1. Mick West

    Mick West Administrator Staff Member

    The experience debunking the angle cut column was interesting, as it showed that even though something may well have been debunked sufficiently for most people years ago, that debunking information is often inaccessible and arguable - sometimes buried in discussion threads. The cut column discussion resulted in an inarguable debunk, with accompanying simple graphics.
    Debunked Angle Cut Column 2.

    So I'd like to do the same with one of the most popular claims - that thermite was found in the remains of the towers. Specifically with this 2009 paper:


    I took some steps along this road with the post discussing the misidentification of "abundant manganese" via X-EDS, which taught me a little about the X-EDS process and problems. Jones, it seems, had difficulty correctly interpreting X-EDS back in 2006. Perhaps there were similar errors in the 2009 paper?

    I recognize this is old ground, however my goal here is to produce a go-to debunk of the paper using past, synthesizing past debunks with any more recent information, and presenting it in an accessible manner.

    So let's begin by gathering the past debunks, much of which was done deep within the JREF/ISF forum.

    Oy's summary likely refers to these thread on ISF (formally JREF)

    Which in total contain over five thousand posts. Discussion continues to this day in:

    So what I'd like to do is create a simple yet deep summary of the evidence for paint, and present it in a a permanently findable format with nice large graphics, and maybe a short explanatory video.

    More useful references:

    Last edited: Feb 2, 2018
  2. Mick West

    Mick West Administrator Staff Member

  3. Mick West

    Mick West Administrator Staff Member

    I've always wanted some red/grey chips of my own, and it turns out I had them all along.

    Metabunk 2018-02-02 11-51-48.
    That's a steel wheelbarrow that's been outside for years. There's rust, but there's also lots of red paint
    Metabunk 2018-02-02 11-53-16.

    So, I hit it with a hammer.
    Metabunk 2018-02-02 11-53-56.

    Collected a bunch of flakes. Lots of rust, but some chips of paint.
    Metabunk 2018-02-02 11-54-17.

    Found one!
    Metabunk 2018-02-02 11-55-18.

    It's red on one side, and grey on the other:
    Metabunk 2018-02-02 11-58-38.

    Viewed edge on
    Metabunk 2018-02-02 12-02-51.

    It's a red/grey chip, and I've got a wheelbarrow full of them!
    • Like Like x 4
    • Informative Informative x 1
  4. deirdre

    deirdre Moderator Staff Member

    sorry to interrupt.. are ae911 saying these truss things are literally covered in thermite?
    laclede steel.JPG
    • Funny Funny x 1
  5. Mick West

    Mick West Administrator Staff Member

    I bashed off a bunch of pain chips from my red painted steel wheelbarrow and waved a butane flame over them. Result = iron microspheres

    Here's a scale comparison with the Harrit microspheres (left) and mine (right).

    Metabunk 2018-02-02 13-50-05.

    Of note, in both their photos and mine the red layer appears undamaged. Curious, since that's supposed to be the one that's nanothermite. What seems to have happened is the iron oxide layer has "burnt" (perhaps with some of the paint, of some intermediate layer), and created some iron microspheres.

    Metabunk 2018-02-02 14-12-50.

    Added a 50 micron blue line. The range of sizes of shiny microsphere is very similar to those in the Harrit paper.
    Last edited: Jun 30, 2018
    • Informative Informative x 1
  6. Mick West

    Mick West Administrator Staff Member

    Here's a shot at imaging from the side, so you can see the shape a bit better (different chip, they get lost)
    Metabunk 2018-02-02 14-36-54.

    Again, unburnt paint, molten iron.
  7. Oystein

    Oystein Active Member

    I am away this weekend and only checking a few sites before nodding off to sleep.

    Do not conflate the topic "what are these chips" with "what are they not" or "are they thermitic". There are several grave problems with Harrit et al that render their conclusion "thermite" untenable and ridiculous, with no need to come up with a different concusion.

    I love however your micrographs!!!
    • Agree Agree x 1
  8. Mick West

    Mick West Administrator Staff Member

    Agreed, however it's also good to have a reasonable alternative explanation, even if it can't be proven 100%.
  9. Mick West

    Mick West Administrator Staff Member

    I was wondering how to make a paint chip as close as possible to those found in the WTC dust. The obvious candidate is the primer paint on the floor joists (trusses)

    This image above shows how they were mangled in the fall, and this is probably one of the more pristine examples.

