Material observations ⁑ Dernocua

Pouring sodium silicate solution into concentrated oil of lime hardens the surface enough that it doesn’t stick to the bottom of the yogurt cup, and you can pick up the resulting ooze by hand. Placing the ooze on a bed of vermiculite and pouring hot construction sand (300°?) over it produces foamed glass, which grows over a longer period of time than the foamed glass produced from dried waterglass, but with about the same density. However, it sticks to the sand and vermiculite as it grows.

Having burned through a cotton towel when pouring hot sand previously, I used a steel wool pad as a hotpad in this case, which worked fine. Obviously there are drawbacks to using such inflammable materials as insulating refractory, but in this case it didn’t ignite, and it worked fine to protect my fingers.

Having melted the aluminum parts of the previous butane-torch nozzle (the kind that uses 227-gram cans of butane), I bought a somewhat more robust one from the hardware store for US$6, with no evident aluminum or plastic parts and no striker. This can produce a substantially more exciting flame.

Borax

Bought some 700 g borax from the hardware store for some US$2. Smells slightly acidic, suggesting it’s maybe contaminated with boric acid. It successfully cross-links PVA glue (“plasticola”) into an insoluble mass. A hot wire causes it to adhere readily; application of heat causes it to foam up quite a bit, suggesting that it’s quite hydrated, and then collapse down into a glassy layer of, presumably, mostly boria, which seems to be able to mostly dissolve construction sand when reheated. This suggests that it really is borax.

Heating the borax on low heat on the stove on some aluminum foil produces a prodigious amount of white solid foam, similar to the intumescent foam from sodium silicate (see Glass foam), which can be cut with a hot wire. A drop of water on this foam has no immediately obvious effect, but after a few minutes, has dissolved a hole some 30 mm in diameter.

Quartz-halogen bulbs

The internal bulb in one of these halogen household bulbs can be heated to an orange heat in the butane torch without any softening, and even quenching in water does not crack them. This strongly suggests that it really is quartz. Too bad it’s burned out. These bulbs are hard to find nowadays, though some vendors have some remaining stock for under US$1 each, and 150W–500W exterior halogen floodlights still cost only US$2–US$4.

The replacement headlight bulb I got from an auto parts store (US$3) has the same behavior, except that I haven’t tried quenching it, suggesting that it’s also a quartz-halogen bulb. (Also, the box carries a pictorial warning against touching the glass.) This is a Magneti Marelli “H7 12V 55W PX26d”. Cold (10°?) it measures some 7Ω on a shitty multimeter that measures 3Ω short-circuit, demonstrating why household 240V quartz-halogen bulbs would be better for use as RTDs. When reheated to orange-hot, it measures 2Ω but the meter’s short-circuit resistance has dropped to 1Ω, demonstrating that the meter is totally inadequate for this purpose. For heating-element purposes such a bulb ought to be adequate up to nearly 1000°.

Pumice and garden stores

I bought 10 liters of off-white pumice from the garden store for US$2. It’s so porous that it floats, but it’s open-cell enough that sometimes it becomes waterlogged and sinks after a few seconds. It seems to weigh about 3.5 kg. Stone dimensions are on the order of 10–40 mm. It can be heated to orange heat in the blowtorch without apparent change. A hacksaw cuts it almost as easily as wood; a steel wire can bore a hole through it far more easily than through wood, and without cracking it in half. It can be easily sanded on granite or on other pumice stones, yielding a fine floury dust.

Though there is significant variation, it’s too hard to crush between my fingers, break with my fingers, or carve with thumbnails in its default form, though you can drill through it with a bamboo chopstick and enough effort. However, heating it to orange heat in the blowtorch renders it somewhat more fragile (presumably by thermal shock inducing microfractures), so that it can be broken by hand, though still not crushed between my fingers. Simply soaking it in water does not have such a weakening effect, but quenching it in water after such heating weakens the stone to an even greater degree.

One stone weighs 6.7 g. When added to 151.3 g of tap water (including yogurt cup), the reading was 158.1, a reading of 6.8 g. Later the reading was 157.9, and upon depressing it below the surface of the water with wires, 159.1, suggesting 1.2 cc of volume out of the water, or some 7.9 cc of total volume. Upon removing the wet stone from the water, the reading dropped from 158.0 to 150.2, a difference of 7.8 g, and the wet stone weighed 7.7 g, suggesting there was 1.05 g of water in its pore volume. This suggests a total volume, including pore volume around 9.0 cc and thus a dry density of 0.74 g/cc.

I stopped by a grow shop and was disappointed to find very little in the way of pure fertilizers and pesticides, less even than in the local garden store.

Carbonates and phosphates of iron

The garden-store green vitriol looks quite pure, though clumping together like a gorgeous green deliquescent thing, but the solution in tap water is a bit cloudy. Adding an excess of baking soda produces, after a while, voluminous green-brown bubbles that grow quite large and do not pop, overflowing my inadequate yogurt cup. A gray-green gritty slime with a smell like metal precipitates; this is probably a mix of carbonates and hydroxides of iron. Effervescence continues for an hour or more; probably boiling the soda first would have avoided it. Adding diammonium phosphate fertilizer instead in solid form produces no visible reaction at first, but upon heating produces a milky white precipitate, which later proceeds to form convection cells like those in miso soup.

02021-07-29

Iron products

Neither of the insoluble iron preparations have defecated to anything like transparency. Gravity filtration of the carbonate through a coffee filter leaves a dark green-brown mud on the filter and a fairly cloudy filtrate, which defecates. The mud smells like sulfur or metal, but the filtrate smells like dust. A second filtration still leaves a cloudy filtrate.

The phosphate is instead partly white, partly gray, with some solid chunks which presumably result from the green vitriol remineralizing the fertilizer prills without letting them completely dissolve. It has a similar metallic/sulfurous smell, and similarly the filtrate through coffee filters is cloudy. No ammonia smell is detectable. Excitingly, some of the phosphate product required a bit of work with a chopstick to scrape free of the epoxy (?) liner of the aluminum can. But this may have been because I overheated it when heating the mixture. I should maybe recrystallize some of the ammonium phosphate and figure out some of the stoichiometry.

Pumice grinding

Grinding the surface of the pumice with, for example, bamboo, tends to make the surface harder and less porous, because it fills the exposed bubbles with packed powder. When a hole thus drilled with a bamboo chopstick has recently broken through, there is greater risk of breaking the rock by wedging.

Vendors

Nearby vendors selling muriate of lime and similar materials include Kubra Quimica (JUANMANUELCURRA) in Haedo, ROAL4512404 in Ramos Mejía (aimed at making cheeses and yogurts), DUROPAVIMENTO in Ciudadela (aimed at making cheeses), and NAMECO QUIMICA in Villa Bosch (Kaiser 921, 5 km north, doesn’t list products on website), Química Kraff (KRAFF QUIMICA) in Morón (Av Estanislao Zeballos 1820, “cerca de la cancha del club dep morón”), Insumos Nahum in Ciudadela, ACUARIO MAKARIOS in El Palomar (Sargento Cabral 1205? El Rodeo 1416?), Isiquim SA (El Rodeo 1355 in El Palomar), Laboratorios Condisal SA in Lomas del Mirador (Tapalqué 543), Planeta Verde (TIENDAONLINEPLANETAVERDE) in Tapiales, maybe Distriquím (Soldado Caballero 8664?), Biosix SA (S. Ortiz 2498, piso 2, Ramos Mejía, surely the closest, doesn’t list products on website), Victor Eduardo Simo S.A. at Carlos Tejedor 5345 in Caseros (4750 6644, no website, Monica Barraza?, magnesium sulfate and fluorescein and phosphoric acid), Productos Químicos Alcesa) in Haedo (Paraguay 939), Silap (Sialp? Bacterint?) in Haedo (San Luis 661), Astra SA in Haedo (Argerich 536), Inquimec S.R.L in Haedo (Inalican 1154), Guillermo A. Rodríguez (Transquimia Chemist Cia Química) in Haedo (La Fraternidad 590), Ferar Química (B. Márquez 1235, Loma Hermosa), Química Morón (Hipólito Yrigoyen 1227, Morón), Productos Químicos SRL in Morón (H. Yrigoyen 625). Also there’s an Easy at Juan Bautista Alberdi 4950 in Caseros.

02021-07-30

Muriate of lime

Put 108.0 g of water in a 10.9 g cut-off can and started heating it up gently. Too bad I don’t have a thermometer. After heating to boiling it weighed 117.8 g. Adding (hopefully anhydrous) muriate of lime it had all dissolved up to a total weight of 235 g, so the amount of muriate was also 117 g. At 282.0 g total weight some of the crystals remained undissolved, but then did dissolve, with a weight gain to 287.6 g. I added more crystals to 322.1 g and a little bit seems undissolved; gentle heating dissolved them, reducing the weight to 321.9 g. Adding more crystals to get up to 341.9 g leaves some crystals undissolved. Further gentle heating leaves weight at 341.6 and dissolves all crystals; adding more crystals to 366.0 g leaves some crystals undissolved at first. Further gentle heating dissolves all crystals and leaves the weight at 364.1 g; adding more crystals brings the weight to 387.5 g. Further gentle heating dissolves all crystals and leaves the weight at 385.3 g; I added more crystals to 442.2 g.

And then I gave up. I think those dissolved too. I’m letting it cool now; a thin crust of solid has formed at the surface, preventing further evaporation.

After a couple of hours, the material in the can has the appearance of a solid, transparent block, but this is an illusion; once the crust is broken, the inside can be seen to be a mix of rather jagged crystals of muriate of lime and warm, syrupy oil of lime, which stick nicely to the finger.

NightHawkInLight’s intumescent refractory

Now I have borax so I should be able to make resistant carbonizing intumescent refractory materials. One person reports success with 10 tsp cornstarch, 1 tsp baking soda, 4 tsp PVA glue, and no extra water. NightHawkInLight is reported by WP to have switched from this recipe to cornstarch, flour, sugar, and borax, which was less prone to mold; in his video on the improved formulation he explains (3'47"):

At this point he shows his notebook with something like eight alternative recipes:

So, I guess I’ll try his recommended recipe: 40 g flour, 20 g corn starch, 20 g sugar, 20 g borax, and 25 g water.

The total ball of dough was 133.9 g, so I probably lost about 4.5 g that stuck to the mixing container or my hands or whatever. It smelled like cookie dough and had a tendency to kind of slump, probably due to having too much water. I was able to use it to cover the inside of a Monster can with the ends cut off and heat the can from the inside with the hand butane torch, and it did indeed foam up and produce an aroma similar to burnt toast and a carbon foam similar in hardness to insulating firebrick.

After letting the can cool, I turned the flame back on and ran it through the can again while holding it between my fingers; after about 30 seconds, the can became too hot to hold comfortably.

I heated up a small, thin piece (initially on a bed of waterglass foam, then inside the Monster-can “forge”) to see if I could cook it thoroughly into foam to get some kind of read on its material properties. During this process I noticed a second objectionable feature of this composition, aside from the burnt-toast smoke: it has a tendency to continue to smolder for a while after the source of heat is removed. (I wrapped the piece in aluminum foil to put it out without wetting it.) Also, significant cracks were opening up in the lining of the Monster can, although they seem to have been exposing uncooked dough rather than unprotected aluminum.

After all that, the piece turned out to be under 100 mg and so too small for my shitty scale to weigh, even when wet. It floated in water mostly below the surface, similar to the pumice, but it’s hard to tell how much of that is a question of water absorption.

In terms of fragility, it is definitely less fragile than the waterglass foam, and definitely more fragile than pumice.

Heating some more of this stuff with he blowtorch on top of the waterglass foam, I notice that each little black particle has a little crater around it in the waterglass foam, like snow around de-icing salt on a sidewalk. This suggests that either the carbon particles are producing heat to melt the waterglass around them, for example by burning (quite possible), or that the borax in them is fluxing the waterglass foam and allowing it to soften and collapse at a lower temperature (also possible).

One of the resulting chunks of carbon foam ranges from 15 mm in thickness to 25 mm in length and weighs 1.0 g.

Density measurement: I placed a 64.5 g cup of water on the scale. Immersing the foam in it beneath the surface with chopsticks yielded a reading of 68.1 g; removing it produced a reading of 61.8 g; weighing the foam separately yielded 3.7; immersing it a second time produced a reading of 67.8; weighing it a second time produced a weight of 4.0; a final weighing of the cup produced a reading of 61.4. Apparently in the first immersion it absorbed 2.7 g of water (both the difference in the foam and in the water), having displaced 3.6 cc of water; after the second immersion it had absorbed 3.0–3.1 g of water and was displacing 3.3 cc of water. So its total volume was 6.3 cc, including both the pore spaces that filled with water and those that excluded water, but not including spaces from which the water drained immediately upon removal. This is a pretty reasonable volume, given that 15 mm × 20 mm × 25 mm would have been 7.5 cc, and the shape was quite irregular. So the density of the foam as such is about 0.16 g/cc.

The water in this process turned dirty brown and had lots of black crap floating in it, so presumably there was a significant amount of lost mass. On soaking the foam a little longer, the water turned quite black.

After driving the water off from the foam that hasn’t dissolved with the torch, it weighed 800 mg, so only about 20% was lost.

I’ve flattened out a remaining piece of the dough to let it air-dry so that I can test its intumescent properties after drying.

Waterglass and starch intumescent refractory

It occurs to me that a 1:1 molar ratio of carbon sources such as cornstarch with silicon sources such as waterglass might be able to produce a foam that is partly or wholly carborundum (melting point 2830°), which would have mechanical and oxidation-resistance properties superior to carbon foam, and mechanical and melt-resistance properties superior to silicate glass foams. Boron carbide B₄C would probably be superior (harder than carborundum, softer only than CBN and diamond) but I think it’s probably even harder to produce than carborundum; Moissan’s original synthesis involved heating boria with carbon over the melting point of boron carbide, 2763°, though apparently magnesium permits the reaction under much less violent conditions.

Mixing a little waterglass with manioc starch produces a mixture with the consistency of warm chewing gum; I kept adding starch and kneading it in until it seemed like adding more starch would make it fall apart. (But I don’t know whether this is 1:3, 1:1, 3:1, or what.) The resulting rubbery, doughy substance is intumescent under the flame, producing a black foam with similar strength to that of the borax-based recipe above, though perhaps with a smaller cell size, but, as I was hoping, this foam has no tendency to smolder. Also, perhaps unsurprisingly, it stuck rather tightly to the foamed waterglass bed I was trying to heat it on, which the carbon/borax foam did not. I’ve left a fragment of this composition also to dry. While intumescing it produced a scent vaguely reminiscent of pine pitch, which strikes me as slightly alarming.

Both this foam and the waterglass-free version can be very readily cut with a hacksaw.

Waterglass, borax, and starch intumescent refractory

It occurred to me that maybe borax would be compatible with waterglass, and indeed you can mix quite a lot of borax and water into this waterglass solution without coagulating it in the way calcium or magnesium ions do. (I was concerned that perhaps the borate ions would be sufficiently amphiphilic to cross-link the silicate, but apparently not.) Mixing up manioc starch as before with this mixture produces another white doughy mixture which produces intumescent black foam when heated — though there are also bits of white foam, which may have been undissolved borax crystals. I’ve left a couple of these samples also to dry and see how they behave. One I heated immediately; it also has the unnerving pine-resin smell, and the foam it produces is the hardest yet — I can barely carve it with my thumbnail. I ran out of butane for the torch in the process, so I heated the sample for a longer period of time on the kitchen stove’s gas flame. Hacksawing through it reveals that although it has foamed up throughout its thickness, only the outermost 5 mm charred, leaving an inner circle of white uncharred material separated from the fissured outer carbon foam by a brown ring.

I can still break this foam material by hand, but it is hard enough that I cannot crush it between my fingers. Taking a 12-mm-diameter bit of it that weighs in at 0.65 g, I tried repeating the earlier density measurement but screwed up my measurements; I think the water weighs 81.9 g after pulling the wet foam out, 83.6 with it immersed, and the wet foam weighs 1.6 g. Or maybe it’s 80.7 g after the foam is removed. So it absorbed 0.95 g of water and displaced 1.7–2.9 g, which is really a damned wide range, so its volume including pores is 2.65–3.85 cc. Which is totally impossible, because (12mm)³ = 1.728 cc, so it’s somewhere in the range 0.17–0.72 g/cc. Not a good measurement by any stretch of the imagination. All you can say is that it floats if it’s dry. It turns out that it floats if it’s wet, too, barely, so I guess it’s below 0.4 g/cc dry.

So, I’ll have to redo the test with a larger piece, maybe after I get more butane.

The part of the foam that has been entirely charred is dramatically more fragile than the part that hasn’t.

It’s too bad I didn’t weigh out the ingredients, because this formulation in particular would have been pretty great to be able to reproduce.

I think the torch I’m using doesn’t get as hot as the torch I melted — it’s more for heating larger areas — so I haven’t been able to get any of these samples above orange heat. It’ll be good to be able to verify that they don’t lose structural integrity at higher temperatures due to, say, too much borax, too much waterglass, or too much thermal expansion.

The waterglass-based samples I’ve left to dry seem to be leaching waterglass out the bottom.

