Food Anal.

Methods (2016) 9:777–784

DOI 10.1007/s12161-015-0252-1

Evaluation of Sample Preparation Methods for the Determination

of Ca, Cu, Fe, K, and Na in Milk Powder Samples by ICP-OES
Eliézer Q. Oreste 1 & Alexander O. de Souza 1 & Camila C. Pereira 1 &
Meibel T. Lisboa 1 & Mirla Janaína A. Cidade 2,3 & Mariana A. Vieira 1 &
Solange Cadore 3 & Anderson S. Ribeiro 1

Received: 12 April 2015 / Accepted: 30 June 2015 / Published online: 12 July 2015
# Springer Science+Business Media New York 2015

Abstract In this paper, a comparative study between three digestion were 14.0, 0.57, 1.78, 14.65, and 46.08, respective-
acid decomposition procedures of milk powder samples for ly. The decomposition procedure using a cold finger in an
the determination of Ca, Cu, Fe, K, and Na by optical emis- open system is effective, presenting the lowest residual carbon
sion spectrometry with inductively coupled plasma is present- content (6 %), good accuracy, and relative standard deviation
ed. Utilizing a reflux system, it was possible to increase the lower than 5 %, and appears to be a suitable procedure for
temperature of the digester block above the boiling point of sample preparation for routine analysis.
the reaction medium, avoiding the loss of analyte and the
excessive evaporation of the acids during the heating. Keywords Milk powder . Metal . ICP-OES
Conditions of acidity (10 mL HNO3/H2SO4), temperature
(350 °C), and decomposition time (4 h) were defined by eval-
uation of residual carbon content. Acid digestion in an open Introduction
system with a cold finger was compared with a conventional
open system (4 h ate 120 °C), as well as with microwave Milk and milk products are widely consumed around the
digestion (heating ramp to 180 °C by the time 45 min), for world, being an important source of beneficial nutrients for
four commercial samples of milk powder. For the three sam- humans, due to their essential chemical composition (protein
ple preparation procedures, the values of the analyte concen- and mineral content) (Buldini et al. 2002; Inam and Somer
trations in all analyzed samples were similar. The accuracy of 2000). Milk powder is of great importance from an economic
the acid digestion procedures for Ca, Cu, Fe, K, and Na was point of view, because it is widely consumed in retail market,
assessed by the analysis of a certified reference material of and it is also widely consumed in public schools (Ferrão et al.
milk (infant formula), and good agreement between the certi- 2007). Thus, it is of great importance to assess the authenticity
fied values and the measured value was obtained. The limits of of milk and its products for both economic reasons and public
quantification for the acid digestion with reflux system for Ca, health. The most common frauds registered in milk powder
Cu, Fe, K, and Na, in milligram per kilogram, were 7.48, 0.31, occur by the addition of milk whey, maltose, sucrose, starch,
1.0, 8.20, and 25.81, respectively, and for the microwave and caustic soda, diluting the constituents of the milk (Ferrão
et al. 2007; Veloso et al. 2002). Another important aspect that
also needs to be addressed is the concentration of metallic ions
* Anderson S. Ribeiro present in this kind of food, in order to inform consumers if the
[email protected] concentrations are in agreement with the legislation (Korn
et al. 2008; Pereira et al. 2013; Tokalioğlu and Gürbüz 2010).
1
Laboratório de Metrologia Química (LabMeQui), Universidade For elemental determination in milk powder, it is necessary
Federal de Pelotas, 96160-000 Capão do Leão, RS, Brasil to apply an appropriate sample preparation method in order to
2
Departamento de Química, Universidade Federal de Roraima, get accurate results (Pereira et al. 2013; Mitra 2003). One
69304-000 Boa Vista, RR, Brasil method widely employed for the preparation of organic sam-
3
Institute of Chemistry, University of Campinas, P.O. Box 6154, ples is acid decomposition, which involves the use of oxidant
13083-970 Campinas, SP, Brasil acids followed by a heating step that can be performed in an
778 Food Anal. Methods (2016) 9:777–784

