2013

5th IEEE International Workshop

on Advances in Sensors and

Interfaces IWASI

June 13-14, 2013

Bari, Italy

IEEE Catalog Number: CFP13IWI-USB ISBN: 978-1-4799-0040-4

Proceedings of the

2013

5th IEEE International

Workshop on Advances in

Sensors and Interfaces

IWASI

June 13-14, 2013

Bari, Italy

IV

Proceedings

2013 5th

IEEE International Workshop on Advances

in Sensors and Interfaces - IWASI

June 13-14, 2013

Bari, Italy

V

Copyright ©2013 by the Institute of Electrical and Electronics Engineeers, Inc.

All rights reserved

Copyright and Reprint Permission: Abstracting is permitted with credit to the source. Libraries are

permitted to photocopy beyond the limit of U.S. copyright law for private use of patrons those articles in

this volume that carry a code at the bottom of the first page, provided the per-copy fee indicated in the

code is paid through Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923.

For other copying, reprint or republication permission, write to IEEE Copyrights Manager, IEEE Operations

Center, 445 Hoes Lane, Piscataway, NJ 08854. All rights reserved. Copyright © 2013 by IEEE.

IEEE Catalog Number: CFP13IWI-USB

ISBN: 978-1-4799-0040-4

Additional copies of this publication are available from: Curran Associates, Inc. 57 Morehouse Lane, Red

Hock, NY 12571 USA. Phone: (845) 758-0400, Fax: (845) 758-2633; E-mail: curran@proceedings.com.

©2013 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this

material for advertising or promotional purposes or for creating new collective works for resale or

redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must

be obtained from the IEEE.

A Novel Electrochemical Method for Olive Oil

Acidity Determination

Marco Grossi, Bruno Riccò

Department of Electrical Energy and Information

Engineering “Guglielmo Marconi” - DEI

University of Bologna

Bologna, Italy

marco.grossi8@unibo.it; bruno.ricco@unibo.it

Giuseppe Di Lecce, Tullia Gallina Toschi

Department of Agricultural

and Food Sciences - DISTAL

University of Bologna

Cesena, Italy

leccegius@hotmail.com; tullia.gallinatoschi@unibo.it

Abstract—Olive oil is a very important product due to its

nutritional value, sensory and antioxidant properties. National

and international regulations define basic quality parameters

(such as acidity and peroxide index) to discriminate high quality

oils (extra virgin olive oil) from those of lower quality.

The procedures to measure acidity and peroxide index are

manual titration techniques requiring a laboratory environment

and trained personnel.

In this paper a novel technique, based on electrical impedance

spectroscopy, is presented to be suitable for the realization of a

simple, low-cost, mobile instrument allowing such measurements

to be carried out everywhere and by anybody, with substantial

improvements of the current quality control of oil producers of

any dimension.

The presented method has been validated on 39 olive oil

samples, featuring different levels of acidity, peroxide index and

total phenolic content.

The results show that the emulsion conductance is the

electrical parameter best suited to discriminate the oil acidity.

Moreover, since the emulsion conductance is also affected by the

oil storage conditions (i.e. peroxide index), the presented

technique can also be proposed, in a second time, to evaluate

product ageing.

Keywords—olive oil; acidity; quality; impedance; sensor system

I. INTRODUCTION

In the Mediterranean diet, olive oil is the main source of mono-unsaturated fatty acids, and is highly appreciated as for numerous benefic biological effects [1].

Currently, about 95% of the world olive oil production is from countries of the Mediterranean area, with European countries responsible for about 80% of the production and Italy being the second largest olive oil producer after Spain [2].

Olive oil quality is affected by different factors, such as seasonal weather, sanitary state of drupes, ripening stage, storage and extraction method [3].

The most common parameters defining olive oil quality are acidity (express as oleic acid content) and peroxide index.

Olive oil acidity, defined as the amount of fatty acids no longer linked to their parent triglyceride molecules, is measured as percent of oleic acid in 100 grams of oil.

The peroxide index is an indicator of the oil primary oxidation and is expressed as milliequivalent of active oxygen

per kilogram of oil (meq O2/kg oil). If storage conditions are not adequate (essentially due to lack of protection against heat and light), oil oxidation takes place and degrades the product quality.