02021-07-31

The unfired waterglass bodies have hardened noticeably. Also, two of them have split, evidently from contracting as the surface dried.

I heated up one of them on the stove flame; it bubbled up pretty much as before.

Made a giant cookie using the flour/starch/sugar recipe, this time using baking soda instead of borax (of course), and adding some oil and vanilla. Also I added too much water, so I’m frying it in a skillet instead of baking it. First conclusion is that this is far too much leavening for a cookie or pancake. After cooking, it was even more inedible than beforehand; the surface caramelized to the point of being too bitter to eat. I discarded most of it, retaining a piece for future torch testing. Also, I keep belching from the uncooked batter I ate. The kitchen definitely smells more appetizing and less alarming now, though.

Grupo Ecoquimica at Avenida de Mayo 1761 in Ramos Mejía has zinc oxide (AR$4500/kg), epsom salt (AR$220/kg), mesh-80 calcium carbonate (AR$80/kg), borax (AR$220/kg), alum (AR$630/kg), citric acid (AR$290/kg), sulfur (AR$300/kg), bentonite (AR$50/kg), glycerin (AR$500/kg), kaolin (AR$150/kg), magnesium chloride (AR$210/kg), dipropylene glycol (AR$528/kg, see Dipropylene glycol), soy lecithin (AR$600/kg), cerium oxide (AR$13200/kg), boric acid (AR$275/kg), copper sulfate (AR$880/kg), titanium dioxide (AR$2100/kg), electrode gel (AR$360/kg), talc (AR$200/kg), infusorial earth (AR$167/kg), etc. Generally they sell in units of around AR$300-AR$500, so for most materials the limiting factor would be how much I could carry. It’s a few km away. Kubra and Insumos Nahum have propylene glycol, these guys don’t, just dipropylene glycol.

Oh, and it looks like the place I bought muriate of lime at the beginning of the pandemic was Química Industrial Caseros, at Bartolomé Mitre 4405 in Caseros, which is at the intersection with Lisandro de la Torre, which intersects Alvear at Día, not where I had thought. So that’s why I didn’t find it the other day when I went walking up that way: I was a few blocks off.

02021-08-02

The dried inedible pancake fragment, with its extra sugar and oil, and with baking soda instead of borax, can not only smolder but sustain a flame for a while after the torch is removed. Also, it evidently softens enough to deflect noticeably while the torch is on it.

The other dried/hardened samples (waterglass with starch, waterglass with starch and borax, flour/sugar/starch/borax) all produce hard intumescent foams as before. The waterglass-based samples don’t smolder.

The massive waterglass/starch/borax sample cracked in a few places on the outside as its crust shrank while drying.

02021-08-06

I ground up a spoonful of some diammonium phosphate fertilizer using scraps of granite as grindstones, noting an objectionable ammonia odor; sprinkled it onto some construction sand, previously dried with a torch, atop a bed of vermiculite; ground up a spoonful of what I believe to be some aluminum hydroxide, and sprinkled it on top; and heated the whole lot with the torch for a while. I never reached an orange heat for the whole mass, but the fertilizer did bubble out quite a bit of ammonia, and I often reached an orange heat for the surface. At one point I stopped, and the large amount of ammonia that began to escape persuaded me that stopping had been a bad idea, so I flambéd the mass a while longer until it stopped bubbling, about 10 or 20 minutes.

Upon cooling, I had a somewhat porous and brittle but surprisingly strong and hard crust of white material. It did not show any water-solubility or acid or base reaction (that is, upon dripping vinegar on it and letting the vinegar sit for a while, baking soda sprinkled on it would bubble, so it did not neutralize the vinegar; but upon being wetted with water, baking soda sprinkled on it would not bubble, so it did not neutralize the baking soda either.)

It’s probably worth making some better aluminum hydroxide, and trying the three pairwise combinations of the fertilizer, the construction sand, and the aluminum hydroxide.

At this point, despite this process having produced some ammonia and presumably a significant quantity of nitrogen oxides, my lungs were feeling fine. I tried heating a piece of the aluminum phosphate (?) crust thus formed by blowing the torch through the Monster-can “forge” with the borax/flour/etc. insulation described earlier to see if I could get it to a higher temperature and maybe melt it. I did get it to bubble a bit more, but it didn’t seem to melt bodily. After heating it in this way for several minutes, the intumescent insulation was smoldering pretty enthusiastically, and in places was peeling away from the can; after removing the gas flame, it sustained a flame for under a minute, then went back to smoldering. This left the house full of smoke that smelled like burnt toast and left my lungs feeling shitty.

I should probably use one of the non-smoldering, waterglass-based intumescent recipes.

02021-08-07

Still no water-solubility evident from the scrap of aluminum phosphate (?) in water.

Mina’s impression of the phosphate material was that it felt like a “fragment of wall”, that is, fully hardened portland-cement concrete that has exfoliated (maybe because it was improperly adhered stucco). That seems about right, but I think it might be slightly harder than that.

02021-08-08

I mixed some one-component hardware-store silicone (3M 420 transparente, vidrios y ventanas ) with the borax, then mixed by hand, adding more borax until it seemed like it was going to fall apart, to get a translucent ball some 25 mm in diameter, around 00:40. I poked and prodded it to see if it had become elastic or was still just visco-.

The borax is, I think, too coarsely ground to keep the silicone from being sticky; so I got sticky silicone all over my fingers, despite initial attempts to avoid this by using polyethylene bags as gloves. At least I seem to have avoided getting major skin irritation from the acetic acid. I was hoping for more of an Oogoo-style effect, but I realize that I ① haven’t made Oogoo and ② probably would need to have the borax finely ground. The idea at any rate was that maybe the water of crystallization of the borax would serve as a hydroxyl source for the silicone polymerization, allowing it to harden all the way through like Oogoo, and then maybe if the resulting object was heated, the borate would form a borosilicate network with the silicone backbones of the polymers. That evidently didn’t happen.

At 01:00 the ball was still pretty visco- and not very elastic, but at least it wasn’t sticking to my fingers anymore. Likewise at 01:24. The temperature in here is probably around 15°. By 07:30 the ball was mostly returning to its initial shape after being squished, and no longer cracking apart. My impression is that this is very similar to the behavior it would have if I had mixed in sand instead of borax — if it didn’t stick to the sand at all.

(Unfortunately I didn’t bother to note whether it had an acetic-acid smell at the time.)

I’ve been thinking about how to make these tests more rigorous, having just read Kingery’s earthshaking dissertation on phosphate-bonded refractories from 01950; right now I can’t even reliably tell if one mix is twice as strong or weak as another, or contains twice as much sand. It would be great to be able to do quantitative compressive and flexural strength tests, for example, which would require being able to cast bars of reproducible dimensions and fire them at reproducible temperatures. The tests mentioned previously with the phosphate fertilizer involved training a somewhat inconstant butane torch on some material until it stopped bubbling, which took about 20 minutes.

Doing this in a consistent fashion is going to involve setting up some kind of kiln that can maintain a reproducible temperature, for probably considerably longer than 20 minutes, since the temperature has to penetrate to the center of the bar instead of just a millimeter-thick layer. See Pocket kiln for thoughts on this.

Also, it would probably be useful to pre-purify the phosphate fertilizer, maybe using that recrystallization protocol by that African research group. And maybe pre-dissociating it to get 99% crystalline phosphoric acid would be useful for non-intumescent recipes, and also eliminate the problems associated with ammonia emissions. The phosphoric acid would have to be carefully granulated, though, to prevent it from being a serious inhalation hazard, and if it overheated (past 42.35°, assuming no freezing point depression from impurities) it would glom together into just a sticky mass. And I don’t have a prilling tower.

02021-08-09

Attempted to repeat the production of the apparent aluminum phosphate without success. I suspect the sand may have been wet enough to keep the temperature too low, but stoichiometry is another possible culprit. This time I had four areas heated to orange heat at different times: nothing (just construction sand), possible aluminum hydroxide, possible aluminum hydroxide with diammonium phosphate, and just diammonium phosphate. The grey aluminum compound remained inert to the heat in both cases; the phosphate bubbled orange as before, but only produced a cemented mass in the place where it was alone on the sand, not where it was mixed with the aluminum compound. This time, though, the cemented mass was acidic and black, and produced an acid gas while being flambéd. The aluminum compound looked for all the world as if no phosphate had ever been there. The phosphate flames were tinted green, a phenomenon I could barely see and hadn’t noticed before.

The waterglass+borax+starch pieces I’d made previously have become brittle and fragile as they dried out; I can break a massive (12mm) bar of it with my thumb, and upon dropping another piece on the floor it broke. Also, one piece that had initially hardened on a piece of PET is off-white on the air side and pure white on the plastic side, suggesting that the water accumulation I’d observed on that side led to a permanent compositional difference, perhaps the dissolution of all the borax. However, they haven’t developed visible cracks, and they still form an intumescent refractory foam upon flaming. It smolders for a few seconds upon flame removal, though, and upon continued flaming the foam melts somewhat; this suggests both that the composition doesn’t contain enough waterglass to suppress smoldering entirely, and that it doesn’t contain enough carbon to form a continuous carbon network that can remain solid even when the silicate network melts.

This probably suggests that I should consider using way less borax in these mixes. WP says non-alkaline-earth borosilicate glass is 12–13% boria, alkaline-earth types are 8–12% boria and 5% alkaline earths and alumina, and high-borate types are 15–25% boria with lower amounts of alkaline earths and alumina, the remainder being silica in all cases. So even a small amount of borax is probably adequate.

02021-08-10

I attempted to make three crude pieces of pumice flat by rubbing them together by hand for about 20 minutes, without external abrasives. The resulting deviations from flatness over distances of some 30 mm were on the order of 1 mm. In one of the three cases this was large enough to easily detect by touch, which seems to be due to an embedded hard stone; the other two look and feel flat (except for bubbles), even in glancing light, until you test them against a really flat surface. A great deal of white pumice flour covered the floor, its adherent nature posing difficulties for broom cleanup; also, it is slightly gritty between the teeth. Some of it has gotten into my keyboard and mouse buttons.

I filled a plastic shopping bag with methane, but to my disappointment it did not rise into the air, though it did burn nicely, leaving an unobjectionable candle-wax aroma. On calculation it seems to be about a sixteenth of the required volume. See Methane bag.

02021-08-12

The borax silicone from four days ago is rubbery all the way through. Washing it in water reveals that it is spongy and hydrophobic; drops of water can be squeezed out of its void spaces and bead up on the surface; evidently the water is washing the borax out of internal spaces, so it’s like a silicone sponge. Upon washing it in 95% ethanol instead, gritty borax comes out on my hands; a drop of water immediately dissolves it away into slippery wetness, confirming that it’s borax, or at least something alkaline. Cutting it in half with a razor knife makes some white noise as the knife encounters hard grains and leaves crystalline grains of borax on the razor knife and the exposed surfaces; they dissolve immediately in water. So on one hand there seems to be no uncured silicone, as desired, but on the other hand the borate mostly doesn’t seem to have been available to interact with the silicone. I suspect that probably I used enough crystalline borax to make the silicone effectively an open-cell foam, with the pores full of borax, perhaps allowing moisture from the air to penetrate. I can’t smell any acetic acid smell, but it has been four days. Perhaps a better-controlled version of the experiment would put the silicone and borax inside an airtight membrane, such as a party balloon or ziploc bag; this would prevent acetic acid from escaping or moisture from entering.

I mixed some Oogoo in the same way with cassava starch and the same silicone, inside a pair of polyethylene shopping bags. It never did reliably stop being sticky; when there was enough starch on the surface it would be nonsticky, but sufficient kneading would make it sticky again. I have left it to possibly harden. I was hoping I could mix it by kneading the inner shopping bag from outside, but instead I got silicone all over my fingers, which I washed off with dish detergent and water. No skin irritation was evident four days ago, and none is evident now.

After sitting for a while, its surface stopped being sticky at all; further kneading made it a little sticky again, but I didn’t continue kneading long enough to see if it returned to its original stickiness. I just rolled it back into a ball.

A few hours later (4?) the Oogoo thing had thoroughly solidified into a non-sticky smooth white rubber ball, which bounced nicely. Slicing it with the razor knife revealed that it was solidified all the way through (30–35 mm), so the cassava starch did succeed in avoiding the standard problem with household one-component silicone where it only cures in a very thin layer. It still has a noticeable smell of acetic acid.

Even though it’s cured all the way through, it’s still noticeably viscoelastic.

It easily sustains a flame once ignited. Burning it thoroughly with a butane torch converted it to a fragile carbon foam slightly larger than the original rubber, with the surface converted to a white ash, and produced a slight annoying smell of acid. The vermiculite near where it was burned received a gray deposit from the fumes. I wet three of the vermiculite stones in a (not pure polypropylene) bottle cap with a drop of water; baking soda then added did not fizz. So whatever the gray stuff is, it’s not noticeably acidic in water.

02021-08-13

In the end my lungs hurt for about an hour last night from, I guess, the smoke from the burning silicone.

The 3M 420 silicone I’ve been using costs AR$310 for 280 g, which at AR$180/US$ is US$1.72, or US$6.15/kg. The bisected ball from the other day weighs 19.0 g, probably half of which is cassava starch (AR$190/kg or US$1.05/kg), so this ball cost about AR$3 of material.

We have a glass microwave turntable we’ve been using as a plate, but it’s not very stable (it tends to tilt one way or the other, or spin around the middle, because it has three little glass projections near the center) so I’ve cut part of the ball into three little feet and glued them onto the bottom with more of the silicone. We’ll see if that works.

The Monster can of muriate of lime I dissolved/melted on 02021-07-30 has finally accumulated an appreciable amount of liquid on its surface by dehydrating the air.

At about 21:15, in one plastic shopping bag I put about 12.0 g cassava starch and 23.2 g silicone and kneaded them together through the bag, which produced a mass too sticky to remove from the bag; in another I similarly mixed 11.8 g silicone with 21.5 g starch, but despite extensive kneading I cannot get it to cohere into a single mass. It just acts like slightly damp cornstarch, easily crumbling away. I can’t smell the acetic acid anymore, but I have the impression that the lower-silicone crumbly version had a lot less acetic aroma to it at the time.

So I think the silicone-to-starch ratio has to be pretty close to 1:1; neither 2:1 nor 1:2 works well. I might try a third batch where I start with a ball of silicone in a vast oversupply of starch, then progressively knead in more starch, then measure how much starch remains unincorporated once I’ve produced a reasonable product. That seems like it might work better than trying to knead it through the bag. Like how you flour your hands when kneading bread.

After an hour and a half, though, at 22:47, I can peel the 12g-starch, 23.2-g silicone mix off the bag and roll it into a ball, which weighs 30.2 g out of the 35.2 g originally placed. So there’s probably about 5.0 g of silicone (and starch) still stuck to the bag that I didn’t manage to peel off. The ball is still annoyingly sticky to the touch. I can smell the acetic acid again, and it’s sort of chokingly strong.

After another half hour, at 23:13, the ball is no longer fully plastic; it tends to return to its original form after even fairly extreme deformations, so it is no longer kneadable. If this is typical behavior it’s fairly annoying: it went from annoyingly sticky on the hands half an hour ago to unworkable now. I wonder if a larger amount of starch would mitigate this behavior, because the ball last night was a lot more kneadable when reasonably unsticky. But maybe I just got lucky?

I mixed some more sodium silicate (waterglass) with cassava starch on another shopping bag to form a dough, which I tried to use to form a little pocket oven. This was unsuccessful, as over the following 20 minutes the oven gradually slumped down into a flat object; evidently sodium silicate with cassava starch behaves as a very viscous fluid, not a viscoelastic solid. Maybe gelatinizing the starch first would help.

I flambéd a small piece of it with the butane torch. It produced a little bit of smoke, which smelled like pine pitch — Mina reported that it smelled like mentholated wood. Upon long flaming it was able to remain smoldering for a few seconds after the flame source was removed, but not sustain a flame on its own. The carbonizing stuff swelled up like a sopaipilla in the torch flame. I was mostly able to hold it just in my fingers even though it was only a few cm long; at some points I resorted to needlenose pliers, which broke through the surface and exposed the hollow interior.

Mina reports that the sodium silicate bottle has a very strong smell resembling that of toilet cleaner (the kind based on muriatic acid), one which provoked nausea. I can’t smell anything from it, but I suppose there must be a lot of hydroxyls flying around if she can.

In between the hydroxyls from the sodium silicate, the acetic acid from the Oogoo, and the pine-scented smoke from the intumescent, we’re both kind of clearing our throats a lot.

By 00:00 the Oogoo seemed to be fully hardened but of course still preserves a quite strong acetic-acid aroma. I stuck a hardened steel shaft through it.

