Lithium hexafluorophosphate(1-)

Endpoint summary

Key value for chemical safety assessment

Effects on fertility

Description of key information

Available data for the hydrolysis products of LiPF6 are sufficient to allow conclusions on reproductive and developmental toxicity to be drawn.

Effect on fertility: via oral route

Endpoint conclusion:

no study available

Effect on fertility: via inhalation route

Endpoint conclusion:

no study available

Effect on fertility: via dermal route

Endpoint conclusion:

no study available

Additional information

HF

Following absorption of HF, only fluoride (as free F- or bound organofluoride) is systemically distributed with the mammalian body, and the same is also true of soluble inorganic fluoride salts. For this reason, reproductive toxicity test data for sodium fluoride can be used to predict the toxicity of HF to reproduction (HF: EU Risk Assessment Report, 2001).

 

Following detailed review, the EU RAR for HF concluded that available animal test data in which effects on fertility were reported are of questionable reliability for several reasons. NOAEL values were not established in these studies and dose levels were below expected background exposure levels for fluoride, which were not measured. By contrast, a US-FDA 2-generation study in rats and US NTP developmental toxicity studies in rats and rabbits, both using administration of sodium fluoride in drinking water, are considered reliable.

 

In the 2-generation study (Collins et al, 2001a: performed under GLP and using methods in accordance with OECD guideline 416), rats maintained on a low fluoride diet and given drinking water with a background F- level <0.2 mg/l were given water supplemented with NaF at levels up to 250 mg/l (delivering an F- dose equivalent to 12-13 mg F-/kg/day to females, 10-11 mg F-/kg/day to males). Parental (F0) animals were mated after 10 weeks of treatment, males (together with F1 males not used for mating) then being entered into a separate investigation of male fertility effects. After weaning, separate subsets of F1 rats were mated, treated without mating up to Day 90, or culled. In the separate investigation of male fertility, histopathological examination of male reproductive organs and measurement of sperm parameters found no adverse effects of treatment. No significant effects of treatment on male or female reproduction were seen in the F0 or F1 generations: pregnancy rates, F1 survival and development and organs weights (for male and female reproductive organs) did not differ significantly between test and control groups. The only treatment-related morphological changes noted were related to growth or coloration of the teeth after high-level exposure. The authors concluded that intake of water containing NaF at up to 250 ppm, equivalent to 12.8 mg F-/kg/day, had no adverse effect on reproduction (Collins et al, 2001a). This conclusion is expressed as an NOAEL for effects on fertility of about 10 mg F-/kg/day in the EU RAR for HF.

Fluoride (F-)

This is addressed in the HF text above. 

Lithium 

In an OECD 422 repeated dose and reproductive/developmental toxicity study, lithium bromide (5, 20 or 80 mg/kg/day) was given orally to rats for 42 days (males) or from 14 days premating to lactation Day 4 (females) (Japan MHW, 2002). Males showed indications of renal toxicity at 20 mg/kg/day and higher dosage (reduced urine specific gravity, increased urinal and plasma creatinine, reduced relative kidney weights); females showed lesser signs of toxicity, with slightly reduced plasma albumin concentration at 20 mg/kg/day. The only reproductive parameter showing a possible effect was time to copulation after pairing: this was slightly (but not significantly) extended at 20 mg/kg/day and higher dosage, but copulation index and pregnancy rates were unaffected. NOEL values of 20 mg/kg/day (for adult and reproductive toxicity) and 80 mg/kg/day for developmental toxicity were concluded; these lithium bromide dosages included Li at 1.6 mg/kg/day and 6.4 mg/kg/day respectively.

 

Toxicity of lithium to reproduction was also evaluated in the international expert review (Moore et al, 1995). This concluded that available data human data do not permit firm conclusions to be drawn, but noted a report that mice given lithium (via drinking water) at 11.4 mg/kg/day, with resultant serum levels of 0.67mM (4.7 mg/l), produced smaller litters at increased intervals; exposure to double this level of lithium, reproduction was completely inhibited. However the design of this study did not allow determination of an NOAEL or which sex (if only one) was affected. The review concluded a calculated Unlikely Effect Level for reproductive toxicity in humans of 0.22 mgLi/kg.

