A publication of

CHEMICAL ENGINEERING TRANSACTIONS

The Italian Association
VOL. 35, 2013 of Chemical Engineering
www.aidic.it/cet
Guest Editors: Petar Varbanov, Jiří Klemeš, Panos Seferlis, Athanasios I. Papadopoulos, Spyros Voutetakis
Copyright © 2013, AIDIC Servizi S.r.l.,
ISBN 978-88-95608-26-6; ISSN 1974-9791 DOI: 10.3303/CET1335180

Fractionation and Characterization of Insolubles Formed in

Palm Oil Biodiesel
Vladimir Plataa,*, Magda Serranoa, Karoll Tiriaa, Arnold R. Romero Bohórquezb,
Viatcheslav Kafarova, Paola Gauthier-Maradeic, Edgar Castillod
a
Research Center for Sustainable Development in Industry and Energy, Industrial University of Santander, Carrera 27
calle 9, Bucaramanga, Colombia
b
Organic and Biomolecular Chemistry Laboratory, Industrial University of Santander, Carrera 27 calle 9, Bucaramanga,
Colombia
c
INTERFASE, Industrial University of Santander, Carrera 27 calle 9, Bucaramanga
d
Colombian Petroleum Institute ICP-Ecopetrol, Piedecuesta, Colombia
vladimirplata@gmail.com

During the last decade, biodiesel has emerged as a promising alternative to diesel. However, the
sedimentation of some insoluble material in pure biodiesel has become a concern for its producers. This
material easily drops out of solution, settles on the bottom of biodiesel storage tanks and forms deposits
on vehicle fuel filters. In this work, the insoluble material that forms in Colombian palm oil biodiesel was
characterized for the first time. Non-expensive and easy to use Thin Layer Chromatography and Fourier
Transform Infrared Spectroscopy were used as analytical techniques. The insoluble material was found
not to be composed by free steryl glucosides exclusively. Acylated steryl glucosides appeared to be part of
the insoluble material as well. This finding should lead to rethink the strategies that have been developed
to prevent the formation of haze in biodiesel so far, as they have been mainly focused on the removal of
free steryl glucosides from biodiesel.

1. Introduction
During the last decade, biodiesel has emerged as a promising alternative to traditional diesel. It offers
many important advantages such as higher cetane number, lower emissions of most regulated species
and biodegradability (Plata et al. 2010). In Colombia, the local government has implemented a Biofuel
National Program since 2008. This program was designed in order to develop the agricultural sector,
generate permanents job, improve the air quality and replace illicit crops. These facts have made
Colombia one of the leaders of biofuel production in Latin America (Janssen and Rutz, 2011). A mandatory
B10 blend is currently used in almost all the country, and the local government has planned to increase the
percentage of blending in the next years. However, the sedimentation of insoluble material in pure
biodiesel has become a concern. This material can appear at temperature above the biodiesel cloud point
or even if biodiesel stands at room temperature for extended periods. It settles on the bottom of biodiesel
storage tanks and forms deposits on diesel engine injectors (Plata et al. 2012).
Research made to identify the nature of haze has revealed that it may be a mixture of several compounds.
Tang et al. (2008a) detected free and acilated sterols, monoglycerides (MG) and free steryl glucosides
(FSG) in the insoluble material isolated from cotton seed oil biodiesel. Lin et al. (2011) identified soaps in
the sediments collected from the bottom of storage tanks filled with canola oil biodiesel. Research has also
revealed that the composition of the haze may depend on the feedstock oil and technologies used to
process it (refining and transesterification). Tang et al. (2008b) examined the insoluble material isolated
from palm oil biodiesel and registered only signals attributable to MG. On the contrary, Van Hoed et al.
(2008) found that FSG were the major components of the cake collected after the filtration of biodiesel
between storage tanks in palm oil biodiesel facilities. Similar results were reported by Bondioli et al. [6].

