Trends in Analytical Chemistry, Vol. 28, No.

1, 2009 Trends

Determination of metal content

in honey by atomic absorption
and emission spectrometries
Pawel Pohl

Honey contains a variety of metals. Major metals are primarily derived from [7]. The contribution of minerals is rela-
soil and nectar-producing plants, but consideration also needs to be given to tively low and normally accounts for 0.1–
environmental pollution or other anthropogenic sources of metals in honey, 0.2% of nectar honeys. The total mineral
especially of Cd, Cr, Cu, Fe, Ni, Pb and Zn, which may present hazards to content of honeydew honeys can exceed
human health and adversely affect the quality and the safety of honey. 1% [8]. All components are dissolved in
This review surveys the literature from the past 15 years on determination water (13.4–22.9%) to give the product its
of the metal content of honey by atomic absorption and emission spectro- distinctive color and texture (liquid or
metries. It pays particular attention to sample treatment, sample preparation semi-liquid).
and measurement techniques. It also discusses the suitability of information The main concentrations of metals in
on the composition of metals in honey for classifying according to botanical honey derive from soil. They are trans-
and geographical origin or indicating and controlling environmental pollu- ported to honey plants through the root
tion. system, pass to nectar and then to honey
ª 2008 Elsevier Ltd. All rights reserved. produced by bees [9]. Accordingly, the
composition and the content of metals in
Keywords: Atomic absorption spectrometry; Atomic emission spectrometry; Environ-
honey, particularly major and minor
mental pollution; Hazard; Honey; Major metal; Optical emission spectrometry; Sample
preparation; Sample treatment; Trace metal
metals (Ca, K, Mg, Mn, Na), is affected by
the soil composition (geographical origin)
1. Metals in honey determined by geochemical and geological
Pawel Pohl*
features (i.e. volcanic and hydrothermal
Analytical Chemistry Division,
Faculty of Chemistry, Wroclaw Honey is a sweet, viscous substance activity, regional conditions and climatic
University of Technology mainly produced by honey bees (Apis changes in the forage area of the bees)
Wybrzeze Stanislawa mellifera) from nectar of different plant [1,2,8–18]. In honeys from coastal regions
Wyspianskiego 27, 50-370 flowers and secretions from plants or or islands, the sea or the ocean is also a
Wroclaw, Poland
plant-sucking insects [1–3]. Because of its geogenic source of K and Na as a conse-
sweetness, color and flavor, honey is often quence of marine aerosol [8,19]. This non-
used as a sugar substitute, an ingredient floral source can elevate the concentra-
or a natural preservative in hundreds of tions of alkali metals by up to 10 times
manufactured foods. It is also a very that in honeys from other territories [19].
energetic food, possessing valuable nour- The floral type of honey plants, floral
ishing, healing and prophylactic properties density and composition of nectar and
[1,4–6], which all result from its chemical pollen (botanical origin) are also respon-
composition and predominant content of sible for variations in the metal content of
simple sugars [7]. Depending on the honey [1,15–25]. Accordingly, the
botanical origin and the chemical com- assessment of metal concentrations is
position of the nectar of honey-producing helpful for classifying honey according to
plants or secretions, honey contains mix- its botanical and geographical origins
ture of different carbohydrates, including [1,2,10,14,15].
fructose (27.3–44.3%), glucose (22.0– The most abundant metal in honey is K,
*
Tel.: +48 71 320 3445;
40.8%), maltose (2.7–16.0%), sucrose which accounts for 45– 85% of the total
Fax: +48 71 320 2494;
E-mail:
(1.5–3.0%), high sugars (0.1–8.5%), pro- mineral profile [1,10,13–15,17,19,20,23,
[email protected] teins, amino acids, vitamins and minerals 26–35]. Other major metals present in

0165-9936/$ - see front matter ª 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.trac.2008.09.015 117
Trends Trends in Analytical Chemistry, Vol. 28, No. 1, 2009

honey are Na (second most common), Ca and Mg. Cu, quantities of total minerals. Interestingly, the color of
Fe, Zn and Mn are present in intermediate quantities (see dark honeys mostly relates to the concentrations of Cd,
Table 1). Rb, whose content in Spanish honeys is in the Fe and Pb, while the color in light and brown honeys is
range 0.3–3193 lg/g, also belongs to this group associated with the concentrations of Al and Mg [35].
[8,44,45].
Honey also contains trace metals at much lower levels
(<1 lg/g) (see Table 2). 2. Honey as an indicator of environmental
Honey can be regarded as an essential source of metals pollution
in daily diet [6]. However, the intake of metals from
honey is low because it is consumed in small quantities Beside geographical and botanical sources of metals in
[12,27,47]. Typically, daily honey consumption con- hone, the presence of heavy metals can result from
tributes to recommended dietary allowances (RDAs) of anthropogenic factors [55]. The bees forage area covers
many major and minor metals, namely K, Mg, Ca, Na, approximately 7 km2 and includes different environ-
Fe, Mn, Cu and Zn, no more than a few per cent, or even ments, plants and foods. When searching for nectar,
lower (e.g., 0.005–0.5%) [27]. Conversely, heavy metals honeydew, pollen and plant exudates within such a large
present in honey above accepted safety values can cause territory, bees come into a contact with plants, air, water
a certain health risk [6,37,50]. and soil (see Fig. 1). Where there is environmental con-
The mineral composition of honey correlates with its tamination, they also become contaminated and carry
color [6,17,32,35]. Dark and amber honeys (avocado, the pollutants from surroundings into a hive themselves
chestnut, heather) contain higher amounts of certain or in the raw materials collected [2,11,54,55]. In this
major, minor and trace metals (e.g., Al, Ca, Cd, Cu, Fe, K, way, contaminants reach honey and change its compo-
Mg, Mn, Na, Ni, Zn) than pale honeys [6,14,15,17,35]. sition and quality [3,10,39,50].
Honeydew honeys, in particular, have the highest metal Apparently, mines and steelworks, industrial and
content of all honeys [10,17,19,33,35,38,46]. Amber urban areas or highways in or near the bees forage area
light honeys (eucalyptus, thyme) exhibit rather medium can result in an increase of the concentrations of certain

Table 1. Concentrations (in lg/g) of major (K, Na, Ca, Mg) and minor (Fe, Li, Zn, Mn, Cu) metals in honeys of different countries of origin (metals
are in order of appearance in honey)