In a 90-day rat study (Thakur et al, 2003), lithium carbonate was fed to male rats at concentrations up to 1100 mg/kg in diet. At termination, reproductive organs were weighed and processed for histopathological examination; testicular interstitial fluid volume was also measured, together with serum testosterone and investigations of sperm morphology and performance (fertility index following mating). The test dosages employed were found in an earlier, similarly conducted subchronic toxicity study to raise serum Li levels by 0.5 – 1.2 mM (3.5 – 8.3 mg/l): using the recognised default conversion factor to estimate daily intake from dietary concentration of the test substance (0.09 x diet concentration) lithium intakes can be calculated as 8.5, 13.5 and 18.6 mg Li/kg/day. At the two higher dosages testis, epididymis, seminal vesicle and prostate gland weights were reduced (differences from controls being significant only for absolute and not bodyweight-relative weights) and histopathology revealed structural changes in the testes and accessory organs. These were most marked at the highest test dosage, where moderate to severe degeneration was seen. Among sperm parameters investigated, epididymal number, daily production and transit rate were all significantly affected in the two higher dosage groups. Abnormal sperm (%) were significantly increased in all test groups and serum testosterone was reduced (not significantly in the low dose group). Mating with untreated females showed reduced fertility in all test groups, significantly different from controls at the two higher test dosages. No data on systemic toxicity are presented, but from the calculated lithium intakes and reported serum Li concentrations it could be expected that the highest tested dosage would cause significant systemic toxicity (e.g. adverse renal effects): in another study summarised by Moore et al (1995) which gave lithium chloride in drinking water to rats of the same strain, an estimated Li dosage of 22.9 mg/kg/day proved fatal within a few weeks, while 15.3 mg/kg/day produced no general toxicity over a 2-year period.

 

Phosphate

The ubiquity of endogenous phosphate within the body and the Maximum Tolerable Daily Intake level of phosphate for man (70 mg/kg/day as P: FAO/WHO/IPCS, 1982; EFSA, 2008b) clearly demonstrate the low toxicity of phosphoric acid/phosphate. The FAO/WHO Expert Committee report on phosphoric acid, phosphates and polyphosphates (FAO Nutrition Series 48A, 1970) noted:

- a study in which rats were fed phosphoric acid at up to 0.75% in diet over three successive generations with no harmful effects on growth or reproduction

- a similar study using mixed Na2H2P2O7 + KPO3 added at up to 5% into a diet containing 0.47% P . Growth and fertility were not significantly affected at up to 2.5%, but were reduced at 5%.

It is evident that any internal dose of phosphate resulting from LiPF6 exposure at levels not causing frank toxicity due to HF/F- release will not be of concern for toxicity to reproduction (P forming 20% of LiPF6, with F constituting 75%).

 

Short description of key information:
Considering the data available on hydrolysis products of LiPF6:
[1] it is concluded that F- absorbed and systemically distributed following intake of LiPF6 would not cause developmental or reproductive toxicity
[2]it is concluded that systemically distributed Li+, if present at sufficient concentration, could have potential to elicit both developmental and reproductive toxicity. However, this has no relevance for classification or risk assessment of LIPF6: the lowest NOAEL value determined cited earlier for developmental or reproductive toxicity of fluoride (determined in the reliable 2-generation rat study or in reliable rat or rabbit developmental studies ) was 10 mgF-/kg/day, and the maternal toxicity NOAEL is similar or lower. To achieve this level of fluoride from LiPF6 intake would require intake of 13.3 mg of the substance/kg/day: this would correspond to a Lithium intake of only 609 µgLi/kg/day, which is clearly below the reported lithium NOAEL values for developmental or reproductive toxicity. Thus consideration of fluoride, not lithium, toxicity should be applied to LiPF6 classification with regard to effects on reproduction and this does not indicate selective toxicity with regard to fertility or embryo/foetal development
[3] it is evident that any internal dose of phosphate resulting from LiPF6 exposure at levels not causing frank toxicity due to HF/F- release will not be of concern for toxicity to reproduction (P forming 20% of LiPF6, with F constituting 75%).

