Measurement: Food 7 (2022) 100035

Contents lists available at ScienceDirect

Measurement: Food
journal homepage: www.elsevier.com/locate/meafoo

Quality evaluation of different repeatedly heated vegetable oils for

deep-frying of yam fries
Daniel Dodoo a,∗, Francis Adjei a, Samuel Kofi Tulashie a, Kingsley Enoch Adukpoh b,
Redeemer Kofi Agbolegbe c, Kokoutse Gawou a, Gloria Pokuaa Manu d
a
Industrial Chemistry Section, Department of Chemistry, School of Physical Sciences, College of Agriculture and Natural Sciences, University of Cape Coast, Cape Coast,
Central, Ghana
b
Department of Chemistry, Faculty of Physical and Computational Science, College of Science, Kwame Nkrumah University of Science and Technology, Kumasi,
Ashanti Region, Ghana
c
Department of Food Science and Technology, Faculty of Biosciences, College of Science, Kwame Nkrumah University of Science and Technology, Kumasi,
Ashanti Region, Ghana
d
Department of Material Science and Engineering, School of Engineering Sciences, College of Basic and Applied Sciences, University of Ghana, Accra, Ghana

a r t i c l e i n f o a b s t r a c t

Keywords: Many studies have been conducted to evaluate the quality of repeatedly heated oils (RHO) used in deep-frying
Repeatedly heated oils potato fries, but little is known about the quality of RHO used for deep-frying yam fries. Therefore, this study
Vegetable oils investigated the quality of RHO used for deep-frying yam fries, focusing on coconut oil (CNO), sunflower oil (SFO),
Thermal oxidation
and soybean oil (SBO). The oils’ quality was determined by changes in moisture content, acid value (AV), free fatty
Thermal hydrolysis
acid (FFA), iodine value (IV), peroxide value (PV), and total polar compounds (TPC). The oils’ degradation and the
Thermal degradation
presence of primary and secondary oxidation products were evaluated with thermogravimetric analysis (TGA) and
Fourier transform infrared spectroscopy (FTIR), respectively. The TGA curves indicated that the thermal stability
of RHO followed the trend: RSFO (217 °C) > RSBO (187 °C) > RCNO (158 °C). RHO recorded a sharp increase
in AV, FFA, PV, and TPC. The presence of secondary oxidation compounds was registered at a peak intensity of
around 1742 cm−1 using the FTIR technique. A significant decrease (p < 0.05) in IV, PUFAs, 𝛽-carotenes, and
color content were registered, rendering RHO unsuitable for deep-frying of yam fries. Finally, the RHO quality
decreased in the following order: RSFO > RSBO > RCNO, implying that oils with a high degree of saturation are
less susceptible to thermal changes during deep-frying.

1. Introduction The reuse of RHO has been a regular practice in restaurants, industries,
and households with little public awareness of its consequences. Several
Deep-frying is a quick and easy cooking method that involves im- studies have been conducted around the world in which oil consumers
mersing food in oil and heating it above 140°C in the presence of mois- and vendors store RHO for future use [4,5]. According to a pilot sur-
ture and air [1]. During deep-frying, the oil acts as a heat transfer vey conducted in Kuala Lumpur (Malaysia, Asia), approximately 63% of
medium, causing water from the food to be displaced to absorb the deep-fried food vendors nightly store RHO for use on consecutive days
oil [1,2]. Thus, simultaneous heat and mass transfer arise whereby mi- [6]. Reusing RHO has been reported globally to be linked with various
grations between the oil and water occur to produce the desirable and health issues, but discarding these oils after a few deep-frying cycles is
unique quality of the fried food [1,3]. As a result, the quality of the deemed inconvenient from an economic and marketing standpoint [7].
oil used for deep-frying has a significant impact on the texture, flavor, Even so, recent studies have shown that consuming RHO-fried foods is
color, aroma, and nutritional quality of the fried food [3]. This rein- detrimental to human health [5,8].
forces that vegetable oils (VOs) used for deep-frying must be in the best The reuse of RHO accelerates the spoilage of the oils, which wors-
state to produce healthy and nutritious fried foods. However, reusing re- ens to the point where their quality is significantly reduced. As a re-
peatedly heated vegetable oils (RHO) has proven to be an efficient way sult, the food’s initial nutritional components, flavor, color, and tex-
of reducing the amount of oil used in deep-frying. This saves money ture are lost, posing a risk to human health. This is due to thermo-
on buying new oils for each deep-frying cycle during food preparation. chemical reactions such as hydrolysis, oxidation, polymerization, and
degradation that emerge during deep-frying [2]. Thermal hydrolysis re-

∗
Corresponding author.
E-mail address: [email protected] (D. Dodoo).

https://doi.org/10.1016/j.meafoo.2022.100035
Received 9 February 2022; Received in revised form 24 April 2022; Accepted 26 April 2022
2772-2759/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)
D. Dodoo, F. Adjei, S.K. Tulashie et al. Measurement: Food 7 (2022) 100035

sults in hydrolytic compounds (e.g., fatty acids and glycerols); while the
presence of air and moisture results in the isomerization of fatty acids
that accelerate the oil’s oxidation, producing thermally oxidized com-
pounds (e.g., hydroperoxides, aldehydes, etc.), polar and polymerized
lipid compounds in the oil [9]. Similarly, thermal degradation of impor-
tant chemical constituents in oils (polyunsaturated fatty acids (PUFAs),
color pigments, tocopherols, vitamins, etc.) has been observed during
deep-frying [1]. The physical and chemical quality of the oils deterio-
rates due to the thermal changes, and toxic compounds form in fried
foods [1,2]. Their consumption may increase the risk of cardiovascular
diseases, high blood pressure, inflammation of the liver and kidney, and
atherosclerosis [5,8].
There has been limited research on the thermal hydrolysis, thermal
oxidation, thermal degradation, and thermal stability of RHO used for Fig. 1. Image of West African white yam, Dioscorea rotundata.
deep-frying yam (Dioscorea) fries. The quality of RHO used in deep-
frying potato fries (Solanum tuberosum) has been the subject of numer-
ous studies [3,10,11]. However, using yams in the production of fries and sealed box at a room temperature of 25°C away from sunlight. Stan-
can (a) improve its global distribution network, (b) reduce the heavy dard chlorophyll a was acquired from Sigma Chemical Company (St.
reliance on potatoes, and (c) contribute significantly to food security, Louis, MO, USA).
particularly in Africa, by providing income through exportation and uti-
lization in countries with fewer potato reserves [12,13]. Yams contain
2.2. Method
a variety of nutritional and non-nutritional compounds that have been
shown to have numerous health benefits [12,13]. As its usage for fries’
2.2.1. Deep-frying process
production expands globally, the quality of the oil used for deep-frying
An aliquot of each vegetable oil type (500 mL) was separated and do-
needs to be evaluated to ensure yam fries are of the highest quality.
nated as the fresh vegetable oil (FVO), while another aliquot (2000 mL)
Such studies are of great interest not only to policymakers and regula-
was used for the deep-frying of yam fries. 2000 mL of each oil (CNO;
tors, nutritionists, food processors, food scientists, and technologists; but
SBO; SFO) was heated in a separate stainless-steel wok placed on a gas
also for consumers to be aware of the thermochemical changes that oc-
stove (ATAG Heating Technology UK Ltd, Kent, UK) by regulating the
cur when RHO is used in deep-frying yam fries. Data obtained from the
knob to about 175 ± 5 °C for 10 min. 12 pieces of yam fries, soaked
quality evaluation of yam fries will be useful in the ongoing and future
in 200 mL of filtered drinking water and dried for 15 min, were deep-
research on food waste conversion into valuable and sustainable prod-
fried in the heated oil for 15 min. The yam fries were rotated occasion-
ucts [14,15]. This could include using it to enhance the production of
ally during the deep-frying process to ensure that all sides were evenly
biolubricants and biodiesels, as road oil to control dust, and as a source
fried and presented a pleasing appearance, color, texture, and aroma.
of nutrients for microalgae cultivation in the production of bioplastics
After 15 min, the yam fries were deep-fried and transferred to a sieve to
[14,16,17].
drain any remaining oil adhered to the yam fries. A sample of 500 mL
Along these lines, this research’s objective was to evaluate the quality
of the heated VOs was set aside and labeled as singly heated vegetable
of RHO used for deep-frying yam fries. The moisture content, acid value
oil (SHO), while the remaining heated oil in the stainless-steel wok was
(AV), free fatty acid (FFA), peroxide value (PV), iodine value (IV), and
used to deep-fry another batch of yam fries. Thus, the SHO was used in
total polar compounds (TPC) were investigated. Likewise, the thermal
deep-frying several batches of fresh yam fries three times to obtain a re-
degradation of the fatty acid compositions was evaluated with gas chro-
peatedly heated oil (RHO) with a cooling interval of at least 30 min. No
matography (GC). The color and 𝛽-carotene content were measured to
fresh oil was added during each frying cycle to compensate for the loss
study their thermal degradation during deep-frying. The Fourier trans-
caused by the yam fries’ uptake of the oil. The deep-frying treatment
form infrared spectroscopy (FTIR) technique was used to identify ther-
was triplicated, and the RHO obtained was stored in dark glass bottles
mally oxidized compounds present in different quantities depending on
for further analysis and characterization. The entire deep-frying of the
the type of oil. The thermal stability and degradation of the RHO were
yam fries is depicted in Fig. 2.
evaluated with thermogravimetric analysis (TGA). Finally, coconut oil
(CNO), sunflower oil (SFO), and soybean oil (SBO) susceptibility to ther-
mochemical changes was evaluated due to their different fatty acid com- 2.3. Physicochemical measurements
positions.
The AV and FFA were evaluated according to the ISO 660:2009 pro-
2. Methodology tocol [18] and expressed in percentages (%). The moisture content and
total polar compounds were evaluated following the Nielsen [19] and
2.1. Materials and storage Schulte [20] protocols and expressed in percentages (%), respectively.
The peroxide value (PV) was analyzed to determine the degree of lipid
The yams, commonly known as the West African white yam, oxidation. The PV was titrimetrically evaluated following the American
Dioscorea rotundata (Fig. 1), were purchased at the Mallam local super- Oil Chemists’ Society (AOCS) Official Method (Cd 8b–90) [21], and the
market (Accra, Ghana). Refined coconut oil (CNO), soybean oil (SBO), recorded values were in the unit of meq/kg. The Iodine value (IV) was
and sunflower oil (SFO) were purchased at a local supermarket in used to evaluate the degree of unsaturation. The IV was analyzed fol-
Kaneshie (Accra, Ghana). The oils were acquired in 1000 mL transparent lowing the AOCS (Cd 1–85) protocol [22], and the unit was given in
polypropylene (PP) plastic bottles, with a total of 5 bottles per oil. The grams of iodine/100g of oil.
retailers stated that the oils were produced fresh under hygienic condi-
tions, without blending with other oils or adding antioxidants. Accord- 2.4. Color and 𝛽-carotene content measurements
ing to the oil manufacturers, their lifespan was 2 to 4 years. To prevent
oxygen from entering, the oils were immediately wrapped in aluminum The color content was determined using a CR-400 Chroma Meter
foil, tightly capped, and sealed with plastic parafilm. To avoid prema- (Minolta Co., Osaka, Japan), which included a measuring head (CR-
ture photochemical reactions, they were quickly stored in a well-secured 400) with an 8 mm diameter measuring area, two standard observers,

