Sciencemadness Discussion Board - Making borosilice fritted glass for distillation

Making borosilice fritted glass for distillation

semiconductive - 23-4-2018 at 14:25

Does any one know of a simple way (closer to foolproof) to make a borosilicate glass frit?

I'm wanting to modify a vacuum distillation tap to protect the air/vacuum pump from polar chemicals using a frit barrier and hydrophobic coating.

I'm concerned that if I just crush up borosilicate glass vials into frit and try to melt/bond it with a oxy-acetyline torch, I will either have loose glass or fused air-tight as it's hard to heat everything perfectly.

I was hoping someone might have experience with a mixture that would allow me to make reasonably consistent (amateur) glass frit/porous glass plate.

Perhaps a method like mixing water glass, borax, and some kind of pore-maker chemical; eg: table salt or another salt which would not fuse with the glass so that the glass can be torched heavily with the salt present -- but then cooled and rinsed to produce "holes"?

I'm wanting to make and bond a borosilicate frit to a borosilicate vacuum take off (eg: having same coefficient of expansion) . See photo:

driptube.png - 26kB

I'm wanting to make the frit in the green marked area outside the drip tube but inside the adapter.

Background/reason why:

I've found that I can distill H2SO4 at much lower temperatures by making a reflux column with a condenser tube Tee at the bottom and a standard consensor below. The condenser tube is kept just above the boiling temperature of water. Then I pump air into the boiler flask blowing it on the surface of the acid and then evacuate the same air after the collection flask. The result is similar to steam distillation, but with most of the water staying in the boiler flask and reflux column. I'm able to get reasonable quantities of concentrated and distilled H2SO4 at 150C or below. I can then reflux the end product to make it more anhydrous as needed. This technique is probably useful for other chemicals as well, so I'd like to make a general purpose apparatus.

The only issue I had was that I need to keep as much of the sulfuric vapors out of the air pump as possible while recycling the air.

I can reduce fumes getting to the vacuum pump by putting a glass bead on the vacuum takeoff valve adapter to "plug" (not seal) the drip pipe. See the red dot in the picture. That's where I put a glass bead. As soon as moisture covers the bead, air pressure will push droplets down the drip pipe and they will collect as liquid. Almost no sulfuric fumes remain in the air. However, after sitting for a long while ... some of the liquid will re-vaporize and since the vacuum take-off is completely open -- some vapor goes into the pump.

I thought of just making a sodium carbonate trap before the air pump; but would like to try a permanent -- no maintenance method -- to remove as much acid as possible before a chemical trap.

My thought is that the ideal , non-exhaustible, pre-filter might just be a glass frit (similar to the glass bead) . The frit would again condense fumes out of the air because it's hydrophilic. To prevent reasonably strong acid from being forced through the frit by vacuum , I can just dissolve a hydrophobic substance like silicone caulk in MEK and coat the top of the frit with it. A thin coating of silicone will pass gasses but not liquids. eg: silicone will not pass water or acids -- but it will pass air, slowly.

aga - 23-4-2018 at 14:48

Conc sulph seriously cannot be worth this amount of effort.

semiconductive - 23-4-2018 at 14:59

You're correct; I can make concentrated sulphuric by just concentrating battery acid from NAPA. But making the acid isn't the only reason for distilling it ... There are several other reasons; such as recovery of sulfuric from a reaction and not putting waste acid in a landfill, not flushing it down my septic system; etc. Being able to recycle sulfuric at low $$ is a good thing to be able to do.

There are other chemicals that this apparatus will recycle as well. Anything that lowers the heat required to recycle chemicals also reduces bills. esp. as I scale the process up.

j_sum1 - 23-4-2018 at 15:05

Really good question.
There are a couple of glass blowers here who might have some specific knowledge. Let me give my (totally unsupported) thoughts.

1. I think you will need uniform heating -- no flame which will give localised hot spots and hence melting. You need to sinter the glass frit while keeping the remainder of the structure unchanged. This will mean a kiln.
2. I would assume that the composition of the frit is different from the borosilicate. You do not want the structural glass softening and deforming. You might get some progress with sodalime glass and controlling the temperature. You might also try a mix of crushed sodalime and borosilicate powders.
3. You might get some success with getting piece of frit and sticking it in place with epoxy putty. If it works it will be an easier solution -- even if you need to replace it every few runs.
4. My first attempt would be to see if I could achieve the required result using a rolled up wad of glass wool. That may well be the simplest solution of all.

j_sum1 - 23-4-2018 at 15:14

Or buy this and use it with other adapters (female to female) to get the configuration you want.
https://www.ebay.com/itm/A-C-Glass-Medium-Fritted-Vertical-V...

Edit
In closer look, not exactly what you want. But you might be able to find something.
Alternatively, if you are ok with melting glass, it might be easier to dremmel and weld bits of glassware together. This could be better than making your own frit.

[Edited on 23-4-2018 by j_sum1]

semiconductive - 23-4-2018 at 15:47

Wow, I hadn't seen that before on ebay. I'm using 24/40 ... so I would need to buy the $39 version. but....
Unfortunately, the frit on the adapter you found is on the end where the glass bead normally gets put. (red mark on my diagram).
That only solves the easy part of the problem...
The vacuum take-off is still open to fumes ... so it's not an off the shelf solution.

If I could find a cheap piece of frit the exact right size -- rather than $89/random sized fritted disc, I could probably just use water-glass to bond it to the body.
But I don't see any easy way to locate a borosilicate frit disk of exactly the right size with a stem hole in it; and I'm not sure how difficult it is to drill a hole in a frit filter plate.
Otherwise, I do know water glass dehydrates and becomes silicon dioxide at a much lower temperature than the melting point of borosilicate.
Once dehydrated and washed with strong acid, burnt silica gel is chemically inert.
So, in theory waterglass would be an easy way to weld a borosilicate frit to a borosilicate body using a torch without warping the body.

I can make a kiln, if necessary. That's actually not hard to do. But, I'd like to know an actual glass blowers advice before trying that. I want to be sure to succeed if I go to that much work.

Edit: I saw your edit a moment ago;

I will look for glass wool; that's probably a good alternative if it works.

As to cutting glass and welding it:
Thick glass tends to break strangely when I score it. I tried making an adapter with an extra long european joint short by scoring the joint and breaking it -- but it broke wrong. I would really appreciate knowing if there is an abrasive or way using a dremmel to cut glass at home with less risk than scoring glass.

In order to weld I must first cut off the existing vacuum tube in order to be able to weld a frit filter / buchner funnel / chromatography column in it's place. It would need to be welded such that liquid condensate refluxes back into the collector jar.

Being able to cut odd shaped glass reliably is important to me; I'd love to learn the skill if anyone knows how.

[Edited on 24-4-2018 by semiconductive]

Heavy Walter - 23-4-2018 at 16:17

I work with a glassblower who prepare its frits.
He grinds in a ball mill pieces of borosilicate glass.
Depending on the desired porosity he choses a particle fraction using sieves.
Then he fills an iron mold –cylindrical, typically 5 mm height, previously brushed with colloidal graphite- with the selected fraction.
Then puts it in a kiln, at some specific temperature depending on pore size for a specific time.
Then he seals the frit to a joint.
Having other questions, I can ask him.

[Edited on 24-4-2018 by Heavy Walter]

RogueRose - 23-4-2018 at 16:43

I don't think it should be too difficult as I've watched a glass blower do something similar. I've been keeping my broken boro (quality Kimax & professional Pyrex) for things like this. It will all depend on what type of boro you are working with but a good rule of thumb is that it has a MP of about 1500F and an annealing point of about 1,020F.

You have to make sure that you aren't working with quartz glass as that has a much higher melting point and if your frit is quartz, then your "funnel" is going to melt well before if it is boro. Now Wiki does list that lampworking/scientific boro glass has a melting point of 3000F, which I think is totally incorrect b/c this is just under the MP of SiO2 at 3155F. Now maybe the glass they are talking about is a 99-99.5% SiO2 & .5-1% B. Now "fused silica", which is what quartz glass is also often called, is stated as having a MP of 3000F.

