Sciencemadness Discussion Board - Infrared Heat Source for Distillation - Decomposition: Yes or No? Why?

Infrared Heat Source for Distillation - Decomposition: Yes or No? Why?

Pyrotrons - 29-10-2013 at 19:37

Hi All,

I've come to the conclusion, that an infrared heat lamp is an exceptionally good way to heat large round-bottom flasks for distillation. However, there might be something that I'm missing, and that is, possible decomposition of SOME compounds due to high-energy IR. Also, there is probably a possibility that the substance does not absorb IR - hence minimal or no heating (but an easy and fix for this would be to paint the bottom of the flask black).

I searched Google and this site for 1.5 hours on this, and came up with next to nothing. This almost leads me to believe that heat lamps don't work, but for purposes of Ethanol distillation, they are absolutely (pun intended) fantastic in my lab.

My question is, how do I determine if a compound will decompose under conditions of strong IR light?? Ideally there will be a discussion of wavelength, power, absorption spectra, and examples of compounds that do and do not decompose. My heat lamp is 250 watts, 120VAC. The front glass of the lamp is a red cutoff filter - it's presence strips the light coming from the filament of any wavelengths higher than "deep red". It definitely would kill any UV...which as I understand it, is up there with the real wavelengths of interest (~400nm).

Thanks in advance.

---------------------------------------

My spiel on heat lamps, specifically pertaining to those types possessing a red-colored glass filter, to stop visible and ultraviolet light:

An infrared heat lamp:

- Is infinitely controllable via electronic controller or a simple light dimmer

- Comes up to full operating temp in less than a second due to very low thermal mass. There is no waiting for warm-up or cool-down. Compare this to a sand bath or heating mantle.

- Does not contain a big bath of high-energy hydrocarbon fuels (as in an oil bath) at elevated temperatures.

- Is VERY gentle on the glass. I have measured multiple times that the CONTENTS of my flasks are sometimes HOTTER THAN THE OUTSIDE OF THE GLASS. I have access to a small thermal imaging system, the glass is heated perfectly evenly as long as the IR lamp is not too close to the flask.

- Is seemingly quite efficient, directing most of the energy into the flask contents.

Oscilllator - 30-10-2013 at 02:00

All infrared radiation has a lower energy than visible radiation, since the frequency of infrared is lower than that of visible light. This is stated by the formula energy=frequency x planks constant. So if the compound is not decomposed by visible light, I don't see how it could be decomposed by something with a lower energy.

In general though, this seems like a fantastic idea! What wattage lamp do you use, and what temperatures can you get with it? I would love to see some photos of your setup!

[Edited on 30-10-2013 by Oscilllator]

deltaH - 30-10-2013 at 02:32

Hi Pyrotrons, lovely idea! At infrared wavelengths there is no chance of radical formation, simply thermal heating, so you only have a chance of any decomposition if you overheat and that is LESS likely with infrared as you heat from the outside in but over greater volume as you organic absorbs the light over some short distance.

That said, there is one thing to be aware of. If there is something dark in your flask and your power is very high, then the dark bit will absorb much more light, get hotter and form a hot spot that can decompose things to make more dark matter and so this runs away... but this is only a problem at very high light intensity. Obviously not an issue for distilling solvents and the like, but may be an issue for messy reaction mixtures where there may be suspensions and chances of other things forming of a hot spot forms on the inner surface of the flask, in such a case, I'd exercise caution with such a system.

Pyrotrons - 30-10-2013 at 03:55

Wow, thanks for your thoughts and comments.

Lamp Data:
http://www.ecat.lighting.philips.com/l/lamps/incandescent-la...

(Or just type in "Philips Red R40" on Google, comes up first hit)

Oscillator, I did a quick test for you (and me). Two of these lamps, shining at the center point inside a 1L round-bottom flask, with the front of the lamp glass spaced 5cm from the flask, heated 200mL of distilled water at a rate of 5 degrees C per minute. I only tested it from 41 to 51 degrees C.

