Ivan Kminek
Muse
- Joined
- Jun 15, 2011
- Messages
- 906
To Oystein, Chris, Sunstealer, Almond, Basquearch and anybody who may be interested:
(In advance: Sorry for very long contribution)
Yesterday evening, I spent another 3 hours looking at Jim Millette’s results and here are my further remarks. Remember that I am layman in XEDS etc. And it is fully possible that I misunderstood or mismatched something.
My basic assumptions were: according to our discussion in the Paint thread, Laclede paint should prevail in the WTC dust as “red-gray chips”, but Tnemec paint could be also present and other “minor” red paints as well. Problem is that Jim Millette was not informed about the second major WTC red primer we “discovered” in NCSTAR 1-6b report and he basically thinks that all red chips may be the same paint (similarly as Harrit et al thought that all red chips are nanothermite).
Jim Millette does not claim this, but his conclusion is: “The red/gray chips found in the WTC dust at four sites in New York City are consistent with a carbon steel coated with an epoxy resin that contains primarily iron oxide and kaolin clay pigments.” In fact, some chips still can be e.g. Tnemec paint which does not contain epoxy, but alkyd-linseed binder instead.
My main question was: could be the most of red chips found by Jim the chips of Laclede paint (as we basically expected) or are there some clues that some chips can be Tnemec or other paint? The most important here are XEDS spectra, but as I was afraid, their interpretation is not unambiguous, because some peaks can be attributed at least partially to surface contamination, the heights of main peaks can vary for numerous reasons etc.
Initially, I planned some “detective work” in sense: let me choose all chips which contain simultaneously epoxy binder, typical kaolinite stacking platelets and nanosized iron oxide: all such chips must be Laclede paint, no matter how their XEDS look (what minor elements which are not expected in Laclede are detected in the spectra). This could be good beginning for further analysis. But, there are not enough data in the paper to start with such elimination/confirmation of “non-Laclede paint”.
(Btw, I have to complain here that notation of Jim′s samples is quite complex, not really clear and not really unified. )
Lately, I decided to establish something like XEDS “fingerprints” and “markers” of Laclede and Tnemec paints.
XEDS fingerprints:
It’s rather ironic, but I’m forced to regard XEDS spectra of freshly-cut Bentham chips (a) o (d) as a “fingerprint” of Laclede paint, since they very well correspond to the Almond’s Monte Carlo simulation of the spectra and no trace elements attributable to contaminants are present.
As for Tnemec, I have still to consider the XEDS spectra taken by Steven Jones and spectra of MEK chip in Bentham paper (see Oystein’s blog) as a kind of “Tnemec fingerprint”, although they are apparently too big peaks of calcium in them (probably because of quite massive contamination with calcium stuffs).
Looking at the XEDS spectra of red layers it is clear that even the ratios between the main peaks of Fe and Si/Al greatly vary, which is normal in many cases since XEDS probe was intentionally “focused” to very small areas in which some pigment prevails. But, the identification of the paint is even more difficult because of this.
XEDS Markers:
I think that there is one very good “marker” of Laclede paint in the XEDS: Si and Al peaks are roughly the same (because Si and Al with similar atomic numbers are present in equimolar ratio and in the same pigment particles, i.e. kaolinite). I will call this marker “Si/Al”.
On the other hand, Si peak in Tnemec paint should be substantially higher than Al peak, since silicon compounds strongly prevailed in comparison with Al compounds (mostly “calcium silicates and aluminates”, according to specification). Another important Tnemec marker can be seen in any zinc peak (from zinc chromate) and calcium peak should be quite large (but not as so big as in uncleaned “MEK chip”), whereas it should be missing in the clean Laclede paint. Concerning chromium (if any detected), it was present in both Laclede and Tnemec paint, so it is not element good for paint identification. Notably, some magnesium was present in Tnemec (in talc), but not in Laclede. But, just magnesium peaks ”meaning” is rather questionable and unclear for me. I will not consider some trace elements like Ti, since its peaks are always very little or barely visible in XEDS spectra.
Comments on some data, as regards attempts to distinguish between Laclede and Tnemec chips:
In the further text, I will comment what I see in some relevant figures and spectra, starting with Appendices which contain the whole sets of available data.
