Chlorine

Endpoint summary

Additional information

The active principle of chlorine and sodium hypochlorite in aqueous solutions is hypochlorite acid. The hypochlorite acid (HOCl) is in equilibrium with hypochlorite anion (OCl-) and chlorine. The equilibrium depends on the pH value: below pH 4 chlorine is available, in the neutral pH range hypochlorous acid is the predominant species and at pH values higher then 10, the only species present is the hypochlorite ion.

The sum of these species (hypochlorite ion + hypochlorous acid + chlorine) is defined as active chlorine or available chlorine. For the chemical reactivity in aqueous solution with the same available chlorine concentrations and the same pH conditions, it is irrelevant whether available chlorine is generated from chlorine gas or sodium hypochlorite. Therefore, all studies investigating aqueous solutions of chlorine or hypochlorite salts can be used for evaluation and assessment of both substances For more information see hydrolysis as a function of pH and identification of breakdown products (IUCLID5 section 5.1.2). Consequently, using data generated with sodium hypochlorite or its salts for the assessment of chlorine in aquatic systems, is scientifically justified.

Stability in aqueous solutions

Sodium hypochlorite is unstable in aqueous solutions. The main parameters that need taken care of when describing aquatic toxicity tests for sodium hypochlorite are:

(a) the pH and temperature

The experimental pH and temperature have an important role in the evaluation of aquatic toxicity data as they determine the chemical species present. In water, the hypochlorite ion is in equilibrium with hypochlorous acid, the ratio of each species being pH and temperature dependent. The amount of hypochlorous acid decreases as pH and temperature increase (Hellawell, 1986). At pH 7.0 70% is HOCl whereas at pH 8.0 80% is ClO- when the temperature is 25°C (Taylor, 1993). Therefore in freshwater (characteristic pH range of 6.5-7.2) the dominant form is the more toxic HOCl while at estuarine pH’s (7.5-8.2) the reverse is true (Gentile et al., 1976). At pH values above 4.0, Cl2 does not exist - hence when authors have reported that they used chlorine in aquatic toxicity studies conducted at a pH of about pH 7.0, the chemical species present are a mixture of HOCl and ClO-. Temperature also affects the equilibrium between hypochlorous acid and hypochlorite, but to a much lesser extent than pH, with the ionization content (pKa) for hypochlorous acid decreasing from 7.75 at 5°C to 7.63 at 15°C and 7.54 at 25°C. Aquatic organisms are generally more sensitive to chlorine at higher temperature; an inverse relationship between LC50 and temperature was found in fish (Brooks and Seegert, 1977; Scott and Middaugh, 1977). It has been demonstrated that a thermal stress combined with exposure to chlorinated compounds produces a synergetic effect (Thatcher et al., 1976).

(b) the characteristics of the test media

Some characteristics of the test media may also affect the evaluation of aquatic toxicity data e.g. presence of ammonia and organic compounds, hardness and salinity. In clean aqueous test media the chlorine present is likely to exist as hypochlorous acid/hypochlorite ion (i.e. detectable as FAC) decaying rapidly due to reduction and photolysis, e.g. half-life < 2 hours for both hypochlorous acid and the hypochlorite ion (Taylor, 1993). In natural water there will be more interactions with nitrogen and organic compounds so that the hypochlorous acid/hypochlorite ion decays faster and the chlorine present is likely to be a mixture of FAC and chloramines (Ewell et al, 1986; Klerks and Fraleigh, 1991). The products formed from chlorination of natural water are also a function of salinity. Since sea water typically contains 60 mg/kg bromide, bromination rather than chlorination may predominate as salinity increases (Scott et al., 1980). In sewage the decay of hypochlorous acid is very rapid, with nearly all the chlorine present being as chloramines.

The effect of the hardness of the test media on the aquatic toxicity of sodium hypochlorite has not been systematically studied. However it is not considered to be substantial as the dissociation of calcium hypochlorite to the hypochlorite ion is reported as reversible suggesting that the availability of the hypochlorite ion in test media would not be reduced by the presence of calcium ions. Some authors report that variation caused by changes in hardness is less than a factor of 3 in some fish species.

(c) the measured test concentration

Sodium hypochlorite is rapidly hydrolysed in water and hypochlorous acid/hypochlorite ion concentrations can decay over the duration of the test. Therefore the initial dosed concentration is not representative of the concentration to which the test organisms have been exposed for the duration of the test.

Summary Aquatic toxicity:

In the scientific literature of the last 30 years, a great number of short and long term aquatic toxicity studies on sodium hypochlorite, conducted with fresh and saltwater organisms belonging to all trophic levels, have been reported. In spite of the high number of studies conducted, only a few provide reliable information useful for the assessment. This because most of the studies were carried out in the 70’s and 80’s and were tailored to answer specific questions such as, for example the efficacy of sodium hypochlorite as biofouling agent, the effect of temperature stress and/or of intermittent exposure of very short pulse additions of hypochlorite. Hence, their evaluation is not straightforward. From many of these studies, although good and reliable, we were not able to retrieve the proper end-point requested for derivation of PNEC (e.g. the NOEC for long-term studies), so that, they were used as additional information in support of valid toxicity data (rated 1 or 2) used in the PNEC calculation. In addition, it should be noted that, besides one new acute toxicity to daphnia study, none of the evaluated experiments were conducted according to GLP, due to the fact that most of them were carried out long before the requirements of REACH. Short and long term toxicity data are available for microorganisms, algae, invertebrates and fish but not equally distributed between fresh and saltwater environments. The evaluation and comparison of toxicity data are made difficult by the complexity of sodium hypochlorite chemistry in water and by the different analytical methods used to measure its concentration. TRC is a measure of both free and combined chlorine (such as chloramines). It is difficult to separate the contribution to toxicity of the free available chlorine such as HOCl/- OCl from that of the combined chlorine species. Also, the relative amounts of the different chlorine species will be different for each test, due to pH and other medium related effects, test duration and ammonium level effects, etc. For those studies where the percentage of FAC (free available chlorine) out of the TRC had been measured and reported the toxicity endpoints have also been expressed as concentration of FAC/l. In chlorinated salt water, toxicant concentration is often expressed as Total Residual Oxidant (TRO) or Chlorine Produced Oxidants (CPO), which include, in addition to free end combined chlorine, also other oxidative products. This makes the assessment task and the comparability of fresh and salt water toxicity data even harder. In the following we summarise the results for the two aquatic habitats.

