- Email: [email protected]
Multicriteria decision analysis in ranking of analytical procedures for aldrin determination in water
Journal of Chromatography A, 1387 (2015) 116–122
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Multicriteria decision analysis in ranking of analytical procedures for aldrin determination in water Marek Tobiszewski a,∗ , Aleksander Orłowski b a b
Department of Analytical Chemistry, Chemical Faculty, Gda´ nsk University of Technology, 11/12 G. Narutowicza St., 80-233 Gda´ nsk, Poland Department of Management, Faculty of Management and Economics, Gda´ nsk University of Technology, 11/12 G. Narutowicza St., 80-233 Gda´ nsk, Poland
a r t i c l e
i n f o
Article history: Received 10 January 2015 Received in revised form 2 February 2015 Accepted 3 February 2015 Available online 11 February 2015 Keywords: Green analytical chemistry Decision making Environmental impact assessment Analytical methodologies PROMETHEE Multi-criteria decision analysis
a b s t r a c t The study presents the possibility of multi-criteria decision analysis (MCDA) application when choosing analytical procedures with low environmental impact. A type of MCDA, Preference Ranking Organization Method for Enrichment Evaluations (PROMETHEE), was chosen as versatile tool that meets all the analytical chemists – decision makers requirements. Twenty five analytical procedures for aldrin determination in water samples (as an example) were selected as input alternatives to MCDA analysis. Nine different criteria describing the alternatives were chosen from different groups – metrological, economical and the most importantly – environmental impact. The weights for each criterion were obtained from questionnaires that were sent to experts, giving three different scenarios for MCDA results. The results of analysis show that PROMETHEE is very promising tool to choose the analytical procedure with respect to its greenness. The rankings for all three scenarios placed solid phase microextraction and liquid phase microextraction – based procedures high, while liquid–liquid extraction, solid phase extraction and stir bar sorptive extraction – based procedures were placed low in the ranking. The results show that although some of the experts do not intentionally choose green analytical chemistry procedures, their MCDA choice is in accordance with green chemistry principles. The PROMETHEE ranking results were compared with more widely accepted green analytical chemistry tools – NEMI and Eco-Scale. As PROMETHEE involved more different factors than NEMI, the assessment results were only weakly correlated. Oppositely, the results of Eco-Scale assessment were well-correlated as both methodologies involved similar criteria of assessment. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Green analytical chemistry is laboratory philosophy that recognizes the certain analytes determination process as an environmental problem. During analytical procedure, certain compounds can be emitted to the environment [1]. Application of green analytical chemistry means the reduction of procedural environmental impact [2]. The most common methods of reducing negative influence is the application of miniaturized extraction techniques at the stage of analytical sample preparation and miniaturization of analytical instruments [3,4]. The other approach is elimination of sample preparation step by application of direct analytical techniques, which in fact can be applied rather seldom [5,6]. These solutions in greening the analytical procedures are relatively well established and have been reviewed exhaustively [7–10].
∗ Corresponding author at. Tel.: +48583472194. E-mail address: [email protected] (M. Tobiszewski). http://dx.doi.org/10.1016/j.chroma.2015.02.009 0021-9673/© 2015 Elsevier B.V. All rights reserved.
One of the main problems in the field of green analytical chemistry is the assessment of analytical procedures greenness. There are several still very vital questions. How to measure the procedural greenness? How to compare the environmental impact of methodologies to choose the most environmentally friendly one? Can an universal procedure be developed to help or guide in procedure selection for the given purpose? There are few assessment methods available, each of them has its own advantages and disadvantages. The only assessment procedure, that is brought into a broader laboratory practice, is labeling with NEMI symbols [11]. NEMI symbol is easy to read circle with its fields filled green (or not) if certain assumptions are fulfilled. These are generation of wastes in the amount lesser than 50 g, no corrosive environment is present and there are no hazardous or dangerous chemicals used during analytical procedure. The main drawback in NEMI application is the need to search the hazardous substances lists to check if any of chemicals used in the procedure is present on these lists, which is time-consuming task. The other NEMI drawback is that the information carried by the NEMI symbol is not quantitative. This problem is partially solved when NEMI modification
M. Tobiszewski, A. Orłowski / J. Chromatogr. A 1387 (2015) 116–122
Stating the decision problem Choosing the procedure for aldrin in water determination Selection of alternatives Searching for adequate procedures Selection of criteria Metrological, economical, environmental criteria Weights given by experts As a result of answers to questionnaire
Weighting the criteria Setting the relative importance of each criterion
117
when concerning PROMETHEE as the tool in assessment of green analytical chemistry metrics. It seems to be important to mention that there were several rankings of MCDA methods published [20] and PROMETHEE is concerned to be one of the top in alternatives ranking, easiness of use and potential of future development. The aim of the study is to investigate the possibility of application of multicriteria decision analysis (PROMETHEE) as the analytical procedure assessment tool in the light green analytical chemistry. The second aim is to show that environmental impact aspects should be considered when making decision of choosing the analytical procedure for the given purpose. Some remarks on choosing the best procedure will be also provided.
PROMETHEE analysis With ranking of analytical procedures as the result Choosing the best alternative The first from the ranking Fig. 1. The scheme of analytical methodologies assessment procedure with multicriteria decision PROMETHEE technique.
proposed by de la Guardia and Armenta is applied [12]. In such a case NEMI symbol carries the semi-quantitative information about threat in each of its categories. The other approach in assessment of analytical procedures greenness is application of Eco-Scale [13]. The Eco-Scale involves many factors that can have negative impact on the environment such as the amount and the type of any chemical used in the procedure, the mass of generated waste, the occupational exposure and energy consumption. The result of Eco-Scale application is the number lower than one hundred, the closer the values to one hundred the more green analysis. The main disadvantage of application of Eco-Scale is the fact that the result is the number which in fact involves many parameters but it does not carry any information about the nature of threats related to the analytical procedure. The latest approach, that gives the possibility to classify the analytical procedures taking into consideration their environmental impact, is the application of multivariate statistical techniques [14]. The techniques such as self-organizing maps allow to group the analytical procedures according to their similarity when environmental, metrological and technical parameters, describing the analytical procedure, are considered. The drawback of multivariate statistics application is tedious assessment procedure and that it might be applied to compare limited amount of procedures [15]. It is clear that green analytical chemistry still needs the reliable procedural assessment methods. Multicriteria decision analysis is used in environmental science and environmental management [16]. It has been successfully applied to select proper remediation procedure for contaminated mine treatment [17], method of water disinfection [18] or choosing priority areas for environmental monitoring [19]. All these examples involve social, environmental, economical, technological criteria during the decision process. The Preference Ranking Organization Method for Enrichment Evaluations (PROMETHEE) methods are chosen from available multicriteria decision analysis methods because of two reasons. First of all they are easy to apply compared to other methods (such as Elimination and Choice Expressing Reality – ELECTRE or Analytic Hierarchy Process – AHP) because PROMETHEE methods require fewer parameters from decision makers. They also rank alternatives as well as identify the best alternative, whereas the previously mentioned utility theory methods only identify the best alternatives [17]. Ranking alternatives seems to be the most important because there are 25 alternatives taken into consideration and it is not only important to know which is the best among all but also to make it possible to compare all the alternatives. It is important aspect
2. Materials and methods The general scheme of the presented multicriteria decision analysis assessment approach is shown in Fig. 1. Briefly, after stating the decision problem, the set of alternatives (the possible options) is identified. Then the important aspects – criteria, have to be defined, together with their relative importance by assigning weights to the criteria.
2.1. The PROMETHEE method The PROMETHEE (Preference Ranking Organization Method for Enrichment Evaluations) is one of the leading MCDA (Multi-Criteria Decision Aid or Analysis) methods that was developed by Brans and Vincke in 1985 [21]. The methodology is successfully used in various areas such as business, public administration, industrial location, medicine or tourism [22]. PROMETHEE is a family of outranking methods that consist of PROMETHEE I which generates partial ranking, PROMETHEE II for complete ranking of alternatives, PROMETHEE III for ranking based on interval, PROMETHEE IV for ranking (partial or complete) of alternatives set of viable solutions is continuous, PROMETHEE V for problem with segmentation constraints, PROMETHEE VI for human brain representation. There are also additional methods: PROMETHEE GDSS for group-decision making, PROMETHEE TRI used when sorting problems appear and PROMETHEE CLUSTER for nominal classification [22]. In the presented paper PROMETHEE II is going to be applied, because the ranking of many alternatives is the most desired analysis output from our point of view. The principle of PROMETHEE II is based on pairwise comparison of alternatives along each criterion. To implement the methodology it is required to have two types of information: weights of the criteria and the preference function. In order to obtain information within the criteria, a preference function for each criterion, expressing the difference in performance of alternative a over alternative b must be identified, adopting as a result the pairwise comparison approach [20]. It is also requested to have clear and understandable (by both decision-maker and analysts) information between the criteria and information within each criterion. When A of n alternatives (a1 , a2 , . . ., an ) have to be ranked and G of k criteria (g1 , g2 , . . ., gk ) have to be maximized (or minimized) the resulting problem has the form as shown [17]: max{g1 (a), g2 (a), . . ., gk (a)|a ∈ A} Some criteria might be maximized whereas others might be minimized, the goal of the decision-maker is to identify an alternative which is optimizing all criteria. The detailed mathematical description of PROMETHEE was presented by its developers Brans and Vincke in [21].
118
M. Tobiszewski, A. Orłowski / J. Chromatogr. A 1387 (2015) 116–122
Table 1 The criteria being input data to the multicriteria decision analysis. No.
Criterion
Description
Preference function
1
Variable with linguistic description: 0 – hardly available; 0.25 – poorly available; 0.5 – moderately available; 0.75 – rather available; 1 – easily available. The lowest concentration of analyte that can be detected. Expressed as coefficient of variance. The percentage of analyte that is recovered with a procedure.
The higher the better Range 0–1
2 3 4
Availability of analytical equipment in the laboratory needed for performing the analytical procedure Limits of detection Precision of the methodology Recovery of analytes
5
The number of other analytes
6 7 8
Time of analysis Amount of organic solvent The hazards and toxicity of the solvents and reagents applied in the procedure The amount of solid wastes
The number of other analytes apart from aldrin that can be determined in the single run. The total time to prepare and analyze the analytical sample. Total amount of organic solvents that are used in the procedure. The sum of all hazards connected application of given solvent times its amount. The total amount of solid wastes that are generated in the analytical run.
