Quantification of Nonanal and Oleic Acid Formed during the Ozonolysis of Vegetable Oil Free Fatty Acids or Fatty Acid Methyl Esters

Article abstract of DOI:10.1007/s11746-015-2780-7

The ozonolysis of unsaturated lipids is a process that has been used to generate aldehydes, acids, alcohols, and other biobased chemical intermediates. Reported here is a method that can be used to measure the formation of nonanal and oleic acid during the ozonolysis of unsaturated vegetable oil fatty acids or their methyl esters to indicate the extent of the ozonolysis reaction. Derivatization was performed using boron trifluoride in methanol solution to transform nonanal and oleic acid into nonanal dimethyl acetal and oleic acid methyl ester, respectively. Undecanal and 10-heptadecenoic acid were used as internal standards and separation was performed using gas chromatography coupled with a flame ionization detector. The method was validated by performing a standard addition procedure in which nonanal or oleic acid standards were spiked into samples collected during the ozonolysis of oleic acid or canola oil fatty acid methyl ester (FAME). Linear regression results indicated that the measured nonanal and oleic acid are in good agreement with the actual amounts of nonanal and oleic acid added to the sample with at least 98 % recovery. The application of the method was demonstrated by the successful measurement of nonanal and oleic acid formed throughout the ozonolysis process for high oleic canola oil FAME.

Full text of DOI:10.1007/s11746-015-2780-7

J Am Oil Chem Soc (2016) 93:303–310
DOI 10.1007/s11746-015-2780-7
ORIGINAL PAPER
Quantification of Nonanal and Oleic Acid Formed During the
Ozonolysis of Vegetable Oil Free Fatty Acids or Fatty Acid Methyl
Esters
M. H. Tavassoli‑Kafrani1 · P. Foley2 · E. Kharraz1 · J. M. Curtis1
Received: 16 September 2015 / Revised: 20 November 2015 / Accepted: 12 December 2015 / Published online: 30 December 2015
© AOCS 2015
Abstract The ozonolysis of unsaturated lipids is a process
that has been used to generate aldehydes, acids, alcohols, and
other biobased chemical intermediates. Reported here is a
method that can be used to measure the formation of nonanal
and oleic acid during the ozonolysis of unsaturated vegetable
oil fatty acids or their methyl esters to indicate the extent of
the ozonolysis reaction. Derivatization was performed using
boron trifluoride in methanol solution to transform nonanal
and oleic acid into nonanal dimethyl acetal and oleic acid
methyl ester, respectively. Undecanal and 10-heptadecenoic
acid were used as internal standards and separation was
performed using gas chromatography coupled with a flame
ionization detector. The method was validated by perform-
ing a standard addition procedure in which nonanal or oleic
acid standards were spiked into samples collected during the
ozonolysis of oleic acid or canola oil fatty acid methyl ester
(FAME). Linear regression results indicated that the meas-
ured nonanal and oleic acid are in good agreement with the
actual amounts of nonanal and oleic acid added to the sample
with at least 98 % recovery. The application of the method
was demonstrated by the successful measurement of nonanal
and oleic acid formed throughout the ozonolysis process for
high oleic canola oil FAME.
Introduction
Environmental concerns regarding the use of petrochemicals
as feedstocks for polymer production and the rapid depletion
of this resource have led researchers to look for alternatives
[1, 2]. Vegetable oils are one such potential alternative to pet-
rochemical feedstocks for polymer production [1]. The pres-
ence of a significant degree of unsaturation in vegetable oil
allows for various chemical modifications such as hydrogen-
ation, epoxidation, and ozonolysis to be performed. Ozone
has been used to cleave fatty acids at double bond positions
resulting in the formation of different compounds, the major
ones being carboxylic acids, aldehydes, ketones, and alco-
hols [3]. The mechanism of ozonolysis of vegetable oils has
been studied in detail [4]. Briefly, ozone reacts with double
bonds and forms an unstable intermediate compound named
1,2,3-trioxolane (molozonide) which decomposes into an
aldehyde and carbonyl oxide (Scheme 1). These products can
engage in further reactions and yield other products such as
oligomeric peroxides [5] and secondary aldehydes [6]. Ozo-
nolysis has been used to produce polyols from unsaturated
vegetable oil triacylglycerols by oxidizing double bonds into
aldehyde groups and then reducing them to hydroxyl groups
[7–9]. In addition, aldehydes produced in ozonolysis can be
used in the fragrance industries [10, 11].
