Principle Investigator: Dr. Min Soo Lim
Tribology is the science and technology of two interacting surfaces in relative motion, such as friction, wear, adhesion and lubrication . The significance of tribology has been recognized recently because of economic reasons  and has emerged as an independent area of science and technology.
The continuing desire to miniaturize devices demands a complete understanding of tribological properties of surfaces of moving components on an atomic or molecular level and moves conventional tribology to an ultra small scale, referred to as nanotribology. A few examples of its application are magnetic storage devices such as computer hard disk drives (HDD) , microelectromechanical systems (MEMS) , and chemical mechanical planarization (CMP) [5-7] processes in semiconductor fabrication.
The reduction of the distance between the rotating disk and read-write head to several nanometers in up-to-date high density magnetic storage, and dimensional reduction of moving components of MEMS to sub-micron levels demands lubricant films to be as thin as a single molecular layer with greatly enhanced lubricity. The importance of the influence of chemical composition and molecular structure of such molecularly thin lubricants on tribological properties cannot be overemphasized and should be investigated for the development of new lubricant systems suitable for modern-day devices, leading to their prolonged lifetime and cost reduction.
Planarizing a wafer surface with a good control of wafer planarity and material removal rate is a critical step in the chip fabrication process. CMP process has been widely adapted in the semiconductor industry to planarize each level using slurry chemistry and mechanics for multi-level interconnection. As device size continues to shrink, the control of slurry chemistry plays a more important role in the CMP process. Fundamental research on the influence of slurry chemistry on planarization efficiency needs to be done to optimize the process, ultimately reduce the cost of ownership, and result in a greater quantity of wafer products of good quality and a reduced quantity of toxic slurry waste.
An atomic force microscope (AFM) is an indispensable tool to explore atomic scale friction, wear, and lubrication of surfaces, because it models a tribological contact by using a probe tip as a model of a single microasperity. AFM also produces images (images of topography, friction force, magnetic force, chemical force, etc…) of a sample surface by scanning an unmodified or modified sharp tip mounted on a flexible cantilever with a spring constant typically between 0.1 and 100 N/m over the sample surface at fixed or variable loads.
Exploiting semiconductor device fabrication technologies and developing new synthesis and extraction methods of self-assembly materials are essential for the proposed research. In this regard, interdisciplinary research efforts will be established in collaboration with chemists, material scientists, and electrical/computer science engineers. Described below are a few examples of research areas that I would like to investigate in my research group with such combined efforts.
In the following sections, I will describe briefly measurement methodology of interfacial forces (friction and adhesion), and two research plans to investigate tribological properties of bare and lubricated surfaces in air and in liquid.
II. Manipulation of interfacial forces
1. Surface topography
The magnitude of each signal from the four individual sectors of the quadrant photo detector is summed or subtracted electronically to produce normal and lateral force signals . When the tip is far away from the sample surface, the reflected laser light hits the center or equilibrium position of a photo detector, producing zero value of normal or lateral force. However, contact of the tip with the sample surface causes the cantilever to deflect normally or laterally, which generates non-zero values of normal or lateral force.
Topography of the sample surface is obtained under feedback control (vertical positioning mechanism) by plotting the piezo voltage required to hold the fixed normal force, as a function of position across the surface. Surface topography is, in other words, a 3-D display of equal normal force over the surface. Since topographic images are collected simultaneously with their corresponding lateral force images, surface morphology can be correlated to the frictional properties of surface. Lateral force images can also resolve domain structures which are not revealed by topographic images.
2. Friction-load plot
The signals from normal and lateral deflections of the cantilever are monitored and recorded independently and simultaneously from the photo detector . A voltage ramp will be applied to the piezoelectric Z scan sector to load/unload the sample during scanning. Signals from both normal and lateral deflections of the cantilever will be obtained simultaneously. The friction-load plot will be produced by plotting frictional force as a function of the corresponding normal force or load (Figure 1). The slopes of the data curves correspond to ‘coefficients of friction’ and represent frictional properties of a given sample surface.
3. Force-distance curve
Force-distance curves will be collected by monitoring normal deflection of the cantilever versus vertical separation distance as the tip and the surface are brought into contact and subsequently separated. Adhesion or the pull-off force (i in Figure 2) is the force required to separate the tip completely from the sample surface after contact and can be correlated with the interfacial adhesive properties between a probe tip and a given sample surface.
