Is thomsonite a zeolite?
Yes. Thomsonite is a zeolite: two IMA species, thomsonite-Ca and -Sr, with silicon and aluminium one to one. The TSi ladder, and how it differs from natrolite.
Yes. Thomsonite is a zeolite: the 1997 IMA zeolite report accepts it and both Handbook sheets, thomsonite-Ca and thomsonite-Sr, print Zeolite group. Its framework is half silicon, half aluminium: TSi 0.500 in the IMA formula NaCa2(Al5Si5)O20·6H2O, against 0.600 for natrolite. Thomsonite-Ca is hardness 5 to 5.5, density 2.23 to 2.39.
In short
- The definition and the verdict. Coombs et al., Canadian Mineralogist 35 (1997), the IMA zeolite report, define a zeolite mineral by a framework of linked tetrahedra with open channels and cages, usually holding water and exchangeable cations, dehydrating mostly below about 400 °C, largely reversibly. The report lists thomsonite among its accepted zeolites, with framework code THO, and both Handbook of Mineralogy thomsonite sheets print Zeolite group.
- Two species now. A script over the September 2026 IMA-CNMNC Master List finds two rows: thomsonite-Ca, NaCa2(Al5Si5)O20·6H2O, status Rn, 1997 s.p., United Kingdom, and thomsonite-Sr, NaSr2(Al5Si5)O20·6-7H2O, A, 2000-025, Russia. The 1997 report lists a single thomsonite. The thomsonite-Sr sheet’s analysis has Sr1.42 against Ca0.62.
- The TSi ladder. The report gives each zeolite a TSi, the proportion of tetrahedral sites occupied by Si. Thomsonite’s printed range is 0.50 to 0.56; of the 11 zeolites tabled below, its lower bound is the lowest, shared at 0.50 with willhendersonite. Natrolite, mesolite and scolecite sit at 0.59 to 0.62, clinoptilolite at 0.80 to 0.84. Worked on the IMA formula, Al5Si5 gives 0.500.
- Documented failures. TSi is an analysis, not a bench test. The sheet’s Old Kilpatrick analysis works out at 0.497, a hair under the report’s 0.50, and its Goble analysis at 0.552. On the balance a 20.00 g piece reads 11.03 to 11.63 g, overlapping natrolite at 10.91 to 11.15 g and scolecite at 11.11 to 11.27 g.
- When the answer changes. The report says thomsonite shares chain units with the natrolite framework but cross-links them differently, so a fibrous thomsonite is not a natrolite. prehnite, on thomsonite’s association line, is not a zeolite: see is prehnite a zeolite. The chabazite question is on is chabazite a zeolite.
| Zeolite | IMA formula, September 2026 | TSi from the IMA formula | TSi range printed in the 1997 report | Density (sheet); 20.00 g in water | Report framework code |
|---|---|---|---|---|---|
| Thomsonite (-Ca and -Sr) | NaCa2(Al5Si5)O20·6H2O; NaSr2(Al5Si5)O20·6-7H2O | 0.500 | 0.50 to 0.56 | -Ca 2.23 to 2.39; 11.03 to 11.63 g. -Sr 2.47; 11.90 g | THO |
| Willhendersonite | KCa(Si3Al3)O12·5H2O | 0.500 | 0.50, 0.51 | 2.10 to 2.18; 10.48 to 10.83 g | CHA |
| Gismondine (now -Ca, -Sr) | Ca2(Si4Al4)O16·8H2O (-Ca) | 0.500 | 0.51 to 0.54 | No sheet consulted | GIS |
| Gonnardite | (Na,Ca)2(Si,Al)5O10·3H2O | Not fixed by the formula | 0.52 to 0.59 | 2.25 to 2.36; 11.11 to 11.53 g | NAT |
| Natrolite | Na2(Si3Al2)O10·2H2O | 0.600 | 0.59 to 0.62 | 2.20 to 2.26; 10.91 to 11.15 g | NAT |
| Mesolite | Na2Ca2(Si9Al6)O30·8H2O | 0.600 | 0.59 to 0.62 | 2.26; 11.15 g | NAT |
| Scolecite | Ca(Si3Al2)O10·3H2O | 0.600 | 0.60 to 0.62 | 2.25 to 2.29; 11.11 to 11.27 g | NAT |
| Analcime | Na(AlSi2O6)·H2O | 0.667 | 0.59 to 0.73 | 2.24 to 2.29; 11.07 to 11.27 g | ANA |
| Chabazite (series) | Ca2[Al4Si8O24]·13H2O (-Ca) | 0.667 (-Ca); 0.750 (-Mg) | 0.58 to 0.81 | -Ca 2.05 to 2.20; 10.24 to 10.91 g | CHA |
| Heulandite-Ca | (Ca,Na,K)5(Si27Al9)O72·26H2O | 0.750 | 0.71 to 0.80 | 2.10 to 2.20 (heulandite sheet); 10.48 to 10.91 g | HEU |
| Clinoptilolite (series) | K6(Si30Al6)O72·20H2O (-K) | 0.833 | 0.80 to 0.84 | No sheet consulted | HEU |
What is a zeolite, and does thomsonite qualify?
Thomsonite qualifies: the IMA zeolite report accepts it and both Handbook thomsonite sheets print Zeolite group. Coombs et al., Canadian Mineralogist 35 (1997) define a zeolite mineral as a crystalline substance with a framework of linked tetrahedra, each of four oxygen atoms around a cation, containing open cavities in the form of channels and cages, usually occupied by water and exchangeable extra-framework cations, with dehydration mostly below about 400 °C and largely reversible. Thomsonite has its own entry in the report’s list of accepted species, written Ca2Na[Al5Si5O20]·6H2O, framework code THO.
