Industrial Crops and Products 94 (2016) 178188

Contents lists available at ScienceDirect

Industrial Crops and Products

journal homepage: www.elsevier.com/locate/indcrop

Composition and anti-bacterial activity analysis of citronella oil

obtained by hydrodistillation: Process optimization study
Robinson Timung a , Chitta Ranjan Barik a,b , Sukumar Purohit a,b , Vaibhav V. Goud a,b,
a
Department of Chemical Engineering, Indian Institute of Technology Guwahati, Assam 781039, India
b
Centre for Energy, Indian Institute of Technology Guwahati, Assam 781039, India

a r t i c l e i n f o a b s t r a c t

Article history: Hydro-distillation was performed to determine the quantity of essential oil in different parts of Java
Received 7 April 2016 Citronella (Cymbopogon winterianus Jowitt) plant. Parameters affecting the oil extraction process were
Received in revised form 10 August 2016 evaluated using Response Surface Methodology. The generated second order polynomial model was
Accepted 12 August 2016
highly signicant with R2 = 0.9744 and P < 0.0001. Highest yield of 2.38% vw1 was obtained using the
Available online 29 August 2016
leaves part at distillation time of 180 min, which was about 40% more than stems part and 17% more than
whole aerial parts. Higher essential oil content of fresh plant (2.43% on a dry weight basis) compared to
Keywords:
dried plant (2.12% on a dry weight basis) indicates that the moisture content has a signicant effect on
Citronella oil
Response surface methodology
oil yield. GC/MS analysis of the oil extracted from leaves part revealed about 95% of commercially impor-
Hydrosol tant compounds viz. citronellal (55.23%), geraniol (26.29%) and citronellol (13.41%), indicated very high
Anti-bacterial activity quality citronella oil. The oil was found effective against all the six tested bacteria strains namely Bacillus
Physico-chemical characterization subtilis, Staphylococcus epidermidis, Staphylococcus aureus, Pseudomonas aeruginosa, Klebsiella pneumonia
and Escherichia coli, which revealed that citronella oil could be applied for the treatment of infections
caused by various gram positive or negative pathogenic organism. The hydrosol obtained after hydro-
distillation was analyzed using HPLC and was found to contain sugar alcohols such as adonitol, erythritol
and mannitol which could be utilized in nutraceutical industries.
2016 Elsevier B.V. All rights reserved.

1. Introduction et al., 2013). Even though the essential oil content in the plant parts
is very low (typically 18% of total weight of the plant), but has high
Essential oils are essences of aromatic plant species obtained by industrial importance and market value. Therefore, it is important
the hydrodistillation or steam distillation of whole plants or from to optimize the extraction process parameters in order to obtain a
certain parts such as owers, fruits, leaves, roots, barks and seeds high yield and good-quality oil (Luthria and Natarajan, 2009).
(Muazu et al., 2012). Different methods of extraction like distilla- Citronella oil is an essential oil known for its natural insect
tion (hydro and steam), solvent extraction and supercritical uid repellent property and is of great interest for pharmaceuti-
extraction can be used to extract essences or volatiles. However, cal and fragrance industry (Wany et al., 2013). It has various
the quality and quantity of the oil yield depend on extraction tech- therapeutic uses as analgesic, anticonvulsant, anxiolytic, etc.
nique used (Chanthai et al., 2012). Essences or volatiles obtained and is a favorable agent towards anti-fungal, anti-bacterial,
by the solvent extraction may contain traces of the solvent which anti-parasitic and nematicidal activities (Almeida et al., 2001,
may have deleterious effect. On the other hand, supercritical uid 2004). In traditional practice, Citronella oil has been used as an
extraction is although an efcient technique to produce high yield antipyretic, aromatic tea, vermifuge, diuretic and in mental ill-
and good-quality essences or volatiles, but the initial cost of invest- ness (Wany et al., 2014). Citronella grass has been traditionally
ment is very high. Hence, distillation methods which are relatively used in Thailand to treat stomach ache, indigestion, intesti-
simple, cheap, environment friendly, produce good quality oil are nal cramps, irritable bowel, atulence, gastritis and as a blood
generally preferred for essential oil extraction from grasses (Manaf tonic (Salguero, 2003). Citronella essential oil is obtained from
the citronella plant which is an aromatic grass belonging to
the Kingdom- Plantae, Clades-Angiosperms, Monocots and Com-
melinids, Order-Poales, Family-Poaceae, Subfamily-Panicoideae,
Corresponding author at: Department of Chemical Engineering, Indian Institute
Tribe-Andropogoneae, Sub tribe- Andropogoninae and Genus-
of Technology Guwahati, Assam 781039, India.
Cymbopogon Spreng, according to the APG III system of owering
E-mail address: [email protected] (V.V. Goud).

http://dx.doi.org/10.1016/j.indcrop.2016.08.021
0926-6690/ 2016 Elsevier B.V. All rights reserved.
 R. Timung et al. / Industrial Crops and Products 94 (2016) 178188 179

plant classication (APG III, 2009). From the genus Cymbopogon

two species are commercially categorized: (1) Ceylon Citronella
(Cymbopogon nardus Rendle) and (2) Java Citronella (Cymbopogon
winterianus Jowitt) (Wijesekara, Wijesekara, 1973). Java Citronella
oil containing about 85% of commercially important compounds
like citronellal, geraniol and citronellol is considered to be more
superior to that of Ceylon Citronella oil, which contains only 5565%
of these three compounds (Wijesekara, 1973; Wany et al., 2014).
Therefore, Java type is more widely cultivated in tropical and sub-
tropical countries like India, Sri Lanka, Malaysia, Taiwan, Ecuador,
Madagascar, Guatemala, Honduras, Brazil, Argentina, Mexico and
West Indies (Wany et al., 2013). The North Eastern part of India is
one of the most important biodiversity hotspot having diverse agro
climatic zones, which provide ample conditions for the cultivation
of Java Citronella. Major Java Citronella oil-producing states of India
are Assam, Karnataka, Uttar Pradesh, Madhya Pradesh, Maharash-
tra, Tamil Nadu and West Bengal (Wany et al., 2013). The region
experiencing well distributed rainfall all over the year is considered
favorable for the cultivation of citronella plant. Sandy loamy soil
with high moisture content, but without water logging is the appro-
priate soil condition for plant growth. The best planting period is
the rainy season, and the most ideal season of harvesting is sum- Fig. 1. Representation of different parts of citronella plant.
mer and the early winter. Citronella plants can be harvested 34
times annually with the interval of 2.53 months, starting the rst
harvest six months after plantation. However, harvest during the
late winter signicantly reduces the oil yield (Joy et al., 2001; Blank
et al., 2007). In order to extend the shelf life of the plant mate-
rial, post harvest drying is important, which reduces the moisture
content and prevents the microbial and enzymatic activity (Rocha
et al., 2011). Although biosynthesis of essential oil is genetically
regulated, but still environmental factors also have a high impact.
Citronella plants grown in the North Eastern region of India are rich
in Citronellal content (Ahmed, 2005).
Although a lot of work has been reported on the citronella
plant, but limited literature is available on the evaluation of oil
content and quality from different parts of the plant and optimiza-
tion of the oil extraction using hydro-distillation technique as well
as evaluation of the anti-bacterial activity of the oil. Hence, the
present study has the objectives of optimization of essential oil
extraction from different parts of the Java Citronella plant using
hydro-distillation technique, physico-chemical characterization of
extracted citronella oil and to evaluate the anti-bacterial activity of
the extracted oil on some selected bacterial strains.
Fig. 2. Effect of distillation time on percentage oil yield.

