A candidate for "dark matter", predicted and now apparently detected to have both a higher thermal conductivity and higher diffusion rate than H2 is purported in this PowerPoint Presentation file. From that presentation, this output of a Thermal Conductivity Detector (run at 60C) of a HP 5890 Series II gas chromatograph using an Agilent molecular sieve column with Helium carrier gas is puzzling. The Hydrogen peak was establish by a control run. However, in the experimental run there appears a mysterious feature annotated with "???" indicating some sort of gas (that is not Helium) with a higher diffusion rate than Hydrogen.

Mystery features ??? ???

A TCD temperature of 60C was chosen to ensure hydrogen would not show up as a negative peak as indicated by Figure 2 of "Thermal Conductivity Detector Analysis of Hydrogen Using Helium Carrier Gas and HayeSep® D Columns", by K Snavely and B Subramaniam.

enter image description here

What are plausible candidates for this feature?

For context, here is a control run:

90H2+10CH4 Control

Here is the corresponding experimental run from which the above detail screenshot is taken:

enter image description here

After the reaction, the sealed vessel stood for over 24 hours. This TCD graph obtained. Although a small peak associated with Nitrogen appeared, due to atmospheric ingress to the vessel, the Hydrogen peak no longer exhibited the "???" features:

enter image description here

  • $\begingroup$ Just fyi, before you waste time on this topic. The user James Bowery is a disciple of Randell Mills. He has posted here to push the conversation into the "Could the anomaly be Hydrinos?" direction. Link to the BLP subreddit Another link where he admits using "stealth mode" $\endgroup$ – Steve Miller Mar 7 at 19:07
  • $\begingroup$ I'm running what is called a "blind" review to avoid precisely the kind of prejudice you have exhibited by calling me a "disciple". The data and procedures stand or fall on their merits. If you want to accuse Randall Mills of scientific fraud, that is another matter. People tend to be polarized on controversial topics so my attempt at a blind review is entirely reasonable, and calling it "stealth mode" to those who are polarized toward Mills is reasonably politic. $\endgroup$ – James Bowery Mar 7 at 21:17
  • $\begingroup$ The data is still there. No amount of prejudice will change the value of it. You will either find someone who can confirm that Hydrinos have been found or you will not. Good luck. $\endgroup$ – Steve Miller Mar 7 at 22:34
  • $\begingroup$ Well your likely-successful attempt to prejudice the review is too late in any event. I've got 2 alternative explanations and have put them into perspective with the discovery of a paper -- on my own -- that addresses one of them (now cited) regarding TCD operating temperature and the other (overloading analyte) to be critically addressed. $\endgroup$ – James Bowery Mar 7 at 22:41
  • $\begingroup$ Oh, and don't sell yourself short: When topics are so controversial people's careers can be terminated and mortgages foreclosed for merely looking into them, making blind reviews impossible may foreclose discovery of the truth -- yea or nay. The fact that you're too late by maybe 16 hours in doing your worst should not count against you. You did try. $\endgroup$ – James Bowery Mar 7 at 22:45

It would be nice to have a full chromatogram rather than a cropped one to see how other peaks look like. The phenomenon of odd behavior of hydrogen with helium as carrier gas with TCD detector in quite well known, but I am not sure if it has been investigated in detail in the literature or not. Hydrogen has higher thermal conductivity than helium. Ideally when you inject $\ce{H2}$, one should see a negative dip as you are seeing. However people also see a split peak. Yours is a mixture of split and a dip. The rest of the peaks should appear normal. If you change the carrier gas, this problem will go away (for $\ce{H2}$) but you lose sensitivity for other analytes as helium has the highest thermal conductivity after hydrogen in gas phase.

If you change the polarity of the detector, you would see this phenomenon for $\ce{H2}$ reverted and other analytes would appear as negative peaks.

