Fuel 139 (2015) 248256

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

A preliminary study for the photolysis behavior of biodiesel and its

blends with petroleum oil in simulated freshwater
Zeyu Yang a,b,, Bruce P. Hollebone a, Zhendi Wang a, Chun Yang a, Carl Brown a, Gong Zhang a,
Mike Landriault a, Xinchao Ruan c,
a

Emergencies Science and Technology Section (ESTS), Science and Technology Branch, Environment Canada, Ottawa, ON, Canada
Key Laboratory of Catalysis and Materials Science of the State Ethnic Affairs Commission & Ministry of Education, Key Laboratory of Analytical Chemistry of the State Ethnic
Affairs Commission, College of Chemistry and Materials Science, South-Central University for Nationalities, Wuhan 430074, PR China
c
The Research Center of Environmental Science, Wuhan Textile University, Fangzhi Road, Hongshan District, Wuhan 430073, PR China
b

h i g h l i g h t s
 For FAMEs, higher degree of saturation resulted in lower transformation.
 Water matrices, initial concentration of biodiesel affected degradation slightly.
 Biodiesel source effect can be neglected.
 Presence of biodiesel stabilized the small oil droplets in aqueous phase.
 Biodiesel inhibited the degradation of some heavy hydrocarbons.

a r t i c l e

i n f o

Article history:
Received 13 June 2014
Received in revised form 11 August 2014
Accepted 26 August 2014
Available online 6 September 2014
Keywords:
Photolysis
Biodiesel
Diesel
Blend
Freshwater

a b s t r a c t
With the increasing use of biodiesel and its blends with petroleum fuel, the corresponding environmental
issues also occur during its production, application and transportation. The photolysis behavior for
biodiesel and the impacts of biodiesel on the photo-oxidation of petroleum hydrocarbons in simulated
freshwater was studied by irradiated with ultra violet (UV) and simulated sunlight in the present study.
The results indicated that the photolysis rates of fatty acid methyl esters (FAMEs) were mainly depended
on their degree of saturation, slightly on water matrices and the initial concentration of biodiesel,
regardless of biodiesel sources. Similar results were observed for total organic carbon (TOC) removal
rates; however, TOC removal rates were slightly dependent on the initial concentration of biodiesel.
The presence of humic acid and pyrogallic acid or lake water matrices slightly inhibited the removal rates
of TOC. The photolysis rates of individual petroleum hydrocarbons with and without the presence of
biodiesel followed similar rules. In brief, alkanes with light molecular weights were transformed faster
than those with heavy molecules, the removal of polycyclic aromatic hydrocarbons (PAHs) were more
signicantly than alkanes, and the removal of alkylated PAHs (APAHs) increased concurrently with the
alkylation level in each family. The presence of biodiesel only inhibited the photolysis of some heavy
alkanes and PAHs, not for all other petroleum hydrocarbons. Biodiesel, as a surfactant-like material, could
stabilize small oil droplets initially formed by agitation, therefore, these droplets experience longer
lifetimes in the water phase before re-aggregating into larger globules and rising to the surface. The
apparent solubility of petroleum hydrocarbons, especially for those with heavier molecular weights,
has been enhanced in the presence of FAMEs. In this scenario, light needs to penetrate water phase to
degrade these targets compared with diesel alone. The direct contacting opportunities between UV light
and targets, and radicals produced to attack targets were reduced, which nally resulted in the inhibited
photolysis rates of some heavy molecular weight hydrocarbons.
2014 Elsevier Ltd. All rights reserved.

Corresponding authors. Address: Emergencies Science and Technology Section

(ESTS), Science and Technology Branch, Environment Canada, 335 River Road,
Ottawa, ON K1A 0H3, Canada. Tel.: +1 (613)990 3219; fax: +1 (613)991 9485
(Z. Yang), The Research Center of Environmental Science, Wuhan Textile University,
Fangzhi Road, Hongshan District, Wuhan 430073, PR China (X. Ruan).
E-mail addresses: zeyu.yang@ec.gc.ca (Z. Yang), ruanxc72@163.com (X. Ruan).
http://dx.doi.org/10.1016/j.fuel.2014.08.061
0016-2361/ 2014 Elsevier Ltd. All rights reserved.

1. Introduction
Biodiesel consists of a mixture of long-chain fatty acid alkyl
esters processed from biological triglycerides. It has received
intensive attention as one of the most signicant supplement fuels

Z. Yang et al. / Fuel 139 (2015) 248256

and alternatives to petroleum diesel fuel, because of its comparable engine performance and environmental characteristics.
Spill accidents happen more often than ever with the increasing
use of biodiesel and its blends with petroleum fuel. Biodiesel spills
tend to spread and form a sheer thick layer of clear to light milky
white color on top of sea water, and fatty acid methyl esters
(FAMEs) can be easily broken down to generate free fatty acids
and methanol after spillage [1,2]. Corresponding adverse effects
will occur to ecosystems and human health. Therefore, it is necessary to understand the fate of biodiesel exposed to the environment, and the dominant degradation mechanisms, for developing
suitable remedial action and identifying the source in the case of
a fuel spill.
Once biodiesel/its blend with petroleum oil spills into the environment, various weathering processes, for example, biodegradation, photo-oxidation, dissolution and evaporation, will occur.
Because biodiesel components are more readily biodegraded than
fossil fuels, many studies have focused on microbial degradation
for remediation purposes [38]. Among them, a remediation study
has shown that fatty acid methyl esters in biodiesel are degraded
much faster than most components in fossil diesel [4]. This corresponds very well with the fact that biodiesel is more readily biodegradable compared with fossil diesel [3].
The contribution of biodiesel to the physical property of diesel
has been investigated by several researchers [4,9]. It was found
that the evaporation rate of fossil diesel components was not
affected by the presence of FAMEs in simulated seawater [4]. However, the physical mobility of heavy oil mixtures was enhanced by
FAMEs amendment in sand column [5]. Recently, one of our
research has reported that the evaporation process and stability
of diesel would not alter with the presence of biodiesel in a simulated ambient condition [10].
The contribution of biodiesel to the biodegradation of diesel
suggested that the presence of biodiesel could facilitate the biodegradation of some petroleum hydrocarbons due to microbial
growth promotion [5,9] and increase bioavailability (emulsion)
[11]. Other researchers, however, have found that the presence
of biodiesel did not accelerate the biodegradation of hydrocarbons [4,12,13]. For example, the degradation of benzene and toluene in anoxic and hypoxic conditions was hindered by the
presence of biodiesel [12]. It seemed that the relatively high viscosity of biodiesel limited the migration potential of target hydrocarbons, resulting in their relatively slow natural attenuation
process.
Among the various weathering processes, photolysis behavior
of fuel is one of the important factors to control their transformation and fate in the environment. It is well known that the photolysis behavior of organic compounds in aquatic phase is based on
the ability of solar/UV radiation to attack the target compounds.
As a consequence, photochemical processes may take place, in
which different transient species are generated: e.g., photo-ionization, radicals generated by bond homolysis or bond heterolysis, as
well as a number of photo-physical processes (uorescence, phosphorescence, etc.) [14]. The photo-oxidation behavior of petroleum
in water has attracted much attention in the past years [1518].
Generally, the photolysis rates of petroleum hydrocarbons depend
on the composition and physical properties of the exposed parent
oil, wavelength, turbidity levels of sample, suspended particulate
matter concentration, and water matrices [17,19,20]. Till yet, different mechanisms for the photo-oxidation of petroleum and
FAMEs have been described, including free-radical oxidation in
the presence of oxygen, singlet oxygen initiation of hydro peroxide
formation, and ground state triplet oxygen combining with free
radicals to form peroxides [17,18,21]. Recently, the photochemical
process of crude oil has also become better characterized due to
the development of analytical technologies [16,22].

