The visible and ultraviolet spectra of molecules and ions are associated only with transition between electronic energy levels of certain types of groups of atoms within the molecule and do not characterize the molecule as a whole. In contrast, as we have already seen, absorption of radiation in the infra red region is sufficient to stimulate the whole molecule under investigation to undergo vibrational and rotational changes which do reflect the whole structural entity. The much larger energies involved in the absorption of radiation in the ultraviolet region cause electronic as well as associated vibrational and rotational changes. Since complete resolution of the electronic and and closely spased vibrational and rotational bands is not possible.
In contemporary practice the ultraviolet spectrum of particular compound is recordered in conjunction with other spectral data such as infra red in an attempt to deduce its molecular structure. Ultraviolet spectra, however, do not furnish prime information as such but tend to act as complimentary or even supplementary evidence to infra red. In discussing the following spectra it has been assumed, therefore, that functional groups have been successfully detected and we shall concern ourselves with ascertaining substitution patterns.
Applications of Ultraviolet Spectra
– Kinetic studies
– Structural differentiation
– Decomposition products
In this paper we will review abot the internationally jounal. The journal tittel is FUEL PARK WASTEWATER MONITORING WITH UV-VIS SPECTRA AND PARTIAL LEAST SQUARES MODELS.
The journal explaind about the UV-VIS analysis to monitor the fuel park wastewater. We know that the wastewater is big problem in our life and it can be measure by any way. For example we can measure the BOD5 (Biochemical Oxygen Demand). In BOD5, we analyse the oxygen rate in the water, and it is very important to know the living of the bioorganism in this water. If the BOD5 is high so the OD (oxygen demand) is low, it is mean that the polution is high.
The other wastewater analysis is COD (Chemical Oxygen Demand). It is slightly same with the BOD5 but in this analysis, we analyse the chemical content of the wastewater.
And other wastewater analysis is TSS (Total Suspended Suspension).
So, from this journal, the UV-VIS analysis is done for the 4 parameters, they are BOD5, COD, TSS and OG. But for OG, it is difficult to make the PLS calibration for OG. So we will explain about BOD5, COD, and TSS.
Given the high development and application potentials for on-line (automated sample extraction and delivery to the analyzer) and in-line (in situ analysis using a probe inside the process) measurements in wastewater quality monitoring, UV-Visible (UV-Vis) spectrophotometry can be very useful in this area, contributing to the correct operation of the treatment systems. UV-Vis spectrophotometry is fast and simple method that has been used for wastewater quality evaluation and organic matrix composition identification, since most organic compounds and a few soluble minerals (such as nitrates) absorb in the UV-Vis region. The reported quantitative environmental applications of UV-Vis spectrophotometry include the estimation of organic matter and nitrate in wastewaters, and determination of polycyclic aromatic hydrocarbons (PAH) in soils. In the present work UV-Vis spectra of samples collected in a fuel park Wastewater Treatment Plant (WWTP), after biological treatment, were acquired and used for the attempted development of Partial Least Squares (PLS) calibration models for four environmental monitoring parameters, namely, Chemical Oxygen Demand (COD), 5-day Biochemical Oxygen Demand (BOD5), Total Suspended Solids (TSS) and Oil & Grease (OG).
Companhia Logística de Combustíveis (CLC) is a logistic fuel company responsible for the operation of the multi-product pipeline between the Sines oil refinery and the Aveiras fuel park, Portugal. CLC’s activities involve the transport, storage and expedition of gaseous and liquid fuels, their clients being the company shareholders.
Wastewaters generated at the Aveiras fuel park include domestic effluents, liquid effluents with high hydrocarbon content due to occasional spills and aqueous effluents with detergents and lubricants, mainly from the gas cylinder washing zone.
The samples used in the present work were collected weekly at the discharge line of the WWTP biological step and analyzed at the Instituto Superior Técnico Laboratory of Analyses (LAIST, certified laboratory).
We use the UV-VIS analysis, UV-VIS Specord 200 Spectrophotometer. The analysis is done at the 190 to 500 nm wavelenght in a quartz cell of 10 mm path length, raw, and after appropiate dilution in order to overcome signal saturisation in the UV region.
So in this step, we collet the wastewater sample fron the CLC and use the UV-VIS to analyses based the 3 parameters, BOD5, COD and TSS.
PLS model development
The PLS regression method is the most widely used method for multivariate calibration and corresponds to a guided decomposition model where the dependent variables, Y, intervene directly in the decomposition of the independent variables, X. The purpose of this method is to determine a small number of latent factors that can predict Y (analytical data) using the data in X (spectra) as efficiently as possible. The UV-Vis spectral data obtained from CLC WWTP samples and the correspondent analytical reference data were exported to Grams 32/AI software extended with the PLSplus/IQ module (Thermo Galactic, USA) for PLS calibration development.
