Chemosphere 107 (2014) 23–30
Contents lists available at ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
Characteristics of volatile organic compounds emission profiles from hot road bitumens Grzegorz Boczkaj a,⇑, Andrzej Przyjazny b, Marian Kamin´ski a a b
Gdansk University of Technology, Chemical Faculty, Department of Chemical and Process Engineering, G. Narutowicza St. 11/12, 80-233 Gdansk, Poland Kettering University, 1700 University Avenue, Flint, MI 48504, USA
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
New procedure for comparison of
VOC emission profiles from hot bituminous materials. Very complex chemical composition of bitumen fumes. The content of VOC strongly depends on the degree of bitumen oxidation. Useful tools for optimization of the bitumen manufacturing processes.
a r t i c l e
i n f o
Article history: Received 15 October 2013 Received in revised form 17 February 2014 Accepted 19 February 2014
Handling Editor: I. Cousins Keywords: Asphalt GC–MS Volatile organic compounds (VOCs) Bitumen fumes NPD Pollutants
a b s t r a c t A procedure for the investigation and comparison of volatile organic compounds (VOCs) emission profiles to the atmosphere from road bitumens with various degrees of oxidation is proposed. The procedure makes use of headspace analysis and gas chromatography with universal as well as selective detection, including gas chromatography–mass spectrometry (GC–MS). The studies revealed that so-called vacuum residue, which is the main component of the charge, contains variable VOC concentrations, from trace to relatively high ones, depending on the extent of thermal cracking in the boiler of the vacuum distillation column. The VOC content in the oxidation product, so-called oxidized paving bitumen, is similarly varied. There are major differences in VOC emission profiles between vacuum residue and oxidized bitumens undergoing thermal cracking. The VOC content in oxidized bitumens, which did not undergo thermal cracking, increases with the degree of oxidation of bitumens. The studies revealed that the total VOC content increases from about 120 ppm for the raw vacuum residue to about 1900 ppm for so-called bitumen 35/ 50. The amount of volatile sulfur compounds (VSCs) in the volatile fraction of fumes of oxidized bitumens increases with the degree of oxidation of bitumen and constitutes from 0.34% to 3.66% (w/w). The contribution of volatile nitrogen compounds (VNCs) to total VOC content remains constant for the investigated types of bitumens (from 0.16 to 0.28% (w/w) of total VOCs). The results of these studies can also find use during the selection of appropriate bitumen additives to minimize their malodorousness. The obtained data append the existing knowledge on VOC emission from oxidized bitumens. They should be included in reports on the environmental impact of facilities in which hot bitumen binders are used. Ó 2014 Elsevier Ltd. All rights reserved.
⇑ Corresponding author. Tel.: +48 697970303; fax: +48 58 347 26 94. E-mail addresses: [email protected] (G. Boczkaj), [email protected] (A. Przyjazny). http://dx.doi.org/10.1016/j.chemosphere.2014.02.070 0045-6535/Ó 2014 Elsevier Ltd. All rights reserved.
24
G. Boczkaj et al. / Chemosphere 107 (2014) 23–30
1. Introduction The occurrence of volatile organic compounds in bitumen binders constitutes a serious technological and environmental problem, taking place from the moment of their formation during production until their utilization in road paving and other applications. Over 250 examples of bitumen application have been reported (Read and Whiteoak, 2003). In most cases bitumens have to be heated prior to use in order to lower viscosity enabling their application. The production of bitumens involves oxidation of residue from vacuum distillation of petroleum to obtain products with desired properties. During operations yielding the raw material for bitumen production (vacuum distillation) as well as the process of oxidation with hot air, the bitumen mass deposited on heating elements undergoes partial thermal cracking, which results in the formation of unsaturated and aromatic compounds as well as hydrogen sulfide, water vapor, carbonyl sulfide, carbon disulfide and others. These compounds undergo further transformations yielding a variety of volatile compounds: ketones, aldehydes, organic acids, phenols and their derivatives, as well as organosulfur and organonitrogen compounds. A fraction of the resulting volatile compounds is removed from the reactor with hot air, yielding so-called exhaust gases, which undergo scrubbing in a basic aqueous solution or absorption in wash oil (Boczkaj et al., 2010, 2014). VOCs remaining in the bitumen mass are emitted during storage (Deygout, 2011; Davie et al., 1993a, 1993b; Trumbore, 1999), transport (EPA, 2000a), making mineral–bitumen mixtures (EPA, 2000b; Ventura et al., 2007; Ruehl et al., 2007; Gasthauer et al., 2008; Chauhan et al., 2010; Jullien et al., 2010), roof impregnation (Rogge et al., 1997; Trumbore, 1998; Franzen and Trumbore, 2000; Kriech et al., 2004; Ruehl et al., 2007; Parker et al., 2011) and road paving (Brandt et al., 1985; Monarca et al., 1987; Greenspan et al., 1995; Hicks, 1995; Gamble et al., 1999; Burstyn and Kromhout, 2000; Burstyn et al., 2003; Posniak, 2005; Preiss et al., 2006; Hugener et al., 2010; Breuer et al., 2011; Spickenheuer et al., 2011; Raulf-Heimsoth et al., 2011). The studies on characteristics of fume emissions from hot bitumen binders under real conditions published thus far have mostly dealt with control of total parameters, i.e. total VOC content, total PAHs, particulate matter (PM), benzene soluble matter (BSM). Hydrogen sulfide emission has also been determined. Individual polycyclic aromatic compounds have been determined in bitumen fume condensate (Blomberg et al., 1999; Brandt et al., 2000; Burstyn et al., 2002; Ventura et al., 2007; Rasoulzadeh et al., 2011; Trumbore et al., 2011). A detailed investigation of the determination of unsubstituted PAHs, considered to be carcinogenic, revealed that for bitumen samples the analytical problem is very complex and requires a two-step sample preparation procedure prior to the final determination step by chromatographic analysis (Gilgenast et al., 2011). The distribution of distillation temperature of fume condensate characterizing their volatility range can be determined using simulated distillation (SIMDIS) technique (Kriech et al., 1999; Ekstrom et al., 2001; Preiss et al., 2006; Kriech et al., 2007). For hydrocarbon mixtures containing significant amounts of aromatic compounds more accurate results are obtained by empty-column gas chromatography (EC-GC) (Boczkaj et al., 2011; Boczkaj and Kamin´ski, 2013). Relatively little is known on detailed characteristics of the presence of VOCs in bitumen fumes. So far, despite information on the possibility of formation of hydrogen sulfide and volatile sulfur compounds (VSCs) during the production of oxidized bitumens (Davie et al., 1993a; Boczkaj et al., 2010; Deygout, 2011; Davie et al., 1993b), no papers are available on the identification of VSCs released at elevated temperatures from bituminous materials. The information on the presence of volatile nitrogen compounds (VNCs) in bitumen fumes is also scarce. Screening tests for VOCs
were carried out using static headspace analysis combined with solid-phase microextraction/gas chromatography/mass spectrometry (SHS–SPME–GC–MS). Using this technique, 25 polycyclic aromatic compounds (PACs) occurring in bitumen fumes were identified (Agozzino et al., 1999). Determination of individual components in very complex hydrocarbon matrices by one-dimensional gas chromatography often fails (Boczkaj et al., 2013). The SHS–SPME–GC–MS technique was also used for the identification of volatile components of asphalt release agents (Tang and Isacsson, 2005; Tang and Isacsson, 2006) and bituminous sealants (Tang et al., 2006). Emission of VOCs from hot bitumens has several consequences. One of the most important environmental problems is the direct effect of fumes on workers pouring asphalt, making and transporting asphalt mixes and road construction. Such an exposure is of a long-term nature. The effect of bitumen fumes on human health has been investigated in numerous studies. The emission of VOCs from bitumens also impacts the quality of air at the emission site. In addition, global impact of VOC emissions from bitumens should be considered: pollution of the atmosphere by volatile organic compounds and their transformation products, especially the depletion of the ozone layer and formation of secondary aerosols (Andreae and Crutzen, 1997). This paper presents a new procedure for the comparison of VOC emission profiles from hot bituminous materials. The studies enabled detailed characteristics of volatile fraction of bitumen fumes.