Origin Ka,b,c Naa,b,c Caa,b,c Mgb,c Feb,c,d Lia,b,c Znb,c,d Mnb,c,d Cub,c,d Ref.
Brazil 167–3353 139–509 7.0–237 ND–373 ND–15.0 – ND–14.0 ND–24.0 ND [18,36]
Chile – – – – 0.10–7.66 – 0.01–4.93 0.01–6.97 0.06–4.32 [37]
Czech – – 11.2–142 18.4–89.0 – – 0.40–3.42 1.16–8.99 0.11–0.88 [38]
Egypt# 213–15550 378–2550 – 102–1325 58–3690 – 5.0–9.3 0.5–5.7 1.00–1.75 [12]
France – – 3.0–108 1.4–110 0.1–10.0 0.02– 0.04–5.96 0.06–10.3 0.03–2.30 [22]
0.24
Greece – – – – – – – 0.11–7.22 ND–0.48 [39]
Hungary – – – – – – – – 0.04–0.44 [24]
India 490–932 97.9–304 32.6–84.6 – 3.60–28.4 – 1.10–29.0 – 1.06–2.90 [30,40]
Ireland 410–714 41.3–196 74.9–175 18.9–53.3 1.70–36.3 – 1.60–22.5 0.90–10.2 1.00–2.30 [32]
Italy 147–4136 6.10–232 9.10–409 3.90–159 0.30–35.1 – 0.60–3.66 0.08–16.9 0.14–5.90 [2,17,27,
34,40]
Macedonia 169–3323 5.90–150 4.10–170 4.40–182 0.03–7.00 – 0.31–15.0 0.16–82.0 0.02–5.90 [9]
Poland 7.70–3659 0.38–89.6 3.30–159 1.10–19.8 ND–16.1 – ND–22.3 – ND–1.82 [4,33]
Romania – – – – 2.04–2.91 – 0.20–7.50 1.75–2.23 – [5,11]
Saudi 9.3–1367 10.0–133 1.50–27.7 18.4–23.2 0.31–8.39 – 0.20–3.00 0.03–0.37 0.21–0.60 [1,6,21]
Arabia
Slovenia 2910 – – – <4.8 – 5.1 6.3 1.4 [41]
Spain 14.0–6785 11.0–1221 13.8–341 3.60–230 ND–60.0 0.47– 0.07–19.1 0.01–27.0 0.04–7.80 [3,8,13–15,
35.4 19,20,28,29,
31,35,42–47]
Turkey 143–6029 9.30–172 3.30–900 2.00–111 0.04–19.7 0.30–1.48 ND–20.2 ND–74.2 ND–3.50 [10,16,23,
25,48–51]
a
Determined by flame emission photometry or spectrometry.
b
Determined by inductively coupled plasma optical emission spectrometry (ICP–OES).
c
Determined by flame atomic absorption spectrometry (FAAS).
d
Determined by electrothermal atomic absorption spectrometry (ET-AAS).
#
Metal content given in reference to dry honey weight; ND, Not detected.

118 http://www.elsevier.com/locate/trac
Trends in Analytical Chemistry, Vol. 28, No. 1, 2009 Trends

Table 2. Concentrations (in lg/g) of trace metals in honeys of different countries of origin

Origin Ala,c Cda,b,c Coa,b,c Cra,b,c Nia,b,c Pba,b,c Ref.

Brazil – ND – – – ND [36]
Chile 0.04–22.1 0.01–0.05 0.01–0.62 0.01–1.98 0.01–1.48 0.01–0.11 [37]
Czech 0.61–20.2 – – – 0.06–1.53 – [38]
Egypt# – 0.01–0.50 1.75–3.20 – 1.25–4.10 4.20–9.30 [12]
France 0.05–1.44 – 0.03–0.25 0.05–0.52 – – [22]
Greece – 0.089–0.222 0.010–0.087 – ND – [39]
Hungary – 0.001–0.003 – 0.002–0.109 – 0.012–0.163 [24]
India – <0.050–0.550 <0.25 – <0.05–0.65 <0.050–0.920 [40]
Italy – 0.001–0.305 0.002–0.057 0.008–0.102 <0.05–2.76 0.003–0.620 [17,40,52]
Poland – 0.008–0.027 – – – 0.02–0.07 [4]
Romania – 0.008–0.032 – 0.03–0.71 0.90–2.03 0.03–0.18 [5,11]
Saudi Arabia – ND–0.037 – – – 0.030–0.240 [6,21]
Slovenia – – – 0.4 <1.3 <2.0 [41]
Spain <0.1–132 <0.001–0.365 <0.01–0.72 0.01–4.48 0.01–3.37 ND–1.20 [3,8,13,31,35,42,44,46,47,53,54]
Turkey 0.004–28.7 <0.001–0.340 0.001–1.60 0.001–0.940 ND–2.17 ND–1.20 [10,16,23,25,48–51]
a
Determined by inductively coupled plasma optical emission spectrometry (ICP–OES).
b
Determined by flame atomic absorption spectrometry (FAAS).
c
Determined by electrothermal atomic absorption spectrometry (ET-AAS).
#
Metal content given in reference to dry honey weight; ND, Not detected.

sensitive bio-indicator for monitoring and control pur-

poses is questioned, and its usefulness is limited to cases
involving very heavy environmental contamination
[52]. Low concentrations of trace metals and their great
variability, according to floral origin, and seasonal and
climatic changes, also mean that honey may not be a
sensitive biological matrix for monitoring anthropogenic
differences and heavy-metal sources in the environment
[37,50].
Honey can also be contaminated with some transition
metals during its processing (see Fig. 2). Sources of
contamination are beekeepers, equipment and tools used
Figure 1. Natural and anthropogenic sources of metals in honey. by them and processing environment. Due to contact
with honey, some metals (e.g., Al, Cd, Co, Cr, Cu, Fe, Pb,
Ni and Zn) can be released from materials (stainless steel,
metals in honey (Al, Ba, Ca, Cd, Cu, Mg, Mn, Ni, Pb, Pd galvanized steel and aluminum) of the equipment used
or Zn) due to pollution by chemical wastes and exhaust for harvesting, producing and preparing honey (e.g.,
fumes [4,5,9,10,16,50]. The degree of contamination extraction, centrifugation, ripening, and containers em-
depends on the floral origin of plants visited by bees. For ployed for settling, storage and shipment) [2,9,10,16,
example, honeys derived from caduceus trees are gen- 18,23,25,46,49,50]. The natural acidity of honey (pH
erally less contaminated than honeys of other botanical 3.5–4.8) and its corrosive effect on the surface of tools
provenance [22]. Honeys produced from nectar of aro- and containers enhance the extent of this process
matic plants are characterized by extremely high con- [2,9,17,25,28]. Metallic contaminants may also be
centrations of heavy metals because these plants tend to introduced with substances supplied for nutrition of bees
concentrate pollutants more easily than herbaceous (e.g., syrup-feed honeys exhibit higher concentrations of
plants [22]. Cd, Co, Fe, K, Mg, Mn, Pb, Na, which are contaminants
As a result, honey can be considered a biomarker for from industrial processing of syrups) [12].
environmental pollution and can accumulatively indi-
cate the level of air, water, plant and soil contamination
over the bees forage area [3–5,9–12,17,18,32,39, 3. Classifying honey by metal content
47,52,54,56]. Investigation of trace metals in honey can
help estimate environmental quality in different regions The demand for classifying the geographical and
and countries. By contrast, due to weak evidence of botanical origins of honey is obvious [14,36,45,57]. It is
metal bioaccumulation, the reliability of honey as a contained in certain national legislation or local

http://www.elsevier.com/locate/trac 119
Trends Trends in Analytical Chemistry, Vol. 28, No. 1, 2009

Figure 2. Honey processing and possible sources of contamination.