2. Materials and methods

2.1. Materials

Java Citronella (Cymbopogon winterianus Jowitt) was collected

from Karbi Anglong, Assam, North-East India in the month of
December (winter season). Leaf lamina or leaf blades were sepa-
rated from the whole aerial parts and were used as leaves part.
The remaining culm and leaf sheaths were considered as stem and
leaf sheath part (Fig. 1). Whole aerial plants were initially shade
dried in the semi-dark room for 1015 days and then chopped
into small pieces. The plant sample is then stored separately in
zipped plastic bags at an ambient temperature. Millipore water
was obtained using Millipore water synthesis unit bearing Model
No: Elix-3, Milli Q; Make: Millipore, USA. 3451-Clevenger appara-
tus (1000 ml capacity) purchased from Borosil was used. All the
bacterial strains (three gram positive and three gram negative)
were purchased from MTCC, Chandigarh. All the chemicals used for
analysis were obtained from Merck India Pvt. Ltd. Analytical grade Fig. 3. Effect of water volume on percentage oil yield.
potassium bromide and HPLC grade hexane was used. The standard
180 R. Timung et al. / Industrial Crops and Products 94 (2016) 178188

determine and dene the levels of an independent factor. A three

level-three factor design was employed. Considering 5 central
points, total of 17 experimental runs were generated consisting
6 factorial points and 6 axial points. A second-order polynomial
model was generated to express the responses as a function of
independent factors. Taking Y as predicted response (oil yield), 0
as the model constant and 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 and 9
as coefcients associated with factor A, B, C and their respective
interactions; the model equation was generated as given in Eq. (3).

%Y = 0 +1 A+2 B+3 C+4 AB+5 AC+6 BC+7 A2 +8 B2 +9 C2 (3)

2.2.4. Sample analysis using FTIR

IR spectra of citronella oil (acquired from leaves part at high-
est oil yield) was obtained from FTIR spectrometer IR Afnity1
(Shimadzu Corporation, Japan), measured in the range between
4000400 cm1 with resolution of 2 cm1 and 30 scans per spec-
trum. Potassium bromide (KBr) was used for reference. The KBr
Fig. 4. Effect of plant part used on percentage oil yield. pellet was prepared and about 20 l citronella oil was spread over
the pellet and then directly taken for analysis. Baseline corrections
and variation of frequency was applied using Shimadzu IR solution
sugars and sugar alcohols used in HPLC analysis were obtained from
1.5 software supplied with the equipment.
Himedia and Sigma Aldrich.

2.2. Methods 2.2.5. GC/MS analysis

The qualitative and quantitative analysis of the extracted cit-
2.2.1. Moisture content analysis of citronella biomass ronella oil was carried out using Perkin Elmer Clarus 680 GC/600C
The moisture content in the plant material (whole aerial) was mass selective detector (MSD) system. Fused silica HP-5MS col-
estimated by oven drying the sample at 105 2 C for 24 h as umn cross-linked with 5% phenylmethylsiloxane of 60 m length
reported in CEN/TC 335, (2004). The moisture contents in the plant and internal diameter of 0.25 mm was used. The initial oven tem-
samples were estimated according to the following relation (Eq. perature was 50 C for 2 min and then ramp at 5 C min1 up to
(1)). 260 C. Injector and transfer line temperature was 250 C and 200 C
Initial weight Final oven dried weight respectively. The holding time was 2 min and the solvent delay time
% Moisture Content = 100 (1)
Initial weight was 8 min. Helium was used as a carrier gas at 1 ml min1 . 2 l
volume of oil sample (2% solution of citronella oil in hexane) was
2.2.2. Extraction procedure injected. Split ratio 50:1 and scan range of 50600 Da was used.
Variable amount (1030 g) of different parts of plant material The identication of constituents was done on the basis of their
and Millipore water (500 ml) were charged into 1000 ml capacity retention time and mass spectra library search (NIST).
round bottom ask of Clevenger apparatus for hydro-distillation.
Heating was provided using a heating mantle (Ikon instruments,
2.2.6. SEM analysis
Delhi-95). Distillation time was accounted from the moment water
Plant samples before and after treatment was grinded and
starts boiling. The oil was collected, dried over anhydrous sodium
sieved to 1 mm size using 16 BSS mesh. The powdered sample
sulphate and stored in an amber bottle at 4 C until used. Extrac-
was sprayed over a conductive carbon tape attached to the sam-
tions were performed at least three times, and the mean values
ple holder. As the sample was non-conducting, sputter gold coating
were reported. The oil yield was calculated from the relation as
was performed to prevent charging. All the specimens were exam-
given in Eq. (2).
  ined using LEO 1430VP Scanning Electron Microscope under the
Volume in ml of extracted oil
% Yield v.w-1 = 100 (2) vacuum conditions with an accelerating voltage of 8 kV and 10 kV
Weight in gram of the Citronella plant sample
at the working distance of 15 mm. The image was taken at the
2.2.3. Design of experiment (DOE) using response surface magnication of 1.78 KX.
methodology
Response surface methodology (RSM) is a mathematical and 2.2.7. HPLC analysis
statistical technique useful for modeling and analyzing the effect High performance liquid chromatography (HPLC) is an ef-
of several quantitative variables in the response of interest cient technique for separation, identication and quantication
(Montgomery, 2005). This methodology saves time and effort by of each component in a liquid mixture. SUPELCOGEL Ca HPLC
reducing the number of experimental runs and gives optimized and column provides an excellent separation of monosaccharide and
statistically signicant results. Hence it is widely used in various sugar alcohols. The hydrosol liquid obtained after hydro-distillation
elds of engineering and applied sciences to optimize the pro- was ltered using 0.2 m nylon membrane lter paper for com-
cess variables (Galadima et al., 2012; Peng et al., 2012; Betiku and position analysis using HPLC (PerkinElmer Series 200) equipped
Adepoju, 2013). The most common experimental design used are with RI detector, vacuum degasser and network chromatogra-
Central Composite Design (CCD), Box-Behnken (BBD) and Doehlert phy interface (NCI) 900. SUPECOGEL Ca column with dimensions
design. of 300 mm length and 7.8 mm inner diameter connected to Ca*
Box-Behnken design using Design-expert software version guard column purchased from Sigma-Aldrich was used. HPLC was
7.0.0 (Stat-Ease, inc., Minneapolis) was employed to evaluate the operated at 80 C maintained using a column oven for the run
independent and interaction effects of process parameters such as time of 45 min. Milli-Q water was used as mobile phase at a ow
distillation time, plant parts and solute to a solvent ratio towards rate of 0.5 ml min1 . The peaks obtained were identied from the
citronella oil yield. Preliminary experiments were performed to retention time of the standard sample and quantied using built
 R. Timung et al. / Industrial Crops and Products 94 (2016) 178188 181

