LWT - Food Science and Technology 97 (2018) 45–52

Contents lists available at ScienceDirect

LWT - Food Science and Technology

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

Cellulose acetate and adsorbents supported on cellulose fiber extracted from T

waxy corn husks for improving shelf life of frying oil
Patchimaporn Udomkuna,∗, Bhundit Innawongb, Nitiporn Jumrusjumroendeeb
a
International Institute of Tropical Agriculture (IITA), Bujumbura, Burundi
b
Silpakorn University, Faculty of Engineering and Industrial Technology, Department of Food Technology, Thailand

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

Keywords: The efficacy of cellulose paper consolidated with cellulose acetate and mixed adsorbents (bentonite: activated
Cellulose acetylation clay: celite = 37.5: 50: 12.5 g with 1 g citric acid/100 g mixed adsorbents) on the physico-chemical properties of
Adsorbent combination oil was evaluated during deep-fat frying of chicken nuggets for 7 days. Cellulose fiber and cellulose acetate were
Filter paper produced from waxy corn husks. Using adsorbents supported on cellulose fiber with or without cellulose acetate
Palm oil
resulted in lower poly-unsaturated fatty acids (PUFAs), trans-fatty acid, free fatty acid (FFA), peroxide value
Fatty acid profile
(PV), and total polar materials (TPMs) compared with results from commercial filter paper (control). The L*, a*,
Agricultural waste
and b* changes of treated oils were better than in control. Limiting the polarity in cellulose acetate might
influence TPMs reduction but was more effective in overall quality improvement. The PUFAs, total trans-fatty
acid, FFA, PV, and TPMs decreased by 1.8, 7.4, 37.0, 133.3, and 20.5%, respectively when compared with
control. Therefore, the use of cellulose paper and cellulose acetate obtained from corn husks was found to
significantly lengthen the life cycle of frying oil.

1. Introduction
Deep-frying food in hot oil (150–190 °C is a method commonly used
in preparing meals, snacks, and fast food from frozen foodstuffs.
Generally, high temperatures create the desirable and unique flavor,
color, and texture of fried foods. However, frying leads to several oxi-
dative and thermal reactions resulting in the alteration of the physico-
chemical, nutritional, and organoleptic properties of the oil (Santos,
Molina-Garcia, Cunha, & Casal, 2018; Che Man & Jaswir, 2000) and
some of these byproducts such as free fatty acid (FFA), alcohols, cyclic
compounds, dimers, and polymers (Tabee, Jagerstad, & Dutta, 2009)
have been identified as crucial factors contributing to several health
risks (Seppanen & Sarri-Csallany, 2002; Romero, Bastida, & Sanchez-
Muniz, 2006, although other commonly identified compounds might
not be factors determining biological toxicity (Totani, Burenjargal,
Yawata, & Ojiri, 2008). Some crucial health problems after exposure to
deteriorated frying oil have been reported such as metabolic alterations
(Chen et al., 2008; Chiang et al., 2011), atherosclerosis (Kummerow,
2013; Xian et al., 2012), hypertension (Jaarin, Mustafa, & Leong,
2011), coronary heart disease (Kummerow, 2013), and cancer
(Srivastava et al., 2010; Chopra & Schrenk, 2011). Thus, controlling the
quality of frying oil is important for health and economic concerns.
In fried food industries, where the same food is usually fried daily in
large amounts for many hours, the frying oil is replaced periodically,
treated, or discarded to maintain quality. Nonetheless, some amounts of
used oils remain in the fryer and are adsorbed by every single fried
product. To reduce the detrimental effects and to prolong the usable life
of frying oil, several scientific studies have addressed the safety of the
used oils (Somnuk, Innawong, & Tirawattannawanich, 2013) and
claimed that regular cleaning and maintenance of equipment, good
quality frying fat, and proper frying conditions are required (Stevenson,
Vaisey-Jenser, & Eskin, 1984).
Filtration has been also used to recover deteriorated frying oil after
use, thereby extending its lifespan. Passive filtration, wherein the used
oil filtrates pass through a filter paper or cloth, is used to slow oil
breakdown by limiting the surface area (insoluble particles) available
for attachment, thereby reducing oil deterioration (Akoh & Reynolds,
2001). However, although coarse particles can be separated in this
manner, very fine particles, chemical products, and/or water remain in
the oil. There is intense interest in active filtration through the ad-
sorptive property of single or several natural and synthetic adsorbents
(Bhattacharya, Sajilata, Tiwari, & Singhal, 2008; Lin, Akoh, & Reynolds,
2001). When these adsorbents are in contact with oil, some of the polar
degradation products, such as peroxides, aldehydes, ketones, hydro-
peroxides, polymers, and oxidized monomers that occur during deep
frying, are held on the adsorbents by physical and/or chemical forces.

∗
Corresponding author.
E-mail address: P.Udomkun@cgiar.org (P. Udomkun).