Justification for selection of Effect on fertility via oral route:
Lithium hexafluorophosphate is reactive and unstable in water and air. Reaction in contact with water proceeds rapidly, with release of hydrogen fluoride (forming hydrofluoric acid). Such local generation of hydrogen fluoride/hydrofluoric acid at the site of contact with skin or other membranes, with consequent potential for serious local tissue damage, is a major cause of the observed corrosivity of the substance, and secondary tissue necrosis due to localised free fluoride ion concentrations is also a likely contributor to this (Kirkpatrick, Enion and Burns, 1995): delayed onset of deep tissue damage and pain is known after skin contact with HF solutions below 20% in concentration (US ATDSR, 2001). Ethical and practical reasons therefore make it inappropriate to conduct reproductive toxicity testing of lithium hexafluorophosphate in animals; since information is available on systemic toxicity of the ultimate hydrolysis products hydrogen fluoride, lithium/Li+, fluoride/F- and phosphoric acid/phosphate, such testing is also scientifically unnecessary.

Justification for selection of Effect on fertility via inhalation route:
Lithium hexafluorophosphate is reactive and unstable in water and air. Reaction in contact with water proceeds rapidly, with release of hydrogen fluoride (forming hydrofluoric acid). Such local generation of hydrogen fluoride/hydrofluoric acid at the site of contact with skin or other membranes, with consequent potential for serious local tissue damage, is a major cause of the observed corrosivity of the substance, and secondary tissue necrosis due to localised free fluoride ion concentrations is also a likely contributor to this (Kirkpatrick, Enion and Burns, 1995): delayed onset of deep tissue damage and pain is known after skin contact with HF solutions below 20% in concentration (US ATDSR, 2001). Ethical and practical reasons therefore make it inappropriate to conduct reproductive toxicity testing of lithium hexafluorophosphate in animals; since information is available on systemic toxicity of the ultimate hydrolysis products hydrogen fluoride, lithium/Li+, fluoride/F- and phosphoric acid/phosphate, such testing is also scientifically unnecessary.

Justification for selection of Effect on fertility via dermal route:
Lithium hexafluorophosphate is reactive and unstable in water and air. Reaction in contact with water proceeds rapidly, with release of hydrogen fluoride (forming hydrofluoric acid). Such local generation of hydrogen fluoride/hydrofluoric acid at the site of contact with skin or other membranes, with consequent potential for serious local tissue damage, is a major cause of the observed corrosivity of the substance, and secondary tissue necrosis due to localised free fluoride ion concentrations is also a likely contributor to this (Kirkpatrick, Enion and Burns, 1995): delayed onset of deep tissue damage and pain is known after skin contact with HF solutions below 20% in concentration (US ATDSR, 2001). Ethical and practical reasons therefore make it inappropriate to conduct reproductive toxicity testing of lithium hexafluorophosphate in animals; since information is available on systemic toxicity of the ultimate hydrolysis products hydrogen fluoride, lithium/Li+, fluoride/F- and phosphoric acid/phosphate, such testing is also scientifically unnecessary.

Effects on developmental toxicity

Description of key information

Considering the data available on hydrolysis products of LiPF6:
[1] it is concluded that F- absorbed and sytemically distributed following intake of LiPF6 would not cause developmental or reproductive toxicity
[2]it is concluded that systemically distributed Li+, if present at sufficient concentration, could have potential to elicit both developmental and reproductive toxicity.  However, this has no relevance for classification or risk assessment of LIPF6: the lowest NOAEL value determined cited earlier for developmental or reproductive toxicity of fluoride (determined in the reliable 2-generation rat study or in reliable rat or rabbit developmental studies ) was 10 mgF-/kg/day, and the maternal toxicity NOAEL is similar or lower.  To achieve this level of fluoride from LiPF6 intake would require intake of 13.3 mg of the substance/kg/day: this would correspond to a Lithium intake of only 609 µgLi/kg/day, which is clearly below the reported lithium NOAEL values for developmental or reproductive toxicity.  Thus consideration of fluoride, not lithium, toxicity should be applied to LiPF6 classification with regard to effects on reproduction and this does not indicate selective toxicity with regard to fertility or embryo/foetal development
[3] it is evident that any internal dose of phosphate resulting from LiPF6 exposure at levels not causing frank toxicity due to HF/F- release will not be of concern for toxicity to reproduction (P forming 20% of LiPF6, with F constituting 75%).