open or closed system (Oliveira, 2003). Considering the open and high sensitivity, as well as being widely used for analysis
system, analyte and/or reagent loss by volatilization may be of milk powder samples (Ikem et al. 2002; McKinstry et al. 1999;
observed due to the high temperatures usually employed in Pereira et al. 2013; Ribeiro et al. 2003). Nevertheless, in these
this system, as well as the higher risk of contamination and techniques, a very high residual carbon content can generate
blank values (Castro et al. 2009; Oliveira, 2003). These pecu- carbon deposits in sampler cones of ICP-MS, as well as instabil-
liarities are not observed when utilizing a closed system, such ities in the plasma, affecting the accuracy of the technique.
as microwave oven. However, the microwave system requires In this context, the objective of this study was to evaluate a
the use of a small sample mass to prevent the pressure exceed- sample preparation procedure for milk using conventional ac-
ing the safety level. Thus, it is essential that the equipment id decomposition in an open system with a cold finger, eval-
presents not only a complete security system to avoid explo- uating the effect of acid and its mixtures, time and temperature
sions, including temperature and pressure control, but also an of decomposition, and residual carbon content. The results
exhaust system; these requirements may make the system obtained were compared with those of conventional acid de-
rather costly (Korn et al. 2008; Nemati et al. 2010; Pereira composition in open system and also in closed system (micro-
et al. 2013; Oliveira 2003). wave digestion) for the determination of Ca, Cu, Fe, K, and
Recently, there have been reports of the use of reflux sys- Na by ICP-OES.
tems for acid decomposition in open systems, obtaining satis-
factory results for the determination of several elements (in-
cluding volatiles) in various kinds of samples (Ferreira et al. Experimental
2013; Pinheiro et al. 2014; Oreste et al. 2013a; Oreste et al.
2013b). The reflux system consists of a cold finger inserted in a Instrumentation
digestion vessel for sample preparation using an open system
with conventional heating. The reflux system efficiency was The analytical measurements were made with a simultaneous
compared with the conventional open system by Oreste et al. ICP-OES, model Optima 3000 DV (PerkinElmer, Norwalk,
(2013a) for decomposition of biological samples and determi- CT, USA), equipped with a peristaltic pump, a cross-flow neb-
nation of Hg, an extremely volatile element. The authors ob- ulizer coupled to a Scott camera, and a ceramic injector tube
served that the heating of the digester tube at 120 °C promotes with 2.0-mm internal diameter. This instrument has a solid-
the dryness of this tube in the absence of the cold finger. state segmented array charge coupled device (SCD) detector
Furthermore, the use of the reflux system avoids losses of Hg and can operate in both radial and axial torch configurations.
when the temperature of the digester block was maintained at The entire system was controlled with PE Winlab software.
120 °C, and the results showed a good compliance between the Argon with a purity of 99.996 % (White Martins, São Paulo,
certified values and the measured values for Hg. Additionally, SP, Brazil) was used in the analytical measurements. The in-
analyte losses were observed when the open system was used, strumental conditions used are presented in Table 1.
at the same temperature, generating unsatisfactory results. A Shimadzu TOC-5000 TOC analyzer (Shimadzu GmbH
An important factor in evaluating the efficiency of acid Europe, Duisburg, Germany) was used for the determination
decomposition is the percentage of residual carbon, which of the residual organic carbon content. The determination of
also helps to avoid interference during the instrumental mea-
surement, since this residue can affect the performance of Table 1 Instrumental parameters for determinations by ICP-OES
some analytical techniques, mainly those that use plasma Parameter Condition
(Gouveia et al. 2001; Wasilewska et al. 2002). Oreste et al.
(2013a) report that the total organic carbon (TOC) for reflux RF power (kW) 1.3
system was higher than conventional open system under the Plasma argon flow rate (L min−1) 15.0
same conditions used for acid decomposition with HNO3 on- Nebulization argon flow rate (L min−1) 0.8
ly. However, additional studies including important matters Auxiliary argon flow rate (L min−1) 0.5
such as time, temperature, and/or acid mixtures for sample Sample flow rate (mL min−1) 1.0
preparation were not performed in order to decrease the resid- Integration time (s) 1 (min)–5 (max)
ual carbon of the solutions resulting from sample preparation. Read delay (s) 30
Various analytical techniques based on atomic spectrometry Emission wavelengths (nm) Ca (317.933) (II)a;
have been reported for the determination of metals in food sam- Fe (238.204) (II)a;
ples (Korn et al. 2008; Fernandes et al. 2015; Lara et al. 2014; Cu (324.752) (I)b;
Na (330.237) (II)b;
Freschi et al. 2012). However, plasma-based techniques such as K (766.490) (I)a
inductively coupled plasma optical emission spectrometry
a
(ICP-OES) or inductively coupled plasma mass spectrometry Radial mode
b
(ICP-MS) are preferred due to multielemental characteristics Axial mode
Food Anal. Methods (2016) 9:777–784 779

