Registration Dossier

Diss Factsheets

Administrative data

Description of key information

Key value for chemical safety assessment

Repeated dose toxicity: via oral route - systemic effects

Endpoint conclusion
Endpoint conclusion:
no adverse effect observed
Dose descriptor:

Additional information

Introductionto read-across matrix

A comprehensive data gap analysis was conducted for the entire substance portfolio of the REACH Metal Carboxylates Consortium (RMC), covering 10 metal carboxylates in total. This literature screening effort included:


  • all available proprietary studies from the REACH Metal Carboxylates Consortium (RMC)
  • detailed literature searches in online databases
  • screening of human health review articles
  • rigorous quality and reliability screening according to Klimisch criteria, where those criteria apply


During the literature search and data gap analysis it became obvious that the overall database on substance-specific human health hazard data for the metal carboxylates is too scant to cover all REACH endpoints. Therefore, the remaining data gaps had to be covered by either experimental testing or read-across from similar substances.


Selected endpoints for the human health hazard assessment are addressed by read-across, using a combination of data on the organic acid counterion and the metal (or one of its readily soluble salts). This way forward is acceptable, since metal carboxylates dissociate to the organic anion and the metal cation upon dissolution in aqueous media. No indications of complexation or masking of the metal ion through the organic acid were apparent during the water solubility tests (please refer to the water solubility data in section of the IUCLID and chapter of the CSR). Once the individual constituents of the metal carboxylate become bioavailable (i.e. in the acidic environment in the gastric passage or after phagocytosis by pulmonary macrophages), the “overall” toxicity of the dissociated metal carboxylate can be described by the toxicity of the “individual” constituents. Since synergistic effects are not expected for this group of metal carboxylates, the human health hazard assessment consists of an individual assessment of the metal cation and the organic anion.


The hazard information of the individual constituents was obtained from existing REACH registration dossiers via a license-to-use obtained by the lead registrant. These registration dossiers were submitted to ECHA in 2010 as full registration dossiers, and are thus considered to contain relevant and reliable information for all human health endpoints. All lead-registrant dossiers were checked for completeness and accepted by ECHA, i.e. a registration number was assigned.


Oxygen-free inorganic zirconium(IV) substances (ZrX4) are highly hygroscopic and rapidly decompose in damp atmosphere or water under formation of zirconyl compounds (ZrOX2and ZrO(OH)X). These zirconyl compounds further show an ageing under formation of zirconium dioxide (ZrO2). The zirconium dioxide is an inert metal oxide with a very low water solubility (<55µg/L) and is also the naturally occurring mineral baddeleyite.

The water solubility test (according to OECD 105 and under GLP, as presented under the respective endpoint in this dossier) with zirconium propionate confirmed the above described decomposition of inorganic zirconium salts under formation of insoluble zirconium dioxide. After 24h stirring at a loading of 100mg/L, the zirconium concentration of the samples was at 11.89 µg Zr/L.

It is concluded that localtoxicological effects of inorganic zirconium salts might be exerted solely due to the caustic properties of the concentrated resulting acid in the decomposition reaction (in this case propionic acid). Potential, local effects caused by zirconium propionate are addressed by substance-specific test data, whereas systemic effects are addressed by reading-across to the decomposition products, namely the propionate anion and zirconium dioxide.


Based on the above information, unrestricted read-across is considered feasible and justified.


Although the term „constituent“ within the REACH context is defined as substance (also being part of a mixture), the term constituent within this hazard assessment is meant to describe either part of the metal carboxylate salt, i.e. anion or cation.


Repeated dose toxicity

No repeated dose toxicity study with zirconium propionate is available, thus the repeated dose toxicity will be addressed with existing data on the dissociation products zirconium and propionic acid as detailed in the table below.


Table: Summary of repeated dose toxicity data of zirconium propionate and the individual constituents.


Zirconium dioxide
(CAS# 1314-23-4)

Propionic acid

(CAS# 79-09-4)

Zirconium propionate(CAS# 84057-80-7)

Repeated dose toxicity – oral

NOAEL(90day,oral,rat)> 3150 mg ZrO2/kg bw/day


NOAEL(90day,oral,rat)> 2331 mg Zr/kg bw/day*

NOAEL (chronic, rats)= 2640 mg/kg bw**

no data

Repeated dose toxicity – inhalation

NOAEC(60day, cat, dog, guinea pig, rabbit, rat)> 15.4 mg ZrO2/m³


NOAEC(60day, cat, dog, guinea pig, rabbit, rat)> 11.4 mg Zr/m³


NOAEC(30day, dog, rabbit, rat)> 100.8 mg ZrO2/m³


NOAEC(30day, dog, rabbit, rat)> 74.6 mg Zr/m³

No data

No data

* Identified as most sensitive endpoint in the registration dossier for zirconium, thus has been used for the DNEL derivation of this substance.

