Talanta 77 (2008) 593–605

Contents lists available at ScienceDirect

Talanta
journal homepage: www.elsevier.com/locate/talanta

Review

Critical review on analytical methods for biodiesel characterization

Marcos Roberto Monteiro a,∗ , Alessandra Regina Pepe Ambrozin a ,
Luciano Morais Lião b , Antonio Gilberto Ferreira c
a
Laboratório de Combustíveis, Centro de Caracterização e Desenvolvimento de Materiais, Departamento de Engenharia de Materiais,
Universidade Federal de São Carlos, 13560-971 São Carlos - SP, Brazil
b
Instituto de Química, Universidade Federal de Goiás, Campus Samambaia, 74001-970 Goiânia - GO, Brazil
c
Departamento de Química, Universidade Federal de São Carlos, 13560-905 São Carlos - SP, Brazil

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

Article history: Biodiesel is an alternative fuel composed of mono-alkyl esters and obtained mainly from the base-
Received 11 March 2008 catalyzed transesterification reaction of oils or fats. Its use (pure or blended) does not demand any
Received in revised form 30 June 2008 modification in the diesel engine and in the existing fuel distribution and storage infrastructure. Moreover,
Accepted 1 July 2008
biodiesel has a high energetic yield, fixes the solar energy and contains insignificant amounts of sulphur.
Available online 9 July 2008
Therefore, biodiesel is currently the best substitute for fossil diesel fuel.
Besides mono-alkyl esters, glycerol (main co-product), alcohol, catalyst, free fatty acids, tri-, di- and
Keywords:
monoglycerides compose the final mixture of biodiesel production process. These and other kinds of
Biodiesel
Fatty mono-alkyl esters
contaminants can lead to severe operational and environmental problems. Therefore, the quality con-
Analytical methods trol of biodiesel is greatly significant to the success of its commercialization and market acceptance.
Spectroscopy Some important issues on the biodiesel quality control involve the monitoring of transesterification
Chromatography reaction, the quantification of mono-alkyl esters and free- and bonded glycerol as well as determina-
tion of residual catalysts and alcohol. Moreover, the determination of blend levels is another key aspect
of biodiesel analyses. Chromatography and spectroscopy are the analytical methods most used for the
biodiesel characterization, but procedures based on physical properties are also available.
Previously, a review on analytical methods used to evaluate biodiesel quality was written by Knothe.
Due to the importance of this field, we made an update of Knothes’ review. Therefore, in this paper, we will
describe new developments in biodiesel analyses and some references showed in Knothes’ paper. Specially,
we will describe analytical methods used for quantification of glycerol, mono-, di-, triglycerides, methanol,
water, Na, K, P, and steroids in biodiesel or along the transesterification reaction. Also, the determination
of biodiesel content in blends and some physicochemical parameters are discussed. At the end, we will
assess the available techniques and point out some improvements on analytical methods for biodiesel
characterization.
© 2008 Published by Elsevier B.V.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
2. General considerations on biodiesel analysis techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
3. Monitoring of the transesterification reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
4. Determination of fatty mono-alkyl esters in blends or pure biodiesel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597
5. Determination of free and total glycerol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599
6. Determination of other compounds and parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600
6.1. Water content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600
6.2. Methanol content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600
6.3. Steroids content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 601
6.4. Metals and metalloids content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 601

∗ Corresponding author. Tel.: +55 16 3351 8844; fax: +55 16 3351 8850.
E-mail address: [email protected] (M.R. Monteiro).

0039-9140/$ – see front matter © 2008 Published by Elsevier B.V.

doi:10.1016/j.talanta.2008.07.001
594 M.R. Monteiro et al. / Talanta 77 (2008) 593–605

6.5. Biodiesel oxidation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602

6.6. Biodiesel thermal stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
7. Physicochemical parameters and biodiesel chemical composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
8. Final considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604

1. Introduction confidence. In fact, EN 14214 and ASTM D 6751 biodiesel standards

[12,13] were already established in Europe and USA, respectively.
The petroleum market instability, limited availability of crude These standards are usually employed as references to establish
oil, and mainly the serious impact of the use of petroleum-based other standards and their analysis. Therefore, ASTM D 6751 and EN
fuels on the environment have currently stimulated the spread of 14214 specifications as well as their analysis methods for biodiesel
alternative fuels [1]. Many researches have been held on biodiesel, are illustrated in Table 1. In Brazil, Resolution number 7 (from
an alternative fuel for diesel engines that has recently obtained 19/3/2008) of the National Agency of Petroleum, Natural Gas and
increasing attention worldwide [2]. This fuel is the best substitute Biofuels (ANP) supplies the specification of the pure biodiesel
for diesel due to its physical properties, which are close to those of (B100) and the methodologies for its characterization. In fact, the
fossil diesel. Therefore, its use (neat or blended) does not demand Brazilian analyses are performed in accordance with the standards
any modification in the diesel engine and in the existing fuel dis- of the American Society for Testing and Materials (ASTM) and
tribution and storage infrastructure [3]. Besides that, biodiesel has Brazilian Association of Technical Standards (ABNT) [14].
a high energetic yield, fixes the solar energy and contains insignif- Therefore, the quality control of biodiesel is greatly important
icant amounts of sulphur [4]. Nevertheless, cold-flow properties, to the successful commercialization of this fuel and its blends [6].
NOx emissions, and oxidative stability are issues that have to be In this point, it is worth to note that the type (chain length, degree
overcome. At the moment, the main problem of biodiesel commer- of unsaturation and presence of other chemical functions) and con-
cialization is its price, which is higher than for fossil diesel [5,6]. centration of fatty esters as well as the structure of the ester moiety
Actually, the average price per gallon of diesel in the USA in June derived from the alcohol have an outstanding effect on biodiesel
2006 was US $2.98, while the cost of pure biodiesel was US $3.76 properties, which will also influence its storage and oxidation [4,11].
[7]. Moreover, it is worth to mention that the economical utilization Besides, biodiesel composition is completely dependent on the
of glycerine, the main co-product of biodiesel, is also an important source used to produce it [1]. Also, since some contaminants come
aspect for the feasible commercialization of this alternative fuel up from the transesterification reaction, it is important to monitor
[8]. the status of biodiesel production to recognize and correct prob-
According to the Brazilian Biodiesel Programme, biodiesel is lems at an early stage [15]. Fernando et al. [16] describe complete
defined as “a fuel obtained from mixtures, in different proportions, reaction to mono-alkyl esters, the removal of free glycerine, cata-
of fossil diesel and alkyl esters of vegetable oils or animal fats”. Thus, lysts, alcohol, and free fatty acids in biodiesel as critical issues of
biodiesel, under the chemical point of view, is composed mainly by the quality control. Moreover, the determination of blend levels
fatty acids mono-alkyl esters. Transesterification is a widespread is another key point of biodiesel analyses since its use has been
process used for obtaining biodiesel and involves the catalyzed increasing considerably [15].
reaction of triglycerides (major compounds of oils and fats) and The parameters, which are used to define the quality of biodiesel,
short-chain alcohols such as methanol and ethanol. Most trans- can be divided in two groups [10]. One of them is also used for min-
esterification industrial processes employ alkaline catalysis (KOH, eral diesel, and the second describes the composition and purity
NaOH, and NaOCH3 ) and methanol. The utilization of potassium of fatty esters. The former includes, for example, density, viscosity,
hydroxide is more suitable since, at the end of reaction, the mix- flash point, sulphur %, Conradson carbon residue, sulphate ash %,
ture can be neutralized with phosphoric acid to afford potassium cetane number, and acid number. The latter comprises, for exam-
phosphate, which is a fertilizer. In Brazil, the use of ethanol is more ple, methanol, free glycerol, total glycerol, phosphorus contents,
advantageous due to its large-scale production. Furthermore, the water and esters content as well as others properties described in
whole biodiesel manufacture process is independent of petroleum Table 1.
because ethanol is obtained from biomass [8]. The designation of Chromatography and spectroscopy are the most used analytical
pure biodiesel is B100 (100% fatty acid alkyl esters). However, this methods on biodiesel analyses, but procedures based on physical
fuel could be also used in blends with fossil diesel and the BXX properties are also available [6]. Moreover, it is worth to mention
abbreviation will indicate the volume (in percent) of B100 in the that most chromatographic analyses, mainly GC, have been applied
mixture. Consequently, B2 is constituted by 2% of B100, and 98% of to methyl and not to ethyl esters [15].
diesel [4]. In Brazil, B2 will be mandatory in 2008 and B5 in 2013 Previously, a review on analytical methods used for biodiesel
[9]. analysis was published [15]. It showed the main techniques devel-
As stated before, the transesterification reaction affords fatty oped until 2006. Due to the importance of evaluating biodiesel
esters. However, the final mixture is also composed of glycerol quality, we made an update of biodiesel references described by
(main co-product), alcohol, catalyst, tri-, di- and monoglycerides Knothe [15]. Therefore, in our review, we will describe new tech-
[4], as well as free fatty acids. These and other contaminants of niques as well as methods developed before 2006. Particularly,
biodiesel can lead to severe operational and environmental prob- analytical methods used for quantification of glycerol, mono-, di-,
lems. Therefore, standards that limit the amount of contaminants triglycerides, methanol, water, Na, K, P, and steroids in biodiesel
in biodiesel fuel are necessary. Additionally, the establishment or along the transesterification reaction will be described. Also,
of standards is also required to the description of the product the determination of biodiesel content in blends and some physic-
quality [10]. Knothe [11] states that the successful introduction ochemical parameters are discussed. At the end, we will assess
and commercialization of biodiesel has been accompanied by the the available techniques and discuss some improves on analytical
development of standards to ensure high product quality and user methods for biodiesel characterization.
 M.R. Monteiro et al. / Talanta 77 (2008) 593–605 595

Table 1
Biodiesel specifications according to ASTM D6751, and EN 14214 standards

Property ASTM D 6751 EN 14214

Test method Limits Test method Limits

Ester content – – EN 14103 96.5% (mol mol−1 ) min

Linolenic acid content – – EN 14103 12.0% (mol mol−1 ) max
Content of FAMEa with ≥4 double bonds – – – 1.0% (mol mol−1 ) max
MAGb content – – EN 14105 0.80% (mol mol−1 ) max
DAGc content – – EN 14105 0.20% (mol mol−1 l) max
TAGd content – – EN 14105 0.20% (mol mol−1 ) max
Free glycerine ASTM D 6584 0.020% (w/w) max EN 14105 0.020% (mol mol−1 ) max
Total glycerine ASTM D 6584 0.240% (w/w) max EN 14105 0.25% (mol mol−1 ) max
Water and sediment or water content ASTM D 2709 0.050% (v/v) max EN ISO 12937 500 mg kg−1 max
Methanol content – – EN 14110 0.20% (mol mol−1 ) max
(Na + K) content UOP 391 5.0 mg kg−1 max EN 14108 5.0 mg kg−1 max
(Ca + Mg) content – – prEN 14538 5.0 mg kg−1 max
P content ASTM D 4951 0.001% (w/w) max EN 14107 10.0 mg kg−1 max
Oxidative stability (110 ◦ C) – – EN 14112 6 h min
Density (15 ◦ C) – – EN ISO 3675 860–900 kg m−3
Kinematic viscosity or viscosity (40 ◦ C) ASTM D 445 1.9–6.0 mm2 s−1 EN ISO 3104 3.5–5.0 mm2 s−1
Flash point ASTM D 93 130 ◦ C min EN ISO 3679 120 ◦ C min
Cloud point ASTM D 2500 Not specified – –
Sulphur content ASTM D 5453 0.05% (w/w) max EN ISO 20864 10.0 mg kg−1 max
Carbon residue ASTM D 4530 0.050% (w/w) max EN ISO 10370 0.30% (mol mol−1 ) max
Cetane number ASTM D 613 47 min EN ISO 5165 51 min
Sulphated ash ASTM D 874 0.020% (w/w) max ISO 3987 0.02% (mol mol−1 ) max
Total contamination – – EN 12662 24 mg kg−1 max
Copper strip corrosion (3 h, 50 ◦ C) ASTM D 130 No. 3 max EN ISO 2160 1 (degree of corrosion)
Acid number or acid value ASTM D 664 0.50 mg KOH g−1 max EN 14104 0.50 mg KOH g−1 max
Iodine value – – EN 14111 120 g I2 ·100 g−1 max
Distillation temperature (90% recovered) ASTM D 1160 360 ◦ C max – –
a
FAME = fatty acid methyl esters.
b
MAG = monoacylglycerols.
c
DAG = diacylglycerols.
d
TAG = triacylglycerols.