directives, which require strict honey labeling (e.g., dew). Hence, metal content is highly indicative of a given
European Union countries need to specify floral origin, area and useful for characterizing and identifying the
source, organoleptic characteristics, physicochemical origins of a honey [8,14,18,46]. For the same reasons,
properties and regional, territorial or topographical ori- mineral composition can also be used to detect adulter-
gins) [45,57]. This practice assures that the quality, the ations of honey (i.e. dilution with water, addition of
properties and the origin of honey in the market comply sugars and syrups, or blending) [12,17].
with the information stated on the label. It also combats
consumer and economic fraud resulting from mixing 3.1. Chemometrics
high-value, original honeys with others, usually of much Statistical data-reduction and visualization techniques,
lower grade and quality [15,44]. Honey is frequently including cluster analysis (CA) [3,8,12,13,15,18,22,23,
subjected to other forbidden practices (e.g., dilution with 33,38,44,45], principal component analysis (PCA)
water or addition of syrups) [15]. In the latter case, [3,8,13–15,17–19,22,34,37,44] or factor analysis (FA)
syrups can be added directly to the honey after har- [9,22], are frequently used to discriminate honey
vesting or supplied as bees feed during the harvest. according to its type and origin. The same data related to
As mentioned before, identification of honey authen- the concentration of metals can be analyzed by pattern
ticity is important in differentiating natural, original recognition and classification procedures, namely linear
honeys manufactured in a certain region by traditional and square discriminant analysis (LDA, SDA) [3,8,13–
procedures from those that are mislabeled, artificial, 15,19,34,35,44,46], K nearest neighbor (KNN) analysis
highly processed, adulterated or otherwise fraudulent. [3,44], soft independent modeling of class analogy
Evidence of geographical location and honey-plant (SIMCA) [3,44], artificial neural networks (ANN)
genus can be obtained by studying pollen loads and [3,19,45], and partial last squares (PLS) modeling [45] or
components within honey and recognizing different logistic regression [8]. Simple analysis of variance
pollen sources (melissopalynology) [17]. (ANOVA) [14,16,17,33,34,47] or t test [20,27] can also
Analysis of minor components of honey, especially be convenient to establish the variables that mean values
metal content, can be a way of recognizing and dis- statistically differ from others and represent honeys
criminating honey, as well as detecting potential fraud in originating from certain regions or botanical species.
its stated origin [3,8,12–15,17–19,22,23,33,34,37, Such chemometric analysis of the metal-content pro-
38,44–46], because metals are good descriptors of floral file of honey is a helpful approach to authenticating
and geographical origins. They are stable in honey over genuine brand honeys, quickly differentiating between
a long time, their concentrations depend on the type of various types of honey and classifying honey according
soil in which nectar-producing plants grow and on to botanical and geographical origins (e.g., unifloral or
vegetation conditions in the forage area [14,19,44–46]. nectar and honeydew, unifloral and artificial feed, and
An excess or a deficiency of certain metals in soil and coastal and mountain) [8,12–15,17–19,33,38,45,46].
water is usually reflected in the mineral composition of Confirmation of genuineness has a special economic
plants visited by bees and the raw materials that they significance because it can help preserve the identity and
collect to produce honey (i.e. nectar, pollen and honey- the quality of honey, and may also protect potential

120 http://www.elsevier.com/locate/trac
Trends in Analytical Chemistry, Vol. 28, No. 1, 2009 Trends

consumers from deception by detecting fraud, adultera- burner heads or the nebulizers of the spectrometers used
tion and faking of honey [45]. [39,43].
Ca [3,8,13–15,18,34], K [3,8,15,18,34,44,45], Mg
[3,8,13–15,34], Li [3,13,34,44], Na [8,18], Rb [44,45] 4.1. Initial sample preparation
and Sr [8] are the variables that provide the most effi- Monoflower, multi-flower and honeydew honeys for
cient classification of honey. The presence of these analysis have been purchased in local markets, stores or
metals in honey is primarily associated with floral origin, cooperatives [1,8,10,11,15,22,24,26,30,36,39,40,42–
soil characteristics and agricultural practices. Fe 44,58]. Raw and processed flower and honeydew honeys
[3,8,18,44,45], Zn [3,8,15,18,44,45], Cu [3,15,38], Cr after extraction have been supplied by beekeepers or
[3,13], Ni [3,14], and Al [13], Cd [3] or Pb [3] are also their associations [2–4,6,8–14,16–21,23,25,27–35,37,
highly discriminatory, but their presence in honey is 38,41,44–48,50–53,56].
very related to environmental pollution. The use of all Honey has been kept in original glass jars, bottles or
available metals for analyzing honey is pointless because buckets used for transport [1,10,22,44] or collected and
some of them carry little or no discriminating informa- stored in glass and plastic vessels, bottles, jars or holders
tion and introduce noise. The number of variables can [2,3,6,9,12,13,16–18,23,24,27–29,31–35,42,44,45,49,
therefore be minimized (3–9 features) to enable correct 50,53]. Decontaminated plastics bottles or polyethylene
classification. The key features selected contain the most tubes have been used to prevent contamination from
information for classification, so they increase reliability glass and metal [19,40,47,52,56]. Representative honey
in categorizing honey and improve data visualization in samples have been kept in the dark [3,12,18,24,
two-dimensional plots [44,46]. 33,40,42,49,53] at 3–5C [1,3,6,12,19,23,27–29,33,
Sometimes, mineral content unaccompanied by other 34,44,45,50,53] or lower temperatures (i.e. 0C [13,
physico-chemical variables may be insufficient to dis- 31,35] or –18C [52]). Honey has also been stored
criminate consistently between types and classes of at room temperature until analysis [2,10,16,18,24,
honey being considered [14,19]. Usually, additional 32,49].
parameters have to be determined to improve the Before sampling, the honey containers have been
description and the classification of honey [namely gently heated to 40–50C [2,17,24,38–40,42,56] in
acidity, water content, ash content, conductivity, water baths, stirred with glass sticks, or additionally
organics content (i.e. sugars and hydroxy methylfur- subjected to sonication [2,17,39,56]. This treatment
furaldehyde)] [14,19,28,29,34,38,46]. The water con- decreased honey viscosity, homogenized the material
tent is associated with climatic conditions, processing being analyzed and facilitated the handling of sample
techniques and storage conditions, the season of year aliquots.
and the degree of maturity of the honey. The pH affects The mass of honey is critical for analysis. For small
texture, stability and shelf life of honey, while the elec- sample portions, the heterogeneity of a raw material is
trical conductivity is associated with the presence of significant and may result in a high inconsistency of the
minerals, organic acids and proteins [28,34]. In combi- results, especially when more sub-samples of the same
nation with metal content, all these selected parameters material are analyzed [2]. The analysis of a homoge-
exhibit great variability as a result of different floral and neous bulk-fluid material is more convenient. It has been
geographical origins of honey and are convenient in prepared by dissolving a large mass of honey (e.g., 10 g
differentiating it using chemometric data analysis. or 20 g, in water at a ratio 2:1 (g/ml) [2,58] or 1:1 (g/
ml) [36]). The sample fluid prepared has been heated at
50C [2,58] and later subjected to sonication [2,36,58].
4. Sample treatment A portion of such dense honey solution was subse-
quently digested or diluted prior to the measurements.
Although honey is soluble in water and metals can di-
rectly be determined in solutions prepared by dissolving
honey samples in water or acidic solutions, samples are 4.2. Sample decomposition
usually decomposed before measurements. The resulting Honey is subjected to different mineralization procedures
residues (ashes and digests) are re-dissolved in water or before analysis of the presence of metals by atomic
acid solutions, and the mineral components are trans- absorption and emission spectrometries. Typically, high-
ferred into solution. This treatment disrupts the organic temperature, dry ashing or wet-acid digestion have been
matrix of honey and extracts metal species. In addition, used to disrupt the organic substances and release
it prevents variations in physico-chemical properties of metals from the complex sample matrix into solution,
standard and sample solutions, as well as numerous suitable for introduction into atomizers (flame, graphite
physical and chemical interferences that can occur in furnace) and excitation sources (flame, inductively
the flame or in the plasma. It also eliminates accumu- coupled plasma). However, this sample treatment is time
lation and deposition of carbonaceous residues in the consuming and includes stages that may be potential

http://www.elsevier.com/locate/trac 121
Trends Trends in Analytical Chemistry, Vol. 28, No. 1, 2009

