Sulphur hexafluoride

Endpoint summary

Description of key information

Available animal and human data indicate that acute toxicity of SF6 is very low. Although most available animal acute toxicity studies have not been performed in accordance with currently regulatory guidelines, based on a weight of evidence approach in combination with available human studies they are considered to be sufficient to cover this endpoint.
In the limited acute inhalation studies conducted with rats which were exposed up to 80% SF6, no deaths or adverse effects clearly attributable to SF6 were recorded. Also no mortalities or adverse effects were noted in a GLP-compliant 28-day repeated dose toxicity study with rats, exposed to 302687 mg/m3 SF6 for 6 hours daily. No significant adverse effects were recorded in several studies in humans acutely exposed to an atmosphere containing up to 80% SF6, although slight anaesthetic effects and slight signs of discomfort, such as coolness in the upper respiratory tract and the occurrence of voice deepening (which is likely due to the higher density of SF6 in comparison with air), were observed.

Key value for chemical safety assessment

Acute toxicity: via oral route

Endpoint conclusion

Endpoint conclusion:

no study available

Acute toxicity: via inhalation route

Endpoint conclusion

Endpoint conclusion:

no adverse effect observed

Acute toxicity: via dermal route

Endpoint conclusion

Endpoint conclusion:

no study available

Additional information

Animal data

As SF6 is a gas, the performance of oral and dermal acute toxicity studies is technically not feasible. Lester and Greenberg (1950) exposed rats to various concentration of SF6 by inhalation. Adult white rats (sex and strain not specified) were exposed to various concentration of sulphur hexafluoride in a glass chamber. Six rats were exposed to 17% SF6 for 18 hrs. Three and two animals were exposed to 80% of the test substance for 2 and 23 hours, respectively. None of these animals showed any clinical sign of irritation or toxic effect during and at the end of the exposure. According to the authors, about 50 other rats were exposed to 80% SF6 and 20% O2 for periods of 16 to 24 hours without any indication of adverse effects.

White rats (sex, strain and number not specified) were subjected for a single exposure period (time of exposure not specified) to 1500, 300 and 62.9 g/m3 (25, 5 and 1%) SF6 via inhalation by Sanotskii et al. (1981). Clinical signs, neurobehavioral observations, blood analyses and hepatic and renal functions were studied. Gonadal effects and bone marrow genotoxicity were also assessed. No mortality was recorded in this study. Neurobehavioral effects were observed in rats exposed to 1500 g/m3. However these data were affected by a low level of O2 in the exposure atmosphere. Disturbance of short and long-term memory was observed in rats exposed to 300 g/m3. No changes were observed in rats exposed to 62.9 g/m3 SF6. No other effects were observed in this study.

The toxicological inertness of SF6 was assessed by Specht and Burbach (1951). In this study 17 white rats (sex and species not specified) were exposed to SF6/O2 mixture at 72%/28% (v/v) ratio for 6 hours. 40 days after the first exposure 8/17 animals were exposed to SF6/O2 78%/22% (v/v) for 6 hours. No deaths or any other adverse effect were recorded in this study.

The value of the three available studies is limited (rated separately with a Klimisch code of 4). However, based on a weight of evidence approach, these studies in combination with the human data on acute inhalation toxicity (see section 5.2.2) are considered sufficient to cover the endpoint acute toxicity by inhalation.

Also no adverse effects or mortalities were reported in the available 28-day repeated dose toxicity study with reproductive and developmental screening, performed according to OECD guidelines and under GLP, in which Wistar rats were exposed for 6 hours/day to an analytical concentration of 302687 mg/m³ SF6 (target concentration of 50000 ppm), confirming the very low acute toxicity of the substance.

Human data

Six human studies (some have limitations in reporting) and one case report were available for assessment. In combination with the animal acute inhalation toxicity data, these studies are considered sufficient to cover the endpoint acute inhalation toxicity.

Esch et al.(1988) studied the effect of carrier gas composition on particle deposition in the airways. For this purpose, 2 male and 2 female volunteers were exposed to 1 litre of 80%/20% mixture of SF6/O2. Triphenyl phosphate was used for aerosol generation. No signs of toxicity or adverse effects were observed in this study for human volunteers exposed to 80% SF6.