2. General considerations on biodiesel analysis techniques HPLC analysis is less employed in biodiesel characterization,
but the analysis time is shorter than GC one and sample deriva-
Several factors such as composition of feedstock (oil or fat), tization is not needed. Moreover, this technique can be applied
production process (reaction and purification steps), storage and to biodiesel from different feedstock and it is more appropriate
handling can influence biodiesel fuel quality. The evaluation of for blend analysis than GC. Several detectors for HPLC biodiesel
biodiesel quality is achieved through the determination of chem- analysis are described, among them evaporative light scattering
ical composition and physical properties of the fuel. In fact, some detection (ELSD) is quite suitable.
contaminants and other minor components are the major issues Spectroscopic analyses are most employed for monitoring the
in the quality of biodiesel. As stated before, glycerol, mono-, di-, transesterification reaction and for the determination of blend
triglycerides, alcohol, catalysts and free fatty acid are compounds level. Nuclear magnetic resonance (NMR) spectroscopy and sev-
that could be present in biodiesel. Moreover, the biodiesel composi- eral techniques based on infrared spectroscopy (IR) are commonly
tion could be modified along the storage and handling. For example, applied to these analyses. NMR is an excellent technique, but the
biodiesel can absorb water or undergoes oxidation. Therefore, these instrumentation and maintenance costs are relatively high.
parameters and their analytical methods are addressed in stan- In the next sections, we will show some analytical methods
dards as well as alternative methodologies have been investigated. described for the determination of several compounds in biodiesel.
Additionally, the monitoring of transesterification reaction and the
blend level determination are significant aspects on biodiesel anal- 3. Monitoring of the transesterification reaction
ysis.
The most important parameters of biodiesel (fatty mono-alkyl As stated before, the transesterification monitoring is an impor-
esters, fatty acids, glycerol and their acyl derivatives) are commonly tant issue to biodiesel quality control since some contaminants
analyzed by gas chromatography (GC) and high-performance liq- arise from this reaction. Then, such monitoring allows recognizing
uid chromatography (HPLC). In fact, GC has been the most used and correcting problems at an early stage.
technique due to its high accuracy for the quantification of minor The first method for monitoring the transesterification reaction
components. However, baseline drift, overlapping signals, and of vegetable oils was developed by Freedman et al. [17]. For such,
aging of standards and samples can destructively affect the GC they employed thin layer chromatography (TLC) with FID. Besides
accuracy. Moreover, GC analyses frequently require sample deriva- fatty esters, this method allows to analyze tri- di-, and monoglyc-
tization, mainly to afford trimethylsilyl derivatives of the hydroxyl erides. Moreover, the analysis time is quite short; 30 samples could
groups. Although this procedure improves chromatographic sep- be analyzed in 2–3 h. However, this method shows lower accu-
aration, it also increases the analysis time. Flame ionization racy, sensitivity to humidity, material discrepancies as well as fairly
detection (FID) is the most widespread detector used in GC, but high cost of the instrument. Subsequently, other TLC on silica gel
the utilization of mass spectrometer has increased. The latter methodology was developed [18]. In this method, the area of tri-
eliminates any ambiguities about the identification of the eluting acylglycerol spot of mixture is compared to a standard. However,
materials, but their quantification could be affected. the analysis is only qualitative and does not allow the exact deter-
596 M.R. Monteiro et al. / Talanta 77 (2008) 593–605

mination of the degree of conversion. Actually, TLC is quite used for quicker and simpler than GC and HPLC. Moreover, a small amount
qualitative analyses, especially, for the evaluation of the oil conver- of sample is required and it could be analyzed without a pre-
sion, since it is fast and effective [4]. purification process. After that, Morgenstern et al. [28] developed
Afterward, Freedman et al. [19] developed the first GC method- a 1 H NMR method to monitor the transesterification reaction of
ology to monitor fatty acids, tri-, di-, and monoglycerides in soybean oil. Through NMR analysis, they were able to establish
the transesterification reaction of soybean oil. In this method, the average degree of fatty acid unsaturation and methyl esters in
before performing the analyses, mono- and diglycerides have to be biodiesel.
silylated with N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA). Other alternative for monitoring the transesterification of
Such procedure affords the trimethylsilyl derivatives, allowing a oils/fats is IR spectroscopy. The first reported work in this area is
better separation and tailing reduction. In this first GC work, from Knothe [29,30], which developed a fiber-optic near infrared
authors used a short fused silica capillary column (1.8 m; 100% (NIR) method to monitor the transesterification reaction of soy-
dimethylpolysiloxane), and tridecanoin as internal standard. The bean oil, which was based on the differences in the NIR spectra
complete separation of acylglycerols and fatty esters was obtained at 6005 and 4425–4430 cm−1 , where fatty esters display peaks
in a run time of 12 min. After that, other papers [20–22] described and triglycerides exhibit shoulders. Thus, these peaks were chosen
a GC–FID method with a packed column to evaluate the conversion for monitoring the reaction. In fact, the former peak was used for
of rapeseed oil to fatty ethyl or methyl esters. The evaluation was quantification since it is more suitable. Besides the transesterifi-
performed through the peak areas of esters during the reaction. cation monitoring, the biodiesel quality can be assessed through
The advantages of this method were the requirement of already this methodology by the correlation with other analytical data,
available compounds and the use of cheap columns. as pointed by Knothe [30], which correlated the NIR results with
The first HPLC method for monitoring transesterification was 1 H NMR spectroscopy. This kind of procedure is a way of cross-

developed by Trathnigg and Mittelbach [23]. They describe a checking the results. In addition, NIR method can also be used for
HPLC methodology with density detection (DD), which allows the quantification of methanol in biodiesel, which can be an alter-
the determination of the overall content of tri-, di-, and mono- native to flash point evolution of biodiesel. Finally, the NIR method
glycerides in biodiesel samples from methanolysis mixtures as could be used to monitor the reaction of several oils as well as
well as the methyl esters detection. The analyses were performed ensure the quality control of biodiesel when employed with other
through coupling a cyano-modified silica with two-gel perme- analytical techniques. Another work reports a FT-IR method and a
ation chromatography (GPC)-columns, and using an isocratic eluent multivariate approach (PLS, partial least squares regression), which
(chloroform/ethanol 0.6%). According to the authors, this method uses the ester peak at 1700–1800 cm−1 to monitor the ethanoly-
is simple and reliable. Subsequently, Holcapek et al. [24] employed sis of degummed soybean oil [31]. Such method was considered
reversed-phase (RP)-HPLC with several detection methods (UV, fast and accurate to predict reaction yields. Later, Siatis et al. [32]
at 205 nm; ELSD; and APCI–MS, atmospheric pressure chemical developed a FT-IR method to monitor the ultrasonically assisted
ionization–mass spectrometry) to monitor the transesterification extraction transesterification of seed and seed cakes from cot-
of rapeseed oil to methyl esters as well as to quantify the residual ton, sunflower, sesame, and Cynara cardulus seeds. The method
content of triacylglycerols. These detection methods are suitable affords the simultaneous determination of fatty acid methyl ester
for the analysis of complex mixtures due to their compatibility and triglycerides. They used PLS algorithm and 1300–1060 cm−1
with gradient elution, which is necessary to good resolution of spectral region to determine the percentage of fatty ester in the
methyl esters, mono-, di-, and triacylglycerols. The main disadvan- mixture of reaction. The most characteristic peak at 1200 cm−1
tage of the UV detection is a weak absorbance of acylglycerols at is that related to O CH3 vibrations in the methyl fatty esters.
wavelengths higher than 220 nm, besides the non-quantification of Subsequently, Ghesti et al. [33] described a FT-Raman spectro-
saturates. However, the sensitivity of APCI–MS and ELSD decreases scopic method to monitor and quantify the transesterification of
strongly with the increasing of double bonds in acylglycerols. In soybean oil. The differences between the vegetable oil Raman
conclusion, APCI–MS was considered the most suitable method for spectrum and the fatty ethyl esters spectrum were observed in
biodiesel analysis. bands at 2932 cm−1 (CH2 ), 861 cm−1 (R C O and C C ), 372 cm−1
Only in 1995, Gelbard et al. [25] described the first work on (ıCO O C ), as well by the displacement of the C O band from 1748
the utilization of nuclear magnetic resonance, particularly 1 H NMR, to 1739 cm−1 . They employed uni- and multivariate analysis meth-
for monitoring the yield of transesterification reaction. The peaks ods to build analytical curves and to check the method. Using an
of methylene group adjacent to ester moiety in triacylglycerols internal normalization standard ( CH band), the best results were
(␣-CH2 , 2.3 ppm, t) and the methoxy group in the esters (OCH3 , achieved by Raman/PLS calibration models. Afterward, the same
3.7 ppm, s) were used to follow the reaction progress. The con- authors [34] correlated this method with a new NMR technique
version was calculated from the areas of those peaks, using an developed for monitoring the ethanolysis of soybean oil but that
equation: C = 100 × (2AOCH3 /3A˛-CH2 ). The authors state that this could be extended to the transesterification of other vegetable oils.
method is faster and simpler than chromatographic ones. How- Also, a FT-IR method to assess the extent of transesterification reac-
ever, instrumentation and maintenance costs are relatively high tion of oils was recently described [35]. The author states that
and must be evaluated. Another work with NMR spectroscopy methyl peak areas intensities and positions could be employed to
describes the use of 13 C NMR for monitoring the rapeseed oil monitor the reaction. Finally, mid-IR with a multivariate approach
transesterification [26]. The signal at 14.5 ppm of the terminal was used to monitor on line the transesterification of soybean
methyl groups, which are not affected by reaction, was chosen as oil with ethanol [36]. For such, a cylindrical reflectance cell of
internal standard, and the glyceridic carbons at 62–71 ppm along PbSe and the range of 3707–814 cm−1 were employed. The mon-
with methoxy carbon of fatty esters at 51 ppm were selected to itoring was carried out for 12 min. The yield of the reaction was
determine the conversion rate. After that, a 1 H NMR methodol- also achieved through 1 H NMR analysis of samples collected from
ogy to monitor the soybean oil ethanolysis as well as to quantify the reactor. The proposed method could be used in the process
the content of fatty ethyl esters in mixtures of biodiesel and oil control and for reaction optimization since it is fast and shows
was developed [27]. The region of 4.05–4.40 ppm (ester ethoxy low-cost.
and glycerol methylene hydrogens) was chosen for the quantifi- Also, GPC (or SEC-size exclusion chromatography) is used for
cation of the reaction. The authors [27] state that the method is monitoring the transesterification. Dubé et al. [37] suggested two
 M.R. Monteiro et al. / Talanta 77 (2008) 593–605 597

