INRiM contributions in the characterization and certification of reference materials

Measurement 44 (2011) 1381–1388

Contents lists available at ScienceDirect

Measurement journal homepage: www.elsevier.com/locate/measurement

INRiM contributions in the characterization and certification of reference materials Francesca Durbiano a,⇑, Elena Amico di Meane a, Luigi Bergamaschi b, Chiara Boveri a, Laura Giordani b, Michela Sega a a b

Istituto Nazionale di Ricerca Metrologica – I.N.Ri.M., Strada delle Cacce 91, I-10135 Torino, Italy Istituto Nazionale di Ricerca Metrologica – I.N.Ri.M., Unità di Radiochimica, Viale Taramelli 12, I-27100 Pavia, Italy

a r t i c l e

i n f o

Article history: Received 20 December 2010 Received in revised form 4 March 2011 Accepted 5 May 2011 Available online 18 May 2011 Keywords: Reference material Metrology in chemistry Measurement uncertainty Gas mixture Electrolytic conductivity Trace elements determination

a b s t r a c t The National Institute of Metrological Research (INRiM), in Italy, has among its duties the development, the maintenance and the dissemination of primary standards for establishing a correct traceability chain at national level. For this reason, INRiM has been developing research activities on reference materials, mainly in the chemical field, dealing both with the characterization and the certification of reference materials. The present paper focuses on the certified reference solutions for electrolytic conductivity, on the primary reference gas mixtures, and on INRiM contribution in the certification of various reference materials by determining trace element mass fraction by means of Instrumental Neutron Activation Analysis. The description of the main features of the produced reference materials, of the used procedures, and of the uncertainty evaluation are given. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction Analyses in the environmental as well as in food and nutritional areas imply the determination of various analytes, sometimes at very low concentration, in different and complex matrices. Accurate and reliable, hence traceable, measurement results are needed as a lot of important decisions regarding public or individual health, environmental protection and international trade are based on the results of analytical measurements. Due to the lack of primary methods applicable in routine chemical measurements, metrological traceability of measurement results can be achieved by a proper use of suitable Certified Reference Materials (CRMs) which had been characterized by means of highly reliable methods. The use of CRMs enables

⇑ Corresponding author. Tel.: +39 011 3919316; fax: +39 011 346384. E-mail addresses: [email protected] (F. Durbiano), l.bergamaschi @inrim.it (L. Bergamaschi), [email protected] (L. Giordani), m.sega@ inrim.it (M. Sega). 0263-2241/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.measurement.2011.05.003

the transfer of the values of measured or assigned properties between National Metrology Institute (NMIs) and testing and calibration laboratories. Traceability is a relatively new concept in the field of chemical measurement and the values carried by very few chemical reference materials are explicitly traceable to the units of the International System (SI). Such materials are widely used, e.g. for the calibration of measuring equipment and for the evaluation or validation of measurement procedures. The COMAR database contains information on more than 10000 reference materials (RMs) and CRMs [1]. The demand for new reference materials of higher quality is increasing as a consequence of various factors: the better performances of measuring equipment, the requirement for more accurate and reliable data in scientific and technological fields and in the measurements related to the quality of life, the stricter recommendations of current regulations. It is, therefore, not only necessary for RM producers to provide information about their materials in the form of reports, certificates and statements, but also to demonstrate their competence in producing RMs of appropriate quality. A NMI has among

