Article

pubs.acs.org/JAFC

Insecticidal Activity of Bio-oil from the Pyrolysis of Straw from

Brassica spp.
Liu Suqi,†,‡ Luis Caceres,‡,§ Katie Schieck,‡ Christina J. Booker,§ Brian M. McGarvey,‡ Ken K.-C. Yeung,§
Stephane Pariente,∥ Cedric Briens,∥ Franco Berruti,∥ and Ian M. Scott*,‡
†
Shanxi Agricultural University, Taigu, Jinzhong, Shanxi 030801, People’s Republic of China
‡
Southern Crop Protection and Food Research Centre, Agriculture and Agri-Food Canada, London, Ontario N5V 4T3, Canada
§
Departments of Biochemistry and Chemistry, and ∥Institute for Chemicals and Fuels from Alternative Resources (ICFAR), Faculty
of Engineering, University of Western Ontario, London, Ontario N6A 5B8, Canada

ABSTRACT: Agricultural crop residues can be converted through thermochemical pyrolysis to bio-oil, a sustainable source of
biofuel and biochemicals. The pyrolysis bio-oil is known to contain many chemicals, some of which have insecticidal activity and
can be a potential source of value-added pest control products. Brassicacae crops, cabbage, broccoli, and mustards, contain
glucosinolates and isocyanates, compounds with recognized anti-herbivore activity. In Canada, canola Brassica napus straw is
available from over 6 000 000 ha and mustard Brassica carinata and Brassica juncea straw is available from 200 000 ha. The straw
can be converted by microbial lignocellulosic enzymes as a substrate for bioethanol production but can also be converted to bio-
oil by thermochemical means. Straw from all three species was pyrolyzed, and the insecticidal components in the bio-oil were
isolated by bioassay-guided solvent fractionation. Of particular interest were the mustard straw bio-oil aqueous fractions with
insecticidal and feeding repellent activity to Colorado potato beetle larvae. Aqueous fractions further analyzed for active
compounds were found not to contain many of the undesirable phenol compounds, which were previously found in other bio-
oils seen in the dichloromethane (DCM) and ethyl acetate (EA) solvent phases of the present study. Identified within the most
polar fractions were hexadecanoic and octadecanoic fatty acids, indicating that separation of these compounds during bio-oil
production may provide a source of effective insecticidal compounds.
KEYWORDS: pyrolysis, bio-oil, Brassica, mustard, Colorado potato beetle, fatty acids

■ INTRODUCTION
Plant natural products include a vast number of compounds
with biological activity against pathogens and insect pests. Of
particular interest are phytochemicals produced by crop plants
that provide resistance to disease or herbivores. One plant
family that has proven to have unique chemical defenses is the
Brassicaceae, including cabbage, broccoli, and oilseed rape or
canola.1 Glucosinolates, 120 of which have been identified, are
present as constitutive defenses in Brassicaceae. When attacked
by herbivores, glucosinolates are hydrolyzed by the enzyme
myrosinase and converted to more toxic and pungent products,
isothiocyanates, nitriles, and oxazolidinethiones, that can
protect the plant from insects. Because the production of
these compounds can protect these plants from herbivores,
there has been interest in applying the material as a green
manure and biopesticide.2−4
A large source of Brassicaceae plants in Canada is the yellow
and oriental mustards, Brassica hirta and Brassica juncea,
respectively. Mustard is grown in large acreages in the western
provinces for seed and oil to produce emulsifiers, with annual
average plantings between 1998 and 2007 of 130 000−328 000
ha.5 Mustard requires a relatively short growing season, 80−85
days for yellow mustard and 90−95 days for oriental mustard.
When grown under the hot, dry, summer weather of western
Canada, the oil content is reduced but the concentrations of
protein and glucosinolates are increased. Another source of
Brassica plant material in Canada is Ethiopian mustard, Brassica
carinata and canola Brassica napus. Also known as rape seed, B.
napus was originally used as a livestock feed, but toxicity
occurred because of the presence of glucosinolates and erucic
acid in the seeds.6,7 Canola was developed in Canada as a
cultivar of B. napus by breeding with Brassica campestris to
reduce the levels of these compounds in the oil and meal.8
Canola seed contains 42−43% oil, which is used as an edible
vegetable oil, and the remaining meal is turned into a high
protein animal feed at higher concentrations than the original
rape seed could be added. Canada produces 12.5 million tonnes
of canola seed per year,9 and the seed, oil, and meal are widely
traded. Canada also produces on average 178 000 tonnes of
mustard seed in the western provinces.10 Brassica seed oil has
been identified as a source of biodiesel and potential biorefinery
and bioindustrial platform for producing long-chain fatty-acid-
enhanced oils.11 Uses for the stalk or straw of the Brassicaceae
crops that are often left in the field include biomass for biofuels,
converted by either enzymatic hydrolysis12 or distillation.13 The
remaining mustard seed meal after the oil is removed has been
developed as a potent biopesticide. When the meal is mixed
with water, the glucosinolates and myrosinase can interact and
form isothiocyanates that can repel or even kill soil pathogens
Received: January 9, 2014
Revised: April 1, 2014
Accepted: April 3, 2014
Published: April 3, 2014