2
D. Dodoo, F. Adjei, S.K. Tulashie et al. Measurement: Food 7 (2022) 100035

Fig. 2. A schematic illustration of the deep-frying of yam fries.

and a data processor. The Chroma meter was calibrated on the Hunter- 2.7. GC analysis
lab color space system with a white tile (D65:Y = 85.1, x = 0.3168,
y = 0.3236). The color content was measured in triplicate according to a The preparation of the fatty acid methyl esters (FAMEs) followed the
similar protocol described by Romaniello et al. [23] and was expressed European Method EN14103:2011 [25] with nonadecanoic acid methyl
in the Hunter and CIELAB color parameters: L/L∗ (lightness), +a/+a∗ ester (Sigma Aldrich Chemical Co., Missouri, USA). The nonadecanoic
(redness), and +b/+b∗ (yellowness). Similarly, the 𝛽-carotene content acid methyl ester was the internal standard for calibration and quanti-
was determined using a double beam UV-Vis (Ultraviolet-Visible) spec- tative analysis. The FAMEs contents of oils were detected qualitatively
trophotometer (Perkin Elmer, Waltham, Massachusetts, USA). Before and quantitatively with the Shimadzu GC-2010 plus (Shimadzu Scien-
any measurements, the oil samples were centrifuged for 30 min at tific Instruments, Columbia, USA) gas chromatograph (GC) and an Auto-
5000 rpm to reduce the interference of suspended particles. The 𝛽- Injector (AOC-20i). The GC instrument was equipped with a capillary
carotene content (ppm) was measured in triplicate according to the column VF-5ms (5% phenyl, 95% dimethylpolysiloxane) of 30 m length,
method described by Cayuela et al. [24] at 450 nm. 0.25 mm inner diameter, and 0.25 μm film thickness. An aliquot of 1 μL
of the sample was injected into the GC with an Auto-Injector AOC 20i,
2.5. FTIR analysis keeping the injector temperature at 250 °C. The injection mode was
set to split mode with a split ratio of 1:10. Nitrogen gas (99.9%) was
The infrared (IR) spectra of the VOs were obtained using an FTIR used as a carrier gas at a constant flow rate of 1 mL/min. The column
Bruker Alpha FTIR spectrometer equipped with platinum attenuated to- temperature was initially set at 100 °C (hold time, 0 min), increasing
tal reflectance (ATR-FTIR, Bruker, Karlsruhe, Germany). The analysis by 6 °C/min to 250 °C (hold time, 25 min) and a total program time of
was performed at the KNUST central laboratory in Kumasi, Ghana, at 50 min. The flame ionization detector (FID) was set to 260 °C during the
a room temperature of 25 °C. The ATR-FTIR diamond crystal and all analysis. The FAs were identified by comparing their retention times to
accessories were thoroughly cleaned with isopropanol between samples a reference standard (C8 – C24 FAMEs Mixture, SUPELCO, Bellefonte,
and background scans. Each VOs sample of 1 mL was placed on the PA, USA). The internal standard was used to determine each FAME’s
surface of the diamond ATR crystal in the ATR accessory chamber, and response factor in the reference standard for quantification.
sample spectra were collected. The spectra were measured from 4000
to 400 cm−1 with a scanning time of 32 s at a spectral resolution of 2.8. Statistical analysis
4 cm−1 . The spectra were obtained using OPUS software (Bruker, Karl-
sruhe, Germany). The statistical analysis was carried out using OriginPro (version
2020, Northampton, MA, USA) and Microsoft Excel®. Each sample was
2.6. TGA analysis repeated three times, and the mean values were presented to guarantee
reproducibility. Each replicate was expressed as the mean ± SD, and the
The TGA analysis of the oils was performed at the University of one-way analysis of variance (ANOVA) was to compare the mean values.
Ghana laboratory (Legon, Ghana). The SDT-Q600 Simultaneous TGA Se- The significant difference was defined by the Duncan test (p < 0.05).
ries 0600–1786 (TA Instruments, Delaware, USA) instrument was used
for the TGA analysis. The calibration of the TGA instrument was con- 3. Results and discussion
ducted with pure indium, and the baseline was obtained with an empty
alumina pan used as a reference. 8 mg of the oil was weighed into an alu- 3.1. Thermally hydrolyzed products
mina sample pan. The sample pans were placed in the sample chamber
and heated until 600 °C with a heating rate of 20 °C/min under nitrogen When food materials are added to heated VOs during deep-frying,
flow and ambient air. the air and moisture in the oil or introduced by the food materials cause

3
D. Dodoo, F. Adjei, S.K. Tulashie et al. Measurement: Food 7 (2022) 100035

Fig. 3. (a) Moisture content, (b) acid value, and (c) free fatty acids of fresh and heated vegetable oils.