Please cite this article as: Plata V., Serrano M., Tiria K., Romero Bohorquez A.R., Kafarov V., Gauthier-Maradei P., Castillo E., 2013,
Fractionation and characterization of insolubles formed in palm oil biodiesel, Chemical Engineering Transactions, 35, 1081-1086
DOI:10.3303/CET1335180
1081
Until now, most of the research has focused on the characterization of haze isolated from biodiesel
produced in Europe and the USA. To the best of our knowledge, the insoluble material that forms in
Colombian palm oil biodiesel has not been characterized so far. In this regard, a sample of haze isolated
from biodiesel produced in Ecodiesel Colombia S.A. (one of the largest providers of palm oil biodiesel in
Colombia) was characterized in this paper. It was first purified by successive washes with hexane, then
fractionated by column chromatography and finally characterized. Thin layer chromatography (TLC) and
Fourier transform infrared spectroscopy (FTIR) were used as analytical techniques. As a result, the haze
was found to be mainly composed by MG. FSG were also identified. This finding should lead to rethink the
strategies that have been developed to prevent the formation of haze in biodiesel so far since they have
been mainly focused on the removal of FSG from biodiesel.

2. Materials and methods

2.1 Preparation
The haze was taken from turbid biodiesel stored in plastic containers for a month at about 20 °C. 1 g of
sample was washed with 10 ml of hexane in order to remove the residual biodiesel contained therein. The
slurry formed was stirred in a Heidolph Unimax 2010 shaker for 1 min at 300 rpm and then centrifuged for
15 min at 2,100 rpm and 25 ° C using a Hettich Universal 320R centrifuge. The supernatant was removed
after centrifugation and the solid extracted was placed in an oven at 80 °C for 30 min and then macerated.
This solid was denominated Hz.
2.2 Fractionation
Fractionation of the haze was performed by column chromatography using a 0.7 x 0.3 m glass column
packed with approximately 55 g of silica gel 60 (0.063-0.200 mm, Merck ref no. 1.07734.1000) previously
activated in a oven at 80 °C for 4 h. The silica gel was rinsed with three portions of hexane and the solvent
was drained off until it reached the top of the silica. 1 g of sample was dissolved in chloroform/methanol
(85:15 vol/vol), and silica gel was added to this solution. The slurry formed was rotoevaporated until
dryness, and the solid obtained was introduced onto the column. The column was then washed
successively with hexane/ethyl acetate, pure ethyl acetate, ethyl acetate/methanol and pure methanol. The
fractionation was monitored by TLC, and the fractions collected were rotoevaporated until dryness and
weighted.
2.3 Characterization
Two simple TLC evaluations were first performed on TLC silica gel 60 F254 plates purchased from
MERCK (ref. no. 1.05715.0001). Samples of the fractions collected and reference standards of FSG and
monopalmitin were applied on two different TLC plates. The plates were developed with hexane/ethyl
acetate (1:1 vol/vol) mixture in one evaluation and ethyl acetate/methanol (10:1 vol/vol) mixture in the
other evaluation. Characterization by FTIR was performed using a Bruker Tensor 27 spectrometer
equipped with an attenuated total reflectance accessory (with diamond crystal). All the spectra were
-1
obtained in the wave number range of 4,000-650 cm , using an average of 16 scans with a spectral
-1
resolution of 4 cm and air as the background. All chemical reagents used were analytical grade. The
standard of FSG was purchased from Matreya LLC. The standard of monopalmitin was purchased from
Sigma Aldrich Inc. Detection was carried out under iodine.

3. Results and discussion

3.1 Preparation and fractionation

A “washing and drying” yield, defined as the ratio between the weight of solid and the weight of unwashed
haze, higher than 15 wt% was not achieved. This result was consistent with that reported by Moreau et al
(2008), who analyzed samples of haze isolated from a variety of locations in the soybean oil biodiesel
supply chain. These authors found that most of the samples were mainly composed by fatty acid esters.
They also encountered that the amount of FSG ranged between 0 and 68 %w/w.
Two fractions were obtained during the fractionation. The first fraction, denominated F1Hz, was obtained
after washing the column successively with hexane/ethyl acetate (5:1, 1:1, 1:10 vol/vol), pure ethyl acetate
and ethyl acetate/methanol (10:1 vol/vol). The second fraction, denominated F2Hz, was obtained after
using ethyl acetate/methanol (10:1, 1:1 vol/vol) and pure methanol. The amount of solid collected was
0.7668 mg for F1Hz and 0.1225 mg for F2Hz.

