Sciencemadness Discussion Board

Distillation tails at lower temperature. (Damn. I thought I fully understood distillation.)

j_sum1 - 2-9-2022 at 23:50

See thread title.

I have been distilling solvents recently to obtain some fractions that are clean and useful for me. Latest project has been on kerosene to obtain a hyrdocarbon mix that is mostly aliphitic and saturated. Desired fraction is 150-180°C

Procedure:
  • Distil store-bought kero and collect from 120°C up to approx 190°C
  • Make a solution of potassium permanganate. Shake this into the distillaterepeatedly and allow to react with unsaturated hydrocarbons until purple pigment is persist in the aqueous phase.
  • Decant and use sep funnel to isolate organic phase (slightly yellowed)
  • Add anhydrous MgSO4 and distil

    It is this second distillation that has posed some questions for me.
    The first fraction is a few mL from about 70°C to just below 100°C. Evidently this contains water. Presumably some azeotropes there.
    I collected a second fraction that began at 126°C up to 150°C
    I then switched receivers and collected a fraction up to 185°
    At this point I expected the temperature to keep rising. The temperature however dropped significantly and the final fraction, approximately a third of the total volume, was collected from 185° to 130°.

    This was not a one-off occurance. The same temperature decrease occured in another batch that was distilled a second time but was not treated with the permanganate.

    Like I said, I thought my understanding of distillatioon was ok, but I do not understand why this low boiling fraction is coming over after the high-boiling fraction.
    If it makes a difference, the distillation rate was fairly steady and quite rapid -- at approximately 0.25-0.5mL per second.

    Help me to understand.


    Sulaiman - 3-9-2022 at 00:10

    Guess : lower gas flow rate for higher bp fraction allowing thermometer to read low.
    Try measuring temperature at a physically lower point,
    eg just above the boiling liquid.
    EDIT : or in the top of the column if fractional distillation.

    [Edited on 3-9-2022 by Sulaiman]

    j_sum1 - 3-9-2022 at 01:18

    Yeah. That was my first thought. But with such a large last fraction, such a high distillation rate, and no appreciable drop in rate, that seems unlikely.

    Rainwater - 3-9-2022 at 04:10

    Might be a secondary reaction.
    You can confirm the boiling point of the third fraction by measuring the temperature of the boiling liquid directly.

    BromicAcid - 3-9-2022 at 04:24

    My guess would be you finished distilling off the volatile components in that temperature range. You have a lull before the next component in series starts coming over even though you have material in your pot. I know these things are usually fractions from x-y temperature but there can be dead zones where there are no hydrocarbons. Keep cranking up the heat and you should hit the next one inline.

    j_sum1 - 3-9-2022 at 04:43

    Quote: Originally posted by Rainwater  
    Might be a secondary reaction.
    You can confirm the boiling point of the third fraction by measuring the temperature of the boiling liquid directly.

    That's a good idea. I can't think what reaction it would be (maybe a decomposition), but testing the mixture is sensible.
    Quote: Originally posted by BromicAcid  
    My guess would be you finished distilling off the volatile components in that temperature range. You have a lull before the next component in series starts coming over even though you have material in your pot. I know these things are usually fractions from x-y temperature but there can be dead zones where there are no hydrocarbons. Keep cranking up the heat and you should hit the next one inline.

    Dead-zones I understand, but this was different. There was no observable change in the rate of material coming over.



    FTR, the fraction I was after (third fraction) appears crystal clear and bright with a high refractive index. It has nearly no odour. I am happy with this.

    The very first fraction (only 3mL from 500) was a little cloudy -- hence my conclusion that it contained water.
    Fraction 2 (120-150) has a little odour -- a slightly turpentine smell and is clear. It has since dropped a small amount of water that has settled to the bottom of the beaker.
    The fourth fraction smells more than the third: a slightly diesel-like aroma. It is also clear but does not seem as bright and refactive as fraction 3.

    THe residue in the flask was only about 5mL. It was quite oily in appearance (unsurprisingly.) I was expecting something a bit tarry from the oxidation products. It was only a slightly darker shade of yellow than the mixture I started with in the second distillation. So, I am not sure if anything had decomposed and come over in fraction 4.

    I guess the next thing to do is some tests on the 4th fraction to find its BP and also discover if it has anything with unsaturated bonds.

    Tsjerk - 3-9-2022 at 04:45

    Quote: Originally posted by Rainwater  
    Might be a secondary reaction.
    You can confirm the boiling point of the third fraction by measuring the temperature of the boiling liquid directly.


    Can't you be cracking something indeed? When you say it first didn't come over, but later it did, with high volume...

    [Edited on 3-9-2022 by Tsjerk]

    j_sum1 - 3-9-2022 at 04:52

    Cracking was not something I expected. I thought some kind of catalyst and a higher temperature would be required.
    But, it is something I can test for.