249

Biodiesel, produced from different feedstocks, is generally

spiked with different antioxidants to extend its storage stability
[23]. The environmental factors affecting the fate of biodiesel and
its blends with diesel at spillage site may vary with the physical
and chemical properties of the spilled contaminants, the environmental matrices, climate, weather, topography and hydrology.
Unfortunately, the impacts of environment matrices and physicchemical properties of biodiesel on the photolytic behavior of biodiesel have not been extensively studied till yet. The photolytic
behavior for individual petroleum hydrocarbons in biodiesel and
diesel blends has not been fully understood either.
The purpose of the present study is to investigate the effect of
physic-chemical properties of biodiesel and environmental matrices on the photolysis of biodiesel. The impact of biodiesel on the
photolytic behavior of individual petroleum hydrocarbons would
also be discussed. Biodiesel samples with different sources and initial concentration spiked into ultrapure (UP) water, and several
simulated freshwater matrices (UP water, UP water with humic
acid (HA) or pyrogallic acid (PY), and lake water) were irradiated
by UV and simulated sunlight rstly. The removal rates of main
FAMEs and total organic carbon (TOC) were measured and compared to evaluate the signicant contributors to the photolysis of
biodiesel in freshwater. Similarly, the depletion rates of petroleum
hydrocarbons with and without the presence of biodiesel after irradiated by UV light were measured and compared to investigate the
inuence of the presence of biodiesel on the photolytic behavior of
petroleum hydrocarbons.
2. Experimental procedures
2.1. Chemicals and materials
Solvents including hexane, dichloromethane (DCM) and acetone were supplied by Spectrum Chemicals (Gardena, CA, USA) at
the highest purity and used without further purication. Silica
gel (100200 meshes), humic acid and pyrogallic acid were purchased from Tianjin Chemical Co., Ltd. (Tianjin, China).
Normal alkane calibration standards from n-C9 to n-C40,
5a-androstane, and polycyclic aromatic hydrocarbon (PAH) calibration certied standard mixtures were purchased from Restek
(Bellefonte, PA, USA) and the US National Institute of Standards
and Technology (NIST, Gaithersburg, MD, USA), respectively.
Deuterated internal and surrogates including [2H14] terphenyl
(terphenyl-d14), [2H50] n-C24(C24D50), and PAH surrogate standards
including [2H8] naphthalene (naphthalene-d8), [2H10] acenaphthene (acenaphthene-d10), [2H10] phenanthrene (phenanthrened10), [2H12] benz[a]anthracene (benz[a]anthracene-d12), and
[2H12] perylene (perylene-d12), were supplied by Supelco (Bellefonte, PA, USA).
FAME mixtures, including 14 FAME standards ranged from C6 to
C24 with different saturated degree, surrogate of 13-methyl,
methyl myristate (13-methyl, C14:0), and internal standard of
methyl heptadecanoate (C17:0) purchased from SigmaAldrich
(Bellefonte, PA, USA) were employed for identifying and quantifying FAMEs. Detailed chemical information for FAME standards
used for calibration and identication are shown in Ref. [24].
2.2. Preparing fuel mixtures
Diesel purchased from a gas station in Wuhan, China, was used
in this study as the reference oil to be blended with biodiesel.
Three pure biodiesel samples sourced from soybean oil, canola
oil and animal fat was designated as Bsoy, Bca and Ban, respectively.
All samples were diluted with hexane to stock solution with a nal
concentration of 80 mg/mL.

