INFORMATION TECHNOLOGY

Ultra-fast multi-dimensional imaging: Simultaneously measure coordinates of more than 1,000


Recently, Jungwon Kim’s team from the Korea Advanced Institute of Science and Technology demonstrated a camera capable of simultaneously measuring the time-of-flight variation of more than 1,000 spatial coordinates by using electro-optical sampling using a frequency comb. This research result can accurately and quickly image complex structures and dynamic processes in 3D equipment, which has high research and application value.

The article was published in Light: Science & Applications in the top international academic journal “Massively parallel electro-optic sampling of space-encoded optical pulses for ultrafast multidimensional imaging”, Yongjin Na is the first author of the paper.

Optical imaging and measurement technology is very important in modern science and technology, and its application range and its wide range, such as vibration mode measurement, in vivo biomedical imaging and the development of autonomous driving technology, are inseparable from the progress of optical imaging and measurement technology. In particular, the rapid and accurate imaging of the surface contours of micro- and nanoscale devices plays a key role in studying the static and dynamic properties of such devices.

In terms of static properties, dimensional measurements with higher dynamic range and higher data volumes over larger wafer areas are becoming increasingly important to the semiconductor industry. Until now, the use of interferometry and confocal microscopy to achieve surface measurements has been used, however, these methods still have considerable limitations in terms of measurement range (typically less than a few microns) and speed (usually hundreds of seconds to complete imaging).

In terms of dynamic properties, accurate characterization of vibration and dynamic behavior in micro- and nanomechanical devices is critical to understanding fundamental physics and advancing its applications. In particular, the newly discovered nonlinear, transient, and complex mechanical dynamics, such as non-resonant vibrations in micro- and nanomechanical resonators, pulsed optomechanics, and more, require real-time surface deformation imaging with finer axial and lateral resolution, higher speeds, and higher dynamic ranges. In this regard, coherent interferometers and white light interferometers are widely used due to their nanoscale axial resolution and reliability. However, these techniques also have limitations such as submicron blurred range and low dynamic imaging rates.

Therefore, whether static or dynamic microscopic surface imaging, there are defects in imaging quality and rate, and if this problem can be solved, it will undoubtedly make optical measurement technology and micromechanical technology a considerable step forward.

In this study, the researchers built a novel line-scan time-of-flight (TOF) imaging technique based on electro-optical sampling that captures the static and dynamic characteristics of microdevices with high dynamic range (Figure 1). This method achieves high pixel rates (up to 260 megapixels per second), high axial resolution (down to 330 pm) (Figures 2, 3), and high dynamic range (up to 126dB) (Figure 4), as well as simultaneous detection of TOF ranges of more than 1000 spatial coordinates over a field of view (FOV) of several millimeters. This unprecedented performance advantage is extremely important for the development of the field of micrometrology not only for fast and precise imaging of complex structures without much prior knowledge, but also for the observation of fast and non-repetitive mechanical movements in miniature devices and mechanical resonators in real time.

Figure 1: The working principle of a line scan TOF camera based on electro-optical sampling. Use mode-locked Er-fiber oscillator as the source for optical frequency combs. An ultra-low jitter photocurrent pulse generated by a MUTC photodiode is used to generate a time gauge. During target imaging, light pulses are extended and spectrally dispersed for space-wavelength encoding. After reflection from the target object, the TOF-encoded sub-pulses are collected and converted from TOF to intensity in EOS-TD. Finally, the EOS-TD output spectrum is analyzed using a line scan camera to simultaneously reconstruct TOF information for more than 1000 spatial points.

Figure 2: Analysis of axial and lateral resolution of the line scan TOF method. (a) Measurement of TOF accuracy, overlapping Allan deviation as a function of acquisition time. The three MUTC photodiode bias voltages of 4 V, 8 V, and 16 V (unsaturated and saturated camera conditions) are shown. Illustration: The normalized EOS-TD output relative to the relative time between the rising edge of the optical pulse and the photocurrent pulse shows the measurable ranges of 3 mm, 1.6 mm, 1.2 mm, and 0.4 mm for 4 V, 8 V, and 16 V (unsaturated) and 16 V (saturated) bias voltages, respectively. (b) TOF accuracy measurement at each pixel position at 10 ms acquisition time (16 V bias, unsaturated camera). (c) Beam profile measured when focusing at a focal length of 30 mm. (d) Microscopic imaging of resolution targets. Magnified images of groups 6 and 7 are shown in the upper right image (red box in the left panel). As shown by the single-line scan traces, the three bars of element 6 in group 6 have a contrast ratio of approximately 23%, resulting in a lateral resolution of approximately 114 lp/mm (4.38 μm).

Figure 3: 3D surface contour imaging results. (a) Surface contour imaging of two measuring blocks of the same material (chromium carbide). As shown in the cross-sectional view from points A to A, step heights of 300 μm can be clearly measured. The gray area represents the edge of the measuring block, where TOF is fuzzy due to reflections from both surfaces. The step height (between points I and II) was determined at 300.029 μm, the repetition error (the standard deviation value for 100 consecutive measurements at 100 μs acquisition time) was 31 nm, and the error from the calibration interferometer result was +31 nm. (b) Imaging results of different material assembly; Two steel measuring blocks attached to the ceramic optical plane. The measured 500 μm order (II – I) has a repetition error of 93 nm (100 μs acquisition time) and a -22 nm error of the calibration interferometer results. (c) Surface contouring of complex periodic structures (silicon sample coated with 100 nm thick silver). A pair of f = 60 mm lenses is used for better spatial resolution. The histogram of the TOF point in Region I shows an average height difference of 10.039 μm with an error of -14 nm compared to the confocal microscopy results. Illustration: Microscopic image of the sample (2.5X).

Figure 4: Dynamic imaging results. (a) Interaction between two mirrors with PZT attached. f = 75 mm lens for horizontal FOV of approximately 10 mm. The drive duration of the two PZTs is about 100 milliseconds and the latency is about 25 milliseconds. The figure below shows the TOF trace reconstructed at the beginning of modulation, interaction transient, steady-state, and modulation end. (b) Real-time observation of the bending modal shape of a MEMS bridge. 14 beams with a beam size of about 8 μm and a FOV of about 880 μm are incident along the long side of the bridge. The resonant motion of the first five bending modes (from 4.0 kHz to 80.9 kHz) was measured. TOF at 14 local locations is represented by red dots, and the TOF curve between the points is interpolated by the spline method. Imaging results from a scanning electron microscope are shown in the illustration.

The article was recently published in the top international academic journal Light: Science & Applications, entitled “Efficient dispersion modeling in optical multimode fiber”, with Szu-Yu Lee as the first author and Professor Martin Villiger as the corresponding author of the paper. (Source: LightScience Applications WeChat public account)

Related paper information:https://www.nature.com/articles/s41377-023-01077-7

Special statement: This article is reproduced only for the need to disseminate information, and does not mean to represent the views of this website or confirm the authenticity of its content; If other media, websites or individuals reprint and use from this website, they must retain the “source” indicated on this website and bear their own legal responsibilities such as copyright; If the author does not wish to be reprinted or contact the reprint fee, please contact us.



Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button