Ecotoxicology and Environmental Safety 120 (2015) 169–173
Contents lists available at ScienceDirect
Ecotoxicology and Environmental Safety journal homepage: www.elsevier.com/locate/ecoenv
Scoring of solvents used in analytical laboratories by their toxicological and exposure hazards Marek Tobiszewski n, Jacek Namieśnik Department of Analytical Chemistry, Chemical Faculty, Gdańsk University of Technology (GUT), 11/12 Gabriela Narutowicza Street, 80-233 Gdańsk, Poland
art ic l e i nf o
a b s t r a c t
Article history: Received 30 March 2015 Received in revised form 25 May 2015 Accepted 27 May 2015
Green analytical chemistry, although a well recognised concept, still lacks reliable environmental impact assessment procedures. This article describes scoring of solvents, frequently used in analytical laboratories, with CHEMS-1 model. The model uses toxicological and exposure data to calculate hazard values related to the utilisation of solvents. The original model was modified to incorporate hazards related to the volatility of chemicals. The scoring of hazard values showed that polar solvents are less hazardous. The scoring results were applied to assess the total hazard values in terms of solvent consumption. The hazard scores calculated for each chemical were multiplied by the volumes of solvent used during the analytical procedure. The results show that calculation of total procedural hazard values is valuable in the green analytical chemistry assessment procedure. Moreover, the assessment procedure can be combined with other procedural greenness assessment methods. & 2015 Elsevier Inc. All rights reserved.
Keywords: Green analytical chemistry Chemicals toxicity Hazard assessment Occupational exposure
1. Introduction Analytical chemistry laboratories consume high-purity solvents, which are widely applied in procedures as extraction agents, cleaning agents and mobile phases in liquid chromatography (Mohamed, 2015). Solvents mentioned above come from a variety of chemical groups, from polar alcohols to nonpolar aliphatic or aromatic hydrocarbons and halogenated solvents. Their physicochemical properties are relatively diverse, similarly to their inhalation and oral toxicities, influence on the environment (especially aquatic ecosystem) and the exposure parameters. Preparation of analytical sample solvents consumption, related to this analytical step, are concerned as the most polluting and hazardous of the whole analytical procedure (Spietelun et al., 2013; Li et al., 2013). Green analytical chemistry is the concept that recently finds more acceptance among the analysts. It assumes that the results of chemical measurements can be obtained with minimising the consumption of chemicals, reduction of emission and exposure without sacrificing the quality of analytical parameters (Armenta et al., 2008; Moros et al., 2010). Environmental impact is the factor that makes statistical difference between the analytical n
Corresponding author. E-mail address: [email protected] (M. Tobiszewski).
http://dx.doi.org/10.1016/j.ecoenv.2015.05.043 0147-6513/& 2015 Elsevier Inc. All rights reserved.
procedures, while parameters like limits of detection, recoveries and precisions are in most of the cases at comparable levels (Tobiszewski et al., 2013). Green analytical chemistry favours the procedures that minimise their impact by miniaturisation, application of solventless extraction techniques, greener reagents and alternative solvents. Only few propositions of techniques are known for assessment of the environmental impact or “greenness” of analytical procedures and these include NEMI labelling (Keith et al., 2007), analytical Eco-scale (Gałuszka et al., 2012), application of HPLC-EAT tool (Gaber et al., 2011) or multivariate statistics (Tobiszewski et al., 2013). The details of techniques mentioned are accessible in the references and have been discussed recently (Tobiszewski and Namiesnik, 2014). All of them are characterised by severe drawbacks, like tedious assessment procedures, lack of data availability as well as lack of information about the structure or nature of the hazards carried by application of the procedure. Still, this field requires new solutions to give easy-to-use, based on easily available information and clear-to-read “greenness” assessment procedures. The aim of the study is to perform scoring of the solvents commonly used in analytical laboratories with CHEMS-1 model (Swanson et al., 1997). Such scores can be useful at initial stages of procedure development, when solvents are selected. The scores can be also successfully used as one of the factors when assessing greenness of analytical procedures.
170
M. Tobiszewski, J. Namieśnik / Ecotoxicology and Environmental Safety 120 (2015) 169–173
2. Materials and methods The CHEMS-1 procedure was developed for relative scoring and ranking of chemicals with respect to their hazard. For the detailed discussion about CHEMS-1 the reader should kindly refer to the source article (Swanson et al., 1997) but brief introduction to the scoring procedure is presented below. Although the CHEMS-1 model was developed to score the risk of using chemicals in industry, we find such approach very attractive for scoring of solvents used in analytical laboratories. The procedure considers toxicity of chemicals to human and the environment as well as chemical exposure potential. The input data are toxicological and ecotoxicological endpoints collected from material safety data sheets (MSDS). Such data sources are easily available, which is very important to perform the chemical scoring by potential user. The assessment procedure involves transformation and scaling of the numerical values to give similar weights to each parameter. For non-toxic compounds the value for respective hazard values is set to 0, while for very toxic compounds the hazard value is set to maximum value of 5. Parameters included in the algorithm are hazards related to oral toxicity (HVORAL), inhalation toxicity (HVINH), carcinogenicity (HVCAR), other hazardous effects (HVHE), aquatic acute toxicity (HVFA), aquatic chronic toxicity (HVFC) and exposure-related parameters like biodegrability (HVBOD), hydrolysis (HVHYD) and bioconcentration (HVBCF). The algorithm to assess the total hazard (tHV) by CHEMS-1 algorithm is multiplication of the sum of hazards related to toxicity by hazards related to exposure factors (Swanson et al., 1997)
different physiochemical properties, that are commonly used in analytical chemistry laboratories. They find application as extraction solvents in different modes of liquid extraction, auxiliary solvents and mobile phases in liquid chromatography. Table 1 shows the scoring of solvents with original and modified CHEMS-1. The values of total hazard and total analytical hazard values are the highest for benzene and chlorinated solvents. These compounds posses high oral, inhalation and aquatic toxicities, are categorised as carcinogens and possess other specific human health effects. Moreover their exposure factor is high as they are relatively persistent in the environment. Boiling points of investigated compounds are mainly below 100 °C, so their analytical exposure factor is also high. In analytical laboratories they are usually used as extraction solvents in various modes of liquid–liquid extraction or solid–liquid extraction. The least hazardous organic solvents used in analytical laboratories are polar ones, like short chain alcohols, ethers, acetone and ethyl acetate. Their toxicities are comparatively low and they are readily biodegradable, hydrolysable and non-bioaccumulative. In analytical laboratories polar organic solvents are mainly used as Table 1 The results of total hazard values (tHV) and total analytical hazard values (taHV) calculation. Compound name