250

Z. Yang et al. / Fuel 139 (2015) 248256

2.3. Exposure experiments

2.3.1. Biodiesel exposure
The diagram of UV or visible light exposure for biodiesel and its
blends with diesel is depicted in Fig. S1. In detail, exposures were
conducted by pouring 130 mL of ultrapure water (with a pH of 7,
18.2 MX of resistivity, 0.01 lg/mL of TOC, and undetectable cations, anions and total dissolved solids) into a 7-cm (i. d.)  10-cm
(height) glass reactor. Biodiesel diluted by acetone was added into
the water phase to a nal concentration ranged from 26.5, 62.5 to
125 lg/mL. Water samples were then exposed to UV light with a
light intensity of 33.7 mW/cm2, and Xenon lamp as a simulated
sunlight system with a light intensity of 55.8 mW/cm2. Because
light irradiation will heat the exposed water, an outside cooling
water recycle was employed to avoid the alternations of temperature during UV and sunlight exposure. The temperature in the irradiated water was maintained at 25 2 C. Biodiesel with different
sources and varied spiking concentration, and different water
matrices were tested to assess the factors affecting the photolysis
behavior of biodiesel. The water matrices included UP water, UP
water spiked with 10 lg/mL of humic acid or 8 lg/mL of pyrogallic
acid, and lake water. It is noted that the original TOC of water with
pyrogallic acid was determined to be close to 40 lg/mL. It began to
degrade after 2 h UV irradiation, and decreased to 16 lg/mL after
5 h. The TOC of water with humic acid was determined to be
10.4 lg/mL with a standard deviation of 0.9 lg/mL. It almost kept
constant within 5 h UV irradiation. The TOC and total nitrogen of
lake water was determined to be 18.7 0.6 and 11.0 0.8 lg/mL,
respectively. The UV irradiation within 5 h did not alter the TOC
for lake water. All TOC data were corrected based on their corresponding blank sample after exposed to UV light at same time.
For every test, two same samples were prepared. One sample
was exposed to light; the other one was kept at room temperature
in a dark area as a control sample to avoid microbial degradation
during UV/sunlight exposure. Sub-samples were taken at 0, 0.5,
1, 1.5, 2, 3, 4, and 5 h, respectively, for further analysis.
Similarly, the simulated solar irradiation within 24 h was tested
for all above mentioned conditions. In this scenario, samples were
taken at 0, 4, 8, 12, 16, 20 and 24 h, respectively, for further
analysis.
2.3.2. Exposure of biodiesel and diesel blends
The effect of the presence of biodiesel on the photolysis of individual petroleum hydrocarbons was evaluated by exposing diesel
and the blends of biodiesel and diesel with a mass ratio of 1:1 to
UV for 1, 2, 3, 4, and 5 h, respectively. The spiked concentration
of diesel and the blends was set as 40 lg/mL. After exposure, all
irradiated water samples were taken for liquidliquid extraction
(LLE), and DCM was used to rinse the glassware of the reactor for
further analysis. To eliminate the biodegradation and evaporation
loss, the one-to-one control samples were set up in the darkness
to act as the original standard without UV exposure for estimating
depletion rates of individual petroleum hydrocarbons.
2.4. Analytical procedures
2.4.1. Liquidliquid extraction (LLE)
For pure biodiesel samples, ten milliliter of water samples
spiked with 100 lL of 200 lg/mL 13-methyl, C14:0 as surrogate
were extracted with DCM three times for FAME analysis. The
extracts were combined and concentrated to about 0.9 mL, then
100 lL of 200 lg/mL C17:0 was added as the internal standard
for GC/MS analysis.
For blended samples, all water samples spiked with 100 lL
200 lg/mL of C24D50 (surrogate for n-alkanes), 100 lL 10 lg/mL
of PAH surrogate standards (surrogates for PAHs and their

alkylated homologues), and 100 lL 200 lg/mL of 13-methyl,

C14:0 (Surrogate for FAMEs), were extracted with DCM three
times. The LLE extracts combined with above glassware rinsing
solution were concentrated and solvent exchanged to hexane for
the following fractionation procedures.
2.4.2. Column chromatographic fractionation
For pure biodiesel, the above concentrated extracts were prepared with hexane to a nal concentration of approximate
200 lg/mL, and spiked with methyl heptadecanoate (C17:0) with
a nal concentration of 20 lg/mL as internal standard for direct
GC/MS analysis of FAMEs. However, for the blended biodiesel
and diesel extracts, column clean-up procedures were performed
prior to GC/MS analysis. This fractionation procedure is to fractionate extracts into petroleum hydrocarbons and FAMEs based on
their polarity difference, and avoid the interference of FAMEs during GC/MS analysis. Column clean-up and fractionation procedures
were adapted from one of our methods for chemical ngerprinting
of blends of biodiesel and diesel [24]. In brief, a 3-gram silica gel
column topped with a 1-cm layer of anhydrous sodium sulfate
was preconditioned with about 20 mL of hexane. Hexane (12 mL)
was used to elute aliphatic hydrocarbons (F1), and 15 mL mixture
of 1:1 hexane: DCM by volume was used to elute aromatic hydrocarbons (F2). F1 and F2 fractions were nally concentrated to about
1 mL, spiked with 100 lL of 200 lg/mL of 5a-androstane and
10 lg/mL of d14-terphenyl as the internal standards for GCMS
analysis of n-alkanes, petroleum-characteristic alkylated PAH
homologues and other US Environmental Protection Agency (US
EPA) priority PAHs, respectively. It is noted that the variation of
FAMEs in blends was not investigated in the present study, so
the elution of FAMEs was not mentioned here.
2.4.3. GC/MS and TOC analysis
Characterizations of n-alkanes, PAHs, petroleum biomarkers,
and FAMEs were performed on an Agilent 6890 GC system interfaced to an Agilent 5973 mass spectrometer. The n-alkanes, PAHs
and petroleum biomarker compounds were separated on a HP-5
MS capillary column (30 m  0.25 mm I.D., 0.25 lm lm thickness)
with the following temperature program: 50 C for 2 min, heated
to 300 C at 6 C/min and held for 15 min at 300 C.
A DB-225 MS GC column (30 m  0.25 mm i. d., lm thickness:
0.25 lm) from Agilent was employed to separate FAME and sterol
compounds with the following settings: initial temperature of
50 C, held for 1 min, then increased at a rate of 7 C/min to
185 C, held for 10 min, then increased at a rate of 15 C/min to
230 C, with a nal hold for 5 min.
For all GC/MS analysis, ultrahigh purity helium was employed
as carrier gas at a ow rate of 1 mL/min. The temperatures of injector, transfer line, ion source, and MS quadrupole analyzer were
held at 280, 280, 230, and 150 C, respectively. Samples were
injected in splitless/split mode. The mass-selective detector
(MSD) was operated at an electron impact mode (70 eV) for
selected ion monitoring (SIM) runs.
For TOC analysis, samples were acidied by hydrochloric acid,
and total inorganic carbon was removed by purging the acidied
sample with an inert gas prior to TOC measurement. All acidied
samples were rinsed with UP water till a pH of 7. Then organic carbon present in the pre-treated samples was oxidized to carbon
dioxide by catalytic combustion in the TOC analyzer. The relative
standards for TOC analysis were less than 10% for all
measurements.
2.5. Data analysis
Except for the analysis of the variation of FAMEs for pure
biodiesel and petroleum hydrocarbons for blends by GC/MS, the

251

Z. Yang et al. / Fuel 139 (2015) 248256

TOC%

TOC0  TOCt
 100
TOC 0

where TOC0 and TOCt are the measured TOC values at the beginning
and pre-determined reaction time, respectively. Similarly, the
transformation rates of FAMEs and petroleum hydrocarbons were
estimated from the following Eq. (2), where Cc and Ct are the measured concentration for one of the targets at the control and UV
exposure samples at pre-determined time points, respectively.
Therefore, all data reported in this study were control corrected.

