C G Green Chemistry August 1999 G105 F E AT U R E ead usually enters the environment by one of four routes—primary or secondary smelting, fabrication processes or paint manufacture, disposal or discarding of unwanted lead-containing materials and combustion of coal and other fuels, especially leaded gasoline. Lead emission from industrial processing is subject to stringent regulation of acceptable emission levels and safe working practice and industrial emission is usually centred on a specific locality and so largely confined to relatively small geographical regions.Lead from automobile exhaust In the mid-1980s approximately 85% of all lead discharged into the atmosphere originated from automobile exhaust.1 Alkyl lead compounds, especially tetramethyl and tetraethyl lead, have been accepted as antiknock agents in gasoline since tetraethyl lead was recognised as an effective antiknock agent in 1921.2 The environmental health implications of lead in the atmosphere were brought into question3 as early as 1922 but, from 1923, when it first went on sale, consumption of leaded fuel soared.A low-level background concentration of lead, arising through mobilisation from lead ores, is naturally present in the environment but anthropogenic lead emission has raised atmospheric lead levels significantly since leaded fuel became widely used.Comparison of the Northern Hemisphere air lead levels with those of the Southern Hemisphere (where there is a substantially lower anthropogenic contribution) shows that the Northern Hemisphere has up to 10 times as much airborne lead as the Southern—0.05–0.20 µg Pb m–3 compared with 0.02 µg m–3.The natural atmospheric concentration of lead arising through airborne particulates and gaseous diffusion should, it is calculated,4 be as low as 5 3 10–4 µg m–3. The mean lead concentration in the earth’s crust is just 16 ppm5 and it has been estimated that, in prehistory, the airborne concentration was 4 3 10–4 µg m–3 .In North America in the mid 1980s levels in even the most remote areas were around 10 3 10–4 µg m–3 and up to 10 µg m–3 in urban areas. Although substantially higher than natural background levels this represented a decrease from the levels of the 1960s and 1970s when leaded gasoline consumption was higher. In 1967, atmospheric lead levels in large US cities averaged 10–35 µg m–3—an isolated case of ca. 70 µg m –3 was measured near Hollywood, California.6 In dust collected in Birmingham, England from 1972–75 lead was found at a concentration of ca. 970 ppm.7 The particle size strongly influences environmental lead burdens—the smallest particles contain the greatest lead concentrations. The physiological effects of lead pollution are well documented. An atmospheric lead concentration of 1 µg m–3 can produce a rise of 1.0 µg 100 cm–3 in blood and a lead concentration in the soil of 1000 ppm results in 0.6 µg 100 cm–3 increase.8 The ‘natural’ average lead content4 is 0.25 µg 100 cm–3 and the Lead at the roadside Kam Kit Lam, Gerry Ottewill, Brian Plunkett and Frank Walsh at the University of Portsmouth describe the change to roadside vegetation in southern England resulting from the switch from leaded to unleaded fuel * L *The authors gratefully acknowledge Hampshire Wildlife Trust for their permission to sample on land under their management.F E AT U R E C G G106 Green Chemistry August 1999 distances of 0 m and 80 m, respectively, from the roadside.Lead content in washed samples averaged 55% of that in the unwashed, ranging from 2.31 (±0.38) to 16.8 (±0.1) µg Pb g–1 at 0 m and 80 m distances, respectively.The distance profiles for washed and unwashed samples ‘acceptable’ blood lead level,8 70–80 µg 100 cm–3 in adults. For children1 25 µg 100 cm–3 is considered the upper limit ‘normal’. Over the period 1976–1980, as use of leaded fuel fell, mean blood levels began to decrease.5 Consumption of leaded gasoline began to decline only when legislation was passed in the US enforcing the reduction of exhaust emissions—50 years after we started to use it! The capital cost to the US refineries to produce 100% lead free petrol was almost 15 000 million U.S.dollars,9 but by 1985, unleaded fuel had the major market share, 78% of the total compared with only 25% in 1977.We have looked at lead distribution on the vegetation alongside a length of the A27 major roadway in three different years, 1978, 1984 and 1994, measuring the lead concentrations and examining their variation over time, distance from the road and vegetation type.