residual acidity was performed using acid–-base titration with

NaOH solution, and phenolphthalein was used as indicator.
The result was expressed as % of the original acid that remains
in final digest. Samples were weighed using an analytical
balance model 2140 from Ohaus Adventurer, USA, with a
resolution of 0.1 mg and tare maximum of 210 g. For the
sample acid digestion, a heated digester block was used
(MA-4025 model, Marconi, Brazil). In each digester tube, a
cold finger was introduced to avoid loss by volatilization of
analytes and reagents, as shown in Fig. 1, based in a previous
publication by our research group (Oreste et al. 2013a). A
microwave digestion system ETHOS 1 (Milestone, Sorisole,
BG, Italy), equipped with closed polytetrafluorethylene
(PTFE) vessels and sensors for temperature and pressure con-
trol, was also used for sample treatment.

Materials and Reagents

Analytical reagent-grade materials were used for all of the ex-

periments. All solutions were prepared using high-purity water
having a resistivity of 18.3 MΩ cm obtained from a Direct-Q 3
Water Purification System (Millipore Corporation, Bedford,
MA, USA). Working reference solutions of Ca, Cu, Fe, K,
and Na were prepared by appropriate dilutions of a stock solu-
tion containing 1 g L−1 (Cu and Fe) or 4 g L−1 (Fe, K, and Na)
in 1 % (v/v) HNO3. The nitric acid was purified by doubly sub-
boiling distillation in a quartz system MA-075 (Marconi,
Brazil). Sulfuric acid (Sigma-Aldrich, Germany) and hydrogen
peroxide 30 % (v/v) (Merck, Darmstadt, Germany) were also
used. All glass apparatus was conventionally washed before
used, soaked in 10 % (v/v) HNO3 for at least 48 h, and then
rinsed with ultrapure water prior to use.

Samples and Procedures

For the development of the proposed procedure and determina-

tion of analyte concentrations, four samples of milk powder were
used: two for adult consumption (whole and skimmed) and two
for children consumption all acquired in a local market. To eval-
uate the accuracy, SRM 1846 (Infant Formula), produced by the
National Institute of Standards and Technology (NIST,
Gaithersburg, MD, USA), was used. For analysis by ICP-OES,
the samples were prepared using three different procedures in
order to compare the results, as described below. Fig. 1 Schematic diagram of cold finger coupled to glass digester tube.
1: water inlet; 2: water outlet; 3: end cap of PTFE; 4: outlet to gas; 5:
Acid Decomposition in an Open System with a Cold Finger reaction flask
(Procedure 1)
temperature, the final volume of 40 mL was completed with
Approximately 1 g of milk powder sample was weighed in ultrapure water for subsequent analysis. For each sample,
the digestion tube, and 10.0 mL of acid mixture HNO3/ this procedure was carried out in triplicate. The conditions
H2SO4 (3:1) was added. The reflux system was coupled on used in this procedure (acid mixture, time, and temperature
the tubes, and the mixture was heated in a digester block at for decomposition) were previously optimized as a function
350 °C for 4 h in this temperature. After cooling to room of residual organic carbon content.
780 Food Anal. Methods (2016) 9:777–784

30
Conventional Open System (Procedure 2) HNO3
HNO3/HCl (3:1)
25 HNO3/H2SO4 (3:1)
This procedure was carried out using the same amount of HNO3/HCl (1:3)
sample and acid mixture of procedure 1. However, because 20

