Ethylene dibromide in urban air

Atmospheric Enrwonmenr @ Pergamon Press Ltd

CC@-6981/78/1201~2383

Vol. 12, pp. 2383-2387. 1978. Printed in Great Britam.

ETHYLENE

DIBROMIDE

$02.00/0

IN URBAN AIR

P. LEINSTER, R. PERRYand R. J. YOUNG* Public Health and Water Resource Eng. Section, Imperial College of Science and Technology, London S.W.7., U.K. (First received 25 August 1977 and in final form 9 June 1978)

Abstract-Ethylene dibromide (EDB) has recently been reported by the National Cancer Institute U.S.A., to be a potential carcinogen. Its commercial use is predominantly as a scavenging agent for lead in petrol. A procedure has been developed for the rapid sampling of ethylene dibromide in ambient air followed by analysis using gas chromatography with an electron capture detector. Ambient levels in London air were found to be in the range 0.001-0.17 pg m-s and on a garage forecourt levels of 1.2 and 1.8 pg m-s were determined. Ethylene dibromide was also measured in car exhaust and a calculation is included relating levels of organic lead to those of ethylene dibromide.

INTRODUCTION Ethylene dibromide (EDB) and ethylene dichloride (EDC) are used mostly as petroleum additives. In this application these compounds should more correctly be known as 1,Zdibromoethane and 1,2-dichloroethane; and their function in lead-alkyl containing petrols is to act as scavengers for the lead oxide released in the combustion process. For quite some time there has been concern about the potential toxic hazards of chlorinated. organic compounds used as insecticides. More recently some degree of attention has been transferred to the lower molecular weight organohalogen species which are produced and used industrially on a vast scale. McConnell et al. (1975) have considered the problems associated with the chloroderivatives of methane, ethane and ethylene which were found to be widely distributed in the atmosphere at a background level in the ppb range (1 part in 109). In the report of a “workshop” held in the U.S.A. (Rasmussen et al., 1976 and references therein) the authors concluded that there are currently four practical methods for the measurement of environmental concentrations of the halogenated hydrocarbons. These are: (1) Gas chromatography with an electron capture detector. (2) Gas chromatography coupled with mass spectrometry. (3) Long path-length infra-red absorption spectroscopy. (4) Infra-red solar spectroscopy. Investigations into the levels of low mol. wt brominated compounds in the atmosphere have been limited; bromoform (Pellizzari et al., 1976), methyl bromide (Grimsrud and Rasmussen, 1975) and 1,2dibromoethane (Going and Long, 1975) and (Going

l

Author to whom correspondence should be addressed.

and Spigarelli, 1976) being the only compounds detected. It is apparent that the distribution and character of potential sources of EDB and EDC are very different to those for other halogenated compounds. Particularly in urban areas the potential for localized concentrations well above background levels resulting from the accumulation of vehicles or proximity to garages needed investigation. The concentration of EDB in petrol is variable but the sum of EDB and EDC added is on a mole-to-mole basis with organic lead. A typical fuel contains 0.5 g Pb l- ’ with an atom ratio of Pb:Cl:Br of 1:2:1. In October 1974 the National Cancer Institute, U.S.A., issued a memorandum of alert on EDB having previously reported (Olson er al., 1973) experiments to determine the carcinogenicity of EDB to rats and mice. A preliminary study has been conducted by the EPA (Going and Long, 1975) in the U.S.A. to determine the levels of EDB in both urban air and in the vicinity of manufacturing plants. The levels reported in this and a more extensive U.S. NTIS report (Johns, 1976) were several orders of magnitude below the threshold limit value of 2Oppm currently applied by the American Conference of Governmental Industrial Hygienists (1976). Although EDB does not appear to be a substance posing immediate environmental hazards, there is available only limited information, upon which assessments may be based of the potential consequences of long-term low-level exposures or possible synergistic effects. Thus it is important to determine ambient levels of EDB in urban air. When heated with water EDB hydrolyses to ethylene glycol and bromoethanol. Under neutral conditions and ambient temperature the half-life of this reaction is S-10 days. EDB is resistant to atmospheric oxidation by peroxides and ozone, typically the half-life for these reactions is in excess of 100 days. EDB may also react with hydroxyl radicals and this reaction may be significantly faster than hydryolysis.