https://doi.org/10.1016/j.lwt.2018.06.035
Received 13 February 2018; Received in revised form 17 June 2018; Accepted 18 June 2018
Available online 18 June 2018
0023-6438/ © 2018 Elsevier Ltd. All rights reserved.
P. Udomkun et al.
The main parameters that affect adsorption efficiency are contact time,
temperature, and the amount of adsorbent per treatment (Bheemreddy,
Chinnan, Pannu, & Reynolds, 2002; Miyagi & Nakajima, 2003; Sonkaew
& Chaisena, 2012). The most popular adsorbents for the regeneration of
frying oils are activated carbon, magnesium oxide, diatomaceous earth,
bleaching earth, calcium silicate, and various forms of silica.
Many reports are available on the effect of adsorbents in controlling
oil quality. For example, Mancini-Filho, Smith, Creveling, and Al-
Shaikh (1986) showed that the dielectric constant changes (DCC) ob-
viously improve with the utilization of bleaching clay and charcoal.
McNeill, Kakuda, and Kamel (1986) also examined the feasibility of
various mixtures between activated carbon and silica for improving the
quality of used canola oil. Oil treated with mixed adsorbents was re-
ported to show more effective reduction in acid values (AV), saturated
and unsaturated carbonyl contents, peroxide values (PV), total polar
components (TPC), and photometric color than the control (without
treatment). These results might be due to the greater capability and
higher adsorption power of mixed adsorbents to react with the chemical
compounds in the used oils. In addition, Lin et al. (2001) reported re-
ductions of final values in used oil by a mixture of synthetic chemical
absorbents (3% Hubersorb600, 2% Frypowder, and 3% Magnesol) after
daily treatment for 4 days compared to the untreated control.
The need for environmentally friendly materials and processes has
motivated much research in the use of raw materials from agricultural
waste because this represents an abundant, inexpensive, and readily
available source. Cellulose is one of the useful fibers from agricultural
waste and is widely used in many industries owing to the presence of
functional groups such as hydroxyl, carboxyl, methoxy, phenols, etc.
(Hassanein & Koumanova, 2010). Apart from cellulose fiber, cellulose
acetate is an esterified cellulose derivative widely known for its nu-
merous applications in membrane separation processes, production of
matrices for controlled drug release (membranes or particles), protec-
tion of optical films, cigarette filters, among other applications (da Silva
et al., 2017). Cellulose acetate is easy to process and biocompatible,
with the ability to change its hydrophilic/hydrophobic characteristics
through acetylation/deacetylation (Zhou et al., 2016). Hence this
polymer presents attractive properties for its use in the incorporation of
active compounds such as adsorbents.
The preliminary study showed that the quality of cellulose fiber and
cellulose acetate obtained from waxy corn husks was better than that
from sweet corn husks, rice straw, and bagasse owing to higher holo-
cellulose and alpha-cellulose contents. This indicated its potential for
extracting cellulose in the process of frying oil filtration. Therefore, this
study aimed at improving the overall quality of frying oil by using
cellulose filter paper together with cellulose acetate from waxy corn
husks and mixed adsorbents. The physico-chemical characteristics were
studied in detail to evaluate the potential of value-added products from
agricultural waste for improving the quality of used oil.
2. Materials and methods
2.1. Sample preparation
Waxy corn husks from Nakhon Pathom province, Thailand, were
boiled in water at 100 °C for 30 min and then hot air dried at 60 °C for
24 h to moisture content of 7.5 ± 0.5 g water/100 g. The dried sample
was cut into pieces of 3–5 cm prior to the steam explosion process.
2.2. Steam explosion process
The waxy corn husk sample was loaded into a 2.5 L stainless steel
reactor (model 316, Nitto Koatsu Co. Ltd., Tsukuba, Japan) and treated
with saturated steam at 215 °C and a pressure of 21 × 106 Pa for 5 min.
The steam exploded sample was separated by filtration, thoroughly
washed with water, centrifuged to eliminate fluid and then hot air dried
at 70 °C for 24 h to moisture content of 6.5 ± 0.5 g water/100 g.
46
LWT - Food Science and Technology 97 (2018) 45–52
2.3. Cellulose acetate preparation
To extract cellulose, 25 g of sample after steam explosion was first
soaked in 500 mL of 5 g/L NaOH solution then placed in a water bath at
80–90 °C for 1 h; subsequently, it was cleaned with water and hot air
dried at 55 °C for 24 h. The dried sample was ground to obtain cellulose
powder and sieved with 250 μm mesh size.
Cellulose acetate was produced by modifying the method of Biswas,
Saha, Lawton, Shagren, and Willett (2005) that involved weighing 2 g
of the cellulose powder into a 100-mL round bottomed flask with
0.5 mL acetic acid, 5 g acetic anhydride, 30 mL methylene chloride, and
0.04 mL of 18.4 mol/L concentrated sulfuric acid. The mixture was
shaken at 80 °C for 4 h using a water bath. The sample was cooled at
room temperature then filtered through Whatman No. 1 paper. The
extracted was mixed with 60 mL chloroform, stirred at room tempera-
ture for 30 min, and filtered under vacuum conditions. The collected
solution was entirely evaporated to dryness in a rotary evaporator
(model V-500, Büchi, Postfach, Switzerland). After that, 20-mL ethyl
alcohol was added to dissolve cellulose acetate before vacuum eva-
poration. The collected cellulose acetate on the filter paper was hot air
dried at 80 °C for 24 h.
2.4. Preparation of filter paper
In this study, four different adsorbents were used: celite (Celite®
coarse 545, Flukachemical, Buchs, Switzerland); activated clay (Power
Dry Co., Ltd., Bangkok, Thailand); bentonite (Sigma Aldrich Chemica
GmbH, Hamburg, Germany); and citric acid. Celite is a white, odorless
powder composed of silicon dioxide (SiO2). Activated clay is gray
granules, odorless, insoluble in water, and composed of calcium alu-
mino silicate. Bentonite is an odorless, light-brown powder composed of
silicon dioxide (SiO2) and aluminum dioxide (Al2O3). The adsorbent
combinations consisted of bentonite, activated clay, and celite at the
ratio 37.5: 50: 12.5 g with 1 g citric acid/100 g mixed adsorbents. The
selection of the types of adsorbent and their appropriate ratio was based
on the cost and efficiency of adsorbents in the preliminary study.
To prepare the cellulose filter paper for three different treatments,
cellulose extracted from waxy corn husks was mixed with distilled
water in the ratio of 1:35 (g/mL) using a blender (model DDF 341,
Moulinex, Paris, France). The control was Whatman No. 4 filter paper
