Design and implementation of a portable diagnostic system for Thomson scattering and optical emission spectroscopy measurements

Thomson scattering (TS) is a trusted diagnostic for measuring the key plasma parameters of electron density and temperature and has been so for many decades. Since the first application of the TS measurement technique,¹ TS diagnostics have been instrumental in the study of plasmas, from verifying the first tokamak plasmas to many high- and low-temperature plasma applications.^2–7 TS diagnostics allow for a localized measurement of the plasma. A focused probe laser beam with a fixed wavelength interacts with free electrons and produces scattered photons. The scattered photons collected in the sample volume defined by the measurement line of sight can be co-located anywhere within the experimental device, given the availability of appropriate access ports. The intersection volume can be made very small, enabling high spatial resolution. Fans of lines of sight can be aligned to the laser path as it transects the plasma volume, resulting in simultaneous multi-point spatial profiles. The temporal resolution is defined by the laser pulse duration, which is typically <10 ns.

While TS is a well-established technique, its implementation on plasma devices is as variable as the devices themselves.^3,4,6–10 Some of this variability is driven by the ranges of n_e and T_e that are encountered in differing plasma devices. The advancement of Nd:YAG laser technology has reduced the footprint of the laser system, and through the rigors of industrial application, high energy “YAGs” have become turn-key systems for scientific TS applications. The infrared 1064 nm central wavelength of YAG lasers is better suited to filter polychrometer arrays of detectors, which are reliable but relatively expensive to implement more than ∼five wavelength filters for each spatial line of sight and typically require temperature-controlled environments for stability.¹¹ Recently, harmonic generators (again driven by industrial applications) have become robust, and frequency-doubled YAGs operating at green 532 nm enable the emergence of TS applications that take advantage of conventional visible spectroscopy hardware.^12–14 Visible spectrometers and detectors are relatively inexpensive, sensitive, and robust. This allows the TS system to be built out of largely off-the-shelf components with high sensitivity and at relatively low expense.

Moreover, in the absence of the TS laser excitation, the visible spectroscopy hardware can be used to measure the spectral emission lines of plasma ions. Many TS systems, designed to always look near the laser wavelength, can only examine a very narrow range of plasma ion emission lines. The implementation of the new diagnostic system discussed here has multiple gratings of various line spacing, thus allowing for a wide variety of resolving power. Moreover, the diagnostic system can be tuned to look at other spectral regions of interest and uses diffraction gratings with a range of resolving powers. Recently, spectrometer platforms have become commercially available, which provide high throughput of light with a tunable grating. Using the spectrometer system described here, for example, Thomson scattering and Optical Emission Spectroscopy (OES) diagnostics can be utilized simultaneously.

The OES diagnostic is an integral part of this diagnostic system, which collects the line intensities of excited plasma ions and impurities. OES techniques can simultaneously measure the peak intensities of helium emission lines, for example, which are evaluated using a collision-radiative model (CRM) to estimate n_e and T_e.¹⁵ The availability of multiple gratings provides a wide range of measurements of the ion and electron temperatures.

With the advancements in laser and detector technologies, the concept of a modular TS and OES diagnostic system has been funded by the Advanced Research Projects Agency-Energy (ARPA-E). Once completed, the portable diagnostic package (PDP) will travel to several potentially transformational fusion-relevant concept devices to measure fundamental plasma parameters.

This paper discusses the design and setup of the PDP. It will discuss various components used, system calibration, and stray laser light intensity optimization. It also discusses the PDP implementation on an Electrothermal (ET) arc source,^16,17 along with preliminary results from OES.

The PDP was developed to create a high-performance, portable system for both TS and OES, which can be implemented on multiple fusion-relevant experiments. The system is considered very portable as it can be deployed in approximately one week, after arriving at the host location. The high-performance spectroscopic system relies heavily on a range of gratings, which allows for both high resolution and adequate free-spectral ranges over a wide variety of plasma parameters. The range of spectral resolutions also allows for OES measurements from a variety of emission lines to be taken simultaneously on the same photo-sensor. Additionally, multichannel TS broadened spectra can be measured in a single photo-sensor. The high-sensitivity photo-sensors enable a very high total dynamic range of 2¹⁶ = 65 536 counts and an amplification factor of 1–100.

Table I shows the parameters the PDP will measure along with the expected measurement range and temporal resolution and spatial positions.