† Lead in grass samples Figure 1 shows the lead content of grass samples as a function of distance from the roadside.The lead content in unwashed grass samples collected in 1978 is shown in Figure 1(a). It ranged from 77 µg Pb g–1 to 45 µg Pb g–1 at distances of 37 m and 50 m, respectively, with distinct maxima and minima at distances of approximately 20 and 50 m. The distance profile for the 1984 samples [Figure 1(b)] shows a more linear pattern falling from 160 µg Pb g–1 at 10 m to 15 µg Pb g–1 at a distance of 60 m.Results for the 1994 analyses, illustrated in Figure 1(c) are reported for both washed and unwashed samples. The lead in unwashed grass had a concentration range from 5.22 (±1.05) µg Pb g–1 to 35.0 (±0.9) µg Pb g–1 at both fall from a maximum at the road edge to a shallow minimum at 20 m. A second maximum occurs at about 40 m where the concentration is almost as high as at the roadside.From 40 m to 80 m overall the lead content drops. The unwashed sample profile has a much Figure 1. Lead content of grass samples as a function of distance from the side of the A27 roadway: (a) 1978 data; (b) 1984 data; (c) 1994 data. † The sampling sites lay beside a busy section of the A27 trunk road running through the Farlington Marshes Nature Reserve, Hampshire, England.A series of posts, 1.5 m in height, was erected in a line perpendicular to the carriageway. The posts were placed at 10 m intervals up to a total distance of 90 m from the roadside. Vegetation samples consisted of grass, weed and moss. The species of grass were, primarily, False-brome (Brachypodium sylvaticum), foxtail (Alopecurus geniculatus) and Yorkshire fog (Holcus lanatus).Weed samples were silverweed (Potentilla anserina); the moss was Sphagnum acubifolium aggregate.10 All samples were air-dried and then oven-dried for 36 h at 110 ºC. The dried material was ground into a powder and portions of the powdered vegetation (0.5–1.5 g) were digested for 2 h in a 4:1 v/v mixture of aqueous Analar nitric and perchloric acids (25 cm3).The digest was filtered and diluted twofold using 1% nitric acid. Samples were analysed, in triplicate, using flame atomic absorption spectroscopy. The lower detection limit was 0.02 µg Pb cm–3 with a concentration range linear up to 20 µg Pb cm–3 .Green Chemistry August 1999 G107 C G F E AT U R E shallower maxima and minima, particularly evident around 40 m.The decrease in the lead content at 40 m is around 70% with washing. At 60 m the reduction due to washing is much less— only about 2%. Lead in silverweed samples Figure 2 shows distance profiles for washed and unwashed samples of silverweed collected in 1994. Concentrations ranged from 1.7 (±1.1) µg Pb g–1 at a 20 m distance to 6.4 (±2.0) µg Pb g–1 at 80 m in unwashed silverweed and from 0.83 (±0.21) µg Pb g–1 to 4.7 (±1.3) µg Pb g–1 at corresponding distances in washed silverweed samples.As for grass, washing produced a reduction in lead content but a much smaller reduction than in the silverweed. Washed samples of silver weed were found to contain approximately 70% of the lead content of unwashed ones, a higher percentage than the 55% that remained in the washed grass samples.Washing, which removes particulate lead effectively, appears to have a much greater effect on grass than on silverweed. A multiple t test (8 degrees of freedom; 95% confidence level) confirms a statistically significant difference in lead content between washed and unwashed samples of grass (texpt = 7.96; tcrit = 2.31).A similar statistical test (7 degrees of freedom; 95% confidence level), on results from silverweed samples, indicates no statistically significant difference between washed and unwashed specimens (texpt = 0.68; tcrit = 2.37). Profiles for washed and unwashed samples follow similar overall shapes falling from a maximum at the roadside to an overall minimum at ca. 80 m. Within this, a smaller maximum appears at 40 m.As for grass samples, the greatest difference between washed and unwashed sample lead content (2.1 µg Pb g–1, 39%) occurs at 40 m. At a distance of 70 m from the road no difference at all is discernible. At the sampling site hedgerows lined the roadside. The shelter provided by these explains the reduction in grass lead content that occurs in the initial 20 m from the road.Interestingly, the drop in weed lead concentration is delayed, occurring between 20 and 30 m, probably due to shielding by higher grasses in the open field. The maxima at around 40 m are due to the dispersion pattern of automobile