without the addition of adsorbents and cellulose acetate. Treatment I
was cellulose paper with the addition of 35 g mixed adsorbents/100 g
cellulose paper. Treatment II was cellulose paper with the addition of
30 g mixed adsorbents and 5 g cellulose acetate/100 g cellulose paper.
The filter paper with a mixture of cellulose suspension with and without
adsorbents and cellulose acetate was spread on a sieve, dried at room
temperature for 24 h, removed from the sieve and prepared for testing.
The scanning electron micrographs of cellulose paper with and without
adsorbents are shown in Fig. 1.
2.5. Frying experiments
Chicken nuggets were purchased from CPF Food Products Co. Ltd.,
Lopburi, Thailand and palm oil was obtained from Morakot Industry
Co. Ltd., Bangkok, Thailand. The frying process used a commercial
dual-unit electric batch fryer (model IV, Protech Co. Ltd., Bangkok,
Thailand) with a capacity of 5 L of frying oil. Before frying, fresh oil was
preheated at 175 °C for 10 min to simulate normal frying conditions.
The oil was maintained at ± 5 °C of the set temperature (175 °C) using a
programmable temperature controller. No fresh oil was added during
frying. However, at the beginning of the next frying date, a small
amount of fresh oil (about 200 mL/day) might be added in the fryer to
keep the initial oil level constant.
Each batch of 100 g chicken nuggets was randomly fried at a setting
temperature of 175 °C in 2 L of heated oil. For each cycle, the samples
were placed in a stainless steel basket to keep them submerged and
P. Udomkun et al.
Fig. 1. Scanning electron micrographs at 200 and 1000 μm of cellulose paper (a) without adsorbents, (b) with 35 g mixed adsorbents/100 g cellulose paper, and (c)
with 30 g mixed adsorbents and 5 g cellulose acetate/100 g cellulose paper.
fried for 3 min. The fried samples were immediately withdrawn from
the oil and cooled to the ambient temperature. The temperature of oil
was kept at 175 °C for the rest of cycle even if food materials were being
unloaded. The gap between each cycle provided time for preparation of
the next batch and allowed the temperature of the fryer to build back
up. Each batch of chicken nuggets was fried at intervals of 20 min.
Each treatment was conducted for 7 days to simulate industrial
commercial frying. The fryer was switched on for 6 h/day. Sequential
daily frying of 126 batches of chicken nuggets matched the same
amount of frying load/oil used in this study.
2.6. Oil filtration and sample collection
Every day at intervals of 3 and 6 h the hot oil from the fryer pot was
collected into a container and filtered to remove food crumbs and batter
sediment under vacuum conditions with three different filter papers.
The collected samples were kept individually in closed containers with
the caps sealed off to prevent oxidation, then air-cooled in a dark room
(at approximately 7 °C) for 1 h. Thereafter, samples were kept at room
temperature until further physico-chemical analyses. The fatty acid
profile, FFA, PV, TPMs, and color (L*, a*, b*) parameters were mea-
sured with regard to expressing the rate of deterioration.
2.7. Quality analyses
2.7.1. Fatty acid profile
To determine the fatty acid profile, fatty acid methyl esters (FAMEs)
were prepared by esterification with potassium hydroxide following the
standard IUPAC method 2.301 (IUPAC, 1979) with minor modification.
According to the method, 10 g of oil sample was weighed into a tube
and dissolved in 5 mL of hexane. Then, 0.1 mL of 2 mol/L methanolic
potassium hydroxide solution was added and the tube was capped
tightly and shaken for 30 s. This solution was centrifuged at 3424 g-
force for 5 min (model Z326K, Hermle Labortechnik, Wehingen, Ger-
many). Afterwards, the upper layer of sample was transferred to a GC
47
LWT - Food Science and Technology 97 (2018) 45–52
vial prior to analysis.
The GC-MS analysis was performed using a gas chromatograph-mass
spectrometer (Agilent 6890N and G2579A mass-selective detector,
Agilent Technologies, New York, USA) equipped with Rt-2560
Biscyanopropylsiloxane capillary column (100 m × 0.25 mm
i.d. × 0.25 μm). The initial column temperature was maintained at
140 °C for 2 min upon injection, then increased at a rate 4 °C min−1 to
230 °C for 5 min. The injector and detector temperatures were 240 and
260 °C. The split ratio was 1:50 and purified helium gas (purity 99.99%,
passed through a molecular sieve and an oxygen trap) at a flow rate of
0.7 mL/min was used as the GC carrier gas. The mass spectrometer was
operated in the electron ionization (EI) mode with an electron energy of
70 eV with ion source temperature of 230 °C, analyzer temperature of
150 °C, and a mass range m/z 50–550. Peak identification of the fatty
acids was obtained by comparison of retention indices, by matching the
unknown spectra with known standards. In this study, standard methyl
esters of lauric (C12:0), myristic (C14:0), palmitic (C16:0), stearic
(C18:0), oleic (C18:1 n-9cis), linoleic (C18:2 n-9,12cis), linolenic
(C18:3 n-9,12,15cis), elaidic (C18:1 n-9tran), linolelaidic (C18:2 n-
9,12tran), and linolenelaidic (C18:3 n-9,12,15tran) acids were pur-
chased from Sigma (St. Louis, Missouri, USA) and used as GC-MS li-
braries results.
2.7.2. Free fatty acid (FFA)
The percentages of FFA were analyzed by the titration method as
described in AOCS official method Ca 5a-40 (AOCS, 1990). Briefly, 1 g
of oil sample was precisely weighed into an Erlenmeyer flask. Ten
milliliters of 95 mL/L ethanol and a phenolphthalein indicator were
added to the flask with volumetric dispensers and strongly shaken. The
mixture was then titrated against 0.1 mol/L sodium hydroxide solution
until a consistent pink color persisted for at least 30 s. Weight percen-
tage of FFA was calculated on oleic basis. The test was repeated three
times.
P. Udomkun et al.
2.7.3. Peroxide value (PV)
The PV was determined according to the standard AOCS official
method Cd 8–53 (AOCS, 1990). About 2 g of frying sample was dis-
solved in 30 mL of a mixture of 2 L chloroform/3L acetic acid and then
1 mL of freshly prepared saturated potassium iodide solution was added
to react in darkness for 3 min. Subsequently, 30 mL of distilled water
and 1–2 drops of 1 g/100 g starch solution were added and the liberated
iodine was titrated with a 0.1 mol/L sodium thiosulfate solution. The
PV was expressed as milli-equivalents of active oxygen per kg of oil
(meq O2/kg) and the test was repeated three times.
2.7.4. Total polar materials (TPMs)
The content of TPMs in oil sample was analyzed using AOCS official
method Cd 20–91 (AOCS, 1990). A glass column (length 35 cm, dia-
meter 2.1 cm) was used for chromatography and a mixture of petroleum
