We constructed a double-modulation, reflection-type terahertz (THz) ellipsometer for precise measurement of the thickness of a paint film which is coated on a metal surface and which is not transparent to visible or mid-infrared light. The double-modulation technique enabled us to directly obtain two ellipsometric parameters, Δ(ω) and Ψ(ω), as a function of angular frequency, ω, with a single measurement while reducing flicker noise due to a pump laser. The bias voltage of a photoconductive antenna (PCA) used as a THz pulse emitter was modulated at 100 kHz, and a first lock-in amplifier (LA1) was connected to the output of an electro-optic (EO) signal-sampling unit. In addition, a wire-grid polarizer (WGP) was rotated at 100 Hz to conduct polarization modulation with a frequency of 200 Hz. The output signal from LA1 was fed into a second lock-in amplifier (LA2) that worked in synchronization with the rotating WGP (RWGP). By operating LA2 in a quadrature phase-detection mode, we were able to obtain in-phase and out-of-phase signals simultaneously, from which the two ellipsometric parameters for an isotropic sample could be derived at the same time while cancelling common-mode noise. The lower detection limit of the thickness measurement and the relative standard deviation (RSD) of a black paint film coated on an aluminum substrate were 4.3 µm and 1.4%, respectively. The possibility of determining all elements of the Jones matrix for an anisotropic material is also discussed.
© 2014 Optical Society of America
The terahertz (THz) frequency band, ranging from around 0.1 THz to several tens of THz, is located at the boundary region between the electric- and optical-frequency regions, or at the intermediate region between the microwave and the mid-infrared wavelength regions. THz waves exhibit promising properties, such as propagation in free-space, high transmittance through materials, low scattering, and non-invasiveness to living bodies. These properties are of use not only for nondestructive testing of industrial materials but also for electronics, photonics, materials science, and medical applications . One of the most practical applications in the industrial field is non-contact and in-line measurement of the thickness of a paint film on cars, airplanes, ships, and other transportation means. Although such measurement would appear to be simple, it is not always easy to determine coating thicknesses of several micrometers or less with precision. The reason is that the paint material is usually not transparent to light in the optical-frequency region. Conventional methods , such as ultrasonic methods, eddy-current methods, and electromagnetic-induction methods, are hard to use for measuring a wet paint film; these methods are “destructive” ones, which means that one has to contact the sensor probe against the sample.
In light of this situation, THz-wave techniques have been introduced. For thickness measurement, a time-of-flight (TOF) technique based on THz time-domain spectroscopy (THz-TDS) has been developed . In THz-TDS, a photoconductive antenna (PCA) is used as an excitation light source, and the time difference of the two THz pulses that are reflected from the surface of a paint film and the film–metal interface is measured. Therefore, it is difficult to apply the TOF method to the measurement of a thin paint coating because of the large temporal overlapping of the two pulses. For example, when the pulse width of the THz pulse and the group refractive index are 1.2 ps and 1.6, respectively, the measurable thickness of the paint film is 100 µm or more. Although a numerical deconvolution technique and a mathematical multivariate method have been proposed to determine the thickness of such thin films [4–6], they are still difficult to apply to the measurement of film thicknesses of less than 20 µm. In addition, one has to know the frequency-dependent group-refractive-index of the paint film in advance.
For such a situation, the ellipsometric technique seems promising . Although a THz ellipsometer employing a backward wave oscillator  and one using synchrotron radiation  as the THz light source have reported, simultaneous broadband measurement is difficult. Therefore, a frequency-domain (FD)-THz ellipsometer is usually employed . With a FD-THz ellipsometer, one can obtain the spectroscopic information all at once; a time-series waveform of the electric-field strength of the THz pulse is Fourier transformed, which allows the frequency dependence of the ellipsometric parameters, ∆ and Ψ, to be calculated directly .