© 2014 American Chemical Society 3610 dx.doi.org/10.1021/jf500048t | J. Agric. Food Chem. 2014, 62, 3610−3618
Journal of Agricultural and Food Chemistry
and pest insects.4 The breakdown products of glucosinolates
include allyl isothiocyanate (AITC), allylcyanide, and 1-cyano-
2-hydroxy-3-butene, which have also been proposed for
fumigation of stored product pests because of the volatile
nature of these compounds.3 Oriental mustard, B. juncea, has
been developed into a biopesticide that can control nematodes,
a pest that leads to $78 billion U.S. in crop losses. AITC is the
principal product formed from allylglucosinolate (sinigrin) that
is found in B. carinata, B. juncea, and B. napus, with the highest
concentrations found in the seeds of B. juncea. The seeds and
bran or husk contain 5 and 2% sinigrin, respectively. Indian
mustard, B. juncea, contains glucosinolates that were shown to
reduce herbivore feeding and the associated damage to the
plant.14 The focus of the current project was to extract these
known compounds for biopesticide use through an alternative
process, thermochemical conversion.
Prior to this study, it was determined that many natural
products, including glucosinolates, are stable at moderately high
temperatures. Pyrolysis is the thermochemical process that can
convert abundant agricultural biomass residues into gases, bio-
oils, and char by heating under the absence of oxygen, and the
products from pyrolysis are recognized as a CO2-neutral
alternative compared to fossil fuels.15 The use of these end
products includes fuels, chemicals, and fertilizers. Previous
research by our group determined that pyrolysis products
produced from a variety of crop residues are active as pesticides,
and the chemical content was analyzed. Bio-oil from tobacco
stalks and leaves,15,16 tomato leaves and stalks,17 coffee
grounds,18 and cellulose, hemicellulose, and lignin19 had
antimicrobial and insecticidal activity. The separation of the
bio-oil components responsible for pesticidal activity has
identified many small molecules, including phenols and their
derivatives. However, these compounds may not be desirable
for crop protection because of the broad toxic effects of phenol
derivatives.20
For this reason, the current project focused on specific
compounds and biomass with known pesticidal properties and
applied pyrolysis as a unique method for separation and
isolation. In the present study, dried straw from B. napus, B.
carinata, and B. juncea, as a potential source of the known anti-
herbivore compounds and as a source of other insecticide
material, was converted by thermochemical means through the
process of fast pyrolysis. The resulting bio-oils were evaluated
as a source of material for pesticide use and then further
fractionated to isolate the active compounds. The hypothesis of
the project was that glucosinolates or isothiocyanates present in
the Brassicaceae straw material would contribute to the
insecticidal activity of extracts. In contrast, it was other
compounds, including fatty acids, hexadecanoic and octadeca-
noic acids, that were isolated in the active fractions and
determined to be potentially responsible for both acute toxicity
and feeding deterrence.
■
MATERIALS AND METHODS
Biomass and Insects. The canola B. napus and mustard B. juncea
and B. carinata straw were collected from Agriculture and Agri-Food
Canada (AAFC), Saskatoon, Saskatchewan, Canada, research fields in
October 2008 (batch 1). In 2010, B. juncea was collected from the
same research fields to provide a second year of biomass for a
comparison to the 2008 findings (batch 2). The field-dried straw was
ground using a blender/mixing mill and sieved to a Sauter mean
diameter of 60 μm for pyrolysis.
Colorado potato beetle (CPB), Leptinotarsa decemlineata Say
(Coleoptera: Chrysomelidae), was maintained without exposure to
3611
Article
insecticides for over 100 generations at the AAFC, London, Ontario,
Canada, laboratory. The colony was reared at an average temperature
of 25 °C, 50% relative humidity (RH), and a photoperiod of 16:8
light/dark on greenhouse grown potato plants (var. Kennebec).
Bioassay Method. A potato leaf disc bioassay was used to test the
insecticidal activity of the original bio-oil, and extracts were prepared
through aqueous and solvent separations or high-performance liquid
chromatography (HPLC) fractionation. A 50 μL sample of bio-oil
extract was spread on the top of a 4 cm diameter potato leaf disc, and a
100 μL sample of bio-oil extract was spread on the bottom of the leaf
disc. All leaf discs were held in the fume hood for 20 min to ensure
adequate drying before transferring the disc to a 4.25 cm Gelman Petri
dish with a moistened (30 μL of water) filter paper to maintain
humidity. Control leaf discs were treated with the same volume of
reverse osmosis (RO) water or acetone as a check for the aqueous- or
organic-phase bio-oil, respectively. Five first instar CPB larvae were
placed on each leaf disc, and the Petri dishes were put into a holding
chamber for 3 days (25 °C, 50% RH, and 16:8 light/dark). All trials
were repeated 3 times. Mortality and leaf consumption was assessed at
24, 48, and 72 h. Fatty acid, octadecanoic and hexadecanoic acids, and
methyl ester bioassays were conducted with similar volumes of each
compound dissolved in acetone at 1 and 10 mg/mL concentrations
and applied to potato leaf discs, as described above.
Pyrolysis and Bio-oil Preparation. The canola and mustard bio-
oils were produced by pyrolysis at 300 and 500 °C. Bio-oil produced in
the first batch was from all three Brassica species, while the second
batch was only B. juncea bio-oil. Further purification of the aqueous
fraction of the first batch mustard straw bio-oils focused on B. juncea
bio-oil because of the greater quantities of dry B. juncea bio-oil
available for testing. Pyrolysis was carried out using a fluidized-bed
pilot plant with a 0.078 m diameter atmospheric fluidized-bed reactor,
as described by Booker et al.15 The dried, ground straw was injected
into the reactor and produced vapors that exited at the top of the
reactor and were collected in a bio-oil-condensing system consisting of
a cyclonic condenser (aqueous and organic phases) immersed in
chilled water and an electrostatic precipitator (ESP) that charged fine
droplets escaping from the condenser and then used an electric field to
collect them. The bio-oil collected from the condenser contains both
polar and nonpolar compounds, while the ESP bio-oil is one-phase,
which contains nonpolar compounds that dissolves in acetone. The
condenser and ESP were weighed before and after each run to obtain
an accurate liquid yield. The amount of bio-oil collected in the
condenser and ESP was typically greater after 500 °C pyrolysis
compared to 300 °C. The 500 °C pyrolysis of ground B. juncea straw,
the condenser, and ESP yields were 24 and 14%, respectively, for a
total yield of 38% (all reported yields are on a mass basis). Reducing
the pyrolysis temperature to 300 °C decreased the condenser and ESP
yields to 17 and 6%, respectively, for a total yield of 23%. The same
reactor conditions yielded greater amounts of bio-oil when tobacco
leaf and coffee grounds were pyrolyzed at 500 °C, a 43% yield,15,18 and
similar yields were obtained with tobacco leaf using a batch or slow
pyrolysis.21
The bio-oil from the condenser was passed through a 9 cm
diameter, qualitative filter paper (Whatman, U.K.) and then separated
into aqueous and organic phases. Filtering removed solid material
(char and other precipitates) that was not soluble in either water or
organic solvent. All three phases (condenser aqueous and organic and
ESP phases) were dried and then redissolved in either water (aqueous
