Storage modeling standard Cr(VI) solution based on comparison of time regression value with UV-Vis spectrophotometry

: Chromium (Cr) compounds are found in nature in two forms, Cr (III) and Cr (VI). Analysis of Cr (VI) was determined using spectrophotometer UV-vis. A standard solution must be fresh as a standard that is usually not used repeatedly in the measurement. This study aims to determine whether the standard solution of Cr (VI) be used for a certain period until it must be disposed of. This idea is based on Cr (VI) compounds that were categorized as hazardous and toxic materials. In this study, the shelf life of standard Cr (VI) was carried out with controlled temperature and humidity in a refrigerator at 4 ℃ during the test period of two days. In this study, there was found that every hour of storage of standard Cr (VI) will observe a change in the gradient value of the regression. The model then compared with p-value resulted in only up to 7 hours of storage with a deviation of regression slope and intercept compared with initial time (T 0 –T 7 ) is 0.0062 and 0.0019 which and its increases over time. From the result, it can be concluded that standard regression can be used as a model for determining the maximum storage time by comparing the regression values each time so that the seven hours storage time is the limit where the Cr (VI) that can no longer be used as a measurement standard. It concerned that standard regression could be used as a model for determining the maximum storage time by comparing the regression values each time.


Artikel
Storage modeling standard Cr(VI) solution based on comparison of time regression value with UV-Vis spectrophotometry Elfiwani and Fachrurrazie*

■ INTRODUCTION
Chromium (Cr) is one of the mineral metals whose existence is contained in the layers of the earth. It is found at minuscule concentrations in ores, animals, plants, volcanic ashes, and fumes. Hexavalent Cr is a globally distinguished environmental pollutant with potential as carcinogenic, teratogenic, and mutagenic consequences [1].
Epidemiological studies show that inhalation of Cr (VI) in particulate matter (PM) causes lung and nasal cancer. Induced nasal carcinoma and DNA damage have also been reported in workers exposed to Cr (VI) after roughly eight years of exposure in electroplating processes. On the other hand, oral exposure studies report serious reproductive effects, such as alteration of semen which decreases sperm count [2].
As an outgrowth of anthropogenic activities, there occurs a significant hastening in the release of Cr to the environment. An excessive amount of wastewater contains Cr (VI) formed by the various activities [3]. Depending on its valence, Cr can be unless poisonous or advantageous. Industrial effluents carry significant amounts of Cr. Cr (VI) is one of the prominent polluting elements in the environment, the forms of Cr (VI) chromate and dichromate are extremely soluble under all pH conditions, but they could precipitate with divalent cations [4].
Experimental work has shown that aqueous Cr (VI) production rates are primarily controlled by the solubility and dissolution rates of Cr (III) from host minerals as well as the solution pH. When manganese-iron oxides co-exist with Cr (III)-bearing minerals, Cr (III) oxidation by Mn(V)-oxides can occur on the mineral surface, yielding Mn (II) and Cr (VI), which may then be mobilized into the aquatic phase [5].
Due to the extensively varying toxicity of Cr, total Cr measurement offends to be used to circumscribe the actual environmental impact. Thus, chemical speciation of Cr in environmental samples expect precisely assess pollution levels [6]. Accordingly, the determination of Cr (VI) at trace levels accurately has received considerable attention. In this study, we modelled the stability of Cr standard solution to define when the precision of the standard was diminishing over time.  Based on the visual colour, we found that no significant difference in the visual appearance of the Cr (VI) day 0 standard with the 32 hours storage at 4oC standards as shown in Figure  1. Nonetheless, this interpretation cannot conclude as a reference. From this finding, we needed a statistical method for the test to find when there was a distinct value between initial time measurement (T0) with storage one. The biasness of the models was need to be tested using t-test, Root mean square error (RMSE) is also a good measure of how accurately the model predicts the response. RMSE is also most important criterion for fitness of model, if the main purpose of the model is predicting [7].
The investigation between the two regression models was conducted according to the last paper [8]. The two regression was compared in a two-step, first, we find where the p-value of the two-regression model with a parallel hypothesis or not, which mean the slope of the two-regression model was not different enough with a significant error of 5%. second, finding the p-value of two regression models that match each other, which is after passing the first hypothesis the two regression intercepts are also not different enough with a significant error of 5%. Figure 2 shows the standard Cr regression graph (VI) from 0 to 32 hours storage, there was a slight difference in the slope of each resulting regression but from the correlation coefficient (R 2 ), there was no distinction. The R 2 value of the measurement still produces a value of >0.9998 this condition means all of the regression still produce valid data. the intercept also shows a slight fluctuation there is no significant fluctuation in the regression equation and the deviation of regression slope and intercept compared with initial time (T0-T7) is 0.0062 and 0.0019 respectively. This could be occurred by the instability complex between the Cr (VI) with 1,5diphenylcarbazide. Interferences due to cationic and anionic species were could also be major problem for spectrophotometry technique, significant interferences were reported for iron (III), nevertheless, Cr (III), iron (III), copper (II) and mercury (II) did not interfere significantly up to their concentration of 10 mgl -1 the interference was considered significant if the observed change of Cr signal was higher that 5% [9]. Based on the regression results, we also compared the measurement results of the regression slope value vs time measurement (Figure 3). This could be seen from seven hours storage of standard solution there has been a decrease in the slope value of the standard regression curve, from 0.8044 to 0.7627 in only 32 hours. from that result, it would greatly affect measurement sensitivity thus accuracy would be affected too.
The pH of the acid complex formed so for the Cr (VI) was pH 1-1,5 and if it was not optimally coordinated, although the resulting absorbance showed a large absorbance, The optimum wavelength for the metal ion complex Cr (III) at pH 8 [10]. During analyses, the pH of samples analysis was adjusted to 9-9.5. At this pH Cr exists as CrO4 2− species. The adjustment of pH of samples to 2 ensured that Cr exists as HCrO 4− and Cr2O7 2− [11].  Performing a regression comparison between fresh to 4, 7, 25, 29, and 32 hours storage, we modelled so we could know when the standard solution was different from fresh standards solution regression result. In Figure 4, we got the p-value of the regression parallel test the critical point of the probability of rejecting H0 has occurred when we store the standard over seven hours. when H0 is accepted it means the regression curve was parallel with the fresh standard solution. on the p-value intercept test ( Figure 5), we found the p-value test was occurred not over four hours storage standard regression which was near the critical point probability of rejecting H0. eventually in four hours regression p-value which was 0.0775>0.05 the H0 was accepted in this test it means the regression curve was parallel and fit with fresh regression standard and not differ significantly. It also shows that in this test the measurement results from the standard test over four hours storage in 4 o C cannot be used again as an standard.