Keywords Ozonolysis · Nonanal · Vegetable oils · Gas
chromatography · Dimethyl acetal derivative
Oleic acid is a major fatty acid component of many veg-
etable oils which can be ozonized to produce nonanal as a
result of the presence of a double bond at the Δ9 position.
Hence, a method for the measurement of oleic acid and
nonanal concentration during the ozonolysis of oleic acid-
containing oils would provide an indication of the extent of
the ozonolysis reaction. Gas chromatography (GC) provides
a high resolution and potentially rapid separation for many
lipid derivatives. However, the direct analysis of ozonides
by high temperature GC may result in the decomposition of
* J. M. Curtis
jcurtis1@ualberta.ca
1
Lipid Chemistry Group, Department of Agricultural, Food
and Nutritional Science, University of Alberta, Edmonton,
AB T6G 2P5, Canada
2
P2 Science Inc., 4 Research Drive, Woodbridge, CT 06525,
USA
1 3

304
J Am Oil Chem Soc (2016) 93:303–310
Scheme 1 Mechanism of
ozonolysis for unsaturated fatty
acids. Adapted from [4–6, 13]
ozonides into aldehydes and/or acids and therefore lead to
inaccurate aldehyde values. Additional challenges in quanti-
tative ozonolysis reaction monitoring by GC include (a) fatty
acids can be difficult to quantify under a number of ozonoly-
sis conditions given that they may be ionized and therefore
be under-represented without acid treatment; (b) certain high
molecular weight esters (e.g., diglycerides) and oligom-
ers (e.g., ozonide oligomers) may be undetectable without
derivatization; and, (c) performing direct GC analysis of
aldehydes from ozonolysis without derivatization results in
poor reproducibility, e.g., standard deviations of up to 19 %
have been reported for nonanal measurement by direct GC
[11, 12]. It should be noted that this last factor may not be
entirely due to GC analysis itself but also to the difficulties
associated with sampling from unstable emulsions.
In this work we describe the conversion of aldehyde
and carboxyl groups formed in the ozonolysis of unsatu-
rated lipids into their dimethyl acetal and ester derivatives,
respectively. These derivatives are more stable in the high
temperatures used in GC analysis. A specific target for this
work is the development of an accurate and reliable method
for the measurement of nonanal and residual oleic acid
from the ozonolysis of unsaturated fatty acids and fatty
acid methyl esters (FAME).
Aldrich, respectively. Sodium bicarbonate and sodium sul-
fate were obtained from Sigma Aldrich. Oleic acid (90 %)
used for the ozonolysis was of technical grade and obtained
from Sigma Aldrich. All chemicals and solvents were used
without additional purification.
Ozonolysis and Sampling
A mixture (600 g) of canola oil FAME or oleic acid (90 %) in
water was prepared at the ratio of 1:1 (w/w) and transferred
into a 2-L stainless steel ozonolysis reactor equipped with
temperature controlling unit (Refrigerated/Heating Circula-
tor, Julabo F25) and a speed controlled rotor unit. Ozone at a
concentration of 50 g/m3 (ozone/oxygen mixture) was gener-
ated by passing dry oxygen (99.6 %, Praxair Canada Inc.)
through an ATLAS100 ozone generator (model ATLAS100,
Absolute Ozone, Canada). Ozone was introduced into the
reactor as finely dispersed bubbles using a purpose-built coil
with holes located at the bottom of the reactor. The ozonol-
ysis of canola oil FAME or oleic acid (90 %) was contin-
ued for 300 min using an ozone flow rate of 6.5 L/min at
an oxygen pressure of 7 psi. The temperature and agitation
speed were kept at 0 °C and 800 5 rpm during ozonolysis,
respectively. Every 10 min, a 2-mL sample of the ozonolysis
mixture was transferred into a 10-mL glass vial with screw
cap. Then 5 mL salty water (3 %) was added to the sample
and was shaken for 1 min in a vortex mixer. The sample was
then centrifuged (2500 g) for 10 min. Around 0.5 mL of the
top oily layer was transferred to a 2-mL vial with cap and
stored at −18 °C before analysis with GC–FID.