and diethyl ether in the ratio of 87:13 (mL/mL) was prepared as an
eluent. About 2.5 g of sample was loaded into the packed column and
the non-polar fraction was eluted. The content of TPMs (%) was cal-
culated as the mass fraction of the total polar compounds in the oil
sample. The test was repeated three times.
2.7.5. Color
A Tintometer (model PFX190, Lovibond, Amesbury, UK) was used
to determine oil color. The L*, a*, and b* values of the degraded oil
were measured. Three color readings were taken from each oil sample
and the average was used for analysis.
2.7.6. Evaluation of the efficiency of filtration
The performance of filtration treatments in improving oil quality
was determined by calculating the percentage improvement (PI) of the
treated oil over the untreated oil (Equation (1))
Value of untreated oil − Value of treated oil
PI= × 100
Value of untreated oil (1)
2.8. Statistical test
Statistical analyses were performed using General Linear Model
Program (GLM) to test the effect of treatment on the qualities of frying
oils. Least Significant Difference (LSD) was used to estimate the sig-
nificant differences among the means of each treatment at 5% of the
probability level using SAS program (SAS, 2000).
3. Results and discussion
3.1. Fatty acid composition
The main fatty acids identified in fresh frying oil were palmitic
(43.4%), oleic (38.4%), and linoleic (9.7%) acids (Table 1). Fresh oil
contained almost equal amounts of saturated fatty acid (SFA) and un-
saturated fatty acids (UFA). Oleic acid was major mono-unsaturated
fatty acid (MUFA), while linoleic and linolenic acids were the main
poly-unsaturated fatty acids (PUFA) in fresh oil. The total cis-fatty acids
contributed 48.6% and total trans-fatty acids contributed 1.37%.
A progressive change in fatty acid profiles of all treatments was
observed throughout the frying period (Table 1). In control, total SFAs
were increased (3.8%) while total UFAs were decreased (7.3%) during
frying. The decrease in MUFA and PUFAs might be attributed to the
destruction of double bonds by oxidation and polymerization (Choe &
Min, 2007), whereas SFAs are more stable (Arslan, Şapçi, Duru, & Kara,
2017). In addition, the content of total cis-fatty acids was reduced by
4.3% after 42 h of frying. The deterioration of cis-linolenic acid was
more pronounced, with its contents being decreased by 31.9%, when
compared to cis-oleic (4.1%) and cis-linoleic (3.7%). This result is si-
milar to a study of Aladedunye and Przybylski (2009) who reported that
cis-linolenic acid decreased by 47.1% and cis-linoleic acid by13.3%,
48
LWT - Food Science and Technology 97 (2018) 45–52
when fresh canola oil was heated at 215 °C for 7 days. In the case of
trans-fatty acids, an obvious increase (18.9%) was observed when
compared to their initial frying oil samples. Saguy and Dana (2003)
indicated that the exchange of fatty acids between the fried food and
the oil as well as the high temperature and prolonged frying process
may cause an increase of trans-fatty acid contents in frying oil. Chen
et al. (2014) also indicated that the formation of trans-isomer at high
temperatures is associated with a specific amount of energy required to
transfer double bonds from cis-to trans-configuration and when the
numbers of cis-double bond decrease, the activation energy for iso-
merization increases. Moreover, Normand, Eskin, and Przybylski
(2001) suggested that the ratio of cis-linoleic acid to palmitic acid can
be used to evaluate oil deterioration. In this study, a decrease in cis-
linoleic acid to palmitic acid ratio was found during frying.
As active adsorbents with or without cellulose acetate applied, total
SFAs decreased by 1.2% for Treatments I and II at the end of 42 h of
frying operations (Table 1). Total MUFAs increased by 2.0% for
Treatment I and by 2.1% for Treatment II, while total PUFAs were
decreased 1.7% for Treatment I and 1.8% for Treatment II when com-
pared to control. In addition, total cis-fatty acids were increased by 1.5
for Treatment I and 1.6% for Treatment II after 42 h of frying. Inter-
estingly, a progressive reduction of total trans-fatty acids was found in
Treatment I (6.1%) and Treatment II (7.4%). When compared to con-
trol, a slight different in the cis-linoleic acid to palmitic acid ratio of
both treatments was pronounced.
3.2. Free fatty acid formation
In all treatments FFA content gradually increased over 0–42 h of
frying (Fig. 2). This acidity is mainly formed through hydrolysis reac-
tion wherein water, in the form of vapor being released from the inside
and surfaces of chicken nuggets, consequently reacted with triglycer-
ides and formed di- and mono-acylglycerols, glycerol, and FFA. In ad-
dition, Chen, Chiu, Cheng, Hsu, and Kuo (2013) reported that the lipase
in palm fruit might decompose the oil to produce FFA. Furthermore,
Kun (1990) indicated that the change of FFA content could also be
caused by further oxidation of the secondary products formed during
frying.
In the control sample, FFA was 0.2% and reached 0.8% after 42 h of
frying. Treatments I and II significantly (P < 0.05) reduced FFA con-
tents when compared with the control. The average PI in the treatments
was 35.1% for Treatment I and 37.0% for Treatment II. It could be
deduced that the daily application of active adsorbents, with or without
cellulose acetate in frying oil, could significantly reduce the accumu-
lation of FFA. In general, active filtration with activated clay could
remove impurities by occlusion and adsorption based on their chemical
properties. Meanwhile celite is a form of silica oxide, which is capable
of adsorbing FFA compounds through an interaction with the exposed
silinol group. However, the application of cellulose acetate in
Treatment II did not significantly reduce FFA in frying oil when com-
pared to Treatment I.
3.3. The changes in peroxide value
The PV is used as an indicator of the primary level of oil oxidation
(Frankel, 2005, pp. 1–450). It is represented by the amount of hydro-
peroxides generated from FFA radicals through thermal oxidation
processes (Tyagi & Vasishtha, 1996; Che Man & Jaswir, 2000). This
ambient oxygen attacked the double bonds of the oil structure. In-
directly, it was measured by the degree of initial oxidation of fats and
oils. After 42 h of frying, the PV was 0.95 meq peroxide/kg in control.
Treatment I (81.0%) and Treatment II (133.3%) showed excellent po-
tential to reduce PV significantly when compared with passive filtration
(control) (Fig. 3). This result is in agreement with the study of
Udomkun, Innawong, Siasakul, and Okafor (2018) who reported that
mixed adsorbents could reduce the PV in soybean oil during frying of
 P. Udomkun et al.