CAS number
tHV
Hydrocarbons
Pentane Hexane Cyclohexane Heptane Isooctane Benzene Toluene Xylenes
109-66-0 110-54-3 110-82-7 142-82-5 540-84-1 71-43-2 108-88-3 1330-20-7
23.4 36.4 53.8 81.4 57.9 79.8 15.2 20.9 57.1 79.2 84 122 43.9 60.5 51.5 68.2
Alcohols
Methanol Ethanol Isopropanol Heptanol Octanol Nonanol Benzyl alcohol Diethyl ether Methyl tert-butyl ether Tetrahydrofuran
67-56-1 64-17-5 67-63-0 111-70-6 111-87-5 143-08-8 100-51-6 60-29-7 1634-04-4 109-99-9
8.8 4.1 3.9 22.9 38.3 39.8 45.2 8.7 2.7 9.1
15.7 7.2 6.8 28.6 45.7 46.7 55.1 16 3.8 14.9
Aldehydes
Furfural Benzaldehyde
98-08-1 100-52-7
47.7 56.9
69.2 77.5
The algorithm of total analytical hazard value (taHV) calculation is like follows:
Ketones
Acetone
67-64-1
1.5
2.6
taHV=(HVORAL + HVINH + HVCAR + HVHE + HVORAL + HVFA
Organic acids
Formic acid Acetic acid Propionic acid
64-18-6 64-19-7 79-19-4
25.8 1.3 27
43 2.1 41.2
Esters
Ethyl acetate
141-78-6
5
7.3
Chlorinated hydrocarbons
Dichloromethane Chloroform Carbon tetrachloride Trichloroethylene Tetrachloroethylene 1.1.1-Trichloroethane 1.1.2.2-Tetrachloroethane Chlorobenzene
75-09-2 67-66-3 56-23-5 79-01-6 127-18-4 71-55-6 79-34-5 108-90-7
39.3 70.7 80 90.9 82.7 34.7 76.9 58
59.8 103.8 109.7 125.1 107.7 49 97.3 75.9
Others
Acetonitrile Carbon disulphide
75-05-8 75-15-0
18.8 61.6
26.8 89.6
tHV=(HVORAL + HVINH + HVCAR + HVHE + HVORAL + HVFA + HVFC)*(HVBOD + HVHYD + HVBCF) The main exposure pathway when solvents are applied in analytical laboratories is inhalation. The original scoring algorithm does not have any term related to exposure via inhalation. The hazard value related to exposure via inhalation could be calculated based on boiling point (BP) or vapour pressure. Both parameters are easily accessible (Mackay et al., 2006) but the operation with boiling points seem to be more convenient. Very volatile solvents (BPo 50 °C) are scored by 2.5 while semi-volatile solvents (BP4 200 °C) are scored by 1 hazard value. The hazard value, related to the volatility (HVVOL), was calculated with the following formulas:
HVVOL = 3–0.01*BP
Ethers
taHV
for 50°C ≤ BP < 200 °C
HVVOL = 1
for BP ≥ 200 °C
HVVOL = 2.5
for BP < 50 °C
+ HVFC)*(HVBOD + HVHYD + HVBCF + HVVOL) Original algorithm included the possibility of weighting hazard values (Swanson et al., 1997). In this study, however, we apply no weights, as we assume each hazard is equally important. Weighting of hazard values can be applied if there is one dominating exposure pathway or some data are significantly more or less reliable than other.
3. Results and discussion 3.1. Scoring of solvents The algorithm was run for 34 different solvents, characterised by
M. Tobiszewski, J. Namieśnik / Ecotoxicology and Environmental Safety 120 (2015) 169–173
constituents of mobile phases in HPLC but also sometimes are applied as extraction agents or auxiliary solvents. Main reason of result variation in obtaining solvent scoring is due to data in MSDS. First of all the toxicity endpoint values included in the algorithm were the lowest available (found) but there is also the possibility that other sources may give the endpoints of lower values. This fact can result in slightly different results of hazard values calculation. Second reason is that some information was not available in the MSDS. The NOEL values were calculated with previously described equations for the most of solvents (Swanson et al., 1997). Determination of inhalation toxicities values for benzaldehyde and propionic acid occurred to be a problem as these values were missing. In such case of missing parameter the hazard value is set to maximum, which is in agreement with precautionary principle.
3.2. Volume weighted scoring The results of solvents scoring can be successfully applied in assessment of analytical procedures. To perform such analysis the analytical total hazard value (taHV) score related to every solvent utilisation is multiplied by the unit of volume, in case of analytical procedure we suggest the volume of 1 mL
vtaHV=taHV*V As analytical procedures involve highly variable amounts of consumed solvents, from single microlitres to hundreds of millilitres, therefore the multiplier equal to one millilitre seems to be appropriate, since it is in the middle of this range. The value of volume-weighted total analytical hazard value for given solvent is obtained. To obtain procedure hazard value (pHV) it is required to sum total volume-weighted analytical hazard values, calculated for all n solvents consumed during analysis
pHV=vtaHV1 + vtaHV2 + ··· + vtaHVn
171
3.3. Case studies The applicability of procedural hazard values in assessment of solvent impact during the whole analytical methodology will be shown on four short case studies. All the assessed procedures are dedicated to polycyclic aromatic hydrocarbons determination in sediment samples but are based on different extraction approaches. Two of them are based on gas chromatographic separation, whereas the two other on liquid chromatographic separation. The schemes of analytical procedures are shown in Fig. 1. For the clarity of information the volumes of each solvent used at every procedure step are shown together with GHS hazard pictograms related to respective solvents. Fig. 1A shows the scheme of PLE–GC–MS procedure. During the PLE step the mixture of hexane and dichloromethane is used, resulting in the consumption of 17.6 mL of hexane and 4.4 mL of dichloromethane. The scores of taHV for both compounds are taken from Table 2. In this particular case the value of pHV is calculated in the following way:
pHV=17.6*81.35 + 4.4*59.79 = 1695 Fig. 1B shows the amount of solvents consumed during MAE sediment sample preparation followed by SPE. In this case 35.1 mL of relatively nontoxic methanol and 13 mL of hazardous dichloromethane are consumed. In this case the procedural hazard value is calculated
pHV=35.1*15.69 + 13*59.79 = 1328 The scheme of VAE–DLLME sample preparation followed by HPLC analysis is shown in Fig. 1C. Sub-millilitre volume of dichloromethane used results in small volume-weighted total analytical hazard (vtaHV) for this compound. Acetonitrile, which is used both at the stage of sample preparation and final determination, results totally in amount of 26.14 mL. The vtaHV of acetonitrile used as mobile phase in HPLC is 92% of pHV. Water as constituent of mobile phase is nontoxic and is omitted in the assessment procedure
Fig. 1. The scheme showing the amount of solvents consumed during: A – pressurised liquid extraction–gas chromatography–mass spectrometric detection (Choi et al., 2014); B – microwave assisted extraction–solid phase extraction–gas chromatography–mass spectrometric analysis (Itoh et al., 2008); C – vortex assisted extraction followed by dispersive liquid–liquid microextraction and high performance liquid chromatography (Leng et al., 2012); D – ultrasound assisted extraction with high performance liquid chromatographic procedure (Peng et al., 2012), all for the determination of PAHs in sediment samples.
