Origin of the paint that was found as red-gray chips - any ideas?

[Oystein]

You seem to think differently here;

http://www.internationalskeptics.com/forums/showpost.php?p=7466108&postcount=115



MM

Keep reading the thread from that post onward.

One difference between truthers and rational people I consistently observe around here is that we are willing and able to learn.

 

[Oystein]

How predictable: Miragememories seems to be running away from pertinent, inconvenient questions.

To wit:

In post 735 I repeated about 13 questions (a couple of which may be redundant, okay, so make that 9 or 10 separate questions) I asked of him in this thread during a time when MM was actively participating, and which he hasn't answered so far.

MM posted three posts later (# 738), and again two posts later (#740), so we know he opened this page.

Could it be that he ignored the unanswered questions on purpose?

MM, please show some integrity and honesty and try to answer the open questions I summarized in post 735!

 

[Ivan Kminek]

Oystein: It is perhaps better that MM is not responding (At least for me...)

For a while, back to two unsolved matters of our indirect detective work:

1) Temperature of thermal-oxidative degradation. Considering that "Bentham chips" (a) to (d) contained epoxy binder, why exotherms observed in DSC machine peaked at ca 420 - 430 degrees C, whereas for typical (freshly prepared ) epoxies, the fastest thermal-oxidative degradation ("burning" in fact) is observed usually at ca 380 degrees C?
As I already wrote here, it could be attributed (among others) to the fact, that Bentham chips were almost 40 years old.
Unfortunately, I have not found any paper in which TGA (less the DSC) was measured for fresh and aged epoxy resins. Many papers investigating an aging of epoxies exist, but they usually follow changes in mechanical or other properties important for applications, and/or some spectral properties (e.g. infrared spectra) etc.
This morning, I have visited our department dealing with polymer degradation (Dr. Jana Kovarova). They confirmed my assumption that if a thin layer of epoxy is significantly "pre-degraded" (e.g. pre-oxidized) because of very long contact of with air/humidity, it can be oxidized/degraded/(or even ignited) at higher temperatures during measurements of the thermal behavior. But, there are several other factors playing role. E.g., if the aging leads to the epoxy network scissions (not frequent), a main deal of degradation can be shifted to lower temperatures. And quite opposite, if the aging leads to additional crosslinking (more frequent), caused by e.g. oxygen, water or by unreacted epoxy/amino groups, thermal-oxidative degradation can be shifted to higher temperatures. Therefore, no general rule can be expected here.

2) Nitrogen XEDS signals in epoxies. You wrote here that XEDS spectra of Laclede paint chips should show nitrogen peaks (between carbon and oxygen peaks) because epoxy binder contains nitrogen atoms from curing agent (amine). I gave you several links to papers showing XEDS spectra of epoxies cured by amines, where even the minute peaks of nitrogen were not visible (e.g. http://www.geos.ed.ac.uk/facilities/ionprobe/EpoxyResins/Composition.html).
Yesterday, I found another papers showing XEDS spectra of polyamides and polyimides, which are also polymers containing nitrogen atoms. Again, no nitrogen signals are visible anywhere, eg. http://esmat.esa.int/Materials_News/ISME09/pdf/10-In-flight/S12 - Yamanaka.pdf, Fig. 4.
I still do not know why.
I think that I asked Sunstealer and/or Almond, how it is with nitrogen XEDS signals (perhaps it is a matter dependent on the device and its detector window), but no answer yet...

 

[Miragememories]

"Temperature of thermal-oxidative degradation....They confirmed my assumption that if a thin layer of epoxy is significantly "pre-degraded" (e.g. pre-oxidized) because of very long contact of with air/humidity, it can be oxidized/degraded/(or even ignited) at higher temperatures during measurements of the thermal behavior. But, there are several other factors playing role. E.g., if the aging leads to the epoxy network scissions (not frequent), a main deal of degradation can be shifted to lower temperatures.

And quite opposite, if the aging leads to additional crosslinking (more frequent), caused by e.g. oxygen, water or by unreacted epoxy/amino groups, thermal-oxidative degradation can be shifted to higher temperatures. Therefore, no general rule can be expected here."

formatting changes are mine

Makes sense.

