Sciencemadness Discussion Board

quartz tube, how hot can I go?

woelen - 4-11-2007 at 12:33

I just won a quartz tube on eBay (70 cm, inner diameter 12 mm) for just EUR 6 :). I intend to do experiments with hot reactions, leading gases over powdered metals and that kind of things. The excess length then can be used for condensing volatile products.

This, however, all is quite new to me, and I am just wondering, how hot such a tube may get. Do I need to take special precautions in order to prevent cracking of the tube? This is a one-chance lucky thing for me, and I don't want to break the tube.

Any suggestions, ideas, and tips?

The_Davster - 4-11-2007 at 12:45

I have worked quartz with an oxyhydrogen flame. If you intend to work it be sure to use welding goggles, as the light given off is very very bright.

I have also used it in furnaces up to 1100 C or 1000C without problems, I think it runs into problems around 1500C though with softening occuring.
The stuff seems to have no problems with abuse either, I regularly took sealed tubes at 900C with a vaccuum inside and dropped them into room temp water with no cracking.

Antwain - 4-11-2007 at 12:48

I think that the glass tube in the furnace I used at uni to make YBCO was silica (It wasn't sodium glass and even borosilicate would have given out I think, which leaves silica). The furnace was rated, and thermostatted to 1200*C.

Fleaker - 4-11-2007 at 13:35

1200*C is about the limit for quartz in my opinion. Just don't use it with Group 1 or 2 oxides/hydroxides.

garage chemist - 5-11-2007 at 01:54

The company (Quarzschmelze Ilmenau, "Qsil") from which I got my quartz tubes states that maximum continous use temperature for quartz is 1100°C, and short-term (non-evacuated) up to 1300°C.
But you will greatly extend the lifetime of your tube if you do not exceed 1000°C, since crystallization of the quartz (devitrification) becomes noticeable above 1000°C and can quickly destroy the tube (it becomes opaque, porous and brittle and is not gastight anymore).
For this reason you should also always clean your tube thoroughly with alcohol before heating, and not touch the cleaned tube with bare hands since traces of sweat (contains alkali salts) act as crystallization seeds at high temperature. Fused quartz is a supercooled melt, after all- thermodynamically metastable, and prone to crystallization if given the chance.

There is also a whole range of substances (in addition to the mentioned group 1 and 2 compounds) that destroy quartz at high temperatures, e.g. PbSO4, some metal oxides, some fluorine compounds etc... Read up on handling of quartz labware.


Woelen, do you have a tube furnace? I am in the process of building a new large tube furnace for up to 1300°C, but the project had to be put on a halt due to not being able to find a suitable ceramic tube.
(My old Ytong tube furnace, documented on versuchschemie, unfortunately broke- Ytong is unsuitable for high temperature use, sadly- it insulated so well).
On the other side, I have now found a supplier for 1400°C-rated ceramic fiber blankets, which I will use as insulation. The only thing I really need is the ceramic tube, I already have the Kanthal A1 wire and the thyristor circuit for power regulation.
The finished furnace will be built completely without any types of cement or castable.
There is a section on tube furnace construction in Brauer, and I will follow those suggestions- wind the wire on the ceramic tube, coat with paste of MgO and Al2O3, dry, and install it in a casing with the insulation. Brauer suggests loosely packed MgO, I will use the ceramic fiber blankets.

I have completed a small (22cm length, accepts 25mm tubes) low-tech tube furnace that uses 0,5mm Nichrome wire for heating (max. wire temperature for nichrome is 1200°C, which translates to 1000°C maximum furnace temperature for long life of the wire) using an old Acros acetaldehyde can as the outer casing, and two different cement mixtures for structure and insulation (inner side, in contact with the wire: ciment fondu + chamotte grog dense refractory, insulation: same formula, but with 3 parts perlite added), and no ceramic tube.
It was quite simple to build, I'll take a few pictures soon. The principle is not applicable to larger furnaces though. It will be used mostly for ketene production and gas-solid reactions since I only have a pyrex tube for it that I cant take above 700°C.

I also have a large (1m length, 28mm OD, 1,5mm WT) quartz tube with two NS29 ground-glass joints at both sides which will be used with the big tube furnace for high-temperature projects like CS2 production.
I had this tube custom made- the tube and joints were ordered directly from Qsil, and I had the glassblower put them together.
It did cost about 100€ total, including the money for the glassblower- if I had the glassblower order the parts, it would have cost nearly two times as much since he orders from a more expensive supplier.

I will gladly help you with advice if you want to build a tube furnace- its a highly useful piece of equipment, and reasonably easy to construct yourself. And they can be enormously expensive if you wanted to buy a new commercial model.

YT2095 - 5-11-2007 at 02:01

12AX7 might be the guy to help out with heating, Induction heating sounds idealy suited to this app.

garage chemist - 5-11-2007 at 02:52

Inductively heated tube furnace? Sounds interesting, but I've never heard of such an idea before. What conduvtive material would be used to absorb the magnetic field, and could this material not be used for direct electric heating then?
I dont think metal powders sufficiently absorb magnetic fields to heat them directly, though it would be an interesting method for compact metal pieces.