sources of analyte loss due to sample overheating or  (1 + 1) concentrated HNO3 and HClO4 mixture (for
analyte gain because of various contaminations. 3 h to near dryness [12]); or,
 concentrated HNO3 followed by addition of (2+1) con-
4.2.1. Dry ashing. High-temperature dry ashing is the centrated HNO3 and HClO4 mixture [15].
most common sample pre-treatment used before the The resulting digests or residues left are directly mixed
determination of metals in honey by means of AAS with water. A 1 mol/l HCl solution can be used instead
methods. In the preparation procedure given by the [12], or concentrated HNO3 can be added and then
Association of Official Analytical Chemists (AOAC, Offi- water for dilution [15]. A microwave oven can also
cial Methods of Analysis, issued in 1984, 1990, 2000) decompose the honey in concentrated HNO3 in special
relatively large portions of honey (e.g., 20 g burning cups [23].
[1,20,26,44,48] or 5 g [6,13,19,28,29,32,35,37]) are Wet digestions with the aid of sealed PTFE bombs or
taken for analysis to concentrate the metals present and high-temperature, high-pressure containers of micro-
to decrease the risk of sample heterogeneity. However, wave-oven systems are also used to decompose honey.
the whole procedure can be protracted and last a few Concerning the volatility of some Cd, Ni, Pb, Ni, and Zn
days. Honey samples, initially dried at 100C, are species at high temperatures, microwave-assisted, wet-
incinerated in quartz or platinum crucibles at 550C or digestion procedures seem to be more reliable than dry
600C until constant mass. Samples can additionally be ashing or ashings with Mg(NO3)2, NH4NO3 additives
exposed to infrared (IR)-lamp radiation during calcina- [50,56]. In addition, losses of metal species are elimi-
tion to prevent vigorous foaming [1,20,26,44,48]. Next, nated, chances of sample contamination are less, while
5 ml of 0.1 mol/l HCl [1,6,19,20,26,32,44,48] or HNO3 quantities of reagents and preparation time are reduced.
[13,28,29,31,35] solutions are added to the resulting In this case, relatively smaller sample portions (i.e., 0.2–
ashes, being 0.2–0.6% (w/w) of the initial masses, and 0.3 g [21,56], 0.5 g [9,40], 0.6–0.7 g [27,34], 1 g
the samples are heated to almost complete dryness. The [2,17,24,25,33,36,40,50] or 2 g [59]) are digested
residues left are dissolved in 10 ml of the same acid using concentrated HNO3 and H2O2 mixtures at ratios of
solution and made up with water to 100 ml [1,20,26, 1:1 [9,21,36,56], 3:1 [25,50], 4:1 [2,17,34,59] and 5:1
44,48], 50 ml [19,32] or 25 ml [13,28,29,31,35]. If the [24], or concentrated HNO3 only [33,40]. The resulting
final volume of ash solutions is smaller (e.g., 10 ml), the digests are diluted with water to 50 ml [9,34], 25 ml
volume of the second portion of HCl solution added is [2,36,59], 10 ml [21,24,25,56] or 5 ml [40,50].
also smaller (e.g., 2 ml) [6]. Finally, 10% or 20% (w/v)
honey solutions are obtained prior to the analysis. 4.3. Direct analysis
Other ashing and ash-dissolution-preparation proce- Although, direct analysis of honey samples is performed
dures are also applied, but they differ from the method only occasionally, it certainly minimizes the risk of
described above by sample mass taken for analysis, contamination or loss of analytes due to prolonged
ashing conditions and ash-dissolution preparation (see treatment of samples or their incomplete digestion.
Table 3). When portions of honey <1 g are ashed, Samples are often dissolved and diluted using only water
foaming is negligible and preparation time is shortened (for details, see Table 4) and such sample preparation
to up to 12 hours [41–43]. obtains the results as reliable as those with microwave-
assisted, wet-acid digestions [59]. Obviously, direct
4.2.2. Wet digestion. In classical wet-acid digestions analysis reduces the time of analysis [59]. Before mea-
(open-vessel system), samples of honey are incinerated surement, the sample solutions prepared can be soni-
with the aid of oxidizing reagents or their mixtures. The cated [42] or shaken [43] to improve their consistency.
procedures are performed using glass or PTFE beakers, Filtration removes beeswax impurities [36].
Kjeldahl flasks or porcelain crucibles. Mild heating
(<100C) and water baths are usually employed to avoid
sample spattering and foaming. When honey is initially 5. Atomic spectrometry
dried at 105C for 24 h, a hot plate is used and higher
temperatures can be reached [12]. Normally, samples of Concerning the nutritional value and the potential effect
1–2 g [15,22,37,50], 3 g [58] or 5 g [30] are digested. of honey on human well-being, analysis is especially
The following reagents are applied for sample disrup- significant for consumers because it guarantees the
tion and solubilization: general safety and the wholesomeness of honey. Besides,
 a 9 mol/l HNO3 solution (at 60C for 6–7 h [22]); quality control of honey and detection of its adulteration
 concentrated HNO3, followed by addition of concen- and possible contamination by heavy metals are
trated H2O2 (at 90C for 1 h [58]); important for beekeepers. There are special food-safety
 (1 + 1) concentrated HNO3 and H2O2 mixture (for 4 h regulations and standards (e.g., Codex Alimentarius sets
to dryness [50]); down admissible levels for heavy metals in honey and
 concentrated HNO3 and H2SO4 mixture [30]; honey products [6]).

122 http://www.elsevier.com/locate/trac
Trends in Analytical Chemistry, Vol. 28, No. 1, 2009 Trends

Table 3. Procedures used for honey calcination and ash dissolution

SM, g HC, % Description Ref.