phase) or acetone (organic and ESP phases) to make a solution with a
concentration of 30 or 50 mg/mL. The bioactivity was tested by the
CPB−potato leaf dip bioassay.
Solvent Extraction. A selected bio-oil phase, 3 g, was separated by
mixing equal parts water and dichloromethane (DCM) (the ratio of
bio-oil/solvent was 1:50). A residue that did not dissolve in water,
ethanol, methanol, or DCM was partially dissolved in acetone. When
tested with the insect bioassay, no activity was observed with this layer.
The original bio-oil phase, fraction A, was separated into the water
phase, fraction B, and the DCM phase, fraction C (Figure 1). Fractions
B and C were evaporated to dryness at 50 °C and redissolved with
water (fraction B) and acetone (fraction C) to make a solution of 30
dx.doi.org/10.1021/jf500048t | J. Agric. Food Chem. 2014, 62, 3610−3618
Journal of Agricultural and Food Chemistry
Figure 1. Brassica bio-oil purification diagram. All bio-oil phases and
liquid−liquid separations colored blue indicate separations and
fractions that caused high CPB larval mortality and/or feeding
deterrence.
mg of dried B or C/mL, respectively, and the bioactivity was tested by
the CPB−potato leaf disc bioassay.
The batch 1 bio-oil fraction B was further separated with water and
ethyl acetate (EA) using equal parts of each (1:50 bio-oil/water−
solvent) to create fraction D (water−water phase) and fraction E
(water−EA phase). According to the bioassay results of these fractions,
only fraction D was considered for further fractionation and
purification. The second batch bio-oil fraction B was not separated
by water−EA but freeze-dried at −80 °C overnight, and the final
weight and concentration were calculated. Tests with the insect
bioassay were performed with 30 and 50 mg/mL concentrations
redissolved in water. The batch 2 aqueous fraction B for both the 300
and 500 °C bio-oils was measured in a preweighed vial and dried using
a stream of nitrogen. The samples were redissolved in water to a
concentration of 50 mg/mL and tested with the insect bioassay.
Gas Chromatography−Mass Spectrometry (GC−MS) Anal-
ysis. A HP 6890 Series gas chromatography system with a 5973 mass
selective detector was used to analyze the various extracted bio-oil
fractions. The experiments were performed on a HP-5MS, 30 m
column, with an inner diameter of 0.25 mm and a film of 0.25 μm
(Agilent Technologies). A 1 μL sample with a pulsed splitless injection
at 25 psi until 0.5 min was followed by purge flow to split vent 40 mL/
min at 1 m. The injector and auxiliary temperatures were maintained
at 300 °C. The oven temperature began at 60 °C for 2 min, then was
increased at 10 °C/min to 280 °C, and was held for 6 min. A mass/
charge scan range of 50−300 at a rate of 2.98 scans/s was used.16 The
National Institute of Standards and Technology (NIST) 2008 library
was used to identify the peaks found in the GC−MS chromatograms
for the Brassica bio-oil liquid−liquid phases. A total of 10 standards
were obtained to confirm the presence of the phenols previously found
in bio-oil fractions.16 These included phenol, 3-methylphenol, 4-
3612
Article
methylphenol, 4-ethylphenol, 2-methylphenol, indole, D-limonene, 2,4-
dimethylphenol, 3,4-dimethylphenol, and 2,3-dimethylphenol.
HPLC Fractionation and Isolation: Purification for Aqueous
Fraction D. To further separate the chemical components in the
batch 1 bio-oil fraction D, an Agilent 1200 Series HPLC was used with
a Waters symmetry semi-preparative C18 column (7 μm, 7.8 × 300
mm). This HPLC instrument was equipped with a photodiode array
detector and ChemStation software, which was used for the data
acquisition and analysis. Injections (200 μL) were made using an
Agilent 1200 autosampler. The binary mobile phase was composed of
water and acetonitrile (AcN). The flow rate was kept constant at 3
mL/min for a total run time of 70 min. The system was run with the
following gradient program: 100% water was run for 10 min, then
linearly changed to 80% water/20% AcN over 10 min, then linearly
changed to 50% water/50% AcN over 25 min, then linearly changed to
100% AcN over 5 min, where it was held for 10 min, and then brought
back to 100% water in 1 min, where it was held for 9 min. The time-
based trigger mode was used to collect 2 min fractions starting at 2
min. The concentration of fraction D was 30 mg/mL with 50
injections for a total of 10 mL of fraction D. The 34 fractions collected
were combined into 10 subfractions based on chromatogram peak
occurrences (Figure 1). All subfractions were evaporated to dryness,
and then 10 mL of distilled water was added to redissolve samples for
use in the leaf disc bioassay.
Purification of Aqueous Subfraction F1. According to the
bioassay results, the most polar fraction, F1, was the most active.
Further subfractionation was used to resolve the F1 compounds with
the same semi-preparative column as above. Samples were run
isocratically with 98% water/2% AcN for a total run time of 20 min.
The injection volume was 100 μL, and 50 injections were run.
Between 1 and 7 min, the flow rate was linearly decreased from 1 to
0.7 mL/min and held at 15 min. The trigger mode of the collector was
peak-based (from 8.5 to 11 min) and time-based (from 11 to 15 min,
with time slices of 1 min). The purification of F1 produced only five
peaks, and nine fractions were collected (Figure 1). All fractions were
evaporated to dryness, and 5 mL of water was added to redissolve for
the bioassay. The insecticidal activity was found primarily in fraction 1-
5 and partly in fraction 1-4.
Aqueous F1-5 Subfractionation. To isolate the individual peaks
in fraction F1-5, a Water’s amino semi-preparative column was used
(Nova NH2, 4 μm, 7.8 × 250 mm) to collect the F1-5 fractions. A 100
μL injection volume for a total of 40 injections was used, and the flow
rate was 3 mL/min for a total run time of 26 min. The system was run
with the following gradient program: 5% water/95% AcN isocratic for
3 min, then increased to 90% water/10% AcN at 15 min, and then
brought back to 5% water/95% AcN in 1 min, where it was held for 10
min. The trigger mode of the fraction collector was peak-based (from
4 to 9 min, with the maximum peak duration of 1 min) and time-based
(from 14 to 26 min, with time slices of 2 min). This purification of F1-
5 with the same program produced four peaks collected over 10
fractions, which were dried and then redissolved into 4 mL of water.
GC−MS Analysis. The aqueous fractions collected from the amino
semi-preparative column were analyzed using a 7890A GC system
(5975C MSD, Agilent Technologies) with a HP-5MS, 30 m × 250 μm
× 0.25 μm, column (Agilent). The oven temperature began at 70 °C
for 5 min and then increased at 5 °C/min to 290 °C, where it was held
for 10 min. A mass/charge scan range of 40−1050 was used at a rate of
1.49 scans/s. Samples were prepared by drying 20 μL of the 50 mg/
mL bio-oils in GC vials and preparing a 1 mg/mL solution. The bio-oil
solution was filtered using a 0.45 μm nylon CHROMSPEC filter, and
100 μL of the bio-oil solution was added to a GC vial, with 10 μL of
0.2 mg/mL ribitol being added. Solutions were dried with N2 for 5 min
and derivatized by adding 80 μL of methoxyamine hydrochloride
(MOX, 20 mg/mL in pyridine, Sigma) to the vial and heating at 50 °C
for 1.5 h in a heating block, after which 80 μL of N-methyl-N-
trimethylsilyltrifluoroacetamide (MSTFA, Sigma) was added with
heating at 50 °C for 0.5 h.
Fatty acid standards, hexadecanoic (palmitic) acid (Sigma) and
octadecanoic (stearic) acid (Eastman Organic Chemicals), were
dissolved in acetone at concentrations of 1, 5, 15, and 50 mg/L, and
dx.doi.org/10.1021/jf500048t | J. Agric. Food Chem. 2014, 62, 3610−3618
Journal of Agricultural and Food Chemistry Article