Table 1
Changes of fatty acid profile in control and treated used oil during 42 h of frying.
Time (h) Sample Fatty acid (%)

Lauric Myristic Palmitic Stearic Oleic Elaidic Linoleic acid Linoelaidic Linolenic Linolenelaidic Total Total mono- Total poly- Total Total
acid acid acid acid acid acid (C18:2 n- acid (C18:2 n- (C18:3 n- (C18:3 n- saturated fatty unsaturated unsaturated cis-fatty trans-
(C12:0) (C14:0) (C16:0) (C18:0) (C18:1 n- (C18:1 n- 9,12cis) 9,12tran) 9,12,15cis) 9,12,15tran) acids (SFAs) fatty acids fatty acids acids fatty
9cis) 9tran) (MUFAs) (PUFAs) acids

0 Fresh 0.27 1.24 43.78 4.65 38.38 1.23 9.74(0.11) 0.13 (0.01) 0.47 (0.03) 0.01 (0.00) 49.94 (0.87) 39.61 (0.75) 10.35 (0.05) 48.59 1.37
(0.02) (0.03) (1.12) (0.05) (0.52) (0.02) (1.02) (0.08)
3 Controla 0.32 1.29 43.87 4.76 38.21 1.25 9.65 (0.03) 0.16 (0.00) 0.42 (0.02) 0.03 (0.00) 50.24 (1.02) 39.46 (0.68) 10.26 (0.04) 48.28 1.44
(0.02) (0.04) (0.56) (0.08) (0.18) (0.02) (0.82) (0.06)
Treatment Ib 0.29 1.26 43.8 4.70 38.27 1.24 9.69 (0.09) 0.14 (0.01) 0.42 (0.03) 0.02 (0.00) 50.05 (0.86) 39.51 (0.55) 10.27 (0.05) 48.38 1.40
(0.01) (0.02) (0.63) (0.03) (0.35) (0.01) (0.92) (0.02)
c
Treatment II 0.30 1.27 43.81 4.68 38.25 1.24 9.71 (0.08) 0.15 (0.02) 0.42 (0.03) 0.02 (0.00) 50.06 (1.08) 39.49 (0.86) 10.30 (0.12) 48.38 1.41
(0.02) (0.01) (1.02) (0.03) (0.18) (0.03) (0.90) (0.05)

49
21 Control 0.35 1.45 44.12 4.90 37.75 1.30 9.50 (0.05) 0.18 (0.01) 0.39 (0.01) 0.05 (0.00) 50.82(0.78) 39.05 (0.62) 10.12 (0.09) 47.64 1.53
(0.03) (0.03) (0.75) (0.04) (0.20) (0.03) (1.00) (0.05)
Treatment I 0.28 1.36 44.08 4.85 37.92 1.27 9.42 (0.04) 0.16 (0.02) 0.36 (0.02) 0.03 (0.00) 50.57 (1.05) 39.19 (0.82) 9.97 (0.16) 47.70 1.46
(0.01) (0.04) (0.81) (0.02) (0.16) (0.04) (0.98) (0.04)
Treatment II 0.29 1.38 44.1 4.84 37.98 1.28 9.44 (0.05) 0.15 (0.01) 0.37 (0.01) 0.04 (0.00) 50.61 (0.94) 39.26 (0.59) 10.00 (0.05) 47.79 1.47
(0.02) (0.04) (0.62) (0.04) (0.12) (0.02) (0.86) (0.03)
42 Control 0.38 1.57 44.78 5.12 36.80 1.35 9.38 (0.08) 0.21 (0.01) 0.32 (0.02) 0.07 (0.00) 51.85 (1.08) 38.15 (0.76) 9.98 (0.14) 46.50 1.63
(0.02) (0.05) (1.03) (0.05) (0.31) (0.03) (0.84) (0.08)
Treatment I 0.30 1.42 44.63 4.90 37.62 1.30 9.30 (0.16) 0.18 (0.02) 0.28 (0.01) 0.05 (0.00) 51.25 (1.11) 38.92 (0.66) 9.81 (0.10) 47.20 1.53
(0.01) (0.02) (0.75) (0.03) (0.18) (0.02) (0.78) (0.06)
Treatment II 0.32 1.44 44.56 4.92 37.65 1.30 9.32 (0.12) 0.17 (0.01) 0.27 (0.02) 0.04 (0.00) 51.24 (1.03) 38.95 (0.84) 9.80 (0.12) 47.24 1.51
(0.02) (0.02) (0.92) (0.04) (0.12) (0.04) (1.03) (0.04)

Numerical number in the table presented x (SD) for triplicate determination.