15 15 The sample is calcined at 550C; the resulting ash is dissolved in 5 ml of 2 mol/l HNO3 and 5 ml of [46]
2 mol/l HCl; the solution is made up with water to 100 ml
10 20 The sample is dried at 105C for 2 days and next ashed at 450C or 600C; the resulting ash is [10]
extracted with concentrated HCl and HNO3 mixture; the solution is diluted with water to 50 ml;
solid Mg(NO3)2 Æ 6H2O can be added to the sample to aid calcination, a small portion of
concentrated HNO3 can be added at the end of ashing to complete the process
10 100 The sample is dried at 100C and then ashed at 500C; the resulting ash is dissolved in 10 ml of [49]
concentrated HClO4 and HNO3 mixture
10 100 The sample is heated at 550C; the resulting ash is dissolved in a 6 mol/l HCl solution and then [14]
heated to dryness; the residue is made up with water to 10 ml
10 – The sample is incinerated at 450C and next the residue is dissolved in concentrated HNO3 [4]
6-8 30-40 The sample is desiccated at 105C for 2 h and then calcined at 450C; the resulting ash is dissolved [52]
in 1 ml of concentrated HNO3 and diluted with water to 20 ml
5 20 The sample is heated at 60C (for 2 h), 350C (for 2 h) and 500C (12 h); the resulting ash is dissolved [8]
in 3 ml of concentrated HNO3 and then diluted with water to 25 ml
5 20 The sample is caramelized and then ashed at 700C for 3 h; the resulting ash is dissolved in 5 ml of a [5]
(1+6) HNO3 solution and heated to evaporate half the volume; the resulting solution is diluted with
water to 25 ml
5 10 The sample is dried at 60C and next at 80C for 12 h, and then burned to ash at 450C; the resulting [47]
ash is dissolved in a 0.8 mol/l HNO3 solution to 50 ml
3 15 1 ml of concentrated HCl is added to the sample and heated at 170C, 230C, 290C and 380C [38]
(each step for 1 h), the residue is ashed at 500C (for 16 h); to the resulting ash, 0.5 ml of
concentrated HNO3 is added, and it is heated to dryness; the residue is leached using 5 ml of diluted
HNO3 with 0.5 ml of concentrated HNO3 and diluted with water to 20 ml
2 20 The sample is ashed at 550C; the resulting ash is dissolved in a small volume of a 1 mol/l HCl [3]
solution and diluted with water to 10 ml
2 8 The sample is dried for 1 h and then ashed at 550C for 8 h; the resulting residue is dissolved in 5 ml [18]
of concentrated HCl and made up with water to 25 ml
2 7 The sample is mixed with magnesium acetate, dried at 110C (for 2 h) and then ashed at 500C; the [16]
resulting ash is extracted with a 2 mol/l HNO3 solution and diluted with water to 30 ml
1 10 The sample is ashed at 450C; the resulting residue is dissolved in 5 ml of a 4.5 mol/l HNO3 solution [50]
by heating; the solution is made up with water to 10 ml
1 10 The sample is irradiated with an IR lamp for 1 h after adding 0.1 ml of concentrated HNO3; next the [43]
sample is ashed at 500C for 8 h; the residue is dissolved in 1 ml of concentrated HNO3 and diluted
with water to 10 ml
1 10 The sample is heated at 450C for 16 h; the resulting ash is treated with 4 ml of concentrated HNO3, [42]
heated to dryness and then again treated with 2 ml of concentrated HNO3 and 1 ml of concentrated
HClO4, and heated to near dryness; the residue is diluted with water to 10 ml
1 4 The sample is heated at 150C, 250C, 350C and 550C (each step for 2 h); several drops of a [56]
7.1 mol/l H2SO4 solution are added to the resulting ash, and the solution is diluted with water to 25 ml
1 4 The sample is dried at 105C for 2 h after adding 1 ml of 2 mol/l Mg(NO3) Æ 6H2O or 6 mol/l [56]
NH4NO3 solutions; next, the sample is ashed at 450C for 4–5 h, 1 ml of a 10 mol/l HNO3 solution is
added to the resulting ash and it is heated at 450C for 1 h; the solution left is diluted with water to
25 ml
1 4 The sample is allowed to stand overnight after adding 2–3 ml of a 7.1 mol/l H2SO4 solution; next, the [56]
sample is ashed at 450C for 4–5 h; the resulting residue is dissolved and diluted with water to 25 ml
0.5–1 – The sample is ashed at 450C after adding a small portion of a (1+4) H2SO4 solution; the residue is [41]
dissolved in 0.5 ml of a (1+1) HCl solution and diluted with water to the volume

SM, Sample mass; HC, Honey concentration in final solution.

Indeed, determinations of essential and toxic trace electrothermal atomic absorption spectrometry (ET-
metals in honey are critical for human health and AAS) and inductively coupled plasma optical emission
environmental monitoring. However, analysis of honey spectrometry (ICP-OES), are the most favored in ana-
is difficult and challenging due to the complex organic lyzing honey for the presence of metals. For that reason,
matter of the matrix [24]. Measurements of metals in this review focuses only on these techniques. Although
this kind of sample usually require accurate, sensitive other instrumental techniques are also used {e.g.,
and robust analytical methods [14,22,24,39]. inductively coupled plasma mass spectrometry (ICP-MS)
It is evident that atomic spectrometric methods, [2,33,59–61], ion chromatography [40], total reflection
including flame atomic absorption spectrometry (FAAS), X-ray fluorescence spectrometry (TXRF) [41,62],

http://www.elsevier.com/locate/trac 123
Trends Trends in Analytical Chemistry, Vol. 28, No. 1, 2009

Table 4. Procedures for preparing honey samples without mineralization

SM, g FV, ml Description Ref.

20 1000 The sample is dissolved and diluted with water [39]
1–5 50 The sample is dissolved and diluted using a diluent solution (0.55 mol/l HCl with 6 g/l [43]
La(NO3)3 Æ 6H2O)
1–5 or 0.1–0.2 25 0.25 ml of concentrated HNO3, followed by 2.5 ml of concentrated H2O2 and 0.025 g of [42]
NH4H2PO4 are added to the sample; after dilution with water, the resulting solution is sonicated for
5 min and then magnetically stirred
2.5 25 The sample is dissolved and diluted with a 0.1 mol/l HNO3 solution [8]
1.2 25 1.2 ml of concentrated HNO3 is added to the sample and the mixture is sonicated for 20 min; the [18]
resulting extract is diluted with water
1 100 The sample is dissolved and diluted with water [11]
1 25 0.5 ml of concentrated HNO3 are added to the sample solution (1/1 honey-water mixture) and then [36]
diluted with water
1 5 The sample is dissolved and diluted with water [3,53,54]
0.5 10 0.5 ml of concentrated HNO3 and 1 ml of concentrated H2O2 solutions are added to the sample and [24]
then sonicated for 10 min; the resulting mixture is diluted with water
0.2–0.4 10 The sample is dissolved and diluted with a 2 mol/l HNO3 solution, the resulting solution is mixed [56]
using an ultrasonic shaker
– 25 The whole sample (ampoule containing 5 g of dissolved honey CRM) is diluted with water [59]

SM, Sample mass; FV, Final volume of sample solution.

potentiometric stripping analysis [63] or anodic strip- and Mg [6,8–10,12,18,25,27,30,32–34,38,43–45,48,

ping voltammetry [64]}, they are in the minority. 49,51]), which are the most abundant minerals in
honey. Concentrations of these metals are so high that
5.1. FAAS sample solutions can be diluted even 50–500 times to
The main advantages of FAAS with atomization in a match the dynamic ranges of calibration curves
widespread air-acetylene flame are low operational costs [10,20,26,43]. This practice also minimizes or com-
and reasonably good analytical performance [39,43]. pletely eliminates possible chemical interferences.
The possibility of only sequential analyses and the Beside major metals, minor and some trace metals can
narrow ranges of linear response can be regarded as be determined by applying FAAS (namely Fe [1,6,8,10,
disadvantages of this method [41]. 12,18,20,25,26,30,32–34,44,45,48–51], Cu [6,8,10,
Using FAAS for the determination of metals in honey, 12,18,20,30,32,33,38,44,45,48,49,51], Zn [5,6,8–10,12,
the samples are usually mineralized in order to avoid 18,25,30,32,33,38,43–46,48,50,51], Mn [6,10,12,18,
possible matrix-related interferences [1,5,6,9,10,12,18, 25,32,38,44,45,48,49,51], Co [12,44,48], Ni [12,44,
20,25–27,30,32,38,41,43–46,48,49,51]. In this case, 48,49], Sr [8,12], Cd [6,12,49] and Pb [6,12]). Due to
aqueous external standard solutions and simple linear low concentrations of these metals, the sample solutions
regression can be used for calibration and quantification, analyzed are usually undiluted. In some cases, the met-
respectively [6,8,10,18,27,30,32,34,38,43,49]. Other- als measured are initially enriched from solutions before
wise, multiple linear regression can be applied to mini- analysis [e.g., Cd and Pb are complexed by ammonium
mize the effects of chemical interferences in the flame pyrrolidine dithiocarbamate (APDC) and extracted into
[20,26]. methyl isobutyl ketone (IBMK)] [8].
Direct analysis of sample solutions prepared by dis- Ionization of Ca and Mg is avoided by adding NaCl
solving honey in water containing small amounts of solution to sample solutions [10] or using a fuel-rich
HNO3 is quite rare [8,18,65]. In this case, high organic- flame [8]. Chemical interferences from phosphates and
matter content of samples hinders calibration against other anions in measurements of Ca, Mg, Fe, Cu, Zn and
aqueous standards because the physico-chemical prop- Mn are minimized by adding La(NO3)3 or LaCl3 solutions
erties of the solutions measured differ considerably [65]. to sample solutions [10,32,38]. For the same reason, La
This drawback can be alleviated by decreasing the con- modifiers can also be present in carrier solution of the FI
centration of organic compounds supplied to the nebu- manifold [8] or diluting solution of the automatic dilu-
lizers by using on-line automatic dilutors [43] or tion system [43]. Table 5 shows the amounts and the
discrete-sample-introduction manifolds [i.e. flow injec- concentrations of the modifiers to be added.
tion (FI)] [8,66].
FAAS is commonly used to measure the concentra- 5.2. ET-AAS
tions of alkali and alkaline-earth metals (i.e. K [1,9,10, The sensitivity offered by ET-AAS is needed for the deter-
12,20,25–27,30,32,34,38,41], Na [1,9,12,20,26,27, mination of trace quantities of Al [16,25,46,50], Cd [3,4,
30,32,34], Ca [1,8–10,18,25,27,30,32–34,38,43,51] 9,24,25,40,42,46,47,50,52,53,58], Co [16,25,40,46],

