Abstract for Poster 15

 

 

Absorption from contaminated soil into skin and silicone rubber membranes

S.E. Deglin, D.L. Macalady, A.L. Bunge*
Colorado School of Mines, Golden, CO, United States

Background

The potential health risk from dermal exposure to contaminated soils depends on the amount absorbed and the toxicological potency. Dermal absorption measurements from soils usually are reported as the percent of the amount of chemical applied (i.e., the % absorption), which is the product of the contaminant concentration on the soil (C_soil) and the amount of soil on the skin (M_soil /A). In actual exposures the amount of soil on skin is less than required to completely cover the exposed area. Despite this, absorption experiments are almost always conducted with enough soil to cover the skin surface with several layers. If chemical from the extra soil layers does not migrate to the skin, then the % absorption in an actual exposure could be much larger than observed in these multi-layer experiments. Thus, the goal of this study is to better understand the effect of M_soil /A on the % absorption.

Method

The permeation of 4-cyanophenol (4CP, molecular weight = 119, octanol-water partition coefficient = 40) from soil was measured through human skin or silicone rubber membranes (SRM) used as a surrogate for skin.  The SRM from Samco Silicone Products (Nuneaton, Warwickshire, UK) was about 360 mm in thickness.  Split-thickness, human skin collected and frozen within 24 h of death was acquired from NDRI (Philadelphia, PA) and kept frozen at < -60^oC until used.  SRM or skin was clamped into Franz-type vertical diffusion cells (PermeGear, Bethlehem, PA) and equilibrated with the receptor solution for several hours.  The diffusion area was 0.64 cm² and the average temperature of the cells was 32^oC.  The receptor solution, degassed deionized water, was introduced to the diffusion cell using a peristaltic pump (Ismatec, Cole Parmer Instrument Co., Chicago, IL). The solution from the receptor chamber was collected either continuously in 1-h intervals when the flow was constant (either 1 or 6 mL h^-1 for applications of soil and aqueous solution, respectively), or periodically by pumping 30 mL h^-1 for one minute every two hours (i.e., 0.5 mL).  The latter sampling protocol was useful when the flux was small (as it is for soil on skin).  The volume of the receptor chamber and outlet tube was determined to be < 0.2 mL. 

The soil in this study was the < 250 mm fraction of a clay loam collected from Ft. Collins, CO. which contained approximately 1 weight percent organic carbon.  This soil was contaminated with 4CP to a concentration (C_soil) of 300 mg g^-1 by mixing 30 g of soil with 5 mL of 4CP in acetone.  The acetone was then removed by evaporation.  The amount of 4CP on the soil was then confirmed by extraction into aceto­nitrile and quantification of the extract by GC-FID.¹ Following a proce­dure described elsewhere,¹ the solubility limit of 4CP in this soil, which had been aged for about 2 years, was determined to be approxi­mately 150 mg g^-1. Thus, the soil used in these experiments is supersaturated, meaning that an excess of 4CP is present. 

Experiments were conducted with the whole contaminated soil (i.e., all particles < 250 mm) or with the 38 to 63 mm fraction of the whole contaminated soil. Soil was applied in an amount (40 mg cm^-2) that produced multiple layers or in smaller amounts between 0.2 and 8 mg cm^-2.  In a few experiments, 40 mg cm^-2 of soil was applied and then dumped off leaving a thin layer of particles that adhered to the skin or SRM. The mass loading of this adhering layer was approximately 1 to 2 mg cm^-2. Often, after a specified exposure time, the soil was removed, the membrane was rinsed with de-ionized water and then covered with a saturated aqueous solution of 4CP. After this, the experiment was continued for another 6 or 8 h.  For each diffusion cell, the flux of chemical through the skin or SRM was monitored by HPLC analysis of the receptor solution collected over time. 

Results

Silicone Rubber Membrane (SRM)

Early in the SRM experiments, the fluxes from the saturated solution and from super­saturated soil are similar, as we might expect since the thermodynamic driving force for transfer through the SRM are the same.  However, the flux from soil decreased in time, unlike the saturated solution, for which flux was constant.  This decrease was even larger for the adhering layer (Fig. 1).

 

After 12 h, about 67% of the 4CP from the adhering layer had penetrated the SRM. In the experiment with 40 mg cm^-2 of applied soil, there was a 5 to 10-fold decrease over the 12 h exposure even though only about 6% of the chemical had penetrated the SRM. This suggests that there was limited transfer of chemical from soil particles that do not have direct contact with SRM. 

