Sciencemadness Discussion Board

adapting a domestic MW

contrived - 28-9-2007 at 22:18

We want to try using a domestic MW oven to run some reax. The plan is to cut a hole in the top large enough to handle a condenser and hoses. We plan to set the machine up after the alteration and test for stray emissions. If the space arounbd the machine is clean enough we'll try some simple extractions according to the information in several good books we found in the library. Then we plan on doing some sort of Hanschz synthesis involving glycerol in oil to try and make the glycerols quantifiable with vis-spec by turning them into heterocycles. This is another item we found in one of the MY-chem texts. (I can cite this if anyone wants to know) We're hoping someone here has tried this alteration and use of domestic equipment before. We couldn't find it in the threads but that doesn't mean its not here.

Sauron - 28-9-2007 at 22:25

There is an excellent article in the Recent Journal Articles of Interest section, on the best choice of MW for this purpose and the modifications to be made to it. Use the search engine. Why reinvent the wheel?

DerAlte - 29-9-2007 at 09:55

It is foolhardy to cut a hole of any size in a microwave oven. If the hole perimeter approaches a wavelength (about 4.8 inches 12 cm) the aperture will radiate almost the entire available power. This is several hundred watts.

A circular hole of roughly 3.8cm (1.5 inch) or a square hole of 3 cm (1.2 inches) will be sufficient. If filled partially with dielectric or conducting material a lesser size will do the same. An off-resonant iris will still radiate substantial amounts of power. A thin slit will arc over.


Der ALte

Bromine - 4-10-2007 at 10:36

What if you put whole setup and magnetron in a bigger faradey cage, would it be safe?

chemrox - 4-10-2007 at 12:31

@Bromine I'd like to know too- enclose with hardware cloth and then ground the whole thing?

@DerAlte- could one expand that hole to any diameter short of a wavelength? Since this is my lab he's building it in I am a bit concerned. You said 12 cm and all we need is a hole 5-6 cm diam. We have library books coming out our ears and it looks do-able. We've got a multimeter with a radiation attachment to look for stray MW. We want to run a condenser out the top and we could use short extension to a condenser and reduce the diam to 4 cm if that's important. Please respond with any thoughts or info before he starts cutting. I saw a box of hole saws when I came in this morning.

@Sauron - I'll search again, I would really like to catch that thread and refs.

Twospoons - 4-10-2007 at 14:15

Any size hole will radiate! You need to get down to roughly 1/20th wavelength for the radiation to be siginificantly attenuated.
What is 'hardware cloth'? You should use a solid sheet metal enclosure for your Faraday cage - all seams should be soldered, so it will have to be made of brass or copper sheet. Ideally any holes should be less than 1cm diameter, with a microwave absorber sleeve to catch any strays (could be a ferrite or graphite block )

UnintentionalChaos - 4-10-2007 at 15:29

"Hardware cloth" is wire mesh. Solid is really unnecessary to stop microwaves...hence "cage" not "box" The "screen across the window of the microwave should give you a good size to shoot for.

DerAlte - 4-10-2007 at 16:15


As indicated in the prior two posts, all holes do radiate but it they are small enough (as in the metal screens on the doors) the effect is small.

The efficiency of a magnetron at 2.45 GHz is at least 50%, so an oven consuming 1Kw is capable of putting out 500 W of RF energy. Larger micowave ovens may have as much as 1 Kw available RF power. A crude computation suggests that at 1M (3ft) such an over will give an averaged power density of 1000/6 =166 w/m^2 from a resonant hole roughly 12 cm in perimeter. This is about three times the limit for organ damage (esp. eyes and brain) for an exposure of 5 minutes. And you will feel the heat on a hand at this distance. Much closer will fry it.

I do not have any data for the exact size of hole - it depends on the shape and the thickness of material. A slit would be typically about 0.45 wavelengths long to resonate ( it is the electromagnet compliment of a dipole): 5.5 cm or 2.2 inches long. Any hole with a perimeter above 0.1 wavelength (1.25 cms.} will radiate substantially.

The recommendended limits are around 1/5 to 1/10 of that quoted above.

Do a lot of research before cutting a hole and ten times as much before switching the oven on.


Der Alte (old radar engineer)

[Edited on 4-10-2007 by DerAlte]

leu - 4-10-2007 at 18:07

You will need to attach a sufficient length of copper tubing extending out of the top for the condenser to keep the microwave radiation from escaping :cool:

len1 - 4-10-2007 at 19:12

I have a formula giving total power radiated from a short slot as

(4 Pi / 3) (L/lambda)^2 V^2/Z

where L is the slot length (< wavelength), V is the maximum microwave potential, and Z resistance of space i.e. 377ohms.

