THE IMPORTANCE OF METROLOGY AND STANDARDISATION FOR MICRO-SYSTEMS TECHNOLOGY R. K. Leach Centre for Basic, Thermal & Length Metrology, National Physical Laboratory, TEDDINGTON, MIDDLESEX TW11 0LW UK, 0044 208 943 6303, (
[email protected]) This paper discusses the standardisation and metrology issues that are required to enhance the commercial viability and quality of micro-components and micro-systems. It is asserted, as with macro-scale engineering and production, that the key to the successful manufacture of micro-scale devices is proper process control using various metrology tools, without which, useable yields will be low and failure modes poorly understood. Consequently, poor yields can lead to slow and often costly adoption of these devices. The paper discusses the dearth of instruments that have a measuring range that lies between the fledgling tools of nanotechnology (for example AFMs and their variants) and the metrology tools of conventional manufacturing (for example CMMs) that are vital to the application of metrology at the micro-scale. A real-life example is given to illustrate how metrology can be applied to enhance a micro-manufacturing process. The results of a recent comparison of commercial instruments are presented. 1. Introduction Micro-scale components are becoming increasingly important in a wide range of industrial applications, including digital projectors containing millions of micro-mirrors and micro-scale motion sensors for car airbags. Many other examples of microsystems technology promise to have a major impact on our daily lives, such as palm-sized high-definition displays and grain-sized implantable medical devices. These systems will work well only if they are built with micrometrelevel accuracy. As with macro-scale engineering and production, the key to the successful manufacture of microscale devices and products is proper process control [1]. Without this control the fraction of useable devices will be low and failure modes will be poorly understood leading to costly modifications and slow market uptake. Dimensional metrology is fundamental to understanding and controlling production processes at the micrometre level. However, while the metrology and standards infrastructure exists in traditional manufacturing industries, the need for traceable metrology of micro-scale components and devices has not been given the attention it urgently needs. One example in which improved metrology could prove to be highly beneficial comes from the commercial production of high-accuracy pressure measuring instruments. Druck Holdings in the UK use micro-machined silicon diaphragms as the transducer elements in their range of pressure sensors. These sensors are used in many applications ranging from laboratory-environment monitoring and marine sensing to automotive production and aerospace altimetry. The diaphragms are machined using methods borrowed from the semiconductor industry and are then inspected by human operators who decide if each component is fit for purpose. Each suitable diaphragm is then interfaced with signal-processing electronics to optimize the response function of the sensor. With proper metrology, these costly manual inspection and optimization processes could be completely avoided. Measurements of the 3-D structure of the individual diaphragms could lead to significant
cost savings. However, as yet, no measurement system that can traceably measure the diaphragms is commercially available. 2. Current tools for the metrology of micro-systems Engineering at conventional scales (metres to millimetres) benefits from the availability of a wide range of metrology tools, ranging from specialist devices designed to measure just one particular feature to more versatile tools such as coordinate measuring machines (CMMs). CMMs measure the spatial co-ordinates of points on the surface of an object, with either a contacting or noncontacting probe and typical coordinate measuring machines can survey objects up to a few metres in size with an accuracy, for the best CMMs, of the order of a few micrometres. At the nanometre scale, the closest analogues of CMMs are scanning probe microscopes. Although these instruments can achieve accuracies of a few nanometres, their range is limited and they tend to be restricted in their height measurement. However, between the conventional and the nano-scale instrumentation, there is a significant lack of metrology tools that are capable of measuring 3-D objects. The following section describes some of the commercially available tools for metrology of micro-systems. 2.1 Stylus instruments There are a number of commercially-available instruments that operate by dragging a (usually diamond-tipped) stylus across a surface and measuring the vertical displacement of the stylus as it encounters topographical features (see [2], [3] and [4]). When such an instrument is used to measure the surface texture of planar surfaces in one dimension these instruments are backed up by a well-established standards infrastructure [5], but it is difficult to use such instruments to measure the high aspect ratio structures that are commonly encountered in micro-systems. Other limiting factors of such instruments include: the finite size of the stylus tip (the smallest reported radius of curvature is 50 nm, but it is more usually around 2 µm); the need to contact the surface, and therefore possibly cause damage; and their slow operating speed when measuring complex features. However, stylus instruments are the only instruments that have a route to traceability to the definition of the metre that can be easily demonstrated [6]. 2.2 Scanning probe microscopes The most common form of a scanning probe microscope (SPM) used to measure microstructures is the atomic force microscope (AFM). AFM images are obtained by measurement of the force on a sharp tip created by the proximity to the surface of a sample. The tip is usually at the end of a cantilever arm, the deflections of which can be measured using some optical means. As the tip is scanned over the surface in a raster pattern the force is kept constant with a feedback system and thus the tip follows the contours of the surface. Tip radii are usually of the order of 10 nm and forces are of the order of nano-newtons. To a lesser extent, AFMs suffer from the same drawbacks as the larger stylus instruments, and a great deal of work has been carried out to create a standards infrastructure for these instruments [7]. Whilst AFMs can have resolutions of less than a nanometre in x, y and z, they are limited to short scan ranges (usually less than 100 µm in x and y and 5 µm in z). Recent comparisons show large discrepancies when comparing AFM results [8], [9]. Another SPM method that has been applied to microstructures is scanning near-field optical microscopy [10]. This method has been commercially combined with a confocal microscope to increase its range. However, there are few methods that can measure truly 3-D features.