exhaust. Sunlight and traffic heat the air immediately above the road surface. The warmed air rises carrying automobile exhaust emission which is dispersed along a line between the vertical and horizontal planes.Thermal currents, wind speed and direction also influence the dispersion and the fact that the road here is approximately 5 m above the surrounding field is probably a contributory factor. The degree of lead removal is closely related to the surface texture of the leaves. A rougher surface, such as that of the silverweed, can retain particulate material better than the smooth surface of the blades of grass and the rougher leaf holds the deposited material more tenaciously when washed.The relationship we observed between lead retention and leaf texture agrees with results found by other authors.11 It was shown,11 using radioactively labelled petrol, that rough or hairy leaves such as white poplar retain up to 8 times as much lead as smooth ones like laurel.Hazel, oak, birch and ash trees were examined at distances up to 50 m from the edge of the M25. The lead concentration decreased exponentially for hairy, retentive hazel leaves but smoother oak leaves had lower surface lead levels that varied differently with distance. Leaves closest to the motorway showed lower levels than those more remote, possibly due to exposure to the adverse weather as Figure 2.Lead content of weed samples as a function of distance from the side of the A27 roadway (1994 data).well as to their lower retentive capacity. Our results indicate that, on average, the grass at a given distance from the road has a greater deposited lead content than silverweed at an equivalent distance.The relative concentrations in grass and silverweed result from a combination of two conflicting effects, surface texture and duration of exposure to exhaust emission. The rougher texture of the silverweed leaf allows adhesion of particulates and so tends to raise the lead levels in silverweed relative to those in grass, but the grass is present throughout the year. Its persistent exposure to the automobile emissions compared with the seasonal exposure of the silverweed results, overall, in accumulation of more lead particulates, i.e.the length of exposure outweighs the differences due to surface roughness. The balance of these two effects needs close monitoring, allowing for seasonal trends in automobile use and emission dispersion and seasonal plant growth patterns with analysis of foliage at different stages of development. The age of foliage influences the degree of metal retention.12,13 Marked seasonal variation in retained lead has been observed,14 with a maximum in autumn when leaves have been exposed for the longest time.The longer exposure clearly outweighs the growth dilution effect.Other factors also influence lead uptake by vegetation. Different portions of a plant show very different levels of lead retention. In coniferous forest,15 twigs retain more lead (approximately 28 mg kg–1) than tree bark (23 mg kg–1) which, in turn, retains more than the foliage (3 mg kg–1) and the highest concentration has been found in tree roots.16 Metals uptake by plants also depends on the nature of the soil and the height of the growth site.Lead concentrations in clouds and fog are usually higher than concentrations in rain and so, since high elevation areas usually have more precipitation and high rates of interception of cloud water, lead deposition may increase at high elevations. Lead in moss Virtually all particulate matter suspended in the air has a diameter of less than 20 µm.Metals suspended in this way can be deposited on the ground surface by rain out or wash out or simply settling under gravity. Alternatively, they may be intercepted by vegetation, by soil or by a water surface. Sphagnum moss obtains mineral nutrients from the air and can, therefore, be used to monitor airborne metal levels. A study of heavy metal retention in moss17 showed that lead and copper are retained more strongly than Ni, Co, Zn or Mn, a characteristic that means moss bags can be exploited as a natural field gauge to measure the deposition of lead to ground vegetation.We studied the relationship between lead deposition rate and distance from the roadside using bags of sphagnum moss placed at regular intervals over a 90 m length by the A27 in Hampshire.Figures 3(a), (b) and (c) show plots of mean lead deposition rate on moss [mean lead content (mg Pb y–1) divided by the exposure area (m2)] as a function of distance. The 