1382

F. Durbiano et al. / Measurement 44 (2011) 1381–1388

its duties the development, the maintenance and the dissemination of primary standards for establishing a correct traceability chain at national level. At the same time, it must act as a connection between national and international metrology. For these reasons, the National Institute of Metrological Research (INRiM), being one of the two Italian NMIs, has been developing research activities on reference materials, mainly in the chemical field. The present paper focuses on the preparation of CRMs for electrolytic conductivity and gas analysis, and on INRiM contribution in the characterization and certification of various reference materials by determining trace metals by means of Instrumental Neutron Activation Analysis (INAA). The reliability of the certified values of reference materials can be established either when they are directly related to a reference through a documented unbroken chain of calibrations, or, when a higher metrological reference does not exist, through the definition of a consensus value arisen from different and independent measurement techniques. In the paper examples for these different operational modes are described. The uncertainty evaluation, according to [2] is also presented. The realization of the above CRMs is carried out following specific documents for reference material producers [3,4]. In addition, INRiM participation with success to related international comparisons underpins their value and mutual comparability with similar CRMs produced by other NMIs [5–11]. The certificated values, in terms of stability and uncertainties are comparable with the ones obtained by other NMIs and are satisfactory for the main user requirements. It is noteworthy that the proposed INRiM reference materials are produced and characterized by means of primary methods [12–14], which are at the top level in terms of metrological features.

2. Materials and methods 2.1. Electrolytic conductivity Reference materials accompanied by a certificate for electrolytic conductivity are at present prepared at INRiM. They are solutions composed by analytical grade pure water [15] and potassium chloride salt in different concentrations in equilibrium with atmospheric carbon dioxide. At international level the employment of KCl has been chosen for its specific properties; it is hardly hygroscopic, chemically inert, and commercially available with high purity at low cost. Moreover, in solution it dissolves completely and steadily in K+ and Cl ions. The electrolytic conductivity is a sum parameter which allows to provide quickly and at low cost information on ion concentration in the solutions. If very low conductivity values (less than 1 lS cm1) are considered, this parameter enables to check the purity of water. Therefore it is a very useful and exploited parameter in pharmaceutical, environmental and microelectronic sectors. However, reference materials with so low conductivity values are not yet in general commercially available because they suppose a measurement system and a store facilities in condition to prevent entirely the water and the material

contamination. INRiM, as other NMIs in other countries, are working on this project. The CRM permits to the operator to calibrate the conductivity meters at the company, with a suitable periodicity chosen by the operator himself. Typically, the CRM is provided in a 500 ml Pyrex bottles, with an hermetically closed cap with a seal, and equipped with label and analysis certificate. The electrolytic conductivity value at INRiM is determined using the primary measurement system based on the resistance measurement and on the cell geometry, and therefore traceable to the SI units [9]. The electrolytic conductivity measurement capability at INRiM was tested in turn by the participation in three International Measurement Comparisons [5–7]. As an alternative the conductivity value is obtained using the secondary measurement system calibrated by comparison with the primary so that the traceability to the units of the SI is maintained. The solution is prepared dissolving different amounts of KCl in analytical grade pure water depending on the aimed conductivity values, and afterward it is distributed in bottles. After the conductivity value has been determined, in order to define an associated uncertainty value, a homogeneity study and a stability study are carried out. Even if the solutions are expected to be homogeneous on physical (thermodynamic) grounds, assessments of the betweenbottle inhomogeneity were necessary to demonstrate that the batch of bottles was sufficiently homogeneous and to verify the correct washing process of the bottles. In fact the solution must keep the same conductivity value after it has been divided in the bottles. The homogeneity study [16] of solution among the bottles is carried out with a high resolution conductivity meter (WTW InoLab TetraCon 325) which allows to execute a fast and reliable measurement. Considering a replenishment order, the bottles selected for the homogeneity study are typically the first one, the one in the middle of the batch, and the last one. The measurement is carried out taking a solution sample from each bottle belonging to the same solution batch, and determining the conductivity value with the conductivity meter. The variability of the conductivity value obtained among the bottles is then accepted only if it is lower than the conductivity meter resolution. In this case the homogeneity contribution results negligible in the uncertainty budget. Since the solution can be polluted during the time due to the air contact (contamination by CO2), or because of a desorption from the bottle glass, long-term stability studies of different solutions are carried out in order to establish how acceptable is the value drift within a specific uncertainty. The long-term stability concerns the remaining instability of conductivity values of the solution under specified storage conditions. These conditions regard mainly the temperature and the storage state. Conductivity is strongly influenced by temperature, for example at 25 °C the temperature coefficient is approximately 2.0% per °C. A reference temperature of 25 °C has been chosen for providing reference materials because of the solution stability. Moreover, the solution should be kept in the original, unopened bottle, sealed, and to shelter from the light for minimising the possibility of pollution. For what concerns the

