Initial Fabrication and Characterization of Chemically-Etched Silicon Slits for KOSMOS

KOSMOS is a low-resolution, long-slit, optical spectrograph that has been upgraded at the University of Washington for its move from Kitt Peak National Observatory's Mayall 4m telescope to the Apache Point Observatory's ARC 3.5m telescope. One of the additions to KOSMOS is a slitviewer, which requires the fabrication of reflective slits, as KOSMOS previously used matte slits machined via wire EDM. We explore a novel method of slit fabrication using nanofabrication methods and compare the slit edge roughness, width uniformity, and the resulting scattering of the new fabricated slits to the original slits. We find the kerf surface of the chemically-etched reflective silicon slits are generally smoother than the machined matte slits, with an upper limit average roughness of 0.42 $\pm$ 0.03 $\mu$m versus 1.06 $\pm$ 0.04 $\mu$m respectively. The etched slits have width standard deviations of 6 $\pm$ 3 $\mu$m versus 10 $\pm$ 6 $\mu$m, respectively. The scattering for the chemically-etched slits is higher than that of the machined slits, showing that the reflectivity is the major contributor to scattering, not the roughness. This scattering, however, can be effectively reduced to zero with proper background subtraction. As slit widths increase, scattering increases for both types of slits, as expected. Future work will consist of testing and comparing the throughput and spectrophotometric data quality of these nanofabricated slits to the machined slits with on-sky data, in addition to making the etched slits more robust against breakage and finalizing the slit manufacturing process.

telescope. 2 In April 2018, the Dark Energy Survey Instrument (DESI) became the primary instrument for the 4m Mayall telescope at KPNO. The installation of DESI required the telescope to be taken apart and reconfigured to accommodate this new instrument, retiring KOSMOS and other 4m instruments.
KOSMOS is now seeing second light at Apache Point Observatory's (APO) ARC 3.5m telescope. A general purpose low-resolution optical spectrograph is an important tool for science targets that may be faint or not well-characterized. One of the initial instruments commissioned for the ARC 3.5m was the Dual Imaging Spectrograph 3 (DIS), which has been heavily used as the low resolution (R ∼ 1000-7000) optical spectrograph at APO since the early 1990s. Now that it is approximately 30 years old, it is currently struggling with some deterioration of performance as a function of age. KOSMOS updates the telescope's optical spectroscopic capability, after being modified for the APO observing community.
Additions to KOSMOS include a slitviewer camera, internal calibration lamps, a second internal electronics box to support the aforementioned modifications, and a Nasymth Port (NA2) adapter (Kadlec et al, in prep). These features will improve the usability of KOSMOS, particularly as APO has shifted over time to enhance capabilities to support time domain follow up and other time-specified observations. Having internal calibration lamps, for example, allows observers to run calibrations at the pointing for improved wavelength calibration, or outside of their observing windows while the instrument is off the telescope. The slitviewer is essential for ensuring the target is on the slit without using valuable observing time to switch between imaging and spectroscopic modes. Along with its use as a target acquisition camera, the slitviewer integrates with the field and boresight guiding software to function as an on-axis guider. Furthermore, the long slit on KOSMOS will allow users to take spectra of multiple aligned objects simultaneously.
Of the many modifications made to KOSMOS, this paper focuses on the requirement of a library of reflective slit masks. Prior to the addition of the slitviewer, KOSMOS utilized matte, machined slits to minimize scattering. Reflective slits redirect the photons that would have been scattered by matte slits towards the slitviewer camera, giving users the ability to know where they are in the sky. With new slits required for the retrofit of KOSMOS, we had the opportunity to test and document an alternative way of manufacturing slits -via wet chemical-etching, which is the preferred method of selective material removal in the semiconductor industry. As such, the paper adopts the same technology and investigates the feasibility of this method of fabrication for slits.
The process outlined in this paper was the pilot manufacturing for KOSMOS's reflective slits and will be improved upon in future iterations of slits. In the process of testing these new slits, we were able compare the slit edge roughness and uniformity of the slits fabricated via chemical-etching and wire EDM (electrical discharge machining). This allows us to test whether the roughness of the slit is an important factor in achieving higher spectrophotometric precision and throughput.