124 http://www.elsevier.com/locate/trac
Trends in Analytical Chemistry, Vol. 28, No. 1, 2009 Trends

Table 5. Modifiers used in analysis of honey by FAAS and FAES

Method Analyte Concentration Amount of modifier solution added Ref.

CsCl
FAESa K, Na 25 g/l – [33]

KCl
FAESb Li, Na, Rb 2.5 g/l 10/1c or 20/1c [8]

LaCl3
FAASa Cu, Fe, Mn, Zn 2 g/l 2 ml to 8 ml of sample solution [10]
FAASa Ca, Mg 1 g/l 1.5 ml to 50 ml of sample solution [32]

La(NO3)3
FAASa Ca, Mg 16 g/l 3 ml to 2 ml of sample solution (Ca) or 1 ml to 9 ml of sample solution (Mg) [38]
FAASb,d Ca, Mg, Zn 6 g/l with 2% HCl – [43]

La2O3
FAASb Mg 1 g/l with 0.65% HNO3 20/1c [8]

NaCl
FAASa Ca, Mg 1 g/l 3 ml to 1 ml of sample solution [10]
FAESb K 5 g/l 10/1c [8]
a
Modifier added to sample solutions.
b
Modifier present in carrier or diluting solutions.
c
Carrier to sample solutions flow rate.
d
Modifier solution used to dissolve honey samples.

Cr [3,16,24,25,46,50,52], Cu [3,9,16, 24,25,27,34, propriate distribution in the graphite tubes. Usually, 4–

40,42,46,50], Ni [3,16,25,40,46,50,58], Pb [3,4,24, 6% (m/v) concentrations of honey in sample solutions
25,40,42,47,50,52,54,58], Zn [4,16,27,40,42,58] and are recommended because this honey dilution is not too
V [58]. Minor metals (e.g., Fe [9,16,27,40,46,58] and high and the pyrolysis products do not remain in the
Mn [9,16,27,34,46,50,58]) are also measured using graphite tubes [24].
ET-AAS. Direct calibration with aqueous standards is employed
Honey samples are mineralized because the carbo- to quantify metals in sample solutions being analyzed
hydrate-rich sample matrix is responsible for gradual [3,16,24,34,42,53,54]. If matrix effects are observed,
accumulation of carbonaceous residues in graphite despite optimal honey dilution or for other reasons, the
tubes, leading to deterioration in the sensitivity and the use of aqueous standards for calibration is hampered
precision of ET-AAS measurements [4,9,16,25,27,34, [42]. Hence, standard addition is suggested [3,52,40,53,
40,46,47,50,52,58]. Analysis of honey without initial 54,58] or matching matrix-standard solutions contain-
sample treatment, except for dissolution in water, is ing fructose are used [42]. As in case of FAAS, the short
rather rare [3,24,42,53,54]. Despite in situ ashing linearity ranges of ET-AAS requires adequate sample
(calcination) at 650–1200C, which is incorporated into dilution [24,40].
a furnace-heating program to disrupt the organic matrix Different chemical modifiers are widely applied in ET-
of honey, the residues accumulate inside the graphite AAS to reduce or to remove matrix interferences and to
tubes when the concentration of honey in the solutions control atomization. Additions of NH4H2PO4 in the case
prepared exceeds several per cent (w/v). As a result, of Cd, Pb and Zn [3,42,50,53,54], Mg(NO3)2 for Al, Cd,
atomization profiles (absorption signals) deteriorate and Co, Cr, Cu, Fe, Mn, Ni and Zn [40,50,58], Pd(NO3)2 for
background levels increase, because of the absorption of Cu, Cr and Ni [3], or mixtures of NH4H2PO4 and
light by molecular species or dispersion by particulates Mg(NO3)2 or Pd(NO3)2 and Mg(NO3)2 for Cd, Cr, Cu
[42]. To alleviate these problems and to prolong the and Pb [3,24,40,47,50,53,54] prevent metal loss and
lifetime of atomizers, sample is disrupted in a heating supply steady isothermal conditions of pyrolysis and
cycle in the presence of H2O2 and HNO3. Both oxidizing atomization in graphite tubes. Modifier solutions are
reagents are initially added to the sample solutions initially pre-mixed with the sample solutions prior
[24,42] or mixed with them in the autosampler cups to injection or, more frequently, they are added into
[3,53,54]. In the latter case, Triton X-100 is added to the graphite tubes together with the samples (see
sample solutions to reduce solution tension and inap- Table 6).

http://www.elsevier.com/locate/trac 125
Trends Trends in Analytical Chemistry, Vol. 28, No. 1, 2009

Table 6. Matrix modifiers used in analysis of honey by ET-AAS (A added to the sample, B added to the graphite tube)

Analyte Modifier mass MV IV Ref.

NH4H2PO4
A Cd, Pb 2 mg/1 ml (Cd), 1 mg/1 ml (Pb) 200 ll (Cd), 100 ll (Pb) 30 ll [3,53,54]
A Cu, Cd, Pb, Zn 25 mg/ 25 ml – 20 ll [42]
B Pb 200 lg 10 ll 20 ll [50]

Mg(NO3)2
B Al, Cr, Cu, Ni, Mn 50 lg 10 ll 20 ll [50]
B Cd 10 lg 10 ll 10 ll [58]
B Co, Fe, Ni 50 lg 10 ll 20 ll [40]
B Zn 6 lg 10 ll 20 ll [40]

Pd(NO3)2
B Cr, Cu, Ni 14 lg 18 ll 12 ll [3]

NH4H2PO4 (1) – Mg(NO3)2 (2)

B Cd, Pb 20 lg (1), 1 lg (2) 10 ll 20 ll [40]
– Cd, Pb – – 15 ll (Cd), 25 ll (Pb) [47]

Pd(NO3)2 (1) – Mg(NO3)2 (2)

B Cd 22 lg (1), 4 lg (2) 9 ll (1), 9 ll (2) 12 ll [3,53]
B Cd 15 lg (1), 10 lg (2) 10 ll 20 ll [50]
B Cd, Cr, Cu, Pb 5 lg (1), 3 lg (2) 5 ll 20 ll [24]
B Pb 18 lg (1), 18 lg (2) 9 ll (1), 9 ll (2) 12 ll [3,54]

MV, Volume of modifier solution; IV, Sample injection volume.