F. Durbiano et al. / Measurement 44 (2011) 1381–1388

short-term stability each bottle should be treated as a single-use standard. The instructions for use, given in the certificate, assert to use the solution immediately after the bottle is opened to avoid evaporation of the solution being measured. As in the homogeneity study also in this study the WTW conductivity meter is used. The measurements are carried out taking different samples from the same bottles at regular intervals. The variance obtained by the measurements with the conductivity meter is one of the uncertainty contributions assigned to the solution conductivity value. The evaluation of all the uncertainty contributions is undertaken considering the homogeneity and stability contributions as well as the ones connected to the primary or secondary measurement system and to the temperature. The analysis certificate associated to the reference solution is delivered according to the requirements of reference documents [17]. 2.2. Gas mixtures INRiM has developed a research activity on the preparation of primary reference gas mixtures by the gravimetric method [13,14], a primary method consisting in the introduction in two or more steps of determined amounts of gases (parent gases, which can be either pure gases or gas mixtures of known composition) in an evacuated cylinder. The mass of gas introduced in each step is measured accurately. The mass fraction and then the mole fraction of each component is calculated from weighing data. Small uncertainty, direct relation to SI units, realization of a virtually infinite number of mixtures and the easiness of cylinder transportation, allowing their use in field, are the advantages of this method. A facility for evacuating and filling cylinders and a device for high precision weighing were realized at INRiM. Mixtures are prepared in aluminium alloy cylinders having a volume of 5 l. Before the filling steps, they are previously evacuated and treated to eliminate impurities. The weighing of cylinders is a critical step, since the determination of a small mass of gas introduced in a cylinder, which has large mass and volume, requires attention to achieve the requested accuracy. For this purpose, the device for high accuracy weighing was planned to automatize the exchange of two cylinders, the one in which the gas mixture is prepared and a reference one. The mass comparison of the two cylinders reduces the correction due to the buoyancy effect. Calibrated mass standards are used to keep the mass difference between the two cylinders within 1 g, to optimize the mass comparator performance [14,18]. After the preparation, mixtures are homogenized by mechanical rolling and then analytically checked in order to verify the whole preparation process. Mixtures stability is investigated by replicating the analytical measurements at regular intervals. The homogeneity of this kind of mixtures is not a critical parameter. In principle, a stratification of the components due to their different physical properties could happen, but the mixtures, once prepared, are contained in cylinders at high pressure, hence it is not possible to sample them at different cylinder heights to verify if a stratification occurs. For this reason, to