Transformation %

Cc  Ct
 100
Cc

120
Bca
Ban
Bsoy

100

TOC removal rate (%)

variation of TOC after exposure to UV or visible light was also

recorded at the determined time points. The removal rate of TOC
can be expressed as TOC (%); it can be described as the following
Eq. (1):

80
60
40
20
0

UV irradiation time (h)

Fig. 1. Transformation of different biodiesel formulation vs. UV exposure time.

3. Results and discussion

The dominant FAMEs detected in Bca, Ban and Bsoy have a carbon
number of 18. Methyl oleate (cis-9, C18:1) comprises 64.0 4.2%
(m/m) of Bca. The other FAMEs from high to low concentration in
the Bca sample are methyl linoleate (cis-9, cis-12, C18:2), methyl
linolenate (cis-9, cis-12, cis-15, C18:3), methyl palmitate (C16:0),
methyl vaccinate (cis-11, C18:1), and methyl stearate (C18:0).
Methyl linoleate (cis-9, cis-12, C18:2) makes up 41.8 2.0% (m/
m) of the Bsoy sample. Similarly, the other FAMEs for the Bsoy from
high to low abundance are ranked as: cis-9, C18:1, C16:0, cis-9,
cis-12, cis-15, C18:3 and C18:0. For Ban, the FAMEs distribution
from high to low is followed as: cis-9, C18:1 (3542%) > cis-9,
cis-12, C18:2 or C18:0 (8.011.2%) > C16:0 (5.27.5%). The total
FAME composition in Bca, Bsoy and Ban were measured to be
102 4%, 95 3% and 96 4%, respectively. The chemical composition of all these biodiesel samples did not show signicant
variation in all control blank samples.
The chemical composition of the present used diesel is depicted
in Fig. S2. It can be seen that the distribution proles of n-alkanes
range from C9 to C30 with a bell shape, and their concentration
locates from 21.2 to 14,563 ng/g. The measured alkylated polycyclic aromatic hydrocarbons (APAHs) included 04 alkylated
naphthalene, phenanthrene, uorene and chrysene, they were
abbreviated as Ci-N, P, F and C (i = 04), respectively. It can be seen
that Ci-N are the most dominant components, followed by Ci-F,
Ci-P, and Ci-C. For all control blank samples, most of hydrocarbons
did not change signicantly, but some of light molecular weight
hydrocarbons (e.g., n-C9, n-C10, and the naphthalene family)
showed a decreasing tendency with the extension of exposed to
air. It is obvious that the loss is ascribed to the evaporation of light
molecular hydrocarbons.
3.2. Impact of biodiesel source on TOC removal rates
To investigate the effect of biodiesel source on its photolysis
behavior, the variation of TOC removal rates for biodiesel sourced
from soybean oil, canola oil and animal fat with exposure time is
depicted in Fig. 1. It can be seen that the accumulated TOC removal
rates increased with the irradiation time regardless of biodiesel
source. The nal total TOC removal rates reached up to 87.0% with
a relative standard derivation of 5.2% after 5 h UV irradiation.
Therefore, FAMEs and most of the produced intermediates have
been degraded within 5 h by UV irradiation. The accumulated
TOC removal rates followed the zero order kinetics with

correlation coefcients ranged from 0.945 to 0.971, the average

slope of the linear regression varied from 11.0, 11.8 to 12.6 for
Bca, Ban and Bsoy, respectively. As the slope represents the degradation rates of different biodiesel, it seems that Bsoy was degraded
faster than Ban and Bca. It is reasonable because the main FAME
component in Bsoy is cis-9, cis-12, C18:2; the most dominant component is cis-9, C18:1 in Ban and Bca. However, the difference of the
slopes can be neglected due to experimental errors. Therefore, the
contribution of biodiesel sources to the mineralization of biodiesel
can be neglected in the present study.
3.3. Effect of Bsoy concentration on TOC removal rate
As the removal of TOC is independent on the biodiesel source,
Bsoy was selected as the representative biodiesel to study its photolysis behavior in the following sections. The concentration of Bsoy
ranged from 25.0, 62.5 to 125 lg/mL was investigated to explore
the dependence of the TOC removal rates on the loading concentration of biodiesel in water (Fig. 2). It can be seen that all the TOC
removal rates increased with the extension of exposure time. However, they increased with the decrease of the initial Bsoy loading
concentration when the irradiation time was less than 2 h. After
that, similar TOC removal rates were observed for the above three
loading amounts. It is obvious that the energy to degrade FAMEs
and their produced intermediates are limited, therefore, the high
initial concentration of Bsoy resulted in the relatively low TOC

120
25 ppm
62.5 ppm
125 ppm

100

TOC removal rate (%)

3.1. Chemical composition analysis of biodiesel and diesel

80
60
40
20
0

UV irradiation time (h)

Fig. 2. Effect of Bsoy concentration on TOC removal rate.

