**3. Direct determination**

The direct determination of metals in oils can be carried out by sample solubilization in an organic solvent, an emulsification procedure in aqueous solutions in the presence of emulsifiers such as Triton X-100 or a solid sampling strategy.

#### **3.1 Dilution with organic solvent**

The procedure of the dilution with organic solvents is an easy way to sample pretreatment before detection, but has some requirements: special devices for sample introduction e. g. for FAAS (Bettinelli et al., 1995), the addition of oxygen as an auxiliary gas in ICP-OES or ICP-MS (Costa et al., 2001). The volatile organic solvents have been directly introduced into ICPs for many years, but this can cause plasma instability, less sensitivity, less precision and high cost. Al, Cr, Cd, Cu, Fe, Mn, Ni and Pb contents of olive oil were investigated using diethyl ether, methyl isobutyl ketone (MIBK), xylene, heptane, 1,4-dioxane as solvent and N,N-hexamethylenedithiocarbamic acid, hexamethyleneammonium (HMDC-HMA) salt as a modifier by ETAAS (Karadjova et al., 1998). A transverse heated filter atomizer (THFA) was employed for the direct determination of Cd and Pb in olive oil after sample dilution with nheptane (Canario & Katskov, 2005). Moreover, Martin-Polvillo et al. (1994) and List et al. (1971) determined trace elements in edible oils based on the direct aspiration of the samples, diluted in MIBK. In another research, the mixture of 2%lecithin-cyclohexane was used to introduce the oil samples to a polarized Zeeman GFAAS (Chen et al., 1999). Van Dalen was

Metal Determinations in Olive Oil 97

pretreatment procedures in common like: wet digestion, dry ashing, extraction and dilution with organic solvent in order to eliminate hard organic matrix. In the processing large numbers of samples, flow injection analysis (FIA) systems can be preferred for sample pretreatment. The FIA system for oil analysis is frequently based on the on-line preparation of oil-in-water emulsions by using ultrasonic bath with serious drawbacks in efficient preparation of stable emulsions. By this way, more concentrated emulsions (high oil concentration) can be introduced into the plasma and thereby the LODs were improved. A limited number of researches related to metal determination in oils by FIA systems have been presented. Jimenez et al. succeeded multi-element determination in virgin olive oil by flow injection ICP-MS using with HNO3 and Triton X-100 as emulsifying agents (Jimenez et al., 2003). A magnetic-stirring micro-chamber has been developed for on-line emulsification and has been successfully employed by Anthemidis et al. to detect Ag, Al, B, Ba, Bi, Ca, Cd, Co, Cr, Cu, Fe, Ga, In, Mg, Mn, Ni, Pb, Tl and Zn in olive oil using flow injection ICP-OES (Anthemidis et al., 2005). The low concentration of analyte in the sample analysed and difficulty of obtaining stable emulsions with rich oil content were reported as the main problems. On-line emulsion preparation procedure was suggested as simpler, more effective, less time consuming, less labor intensive, less matrix interferences and less contamination risk over the other direct sample introducing procedures. The direct determination of Cu and Fe in edible oils based on the flow injection standard addition method by FAAS was performed

As mentioned above, various pretreatment procedure and detection techniques have been employed for the total determination of metals in olive oil. The researchers have dealt with metallic contents of olive oils during last few decades. As can be seen in Table 6, the

Fe 12.5\* 139.0 (Anthemidis et al., 2005; Benincasa et al., 2007; Buldini et

Cu 1.7\* 4.51 (Angioni et al., 2006; Anthemidis et al., 2005; Buldini et al.,

Ni 10.6\* 2.26 (Benincasa et al., 2007; Buldini et al., 1997; Calapaj et al.,

2011; Zeiner et al., 2005) Zn 0.6\* 4.61 (Angioni et al., 2006; Cindric et al., 2007; Lo Coco et al., 2003;

al., 1997; Calapaj et al., 1988; Cindric et al., 2007); De Leonardis et al., 2000; Llorent-Martinez et al., 2011a, 2011b; Martin-Polvillo et al., 1994; Mendil et al., 2009; Nunes et al.,

1997; Calapaj et al., 1988; Castillo et al., 1999; Cindric et al., 2007; De Leonardis et al., 2000; Galeano Diaz et al., 2006; Jimenez et al., 2003; Karadjova et al., 1998; Llorent-Martinez et al., 2011a, 2001b; Martin-Polvillo et al., 1994; Mendil et al., 2009; Nunes et al., 2011; Pehlivan et al., 2008; Zeiner et

1988; Castillo et al., 1999; Cindric et al., 2007; Nunes et al.,

Mendil et al., 2009; Nunes et al., 2011; Zeiner et al., 2005)

2011; Pehlivan et al., 2008; Zeiner et al., 2005)

without sample dilution in a previous study (Carbonell et al., 1991 ).

concentration range of total amount is given for many metals.