1383

check the homogeneity, NDIR measurements were carried out on cylinders before and after their mechanical rolling, which is a common practice for homogenizing a mixture contained in a cylinder. The results did not show a significant difference. In addition, no correlation was found between the analytical response and the residual gas pressure inside the cylinders. The mixtures preparation process was validated by the participation in International Measurement Comparisons, an example of which is reported in [13]. 2.3. Trace element analysis INRiM activities in the inorganic chemistry field are mainly focused on trace elements analysis. Samples of environmental, biological, forensic and industrial origin are usually studied [19,20]. For all these matrices the development of appropriate and defined analytical methods is normally carried out and tested with meticulous consideration on the operation steps which imply possibility of loss or contamination of the samples. The field of analyzable matrices is really wide and it deals with, for example, environment (soils, air particulate matter, botanical materials), health (biological fluids and tissues, cells, food), materials (polymers, plastics, ceramics, metals, ores, alloys). INAA and Graphite Furnace-Atomic Absorption Spectroscopy (GF-AAS) are employed as analytical techniques. However, in trace elements determination, INAA has been used as the main one. It is based upon the conversion of stable atomic nuclei in radioactive ones by irradiation with neutrons. The induced radioactivity causes gamma radiation emission. The subsequent spectrometry of gamma radiation is carried out using High Purity Germanium (HPGe) detectors coupled with computerized multichannel analysers. After this, by acquiring and elaborating gamma emission data from characteristic isotopic neutron reaction, a qualitative and quantitative analysis is performed. Sample irradiations are carried out in the Triga Mark II Nuclear Reactor of the University of Pavia. Irradiation time is selected on the basis of the decay life time of the analysed elements. It may vary from 30 to 500 s for short life elements and from 4 to 40 h for medium and long life elements; it depends also on the samples matrix and the neutron flux. Table 1, shows elements that are usually analysed at INRiM, the corresponding radionuclides produced by neutron irradiation, the mass ranges which can be determined, the gamma energy and half-life time of each radionuclide. Samples to be characterized or certificated for the elemental analysis undergo to different analytical steps. For the determination of the moisture content, some samples are weighed before and after drying: this allows to get the concentration values referred to the dry samples weight. Usually the drying protocol is suggested by the RM producer. Thus, different aliquots of samples are weighed and the minimum mass intake is determined by the homogeneity test carried out by the producer. The mass traceability to SI is achieved using a calibrated analytical balance. Samples are subsequently sealed in polyethylene or quartz vials (if the irradiation exceeds 6 h) and

1384

F. Durbiano et al. / Measurement 44 (2011) 1381–1388

Table 1 List of elements usually analysed at INRiM. For each element the corresponding radionuclide produced by neutron irradiation, the mass range which can be determined, the gamma energy and half-life time of the radionuclide are reported. Element Short live elements Magnesium Vanadium Manganese Copper Selenium Medium live elements Zinc Arsenic Bromine Molybdenum Cadmium Gold Long live elements Scandium Chromium Nickel Iron Cobalt Zinc Selenium Rubidium Antimony Mercury

Radionuclide

Mass interval (mg/kg)

Energy (keV)

t1/2

27

Mg V Mn 66 Cu 77m Se

10–1000 0.1–500 0.1–500 10–500 0.1–500

1014.1 1434.4 846.8; 1810.7 10.39 161.9

9.5 min 3.75 min 2.58 h 5.10 min 17.4 s

69m

Zn As 82 Br 99 Mo 115 Cd 198 Au

10–1000 0.5–500 0.5–500 5–300 10–500 0.01–10

438.7 559.2; 657 554.5; 776.7 140.6; 181 336; 527.7 411.8

13.8 h 26.4 h 35.34 h 66.7 h 2.3 d 2.67 h

46

0.1–500 0.1–500 50–1000 10–1000 0.1–300 10–1000 0.1–500 1–500 0.1–500 0.1–300

889.4; 1120.3 320 810.3 1098.6 1173.1; 1332.4 1115.4 264.7 1076.6 1690.7 279.1

83.9 d 27.8 d 71.3 d 45.1 d 5.24 y 245 d 119.8 d 18.66 d 60.9 d 46.9 d

52 56

76

Sc Cr Ni(58Co) 59 Fe 60 Co 65 Zn 75 Se 86 Rb 124 Sb 203 Hg 51

irradiated in the nuclear reactor. Empty vials are also irradiated for blank measurements. In order to obtain a quantitative determination, an external calibration is carried out by means of certified standard solutions gravimetrically diluted and deposited onto filter paper rolled up as a cylinder, inserted in the vials and evaporated to dryness in a fume hood. In this way calibration solutions and samples (typically about 200–300 mg) have the most closest geometry. Samples geometry is actually one of the most relevant parameter in the uncertainty budget. Furthermore, as an additional reliability control of the whole analytical process, different CRMs undergo to the same analytical procedure. They are selected on the basis of matrix similarity and certified concentrations close to those expected in samples. For instance, in the international comparison CCQM-P118 ‘‘Toxic metals in algae’’, INRiM used as CRMs for validation and for recovery evaluation the NIST SRM 1566b ‘‘Oyster tissue’’ having a different matrix, but low spectral interferences, the NIST SRM 1570a ‘‘Spinach leaves’’ and the NIST SRM 1573a ‘‘Tomato leaves’’ having matrix and certified values very close to the samples. Samples, standards and blanks are inserted in the same irradiation container, a neutron flux monitor (Al/Co wire 0.1% Co) is also inserted to take into account a possible variation of the neutron flux inside the container and to calculate the corresponding uncertainty contribution. The container is transferred in the reactor for the irradiation; gamma spectrometry is then carried out on the irradiated samples by HPGe detectors. The distance between samples and detector is adjusted depending on the activity of the samples, the more is the distance the less is the associated uncertainty but the efficiency of the detector decreases.