2. Materials and methods 2.1. Materials Please see Supplementary Data – Section S1. 2.2. Apparatus Please see Supplementary Data – Section S2. 2.3. Procedure 2.3.1. Identification of volatile components by dynamic headspace and gas chromatography–mass spectrometry (DHS–GC–MS) Samples of bituminous materials (ca. 0.1 g weighed accurately on an analytical balance to convert GC peak areas per unit mass of bitumen) were placed in 22-mL headspace vials, which were then capped with caps with a PTFE-lined silicone septa using a crimper. Two fused silica capillaries were then introduced through the septum – one of them fed helium purging the bitumen sample (1 cm above the bituminous material – the intended position of the capillary relative to the vial bottom was marked with a marker on the outer wall of the vial prior to the experiment) while the other one transported the gas with the analytes to a sorbent trap (please see Supplementary Data – Fig. S1). The sample was placed in a heating block and thermostatted for 30 min at 180 °C. Next, the flow of the purge gas was turned on and the volatile analytes released from bitumen were passed to the sorbent trap maintained at 30 °C. The purging was carried out for 11 min. During thermal desorption of the analytes trap was initially heated to 260 °C, and the analyte desorption from the trap was performed at 270 °C for 4 min. The desorbed analytes were passed through a fused silica transfer line heated to 200 °C directly to the gas chromatograph. For conditions of separation and detection please refer to Supplementary Data – Section S.3.1.
G. Boczkaj et al. / Chemosphere 107 (2014) 23–30
2.3.2. Identification of volatile sulfur compounds using static headspace and gas chromatography with pulsed flame photometric detector (SHS–GC-PFPD) Samples of bituminous materials (ca. 0.1 g weighed accurately on an analytical balance to convert GC peak areas per unit mass of bitumen) were thermostatted in 22-mL capped headspace vials at 180 °C for 30 min. Next, 0.5-mL headspace samples were collected using a heated gastight syringe. The samples were immediately introduced into the injection port of gas chromatograph. For conditions of separation, detection and quantification procedure please refer Supplementary Data – Section S.3.2. 2.3.3. Identification of volatile nitrogen compounds by static headspace and gas chromatography with nitrogen-phosphorus detector (SHS–GC-NPD) Samples of bituminous materials (ca. 0.1 g weighed accurately on an analytical balance to convert GC peak areas per unit mass of bitumen) were thermostatted in 22-mL capped headspace vials at 180 °C for 30 min. Next, 0.5-mL headspace samples were collected from the vials using a heated gastight syringe. The samples were immediately injected into the injection port of gas chromatograph. For conditions of separation, detection and quantification procedure please refer to Supplementary Data – Section S.3.3. 2.3.4. Determination of total content of volatile organic compounds by static headspace and gas chromatography with flame ionization detector (SHS–GC-FID) Samples of bituminous materials (ca. 0.1 g weighed accurately on an analytical balance to convert GC peak areas per unit mass of bitumen) were thermostatted in 22-mL capped headspace vials at 180 °C for 30 min. Next, 0.5-mL headspace samples were collected from the vials using a heated gastight syringe. The samples were immediately injected into the injection port of gas chromatograph. For conditions of separation, detection and quantification procedure please refer to Supplementary Data – Section S.3.4. 2.3.5. Analytical characteristics of the procedure for the determination of total VOCs Please see Supplementary Data – Section S.3.5. 2.3.6. Comparison of emission profiles of the investigated bituminous materials using DHS–GC–MS DHS–GC–MS chromatograms recorded in SCAN mode were used for comparison purposes. Individual chromatographic peaks were identified by comparison of their mass spectra with those in the NIST and Wiley libraries. Two characteristic mass-to-charge ratio values were selected for each compound (m/zid and m/zint). This formed the basis for VOC identification. Integration of chromatographic peaks was based on the detector signal counted only for the selected ions for each compounds (m/zint). The results were compared for the investigated bitumen samples. 3. Results and discussion Literature search revealed that the information on volatile chemical compounds present in bitumen fumes is scarce. One of the objectives of the present work was to identify as many volatile chemical compounds released from hot bitumens with a varying degree of oxidation as possible, as well as the development of a procedure allowing comparison of emission profiles for various bituminous materials. Among heavy hydrocarbon compounds present in residue from vacuum distillation of petroleum there occur, in amounts up to several percent, compounds containing heteroatoms – sulfur
25
(mostly dibenzothiophene substituted with aliphatic and alicyclic groups as well as other polycyclic aromatic hydrocarbons containing sulfur atoms), nitrogen (benzopyrrole and dibenzopyrrole substituted with aliphatic and alicyclic groups as well as compounds with a larger number of aromatic rings), and oxygen (analogous to sulfur compounds structures of dibenzofuran type and compounds with a higher molar mass). During thermal cracking of vacuum residue, unsaturated compounds (olefins) and aromatic compounds, including BTEX and PAHs (especially pyrene and chrysene) are formed. Thermal decomposition of compounds containing sulfur and nitrogen results in formation of volatile organic compounds containing these heteroatoms as well as inorganic compounds, i.e. hydrogen sulfide and ammonia. Among VSCs, the presence of thiols, dithiols, sulfides, disulfides, and thiophene derivatives can be expected (Boczkaj et al., 2010). Thus far, the presence of hydrogen sulfide as well as CO and COS in bitumen fumes has been confirmed. Thermal decomposition of nitrogen-containing compounds can lead to the formation of aliphatic and aromatic amines, including environmentally important pyridine and its derivatives as well as ammonia. The presence of oxygen-containing VOCs results in part from thermal decomposition of high-molecular-weight oxygen-containing compounds (aliphatic ethers and furan, dioxane and their derivatives), but primarily from oxidation of unsaturated compounds formed during thermal cracking, which leads to formation of mostly ketones, alcohols, aldehydes, and carboxylic acids.