"Nitrogen XEDS signals in epoxies. You wrote here that XEDS spectra of Laclede paint chips should show nitrogen peaks (between carbon and oxygen peaks) because epoxy binder contains nitrogen atoms from curing agent (amine). I gave you several links to papers showing XEDS spectra of epoxies cured by amines, where even the minute peaks of nitrogen were not visible (e.g. http://www.geos.ed.ac.uk/facilities/...mposition.html).

Yesterday, I found another papers showing XEDS spectra of polyamides and polyimides, which are also polymers containing nitrogen atoms. Again, no nitrogen signals are visible anywhere[/b], eg. http://esmat.esa.int/Materials_News/...20Yamanaka.pdf, Fig. 4. I still do not know why...

formatting changes are mine

Why?

Maybe because Oystein's expectations are wrong.

MM

 

[Oystein]

No worries, Ivan, MM will not reply to any of the pertinent questions (he must know full well that any honest answer will expose the fraud that he is), and I promise I will try my best not to respond to any further evasions


Thanks for re-raising some open questions. Here are more:

3) We have seen small signals for elements that we did not expect from the theoretical composition of LaClede primer:
- Na, S, K and Ca in chip c (Fig. 7 of Bentham paper)
- Na and K associated with both hematite and kaolinite particles in an unidentified sample (Fig. 11 ibid)
- S and Ca found post-DSC residue in Fe-rich sphere from some unidentified sample (Fig. 21 ibid)
- S, Ca and Ti (!?) in another post-DSC sphere from some unidentified sample (Fig. 25 ibid)
- S and Ca in chip a (Fig. 5b of Harrit's "Why the red/gray chips are not primer paint")
- Marc Basile quantified 0-0.22% Na, 0.22-0.25%S, 0.4-0.5% K, 0.87-0.96% Ca in his youtubed presentation. Not sure where that sample and the numbers came from - borrowed from Harrit, or his own samples and work.

Now, I'd be very careful with all the unidentified samples; these could be anything. Presence of Ti for example raises an alarm with me as that could well be a third kind of paint.

I am most concerned about Ca and S, which pop up in chips a and c, and especially the nearly 1% Ca in Basile's presentation.

I think the most likely source of Ca (and Na and K) is impurities in the natural pigments, especially kaolinite - if these pigments are indeed natural. However, in Basile's samples, Ca content is roughly 50% and 75% of the Al content; too much to merely be an impurity!

S could also be a component of natural minerals, or play a minor role in the epoxy, but that's comjecture that I am certainly not happy with


4) Observed thickness of red layer is smaller than specified (1mil +/- 0.2 mil = 25µm +/- 5µm). Did LaClede cut corners? Or did the layer shrink during 40 years?



All four questions could be answered quite definitively by experiment if we could find a nice, unburned piece of actual floor joist.

 

[Oystein]

formatting changes are mine

Makes sense.


formatting changes are mine

Why?

Maybe because Oystein's expectations are wrong.

MM

I hope you are grasping the concept that we, as skeptics, raise these concerns ourselves because we always doubt our own theories and arguments, as a matter of principle. I realise that this concept is extremely foreign to Truthers.

So you say maybe my expectations are wrong? Hell no! They are definitely wrong! Obviously! See, I assume good faith and consider the data presented by Harrit, Basile etc. to be actual and reliable (within the accuracy of the methods that were used, of course).

My expectation is this:
I expect that the red layer is LaClede primer (1) which consists of up to 71.5% epoxy amine (2), an organic binder that contains roundabout 10% nitrogen by weight (3), and this N should show up in the XEDS spectra (4).

My expectation is definitely wrong, because N doesn't show up in the XEDS graphs. Big question is: Why is my expectation wrong? Could be that (1) or (2) or (3) or (4) is wrong, no? So yes, I am open to the possibility that my expaction of LaClede primer is wrong. Could be. Could also be that LaClede didn't use the epoxy amine binder that was specified, and instead used some non-amine epoxy. Could also be that my expectation of the amount of N that is still present in cured and aged epoxy amine is significantly too high. Could also be that for some technical reason, the XEDS would not show an identifyable peak for N even though it is present.

Now MM, can you shed some light here, and help? Which of the 4 assumptions that I put into my expectation is wrong, and why?

Please be so kind and STFU if you don't have anything substantial to say on this question, just as you were perfectly able to STFU about 13 other questions asked directly of you.