■ CONCLUSION
This study concluded that the storage of the standard series of Cr (VI) for more than four hours after being prepared as a standard solution could cause decreasing slope value of the regression curve and affect the sensitivity of the measurement of Cr (VI). By modeling the p-value test vs time measurement, we could predict the critical point when the event that p-value <0.05 was occurred and caused reject H0 condition. There was a need for further studies that study the hourly changes in the regression slope of another unstable standard that could be determined when no longer be used from the time of its making.

Reagents and materials
All chemicals and solvents used were of pro analysis (pa) reagent grade quality were used without further purification with brand Merck grade produced in Germany. Doubly deionized distilled water was used throughout. Standard K2Cr2O7, 1,5-diphenylcarbazide, H2SO4, and acetone were used.

Apparatus
Equipment used to comprise this study were a refrigerator, laboratory glassware, pH indicator paper, and Shimadzu analytical balance. Shimadzu UV-Vis Spectrophotometer UV-1800 type double beam UV-vis spectrophotometers with quartz cell used for the solution electronic spectra and the absorbance of the produced complex ion-associates in the organic solvent.

Cr (VI) Standard Solution
The reagent was made by weighing 250 mg of 1,5-diphenylcarbazide and then dissolved in 50 mL of acetone with the dark container. Two litres of doubly deionized distilled water added a few drops of 4 N H2SO4 adjusted to pH of 1.5 with pH indicator paper. 0.0283gram K2Cr2O7 weighed and dissolved in 100 ml volumetric flask with doubly deionized distilled water. The standard series 0.1; 0.2; 0.4; 0.6; 0.8; and 1.0 ppm was allowed to stand for 5-10 minutes and then measure the absorbance with a UV-Vis Spectrophotometer at a wavelength of 543 nm.

Standard Cr (VI) measurement
The standard series was measured then stored in the refrigerator at 4 o C. Absorbance measurements carried out with intervals 0, 4, 7, 25, 29, and 32 hours from the initial reading.

182
(3) third, calculate the value of the significance level from the absolute value of the test statistic t = | t | using t distribution for two tails with degrees of freedom df= n1+n2-4 the significance value is known as p-value, which is the critical point of the probability of rejecting H0. And fourth draw statistical conclusions with a significance value of 5%, so that if the p-value <5% means that H0 is rejected.
There are also five steps statistical test to find out whether α1=α2: first determine the null hypothesis (H0) and its alternative hypothesis (H1): H0: α1=α2 H1: α1≠α2 second, calculate the combined slope value (bg) according to the formula: (6) third, calculate the t-test statistics using the following formula: fourth, calculating the value of the significance level (p-value) from the absolute value of the t-test statistic = | t | using Student's t distribution for two tails with degrees of freedom df= n1+n2-3 the significance value is known as p-value, which is the critical point of the probability of rejecting H0, and fifth step, draw statistical conclusions with a significance value of 5%, so that if the p-value <5% means that H0 is rejected.
After statistical test between regression slope and intercept completed, we modelled the p-value the into a time frame of measurement with intervals 0, 4, 7, 25, 29, and 32 hours.

■ DECLARATION
There is no conflict of interest from authors for this research.

■ ACKNOWLEDGEMENT
Our research team expressed gratitude and sincere appreciation to AKA Polytechnic Bogor that accommodated the laboratory, equipment, and material to ensure this research was done great.