distinct methods for monitoring the biodiesel production from umn [43]. For such, the analyses were carried out with selected
waste canola frying oil: GPC with refractive index detector (RID), ion monitoring (SIM) mode and the sample was submitted to a
and attenuated total reflectance (ATR)–FT-IR spectroscopy. The silylation reaction. The authors stated that the method also allows
decreasing of the peak at 1378 cm−1 (terminal methyl and OCH2 in distinguishing between un- and distillated products. After that, the
acylglycerols) in the ATR–FT-IR spectrum was used for monitoring same authors described the quantification of fatty methyl esters
the biodiesel synthesis. The results of the analyses of 100 biodiesel in biodiesel/diesel blends using GC–FID [44]. The process consists
samples by both methods showed very small differences. Thus, in the separation of biodiesel from diesel on silica cartridge with
the authors consider that the methodologies could be equivalent. hexane/diethyl ether, and thus in the GC analysis. Additionally,
However, the GPC–RID methodology allows only the quantification a relationship was established between saponification value (SV)
of monoglycerides and fatty methyl esters, while the ATR–FT-IR and percent of biodiesel in blends, which could also be useful for
just consents the quantification of the sum of mono-, di- and such determinations. In a subsequent work of these authors [45],
triglycerides. Finally, Arzamendi et al. [38] described a method they described a GC–FID for the determination of biodiesel lev-
based on SEC to determine simultaneously the total amount of els in biodiesel/diesel blends, which differs from the former on the
tri-, di-, monoglycerides, fatty acid methyl esters, free glycerol and sample pre-treatment. This new methodology consists on a prelim-
methanol in samples from the transesterification reaction of sun- inary acetylation of the sample followed by a separation process
flower oil. Two Styragel® HR0.5 and HR2 columns with RID and using a silica cartridge and hexane as solvent. The authors state
viscometer detectors as well as THF solvent were employed for the that the results are very encouraging in terms of precision and
analyses. According to the authors, the method is simple, robust, accuracy. Additionally, capillary GC–FID was also employed for the
relatively fast, it may be conducted at room temperature and it gives identification of fatty acid methyl ester composition of biodiesel
accurate and reproducible results. samples from six edible vegetable oils (rapeseed, peanut, corn,
Other techniques were also proposed. Xie and Li [39] demon- cotton-seed, sesame and soybean oil) as well as for the determi-
strated a method for monitoring the progress and the end-point of nation of the fatty acid composition of those oils [46]. A similar
the transesterification of soybean oil and to determine the yield work with GC–FID, employing HP-1 wide-bore column, was also
of the reaction, without the necessity of sample derivatization. described [47]. According to the authors, this method is simple,
Such method is based on hydroxyl content or refractive index mea- rapid, and accurate. Subsequently, Li et al. [48] developed a sim-
surements of transesterified mixture since the compounds of such ple and rapid GC–FID method for the quantification of fatty acid
blend (glycerin, mono-, di-, triglycerides, and fatty methyl esters) methyl esters and glycerides in biodiesel without the need of sam-
have significantly differences in those properties. The hydroxyl con- ple derivatization. This could be achieving due to the use of a
tent were measured according to AOCS Official Method Cd 13–60, non-polar and high temperature-resistance capillary column. After
which consists in an acetylating reaction with acetic anhydride that, another GC–FID method for qualitative and quantitative anal-
in pyridine followed by a titration with potassium hydroxide. The ysis of fatty acid methyl esters in biodiesel, which employs a polar
refractive measurements were performed after washing the trans- capillary column (J&W INNOVAX), was developed [40]. They stated
esterified mixture three times with NaCl solution. Authors believe that the method has high accuracy and precision, and it does
that this methodology is advantageous and useful for control pro- not require special preliminary sample preparation. However, this
cess, while it is simple, rapid, and inexpensive. Also, the use of method requires the use of several internal standards (lauric, myris-
an acoustic wave solid-state viscometer (ViSmartTM ) to monitor tic, palmitic, stearic, oleic, linoleic, and linolenic esters). An accurate
the transesterification reaction is described [7]. The progress of GC–FID method is also described for evaluating fatty methyl esters
the reaction is evidently indicated for the decreasing viscosity in biodiesel [49]. According to the authors, recovery and standard
of the mixture. The viscometer was able to detect the end-point deviations were satisfactory. Finally, Schober et al. [50] developed
of the reaction in pilot-scale and could therefore be used in the a GC–FID method for quantifying the methyl esters in biodiesel,
future to monitor the batch production process of biodiesel. The which is based on EN 14103. The latter does not include heptade-
main advantage of the use of this kind of viscometer is that it canoic acid ester that is found in animal fats. Thus, the authors
does not require an extra step of measuring the density like oth- showed that the amount of esters is increased in 2–7 wt.% by the
ers. use of their methodology. Moreover, such method allows the deter-
mination of short-chain fatty acid esters (C8 –C12 ), which occur in
4. Determination of fatty mono-alkyl esters in blends or some biodiesel samples from coconut and palm oil. In these cases,
pure biodiesel the ester content of such biodiesel differed by >40 wt.%. The authors
suggest the use of fresh standard solutions, since the stability of the
Pure biodiesel is mostly composed of fatty mono-alkyl esters, methyl heptadecanoate (internal standard) influences the values of
usually by ethyl or methyl esters. Best engine performance requires ester content.
98.8% of esters minimum content [40]. Another method described for evaluating the methyl ester con-
In the literature, there are many reports about esters determi- tent of biodiesel samples is based on viscosity measurements
nation in blends, and pure biodiesel. Frequently, spectroscopy and [51]. There is a correlation between the content of esters and the
chromatography have been used for assessing biodiesel quality and viscosity: the higher the viscosity, the lower the esters content.
for monitoring transesterification reaction, as previously discussed. According to these authors, since the method is quick and sim-
For analyzing fatty methyl ester content in pure biodiesel, there is ple, it is especially suitable for process control purposes, such as,
the EN 14103 standard method, which employs GC–FID. However, transesterification monitoring.
for determining the blend level of biodiesel in diesel, chromatog- After that, Fillieres et al. [52] employed a high-performance
raphy seems to be less suitable due to the complexity of diesel size exclusion chromatography (HPSEC or SEC) for the quantifi-
composition [41]. Currently, the most widely used technique is cation of ethyl esters, mono-, di-, and triglycerides, and glycerol
medium FT-IR spectroscopy, which is also the base of the European in biodiesel from rapeseed oil. Their purpose was to evaluate the
standard reference method [42]. influence of different parameters on the transesterification reac-
The first work for the quantification of methyl esters, as well as tion. After that, another GPC–RID method for quantifying methyl
mono-, di- and triglycerides in pure biodiesel, describes the use esters, tri-, di-, and monoglycerides, glycerol in biodiesel samples
of GC–MS with a (5%-phenyl)-methylpolysiloxane capillary col- from palm oil was developed [53]. Such methodology employs two
598 M.R. Monteiro et al. / Talanta 77 (2008) 593–605

Phenogel columns and tetrahydrofuran as solvent. Moreover, the palmitic, stearic, oleic, and linolenic) in biodiesel. Determina-
sample preparation is simple, involving only dilution and neutral- tion of esters content showed errors lower than those obtained
ization with HCl. through reference method. After that, Soares et al. [62] developed
A HPLC method was described for quantifying fatty ethyl, iso- a PLS–ATR–FT-IR method to quantify raw oil, from 1 to 40% (v/v), in
propyl, 2-butyl, and isobutyl esters as well as fatty acids, tri-, di-, different biodiesel samples. Such calibration model could preview
and monoglycerides in biodiesel samples from fats, oils, and recy- with 95% of significance the samples analyzed.
cled greases [54]. The process involves the use of an ELSD detector A recent work [63] suggested the evaluation of biodiesel qual-
and a gradient elution profile (hexane and tert-butyl ether, both ity through determination of oil content by an optical fiber sensor.
with 0.4% acetic acid). This method is based on a previous method- Such method showed errors of 0.4 and 2.6% for pure biodiesel and
ology described in the literature, but it showed a reduction of 25% soybean oil, respectively. This fiber sensor, which is a compact and
in analysis time. Succeeding the first HPLC work, Holcapek and Jan- cheap device, has high sensitivity and allows the ease assessment
dera [55] developed a HPLC–APCI–MS method for the identification of biodiesel quality.
of fatty methyl esters, mono-, di-, and triacylglycerols in differ- The blend level of biodiesel in diesel could be determined
ent kinds of biodiesel from rapeseed, sunflower, soybean, palm, through the use of both NIR and 1 H NMR techniques [41]. Soy-
macadamia, almond, poppy seed, and hazelnut oils. The authors bean biodiesel was employed in this study. Two spectral regions
emphasize that the method can be used with any kind of biodiesel. (6.005 and 4.800–4.600 cm−1 ) were chosen to the NIR analyses.
Subsequently, HPLC method for quantifying soy biodiesel (1–30%, Moreover, in 1 H NMR procedure were used the peaks of the methyl
v/v) in diesel was developed [42]. This analytical methodology ester moiety (3.6–3.7 ppm), the clusters of peaks (0.8–3.0 ppm)
could also be used for the determination of the same amount of from the methylene and terminal methyl protons of the hydro-
triglycerides in diesel. The methodology employs silica column carbon moieties in biodiesel and diesel, and the olefinic hydrogen
with isocratic elution (hexane 90% and methyl t-butyl ether 10%) (5.3–5.4 ppm) in biodiesel for determining blend levels. Both meth-
and UV or ELSD detectors. The authors state that ELSD detector is ods were in good agreement, and the NIR methodology is easy and
preferred over UV because its response is directly related to the rapid.
mass of solute injected. On the other hand, UV detector’s response Additionally, Birova et al. [64] described two methods to evalu-
is proportional to the number of double bonds. However, there ate the fatty methyl esters content in biodiesel/diesel fuel blends.
is no statistical difference between the analyses of both detec- One of them is based on the ester number (difference between the
tors. Therefore, Foglia et al. [42] concluded that HPLC is a fast saponification and the acid values) and the other one uses the IR
technique that can analyze blends of biodiesel produced from var- spectroscopy, specially the measurement of carbonyl band inten-
ious feedstocks in petrodiesel. Subsequently, Kaminski et al. [56] sity. According to the authors, both methods are suitable to the
described a procedure that enables the simultaneous determina- desirable determination. Oliveira et al. [65] also developed an ana-
tion of aromatic hydrocarbons, total content of polycyclic aromatic lytical method for the identification of methyl esters in biodiesel
hydrocarbons, and fatty acid methyl esters in diesel, which contain blends by ATR–FT-IR and NIR spectroscopies combined to PLS and
up to 30% of these esters. A normal-phase separation (LiChro- ANN (artificial neural network) analysis. Through these methodolo-
spher NH2 5 ␮m column), using n-heptane as mobile phase, was gies were analyzed blends of one type of biodiesel and diesel (Group
employed. Besides that, two HPLC detectors connected in series I) besides mixtures of three kinds of biodiesel and diesel (Group II).
were used. The first one was UV–DAD (ultraviolet diode array detec- The biodiesel was obtained from different vegetable oils: soybean,
tor) followed by RID. The UV detection at 260 nm is preferred as soybean fried, babassu, and dende. According to the results, PLS
it provides higher accuracy and precision than RID. According to model based on NIR was more precise and accurate than the one
the authors, this method with detectors in series is advantageous based on ATR–FT-IR for the analysis of Group I samples. However,
since it provides correct and precise determination of the backflush NIR and ATR–FT-IR ANN models for Group I presented similar pre-
point, confirmation of the fatty acid methyl esters, and detection cisions and accuracies. For Group II, both PLS methods (NIR and
of the presence of resins, which occur because of long-term stor- ATR–FT-IR) had similar accuracies while the precision of PLS NIR
age. Additionally, HPLC and GC could also be used together for was better. However, ANN models for Group II presented almost
the quantification of ethanol, fatty ethyl esters, mono-, di-, and equivalent precisions and accuracies. The authors mentioned that
triglycerides in ethanol biodiesel from waste cooking oil [57]. A new these methods are faster than the chromatographic ones and sam-
RP–HPLC–UV method was recently developed in order to determine ple integrity is preserved. Consequently, they could be used to
fatty acid methyl esters, free fatty acids, mono-, di-, and triacylglyc- determine different methyl esters contents in biodiesel blends. A
erols in biodiesel [58]. Detection was mostly affected by flow rate similar work was described by Pimentel et al. [66]. However, in
and gradient end time. Such method allowed the identification of such work, PLS multivariable calibration models based on mid-FT-
all components in less than 30 min. Also, a high speed LC–MS sys- IR and NIR spectroscopy were developed in order to determine the
tem, which employs electrospray ionization (ESI) and a patented level of biodiesel and/or raw vegetable oils in blends with fossil
cone-wash feature, was described as a efficiency way to identify diesel. Additionally, PCA (principal component analysis) methods
fatty methyl esters and other compounds in biodiesel [59]. The were employed to fast identification of diesel samples contami-
main advantage of this technique is the reduction of the analysis nated with raw vegetable oils. Soybean, castor and used frying
time. oils as well as their biodiesel were employed in the preparation
Supercritical fluid chromatography (SFC) was also used to deter- of diesel blends. Subsequently, Oliveira et al. [67] showed that NIR
mine fatty acid methyl esters, fatty acids and glycerol in biodiesel and Raman spectroscopy combined with chemometric methods are
samples [60]. This is a preliminary study that allowed the analysis quite useful to identify residual oils in biodiesel/diesel blends. PLS,
of these compounds in less than 5 min. The experiments were car- principal component regression (PCR), and ANN calibration mod-
ried out in a SFC–MS–UV–ELSD system, using a C-18 column and els were applied to NIR and FT-Raman spectroscopic data aiming
isocratic elution. to determine adulterations of B2 and B5 blends with vegetable oils.
Recently, several works have reported the use of IR spectroscopy Among the designed models, FT-Raman ANN showed the best accu-
to assess the biodiesel quality control. For example, Baptista et al. racy (0.03%, w/w). Also, Aliske et al. [68] developed a FT-IR method
[61] described a PLS-NIR methodology, which allowed to deter- to the determination of biodiesel and diesel mixtures. The method
mine esters content and some specific acid methyl esters (myristic, covers the full ranges of mixture (0–100%) and it employs the
 M.R. Monteiro et al. / Talanta 77 (2008) 593–605 599