The European Commission Regulation No. 2568/91 and subsequent amendments impose a maximum acidity of 0.8 g oleic acid/100 g oil and a maximum peroxide index of 20 meq O2/kg oil, for extra-virgin olive oil [4].

The standard techniques to measure acidity and peroxide index in olive oil are titration procedures that can be carried out only by skilled personnel in a laboratory environment.

Innovative procedures allowing fast and in-situ determination of olive oil quality parameters are currently investigated by different research groups and several techniques have been proposed.

One of the most promising ones is based on Near-Infrared (NIR) spectroscopy, since studies have shown the possibility to estimate acidity and peroxide index by optical spectroscopy analysis of the transmission spectra in the wavenumber range 4541 to 11726 cm

-1 [5]. NIR technique has also been used to

detect adulteration of extra-virgin olive oil with lower quality products [6]. Furthermore, a recently proposed method, called spectral nephelometry, combining absorption spectroscopy and nephelometry, has been used to obtain a “fingerprint” of the oil allowing to distinguish olive oils of different geographical origins [7].

All the techniques based on light interaction with the sample are very promising since they allow fast, non destructive measurements with minimal preparation of the sample. However, the instrumentation needed is quite expensive and all the procedures must be separately calibrated for different types of olive oil (produced from olives of different varieties, different geographical origin, etc).

A different approach, microwave dielectric spectroscopy [8][9], is based on dielectric measurements allowing fast discrimination of different acidity levels. However, to achieve a good accuracy test signals of high frequencies (hundreds of MHz) are necessary, and these, in turn, require expensive instrumentation (time domain reflectometry). A pH-metric technique for olive oil acidity determination (by extraction of free fatty acids into a solvent phase) has also been proposed that does not require titration [10]. While the technique provides fast response (within a minute), the pH-

135

Fig. 1. The sensor of this work and the empirical electrical model used to interpret the results (a). The experimental set-up used to develop, optimize and validate

the technique of this work (b). The Nyquist plot for both the aqueous solution and the emulsion (c).

meter with an aqueous reference electrode is not the most economical solution and the used reagents (0.20 M triethanolamine in a 1:1 solution of water and isopropanol) can add costs for the dispose after the measure.

In this such a context, this paper presents a novel technique, based on Electrical Impedance Spectroscopy (EIS), that is shown capable of measuring olive oil acidity and suitable for the realization of a simple, low-cost, mobile instrument allowing measurements to be carried out everywhere and by anybody, with substantial improvements in the current situation of oil producers of any dimension.

EIS is a very promising technique that stimulates the sample under test (SUT) with a sinusoidal test signal in a broad range of frequencies and measures the SUT electrical response (|Z| and Arg(Z)) [11]. This technique has been used in recent years for the characterization of different food product characteristics, such as, in particular: water and lipid content in meat [12], dilution factor in apple puree [13], ripening stage of different varieties of fruits [14][15], bacterial concentration in ice-cream [16][17], raw milk [18], beer [19] and water [20].

In the application presented in this work, EIS is used with SUT of olive oil emulsified with a hydroalcoholic solution, tested with signal frequencies ranging from 20 Hz to 2 MHz. The results show that the method is capable of discriminating olive oils of different acidity levels and peroxide indexes, thus allowing to distinguish between extra virgin olive oil and oils of lower quality.

Moreover, the relatively low frequencies used in the measurements allow the realization of low cost, portable

instruments (embedded sensor systems), thus opening the road to distributed quality control procedures, affordable also by small oil producers, with obvious benefits in terms of product quality.

II. MATERIALS AND METHODS

A set of 39 olive oil samples (bought from local markets as well as taken from italian mills) have been tested to investigate the correlation between product electrical characteristics and quality parameters (acidity and peroxide index).

A. Electrochemical sensor

Virgin olive oil contains a limited amount (1-2%) of chemicals known as “minor polar components” (normally indicated as CMP), including more than 230 different molecules, that are soluble in a hydrophilic solution. The principal CMP components are: free fatty acids, monoacylglycerols, diacylglycerols, aliphatic alcohols di-and triterpene, sterols and hydrophilic antioxidants [21].