172
M. Tobiszewski, J. Namieśnik / Ecotoxicology and Environmental Safety 120 (2015) 169–173
Table 2 Examples of solvent consumption hazard values scoring for different analytical procedures. Analytical technique
Analytes
Matrix
Solvents consumed
Procedural hazard value
Reference
PLE–GC–MS/MS
PBDEs
Indoor dust
17 mL Dichloromethane
1016
(Mercier et al., 2014)
DLLME–IC–FD
Alkylphenols
Water
1.5 mL Acetone; 0.05 mL trichloroethylene 6.53 mL Acetonitrile 21.6 Methanol
524
(Zgoła-Grześkowiak, 2010)
P&T–GC–ECD
THMs
Seawater
–
0
(Allonier et al., 2000)
DLLME–GC–MS
Biogenic amines
Beer
1 mL Acetonitrile 0.35 mL Toluene 0.075 mL Methanol
49
(Almeida et al., 2012)
MAE–DLLME–GC–MS
PAHs
Grilled meat
5 mL Ethanol 3 mL Acetic acid 0.3 mL Acetone 0.08 mL Tetrachloroethylene
51
(Kamankesh et al., 2015)
HS–SPME–GC–NPD
Venlafaxine
Whole blood
–
0
(Mastrogianni et al., 2012)
SBSE–HPLC–MS–MS
Pesticides
Water
10.1 mL Methanol 7.1 mL Acetonitrile 5 mL Dichloromethane 0.004 mL Formic acid
658
(Margoum et al., 2013)
EPA method 550 LLE–LC–UV
PAHs
Water
180 mL Dichloromethane 34.1 mL Acetonitrile
11,667
(Hodgeson, 1990)
pHV=26.14*26.83 + 0.08*59.79 = 706 Fig. 1D shows the amount of solvents used during USE-HPLC sediment sample analysis with extract clean-up. This procedure requires large amount of hexane, what is reflected in high pHV score. The methanol used as HPLC mobile phase contributes to 6% of the pHV
pHV=44.225*15.69 + 18*59.79 + 53*81.35 = 6082 In Table 2 more examples of calculated pHV are listed. The table gathers different examples of techniques, analytes and matrices to give the overview on the amounts of solvents consumed and their hazard values.
The latest drawback can be ignored when the calculated procedural hazard values are input data to other assessment procedures. The Eco-scale assessment procedure (Gałuszka et al., 2012) requires the amount of solvent and its toxicity as input data. The amount is given in three intervals: below 10 mL, in the range of 10–100 mL and above 100 mL. The toxicity is expressed as the amount of hazard pictograms related to given solvent. The proposed procedural solvent consumption hazard value is the parameter that after weighting can be incorporated to Eco-scale. Procedural hazard values can be reliable input data to other assessment procedures based on multivariate statistics (Tobiszewski et al., 2014) and multicriteria decision analysis (Tobiszewski and Orłowski, 2015).
3.4. Utility of solvents scoring The calculated analytical hazard values can be utilised in at least three ways. They can give the general information about hazard when solvents are selected during procedure design and optimisation. Environmental issues can be one of the criteria when preliminary solvent selection is performed. The second application of pHV is being assessment procedure itself. As the assessment procedure is taken into account, the calculation of pHV considers hazards related to solvents utilization, which is the most environmentally problematic issue. It describes the hazards related to each solvent in details. Apart from toxicity via different exposure to terrestrial and aquatic organisms it involves consideration of persistence of solvents in the environment. Finally, the modification of tHV to taHV by concerning the volatility of solvents is useful as the analyst exposure factor is directly involved in hazard value calculation. The main drawbacks of the proposed assessment procedure is that other environmental factors are not considered, such as energy consumption, solid wastes generation, corrosiveness of the environment or the managerial strategies to deal with wastes (including solvents themselves).
4. Conclusions The CHEMS-1 scoring of chemicals can be a useful tool in green analytical chemistry. The modification of the assessment procedure by incorporation of solvents volatility was done to better fit the exposure of analysts. The solvents used are characterised by highly variable hazards degree, generally lower with increased polarity. The volume-weighted hazard values are valuable tool in assessment of procedural environmental impact in the view of solvents consumption. The presented case studies have proved that this tool can be successfully applied in comparative studies. The calculation procedure takes into consideration parameters such as: the amount of solvent, its toxicity towards terrestrial and aquatic organisms, the persistence of the chemical, so it should be considered as comprehensive risk assessment tool. Finally, the calculation of procedural hazard values can be reliable input data to more advanced green analytical chemistry assessment procedures.
M. Tobiszewski, J. Namieśnik / Ecotoxicology and Environmental Safety 120 (2015) 169–173
Acknowledgements Marek Tobiszewski is grateful for the support from the programme the Center for Advanced Studies – the development of interdisciplinary doctoral studies at the Gdansk University of Technology in the key areas of the Europe 2020 Strategy (POKL04.03.0000-238/12).