To estimate the water content of the liquid that had formed on the muriate salt, I took a shiny steel bowl (nickel-plated?) weighing 42.3 g and poured the salt solution out of the Monster can, bringing the weight to 47.3 g, so about 5.0 g of liquid had formed. Then I heated the bowl on the stove until it was apparently just dry white deposits of muriate of lime, though a little hissing was still audible. At this point the bowl weighed 45.0 g, so only about 2.7 g of muriate of lime remained; 2.3 g of water had boiled away. (Also, a little bit had been thrown out of the bowl by the violent sizzling bubbling — white specks were visible all around the rim of the bowl, so some must have also overshot.) After strongly heating the bowl on the stove a few minutes more, bringing parts of its floor to red heat while covering the top of it partly with aluminum foil, turning parts of the shiny nickel (?) black, producing a strong smell of metal, it still weighed 45.0 g, so if more water was lost, it was compensated by oxidation of the bowl. After washing the bowl with water, during which a bunch of the black stuff came off onto my fingers (NiCl₂? But that would be green and water-soluble. CrCl₃? Would be purple (but maybe I wouldn’t notice the difference)) and drying it, the bowl weighed 42.2 g, so evidently there wasn’t a lot of oxidation going on, so probably the white stuff from when it boiled dry was pretty much just the anhydrous salt, as you’d expect.

That would mean 2.3 g of water was enough to hydrate and dissolve 2.7 g of anhydrous muriate of lime, so something is clearly wrong here! According to Heating a shower tank with portable TCES? the solid hexahydrate should weigh 219.07/110.98 of the anhydrate form, or 5.3 g/2.7 g, so 2.6 g of water should have been required just to get to the hexahydrate. Moreover Wikipedia says the hexahydrate only dissolves 81.1 g per 100 mℓ of water at 25°, so to dissolve those 5.3 g we should have needed 11.9 g of total solution, another 6.6 g of water, or more at temperatures below 25°. It’s definitely cold enough here for the hexahydrate to solidify.

I’m not sure where my error is exactly. The Wikipedia numbers predict that 11.9 g of saturated hexahydrate solution should have been required to boil down to 2.7 g, not 5.0 g. Given how exactly the scale was able to reproduce previous measurements, it hardly seems likely that it had a 6.9 g error in measuring the weight of the solution I dripped into it, even if maybe it does have a little bit of silicone on the table under it. It measures the blob of Oogoo at 30.8 g, the one it measured as 30.2 g a few hours ago, but also I think a little more Oogoo got stuck to the ball after I initially weighed it. Maybe this stuff isn’t really muriate of lime — but then, what else could it be?

I should probably see whether, say, 20 g of salt from the bag loses mass if I heat it to red heat, and then how much water it takes to start to dissolve it at room temperature, which should tell me how much mass the hexahydrate has.

Interestingly, when I pulled the shaft out of the Oogoo to weigh it, it left a perfectly round hole through the middle of the blob. I guess it’s still flowing a little bit (or creeping a lot, depending on how you look at it). It’s 01:22, and I’ve stuck the original shaft back in, and I’m sticking a second shaft through it to see if the hole it made remains open, or round, when I pull it out tomorrow.

02021-08-14

The Oogoo is noticeably harder this afternoon. The second shaft left a hole smaller than the first, about half the diameter (maybe 1mm instead of 2mm; the calipers say the shafts are 2.94–2.98 mm, so I think that even the first hole is smaller than the shaft that made it); I’ve taken both shafts out and inserted one of them to make a third hole. It still smells of acetic acid.

The waterglass starch dough has mostly hardened in the position I left it in. It occurred to me that last night it was flowing liquidly when left alone for long periods of time, while breaking brittlely when I tried to form it rapidly, which is pretty much the shear-thickening behavior you’d expect for a suspension of starch in water, just at lower flow rates.

We found a stone mortar that had been missing. It was in the kitchen cabinet. Unfortunately, vinegar testing on the bottom shows that it’s marble, not agate, so it will contaminate anything ground in it with marble dust. Nevertheless I will grind some of the borax. Agate mortars are available for around US$800, which seems excessive when a chunk of agate of the right size is only US$30.

I used the mortar to grind some borax to a mostly flour-like consistency. I weighed it into the same bowl I used yesterday, which now weighs 42.3 g; after taring the bowl, the borax weighs 52.5 g. I took a Nutella jar which weighs 30.2 g; upon adding the borax, it weighs 82.6 g, giving a 52.4 g weight of finely ground borax. I’d probably need to screen the borax to separate out the remaining large crystals I missed in the grinding.

The borax tastes like baking soda (something I haven’t been brave enough to try previously) but does not bubble in vinegar. After adding vinegar to it, baking soda that is added still fizzes, but very slowly, which is consistent with the borax having converted the vinegar into sodium acetate and much weaker boric acid.

To try Oogoo with a little of this new finer borax, I mixed 10.2 g of cassava starch with 0.5 g of this flour-ground borax in the bowl, then transferred it to a shopping bag. In another bag, I mixed 10.3 g of cassava starch with 2.0 g of the same borax. To the shopping bag with the 0.5 g of borax I added 10.4 g of the same silicone; to the other bag with the 2.0 g of borax I added 13.2 g. In both cases I smooshed them around by kneading them through the bag until they seemed pretty uniform. Now it’s 01:48 by the laptop clock.

The idea is that the borax may do one or more of: provide borate ions that cross-link the silicone to make it harder; neutralize the acetic acid into sodium acetate, converting the borax itself into relatively inert boric acid, thus reducing the annoying outgassing; or release extra water, speeding the acetic cure. And by being more finely ground, maybe it will work better for these things than last time, and also not add so much porosity to the final material as my first borax silicone test on 02021-08-08.

At only 01:56 (8 minutes later) the two Oogoo batches are noticeably harder. I can already peel them off the bags. The ball from the other bag (with 2.0 g of borax) weighs 25.8 g, rather than the 2+13.2+10.3=25.5 g expected, suggesting a weighing error. The ball from the shopping bag weighs 20.9 g rather than the 0.5+10.4+10.2=21.1 g expected. I think I left more of the silicone stuck to the very thin shopping bag because it was harder to peel the bag away from the silicone, perhaps because it was much more wrinkly. But I guess it probably wasn’t more than about 3%.

It’s 02:09 now, 21 minutes after initial mixing, and the two chunks of Oogoo are very moldable and kneadable and still kind of annoyingly sticky, but at least they’re not leaving chunks of silicone stuck all over my fingers any more. They both have a strong acetic acid smell. So far, there’s no evidence that the borax has had any effect. That does sort of mean that it was ground finely enough to eliminate the obvious grittiness I had in the first borax silicone test. I took advantage of the kneadability to fold each one in half 20 times to mix them more homogeneously.

At 02:20, 32 minutes after initial mixing, they’re still very moldable and kneadable, and noticeably less sticky. The two versions seem pretty identical though. Most of the acetic-acid smell seems to be gone from the shopping-bag ball, the one with only 0.5 g of borax. Both balls look wet when being left without kneading for a few minutes, suggesting that the silicone is smoothing out rough surfaces with viscous flow. Kneading makes them look dry for a little while.

I think I’ll try kneading some construction sand into part of one of the balls. I dried some sand in the bowl, which weighed 70.7 g with the dry sand, and thus this is about 28.4 g of sand. I’m pulling off about half the other-bag ball, 11.9 of 25.8 g of Oogoo. After having kneaded in sand to 21.4 g (presumably 9.5 g of sand) it was starting to have noticeably less structural integrity, but I managed to incorporate essentially all the sand and still have a brown plastic mass weighing 39.0 g. I must have lost about 1.3 g of sand along the way. Mina commented that it felt like whole-wheat flour.

As plastic molding compounds go, this sand mix is not quite as cheap as pottery clay, but it’s probably watertight once it hardens at room temperature, and it wouldn’t be surprising if it were enormously more impact-resistant as well, though weaker and softer. The proportions are, I think:

10.3 g cassava starch × 11.9/25.8 = 4.8 g (12%)
2.0 g borax × 11.9/25.8 = 0.92 g (2.4%)
13.2 g silicone × 11.9/25.8 = 6.1 g (16%)
27.1 g sand (70%)
39.0 g total

1 kg of the stuff would cost 2.1¢ for the sand (US$0.03/kg according to Potential local sources and prices of refractory materials), 98¢ for the silicone, 7¢ for the borax at US$3/kg, and 12¢ for the cassava starch, for a total of US$1.19/kg.

It’s now 02:52 by my laptop clock. It’s still possible to fold the Oogoo over on itself an arbitrary number of times, but there’s noticeably more resistance, and quite a bit of elasticity, where it tends to return partway to its previous form. So at 64 minutes past initial mixing it’s already past its prime for molding.

02021-08-15

My hands and upper respiratory system were a little irritated last night; some of my fingers itched. I suspect exposure to acetic acid exacerbated by a lot of handwashing was to blame.

All three Oogoo samples from last night are fairly hard, as rubbers go. The sand-filled silicone is noticeably harder than the other two, which otherwise seem pretty similar. An acetic-acid smell is evident from all three.

Attempting to cut the sand-filled silicone with a razor knife quickly destroy the knife, within a few millimeters. I think that if I were to push it hard enough I could probably get it to keep cutting, because I can cut the stuff with my thumbnail. However, it’s rubbery enough that when I let go of the cut piece, it flaps back into place. Note that this is very different from the behavior I predicted for sand-filled silicone on 02021-08-08, when the borax didn’t stick to the silicone.

The sand-filled silicone definitely wets with water better than the other silicones. I haven’t measured a contact angle or anything, though. When heated up to 100° in a boiling-water bath, it wets completely with water.

At 19:55, I’ve been boiling the sand-filled silicone for about half an hour. It seems slightly changed: the surface feels slimy and leaves a slimy substance on the fingers, and it no longer smells of vinegar. But it feels about as hard as before.

I think the seal made between the 2.94-mm ground hardened steel shaft and the 2-mm-or-so hole it made in the 02021-08-13 Oogoo (2 hours and 45 minutes after being mixed) amounts to a usable sliding seal. It’s easy to spin the shaft with the fingers when it’s pressed through the silicone, and I’m reasonably sure that the hydrophobic nature of the silicone would keep water from sneaking past it. You’d want to use a thinner silicone washer in practice than the 30+ mm. A casual test with a drop of water on the slice of that block of Oogoo suggests that this would work, but of course that doesn’t tell us anything about how much pressure the seal could withstand.

Mina discovered that the Oogoo from 02021-08-13 is capable of erasing pencil marks from paper, much to my surprise. In fact, it seems to work better than a normal pencil eraser, although the graphite remains on the surface of the silicone instead of being carried away in crumbs. But getting the stuff to flake off in crumbs is presumably just a matter of adding enough of some kind of friable filler or filler with poor adhesion to the silicone, like the borax-filled silicone on 02021-08-08 that ended up making a fragile sponge.

At 20:22 the sand-filled silicone has been out of the boiling water for nearly half an hour but still feels wet. I suspect I gelatinized the remaining starch in its surface layer, and that’s now retaining water, and smears off on your fingers when you rub it on them. Rubbed on a wine bottle, it leaves a visible smear which does not wipe off with a dry paper towel, but does wipe off with a wet one, suggesting that it is indeed some kind of water-soluble thing like gelatinized starch. It still feels the same way another hour later, at 21:31. I wonder if it will grow mold (you’d think the borax would prevent that, but the borax might have been leached out of the surface layer by the boiling).

This slimy water-soluble stuff makes me wonder if maybe with so much starch the stuff might not be waterproof even at room temperature. It might be worth trying a wider range of ratios and fillers.

A permanent marker (Du Hu brand, UPC 9-930691-011013) easily marks the Oogoo (tried on the two sand-free samples from last night and the one from the other day), but it easily rubs off. A smiley face thus drawn on one of the Oogoo balls rubbed onto paper transfers onto the paper, suggesting the possibility of flexographic offset printing with this material.

By 02:41 the sand-filled Oogoo had stopped feeling wet on the top side (and stopped depositing water-soluble slime on wine bottles when rubbed) but still felt wet on the bottom (and still deposited slime).

02021-08-16

The 39 grams of sand-filled Oogoo had dried out by this afternoon. I pounded the crap out of it with a hammer against some bricks (maybe 20 hammer blows of 50 J each); it crumbled somewhat and became softer and porous, but mostly retained its structural integrity. The only pieces that came off were due to a couple of hammer blows where the corner of the hammer punched all the way through it into the brick on the other side. The surfaces are now whiter and rougher. It now weighs 38.0 g, and I regret not having weighed it before pounding it with the hammer, because I suspect that most of that weight loss was from boiling starch out of its surface, not from hammering damage.

This seems like noticeably worse impact toughness than I would normally expect from this kind of silicone, but then again, this object is only 16% silicone by weight. It’s kind of in the ballpark for what I’d expect from non-ferrous metals, although of course this impact toughness was achieved mostly through large deformations rather than large stiffness.

As I suspected yesterday, I can cut through the stuff with a razor knife just by pushing it hard enough through the material if I carefully avoid sawing back and forth, which immediately destroys the edge of the razor knife. By contrast, the knife edge is mostly intact when I just push through, though not without nicks from the sand. Pushing hard even works when the knife edge has been thus dulled by sawing, although it requires more force.

Using the unsharpened back of the razor knife, I can cut through the pumice quite readily — not quite as readily as cutting soft materials like cardboard or silicone, but very nearly. This further reinforces my impression that Foams Are A Miracle.

02021-08-19

The Oogoo samples from 02021-08-15 are of course fully hardened through, and can easily be cut with a razor knife. Perhaps more interestingly, with some sawing, they can be cut with a tin can lid as well, which takes more effort and leaves a scalloped cut. An acetic-acid smell is evident upon cutting, but it is not chokingly strong.

I got a new, hopefully less melty blowtorch head, made of brass (US$14), for those 8-ounce (227-g) butane bottles.

I also got some 5-mm carbon arc cutting electrodes (11.2 g, 5.11–5.19 mm diameter at the short unclad tip, 5.20–5.32 mm diameter in the copper-clad section, 306 mm length, thus average density around 1.69 g/cc). Treated as pencil leads, they leave only very light marks on paper (about 4H or 6H pencil lead hardness, say.). With significant difficulty I can break the end off the carbon rod with my fingers, revealing an earthy fracture, dull black with a few tiny bright particles, probably cleaved graphite grains. The sparkly carbon cladding resists water, ethanol, and nail polish diluent (probably ethyl acetate) with no damage; when heated, it forms iridescent coatings and then dull black copper oxide; and my multimeter can measure no resistance (though it bottoms out at about ½Ω), so it may be just copper plating rather than any kind of paint. The ohmmeter measures the graphite itself as highly conductive, but with significant contact resistance of a few hundred milliohms. Measuring the resistivity of the carbon rod itself would require removing the copper from it.

I heated one of the electrodes to orange heat with the brass torch. Mina used the hot end to burn holes in a cardboard box, which curiously smoldered and went out rather than igniting.

I flamed one of the waterglass/starch/borax samples from before, and it intumesced lightly as before.

I took one of the old waterglass-foam samples and melted it a bit with the new brass torch, which was relatively easy. Then I poured a bit of oil of lime (it keeps accumulating in the Monster can) onto the sample and tried again; the oil of lime seems to have been effective at increasing the heat resistance of the foam by, presumably, converting it to amorphous calcium silicate; now it apparently suffers no degradation even at white heat. Since muriate of lime melts at 772–5° the white-hot solid residuum cannot be merely muriate of lime.

The process filled the kitchen with a cockroach-like aroma, which, as Mina pointed out, is almost surely smoke from the bamboo chopsticks I was using to hold the waterglass foam sample in front of the flame.

I tried washing the sample with water after heating to remove the salt, which seems to have caused it to crumble without removing all the salt, because it still left a sticky residue on my hands after washing (more like oil of lime than plain salt) and created a white efflorescence on subsequent heating. This is probably a problem that could be solved by sufficient heating time and washing time.

You could also paint the waterglass directly onto the structure, but you might probably need to paint several coats, waiting for each one to dry in between, to get enough thickness. It would be pretty cool if you could include the oil of lime between the layers of waterglass instead of applying it afterwards, so that’s probably worth a try to see if I can get reasonably low densities that way.

It’s likely that such a low-density foam will have a more open-cell structure and thus more convection through it than the denser closed-cell foams that firebrick normally consists of, and therefore a lower thermal insulance. Even then, though, it might have better insulance per mass, so it might still be useful for portable pocket ovens.

The pieces of waterglass/starch dough from 02021-08-13 have thoroughly hardened. I flamed one of them with the brass blowtorch for a few minutes, and it intumesced, swelling by perhaps a factor of 2, forming a black foam which mostly did not melt, maybe just a little bit around the edges. Both the uncharred material and the charred material take noticeable effort to break by hand; the charred material is a fine foam with mostly bubbles around the size of 100 μm which can be indented with the thumbnail with some effort, while the uncharred material does not appear to be a foam and cannot be thus indented. It did smolder for a few seconds after the flame was removed, producing the pine-pitch-like smell I mentioned previously.

I had previously mixed up calcined alabaster powder with a retail baking powder (double-acting I think, and including bicarbonate, but I don’t know what else) and baked it in the oven into a white biscuit. It had formed an open-cell foam with about 50% porosity and pore sizes on the order of 2 mm. I took half of this and heated part of it with the brass torch to white heat for 7 minutes. This formed a black circle, with a dark gray circle inside of it, with a 20-mm-radius white circle inside of that, with increasing cracking visible in the gray and white parts. Upon removal of heat, it continued to glow visibly through the cracks for a minute or so. The white part flaked away somewhat under finger pressure; on breaking the biscuit, it seemed to have penetrated some 5mm deep. Unexpectedly, most of the biscuit thickness had turned gray, and it fell apart in my hand.

