A Dissolved-Particulate Baseline Study of PAHs in Particulate-Laden Water

Polyaromatic hydrocarbons (PAHs) are of great environmental concern due to their carcinogenic impact. When water samples are extracted and analyzed for the PAHs, many methods require the water sample to first be filtered, such that the particulate matter is removed. Filtration is required when conventional solid-phase extraction (SPE) cartridges are used for extracting the sample because SPE cartridges will plug easily when any particulate matter is present. With liquid–liquid extraction (LLE) techniques, emulsions typically create problems that could impact the proper extraction of the suspended material adversely.

However, compounds such as polychlorinated biphenyls (PCBs) and PAHs will come out of the dissolved phase and adsorb onto the surface of the suspended particulate matter in the water column. Therefore, to determine the actual concentration of PAHs in the water sample, the "whole" water sample with the suspended particulate matter should be processed. As many water samples extracted for PAHs are surface water, river water, ground water, waste water, or water with suspended material, it is desirable to have an analytical method that can handle the "whole" water sample. SPE in a 47-mm disk format is a proven extraction method for extracting particulate-laden water samples. SPE disks provide fast flow rates with no breakthrough of analtyes. Aqueous samples of 1 L and greater can be processed quickly and effectively, providing excellent recoveries of PAH compounds at lower detection limits.

The International Organization for Standardization (ISO), a worldwide federation of national standards solicited their technical committees to develop a new ISO method for analyzing PAHs in particulate-laden water. The IWW Water Centre Institute, located in Muelheim an der Ruhr, Germany, offered to participate in the development of a method to explore the possibilities of using SPE disks for the analysis of 16 PAHs in drinking water, ground water, waste water, and surface water. The study conducted by IWW had several goals:

• Develop an SPE disk method that could handle "whole" water samples (including particulates and suspended matter).

• Due to possible partitioning of PAHs between the dissolved and particulate phase of the water column, to determine the effectiveness of the extraction technique for PAHs when the particulate matter and the SPE disk are both extracted and eluted with the SPE extracting solvent.

• Determine the total amount of particulate matter that can be handled by the SPE disk.

• Determine the best nonhalogenated solvent to be used for extracting the SPE disk.

• Determine if a method could be developed that did not require a solvent concentration step.

• Develop a fully automated SPE disk method.

Experiments were carried out using both spiked drinking water and spiked surface water coming from natural water bodies (river water). These experiments would examine the ability of "whole" water samples to be extracted and confirm if SPE would be sufficient to extract the sediment particles properly and remove any PAHs adsorbed onto the sediment particles.

The best recoveries for this work were achieved with reversed-phase SPE disks used with an automated extraction system. This combination worked well for many surface water samples spiked with natural sediments up to a level of 1000 mg/L. Even with this level of suspended material, 1000-mL samples could still be processed in less than 20 min, without ever plugging a disk.

Before starting the planned work on surface water samples spiked with natural sediments, the following preliminary work was done:

• Calculate recoveries of PAH-spiked drinking water samples not containing any suspended particulate matter (without any solvent concentration step).

• Explore potential adverse effects of solvent extract concentration, that is, potential losses of target compounds.

• Determine PAH recoveries from the extraction of certified dried natural sediments (not spiked to the water sample).

Instrumentation

An SPE-DEX 4790 automated extractor system (Horizon Technology, Salem, New Hampshire) was used with C18 Speedisk SPE disks (JT Baker, Phillipsburg, New Jersey) and a model 6890 gas chromatography–mass spectrometry (GC–MS) system (Agilent Technologies, Palo Alto, California).

Method Summary

In the extraction method, 1000-mL water samples were used. The samples were spiked to a 0.10-μg/L concentration. The sample bottles were placed onto the extractor and the method was started. After 20 min, approximately 9-mL of extract per sample was collected. Some extracts were concentrated and others were not, to determine the impact of the concentration step. The sample extracts were then analyzed by GC–MS.

Results

Spiked drinking water samples: All water samples were processed using the method shown in Table I.