Experimental Procedures
Materials
Pure nonanal (97 %), undecanal (97 %), oleic acid (99 %),
10-heptadecenoic acid (99 %), boron trifluoride solution
(12 % w/w in methanol), and NaCl (99.5 %) were pur-
chased from Sigma Aldrich. Heptane and methanol were
of HPLC grade and purchased from Caledon and Sigma
Derivatization
Samples were removed from the freezer and allowed to
melt at room temperature. Then about 40 mg of sample and
1 3

J Am Oil Chem Soc (2016) 93:303–310
305
20 mg of undecanal and 10-heptadecenoic acid as inter-
nal standards were accurately weighed into a 40-mL tube
with a screw cap. Boron trifluoride solution (5 mL, 12 %
in methanol) and methanol (5 mL) were added into the
tube and the solution was mixed for 1 min. The tube was
then transferred into a 75 °C water bath for 5 min. After
this period, the tube was removed from the water bath and
cooled to room temperature. Heptane (5 mL) was added
and the tube was returned to the 75 °C water bath for a fur-
ther minute. The solution was allowed to cool to ambient
temperature and 5 mL of salty water (3 % sodium chloride
in water) was then added. The tube was shaken for 1 min
on a vortex mixer after which the upper layer (organic) was
transferred into another tube and washed with salty water a
further two times. The organic layer was dried with sodium
sulfate and then centrifuged for 2 min to remove sodium
sulfate. The organic phase was stored in a 10-mL vial with
a screw cap and was diluted with heptane to give a final
concentration of close to 0.3 mg/mL of each internal stand-
ard for subsequent GC analysis.
in methanol as the mobile phase at a flow rate of 200 µL/
min for 2 min. The temperature of the ion source was set
at 400 °C and the mass spectra were obtained over range of
m/z 50–1000. The ion spray voltage, focus potential (FP),
declustering potential (DP), and DP2 potential were 5000,
200, 50, and 10 V, respectively. Nitrogen was used for aux-
iliary gas at 40, nebulizing gas at 50, and curtain gas at 30,
all in arbitrary units.
Calibrations
A series of calibration solutions of nonanal and oleic acid
ranging from 0.1 to 1.0 mg/mL with 0.1 mg/mL inter-
vals and containing 0.3 mg/mL of both internal standards
(undecanal and 10-heptadecenoic acid) were prepared. In
brief, amounts between 5 and 35 mg each of nonanal and
oleic acid at 5-mg intervals were weighed into a 40-mL
tube. Internal standards were added and the mixtures were
derivatized and finally diluted in heptane. The solutions
were injected into the GC–FID and the peak areas of deri-
vatized nonanal, oleic acid, undecanal, and 10-heptadece-
noic acid were measured. The calibration curve for nonanal
was developed by plotting the area ratios of nonanal/unde-
canal against the concentration ratios of nonanal/undecanal
for each standard solution. The calibration curve for oleic
acid was developed in a similar way to the nonanal but
using 10-heptadecenoic acid as internal standard. Linear
regressions were performed in each plot and the equations
obtained were used to quantify nonanal and oleic acid.
GC Analysis
The GC instrument used in this study was an Agilent Tech-
nologies GC (model 6890N, USA) coupled with flame ion-
ization detector (FID). Separation was performed on an HP
5 capillary column (30 m × 0.32 mm ID, film thickness
0.25 μm, SGE Analytical Science Pty Ltd, USA) using
helium gas as the mobile phase at a flow rate of 1.8 mL/
min. The oven temperature was programmed to start at
100 °C, then rise to 300 °C at a rate of 20 °C/min giving
a total run time of 10 min. The injection was carried out
at a split ratio of 20:1 with injection volume of 1 μL. The
temperatures of the injector and detector were both kept at
275 °C.