Now to approximately relate maximum V in the microwave to its total power use V^2/Z = P, so the power radiated by a thin short slot cut in a microwave of total power P is

W = P 4 (L/lambda)^2

So having L tending to half wavelength (at 2.4GHz this is 12cm) will have all power radiated (here the approx breaks down). Having L = 3cm will radiate a quarter of the power, which is still too much. 2cm will radiate approx 10%.

Attaching a conducting waveguide to the slot, such that it can not support any mode propagation (width << 6cm) of several wavelengths length, will ensure no power is lost for slots in the range 2 - 6 cm.

Im surprised by an efficiency of only 50% for the magnetrons. This would mean that for an 800W oven, 300-400W will be lost in the magnetron, something its heatsink doesnt seem capable of dissipating. I would judge the magnetron in todays MW disspates 100W maximum and so is 80-90% efficient.

chemrox - 4-10-2007 at 20:14

Originally posted by leu
You will need to attach a sufficient length of copper tubing extending out of the top for the condenser to keep the microwave radiation from escaping :cool:

Is that a 'waveguide' so to speak? I will insist we keep doing reserach before cutting any holes. My friend is as impetuous as I am verbose.. que lastima.

DerAlte - 4-10-2007 at 20:56


Your radiation formula may give a realistic approximation. I assume it is derived from standard dipole theory.

With regard to efficiency, (IIRC from a long stretched memory from my M.Sc. course) a very simplistic formula for maximum theoretic efficiency is based on the kinetic energy available from the electrons in their spiral path to the anode of a magnetron. The higher the accelerating voltage the better the efficiency. The theoretical efficiency on this basis is indeed 80+% for high anode voltages. ( But commercial microwave ovens have to use reasonablly low voltages of the order of 5Kv)

However, you cannot slow the electrons to zero velocity and the practical efficiency is around 65% for a practical resonator structure.. If you couple that with the power supply efficiency and add the auxiliaries, this will probably be near my 50% estimate. I also doubt that any real tricks have been pulled to increase the efficiency of magnetrons since I was last associated with their usage in the 1960's when we assumed 65% as a working value. Magnetrons were fairly mature at the end of WWII.

Use of cutoff waveguide or choke joints implies a knowledge probably not avaiable to the average amateur. So does absorbing material; I see fires resulting through ignorance. Amateurs who don't know what they are doing should not play with 500 W of RF or microwave power, or do so at their own risk. Apart from electrical considerations, the ever present danger of superheating ionic liquids remains, especially where inflammable solvents and/or productsare in use. And for non-ionic solvents, heating will be minimal except at the glass. My strong advice is, research heavily before you attempt any like this. Use a traditional heating mantle!

Der Alte.

Sauron - 4-10-2007 at 21:20

Did anyone find the article I referred to in one of the earliest posts?

A lot of this talk is answered in that file IIRC.

I am posting it here and then will go re-read it and comment.

The authors state that leakage testing indicated full compliance with legislated standards.

< 1 mW/cm2 for new unit

<5 mW/cm2 in use.

This is a 1992 article and the Sharp model employed is likely out of the market but an equivalent Sharp model may be available.

The model selected was a 1200 W dual magnetron inverter-controlled design.

[Edited on 5-10-2007 by Sauron]

Attachment: MWmods.pdf (427kB)
This file has been downloaded 1084 times

Nixie - 14-11-2007 at 14:00

I'm curious, what's the point of using a microwave for this? Microwave efficiency isn't that much better than using a say heating element.

FrankRizzo - 14-11-2007 at 14:49

It's due to the nature of the microwaves interacting with the feedstock. Some reactions proceed much more efficiently with MW energy.

Nixie - 14-11-2007 at 15:39

Originally posted by FrankRizzo
It's due to the nature of the microwaves interacting with the feedstock. Some reactions proceed much more efficiently with MW energy.

I was under the impression that food in the microwave, for example, has no other effect upon it than would be gotten by simple heating. Isn't the assumed safety of microwaves in terms of food processing predicated on that?

[Edited on 14-11-2007 by Nixie]

FrankRizzo - 14-11-2007 at 16:16

You're absolutely right in that it's still a basic heating operation, but unlike oil baths and hotplates the heat is conducted to almost the entire feedstock at the same time. This uniform heating leads to much higher yields and purer end products due to reduced side -reactions at the heating interface.

not_important - 14-11-2007 at 20:08

In many liquids microwave heating can cause localised superheating, above the temperatures that would be reached by external heating.

