Open Access

Validation and measurement uncertainty evaluation of the ICP-OES method for the multi-elemental determination of essential and nonessential elements from medicinal plants and their aqueous extracts

  • Marin Senila1Email author,
  • Andreja Drolc2,
  • Albin Pintar2,
  • Lacrimioara Senila1 and
  • Erika Levei1
Journal of Analytical Science and Technology20145:37

https://doi.org/10.1186/s40543-014-0037-y

Received: 30 March 2014

Accepted: 9 July 2014

Published: 14 August 2014

Abstract

Background

The paper presents the development, validation, and evaluation of measurement uncertainty of a method for quantitative determination of essential and nonessential elements in medicinal plants and their aqueous extracts by using inductively coupled plasma optical emission spectrometry.

Methods

The detailed validation of the analytical procedure and calculation of the measurement uncertainty budget allowed the recognition of the methods' critical points.

Results

The obtained limit of quantification, repeatability, and measurement uncertainty were satisfactory. The trueness of the method was verified by recovery estimation using certified reference materials. The recovery rates of all metals were between 95% and 105%.

Conclusions

The paper presents for the first time all the steps needed to evaluate the measurement uncertainty and validate the determination method of selected elements in medicinal plants and their aqueous extracts. In summary, the obtained results demonstrate that the method can be applied effectively for the designed purpose.

Keywords

ICP-OESMedicinal plantsMulti-elemental analysisValidationMeasurement uncertainty

Background

The inductively coupled plasma optical emission spectrometry (ICP-OES) is a strong tool for the determination of various elements in liquid and solid samples. Elevated concentrations of essential elements (e.g., Fe, Mn, Zn, Cr, Cu) and low concentrations of nonessential elements (e.g., Cd, Ni, As) may present a potential hazard for human health. When preparing tea by infusion of plants, metals can be leached into the water and consumed by humans. Therefore, the metal contents in the plant infusion should comply with the limit values set by the Drinking Water Directive (Council of the European Union [1998]).

Numerous plant species used as remedies in traditional medicine are grown as spontaneous flora (Chuparina and Aisueva [2011]). In Romania, among the most used medicinal plants in folk medicine are chamomile (Matricaria recutita), milfoil (Achillea millefolium), rattle (Hypericum perforatum), brotherwort (Thymus serpyllum), pot marigold (Calendula officinalis), linden (Tilia platyphyllos), and peppermint (Mentha piperita). Chamomile is used for its anxiolytic, antiseptic, and anti-inflammatory properties, while milfoil is used for its strong astringent effect and to treat a variety of illnesses and disorders from stomach aches to circulatory disorders. Rattle is usually used to treat digestive and neurological disorders; brotherwort has antiseptic properties and is used for acne and allergies treatment; pot marigold has anti-inflammatory and antitumor properties; linden is used for its calming effects and to ease cold and flu symptoms; while peppermint is used to relieve stomach aches, nausea, fever, stress, and to boost the immune system.

Since, all over the world, there are numerous metal-polluted sites (European Environment Agency [2007]), metals from soil can be transferred to the plants (Moreno-Jimenez et al. [2009]; Malandrino et al. [2011]; Senila et al. [2012]; Rodrigues et al. [2012]) and may have adverse effects on consumers' health, as local residents use the plants in their diet, mainly for tea preparation. Consequently, there is a need to develop reliable methods for the determination of metals in medicinal plants and their aqueous extracts.

The ICP-OES method has become a routine analytical technique for metal determination; however, the information related to method validation are scarce, and research on this field is still needed (Mermet [2005]). Several studies present the use of ICP-OES for metals determination in tea or other food samples (Mitic et al. [2012]; Froes et al. [2014]). For consistent interpretation of the measurement results, it is necessary to evaluate the confidence that can be placed in, therefore, the presentation of an analytical result which must be accompanied by indication of the data quality. This information is essential for the interpretation of the analytical result (Kessel [2002]; Drolc and Pintar [2011]). Method validation is an essential component of the measures that a laboratory should implement in order to produce reliable analytical data (EURACHEM [1998]). Besides common method performance characteristics obtained in the validation process, testing laboratories shall have and apply procedures for estimating the uncertainty of measurements (International Organization for Standardization [2005]). This clearly means that the analytical result cannot be viewed only as a separate value. The International Organization for Standardization (ISO) guide (International Organization for Standardization [1995]) recommends the calculation of uncertainty using a model equation, based on its uncertainty components, and by using the law of propagation of uncertainty in order to combine them into uncertainty. It has subsequently been interpreted for analytical chemistry (Ellison et al. [2012]). There are several possibilities to estimate the uncertainty, as reported in the literature (Ellison et al. [2012]; International Organization for Standardization [1995]; Magnusson et al. [2012]; Baralkiewicz et al. [2013]). The measurement uncertainty is estimated mainly by the top-down or bottom-up approaches. In the top-down approach, the major sources of uncertainty are identified and evaluated, while in the bottom-up approach, all the uncertainty sources are systematically evaluated and only those with significant contributions are used to derive the measurement uncertainty. The top-down approach is time-consuming and requires extensive knowledge of the analytical procedure, but it enables identification of major uncertainty sources and consequently reduction of total measurement uncertainty. Another relatively quick and easy way of uncertainty estimation is the in-house validation that includes the determination of the method performance parameters (Baralkiewicz et al. [2013]).

