AN fuels, medical purposes, lubrication, manufacture of





1.1.  Theoretical background


Oils are an essential part of
our daily life. They are viscous substances with high carbon and hydrogen
content. They are of animal, vegetable or petrochemical origin and are used for
fuels, medical purposes, lubrication, manufacture of many types of materials
and most important for this investigation, they are used for food. The oils
used for food or the preparation of food are for example sunflower, olive,
pumpkin, palm, soybean, and canola oil (Wikipedia, 2017). All these oils have similar
physical and chemical properties, which suggests that their chemical
composition is similar. Oils are built from glycerol and fatty acids, which can
in short be divided on saturated and unsaturated fatty acids. Saturated fatty
acids contain no double bonds, which means that they are more unreactive than
unsaturated fatty acids. These contain, in addition to single bonds, also
double bonds. They can be mono- (one double bond) or polyunsaturated (multiple
double bonds). A double bond is built from a sigma and a pi bond. Sigma bonds
are a result of a heads-on overlap between different orbitals. Pi bonds on the
other hand, are a result of a sideways overlap between the non-bonding
dumb-bell shaped orbital, which results in a cloud of electron density
concentrated above and below the sigma bond. This cloud of electron density
makes the bonds more reactive and more susceptible to a process called
oxidation. There are several factors affecting the rate of oxidation. These
include temperature, light, availability of oxygen, and the presence of
humidity and metals such as iron (Miller, 2014).


For this investigation, extra
sunflower and extra virgin olive oil were used. There are a few differences in the vitamins that are
present in each oil. Olive oil in general contains more vitamin K, whereas
sunflower oil in general contains more vitamin E. Also, sunflower oil contains
only traces of different minerals, whereas olive oil contains minerals such as
iron, potassium, sodium and calcium (Melodie, 2017). However, the most important differences for this investigation
occur in fat differences between both oils. The crucial part is the content of
mono- and polyunsaturated fatty acids. The sunflower oil contains more
monounsaturated fatty acids, but less polyunsaturated fatty acids than olive
oil. This means that olive oil is more susceptible to oxidation than sunflower
oil. Because of this and the fact that UV light is also a factor affecting the rate
of oil oxidation, the rate of oxidation of both oils will be investigated at
different lengths of exposure to UV light.


When an old oil smells and has
an odd flavour, it is due to the process of oxidation. Oil oxidation is a
series of chemical reactions that degrade the quality of an oil. One of the
ways of measuring it is to determine its peroxide value. The peroxide value is
a number that expresses the quantity of hydroperoxide (denoted as ROOH)
contained in 1000g of the substance. The lower the peroxide value the higher
the quality of oils. The whole theoretical background behind the determination
will be explained in the following section.


The process of determination
involves mixing the oil with glacial acetic acid, chloroform, potassium iodide,
starch and titrating this solution with sodium thiosulfate solution (Na2S2O3).
But in order to even determine the amount of hydroperoxide, the first condition
is its generation and this occurs in a process called auto-oxidation. This is a
spontaneous non-enzymatic reaction with oxygen that among other substances
produces hydroperoxide. It is a chain reaction of free radicals and consists of
3 stages (Choe and Min, 2006):

·    Initiation: auto-oxidation of oils requires fatty acids (denoted as
RH) to be in radical forms. Since the reaction with atmospheric oxygen is
thermodynamically unfavourable, a hydrogen atom first has to be removed in
order to produce hydrogen (denoted as H·) and fatty acid radicals (denoted as R·). The most easily removed hydrogen atoms are near the
double bonds of a fatty acid, especially if this hydrogen atom is bonded to a
carbon between 2 double bonds. The scheme below shows the structure of a fatty
acid (the red circle shows where the hydrogen atom is most easily removed and
where the initiation could take place):




The initiation process can be induced by
heat, metal catalysts or in the case of this investigation, by UV light:


à R·
+ H·


·    Propagation: the fatty acid radical can now react with atmospheric
oxygen and a peroxy radical (denoted as ROO·) is produced. This occurs very quickly at normal oxygen
pressure. Consequently, since the concentration of the fatty acid radicals is
much lower than that of peroxy radicals, the peroxy radicals remove hydrogen
from other fatty acids. A hydroperoxide (ROOH) and another fatty acid radical
is formed:


R· + O2 à ROO·

ROO· + RH à


The hydroperoxides are
products of primary oxidation and are stable at room temperature and in the
absence of metals. However, at higher temperatures and in the presence of
metals, secondary oxidation occurs and they are decomposed to aldehydes,
ketones, acids, alcohols etc. These substances give the oil an odd smell and a
rancid flavour.