  • 1
    $\begingroup$ I've added 2 full chromatogram graphs for context. $\endgroup$ – James Bowery Mar 3 at 16:49
  • 1
    $\begingroup$ That makes perfect sense, because the rest of the peaks are positive and only H2 is messing. The answer is still valid. $\endgroup$ – M. Farooq Mar 3 at 17:03
  • 1
    $\begingroup$ The preceding positive spike is higher than the following positive peak for hydrogen. Is this what you are referring to as an apparent split peak for hydrogen? If one should expect to see a negative dip, why might it appear only in the experimental run and not the control run? $\endgroup$ – James Bowery Mar 3 at 17:03
  • 1
    $\begingroup$ Just saw that you added more info. What is the difference between control and experimental runs? I mean what are you injecting in each case. The "misbehavior" of H2 peak should be seen with the injection of pure H2 as well as the sample containing H2. Do you know the dead time (void time) of the column? It may be possible that something is co-eluting with H2 in the experimental part. $\endgroup$ – M. Farooq Mar 3 at 17:11

Hydrogen is the only element with thermal conductivity greater than helium ($0.182$ vs $\pu{0.151 W\,m^{-1}K^{-1}}$ at $\pu{25 ^{\circ}C}$) as M. Farooq pointed out (cf., thermal conductivity of $\ce{N2}$ is $\pu{0.026 W\,m^{-1}K^{-1}}$ at $\pu{25 ^{\circ}C}$). The mixtures of $\ce{H2}$ in $\ce{He}$ at moderate temperatures exhibit varying thermal conductivities, more or less than either component alone. Thus, if you are analyzing for $\ce{H2}$ with $\ce{He}$ as your carrier gas, a peak of $\ce{H2}$ may appear as positive, negative, or as a split one (This may be the case for the difference between your reference and experimental runs).

One solution for this problem is to operate the detector at higher temperatures, to say, from $\pu{250 ^{\circ}C}$ to $\pu{300 ^{\circ}C}$: You may analyze a known range of hydrogen concentrations using your said method until you find the correct detector operating temperature. To do so, you may need to increase the operating temperature several runs, until the hydrogen peak exhibits normal shape. And ensure it is always in the same direction regardless of $\ce{H2}$ concentration. This newfound detector temperature also ensures high sensitivity because of the higher value.

  • $\begingroup$ These issues raise the question as to whether there exist validated models of binary gas mixture thermal conductivity. The result of my search for this came up with this relatively recent paper: thermalfluidscentral.org/journals/index.php/Heat_Mass_Transfer/… $\endgroup$ – James Bowery Mar 4 at 14:38
  • $\begingroup$ The TCD ran at 60C (which I added to the question). However, the graph that I added showed that by merely permitting the reaction vessel to stand idle after the reaction (over 24 hours) the ??? features disappeared and the graph reverted to the control except for a slight N2 ingress (the CH4 peak apparently being cropped out). $\endgroup$ – James Bowery Mar 5 at 3:10
  • $\begingroup$ 1) Please mention the dead time or void time of your chromatograph. It is the time it takes for an unretained molecule to travel from the injector to the detector. Is H2 eluting at the dead time? Without knowing this information, all this discuss can become meaningless. 2) Your detector settings, split ratio are not suitable either. Why there is so much noise, why the baseline is oscillating? $\endgroup$ – M. Farooq Mar 5 at 6:16
  • $\begingroup$ @M. Farooq: I think high noise duto the detector temperature, which is operating at $\pu{60 ^{\circ}C}$. I still think OP need to increase that temp to a higher value (probably over $\pu{250 ^{\circ}C}$ since autoignition temperature of $\ce{H2}$ is way above $\pu{500 ^{\circ}C}$) to get better sensitivity unless there is a reason to operate at $\pu{60 ^{\circ}C}$. $\endgroup$ – Mathew Mahindaratne Mar 6 at 22:29
  • $\begingroup$ @MathewMahindaratne, I agree, the settings need to be optimized. $\endgroup$ – M. Farooq Mar 6 at 23:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.