The areal data density of magnetic recording hard disk drives (HDDs) increases year by year, following a trend similar to Moore’s law. However, the increase is not unbounded and there are some physical limits. As the density increases, the size of each magnetic grain shrinks. Finally the magnetic grain will be no longer thermally stable due to what is termed superparamagnetism. Above this point, the magnetic storage would be not reliable because the magnetic grains’ orientations fluctuate randomly. To increase magnetic recording density to more than 1 Tb/in2 and break this limit, heat assisted magnetic recording (HAMR) is proposed. In HAMR systems, a more thermally stable magnetic material, one with higher coercivity, will be used as a recording layer. But the coercivity of this material at room temperature is so high that it is difficult for the writer to switch the magnetic orientation with current magnetic transducers. However, the coercivity drops sharply if the temperature is raised close to the Curie temperature. In HAMR systems, a laser is proposed as the means to heat the disk to the Curie point. Simultaneously the magnetic field is applied from the writer to switch the magnetic bits. The success of the magnetic switching is very sensitive to the media temperature [2]. If the temperature is too low compared with the Curie point, it will not be able to write any information into the media. Conversely, heating the media over the Curie point requires more energy and may bring a greater challenge for the head disk interface (HDI). It is very important to understand the local temperature distribution during the laser heating and to calibrate the laser power input for HAMR writing. Some work has been done to evaluate the temperature increase using both numerical and experimental methods [3, 4]. Tagawa et.al. observed the disk refractive index change during laser heating and compared it with the change under conventional oven heating. This is a good method to calibrate the laser power and get the average temperature but it has some limitations for getting the accurate temperature distributions because of the averaging effect for the refractive index measurement by ellipsometry.