References Armenta, S., Garrigues, S., de la Guardia, M., 2008. Green analytical chemistry. TrAC Trends Anal. Chem. 27, 497–511. Allonier, A.S., Khalanski, M., Bermond, A., Camel, V., 2000. Determination of trihalomethanes in chlorinated sea water samples using a purge-and-trap system coupled to gas chromatography. Talanta 51, 467–477. Almeida, C., Fernandes, J.O., Cunha, S.C., 2012. A novel dispersive liquid–liquid microextraction (DLLME) gas chromatography–mass spectrometry (GC–MS) method for the determination of eighteen biogenic amines in beer. Food Control 25, 380–388. Choi, M., Kim, Y.-J., Lee, I.-S., Choi, H.-G., 2014. Development of a one-step integrated pressurized liquid extractionand cleanup method for determining polycyclic aromatichydrocarbons in marine sediments. J. Chromatogr. A 1340, 8–14. Hodgeson, J.W., 1990. Determination of Polycyclic Aromatic Hydrocarbons in Drinking Water by Liquid–Liquid Extraction and HPLC with Coupled Ultraviolet and Fluorescence Detection. EPA, U.S., EPA Method 550. Gałuszka, A., Migaszewski, Z.M., Konieczka, P., Namieśnik, J., 2012. Analytical Ecoscale for assessing the greenness of analytical procedures. TrAC Trends Anal. Chem. 37, 61–72. Gaber, Y., Tornvall, U., Kumar, M.A., Amin, Magdy Ali, Hatti-Kaul, Rajni, 2011. HPLCEAT (Environmental Assessment Tool): a tool for profiling safety, health and environmental impacts of liquid chromatography methods. Green Chem. 13, 2021. Itoh, N., Numata, M., Yarita, T., 2008. Alkaline extraction in combination with microwave-assisted extraction followed by solid-phase extraction treatment for polycyclic aromatic hydrocarbons in a sediment sample. Anal. Chim. Acta 615, 47–53. Keith, L.H., Gron, L.U., Young, J.L., 2007. Green analytical methodologies. Chem. Rev. 107, 2695–2708. Kamankesh, M., Mohammadi, A., Hosseini, H., Tehrani, Z. Modarres, 2015. Rapid determination of polycyclic aromatic hydrocarbons in grilled meat using microwave-assisted extraction and dispersive liquid–liquid microextraction coupled to gas chromatography–mass spektrometry. Meat Sci. 103, 61–67. Li, Y., Fabiano-Tixier, A.S., Vian, M.A., Chemat, F., 2013. Solvent-free microwave extraction of bioactive compounds provides a tool for green analytical chemistry. TrAC Trends Anal. Chem. 47, 1–11.
173
Leng, G., Lui, G., Chen, Y., Yin, H., Dan, D., 2012. Vortex-assisted extraction combined with dispersive liquid–liquid microextraction for the determination of polycyclic aromatic hydrocarbons in sediment by high performance liquid chromatography. J. Sep. Sci. 35, 2796–2804. Mohamed, H.M., 2015. Green, environment-friendly, analytical tools give insights in pharmaceutical and cosmetics analysis. TrAC Trends Anal. Chem. 66, 176–192. Moros, J., Garrigues, S., de la Guardia, M., 2010. Vibrational spectroscopy provides a green tool for multi-component analysis. TrAC Trends Anal. Chem. 29, 578–591. Mackay, D., Ying Shiu, W., Ma, K.-C., Chi Lee, S., 2006. Handbook of Physical-Chemical Properties and Environmental Fate for Organic Chemicals. Taylor and Francis, Boca Raton, Florida. Mercier, F., Gilles, E., Saramito, G., Glorennec, P., Le Bot, B., 2014. A multi-residue method for the simultaneous analysis in indoor dust of several classes of semivolatile organic compounds by pressurized liquid extraction and gas chromatography/tandem mass spectrometry. J. Chromatogr. A 1336, 101–111. Mastrogianni, O., Theodoridis, G., Spagou, K., Violante, D., Henriques, T., Pouliopoulos, A., Psaroulis, K., Tsoukali, H., Raikos, N., 2012. Determination of venlafaxine in post-mortem whole blood by HS-SPME and GC–NPD. Forensic Sci. Int. 215, 105–109. Margoum, C., Guillemain, C., Yang, X., Coquery, M., 2013. Stir bar sorptive extraction coupled to liquid chromatography–tandem mass spectrometry for the determination of pesticides in water samples: method validation and measurement uncertainty. Talanta 116, 1–7. Peng, X., Yan, G., Li, X., Guo, X., Zhou, X., Wang, Y., 2012. Optimization of ultrasonic extraction and clean-up protocol for the determination of polycyclic aromatic hydrocarbons in marine sediments by high-performance liquid chromatography coupled with fluorescence detection. J. Ocean Univ. China 11, 331–338. Spietelun, A., Marcinkowski, Ł., de la Guardia, M., Namieśnik, J., 2013. Recent developments and future trends in solid phase microextraction techniques towards green analytical chemistry. J. Chromatogr. A 1321, 1–13. Swanson, M.B., Davis, G.A., Kincaid, L.E., Schultz, T.W., Bartmess, J.E., Jones, S.L., George, E. Lou, 1997. A screening method for ranking and scoring chemicals by potential human health and environmental impacts. Environ. Toxicol. Chem. 16, 372–383. Tobiszewski, M., Tsakovski, S., Simeonov, V., Namieśnik, J., 2013. Application of multivariate statistics in assessment of green analytical chemistry parameters of analytical methodologies. Green Chem. 15, 1615. Tobiszewski, M., Namiesnik, J., 2014. Developments in green chromatography. LC GC Eur. 27, 405–408. Tobiszewski, M., Tsakovski, S., Simeonov, V., Namieśnik, J., 2014. Multivariate statistical comparison of analytical procedures for benzene and phenol determination with respect to their environmental impact. Talanta 130, 449–455. Tobiszewski, M., Orłowski, A., 2015. Multicriteria decision analysis in ranking of analytical procedures for aldrin determination in water. J. Chromatogr. A 1387, 116–122. Zgoła-Grześkowiak, A., 2010. Dispersive liquid–liquid microextraction applied to isolation and concentration of alkylphenols and their short-chained ethoxylates in water samples. J. Chromatogr. A 1217, 1761–1766.
