(First draft prepared by Dr Stuart Dobson, Institute of Terrestrial Ecology, Huntingdon, United Kingdom, and Mr Richard Cary, Health and Safety Executive, Liverpool, United Kingdom. Published under the joint sponsorship of the United Nations Environment Programme, the International Labour Organization, and the World Health Organization, and produced within the framework of the Inter-Organization Programme for the Sound Management of Chemicals.)
World Health Organization Geneva, 2002

1. Effects on humans.
Very few data are available relating to single exposures in humans. From the reports that are available (Dalhamn, 1957; Gloemme & Lundgren, 1957; Kennedy et
al., 1991; Salisbury et al., 1991; Anon, 1997), it would appear that single high-level exposures may lead to eye irritation, respiratory tract lesions, and possibly permanent impairment of lung function. However, the quality of the data available is poor, often involving mixed exposures with other irritant gases, such as chlorine or sulfur dioxide, and there is no dose–response information.
- Drinking-water studies.
As with animal studies using this route of administration, human studies using drinking-water administration are of limited value in relation to occupational considerations; the inhalation and dermal routes would be expected to be the main routes of exposure. The following studies are summarized to help complete the toxicological profile for chlorine dioxide.
In a series of extensive human volunteer studies on water disinfectants, groups of 10 males received aqueous chlorine dioxide in drinking-water by a range of different protocols (a sequence of rising concentrations of up to around 0.34 mg/kg body weight over a 16-day period, approximately 0.035 mg/kg body weight on every third day for 12 weeks, or approximately 3.6 × 10–5 mg aqueous chlorine dioxide/kg body weight per day daily for 12 weeks) (Lubbers et al., 1982, 1984; Lubbers & Bianchine, 1984). Observations included physical examination (blood pressure, respiration rate, pulse, oral temperature, and electrocardiography), extensive blood biochemistry, haematology, and urinalysis, and the subjective recording of taste. There were no significant adverse effects recorded for any of the parameters measured.
A prospective epidemiological survey was per- formed on a group of 197 people exposed to chlorine dioxide-treated drinking-water on a seasonal basis (Michael et al., 1981). Haematology and blood bio- chemistry samples were taken before and after a 12-week chlorine dioxide exposure period. Reliable quantification of exposure was almost impossible due to the difficulties associated with estimating water consumption and the rapid decay of aqueous chlorine dioxide. There were no significant changes as a result of chlorine dioxide exposure in any of the parameters recorded.
In a retrospective study, hospital records relating to the morbidity and mortality of infants born between 1940 and 1955 were studied from a community in the USA (Tuthill et al., 1982). Tap water was treated with chlorine dioxide between 1944 and 1958, and compari- sons were made with a nearby community, which, in part, used the same three hospital facilities and apparently did not receive chlorine dioxide-treated tap water. There were no clear demographic differences between the populations studied. A statistically significant increase in premature births was noted among members of the community that received chlorine dioxide-treated tap water. However, the identification of prematurity was on the basis of the physician’s assessment, there were no objective measures, and the proportion of premature births differed markedly between hospitals. There were no other significant differences in the condition of neonates between the two communities. Due to the lack of information on the extent of chlorine dioxide exposure, the uncertainties attached to the diagnoses of prematurity at the hospitals, and lack of adequate consideration of confounding factors such as smoking and socioeconomic status, no conclusions can be drawn from this study.
2. Effects on other organisms in the laboratory and field.
An EC50 for inactivation of Cryptosporidium parvum, a protozoan parasite that can infect the diges- tive tract of humans and other warm-blooded animals, was measured at 1.3 mg/litre; parasite inactivation was monitored by infectivity (Korich et al., 1990).
Spores of the giant kelp (Macrocystis pyrifera) were exposed to nominal concentrations of chlorine dioxide for 48 h at 15 °C with constant illumination by cool fluorescent lamps. A no-observed-effect concen- tration (NOEC) was determined at 2.5 mg/litre, with lowest-observed-effect concentrations (LOECs) for germination and germ tube length at 25 and 250 mg/litre, respectively (Hose et al., 1989).