I would take your scrap glass, crush it w/ mortar and pestle and then sift it with different size screens/strainers, etc. you can make something for small particles by using a needle or pin and poking small holes in thick paper and then using that to sift the finer particles through, I've done simialr and it is slow, but it will work. for the size you have it may not take a lot of time.

You might also want to try a coffee grinder for a few short bursts. I'd also consider taking something like some ball bearings of nuts or even bolts (ones of moderate size of 1/4" diam minimum for bolts), put them in a sealable container with your fine-ish glass, close lid, then shake/swirl vigorously. This should help break off and round off some of the very sharp edges which can really cause some problems by making large spaces between other pieces. Don't go overboard unless you want finer particles. You should be able to feel a big difference in the texture after doing something like this. Of course a ball mill would work as well, but the other way allows for more visual control I would think.

If you have never worked with melting glass then try to melt and bend some tubing if you have it, old broken lab glass, any light bulbs of halogen, metal halide (high pressure sodium or Mercury Vapor), though many halogen may be quartz (most of mine are).

Once you have the particle sizes you want, you can fill the entire bottom with the fine glass. When you apply a blow torch to the outside (to heat outer glass evenly) this expands the glass and then you can apply the heat to the top layer on the inside until it gets a very slight orange (best done in low light conditions). This will make a layer that should hold in place if done correctly and allow most all the glass underneath to fall out. This could be a thin layer and that is good as you can place more on top of this and heat again to give a slightly thicker layer.

Without better pics of how the adapter looks from the bottom, I don't know what else to say. I think you should be able to do this with a standard propane blow torch with no problem (as long as it is standard borosillicate). I'd be very hesitant to use any gas that is hotter. I have used the least expensive benzomatic propane torch with a not-so aggressive flame head (not extremely focused), with great success joining1.5" diameter glass tubing to each other (I use one torch for each tube, then one when pressing together to heat).

semiconductive - 23-4-2018 at 17:50

@Heavy Walter: I think I can do that if you give me the specifics I need to know.
The frit filter I'm trying to make shouldn't be very fine eg: A. Coarser frit makes liquid flow easier and drip out, so a typical coarse buchner funnel kind of frit, or even a little coarser would be OK.

I have a rock tumbler that can be used as a ball mill.
The plastic, though, won't work well. It'll get mixed in the sharp glass.

However, I can put borosilicate glass vials in an iron can with an iron bolt and place it on the rock tumbler.
if I use Iron, maybe a rare earth magnet will be able to get rid of iron dust found in the glass after milling -- unless there is something better your friend recommends.
I can also weld a second bolt to the side of the can, which will lift and drop the glass and iron bolt repeatedly.
I'm not sure what size can I need to use, but I'm sure that can be made to smash glass up into all sizes of frit.

I've made sifters before. I used wire screen and plastic plant pot bottoms as a frame. I epoxy the mesh to the pot, and cut a hole in it to let the dust out.
American window screens are about 20 mesh; 1/20 = 0.005 inch ~= 1.27mm
Is that a practical mesh for smallest particles in a coarse frit filter ? or is it too fine?
Secondly, what size mesh should I buy to get rid of glass particles that are too big?

To make the sintering work,
I imagine that your friend waits until the kiln has gotten to some minimum temperature before he begins timing the sintering.
Then he measures a time and has some kind of temperature profile he likes.

I have thermocouples and a computer control, so I can measure temperature to 0.1C and control the heating and cooling of a kiln very repeatably.
But I need to know how quickly it needs to heat up, and to what temperature(s), and how long it's to keep it there, and finally how quickly to cool down.
I appreciate any information you can provide; as that's the difference between being able to do the project this way ... and not!

[Edited on 24-4-2018 by semiconductive]

semiconductive - 23-4-2018 at 18:04

Quote: Originally posted by RogueRose

You have to make sure that you aren't working with quartz glass as that has a much higher melting point and if your frit is quartz, then your "funnel" is going to melt well before if it is boro.

I have borosilicate glass tubing, B3.3 I think it was advertised as. I also have vials that have the temperature coefficient of borosilicate glass when I measured them. I seriously doubt they are fused quartz, because they are quite inexpensive. This is also why I was thinking to use water-glass and borax, because those two things together if mixed in the right proportions ought to make borosilicate glass. Water glass is sodium silicate, and borax is sodium borate. Borosilicate glass, commercially, always has sodium impurity but not calcium. So, I ought to be able to make borosilicate glass if I knew the right ratio to mix.

Quote:

If you have never worked with melting glass then try to melt and bend some tubing if you have it, old broken lab glass, any light bulbs of halogen, metal halide (high pressure sodium or Mercury Vapor), though many halogen may be quartz (most of mine are).

I've gotten defective joints from china, before, that had wrong angles on them and heated them with my oxy-mapp gas torch and a Harris precision torch body. The flame can be turned down very accurately and spread to control average temperature. It made bending borosilicate glass simple. So, I'm not to worried about bending and welding; I'm just not sure I will be able to tell when a frit is melted properly or too much. Eg: the "art" side of things isn't my specialty. That's why I'd rather do something that has a foolproof safety, or else build a precision computer controlled device to make the frit correctly on the first try.

Quote:

Once you have the particle sizes you want, you can fill the entire bottom with the fine glass. When you apply a blow torch to the outside (to heat outer glass evenly) this expands the glass and then you can apply the heat to the top layer on the inside until it gets a very slight orange (best done in low light conditions). This will make a layer that should hold in place if done correctly and allow most all the glass underneath to fall out. This could be a thin layer and that is good as you can place more on top of this and heat again to give a slightly thicker layer.

I can try that on top of a fire brick with the frit before ever risking the adapter; but I have a sinking feeling I won't get it right. If there is loose glas on the bottom after sintering as you suggest, then if I make more layers I might not get both layers to fully sinter or go too far. The loose glass in the middle could be a problem.

Here's what the bottom looks like. Just a drip tube sticking out of a pipe that has a vacuum port on the side.

[Edited on 24-4-2018 by semiconductive]

Dr.Bob - 23-4-2018 at 18:25

I have some various 24/40 m/f jointed tubes with a frit in the middle. Not perfect, but might work. I'll look for anything like what you posted, but I might have something close.

Bert - 23-4-2018 at 18:59

Ace glass is pretty good (excepting the prices- eeep!) equipment in my experience.

They make their fritted glass filters from chopped glass fibers. Never tried this, I have cut glass and done small fusing and glass blowing operations only (and poorly at that).

The sincerest form of flattery is plagiarism-

http://www.adamschittenden.com/frit_info.html

Please do report on your progress,.if you ever get so far as making such tools.

semiconductive - 23-4-2018 at 20:55

Quote: Originally posted by Bert

The sincerest form of flattery is plagiarism-

Thanks!
There is nteresting information on that link ... but its a bit incomplete / stumping.
I'm planning on using an aquarium air pump, as it's cheap, light, low-power, and the rubber diapragm can be coated with silicone to improve chemical resistance a little bit.

So, I know my pressures are limited to around 30millibar and 250mL/minute airflow.
There should be a relationship between frit size and resistance to air that allows me to predict how wide and thick various glass frits need to be in order to stay under 30millibars pressure.

I know an analogy that let's me relate flow rate to a crude figure of merit -- the "resistance" or relative resistance of a frit.
In electronics, electron flow in a conductor is similar to a fluid through a porous surface. There are obstructions causing collisions and "resistance" to flow. So, The exact same proportionality formula ought to apply to glass frits.

Flow is proportional to pressure / "pore_resistance" * ( diameter**2 / thickness )

However, Adams/chitten doesn't show the thickness of the frits that they are measuring in their guideline chart. If I assume the discs are all the same thickness, the proportion chart should be computed as the square of (size/30mm).

http://www.adamschittenden.com/Robu%20flowrate%20graphs.pdf

diameter AdCh theory difference.
10mm 0.13 0.11 84%
20mm 0.55 0.44 80%
30mm 1.0 1.00 100% ** nominal size and factor ( 30/30 )**2 = 1
40mm 1.5 1.78 119%
50mm 2.5 2.78 111%
120mm 9.7 16.0 165%
250mm 17.4 69..44 400%
450mm 32.5 225.00 692%

However, the formula doesn't work ... so either it's wrong, or the thicknesses of the frit discs generally get thicker as the discs get wider.