252

Z. Yang et al. / Fuel 139 (2015) 248256

removal at the beginning of the reaction. However, the energy was

enough to remove the residual targets, when most of FAMEs and
intermediates have been removed after 2 h exposure. Linear
regression of the kinetics of TOC removal rates indicated that the
correlation coefcients ranged from 0.940 to 0.982; and average
slopes varied from 13.1, 12.6 to 12.6 for 25.0, 62.5 and 125 lg/
mL of Bsoy, respectively. It is obvious that the TOC removal processes obey the zero kinetics. The spiking level of 25.0 lg/mL
was degraded faster than the other two levels. However, they did
not change signicantly when Bsoy concentration increased from
62.5 to 125 lg/mL. Considering the analytical detection limits of
FAMEs, a concentration of 62.5 lg/mL was selected in the following sections.
3.4. Effect of sample matrices on TOC removal rates
Humic acid and PY, representing NOM and synthetic antioxidant, respectively, were spiked into UP water to test the effect of
sample matrices on TOC removal efciency. A natural lake freshwater was also selected as the representative of real water matrices. The background for these sample matrices have been
reported in Section 2.3.1. It can be found that although similar levels of humic acid and PY have been spiked, samples with humic
acid were determined to own lower TOC than PY. It is reasonable
because humic acid is a sodium salt, which is water soluble, and
contains around 20% inorganic residue. The variation of TOC
removal rates with different sample matrices is depicted in
Fig. 3. Similarly, all TOC removal rates increased greatly with the
extension of UV exposure till 34 h, a relatively slight increasing
tendency was observed after 4 h exposure. The nal TOC removal
rates reached up to 81.293.1% after 5 h UV exposure. However,
the TOC removal rates also slightly depended on the sample matrices. In detail, the TOC removal rates in UP water were generally
higher than the other sample matrices, followed by UP water with
PY and lake water, then UP water with humic acid. The linear
regression of the kinetics of TOC removal rates indicated that the
correlation coefcients ranged from 0.952 to 0.981, and average
slopes/degradation rates per unit time varied from 12.6, 11.9,
11.8 to 11.0 for UP water, UP water with PY/lake water and UP
water with humic acid, respectively.
It is obvious that the presence of other materials in water
slightly inhibited the photo-degradation of biodiesel and their
intermediates, because these chemicals may absorb part of light
or scavenge some free radicals [17]. The chemical composition in
lake water should be more complex than all the other three samples, however, similar TOC removal rates were observed in lake

100

TOC removal rates (%)

80
60
40

20
UP water
UP water with HA
UP water with PY
Lake water

Irradiation time
Fig. 3. Effect of sample matrices on TOC removal rates.

water and UP water with PY. The presence of humic acid in UP

water inhibited the photo-degradation behavior of biodiesel and
its intermediates more obviously. It was reported that humic materials are UV-absorbing species, which can compete UV with target
pollutants to reduce the efciency of PAH degradation [25]. Therefore, humic acid may compete UV light from FAMEs and/or their
intermediates, thus resulting in less TOC removal rates. Again, Mill
et al. [26] found that humic acid, as a light screening material,
decreased the light intensity in water, resulting in the decreased
degradation rates of polycyclic aromatic hydrocarbons.
3.5. Factors affecting the photolysis of individual FAME
Except for measuring the variation of TOC, the variation of several representative FAMEs was also recorded. The effects of sample
matrices on the removal of the representative FAMEs are shown in
Fig. 4. It can be seen that the FAME photolysis rates are mainly
dependent on their degree of saturation. However, sample matrices also affect their transformation rates. Firstly, the saturated level
of FAMEs controls their transformation rates. In detail, around 90
100% cis-9, cis-12, C18:2 was transformed within 1.5 h UV irradiation, then it almost kept constant after 2 h. Cis-9, C18:1 was
removed by 6080% within 1.5 h, it continued to be transformed
even after 5 h irradiation. Both of them have been transformed
90100% after 5 h UV exposure. It is noted that the variations of
cis-9, cis-12, cis-15, C18:3 were not monitored because its content
in the tested biodiesel are relatively low. It can be detected in the
original samples; however, it disappeared after 0.5 h UV irradiation, which indicates that cis-9, cis-12, cis-15, C18:3 can be transformed by UV irradiation very easily. Similarly, C16:0 and C18:0
were removed very quickly at the initial UV exposure stage, followed by a slow removal process till they were removed about
4060% after 5 h UV irradiation. It is obvious that the saturated
FAMEs of C16:0 and C18:0 are more photo stable compared with
cis-9, cis-12, C18:2 and cis-9, C18:1. Similar conclusions have been
reported by Khoury et al. [1].
Secondly, samples matrices also contributed to the photolysis of
FAMEs. Generally, FAMEs in lake water exhibited the lowest
removal rates, followed by UP water plus PY/humic acid, and then
in UP water. This conclusion is consistent with the variation rules
of TOC removal rates. For different FAME congeners, it seems that
C16:0 and C18:0 were affected by sample matrices more signicantly than cis-9, cis-12, C18:2 and cis-9, C18:1. It is reasonable
as unsaturated FAMEs are more vulnerable to photo-oxidation
than saturated ones [1]. Therefore, even sample matrices partially
absorb light or scavenge part of free radicals [17,25,26], the residual energy and radicals are strong enough to break the unsaturated
FAMEs, but not for saturated ones.
Fig. 4 also demonstrates that FAMEs, especially for unsaturated
FAMEs, were transformed very quickly at the initial UV irradiation,
followed by a platform or a slow increasing photolysis process.
However, the removal rates of TOC (Figs. 13) increased with the
extension of UV irradiation, no obvious platform was formed for
the removal rates of TOC. It is reasonable because cis-9, 12,
C18:2 and cis-9, C18:1 should have been transformed into some
intermediates prior to mineralization to carbon dioxide and water.
Although they could not be detected after they have been damaged
to other intermediates, these intermediates still contributed to the
measured TOC results. Therefore, different removal proles have
been achieved.
Similarly, a simulated sunlight irradiation within 24 h was
tested for all above mentioned scenarios. Unfortunately, the TOC
for all simulated samples did not decrease signicantly within
24 h irradiation. Individual FAME did not exhibit signicant
decrease either. Khoury et al. [1] studied the photolysis behavior
of FAMEs by exposure seawater with a oating layer of biodiesel

253

Z. Yang et al. / Fuel 139 (2015) 248256

120

120

cis-9, 12, C18:2

cis-9, C18:1
100

100

80
80
60
60

FAME photolysis rates (%)

40
UP water
UP water + HA
UP water + PY
Lake water

40

20

20

0
0

80

80

C16:0

C18:0

60

60

40

40

20

20

0
0

UV irradiation time (h)

UV irradiation time (h)

Fig. 4. Effect of sample matrices on the degradation rates of several representative FAMEs.