·    Termination of
auto-oxidation: when two free radicals
react, the auto-oxidation is terminated. There are two options of termination
reactions, where either esters (denoted as ROOR) or longer chains of fatty
acids are produced (denoted as RR):


ROO· + R·

R· + R·
à RR


When the hydroperoxide is
formed, we can effectively determine the oil’s peroxide value. Even though the
method is explained as the determination of the amount of hydroperoxide in 1000
g of the substance, the amount of hydroperoxide is determined indirectly by
measuring the quantity of iodine liberated from potassium iodide. This occurs
in the first step of the method called the generation of iodine. Starch is an
indicator for iodine and is therefore added to the mixture and it turns purple.
CH3COOH is the glacial acetic acid added to the oil before




ROOH + 2HI à ROH + H2O
+ I2 + starch indicator


The next step of the method is the
titration step. The mixture is titrated with an experimentally predetermined
concentration of sodium thiosulfate. The reaction explaining how this process
happens is as follows:


+ 2Na2S2O3 à Na2S4O6
+ 2NaI (colourless)



In short, the volume of sodium
thiosulfate proportional to the amount of iodine formed and this is
proportional to the amount of hydroperoxide formed:


V (Na2S2O3)
a I2 formed a
the amount of hydroperoxide


It was assumed
that iodine reacted with all the formed hydroperoxides so that the final step of the
method, the calculation of the peroxide value, could be done. This is done
according to the following equation:




Where VT is the volume of the sodium thiosulfate
(in mL) consumed in the actual test, VB
is the volume of the sodium thiosulfate (in mL) consumed in the blank
test, c is the concentration of the
sodium thiosulfate solution (in mol/L) and m
is the mass of the sample (in g) used.




2.1.  Research question


Based on all the
information about oil a research question as follows was formed: How
does the length of exposure to UV light affect the oxidation of extra virgin olive
and extra sunflower oil determined using peroxide value?


The hypotheses
involved in this investigation are:

·    The longer the extra
virgin olive and the extra sunflower oil are exposed to the UV light, the
higher the peroxide value.

·    Extra virgin olive oil will have higher peroxide values
than extra sunflower oil.






2.2.  Variables


Controlled variables: the experimentation was carried out in the same room and at
approximately the same room temperature. The chemicals used were exactly the
same. This was achieved by mixing the glacial acetic acid and the chloroform in
the right ratios and the quantity was calculated prior to the start of
experimentation. The length of exposure of oils to UV light was exactly the
same. This was made sure by timing the length they spent in a cupboard. While
being in the cupboard, the oils were only affected by the UV light. The UV
light itself was also the same for all trials. Furthermore, the type of
air-sealed transparent flasks, in which the oils were exposed to UV light, were
exactly the same.


Independent variables: the length of exposure to UV light, the type of oil


Dependent variables: the peroxide value of each oil



2.3.  Method


The method used
in this experiment was titration with sodium thiosulfate and later calculation
of the peroxide value of extra sunflower and extra virgin olive oil. The
peroxide value of the both oils was measured every 30 minutes of exposure to UV
light, the first sample not being exposed to UV light. Titration of each sample
was performed 3 times in order to gain the most accurate results. The detailed
experimental procedure is explained in section 2.3.2.


2.3.1. Apparatus used

Extra virgin olive and extra sunflower

250 mL conical flasks

Two 40 mL beakers

Two funnels

Two 1000 mL volumetric flasks (± 0.4 mL)

Two 250 mL (± 1 mL) and one 30 mL (± 0.5 mL)
measuring cylinders

Two 50 mL burettes (± 0.05 mL)

25 mL pipette (± 0.05 mL)

Two 3 mL droppers

Fume cupboard

Top pan balance (± 0.01 g) (KERN & SOHN GmbH, Balingen, Germany, 2016)


2.3.2. Experimental procedure


Firstly, a preliminary test
with both oils is carried out to determine the appropriate amounts and
concentrations of reagents.

o  The most suitable concentration of sodium thiosulfate is determined
to be 0.01 mol/dm3 for the extra sunflower oil and 0.05 mol/dm3
for the extra virgin olive oil.

Secondly, a blank test (without
the oil) is carried out. No iodine is formed, so the solution does not turn purple
(as with the oil), indicating that there are no fatty acids present needed for
the reaction. The volume of the sodium thiosulfate needed to react with iodine
is therefore taken to be zero.