Here are the notations of dust samples from 4 locations in Manhattan. Chips separated from those samples using magnet have basically the same notations (with the number of particle added), in which I just deleted originally present “pre-numbers” 9119 (number of this project of MVA).
MVA # Date Collected Sample Location Map No.
4808-L1616 28 September 2001 22 Cortlandt St. 1
4795-L1560 22 September 2001 Murray & Church St. 2
5230-M3451 15-16 September 2001 49 Ann St. 3
9119-X0135 07 October 2001 33 Maiden Lane 4
Appendix C
It contains FTIR spectra confirming the presence of epoxy resin and kaolinite for samples:
4795/1560-red-atr-0001
4808/1616-red-atr-0001
5230m3451-red-atr-0001
9119X0135-red-atr-0001
We have only these 4 spectra for further discussion, with no possibility to attribute them to the individual chips (at least I think so). At least we know that epoxy binder was found in chips from all 4 dust samples.
Appendix E
It contains data after low temperature ashing, which released pigment particles. Kaolinite and iron oxide particles, basically consistent with pigments in Bentham chips (a) to (d), were found in samples:
4795/1560-2
4808L1616-2-60KX
4808/1616-2a
5230m3451
F9119x0135
Appendix F
It contains micrographs taken on ultrathin “plates” of red layers cut by microtome on chip(s) (?) from sample F9119x0135. Again, it is a proof of kaolinite and iron oxide in this one sample. Similar results follow from muffle ashing (400 degrees C for 1 h), see Fig. 17, sample 5230m3451.
Appendix D
It contains XEDS taken on cross-sections of red layers. I like this appendix
) Basically, it shows mostly spectra consistent quite well with Laclede paint. In some cases, tiny calcium peaks are marked. Here is some list with tentative attributions to paint:
9119L1560(1) Laclede
9119L1560(2) Laclede
9119L1560(3) Laclede
9119L1616(3) Laclede
9119L1616(1) Laclede
9119L1616(2) Laclede
9119-5230m3451B can be Laclede, but distinct Ca peak and some S (as contaminants?)
9119X0135(3) Laclede
9119X0135(1) Laclede, some Ti is visible
9119X0135(2) Laclede
Appendix B
It contains XDES spectra on red layer surfaces (mostly washed with water; btw is this procedure good enough for removing surface contaminants? I do not know).
I do not like this appendix, since I cannot make really clear conclusions as regards attributions to some paint.
Main problems I see as a layman (without proper experience):
- There is invariably some little peak of Mg, which should not be in Laclede paint; could it be contaminant?
- There is always little to medium peak of Ca, which does not belong to Laclede. Could it be contaminant?
Since there is a lot of XEDS spectra in this Appendix, just some summary. I will not consider spectra taken on unwashed surfaces of red layers (16 chips from Sample M3451, 49 Ann St.).
- Regarding Ca and Mg as possible contaminants, most of chips could be Laclede paint, namely because of “proper” Si/Al marker.
- There are some chips which can be considered as Tnemec because of proper Si/Al marker, medium Ca peak and visible Zn peak:
Sample L1560, particle 2
Sample X0135, particles 2, 3, 5, 6 (but for particle 5, Si/Al marker does not really correspond).
As concerns the main text:
Fig. 7 shows XEDS of three red layers from sample 5230m345, photographed on Fig. 6. Basically, spectra seem to correspond to Laclede paint, but Ca peaks are quite distinct (contaminant?). It is not clear if chips were washed.
My partial conclusions:
- Especially because of Ca and Mg peaks, which can come from contaminants, it is not possible to attribute many XEDS spectra in Appendix B unambiguously to some paint.
- Some XEDS spectra correspond quite well or very well to Laclede paint (especially in Appendix D), few spectra can be attributed to Tnemec (in Appendix B).
- Jim Millette made me happy
, since he proved epoxy binder in some chips, as well as kaolinite and nanosized iron oxide in some chips, which is an experimental proof of Laclede paint I “discovered”. On the other hand, the results of his study do not allow some clear “distribution” of samples between just Laclede and Tnemec paint
.
- Of course, also some other paints could be analyzed/found in the study, but I am simply not sure at present.