( µg g-1 ) (\* ng g-1) References

al., 2005)

Concentration

Minimum Maximum

Metal

also used lecithin and the organopalladium modifier solutions for the injection of the edible oils (Van Dalen, 1996).

#### **3.2 Emulsification**

Taking into account parameters such as economy, safety, environment, time, and low risk of contamination, emulsification appears beneficial over microwave assisted acid digestion. On the other hand, optimization of the particle size effect, slurry concentration and homogeneity are necessary in order to obtain good precision and recoveries with slurry techniques. In spite of optimization, complete destruction of the sample matrix in plasma and then liberation of analyte from the sample matrix are not always succeeded, causes unsatisfactory results. An alternative technique for introduction of oil sample directly into ICP is the on-line emulsification (Anthemidis et al., 2005). Direct introduction of oil samples in the form of emulsion into ICP facilitates the spray chamber and plasma torch owing to no need of extra oxygen or sophisticated desolvation device. In such a case, the use of stable emulsions with proper surfactant concentration is very important (Anthemidis et al., 2005).

Emulsification as sample preparation has been performed for the determination of trace metals in vegetable oils by ICP-OES (De Souza et al., 2005; Murillo et al., 1999), ICP-MS (Castillo et al., 1999; Jimenez et al., 2003), FAAS (List et al., 1971) and GF-AAS (Lendinez et al., 2001). Additionally, the use of microemulsion as sample preparation for vegetable oil analysis by High-Resolution Continuum Source FAAS (HR-CS FAAS) has been described by Nunes et al. (2011). The determination of Zn, Cd and Pb in vegetable oils by electrothermal vaporization in combination with ICP-MS (ETV-ICP-MS) was described in literature (Huang et al., 2001).

#### **3.3 Direct solid sampling**

Direct introduction of oil samples into the graphite furnace by solid sampling strategy is rarely used, providing an alternative methodology. Due to technical improvements in spectrophotometer and software capabilities of modern instrumentation, this method has not been entirely accepted (Sardans et al., 2010). Direct solid sampling has some advantages such as no sample dilution, satisfactory LOD levels, calibration probability with aqueous analytical solutions, simple analysis and no sample digestion or extraction. Other advantages of this method are reduced time and cost, required little amount of sample and the achievement of high sensitivity. Additionally, it reduces the risk of contamination due to the nonexistence of sample preparation and use of chemical reagents. Some disputes against the method are the difficulty of introducing small sample masses, faulty measurement of the results due to the heterogeneity of some natural samples and the limiting linear working range of AAS (Sardans et al., 2010). Despite these restrictions, direct solid sampling is a reasonable alternative for the determination of the total content of metals in oils, since it needs almost no sample preparation. A method for the direct determination of Ni and Cu in vegetable oils by GFAAS using the solid sampling strategy has been reported without sample dilution by Matos Reyes et al. (2006).

#### **3.4 Flow injection**

Various detection techniques like ETAAS, FAAS, ICP-OES, ICP-MS, voltammetry have been utilized for metal determination in oils. However, all of them have the need for sample

also used lecithin and the organopalladium modifier solutions for the injection of the edible

Taking into account parameters such as economy, safety, environment, time, and low risk of contamination, emulsification appears beneficial over microwave assisted acid digestion. On the other hand, optimization of the particle size effect, slurry concentration and homogeneity are necessary in order to obtain good precision and recoveries with slurry techniques. In spite of optimization, complete destruction of the sample matrix in plasma and then liberation of analyte from the sample matrix are not always succeeded, causes unsatisfactory results. An alternative technique for introduction of oil sample directly into ICP is the on-line emulsification (Anthemidis et al., 2005). Direct introduction of oil samples in the form of emulsion into ICP facilitates the spray chamber and plasma torch owing to no need of extra oxygen or sophisticated desolvation device. In such a case, the use of stable emulsions with proper surfactant concentration is very important (Anthemidis et al., 2005). Emulsification as sample preparation has been performed for the determination of trace metals in vegetable oils by ICP-OES (De Souza et al., 2005; Murillo et al., 1999), ICP-MS (Castillo et al., 1999; Jimenez et al., 2003), FAAS (List et al., 1971) and GF-AAS (Lendinez et al., 2001). Additionally, the use of microemulsion as sample preparation for vegetable oil analysis by High-Resolution Continuum Source FAAS (HR-CS FAAS) has been described by Nunes et al. (2011). The determination of Zn, Cd and Pb in vegetable oils by electrothermal vaporization in combination with ICP-MS (ETV-ICP-MS) was described in literature (Huang et al., 2001).

