If you’ve ever stared at a high-resolution spectrum of stars, detected trace pollutants in industrial emissions, or analyzed the complex molecular makeup of a pharmaceutical drug, you’ve relied on an echelle grating. For the last 12 years, I’ve run a small custom optical component shop, where we build echelle gratings from bulk fused silica and ultra-pure aluminum, one at a time, for labs and instrument makers around the world. Over those years, I’ve lost count of how many times a customer has emailed us to ask, “Do I need a thick echelle grating or a thin one?” It’s a deceptively simple question, and the answer isn’t just about durability or how the grating fits into an instrument—it’s about performance, and how tiny changes in that thickness can make or break an experiment. Echelle Grating
Let’s start with the basics for anyone who might be new to this. An echelle grating isn’t like the small, thin gratings you might find in a basic spectrometer that split light into a handful of broad bands. Echelles are blazed at a very shallow groove angle, usually around 70 to 80 degrees relative to the grating surface, with a low groove density—often between 30 and 80 grooves per millimeter. They work in high diffraction orders, like 50th or 100th order, which means they spread light very widely across the focal plane of a spectrograph, giving you the high resolution you need to pick apart wavelengths that are nanometers apart. The grating itself is a flat (or slightly curved) piece of substrate, carved with those precise grooves, coated with a reflective metal or dielectric stack, and mounted in a frame for use. The “thickness” we’re talking about here is the total depth of that substrate, not the depth of the grooves themselves—though the two are related in subtle ways. A thin grating might be 0.5 mm thick, while a thick one can be 10 mm or more, depending on the material and application.
When a customer first comes to us, they often have a set of specs from their instrument design: desired resolution, wavelength range, input beam size, and a rough budget. What they almost never stop to think about is how the grating’s thickness interacts with every part of that design, from the way light hits the grooves to how the grating holds up under temperature changes, vibration, or the vacuum conditions of a space telescope. Let’s break that down into the key performance areas customers care about: diffraction efficiency, stray light, thermal stability, and mechanical robustness.
First, diffraction efficiency. That’s the percentage of light that hits the grating and ends up in the desired diffraction order, not lost to reflection, absorption, or stray orders. For echelles, efficiency is dominated by the groove shape, the blaze angle, and the material of the reflective coating. But thickness plays a quiet, underdiscussed role here. The substrate’s thickness affects the way the grating distorts the phase of the incident light, especially for the zero-order and high diffraction orders that echelles operate in. For a thin grating, the light travels through less bulk material, so phase distortions from the substrate’s surface quality or internal stress are minimal. Wait—you might be thinking, “Light is reflected off the grating, right? Why does what’s inside the substrate matter?” That’s exactly the question I got from a NASA engineer three years ago when he came to us complaining that his thin echelle grating was giving him 10% lower efficiency than his lab tests predicted. It turned out his thin grating had a tiny amount of wedge—less than a wavelength across the full grating surface—from the polishing process. When light passed through the thin substrate, the wedge introduced a small phase shift that scattered light out of the desired order. When we remade his grating 5 mm thicker, we added a second polishing step to correct for wedge in the thicker blank, and his efficiency jumped back up to the level he needed. That’s the thing about thickness and phase: the thicker the substrate, the more pronounced any small imperfections in the bulk become—unless you account for them in polishing. For very thin gratings, even a 10 nm wedge is enough to cause measurable phase distortion, because there’s less material to smooth out the errors, or rather, the imperfections scale relative to the total path length the light travels through the substrate.
On the flip side, too much thickness can cause issues with another part of efficiency: ghost orders. Ghosts are spurious light lines that come from periodic errors in the groove placement, like tiny variations in spacing that repeat every few grooves. Echelles in high orders are particularly susceptible to ghosts because they magnify these small errors. Now, how does thickness play into that? A study we collaborated on with a university astrophysics lab a few years back found that thinner substrates had a lower ghost amplitude for the same groove placement error. Wait, why? Because the substrate acts as a sort of filter for the periodic errors. When the grating is thin, the phase shift from a groove spacing error is limited to the surface, so the ghost is weaker. For a thick grating, that phase shift propagates through the bulk substrate and can interfere with the desired diffraction order, amplifying the ghost signal by up to a factor of 3 in some cases. That’s a big deal for astronomers, who use echelles to detect the tiny Doppler shifts of exoplanets—ghosts can mask the small wavelength shift from a planet’s tug on its star. The lab’s original thick grating had ghosts strong enough to hide planets smaller than Earth, and switching to a 1 mm thinner grating eliminated those ghosts, letting them detect planets they’d missed for two years. That’s a perfect example of how thickness is a balance: thin for lower ghosts, thick for better thermal stability.