The diagnostic components of the PDP include a Quantel Q-smart 1500 amplified Nd:YAG laser, two Princeton Instruments SCT320 spectrometers each with a three grating turret, and two Princeton Instruments PI-MAX4 intensified cameras. The laser produces 1500 mJ of energy per pulse at 1064 nm at 10 Hz, which is then frequency-doubled to produce up to 850 mJ of 532 nm light. There are six different gratings between 150 and 2400 g/mm. These various gratings will allow for ion temperatures below 10 eV to be resolved and electron temperatures up to ∼1 keV. The intensified CCD camera, PI-MAX 4:1024f, is equipped with a Gen III intensifier. The imaging array of the camera consists of 1024 × 1024 pixel². The intensifier can be gated at ∼>2 ns, which is sufficient to adequately capture the PDP TS laser (<10 ns). The capability to change the intensifier gain adds customizability to resolve the TS spectrum and line intensities for the OES measurements. Additionally, the capability to bin the sensor into multiple channels for reduced read-noise allows data collection from distributed spatial locations using a fiber-bundle array.

To be portable, a modular utility cart (dimension: 51 × 40 × 24 in.³) has been set up to mount two spectrometers and two cameras on the top two shelves, as shown in Fig. 1. The bottom shelf has the workstation and data acquisition chassis of National Instruments. The lower shelf also accommodates the fiber bundle transfer chassis. The Q-smart 1500 laser is not part of this assembly. It is intended to be mounted separately, i.e., to the plasma device itself.

The collection optics includes a custom-made 11 × 3 fiber bundle where each fiber is 800 µm in diameter and has a numerical aperture (NA) of 0.12. These low-OH, stainless steel monocoil fibers are ruggedized to enable deployment in difficult situations. Each spectrometer can accommodate 11 input fibers, which leaves 11 collection fibers for utilization by the host facility. Commercial camera lenses will be used to collect light. Each host facility may have different requirements, so a commercially available Nikon 18–105 mm lens with a variable focal length will be used for the collection lens. On the ET-Arc source, a Nikon lens with a fixed focal length of 50 mm has been used.

The portable diagnostic system is being tested on an ET-Arc plasma source at Oak Ridge National Laboratory (ORNL). Figure 2 shows a model of the ET-Arc source, including the location of the path of the TS laser and the viewport. In Fig. 2, the plasma is injected from the top (yellow) port of the vacuum chamber and interacts with the target surface (green). The laser traverses horizontally outward from the page, remaining parallel to the target surface. The viewport for the collection optics is located at the bottom of the chamber.

Using Eq. (1), the expected number of scattered photons n_pe detected from the TS signal at the detector can be calculated as follows:

where E is the laser photon energy, h is Plank’s constant, ν is the laser frequency, ΔL is the scattering length, Ω is the solid angle, $dσdΩTS$ is the differential TS cross section, and η_transmission and η_quantum are transmission and quantum efficiencies. In the ET-Arc source implementation, the observation geometry solid angle, Ω, is ∼ 0.017 str. At the laser energy of 820 mJ and n_e = 1 × 10¹⁹ m⁻³, the total scattered photoelectron yield is estimated to be the total scattered photoelectron yield to be ∼800.

During the TS measurement, stray laser light can become a problem when trying to resolve the scattered light spectra due to the relatively small TS cross section. Thus, straylight mitigation is very important to the success of these measurements. When performing calibration measurements on the ET-Arc source, several techniques were used to reduce the straylight, and the results of these techniques can be seen in Fig. 3. Figure 3 shows a 3D plot of the normalized average intensity (z axis) as a function of various conditions (x axis) for all 11 channels (y axis). There are four different system configurations, which are compared to the maximum baseline straylight without any mitigation (represented by N/A in Fig. 3).

For condition 1 (C1), we added a polarizer at the collection optics that reduced the straylight nearly by 50% as expected. Next, during C2, we rotated the direction of the polarizer and observed the straylight intensity to decrease. Furthermore, in C3, we used the polarizer and its orientation from C1 and did a laser beam walk using two mirrors where the laser beam was optimized to pass close to the center of the mirrors and the Brewster window. This combination helped the straylight intensity to be reduced by about a factor of 3. Finally, in C4, condition 3 was repeated using one iris where only a marginal change was observed. Without precise alignment targets in the laser path, laser beam walking is subjective and can increase the straylight, for example, if the majority of the straylight is coming from something non-uniform such as dust in the chamber or imperfections on the entrance windows.

To determine the instrument function, the spectrometers are set to collect laser light. Due to the spectrally narrow nature of laser light emission, all broadening measured is assumed to be due to the instrument function. To perform these measurements, the grating is rotated so that the laser light falls onto several locations on the photo-sensor. Each spectrum was fitted with a Gaussian function to get the full-width half maximum (FWHM) for each channel.