Effect on developmental toxicity: via oral route

Endpoint conclusion:

no study available

Effect on developmental toxicity: via inhalation route

Endpoint conclusion:

no study available

Effect on developmental toxicity: via dermal route

Endpoint conclusion:

no study available

Additional information

HF

Data from a rat 2-generation study were used to evaluate the developmental toxicity of sodium fluoride (Collins et al, 2001b). At Day 20 of gestation, a subset of F0 females and their litters (F1) from each study group were examined: implant status, foetal weight, length and sex plus morphology were recorded. F1 females and their litters were later similarly examined, with soft-tissue and skeletal development of F2 foetuses being recorded. No effects of treatment on pre-implantation loss, foetal viability, implant deaths, sex distribution or visceral development were recorded. Decreased ossification of the hyoid bone, significant only in F2 foetuses of the 250 mg NaF/l group and when analysed on a total pups (not per litter) basis, made the authors consider this an effect level for developmental toxicity. No definitive evidence of maternal toxicity was seen: although F0 female weight gains were reduced at the two highest exposure levels (175 and 250 ppm NaF), this was seen only at 175 ppm and not at 250 ppm in F1 females. NOAEL values for developmental toxicity of 11.7 mgF-/kg.day or 12.7 mgF-/kg/day for the F0 and F1 generation respectively were concluded in the EU RAR for HF.

 

In an earlier study of developmental toxicity, the same researchers used a test protocol comparable to OECD method 414, giving mated female rats drinking water containing NaF at up to 250 ppm (and low-fluoride feed). After termination on gestation day 20, implant and resorption sites plus numbers of live and dead foetuses were recorded. Live foetuses were then measured and examined for external and internal (skeletal and soft tissue) abnormalities. Indications of maternal toxicity at the maximum (250 ppm) exposure level included reduced food and water intake and reduced weight gain; in this treatment group only the mean numbers of foetuses/litter having three or more skeletal variations were significantly increased. Calculated NaF intake at the maximum exposure level was 25.1 mg/kg/day, which corresponds to 11.4 mg F-/kg/day (Collins et al, 1995). This represents the NOAEL for F- in the study.

 

Studies performed under the US National Toxicology program investigated developmental toxicity in rats and rabbits using methods comparable to OECD guideline 414, with NaF at levels up to 300 ppm (rats) or 400 ppm (rabbits) in drinking water being given on Days 6-15 (rats) or 6-19 (rabbits) of gestation (US NTP 1994 and 1993). Fluoride levels in the water itself were <0.6 ppm; taking into account fluoride intake from food, maximum daily doses of F- in test groups were approximately 13 mg/kg/day (rats) and 14 mg/kg/day (rabbits). There was minimal evidence of maternal toxicity (transient weight loss) in both species at the highest exposure level, but no indication of developmental toxicity (no effects on post-implantation loss, foetal weights or malformations). Concluded NOAEL values, expressed as NaF concentration and F- dosage, were:

- rat maternal NOAEL 150 ppm (9 mg F-/kg/day)

- rat developmental toxicity NOAEL 300 ppm (13 mg F-/kg/day)*

- rabbit maternal NOAEL 200 ppm (9 mg F-/kg/day)

- rabbit developmental toxicity NOAEL 400 ppm (14 mg F-/kg/day)**

* NOAEL cited as 12.3 mg F-/kg/day in the HF: EU RAR summary of this study

** NOAEL cited as 13.2 mg F-/kg/day in the HF: EU RAR summary of this study.

Measurement of serum fluoride levels in rabbits showed that the maternal and developmental NOAEL values corresponded to 0.39 +/- 0.14 mg/l and 0.70 +/- 0.33 mg/l respectively.