Drinking water samples (1 L) were spiked at a concentration of 0.1 μg/L and extracted. To maintain consistent conditions, a set water sample flow rate of 50 mL/min was maintained by adjusting the vacuum level. As one of the goals was to examine the impact of solvent evaporation, the rinse solvent volumes was adjusted so each solvent rinse delivered approximately 3 mL of solvent. With three rinses, this gave a final solvent volume of approximately 9 mL. The final volume was adjusted to 10 mL, and the extract analyzed.

Table I.

The results from these experiments (see Table II) show high recoveries for the 16 PAHs under investigation. Even naphthalene showed recoveries of 75% and a low repeatability coefficient of variation; however, minor losses could be traced back to the 7-min air-dry time of the SPE disk. A long air-dry time can lead to volatilization and oxidation of the PAH compounds.

Table II: Recovery of PAH spiked drinking water, not containing any suspended matter (starting volume: 1000ml, conc = 0.1 μg/l, conditioning and elution with acetone, final volume 10 ml, no solvent concentration).

Impact of solvent concentration: To determine if there could be substantial losses of individual analytes during the solvent concentrating process, several nonhalogenated solvents were spiked with all 16 PAHs. The concentrating process was carried out with a 10-mL sample (starting) volume spiked to a concentration level of 0.005 μg/mL using a gentle stream of nitrogen and temperatures below 20 °C. The final solvent volume was 1 mL. Both acetone and hexane containing 5% (v/v) ethyl acetate appeared to provide the best results.

Results from these experiments with a final volume of 1 mL are shown in Table III. It can be seen from these results that if concentration to 1.0 mL is necessary, better recoveries can be achieved by using hexane–5% ethyl acetate.

Table III: Recovery of concentration procedure using different solvents (starting volume: 10 ml, conc = 0.005 ug/ml, final volume 1 ml, a: acetone, b: hexane / 5% ethyl acetate).

To determine the impact on recoveries of stopping the concentration at a final volume of 5 mL rather than 1.0 mL, another set of experiments was run. Each experiment was carried out four times, and the results are shown in Table IV. The results include the repeatability coefficient of variation.

It can be seen from the results of Table III and Table IV that high recoveries will be achieved for all analytes under investigation. However, by stopping the concentration at 5 mL rather than 1.0 mL, there is a significant improvement in the recovery values. As great care is essential to avoid losses of the readily volatile naphthalene and acenaphthylene during the concentration step, eliminating the solvent-concentration step was chosen. In addition, temperatures during the concentration step should not exceed 20 °C and the final volume of the concentration step should not go below the 5-mL mark. In those cases, both solvents can be used.

Table IV: Recoveries of concentration procedure using different solvents (starting volume: 10 ml, conc = 0.005 μg/l, final volume 5 ml, a: acetone, b: hexane/5% ethyl acetate).

Extraction of certified dried natural sediments (not spiked into water sample): Table V shows the results of direct extraction of dried natural sediment by using acetone and hexane–5% ethyl acetate. The extractions were performed in vessels fitted with a magnetic stirrer, with 0.5 g sediment, and each with 5 mL of each solvent. The sediment used was EC-3 (A Lake Ontario Sediment for Toxic Organics) — National Water Research Institute, Canada 1999 certified reference values (see Table V). It can be seen from the results that both solvents lead to acceptable extraction yields. Most measured values lie exactly within deviations of EC-3 uncertainty limits, or at least within a range of 15% above or below. However, the measured amount of PAH using acetone is a bit higher than that when using hexane–5% ethyl acetate. The EC-3 natural dried sediment was later used for spiking purposes, as described in the following section.

Table V: PAH extraction yields calculated from the direct extraction of 0.5 g of the dried certified sediment of Lake Ontario (EC-3) by using 5 ml of both acetone and hexane containing 5% ethyl acetate. Deviations from the certified reference value are given as a percentage (!!: deviation within EC-3 uncertainty limits).