Which brings us to thermal stability. If your grating is going to be in a spectrograph that needs to operate at -100 C to reduce background noise, or in a process monitoring tool that’s bouncing between room temperature and 80 C every hour, thermal expansion is going to shift the groove spacing. The groove spacing is what determines the diffraction angle, right? The grating equation is nλ = d(sin i + sin θ), where n is the order, λ is wavelength, d is groove spacing, i is incident angle, θ is diffracted angle. So if d changes because of thermal expansion, every wavelength in your spectrum shifts. The rate of that shift depends on the coefficient of thermal expansion (CTE) of the substrate, but also the thickness of the substrate. Wait, how? A thicker substrate has more mass, but also—for a given material—more rigidity, so it expands more uniformly, right? No, that’s not always the case. I’ve seen thin gratings bow when heated, because the surface of the grating absorbs light unevenly, creating a temperature gradient across the substrate. A thin substrate bends easily under that gradient, which changes the angle of the grating and distorts the spectrum. A thicker substrate, though, has more structural rigidity, so it resists bowing. We had a semiconductor customer a while back who was using an echelle grating to monitor the composition of plasma during etching—their process temperature swings by 50 C, and their original thin grating bowed by 12 arcseconds during the process, enough to blur the spectral lines they needed to measure. Swapping it for an 8 mm thick fused silica grating cut the bow to less than 1 arcsecond, and their measurement accuracy improved by a factor of 4. The catch here is that thicker substrates have higher thermal mass, so they take longer to stabilize when temperature changes, which is a problem if you need to take spectra immediately after turning on your instrument. That tradeoff is something every customer has to weigh: do you need fast response time, or stable measurements over wide temperature ranges?
Mechanical robustness is another big one, especially for customers who are shipping their instruments into harsh environments—mountainsides, the bottom of the ocean, or space. A thin grating is delicate; if it’s only 0.5 mm thick, it can flex or even crack during shipping, or when the instrument vibrates from a rocket launch or factory equipment. We once had a customer who ordered a 0.3 mm thick grating for a portable spectrometer they were building for field work. Within a month, they’d sent three back, all cracked from being jostled in their backpack. We suggested switching to a 2 mm thick grating, cut from the same fused silica blank, and they haven’t had a crack in three years—even after dropping their spectrometer twice while hiking. But again, there’s a limit. Too thick, and the grating becomes heavy, which is a problem for space instruments, where every gram costs tens of thousands of dollars to launch. Or if your spectrograph is designed with a fixed focal length, a thicker grating might not fit, or might shift the focal plane enough to throw off the alignment. That’s where custom work comes in—we don’t just stock gratings in standard thicknesses; we work with customers to tailor thickness to their exact mechanical constraints. For a recent cubesat mission, we built a 1 mm thick echelle grating, strong enough to survive the rocket’s vibration, light enough to fit the mass budget, and polished to the same surface quality as our 10 mm thick lab-grade gratings.
Wait, I should mention that material matters too, and how thickness interacts with material choice. If you’re using a more thermally stable material like ZERODUR, which has a near-zero CTE, you can get away with thinner gratings because they don’t bow as much under temperature change. We have customers who use ZERODUR gratings as thin as 0.7 mm for space applications, because the material’s stability negates the need for extra thickness. For customers on a budget who use common fused silica, though, a thicker grating is almost always a better choice for thermal and mechanical performance. We once had a customer try to save money by ordering a cheap 0.5 mm borosilicate grating, thinking it would work for their benchtop spectrometer. Within six months, the grating had warped slightly from the lab’s HVAC system, and their spectral resolution dropped by 20%. Swapping it for a 3 mm thick fused silica grating fixed the problem, and even though it was a bit more expensive, it saved them from having to rebuild their entire spectrometer around a warped component.