a
Control: The oil was filtered every 3 h through normal filter paper for 7 days.
b
Treatment I: The oil was filtered every 3 h through cellulose filter paper mixed with adsorbents for 7 days.
c
Treatment II: The oil was filtered every 3 h through cellulose filter paper mixed with adsorbents and cellulose acetate for 7 days.
LWT - Food Science and Technology 97 (2018) 45–52
P. Udomkun et al.
Fig. 2. Changes of free fatty acid (% FFA) in control and treated used oil during
42 h of frying; ● = Control, ■ = Treatment I, and ▲ = Treatment II.
Fig. 3. Changes of PV in control and treated used oil during 42 h of frying;
● = Control, ■ = Treatment I and ▲ = Treatment II.
Fig. 4. Changes of total polar materials (%TPM) in control and treated used oil
during 42 h of frying; ● = Control, ■ = Treatment I, and ▲ = Treatment II.
chicken drumsticks for 36 h.
Although the process of active filtration was applied, the PV fluc-
tuated in all treatments because peroxides are unstable organic com-
pounds formed from triglycerides, particularly under conditions with
high temperatures. The double bonds in the oil molecules were broken
and new substances formed in the fryer (Moreira, Castell-Perez, &
Barrufet, 1999, p. 350). Oil was also oxidized to form a great variety of
oxidative substances such as volatiles (aldehydes, ketones, alcohols,
some esters, and some acids) and non-volatile compounds (dimers and
polymers, polar and non-polar compounds) causing the reduction of PV
(Innawong, Mallikarjunan, Irudayaraj, & Marcy, 2004).
Fatty acids in frying oil could be oxidized to form hydro-peroxides
and then undergo an intra-molecular reaction, leading to the formation
of a cyclic compound such as cyclo-hexene. This product of degradation
50
LWT - Food Science and Technology 97 (2018) 45–52
Fig. 5. Changes of lightness (L*) in control and treated used oil during 42 h of
frying; ● = Control, ■ = Treatment I and ▲ = Treatment II.
Fig. 6. Changes of redness (a*) in control and treated used oil during 42 h of
frying: ● = Control, ■ = Treatment I and ▲ = Treatment II.
Fig. 7. Changes of yellowness (b*) in control and treated used oil during 42 h of
frying; ● = Control, ■ = Treatment I and ▲ = Treatment II.
may be further oxidized to form benzene, which in turn reacts with a C4
compound to form naphthalene and other polycyclic aromatic hydro-
carbons (Pitts, 1983). Tyagi and Vasishtha (1996) stated that higher
temperatures also had a stronger tendency to decompose the peroxides.
It is interesting that carbonyl compounds formed during oxidation
process would react with amino acids (especially asparagines), produce
acrylamide, and decrease the nutritional value and safety of foods
(Choe & Min, 2007).
3.4. An alteration in total polar materials
The presence of TPMs indicates degradation of oils and the break-
down of triglycerides, mainly resulting in the formation of dimer tri-
glycerides with a molecular weight of approximately 1800 and mono-
P. Udomkun et al.
glycerides (Billek, Cuhr, & Waibel, 1978; Paradis & Nawar, 1981).
Blumenthal (1991) referred to all oxidized and dimerized triglycerides,
FFAs, mono- and di-glycerides, sterols, carotenoids, antioxidants, anti-
foamers, crystal inhibitors, bleaching earth, filteraid, hydrogenation
catalyst residues, soaps, residues of chlorophyll, and phospholipids, and
other materials that are soluble, emulsified, or present as suspended
particulates in the frying oil as polar materials (Romano, Giordano,
Vitiello, Le Grottaglie, & Musso, 2012). TPMs are considered the most
reliable method for evaluating oil deterioration (Fritch, 1981). Many
countries, including Thailand, have set 25% as the maximum level of
polar compounds while Billek et al. (1978) stated that any heated oil
with ≥27% polar materials should be discarded.
In this study, TPMs gradually increased in all treatments over time
during 42 h of frying. The TPMs of control increased faster than in
treated oils with an initial content of 4.5% and was up to 28.5% at the
end of 42 h of frying operations (Fig. 4). The quality of treated oil was
improved in Treatment I by 22.4% and in Treatment II by 20.5%. A
reduction of TPMs after the use of active filtration of mixed adsorbents
with and without cellulose acetate process can be attributed to the polar
nature and surface area of adsorbents. Bhattacharya et al. (2008) re-
ported that binary and quaternary adsorbent treatments were better
than the use of single adsorbents in improving the quality of palm oils
used continuously for deep-fat frying of legume-based snacks for 30 and
36 h. These results might be due to the greater capability and higher
adsorption power of mixed adsorbents to react with the chemical
compounds in the used oils. Zhou et al. (2016) stated that cellulose
acetylation dramatically alters the surface characteristics of the product
by replacing the hydroxyl group (-OH) with acetyl group (-OCOCH3),
altering the strongly hydrophilic nature of the cellulose by decreasing
its polarity. This explanation might support a lower reduction of TPMs
in Treatment II when compared with Treatment I.
3.5. Color development
Color is generally used to indicate the physical quality of frying oil
and fried products (Bheemreddy et al., 2002). Changes in color po-
tentially develop not only from the dissolution of some products from
the Maillard reaction owing to the interaction between sugar and amino
acid but also from the formation of tar and darker decomposition
products with regard to oxidation, pyrolysis, and polymerization (Lin
et al., 2001; Sunisa, Worapong, Sunisa, Saowaluck, & Saowakon, 2012).
The intensity of browning by the Maillard reaction is primarily corre-
lated with the losses of lysine, histidine, and methionine. The reaction
between epoxyalkenals and proteins produces polypyrrolic polymers as