TOC (%)
it is an open system and to prevent dryness of the reagents in 15
the digestion tube, the heating temperature of the digester
block did not exceed 120 °C (boiling temperature of HNO3). 10
For this method, the same decomposition time described in 5
procedure 1 (4 h) was also adopted. After cooling at room
temperature, the solutions were made up to 40 mL with ultra- 0
pure water for subsequent analysis. For each sample, this pro- 100 150 200 250 300 350
o
cedure was carried out in triplicate. Temperature ( C)
Fig. 2 Effect of acid mixture and heating temperature of the digester
block for decomposition of milk powder samples using a reflux system
Microwave Digestion (Procedure 3)
As can be seen in Fig. 2, the effect of the heating temper-
For this procedure, approximately 350 mg of sample was ac-
ature of the digester block for the acid mixture HNO3/HCl
curately weighed into PTFE vessels and 3 mL of HNO3 65 %
(1:3) was performed only up to 150 °C, as a result of the
(v/v), 1 mL of H2O2 30 % (v/v), and 1 mL of deionized water
solution showing a high turbulence, with projection of the
were added (Ribeiro et al. 2003). The closed vessels were then
reaction in the digester tube at higher temperatures. The results
placed inside the microwave oven, and the decomposition was
for the acid mixture HCl/HNO3 (1:3) showed no significant
carried out according to the heating program, in which the
differences in the temperature range studied, giving an aver-
sample was initially heated from room temperature to 80 °C
age residual carbon content of approximately 19 %. When
in 4 min (hold for 6 min), then to 120 °C in 4 min (hold 6 min),
only HNO3 was used as oxidant medium to prepare the sam-
and finally to 180 °C in 10 min (hold for 15 min).
ples, there was a decrease in the residual carbon content along
Subsequently, the samples were left to cool, transferred to
the temperature range studied. This decrease was lower when
individual 25.0-mL volumetric flasks, and diluted with deion-
compared to the efficiency obtained for acid mixture HNO3/
ized water for subsequent analysis. For each sample, this pro-
H2SO4 (3:1), in which the higher temperature studied (350 °C)
cedure was carried out in triplicate.
presented a residual carbon content of approximately 6 %,
which is extremely low for the decomposition of samples
using open system. This value was close to those obtained
Results and Discussion with the microwave system, in the range of 6.8 %.
Thus, the acidic mixture HNO3/H2SO4 and a temperature
Effect of Acidity and Heating Temperature of 350 °C were the best conditions for the decomposition
process of milk powder samples. This acidic mixture was also
Many methods involving atomic spectrometric techniques for employed by Santos et al. 2005 for the decomposition of liq-
the determination of elements in solid samples require the uid milk and subsequent determination of eight metallic ele-
sample decomposition. So, it is important to know the relevant ments by ICP-OES. In this study, the authors used 4-mL mix-
factors for the complete decomposition of samples, particular- ture of HNO3/H2SO4 (3:1) for the decomposition of 2.5-mL
ly to know about the interferences resulting from the residual sample in a focused microwave oven, obtaining residual car-
carbon content after incomplete sample decomposition be- bon contents lower than 2 %. Moreover, other studies can also
cause this can affect the accuracy and repeatability of the be found in the literature using this acidic mixture, but in
entire analytical procedure (Wasilewska et al. 2002). In this different proportions for the decomposition of vegetables
way, the effect of the acidity and of heating temperature of the (Momen et al. 2006), bean (Costa et al. 2006), bovine liver
digester block on the decomposition of milk powder samples (Trevizan et al. 2003), and nuts (Momen et al. 2007).
was studied as a function of residual carbon content in order to In order to compare the acid decomposition using the reflux
evaluate the efficiency of decomposition of these samples system, the optimized conditions of acidic mixture and tem-
with the use of the cold finger system. For this, the use of perature were also applied to the conventional open system
HNO3 and their mixtures HNO 3/HCl (1:3), HCl/HNO3 (procedure 2). However, when temperatures above 120 °C
(1:3), and HNO3/H2SO4 (3:1) was evaluated. Each condition were applied to the digester block, it was observed that the
was subjected to different temperatures in the digester block HNO3 evaporated quickly and the remaining H2SO4 caused
that ranged from 100 to 350 °C and held for 3 h. Figure 2 the carbonization of the milk samples. Based on this fact, it
shows the results. was evident the efficiency of the reflux system in
Food Anal. Methods (2016) 9:777–784 781

25
acidity of the solutions of 2.18±0.01 mol L−1, equivalent to a
residual acidity of 53 %.
TOC (%) 20