The SUT electrical characteristics have been measured in an emulsion of olive oil with a hydroalcoholic solution. Ethanol has been chosen to adjust the polarity of the solution and preferred to methanol because it is less toxic. The electrochemical sensor hosting the emulsion (Fig. 1 a) is a 50 ml tube modified to feature a couple of cap-shaped stainless steel electrodes (diameter 6 mm, spaced by 12 mm) required for the electrical characterization. Different ratios oil/aqueous solution have been tested and the results are reported in Section III. The oil sample and the solution

136

Fig. 2. The emulsion conductance Gm plotted vs. oil acidity in case of emulsion realized with 15 ml oil and 15 ml aqueous solution (a), 5 ml oil and 15 ml

aqueous solution (b), 1 ml oil and 15 ml aqueous solution (c).

are mixed and stirred to create the emulsion. The sensor is placed in a thermal incubator at 20 °C and measurements are finally done.

B. Measurement setup

The measurement set-up (Fig. 1 b), making use of benchtop instruments is used to: a) validate the characterization procedure presented in this work; b) determine the essential features of the low-cost, portable, embedded system representing the final goal of this work.

The thermal incubator Binder APT KB 53 is set to 20 °C to guarantee that all samples are tested at the same controlled temperature. The LCR meter Agilent E4980A, controlled by a PC laptop through a USB interface, measures the sample electrical parameters |Z| and Arg(Z) over the frequency range 20 Hz – 2 MHz by applying to the sensor electrodes a sinusoidal voltage of 100 mV. The PC system acquires the data from the LCR meter and store them on the hard disk. All the programs for instrumentation control have been developed with LabVIEW (National Instruments, USA).

C. Equivalent circuit model

Both the aqueous solution and the emulsion have been

electrical characterized with the measurement set-up

presented in the previous Subsection. The Nyquist diagrams,

i.e. Im(Z) plotted vs Re(Z) for the entire frequency range,

are shown in Fig. 1 (c). As can be seen, in both cases the

plot is a semi-circle with the center on the x-axis and both

Re(Z) and Im(Z) tend to 0 as ω → +∞, indicating that the

sensor-sample system is well described by an equivalent

circuit formed by a conductance in parallel with a

capacitance. Thus:

mm CjGZ

Y ω+== 1

(1)

( )22

mm

mm

CG

CjGZ

ω

ω

+

−= (2)

In these equations, Gm and Cm take into account the sample conductivity and dielectric properties, respectively. The electrical parameters have been calculated by fitting the equivalent circuit of Fig. 1 (a) using Multiple Electrochemical Impedance Spectra Parametrization (MEISP) v3.0 by Kumho Chemical Laboratories. The aqueous solution is characterized by values of Gm 0.711 ± 0.038 µS and Cm 7.396 ± 0.188 pF.

As the solution is mixed with the oil and the emulsion is created, Gm exhibits a strong increase (essentially due to the free fatty acids as well as other compounds of the sample), while Cm decreases.

Measurements on a large types of olive oil samples have shown that Cm is almost always the same, as expected since the dielectric properties of vegetable oils at low

137

TABLE I. SENSOR PERFORMANCE AS FUNCTION OF THE OIL / AQUEOUS SOLUTION RATIO.

Sensor performance Oil to aqueous solution ratio

Sensitivity (µS/%) Error in acidity estimation (%) Non linearity (%)

15 ml : 15 ml 0.21 1.528 16.26

5 ml : 15 ml 0.82 0.424 7.98

1 ml : 15 ml 1.34 0.0756 9.87

frequencies (< 10 MHz) are known not to vary significantly from one case to the other.

On the contrary, Gm turns out to be well suited for acidity determination.

Even if the current investigations have been carried out on the full spectrum range 20Hz-2MHz, data analysis has shown that, as expected from (1), Re(Y) is constant in the frequency range 40Hz-40kHz and its value can be used to estimate Gm with an error < 0.5% compared to the value calculated using the full frequency spectrum. Therefore, Gm

can be reliably calculated from measurements utilizing a single frequency, and this, of course, allows to greatly simplify the design of an embedded system for oil acidity measurements to be used in production environment.

D. Reference methods

All the tested samples have been analyzed using the reference methods as defined by the European Commission Regulation No. 2568/91. These are manual titration procedures for the determination of acidity and peroxide index in olive oil [4].