In another document, AdCh shows that "coarse" frits have average final pore sizes of 60 microns, eg: 0.06mm -- which is about 1/20th the size of a window screen hole ~1.3mm. However, I don't know how the average pore size is related to the average glass particle size or how to compute that ....

AdCh does explain that coarse frits are about 40% (by volume) empty space.

There is probably a way to estimate particle size from that figure ; I just don't know it off the top of my head.

Edit:
The chart AdCh shows is actually from ROBU glass in Germany. I found a ROBU thickness specification, and frit discs do increase in thickness as the diameter increases.

30mm wide is 3.5mm thick; so the difference percentage for selected diameters can be computed based on 3.5mm nominal thickness; Here's a table of theoretical difference, vs. difference found in previous chart:

http://www.robuglas.com/en/sintered-glass-filter/filter-disc...

10mm 2.8/3.5 = 80% vs 84%
20mm 3.0/3.5 = 85% vs 80%
30mm 3.5/2.5 = 100% vs 100%
40mm 4.5/3.5 = 128% vs 119%
50mm 4.5/3.5 = 128% vs 111%
120mm 8.5/3.5 = 243% vs 165%
250mm 19/3.5 = 542% vs 400%

The general trend is correct, but there are fairly large random errors on top of a systematic one.

It's hard to tell what is causing the difference in percentage thickness from theoretical, but the correlation with thickness suggests that as frit gets thicker, the pore sizes also get larger. I imagine that the particles near the surface of the frit might be more melted together than those farther from the surface; so the thicker the glass, the less well sintered the particles would be near the center of the filter.

So, I think my formula is probably correct; but process variations vs. thickness are an unknown (on top of accuracy of measurement being questionable, because manufacturing tolerances are not specified.)

Anyone else have any knowledge that I might have overlooked as to why the frits vary?

[Edited on 24-4-2018 by semiconductive]

wg48 - 24-4-2018 at 03:01

As others have said it will be tricky to heat a pad of glass powder with a flame such that it only fuses the points of contact of the particles but not so much that it melts in to a pad of glass with air bubbles. It will be much easer with a small temperature controlled furnace but some experimentation will be required to get the correct time temperature. A programmable ramp and soak type controller would be ideal.

Perhaps I do not understand what you proposing in your distillation set up. You seem to be suggesting that by blowing air over the surface of hot dilute sulphuric acid it will cause more concentrated acid to be volatised and leave more diluted acid in the flask. I would expect the vapour would be more dilute acid or mostly water and what is left in the boiling flask to be more concentrated acid.

Incidentally it usual to bubble the purge gas through the solution and that will have the added advantage of reducing bumbing.

How does the frit stop sulphuric acid vapour? Yes it may capture sulphuric acid mist (droplets of acid) but not vapour or did you mean mist when you used the term vapour?

highpower48 - 24-4-2018 at 08:02

Doesn't Deschem offer to make custom glassware. Not sure but that might be easiest and cheapest way to go. If you don't mind Chinese glassware.

semiconductive - 24-4-2018 at 08:20

Quote: Originally posted by wg48

As others have said it will be tricky to heat a pad of glass powder with a flame such that it only fuses the points of contact of the particles but not so much that it melts in to a pad of glass with air bubbles. It will be much easer with a small temperature controlled furnace but some experimentation will be required to get the correct time temperature. A programmable ramp and soak type controller would be ideal.

I have a program, written in Python, that has a PID servo type loop. It takes a list of times and temperatures and linearly interpolates between the list elements to make temperature/time setpoints. I really hate doing a lot of experimenting when there should be a lot of people in the world who already know the answer. Glass making for chemistry is something that lots of people have known in the past. Perhaps we're loosing knowledge due to industrialization of labware.

I can make any piece wise temperature profile and repeat it, so long as the diffusion time of kiln-heat is small compared to the programmed ramp rate. Eg: the smaller the kiln, the more precisely the temperature will be controlled.

The loop feedback time is three seconds.

Quote:

Perhaps I do not understand what you proposing in your distillation set up. You seem to be suggesting that by blowing air over the surface of hot dilute sulphuric acid it will cause more concentrated acid to be volatised and leave more diluted acid in the flask.

No, not quite. The volatiles will be at whatever concentration the moisture content allows.

Quote:

I would expect the vapour would be more dilute acid or mostly water and what is left in the boiling flask to be more concentrated acid.

Yes, that's probable.

Quote:

Incidentally it usual to bubble the purge gas through the solution and that will have the added advantage of reducing bumbing.

That would not be a good idea, as there are liquid impurities in the boiler. If I bubble air through the liquid, droplets of acid with impurities will be splashed into the air stream as bubbles pop. I have thought about making a straw like shape out of porous frit with larger holes in it. The frit would "wick" the boiler liquid up and give more surface area of boiler liquid for the air to come into contact with, but without splashing or bubbling.

Even without a "wick"; There is no bumping because the sulfuric acid is not boiling. At 125 to 150C, we simply have humid air in laminar flow across the steaming liquid's surface. Water will steam out slowly, but not boil, because the acid has raised the boiling point of water significantly.

In an actual experiment; The air temperature was actually close to 25C when it first comes in contact with the boiler acid. ( I intend to pre-heat the air in future versions of the still). So, there was a mixture of water droplet (fog) coming into contact with the acid and steam as the acid heats the air. As the air warms, the steam droplets disappear.

Quote:

How does the frit stop sulphuric acid vapour? Yes it may capture sulphuric acid mist (droplets of acid) but not vapour or did you mean mist when you used the term vapour?

You've overlooked a couple of things.
The frit after the collection flask is below the boiling point of water. Therefore, dilute water condenses on the frit. Whether the acid vapor is droplets or individual molecules of acid is unknown and likely a mixture. However, in both cases ... the acid will either condense on the cool frit, or it will bond to the water that is present on the frit.
I want a large surface area (thick frit) if possible to maximize the number of times humid air bounces or comes into contact with a hydrophilic surface.

Because the frit is polar, I expect it to attract water molecules that are polar faster than condensation would happen on a cool hydrophobic surface. Once a cool water film exists, it will collect acid vapor or droplets equally efficiently. Because H2SO4 is hygroscophic (it a great dessicator) H2SO4 will "stick" to the water film preferably to staying as vapor in air.

A human nose is capable of detecting sulfuric acid vapors in the part per million. Even a glass bead with a water film on it is able to make the odor so faint as to be nearly undetectable for several hours. So, I'm gambling that a glass frit will be at least as effective.
If I do not put the glass bead on the opening, however, there are open spaces where air can get into the collection jar without coming into contact with a water film. In that case, the smell of sulfuric acid is chokingly strong. Even though I would expect the air to touch the condensor glass, or the drip tube glass, a significant amount of air does not. It has to be forced into contact by a glass bead which has only microscopic sized gap.

The other thing you are overlooking is that there is a reflux column above the boiler (vigeruex). The vast majority of the water vapor condenses on that reflux column and drips back into the boiler from the center of the vigerux. The top of the reflux column is capped with a marble.... so very little acid or water escapes. Since the "condenser" three way joint / tap is below and leaves perpendicular to the reflux column; no reflux drips into the condenser. The condenser is kept above 100C and is oriented vertically after the three way tap. That causes rising steam to go BACK up to the reflux column rather than down the condenser and into the collection jar. A certain amount of water vapor will be forced down the condenser by the slow moving air, but the concentration of the acid essentially increases as it goes down the heated "condensation" column. The "condenser" column is essentially an acid reflux column in reverse. Both the reflux column and the condenser column could probably be made more efficient by packing them.