15
Analytical Results
10
Analytical features for the calibration curves for Ca, Cu, Fe,
5 K, and Na using different sample treatments are shown in
Table 2. Good linear correlation coefficients in the curves
0 were obtained (R>0.99) independent of the method used for
30 60 90 120 150 180 210 240 sample preparation. Calibration curves were prepared using
Time (min) aqueous standard solutions in 2 % (v/v) HNO3 and applied
Fig. 3 Effect of heating time on the decomposition of milk powder independent of sample preparation procedure. Considering
samples using a reflux system and acidic mixture HNO3/H2SO4 (3:1) that the signals obtained in this study were stable for all
analytes, internal standard was not used during the measure-
condensation of acid vapors generated inside the digestion ments. The limit of detection (LOD) and limit of quantifica-
tube, taking full advantage the ability of decomposition and tion (LOQ) were calculated as being three or ten times the
minimizing loss of reagents and analytes when high tempera- standard deviation of ten measurements of the blank, for each
ture heating is applied in the digester block. procedure divided by the slope of the calibration curve, con-
sidering the mass and the sample dilution factor used.
Effect of Heating Time The limits of detection and quantification were identical for
procedures 1 and 2, since the same samples of mass and final
The effect of heating time in the acidic decomposition using volume of solutions were used for these procedures, and these
reflux system was studied as a function of residual carbon values are different from those found in procedure 3, consid-
content. For this, the milk powder samples were prepared ering that the sample mass and final volume of the solutions
using acid mixtures of HNO3/H2SO4 (3:1) and 350 °C at the were lower. LOD and LOQ values for the procedures 1 and 2
digester block. The decomposition time was evaluated from were lower, since it is possible to use a larger mass of sample
30 to 240 min, and the obtained results are shown in Fig. 3. because the systems operate at atmospheric pressure. Due to
In accordance to Fig. 3, a high residual carbon content was this fact, it becomes a limitation for closed systems on account
observed during 30 min of decomposition, showing that in of when using a larger sample mass, an increased internal
this time, the complete decomposition has not yet been pressure of the reaction flask may occur, increasing the risk
achieved. Above that time, there was a considerable decrease of explosion or opening the septum of security, loss of re-
in the percentage of residual carbon content, and similar re- agents and analytes, etc. Thus, this limitation is directly
sults were observed between 150 and 240 min. Thus, the time reflected in the limits of detections for the closed procedure
of 4 h (240 min) was selected for further studies in order to studied as shown in Table 2. A certified reference material of
ensure higher efficiency of decomposition for milk powder infant formula (CRM 1846) was used to evaluate the efficien-
samples and final solutions without residues, using an open cy of the digestion using the cold finger system and also of
system with a cold finger. In addition, it was observed a stan- other procedures employed (Table 3). The application of the t
dard deviation of 0.95, which correspond to a relative standard test with confidence level of 95 % showed good compliance
deviation of 15 %, indicating a good precision of the proce- between the measured values obtained with different sample
dure. Under these conditions, it was possible to obtain a final treatments and the certified values, proving the accuracy of the

Table 2 Analytical features for

the determination of Ca, Cu, Fe, Analyte Linear range (mg L−1) s (L mg−1) R Procedures 1 and 2 Procedure 3
K, and Na by ICP-OES in milk
powder samples LOD LOQ LOD LOQ

Ca 5–200 1855 0.9999 2.35 7.48 4.20 14.00

Cu 0.01–1.00 8225 0.9999 0.09 0.31 0.16 0.57
Fe 0.5–5.0 1389 0.9997 0.30 1.00 0.53 1.78
K 5–200 1548 0.9999 2.46 8.20 4.39 14.65
Na 5–200 244.3 0.9985 7.74 25.81 13.83 46.08

s sensitivity of the calibration curve, LOD (mg kg−1 ) limit of detection, LOQ (mg kg−1 ) limit of quantification, R
correlation coefficient
782 Food Anal. Methods (2016) 9:777–784