100 μL of each standard solution was added to 1.5 mL amber GC vials.

To each one of the vials containing the standard solutions, 10 μL of
0.2 mg/mL ribitol was added as an internal standard. Solutions were
dried with N2 for 5 min and derivatized as described above.
Statistical Analyses. Statistical analysis was undertaken using
Graphpad Prism 5.01 software, and the log-transformed and arcsine-
transformed data were subjected to the general linear model procedure
to test the significant differences between individual fractions using
two-way analysis of variation (ANOVA) and Bonferroni post-tests.

■ RESULTS AND DISCUSSION

Insecticide Bioassay-Guided Separation of Brassica
Bio-oil. The CPB bioassay determined that the batch 1
mustard straw condenser bio-oil produced at 300 °C was more
active than the oil produced at 500 °C, despite the higher yield
at 500 °C (data not shown). To isolate and identify the more
active components, the remaining tests focused only on the 300
°C bio-oil, where the condenser bio-oil was separated into
aqueous and solvent phases and compared to the ESP solvent-
soluble phase.
From the bioassay results, the aqueous bio-oil phases of all
three Brassica species collected in the condenser were just as
active as the ESP solvent phase because they all produced 100%
CPB mortality after 72 h at the 30 mg/mL concentration. The
condenser organic phase contained less active components
because there was <20% mortality for the two mustard species
bio-oils (Figure 2). On the basis of these results, further

Figure 3. Percent mortality and percent leaf consumption ± standard

error for CPB larvae after 24, 48, and 72 h of exposure to 3 and 30
mg/mL B. juncea aqueous (fraction B) and organic (fraction C)
separations. Bars with a letter are significantly different (p < 0.05; a, b,
and c for 24, 48, and 72 h, respectively) compared to control mortality
and leaf consumption at the same time.

purified with water (fraction D dissolved in water) and EA

Figure 2. Percent mortality ± standard error for CPB larvae after 24,
48, and 72 h of exposure to condenser aqueous, organic, and ESP
phases of B. carinata (BC), B. juncea (BJ), and B. napus (BN) bio-oil
(30 mg/mL) produced with 300 °C pyrolysis in batch 1. Bars with a
letter are significantly different (p < 0.05; a, b, and c for 24, 48, and 72
h, respectively) compared to control mortality at the same time.
examination focused on the aqueous B. juncea bio-oil phase
(fraction A of Figure 1) produced in batches 1 and 2 because of
its high availability. Fraction A was separated by liquid−liquid
extraction with water and DCM and produced a water fraction
(fraction B of Figure 1) and solvent fraction (fraction C of
Figure 1). Both fractions B and C at 30 mg/mL produced
significantly greater CPB mortality compared to controls after
48 h (p < 0.05) and reduced feeding at 24 h with relatively low
mortality [degrees of freedom (df) = 14; F = 34.1; p < 0.05;
Figure 3]. Fraction B was selected for further study because it
produced 25% greater mortality at 30 mg/mL than fraction C
(df = 14; F = 5.1; p < 0.05; Figure 3). Fraction B was further
3613
(fraction E dissolved in acetone), and bioassays determined
that fraction D produced 100% greater mortality at 30 mg/mL
compared to fraction E (df = 16; F = 7.3; p < 0.05; Figure 4).
There was no significant difference in mortality for 30 mg/
mL fraction E treatments compared to the solvent control (p >
0.05), but feeding was reduced at 48 h (p < 0.001; Figure 4).
GC−MS analyses of the separated bio-oil phases for B. juncea
batch 1 determined that fraction C contained phenols, such as
2-methoxyphenol and 2,6-dimethoxyphenol, while fraction E
contained a large proportion of phenols, including catechin, p-
dihydroxybenzene, and other benzene compounds (Table 1).
Hydroquinone, also present in fraction E, is one of several
that may be considered genotoxic and mutagenic, and 1,2-
benzenediol or catechol is a likely human and animal
carcinogen.20 In the redissolved condenser solid and the ESP
bio-oil, phenol compounds were also predominant. All phenol
derivatives are undesirable contaminants, and it would be best
to avoid these types of compounds in a biopesticide because
potential health and environmental problems could occur with
the necessary concentrations required for insect control.
Fraction D, the aqueous phase, did not contain phenol-related
compounds but did contain 1,1-dimethylhydrazine, considered
to be carcinogenic and a skin hazard.20 However, this
compound was not measured in the final aqueous fraction of
batch 2 (Tables 2 and 3).
dx.doi.org/10.1021/jf500048t | J. Agric. Food Chem. 2014, 62, 3610−3618
Journal of Agricultural and Food Chemistry Article

