Chemistry Food Acids

Rationale

For thousands of years, people have been using fats/oils for cooking, these acting as a medium between the heat energy-source and the food being cooked. In modern times, science has come to experiment and found useful discoveries about these substances we humans dearly love, discoveries which has benefitted us in terms of our health. The purpose of this investigation is to find out how stable unsaturated fats are compared to saturated fat when these are exposed to great heat-energy and more specifically water. By this information, this can be used as a tool by understanding how certain fats or oils behave at certain conditions and the chemistry behind it, and improve our awareness when it comes to selecting the “right” type of oil in our cooking. Fats and oils (lipids) are hydrophobic organic compounds, which are formed as a result of triesters of glycerol and three fatty acid chains (carboxylic-acids). These undergo a condensation reaction to form a triglyceride-molecule and the by-product, three molecules of water [1]. The general structure of triglyceride can be presented below:

Where R, R’ and R’’ represent particular carbon fatty acid chains. The fatty acid composition is the most important characteristic of the triglyceride. The physical and chemical properties of particular triglycerides depend on the nature and arrangement of these fatty acid residues, and how these react with each other and external substances. In the saturated fatty acid-molecule there are no double-bonded carbons. Hence, all the carbons in the structure are saturated with hydrogen-atoms. All of these hydrogen atoms attached to the backbone keep saturated fats intact, but also help protect it against oxidation and the binding of free radicals (molecules which have unshared pair of electron on their outer valence-shell, making these very reactive) [2]. In contrast, the unsaturated fatty acid-molecule has at least one double-bonded carbon in the structure. In the unsaturated fatty acid’s case, two types exist: mono-unsaturated fatty acid and poly-unsaturated fatty acid.

The difference between these is that the mono-unsaturated fatty acid chain only contain one cis double-bond between two carbons, hence the term mono. These types have less than the maximum possible number of hydrogen atoms in their molecule. As a result, these become more reactive to free radicals and the process oxidation (lipid peroxidation, See Figure 1.0 and 1.05), and increase the probability of a particular oil of becoming rancid. This changes the chemical properties of the molecules, particularly its odour, taste and its safety for consumption. As the poly-unsaturated fatty acids contain more than one cis double-bond, these are even more reactive to the external environment (the more double-bonds a fatty acid chain contains, the more reactive it is to light, oxygen and heat [3]) and more importantly; lipid peroxidation (See Figure 1.0 and 1.05) [5.6] As a result, the properties between these differ slightly in terms of boiling-point and melting-point; one is more stable than the other.

Above: Figure 1.0: Mechanism involved in the formation of oxirane ring (epoxide): direct attack by a hydroperoxyl radical on the vinylic carbons, leading to the formation of an oxirane group and an alkoxyl radical [5]

Figure 1.05: As a result of the oxirane-ring, the particular section of the fatty acid chain becomes further oxidized to an aldehyde. This breaks the chain, resulting in shorter fatty acid chains [6] Examples of the mentioned kinds of fatty acid chains can be illustrated below:

Figure 1.1: The structure of a typical saturated triglyceride As saturated fats do not contain any double-bonds, they tend to be very stable in comparison to the other mentioned types – contributing to a greater melting point. Hence, solid fats are mainly composed of saturated fatty acids. The straight, crisscross molecular structure of saturated fatty acids (Figure 1.0) makes it easy for these to align on top of each other, with the presence of dispersion force within and between corresponding fatty acid chain, causing these to have an adhesive, compact structure in general. With the non-polar “carbon-wall” (glycerol portion of the structure) and the forces, these further resist water-molecules of reaching their main target: the ester-bond (C-O bond) connecting the alcohol and carboxyl-group (Figure 1.2 below). Consequently, this contributes to the strong stability of saturated fats.