Motivation
Wire EDM is a common method of fabricating spectrograph slits. This is an electrothermal production process where a thin single strand metal wire cuts through metal by the use of heat from electrical sparks. This method has been one of the most consistent, precise, and cost effective methods of machining slits. 4 Despite this, there is still room for improving the consistency and uniformity of these slits. The slit edge roughness of the wire EDM slits from DIS ( Figure 1) is visually apparent, especially upon comparison to the chemically-etched slits in Figure 11. Typical roughness as a result of wire EDM is on the order of micron scales, 5 while typical roughness from wet-etching is on the scale of nanometers. 6 With nanofabrication facilities available at many re-search universities (including the University of Washington), we wanted to test the feasibility of the semiconductor device fabrication methods for producing ultra-flat reflective slits as demonstrated in Ref. 7 and compare their uniformity and performance. Many of these fabrication processes and their benefits and constraints as applied to astronomical instruments are not well-documented. As these techniques have become more widely available, in this paper, we attempt to better constrain requirements, make measurements, and document their impact on astronomical data.
Below we present the manufacturing and characterization of chemically-etched reflective slits, a required upgrade to accommodate the new slitviewing capability of KOSMOS at APO. In Section 2, we discuss the design approach for the new slits. In Section 3, we discuss the manufacturing process for the nanofabricated slits. In Section 4, we discuss the acquisition and processing of data taken via microscope and in lab with KOSMOS. In Section 5, we present the slit edge roughness and width results of the matte and reflective slits and their impact on the data.

Design
Until recently, the functional field of view (FOV) at APO was 6.1 arcminutes. Recent upgrades to the baffles increased this to 8 arcminutes. To align with these upgrades and acquire data across the entire FOV, we attempted to make the slits 80mm long (8.6 arcminutes). However, this made the wafers fragile. The stress of its own weight would break the wafers along the slits while sitting in their carriers in between fabrication steps. Of the 75 slits fabricated, all 25 8.6 arcminute-long wafers broke during various steps of the manufacturing process. Each 525 µm thick, 100 mm long wafer had a remaining 5 mm on each side of the slit, which was simply not enough to support the the sizable through-wafer hole in the center. Unfortunately, dicing the wafer to reduce strain prior to the etching was not an option due to constraints of the tools. To create more stability, we shortened the length of the slits to be the length of the original KOSMOS slits, 64 mm long (6.8 arcmins; Figure 2). This is consistent with the size limitations due to the current shutter 90mm in diameter (Kadlec et al., in prep).
Selection of desired slit widths goes from 0.5" to 20" (29 µm to 1169.6 µm), with these limits Fig 2 The thin lines of the rectangle outline the dicing lanes where the wafer is diced. The slit length is indicated, but slit width is not as the width varies depending on design specifications. For this work, achieved slit widths were 0.36" to 7.3" (64 µm to 1250 µm). The image is not to scale. chosen based on being seeing-limited at 0.5" and 20" for a user-specific project. However, due to an error with the plate scales of the instrument and telescope, the achieved widths have become 0.36" to 7.3" (64 µm to 1250 µm); Table 4). Since the matte slits from the previous iteration of KOSMOS are from 0.58" to 2.92" (91 µm to 456 µm); Table 5), this selection of reflective slit widths still expands upon the previously available slit widths for KOSMOS.

