Dispersion Leverage Coronagraph (DLC)

• Dispersion Leverage Coronagraph (DLC) is a family of techniques that leverage wavelength dispersion to distinguish faint companion signals from overwhelming stellar light.
• DLC implementations combine high-contrast imaging, high-resolution spectroscopy, and single-mode fiber filtering to separate planetary spectral features from residual speckles.
• Key trade-offs in DLC designs involve balancing spectral resolution, fiber coupling efficiency, and adaptive optics performance to optimize exoplanet characterization.

Searching arXiv for the cited DLC/HDC papers to ground the article in the current literature. Dispersion Leverage Coronagraph (DLC) denotes a family of coronagraphic concepts in which wavelength dispersion is used as an additional discriminant, rather than treating coronagraphy as a purely spatial nulling problem. In one widely used formulation, DLC is instantiated as High-Dispersion Coronagraphy (HDC), which combines high-contrast starlight suppression, single-mode spatial filtering, and high-spectral-resolution spectroscopy so that residual stellar speckles are separated from planet signals in spectral and Doppler space. In other formulations, dispersion is built directly into the coronagraphic optic, as in broadband scalar vortex schemes and in nulling interferometric designs for Primary Objective Grating telescopes. Across these usages, the common idea is that dispersion is not merely tolerated but actively leveraged to preserve an achromatic null, relax raw contrast requirements, or improve discrimination between stellar leakage and companion light (Wang et al., 2017, Mawet et al., 2017, Errmann et al., 2013, Swordy et al., 15 Jul 2025).

1. Terminology and conceptual scope

The literature uses “Dispersion Leverage Coronagraph” in related but non-identical ways. In the HDC usage, the coronagraph acts as a spatial filter isolating the planet, while the spectrograph acts as a spectral filter, leveraging the very different high-resolution line structure and Doppler shifts of planet versus star. In the broadband scalar vortex usage, dispersion is balanced between phase gratings so that chromatic angular dispersion is canceled while the vortex phase remains. In the Primary Objective Grating usage, a two-arm nuller exploits the spectral mapping created by large dispersive primary gratings so that pupil inversion aligns the stellar spectrum and nulls it for all wavelengths simultaneously (Wang et al., 2017, Errmann et al., 2013, Swordy et al., 15 Jul 2025).

Usage in the literature	Defining mechanism	Representative parameters from the literature
HDC / DLC	Coronagraphy plus high-resolution spectroscopy	$R \gtrsim 10^4$ to $10^5$; raw suppression relaxed to $10^{-4}$ in some ground-based cases
Broadband scalar vortex	Grating plus CGH cancel chromatic tilt and retain vortex phase	$\Delta \lambda = 120\,\mathrm{nm}$ centered at $700\,\mathrm{nm}$; inner working angle $\sim \lambda/D$
POG-based DLC	AIC-like nulling adapted to dispersive primary gratings	Achromatic on-axis null across the focal plane; secondary disperser at $R \approx 10^5$ to $10^6$ in the DICER case

This multiplicity of meanings is not a contradiction. It suggests that “dispersion leverage” is best understood as a design principle: dispersion is arranged so that the unwanted stellar term becomes easier to reject than the companion term. What differs across implementations is whether that leverage is realized in a spectrograph, a vortex generator, a pupil-remapping coronagraph, or an interferometric nuller.

2. Spectral leverage in high-dispersion coronagraphy

In HDC, the central objective is to combine high-contrast imaging techniques such as adaptive optics and wavefront control plus coronagraphy with high spectral resolution spectroscopy. At high spectral resolving power, the quasi-static speckle noise from the star appears as a smooth continuum, while the planet’s narrow molecular lines stand out and Doppler shift in time. Mawet et al. summarize the scaling as

$$S/N \simeq \eta \cdot S_p \cdot \sqrt{N_{\mathrm{lines}}}\ /\ \sqrt{S_\star/K + \sigma^2_{\mathrm{bg}} + \sigma^2_{\mathrm{rn}} + \sigma^2_{\mathrm{dark}}}.$$

Here $S_p$ and $10^5$0 are the planet and star photo-electron rates, $10^5$1 is the coronagraphic raw suppression, $10^5$2 is the total throughput from planet to spectrograph, and $10^5$3 is the multiplexing gain of many resolved lines (Mawet et al., 2017).