Table 2. GC−MS Characterization of Batch 2, 300 °C Bio-

oil Aqueous Phase with a Quality Match of ≥80% by NIST
2007
peak retention time (min) area (%) NIST assignment
1 11.42 1.29 L-(+)-lactic acid
2 11.91 1.93 glycolic acid
3 12.65 0.17 2-aminoethanethiol
4 17.67 2.29 1-hexadecanol
5 17.77 1.64 glycerol
6 19.37 0.40 glyceric acid
7 19.88 0.44 citraconic acid 1
8 22.78 1.32 threose 1
9 22.92 0.93 threose 2
10 27.96 0.31 ribose
11 33.44 0.14 D-(+)-altrose 1
12 36.60 0.36 palmitic acid
13 40.40 0.22 stearic acid

Table 3. GC−MS Characterization of Batch 2, 500 °C Bio-

oil Aqueous Phase with a Quality Match of ≥80% by NIST
2007
peak retention time (min) area (%) NIST assignment
a 10.68 0.23 2-hydroxypiridine
b 11.15 0.28 pyruvic acid
c 11.41 0.77 L-(+)-lactic acid
d 11.91 1.35 glycolic acid
e 17.66 1.28 1-hexadecanol
f 17.77 1.91 glycerol
Figure 4. Percent mortality and percent leaf consumption ± standard g 19.37 0.23 glyceric acid
error for CPB larvae after 24, 48, and 72 h of exposure to 3 and 30
h 19.44 0.22 succinic acid
mg/mL B. juncea aqueous−aqueous (fraction D) and aqueous−
organic (fraction E) separations. Bars with a letter are significantly i 19.88 0.37 citraconic acid
different (p < 0.05; a, b, and c for 24, 48, and 72 h, respectively) j 22.92 0.61 threose 2
compared to control mortality and leaf consumption at the same time. k 27.95 0.47 D-lixose 2

Bars with an asterisk indicate a significant difference between l 36.56 0.24 palmitic acid
treatments at the same time period within each separation phase m 40.37 0.24 steric acid
(aqueous or solvent).