Manufacturing
The reflective slits for KOSMOS are fabricated using photolithography, wet chemical etching, and vapor deposition, which is a typical process for fabricating semiconductor devices. Based on Atalla's work on surface passivation by thermal oxidation, 8 Hoerni patented the planar process in 1959. 9 The planar process is a manufacturing process and serves as the basis for small-scale and mass production of integrated circuits in the semiconductor industry, as well as the groundwork for this project. It is composed of photolithography, wet-chemical development, and metal deposition.
To optimize for mass production, the industry uses standard silicon wafers and various applications of the planar process. This process is widely-used for mechanical and electrical silicon structures, such as membranes, nozzles, diaphragms, and trenches, due to the nanometer-scale precision that photolithography allows in patterning. 10,11 Because this process is standard, nanofabrication facilities are widely available at the university level.
Bulk micromachining is another class of processes for micromechanical electronics systems (MEMS) used in our manufacturing process. For bulk processes, substrate or bulk is removed via etching (either dry or wet) to create microstructures, such as cavities or through-wafer holes, both of which we utilize in our process. Wet etching, where the etchants are liquid, has many benefits over dry etching, where the etchants are plasma. Wet etching is faster, cheaper, more anisotropic, and lower in chemical waste output. 12 In the case of the slits for KOSMOS, we are using a 35% concentration of potassium hydroxide (KOH) at 100 • C to etch the silicon. KOHetching is a well-documented anisotropic Si etch process, which takes advantage of the etching ratio between the crystal planes in the silicon lattice, specifically the 100 and 111 planes. 13 Because KOH preferentially etches the 100 plane and minimally etches the 111 plane, the 111 plane is exposed as the KOH etches through the wafer ( Figure 3). This creates a rectangular-shaped cavity with slanted sidewalls, which are the 111 planes. Because of this crystalline structure of silicon, the surface of each wafer is atomically flat, and through anisotropic etching with KOH, the etched 111 surfaces have an estimated roughness of 3-5 nm. 6 In Section 4, we observe the maximum roughness of this plane for the slits, with results summarized in Table 2.
We explored and refined a known method of KOH (Potassium Hydroxide) silicon etching to precisely etch through silicon wafers. The method below was initially developed and tested by engineers at the Washington Nanofabrication Facility (WNF) and Ref. 14, based on work done by Ref. 7. For specifics on tools used, their specifications, etc., see the manual. 15 Below we present a ICP-F etches the exposed silicon nitride. f) EKC bath to remove photoresist. g) Pattern from photoresist transferred onto the nitride. h) KOH bath to etch exposed silicon. i) Slit and dicing lanes etched into silicon wafer. j) Nitride at the bottom of wafer etched off. k) Aluminum evaporated onto wafer. l) Silicon dioxide deposited onto aluminum. m) Light can now go through the slit and have whatever light that doesn't go through reflected back. n) Dicing the extra parts of the wafer off. o) Now completed slit ready for mounting. Silicon in gray; Nitride in blue; Silicon dioxide in teal; Photoresist in dark orange; Aluminum in orange; Developer in light blue. Note -steps j to l were switched in the initial testing of this process such that the reflective side was the wider side. After testing at APO, future slits will be manufactured as described in this diagram. general overview of the approach and results.

Photolithography and Pattern Transfer
The processing begins with 525 µm thick, 100 mm diameter silicon wafers coated with 25 µm LPCVD silicon nitride on both sides by Rogue Valley Microdevices 1 (Figure 4a). The nitride acts as a mask to protect the silicon from dissolving in the KOH bath, as silicon nitride etching by KOH is negligible (< 1 nm/hour) if etched at all. To etch the nitride such that the silicon beneath it can be exposed, we used photolithography to create a nitride mask on one side of the wafer for etching the silicon. First, we coated the top of the wafer with 1.5 µm of AZ 1512 photoresist (Figure 4b). We then used a 175W 405nm laser with a 20mm write head (Heidelberg DWL 66+ laser lithography tool) with edge roughness of 0.11 µm 16 to expose the photoresist (Figure 4c) with the pattern in Figure 2. We chose the 20mm write head due to its faster write speed and because the resulting edge roughness is small compared to the width of the slit.
Then we developed the photoresist with AD10, vacuum baked at 100 • C for 60 seconds ( Figure   4d). This developed photoresist acts as a mask, allowing us to etch the nitride off where the slit will be. We use ICP (inductively coupled plasma) to remove the nitride with a CHF3-O2 plasma at 20 • C ( Figure 4e). Bare silicon is thus exposed where the slit and dicing lanes will be. We removed the photoresist in an EKC bath ( Figure 4f).

KOH Etching and Plasma Etching
With a nitride mask protecting the parts of the silicon wafer we do not want removed (Figure 4g), we used the KOH to etch the silicon to create the slit ( Figure 4h). In our process, this took between 5-14 hours (further discussed in Section 5.4). In ideal conditions, a 35% KOH solution at 100 • C should etch this in 2 hours. 12,17 The remaining nitride on the back of the wafer is removed via the reactive ion etcher (Figure 4j).