On repeating the process, I noted a slight acid smell, and my lungs became irritated, perhaps due to failing to maintain a judicious distance from the proceedings and producing a bit of vitriol from the alabaster.

I used the torch on the Monster can I’d previously (02021-07-30) lined with borax/flour/starch/water intumescent refractory to heat up some eggshells to a yellow or white heat for a couple of minutes. Unfortunately, the refractory material began to smolder, and the paint on the outside of the can started to change color, indicating that it was burning, so I aborted. There was also a distinct smell of ammonia, so perhaps the can was sitting on top of some undecomposed diammonium phosphate from a previous test. (Also, a smell of burning plastic from the can.) The lining had broken apart into many pieces, exposing much of the walls, and it continued to smolder for several minutes, even after I dumped it out of the can, so, again, the borax was not as effective at preventing combustion as one might hope.

I picked the eggshells out of the smoldering rubble with chopsticks (some of them had stuck slightly to the carbon foam) and dropped them into a plastic yogurt cup, then dripped a few drops of water into it. There was no visible reaction or evident heating, but after a few minutes the water looked milky. This seemed promising. The milkiness was due to tiny white particles. However, several drops of this milky liquid were unable to neutralize a couple of drops of vinegar to the point where baking soda wouldn’t fizz, so the milkiness probably wasn’t slaked lime as I was hoping it was.

Properly calcining lime probably requires longer heating, maybe 15 minutes to 15 hours. 15 days is more traditional, but I think I can do better than that.

I had the remaining fragments of waterglass foam floating in water for a few hours. They did not crumble any further, so whatever caused them to crumble earlier, it probably wasn’t the water — either it was my fingers’ roughness, or the process of boiling water out of them, or maybe there were parts subject to water attack and parts that weren’t.

I placed them on a bed of vermiculite and heated them with the brass torch to white heat. Once white heat was reached, they slowly melted, but they retained substantial foam structure for several minutes of such heating. (As before, untreated waterglass foam disintegrates upon being heated to merely orange heat in well under a minute, collapsing down to a glob of glass.) I don’t know if this is because the calcium treatment was incomplete or because lime silica glass can’t withstand the temperatures involved. The vermiculite remained unaffected, so the temperature wasn’t exceeding its melting point.

Some acid gas was evolved from the vermiculite, so I think it may be contaminated from a previous test.

02021-08-20

Mixed up some new Oogoo, this time with coloring. In a 126.5-gram glass jar Mina mixed 11.9 grams of dish detergent (see below) with a tiny amount of food coloring paste (see below). This produced a brilliant dark violet color. On weighing it afterwards, the scale read 137.9 g, which is 0.5 g less than before adding the coloring (so there’s only 11.4 g of detergent/coloring mix); I think there are about 0.5 grams of detergent mix left on the Q-tips.

For the Oogoo, I mixed 27.6 g of the same 3M 420 silicone as before with 27.6 g of cassava starch inside a tiny plastic shopping bag by kneading it through the bag. I think a Ziploc bag would be a much better approach; the thin (10-μm?) bag developed a hole in it during the kneading process. After it was thoroughly kneaded, we waited 9 minutes, then kneaded the resulting deposits into a 43.5-gram ball of Oogoo. To this we added 2.9 grams of the detergent/coloring mix, which I kneaded thoroughly together, getting a uniform light lilac color.

The dish-detergent-added Oogoo (≈6% commercial detergent mix, maybe ≈1% sodium lauryl sulfate and sodium laureth sulfate, but we don’t really know) was noticeably softer during molding and seemed noticeably less prone to stick to our hands, but also seemed to be less cohesive during kneading. (I now realize I should have kept some aside without detergent to compare post-cure properties.) I’m astonished that it was able to mix with the silicone at all; Mina told me it would work but I didn’t really believe her. I thought the water either wouldn’t mix in or would immediately polymerize all the silicone. (Surely the silicone polymerization does consume some of the water from the detergent.)

Unfortunately I neglected to weigh the ball after mixing the detergent into it, which would have been a useful clue as to how much of the detergent and coloring went into the Oogoo and how much soaked into my hands instead.

Mina shaped most of the ball into an eight-petaled flower, then painted part of the outside with food coloring/detergent mix left over on the Q-tips. I shaped a small mushroom from it.

I also neglected to measure how long it took to harden the silicone, but it did seem to be quicker than on other occasions.

Dish detergent: “5× concentrated”, lemon scent, Magistral UPC 7-500435-137900; biodegradable active surfactants alquilsulfato de sodio, alquiletoxisulfato de sodio y óxido de amina; other ingredients: etanol, agua, coadyuvantes, agente de limpieza, conservantes, colorante y fragancia, which is roughly as nonspecific as you could possibly get. “Materia activa minima 20%” might mean it’s 20% surfactants, or it might not.

Food coloring paste: Fleibor.com.ar; Azul T: azul brillante al 3.17%, bottled April 25, 02016, ingredients: azul indigotina, azul patente, tartrazina, propilen glicol, glicerina, azúcar, dióxido de silicio, which is a pretty intense cyan, and also Violeta L: azul indigotina y amaranto al 14.51%, bottled January 4, 02016, ingredients: entrocina, amaranto, azul indigotina, azul patente, propilen glicol, glicerina, azúcar, dióxido de silicio, which is a pretty intense magenta.

At 17:20, the color/detergent mix painted onto the outside of the flower is still wet, but the Oogoo is quite firmly set.

At 17:26 I lined a dry steel bowl with aluminum foil (the 10μm stuff I might have mentioned previously, maybe Ecobol shitty store brand; folded in half 5 times to make 32 layers, I measure 0.32 mm with the calipers) and poured waterglass into it, then covered the top with aluminum foil and put it on low heat, the intent being to make some more waterglass foam, but this time with an aluminum foil backing on one side. At 17:31 and 17:37 and 17:45 it’s crackling in a threatening fashion. At 17:51 it has died down quite a bit. At 18:05 it seems to have stopped completely, and the top aluminum foil is distended upwards; I can feel a hot, hard round shape through it, which is presumably waterglass foam, suggesting that this approach to foaming waterglass has actually been enormously more successful than I had imagined it could be. I might have had 2 mm of waterglass in the bottom of the bowl and now it’s apparently like 70 mm tall! That suggests a considerably larger void fraction than I’ve been able to achieve in the past.

I turned it off at 18:56 after cooking on low heat for an hour and a half, at which point it was maybe making some tiny little crackles you could barely hear if you put your ear next to it. I’m eager to see what the foam structure looks like, but I want to let it cool slowly to reduce cracking from thermal deltas (I can hear slight crackling, which is probably precisely that). The whole bowl, foil, foam, and all, weighs 57.6 g, but I’m not sure if this is the same bowl I previously weighed empty at 42.3 g or a different one. It’s about 170 mm in diameter. If the loaf’s volume is ⅓ of the cylinder it fits inside, ø170 mm × 70 mm, that would be 530 mℓ. If it really weighs 15.3 g and occupies 530 mℓ, that’s 29 mg/cc, half the density of the foams I made previously. I wonder if it’s just one giant bubble, though.

At 19:24 it seems to be silent and the hard shape distending the top of the foil is cool to the touch.

At 19:27 there is only the occasional click from it. I removed it from the bowl. It weighed 15.5 g. I had hoped it was filling the whole interior of the bowl, but that’s not what happened at all; it’s probably only about the same density as the previous samples. Because the heat was being applied from below, the bottom part of the waterglass, stuck to the lower foil, hardened first. Then, as the rest of it gradually hardened, the upper part expanded, causing the waterglass layer to curl and form a large empty space below it, some 30 mm deep and 70 mm across. The overall mass of waterglass froth, which looks very much like soap-bubble froth from well-shampooed hair, is some 90 mm across.

With some care I was able to peel almost all of the upper aluminum off of it, tearing off aluminum tags in only three places (removed successfully with the fingers), and leaving invisible fragments of waterglass on the aluminum in a few more places that can be felt with the fingers. The outer soap-bubble-like film is very fragile, and on a couple of cases as I weighed it and opened the package, I have seen bubbles of glass floating around in the air. A small puff of air from my lips can shatter some of the foam into glistening fragments.

On the tongue the delicate bubble surfaces feel a bit like gelatin bubbles, but have no taste, leaving a slight slipperiness and tingling for a few minutes like some soaps. (At this point it’s worth mentioning that this is not a very alkaline waterglass formulation; it has a lot of silicon per sodium.) When touching the denser parts of the froth with my tongue, they do not immediately dissolve, but remain solid, as you would expect for sodium silicates. However, nothing gritty comes off in the mouth.

The aluminum foil I peeled off the top weighs 1.7 g, and the remaining waterglass foam and bottom foil together weigh 13.8 g. Since the top and bottom foil are about the same size this suggests that the foam alone weighs about 12.1 g.

Concerned that the tiny glass flakes from the bubbles might cause skin irritation, I rubbed the froth lightly on my slighly sweat-damp left thigh to deposit a bunch of shiny bubble tops, then rubbed it around thoroughly until the deposit was just a fine glitter, so that if the microscopic sharp glass edges causes irritation mechanically (like fiberglass) I will be able to observe it there. Either I will get a hell of a rash on my thigh or I can tell people I’m Edward Cullen.

Kinesthetically the foam feels like it’s about 25 or 30 mm thick in the middle, maybe 15 mm thick near the edge. This permits a crude volume estimate as a 120-mm-diameter circle (due to the edge curling) that’s 20 mm thick, which would be 230 mℓ. 13.8 g in 230 mℓ would be 61 mg/cc, which is about the same density I’ve gotten in the past. This might be slightly less dense because of the bottom foil.

(Later measurements, after cutting, show that it was about 20 mm thick in the middle and about 10 mm thick near the edge, occasionally as much as 15, but 120 mm is about right.)

I heated one side of the waterglass mass with the brass torch and, as before, the surface immediately melted at orange heat, accompanied by a great deal of crackling and a few flakes flying away in the air, then settling nearby. After 20 seconds of heating with the torch, it looked like styrofoam whose surface has been eaten by gasoline or acetone. I poured some saturated oil of lime over another part of it, waited about 20 seconds, heated it gently with the torch at a distance for about 30 seconds to drive off the water, and then heated the lime-treated surface strongly for about 20 seconds. The treated surface melted only very slightly at yellow-white heat. This left a white efflorescence. Touching it with my fingers leaves the sticky oil-of-lime feeling on them, so I suspect that only some of the oil of lime was consumed.

I tried to film this with my cellphone but can’t figure out how to get the exposure right in video, so the blowtorch-affected area is just white. The video has some other problems, like being 310 MB. I was able to take some stills, especially after I learned about the voice-activation feature of the cellphone.

There was some of the “cockroach” smell from heating the lime-treated part of the waterglass, even though no chopsticks were present, and maybe also some acid gas. This suggests that I might be smelling some kind of muriate or calcium product, or possibly something that’s present as an impurity in my industrial-desiccant source.

After all this it weighs 16.4 g, which suggests that the oil of lime added about 2.6 g, including the part that’s still wet on the still-attached aluminum foil. Removing as much as possible of the aluminum foil reduces the weight to 14.2 g. In the process I noticed that the bottom aluminum foil had actually torn in a few places, looking rather like stretch marks from rapid weight gain. Also, it’s not very firmly adherent; it’s possible to peel it off in bits in places. The waterglass attached to the aluminum foil is noticeably porous, though much less than the surface; I can push my thumbnail through the foil into the waterglass.

At 20:52 my thigh is still glittery and not irritated. It’s been maybe half an hour since I applied the potential irritant, but if I recall correctly from my childhood, skin irritation from fiberglass can take hours to develop. At 21:22 still no irritation.

I passed some water from the faucet over one side of it to see if it would have a similar effect to the fire, but it seems to have only affected a few of the largest and most delicate bubbles.

I cut it in half with a razor knife, mostly using the back of the blade. For this it was convenient to hold the foam between my thumb on the aluminum-foil backing and a finger on the untreated blowtorch-melted area, which is partly covered by a network of dense glass which can withstand finger pressure without crumbling. Cutting through the area that had been treated with oil of lime and then melted was much more difficult. I dropped half of it on the floor from a standing position when I finished; a small piece broke off from the impact.

With my teeth, I took a tiny bite of one of the areas affected by the water, about 21:34. The material is gritty between my teeth, like toothpaste, rather than dissolving immediately upon being reduced to fine powder. No large sand-like grains are present. It’s still gritty at 21:36. Only a little grittiness is left at 21:44 but probably because I swallowed the grit rather than because it dissolved.

If this foam is 15 mm thick, 120 mm in diameter, and weighed 12.1 g (the guess above), it’s 71 mg/cc, which is a little denser than some I’d made in the past (for example by rapid heating). According to Potential local sources and prices of refractory materials, liquid waterglass costs US$2/kg, which probably means the solid form costs about US$6/kg (I missed an opportunity to find this out today by not weighing the bowl before cooking the foam). At 71 mg/cc that would be US$430/m³ or 42.6¢/ℓ, which is slightly more expensive per volume as other lightweight insulating mineral aggregates mentioned in that file such as pumice (17¢/ℓ), LECA (29¢/ℓ or 12¢/ℓ for construction), rock wool (31¢/ℓ), perlite (38¢/ℓ), and vermiculite (23¢/ℓ), though these are mostly denser: pumice is 400 mg/cc, LECA is 1200 mg/cc, rock wool is 100 mg/cc, perlite is 128 mg/cc, and vermiculite is 60–160 mg/cc. Plausibly, this waterglass foam could provide better insulation than these other materials, being lighter in weight than most of them, and being a solid mostly closed-cell foam rather than an open-cell foam or a loose particulate aggregate.

This experience suggests a few ways of making sandwich panels out of waterglass foam:

Part of my intent with this sample was originally to carve it into some kind of interesting shape and then melt the shape’s surface to make it sturdy (and maybe also cover it in aluminum foil), but I got caught up in other stuff I guess.

I cut out a small cuboid to measure density. It floats in water without getting waterlogged or sinking, which (in combination with its aggressively hydrophilic nature) suggests that most of the structure is closed-cell foam, which is good news for use as insulation. The outermost part does seem to absorb water; perhaps its cells were open when it formed, or perhaps they were originally closed but broke easily. After observing this, I realized I was an idiot, because I hadn’t weighed the cube dry, so I rinsed the cube and converted the bowl into a dehydration chamber: I stuck the cube in the bowl on top of a plastic lid on top of some anhydrous muriate of lime. The idea is that if I dry it by just heating it up, I’d possibly damage its foam structure.

I tried draping some aluminum foil over the top of one of the cut pieces of the waterglass foam and melting it on with the brass torch, but this didn’t work very well; the foam under the melted aluminum foil had a tendency to rip it apart, and aluminum foil after having been melted tends to break easily, and it looks dull and wrinkly. By contrast, the original underside foil remained fairly shiny. I think a better way to put an aluminum-foil surface on a shape carved from this foam might be to coat one side of the foil with liquid waterglass, then wrap it around the foam, then cook it.

This process also evolved some acid gas, I think from the vermiculite bed I was resting the foam on for this process.

02021-08-21

The cuboid, probably mostly dehydrated, weighs 900 mg. I really should get a 10-mg-resolution scale! Or I should have cut out a much larger cuboid.

I wonder if I could get waterglass to gel with some agar? Agar is supposed to gel at pH from 2.5–10 and concentrations in the 0.5%-1% range. Then maybe I could make some aluminum-foil waterglass tape with the waterglass gelled with agar, then wrap it to form a thing, then let it dry or treat it with polyvalent cations or with CO₂ before heating. I don’t think I have any agar, though.

With great difficulty, I bit through one of the Oogoo samples from 02021-08-15 — the borax one, I suspect, from the taste and the soapy tongue feel afterwards. A slight acetic-acid smell is still evident. After being rinsed, it’s relatively straightforward to tear the piece the rest of the way, which is also true with an unrinsed piece that I cut partway through with a razor knife. However, the tearing process is only easy if done very slowly; the fibers of PDMS bridging the growing crack gradually yield, neck, and break.

I folded a bag from aluminum foil (maybe 20 mm × 150 mm) and poured a few grams of waterglass into it. On heating it with the brass torch on a bed of vermiculite, it swelled and crackled, and the waterglass inside foamed up and eventually hardened, and then the aluminum on top melted off. I added another torch, double-fisting the butane. Eventually it seemed to become inactive, so I turned the torches off and turned it over (the bottom was still shiny foil), then heated it some more. Upon turning it a second time the vermiculite stuck to it in a syrupy mass, probably because the waterglass had poured out through the melted-open holes in the aluminum into the vermiculite. Heating the waterglass-cemented vermiculite cemented it together. As usual, the foam melted and collapsed rapidly whenever the torches brought it to an orange heat.