2383

P. L~INSTER,R.

2384

F%RRY

There are several differences between the U.S.A. and U.K. situations which make it desirable that levels of EDB should be determined in this country : (1) There are no evaporative or combustion emission controls in the U.K. (2) The structure of the vehicle population is different, as are driving conditions and modes of operation. Furthermore, workers in the U.S.A. stated that : “EDB is not expected to be emitted in the exhaust because it is not expected to survive the combustion process”. Preliminary studies in this laboratory have shown this assumption to be invalid.

EXPERIMENTAL

EDB was determined in the atmosphere by adsorption of the compound onto a short column of Chromosorb 102. The adsorbed material was subsequently heat desorbed and analysed by gas chromatography using electron capture detection. Heat desorption was preferred to solvent extraction for several reasons. If solvent extraction is used as a method of removal of the adsorbed material considerable sample dilution occurs. Micro-Soxhlet extractions have been used as a method of reducing the dilution effect but this technique still suffers from the loss of some of the more volatile compounds. Ambient levels of EDB necessitated total desorption of the adsorbed material into the gas chromatograph if rapid analyses were to be achieved and sensitivity optimized. Sampling tubes

Chromosorb 102 (0.35g) was packed into 4 x f in o.d. stainless steel tubes, which were purged with helium at 220°C and a flow rate of 20 ml min- ’ for 24 h. After this time the Chromosorb was repacked to overcome problems of preferential gas channelling resulting from polymer shrinkage, and purged for a further 12 h. Sampling procedure The sampling tube was connected between a glassfibre inlet filter and a Du Pont P200 constant flow sampling pump. Using this pump a known volume of air could be sampled with confidence. The sample collection rate was automatically regulated to ensure that a preset flow rate in the range 50-200 ml min- i was maintained to within 5X The flow rate setting was independent of the collector media pressure drop characteristics which was essential in field use where a variety of tubes was used, each with a different pressure drop. Air was sampled at 100 ml min-’ for one hour. Desorption and analysis were carried out by connecting sample tubes via a Carle Model 4300 valve oven into the carrier gas line of the gas chromatograph just prior to the injection port. Following the connection of the tube to the valve the whole system was warmed to 200°C whilst carrier gas flowed through a closed loop on the valve. When this temperature had been reached the carrier gas was diverted to flow through the tube. The carrier gas was allowed to flow continuously through the sample tube to achieve complete desorption of the adsorbed material. Calibration The electron capture detector was calibrated for picogram quantities of EDB by direct injection of standard solutions in spectroscopic hexane and a detection limit of 0.5 pa-/injection was established (signal to noise 2.5 : 1). Similar solutions were injected into nitrogen gas streams prior to sampling plugs and were warmed to aid volatilization; analysis of these tubes indicated high efhciencies of both trapping and recovery of EDB even at picogram levels (recovery > 95%). Calibration

and R. J.

YOUNG

curves of detector response obtained for, (i) direct injection, and (ii) recovery from sampling plugs on thermal desorptiob showed good agreement. The sampling and analytical protocol was validated for higher concentrations (> 1.2 pg rnv3) using a permeation tube oven. The permeation rate was 0.12 pg mitt-’ and the flow rate through the chamber 100 ml min-’ which was subsequently diluted to 11. min-’ giving a concentration of 12Opg m-s. Sampling flow rates were varied between 10 and 100 ml mitt-’ for periods of between 6min and 1 h. No “breakthrough” was observed unto a second sampling plug connected in series after the sampling of 20 1. at a concentration of 120 pg rnv3. A calibration curve was plotted of detector response against EDB concentration (in the range equivalent to O.OOl-12Opg rnm3) using data obtained from thermal desorption of sampling plugs “spiked” by both the permeation tube and hexane solution procedures. A maximum deviation of + 10% was observed from the line of best fit. Gas chromatography