In some cases microwave heating changes the product(s) or greatly shifts the ratios of products. While there may be newer research on this, I remember that in some cases the researchers had no conclusion as to why this change in products occurred.

Nixie - 14-11-2007 at 21:59

FrankRizzo and not_important give opposing accounts of the case, the first one saying even heating, the second one localized...

not_important - 14-11-2007 at 22:31


Abstract: The purpose of the study is to investigate the microwave superheating phenomena and their impact in homogeneous chemical reaction by using microwave heating. The specific mode of microwave heating is also used to improve chemical process by the selective and volumetric thermal effect of microwaves on the heterogeneous catalytic reactions.

[Edited on 15-11-2007 by not_important]

LSD25 - 2-2-2008 at 07:15

Look at the other topic (Microwave plasma generator) and see what they used -

(1) They enclosed everything which passed through the skin of the microwave with a faraday cage (fine stainless mesh)

(2) They used a lot of PTFE (& Similar) as well as brass etc. to support the reagents, to collect the gaseous products, etc.

Also examine this patent (MW Batch Reactor):

The choice of substrates upon which to demonstrate the applicability of the procedure may offer some hope to some here.

I am especially interested in this novel use of a cold-finger apparatus - may be another use for the vycor tube from a tungsten-halide lamp, eh? How the fuck does one source the heat sensor's for MW Applications?

chemrox - 2-2-2008 at 21:17

@Sauron- thanks for re-posting the modifications article. When we get caught up we're going to write to some of the labs working with MW and see if anyone's made the kind of mods we want to try. We want a port to allow a reflux condenser rather than trying to ventilate the vapors awayfrom the internal systems.

@not_important- thanks for the refs .. some things to remember and ponder

[Edited on 2-2-2008 by chemrox]

franklyn - 3-6-2008 at 08:51


ShadowWarrior4444 - 3-6-2008 at 14:09

A microwave reaction setup akin to what you desire to make: (with a picture!)

Another article, with a more useful picture:

I personally recall that a small hole (24/40) drilled in the top of the microwave will not radiate appreciably. If you are extra-ordinarily concerned about microwave radiaion, you can install a tantalum grid between the reaction vessel and the outflow pipe. Surrounding the outflow pipe with a copper grid will also help.

I should note the obvious though, that if you have a condenser directly above the hole, the water will absorb any stray microwaves.

jarynth - 4-9-2008 at 15:31

Losses will be inevitable with any hole large enough to allow the insertion of tubing of useful capacity. You don't want to enclose the water condenser in the 'cage' together with the reactor and feedstock. If you can't reflect all RF escaping te hole back into the original MW oven chamber (eg by interposing a metal grid transversal to the reflux connector to close the 'cage'), that energy can be written off and the next thing to be concerned about is to protect yourself from the RF by building a 'cage' around yourself! A large box with fine mesh walls will let you work safely, observe the experiment and also control it electronically through wiring etc. Just be sure to work in an environment where your experiment won't disrupt wireless appliances too noticeably.

[Edited on 4-9-2008 by jarynth]

Not to mention how useful this would also prove in other HV experiments.

[Edited on 4-9-2008 by jarynth]

chief - 9-10-2008 at 07:55

Here a link upon reaching 1300 [Cels] within a standard-microwave-oven,within 5 minutes, using "activated coal" as susceptor :

Maybe charcoal would work too ?

Nixie - 9-10-2008 at 10:38

Why does it need to be activated?

chief - 9-10-2008 at 11:27

I don't know if "activated" does better; I now tried a cup full of charcoal for 2 minutes: It heated up (but did not glow, neither catched fire) and now is 2 half-cups; but it also had some solid NaCl-melt on the bottom, from a previous run, so maybe the cup was weak already ...

FrankRizzo - 9-10-2008 at 13:41

I *believe* that graphite will work. Maybe paint it on to the inner surface of the crucible with sodium silicate solution?

Nixie - 9-10-2008 at 15:01

I remember crushing up a silicon carbide grinding wheel and that worked best (bonded with clay). Second place was a ferrite yoke from an old TV. Serious insulation is critical to get something really hot because the losses to the environment are huge when you get a decent temperature going (and don't forget to block the fan grille on the inside of the oven cavity).

[Edited on 9-10-2008 by Nixie]

FrankRizzo - 9-10-2008 at 16:13

Silicon carbide can also be found in an already powdered form as well. Any online lapidary supply will have it in various particle sizes for use as a grinding grit.

Carbon would seem to be the better material as it will withstand much higher temperatures.