Tube furnaces are usually heated with Nichrome wire for max. 1100°C, Kanthal A1 for max. 1300°C, platinum or Pt/Rh wire for 1500°C (Brauer mentions Pt and Pt/Rh wire for small furnaces with the wire coil inside the ceramic tube- this would be prohibitively expensive nowadays due to the high price of platinum) or Mo, W and Ta wire under protective atmosphere (the metals would oxidise otherwise) for 1500°C.

Non-metallic heating elements are silicon carbide rods (rather short lifespan) for 1500/1600°C, molybdenum disilicide rods (Super Kanthal heating elements) for 1600/1700°C and special doped ZrO2 or other doped conductive ceramics for up to 1800°C (such doped ceramic heating elements need preheating since their resistance is too high at temps below a few hundred °C).

And then there are various vacuum furnace technologies with W sheets or inductively heated doped ceramic crucibles, and the graphite tube furnace (Tammann furnace, heated by direct current passage through the graphite tube) for temperatures in excess of 2000°C.

[Edited on 5-11-2007 by garage chemist]

12AX7 - 5-11-2007 at 09:14

Lots of succeptors are workable. Graphite, molybdenum, tungsten, SiC, WC to name a few.

No electrical connections are necessary, making it ideally suited for contact-free heating, especially if the material you're heating has the unfortunate condition of not having contact (molten metal for instance). No vacuum seals to bother with. You can even do levitation heating.

Tim

Fleaker - 5-11-2007 at 09:49

The levitation part is particularly nice as it is zero contamination and can be carried out in high vacuum.

@Garage chemist, I suggest you look online for a pottery supply house that carries lightweight kiln brick. They are very easily sculpted with even a bread knife and insulate very well. Over them you can put a zirconia wash which will reflect much of the radiation, increasing your efficiency and making the interior much more durable. It is called ITC-100 and I have a source for it. You can also do the same thing with kaowool, which can be rolled up into a tube, dipped into a zirconia rigidizer, and ceramic pegs used to hold the heating elements. If you want to put metal tubes in there, I would also talk to a potter about having him make a kaolin (essentially a high alumina clay) tube a few millimeters thick for you. This can go over the heating elements and protect from damage and accidental shorts.

I have 1200*C continuous operating kaowool, but it is good for 1300C in the conditions you would be using. There are also 1500*C continuous chrome spinel based ones, but they are expensive about 10 USD/sq foot, if you can find it out of a case or pallet lot.

garage chemist - 5-11-2007 at 11:55

You suggest lightweight firebrick as thermal insulation for the tube?
I thought about that, and the shop I want to buy the ceramic wool from also carries lightweight firebrick, but it would be difficult to make a fitting shell for the 50cm long tube out of the bricks. Also, this would require a square metal housing for the tube furnace, while ceramic wool can simply be wound around the tube to a thickness of, say, 5cm or 7,5cm and then put into a metal oven pipe as casing.
So no need to build a special casing, only the ends will require some tinkering with sheet metal and cut firebrick to seal off the ceramic wool to prevent its slow disintegration.

The lightweight firebrick would come no bit cheaper at all than the ceramic wool, since many firebricks would be needed and there would be a lot of waste due to the sawing to size. Not to mention the work it would be to saw a halfpipe shape into all those firebricks.
No such problems with ceramic wool.
I have no source for a rigidizer though. Is it really needed when the wool is encased in a metal housing and not subject to mechanic movement? I would not use such a product unless really necessary due to better serviceability (wire replacement) of the furnace without it.


Is there something that specifically rules against ceramic wool as insulation for a tube furnace? Does it become brittle and crumbly with time and use, even the 1400°C rated wool?
I am aware of the health risks of ceramic fiber products and will wear a dust mask when working with ceramic wool in the open.

By the way, today afternoon I found a supplier that is willing to sell me two 32mm ID, 40mm OD and 500mm length pythagoras tubes! Work on the tube furnace can be resumed once they arrive. I also still need the Al2O3 and MgO for the wire embedding paste, and the ceramic wool.

Pythagoras appears to be a sort of technical hard porcelain, 60% Al2O3, 37% SiO2, 3% K2O. It is good for up to 1400°C according to the manufacturer, perfect for me, and mechanically more stable than sintered Al2O3.
The tubes are remarkably expensive (EUR 55,70 per tube), costing more than quartz tubes of similar size. Al2O3 tubes would be even more expensive though, and not better for this application since I dont need their even higher heat resistance.

What worries me is the low thermal conductivity of pythagoras: 2W/m*K, the lowest of all technical dense refractory ceramics.
I hope that this does not lead to temperature gradients greater than 100°C between outside and inside of the tube (4mm wall thickness) since the wire is good up to 1400°C and I want at least 1200°C inside the pipe. Can someone estimate how long a temperature gradient of 100°C across the 4mm wall would take to dissipate when no heat is removed on the inside of the tube?
I would be prudent to slow down further heating once 1000°C are reached in the tube to avoid the wire going above the maximum design temperature of 1400°C and failing.