In spite of a several papers published on the topic, there is a lack of fully validated methods for metal determination in medicinal plants and their extracts. The purpose of the present work was to perform a detailed validation of the analytical procedure and estimate the measurement uncertainty budget for determination of some essential (Fe, Mn, Zn, Cr, Cu, Al, Mg) and toxic (Pb, Cd, Ni, As) elements in the medicinal plants and their aqueous extracts. The method was validated according to the international guidelines ISO/IEC 17025:2005 (International Organization for Standardization [2005]). The assessment of uncertainty was carried out using modelling approach and a full combined uncertainty calculation, including possible sources of uncertainty.

Methods

Instrumentation

Analyses were carried out using a dual viewing inductively coupled plasma optical emission spectrometer Optima 5300DV (PerkinElmer, Waltham, MA, USA) coupled to an ultrasonic nebulizer CETAC 6000AT+ (CETAC, Omaha, NE, USA). The operating conditions employed for ICP-OES determination were 1,300 W RF power, 15 L min−1 plasma flow, 2.0 L min−1 auxiliary flow, 0.8 L min−1 nebulizer flow, and 1.5 mL min−1 sample uptake rate. Axial view was used for metals determination, while 2-point background correction and six replicates were used to measure the analytical signal. In order to eliminate the memory effect caused by the use of ultrasonic nebulization, the delay time for washing between samples and signal measurement was set to 180 s. The measurement of a blank solution after measuring 1 mg L−1 calibration standard indicated the lack of memory effect. High-purity argon (99.995%) supplied by Linde Gas SRL (Timis, Timisoara, Romania) was used to sustain plasma and as carrier gas. A closed-vessel MWS-3+ microwave system (Berghof, Germany) with temperature control mode was used for sample digestion. All PTFE digestion vessels were previously cleaned in a bath of 10% (v/v) nitric solution for 48 h to avoid cross-contamination.

Reagents and CRMs

Multi-elemental solutions of 1,000 mg L−1 ICP Standard Certipur® (Merck, Darmstadt, Germany) containing the analysed elements (Fe, Mn, Al, Mg, Pb, Zn, Cr, Cu, Cd, Ni, and As) were used for calibration. Analytically graded 65% HNO3 and 30% H2O2 (Merck, Germany) were used for sample digestion. Ultrapure water obtained by a Milli-Q system (Millipore, Molsheim, France) was used for all dilutions and infusions. For metals' determination in plants and their aqueous extracts, the calibration standards were prepared by diluting the reference multi-elemental standard solution in 8% (v/v) nitric acid and 0.5% (v/v) nitric acid, respectively, in order to assure the similar concentration of nitric acid in samples and in calibration standards.

A vegetable certified reference material (CRM) IAEA-359 Cabbage (IAEA, Austria) and a water CRM trace metals 1-WP QC11132 (Sigma Aldrich, Steinheim, Germany) were used for the quality control of metals' determination.

Plant samples

Seven medicinal plants (chamomile, milfoil, rattle, brotherwort, pot marigold, linden, peppermint) were randomly collected from spontaneous flora grown in NW Romania. Three specimens of each plant species were intensely rinsed with tap water and distilled water and dried in an oven at 40°C until weight is constant. In order to accelerate the digestion process, samples were grinded to powder with a kitchen mixer grinder and sieved through a 100-μm mesh. Five sub-samples from each plant species were used for analysis.

Microwave digestion procedure

An amount of 0.5 g of plant powder was weighted into dry, clean PTFE vessels then 6 mL of HNO3 and 2 mL of H2O2 were added. Vessels content were mixed and kept at room temperature for 12 h, then the vessels were introduced in the microwave digestion system and digested using a four-step digestion program: (1) 5 min at 280 W, (2) 5 min at 700 W, (3) 10 min at 1,050 W, (4) 1 min at 0 W. The resulting solutions were cooled, diluted to 50 mL with distilled water, filtered, and then analysed by ICP-OES. In order to evaluate the accuracy of the method, the vegetable CRM was analysed in the same experimental conditions as the samples.

Aqueous extracts

An amount of 1 g of plant powder was prepared for infusion in 200 mL of boiling ultrapure water for approximately 10 min. The obtained infusions were filtered, evaporated to approximately 10 mL on a hot plate, then 1 mL of HNO3 was added and the samples were digested in the microwave digestion system using the same digestion program as for solid samples. After cooling, the obtained solutions were filtered and diluted with ultrapure water in 100 mL volumetric flasks and analysed by ICP-OES.

Results and discussion

Method validation

The validation of the analytical procedure for quantitative determination of elements in medicinal plants and their aqueous extracts was performed by evaluating selectivity, working and linear ranges, limit of detection (LoD), limit of quantification (LoQ), trueness, and precision (repeatability and reproducibility) (EURACHEM [1998]).

