Skip to sections.
The manufacture of millimeter-size industrial components requires micrometer-order processing precision. Inspection of these components requires even greater precision, in the form of nanometer precision measurement technology. In the growing field of ultra-micro processing, the Optical Radiation Pressure Microprobe Unit manufactured by Nikon Engineering will prove indispensable. This device utilizes laser trapping technology for “capturing” an object using light, and it can accurately measure the shape of ultra-fine components.
In this interview we asked Professor Yasuhiro Takaya of Osaka University’s Graduate School of Engineering about such topics as the principle of laser trapping using optical radiation pressure, the scope of application of the microprobe unit, the nano-CMM coordinate measuring machine—a 3-dimensional coordinate measurement system, and anticipated future developments.
“Capturing” an object refers to the ability to move it by applying a force or to lift it in defiance of gravity. Force is applied to the object in order to capture it, and light is used to apply this force. The lighting that is now illuminating us all is in fact applying an extremely weak force. The effect that light exerts on the surface area of a target object in the micro domain is greater than those exerted by the object’s own volume and by gravity. Hence, despite the fact that light only exerts an extremely weak force on the object, it is sufficiently powerful to move it.
This was predicted by Isaac Newton over 300 years ago. James Clerk Maxwell subsequently turned this into theory in 1864 when he published equations that revealed light possesses energy. However, more than 100 years passed before light was actually used to raise an object. Since the energy in light is very weak, moving an object requires that the light be focused into a small area by means of a lens, and this was made possible by the development of the laser. Successfully demonstrated experimentally in 1971, lasers were probably not that widespread at the time. In their experiments*1, Arthur Ashkin’s team succeeded in raising (levitating) an object in a vacuum by striking it with a laser from beneath.
In our current experiments we raise an object in air by striking it with a laser from above. However, the basic method of focusing the laser light using a lens and striking the object remains essentially unchanged to this day.
When an object changes its course of direction, it applies a force to the corresponding object from which it rebounds.
When the course of direction of light changes due to reflection or refraction, the light likewise exerts a motive force on each corresponding object. This force is known as “optical radiation pressure.”
Based on the principle of laser trapping, to put it simply, it represents the skillful use of the particle-holding properties of light. For example, when a ball strikes a wall and rebounds, its direction of travel changes, and at the same time it exerts a force on the wall. Light is just the same. For example, if light strikes a transparent glass pane, it will be refracted and enter the glass. In addition, some of the light will be reflected from the surface. It is probably easy to understand that, in the case of reflection, a force is acting on the glass just like a ball. However, since refraction also alters the direction of light, a force is acting in this case too. This force is referred to as “optical radiation pressure.”
If the object is a level surface, it will be pushed down by light from above. However, in these experiments we are using minute particles that are spherical. The path of the light changes twice, with the reflection and refraction that occur at the particle surface, and with reflection and refraction as the light re-emerges from the particle. The vectors of the optical radiation pressure from each change of direction combine to become the force that raises the sphere, and essentially the minute particle moves towards the focus of the laser light. Since gravity is also acting, stability is achieved at the point where the raising force and gravity cancel one another out (slightly below the focal point), and this is the trap point. This position varies according to the power of the laser, the particle’s weight, and the numerical aperture (N.A. value) of the lens. However, we are going further and taking into account factors such as the refractive indices of the air and the particle in our calculations, ascertaining the type of conditions that will increase the lifting force (trap force), and verifying them experimentally. The results of these experiments are now also being used in the unit which Nikon has manufactured for us.
Laser trapping in air has rarely been investigated overseas, and there have only been a few cases in Japan. That is to say, it is even more difficult to conduct experiments in air, than in a vacuum or in water. In air, for some reason there is an adsorption force or viscosity acting between the substrate on which the particle is placed and the particle itself. This powerful force strains to tear the particle from the glass substrate. We have been conducting these basic experiments for approximately 10 years now; however, only for the past five or six years have we been able to smoothly achieve levitation in air.
A number of methods exist; initially, however, we tried the technique of striking the adhering surface with a strong pulse laser. With this technique a powerful laser light is first used to alter the properties of the contact area, and the object is then raised. We started using this method in June and were finally successful with it six months later in December. We later realized that the amount of moisture in the air somehow affects the adsorption force. We believe that this is why this method did not work well during the rainy season but was successful in December, when the air was dry.
In order to make the apparatus smaller, we subsequently switched from a method employing a large pulse laser device to a technique for tearing the particle from the glass substrate by using ultrasound to make the stage vibrate. Even with this method, the particle still moved with the glass substrate, so after all we have been going through approximately six months of repetitive trial and error, tinkering with frequency, amplitude, and the like.
At one point we were unsuccessful in placing particles on the stage, in that we could not disperse them evenly and they tended to cluster. Nevertheless, when we went ahead and vibrated the glass substrate, the particles still rose. The particles piled up on the glass substrate, and as the vibration spread belatedly from one individual particle to the next, the uppermost particles became more easily detachable. This method for trapping the particles with a laser exhibited a high level of repeatability and now enjoys a success rate of almost 100 percent. But for this chance failure with the clustering, we would probably have had to expend even more effort.
