Over 700 disinfection byproducts (DBPs) have been identified, but they account for only ∼30% of total organic halogen (TOX). Extracting disinfected water is necessary to assess the overall toxicity of both known and unknown DBPs. Commonly used DBP extraction methods include liquid–liquid extraction (LLE) and solid-phase extraction (SPE), which may use either XAD resins or other polymeric sorbents. With few exceptions, DBP recoveries have not been quantified. We compared recoveries by LLE, XAD resins, and a mixture of Phenomenex Sepra SPE sorbents (hereafter SPE) for (semi-)­volatile DBPs and nonvolatile model compounds at the 1-L scale. We scaled up the three methods to extract DBPs in 10 L of chlorinated creek waters. For (semi-)­volatile DBPs, XAD resulted in lower recoveries than LLE and SPE at both 1- and 10-L scales. At the 10-L scale, recovery of certain trihalomethanes and trihalogenated haloacetic acids by XAD was negligible, while recovery of other (semi-)­volatile DBPs extracted by XAD (<30%) was lower than by SPE or LLE (30–60%). TOX recovery at the 10-L scale was generally similar by the three extraction methods. The low TOX recovery (<30%) indicates that the toxicity assessed by bioassays predominantly reflects the contribution of the nonvolatile, hydrophobic fraction of DBPs.

No abstract

Iodized table salt provides iodide that is essential for health. However, during cooking, we found that chloramine residuals in tap water can react with iodide in table salt and organic matter in pasta to form iodinated disinfection byproducts (I-DBPs). While naturally occurring iodide in source waters is known to react with chloramine and dissolved organic carbon (e.g., humic acid) during the treatment of drinking water, this is the first study to investigate I-DBP formation from cooking real food with iodized table salt and chloraminated tap water. Matrix effects from the pasta posed an analytical challenge, necessitating the development of a new method for sensitive and reproducible measurements. The optimized method utilized sample cleanup with Captiva EMR-Lipid sorbent, extraction with ethyl acetate, standard addition calibration, and analysis using gas chromatography (GC)–mass spectrometry (MS)/MS. Using this method, seven I-DBPs, including six iodo-trihalomethanes (I-THMs) and iodoacetonitrile, were detected when iodized table salt was used to cook pasta, while no I-DBPs were formed with Kosher or Himalayan salts. Total I-THM levels of 11.1 ng/g in pasta combined with cooking water were measured, with triiodomethane and chlorodiiodomethane dominant, at 6.7 and 1.3 ng/g, respectively. Calculated cytotoxicity and genotoxicity of I-THMs for the pasta with cooking water were 126- and 18-fold, respectively, compared to the corresponding chloraminated tap water. However, when the cooked pasta was separated (strained) from the pasta water, chlorodiiodomethane was the dominant I-THM, and lower levels of total I-THMs (retaining 30% of the I-THMs) and calculated toxicity were observed. This study highlights an overlooked source of exposure to toxic I-DBPs. At the same time, the formation of I-DBPs can be avoided by boiling the pasta without a lid and adding iodized salt after cooking.