Selectivity

Selectivity is the ability of a method to accurately quantify the analyte in the presence of interferences, under the stated conditions of the assay for the sample matrix being studied (EURACHEM [1998]). The selectivity in the case of ICP-OES method is related to possible interferences of the emission spectrum at specific wavelengths. The emission lines used for quantitation of each element, based on known interferences and baseline signal at selected wavelengths observed empirically during the measurements, are presented in Table 1. Matrix effects were studied by standard addition method, by adding a spike of 1 mg L−1 of each element to the original samples. The recoveries were within 90% and 110% for all the studied elements.
Table 1

Wavelengths for selected elements, LoD, and LoQ in aqueous extracts and in dry plants

Element

Wavelength (nm)

Plant aqueous extractsa

Plant dry massb

LoD (mg L−1)

LoQ (mg L−1)

Target valuec(mg L−1)

LoD (mg kg−1)

LoQ (mg kg−1)

Target valued(mg kg−1)

Cd

228.805

0.13 × 10−3

0.43 × 10−3

0.50 × 10−3

0.019

0.063

0.05 × 10−3

Cr

267.713

0.60 × 10−3

2.00 × 10−3

5.00 × 10−3

0.075

0.25

-

Cu

327.398

0.75 × 10−3

2.50 × 10−3

200 × 10−3

0.12

0.40

-

Fe

238.205

2.24 × 10−3

7.46 × 10−3

20 × 10−3

0.25

0.83

-

Al

396.153

1.59 × 10−3

5.30 × 10−3

20 × 10−3

0.20

0.67

-

Mg

258.213

0.60 × 10−3

2.00 × 10−3

-

0.11

0.37

-

Pb

220.355

0.29 × 10−3

0.97 × 10−3

1.00 × 10−3

0.043

0.14

0.5 × 10−3

Mn

257.611

0.25 × 10−3

0.83 × 10−3

5.00 × 10−3

0.030

0.10

-

Ni

231.604

0.57 × 10−3

1.90 × 10−3

2.00 × 10−3

0.13

0.43

-

Zn

213.856

1.52 × 10−3

5.06 × 10−3

-

0.22

0.73

-

As

193.759

0.30 × 10−3

1.00 × 10−3

1.00 × 10−3

0.045

0.15

-

aCalculated for the extraction method (1 g of plant digested with nitric acid and perhydrol in 100 volumetric flask); bcalculated for the extraction method (0.5 g of plant extracted in water in 100 volumetric flask); c10% of the limit values according to Drinking Water Directive; d10% of the limit values according to the European Pharmacopeia.

LoD and LoQ in aqueous extracts and in dry plants

The LoD indicates the level at which detection becomes problematic, while LoQ is the lowest concentration of the analyte that can be determined with an acceptable level of repeatability, precision, and trueness. LoD was estimated from the calibration function for a signal equal to the net signal of blank and three times its standard deviation, while LoQ was estimated from the calibration function for a signal equal to the net signal of blank and ten times its standard deviation (EURACHEM [1998]; Miller and Miller [2000]). Standard deviation of the blank resulted from the analysis of ten independent reagent blank solutions, each measured once on the same day. As the metal content in tea is not legislated, the performance criteria targeted for the LoD for aqueous plant extracts were 10% of the limit values (μg L−1) for drinking water: As - 10; Cd - 5; Cr - 50; Cu - 2,000; Pb - 10; Ni - 20; Fe - 200; Al - 200; and Mn - 50 (Council of the European Union [1998]). The European Pharmacopeia (Council of Europe [2011]) proposed a limit of 5 mg kg−1 for Pb and 0.5 mg kg−1 for Cd in herbal drugs. For plant samples, the performance criteria targeted for the LoD were 10% of these values. Data in Table 1 showed that the performance targets were achieved by our methods. In order to experimentally confirm LoQ, six standard solutions with concentrations close to the LoQ were prepared and analysed. The targeted repeatability expressed as relative standard deviation (RSD) and targeted recovery were 20% and 90% to 115%, respectively. The measured RSD and recovery are presented in Table 2.
Table 2

Confirmation of LoQ in aqueous extracts and in plant dry mass

Element

Plant aqueous extracts

Plant dry mass

RSD (%)

Recovery (%)

RSD (%)

Recovery (%)

Cd

9.61

95.6

10.5

98.9

Cr

11.8

91.2

10.2

95.6

Cu

8.86

104

9.95

101

Fe

15.2

112

12.6

115

Al

9.12

104

9.05

110

Mg

8.24

98.6

8.87

96.6

Pb

11.6

94.6

12.2

97.2

Mn

10.1

108

8.96

110

Ni

14.2

114

15.1

104

Zn

8.89

98.8

10.4

96.3

As

14.5

106

12.8

112

Working and linear range

Working range is the range of analyte concentrations over which the method is linear. At the lower end of the concentration range, the limiting factor is LoQ, while at the upper end limitations are imposed by various effects depending on the instrument response. Although generally, three or four calibration standards are used to evaluate the linear range of ICP-OES method in order to evaluate the appropriate measurement uncertainty budget; in our study, linearity was evaluated from the regression function of calibration using eight standards, the lowest concentration close to the LoQ, while the others were 0.05, 0.10, 0.20, 0.40, 0.60, 0.80, and 1.00 mg L−1 for each element. The fit for purpose working range was selected to be between LoQ of each element and 1.00 mg L−1.