Contents lists available at ScienceDirect
Ecotoxicology and Environmental Safety journal homepage: www.elsevier.com/locate/ecoenv
Scoring of solvents used in analytical laboratories by their toxicological and exposure hazards Marek Tobiszewski n, Jacek Namieśnik Department of Analytical Chemistry, Chemical Faculty, Gdańsk University of Technology (GUT), 11/12 Gabriela Narutowicza Street, 80-233 Gdańsk, Poland
art ic l e i nf o
a b s t r a c t
Article history: Received 30 March 2015 Received in revised form 25 May 2015 Accepted 27 May 2015
Green analytical chemistry, although a well recognised concept, still lacks reliable environmental impact assessment procedures. This article describes scoring of solvents, frequently used in analytical laboratories, with CHEMS-1 model. The model uses toxicological and exposure data to calculate hazard values related to the utilisation of solvents. The original model was modified to incorporate hazards related to the volatility of chemicals. The scoring of hazard values showed that polar solvents are less hazardous. The scoring results were applied to assess the total hazard values in terms of solvent consumption. The hazard scores calculated for each chemical were multiplied by the volumes of solvent used during the analytical procedure. The results show that calculation of total procedural hazard values is valuable in the green analytical chemistry assessment procedure. Moreover, the assessment procedure can be combined with other procedural greenness assessment methods. & 2015 Elsevier Inc. All rights reserved.
Keywords: Green analytical chemistry Chemicals toxicity Hazard assessment Occupational exposure
1. Introduction Analytical chemistry laboratories consume high-purity solvents, which are widely applied in procedures as extraction agents, cleaning agents and mobile phases in liquid chromatography (Mohamed, 2015). Solvents mentioned above come from a variety of chemical groups, from polar alcohols to nonpolar aliphatic or aromatic hydrocarbons and halogenated solvents. Their physicochemical properties are relatively diverse, similarly to their inhalation and oral toxicities, influence on the environment (especially aquatic ecosystem) and the exposure parameters. Preparation of analytical sample solvents consumption, related to this analytical step, are concerned as the most polluting and hazardous of the whole analytical procedure (Spietelun et al., 2013; Li et al., 2013). Green analytical chemistry is the concept that recently finds more acceptance among the analysts. It assumes that the results of chemical measurements can be obtained with minimising the consumption of chemicals, reduction of emission and exposure without sacrificing the quality of analytical parameters (Armenta et al., 2008; Moros et al., 2010). Environmental impact is the factor that makes statistical difference between the analytical n
Corresponding author. E-mail address: [email protected] (M. Tobiszewski).
http://dx.doi.org/10.1016/j.ecoenv.2015.05.043 0147-6513/& 2015 Elsevier Inc. All rights reserved.
procedures, while parameters like limits of detection, recoveries and precisions are in most of the cases at comparable levels (Tobiszewski et al., 2013). Green analytical chemistry favours the procedures that minimise their impact by miniaturisation, application of solventless extraction techniques, greener reagents and alternative solvents. Only few propositions of techniques are known for assessment of the environmental impact or “greenness” of analytical procedures and these include NEMI labelling (Keith et al., 2007), analytical Eco-scale (Gałuszka et al., 2012), application of HPLC-EAT tool (Gaber et al., 2011) or multivariate statistics (Tobiszewski et al., 2013). The details of techniques mentioned are accessible in the references and have been discussed recently (Tobiszewski and Namiesnik, 2014). All of them are characterised by severe drawbacks, like tedious assessment procedures, lack of data availability as well as lack of information about the structure or nature of the hazards carried by application of the procedure. Still, this field requires new solutions to give easy-to-use, based on easily available information and clear-to-read “greenness” assessment procedures. The aim of the study is to perform scoring of the solvents commonly used in analytical laboratories with CHEMS-1 model (Swanson et al., 1997). Such scores can be useful at initial stages of procedure development, when solvents are selected. The scores can be also successfully used as one of the factors when assessing greenness of analytical procedures.
170
M. Tobiszewski, J. Namieśnik / Ecotoxicology and Environmental Safety 120 (2015) 169–173
2. Materials and methods The CHEMS-1 procedure was developed for relative scoring and ranking of chemicals with respect to their hazard. For the detailed discussion about CHEMS-1 the reader should kindly refer to the source article (Swanson et al., 1997) but brief introduction to the scoring procedure is presented below. Although the CHEMS-1 model was developed to score the risk of using chemicals in industry, we find such approach very attractive for scoring of solvents used in analytical laboratories. The procedure considers toxicity of chemicals to human and the environment as well as chemical exposure potential. The input data are toxicological and ecotoxicological endpoints collected from material safety data sheets (MSDS). Such data sources are easily available, which is very important to perform the chemical scoring by potential user. The assessment procedure involves transformation and scaling of the numerical values to give similar weights to each parameter. For non-toxic compounds the value for respective hazard values is set to 0, while for very toxic compounds the hazard value is set to maximum value of 5. Parameters included in the algorithm are hazards related to oral toxicity (HVORAL), inhalation toxicity (HVINH), carcinogenicity (HVCAR), other hazardous effects (HVHE), aquatic acute toxicity (HVFA), aquatic chronic toxicity (HVFC) and exposure-related parameters like biodegrability (HVBOD), hydrolysis (HVHYD) and bioconcentration (HVBCF). The algorithm to assess the total hazard (tHV) by CHEMS-1 algorithm is multiplication of the sum of hazards related to toxicity by hazards related to exposure factors (Swanson et al., 1997)
different physiochemical properties, that are commonly used in analytical chemistry laboratories. They find application as extraction solvents in different modes of liquid extraction, auxiliary solvents and mobile phases in liquid chromatography. Table 1 shows the scoring of solvents with original and modified CHEMS-1. The values of total hazard and total analytical hazard values are the highest for benzene and chlorinated solvents. These compounds posses high oral, inhalation and aquatic toxicities, are categorised as carcinogens and possess other specific human health effects. Moreover their exposure factor is high as they are relatively persistent in the environment. Boiling points of investigated compounds are mainly below 100 °C, so their analytical exposure factor is also high. In analytical laboratories they are usually used as extraction solvents in various modes of liquid–liquid extraction or solid–liquid extraction. The least hazardous organic solvents used in analytical laboratories are polar ones, like short chain alcohols, ethers, acetone and ethyl acetate. Their toxicities are comparatively low and they are readily biodegradable, hydrolysable and non-bioaccumulative. In analytical laboratories polar organic solvents are mainly used as Table 1 The results of total hazard values (tHV) and total analytical hazard values (taHV) calculation. Compound name