to sunlight. It was reported that cis-9, cis-12, cis-15, C18:3 was

removed completely after 3 days sunlight exposure, cis-9, cis-12,
C18:2 decreased dramatically in 3 days and disappeared totally
after 10 days. Therefore, 24 h sunlight irradiation was not long
enough for studying the photolysis behavior of biodiesel. The further investigation will be performed by extending exposure time
longer than 24 h for sunlight exposure.
3.6. Photolysis of diesel and blends of biodiesel and diesel
3.6.1. Removal kinetics of representative petroleum hydrocarbons
As mentioned in the above sections, most of FAMEs can be
removed within 5 h by UV irradiation. The removal kinetics of
petroleum hydrocarbons with and without the presence of biodiesel was simulated within 5 h UV exposure (Figs. 5 and 6, the original GC/MS data are shown Tables S1 and S2 in the Supporting
Material). Where the left panel depicts the variation of alkanes in
diesel samples, right panel demonstrates those in the blends of
biodiesel and diesel. It is obvious that all alkanes showed an
increasing removal rates with the extension of UV exposure in
two simulated samples, although some light molecular weight
alkanes almost completely removed within 1 h exposure. It seems
that alkanes with long carbon chain length are more reluctant to be
transformed compared with short ones. Unlike bio-degradation,
where n-alkanes can be degraded easier than branched alkanes,
cyclic alkanes, substituted/unsubstituted aromatics. For example,
n-C17 can be depleted preferably to pristine, and n-C18 is preferable
to phytane through biodegradation [27]. Herein, the pairs of n-C17/
pristine and n-C18/phytane were transformed at similar rate, which
suggests photo-oxidation treatment gives a completely different
result compared with biodegradation. Fig. 5 also illustrates that

the depletion of alkanes in blended samples is more random than

diesel sample. This suggests the presence of biodiesel affects the
photo-oxidation of alkanes (see detailed discussion in Section
3.6.2). It is noted that several depletion rates are lower than 0%,
which are acceptable as they locate in the limits of experimental
errors.
Similarly, the photolytic kinetics of representative alkylated
PAHs in diesel and blended samples are shown in Fig. 6. The photolysis rates for most of APAHs increased with the extension of UV
irradiation within 12 h, but the family of naphthalenes (Ci-N, i = 04)
were transformed quicker than the other high molecular weight
PAHs. More random variation was also observed for Ci-N series,
this discrepancy can be partially ascribed to the random evaporation loss of these targets. For most of phenanthrene (Ci-P) and uorene (Ci-F) families, they were transformed dramatically at the
rst 2 h exposure in both diesel and blended samples, followed
by an almost consistent photolysis rates in the following 3 h exposure. It can be found that APAHs were depleted more signicantly
than alkanes in the same sample by comparing Figs. 5 with 6. In
detail, almost 100% APAHs, but 80100% of alkanes have been
transformed in diesel sample after 5 h UV exposure. In blended
samples, all APAHs, except for C0-F, have been removed by 80
100%, but the removal rates of alkanes ranged from 20% to 100%.
It is obvious that aromatic hydrocarbons are more sensitive to
photo-oxidation than alkanes. In diesel samples, the photolysis
efciency was generally increased concurrently with the alkylation
level in each family of APAHs after 1 h UV exposure. It is reasonable
as the alkyl groups enrich the electron density of the p system of
the molecule, facilitating electron excitation and its resulting photolysis [28,29]. This effect should also be enhanced by the number
of aromatic rings in the molecule [30], but higher depletion of

254

Z. Yang et al. / Fuel 139 (2015) 248256

(a)

(b)

120

120

100

100

80

80

60

60
n-C10
n-C12
n-C14
n-C16
n-C17
Pristane

40

Photolysis rate (%)

20
0

n-C10
n-C12
n-C14
n-C16
n-C17
Pristane

40
20
0

-20

-20
0

120

120
n-C18
Phytane
n-C22
n-C26
n-C30

100
80
60

80
60

40

40

20

20

-20

-20
0

n-C18
Phytane
n-C22
n-C26
n-C30

100

UV exposure time (h)

Fig. 5. Photolysis kinetics of representative n-alkanes for diesel and the blends of diesel and biodiesel. (a), Diesel; (b), blends.

Naphthalene family
120

120

100

100

80

80
C0-N
2-ME N
1-ME N
C2-N
C3-N
C4-N

60
40

60
40

20

20
0

Photolysis rate (%)

Phenanthrene family
120

120

100

100

80

80
C0-P
C1-P
C2-P
C3-P
C4-P

60
40

60
40

20

20
0

Fluorene family
120

120

100

100

80

80

60

60
C0-F
C1-F
C2-F
C3-F

40

40

20

20
0

UV exposure time (h)

Fig. 6. Photolysis kinetics of representative APAHs for diesel and the blend of diesel and biodiesel. Left panel, diesel; right panel, blends.

Z. Yang et al. / Fuel 139 (2015) 248256

140

(a)
120

40 g/mL diesel
40 g/mL blend
20 g/mL blend

100
80
60
40

0
n-C10
n-C11
n-C12
n-C13
n-C14
n-C15
n-C16
n-C17
Pristane
n-C18
Phytane
n-C19
n-C20
n-C21
n-C22
n-C23
n-C24
n-C25
n-C26
n-C27
n-C28
n-C29
n-C30
n-C31
n-C32

Photolysis rate (%)

20

Individual n-alkanes
140

(b)
120
100
80
60
40
20

C0-N
2-ME N
1-ME N
C2-N
C3-N
C4-N
C0-P
C1-P
C2-P
C3-P
C4-P
Co-F
C1-F
C2-F
C3-F
Chry
C1-C
C2-C
C3-C

Individual APAHs
Fig. 7. Effect of biodiesel on the photolysis rates of individual petroleum hydrocarbons. Conditions: 5 h UV irradiation. Blends of diesel and Bsoy (1:1, m/m) with a
nal concentration of 40 lg/mL and 20 lg/mL in water phase. (a), n-Alkanes; (b),
alkylated PAHs. Where Ci-N, Ci-P, Ci-F and Ci-C represent the 04 alkylated
naphthalene, phenanthrene, uorene and chrysene, respectively.

naphthalene families have been observed due to the evaporation

loss during the test, which can be conrmed by the control samples
in the dark (Table S2). For blended samples, the photolysis extent
varied randomly. This also indicates the presence of biodiesel
affects the photo-oxidation of petroleum hydrocarbons. The
detailed discussion involving in the contribution of biodiesel to
the photolysis of petroleum hydrocarbons will be discussed in
the following section.