carbonyl peak present only in biodiesel spectrum for the quantifica- Since the limits of bonded and free glycerol are very low, there is
tion. This methodology stands for a simple way to perform quality a need for precise and reliable analytical methods for both determi-
control and monitoring of biodiesel–diesel blends. It is worth to nations [10]. In the literature, there are several procedures for the
notice that the method was developed using ethyl biodiesel from determination of free and total glycerol in biodiesel and its blends,
soybean oil and diesel blends. Recently, Guarieiro et al. [69] also such as HPLC and GC methods. Usually capillary GC is used for it
described a FT-IR method to determine biodiesel content in diesel [75]. According to Mittelbach [10], only GC meets all requirements
blends through area measurement of the peak at 1754 cm−1 . The for the determination of low contents of mono-, di- and triglyc-
method is fast, low-cost and it allows the determination of biodiesel erides in biodiesel. However, the purity of standard substances
content upper than 0.1%. must be checked by HPLC since such feature can lead to inaccurate
One spectrophotometric technique was also established for analyses.
the determination of biodiesel content in blends [70]. The spec- The first method, described for determining the total glycerol
trophotometric features in the wavelength range of 190–1100 nm in biodiesel, is based on enzymatic procedure [77]. The process
of diesel/biodiesel blends were investigated. Biodiesel from six dif- involves solid-phase extraction, saponification reaction, followed
ferent feedstocks and five distinct diesel samples were used for such by enzymatic analysis of the sample. This method, which was devel-
study. The authors concluded that UV spectroscopy is a reliable and oped for the rapeseed biodiesel, does not distinguish among mono-,
reasonable method for blend level detection of any biodiesel and di-, and triglycerides. Moreover, it is relatively complex and it shows
different diesel samples. After that, Sastry et al. [71] also described fairly low reproducibility. Afterward, the Sigma–Aldrich Fine Chem-
the use of spectroscopic and conventional methods for the deter- icals developed a kit (BQP-02) for enzymatic determination of free
mination of biodiesel in biodiesel/diesel blends. The authors state and total glycerin in biodiesel. The method is based first on the
that such methods are simple, fast and reliable. enzymatic reaction of free and bonded glycerol and then on the
A practical procedure using hydrometry was also developed for spectrometric measurement of obtained colored product.
the determination of biodiesel content in blends [72]. In this work, Subsequently, Bondioli et al. [78] described a GC–FID method
several procedures and instruments for determining biodiesel for the determination of free glycerol in biodiesel. Sample deriva-
blends were evaluated at different temperatures. Specific gravity tization was not necessary and a glass column (2 m × 4 mm i.d.),
(hydrometer), density (balance/volumetric flask), electromagnetic loaded with Chromosorb 101, was employed. The method is only
absorbance (spectrophotometer) and viscosity (zahn viscometer) reliable for rapeseed oil, sunflower-oil, and soybean biodiesel and
of different blends were measured. In fact, diesel, B5, B10, B20 it is suitable if the quantity of free glycerol present in the sam-
and B100 from distinct feedstocks (canola, sunflower, soybean and ple is higher than 0.02%. Most of the reports on GC utilization
corn) were evaluated. According to the authors, the best method for biodiesel analysis describe the use of FID [15]. However, there
is hydrometry due to its simplicity, accuracy and low instrument is a GC–MS method for the quantification of mono-, di-, and
cost. triglycerides in methyl biodiesel fuels [79], which employs a (5%
Recently, Corgozinho et al. [73] showed that synchronous flu- phenyl)-methylpolysiloxane column (10 m × 0.25 mm i.d.) and MS
orescence spectroscopy (SFS) and chemometrics are useful to in SIM mode. The analysis is performed after silylation with BSTFA.
identify and quantify residual oils in biodiesel/diesel blends. Spe- According to the authors, the method gives excellent quantifica-
cially, PLS, PCA and Linear Discriminant Analysis (LDA) applied to tion data. Mittelbach [80] also described a GC method with FID
spectrofluorimetry data allowed both discrimination and quan- or MS detection for the determination of free glycerol in methyl
tification of vegetable oil in diesel and B2 blends. According to biodiesel, using a DB-5 column (60 m × 0.25 mm). The methodol-
the authors, the method is simple, fast, accurate and reliable ogy is quite similar to that described by Plank and Lorbeer [79],
to detect adulteration of diesel and B2 blends with vegetable involving sample derivatization with BSTFA, and analysis by GC–MS
oils. in SIM mode. The author states that such method is more sensitive
Also, a radiocarbon-based method was developed to quantify and quicker than the others that had been described. Subsequently,
biodiesel in diesel blends with accuracy of ±1% [74]. Such method Mittelbach et al. [81] improved the GC–FID, GC–MS method [80],
does not require any knowledge about biodiesel type or diesel which allowed the simultaneous determination of free glycerol and
components as well as a calibration curve. However, cost and turn- methanol in biodiesel. After that, an important GC–FID procedure
around-time of analysis are critical disadvantages of this method. was developed [75], which later became the EN 14105 and ASTM
6584 methods. Such analytical method allows the simultaneous
5. Determination of free and total glycerol determination of glycerol, mono-, di-, and triglycerides in vegetable
oil methyl esters predominantly consisting of C18 methyl esters,
Glycerol is the major by-product in the manufacturing of such as biodiesel from rapeseed, sunflower, soybean and used
biodiesel. Its removal is desirable since it can cause damage to the frying oil. The method was mainly developed to the methyl rape-
engine and hazardous emissions. As a consequence, quality control seed biodiesel. However, it could not be applied to methyl esters
of this compound is essential [75]. According to ANP legislation, obtained by transesterification of lauric oils without modifications.
0.02% of free glycerol is the maximum quantity permitted [2]. More- The determination of all classes of compounds was achieved by
over, European and US standards specify the tolerate limits of free the silylation of the free hydroxyl groups, employing N-methyl-
and total glycerol (sum of glycerol, mono-, di- and triacylglycerols) N-trimethylsilyltrifluoroacetamide (MSTFA), followed by capillary
(Table 1). Such compounds can contaminate biodiesel samples due GC analysis in a DB-5 column (10 m × 0.32 mm). Moreover, the use
to incomplete transesterification and insufficient purification. In of 1,2,4-butanetriol and tricaprin as internal standards allowed for
fact, during the biodiesel production, washing steps can easily reliable quantitative analysis within a run time of 30 min. Therefore,
remove free glycerol, while a low content of glycerides can only this method is suited for the quality control of biodiesel. However,
be achieved by the use of suitable catalysts and reaction conditions Bondioli and Della Bella [76] state that the use of an alternative
or by further distillation of the product [10]. Also, some distilled method sometimes is necessary for the efficient control of biodiesel
biodiesel samples may contain free glycerol distilled as a head prod- since the presence of trace of volatile products in some samples may
uct of this unit operation [76]. Finally, Knothe [30] and Mittelbach cause interference in the Plank’s method. Ruppel and Hall [82] also
[10] propose that the amount of glycerol and glycerides is a major evaluated the ASTM D 6584 method and consider that the method is
factor in determining fuel quality. simple, sensitive and reliable, and it requires only a small amount of
600 M.R. Monteiro et al. / Talanta 77 (2008) 593–605