3. Results 3.1. Electrolytic conductivity Currently INRiM develops CRMs for electrolytic conductivity in the range between 50 lS cm1 and 1.3 S m1 with an associated relative uncertainty lower than 0.7%, at a temperature of 25 °C. The model equation for evaluating the conductivity value at the reference temperature has been expressed as:

kSol ¼ C 

1 þ dkHomogBB þ dkSt R  ½1  aT  ðT  T Ref Þ

ð1Þ

where the first term, kSol, represents the electrolytic conductivity value obtained by the batch characterization with the reference cell, C is the geometric constant, characteristic of the cell, R is the measured resistance, aT is the temperature coefficient, T is the temperature, TRef is the reference temperature, dkHomogB–B is the contribution associated to the homogeneity of the solution between the bottles, and, kSt, is the contribution due to the stability of the solution verified with measurements carried out during its validity period. Fig. 1 shows the uncertainty contributions for a solution at 25 °C with the electrolytic conductivity of 214.9 lS cm1. The combined standard uncertainty has been evaluated as 6.7  105 S m1 and the expanded uncertainty as 1.3  104 S m1. The expanded uncertainty, here reported, has been obtained by multiplying the combined standard uncertainty by a coverage factor k = 2, corresponding to an interval having a level of confidence of approximately 95%. Numerous bottles of the same solution have been used in the stability studies. The variance of the whole set of measurements during

F. Durbiano et al. / Measurement 44 (2011) 1381–1388

1385

Fig. 1. Example of uncertainty budget for a solution having electrolytic conductivity of 214.9 lS cm1 at 25 °C. The contributions due to solution stability, homogeneity among bottles, temperature, temperature coefficient, repeatability of resistance measurements, and cell constant are reported.

for example 6 months has been evaluated considering the maximum range within which the conductivity values are included and assigning a rectangular distribution. Depending on the conductivity range, the stability study and the associated uncertainty contribution are useful to evaluate the solution expiry date. For high conductivity solutions (more than 1400 lS cm1) the uncertainty value is assigned considering a solution stability of 1 year; while for low conductivity solutions the stability is guaranteed for 6 months. Currently, the uncertainty contribution due to the homogeneity among bottles is negligible; instead, the contribution due to stability is among the dominant ones. For each provided CRM, a bottle is kept in the laboratory in order to enable a periodical check of the conductivity value in case of customer complaint.

lytes and on purity of parent gases, uver is the contribution due to the analytical verification of the mixture and ustab comes from the mixture stability. Each term takes into account the possible covariances between the different input quantities. The term ustab is evaluated as the standard deviation of the differences among the analytical values, obtained during the verifications by Non Dispersive Infrared spectroscopy (NDIR) for the stability assessment, and the preparation weighing data in 2-year time. In Fig. 2 an example of uncertainty budget for a mixture of CO2 at the mole fraction of 11.99  102 mol mol1 in matrix of N2 is reported. As it can be seen, the contribution of the gravimetric preparation is negligible if compared with the others. The large contribution of the two other sources is mainly due to the NDIR photometer used for the analyses.