Table 3 Analytical results for

Ca, Cu, Fe, K, and Na in certified Analyte Label (mg kg−1) Concentration, x±SD (RSD) mg kg−1 (%)
reference material (CRM 1846)
after different sample treatments Procedure 1 Procedure 2 Procedure 3
(n=3)
Whole milk (adult)
Ca 9461 8694±601 (6.9) 8919±366 (4.1) 9188±138 (1.5)
Cu n.i. 0.33±0.04 (12.2) 0.32±0.02 (6.3) 0.40±0.08 (20.0)
Fe 200 223±14 (6.3) 225±8 (3.5) 226±2 (0.9)
K n.i. 11,154±519 (4.6) 11,423±350 (3.1) 11,687±91 (0.7)
Na 3653 3148±72 (2.3) 3318±155 (4.7) 3200±44 (1.4)
Skim milk (adult)
Ca 25,000 21,690±200 (0.9) 20,429±1227 (6.0) 22,239±506 (2.3)
Cu n.i. 0.47±0.02 (4.2) 0.46±0.08 (17.4) 0.52±0.04 (7.7)
Fe n.i. 9.27±1.56 (16.8) 8.34±0.13 (1.5) 9.32±1.23 (13.2)
K n.i. 15,336±79 (0.5) 15,362±1474 (9.5) 15,652±593 (3.8)
Na 4700.00 4535±28 (0.6) 4295±279 (6.5) 4679±149 (3.1)
Infant formula 1
Ca 8310 7215±185 (2.6) 7560±26 (0.3) 7653±130 (1.7)
Cu n.i. 0.18±0.01 (5.5) 0.16±0.04 (25.0) 0.24±0.06 (25.0)
Fe 69 68±2 (2.9) 70.8±0.4 (0.6) 70.2±1.5 (2.1)
K 7344 6120±497 (8.1) 6881±135 (2.0) 6215±224 (3.6)
Na 2517 1910±115 (6.0) 2092±7 (0.3) 2039±18 (0.9)
Infant formula 2
Ca 3200 2768±283 (10.2) 2650±56 (2.1) 2797±143 (5.1)
Cu 4 4.2±0.1 (2.4) 4.2±0.1 (2.4) 4.4±0.1 (2.3)
Fe 55 58.9±2.4 (4.1) 57±4 (7.0) 57.4±0.7 (1.2)
K 4800 3335±55 (1.6) 3380±34 (1.0) 3320±205 (6.2)
Na 1300 1270±44 (3.5) 1254±20 (1.6) 1343±17 (1.3)

Values in milligram per kilogram

n.i. not informed

results for all analytes studied and also that there is no matrix were submitted to a statistical paired t test with a 95 % confi-
effect. Silva et al. (2004) analyzed the bovine milk sample and dence level, and in general, no significant differences between
also does not find problems with matrix effects. the results were observed, except for Ca in the infant formula
In order to obtain information about the concentration of sample 1 and for Na in the infant formula sample 2, which
the analytes in different milk powder samples, the same was showed a discrepancy between the found values.
treated using the three procedures and the results obtained are In general, all concentration values found were closed to
shown in Table 4. those reported in the label of the samples of powder milk,
Concentrations obtained from the analysis of milk powder which confirms no adulteration for these analytes, and hence
samples using the three procedures of sample preparation allowing them to be suitable for marketing and consumption.

Table 4 Analytical results for the determination of Ca, Cu, Fe, K, and Na in milk powder samples by ICP-OES (n=3)

Certified value Found value, x±SD (RSD) (%)

Procedure 1 Rec. (%) Procedure 2 Rec. (%) Procedure 3 Rec. (%) (%)

Ca 3670±200 3365±178 (5.3) 91.7 3391±52 (1.5) 92.4 3455±119 (3.4) 94.2
Cu 5.04±0.27 4.73±0.14 (3.0) 93.8 4.74±0.06 (1.3) 94.0 4.64±0.13 (2.8) 92.1
Fe 63.1±4.0 58.2±2.4 (4.1) 92.2 58.7±0.7 (1.2) 93.1 58.3±5.3 (9.1) 92.4
K 7160±380 6763±284 (4.2) 94.4 6764±103 (1.5) 94.5 6814±271 (4.0) 95.2
Na 2310±130 2198±87 (3.9) 95.2 2161±20 (0.9) 93.5 2133±51 (2.4) 92.3
Food Anal. Methods (2016) 9:777–784 783