Fig. 5. 3D response plot to show the interaction effects of process variables on% oil yield: (a) Distillation Time and solute-solvent ratio; (b) Distillation Time and Plant part;
(c) Plant part and solute-solvent ratio.

Fig. 6. Comparison of the actual and predicted percentage oil yield.

Fig. 7. FTIR spectra of hydro-distilled citronella oil.

calibration plots standardized with at least 5 points of different

concentrations. pneumonia and Escherichia coli by agar well diffusion method as
reported by Naik et al. (2010) with minor modications. All the bac-
2.2.8. Preparation of bacterial strains and oil terial strains were revived on respective culture medium at their
Anti-bacterial nature of citronella oil was evaluated against ambient temperature, and glycerol stocks of the same were pre-
three gram-positive bacterial strains namely Bacillus subtilis, pared and stored at 80 C for further use. The oil sample was
Staphylococcus epidermidis and Staphylococcus aureus, and three ltered and sterilized through 0.2 m lter and mixed at different
gram-negative strains namely Pseudomonas aeruginosa, Klebsiella ratio (10%, 20%, 30%, 40% and 50% v.v1 ) with previously sterilized
182 R. Timung et al. / Industrial Crops and Products 94 (2016) 178188

Table 1
Effect of moisture content on oil yield.

State of plant Initial wt. of MoistureContent Final wt. of plant Time Vol. of oil %Yield (vw1 )
(Whole aerial plant (g) (wt%) (dry wt.) (g) (min) (ml) Dry wt. basis
part)

Fresh 30 76.01 0.14 7.20 180 0.175 2.43

Shade Dried 30 9.43 0.11 27.16 180 0.575 2.12

Table 2
Factors and coded factors levels for Box-Behnken design.

Factor Low Medium High

1 0 1

A Distillation Time (min) 60 120 180

B Solute/Solvent (g/500 ml) 10 20 30
C Plant part Stems with Leaf sheaths (S) Whole aerial parts (L + S) Leaves (L)

tween 80 for the assay. Single colony of each bacterial strain was From the aforementioned discussion, it was clear that the drying
inoculated into the particular culture broth medium and grown at process affected the active ingredients of essential oil. Therefore,
respective temperature for 2448 h to get the fully grown bacte- in the present work, plant materials were shade dried in a semi
ria. 300 l of the fully grown active bacteria were uniformly spread dark room. Moreover, considering the traditional practice of distil-
over the Luria bertani agar, nutrient agar and tryptone soya agar lation of citronella plant and non-availability of the fresh material
plates by using sterilized spreaders. All plates were left for 1 h to on regular basis, shade dried citronella plant parts were used in
dry. Later a single well per plate was made with the help of a cork optimization studies by the hydro-distillation process.
borer at the centre of all plates. Then 50 l of each concentration of
citronella oil-tween 80 mixtures were lled aseptically into every 3.2. Preliminary experimental results
plate, where only tween 80 was lled in the wells of control plates.
All the plates were left for 1 h to get the oil sample diffuse prop- The preliminary experiments were carried out to evaluate the
erly and later they were kept in the incubator for 2448 h in their effect of different parameters such as parts of citronella plant,
respective temperatures. Finally the zone of inhibition by the oil solute-solvent ratio and distillation time on oil yield. The oil yield
for different bacterial strains was measured by using a geometric increased with an increase in extraction time, and almost all the
scale. oil was extracted within 2 h of the extraction period (Fig. 2). Fur-
ther increase in extraction time up to 3 h increases the oil yield by
only 15% of the total extracted oil and beyond 3 h deteriorates the
oil quality (Ahmed, 2005). Comparison of the oil yield from differ-
3. Results and discussion ent parts of the plants revealed that leaves contained about 40%
and 17% more essential oil than the stems and whole aerial parts
3.1. Effect of moisture content on oil yield respectively. On the other hand, whole aerial parts contain approx-
imately 21% more essential oil than the stems part. The volume of
The moisture content of fresh and dried sample (whole aerial water used in a hydro-distillation process showed an insignicant
citronella plant) was found to be 76.01 0.14% and 9.43 0.11% effect on the oil yield within the studied range (Fig. 3). However,
respectively. On a dry weight basis, the fresh citronella plant sufcient quantity of water was maintained throughout the study
showed higher oil yield (2.43%) than a dried sample (2.12%) at 3 h to avoid insufcient evaporation of essential oil and burning of
distillation time (Table 1). These results revealed that oil yield is plant material. Similarly, excessive water was also avoided as more
greatly affected by the moisture content of the plant sample. During water volume may require more heat energy and time. The yield of
drying of the plant sample, volatile compounds might percolate to essential oil obtained from different parts of citronella plant such
the surface and evaporate along with water resulted in a decreased as leaves, stems with leaf sheaths, and whole aerial part at 180 min
oil content. Cassel and Vargas, (2006) obtained 0.78% of citronella was found to be 2.38%, 1.58%, and 1.92% respectively (Fig. 4).
oil yield for the distillation time of 4 h using dry sample and 0.94%
yield using the fresh sample. In the drying process, temperature is 3.3. Box behnken design (BBD)
the most important parameter to preserve the essential oil intact
in the plant materials. Drying methods signicantly affect the oil According to the preliminary experimental results, the matrix
content and composition. Jalal et al. (2009) evaluated the effect of was designed for the low and high level of each factor. The coded
drying on oil yield in Rosemary leaves and discovered that shade levels of the independent factors are given in Table 2. The design
dried sample (1.8%) showed higher oil content than the sun-dried matrix and their corresponding experimental and predicted yields
(1.63%) and 45 C oven dried sample (1.5%). Rocha et al. (2000) stud- are shown in Table 3.
ied the air-drying of Java Citronella at ve different temperatures
viz. 30, 40, 50, 60, 70 C and found that the best air-drying condition 3.3.1. Effect of factors
for Java Citronella plant was at 60 C, where the highest oil yield was The parameters affecting the oil yield can be determined from
obtained without affecting its quality. Kakaraparthi et al. (2014) the analysis of variance (Table 4 and Table 5). The large F-value and
observed that in the process of shade-drying and sun-drying, the corresponding low P-value in Table 4 depicted that the factor had
oil content in Palmarosa increased up to 79 days after harvesting a signicant effect on the response (oil yield), whereas the factors
and decreased thereafter. with P-value > 0.05 showed an insignicant effect on the response.
 R. Timung et al. / Industrial Crops and Products 94 (2016) 178188 183