(It occurred to me that the vermiculite might be cemented with moist phosphoric acid from some previous tests, and there was indeed some acid-gas smell present, but I haven’t actually seen any syrupy phosphoric acid going around. I got some of the syrupy vermiculite stuff on my fingers and didn’t wash it off for several minutes, but have no burns, so it was probably waterglass.)

I was hoping that I could get a chunk of waterglass foam entirely covered with aluminum foil this way, but I think for that to work, I will need to heat it more gently — maybe quickly, as with hot sand, but to a temperature below the melting point of the aluminum.

I never did get any detectable skin irritation from rubbing my thigh with waterglass micro-bubble fragments last night.

The food coloring painted with detergent onto the flower Mina made yesterday is still wet, presumably due to the detergent. The mushroom I made is still much, much softer, especially at large deformations, than the other Oogoo I’ve made in the past without color or dish detergent, and so is the flower. Upon inflicting large deformations on them with my fingernail, I observe some white discoloration, which is probably permanent. I suspect that the dish detergent is acting as a sort of plasticizer. The feeling is sort of in the neighborhood of stretchy human skin, an illusion that becomes more pronounced when I rub the mushroom with cassava starch (inspired by vague memories that the “Cyberskin” skin simulant is a silicone with talcum powder added.)

02021-08-22

I broke apart one of the chunks of cemented vermiculite by hand, finding white foam between the vermiculite grains, which I think confirms that it was leaked waterglass and not contaminating phosphoric acid. This is more evident when I cut it with a razor knife instead of breaking it by hand, because breakage by hand tends to cleave it along cleavage lines within the weak vermiculite grains.

Bought more dish detergent, the Unilever Cif version, also 20% surfactants, except they tell you what they are. AR$175 (US$1) for 500 mℓ, which works out to about US$10/kg for the sodium laureth sulfate etc., which seems a bit pricey to me.

02021-08-23

I checked MercadoLibre again today. Grupo Ecoquimica has raised all their prices by about 25–50%, presumably to compensate for recent inflation; for example, alum is AR$850/kg (US$4.70/kg) instead of AR$630/kg, and bentonite is AR$75/kg (42¢/kg) instead of AR$50/kg. It’s about 4 km away. I was planning to go, but didn’t.

The PVA glue I crosslinked with borax on 02021-07-28 has quite thoroughly dried out. I cut the chunk in half with a hacksaw, which took a couple of minutes (it’s about 20mm in diameter) and felt a lot like cutting a solid hard plastic, with no melting on the saw blade, a thermoset. It’s hard all the way through, and it smells slightly of PVA inside. This made me wonder whether it was in fact a thermoset.

Shibayama, Yoshizawa, Kurokawa, Fujiwara, and Nomura published a paper in 01988 which suggests that this slime definitely can gel, and higher pH raises the gel’s melting temperature, though they mostly measured gels that melted around 75° with 0.8–2.5 × 10⁻² mol/ℓ of boric acid. However, they propose that the borate is only covalently bonded to one of the PVA chains, while the other chelates a sodium ion, rather than forming a di-diol cross-linking bond, and neither that mechanism nor their melting point plots are particularly promising for the prospects of this material being a thermoset.

I heated a small piece of it on vermiculite with the brass torch. Its surface intumesced slightly with white bubbles at gentle heating, and it produced a terrible smell reminiscent of burning polystyrene, and a little bit of white smoke. Stronger heating charred the surface and produced a little flame, which self-extinguished in a second or two and did not continue smoldering. The uncharred material had stuck to the vermiculite and become noticeably plastic; I could pull it apart with my fingers. I conclude that it is not a thermoset, just another thermoplastic (if a particularly hard one, and one that forms a nice gel with water).

Even if it’s a thermoplastic, though, you could still use a PVA solution as a binder for 3-D printing in a powder bed of filler doped with some kind of borate.

A low-concentration thermoset hydrogel would be appealing for the waterglass-tape application mentioned earlier, a great improvement on agar, as well as for molding things with the appropriate fillers and for preceramic polymers and for producing carbon foam (or fibers). All I have available at the moment is wheat gluten; the internet suggests chitosan cross-linked with β-glycerophosphate, dicarboxylic acids such as citrate, glutaraldehyde, divinyl sulfone, epichlorohydrin, or electron beams as possibilities, or cellulose or chitin dissolved in more exotic solvents, which I guess is how rayon/cellophane is made. It seems likely that most materials that could covalently cross-link common biopolymers like starch or chitosan would be pretty toxic; apparently epichlorohydrin, monosodium phosphate, sodium trimetaphosphate, and sodium tripolyphosphate (sodium triphosphate) are commonly used for cross-linking starches. (Could the obesity pandemic be largely caused by modified food starch?)

02021-08-24

I heated a bit of baling wire in the brass torch to orange heat to melt some holes in the five bottom lobes of a plastic soft-drink bottle (Tomasso cola) to turn it into a gardening pot, which yielded curls of white smoke with a burning-plastic smell, but no flames. I burned the remaining PET off the wire with the torch, then set it down to cool on a piece of the waterglass foam, to which it did not noticeably adhere. Although it’s only 57 mm long and about 1.5 mm diameter, the handle end didn’t get hot throughout the whole process. This is the same wire I had Mina use to burn the mole off my forearm, but she had it gripped in the vise-grips rather than her hand.

I notice that the off-balance weight of the brass torch is making the butane can a little unstable now that it’s nearly empty.

It occurs to me that maybe some metal hydroxide could serve as the hydroxyl donor for the polymerization of acetic-cure silicone, and copper hydroxide (which I’ve made previously by electrolysis) seems like an ideal candidate because of its laid-back enthalpy of formation (-225 kJ/mol OH) and associated low decomposition temperature. By contrast, sodium hydroxide is -425.8 kJ/mol and doesn’t boil (decomposing into the elements, I think) until 1388°, and aluminum trihydroxide (78.00 g/mol, 2.42 g/cc, gibbsite, dehydrates in the range 180°–300°) is -1277 kJ/mol, or -426 kJ/mol OH, same as the sodium salt.

Hmm, https://iupac.github.io/SolubilityDataSeries/ is an intriguing URL! Apparently IUPAC has put their solubility data series on GitHub, as well as a number of other datasets, although the link to the source repository seems to be broken. Copper hydroxides are in volume 23. About half of the volumes have analogous URLs; many others (5, 6, 16, 17, 27, 28, 45, 46, 53, and 67–104) are missing.

IUPAC’s solubility data is from 01986 and suggests using lye or ammonia to solubilize copper hydroxide. It has this lovely epigraph at the beginning of the Foreword:

02021-08-25

I bought a kg of electrical copper from the recycler around the corner for AR$1200 (US$6.70/kg). Some of it is very fine (presumably enameled wire from CRT yokes) while other parts are stranded, thicker, and uninsulated, measuring 0Ω on the multimeter. I cut a 1-meter section of a single strand of the uninsulated cable, which measures 0.51–0.57 mm in diameter with the calipers at different places along the strand, thus having a volume of about 0.20–0.26 cc. It weighs 2.1 g, which would put it in the density range 8.2–10.3 g/cc. It can be easily bent by hand.

It melts on the bed of vermiculite at a red-orange heat from the butane torches, a heat which is difficult to attain without placing some heat-reflecting waterglass foam behind the wire, at which point the waterglass also starts melting. Upon heating, it forms a hard, adherent black oxide coating. The melting is sharply defined, so that an intact wire forms a small sphere of liquid at the end of it. No visible fumes are emitted.

All of these characteristics are consistent with the metal really being copper. The density is in the right range; measuring the density of a larger piece of metal, or using a less shitty scale, would give me a more precise number. Zinc (from brass) would boil at 907°, emitting white fumes, and bronze would partly melt around 798°, or, in some flavors, completely melt at below 750°. Either alloy would melt gradually rather than suddenly, and both would probably be harder.

The melted copper wetted the vermiculite grains enough to stick firmly to them, suggesting that it ought to be possible to use copper (in a reducing atmosphere!) to repair broken pottery in the same way as silver and gold. To confirm this, I tried melting a little more copper wire on a bed of sand, and indeed the little ball of melted copper got sand stuck all over it, much of which I could not remove with my thumbnail. The effort abraded away much of my thumbnail stickout, in fact. I attempted to test the hypothesis directly by putting a couple of broken bits of ceramic floor tile with copper wire between them into an oven improvised from bits of waterglass foam, but although the glaze on one of the tile bits melted and flowed a bit, I wasn’t able to get them hot enough with the torches to melt the copper. They glowed orange. I was able to melt copper wire onto the bottom of one of the tiles by itself, a little, but it was relatively easy to dislodge after it cooled.

In the process I lost the aluminum foil on the bottom side of one of the waterglass pieces from the loaf I baked the other day; it melted away, along with a good portion of the thickness of the waterglass. Some of the waterglass also melted onto the top of one of the ceramic tile fragments, though I was able to dislodge it later.

The flames shooting out of the gaps in the oven were bright green, which might be from something in the ceramic (especially its glaze) or (less likely) the copper or waterglass.

During this process there were some acid fumes and some that smelled a bit like some kind of burning plastic, so I probably ought to throw out this vermiculite bed. Now I have both slightly irritated lungs and throat and a bit of a headache.

It occurs to me that if I can disperse a fine dust of a water-insoluble inert source of polyvalent cations in the waterglass, something that remains fairly inert at the low temperatures needed to soften and foam up the waterglass but then gives up its ions when the waterglass is facing real fire, it would probably work better than just applying oil of lime to the surface of the foam. Candidates include talc, aluminum trihydroxide, bentonite, copper oxides or hydroxide, finely ground enstatite, dehydrated borax, chalk, kaolin, rutile, or zincite. Amorphous silica such as diatomaceous earth might also work, but in a different way, by dissolving into the warm anhydrous sodium silicate melt and merely diluting the sodium rather than displacing it.

It occurred to me that an acid electrolyte like vinegar is not going to be useful for producing copper hydroxide, because it will convert essentially all the copper hydroxide into (soluble) copper acetate. So I resolved to get a vitriol electrolyte, but what I have handy is green vitriol, which might produce undesirable deposits of its own, maybe insoluble iron hydroxide, contaminating the copper hydroxide.

So I heated up some baking soda in tap water on the stove; it started bubbling as soon as I started applying heat, indicating that it was decomposing to soda ash. I added a little unheated baking soda to a near-transparent green solution of iron sulfate fertilizer in tap water, resulting in the formation of the usual nasty green muck with lots of bubbles. After the baking soda had boiled for a while, spoonfuls of the solution still bubbled enthusiastically when vinegar was dripped into the spoon. But I added the hot solution to the iron etc. mix and got only a little more bubbling, presumably due to the accelerated decomposition of the remaining baking soda. Bubbling slowed and the solution began to defecate after a few minutes.

Now I realize that because I didn’t measure anything my solution contains an unknown mix of soda and sal mirabilis, plus the nasty green iron carbonate that’s precipitating out. If I use that to electrolyze the copper I’ll end up with verditer, which is the usual blue-green pigment and which decomposes to tenorite on heating, so I should be satisfied with that. But in that case I might as well just use the soda as the electrolyte directly without the sal mirabilis.

I turned off the flame for a while, allowing the soda to cool a bit and crystallize into a single large mass which broke up and partly dissolved readily upon the addition of more water. Reheating produced no further bubbles, indicating that the baking soda was essentially all decomposed. There was a small black speck floating in it, probably from the defective faucet. I spooned a bit of the menstruum into the defecated vitriol product, which precipitated no further visible solids, indicating an excess of soda. Then I decided to dehydrate the soda, for which purpose I added aluminum foil over the top of the container. After a few minutes of further heating, pops were heard, so I lowered the flame further. Pops continued; upon inspection the soda has crystallized into a mostly porous mass, which has a few dozen holes in it where steam explosion has blasted fragments elsewhere.

The internet suggests that the green material may be iron(II,III) hydroxycarbonate, “carbonate green rust”. (On previous occasions, after filtration and drying, it turned into a very fine purple-brown powder.). After a few minutes it has defecated into a well-defined layer at the bottom of the jar (about 12 mm deep out of 60 mm, if the glass is 3 mm thick), which appears solid but is actually a fine suspension that offers no resistance to being stirred with a knife. The water above it is green and cloudy, but basically transparent.

I heated a piece of brass-wool dishwashing sponge (<500mg) with the brass torch as a control, to see if it behaved the same as the copper. It seemed to be, if anything, harder to melt, but it did emit a little white smoke, and did not ball up due to crossing sharply-defined melting points, but rather wilted. It did form the same kind of black oxide coating, though. However, a white surface deposit (presumably philosopher’s wool) was evident on the lower layers of the result. The burned sponge was crumbly rather than solid as the melted copper wire had been, perhaps because it was thin enough to oxidize most of the material rather than just a superficial coating.

After a few minutes more of heating, the pops and hissing from the soda ceased, and I was left with a bright white mass of soda, which had crept up the walls a bit. It had delaminated from both the floor and walls of the thin stainless steel pot, and indeed evidently detached from the pot in a single block. As it cooled, it crackled softly.

I flaked some soda off the wall and dripped some oil of lime on it in my hand, then let it to soak in my hand for a few seconds. No reaction was evident, not even warming, but afterwards the flakes took multiple washings over several minutes, clouding the water each time, to dissolve completely in water. By contrast, other similar flakes mostly dissolved in water in a few seconds and entirely in a minute or two. This suggests that fairly insoluble chalk was formed, though not enough in a few seconds to withstand multiple washings; chalk is not as strong as the phosphates of calcium, but still a viable building material. If you wanted to do this with an inkjet printer, calcium dinitrate might be a less corrosive source of calcium ions than oil of lime.

Using a cut-off end of construction I-beam as a handheld anvil, I was able to hammer one of the balls of copper flat, to perhaps 10× its original area and 1/10 of its original thickness, without any difficulty, although it cracked a bit around the edges.

96% ethanol is sufficient to remove residual adhesive from Emeth Línea Gourmet jam jar labels, though it requires several seconds of soaking and usually multiple tries.

I ground up the dried soda in a marble mortar and stored it in another Emeth jar.

Oh, I realize I never recorded the ingredients of the Cif dishwashing detergent, whose bottle looks like it’s made in the same factory as the Magistral, maybe with the same blow molds: the surfactants, both anionic, are linear sodium lauryl ether sulfate and sodium alkylbenzenesulfonate. The other ingredients are “pH regulating agent, coadjuvant, viscosant, preservatives, sequestrant, colorants, and perfume.” Pretty shitty labeling if you ask me.

Mercado Libre offers 5 kg of technical grade sodium lauryl sulfate for AR$7430 (US$41, US$8.30/kg), which is pretty close to the US$10 I was guessing I was paying at the supermarket, without having to dump the formaldehyde in it myself.

Because the Wikipedia article on green rust mentions that it normally oxidizes to iron oxyhydroxide, which is not green but more brown, and because the stuff I’d made previously turned purple-brown, I sucked a little of the green muck into a tube and put it in a bowl, then added some drops of 30-volumes hydrogen peroxide to it. It immediately turned brown-black and foamed up a lot, which is consistent with the identification as carbonate green rust. It would be better to use an oxidizer that isn’t itself inclined to foam up, I guess, so I could be sure that it’s really releasing the carbonate.

The weird thing is that the syntheses described in Wikipedia are a lot more involved than just dumping an excess of soda ash on some green vitriol in a jar, so I wonder if there are impurities in my fertilizer that predispose it to this. Or maybe this hydroxycarbonate isn’t very pure.

The hydroxycarbonate was a little stinky, same as last time. I wonder if that means the green vitriol is slightly contaminated with, say, iron sulfide.

02021-08-26

The hydroxycarbonate has entirely defecated, leaving what seems to be a perfectly clear menstruum, except that some of it is stuck to the walls of the jar. Siphoning this water out may usefully substitute for a stage of filtering if the objective is getting the solutes.

It occurred to me that in fact electro-etching of copper in soda ash solution may not work, precisely because copper carbonate is insoluble. If so, the sal mirabilis in that jar could be crucial.

02021-08-29

The cuboid of foam I put in to dehydrate a few days ago now weighs 700 mg, which I guess means that it wasn’t fully dehydrated at 900 mg. The bowl seems to have remained reasonably well sealed with the foil; the flakes of muriate of lime that were slightly damp before have now become almost dry, just slightly shiny, and very firmly stuck to the bottom of the bowl, and slightly discolored with what looks like iron oxide rust. Upon immersing the cuboid in tared water the scale read 7.0, indicating that its extra-pore volume is 7.0 cc; upon removing it from the water, the scale read -1.1, indicating that 1.1 cc of water had entered its pores; upon weighing the wet cube I get 1.8, which is consistent with 1.1 g of water plus 0.7 g of cube. So this foam cuboid, when dry, is actually about 8.1 cc and 86 mg/cc. Upon leaving it on the scale a little while, turning the scale back on, and picking it off, the scale reads -1.7 g, so about 100 mg of water seeped out of the foam; upon drying the scale with a paper towel (should have used a sample boat!) I get -1.8 g as expected.

Unfortunately I didn’t weigh the bowl of muriate of lime before adding water to it to get rid of the muriate, so I can’t heat up the muriate to see how much water I drive out of it (most of which was presumably obtained from the cuboid).