9
2.2. The alternatives All the analytical procedures are dedicated to determine aldrin in water samples. The data are collected by searching the scientific databases. The required information (listed in Table 1) was extracted from articles or standard procedures. The databases that were searched are ACS, RSC, ScienceDirect, Springerlink, Wiley. If any of the information was not available in the source article such procedure was excluded from further consideration. Most of the criteria are assessed with the values that are directly taken from the literature data. The only criterion that has linguistic description is the availability of the laboratory equipment. The translation of this criterion from linguistic description to numerical value is presented in Table 1. The availability estimation for the analytical equipment is arbitrary and was done by ourselves. All the preference functions for all the criteria seem to be obvious and easy to be implemented in PROMETHEE. Only the recovery value criterion is non-standard one. We have taken the absolute value of the difference between the analytical procedure recovery value and 100%. In this sense the recovery values of 95% and 105% are equally good. The analytical procedures included in the analysis are based on various sample preparation techniques but all of them are based on gas chromatographic separation. The alternatives for multicriteria decision analysis are presented in Table 2. As an assumption at the data preprocessing step, all methodologies with LOD > 10 ng L−1 were excluded from further consideration, as they do not meet the requirement set by Maximum Allowable Concentration for aldrin in water samples.
The lower the better The lower the better The closer to 100% the better The higher the better The lower the better The lower the better The lower the better The lower the better
Table 2 The alternatives considered during multicriteria decision making. Alternative No.
Methodology abbreviation
Reference
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
6630 – LLE–GC–ECD 505 LLE–GC–ECD O-1104 LLE–GC–ECD SPE–GC–GC–TOF–MS SPE–GC–MS SPE–GC–MS 508.1 SPE–GC–ECD SPE–GC–ECD On-line-SPE–GC–MS On-line-SPE–PTV–GC–MS MASE–LVI–GC–MS HS–SPME–GC–MS–MS SPME–GC–MS–MS SPME–GC–ECD SPME–GC–MS SBSE–TD–GC–GC–MS SBSE–GC–GC–TOF–MS SBSE–TD–GC–MS DLLME–GC–MS DLLME–GC–MS LPME–GC–ECD HF–LPME–GC–MS–MS DHS–TEH–LPME–GC–MS–MS MLLE–LVOCI–GC–ECD MLLE–LVI–NICI–MS
[23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [35] [44] [45] [46]
DHS – dynamic headspace; DLLME – dispersive liquid–liquid microextraction; ECD – electron capture detection; GC – gas chromatography; HF – hollow fiber; HS – headspace; LLE – liquid–liquid extaction; LPME – liquid phase microextraction; MASE – membrane-assisted solvent extraction; MLLE – micro liquid–liquid extraction; MS – mass spectrometry; PTV – programmed-temperature vaporization; SBSE – stir-bar sorptive extraction, SPE – solid phase extraction; SPME – solid-phase microextraction; TD – thermal desorption; THE – time-extended helix liquid-phase microextraction; TOF – time-of-flight.
2.3. The questionnaire The questionnaire was created to establish weights for each criterion. The weight of a criterion is a positive number that represents the criterion relative importance [47]. The higher the weight, the more important is the criterion [48]. The questionnaire was designed in such a way that the importance of each criterion (listed in Table 1) could be assessed verbally in seven – point scale from “very low” trough “low”, “medium low” “medium”, “medium high”, “high” to “very high”. Such an approach allows to easily fill the questionnaire by the expert. Additionally, the questionnaire contained the question about expert’s attitude towards green analytical chemistry with three possible answers – (a) I always choose green analytical methodologies, (b) greenness of analytical methodology is minor aspect, I choose greener methodology only if it is cheaper, faster and has beneficial metrological parameters, (c) I do not consider the environmental impact of the analytical procedure. The question was important for the assessment of
PROMETHEE as the tool in green analytical chemistry metrics and gives information about possible scenarios. Such questionnaires were sent to experts who were asked to assess the importance of each criterion when choosing analytical procedure. 3. Results and discussion 3.1. Assigning weights for the criteria To perform the PROMETHEE analysis it was necessary to define the importance of criteria which are going to be included in analysis. There are nine criteria (Table 1) which are taken into consideration in the process of selecting the best alternative and ranking the remaining ones. Assigning weights to criteria involves priorities of the decision-maker because not all criteria have equal importance. Due to this fact it was necessary to assign the
M. Tobiszewski, A. Orłowski / J. Chromatogr. A 1387 (2015) 116–122
119
Table 3 The results of the questionnaires together with assigned weights. Criterion
1 2 3 4 5 6 7 8 9 Total
Questionnaire 1
Questionnaire 2
Questionnaire 3
Mean of questionnaires
Verbal description
Points
Weight
Verbal description
Points
Weight
Verbal description
Points
Weight
Verbal description
Points
Weight
High High High Medium Medium low Medium Very high Very high High
6 6 6 4 3 4 7 7 6 49
0.122 0.122 0.122 0.081 0.061 0.081 0.142 0.142 0.122 1
Very high High High Very high High Medium high Medium Medium Medium
7 6 6 7 6 5 4 4 4 49
0.142 0.122 0.122 0.142 0.122 0.102 0.081 0.081 0.081 1
Very high High Very high High High High Low Very low Medium high
7 6 7 6 6 6 2 1 5 46
0.152 0.13 0.152 0.13 0.13 0.13 0.043 0.021 0.108 1
Very high High High High Medium high Medium high Medium Medium Medium high
7 6 6 6 5 5 4 4 5 48
0.146 0.125 0.125 0.125 0.104 0.104 0.083 0.083 0.104 1
numerical value to the importance of each criteria. Several methods of assigning weights were considered by authors among them making decision independently basing on authors’ experience or using experts’ knowledge. Having in mind the importance of the decision (influence on the final results) we decided to use experts’ knowledge. It is assumed that experts have wide knowledge and they are objective in their jurisdictions–they were not informed about the details of the case but were just asked to assign importance to the criteria. The experts are analytical chemists, who have great experience with various analytical procedures. We have tried to select experts with different approach to the green analytical chemistry philosophy. Because there was just a need to evaluate the criteria, typical interview with experts seemed not to be necessary. It was decided to prepare a dedicated questionnaire (Section 2.3) that was sent to the experts. Designing the questionnaire we have considered the manual assigning of numerical values of weights by experts. This would mean that experts have to give fraction numbers to each criteria and they should sum up to 1. Such an approach gives confusing questionnaire to be filled by experts and there is a possibility that one of the criterion obtains weight considerably higher than remaining ones, what brings multicriteria analysis to virtually single-criterion analysis. We wanted to avoid such situation, however such an approach is also possible. We have decided for easier, for experts, approach – giving descriptive answers to describe criteria importance. Next, it was necessary to transform the experts’ answers to the weights to be used in the analysis. In PROMETHEE method (other than in e.g. AHP) there is no precisely described way of setting weighs of criteria [49]. We considered two methods of setting weights, for both there is no objection to consider normalized weights (sum = 1). Constant value difference between weights – this method presents constant degree between each of the weights representation. The method was implemented (data not published) but has an important disadvantage: it does not take into consideration the expert’s tendencies to either very high or very low evaluation of criteria. ‘In the second method to each possible answer there are numerical values (points) assigned, from 1 assigned to “very low” to 7 for “very high”. Basing on the answers given by the expert the total sum of values of answers is generated. Next the value of each answer is divided by the total sum of points and as the result the weights are generated. Next there was a need to decide how the weights are going to be interpreted. It is possible to generate the average answer for all three experts or to prepare PROMETHEE analysis for individual weights by each expert and threat the results as scenario. Both methods were tried, the results for the scenarios and mean answers for the scenarios are presented in Table 3. Referring to the selection of experts, none of them was seen as more important so the importance of their answers was equal. Basing on the answers received from experts, it was possible to use an
average importance (based on arithmetic mean) for each criterion (Table 3). It brings the opportunity to present the knowledge from wider perspective (three different perspectives from three experts) and to compare results gained from different experts (different points of view). The expert filling the questionnaire 1 indicated that he always chooses the analytical procedures according to green analytical chemistry principles. The remaining two experts admitted that they choose green analytical chemistry procedures when other decision criteria are met. 3.2. The ranking with PROMETHEE Table 4 presents the results of analytical methodologies for aldrin determination ranking with PROMETHEE method. The results for the respective scenarios do not differ significantly, the rankings are different in details only. Generally, the first 10 analytical procedures in rankings are based on SPME or various modes of liquid–liquid microextractions, with either mass spectrometric of electron capture detection. These techniques are characterized by good analytical parameters, minute amounts of solvents consumption and were assessed as requiring rather available equipment. In the second part of ranking procedures based on SBSE, SPE and LLE sample preparation and procedures involving more sophisticated apparatus like TOF–MS or MS–MS detectors are placed. These kind of detectors in these study were assessed as less available than ECD or MS. The procedures involving SPE or LLE require considerable amounts of solvents that are often toxic and usually produce solid waste. The PROMETHEE algorithm allowed to rank the analytical procedures according to their greenness but also taking into account other factors. The results of ranking are in agreement with widely accepted reception of analytical procedures greenness. What is worth mentioning the standard procedures were ranked very low. This suggests that other alternatives, due to their attractive features, would be preferentially chosen by analysts over these standard procedures. None of the experts indicated that the environmental impact of the procedure, when making selection decision, is not considered. The results of the ranking, for expert who admitted that he always chooses green analytical chemistry procedures, are in close agreement to those for experts that choose green procedures when other factors are beneficial. This shows that analyst can choose green analytical procedures without sacrificing other analytical procedure parameters, such as metrological ones. The second conclusion is that the decision makers who choose analytical procedures, taking into account other than environmental factors, unwittingly choose procedures that meet green analytical chemistry standards. The application of PROMETHEE as green analytical chemistry assessment tool has the advantage of possibility of including other parameters. In this study, apart from metrological and environmental ones, we have included the “availability of equipment” criterion.