The anaesthetic potential of SF6 and nitrous oxide were compared by Virtue and Weaver (1952). 20 young adults were exposed to 79% SF6 (21% O2) and 30% N2O for about ten minutes. The earliest subjective sensation appeared when SF6 reached 55%. The experienced sensations were analgesia, sleepiness and deeping quality of the voice (this effect is likely due to the higher density of SF6 in comparison with air). Subjects showed a faster recovery with SF6 than with nitrous oxide. Four subjects felt slightly dyspnoeic. However, no other adverse effect was observed in subjects exposed to SF6. First anaesthetic effects were experienced at 22% SF6.

No signs of discomfort were observed in 58 voluntary subjects exposed to 70% SF6 in a study on the differences in respiratory flow rates of smoker (n=29) versus non-smoker subjects (n=29) apart from a temporary deepening of the voice in the exposed subjects (Brooks et al. 1978).

Specht and Brubach (1951) voluntarily subjected themselves to breathe a mixture of SF6/O2 (78%/22%) for short periods (exposure periods not specified). Coolness in the upper respiratory tract and the occurrence of voice deepening was reported as the first effect related to inhalation of SF6. After 20-30 breaths of the mixture at normal rate and volume a slight to marked vertigo was experienced, disappearing after few minutes on breathing air. No other adverse effect was reported.

The narcotic influence of SF6 on mental and psychomotor performance has been studied in 9 subjects at normal atmospheric pressure (Östlund et al., 1992). Control experiments were performed with air and with nitrous oxide. Psychomotor, perceptual and cognitive abilities were assessed using a computerized test battery. Subjects were exposed to air and six different normoxic gas mixtures: 13, 26, and 39% N2O, and 39, 59, and 79% SF6. Significant performance impairments were found with 13% N2O and gradual further impairment with 26, and 39% N2O. During exposure to 39, 59, and 79% SF6over-all performance was impaired by 5, 10, and 18%, respectively. Impairment was significant with 59 and 79% SF₆. The results indicate that the relative narcotic potency of SF6: N2O is about 1:4 in humans. It is concluded that a normoxic SF6 -O2 mixture can be inhaled for lung function studies without any harmful effects and that the short-lasting narcotic effect, although detectable with a test battery, would not impair the ability of the subject to perform simple breathing procedures.

In another study of Östlund a comparable relative narcotic potency between SF6 and N2O was observed (Östlund et al, 1994). Impairments of psychomotor, perceptual, and cognitive abilities were determined in nine male subjects exposed to inhaled SF₆partial pressures of 0, 52, 104, and 156 kPa and to inhaled N2 partial pressures of 103, 575, 825, and 1,075 kPa. Also data from a previous study with inhaled N2O partial pressures of 0, 13, 26, and 39 kPa were included. With the highest gas concentrations, performances were reduced by 41-57%. Effective doses for 20% performance impairment were 830, 97, and 21.5 kPa for N2, SF6, and N2O, respectively, yielding relative narcotic potencies of 1.0:8.5:39. The order of narcotic potencies is the same as for the lipid solubility of the three gases. In contrast, the order of increasing tendency for hydrate formation (decreasing hydrate dissociation pressure) for the three gases is N2, N2O, and SF6. Thus, mild to moderate inert gas narcosis in humans shows the same positive relationship to lipid solubility as was shown in previous animal models that utilized much deeper levels of anaesthesia.

Two electrical equipment maintenance workers were accidentally exposed to SF6. Both subject fell unconscious after a few minutes and were rescued after about 30 minutes of exposure. SO2F2was the only SF6 by-product detected into the tower, at a concentration of 50 ppm. 1 subject developed a florid pulmonary oedema on 90 minutes after the end of the exposure (probably caused by toxic by-product formed by sparks or electrical arches), whereas the other exposed worker did not complain of any respiratory effect. Both subjects recovered within 3 days (Pilling and Jones, 1988).

Justification for classification or non-classification

Based on the absence of mortality and adverse effects at a SF6 exposure concentration as high as 302687 mg/m3/6 h, classification of SF6 for acute toxicity is not warranted in accordance with EU Classification, Labelling and Packaging of Substances and Mixtures (CLP) Regulation (EC) No. 1272/2008.