1. What Is Dissolved Silica and Where Does It Come From?
Dissolved silica refers to silicon dioxide (SiO₂) present in water as silicic acid, Si(OH)₄, a weak, uncharged molecule that remains invisible and tasteless at typical drinking-water concentrations. In natural systems, silica originates from the slow weathering of silicate minerals—feldspars, quartz, micas—in igneous and sedimentary rocks; rainwater percolating through granite, sandstone or basalt dissolves tiny amounts of SiO₂, eventually delivering it to aquifers and surface reservoirs. UK geology produces a wide gradient: soft, low-silica supplies in areas of Carboniferous limestone or slate (north-west England, much of Wales), and higher-silica water in regions underlain by Old Red Sandstone (the Midlands, parts of Scotland) or granite intrusions (Cornwall, the Cairngorms). Typical concentrations in treated mains water range from below detection (<0.5 mg/l) to about 20 mg/l SiO₂, well beneath the 50–100 mg/l found in some geothermal or volcanic spring waters.
Beyond geology, silica enters brewing water from man-made sources. Older asbestos-cement water mains, still present in parts of the UK distribution network, release chrysotile fibres and dissolved silica as they degrade; although these pipes are gradually being replaced, localised spikes can occur. Electric kettles with quartz or glass heating elements, and glass-lined boilers, leach small amounts of silica when water is held at high temperature for extended periods—one study of laboratory glassware found SiO₂ concentrations rising from 2 to 8 mg/l after ten successive boiling cycles in Pyrex. Reboiling the same water amplifies this effect, because each cycle evaporates pure H₂O and concentrates dissolved solids, including silica. For most households the increment is trivial, but users of large hot-water dispensers or commercial urn systems may see cumulative increases if fresh water is not regularly cycled through.
2. The Chemistry of Silicic Acid and Polymerisation
At pH 7 and room temperature, monomeric silicic acid (Si(OH)₄) is the dominant form below about 100 mg/l SiO₂; it carries no net charge, does not complex strongly with metal ions, and behaves as a weak acid with a pKa near 9.8, meaning it remains largely protonated and un-ionised across the pH range of brewing water (typically 6.5–8.5). This molecular simplicity conceals a dynamic equilibrium: silicic acid molecules slowly condense via dehydration reactions—two Si(OH)₄ units lose a water molecule to form a Si–O–Si bridge—creating dimers, trimers, and eventually colloidal polymers. The rate of polymerisation is highly sensitive to concentration, temperature and pH. Above ~100 mg/l or at elevated temperature, condensation accelerates; in boiling water, even moderate silica concentrations (20–30 mg/l) begin to form nanometre-scale colloidal particles within minutes, although the process is reversible on cooling if concentrations remain below saturation.
Tea brewing sits in an intermediate zone. Most UK tap water delivers 5–15 mg/l SiO₂ into the kettle, well below the polymerisation threshold; heating to 95–100 °C nudges the equilibrium towards condensation but does not instantaneously produce colloid. However, if the same water is reboiled multiple times or held at temperature in a hot-water dispenser, silica concentration can creep upwards through evaporation, and small colloidal nuclei begin to form. These colloidal silica particles, typically 1–10 nm in diameter, scatter light weakly (contributing to faint turbidity) and carry a slight negative surface charge at neutral pH, which allows them to interact electrostatically with positively charged or amphiphilic molecules in tea liquor—most notably polyphenol–protein aggregates and certain volatile aroma compounds adsorbed onto organic colloids.
The distinction between monomeric and polymeric silica matters because only the latter significantly affects mouthfeel and clarity. Dissolved Si(OH)₄ passes through the palate without sensory trace; colloidal silica, by contrast, has measurable surface area and can adsorb onto the tongue's mucin layer, modulating the perception of astringency and body. This is why two waters with identical total SiO₂ by analysis can behave differently in the cup: one freshly drawn and boiled once (mostly monomeric), the other reboiled or sourced from a high-temperature system (partly colloidal).
Polymerisation Is Reversible—Until It Isn't
Colloidal silica formed by gentle heating will often revert to monomeric Si(OH)₄ on cooling if concentrations stay below ~40 mg/l. Above that, or with prolonged boiling, irreversible gel networks can form—the origin of scale-like deposits in industrial boilers, though rarely visible in home kettles.
