NASA-supported researchers studying a starshade to be used with a space telescope ran into an unexpected problem.
To directly image a planet outside the solar system, you first have to block the light from the star it orbits. A star emits its own light, while a planet reflects only a small fraction of that light. If even a little starlight leaks into the telescope, the planet's faint signal is buried.
One device being developed for this purpose is the starshade. It is a giant flower-shaped screen positioned far from a space telescope so that it blocks the light of a distant star before that light reaches the telescope. No starshade has yet been deployed on an actual observing mission; the technology is still being developed and tested for future exoplanet observations.
But another source of light creates trouble: light from our own Sun.
If sunlight hits the edges of the starshade's petals and scatters, some of it can reach the telescope. NASA-supported researchers sharpened those edges to a thickness of about 300 nanometers, only a few hundredths to a few hundredths of the width of a human hair.
Even then, the scattered light was too strong.
So why not make the edge black?
The problem was how conventional ultra-black coatings become black. Coatings based on carbon nanotubes or three-dimensional microstructures create structures of a certain height above the surface to trap light. They can be extremely black, but they can also be several micrometers thick. Applied to a 300-nanometer edge, they would blunt the carefully fabricated tip, increasing edge-scattered light again.
A starshade needed a coating that was both extremely black and extremely thin.
The coating developed by ZeCoat used a different approach. Instead of tall microstructures, it stacked multiple very thin, partially transmitting metal layers with dielectric glass layers. By tuning the thicknesses precisely, light forms standing waves inside nanoscale cavities, while the metal layers absorb the optical energy.
If a carbon-nanotube coating makes light wander through a deep forest, this method is more like keeping light bouncing inside a very thin layered cavity.
The coating was about one hundredth as thick as the coatings previously considered. Applied to a 50-centimeter-long prototype starshade edge, it reduced reflected light by about a factor of 20.1
The sharp geometry was preserved while the needed absorption performance was achieved.
A carbon forest that absorbed 99.995%
That does not mean records for lowest reflectance are unimportant.
In 2019, MIT researchers were growing carbon nanotubes on aluminum when they discovered an unusually black surface. MIT reported that it absorbed at least 99.995% of incident light. At the time, the amount of light it reflected was roughly one-tenth that of other known ultra-black materials.
The principle was similar to that of the early Vantablack discussed in Part 1. Light entering the gaps among many carbon nanotubes changes direction repeatedly and loses opportunities to escape.
Interestingly, the researchers were not trying to set an ultra-black record. They were studying a way to break down aluminum's native oxide layer so that carbon nanotubes could bond directly to the metal, and they discovered the unexpected optical absorption performance along the way.2
Does that mean any material with a number lower than 99.995% is always less black?
Change the wavelength, and the answer changes.
A surface that looks black to humans can look bright to a sensor
Japan's Koyo Orient makes Musou Black, an ultra-black acrylic paint that can be applied with a brush or airbrush.
Light behaves differently at different wavelengths. Visible light for humans lies roughly between 380 and 780 nanometers. According to the manufacturer's data, Musou Black absorbs about 99.4% of visible light, and at a wavelength of 550 nanometers it reflects about 0.6%.
At longer wavelengths, the result changes. Reflectance rises to 2.0% at 950 nanometers and 10.7% at 1,500 nanometers. Both wavelengths fall in the near-infrared region, invisible to human eyes.
To a person, the surface looks very black. But to a sensor operating at 1,500 nanometers, it reflects roughly one-tenth of the incoming light.
Koyo Orient developed other ultra-black materials for infrared equipment. Its IR Flock Sheet has a reflectance of about 0.3% at both 550 and 950 nanometers, and about 0.5% even at 1,500 nanometers. Key applications include LiDAR, machine vision, and infrared sensor testing.3
To human eyes, both are black.
To a machine, they are not the same black.
If it is too black, even the shape of a car disappears
Even if you fix the wavelength, there is another complication in the meaning of “black.” The same surface can look different depending on the direction from which light arrives and the direction in which it leaves.
In 2019, BMW coated an X6 with a Vantablack-family material. But it did not choose the blackest Vantablack.
The material was spray-applied VBx2. Its total hemispherical reflectance (THR)—the total amount of light reflected in all directions—is about 1%. That is still extremely black, but it leaves more reflected light than the original Vantablack.
