I actually managed to find a professor of fire chemistry who agreed to talk with me about the question of the formation of iron-rich spheres in the WTC fires. And he even referred me to a SECOND fire chemist who had some more answers. This should lay to rest the question of whether iron-rich spheres can be created at ONLY the bulk melting point of iron.
He explained that a typical office fire might burn at, say, 1100-1800F bulk
temperature. But, the adiabatic
flame temperature, the theoretical upper limit of burning materials starting at room temperature, is much higher. For hydrocarbon fuels it is 3600-4500F, for wood it is almost 3500F. If you use laser to focus in on the flame temperature as they did in these published papers (below), observable local instantaneous temperatures can actually approach the adiabatic temperature. Any tiny flake of rust exposed to such micro near-adiabatic temperatures could easily hit the melting point of iron. We can expect huge blobs of melted iron when the bulk temperature exceeds its melting point, but only tiny iron micropsheres when small flakes of rust fly around in these smaller areas where the temperatures are much higher. And that is exactly what was found at the World Trade Center site!
Adiabatic flame temperature: for hydrocarbon fuels it is 3600-4500F, for wood it is almost 3500F. (Wikipedia) Laser focuses in on the flame temperature. Any tiny flake of rust exposed to such micro-scale near-adiabatic flame temperatures could surpass iron’s melting point.
(see figure 4)
Barlow, R. S. and Frank, J. H., Proc. Combust. Inst. 27:1087-1095 (1998)
Barlow, R. S., Frank, J. H., A. N. Karpetis, and Chen, J.-Y., "Piloted Methane/Air Jet Flames: Scalar Structure and Transport Effects," Combust. Flame 143:433-449 (2005).
Schneider, Ch., Dreizler, A., Janicka, J., "Flow Field Measurements of Stable and Locally Extinguishing Hydrocarbon-Fuelled Jet Flames," Combust. Flame 135:185-190 (2003).
You can also use as a reference showing temperatures in the vicinity of 2000C a paper out of Purdue:
Thariyan, MP, Ananthanarayanan, V. Bhuiyan, AH, Naik, SV, Gore, JP, and Lucht, RP, “Dual-pump CARS temperature and major species concentration measurements in counter-flow methane flames using narrowband pump and broadband Stokes lasers,” Combustion and Flame, 150, 1390-1399, (2010).
The second fire chemist said, "I am a physical chemist, and some of my recent research has focused on the reduction-oxidation properties of iron oxides at high temperature..."
"There are several oxides of iron, and each has a different melting point. The lowest of these is FeO (also called wüstite) at ~ 1377 C, i.e. lower than the melting point of Fe (~ 1535 C). Rust (Fe2O3, or hematite) decomposes/reduces around 1566 C in air, but under partially reducing atmosphere this can shift to much lower temperature. For instance, in an incompletely-burned fuel fire with hydrocarbons and CO, reduction of iron oxides to metallic iron can occur << 1000 C (I’ve seen it at 600 C under a dilute hydrogen stream). Stating from hematite the reduction sequence would be as follows: Fe2O3 – Fe3O4 – FeO – Fe, thus as rust is reduced it reaches FeO (lowest melting phase) before Fe. Could it be that the “iron rich” microspheres are FeO, or a mixture of Fe and Fe3O4 from the disproportionation reaction of FeO (4FeO = Fe + Fe3O4)?"
Other people I have talked with have been unable to provide proof of this, but collectively they have proposed other possible explanations and hypotheses about how iron-rich microspheres could be created at less-than-melting temps: There are processes by which rust flakes can be made to congeal into iron-rich microspheres at lower-than-melting temperatures in an office fire (such as with CO in an oxygen starved part of a fire, sintering, annealing, eutectic mix of aluminum and steel, other processes) that can cause these iron-rich microspheres to form. Myn concern is that some of these may have commercial or scientific applications that don't apply in regular office fires. But the near-adibiatic tempertures of highly localized parts of fires as measured by laser technology at Sandia National Lab (and the tiny rust flakes that turn into iron-rich spheres inside these areas) and for that matter, the micro-thermitic reaction that is created when rust and aluminum collide as they did in the 9/11 collapases, convinces med that thermite is NOT necessary to explain this phenomenon.