Titration is a common laboratory method of quantitative chemical analysis used to determine the unknown concentration of an identified analyte. A reagent (the titrant) is added to a solution composed of the oil sample and chemicals (the titrand) until the color of the titrand solution changes (end-point of titration). The concentration of the analyte is estimated from the amount of used titrant.

In the case of acidity determination, the titrant is sodium hydroxide NaOH 0.1 M, while the titrand is a solution composed of an alcohol-ether mix, phenolphthalein (as indicator) and the oil sample.

Instead, for the determination of peroxide index, the titrant is sodium thiosulphate Na2S2O3 and the titrand a mix of acetic acid, chloroform, saturated solution of potassium iodide KI, starch (as indicator) and the oil sample.

III. RESULTS AND DISCUSSION

The measurement has been initially optimized for acidity determination, finding the best oil – aqueous solution ratio for the purpose. Successively, the peroxide index, of interest to investigate the effects of the oil storage conditions, has been considered. Finally, a set of 39 olive oils (purchased from markets as well as taken from Italian olive oil mills) have been tested.

A. Choice of the oil/aqueous solution ratio

Three different ratios oil/aqueous solution for the emulsion have been investigated to achieve the best accuracy in acidity determination: (a) 15 ml oil and 15 ml

solution, (b) 5 ml oil and 15 ml solution, (c) 1 ml oil and 15 ml solution.

A commercial olive oil has been modified with the addition of different concentration of oleic acid to obtain samples that differ only for their acidity, while presenting the same remaining properties (peroxide index, polyphenol content, etc.). Four different test samples were created, respectively with acidity 0.35%, 0.75%, 1.25% and 2.25%. All measurements have been carried out in triplicate and both mean value and standard deviation have been calculated.

The measured values of Gm plotted vs the sample acidity

are shown in Fig. 2. As can be seen, in all cases Gm

increases with the sample acidity. The best results are

obtained for case (c), i.e. the emulsion realized mixing 1 ml

of oil with 15 ml of aqueous solution. In fact, as also pointed

out in Table 1, the sensitivity (variation of Gm in µS for 1%

variation of the sample acidity) is almost twice as large as

that of case (b) and more than six times higher than case (a).

The same holds for the error in acidity estimation, with

case (c) achieving a very low value (0.0756%), more than

15 times lower than case (a).

As for the sensor linear response, case (b) and (c) are

comparable and much better than (a). Thus, in the following, all the measurements, have been

carried out with emulsions created from 1 ml of olive oil and 15 ml of aqueous solution.

B. Effects of the storage conditions

To investigate if and how Gm can be used to investigate, and possibly monitor, the effects of olive oil storage conditions the following stress test has been carried out. A commercial extra-virgin olive oil sample has been subjected to UV radiation for different amount of time and the

Fig. 3. The emulsion conductance Gm plotted vs. UV exposure time.

138

Fig. 4. The emulsion conductance Gm plotted vs. oil acidity for a subset of olive oils featuring peroxide index < 20 (a). A scatter plot of the full set of tested oils

plotted as function of acidity and peroxide index. The samples inside the dashed area are extra-virgin olive oils (b).

conductance of the emulsion has been measured before the exposition to UV as well as after 1, 2 and 3 weeks, respectively. The results shown in Fig. 3 indicate that Gm

has a strong increase after the first week of UV exposure, while it features negligible variations successively. Under UV exposure, a strong increase of both primary and secondary oxidation products takes place [21], resulting in an increase of the emulsion conductance. In particular, Fig. 3 shows that Gm increases by 125.6%, 131.3% and 126.9% after 1, 2 and 3 weeks, respectively (compared to the value measured before the UV stress test).

C. Results from different oil blends

A set of 39 olive oils have been tested that differ substantially for the olive cultivar, acidity, peroxide index and total phenolic content. The tested oil samples feature acidity values from 0.1% to 2.25%, peroxide index from 4 to 75 meq O2/kg oil and phenolic content from 30 to 700 mg gallic acid/kg oil.

In Fig. 4 (a) Gm is plotted vs. the acidity value for the oil subset featuring peroxide index < 20 (i.e. samples in good storage conditions). These data clearly show that the emulsion conductance Gm increases with oil acidity. In particular, the calculated linear regression line equation is Gm = 1.75·acidity + 1.43 and the determination factor is R

2

= 0.748. A conductance threshold can then be set so that all samples featuring Gm < 2.75 µS are characterized by acidity < 0.8%. Similarly, all samples featuring Gm > 3.25 µS are characterized by acidity > 0.8%, while in the conductance range 2.75-3.25 µS the samples acidity is close to 0.8% (0.68% to 0.85%).