In general, good precision was obtained with average of rela- References

tive standard deviation (RSD) of 5 % for all procedures
employed. For some samples, the RSD values were high Buldini PL, Cavalli S, Sharma JL (2002) Matrix removal for the ion
(>10.0 %), because the concentrations found were close to chromatographic determination of some trace elements in milk.
Microchem J 72:277–284
the LOQs, and inevitably, the precision was impaired. For
Castro JT, Santos EC, Santos WPC, Costa LM, Korn M, Nóbrega JA,
the open system with cold finger, the RSD was lower than Korn MGA (2009) A critical evaluation of digestion procedures for
11.3 %. coffee samples using diluted nitric acid in closed vessels for induc-
tively coupled plasma optical emission spectrometry. Talanta 78:
1378–1382
Costa LM, Korn MGA, Castro JT, Santos WPC, Carvalho EV, Nogueira
Conclusions ARA (2006) Planejamento fatorial aplicado à digestão de amostras
de feijão assistida por radiação micro-ondas. Quim Nov. 29:149–
The acid decomposition using a reflux system was evaluated 152
as a sample preparation procedure for determination of Ca, Fernandes TAP, Brito JAA, Gonçalves LML (2015) Analysis of
micronutrients and heavy metals in Portuguese infant milk powders
Cu, Fe, K, and Na in milk powder samples by ICP-OES.
by wavelength dispersive X-ray fluorescence spectrometry
The obtained results were compared with those of convention- (WDXRF). Food Anal Methods 8:52–57
al methods of sample preparation and validated with satisfac- Ferrão MF, Mello C, Borin A, Maretto DA, Poppi RJ (2007) LS-SVM:
tory results for the analysis of CRM. However, lower values of uma nova ferramenta quimiométrica para regressão multivariada.
LOD and LOQ were obtained for procedures 1 and 2, which Comparação de modelos de regressão LS-SVM e PLS na
quantificação de adulterantes em leite em pó empregando NIR.
are open digestion systems and have no limitations regarding Quim Nov. 30:852–859
the use of larger sample masses during the preparation step. Ferreira SLC, Silva LOB, Santana FA, Junior MMS, Matos GD, Santos
Based on these studies to determine the residual carbon WNL (2013) A review of reflux systems using cold finger for sam-
content, it was observed that procedures 1 and 3 presented a ple preparation in the determination of volatile elements. Microchem
J 106:307–310
content of approximately 6 %, indicating that the acidic mix-
Freschi GPG, Fortunato FM, Freschi CD, Neto JAG (2012) Simultaneous
ture and the conditions of time and temperature used showed and direct determination of As, Bi, Pb, Sb, and Se and Co, Cr, Cu,
excellent efficiency for the decomposition of milk powder Fe, and Mn in milk by electrothermal atomic absorption spectrom-
samples using the procedure with reflux system, in compari- etry. Food Anal Methods 5:861–866
son to a closed system that utilizes the microwave oven. The Gouveia ST, Silva FV, Costa LM, Nogueira ARA, Nóbrega JA (2001)
Determination of residual carbon by inductively coupled plasma
final acidity was 2.18 mol/L that means that the losses of acid optical emission spectrometry with axial and radial view configura-
by evaporation were minimized. Thus, the reflux system tions. Anal Chim Acta 445:269–275
showed, once again, its technical features to minimize analyte Ikem A, Nwankwoala A, Odueyungbo S, Nyavor K, Egiebor N (2002)
loss and/or acid by volatilization combined with simplicity Levels of 26 elements in infant formula from USA, UK, and Nigeria
by microwave digestion and ICP-OES. Food Chem 77:439–447
and low operating cost. The use of reflux system in the digest-
Inam R, Somer G (2000) Direct method for the determination of selenium
er block proved to be an excellent alternative for laboratories and lead in cow's milk by differential pulse stripping voltammetry.
that do not have a microwave oven. Certainly, the procedure Food Chem 69:345–350
may be expanded for the determination of other analytes as Korn MGA, Boa Morte ES, Santos DCM, Castro JT, Barbosa JTP,
well as the analysis of other food samples by atomic spectrom- Teixeira AP, Fernandes AP, Welz B, Santos WPC, Santos EBGN,
Korn M (2008) Sample preparation for the determination of metals
etry techniques. in food samples using spectroanalytical methods—a review. Appl
Spectrosc Rev 43:67–92
Acknowledgments The authors gratefully acknowledge the Conselho Lara PCP, Silveira JN, Neto WB, Beinner MA, Da silva JBB (2014) Use
Nacional de Desenvolvimento Científico e Tecnológico (CNPq), the of multivariate optimization to develop methods for direct copper
Fundação de Amparo à Pesquisa do Estado de São Paulo, and and lead determination in breast milk by graphite furnace atomic
CNPq/CAPES (Projeto Casadinho no. 552197/2011-4) for their financial absorption spectrometry. Food Anal Methods 7:790–797
support and scholarships. McKinstry PJ, Indyk HE, Kim ND (1999) The determination of major
and minor elements in milk and infant formula by slurry nebuliza-
Conflict of Interest Eliézer Quadro Oreste has no conflict of interest. tion and inductively coupled plasma–optical emission spectrometry
Alexander Ossanes de Souza has no conflict of interest. Camila Côrrea (ICP-OES). Food Chem 65:245–252
Pereira has no conflict of interest. Meibel Teixeira Lisboa has no conflict Mitra S (2003) Sample preparation techniques in analytical chemistry.
of interest. Mirla Janaína A. Cidade has no conflict of interest. Mariana Wiley, New Jersey
Antunes Vieira has no conflict of interest. Solange Cadore has no conflict Momen AA, Zachariadis GA, Anthemidis AN, Stratis JA (2006)
of interest. Anderson Schwingel Ribeiro has no conflict of interest. Investigation of four digestion procedures for multi-element deter-
mination of toxic and nutrient elements in legumes by inductively
Ethical Approval This article does not contain any studies with human coupled plasma-optical emission spectrometry. Anal Chim Acta
or animal subjects. 565:81–88
Momen AA, Zachariadis GA, Anthemidis AN, Stratis JA (2007) Use of
Informed Consent Informed consent was obtained from all individual fractional factorial design for optimization of digestion procedures
participants included in the study. followed by multi-element determination of essential and non-
784 Food Anal. Methods (2016) 9:777–784