 

[Ivan Kminek]

formatting changes are mine

Makes sense.


formatting changes are mine

Why?

Maybe because Oystein's expectations are wrong.

MM

MM, ad) your first point: you know, this is just normal in real science: some processes, e.g. degradation of epoxies, are so complex that no general rules can be established in some respects.

Ad) your second point (Maybe because Oystein's expectations are wrong): you apparently again do not understand.
1) Oystein wondered why there were no signals of nitrogen in XEDS spectra of Bentham chips, if they were particles of Laclede paint and therefore must contain epoxy binder cured with some amine (nitrogen compounds).
2) I wrote him that I do not know (I am a "novice" in XEDS), but also published XEDS spectra of epoxies cured by amines (and also other polymers containing nitrogen) do not contain peaks of nitrogen. So: although these polymers for sure contain nitrogen, its peak is not visible/hidden/too weak in XEDS for some reason (unknown for me)...

 

[NoahFence]

delete - wrong truther

 

[The Almond]

[...]
2) I wrote him that I do not know (I am a "novice" in XEDS), but also published XEDS spectra of epoxies cured by amines (and also other polymers containing nitrogen) do not contain peaks of nitrogen. So: although these polymers for sure contain nitrogen, its peak is not visible/hidden/too weak in XEDS for some reason (unknown for me)...

The mass absorption coefficient for nitrogen X-rays in a carbon rich matrix is about 25,000 cm^2/g. Any N x-rays are going to be absorbed very strongly by the matrix. Further, at 15 keV accelerating voltage, the ionization potential of N k-shell electrons is basically 0. It is entirely expected that you won't see N peaks in this material.

 

[sheeplesnshills]

From page 4 of this thread - http://www.internationalskeptics.com/forums/showpost.php?p=7475372&postcount=143

So they couldn't use the exact same formulation of paint because it's no longer manufactured.

It's a moot point.

It was only of interest in as much as they wanted to see if it affected the fire proofing adhesion. As you say its a moot point as it was well known that the adhesion was poor at best and the impact and explosion would have removed much of the fireproofing in any case. The paint type and ignition point etc would have had little effect, if any, on how the fire affected the trusses.

 

[Ivan Kminek]

The mass absorption coefficient for nitrogen X-rays in a carbon rich matrix is about 25,000 cm^2/g. Any N x-rays are going to be absorbed very strongly by the matrix. Further, at 15 keV accelerating voltage, the ionization potential of N k-shell electrons is basically 0. It is entirely expected that you won't see N peaks in this material.

Almond, thank you very much

 

[Oystein]

The mass absorption coefficient for nitrogen X-rays in a carbon rich matrix is about 25,000 cm^2/g. Any N x-rays are going to be absorbed very strongly by the matrix. Further, at 15 keV accelerating voltage, the ionization potential of N k-shell electrons is basically 0. It is entirely expected that you won't see N peaks in this material.

Thanks a lot!

But...

I don't have a feel for how significant this is. Can you explain this a little and put into perspective, for example by giving examples of the mass absorption coefficient for other pertinent X-rays in a carbon rich matrix, such as those for Fe, Al, Si, O, Cr, Sr, Ca, S - and how that coefficient corresponds to peak size? Double the coefficient -> half the peak or what?

Also, would your monte carlo simulations take into account this absorption and the ionization potential? I always wanted to ask you to m-c-sim hypothetical LaClede primer as per specification. What would you need for that as input? Mass proportions of the elements in the mix? Molar proportions?

Thanks!!

 

[The Almond]

Thanks a lot!

But...

I don't have a feel for how significant this is. Can you explain this a little and put into perspective, for example by giving examples of the mass absorption coefficient for other pertinent X-rays in a carbon rich matrix, such as those for Fe, Al, Si, O, Cr, Sr, Ca, S - and how that coefficient corresponds to peak size? Double the coefficient -> half the peak or what?

Also, would your monte carlo simulations take into account this absorption and the ionization potential? I always wanted to ask you to m-c-sim hypothetical LaClede primer as per specification. What would you need for that as input? Mass proportions of the elements in the mix? Molar proportions?

Thanks!!