Figure 1.2: The red highlighted areas are the ester-bonds the water-molecules target to hydrolyse. These hydrolyse a particular triglyceride into glycerol and carboxylic acids (free fatty acids). However, in mono-unsaturated fatty acids the presence of the cis double-bond causes the molecule to bend, having a slightly different arrangement. The cis double results in hydrogen-atoms being on the same side of the bond, and the carbon-groups on
the opposite, as following (Figure 1.3):

Figure 1.3: Cis double-bond.
This result the fatty acid chains having an irregular structured (Figure 1.4) in comparison to the saturated fats:

Figure 1.4: The structure of a typical mono-unsaturated triglyceride Observing Figure 1.4, it can be seen that the distance between the top fatty acid chain and the second is much greater in comparison to the distance between the two bottoms. Thus, the dispersion-force is weaker as a result of this; structure is less stable. Furthermore, the alignment of mono-unsaturated fats becomes more complicated as a result of this arrangement; they do not superimpose well. As a result of these factors, this simplifies for H20-molecules of approaching and breaking the ester-bond; hence, more fatty-acids are separated. In poly-unsaturated fatty acids, the presence of at least two cis-double bonds causes the structure of the fatty acid chain to be even more complicated (Figure 1.4):

Figure 1.5: The structure of a typical poly-unsaturated triglyceride Observing Figure 1.5, it can be seen that the distance between the bottom two fatty acid chains is even more greater than what it has been seen in Figure 1.1 and Figure 1.4. Also, these will ineffectively superimpose, and stack together. Thus, this simplifies it even further for H20-molecules to approach and break the ester-bonds. Consequently, more fatty acid chains will separate, thus increasing the acidity-concentration of the oil. By using this information, it has been hypothesised that unsaturated fats will hydrolyse more, to produce greater acidity, in comparison to saturated fats. It is important to understand that poly-unsaturated fats will hydrolyse the most of these different types, due to the greater number of double-bonds and its complicated structure.

Topic: Usage of different types of oils in cooking.

Aim: To measure the free fatty acid-concentration of different types of oils after exposed to heat and water. Hypothesis: Unsaturated fats are more likely to hydrolyse during cooking than saturated fats of similar chain-length. Materials

Refer to Appendix 1.
Method

In this experiment, different types of oils were fried at the temperature range 170-190 degrees Celsius, at the same time being exposed to a particular amount of water. Periodic samples were taken in each frying to later be used to determine the change in acidity, via titration. The methods are described in details in Appendix 2. These were the controlled variables in the experiment:

* Temperature
* Amount of oil fried
* Amount of wet cloths exposed to the oils
The different types of oils (saturated, mono-unsaturated, poly-unsaturated; coconut oil, olive oil and sunflower oil) were the independent variables of the experiment.

Results

At the end of frying the oils, the amount of water diffused into the cloths was calculated by subtracting the initial amount of water in the beaker with the final amount of water; as mentioned in Method. In each case, this was then divided by the number of wet cloths exposed to the oil (40): Amount of water contained in each cloth, average (grams)|

Sunflower oil (poly-unsaturated)*| Olive oil (mono-unsaturated)| Coconut oil (saturated)| 4.08 (first 20) ± 0.05 (±1.23 %)**| 4.56 ± 0.05 (±1.1%)**| 3.64 ± 0.05 (±1.37%)**| 4.82 ± 0.05 (±1.04 %) (remaining 20)| |

Table 1.1

*Initially, it was thought that exposing the oil to 20 wet cloths would be enough; however, this was seen as insufficient. As a result, another 20 clothes had to be soaked in 180 grams of water. The amount of water present in these 20 cloths was slightly different to previous, as seen on Table 1.1. ** Relative error placed brackets. This will be apparent in all tables and calculations. Using this information, the amount of water exposed to each sample could then be calculated, by multiplying the number of cloths a particular oil-sample was exposed to by the corresponding values in Table 1.1: Amount of water exposed to each oil-sample, average (grams)| Sample| Sunflower oil| Olive oil| Coconut oil|