Thanks to anybody who was able to read at least partially this text
Any comments or help from experts (Sunstealer, Almond) are of course highly welcome.
(In advance: Sorry for very long contribution)
Yesterday evening, I spent another 3 hours looking at Jim Millette’s results and here are my further remarks. Remember that I am layman in XEDS etc. And it is fully possible that I misunderstood or mismatched something.
My basic assumptions were: according to our discussion in the Paint thread, Laclede paint should prevail in the WTC dust as “red-gray chips”, but Tnemec paint could be also present and other “minor” red paints as well. Problem is that Jim Millette was not informed about the second major WTC red primer we “discovered” in NCSTAR 1-6b report and he basically thinks that all red chips may be the same paint (similarly as Harrit et al thought that all red chips are nanothermite).
Jim Millette does not claim this, but his conclusion is: “The red/gray chips found in the WTC dust at four sites in New York City are consistent with a carbon steel coated with an epoxy resin that contains primarily iron oxide and kaolin clay pigments.” In fact, some chips still can be e.g. Tnemec paint which does not contain epoxy, but alkyd-linseed binder instead.
My main question was: could be the most of red chips found by Jim the chips of Laclede paint (as we basically expected) or are there some clues that some chips can be Tnemec or other paint? The most important here are XEDS spectra, but as I was afraid, their interpretation is not unambiguous, because some peaks can be attributed at least partially to surface contamination, the heights of main peaks can vary for numerous reasons etc.
Initially, I planned some “detective work” in sense: let me choose all chips which contain simultaneously epoxy binder, typical kaolinite stacking platelets and nanosized iron oxide: all such chips must be Laclede paint, no matter how their XEDS look (what minor elements which are not expected in Laclede are detected in the spectra). This could be good beginning for further analysis. But, there are not enough data in the paper to start with such elimination/confirmation of “non-Laclede paint”.
(Btw, I have to complain here that notation of Jim′s samples is quite complex, not really clear and not really unified. )
Lately, I decided to establish something like XEDS “fingerprints” and “markers” of Laclede and Tnemec paints.
XEDS fingerprints:
It’s rather ironic, but I’m forced to regard XEDS spectra of freshly-cut Bentham chips (a) o (d) as a “fingerprint” of Laclede paint, since they very well correspond to the Almond’s Monte Carlo simulation of the spectra and no trace elements attributable to contaminants are present.
As for Tnemec, I have still to consider the XEDS spectra taken by Steven Jones and spectra of MEK chip in Bentham paper (see Oystein’s blog) as a kind of “Tnemec fingerprint”, although they are apparently too big peaks of calcium in them (probably because of quite massive contamination with calcium stuffs).
Looking at the XEDS spectra of red layers it is clear that even the ratios between the main peaks of Fe and Si/Al greatly vary, which is normal in many cases since XEDS probe was intentionally “focused” to very small areas in which some pigment prevails. But, the identification of the paint is even more difficult because of this.
XEDS Markers:
I think that there is one very good “marker” of Laclede paint in the XEDS: Si and Al peaks are roughly the same (because Si and Al with similar atomic numbers are present in equimolar ratio and in the same pigment particles, i.e. kaolinite). I will call this marker “Si/Al”.
On the other hand, Si peak in Tnemec paint should be substantially higher than Al peak, since silicon compounds strongly prevailed in comparison with Al compounds (mostly “calcium silicates and aluminates”, according to specification). Another important Tnemec marker can be seen in any zinc peak (from zinc chromate) and calcium peak should be quite large (but not as so big as in uncleaned “MEK chip”), whereas it should be missing in the clean Laclede paint. Concerning chromium (if any detected), it was present in both Laclede and Tnemec paint, so it is not element good for paint identification. Notably, some magnesium was present in Tnemec (in talc), but not in Laclede. But, just magnesium peaks ”meaning” is rather questionable and unclear for me. I will not consider some trace elements like Ti, since its peaks are always very little or barely visible in XEDS spectra.
Comments on some data, as regards attempts to distinguish between Laclede and Tnemec chips:
In the further text, I will comment what I see in some relevant figures and spectra, starting with Appendices which contain the whole sets of available data.
