Sodium Metasilicate, Anhydrous [6834-92-0], Sodium Metasilicate Pentahydrate [10213-79-3], and Sodium Metasilicate Nonahydrate [13517-24-3] Review of Toxicological Literature, January 2002 (sodium metasilicate is sodium silicate, aka #toxicology #waterglass.) Notes that you can recover the metasilicate from ordinary silica gel by adding an alkali, such as sodium hydroxide, which sounds a lot easier than the silica-sodium-carbonate reaction at 1400°. Also notes that it’s sufficiently alkaline to be corrosive to many metals.
is an MSDS for #waterglass giving its pH as 12.4 in a 1% aqueous solution in the nonahydrate form.
#waterglass nonahydrate is “efflorescent sticky crystals”, while amorphous and pentahydrate are pellets. “Sodium metasilicate, International Programme on Chemical Safety, Poisons Information Monograph 500” #toxicology
The “abstracting abstract machines” paper by Van Horn and Might, introducing the AAM reduction formalism by transforming the #CEK machine (etc.) into an abstract interpretation of itself. (Also the CM machine.)
Search in the list monad in #Python and Perl.
Talc is hydrated magnesium silicate, H₂Mg₃(SiO₃)₄, so I guess that’s what you get if you precipitate #waterglass with magnesium salts. (Or not; see below.) #minerals
"Larnite" is calcium silicate, so I guess that’s what you get if you precipitate #waterglass with calcium salts. It’s Mohs hardness 6, a lot harder than talc. Occurring as a phase in concrete, it’s called “belite”. “readily prepared from reactive CaO and SiO₂ at 300 °C.” #minerals
"Olivine" is magnesium iron silicate, which is I guess what you get if you precipitate #waterglass with magnesium and iron salts? Or will the hydration interfere with that and give you a softer and weaker material, like talc? #minerals
No, wait, "forsterite" (an #olivine; Mohs hardness 7! Gemstone peridot!) is magnesium silicate, Mg₂SiO₄. I thought that was talc? Apparently talc is hydrated, so I’m guessing that precipitating #waterglass with magnesium is going to give you something more like talc than like forsterite, which also has a 2:1 Mg:Si ratio, compared to talc’s 3:4 Mg:Si ratio; so baking out the water probably wouldn’t be enough. #Refractory #materials: doesn’t melt until 1900°! (And iron ions, but only iron(II) ions, can replace the magnesium in any concentration.) But wait! There’s #enstatite! #materials #minerals
"Enstatite", a pyroxene, is yet another magnesium silicate mineral! It’s MgSiO₃, with a 1:1 Mg:Si ratio, in between forsterite’s 2:1 and talc’s 3:4, and it’s not hydrated. It’s around 5-6 Mohs hardness, and forsterite reacts with quartz to form it; apparently it’s possible to hydrate it and turn it into talc. There’s a pyroxene spectrum from enstatite to ferrosilite, FeSiO₃ #materials #minerals
Oh my. You can fire #talc to turn it into #enstatite and cristobalite at 1000° to 1200°. Be still, my beating heart! So you could maybe precipitate #waterglass with magnesium to form talc, machine the talc to precise dimensions, and then fire it to dehydrate it and precipitation-harden it into enstatite and silica? And then it’s refractory enough to cast silver in? That’s AMAZING. Can you also reinforce it with some kind of refractory fiber, like basalt fiber (declassified 1995!), as you precipitate it, so that it will have reasonable tensile strength? #minerals
YES, apparently you CAN precipitate #waterglass with magnesium salts. But you get AMORPHOUS magnesium silicate, as a fine powder. #materials
Hmm, you can synthesize nitric acid by heating magnesium nitrate (produced, perhaps, by double-exchanging nitrate anions with some other magnesium salt?) and absorbing the resulting nitrogen oxides in water, without using any strong acids. #materials
“#Forsterite, (Mg2SiO4) because of its exceptionally high fracture toughness, which is close to that of cortical bones, has been nominated as a possible successor to calcium phosphate bioceramics.”
“Lime #olivine” (which I guess is a calcium analogue of forsterite) “does not hydrate, and is avoided in #cement manufacture”, but apparently heating it above 500° “cures” that “problem”. Also called #larnite. Also, this page explains that forsterite weathers to serpentine and brucite in nature.
Hmm, magnesium hydroxide as a mineral is used as a flame retardant because it releases water when heated, and its formation is what destroys concrete in contact with seawater.
