R. Shrestha, K. M. Lee and T. Y. Choi – University of North Texas
D. S. Kim – POSTECH
Due to the required experimental characterization complexity for the thermal conductivity measurement of thin films, accurate measurement of such materials has been difficult to achieve for thicknesses less than 1 µm. For example, conventional Fourier heater plate methods have proven inapplicable as the thermal resistance between the film and the plate is the dominating measured quantity, rather than the thermal resistance of the thin film itself . Furthermore, when large temperature sensors (smallest ~50 µm) are used, the conductive heat loss through the sensors may be larger than the heat flow through the thin film, making it difficult to measure the film thermal conductivity. In an effort to resolve these issues, laser source heating, rather than a heating strip, was employed in combination with a high sensitivity, high resolution glass-based micropipette thermal sensor. This laser heating and micro scale sensor combination allows for accurate measurement of ±0.01oC of temperature differences while minimizing heat loss through the sensing tip and eliminating thermal resistance issues between heating strips and thin film interfaces. Such thermal response measurements have previously been demonstrated using pipette based sensor technology [2-4]; however, these micropipette sensors have had limitations regarding measurement accuracy and reliability (e.g., deterioration of the sensor tip resulting from repeated use). The complex fabrication process, cost effectiveness and small thermo power of the sensors were also potential limitations.
FABRICATION OF THE SENSOR
In order to overcome such limitations, a novel glass micropipette sensor was developed to permit accurate measurement of thin film thermal conductivity. A pipette puller (P-97, Sutter Instruments) was programmed according to a known recipe for creating patch pipettes with a tip size of approximately 1 µm and 5 to 7 mm taper length. It was used to pull a thick-wall borosilicate glass tube, with 1.5 mm outer diameter and 0.86 mm inner diameter, into a micropipette. Similar pipettes are often used in various biological applications for injecting solutions into biological tissue. The pulled pipette was filled with a lead-free soldering alloy composed of mostly tin (Sn) through an injection molding process in conjunction with localized heating of the material. The injection molding was accomplished by mechanical pressurization (or pushing) of molten metal at the upper part of the pipette while simultaneously heating the lower part near the pipette tip with an electronic soldering gun maintained at around 250 oC.
Followed by the molding process, the pipette tip was beveled in order to remove any unwanted extruding metal. This step is particularly important to assure a smooth and continuous contact between the two metals in the formation of a thermocouple. Therefore, the BV-10 micropipette beveller (Sutter Instruments) which was designed for beveling micropipettes with tip diameters between 0.1 and 50 μm was used to sharpen and smoothen the pipette tip. After beveling, the pipette was cleaned with ethyl alcohol using an ultrasonic cleaner. Following the cleaning process, a sputtering technique was used to coat thin films of nickel on the outer surface of the glass and thus form a Ni-Sn alloy junction at the beveled tip. This process was successfully used to develop a thermal sensor with micrometer sized sensor tip (Figure 1). The size of the sensor in the study is only a few microns. Thus, the sensor can detect temperature distribution of a few micron sized area. Spatial resolution of the sensor is determined by the size of the sensor. Because of micron-scale size, the lost power transmitted through the sensor tip turns out to be negligible as compared to the absorbed laser power. This fact was confirmed by accurate measurement of thermal conductivity of a stainless steel thin stripe (not published).
CALIBRATION OF THE SENSOR
Calibration of the fabricated sensor was conducted with a water-filled chamber maintained at constant temperature with an accuracy of ±0.01oC. A high-precision digital thermometer and the fabricated sensor were immersed into the water bath to read out an actual temperature of the water. During the calibration process, the cold junction (Ni-copper(Cu) and Sn-Cu junction; Cu is used to be a lead wire from the voltmeter) of the sensor was maintained at a constant temperature (e.g. 24.5 oC) slightly above room temperature using another in-house isothermal block made of aluminum. This was done so that any additional unwanted thermocouple effects from the cold junction could be removed. The voltage generated by the sensor was recorded by a voltmeter (Nano Voltmeter, Keithley 2182) and compared to the temperature read by the thermometer. In this way, variation of voltage with temperature difference (ΔV/ΔT) was obtained for the thermoelectric power, i.e. Seebeck coefficient. The standard deviation in the voltage measurement was less than 0.018 μV which is equivalent to temperature rise of 0.002 oC, which is one order of magnitude lower than the temperature measurement accuracy of 0.01 oC.
