The Anatomy of a Microcalorimeter

Previous page:

What Is Microcalorimetry?

Using standard photolithography on a Si chip, my research group at UCSD fabricated a thin-film microcalorimeter capable of measuring microgram samples from 1-800 K. Such a device is far more sensitive than any commercial counterpart, and over a broader temperature range. For micrograms of metallic samples, for instance, there is just no other way to measure the heat capacity at and above room temperature.

We made (and probably still make) them at the Berkeley Microfabrication Laboratory in the EECS Department at UCB. This facility is a clean room for silicon wafer processing.

The details of this device are published: Denlinger et. al., "Thin film microcalorimeter for heat capacity measurements from 1.5 to 800 K," Rev. Sci. Instrum. vol. 65, 946 (1994).

Below is a scaled-down summary of the description of the calorimeter that appears in the above paper. I intend it to be readable by people with an undergraduate knowledge of physics and chemistry. I promise that in simplifying the language, I will not say anything that is scientifically untrue, nor will I omit necessary facts. If you want a rigorous, high-level explanation of the microcalorimeter and our data-- well, that's why we wrote the paper.

Above are front- and side-view diagrams of our device. It is 1 square centimeter in size. The key feature to notice first is the white space which is 0.5 cm by 0.5 cm (ie, the central region, more clearly visible on the side view). This is a very thin membrane made of amorphous silicon-nitride, something like a drumhead across the Si frame. It's only about 1200 atoms thick, and appears transparent to the eye like a little window. The reason it is so important is that it serves to thermally isolate everything on it from the Si frame without providing alot of thermal mass itself. I'll come back to this point later, but for now let me just mention that the biggest problem in calorimetry is reducing the "addenda"-- the extra contribution to the heat capacity due to the measurement apparatus itself, which can swamp a tiny signal from the sample. The membrane is what makes our device so incredibly sensitive.

The zigzag features in the middle are made of platinum (Pt) -- one is a heater, the other a thermometer. There is also a matching Pt thermometer on the Si frame. The lines connecting the Pt features are leads made of a gold-paladium (Au-Pd) alloy. The small rectangular features which, like the Pt, have one part on the membrane and a matching part on the frame, are also thermometers. They are made of either polycrystalline silicon with boron atoms implanted in it or a niobuim-silicon (Nb-Si) alloy, and are also linked by Au-Pd leads. All of these features are thin films evaporated directly onto the device surface (ie, they are an integrated circuit).

The small square in the center which borders the zigzag represents the sample to be measured. It, too, is evaporated onto the device, but it goes on the back. The side view makes this clearer-- we deposit through a "shadow mask" to block out all the area except where we want the sample to go. (Such a procedure is not possible for the small, close-together features on the front; to fabricate those, we use photolithography, which I won't go into in detail). The sample has to go on the back so it doesn't short out the heater and thermometers on the front. Since the membrane is so thin, the sample is still in thermal contact with the electronics on the front, just not electrical contact.

Once all the layers are on the device, we attach wires to the contact pads (the squares around the edge of the device-- there are a total of 11) and mount the device in a clean chamber which will be evacuated, then cooled or heated to the temperature range we want to measure. As mentioned above, the device operates from 1-800 K (-272 to 527 C); most of our measurements span the range from 4.2 K (-269 C) to 250 C.

Next pages:

Return to the Main Microcalorimetry Page