[Edited on 24-4-2018 by semiconductive]

Heavy Walter - 24-4-2018 at 10:08

Hi semi

In case of a coarse frit -the largest porosity used in labs- he employs:
Sieve sizes: 20 and 40. All that passes through the 40, goes to be the frit.
Temperature & time: 850 ° C during 15 min
Good luck!

semiconductive - 24-4-2018 at 10:24

Quote: Originally posted by Heavy Walter

Hi semi

In case of a coarse frit -the largest porosity used in labs- he employs:
Sieve sizes: 20 and 40. All that passes through the 40, goes to be the frit.
Temperature & time: 850 ° C during 15 min
Good luck!

:S

If I use everything that goes through the 40, the smaller hole size, why would i bother using a 20 mesh? Doesn't everything that goes through 40 also go through 20 mesh ?

Do you mean, everything that goes passes through the 20 to the 40, but not through the 40 is used ?

Thanks!

semiconductive - 24-4-2018 at 10:34

Quote: Originally posted by highpower48

Doesn't Deschem offer to make custom glassware. Not sure but that might be easiest and cheapest way to go. If you don't mind Chinese glassware.

No, I don't mind. Deschem's glassware looks green in the pictures ... which is odd for borosilicate glass. Lime glass tends to look green.

I'll try sending him a private message on ebay, and see what custom glass costs.

Heavy Walter - 24-4-2018 at 10:50

Hi Semi

You are right, I told it wrongly.
You have to use what rests into the 40 sieve.

Sulaiman - 24-4-2018 at 10:57

I'm dubious of the functionality of the device,
but I would try using a wad of glass wool to test the concept first,
just packed, not melted.

semiconductive - 24-4-2018 at 12:36

Quote: Originally posted by Sulaiman

I'm dubious of the functionality of the device,
but I would try using a wad of glass wool to test the concept first,
just packed, not melted.

I dont have glass wool yet, and I'm ordering 40 mesh screen. So If you have wool ... I would appreciate your trying that route if you would.

I'm still going to work on making glass frit, and am looking up information, while I wait for orders to be delivered.

I found an article with quite a bit of useful information on making efficient glass frit filters.

https://www.mvm.kit.edu/img/content/AME_Veroeffentlichung_Se...

[Edited on 24-4-2018 by semiconductive]

aga - 24-4-2018 at 12:49

In the end, you just spend time making money to Buy things that other people make all day long, rather than try to make them yourself.

In a survival situation you'd filter with bits of grass, so knowing how to make a glass fritt would not be of much use.

Knowing the exact process to make a fritt would be interesting.

Has anyone tried to contact a frit-maker ?

Last time i suggested phoning a manufacturer, they ended up with Free advice and a Free sample.

For those who do not know, 'phoning' is this Old thing that people used to do.
You use the 'phone' app on your smartphone and punch in the number of the company.

Sometimes an actual person starts speaking to you.

Crazy i know, but it happens.

RogueRose - 24-4-2018 at 13:20

Haven't read the whole thread, but you can order glass frit in different mesh sizes from glassblowing supply houses and it is pretty cheaply. I think I remember something like 4-6oz for about $8 IIRC. Do a search for borosillicate tubes and glassblowing supplies and you should probably find a site that offers this. I remember the place I saw even had some extremely cheap at a few $ for like 1/2lb but it may have not been screened for uniform size.

wg48 - 24-4-2018 at 13:51

semiconductive@

You seem to want to use the frit as a condenser surface and a molecular sieve. A condenser and or 500g of molecular sieve will work much better or just wash bottle filled with a kilo of dry sand, crushed glass or sodium carbonate granules. .

Ok so your using the condenser as a heated unpacked fractionating column. So you expect the air and sulphuric acid/water vapour to flow in at the top at the same time that water vapour flows out of the top. Why would that happen with the air flow and a condenser that must be cooler than the vapour or none of it will condense?

I don't think it will work. But try it with glass wool. Wash the pump out quickly at the end and do report back.

semiconductive - 24-4-2018 at 14:15

Quote: Originally posted by wg48

semiconductive@

Ok so your using the condenser as a heated unpacked fractionating column.

yes ...

Quote:

So you expect the air and sulphuric acid/water vapour to flow in at the top at the same time that water vapour flows out of the top. Why would that happen with the air flow and a condenser that must be cooler than the vapour or none of it will condense?

Because sulphuric acid condenses at a temperature well above the boiling point of water. The condenser is held at around 100C .... very strong sulfuric acid *will* condense at that temperature, but water does not condense. So, the principle I'm using is the same as a reflux vigereux colum ... a hotter part of fractioning column has stronger sulfuric condensing, while a cooler part has weaker sulfuric condensing.

Quote:

I don't think it will work.

I already know that the heated condenser does work. I distilled over 200mL of rooto sulfuric acid using the set-up and a silicon carbide ball bearing as a "glass" bead. Unfortunately, some of the carbon was etched... but the impurities in rooto did not make it into the collection flask. There are probably more efficient ways (and complicated) to design a still, by moving the vigereux column after the collection bottle -- with some kind of reflux trap to route all the reflux back to the boiler flask. But, I haven't figured that out yet.

Quote:

But try it with glass wool. Wash the pump out quickly at the end and do report back.

I do not have glass wool to test the cold trap after the collection bottle, but I have ordered it. There are going to be several days (possibly weeks) between now and any tests I can do. Glass wool can still leak as there is going to be a scrunched up spot where the wrapped wool comes back on itself. Holes along the seam are inevitable. But, I will still try it.

The idea with sand is an interesting possibility. However, isn't regular sand impure silica? It seems to me that reflux from a sand trap would have unknown contaminants in it. The only place I could put a sand trap is after the vacuum tap ... and that's no different than using a neutralizing chemical... So, it seems like a waste of effort and chemicals when unfired glass frit seems easy to make using a pin mill.

[Edited on 24-4-2018 by semiconductive]

Chemetix - 24-4-2018 at 15:33

I occasionally make a frit for some glassware applications, but the commercially made ones are usually the most economical way to go. If you want a longer section of frit than the standard thin discs then my technique has some merit.

Getting a frit to work is quite a process, involving many steps and well calibrated equipment. Such as a furnace and a mold that can be inserted and withdrawn with precise timing. Even when you have a frit, getting them sealed glass is another headache, it takes a good glassblower with considerable experience to get it right without cracking on you, and that's with a good reliable brand of frit with good QC.

Glass wool and bits of crushed glass or beads as packing is the way to go here.

wg48 - 24-4-2018 at 16:42

Quote:

Quote: Originally posted by semiconductive

...

Quote:

We can agree on the temperature at which different strengths of sulfuric acid condense.

You did not answer my question directly: Do you expect the air and sulphuric acid/water vapour to flow in to the top of the column at the same time that vapour with less sulphuric acid flows out of the top of the column?

You have already tried this set up and it worked?

Bert - 24-4-2018 at 17:51

A fter looking back through the whole thread?

Please do report on an actual experiment.

semiconductive - 24-4-2018 at 18:54

Quote: Originally posted by wg48

You did not answer my question directly: Do you expect the air and sulphuric acid/water vapour to flow in to the top of the column at the same time that vapour with less sulphuric acid flows out of the top of the column?

You have already tried this set up and it worked?

yes, it works. I said I already did an experiment -- but NOT with a trap before the vacuum takeoff and glass frit or floss, but only with a silicon carbide bearing where the "red" mark is in the original diagram.

I have two experimental designs, the first experiment is already posted on sciencemadness.com with photos. The second is better, but I don't have photos.

In my second experiment, the allihn condensor is oriented vertically.
There are swirls near the top of the condenser where the cooler air from the vigerux column hits the warmer air from the condenser. It's a bit like a tornado, but very slow. The cooler air appears to go down the center of the allihn condenser, while the warmer air wants to go up the sides. As the cool air warms up, I think it must expand and hit the sides of the (allihnn) condenser. That causes water/acid droplets form on the glass. Again, the allihn condenser is vertical... but even if it were at an an angle, I still think it would work. The cooler air would sink to the bottom, while the hotter air would move up to the top of the condenser. My first experiment used an angled allihn condensor with towels for insulation.