The entire assembly during basic experimentation. As can be seen, it takes up the entire surface of the desk. The long, white box-shaped device is a YAG laser vibrator.
Optical radiation pressure microprobe units mounted on a nano-CMM
The two upper boxes in the center are optical radiation pressure microprobe units manufactured in the course of this project. The entire assembly enables measurement to be carried out at the nm level.
The internal structure of an optical radiation pressure micro-probe unit
From the upper right corner, laser light for optical trapping is radiated and led to the lower left objective lens. The left side is the optical system for detection and observation.
At the basic experimental stage, it consisted of an optical system that focused and controlled the laser and a mechanism for raising the particles that were on the stage. Since the basic experiments required a high-output pulse laser (or YAG laser), as explained previously, the assembly as a whole was huge and covered an entire table.
The current optical radiation pressure microprobe unit is approximately the size of an A4 file. Since the technique of vibrating the stage is used to lift the particles, a 1064-nm-wavelength infrared CW laser (or fiber laser) is employed for the sole purpose of laser trapping. Luckily, a compact fiber laser came out that was of exactly the same wavelength as the pulse laser that we were using in our basic experiments.
The device overall is comprised of three mechanisms: a laser-trapping mechanism, a structure for operating a captured particle as a probe, and a mechanism for observing the action of the probe and detecting its position. Nikon put great effort into reducing the device’s footprint from the size of an entire table to A4 size.
As it was assumed that this device would be incorporated into larger mechanisms as a sensor, it needed to be as compact as possible. It was initially envisaged that the finished device would be approximately the same size as a sausage. We thought that it would be good if we could make it about the size of a fish sausage (laughs). As a practical matter, with precision at a premium, a long focus lens system is required; however, that would be difficult at this stage. Although we came up with the basic design here, Nikon put great effort into the design of the optical system, just to get the device down to its current size.
We do possess the research equipment and technology to build a large optical system. However, once you actually try to build the device, implementation requires various skills, such as design ability, mechanical engineering capability, and manufacturing capability. A university lacks this kind of expertise. Nikon, after all, seemed to be the best option. We did sound out other manufacturers; however, since this was a device that no-one had ever built before, we didn’t know how it would work until it had been completed. The truth is that only Nikon has a proven track record in the development of this kind of device and fulfills the conditions that I mentioned earlier. Thanks to Nikon, it works perfectly. This is the first device of its size ever built that can trap an object in air.
A certain level of manufacturing accuracy is required in order for industrial product components to fulfill their functions, and this level of manufacturing accuracy requires a certain level of measurement accuracy. For example, the technology now exists for both manufacturing and measuring the many 10 cm components that are used in car production and other sectors. However, making the 1 mm machine components that are used in cell phones requires µm precision manufacturing and nm-precision to verify whether or not they have been manufactured according to their design. Manufacturing requires precision that is four orders of magnitude higher, while measurement requires precision that is a further two orders of magnitude higher still—that is, of the order of 10–6 x the component dimensions. This is what makes research into measurement such a tough business (laughs).
The horizontal axis shows the size of manufacturing components while the vertical axis shows the measurement precision required. For example, components with basic dimensions of 1 m require a processing precision four orders of magnitude higher (at 0.1 mm) and a measurement precision a further two orders of magnitude higher (at 1 µm). Similarly, 1 mm scale micro-components require a measurement precision of 1–10 nm.
Optical radiation pressure microprobe units are currently incorporated into the nano-CMM (coordinate measuring machine), which employs a 3-dimensional coordinate measurement system. This device strikes the object to be measured using a probe, takes the coordinates of every point on the object, and reconstructs its 3-dimensional shape using the accumulated coordinate data. This type of measurement could probably be ascertained from photographic imagery; however, photographs would not reveal an object’s depth. To a certain extent the depth of a perpendicular slope or a hole could be found by rotating the measurement object; however, this would then require precise rotational control. This approach is inadequate for the measurement of microscopic objects.
In the past there have been CMMs that could measure an object with dimensions of several cm to µm-level precision by physically contacting it with a probe with a diameter of the order of 0.1 mm. However, there were problems in the form of frictional resistance caused by the contact and in deformation of the probe itself, so that even measuring objects with dimensions of several cm proved difficult. Since the requirement is now for measurement of mm-class microscopic components, the precision requirements dictate that a probe no larger than 10 µm be used. This is why laser trapping has been used.
An 8 µm diameter particle (a glass sphere) is captured by laser trapping for use as a probe and brought up close to the object to be measured. The object’s surface position is ascertained by reading the changes in the reflection of light from the particle as the probe makes contact with the object. This also entails various problems, however. Since the probe is extremely small, it is affected by the thermal motion (or Brownian motion) of air molecules. Merely bringing it up close to the measurement object causes significant movement of the probe, such that the position at which contact with the object is made cannot be ascertained. In addition, it has been known for the probe to adsorb to the measurement object. Initially, we did not know what to do about these problems; however, we solved them by vibrating the probe and reading the resulting changes in vibration frequency and amplitude.