a thermochemical reaction that results in thermal hydrolysis of the oil the oil caused by increasing the number of frying cycles [29]. As the
[26–28]. Thermal hydrolysis is a reversible thermochemical reaction temperature rises, the water in the frying medium evaporates, reducing
stimulated by heat in the hydrolysis of triacylglycerols (TAGs) that pro- the moisture content of the oil. This outcome has also been reported by
duces thermally oxidized products [28]. In thermal hydrolysis, the mois- Zhang et al. [29] and Erickson [8]. According to the authors, increasing
ture existing in the oil as weak nucleophiles attacks the ester linkage the number of deep-frying cycles evaporated the majority of the water
of TAGs to produce diacylglycerols (DAGs), monoacylglycerols (MAGs), in the frying medium, lowering the moisture content.
glycerol, and free fatty acids (FFA) [2]. The thermally hydrolyzed prod-
ucts in oils are often measured to check the oil’s quality. Hence, a free 3.1.2. Acid value
fatty acid test was conducted to evaluate the quality of the RHO used The acid values of the oils were measured and are presented in
for deep-frying yam fries. Likewise, the moisture content and acid value Fig. 3(b). The variation in acid values shown in Fig. 3(b) indicated that
were also ascertained to monitor the thermal hydrolysis of RHO. repeatedly heating the oil had a significant effect on their AV. The AV in
the RHO was significantly (p < 0.05) increased compared with the SHO
3.1.1. Moisture content and FVO. The initial acid values (%) recorded for FCNO, FSFO, and FSBO
Thermal hydrolysis of oils is activated by heat, air, and moisture, were 2.04 ± 0.11, 0.31 ± 0.12, and 0.24 ± 0.09, respectively. After a sin-
which initiates a thermochemical reaction that leads to the oil’s deteri- gle deep-frying cycle, no significant changes (p > 0.05) in AV (%) were
oration. The moisture content is an indirect measure of the oil’s quality observed for the SHO (SCNO: 2.13 ± 0.11; SSHO: 0.34 ± 0.10; SSBO:
and resistance to thermal oxidation. Fig. 3(a) shows the moisture con- 0.27 ± 0.11). The AV in RHO, on the other hand, rose to 2.65 ± 0.06
tent of the oils before and after deep-frying. The initial moisture con- (RCNO), 1.01 ± 0.09 (RSFO), and 0.61 ± 0.10 (RSBO). Regardless of the
tent (%) was 0.58 ± 0.07, 0.44 ± 0.09, and 0.22 ± 0.05, for CNO, SFO, oil type, the AV increased as the number of deep-frying cycles increased.
and SBO, respectively. CNO had the highest moisture content, which The increasing rate accelerated as the number of frying cycles increased,
makes it more susceptible to thermal hydrolysis. The moisture content increasing the acidity of the oil. According to Zhang et al. [29], this trend
of the oils was reduced at varying rates depending on the type of oil can partially be attributed to the hydrolysis of TAGs and carboxyl groups
and the number of frying cycles. After a single deep-frying cycle, there into oxidized and polymerized products during deep-frying. VOs are pri-
was no significant (p > 0.05) change in moisture content for SHO, but marily composed of triglycerides that are disintegrated during thermal
there was a significant (p < 0.05) loss of moisture content for RHO. Re- hydrolysis to produce fatty acids. Hence, the AV of the oil represents
peatedly heating of the oils dropped the initial moisture content (%) of the degree to which the triglycerides in the oil have been thermally
the oils to 0.40 ± 0.11 (RCNO), 0.19 ± 0.10 (RSFO), and 0.07 ± 0.12 hydrolyzed, and it increases as thermal hydrolysis progresses [29]. A
(RSBO). After deep-frying, about 31% (RCNO), 57% (RSFO), and 68% related study by Li et al. [30] investigated the acidity of oils used for
(RSBO) of the oils’ initial moisture content (%) were lost. This is due deep-frying potato chips. It was revealed that increasing the number of
to the increased evaporation of water during the thermal hydrolysis of deep-frying cycles increased the oil’s acidity.

4
D. Dodoo, F. Adjei, S.K. Tulashie et al. Measurement: Food 7 (2022) 100035

3.1.3. Free fatty acids amounts of primary and secondary oxidation products [31]. The degra-
FFA is a significant criterion for determining the quality of VOs suit- dation of unsaturation content is advanced in oils with a high degree of
able for both edible and non-edible applications. FFA is frequently re- unsaturation, as evident in Fig. 4(a). The RSBO with the highest degree
duced in oils after refining; as a result, the FFA obtained for the fresh of unsaturation showed a significant (p < 0.05) decrease in IV compared
VOs was quite low, as shown in Fig. 3(c). The initial FFA (%) in the fresh with RCNO.
oils was about 1.03 ± 0.08 (FCNO), 0.13 ± 0.13 (FSFO), and 0.10 ± 0.09 Nonetheless, a reduction in the total UFAs in oils after deep-frying
(FSBO), which rose with an increase in the number of frying cycles. The is accompanied by an increase in the formation of hydroperoxides [31].
FFA (%) remained nearly unchanged in SHO, while it increased signifi- Therefore, to observe the formation of the hydroperoxides, the peroxide
cantly (p < 0.05) in RHO. The free fatty acids (%) recorded for RHO were value (PV) was measured and presented in Fig. 4(b). The PV is an impor-
1.43 ± 0.09 (RCNO), 0.39 ± 0.11 (RSFO), and 0.36 ± 0.11 (RSBO). The tant parameter for assessing the formation and breakdown of hydroper-
increase in FFA with increasing deep-frying cycles is caused by TAGs oxides during thermal oxidation. Although the PV alone is insufficient to
hydrolysis and hydroperoxide decomposition during the primary oxida- ascertain the oil’s thermal oxidation process, it is a useful measurement
tion stage in oils [10,28]. An increase in FFA after deep-frying implies for monitoring its oxidation during a particular period. In Fig. 4(b), the
that the oils are highly unstable and may quickly oxidize to produce PV in the RHO significantly (p < 0.05) increased compared with SHO
harmful oxidized compounds in oils, thereby affecting the quality of the and FVO. In the fresh oils, the initial PV (meq/kg) obtained for FCNO,
fried food [10,28]. The findings are consistent with those reported by FSFO, and FSBO was 1.98 ± 0.09, 3.21 ± 0.10, and 4.54 ± 0.13, respec-
Kaur et al. [28], who reported an increase in FFA when different oils tively. After a single deep-frying cycle, no significant changes (p > 0.05)
were repeatedly used for deep-frying chickpea splits. in PV (meq/kg) were observed for the SHO (SCNO: 1.97 ± 0.10; SSHO:
Overall, the amount of AV and FFA in waste cooking oils is critical in 3.20 ± 0.13; SSBO: 4.52 ± 0.11). The PV in RHO, on the other hand,
determining whether an alkali or acid catalytic method should be used recorded 2.53 ± 0.09 (RCNO), 5.24 ± 0.11 (RSFO), and 7.12 ± 0.13
to produce biodiesels and biolubricants [14,16]. Hence, the AV and FFA (RSBO). From Fig. 4(b), it was clear that the number of deep-frying cy-
data from this study may be useful to researchers interested in producing cles influenced the level of peroxides in oils. Increasing the number of
biodiesel from waste cooking oils used to deep-fry yam fries. frying cycles could have quickened the thermal oxidation process, ac-
celerating the hydroperoxides’ formation. The singly heated oils showed
3.2. Thermally oxidized products a nominal increase in PV compared with the repeatedly heated oils.
Among the repeatedly heated oils, RSBO recorded the highest PV. In all,
Any fair amount of dissolved oxygen in oils or oxygen introduced the PV results corroborate well with other PV data reported by Jaarin &
when food materials are added to the oil can automatically initiate an Kamisah [9], who observed a rise in PV for repeatedly heated palm oil
oxidation reaction at room temperature, a process known as autoxida- (PO) and soy oil (SO). In their study, the five-times heated PO and SO
tion [31]. Theoretically, autoxidation is a free radical chain reaction in increased from 2.1 ± 0.11 to 9 ± 0.10 meq/kg and from 5.2 ± 0.15 to
which the initiation steps can be catalyzed by a variety of oxidation 11.6 ± 0.20, respectively.
initiators (light, heat, metals, etc.) [1]. In deep-frying, RHO consecu-
tive heating at high temperatures, can catalyze and accelerate the oxi- 3.2.2. Total polar compounds
dation reaction, rapidly forming free radicals [1,8]. This subsequently The total polar compounds (TPC) indicate the total amount of degra-
leads to hydroperoxides developing at the primary stage of lipid oxida- dation compounds in deep-frying oils. The total polar compounds in-
tion, which decompose quickly under further heating into the secondary clude polymers of TAGs, FFA, hydroperoxides, carbonyl compounds,
oxidation products (aldehydes, ketones, alcohol, etc.). Therefore, both and others formed due to the thermal hydrolysis and oxidation of oil
the oxidation reactions and decay of hydroperoxides are advanced with when used in deep-frying food. The TPC was measured and presented
heat, leading to the oxidation type referred to as thermal oxidation [31]. in Fig. 4(c). Fig. 4(c) shows that TPC advanced with the number of
Thermally oxidized products, which result from thermal oxidation, have frying cycles due to thermochemical reactions promoted by the fry-
an impact on oil quality and are frequently used to monitor oil quality. ing conditions. The initial TPC (%) of the fresh oils was 1.84 ± 0.09,
This study suspected that deep-frying yam fries with RHO might con- 3.17 ± 0.07, and 3.59 ± 0.07 in the FCNO, FSFO, and FSBO, respectively.
tain some thermally oxidized products due to their continuous heating Insignificant changes (p > 0.05) were observed in SHO. At the end of the
at around 175 °C for a longer period of time without refilling with fresh deep-frying, the TPC (%) rose rapidly in the repeatedly heated oils to
oils. As a result, the peroxide value (PV) and polar compounds in RHO 16.34 ± 0.07 (RCNO), 18.78 ± 0.08 (RSFO), and 19.52 ± 0.06 (RSBO).
were measured to assess their quality. At the same time, FTIR analysis Though many reports and standards [21,30] advocate that the highest
was used to identify the secondary oxidation products. Iodine value (IV) level of TPC above which oil should be suitable for deep-frying is 24–
was also measured to monitor the changes in the unsaturation content 27%; the rapid increase in TPC suggests that increasing the number of
in RHO during their thermal oxidation. frying cycles is accompanied by an increase in TPC during deep-frying.
In fact, a study by Li et al. [30] observed that the TPC in frying oils
3.2.1. Iodine and peroxide value increases linearly with frying time. Their study noted a TPC of around
The iodine value was measured to monitor the changes in the degree 33.76% after 40 h of repeatedly heating the oil, exceeding the TPC stan-
of unsaturation of the oils after deep-frying. The results are presented dards of about 24–27%. Therefore, repeatedly heating oils beyond the
in Fig. 4(a), comparing the IV of FVO, SHO, and RHO. In Fig. 4(a), four frying cycles reported in this study could likely result in higher TPC
the initial IV (g of iodine/100g of oil) obtained for the FCNO, FSFO, levels exceeding the recommended standards. This shows that limiting
and FSBO was 7.09 ± 0.08, 121 ± 0.10, and 124 ± 0.08, respectively. the use of RHO for deep-frying could reduce the amount of TPC in the oil
The higher IV for fresh SBO and SFO is due to their high unsaturated and the fried food. Despite this, the increase in total polar compounds
fatty acids (UFAs), as registered in the GC analysis (Table 2). Neverthe- in RSBO was relatively faster than in RSFO and RCNO.
less, no significant changes (p > 0.05) in IV (g of iodine/100 g of oil)
were observed for the SHO (SCNO: 7.07 ± 0.12; SSHO: 119 ± 0.09; 3.2.3. FTIR studies
SSBO: 122 ± 0.11). Whereas a significant decrease (p < 0.05) in IV The FTIR measurement was conducted to identify any thermally ox-
was recorded after deep-frying the yam fries with RHO. The IV (g of idized compounds that emerge before and after deep-frying yam fries.
iodine/100g of oil) was decreased to 4.91 ± 0.11 (RCNO), 89 ± 0.12 The FTIR measurement of the fresh VOs was first determined, and the
(RSFO), and 91 ± 0.11 (RSBO), possibly as a result of thermal oxida- spectra of each oil type are shown in Fig. 5. The spectra in Fig. 5 show
tion after deep-frying. Thermal oxidation reactions are characterized by only minor molecular differences in the oil, with a few changes in ab-
a drop in the total UFAs in oils with an accompanying increase in the sorption band intensity being observed. These variations are caused by