What the sheet records. The thomsonite-Ca sheet: orthorhombic, pseudotetragonal; crystals commonly prismatic, acicular or bladed, flattened on {010}, elongated and striated parallel to [001], to 12 cm; in radiated spherical or columnar aggregates; smooth, globular, botryoidal; compact, massive. Twinning on {110}. Cleavage perfect on {010}, good on {100}; brittle. Hardness 5 to 5.5; density 2.23 to 2.39; pyroelectric. White, yellowish, pink, brown or greenish, may be concentrically zoned; vitreous to pearly. Biaxial (+), α = 1.497 to 1.530, γ = 1.518 to 1.544.
Where it grows. In amygdules and fractures in mafic igneous rocks, typically basalts, and in some alkalic and hypabyssal rocks and contact zones; an authigenic cement in some sandstones. Associates are zeolites, calcite, prehnite, datolite and quartz. The sheet calls it a common zeolite and names good crystals from Old Kilpatrick, Dumbartonshire, and Bishopton, Renfrew, Scotland; Mt. Monzoni, Italy; Kadan and Ústí nad Lábem, Czech Republic; the Faeroe Islands; Paterson, New Jersey; Table Mountain, Colorado; several Oregon localities; and Flinders, Victoria. The name honours Dr. Thomas Thomson, Scottish chemist and mineralogist of Glasgow, who first analyzed it.
How much silicon is in thomsonite? The TSi ladder
The TSi ladder: thomsonite’s framework is about half silicon and half aluminium, the bottom rung of the 11 zeolites tabled here. The 1997 report gives each species a TSi, the proportion of tetrahedron sites occupied by Si atoms, from published analyses. It prints 0.50 to 0.56 for thomsonite, 0.50 and 0.51 for willhendersonite, 0.51 to 0.54 for gismondine, 0.52 to 0.59 for gonnardite, 0.59 to 0.62 for natrolite and mesolite, 0.60 to 0.62 for scolecite, 0.59 to 0.73 for analcime, 0.58 to 0.81 for the chabazite series, 0.71 to 0.80 for heulandite-Ca and 0.80 to 0.84 for clinoptilolite. Sorted by script, thomsonite and willhendersonite share the lowest lower bound in that set.
Worked from the formulas. Thomsonite-Ca and thomsonite-Sr both carry (Al5Si5) on the September 2026 IMA-CNMNC Master List: TSi 0.500, Si/Al 1.00. Natrolite’s (Si3Al2) gives 0.600, chabazite-Ca’s [Al4Si8] 0.667, heulandite-Ca’s (Si27Al9) 0.750, clinoptilolite-K’s (Si30Al6) 0.833. The report describes thomsonite as varying approximately as Na4+x(Ca,Sr)8−x[Al20−xSi20+xO80]·24H2O, with x from about 0 to 2: more silicon comes with more sodium.
Worked on the sheets’ analyses, with a documented failure. The thomsonite-Ca sheet’s Old Kilpatrick, Scotland analysis corresponds to Na1.04Ca2.00Al5.03Si4.97O20: TSi 0.497, just under the report’s printed 0.50. Its Goble, Oregon microprobe analysis, Na1.09Ca1.70Al4.48Si5.52O20, gives 0.552, near the top. The thomsonite-Sr sheet’s analysis, Si5.05Al4.91, gives 0.507. The ladder places thomsonite chemically; it cannot be read without an analysis, which is why it is a classification fact and not a field test.
Is thomsonite the same as natrolite? Related chains, different framework
No. The report says thomsonite has chains with a repeat unit of five tetrahedra as in the NAT structure type, but cross-linked in a different way. Natrolite, mesolite, scolecite and gonnardite carry the code NAT in the report; thomsonite carries THO. Both families make fibrous to acicular radiating sprays in basalt, which is why they are confused; the natrolite-group comparison is on natrolite vs scolecite.
On the bench, documented failures. A 20.00 g piece of thomsonite-Ca reads 11.03 to 11.63 g in water; natrolite reads 10.91 to 11.15 g, mesolite 11.15 g and scolecite 11.11 to 11.27 g, all overlapping. Hardness 5 to 5.5 matches natrolite’s and scolecite’s. Cleavage is the better clue on a crystal large enough to show it: thomsonite perfect on {010} and good on {100}, the natrolite sheet perfect on {110}. The weighing method is on how to measure specific gravity at home.
Under UV. Neither thomsonite sheet prints fluorescence, by a ligature-aware search checked on the hemimorphite and willemite sheets; the natrolite and scolecite sheets do print it. Absence on a sheet is not a test, and this page gives no colour.
When the answer changes: thomsonite-Ca, thomsonite-Sr and old names
One name in 1997, two species now. The 1997 report lists thomsonite as one species, noting extensive variation in Na:(Ca + Sr). The September 2026 IMA list carries thomsonite-Ca, status Rn, 1997 s.p., and thomsonite-Sr, A, 2000-025, Russia; the sheets say the two form an isomorphous series. The thomsonite-Sr sheet’s analysis, from Mt. Rasvumchorr, Khibiny massif, Kola Peninsula, prints Sr1.42Ca0.62, and its name alludes to the predominance of Sr rather than Ca. Gismondine has since split the same way: gismondine-Ca, Rn 1997 s.p., and gismondine-Sr, A, 2021-043, Israel.
What a label should say. A thomsonite label without a suffix records the series; the Ca suffix needs an analysis showing calcium ahead of strontium. Both analyses on the thomsonite-Ca sheet show it, but they describe those specimens only. The same reasoning for chabazite is on is chabazite a zeolite and chabazite; zeolites in general on which minerals are zeolites; stilbite and analcime are covered as species. The classification questions are indexed at the collecting guides. Nothing on this site is offered for sale; the wanted list is the only commercial route.