Validation
In order to validate the method, two procedures were per-
formed. In the first procedure, oleic acid and high oleic can-
ola oil FAME were ozonized and samples were collected
over the course of the reaction. The amount of nonanal and
oleic acid was measured by GC–FID after derivatization.
In the second procedure, either oleic acid or nonanal
was added to the ozonolysis products extracted after 60
and 300 min of the ozonolysis of free fatty acids or FAME,
as described in the “Experimental” section. Briefly, 40
samples were collected from the ozonolysis mixtures (10
samples for fatty acid and 10 samples for FAME mix-
tures for each of the two time points). To each of the 10
samples collected after 60 min of ozonolysis, a series of
additions of pure nonanal (98 %) were made. These were
chosen to result in an increase in the final nonanal concen-
tration of 0–225 mmol/100 g reaction mixture, at incre-
ments of around 30 mmol/100 g. To each of the 10 sam-
ples collected after 300 min, 0–80 mmol of oleic acid
was added per 100 g reaction mixture, at increments of
around 10 mmol/100 g. Both the oleic acid and nonanal
spiked samples were shaken vigorously for 2 min and the
GC/MS Analysis and Mass Spectrometry
The GC/MS system used in this work was an Agilent GC/
MS with electron ionization (EI) with triple-Axis Detector
(Agilent technologies 5975C inert XL EI/CI MSD, Santa
Clara CA, USA). The scan range was m/z 50–1000 at a
scan speed of 1.55 scans/s with an electron multiplier volt-
age of 1282 V. The separation was performed on the same
column and using the same conditions as used for GC/FID
analysis.
For mass spectrometry, a QSTAR Elite (Applied Biosys-
tems/MDS Sciex, Concord, ON, Canada) mass spectrom-
eter in positive electrospray ionization (ESI) with time of
flight (TOF) mass analyzer was used. About 2 µL of deri-
vatized product (in a heptane solution at concentration of
ca. 0.1 mg/mL) was introduced into the mass spectrom-
eter by flow injection using 10 mmol ammonium formate
1 3

306
J Am Oil Chem Soc (2016) 93:303–310
Fig. 1 GC chromatogram of
derivatized nonanal, undecanal,
10-heptadecenoic acid, and
oleic acid in a heptane solution
(dark line) and the ozonolysis
sample taken after 120 min
ozonolysis of high oleic canola
oil fatty acid methyl esters (gray
line). The chromatogram of the
ozonolysis sample has been
shifted to the right and up for
clarity
concentrations of nonanal and oleic acid were then ana-
lyzed by GC/FID. The total nonanal or oleic acid amount
experimentally determined at each time point was plotted
against the added nonanal or oleic acid amount and linear
regression was performed for each plot.
derivatization products from nonanal, undecanal, 10-hep-
tadecenoic acid, and oleic acid were well separated with
retention times of 3.06, 4.22, 7.08, and 7.53 min, respec-
tively. For comparison, the GC chromatogram of the FAME
of high oleic canola oil ozonolysis mixture is presented in
Fig. 1 (gray line, offset to the right) and it can be seen that
this separation is suitable for quantitative measurement of
these analytes.
Extraction of Ozonides
In order to test the hypothesis that free aldehydes and acids
are not formed from ozonide compounds as a result of the
derivatization step, ozonides were extracted from a reac-
tion mixture formed by the ozonolysis of oleic acid in hep-
tane. Briefly, around 2 g oleic acid was dissolved in 20 mL
heptane and was transferred into a 50-mL glass ozonolysis
reactor with magnet stirrer. Ozonolysis was carried out for
90 min by passing ozone at concentration of 50 g/m3 with
a flow rate of 1 L/min and temperature maintained at 0 °C.
After completion of ozonolysis, the mixture was trans-
ferred into a separating funnel and the lower layer contain-
ing the ozonides was collected. The collected viscous layer
was then washed with heptane twice and finally dried under
a nitrogen stream and stored at −18 °C prior to analysis by
1H NMR spectroscopy.