essential elements in nuts using ICP-OES technique. Talanta 71: Ribeiro AS, Moretto AL, Arruda MAZ, Cadore S (2003) Analysis of
443–451 powdered coffee and milk by ICP OES after sample treatment with
Nemati K, Bakar NKA, Abas MRB, Sobhanzadeh E, Low KH (2010) tetramethylammonium hydroxide. Microchim Acta 141:149–155
Comparative study on open system digestion and microwave Santos DM, Pedroso MM, Costa LM, Nogueira ARA, Nóbrega JA
assisted digestion methods for metal determination in shrimp sludge (2005) A new procedure for bovine milk digestion in a focused
compost. J Hazard Mater 182:453–459 microwave oven: gradual sample addition to pre-heated acid.
Oliveira E (2003) Sample preparation for atomic spectrometry: evolution Talanta 65:505–510
and future trends. J Braz Chem Soc 14:174–182 Silva JCJ, Santos DM, Cadore S, Nobrega JA, Baccan N (2004)
Oreste EQ, Jesus A, Oliveira RM, Silva MM, Vieira MA, Ribeiro AS Evaluation of inductively coupled plasma optical emission spec-
(2013a) New design of cold finger for sample preparation in open trometers with axial configuration: interfaces with end-on gas and
system: determination of Hg in biological samples by CV-AAS. shear gas. Microchem J 77:185–190
Microchem J 109:5–9 Tokalioğlu Ş, Gürbüz F (2010) Selective determination of copper and iron
Oreste EQ, Oliveira RM, Nunes AM, Vieira MA, Ribeiro AS (2013b) in various food samples by the solid phase extraction. Food Chem
Sample preparation methods for determination of Cd, Pb and Sn in 123:183–187
meat samples by GF AAS: use of acid digestion associated with a Trevizan LC, Nogueira AR, Nóbrega JA (2003) Single vessel procedure
cold finger apparatus versus solubilization methods. Anal Methods for acid vapor partial digestion of bovine liver in a focused micro-
5:1590–1595 wave: multielement determination by ICP-OES. Talanta 61:81–86
Pereira JSF, Pereira LSF, Schmidt L, Moreira CM, Barin JS, Flores EMM Veloso ACA, Teixeira N, Ferreira IMPLVO, Ferreira MA (2002)
(2013) Metals determination in milk powder samples for adult and Detecção de adulterações em produtos alimentares contendo leite
infant nutrition after focused-microwave induced combustion. e/ou proteínas lácteas. Quim Nov. 25:609–615
Microchem J 109:29–35 Wasilewska M, Goessler W, Zischka M, Maichin B, Knapp G (2002)
Pinheiro ACA, Lisboa MT, Ribeiro AS, Nunes AM, Yamasaki A (2014) Efficiency of oxidation in wet digestion procedures and influence
Avaliação da mineralização de arroz em sistema de refluxo para a from the residual organic carbon content on selected techniques for
determinação de Cu, Fe, Mn e Zn por F AAS. Quim Nov. 37:6–9 determination of trace elements. J Anal At Spectrom 17:1121–1125