Wang et al. express the same principle through single-channel and cross-correlation scalings. In one spectral channel,

$10^5$4

and for $10^5$5 resolved lines,

$10^5$6

For fixed overall bandwidth, $10^5$7, so $10^5$8 grows approximately as $10^5$9 in the photon-noise limit, or even as $10^{-4}$0 if the speckle noise truly goes away linearly. The associated relaxation of the raw suppression requirement is written as

$10^{-4}$1

This is the formal basis for the claim that dispersion leverage can trade spatial suppression for spectral filtering (Wang et al., 2017).

A recurrent misconception is that spectral leverage is simply “more resolution is always better.” The trade studies do not support that proposition in a universal form. For ground-based $10^{-4}$2 telescopes, high $10^{-4}$3 remains favorable because stellar photon flux per pixel is large and detector noise is sub-dominant. For space concepts, detector noise and speckle chromatic noise produce finite optima: $10^{-4}$4 for HabEx and $10^{-4}$5 for LUVOIR in the simulations reported by Wang et al. (Wang et al., 2017).

3. Active single-mode fiber injection and on-fiber wavefront control

A key HDC implementation step is the active single-mode fiber injection unit demonstrated by Mawet et al. Its layout comprises an actuated Tip–Tilt Mirror upstream of the fiber, a beamsplitter or dichroic that sends approximately $10^{-4}$6 of the science beam to a tracking camera, a corner-cube retroreflector and a back-injected calibration source used to locate the precise position of the single-mode fiber tip on the tracking camera, and a single-mode fiber mounted on a three-axis alignment stage. During acquisition, the calibration laser back-illuminates the fiber tip; its reflection off the corner cube marks the fiber location; and a closed feedback loop adjusts the Tip–Tilt Mirror until the planet PSF and the fiber beacon coincide on the tracking camera (Mawet et al., 2017).

The fiber-coupling efficiency is the overlap integral between the incident focal-plane field and the SMF fundamental mode:

$10^{-4}$7

For an unobstructed circular pupil and a perfect Airy-to-Gaussian match, $10^{-4}$8. In the laboratory, Mawet et al. report $10^{-4}$9–$\Delta \lambda = 120\,\mathrm{nm}$0 at $\Delta \lambda = 120\,\mathrm{nm}$1, with the shortfall attributed to residual aberrations and Fresnel or interface losses; on sky with Subaru/SCExAO they quote $\Delta \lambda = 120\,\mathrm{nm}$2 in $\Delta \lambda = 120\,\mathrm{nm}$3-band. The same system achieves blind offsets good to $\Delta \lambda = 120\,\mathrm{nm}$4 within a few seconds through a calibrated mapping from camera pixels to Tip–Tilt Mirror voltages (Mawet et al., 2017).

Residual stellar leakage that overlaps the fiber is attacked by coherent modulation and speckle nulling. A sinewave ripple placed on the deformable mirror,

$\Delta \lambda = 120\,\mathrm{nm}$5

creates two anti-speckles at $\Delta \lambda = 120\,\mathrm{nm}$6 in the focal plane. The experiment measures the coupled stellar leakage through the SMF and photodiode and iteratively adjusts $\Delta \lambda = 120\,\mathrm{nm}$7, accepting a new DM shape if the coupled power decreases. Because the SMF mode weights the overlap integral, the reported monochromatic on-fiber suppression on individual speckles at $\Delta \lambda = 120\,\mathrm{nm}$8 exceeds $\Delta \lambda = 120\,\mathrm{nm}$9, while image-based speckle nulling on the tracking camera yields only $700\,\mathrm{nm}$0–$700\,\mathrm{nm}$1. In a broadband simulation over $700\,\mathrm{nm}$2 at $700\,\mathrm{nm}$3 with $700\,\mathrm{nm}$4 rms aberrations, the reported suppression is $700\,\mathrm{nm}$5 across the band with no planet throughput penalty; in that run a $700\,\mathrm{nm}$6 gain is noted (Mawet et al., 2017).

4. Broadband vortex and PIAA realizations

Errmann, Minardi, and Pertsch present a broadband scalar vortex coronagraph that realizes dispersion leverage by using two phase-only elements in series: a grating $700\,\mathrm{nm}$7 that angularly disperses each wavelength and a computer-generated hologram combining the same grating period with a charge-$700\,\mathrm{nm}$8 phase singularity. In the $700\,\mathrm{nm}$9 diffraction order, the CGH adds the azimuthal phase term $\sim \lambda/D$0 and re-diffracts each wavelength so that the net propagation angle is zero. The combined on-axis phase is therefore

$\sim \lambda/D$1

because the two linear-ramp terms cancel. This arrangement generates a scalar optical vortex of fixed topological charge over a broad band, so that starlight can be nulled at small angles (Errmann et al., 2013).