1082
3.2 Characterization
Hz, F1Hz and the monopalmitin standard marked a single spot with a retention factor (Rf) of about 0.35
when the TLC plate was developed with the hexane/ethyl acetate mixture (see Figure1, number 1, 2 and 5
on the left TLC plate). On the contrary, F2Hz and FSG standard did not mark any spots. When the TLC
was developed with the ethyl acetate/ methanol mixture, the spots previously marked by both F1Hz and
monopalmitin standard increased their Rf to nearly 0.91 (number 1, 2 and 5 on the right TLC plate).
However, in this case, both F2Hz and FSG standard marked a single spot with Rf of about 0.63 (number 3,
4 and 6 on the same TLC plate). These results suggested that the haze was composed by a mixture of
MG, which is consistent with the report of Tang et al. (2008b), and also by FSG, which is consistent with
the finding of Bondioli et al. (2008) and Van Hoed et al. (2008). Monopalmitin was selected because
saturated monoglycerides have been reported as the main glyceride compounds capable of forming haze
in biodiesel (Farahani et al. 2011), and palm oil biodiesel is mainly composed by palmitic acid (about 45
%w/w (Prada et al. 2011)).
Figure 2 shows the FTIR spectrum obtained for the monopalmitin standard. A series of strong bands,
which arises from twisting and wagging bending of –CH2, was registered in the region usually referred as
-1
the “finger-print” (1,300-900 cm ). A doublet for stretching of –OH, sharp absorption peaks attributable to
asymmetric and symmetric stretching of –CH2, a strong band that arises from stretching of –C=O in
carboxyl and a sharp absorption peak attributable to rocking bending of –CH2 were also registered in the
-1
spectrum at 3,299, 2,916, 2,850, 1,730 and 720 cm . Similar strong absorption peaks were registered on
the spectrum of F1Hz (see Figure 3). All these signals are characteristic of glycerides (Tang et al. 2008a).
Figure 4 shows the FTIR spectrum obtained for the FSG standard. This spectrum displayed a broad
-1
absorption band instead of a doublet for stretching of –OH at 3,373 cm , less sharp absorption peaks that
-1
arise from asymmetric and symmetric stretching of-CH2 at 2,932 and 2868 cm , respectively, and signals
attributable to asymmetric and symmetric stretching of C–O–C and asymmetric stretching of O–C–C in
-1
glucoside moiety at 1,070, 1018 and 1,165 cm , respectively. Similar absorption peaks were registered on
the spectrum of F2Hz (see Figure 5). All these signals are characteristic of steryl glucosides (Bondioli et al.
2008; Van Hoed et al. 2008).
-1
An additional absorption peak registered on the F2Hz spectrum at 1,733 cm was attributed to the
presence of ASG. In a previous work, Lacoste et al. (2009) reported that these compounds can coelute
with the FSG present in soybean oil and palm oil when these oils were purified by column
chromatography.

4. Conclusions
The insoluble material that forms in Colombian palm oil biodiesel, characterized in this work for the first
time, consists in a mixture of monoglycerides and free steryl glucosides. More than three quarters of its
mass consisted in monoglycerides, which are probably mostly monopalmitin. Acylated steryl glucosides
appeared to be part of the haze as well, but more research is needed to confirm this preliminary finding. In
any case, these results should lead to rethink the strategies that have been developed to prevent the
formation of haze in biodiesel so far, as they have been mainly focused on the removal of free steryl
glucosides from biodiesel.

Figure 1: Scheme of TLC evaluations. Left TLC plate: developed with hexane/ethyl acetate. Right TLC
plate: developed with ethyl acetate/ methanol. 1. Hz1 first spot, 2. F1Hz, 3. F2Hz, 4. FSG standard, 5.
Monopalmitin standard, 6. Hz1 second spot

1083
1084
Transmittance [%] Transmittance [%]
20 30 40 50 60 70 80 90 100 20 30 40 50 60 70 80 90 100

3500
3500
3295.94 3295.94

Figure 3: FTIR spectrum of F1Hz

3000
3000
2915.38 2915.38
2850.17 2850.17

2500
2500

Figure 2: FTIR spectrum of the monopalmitin standard

2000
2000

Wavenumber cm-1
Wavenumber cm-1
1730.29 1730.29

1500
1467.92
1500
1467.92
1417.86 1417.86
1391.60 1391.60
1306.37 1306.37
1288.36 1288.36
1264.81 1264.81
1245.10 1245.10
1219.94