To get relief from gastric reflux, I drank a little dilute washing-soda solution earlier. It was bitter, left my tongue sort of tingling in a soap-like fashion, and left my throat more irritated than before, so I probably should stick to baking soda. (I took some baking soda a little bit later.) I was thinking it might be interesting to try to saponify some butter with that soda, but apparently this involves boiling it for 2–5 days, so I don’t think I’ll do that. Maybe I’ll make some soap with lye, but not tonight.

Maybe I can make boric acid (61.83 g/mol). The usual borax is the decahydrate (381.33 g/mol) and decomposes to the anhydrous form (201.22 g/mol) at 75°, which then doesn’t melt until 743°, a temperature with a deep-red glow. In water you can dissolve 3.17 g/100 mℓ, I guess at room temperature. Boric acid, by contrast, doesn’t have a hydrated form and melts at 170.9° (well, dehydrates to the almost insoluble metaboric acid, which melts at 236°), but dissolves about 2.52 g at 0°, 4.72 g/100 mℓ at 20°, rising to 27.5 g/100 mℓ at 100°. I don’t really understand pKa, but I think that because boric acid’s (first) pKa is 9.24, acetic acid (60.052 g/mol, pKa=4.756) should be able to convert all but about 0.003% of the borate in borax into boric acid by stealing off its sodium for the highly soluble sodium acetate (82.034 g/mol, 119 g/100 mℓ at 0°, 123.3 g/100 mℓ at 20°, 162.9 g/100 mℓ at 100°).

The stoichiometry here is a little confusing. I need one acetate per sodium, and borax (Na₂B₄O₇·10H₂O) has two borons per sodium, so I guess I have two borons per acetate. But that tetraborate group needs some extra hydrogen to break it up into loose boric acid molecules, namely three hydrogens per boron, plus some extra oxygens (5 oxygens per tetraborate) but acetic acid only provides one hydrogen per acetate and no oxygen. So I guess we have to get our other hydrogens and oxygens from the water: Na₂B₄O₇·10H₂O + 2CH₃COOH → 5H₂O + 2NaCH₃COO + 4B(OH)₃.

So if I get a reaction product that crystallizes at anything less than overwhelming concentrations, even at 0°, it’s either borax or boric acid, and if the solution is still acidic, it’s boric acid. And ideally I should be able to get about 90% of the boric acid to crystallize if I use a minimal amount of water to dissolve it. (Even if I were to crystallize it just by cooling down an excessive amount of solution, the yield ought to be above 50% as long as I have less than 100 mℓ of water per 5.04 g of boric acid.) Anhydrous sodium acetate’s boiling point is listed as 881.4°, so as long as I don’t boil the stuff dry (and then overheat it to red hot), it ought to be stable.

I don’t have a simple way to find out how much acetic acid is in my vinegar, though, so I’d have to add an excess; acetic acid is entirely miscible with water and boils at only 118°.

I guess I could see roughly how much vinegar is needed to neutralize, say, a gram of baking soda (84.0066 g/mol). (This would be a lot easier with a better scale, but I guess I’ll just be very approximate instead of carefully titrating.)

I added 1.0 g of baking soda to a tray (which initially weighed 2.2 g). I added vinegar to a cup that initially weighed 8.2 g, with a final weight of 53.7 g, so 45.5 g of vinegar. By the time the cup weighed 35.5 g (27.3 g of vinegar) the baking soda seemed to all be neutralized (by the 18.2 g of vinegar I added), but not much before that. Let’s say somewhere between 13 g and 18.2 g.

In theory this is NaHCO₃ + CH₃COOH → CO₂ + H₂O + NaCH₃COO, a 1:1 mole ratio between acetic acid and baking soda, so the 1.0 g of baking soda corresponds to 60.052/84.0066 = 0.715 g of acetic acid, but given the weighing imprecision of the scale, somewhere between 0.68 and 0.75 g. Continuing the interval arithmetic, that puts us between 3.7% and 5.8% acetic acid in this vinegar (Alcazar vinagre de alcohol, acidez 5%, UPC 7-790130-000294). I could probably take a second measurement with more careful titration to nail that down to something like “4.3% to 5.3%” but I’m not going to, because this is precise enough to allow me to ensure an excess of acetic acid with the borax without adding an absurd amount of extra water, and I can harmlessly boil off any extra water without losing any boric acid.

So suppose I want 100 g of boric acid, B(OH)₃. I’d need 364 g of water to dissolve it in. This bowl I used to dehydrate the glass foam weighs 42.5 g empty (it’s probably the bowl that weighed 42.3 g before) and 562 g with an unreasonably large amount of water in it, an amount almost certain to spill. But looking at the equation above, every mole of borax (381.33 g) requires two moles of acetic acid (120.104 g) to neutralize it, producing four moles of boric acid (247.32 g) and two moles of sodium acetate (164.068 g). But 120.104 g of acetic acid at my lower-bound 3.7% concentration would be 3300 g of vinegar, and I have less than 500 g here, and at any rate if I used that much it would spill everywhere when it came to a boil.

So suppose I use a more reasonable 200 g of vinegar, providing at least 7.4 g of acetic acid. That should suffice to neutralize 23 g of borax, producing 15 g of boric acid and 10.1 g of sodium acetate. If I then just stick it in the freezer, the 193 g of water will only be able to dissolve 4.9 g of it, for a theoretical 67% yield, but if I boil it down to 100 g or 50 g or something, I should be able to improve that significantly.

I weighed out 20.0 g of borax into an improvised aluminum-foil sample boat, put it in the bowl, and weighed out 200 g of vinegar, then put it on a low flame. Some of the borax seemed to dissolve in the vinegar at room temperature, but, as predicted, not most of it. It finished dissolving at about 01:17 as the vinegar warmed up a bit, at a point where it was too hot to hold my finger in but did not burn my finger in a second or two, perhaps 40–45°. I continued heating it on low heat a while longer, then decided to go for broke and boil it rapidly with an aluminum-foil lid. At 01:23 it weighed 244 g without the foil lid, 246 g with it. At 01:25 it weighed 232 g, and the foil seemed to be doing a reasonable job at preventing liquid drops from spraying out.

At 01:30 it had boiled down to 167 g (remember, 43 g is the bowl, and 1 g is the lid). I took off the lid to see how it looked, like if anything had precipitated, but it just smelled very acidic and looked clear. Unfortunately I tore the lid, which reduced the weight to 163 g, so about 3 g of condensation and spray went with the previous lid. I made a new lid (also about 1 g) and returned it to the fire.

7 minutes to drop from 244 g to 167 g is 11.0 g per minute, or more realistically 8–16 g per minute. If we assume that basically all the power is going into boiling water, at water’s enthalpy of vaporization of 40 kJ/mol and 18 g/mol, this is about 400 watts.

At 01:36 it was starting to sound syrupy, so I turned it off. It weighed 98 g, 96 g without the lid, which stuck a little to the bowl as I tried to remove it. Some drops that had splashed onto the side of the bowl had crystallized into white spots, but they dissolved rapidly when I sloshed the hot syrup over them. I put the mix into the freezer. There was a strong smell of acid.

96 g should be about 53 g of solution, of which about 10.1 g should be sodium acetate, about 13 g boric acid, some undetermined but fairly significant amount should be acetic acid, some insignificant amount should be borax, and the rest (less than 29.9 g) should be water.

At 01:45 I looked in the freezer; there is no visible precipitation yet, except possibly some dust floating in the syrup. Its viscosity seems unchanged.

At 01:56 there was a lot of translucent white crystallization on the bottom and some more floating around in the syrup, which actually seems maybe a bit less syrupy now. I made a gravity filtration setup out of a couple of Monster cans and two coffee filters and stuck it in the freezer to cool. Empty, it weighed 27.6 g, but then I realized ⓐ it actually had a drop of water in it and ⓑ was too tall to fit in the freezer and ⓒ at any rate I was going to be filtering 40 mℓ of solution and not 400 mℓ so I could cut it down a bit. So I cut it down to 24.0 g.

At 02:06 the fluid is quite milky from crystallization and seems to be about the same viscosity as before, and it still smells intensely of vinegar. My plan is to dump the liquid into the filtration funnel, then rinse the crystals stuck to the bowl with some 0° water, passing that water also through the filtration funnel, and then set both the bowl and the filters out to dry under a loose covering for a few days.

At 02:18 I did the filtering, but didn’t do a very good job of record keeping because I was trying to do it all before things warmed up and increased solubility. If I recorded this correctly, though, the filter funnel setup initially weighed 24.1 g, and I think the bowl initially weighed 77.8 g. After pouring the milky menstruum into the filter funnel, it weighed 38.9 g, but I didn’t record the bowl’s weight; it must have been about 63.0 g. I added some 0° water to the bowl, bringing its weight up to 86.0 g, which seems like a lot more water than I should have added; after I poured that into the filter funnel too, the bowl was down to 72.7 g, and then I knocked over the filter funnel, spilling some of the menstruum onto the table and bringing its weight down to 45.2 g. I added more water to the bowl, bringing it up to 80.2 g, and dumped that into the funnel, reducing the bowl’s weight to 69.5 g, and bringing the funnel’s weight up to 58.4 g. I should have weighed the water bottle at the beginning, too!

Afterwards the water bottle weighed 514 g; it’s a nominally 500 mℓ Coke bottle, of a type which weighs 24 g including the PCO1881 cap. On refilling it until only a small bubble is left, as before, it weighs 555 g. So the total amount of cold water I added was somewhere in the neighborhood of 30–45 g.

Now I realize I should add some extra rinse water to the filter funnel to wash water-solubles out of the filters. I first reduced the weight of the cold-water bottle so I could weigh it on the scale at 390.0 g, and on adding some of it to the funnel apparatus, its weight increased from 58.4 g to 105.2 g, leaving the bottle weighing 345.2 g, so I transferred 44.8 g or 46.8 g of water from one to the other. One of those numbers must be wrong, because I doubt I spilled 2 g of water or got it stuck to my hands, but on weighing the filter apparatus again it weighs 105.0 g, and the bottle still weighs 345.2 g. So, I don’t know what went wrong, but it’s probably safe to say I added on the order of 44–47 g of water to rinse the funnel.

105.0 g amounts to 66.1 g of rinse water, which will probably have stolen about 1.7 g of the boric-acid product from the funnel. It’s a miracle there’s anything solid left in that coffee filter, but there is.

To see if the wet product in the bowl is really boric acid, I spooned about 0.2 g of it with a chopstick onto a 3.9-g bottom of a Monster can and turned on the stove. At first it dissolved a bit as the water heated, then formed a white apparently amorphous foam as the water boiled. I scraped a little powder off the foam with a chopstick, which landed on the aluminum. No further changes were seen to the foam or the powder as I turned up the heat, though the epoxy inner liner of the Monster can started to smoke white, and my lungs started to criticize me. I blew a butane torch over the top of it to heat it further and hopefully burn off the smoke, but this mostly had the effect of heating parts of the foam to orange-hot and melting the aluminum underneath it, at which point I turned off the fires and opened the window.

This is uninspiring, more like what I had seen previously from heating borax (in a less stupid and dangerous way, using aluminum foil) than what I was hoping for from boric acid. There was never a point where it convincingly melted, even though it got to red heat. The only detectable difference is that it’s black instead of white (from the smoke from the charring epoxy) and it’s perhaps a bit harder (but maybe I reached borax’s melting point and thus collapsed superficial bubbles, reducing porosity).

The remaining prediction to test, I guess, is that, after heating the “boric acid” enough to convert it to metaboric acid or boria, it should not dissolve in water; the solubility of the borax foams, by contrast, was one of their most striking properties. Accordingly, I added most of the blackened remains of the “boric acid” foam to a cup of warm (≈40°) water. In a minute or so, they dissolved into black sludge, which breaks up when the water is stirred. So I think what I have here is still just borax, contaminated with the toxic decomposition byproducts of epoxy. I threw it away.

Other candidate tests suggested in some random lab handout include a flame test (which should be bright yellow for borax, transparent green for boric acid) and conductivity in distilled water (which I think ought to be very high for borax and very low for boric acid, though the lab worksheet doesn’t actually say, instead being phrased as an exercise). The lab procedure used 1M vitriol rather than vinegar but was otherwise equivalent.

I tried the flame test and got a bright yellow flame that looked like sodium. However, I also got yellow flames from my iron wire and carbon welding electrode when they were supposedly clean, though maybe the yellow was a little less bright. I also contaminated the bowl of sample with the carbon welding electrode; when I dipped it in while hot, it scattered carbon particles around.

A much simpler test would be to heat dry “boric acid” past 200°, weighing it before and afterwards. If it’s boric acid, which is 61.83 g/mol, heating to even a few hundred degrees will only drive off the hydrogens, reducing its weight by about 4%. If it’s borax, heating it past 75° will dehydrate it from 381.38 g/mol to 201.22 g/mol, a 47% reduction in weight. But I cannot do that until it is dry. A differential scanning calorimetry/thermogravimetric analysis rig would be extremely helpful for this kind of thing.

It’s 03:40. The filtering apparatus has mostly frozen; the funnel part has a little icicle hanging down, and the receptacle in the bottom is mostly a block of ice, with a little of some kind of syrupy liquid on top of it, which is plausibly a low-melting solution of water, acetic acid, borax, boric acid, and/or sodium acetate. I guess that means it’s not doing any more filtering at 0° or -20° or whatever. So I took it out of the freezer and left it to thaw on the table, with a little aluminum foil over the top to keep dust from falling in. The idea is that when it melts, a little more filtration will happen, and then I can remove the filters from the funnel and lay them out to dry.

The old article, however, does mention acidifying with acetic acid — but followed by a stage of evaporating to dryness, then volatilizing the result with methanol! Methanol seems to be available for AR$190/ℓ (US$1.06/ℓ) as a gasoline performance additive for hot-rodding, so maybe I can find it at an auto parts store, but it seems like it’s a more specialty product than in the US, and the HEET brand name (for its use as a gas-line antifreeze) is nonexistent.

However, Gooch seems to be concerned not with purifying boric acid but with measuring the amount present, so he ends up turning it into calcium borate to weigh it. On the other hand, he has a lot of helpful tips about this sort of work, mentioning for example “the exceedingly delicate test with turmeric” for detecting residual boric acid. This is mentioned in the ScienceMadness Wiki:

It seems like Gooch was forming (and distilling) trimethyl borate, which boils at 68°, but had no way to find that out.

It looks like boric acid is highly soluble in methanol and ethanol (even without reacting with them), while borax is only slightly soluble in ethanol. So that might be a thing to try.

I took out another sample of 0.2 g or so from the bowl and dripped 96% ethanol on it on a scrap of aluminum foil, letting the somewhat milky liquid run onto a different scrap. I set the second scrap on fire, and got what looked like a pretty normal yellow blackbody diffusion flame with a green aura around it, which at least vaguely suggests some boric acid might be present. Eventually the alcohol burned away, leaving a white residue. On heating this foil on the stove, the white residue turned black, even though there was no epoxy available to contaminate it. I suspect that this was sodium acetate being charred, which means that I didn’t rinse the bowl very well. (Gooch mentions that sodium acetate is soluble in methanol; apparently ethanol also dissolves 5.3 g/100 mℓ of the trihydrate.)

02021-08-31

The borax/vinegar material in the bowl has solidified and smells strongly of vinegar, indicating again that it was inadequately washed. The filtration funnel still has sediment in the filters; the menstruum it filtered remains clear (but is hardly exposed to evaporation).

Yesterday I ate a can of mackerel and washed it out, but it retained a strong fish smell. I heated it full of water and sodium percarbonate on an electric burner, which produced froth. There was too much sodium percarbonate to dissolve, which restricted the flow of water near the bottom; this may have been a factor in the plastic can lining bubbling up along the bottom where it was over the heating element. It still smelled strongly of fish. I left it full of dilute household ammonia overnight, but today it still smells strongly of fish. Now I have put it full of dilute household bleach (taking care to rinse the ammonia out first).

To a cup cut from the bottom of a Monster bottle weighing 8.1 grams I added 40.4 g of the somewhat wet copperas fertilizer, removing a large crystal that somehow got in there (presumably of copperas, but possibly something else). 48.4 g was the final weight, so maybe that was more like 40.35 g. I added 151.2 g of tap water, which mostly dissolved the copperas immediately, though the solution was a bit murky. Applying gentle heating does not seem to remove the cloudiness, so it’s probably insoluble particulate contamination, maybe the pyrites I suggested earlier might be aromatizing my iron precipitation products. Actually, gentle heating seems to have turned it an opaque green-brown! But there was no acidic smell, so the heat surely isn’t decomposing sulfate ions.

Now I realize I should have tried heating it to dryness first to quantify how much water there was in it; this ought be the heptahydrate plus a bit.

To a yogurt cup weighing 7.1 g I added an egg white; afterwards it weighed 41.6 g, suggesting that I have 34.5 g of egg white here.

I turned another Monster can into a filter-funnel setup and filtered the warm copperas solution through it, spilling a little. Much to my surprise, there’s a green translucent mud at the bottom, apparently copperas crystals. The filtrate seems to be just as opaque as the solution, so evidently this coffee filter is not doing a good job at filtering out the particles, which I now realize may be precipitated copperas.