5
D. Dodoo, F. Adjei, S.K. Tulashie et al. Measurement: Food 7 (2022) 100035

Fig. 4. (a) Iodine value, (b) peroxide value, and (c) total polar compounds of fresh and heated vegetable oils.

Fig. 5. Initial FTIR spectra of fresh vegetable

oils.

6
D. Dodoo, F. Adjei, S.K. Tulashie et al. Measurement: Food 7 (2022) 100035

Fig. 6. The spectral changes in the cis-double bonds of repeatedly heated oils (RCNO; RSBO; RSFO) in comparison to fresh oils (FCNO; FSBO; FSFO). The spectral
changes were due to decrease in unsaturated fatty acids contents at 722 cm−1 (a, b) and 3009 cm−1 (c, d).

differences in the levels of unsaturated fatty acids (UFAs) and saturated double bonds imply that this oil has a higher degree of unsaturation,
fatty acids (SFAs) in the various oils. For instance, significant spectral often showing stronger absorption intensities in their spectra. For this
differences were observed at the 3006 ± 3 cm−1 and 720 ± 5 cm−1 study, the FSBO and FSFO showed the highest peak intensities (Fig. 6(a)
bands, which were assigned to the presence of the cis-double bonds in and (c)), suggesting a higher degree of unsaturation, while the FCNO
the UFAs. The SFO and SBO are characterized by a higher level of UFAs, recorded the weakest peak intensity. In Fig. 6(b) and (d), the peak in-
showing more elevated absorption bands of around 3006 ± 3 cm−1 and tensities were found to decrease in RHO due to the decomposition of the
720 ± 5 cm−1 than the CNO with low UFAs. Overall, the spectra of cis-olefinic double bonds in UFAs during thermal oxidation. The peaks
the oils and their respective bands’ assignments portrayed the typical were more reduced in the RSBO and RSFO, denoting a higher degrada-
absorption bands already registered in the literature for each oil type tion of the cis-olefinic double bonds in the UFAs of these oils. One way to
[32,33]. further establish this is to compare it with the iodine value obtained in
On the one hand, the FTIR measurements of RHO revealed spectral this work, reflecting the degree of unsaturation in the VOs. The IV also
changes that differed from those of the fresh oils. These changes have decreased due to the loss of UFAs and followed the exact order shown in
been reported in the literature [34] to result from the degradation of Fig. 6(b) and (d) as; RSFO > RSBO > RCNO. Poiana et al. [34] reported
the UFAs during the oil’s thermal oxidation, leading to the formation of a decrease in absorbance at 722 cm−1 and 3009 cm−1 of SBO due to the
thermally oxidized compounds. Hence, the specific IR regions likely to loss of the cis-disubstituted olefins groups during oxidation.
change during thermal oxidation were enlarged to monitor the thermo- In the FTIR measurement of oils, three prominent absorption bands
oxidative behavior of RHO. are useful to detect the presence of thermally oxidized compounds.
Fig. 6(a)–(d) present the bands’ shifts in the IR range from 702– The first band is located around 3400–3450 cm−1 , which is associated
742 cm−1 and 2980–3030 cm−1 of FVO and RHO. The peak intensities with the –OH vibrational stretch of the hydroperoxide groups; while
at 3006 ± 3 cm−1 and 720 ± 5 cm−1 bands correspond to the vibra- the bands located around 1155 ± 2 and 1742 ± 3 cm−1 are associated
tions of the cis-olefinic double bonds in UFAs [35]. The position of these with the vibrational stretching of the ester triglyceride carbonyl groups.
peaks varies greatly, depending on the amount of cis-olefinic double These absorption band regions have been magnified and presented in
bonds present in UFAs found in oils.. Oils containing many cis-olefinic Figs. 7 and 8 to demonstrate their strong presence in RHO compared

7
D. Dodoo, F. Adjei, S.K. Tulashie et al. Measurement: Food 7 (2022) 100035

Fig. 7. Magnified region of hydroperoxides absorption

band in FTIR spectra of repeatedly heated oils (RCNO;
RSBO; RSFO) in comparison to fresh oils (FCNO; FSBO;
FSFO). The dash and solid lines represent the repeat-
edly heated oils and fresh oils, respectively.

with FVO. In Fig. 7, the appearance of this band was absent in the fresh oxides’ decomposition in SBO and SFO after prolonged deep-frying of
oils (bold lines) before the oils were used for deep-frying. This can be chickpea splits.
attributed to the negligible amount of hydroperoxides in the fresh oils,
producing a very weak band that could scarcely be noticed. Yet, ther- 3.3. Thermally degraded products
mal oxidization occurred after the RHO was continually used for deep-
frying, producing hydroperoxides at the primary oxidation stage. The Deep-frying of foods involves the thermal processing of food by grad-
hydroperoxides formed (dash lines) were more pronounced in RSBO and ually heating the frying medium until the fried food is in a suitable and
RSFO than in RCNO (Fig. 7). This is because the thermo-oxidative sus- acceptable form for consumption. As the frying advances, certain con-
ceptibility of oils depends heavily on their degree of unsaturation. The stituents in oils, such as the nutrients and food colorant, which are li-
RCNO with high SFAs was more stable during the thermal oxidation pro- able to heat change, are sequentially impacted [8]. Hence, repeatedly
cess, producing weaker absorption bands. Whereas the RSFO and RSBO deep-frying food without replacing the oils after each deep-frying cy-
showed a more potent and broad absorption band because they quickly cle may deteriorate certain constituents in oils that decompose under
oxidize and rapidly generate hydroperoxides due to their high degree of excessive and consecutive heating. This thermally degrades key oil con-
unsaturation. Along these lines, the outcome is consistent with the per- stituents, including vitamins, antioxidants, food colorants, PUFAs, and
oxide values reported for RSBO and RSFO in this study, confirming that others, affecting the oil’s nutritional and sensory quality [29,31]. Hence,
unsaturated oils are more prone to thermal oxidation than saturated oils. this study assesses the nutritional quality and the color of the investi-
A similar study by Xu et al. [36] compared oxidized and unoxidized VOs gated oils by measuring the color contents, fatty acid compositions, and
(SFO and SBO) in IR spectra in the range of 3750–3150 cm−1 . In their the amount of 𝛽-carotenes.
work, the unoxidized SFO and SBO showed a very weak band, while a
strong absorption band was observed for the oxidized SFO and SBO. 3.4. Changes in beta-carotene
Nevertheless, the hydroperoxides can be decomposed into sec-
ondary oxidation products, causing an absorption around 1155 ± 2 and Fig. 9 shows the 𝛽-carotene content of the analyzed oils. The ini-
1742 ± 3 cm−1 , as depicted in Fig. 8(a)–(d). In Fig. 8(a)–(c), the bands tial 𝛽-carotenes (ppm) obtained for the FCNO, FSBO, and FSFO were
at 1157 and 1745 cm−1 were less broadened with shorter peaks in FVO 3.43 ± 0.09, 3.01 ± 0.09, and 8.05 ± 0.10, respectively. After the first
compared with those revealed in RHO (Fig. 8(b)–(d)). The increase and deep-frying cycle, the initial 𝛽-carotenes (ppm) in the SCNO, SSFO, and
broadening of the intensity peaks in RHO arise when the carbonyl es- SSBO decreased by 6.99%, 5.65%, and 1.86%, respectively. In contrast,
ter functional groups are hydrolyzed into aldehydes, ketones, and other a rapid decrease was observed for RHO. RCNO, RSFO, and RSBO fell
secondary thermally oxidized compounds after deep-frying [28]. This by 49.11%, 49.71%, and 41.92%, respectively. The outcome was not
broadening is wider when the degree of unsaturation of the oil is high. unexpected given the many studies [37,38] that reported 𝛽-carotenes
Hence, the oxidation of the oil is more advanced, forming high levels of as thermolabile pigments in oils that are unstable and decompose when
secondary oxidation products (aldehydes, ketones, alcohol, etc.). On this heated at temperatures beyond their threshold. Carotenoid pigments are
ground, the RSBO showed the highest broadening, followed by the RSFO made up of conjugated double bonds that are easily cleaved when oils
and RCNO. Similarly, Kaur et al. [28] also associated the variations at are heated to high temperatures repeatedly. Their apparent degradation
1157 cm−1 and 1745 cm−1 with the ketones’ formation and hydroper- in RHO may therefore be linked to the repeated heating of oils, which