The measured laboratory performance is a constant peak-to-peak attenuation below $\sim \lambda/D$2 over a bandwidth of $\sim \lambda/D$3 centered at $\sim \lambda/D$4, with a more detailed null depth reported as $\sim \lambda/D$5 over $\sim \lambda/D$6–$\sim \lambda/D$7. An inner working angle of $\sim \lambda/D$8 is demonstrated along with a raw contrast of $\sim \lambda/D$9 magnitudes at $R \approx 10^5$0. The vortex unit throughput is reported as $R \approx 10^5$1, constant over $R \approx 10^5$2–$R \approx 10^5$3, and the charge $R \approx 10^5$4 design is chosen for the smallest inner working angle $R \approx 10^5$5 (Errmann et al., 2013).

A different small-angle realization appears in RISTRETTO, which is explicitly designed to enable HDC at $R \approx 10^5$6. Its coronagraphic IFU is based on a modified version of the PIAA apodizer, allowing nulling on the first diffraction ring. The instrument combines an extreme adaptive optics system, a coronagraphic Integral Field Unit, and a diffraction-limited spectrograph with $R \approx 10^5$7 over $R \approx 10^5$8–$R \approx 10^5$9. For the proposed design, the reported potential performance is $10^6$0 coupling and $10^6$1 contrast at $10^6$2 in median seeing conditions. The corresponding AO requirements include global WFE $10^6$3 RMS, low-order WFE $10^6$4 RMS, at least $10^6$5 actuators across the pupil, and loop rate $10^6$6; end-to-end OOMAO simulations for a $10^6$7 DM and $10^6$8 pyramid WFS give $10^6$9 and $$S/N \simeq \eta \cdot S_p \cdot \sqrt{N_{\mathrm{lines}}}\ /\ \sqrt{S_\star/K + \sigma^2_{\mathrm{bg}} + \sigma^2_{\mathrm{rn}} + \sigma^2_{\mathrm{dark}}}.$$0 (Blind et al., 2022).

5. DLC as an achromatic nuller for Primary Objective Grating telescopes

In the 2025 formulation, DLC is a novel variation of the Achromatic Interfero Coronagraph designed specifically for optical systems featuring large, dispersive Primary Objective Gratings. The optical train contains two identical linear diffraction gratings of length $$S/N \simeq \eta \cdot S_p \cdot \sqrt{N_{\mathrm{lines}}}\ /\ \sqrt{S_\star/K + \sigma^2_{\mathrm{bg}} + \sigma^2_{\mathrm{rn}} + \sigma^2_{\mathrm{dark}}}.$$1 and width $$S/N \simeq \eta \cdot S_p \cdot \sqrt{N_{\mathrm{lines}}}\ /\ \sqrt{S_\star/K + \sigma^2_{\mathrm{bg}} + \sigma^2_{\mathrm{rn}} + \sigma^2_{\mathrm{dark}}}.$$2, two secondary telescopes, and a nulling periscope interferometer in which one arm acquires a $$S/N \simeq \eta \cdot S_p \cdot \sqrt{N_{\mathrm{lines}}}\ /\ \sqrt{S_\star/K + \sigma^2_{\mathrm{bg}} + \sigma^2_{\mathrm{rn}} + \sigma^2_{\mathrm{dark}}}.$$3 phase shift and pupil inversion before recombination on a $$S/N \simeq \eta \cdot S_p \cdot \sqrt{N_{\mathrm{lines}}}\ /\ \sqrt{S_\star/K + \sigma^2_{\mathrm{bg}} + \sigma^2_{\mathrm{rn}} + \sigma^2_{\mathrm{dark}}}.$$4 beamsplitter. Constructive-interference light is dumped, while the destructive-interference port is re-imaged onto a curved focal surface. In the DICER use case, a bank of high-resolution immersion gratings with $$S/N \simeq \eta \cdot S_p \cdot \sqrt{N_{\mathrm{lines}}}\ /\ \sqrt{S_\star/K + \sigma^2_{\mathrm{bg}} + \sigma^2_{\mathrm{rn}} + \sigma^2_{\mathrm{dark}}}.$$5 to $$S/N \simeq \eta \cdot S_p \cdot \sqrt{N_{\mathrm{lines}}}\ /\ \sqrt{S_\star/K + \sigma^2_{\mathrm{bg}} + \sigma^2_{\mathrm{rn}} + \sigma^2_{\mathrm{dark}}}.$$6 is then placed in the focal beam so that each detector pixel simultaneously measures focal-plane position and wavelength (Swordy et al., 15 Jul 2025).