3.1. Identification of VOCs using DHS–GC–MS Fumes emitted from bitumens have very complex chemical composition. A variety of volatile chemical compounds released from hot bituminous binders often results in coelution of analytes if the separation system has an insufficient resolution. This problem has been largely eliminated by proper selection of the stationary phase and the use of 60-m high-resolution capillary columns in GC–MS. To ensure adequate detection sensitivity for VOCs, the dynamic headspace analysis was used in this work. Since significant emission of VOCs from bituminous materials takes place only upon heating, the dynamic headspace system was modified to allow trapping of the analytes released from samples thermostatted at an elevated temperature. Examples of chromatograms of fumes of vacuum residue are illustrated in Figs. S2 and S3 (Supplementary Data), while Figs. S4–S9 show chromatograms of fumes of oxidized bitumens with an increasing degree of oxidation 160/220, 50/70 and 35/50 (most oxidized bitumen among the investigated materials) recorded in the total ion current (TIC) mode using the DHS–GC–MS procedure. To facilitate interpretation, each chromatogram was divided into two 15-min segments. Inspection of the chromatograms reveals that bitumen 35/50 in contrast with vacuum residue emits at an elevated temperature a wide range of VOCs absent originally from vacuum residue, mostly ketones, aldehydes, aromatic compounds (including BTEX), as well as alkanes, alkenes, and cycloalkanes. Bitumens with an intermediate degree of oxidation are characterized by lower emission of VOCs. Some identified VOCs occur in fumes of all of the investigated bituminous materials: these are saturated hydrocarbons. The content of aromatic compounds as well as individual aldehydes and ketones increases with an increase of degree of conversion of vacuum residue. This is associated with a longer residence of the charge in oxidation reactors, and hence with a longer period of thermal cracking, which also takes place during bitumen oxidation. The content of volatile compounds in the binder depends on the degree of bitumen oxidation and the technology of production and preparation of the final product, which can be composed of only the charge oxidized to yield
26
G. Boczkaj et al. / Chemosphere 107 (2014) 23–30
the product of desired properties, or a blend of more and less oxidized bitumens and raw vacuum residue. A broad spectrum of volatile chemical compounds emitted from hot bitumens makes emission profiles a preferred method of comparison of emissions from various bituminous materials. In this work, VOC emission profiles of four types of bituminous binders – raw vacuum residue and three oxidized bitumens, were compared. Chemical compounds compiled in Table 1 were identified based on their retention times assigned to the analytes previously identified by GC–MS, detected in various types of bitumens. The ratio of m/z (mass-to-charge) values for two characteristic ions for each analyte (marked in the table as m/zint and m/zid) was used to confirm identification. Integration of chromatographic peaks was performed for the signal recorded for the selected ion – m/zint. Volatile organic compounds listed in Table 1 were classified on the basis of their chemical structure into nine groups of chemical compounds. The VOC emission profiles determined in this way are compared in Fig. 1. It should be noted that these emission
profiles do not contain VNCs or VSCs, as these classes of compounds could not be identified by DHS–GC–MS. Thus, these emission profiles reflect the composition of fumes only with respect to the analytes present at the highest concentrations. Only the use of a high-resolution system (capillary GC) and a highly selective detection method allows the detection of VNCs and VSCS compounds in such a complex matrix as bitumen fumes. Oxidation of raw vacuum residue results in conversion of some aromatic compounds to resins and resins to asphaltenes. As a result of these conversions, oxidized bitumen has better operational parameters (Read and Whiteoak, 2003). Bitumen 35/50 is the material most oxidized. The final composition, and thus properties of the product, can be adjusted by blending more and less converted bitumens to meet the requirements of product quality certificate. Inspection of the data in Fig. 1 and Tables 1–3 and S3 (Supplementary Data) reveals that the VOC content in bitumens increases with the degree of oxidation of vacuum residue.