3. Silica, Polyphenols and the Formation of Tea Cream
One of the most visible manifestations of silica chemistry in tea is its role in accelerating 'tea cream'—the cloudy precipitate that forms when black tea cools. The classic mechanism involves thearubigins and other oxidised polyphenols complexing with residual proteins and caffeine; as temperature drops, these complexes become less soluble and aggregate into micrometre-scale particles that scatter light, creating haze or sediment. Silica does not directly bind polyphenols in the way that casein proteins do, but colloidal silica particles act as nucleation sites and surface scaffolds: their negatively charged silanol (Si–OH) groups attract positively charged patches on polyphenol–protein aggregates, drawing multiple complexes together and speeding precipitation.
This effect is concentration-dependent. In controlled experiments with model black-tea extracts, water containing 5 mg/l SiO₂ produced faint cream after several hours of refrigeration; at 15 mg/l the same extract turned noticeably cloudy within 30 minutes, and at 25 mg/l a visible sediment formed even at room temperature. The particles themselves are composite: a core of tannin–protein complex decorated with adsorbed silica colloid, sometimes incorporating trace calcium or magnesium if hardness is present. Because mineral balance and silica both influence cream formation, their interaction is non-linear—high-silica, low-hardness water can produce more dramatic cream than moderately hard, low-silica water, because calcium competes for binding sites on polyphenols and can inhibit the initial aggregation that silica later amplifies.
For the drinker, this means that high-silica water tends to produce clearer, brighter liquor immediately after brewing (because polyphenol aggregates are swept into larger, faster-settling particles) but cloudier tea on standing or cooling. The texture also shifts: teas brewed in silica-rich water often feel smoother and less puckeringly astringent, because a fraction of the soluble tannins has been sequestered onto colloidal surfaces before reaching the tongue. This is particularly noticeable in robust black teas with high thearubigin content; delicate greens and whites, which contain fewer oxidised polyphenols, show smaller sensory differences unless silica concentrations exceed 20 mg/l.
Silica Smooths, Calcium Sharpens
In side-by-side tastings, brewers often describe high-silica water as 'rounder' and high-calcium water as 'brighter' or 'cleaner'. The silica effect is subtle—a shift in texture rather than flavour—but becomes unmistakable in long gongfu sessions where mouthfeel nuances accumulate over multiple infusions.
4. Mouthfeel, Astringency and Ionic-Strength Buffering
Astringency—the dry, puckering sensation elicited by tea polyphenols—arises when tannins bind to salivary proteins (particularly proline-rich proteins) and mucins on the tongue, precipitating them and reducing lubrication. The intensity of this binding depends on ionic strength, pH and the availability of competing surfaces. Colloidal silica particles, with their high surface-area-to-mass ratio and weakly charged silanol groups, provide an alternative adsorption target: polyphenols can hydrogen-bond to surface Si–OH, and salivary proteins can adsorb onto the silica–polyphenol composite. The net effect is a reduction in the free tannin available to precipitate mucins directly, which translates sensorially into a softer, less astringent palate.
This buffering mechanism is distinct from the role of calcium and magnesium. Hardness ions enhance extraction of polyphenols (by neutralising negative charges on leaf cell walls) and can increase perceived astringency at high concentrations; silica, by contrast, mops up extracted polyphenols after they enter solution, moderating the harshness without necessarily reducing total polyphenol yield. The interplay is subtle: in moderately hard water (100–150 mg/l CaCO₃) with low silica, teas may taste bold and brisk; add 15 mg/l SiO₂ to the same hardness profile, and the brew often feels fuller-bodied yet gentler, with less of the 'grip' that some drinkers prize and others find excessive.
Quantifying mouthfeel is notoriously difficult, but trained panels have identified a sensory threshold for silica near 10 mg/l SiO₂ in green and oolong teas, and closer to 15 mg/l in robust blacks. Below these values, most tasters cannot reliably distinguish silica's contribution from background mineral effects; above them, descriptors such as 'silky', 'coating' and 'round' become statistically more frequent. The mechanism also depends on contact time: brief infusions (1–2 minutes) leave insufficient opportunity for silica–polyphenol adsorption to reach equilibrium, so mouthfeel differences are minimal; longer steeps (4–5 minutes) allow the colloids to mature and bind, magnifying the textural shift.
