Terahertz-driven polymerization of resists in nanoantennas

Plasmon-mediated polymerization has been intensively studied for various applications including nanolithography, near-field mapping, and selective functionalization. However, these studies have been limited from the near-infrared to the ultraviolet regime. Here, we report a resist polymerization using intense terahertz pulses and various nanoantennas. The resist is polymerized near the nanoantennas, where giant field enhancement occurs. We experimentally show that the physical origin of the cross-linking is a terahertz electron emission from the nanoantenna, rather than multiphoton absorption. Our work extends nano-photochemistry into the terahertz frequencies.

For the intense THz light illumination, we use an amplifier-based THz system. The THz light is generated via pulse-front-tilted optical rectification of a prism-cut lithium niobate (LiNbO 3 ) crystal using femtosecond pulses from a Ti: sapphire regenerative amplifier (repetition rate: 1 KHz). The maximum electric field is 400 kV/cm at the focal point in the air. The peak frequency of the generated terahertz waves is 0.8 THz. The incoming THz wave illuminates the sample side first.
We illuminate intense THz pulses to resist/antenna structure for 2 hours. The polarization direction of THz wave is perpendicular to the long axis of the slot antenna. The experiment is carried out in dark ambient environment in order to prevent cross-linking of the resist by ultraviolet light. As a result, resist polymerization occur in the slot antenna, as shown in Fig. 2a. For parallel polarization, the resist polymerization does not occur. As can be seen in Fig. 2b,c, the resist cross-linked antennas are distinctly different from the antennas without resist. We cannot observe any electromigration effect. We confirm that the resist is cross-linked only inside the antenna. That is, local field enhancement is an important factor for THz-driven polymerization. However, the photon energy of a terahertz wave is several meV, which is much smaller than the photon energy of ultraviolet light. It is very unlikely to induce multiphoton polymerization using terahertz field enhancement. Therefore, we infer that the process of resist cross-linking is different. As mentioned before, various experiments have been reported that electrons can be emitted when a strong terahertz wave is an incident on a metal structure. Therefore, we have inferred that polymerization is caused by terahertz electron emission.
To confirm our assumption, we fabricate terahertz-resonant bowtie nanoantennas. Bowtie nanoantennas are made by using electron beam lithography and ion milling. The period of the antennas is 360 μm, length of the long axis is 170 μm, length of the short axis is 85 μm, the gap size is 500 nm, respectively. Resist polymerization occurs again after 2-hour exposure to high-power terahertz waves. Cross-linking occurs only at the  center of the bowtie antenna, where giant field enhancement occurs (Fig. 3a). To analyze the characteristics of the bowtie antenna, we perform finite element method simulation (COMSOL Multiphysics 5.3). The simulated electric field profile at the resonance frequency is shown in Fig. 3b. In this simulation, the gap size of the antenna is 1 μm. At the center of the bowtie antenna, the field enhancement factor reaches ~300. We measure the electric current while illuminating a terahertz wave to the sample to see directly if the electrons are emitted. Each pole of bowtie antennas is isolated, so electric current measurements can be made by using probe station 32 . In the measurement, we use 1 μm gap bowtie antenna, to match with simulation. The incoming terahertz wave illuminates the substrate side first. The intensity of the terahertz wave is adjusted using two wire grid polarizers. The polarization direction of the terahertz wave is not changed. We observe that the magnitude of the current exponentially increases as the amplitude of the THz wave increases (Fig. 3c, black dot). We analyze the data using Fowler-Nordheim plot and obtain a linear graph (Fig. 3d, black dot). When electrons are emitted from the metal by intense electric field, a current density follows Fowler-Nordheim equation 33 . The field emission current density J is given by the form 21 where a = 1.53 × 10 −6 AeVV −2 and b 6 829 10 eV Vm , are first and second Fowler-Nordheim constants, respectively, α M is the area efficiency of emission, λ C is a characteristic supply correction factor, v is a correction factor associated with the barrier shape, Φ is the work function of the metal (in our case, the metal is gold.), β is the field enhancement factor, and E 0 is the incident electric field. In our case, the barrier shape is triangular, so the factor α M , λ C , v becomes α M = λ C = v = 1 21 . We calculate theoretical electric current by using this equation and plot in Fig. 3c,d (red line). The calculation is consistent with the experimental data. In Fowler-Nordheim plot (Fig. 3d), the slope of the experimental data and theoretical data are almost identical. Considering the factor β Φ vb / 3 2 , the slope is a function of field enhancement factor. That is, the field enhancement factor calculated by the COMSOL