3.2. Gas mixtures 3.3. Trace element analysis At present, primary mixtures of carbon dioxide (CO2) in matrices of nitrogen (N2) and of synthetic air at automotive emission level (10–14% mol mol1) and at ambient level (200–400 lmol mol1) are prepared at INRiM, having standard uncertainties deriving from preparation of some lmol mol–1. The choice of carbon dioxide as analyte derives both from its environmental relevance as a greenhouse gas and from its role in some metrological measurements (buoyancy effect in calibration of mass standards, air refraction index in length measurements, solubility in water in electrolytic conductivity). Uncertainty on mixture concentration, uC, is calculated (Eq. (2)) according to the uncertainty propagation law [2]:

uC ¼

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u2grav þ u2ver þ u2stab

ð2Þ

where ugrav derives from the gravimetric preparation and comprises also the uncertainty on molar masses of the ana-

Environmental, biological, forensic, and industrial samples are usually complex matrices. Therefore, accurate measurement of element concentrations is really difficult to obtain and the conversion process of these matrices to CRMs results often critical. In these cases, a possible way to obtain reliable values, is to use and compare two or more analytical techniques based on different physical principles. This procedure is used at international level when a direct traceability to SI cannot be established. Following this approach INAA is often required for characterization and certification campaign of RMs. For some elements, often in the group of the rare earths, INAA is the only helpful technique at low concentration level. INAA presents some important properties as high sensitivity, opportunity of multielemental analyses, high accuracy and precision. If compared with other techniques commonly used for the determination of trace elements in

1386

F. Durbiano et al. / Measurement 44 (2011) 1381–1388

Fig. 2. Example of uncertainty budget for a mixture of CO2 at the mole fraction of 11.99  102 mol/mol in matrix of N2. The contributions due to gravimetric preparation, analytical verification of the mixture and its stability are reported.

various matrices it presents many advantages: it can be used for the analysis of micro-samples (few milligrams), it is a non-destructive analytical technique, chemical state and physical form of analytes do not affect the analysis, and elements such as H, C, N, O, do not affect the determination of others. In addition, there are many adjustable parameters for optimizing the experimental design. Furthermore, it requires less complex operations (dissolution, digestion or other chemical treatments of the samples). This reduces the uncertainty sources and consequently a complete and well defined uncertainty budget for INAA is possible. For its features, CCQM has declared that INAA has a ‘‘similar status’’ as well as the other primary ratio methods for elemental analysis [21]. The calibration and measurement capabilities [22], verified through the participation to international comparisons, allow INRiM to take part in the characterization of CRMs related to human health, nutrition, environment as well as to technological field (high purity materials). Particularly relevant is INRiM participation in the certification of NIST SRM 2783, a simulated urban aerosol

um ¼

‘‘Sediment’’ were analysed in order to define the elemental mass fraction. The above mentioned RMs and similar ones, for example produced by ISPRA, are used in the regional field laboratory as references for their analysis, implementing the dissemination. For the evaluation of the combined standard uncertainty in INAA analyses, typical chemical analysis contributions are considered and about 30 different parameters can be added specifically for the INAA, but only the most significant are really taken into account. They concern sample mass (sm), standards mass (stm), counting statistics for samples (cs), counting statistics for standards (cst), blanks (bk), neutron flux and irradiation geometry (nf), counting geometry (cg), pulse pileup and life-time correction (lt), peak integration (pi), recovery (rec) and repeatability (rep). An example of uncertainty budget for the determination of the mass fraction of chromium (Cr) in polypropylene, at a concentration of about 250 mg/kg, is presented in Fig. 3. The combined standard uncertainty of the mass of the element in the sample is calculated from Eq. (3), according to the uncertainty propagation law [2]:

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u2sm þ u2stm þ u2cs þ u2cst þ u2bk þ u2nf þ u2cg þ u2lt þ u2pi þ u2rec þ u2rep

(PM. 2.5) deposited onto filter media, prepared by the US metrological institute and characterized for elemental content by various expert laboratories with different techniques. INRiM took part in the characterization campaigns of candidates RMs produced and distributed by the Italian Institute for Environmental Protection and Research (ISPRA). RM014 ‘‘Polluted soil-A’’, and RM 021

ð3Þ

The experience acquired in developing reliable analytical methods allows to transfer the metrological approach in the applied research fields. Here the activity has been focused on air quality (pollution in urban, industrial, remote areas; bio monitoring studies) as well as on human health (i.e. toxicological role of trace elements in neurodegenerative disease like Parkinson) and on industrial materials.

