A preprint study of 11 low-amplitude, totally eclipsing contact binaries reports that the uneven brightness pattern between eclipses can reverse during a single TESS observing sector. In J011323 and J002747, the measured difference between the two out-of-eclipse brightness peaks changed from -0.0132 to +0.0131. The authors interpret the changing pattern as consistent with starspot migration on the more massive component.
A changing pattern in the light
The paper calls this unevenness the O'Connell effect. It is simply the difference between the first and second brightness maxima outside an eclipse. A change from a negative value to a positive one means the order of the two peaks has reversed, which is what happened in the two TESS cases.
The researchers combined multiband photometry collected from 2022 to 2024 at four ground-based facilities with public photometry from TESS, ZTF, SuperWASP, CRTS and ASAS-SN. All of the light curves were modeled with PHOEBE version 2.4, while CBLA supplied starting estimates for the mass ratio, inclination, fill-out factor, temperature ratio and spot parameters.
Final fits used emcee, an MCMC method that samples many possible parameter combinations, with Gaussian priors. More than 2,000 iterations were run for each target, and convergence required the iteration count to exceed ten times the integrated autocorrelation time. The final model set any third-light contribution to zero and assigned the spot to star 1.
Activity offered no simple answer
Spectral analysis added a second line of evidence, but it did not produce a simple match. After subtraction of matched inactive reference spectra, H-alpha excess was detected in six systems: J024047, J040106, J003357, J004106, J085750 and J091721. It was not detected in five: J011323, DK Ari, J025641, J232802 and J002747.
Because the photometry and spectroscopy were not taken at the same time, the comparison could not establish a direct link between chromospheric activity and the O'Connell effect. The study therefore leaves the relationship open: its TESS analysis points to spot migration as a consistent interpretation, while its spectral result shows that chromospheric excess was detectable in only a subset of the sample.
The orbital clock moved both ways
Eclipse-timing analysis indicated long-term period increases in six systems and decreases in five, with J085750 showing the weakest decreasing trend. For that system, the reported fitted rate was negative 0.35 plus or minus 0.11 times 10 to the minus seventh days per year.
These period shifts were used as diagnostics for combinations of mass transfer and angular-momentum loss, rather than as a single explanation for every system. That framing matters because the study's evolutionary conclusions depend on which transfer and loss processes are included in the model.
Extreme systems with estimated dimensions
Photometric solutions put all 11 targets in the paper's extremely low mass ratio contact-binary category, with binary mass ratios below 0.15. Two systems, J011323 and J024047, were medium-contact; nine were deep-contact and none was shallow-contact. Using nominal corrected component temperatures, the authors classified five as W-subtype and the others as A-subtype.
The component masses and radii should be read as estimates from the photometric analysis, not as dynamical measurements. The formal errors listed for them also omit additional systematic uncertainties.
Stable by one test, difficult to evolve
On the paper's adopted dynamical test, all 11 systems remained below the Darwin-instability threshold: the ratio of spin angular momentum to orbital angular momentum was less than one third. J040106 came closest, at 0.26 plus or minus 0.07, and the systems were judged dynamically stable within the quoted uncertainties.
The formation picture was less settled. Integrations found that saturated magnetic braking plus gravitational radiation alone was insufficient to reach the observed angular-momentum level before the accretor reached TAMS, the end point of its modeled main-sequence track. Even after the authors imposed residual magnetic braking at 20 percent for stars more massive than 1.35 solar masses and allowed non-conservative mass transfer, the tracks still failed to reproduce the shortest-period targets, including J040106.
The authors infer that additional angular-momentum loss may be needed within the adopted framework, while stressing that this conclusion depends on assumptions about mass transfer, energy transfer and radius evolution.
What the preprint leaves open
The manuscript is an arXiv preprint, arXiv:2608.28066v1, dated 28 August 2026 and marked as submitted to ApJ. Its conclusions remain tied to this selected set of 11 systems and to the observations and model prescriptions used in the analysis.
Paper data and sources
Original title: Extremely Low Mass Ratio Contact Binaries. III. Photometric and Spectroscopic Investigations of Eleven Systems
Authors: Youmo Lai, Kai Li, Raul Michel Murillo et al.
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-28
DOI: Not available
Original paper · Full text