[Edited on 5-11-2007 by garage chemist]

woelen - 5-11-2007 at 12:22

Thanks for all the responses. Currently I am not considering building a real tube furnace. I was thinking more in terms of heating part of the tube with a flame and putting some material at that place, while at the same time, leading a suitable gas over the material. One experiment I want to do is mixing powdered Nb and leading NbCl5 over it, in order to get blue NbCl4. Another experiment which sounds very interesting to me is making WCl5, using a similar setup. Such experiments require heating to several 100's of Celsius and the leading of gas (e.g. Cl2) over the material.

For this purpose, a thin tube of 12 mm inner diameter is perfect, it allows me to lead gas over the materials without excessive losses and it allows me to do experiments on a microscale.

It is good to know that all temperatures, which I have in mind are not a problem at all. You all are speaking of 1000 C or higher, I do not expect that at any time I will go that hot.

Maya - 5-11-2007 at 17:20

<<< It is good to know that all temperatures, which I have in mind are not a problem at all. You all are speaking of 1000 C or higher, I do not expect that at any time I will go that hot. >>>


Well, you did title this " ...How hot can I go"

not accurate to your purpose then.
A search will reveal these quartz tubes are very robust BTW. They can go straight from hot to cold with no damage.

Incidently. I have 16X 36" long by 25mm OD quartz tubes and a marshall tube furnace rated to 1200 deg which is maybe 32" long and 45 mm ID.

Any ideas on how to not get the ends of the tubes too hot? ( so as to pass inert gas thru it ).

Maybe copper coils with water circulating at the ends? Otherwise I'll have to get rid of this furnace and get a smaller one. didn't realize it was this big when I bought it...

12AX7 - 5-11-2007 at 18:59

Quote:
Originally posted by garage chemist
What worries me is the low thermal conductivity of pythagoras: 2W/m*K, the lowest of all technical dense refractory ceramics.
I hope that this does not lead to temperature gradients greater than 100°C between outside and inside of the tube (4mm wall thickness) since the wire is good up to 1400°C and I want at least 1200°C inside the pipe. Can someone estimate how long a temperature gradient of 100°C across the 4mm wall would take to dissipate when no heat is removed on the inside of the tube?


In a closed system, approximately infinity.

What limits are you proposing? Within a gradient of 1°C? 0.1? 1 time constant or 5? What is the matter inside the tube, is it exothermic, endothermic or neither, and what's its heat capacity? What's on the outside? Heat bath? Insulation? How good of either?

...In general, when you're heating things, just give it time. A metal-melting furnace might be at full throttle for half an hour, or a more efficient (or more heavily fueled..) one for 15 minutes or less. But that same furnace can be operated at perhaps 1/50th of that power level and still reach incandescent temperatures, if given an hour or two soak time to reach equilibrium.

Tim

Magpie - 5-11-2007 at 19:33

Maya, rather than getting a new furnace why not just get some new tubes. Or possibly, some special end fittings as shown here might work with your existing tubes.

http://www.sentrotech.com/tube.php

Magpie - 6-11-2007 at 12:43

Garage chemist I took a crack at calculating the time that would be needed to heat your tube, using the following assumptions:

k=2W/M-K
Cp = 0.3 cal/(g-C)
density = 1.8g/cm3
r1=16 mm
r2=20 mm
tube outer surface initially 100C above tube inner surface.

dq=mCpdT (eqn 1)
dq/dt=2piLk(T2-T1)/[ln(r2/r1)] (eqn 2)

After setting these two equations equal by eliminating dq, and integrating for time, t, and temperature T, the simplified result is:

t = 0.3(ln100-lnX) (in minutes)

where X = how close tube inner surface temperature approaches that of the outer surface, deg C.

For example, if X is chosen as 0.1 degree C, then t = 2.1 min.

Does this seem reasonable?

chemrox - 6-11-2007 at 16:17

Quote:
Originally posted by FleakerJust don't use it with Group 1 or 2 oxides/hydroxides.


What happens? Does that make the qtz into feldspar? (which would cause disintegration as well..)

Fleaker - 6-11-2007 at 18:08

It frosts and etches the quartz, weakening it. I'm too lazy to think of exactly what happens, but it does flux the SiO2. I just know I have seen it happen when someone else in the lab was making SrO from Sr(OH)2.


I'll U2U the source for the ITC-100, it is very expensive, but this fellow I know is the best price anywhere. However, it is very efficient with electric furnaces and offers improved heat up times. I also suggest you use a piece of ceramic wool to plug both ends of the furnace once it is heating up (drastically reduces heat up time).

The kiln brick is so soft that you can make a hemisphere in it in probably under a minute with an aggressive rasp for woodworking. There would be material loss, and you are correct in stating that it is equal in expense to the kaowool per unit of area, if not more so. A shell would be easy to build with basic sheet metal skills. If you really wanted, you could duct tape it together even!
I did suggest ceramic fiber, and no, a rigidizer is not entirely necessary, just preferred as it lengthens the life of your fiber product (yes it does become dusty once fired).

To avoid itching badly from it, spray it down with water, and do wear a mask!