Total energy input over the course of the test: 500W for two minutes.
500W for 2 minutes = 1000W for one minute
1000W for 60 seconds = 60 kilojoules (kJ)

Heat capacity of water: 4.186J will raise 1 gram (or 1mL) by 1 Deg. C
So, 4.186kJ should raise 1 Liter by 1 Deg. C
60kJ should raise 1L by 14.33 Deg. C

60kJ should also raise 200mL by 71.65 Degrees C.
In comparison, it only raised it by 10 degrees.

Efficiency: About 14%. Seems low, but doesn't it always.

Compare to the efficiency of a bunsen burner : )

Ok now I feel better.

deltaH: Very good point. I can definitely see that happening with thick mixtures as well.

EDIT: Added the following:

Oscillator, it's true that longer wavelength (lower frequency) light possesses less energy, however, what if there happens to be an absorption band in that particular compound that absorbs that (lower energy light) and decomposes it, because it's higher than some threshold? I'm just speculating, I don't even know if such a "threshold" even exists.

[Edited on 30-10-2013 by Pyrotrons]

Oscilllator - 31-10-2013 at 01:06

Pyrotrons it is definitely true that different compounds (or more precisely, different bonds) have absorption bands in different parts of the IR spectrum. This is the basis for infrared spectroscopy after all. It is also true that any given bond will have a particular energy. Perhaps if a photon of that particular energy hit the bond it might cause the bond to break, but I imagine that these bond energies would correspond to wavelengths of light well into the UV spectrum. In any case, my knowledge on this subject is limited but I am almost certain that you have nothing to fear from decomposition of your reactants.

With regards to your efficiency, I am sure it is possible to get it much higher than 14% with a minimum of effort. Am I correct in saying you just pointed the lamp at the round-bottom flask, without any kind of enclosure? If this is the case, then I am sure the efficiency could be improved by placing the lamp at the bottom of a shiny tube (I am thinking a large tin can) to direct the light at the water. Also, it quite probable that the transparency of the water played a big role as well. Perhaps adding some blue/purple dye to the water (KMnO4?) would increase the efficiency, although of course this might not be practical for real reactions.

Pyrotrons - 31-10-2013 at 02:35

Yes, I just pointed the lamps at the flask, with no reflector. The lamps however, have an internal reflector, a metallization that is very effective. Hardly any light or heat comes out of the sides. At 10% power, you can't hold your hand within 5cm of the front for more than a second or two.

I think you're right about the light passing straight through the flask and contents.

There is 870 Degree C paint (ceramic-based) here, available in black, that is used for barbeque grills, ovens, and the like. I'll spray a patch of that on the bottom of the flask, I bet it really cooks. I might have to start being careful with the glass (temperature gradients). If I do this, I'll put the thermal imager back on it and post the results.

chemrox - 31-10-2013 at 10:01

It may not be the best idea. It has the possible drawback of heating focused areas within the flask. I had a professor that advocated oil baths over heating mantles to improve yields for the same reason.

Pyrotrons - 1-11-2013 at 18:32

Hmm, good point.

Wet sanding the bottom of the flask with 1000 grit sandpaper should diffuse the light. Or using multiple smaller lamps Vs. one large one.

Not that I want to do either one of those.

jwpa17 - 2-11-2013 at 18:37

Pyrotrons -
Your IR heating idea is good, and will heat efficiently, particularly if you can couple strong IR absorbers to the flask - like the thermally resistant paint.
However, it doesn't necessarily prevent thermal decomposition. When you dump energy into a molecule, whether by collisions due to thermal energy (such as conventional heating) or by collisions due to absorbing IR radiation, it is distributed across or in the molecule in a particular fashion. You can get differences if you put that energy in in different fashions - like if the IR is absorbed by a particular bond. But that would be very difficult with a lamp such as you indicate. But that's one of the reasons you can get unusual selectivity with photochemistry.

Pyrotrons - 3-11-2013 at 01:58

Thanks jwpa17, really good science.