CAS number
tHV
Hydrocarbons
Pentane Hexane Cyclohexane Heptane Isooctane Benzene Toluene Xylenes
109-66-0 110-54-3 110-82-7 142-82-5 540-84-1 71-43-2 108-88-3 1330-20-7
23.4 36.4 53.8 81.4 57.9 79.8 15.2 20.9 57.1 79.2 84 122 43.9 60.5 51.5 68.2
Alcohols
Methanol Ethanol Isopropanol Heptanol Octanol Nonanol Benzyl alcohol Diethyl ether Methyl tert-butyl ether Tetrahydrofuran
67-56-1 64-17-5 67-63-0 111-70-6 111-87-5 143-08-8 100-51-6 60-29-7 1634-04-4 109-99-9
8.8 4.1 3.9 22.9 38.3 39.8 45.2 8.7 2.7 9.1
15.7 7.2 6.8 28.6 45.7 46.7 55.1 16 3.8 14.9
Aldehydes
Furfural Benzaldehyde
98-08-1 100-52-7
47.7 56.9
69.2 77.5
The algorithm of total analytical hazard value (taHV) calculation is like follows:
Ketones
Acetone
67-64-1
1.5
2.6
taHV=(HVORAL + HVINH + HVCAR + HVHE + HVORAL + HVFA
Organic acids
Formic acid Acetic acid Propionic acid
64-18-6 64-19-7 79-19-4
25.8 1.3 27
43 2.1 41.2
Esters
Ethyl acetate
141-78-6
5
7.3
Chlorinated hydrocarbons
Dichloromethane Chloroform Carbon tetrachloride Trichloroethylene Tetrachloroethylene 1.1.1-Trichloroethane 1.1.2.2-Tetrachloroethane Chlorobenzene
75-09-2 67-66-3 56-23-5 79-01-6 127-18-4 71-55-6 79-34-5 108-90-7
39.3 70.7 80 90.9 82.7 34.7 76.9 58
59.8 103.8 109.7 125.1 107.7 49 97.3 75.9
Others
Acetonitrile Carbon disulphide
75-05-8 75-15-0
18.8 61.6
26.8 89.6
tHV=(HVORAL + HVINH + HVCAR + HVHE + HVORAL + HVFA + HVFC)*(HVBOD + HVHYD + HVBCF) The main exposure pathway when solvents are applied in analytical laboratories is inhalation. The original scoring algorithm does not have any term related to exposure via inhalation. The hazard value related to exposure via inhalation could be calculated based on boiling point (BP) or vapour pressure. Both parameters are easily accessible (Mackay et al., 2006) but the operation with boiling points seem to be more convenient. Very volatile solvents (BPo 50 °C) are scored by 2.5 while semi-volatile solvents (BP4 200 °C) are scored by 1 hazard value. The hazard value, related to the volatility (HVVOL), was calculated with the following formulas:
HVVOL = 3–0.01*BP
Ethers
taHV
for 50°C ≤ BP < 200 °C
HVVOL = 1
for BP ≥ 200 °C
HVVOL = 2.5
for BP < 50 °C
+ HVFC)*(HVBOD + HVHYD + HVBCF + HVVOL) Original algorithm included the possibility of weighting hazard values (Swanson et al., 1997). In this study, however, we apply no weights, as we assume each hazard is equally important. Weighting of hazard values can be applied if there is one dominating exposure pathway or some data are significantly more or less reliable than other.
3. Results and discussion 3.1. Scoring of solvents The algorithm was run for 34 different solvents, characterised by
M. Tobiszewski, J. Namieśnik / Ecotoxicology and Environmental Safety 120 (2015) 169–173
constituents of mobile phases in HPLC but also sometimes are applied as extraction agents or auxiliary solvents. Main reason of result variation in obtaining solvent scoring is due to data in MSDS. First of all the toxicity endpoint values included in the algorithm were the lowest available (found) but there is also the possibility that other sources may give the endpoints of lower values. This fact can result in slightly different results of hazard values calculation. Second reason is that some information was not available in the MSDS. The NOEL values were calculated with previously described equations for the most of solvents (Swanson et al., 1997). Determination of inhalation toxicities values for benzaldehyde and propionic acid occurred to be a problem as these values were missing. In such case of missing parameter the hazard value is set to maximum, which is in agreement with precautionary principle.
3.2. Volume weighted scoring The results of solvents scoring can be successfully applied in assessment of analytical procedures. To perform such analysis the analytical total hazard value (taHV) score related to every solvent utilisation is multiplied by the unit of volume, in case of analytical procedure we suggest the volume of 1 mL
vtaHV=taHV*V As analytical procedures involve highly variable amounts of consumed solvents, from single microlitres to hundreds of millilitres, therefore the multiplier equal to one millilitre seems to be appropriate, since it is in the middle of this range. The value of volume-weighted total analytical hazard value for given solvent is obtained. To obtain procedure hazard value (pHV) it is required to sum total volume-weighted analytical hazard values, calculated for all n solvents consumed during analysis
pHV=vtaHV1 + vtaHV2 + ··· + vtaHVn
171
3.3. Case studies The applicability of procedural hazard values in assessment of solvent impact during the whole analytical methodology will be shown on four short case studies. All the assessed procedures are dedicated to polycyclic aromatic hydrocarbons determination in sediment samples but are based on different extraction approaches. Two of them are based on gas chromatographic separation, whereas the two other on liquid chromatographic separation. The schemes of analytical procedures are shown in Fig. 1. For the clarity of information the volumes of each solvent used at every procedure step are shown together with GHS hazard pictograms related to respective solvents. Fig. 1A shows the scheme of PLE–GC–MS procedure. During the PLE step the mixture of hexane and dichloromethane is used, resulting in the consumption of 17.6 mL of hexane and 4.4 mL of dichloromethane. The scores of taHV for both compounds are taken from Table 2. In this particular case the value of pHV is calculated in the following way:
pHV=17.6*81.35 + 4.4*59.79 = 1695 Fig. 1B shows the amount of solvents consumed during MAE sediment sample preparation followed by SPE. In this case 35.1 mL of relatively nontoxic methanol and 13 mL of hazardous dichloromethane are consumed. In this case the procedural hazard value is calculated
pHV=35.1*15.69 + 13*59.79 = 1328 The scheme of VAE–DLLME sample preparation followed by HPLC analysis is shown in Fig. 1C. Sub-millilitre volume of dichloromethane used results in small volume-weighted total analytical hazard (vtaHV) for this compound. Acetonitrile, which is used both at the stage of sample preparation and final determination, results totally in amount of 26.14 mL. The vtaHV of acetonitrile used as mobile phase in HPLC is 92% of pHV. Water as constituent of mobile phase is nontoxic and is omitted in the assessment procedure
Fig. 1. The scheme showing the amount of solvents consumed during: A – pressurised liquid extraction–gas chromatography–mass spectrometric detection (Choi et al., 2014); B – microwave assisted extraction–solid phase extraction–gas chromatography–mass spectrometric analysis (Itoh et al., 2008); C – vortex assisted extraction followed by dispersive liquid–liquid microextraction and high performance liquid chromatography (Leng et al., 2012); D – ultrasound assisted extraction with high performance liquid chromatographic procedure (Peng et al., 2012), all for the determination of PAHs in sediment samples.