In an FD-THz ellipsometer, an essential point for precise measurement is to enhance the signal-to-noise ratio (SNR). At present, however, SNR enhancement is limited by low-frequency flicker noise, due to the femtosecond laser used as a pump (and/or a probe) for the PCA used as the THz pulse emitter (and/or detector). Therefore, a lock-in light-detection method is commonly used, where the THz pulse trains incident on the detector should be modulated at as high a frequency as possible.
In the ellipsometric measurements, one has to derive the complex-amplitude reflection ratio ρ, which is defined as the ratio of the complex-amplitude reflectivity of the p-polarized light to that of the s-polarized one. Therefore, two measurements are required, and two scans of a mechanical time-delay stage are necessary , which is apt to introduce errors due to changes in the sample and/or the instrument conditions as time passes. In order to eliminate such errors, a single-scan measurement is desirable. Therefore, a polarization modulation technique in combination with a quadrature-phase-detection method should be introduced. As described later, a two-phase lock-in amplifier (LA) operated in the quadrature-phase-detection mode enables us to obtain two signals simultaneously: one is an in-phase signal and the other is an out-of-phase signal, from which we can derive the two ellipsometric parameters simultaneously. Cancelation of the common-mode noise in the measurement is another bonus. One way to conduct polarization modulation in the THz frequency range is to use a mechanically rotating wire-grid polarizer (RWGP). However, the upper modulation frequency is limited to 30 Hz [11–13]. At such a low frequency, the flicker noise cannot be eliminated sufficiently.
To this end, in order to construct an FD-THz ellipsometer, we have to employ a double-modulation method; in this method, combined use of high-frequency modulation of the bias voltage of the PCA and relatively low-frequency polarization modulation by the RWGP is indispensable. The double-modulation method retains the advantages of both methods while compensating for their disadvantages. To carry out double modulation, the output signal from the detector should be fed into two LAs connected in tandem: one LA works in synchronization with the modulation frequency of the bias voltage for the PCA with a time constant τ1, and the other LA works in synchronization with the polarization modulation frequency with a time constant τ2. The two time constants should be set so that τ1 < τ2, and τ2 should be at least one order of magnitude larger than τ1. Such a double-modulation technique has been used for spectroscopy in the optical-frequency region  and has also been considered for the THz region . A trial to measure polarization Stokes parameters also has been reported recently . However, the former system is based on the THz-TDS method for obtaining the optical constants of materials through transmission measurements. The latter one is a transmission type-polarimeter for measuring the optical constants. Neither instrument was intended for thickness measurement of paint films, and the instruments were not described in sufficient detail. In addition, the modulation frequency was not so high, and quadrature-phase detection has not been introduced yet. As far as polarization measurements are concerned, new ideas have been proposed, such as a PCA emitter (and/or detector) that has multiple electrodes [16–18]. A method that uses a rotating electro-optical (EO) crystal has also been reported . Although the proposed methods work well, specially designed components or devices are required.
In the present paper, we focus on the construction of a double-modulation reflection-type THz ellipsometer for measuring the thickness of a paint film coated on a metal surface. We also discuss ellipsometric measurement of Jones matrixes for anisotropic materials.
Figure 1 shows a schematic diagram of the double-modulation reflection-type THz ellipsometer, and Fig. 2 shows a timing diagram explaining its operation. A laser beam from a mode-locked Ti:sapphire laser (Spectra-Physics, Mai Tai; wavelength, 800 nm; pulse width, 100 fs; repetition frequency, 80 MHz) was divided into a pump beam and a probe beam by a 9:1 beam splitter (BS). The pump beam, shown in Fig. 2(a), was focused on the PCA (Gigaoptics GmbH, TERA-SED) for emitting THz pulses. The bias voltage of the PCA was modulated by a 20 Vpp modulation signal, as shown in Fig. 2(b), where the duty ratio and the frequency of the modulation signal were 50% and 100 kHz, respectively. The THz pulse trains emitted from the PCA, as shown in Fig. 2(c), were collimated by a lens (TL1; focal length, 50 mm) and passed through four wire-grid polarizers (WGP1–4; frequency range, 0.1–1.5 THz; wire diameter, 50 µm; spacing, 125 µm) sequentially and a rotational WGP (RWGP; frequency range, 0.1–1.5 THz; wire diameter, 50 µm; spacing, 125 µm; motor, model HM2669E18H, Technohands Co., Ltd.). The orientation angles (or transmission axes) of WGP1, WGP2, RWGP, WGP3, and WGP4 are indicated by θ1, θP, θR, θA, and θ4, respectively, where θ = 0° means p-polarized light (parallel to the plane of the figure), and θ = 90° means s-polarized light (perpendicular to the plane of the figure). For the measurement of the anisotropic sample, we set θ1 = 45° and θ4 = −45°. The role of WGP1 was to fix the polarization of the incident light at 45°. WGP4 was used for EO sampling, described later.