I ran a fractional distillation of 2-Propanol & H2O today just for the hell of it, utilizing a 300mm Vigreux. While rising gasses were first holding up, I noticed that I could stop the condensation front DEAD where I wanted it on the column with just the right touch of the light dimmer. Then I could move it up a couple of inches. Down an inch. Up 5. All in the course of about a minute...pretty damn good resolution. Should be able to get fast and excellent results with this and a micro-controller AND an RTD (Resistor Temperature-Detector)...with temp. resolution of 0.1 Deg C, at the still head. But I suppose any heating scheme could do that.

I'll hopefully try the black paint tomorrow. IIRC I've used it before (not on flasks) and it sticks to glass, but we'll find out. Anyhow I'll need to be careful of flask bottom temperatures, 500W of power isn't that much, but if it does somehow get accidentally focused... (as chemrox was saying, although I believe he was referring to a lens-effect, with the flask itself focusing power to a single point inside the reaction mixture).

+1000 Deg. C black paint for seven bucks:

https://www.rustoleum.com/CBGProduct.asp?pid=371

watson.fawkes - 3-11-2013 at 08:58

Quote: Originally posted by Pyrotrons

While rising gasses were first holding up, I noticed that I could stop the condensation front DEAD where I wanted it on the column with just the right touch of the light dimmer. Then I could move it up a couple of inches. Down an inch. Up 5. All in the course of about a minute...pretty damn good resolution.

The main reason for this good ability to control is the low thermal mass of the heating element. It will significantly lower response times. This seems to be the main advantage of this method. One way to use it is as controllable heat and not the entire heat source. Use any traditional heat source to bring the vessel up to near the temperature wanted but slightly lower, maybe 5 °C. This can be a high thermal mass heater such as a bath of some sort. Then use radiative heat as a controllable supplement. I imagine this would put the limits on control on the resolution of the temperature sensor.

Using black spray paint is tantamount to using an external susceptor and may blunt the efficacy of this method. The reason is that you are now heating the glass first and not the vessel contents, raising the thermal mass of the target, meaning slower response. On the other hand, using an internal susceptor might work better, basically an inert black rock. What that's made of would depend up on the reagents in the vessel. And since that would make hot spots on the susceptor, you'd need stirring. Perhaps combine both and make the stirring paddle into a susceptor.

Pyrotrons - 3-11-2013 at 20:41

The high thermal-mass bath at the bottom would be fairly tightly coupled to the flask contents. This would require the low thermal-mass IR light to raise the temperature of the flask contents AND the bath, resulting in a slow response time.

[Edited on 4-11-2013 by Pyrotrons]

watson.fawkes - 3-11-2013 at 21:42

Quote: Originally posted by Pyrotrons

The high thermal-mass bath at the bottom would be fairly tightly coupled to the flask contents.

Not particularly. Glass isn't exactly a great thermal conductor. It's used in the lab because of its chemical inertness, not because of its stellar thermal properties. And on top of that, there would be a heat flow outward from the flask to the bath, but not with a huge temperature gradient, so not a high heat transfer rate. This assumes that you can differentially couple thermal radiation to the flask to begin with, which seems to work.

Pyrotrons - 4-11-2013 at 00:43

Surface area can more than make up for lack of thermal conductivity. Placing a bath under such a large surface area of the thin glass of an RBF will surely increase the system thermal mass. I'm sticking with the single low-mass high-power IR source underneath.

And maybe some nice black paint, if I can find the time Aaarrgghhh...

Pyrotrons - 28-11-2013 at 19:57

Dammit, I just made a decent post about the results, then lost it to some dumbshit that designed the keyboard on this laptop.

I suspended an LM35 temperature sensor in the middle of a 500mL RBF containing ~200mL of 91% IPA. The LM35 is a little bug in TO-92 package that looks just like a transistor (like 2N2222, 2N4401, 3904, etc). Heating of the flask was accomplished by a single 250W lamp as described in the above posts. Heat control was by Arduino (with ridiculously simple code) and a makeshift solid-state relay I made from a big 'ol MOSFET laying around.