Fig. 8. GC/MS chromatogram of hydro-distilled citronella oil.

Fig. 9. Electron micrograph of citronella plant; (a) before hydro-distillation and (b) after hydro-distillation.
184 R. Timung et al. / Industrial Crops and Products 94 (2016) 178188

Fig. 10. Comparison of HPLC chromatogram of hydrosol obtained after hydro-distillation from different parts of citronella plant.

Table 3
Box-Behnken design of the extraction process variables using coded factors and its corrresponding experimental and predicted responses.

Std. order Run order Distillation time(A) Solute/Solvent (B) Plant Part(C) Oil Yield (%)

Expt. data Pred. data

1 1 1.00 1.00 0.00 1.50 1.46

2 4 1.00 1.00 0.00 2.00 1.94
3 9 1.00 1.00 0.00 1.33 1.39
4 2 1.00 1.00 0.00 1.92 1.96
5 6 1.00 0.00 1.00 1.13 1.12
6 17 1.00 0.00 1.00 1.50 1.51
7 8 1.00 0.00 1.00 1.75 1.74
8 14 1.00 0.00 1.00 2.38 2.39
9 11 0.00 1.00 1.00 1.50 1.55
10 7 0.00 1.00 1.00 1.58 1.53
11 13 0.00 1.00 1.00 2.25 2.30
12 5 0.00 1.00 1.00 2.33 2.28
13 10 0.00 0.00 0.00 2.00 2.00
14 16 0.00 0.00 0.00 2.00 2.00
15 3 0.00 0.00 0.00 2.13 2.00
16 15 0.00 0.00 0.00 1.88 2.00
17 12 0.00 0.00 0.00 2.00 2.00
 R. Timung et al. / Industrial Crops and Products 94 (2016) 178188 185

Table 4
Analysis of variances (ANOVA) for response surface quadratic model.

Source Sum of squares DF Mean square F-value P-value

Model 2.03 9 0.23 29.58 <0.0001

A: Distillation time 0.55 1 0.55 77.70 <0.0001
B: Solute/Solvent 1.012E-003 1 1.012E-003 0.13 0.7261
C: Plant Part 1.13 1 1.13 147.73 <0.0001
AB 2.025E-003 1 2.025E-003 0.27 0.6220
AC 0.017 1 0.017 2.22 0.1799
BC 0.000 1 0.000 0.000 1.0000
A2 0.31 1 0.31 40.23 0.0004
B2 8.432E-003 1 8.432E-003 1.11 0.3276
C2 7.516E-003 1 7.516E-003 0.99 0.3536
Residual 0.053 7 7.615E-003
Lack of t 0.022 3 7.342E-003 0.94 0.5006
Pure error 0.031 4 7.820E-003
Cor Total 2.08 16

Table 5
Signicance of the response quadratic model.

Factor Coefcient DF Std. error 95% CI low 95% CI high VIF

Intercept 2.00 1 0.039 1.91 2.09

A: Distillation time 0.26 1 0.031 0.19 0.33 1.00
B: solute/solvent 0.011 1 0.031 0.084 0.062 1.00
C: Plant part 0.38 1 0.031 0.30 0.45 1.00
AB 0.023 1 0.044 0.081 0.13 1.00
AC 0.065 1 0.044 0.038 0.17 1.00
BC 0.000 1 0.044 0.10 0.10 1.00

