Nanostructured copper DOI: 10.1002/smll.200700991 Nanostructured Copper Interfaces for Enhanced Boiling** Chen Li, y Zuankai Wang, y Pei-I Wang, Yoav Peles, Nikhil Koratkar, * and G. P. Peterson* Phase change through boiling is used in a variety of heat-transfer and chemical reaction applications. [1–7] The state of the art in nucleate boiling has focused on increasing the density of bubble nucleation using porous structures and microchannels [8–12] with characteristic sizes of tens of micro- meters. Traditionally, it is thought that nanoscale surfaces will not improve boiling heat transfer, since the bubble nucleation process is not expected to be enhanced by such small cavities. [13–15] In the experiments reported here, we observed unexpected enhancements in boiling performance for a nanostructured copper (Cu) surface formed by the deposition of Cu nanorods on a Cu substrate. Moreover, we observed striking differences in the dynamics of bubble nucleation and release from the Cu nanorods, including smaller bubble diameters, higher bubble release frequencies, and an approxi- mately 30-fold increase in the density of active bubble nucleation sites. It appears that the ability of the Cu surface with nanorods to generate stable nucleation of bubbles at low superheated temperatures results from a synergistic coupling effect between the nanoscale gas cavities (or nanobubbles [16–18] ) formed within the nanorod interstices and micrometer-scale defects (voids) that form on the film surface during nanorod deposition. For such a coupled system, the interconnected nanoscale gas cavities stabilize (or feed) bubble nucleation at the microscale defect sites. This is distinct from conventional-scale boiling surfaces, since for the nanostructured surface the bubble nucleation stability is provided by features with orders-of-magnitude smaller scales than the cavity-mouth openings. Cu nanorods were deposited on a polished Cu substrate by oblique-angle deposition. Cu was selected for this study because it is widely used in heat-transfer applications. [19,20] In oblique-angle deposition, a flux of Cu atoms is incident on the substrate at a large incidence angle (>858; see Figure 1a). This results in the formation of isolated nanorods due to the atomic shadowing effect [21,22] during growth, through which some of the incident atoms are not able to reach the substrate because of concurrent growth of parallel structures. Since the substrate was not rotated during the deposition process, the rods grew inclined [21] (a  608) towards the direction of the incident flux, as shown schematically in Figure 1a. The depositions were performed in an electron-beam evaporator with base pressure of 10 7 Torr. Scanning electron microscopy (SEM) characterization was performed (Figure 1b) to study the structure of the deposited nanorods. The diameter of the nanorods increased from a few nanometers at the base to about 40–50 nm at the tip. The average tip-to-tip spacing between the nanorods is 50 nm and the nanorod height is 450 nm. The presence of defects (or voids) is also noticeable in the SEM image. These voids are an artifact of the uneven height distribution of the Cu substrate, which causes certain regions (for example, in valleys) to be shadowed out [22] during the oblique-angle deposition. These micrometer-scale voids give the Cu nanorod surface a combined nano- and microscale surface topography. The density of microscale defects on the Cu surface with nanorods (10 4 defects mm 2 ) was comparable to that of the plain Cu surface. Prior to performing the boiling heat-transfer experiments, we tested the wetting properties of the nanorod samples in comparison to the polished Cu substrate. For this, 2-mL droplets of distilled water were deposited on the substrates and the static contact angle was measured. The results (Figure 1c) indicate a reduction in the macroscopic water contact angle from approximately 55 to 38.58 due to the enhanced roughness [23] caused by the nanorod structures. The test facility used to evaluate the effect of nanorod deposition on the boiling performance (Figure 2a) consists of a Cu block, liquid chamber, liquid supply system, and heating system. The Cu block is thermally insulated to assure nearly one-dimensional (1D) thermal conduction. Three K-type thermocouples (TC 1 , TC 2 , and TC 3 ; Figure 2a) are soldered onto the Cu block at a separation of 10 mm to monitor the temperature profile along the axial direction, and hence the 1D heat flux can be accurately estimated by Fourier Law. [24] The TC 1 sensor, which is 0.5 mm below the boiling surface, is used to estimate the heating wall temperature, while sensor TC 4 provides the ambient fluid temperature. The bubble visualization system comprises a Motion Scope high-speed charge-coupled device (CCD) camera with 640  480 resolution and a maximum frame rate of 2000 frames s 1 , an Olympus microscope (U-TV1X-2) with magnification from 60 to 1000, and an Agilent data acquisition system. The Redlake image-processing program enables precise analysis of bubble release size and release frequency. In the boiling tests, the power to the heater was increased and the process repeated until the critical heat flux (CHF) was communications [  ] Prof. N. Koratkar, Z. Wang, Prof. Y. Peles Department of Mechanical, Aerospace, and Nuclear Engineering Rensselaer Polytechnic Institute Troy, NY 12180 (USA) E-mail: koratn@rpi.edu Prof. G. P. Peterson, Dr. C. Li Department of Mechanical Engineering University of Colorado Boulder, CO 80309 (USA) E-mail: bud.peterson@colorado.edu Dr. P.-I. Wang Department of Physics, Applied Physics, and Astronomy Rensselaer Polytechnic Institute Troy, NY 12180 (USA) [y] These authors contributed equally to this work. [  ] The authors acknowledge the support of the National Science Foundation under award CEBT-0721246 to G.P.P. and awards ECS 0403789 and ECS 0506738 to N.K. : Supporting Information is available on the WWW under http:// www.small-journal.com or from the author. 1084 ß 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim small 2008, 4, No. 8, 1084–1088