simulation is in good agreement with the experimental data. Therefore, we are able to confirm the terahertz photoemission directly through experiments, COMSOL simulation, and theoretical calculation. The electron emission occurs at the end of the bowtie antenna, which is similar to the hot electron emission at visible frequencies 34 . We also examine whether the amount of electrons emitted from the metals is sufficient to induce the resist polymerization. We obtain the electron doses from the measured and calculated electric current. In Fig. 3c, we plot electron doses as a function of the amplitude of the incident electric field. The exposure time is fixed as 2 hours. The calculated electron doses are about 10 5~1 0 6 electrons/  2 , which is much higher value comparing the one in the previous report using femtosecond laser pulse (10 4~1 0 5 electrons/nm 2 ) 16 . We, therefore, conclude that the amount of electrons emitted by the terahertz wave is sufficient to induce the resist polymerization. Note that measurable field emission starts at 250 kV/cm. Considering the field enhancement factor 300, near-field reaches at 7.5 V/nm, which is much larger value than the threshold of 1 V/nm in the previous report by Jingdi Zhang et al. 22 . This is due to the difference between vacuum and atmospheric pressure. In atmospheric pressure, measurable field emission threshold becomes much higher than that in vacuum 35,36 . It may be because the electron transport between two metals can be affected by collision with molecules in the air 36 . In the ref. 35 , field emission threshold of atmospheric pressure becomes 7.5 times higher than that of vacuum. In our case, the threshold is also 7.5 times higher than that in the vacuum, which is similar to the ref. 35 .
We also conduct a resist polymerization experiment using a square-shaped slot antenna to investigate the applicability of terahertz-driven polymerization. The antenna is fabricated by using FIB, the side length of the antenna is 60 μm, and the width of the gap is 400 nm. The antenna is made on a glass substrate. First, in order to investigate the characteristics of the square-shaped slot antenna, terahertz time-domain spectroscopy (THz-TDS) is performed. We perform THz-TDS using an oscillator-based system. A femtosecond Ti: sapphire laser with 780 nm center wavelength, 80 MHz repetition rate, and 130 fs pulse width (Mira 900 and Verdi V5, Coherent) is divided into pump and probe beam. Then pump beam and a biased GaAs emitter are used to generate single-cycle THz pulses. The pulses are focused with off-axis parabolic mirrors (NA = 0. 25) and incident normally in the sample. Transmitted THz wave is collected by parabolic mirrors (NA = 0.32) and detected by using (110)-oriented ZnTe crystal via electro-optic sampling. In this experiment, the incoming THz wave illuminates the substrate side first. Although the direction of the incident THz wave has changed, the transmission characteristics do not change, as previously reported 37 . The transmitted signal from the sample is normalized to the signal from the substrate. The normalized amplitude is about 0.03 at resonance and frequency is 0.8 THz. We determine the field enhancement factor using experimental values and Kirchhoff integral formalism 38 . Field enhancement factor is about 100 at 0.8 THz (Fig. 4a). We spin-coat resists on square-shaped slot antennas and expose intense terahertz pulses for 4 hours. After development, we can observe cross-linked resist not only inside the slot antenna but also outside. In addition, resist polymerization is observed only on the side perpendicular to the polarization direction of the terahertz waves. A comparison of Fig. 4b,c clearly shows the difference. The profile of cross-linked resist is similar to the near-field distribution in the terahertz nano-slot antenna 39 . Since photoemission is closely related to the local field enhancement factor, we can say that this is a spatial mapping of the photoemission. Note that resist polymerization occurs only inside the nano-slot antenna for 2-hour exposure time (Fig. 2c). For the sample in Fig. 4b, the 2-hour exposed resist is also polymerized only inside the nano-slot antenna. That is, by controlling the exposure time, we are able to perform not only a nanoscale chemistry and but also a photoemission mapping. To summarize, we are able to induce nanoscale chemical reactions using tunneling electrons generated by intense terahertz waves, and, show its potential application -photoemission mapping.
In conclusion, we demonstrate THz-driven resist polymerization using nanoantennas. The resist is cross-linked near the area where local field enhancement occurs in the nanoantenna. The enhanced THz waves in the nanoantenna induce photoemission and the electrons ejected from the metal cause the polymerization of the resist. Electric current measurement shows that the origin of cross-linking is photoemission, not multiphoton absorption. Our data shows remarkable results opening new possibilities for approaching nano-photochemistry with terahertz optics. Our efforts combining terahertz optics and nanochemistry can present great potential for nanogap-enhanced spectroscopy, nanolithography, near-field mapping, and modulation, etc.