Why are you so concerned that it wouldn't work ?
Are you imagining that the air is moving fast?

There are photos in another thread of my first experiment (link below). The insulation has been removed from the horizontal Allihn condensor. eg: so you could see the condenser in the picture and not a towel. I didn't know how to use the allihn condensor at the time, so it wasn't hooked up vertically yet. Even so, the 150C boiler temperature kept the top portion of the condensor hot since I had no water flow in the cooling jacket. The allihn condensor was effectively insulated by the jacket with no water in it. I think my thermometer measured 80C when I stuck it inside the exhaust port of the condensor near the top in the shown experiment. After wrapping the connector and top of the tube with a towel (not shown) the temperature went up. The bottom half of the allihn condensor tube formed water/acid droplets while the top half stayed dry. An angled allihn condenser is NOT as effective (convection current wise) as one mounted vertically; But it still has a certain amount of convection currents happening when sufficiently insulated. The strength of the acid in the alliihn condenser is enough to dehydrate cotton. It's not at all "weak" acid. The water at the top of the vigerux column, however, will not dehydrate cotton.

Quote: Originally posted by Bert

A fter looking back through the whole thread?
Please do report on an actual experiment.

So, I clearly did answer your question @ wg48 ... you're just having a hard time believing me ?
( And superadministrator, Burt, too? or is that just good advice? )

Look at the following link, this is an earlier example before I knew how to use the allihn condensor properly: Notice the washcloth that got dripped on, and the paper envelope, have both been turned to carbon on the drip spots.

https://www.sciencemadness.org/whisper/viewthread.php?tid=79...

The fish tank pump is the blue one on top of the oscilloscope. I took the insulating towels off the allihn condensor to take the photos. The silicon carbide bearing is clearly visible in the second photo. I didn't have a third vacuum take off valve at the time, so I used a sepratory funnel and a marble with a bit of water as a sulfuric acid trap.

The vertical allihn condensor design for the second experiment is hardly more sophisticated than the first experiment I show in that photo, the only differences are a hot air soldering iron tip is preheating air forced into the bottom of the allihn condenser. I also figured out how to make plaster of paris + pearlite insulator which doesn't get ruined by acid like towels do. So insulating jackets are acid tolerant. And I have a bicarbonate soda trap before the fishtank air pump's inlet, where I used marine grade epoxy to hook a pipe over the hole where the pump sucks air. So my fishtank pump now has an inlet and outlet pipe, and an adaptor replaces the sepratory funnel so I can return air to the fishtank pump.

To get a feel for the airflow, which seems to be the mental issue people are having....

The cross section of 24/40 glassware is over 16x the area of fish-tank air hose. If you put a fish tank air hose from a pump for 2 to 5 gallon fish tank into a steam from a pan or dust cloud from chalk, you can easily see yourself that the air does not move very fast and that the cloud will "swirl" around the air from a 1/4 inch hose. In 24/40 glassware, the air could only be moving 1/16th the speed of what's in the hose -- and that's divided in half because more than half goes up the vigereux. There is plenty of room for convection air currents to form as long as there are temperature differentials of 10 degrees or so.

Again, when I don't put a glass bead inside the top chamber of the vacuum take off adapter (red color), then the cooler air makes it to the bottom of the allihn condenser (300mm long) still laden with sulfuric fumes. So, it's quite possible the experiment wouldn't work if I set the system up even slightly differently.

Everything I am telling you is obvious, and you can repeat the experiment yourself.

I don't know if the information people are giving me is accurate, but I do know what I have done. I will be happy to post mail tracking on the stuff I ordered, to prove I'm not yanking your chains. Your time might be better spent finding out if the guy from argentina is really giving me straight answers that are useful. People seem divided on whether he's given me enough information to reliably make a frit filter ; and I'm not sure either.

I get real tired of having my integrity attacked; my first thread on this site got moved to detritus because of judgmental suspicions, and the day will come where the whole world will know the person who judged me was very wrong. I'm still very sore about that. I spend more time defending myself than doing experiments when people act like Sir Isaac newton who set science back 300 years. He was a know-it-all, and was "dubious" about someone else's experiment. When certain people come into power -- their intolerance is a problem.

[Edited on 25-4-2018 by semiconductive]

Sulaiman - 25-4-2018 at 00:43

Quote: Originally posted by semiconductive

I get real tired of having my integrity attacked; my first thread on this site got moved to detritus because of judgmental suspicions, and the day will come where the whole world will know the person who judged me was very wrong. I'm still very sore about that. I spend more time defending myself than doing experiments when people act like Sir Isaac newton who set science back 300 years. He was a know-it-all, and was "dubious" about someone else's experiment. When certain people come into power -- their intolerance is a problem.
[Edited on 25-4-2018 by semiconductive]

I apologise if my use of the word 'dubious' offended you when I wrote
"I'm dubious of the functionality of the device"
I used the word meaning 'doubtful' - of the device - not your integrity.

You have proposed a completely novel setup operating on principles not completely/clearly understood by me,
and I suspect that your frit will 'choke'
so what is wrong in warning you of problems before you encounter them ?

Reading the posts above it seems to me that you equate criticising the concept with criticism of yourself,
that is just your ego causing you trouble.
You may do better if you ask why there is doubt rather than get upset about it

j_sum1 - 25-4-2018 at 01:49

I confess too that I am not fully folowing the process that you are suggesting. But I have not read every post in detail.

I would advise listening to chemetix. He does scientific glassblowing for a living and knows what he is talking about.

semiconductive - 25-4-2018 at 06:28

Quote: Originally posted by j_sum1

I confess too that I am not fully folowing the process that you are suggesting. But I have not read every post in detail.

I would advise listening to chemetix. He does scientific glassblowing for a living and knows what he is talking about.

Thanks, I appreciate you seconding chemtex.
That's very helpful.

@i_sum1, @chemtex
I ordered the glass floss, yesterday. Borosilicate floss is cheap, but I live in the country and mail order of the lower cost stuff takes time. I'll post pics when I try that method but be patient, it will likely be weeks before stuff arrives.

Sorry, my posts are so long. I never learned to condense ideas without making them too confusing. The more frustrated I am, the more verbose I get.
Hang in there chemtix, I'd love to pick your brains a little.

I think I need to show how advice in this thread compares to my budget reality ... but I want to emphasize that I don't want to try only the cheapest solution; I'm wanting the "middle" road with reasonable chances of success and lower costs/even if longer time.

When it comes to buying glass frit discs off the shelf, China is generally cheaper than American made. But buying pre-made frits doesn't seem practical for most amateurs. The lowest ebay/china introductory price is still $89.

https://www.ebay.com/itm/10mm-150mm-Sintered-Glass-Discs-Fri...

I can do a lot of home-brew experiments for $89.
I've done lots of searches ... and there *are* random sized frit discs sold by individuals (used) for $8, or teflon gaskted frits for buchner funnels at around $25 each; but, basically, glass frits are custom made/modified things for specific applications. I can't buy a generic disc for the inside lip of 24/40 adapters with a hole in the center for a drip tube.

$89 is a lot of money to buy a generic frit in order experimentally drill it ... when I don't even know what drill will work reliably.

Chemtex didn't offer to make a disc for me, So I assume that his earning a living prefers not to make frit disks. eg: He would have to charge a lot because it's a "pain" to do, and there are other unspoken snags and price raisers like bulk purchasing prices.

I kind of figured that was the case when I started the thread; That's why I'd like to ask for advice, again... Let me re-iterate my capabilities which are probably sufficient, if the right professional would guide me.

I have several computer based temperature controllers and kilns. Most amaateurs don't ... but they aren't hard to make. I can make them.
One example of my kilns is the alumina cement lined toaster overn in the background of the distillation apparatus photos. (left side, black toaster oven)

https://www.sciencemadness.org/whisper/viewthread.php?tid=79...