8
D. Dodoo, F. Adjei, S.K. Tulashie et al. Measurement: Food 7 (2022) 100035

Fig. 8. The spectral changes in the carbonyl groups of the repeatedly heated oils (RCNO; RSBO; RSFO) in comparison to fresh oils (FCNO; FSBO; FSFO). The spectral
changes were due to the formation of secondary oxidation products (aldehydes, ketones, and others) at 1157 cm−1 (a, b) and 1743 cm−1 (c, d).

Fig. 9. Beta-carotene of fresh and heated vegetable

oils.

9
D. Dodoo, F. Adjei, S.K. Tulashie et al. Measurement: Food 7 (2022) 100035

Table 1
The color contents in fresh and heated vegetable oils.

Frying Hunter scale (L a b) CIELAB scale (L∗ a∗ b∗ )

VOs Cycle (s)
L a b L∗ a∗ b∗

CNO 0 84.60 ± 08 0.79 ± 08 2.68 ± 11 87.80 ± 08 0.47 ± 07 8.07 ± 10

1 81.12 ± 06 0.77 ± 09 2.52 ± 12 83.30 ± 09 0.45 ± 09 8.03 ± 11
4 59.70 ± 08 0.58 ± 06 1.41 ± 10 63.61 ± 04 0.30 ± 06 5.02 ± 08
SFO 0 84.60 ± 07 0.74 ± 08 3.57 ± 12 88.50 ± 07 0.66 ± 11 8.92 ± 09
1 84.02 ± 08 0.72 ± 06 3.52 ± 13 87.10 ± 06 0.60 ± 13 8.87 ± 08
4 57.90 ± 13 0.53 ± 11 1.80 ± 07 59.12 ± 04 0.45 ± 10 5.17 ± 04
SBO 0 85.80 ± 08 1.03 ± 09 3.77 ± 08 88.70 ± 03 0.70 ± 06 9.12 ± 13
1 85.30 ± 05 0.98 ± 07 3.70 ± 07 88.22 ± 05 0.66 ± 10 9.07 ± 12
4 55.41 ± 03 0.69 ± 05 1.89 ± 12 58.81 ± 08 0.49 ± 08 5.15 ± 07

Abbreviations: VOs, vegetable oils; CNO, coconut oil; SFO, sunflower oil; SBO, soybean oil; L/L∗
represents the lightness of the oil; a/a∗ given in positive values denotes the red color pigment in
the oils; b/b∗ given in positive values represents the yellow color pigment in the oils.

causes them to deteriorate. This outcome is strongly supported by a re- were abundant in UFAs, mainly oleic (C18:1) and linoleic (C18:2) acids.
lated study on the changes in 𝛽-carotenes of VOs [39]. In all, 𝛽-carotenes Although both the FSFO (63.29%) and the FSBO (51.46%) were high in
are essential color pigments in oils, as well as a source of minor vitamins linoleic acids (C18:2), the FSBO (7.58%) had a higher percentage of
and antioxidants. This contributes to the nutritional quality and protects linolenic acids (C18:3) than the FSFO (0.48%).. Overall, the fatty acid
the oil from photosensitized oxidation [39]. Hence, the use of RHO must compositions in the fresh oils were a natural blend of SFAs and UFAs.
be minimized to prevent their loss in the oil and the fried food. As revealed in Table 2, the FCNO (92.11%) contained a high portion of
3.3.2. Changes in Color Content SFAs, while FSFO (89.01%) and FSBO (83.33%) recorded a high amount
Oil color is a critical quality parameter that is commonly used in of UFAs. The fatty acid profile reported in this study agrees well with
restaurants, industries, and homes to monitor oil deterioration. An oil’s those registered in the literature [42,43]. However, there were substan-
fresh color disappearance may imply a problem during its usage and ap- tial changes when the number of deep-frying cycles was increased.
plication. In this study, the investigated oils’ color contents were deter- In Table 2, the thermal oxidation of RHO caused a change in their
mined and expressed on two scales presented in Table 1. Table. 1 shows level of fatty acid composition compared with SHO. In thermal ox-
the color contents of the analyzed oils in hunter (L a b) and CIELAB (L∗ idation, the fatty acids present in oils undergo the most significant
a∗ b∗ ) color scales. The L/L∗ values depict the darkness or lightness of changes throughout the oxidation process. The oleic (C18:1) and linoleic
the oil in the range of 0 to 100, while the a/a∗ and b/b∗ positive values (C18:2) acids quickly decompose under heat treatment to produce radi-
indicate the red and yellow color pigments in the oil, respectively. These cals (free radicals) that propagate during oxidation. However, the stearic
three designations were used to describe the oil’s color before and after (C18:0) acids rarely participate in the oxidation process due to their
deep-frying of the yam fries. From Table 1, the initial color contents of high oxidative stability [32], as shown in Table 2. The SHO shown
the fresh oils before deep-frying possess sufficient lightness (L/L∗ ), red in Table 2 showed no appreciable differences in the amount of stearic
(a/a∗ ), and yellow (b/b∗ ) color contents. The L a b positive values of (C18:0), oleic (C18:1), and linoleic (C18:2) acids when compared with
the initial FCNO were 84.60 ± 08, 0.79 ± 08, and 2.68 ± 11, that for the fresh oils. On the contrary, there was a slight increase (p < 0.05) in
FSFO were 84.60 ± 07, 0.74 ± 08, and 3.57 ± 12, while FSBO recorded the amount of stearic (C18:0) acids and a decrease in oleic (C18:1) and
85.80 ± 08, 1.03 ± 09, and 3.77 ± 08. Likewise, the L∗ a∗ b∗ positive linoleic (C18:2) acids in RHO. The higher the oleic (C18:1) and linoleic
values of initial FCNO recorded 87.80 ± 08, 0.47 ± 07, and 8.07 ± 10, (C18:2) acids in the oil type, the higher the quantitative loss. As a re-
that for FSFO were 88.50 ± 07, 0.66 ± 11, and 8.92 ± 09, while FSBO sult, the oleic (C18:1) and linoleic (C18:2) acids in RSBO and RSFO were
gave 88.70 ± 03, 0.70 ± 06, and 9.12 ± 13. After the first deep-frying rapidly reduced than RCNO.
cycle, no significant (p > 0.05) change in the color was observed for These changes in the fatty acids are due to their varied susceptibility
SHO. On the one hand, the color of RHO was found to decrease sig- to hydrogen abstraction during the oxidation process. At the initial stage
nificantly. The lightness (L) of RCNO, RSFO, and RSBO decreased by of thermal oxidation, withdrawing a hydrogen atom from the substrate
29.41%, 31.62%, and 35.41%, respectively. This trend was also consis- (fatty acids) is needed in creating the free radicals or radical ions to
tent in the CIELAB (L∗ ) color scale, with the L∗ positive values of RCNO, initiate the oxidation [44]. The hydrogen removal is held energetically
RSFO, and RSBO decreasing by 27.62%, 33.21%, and 33.72%, respec- by the fatty acids type and the bond strength of the hydrogen atom
tively. The yellow color pigments were intensely lost in RHO, decreasing adjacent to the 𝛼–CH2 in the fatty acid’s structure [45]. The amounts
by 47.43% (RCNO), 49.62% (RSFO), and 49.92% (RSBO) in the Hunter of stearic (C18:0), oleic (C18:1), and linoleic (C18:2) acids in VOs have
color scale (L); while by 37.81% (RCNO), 42.12% (RSFO), and 43.53% severally been reported to have different bond strengths, affecting their
(RSBO) in the CIELAB scale. This may be due to the thermal degrada- stability towards oxidation [45,46]. Min and Boff [44] reported that the
tion of the color pigments when the oils are repeatedly heated for a hydrogen bond strength in SFAs is higher than in UFAs. For instance,
consecutive period. At high deep-frying cycles, non-enzymatic brown- about 100 kcal/mol of energy is required to remove hydrogen from the
ing reactions occur, resulting in the darkening of the oils. Our results are 𝛼–CH2 in the SFAs, while ≤ 80 kcal/mol for the UFAs. Therefore, the
similar to the work of de Almeida et al. [40], reporting the changes in hydrogen positioned close to the double bonds found in the UFAs is easy
the color pigments of crude palm oil after deep-frying Akara (fried-bean to remove due to low energy requirements. Easy removal of hydrogen
paste). suggests a rapid initiation of thermal oxidation, implying that the oleic
and linoleic acids containing a high number of double bonds in their
chemical structure are most likely to change during thermal oxidation.
3.4.1. Changes in fatty acid compositions In contrast, stearic acids are the most stable against thermal oxidation,
GC studies. The fatty acid compositions of the oils were identified us- possessing no double bonds in their chemical structure.
ing GC and are presented in Table 2. Before the deep-frying, the dom- However, a few saturated fatty acids (myristic acid, palmitic acid,
inant fatty acid in FCNO was lauric acid (C12:0) [41], reaching about and stearic acid) were found to slightly increase during deep-frying. It
48.50%, while the other SFAs were roughly 43.61%. The FSFO and FSBO