2.3 Optical instruments There are many optical methods that are now being applied to the measurement and inspection of microstructures. Amongst others these include scanning white light interferometry [11], confocal microscopy [12], systems using digital micro-mirror devices [13] and holographic microscopic interferometry [14]. Whilst optical methods are fast and non-contact, they suffer from a number of problems that make them difficult to calibrate and this is discussed in the next section [15].
Ra/nm
3 A recent comparison It is widely recognised in the scientific literature that significantly different results are obtained from different measurement systems measuring the same features on the surface – unfortunately this fact is often overlooked in practice. There is no reason to suppose, for example, that optical radiation associated with a non-contacting sensor 300 penetrates to exactly the 250 same depth below a surface as a contacting stylus will 200 penetrate elastically. Optical 150 and stylus probes happen to agree well on many types of 100 surfaces, but extra care must be taken when interpreting 50 the measurement data 0 towards the extremes of the NS IV Stylus 1 Stylus 2 Stylus 3 Inter 1 Inter 2 Inter 3 operational range of an instrument. Optical probes, Figure 1 Results with an 8 mm nickel sinusoid (note that the 2455 nm result for example, do not perform is off the scale) well when large localsurface slopes are present that may well be within the tracking capability of most stylus instruments. A recent comparison [15] of profiling measurements in the UK has shown a wide variation in results for laboratories with different, and in some cases identical, measuring systems. In one instance, an artefact with a sinusoidal profile, with an 8 µm period and 150 nm Ra, was measured as having an Ra that varied from 20 nm to 2455 nm! The comparison exercise included using stylus instruments, interferometers and an instrument with an auto-focus system. Figure 1 shows the results of the comparison. It is sobering to consider that if these instruments are struggling to achieve comparable results with well-defined structures in 2-D, what is happening when they are measuring complex microstructures? 4 Small CMMs In order to close the metrology gap discussed in the introduction to section 2, many groups have begun to develop small CMMs (for a recent review see [16]). NPL has developed a small CMM [17] that operates over a range up to 50 mm along the x, y and z axes, and has an accuracy of 50 nm. It therefore extends the metrology capabilities of CMMs to measure smaller, more delicate, objects, while achieving measurement uncertainty that is an order of magnitude smaller. The new instrument comprises a novel contact probe that applies a tiny force to the component so that measurements can be made without significantly damaging the surface. The tip of the probe is
monitored by six laser interferometers, which measure both its displacement and rotation with high precision. Any imperfections in the structure of the machine are compensated for by a selfcalibration technique, which makes use of the redundancy of information from the six laser interferometers and other angle-sensitive devices mounted on the machine. In this way, a 3-D vector model of the positions and alignment axes of the laser interferometers can be derived, and any errors in the flatness of the optical components can be mapped. 5 A future look In spite of the technical advances made so far, there is still a major gap in the available instrumentation that NPL plans to close in the near future. NPL is aiming to develop an even smaller instrument - a truly micro CMM that will be able to measure complex three-dimensional microstructures with nanometric accuracy. This will involve the development of tiny probes, using micro-machined parts and, perhaps nanotubes as the sensing elements. NPL is also collaborating with the Institute für Mikrotechnik, Mainz (Germany) and the Central Microstructure Facility (UK) in an EU-funded project (MICROSPEC) to review the standardisation issues that face micro-systems technology. This will run an International Workshop (around February 2003) and produce a Roadmap. Once the appropriate measurement standards and instruments have been established, calibrated commercial devices can be used more effectively on the multitude of microstructures that are being developed. These devices will pave the way for formal standards in microsystems technology that will eventually improve future micro-scale products and the quality of life for everyone. Acknowledgements The work on the Small CMM was funded by the UK National Measurement System Length Programme 1999-2002. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.
T. Pfeifer, R. Freudenberg, G. Dussler, B. Bröcher, Measurement 30, 1, (2001) T. Thomas, Rough surfaces (Imperial College Press) (1999) D. J. Whitehouse, Handbook of surface metrology (IOP Publishing) (1994) S. T. Smith, Industrial metrology: surfaces and roundness, (Springer) (2002) R. K. Leach, The measurement of surface texture using stylus instruments, NPL Measurement good practice guide No. 37 (2001) R. K. Leach, Meas. Sci. Technol. 11, 1162, (2000) J. Garnaes, et. al., Proc. Xth Int. Colloq. On Surfaces, Chemnitz, Germany, 31st Jan – 2 Feb, 277, (2000) M. Senoner, et. al., Proc. 3rd Int. euspen Conf., Eindhoven, Netherlands, 26th – 30th May, 773, (2002) R. K. Leach, Proc. Large Lenses & Prisms, 27-30 March, UCL, UK, (2001) B. T. Rosner, Rev. Sci. Instrum., 73, 2505, (2002) M. Hill, et. al., Opt. Lasers Eng., 36, 169, (2001) H. J. Tiziani, T. Haist, S. Reuter, Opt. Lasers, Eng., 36, 403, (2001) F. Bitte, G. Dussler, T. Pfeifer, Opt. Lasers Eng., 36, 155, (2001) G. Wernicke, et. al., Opt. Lasers Eng., 36, 475, (2001) R. K. Leach, NPL Report, CBTLM 15, 1, (2002) S. Cao, et. al., Microsys. Technol., 8, 3, (2002) G. N. Peggs, A.J. Lewis, S. Oldfield, Ann. CIRP, 48, 417, (1999)