F. Durbiano et al. / Measurement 44 (2011) 1381–1388

1387

Fig. 3. Example of uncertainty budget for the determination of Cr in polypropylene at a concentration of about 250 mg/kg. The contributions due to sample mass, standards mass, counting statistics for samples, counting statistics for standards, blanks, neutron flux and irradiation geometry, counting geometry, pulse pileup and life-time correction, peak integration, recovery and repeatability are reported.

4. Conclusions

References

INRiM activities related to CRMs, even if relatively recent, are encouraging. In order to deal with the needs of establishing a correct traceability chain to allow the metrological activities of the national bodies, accredited and field laboratories involved in various sectors related to the environment, food and nutrition, other developments are in progress. The declaration of new CMCs on the BIPM database related to CRMs production is foreseen. Recently INRiM attended an international comparison about seawater salinity measurement [23] to expand INRiM measurement capabilities to high conductivity values (more than 5 S m1) and to offer new CRMs of interest for the marine community. Moreover the development of a flow system with a closed-circuit which allows to avoid pure water contamination is in progress. This new measurement system will permit to expand the measurable conductivity range to values lower than 1 lS cm1, and to produce the correspondent CRMs in order to satisfy the growing needs in the pharmaceutical sector. As for CRMs for gas analysis, the preparation of gravimetric mixtures of nitrogen oxides in N2 and in synthetic air has just started and it will be extended to other analytes, such as carbon monoxide, methane, propane and their combination to meet traceability needs at national level. Purity of parent gases is also under investigation. Some improvements are still needed to reduce the contributions due to analytical verification and stability. Furthermore, the analytical methods developed for the reference material certification by INAA will be applied to present and future studies, in particular related to element determination in biological fluid, in air particulate matter, in brain tissues to study neurodegenerative diseases, and to other studies about health and environment.