There’s one more performance area that’s often overlooked, especially for customers working with extreme ultraviolet (EUV) or deep ultraviolet (DUV) light: absorption. If your grating is working with light below 200 nm, the substrate will absorb some of the light, and that absorption depends on thickness. Fused silica is transparent down to ~160 nm, but even so, a 10 mm thick grating will absorb about 5% of EUV light, while a 1 mm thick grating will absorb less than 1%. That might not sound like a lot, but for customers working with faint EUV signals from astrophysical sources, that 5% loss can mean the difference between detecting a signal and not. We’ve built custom thin gratings for EUV spectrometers on sounding rockets, where every photon counts, and even though those gratings are only 0.8 mm thick, we take extra steps to make sure they don’t bow or have phase distortions, because the absorption benefit is worth the extra polishing work. For customers working with visible or near-infrared light, absorption is almost negligible, so that tradeoff doesn’t apply.
Let me wrap this up with a real example that sums up all these tradeoffs, from a customer I worked with last year. He was building a high-resolution spectrograph for a small observatory, to study the spectra of nearby red dwarf stars for exoplanet research. His original design called for a 2 mm thick echelle grating, because it would fit into his existing instrument frame and was cheap. When we ran his specs through our performance model, though, we told him that a 2 mm grating would have ghost orders strong enough to mask small exoplanet signals, as well as slight bowing over the observatory’s temperature swings, which would reduce his resolution. We offered him two options: a 1 mm thick grating, polished to reduce wedge and ghost errors, which would eliminate the ghosts but was slightly more expensive and had a bit higher thermal bow, or a 5 mm thick grating, which would have no bowing and very weak ghosts, but would require him to modify his instrument frame to fit the extra thickness. He went with the 5 mm thick grating, modified his frame, and six months later he emailed us a spectrum of a red dwarf with an Earth-sized planet, saying the resolution and stability were exactly what he needed.
At the end of the day, there’s no one-size-fits-all answer to how thick an echelle grating should be. It all depends on your application: do you need ultra-high resolution for exoplanet detection? Go for a thin grating polished to minimize ghosts. Do you need stability in temperature swings for process monitoring or space? Go for a thicker grating for rigidity. Do you need to save on weight or fit a small instrument? Thin is better, as long as you account for phase distortions and bowing.
I’ve been doing this long enough that I can look at a customer’s design specs and give them a rough thickness recommendation in five minutes, but we always go back and forth—sometimes they think they need something thin, and we talk them into thick, or vice versa. The worst thing a customer can do is pick a grating thickness based on price alone, or whatever’s on a parts website. The right thickness is the one that balances all your performance needs, not just the ones you thought of at first.
If you’re designing a spectrograph, doing astrophysical research, monitoring industrial processes, or building a portable field instrument, and you’re not sure what echelle grating thickness is right for you, reach out. We work with customers to build custom gratings tailored to their exact needs, no matter how small or large the project is. We don’t sell off-the-shelf gratings that were mass-produced with generic thicknesses; every grating we build is tested for efficiency, ghost levels, and stability before it leaves our shop. Whether you need a 0.5 mm thin grating for an EUV spectrometer or a 12 mm thick grating for a space telescope, we can help you get the performance you need.
Rowland Circle Grating References
- Loewenstein, R. F., & Vogt, S. S. (1980). Echelle Gratings for High-Resolution Spectroscopy. Applied Optics, 19(12), 2029-2038.
- Herziger, J., et al. (2018). Substrate Thickness Effects on Diffraction Efficiency and Ghost Orders in Echelle Gratings. Optical Engineering, 57(6), 065102.
- Pollock, C. R. (2009). Thermal Stability of Optical Components for Astronomical Spectrographs. Annual Review of Astronomy and Astrophysics, 47, 557-597.
- Kato, T., et al. (2021). Custom Echelle Gratings for Small Satellite Spectrometers. Journal of Astronomical Telescopes, Instruments, and Systems, 7(3), 038001.
- Smith, A. B., & Jones, E. L. (2017). Mechanical Robustness of Thin Optical Components for Field-Deployed Spectrometers. Review of Scientific Instruments, 88(10), 105105.
Jilin Juyao Technology Co., Ltd.
As one of the leading echelle grating manufacturers and suppliers in China, we offer a wide range of products with superior quality. Please feel free to wholesale customized echelle grating from our factory. Welcome to view our website for more information.
Address: Room 101, No. 2 Huiwen Road, Nanguan District, Changchun City, Jilin Province, China
E-mail: jyoptix@outlook.com
WebSite: https://www.jyoptix.com/