Table 1 Comparison of responses for individual VOCs in fumes released from hot bitumens. Compound
m/zid
m/zint
Bitumen type 35/50
50/70
160/220
VR
14 450 0 0 967 451 0 740 678 0 0 0 679 192 429 584 0 0 321 0 963 0 160 7487 86 0 4136 4003 663 715 715 604 1159 640 2142 1027 650 981 15 254 0 1022 1274 1549 1765 2746
0 0 0 0 0 0 320 0 0 0 0 421 0 0 655 0 0 0 0 458 1304 0 10 677 0 1177 1463 1520 0 234 234 0 0 212 364 324 406 359 5301 0 386 401 845 263 0
Peak area (arbitrary units) 2-Propanoneb Isobutyraldehyde 2-Butenal Butanal 2-Butanone Acetic acida Hexane 2-Pentenal 3-Methylbutanal 3-Ethyl-2,5-dihydrofuran Benzene 3-Hexenea 2-Metlylpentane Pentanal Heptane 2-Hexyn-1-ola 2-Ethyl-2-butenala 2-Methyl-2-pentenala 2,4-Pentadienal Toluene 3-Ethylhexane 2,4-Hexadienal Octane 1,2-Dimethylcyclohexane 1,1,2-Trimethylcyclohexane Ethylbenzene 2-Hexenal Heptanal Nonane 2,4-Dimethylheptane 6-Methyl-2-heptanone Octanal Decane Isopropylbenzene 4-Methylbenzaldehyde Nonanal Undecane 4-Ethyl-1,2-dimethylbenzene Cyclododecane Decanal Dodecane 1-Ethyl-4-(1-methylethyl)benzene Tridecane Tetradecane
43 43 69 57 57 45 57 83 71 83 78 69 56 58 71 83 83 83 53 92 84 81 71 97 111 91 83 86 99 71 110 100 85 105 120 95 85 119 111 95 113 133 71 71
58 72 70 72 72 60 86 84 86 98 77 84 57 86 100 98 98 98 82 91 85 96 85 112 126 106 98 96 128 85 128 110 142 120 119 98 156 134 140 112 170 148 85 85
45 171 1142 834 4057 2349 7939 1923 2648 480 1238 2469 1230 975 196 1534 486 528 402 2621 3887 2467 1206 12 513 761 236 10 274 6489 3581 1854 1854 2422 2780 2198 3608 3738 2436 2506 18 184 398 1578 3460 6578 2339 10 110
42 912 867 602 3099 1822 6005 2023 2422 329 720 2176 1127 333 1852 1522 367 510 405 2262 3305 2320 786 11 491 654 134 9412 6311 2842 1730 1730 1957 2102 1953 3122 224 1739 2366 18 660 0 234 2762 5123 1233 8769
m/zint and m/zid values were used to confirm identification. Integration of chromatographic peaks was performed for the signal recorded for the selected ion – m/zint. a The m/zid to m/zint ratio differed by 10–20% from the value assumed at the identification step. b The m/zid to m/zint differed by 10–30% from the value assumed at the identification step. For the remaining analytes, the differences in the m/zid–m/zint ratio did not exceed 10% of the value assumed during identification.
27
G. Boczkaj et al. / Chemosphere 107 (2014) 23–30
Fig. 1. Comparison of VOC emission profiles from hot bitumens. Due to large uncertainty of the 2-propanone peak area, this compound was not included in the above profiles.
3.2. Identification and distribution of concentrations of volatile sulfur compounds (VSCs) in bitumen fumes using SHS–GC-PFPD The presence of volatile sulfur compounds in bitumen fumes is an important problem due to their malodorousness and, in case of hydrogen sulfide, also toxicity. This paper describes an attempt at identification and determination of VSCs in bitumen fumes. Thus far, very few papers have dealt with VSCs in samples of this type. Static headspace analysis, which allows analysis of samples with complex and nonvolatile matrices without the need for time-consuming sample preparation steps, was used in the investigations. A pulsed flame photometric detector (PFPD) ensures the required sensitivity and selectivity which are necessary in case of analysis of materials with complex matrices. The results of VSC determinations are listed in Table 2 as the headspace concentration calculated on a per sulfur basis. Such an approach allows the determination of even unidentified volatile sulfur compounds on the basis of average response factor of the detector. The pulsed flame photometric detector has an almost equimolar response to sulfur; thus, determining identified VSCs and calculating the concentration of unidentified compounds the total VSC content can be estimated. However, this can only be done on a per sulfur basis since the concentration expressed in mg/m3 of individual
compounds depends on the structure of their molecules and the number of carbon atoms in a chain (e.g., thiols or sulfides). An example of SHS–GC-PFPD chromatogram of fumes of bitumen 35/50 is shown in Fig. S10. Identified volatile sulfur compounds and their headspace concentrations for the investigated bituminous materials, i.e. raw vacuum residue and bitumens 160/220, 50/70 and 35/50, are listed in Table 2. The investigations revealed the occurrence of VSCs in the fumes of oxidized bitumens and their absence in raw vacuum residue. This indicates that thermal decomposition of heavy sulfur compounds takes place primarily during oxidation of bitumens as a result of overheating of raw material on heating elements. The largest number of VSCs was detected in the most oxidized bitumen, i.e. 35/50. A smaller number of VSCs were detected in bitumens 160/220 and 50/70. Concentrations of identified VSCs are low, but for each one of them they are higher than the olfactory threshold. Total VSCs content increases from 1.2 ppm for bitumen 160/220 (least oxidized) to 68.6 ppm for bitumen 35/50. The emission, calculated as the amount of VSCs (on a per sulfur basis) released to the headspace (vial volume was 22 mL) per gram of bitumen ranged from 0.34 to 19.72 lg s/g bitumen for bitumen 160/220 and 35/50, respectively. Consequently, emission of fumes consisting solely of VSCs at the determined concentration level would constitute olfactory nuisance at the site of their emission. In the present case, the emitted fumes are a much more complex mixture, whose odor is the resultant of interactions of individual odorants with the sense of smell, which can include synergistic effects as well as partial masking of odorous effects of individual analytes. 3.3. Identification and distribution of concentrations of volatile nitrogen compounds (VNCs) in bitumen fumes using SHS–GC-NPD Similarly to VSCs discussed above, bitumen fumes were analyzed for the presence of volatile nitrogen compounds. A SHS–GC-NPD chromatogram of bitumen 35/50 is shown in Fig. S11. Volatile nitrogen compounds were detected only in the fumes of oxidized bitumens. Eight volatile nitrogen compounds were detected in the most oxidized bitumen and three of them were identified by their retention times. It follows from Fig. S11 that the identified components – pyridine and alkylpyridines – constitute main VNCs released from bitumens at elevated temperatures. Headspace concentrations of the identified VNCs in the investigated materials are compiled in Table 3. Total headspace concentration of VNCs in oxidized bitumens varies from 0.56 ppm do 5.25 ppm, which corresponds to VNCs emission from 0.24 to 4.18 lg VNCs/g bitumen.