Analysis by Mass Spectroscopy
GC/MS with EI was performed (data not shown) to iden-
tify the products of BF3/methanol derivatization of nona-
nal, oleic acid, 10-heptadecenoic acid, and undecanal. A
National Institute of Standards and Technology (NIST)
library search showed that the possible derivatization
products are nonanal dimethyl acetal, undecanal dimethyl
acetal, and methyl esters of 10-heptadecenoic acid and oleic
acids. Since the fragmentation in EI is high, the molecu-
lar ions of the products were not found in the EI spectrum
but in each case an ion representing [M − CH3O·]+ was
observed. In order to confirm the products of derivatization,
nonanal, undecanal, 10-heptadecenoic acid, and oleic acid
after derivatization were analyzed by flow injection posi-
tive ion ESI mass spectrometry. Figure 2 shows the ESI(+)
mass spectrum of the undecanal and oleic acid after deri-
vatization with BF3/methanol. The ions at m/z 239.200 and
m/z 319.264 correspond to the sodium ion adducts of unde-
canal dimethyl acetal ([C13H28O2 + Na]+) and oleic acid
methyl ester ([C19H36O2 + Na]+), respectively whilst the
ion at m/z 314.307 is the ammonium ion adduct of oleic
acid methyl ester ([C19H36O2 + NH4]+). The ion at m/z
185.193 ([C12H25O]+) is consistent with the loss of sodium
methoxide from the sodium adduct ion of undecanal dime-
thyl acetal (m/z 239.2). The results obtained for derivat-
ized nonanal and 10-heptadecenoic acid were analogous to
results obtained for derivatized undecanal and oleic acid.
According to the results obtained by GC/MS and ESI
mass spectrometry, nonanal dimethyl acetal and undeca-
nal dimethyl acetal were identified as the products of the
1H NMR Spectroscopy
A solution of the ozonide sample in deuterated (D6) ace-
1
tone at a concentration of 15 mg/mL was prepared for H
NMR analysis using a 400-MHz Varian Inova 400-MR
NMR. The spectrum was obtained at room temperature.
Results and Discussion
GC Separation
Figure 1 (dark line) shows the GC–FID chromatogram of
standards of nonanal, undecanal, 10-heptadecenoic acid,
and oleic acid after derivatization with BF3 and dilution
with heptane to a concentration of 0.3–0.4 mg/mL. The
1 3

J Am Oil Chem Soc (2016) 93:303–310
307
Fig. 2 Flow injection ESI(+) mass spectrum of products of derivatization of a undecanal and b oleic acid with BF3/methanol
derivatization of nonanal and undecanal by BF3/methanol.
Similarly, 10-heptadecenoic methyl ester and oleic acid
methyl ester were identified as the derivatization products
from 10-heptadecenoic acid and oleic acid. Hence, the deri-
vatization reaction can be summarized as given Scheme 2.
1H NMR Spectroscopy
In order to quantify the free aldehydes and acids that are
formed as a result of ozonolysis, it is necessary to confirm
that during the derivatization step (Scheme 2), ozonide
compounds that are present are not reduced to aldehydes
and acids, since this would result in an overestimation in
the measurement of free aldehydes and acids. In order to
Scheme 2 Derivatization of aldehydes and acids with methanol
prove this, ozonolysis of oleic acid in heptane was carried
out for 90 min and then ozonide compounds, the major-
ity of which become insoluble in heptane, were separated
from the ozonolysis mixture. The extracted material was
1
first analyzed by H NMR spectroscopy. As it can be seen
in Fig. 3, resonances at 5.11 and 5.17 ppm were clearly
1
seen in the H NMR spectrum, confirming the presence of
ozonides [14, 15]. Then, a portion of this ozonide sample
was derivatized using the above procedures and analyzed
by GC–FID. The duplicate GC–FID results of the dimethyl
acetate derivatives indicate that the nonanal concentration
is less than 2 % by weight in the ozonide. Hence, there is
not a significant conversion of ozonide to aldehyde during
derivatization and it can be assumed that nonanal which is
formed during ozonolysis under some reaction conditions
(such as in the presence of water) can be quantified via
Fig. 3 1H NMR spectrum of extracted materials from ozonoly-
sis mixture of oleic acid. Resonances at 5.1 and 5.17 ppm refer to
ozonide
dimethyl acetate derivatives.