Similarly, the coefcient of the factor with 95% CI low and 95% CI 3.3.2. Model equation
high both having positive values as represented in Table 5 showed The model equation developed for the extraction of essential
signicant effect on the response. It can be clearly observed that A oil from the citronella plant using a hydro-distillation technique in
(distillation time), C (plant part) has high individual effect on the terms of both experimental and coded values are given in Eqs. (4)
response, and A2 with P-value 0.0004 has a signicant effect on the and (5) respectively.
response indicating the oil yield was increasing with an increase Y = 2.00+0.26A0.01B+0.38C+0.02AB+0.07AC+0.00BC0.27A2 0.04B2 0.04C2 (4)
of the distillation time. The other factors and their interactions Y = 2.00+0.26A0.01B+0.38C+0.02AB+0.07AC+0.00BC0.27A2 0.05B2 0.04C2 (5)
showed an insignicant effect on the oil yield.
The highest and lowest oil yield from every part was obtained at 3.3.3. Validation of the model
3 h and 1 h distillation time respectively. Experimentally obtained The analysis of variance showed that the experimental data was
highest and lowest yields for the leaves sample were 2.38% and best tted to the second-order polynomial model. The signicance
1.75% respectively. For whole aerial parts, the experimentally of the model was checked by the determination coefcient (R2 ). For
obtained highest yield was 2% and lowest was 1.33%. Likewise, a good t of a model, R2 value should be a miniMum of 0.80 (Guan
for stems with leaf sheaths sample, the highest yield was 1.58% and Yao, 2008; Joglekar and May, 1987). The statistical parame-
whereas the lowest yield was 1.13%. Blank et al. (2007) reported ters of the model such as R2 = 0.9744, standard deviation = 0.087,
that dry leaves harvested at different day time during the winter F-value of 29.58 and P-value <0.0001 showed a highly signicant
season showed oil yield as 2.332.67%, whereas leaves harvested model. The true coefcient (2.00) of the model intercepts with pos-
during spring season yielded up to 3.55% oil for 3 h of a hydro- itive values of both 95% CI low (1.91) and 95% CI high (2.09) also
distillation period. The plant material used in this study was also depicted the model as highly signicant. The P-value of 0.5006 for
collected during the winter season and gave almost similar oil yield Lack of t indicated that, lack of t, which was not signicant rel-
percentage (i.e. 2.38%). Wany et al. (2014) reported yield of 1% by ative to the pure error revealing a good tted model. Hence, the
applying hydro-distillation to fresh sample and 0.7% by implement- model can be used to navigate the design space. The comparison
ing steam distillation to the dry sample for 810 h. of the actual and predicted oil yield as represented in Fig. 6 also
Fig. 5 represents the 3D response surface plots for interaction exhibited a good tness of the data to the model.
effect of independent factors on the oil yield. The effect of Mutual
interactions of individual factors on oil yield can be observed from 3.4. FTIR analysis
the curvature nature of surface plots. There was a minuscule inter-
action between plant part and distillation time (AC), as well as The qualitative analysis of different organic compounds can be
between distillation time and solute to the solvent ratio (AB), but ascertained from the characteristics vibrational bands appeared in
no Mutual interaction between the plant part and solute to the sol- the infrared spectral region at a particular frequency inuenced
vent ratio (BC), which can be observed from the response plots as by specic functional groups. The percentage transmittance corre-
well as from Tables 4 and 5. sponding with the wave number is deduced in the attenuated total
reectance IR spectra as shown in Fig. 7.
There was an intense broad peak within the range of
36003200 cm1 particularly at 3365.78 cm1 corresponding to
186 R. Timung et al. / Industrial Crops and Products 94 (2016) 178188

the polymeric hydroxyl (OH) group. Another intense and bifur- the extracted oil from leaves and whole aerial parts with 94.94% and
cated peak within the range of 29352915 cm1 which corresponds 89.15% respectively, indicated that the extracted oil was of superior
to the C H methyl and methylene asymmetric stretch, mostly quality. Differences in the composition of oil may be driven by vari-
aliphatic alkyl groups were observed. The medium peak at ous factors such as variation in distillation time, genetical makeup,
2719.63 cm1 validated a terminal aldehydic C H stretch of car- geographical locations and environmental conditions (Sarma et al.,
bonyl compound. Another distinct and sharp peak within the 2001; Sarma, 2002; Cannon et al., 2013). The overall compositional
range of 17501705 cm1 signied aldo, keto, estero and or distribution of extracted Java Citronella oil identied using GC/MS
acido (C O) stretch. A strong and relatively narrow absorp- is given in Table 6.
tion peak at 1668.43 cm1 contributed to olenic unsaturated
C C group. The sharp and strong peaks were observed for 3.6. SEM analysis
methylene C H (14851445 cm1 ), methyl C H symmetric
bend (13801371 cm1 ) and aryl-O H stretch (12701230 cm1 ). SEM analysis of the sample showed a distinguishable change
Moreover, C O bend (11401050 cm1 ), simple OH stretch in the structure of the citronella plant material after hydro-
(12001000 cm1 ) and CH CH trans-unsaturated (910860 cm1 ) distillation. Fig. 9(a) represents the micrograph of untreated plant
functional groups with medium peaks were also observed. A material. The smooth and linear surfaces of the sample indicated
medium peak depicting di or tri-substituted alkenes (C H) stretch that the cells of the plant material were intact. In the process of
was detected at 825.53 cm1 . Minor vibrations within the range of hydro-distillation when the plant material was subjected to heat
750660 cm1 attributed to the presence of aromatic, vinyl C H the cellular content of the plant tissues gets distorted. Extensive
group. These results were in absolute accord with the previous thermal stress on the oil glands boosted the localized high pressure
work carried out by Wany et al. (2014). to cause rapid expansion of oil glands. When it exceeded the critical
coefcient of expansion, the oil glands burst and released the oil.
3.5. GC/MS analysis The severe mechanical strain induced due to rapid disintegration
and vehement vaporization of water caused fast dehydration lead-
The compositional analysis of hydro-distilled citronella oil was ing the plant cells, including the oil glands to collapse or crumble
studied based on the peaks obtained in the GC chromatogram. The promptly. Ferhat et al. (2006) also observed similar kind of defor-
oil extracted from different parts (leaves, stems and whole aerial) mation in the SEM analysis of hydro-distilled residue of orange peel.
obtained at the highest yield were analyzed using GC/MS. Total The signicant deformation on the external surface of the citronella
of 50 compounds shown at a different retention time were gener- plant material after hydro-distillation can be observed as shown in
ated according to the electronic integration of a chromatogram plot. Fig. 9(b).
Out of these 50, a total of 13 distinct peaks can be observed in the
chromatogram of leaves sample (Fig. 8) constituting 99.21% of the 3.7. HPLC analysis
total composition which were identied as citronellal, citronellol,
geraniol, 3-hydroxy-5-isopropyl-2-methylbenzo-1-4-quinone, cit- The hydrosol liquids obtained after distillation of aromatic
ronellyl propanoate, limonene, -elemene, patchoulane, linalool, plants are not generally analyzed. The signicance of such liquid
germacrene d, -cadinene, germacrene A and germacrene d-4-ol is therefore, not efciently determined. Therefore, in the present
respectively. Other 37 trace compounds constituting remaining study, HPLC analysis of the hydrosol obtained at highest oil yield
0.79% were not identied. Whereas, whole aerial parts and stems were performed to determine their composition. The hydrosol
parts had these 13 compounds with 95.67% and 81.20% of the obtained after hydro-distillation of citronella plant contains a
total composition of the oil respectively. This implied that the minuscule amount of pentose sugar and sugar alcohols such as
composition of other trace compounds of oil were dispersed at a xylose, adonitol, erythritol and mannitol (Fig. 10). The composi-
higher percentage in the whole aerial parts and more distributed tion of sugars was found to be different from the parts of plant
in the stems part. The rst major peak obtained was citronellal at materials. The leaves part contained 0.07 mg ml1 of erythritol and
26.71 min suggesting citronellal as the most volatile compound in 0.03 mg ml1 of mannitol. Hydrosol of whole aerial parts consists of
citronella oil and the last major visible peak was germacrene d-4- 0.09 mg ml1 xylose, 0.63 mg ml1 adonitol, 0.03 mg ml1 erythri-
ol (39.34 min). Silva et al. (2011) and Manaf et al. (2013), in their tol and 0.02 mg ml1 mannitol. Stems of citronella plant showed
study also reported citronellal as the rst major peak. Higher area high quantity of erythritol (4.49 mg ml1 ) relative to leaves part and
percentage of the peak signied higher percentage of the corre- whole aerial parts. These sugar alcohols are of great importance for
sponding compound in the oil. The chromatogram plot obtained human health and medication. Erythritol is a useful component for
from leaves part revealed a high concentration of commercially food industry and oral health. Mannitol has diuretic effect, which
important major three compounds i.e. citronellal (55.23%), geran- is useful for the kidney. This composition analysis of hydrosol evi-
iol (26.29%) and citronellol (13.41%) accounting for 94.94% of total denced its high importance and projected extensive scope for its
constituents. Whereas, whole aerial parts had 89.15% of these three further utilization.
compounds and the stems part constituted only 69.84%. This com-
positional distribution analysis indicated that the oil obtained from 3.8. Physical properties of citronella oil
leaves part was of good quality compared to whole aerial parts
and stems parts. Cassel and Vargas, (2006) during their study on The physical properties of the extracted citronella oil were eval-
essential oil extraction from Cymbopogon winterianus using steam uated for the oil obtained from the leaves sample at highest oil yield.
distillation reported the composition as citronellal (35.90%), cit- The colorless to pale yellow colored citronella oil has a character-
ronellol (5.2%) and geraniol (20.90%) quantifying 62% of the total istic citronellal aroma. The oil obtained from a particular citronella
oil composition. Beneti et al. (2011) obtained 71.32% constituting cultivar used to have its unique constituents strongly inuenc-
citronellal (40.23%), citronellol (13.39%) and geraniol (17.70%) in ing the overall weight of the oil and degrees at which they tend
their study on fractionation of citronella (Cymbopogon winterianus) to refract light when a beam of light passes through it. These
essential oil. Similarly other researchers also reported these three characteristics can be determined by evaluating the refractive
major constituents of citronella oil in the range of 6075% (Songkro index, optical rotation and specic gravity, which help to evalu-
et al., 2012; Li et al., 2013; Pinheiro et al., 2013). In this study, the ate, whether the citronella oil has been adulterated. The refractive
composition of these three commercially important compounds in index of the oil was measured at 20 C using a Refractometer
 R. Timung et al. / Industrial Crops and Products 94 (2016) 178188 187