    Rainwater - 3-9-2022 at 06:10

    Standard procedure. List all the possibilities and start crossing them out one by one.

    Ideas

    1)Being a refined product, theirs no doubt that at some point in the production of the kerosene a catalyst was used. Could be a trace amount in your reagents.

    2) extra high heat input is blowing through fractions.
    Distillate temp = 180c
    Boiling flask temp = 350c
    There has not been mention how hot your heating was. This may not be an issue

    1A) Temps that high could cause some cracking if a catalyst is present. But not the volume described. Tho there isn't a time reference.

    3)Multiple azeotropes could be present. Stabilizers, dyes, etc. Fuel oils are not very pure
    50%A + 25%B + 25%C = boiling point 180
    20%A + 40%B + 40%C = boiling point 160
    50%B + 50%C = boiling point 140
    25%B + 75%C = Bp 130
    Etc.

    3) maybe a TLC plate could be used to separate the different components?
    I have limited experience with TLC, not sure if there is a better method

    4) faulty equipment.
    At 0-200 i dought that a thermocouple would read inaccurate if its functional, i know that 600-1400, if not properly rated, will read inaccurate or outright fail

    Tsjerk - 3-9-2022 at 07:54

    Quote: Originally posted by j_sum1  
    Cracking was not something I expected. I thought some kind of catalyst and a higher temperature would be required.
    But, it is something I can test for.


    I also didn't think it would be that easy, but maybe industry uses catalysts because it is more efficient? Easy to test though as you say. Would be an interesting find.

    j_sum1 - 3-9-2022 at 18:31

    Tested with KMnO4.
    No reaction at room temp. Oxidised with some difficulty in Bunsen flame. But then, any hydrocarbon will do that with an oxidant as aggressive as high temperature permanganate.
    So, I guess that rules out cracking as an explanation.

    I have several more batches to run so I will monitor the distillation temperature more closely.

    Sulaiman - 3-9-2022 at 18:56

    The (Rainwater) suggestion to measure the bp of your products seems the simplest and most logical next step.

    j_sum1 - 3-9-2022 at 19:51

    Ok. Mystery solved. I am not going crazy.

    The most likely explanation is that the thermocouple moved and I did not notice it.
    I use a K-type thermocouple in a thermometer well. And I usually secure it in place with a cotton ball or scrunched up piece of alfoil. This time around I used a small piece of paper towel and it was noticeably more mobile.
    This also explains my observing the same phenomenon twice.

    I redistilled the tails to watch the temperature profile.
    Boil began at 182°C as measured with infra red thermometer.
    Temperature at the still head was 185°C when steady distillation started (at a rate comparable with what I was distilling before.)
    Temperature at the still head peaked at 201°C and dropped for the last few mL once the flow rate slowed.
    This all seems entirely consistent with what would be expected.

    There are some lessons in there about thermometer placement and the reliability of readings. Good to have the reminder.

    j_sum1 - 3-9-2022 at 19:54

    I was prepared to consider the effect of residual MnO2 particles acting as a catalyst. They could have been what gave the mixture a slight yellow colour. Then, during distillation, initiated cracking.
    Evidence seems to refute this hypothesis.

    Sulaiman - 3-9-2022 at 20:36

    lubricating or cooking oil gives a more accurate reading at the cost of slower response times,
    ok to 200C.
    A short thermo-well/pocket is very sensitive to ambient temperatures (giving a low reading)
    and glass temperature (higher or lower reading depending upon setup)

    My best setup so far is a k-type thermocouple
    in a 7mm od glass tube
    with the end of the tube melted and squeezed (with pliers) to seal the tube and grip the thermocouple junction.
    I use a 19/23 or 24/29 ptfe overhead stirrer seal/stopper.
    Quite quick response, adjustable insertion depth and very good chemical resistance.


    PS I had very variable remperature readings when measuring the liquid and vapour temperatures
    using a side arm of a boiling flask during distillations of ethanol.
    When the liquid level in the flask reduced to below the level of the heating wires in my mantle,
    the glass of the flask above the liquid level was 10's oC higher than the temperature of the liquid.
    I popped an alcohol thermometer due to this.

    [Edited on 4-9-2022 by Sulaiman]

    BromicAcid - 4-9-2022 at 05:13

    Of course 185°C is around the temperature where dicyclopentadiene starts to really crack giving cyclopentadiene in the distillate. I know you said you solved the mystery so this is no very likely, but it certainly has a distinct odor.

    j_sum1 - 4-9-2022 at 13:35

    Hmmm. Should not be any dicyclopentadiene after treating with permanganate. Certainly not enough to account for 30% of my mixture.

    One of the things I love about this site us that you pose a little problem and you end up with a wealth of knowkedge on tangential topics. So enriching.