Dillon Lee ‘08, Curtis Hansen ‘10, Shahen Huda ‘10, Jie Xu ‘10
A wide variety of substances are currently used to disinfect the water in swimming pools, including many containing chlorine and bromine, which can react to form potentially harmful disinfectant by-products (DBPs). In order to determine the concentrations at which these occur in swimming pool water, samples were taken from a combination of public and private swimming pools in the Hanover, New Hampshire area. Samples were analyzed for the presence of two DBPs, chloroform and bromoform, by utilizing solid-phase microextraction (SPME) followed by gas chromatography-mass spectrometry (GCMS). These techniques allowed for the detection and quantification of these compounds at low concentrations. It was found that the pools contained chloroform at a maximum concentration of 1.28 ppm; only one pool contained bromoform, but its concentration could not be successfully quantified. Additionally, although the concentration of chloroform did not vary significantly based on depth, bromoform levels were significantly elevated at a depth of approximately 1.5 m compared to surface levels. The identification and quantification of these chemicals serves to inform the public about health hazards associated with these common disinfectants.
As of 2007, there were over eight million private pools in the United States alone, and their maintenance cost reached about four billion dollars (1). Though swimming pools have become a staple of backyard activity enjoyed by nearly everyone, their maintenance introduces many potential hazards. Factors that must be monitored include pH, water hardness, total alkalinity, and total dissolved solids. Even with the most careful efforts, humic and fulvic materials from hair, lotion, saliva, skin, or urine inevitably react with disinfectants. In varying degrees, these reactions form hundreds of undesirable disinfectant by-products (DBPs) such as haloforms, halogenated carboxylic acids, halofuranones, and haloketones. This experiment focuses on the formation of the most concentrated compounds, bromoform and chloroform, in the pool waters around Hanover, New Hampshire, USA (2).
Chlorine has been the main water disinfectant for many years. It is effective against waterborne illnesses like cholera and typhoid, as well as other harmful microorganisms. However, it can also form undesirable DBPs over time; some are shown in Figure 1. According to a research study conducted by Bull and Kopfler, some haloforms have been linked to cancer growth in laboratory animals (3). In response to the conclusive results from a number of studies, the U.S. Environmental Protection Agency (EPA) lowered the maximum contaminant level to 80 µg/L for total haloforms in drinking water in 2003 (4). Another federal agency, the Occupational Safety and Health Administration (OSHA) has set the exposure limit as 0.5 ppm for bromoform and 50 ppm for chloroform, with a recommendation of no more than 2 ppm of chloroform (5-6). EPA regulations in general aim to keep the number of excess cancer incidences to between 10-4 and 10-6 to provide a reasonable safeguard against cancer (7).
The target molecules of this project, bromoform and chloroform, are easily absorbed by both the body and the environment. They typically enter the environment through disposal of the disinfected water, or as vapor because of their volatility. Vapors can remain in the air with a half-life of one to two months. At the surface of the water, the haloforms decompose slowly due to the greater availability of oxygen. Under anaerobic conditions, commonly found in underground water sources, degradation of these molecules occurs more rapidly. Although these compounds are mobile in soil and seep into ground water, there is no evidence of their bioaccumulation in fish. Likewise, they easily enter the human system through inhalation, absorption, and ingestion, but 50-90% is removed within eight hours (8).
Experimental methods were primarily adapted and taken from Hardee et al. (10). First, a calibration curve was made for chloroform concentrations ranging from 15 to 40 ppm via serial dilutions with deionized water. The chloroform was originally 200 µg/L in methanol (Supelco, Bellefonte, PA). A mass of 1.75 g of sodium chloride (NaCl) was introduced to each amber vial of chloroform solution along with a magnetic stirring bar. The addition of salt enhances the partitioning of an organic compound to favor the organic phase on the fiber (10).