III. Manipulation of interfacial properties of solid surfaces with functionalized self-assembled monolayers (SAMs)
Molecular-scale boundary lubrication is inevitable for many modern day devices due to micrometer-/nanometer- sized shrinkage of their geometrical dimensions, including high-density computer hard disk drives (HDD) and microelectromechanical systems (MEMS). Increasing the capacity of hard disks can be achieved partly by decreasing the fly-height of the read-write head over the disk surface. The fly-height of up-to-date hard disk drives already reaches to ~ 5 nm, leading to more frequent high-speed contacts of the read-write head with the disk surface and probable damage to the surface. Consequently, the lubricant film should be molecularly thin and should be enhanced with lubricity.
MEMS are yet to be realized in many practical applications due to serious problems caused by friction, wear, and stiction. Friction, wear, and stiction between ultra small moving components generally made of polysilicon films should be reduced effectively with proper lubricants, leading to high-speed performance and the prolonged life time of MEMS. Chemical and structural modification of the molecularly thin lubricants would influence their lubricity greatly and that demands fundamental study on the tribological properties of lubricants.
A self-assembled monolayer (SAM) of organic molecules is a good candidate to serve this demand . A SAM is a two-dimensional molecular array that is spontaneously organized by adsorption of amphiphilic organic molecules on a solid inorganic surface. Several good examples include alkanethiols on noble metals such as gold and silver, alkyl carboxylic acids on metal oxides, and alkyl silanes on SiO2 surfaces. The intermolecular interactions of the SAM can provide highly ordered monolayers with nearly defect free coverage under the appropriate conditions while the head group can provide strong binding with the surface.
The interfacial properties of SAMs such as friction and adhesion are decisive criteria for good lubricants which are significantly dependent upon their molecular structure, chain length, packing order, packing density, chemical composition, and molecular termination. Figure 2 is an example of data that was obtained several years ago and it signifies the influence of molecular termination of a SAM on the interfacial frictional force on a nano-Newton (nN) scale. SAMs of methyl- and trifluoromethyl- terminated hexadecanethiols were formed on Au(111). Interfacial friction was measured from each of the SAM and corresponding friction-load plot was made as shown in the Figure 2. It shows that fluorination of the SAM results in an increase of interfacial friction approximately by the factor of two in comparison with its hydrogenated counterpart. It further emphasizes that simple chemical modification of the SAM provided better lubricity than that of the unmodified.
As a starting point and in an effort to elucidate the origin of the tribological nature of lubricants in chemical aspects, a variety of organo-silanes with different structure, chain length, and terminal groups will be systematically grown on silicon surfaces. A series of friction and adhesion measurements will be performed on the SAMs of the organo-silanes.
IV. Manipulation of interfacial forces of solid surfaces by tailoring chemical conditions in the liquid phase
Miniaturization of device dimensions and the related need to interconnect and embed an increasing number of functional components on a single chip gives rise to the construction of multilevel interconnections on planarized levels in the integrated circuit (IC) fabrication process. Chemical mechanical planarization (CMP)  is a wafer planarization technique that removes very thin layers of materials by pressurizing a polishing pad against a wafer and simultaneously adding slurry onto the interface. Abrasives and chemicals in the slurry remove materials mechanically and chemically, respectively until desired planarity is reached at each level. As device size continues to shrink, the control of slurry chemistry (pH, concentration, reactivity, ect.) in comparison with the control of mechanical action of abrasives plays an even more important role in the CMP process. Figure 3 is another example of data that was obtained several years ago and it shows influence of solution pH on the interfacial friction and adhesion of a tungsten surface. It indicates that friction and adhesion of an interface embodied in an acidic or basic environment are influenced by chemical changes related to the protonation/deprotonation of the contacting surfaces, specifically through modifications of the interfacial shear strength.
Slight modification of chemical conditions in the CMP slurry can be significant enough to influence the success or failure of the planarization processes. However, little has been done to understand the influence of chemicals and interplay between chemicals and abrasives on the material removal.