Table 2 Comparison of measured headspace concentrations of VSCs of oxidized bitumens. Compound
Hydrogen sulfide Ethanethiol Carbon disulfide 2-Propanethiol 1-Propanethiol 3-Methylthiophene 2-Ethylthiophene Dipropyl sulfide 1,4-Butanethiol 1-Heptanethiol Sum of unidentified VSCs
Bitumen 160/220
Bitumen 50/70
Bitumen 35/50
Headspace concentration (mgS/m3) (ppm)
RSD (%)
Headspace concentration (mgS/m3) (ppm)
RSD (%)
Headspace concentration (mgS/m3) (ppm)
RSD (%)
1.12 (0.86)
Contents lists available at ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
Characteristics of volatile organic compounds emission profiles from hot road bitumens Grzegorz Boczkaj a,⇑, Andrzej Przyjazny b, Marian Kamin´ski a a b
Gdansk University of Technology, Chemical Faculty, Department of Chemical and Process Engineering, G. Narutowicza St. 11/12, 80-233 Gdansk, Poland Kettering University, 1700 University Avenue, Flint, MI 48504, USA
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
New procedure for comparison of
VOC emission profiles from hot bituminous materials. Very complex chemical composition of bitumen fumes. The content of VOC strongly depends on the degree of bitumen oxidation. Useful tools for optimization of the bitumen manufacturing processes.
a r t i c l e
i n f o
Article history: Received 15 October 2013 Received in revised form 17 February 2014 Accepted 19 February 2014
Handling Editor: I. Cousins Keywords: Asphalt GC–MS Volatile organic compounds (VOCs) Bitumen fumes NPD Pollutants
a b s t r a c t A procedure for the investigation and comparison of volatile organic compounds (VOCs) emission profiles to the atmosphere from road bitumens with various degrees of oxidation is proposed. The procedure makes use of headspace analysis and gas chromatography with universal as well as selective detection, including gas chromatography–mass spectrometry (GC–MS). The studies revealed that so-called vacuum residue, which is the main component of the charge, contains variable VOC concentrations, from trace to relatively high ones, depending on the extent of thermal cracking in the boiler of the vacuum distillation column. The VOC content in the oxidation product, so-called oxidized paving bitumen, is similarly varied. There are major differences in VOC emission profiles between vacuum residue and oxidized bitumens undergoing thermal cracking. The VOC content in oxidized bitumens, which did not undergo thermal cracking, increases with the degree of oxidation of bitumens. The studies revealed that the total VOC content increases from about 120 ppm for the raw vacuum residue to about 1900 ppm for so-called bitumen 35/ 50. The amount of volatile sulfur compounds (VSCs) in the volatile fraction of fumes of oxidized bitumens increases with the degree of oxidation of bitumen and constitutes from 0.34% to 3.66% (w/w). The contribution of volatile nitrogen compounds (VNCs) to total VOC content remains constant for the investigated types of bitumens (from 0.16 to 0.28% (w/w) of total VOCs). The results of these studies can also find use during the selection of appropriate bitumen additives to minimize their malodorousness. The obtained data append the existing knowledge on VOC emission from oxidized bitumens. They should be included in reports on the environmental impact of facilities in which hot bitumen binders are used. Ó 2014 Elsevier Ltd. All rights reserved.
⇑ Corresponding author. Tel.: +48 697970303; fax: +48 58 347 26 94. E-mail addresses: [email protected] (G. Boczkaj), [email protected] (A. Przyjazny). http://dx.doi.org/10.1016/j.chemosphere.2014.02.070 0045-6535/Ó 2014 Elsevier Ltd. All rights reserved.
24
G. Boczkaj et al. / Chemosphere 107 (2014) 23–30
1. Introduction The occurrence of volatile organic compounds in bitumen binders constitutes a serious technological and environmental problem, taking place from the moment of their formation during production until their utilization in road paving and other applications. Over 250 examples of bitumen application have been reported (Read and Whiteoak, 2003). In most cases bitumens have to be heated prior to use in order to lower viscosity enabling their application. The production of bitumens involves oxidation of residue from vacuum distillation of petroleum to obtain products with desired properties. During operations yielding the raw material for bitumen production (vacuum distillation) as well as the process of oxidation with hot air, the bitumen mass deposited on heating elements undergoes partial thermal cracking, which results in the formation of unsaturated and aromatic compounds as well as hydrogen sulfide, water vapor, carbonyl sulfide, carbon disulfide and others. These compounds undergo further transformations yielding a variety of volatile compounds: ketones, aldehydes, organic acids, phenols and their derivatives, as well as organosulfur and organonitrogen compounds. A fraction of the resulting volatile compounds is removed from the reactor with hot air, yielding so-called exhaust gases, which undergo scrubbing in a basic aqueous solution or absorption in wash oil (Boczkaj et al., 2010, 2014). VOCs remaining in the bitumen mass are emitted during storage (Deygout, 2011; Davie et al., 1993a, 1993b; Trumbore, 1999), transport (EPA, 2000a), making mineral–bitumen mixtures (EPA, 2000b; Ventura et al., 2007; Ruehl et al., 2007; Gasthauer et al., 2008; Chauhan et al., 2010; Jullien et al., 2010), roof impregnation (Rogge et al., 1997; Trumbore, 1998; Franzen and Trumbore, 2000; Kriech et al., 2004; Ruehl et al., 2007; Parker et al., 2011) and road paving (Brandt et al., 1985; Monarca et al., 1987; Greenspan et al., 1995; Hicks, 1995; Gamble et al., 1999; Burstyn and Kromhout, 2000; Burstyn et al., 2003; Posniak, 2005; Preiss et al., 2006; Hugener et al., 2010; Breuer et al., 2011; Spickenheuer et al., 2011; Raulf-Heimsoth et al., 2011). The studies on characteristics of fume emissions from hot bitumen binders under real conditions published thus far have mostly dealt with control of total parameters, i.e. total VOC content, total PAHs, particulate matter (PM), benzene soluble matter (BSM). Hydrogen sulfide emission has also been determined. Individual polycyclic aromatic compounds have been determined in bitumen fume condensate (Blomberg et al., 1999; Brandt et al., 2000; Burstyn et al., 2002; Ventura et al., 2007; Rasoulzadeh et al., 2011; Trumbore et al., 2011). A detailed investigation of the determination of unsubstituted PAHs, considered to be carcinogenic, revealed that for bitumen samples the analytical problem is very complex and requires a two-step sample preparation procedure prior to the final determination step by chromatographic analysis (Gilgenast et al., 2011). The distribution of distillation temperature of fume condensate characterizing their volatility range can be determined using simulated distillation (SIMDIS) technique (Kriech et al., 1999; Ekstrom et al., 2001; Preiss et al., 2006; Kriech et al., 2007). For hydrocarbon mixtures containing significant amounts of aromatic compounds more accurate results are obtained by empty-column gas chromatography (EC-GC) (Boczkaj et al., 2011; Boczkaj and Kamin´ski, 2013). Relatively little is known on detailed characteristics of the presence of VOCs in bitumen fumes. So far, despite information on the possibility of formation of hydrogen sulfide and volatile sulfur compounds (VSCs) during the production of oxidized bitumens (Davie et al., 1993a; Boczkaj et al., 2010; Deygout, 2011; Davie et al., 1993b), no papers are available on the identification of VSCs released at elevated temperatures from bituminous materials. The information on the presence of volatile nitrogen compounds (VNCs) in bitumen fumes is also scarce. Screening tests for VOCs
were carried out using static headspace analysis combined with solid-phase microextraction/gas chromatography/mass spectrometry (SHS–SPME–GC–MS). Using this technique, 25 polycyclic aromatic compounds (PACs) occurring in bitumen fumes were identified (Agozzino et al., 1999). Determination of individual components in very complex hydrocarbon matrices by one-dimensional gas chromatography often fails (Boczkaj et al., 2013). The SHS–SPME–GC–MS technique was also used for the identification of volatile components of asphalt release agents (Tang and Isacsson, 2005; Tang and Isacsson, 2006) and bituminous sealants (Tang et al., 2006). Emission of VOCs from hot bitumens has several consequences. One of the most important environmental problems is the direct effect of fumes on workers pouring asphalt, making and transporting asphalt mixes and road construction. Such an exposure is of a long-term nature. The effect of bitumen fumes on human health has been investigated in numerous studies. The emission of VOCs from bitumens also impacts the quality of air at the emission site. In addition, global impact of VOC emissions from bitumens should be considered: pollution of the atmosphere by volatile organic compounds and their transformation products, especially the depletion of the ozone layer and formation of secondary aerosols (Andreae and Crutzen, 1997). This paper presents a new procedure for the comparison of VOC emission profiles from hot bituminous materials. The studies enabled detailed characteristics of volatile fraction of bitumen fumes.