It has to be emphasized that, due to its high reactivity with many chemicals present in (natural) waters: metals (e.g. Mn, Fe), inorganics (e.g. ammonium, sulphide) and organics (organic nitrogen, humic and fulvic acids), it is virtually impossible to carry out long term studies for NOECs derivation. The lower is the tested concentration, higher is the amount of it reacting at once with above mentioned substances. As it is not appropriate to carry out any fresh or marine water ecotoxicity test in ultra-pure water that does not sustain life for prolonged duration, the decay of hypochlorite ion by reaction with substances present in water cannot be circumvented in practice. Therefore, there are no pertinent long term NOECs based on standard tests to derive PNECs. PNEC are derived from field experiments carried out under flow-through conditions.

With respect to testing on algae, there are additional constraints due to photolysis (half-life 12 -60 min, Nowell & Hoigné, 1992) and the impossibility to adopt a flow through design which prevent to obtain useful concentration/effect relationships. As indicated un the CLP Guidance, use is made of a vascular plant test for classification purpose.

Freshwater

For short term toxicity valid acute data are available only for invertebrates (daphnia 48 h LC50 = 0.141 mg active chlorine/L, continuous, flow-through exposure). In the searched literature, adequate standard acute tests with fish are surprisingly lacking, as in preference many reliable studies have been performed under intermittent exposure. From these latter studies, the trout was shown to be the most sensitive species; three 40 minutes pulses per day produced an LC50 = 60 μgTRC/l after 96h and an LC50 μg/l = 33 μgTRC/l after 168h. As the intermittent exposure regime is known to have a lower impact on animals, permitting a certain degree of recovery, than continuous exposure, we can expect that the 96h LC50 for fish, exposed in a standard test, would be < 60 μgTRC/l. For algae the very few data examined were judged not valid. A freshwater vascular plant test gives an ErC50 = 0.1 mg/L.

For long-term toxicity, no valid NOEC values from standard long-term tests on freshwater species are available.

In the new state of the art 48–hr acute toxicity of sodium hypochlorite to Daphnia magna flow through conditions were adopted(Wildlife International, 2009). Daphnids were exposed to control and test chemical at nominal concentration of 12.5, 25, 50, 100, 200 and 400 µg active chlorine/L for 48 hr. Mortality/immobilization and sublethal effects were observed daily. The 48 – hr EC50 was 141 µg active chlorine/L.

Results from a microcosm study on periphytic community provide a NOEC for algae which can be used for the assessment: 7d NOEC = 3 μgTRC/l, corresponding to 2.1 μgFAC/l. There are supportive data from another microcosm study by same authors showing that a 28d exposure to the same hypochlorite concentration causes 50% reduction in algal biomass (28d EC50 = 2.1 μgFAC/l). For fish there are supportive information from a field study indicating a 134d NOEC ≥ 5 μgTRC/l, i.e. fish appear somewhat less sensitive than algae after long term exposure. The results of supporting information from a mesocosm study with daily pulse chlorine contamination (that could potentially underestimate the long-term toxicity of sodium hypochlorite) provide indication that zooplankton density (24d NOEC = 1.5 μgTRC (or FAC)/l) is an endpoint much more sensitive than algal or chlorophyll a reduction (24d NOEC = 79 μgTRC (or FAC)/l). Supportive information on bivalves only suggest for this group a NOEC << 50 μgTRC/l.

The ECHA Guidance documents give guidance on which assessment factors should be applied to derive a PNEC when different kinds and types of data are available. The minimum data set required (at least one short-term L(E)C50 from each of three trophic levels) is met and algae is determined to be the most sensitive trophic level. Long-term valid (with restriction) data are available for one trophic level, i.e. primary producers (algae). Indications of long-term toxicity to fish and molluscs suggest that these groups are less sensitive than algae.

Marine water

Valid short-term toxicity data have been collected for molluscs, copepods and fish, which show a similar sensitivity: (resp. 48h-EC50 = 26 μgTRC/l , = 29 µg/L and 96h LC50 = 32 μg TRO/l). Supportive information indicate that acute toxicity is higher for fish eggs: 48h EC50 = 8 μg TRC/l. No data for algae are available, so that it is not possible to determine which trophic level is the most sensitive in short-term tests.

For the same groups, molluscs and fish, long-term toxicity data have also been found. Molluscs (15d NOEC = 7 μgTRO/l) are more sensitive than fish fry (28d NOEC (fry survival) = 40 μgCPO/l) For crustaceans and algae, adequate data for risk assessment are again not available. Anyhow a study conducted on periphytic community under intermittent exposure provides a 21d EC50 laying between 1 and 10 μgTRC/l, suggesting that following continuous exposure the NOEC for algae is likely to be lower than that.

Conclusion

The aquatic PNECs were derived from most sensitive endpoint for the aquatic compartment, algal biomass in microcosm experiement. The test revealed a NOEC (7d) of 0.0021 mg/L (Cairns, 1990). An assessment factor of 10 was applied (Guidance on Information Requirement and Chemical Safety Assessment, May 2008).