10
D. Dodoo, F. Adjei, S.K. Tulashie et al. Measurement: Food 7 (2022) 100035

is unclear why these SFAs showed this trend with an increase in the num-
ber of frying cycles. One reason may be attributed to the different types
of secondary oxidation products formed at the latter stage of an oxida-
tion process. Depending on the kind of UFAs and the oxidation pathway,
various types of secondary oxidation products can exist in the VOs af-
ter oxidation [47,48]. Thus, at the termination stage of the oxidation
Linoleic acid (C18:2) process, the oxidized hydroperoxide groups are cleaved by homolysis

Data are given as mean values ± S.D, n = 3, S.D; Standard Deviation, The VOs, CNO, SBO and SFO represent the vegetable oils, coconut oil, soybean oil and sunflower oil.
to yield the secondary oxidation products [49]. The resulting secondary
oxidized products (fatty acids) may either be unsaturated or saturated

66.50 ± 0.05
65.11 ± 0.05
61.90 ± 0.05
64.79 ± 0.09
64.08 ± 0.09
59.50 ± 0.09
depending on the location of the carbon-carbon bond which encoun-
2.33 ± 0.04
2.14 ± 0.04
1.65 ± 0.04
tered the cleavage [49]. The formation of saturated oxidized fatty acids
may increase the level of saturation in the VOs and vice versa. In two
previous studies by Frankel et al. [50] and Min and Akoh [46], the au-
thors reported the formation of different saturated secondary (volatile)
Oleic acid (C18:1)

oxidized products, such as saturated oxidized aldehydes and ketones.

29.12 ± 0.02
28.50 ± 0.02
26.40 ± 0.02
25.82 ± 0.03
25.70 ± 0.03
23.90 ± 0.03
Similar results with an increase in the amount of stearic (C18:0) acids
7.07 ± 0.03
6.92 ± 0.03
6.85 ± 0.03

while a decrease in oleic (C18:1) and linoleic (C18:2) acids with an in-
crease in frying oil cycles have also been reported by Kaur et al. [28].
In all, the PUFAs were the major components lost in RHO compared
with FVO and SHO, as shown in Fig. 10(a) and (b). Hence, one may
Stearic acid (C18:0)

suspect from Fig. 10(a) and (b) that when oils are repeatedly used in
the deep-frying of yam fries, their PUFAs deteriorate, which may be
1.01 ± 0.03
1.17 ± 0.03
2.56 ± 0.03
1.01 ± 0.03
4.09 ± 0.03
3.34 ± 0.03
2.01 ± 0.05
2.46 ± 0.05
4.14 ± 0.05

lost in the fried food, affecting their nutritional quality. The PUFAs are
chemical constituents with important chemical, biological, physical, and
structural roles in the body, serving as a critical energy source. Since
the human body cannot synthesize all the essential fatty acids required,
Palmitic acid (C16:0)

foods must supply essential fatty acids, especially from the PUFAs in
VOs. Hence, their loss in oils reduces the fried food’s nutritional quality,
resulting in a nutritional loss in consumers’ diets.
10.80 ± 0.10
7.18 ± 0.11
7.53 ± 0.11
8.62 ± 0.11
3.11 ± 0.08
6.48 ± 0.08
6.77 ± 0.08
6.10 ± 0.10
6.78 ± 0.10

3.4.2. TGA studies

A thermogravimetric analysis (TGA) was conducted to evaluate the
thermal decomposition and weight loss (%) in repeatedly heated oils
Myristic acid (C14:0)

compared with the fresh oils. The thermal behavior was measured as a
function of temperature (°C) and is presented in Fig. 11. The TGA curves
17.90 ± 0.06
18.08 ± 0.06
19.70 ± 0.06

(Fig. 11) reveal that the thermal behavior of the oils reported a plateau,
0.08 ± 0.04
0.11 ± 0.04
0.11 ± 0.04
0.01 ± 0.02
0.01 ± 0.02
0.09 ± 0.02

with the thermal decomposition of the oils terminating around 460 °C.
From Fig. 11, the initial decomposition temperature (Tonset ) estimated
from TGA curves indicated that the thermal stability of the fresh oils
followed the trend: FSFO (256 °C) > FSBO (215 °C) > FCNO (179°C).
The fatty acid compositions (%) in fresh and heated vegetable oilsa .

Lauric acid (C12:0)

This ranking suggests that the FSFO has the highest thermal stability
than FSBO and FCNO. The FCNO, on the other hand, recorded the lowest
46.91 ± 0.08
47.01 ± 0.08
47.10 ± 0.08

Tonset , insinuating that the oil decomposed at the lowest temperature.

The Tonset signifies the oil’s resistance to thermal degradation, implying
that FCNO is more susceptible to thermal degradation than FSBO and
–
–
–
–
–
–

FSFO. To detect any thermal changes, the TGA data of the fresh oils
Caprylic acid (C8:0)

were compared to those of the repeatedly heated oils.

In the TGA curves (Fig. 11), three thermal decomposition stages
were observed in the range of 200°C to 460°C, with little-to-no residue
8.11 ± 0.03
8.11 ± 0.03
8.13 ± 0.03

remaining at 500°C. Generally, the thermal decomposition of oils occurs

in three stages, associated with the decomposition of PUFAs, MUFAs,
and SFAs [35,51]. The first stage of the thermal decomposition of oils
–
–
–
–
–
–

is crucial in the thermal stability study, as the decomposition of PUFAs

Frying Cycle(s)

begins in this step [52]. The Tonset obtained from TGA shows that the
first decomposition stage in the fresh oils occurred around 179–311 °C
(FCNO), 215–372 °C (FSBO), and 256–409°C (FSFO). The second stage
associated with the decomposition of the MUFAs surfaced around 311–
0
1
4
0
1
4
0
1
4

330 °C (FCNO), 372–405°C (FSBO), and 409–418°C (FSFO). The final

Table 2

CNO

SBO
VOs

SFO

thermal decomposition stage occurred at temperature range of about

a

330–382°C (FCNO), 405–430°C (FSBO), and 418–435°C (FSFO), corre-

sponding to the thermal decomposition of SFAs. In contrast, the RHO
was more thermally decomposed and unstable than the FVO. The three
decomposition stages of RHO were found to occur at lower tempera-
tures than the fresh ones. Thus, the first thermal decomposition stage
was around 158–280 °C (RCNO), 187–351 °C (RSBO), and 217–378°C
(RSFO); the second stage occurred at about 280–305 °C (RCNO), 351–

11
D. Dodoo, F. Adjei, S.K. Tulashie et al. Measurement: Food 7 (2022) 100035

Fig. 10. The polyunsaturated fatty acids of

fresh and heated vegetable oils: (a) soybean oil
and sunflower oil; (b) coconut oil.