120
M. Tobiszewski, A. Orłowski / J. Chromatogr. A 1387 (2015) 116–122
Table 4 The results of PROMETHEE ranking of analytical procedures for different scenarios, based on questionnaires answers. As the abbreviations for some procedures are doubled, the alternatives are presented with numbers from Table 2. Rank
Scenario 1
Scenario 2
Scenario 3
Mean of scenarios
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
14. SPME–GC–ECD 24. MLLE–LVOCI–GC–ECD 19. DLLME–GC–MS 15. SPME–GC–MS 12. HS–SPME–GC–MS–MS 25. MLLE–LVI–NICI–MS 20. DLLME–GC–MS 23. DHS–TEH–LPME–GC–MS–MS 13. SPME–GC–MS–MS 21. LPME–GC–ECD 10. On-line-SPE–PTV–GC–MS 2. 505 LLE–GC–ECD 16. SBSE–TD–GC–GC–MS 9. On-line-SPE–GC–MS 18. SBSE–TD–GC–MS 8. SPE–GC–ECD 22. HF–LPME–GC–MS–MS 6. SPE–GC–MS 4. SPE–GC–GC–TOF–MS 11. MASE–LVI–GC–MS 17. SBSE–GC–GC–TOF–MS 5. SPE–GC–MS 7. 508.1 SPE–GC–ECD 1. 6630 LLE–GC–ECD 3. O-1104 LLE–GC–ECD
14. SPME–GC–ECD 24. MLLE–LVOCI–GC–ECD 19. DLLME–GC–MS 25. MLLE–LVI–NICI–MS 15. SPME–GC–MS 20. DLLME–GC–MS 12. HS–SPME–GC–MS–MS 13. SPME–GC–MS–MS 23. DHS–TEH–LPME–GC–MS–MS 2. 505 LLE–GC–ECD 21. LPME–GC–ECD 10. On-line-SPE–PTV–GC–MS 18. SBSE–TD–GC–MS 8. SPE–GC–ECD 22. HF–LPME–GC–MS–MS 11. MASE–LVI–GC–MS 16. SBSE–TD–GC–GC–MS 6. SPE–GC–MS 9. On-line-SPE–GC–MS 4. SPE–GC–GC–TOF–MS 5. SPE–GC–MS 1. 6630 LLE–GC–ECD 7. 508.1 SPE–GC–ECD 17. SBSE–GC–GC–TOF–MS 3. O-1104 LLE–GC–ECD
24. MLLE–LVOCI–GC–ECD 14. SPME–GC–ECD 19. DLLME–GC–MS 25. MLLE–LVI–NICI–MS 20. DLLME–GC–MS 15. SPME–GC–MS 12. HS–SPME–GC–MS–MS 23. DHS–TEH–LPME–GC–MS–MS 13. SPME–GC–MS–MS 10. On-line-SPE–PTV–GC–MS 21. LPME–GC–ECD 2. 505 LLE–GC–ECD 18. SBSE–TD–GC–MS 8. SPE–GC–ECD 11. MASE–LVI–GC–MS 4. SPE–GC–GC–TOF–MS 6. SPE–GC–MS 16. SBSE–TD–GC–GC–MS 9. On-line-SPE–GC–MS 22. HF–LPME–GC–MS–MS 5. SPE–GC–MS 1. 6630 LLE–GC–ECD 7. 508.1 SPE–GC–ECD 3. O-1104 LLE–GC–ECD 17. SBSE–GC–GC–TOF–MS
14. SPME–GC–ECD 24. MLLE–LVOCI–GC–ECD 19. DLLME–GC–MS 25. MLLE–LVI–NICI–MS 15. SPME–GC–MS 20. DLLME–GC–MS 12. HS–SPME–GC–MS–MS 13. SPME–GC–MS–MS 23. DHS–TEH–LPME–GC–MS–MS 21. LPME–GC–ECD 2. 505 LLE–GC–ECD 10. On-line-SPE–PTV–GC–MS 18. SBSE–TD–GC–MS 8. SPE–GC–ECD 16. SBSE–TD–GC–GC–MS 9. On-line-SPE–GC–MS 22. HF–LPME–GC–MS–MS 11. MASE–LVI–GC–MS 6. SPE–GC–MS 4. SPE–GC–GC–TOF–MS 5. SPE–GC–MS 1. 6630 LLE–GC–ECD 7. 508.1 SPE–GC–ECD 17. SBSE–GC–GC–TOF–MS 3. O-1104 LLE–GC–ECD
The other criteria that can be included might be: cost of reagents per sample, apparatus cost, the ability of laboratory staff to perform given analysis (by Bernoulli or fuzzy description) or sample analysis throughput. In more advanced multicriteria decision analysis risk assessment or reagents delivery chain can be included. The criteria can be described by discrete or fuzzy numbers, “yes” or “no” answers or verbal description can be applied and translated into numbers, what makes the MCDA procedure more convenient to use. The PROMETHEE V, another type of PROMETHEE methodology, enables user to define constraints which is used in selecting a best subset of action. Such constraint might be for example maximum budget defined for the project which seems to be important in the process of selecting analytic procedure for every individual decision-maker. As general, in PROMETHEE it is fast and easy to change the weights which practically will be important in adapting the process of choosing analytical procedure to individual case. The individual user can also change his preferences according to his needs by modifying weights. For example, when large amount of samples has to be analyzed in short time, the analysis time becomes the most important factor and high weight is assigned to this criterion. PROMETHEE enables also the possibility of visualization of the results. The PROMETHEE I enables to assess quantitatively, how each criterion of the alternative influences the score. 3.3. The comparison of the results with other assessment methodologies The most widely accepted green analytical chemistry metrics tool is NEMI labeling. Fig. 2a shows the correlation of NEMI score with PROMETHEE ranking. There is virtually no correlation between the results with these two assessment tools. NEMI symbols take into consideration 4 parameters only in Bernoulli distribution – each criterion is met or not. PROMETHEE ranking, in this case, takes into consideration 9 criteria and the numerical values for each criteria can be any number from a given range. This results in fact that the NEMI results have considerably lower “resolution” as
they can be natural numbers from 0 to 4. The second reason for such lack of correlation is the fact that during PROMETHEE ranking other than environmental factors, such as “availability of equipment” or “analysis time”, have been considered. The second tool in analytical procedures environmental impact assessment that gains in interest is Eco-scale. The correlation of Eco-scale with PROMETHEE results is more legible than in case of NEMI labeling results (Fig. 2b). Eco-scale in its procedure involves more parameters than NEMI and is more similar to our case of PROMETHEE ranking. Eco-scale quantitatively considers the amount and the toxicity of each reagent consumed in analytical procedure. The analytical procedure (alternative in MCDA nomenclature), that is the mostly outlying point, is the procedure based on SBSEGC × GC-TOF-MS. It is ranked at 24 position, while its NEMI score is 4 and Eco-scale score is 72. This procedure is based on extraction with SBSE followed by thermal desorption and analysis with two dimensional gas chromatography and time-of-flight mass spectrometer. The procedure involves small amount of reagents, so NEMI and Eco-scale scores are high. On the other hand sophisticated equipment needed, relatively poor metrological merits result in such a low PROMETHEE ranking. In contrast to NEMI and Eco-scale, PROMETHEE has the advantage that it allows to choose the greenest option among alternatives (at least two analytical procedures) by direct comparison. It gives possibility to include in the assessment procedure other criteria than those included in the NEMI and Eco-scale techniques. Some users could find the application of weights for the criteria confusing, as they may change significantly when different analytes and matrices are considered. We see user-defined criteria weights as advantage, because such approach makes the PROMETHEEbased assessment procedure very specific. User can modify the assessment procedure to his needs by choosing and weighting criteria. Concerning PROMETHEE as standard procedure in greenness assessment would require defined approach to choosing criteria and assigning weights. PROMETHEE as assessment procedure should be considered as rather tedious but its result is very easy to be interpreted.
M. Tobiszewski, A. Orłowski / J. Chromatogr. A 1387 (2015) 116–122
121
Fig. 2. NEMI (a) and Eco-scale (b) results correlation to PROMETHEE mean ranking.
4. Conclusions
References
The PROMETHEE procedure applied in selection of the most appropriate analytical methodologies seems to be useful tool. It allows user to select the best solution in proper way concerning various, often contradictory, criteria. If some of the criteria are describing environmental impact of analytical procedure, then the PROMETHEE becomes very useful tool in assessment of green analytical chemistry metrics. Although some of the presented scenarios did not consider environmental criteria as important, the results of PROMETHEE ranking allowed to choose the analytical procedures with low environmental impact. The presented case shown that although the given weights are slightly different for every expert the ranking results are very similar. The analysts who claim not to follow green analytical chemistry path, choose the analytical procedures with low environmental impact. PROMETHEE as green analytical chemistry tool has the key advantage: it includes user or decision maker personal preferences which allows to select criteria and their weights. It has the great potential to include other criteria important for every decision maker. It also has the potential to be combined with other tools such as life-cycle assessment.
[1] S. Armenta, S. Garrigues, M. de la Guardia, Green analytical chemistry, TrAC Trends Anal. Chem. 27 (2008) 497–511. ´ [2] M. Tobiszewski, A. Mechlinska, J. Namie´snik, Green analytical chemistry—theory and practice, Chem. Soc. Rev. 39 (2010) 2869–2878. [3] A. Spietelun, Ł. Marcinkowski, M. de la Guardia, J. Namie´snik, Recent developments and future trends in solid phase microextraction techniques towards green analytical chemistry, J. Chromatogr. A 1321 (2013) 1–13. [4] A. Spietelun, Ł. Marcinkowski, M. de la Guardia, J. Namie´snik, Green aspects, developments and perspectives of liquid phase microextraction techniques, Talanta 119 (2014) 34–45. [5] A. Mir-Marqués, S. Garrigues, M.L. Cervera, M. de la Guardia, Direct determination of minerals in human diets by infrared spectroscopy and X-ray fluorescence, Microchem. J. 117 (2014) 156–163. [6] M. Tobiszewski, J. Namie´snik, Direct chromatographic methods in the context of green analytical chemistry, TrAC Trends Anal. Chem. 35 (2012) 67–73. [7] H. Shaaban, T. Górecki, Current trends in green liquid chromatography for the analysis of pharmaceutically active compounds in the environmental water compartments, Talanta 132 (2015) 739–752. [8] M. Kaljurand, M. Koel, Recent advancements on greening analytical separation, Crit. Rev. Anal. Chem. 41 (2011) 2–20. [9] M. Koel, M. Kaljurand, Application of the principles of green chemistry in analytical chemistry, Pure Appl. Chem. 78 (2006) 1993–2002. [10] M. Farré, S. Pérez, C. Gonc¸alves, M.F. Alpendurada, D. Barceló, Green analytical chemistry in the determination of organic pollutants in the aquatic environment, TrAC Trends Anal. Chem. 29 (2010) 1347–1362. [11] L.H. Keith, L.U. Gron, J.L. Young, Green analytical methodologies, Chem. Rev. 107 (2007) 2695–2708. [12] S. Garrigues, M. de la Guardia, Handbook of Green Analytical Chemistry, Wiley, 2012. [13] A. Gałuszka, Z.M. Migaszewski, P. Konieczka, J. Namie´snik, Analytical Eco-Scale for assessing the greenness of analytical procedures, TrAC Trends Anal. Chem. 37 (2012) 61–72. [14] M. Tobiszewski, S. Tsakovski, V. Simeonov, J. Namie´snik, Application of multivariate statistics in assessment of green analytical chemistry parameters of analytical methodologies, Green Chem. 15 (2013) 1615–1623.