5.3. Emission spectrometry means that external aqueous solutions are often em-
Using the same equipment as for FAAS, but measuring ployed for calibration and quantification of metals
the optical emission (FAES), K [8,10,14,18,33,44,45], [2,3,11,19,22,28,29,31,35,37–39,56]. The acid con-
Na [8,14,18,33,44,45], Li [8,44,45] or Rb [8,44,45] tent in calibration solutions should match the acidity of
can be determined. Otherwise, separate flame-emission the prepared samples [56]. Standard additions [15,59] or
photometers are used, and Na [46,48,51], K [46,48,51] internal standards (In or Y) [36,59] are used in some
or Ca [48] concentrations are measured. Unfortunately, cases and improve the analytical results by increasing
the determinations of alkali metals may be accompanied precision. Matching matrix standard solutions contain-
by ionization interferences, so FI can be used to reduce ing glucose and fructose mixtures, sucrose or honey can
the amounts of sample while adding ionization sup- also be applied [39].
pressors to the carrier solution (i.e. KCl for Na, Li and Rb,
or NaCl for K) [8]. A solution of CsCl can be added to 5.4. Measurement quality assurance
sample solutions for the same purpose when determining Although two certified reference material (CRM) candi-
Na and K [33] (see Table 5). dates for trace elements in honey have been tested [59],
Inductively coupled plasma optical emission spec- a CRM is still not available commercially due to problems
trometry (ICP-OES) is quite often applied to multi-ele- with the homogeneity of honey. These CRM candidates,
mental analyses of honey [2,3,11,13–15,17,19,21,22, prepared by the Institute of Reference Materials and
28,29,31,35–39,41,47,56,59]. The method has ex- Measurements of the Joint Research Center of European
tended dynamic ranges (several orders of magnitude), Union (EC-JRC-IRMM), were dense aqueous solutions of
high-power detection, high sensitivity and high preci- eucalyptus (70%) and robinia (80%) ripened honeys
sion, low background levels and freedom from chemical placed in sealed ampoules purged with Ar.
interferences. Validation of the analytical procedures and the
Honey samples are digested before measurements methods used for honey analysis have been established
using dry ashing [3,13,19,28,29,31,35,37,38,41,47] in a different way. Other CRMs with high carbohydrate
and wet-acid digestion in open or closed vessel systems content {e.g., apple leaves (NBS 1515) [25,27,34,47,
[2,12,17,21,22,37,56,59]. 50,52], brown bread (BCR 191) [59], corn (NBS 8413)
Direct analysis of samples dissolved and diluted with [17], wheat (IPE 684) [38], wholemeal flour (BCR 189)
water is also undertaken [11,14,36,39,56]. [58] or Antarctic krill (MURST-ISS-20) [34]} are com-
Matrix decomposition or sample dilution, which has monly used. For the same purpose, recoveries from
no significant effect on the performance of ICP-OES, samples spiked with known amounts of metals at

126 http://www.elsevier.com/locate/trac
Trends in Analytical Chemistry, Vol. 28, No. 1, 2009 Trends

different concentrations have been evaluated [9,15, Fe, Mg and Zn in honey [67–69]. Other techniques (e.g.,
24,25,36,39,50,53,54,58]. The results obtained using ultrafiltration or size-exclusion chromatography) might
the calibration with external aqueous standard solutions also be expected to be helpful in separating the fractions
have often been compared with those achieved using the of metal species associated with organic compounds of
standard-additions method that enables control of pos- different molecular weights. In addition, there could be a
sible matrix effects [42,43,53,54,58]. The conformity much greater contribution from procedures for direct
between the results shows that the matrix effects are analysis of honey, involving minimal sample treatment
absent, while the new methods of analysis, including and preparation.
sample-preparation treatments and calibration strate-
gies, are valid.
Comparisons of the results obtained without sample Acknowledgment
manipulation or treatment with those based on the P. Pohl acknowledges receiving the Return Fellowship of
conventional sample-calcination and ash-dissolution the Alexander von Humboldt Foundation in 2007.
procedures [42,43] or using another measurement
method [i.e. ion-selective electrodes (ISEs)] [12], have
also had a role in quality assurance and control. References
[1] H.M. Abu-Tarboush, H.A. Al-Kahtani, M.S. El-Sarrage, Food
Chem. 46 (1993) 13.
[2] S. Caroli, G. Forte, A.L. Iamiceli, B. Galoppi, Talanta 50 (1999)
6. Conclusions 327.
[3] J.C. Rodriguez Garcia, R. Iglesias Rodriguez, R.M. Pena Crecente,
Interest in the analysis of honey due to its metal content J. Barciela Garcia, S. Garcia Martin, C. Herrero Latorre, J. Agric.
Food Chem. 54 (2006) 7206.
has significantly increased over the past 15 years. In a [4] P. Przybylowski, A. Wilczynska, Food Chem. 74 (2001) 289.
number of studies, assessment of the levels of different [5] I. Bratu, C. Georgescu, J. Cent. Eur. Agric. 6 (2005) 467.
metals, especially heavy metals, has been a general [6] K.A. Osman, M.A. Al-Doghairi, S. Al-Rehiayani, M.I.D. Helal,
objective aimed at characterizing various honeys in view J. Food Agric. Environ. 5 (2007) 142.
of the implications for their safety. Such information is [7] H.D. Belitz, W. Grosch, P. Schieberle, Food Chemistry, Springer-
Verlag, 2004 pp. 88–891.
important to beekeepers, as it helps them to avoid pos- [8] O.M. Hernandez, J.M.G. Fraga, A.I. Jimenez, F. Jimenez, J.J. Arias,
sible contamination during honey processing, and to Food Chem. 93 (2005) 449.
consumers, as it assures them that the product is of good [9] E. Stankovska, T. Stafilov, R. Sajn, Environ. Monit. Assess. 142
quality. (2008) 117.
Since soil and plants are natural sources that have a [10] A. Uren, A. Serifoglu, Y. Sarikahya, Food Chem. 61 (1998) 185.
[11] N. Matei, S. Birghila, S. Dobrinas, P. Capota, Acta Chim. Slov. 51
great influence on the mineral composition of honey, (2004) 169.
information on the metal profile is suitable for categoriz- [12] M.N. Rashed, M.E. Soltan, J. Food Compos. Anal. 17 (2004) 725.
ing honeys according to their floral and geographical [13] A. Terrab, D. Hernanz, F.J. Heredia, J. Agric. Food Chem. 52
provenance. However, other sources of metals in honey (2004) 3441.
(e.g., environmental pollution, apiculture practices and [14] M.J. Nozal Nalda, J.L. Bernal Yague, J.C. Diego Calva, M.T. Martin
Gomez, Anal. Bioanal. Chem. 382 (2005) 311.
honey processing) should also be considered in classifying [15] R. Fernandez Torres, J.L. Perez Bernal, M.A. Bello Lopez,
and identifying honey using chemometric techniques. M. Callejon Mochon, J.C. Jimenez Sanchez, A. Guiraum Perez,
Apart from nectar and other plant exudates, bees also Talanta 65 (2005) 686.
interact with other parts of the environment, from which [16] E. Yarsan, F. Karacal, I.G. Ibrahim, B. Dikmen, A. Koksal,
they collect dusts, aerosols and other deposits, so it is really Y.K. Das, Bull. Environ. Contam. Toxicol. 79 (2007) 255.
[17] A. Pisani, G. Protano, F. Riccobono, Food Chem. 107 (2008)
advisable to monitor honey to assess the pollution level in 1553.
industrial and urban areas where hives are located. [18] J. Soares dos Santos, N. Soares dos Santos, M.L. Pires dos Santos,
Unfortunately, the overwhelming majority of works S. Novaes dos Santos, J.J. de Jesus Lacerda, J. Braz. Chem. Soc. 19
on the determination of metals in honey deals with total (2008) 502.
concentrations measured after complete disruption of the [19] A. Terrab, A.G. Gonzalez, M.J. Diez, F.J. Heredia, J. Sci. Food Agric.
83 (2003) 637.
samples. However, this approach is ineffective in [20] J.L. Rodriguez-Otero, P. Paseiro, J. Simal, A. Cepeda, Food Chem.
retrieving information on the specific physico-chemical 49 (1994) 169.
forms of metals present in honey. Demand for investi- [21] A.S. Al-Khalifa, I.A. Al-Arify, Food Chem. 67 (1999) 21.
gation of the functions of metal species and their actual [22] J. Devillers, J.C. Dore, M. Marenco, F. Poirier-Duchene, N. Galand,
impact on human health could result in studies on metal C. Viel, J. Agric. Food Chem. 50 (2002) 5998.
[23] Y. Bagci, D. Arslan, M.M. Ozcan, N. Dursun, Int. J. Food Sci. Nutr.
speciation in honey and evaluation of metal bioavail- 58 (2007) 567.
ability and absorbability. [24] Z. Ajtony, L. Bencs, R. Haraszi, J. Szigeti, N. Szoboszlai, Talanta 71
So far, the only attempt at speciation, based on solid- (2007) 683.
phase extraction (SPE) with different ion exchangers, [25] S. Silici, O.D. Uluozlu, M. Tuzen, M. Soylak, J. Hazard. Mater. 156
proposed in the literature brought some insight into Ca, (2008) 612.