Fig. 11. Thermogravimetric analysis of repeatedly

heated oils (RCNO; RSBO; RSFO) in comparison to
fresh oils (FCNO; FSBO; FSFO). The dash and solid lines
represent the repeatedly heated oils and fresh oils, re-
spectively.

392 °C (RSBO), and 378–410 °C (RSFO); and the final stage happened thermal oxidation occurs as the oil is heated repeatedly in the presence
about 305–360°C (RCNO), 392–420 °C (RSBO), and 410–430 °C (RSFO). of air, producing hydroperoxide products [8]. The decomposition of the
The results show that the thermal decomposition of RHO occurred pre- hydroperoxide products generates volatile decomposition compounds
maturely at lower temperatures at all stages than FVOs. This is caused (light hydrocarbons, ketones, aldehydes, alcohols, and esters) in the oil
by the increased levels of volatile compounds in RHO, which are ther-
mally degraded at lower temperatures. When yam fries are deep-fried,

12
D. Dodoo, F. Adjei, S.K. Tulashie et al. Measurement: Food 7 (2022) 100035

[49]. The presence of volatile decomposition compounds increases the References

volatility of RHO and causes the oil to readily vaporize at earlier decom-
position temperatures [49,52]. Hence, the TGA (Fig. 11) curve reveals [1] E. Choe, D.B. Min, Chemistry of deep-fat frying oils, J. Food Sci. (2007),
that repeatedly heating oil raises its volatility, causing the oil to ther- doi:10.1111/j.1750-3841.2007.00352.x.
mally decompose at lower temperatures. [2] S. Asokapandian, G.J. Swamy, H. Hajjul, Deep fat frying of foods: A critical
review on process and product parameters, Crit. Rev. Food Sci. Nutr. (2019),
doi:10.1080/10408398.2019.1688761.
4. Conclusions [3] E. Akil, V.N. Castelo-Branco, A.M.M. Costa, A.L. Do Amaral Vendramini, V. Calado,
A.G. Torres, Oxidative stability and changes in chemical composition of extra virgin
Deep-frying involves the absorption of frying oils into food materials olive oils after short-term deep-frying of French fries, JAOCS J. Am. Oil Chem. Soc.
(2015), doi:10.1007/s11746-015-2599-2.
as well as adsorption on their surface; as a result, the quality of the oil [4] M. Sivananthan, M. Elamaran, K. Jasmin, R. Kushha, Level of knowledge, attitude
is critical, as it greatly influences the quality of fried food. This necessi- and practice of food outlet operators in Raub, Pahang morning market regarding
tates an in-depth investigation into the quality of RHO commonly used the usage of repeatedly heated cooking oil-A cross sectional study, Int. J. Biomol.
Biomed. (2013), doi:10.6084/M9.FIGSHARE.1404920.
in restaurants, homes, industries, and fast-food chain sectors for deep- [5] R. Perumalla Venkata, R. Subramanyam, Evaluation of the deleterious health ef-
frying foods. Therefore, this study investigated the quality of RHO for fects of consumption of repeatedly heated vegetable oil, Toxicol. Rep. (2016),
deep-frying yam fries. The study observed that the quality of RHO, par- doi:10.1016/j.toxrep.2016.08.003.
[6] A. Abdullah, M.S. Suondoh, C.S. Xuan, N.A. Patah, K. Mokhtar, N.A.M. Fahami,
ticularly coconut oil, sunflower oil, and soybean oil, used for deep-frying Q.H.M. Saad, K. Yusof, K. Jaarin, Level of awareness amongst the general public
deteriorated due to thermochemical reactions promoted by the frying regarding usage of repeatedly heated cooking oil in Kuala Lumpur, Malaysia, Int.
conditions. As a result, there was a significant (p < 0.05) increase in AV, Med. J. (2010).
[7] P.S. Vankar, R.S. Sahu, Regeneration of used soybean frying oils with rampad ad-
FFA, PV, TPC, and secondary oxidation compounds, while a severe de-
sorbent, Electron. J. Environ. Agric. Food Chem. (2011).
crease in the iodine value, PUFAs, 𝛽-carotenes, and color content were [8] M.D. Erickson, Deep Frying, Elsevier, 2007, doi:10.1016/C2015-0-02457-1.
observed in the oils. This could lead to the spoilage of the oils, affecting [9] K. Jaarin, Y. Kamisah, Repeatedly heated vegetable oils and lipid peroxidation, Lipid
the fried yam’s nutritional value, flavor, color, and texture, rendering Peroxidation, InTech, 2012, doi:10.5772/46076.
[10] S. Multari, A. Marsol-Vall, P. Heponiemi, J.-P. Suomela, B. Yang, Changes in the
them unacceptable for human consumption. Our findings suggest that volatile profile, fatty acid composition and other markers of lipid oxidation of six
guidelines for proper deep-frying practices should be implemented in different vegetable oils during short-term deep-frying, Food Res. Int. 122 (2019)
all restaurants, homes, industries, and fast-food chains. Such appropri- 318–329, doi:10.1016/j.foodres.2019.04.026.
[11] E. Aydinkaptan, I.B. Mazi, Monitoring the physicochemical features of sunflower oil
ate deep-frying practices include: (a) topping up the oil lost after deep- and French fries during repeated microwave frying and deep-fat frying, Grasas y
frying, (b) using oils with lower levels of unsaturated fatty acids (UFAs), Aceites (2017), doi:10.3989/gya.1162162.
which are less susceptible to thermal hydrolysis and oxidation, and (c) [12] J.E. Obidiegwu, J.B. Lyons, C.A. Chilaka, The Dioscorea Genus (Yam)–an Appraisal
of Nutritional and Therapeutic Potentials, Foods, 2020, doi:10.3390/foods9091304.
using fresh oils if the cost of the oil is affordable for each deep-frying [13] A comparative quality evaluation of white yam (Dioscorea Rotundata) and Wa-
cycle. Research in this area could be expanded to include other types ter Yam (Dioscorea Alata) chips as African fries, Adv. Nutr. Food Sci. (2017),
of cooking oil while taking additional oil quality parameters, such as doi:10.33140/anfs/02/01/00005.
[14] B.H.H. Goh, C.T. Chong, Y. Ge, H.C. Ong, J.-H. Ng, B. Tian, V. Ashokkumar, S. Lim,
trans-fatty acids, p-anisidine, antioxidants, atherogenic (AI) and throm- T. Seljak, V. Józsa, Progress in utilisation of waste cooking oil for sustainable
bogenic (TI) indexes, into account. biodiesel and biojet fuel production, Energy Convers. Manag. 223 (2020) 113296,
doi:10.1016/j.enconman.2020.113296.
Funding [15] J.W.R. Chong, K.S. Khoo, G.Y. Yew, W.H. Leong, J.W. Lim, M.K. Lam, Y.C. Ho,
H.S. Ng, H.S.H. Munawaroh, P.L. Show, Advances in production of bioplastics by
microalgae using food waste hydrolysate and wastewater: a review, Bioresour. Tech-
The research did not receive any funding. nol. (2021), doi:10.1016/j.biortech.2021.125947.
[16] G. Appiah, S.K. Tulashie, E.E.A. Akpari, E.R. Rene, D. Dodoo, Biolubricant produc-
Declaration of Competing Interest tion via esterification and transesterification processes: current updates and perspec-
tives, Int. J. Energy Res. 46 (2022) 3860–3890, doi:10.1002/er.7453.
[17] J.W.R. Chong, G.Y. Yew, K.S. Khoo, S.H. Ho, P.L. Show, Recent advances on food
The authors declare that they have no known competing financial waste pretreatment technology via microalgae for source of polyhydroxyalkanoates,
interests or personal relationships that could have appeared to influence J. Environ. Manage. (2021), doi:10.1016/j.jenvman.2021.112782.
[18] International Standard Organization, Animal and vegetable fats and oils - determi-
the work reported in this paper.
nation of acid value and acidity (ISO 660:2009), Int. Stand. (2010).
[19] S.S. Nielsen, Moisture content determination, in: 2017: pp. 105–115. 10.1007/978-
CRediT authorship contribution statement 3-319-44127-6_10.
[20] E. Schulte, Micromethod for the gravimetric determination of polar components
in frying fats with ready for use columns, Eur. J. Lipid Sci. Technol. (2000)
Daniel Dodoo: Conceptualization, Project administration, Supervi- 10.1002/1438-9312(200009)102:8/9<574::aid-ejlt574>3.0.co;2-f.
sion, Investigation, Resources, Methodology, Writing – original draft, [21] AOCS official method Cd 8b-90, peroxide value acetic acid-isooctane method, Off.
Writing – review & editing, Formal analysis, Validation, Visualiza- Methods Recomm. Pract. Am. Oil Chem. Soc. (2003).
[22] AOCS Official AOCS, Method Cd 1c-85. Calculated iodine value, Off. Methods
tion. Francis Adjei: Methodology, Resources, Data curation, Investi-
Recomm. Pract. Am. Oil Chem. Soc. (2009).
gation, Writing – review & editing. Samuel Kofi Tulashie: Project ad- [23] R. Romaniello, A. Leone, G. Peri, Measurement of food colour in L∗ a∗ b∗ Units from
ministration, Supervision, Writing – review & editing. Kingsley Enoch RGB digital image using least squares support vector machine regression, J. Agric.
Adukpoh: Methodology, Resources, Data curation, Investigation, Writ- Eng. (2015), doi:10.4081/jae.2015.482.
[24] J.A. Cayuela, K. Yousfi, M. Carmen Martínez, J.M. García, Rapid determination of
ing – review & editing. Redeemer Kofi Agbolegbe: Methodology, olive oil chlorophylls and carotenoids by using visible spectroscopy, JAOCS J. Am.
Resources, Data curation, Writing – review & editing. Kokoutse Oil Chem. Soc. (2014), doi:10.1007/s11746-014-2515-1.
Gawou: Investigation, Formal analysis, Writing – review & editing. [25] J.D. McCurry, GC Analysis of Total Fatty Acid Methyl Esters (FAME) and Methyl
Linolenate in Biodiesel Using the Revised EN14103 : 2011 Method, Agil. Technol,
Gloria Pokuaa Manu: Methodology, Data curation. 2011.
[26] J.K. Quansah, F.K. Saalia, L. Abbey, G.A Annor, Performance of yam as an alternative
Acknowledgements to frozen potato French fries, Nat. Sci. (2010).
[27] Y. Touré, C. Nindjin, Y. Brostaux, G. Amani, M. Sindic, Development of frozen-fried
yam slices: optimization of the processing conditions, Afr. J. Food Agric. Nutr. Dev.
The authors are thankful to the Kwame Nkrumah University of Sci- (2013), doi:10.18697/ajfand.57.11490.
ence and Technology, (Kumasi, Ghana) and the University of Ghana (Ac- [28] A. Kaur, B. Singh, A. Kaur, N. Singh, Changes in chemical properties and oxida-
cra, Ghana) for assistance with the characterization and access to their tive stability of refined vegetable oils during short-term deep-frying cycles, J. Food
Process. Preserv. (2020) 44, doi:10.1111/jfpp.14445.
laboratories. Daniel Dodoo gratefully acknowledges Thomas Dodoo [29] Q. Zhang, A.S.M. Saleh, J. Chen, Q. Shen, Chemical alterations taken place during
(University of Mines and Technology, Ghana) for his graphical design deep-fat frying based on certain reaction products: a review, Chem. Phys. Lipids.
support. 165 (2012) 662–681, doi:10.1016/j.chemphyslip.2012.07.002.

