Scanning thermal conduction microscopy with dual cantilever resistive probe

The thermal properties of very small things differ from those of the larger objects with which we are familiar. For example, the size of a small object such as a carbon nanotube is comparable to the wavelength of the sound waves which transport heat ("acoustic phonons"). The thermal conduction of a carbon nanotube is therefore defined by quantum effects. At the same time these nanomaterials are increasingly used to make useful objects, such as composites, transistors and lasers. The thermal behaviour of these useful devices and materials is extremely important: A hot laser will be inefficient and will fail in a short time. A composite material will very often have an important thermal specification to meet, as well as being required to be strong, light and tough. The methods currently used to measure the thermal properties of materials at the nanoscale are inadequate.

This project is concerned with the development of a new technology for the measurement of thermal properties at the nanoscale. It follows on from the successful development of quantitative thermometers at Glasgow which are used in "Scanning Thermal Microscopes" (SThM). These nano-thermometers incorporate a small resistance thermometer on the end of an "Atomic Force Microscope" (AFM) cantilever probe. Although they have proved to be very capable as a means to measure temperature at the nano-scale, they are bad at measuring thermal conductivity. This is because measurement of thermal conductivity requires both a thermometer and a heater, and the requirements of the two are very different: A thermometer must not get hot (or it would change the temperature of the sample) and it must couple heat out of the sample as weakly as possible (or it would cool the sample down). The requirements for a heater are the exact opposite. It must be capable of becoming as hot as necessary and also be strongly coupled to the sample, so as to inject a significant power into the sample. The solution proposed is to use a similar type of thermometer to that previously described, and to integrate it with a newly - designed heater, situated at a known distance from the thermometer. Both the thermometer and heater need to be held in contact with the sample as they are scanned over the it to measure changes in thermal properties. This is accomplished by use of a micromachined AFM cantilever with two tips, which maintains a constant lateral separation between the heater and thermometer elements whilst permitting them to "ride" over bumps independently. The materials used to fabricate heater and thermometer, along with the shape of the tip used to couple them to the surface will be independently optimised.

A sensor is not, by itself, a measurement system. The measurement must be made in a controlled environment. In this case a vacuum is used to minimise the surface water film which degrades spatial resolution and to prevent direct coupling of heat from heater to thermometer by conduction through the air. A system for calibration is required to ensure that measurements are quantitative, not just pretty pictures. Electronics needs to be constructed to control the power input to the sample and to make a sensitive measurement of the resulting temperature change. Lastly the sensor must be brought into controlled contact with the sample and moved around, so as to map the changes in thermal properties of the sample as a function of position. This project seeks to accomplish all of these things. Development of the measurement system itself will be driven by the measurement of an important set of three materials which are expected to have extreme thermal properties. These are nanocrystalline diamond, carbon nanotubes and graphene films. All of these materials are made from carbon, but have different dimensionality. All of them are used in practical applications in which their thermal properties are important.

Measurement technology is one of the best ways for "small science" to make a large impact in the wider community.

The proposal will develop a new measurement technology. The use of microfabrication technology to implement the sensor will facilitate its rapid incorporation by manufacturers of scanned-probe microscopes, since the majority of the system is common to all scanned probe techniques. In this case the benefit will be one of a direct increase of sales of instruments.

New methods for the interrogation of materials are likely to confer extremely broad benefits to the commercial sector. In this case, anyone who develops materials or systems in which the thermal properties of small systems are important is likely to benefit from the ability to measure, and hence optimise, thermal conductivity at the nanoscale. This includes all active sensors (gyroscopes, accelerometers, implantable biosensors), where the devices used have voids or significant heterogeneity which make thermal (heatsinking) design difficult. MEMS structures are often thermally actuated, in which case the need to determine the thermal properties of the layers used is even more critical. Another large class of devices in which thermal conductivity measurement on small length scales will be important is that of power constrained electronic systems. These include high performance processors, high frequency and power semiconductors, all of which are generally operated in a thermally limited regime. Optoelectronic devices such as those used in telecommunications, spectroscopy and remote sensing are all exquisitely sensitive to thermal conduction variation, either because the power output is thermally limited, or because temperature control is used for tuning, modulation etc. The need to generate power from alternative energy sources is strategically and economically important. Modern high performance nanostructured solar cells used with concentrating optics are limited by maximum permitted temperature rise. Thermoelectric materials used in energy scavenging or geothermal power generation are also likely to use nanostructuring separately to optimise thermal conductivity and electrical conductivity.

The ability to optimise thermal performance of devices and systems at all length scales will enable more rational and efficient design. For example heatsinking may be reduced at the same time that local temperature rises are minimised: Systems will be smaller, cheaper to make and more reliable. This gives a straightforward commercial advantage to any system manufacturer. The ability to measure the thermal properties of new materials, which may not be available as bulk (e.g. strained epilayers) will allow their early incorporation into new products: By measuring elemental properties and then simulating performance, instead of building "all-up" systems and testing them, the time to development (and hence market) is greatly shortened.

Direct exploitation of prototype devices by collaborators will occur before the end of the project. Since a manufacturing and distribution channel for SThM thermometer probes already exists the time taken to implement commercial measurement systems will be limited solely by the time taken by the microscope OEM, typically two years after decision to develop. This research is the fundamental gate to the development of quantitative thermal conductivity instrumentation at the nanoscale worldwide; As far as is known by the proposers no comparable system is being developed anywhere else.

The development of instrumentation and the characterisation of materials are both highly useful skills in both the academic and commercial sectors. By providing a developed thermal conductivity subsystem it is hoped that collaborators will be encouraged to develop their own solutions for measurement. The proposed work is likely to drive the development of these valuable skills in several institutions.