The THz pulses were focused on the sample with an incident angle ϕ of 60° by an off-axis parabolic mirror (OAP; focal length, 101.6 mm). The THz pulses reflected from the sample were collimated again by another OAP and then passed through the RWGP, which was rotated at a frequency f = 100 Hz. Thus, the frequency of the polarization modulation was 2f = 200 Hz. Figures 2(d) and 2(e) show the waveforms of the THz pulse trains after passing through the RWGP and WGP3, respectively. To clarify the illustration, the waveforms drawn in Fig. 2 are schematic ones obtained with no paint coating on the metal and θP = θA = 0°. Typical orientation angles of the RWGP as time passed are depicted at the lower part of the waveform in Fig. 2(d).
In order to detect the pulsed THz light, we used the EO sampling technique [20, 21]. The THz pulses were focused into a ZnTe EO crystal. The probe beam reflected by the BS, after being polarized, was also incident on the same crystal. A slight change in its polarization state by the Pockels effect due the electric field of the THz pulses was detected by a balanced light detector (New Focus Corp., Model 2007) followed by a quarter-wave plate (QWP) and a Rochon prism (RP). The output signal from the detector was fed into the LA1 (NF Corp., LI5640), whose reference signal was derived from the PCA bias modulation-signal shown in Fig. 2(b). The output signal from the LA1, shown in Fig. 2(f), was fed again into the LA2 (NF Corp., LI5640), which worked in synchronization with the phase modulation signal, so that quadrature phase detection was carried out. Figures 2(g) and 2(h) show the in-phase and out-of-phase reference signals used for the quadrature-phase detection procedure, respectively.
The in-phase signal obtained from the LA2 was proportional to the difference between the two electric-field strengths when θR = 0° and when θR = 90° for the output of LA1 shown in Fig. 2(f). Similarly, the out-of-phase signal was proportional to the difference between the two electric-field strengths when θR = 45° and when θR = −45°. The two component signals were recorded as a function of the time delay given by the movement of the mechanical stage. Finally, by Fourier transformation, we could obtain two complex electric-field components in the frequency domain: the in-phase component, Ein(ω), and the out-of-phase component, Eout(ω).
In the present system, the incident angle was fixed at ϕ = 60°, which was near the Brewster angle of ϕ = 61.3° for the air/paint film interface, where the refractive index of the paint film at f = 1.02 THz was assumed to be n = 1.83. The time span of the optical delay was 27.3 ps, and its step interval was 53.4 fs, which corresponded to a frequency resolution of ∆f = 36.6 GHz. The overall frequency range in the present system was limited by that of the WGP, which was from 0.1 to 1.5 THz. The time constants of the LA1 and the LA2 were τ1 = 300 µs and τ2 = 300 ms, respectively.
3. Ellipsometric measurements
3.1 Quadrature signal detection
In this section, we derive mathematical expressions for the output signals obtained from the LA2 in the quadrature-phase detection method for several combinations of θP and θA; these expressions will be used in the next section.