sample preparation. Subsequently, the ASTM D 6584 method was quick, economical and sufficiently reliable. They also demonstrated
improved through the use of capillary flow technology [83] and the feasibility of this procedure applied to a diesel fuel/biodiesel
high temperature GC columns [84–86]. An alternative methodol- blends (5 and 20%). However, more tests are necessary to evaluate
ogy for the chemical derivatization of glycerides, which employed the robustness and the suitability of the procedure to this kind of
fluorinated acid anhydrides and affords perfluoro alkyl esters, was matrix.
described [87]. Such procedure can reduce both the analysis time Finally, Gonçalves Filho and Micke [2] developed a capillary elec-
and cost. After derivatization, the authors used GC–MS technique trophoresis method (EC–DAD) for the quantification of free glycerol
and FT-IR spectroscopy to measure the amount of mono- and in biodiesel. Before the analysis, the reaction between glycerol and
diglycerides in biodiesel. periodate (HIO4 ) is carried out, in less than 2 min, to afford iodate
The first HPLC procedure for the analysis of glycerol was (HIO3 ). Commercial biodiesel from chicken fat, soybean, and castor
described in the work of Lozano et al. [88]. They employed a HPLC oil were analyzed by this methodology. All samples had a value less
method with pulsed amperometric detection (PAD) to measure free than the specified by ANP. According to the authors, this method-
glycerol in methyl or ethyl biofuels. This method also allows the ology for the extraction and analysis of free glycerol in biodiesel is
detection of residual alcohol in the biodiesel sample. According fast, simple, and reliable.
to the authors, the method is simple, rapid and accurate. In fact,
the sensitivity seems to be the major advantage of this method. 6. Determination of other compounds and parameters
In a subsequent work, Sala and Bondioli [89] evaluated two dis-
tinct methodologies for determining glycerol. One of the methods 6.1. Water content
employed a titration with periodate and the other one was based
on HPLC technique. From the results, the authors concluded that The water content in biodiesel is an important factor in the
the periodate method showed very good precision and accuracy, quality control. Water can promote microbial growth, lead to tank
which were improved by the use of a potentiometric titration, corrosion, participate in the formation of emulsions, and cause
and HPLC analysis is less time-consuming and it supplies more hydrolysis or hydrolytic oxidation. Therefore, the content of water
information about the sample. Also, an HPLC–RID methodology for is limited to 0.05% (w/w) according to EN 14214 and ASTM D 6751
the determination of free glycerol in biodiesel is described [90]. standards (Table 1). Moreover, such standards establish the use
Before the analysis, the sample has to be extracted with water. of centrifugation or Karl–Fischer titration for determining water
This method is reliable and faster than GC ones, giving the same content in biodiesel. A procedure based on the latter method
analytical results. Finally, Foglia et al. [91] described a comparison was recently described [94]. Besides that, Felizardo et al. [95,96]
between high temperature gas chromatography (HTGC) and HPLC developed a NIR method together with PLS and PCR analyses to
for the determination of bound glycerol in soybean and rapeseed determine water and methanol content in biodiesel. The authors
biodiesel. HTGC–MS and HPLC–ELSD showed similar results. How- emphasize that the use of a pre-processing method, such as OCS
ever, in the point of view of operation, HPLC is better than HTGC (orthogonal signal correction), is especially important in the devel-
because it does not require sample derivatization, it has shorter opment of PLS and PCR models employed for water and methanol
analysis time (as proposed before) and it is directly applicable to determination. They also consider that the use of NIR spectroscopy,
most biodiesel fuels. Besides that, HPLC shows versatility in ana- in combination with multivariate calibration, is a promising tech-
lyzing biodiesel from different feedstocks and several alkyl esters nique to assess the biodiesel quality. Finally, Todd et al. [97] describe
(methyl, ethyl, isopropyl). a novel direct sampling mass spectrometry membrane to measure
The combination of HPLC and GC is also reported to the free or the water content in biodiesel. This method could be used for mon-
total glycerol analysis in biodiesel. This practice reduces the com- itoring the synthesis process of biodiesel.
plexity of the GC chromatograms and it allows more reliable peaks
assignments. Lechner et al. [92] developed a LC–GC method for 6.2. Methanol content
determining mono-, di-, and triacylglycerols in methyl biodiesel
samples, which has 52 min of run time. Before the analyses, samples The residual methanol in biodiesel can cause corrosion of met-
are submitted to acetylation and then analyzed through HPLC–DAD, als, mainly of aluminium, and decrease the biodiesel flash point.
and GC–FID, employing a DB-5 column (10 m × 0.32 mm i.d.). Besides, it is responsible for cetane number and lubricity decreas-
In a recent work, Catharino et al. [93] developed a MS method, ing of fuel. ASTM D 6751 limits indirectly the methanol content
which employs direct infusion ESI and it allows the determina- through the flash point minimum value. However, the EN 14214
tion of residual glycerol, mono-, di-, and triglycerides, besides standard, beside the flash point, establishes 0.2% (w/w) as the
fingerprinting typification of biodiesel, alcohol identification, and maximum content of methanol (Table 1). Previously, we described
monitoring of degradation and adulteration. Typification and analytical methods [38,81,95,96] that allow, beside the determi-
biodiesel degradation were performed in the negative ion mode nation of other compounds, the quantification of methanol in
(ESI−), while other parameters were determined using positive ion- biodiesel. Moreover, Fang and Zeng [98] developed a UV spec-
ization (ESI+). This technique seems to be very useful since it allows troscopy procedure for the determination of methanol in biodiesel
both typification and fast screening of some important parameters samples. According to the authors, this method was reasonable
related to biodiesel quality. and it had good reproducibility and accuracy. Subsequently, Li
A spectrophotometric method based on measurements of 3,5- et al. [99] described a GC–FID method for the determination
diacetyl-1,4-dihydrolutidine at 410 nm was also developed for of small amounts of methanol in biodiesel. Such methodology
the glycerol determination [76]. This compound is obtained from employs two columns (a pre-column and a polar PEG-20M one)
Hantzsch reaction of glycerol, which consists in two successive with pressure backflush system and it was used for the anal-
reactions: periodate oxidation of free glycerol affords formalde- ysis of 8 different biodiesel samples. Additionally, the authors
hyde and then the reaction of this compound and acetylacetone discuss several parameters that affect fatty acid methyl esters,
in the presence of ammonium acetate gives the desirable product. free fatty acids, mono-, di-, triglycerides, and methanol determi-
The 3,5-diacetyl-1,4-dihydrolutidine has a very high absorption at nations in biodiesel. After that, Araujo et al. [100] developed a
410 nm, increasing the potential to get a very low detection limit fast and reliable flow methodology, which employs a microporous
for glycerol. According to the authors, this methodology is simple, hydrophilic membrane to extract methanol from biodiesel, and it
 M.R. Monteiro et al. / Talanta 77 (2008) 593–605 601

determines methanol content by UV measurements (at 240 nm), In the first work described, the determination of Ca, Cl, K, Mg,
after derivatization with alcohol oxidase (AOD), soluble peroxi- Na, and P in biodiesel was carried out through inductively couple
dase, and 2,2 -azinobis(3-ethylbenzothiazoline-6-sulphonic acid) plasma optical emission spectrometry (ICP-OES) [106]. This tech-
(ABTS). Such methodology showed good precision, with a relative nique is absolutely suitable to control the production of biodiesel
standard deviation <5.0% (n = 10) and detection capacity of 0.211% and the quality of final product since it has excellent analytical
(w/w). The results obtained with this method showed good cor- properties such as multi-element capability, higher power of detec-
relation with those furnished by GC. However, the flow method tion, high precision and short analysis time. The monitoring of Ca, K,
seems to be more environmental friendly and cost-effective than Mg, and Na amount is necessary due to their ability to form undesir-
the GC reference one. Other method to determine methanol con- able compounds in the engines. In addition, the evaluation of Na and
tent in biodiesel was recently proposed by Paraschivescu et al. K is also important since some production processes employ KOH
[101]. It involves a headspace solid-phase microextraction, using or NaOH as catalysts. Additionally, environmental issues justify the
a carboxen-polydimethylsiloxane SPME fiber, and a subsequent control of Cl content. The authors of this first work also determined
GC–FID analysis, employing a HP-5 capillary column. Such method that oxygen addition to the outer, intermediate or nebulizer gas or
showed good reproducibility and recovery, allowing analysis of argon addition to nebulizer gas reduce the background emission of
methanol in concentrations lower than those imposed by the stan- the spectral lines of Na and K. The most significant improvement
dard specifications. of background ratio is acquired when oxygen or argon is added to
nebulizer gas. In fact, the addition of an argon–oxygen mixture to
the nebulizer gas improves significantly the detection of Na and K.
6.3. Steroids content
The sensitivity of these elements searches out the ␮g kg−1 range.
Finally, the authors state that the method demonstrates linear cali-
Sterols are natural non-glyceridic compounds that occur in veg-
bration range and short analysis time with high sample throughput.
etable oils, and can arise in biodiesel due to their solubility in
Besides that, chemical and spectral interferences are less encoun-
such fuel. Consequently, their amount may influence biodiesel
tered than in AAS (atomic absorption spectrometry). Another work
quality. The GC–FID and on-line LC–GC methods are described
that employs ICP-OES for simultaneous determination of Ca, P, Mg,
to determine the sterols and their esters in biodiesel. The LC–GC
K and Na was described by dos Santos et al. [107]. In contrast to
is more recommended because of additional information, short
other methods, this procedure uses a less toxic solvent (ethanol)
analysis time, and reproducibility [15]. In fact, Plank and Lor-
for dilution of the biodiesel sample. Woods and Fryer [108] also
beer [102–104] have some works concerning sterols analysis in
used ICP to determine several inorganic species in biodiesel sam-
biodiesel. In their first work [102], they describe a GC–FID method
ples. However, MS detection and an octopole reaction system were
for determining free and esterified sterols in methyl biodiesel from
used, allowing measurements at levels below those obtained by
rapeseed oil. Before the analysis, free sterols were silylated with
ICP-OES. Besides that, the simultaneous analysis of a wide variety
BSTFA, which hold 1% of trimethylchlorosilane (Me3 SiCl), and thus
of elements is a valuable feature of ICP–MS. Nowka [109] devel-
they were analyzed in a (5% phenyl)methylpolysiloxane column.
oped a flame atomic absorption spectrometry (FAAS) method for
Brassicasterol, ␤-sitosterol, campesterol, cholesterol, stigmasterol,
the determination of Na and K in biodiesel. In this method, air was
and 5-avenasterol as well their esters were identified in rapeseed
used as oxidant aiming at avoiding the baseline shift of K. Moreover,
biodiesel. After that, Plank and Lorbeer [103] developed a simi-
interferences of the particle emissions from the matrix were min-
lar GC–FID for determining free sterols in rapeseed methyl esters
imized by the use of a suitable technique. Finally, the method was
as well as an online LC–GC method to determine free sterols and
compared with other techniques and the results correlated well.
sterol esters in the same samples. The former analyses were carried
Furthermore, a photometry method for the determination of phos-
out after saponification. The results showed the total sterol content
phorous in biodiesel samples was developed [110]. This method is
(0.70–0.81 wt.%) consisted of cholesterol, brassicasterol, campes-
based on phosphomolybdenum blue photometry and it involves
terol, stigmasterol, ␤-sitosterol, and 5-avenasterol. Moreover, by
carbonization and sample transformation in ash before the analy-
online LC–GC, 0.24–0.34 wt.% of free sterols and 0.55–0.71 wt.% of
sis. A similar methodology was described previously, but this new
sterol esters were found. Subsequently, the same authors [104]
one has shorter analytical time and the sample size is only 1/5 of
improved the LC–GC method for the analysis of free and esterified
the previous one. The methodology was employed for analyzing 6
sterols in five different kinds of biodiesel from rapeseed, soybean,
biodiesel samples from distinct feedstocks (rapeseed, peanut, corn,
sunflower, high-oleic sunflower, and used frying oil. The analyses
cottonseed, sesame and soybean oil). The phosphorus content of
were carried out without saponification, but the free sterols were
these six samples was less than 5 mg kg−1 . The review of Korn et
silylated with MSTFA.
al. [105] shows the use of flame emission spectrometry (FAES) for
the determination of Na and K in biodiesel. Recently, de Jesus et
6.4. Metals and metalloids content al. [111] have also developed a FAAS method to determine Na and
K in biodiesel that employs a water-in-oil microemulsion as sam-
Metals and metalloids are also important issues in biodiesel ple preparation. Such method showed detection limits of 0.1 and
quality control, since high contents can cause environmental prob- 0.06 ␮g g−1 for Na and K, which are two times better than those
lems or damage the engines. ASTM D 6751 and EN 14214 standards obtained according to EN 14108 and EN 14109 norms. Moreover,
of biodiesel limit to 5.0 mg kg−1 the amount of Na and K. Indi- microemulsions procedure has several advantages: high stability
rectly, these contents are also restricted through the sulphated ash and easy handle of sample and standard microemulsions; and,
value. On the other hand, the quantity of phosphorous is limited no need of using organometallic standards and carcinogenic sol-
to 10.0 mg kg−1 , Ca and Mg to 5.0 mg kg−1 , and S to 10.0 mg kg−1 , vents.
according to EN 14214 (Table 1). Also, there are some works Finally, Castilho and Stradiotto [112] developed a poten-
described in the literature, which employ mainly atomic absorp- tiometric method to determine K ions in biodiesel, using a
tion methodologies for the determination of metals and metalloids nickel hexacyanoferrate-modified electrode. Such method allowed
in biodiesel. In fact, there is a recent review [105] on the atomic the determination of K content in the concentration range of
spectrometric methods to determine metals and metalloids in fuels 4.0 × 10−5 to 1.0 × 10−2 mol L−1 , showing a detection limit of
like biodiesel. 1.9 × 10−5 mol L−1 . The obtained results were similar to those
602 M.R. Monteiro et al. / Talanta 77 (2008) 593–605