    Here's a specification for the paint used.
    http://ws680.nist.gov/publication/get_pdf.cfm?pub_id=101042 (appendix B)

    Metabunk 2018-02-27 08-52-06.

    So there's several issues
    1. Paint. Is there anything on the market I can get a can of that's a close match? Can I mix it from other things?
    2. Steel. What should I apply it to? A piece of A36 uncoated steel sheeting? How to clean the surface?
    3. Application. Presumably spray-on? What does "bake" mean? Heat the steel to 350°F before or after application? How to get it 1 mil (1/1000") thick?
    4. Aging. How long should I leave it? Should I simulate any weathering, heating cooling cycles?
    5. Flaking. How do I make flakes? Bend the steel? Hit it with a sledgehammer? Heat it? All three, and more?
    6. Testing. Beyond heating the flakes, checking for magnetism, and looking at them under a microscope, wha tother tests can I do to see if they match (or don't match) tests done on WTC dust flakes by Harrit, et al?
    7. Other. ???
    • Like Like x 1
  10. Mick West

    Mick West Administrator Staff Member

    #1 Paint,

    #3 applicaiton
    B50NV12 is the red one, B50NV11 is grey.
    • Like Like x 1
  11. Jedo

    Jedo Member

    There are several references in the Harrit et al. paper where they compared their red/grey chips with paint flakes:

    I can't find anything where they specify which paint they used, except in reference 31, which reads

    So it seems they tested at least for Zinc ferrite corrosion protection paint.

    However, as their data shows and as @Oystein correctly mentioned, they probably should have tested it against paint containing aluminium silicates. They have no mention that they used such.


    So other tests you could do are:

    1. put it in methyl ethyl ketone solution for 55 hours +
    2. try to burn the paint chips with a oxyacetylene flame
    3. test the conductivity / resistance
    4. try to ignite them by heating them up to 435°C, and
    5. try to obtain microspheres when heated to 700°C
    Last edited: Feb 27, 2018
  12. Oystein

    Oystein Active Member

    B50NV12 (B50AV12?) differs from the LaClede shop primer in several regards:
    a) It's Alkyd based, not Epoxy
    b) It has a much higher percentage of solids (pigments): 78 wt-% vs. 28.5 wt-% (this may be due to the application process, spray-on vs. electro-deposition)
    c) No chromate (LaClede has 1.1 wt-% strontium chromate, which is elusive in Millette's study, but has some blips in Harrit's and Farrer's apocryphal data)
    d) It's brownisg-red, rather than orange-red, indicating a larger particle size of the iron oxide pigment

    It may have been an acceptable replacement in NIST's fire test, but may be of rather limited value for comparison with Harrit's chips.

    Better go searching for a recipe that has epoxy binder and also some kaolin.
    • Useful Useful x 1
  13. Mick West

    Mick West Administrator Staff Member

    I wonder if electrodeposition of paint is at all responsible for the two layers? (red and grey). It seems like this would be a tricky process to duplicate in my garage.

    Here's what seems to be something similar:
    Metabunk 2018-02-27 11-14-15.

    It would also seem like this process would explain some of the physical and chemical differences between red/grey WTC chips, and just "paint"

    Is it possible that Laclede might have used a two stage coating process? A grey oxide layer then a red oxide/pigment layer? If you look at electrocoating in automobile manufacturing they use multiple layers.
  14. Mick West

    Mick West Administrator Staff Member

    So this question is answered by the above. Not spray on, and baked AFTER the electrodeposition.
  15. Jedo

    Jedo Member

    So by electrodeposition they mean the following process, where the paint is in solution, and the object completely dipped into it?

    There is also a simpler version of it by using electrostatic charging of the paint particles as the leave the spray nozzle:

    or here with an demonstration with / without potential:

    On the Wikipedia page on Electrophoric deposition they list some advantages they got from a paper that deals with electrodeposition for nanostructures; there they have the following advantages listed:
    So it seems there is no advantage listed from "electrochemical reactions" as cited above, besides the densification of the paint (which is equal to 'free from porosity' I suppose). For I interpret the electrochemical reaction they mention there as simply the hydrolysis of the salt into an acid and a base (on different electrodes), or just having the acid fall out and leaving the solution basic.

    I would first spray it with an air spray without electrostatic charging simply.
  16. Jedo

    Jedo Member

    Does the Spray-Applied Fire Resistive Material (SFRM) play any role here?