Direct introduction of oil samples into the graphite furnace by solid sampling strategy is rarely used, providing an alternative methodology. Due to technical improvements in spectrophotometer and software capabilities of modern instrumentation, this method has not been entirely accepted (Sardans et al., 2010). Direct solid sampling has some advantages such as no sample dilution, satisfactory LOD levels, calibration probability with aqueous analytical solutions, simple analysis and no sample digestion or extraction. Other advantages of this method are reduced time and cost, required little amount of sample and the achievement of high sensitivity. Additionally, it reduces the risk of contamination due to the nonexistence of sample preparation and use of chemical reagents. Some disputes against the method are the difficulty of introducing small sample masses, faulty measurement of the results due to the heterogeneity of some natural samples and the limiting linear working range of AAS (Sardans et al., 2010). Despite these restrictions, direct solid sampling is a reasonable alternative for the determination of the total content of metals in oils, since it needs almost no sample preparation. A method for the direct determination of Ni and Cu in vegetable oils by GFAAS using the solid sampling strategy has been

Various detection techniques like ETAAS, FAAS, ICP-OES, ICP-MS, voltammetry have been utilized for metal determination in oils. However, all of them have the need for sample

reported without sample dilution by Matos Reyes et al. (2006).

oils (Van Dalen, 1996).

**3.3 Direct solid sampling** 

**3.4 Flow injection** 

**3.2 Emulsification** 

pretreatment procedures in common like: wet digestion, dry ashing, extraction and dilution with organic solvent in order to eliminate hard organic matrix. In the processing large numbers of samples, flow injection analysis (FIA) systems can be preferred for sample pretreatment. The FIA system for oil analysis is frequently based on the on-line preparation of oil-in-water emulsions by using ultrasonic bath with serious drawbacks in efficient preparation of stable emulsions. By this way, more concentrated emulsions (high oil concentration) can be introduced into the plasma and thereby the LODs were improved. A limited number of researches related to metal determination in oils by FIA systems have been presented. Jimenez et al. succeeded multi-element determination in virgin olive oil by flow injection ICP-MS using with HNO3 and Triton X-100 as emulsifying agents (Jimenez et al., 2003). A magnetic-stirring micro-chamber has been developed for on-line emulsification and has been successfully employed by Anthemidis et al. to detect Ag, Al, B, Ba, Bi, Ca, Cd, Co, Cr, Cu, Fe, Ga, In, Mg, Mn, Ni, Pb, Tl and Zn in olive oil using flow injection ICP-OES (Anthemidis et al., 2005). The low concentration of analyte in the sample analysed and difficulty of obtaining stable emulsions with rich oil content were reported as the main problems. On-line emulsion preparation procedure was suggested as simpler, more effective, less time consuming, less labor intensive, less matrix interferences and less contamination risk over the other direct sample introducing procedures. The direct determination of Cu and Fe in edible oils based on the flow injection standard addition method by FAAS was performed without sample dilution in a previous study (Carbonell et al., 1991 ).


As mentioned above, various pretreatment procedure and detection techniques have been employed for the total determination of metals in olive oil. The researchers have dealt with metallic contents of olive oils during last few decades. As can be seen in Table 6, the concentration range of total amount is given for many metals.


Metal Determinations in Olive Oil 99

Fractionation was defined as "the process of classification of analyte or a group of analytes from a certain sample according to physical (e.g., size, solubility) or chemical (e.g., bonding, reactivity) properties", and speciation of an element was also defined as "distribution of an element amongst defined chemical species in a system" by Templeton et al. (2000). The physicochemical form of an element, i.e. the actual species found in exposure medium and in the different body fractions, is frequently determinant in the evaluation of its bioavailability and toxicity (Flaten, 2002). An element can be found in various species: anionic or cationic inorganic forms, inorganic compounds, complex compounds with protein, peptide etc. Some organometallic compounds are much more toxic than the ions of the corresponding inorganic compounds. Hg, Pb and Sn obey this rule, for example, methyl-Hg and inorganic Hg are both toxic, but methyl-Hg show more toxicity than other (Templeton et al., 2000). In contrast to this, in the case of As and Se, most organo-arsenicals are less toxic than inorganic As species, organic forms of Se are ordinarily less toxic than

The determination of the total amount of an element in samples cannot give adequate information for understanding its bioavailability or toxicity, that's why the fractionation and speciation of metals in oils are increasingly gaining importance. The fractionation and speciation analysis are more informative than total element determinations for all type of