acquired by flame photometry (good reproducibility and low stan- the decomposition of esters at 259, 317, 431, and 516 ◦ C. Addition-
dard deviations). ally, there was established the calorific capacity of castor biodiesel
(1.855–2.179 J/g K) in the 55–125 ◦ C range of temperature. Regard-
6.5. Biodiesel oxidation ing physicochemical analysis, castor biodiesel shows higher flash
point and viscosity than diesel. The high flash point ensures more
The presence of light, high temperature, metal, the mate- security in the handling and storage and the high viscosity can be
rial of the container, and other extraneous materials can affect corrected through the use of blends. Recently, thermal behavior of
the quality of biodiesel. Biodiesel oxidation leads to a variety of babassu biodiesel was investigated [115]. TG and DTA curves were
species including shorter-chain fatty acids and aldehydes as well obtained in air and nitrogen. Babassu biodiesel was stable up to
as higher-molecular-weight species through oxidative polymeriza- 52 ◦ C in air and 60 ◦ C in nitrogen. DTA curves obtained in air showed
tion. Besides that, the fatty acid profile of some feedstocks for a high number of decomposition steps.
biodiesel production can also affect the oxidative stability of this
fuel. Therefore, this parameter is one of the major issues affecting 7. Physicochemical parameters and biodiesel chemical
the use of biodiesel [15,113]. In fact, Bouaid et al. [114] propose that composition
the resistance to oxidative degradation during storage is an essen-
tial issue for the successful development and viability of alternative Several physical properties are limited by ASTM D 6751 and
fuels such as biodiesel. EN 14214 standards in order to ensure the biodiesel fuel quality
The Rancimat and AOCS Oil Stability Index methods are used for (Table 1). Recently, two papers [10,15] have discussed intensely such
determining the oxidative stability of biodiesel. Knothe [15] pro- parameters. Adding to the contribution of these papers, we intend
poses that, with the inclusion of specification for this measurement to describe only the works that correlate physical parameters with
in EN 14214, the iodine value will be not necessary anymore. biodiesel chemical composition. Flash point, kinematic viscosity,
One example of biodiesel oxidation study is described in the ash content, carbon residue, and acid number are the main proper-
paper of Bouaid et al. [114]. They investigated the oxidation stabil- ties that could be associated with biodiesel composition. In fact,
ity of three biodiesel during the storage. Distinct types of biodiesel flash point strictly corresponds to the amount of methanol and
obtained from sunflower, Brassica carinata, and used frying oils the viscosity correlates with the content of unreacted triglycerides,
were stored in white and amber glass containers at room tem- or with existing undesirable materials such as crude vegetable oil
perature for a period of 30-months. After regular intervals, some in biodiesel [10,116]. Therefore, the viscosity depends on the fatty
physicochemical parameters such as acid value, peroxide value, acid composition of the oil/fat from which biodiesel is made, as
viscosity, iodine value and insoluble impurities were measured. well as on the extent of oxidation and polymerization of biodiesel
Results showed that the iodine value decreased with increasing [3]. Moreover, kinematic viscosity is useful for monitoring the fuel
storage time, but the other parameters increased through the stor- quality of biodiesel during storage, and the ash content of biodiesel
age. Therefore, all kinds of biodiesel were very stable because the indicates the residual Na or K from the catalyst. According to Mit-
increase of the three parameters was not fast. However, there is telbach [10], the carbon residue is the most important indicator
deterioration of the biodiesel after 12 months of storage and the for the quality of biodiesel since it corresponds strictly to the con-
specification limits of the parameters studied was exceeded after tent of glycerides, free fatty acids, soaps, remaining catalysts, and
this period. Besides that, samples exposed to daylight degrade other impurities. However, Mahajan et al. [117] state that one of
faster than the other fuels. Another conclusion is that the oxida- the most important features of biodiesel is the acid number, which
tive stability is more strongly influenced by the presence of small represents almost exclusively the fatty acid content.
amounts of more highly unsaturated fatty acid compounds than by In the literature, there is a work that describes a method to pre-
increasing quantities thereof. The results obtained suggested that dict the biodiesel viscosity from the knowledge of its fatty acid
it is necessary to limit access to oxygen and exposure to light and composition [118]. The method is reliable for fatty acid methyl and
moisture in order to obtain a highly stable biodiesel. ethyl esters and it was checked in methyl biodiesel from canola,
coconut, palm, peanut, and soy oil. It was identified that the vis-
6.6. Biodiesel thermal stability cosity, which is the most significant property to affect the biodiesel
as a fuel, reduces with the increase in unsaturation and it is also
In the last few years, thermal analyses (thermogravimetry, TG; affected by small amounts of glycerides. Moreover, the viscosities
differential scanning calorimetry, DSC; differential thermal analy- of saturated ethyl esters (C8 –C18 ) were slightly higher than those
sis, DTA) have become very important for supplying data that can be for the correspond methyl esters. Finally, the method was employed
useful, for example, for the establishment of thermal stability [14]. to predict the viscosities of 15 biodiesel samples. A 100% difference
Such techniques have also been employed for biodiesel characteri- in viscosity range was observed; the rapeseed methyl biodiesel had
zation. Dantas et al. [8] described the characterization of methyl the highest predicted viscosity (4.72 mPA s) and coconut biodiesel
and ethyl corn biodiesel. They employed TG to verify the influ- had the lowest (2.25 mPA s). Subsequently, Tat and Van Gerpen [3]
ence of the heating rate on the biodiesel thermogravimetric profile. described kinematic viscosity data of biodiesel and its blends (B20,
Moreover, physicochemical analysis was used to demonstrate that B50, and B75) with No. 1 and 2 diesel fuels from −20 to 100 ◦ C.
both biodiesels meet the specifications of the ANP standards, as The measurements were carried out according to ASTM D 445-88.
well as GC–FID, 1 H NMR, and FT-IR were utilized to monitor the The results showed that viscosity quickly increases as the tem-
transesterification reaction. After that, the physicochemical and perature decreases, and that biodiesel and its blends demonstrate
thermoanalytical (TG and DSC) characterization of the biodiesel temperature-dependent behavior similar to diesel, despite the fact
obtained from castor oil was described [14]. The volatilization tem- that viscosity of biodiesel is higher. Moreover, the viscosity differ-
peratures of this biodiesel are very close to those of fossil diesel. ence among the blends with No. 2 diesel was less than those of No. 1
The TG curve of castor biodiesel presented two stages of thermal diesel/biodiesel blends, since the former diesel has viscosity closer
decomposition at 150–334 ◦ C and 334–513 ◦ C with mass losses of to biodiesel. Finally, the authors established a blending equation
97 and 3%, which were related to volatilization and/or decomposi- that allows the estimation of the kinematic viscosity as a function
tion of methyl esters. On the other hand, the calorimetric curve of of the biodiesel fraction. Recently, Froehner et al. [119] established
castor biodiesel showed four exothermic transitions attributed to a method that relates density and ethyl esters content in biodiesel.
 M.R. Monteiro et al. / Talanta 77 (2008) 593–605 603