Metallization and Assembly
We then evaporated (through physical vapor deposition) 1 µm on aluminum onto the side of the wafer with the narrow end of the slit to make it more reflective (Figure 4k). To protect the alu-minum from oxidation (which causes the aluminum to become dull and less reflective), we used plasma-enhanced vapor deposition (PECVD) to deposit 100 nm of silicon dioxide onto the wafer ( Figure 4l). Lastly, we diced the wafers into rectangles for the mounts using a Disco diamond dicing saw (Figure 4n; Figure 2). Three of each size slit were mounted for use, and the remaining are kept as extras.

Data and Analysis
To understand whether and/or how the uniformity in width across each slit and the average roughness of the slits impact spectral data, we must first characterize the width and roughness of each slit. However, due to the steep, deep sidewalls of the slits, which are very narrow, it is not possible to use traditional tools to measure the roughness of the slits, such as a profilometer. Instead, we must image the slits and define roughness and width from imaged data.
After measuring the characteristics of each slit, we used KOSMOS in lab to measure the potential impacts, scattering in particular, of the roughness and uniformity on spectral data in a controlled setting. Finally, in our future work, we will use KOSMOS on sky to understand the performance of the slits. We will characterize the throughput and the spectrophotometric data quality as a function of position on the slit, in conditions that observers will typically experience.

Acquisition -Microscopic Data
For each slit, ten sections along the slit were imaged randomly (Figure 5a), taken on an optical Leica DM compound microscope, with the slit being illuminated from below with the light pointing through the slit towards the camera. For chemically-etched slits with slit widths between 60 µm and 400 µm, images were taken at magnifications of 5x, 10x, and 20x. Images were taken with the light illuminating through the narrow side of the slits and the wide side of the slits to test whether there is a measurable difference in the widths of the slit. However, as the narrow side of the slit, which is its targeted width, sets the amount of light coming into the slit, the measured slit widths ended up being the same on both sides. For etched slits with slit widths larger than 1 mm, images were taken at 5x, as the slit is too wide for imaging at higher magnifications. For wire EDM slits, data were taken with the 10x objective to allow for standardized comparisons between all slits.
Using numerical apertures and and estimated peak wavelength to be at 550 nm, we obtain the resolution of images at each objective. We also imaged a reticle at 5x, 10x, and 20x magnifications to find the conversion of pixels to microns Table (1). Due to the low spatial resolution of the optical microscope and unresolved structure, these measurements set upper limits on the roughness of the slits and high errors on the measured width.

Data Analysis -Microscopic Data
Images were sharpened using Gaussian unsharp masking (Figure 5b), a common digital signal processing technique used to enhance the intensity at the edges. 18 First, we smooth (Equation 1) the original image by convolving a Gaussian kernel (Equation 2), which is a discretized approximation of a 2-D Gaussian-in this case in a 5x5 matrix which acts as a highpass filter, with the original image. Because the smoothing is done at a smaller spatial scale than the rough features (more than 5x5 pixels), resolution of the slit edges are not impacted. (2) This is a vital step for allowing Canny, the edge detection algorithm used and described below, to properly detect the edges of the slit.  The roughness is determined by the mean residual (absolute value of the difference) between the edge detected and the best-fit line determined by HoughTransformP along the sampled image.
The slit width is the average distance between the two lines detected. Due to the reflective slits being angled at 15 • for the slitviewer, the effective slit width is slightly smaller (by cos(15 • )) than the measured width. The width uniformity is the standard deviation of the measured width along the length of the slit. These roughness and width measurements for both matte and reflective slits, as well as the effective width for the reflective slits, are presented in the Section 5.

Acquisition -KOSMOS Data -In Lab
For each mounted slit, both chemically-etched and wire EDM, spectra were taken using the krypton, neon, and argon internal calibration lamps on simultaneously, alternating between red and blue grisms. By having all internal calibration lamps on, more emission lines were present across the entire optical bandpass and minimized the need to wait for warm up times of the individual calibration lamps for each set of observations. Both red and blue grisms were separately used in observations to gauge impacts on throughput throughout the entire bandpass. Darks, or images taken without opening the shutter, were taken to characterize the thermal noise of the detector.

Data Analysis-KOSMOS Data -In Lab
Data were reduced by subtracting a median-combined master dark, where all the darks taken were median-combined, from spectral data. From this reduced data, we divided by exposure time to were taken with both the red and blue dispersers for each mounted slit, the background and slit widths obtained from the separate red and blue disperser frames were averaged.