The Jones polarization-matrix for the ellipsometric measurement is given byEq. (1), E0(ω) is the electric-field strength at the angular frequency ω; Ep(θR, ω) and Es(θR, ω) stand for the p- and the s-polarized light components, respectively, when θ = θR; and Pθ (θ = θP, θR, or θA) represents the Jones matrix for the polarizer with orientation angle θ, which is given by
The Jones matrix, S, for the sample is given by
In Eq. (1), Es(θR, ω) = 0 when θA = 0þ. Then, the output signal from the LA1 is proportional to Es(θR, ω) only. Also, Ep(θR, ω) = 0 when θA = 90°. Then, the output signal from the LA1 is proportional to Es(θR, ω) only. Therefore, the in-phase signal Ein(ω) is given byEq. (4), when θA = 0°, i = p, and when θA = 90°, i = s. Similarly, the out-of-phase signal Eout(ω) is given byTable 1 summarizes the calculated results. Note that the non-diagonal elements of the Jones matrix, S, become zero for the case of an isotropic sample: = = 0.
3.2 Measurement of an isotropic sample
In reflection-type ellipsometry of an isotropic sample, the ratio ρ is measured, which is the ratio of the complex amplitude reflectivity of the p-polarized light to that of the s-polarized light:Table 1:Eq. (6) as a function of ω.
Similarly, when (θP, θA) = (45°, 90°), the following relation can be derived:22].
3.3 Measurement of an anisotropic sample
For the case of an anisotropic sample, the following ratio ρij should be measured:Table 1:
The four elements of the Jones matrix can be determined by another two combinations of setups: (i) (θP, θA) = (0°, 0°) and (90°, 0°), and (ii) (θP, θA) = (0°, 90°) and (90°, 90°). In these two setups, although the condition is still required, the WGP4 can be removed. This is because the orientation angle θA is kept constant for the two sequential measurements.
4.1 Measurement of the complex refractive index of a Si substrate
In order to compare the measurement precision of the proposed double-modulation THz ellipsometer with that of a single-modulation one without the polarization modulation technique, we measured the complex refractive index, , where n(ω) is the refractive index and κ(ω) is the extinction coefficient, of a high-purity, polycrystalline Si substrate. The value of for such an isotropic sample without multiple reflections can be obtained from the following relation :
Figures 3(a) and 3(b) show n(ω) and κ(ω), respectively, which were obtained from the double-modulation ellipsometer. Error bars attached to the plot indicate the standard deviation obtained from twenty repeated measurements. Figures 3(c) and 3(d) show those measured by the single-modulation ellipsometer. The horizontal solid lines in the figure show a value reported in the literature for undoped Si , namely, . The measured values agreed well with the literature ones. The averaged relative standard deviation (RSD) over the frequency range from 0.1 to 1.5 Hz was 2.6%. By introducing the double-modulation technique, the RSD value was reduced by two times in comparison with that without the double-modulation technique.
4.2 Measurement of the complex refractive index of a paint coating
Next, we measured a black lacquer paint film coated on an Al substrate. Because the complex refractive index of the paint film in the THz frequency range is not known, we prepared a similar but thicker paint film to measure the value accurately. The thickness of the coating was d = 170 ± 1.5 µm, which was obtained from thirty measurements by using an eddy-current thickness meter (LZ-300C, Ketto Kagaku).