Initial (0 wet cloth)| 0| 0| 0|
1 (10 wet cloths)| 40.8 (±1.23 %)| 45.6 (±1.1%)| 36.4 (±1.37%)| 2 (20 wet cloths)| 81.6 (±1.23 %)| 91.2 (±1.1%)| 72.8 (±1.37%)| 3 (30 wet cloths)| 129.8 (±1.04 %)| 136.8 (±1.1%)| 109.2 (±1.37%)| 4 (40 wet cloths)| 178 (±1.04 %)| 182.4 (±1.1%)| 145 (±1.37%)| Total| 178 (±1.04 %)| 182.4 (±1.1%)| 145 (±1.37%)|

Table 1.2
The end-point titre-value required to neutralise each sample was also be formulated into Table 1.3: The end-point titre-value of each oil-sample (mL) NaOH|
Sample| Sunflower oil| Olive oil| Coconut oil|
Initial | 0.7 ± 0.05 (±7.14%)| 2.3 ± 0.05 (±2.17%)| 1.9 ± 0.05 (±2.63%)| 1 | 0.7 ± 0.05 (±7.14%)| 2.2 ± 0.05 (±2.27%)| 1.9 ± 0.05 (±2.63%)| 2 | 0.75 ± 0.05 (±6.67%)| 2.3 ± 0.05 (±2.17%)| 2.2 ± 0.05 (±2.27%)| 3 | 1.0 ± 0.05 (±5%)| 2.4 ± 0.05 (±2.08%)| 1.9 ± 0.05 (±2.63%)| 4| 1.8 ± 0.05 (±2.78%)| 3.0 ± 0.05 (±1.67%)| 2.0 ± 0.05 (±2.5%)| Table 1.3

This could then be formulated into a more complex table, including the mass of each sample:

Types of oils|

Sunflower oil (poly-unsaturated)| Olive oil (mono-unsaturated)| Coconut oil (saturated)| Sample| Mass of Sample (grams)| Amount of H20 exposed to sample (grams)| Titre NaOH (mL)| Sample| Mass of Sample (grams)| Amount of H20 exposed to sample (grams)| TitreNaOH(mL)| Sample| Mass of Sample (grams)| Amount of H20 exposed to sample (grams)| TitreNaOH (mL)| Initial| 33.8 (±0.15%) | 0| 0.7 ± 0.05 (±7.14%)| Initial| 21.4 (±0.23%)| 0| 2.3 ± 0.05 (±2.17%)| Initial| 17.7 (±0.28%)| 0| 1.9 ± 0.05 (±2.63%)| 1| 16.8(±0.3%) | 40.8 (±1.23 %)| 0.7 ± 0.05 (±7.14%)| 1| 20.0 (±0.25%)| 45.6 (±1.1%)| 2.2 ± 0.05 (±2.27%)| 1| 16.5 (±0.30%)| 36.4 (±1.37%)| 1.9 ± 0.05 (±2.63%)| 2| 17 (±0.29%)| 81.6 (±1.23 %)| 0.75 ± 0.05 (±6.67%)| 2| 20.3 (±0.25%)| 91.2 (±1.1%)| 2.3 ± 0.05 (±2.17%)| 2| 17.4 (±0.29%)| 72.8 (±1.37%)| 2.2 ± 0.05 (±2.27%)| 3| 18.4 (±0.27%)| 129.8 (±1.23 %)| 1.0 ± 0.05 (±5%)| 3| 21.3 (±0.23%)| 136.8 (±1.1%)| 2.4 ± 0.05 (±2.08%)| 3| 18.8 (±0.27%)| 109.2 (±1.37%)| 1.9 ± 0.05 (±2.63%)| 4| 20.5 (±0.24%) | 178 (±1.23 %)| 1.8 ± 0.05 (±2.78%)| 4| 26.5 (±0.19%) | 182.4 (±1.1%)| 3.0 ± 0.05 (±1.67%)| 4| 23.4 (±0.21%)| 145 (±1.37%)| 2.0 ± 0.05 (±2.5%)| Table 1.4