    In the WTC they used parlty asbestos, partly mineral wool. But the thickness was up to 3/4". So probably not…
  17. Mick West

    Mick West Administrator Staff Member

    My poor wheelbarrow is taking a bit of a pounding in an attempt to make some flakes. I suspect it is electrocoated, as it's sure hard to get anything off it.
    Metabunk 2018-02-27 12-22-22. Metabunk 2018-02-27 12-22-58.

    What I collected was basically dust, nothing over 1mm
    Metabunk 2018-02-27 12-25-15.

    Source: https://www.youtube.com/watch?v=j-8gY-YBtmw
    • Like Like x 1
  18. Oystein

    Oystein Active Member


    First, we have the specification, and it specifies only one paint.
    Secondly, NIST did not think two layers were applied or needed.
    Thirdly, the SEM and EDS work by Harrit/Jones/Farrer and also Millette shows the gray layer as iron/iron oxide with bits of carbon, manganese, silicon - a composition consistent with structural steel - which is solid and homogenous, quite different from the epoxy+pigments red layer.
  19. Mick West

    Mick West Administrator Staff Member

    So it seems:
    NCSTAR 1-A1
    Or does it?
    We only have the specification of the Laclede paint, but it looks like there's also a possible "PPG red power primer RF-2184 replenishing material"?

    The "sections" referred to above are from this:
    "PONYA (Port of New York Authority). 1967. Fabricated Steel Floor Trusses, Bridging, Beams and Bracing for Prefabricated Floor Units for North and South Towers. World Trade Center Contract
    WTC-221.00. (WTCI-71-I)."
  20. Oystein

    Oystein Active Member

    This is pretty clear IMO: The steel was to be painted with one paint: A red chromate paint. This paint was specified to adhere (either) to the PPG company standard OR to the LaClede in-house standard. It may be possible that some proportion of the steel trusses was painted with the PPG paint instead of the LaClede paint, but there is no mention whatsoever of two layers, or any gray paint.
  21. Mick West

    Mick West Administrator Staff Member

    I agree with the one layer of paint. I was wondering with this though if this PPG paint might account for some of the differences between paint chips (if such a difference has in fact been conclusively shown)
  22. Efftup

    Efftup Senior Member

    Certainly Tnemec Primer has been mentioned repeatedly in discussions and I have certainly seen discussion that there was likely more than one different primer on chips. Millette was not given access to Harrit's chips so had to match by xeds data. He discarded chips that didn't match.Harrit apparently had at least one with different primer. It is of course hard to tell as Harrit's details are a little sloppy. In the paper there are 4 samples mentioned 1. Mckinlay, 2 . Delassio 3. Intermont and 4. White. Then XEDS spectra are shown a,b, c& d apparently corresponding to 1 through 4 but on DSC tests there are 2 Mckinlay samples and no Delassio so these appear to be different chips. Also the peaks are lower for the 2 Mckinlay samples on the DSC and on XEDS chip c shows elements not found in the other 3.
  23. Tomi

    Tomi Member

    The paint had no real function other than to make the steel look nice and new. It is sometimes called a holding primer or a works primer. The steel after erection is in an internal environment so there is no risk of corrosion.

    Red oxide primers are still used today.

    I would expect that the bar hoist floors would be made at a different location to the heavy steel columns. And it would be probable/possible to have a different paint supplier and even a different composition. If it is a big job there may be multiple fabrication centres all of whom would have there own version of a red chromate primer.

    It is not an important part of the work and as such there is less control on the outputs.

    It would be very unusual to have more than one coat since you are not trying to provide corrosion protection you are only trying to provide a uniform coating. Of course painting bar joists is quite difficult, as is fire protecting them, and you may need multiple passes to get a uniform surface.

    And it would be quite normal to paint over mill scale and light rust and shop welds
  24. Oystein

    Oystein Active Member

    The paper actually documents that they were looking at at least 6 (six!) different materials, judging from the elemental composition:
    Material 1 - the 4 chips analysed in Figures 5-11 - has a red layer with only the elements C, O, Si, Al and Fe in more than trace amounts (they also have traces of Cr, Sr, Na, Ti and others), with the extra property that Al and Si appear in roughly equal amounts (nearly equal peak hights, which in this case of elements that are neighbors in the periodic table translates quite directly to amounts).
    Material 2 - the chip analysed in Figures 12-18, which has been subjected to MEK solvent - additionally contains Zn and Mg, and very likely Ca.
    Material 3 is one that has significant Ti in its ash and thus in its composition - an element that was not significant in the first two materials
    Material 4 is the multi-layered chip, which contains significant Pb in the red layer, which none of the previous three materials exhibit
    Material 5 is a chip where the gray layer is not mostly made up of iron/iron oxide but of carbon
    Material 6 (and likely 7) are evidenced when the text body says that they found chips with significant Cu and Ba, neither which any of the other materials contain. It is not clear whether Cu and Ba were found within the same specimens, or in different specimens.