APPLICATION OF THE SENSOR
As an application for the sensor, thermal conductivity measurements of a single walled carbon nanotube (SWNT) film suspended over a hole in a poly-carbonate substrate were made. The CNT film consisted of SWNTs randomly oriented in the plane of the film, without any matrix material (such as a polymer) being present. To produce a film of SWNT, a vacuum filtration method was employed which involves vacuum filtering a dilute suspension of nanotubes in a solvent over a porous alumina filtration membrane (Whatman, 200 nm pore size, 47 mm diameter). A schematic of the measurement setup is shown in Figure 2. Antireflection coated lenses were used to collimate and expand a ray of beam which was then diverted at 90◦ by a beam splitter toward the CNT film. A 20× super-long working distance objective lens (Mitutoyo MPlan Apo SL) converted the diverted beam to a Gaussian beam profile of 3 μm spot size on the CNT film. A class B laser at 532 nm was irradiated at the center of the suspended, 50-μm CNT film (thickness of 100 nm). The charged-coupled device (CCD) camera as shown in Figure 2 was used to obtain clear images of laser spot and the sensor tip. The inset in Figure 2 shows CCD images of (a) laser shined at the center of the film and (b) the pipette sensor positioned on the film.
The temperature difference at two radial positions was measured using the pipette sensor with tip size approximately 3.5 μm, and Seebeck coefficients of 5.67 μV/oC. The same procedure was repeated using a different pipette sensor with the same tip size and Seebeck coefficient of 7.44 μV/oC, to enhance the measurement reproducibility. The voltages at the two different locations, approximately 8 μm apart in radial directions, were measured with known Seebeck coefficients, and then converted to temperature difference. The power absorbed by the CNT film was determined to be approximately 75% of the total irradiated power. The absorbed power was calculated according to the relationship Pabsorbed= Pincident − Preflected − Ptransmitted. The thermal conductivity was determined to be 73.4 W/moC using equation 1, which is derived for one dimensional radial heat conduction in a cylinder.
where is the power absorbed, k is the thermal conductivity of CNT film, t is the thickness of CNT film, r1 and r2 are radii of two locations, and T1 and T2 are the temperatures at two radial positions.
Measurement of thermal conductivity of SWNT film has been rarely made elsewhere except for this study. However, measurements of thermal conductivity of multi-walled carbon nanotube (MWNT) films were made by other groups which showed varying values from 1.52 W/moC to 83.0 W/moC [5-8]. In addition, we have performed a thermal conductivity measurement at room temperature for a thin (~1 µm thick) stainless steel stripe using micropipette thermal sensors (not published). Repeated measurement results for stainless steel stripes revealed good agreement with literature value.
The developed technology can be applicable to electronic devices and integrated micro/nano-electro-mechanical systems (MEMS and NEMS) due to the feasibility in measuring temperature at a highly localized hot spots with high spatial (few micron), and temperature resolutions (<0.01oC). Moreover, usefulness of the micro thermocouple sensor is extended to biological purposes; intramembranous cell temperature measurements have previously been made by our group to assess cell temperature rise due to laser irradiation .
 Volklein, F and Kessler, D.-P. E., Methods for the Measurement of Thermal Conductivity and Thermal Diffusivity of Very Thin Films and Foils. Measurement, 1987, Vol. 5(1): pp. 38-45.
 Fish, G., Bouevitch, O., Kokotov, S., Lieberman, K., Palanker, D., Turovets, I., and Lewis, A., Ultrafast Response Micropipette-based Submicrometer Thermocouple, Rev Sci. Instrum., 1995, Vol. 66 (5): pp. 3300-3306.
 Watanabe, M. S., Kakuta, N., Mabuchi, K. and Yamada, Y., Micro-thermocouple Probe for Measurement of Cellular Thermal Responses, Proc. 27th Ann. Conf. 2005 IEEE EMB, pp. 4858-4861.
 Kakuta, N., Suzuki, T., Saito, T., Nishimura, H. and Mabuchi, K., Measurement of Microscale Bio-thermal Responses by Means of Micro-thermocouple Probe, EMBS International Conference, 2001, Istanbul, Turkey: IEEE, 3114-3117.
 Sinha, S., Barjami, S., Iannacchione, G., Schwab, A., and Muench, G., Off-axis Thermal Properties of Carbon Nanotube Films, J. Nanopart. Res., 2005, Vol. 7: pp. 651-657.
 Che, J., Cagin, T., and Goddard III, W. A., Thermal Conductivity of Carbon Nanotubes,
Nanotechnol., 2000, Vol. 11: pp.65-69.
 Yang, D. J., Zhang, Q., Chen, G., Yoon, S. F., Ahn, J., Wang, S. G., Zhou, Q., Wang, Q., and Li, J. Q., Thermal Conductivity of Multiwalled Carbon Nanotubes, Phys. Rev. B, 2002, Vol. 66: p. 165440.
 Hu, X. J., Padilla, A. A., Xu, J., Fisher, T. S., and Goodson, K. E., 3 omega Measurements of the Thermal Conductivity of Vertically Oriented Carbon Nanotubes on Silicon, J. Heat Transfer, 2006, Vol. 128: pp. 1109-1113.
 Shrestha, R., Choi, T. Y., Chang, W., Kim, D., A High-Precision Micropipette Sensor for Cellular-Level, Sensors, 2011, Vol. 11: pp. 8826-8835