Results and Discussion
Due to the wide variety of scientific drivers for instruments, which are infrequently built, it is difficult to disentangle oral history from holdovers from older technologies. Below we discuss the results of a thorough examination of slit roughness and uniformity, as we want to be as cautious as possible about reducing possible sources of background and increasing throughput in science data.
As new technologies become available, we have taken the opportunity to measure and compare to older techniques. By characterizing the machined and chemically-etched slits, we can fully understand the performance limits of these slits. This way, we have all the information to judge the trade-offs and what can be improved and what we should continue to use.

Slit Results-Width Uniformity and Maximum Roughness
By comparing the results in Tables 2 and 3 To examine the roughness as a function of width for both methods of fabrication, we fit roughness versus width in Figure 10. The roughness of the slits do not depend on widths in both the machined and etched slits. Additionally, the wire EDM slits are consistently rougher than chemically-etched slits, no matter the slit width.
In comparing Tables 4 and 5, the chemically-etched slits also are more uniform in width   throughout the individual slits, as the error is more consistent and generally smaller than machined slit width error. This is due to the 0.11 µm precision of the laser lithography tool, which sets the shape of the slit. Under ideal conditions, the KOH etching sets the size of the slits and will etch a smooth edge in the silicon. However, due to the remaining nitride on the slit (Figure 11) as In the testing of this process, we found a large disparity between the targeted slit width and the measured slit width of the chemically-etched slits. This comes from the unpredictability of our wet-etching process, which was missing a step of removing silicon dioxide prior to the KOH etch, discussed in Section 5.4. Instead of grouping them by targeted slit width, we chose to demarcate the slits by their functional measured slit width for practical purposes in Table 4.

Slit Results -Scattering
In examining Tables 6 and 7, the scattering of the chemically-etched slits is higher on average than that of the machined slits. Since the etched slits are smoother than the machined slits, the reflectivity, not the roughness, is the major contributor to the observed scattering. The scattering increases as slit width increases for both matte and reflective slits and is greater for reflective slits ( Figure 12). To normalize the scatter by slit width and compare the impact of reflectivity, the scatter was divided by the width, giving units of ADU/pix/s/micron. The scatter/width for matte slits including continuum is 0.09 ± 0.5 ADU/pix/s/micron and 0.13 ± 0.3 ADU/pix/s/micron for reflective slits. For light outside the slit, the scatter/width for matte slits is -0.0004 ± 0.03 ADU/pix/s/micron and 0.04 ± 0.02 ADU/pix/s/micron for reflective slits. In both scenarios, the scatter using the reflective slits is higher, but with proper background subtraction, the impacts of the scatter can be minimized.

Advantages and Disadvantages of Wet Chemical Etching for Slit Fabrication
The primary advantages of using wet chemical etching to make slits are that they are less rough than that of wire EDM slits. There are many academic labs with nanofabrication tools available, making this fabrication method fairly available. The process for setting the widths and features of the slits is also easy, as it just requires editing a CAD file. In addition, as mentioned in Ref. 7, these slits are ultra-flat, which makes them more ideal mirrors for the purposes of the slitviewer and minimizes any distortions due to warping or uneven polishing.
The disadvantage for this specific project was the unpredictability in the etching process. The etch rate was inconsistent across wafers in the bath due to native oxide on the silicon slowing down the KOH etch. A way to mitigate these inconsistencies in future work would be to use BOE (buffered oxide etchant) to etch away the silicon oxide that grew after etching the nitride off, as done in Ref. 25. This is expected to constrain the etch times to be closer to the theoretical through etch time of 2 hours at 35% KOH at 100 • C, 17 as opposed to the 5-14 hours that it took for this project. Another factor that contributed to the difficulties of this process were that labels with the different slit sizes in the etch process would etch away, making distinguishing different slits difficult. A suggested alternative method of labeling would be making multiple marks in the etching process that do not blend together. Lastly, the dicing lanes would not always fully etch through, though ultimately with the way the dicing saw was set up, it did not matter too much.
Overall, this is an effective method for those who need reflective, highly uniform, and flat slits with multiple custom sizes provided that there is the addition of the step of a BOE etch to remove the silicon dioxide prior to the KOH etch to reduce unpredictability in sizing. The lead time is longer, as the uniqueness of each project in these academic labs make the process look vastly different for each user and there are a large number of tools to learn to use. It is possible for this process to be outsourced to contractors at facilities with equipment to handle different sized wafers, though this has not been considered at this point due to the process still being tested. Fortunately many of these steps can be done in batches, cutting some of the time it takes for processing. Most of the uncertainty in etch time can be minimized by doing a quick BOE dip prior to the KOH etch, as well as considering the amount of silicon that will be exposed and etched at a time.