To the can with the sparkling green sludge, now weighing 23.3 g (16.2 g sludge) I added 81.1 g of water. The crystals began to dissolve immediately, but were not completely dissolved after a few minutes when I spilled the whole thing on the floor. This, however, revealed that some of the crystals were stuck together in pale blue-green granular masses, presumably from when I was heating them; the sludge had hidden them previously. These masses broke into a few pieces when the can hit the ground. The can, unspilled green water, and crystals weigh 25.8 g (17.7 g sludge and water); I added another 21.3 g of tap water to try to dissolve them.

So evidently what happened was that I prepared a saturated solution of copperas with some minor particulate contamination, then evaporated it (failing to increase the solubility enough to desaturate) to precipitate lots of tiny copperas crystals, which are now clogging my coffee filter. But the 151.2 g of tap water I added was evidently capable of dissolving at least 40.4-16.2 = 24.4 g of the copperas crystals, and probably more, since some of that 16.2 g of sludge was water. This works out to at least 16.1 g/100 mℓ but less than 26.7 g/100 mℓ; WP says its solubility in water is 20.5 g/100 mℓ at 10° and 29.51 g/100 mℓ at 25°, ranging up to 51.35 g/100 mℓ at 54°. So I probably shouldn’t have been so ambitious with the quantities.

The granular masses and crystals to which I added the 21.3 g of water mostly dissolved, so I poured the cloudy green solution to the funnel, leaving only a single crystalline mass stuck to the side (dropping the mass from 46.8 g, including the can, by 38.1 g, to 8.6 or 8.7 g) and added another 24.4 g of tap water, which rapidly diminished the size of the polycrystalline chunk, which dissolved in a minute or two. The pale green solution is still fairly cloudy, but I added it to the filter. The empty can weighs 8.2 g and has some green deposits on its walls.

I poured the 228.8 of filtered liquid back in to the 8.1 g (!) empty can and thence back through the filter funnel, except for some which I lost. I’m hoping that the particles clogging the filter will provide me with finer filtration this time around.

It looks slightly clearer maybe. I put 1.9 g of the egg white into the empty can and added 7.5 g of the slightly clearer filtrate, for a total weight of 17.6 g. Yellow-brown, slimy transparent masses started to coagulate in the solution; I was hoping for a precipitate but I was not expecting it to be either brown, transparent, or slimy, but rather hard and granular, with an octahedral crystal habit.

There is still considerably more particulate in the filtrate in the funnel setup than in the copperas I’d started with, although I’ve surely added much more water than evaporated, so I’m guessing that whatever particulate formed when I heated the copperas solution, it wasn’t just precipitated copperas. It might conceivably be that I oxidized some of the copperas, which would maybe explain the browner color (it has a browner green color) but that shouldn’t precipitate it; rather the contrary, WP gives ferric sulfate’s solubility as 25.6g/100 mℓ.

The particulate in the funnel is a yellow-brown, maybe even a baby-poop yellow, very different from the blue-green of the copperas itself. After 2½ filtrations through the same funnel the solution seems to be getting a little clearer.

I added a spoonful of salt and nine spoonfuls of vinegar to half a cup of milk, which began to curdle into tiny curds. I think it started to thicken around six spoonfuls; at the end, it was very sour. I set up another Monster can as a filtration funnel with a coffee filter and used that to filter the curds from the whey.

I tried to remove the full coffee filter from the funnel, but it split, and I had to start again. I spilled a little of the cheese. The restarted filtration yielded a great deal of acid whey, somewhat curded at first but later almost perfectly clear, which I neutralized with baking soda and drank, despite its excessive saltiness and perhaps excess of bicarbonate, which I then tried to correct with a little lemon juice. Finally I diluted it with Diet Sprite. A white powder that looks like baking soda was visible at the bottom of the cloudy liquid, but the liquid was drinkable without difficulty. Indeed, on adding vinegar to the white powder, it fizzes! So it was at least partly undissolved baking soda.

I managed to get a little cheese off the first, broken filter, and scraped it off a plate. It was quite agreeable, despite the strong vinegar flavor. This is definitely a process to keep in mind if I ever have a fridge full of milk in a power outage. Presumably I can leach the vinegar out once I have it solid; I wonder if I could preserve the milk with acid and sugar (and, say, chocolate) rather than salt? Maybe some citrus terpenes?

Something has gone wrong with the water supply, so I suppose I will allow the various materials to repose until tomorrow; I no longer have a way to wash my hands after handling them. Hopefully the egg white will not putrefy. The little funnel full of cheese is gradually dripping at about 2 Hz.

A few hours later the water supply is back. After 4 filtrations the filtrate is clear, looking like fake apple juice — still not the blue-green color of copperas crystals, but definitely yellow-green rather than brown.

02021-09-01

To the can where I had combined egg white with copperas before, which looks like brown phlegm I coughed up once when I was sick, I added the remaining 29.0 g of egg white, then 100.0 g of the copperas solution. More ropy brown precipitate formed immediately.

Soaking the mackerel can in dilute bleach since yesterday seems to have effectively deodorized it, which ammonia and (separately) sodium percarbonate failed to do. There is a little chlorine odor left.

Adding oil of lime to saturated clear solution of baking soda produces effervescence and a precipitate, just as with ferrous sulfate and baking soda; this despite WP claiming baking soda’s solubility is 9.6 g/100 mℓ and calcium bicarbonate’s is 16.6 g/100 mℓ. The precipitate appears white, being presumably chalk. I suppose that doing this in a dilute enough solution might produce large calcite crystals. See Fast electrolytic mineral accretion (seacrete) for digital fabrication?. I diluted the results with water, and the chalk settled within a few hours, but remained impalpably fine. It adhered to the surface of the aluminum can a bit and did not immediately dissolve in vinegar, but did dissolve with mild heating of the vinegar.

By contrast, adding oil of lime to Diet 7-Up (not Sprite) produces immediate effervescence but no precipitate, just as you would expect: even though 7-Up is a saturated solution of CO₂, it does not react to form CaCO₃.

The large dark green crystal I fished out of the copperas yesterday has now dried to a sparkling light green; it looks like someone sprinkled sugar on wasabi.

02021-09-02

The cheese is no longer wet, just moist. It is delicious, similar to an acidic ricotta.

The label of Doreé [sic] Capilar 30-volumes hydrogen peroxide (UPC 7-794050-007050) peels off the polyolefin bottle without leaving an adhesive residue. 95% ethanol removes the lot number and expiration date. Now I have a 100mℓ hermetic polyolefin bottle.

The baby-poop brown precipitate from the copperas on the filter has now turned brown. Adding 30-volumes hydrogen peroxide to it does not change its color. This suggests it may have oxidized from ferrous to ferric when I was warming it up, thus reducing its solubility (?).

02021-09-03

The dried bowl of borax vinegar is actually sort of soft and slushy. I placed a little on a pebble of pumice, skewered on baling wire, and heated it with the brass torch; it melted rapidly down into the pores of the pumice, and with further heating converted the pumice surface into porous glass. There was a strong smell of vinegar. The vinegar charred into little balls of black on the surface of the pumice. Then the piece of pumice cracked, and a couple of flakes gradually hinged away from the main block and then fell. A third crack remained open by a couple of millimeters. Did the acetic acid attack the pumice at high temperature, opening or widening cracks? Did the borax?

I placed some oil of lime on the pumice, which soaked in. On further heating I saw an orange flame, and the pumice cracked again, right through the middle, held together only by the baling wire.

I took an additional piece of pumice and dunked it in tap water, then heated it in the same way, alternating between dunking and heating a few tims. It did not visibly crack, but on tapping it on a ceramic plate afterwards, a piece fell off, evidently having been previously severed by an invisible crack.

The surface of the pumice reached orange and yellow heat during this procedure, but only melted where borax had been applied.

There was a bit of the “cockroach smell” I mentioned previously from heating the oil of lime, and also an acid smell, which might have been just more acetic acid.

The end of the baling wire was covered with borate glass from previously, which liquefied and smoothed out in the flame. It was still warm enough to sizzle when I added the oil of lime; thereafter it bubbled and had a pebbly surface in the flame, as if the glassy calcium borate (or whatever it was!) was only yielding up its water at orange heat. Calcium borate could be pretty interesting: colemanite (1:3 Ca:B mole ratio) is Mohs 4.5, piezoelectric, and fluorescent yellow, and nobleite (1:6 Ca:B) is Mohs 3. Sadly neither Wikipedia nor Mindat bothers to mention whether either of them is water-soluble, though colemanite mixed with ulexite is a popular pottery glaze ingredient known as Gerstley Borate. The US Borax MSDS for it describes it as “sparingly soluble”.

02021-09-04

02021-09-06

I hammered flat some 60/40 Radio Shack lead/tin rosin-core solder (barcode 40293-11311, Radio Shack part number I think 64-007 E) and stuck it in bleach (Ayudín agua lavandina común, 25 g Cl/ℓ, UPC 7-793253-000400). Also, a piece of aluminum foil. After 30 minutes no tarnishing was visible on either but the aluminum foil has some small bubbles; I suspect it may be forming aluminum chloride. I don’t know how much chlorine is left in this bleach; it’s been sitting around the house for quite a while.

The can of dissolved copperas has been sitting open since 02021-08-31 and has developed a non-metallic yellow-brown deposit at the bottom, though the liquid is still clear.

I took a cut-off Speed can weighing 6.5 g and added 19.7 g of the copperas solution to it, which appeared perfectly transparent, in the sense of not being cloudy, but yellow. Then I added 7.4 g of 30-volumes hydrogen peroxide, which immediately began to effervesce and turned the solution apparently black and opaque. A little while later it was quite warm, red-brown, and perfectly transparent (in the same sense) and weighed 33.4 g, gradually decreasing to 33.2 g (unless that’s an error in my scale). I conclude that I have made ferric sulfate (since none of the iron oxides, hydroxides, or oxyhydroxides are soluble) at a low enough concentration to not precipitate.

The mass loss so far is 400 mg, which is presumably a combination of lost oxygen and evaporated water.

I placed the Speed can out of the way with a loosely-fitting aluminum-foil hat to allow the solution to dry out without too much dust falling in. NIH says ferric sulfate should appear as “a yellow crystalline solid or a grayish-white powder”. If it takes more than a week, I’ll try putting it in a sealed chamber with some muriate of lime and let them fight over the water in the air; this is probably going to be necessary because NIH also describes it as “deliquescent”, and “decomp in hot water”. A pH-1 solution of the stuff is apparently used in Pakistan as a dental hemostat: “Known as the “classic” hemostatic agent”! It seems to be the right color from all the pictures, too.

It’s been about 8 hours, and the bleach has eaten the aluminum foil I stuck in it. The lead-tin solder is still shiny, but there’s a white deposit on some of it; I suspect this might have resulted from galvanic contact with the aluminum foil. Oops.

The sodium silicate I poured into the sand two days ago soaked through about 20 mm of sand and formed a solid mass, which has not stuck to the polystyrene container it’s in at all. The mass looks a bit wet, and when I squeeze it between my fingers, it yields and fractures, revealing inner surfaces that look even wetter.

In theory I ought to be able to mix a fair bit of muriate of lime with baking soda without getting chalk — until the solution dries out. Chalk’s solubility is supposed to be 0.0013 g/100mℓ, but calcium bicarbonate’s is supposed to be 16.6 g/100mℓ (162.11464 g/mol), more than a thousand times higher. Baking soda’s is supposed to be 9.6 g/100mℓ (84.0066 g/mol), and muriate of lime’s 74.5 g/100mℓ (110.98 g/mol). All these at 20°, which is a little warmer than it is here at the moment.

I guess the saturated solution of calcium bicarb at 20° is 1.02 mol/ℓ. For that we’d need 2.04 mol/ℓ of baking soda, 17.1 g/100 mℓ, which is higher than its solubility (though the potassium or ammonium salts could maybe do it directly), suggesting that the appropriate amount of muriate of lime ought to solvate baking soda. We’d also need 1.02 mol/ℓ of muriate of lime, which would be only 11.3 g/100 mℓ, 6.6 times more dilute than the saturated solution.

Maybe to understand why the large excess of calcium in the oil of lime resulted in precipitation last time I need to understand solubility products a lot better. But I think this means that if I dilute the oil of lime 7:1 and then add saturated bicarb to it, it shouldn’t happen.

So I took a cut-off Monster can weighing 7.3–7.6 g (the scale can’t quite decide) which can hold 227 g of water, and I added 12.1 g of oil of lime to it, bringing the total to 19.7–19.8 g; this should contain about 5.17 g of muriate of lime and 6.93 g of water. Then I added tap water to bring the total to 103.3 g, thus 83.5 g of tap water. So in theory I have 5.17 g of muriate of lime, .0466 mol, dissolved in 90.43 g of tap water, thus about .515 mol/ℓ. I stirred it with a soda straw; a little dust settled to the center of the bottom. So now even a totally saturated baking-soda solution shouldn’t cause effervescence and chalk deposition.

.0466 mol of baking soda should be 3.91 g, which in theory would be contained in 44.6 g of the saturated solution (44.7 g water). So I prepared a saturated solution of baking soda. To the now 103.1 g can (?!) I added 48.4 g of the solution (oops!) and... it turned cloudy with what looked like chalk, but this time completely without effervescence. That was neither the outcome I hoped for (a clear solution) nor the outcome I feared (effervescence and chalk).

So that should have been 135.1 g of water, 5.17 g of muriate of lime, and 3.91 g of baking soda. Oh, oops, that was half the amount of baking soda I was supposed to use...

And, after a minute or two, it started to effervesce. Ugh. I guess it just hadn’t reached the solubility limit of choke-damp, and that was chalk after all. It continued to effervesce for many minutes.

I guess I should actually measure the ingredients and the temperature when I repeat this.

After a while the chalk had largely settled out, except for the abundant dust forming aggregates on the surface of the water; effervescence continued at a lower rate. To get a better look I dripped a drop of diluted dish detergent into the can, which immediately scattered the surface dust to the edges of the container. The water beneath was still pretty cloudy, but I can see down to where the chalk has settled on the bottom.

Since I now realize I added half as much baking soda as I’d intended, I added some more of the saturated baking soda solution. The effervescence increased dramatically again, suggesting that the baking soda was indeed the limiting reagent.

It’s been a few hours, and a yellow-brown powder has precipitated in the “ferric sulfate”, so maybe that’s not what it was, because it certainly hasn’t evaporated that much. Or maybe there was an excess of hydrogen peroxide and now it’s producing insoluble iron oxide or oxyhydroxide from the soluble ferric sulfate in solution.

02021-09-07

The chalk has settled out, leaving clear water with a thin layer of chalk at the surface; maybe the chalk adsorbed the dish detergent. I took a 5.9-gram cut-off Speed can and added 10.0 g of the clear liquid to it with a syringe which I think had previously been used to fill inkwells. No India ink contamination was visible. Then I heated the can on the stove.

The idea here is that if the chalk precipitated, whatever is left over in the solution must be in equilibrium with the chalk. Maybe it’s calcium bicarbonate (and salt), just at a lower concentration than I had thought would be soluble. If so, I should be able to boil it down to chalk (contaminated with salt), which should be recognizable by virtue of being a white precipitate that bubbles and dissolves in vinegar and does not dissolve in hot water. And hopefully there will be enough solids dissolved in 10 grams of the solution that I can weigh it.

Adding a few more drops of dish detergent to the can where the chalk was formed does not disperse the surface chalk, so I guess the chalk didn’t adsorb the detergent, or the surface would have dramatically cleared again as before.

After heating the chalk (? etc.) deposit in the can for a few hours (mostly gently, because it was popping when I tried heating it intensely) the can weighs 4.8 grams. This suggests that the remaining solid deposit weighs -1.1 grams, which is obviously wrong; maybe I had something like 1.1 grams of water on it previously when I first weighed it. Then a flying plastic fragment bounced off the ceiling and knocked it onto the floor, scattering an unknown amount of the precipitate, so I guess I need to start over.

Both “nail enamel diluyent” and “enamel remover” are capable of softening the pressure-sensitive adhesive that held the front polarizing film onto this discarded laptop screen, allowing me to scrape it into giant boogers with my fingernails, but neither actually dissolves it. Surprisingly, one or the other of them did attack the polarizing film itself, damaging it (though it still seems to polarize fine). 96% ethanol works dramatically better, softening the adhesive much more quickly, allowing me to rub it into eraser-crumb strings that peel off the plastic cleanly, and apparently not dissolving the plastic itself. Unfortunately, I ran out. Fortunately, some alcohol-based hand sanitizer gel was enough to finish the job. The plastic is still sort of foggy.

I diluted and tasted the citric acid from the health food store. It dissolves to perfect transparency in water and tastes like citric acid.

02021-09-08

The waterglass-cemented sand from 02021-09-04 is now rock-hard and survives being dropped on the ceramic floor without breaking.