Then, all the substances needed
in the experimentation are prepared based on the calculation of how much of
each substance will be needed. This includes:

o  Mixture of glacial acetic acid and chloroform in 3:2 ratio
respectively. In this investigation 1 L of the mixture was prepared.

o  Saturated mixture of potassium iodide (KI).

o  The starch test solution and sodium thiosulfate solutions of 0.01
mol/dm3 and 0.05 mol/dm3 are prepared by a laboratory
assistant. 1 L of both solutions was prepared for this investigation.

o  The oils are put in a transparent flask and put in a cupboard with
UV lamp.

Lastly, the actual experiment
was carried out:

o  5 g of oil is accurately weighed in a conical flask.

o  25 mL of mixture of glacial acetic acid and chloroform is added and
shaken to dissolve.

o  0.5 mL of saturated potassium iodide solution is added with a

o  The mixture is shaken for exactly 1 minute (timed with a stopwatch).

o  30 mL of distilled water is added.

o  2.5 mL of starch test solution is added with a dropper and the
solution turns purple.

o  It is titrated with sodium thiosulfate until the purple colour is
discharged (until the solution turns back to the same yellow-ish colour as before
the addition of starch).

o  This procedure is repeated 3 times for each sample of the oil.

The peroxide values are
calculated for each of the samples.



2.3.3. Safety and environmental considerations


Because of the
fact that this method was performed with highly corrosive and hazardous glacial
acetic acid and chloroform, protective gloves, goggles and lab coat were worn
at all times. The mixing of these two solutions was also performed in a fume
cupboard to prevent any health risks via inspiration of the fumes. Furthermore,
the titration was performed in a well-ventilated chemistry laboratory.


The volume of
chloroform and glacial acetic acid used was also large and therefore it was
calculated how much of it was needed before the start of experimentation. Since
there are also a lot of different substances mixed in the course of experiment,
they were put in a special bottle to be properly disposed by the lab assistant
without causing damage to the environment.




3.1.  Raw data and calculations


Table 1 and
Table 2 show the experimentally collected values needed for the calculation of
the peroxide value. The peroxide values are already calculated according to the
below explained procedure.


Table 1. All the collected and calculated data for extra
sunflower oil.




2. All the collected and calculated data for extra virgin olive oil.




3.2.  Processing


After the
experimental part all the peroxide values were calculated according to the
already mentioned equation:




VT = the volume of
the sodium thiosulfate consumed in the actual test (mL)

VB = the volume of
the sodium thiosulfate consumed in the blank test (mL)

c = the concentration of the
sodium thiosulfate solution (mol/L)

m = the weight of oil (g)


The volume of the sodium
thiosulfate consumed in the blank test was always zero, because the solution
did not turn purple after the addition of starch test solution. It was
performed one time. Therefore, only the volume of sodium thiosulfate used in the
actual test was used and was denoted by DV.
The actual equation used was consequently:




A graph
for each respective oil was made based on the calculated mean peroxide value
and its relative uncertainty was used for error bars. Graph 1 and Graph 2 show
the peroxide value for each type of oil at different time of exposure to UV



Graph 1.
The effect of the length of exposure to UV light on peroxide value of extra sunflower



Graph 2.
The effect of the length of exposure to UV light on peroxide value of extra
virgin olive oil.

The second
peroxide value in Graph 4 was left out, because it was an outlier with 91.0
mmol/g. The peroxide values presented on both graphs were then connected into a
line, which is shown in Graph 3 and Graph 4.




Graph 3. Peroxide value of
the extra sunflower oil connected into a line to show the effect of the
exposure to UV light on it.




Graph 4. Peroxide value of
the extra virgin olive oil connected into a line to show the effect of the
exposure to UV light on it.

3.3.  Uncertainties


Table 1 and
Table 2 show the data collected and calculated during experimentation. This
section will show how this was done.

The following
calculations done are examples of how all the results were calculated. The
values of extra sunflower oil trail 1 at 0 minutes of UV light exposure will be
taken as an example.