Ten replicates of the lowest and ten at the highest concentration of the working range were measured. To check the homogeneity of variances, the standard deviations (s1) and (s2) of the lowest and the highest concentrations from calibration curves and the PG ratios (s12/s22 or s22/s12) were calculated and compared with the critical value F9;9;0.99 = 5.35. The values for intercept (a), slope (b), correlation coefficient (r2), and PG ratio are presented in Table 3. The experimental data showed that the variances are homogenous; therefore, linear regression curve can be used (International Organization for Standardization [1990]).
Table 3

Calibration curves for working range LoQ to 1.00 mg L −1

Element

avalue

bvalue

r2value

PG

Cd

22,300

901,200

0.9999

1.96

Cr

12,090

868,000

0.9999

2.88

Cu

5,345

1,045,000

0.9997

4.33

Fe

23,700

1,503,900

0.9999

4.18

Al

−11,100

1,922,000

0.9999

3.12

Mg

−5,600

3,459,000

1.0000

1.72

Pb

129

184,800

0.9998

4.56

Mn

141,500

8,094,000

1.0000

1.96

Ni

15,400

111,300

0.9999

4.06

Zn

56,500

1,159,000

0.9999

3.36

As

−33

32,400

0.9999

4.67

Trueness

The most frequent approach to estimate trueness of the method is CRM analysis. Six parallel samples of water and vegetable CRMs were analysed in order to determine the method's trueness (Tables 4 and 5). These results showed that the recoveries for all elements were generally within ±5% of the certified values. The Student's t test confirmed that the obtained recoveries are not significantly different from 100%.
Table 4

Results of analysis of water CRM trace metals 1-WP QC11132 (Sigma Aldrich)

Element

Found value (μg L−1)

s(μg L−1)

Certified value (μg L−1)

UCRM(μg L−1)

Recovery (%)

Cd

390

3.10

385

5.07

101

Cr

862

25.7

864

11.6

99.8

Cu

592

8.00

603

7.80

98.2

Fe

1,350

27.0

1,340

24.4

101

Al

452

12.9

463

11.4

97.6

Pb

908

30.2

929

14.4

97.7

Mn

1,200

7.15

1,200

15.5

100

Ni

1,730

21.1

1,710

20.4

101

Zn

918

8.04

915

16.2

100

As

414

7.25

416

8.25

99.5

Table 5

Results of analysis of IAEA-359 cabbage CRM (IAEA)

Element

Certified content, μg g−1

Found contenta, μg g−1

Cd

0.115 to 0.125

0.116 ± 0.014

Cr

1.24 to 1.36

1.25 ± 0.088

Cu

5.49 to 5.85

5.44 ± 0.41

Fe

144.1 to 151.9

147 ± 4.53

Mn

31.3 to 32.5

32.4 ± 2.15

Ni

1.00 to 1.10

1.03 ± 0.094

Zn

37.9 to 39.3

37.8 ± 1.89

As

0.096 to 0.104

0.100 ± 0.012

Mg

2,110 to 2,210

2,114 ± 23.6

aValues are expressed in microgram per gram dry weight and reported as average ± s; n = 5; 95% confidence level.

Precision

The most common measures of precision are repeatability and reproducibility (Tables 6 and 7), which were estimated considering within and between days variation, respectively. The results obtained in repeatability were conducted on six parallel samples by a single operator using the same equipment. The set targets for concentrations lower than 100 μg L−1 were standard deviation of repeatability (sr) below 10% and limit of repeatability (r) below 28%, while for concentrations higher than 100 μg L−1, sr below 7% and r below 20%.
Table 6

Results from the repeatability study for two levels of concentration

Element

Average (μg L−1)

sr(%)

r(%)

Average (μg L−1)

sr(%)

r(%)

Cd

23.6

6.8

19

211

3.6

10

Cr

24.6

7.1

20

208

3.7

10

Cu

26.1

6.4

18

198

4.4

12

Fe

25.5

9.5

27

221

6.3

18

Al

25.0

6.2

17

213

3.8

11

Mg

25.3

4.1

11

200

2.9

8.1

Pb

24.6

7.1

20

222

3.2

9.0

Mn

24.8

7.4

21

232

4.1

11

Ni

26.3

7.8

22

215

5.2

15

Zn

25.1

6.9

19

225

4.7

13

As

24.3

8.8

25

208

6.5

18

sr, standard deviation of repeatability; r, limit of repeatability (sr × 2.8).

Table 7

Results obtained for the reproducibility by ICP-OES

Element

Average (μg L−1)

sR(%)

R(%)

Cd

107

9.8

27

Cr

98.5

10

28

Cu

101

12

34

Fe

94.8

14

39

Al

96.8

6.9

19

Mg

101

5.8

16

Pb

105

8.6

24

Mn

111

11

31

Ni

105

14

39

Zn

114

8.8

25

As

95.6

16

45

sR, standard deviation of reproducibility; R, limit of reproducibility (sR × 2.8).