That kiln tops out at 500C, but (@chemitix) I can (and have) made kilns for less than $25 that go to 900C at programmable ramp rates and times. I don't mind building a kiln to try a sintering process, *IFF* I am given good instructions / decent estimates of temperature profiles vs. time. eg: just close is fine. But I also don't want to spend a lot of time making something for a process that has little or no hope of succeeding or being repeated by other amateurs. That's why I want an experts advice. I have fired clay ceramics and am familiar with quartz inversion points, pre-heating, soaking, tempering processes, and glazes. So, I will be a very quick learner of anything you have to offer.

I do have alumina hydrate, sodium silicate, boric acid, and borax, with which to make borosilicate glass from scratch; These are all chemicals that amaters can easily get or make from houshold/hardware store chemicals. If I had some guidance, I would willingly spend the time to refine a practical, non-proprietary borosilicate glass formula and share the results.

I think generic borosilicate 3.3 glass AKA: *non-alkaline* borosilicate glass is about
81.1% SiO2
12.5% B2O3
4.2% Na2O
2.2% Al2O3

I'm pretty sure the correct oxides will form if I just maintain the same ratios of Silicon, Boron, Sodium, and Aluminum in precurser chemicals and ignore the hydroxides and oxides. I think that making from scratch would be better than buying random borosilicate glass that might turn out to be 36 or 41 expansion rates, etc.

Whatever process I come up with, at least it would be documented and repeatable by other amateurs. eg: with less russian roulette effects due to salesmen mistakes and brand loyalty or accidentally buying the wrong glass to make frit from.

Do you have any expertise with the glass making itself, know someone who does, or are you forced to buy glass frit in bulk from reliable suppliers at non-amateur level prices?

[Edited on 25-4-2018 by semiconductive]

semiconductive - 25-4-2018 at 08:52

Quote:

I apologise if my use of the word 'dubious' offended you when I wrote
"I'm dubious of the functionality of the device"
I used the word meaning 'doubtful' - of the device - not your integrity.

Thank you, I appreciate that.
Be aware that others have, and will, see such remarks as attacks on my credibility.

Quote:

You have proposed a completely novel setup operating on principles not completely/clearly understood by me,
and I suspect that your frit will 'choke'
so what is wrong in warning you of problems before you encounter them ?

Quote:

that is just your ego causing you trouble.

That's not tactful; you clearly did not read my posts carefully and are shifting blame.
You primary (repeated) criticism was clearly not about the frit, but about air-flow at the top of a condenser where no frit exists.

What is wrong with warning me, is that you had no real evidence; you summarized what I said wrongly, and other people will assume you are correct because most people are lazy and don't read carefully esp. in a long thread. So if I don't respond, your slap at my ego will go unchallenged and rewrite history in most people's minds. Smart people are easy to attack on sciencemadness forums. You have (intentionally or not) also set me up to be attacked by other posters. The EXACT same thing happened in the first thread I made.

I'm an electrical engineer, well above average but didn't graduate cum laude -- I have designed things like electric kilns, hydroponic controls, and small supercomputer clusters for making satellite boards. There are other engineers who make bigger supercomputers than me -- I'm not the best, just because I have one skill. But, people obiously like to attack and discredit any intelligent person who gives background information. The more intelligent a person is, the more others assume they are an egotistical bastard.
I'm an easy hit.

The supermoderator, Burt, has a nice checklist of things to do before criticizing someone's idea. It's at the end of each of his posts as a signature right now. It's written by another smart person. If you still don't "Get" it, search for my first thread ever, on scimadness. See how one person saying "dubious" or "doubtful" totally derailed a thread and got my question condemned. It cost me a lot of wasted time and frustration. I spent a lot of work, and one comment from a suspicious person tore it all down.

I have been told by other, kind, members that there is a troll on scimadness who posts theoretical stuff that isn't true; Many members are hypersensitive to anything that looks suspicious ... therefore when anyone says my idea is "dubious","doubtful", etc. they automatically encourage other members to assume the "troll" is me. I've been there, done that, got the T shirt, and don't want to see it again.

If the reason I want to make a frit filter helps you make positive suggestions about how I might make the filter not "choke" that would be good ... but I don't see how your warning is helpful in any way in 1) encouraging or teaching me (and others) how to make frit filters. 2) in making me more likely to succeed in what I'm doing.

Even if the frit filter doesn't work well in one particular application ... having a thread which teaches how to make a generic frit filter would be useful to other amateurs. So, your warning only has the effect of discouraging me from even trying to do something that will help others, as well. I generally try to learn skills that are useful in many ways, and not just in one experiment.

[Edited on 25-4-2018 by semiconductive]

[Edited on 25-4-2018 by semiconductive]

Chemetix - 25-4-2018 at 16:33

I can make a frit for you, I just thought you were involved in the challenge of trying to make your own, and why not sometimes, but it's a complex task and I didn't want you chasing a goal that will sink more time and money than it's worth. Just for interests sake I've supplied a pic of a typical frit that was from a reputable supplier, cleaned and prepped well, tooled nicely, moon ascendant in Sagittarius and still failed.

Give me a drawing with dimensions and your ideas about the design and I'll give you a quote.

I understand a bit about your design, and it seems like a mist trap to me, if I'm wrong please correct me but the thread is a little messy to follow.

semiconductive - 26-4-2018 at 13:53

Quote: Originally posted by highpower48

Doesn't Deschem offer to make custom glassware. Not sure but that might be easiest and cheapest way to go. If you don't mind Chinese glassware.

Your idea appears to be spot on, thanks.

Deschem just replied to my message on e-bay, they say they can make the whole adapter for $15. So, now I'll try ordering and see if they are actually able to do it.

[Edited on 27-4-2018 by semiconductive]

semiconductive - 26-4-2018 at 15:04

Quote: Originally posted by Chemetix

I can make a frit for you, I just thought you were involved in the challenge of trying to make your own, and why not sometimes, but it's a complex task and I didn't want you chasing a goal that will sink more time and money than it's worth.

I do want the knowledge; but it's for a goal, and not just for knowledge's sake.
There are some glassware I'm not able to buy, and there are no general purpose glassware that can be used to make it easily. I'm not familiar with costs of custom glassware, and any stock items in the U.S. market that are even slightly unusual seem to sell for astronomical prices.

For example; No seller advertises an adapter for European standard to American joints, they just adapt "sizes". I bought a European 24/40 to 19/23 adapters and the 19/23 joint went way too far into an american Pyrex flask. Pyrex measures the fitting diameter halfway down the fitting, but European glassmakes measure fittings as the top.

In theory, I can just "cut" off the extra length on a European 19/xx adapter that penetrates too deeply ... but in reality, scoring and snaping ended up breaking the joint in places not scored.

I made a ceramic mold from a pyrex 19/23 plug; and if I had frit, I could fix the joint. But, I'm just shy of the knowledge needed to repair it. It's that kind of thing that keeps happening as I'm an amateur, that I'd like to be able to repair.

Quote:

Just for interests sake I've supplied a pic of a typical frit that was from a reputable supplier, cleaned and prepped well, tooled nicely, moon ascendant in Sagittarius and still failed.

Give me a drawing with dimensions and your ideas about the design and I'll give you a quote.

I understand a bit about your design, and it seems like a mist trap to me, if I'm wrong please correct me but the thread is a little messy to follow.

More or less, yes; The two frits are to stop mist and fumes. The inlet frit will probably also condense a minor amount of gasses that are hotter than the frit; eg: gasses from the heated condenser are not homogeneous in temperature or composition.

There ought to be no mist inside the collection flask (after the condenser) because the top frit blocks the condenser mists. I expect that only fumes will exist in the collection flask. The lower frit that protects the vacuum take off, then, should only be exposed to low concentrations of strong acid fumes.

That's why I only plan to put a hydrophobic coating on the lower frit. Fumes of sulfuric acid do not like to go through silicone membranes, but air does

I've got to go out, tonight, so I'll make some detailed drawings tomorrow and measure the dimensions of the adapters I have. I basically have three ideas; but they are just variations on the same theme.

If you're able to make non-standard glassware at reasonable prices, there are possibilities that I'd like to ask about. I've been focusing on what can be done by making minimal changes to standard off the shelf items; but as I learn, I've begun to realize that standard stuff has some serious drawbacks.

