Spin-down of the accreting magnetar candidate 4U 0114+65: possible first evidence for a strong coupling regime

Abstract: 4U~0114+65 is a high-mass X-ray binary composed of the B1\,Ia supergiant V*~V662~Cas and one of the slowest known accreting neutron stars, with a spin period of $\sim$9.4 ks. In 2025, its long X-ray pulsations became undetectable in \textit{Swift}/BAT monitoring, motivating a Director's Discretionary Time observation with \textit{XMM-Newton}. We compare this observation with a 2015 \textit{XMM-Newton} observation, when the source was brighter and clearly pulsed, and analyze the long-term spin evolution using \textit{Swift}/BAT data. We performed average and pulse-phase-resolved spectroscopy using the same model as in previous work. The 2025 observation still reveals weak pulsations, with a period of about 9.3 ks, despite their non-detection in \textit{Swift}/BAT. The overall spectral shape remains similar in both epochs, but the luminosity decreased by about one order of magnitude, mainly due to strong suppression of the bulk-motion Comptonization component. Although the absorbing column is higher in 2025, the inferred wind properties remain broadly compatible with those from 2015, suggesting that no major global change in the donor wind is required. Instead, the results point to a substantial reduction in accretion efficiency close to the neutron-star magnetosphere. We propose that 4U~0114+65 may be evolving toward partial centrifugal inhibition in the strong-coupling regime, where the toroidal magnetic-field component is comparable to the poloidal one. If confirmed, this would represent the first observational evidence of this state. Accretion would become progressively less efficient and more intermittent without reaching a fully developed propeller regime. The apparent disappearance of the pulse in long-term hard X-ray monitoring would then result from the lower luminosity and reduced absolute pulsed flux, rather than from the loss of the underlying spin modulation.