Separation of the eluted material was carried out using a Hewlett-Packard 5750 gas chromatograph fitted with a Ni6’ electron capture detector. The column used for routine analysis was 8 ft x 4 o.d. stainless steel packed with 5% Carbowax 20 M on Gas Chrom W. (AWDMCS), the 95% argon 5% methane carrier gas tlow being 30 ml mitt-i with the column operated isothermally at 70°C. EDB was identified by comparing retention time data obtained on five different columns: Carbowax TOM,NGS, OVlOl, 0V7 and ov17. Results

Ambient concentrations were determined at three sites. Typical results of hourly mean concentration are tabulated in Table 1. The sampling sites were: (1) Exhibition Road, London S.W.7. This is a wide dual carriageway carrying approximately 1500 vehicles per hour. The sampling site was 5 m from the roadside at a height of 1 m above ground level. (2) Car Park. There is an open air car park in the centre of the campus. There was little traffic movement during the sampling period. (3) Garage Forecourt. The sampling point was 5 m downwind of the petrol pumps at a height of 1 m. Dynamometer

studies

Using a Clayton Dynamometer a Ford Cortina 16OOccwas driven under standard conditions and the exhaust analysed for various components. Both the ECE driving test cycle (regulation El5 as adopted by the EEC) and constant speed tests were used, and the results are tabulated in Tables 2 and 3 respectively. The ECE driving cycle is a repetitive test comprising four cycles from a cold start. It has been devised to represent typical driving patterns in European towns using European cars. The total raw exhaust gas from the four cycles is passed through a heat exchanger to remove most ofthe moisture and then into two large plastic bags, cycles 1 and 2 into one bag and cycles 3 and 4 into the other. For each bag the following parameters were determined; total hydrocarbons, carbon monoxide, carbon dioxide, oxides of nitrogen and EDB.

ETHYLENE

DIBROMIDE/ORGANIC

LEAD IN AIR

It was assumed that Premium grade petrol contained 0.51 g I-’ of lead, was of molecular formula CsH11.4, specific gravity 0.75, and that additives were present in the following concentrations (mole 1-l): TEL(tetraethy1 lead) 0.57 x 10m3; TML(tetramethyl

2385

Etbyiene dibromide in urban air Table 1. Ethylene dibromide in ambient air Sampling site Date (wind)

EDB Mm-‘)

Exhibition Road August 1976

Temperature (“C)

Time

28-30

11.00-12.00 12.00-13.00 13.00-14.00

4-9

10.00-11.00 11.00-12.00 12.00-13.00 13.00-14.00 14.00-15.00 15.00-16.00 16.00-17.00

4-8

lO.oo- 11.00 11.00-12.00 12.00-13.00 13.00-14.00 14.00-15.00 15.00-16.00 16.00-17.00

0.09

0.08 0.09 Exhibition Road 2 December 1976 (Wind SW O-40 km h-‘)*

0.001 0.001 0.001 0.001 0.03 0.02 0.01

Exhibition Road 3 December 1976 (Wind SW O-10 km b-r)*

0.01 0.04

0.10 ::: 0.17 0.01 Garage forecourt August 1976 Car park August 1976

1.8 1.2

30

14.00-15.00 15.00-16.00

0.02 0.05

30

15.00-16.00 14.00-l 5.00

* Data measured and supplied by the Atmospheric Physics section at Imperial College.

Table 2. Ethylene dibromide and other species produced under ECE driving test cycle conditions

(Z)

co

co2

NO,

(% v/y)

(% v/y)

(ppm)

EDB (ppb)M m-?