This method is quick and simple, allowing the determination of Table 2

Analytical methods described for biodiesel analyses
ester content by a single density measurement.
Also, several parameters were used for the characterization of Compounds Methods References
biodiesel from 100% canola oil, green seed canola oil as well as 1, 3, 4, 5 TLC–FID [17]
processed and unprocessed waste vegetable cooking oils [120]. 1, 2, 3 TLC [18]
According to standard procedures, the densities, viscosities, iodine 3, 4, 5, 11 GC–FID [19]
values, acid numbers, cloud points, pour points, heats of com- 1, 2 GC–FID [20–22]
1, 3, 4, 5 HPLC–DD [23]
bustion, lubricity properties, and thermal properties of the four
1, 3, 4, 5 HPLC–UV, –ELSD, –MS [24]
biodiesel were evaluated. Besides that, the fatty esters and lipid 1, 3, 4, 5 GPC [37]
compositions were determined by GC–FID and HPLC–ELSD meth- 1, 3, 4, 5, 6, 7 SEC–RID, –viscometer [38]
1
ods, respectively. GC analyses were carried out with a DB-FFAP 1, 3 H NMR [25]
1
2, 3 H NMR [27]
column and HPLC analyses employed a gel permeation liquid chro- 1
1 H NMR [28]
matography column with tetrahydrofuran as mobile phase. The 2 1
H NMR [34]
authors concluded that all the four types of oils could be employed 1, 3, 4, 5 13
C NMR [26]
to produce biodiesel. However, 100% canola oil, and green seed 1, 2, 3, 7 NIR [29,30]
canola oil are more suitable since they afford a biodiesel that has 1, 3, 4, 5 ATR–FT-IR [37]
2 FT-IR [31,36]
similar physicochemical properties to diesel. However, green seed
1, 3 FT-IR [32]
canola oil afforded a biodiesel with a low lubricity number. So, it 1 FT-IR [35]
is not appropriate to be used as an additive until further modi- 2 FT-Raman [33,34]
fications. Therefore, the biodiesel from canola oil is the best as a 1, 3, 4, 5, 6 Titration (hydroxyl content) [39]
1, 3, 4, 5, 6 Refractive index measurements [39]
fuel or additive. Later, a similar work was described [11], in which
– Viscometry [7]
the relationship between biodiesel fuel properties and its fatty 1, 3, 4, 5 GC–MS [43]
acid alkyl esters composition was investigated. It was established 1 GC–FID [44]
that structural features such as chain length, degree of unsatu- 1 GC–FID [45]
ration, and branch of the chain, influence the physical and fuel 1 GC–FID [46]
1 GC–FID [47]
properties of biodiesel, mainly cetane number, heat of combustion,
1, 3, 4, 5 GC–FID [48]
melting point, oxidative stability, viscosity, and lubricity. Usually, 1 GC–FID [40]
cetane number, heat of combustion, melting point, and viscosity 1 GC–FID [49]
increase with the increasing chain length and they decrease with 1 GC–FID [50]
the increasing unsaturation. 1 Viscometry [51]
2, 3, 4, 5, 6 GPC (HPSEC) [52]
Subsequently, Imahara et al. [121] developed a model to predict 1, 3, 4, 5, 6 GPC–RID [53]
the cloud point of biodiesel, which is related only to the amount 2, 3, 4, 5, 8, 9, 10, 11 HPLC–ELSD [54]
of saturated methyl esters. They stated that the model permits to 1, 3, 4, 5 HPLC–MS [55]
estimate cloud point of biodiesel made from several oil/fats feed- 1, 3 HPLC–ELSD, –UV [42]
1 HPLC–UV –RID [56]
stocks, thus being a useful tool to determine optimized fatty acid
2, 3, 4, 5, 12 HPLC and GC [57]
methyl ester composition. 1, 3, 4, 5, 11 RP–HPLC–UV [58]
Finally, Fernando et al. [16] studied the relationship between the 1 LC–MS [59]
amount of unconverted triglycerides (or a low content of mono- 1, 6, 11 SFC [60]
alkyl esters) in biodiesel and some biodiesel specifications (flash 1 NIR [61]
22 ATR–FT-IR [62]
point, water and sediment, kinematic viscosity, sulphur content, 22 Fiber sensor [63]
sulphated ash, copper strip corrosion, cetane number, cloud point, 1 NIR, 1 H NMR [41]
carbon residue, acid number, free and total glycerin, phosphorous 1 FT-IR [64]
content, and distillation temperature). According to standard test 1 ATR–FT-IR and NIR [65]
1 FT-IR and NIR [66]
methods, they evaluated B100 and mixtures of biodiesel and soy-
22 NIR and Raman [67]
bean oil (95, 90, 85, 80, and 75%). The results showed that the 2 FT-IR [68]
increase of oil content affects mainly the total glycerin, which fail- 1 FT-IR [69]
ures in acquires its specification. Besides that, there was increase 1 UV–Vis [70]
of viscosity, carbon residue, and flash point as well as decrease of 1 Spectroscopy [71]
1 Hydrometry [72]
cetane number. In spite of the increasing of unconverted triglyc- 22 SFS [73]
erides has affected biodiesel properties, flash point, water and 1 14
C-based method [74]
sediment, sulphur content, sulphated ash, copper strip corrosion, 13 Enzymatic [77]
cloud point, acid number, free glycerin, phosphorous content, and 6, 13 Enzymatic, spectrometric –a
6 GC–FID [78]
distillation temperature agreed to the ASTM D 6751 specifications.
3, 4, 5 GC–MS [79]
6 GC–FID, –MS [80]
8. Final considerations 3, 4, 5, 6 GC–FID [75]
6, 7 GC–FID, –MS [81]
The biodiesel quality control is ensured to the limitation of some 6, 13 GC–FID [83,84–86]
3, 4, 5 HTGC–MS, HPLC–ELSD [91]
contaminants and minor components as well as to the monitoring 4, 5 GC–MS, FT–IR [87]
of transesterification reaction and oxidation process. Several ana- 6 HPLC–PAD [88]
lytical methodologies were described for such control (Table 2). 6 HPLC, titration [89]
The transesterification monitoring is carried out mainly by spec- 6 HPLC–RID [90]
3, 4, 5 LC–GC (HPLC–DAD, GC–FID) [92]
troscopy methods, such as NMR or IR. The same techniques are
3, 4, 5, 6, 7, 12 ESI–MS [93]
employed in blends determinations, which are performed mainly 6 UV [76]
by IR techniques. In contrast, chromatography methods, mainly GC 6 EC–DAD [2]
ones, are commonly used for the quantification of fatty methyl
604 M.R. Monteiro et al. / Talanta 77 (2008) 593–605

Table 2 (Continued ) [12] European Committee for Standardization, EN 14214.

[13] American Society of Testing Materials, ASTM D 6751.
Compounds Methods References [14] M.M. Conceição, R.A. Candeia, F.C. Silva, A.F. Bezerra, V.J. Fernandes Jr., A.G.
14 Karl–Fischer titration [94] Souza, Renew. Sustain. Energy Rev. 11 (2007) 964.
7, 14 NIR [95,96] [15] G. Knothe, J. Am. Oil Chem. Soc. 83 (2006) 823.
[16] S. Fernando, P. Karra, R. Hernandez, S.K. Jha, Energy 32 (2007) 844.
14 MS [97]
[17] B. Freedman, E.H. Pryde, W.F. Kwolek, J. Am. Oil Chem. Soc. 61 (1984) 1215.
7 UV [98]
[18] J. Cvengros, Z. Cvengrosova, C. Hoka, Petrol. Coal 44 (2002) 67.
7 GC–FID [99]
[19] B. Freedman, W.F. Kwolek, E.H. Pryde, J. Am. Oil Chem. Soc. 63 (1986) 1370.
7 UV [100] [20] J. Cvengros, Z. Cvengrosova, J. Am. Oil Chem. Soc. 71 (1994) 1349.
7 GC–FID [101] [21] Z. Cvengrosova, J. Cvengros, M. Hronec, Petrol. Coal 39 (1997) 36.
15 GC–FID [102] [22] Z. Cvengrosova, J. Cvengros, M. Hronec, Petrol. Coal 40 (1998) 97.
16 GC–FID [103] [23] B. Trathnigg, M. Mittelbach, J. Liquid Chromatogr. 13 (1990) 95.
16, 17 LC–GC [103,104] [24] M. Holcapek, P. Jandera, J. Fischer, B. Prokes, J. Chromatogr. A 858 (1999) 13.
18, 19, 20, 21 ICP-OES [106,107] [25] G. Gelbard, O. Bres, R.M. Vargas, F. Vielfaure, U.F. Schuchardt, J. Am. Oil Chem.
– ICP-OES [108] Soc. 72 (1995) 1239.
19, 20 FAAS [109] [26] T. Dimmig, W. Radig, C. Knoll, T. Dittmar, Chemische Technik (Chem. Tech.) 51
21 Spectrophotometry [110] (1999) 326.
19, 20 FAAS [111] [27] P.R. Costa Neto, M.S.B. Caro, L.M. Mazzuco, M.G. Nascimento, J. Am. Oil Chem.
20 Potentiometry [112] Soc. 81 (2004) 1111.
[28] M. Morgenstern, J. Cline, S. Meyer, S. Cataldo, Energy Fuels 20 (2006) 1350.
19, 20 FAES [105]
[29] G. Knothe, J. Am. Oil Chem. Soc. 76 (1999) 795.
2 Density [119]
[30] G. Knothe, J. Am. Oil Chem. Soc. 77 (2000) 489.
a
BQP-02 kit from Sigma–Aldrich. 1 = Fatty acid methyl esters; 2 = Fatty acid [31] G.F. Zagonel, P. Peralta-Zamora, L.P. Ramos, Talanta 63 (2004) 1021.
ethyl esters; 3 = Triglycerides; 4 = Diglycerides; 5 = Monoglycerides; 6 = Glycerol; [32] N.G. Siatis, A.C. Kimbaris, C.S. Pappas, P.A. Tarantilis, M.G. Polissiou, J. Am. Oil
7 = Methanol; 8 = Fatty isopropyl ester; 9 = Fatty 2-butyl ester; 10 = Fatty isobutyl Chem. Soc. 83 (2006) 53.
[33] G.F. Ghesti, J.L. de Macedo, V.S. Braga, A.T.C.P. de Souza, V.C.I. Parente, E.S.
ester; 11 = Fatty acids; 12 = Ethanol; 13 = Total glycerol; 14 = Water; 15 = Total sterols;
Figueredo, I.S. Resck, J.A. Dias, S.C.L. Dias, J. Am. Oil Chem. Soc. 83 (2006) 597.
16 = Free sterol; 17 = Sterol esters; 18 = Ca, Cl, Mg; 19 = Na; 20 = K; 21 = P; 22 = Raw oil.
[34] G.F. Ghesti, J.L. de Macedo, I.S. Resck, J.A. Dias, S.C.L. Dias, Energy Fuels 21
(2007) 2475.
[35] J.R. Barone, 233rd ACS National Meeting, paper AGRO-153, Chicago, March
esters, glycerol and their acyl esters. The amount of water, methanol 25–29, 2007.
and sterols are also established by spectroscopy or chromatography [36] M.G. Trevisan, C.M. Garcia, U. Schuchardt, R.J. Poppi, Talanta 74 (2008) 971.
methodologies, and metals and metalloids mainly through atomic [37] M.A. Dubé, S. Zheng, D.D. McLean, K. Morris, J. Am. Oil Chem. Soc. 81 (2004)
599.
spectrometry.
[38] G. Arzamendi, E. Arguinarena, I. Campo, L.M. Gandiá, Chem. Eng. J. 122 (2006)
In general, classes of compounds are analyzed, not individ- 31.
ual species. Such data are enough to meet the requirements of [39] W. Xie, H. Li, J. Am. Oil Chem. Soc. 83 (2006) 869.
[40] R. Wawrzyniak, W. Wasiak, M. Frackowiak, Chem. Pap. 59 (2005) 449.
biodiesel standards. However, most chromatography and spectro-
[41] G. Knothe, J. Am. Oil Chem. Soc. 78 (2001) 1025.
scopic methods were developed for methyl esters analysis, and [42] T.A. Foglia, K.C. Jones, J.G. Phillips, Chromatographia 62 (2005) 115.
thus they should be modified to the characterization of higher [43] C. Mariani, P. Bondioli, S. Venturini, E. Fedeli, Riv. Ital. Sostanze Grasse 68
esters, such as ethyl esters, which will be increasingly used in (1991) 549.
[44] P. Bondioli, A. Lanzani, E. Fedeli, M. Sala, S. Veronese, Riv. Ital. Sostanze Grasse
the future. Probably, this will be an important challenge for the 71 (1994) 287.
biodiesel analysis field. On the other hand, methods that corre- [45] P. Bondioli, L. Della Bella, Riv. Ital. Sostanze Grasse 80 (2003) 173.
late some easily measured properties with biodiesel quality can [46] M. Wu, G. Wu, Y. Han, P. Zhang, Zhongguo Youzhi 28 (2003) 65.
[47] D. Guo, M. Sheng, S. Luo, Zhongguo Youzhi 29 (2004) 44.
be especially useful. Some of them were recently developed and [48] C. Li, Z. Tang, H. Yang, Fenxi Ceshi Xuebao 24 (2005) 66.
they were described in the last section of this paper. Therefore, the [49] X. Liu, C. Liu, S. Wu, F. Zeng, Q. Hu, Huaxue Yanjiu Yu Yingwong 18 (2006)
development of alternative and practical analytical methods, which 591.
[50] S. Schober, I. Seidl, M. Mittelbach, Eur. J. Lipid Sci. Technol. 108 (2006)
can be used in field analysis along the production process, handling 309.
and storage, will be important issues in the progress of biodiesel [51] P. de Filippis, C. Giavarini, M. Scarsella, M. Sorrentino, J. Am. Oil Chem. Soc. 72
analysis. (1995) 1399.
[52] R. Fillieres, B.B. Mlayah, M. Delmas, J. Am. Oil Chem. Soc. 72 (1995) 427.
[53] D. Darnoko, M. Cheryan, E.G. Perkins, J. Liquid Chromatogr. Relat. Technol. 23
Acknowledgements (2000) 2327.
[54] T.A. Foglia, K.C. Jones, J. Liquid Chromatogr. Relat. Technol. 20 (1997) 1829.
[55] M. Holcapek, P. Jandera, Adv. Mass Spectrom. 15 (2001) 583.
The authors are grateful to Centro de Caracterização e Desen-
[56] M. Kaminski, E. Gilgenast, A. Przyjazny, G. Romanik, J. Chromatogr. A 1122
volvimento de Materiais (CCDM,UFSCar/UNESP), Fundação de (2006) 153.
Amparo à Pesquisa do Estado de São Paulo (FAPESP), Conselho [57] H. Gecol, E. Ergican, S.R. Hiibel, 61st Northwest Regional Meeting of the ACS,
paper RE06-195, Reno, June 25–28, 2006.
Nacional de Desenvolvimento Científico e Tecnológico (CNPq),
[58] G. di Nicola, M. Pacetti, F. Polonara, G. Santori, R. Stryjek, J. Chromatogr. A 1190
and Financiadora de Estudos e Projetos (FINEP) for the financial (2008) 120.
support. [59] J. Huang, R. Chen, D. Cavagnino, E. Long, LC–GC N. Am. Suppl. (2007) 54.
[60] J. Cole, J. Lefler, R. Chen, LC–GC N. Am. Suppl. (2008) 40.
[61] P. Baptista, P. Felizardo, J.C. Menezes, M.J.N. Correia, Anal. Chim. Acta 607
References (2008) 153.
[62] I.P. Soares, T.F. Rezende, R.C. Silva, E.V.R. Castro, I.C.P. Fortes, Energy Fuels 22
[1] E. Stauffer, D. Byron, J. Forensic Sci. 52 (2007) 371. (2008) 2079.
[2] L.C. Gonçalves Filho, G.A. Micke, J. Chromatogr. A 1154 (2007) 477. [63] R. Falate, K. Nike, P.R. da Costa Neto, E. Cação Jr., M. Muller, H.J. Kalinowski, J.L.
[3] M.E. Tan, J.H. Van Gerpen, J. Am. Oil Chem. Soc. 76 (1999) 1511. Fabris, Quim. Nova 30 (2007) 1677.
[4] A.C. Pinto, L.L.N. Guarieiro, M.J.C. Rezende, N.M. Ribeiro, E.A. Torres, W.A. [64] A. Birova, E. Svajdlenka, J. Cvengros, V. Dostalikova, Eur. J. Lipid Sci. Technol.
Lopes, P.A.P. Pereira, J.B. de Andrade, J. Braz. Chem. Soc. 16 (2005) 1313. 104 (2002) 271.
[5] J.M. Encinar, J.F. González, A.R. Reinares, Fuel Process. Technol. 88 (2007) 513. [65] J.S. Oliveira, R. Montalvão, L. Daher, P.A.Z. Suarez, J.C. Rubim, Talanta 69 (2006)
[6] G. Knothe, Trans. ASAE 44 (2001) 193. 1278.
[7] N. Ellis, F. Guan, T. Chen, C. Poon, Chem. Eng. J. 138 (2008) 200. [66] M.F. Pimentel, G.M.G.S. Ribeiro, R.S. da Cruz, L. Stragevitch, J.G.A. Pacheco Filho,
[8] M.B. Dantas, A.A.F. Almeida, M.M. Conceição, V.J. Fernandes Jr., I.M.G. Santos, L.S.G. Teixeira, Microchem. J. 82 (2006) 201.
F.C. Silva, L.E.B. Soledade, A.G. Souza, J. Therm. Anal. Calorim. 87 (2007) 847. [67] F.C.C. Oliveira, C.R.R. Brandão, H.F. Ramalho, L.A.F. da Costa, P.A.Z. Suarez, J.C.
[9] J.R.O. Lima, R.D. da Silva, C.C.M. da Silva, L.S.S. dos Santos, J.R. dos Santos Jr., Rubim, Anal. Chim. Acta 587 (2007) 194.
E.M. Moura, C.V.R. de Moura, Quim. Nova 30 (2007) 600. [68] M.A. Aliske, G.F. Zagonel, B.J. Costa, W. Veiga, C.K. Saul, Fuel 86 (2007) 1461.
[10] M. Mittelbach, Bioresour. Technol. 56 (1996) 7. [69] L.L.N. Guarieiro, A.C. Pinto, P.F. de Aguiar, N.M. Ribeiro, Quim. Nova 31 (2008)
[11] G. Knothe, Fuel Process. Technol. 86 (2005) 1059. 421.
 M.R. Monteiro et al. / Talanta 77 (2008) 593–605 605