I was able to control the condensation front (as visible by vertically streaking lines of solution) with very, very, high accuracy. If I raised the sensor 1cm, the lamp would come on at full power...the vapor front would then reach it...then the system would calm down and regulate temperature as before. The vapor front chases the sensor around perfectly, and even the "swirlies" (where the vapors mix) have a visible effect on the system.

Also interesting, is that the lamp has such low thermal mass that tuning the controller seems to be a non-issue. At first pass (no tuning whatsoever) there was very minimal overshoot when coming up at full-power from 20C to 50C. I think the overshoot that I did see was due to zero stirring within the flask...so the hottest liquid was as the very bottom.

I can't wait to try this with a Pt-film RTD and a good fractionating column, which is in the mail.

** Any thoughts on implementing a control method for stepping through fractions with this would be greatly appreciated. My first idea is:

Full power until "X" degrees away from the first fraction's B.P. (sensed with a sensor that is lower in the still than at the head), then begin distilling continuously at "Y" drops per second. System then eventually recognizes that no more drops are condensing at that temp...so it raises the heating power to start the next fraction. There is a ton missing from this that I've not worked out. The limitation is that there is no human factor to carefully go at "X" drops per second through all the fractions.

How do the commercial microcontroller-controlled, spinning-band fractionating systems do it? Where are their temp sensors located within the system??

------------------------------
Edit: Added the following:

I set my 300mm Vigreux on the 500mL flask, and then snaked the LM35 temp sensor + wires halfway down the middle of it. I put the system on for full power up to 40 Deg. C and watched the condensation front quickly rise to the sensor... then stop dead. Now it's oscillating just a tiny bit around the sensor ; ) I LOVE IT!!!

I put the thermal imager on it, I don't even have words to describe the beauty.

[Edited on 29-11-2013 by Pyrotrons]

Sedit - 31-12-2013 at 21:56

Just a thought but what about a "basket" of IR LED to hold the flask almost like an Infrared heating mantel of sorts. Should minimize light lost. Over all I just think it would be a cool ass looking piece of equipment when completed.

cyanureeves - 1-1-2014 at 08:10

a drop of cold water might be bad news for the hot light bulb.

[Edited on 1-1-2014 by cyanureeves]

IrC - 1-1-2014 at 14:02

Unless your talking lasers, LED's are not going to do well heating a beaker full of solution. What does work extremely well are the lamps from old 35 mm film (slide) projectors (hacker sourced from garage sales, flea markets - you name it) readily available cheap since so many homes in the 50's - 70's had them. Just junk today. Picking out the thick glass filter you have another item useful for other experiments. Also everything up to the line cord to power a bulb is in the old projector including bulb socket. A decent quality blue and above blocking filter made of quartz or glass able to take high temperatures helps reduce problems with higher wavelengths initiating unwanted reactions. Or use the IR block (glass disc) from the projector if you want higher frequencies at reduced thermal energy for use in light activation chemistry.

All in all a cheap old projector is a goldmine, one families trash is a mad scientist treasure. Barring old salvage in your area one could get a socket and replacement bulb. Also useful is a window to protect the hot bulb from liquid contact by a boil over, or from a broken glass container.

You will need a fan for airflow around the bulb in use as well. These projector bulbs do a far better job for chemistry than any type of floodlight and likely save cash on the utility bills for a comparative number of hours of operation. An old slide projector:

This bulb is 500 watts:

http://www.ebay.com/itm/Sylvania-DFR-Projector-Lamp-/2213463...

This 300 watt bulb needs a lamp regulator (the exposure lamp regulator from an old Savin copier works):

http://www.ebay.com/itm/KODAK-FHS-PROJECTOR-PROJECTION-LAMP-...

Just a few thoughts on the subject for what its worth. This type of lamp would be great for test tube experiments where you don't want a wide beam of heat such as that produced by a flood type of heat lamp.

[Edited on 1-1-2014 by IrC]