Measurement uncertainty

Measurement uncertainty was evaluated based on the bottom-up approach (International Organization for Standardization [1995]). All the contributions were obtained from calibration certificates and from statistical analysis of repeated measurements. Trueness of the method was calculated from results of CRM analysis, while repeatability was evaluated from precision experiments. The uncertainty of volumetric operations (volumetric flasks, pipettes) was calculated by using manufacturer data on calibration uncertainty (from certificates), the uncertainty associated with the use of glassware at a temperature different from that of calibration, and the repeatability of volumetric deliveries. Uncertainty of balances was calculated from data obtained from calibration certificates (declared uncertainty) and the repeatability of weighing. After estimation, all sources of uncertainty were combined according to the law of propagation of uncertainties, obtaining the combined standard uncertainty (u(Ca)). The final result was reported as expanded uncertainty (U(Ca)), calculated as U(Ca) = k × u(Ca), where k is the coverage factor, corresponding to a 95% confidence level.

The identified main sources of measurement uncertainty were uncertainty of calibration reference materials (C i ), uncertainty of delivered volumes, uncertainty of measured intensities of the reference solutions (A i ), and recovery of the method (Figure 1).
Figure 1

Cause and effects diagram of uncertainties. The uncertainties in the measurement of mass concentration of elements in aqueous herbal extracts are obtained using ICP-OES.

The contributions of repeatability to the measurement uncertainty were combined into one contribution for the overall experiment and were obtained from the method validation study performed in the laboratory. Recovery accounts for possible interferences in the method when samples of selected matrix are analysed. With these corrections, the concentration of each element (C) in a sample was expressed by the model:
C = A a b 1 R F rep F dil
(1)
where R is the method recovery, Fdil is the dilution factor, Frep is the repeatability factor, A is the area of the sample, while a and b are the linear regression coefficients. The sources of uncertainty and uncertainty components in determining elements are schematically presented in the cause and effects diagram (Figure 1). Calculations were made by using GUM Workbench software version 1.3 (Metrodata GmbH, Grenzach-Wyhlen, Germany) which is a standard application program with the possibility that user can define any model equation in order to enable various uncertainty calculations. The software was checked and validated before use in order to demonstrate that it is suitable for intended use. The results of the measurement uncertainty are listed in Table 8. Results revealed that for all the metals tested in plant aqueous extracts, extended measurement uncertainty is lower than 10% and therefore fulfills the requirements stated in the Drinking Water Directive (Council of the European Union [1998]).
Table 8

Measurement uncertainty of elements determination in plant aqueous extracts and plant dry mass

Element

Measurement uncertainty, % (k = 2)

Plant aqueous extracts

Plant dry mass

Cd

5.8

8.7

Cr

7.5

8.8

Cu

5.7

8.9

Fe

8.2

8.3

Mn

6.2

6.4

Ni

6.6

8.8

Zn

5.9

6.3

As

7.7

9.8

The uncertainty components of Cr concentration in plant aqueous extracts and in plant dry mass as a case study are presented in Table 9.
Table 9

Uncertainty components of Cr in plant aqueous extracts and mass fraction in plant dry mass

Symbol

Unit

Plant aqueous extracts

Plant dry mass

Value

Standard uncertainty

r i

Value

Standard uncertainty

r i

C 1

mg L−1

0.00

0.004

0.5

0.00

0.004

0.5

C 2

mg L−1

0.05

0.004

0.5

0.05

0.004

0.5

C 3

mg L−1

0.10

0.004

0.4

0.10

0.004

0.5

C 4

mg L−1

0.20

0.004

0.4

0.20

0.004

0.4

C 5

mg L−1

0.40

0.004

0.3

0.40

0.004

0.3

C 6

mg L−1

0.60

0.004

0.3

0.60

0.004

0.3

C 7

mg L−1

0.80

0.004

0.0

0.80

0.004

0

C 8

mg L−1

1.00

0.004

0.1

1.00

0.004

0.1

A 1

-

20,190

3,028

0.4

20,190

3,028

0.4

A 2

-

55,030

1,045

0.0

55,030

1,045

0

A 3

-

95,540

840

0.0

95,540

840

0

A 4

-

181,800

1,181

0.0

181,800

1,181

0

A 5

-

355,180

1,882

0.0

355,180

1,882

0

A 6

-

529,800

3,284

0.2

529,800

3,284

0.2

A 7

-

703,882

3,097

0.1

703,882

3,097

0.1

A 8

-

880,011

5,104

0.1

880,011

5,104

0.1

A

-

255,400

1,127

0.1

31,140

4,670

0.1

R

-

1.00

0.015

20.4

1.00

0.027

35.8

F rep

-

1.00

0.029

76.2

1.00

0.029

60.7

C m

g L−1

-

-

-

10

0.001

0.0

Result

 

0.283 mg L−1

0.019 mg L−1 (6.7%, k = 2)

 

2.53 mg kg−1

0.202 mg kg−1 (9.0%, k = 2)

 

C1 to C8 are the concentrations of calibration standard solutions; A1 to A8 are the respective standard solutions; A is the emission intensity of the sample; R is the recovery from CRM; and Frep is repeatability factor; C m is the concentration of measured sample in digested solution.