[70] A. Zawadzki, D. Shrestha, B. He, Trans. ASABE 50 (2007) 1349. [96] P. Felizardo, P. Baptista, M.S. Uva, J.C. Menezes, M.J.N. Correia, J. Near Infrared
[71] G.S.R. Sastry, A.S.R.K. Murthy, P.R. Prasad, Energy Source Part A 28 (2006) Spectrosc. 15 (2007) 97.
1337. [97] M.S. Todd, T. Peters, D.J. LeCaptain, 233rd ACS National Meeting, paper CHED-
[72] J. Duban, R. Turner, 2006 ASAE Annual International Meeting, paper 06-6239, 299, Chicago, March 25–29, 2007.
Portland, July 9–12, 2006. [98] F. Fang, H. Zeng, Jingxi Huagong Zhongjianti 35 (2005) 66.
[73] C.N.C. Corgozinho, V.M.D. Pasa, P.J.S. Barbeira, Talanta 76 (2008) 479. [99] C. Li, H. Yang, L. Wang, S. Tian, Sepu 24 (2006) 524.
[74] C.M. Reddy, J.A. Demello, C.A. Carmichael, E.E. Peacock, L. Xu, J.S. Arey, Environ. [100] A.R.T.S. Araujo, M.L.M.F.S. Saraiva, J.L.F.C. Lima, M.G.A. Korn, Anal. Chim. Acta
Sci. Technol. 42 (2008) 2476. 613 (2008) 177.
[75] C. Plank, E. Lorbeer, J. Chromatogr. A 697 (1995) 461. [101] M.C. Paraschivescu, E.G. Alley, W.T. French, R. Hernandez, K. Armbrust, Biore-
[76] P. Bondioli, L. Della Bella, Eur. J. Lipid Sci. Technol. 107 (2005) 153. sour. Technol. 99 (2008) 5901.
[77] J. Bailer, K. de Hueber, Fresenius’ J. Anal. Chem. 340 (1991) 186. [102] C. Plank, E. Lorbeer, J. High Res. Chromatogr. 16 (1993) 483.
[78] P. Bondioli, C. Mariani, A. Lanzani, E. Fedeli, S. Veronese, Riv. Ital. Sostanze [103] C. Plank, E. Lorbeer, Fett Wiss. Technol. 96 (1994) 379.
Grasse 69 (1992) 7. [104] C. Plank, E. Lorbeer, J. Chromatogr. A 683 (1994) 95.
[79] C. Plank, E. Lorbeer, J. High Res. Chromatogr. 15 (1992) 609. [105] M.G.A. Korn, D.S.S. dos Santos, B. Welz, M.G.R. Vale, A.P. Teixeira, D.C. Lima,
[80] M. Mittelbach, Chromatographia 37 (1993) 623. S.L.C. Ferreira, Talanta 73 (2007) 1.
[81] M. Mittelbach, G. Roth, A. Bergmann, Chromatographia 42 (1996) 431. [106] M. Edlund, H. Visser, P. Heitland, J. Anal. At. Spectrom. 17 (2002) 232.
[82] T. Ruppel, G. Hall, LC–GC N. Am. Suppl. (2007) 53. [107] E.J. dos Santos, A.B. Hermann, E.S. Chaves, W.W.D. Vechiatto, A.C. Schoem-
[83] J.D. McCurry, C. Wang, LC–GC N. Am. Suppl. (2007) 48. berger, V.L.A. Frescura, A.J. Curtius, J. Anal. At. Spectrom. 22 (2007) 1300.
[84] K. Kelly, S. Countryman, N. Nguyen, LC–GC N. Am. Suppl. (2007) 60. [108] G.D. Woods, F.I. Fryer, Anal. Bioanal. Chem. 389 (2007) 753.
[85] N. Nguyen, K. Kelly, S. Countryman, LC–GC N. Am. Suppl. (2008) 36. [109] R. Nowka, GIT Labor-Fachzeitschrift 47 (2003) 518.
[86] B. Burger, G. Stidsen, LC–GC N. Am. Suppl. (2008) 38. [110] M.X. Wu, G.Y. Wu, J. Wang, H. Xuan, Jiangsu Gongye Xueyuan Xuebao 16 (2004)
[87] N.S. Chong, S. Ramamoorthy, C. Ashford, A. Buerstetta, K. Donthula, B.G. Ooi, 23.
233rd ACS National Meeting, paper CELL-080, Chicago, March 25–29, 2007. [111] A. de Jesus, M.M. Silva, M.G.R. Vale, Talanta 74 (2008) 1378.
[88] P. Lozano, N. Chirat, J. Graille, D. Pioch, Fresenius’ J. Anal. Chem. 354 (1996) [112] M.S. Castilho, N.R. Stradiotto, Talanta 74 (2008) 1630.
319. [113] G. Knothe, Eur. J. Lipid Sci. Technol. 108 (2006) 493.
[89] M. Sala, P. Bondioli, Riv. Ital. Sostanze Grasse 75 (1998) 305. [114] A. Bouaid, M. Martinez, J. Aracil, Fuel 86 (2007) 2596.
[90] M. Hajek, F. Skopal, J. Machek, Eur. J. Lipid Sci. Technol. 108 (2006) 666. [115] N.A. Santos, M.L.A. Tavares, R. Rosenhaim, F.C. Silva, V.J. Fernandes Jr., I.M.G.
[91] T.A. Foglia, K.C. Jones, A. Nunez, J.G. Phillips, M. Mittelbach, Chromatographia Santos, A.G. Souza, J. Therm. Anal. Cal. 87 (2007) 649.
60 (2004) 305. [116] P. Bondioli, C. Mariani, E. Fedeli, A.M. Gomez, S. Veronese, Riv. Ital. Sostanze
[92] M. Lechner, C. Plank, E. Lorbeer, J. High Res. Chromatogr. 20 (1997) 581. Grasse 69 (1992) 467.
[93] R.R. Catharino, H.M.S. Milagre, S.A. Saraiva, C.M. Garcia, U. Schuchardt, M.N. [117] S. Mahajan, S.K. Konar, D.G.B. Boocock, J. Am. Oil Chem. Soc. 83 (2006) 567.
Eberlin, R. Augusti, R.C.L. Pereira, M.J.R. Guimarães, G.F. de Sá, J.M.R. Caixeiro, [118] C.A.W. Allen, K.C. Watts, R.G. Ackman, M.J. Pegg, Fuel 78 (1999) 1319.
V. de Souza, Energy Fuels 21 (2007) 3698. [119] S. Froehner, J. Leithold, L.F. Lima Jr., Quim. Nova 30 (2007) 2016.
[94] B. Faas, R. Schlink, LaborPraxis 31 (2007) 34. [120] S.L. Dmytryshyn, A.K. Dalai, S.T. Chaudari, H.K. Mishra, M.J. Reaney, Bioresour.
[95] P. Felizardo, P. Baptista, J.C. Menezes, M.J.N. Correia, Anal. Chim. Acta 595 Technol. 92 (2004) 55.
(2007) 107. [121] H. Imahara, E. Minami, S. Saka, Fuel 85 (2006) 1666.