In Fig. 4 (b), showing the results obtained with the full set of tested oils, each sample is represented by a dot as a function of the acidity and peroxide index (determined by the standard procedures). All the samples inside the dashed area are extra-virgin olive oils (acidity < 0.8% and peroxide index < 20). The samples have been divided in three groups based on the measured electrical conductance: Gm < 2.75 µS ; 2.75 µS < Gm < 3.25 µS; Gm > 3.25 µS. As already mentioned, samples with high acidity feature high values of Gm . However, also samples with low acidity but high peroxide index exhibit high values of Gm.

Taking both these conditions into consideration, all extra virgin olive oil samples are correctly distinguished from those of lower quality by means of Gm measurements, except one sample that is classified as extra-virgin in spite of a peroxide index > 20.

D. Realization as an embedded system

The technique presented above has been validated by means of laboratory instrumentation. It, however, can be easily implemented as a low-cost, easy to use, portable electronic system suitable for distributed and on-line quality control, by large oil producers as well as small oil mills.

In fact, the LCR meter was used essentially to investigate the system electrical model, but for routine use it could be replaced by a data acquisition board working at a single frequency. As for the effect of temperature, they could be experimentally characterized so as to be able to scale results obtained at different temperatures to the same reference value by means of a simple model or stored look-up tables.

In addition, data storage, wireless communication, independent power supply, signal acquisition and analysis can be easily realized by means of standard embedded system technology to obtain an easy-to-use, low cost, mobile instrument.

The implementation of a portable instrument featuring wireless transmission allows rapid collection of data through a sensors network for real-time, quick quality control.

IV. CONCLUSIONS

A novel technique to measure acidity as a principal quality parameter of olive oil has been presented. The method is based on the measurement by electrical impedance spectroscopy of the electrical characteristics of an emulsion between the oil sample and an hydroalcoholic solution.

The results obtained with a significant lot of oil samples of different origin show that the electrical conductance Gm

of the emulsion can be used as a parameter to determine oil acidity. Furthermore, for samples with comparable values of acidity, Gm increases because of stressing storage conditions (resulting in higher peroxide index).

139

In practice, then, the measured values of Gm can be used to discriminate extra virgin olive oils (acidity < 0.8%, peroxide index < 20) from those of lower quality.

In this paper the presented technique has been developed and validated by means of laboratory instrumentation. It, however, can be easily implemented as a low-cost, easy to use, portable electronic system suitable for distributed and on-line quality control, by large oil producers as well as small oil mills.

ACKNOWLEDGMENT

This work has been financially supported by the CESAR Project, RIDIIT program, funded by the Ministry of Economic Development (Italy).

REFERENCES

[1] S. Tulipani, M. Martinez Huelamo, M. Rotches Ribalta, R. Estruch, E. Escribano Ferrer, C. Andres-Lacueva, M. Illan and R.M. Lamuela-Raventós, “Oil matrix effects on plasma exposure and urinary excretion of phenolic compounds from tomato sauces: Evidence from a human pilot study”. Food Chemistry, vol. 130 ,pp. 581-590, 2012.

[2] F. Carbonari and T. Sarnari, “Il mercato internazionale e nazionale dell’olio di oliva. ISMEA, pp. 1-13, 2013.

[3] R. Sacchi, L. Mannina, P. Fiordiponti, P. Barone, L. Paolillo and M. Patumi, “ Characterization of italian extra virgin olive oils using 1H-NMR spectroscopy“ Journal of Agricultural and Food Chemistry, vol. 46, no. 10, pp. 3947-3951, 1998.

[4] EC 1991 Commission Regulation (EEC) 2568/91 of July 11th 1991 on characteristics of olive oil and on the relevant methods of analysis. (1991). Official EC Journal, L248, 0001–0083.

[5] S. Armenta, S. Garrigues and M. de la Guardia, “Determination of edible oil parameters by near infrared spectrometry”, Analytica Chimica Acta, vol. 596, pp. 330-337, 2007.