The proposed research is unique because a small AFM tip (silicon nitride or alumina coated silicon nitride) will be used to simulate a single microasperity contact of a single abrasive particle with the metallic surfaces in systematically controlled aqueous conditions. In an effort to optimize chemical and mechanical conditions of the slurry, the influence of chemical elements and abrasives and interplay between the two on the interfacial forces and material removal will be investigated in-situ in various aqueous conditions with AFM. Earned achievements in the research will be conveyed to CMP industry and will be utilized to optimize/tailor slurry conditions depending on metal (tungsten, copper, aluminum, and so forth) CMP or dielectric CMP processes.
1. Carpick, R. W.; Salmeron, M., Scratching the Surface: Fundamental Investigations of Tribology with Atomic Force Microscopy. Chemical Reviews,1997, 97, 1163 - 1194.
2. We may want strong tribological effects in some cases or weak effects in others, for example, car brakes and tires (friction with less wear is desirable), machining/polishing (friction with wear is desirable), engines, gears, and bearings (less friction and wear is desirable), and joints of dynamic components (lubrication is desirable to reduce friction and wear at the interface). Losses up to 6 % of gross national products, or several hundreds of billion dollars in the United States, result annually because of energy loss and wear associated with tribological events.
3. Perry, S. S.; Mate, C. M.; White, R. L.; Somorjai, G. A., Bonding and tribological properties of perfluorinated lubricants and hydrogenated amorphous carbon films. IEEE Trans. Magn. 1996, 32, (1), 115-121.
4. Maboudian, R.; Howe, R. T., Critical Review: Adhesion in surface micromechanical structures J. Vac. Sci. and Tech. B. 1997, 15, (1), 1 - 20.
5. Lim, M. S.; Perry, S. S.; Galloway, H. C.; Koeck, D. C., Microscopic studies of friction and wear at the benzotriazole/copper interface. Tribology Letters 2003, 14, (4), 261 - 268.
6. Lim, M. S.; Perry, S. S.; Galloway, H. C.; Koeck, D. C., pH-mediated frictional forces at tungsten surfaces in aqueous environments. J. Vac. Sci. Technol. B 2002, 20, (2), 575 - 579.
7. Lim, M. S.; Heide, P. A. W. v. d.; Perry, S. S.; Galloway, H. C.; Koeck, D. C., Microscopic investigations of chemo-mechanical polishing of tungsten. Thin Solid Films 2004, 457, 346 - 353.
8. See appendix
9. Lim, M. S.; Feng, K.; Chen, X.; Wu, N.; Raman, A.; Nightingale, J.; Gawalt, E. S.; Korakakis, D.; Hornak, L. A.; Timperman, A. T., Adsorption and Desorption of Stearic Acid Self-Assembled Monolayers on Aluminum Oxide. Langmuir 2007, 23, 2444 - 2452.
10. CMP is a process of high cost and its market grows rapidly at an annual growth rate of 30 %. It is estimated that one CMP machine with a wafer production rate of 40 wafer/hr and operating at 65 % of capacity would consume $275,000 per year only in CMP slurry without considering additional costs for the disposal of toxic slurry wastes, and deionized water needed for cleaning the processed wafers.
VI. Appendix - Instrumentation
The AFM in this research will be a dual mode system, operational in both ambient and liquid environments and will employ a beam deflection technique for the detection of signal because of its relative simplicity and convenience. In this method, laser light is reflected from the back of a gold-coated cantilever onto a four-quadrant position-sensitive photo detector (PSPD) (Figure 4). The deflection of a tip/cantilever assembly arises from interaction with the sample surface and is monitored by the PSPD. Tip-cantilever assemblies will be purchased from Veeco Metrology.
The components of the AFM body (D in Figure 5) will be either purchased directly from manufacturers (laser diode, photo detector, fluid cell, piezoelectric scanner, tip/cantilever assembly, etc…) or fabricated in the machine shop (sample holder, AFM house, etc…). Control electronics and software will be purchased from RHK Technology and will be used for data collection and process.
Mechanical noise will be blocked by two noise-damping stages, that is, a floating table (E in Figure 5) and three springs (F in Figure 5) that hold and float the AFM body in the air. Noise generated by air current will be further blocked by a glass bell-jar (C in Figure 5). The gas-inlet port (A in Figure 5) will control the relative humidity inside by flowing dry or wet nitrogen gas.