Sunflower oil (poly-unsaturated)| Olive oil (mono-unsaturated)| Coconut oil (saturated)| Sample| Titre NaOH (mL)| Sample| TitreNaOH(mL)| Sample| TitreNaOH (mL)| Initial| 0.7 ± 0.05 (±7.14%)| Initial| 2.3 ± 0.05 (±2.17%)| Initial| 1.9 ± 0.05 (±2.63%)| 1| 0.7 ± 0.05 (±7.14%)| 1| 2.2 ± 0.05 (±2.27%)| 1| 1.9 ± 0.05 (±2.63%)| 2| 0.75 ± 0.05 (±6.67%)| 2| 2.3 ± 0.05 (±2.17%)| 2| 2.2 ± 0.05 (±2.27%)| 3| 1.0 ± 0.05 (±5%)| 3| 2.4 ± 0.05 (±2.08%)| 3| 1.9 ± 0.05 (±2.63%)| 4| 1.8 ± 0.05 (±2.78%)| 4| 3.0 ± 0.05 (±1.67%)| 4| 2.0 ± 0.05 (±2.5%)| Section of Table 1.5

Analysing Table 1.5, it can be seen that the greatest errors occurred when measuring the titre in each titration. This was because, as the 20-30 grams of oil-samples were insignificantly acidic, it would not require great values of titre (NaOH) to neutralize these. Hence, low values of titre were obtained. As the scale of the burette measured only to a tenth of a millilitre, the absolute error (± 0.05 ml) was relatively great to the titre obtained. As a result, this resulted in a greater relative error. To decrease the relative errors, greater oil-samples should be obtained. By this, it would require greater amount of the titrant to neutralize the low-acidic oil-samples, thus; relatively, the absolute error would have less effects on the relative errors. If possible, burettes with finer scale should instead be used to minimise the absolute error, thus, decreasing the relative error.

The metal saucepan that was used to fry the oils and cloths had a great surface area; hence, it was easily exposed to external environment (esp. oxygen, light). Frying at approximately 180 degrees Celsius, the fatty acid chains that had been hydrolysed were volatile in this extreme condition. It is assumed that most of these did not remain in the oils, and was evaporated with the steam. Furthermore, various fatty acid chains may have been further oxidized, especially the unsaturated fats. Thus, instead of the actual fatty acid chains being present in the oil, these were degraded to short-chain compounds. As a result, this further affects the results – providing greater inaccuracy in determining the actual acidity of each sample. To minimise the evaporation of fatty acid chains and the rate of oxidation, during the frying-procedure, the metal saucepan should be enclosed in a condenser, in which the steam contained with fatty acids becomes condensed to liquid form.

This can then be returned onto the frying oil. Hence, this effectively maintains the lost fatty acid chains, results in more reliable measurements. As samples were taken from the oils, the depth of the oil continuously decreased. On two occasions in the experiment, the thermometer was instead of measuring the temperature of the oil at the bottom, measuring the temperature of the oil-surface; presenting misleading values. The temperature of the oil was manually controlled, maintaining 180 degrees Celsius. As a result of the misleading values, the temperature rose nearby to the smoking points of two particular oils (sunflower and olive oil’s case); around 250-300 Celsius. Thus, the fatty acid chains were oxidized and reacting with free radicals at much greater rate than desired. As a result, the results of the acid concentration of each sample may not be accurate as wanted.

To avoid this mistake, the depth of the frying oil should be deep enough for the thermometer to be submerged in the oil. Diagram 1.1: Acid concentration of Sunflower oil (Polyunsaturated) Observing Diagram 1.1 above, a trend is evident: the greater the amount of water the poly-unsaturated fat (sunflower oil) was exposed to, the more acidic it become. Hence, the unstable poly-unsaturated oil barely resisted to the hydrolysis and chemical changes in its structure, as stated in the rationale. Diagram 1.4: Comparison of the relative change in acidity between sunflower (polyunsaturated) and coconut oil (saturated). Observing Diagram 1.4, it can be seen that, in contrast to the unsaturated sunflower oil, the saturated coconut oil has marginally changed in terms of its acid concentration. In fact, it has in overall improved in quality (by dropping in acid-concentration). However, this can be due to measurement errors.