2. Materials and methods 2.1. Materials Please see Supplementary Data – Section S1. 2.2. Apparatus Please see Supplementary Data – Section S2. 2.3. Procedure 2.3.1. Identification of volatile components by dynamic headspace and gas chromatography–mass spectrometry (DHS–GC–MS) Samples of bituminous materials (ca. 0.1 g weighed accurately on an analytical balance to convert GC peak areas per unit mass of bitumen) were placed in 22-mL headspace vials, which were then capped with caps with a PTFE-lined silicone septa using a crimper. Two fused silica capillaries were then introduced through the septum – one of them fed helium purging the bitumen sample (1 cm above the bituminous material – the intended position of the capillary relative to the vial bottom was marked with a marker on the outer wall of the vial prior to the experiment) while the other one transported the gas with the analytes to a sorbent trap (please see Supplementary Data – Fig. S1). The sample was placed in a heating block and thermostatted for 30 min at 180 °C. Next, the flow of the purge gas was turned on and the volatile analytes released from bitumen were passed to the sorbent trap maintained at 30 °C. The purging was carried out for 11 min. During thermal desorption of the analytes trap was initially heated to 260 °C, and the analyte desorption from the trap was performed at 270 °C for 4 min. The desorbed analytes were passed through a fused silica transfer line heated to 200 °C directly to the gas chromatograph. For conditions of separation and detection please refer to Supplementary Data – Section S.3.1.
G. Boczkaj et al. / Chemosphere 107 (2014) 23–30
2.3.2. Identification of volatile sulfur compounds using static headspace and gas chromatography with pulsed flame photometric detector (SHS–GC-PFPD) Samples of bituminous materials (ca. 0.1 g weighed accurately on an analytical balance to convert GC peak areas per unit mass of bitumen) were thermostatted in 22-mL capped headspace vials at 180 °C for 30 min. Next, 0.5-mL headspace samples were collected using a heated gastight syringe. The samples were immediately introduced into the injection port of gas chromatograph. For conditions of separation, detection and quantification procedure please refer Supplementary Data – Section S.3.2. 2.3.3. Identification of volatile nitrogen compounds by static headspace and gas chromatography with nitrogen-phosphorus detector (SHS–GC-NPD) Samples of bituminous materials (ca. 0.1 g weighed accurately on an analytical balance to convert GC peak areas per unit mass of bitumen) were thermostatted in 22-mL capped headspace vials at 180 °C for 30 min. Next, 0.5-mL headspace samples were collected from the vials using a heated gastight syringe. The samples were immediately injected into the injection port of gas chromatograph. For conditions of separation, detection and quantification procedure please refer to Supplementary Data – Section S.3.3. 2.3.4. Determination of total content of volatile organic compounds by static headspace and gas chromatography with flame ionization detector (SHS–GC-FID) Samples of bituminous materials (ca. 0.1 g weighed accurately on an analytical balance to convert GC peak areas per unit mass of bitumen) were thermostatted in 22-mL capped headspace vials at 180 °C for 30 min. Next, 0.5-mL headspace samples were collected from the vials using a heated gastight syringe. The samples were immediately injected into the injection port of gas chromatograph. For conditions of separation, detection and quantification procedure please refer to Supplementary Data – Section S.3.4. 2.3.5. Analytical characteristics of the procedure for the determination of total VOCs Please see Supplementary Data – Section S.3.5. 2.3.6. Comparison of emission profiles of the investigated bituminous materials using DHS–GC–MS DHS–GC–MS chromatograms recorded in SCAN mode were used for comparison purposes. Individual chromatographic peaks were identified by comparison of their mass spectra with those in the NIST and Wiley libraries. Two characteristic mass-to-charge ratio values were selected for each compound (m/zid and m/zint). This formed the basis for VOC identification. Integration of chromatographic peaks was based on the detector signal counted only for the selected ions for each compounds (m/zint). The results were compared for the investigated bitumen samples. 3. Results and discussion Literature search revealed that the information on volatile chemical compounds present in bitumen fumes is scarce. One of the objectives of the present work was to identify as many volatile chemical compounds released from hot bitumens with a varying degree of oxidation as possible, as well as the development of a procedure allowing comparison of emission profiles for various bituminous materials. Among heavy hydrocarbon compounds present in residue from vacuum distillation of petroleum there occur, in amounts up to several percent, compounds containing heteroatoms – sulfur
25
(mostly dibenzothiophene substituted with aliphatic and alicyclic groups as well as other polycyclic aromatic hydrocarbons containing sulfur atoms), nitrogen (benzopyrrole and dibenzopyrrole substituted with aliphatic and alicyclic groups as well as compounds with a larger number of aromatic rings), and oxygen (analogous to sulfur compounds structures of dibenzofuran type and compounds with a higher molar mass). During thermal cracking of vacuum residue, unsaturated compounds (olefins) and aromatic compounds, including BTEX and PAHs (especially pyrene and chrysene) are formed. Thermal decomposition of compounds containing sulfur and nitrogen results in formation of volatile organic compounds containing these heteroatoms as well as inorganic compounds, i.e. hydrogen sulfide and ammonia. Among VSCs, the presence of thiols, dithiols, sulfides, disulfides, and thiophene derivatives can be expected (Boczkaj et al., 2010). Thus far, the presence of hydrogen sulfide as well as CO and COS in bitumen fumes has been confirmed. Thermal decomposition of nitrogen-containing compounds can lead to the formation of aliphatic and aromatic amines, including environmentally important pyridine and its derivatives as well as ammonia. The presence of oxygen-containing VOCs results in part from thermal decomposition of high-molecular-weight oxygen-containing compounds (aliphatic ethers and furan, dioxane and their derivatives), but primarily from oxidation of unsaturated compounds formed during thermal cracking, which leads to formation of mostly ketones, alcohols, aldehydes, and carboxylic acids.