** Identified as most sensitive endpoint in the registration dossier for propionic acid, thus has been used for the DNEL derivation of this substance.



Repeated dose toxicity: oral

No effects were reported after oral administration to rats during 17 weeks of hydrated basic carbonate in form of moist paste containing 20.9% zirconium dioxide from a reliable study (Klimisch 2). The total intake of ZrO2 during the test period is 0, 0.9, 9 and 103.5 g. The equivalent NOAEL for ZrO2 is > 3150 mg/kg bw/day.

A similar study was performed on kittens, but the reported information from that study is limited and thus is just submitted as supporting information.


Repeated dose toxicity: dermal

No reliable studies are available for repeated dose toxicity via the dermal route of exposure. Testing is waived based on the following justification: a short-term (30 days) and sub-chronic (60 days) study are available for the inhalation route of exposure. According to the REACH Regulation, only one route of exposure should be tested for repeated dose toxicity (column 2 adaptation, annex VIII, section 8.6.1). Therefore, it is not necessary to perform a repeated dose toxicity study via the dermal route of exposure.


Repeated dose toxicity: inhalation

Two reliable studies were available for this endpoint (Klimisch 2): a 30 day repeated dose inhalation test in dog, rabbit and rat and a 60 day repeated dose toxicity test in cat, dog, guinea pig, rabbit and rat. No effects were reported in any of the species studied after inhalation of ZrO2 dust (NOAEC > 100.8 mg ZrO2/m3 air in the 30 day study and NOAEC > 15.4 mg ZrO2/m3 air in the 60 day study). These studies are used in a weight of evidence approach and support each other in the fact that no inhalation toxicity was observed after repeated exposure. The 28 days study is covered by the 30 days test. The 60 days study didn't observe effects after repeated inhalation exposure. Therefore the 60 days study is used to cover the 90 days study requirement.


Propionic acid

There are numerous reliable repeat dose toxicity studies in rodents in which propionic acid is administered via the diet. The primary focus of these studies is the local site-of-contact effect of propionic acid on the forestomach epithelial mucosa.


At BASF, groups of 20 male and 20 female Sprague Dawley rats weighing 128-136g were fed diets containing technical grade propionic acid in graduated doses (6,200, 12,500, 25,000 and 50,000 ppm) on daily basis for a period of 90 days. After the administration interval, 10 animals of each sex from the control, 6200 and 50,000 ppm dose groups were maintained for an additional 42 days recovery interval to determine the reversibility of potential effects. During the administration interval, no substance related clinical signs of toxicity occurred. No substance related systemic toxicity was exhibited by the test animals. Haematological and clinical chemistry parameters of the treated animals were within physiological limits and comparable to that of the control animals. Gross pathology revealed no adverse effects. In the fore stomach of the rats, histopathology revealed a dose dependent increase in the incidence and severity of proliferation-acanthosis and retention-hyperkeratosis of the forestomach mucosa was seen from the 12500 ppm dose group and above. These effects were more distinctive in females than in males. Reversibility of these effects was noticed after the 42 day post-exposure-observation-period. Based on these results, the LOAEL (local effects) for this study is 12,500 ppm propionic acid in the diet. The NOAEL (local effects) for this study is 6200 ppm propionic acid in the diet. No systemic effects were noted in the animals. The NOAEL for systemic effect was 50,000 ppm (BASF AG1971).


Similar effects were found in a 28 day feeding study when young Sprague Dawley rats (weighing 123-126 g) (10/sex/dose) were exposed to propionic acid in the diet at concentrations of 0, 10000, 20000 and 50000 ppm. The compound intake calculated from food consumption was 0, 936.0, 1836.0 and 4512.5 mg/kg bw for female animals and 0, 906.0, 1778.0 and 4415.0 mg/kg bw for male animals. During the entire study period, none of the animals showed clinical signs that were related to the administration of the test substance. Body weights and food consumption were comparable in all groups. There were no substance dependent effects on haematology, clinical chemistry and gross pathological parameters. No systemic effects were observed even at the highest doses. A dose dependent increase in the incidence and severity of proliferation-acanthosis and retention-hyperkeratosis of the forestomach mucosa was observed in the 10000 ppm dose group and above. The LOAEL (local effects) for this study was 10000 ppm or 936.0 mg/kg bw for females and 906.0 mg/kg bw for males (BASF AG 1970).


