Fundamental studies to develop certified reference material to calibrate spectrophotometer in the ultraviolet region

Spectrophotometry is the technique used in a great number of laboratories around the world. Quantitative determination of a high number of inorganic, organic and biological species can be made by spectrophotometry using calibrated spectrophotometers. International standards require the use of optical filters to perform the calibration of spectrophotometers. One of the recommended materials is the crystalline potassium dichromate (K2Cr2O7), which is used to prepare solutions in specific concentrations for calibration or verification of spectrophotometers in the ultraviolet (UV) spectral regions. This paper presents the results concerning the fundamental studies for developing a certified reference material (CRM) of crystalline potassium dichromate to be used as standard of spectrophotometers in order to contribute to reliable quantitative analyses.


Introduction
The absorption molecular spectrometry (also called spectrophotometry) in the ultraviolet (UV) spectral regions is very used for quantitative determining of a high number of inorganic, organic and biological species in chemical laboratories, petrochemical, environmental, forensic and clinic around the world. It is based on the transmittance or absorbance measurements carried out in spectrophotometer, which is an instrument that relates the concentration of determined substance with the quantity of the absorbed or transmitted radiation in specific wavelength, that traversed transparent container with length of path defined [1,2] according to the Lambert-Beer´s law as equation (1).

=
(1) Where, is the apparent absorbance; , apparent absorptivity; b, length of path and c, the concentration. The term named apparent was introduced together with the parameters absorbance and absorptivity because the effect related to the buoyancy correction for mass and also internal reflections of light in the quartz cuvette were not considered in the measurements.
For guaranteeing of reliability and metrological traceability from measurement results, the ISO/IEC 17025 standard states that all equipment, which has an important effect in the accuracy and reliability of the measurement results of testing, calibration or sampling, shall be calibrated before being used [3], on the other hand, one of the statements concerning the spectrophotometers in the regional standard ASTM E925 [4] says that it is of own analyst´s responsibility verify and validate either if the instrument is running correctly or if it is capable to provide acceptable analytical results. The verification of the performance of the instrument shall be done with the use of certified reference materials (CRMs). These CRMs must have optical properties very well characterized to be used for that purpose. Currently, the National Institute of Standards & Technology (NIST) provides the SRM 935a [5]. This standard reference material (SRM) is the crystalline potassium dichromate, from which solutions with specific concentrations to the wavelengths in the spectral regions of UV are prepared. These standard solutions are used to evaluate the performance of the spectrophotometers in the UV regions.
The National Institute of Metrology, Quality and Technology (Inmetro) was granted the calibration and measurement capability (CMC) for high purity potassium dichromate. Due to the importance of verifying the performance of the spectrophotometers to regular time interval, the Electrochemistry Laboratory from the Chemistry Metrology Division (Dquim) together with the Radiometry and Photometry Laboratory from the Optical Division (Diopt) are developing the CRM of crystalline potassium dichromate.
The main goal of this work is to present the fundamental studies needed to develop the CRM of crystalline potassium dichromate in Brazil, since this CRM will contribute to guarantee the accuracy and linearity of the absorbance scale to the spectrophotometers, through which analytical measurement results will be provided with metrological traceability and reliability.

Feasibility studies
Initially, solutions of potassium dichromate (K2Cr2O7) in 0.001 mol L 1 perchloric acid (HClO4) medium were prepared. The concentrations of the K2Cr2O7 were 20, 40, 60, 80 and 100 mg kg 1 . These concentrations will provide absorption values in the specific wavelength range needed to verify the performance of the spectrophotometers in the UV regions. With the measurement results of absorbance for each solution, there was calculated the apparent absorptivity as described in equation (1) and the results were compared to the values shown in the SRM 935a certificate.

Stability of the solutions
It was prepared, in addition, 1 L of two new solutions of K2Cr2O7 in 0.001mol L 1 HClO4 either in the lowest level of concentrations, 20 mg kg 1 , or in the highest level, 100 mg kg 1 , for the stability study. This study is needed to verify the bias in the parameter of apparent absorptivity during the time. This study is very important for the consumer to know if the dichromate solutions will remain their optical properties constant along the time.

Batch preparation
A batch contained 64 bottles of the candidate of CRM was prepared. Firstly, 1 kg of the salt of K2Cr2O7 was mixed and manually homogenized; after, 15 g of the material was weighed in an analytical balance and was filled in each glass bottle, which was numbered in the order of filling.