After conditioning a solid-phase microextraction (SPME) syringe (75-µm Carboxen/polydimethylsiloxane fiber, Supelco) for one hour at 250°C, the vapor in the headspace above the chloroform solution was extracted for a time period of 30 minutes at ambient temperature. Samples were sealed using a screw cap containing a PTFE-faced rubber septum. The vials were clamped onto a magnetic stirring plate and the SPME assembly was secured in place above the vial cap (10).
Following a 30 minute extraction period, the fiber was immediately retracted and transferred to the injection port of the gas chromatograph. Chromatographic analysis was performed using a Hewlett-Packard GCD Plus GC–MS system. The initial oven temperature was set at 40°C for 5 minutes, ramped at 7°C/minute to 150°C, and held for 2 minutes. The analytical column was an HP-5, dimethylsiloxane (30 m x 0.10 mm i.d. x 0.2 µm film thickness). Predrilled septa (Supelco) were used and samples were desorbed under splitless conditions with helium as the carrier gas. Quantification of analytes was performed by measuring the peak areas of the analyte in pool water samples with respect to the calibration experiments (10).
Samples were collected in triplicate in amber 15 mL vials, with minimal air inclusion (See Table 2 for sample locations). At all locations except Storrs Pond, Storrs Pool, Ciambra’s pool, and Dartmouth pool, samples were collected at the surface of the water, and at a depth of approximately five feet (1.5 m). At the aforementioned locations, samples were collected only at the surface. Upon collection of the samples, air temperature and surface water temperature were measured, and current weather conditions recorded.
Samples were analyzed as soon as possible after collection, from a couple of hours to a few days. When not in transit or being analyzed, chloroform standards and pool samples were stored with NaCl in 5 mL quantities at 4°C. While in transit, samples were kept in a cooler at a temperature near 0°C for a time period not exceeding 90 minutes.
To quantify the amount of chloroform found in pool water, a calibration curve was established relating absolute concentrations of chloroform to area under the GC peak. The calibration curve constructed from the standards for chloroform in deionized water is presented below.
Seven of the eight pools sampled tested positive for haloforms (six contained chloroform, and one contained bromoform). No pool contained measurable amounts of both haloforms. The concentration of chloroform found in the pool water varied from none detected to as high as 1.28 ppm. Neither of the two natural bodies of water contained any haloforms.
In all cases where data for haloform concentration at deeper water is available, average haloform concentration was greater in deeper water, compared to the surface. Only in Gribble’s pool, however, was this difference statistically significant (tcalc = 6.95, n = 4, CL = 95).
Identification of the compounds was performed using a combination of data from the gas chromatogram as well as the mass spectra. Overall, both chloroform and bromoform exhibited a characteristic mass spectrum, and chloroform was detected at approximately 4.1 minutes by gas chromatography while bromoform was detected around 13.5 minutes.
The presence of chloroform and bromoform was successfully detected at almost all expected locations. The experimental data suggests that haloforms are often found in swimming pools and do not exist in natural water sources. The amount of chloroform detected in pool water varied between levels below the detection limit of GCMS of 0.0 ppm (Dartmouth Pool sample), to a maximum average of 1.3 ppm, found at a depth of five feet in the White River Junction pool. The widespread detection of chloroform in the samples (present in seven of eight pool samples) is hypothesized to be the result of the prevalence of chlorine-based disinfectants. Bromoform was found in only one pool (the Gribble pool). No chloroform or bromoform was detected in either of the natural sources, which was expected, since these were not expected to have any chlorine- or bromine-containing disinfectant additives.
One of the prevailing trends apparent from the results is the consistently higher concentration of haloforms in the samples taken from deeper water. This was true for every pool considered (other than the Dartmouth Pool, which contained no measurable haloforms). However, this difference was only statistically significant at the Gribble Pool, which was also the only location containing bromoform. The general trend directly contradicts the original idea that haloforms should form closer to the surface of the pool because of higher water temperatures. Upon closer inspection of this hypothesis, it should be noted that the water temperature differences at the surface and at a depth of five feet would likely be very low, although no actual measurements were taken. There was greater haloform concentration observed in deeper water than in surface water at all pools regardless of the surface water temperature, which varied from 22.6°C to 31.0°C. However, it does seem likely that comparing pool surface water with water from a much greater depth may result in lesser haloform formation in deeper water, provided the appropriate reagents (organic matter and disinfectant) are both present. Under the tested conditions, alternate explanations are necessary to explain the prevailing trend throughout the data.