From these three developmental toxicity studies, the HF: EU RAR concludes maternal and developmental toxicity NOAELs for fluoride of 11.12 mg F-/kg/day (NOAEL from the Collins 1995 study, as expressed in the HF: EU RAR).

Fluoride (F-)

This is addressed in the HF text above. 

Lithium

Developmental and reproductive toxicity of lithium has been reviewed in detail by a large and international expert committee (Moore at al, 1995).  Among 39 reports of possible relevance to human developmental toxicity, 20 were used for evaluation. Data from the International Register of Lithium Babies suggested an increased risk of cardiac malformation following first trimester lithium exposure. Prospective studies of pregnancy outcomes among women taking lithium medication did not give consistent results, one indicating an excess of cardiac defects in offspring, one showing no increase in malformations compared to a control group. However the former study also found an increase in neonatal deaths. Retrospective studies and reports of clinical cases did not add significantly to the toxicity profile. It was concluded that administration of lithium at doses used for human therapy can cause developmental toxicity.

 

The same review also considered reports of developmental toxicity in animal tests, conducted using rats, mice, rabbits, monkeys and pigs. From a number of different studies in rodents, it concluded that lithium exposure during pregnancy can cause developmental toxicity in rats and mice, including in some cases reduced foetal weight, delayed postnatal development and cleft palate abnormality (in mice – perhaps a species-specific effect). Effective dosages for these effects were in the range 2.71 - 12.67 mmol Li/kg/day (for prenatal toxicity: 18.8 – 87.9 mg Li/kg/day) and 0.41 – 1.64 mg Li/kg/day (postnatal toxicity: 2.8 – 11.4 mg Li/kg/day). However it is notable that these generally occurred in association with maternal toxicity, and the more limited investigations in other species showed no such effect. The review concluded an Unlikely Effect Level for developmental toxicity in humans (set using an added uncertainty factor) of 0.11 mgLi/kg.

 

In an OECD 422 repeated dose and reproductive/developmental toxicity study, lithium bromide (5, 20 or 80 mg/kg/day) was given orally to rats for 42 days (males) or from 14 days premating to lactation Day 4 (females) (Japan MHW, 2002). Males showed indications of renal toxicity at 20 mg/kg/day and higher dosage (reduced urine specific gravity, increased urinal and plasma creatinine, reduced relative kidney weights); females showed lesser signs of toxicity, with slightly reduced plasma albumin concentration at 20 mg/kg/day. The only reproductive parameter showing a possible effect was time to copulation after pairing: this was slightly (but not significantly) extended at 20 mg/kg/day and higher dosage, but copulation index and pregnancy rates were unaffected.  NOEL values of 5 mg/kg/day (for adult and reproductive toxicity) and 80 mg/kg/day for developmental toxicity were concluded; these lithium bromide dosages included Li at 0.4 mg/kg/day and 6.4 mg/kg/day respectively.

Phosphate

The ubiquity of endogenous phosphate within the body and the Maximum Tolerable Daily Intake level of phosphate for man (70 mg/kg/day as P: FAO/WHO/IPCS, 1982; EFSA, 2008b) clearly demonstrate the low toxicity of phosphoric acid/phosphate. The FAO/WHO Expert Committee report on phosphoric acid, phosphates and polyphosphates (FAO Nutrition Series 48A, 1970) noted:

- a study in which rats were fed phosphoric acid at up to 0.75% in diet over three successive generations with no harmful effects on growth or reproduction

- a similar study using mixed Na2H2P2O7 + KPO3 added at up to 5% into a diet containing 0.47% P . Growth and fertility were not significantly affected at up to 2.5%, but were reduced at 5%.