The single-wavelength null-port transmission is written as

$$S/N \simeq \eta \cdot S_p \cdot \sqrt{N_{\mathrm{lines}}}\ /\ \sqrt{S_\star/K + \sigma^2_{\mathrm{bg}} + \sigma^2_{\mathrm{rn}} + \sigma^2_{\mathrm{dark}}}.$$7

When the sky angles satisfy $$S/N \simeq \eta \cdot S_p \cdot \sqrt{N_{\mathrm{lines}}}\ /\ \sqrt{S_\star/K + \sigma^2_{\mathrm{bg}} + \sigma^2_{\mathrm{rn}} + \sigma^2_{\mathrm{dark}}}.$$8, one has $$S/N \simeq \eta \cdot S_p \cdot \sqrt{N_{\mathrm{lines}}}\ /\ \sqrt{S_\star/K + \sigma^2_{\mathrm{bg}} + \sigma^2_{\mathrm{rn}} + \sigma^2_{\mathrm{dark}}}.$$9 and $S_p$0, so $S_p$1 for all $S_p$2. The notable point is that the null is achromatic across the focal plane even though the primary optic is highly dispersive (Swordy et al., 15 Jul 2025).

The paper derives finite-star leakage, OPD tolerance, and jitter requirements. For DICER parameters $S_p$3, the leakage estimate is $S_p$4, corresponding to star-light suppression of order $S_p$5. For a target rejection $S_p$6 at $S_p$7, the path-length requirement is $S_p$8. The jitter tolerances are asymmetric: to achieve $S_p$9, the requirements are $10^5$00 $10^5$01 in the dispersion axis and $10^5$02 $10^5$03 in the orthogonal axis; for $10^5$04, the requirement relaxes to $10^5$05 (Swordy et al., 15 Jul 2025).

6. Performance regimes, trade-offs, and scientific reach

The principal performance claim of HDC-style DLC is relaxation of the raw starlight suppression requirement. Wang et al. report that for ground-based telescopes, HDC observations can detect an Earth-like planet in the habitable zone around an M dwarf star at $10^5$06 starlight suppression level, compared to the $10^5$07 planet/star contrast, so the requirement is relaxed by a factor of $10^5$08. For space-based concepts, the same paper reports a relaxation factor of $10^5$09 for HabEx and $10^5$10 for LUVOIR for a planet with contrast $10^5$11, with detector noise becoming a major limitation at spectral resolutions higher than $10^5$12 (Wang et al., 2017).

The instrument-specific trade spaces differ substantially. In RISTRETTO, the decisive variables are inner working angle, off-axis throughput, low-order wavefront stability, and fiber-coupled raw contrast at $10^5$13. In the POG-based DLC, the decisive variables are instead OPD control, grating line-spacing uniformity, thermal stability, fine-guidance asymmetry, and zodiacal-background control through a second disperser. This suggests that “DLC performance” cannot be summarized by null depth alone; the governing bottleneck depends on whether the leverage is spectral, modal, interferometric, or pupil-remapping in character (Blind et al., 2022, Swordy et al., 15 Jul 2025).

The scientific reach described in the literature is correspondingly broad. HDC is presented as a pathway toward characterizing exoplanet atmospheres across a broad range of masses from giant gaseous planets down to Earth-like planets, including molecular composition, Doppler mapping of temperature, clouds, and wind, and precise measurements of rotational velocities. RISTRETTO targets Prox Cen b and other planets at about $10^5$14 from their star, corresponding to $10^5$15 at $10^5$16. The benchmark DICER simulation reports that the POG-based DLC could plausibly find and characterize approximately $10^5$17 nearby, habitable exoplanets around Sun-like stars in a seven year mission, with about $10^5$18 of the habitable exoplanets within $10^5$19 found in the simulation; the same study estimates roughly $10^5$20 days per planet for $10^5$21 ozone spectroscopy near $10^5$22 for an average Earth analog (Mawet et al., 2017, Swordy et al., 15 Jul 2025).

Taken together, these results define DLC not as a single canonical coronagraph, but as a research program in which dispersion is deliberately structured so that the stellar term becomes easier to reject than the companion term. In HDC that structure is created downstream in the spectrograph and the SMF; in the broadband scalar vortex it is created by matched dispersive phase elements; in RISTRETTO it is combined with PIAA apodization and a fiber-fed IFU; and in the POG architecture it is embedded in the interferometric geometry of the telescope itself.