    It is of course spurious to extend findings on one material to any or all of the others. This alone renders the paper worthless.

    Yes, correct, the 4 chips represented in the DSC data of Figure 19 are a set different from the 4 chips represented in the SE/BSE/XEDS data of Figures 5-11.

    It is unlear which of the six materials listed above are represented by which DS curve, if any.
    It is spurious to lump together the DSC data and the XEDS data, as it is not clear at all they were won from the same material(s). This alone renders the paper worthless.

    Do not confuse yourself too much by focussing on the identity of the four dust samples. Each dust samples could contain any and all primer paints that were released into the total dust. If you look at two chips, it matters little if they came from the same or from two different samples.

    As for the DSC data, I would want to point out that they use NONE of the numbers to form their conclusions! They only look at one quality: That there is an exotherm peak. Somewhere. Of some height. There is no analysis, no discussion of the peak temperature or peak power. There is nothing about the value "425 °C" that makes the conclusion "thermite" likely or even compelling.
    In Figure 29 they compare one of their DSC traces with one trace of actual, experimental nano-thermite. Not only are peak temperature and peak power very different; the curves have MANY differences, and nothing of significance in common, except that both show an exotherm reaction. Which is d'uh.
    What's worse: The four traces in Figure 19 are not very similar, except for the peak temperature. The black and the green curve are somewhat similar, and the red and blue share similarities, but the latter two are much different from the former two.

    It is perfectly clear that peak temperature and power are dominated by the combustion of the organic matrix. Many organic polymers burn in the vicinity of 400 °C or slightly beyond. One such polymer is epoxy (or epoxies - there exist various). In fact, the blue and red curve have properties reminiscent of epoxy.

    Harrit et al note themselves that the specific energy they measured with the DSC data (1.5/3/4.5/7.5 kJ/g) is, in some cases, more than what thermite could possibly do, and they conclude, correctly, that "some" of that energy comes from the matrix.
    They fail to consider that perhaps ALL of the exotherm comes from burning organics.
    Since the gray layer is iron oxide and thus more or less inert - as Harrit et al themselves admit -, but also denser than the gray layer, it follows that the gray layer adds mass, but no energy, to the DSC power flow. Specific energy of the red layer alone must thus be conjectured to be twice or more the values presented. This raises a huge problem: It can be computed that 90% and more of the energy MUST come from the organics, and <10% only could theoretically come from a thermite reaction. This means immediately that the DSC peaks are organic combustion peaks.
    There is no additional thermite peak. It just doesn't happen.
    • Like Like x 1
  25. Oystein

    Oystein Active Member

    The trusses were painted in the factory to protect them from the elements during transport, storage and constrution.

    Yes, and more importantly: Has been used for decades, if not millenia.
    Red oxide pigment is sub-miron sized - i.e. nano-scale - and has been for decades, if not millenia. It is not a high-tech product.

    Yes, this is true. The WTC steel ame from half a dozen or more different steel manufacturers: I remember Pittsburgh Steel for part of the exterior columns, LaClede Steel for the floor trusses. Some other company provided the rest of the exterior columns, two or so companies shared in providing core columns, and then you have steel in the foundations and elsewhere, for which yet other companies were contracted.
    And indeed, different paints were specified: The floor trusses (LaClede) were specified as LaClede's shop primer - epoxy matrix, iron oxide, aluminium silicate, strontium chromate. The exterior colums were to be painted with Tnemec Red 69 or 99, which is alkyd resin with linseed oil, and pigments of iron oxide, zinc chromate, plus talc, silica and calcium aluminates. We do not know what paint(s) were specified for the core columns. We do not know what was specified for WTC7. No doubt other paints were used.

    The trusses were dipped into a bath of paint, direct current applied (anode, cathode), and the paint thus deposited in one application

    Specification demanded that the steel be cleaned before painting.
    • Like Like x 2
  26. Efftup

    Efftup Senior Member

    I am in no confusion as to the worthlessness of the Harrit paper. I just mentioned there was definitely more than one paint sample and the general sloppiness of the Harrit paper.
    As you mentioned, not only do the DSC samples not match each other, but when compared to the Tillotsen Xerogel Nanothermite it is not even close. The Nanothermite peak is way lower, starts at a much higher temperature and is in fact ENDOTHERMIC until about 380 degrees C.