The essential requirement is choose the suitable wavelength. The UV-Vis region used in the developed PLS models was chosen by evaluation of the coefficient of determination (R2) for each wavelength of the data set and selection of the region with higher values for this coefficient.
The average results of each run were calculated and the quality of the results was evaluated by the root mean squared error of cross-validation, RMSECV, according to
where n is the number of samples in the training set, yi the measured value of sample i and ŷ\i the predicted value of sample i when the model is constructed without sample i.
The validation of the final PLS models was attempted with the test set and the difference between the predicted and the measured values for the wastewater parameters was expressed by the root mean squared error of prediction, RMSEP, calculated according to
where n is the number of samples in the test set, yi the measured value of sample i and ŷi the predicted value of sample i.
RESULT AND DISCUSSION
A PLS model using 5 latent factors with 0.92 R2 was achieved for COD (Fig. 1a). The obtained RMSECV was 3.3 mg O2/l and it should be noted that for COD measurements in the range of 5 to 50 mg O2/l the standard analytical method mean squared error (MSE) is 13 mg O2/l (APHA, 1995). For BOD5 a PLS model using 3 latent factors with 0.88 R2 and 3.1 mg O2/l RMSECV was developed (Fig. 1b). The standard analytical method MSE for this parameter is around 11 mg O2/l (detection limit of 2 mg O2/l) (APHA, 1995). The PLS model achieved for TSS used 6 latent factors and had 0.84 R2 and 1.2 mg O2/l RMSECV (Fig. 1c). The detailed results of the obtained PLS models are summarized in Table 1.
The validation of the proposed PLS calibration models was attempted with a set of external samples (test set) and the results for COD, BOD5 and TSS are presented in Fig. 2. The RMSEP was 3.8 mg O2/l for COD, 3.5 mg O2/l for BOD5 and 2.9 mg/l TSS.
The results of the COD PLS model validation presented in Fig. 2a are highly satisfactory. A PLS calibration model for the parameter COD based on UV-Vis spectral analyses of WWTP samples with 9 factors, 0.91 R2 and 23.5 mg O2/l RMSEP was already reported . However, the COD range covered by that PLS model was 50–500 mg O2/l and is not comparable to the low range of values encountered in CLC WWTP samples (26–65 mg O2/l).
From the result above we ca analyse the result.
There are significant deviation if the matrix composition changes. In fact, whereas COD analysis determines almost 100% of organic carbon, the spectrophotometric method includes only the part of the organic carbon matrix that absorbs light in the UV-Vis region and there might be some important fractions of organic carbon present that do not show absorption in the UV-Vis spectra, such as short chain fatty acids, sugars and starch.
Although BOD5 is a commonly used parameter for wastewater quality characterization, several problems are inherent to its standard analytical method, since most of the variables present are difficult to control during the analysis, including seed quality, presence and concentration of natural biological inhibitors, the time taken to complete the analysis, and test method accuracy. The results presented in Fig. 2b for the BOD5 PLS model validation showed higher predicted values than the standard analytical method but with a low RMSEP (3.5 mg O2/l) value when compared to the standard analysis MSE (11 mg O2/l). However, it should be noted that this method was developed and validated with a limited number of samples of a narrow interval of BOD5 values very close to the reference method detection limit. For that reason the presented results can only suggest that the development of PLS calibration models for BOD5 based on the UV-Vis spectra of wastewater samples is a promising technique for real-time estimation of BOD5. This was already attempted by a near infrared (NIR) transmission PLS calibration model with 28.6 mg O2/l RMSECV and 28.7 mg O2/l RMSEP for BOD5 values ranging from 35 to 275 mg O2/l.
The calibration model for TSS was developed with a limited number of samples of a narrow interval of values. The validation results presented in Fig. 2c show that the predicted TSS values can be higher or lower than the concentration values determined by the standard method, suggesting that UV-Vis spectra may contain more information related to the suspended organic matter than the simple standard TSS analysis. In fact, different wastewater samples with the same concentration of organic carbon could be distinguished by UV-Vis analysis, as well as different suspensions with the same TSS concentration.
The preliminary results of the present work showed that UV-Vis spectra of aqueous wastewater samples contain information that can be extracted and used for WWTP monitoring. In the particular case of the treated wastewater samples from CLC, the use of spectrophotometric techniques in the UV-Vis region and their chemometric treatment (through PLS models) revealed a high application potential as a fast, simple and cost-effective (no sample preparation, no reagents consumption) method for COD determination, and also some potential for the parameters BOD5 and TSS, in spite of the models being developed in a narrow interval of values close to the standard analytical methods’ detection limits. The use of PLS UV-Vis calibration models for the fast determination of COD or BOD5 can constitute a highly advantageous alternative to the analyzers installed at CLC (total organic carbon and OG) for the on-line or in-line control of the discharged treated wastewater quality.