Embryos of the purple sea urchin (Strongylocen- trolus purpuratus) were exposed to nominal concentra- tions of chlorine dioxide at 15 °C for 48 h. Abnormalities recorded included pre-hatch malformations, retarded development, post-hatch abnormalities, skeletal malformations, and gut malformations. The NOEC was determined at 25 mg/litre, with a LOEC for malformations at 250 mg/litre (Hose et al., 1989).
Bluegill sunfish (Lepomis macrochirus) and fathead minnow (Pimephales promelas) 96-h LC50 values were reported at 0.15 and 0.02–0.17 mg/litre, respectively. Exposure was by release of chlorine dioxide stock solutions into the test medium for approximately 1 h in each 24 h (Wilde et al., 1983).The NOEC for survival of kelp bass (Paralabrax clathratus) eggs exposed to chlorine dioxide for 48h at 20 °C without aeration was 25 mg/litre (Hose et al., 1989).
A major field incident occurred in Sweden in the early 1980s, when it was recorded that the bladderwrack (Fucus vesiculosus), the major component of brackish water communities in Sweden, had disappeared from an area of 12 km2 (Lindvall & Alm, 1983). It was subsequently demonstrated in laboratory experiments and model ecosystems that chlorate was responsible (Rosemarin et al., 1985; Lehtinen et al., 1988). It was also shown that brown algae of many species are sensitive to chlorate, with a threshold concentration at around 10–20 µg/litre for prolonged exposure (4–5 months) when exposure took place in nitrate-limited brackish water with a salinity of 0.7–0.8‰ (Rosemarin et al., 1994). A requirement to treat wastewater from pulp mills to reduce chlorate (derived from use of chlorine dioxide) to chloride has diminished the problem (Landner et al., 1995). Data on the effects of chlorine dioxide on terrestrial organisms were not available.
3. Effects evaluation.
a. Evaluation of health effects.
- Hazard identification and dose–response assessment.
Toxicokinetic data are limited, although it would seem unlikely that there would be any significant systemic absorption and distribution of intact chlorine dioxide by dermal or inhalation routes. It is possible that other derivatives, such as chlorate, chlorite, and chloride ions, could be absorbed and widely distributed. One study shows that “chlorine” (chemical form not charac- terized) derived from aqueous chlorine dioxide is absorbed by the oral route, with a wide distribution and rapid and extensive elimination. No clear information is available on the identity of metabolites, although break- down products are likely to include, at least initially, chlorites, chlorates, and chloride ions.
Given the reactive nature of chlorine dioxide, it seems likely that health effects would be restricted to local responses. There are no quantitative human data, but chlorine dioxide is very toxic by single inhalation exposure in rats; there were no mortalities following exposure to 16 ppm (45 mg/m3) for 4 h, although pulmon-ary oedema and emphysema were seen in all animals exposed to 16–46 ppm (45–129 mg/m3) chlorine dioxide, the incidence increasing in a dose-related manner. The calculated mean LC50 was 32 ppm (90 mg/m3). In another study, ocular discharge, nosebleeds, pulmonary oedema, and death occurred at 260 ppm (728 mg/m3) for 2 h. Chlorine dioxide is toxic when administered in solution
by a single oral dose to rats; at 40 and 80 mg/kg body weight, animals showed signs of corrosive activity in the stomach and gastrointestinal tract. The calculated oral LD50 was 94 mg/kg body weight.
Data on the eye and respiratory tract irritancy of chlorine dioxide gas are limited in extent. However, there is evidence for eye and respiratory tract irritation in humans associated with unknown airborne levels of chlorine dioxide gas. Severe eye and respiratory tract irritancy has been observed in rats exposed to 260 ppm (728 mg/m3) for 2 h.
There are no reports of skin sensitization or occupational asthma associated with chlorine dioxide.