The reason was the shape of the car. Convex and concave surfaces, faces pointed toward the light, and surfaces tilted away from it all return different amounts of light. Our eyes use those brightness differences to read the contours and three-dimensional shape of the body.
If reflectance is reduced too far, those visual cues disappear as well. Surrey NanoSystems said that if the original Vantablack had been used, the X6 would have lost almost all sense of three-dimensional form. VBx2's 1% reflectance left enough light for viewers to perceive the car's silhouette and curvature.
Only one Vantablack-coated X6 was made, as an exhibition show car. Surrey NanoSystems also said that making VBx2 durable enough for everyday automotive paint would require substantial improvements in durability.4
Reflected light can be noise that needs to be removed, but it can also be information that reveals an object's shape.
You can see this directionality in an ordinary black car under sunlight. Most of the body may look dark, but when you change your viewing position, one surface suddenly flashes bright.
Call the angle at which light reaches the surface the angle of incidence, and the angle from which the surface is viewed the observation angle. Even for the same surface, changing the combination of those two angles changes how much light returns to the eye or sensor.
A detailed way of measuring this is the Bidirectional Reflectance Distribution Function, or BRDF.
The name is long, but the idea is simple. Change the incidence and observation angles and measure how much light returns for each combination. It is effectively a map of how bright a surface becomes when illuminated and viewed from different directions.
A study published in 2026 compared Vantablack, Musou Black paint, Musou black fabric, black velvet, and ordinary black paint using the same apparatus. The researchers measured reflected brightness for 24,488 combinations of incidence and observation angles for each material.
When those brightness values were ordered from dark to bright, each material showed a different pattern. At the median point, Musou's black fabric had the lowest brightness. In other words, its typical reflectance across all the measured angle combinations was especially low.
At the bright end, however, Vantablack performed best. Even at the 99th percentile, near the brightest 1% of measured values, Vantablack had the lowest reflectance. When other materials became relatively bright at particular angles, Vantablack suppressed those bright reflections more strongly.
Put simply, Musou's black fabric was especially dark under typical angle combinations, while Vantablack stayed darkest even under unfavorable angle combinations that produced brighter reflections.
The researchers also used the BRDF data to render computer-generated spheres made from each material and asked 36 participants which surfaces looked darker. As illumination increased, the difference between ordinary black paint and ultra-black materials became larger, and Vantablack and Musou black fabric were rated darkest under all three lighting conditions. Human judgments and physical reflectance measurements matched fairly well.5
A baffle needs a different kind of black
Telescopes and cameras contain black light-blocking structures around lenses and sensors to prevent stray reflections from reaching the detector. These are called baffles.
A baffle has different requirements from a starshade. Instead of preserving an extremely thin sharp edge, it has to suppress light over many directions and a broad wavelength range, and it has to be applied reliably to complex metal components.
A 2025 study by researchers in Korea examined a method of directly growing carbon nanofibers (CNFs) on Al6061 aluminum alloy used in optical equipment.
The way carbon nanofibers absorb light is similar to carbon nanotubes. Light entering the dense carbon structure encounters repeated reflections and is absorbed before it can escape.
The challenge was to create this structure uniformly and robustly on a real aluminum component.
The researchers placed a nickel catalyst inside the porous oxide layer on the aluminum surface, then grew carbon nanofibers directly using thermal chemical vapor deposition (CVD). Instead of attaching a separately manufactured black film, they grew the ultra-black structure itself on the component surface.
The chromium-rich surface layer of black-anodized Al6061 also played a role. The researchers concluded that the chromium-nickel interface helped stabilize the catalyst and contributed to uniform CNF growth as well as the coating's mechanical and ultraviolet stability.
The optimized coating absorbed an average of about 99% of light from 0.3 to 2.5 micrometers, and about 98.5% on average in the infrared region from 2.5 to 15 micrometers. It maintained high absorption over multiple incidence angles.
The surface also repelled water and remained stable under mechanical stress and ultraviolet testing. The result was tailored to a baffle's needs: absorb light across broad wavelengths and directions while surviving on an actual metal component.
The final validation was performed on a real baffle. In LightTools and Zemax optical simulations based on BRDF data, and in experiments at multiple incidence angles, the new coating reduced stray light by more than 10 decibels compared with conventional black anodizing. A decibel expresses a ratio between two quantities on a logarithmic scale; here, it means the stray-light intensity fell to one-tenth or less. The material-level absorption performance translated into reduced stray light in an actual optical system.6
Starshades and baffles both need to eliminate unwanted light, but they require different kinds of black.