One of the proposed explanations for the correlation between depth and haloform concentration is the tendency for organic matter to sink in the pool, thus accumulating at the bottom. Since the haloform-forming reaction occurs between a carbon source (usually human organic matter from swimmers) and disinfectants, a higher production of the haloform occurs where the reactants are in the highest concentration. A buildup of organic matter at the bottom of the pool would naturally cause an increased concentration of haloform at a greater depth. An alternate hypothesis involves the physical properties of the two haloforms and their general volatilities and insolubility in water (5,6). Any haloform formed at the surface of the water would tend to vaporize due to the high volatility, leaving behind lower concentrations at the surface and higher concentrations in deeper water.
The depth effect may have been most pronounced in the Gribble Pool because of the presence of bromoform. Bromoform is significantly more volatile than chloroform, which could account for the statistical significance seen in this one sample. This explanation is the best fit for the existing data, accounting for differences between chloroform and bromoform. However, these explanations are merely hypotheses that were not verified. It is possible that a larger set of bromoform-containing samples would lend more support to this idea. However, the lack of statistically significant data for any of the chloroform-containing samples makes it impossible to rule out the possibility that the trend occurred by chance.
One of the important purposes of this experiment was not only to determine the presence of the various haloformsthat may be byproducts of disinfection of swimming pools, but to quantify the level at which they are present. Haloforms are considered a health hazard, so their presence has significant health ramifications. The Occupational Safety and Health Administration (OSHA) has set forth guidelines with regards to the permitted levels of each chemical in work spaces where skin contact is involved. This limit is 50 ppm for chloroform, with a recommended limit of less than 2 ppm (5,6). The analyzed samples range from no detected chloroform to 1.3 ppm, well below both limits determined by OSHA. There is a limit of 0.5 ppm for bromoform (also determined by OSHA), but the lack of a bromoform standard curve impedes the ability to form conclusions as to the safety of the observed levels of bromoform. While this data indicates that the pools are safe to swim in with regards to haloform concentration, it raises concerns regarding the effectiveness of pool disinfection. It is important to note that chemical usage in the disinfection of pools is, like many other uses of chemicals, a matter of finding the balance of dose that is enough to be effective, without causing harm to the surroundings.
In addition to the OSHA limits on haloforms for skin exposure, there are regulations on air concentration. The limit for chloroform is 240 mg/m3 while the limit for bromoform is 5 mg/m3 (5,6). In order to make an appropriate comparison between the concentration of each haloform in solution and the concentration in the air, it was necessary to use an experimentally determined Henry’s law constant for chloroform (11). It was impossible to quantify the amount of bromoform in the air because the concentration in the solvent could not be calculated. The amount of chloroform in the air above a swimming pool could be approximated, with quantities ranging from no detected chloroform (Dartmouth Pool) to 12 mg/m3 in Storrs Pool. In these calculations, only the surface concentration of chloroform was used because this is the concentration of chloroform that can vaporize since it is situated adjacent to air. These levels are well below the limit of chloroform set by OSHA, but there are no specifications for long-term exposure to these haloforms, which may be significant. Around outdoor pools, this concentration would probably decrease very quickly away from the surface of the pool, but a buildup of haloform vapors could occur in an indoor pool leading to unsafe concentrations. However, Henry’s law only provides an estimate of the amount of the gas in the air. In order to perform better quantifications, it would be necessary to sample the air and quantify the gaseous haloforms directly.