The quality of the available repeated inhalation exposure data in animals is generally poor, such that the information on dose–response must be viewed with some caution. In addition, there is concern that the nasal tissues were not examined, although rhinorrhoea was reported in one study in rats at 15 ppm (42 mg/m3), indicating that the nasal passages may be a target tissue for inhaled chlorine dioxide. Also in rats, no adverse effects were reported at 0.1 ppm (0.28 mg/m3) for 5 h/day for 10 weeks or at 1 ppm (2.8 mg/m3) for 2–7 h/day for 2 months. Lung damage, manifested by small areas of hemorrhagic alveolitis, appears to develop at 2.5 ppm (7.0 mg/m3) or more following repeated exposure for 7 h/day for 1 month and at 10 ppm (28 mg/m3) or more for
15 min twice per day for 4 weeks. Mortalities occurred following exposure at 15 ppm (42 mg/m3) for 15 min, 2 or
4 times per day, for 1 month. In the same exposure regime, there were no adverse effects reported (among the limited observations performed) at 5 ppm (14 mg/m3).
Repeated oral exposure studies are available in humans and animals but are of very little relevance to occupational considerations and were generally of limited design and/or quality. The results show no consistent evidence of thyroid toxicity (which has been most extensively studied) or of other systemic toxicity associated with chlorine dioxide administered in the drinking-water or by gavage.
There is no data available on the effects of repeated dermal exposure and no useful data in relation to chronic exposure or carcinogenicity.
No conclusions can be drawn from genotoxicity studies of chlorine dioxide in bacteria because of limitations in reporting and/or study design. Studies in mammalian cells using aqueous solutions of chlorine dioxide indicate that it is an in vitro mutagen. This activity was not expressed in well conducted studies in vivo in somatic or germ cells. However, given the generally reactive nature of this substance and the fact that positive results have been produced in vitro, there is cause for concern for local “site-of-contact” mutagen- icity, although no studies have been conducted for this end-point.
Well conducted studies in rats have shown that oral exposure at parentally toxic levels does not impair fertility or development. This is consistent with the view that as chlorine dioxide is a reactive gas, it would be unlikely to reach the reproductive organs in significant amounts.
- Criteria for setting tolerable intakes/ concentrations or guidance values for chlorine dioxide gas.
The main health effect in relation to occupational exposure to chlorine dioxide is irritation of the respira- tory tract, skin, and eyes. There is no reliable quantita- tive human data. The animal studies are old and of poor quality, and no long-term studies are available; the likely target tissue, the nasal tract, was not investigated, and studies focused on the lungs. A NOAEL for respiratory tract effects of 1 ppm (2.8 mg/m3) derived from inhalation studies in rats of up to a 2-month duration thus is based on very limited data.
- Sample risk characterization.
The scenario chosen as an example is occupational exposure in the United Kingdom; the available measured occupational exposure data and the exposure levels predicted using the EASE model indicate a maximum likely exposure of 0.1 ppm (0.28 mg/m3), 8-h TWA.
In occupational settings, a pragmatic approach (so-called “margin of safety”) may be used by comparison of NOAELs for the key end-point of concern with the exposure levels achieved under occupational conditions to help determine the adequacy of current practices in terms of protecting human health. Applying this approach for chlorine dioxide, comparison of the predicted exposure level with the NOAEL of 1 ppm (2.8 mg/m3) suggests that there is no cause for concern in relation to the development of irritation of the respiratory tract and of the eyes in workers occupationally exposed to chlorine dioxide..
b. Evaluation of environmental effects.
Insufficient data are available with which to conduct an environmental risk assessment. Chlorine dioxide would be degraded in the environment to yield chlorite and chlorate in water. However, almost all release is to the atmosphere, with decomposition to chlorine and oxygen. The few ecotoxicity data available show that chlorine dioxide can be highly toxic to aquatic organisms; the lowest reported LC50 for fish was 0.02 mg/litre. Chlorate, released in pulp mill wastewaters following use of chlorine dioxide, has been shown to cause major ecological effects on brackish water communities. Brown macroalgae (seaweeds) are particularly sensitive to chlorate following prolonged exposure. The threshold for effects is between 10 and 20 µg/litre.