Calibration
standard for oleic acid. These were chosen since fatty acids
present in vegetable oils commonly contain an even number
of carbons whereas fatty acids containing an odd number of
In order to develop calibration curves, undecanal
(C10H21CHO) was used as internal standard for nonanal
while 10-heptadecenoic acid (C17H32O2) was the internal
1 3

308
J Am Oil Chem Soc (2016) 93:303–310
Fig. 4 Plots of GC–FID peak
area ratios of a nonanal/undeca-
nal versus concentration ratios
of nonanal/undecanal, b oleic
acid/10-heptadecenoic acid ver-
sus concentration ratios of oleic
acid/10-heptadecenoic acid in
standard solutions
carbons are rare. In order to test the formation of undeca-
nal and 10-heptadecenoic acid during the ozonolysis reac-
tion of fatty acids and FAME, a sample was taken from the
ozonolysis mixture of the high oleic acid canola oil FAME
after 120 min ozonolysis, then derivatized with BF3/metha-
nol and analyzed by GC–FID. As it can be seen in Fig. 1,
there is no peak related to the undecanal or 10-heptadece-
noic acid which indicates that the formation of undecanal
and 10-heptadecenoic acid during the ozonolysis reaction
is not detectable and hence they are suitable as internal
standards.
ozonolysis reaction mixture, standard additions of nonanal
and oleic acid were performed.
Oleic acid (ca. 90 %) was ozonized according to the pro-
cedure described earlier and the standard addition of nona-
nal was carried out for a sample collected at the early stage
of ozonolysis (30 min) where the concentration of nona-
nal from the ozonolysis reaction is relatively low. Also,
the standard additions of oleic acid were performed on the
sample collected towards the end of the ozonolysis process
when most of the oleic acid had been consumed. These
samples were then derivatized and analyzed by GC–FID.
The measured nonanal or oleic acid amounts are plotted
against the actual nonanal or oleic acid added to the sample
in Fig. 5. A similar standard addition experiment was also
performed using high oleic canola oil in the FAME form,
with similar additions of nonanal (after 30 min ozonolysis)
and oleic acid (after 90 min ozonolysis).
Figure 5a shows the measured concentration of nona-
nal and oleic acid by GC–FID against actual concentra-
tion of nonanal added to the ozonolysis of FAME of high
oleic acid canola oil collected after 30 min of ozonolysis.
As expected, the measured nonanal concentration propor-
tionally increased as nonanal was added to the sample. At
the same time, the measured oleic acid decreased because
the oleic acid was diluted as nonanal was added to the sam-
ples. Figure 5b shows the changes in measured oleic acid
and nonanal versus added oleic to the sample collected
after 90 min of ozonolysis of fatty acid. The measured oleic
acid increased linearly by adding oleic acid while meas-
ured nonanal decreased. Similar trends were observed for
the oleic acid (ca. 90 %) samples collected after 30 and
90 min of ozonolysis and spiked with nonanal and oleic
acid respectively (Table 1).
Figure 1 (dark line) shows the chromatogram of a hep-
tane solution of nonanal, undecanal, 10-heptadecenoic acid,
and oleic acid after derivatization with BF3/methanol. The
peak area of non-derivatized nonanal, undecanal, 10-hep-
tadecenoic acid, and oleic acid was not significant, indi-
cating that the derivatization reaction is complete within
5 min. Calibration curves of the dimethyl acetal derivative
formed from a pure nonanal standard, and the methyl ester
derivative formed from an oleic acid standard, are shown in
Fig. 4.
These showed high linearity (R2 > 0.998) and were suit-
able for the determination of nonanal and oleic acid.