well as volatile heterocyclic compounds (Hidalgo & Zamora, 2000).
Furthermore, Moreira et al. (1999) proposed that the changed colors in
fried food could also dissolve and tend to darken the frying oil.
The change of oil color as represented by L*, a*, and b* values
during frying is shown in Fig. 5‒7. The average color of the fresh oil
was 79.8 for L*, 3.7 for a*, and 48.7 for b*. After 42 h of frying, the L*
of the control sample decreased to 68.8, while the value of a* increased
to 35.0 and of b* to 71.4. The application of mixed adsorbents with or
without cellulose acetate could improve the color of frying oil. For
example, the PI values increased by 0.7% for L*, 3.4% for a*, and 3.6%
for b* in Treatment I; whereas an improvement of 1.6% for L*, 4.8% for
a*, and 3.9% for b* was observed in Treatment II. Although most ad-
sorbents act as bleaching compounds, the color improvement was low
when compared with the control. This indicated that consolidating
mixed adsorbents into the cellulose filter paper might not be efficient
enough to bleach browning pigments. The previous study of Udomkun
et al. (2018) showed that continuously mixing similar adsorbents in the
soybean oil for 5 min before filtration could bring improvements of
53.2% for L*, 19.1% for a*, and 39.5% for b*.
51
LWT - Food Science and Technology 97 (2018) 45–52
4. Conclusions
Although there are different commercial synthetic adsorbents for
improving frying oil quality the use of some of these is not permitted in
Thailand. This study shows the possibility of using value-added by-
products from agricultural waste in passive/active filtrations for im-
proving physico-chemical qualities of frying oil. The developed cellu-
lose fiber and cellulose acetate are promising alternatives to the
availability and costly conventional methods for eco-friendly mitigation
of toxic substances in frying oil. However, complete replacement of
synthetic adsorbents by extracted natural cellulose might not be pos-
sible. Additional research is needed to understand the absorption ca-
pacity of these adsorbents, therefore the fourier-transformed infrared
spectroscopy (FTIR) analysis might be required. Continued investiga-
tion can make pilot-scale or industrial-scale implications a sustainable
reality.
Acknowledgements
The authors gratefully acknowledge IITA for the opportunity to
prepare this article.
References
Akoh, C.C., & Reynolds, A.E. (2001). Recovery of used frying oils. U.S. patent 6,187,
355 B1.
Aladedunye, F. A., & Przybylski, R. (2009). Degradation and nutritional quality changes
of oil during frying. Journal of the American Oil Chemists’ Society, 86, 149–156.
AOCS (1990). American oil chemist’ society. The official methods and recommended practices
(4th ed.). (Champaign, Illinois, USA).
Arslan, F. N., Şapçi, A. N., Duru, F., & Kara, H. (2017). A study on monitoring of frying
performance and oxidative stability of cottonseed and palm oil blends in comparison
with original oils. International Journal of Food Properties, 20(3), 704–717.
Bhattacharya, A. B., Sajilata, M. G., Tiwari, S. R., & Singhal, R. S. (2008). Regeneration of
thermally polymerized frying oils with adsorbents. Food Chemistry, 110, 562–570.
Bheemreddy, R. M., Chinnan, M. S., Pannu, K. S., & Reynolds, A. E. (2002). Active
treatment of frying oil for enhanced fry-life. Journal of the American Oil Chemists’
Society, 67(4), 1478–1484.
Billek, C., Cuhr, C., & Waibel, J. (1978). Quality assessments of used frying fats: A
comparison of four methods. Journal of the American Oil Chemists’ Society, 55,
728–733.
Biswas, A., Saha, B. C., Lawton, J. W., Shagren, R. L., & Willett, J. L. (2005). Process for
obtaining cellulose acetate from agricultural by-products. Carbohydrate
Polymers, 1–4.
Blumenthal, M. M. (1991). A new look at the chemistry and physics of deep-fat frying.
Food Technology, 45(2), 68–71 94.
Che Man, Y. B., & Jaswir, I. (2000). Effect of rosemary and sage extracts on frying per-
formance of refined, bleached and deodorized (RBD) palm olein during deep-fat
frying. Food Chemistry, 69, 301–307.
Chen, W. A., Chiu, C. P., Cheng, W. C., Hsu, C. K., & Kuo, M. I. (2013). Total polar
compounds and acid values of repeatedly used frying oils measured by standard and
rapid methods. Journal of Food and Drug Analysis, 21(1), 58–65.
Chen, J. W., Wang, S. L., Liao, P. C., Chen, H. Y., Ko, Y. C., & Lee, C. C. (2008).
Relationship between insulin sensitivity and exposure to dioxins and polychlorinated
biphenyls in pregnant women. Environmental Research, 107, 245–253.
Chen, Y., Yang, Y., Nie, S., Yang, X., Wang, Y., Yang, M., et al. (2014). The analysis of
trans fatty acid profiles in deep frying palm oil and chicken fillets with an improved
gas chromatography method. Food Control, 44, 191–197.
Chiang, Y. F., Shaw, H. M., Yang, M. F., Huang, C. Y., Hsieh, C. H., & Chao, P. M. (2011).
Dietary oxidised frying oil causes oxidative damage of pancreatic islets and impair-
ment of insulin secretion, effects associated with vitamin E deficiency. British Journal
of Nutrition, 105, 1311–1319.
Choe, E., & Min, D. B. (2007). Chemistry of deep-fat frying oils. Journal of Food Science,
72(5), R77–R86.
Chopra, M., & Schrenk, D. (2011). Dioxin toxicity, aryl hydrocarbon receptor signaling,
and apoptosis-persistent pollutants affect programmed cell death. Critical Reviews in
Toxicology, 41, 292–320.
Frankel, E. N. (2005). Lipid oxidation (2nd eds.). Bridgwater, England: The Only Press, PJ
Barnes & Associates.
Fritch, C. W. (1981). Measurements of frying fat deterioration: A brief review. Journal of
the American Oil Chemists’ Society, 58, 272–274.
Hassanein, T., & Koumanova, B. (2010). Evaluation of adsorption potential of the agri-
cultural waste wheat straw for Basic Yellow 21. Journal of the University of Chemical
Technology and Metallurgy, 45(4), 407–414.