172
M. Tobiszewski, J. Namieśnik / Ecotoxicology and Environmental Safety 120 (2015) 169–173
Table 2 Examples of solvent consumption hazard values scoring for different analytical procedures. Analytical technique
Analytes
Matrix
Solvents consumed
Procedural hazard value
Reference
PLE–GC–MS/MS
PBDEs
Indoor dust
17 mL Dichloromethane
1016
(Mercier et al., 2014)
DLLME–IC–FD
Alkylphenols
Water
1.5 mL Acetone; 0.05 mL trichloroethylene 6.53 mL Acetonitrile 21.6 Methanol
524
(Zgoła-Grześkowiak, 2010)
P&T–GC–ECD
THMs
Seawater
–
0
(Allonier et al., 2000)
DLLME–GC–MS
Biogenic amines
Beer
1 mL Acetonitrile 0.35 mL Toluene 0.075 mL Methanol
49
(Almeida et al., 2012)
MAE–DLLME–GC–MS
PAHs
Grilled meat
5 mL Ethanol 3 mL Acetic acid 0.3 mL Acetone 0.08 mL Tetrachloroethylene
51
(Kamankesh et al., 2015)
HS–SPME–GC–NPD
Venlafaxine
Whole blood
–
0
(Mastrogianni et al., 2012)
SBSE–HPLC–MS–MS
Pesticides
Water
10.1 mL Methanol 7.1 mL Acetonitrile 5 mL Dichloromethane 0.004 mL Formic acid
658
(Margoum et al., 2013)
EPA method 550 LLE–LC–UV
PAHs
Water
180 mL Dichloromethane 34.1 mL Acetonitrile
11,667
(Hodgeson, 1990)
pHV=26.14*26.83 + 0.08*59.79 = 706 Fig. 1D shows the amount of solvents used during USE-HPLC sediment sample analysis with extract clean-up. This procedure requires large amount of hexane, what is reflected in high pHV score. The methanol used as HPLC mobile phase contributes to 6% of the pHV
pHV=44.225*15.69 + 18*59.79 + 53*81.35 = 6082 In Table 2 more examples of calculated pHV are listed. The table gathers different examples of techniques, analytes and matrices to give the overview on the amounts of solvents consumed and their hazard values.
The latest drawback can be ignored when the calculated procedural hazard values are input data to other assessment procedures. The Eco-scale assessment procedure (Gałuszka et al., 2012) requires the amount of solvent and its toxicity as input data. The amount is given in three intervals: below 10 mL, in the range of 10–100 mL and above 100 mL. The toxicity is expressed as the amount of hazard pictograms related to given solvent. The proposed procedural solvent consumption hazard value is the parameter that after weighting can be incorporated to Eco-scale. Procedural hazard values can be reliable input data to other assessment procedures based on multivariate statistics (Tobiszewski et al., 2014) and multicriteria decision analysis (Tobiszewski and Orłowski, 2015).
3.4. Utility of solvents scoring The calculated analytical hazard values can be utilised in at least three ways. They can give the general information about hazard when solvents are selected during procedure design and optimisation. Environmental issues can be one of the criteria when preliminary solvent selection is performed. The second application of pHV is being assessment procedure itself. As the assessment procedure is taken into account, the calculation of pHV considers hazards related to solvents utilization, which is the most environmentally problematic issue. It describes the hazards related to each solvent in details. Apart from toxicity via different exposure to terrestrial and aquatic organisms it involves consideration of persistence of solvents in the environment. Finally, the modification of tHV to taHV by concerning the volatility of solvents is useful as the analyst exposure factor is directly involved in hazard value calculation. The main drawbacks of the proposed assessment procedure is that other environmental factors are not considered, such as energy consumption, solid wastes generation, corrosiveness of the environment or the managerial strategies to deal with wastes (including solvents themselves).
4. Conclusions The CHEMS-1 scoring of chemicals can be a useful tool in green analytical chemistry. The modification of the assessment procedure by incorporation of solvents volatility was done to better fit the exposure of analysts. The solvents used are characterised by highly variable hazards degree, generally lower with increased polarity. The volume-weighted hazard values are valuable tool in assessment of procedural environmental impact in the view of solvents consumption. The presented case studies have proved that this tool can be successfully applied in comparative studies. The calculation procedure takes into consideration parameters such as: the amount of solvent, its toxicity towards terrestrial and aquatic organisms, the persistence of the chemical, so it should be considered as comprehensive risk assessment tool. Finally, the calculation of procedural hazard values can be reliable input data to more advanced green analytical chemistry assessment procedures.
M. Tobiszewski, J. Namieśnik / Ecotoxicology and Environmental Safety 120 (2015) 169–173
Acknowledgements Marek Tobiszewski is grateful for the support from the programme the Center for Advanced Studies – the development of interdisciplinary doctoral studies at the Gdansk University of Technology in the key areas of the Europe 2020 Strategy (POKL04.03.0000-238/12).
