Seems to be quite normal. I don't know about transition metals, but normally it takes good time to bring any solid sample to equilibrium state after pressure applied; say, several seconds if not minutes. It's easy to provide this equilibrium with DACs, but quite hard with SW.

.

Thanks for your answers,it's proposed that the melting phenomena detected in DAC may be not ture,might a plastic flowing.

Infering melt from speckle patterns at the surface of DACs has given a number of discrepancies in other metals, but Boehler insists on using it.

Hi Kervin:

Your "question" is actually a statement - a statement relying on diminishing evidence.

There is no evidence that a crystalline material melts in different ways. The perceived differences in melting temperatures of 3d metals under high pressure have to be investigated under consideration of the applied experimental techniques. To my knowledge, these techniques have been used:

(1) shock-compression, Laser or projectile impact with analysis of tempearture;

(2) optical, analysis of speckle patterns;

(3) x-ray diffraction, disappearance of diffraction peaks;

(4) synchrotron Mossbauer spectroscopy, disappearance of nuclear resonant signal.

Each of these methods has strenghts and weaknesses. Method (1) is dynamic and the determination of pressure from the Hugoniot is very accurate. Temperature determination however is difficult and far less accurate. Additional problems in the comparison of results occur because results need to be reduced to the relevant temperature which is usually lower than the temperature in the shock-compression experiment.

Methods (2-4) are all static experiments, i.e., pressure and temperature are constant over a time scale of typically seconds. The determination of temperature is very reliable but pressure is often difficult to determine with comparable accuracy.

Method (2) uses the change in speckle patterns on the Laser-heated surface of the sample to infer melting and gets the lowest melting temperatures. This might be due to effects specific to the surface such as pre-melting and unusually high diffusion.

Method (3) studies the volume of Laser-heated sample in the focal spot of an intense x-ray beam usually at a synchrotron radiation facility. When melting occurs in this volume x-ray diffraction peaks that arise from spatial order disappear. More advanced experiments monitor the simultaneous increase in difuse x-ray scattering from the melt.

Method (4) is new (paper attached) and monitores the atomic motion using a particular x-ray scattering technique called synchtron Mossbauer spectroscopy. Contrary to method (3) this approach is sensitive to motion of the atoms rather than spatial order.

At present there is no overlap in p-T space between reported results of dynamic and static methods. This may change in the future. Also recent extrapolations of melting curves from static methods to the higher p-T region of dynamic methods do not support disagreement.

Recently, I'm interested with Fermi surface nesting under pressure. There might be a connection between the nesting and structural phase transition, in my opinion.

Actually, there some papers stated the opinion.

See: http://www.sciencemag.org/content/340/6131/464

A seemingly understated reason, sometimes verified by measurements, for why (in this context) DAC metal melt temperatures are lower than dynamic compression melt temperatures is because of chemistry effects. Metal carbide products form at the surface of the diamond culet. Sometimes carbide and perhaps other products (depending on thermal insulation materials e.g., NaCl, KCl etc.) have significantly lower or even higher melt temperatures compared to the parent metal. Diffusion of carbide products into the DAC cavity can be rapid: this chemistry effect was elucidated by LH-DAC measurements made at ESRF resulting in a PRL publication. The effect of chemistry on the observed melting temperature of laser heated DAC encapsulated metals should be acknowledged in all future publications.

To state the obvious, no experimental method can be thought of as a panacea. Laser speckle, x-ray diffraction, and perhaps the relatively new synchrotron Mossbauer spectroscopy (SMS) have only been used to signal that a transition has occurred. These techniques may signal that a solid-crystalline to solid-amorphous structural transition occurred. They may even signal indirectly that a chemical reaction occurred. If experimentalists using these techniques further increased temperature (at a fixed pressure) then one of three possibilities would be observed: 1) no further transition occurs thus improving the confidence that a solid-liquid transition occurred (assuming no chemistry), 2) a second transition occurred indicating an amorphous-liquid transition occurred (assuming no chemistry occurred), 3) two more transitions occurred one being a solid-solid and the second being solid-amorphous with the third being amorphous to liquid (assuming no chemistry occurred).

It appears that most experimentalists using these DAC diagnostic techniques do not report increasing temperature further beyond a clear transition to, for example, verify that additional transitions do not occur, all the while assuming that chemistry did not occur.

Shockwave measurements are quite good at conveying when the volume of a material changes; however, sometimes even this method can be blind. For example, shock work on formic acid (FA) revealed no shock-induced volume change –contrary to DAC measurements at relatively low T. Theory then elucidated that indeed a crystalline to amorphous (polymeric) transition does occur in FA (agreeing with DAC measurements); however, there is little change in volume. There are now very aggressive efforts in this community to make temperature determinations though the intrinsic physics issues (e.g., knowledge of emissivity) will always remain.