It is probably easy to understand this by looking at the actual form of the probe. Shot using a microscope, this movie shows how the probe is actually floating. The object being measured is a glass sphere 160 µm in diameter.
This image was taken in air during our basic experiments. It would be impossible to shoot such a clear image in water, due to the relative refractive indices. Taken several years ago, this photograph in fact required a considerable amount of time to shoot. Here, the distance between the objective lens and the particle was approximately 2 mm. On the current optical radiation pressure microprobe unit, this distance is now approximately 0.3 mm. The greater this distance, the better; however, this entails numerous as yet unresolved problems.
The image illustrates the particle’s lateral movement. For demonstration purposes, the particle is shown as slowly making large movements. In reality, however, it would probably be impossible to tell that it was moving, given the low amplitude of no more than 100 nanometers and the high vibration frequency of 1 kHz.
It was initially thought by some that if the particle (or probe sphere) were made to make such pronounced movements, it would fly off. However, the viscous resistance of the air prevents this from happening. It was only later that we realized how important a role the air’s viscous resistance plays. Vibrating the particle and bringing it close to the object creates a thin layer of air between the particle and the object, and increases the resistance between them. It is thus now possible to measure an object without making any physical contact with it.
The pyramid-shaped object shows the difficulty of measuring a three dimensional ridge line using conventional image measurement. This is a picture used in demonstrations of the optical radiation pressure microprobe unit. Since the image was taken with a digital camera, it is close to what the naked eye would see. This is how it appears. Although the 8 µm particle (or probe) cannot be seen, bright points seem to be visible and the floating structure is readily apparent. Only the scattered light rays of the normal illumination from the microscope can be seen.
The particle is brought closer to the object in this way and the changing position of the reflected light is read with a resolution of 1 nm, giving the X, Y, and Z coordinates. Although control would be extremely difficult, it is probably possible to move the particle so as to trace over the object to be measured. Implementing this would likely enable high-speed measurement.
It could be used right away for measuring the shape of micro lenses. These are the minute lenses that are used in CD and DVD pickups. Another possibility is the measurement of micro-machine components. In addition, if a vibrating 8 µm probe were to have an even smaller 100 nm object adhere to it, the change in mass would cause the vibration frequency to change as well. Using this principle, it would probably also be possible to analyze ultra-microscopic substances in the air that cause environmental pollution. It can be assumed that if semiconductor lasers were arranged in an array, they were each used to trap a probe, and they were then operated simultaneously, it would be possible to investigate the concentration of a substance in the air in a short space of time. Since semiconductor lasers and the probes are built small, several hundred sensors could be mounted on a rectangular chip with sides of just a few millimeters. I wish someone would make one (laughs).
In addition, a variety of laser-trapping experiments are being conducted. For example, particles like sugar candy balls are also currently being manufactured using nano-technology. There are other experiments in which minute windmill-shaped thin-film diamond objects are being made and then rotated at high speed using optical radiation pressure. In the microscopic nano-level domain there is a whole world of physical phenomena in operation that are as yet not well understood. However, it is possible to make a hollow or dig a groove 8 nm deep in an object using microprobes and micro-tools.
Examples of surface processing using laser trapping.
It is possible to form a gap or hollow several nanometers deep in the surface of an object.
Germany is the most advanced nation in nano-manufacturing and nano-measurement, probably followed by Japan and then the United States. However, there are various things involving laser trapping in air that only our laboratory is doing. We showed pictures of the probe introduced in this interview at a special laser trapping session held in the United States in August 2007, and these elicited an extremely strong reaction.
We are currently conducting experiments involving imparting rotational motion to a particle (or probe). Although it has proved extremely difficult to get a particle to move in an exact circle orbit, we are taking on the challenge in any case. For one thing, the position of a particle that has been trapped by imparting rotational motion to it is very stable. Even more interesting is the fact that when a probe that is moving in a circle orbit is brought close to the object to be measured, the circle orbit becomes an ellipse orbit due to the viscous resistance of the air, and the rotational axis aligns with the normal line from the surface of the object. As a result, both the coordinates and the normal direction can be ascertained simultaneously.
Ascertaining the normal line makes it easy to quickly move the probe so as to trace the object being measured. Since our laboratory building is located close to a civil-engineering laboratory, there are times when civil-engineering machinery is operating during the day, and even when we opt for the peace of the middle of the night, we cannot perform prolonged experiments, as vibration is a strict no-no when nano-precision measurement is being carried out. If we can improve the precision of our measurements, we would next like to reduce the time that it takes to perform measurements.
Thinking back, right from the start of this project we made unreasonable demands of Nikon (laughs). Initially we thought that they might turn us down. This was a device that no one had ever built before and keeping the design small in size must have been extremely hard. On top of that, Nikon seems to have put great effort into manufacturing all the components. Since Nikon was installing equipment devices that we had not used and whose features we did not properly understand, we had Nikon devise an adjustment mechanism that allows equipment to be swapped in and out. Incorporated into other equipment and measurement devices, the optical radiation pressure microprobe unit is likely to be in action soon.
This website features data in Flash format. To view this data requires Adobe Flash Player 7 or higher, which can be downloaded for free.
Posted April 2008