5. Haze, Clarity and Light Scattering in the Cup
Visual clarity is a prized attribute in many tea styles, particularly Japanese greens, high-mountain oolongs and first-flush Darjeelings, where transparency signals freshness and careful processing. Haze arises when particles in the 100–1000 nm size range scatter visible light (Rayleigh or Mie scattering, depending on particle size relative to wavelength). In tea, the usual culprits are polyphenol–protein aggregates, residual cell-wall fragments, and lipid droplets; silica adds another layer by forming composite colloids that persist even after the tea has cooled and traditional cream has settled out.
Colloidal silica particles themselves are too small (~1–10 nm for freshly polymerised silica) to scatter light efficiently, but they promote the aggregation of organic colloids into the 100–500 nm range—exactly the regime of maximum scattering. This is why high-silica water can paradoxically produce both faster cream formation (visible sediment) and persistent, fine haze: large aggregates settle quickly, but a population of intermediate-sized, silica-stabilised colloids remains suspended. The phenomenon is most evident in teas with high residual protein, such as late-harvest or autumn-picked leaves, and in very soft water where competing minerals are scarce.
For brewers seeking maximum clarity, the practical implication is that low-silica water is preferable, especially when combined with reverse-osmosis purification to remove organic precursors and trace metals that also nucleate haze. Conversely, some traditions—Chinese gongfu brewing of aged oolongs, for instance—value a certain opacity as a marker of body and richness; in such contexts, moderate silica (10–15 mg/l) may enhance the desired aesthetic. The key is awareness: haze is not a defect per se, but its origins and controllability matter when judging leaf quality or troubleshooting an unexpectedly cloudy pot.
Filter Before You Brew, Not After
Because silica-induced haze forms during heating and infusion, passing finished tea through a fine strainer or paper filter removes only large particles, not the colloidal fraction. To cut silica at source, use a reverse-osmosis or DI system on your brewing water before it enters the kettle.
6. Testing, Monitoring and Practical Thresholds
Unlike calcium and magnesium, which feature on most water-quality reports and are straightforward to measure with titration kits, dissolved silica rarely appears in consumer-facing data. UK water companies are required to monitor SiO₂ under drinking-water regulations, but the results are typically reported only in annual technical summaries, not on household bills. Home test kits for silica do exist—colorimetric methods based on the molybdate-blue reaction are standard in aquarium and boiler-water monitoring—but they require careful technique (acidification, precise timing, colour comparison) and are less forgiving than hardness-strip tests.
For practical purposes, most UK brewers can estimate silica by cross-referencing their water supplier's annual report with regional geology: areas drawing from Triassic sandstone aquifers (parts of the West Midlands, Cheshire) or granite-dominated catchments (Cornwall, Aberdeenshire) typically see 10–20 mg/l SiO₂; those supplied by Carboniferous limestone or chalk (much of the south-east, the Pennines) are usually below 5 mg/l. If precise control is desired—say, for competition brewing or quality-control work—purchasing a laboratory analysis (many commercial labs offer single-sample SiO₂ testing for £15–25) provides a reliable baseline. Alternatively, activated-carbon filters and ion-exchange softeners do not remove silica, so users of these systems should assume their input SiO₂ passes through unchanged; only reverse-osmosis units (which reject ~90–95 % of dissolved silica) or mixed-bed deionisers offer effective reduction.
The practical threshold for sensory impact sits around 10 mg/l SiO₂ for mouthfeel and 15 mg/l for visible cream enhancement. Below 5 mg/l, silica effects are negligible in all but the most controlled tastings; above 20 mg/l, they become reliably perceptible to casual drinkers. For those who discover high silica in their supply and wish to moderate it, the simplest interventions are (a) using RO or DI water for delicate teas where clarity and unmediated astringency are prized, and (b) embracing the silica profile for robust blacks and aged teas where a smooth, round mouthfeel is welcome. There is no 'correct' silica level—only informed matching of water to tea style and personal preference.