A starshade needs a black coating thin enough not to damage a razor-sharp edge. A baffle needs a durable black surface that can absorb broad wavelengths and many directions on a complex metal part.
The question for measuring black became longer
A 2026 review of ultra-black materials proposed looking at the field through three elements together: material, scene, and characterization.
In other words, you have to consider what the material is made of, where it will be used, and under what conditions it is measured.7
That does not make records such as 99.995% meaningless. In precision optics, where even tiny reflections matter, absolute reflectance remains important.
But many cases can no longer be answered by a single number.
Is the relevant light visible to humans, or infrared? From what direction does it arrive, and in what direction does it leave? Is the sample a flat laboratory coupon or a real component? Once it is installed in the instrument, does it actually reduce the unwanted light?
“What is the blackest black in the world?” is still an entertaining question.
The problem researchers are actually solving is more specific.
Which light are you trying to eliminate, where, and for what purpose?
Notes
1. NASA Science's 2026 technology-development report explains that a starshade must suppress not only the light of a distant star but also scattered light from our own Sun. The same source describes an approximately 300-nanometer blade edge, the thickness problem with conventional CNT and three-dimensional microstructure coatings, ZeCoat's multilayer cavities made from partially transmitting metal and dielectric glass layers, a coating roughly 100 times thinner, and an approximately 20-fold reduction in reflection on a 50-centimeter test edge. NASA Science
2. Kehang Cui and Brian L. Wardle, “Breakdown of Native Oxide Enables Multifunctional, Free-Form Carbon Nanotube–Metal Hierarchical Architectures,” ACS Applied Materials & Interfaces 11(38), 2019, 35212–35220. The paper reports broadband reflectance of 1×10⁻⁵, while MIT introduced the result conservatively as “at least 99.995%” absorption in light of measurement limits. The article uses MIT's conservative figure. ACS/DOI · MIT AeroAstro
3. The approximate 380–780 nanometer range for visible light is aligned with CIE and NIST standards and technical references. The wavelength-dependent total reflectance and applications of Musou Black and IR Flock Sheet are manufacturer data from the Koyo Orient group. These are manufacturer reference values, not independently reverified measurements under identical comparison conditions. CIE · Musou Black · IR Flock Sheet
4. BMW's 2019 X6 was a one-off show car coated with Vantablack-family VBx2. BMW gave VBx2 a total hemispherical reflectance (THR) of 1%, and Ben Jensen explained that using the original Vantablack would have made the car lose its three-dimensional appearance. In the same official interview, he said achieving the durability required for everyday automotive paint would be a major technical challenge. BMW Group
5. Jiří Filip and Radomír Vávra, “How dark is dark? A reflectance and scattering analysis of black materials,” Journal of the Optical Society of America A 43(7), 2026, 1037–1045. The study measured 24,488 valid incidence-observation direction combinations per material. Musou fabric had the lowest 1st and 50th percentile relative luminance, while Vantablack had the lowest 99th percentile. Thirty-six participants rated samples rendered from the measured BRDF data under three illumination conditions, and Vantablack and Musou fabric were rated darkest at all illumination levels. Optica/DOI · arXiv public manuscript
6. Geun Tak Yuk et al., “Direct CVD Growth of Robust Carbon Nanofiber-Based Super Black Coatings on Complex-Curvature and Commercial Anodized Al(6061) for Enhanced Stray-Light Suppression in Optical Baffle Systems,” ACS Applied Engineering Materials 3(9), 2025, 3049–3064. The paper was published online on September 17, 2025 and appeared in the September 26 issue. It reports Ni catalyst integration, a Cr–Ni interface, average absorption of 99% from 300–2500 nanometers and 98.5% from 2500–15,000 nanometers, mechanical and UV stability, hydrophobicity, and more than 10 dB of stray-light suppression compared with conventional black anodizing in an actual baffle. ACS/DOI
7. Peilin Yang et al., “Ultra-Black Materials: Fundamentals, Design, and Applications,” Laser & Photonics Reviews 20(16), 2026, e02136. First published on May 29, 2026, the review organizes ultra-black research using a material–scene–characterization framework and describes a shift from single reflectance figures toward multidimensional characterization. Wiley/DOI