In another study, 7 weeks old male Sprague Dawley rats weighing 110-135 g were given daily diets containing 4% propionic acid (99% pure). Animals were sacrificed on days 2, 4, 7, 10, 14, 22 and 30 (3 animals/time point) and the glandular and non-glandular stomach subjected to macroscopic and histological investigations. Macroscopic examinations revealed thinning of the wall of the forestomach, prominent limiting ridge and local thickening/ulceration of the forestomach. Histopathology revealed that damage was manifested by reduced cellular basophilia and cellular vacuolation, and some reactive/proliferative changes in the limiting ridge (acanthosis and irregular depth of the epithelium, increased numbers of mitotic figures and basal cell hyperplasia). With regard to the onset of effects in relation to the duration of treatment, no macroscopic lesions were detected at days 2, 4 and 7. The majority of effects occurred from day 14 onwards; such a clear pattern was absent from the detailed histopathological findings although several of the observed changes in treated rats occurred somewhat more frequently towards the end of the study and some were absent during the very early periods. The overall incidence of treatment-related pathological changes was maximal (in numerical terms) after 22 days of treatment (OHC Laboratories 1990).


Similar results were also reported in yet another sub-acute study (up to 28 day) in which 7 weeks old Sprague Dawley rats (140-170g) were fed diets containing 4% propionic acid. Five animals in the treatment group and 3 in the control group were sacrificed after 7, 14, 21 and 28 days, respectively. A further group of animals (5 animals, recovery group) was fed test diet for 21 days, and then returned to control diet for 7 days until sacrifice at day 28. Mean body weight and food consumption were consistently lower in treated animals than in controls throughout the period of the study. When recovery animals were taken off test diet and returned to control diet, food consumption and body weight gain increased markedly. Macroscopically, thinning of the stomach wall and nodules on the forestomach mucosa were detected in both treated and recovery animals occurring after 14 days or more of exposure. Treatment-related histopathological lesions of the non-glandular stomach (limiting ridge) seen as early as 7 days post application included; acanthosis, basal cell hyperplasia, epithelial pallor, increased number of mitotic figures and intracellular vacuolation. No systemic toxicity was found. The LOAEL (local effects) for this study was 4000 ppm . The NOAEL for systemic toxicity was 4000 ppm (OHC Laboratories 1991)


In a repeated dose study performed to investigate the role of diea on propionic acid induced lesions, groups of 6 male Wistar rats (ca. 130g) were fed 0% or 8% propionic acid (=> 99% pure) in pelleted diet for 24 wks, 0% or 4% propionic acid, 1% L-carnitine or 4% propionic acid and 1% L-carnitine in powdered diet for 12 wk. Body weight, food consumption were monitored. Gross and histopathology of the stomach were performed after at termination. Stability of the propionic acid in diet was also monitored. Analysis of pellet samples, initially mixed with 8% propionic acid, revealed that at the time of consumption the actual level of propionic acid in the diet varied between 3.1 ± 0.7% (day 4 of storage) and 2.4 ± 0.7% (day 10 of storage). The recovery of 4.3% propionic acid in powdered diet (wet weight = 4% propionic acid) was 4.4 ± 0.3 % at feeding time, 3.5 ± 0.2% after 6 hr and 2.9 ± 0.5% after 24 hr. No treatment related clinical signs of toxicity or mortality occurred. 4% propionic acid in powdered diet or 8% propionic acid in pelleted diet did not influence body weights and body weight gain after 12 and 24 wks, respectively. Animals given 8% propionic acid in pelleted diet for 24 weeks did not develop hyperplasias. Macroscopical and histopathological examinations showed no changes in various parts of the gastric mucosa. After 12 weeks of 4% propionic acid administration in powdered diet, severe changes in the forestomach but not in the glandular stomach were seen. Grossly, crater-like growths with marginal hyperplasias and central ulceration were found in the forestomachs of all animals, particularly in the prefundic region (area proximate to the glandular stomach). The limiting ridge was substantially thickened. No histological examinations were made. Additional supplementation pf 4% propionic acid with L-carnitin did not influence the incidence nor did it influence the severity of the lesions found in the forestomach of the rat. In the forestomachs of rats fed 4% propionic acid in powdered diet, propionic acid accumulation in hyperplasias (1553 ± 508 µg propionic acid/g tissue) was three times higher than that in the remainder of the tissue (479 ± 247 µg propionic acid/g tissue). The propionic acid content decreased progressively towards the glandular parts. Supplementation of 1% L-carnitine to propionic acid powdered diet for 12 wk led to a significant decrease (P < 0.02) in propionic acid accumulation in hyperplastic tissues compared with hyperplasias induced by propionic acid alone, although L-carnitine had no influence on the severity of hyperplastic changes in the forestomachs of rats. In conclusion, the study presents evidence that the type of diet in which propionic acid is administered to rats is of great importance for the development of hyperplasia. Propionic acid or the corresponding CoA ester can accumulate in the forestomach and is preferably accumulated in hyperplasias and propionic acid accumulation does not appear to be directly related to the severity of hyperplastic changes (Bueld et al. 1993). 