Table 1.6: Fatty acid composition of various oils, expressed as percentage [4] It has also been speculated that the fatty acid-composition of the coconut oil may have become volatile and evaporated into the external environment. Referring to the review “Coconut Oil: Chemistry, Production and Its Applications” [4] (Table 1.6), approximately 60% of the fatty acid composition of a regular coconut-oil is of 8 to 12-carbon fatty acid-chains. With such low-molecular mass and size relatively to sunflower oil (approximately 90% fatty acid composition is of 18-carbon chain), the separated fatty acid chains of the coconut oil became easily evaporated into the atmosphere. This also explains why the acidity of the sunflower oil never dropped during the frying; the fatty acid chains were heavier and less volatile.

Consequently, this explains the acidity-drop of the coconut oil. In contrast, the graph also shows that the acidity initially increased. However, this identified anomaly can be due to measurement errors caused by titrating. When the pale coconut-oil was titrated, it was more difficult to trace the end-point of it in comparison to the other oils included in this experiment. Thus, greater amount of titre may have instead been added; exceeding the end-point. During the frying, a particular, noticeable smell associated with rancidity was felt. As coconut-oil is a saturated-oil which theoretically should not become rancid so early, this may have been due to the oxidation of other organic substances that may have been added into the oil for preservative reasons. Also, this could have been due to poor purification by the company producing this product, resulting in more impurity being present in the substance.

Hence, these may have contributed to the smell felt during frying or the slight increase in acidity of the coconut oil. As a result, this may have caused the calculations made to indicate the slight increase in acid concentration. Comparing the change in acidity between sunflower oil and coconut oil, this proves the hypotheses stated: the sunflower oil (unsaturated) has hydrolysed more than the saturated fat, coconut oil; hence, it is more unstable than the latter. Focusing on Diagram 1.1 (below), it can be seen that the acid concentration positively grows in relation to the amount of water exposed to the oil, as mentioned. More specifically, it has been explored that it grows exponentially. Using Microsoft Excel 2010, a trend-line (line of best-fit) has been introduced to fit the data set:

Diagram 1.1: Advanced
The equation of the trendline is:
FFA(m)=2.6882e0.0071m

Where FFA represents free fatty acid-concentration of the sunflower oil (to the factor 10-4 and units moles per 100 grams) and m the amount of water the oil is exposed to during frying (in grams). According to Microsoft Excel, the coefficient of determination (R2) is 0.907. The closer the R2-square value is to 1 in a trendline, the better it represents the actual data. In other words, the trendline in this particular case moderately fits the actual data, thus the model is good. Using this model, the acid concentration of the sunflower oil can be approximated at a certain water-value. However, it is important to understand that this model only applies to the particular sunflower oil that was used in this experiment. This is a major limitation. Nevertheless, more samples of this particular oil must be tested to further see if it fits this model and affirm its strength. Diagram 1.2: Acid-concentration of Olive oil (monounsaturated)

Analysing Diagram 1.2, it can be seen that the acidity-level of the olive-oil has insignificantly changed in either the positive or negative direction. As it was expected that the mono-saturated olive oil would clearly increase in its acid-concentration, reasons for this insignificant change can be discussed. Table 1.6: Fatty acid composition of various oils, expressed as percentage [4] In comparison to coconut-oil’s majority 8 to 12-carbon fatty acid composition [4], approximately 75% of the fatty acid composition of a regular olive oil is of 18-carbon fatty acid chains (Table 1.6) [4]. Although the mono-unsaturated olive oil used in this experiment mainly contains one double-bond (which stimulates instability), the greater number of carbon-atoms still has a greater significance in terms of the dispersion-forces holding the structure collected.