Table 6
Chemical composition of citronella oil identied using GC/MS.

Retention time (min) Compound name Abundance%

Leaves Whole Aerial Stems

26.71 Citronellal 55.24 48.39 54.64

28.91 Citronellol 13.41 24.86 5.03
29.82 Geraniol 26.29 15.89 10.16
30.28 3-Hydroxy-5-isopropyl-2-methylbenzo-1,4-quinone 0.05 0.11 0.56
32.28 Citronellylpropanoate 0.64 3.81 1.05
33.09 Limonene 0.51 0.10 1.15
34.14 -Elemene 0.37 0.11 1.11
35.35 Patchoulane 0.08 0.08 1.03
35.59 Linalool 0.21 0.09 1.09
36.92 Germacrene D 0.15 0.11 0.97
37.55 -Cadinene 0.22 0.10 0.95
38.35 Germacrene A 0.82 0.09 2.58
39.34 Germacrene d-4-ol 1.22 1.93 0.88
Percentage of total compounds identied 99.21 95.67 81.2
Other trace compounds 0.79 4.33 18.8

Table 7
Physical properties of the citronella oil.

Property Experimental value Literature data

New Directions Aromatics Inc. (2014) Natural resourcing specialist in cosmeceutical ingredients (2010)

Appearance Pale yellow to colorless Oily liquid Clear light yellow to brownish liquid. Clear yellow to pale brown
Odor Lemon characteristic Strong lemon-like odor. Characteristic of lemon, herbaceous
Specic gravity 0.9005 @ 25 C 0.8800.922 @ 20 C 08780.895
Refractive index 1.473@ 20 C 1.4661.474 @ 20 C 1.4661.475 @ 25 C
Optical rotation 4.0 @ 25 C 5 to 0
Dynamic Viscosity 4.053 cP@ 25 C

Table 8
Zone of Inhibition of the citronella oil.

Bacterial Strains Zone of Inhibition at different oil concentrations (mm)

10%v.v1 20%v.v1 30%v.v1 40% v.v1 50% v.v1

Gram +ve
Bacillus subtilis 10 1.41 11 11.5 2.12 19 1.41 21
Staphylococcus epidermidis 10.5 0.71 12.5 0.70 13 18 2.82 23 1.41
Staphylococcus aureus 0 0 5 5 6
Gram ve
Pseudomonas aeruginosa 4 1.41 9 10.5 0.70 13.5 2.12 15
Klebsiella pneumonia 0 6 11 17 23
Escherichia coli 0 4 7.5 0.70 10.5 0.70 12