Table 1. GC−MS Analyses of Batch 1 Bio-oil Samples and concentrations of individual compounds, and those phenols
Separated Fractions and their derivatives would be difficult to register because of
their broad toxic effects.20 In the present study, many similar
compounds present in the highest concentration in samples
sample from aqueous and solvent separated phases of the original compounds were found in the solvent phase of the B. juncea
(fraction) bio-oil prepared from B. juncea straw bio-oil collections, but fraction D did not contain these
condensor 2,6-dimethoxyphenol, 2-methoxyphenol, and compounds, even though insecticidal activity was present.
liquid 4-hydroxy-3,5-dimethoxybenzaldehyde Therefore, the insecticidal activity associated with fraction D
(fraction C)
was pursued as a potential source of compounds other than
condensor 1,1-dimethylhydrazine, 2-methylhexadecanal, and
liquid dimethyl sulfide phenols. Therefore, further research was conducted to isolate
(fraction D) the active compounds and analyze for any known harmful
condensor 1,2-benzenediol (catechol), hydroquinone, and chemicals within fraction D.
liquid 3-methoxy-1,2-benzenediol Insecticide Bioassay-Guided HPLC Fractionation of
(fraction E)
the B. juncea Aqueous Bio-oil Phase (Fraction D). To
condensor 2,6-dimethoxy-4-(2-propenyl)-phenol, 2,6-dimethoxyphenol,
solid and 4-hydroxy-3,5-dimethoxybenzaldehyde determine the active compounds in fraction D, further
ESP 2,6-dimethoxyphenol (syringol), separation of the liquid−liquid phase of the B. juncea condenser
2,6-dimethoxy-4-(2-propenyl)-phenol, and bio-oil was accomplished by HPLC fractionation into 10
(E)-9-octadecenoic acid fractions. Subsequent bioassay results indicated that the most
polar, fraction 1, collected between 2 and 4 min (Figure 5),
In a previous work by our group,16 much of the insecticidal produced the greatest mortality relative to the control and
and antimicrobial activities were associated with the solvent other nine fractions (df = 22; F = 41.6; p < 0.05; Figure 6).
phase of the tobacco bio-oil, because of the presence of many Fraction 1 produced the same CPB mortality and reduced
phenol compounds, especially the dialkylated phenols. leaf consumption as the original aqueous phase (df = 22; F =
However, it was concluded that the observed pesticide activity 25.5; p < 0.05; Figure 6). The HPLC chromatogram for peak
of the bio-oil solvent fraction would require multiple areas visible at 205 nm shows a number of partially resolved
combinations of the individual phenols or excessively high peaks within fraction 1 (Figure 5). To separate these peaks,
3614 dx.doi.org/10.1021/jf500048t | J. Agric. Food Chem. 2014, 62, 3610−3618
Journal of Agricultural and Food Chemistry
Figure 5. HPLC chromatogram showing B. juncea aqueous−aqueous
(fraction D) subfractionation and active fraction 1.
Figure 6. Percent mortality and percent leaf consumption ± standard
error for CPB larvae after 24, 48, and 72 h of exposure to 10 HPLC
fractions (F1−F10) of B. juncea aqueous−aqueous (fraction D)
separation. Bars with a letter are significantly different (p < 0.05; a, b,
and c for 24, 48, and 72 h, respectively) compared to control mortality
and leaf consumption at the same time.
HPLC purification of fraction 1 (F1) was again used to isolate
five peaks and nine subfractions were collected. Insecticidal
activity was found primarily in subfraction 1-5 and partially in
subfraction 1-4, where the 48 h CPB mortality was 53.33 and
20%, respectively (Figure 7). Again the subfraction activity was
not significantly different from F1 after 72 h (df = 9 and 54; F =
83.89; p > 0.05).
Both subfractions 1-5 and 1-4 reduced leaf consumption to 5
and 13.3% after 48 h of treatment, respectively (Figure 7), an
effect not significantly different to the activity demonstrated by
F1 (df = 9 and 54; F = 83.89; p > 0.05). Further purification of
F1-5 by HPLC using an amino column indicated the presence
of four peaks, and a further 10 collections were made. The
combined amino column fraction of F1-5 was analyzed using
3615
Article
Figure 7. Percent mortality and percent leaf consumption ± standard
error for CPB larvae after 24, 48, and 72 h of exposure to 10 HPLC
subfractions of the B. juncea aqueous−aqueous (fraction D) F1
fraction. Bars with a letter are significantly different (p < 0.05; a, b, and
c for 24, 48, and 72 h, respectively) compared to control mortality and
leaf consumption at the same time. Bars with an asterisk indicate a
significant difference between treatments at the same time period
within each separation phase (aqueous or solvent).
the Agilent 7890A GC−MS, and three of the major peaks were
identified as fatty acids, hexadecanoic and octadecanoic acids
(palmitic and stearic acids, respectively), and monooctadeca-
noylglycerol based on the NIST 2008 (version 2.0.f) library
match. Unfortunately, the concentration of each fatty acid
could not be calculated; however, the batch 2 condenser
aqueous samples prepared at 300 and 500 °C, as described in
the Materials and Methods, were also determined to contain
hexadecanoic and octadecanoic acids (panels A and B of Figure
8) among other compounds found in the 300 °C (Table 2) and
500 °C (Table 3) bio-oil aqueous fractions. These fractions
were also found to have insecticidal activity at 50 mg/mL,
significantly reducing leaf consumption relative to the controls
(df = 3; F = 110.6; p < 0.0001; Figure 9). No difference in
insecticidal activity was determined between the two temper-
ature bio-oil fractions (p > 0.05), but the concentration of both
fatty acids was 4−5-fold greater in the 500 °C aqueous bio-oil
fraction relative to the 300 °C aqueous bio-oil fraction (254−
309 versus 1023−1282 ppm). Similarly, there was no greater
feeding repellency observed with the 500 °C aqueous fraction
(p > 0.05; Figure 9). It was assumed that the fatty acids found
in aqueous fractions from both 300 and 500 °C pyrolysis likely
originated from the biomass and did not degrade during the
pyrolysis process. However, because the concentration of both
fatty acids was greater in the higher temperature bio-oil in batch
2, it is possible that these compounds are formed under
pyrolysis conditions or released in greater amounts from the
biomass. Unfortunately, this comparison could not be
dx.doi.org/10.1021/jf500048t | J. Agric. Food Chem. 2014, 62, 3610−3618
Journal of Agricultural and Food Chemistry
Figure 8. Total ion chromatogram (TIC) GC−MS, with 3× enhancement of the B. Juncea bio-oil aqueous phase: (A) fraction A produced at 300 °C
pyrolysis and (B) fraction B produced at 500 °C pyrolysis (batch 2).
confirmed with the batch 1 bio-oil. Because there was no
difference in the concentration between hexadecanoic or
octadecanoic acids in the aqueous phase of the batch 2
condenser bio-oil at either pyrolysis temperature, this confirms
that the relative amounts of each are similar under the same
heating conditions. Shin et al.22 found that octadecanoic acid
was stable at 300 °C and up to 370 °C, but unsaturated fatty
