The paper is at http://www.icr.org/research/icc03/pdf/Helium_ICC_7-22-03.pdf. It's been referenced previously on FR, and I have read it.
The creationists discuss three data sets. The first is an old Russian data set of radiation-damaged zircons, where they don't even know how to interpret the data. That's useless, and I'll ignore it. The second is the Nevada data of Reiners et al that I cited. From those data, Reiners obtained an activation energy, which roughly quantitates how easy it is for helium to diffuse. The higher the activation barrier, the slower the diffusions, and it's a very non-linear relationship. The activation energy is obtained using something called the Arrhenius equation, where you plot the natural log of the diffusion constant versus reciprocal temperature, and find the slope, which is the negative of the activation energy divided by the gas constant R. The guys who actually took the data - Reiners et al - claimed 44 kcal/mole. I extracted a slope from Figure 5 of the creationist paper, in which they reproduce Reiners et al's data, applied the usual Arrhenius equation, and got 43.5 kcal/mole, in agreement with the original authors. However, the creationists claim they get an activation energy of 34.4 kcal/mol, and go on to claim the data at the bottom part of the curve, where they have no quoted error, is as low as 29.4 kcal/mol! That's what they need for their young-earth model!
So, wondering why, I went to their appendix C - the report from the 'independent expert' which contains the results for the zircon data for which the creationists claim an activation energy of 34.4 kcal/mol. The expert says, and I quote "The first 14 steps lie on a linear array corresponding to an activation energy of ~ 46 kcal/mol and a closure temperature of ~183ºC assuming a cooling rate of 10ºC/Myr". That's in excellent agreement with Reiners et al! The much lower activation energy was obtained only when he temperature-cycled the zircons. He says the change might be due to temperature damage of the zircons by the experiment, or by anomalous retardation of helium release in the original sample. In either case, the slope that's relevant to helium retention over geological time is the original one, before the crystals has been altered by heating. The creationist's use without comment of the data from crystals that are temperature cycled, without acknowledging that without temperature cycling, the diffusion constants are orders of magnitude lower and in agreement with established work, looks to me like scientific fraud.
Now aren't you glad you baited me into giving the paper a more careful read?
Time Scales and Heterogeneous Structure in Geodynamic Earth Models
Hans-Peter Bunge,* Mark A. Richards, Carolina Lithgow-Bertelloni, John R. Baumgardner, Stephen P. Grand, Barbara A. Romanowicz
Abstract: Computer models of mantle convection constrained by the history of Cenozoic and Mesozoic plate motions explain some deep-mantle structural heterogeneity imaged by seismic tomography, especially those related to subduction. They also reveal a 150- million-year time scale for generating thermal heterogeneity in the mantle, comparable to the record of plate motion reconstructions, so that the problem of unknown initial conditions can be overcome. The pattern of lowermost mantle structure at the core-mantle boundary is controlled by subduction history, although seismic tomography reveals intense large-scale hot (low-velocity) upwelling features not explicitly predicted by the models.
What sort of ethics does a scientist have, when he can on the one hand author papers claiming the earth is no more than 6000 years old, and on the other author another that discusses geological processes over a 150 million year time-scale? Which of the two papers he co-authored does he not actually believe?