[6] D. Ozdemir and B. Ozturk, “Near infrared spectroscopic determination of olive oil adulteration with sunflower and corn oil”, Journal of Food and Drug Analysis, vol. 15, no. 1, pp. 40-47, 2007.

[7] A. G. Mignani, P. R. Smith, L. Ciaccheri, A. Cimato and G. Sani, “Spectral nephelometry for making extravirgin olive oil fingerprints”, Sensors and Actuators B, vol. 90, pp. 157-162, 2003.

[8] A. Cataldo, E. Piuzzi, G. Cannazza and E. De Benedetto, “Dielectric spectroscopy of liquids through a combined approach: evaluation of the metrological performance and feasibility study on vegetable oils”, IEEE Sensors Journal, vol. 9, no. 10, pp. 1226-1233, 2009.

[9] A. Cataldo, E. Piuzzi, G. Cannazza, E. De Benedetto and L. Tarricone, “Quality and anti-adulteration control of vegetable oils through microwave dielectric spectroscopy, Measurement, vol. 43, pp. 1031-1039, 2010.

[10] Ya.I. Tur’yan, O.Yu. Berezin, I. Kuselman and A. Shenhar, “pH-metric determination of acid values in vegetable oils without titration” Journal of the American Oil Chemists Society, vol. 73, no. 3, pp. 295-301, 1996.

[11] B.-Y. Chang and S.-M. Park, “Electrochemical impedance spectroscopy”, Annual Review of Analytical Chemistry, vol. 3, pp. 207-229, 2010.

[12] M. Chanet, C. Riviere and P. Eynard, “Electric impedance spectrometry for the control of manufacturing process of comminuted meat products”, Journal of Food Engineering, vol. 42, pp. 153-159, 1999.

[13] R. Zywica, G. Pierzynowska-Korniak and J. Wojcik, “Application of food products electrical model parameters for evaluation of apple puree dilution”, Journal of Food Engineering, vol. 67, pp. 413-418, 2005.

[14] F.R. Harker and J.H. Maindonald, “Ripening of nectarine fruit: changes in the cell wall, vacuole and membranes detected using electrical impedance measurement”, Plant Physiology, vol. 106, pp. 165-171, 1994.

[15] F.R. Harker and S.H. Forbes, “Ripening and development of chilling injury in persimmon fruit: an electrical impedance study”, New Zealand Journal of Crop and Horticultural Science, vol. 25, pp. 149-157, 1997.

[16] M. Grossi, M. Lanzoni, A. Pompei, R. Lazzarini, D. Matteuzzi and B. Riccò, “Detection of microbial concentration in ice-cream using the impedance technique”, Biosensors and Bioelectronics, vol. 23, pp. 1616-1623, 2008.

[17] M. Grossi, A. Pompei, M. Lanzoni, R. Lazzarini, D. Matteuzzi and B. Riccò, “Total bacterial count in soft-frozen dairy products by impedance biosensor system”, IEEE Sensors Journal, vol. 9, no. 10, pp. 1270-1276, 2009.

[18] M. Grossi, M. Lanzoni, A. Pompei, R. Lazzarini, D. Matteuzzi and B. Riccò, “A portable biosensor system for bacterial concentration measurements in cow’s raw milk”, Proceedings of the 4th IEEE International Workshop on Advances in Sensors and Interfaces (IWASI), Savelletri di Fasano, Italy, pp. 132-137, 2011.

[19] A. Pompei, M. Grossi, M. Lanzoni, G. Perretti, R. Lazzarini, B. Riccò and D. Matteuzzi, “Feasibility of lactobacilli concentration detection in beer by automated impedance technique”, MBAA Technical Quarterly, vol. 49, no. 1, pp. 11-18, 2012.

[20] M. Grossi, R. Lazzarini, M. Lanzoni, A. Pompei, D. Matteuzzi and B. Riccò, “A portable sensor with disposable electrodes for water bacterial quality assessment”, IEEE Sensors Journal, vol. 13, no. 5, pp. 1775-1782, 2013.

[21] E. Boselli, G. Di Lecce, M. Minardi, D. Pacetti and N.G. Frega, “La spettrometria di massa nell’analisi di componenti minori polari dell’olio vergine di oliva”. Rivista Italiana delle Sostanze Grasse, vol 84, pp. 3-14, 2008.

140