[Edited on 27-4-2018 by semiconductive]

semiconductive - 26-4-2018 at 23:19

Quote: Originally posted by Dr.Bob

I have some various 24/40 m/f jointed tubes with a frit in the middle. Not perfect, but might work. I'll look for anything like what you posted, but I might have something close.

@Dr.Bob
I didn't mean to ignore your post, I just missed it. I haven't seen Eurpean/Chineese versions of a fritted straight joint for sale, and I have been looking for them. I've only found mist blockers with straight glass tubing.

Are the joints you have European standard (24mm measured at top of joint, not middle)?
And, if so, how much are you willing to part with them for.... ?

semiconductive - 28-4-2018 at 18:42

Quote: Originally posted by Chemetix

Give me a drawing with dimensions and your ideas about the design and I'll give you a quote.

After measuring the adapters, I noticed a problem. The places where the frits go are after a narrowing in the glass. I'm not sure how I can get the frits into an existing adapter. I think the whole adapter may need to be made as a unit.

But, I'll post the design ... anyway ... in case you have any interest in bidding a complete part. If Dr. Bob's part would work, I may be able to get away with an adapter that just has the lower frit.

To explain how I got the frit thicknesses:

Here's a chart I made from the ROBU(Tm) data I was given earlier in the thread.
It's a curve fit of liquid flow data and nominal frit thicknesses. The ROBU(Tm) charts do have one point that I find suspect; eg: they show water flow rates are about 10x higher than air flow rates for the same pressure. That doesn't seem realistic. But I don't have another source of data to compare against; so I'm using them anyway unless you have better data.

The gnuplot script to generate the plot is appended to the end of the post. It's just a text file. The script can be loaded interactively from gnuplot, ' load "frit.gplt" ', and then the curve fit formulas become available for interactive use. The more important formula is the flow rate calculator for water vs. pressure.
eg, you can interactively type:

print flowW( 16. , 3. , 50. ) - flowW( 8. , 3. , 50.) , flowW( 19.2, 6.8, 50 )
29.5 29.6

And gnuplot will print the flow rates for a 16mm frit, 3mm thick, with 8 mm removed from the center and a 19.2mm diameter frit, 6.8mm thick, also gives a flow rate of 29.6 at 50 millibars. Both of them have approximately a 29 mL/minute flow rate at 50 millibars each. (100 millibars total, which is a typical fishtank airpump pressure. )

If you want to know an approximate frit thickness compatible with ROBU(Tm) discs, just type "print thick( foo_diamter_in_mm )" and gnuplot will give one that's close.

I drew out a complete adapter, but not the diameters of the frits; because those might change depending on what you have on hand to manufacture an adapter with.
I calculated their diameters and thickness from a typical adapter design.
The lower frit is a 16mm washer shape with 8mm hole, 3mm thick. The upper frit is 19.2mm disk, 6.8mm thick.

If you need to make different nominal diameter frits, or drain tube, I can recompute the frit thicknesses needed to get 1/2mL per second flow rate (30mL/minute). Let me know; or you can do it by guessing with the gnuplot script I attached.

The Gnuplot script to generate the plot, and to allow calculation is below:

Attachment: frit.gplt (2kB)
This file has been downloaded 648 times

officescape - 29-4-2018 at 19:06

One trick which makes these glass/ceramic melting projects easier is to use an oxy-hydrogen flame as well as feldspar as a flux.
With these two things you can do a crazy amount of exotic welding and glasswork.
I think you could grind up some glass and use a tiny oxy-hydrogen torch (electrolysis based units, either ebay or homemade) to weld it while its sitting into a small mold.
You could then put it into position, heat the outside of the takeoff and poke dimples to hold it in place.

semiconductive - 2-5-2018 at 17:34

Quote: Originally posted by officescape

One trick which makes these glass/ceramic melting projects easier is to use an oxy-hydrogen flame as well as feldspar as a flux.
With these two things you can do a crazy amount of exotic welding and glasswork.
I think you could grind up some glass and use a tiny oxy-hydrogen torch (electrolysis based units, either ebay or homemade) to weld it while its sitting into a small mold.
You could then put it into position, heat the outside of the takeoff and poke dimples to hold it in place.

I imagine that's true for welding.
s there a specific advantage to oxy-hydrogen with glass welding?

I imagine a sodium feldspar would make a pretty good flux. My only concern is how repeatable a glass I could make with it if making glass from scratch. Each time I buy feldspar, I might get a different stoichiometry.

For simple glass-making in molds; I think an electric kiln using a silicon carbide furnace ignitor for $20, might be easier for me (Im an electrical engineer). Silicon carbide can get crazy hot -- not as hot as oxy-hrdyogen, but it's practical for kiln work.

My brother wants to make an oxy-hydrogen torch. But, I've actually pretty happy with oxy-mapp gas inside a four fire brick oven for simple jobs. It makes a yellow-white heat easily over 1100C. That's plenty to melt borosilicate, although I'm not sure if it will fuse the precursor materials.

I've worked out a formula that I think will make Borosilicate 3.3 glass from hardware store and pharmacy chemicals.

I'm shooting for 81.1% SiO2, 12.5% B2O3, 4.2% Na2O, and 2.2% Al2O3.
Having too much sodium is the biggest problem when choosing precursor materials.

So, this is my first experimental plan:

I'm going to weigh out .0135 moles of oxalic acid ~= 17.019 grams of hardware oxalic acid (dihydrate).
Mix that with diluted waterglass, which will strip off the sodium and precipitate a very fine silica gel as a dust; so I don't have to grind up glass -- and it's chemically pure once washed with only a trace amounts of sodium.

Then I'll mixed the washed silic acid preciptate with:
0.337 grams of hydrated alumina; ( 0.004315 moles Al[OH]3) (can be made with ammonia and aluminum foil).
2.775 grams of 20 Mule team borax. ( sodium borate decahydrate )
0.421 grams of boric acid (to keep the soda content of borax low.)

If I did the math correctly, this will provide the correct molar ratios of Na, Si, B, Al to make borosilicate glass. Hopefully the oxides and hydroxides will change, appropriately, during fusing.

Most of the contents are water soluble, and they can easily be homogenized in a erlenmeyer flask with stir rod. That makes a pretty straight forward mix that should oxidze and fuse into borosilicate glass; and most importantly, it should be repeatable.

Once I try it, I'll post pics and run some tests to see if the formula needs adjusting.

[Edited on 3-5-2018 by semiconductive]

semiconductive - 21-5-2018 at 13:21

Progress report on buying custom glassware:
Deschem has gone silent .... Chemtex hasn't bid ... so I'm not sure custom glassware is even possible. Here's a copy of my conversation with DESCHEM on ebay.
So, it looks like I'm still going to have to make it myself.

semiconductive - 21-5-2018 at 15:06

Progress report on making borosilicate: Partial success in making ultra-fine silica precursor.

I attempted to make a known number of moles of silic acid jelly that would dehydrate to microfine silica particles.
I thought I could do it by titrating sodium metasilicate with a known number of moles of oxalic acid to neutral Ph. That avoids dehydrating the jelly to weigh it, which converts a fair amount of it to silicon dioxide again (undesirable).

For my first attempt, I tried to make 0.1350 moles of silicic acid (eg: enough to make 10g of borosilicate glass) There is a typo in the formula for borosilicate glass I gave two posts back ... The formula shhould say 0.1350 moles, not .01350moles.

Oxalic acid (dihydrate) is 126.064 g/mol, :. 0.1350 is 17.019 grams of oxalic acid.

I weighed oxalic acid from Ace hardware (assumed dihydrate) added to 700ML RI H2O.
I bought Humco, Houshold Farm (TM), brand "sodium silicate" to titrate with.
I added the sodium silicate to boliing and rapidly stirred 700mL oxalic acid solution. In theory, nothing should happen until the solution gets near neutral Ph, at which point the fluid should rapidly produce a silica jelly that is very weak.