Acknowledgement The authors would like to thank the experts for their help by filling the questionnaires.
122
M. Tobiszewski, A. Orłowski / J. Chromatogr. A 1387 (2015) 116–122
[15] M. Tobiszewski, S. Tsakovski, V. Simeonov, J. Namie´snik, Multivariate statistical comparison of analytical procedures for benzene and phenol determination with respect to their environmental impact, Talanta 130 (2014) 449–455. [16] I.B. Huang, J. Keisler, I. Linkov, Multi-criteria decision analysis in environmental sciences: ten years of applications and trends, Sci. Total Environ. 409 (2011) 3578–3594. [17] G.D. Betrie, R. Sadiq, K.A. Morin, S. Tesfamariam, Selection of remedial alternatives for mine sites: a multicriteria decision analysis approach, J. Environ. Manage. 119 (2013) 36–46. [18] M.D. Gomez-Lopez, J. Bayo, M.S. Garcıa-Cascales, J.M. Angosto, Decision support in disinfection technologies for treated wastewater reuse, J. Clean. Prod. 17 (2009) 1504–1511. [19] M. Alvarez-Guerra, J.R. Viguri, N. Voulvoulis, A multicriteria-based methodology for site prioritisation in sediment management, Environ. Int. 35 (2009) 920–930. [20] M. Cinelli, S.R. Coles, K. Kirwan, Analysis of the potentials of multi criteria decision analysis methods to conduct sustainability assessment, Ecol. Indicators 46 (2014) 138–148. [21] J.P. Brans, P. Vincke, A preference ranking organization method: the PROMETHEE method for multiple criteria decision-making, Manage. Sci. 31 (1985) 647–656. [22] M. Behzadian, R.B. Kazemzadeh, A. Albadvi, M. Agdasi, PROMETHEE: a comprehensive literature review on methodologies and applications, Eur. J. Oper. Res. 200 (2010) 198–215. [23] 6630C Standard Methods Online – Standard Methods for the Examination of Water and Wastewater. [24] 505 Methods for the Determination of Organic Compounds in Drinking Water – Supplement III (EPA/600/R-95-131). [25] O-1104 Methods for the Determination of Organic Substances in Water and Fluvial Sediments, Techniques of Water-Resources Investigations of the United States Geological Survey, Book 5, Chapter A3 Edited by R.L. Wershaw, M.J. Fishman, R.R. Grabbe, and L.E. Lowe. [26] V. Matamoros, E. Jover, J.M. Bayona, Part-per-trillion determination of pharmaceuticals, pesticides, and related organic contaminants in river water by solid-phase extraction followed by comprehensive two-dimensional gas chromatography time-of-flight mass spectrometry, Anal. Chem. 82 (2010) 699–706. [27] C.C. Leandro, D.A. Bishop, R.J. Fussell, F.D. Smith, B.J. Keely, Semiautomated determination of pesticides in water using solid phase extraction disks and gas chromatography–mass spectrometry, J. Agric. Food Chem. 54 (2006) 645–649. [28] S. Barreka, C. Cren-Olivéa, L. Wiest, R. Baudot, C. Arnaudguilhema, Marie-Florence Grenier-Loustalot Multi-residue analysis and ultra-trace quantification of 36 priority substances from the EuropeanWater Framework Directive by GC–MS and LC-FLD-MS/MS in surface waters, Talanta 79 (2009) 712–722. [29] 508.1 Methods for the Determination of Organic Compounds in Drinking Water – Supplement III (EPA/600/R-95-131). [30] I. Vassilakis, D. Tsipi, M. Scoullos, Determination of a variety of chemical classes of pesticides in surface and ground waters by off-line solid-phase extraction, gas chromatography with electron-capture and nitrogen–phosphorus detection, and high-performance liquid chromatography with postcolumn derivatization and fluorescence detection, J. Chromatogr. A 823 (1998) 49–58. [31] L. Brossa, R.M. Marce, F. Borrull, E. Pocurull, Application of on-line solid-phase extraction-gas chromatography–mass spectrometry to the determination of endocrine disruptors in water samples, J. Chromatogr. A 963 (2002) 287–294. [32] L. Brossa, R.M. Marce, F. Borrull, E. Pocurull, Determination of endocrinedisrupting compounds in water samples by on-line solid-phase extractionprogrammed-temperature vaporisation–gas chromatography–mass spectrometry, J. Chromatogr. A 998 (2003) 41–50. [33] B. Hauser, M. Schellin, P. Popp, Membrane-assisted solvent extraction of triazines, organochlorine, and organophosphorus compounds in complex samples combined with large-volume injection-gas chromatography/mass spectrometric detection, Anal. Chem. 76 (2004) 6029–6038.
[34] A. Derouiche, M. Ridha Driss, J.-P. Morizur, M.-H. Taphanel, Simultaneous analysis of polychlorinated biphenyls and organochlorine pesticides in water by headspace solid-phase microextraction with gas chromatography–tandem mass spectrometry, J. Chromatogr. A 1138 (2007) 231–243. ˜ [35] A. Garrido Frenich, R. Romero-González, J.L. Martínez Vidal, R. Martínez Ocana, P. Baquero Feria, Comparison of solid phase microextraction and hollow fiber liquid phase microextraction for the determination of pesticides in aqueous samples by gas chromatography triple quadrupole tandem mass spectrometry, Anal. Bioanal. Chem. 399 (2011) 2043–2059. [36] C. Goncalves, M.F. Alpendurada, Multiresidue method for the simultaneous determination of four groups of pesticides in ground and drinking waters, using solid-phase microextraction-gas chromatography with electroncapture and thermionic specific detection, J. Chromatogr. A 968 (2002) 177–190. [37] E. Beceiro-Gonzalez, E. Concha-Grana, A. Guimaraes, C. Goncalves, S. Muniategui-Lorenzo, M.F. Alpendurada, Optimisation and validation of a solidphase microextraction method for simultaneous determination of different types of pesticides in water by gas chromatography–mass spectrometry, J. Chromatogr. A 1141 (2007) 165–173. [38] N. Ochiai, T. Ieda, K. Sasamoto, Y. Takazawa, S. Hashimoto, A. Fushimi, K. Tanabe, Stir bar sorptive extraction and comprehensive two-dimensional gas chromatography coupled to high-resolution time-of-flight mass spectrometry for ultra-trace analysis of organochlorine pesticides in river water, J. Chromatogr. A 1218 (2011) 6851–6860. [39] M.J. Gomez, S. Herrera, D. Sole, E. García-Calvo, A.R. Fernandez-Alba, Automatic searching and evaluation of priority and emerging contaminants in wastewater and river water by stir bar sorptive extraction followed by comprehensive two-dimensional gas chromatography-time-of-flight mass spectrometry, Anal. Chem. 83 (2011) 2638–2647. [40] P. Tölgyessy, B. Vrana, Z. Kras, Development of a screening method for the analysis of organic pollutants in water using dual stir bar sorptive extraction–thermal desorption-gas chromatography–mass spectrometry, Talanta 87 (2011) 152–160. [41] W.-C. Tsai, S.D. Huang, Dispersive liquid–liquid microextraction with little solvent consumption combined with gas chromatography–mass spectrometry for the pretreatment of organochlorine pesticides in aqueous samples, J. Chromatogr. A 1216 (2009) 5171–5175. [42] C. Cortada, L. Vidal, R. Pastor, N. Santiago, A. Canals, Determination of organochlorine pesticides in water samples by dispersive liquid-liquid microextraction coupled to gas chromatography-mass spectrometry, Anal. Chim. Acta 649 (2009) 218–221. [43] H. Farahani, Y. Yamini, S. Shariati, M.R. Khalili-Zanjani, S. Mansour-Baghahi, Development of liquid phase microextraction method based on solidification of floated organic drop for extraction and preconcentration of organochlorine pesticides in water samples, Anal. Chim. Acta 626 (2008) 166–173. [44] S.-P. Huang, P.-S. Chen, S.-D. Huang, Dynamic headspace time-extended helix liquid-phase microextraction, J. Chromatogr. A 1216 (2009) 4347–4353. [45] E. Concha-Grana, G. Fernández-Martínez, V. Fernández-Villarrenaga, M.I. Turnes-Carou, S. Muniategui-Lorenzo, P. López-Mahía, D. Prada-Rodríguez, A study of large-volume on-column injection GC-ECD for the ultratrace analysis of organochlorine pesticides in water, Talanta 78 (2009) 764–771. [46] S. Nakamura, T. Yamagami, S. Daishima, Determination of organochlorine pesticides in river water by gas chromatography-negative-ion chemicalionization mass spectrometry using large volume injection, Analyst 126 (2001) 1658–1662. [47] Visual PROMETHEE 1.4 Manual, www.promethee-gaia.net, 29.12.2014. [48] J.-P. Brans, B. Mareschal, Multiple criteria decision analysis: state of the art surveys, Int. Ser. Oper. Res. Manage. Sci. 78 (2005) 163–186. [49] A. Kucharski, Multicriteria ranking as the way of the evaluation of the Open Pension Funds competitiveness, Ann. UMCS 46 (2011) 324.