Validation
Standard Additions
During the ozonolysis of vegetable oils, ozone reacts with
double bonds present in the oil and produces unstable
1,2,3-trioxolanes, which then decompose to other interme-
diates such as zwitterions (carbonyl oxide dipole) and car-
bonyl fragments. These intermediates participate in further
reactions as ozonolysis proceeds to form more stable ozon-
ides, oligomeric peroxides, and aldehydes [10]. As a result
of the formation of high molecular weight polar products,
the viscosity of the ozonolysis mixture greatly increases
during the reaction. In order to evaluate recoveries for
the extraction of nonanal and oleic acid directly from the
A summary of the standard addition results for both free
fatty acid and FAME samples are given in Table 1. In both
cases, the measured nonanal and oleic acid concentrations
following derivatization show good linearity with oleic acid
and nonanal additions giving R2 values of between 0.982
and 0.996. The slopes obtained by linear regression of plots
1 3

J Am Oil Chem Soc (2016) 93:303–310
309
Fig. 5 Measured nonanal and oleic acid content by GC–FID after derivatization against a added nonanal into high oleic acid canola FAME after
30 min of ozonolysis and b added oleic acid into high oleic acid canola FAME after 90 min ozonolysis
Table 1 Linear regression
Oil type Time Nonanal changes
Slope
Intercept R2
Oleic acid changes
CV Slope
Intercept R2
results of standard addition
of nonanal and oleic acid into
the ozonolysis sample taken at
different times of the ozonolysis
of FAME of high oleic canola
oil and oleic acid fatty acid (ca.
90 %)
SD
SD
CV
FAME
FA
30a
90b
30a
90b
0.985 37.6
0.9989 3.02 2.10 −0.291 203.0
0.9902 0.85 0.99 1.006 118.0
0.9943 5.40 3.73 −0.331 231.6
0.9896 0.74 0.86 0.988 133.6
0.9957 1.74 1.02
0.9869 3.02 2.03
0.9935 2.46 1.25
0.9821 4.14 2.44
−0.285 94.6
0.985 40.5
−0.268 95.8
SD root-mean square error (RMSE), CV coefficient of variation (%)
a
Spiked with nonanal
b
Spiked with oleic acid
of measured nonanal versus added nonanal were similar
for free fatty acid and FAME samples and indicate that the
recovery of nonanal is 98.5 % in both cases. Similarly, the
recovery of oleic acid added to the ozonolysis mixtures
of FAME and fatty acid collected after 90 min were 98.8
and 101 %. These recoveries are close to 100 % recovery
and therefore suitable for the measurement of nonanal and
residual oleic acid in ozonolysis reactions. This data sug-
gests that high recovery of both nonanal and oleic acid is
independent of the starting fatty acid form.
inhomogeneous mixture of ozonolysis products. Taking
representative samples from this mixture without perform-
ing separation can be problematic since about 50 % of
the mixture is water and the polarity of the organic phase
changes during the ozonolysis process. In this method, the
sample used for the analysis was taken from the organic
phase of the ozonolysis mixture. This can reduce the varia-
tions due to taking sample from an emulsion system.
Although this method was tested for the ozonolysis of
fatty acids and methyl esters in the presence of water, the
application of the method is not limited to these conditions.
The method has the potential to be used for the ozonoly-
sis of TAGs in the presence of water or organic solvents.
In this case, in order to have a representative sample, the
mixture of the ozonolysis can be dissolved in an appropri-
ate solvent to make a homogeneous solution and the results
can be reported on the basis of the total ozonolysis mixture.
In addition to the measurement of nonanal and oleic acid,
the method has the potential to be tested for the measure-
ment of other aldehydes such as hexanal or other carbox-
ylic acids such as nonanoic acid and hexanoic acid during
the ozonolysis of vegetable oils and their derivatives.
Quantification of Nonanal and Oleic Acids During the
Ozonolysis Process
The method was tested to measure nonanal and oleic acid
content during the ozonolysis of oleic acid fatty acids (ca.
90 %) and FAME from a high oleic acid canola oil. The
results are presented in Fig. 6 which shows the rate of
decrease in oleic acid during ozonolysis along with the rate
of increase in nonanal concentration.
