Title: Quantum yield and charge diffusion in the Nancy Grace Roman Space Telescope infrared detectors

Abstract: The shear signal required for weak lensing analyses is small, so any detector-level effects which distort astronomical images can contaminate the inferred shear. The Nancy Grace Roman Space Telescope (Roman) will fly a focal plane with 18 Teledyne H4RG-10 near infrared (IR) detector arrays; these have never been used for weak lensing and they present unique instrument calibration challenges. A pair of previous investigations (Hirata & Choi 2020; Choi & Hirata 2020) demonstrated that spatiotemporal correlations of flat fields can effectively separate the brighter-fatter effect (BFE) and interpixel capacitance (IPC). Later work (Freudenburg et al. 2020) introduced a Fourier-space treatment of these correlations which allowed the authors to expand to higher orders in BFE, IPC, and classical nonlinearity (CNL). This work expands the previous formalism to include quantum yield and charge diffusion. We test the updated formalism on simulations and show that we can recover input characterization values. We then apply the formalism to visible and IR flat field data from three Roman flight candidate detectors. We fit a 2D Gaussian model to the charge diffusion at 0.5 $\mu$m wavelength, and find variances of $C_{11} = 0.1066\pm 0.0011$ pix$^2$ in the horizontal direction, $C_{22} = 0.1136\pm 0.0012$ pix$^2$ in the vertical direction, and a covariance of $C_{12} = 0.0001\pm 0.0007$ pix$^2$ (stat) for SCA 20829. Last, we convert the asymmetry of the charge diffusion into an equivalent shear signal, and find a contamination of the shear correlation function to be $\xi_+ \sim 10^{-6}$ for each detector. This exceeds Roman's allotted error budget for the measurement by a factor of $\mathcal{O}(10)$ in power (amplitude squared) but can likely be mitigated through standard methods for fitting the point spread function (PSF) since charge diffusion can be treated as a contribution to the PSF.