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Multicriteria decision analysis in ranking of analytical procedures for aldrin determination in water Marek Tobiszewski a,∗ , Aleksander Orłowski b a b
Department of Analytical Chemistry, Chemical Faculty, Gda´ nsk University of Technology, 11/12 G. Narutowicza St., 80-233 Gda´ nsk, Poland Department of Management, Faculty of Management and Economics, Gda´ nsk University of Technology, 11/12 G. Narutowicza St., 80-233 Gda´ nsk, Poland
a r t i c l e
i n f o
Article history: Received 10 January 2015 Received in revised form 2 February 2015 Accepted 3 February 2015 Available online 11 February 2015 Keywords: Green analytical chemistry Decision making Environmental impact assessment Analytical methodologies PROMETHEE Multi-criteria decision analysis
a b s t r a c t The study presents the possibility of multi-criteria decision analysis (MCDA) application when choosing analytical procedures with low environmental impact. A type of MCDA, Preference Ranking Organization Method for Enrichment Evaluations (PROMETHEE), was chosen as versatile tool that meets all the analytical chemists – decision makers requirements. Twenty five analytical procedures for aldrin determination in water samples (as an example) were selected as input alternatives to MCDA analysis. Nine different criteria describing the alternatives were chosen from different groups – metrological, economical and the most importantly – environmental impact. The weights for each criterion were obtained from questionnaires that were sent to experts, giving three different scenarios for MCDA results. The results of analysis show that PROMETHEE is very promising tool to choose the analytical procedure with respect to its greenness. The rankings for all three scenarios placed solid phase microextraction and liquid phase microextraction – based procedures high, while liquid–liquid extraction, solid phase extraction and stir bar sorptive extraction – based procedures were placed low in the ranking. The results show that although some of the experts do not intentionally choose green analytical chemistry procedures, their MCDA choice is in accordance with green chemistry principles. The PROMETHEE ranking results were compared with more widely accepted green analytical chemistry tools – NEMI and Eco-Scale. As PROMETHEE involved more different factors than NEMI, the assessment results were only weakly correlated. Oppositely, the results of Eco-Scale assessment were well-correlated as both methodologies involved similar criteria of assessment. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Green analytical chemistry is laboratory philosophy that recognizes the certain analytes determination process as an environmental problem. During analytical procedure, certain compounds can be emitted to the environment [1]. Application of green analytical chemistry means the reduction of procedural environmental impact [2]. The most common methods of reducing negative influence is the application of miniaturized extraction techniques at the stage of analytical sample preparation and miniaturization of analytical instruments [3,4]. The other approach is elimination of sample preparation step by application of direct analytical techniques, which in fact can be applied rather seldom [5,6]. These solutions in greening the analytical procedures are relatively well established and have been reviewed exhaustively [7–10].
∗ Corresponding author at. Tel.: +48583472194. E-mail address: [email protected] (M. Tobiszewski). http://dx.doi.org/10.1016/j.chroma.2015.02.009 0021-9673/© 2015 Elsevier B.V. All rights reserved.
One of the main problems in the field of green analytical chemistry is the assessment of analytical procedures greenness. There are several still very vital questions. How to measure the procedural greenness? How to compare the environmental impact of methodologies to choose the most environmentally friendly one? Can an universal procedure be developed to help or guide in procedure selection for the given purpose? There are few assessment methods available, each of them has its own advantages and disadvantages. The only assessment procedure, that is brought into a broader laboratory practice, is labeling with NEMI symbols [11]. NEMI symbol is easy to read circle with its fields filled green (or not) if certain assumptions are fulfilled. These are generation of wastes in the amount lesser than 50 g, no corrosive environment is present and there are no hazardous or dangerous chemicals used during analytical procedure. The main drawback in NEMI application is the need to search the hazardous substances lists to check if any of chemicals used in the procedure is present on these lists, which is time-consuming task. The other NEMI drawback is that the information carried by the NEMI symbol is not quantitative. This problem is partially solved when NEMI modification
M. Tobiszewski, A. Orłowski / J. Chromatogr. A 1387 (2015) 116–122
Stating the decision problem Choosing the procedure for aldrin in water determination Selection of alternatives Searching for adequate procedures Selection of criteria Metrological, economical, environmental criteria Weights given by experts As a result of answers to questionnaire
Weighting the criteria Setting the relative importance of each criterion
117
when concerning PROMETHEE as the tool in assessment of green analytical chemistry metrics. It seems to be important to mention that there were several rankings of MCDA methods published [20] and PROMETHEE is concerned to be one of the top in alternatives ranking, easiness of use and potential of future development. The aim of the study is to investigate the possibility of application of multicriteria decision analysis (PROMETHEE) as the analytical procedure assessment tool in the light green analytical chemistry. The second aim is to show that environmental impact aspects should be considered when making decision of choosing the analytical procedure for the given purpose. Some remarks on choosing the best procedure will be also provided.
PROMETHEE analysis With ranking of analytical procedures as the result Choosing the best alternative The first from the ranking Fig. 1. The scheme of analytical methodologies assessment procedure with multicriteria decision PROMETHEE technique.
proposed by de la Guardia and Armenta is applied [12]. In such a case NEMI symbol carries the semi-quantitative information about threat in each of its categories. The other approach in assessment of analytical procedures greenness is application of Eco-Scale [13]. The Eco-Scale involves many factors that can have negative impact on the environment such as the amount and the type of any chemical used in the procedure, the mass of generated waste, the occupational exposure and energy consumption. The result of Eco-Scale application is the number lower than one hundred, the closer the values to one hundred the more green analysis. The main disadvantage of application of Eco-Scale is the fact that the result is the number which in fact involves many parameters but it does not carry any information about the nature of threats related to the analytical procedure. The latest approach, that gives the possibility to classify the analytical procedures taking into consideration their environmental impact, is the application of multivariate statistical techniques [14]. The techniques such as self-organizing maps allow to group the analytical procedures according to their similarity when environmental, metrological and technical parameters, describing the analytical procedure, are considered. The drawback of multivariate statistics application is tedious assessment procedure and that it might be applied to compare limited amount of procedures [15]. It is clear that green analytical chemistry still needs the reliable procedural assessment methods. Multicriteria decision analysis is used in environmental science and environmental management [16]. It has been successfully applied to select proper remediation procedure for contaminated mine treatment [17], method of water disinfection [18] or choosing priority areas for environmental monitoring [19]. All these examples involve social, environmental, economical, technological criteria during the decision process. The Preference Ranking Organization Method for Enrichment Evaluations (PROMETHEE) methods are chosen from available multicriteria decision analysis methods because of two reasons. First of all they are easy to apply compared to other methods (such as Elimination and Choice Expressing Reality – ELECTRE or Analytic Hierarchy Process – AHP) because PROMETHEE methods require fewer parameters from decision makers. They also rank alternatives as well as identify the best alternative, whereas the previously mentioned utility theory methods only identify the best alternatives [17]. Ranking alternatives seems to be the most important because there are 25 alternatives taken into consideration and it is not only important to know which is the best among all but also to make it possible to compare all the alternatives. It is important aspect
2. Materials and methods The general scheme of the presented multicriteria decision analysis assessment approach is shown in Fig. 1. Briefly, after stating the decision problem, the set of alternatives (the possible options) is identified. Then the important aspects – criteria, have to be defined, together with their relative importance by assigning weights to the criteria.
2.1. The PROMETHEE method The PROMETHEE (Preference Ranking Organization Method for Enrichment Evaluations) is one of the leading MCDA (Multi-Criteria Decision Aid or Analysis) methods that was developed by Brans and Vincke in 1985 [21]. The methodology is successfully used in various areas such as business, public administration, industrial location, medicine or tourism [22]. PROMETHEE is a family of outranking methods that consist of PROMETHEE I which generates partial ranking, PROMETHEE II for complete ranking of alternatives, PROMETHEE III for ranking based on interval, PROMETHEE IV for ranking (partial or complete) of alternatives set of viable solutions is continuous, PROMETHEE V for problem with segmentation constraints, PROMETHEE VI for human brain representation. There are also additional methods: PROMETHEE GDSS for group-decision making, PROMETHEE TRI used when sorting problems appear and PROMETHEE CLUSTER for nominal classification [22]. In the presented paper PROMETHEE II is going to be applied, because the ranking of many alternatives is the most desired analysis output from our point of view. The principle of PROMETHEE II is based on pairwise comparison of alternatives along each criterion. To implement the methodology it is required to have two types of information: weights of the criteria and the preference function. In order to obtain information within the criteria, a preference function for each criterion, expressing the difference in performance of alternative a over alternative b must be identified, adopting as a result the pairwise comparison approach [20]. It is also requested to have clear and understandable (by both decision-maker and analysts) information between the criteria and information within each criterion. When A of n alternatives (a1 , a2 , . . ., an ) have to be ranked and G of k criteria (g1 , g2 , . . ., gk ) have to be maximized (or minimized) the resulting problem has the form as shown [17]: max{g1 (a), g2 (a), . . ., gk (a)|a ∈ A} Some criteria might be maximized whereas others might be minimized, the goal of the decision-maker is to identify an alternative which is optimizing all criteria. The detailed mathematical description of PROMETHEE was presented by its developers Brans and Vincke in [21].
118
M. Tobiszewski, A. Orłowski / J. Chromatogr. A 1387 (2015) 116–122
Table 1 The criteria being input data to the multicriteria decision analysis. No.
Criterion
Description
Preference function
1
Variable with linguistic description: 0 – hardly available; 0.25 – poorly available; 0.5 – moderately available; 0.75 – rather available; 1 – easily available. The lowest concentration of analyte that can be detected. Expressed as coefficient of variance. The percentage of analyte that is recovered with a procedure.
The higher the better Range 0–1
2 3 4
Availability of analytical equipment in the laboratory needed for performing the analytical procedure Limits of detection Precision of the methodology Recovery of analytes
5
The number of other analytes
6 7 8
Time of analysis Amount of organic solvent The hazards and toxicity of the solvents and reagents applied in the procedure The amount of solid wastes
The number of other analytes apart from aldrin that can be determined in the single run. The total time to prepare and analyze the analytical sample. Total amount of organic solvents that are used in the procedure. The sum of all hazards connected application of given solvent times its amount. The total amount of solid wastes that are generated in the analytical run.
