Silicon and Silica

(endorsed 2025)

Guideline

Based on health considerations, the concentration of silicon in drinking water should not exceed 100 mg Si/L (equivalent to 210 mg SiO₂/L).

To minimise an undesirable scale build up on surfaces, the aesthetic value for silica in drinking water should not exceed 80 mg SiO₂/L (equivalent to 37 mg Si/L).

General description

Silicon (Si) is an ubiquitous chemical element mainly present in the environment as insoluble silicates (polyatomic anions containing silicon and oxygen) and silicon dioxide (SiO₂, a chemical compound commonly known as silica). Silicon is found naturally in foods as silica and silicates. High levels of silicon are found in foods derived from plants, particularly cereals, whereas silicon levels are generally lower in sources of food derived from animals. Silicone refers to manufactured polymers, also known as organosiloxanes.

Silica comes in two forms, crystalline and amorphous. Crystalline silica has a low solubility in water (approximately 6 mg SiO₂/L (Krauskoph 1956)) but amorphous silica has a higher water solubility (100-140 mg SiO₂/L (Alexander et al. 1954)). Dissolution of silica in water leads to formation of soluble orthosilicic acid (Si(OH)₄), which is the most bioavailable source of silicon. When the concentrations of Si(OH)₄ increase, polymerisation of the orthosilicic acid occurs, forming polysilicic acids, followed by the formation of colloidal silica where organic or other complex inorganic compounds are also incorporated (Alexander et al. 1954).

Amorphous silica is used as a food additive, in particular as an anticaking agent, but also to clarify beverages, control viscosity and as an antifoaming agent and dough modifier. It is also used as an anti-caking agent and as an excipient in pharmaceuticals for various drug and vitamin preparations (UK EVM 2003).

Silicon copper alloys with various compositions have been developed to induce grain refining and strength increase in alloys. Silicon copper alloys with various compositions have also been developed to produce lead- and arsenic-free copper alloys with good machinability for plumbing product purposes (SLR 2023) and have been identified as a potential replacement for lead copper alloys in plumbing products. In the United States of America, silicon copper alloys are one of the most common substitutes for lead copper alloys available on the market (ABCB 2021). Further information on lead replacements in plumbing products (such as silicon copper alloys) is available in Information Sheet 4.1 – Metal and metalloid chemicals leaching from plumbing products.

The deposition of silica from solutions can occur via various mechanisms. Dissolved silica will react with the surfaces of glass and metals and begin to form a white precipitate. In cases where customer complaints occur due to a scale build up, water hardness and silica concentrations should be investigated to determine the cause.

Silica can be a problem in water treatment due to its ability to cause fouling of reverse osmosis (RO) membranes (Sheikholeslami and Tan 1999; Ning 2002; Sahachaiyunta and Sheikholeslami 2002). This occurs because most of the dissolved silica cannot pass through the RO membrane and becomes supersaturated in the RO concentrate, causing silicates to form in the presence of metals, and these deposit on the membrane surfaces. The silicate then dehydrates, forming hard layers of silica scale on the membrane that reduce the effectiveness of the process and limit water recovery. A suggested industry standard guideline is to limit the concentration of silica within the RO concentrate to ≈ 120 mg/L at 25°C to limit fouling of the RO membrane (Freeman and Majerle 1995). Antiscalants are an option to reduce silica scale formation on RO membranes. Electrodialysis reversal (EDR) systems can be used instead of RO for water treatment where silica scale formation is problematic. The technology is designed to reduce scale formation and, in addition, the uncharged portion of silica passes through the electrodialysis process unchanged, both enabling higher water recoveries.

Colloidal silica may affect ion exchange processes in water treatment. The stability of colloidal silica as an un-ionised compound causes problems in removal using ion exchange resins. High concentrations may also cause fouling of ion exchange units.

Typical values in Australian drinking water

Concentrations of silicon measured by water suppliers in reticulated drinking water supplies around Australia are usually reported as silica (SiO₂). Concentrations of silicon can be calculated by multiplying the concentration of silica by 0.47 (the atomic mass of silicon divided by the molecular mass of silica (i.e. 28 ÷ 60)). For example, in 2019-2020, the Northern Territory reported average concentrations of silica in various water supply systems of 11 to 104 mg silica/L (equating to approximately 5.2 – 49 mg silicon/L). In Western Australia in 2019-2020, mean concentrations of silica in various water supply systems ranged from 0.6 to 90 mg/L (equating to approximately 0.28 – 42 mg silicon/L) (SLR 2023).