Conclusions and Future Work
Chemically-etched slits are overall smoother than the machined slits. The chemically-etched slits and wire EDM slits produce roughly the same scattering within their measurement errors (see Section 5.2). The main advantage of the chemically-etched slits is how customizable and easy for each user to specify, as it doesn't depend on wire-gauge. The edges, particularly around the ends of the slits, are much smoother than those of the wire-EDM slits, as that method makes a rough hole where the wire enters. For short slits, users won't have to worry about poor data quality at the ends of the slits.
Roughness is not a primary factor in scattering. Rather reflectivity impacts scattering far more.
Background in data increases slightly using the reflective chemically-etched slits. Users can choose whether to use matte slits, as the old KOSMOS slits are still available for selection, or reflective slits depending on their observational needs, i.e. whether they need a slitviewer, whether high signal to noise is desired, and whether the slit width needed is available.
In the future, we will be characterizing the throughput with KOSMOS on-sky. By testing on sky settings, we can determine how the slits perform in the actual conditions observers will have to deal with. Since we have determined the machined and etched slits' roughness and width with microscope data and the scattering from the laboratory data, the on sky KOSMOS data will allow us to determine how the edge roughness and reflectance will impact throughput. Based on the preliminary results with laboratory data taken with KOSMOS, our hypothesis is that the higher scattering due to the reflectivity of the slits will decrease the signal to noise ratio of spectral data.
Compared to the large impact of the reflectivity, the slit edge roughness will have a less noticeable effect, similar to the findings in the laboratory measurements. Below are the tests we will be performing.
We will be taking a variety of data from well-characterized stars and measuring the throughput from both machined and etched slits. Following the method outlined in Ref. 26, we will characterize throughput of each slit, by deriving an expected photon rate from the star and comparing that to the measured photon rate on the detector. While these numbers are dependent on the observing conditions, they are sufficient for rough comparisons to the two sets of slits. This methodology allows us to more precisely understand the conditions that the data are taken in, as opposed to those in the manufacturer-provided specification sheets for wavelength calibration. In addition, we will be stepping the standard star across the chemically-etched and wire EDM slits to compare the stability of spectrophotometric data quality across the slits, as there is larger variation in slit width for the machined slits. From the lessons learned from testing this method of slit fabrication, we will bring the process to industry standards with multiple improvements. Improvements in the fabrication process in future work include using a buffered-oxide etchant prior to the potassium hydroxide etch to better control the consistency of etch time and slit widths, longer RIE etch times for the removal of remaining nitride, removal of the dicing tape by peeling instead of with acetone, and the addition of a metal backing substrate to stabilize the fragile slits. Additionally, we will characterize future slits using a scanning electron microscope for improved resolution over the optical microscope used in this paper.
To accommodate the specific needs of the APO user community, there is potential for fabrication and testing of multi-object reflective slits, for example multiple short slits. All that would be required from the user is a CAD file with the design of the slit mask. The lead time for these requests could be a week or two depending on the availability of tools at the WNF. Further testing will assess the feasibility of multiple short slits and spacing required to maintain integrity of the slit.

Acknowledgements
Special thank you to the engineers at the the Washington Nanofabrication Facility who taught this   The thin lines of the rectangle outline the dicing lanes where the wafer is diced.

List of Figures
The slit length is indicated, but slit width is not as the width varies depending on design specifications. For this work, achieved slit widths were 0.36" to 7.3" (64 µm to 1250 µm). The image is not to scale. One-dimensional example of how Canny detects strong edges above the upper threshold and connects the weak edges between the strong edges to make one detected edge. Any pixels with intensity gradient below the lower threshold is not detected as an edge.