The following publication fully covers the answer to your question and I found it very important:
Nano Titanium Dioxide Technical Information Sheet
Nanomaterial Health Hazard Review: Health effects of titanium dioxide nanoparticles
Industrial nanomaterials (also called nanoparticles) are defined as substances that are intentionally produced, manufactured or engineered to have specific properties and one or more dimensions typically between 1 and 100 nanometres. Most industrial nanomaterials come in varying sizes, shapes, and in some cases surface coatings.
Titanium dioxide (TiO2) nanoparticles, like the regular/bulk form of TiO2, exist in three crystalline structures or polymorphs: rutile, anatase and brookite. They can vary in size and shape, can form agglomerates or aggregates and be coated with other materials. Brookite TiO2 is less common than rutile or anatase TiO2. Samples of TiO2 nanoparticles may contain more than one polymorphic form.
All polymorphs of TiO2 have low water solubility. Hazards and biological activity of TiO2 nanoparticles may vary with its:
crystalline form (e.g. anatase TiO2 is considered more reactive than rutile TiO2);
particle size;
particle shape;
agglomerate/aggregate status;
surface properties; and
type of coating.
Industrial uses
TiO2 nanoparticles are reported to be widely used in cosmetics and paint. Rutile TiO2 nanoparticles are often used in sunscreens as ultraviolet absorbers, and anatase TiO2 in photo-catalytic and sensing applications. In Australia, responses to NICNAS Calls for Information on the use of nanomaterials (2006 and 2008) reported the use of TiO2 nanoparticles in cosmetics, secondary sunscreens and surface coatings. The 2006 responses also indicated use in water treatment and domestic products, but these uses were not reported by industry in 2008.
Hazard profile
The information below is a summary of national and international reviews and journal articles published up to July 2012, which are based on a limited number of studies on specific particle sizes of various TiO2 polymorphs. The information sheet was updated in late 2013 to include important national and international reports published after July 2012. Detail and references for all of the studies considered in this review are included in the Appendix under corresponding headings.
It is possible for TiO2 nanoparticles to be manufactured in various forms and sizes. As particle size and polymorph structure are important determinants of the properties of TiO2, the data summarised below may not represent health effects for all polymorphic forms and/or sizes of TiO2 nanoparticles.
Toxicokinetics—Systemic availability
Studies in rodents showing systemic absorption of TiO2 nanoparticles through the gastrointestinal tract, based on doses ~5000 mg/kg bw, are not likely to reflect realistic human exposures.
Repeated dose oral studies indicated bioavailability through the gastrointestinal tract, although the extent of absorption/bioavailability was not quantified. Studies on rats reported accumulation in the liver, spleen, lung and peritoneal tissues for 500 nm particles. For particles 5000 mg/kg bw) for particles ranging from 5 to ~100 nm.
The current weight of evidence for dermal penetration studies indicates that TiO2 nanoparticles do not reach viable skin cells. No acute dermal toxicity data are available to derive an LD50 value. Given that there is no significant dermal penetration, acute dermal toxicity in uncompromised skin is expected to be low.
The available data indicated low acute inhalation toxicity in mice for particles of 5 and 21 nm (LC50 >7.35 mg/m3/4h).
Skin irritation and sensitisation
TiO2 nanoparticles did not cause skin irritation or sensitisation in rodents.
Eye irritation
TiO2 nanoparticles caused mild eye irritation in rodents.
Repeated dose toxicity
A no observed adverse effect level (NOAEL) of 62.5 mg/kg bw/day is reported for anatase TiO2 (~5 nm) in female mice following oral gavage administration for 30 days. Serious damage to the liver (including increased enzyme activities, triglycerides and cholesterol) was observed at doses 125 mg/kg bw/day and above.
Two studies investigated the effects of intragastric administration of anatase TiO2 (~5 nm) in mice for 60 days. In the first study, accumulation of particles in the liver led to fatty degeneration, necrosis, inflammatory cell infiltration and hepatocyte apoptosis at 10 and 50 mg/kg bw/day. The hepatotoxic effects were considered to be due to alterations in gene expression levels and proteins which caused suppressed immunity in mice. Decline of neurobehavioural performance and morphological signs of brain damage in mice were reported in the second study, starting at 5 mg/kg bw/day. The effects at 10 and 50 mg/kg bw/day were attributed to the disturbances in the homeostasis of trace elements, enzymes and neurotransmitters in the mouse brain. The lowest observed adverse effect level (LOAEL) was reported as 5 mg/kg bw/day.
There are several short-term inhalation (4–13 weeks), or instillation studies in rodents with conflicting results. Two recent studies reported no inflammatory effects in mice and rats after inhaling TiO2 nanoparticles (20 nm anatase and brookite mixture and 35–60 nm rutile TiO2) for four weeks. There was evidence that chronic inhalation exposure of rats to high concentrations of TiO2 particles of both nano and non-nanoscale particles could lead to impaired pulmonary clearance due to particle overload. The associated adverse effects could include chronic inflammation, pulmonary damage, fibrosis, and lung tumours.
No repeated dose dermal toxicity studies were available.
Genotoxicity
Conflicting results have been reported in genotoxicity studies. Genotoxic effects secondary to inflammatory response and/or oxidative stress have been observed in both in vitro and in vivo studies.