In order to determine the time point of induction of the forestomach effects in rats, Rodrigues et al (1986) fed groups of 6 weanling male Fischer 344 rats (approx 75g) with diets containing 4% propionic acid (no data on purity) for 9, 15, 21 and 27 days at which point the rats were sacrificed, the stomach examined and prepared for autoradiography. Body weights and food consumption of the test animals were not influenced by propionic acid. The incorporation of methyl-3H-thymidine in to the mucosa of the forestomach was not influenced after 9 and 15 days but was enhanced significantly after 21 and 27 days of treatment (2 fold compared to control). Macroscopic and histological examination of the stomach of animals after 27 days showed mucosal thickening along the lesser curvature and also nodular thickening on the anterior wall of the forestomach (Rodrigues at al.1986). Feeding of Wistar rats, B6C3F1 mice and Syrian golden hamster (5/sex/dose/species) with 4% propionic acid (approx 99% pure) in the diet for a period of 7 days indicated that damaging effects to the forestomach epithelium and the induction of proliferative responses commence as early as 7 days following repeated exposure (Harrison et al 1991). Species differences susceptibility for induction of the damaging site of contact effects were seen with the rat being more sensitive (exhibiting low level damage; acanthosis, vacuolation and oedema) than mice which developed basal cell hyperplasia and rete pegs followed by hamster which exhibited nuclear vacuolations. Supposing that these changes are determinants in the aetiology of tumour formation of the rodents, the rats should yield more tumours than mice and hamsters.


In a dog study satisfying GLP requirements and OECD 409 TG, propionic acid (> 99% purity) was administered via diet to male and female Beagle dogs (4/sex/dose) for approximately 100 days at diet concentrations of 0, 3000, 10000 and 30000 ppm propionic acid. A recovery period of 6 weeks was allocated for the groups (4/sex/dose) receiving 0 and 30000 ppm propionic acid in the diet. No mortality occurred during the administration period. No substance related clinical signs of toxicity occurred. Calculated from food consumption, the mean daily dose administered were 214.2, 718.9, 2056.3 mg/kg bw for males and 225.1, 749.2, 2071.8 mg/kg bw for females. Dogs from the high-dose group displayed a decrease in appetite, which was attributed as a response due to un-palatability of the diet. This decrease in food consumption however did not seem to significantly affect body weights or body weight gains. No systemic effects were observed even at the highest dose. There were no significant changes in haematology, urinalysis, or clinical chemistry parameters that could be attributed to the test material. Necropsy of dogs after the administration interval revealed no gross lesions. Examination of tissues revealed no lesions except point-of-contact diffuse epithelial hyperplasia of the mucosa of the oesophagus in several high dose dogs. This effect was reversible after a 6 week recovery period. The incidence of focal epithelial hyperplasia in lower dose animals was comparable to controls. Based on these results, the LOAEL (local effects) for this study is 30000 ppm propionic acid in the diet or 2056.3 mg/kg bw for males and 2071.8 mg/kg bw for females. The NOAEL (local effects) for this study is 1% propionic acid in the diet, or 718.9 mg/kg bw for males and 749.2 mg/kg bw for females, respectively The NOAEL (systemic effects) was 30000 ppm propionic acid in diet or 2056.3 mg/kg bw for males and 2071.8 mg/kg bw for females (BASF AG, 1988).


Toxicological effects of propionic acid are practically solely due to the caustic properties of the concentrated substance. Propionic acid is practically of no concern for adverse systemic effects. As systemic NOAEL the highest test concentration in the long-term oral toxicity study is used to derive a systemic long term DNEL. This is a protective systemic DNEL based on the highest repeated dose test concentration available.


Zirconium dipropionate

Since no repeated dose toxicity study is available specifically forzirconium propionate, information on the individual constituents zirconium and propionic acid will be used for the hazard assessment and when applicable for the risk characterisation ofzirconium propionate. The NOAEL for systemic effects of propionic acid (NOAEL=2640 mg/kg bw/day, Griem et al. 1985) will be used for the DNEL derivation forzirconium propionate. This approach is justified since the caustic effects observed via oral administration of propionic acid can safely be excluded forzirconium propionate.

Further in case of the zirconium cation inzirconium propionate, the NOAEL of 2331 mg Zr/kg bw/day in repeated dose toxicity (sub-chronic animal data) will be used for the long-term, systemic DNELs.

Justification for selection of repeated dose toxicity via oral route - systemic effects endpoint:

Information from read-across substances:

animal data for zirconium: NOAEL(rat)=2331 mg Zr/kg bw/day

animal data for propionic acid: NOAEL(rat)=2640mg/kg bw/day

Justification for classification or non-classification