The relative uncertainty variance contributions are used to illustrate the relative impact of different uncertainty components. The relative contribution (r i ) of an uncertainty component x i to the combined standard uncertainty is defined as follows:
r i = δy δ x i 2 u x i 2 u y 2
(2)

where u(xi) is the standard uncertainties of the input parameters, and ∂y/∂x i is the sensitivity coefficient.

The importance of uncertainty sources is determined by their quantitative effect on the measurement result. In case of Cr both for extracts and dry plants, the largest contribution comes from u(R) and from repeatability (u(Frep)), while uncertainty contributions from other input quantities are of minor importance.

Results on real samples (aqueous plant extract and dry plant)

The concentrations of essential and nonessential elements in the dry mass of the analysed plant samples are presented in Table 10. The As concentrations were, in all cases, below the LoQ, while Pb and Cd concentrations were below the proposed limits by European Pharmacopoeia of 5 and 0.5 mg kg−1, respectively. Ni concentrations varied between 0.63 and 6.02 mg kg−1. These results were in the same order of magnitude with those reported by other authors for herbal drugs collected in Europe (Razic et al. [2006]; Basgel and Erdemoglu [2006]; Gentscheva et al. [2010]).
Table 10

Contents of metals in dry plant mass

 

Mg

Al

Cd

Cr

Cu

Fe

Pb

Mn

Ni

Zn

As

Chamomile

2,390 ± 18.5

152 ± 8.25

0.084 ± 0.011

2.50 0.22

11.2 ± 0.078

144 ± 5.56

0.90 ± 0.11

219 ± 5.95

3.05 ± 0.23

22.4 ± 1.36

<0.15

Milfoil

1,620 ± 12.6

41.1 ± 3.12

0.25 ± 0.022

0.88 ± 0.08

15.7 ± 0.12

42.1 ± 3.27

0.68 ± 0.063

83.2 ± 4.12

5.14 ± 0.48

46.2 ± 3.17

<0.15

Rattle

1,380 ± 11.9

126 ± 4.88

0.14 ± 0.015

1.55 ± 0.16

19.1 ± 0.15

189 ± 8.48

0.69 ± 0.077

120 ± 9.26

2.87 ± 0.22

55.9 ± 3.38

<0.15

Brotherwort

2,230 ± 12.5

186 ± 10.2

0.35 ± 0.029

3.18 ± 0.31

21.9 ± 0.18

224 ± 11.4

2.23 ± 0.17

296 ± 11.3

6.02 ± 0.23

25.8 ± 1.57

<0.15

Pot marigold

3,120 ± 20.8

172 ± 11.4

0.11 ± 0.010

2.88 ± 0.23

29.1 ± 0.19

234 ± 20.7

1.33 ± 0.13

178 ± 9.59

4.08 ± 0.24

51.1 ± 2.96

<0.15

Linden

1,450 ± 11.8

72.6 ± 3.33

0.071 ± 0.008

1.50 ± 0.11

9.22 ± 0.066

64.9 ± 6.23

0.44 ± 0.042

71.9 ± 3.77

0.63 ± 0.071

18.8 ± 1.07

<0.15

Peppermint

2,660 ± 24.1

144 ± 6.03

0.41 ± 0.036

3.61 ± 0.30

19.9 ± 0.18

306 ± 12.2

2.53 ± 0.21

255 ± 14.4

2.11 ± 0.11

64.4 ± 3.22

<0.15

Values are expressed as milligrams per kilogram (mean ± standard deviation of five replicates).

The contents of essential elements such as Mg (1,450 to 3,120 mg kg−1), Al (41.1 to 186 mg kg−1), Cr (0.88 to 3.61 mg kg−1), Cu (9.22 to 29.1 mg kg−1), Fe (42.1 to 306 mg kg−1), Mn (71.9 to 296 mg kg−1), and Zn (22.4 to 64.4 mg kg−1) were similar with those reported in the literature (Gentscheva et. al. [2010]; Maharia et al. [2010]; Chuparina and Aisueva [2011]; Miranda and Pereira-Filho [2013]).

The content of metals in the aqueous plant extracts offers information about the uptake of these elements by drinking of a cup of tea. The concentration of essential and nonessential elements in the aqueous extracts (Table 11) offers information about the uptake of these elements following tea consumption. As and Cd concentrations were lower that the LoQ in all the analysed samples. Also, Pb concentrations were generally below the LoQ, while Ni concentrations ranged between 2.70 and 18.2 μg L−1, below the maximum value of 20 μg L−1 established for this element by EU Drinking Water Directive 98/83/EC. Also, the concentrations of essential elements that have established maximum values for drinking water were generally below these limits, except manganese extracted from brotherwort which slightly exceeded 50 μg L−1. The higher metal concentrations in plant extracts were found for Mg (1,590 to 7,800 μg L−1), but this element have no maximum admitted limit for drinking water. Our results for the concentrations of Al, Cu, Mg, and Fe are in line with those reported by Froes et al. ([2014]), but Mn concentrations were generally lower, in our case.
Table 11