Treatment of drinking water

Limited data are available on the treatment of drinking water source waters to minimise silicon concentrations.

The removal of silica from waters can be achieved using processes such as cold lime softeners, hot process softeners, macroreticular anion resin, up-flow filters, cross-flow microfiltration and nanofiltration.

Emerging processes in water treatment, such as EDR and high efficiency reverse osmosis (HERO), have some success in silica removal. EDR is based on applying electrochemical separation processes within the water that allow for the removal of charged ions from the water. This process may offer some level of silica removal if the silica is in a charged form; however, colloidal silica, which is likely to be present in large concentrations, will not be removed due to its stability as an un-ionised compound. HERO is based on normal RO processes; however, pre-treatment removes divalent metals that form scale, particularly when (or just when) the operating pH has been increased to 11 to support the process. At this pH, silica is ionised, which increases its solubility and hence eliminates scaling of the membranes. It has also been reported that, at higher pH values, there is improved rejection of silica.

Measurement

Silicon is commonly measured in drinking water by inductively coupled plasma-mass spectrometry, inductively coupled plasma-optical emission spectroscopy or inductively coupled plasma-atomic emission spectroscopy, according to US EPA Methods SW-846, 3005A, 3010A, 3015A, 3051A, 6010, 6020, 6020A and 29. The standard limit of reporting ranges from 0.05 to 0.5 mg/L depending on the test method used (SLR 2023). This silicon concentration can be converted to an equivalent silica concentration by multiplying by 2.1 (which is the molecular mass of silica divided by the atomic mass of silicon (i.e. 60 ÷ 28)).

Silica can be determined by spectrophotometric techniques upon the addition of ammonium molybdate at pH 1.2; this silica is termed “molybdate-reactive” silica (EPA method 370.1, APHA Method 4500.F-SiO₂) (Clesceri et al. 1998). An equivalent silicon concentration can be calculated by multiplying by 0.47.

Health considerations

The toxicological database for silicon is limited. There is little evidence of adverse effects from oral exposure to silicon in humans. Some reports suggest a possible association of renal stones in humans following long-term use of magnesium trisilicate-containing antacids (EFSA 2010; FAO and WHO 1974), which can result in a dose of up to 4 g/day of magnesium trisilicate (EFSA 2018).

A review found that sub-chronic studies in rats identified no treatment-related adverse effects from dietary administration of about 370 mg silicon/kg bw/day as different forms of silicon (Newberne and Wilson 1970). Another review found a no observed adverse effect level of up to 2,500 mg SiO₂/kg/d in rats (equivalent to 1,165 mg Si/kg bw/day) and 3,500 mg SiO₂/kg/d in mice from a 2-year dietary study of micronized silica in rats (Takizawa et al. 1988).

However, in dogs, doses of 370 mg silicon/kg bw/day, when administered orally as sodium silicate or magnesium trisilicate, resulted in renal lesions (Newberne and Wilson 1970).

Renal effects have also been reported in guinea pigs exposed orally to an estimated dose of 16-32 mg silicon/kg bw/day as a suspension of magnesium trisilicate in drinking water (Dobbie and Smith 1982). These species are considered to have a higher sensitivity than humans due to differences in kidney function (EFSA 2018).

The review also found that other toxicological studies conducted in rats with micronised synthetic amorphous silica found no treatment-related adverse effects in these animals (SLR 2023).

Derivation of guideline

The health-based guideline value of 100 mg/L for silicon in drinking water was derived using toxicological data from a study that examined the health effects from silica exposure as follows:

100 mg Si/L=1,165 mg Si/kg bodyweight/day x 70 kg x 0.32 L/day x 100 \text{100 mg Si/L} = \dfrac{\text{1,165 mg Si/kg bodyweight/day x 70 kg x 0.3}}{\text{2 L/day x 100 }}

where:

  • 1,165 Si mg/kg bw/day (equivalent to 2,500 mg SiO₂/kg bw/day) is the no observed adverse effect level (NOAEL) based on a long-term (2 year) dietary study of micronized silica in rats (Takizawa et al. 1988; UK EVM 2003).