Extraction of surface water samples spiked with dried natural sediment (EC-3): The next set of experiments added 500 mg of sediment (EC-3) to 1000-mL surface water samples. The experiments were carried out four times, and acetone was used as the extracting solvent. The goal was to explore the use of a strong, water-soluble extracting solvent like acetone to extract sufficiently the particulate matter and the SPE disk, even when the SPE disk was still slightly wet with residual water. The results show that the selected 7-min air-dry time for the filtered disk was sufficient for sediment amounts up to 750 mg. For greater sediment loadings, a longer air-dry time would be better. This will be explained later in this article.

Table VI shows the results from a 500-mg spiking experiment of a 1000-mL surface water sample (both including and not including a solvent concentration step). The sample with no solvent concentration step had the final extract volume brought to 10 mL, while the other extract was concentrated to 5.0 mL. The results show basically very similar recovery values.

Table VI: PAH extraction yields calculated from surface water (1000 ml) spiked with 500 mg of certified sediment EC-3 of Lake Ontario. Procedure used acetone (a: without solvent concentration, b: including solvent concentration - final volume = 5 ml). PAH concentrations calculated as both water sample (μg/l) and sediment (ng/g). Deviations from the certified reference value are calculated as a percentage (!!: deviation within EC-3 uncertainty limits of Table V).

The study than examined the recovery of samples with both 250 mg and 750 mg of certified sediment added to water samples. The results are shown in Table VII, and also indicate good extraction recoveries. It also should be noted that all extracts were concentrated to 5.0 mL.

Table VII: PAH extraction yields calculated from surface water (1000 ml) spiked with 500 mg of certified sediment EC-3 of Lake Ontario. Procedure used acetone (a: without solvent concentration, b: including solvent concentration - final volume = 5 ml). PAH concentrations calculated as both water sample (μg/l) and sediment (ng/g). Deviations from the certified reference value are calculated as a percentage (!!: deviation within EC-3 uncertainty limits of Table V).

Table VIII shows the recovery values when a 1000-mL sample was extracted that contained 1000 mg of certified sediment. As can be seen with the recovery data, these values were lower than desired. It was later determined that the 7-min air-dry time, after the water sample had filtered through the SPE disk, was insufficient to remove most of the residual water from the disk surface completely. To improve these recoveries, the air-dry time should be increased, but care must be taken to ensure that not too long of a dry time is used. Due to the volatility and the likelihood of oxidizing the PAHs, the proper air-dry time should be determined.

Table VIII: PAH extraction yields calculated from surface water (1000 ml) spiked with 1000 mg of certified sediment EC-3 of Lake Ontario (insufficient disk drying time). Procedure used acetone and a solvent concentration step (final volume: 5 ml). PAH concentrations calculated as both water sample (μg/l) and sediment (ng/g). Deviations from the certified reference value are given as a percentage (!!: deviation within EC-3 uncertainty limits of Table V).

With the automated SPE system used for this work, an approach to solve this problem would be as follows: First, visually inspect the sample bottle and estimate the amount of suspended material present. If the amount appears to be less than the 1000-mg limit, then use the normal method with the 7-min air-dry time. If the amount appears to approach or be greater than 1000 mg, then an SPE method should be used that stops the extraction process after the established 7-min air-dry time.

The operator would then visually examine the SPE disk with the suspended matter retained on the SPE disk surface, and empirically determine the additional air-dry time that might be required. This new method would be downloaded and run. As an example, there could be a series of methods the operator would choose from that would have additional air-dry times; 2 min, 4 min, 6 min, and so forth. Each of these methods also would include the solvent extraction portion of the method. This type of sample handling definitely would improve the recovery of these compounds.

Conclusions

This study shows that a fully automated SPE system can process water samples containing suspended particulate matter up to 1000 mg for the analysis of PAHs. As PAHs are found in both the dissolved and particulate phase of the water sample, it is important to be able to handle "whole" water samples, such that the particulate matter in the water sample is retained on the surface of the SPE disk and extracted, along with the SPE disk. This filtration method ensures that PAHs found in the dissolved phase, and those PAHs adsorbed onto the particulate matter are extracted properly.

Acknowledgments

This work was conducted by Dr. Friedrich Werres and Peter Balsaa from the IWW Rhenish-Westfalian Institute for Water. This institute is affiliated with the University Duisburg, Muelheim an der Ruhr, Germany.