To give you some perspective, the MAC for Fe K (6.36 keV) x-rays in a carbon rich matrix is 8. Even for wimpy Si K (1.74 keV) x-rays, the MAC is only 416. Regarding peak size, that's a bit more complicated. The spectral response from the radiation is a combination of 3 effects:

1) Photoionization of the incoming electron
2) Absorption of the photon by the matrix
3) Detection of the photon by the X-ray detector

As you might guess, none of these are truly linear processes. Broadly speaking, the spectral response will vary with the base 10 logarithm of the MAC. Thus, the MAC's in the 10,000 range will absorb energy approximately 100 times more than those in the 100 range.

The system that we're describing, the epoxy and primer, is probably the best system on earth for absorbing N x-rays, and couple that with a detector that likely has a Moxtek ultra thin window (made out of carbon with a copper support grid), and your chances of seeing N drop from winning the lottery levels to flat out aint gonna happen.

Monte Carlo is advantageous because I can take into account all of the various complexities of the system without requiring an analytical solution. Consider the following situation: You are on a golf course, and you want to know the distance to the next hole. The only problem is, you don't actually know where the hole is, and you can't measure it directly. One potential solution is to take a drunk, blindfolded golfer (me, for instance, on any given Saturday), and have him hit golf balls. You watch the ball, and see if it disappears into a hole. He needs to hit thousands upon thousands of golf balls, but eventually you'll find the hole.

That's Monte Carlo in a nutshell. I fling an imaginary electron into a sample whose composition I know, and watch it bounce around. If it makes an x-ray, I follow the x-ray out of the specimen into the detector, and then count how many times that happens. The problem is that I'm making a lot of assumptions about the system we're proposing. I'm using my x-ray detector, in my electron microprobe, with my geometry. All that you can supply is the composition in weight percents, which is helpful, but it's not the whole story. That's kinda why I was initially hesitant to just model the primer paint. I'm happy to do so in order to illustrate the major features of the spectral response, but it's not exactly a slam dunk, so to speak.

 

[Oystein]

To give you some perspective, the MAC for Fe K (6.36 keV) x-rays in a carbon rich matrix is 8. Even for wimpy Si K (1.74 keV) x-rays, the MAC is only 416.

What is this proximately dependent upon - the keV value? The electron level (K, L...)? A combination? Can you maybe link me to a table or a formula?

The XEDS graphs we have seen so far have nice tall peaks for O, whose K-level clocks in at "whimpy" 0.537 keV. What's the MAC there?

Regarding peak size, that's a bit more complicated. The spectral response from the radiation is a combination of 3 effects:

1) Photoionization of the incoming electron
2) Absorption of the photon by the matrix
3) Detection of the photon by the X-ray detector

Wow, this sounds like techno-babble to me I.o.w. it's pretty over my head.

As you might guess, none of these are truly linear processes. Broadly speaking, the spectral response will vary with the base 10 logarithm of the MAC. Thus, the MAC's in the 10,000 range will absorb energy approximately 100 times more than those in the 100 range.

Urr wait - what you describe there is linear: 10,000 = 100*100 - isn't it? You surely mean 3 times more (1+log(10,000/100))?

The system that we're describing, the epoxy and primer, is probably the best system on earth for absorbing N x-rays, and couple that with a detector that likely has a Moxtek ultra thin window (made out of carbon with a copper support grid), and your chances of seeing N drop from winning the lottery levels to flat out aint gonna happen.

Given your track record, I believe you, but somehow I don't know yet what to write in my paper about this...
Maybe I'll ask you to join us in writing it and provide the part about XEDS details.

Monte Carlo is advantageous because I can take into account all of the various complexities of the system without requiring an analytical solution. Consider the following situation: You are on a golf course, and you want to know the distance to the next hole. The only problem is, you don't actually know where the hole is, and you can't measure it directly. One potential solution is to take a drunk, blindfolded golfer (me, for instance, on any given Saturday), and have him hit golf balls. You watch the ball, and see if it disappears into a hole. He needs to hit thousands upon thousands of golf balls, but eventually you'll find the hole.

That's Monte Carlo in a nutshell.

Yes, I broadly understand what a M-C sim is, I did some back in my grad school time, doing MCs on queuing problems in Management Science.