02021-09-09

Although the waterglass-cemented sand from 02021-09-04 is rock-hard, I can still break it by hand by flexion. Heating it to orange-yellow with the brass torch for a few minutes turns it from brown to gray but doesn’t expand or soften it visibly, though it did settle a bit in the vermiculite bed, I suspect because the vermiculite had some waterglass or phosphoric acid sticking some of the grains together. However, it is now easier to break by hand, and seems to have some internal porosity (it’s gray all the way through), so maybe it did soften during heating; no such porosity was apparent upon breaking before heating. It gives the impression of poorly cured portland-cement mortar: the outer surface where the flame impinged remains intact when I rub my finger across it, but the inner porous mass instead releases some sand. It’s a bit more sparkly than portland-cement mortars usually are, though.

People have reported that mixing rust powder with waterglass affords a less intumescent refractory mix.

I heated a drop of the glycerin on the stove on a sheet of aluminum foil. At first it emitted a white smoke with a smell resembling burning sugar; this could be ignited, at which point it burned with a blurry yellow flame and emitted no further visible smoke. Small bubbles bubbled out of the glycerin as it burned, and at one point it popped and threw smoking drops of glycerin around. After a couple of minutes it finished burning and left no visible residue, but the aluminum foil had mostly melted.

To a 6.3-gram plastic cup (from Tregar yogurt) I added 7.3 g of buff-colored hardware-store litharge (Indalo litargirio, UPC 7-798123-981544) and 2.1 g of glycerin (Indalo, UPC 7-798123-981483), which I mixed for a couple of minutes with a q-tip to a stiff putty, which nevertheless behaves like a (perhaps non-Newtonian) viscous fluid. This should approximate the 3:2 molar ratio mentioned above: 33 millimoles of litharge to 23 millimoles of glycerin. No heating is evident.

I’d tried mixing up this litharge and glycerin before, but the solid mass initially formed fell apart into a powder after a day or two. There aren’t a whole lot of things the ingredients could be other than litharge and glycerin, though; not many things are dense enough to be counterfeit litharge, though plenty are cheap enough.

The polarizing filter obtained from the front of the laptop screen is evidently a linear polarizer with the direction of polarization being the vertical dimension of the screen (the short dimension of the film). My own laptop’s screen (a Lenovo Thinkpad) seems to be linearly polarized at about a 45° angle, and my cellphone screen horizontally.

02021-09-10

The litharge–glycerin cement has set up solid, but it’s not very strong; poking it with a q-tip bent the bottom of the yogurt cup enough to crack it. It’s still shiny as if it were wet; hopefully this means it is nonporous (as it should be) and it will remain solid this time and not crumble. Mina reports that it has a strong smell of paint.

The aluminum foil in bleach left a cloud of small dark particles but otherwise is gone. The solder in bleach has corroded significantly: brown-black, like iron rust, where it was submerged, and a fairly voluminous white that looks like fungus above the water line. If I had to guess, I’d guess the white was stannous oxyhydroxide, stannic oxide, or plumbous chloride, probably not the last since it only formed out of the water. WP says stannous oxyhydroxide is “easily oxidized to stannic oxide by air”; the chlorides of tin are white too, but they’re highly water-soluble. No idea what the brown-black corrosion is.

In the evening, the litharge–glycerin cement is less shiny, with more of a matte look.

I put some litharge, without glycerin, in the bottom of the deodorized fish can and heated it on the stove. No change was visible, but the can started to smell of burning plastic. In order to be able to discard it safely, I added a layer of diammonium phosphate granules on top and continued heating; this produced a noticeable aroma of ammonia and a slight sizzling sound. An aspirator for gas scrubbing would be really helpful for this kind of thing.

After heating it gently for half an hour or so, there was no visible change, but plenty of ammonia, so I decided to resort to more aggressive measures. I put some diammonium phosphate in a cut-off Speed can and heated it with the brass torch, which melted it and released a lot of ammonia vapors. Worse, though, it released a lot of burnt plastic fumes from the Speed can paint. The mass of bubbling phosphoric acid with phosphates of ammonia dissolved in it was black wherever the torch did not heat it to orange heat. After several minutes and a visible haze of white smoke in the living room, I gave up on this approach as well.

After cooling, the Speed can had a noticeable smell of acid, with a mass of what appeared to be carbon foam on top of a whiter porous mass of, presumably, a mix of ammonium phosphates and phosphoric acid. I rinsed it into the fish can with the litharge and fertilizer, moistening the fertilizer (some of which had stuck to the litharge) and added baking soda to the remainder, which fizzed enthusiastically.

The silicone flower Mina made on 02021-08-21 has faded from its original lilac color to a pale cyan. I think the magenta food coloring in its surface has faded from exposure to light; Mina points out that on the bottom, where it’s been exposed to air but not light, it’s still lilac.

02021-09-11

The litharge/phosphate mix is kind of syrupy with white chunks and smells like a rusty engine. The phosphate/baking soda mix smells acid, and on adding more water it resumes bubbling; maybe I didn’t add enough baking soda to fully neutralize the acid, which I guess is less surprising considering that it’s potentially triprotic. The litharge–glycerin cement still seems to be hard and is now even less shiny, like dried “satin” paint.

02021-09-14

The litharge–glycerin cement has mostly remained solid, rather than crumbling under its own power, but can now be crumbled like dried mud under the pressure of a q-tip, which is similar to its previous uninspiring performance and significantly less impressive than I was hoping for. Where it was exposed to light (with or without air) it has darkened from its buff color to a darker, grayer color. It pulls away from the plastic yogurt cup (recycling triangle 6, polystyrene) with only the lightest adhesion. Its “satin” paint luster remains unchanged.

I bit through one of the Oogoo samples I’d previously bitten on 02021-08-21. If there is an acetic-acid smell, it is faint enough that I can’t be sure it’s present. It’s about 18mm thick in one dimension, so probably the points nearest the surface were something like 7mm away from the surface.

I rubbed the permanent marker spot off the surface of another of the Oogoo samples, this one about 28mm in diameter, and cut it in half with a razor knife; it too has no discernible acetic-acid smell. I’m not sure exactly when I made this one.

02021-09-15

The litharge-glycerin cement I unstuck from the yogurt cup yesterday has largely lost the bisque or buff color it previously had underneath, presumably due to being indirectly exposed to the sunlight through the window for a single day. I wonder if you could use this property to make photographs? It would be more useful if you could stabilize it, but possibly you can; plausibly the photoreaction product is something like lead dioxide or metallic lead, either of which would be much less easily dissolved by acids than litharge itself.

I added some more water and baking soda either yesterday or the day before to the ammonia-scented Speed can containing the product of decomposition of the fertilizer. Later, I added some hardware-store phosphoric acid (Desoxidant fosfatizante TF3, UPC 0-723540-593022, water-clear with unspecified additives). This induced effervescence, showing that the baking soda hadn’t been entirely consumed.

02021-10-15

The various wet substances in open containers have dried up. I have thrown out some things whose identity I could no longer remember. The egg white with green vitriol is quite orange.

I tried heating up a flake of whitewash taken from some nearby political graffiti with the brass torch. I was able to heat it to orange-hot, glowing visibly even in direct sunlight. A smell of burning paint (like linseed oil) ensued, suggesting that it was contaminated with non-whitewash paint. The hottest part gradually crumbled away over several minutes.

02021-12-05

A laptop floppy disk casing seemed suspiciously light, leading me to suspect it of being magnesium. But filings cut from it with a hacksaw do not ignite with the brass torch and do not react with vinegar, so it’s probably just aluminum. (At first the vinegar did not wet the filings; some dish detergent solved that problem.)

I have had some coupons of corrugated cardboard from a cardboard box floating in (originally) saturated aqueous muriate of lime for weeks now. The one I haven’t turned over still isn’t wet on top. I added another one to a bowl of just plain water a few days ago; it also isn’t wet on top. Normally I would have expected capillary action to soak them through in a few minutes, so I guess they’re treated with something to make them slightly hydrophobic. I was hoping that the calcium chloride would convert the sodium silicate adhesive into water-insoluble larnite, but in fact the layers of cardboard peeled apart in warm water just as easily as cardboard normally does, and a little rubbing got it thoroughly wet. I’m not sure if the paper is delaminating or if the glue is, but either way the result isn’t a water-stable cardboard as I was hoping.

02021-12-13

No visible reaction at first between the hardware-store phosphoric acid and aluminum foil in a plastic yogurt cup, but after an hour or two slow bubbling was detectable. After some 10 hours most of the foil had dissolved, leaving a slightly gray transparent liquid with some scraps of foil floating on the top of it, so I added more foil. I was expecting the foil to react, but not to dissolve — although both aluminum and phosphate are trivalent, I think of aluminum phosphate as a single material, one that’s extremely water-insoluble. But in fact there also exists aluminum dihydrogenphosphate (aluminum phosphate monobasic, 磷酸二氢铝), which might be soluble; American Elements claims it is, and Alfa says they sell it dissolved 50/50 in water, so it might be difficult to redissolve after it precipitates, or perhaps disproportionate.

Also, I forgot to note this earlier, but the acid had no visible reaction with oil of lime, presumably because the pKa of muriatic acid is about -3.0 and the first pKa of phosphoric is +2.14. I suppose that if I partly neutralize it first with baking soda I should be able to precipitate various phosphates of calcium. Interestingly, https://en.wikipedia.org/wiki/Disodium_phosphate mentions a reaction going the other way as well, preparing disodium phosphate from “dicalcium phosphate” (CaHPO₄) and sodium bisulfate (NaHSO₄), which is sold in hardware stores to lower the pH of your swimming pool.

Sodium bisulfate itself is a GRAS intermediate byproduct of the Mannheim muriatic acid process, “an exothermic reaction that occurs at room temperature”; on heating it transforms to anhydrous (58°), pyrosulfate (280°), and finally sodium sulfate and sulfur trioxide (460°). It’s a dry powder at room temperature that aggressively consumes azania. Pyrosulfuric acid is interesting as a legal, stronger alternative to sulfuric acid. The bisulfate ion is also available from thermal decomposition of azanium sulfate (250°), everybody’s favorite GRAS thermally stable azanium compound — sold for US$1/kg as a fertilizer, and made from gypsum and hartshorn by throwing down chalk.

(Sodium thiosulfate, aka hyposulfite, Na₂S₂O₃, is a different material: it’s a photographic fixer, reducing agent, and antidote to cyanide poisoning together with sodium nitrite.)

(Sodium azanium sulfate dihydrate is “a well known ferroelectric” and presumably the reaction product with azania. Presumably if you heat it you can expel the azanium and regenerate the azanium absorbent.)

Related to other useful materials to avoid witch-hunt persecution, NaMnO₄ (V-2 oxidant, Condy’s Fluid) can reputedly be made from MnO₂, NaClO, and NaOH, with an analogous route available for Ca(MnO₄)₂. MnSO₄ is readily available as fertilizer (US$5/kg); the nitrate (obtained, for example, by reacting with calcium nitrate) can be decomposed at 400° to yield the oxide, which can be purified by anodic deposition. Dehydration of the lye using the oxide supposedly yields Na₂MnO₄, whose disproportionation is another possible route, the one originally used by Condy, but apparently a pentavalent manganese compound is produced additionally or instead.

Another super fun material would be potassium ferrioxalate; hardware stores sell “salt of lemon” to remove rust stains, and I guess it works by complexing the insoluble ferric iron(III) ions, which I guess should produce K₃(Fe(C₂O₄)₃)₂. It’s fluorescent green (literally fluorescent in solution) and decomposed by light or heat (reducing the iron back to ferric), a property that can be used to used to make blueprints as an alternative to ferric ammonium citrate (apparently Michael J. Ware invented this in 01994 and wrote a fascinating book covering this and every other aspect of cyanotyping); the ferric iron then reacted with the potassium ferricyanide to produce insoluble ferric ferrocyanide, Prussian blue.

02021-12-14

The friselina seems like probably the best material for anode bags and similar filtering: it’s damned cheap, probably PET, safer and cheaper than asbestos, perhaps more inert than glass fiber, certainly more inert than any natural fiber, and PET is surpassed in its inertness among the organic polymers only by polyethylene, polypropylene and fluoropolymers, which are more difficult to find in the form of cloth, though coarse polypropylene cloth is available for carpeting and tarps. (Some vendors also offer nonwovens which purport to be both polypropylene and polyester, or blends of the two.) It may also be useful as a lint-free cloth for cleaning.

HEPA air filters are generally meltspun nonwoven polypropylene; this is a possible alternative source for polypropylene fabric for anode bags. Surgical scrubs and lab coats are another possible source.

The reduction of Prussian blue to Prussian white (produced by fading of Prussian blue by light! and reversible by wet air exposure, therefore inexhaustible) needs to steal an electron from somewhere, so you can use it for photographic patterning of a surface by oxidizing other random things in the environment. I don’t think anyone has done this.

02021-12-18

I heard the other day that you can electrodeposit aluminum from aluminum chloride dissolved in molten methylsulfonylmethane, which is easily available as a dietary supplement. (Aluminum chloride by itself won’t work at atmospheric pressure; it sublimes at 180°). I can’t wait to try it.

02021-12-21

Some time ago I deposited Elmer’s Classic Glitter Glue Silver (UPC either 0-026000-185073 or 0-26000-18191, depending on which label you believe; ingredients: deionized water, PVAc emulsion, polyvinyl alcohol, and craft glitter (various sizes and colors); the US marketing label lists no ingredients but apparently Argentine law requires it) on a sheet of LDPE and let it dry. As deposited, the glitter flakes are dispersed almost isotropically, without any preferred orientation, but upon drying (I’m guessing to about a tenth of the original thickness, suggesting about 90% water content by volume) the glitter flakes were mostly parallel to the surface as indicated by their optical properties, ±10° or so. The dried PVA/PVAc mass peeled off the thin polyethylene sheet fairly easily, as you’d expect. (See 3-D printing in poly(vinyl alcohol).) See Xerogel compacting for the implications.

To get a slightly more precise read on the water content than “about 90% by volume”, I took a polypropylene bottlecap weighing 1.7 g and deposited 3.0 g of the glitter glue in it, for a reported total of 5.3 g (!? fuck this scale) and allowed to dry in the sun all afternoon. At night it weighed 4.3 g and was still noticeably soft, though with a tough glittery skin, so I propped it up in front of the air conditioner condenser fan. I hope that will help it dry faster.

I also have a thin membrane of the dried glue (of mysterious origin), thin enough to be almost entirely transparent, which I sprayed with a saturated aqueous solution of borax. Upon drying, it was crinkly and “harder” than before; previously it was stretchy and “soft”, but now it’s stiff and brittle. I rinsed it again with water and it immediately became a soft gel, coming apart a little, and now it is drying again.

And I’ve deposited bits of the glue on some more LDPE, some aluminum foil, a PET bottle, some polystyrene, and some ABS. My expectation is that, when dry, it will have failed to adhere strongly to any of them and will be easily peeled off; “Elmer’s” (Newell Office Brands) advertises that it's “ideal for paper, cloth, craft, etc.”

From Podsiadlo, 9 other authors, and Kotov’s work (“Ultrastrong and Stiff Layered Polymer Nanocomposites”) in 02007, I learned that PVA can be strongly crosslinked with glutaraldehyde; yesterday I learned that glutaraldehyde can be easily bought from medical supply stores as a heavy-duty disinfectant (typically at concentrations of 2–2.5%, lower than the concentrations used for crosslinking). It’s somewhat hazardous to the humans because it’s great at crosslinking proteins and they are made of proteins. Reportedly it does not reduce the solubility of pure starch because what it does is bond amine groups to hydroxyls, so it can crosslink proteins to starches but not starches to starches, but that can’t be the whole story because PVA is nothing but a saturated carbon backbone with hydrogens and hydroxyls, and evidently glutaraldehyde is great at crosslinking it. 10% chitosan apparently is effective at making starch cross-linkable with glutaraldehyde.

02021-12-22

The glitter glue pulled itself free of the LDPE, polystyrene, and aluminum foil while drying. From the ABS and PET it didn’t peel off spontaneously, but it easily peeled off by hand. In the case of the ABS this may be because it was spread thinly over a rougher surface, rather than because it had more adhesion. Also, the polystyrene may have a coating on it, and all the materials may be contaminated with skin oil.

02021-12-23

The glitter glue bottlecap brought in from outside after two days of drying in warm fan air weighs 2.6 g. Upon separating from the bottlecap the relatively dry glue weighs 0.9 g and the bottlecap a reassuring 1.7. This suggests that the 3.0 or 3.6 grams of glitter glue I originally deposited was at least 70% water, but unless it’s still retaining some water inside, it’s not close to 90%.

But honestly it probably is. I should break it up and put it in a desiccator with some muriate of lime.

Topics

Materials (59 notes)
Pricing (35 notes)
Experiment report (14 notes)
Bootstrapping (12 notes)
Aluminum (10 notes)
Phosphates (9 notes)
Foam (9 notes)
Waterglass (8 notes)
Argentina (7 notes)
Refractory (5 notes)
Aluminum foil (5 notes)
Vermiculite (4 notes)
Poly(vinyl alcohol) (PVA) (4 notes)
Heating (4 notes)
Glass (4 notes)
Kingery, the father of modern ceramics (3 notes)
Alabaster (3 notes)
Oogoo (2 notes)
Pumice
Borax