Here are the notations of dust samples from 4 locations in Manhattan. Chips separated from those samples using magnet have basically the same notations (with the number of particle added), in which I just deleted originally present “pre-numbers” 9119 (number of this project of MVA).
MVA # Date Collected Sample Location Map No.
4808-L1616 28 September 2001 22 Cortlandt St. 1
4795-L1560 22 September 2001 Murray & Church St. 2
5230-M3451 15-16 September 2001 49 Ann St. 3
9119-X0135 07 October 2001 33 Maiden Lane 4
Appendix C
It contains FTIR spectra confirming the presence of epoxy resin and kaolinite for samples:
4795/1560-red-atr-0001
4808/1616-red-atr-0001
5230m3451-red-atr-0001
9119X0135-red-atr-0001
We have only these 4 spectra for further discussion, with no possibility to attribute them to the individual chips (at least I think so). At least we know that epoxy binder was found in chips from all 4 dust samples.
Appendix E
It contains data after low temperature ashing, which released pigment particles. Kaolinite and iron oxide particles, basically consistent with pigments in Bentham chips (a) to (d), were found in samples:
4795/1560-2
4808L1616-2-60KX
4808/1616-2a
5230m3451
F9119x0135
Appendix F
It contains micrographs taken on ultrathin “plates” of red layers cut by microtome on chip(s) (?) from sample F9119x0135. Again, it is a proof of kaolinite and iron oxide in this one sample. Similar results follow from muffle ashing (400 degrees C for 1 h), see Fig. 17, sample 5230m3451.
Appendix D
It contains XEDS taken on cross-sections of red layers. I like this appendix
) Basically, it shows mostly spectra consistent quite well with Laclede paint. In some cases, tiny calcium peaks are marked. Here is some list with tentative attributions to paint:9119L1560(1) Laclede
9119L1560(2) Laclede
9119L1560(3) Laclede
9119L1616(3) Laclede
9119L1616(1) Laclede
9119L1616(2) Laclede
9119-5230m3451B can be Laclede, but distinct Ca peak and some S (as contaminants?)
9119X0135(3) Laclede
9119X0135(1) Laclede, some Ti is visible
9119X0135(2) Laclede
Appendix B
It contains XDES spectra on red layer surfaces (mostly washed with water; btw is this procedure good enough for removing surface contaminants? I do not know).
I do not like this appendix, since I cannot make really clear conclusions as regards attributions to some paint.
Main problems I see as a layman (without proper experience):
- There is invariably some little peak of Mg, which should not be in Laclede paint; could it be contaminant?
- There is always little to medium peak of Ca, which does not belong to Laclede. Could it be contaminant?
Since there is a lot of XEDS spectra in this Appendix, just some summary. I will not consider spectra taken on unwashed surfaces of red layers (16 chips from Sample M3451, 49 Ann St.).
- Regarding Ca and Mg as possible contaminants, most of chips could be Laclede paint, namely because of “proper” Si/Al marker.
- There are some chips which can be considered as Tnemec because of proper Si/Al marker, medium Ca peak and visible Zn peak:
Sample L1560, particle 2
Sample X0135, particles 2, 3, 5, 6 (but for particle 5, Si/Al marker does not really correspond).
As concerns the main text:
Fig. 7 shows XEDS of three red layers from sample 5230m345, photographed on Fig. 6. Basically, spectra seem to correspond to Laclede paint, but Ca peaks are quite distinct (contaminant?). It is not clear if chips were washed.
My partial conclusions:
- Especially because of Ca and Mg peaks, which can come from contaminants, it is not possible to attribute many XEDS spectra in Appendix B unambiguously to some paint.
- Some XEDS spectra correspond quite well or very well to Laclede paint (especially in Appendix D), few spectra can be attributed to Tnemec (in Appendix B).
- Jim Millette made me happy
, since he proved epoxy binder in some chips, as well as kaolinite and nanosized iron oxide in some chips, which is an experimental proof of Laclede paint I “discovered”. On the other hand, the results of his study do not allow some clear “distribution” of samples between just Laclede and Tnemec paint. - Of course, also some other paints could be analyzed/found in the study, but I am simply not sure at present.
Thanks to anybody who was able to read at least partially this text
Any comments or help from experts (Sunstealer, Almond) are of course highly welcome.
Last edited:
