Investigations on the positioning accuracy of the Nano Fabrication Machine (NFM-100)

2.1 Nanopositioning and -measuring machines

Based on decades of research in this field, the Institute for Process Measurement and Sensor Technology developed the Nano Measuring Machine (NMM-1) and the Nano Positioning and Measuring Machine (NPMM-200) [3]. With these Nano Measuring Machines it could be demonstrated that the working principle of these 3D-coordinate measuring machines can be especially used to measure 3D-objects [4], [5]. Also, nanofabrication could be demonstrated in a macroscopic range with the NMM-1 [6]. Based on the experience from designing and operating these machines, the Nano Fabrication Machine was developed. As with the NMM-1 and the NPMM-200, the machine table of the NFM-100 is moved, while the measuring tool of these machines is at a fixed position. However, this machine is optimised for tasks in two dimensions and therefore especially wafer-based tasks.

2.2 Tip-based measuring system

As the machine table has no option for movement in the third dimension, the movement in z-direction has to be realised by the mounted measuring tool. At the moment, a tip-based measuring system is attached, which uses self-sensing and self-actuating microcantilevers.

Figure 1

SEM image of the active cantilever integrating thermal bimorph actuation (colored red), piezoresistive readout (colored green) [7], [9].

On the one hand, these microcantilevers, shown in Figure 1, allow the use in AFM mode for the detection of surface topology and associated surface properties. The microcantilevers, which are used in this tip-based measuring system, utilise a wheatstone measuring bridge to detect the deflection of the beam at the clamping point. To induce oscillation, this type of microcantilevers use a thermomechanical actuator. This actuator consists of two different materials (silicon and aluminium) with different coefficients of thermal expansion in shape of a meander [7] (see Fig. 1). Furthermore, these active microcantilevers can be especially utilised for nanofabrication. This novel lithographic technology based on a Fowler-Nordheim type electron emission from a nanotip is called FESPL, which enables cost-effective technology for nanodevices [8]. Nanostructures in the range of sub-10 nm can be realised with this lithography technique and the described tool, which could be successfully demonstrated in [10]. This system also offers more advantages like closed-loop lithography, which includes pre-imaging, overlay alignment, exposure and post-imaging for feature inspection, and also sub-5 nm lithography resolution with sub-nm line edge roughness [8].

3.1 Set-up of the NFM-100

Figure 2

Image of the nanopositioning machine and a mounted tip-based measuring system (left) and detail view of the tip-based measuring system (right) [2].

For positioning, this machine uses a planar direct drive system [11]. There, the movement of the machine table is realised by linear actuators, which are arranged in a single plane and consist of two fixed flat coils which act on a magnet array. By using three planar air bearings as guiding system, a nearly frictionless motion of the table is possible. With the use of three laser interferometers the position of the slider can be tracked, which are fibre-coupled to one stabilised He-Ne-laser. To enable higher stability in terms of thermal stability and noise, differential plane-mirror interferometers are used. For length measurement, two interferometers are arranged in such a way, that their axes intersect virtually in the tip of the tool to guarantee that the Abbe-principle is fulfilled in x- and y-direction [1], which leads to an Abbe-offset l A b b e , i :

Due to this mechanical set-up, there is no guiding of the slider for the rotation around the z-axis. Therefore, this rotation has to be closed-loop controlled. For that reason, a third interferometer is included, which is detecting the rotation of the slider around the z-axis. Both interferometer arms, the measuring and reference arm, directly probe the moving mirror. The rotation φ of the slider around the z-axis leads to an Abbe-error Δ l i , which can be described by the following equation

This concept of interferometers offers the possibility to detect the rotation of the stage independently from the lateral movement [12]. The actuators as well as the interferometers for traking the position of the slider have an angle of 120° between each other (see Fig. 3), which leads to a circular positioning range of Ø 100 mm.

Figure 3

Mechanical and Interferometer set-up of the NFM-100. The position of the slider is tracked by three laser-interferometers in a 120° arrangement. The slider is driven by linear actuators, which consist of flat coils and a magnet array.

The slider of the nanopositioning machine is built on a granite base, which can be seen on the left side in Figure 2. Additionally, the slider and the mirrors of the interferometers are made of quartz. Thus, the thermal and mechanical drift can be reduced to a minimum. Due to the fact, that the granite base is not ideally flat, the deviation has to be compensated by the measuring tool. To allow ultra-precise positioning, the granite base is installed on a vibration-isolated base. In addition, an acoustic noise suppression enclosure is built around the complete machine. To avoid thermal drift, the entire set-up is located in an air-conditioned room.

In order to be able to operate production and measurement processes flexibly and in a short time, control parameters must be readjusted quickly and therefore a rapid control system was applied [13]. This system can also be adapted with different plug-in cards for numerous purposes, so the system can be modified for optimisation, testing or adapting other measurement/fabrication tools [1].

3.2 Measurement and fabrication tool

Usually, a laser focus sensor is used as standard tool for large-area measurements in NPMMs [14], but there is a limit because of its lateral resolution. In contrast to this, tip-based measurement systems offer nanometre resolution. Currently, an AFM is installed as a measuring system, shown on the right side of Figure 2. For the realisation of large-area nanofabrication in the sub-10 nm range, this AFM-/FESPL-system with the nanopositioning machine represents an ideal possibility. The ability to measure and write with these microcantilevers also provides the opportunity to use this tool for both pre- and post-production inspection of structures. New lithographic technologies can be verified on large area to check suitability for wafer size.

4.1 Static operation of the NFM-100

First the position stability of the machine itself is investigated. To show the position stability, the position noise was recorded for a duration of 5 s with a sampling frequency of 10 kHz, which is depicted in Figure 4. The standard deviation of 0.2417 nm shows, that the NFM-100 in closed loop control mode has an outstanding position stability.

Figure 4

Position noise of the machine table with a sampling frequency of 10 kHz. The position was sampled for a duration of 5 s. The standard deviation of the position is 0.2417 nm.

4.2 Dynamic operation of the NFM-100

Figure 5

Positioning of the slider in 100 pm steps with a velocity of 500 pm/s with raw data and data filtered by a moving average filter of a length of 2500. The sampling frequency of the position is 10 kHz.

The resolution of the nanopositioning machine was investigated with steps in the picometre range. As shown in Figure 5 picometre steps were performed with a step distance of 100 pm in x-direction and a velocity of 500 pm/s. The steps in the pm-range can be made visible clearly by subsequent filtering of the measured data. A moving average filter of a length of 2500 is applied. For nanofabrication in the sub-nm range and also for measuring, it is of importance to position the slider with high precision.

Furthermore, it is particularly important in the manufacturing of micro- and nanostructures to guarantee a small deviation from the target position of certain movements over the entire range of motion. For this purpose, circles with a radius of 5 nm were driven at different positions with the machine table. In Table 1 the standard deviation of circles at different positions with a velocity of 10 nm/s are shown. The measurements were repeated 10 times. In general, the standard deviations of the driven circles show, the motion of the nanopositioning machine is stable and below 271 pm over the whole range of motion. The investigations on the repeatability of measurements is important for further processes of nanostructuring on wafer size, such as mix-and-match [15], step and repeat [16] or for any change of tools [17]. In Figure 6 a driven circle at position [0,0] is shown. Additionally, the ideal target circle is depicted in the figure. These results show, the resolution is adequate for sub-10 nm structuring.

Figure 6

Plot of a circular trajectory with 10 revolutions with a radius of 5 nm driven by the machine table with a velocity of 10 nm/s and an acceleration of 200  nm / s 2 at position [0,0].

Figure 7

Linescan of a holographic grating with a structure depth of approximately 50 nm measured by the installed AFM in combination with the NFM-100. The z-signal was measured by the piezo-scanning unit of the AFM, while the sample was moved by the machine in x-direction. The measurement was repeated 10 times and results in a standard deviation of 2.7709 nm in z-direction. The left plot shows the complete scan over a length of 100 µm. In the right part, a detail of 5 µm is shown.

4.3 Combined measurement of NFM-100 and AFM

To show the performance of the combined measurement of the machine and the mounted AFM, a series of measurements of a holographic grating with a period of approximately 420 nm are made. The movement of the sample in x- and y-direction was realised by the slider of the NFM-100. The controlling of the z-direction was performed by the piezo-scanning unit of the AFM, while in x- and y-direction the microcantilever was in a fixed position. The holographic grating was measured 10 times over a length of 100 µm with a velocity of 1 µm/s. The results of these experimental investigations are shown in Figure 7. The standard deviation of these measurements is 2.7709 nm. Additionally, a section of 5 µm is shown. Furthermore, the movement of the slider during these measurements of the holographic grating was tracked by the differential length interferometers. The results of the driven lines are shown in Figure 8. As in Figure 7, the measurement was repeated 10 times. The standard deviation of the straight line movements of 100 µm is 1.0905 nm.

Figure 8

Line movement with signals of the x- and y-interferometers over a length of 100 µm in x-direction driven with the NFM-100 including acceleration and deceleration phase. The measurement was repeated 10 times and results in a standard deviation of 1.0905 nm. Sampling frequency of the position of the machine table is 10 kHz.

To demonstrate the performance over a long range of motion, a linescan of the holographic grating (same as in Figure 7) with a length of 30 mm and a velocity of 5 µm/s was measured without stitching in Figure 9. The tilt of the sample was removed from the original data. Additionally, a section of 5 µm of this linescan is shown (Fig. 9). These presented measurements show the capability of the combination of these systems.

Figure 9

Linescan of a holographic grating (same as shown in Fig. 7) over a length of 30 mm measured by the NFM-100 and the tip-based measuring system (left) and section of 5 µm of the 30 mm-Linescan (right). Measured with a velocity of 5 µm/s.