3.1. Identification of VOCs using DHS–GC–MS Fumes emitted from bitumens have very complex chemical composition. A variety of volatile chemical compounds released from hot bituminous binders often results in coelution of analytes if the separation system has an insufficient resolution. This problem has been largely eliminated by proper selection of the stationary phase and the use of 60-m high-resolution capillary columns in GC–MS. To ensure adequate detection sensitivity for VOCs, the dynamic headspace analysis was used in this work. Since significant emission of VOCs from bituminous materials takes place only upon heating, the dynamic headspace system was modified to allow trapping of the analytes released from samples thermostatted at an elevated temperature. Examples of chromatograms of fumes of vacuum residue are illustrated in Figs. S2 and S3 (Supplementary Data), while Figs. S4–S9 show chromatograms of fumes of oxidized bitumens with an increasing degree of oxidation 160/220, 50/70 and 35/50 (most oxidized bitumen among the investigated materials) recorded in the total ion current (TIC) mode using the DHS–GC–MS procedure. To facilitate interpretation, each chromatogram was divided into two 15-min segments. Inspection of the chromatograms reveals that bitumen 35/50 in contrast with vacuum residue emits at an elevated temperature a wide range of VOCs absent originally from vacuum residue, mostly ketones, aldehydes, aromatic compounds (including BTEX), as well as alkanes, alkenes, and cycloalkanes. Bitumens with an intermediate degree of oxidation are characterized by lower emission of VOCs. Some identified VOCs occur in fumes of all of the investigated bituminous materials: these are saturated hydrocarbons. The content of aromatic compounds as well as individual aldehydes and ketones increases with an increase of degree of conversion of vacuum residue. This is associated with a longer residence of the charge in oxidation reactors, and hence with a longer period of thermal cracking, which also takes place during bitumen oxidation. The content of volatile compounds in the binder depends on the degree of bitumen oxidation and the technology of production and preparation of the final product, which can be composed of only the charge oxidized to yield
26
G. Boczkaj et al. / Chemosphere 107 (2014) 23–30
the product of desired properties, or a blend of more and less oxidized bitumens and raw vacuum residue. A broad spectrum of volatile chemical compounds emitted from hot bitumens makes emission profiles a preferred method of comparison of emissions from various bituminous materials. In this work, VOC emission profiles of four types of bituminous binders – raw vacuum residue and three oxidized bitumens, were compared. Chemical compounds compiled in Table 1 were identified based on their retention times assigned to the analytes previously identified by GC–MS, detected in various types of bitumens. The ratio of m/z (mass-to-charge) values for two characteristic ions for each analyte (marked in the table as m/zint and m/zid) was used to confirm identification. Integration of chromatographic peaks was performed for the signal recorded for the selected ion – m/zint. Volatile organic compounds listed in Table 1 were classified on the basis of their chemical structure into nine groups of chemical compounds. The VOC emission profiles determined in this way are compared in Fig. 1. It should be noted that these emission
profiles do not contain VNCs or VSCs, as these classes of compounds could not be identified by DHS–GC–MS. Thus, these emission profiles reflect the composition of fumes only with respect to the analytes present at the highest concentrations. Only the use of a high-resolution system (capillary GC) and a highly selective detection method allows the detection of VNCs and VSCS compounds in such a complex matrix as bitumen fumes. Oxidation of raw vacuum residue results in conversion of some aromatic compounds to resins and resins to asphaltenes. As a result of these conversions, oxidized bitumen has better operational parameters (Read and Whiteoak, 2003). Bitumen 35/50 is the material most oxidized. The final composition, and thus properties of the product, can be adjusted by blending more and less converted bitumens to meet the requirements of product quality certificate. Inspection of the data in Fig. 1 and Tables 1–3 and S3 (Supplementary Data) reveals that the VOC content in bitumens increases with the degree of oxidation of vacuum residue.