I used litmus paper to monitor the pH of the solution, and when that showed pH 7 ... I stopped adding sodium silicate. The solution boiled all-night with vigorous magnetic stirring. Unfortunately all the water evaporated except 200mL of water or so. (A mistake.)

I then attempted to vacuum filter the jelly that was left. It was very slow, so I tried adding vinegar (5%) to make the sodium oxalate more dilute and soluble. That worked, and I was able to vacuum filter the jelly which was very soft, smooth and of a pasty consistency.
I then rinsed the jelly with methanol, to drive out as much and oxalic acid salt as possible.

Unfortunately, something is wrong with the theory of the experiment. When I tried heating the sodium oxalate solution to dryness to drive off methanol and acetic acid -- in theory, 0.1350 moles of di-sodium oxalate should have been left in the flask. (18.09g @ 134g/mol (wikipedia));

I did not get anywhere near 18 grams of residue. A small portion of the sodium oxalate browned (1.9g) at 125C -- so probably a contaminant.. The rest appeared to be soldium oxalate because it did not dissolve in pure methanol.

But, the total salt residue was only 11grams. I'm not sure what's wrong.

I checked the literature on manufacture of sodium silicate ... I found sodium-metasilicate is seldom precision made in a fixed sodium:silicon stoichiometry. Commercial preps can contain either excess or deficit of sodium hydoxide.

I took silica sand from the beach, and did a test on making sodium silicate using Lye. It is in fact easy to make, but the stoichiometry of silica to sodium does vary depending on how I cooked the material. eg: more or less sand will dissolve in the same amount of lye.

So, I'm still not sure on how to make a specific amount of pure silica gel.

However!

I heat dried a small sample of the jelly, and it reduced to a fine silica glass dust as expected. I would estimate 300 grit or finer dust was made. So, I'm quite happy with the quality of the silica glass made by this method. I just don't know how to make fixed quantities without lots of sampling, dehydrating, and measuring of every silicate I buy.

Note: According to several patents, ammonia silicate sol should be impossible to produce from silica gel. But I needed to wash my vacuum filter, so I tried ammonia anyway -- and to my delight, the jelly dissolves in 10% Ace hardware ammonia.

semiconductive - 29-5-2018 at 10:25

Update on making glass frit and still: End of May 2018.

1) Idea of hardening Vacuum Pump against chemical attack: It's finished, works great for vacuum filtration. Will show photos in next post.

2) Idea of using Glass Floss: The glass floss is here. It will take a few tries before I get it to fit the way I want.

3) Idea of making glass frit, and melting it:
The stainless steel 40 mesh screen has come, and I made a sieve out of it. Total cost ( $3 ).
I was going to use hardware store window screen for 20 mesh, and use that to sift smashed borosilicate glass ; but the harware store stuff was more expensive ($25). And the "patch" size was too small to make a sieve. So I decided to go ahead and order 20 mesh in stainless. Expect another two week delay.

4) Idea of buying custom glassware:
Deschem still hasn't replied to me on ebay ... so I'm beginning to doubt Deschem actually understands English well enough to be able to make custom glassware. I've tried writing other sellers in chineese using google translate; I used google translate, even checking the translation by translating it back into English ... so I know the translation is consistent; always gets chineese sellers to react negatively. So, either the translation is bad ... or it sounds "dumb" to them.
Does anyone out there know chineese, natively? Care to help?

I'm not going to write Deschem in Chineese myself; I'm pretty sure that avenue of glass making is dead without a Chineese American involved as translator.

5) Idea of making borosilicate glass from scratch:
Still workable, but requires lots of experimentation.
I discovered that although freshly precipitated silicic acid from waterglass does dissolve easily in ammonia water ... it will NOT dissolve in amonnia if allowed to dry out in air before redissolving in ammonia. I think this is the probable reason that many sites report that silicic acid can't be dissolved by weaker alakalines than sodium or potassium.... eg: Why cation exchange is supposedly "required" to make an ammonium sol. Once dried, the gel will not re-dissolve in ammonia.

Therefore: Caution, anyone trying to repeat my experiments Allowing waterglass neutralized by acid to dry in a fritted buchner funnel will ruin the funnel.
Nothing short of lye will get it out of the frit, once dried, and that will etch the frit.

Therefore, I bought a 1L porous plate vacuum buchner funnel. I can treat filter paper with silicone (diluted with mek) to make a durable filter that can be thrown away if I ruin it. ( Pictures, later. )

So, at this point ... the biggest problem in making borosilicate glass is getting the stoichometry of silica correct using water glass.
Anyone have suggestions? Right now, I think the best that can be done is precipitate the silica, dry it, weigh it ... then redissolve it in lye; That will give me an exact measurement of silica content in the liquid.

Then I can re-precipitate the sodium silicate; wash it to get rid of sodium... and then re-dissolve the wet silica gel in ammonia.
The process seems like a pain; but I can't think of another way to get a known amount of silica into a liquid solution without sodium present.

semiconductive - 11-6-2018 at 22:14

Aquarium pump chemical hardening (proof of concept):

Here's a photo essay about converting an aquarium pump to be both a mild vacuum pump and recirculator. It can be used for vacuum drying things, distilling things, etc. The pump works about as effectively on a buchner funnel as a water tap venturi. Aquarium pumps don't generate strong vacuum, but it's convenient. On top of not wasting water, I'm able to turn it off and on using a computer.

My first idea was simply to coat a pump with a mix and oil or MEK (methyl ethyl ketone) with silicone. Letting the pump suck up and spit out the mix will protect any metal or rubber parts inside the pump against harsher chemical fumes. Silicone won't withstand pure sulfuric acid, but it can handle even 90% intermittently. This method does work, as long as the silicone is very thin; but NOT if there is a check valve installed. I originally drilled a hole on top of the pump where the check valve was, but then changed my mind and drilled out the side of the pump and bypassed the check valve.

Aliaphatic oils, like lighter fluid or silicone waterproofing fluid (purchased in an aerosol can from the store) will thin silicone just as effectively as MEK and not hurt a check valve; However, the evaporation rate is very slow. This results in the silicone hardening before the oil is released. This is undesirable because it means that the oil is reducing the silicone's integrity. So, I used MEK and got rid of the check valve.

The pump is a whisper 10, 10 gallong aquarium pump.
All aquarium pumps use 3/16 inch ID tubing. Typically, Vinyl.

Pure GE silicone caulk or fishtank silicone is acetoxy cure. Acetoxy silicones are formulated to cure in thick (1/4) slabs. This means they are designed to have 1 micron sized pores in them to allow air vapor to pass through, slowly. They cure to a hazy finish.

OTOH, cheap alkoxy cure silicones (Nanda, From China #704, #705 ) do not cure in thick layers. The clear silicone is optically clear. I believe this means they are not porous on the scale of 200nm or more. Therefore, they ought to be nearly air tight.

I used #704 (white) so that it would show up in the photos of the pump conversion. It's also the stronger silicone. I hand painted the rubber diaphragm, inside and out, with three thin coats of silicone. When the black rubber and metal screw was no longer visible inside the diapgram, I quit.

I'm not sure what causes the white color, and it may reduce chemical resistance. I may attempt to fumigate the pump with fluorine gas to make a protective coating on the surface of the silicone, eg: if I can figure out how to generate fluorine gas. Likewise, the flapper valves in the white pump body are just a piece of flat plastic and could probably be replaced with teflon tape for better chemical resistance.

For this first attempt, I simply glued in clear vinyl tubing (fishtank air hose) with marine epoxy. I then potted the vinyl pipe with silicone (thinned with mek) ... in several 1mm layers.

I will be replacing the vinyl leader hose with 5mm glass pipe, later. The reason is that plastic connectors I get from the fishtank supply house can't handle MEK or strong solvents, and vinyl tubing, softened by MEK, collapses under vacuum.

I plan to change to glass connectors and use platinum cured silicone vacuum hose. Silicone hose with 3/16 ID and 3/8 OD is not expensive and can handle vacuum even after MEK.

Enjoy the photos:

[Edited on 12-6-2018 by semiconductive]