Contents of metals in the aqueous extracts

 

Mg

Al

Cd

Cr

Cu

Fe

Pb

Mn

Ni

Zn

As

Chamomile

6,010 ± 110

115 ± 5.03

<0.43

2.20 ± 0.21

12.8 ± 0.80

45.5 ± 1.64

<0.33

49.0 ± 3.15

11.1 ± 1.10

70.0 ± 8.58

<1.00

Milfoil

1,060 ± 86.2

91.2 ± 6.22

<0.43

<2.00

27.0 ± 1.56

18.8 ± 1.10

<0.33

39.2 ± 3.24

7.15 ± 0.61

103 ± 7.12

<1.00

Rattle

2,220 ± 103

109 ± 4.10

<0.43

2.21 ± 0.22

64.1 ± 2.90

44.2 ± 1.30

<0.33

35.6 ± 2.27

2.70 ± 0.11

185 ± 10.1

<1.00

Brotherwort

4,600 ± 185

190 ± 8.85

<0.43

3.15 ± 0.30

76.2 ± 3.95

27.0 ± 1.81

0.66 ± 0.08

51.5 ± 4.03

18.2 ± 1.20

68.8 ± 5.22

<1.00

Pot marigold

7,800 ± 230

95.6 ± 5.22

<0.43

3.32 ± 0.34

70.0 ± 3.65

34.5 ± 2.16

0.53 ± 0.08

28.6 ± 2.05

9.58 ± 0.74

163 ± 6.96

<1.00

Linden

1,590 ± 81.6

53.9 ± 2.76

<0.43

<2.00

10.6 ± 0.96

22.2 ± 1.31

<0.33

25.9 ± 1.89

3.55 ± 0.24

29.3 ± 2.11

<1.00

Peppermint

5,650 ± 211

127 ± 3.05

<0.43

4.25 ± 0.39

70.7 ± 3.62

51.0 ± 3.10

0.83 ± 0.09

45.5 ± 2.63

6.21 ± 0.53

196 ± 11.2

<1.00

Values are expressed as micrograms per liter (mean ± standard deviation of five replicates) per 1 g of plant extracted in 200 mL water.

By comparing the results presented in Tables 10 and 11, taking into account the mass of dry plants and final volume of infusion, it can be observed that among the analysed elements, Zn, Cu, and Ni were highly extracted in the aqueous extracts (31% to 64%, 23% to 71%, and respectively, 19% to 73%), in function of the plant species, while Pb and Fe had low solubility (below 10%).

Conclusions

A fully validated method for metal analysis in medicinal dry plant mass and its extracts is presented. The fast and accurate ICP-OES method enables the quantification of selected metals in aqueous and dry samples.

The validation results are presented and organized in tables in order to provide an easy overview of the method's performance. The experimentally determined validation parameters for medicinal extracts were then compared to the criteria stated in the Drinking Water Directive. Measurement uncertainty was determined on basis of modelling approach. Detailed uncertainty budget is presented for Cr in aqueous and dry samples. Systematic uncertainty budgets such as these in the design presented facilitate the uncertainty evaluation process and make it easier to compare the contributions of uncertainty components to the total uncertainty budget and offer a tool for improvement of the method performance. In addition, the use of commercial software can facilitate the calculations in order to make the entire process more user-friendly.

In dry plants, the concentrations of Pb and Cd were below the proposed limits by European Pharmacopoeia of 5 and 0.5 mg kg−1, respectively. The concentrations of essential and nonessential elements in tea infusion of the analysed samples were generally lower than the maximum values established by EU Drinking Water Directive 98/83/EC; thus, these tea can be considered safely for consumption. However, depending on the metal pollution in the sites where the medicinal plants are grown and the uptake of metals in these plants, the concentrations of metals in water extracts can determine the exceeding of the limits for drinking water due to the relatively high extractability of metals like Zn, Cu, and Ni.

Declarations

Acknowledgements

The authors gratefully acknowledge the financial support from the Ministry of Education, Science and Sport, from the Metrological Institute of the Republic of Slovenia (MIRS) and Romanian financing authority CNCS-UEFISCDI, Partnership, project VULMIN, Contract No. 52/2012.

Authors’ Affiliations

(1)
INCDO-INOE 2000, Research Institute for Analytical Instrumentation
(2)
Laboratory for Environmental Sciences and Engineering, National Institute of Chemistry