The absolute
uncertainty for ?VT (thiosulfate) is ± 0.1 and is calculated from the sum of absolute uncertainties of VT
(thiosulfate) – before titration and ?VT (thiosulfate) – after


The relative
uncertainty of DV(thiosulfate) was
calculated as follows:




The relative
uncertainty of the mass of oil calculated as follows:




relative uncertainty for the peroxide value was calculated as follows:


this the average peroxide value was calculated for each respective time of





relative uncertainty of the mean peroxide value was calculated as the average
of the relative uncertainties of each respective peroxide value:





The absolute uncertainty of the mean peroxide value was calculated
as a product of the relative uncertainty and the mean peroxide value:



final result is as follows:







The results of experimentation
show that the longer the oil is exposed to UV light, the higher the peroxide
value, thus lower the quality of both oils. This finding confirms the hypothesis
formed in the beginning based on facts about fatty acids and their

The peroxide
value increases exponentially, which is also in line with literature. Reactions
such as a bimolecular reaction between a hydroperoxide (ROOH) and a hydrocarbon
(RH) (shown below)
lead to formation of more free radicals and they increase the rate of
auto-oxidation (Brodnitz, 1968).


+ R· + H2O


Extra virgin
olive oil has a higher peroxide value than extra sunflower oil, which confirms
the hypothesis generated at the beginning. Based on the label on the bottle
itself, it contains more poly-unsaturated fatty acids than the extra sunflower
oil, which means more double bonds for auto-oxidation and thus higher peroxide
value. Not only that, the rate of auto-oxidation of extra virgin olive oil is
also greater, which is shown by a tangent to Graph 3 and Graph 4 at      t = 60 minutes. The extra sunflower oil
peroxide value rate has a tangent with a coefficient of approximately 0.0082,
while the extra virgin olive oil peroxide value rate has a much greater tangent
coefficient of approximately 0.1188. The greater coefficient means greater slope of the tangent,
which in turn means a greater rate of auto-oxidation.


Graph 5. A tangent to the curve at t = 60 minutes for
the peroxide value graph of extra sunflower oil.


Graph 6. A tangent to the curve at t = 60 minutes for
the peroxide value graph of extra virgin olive oil.



This investigation
gives sufficient evidence into oil oxidation caused by UV light and the results
are also supported by some research studies such as Brodnitz (1968) and Choe and Min (2006). This can be considered one of the most significant strengths of
this investigation. The other strength of this investigation is the relatively
small percentage uncertainty of the peroxide values of both oils, since they
are 4% and 1% for the extra sunflower and the extra virgin olive oil
respectively. This means that the results are not only supported by other
research studies, but also accurate.


On the other
hand, there are also some weaknesses. The applicability to oil oxidation in
real life can be questioned, since oils are usually stored in dark bottles to
prevent the light from causing oxidation. Even without darkened bottles, oils
are usually not kept under a source of UV light or under the sun before they
are consumed. The real-life oil oxidation could be investigated by measuring
the peroxide values of different oils over a longer period of time in their normal
environment; usually in a kitchen cupboard in a darkened bottle. Also, as mentioned
before, one value in Graph 4 was left out due to a high deviation from other
values, which had to occur due to an error during the course of
experimentation. All 3 trials at 30 minutes of exposure of the extra virgin
olive oil are higher than expected. Since all 3 trials are high, this is not a
random uncertainty and it most likely did not occur during titration. Therefore,
this was most likely a systematic uncertainty and a consequence of a mistake in
the experimental procedure for this specific specimen of the extra virgin olive
oil. The most significant improvement would be an accurate determination of
this value. This could be made by repeating the whole set of measurements for
the extra virgin olive oil.


Furthermore, this
investigation has some limitations. The most significant ones are all the assumptions
made during the course of experimentation. These were, as mentioned before,
that white light did not induce any oxidation, even though both oils were
exposed to it for a short period of time before the start of each trial and
that the iodine reacted with all the hydroperoxides formed. Lastly, titration
is also a method dependent on human perception and may therefore be subject to



Brodnitz, M. H. (1968) ‘Autoxidation of
saturated fatty acids’, Journal of Agricultural and Food Chemistry,
16(6), pp. 994–999.


Choe, E. and Min, D.
B. (2006) ‘Mechanisms and factors for edible oil oxidation’, Comprehensive
Reviews in Food Science and Food Safety, 5(4), pp. 169–186.


Melodie, A. (2017) What
Is the Difference Between Olive Oil & Sunflower Oil?,
Available at:
(Accessed: 18 November 2017).


Miller, M. (2014)
‘Oxidation of food grade oils’, Plant and food Research, pp. 1–2.
Available at: 101.pdf (Accessed:
1 October 2017).


Wikipedia (2017) Oil.
Available at: (Accessed: 16 November 2017).