Cold start (l) Bag 1 Bag 2

390 300

4.24 3.10

9.80 10.25

720 760

17.2 11.6

134 90

(2) BarI 1 Bag2

410 330

4.70 3.80

10.25 10.90

760 780

19.8 13.0

154 101

290 272

3.03 3.10

11.45 11.00

840 230

10.9 10.2

85 80

Hot start Bag1 Bag 2

Table 3. Ethylene dibromide and other species produced under constant speed test conditions Vehicle speed Idle 10 30 40 50

At @C[C]w

183 231 282 295 213 172 202 mw3 = z

co (% v/v)

CO2 (% v/v)

1.12 4.46 2.08 2.60 1.12 1.05 1.34

11.15 10.00 13.00 12.30 13.00 13.00 12.30

x CC]ppb

EDB (ppb)@ugmw3) 80 200 920 1100 1820 21% 2200

9 10 8 9 5 0.3

70 78 62 70 61 2

2386

P. ~?ER,

R. PERRY and R. J. YOUNG

lead) 1.89 x lo-” ; EDB 1.23 x lo-’ ; and EDC 2.46 x 10-j. Similarly Regular grade petrol was considered to be of composition CsH1s.4, specific gravity 0.73, and to contain : lead 0.40 g I-’ ; TEL 1.95 x 10e3 molel-‘; TML 0.02 x lo-’ molel-‘; EDB 1.09 x 10m3 molel-‘; and EDC 2.18 x 10m3 molel-‘. The bulk usage of petrol in the U.K. was taken to be 70% Premium and 30% Regular. The petrol in circulation was assumed to be a mixture of 70% Premium to 30% Regular and the mole fractions of additives in such a mixture calculated :

DISCUSSION

The levels of EDB determined in London were in the same range 0.01-1.0 pg mm3 but were generally lower

than those reported by the EPA for several cities in the U.S.A. Previous work in this laboratory (Harrison et al., 1974) has included the measurement of organic lead levels at sampling sites close to those used here for the determination of EDB (Exhibition Road). The level of organic lead on hot days was found to be 0.11 pg rnv3. Application of the EDB to organic lead ratio derived Mole above (0.40 w/w) would give a predicted EDB conMixture Regular fractions Premium centration of 0.04-0.05 pg m-‘. Experimentally de 2.98 x lo-* 1.47 x 1o-4 0.85 x lo+ TEL termined levels of EDB on a hot day were 1.99 x 1o-4 2.81 x 1O-4 0.03 x 1o-4 TML 0.08XNW pg me3 ; this difference may be partially 1.77 x 1o-4 1.83 x lo-* 1.66 x 1o-4 EDB accounted for by a contribution to the EDB levels from exhaust emissions. The proximity of the measured and predicted levels generates confidence in the experimenThe vapour pressures of pure TEL, TML and EDB tal methods used; and although it would be unwise to may be derived from the expression : place undue significance on variations in the few data 0.2185a available, features, which would require statistical log,,P = - 7 + b, verification, are discernible. In general the levels of EDB determined on Exhiwhere T is the temperature in degrees absolute and a bition Road on Winter days (temp. 4-9°C) were and b are constants, the values of which were taken to considerably lower than in Summer with a mean on a be as reported in the 54th Edn. of the Handbookof relatively still day of 0.05-0.06 gg rnm3. (This figure Physics and Chemistry. It was assumed that the vapour pressures exerted by the TEL, TML and EDB frac- contains two high levels of 0.17 and O.lO~grn-~ obtained when vehicles were started up under choke tions in petrol would obey Raoult’s Law and theoreticlose to the sampling point.) The lower Winter levels cal values were determined for the mole fractions are the result of a reduction in evaporative losses obtained above at an assumed temperature of 293 K : particularly from standing vehicles (the calculated evaporation rate for EDB at 5°C is less than a third of Vapour that at 30°C). An indication of the magnitude of pressure evaporative losses from standing vehicles is given by Premium Regular Mixture (mm H8) the levels measured in the car park (0.02-0.05 pg me3). TEL 0.25 x 1o-4 0.86 x 1o-4 0.43 x 1o-4 It is also probable, however, that an opposing trend TML 64.07 x 1o-4 0.68 x 1o-4 45.37 x lo-* will be produced by the change in driving conditions in 20.68 x 1O-4 18.76 x 1O-4 20.00 x 1O-4 EDB Winter (e.g. cold starts), which will cause an increase in the EDB content of exhaust emissions (Table 2). Thus, These vapour pressures were readily converted to if the levels of EDB reaching the atmosphere are concentrations in the vapour phase and calculated considered to be produced by unburnt fuel, the relative values are recorded below. The values for TEL and contributions of evaporative and exhaust sources to TML are recorded as lead concentrations in the this total will vary with weather and traffic conditions. vapour phase to enable comparison with experimenIn Summer (high temperatures) evaporative losses will tally determined levels of organic lead : be the most significant contribution to the levels of EDB in the environment, whilst the change in mode of operation generated by the lower Winter temperatures Concentration Premium will cause an increase in the significance of exhaust Regular Mixture (m8 me31 emissions. EDB may also be present in crankcase 0.28 0.97 0.48 blow-by gases, which have been shown to contribute 72.59 0.77 51.35 up to 20% of unburned hydrocarbon emissions for 21.26 19.29 20.56 uncontrolled vehicles. The study has shown that using this technique it is possible to obtain rapid reproThus at 293 K evaporative losses of EDB, TEL and ducible results for levels of EDB. Sampling periods as TML from the 70% Premium to 30% Regular petrol short as 15 min could be used but if desired the mixture assumed to be in circulation, would be sampling time could be extended to many hours. The expected to generate an EDB:organic lead ratio of levels of EDB so far determined in ambient air do not 0.40 w/w. appear to constitute a risk to health.