[1] www.comar.bam.de. [2] JCGM 100:2008, GUM 1995 with Minor Corrections. Evaluation of Measurement Data – Guide to the Expression of Uncertainty in Measurement, first ed., 2008. [3] ISO Guide 34, General Requirements for the Competence of Reference Material Producers, ISO, Geneva, 2009. [4] ILAC G 12, Guidelines for the Requirements for the Competence of Reference Material Producers, 2000. [5] F. Durbiano, L. Callegaro, P.P. Capra, V. D’Elia, Key Comparison CCQM-K36: Electrolytic Conductivity at 0.5 and 0.005 S/m, IEN Measurement Report, IEN Technical Report No. 694, 2005. [6] F. Durbiano, L. Callegaro, P.P. Capra, G. Marullo Reedtz, Pilot Comparison CCQM-P47 on Electrolytic Conductivity of Dilute Solutions, IEN Measurement Report, IEN Technical Report No. 666, 2003. [7] F. Durbiano, E. Ferrara, G. Marullo Reedtz, P.P. Capra, Pilot Study CCQM – P22 on Electrolytic Conductivity, IEN Measurement Report, IEN Technical Report No. 629, 2001. [8] A.M.H. Van der Veen, J.I.T. Van Wijk, R.P. Van Otterloo, R.M. Wessel, E.W.B. de Leer, J. Novak, M. Sega, A. Rakowska, I. Castanheira, A. Botha, S. Musil, EUROMET.QM-K3: automotive emission gas measurements, Metrologia, Tech. Suppl. 39 (2002) 08005-1. [9] A.M.H. Van der Veen, F.N.C. Brinkmann, M. Arnautovic, L. Besley, H.J. Heine, T.L. Esteban, M. Sega, K. Kato, J.S. Kim, A.P. Castorena, A. Rakowska, M.J.T. Milton, F.R. Guenther, R. Francey, E. Dlugokencky, International Comparison CCQM-P41 Greenhouse gases. 1. Measurement capability, Metrologia 44 (1A) (2007) 08002. [10] A.M.H. Van der Veen, F.N.C. Brinkmann, M. Arnautovic, L. Besley, H.J. Heine, T.L. Esteban, M. Sega, K. Kato, J.S. Kim, A.P. Castorena, A. Rakowska, M.J.T. Milton, F.R. Guenther, R. Francey, E. Dlugokencky, International Comparison CCQM-P41 Greenhouse gases. 2. Direct comparison of primary standard gas mixtures, Metrologia 44 (1A) (2007) 08003. [11] R.M. Wessel, A.M.H. van der Veen, P.R. Ziel, P. Steele, R. Langenfelds, M. van der Schoot, D. Smeulders, L. Besley, V.S. da Cunha, Z. Zhou, H. Qiao, H.J. Heine, B. Martin, T. Macé, P.K. Gupta, E. Amico di Meane, M. Sega, F. Rolle, M. Maruyama, K. Kato, N. Matsumoto, J.S. Kim, D.M. Moon, J.B. Lee, F.R. Murillo, C.R. Nambo, M.d.J. Avila Salas, A.P. Castorena, L.A. Konopelko, Y.A. Kustikov, A.V. Kolobova, V.V. Pankratov, O.V. Efremova, S. Musil, F. Chromek, M. Valkova, M.J.T. Milton, G. Vargha, F.R. Guenther, W.R. Miller, A. Botha, J. Tshilongo, I.S. Mokgoro, N. Leshabane, International Comparison CCQM-K52 carbon dioxide in synthetic air, Metrologia 45 (1A) (2008) 08011. [12] F. Brinkmann, N.E. Dam, E. Deák, F. Durbiano, E. Ferrara, J. Fükö, H.D. Jensen, M. Máriássy, R. Shreiner, P. Spitzer, U. Sudmeier, M. Surdu, L. Vyskocil, Accredit. Qual. Assur. 8 (2003) 346–353.

1388

F. Durbiano et al. / Measurement 44 (2011) 1381–1388

[13] E. Amico di Meane, M. Sega, F. Rolle, International Comparison CCQM-K52 Carbon Dioxide in Synthetic Air – I.N.Ri.M. Results, I.N.Ri.M. Technical Report No. 32, 2006. [14] ISO International Standard 6142 – Gas Analysis – Preparation of Calibration Gas Mixtures – Gravimetric Methods, ISO, Geneva, 2001. [15] Standard Test Method for Electrical Conductivity and Resistivity of a Flowing High Purity Water Sample, ASTM-International, D 5391-99, 2009. [16] ISO Guide 35-reference Materials: General and Statistical Principles for Certification, ISO, Geneva, 2006. [17] ISO Guide 31-reference Materials: Contents of Certificates and Labels, ISO, Geneva, 2000. [18] A. Alink, A.M.H. Van der Veen, Metrologia 37 (2000) 641–650.

[19] E. Rizzio, G. Giaveri, L. Bergamaschi, A. Profumo, G. Tartari, M. Gallorini, Developments in Earth Surface Processes, vol. 10, Elsevier, Amsterdam Boston Heidelberg, 2007, pp. 171–183. [20] L. Bergamaschi, E. Rizzio, A. Profumo, M. Gallorini, Determination of trace elements by INAA in urban air particulate matter and transplanted liches, J. Radioanal. Nucl. Chem. 263 (2005) 745–750. [21] Report on the Activity and Management of the International Bureau of Weights and Measures of the 13th and 14th Meeting of the CCQM (19–20 April 2007 and 3–4 April 2008). [22] www.bipm.org. [23] C. Boveri, L. Callegaro, F. Durbiano, EUROMET Electrochemical Analysis WG – Project: 918. Study on the Traceability of Salinity Measurements in Seawater. INRIM Measurement Report, I.N.Ri.M. Technical Report No. 46, 2007.