As an example of the instrument function calibration, the FWHM for the 300 groves/mm grating is shown in Fig. 4. Figure 4 shows the plot of FWHM (in nm) as a function of channels in the spectrometer, where the observed variation in FWHM across the channel is small. Noticeably, in all channels, for the 300 grooves/mm grating, when the laser is positioned on the right (higher wavelength) side of the photodetector, the FWHM consistently showed the maximum width when compared to left- and center-located spectra. The dashed line shows the manufacturer’s expected linewidth for the grating utilized. The instrument function calibration was conducted with other gratings as well; the results are consistent with the expected FWHM for each grating; however, further optimization could be done.

The diagnostic system has been tested on the ET-Arc source during helium plasma discharges. The experiments included a helium Doppler broadening analysis of He II 486.6 nm light to measure T_i and He I line ratios to measure T_e and n_e using OES. From the collisional radiative model (CRM) developed at ORNL and OES, the helium line ratios can be used to measure T_e and n_e. The three Helium neutral lines used in these analyses were 667.8, 706.5, and 728.1 nm. The ratio between 706.5 and 728.1 can be used to determine the temperature, and the ratio between 667.8 and 728.1 can be used to determine the density. Figure 5 shows spectra of the three helium lines required to calculate the ratios. A spectral sensitivity calibration has been performed to ensure that the apparent intensity of these line ratios is accurate. This work is ongoing. The line ratio technique offers a complimentary set of analyses to the TS and OES Doppler broadening so that the PDP can measure n_e, T_e, and T_i in multiple ways.

The ET-arc plasma discharge is roughly 1 ms in duration and is fairly dynamic. The discharge is characterized by a rising current trace of ∼250 µs, followed by a decaying current trace of ∼750 µs. Initial measurements have been made with the PDP OES system on the ET-arc to benchmark the performance of the PDP and examine the spatial and temporal behavior of the arc discharge. Preliminary analysis indicates that at the end of the current rise, the ion temperature is between 16 and 26 eV (as measured by Doppler broadening of the He II emission), while the electron temperature is between 5 and 20 eV and the electron density is 4 × 10¹⁹–1 × 10²⁰ m⁻³ (as measured by He I line ratios). The TS aspect of the PDP is still under development but will be compared directly to these measurements in future work.

A portable diagnostic package supported by ARPA-E is in development at Oak Ridge National Laboratory. This project aims at measuring several fundamental plasma parameters (n_e, T_e, n_i, T_i) for fusion-relevant devices utilizing TS and OES diagnostics. The OES aspects of the PDP have collected Doppler broadened helium lines to measure T_i and broad-spectrum line ratio data to characterize T_e and n_e.

The TS diagnostic, which will be a crucial part of the diagnostic package, is still being implemented. The work to obtain the Thomson scattering measurement using the Q-smart 1500 laser at the ET-Arc source is ongoing. With the available laser energy of up to 850 mJ from Q-smart 1500, electron density resolution in the range of 10¹⁸ m⁻³ plasma should be possible. Rayleigh scattering with nitrogen is planned to determine the electron density from the integrated TS spectrum. Currently, the two collaborative private entities slated for the deployment of the PDP are Princeton Fusion Systems and CT-Fusion. After the proof-of-principle demonstration on the ET-Arc source at ORNL, the diagnostic package will travel to the Princeton Fusion Systems’ Princeton Field Reverse Configuration-2 (PFRC-2) device,¹⁸ located in Princeton, NJ. The portable diagnostics will later travel to the CT-Fusion’s Steady Inductive Helicity Injected Torus-3 (HIT-SI3)^19,20 device in Seattle, WA.

This paper was authored, in part, by UT-Battelle, LLC, under contract with the U.S. DOE. The U.S. government retains and the publisher, by accepting the article for publication, acknowledges that the U.S. government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this article, or allow others to do so, for U.S. government purposes. DOE will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (http://energy.gov/downloads/doe-public-access-plan).

This project was supported by the U.S. Department of Energy, Contract No. D.E.-AC05-00OR22725. The authors greatly appreciate the funding support provided by the Advanced Research Projects Agency-Energy (ARPA-E) to make this work possible.

Grateful appreciation is given to H. B. Ray for assistance in CRM calculations used to interpret He line ratio data into electron temperature and density values.

The data that support the findings of this study are available from the corresponding author upon reasonable request.