The ozonolysis of fatty acids and their methyl esters
in the presence of water results in the formation of an
1 3

310
J Am Oil Chem Soc (2016) 93:303–310
Fig. 6 Measured nonanal and
oleic acid contents during the
ozonolysis of a fatty acid (oleic
acid ca. 90 %) and b high oleic
acid canola FAME
3. McGuire J, Haslam P, Bond G (2006) Ozonolysis in the produc-
tion of chiral fine chemicals. In: Ager D (ed) Handbook of chiral
chemicals. CRC, Taylor & Francis, Boca Raton, pp 164–185
4. Criegee R (1975) Mechanism of ozonolysis. Angew Chem Int
Ed Engl 14:745–752
5. Bailey PS (1958) The reactions of ozone with organic com-
pounds. Chem Rev 58:925–1010
6. Fliszar S, Carles J (1969) A quantitative investigation of ozo-
nolysis reaction. VII. Ozonolyses of phenylethylenes in pres-
ence of oxygen-18-labeled benzaldehyde. J Am Chem Soc
91:2637–2643
7. Petrovic ZS, Zhang W, Javni I (2005) Structure and properties of
polyurethanes prepared from triglyceride polyols by ozonolysis.
Biomacromolecules 6:713–719
8. Narine SS, Kong X, Bouzidi L, Sporns P (2007) Physical proper-
ties of polyurethanes produced from polyols from seed oils: I.
Elastomers. J Am Oil Chem Soc 84:55–63
9. Petrović ZS, Milić J, Xu Y, Cvetković I (2010) A chemical route
to high molecular weight vegetable oil-based polyhydroxyal-
kanoate. Macromolecules 43:4120–4125
10. Sell CS (2014) Manufacture of fragrance ingredients. In: Chem-
istry and the sense of smell. Wiley, London, pp 296–356
11. Omonov TS, Kharraz E, Curtis JM (2011) Ozonolysis of canola
oil: a study of product yields and ozonolysis kinetics in different
solvent systems. J Am Oil Chem Soc 88:689–705
Conclusion
The progress of ozonolysis reactions for vegetable oils and
their derivatives can be followed by the measurement of oleic
acid, which is one of the major fatty acids in many vegeta-
ble oils, and nonanal, which is one of the major aldehydes
formed by ozonolysis. In this work, a method was developed
for the measurement of these compounds using GC–FID. The
results obtained by performing standard additions of nonanal
and oleic acid into actual ozonolysis reaction mixtures col-
lected from the ozonolysis of oleic acid and the methyl esters
of high oleic canola oil indicate that the method can accu-
rately predict the amounts of added nonanal and oleic acid.
In addition, the use of the method was demonstrated by
the measurement of nonanal and oleic acid during the ozo-
nolysis of oleic acid and the methyl esters of high oleic acid
canola oil. It should be noted that the total required time
for analysis using the method described above, including
sampling, derivatization, and GC analysis, is under about
30 min. This is adequate in the context of a reaction running
for many hours (see Fig. 6) but for reaction monitoring and
for determination of the reaction end-point with less than
30 min resolution, a faster method would be desirable.
12. Dumont JM, Kharraz E, Qi H (2013) Production of polyols and
mono-ols from 10 North American vegetable oils by ozonoly-
sis and hydrogenation: a characterization study. Ind Crop Prod
49:830–836
13. Cvetkovic I, Milic J, Ionescu M, Petrovic ZS (2008) Preparation
of 9-hydroxynonanoic acid methyl ester by ozonolysis of vegeta-
ble oils and its polycondensation. Hem Ind 62:319–328
14. Soriano NU Jr, Migo VP, Matsumura M (2003) Ozonation of
sunflower oil: spectroscopic monitoring of the degree of unsatu-
ration. J Am Oil Chem Soc 80:997–1001
Acknowledgments We acknowledge the University of Alberta
Biorefining Conversion Network and Alberta Innovates Biosolutions
for support for this research. Mr. Tavassoli-Kafrani thanks Alberta
Innovates-Technology Futures for providing a scholarship.
15. Sadowska J, Johansson B, Johannessen E, Friman R, Broniarz-
Press L, Rosenholm JB (2008) Characterization of ozonated
vegetable oils by spectroscopic and chromatographic methods.
Chem Phys Lipids 151:85–91
References
1. Gandini A (2008) Polymers from renewable resources: a chal-
lenge for the future of macromolecular materials. Macromol-
ecules 41:9491–9504
2. Mohanty A, Misra M, Drzal L (2002) Sustainable bio-compos-
ites from renewable resources: opportunities and challenges in
the green materials world. J Polym Environ 10:19–26
1 3