References

  1. Baralkiewicz D, Pikosz B, Belter M, Marcinkowska M: Speciation analysis of chromium in drinking water samples by ion-pair reversed-phase HPLC–ICP-MS: validation of the analytical method and evaluation of the uncertainty budget. Accred Qual Assur 2013, 18: 391–401.View ArticleGoogle Scholar
  2. Basgel S, Erdemoglu SB: Determination of mineral and trace elements in some medicinal herbs and their infusions consumed in Turkey. Sci Total Environ 2006, 359: 82–89.View ArticleGoogle Scholar
  3. Chuparina EV, Aisueva TS: Determination of heavy metal levels in medicinal plant Hemerocallis minor Miller by X-ray fluorescence spectrometry. Environ Chem Lett 2011, 9: 19–23.View ArticleGoogle Scholar
  4. Drolc A, Pintar A: Measurement uncertainty evaluation and in-house method validation of the herbicide iodosulfuron-methyl-sodium in water samples by using HPLC analysis. Accredit Qual Assur 2011, 16: 21–29.View ArticleGoogle Scholar
  5. Ellison SLR, Rosslein M, Williams A: EURACHEM/CITAC, quantifying uncertainty in analytical measurement. LGC, Teddington; 2012.Google Scholar
  6. The fitness for purpose of analytical methods. Eurachem LGC, Teddington; 1998.Google Scholar
  7. European Environment Agency (2007) . Accessed 28 May 2013, [http://www.eea.europa.eu/data-and-maps/indicators/#c5=&c7=all&c0=10&b_start=0&c6=progress+in+management+of+contaminated+sites]
  8. European pharmacopoeia. Council of Europe, Strasbourg, France; 2011.Google Scholar
  9. Froes RES, Neto WB, Beinner MA, Nascentes CC, da Silva JBB: Determination of inorganic elements in teas using inductively coupled plasma optical emission spectrometry and classification with exploratory analysis. Food Anal Methods 2014, 7: 540–546.View ArticleGoogle Scholar
  10. Gentscheva GD, Stafilov T, Ivanova EH: Determination of some essential and toxic elements in herbs from Bulgaria and Macedonia using atomic spectrometry. Eurasian J Anal Chem 2010, 5: 104–111.Google Scholar
  11. Water quality–calibration and evaluation of analytical methods and estimation of performance characteristics - part 1: statistical evaluation of the linear calibration function. ISO 8466–1. ISO, Geneva; 1990.Google Scholar
  12. Guide to the expression of uncertainty in measurement (GUM) JCGM 1995, 100: 2008.Google Scholar
  13. General requirements for the competence of testing and calibration laboratories; ISO/IEC 17025. European Committee for Standardization, Brussels; 2005.Google Scholar
  14. Kessel W: Measurement uncertainty according to ISO/BIPM-GUM. Thermochim Acta 2002, 382: 1–16.View ArticleGoogle Scholar
  15. Maharia RS, Dutta RK, Acharya R, Reddy AVR: Heavy metal bioaccumulation in selected medicinal plants collected from Khetri copper mines and comparison with those collected from fertile soil in Haridwar, India. J Environ Sci Heal B 2010, 45: 174–181.View ArticleGoogle Scholar
  16. Malandrino M, Abollino O, Buoso S, Giacomino A, La Gioia C, Mentasti E: Accumulation of heavy metals from contaminated soil to plants and evaluation of soil remediation by vermiculite. Chemosphere 2011, 82: 169–178.View ArticleGoogle Scholar
  17. Mermet JM: Is it still possible, necessary and beneficial to perform research in ICP-atomic emission spectrometry? J Anal Atom Spectrom 2005, 20: 11–16.View ArticleGoogle Scholar
  18. Miller JN, Miller JC: Statistics and chemometrics for analytical chemistry. Prentice Hall, Harlow, London, New York; 2000.Google Scholar
  19. Miranda K, Pereira-Filho ER: Sequential determination of Cd, Cu and Pb in tea leaves by slurry introduction to thermospray flame furnace atomic absorption spectrometry. Food Anal Methods 2013, 6: 1607–1610.View ArticleGoogle Scholar
  20. Mitic S, Obradovic MV, Mitic MN, Kostic DA, Pavlovic AN, Tosic SB, Stojkovic MD: Elemental composition of various sour cherry and table grape cultivars using inductively coupled plasma atomic emission spectrometry method (ICP-OES). Food Anal Methods 2012, 5: 279–286.View ArticleGoogle Scholar
  21. Moreno-Jimenez EM, Penalosa JM, Manzano R, Carpena-Ruiz RO, Gamarra R, Esteban E: Heavy metals distribution in soils surrounding an abandoned mine in NW Madrid (Spain) and their transference to wild flora. J Hazard Mater 2009, 162: 854–859.View ArticleGoogle Scholar
  22. Magnusson B, Näykk T, Hovind H, Krysell M: Guide handbook for calculation of measurement uncertainty in environmental laboratories. Nordtest, Oslo; 2012.Google Scholar
  23. Council directive 98/83/EC on the quality of water intended for human consumption Official Journal of the European Communities 1998, L 330: 32–54.Google Scholar
  24. Razic SS, Dogo SM, Slavkovic LJ: Multivariate characterization of herbal drugs and rhizosphere soil sample according to their metallic content. Microchem J 2006, 84: 93–101.View ArticleGoogle Scholar
  25. Rodrigues SM, Pereira ME, Duarte AC, Romkens PFAM: Soil–plant–animal transfer models to improve soil protection guidelines: a case study from Portugal. Environ Int 2012, 39: 27–37.View ArticleGoogle Scholar
  26. Senila M, Levei EA, Senila LR: Assessment of metals bioavailability to vegetables under field conditions using DGT, single extractions and multivariate statistics. Chem Cent J 2012, 6: 119.View ArticleGoogle Scholar

Copyright

© Senila et al.; licensee Springer. 2014

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.