In general, many works dealing with the total amount of elements in oil samples are reported, but fractionation and/or speciation analysis in vegetable oils are less common in literature. To the best of our knowledge, magnesium fractionation analysis in olive and olive oil was cited firstly in 2004. The improvement of an analytical scheme for fractionation of magnesium in olive products and also the determination of Mg amounts absorbed in stomach and intestine was achieved by Bağdat Yaşar & Güçer (2004). It was reported that 3.37-8.47% of Mg was absorbed in the stomach (ionic and polar groups) and the remaining percentage of Mg was absorbed in the intestine (molecular and complexed structures) in olive oil. As can be seen, the Mg fraction in olive oil is almost absorbable in the intestine. This study can be accepted as a preliminary step for fractionation studies and the fractionation and/or speciation approach for other elements will be described in

Various researchers deal with determination of metals in oils at trace, ultra-trace levels using spectrometric and electrometric techniques. Mentioned detection techniques may be combined with some chromatographic systems. Oils have been analyzed for different metals using atomic absorption spectrometer (FAAS and GFAAS), inductively coupled plasma optical emission spectrometer (ICP-OES), inductively coupled plasma mass spectrometer (ICP-MS). ICP techniques have become more popular since the early 1990s. Although the use of AAS (flame, graphite furnace, hydride generation and cold vapour) has declined during the same period, it is still the most widely used technique (Rose et al.,

**4. Speciation and fractionation** 

Se(IV) (Kot & Namiesnik, 2000).

samples.

the future.

2001).

**5. Detection techniques** 


Table 6. The metal levels for olive oils.

Mn 0.7\* 0.15 (Anthemidis et al., 2005; Benincasa et al., 2007; Calapaj et al.,

Pb 0.42\* 0.032 (Calapaj et al., 1988; Canario & Katskov, 2005; Castillo et al.,

Co 0.23\* 5.45 (Benincasa et al., 2007; Calapaj et al., 1988; Castillo et al., 1999;

Cd 0.6\* 0.15 (Angioni et al., 2006; Benincasa et al., 2007; Calapaj et al.,

Cr 0.012 2.00 (Anthemidis et al., 2005; Benincasa et al., 2007; Calapaj et

Ba 4.9\* 0.7 (Castillo et al., 1999; Jimenez et al., 2003; Llorent-Martinez

Al 0.030 1.11 (Anthemidis et al., 2005; Cindric et al., 2007; Jimenez et al.,

Mg 0.056 4.61 (Bağdat Yaşar & Güçer, 2004; Benincasa et al., 2007; Cindric

Ca 0.63 76.0 (Anthemidis et al., 2005; Benincasa et al., 2007; Cindric et

K 0.05 2.14 (Cindric et al., 2007; Mendil et al., 2009; Zeiner et al., 2005) Na 8.7 38.03 (Cindric et al., 2007; Mendil et al., 2009; Zeiner et al., 2005)

et al., 2007; Mendil et al., 2009; Zeiner et al., 2005)

al., 2007; Mendil et al., 2009; Zeiner et al., 2005)

V 0.005 0.46 (Castillo et al., 1999); (Llorent-Martinez et al., 2011a)

et al., 2011a)

Zeiner et al., 2005)

Ge 0.03 0.04 (Castillo et al., 1999) Zr 0.01 0.04 (Castillo et al., 1999)

Be 0.118\* 0.178\* (Benincasa et al., 2007) Sc 49.94\* 747.9\* (Benincasa et al., 2007) As 1.248\* 26.65\* (Benincasa et al., 2007) Se 1.47\* 6.78\* (Benincasa et al., 2007) Sr 1.52\* 48.9\* (Benincasa et al., 2007) Y 0.082\* 0.331\* (Benincasa et al., 2007) Sb 0.194\* 0.411\* (Benincasa et al., 2007) Sm 0.004\* 0.226\* (Benincasa et al., 2007) Eu 0.004\* 0.021\* (Benincasa et al., 2007) Gd 0.003\* 0.094\* (Benincasa et al., 2007) Sn 0.126 0.159 (Calapaj et al., 1988)

Table 6. The metal levels for olive oils.

et al., 2009; Yağan Aşç et al., 2008)

1988; Castillo et al., 1999; Cindric et al., 2007; Jimenez et al., 2003; Karadjova et al., 1998; Llorent-Martinez et al., 2011a; Mendil et al., 2009; Pehlivan et al., 2008; Zeiner et al., 2005)

1999; Jimenez et al., 2003; Llorent-Martinez et al., 2011a;

Cindric et al., 2007; Mendil et al., 2009; Zeiner et al., 2005)

1988; Canario & Katskov, 2005; Castillo et al., 1999; Mendil

al., 1988; Castillo et al., 1999; Llorent-Martinez et al., 2011a)

2003; Karadjova et al., 1998; Martin-Polvillo et al., 1994;

Mendil et al., 2009; Martin-Polvillo et al., 1994)