Carcinogenicity
The International Agency for Research on Cancer (IARC) has classified TiO2 as possibly carcinogenic to humans (Group 2B), based on increased reports of lung tumours in rats under conditions of impaired lung clearance. The studies reviewed by IARC used a range of crystalline structures and particle sizes of TiO2. Therefore, it is not clear whether the lung tumours were due to size, crystalline structure or other differences. A 2-year inhalation toxicity study with ultrafine TiO2 (10–50 nm) in rats reported lung tumours at a dose of 10 mg/m3 (equivalent to the NOHSC recommended exposure standard for inspirable dust in general) compared with 250 mg/m3 for fine TiO2. This suggests that the nanoparticles may be causing overloading effects at a lower dose compared with fine particles. Further studies are necessary to investigate lung overload versus non-overload conditions in various animal species to make a conclusion on carcinogenicity from inhalation exposure.
Reproductive toxicity
Several reproductive/developmental studies in mice were available. Three reproductive/developmental studies in mice indicated very low translocation/accumulation in the placenta following inhalation of 21 nm coated rutile TiO2. Considerable translocation/accumulation in the placenta was observed following subcutaneous injections of 25–70 nm anatase TiO2, which caused adverse effects in the genital and cranial nervous system of male offspring. These studies indicated that translocation and accumulation of TiO2 nanoparticles into reproductive organs and/or placenta are necessary in order to exert any adverse effects.
Exposure and health effects
Reported uses of TiO2 nanoparticles in Australia are similar to the overseas uses — cosmetic and sunscreen products and surface coatings such as paint.
The size range for TiO2 particles is determined by the intended use. Pigment-grade TiO2 is optimised for visible light scattering, with particle size typically around 300 nm (Kroschwitz & Howe-Grant, 1997), although incidental presence of particles below 100 nm has been shown to occur. For transparent sunscreens, TiO2 particles in the nanoscale, which are too small to scatter visible light, are used. Anatase TiO2 and TiO2 mixtures containing anatase TiO2 at 85% have shown high photo-catalytic activity compared to rutile TiO2 or TiO2 mixtures containing less than 5% anatase TiO2 (SCCS, 2013). In many applications, particles are coated to reduce photo-catalytic effects and to aid dispersion in formulated products by reducing agglomeration. Coatings such as amorphous silicon dioxide may affect toxicological properties, by reducing production of reactive oxygen species. There is limited information indicating that aluminium hydroxide coated nanoparticles are generating free radicals under experimental conditions when exposed to sunlight in the presence of water containing chlorine. Further research is required to assess whether there is a risk to human health (TGA, 2013).
Cosmetics and sunscreens
Cosmetics and sunscreens used by the public will allow exposure through dermal (e.g. body lotions) and inhalation (e.g. cosmetic powders) routes. Limited incidental oral exposure is also possible.
The current weight of evidence suggests that TiO2 nanoparticles applied to the skin do not reach viable skin cells, but remain on the outer layer of the skin (TGA 2009; EPA 2010) and unlikely to cause harm when used as an ingredient in sunscreens (TGA, 2013). On this basis, health concerns are not expected from TiO2 nanoparticles in cosmetics and sunscreens under normal use conditions. However, data are limited on long-term use of sunscreens containing nanoparticles on hairy, damaged or aged skin, or flexural creases.
If photo-catalytic TiO2 nanoparticles (e.g. anatase TiO2) are present in a sunscreen, they can generate reactive oxygen species (ROS) following exposure to ultraviolet light (SCCS, 2013). The interaction of modern sunscreen formulations with surface coatings was investigated using 35 commercial sunscreens in Australia (Barker and Branch, 2008). Testing of ten samples with key compositional variations indicated that five out of these 10 samples contained TiO2 at 2.5 to 9% concentration, with four of these containing anatase and rutile mixture at 3:1 ratio.
Species-specific pulmonary inflammation is reported in animals following inhalation of all forms of TiO2. While IARC has classified TiO2 as 'possibly carcinogenic to humans' based on inhalation studies in rats with particle overload, inhalation exposure levels from use of cosmetic powders containing incidental TiO2 nanoparticles are expected to be low. Therefore, use of cosmetic powders containing TiO2 nanoparticles is not anticipated to cause lung overload, and consequently unlikely to pose adverse health effects in humans. A recent study reported that nanoparticles in cosmetic powders are in a highly agglomerated state and therefore, as would be expected based on the size of primary particles, the predominant deposition will occur in the tracheobronchial and head airways, and not in the alveolar region. However, considering the inflammatory responses observed in repeated dose inhalation toxicity studies in animals, the SCCS does not recommend the use of nano TiO2 in products that may have significant inhalation exposure (e.g. powders or sprayable products).
The incidental oral exposure is not expected to cause adverse health effects given the likely low exposure and low acute oral toxicity.
Based on the hazard information available, there are no significant concerns at this time from the use of TiO2nanoparticles with low photo-catalytic activity in cosmetics and sunscreens.
Surface coatings
Dermal exposure to TiO2 nanoparticles in surface coatings is not expected to pose significant systemic exposure due to the film-forming nature of surface coating formulations.
Inhalation exposure may occur during spray application of surface coatings. In occupational settings, spray applications need to be carried out in accordance with the Safe Work Australia National Guidance Material for Spray Painting [NOHSC (1999)]. Workplaces are required to maintain airborne concentrations of TiO2 below the exposure standard (10 mg/m3over an eight hour period).
On these grounds, additional control measures are not required for the use of TiO2 nanoparticles in surface coatings.
listed in the Australian Inventory of Chemical Substances (AICS); and
not listed in the Standard for the Uniform Scheduling of Medicines and Poisons (SUSMP).
TiO2 (CAS 13463-67-7) is:
listed in the Hazardous Substances Information System (HSIS) with no risk phrases but with an exposure standard of 10 mg/m3 time weighted average (TWA).
TiO2 nanoparticles are not listed separately in any of the above, therefore there are no nanoparticle-specific regulatory requirements in place for TiO2. However, all regulatory requirements applying to regular/bulk TiO2, will apply to TiO2 nanoparticles.