1978 data [Figure 3(a)] show an overall decrease from 217 to 75 mg Pb m–2 y–1 with a local minimum at 23 m where the rate falls to 175 mg Pb m–2 y–1, and a local maximum of 209 mg Pb m–2 y–1 at 37 m.Figure 3. Lead deposition rate on moss samples as a function of distance from the side of the A27 roadway: (a) 1978 data; (b) 1984 data; (c) 1994 data. F E AT U R E C G G108 Green Chemistry August 1999C G F E AT U R E Green Chemistry August 1999 G109 1984 samples [Figure 3(b)] show a similar pattern ranging from 1025 mg Pb m–2 y–1 to 175 mg Pb m–2 y–1 with a local maximum of ca. 560 mg Pb m–2 y–1 and maximum of ca. 735 mg Pb m–2 y–1 at 20 and 35 m, respectively. In the 1994 analyses [Figure 3(c)] the deposition rate initially rises with distance reaching a broad maximum at a distance range of 30–50 m. The extremes of lead deposition rate are at distances of 50 m (a maximum) 300±70 mg Pb m–2 y–1 and 80 m (a minimum) ±25 mg Pb m–2 y–1.Beyond 50 m the deposition rate falls rapidly remaining approximately steady at 165±15 mg Pb m–2 y–1 by about 60 m. In other investigations18–20 the lead deposition rate in air decreased exponentially with distance, distinctly different from the rising and falling concentration we observed. We believe this is a consequence of the environment of the sampling site. A steep slope runs from the road surface down to the sampling posts so that vegetative growth on the slope and the boundary fence offer some screening.The second important factor is the upward dispersal of pollutants. Airborne lead concentration will be underestimated unless deposition rates take into account the small proportion of emitted lead that deposits in the immediate vicinity of the roadway. References 1 R.L. Boeckx, Anal. Chem., 58, 274A. 2 J. O. Nriagu, Sci. Total Environ, 1990, 92, 13. 3 D. Rosner and G. Markowitz, Amer. J. Public Health, 75, 344. 4 Department of the Environment Central Unit in Environmental Pollution, Lead in the environment and its significance to man: A report of an interdepartmental working group on heavy metals pollution Paper No. 2, HMSO, 1974. 5 V. M. Goldschmidt, Ind. Eng. Chem., 1935, 27, 110. 6 V. J. Kopinski and J. B. Upham, Arch. Environ. Health, 1967, 14, 589. 7 A. Archer and R. S. Barratt, Sci. Total Environ., 1976, 6, 275. 8 D. Turner, Chem. Br., 1980, 312. 9 P. L. Dartnell, Chem. Br., 1980, 308. 10 W. Keble-Martin, New Concise British Flora, Bloomsbury Books, London, 1982. 11 P. Little and R. D.Wiffen, Atmos. Environment, 1997, 11, 437. 12 G. H. Heichel and L. Hankin., J. Air Pollut. Control Ass., 1976, 26, 767. 13 F. Valerio, C. Brescianini, S. Lastraidi and S. Coccia, Int. J. Environ. Anal., Chem., 1993, 53, 1. 14 S. T. Y. Tong, Environ. Int., 1991, 17, 31. 15 A. J. Friedland and A. H. Johnson, J. Environ. Qual., 1985, 14, 322. 16 R. S. Turner, A. H. Johnson and D. Wang, Environ.Qual., 1985, 14, 305. 17 A. Ruhling and G. Tyler, Oikos, 1970, 21, 92. 18 R. O. McLean and B. Shields, Environ. Pollut., 1997, 14, 267. 19 N. I. Ward, R. D. Reeves and R. R. Brooks, Environ. Pollut., 1975, 9, 243. 20 A. C. Chamberlain, M. J. Heard, P. Little, D. Newton, A. C. Wells and R. D. Wiffen, Atomic Energy Research Establishment, AERE-R 9198, 1978. Conclusions Our results show, as expected, that vegetation lead levels decrease with distance from the roadside.At a distance greater than 20–30 m deposition falls approximately to background level. The 1978 and 1984 studies, made in the same area, showed as expected, that the lead content in the roadside grass increased with traffic density. The traffic volume has expanded considerably recently following upgrading of the road to an 8 lane highway but, over the same period, the use of unleaded fuel has increased and so a compensation effect was anticipated. Lead deposition is strongly dependent on the species of vegetation owing to differences in surface texture and to seasonality of growth and its interrelationship with the seasonal patterns in road traffic. Reduction in lead emissions will undoubtedly result in less environmental damage in future years but the continuing damage resulting from the lead that has already entered the environment is often underestimated, especially in terms of the lead entering the soil. The distribution of lead in the roadside soil has been studied and will be reported in another paper. Future monitoring of the site is also planned. ‘Reduction in lead emissions will reduce environmental damage—but lead already in the environment will cause continuing damage’