9
2.2. The alternatives All the analytical procedures are dedicated to determine aldrin in water samples. The data are collected by searching the scientific databases. The required information (listed in Table 1) was extracted from articles or standard procedures. The databases that were searched are ACS, RSC, ScienceDirect, Springerlink, Wiley. If any of the information was not available in the source article such procedure was excluded from further consideration. Most of the criteria are assessed with the values that are directly taken from the literature data. The only criterion that has linguistic description is the availability of the laboratory equipment. The translation of this criterion from linguistic description to numerical value is presented in Table 1. The availability estimation for the analytical equipment is arbitrary and was done by ourselves. All the preference functions for all the criteria seem to be obvious and easy to be implemented in PROMETHEE. Only the recovery value criterion is non-standard one. We have taken the absolute value of the difference between the analytical procedure recovery value and 100%. In this sense the recovery values of 95% and 105% are equally good. The analytical procedures included in the analysis are based on various sample preparation techniques but all of them are based on gas chromatographic separation. The alternatives for multicriteria decision analysis are presented in Table 2. As an assumption at the data preprocessing step, all methodologies with LOD > 10 ng L−1 were excluded from further consideration, as they do not meet the requirement set by Maximum Allowable Concentration for aldrin in water samples.
The lower the better The lower the better The closer to 100% the better The higher the better The lower the better The lower the better The lower the better The lower the better
Table 2 The alternatives considered during multicriteria decision making. Alternative No.
Methodology abbreviation
Reference
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
6630 – LLE–GC–ECD 505 LLE–GC–ECD O-1104 LLE–GC–ECD SPE–GC–GC–TOF–MS SPE–GC–MS SPE–GC–MS 508.1 SPE–GC–ECD SPE–GC–ECD On-line-SPE–GC–MS On-line-SPE–PTV–GC–MS MASE–LVI–GC–MS HS–SPME–GC–MS–MS SPME–GC–MS–MS SPME–GC–ECD SPME–GC–MS SBSE–TD–GC–GC–MS SBSE–GC–GC–TOF–MS SBSE–TD–GC–MS DLLME–GC–MS DLLME–GC–MS LPME–GC–ECD HF–LPME–GC–MS–MS DHS–TEH–LPME–GC–MS–MS MLLE–LVOCI–GC–ECD MLLE–LVI–NICI–MS
[23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [35] [44] [45] [46]
DHS – dynamic headspace; DLLME – dispersive liquid–liquid microextraction; ECD – electron capture detection; GC – gas chromatography; HF – hollow fiber; HS – headspace; LLE – liquid–liquid extaction; LPME – liquid phase microextraction; MASE – membrane-assisted solvent extraction; MLLE – micro liquid–liquid extraction; MS – mass spectrometry; PTV – programmed-temperature vaporization; SBSE – stir-bar sorptive extraction, SPE – solid phase extraction; SPME – solid-phase microextraction; TD – thermal desorption; THE – time-extended helix liquid-phase microextraction; TOF – time-of-flight.
2.3. The questionnaire The questionnaire was created to establish weights for each criterion. The weight of a criterion is a positive number that represents the criterion relative importance [47]. The higher the weight, the more important is the criterion [48]. The questionnaire was designed in such a way that the importance of each criterion (listed in Table 1) could be assessed verbally in seven – point scale from “very low” trough “low”, “medium low” “medium”, “medium high”, “high” to “very high”. Such an approach allows to easily fill the questionnaire by the expert. Additionally, the questionnaire contained the question about expert’s attitude towards green analytical chemistry with three possible answers – (a) I always choose green analytical methodologies, (b) greenness of analytical methodology is minor aspect, I choose greener methodology only if it is cheaper, faster and has beneficial metrological parameters, (c) I do not consider the environmental impact of the analytical procedure. The question was important for the assessment of
PROMETHEE as the tool in green analytical chemistry metrics and gives information about possible scenarios. Such questionnaires were sent to experts who were asked to assess the importance of each criterion when choosing analytical procedure. 3. Results and discussion 3.1. Assigning weights for the criteria To perform the PROMETHEE analysis it was necessary to define the importance of criteria which are going to be included in analysis. There are nine criteria (Table 1) which are taken into consideration in the process of selecting the best alternative and ranking the remaining ones. Assigning weights to criteria involves priorities of the decision-maker because not all criteria have equal importance. Due to this fact it was necessary to assign the
M. Tobiszewski, A. Orłowski / J. Chromatogr. A 1387 (2015) 116–122
119
Table 3 The results of the questionnaires together with assigned weights. Criterion
1 2 3 4 5 6 7 8 9 Total
Questionnaire 1
Questionnaire 2
Questionnaire 3
Mean of questionnaires
Verbal description
Points
Weight
Verbal description
Points
Weight
Verbal description
Points
Weight
Verbal description
Points
Weight
High High High Medium Medium low Medium Very high Very high High
6 6 6 4 3 4 7 7 6 49
0.122 0.122 0.122 0.081 0.061 0.081 0.142 0.142 0.122 1
Very high High High Very high High Medium high Medium Medium Medium
7 6 6 7 6 5 4 4 4 49
0.142 0.122 0.122 0.142 0.122 0.102 0.081 0.081 0.081 1
Very high High Very high High High High Low Very low Medium high
7 6 7 6 6 6 2 1 5 46
0.152 0.13 0.152 0.13 0.13 0.13 0.043 0.021 0.108 1
Very high High High High Medium high Medium high Medium Medium Medium high
7 6 6 6 5 5 4 4 5 48
0.146 0.125 0.125 0.125 0.104 0.104 0.083 0.083 0.104 1
numerical value to the importance of each criteria. Several methods of assigning weights were considered by authors among them making decision independently basing on authors’ experience or using experts’ knowledge. Having in mind the importance of the decision (influence on the final results) we decided to use experts’ knowledge. It is assumed that experts have wide knowledge and they are objective in their jurisdictions–they were not informed about the details of the case but were just asked to assign importance to the criteria. The experts are analytical chemists, who have great experience with various analytical procedures. We have tried to select experts with different approach to the green analytical chemistry philosophy. Because there was just a need to evaluate the criteria, typical interview with experts seemed not to be necessary. It was decided to prepare a dedicated questionnaire (Section 2.3) that was sent to the experts. Designing the questionnaire we have considered the manual assigning of numerical values of weights by experts. This would mean that experts have to give fraction numbers to each criteria and they should sum up to 1. Such an approach gives confusing questionnaire to be filled by experts and there is a possibility that one of the criterion obtains weight considerably higher than remaining ones, what brings multicriteria analysis to virtually single-criterion analysis. We wanted to avoid such situation, however such an approach is also possible. We have decided for easier, for experts, approach – giving descriptive answers to describe criteria importance. Next, it was necessary to transform the experts’ answers to the weights to be used in the analysis. In PROMETHEE method (other than in e.g. AHP) there is no precisely described way of setting weighs of criteria [49]. We considered two methods of setting weights, for both there is no objection to consider normalized weights (sum = 1). Constant value difference between weights – this method presents constant degree between each of the weights representation. The method was implemented (data not published) but has an important disadvantage: it does not take into consideration the expert’s tendencies to either very high or very low evaluation of criteria. ‘In the second method to each possible answer there are numerical values (points) assigned, from 1 assigned to “very low” to 7 for “very high”. Basing on the answers given by the expert the total sum of values of answers is generated. Next the value of each answer is divided by the total sum of points and as the result the weights are generated. Next there was a need to decide how the weights are going to be interpreted. It is possible to generate the average answer for all three experts or to prepare PROMETHEE analysis for individual weights by each expert and threat the results as scenario. Both methods were tried, the results for the scenarios and mean answers for the scenarios are presented in Table 3. Referring to the selection of experts, none of them was seen as more important so the importance of their answers was equal. Basing on the answers received from experts, it was possible to use an
average importance (based on arithmetic mean) for each criterion (Table 3). It brings the opportunity to present the knowledge from wider perspective (three different perspectives from three experts) and to compare results gained from different experts (different points of view). The expert filling the questionnaire 1 indicated that he always chooses the analytical procedures according to green analytical chemistry principles. The remaining two experts admitted that they choose green analytical chemistry procedures when other decision criteria are met. 3.2. The ranking with PROMETHEE Table 4 presents the results of analytical methodologies for aldrin determination ranking with PROMETHEE method. The results for the respective scenarios do not differ significantly, the rankings are different in details only. Generally, the first 10 analytical procedures in rankings are based on SPME or various modes of liquid–liquid microextractions, with either mass spectrometric of electron capture detection. These techniques are characterized by good analytical parameters, minute amounts of solvents consumption and were assessed as requiring rather available equipment. In the second part of ranking procedures based on SBSE, SPE and LLE sample preparation and procedures involving more sophisticated apparatus like TOF–MS or MS–MS detectors are placed. These kind of detectors in these study were assessed as less available than ECD or MS. The procedures involving SPE or LLE require considerable amounts of solvents that are often toxic and usually produce solid waste. The PROMETHEE algorithm allowed to rank the analytical procedures according to their greenness but also taking into account other factors. The results of ranking are in agreement with widely accepted reception of analytical procedures greenness. What is worth mentioning the standard procedures were ranked very low. This suggests that other alternatives, due to their attractive features, would be preferentially chosen by analysts over these standard procedures. None of the experts indicated that the environmental impact of the procedure, when making selection decision, is not considered. The results of the ranking, for expert who admitted that he always chooses green analytical chemistry procedures, are in close agreement to those for experts that choose green procedures when other factors are beneficial. This shows that analyst can choose green analytical procedures without sacrificing other analytical procedure parameters, such as metrological ones. The second conclusion is that the decision makers who choose analytical procedures, taking into account other than environmental factors, unwittingly choose procedures that meet green analytical chemistry standards. The application of PROMETHEE as green analytical chemistry assessment tool has the advantage of possibility of including other parameters. In this study, apart from metrological and environmental ones, we have included the “availability of equipment” criterion.