Table 1 Comparison of responses for individual VOCs in fumes released from hot bitumens. Compound
m/zid
m/zint
Bitumen type 35/50
50/70
160/220
VR
14 450 0 0 967 451 0 740 678 0 0 0 679 192 429 584 0 0 321 0 963 0 160 7487 86 0 4136 4003 663 715 715 604 1159 640 2142 1027 650 981 15 254 0 1022 1274 1549 1765 2746
0 0 0 0 0 0 320 0 0 0 0 421 0 0 655 0 0 0 0 458 1304 0 10 677 0 1177 1463 1520 0 234 234 0 0 212 364 324 406 359 5301 0 386 401 845 263 0
Peak area (arbitrary units) 2-Propanoneb Isobutyraldehyde 2-Butenal Butanal 2-Butanone Acetic acida Hexane 2-Pentenal 3-Methylbutanal 3-Ethyl-2,5-dihydrofuran Benzene 3-Hexenea 2-Metlylpentane Pentanal Heptane 2-Hexyn-1-ola 2-Ethyl-2-butenala 2-Methyl-2-pentenala 2,4-Pentadienal Toluene 3-Ethylhexane 2,4-Hexadienal Octane 1,2-Dimethylcyclohexane 1,1,2-Trimethylcyclohexane Ethylbenzene 2-Hexenal Heptanal Nonane 2,4-Dimethylheptane 6-Methyl-2-heptanone Octanal Decane Isopropylbenzene 4-Methylbenzaldehyde Nonanal Undecane 4-Ethyl-1,2-dimethylbenzene Cyclododecane Decanal Dodecane 1-Ethyl-4-(1-methylethyl)benzene Tridecane Tetradecane
43 43 69 57 57 45 57 83 71 83 78 69 56 58 71 83 83 83 53 92 84 81 71 97 111 91 83 86 99 71 110 100 85 105 120 95 85 119 111 95 113 133 71 71
58 72 70 72 72 60 86 84 86 98 77 84 57 86 100 98 98 98 82 91 85 96 85 112 126 106 98 96 128 85 128 110 142 120 119 98 156 134 140 112 170 148 85 85
45 171 1142 834 4057 2349 7939 1923 2648 480 1238 2469 1230 975 196 1534 486 528 402 2621 3887 2467 1206 12 513 761 236 10 274 6489 3581 1854 1854 2422 2780 2198 3608 3738 2436 2506 18 184 398 1578 3460 6578 2339 10 110
42 912 867 602 3099 1822 6005 2023 2422 329 720 2176 1127 333 1852 1522 367 510 405 2262 3305 2320 786 11 491 654 134 9412 6311 2842 1730 1730 1957 2102 1953 3122 224 1739 2366 18 660 0 234 2762 5123 1233 8769
m/zint and m/zid values were used to confirm identification. Integration of chromatographic peaks was performed for the signal recorded for the selected ion – m/zint. a The m/zid to m/zint ratio differed by 10–20% from the value assumed at the identification step. b The m/zid to m/zint differed by 10–30% from the value assumed at the identification step. For the remaining analytes, the differences in the m/zid–m/zint ratio did not exceed 10% of the value assumed during identification.
27
G. Boczkaj et al. / Chemosphere 107 (2014) 23–30
Fig. 1. Comparison of VOC emission profiles from hot bitumens. Due to large uncertainty of the 2-propanone peak area, this compound was not included in the above profiles.
3.2. Identification and distribution of concentrations of volatile sulfur compounds (VSCs) in bitumen fumes using SHS–GC-PFPD The presence of volatile sulfur compounds in bitumen fumes is an important problem due to their malodorousness and, in case of hydrogen sulfide, also toxicity. This paper describes an attempt at identification and determination of VSCs in bitumen fumes. Thus far, very few papers have dealt with VSCs in samples of this type. Static headspace analysis, which allows analysis of samples with complex and nonvolatile matrices without the need for time-consuming sample preparation steps, was used in the investigations. A pulsed flame photometric detector (PFPD) ensures the required sensitivity and selectivity which are necessary in case of analysis of materials with complex matrices. The results of VSC determinations are listed in Table 2 as the headspace concentration calculated on a per sulfur basis. Such an approach allows the determination of even unidentified volatile sulfur compounds on the basis of average response factor of the detector. The pulsed flame photometric detector has an almost equimolar response to sulfur; thus, determining identified VSCs and calculating the concentration of unidentified compounds the total VSC content can be estimated. However, this can only be done on a per sulfur basis since the concentration expressed in mg/m3 of individual
compounds depends on the structure of their molecules and the number of carbon atoms in a chain (e.g., thiols or sulfides). An example of SHS–GC-PFPD chromatogram of fumes of bitumen 35/50 is shown in Fig. S10. Identified volatile sulfur compounds and their headspace concentrations for the investigated bituminous materials, i.e. raw vacuum residue and bitumens 160/220, 50/70 and 35/50, are listed in Table 2. The investigations revealed the occurrence of VSCs in the fumes of oxidized bitumens and their absence in raw vacuum residue. This indicates that thermal decomposition of heavy sulfur compounds takes place primarily during oxidation of bitumens as a result of overheating of raw material on heating elements. The largest number of VSCs was detected in the most oxidized bitumen, i.e. 35/50. A smaller number of VSCs were detected in bitumens 160/220 and 50/70. Concentrations of identified VSCs are low, but for each one of them they are higher than the olfactory threshold. Total VSCs content increases from 1.2 ppm for bitumen 160/220 (least oxidized) to 68.6 ppm for bitumen 35/50. The emission, calculated as the amount of VSCs (on a per sulfur basis) released to the headspace (vial volume was 22 mL) per gram of bitumen ranged from 0.34 to 19.72 lg s/g bitumen for bitumen 160/220 and 35/50, respectively. Consequently, emission of fumes consisting solely of VSCs at the determined concentration level would constitute olfactory nuisance at the site of their emission. In the present case, the emitted fumes are a much more complex mixture, whose odor is the resultant of interactions of individual odorants with the sense of smell, which can include synergistic effects as well as partial masking of odorous effects of individual analytes. 3.3. Identification and distribution of concentrations of volatile nitrogen compounds (VNCs) in bitumen fumes using SHS–GC-NPD Similarly to VSCs discussed above, bitumen fumes were analyzed for the presence of volatile nitrogen compounds. A SHS–GC-NPD chromatogram of bitumen 35/50 is shown in Fig. S11. Volatile nitrogen compounds were detected only in the fumes of oxidized bitumens. Eight volatile nitrogen compounds were detected in the most oxidized bitumen and three of them were identified by their retention times. It follows from Fig. S11 that the identified components – pyridine and alkylpyridines – constitute main VNCs released from bitumens at elevated temperatures. Headspace concentrations of the identified VNCs in the investigated materials are compiled in Table 3. Total headspace concentration of VNCs in oxidized bitumens varies from 0.56 ppm do 5.25 ppm, which corresponds to VNCs emission from 0.24 to 4.18 lg VNCs/g bitumen.
Table 2 Comparison of measured headspace concentrations of VSCs of oxidized bitumens. Compound
Hydrogen sulfide Ethanethiol Carbon disulfide 2-Propanethiol 1-Propanethiol 3-Methylthiophene 2-Ethylthiophene Dipropyl sulfide 1,4-Butanethiol 1-Heptanethiol Sum of unidentified VSCs
Bitumen 160/220
Bitumen 50/70
Bitumen 35/50
Headspace concentration (mgS/m3) (ppm)
RSD (%)
Headspace concentration (mgS/m3) (ppm)
RSD (%)
Headspace concentration (mgS/m3) (ppm)
RSD (%)
1.12 (0.86)