having temperature indicator and intensity controller (Advanced and antioxidant properties of 17 commercial essential oils also
Research Instruments Company, New Delhi). Optical rotation of observed the inhibition of growth of different pathogenic bacteria
the oil was measured by taking 10 ml of 1% citronella oil in hex- by citronella oil. The susceptibility of the oil towards Bacillus sub-
ane, in a 100 mm glass sample cell at 25 C and 589 nm, using tilis and Staphylococcus epidermidis was high as compared to other
an automatic polarimeter (Rudolph Research Analytical, Auto pol selected strains, as the zone of inhibition was more than 10 mm
I). Specic gravity was determined using a specic gravity bottle even at lowest concentration (10% v.v1 ) of the oil considered. The
(accuracy as per I.S. 5717, Borosil). The optical rotation value (4.0 increased in anti-bacterial activity was observed with an increase
@ 25 C) conrmed that the extracted citronella oil was optically oil concentrations in all the strains irrespective of being gram pos-
active and had a levorotary behavior. The dynamic viscosity of the itive or gram negative. Similar pattern was observed by Naik et al.
oil was found to be 4.053 Centipoise (cP) at 25 C. The specic grav- (2010) in their study on the anti-bacterial activity of lemongrass oil.
ity and refractive index of the extracted citronella oil were 0.9005 On the other hand, Victoria et al. (2012) reported that aldehyde (e.g.
and 1.473 respectively and were comparable with the literature citronellal) and alcohol (e.g. citronellol) have same effects on both
data (Natural resourcing specialist in cosmeceutical ingredients, gram-positive and gram-negative bacteria. MaxiMum zone of inhi-
2010; New Directions Aromatics Inc., 2014) as given in Table 7. bition was around 23 mm, observed for Staphylococcus epidermidis
and Klebsiella pneumonia at highest concentration of citronella oil
(50% v.v1 ). The anti-bacterial activity of citronella oil is attributed
3.9. Anti-bacterial activity of citronella oil
due to the major components in the oil viz. citronellal, citronellol
and geraniol as reported by Lertsatitthanakorn et al. (2008). Hence,
Citronella oil was found effective against all the six tested bac-
citronella oil with selection of specic concentration could act as
terial strains which could be observed from the developed zone
anti-bacterial supplement for the treatment of various bacterial
of inhibition as presented in Table 8. Teixeira et al. (2013) in
infections.
their study on the chemical composition, antibacterial activity
188 R. Timung et al. / Industrial Crops and Products 94 (2016) 178188