120
M. Tobiszewski, A. Orłowski / J. Chromatogr. A 1387 (2015) 116–122
Table 4 The results of PROMETHEE ranking of analytical procedures for different scenarios, based on questionnaires answers. As the abbreviations for some procedures are doubled, the alternatives are presented with numbers from Table 2. Rank
Scenario 1
Scenario 2
Scenario 3
Mean of scenarios
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
14. SPME–GC–ECD 24. MLLE–LVOCI–GC–ECD 19. DLLME–GC–MS 15. SPME–GC–MS 12. HS–SPME–GC–MS–MS 25. MLLE–LVI–NICI–MS 20. DLLME–GC–MS 23. DHS–TEH–LPME–GC–MS–MS 13. SPME–GC–MS–MS 21. LPME–GC–ECD 10. On-line-SPE–PTV–GC–MS 2. 505 LLE–GC–ECD 16. SBSE–TD–GC–GC–MS 9. On-line-SPE–GC–MS 18. SBSE–TD–GC–MS 8. SPE–GC–ECD 22. HF–LPME–GC–MS–MS 6. SPE–GC–MS 4. SPE–GC–GC–TOF–MS 11. MASE–LVI–GC–MS 17. SBSE–GC–GC–TOF–MS 5. SPE–GC–MS 7. 508.1 SPE–GC–ECD 1. 6630 LLE–GC–ECD 3. O-1104 LLE–GC–ECD
14. SPME–GC–ECD 24. MLLE–LVOCI–GC–ECD 19. DLLME–GC–MS 25. MLLE–LVI–NICI–MS 15. SPME–GC–MS 20. DLLME–GC–MS 12. HS–SPME–GC–MS–MS 13. SPME–GC–MS–MS 23. DHS–TEH–LPME–GC–MS–MS 2. 505 LLE–GC–ECD 21. LPME–GC–ECD 10. On-line-SPE–PTV–GC–MS 18. SBSE–TD–GC–MS 8. SPE–GC–ECD 22. HF–LPME–GC–MS–MS 11. MASE–LVI–GC–MS 16. SBSE–TD–GC–GC–MS 6. SPE–GC–MS 9. On-line-SPE–GC–MS 4. SPE–GC–GC–TOF–MS 5. SPE–GC–MS 1. 6630 LLE–GC–ECD 7. 508.1 SPE–GC–ECD 17. SBSE–GC–GC–TOF–MS 3. O-1104 LLE–GC–ECD
24. MLLE–LVOCI–GC–ECD 14. SPME–GC–ECD 19. DLLME–GC–MS 25. MLLE–LVI–NICI–MS 20. DLLME–GC–MS 15. SPME–GC–MS 12. HS–SPME–GC–MS–MS 23. DHS–TEH–LPME–GC–MS–MS 13. SPME–GC–MS–MS 10. On-line-SPE–PTV–GC–MS 21. LPME–GC–ECD 2. 505 LLE–GC–ECD 18. SBSE–TD–GC–MS 8. SPE–GC–ECD 11. MASE–LVI–GC–MS 4. SPE–GC–GC–TOF–MS 6. SPE–GC–MS 16. SBSE–TD–GC–GC–MS 9. On-line-SPE–GC–MS 22. HF–LPME–GC–MS–MS 5. SPE–GC–MS 1. 6630 LLE–GC–ECD 7. 508.1 SPE–GC–ECD 3. O-1104 LLE–GC–ECD 17. SBSE–GC–GC–TOF–MS
14. SPME–GC–ECD 24. MLLE–LVOCI–GC–ECD 19. DLLME–GC–MS 25. MLLE–LVI–NICI–MS 15. SPME–GC–MS 20. DLLME–GC–MS 12. HS–SPME–GC–MS–MS 13. SPME–GC–MS–MS 23. DHS–TEH–LPME–GC–MS–MS 21. LPME–GC–ECD 2. 505 LLE–GC–ECD 10. On-line-SPE–PTV–GC–MS 18. SBSE–TD–GC–MS 8. SPE–GC–ECD 16. SBSE–TD–GC–GC–MS 9. On-line-SPE–GC–MS 22. HF–LPME–GC–MS–MS 11. MASE–LVI–GC–MS 6. SPE–GC–MS 4. SPE–GC–GC–TOF–MS 5. SPE–GC–MS 1. 6630 LLE–GC–ECD 7. 508.1 SPE–GC–ECD 17. SBSE–GC–GC–TOF–MS 3. O-1104 LLE–GC–ECD
The other criteria that can be included might be: cost of reagents per sample, apparatus cost, the ability of laboratory staff to perform given analysis (by Bernoulli or fuzzy description) or sample analysis throughput. In more advanced multicriteria decision analysis risk assessment or reagents delivery chain can be included. The criteria can be described by discrete or fuzzy numbers, “yes” or “no” answers or verbal description can be applied and translated into numbers, what makes the MCDA procedure more convenient to use. The PROMETHEE V, another type of PROMETHEE methodology, enables user to define constraints which is used in selecting a best subset of action. Such constraint might be for example maximum budget defined for the project which seems to be important in the process of selecting analytic procedure for every individual decision-maker. As general, in PROMETHEE it is fast and easy to change the weights which practically will be important in adapting the process of choosing analytical procedure to individual case. The individual user can also change his preferences according to his needs by modifying weights. For example, when large amount of samples has to be analyzed in short time, the analysis time becomes the most important factor and high weight is assigned to this criterion. PROMETHEE enables also the possibility of visualization of the results. The PROMETHEE I enables to assess quantitatively, how each criterion of the alternative influences the score. 3.3. The comparison of the results with other assessment methodologies The most widely accepted green analytical chemistry metrics tool is NEMI labeling. Fig. 2a shows the correlation of NEMI score with PROMETHEE ranking. There is virtually no correlation between the results with these two assessment tools. NEMI symbols take into consideration 4 parameters only in Bernoulli distribution – each criterion is met or not. PROMETHEE ranking, in this case, takes into consideration 9 criteria and the numerical values for each criteria can be any number from a given range. This results in fact that the NEMI results have considerably lower “resolution” as
they can be natural numbers from 0 to 4. The second reason for such lack of correlation is the fact that during PROMETHEE ranking other than environmental factors, such as “availability of equipment” or “analysis time”, have been considered. The second tool in analytical procedures environmental impact assessment that gains in interest is Eco-scale. The correlation of Eco-scale with PROMETHEE results is more legible than in case of NEMI labeling results (Fig. 2b). Eco-scale in its procedure involves more parameters than NEMI and is more similar to our case of PROMETHEE ranking. Eco-scale quantitatively considers the amount and the toxicity of each reagent consumed in analytical procedure. The analytical procedure (alternative in MCDA nomenclature), that is the mostly outlying point, is the procedure based on SBSEGC × GC-TOF-MS. It is ranked at 24 position, while its NEMI score is 4 and Eco-scale score is 72. This procedure is based on extraction with SBSE followed by thermal desorption and analysis with two dimensional gas chromatography and time-of-flight mass spectrometer. The procedure involves small amount of reagents, so NEMI and Eco-scale scores are high. On the other hand sophisticated equipment needed, relatively poor metrological merits result in such a low PROMETHEE ranking. In contrast to NEMI and Eco-scale, PROMETHEE has the advantage that it allows to choose the greenest option among alternatives (at least two analytical procedures) by direct comparison. It gives possibility to include in the assessment procedure other criteria than those included in the NEMI and Eco-scale techniques. Some users could find the application of weights for the criteria confusing, as they may change significantly when different analytes and matrices are considered. We see user-defined criteria weights as advantage, because such approach makes the PROMETHEEbased assessment procedure very specific. User can modify the assessment procedure to his needs by choosing and weighting criteria. Concerning PROMETHEE as standard procedure in greenness assessment would require defined approach to choosing criteria and assigning weights. PROMETHEE as assessment procedure should be considered as rather tedious but its result is very easy to be interpreted.
M. Tobiszewski, A. Orłowski / J. Chromatogr. A 1387 (2015) 116–122
121
Fig. 2. NEMI (a) and Eco-scale (b) results correlation to PROMETHEE mean ranking.
4. Conclusions
References
The PROMETHEE procedure applied in selection of the most appropriate analytical methodologies seems to be useful tool. It allows user to select the best solution in proper way concerning various, often contradictory, criteria. If some of the criteria are describing environmental impact of analytical procedure, then the PROMETHEE becomes very useful tool in assessment of green analytical chemistry metrics. Although some of the presented scenarios did not consider environmental criteria as important, the results of PROMETHEE ranking allowed to choose the analytical procedures with low environmental impact. The presented case shown that although the given weights are slightly different for every expert the ranking results are very similar. The analysts who claim not to follow green analytical chemistry path, choose the analytical procedures with low environmental impact. PROMETHEE as green analytical chemistry tool has the key advantage: it includes user or decision maker personal preferences which allows to select criteria and their weights. It has the great potential to include other criteria important for every decision maker. It also has the potential to be combined with other tools such as life-cycle assessment.
[1] S. Armenta, S. Garrigues, M. de la Guardia, Green analytical chemistry, TrAC Trends Anal. Chem. 27 (2008) 497–511. ´ [2] M. Tobiszewski, A. Mechlinska, J. Namie´snik, Green analytical chemistry—theory and practice, Chem. Soc. Rev. 39 (2010) 2869–2878. [3] A. Spietelun, Ł. Marcinkowski, M. de la Guardia, J. Namie´snik, Recent developments and future trends in solid phase microextraction techniques towards green analytical chemistry, J. Chromatogr. A 1321 (2013) 1–13. [4] A. Spietelun, Ł. Marcinkowski, M. de la Guardia, J. Namie´snik, Green aspects, developments and perspectives of liquid phase microextraction techniques, Talanta 119 (2014) 34–45. [5] A. Mir-Marqués, S. Garrigues, M.L. Cervera, M. de la Guardia, Direct determination of minerals in human diets by infrared spectroscopy and X-ray fluorescence, Microchem. J. 117 (2014) 156–163. [6] M. Tobiszewski, J. Namie´snik, Direct chromatographic methods in the context of green analytical chemistry, TrAC Trends Anal. Chem. 35 (2012) 67–73. [7] H. Shaaban, T. Górecki, Current trends in green liquid chromatography for the analysis of pharmaceutically active compounds in the environmental water compartments, Talanta 132 (2015) 739–752. [8] M. Kaljurand, M. Koel, Recent advancements on greening analytical separation, Crit. Rev. Anal. Chem. 41 (2011) 2–20. [9] M. Koel, M. Kaljurand, Application of the principles of green chemistry in analytical chemistry, Pure Appl. Chem. 78 (2006) 1993–2002. [10] M. Farré, S. Pérez, C. Gonc¸alves, M.F. Alpendurada, D. Barceló, Green analytical chemistry in the determination of organic pollutants in the aquatic environment, TrAC Trends Anal. Chem. 29 (2010) 1347–1362. [11] L.H. Keith, L.U. Gron, J.L. Young, Green analytical methodologies, Chem. Rev. 107 (2007) 2695–2708. [12] S. Garrigues, M. de la Guardia, Handbook of Green Analytical Chemistry, Wiley, 2012. [13] A. Gałuszka, Z.M. Migaszewski, P. Konieczka, J. Namie´snik, Analytical Eco-Scale for assessing the greenness of analytical procedures, TrAC Trends Anal. Chem. 37 (2012) 61–72. [14] M. Tobiszewski, S. Tsakovski, V. Simeonov, J. Namie´snik, Application of multivariate statistics in assessment of green analytical chemistry parameters of analytical methodologies, Green Chem. 15 (2013) 1615–1623.