The weathering reactions of an #olivine like #forsterite and fayalite into serpentine in the presence of water, and also pyroxenes like #enstatite into talc and magnesian chlorite. Serpentine’s Mg:Si ratio is 3:2, in between forsterite’s 2:1 and enstatite’s 1:1.
Tetra-ethoxy silane, or "ethyl silicate" or "tetraethyl orthosilicate", forms silica upon exposure to water. You could use this to 3-D print silica directly by squirting it into water. It’s very hazardous, though, because it forms silica when the vapors get into your eyes. Apparently, in one of the most bizarre features of #toxicology, it’s safe to drink. Methyl silicate also exists but is even more dangerous because it hydrolyzes even faster.
Somebody back in the 1960s got a patent on using alkyl silicates related to #ethyl-silicate as a hydraulic fluid. This sounds like a terrible idea but they were actually talking about pretty long chain alkanes with 12 to 40 carbons per silicon, mixed with 1–30% ethylmethylsiloxane to thicken them, and those should be more resistant to the hydrolysis that makes ethyl silicate so dangerous. #patents
Oh wow, one of the hazards of #waterglass is that it “can etch glass if not promptly removed”. Also helpful: “Do not store in aluminum, fiberglass, copper, brass, zinc, or galvanized containers.” Well, I guess that leaves steel, iron, and polypropylene, huh? pH of 11.3 with a 37.5%-by-weight solution of sodium silicate.
More details on #ethyl-silicate as a silica-producing binder — apparently you can keep it liquid for quite a while by controlling pH. Suggests using it as a binder in refractories or to bind zinc dust to a surface for a corrosion-resistant coating!
OSHA PEL for #ethyl-silicate is 100 ppm; odor threshold is 3.6 ppm; 700 ppm is immediately dangerous.
Amateur chemists discussing how to produce ammonium silicate.
how to use #waterglass as a binder or coating.
Goodness, this "potassium silicate" sounds a great deal easier to produce than #waterglass; potash plus silica at 90°C! And it should still work as a silicate donor for insoluble silicates like calcium or magnesium silicate.
PQ says a NASA patent discloses how to produce a high-mole-ratio #potassium-silicate, patent 4,162,169. #patents
Hmm, #potassium-silicate paint is already a thing! Like, since 1878.
apparently you can dissolve #potassium-silicate in water more easily than sodium-silicate #waterglass, which usually requires pressure cookers to dissolve it. Also apparently potassium-silicate binders make mortar #thixotropic!
Common rust is iron(III) oxide (ferric oxide), so it won’t work to cross-link #waterglass, but magnetite has some iron(II) (ferrous) in it, and wüstite (FeO) is pure iron(II) oxide (ferrous oxide).
Oh interesting, actually very fine synthetic "magnetite" dust is a common thing, like for ferrofluids. This is promising for my plans to use magnetite to precipitate silicate.
handy farm devices and how to make them, from 1909.
Glauber, the dude who discovered Glauber’s Salt, precipitated #potassium-silicate using ferrous chloride — in 1646! I guess the “chemical garden” he produced was made of fayalite #olivine — or does fayalite have a softer hydrated version, like talc? (Apparently serpintinization turns fayalite into magnetite, silica, and hydrogen, so probably not.)
this explains the whole #waterglass insolubilization thing.
high-temperature properties of metal nitrates and nitrites.
#particle-filters for #speech-recognition. Specifically formant tracking. Zheng and Hasegawa-Johnson 2003.
open-source robot-audition software called "HARK" that does #beamforming and automatic #speech-recognition.
a 2012 (?) #introduction to #particle-filters, with lots of math and nothing about applications.
a book chapter about #speech-recognition with #particle-filters.
a bunch of #ebook materials, perhaps mostly under CC-BY. 2481 books totaling 39924 book chapters. All the books seem to be kind of farragos of unrelated research results, like issues of a periodical, and they seem to be unedited and barely literate in English: “DFT plays an important role in the implementation of algorithms into the systems for digital signal processing. The usage of the FFT algorithm significantly reduces the computation time of Fourier transformation. If you need to detect one or more tones in the audio signal or only one or a few frequencies, there is a lot faster method. The Goertzel's algorithm allows decoding of a tone (frequency) with much less processor load compared to the Fast Fourier Transform (FFT).”
stuff about compensating for frequency-selective hearing loss algorithmically.
The Seychelles actually imprisons a larger fraction of their population than the US does now. #human-rights
In Argentina we have 64 288 people imprisoned, 154 per 100k population, of which half are pre-trial detainees. #human-rights
The US State Department's report on Argentina's #humanrights situation.