Ethylene dibromide in urban air Acknowledgements - We thank the S.R.C. for a studentship (P.L.), the Associated Gctel Company Limited for the use of dynamometer facilities and Dr. T. M. Li for valuable discussion of results.

REFERENCES

Going J. E. and Long S. (1975) Sampling and analysis of selected toxic substance-s. Task II-Ethylene dibromide. EPA 560/6-75-001. Going J. E. and Spigarelli J. L. (1976) Sampling and analysis of selected toxic substances. Task IV-Ethylene dibromide. EPA 560/6-76-02 1. Grimsrud E. P. and Rasmussen R. A. (1975) Survey and analysis of halocarbons in the atmosphere by gas chromatography-mass spectrometry. Atmospheric Enoironfnent 9, 1014-1017. Harrison R. M., Perry R. and Slater D. M. (1974) An

2387

adsorption technique for the determination of organic lead in street air. Atmospheric Enuironment 8, 1187-1194. Johns R. (1976) Air-pollution assessment of ethylene dibromide. U.S.NTIS PB256 736. McConnell G., Ferguson D. M. and Pearson C. R. (1975) Chlorinated hydrocarbons in the environment. Endeouour 34, 13-18. Olson W. A., Haberman R. T., Weisburger E. K., Ward J. W. and Weisburger J. M. (1973) Induction of stomach cancer in rats and mice by halogenated aliphatic fumigants. J. Nat. Cancer Inst. 54, 1963-1965. Pellizzari E. D., Bunch T. E., Berkley R. E. and McRae J. (1976). Determination of trace hazardous organic vapour pollutants in ambient atmospheres by gas chromatography-mass spectrometry-computer. Analyt. Chem. 48,803-807. Rasmussen R. A., Pierotti D. J. and Krasnec J. (1976) Analysis of halocarbons in the atmosphere. Report of a Workshop 69th Meeting Air Pollution Control Assoc. June 27-July 1, 1976.