基于光频梳的太赫兹器件面形测量技术研究

Objective Microwave and terahertz (THz) devices, such as antennas and reflectors, are useful in communication, navigation, and remote sensing. The laser tracking and measuring technology plays an important role in characterizing microwave/THz devices, such as surface measuring. Laser tracking and measuring, which is based on laser interferometry, provides high spatial resolution and accuracy for precise surface topography measurements. However, it has limitations such as a limited dynamic range, a low update rate, and difficulty measuring absolute distance. A coherent light source with significantly advanced laser sensing and ranging is provided by an optical frequency comb, which is made up of a large number of equally spaced optical frequency lines (or comb teeth) . Dual-comb-ranging interferometry, which uses two asynchronous optical combs to measure absolute distance, provides high precision, large ambiguity range, and high speed. This method, which does not use moving mirrors, demonstrates significant advantages in laser ranging and holographic imaging. The measurement process, however, can be time-consuming due to a camera's limited frame readout speed. As a result, the benefits of dual-comb technology have not been fully exploited in imaging applications. Furthermore, dual-comb interferometry is yet to be used to characterize THz devices. In this paper, the dual-comb technique, combined with high-speed two-dimensional scanning mirrors, is applied to the surface measuring of THz devices. The measurement time per pixel for a three-dimensional (3D) image with 2.5 x 10(5) pixel is about 8 mu s. The longitudinal measurement error is 1.3 mu m (reduced to 5 nm at 0.1 s integration time), and the ambiguity range can reach 300 m. The experimental results show that dual-comb imaging is a useful and reliable tool for measuring the surface of microwave/THz devices with high precision. Methods Two near-infrared electro-optic combs (EOC 1 and EOC 2) are used for constructing a dual-comb interferometer (Fig. 1). Each EOC contains a Fabry-Perot phase modulator driven by a radio frequency (RF) signal generator. The repetition frequencies of the two combs are set to f(r1) =25 GHz and f(r2) = f(r1)( )+ Delta f=25 GHz 0.5 MHz. Thus, the frequency of the n th comb line of the two combs can be expressed as f(cw) + n . f(r1) and f(cw) + n . f(r2), respectively, where f(r2) = f(r1) + Delta f, f(cw) is the repetition rate of the continuous laser, and n is an integer, ranging from N to N. Here, 2N+1 are the total number of comb lines for a single comb. The spectral coverage of the two combs is similar, spanning from 1542 nm to 1568 nm at - 20 dB (Fig. 1), corresponding to 3.22 THz in spectrum width and 129 comb lines. The EOC 1 has an output power of 11 mW, while the EOC 2 has a power of 2.7 mW. The detection light is sent to a high-speed scanning galvanometer (GM) after passing through a fiber optic circulator (Cir) and a collimator (Col), and then hits the surface of a sample (S) that is placed on a 3D translation platform for adjusting the longitudinal (z-axis) displacement and expanding the imaging view. The reflected light propagates along the original beam path and then combines the local oscillator light (EOC 2) with a 50/50 fiber beam splitter (BS). The common-mode noise is then suppressed using the balanced photo-detector (BPD 1, bandwidth of 1.6 GHz). A data acquisition card (DAC) records the detection signal at a sample rate of 2.5 x 10(8) Sample/s. Note that, to avoid frequency aliasing in dual-comb measurement, an acoustic-optical modulator (AOM) is used to introduce a frequency offset at f(AOM) 40 MHz) to the local comb (EOC 2). Thus, the beat frequency signal in the RF domain can be expressed as f(AOM) + n . Delta f. It should also be noted that the optical fibers of the reference and measuring arms have similar lengths when placed in the same environment to minimize the fluctuation introduced by the fibers. Results and Discussions An ellipsoidal THz reflector' s imaging results (sample 1) are shown (Fig. 7). The sample projection plane has a diameter of 100 mm and a depth of 6 mm. The mean signal-to-noise ratio of each pixel is about 2000 +/- 400, and the measurement error in the z-axis direction is about 1.3 m, depending on the smoothness and shape of the sample surface (Fig. 7). We perform 3D surface fitting based on the data and the results ( Fig. 7) show that the sample surface flatness is 34.78 mu m with a fitting error of 19.84 mu m. Another sample, a planar THz antenna (sample 2), with a diameter of 80 mm and a concave pit in the center with a diameter of 20 mm and a depth of 0.3 mm, is measured (Fig. 8). A flat area of the surface is fitted, yielding the line smoothness of 11.2 mu m, the plane flatness of 5.85 mu m, and the fitting error of 0.02 mu m(2) (Fig. 8), which are consistent with the results obtained by a contour measuring system, e. g. , plane flatness of (7.1 +/- 2.0) mu m. The uncertainty of a single-point measurement in the experiment is within +/- 10 nm (measurement time of 0.1 s) , but the scanning motion of the galvanometer introduces additional fluctuations into the measurement, resulting in a measurement error of (1.3 +/- 0.2)mu m (acquisition time of 8 mu s) for 3D imaging. Nonetheless, the dual-comb method holds great promise for the surface measuring of microwave/THz devices due to its high data update rate (125 kHz) and micrometer-level precision. Conclusions The dual-comb time-of-flight method is used in this paper to measure the three-dimensional shapes of small aperture terahertz antennas and reflectors. The measurement time for a single pixel in the experiment is 8 mu s, and the longitudinal measurement error is 1.3 mu m. When compared to the traditional laser interferometry, the dual-comb method has the advantages of a large ambiguity range, high measurement accuracy, and a fast data update rate, making it promising for surface measurements of terahertz reflectors and microwave antennas.