The fatty acid chains would be less volatile as a result of greater carbon-atoms contributing to great molecular mass and size. Thus, it may still remain stable under extreme conditions for a longer time of period. This explains why the acidity of the olive oil stayed constant throughout the high heat and water-exposure. Diagram 1.5: Comparison of the relative change in acidity between olive (monounsaturated) and coconut oil (saturated) By analysing Diagram 1.5, it can be seen that the acidity of the olive-oil relatively remained unchanged to the coconut oil. Hence, according to the results, olive oil did not hydrolyse as greatly expected, and in fact, maintained relatively the same acidity-level as coconut oil. Consequently, this refutes the hypothesis stated. Conclusion

In conclusion, it can be seen that saturated fats are more stable than poly-unsaturated fats during cooking. However, the general hypothesis “Unsaturated fats are more likely to hydrolyse during cooking than saturated fats of similar chain-length” remains inconclusive as this hypothesis has not been proved in coconut oil (saturated) vs. olive oil (monounsaturated).

Recommendations

It can be said that the olive oil may not have been exposed to enough water to be able to hydrolyse to a clear degree. Hydrolysing and degrading the strong bonded fatty acid composition of olive oil would require more than 200 grams of water and frying of 30 minutes (at 180 degrees Celsius). To verify this, an extended experiment should be performed to assist on proving, or refuting the hypothesis stated. This shall include exposing oils to greater amount of water at longer time-period to observe significant changes that may be noticed and discussed (hence, instead of exposing a certain oil to 40 wet cloths, this should be extended to 100+).

Furthermore, more types of particular oil (more types of saturated oil; i.e. palm oil or unsaturated oils) should be included in this experiment to further verify whether saturated fats are or are not more stable than unsaturated fats during exposure to water. All oils should be categorised to similar carbon chain-length to keep the variables the same (if possible); enhancing the quality of results. Moreover, the method in determining the end-point (by titration) of each sample should be modified as it was very difficult to determine these of oil and alcohol-samples. Instead, using the spectrophotometer would be a better option as this would provide greater inaccuracy. By following the recommendations stated, this hypothesis may be proved or refuted in a more conclusive manner.

Appendix 1 (Material-list):
For frying:

* Balance
* Metal saucepan
* 120 pieces of towel cloth (5 cm x 5 cm)
* Thermometer
* Digital Multimeter (connected to the thermometer, used to obtain the temperature-value) * Stand with holder and clamp
* A large scoop
* 2000 W hot plate
* Hot cover
* 500 ml of sunflower oil (56% poly-unsaturated)
* 500 ml of olive oil (83% mono-unsaturated)
* 500 ml of coco-nut oil (87% saturated)
* Tap-water
* 250 mL Beakers (3)
* 1000 mL Beaker (1)
* Test tubes (15)
* Test-tube-rack
* Plastic foil
* Tong

For titrating:
*
* Stand with holder and clamp
* Burette
* Shellite (at least 500 ml)
* Isopropanol (at least 500 ml)
* 0.11M NaOH (Sodium hydroxide)
* Samples of sunflower-oil (5)
* Samples of olive-oil (5)
* Samples of coco nut-oil (5)
* Balance
* Measuring cylinder (2)
* Phenolphthalein
* Funnel
* Erlenmeyer flask (5)

During the experimental procedures (obligation):
* Safety goggles
* Apron
Appendix 2 (Method)
Frying:

1. 120 pieces of 5 cm x 5 cm towel-cloth was prepared, 40 cloths to be used for each oil-type. 2. The 40 pieces of cloth was soaked in a 250 ml-beaker of (approximately) 180 grams of water for 2-3 minutes. 3. The hot plate was preheated to the temperature-range 170-190 degrees Celsius with 500 grams of sunflower oil. 4. When the temperature was reached, the oil was fried with initially 10 wet cloths.

