Passive control of a high-energy carbon dioxide-pulsed vectorial vortex beam

High-energy pulsed carbon dioxide (CO₂) lasers are used in micromachining for material cutting, drilling, welding, and re-forming. To achieve precise control of angular-variant polarization and phase, one well-known technique is to induce radial polarization of a vectorial vortex beam (one which is cylindrical, and axially symmetric).¹ For practical applications, radial polarization achieves levels of surface smoothness and feature sharpness that are not attainable when using uniform linear and circular polarization.² However, although it is possible to achieve active control of beam polarization using special resonators, this approach requires complicated alignment and a large laser system. To further the development of this technology, it is desirable to have a simple device that operates using a variety of light sources and enables a beam profile with high power density. For active control of beam polarization, electrically or mechanically (and not just generation of the vectorial vortex beam), it is also important to achieve a single-shot observation method in real time. Furthermore, to date there has been no demonstration of a simple device using the passive method (which enables varied output polarization depending on the input) for a short-pulse, high power density CO₂ laser beam that avoids damaging the polarization monitor.

We sought passive control of a high-energy mid-IR pulsed vectorial vortex beam using an achromatic axially symmetric zinc selenide (ZnSe) waveplate (a device that alters the polarization of light passing through it). By exploiting internal Fresnel reflections, we aimed to generate a vectorial vortex beam³ and determine its polarization.^{4, 5}

Figure 1(a) shows our setup for passive control of high-energy pulsed vectorial vortex beams. We used a 10Hz Q-switched CO₂ laser at a wavelength of 10.6μm to generate the vectorial vortex beam with sub-10mJ energy per pulse at a pulse duration of 15ns. The achromatic axially symmetric waveplate enables conversion of the arbitrary polarized beam with a Gaussian profile into the vectorial vortex beam with a ‘doughnut’ profile,⁶ and is fabricated in ZnSe to operate in the mid-IR region. Figure 1(b) shows the phase retardance properties along the slope angles of the waveplate at wavelengths of 0.63 and 10.6μm. The waveplate at the slope angle of 65° gives a phase retardance of 90°. We estimated the transmitted coefficients—shown in Figure 1(c) as the broken line—by Lambert-Beer's law (which describes attenuation of light in terms of the material through which it travels). The phase retardance—the solid line in Figure 1(c)—is achromatically controlled below 3° at the slope angle of 65° in the spectral range 0.55–15μm.

Figure 1. (a) Passive control of an intense mid-IR laser pulse vectorial vortex carbon dioxide (CO₂) beam by a zinc selenide (ZnSe) waveplate. (b) Slope angle properties of phase retardance at the wavelengths of 0.63 and 10.6μm. (c) Wavelength dependence of phase retardance (solid line) and transmitted coefficient (dashed line).

We observed that the incident beam diameter is dependent on the transmitted coefficients of the ZnSe waveplate, and the beam diameter was optimized at 4mm at a maximum transmitted coefficient of 58.7%, as shown in Figure 2(a). The effective diameter d_in of the waveplate was scaled to d_in≤0.42d₁, where d₁ was the input diameter of the waveplate (which we evaluated by ray tracing methods). The transmitted coefficient decreased when the beam diameter increased to greater than 4mm because of optical misalignment inside the ZnSe device. However, for an input beam diameter d<4mm, the transmitted coefficient also decreased as a result of the diffraction effect caused by the aperture controlling the input beam diameter. Figure 2(b) shows the output energy as a function of the input energy, resulting in a maximum output energy of 2.6mJ per pulse. The converted pulse energy would be sufficient for micromachining applications.

Figure 2. (a) Transmitted coefficient while changing the diameter of the incident beam. (b) Transmitted coefficient of the passive ZnSe waveplate.

We determined the single-shot polarization state using an inverse source problem for a converted pulsed CO₂ vectorial vortex beam. Figure 3(a) shows the polarization distribution after passing through the ZnSe waveplate. We sketched ellipsometric parameters on the intensity distribution of the vectorial vortex beam at angles of 15°. The vectorial vortex beam kept the polarization distribution with the degree of polarization (DOP) of 1.08. In addition, we evaluated the ellipticity and its azimuth around the beam—see Figure 3(b)—and reproduced these values using a rotating analyzer to cross-check. According to the polarization parameters, the converted beam had manipulated the angular-variant polarization. This approach provides a universal concept for passive control not only to generate a vectorial vortex beam but also to conduct its single-shot polarization determination at spectral regions from UV to THz.

Figure 3. (a) Polarization distribution on the vectorial vortex beam. (b) The single-shot polarization determination decides ellipticity (solid line) and its azimuth (dashed line) along the angle on the beam. a.u.: Arbitrary units.

In summary, we have demonstrated passive control to produce a 10.6μm, sub-10mJ, 15ns vectorial vortex beam and to determine its polarization by an achromatic axially symmetric ZnSe waveplate. The maximum throughput was 58.7% with the input beam diameter of 4mm. This solution potentially solves the limitations of low-damage threshold, material absorption, and narrow spectral bandwidth for micromachining. Our future work will focus on producing radial polarization with flat phase (no optical vortex beam)³ for micromachining applications that require surface smoothness and feature sharpness.