In a similar experiment, 40 mg cm^-2 of soil was added to the adher­ing layer after 15 h and the diffusion cells that originally had 40 mg cm^-2 of soil were shaken at 7, 10 and 15 h (Fig 2). The addition of soil to the adhering layer increased the flux almost to the value early in the experiment. A similar, but less drama­tic response occurred when the mul­tiple layers of the 40 mg cm^-2 appli­cation was redistributed by shaking.

Fig. 3 shows the cumulative mass of 4CP that permeated through SRM in 8 hours plotted as a function of the mass of whole soil on the skin. Chemical penetration through the SRM increased with soil loading until the loading was ~8 mg cm^-2, after which further increases in soil loading had no effect.  This is consistent with the idea that increas­ing the soil loading above the amount required to completely cover the surface does not increase dermal absorption.  For the sieved soil, loading had no affect when it was larger than ~2 m cm^-2 (data not shown).

Figure 4 shows the cumulative mass of 4CP that penetrated through the SRM from the sieved soil (38-63 mm) and whole soil (< 250 mm) as a function of time and soil loading in experiments conducted on two different days.  The results for experiments repeated on both days were similar (Fig. 4a), indicating reasonable day-to-day reproducibility. As shown in Fig. 4b, the penetration rate through SRM was greater from the sieved soil fraction than from the whole soil. Penetra­tion through SRM from the sieved soil applied with a loading of 0.8 mg cm^-2 was similar to that obtained for 3.1 mg cm^-2 of the whole soil and was nearly twice as much as from 1.6 mg cm^-2 of the whole soil.  This suggests that the adhering layer consists mostly of small particles. Independent determination of 4CP concentration on the whole and sieved soil fraction showed no significant difference.  This would indicate that increased 4CP penetration from the sieved soil was not due to an increased amount of 4CP on the smaller particles.

Skin

Figure 5 shows the results form an experiment, similar to that shown in Fig. 1, in which the contaminated soil was replaced after 12 h by saturated aqueous solution. From saturated water, the flux after 8 h was approximately 50% larger through skin than through the SRM. However, much different than SRM, flux through skin from the contami­nated soil was less than 10% of that from saturated water. Because absorption into skin from the contaminated soil was small, contaminant concentration on the soil changed only a little and consequently, the flux did not decrease in time as was observed in the SRM experiments. The reason for the much smaller flux from the contaminated soil compared to the saturated aqueous solution is unknown, although one possibility is that skin is heterogeneous while SRM is not.

Interestingly, the flux from saturated water reached steady state in 6 h in the SRM experiments, but continued to increase over 8 h in the skin experiments. By contrast, steady state was easily established in 4 h in experiments of flux through skin from saturated water applied after a previous application of 38-63 mm sieved pure 4CP powder (Fig. 6). It seems that the cleaning procedure leaves soil particles that are re­leased slowly from the skin surface.

Comparing Figs. 5 and 6, it is evident that the flux through skin from the contaminated soil is almost the same as from the pure powder, perhaps because this heavily contaminated soil is similar to pure powder sieved to the same particle size. It is important to stress that flux through skin from the contami­nated soil has a much greater variability than was observed in the SRM experiments. A 12-replicate experiment with 1.5 mg cm^-2 of the sieved soil had a coefficient of variation of 50% in the cumulative amount of 4CP that penetrated the skin in 8 h. As a result, a plot like Fig. 4 prepared for skin showed no apparent correla­tion between flux and soil loading. 

Conclusions

Based on the observations described above, there appears to be little transfer of chemical from soil particles that do not have direct contact with the membrane or skin.  Although SRM is commonly used as a surrogate for skin, absorption through skin shows some important differences compared to SRM. For SRM, the flux from contaminated soil is almost the same as from the saturated aqueous solution and pure powder, until absorption into the SRM reduces the concentration of chemical in the soil layer that is in contact with the SRM. Bringing fresh soil into contact the SRM surface increases the flux to its earlier value. For skin, the flux from contaminated soil is only about 10% of that from saturated solution, which is too small to cause the flux to decrease during the experimental times studied. For whole soil on the SRM studied, absorption increased to a maximum value at a soil loading of about 6 mg cm^-2, above which absorption did not increase further. A similar relationship between soil loading and absorption could not be established because of the large variability in the skin experiments. 

We acknowledge support from the US EPA.

References

1.        C.-P. Chen, D.L. Macalady and A.L. Bunge, A New, Simplified Approach for Partition Coefficient Determinations in Soils, In preparation.

2.        W.J. Romonchuk and A.L. Bunge, Absorption of 4-Cyanophenol and Methyl Paraben from Powder and Saturated Aqueous Solution into Silicone Rubber Membranes and Human Skin, In preparation.

 

Content last modified: 15 May 2005