3.6.2. Contribution of biodiesel to the photolysis of individual

petroleum hydrocarbons
Although data at different time points within 5 h UV exposure
are available in the present study, only data with 5 h irradiation
are compared to investigate the effect of the presence of biodiesel
on photolysis rates of individual petroleum hydrocarbons (Fig. 7).
It can be seen that the removal of individual petroleum hydrocarbons with and without the presence of biodiesel exhibits with
some discrepancy. In detail, the photolysis for most of alkanes
was inhibited, especially for low and high molecular ones, but
not for middle ones (e.g., n-C13n-C17). For APAHs, the effect of biodiesel is not as signicant as alkanes, but obvious inhibition effect
can be found for uorene and chrysene series, especially for

255

chrysene. As the discussion in the upper section, the presence of

aromatic structure of PAHs enhances the electron excitement during UV irradiation, which makes them to be more favorable to be
transformed compared with alkanes [28,29]. Therefore, the photolysis of most of alkanes was signicantly inhibited, but not for
APAHs with the presence of biodiesel.
The potential reason for the inhibition effects of biodiesel on
diesel can be explained by the following two possibilities. Firstly,
the dominant components of biodiesel are fatty acid alkyl esters,
one end of these components are hydrophobic alkyl groups, the
other end is hydrophilic ester groups. They exhibit with the characters of surfactants when biodiesel is spiked into water. The presence of FAMEs could stabilize oil droplets in the water phase by
decreasing the oilwater surface tension and therefore reducing
oil droplet re-aggregation. These droplets experience longer lifetimes in the water phase before re-aggregating into larger globules
and rising to the surface. Therefore, oil stabilization by FAMEs may
increase the rate of petroleum hydrocarbons, especially for those
with heavier molecular weight, dissolution into water phase, due
to the increased surface area to volume ratio of smaller oil droplets
[4]. The direct contact probabilities between UV light and targets,
and radicals to attack targets would be reduced signicantly when
more petroleum hydrocarbons dissolution into water phase.
Accordingly, the photolysis of some heavy petroleum hydrocarbons in the presence of biodiesel was signicantly inhibited compared with pure diesel.
One home-made biodiesel without complete purication was
utilized to verify the above assumption rstly. More obvious inhibition effects have been observed in this scenario. As the removal
of glycerol is limited in this product, obvious micelles have been
visualized once it was spiked into aqueous phase. It is obvious that
more surfactant-like materials present in this product than the
commercial Bsoy, which resulted in more petroleum hydrocarbons
dispersing into water phase, and thus more signicant inhibition
effects were observed compared with the above commercial
biodiesel.
Based on this explanation concerning with the dissolution of
petroleum hydrocarbons into water phase, the inhibited photolysis
rates of alkanes with light molecular weights is an exception. These
alkanes may have been generated by photo-oxidative cleavage of
the alkyl side chains present in unresolved compound mixture
(e.g., the photo-oxidation of nonylbenzene [31], or the photo-oxidative decarboxylation of n-alkanoic acids or by the Norrish type
1 rearrangement via an alkyl radical of ketones [27], which can
be generated by the photo-oxidation of n-alkanes.
As mentioned in Section 2.3.2, 40 lg/mL of diesel and blend
samples were used to investigate the impact of biodiesel on the
photolysis of petroleum hydrocarbons. This means extra 40 lg/
mL of biodiesel presented in blends, except for 40 lg/mL of diesel.
Accordingly, the second possible reason for the inhibition effects
may be ascribed to the competitive consumption of UV light from
FAMEs. Blends with a concentration of 20 lg/mL (the mixture of
20 lg/mL biodiesel and 20 lg/mL diesel) are prepared to verify this
consumption (Fig. 7). Comparing the blends with two spiking levels, it can be seen that petroleum hydrocarbons with 20 lg/mL
spiking dosage were depleted more signicantly than 40 lg/mL.
However, comparing the 20 lg/mL of blends with the 40 lg/mL
of diesel samples, it can be found that the photolysis rates of petroleum hydrocarbons in the 20 lg/mL of blends were lower than in
40 lg/mL of diesel. As FAMEs are more vulnerable to be photo-oxidized than petroleum hydrocarbons, the photolysis rates of petroleum hydrocarbons in 20 lg/mL of blends should have been
removed more signicantly than that in 40 lg/mL of pure diesel
if the competition of FMAEs is the main reason for the inhibited
effects. This reversed phenomenon indicates that the competition
of UV light from FAMEs can be neglected. Therefore, the dominant