I fling an imaginary electron into a sample whose composition I know, and watch it bounce around. If it makes an x-ray, I follow the x-ray out of the specimen into the detector, and then count how many times that happens. The problem is that I'm making a lot of assumptions about the system we're proposing. I'm using my x-ray detector, in my electron microprobe, with my geometry. All that you can supply is the composition in weight percents, which is helpful, but it's not the whole story. That's kinda why I was initially hesitant to just model the primer paint. I'm happy to do so in order to illustrate the major features of the spectral response, but it's not exactly a slam dunk, so to speak.

Alright, here it is, for the red layer (LaClede paint as specified, with some hopefully reasonable assumptions about the epoxy):
C: 48,67%
O: 21,12%
Fe: 10,96%
H: 6,80%
N: 6,68%
Si: 2,54%
Al: 2,44%
Sr: 0,49%
Cr: 0,29%

Haven't looked at the chemistry of oxidized steel surfaces yet so I don't have the numbers for the gray layers (basically Fe + 1% Mn, + O).

 

[tsig]

Context?

It shows that even the NIST, who were aware of the LaClede primer, considered the Tnemic primer paint to be the one to use for their testing.

It also shows that the NIST, in their professional opinion, unlike your amateur opinion, did not consider the LaClede primer to have a formulation worthy of serious attention.

MM

I thought you didn't agree with NIST.

 

[tsig]

Well I think Ivan did imply that.



Makes nice copy. The old BYU football stadium paint mockery, but as I've shown, all you have is an email extract produced by an anonymous source that flies in the face of an on-the-record scientist who claim he tested genuine WTC paint. And also ignores the possibility that the Dr. Jones knew that the BYU stadium used the identical paint formulation. I don't know about the credibility of your anonymous source, (well I do know he is an Official Story supporter), but I'm betting my money on the guy with the professional credentials and the cajones to put his real name out front-and-center.



??

A very real possibility of what exactly?

That the WTC floor trusses were coated with a primer paint that immediately ignited during an office fire?

Time to stop the desperate speculation Oystein and provide some scientific test proof to support your "amazing" claims!
MM

Really?

 

[tsig]

Ah yes kaolin, a substance found just about everywhere.

I'm an atheist, and while I'm willing to agree that most religious-based writings are crap, I'm not such a fool as as to hold a belief that all religious scientists are incompetent.

Something tells me that even if Dr. Harrit had tested the red chips in an inert gas or a vacuum, that you still would find reason to disagree with a thermite finding.
Ah right. You covered that scenario by demanding he prove conclusively that thermite was used against the WTC structural elements. Elements that have largely been turned into refrigerators now.

And damn, because we have no audio recordings that conclusively show the use of conventional explosives...

And heaven forbid anyone, especially Richard Gage, ever show you a cardboard illustration, when only a full scale built-to-the-same-specifications-and-appropriately-aged model of WTC will suffice for you.

Sweet dreams lefty.

MM

What would that something be?

 

[tsig]

Obviously?

Their objective was to use the most commonly occurring steel primer paint, Tnemec primer paint.




I am saying that the NIST considered the LaClede primer paint to be inconsequential to their test purposes and likely they never suspected, unlike you apparently, that it was volatile at 430 C.
Since the NIST were aware of the different formulations between the two paint types, I would have to believe that they would have reacted had they noticed anything unusual.

Or do you think a primer paint that ignites in the middle of their heat testing range would not be deemed unusual?

MM

This isn't helping your case.

 

[Oystein]

This isn't helping your case.

Do you feel that your one-liners will prompt MM to help our case, which is still "Find and prove the Origin of the paint that was found as red-gray chips"?

 

[leftysergeant]

4) Observed thickness of red layer is smaller than specified (1mil +/- 0.2 mil = 25µm +/- 5µm). Did LaClede cut corners? Or did the layer shrink during 40 years?

Are we certain that the clowns measured correctly?

All four questions could be answered quite definitively by experiment if we could find a nice, unburned piece of actual floor joist.

Given that they fell mostly straight down into the foot print of the towers (and thus into the great smoldering pire pit, I do not hold out any great hope of collecting a large sample at this late date.

ETA: I just did a little Googling on titanium. Titanium isopropoxide is used as a catalyst in "Sharpless epoxidation." Whether that has anything to do with the epoxy in LaClede is a little beyond me right now, but maybe Ivan can add something on that point.