a concentrated HNO
solution with 1 mL of a 30%, (m/v) H
solution were added. Samples were subjected to the microwaveheating with a maximum power of 600 W for 45 min. Aftercooling, colourless solutions were quantitatively transferredinto 25-mL volumetric ﬂasks and made up to the volume withdeionised water.
Hot plate heating digestion (P2)
: 0.5 g of coffee samples wereplaced into 150-mL Pyrex beakers and 10 mL of a concentratedHNO
solution was added. Solutions were heated on a hot plateat 85
C for about 3 h. Then they were cooled and 5 mL of a 30%(m/v) H
solution was added. The heating was continueduntil clear solutions were obtained and their volumes werereduced to about 2 mL. Resulting aliquots were quantitativelytransferred into 25-mL volumetric ﬂasks and made up to thevolume with deionised water.
Solubilisation in aqua regia (P3)
: 0.5 g of coffee samples wereweighed into 30-mL polypropylene (PP) centrifuge tubes and2 mL of
was added. Resulting mixtures were soni-cated in an ultrasonic bath for 15 min, then made up to 25 mL with deionised water and ﬁnally centrifuged for 10 min at12,000 rpm.
:0.5 gofcoffeesampleswereweighedinto 30-mL PP centrifuge tubes and 0.5 mL of a 25% (m/v) TMAHaqueous solution was added. Tubes were shaken at 80
C for30 min,cooled,madeupto25 mLwithdeionisedwaterandthencentrifuged for 10 min at 12,000 rpm.
Dissolution in HNO
: 1.0 g of coffee samples were weightedinto 30-mL PP centrifuge tubes and dissolved in 10mL of a 2.0%(m/v)HNO
solutionadded.Resultingsamplesolutionswerecen-trifugedfor10 min at12,000 rpm toremoveany solid particles.
Dissolution in H
: 1.0 g of coffee samples were weightedinto 30-mL PP centrifuge tubes and dissolved in 10 mL of H
Oadded. Resulting sample solutions were centrifuged for10 min at 12,000 rpm to remove any solid particles.Until the analysis, sample solutions were kept at 4
C. The sam-ple of each instant slim coffee brand was analysed in triplicate(
Slim Coffee Classic
was used for the comparison of differentsample preparation procedures. With each set of sample solutions,blank sample solutions were prepared and subjected to all sampletreatments to correct ﬁnal results. The chosen sample preparationprocedure was applied then to other coffee products prior to thedetermination of 13 elements in these samples by ICP OES againstthe external calibration with simple standard solutions.
An inductively coupled plasma sequential spectrometer JY38S(Jobin Yvon, France) was used to measure total concentrations of Ba, Ca, Cd, Cr, Cu, Fe, Mg, Mn, Ni, P, Pb, Sr, Zn in all solutions. Oper-ating parameters were as follows: 1.2 kW of a RF power,15 L min
of a plasma ﬂow rate, 0.2 L min
of a sheath-gas ﬂowrate and 0.38 L min
of a carrier-gas ﬂow rate. Sample solutionswere introduced into the plasma using a V-groove nebuliser anda Scott type spray chamber at a ﬂow rate of 1.1 mL min
. Analyt-ical lines of Ba II 233.53 nm, Ca I 317.93 nm, Cd II 228.80 nm, Cr II267.72 nm, Cu I 324.75 nm, Fe II 259.94 nm, Mg I 285.21 nm, Mn II259.37 nm, Ni II 221.65 nm, P I 214.65 nm, Pb II 220.35 nm, Sr II407.77 nm and Zn I 213.86 nm were measured.A Milestone (Italy) high-pressure microwave digestion system(MLS-1200 MEGA), equipped with a rotor MDR 300/10, was usedfor the microwave-assisted sample decomposition. An Elpin (Po-land) thermostatic water bath shaker (type 357) and a JP Selecta(Spain) ultrasonic cleaning bath with heating (UltrasonsH) wereused for experiments with
and TMAH based samplesolubilisations.AMPW-350centrifuge(MPWMed.Instruments,Po-land)wasusedtoacceleratetheseparationofliquidphasesfromso-lid particles.
2.5. Statistical methods
Statistical analysis of the obtained results was performedaccording to
-test at the 95% signiﬁcance level. The degree of therelationship between concentration of elements in analysed slimcoffees was evaluated by calculating Pearson’s
3. Results and discussion
3.1. Comparison of sample preparation procedures
The analytical characteristic of sample preparation procedurescompared here was evaluated by assessing the precision and theaccuracy of ICP OES measurements of Ba, Ca, Cd, Cr, Cu, Fe, Mg,Mn, Ni, P, Pb, Sr and Zn concentrations in sample solutions of one selected instant coffee brand (
Slim Coffee Classic
). The relativestandard deviation (RSD) for three independently prepared sam-ples was used to express the precision of the determination of ele-ment concentrations. The accuracy was evaluated by comparingtotal concentrations of elements obtained using different samplepreparation procedures (P2, P3, P4, P5, P6) with those achievedusing the microwave-assisted wet digestion procedure (P1), re-garded as a well-established tool that provides the total sampledecomposition without losses of analytes and treated as a refer-ence method (Kingston & Jassie, 1988). In addition, the recoverystudy was carried out.
The accuracy of examined sample preparation procedures wasveriﬁed by the recovery test. Accordingly, samples of
were spiked with a multi-element standard solution attwo different concentrations (0.5 and 1.0
) and subjectedtoallsamplepreparationprocedures.Experimentswereperformedin triplicate and resulting sample solutions were analysed to assessrecoveries of elements added. It was established that, indepen-dently of the spike level, recoveries are, in general, between96.5–106% (P1), 93.7–108% (P2), 97.9–104% (P3), 88.9–123% (P4),79.1–105% (P5) and 84.7–116% (P6). As it can be seen, results forthree ﬁrst procedures, i.e., microwave-assisted (P1) and hot-plate(P2) digestions in the mixture of HNO
and the solubili-sation in
(P3), show the most adequate accuracy andprovide quantitatively recoveries for all elements studied. Addi-tionally, for these procedures, slopes of standard additions’ calibra-tion curves are comparable to those found by the externalcalibration curves with simple standard solutions. Hence, in prac-tice, it can be concluded that matrix effects are not observed forthese procedures. Recovery values for other procedures investi-gated here, i.e., the solubilisation in TMAH (P4), and dissolutionsin a diluted HNO
solution (P5) and H
O only (P6), indicate thatthey are not suitable for direct ICP OES measurements of elementconcentrations in resulting sample solutions.Elemental contents of coffee sample obtained using differentsampletreatmentsarelistedinTable1. Consideringconcentrationsofallelementsstudied,itwasfoundthattheapplicationofthesolu-bilisationin
-testatthe95%levelofthesigniﬁcance,whichwasemployedtocompareconcentrationsofelementsresultedfromdifferentsampleprepara-tionprocedures to those resulted from the microwave-assisted wetdigestion(P1),differencesbetweenresultsobtainedusingtherefer-enceprocedure(P1)andthoseachievedusingthesecond(P2)orthe
A. Szymczycha-Madeja et al./Food Chemistry 146 (2014) 220–225