http://www.elsevier.com/locate/trac 127
Trends Trends in Analytical Chemistry, Vol. 28, No. 1, 2009

[26] J.L. Rodriguez-Otero, P. Paseiro, J. Simal, L. Terradillos, A. Cepeda, [47] I. Frias, C. Rubio, T. Gonzalez Iglesias, A.J. Gutierrez, D. Gonzalez
J. Apic. Res. 31 (1992) 65. Weller, A. Hardisson, Bull. Environ. Contam. Toxicol. 80 (2008)
[27] M.E. Conti, Food Control 11 (2000) 459. 30.
[28] A. Terrab, A.F. Recamales, D. Hernanz, F.J. Heredia, Food Chem. [48] H. Yilmaz, O. Yavuz, Food Chem. 65 (1999) 475.
88 (2004) 537. [49] F. Erbilir, O. Erdogrul, Environ. Monit. Assess. 109 (2005) 181.
[29] A. Terrab, F.J. Heredia, J. Sci. Food Agric. 84 (2004) 1801. [50] M. Tuzen, S. Silici, D. Mendil, M. Soylak, Food Chem. 103 (2007)
[30] V. Nanda, B.C. Sarkar, H.K. Sharma, A.S. Bawa, J. Food Compos. 325.
Anal. 16 (2003) 613. [51] M. Kucuk, S. Kolayli, S. Karaoglu, E. Ulusoy, C. Baltaci, F. Candan,
[31] A. Terrab, A.F. Recamales, M.L. Gonzalez-Miret, F.J. Heredia, Food Food Chem. 100 (2007) 526.
Chem. 92 (2005) 305. [52] M.E. Conti, F. Botre, Environ. Monit. Assess. 69 (2001) 267.
[32] G. Downey, K. Hussey, J.D. Kelly, T.F. Walshe, P.G. Martin, Food [53] J.C. Rodriguez Garcia, J. Barciela Garcia, C. Herrero Latorre,
Chem. 91 (2005) 347. M. Freire Rodriguez, S. Garcia Martin, R.M. Pena Crecente,
[33] M. Madejczyk, D. Baralkiewicz, Anal. Chim. Acta 617 (2008) 11. Talanta 61 (2003) 509.
[34] M.E. Conti, J. Stripeikis, L. Campanella, D. Cucina, M.B. Tudino, [54] J.C. Rodriguez Garcia, J. Barciela Garcia, C. Herrero Latorre, S. Garcia
Chem. Central J. 1 (2007) 14. Martin, R.M. Pena Crecente, Food Chem. 91 (2005) 435.
[35] M.L. Gonzalez Miret, A. Terrab, D. Hernanz, M.A. Fernandez [55] S. Bogdanov, Apidologie 37 (2006) 1.
Recamales, F.J. Heredia, J. Agric. Food Chem. 53 (2005) 2574. [56] P. Fodor, E. Molnar, Microchim. Acta 112 (1993) 113.
[36] T.M.F.F. Mendes, S. Nivaldo Baccan, S. Cadore, J. Braz. Chem. Soc. [57] E. Anklam, Food Chem. 63 (1998) 549.
17 (2006) 168. [58] M. Taddia, A. Musiani, S. Schiavi, Ann. Chim. (Rome) 94 (2004)
[37] C. Fredes, G. Montenegro, Cienc. Investig. Agrar. 33 (2006) 50. 107.
[38] J. Lachman, D. Kolihova, D. Miholova, J. Kosata, D. Titera, K. Kult, [59] S. Caroli, G. Forte, M. Alessandrelli, R. Cresti, M. Spagnoli, S. DLlio,
Food Chem. 101 (2007) 973. J.Pauwels,G.N.Kramer,Microchem.J.67(2000)227.
[39] M.D. Ioannidou, G.A. Zachariadis, A.N. Anthemidis, J.A. Stratis, [60] A.P. Packer, M.F. Gine, Spectrochim. Acta, Part B 56 (2001)
Talanta 65 (2005) 92. 69.
[40] P.L. Buldini, S. Cavalli, A. Mevoli, J.L. Sharma, Food Chem. 73 [61] S. Bogdanov, M. Haldimann, W. Luginbuhl, P. Gallman, J. Apic.
(2001) 487. Res. Bee World 46 (2007) 269.
[41] P. Kump, M. Necemer, J. Snajder, Spectrochim. Acta Part B 51 [62] T. Golob, U. Dobersek, P. Kump, M. Necemer, Food Chem. 91
(1996) 499. (2005) 593.
[42] P. Vinas, I. Lopez Garcia, M. Lanzon, M. Hernandez Cordoba, [63] E. Munoz, S. Palmero, Food Chem. 94 (2006) 478.
J. Agric. Food Chem. 45 (1997) 3952. [64] G. Sanna, M.I. Pilo, P.C. Pin, A. Tapparo, R. Seeber, Anal. Chim.
[43] I. Lopez Garcia, P. Vinas, C. Blanco, M. Hernandez Cordoba, Acta 415 (2000) 165.
Talanta 49 (1999) 597. [65] F. Salinas, V. Montero de Espinosa, E. Osorio, M. Lozano, Rev. Esp.
[44] M.J. Latorre, R. Pena, C. Pita, A. Botana, S. Garcia, C. Herrero, Cienc. Tecnol. Aliment. 34 (1994) 441.
Food Chem. 66 (1999) 263. [66] E. Sauri-Duch, J.C. Hernandez-Chavez, Rev. Esp. Cienc. Tecnol.
[45] M.J. Latorre, R. Pena, S. Garcia, C. Herrero, Analyst (Cambridge, Aliment. 34 (1994) 450.
UK) 125 (2000) 307. [67] P. Pohl, B. Prusisz, Talanta 69 (2006) 1227.
[46] A.M. Gonzalez Paramas, J.A. Gomez Barez, R.J. Garcia Villanova, [68] P. Pohl, B. Prusisz, Talanta 70 (2007) 411.
T. Rivas Pala, R. Ardanuy Albajar, J. Sanchez Sanchez, J. Sci. Food [69] P. Pohl, B. Prusisz, Can. J. Anal. Sci. Spectrosc. 52 (2007)
Agric. 80 (2000) 157. 207.

128 http://www.elsevier.com/locate/trac