4. Conclusions Guan, X., Yao, H., 2008. Optimization of Viscozyme L-assisted extraction of oat
bran protein using response surface methodology. Food Chem. 106, 345351.
Jalal, K., Rahmat, M., Mohammad, F.T., Himan, N., 2009. Inuence of drying
The results from the study revealed that the essential oil methods, extraction time, and organ type on essential oil content of rosemary
form the citronella plant can be easily extracted using hydro- (Rosmarinus ofcinalis L.). Nat. Sci. 7, 4244.
distillation technique. High regression coefcients of the response Joglekar, A.M., May, A.T., 1987. Product excellence through design of experiments.
Cereal Foods World 32, 857868.Joy, P.P., Thomas, J., Mathew, S., Joseph Jose,
(R2 = 0.9744) with Model P-value 0.05 showed that the developed G., 2001. Journal of Aromatic plants. In: Bose, T.K., Kabir, J., Das, P., Joy, P.P.
model tted well to the experimental data. Hence, Box-Behnken (Eds.), Tropical Horticulture, vol. 2. NayaProkash, Calcutta, pp. 633733.
design of response surface methodologies can be applied to study Kakaraparthi, P.S., Rajput, D.K., Arigari, N.K., Nivedita, 2014. Inuence of different
methods and time of post-harvest drying on the essential oil content and
the citronella oil extraction process using hydro-distillation. The
composition in Palmarosa (Cymbopogon Martinii (Roxb.) Wats Var. Motia
leaves of a citronella plant contain more essential oil and compar- Burk.). Indian J. Appl. Res. 4, 59.
atively better quality than stems or whole aerial parts. Therefore, Lertsatitthanakorn, P., Taweechaisupapong, S., Aromdee Khunkitti, W., 2008.
Antibacterial activity of citronella oil solid lipid particles in oleogel against
citronella oil should be extracted from leaf part in order to obtain
Propionibacterium acnes and its chemical stability. Int. J. Essent. Oil Ther. 2,
high quantity and quality. The high percentage (about 95%) of com- 167171.
mercially important compounds such as citronellal, geraniol and Li, W.R., Shi, Q.S., Ouyang, Y.S., Chen, Y.B., Duan, S.S., 2013. Antifungal effects of
citronellol obtained in the extracted citronella oil from leaf part, Citronella oil against aspergillus niger ATCC 16404. Appl. Microbiol.
Biotechnol. 97, 74837492.
and around 89% from whole aerial parts indicate a superior quality Luthria, D.L., Natarajan, S., 2009. Inuence of sample preparation on the assay of
of citronella oil suggesting that the geographical locations and envi- isoavones. J. Med. Plants Res. 75, 704710.
ronmental conditions of Karbi Anglong, Assam, North East India Manaf, M.A., Mustapa, A.N., Mustapa, K., 2013. Supercritical uid extraction of
Citronella oil from Cymbopogon nardus and its optimization. IEEE Bus. Eng. Ind.
is exceptionally favorable to the growth of citronella plant. The Appl. Colloq. (BEIAC), 7378, http://dx.doi.org/10.1109/BEIAC.2013.6560229.
positive results of inhibition by the oil revealed that citronella oil Montgomery, D.C., 2005. Design and Analysis of Experiments: Response Surface
possesses a promising anti-bacterial activity and could be utilized Method and Designs. John Wiley and Sons, Inc, New Jersey.
Muazu, K., Mohammed-Dabo, I.A., Waziri, S.M., 2012. Development of
for various bacterial infections. Hydrosol can be used to produce mathematical model for the prediction of essential oil extraction from
sugar alcohols such as erythritol, mannitol, etc., which can be fur- eucalyptus citriodora leave. J. Basic.Appl. Sci. Res. 2, 22982306.
ther utilized in nutraceutical and pharmaceutical industries. Naik, M.I., Fomda, B.A., Jaykumar, E., Bhat, J.A., 2010. Antibacterial activity of
lemongrass (Cymbopogon citratus) oil agaist some selected pathogenic
bacterias. Asian Pacic J. Trop. Med., 535538.
Acknowledgements 2010. Material Safety Data Sheet of Citronella Essential Oil, Physical and Chemical
Properties. 341 Christian Street, Oxford, CT 06478, USA 01/02/2010, http://
www.naturalsourcing.com/msds/MSDS Essential Oil Citronella.pdf. (accessed
The authors wish to thank the Central Instrument Facilities of
25.05.2015).
Indian Institute of Technology Guwahati for providing facilities for Material safety data sheet of Citronella (Java) Essential Oil, 6781-Columbus Road,
the SEM analysis. Mississauga, Canada http://www.newdirectionsaromatics.com/msds/MSDS
Citronella%28Java%29EssentialOil.pdf (accessed 25.05.2015).
Peng, T.Y., Don, M.M., Tahrel, M.A., 2012. Optimization and kinetics studies on the
References extraction of essential oil from Zingiber cassumunar. J. Phys. Sci. 23, 6582.
Pinheiro, P.F., Queiroz, V.T., Rondelli, V.M., Costa, A.V., Marcelino, T.P., Pratissoli, D.,
APG III, 2009. An update of the Angiosperm Phylogeny Group classication for the 2013. Insecticidal activity of Citronella grass essential oil on Frankliniella
orders and families of owering plants: APG III. Bot. J. Linn. Soc. 161, 105121. schultzei and Myzus persicae. Sci. Agrotecnol. Lavras 37, 138144.
Ahmed, M., 2005. Hand Book on Medicinal and Aromatic Plants (Package of Rocha, S.F.R., Ming, L.C., Marques, M.O.M., 2000. Inuence of ve drying
Practices), 1st ed. North Eastern Development Finance Corporation Ltd. temperatures on the yield and composition of essential oil of Citronella
(NEDFi), Guwahati, Assam, pp. 1318. (Cymbopogon winterianus Jowitt). J. Med. Plants 3, 7378.
Almeida, R.N., Navarro, D.S., Barbosa-Filho, J.M., 2001. Plants with central analgesic Rocha, R.P., Melo, E.C., Radunz, L.L., 2011. Inuence of drying process on the quality
activity. Phytomedicine 8, 310322. of medicinal plants: a review. J. Med. Plants Res. 5, 70767084.
Almeida, R.N., Motta, S.C., Faturi, C.B., Catallani, B., Leite, J.R., 2004. Anxiolytic like Salguero, C.P., 2003. A Thai Herbal: Traditional Recipes for Health and Harmony,
effect of rose oil inhalation on the levated plus maze test in rats. Pharmacol. 1st ed. Findhorn Press, Findhorn Forres, Scotland.
Biochem. Behav. 77, 361364. Sarma, T.C., Sharma, R.K., Adhikari, R.K., Saha, B.N., 2001. Effect of altitude and age
Beneti, S.C., Rosset, E., Corazza, M.L., Frizzo, C.D., Luccio, M.D., Oliveira, J.V., 2011. on herb yield, oil and its major constituents of Java Citronella (Cymbopogon
Fractionation of Citronella (Cymbopogon winterianus) essential oil and winterianus Jowitt.). J. Essent. Oil Bear. Plants 4, 7781.
concentrated orange oil phase by batch vacuum distillation. J. Food Eng. 102, Sarma, T.C., 2002. Variation in oil and its major constituents due to season and
348354. stage of the crop in Java Citronella (Cymbopogon winterianus Jowitt.). J. Spices
Betiku, E., Adepoju, T.F., 2013. Sorrel (Hibiscus sabdariffa) seed oil extraction Aromat. Crops 11, 97100.
optimization and quality characterization. Am. Chem. Sci. J. 3, 449458. Silva, C.F., Moura, F., Mendes, M.F., Pessoa, F.L.P., 2011. Extraction of Citronella
Blank, A.F., Costa, A.G., Arrigoni-Blank, M.F., Cavalcanti, S.C.H., Alves, P.B., Innecco, (Cymbopogon nardus) essential oil using supercritical CO2 : experimental data
R., Ehlert, P.A.D., Sousa, I.F., 2007. Inuence of season, harvest time and drying and mathematical modelling. J. Chem. Eng. 28, 343350.
on Java Citronella (Cymbopogon winterianus Jowitt) volatile oil. Braz. J. Songkro, S., Hayook, N., Jaisawang, J., Maneenuan, D., Chuchome, T., Kaewnoppara,
Pharmacogn. 17, 557564. N., 2012. Investigation of inclusion complexes of Citronella oil, Citronellal and
CEN/TC 335., 2004. Solid Biofuels Determination of moisture content Oven dry Citronellol with b-cyclodextrin for mosquito repellent. J. Inclusion Phenom.
method Part 3: Moisture in general analysis sample. Macrocyclic Chem. 72, 339355.
Cannon, J.B., Cantrell, C.L., Astatkie, T., Zheljazkov, V.D., 2013. Modication of yield Teixeira, B., Marques, A., Ramos, C., Neng, N.R., Nogueira, J.M.F., Saraiva, J.A., Nunes,
and composition of essential oils by distillation time. Ind. Crops Prod. 41, M.L., 2013. Chemical composition and antibacterial and antioxidant properties
214220. of commercial essential oils. Ind. Crops Prod. 43, 587595.
Cassel, E., Vargas, R.M.F., 2006. Experiments and modeling of Cymbopogon Victoria, F.N., Radatz, C.S., Sachini, M., Jacob, R.G., Alves, D., Savegnago, L., Perin, G.,
winterianus essential oil extraction by steam distillation. J. Mex. Chem. Soc. 50, Motta, A.S., Silva, W.P., Lenardao, E.J., 2012. Further analysis of the
126129. antimicrobial activity of -phenylseleno citronellal and -phenylseleno
Chanthai, S., Prachakoli, S., Ruangviriyachai, C., 2012. Inuence of extraction citronellol. Food Control 23, 9599.
methodologies on the analysis of ve major volatile aromatic compounds of Wany, A., Jha, S., Nigam, V.K., Pandey, D.M., 2013. Chemical analysis and
Citronella grass (Cymbopogon nardus) and lemongrass (Cymbopogon citratus) therapeutic uses of Citronella oil from Cymbopogon winterianus: a short
grown in Thailand. J. Assoc. Anal. Communities Int. 95, 763772. review. Int. J. Adv. Res. 1, 504552.
Ferhat, M.A., Meklati, B.Y., Smadja, J., Chemat, F., 2006. An improved microwave Wany, A., Kumar, A., Nallapeta, S., Jha, S., Nigam, V.K., Pandey, D.M., 2014.
Clevenger apparatus for distillation of essential oils from orange peel. J. Extraction and characterization of essential oil components based on Geraniol
Chromatogr. A 1112, 121126. and Citronellol from Java Citronella (Cymbopogon winterianus Jowitt). Plant
Galadima, M.S., Ahmed, A.S., Olawale, A.S., Bugaje, I.M., 2012. Optimization of Growth Regul. 73, 133145.
steam distillation of essential oil of Eucalyptus tereticornis by response surface Wijesekara, R.O.B., 1973. The chemical composition and analysis of Citronella oil. J.
methodology. Niger. J. Basic Appl. Sci. 20, 368372. Nat. Sci. Counc. Srilanka 1, 6781.