  • 70 kg is taken as the average weight of an adult.

  • 0.3 is the proportion of total daily intake attributable to the consumption of water. European data suggests that adults may ingest up to 50 mg of silicon per day from food and up to 500 mg of silicon per day from supplements, amounting to a daily intake of up to approximately 8 mg/kg bw (UK EVM 2003). 30% of the safe upper level for supplements (derived by the UK EVM) has been calculated as the relative contribution of drinking water to the total daily intake of silicon.

  • 2 L/day is the average amount of water consumed by an adult.

  • 100 is the safety factor in using results of an animal study as a basis for human exposure (10 for interspecies extrapolation, 10 for intraspecies variations).

  • The calculated value of 122 mg/L is rounded to a final health-based guideline value of 100 mg/L for silicon as per the rounding conventions described in Chapter 6 and is converted to an equivalent silica concentration by multiplying by 2.1.

The aesthetic guideline value for silica of 80 mg SiO₂/L is based on the solubility of amorphous silica being between 100 and 140 mg/L at 25°C. The solubility of silica is affected by both temperature and pH. Establishing a lower value should reduce the possibility of formation of silica scaling.

Review history

This fact sheet was developed based on a review of the available evidence completed in 2023 (SLR 2023; see Administrative Report for more information).

References

ABCB (2021). Lead in plumbing products in contact with drinking water Final Regulation Impact Statement Australian Building Codes Board, July 2021.

Alexander GB, Heston WB, Iler RK (1954). The solubility of amorphous silica in water. The Journal of Physical Chemistry, 58(6):453-455.

Clesceri LS, Greenberg AE, Eaton AD (1998). Standard Methods for the Examination of Water and Wastewater, 20th Edition. American Public Health Association, Washington.

Dobbie JW, Smith MJB (1982). Silicate nephrotoxicity in the experimental animal: the missing factor in analgesic nephropathy. Scottish Medical Journal, 27(1):10–16.

EFSA (2010). Selected trace and ultratrace elements: Biological role, content in feed and requirements in animal nutrition – Elements for risk assessment. EFSA Supporting Publications, European Food Safety Authority. 7:68E.

EFSA (2018). Re-evaluation of calcium silicate (E 552), magnesium silicate (E 553a(i)), magnesium trisilicate (E 553a(ii)) and talc (E 553b) as food additives. EFSA Journal 16(8): e05375.

FAO and WHO (1974). Silicon Dioxide and Certain Silicates, Food Agriculture Organization of the United Nations and World Health Organization.

Freeman SDN, Majerle RJ (1995). Silica fouling revisited. Desalination, 103:113 – 115.

Newberne PM, Wilson RB (1970). Renal damage associated with silicon compounds in dogs. Proceedings of the National Academy of Sciences of the United States of America, 65(4):872–875.

Ning R (2002). Discussion of silica speciation, fouling, control and maximum reduction. Desalination, 151:67-73.

Sahachaiyunta P, Sheikholeslami R (2002). Effect of several inorganic species on silica fouling in RO membranes. Desalination, 144: 373-378.

Sheikholeslami R, Tan S (1999). Effects of water quality on silica fouling of desalination plants. Desalination, 126:267-280.

SLR (2023). Evidence Evaluations for Australian Drinking Water Guidelines Chemical Fact Sheets – Lead Replacements in Plumbing – Silicon Evaluation Report. SLR Consulting Australia Pty Ltd. Report prepared for the National Health and Medical Research Council, December 2023.

Takizawa Y, Hirasawa F, Noritomi E, Aida M, Tsunoda H, Uesugi S (1988). Oral ingestion of syloid to mice and rats and its chronic toxicity and carcinogenicity. Acta Medica et Biologica, 36(1)27–56.

UK EVM (2003). Safe upper limits for vitamins and minerals. UK Expert Group on Vitamins and Minerals.

NOTE: Important general information is contained in PART II, Chapter 6.

Last updated