acids, such as oleic and linoleic acids, are less stable. This
suggests that higher temperatures, such as 500 °C, would
degrade fatty acids faster than 300 °C, but in this study, the
short pyrolysis residence time in the bubbling-bed reactor may
not be a sufficient exposure to degrade the compounds. Higher
pyrolysis temperatures will create increasingly smaller com-
pounds because the reaction breaks down large molecules
present in the biomass to form gases and other compounds,
such as phenols.
The presence of fatty acids in the aqueous fractions may
explain the insecticidal activity, both toxicity and feeding
deterrence. To confirm whether the fatty acids were active in
the 1−10 mg/mL range, a bioassay with standards for both
3616
Article
fatty acids and their respective methyl esters was used to
measure the feeding deterrent effect. No significant effect on
leaf consumption relative to controls was determined, even at
the higher concentration (p > 0.05), except with the methyl
ester of hexadecanoic acid at 10 mg/mL (df = 4; F = 8.18; p <
0.005; Figure 10).
Therefore, the two fatty acids are not solely responsible for
the insecticidal activity observed with the batch 2 bio-oil but
could have contributed to the activity observed with the
aqueous bio-oil fractions generated in batch 1 (concentrations
appeared higher but were not determined). Fatty acids have
been documented to have insecticidal activity against a number
of insect pests. As early as the 1850s, whale oil soap was found
to be the cheapest and most effective insecticide for the rose
chafer Macrodactylus subspinosus.23 Since then, other fatty acids,
such as oleic acid, lauric acid, etc., have been used to control a
variety of insect pests over the years, as reviewed by Kabara.23
More recent studies have documented that pure myristic acid at
50 μg/μL produced 53% mortality with wheat weevil Sotophilus
granarius, which were topically treated with 1 μL of solution per
dx.doi.org/10.1021/jf500048t | J. Agric. Food Chem. 2014, 62, 3610−3618
Journal of Agricultural and Food Chemistry
Figure 9. Percent mortality and percent leaf consumption ± standard
error for CPB larvae after 72 h of exposure to 50 mg/mL B. juncea
aqueous phase produced with 300 and 500 °C pyrolysis (batch 2).
Bars with an asterisk are significantly different (p < 0.001) compared
to respective control leaf consumption.
Figure 10. Percent mortality and percent leaf consumption ± standard
error for CPB larvae after 72 h of exposure to 10 mg/mL hexadecanoic
and octadecanoic acids and their methyl esters.
insect,24 and that lauric acid achieved 100% mortality against
cotton aphids Aphis gossypii, with a 0.5% (w/w) concen-
tration.25 In the latter case, the toxicity to the aphids was
probably related to suffocation and/or desiccation through
changes to the cuticle and cell permeability. Plant extracts that
contain fatty acids with insecticidal activity include the Egyptian
Halocnemon strobilacium (Chenopodiaceae).26 The crude
extract contained almost 70% fatty acids and their esters, with
the largest proportion being oleic acid ester (38%) and
octadecanoic acid (16%). Bioassays with individual fatty acids,
linolenic and linoleic acids, were found to have similar larvicidal
activity against the fall armyworm Spodoptera frugiperda.27 Both
linolenic and linoleic acids were found in the largest proportion
in hexane extracts of Ricinus communis (Euphorbiaceae) leaves,
48 and 15%, respectively, while hexadecanoic and octadecanoic
acids were lower (13 and 1.7%, respectively). However, as the
work of Shin et al.22 found, oleic and linoleic acids will not
3617
Article
survive higher pyrolysis temperatures, perhaps suggesting that
the hexadecanoic and octadecanoic acids would be the likely
fatty acids to be found in the bio-oil. In the present study,
hexadecanoic methyl ester was more active against the CPB
than hexadecanoic acid or the octadecanoic methyl ester. The
methyl esters of hexadecanoic acid were active against cabbage
looper Trichoplusia ni as both toxins and repellents.28 Herein,
GC−MS analysis of the two aqueous fractions in both the 300
and 500 °C bio-oils identified many other compounds based on
NIST 2008 library matches (Tables 2 and 3), a combination of
which could synergize the activity of the fatty acids but was not
tested. Sinigrin but not AITC was detected in most of the
Soxhlet extracts of B. carinata seed, B. juncea straw, and the bio-
oil aqueous phase for B. juncea (data not shown). However,
sinigrin elutes later than fraction 1 with the HPLC method
applied, and no insecticidal activity was associated with the
fractions containing sinigrin; therefore, it is not responsible for
the activity observed in fraction F1-5.
Conversion of the fatty acids isolated in the aqueous bio-oil
fractions to methyl esters by esterification will provide a more
valuable pesticide material. These interesting findings indicate
that the polar phase (fraction D) of pyrolysis bio-oils is a
promising use of agricultural residue biomass that can be
converted by a thermochemical process. It has yet to be
determined whether or not the fatty acid yield is affected by
pyrolysis, because this would suggest whether the starting
material is crucial to the final products. However, plant material
that contains high levels of fatty acids to begin with affords a
more promising source of biomass to extract these compounds
for insecticidal application. Considering the possible toxic
compounds produced during pyrolysis, any future pyrolysis bio-
oil products should be tested for non-target and environmental
toxicity prior to their recommendation for use in agriculture.
■ AUTHOR INFORMATION
Corresponding Author
*Telephone: 1-519-457-1470, ext. 281. Fax: 519-457-3997. E-
mail: ian.scott@agr.gc.ca.
Notes
The authors declare no competing financial interest.
■ ACKNOWLEDGMENTS
The funding for this research was provided by the Shanxi
Scholarship Council of China (2010064) and Agriculture and
Agri-Food Canada (AAFC). The authors gratefully acknowl-
edge the technical assistance of B. Pocs, D. MacArthur, and A.
Majkrzak. Brassicacae straw samples were supplied by K. Falk,
AAFC, Saskatoon, Saskatchewan, Canada, and partial funding
was provided by Agricultural Bioproducts Innovation Program
(ABIP), AAFC.
■
REFERENCES
(1) Hopkins, R. J.; van Dam, N. M.; van Loon, J. J. A. Role of
glucosinolates in insect−plant relationships and multitrophic inter-
actions. Annu. Rev. Entomol. 2009, 54, 57−83.
(2) United States Environmental Protection Agency (U.S. EPA).
Biopesticides Registration Action Document. Oriental Mustard Seed
(OMS). PC Code 014921; Office of Pesticide Programs, Biopesticides
and Pollution Prevention Division, U.S. EPA: Washington, D.C., Dec
17, 2008; p 15.
(3) Tsao, R.; Yu, Q.; Friesen, I.; Potter, J.; Chia, M. Factors affecting
the dissolution and degradation of oriental mustard-derived sinigrin
dx.doi.org/10.1021/jf500048t | J. Agric. Food Chem. 2014, 62, 3610−3618
Journal of Agricultural and Food Chemistry Article