256

Z. Yang et al. / Fuel 139 (2015) 248256

possibility for the inhibited photolysis behavior for some heavy

molecular weight petroleum hydrocarbons can be ascribed to their
increased apparent solubility in the aqueous phase with the presence of FAMEs.
4. Conclusions
The photolysis of FAMEs and their mineralization efciency in
freshwater mainly depended on their degree of saturation, slightly
on water matrices and their initial concentration, regardless of
biodiesel sources. The presence of biodiesel stabilized small oil
droplets in the aqueous phase to increase the apparent solubility
of petroleum hydrocarbons, especially for heavy molecular weight
ones. Therefore, the direct contacting opportunities between UV
light and targets and radicals produced to attack targets were
reduced, which nally resulted in the inhibited photolysis rates
for some heavy molecular weight hydrocarbons.
Acknowledgement
This work was funded and supported by the National Natural
Science Foundation of China (Project No. 41373133), and the
Ministry of Science and Technology of China (No. 2012ZX07503003-002).
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.fuel.2014.08.061.
References
[1] Khoury RR, Ebrahimi D, Hejazi L, Bucknall MP, Pickford R, Hibbert DB.
Degradation of fatty acid methyl esters in biodiesels exposed to sunlight and
seawater. Fuel 2011;90:267783.
[2] Eaton S, Bartlett K, Pourfarzam M. Mammalian mitochondrial beta-oxidation.
Biochem J 1996;320:345.
[3] Zhang X, Peterson CL, Reece D, Haws R, Mller G. Biodegradability of biodiesel
in the aquatic environment. Am Soc of Agric Eng 1998;41. 14230-11430.
[4] DeMello JA, Carmichael CA, Peacock EE, Nelson RK, Samuel Arey J, Reddy CM.
Biodegradation and environmental behavior of biodiesel mixtures in the sea:
an initial study. Mar Pollut Bull 2007;54:894904.
[5] Miller NJ, Mudge SM. The effect of biodiesel on the rate of removal and
weathering characteristics of crude oil within articial sand columns. Spill Sci
Technol Bull 1997;4:1733.
[6] Glia Pereira M, Mudge SM. Cleaning oiled shores: laboratory experiments
testing the potential use of vegetable oil biodiesels. Chemosphere
2004;54:297304.
[7] Ferndez-lvarez P, Vila J, Garrido JM, Grifoll M, Feijoo G, Lema JM. Evaluation
of biodiesel as bioremediation agent for the treatment of the shore affected by
the heavy oil spill of the Prestige. J Hazard Mater 2007;147:91422.
[8] Fernndez-lvarez P, Vila J, Garrido-Fernndez JM, Grifoll M, Lema JM. Trials of
bioremediation on a beach affected by the heavy oil spill of the prestige. J
Hazard Mater 2006;137:152331.

[9] Mudge SM, Pereira G. Simulating the biodegradation of crude oil with biodiesel
preliminary results. Spill Sci Technol Bull 1999;5:3535.
[10] Yang Z, Hollebone B, Wang Z, Yang C, Landriault M. Effect of storage period on
the dominant weathering processes of biodiesel and its blends with diesel in
ambient conditions. Fuel 2013;14:34250.
[11] Taylor LT, Jones DM. Bioremediation of coal tar PAH in soils using biodiesel.
Chemosphere 2001;44:11316.
[12] Corseuil HX, Monier AL, Gomes APN, Chiaranda HS, do Rosario M, Alvarez PJJ.
Biodegradation of soybean and castor oil biodiesel: implications on the natural
attenuation of monoaromatic hydrocarbons in groundwater. Ground Water
Monit R 2011;31:1118.
[13] Owsianiak M, Chrzanowski L, Szulc A, Staniewski J, Olszanowski A, OlejnikSchmidt AK, et al. Biodegradation of diesel/biodiesel blends by a consortium of
hydrocarbon degraders: effect of the type of blend and the addition of
biosurfactants. Bioresour Technol 2009;100:1497500.
[14] Burrows HD, Canle LM, Santaballa JA, Steenken S. Reaction pathways and
mechanisms of photodegradation of pesticides. J Photochem Photobiol B: Biol
2002;67:71108.
[15] Garett RM, Pickering JJ, Haith CE, Prince RC. Photooxidation of crude oils.
Environ Sci Technol 1998;32:371923.
[16] Grifths MT, Da Campo R, O Connor PB, Barrow MP. Throwing light on
petroleum: simulated exposure of crude oil to sunlight and characterization
using atmospheric pressure photoionization furrier transform ion cyclotron
resonance mass spectrometry. Anal Chem 2014;86:52734.
[17] Payne JR, Phillips CR. Photochemistry of petroleum in water. Environ Sci
Technol 1985;19:56979.
[18] Ray PZ, Tarr MA. Petroleum lms exposed to sunlight produce hydroxyl
radical. Chemosphere 2014;103:2207.
[19] D Auria M, Emanuele L, Racioppi R, Vincenzina V. Photochemical degradation
of crude oil: comparison between direct irradiation, photocatalysis, and
photocatalysis on zeolite. J Hazard Mater 2009;164:328.
[20] Bobinger S, Andersson JT. Photooxidation products of polycyclic aromatic
compounds containing sulfur. Environ Sci Technol 2009;43:811925.
[21] Chacon JN, McLearie J, Sinclair RS. Singlet oxygen yields and radicals
contributions in the dye-sensitized photo-oxidation in methanol of esters of
polyunsaturated fatty acids (oleic, linoleic, linolenic and arachidonic).
Photochem Photobiol 1998;47:64755.
[22] Pesarini PF, de Souza RGS, Corra J, Nicodem DE, de Lucas NC. Asphaltene
concentration and compositional alterations upon solar irradiation of
petroleum. J Photochem Photobiol A: Chem 2010;214:4853.
[23] Dunn RO. Effect of antioxidants on the oxidative stability of methyl soyate
(biodiesel). Fuel Proces Technol 2005;86:107185.
[24] Yang Z, Hollebone B, Wang Z, Yang C, Landriault M. Method development for
ngerprinting of biodiesel blends by solid-phase extraction and gas
chromatographymass spectrometry. J Sep Sci 2011;34:325364.
[25] Shemer H, Linden KG. Aqueous photodegradation and toxicity of the polycyclic
aromatic hydrocarbons uorene, dibenzofuran and dibenzothiophene. Water
Res 2007;41:85361.
[26] Mill T, Mabey W, Lan B, Baraze A. Photolysis of polycyclic aromatic
hydrocarbons in water. Chemosphere 1981;10:128190.
[27] Dutta TK, Harayama S. Fate of crude oil by the combination of photooxidation
and biodegradation. Environ Sci Technol 2000;34:15005.
[28] Radovic J, Aeppli C, Nelson RK, Jimenez N, Reddy CM. Assessment of
photochemical processes in marine oil spill ngerprinting. Mar Pollut Bull
2014;79:26877.
[29] DAuria M, Emanuele L, Racioppi R, Velluzzi V. Photochemical degradation of
crude oil: comparison between direct irradiation, photocatalysis, and
photocatalysis on zeolite. J Hazard Mater 2009;164:328.
[30] Garrett RM, Pickering IJ, Haith CE, Prince RC. Photooxidation of crude oils.
Environ Sci Technol 1998;32:371923.
[31] Rontani JF, Bonin P, Giusti G. Mechanistic study of interactions between photooxidation and biodegradation of n-nonylbenzene in seawater. Mar Chem
1987;22:112.