5. A sample of approximately 20 mL was obtained after the initial 10 wet clothes ceased sizzling, and poured into a test-tube (covered with plastic-foil to minimise contamination from the external environment. 6. These oily cloths were removed and put into one of the 250 mL-beaker. 7. 10 wet-cloths were continuously added, using the tong. 8. Step 4-6 was repeated until the sunflower oil had been exposed to 40 wet cloths. 9. The amount of water diffused into the cloths was calculated by measuring the amount of water in the beaker before (180 grams) and after. 10. Step 2-9 was repeated with the other oils involved in this experiment. Blank titration:

11. The burette was set up
12. This was rinsed with 0.11 M of NaOH (the titrant).
13. The elevation of the burette was adjusted until the meniscus became eye-levelled. 14. The burette was filled with 0.11 M NaOH until it reached the meniscus. 15. 30 mL of Shellite and 2-propanol was mixed in a Erlenmeyer flask. 16. 10 drops of phenolphthalein was added into the Erlenmeyer flask. 17. The NaOH was added into the organic-solution. This resulted in a colour change (pink). The Erlenmeyer flask was swirled until the colour disappeared, thus indicating the solution had been neutralised. 18. The blank-titre was recorded.

Titration of oil-samples:

19. Step 10-11 was repeated 20. The initial sample of sunflower oil was weighed poured into the Erlenmeyer flask. 21. This was then dissolved in 30 mL of Shellite. The solution had to be vigorously shaken. 22. 30 mL of 2-propanol was added into the organic-solution. This had to be vigorously shaken, to extract the fatty acids into the polar phase. 23. 10 drops of phenolphthalein was added into the Erlenmeyer flask. 24. The NaOH was added into the organic-solution. This resulted in a colour change (pink). The Erlenmeyer flask was swirled until the colour disappeared, thus indicating the solution had been neutralised. 25. The titre was recorded and subtracted with the blank-titre obtained earlier. 26. Necessary calculations were performed in order to obtain the acid-concentration of the sample. 27. Step 19-26 was repeated with the remaining samples of the sunflower oil. 28. Step 19-27 was repeated with the other oils involved in this experiment. Note: A blank-titration had to be performed before each titration of all oils as these were performed on separate days/time-periods. This was necessary as the properties of the Shellite and 2-propanol may have changed over time.

Bibliography:
[1]:
Wilbraham, Anthony C., Dennis D. Staley, Michael S. Matta, and Edward L. Waterman. “Chapter 24.4: Lipids.” Chemistry. Boston: Prentice Hall, 2005. 775+. Print. [2]:
“CHECKING YOUR OIL: THE DEFINITIVE GUIDE TO COOKING WITH FAT.” Caveman Doctor. Ed. Sayra Lopez. Caveman Doctor, 27 May 2012. Web. 26 May 2013. <http://www.cavemandoctor.com/2012/05/27/checking-your-oil-the-definitive-guide-to-cooking-with-fat/>. [3]:

Watson, Alan. “”Saturated” Means “Stable” – Nothing Else.” Diet Heart News. Diet Heart News, 30 Jan. 2012. Web. 23 May 2013. <http://dietheartnews.com/2012/01/saturated-means-stable-nothing-else/>. [4]:

Raj, Gaurav, Ajit Singh Bhatnagar, Prasanth Kumar, Preeti Chandrashekar, and Gopala Krishna. Coconut Oil: Chemistry, Production and Its Applications – A Review. Rep. Department of Lipid Science & Traditional Foods, Central Food Technological Research Institute, n.d. Web. 27 May 2013. <http://coconutboard.nic.in/English-Article-Gopalakrishna-CFTRI.pdf>. [5]

Dobarganes, M. Carmen. “The AOCS Lipid Library.” Formation of Epoxy-, Keto- and Hydroxy-Fatty Acids in Frying Oils. Instituto De La Grasa (CSIC), n.d. Web. 01 June 2013. <http://lipidlibrary.aocs.org/frying/c-epoxy/index.htm>. [6]

Arnold, C. “Secondary Peroxidation Products From Fatty Acids or More Complex Lipids.”Cyber Lipid. N.p., n.d. Web. 01 June 2013. <http://www.cyberlipid.org/perox/oxid0009.htm>.