7. Interactions with Other Water Chemistry Variables
Silica does not operate in isolation; its effects intertwine with hardness, alkalinity, dissolved oxygen and carbonation to shape the final cup. Calcium and magnesium compete with silica for binding sites on polyphenols and proteins: in very hard water (>200 mg/l CaCO₃), calcium–tannin complexes dominate, and the additional presence of 15 mg/l SiO₂ may shift mouthfeel only marginally. Conversely, in very soft water (<50 mg/l), where few competing ions exist, even 10 mg/l silica can produce a noticeable textural change because it becomes the primary surface for polyphenol adsorption.
Alkalinity (bicarbonate) also modulates silica behaviour. High bicarbonate raises brewing pH, which partially deprotonates silicic acid (increasing the fraction of SiO(OH)₃⁻ ions) and accelerates polymerisation; this is one reason why hard, alkaline water in London or Birmingham sometimes produces cloudier tea than softer, lower-alkalinity supplies with similar silica content. Hydrogen bonding between water molecules and polyphenols is likewise affected by silica: colloidal particles disrupt the local water structure around tannin molecules, altering the energetics of extraction and precipitation in ways that are difficult to predict without direct experimentation.
Finally, vessel material introduces another variable. Ceramic glazes are silicate-based and can leach trace silica into tea during prolonged contact, especially if the glaze is alkaline or incompletely vitrified; iron and copper vessels contribute metal ions that co-precipitate with silica–polyphenol complexes, sometimes producing dramatic colour shifts or sediment. Glass and high-fired porcelain are the most inert options for brewers seeking to isolate the contribution of water silica alone, though even these are not entirely passive—glass kettles themselves can be a source if used repeatedly without fresh water changes.
Map Your Variables One at a Time
If you suspect silica is affecting your brew, hold hardness and alkalinity constant (by using RO water re-mineralised to a fixed recipe) and vary only the SiO₂ addition. This isolates the effect and prevents confounding with calcium or bicarbonate shifts that often accompany natural silica variation.
8. Sourcing, Adjusting and Choosing Water for UK Brewers
For UK tea drinkers who wish to control silica, the options fall into three tiers. First, accept the local supply as-is and choose teas that suit it: high-silica water from sandstone or granite aquifers pairs beautifully with malty Assams, roasted oolongs and aged shengs, where a smooth, coating mouthfeel complements the leaf character; low-silica chalk or limestone water is ideal for bright Darjeelings, crisp Japanese greens and high-mountain Taiwanese oolongs that benefit from unmediated astringency and maximum clarity. This approach requires no equipment beyond a basic understanding of regional geology and the willingness to experiment.
Second, reduce silica by installing a point-of-use reverse-osmosis system or purchasing pre-treated low-mineral bottled water, then selectively re-mineralise with calcium and magnesium salts (as detailed in our minerals guide) to craft a bespoke brewing profile. RO permeate typically contains <1 mg/l SiO₂, giving a blank canvas; adding back 50–100 mg/l hardness without silica yields water that extracts efficiently yet remains astringent and clear—a favourite among competition brewers and specialty importers. The capital cost (£150–400 for a domestic RO unit) and ongoing filter replacement make this economically viable only for serious enthusiasts, but the control is unparalleled.
Third, for those unwilling to invest in RO but seeking lower silica, standard jug filters and charcoal cartridges offer little help—activated carbon adsorbs chlorine and organics but leaves dissolved SiO₂ untouched, and ion-exchange resins target hardness ions, not silicic acid. The only jug-scale option is a mixed-bed deioniser cartridge (sometimes sold for car-battery top-up or laboratory use), which strips all ions, including silica, producing ultra-pure water that must then be re-mineralised; this is fiddly and rarely cost-effective for daily brewing. A pragmatic middle path is to blend: use bottled spring water known to be low in silica (Buxton, for instance, reports ~2 mg/l) for delicate teas, and tap water for robust blacks where silica's smoothing effect is welcome.
Finally, remember that kettle and vessel choice matter as much as source water. Avoid reboiling the same charge repeatedly, which concentrates silica and nudges polymerisation; if using a large hot-water dispenser, establish a routine to drain and refill weekly. Favour stainless-steel or enamelled-steel kettles over glass if you wish to minimise leached silica, and store water in glass or HDPE containers rather than prolonged contact with ceramic, which can release additional SiO₂. With these adjustments, even high-silica tap water can be managed, and the overlooked variable of dissolved silica transformed from mystery to tool.
Comments