13
D. Dodoo, F. Adjei, S.K. Tulashie et al. Measurement: Food 7 (2022) 100035

[30] J. Li, W. Cai, D. Sun, Y. Liu, A quick method for determining total polar com- nol. 38 (2017) 142–147, doi:10.1590/1678-457x.02217.
pounds of frying oils using electric conductivity, Food Anal. Methods (2016), [41] M. Akrofi Anang, D. Dodoo, B.-S. Ntiri, R. Zugle, G. Ellen Fynn, Improving the quality
doi:10.1007/s12161-015-0324-2. of locally produced vegetable oils in ghana using zeolite ZSM-11, Chem. Sci. Int. J.
[31] C. Turek, F.C. Stintzing, Stability of essential oils: a review, Compr. Rev. Food Sci. (2020) 1–13, doi:10.9734/csji/2019/v28i430150.
Food Saf. (2013), doi:10.1111/1541-4337.12006. [42] N. Vlachos, Y. Skopelitis, M. Psaroudaki, V. Konstantinidou, A. Chatzilazarou,
[32] D. Dodoo, S.K. Tulashie, T. Dodoo, F. Kwaw, Assessing the effects of sunlight on the E. Tegou, Applications of Fourier transform-infrared spectroscopy to edible oils,
photooxidation of tropical oils with experimental and computational approaches, Anal. Chim. Acta. (2006), doi:10.1016/j.aca.2006.05.034.
JAOCS, J. Am. Oil Chem. Soc. (2021), doi:10.1002/aocs.12478. [43] F.D. Gunstone, Production and trade of vegetable oils, 2011.
[33] M.D. Guillén, E. Goicoechea, Detection of primary and secondary oxidation prod- 10.1002/9781444339925.ch1.
ucts by Fourier transform infrared spectroscopy (FTIR) and 1H nuclear magnetic [44] D.B. Min, J.M. Boff, Chemistry and reaction of singlet oxygen in foods, Compr. Rev.
resonance (NMR) in sunflower oil during storage, J. Agric. Food Chem. 55 (2007) Food Sci. Food Saf. (2002), doi:10.1111/j.1541-4337.2002.tb00007.x.
10729–10736, doi:10.1021/jf071712c. [45] J.M. DeMan, Principles of food chemistry, 1999.
[34] M.A. Poiana, E. Alexa, M.F. Munteanu, R. Gligor, D. Moigradean, C. Mateescu, Use of [46] C.C. Akoh, D.B. Min, Food Lipids Chemistry, Nutrition, and Biotechnology Second
ATR-FTIR spectroscopy to detect the changes in extra virgin olive oil by adulteration Edition, Revised and Expanded, 2002.
with soybean oil and high temperature heat treatment, Open Chem 13 (2015) 689– [47] E. Choe, D.B. Min, Mechanisms of Antioxidants in the Oxidation of Foods, Compr.
698, doi:10.1515/chem-2015-0110. Rev. Food Sci. Food Saf. (2009), doi:10.1111/j.1541-4337.2009.00085.x.
[35] E. Zahir, R. Saeed, M.A. Hameed, A. Yousuf, Study of physicochemical [48] E.N. Frankel, Secondary products of lipid oxidation, Chem. Phys. Lipids. (1987),
properties of edible oil and evaluation of frying oil quality by Fourier doi:10.1016/0009-3084(87)90045-4.
Transform-Infrared (FT-IR) Spectroscopy, Arab. J. Chem. 10 (2017) S3870–S3876, [49] E.N. Frankel, Hydroperoxide decomposition, in: Lipid Oxid., 2012.
doi:10.1016/j.arabjc.2014.05.025. 10.1533/9780857097927.67.
[36] L. Xu, T. Fei, Q. Li, X. Yu, L. Liu, Qualitative analysis of edible oil oxidation by FTIR [50] E.N. Frankel, W.E. Neff, E. Selke, Analysis of autoxidized fats by gas
spectroscopy using a mesh “cell, Anal. Methods (2015), doi:10.1039/c5ay00438a. chromatography-mass spectrometry: VII. Volatile thermal decomposition products
[37] A. Singh, S. Ahmad, A. Ahmad, Green extraction methods and environmental appli- of pure hydroperoxides from autoxidized and photosensitized oxidized methyl
cations of carotenoids -a review, RSC Adv., 2015, doi:10.1039/c5ra10243j. oleate, linoleate and linolenate, Lipids (1981), doi:10.1007/BF02534950.
[38] N. Achir, V.A. Randrianatoandro, P. Bohuon, A. Laffargue, S. Avallone, Kinetic study [51] S.K. Tulashie, D. Dodoo, G. Appiah, F. Kotoka, K.E. Adukpoh, Oil produced from
of 𝛽-carotene and lutein degradation in oils during heat treatment, Eur. J. Lipid Sci. Ghana cocoa bean for potential industrial applications, Ind. Crops Prod. 177 (2022)
Technol. (2010), doi:10.1002/ejlt.200900165. 114426, doi:10.1016/j.indcrop.2021.114426.
[39] C. Sánchez, A.B. Baranda, I.M. De Marañón, The effect of high pressure and high [52] A. Gouveia De Souza, J.C. Oliveira Santos, M.M. Conceição, M.C. Dantas Silva,
temperature processing on carotenoids and chlorophylls content in some vegetables, S. Prasad, A thermoanalytic and kinetic study of sunflower oil, Braz. J. Chem. Eng.
Food Chem. (2014), doi:10.1016/j.foodchem.2014.04.041. (2004), doi:10.1590/s0104-66322004000200017.
[40] D.T. de Almeida, F.M. Curvelo, M.M. Costa, T.V. Viana, P.C. de Lima, Oxidative
stability of crude palm oil after deep frying akara (Fried Bean Paste), Food Sci. Tech-

14