It is evident that any internal dose of phosphate resulting from LiPF6 exposure at levels not causing frank toxicity due to HF/F- release will not be of concern for toxicity to reproduction (P forming 20% of LiPF6, with F constituting 75%).

 

Justification for selection of Effect on developmental toxicity: via oral route:
Lithium hexafluorophosphate is reactive and unstable in water and air. Reaction in contact with water proceeds rapidly, with release of hydrogen fluoride (forming hydrofluoric acid). Such local generation of hydrogen fluoride/hydrofluoric acid at the site of contact with skin or other membranes, with consequent potential for serious local tissue damage, is a major cause of the observed corrosivity of the substance, and secondary tissue necrosis due to localised free fluoride ion concentrations is also a likely contributor to this (Kirkpatrick, Enion and Burns, 1995): delayed onset of deep tissue damage and pain is known after skin contact with HF solutions below 20% in concentration (US ATDSR, 2001). Ethical and practical reasons therefore make it inappropriate to conduct reproductive toxicity testing of lithium hexafluorophosphate in animals; since information is available on systemic toxicity of the ultimate hydrolysis products hydrogen fluoride, lithium/Li+, fluoride/F- and phosphoric acid/phosphate, such testing is also scientifically unnecessary.

Justification for selection of Effect on developmental toxicity: via inhalation route:
Lithium hexafluorophosphate is reactive and unstable in water and air. Reaction in contact with water proceeds rapidly, with release of hydrogen fluoride (forming hydrofluoric acid). Such local generation of hydrogen fluoride/hydrofluoric acid at the site of contact with skin or other membranes, with consequent potential for serious local tissue damage, is a major cause of the observed corrosivity of the substance, and secondary tissue necrosis due to localised free fluoride ion concentrations is also a likely contributor to this (Kirkpatrick, Enion and Burns, 1995): delayed onset of deep tissue damage and pain is known after skin contact with HF solutions below 20% in concentration (US ATDSR, 2001). Ethical and practical reasons therefore make it inappropriate to conduct reproductive toxicity testing of lithium hexafluorophosphate in animals; since information is available on systemic toxicity of the ultimate hydrolysis products hydrogen fluoride, lithium/Li+, fluoride/F- and phosphoric acid/phosphate, such testing is also scientifically unnecessary.

Justification for selection of Effect on developmental toxicity: via dermal route:
Lithium hexafluorophosphate is reactive and unstable in water and air. Reaction in contact with water proceeds rapidly, with release of hydrogen fluoride (forming hydrofluoric acid). Such local generation of hydrogen fluoride/hydrofluoric acid at the site of contact with skin or other membranes, with consequent potential for serious local tissue damage, is a major cause of the observed corrosivity of the substance, and secondary tissue necrosis due to localised free fluoride ion concentrations is also a likely contributor to this (Kirkpatrick, Enion and Burns, 1995): delayed onset of deep tissue damage and pain is known after skin contact with HF solutions below 20% in concentration (US ATDSR, 2001). Ethical and practical reasons therefore make it inappropriate to conduct reproductive toxicity testing of lithium hexafluorophosphate in animals; since information is available on systemic toxicity of the ultimate hydrolysis products hydrogen fluoride, lithium/Li+, fluoride/F- and phosphoric acid/phosphate, such testing is also scientifically unnecessary.

Justification for classification or non-classification

Justification for no classification

Among the degradants of LiPF6 which could be systemically distributed following exposure to the substance, only HF/F- has been identified as of possible concern for reproductive toxicity (rat NOAEL for fertility 10 mgF-/kg/day). It is known that serum or plasma fluoride levels in man are more readily elevated following exposure than those of rats, but this dosage is far above the human tolerated intake/reference dose/NOEL values for F- of 30-100 ugF-/kg/day. Hence no specific antifertility effect of long-term fluoride intake, in the absence of systemic toxicity, can be expected in man. Accordingly, no classification in respect of reproductive toxicity is appropriate for LiPF6 and calculation of DNEL values should be based on repeated dose toxicity information rather than data on toxicity to reproduction.