There were several limitations and difficulties over the course of this experiment, leading to potential inaccuracies in the data. First, the unpredictability of the environment made it impossible to ensure identical weather and temperature conditions between collection sites, even if the samples were collected within a short time frame. Compensating for such differences, or even predicting what they might be, would be difficult. Numerous potentially significant variables exist. For example, the number of people using the pool prior to sample collection could greatly affect the level of organic matter in the pool and thus the haloform-producing reactions. Another significant source of variability between samples was due to the logistical difficulties of sampling and analysis by GCMS. The equipment limitations allowed only a single sample to be analyzed at a time whereas multiple samples were collected in the field per sampling location. Previous studies have indicated that samples should be run within the hour that they are collected, but this proved to be impossible. Instead, samples were often stored for up to one week in the amber glass vials with sodium chloride at 4 °C before they were run. It is likely that some of the haloform in the sample was lost during storage due to the volatility of these chemicals, but it would be difficult to quantify any amount lost. Additionally, more relevant standard curves made with smaller concentrations of chloroform and bromoform would have increased accuracy, since no samples contained more than 2 ppm of chloroform, yet the standard curve of chloroform ranged from 10 – 40 ppm. In the future, a more relevant range of standards should be run – the reduced haloform concentration may allow for a more accurate calibration curve and may also alleviate the problem regarding solubility of the chemicals. Though the conditions were not exactly ideal, the qualitative nature of this experiment meant that fine control over these variables was largely unnecessary.
In conclusion, despite the health risks of haloforms in general, the amount of chloroform found in pools is generally far too low to present any danger to swimmers. Therefore, although it has been shown that disinfectant by-products do form in detectable concentrations in most pools, it appears that the benefits of disinfection far outweigh the risks posed by its by-products.
We would like to thank professors Gordon Gribble and Siobhan Milde for their valuable feedback and support for this project through their time and resources. We would also like to thank Charlie and Rita Ciambra for their constant laboratory assistance.
1. C. Salter, D. J. Langhus, Chem. Ed. 84, 1124-1128 (2007).
2. Drinking Water Disinfection By-Products. Encyclopedia of Environmental Analysis and Remediation, 1st ed; Wiley-Interscience: New York, N.Y., 1998; Vol. 3, pp. 1398-1421.
3. R. J. Bull, F. C. Kopfler, Health Effects of Disinfectants and Disinfection By-products, AWWA Research Foundation; Denver, C.O., 1991.
4. Environmental Protection Agency. National Primary Drinking Water Regulations: Disinfectants and Disinfection Byproducts; 1998; Vol. 63, No. 241, pp. 69389-69476. http://epa.gov/OGWDW/mdbp/dbpfr.html (accessed August 9, 2008).
5. U.S. Department of Health and Human Services. Occupational Safety and Health Guideline for Bromoform. [Online] 1992; p. 2. http://www.cdc.gov/niosh/pdfs/0066-rev.pdf (accessed August 9, 2008).
6. U.S. Department of Health and Human Services. Occupational Safety and Health Guideline for Chloroform. [Online] 1978; p. 1. http://www.cdc.gov/niosh/pdfs/0127.pdf (accessed August 9, 2008)
7. Chlorine Disinfects Drinking Water – What are Trihalomethanes (THMs)?. http://www.wssc.dst.md.us/service/THM_FAQs.pdf (accessed August 9, 2008).
8. Department of Health and Human Services – Agency for Toxic Substances and Disease Registry. Public Health Statement for Bromoform and Chlorodibromomethane. [Online] 2005; http://www.atsdr.cdc.gov/toxprofiles/phs130.html#bookmark01 (accessed August 12, 2008)
9. Handbook of Water Analysis; Nollet, M. L., Ed.; CRC: New York, N.Y., 2000; pp. 79-93.
10. J. R. Hardee, J. Long, J. J. Otts, Chem. Ed. 79, 633-634 (2002).
11. U.S. Environmental Protection Agency. EPA Online Tools for Site Assessment. [Online] 2006; http://www.epa.gov/athens/learn2model/part-two/onsite/esthenry.htm (accessed August 16, 2008).
This entry was posted on Friday, May 22nd, 2009 at 8:51 pm and is filed under Spring 2009. You can follow any responses to this entry through the RSS 2.0 feed. You can skip to the end and leave a response. Pinging is currently not allowed.