A New Laser Heating and Temperature Measurement System

The laser-heated diamond anvil cell enables measurements of material properties in situ at high pressures and temperatures. An ideal measurement consists of a sample subjected to uniform pressure throughout the sample chamber. However, due to the small size of the samples and the high thermal conductivity of the diamond anvils temperature gradients are always present. Therefore, the gradient must be measured to be able to interpret results from the laser heated diamond cell, and to evaluate methods employed to reduce temperature gradients. In order to characterize these gradients, we have designed and installed a new laser heating and temperature measurement system at UCLA. This system specializes in a precise, quantitative measurement of the radial temperature gradient. Our design is anchored by a 20 Watt diode-pumped Nd:YLF laser for heating and measures temperature by collecting spectral intensity information from the hotspot. The temperature and the associated gradient are measured simultaneously using two cameras: 1) a high speed CCD camera coupled with a 150 mm imaging spectroradiometer that measures temperature, 2) a high dynamic range CCD camera that measures two-dimensional intensity distribution, at a variety of chosen spectral ranges. Information from both cameras is combined to provide the two-dimensional temperature distribution of the hotspot. The spectral intensity of a Gaussian-shaped hotspot has a maximum in the center part of the hotspot, and the sample temperature and therefore spectral intensity decreases radially from the center as proscribed by the thermal properties of the experiment and Planck's law. The intensity is a much more sensitive indication of temperature change. For example, in the case of a hotspot with a 2000K peak temperature, a 15% decrease in intensity at a given wavelength (e.g. 500 nm) corresponds to a temperature drop of 20 K, a 1% decrease, well below most estimates of precision in temperature measurement of laser heated spots. Therefore, relative spectral intensity provides a more reliable estimate of two-dimensional temperature gradients in high P, T experiments. In addition, this method bypasses most additional errors induced by chromatic aberration (a wavelength dependent focusing effect) of the objective lens. As a test of the ability of this system to measure temperature gradients, we have measured the thermal diffusivity anisotropy of highly oriented graphite. A graphite sample oriented perpendicular to the basal plane will create an elliptical hotspot from a radially symmetric laser power input due to faster heat transport along the basal plane (D~1.10x10-3 m2/sec) compared with the c direction (D~ 4.6x10-6 m2/sec). Therefore, the ellipticity of the hotspot is a direct measure of the systems ability to measure thermal diffusivity. Side by side comparison of the high dynamic range camera and the imaging spectroradiometer shows that the high dynamic range attains better measurements of the temperature gradients of a graphite hotspot. For example, at 75% of the peak temperature, the high dynamic range camera measured an ellipticity of 2.8, while the imaging spectroradiometer was smeared out to 2.2. We will present preliminary results from the laser-heating system including calibration information new data on the high P,T behavior of a metals, silicates, oxides, and ices.