References Armenta, S., Garrigues, S., de la Guardia, M., 2008. Green analytical chemistry. TrAC Trends Anal. Chem. 27, 497–511. Allonier, A.S., Khalanski, M., Bermond, A., Camel, V., 2000. Determination of trihalomethanes in chlorinated sea water samples using a purge-and-trap system coupled to gas chromatography. Talanta 51, 467–477. Almeida, C., Fernandes, J.O., Cunha, S.C., 2012. A novel dispersive liquid–liquid microextraction (DLLME) gas chromatography–mass spectrometry (GC–MS) method for the determination of eighteen biogenic amines in beer. Food Control 25, 380–388. Choi, M., Kim, Y.-J., Lee, I.-S., Choi, H.-G., 2014. Development of a one-step integrated pressurized liquid extractionand cleanup method for determining polycyclic aromatichydrocarbons in marine sediments. J. Chromatogr. A 1340, 8–14. Hodgeson, J.W., 1990. Determination of Polycyclic Aromatic Hydrocarbons in Drinking Water by Liquid–Liquid Extraction and HPLC with Coupled Ultraviolet and Fluorescence Detection. EPA, U.S., EPA Method 550. Gałuszka, A., Migaszewski, Z.M., Konieczka, P., Namieśnik, J., 2012. Analytical Ecoscale for assessing the greenness of analytical procedures. TrAC Trends Anal. Chem. 37, 61–72. Gaber, Y., Tornvall, U., Kumar, M.A., Amin, Magdy Ali, Hatti-Kaul, Rajni, 2011. HPLCEAT (Environmental Assessment Tool): a tool for profiling safety, health and environmental impacts of liquid chromatography methods. Green Chem. 13, 2021. Itoh, N., Numata, M., Yarita, T., 2008. Alkaline extraction in combination with microwave-assisted extraction followed by solid-phase extraction treatment for polycyclic aromatic hydrocarbons in a sediment sample. Anal. Chim. Acta 615, 47–53. Keith, L.H., Gron, L.U., Young, J.L., 2007. Green analytical methodologies. Chem. Rev. 107, 2695–2708. Kamankesh, M., Mohammadi, A., Hosseini, H., Tehrani, Z. Modarres, 2015. Rapid determination of polycyclic aromatic hydrocarbons in grilled meat using microwave-assisted extraction and dispersive liquid–liquid microextraction coupled to gas chromatography–mass spektrometry. Meat Sci. 103, 61–67. Li, Y., Fabiano-Tixier, A.S., Vian, M.A., Chemat, F., 2013. Solvent-free microwave extraction of bioactive compounds provides a tool for green analytical chemistry. TrAC Trends Anal. Chem. 47, 1–11.
173
Leng, G., Lui, G., Chen, Y., Yin, H., Dan, D., 2012. Vortex-assisted extraction combined with dispersive liquid–liquid microextraction for the determination of polycyclic aromatic hydrocarbons in sediment by high performance liquid chromatography. J. Sep. Sci. 35, 2796–2804. Mohamed, H.M., 2015. Green, environment-friendly, analytical tools give insights in pharmaceutical and cosmetics analysis. TrAC Trends Anal. Chem. 66, 176–192. Moros, J., Garrigues, S., de la Guardia, M., 2010. Vibrational spectroscopy provides a green tool for multi-component analysis. TrAC Trends Anal. Chem. 29, 578–591. Mackay, D., Ying Shiu, W., Ma, K.-C., Chi Lee, S., 2006. Handbook of Physical-Chemical Properties and Environmental Fate for Organic Chemicals. Taylor and Francis, Boca Raton, Florida. Mercier, F., Gilles, E., Saramito, G., Glorennec, P., Le Bot, B., 2014. A multi-residue method for the simultaneous analysis in indoor dust of several classes of semivolatile organic compounds by pressurized liquid extraction and gas chromatography/tandem mass spectrometry. J. Chromatogr. A 1336, 101–111. Mastrogianni, O., Theodoridis, G., Spagou, K., Violante, D., Henriques, T., Pouliopoulos, A., Psaroulis, K., Tsoukali, H., Raikos, N., 2012. Determination of venlafaxine in post-mortem whole blood by HS-SPME and GC–NPD. Forensic Sci. Int. 215, 105–109. Margoum, C., Guillemain, C., Yang, X., Coquery, M., 2013. Stir bar sorptive extraction coupled to liquid chromatography–tandem mass spectrometry for the determination of pesticides in water samples: method validation and measurement uncertainty. Talanta 116, 1–7. Peng, X., Yan, G., Li, X., Guo, X., Zhou, X., Wang, Y., 2012. Optimization of ultrasonic extraction and clean-up protocol for the determination of polycyclic aromatic hydrocarbons in marine sediments by high-performance liquid chromatography coupled with fluorescence detection. J. Ocean Univ. China 11, 331–338. Spietelun, A., Marcinkowski, Ł., de la Guardia, M., Namieśnik, J., 2013. Recent developments and future trends in solid phase microextraction techniques towards green analytical chemistry. J. Chromatogr. A 1321, 1–13. Swanson, M.B., Davis, G.A., Kincaid, L.E., Schultz, T.W., Bartmess, J.E., Jones, S.L., George, E. Lou, 1997. A screening method for ranking and scoring chemicals by potential human health and environmental impacts. Environ. Toxicol. Chem. 16, 372–383. Tobiszewski, M., Tsakovski, S., Simeonov, V., Namieśnik, J., 2013. Application of multivariate statistics in assessment of green analytical chemistry parameters of analytical methodologies. Green Chem. 15, 1615. Tobiszewski, M., Namiesnik, J., 2014. Developments in green chromatography. LC GC Eur. 27, 405–408. Tobiszewski, M., Tsakovski, S., Simeonov, V., Namieśnik, J., 2014. Multivariate statistical comparison of analytical procedures for benzene and phenol determination with respect to their environmental impact. Talanta 130, 449–455. Tobiszewski, M., Orłowski, A., 2015. Multicriteria decision analysis in ranking of analytical procedures for aldrin determination in water. J. Chromatogr. A 1387, 116–122. Zgoła-Grześkowiak, A., 2010. Dispersive liquid–liquid microextraction applied to isolation and concentration of alkylphenols and their short-chained ethoxylates in water samples. J. Chromatogr. A 1217, 1761–1766.