Figures 4(a) and 4(b) show the in-phase time-domain waveform, Ein(ω), and the out-of-phase time-domain waveform, Eout(ω), respectively, which were obtained from the setup with (θP, θA) = (45°, 0°). Figures 4(c) and 4(d) show the amplitude and the phase spectrum obtained from Fourier transformation of Ein(ω) and Eout(ω), respectively, from which the ellipsometric parameters Ψ(ω) and ∆(ω) were calculated, as shown in Figs. 4(e) and 4(f), respectively. Finally, the real part, n(ω), and the imaginary part, κ(ω), of the complex refractive index, , of the paint film were estimated from Ψ(ω) and ∆(ω), as shown in Figs. 4(g) and 4(h), respectively. Here, we used a three-layer (air/film/Al substrate) optical model for the estimation. In the calculation, the complex refractive index of the Al substrate was derived from the Drude model :
4.3 Measurement of the thickness of a thin paint coating
In the previous section, we obtained the complex refractive index of the paint film as a function of ω. Using these values, we measured the thicknesses of thin paint coatings. The left column and the right column in Fig. 5 show Ψ(ω) and Δ(ω), respectively, which were measured for five thicknesses of paint coatings: (a) and (b) d = 102 µm, (c) and (d) d = 20.0 µm, (e) and (f) d = 10.4 µm, (g) and (h) d = 5.3 µm, and (i) and (j) d = 4.3 µm. The solid lines in the individual plots are mathematically fitted results obtained by using a three-layer model. Table 2 summarizes the estimated thicknesses. For reference, thickness values measured by an eddy-current meter thirty times are also shown. Both values agreed well. For the thickness of the 4.3 µm-thick paint film, the RSD in twenty repeated measurements using the double-modulation THz ellipsometer was 1.4%.
We constructed a double-modulation reflection-type THz ellipsometer for measuring the thickness of a paint film coated on a metal surface. The double-modulation technique, together with a quadrature-phase detection method using a two-phase lock-in amplifier, enabled us to obtain the ellipsometric parameters Ψ(ω) and Δ(ω) at the same time with a single measurement. For a 4.3 µm-thick paint coating, the relative standard deviation (RSD) value in twenty measurements was 1.4%. We also discussed the possibility of measuring all elements of the Jones matrix for an anisotropic material. Although we have measured single-layer films with good precision, in future work we will attempt to measure the thicknesses of multi-layer coatings and to analyze inhomogeneous paint films that include impurity materials such as metal flakes.
This work was supported by a Grand-in-Aid for Scientific Research B (No. 26289066) from the Japan Society for the Promotion of Science. The authors are grateful to Dr. Shinichi Watanabe of Keio University for fruitful discussions on rotational wire-grid polarizer. We also thank Mr. Hiroyasu Furukawa of Nippon Steel and Sumitomo Metal Corporation, Japan, for preparation of the paint samples.
References and links
1. D. Mittleman, Sensing with THz Radiation (Springer, 2003).
2. D. E. Bray and D. McBride, Nondestructive Testing Techniques (John Wiley & Sons, 1992).
3. T. Yasui, T. Yasuda, K. Sawanaka, and T. Araki, “Terahertz paintmeter for noncontact monitoring of thickness and drying progress in paint film,” Appl. Opt. 44(32), 6849–6856 (2005). [CrossRef] [PubMed]
4. T. Yasuda, T. Iwata, T. Araki, and T. Yasui, “Improvement of minimum paint film thickness for THz paint meters by multiple-regression analysis,” Appl. Opt. 46(30), 7518–7526 (2007). [CrossRef] [PubMed]
5. T. Yasui, Y. Kabetani, Y. Ohgi, S. Yokoyama, and T. Araki, “Absolute distance measurement of optically rough objects using asynchronous-optical-sampling terahertz impulse ranging,” Appl. Opt. 49(28), 5262–5270 (2010). [CrossRef] [PubMed]
6. T. Iwata, S. Yoshioka, S. Nakamura, Y. Mizutani, and T. Yasui, “Prediction of the thickness of a thin paint film by applying a modified partial-least-squares-1 method to data obtained in terahertz reflectometry,” J. Infrared Millim. Terahertz Waves 34(10), 646–659 (2013). [CrossRef]
7. K. Vedam, “Spectroscopic ellipsometry: a historical overview,” Thin Solid Films 313–314, 1–9 (1998). [CrossRef]
8. T. Hofmann, U. Schade, C. M. Herzinger, P. Esquinazi, and M. Schubert, “Terahertz magneto-optic generalized ellipsometry using synchrotron and blackbody radiation,” Rev. Sci. Instrum. 77(6), 063902 (2006). [CrossRef]
9. T. Hofmann, C. M. Herzinger, T. E. Tiwald, J. A. Woollam, and M. Schubert, “Hole diffusion profile in a p–p+ silicon homojunction determined by terahertz and midinfrared spectroscopic ellipsometry,” Appl. Phys. Lett. 95(3), 032102 (2009). [CrossRef]
10. T. Nagashima and M. Hangyo, “Measurement of complex optical constants of a highly doped Si wafer using terahertz ellipsometry,” Appl. Phys. Lett. 79(24), 3917–3919 (2001). [CrossRef]
11. D. K. George, A. V. Stier, C. T. Ellis, B. D. McCombe, J. Cerne, and A. G. Markelz, “Terahertz magneto-optical polarization modulation spectroscopy,” J. Opt. Soc. Am. B 29(6), 1406–1412 (2012). [CrossRef]
13. D. J. Aschaffenburg, M. R. C. Williams, D. Talbayev, D. F. Santavicca, D. E. Prober, and C. A. Schmuttenmaer, “Efficient measurement of broadband terahertz optical activity,” Appl. Phys. Lett. 100(24), 241114 (2012). [CrossRef]
14. W. G. Golden, D. S. Dunn, and J. Overend, “A method for measuring infrared reflection—Absorption spectra of molecules adsorbed on low-area surfaces at monolayer and submonolayer concentrations,” J. Catal. 71(2), 395–404 (1981). [CrossRef]
15. S. P. Mickan, K.-S. Lee, T.-M. Lu, J. Munch, D. Abbott, and X.-C. Zhang, “Double modulated differential THz-TDS for thin film dielectric characterization,” Microelectron. J. 33(12), 1033–1042 (2002). [CrossRef]
16. E. Castro-Camus, J. Lloyd-Hughes, M. B. Johnston, M. D. Fraser, H. H. Tan, and C. Jagadish, “Polarization sensitive terahertz detection by multicontact photoconductive receivers,” Appl. Phys. Lett. 86(25), 254102 (2005). [CrossRef]
17. Y. Hirota, R. Hattori, M. Tani, and M. Hangyo, “Polarization modulation of terahertz electromagnetic radiation by four-contact photoconductive antenna,” Opt. Express 14(10), 4486–4493 (2006). [CrossRef] [PubMed]
18. H. Makabe, Y. Hirota, M. Tani, and M. Hangyo, “Polarization state measurement of terahertz electromagnetic radiation by three-contact photoconductive antenna,” Opt. Express 15(18), 11650–11657 (2007). [CrossRef] [PubMed]
19. N. Yasumatsu and S. Watanabe, “Precise real-time polarization measurement of terahertz electromagnetic waves by a spinning electro-optic sensor,” Rev. Sci. Instrum. 83(2), 023104 (2012). [CrossRef] [PubMed]
20. Q. Wu and X.-C. Zhang, “Free-space electro-optic sampling of terahertz beams,” Appl. Phys. Lett. 67(24), 3523–3525 (1995). [CrossRef]
21. Q. Wu, T. D. Hewitt, and X.-C. Zhang, “Two-dimensional electro-optic imaging of THz beams,” Appl. Phys. Lett. 69(8), 1026–1028 (1996). [CrossRef]
22. R. M. A. Azzam and N. M. Bashara, Ellipsometry and Polarized Light (North-Holland Personal Library, 1989).
23. J. D. Holm and K. S. Champlin, “Microwave conductivity of silicon and germanium,” J. Appl. Phys. 39(1), 275–284 (1968). [CrossRef]
24. M. A. Ordal, L. L. Long, R. J. Bell, S. E. Bell, R. R. Bell, R. W. Alexander Jr, and C. A. Ward, “Optical properties of the metals Al, Co, Cu, Au, Fe, Pb, Ni, Pd, Pt, Ag, Ti, and W in the infrared and far infrared,” Appl. Opt. 22(7), 1099–1119 (1983). [CrossRef] [PubMed]