and allyl isothiocyanate in aqueous media. J. Agric. Food Chem. 2000, hemicellulose-pyrolyzed bio-oil combinations: Estimating pesticide
48, 1898−1902. resources. J. Anal. Appl. Pyrolysis 2013, 99, 211−216.
(4) Yu, Q.; Tsao, R.; Chiba, M.; Potter, J. Selective nematicidal (20) International Agency for Research on Cancer (IARC). Re-
activity of allyl isothiocyanate. J. Food Agric. Environ. 2007, 5, 374− evaluation of some organic chemicals, hydrazine and hydrogen
379. peroxide. IARC Monogr. Eval. Carcinog. Risks Hum. 1999, 71, 1586.
(5) Agriculture and Agri-Food Canada (AAFC). Profile. Canada’s (21) Hossain, M. M. New approaches to obtain pesticides from
Mustard Seed Industry. Mustard Seed Statistics; AAFC: Ottawa, Ontario, biomass pyrolysis. Master’s Thesis, Department of Chemical and
Canada, 2013; http://www.agr.gc.ca/eng/industry-markets-and-trade/ Biochemical Engineering, Faculty of Engineering, Western University,
statistics-and-market-information/by-product-sector/crops/pulses- London, Ontario, Canada, 2011; p 95.
and-special-crops-canadian-industry/mustard-seed/?id= (22) Shin, H.-Y.; Ryu, J.-H.; Park, S.-Y.; Bae, S.-Y. Thermal stability
1175116081724 (accessed Oct 29, 2013). of fatty acids in subcritical water. J. Anal. Appl. Pyrolysis 2012, 98, 250−
(6) Fenwick, G. R.; Heaney, R. K. Glucosinolates and their 253.
breakdown products in cruciferous crops, foods and feeding stuffs. (23) Kabara, J. J. Fatty acids and esters as antimicrobial/insecticidal
Food Chem. 1983, 11, 249−271. agents. In Ecology and Metabolism of Plant Lipids; Fuller, G., Nes, D.,
(7) Stefansson, B. R.; Hougen, F. W. Selection of rape plants Eds.; American Chemical Society (ACS): Washington, D.C., 1987;
(Brassica napus) with seed oil practically free from erucic acid. Can. J. ACS Symposium Series, Vol. 325, Chapter 14, pp 220−238.
Plant Sci. 1964, 44, 359−264. (24) Abay, G.; Altun, M.; Karakoc, O. C.; Gul, F.; Demirtas, I.
(8) Cheeke, P. R.; Shull, L. R. Natural Toxicants in Feeds and Insecticidal activity of fatty acid-rich Turkish bryophyte extracts against
Poisonous Plants; AVI Publishing Company, Inc.: Westport, CT, 1985; Sitophilus granarius (Coleoptera: Curculionidae). Comb. Chem. High
p 492. Throughput Screening 2013, 16, 806−816.
(9) Canola Council of Canada. Estimated Acreage and Percentage of (25) Mohamad, S. F. S.; Mohamad, S.; Aziz, A. A. The susceptibility
of aphids, Aphis gossypii Glover to lauric acid based natural pesticide.
HT and Conventional Canola; Canola Council of Canada: Winnipeg,
Procedia Eng. 2013, 53, 20−28.
Manitoba, Canada, Dec 3, 2010; http://www.canolacouncil.org/
(26) Abdallah, S. A.; Barakat, A. A.; Badawy, H. M. A.; Mansour, F.
markets-stats/statistics/estimated-acreage-and-percentage/ (accessed
A.; Solimann, M. M. M. Toxicological and phytochemical studies of
Oct 29, 2013). wild plant Halocnemon strobilacium crude extracts and their
(10) Canadian Grain Commission. Quality of Western Canadian components against Aphis craccivora Koch. J. Appl. Sci. Res. 2009, 5,
Mustard; Canadian Grain Commission: Winnipeg, Manitoba, Canada, 699−705.
2010; http://www.grainscanada.gc.ca/mustard-moutarde/harvest- (27) Ramos-López, M. A.; González-Chávez, M. M.; Cárdenas-
recolte/2010/hqm10-qrm10-3-eng.html (accessed Oct 29, 2013). Ortega, N. C.; Zavala-Sánchez, M. A.; Pérez, G. S. Activity of the main
(11) Taylor, D. C.; Falk, K. C.; Palmer, C. D.; Hammerlindl, J.; Babic, fatty acid components of the hexane leaf extract of Ricinus communis
V.; Mietkiewska, E.; Jadhav, A.; Marillia, E.-F.; Francis, T.; Hoffman, against Spodoptera frugiperda. Afr. J. Biotechnol. 2012, 11, 4274−4278.
T.; Giblin, E. M.; Katavic, V.; Keller, W. A. Brassica carinataA new (28) Antonius, G. F.; Meyer, J. E.; Rogers, J. A.; Hu, Y.-H. Growing
molecular farming platform for delivering bio-industrial oil feedstocks: hot pepper for cabbage looper, Trichoplsia ni (Hübner), and spider
Case studies of genetic modifications to improve very long-chain fatty mite, Tetranychus urticae (Koch) control. J. Environ. Sci. Health, Part B
acid and oil content in seeds. Biofuels, Bioprod. Biorefin. 2010, 4, 538− 2007, 42, 559−567.
561.
(12) Pal, S.; Banik, S. P.; Khowala, S. Mustard stalk and straw: A new
source for production of lignocellulolytic enzymes by the fungus
Termitomyces clypeatus and as a substrate for saccharification. Ind.
Crops Prod. 2013, 41, 283−288.
(13) Stamigna, C.; Chiaretti, D.; Chiaretti, E.; Prosini, P. P. Oil and
furfural recovery from Brassica carinata. Biomass Bioenergy 2012, 39,
478−483.
(14) Mathur, V.; Ganta, S.; Raaijmakers, C. E.; Reddy, A. S.; Vet, L.
E. M.; van Dam, N. M. Temporal dynamics of herbivore-induced
responses in Brassica juncea and their effect on generalist and specialist
herbivores. Entomol. Exp. Appl. 2011, 139, 215−225.
(15) Booker, C. J.; Bedmutha, R.; Vogel, T.; Gloor, A.; Xu, R.;
Ferrante, L.; Yeung, K.K.-C.; Scott, I. M.; Conn, K. L.; Berruti, F.;
Briens, C. Experimental investigations into the insecticidal, fungicidal,
and bactericidal properties of pyrolysis bio-oil from tobacco leaves
using a fluidized bed pilot plant. Ind. Eng. Chem. Res. 2010, 49, 10074−
10079.
(16) Booker, C. J.; Bedmutha, R.; Scott, I. M.; Conn, K.; Berruti, F.;
Briens, C.; Yeung, K.K.-C. Bioenergy II: characterization of pesticide
properties of tobacco bio-oil. Int. J. Chem. Reactor Eng. 2010, 8, A26.
(17) Caceres, L. A.; McGarvey, B. D.; Scott, I. M.; Conn, K.; Berruti,
F.; Briens, C.; Yeung, K. K. Pesticidal properties of tomato plant
residue bio-oil. 2011 Pest Management Research Report2011 Growing
Season; Agriculture and Agri-Food Canada (AAFC): Ottawa, Ontario,
Canada, May 2012; pp 50−59, http://www.cps-cp.ca/publications.
shtml.
(18) Bedmutha, R.; Booker, C. J.; Ferrante, L.; Briens, C.; Berruti, F.;
Yeung, K.K.-C.; Scott, I.; Conn, K. Insecticidal and bactericidal
characteristics of the bio-oil from the fast pyrolysis of coffee grounds. J.
Anal. Appl. Pyrolysis 2011, 90, 224−231.
(19) Hossain, M. M.; Scott, I. M.; McGarvey, B. D.; Conn, K.;
Ferrante, L.; Berruti, F.; Briens, C. Toxicity of lignin, cellulose and

3618 dx.doi.org/10.1021/jf500048t | J. Agric. Food Chem. 2014, 62, 3610−3618