.
from bread residues. Biomass Bioenergy 2008;32:333–7. https://doi.org/10.1016/ [25] Sükrü Demirci A, Palabıyık I, Gümüs T, Özalp Ş. Waste bread as a biomass source:
J.BIOMBIOE.2007.10.007. optimization of enzymatic hydrolysis and relation between rheological behavior
[11] Narisetty V, Cox R, Willoughby N, Aktas E, Tiwari B, Singh Matharu A, Salonitis K, and glucose yield. Waste and Biomass Valorization 2017;8:775–82. https://doi.
Kumar V. Recycling bread waste into chemical building using a circular biorefining org/10.1007/S12649-016-9601-6/TABLES/2.
approach. Sustain Energy Fuels 2021;5:4842–9. https://doi.org/10.1039/ [26] Haque MA, Kachrimanidou V, Koutinas A, Lin CSK. Valorization of bakery waste
d1se00575h. for biocolorant and enzyme production by monascus purpureus. J Biotechnol 2016;
[12] Pietrzak W, Kawa-Rygielska J. Ethanol fermentation of waste bread using granular 231:55–64. https://doi.org/10.1016/j.jbiotec.2016.05.003.
starch hydrolyzing enzyme: effect of raw material pretreatment. Fuel 2014;134: [27] Narisetty V, Nagarajan S, Gadkari S, Ranade VV, Zhang J, Patchigolla K,
250–6. https://doi.org/10.1016/j.fuel.2014.05.081. Bhatnagar A, Kumar Awasthi M, Pandey A, Kumar V. Process optimization for
[13] Ketola A, Shishkova V, Kainulainen K, Kaarnisto A. AIM2Flourish | turning food recycling of bread waste into bioethanol and biomethane: a circular economy
waste into a better future. Available online: https://aim2flourish.com/innovation approach. Energy Convers Manag 2022:266. https://doi.org/10.1016/j.
s/turning-food-waste-into-a-better-future. [Accessed 14 November 2022]. enconman.2022.115784.
[14] Ferreira S, Duarte AP, Ribeiro MHL, Queiroz JA, Domingues FC. Response surface [28] Guadalupe-Daqui M, Goodrich-Schneider RM, Sarnoski PJ, Carriglio JC, Sims CA,
optimization of enzymatic hydrolysis of cistus ladanifer and cytisus striatus for Pearson BJ, MacIntosh AJ. The effect of CO2 concentration on yeast fermentation:
bioethanol production. Biochem Eng J 2009;45:192–200. https://doi.org/ rates, metabolic products, and yeast stress indicators. J Ind Microbiol Biotechnol
10.1016/J.BEJ.2009.03.012. 2023;50:1. https://doi.org/10.1093/JIMB/KUAD001.
[15] Kádár Z, Szengyel Z, Réczey K. Simultaneous saccharification and fermentation [29] Pietrzak W, Kawa-Rygielska J. Simultaneous saccharification and ethanol
(SSF) of industrial wastes for the production of ethanol. Ind. Crops Prod. 2004;20: fermentation of waste wheat-rye bread at very high solids loading: effect of
103–10. https://doi.org/10.1016/J.INDCROP.2003.12.015. enzymatic liquefaction conditions. Fuel 2015;147:236–42. https://doi.org/
[16] Sluiter A, Hames B, Hyman D, Payne C, Ruiz R, Scarlata C, Sluiter J, Templeton D, 10.1016/j.fuel.2015.01.057.
Wolfe J. Determination of total solids in biomass and total dissolved solids in liquid [30] Kawa-Rygielska J, Pietrzak W, Lennartsson PR. High-efficiency conversion of bread
process samples laboratory analytical procedure (LAP) issue date: 3/31/2008. residues to ethanol and edible biomass using filamentous fungi at high solids
2008. loading: a biorefinery approach. Appl Sci 2022:12. https://doi.org/10.3390/
[17] Sluiter A, Hames B, Ruiz R, Scarlata C, Sluiter J, Templeton D, Crocker D. app12136405.
Determination of structural carbohydrates and lignin in biomass: laboratory [31] Walker GM, Basso TO. Mitigating stress in industrial yeasts. Fungal Biol 2020;124:
analytical procedure (LAP); issue date: April 2008; Revision date: July 2011. 2011. 387–97. https://doi.org/10.1016/J.FUNBIO.2019.10.010.
Version 07-08-2011). [32] Moodley P, Gueguim Kana EB. Bioethanol production from sugarcane leaf waste:
[18] Michel K, Sluiter J, Payne C, Ness R, Thornton B, Reed M, Schwartz A, Wolfrum E. effect of various optimized pretreatments and fermentation conditions on process
Determination of cellulosic glucan content in starch containing feedstocks. kinetics. Biotechnol. Reports 2019;22. https://doi.org/10.1016/j.btre.2019.
Laboratory Analytical Procedure (LAP) February 26, 2021;2021. Issue Date:. e00329.
[19] Moodley P, Gueguim Kana EB. Bioethanol production from sugarcane leaf waste: [33] Srimachai T, Nuithitikul K, O-Thong S, Kongjan P, Panpong K. Optimization and
effect of various optimized pretreatments and fermentation conditions on process kinetic modeling of ethanol production from oil palm frond juice in batch
kinetics. Biotechnol. Rep. (Amst). 2019;22:e00329. https://doi.org/10.1016/j. fermentation. In: Proceedings of the energy procedia. vol. 79; 2015. p. 111–8.
btre.2019.e00329. [34] Sudiyani Y, Styarini D, Triwahyuni E, Sudiyarmanto, Sembiring KC, Aristiawan Y,
[20] Felekis V, Stavraki C, Malamis D, Mai S, Barampouti EM. Optimisation of Abimanyu H, Han MH. Utilization of biomass waste empty fruit bunch fiber of
bioethanol production in a potato processing industry. Fermentation 2023;9(2). palm oil for bioethanol production using pilot - scale unit. Energy Proc 2013;32:
https://doi.org/10.3390/fermentation9020103. 31–8. https://doi.org/10.1016/J.EGYPRO.2013.05.005.
[21] Miller GL. Use of dinitrosalicylic acid reagent for determination of reducing sugar. [35] Yan S, Chen X, Wu J, Wang P. Pilot-scale production of fuel ethanol from
Anal Chem 1959;31:426–8. https://doi.org/10.1021/ac60147a030. concentrated food waste hydrolysates using Saccharomyces cerevisiae H058.
Bioproc Biosyst Eng 2013;36:937–46. https://doi.org/10.1007/s00449-012-0827-
9.