1000
1219.94
1000

1196.30 1196.30
1174.57 1174.57
1100.88 1100.88
1047.45 1047.45
990.65 990.65
944.31 944.31
719.44

500
719.44
500

660.67 660.67
 Transmittance [%]
65 70 75 80 85 90 95 100
Transmittance [%]
20 30 40 50 60 70 80 90 100

3500
3372.88

3500
3384.79

Figure 5: FTIR spectrum of F2Hz

3000
2932.22
2868.42

3000
2958.55
2920.23
2851.95

2500

2500
2113.16

2000
Wavenumber cm-1

F2 HZ2 22 NOVIEMBRE

2000
Wavenumber cm-1
Figure 4: FTIR spectrum of the free steryl glucosides standard
1733.04 1649.16

1500
1460.33

1500
1462.92 1369.89
1368.03
1247.59 1164.91
1198.30 1105.56
1168.15 1070.87
1104.97
1000

1018.66
1070.67

1000
1018.78
928.30 800.38
720.62
663.57 623.56
624.32
527.29 527.46
500

500.31

500
421.39
422.18

1085
Acknowledgments
The authors thank Ecodiesel Colombia S.A., especially Fernando Córdoba and Gloria Suarez for providing
the sample for this study. The authors also acknowledge the complete support for this research provided
by COLCIENCIAS and the Colombian Petroleum Institute ICP-Ecopetrol.

References
Bondioli P., Cortesi N., Mariani C., 2008, Identification and quantification of steryl glucosides in biodiesel,
Eur. J. Lipid Sci. Technol. 110, 120-126.
Farahani M., Pagé D.J.Y.S., Turingia M.P., 2011, Sedimentation in biodiesel and ultra low sulfur diesel fuel
blends, Fuel. 90, 951-957.
Janssen R., Rutz D.D., 2011, Sustainability of biofuels in Latin America: Risks and opportunities, Energy
Policy. 39, 5717-5725.
Lacoste F., Dejean F., Griffon H., Rouquette C., 2009, Quantification of free and esterified sterylglucosides
in vegetable oils and biodiesel, Eur. J. Lipid Sci. Technol. 111, 822-828.
Lin H., Haagenson D.M., Wiesenborn D.P., Pryor S.W., 2011, Effect of trace contaminants on cold soak
filterability of canola biodiesel, Fuel. 90, 1771-1777.
Moreau R., Scott K., Hass M., 2008, The identification and quantification of steryl glucosides in precipitates
from comercial biodiesel. JAOCS. 85, 761-770.
Plata V., Kafarov V., Moreno N., 2010, Optimization of third generation biofuels production: Biodiesel from
microalgae oil by homogeneous transesterification, Chemical Engineering Transactions, 21, 1201-
1206.
Plata V., Kafarov V., Castillo E., 2012, Improving the low-temperature properties and filterability of
biodiesel, Chemical Engineering Transactions, 29, 1243-1248.
Prada F., Ayala-Diaz I., Delgado W., Ruiz-Romero R., Romero H.M., 2011, Effect of fruit ripening on
content and chemical composition of oil from three oil palm cultivars (elaeis guineensis jacq.) grown in
Colombia, J. Agric. Food Chem. 59, 10136-10142.
Tang H., Salley O.S., Simon K.Y., 2008a, Fuel properties and precipitate formation at low temperature in
soy-, cottonseed-, and poultry fat-based biodiesel blends, Fuel. 87, 3006-3017.
Tang H., De Guzman R.C., Salley O.S., Simon K.Y., 2008b, Formation of insolubles in palm oil-, yellow
grease-,and soybean oil-based biodiesel blends after cold soaking at 4°C, JAOCS, 85, 1173-1182.
Van Hoed V., Zyaykina N., De Greyt W., Maes J., Verhe´ R., Demeestere K., 2008, Identification and
Occurrence of SterylGlucosidesin Palm and Soy Biodiesel, JAOCS. 85, 701-709.

1086