Acknowledgement The authors would like to thank the experts for their help by filling the questionnaires.
122
M. Tobiszewski, A. Orłowski / J. Chromatogr. A 1387 (2015) 116–122
[15] M. Tobiszewski, S. Tsakovski, V. Simeonov, J. Namie´snik, Multivariate statistical comparison of analytical procedures for benzene and phenol determination with respect to their environmental impact, Talanta 130 (2014) 449–455. [16] I.B. Huang, J. Keisler, I. Linkov, Multi-criteria decision analysis in environmental sciences: ten years of applications and trends, Sci. Total Environ. 409 (2011) 3578–3594. [17] G.D. Betrie, R. Sadiq, K.A. Morin, S. Tesfamariam, Selection of remedial alternatives for mine sites: a multicriteria decision analysis approach, J. Environ. Manage. 119 (2013) 36–46. [18] M.D. Gomez-Lopez, J. Bayo, M.S. Garcıa-Cascales, J.M. Angosto, Decision support in disinfection technologies for treated wastewater reuse, J. Clean. Prod. 17 (2009) 1504–1511. [19] M. Alvarez-Guerra, J.R. Viguri, N. Voulvoulis, A multicriteria-based methodology for site prioritisation in sediment management, Environ. Int. 35 (2009) 920–930. [20] M. Cinelli, S.R. Coles, K. Kirwan, Analysis of the potentials of multi criteria decision analysis methods to conduct sustainability assessment, Ecol. Indicators 46 (2014) 138–148. [21] J.P. Brans, P. Vincke, A preference ranking organization method: the PROMETHEE method for multiple criteria decision-making, Manage. Sci. 31 (1985) 647–656. [22] M. Behzadian, R.B. Kazemzadeh, A. Albadvi, M. Agdasi, PROMETHEE: a comprehensive literature review on methodologies and applications, Eur. J. Oper. Res. 200 (2010) 198–215. [23] 6630C Standard Methods Online – Standard Methods for the Examination of Water and Wastewater. [24] 505 Methods for the Determination of Organic Compounds in Drinking Water – Supplement III (EPA/600/R-95-131). [25] O-1104 Methods for the Determination of Organic Substances in Water and Fluvial Sediments, Techniques of Water-Resources Investigations of the United States Geological Survey, Book 5, Chapter A3 Edited by R.L. Wershaw, M.J. Fishman, R.R. Grabbe, and L.E. Lowe. [26] V. Matamoros, E. Jover, J.M. Bayona, Part-per-trillion determination of pharmaceuticals, pesticides, and related organic contaminants in river water by solid-phase extraction followed by comprehensive two-dimensional gas chromatography time-of-flight mass spectrometry, Anal. Chem. 82 (2010) 699–706. [27] C.C. Leandro, D.A. Bishop, R.J. Fussell, F.D. Smith, B.J. Keely, Semiautomated determination of pesticides in water using solid phase extraction disks and gas chromatography–mass spectrometry, J. Agric. Food Chem. 54 (2006) 645–649. [28] S. Barreka, C. Cren-Olivéa, L. Wiest, R. Baudot, C. Arnaudguilhema, Marie-Florence Grenier-Loustalot Multi-residue analysis and ultra-trace quantification of 36 priority substances from the EuropeanWater Framework Directive by GC–MS and LC-FLD-MS/MS in surface waters, Talanta 79 (2009) 712–722. [29] 508.1 Methods for the Determination of Organic Compounds in Drinking Water – Supplement III (EPA/600/R-95-131). [30] I. Vassilakis, D. Tsipi, M. Scoullos, Determination of a variety of chemical classes of pesticides in surface and ground waters by off-line solid-phase extraction, gas chromatography with electron-capture and nitrogen–phosphorus detection, and high-performance liquid chromatography with postcolumn derivatization and fluorescence detection, J. Chromatogr. A 823 (1998) 49–58. [31] L. Brossa, R.M. Marce, F. Borrull, E. Pocurull, Application of on-line solid-phase extraction-gas chromatography–mass spectrometry to the determination of endocrine disruptors in water samples, J. Chromatogr. A 963 (2002) 287–294. [32] L. Brossa, R.M. Marce, F. Borrull, E. Pocurull, Determination of endocrinedisrupting compounds in water samples by on-line solid-phase extractionprogrammed-temperature vaporisation–gas chromatography–mass spectrometry, J. Chromatogr. A 998 (2003) 41–50. [33] B. Hauser, M. Schellin, P. Popp, Membrane-assisted solvent extraction of triazines, organochlorine, and organophosphorus compounds in complex samples combined with large-volume injection-gas chromatography/mass spectrometric detection, Anal. Chem. 76 (2004) 6029–6038.
[34] A. Derouiche, M. Ridha Driss, J.-P. Morizur, M.-H. Taphanel, Simultaneous analysis of polychlorinated biphenyls and organochlorine pesticides in water by headspace solid-phase microextraction with gas chromatography–tandem mass spectrometry, J. Chromatogr. A 1138 (2007) 231–243. ˜ [35] A. Garrido Frenich, R. Romero-González, J.L. Martínez Vidal, R. Martínez Ocana, P. Baquero Feria, Comparison of solid phase microextraction and hollow fiber liquid phase microextraction for the determination of pesticides in aqueous samples by gas chromatography triple quadrupole tandem mass spectrometry, Anal. Bioanal. Chem. 399 (2011) 2043–2059. [36] C. Goncalves, M.F. Alpendurada, Multiresidue method for the simultaneous determination of four groups of pesticides in ground and drinking waters, using solid-phase microextraction-gas chromatography with electroncapture and thermionic specific detection, J. Chromatogr. A 968 (2002) 177–190. [37] E. Beceiro-Gonzalez, E. Concha-Grana, A. Guimaraes, C. Goncalves, S. Muniategui-Lorenzo, M.F. Alpendurada, Optimisation and validation of a solidphase microextraction method for simultaneous determination of different types of pesticides in water by gas chromatography–mass spectrometry, J. Chromatogr. A 1141 (2007) 165–173. [38] N. Ochiai, T. Ieda, K. Sasamoto, Y. Takazawa, S. Hashimoto, A. Fushimi, K. Tanabe, Stir bar sorptive extraction and comprehensive two-dimensional gas chromatography coupled to high-resolution time-of-flight mass spectrometry for ultra-trace analysis of organochlorine pesticides in river water, J. Chromatogr. A 1218 (2011) 6851–6860. [39] M.J. Gomez, S. Herrera, D. Sole, E. García-Calvo, A.R. Fernandez-Alba, Automatic searching and evaluation of priority and emerging contaminants in wastewater and river water by stir bar sorptive extraction followed by comprehensive two-dimensional gas chromatography-time-of-flight mass spectrometry, Anal. Chem. 83 (2011) 2638–2647. [40] P. Tölgyessy, B. Vrana, Z. Kras, Development of a screening method for the analysis of organic pollutants in water using dual stir bar sorptive extraction–thermal desorption-gas chromatography–mass spectrometry, Talanta 87 (2011) 152–160. [41] W.-C. Tsai, S.D. Huang, Dispersive liquid–liquid microextraction with little solvent consumption combined with gas chromatography–mass spectrometry for the pretreatment of organochlorine pesticides in aqueous samples, J. Chromatogr. A 1216 (2009) 5171–5175. [42] C. Cortada, L. Vidal, R. Pastor, N. Santiago, A. Canals, Determination of organochlorine pesticides in water samples by dispersive liquid-liquid microextraction coupled to gas chromatography-mass spectrometry, Anal. Chim. Acta 649 (2009) 218–221. [43] H. Farahani, Y. Yamini, S. Shariati, M.R. Khalili-Zanjani, S. Mansour-Baghahi, Development of liquid phase microextraction method based on solidification of floated organic drop for extraction and preconcentration of organochlorine pesticides in water samples, Anal. Chim. Acta 626 (2008) 166–173. [44] S.-P. Huang, P.-S. Chen, S.-D. Huang, Dynamic headspace time-extended helix liquid-phase microextraction, J. Chromatogr. A 1216 (2009) 4347–4353. [45] E. Concha-Grana, G. Fernández-Martínez, V. Fernández-Villarrenaga, M.I. Turnes-Carou, S. Muniategui-Lorenzo, P. López-Mahía, D. Prada-Rodríguez, A study of large-volume on-column injection GC-ECD for the ultratrace analysis of organochlorine pesticides in water, Talanta 78 (2009) 764–771. [46] S. Nakamura, T. Yamagami, S. Daishima, Determination of organochlorine pesticides in river water by gas chromatography-negative-ion chemicalionization mass spectrometry using large volume injection, Analyst 126 (2001) 1658–1662. [47] Visual PROMETHEE 1.4 Manual, www.promethee-gaia.net, 29.12.2014. [48] J.-P. Brans, B. Mareschal, Multiple criteria decision analysis: state of the art surveys, Int. Ser. Oper. Res. Manage. Sci. 78 (2005) 163–186. [49] A. Kucharski, Multicriteria ranking as the way of the evaluation of the Open Pension Funds competitiveness, Ann. UMCS 46 (2011) 324.