    Looks like the papers authors were trying to "prove" something instead of TESTING a hypothesis and ignored the fact that their results completely did NOT match what they expected.
    The Energy peak which they claim is a violent sudden release of energy is a spread of 60 degrees or more and as the DSC would normally raise the temperature by about 10 deg/min then the Sudden release took at least 6 minutes to happen.
    As you already mentioned, the fact that they put the whole chip in the DSC and not just the red layer meant their results would have been useless right from the start. Others have also pointed out is the WRONG test to do to determine composition and why Millette did NOT do that test.
  27. Oystein

    Oystein Active Member

    All correct and relevant, Eftup :cool:
  28. Mick West

    Mick West Administrator Staff Member

    This article shows another primer paint composition:
    https://11-settembre.blogspot.com/2009/04/active-thermitic-material-claimed-in.html (http://archive.is/xJHJN)

    With some discussion about how this might account for the red flakes and some of the observations
    Last edited: Nov 16, 2018
  29. Oystein

    Oystein Active Member

    The speculation about the gray layer being "linked to a green corrosion-proofing paint (Tnemec Green Metal Primer, page 303)" is almost certainly false in most instances of red-gray chips.

    Here are four XEDS graphs of the four chips Harrit et al analyse in their first part (Fi. 5-11):

    ActiveThermiticMaterial_Fig06_ab. ActiveThermiticMaterial_Fig06_cd.

    These are essentially Fe and O only - the peaks for C indicate a small percentage, consistent with carbon steel. (a), (b) and (d) also have small but unlabeled signals at 5.9 keV - that's the K-alpha level of manganese. And of course manganese is part of most of the structural steel alloys used at the WTC.

    The gray layer generally is oxidized structural steel.
    There is no hint of any other element at least in these four specimens, and thus no hint of, e.g., green pigments, or any other pigments, really. No zinc, no chromium, no lead, no copper, no nothing.
  30. Oystein

    Oystein Active Member

    Diatomaceous silica is silica formed by diatoms - unicellular algae that have a surprisingly massive imprint on the planet's ecology. Diatomaceous earth is mmade up of the deposited silica shells of ancient (fossile) diatoms, with particle sizes typically in the range from 10 to 200 μm. See Wikipedia: https://en.wikipedia.org/wiki/Diatom#Evolution_and_fossil_record
    These shells have organic structures and would stand out under a microscope of fitting magnification!

    Harrit et al show the micro-structures of the red paint layers of only four chips, and they have specifically selected these four chips foor their very similar properties.

    ActiveThermiticMaterial_Fig04. ActiveThermiticMaterial_Fig05. ActiveThermiticMaterial_Fig08.
    There is no diatomaceous silica in these red layers.

    The reason is simple: These four chips are LaClede shop primer on floor truss steel; and the LaClede paint recipe has no silica.

    Diatomaceous silica is specified for Tnemec Red 99 - to be painted on the perimeter panels. Harrit et al are looking at a chip of Tnemec 99, in Figures 12-18, but surprisingly failed to include an image of the microstructure. One might suspect that they took micrographs in ranges of 1 to 100 µg, but consciously decided against showing them, because they would contain diatom shells, but no hexagonal kaolin platelets so typical for chips (a) to (d).
  31. Mick West

    Mick West Administrator Staff Member

    Some more paint chips from my wheelbarrow.

    Metabunk 2018-10-15 10-54-58. Metabunk 2018-10-15 10-55-34. Metabunk 2018-10-15 10-55-51. Metabunk 2018-10-15 10-56-09. Metabunk 2018-10-15 10-56-27. Metabunk 2018-10-15 10-56-45.
  32. Mick West

    Mick West Administrator Staff Member

    This is a "beam" (probably a section of a column, based on the thickness of the steel) from the World Trade Center site, in Ypsilanti, MI.


    Metabunk 2018-11-16 05-32-03.
    It strikes me as a good potential source of paint chips for comparison. You would not probably need to scrape any off, just swipe a medium strength magnet over the bottom.

    It might not be that useful, as the rusting is probably not typical. Like my wheelbarrow chips, the oxide layer is more orange than the cleaner chips seen in WTC dust, which partly formed with more mechanical assistance.
    Last edited: Nov 16, 2018