Hidalgo, F. J., & Zamora, R. (2000). The role of lipids in non-enzymatic browning. Grasas
y Aceites, 51, 37–51.
Innawong, B., Mallikarjunan, P., Irudayaraj, J., & Marcy, J. E. (2004). The determination
of frying oil quality using Fourier transformation infrared attenuated total
P. Udomkun et al.
reflectance. LWT-Food Science and Technology, 37, 23–28.
IUPAC (1979). International union of pure and applied chemistry (6thed.). Oxford, UK:
Pergamon Press.
Jaarin, K., Mustafa, M. R., & Leong, X. F. (2011). The effects of heated vegetable oils on
blood pressure in rats. Clinics, 66, 2125–2132.
Kummerow, F. A. (2013). Interaction between sphingomyelin and oxysterols contributes
to atherosclerosis and sudden death. American Journal of Cardiovascular Disease, 3,
17–26.
Kun, T. Y. (1990). Improvements in the frying quality of vegetable oil by blending with palm
olein. Bangi, Selangor, Malaysia: Palm Oil Research Institute of Malaysia (PORIM)
Palm oil development No. 15.
Lin, S., Akoh, C. C., & Reynolds, A. E. (2001). Recovery of used frying oils with adsorbent
combinations: Refrying and frequent oil replenishment. Food Research International,
34, 159–166.
Mancini-Filho, J., Smith, L. M., Creveling, R. K., & Al-Shaikh, H. F. (1986). Effects of
selected chemical treatments on quality of fat used for deep frying. Journal of the
American Oil Chemists’ Society, 63(11), 1452–1456.
McNeill, J., Kakuda, Y., & Kamel, B. (1986). Improving the quality of Used Frying oils by
treatment with activated carbon and silica. Journal of the American Oil Chemists’
Society, 63(12), 1564–1567.
Miyagi, A., & Nakajima, M. (2003). Regeneration of used frying oils using adsorption
processing. Journal of the American Oil Chemists’ Society, 80, 91–96.
Moreira, R. G., Castell-Perez, E. M., & Barrufet, M. A. (1999). Deep-fat frying: Fundamentals
and applications. Gaithersburg, Maryland, USA: Aspen Publishers, Inc.
Normand, L., Eskin, N. A. M., & Przybylski, R. (2001). Effect of tocopherols on the frying
stability of regular and modified canola oils. Journal of the American Oil Chemists’
Society, 78(4), 369–373.
Paradis, A. J., & Nawar, W. W. (1981). Evaluation of new methods for the assessment of
used frying oils. Journal of Food Science, 46, 449–451.
Pitts, J. N. (1983). Formation and fate of gaseous and particulate mutagens and carci-
nogens in real and simulated atmosphere. Environmental Health Perspectives, 47,
115–140.
Romano, R., Giordano, A., Vitiello, S., Le Grottaglie, L., & Musso, S. S. (2012).
Comparison of the frying performance of olive oil and palm superolein. Journal of
Food Science, 77(5), C519–C531.
Romero, A., Bastida, S., & Sanchez-Muniz, F. J. (2006). Cyclic fatty acid monomer for-
mation in domestic frying of frozen foods in sunflower oil and high oleic acid sun-
flower oil without oil replenishment. Food Chemistry and Toxicology, 44, 1674–1681.
Saguy, S., & Dana, D. (2003). Integrated approach to deep fat frying: Engineering, nu-
trition, health, and consumer aspects. Journal of Food Engineering, 56(2–3), 143–152.
52
LWT - Food Science and Technology 97 (2018) 45–52
Santos, C. S. P., Molina-Garcia, L., Cunha, S. C., & Casal, S. (2018). Fried potatoes: Impact
of prolonged frying in mono-unsaturated oil. Food Chemistry, 243, 192–201.
SAS. (2000). SAS Institute Inc., Cary, NC, USA.
Seppanen, C. M., & Sarri-Csallany, A. (2002). Formation of 4-hydroxynonenal, a toxic
aldehyde, in soybean oil at frying temperature. Journal of the American Oil Chemists’
Society, 79, 1033–1038.
Srivastava, S., Singh, M., George, J., Bhui, K., Murari Saxena, A., & Shukla, Y. (2010).
Genotoxic and carcinogenic risks associated with the dietary consumption of re-
peatedly heated coconut oil. British Journal of Nutrition, 104, 1343–1352.
da Silva, F. B., de Morais Júnior, W. G., da Silva, C. V., Vieira, A. T., Batista, A. C. F., de
Faria, A. M., et al. (2017). Preparation and characterization of cellulose triacetate as
support for lecitase ultra immobilization. Molecules, 22(1930), 1–19.
Somnuk, C., Innawong, B., & Tirawattanawanich, C. (2013). Cytotoxicity of used frying
oil recovered by different adsorbents. Kasetsart Journal, 47, 874–884.
Sonkaew, R., & Chaisena, A. (2012). Regeneration of used frying palm oil by inorganic
adsorbent combinations. Asian Journal of Chemistry, 24, 677–682.
Stevenson, S. G., Vaisey-Jenser, M., & Eskin, N. A. M. (1984). Quality control in the use of
deep frying oil. Journal of the American Oil Chemists’ Society, 61, 1102–1108.
Sunisa, W., Worapong, U., Sunisa, S., Saowaluck, J., & Saowakon, W. (2012). Quality
changes of chicken frying oil as affected by frying conditions. International Food
Research Journal, 18, 615–620.
Tabee, E., Jagerstad, M., & Dutta, P. C. (2009). Frying quality characteristics of French
fries prepared in refined olive oil and palm olein. Journal of the American Oil Chemists’
Society, 86, 885–893.
Totani, N., Burenjargal, M., Yawata, M., & Ojiri, Y. (2008). Chemical properties and cy-
totoxicity of thermally oxidized oil. Journal of Oleo Science, 57, 153–160.
Tyagi, V. K., & Vasishtha, A. K. (1996). Changes in the characteristics and composition of
oils during deep-fat frying. Journal of the American Oil Chemists’ Society, 73(4),
499–506.
Udomkun, P., Innawong, B., Siasakul, C., & Okafor, C. (2018). Utilization of mixed ad-
sorbents to extend frying oil life cycle in poultry processing. Food Chemistry, 248,
225–229.
Xian, T. K., Omar, N. A., Ying, L. W., Hamzah, A., Raj, S., Jaarin, K., et al. (2012).
Reheated palm oil consumption and risk of atherosclerosis: Evidence at ultra-
structural level. Evidence-based Complementary and Alternative Medicine, 2012,
828170.
Zhou, X., Lin, X., White, K. L., Lin, S., Wu, H., Cao, S., et al. (2016). Effect of the degree of
substitution on the hydrophobicity of acetylated cellulose for production of liquid
marbles. Cellulose, 23(1), 811–821.