Reagents
High purity K2Cr2O7 (99.5%) from Sigma-Aldrich and HClO4analytical purity (70-72%) from Merck were used. All solutions were prepared using deionized water from Elga ® model Option-Q with electrolytic conductivity lower or equal to 1.0 µS cm 1 . The apparent absorbances were measured by UV/VIS Spectrophotometer Lambda 25, using quartz cell both from Perkin Elmer ® with length of path of 1.0 cm. The K2Cr2O7 solutions were prepared weighing the material using analytical balance both from Mettler Toledo ® with resolution 0.001 g or a microbalance with resolution of 0.1µg, when necessary.

Methodology
The utilized methodology was as follows: Initially, K2Cr2O7 solutions were prepared in 0.001 mol L 1 HClO4 using the CRM of potassium dichromate from Inmetro (MRC 8477) in the following concentrations: 20 mg kg 1 , 40 mg kg 1 , 60 mg kg 1 , 80 mg kg 1 , and 100 mg kg 1 . After then, measurements of apparent absorbance in triplicate were done in the specific wavelengths () in the UV regions: 235 nm (1 nm = 10 9 m), 257 nm, 313 nm, 345 nm and 350 nm. Finally, from the apparent absorbance values, the apparent absorptivities were calculated by using equation (1) for each wavelength measured. Figure 1 shows the absorption spectrum for specific concentrations of the solutions (range 20-100 mol kg 1 ). It can be seen that the apparent absorbance increases in each wavelength measured as soon as the concentration is raised. The lowest values measured were obtained for  = 313 nm, as showed in the literature [4,5].  Table1.Apparent absorptivity and the relative error considering this work and the SRM 935a.

Results and discussion
1 Apparent absorptivity of this work divided by the apparent absorptivity from NIST.
Although the spectral bandwidth (SBW) used in this work was constant and equal to 1 nm, the results were comparable do the results showed in the NIST's certificate. In the certificate SRM 935a the SBWs varied in the range of 0.8 to 1.2 nm. On the other hand, the ASTM E-925 standard recommends that the SBW should not be higher than 2 nm, therefore the comparability of the measurement results can be acceptable.
From table 1, it can be seen that the higher relative error was around 5%, due to the lower concentration used, since this concentration has lower absorptions and, thus, can present deviations from Lambert-Beer law. Therefore, these results show that it is viable to develop this new CRM of crystalline potassium dichromate in Brazil.
For 5 months, the stability of the K2Cr2O7 solutions in 0.001 mol L 1 HClO4 medium in the concentrations of 20 mg kg 1 and 100 mg kg 1 is showed in figure 2. The wavelength of 235 nm was considered as an example. All the data of apparent absorptivity was considered normally distributed by Shapiro-Wilk test. By regression analysis, the values of apparent absorptivity measured in the solutions were considered stable (p-value > 0.05) for a confidence level of 95% [6]. Because of that, the K2Cr2O7 solutions can be considered stable for 5 months.
Considering the results of feasibility studies, a batch of a candidate CRM of crystalline potassium dichromate was prepared. Four bottles were at random selected from the batch for the studies of homogeneity, characterization and stability of the solutions. Both homogeneity and characterization studies are being performed. The uncertainty related to the both studies will be added to the uncertainty of the CRM. The absorbance of the solutions in five different concentrations of K2Cr2O7 was chosen as the property value. The inhomogeneity of the batch will be inferred by using the statistical test of One-way ANOVA. The characterization of the candidate CRM will be carried out by a comparison test performed by two laboratories.
The process of certification of the batch of crystalline K2Cr2O7 will be subject to a next work.

Conclusions
Spectrophotometry is an analytical technique very well used in diverse laboratories in the country. With the development of the CRM of crystalline potassium dichromate, the laboratories can guarantee better accuracy in the spectrophotometric analyses in diverse areas, for example, the petrochemical laboratories for analyzing some quality parameters of biofuels needed to be in accordance with the specifications of National Agency of Oil, Natural Gas and Biofuels (ANP).

Aknowledgements
One of the authors (F C da Conceição) wants to thank to Petrobras for granting a fellowship (PRH-PB nº 103).