# Catalog of Candidates for Quasars at 3 < z < 5.5 Selected among X-Ray Sources from the 3XMM-DR4 Survey of the XMM-Newton Observatory

## TL;DR

This paper presents a catalog of 903 high-redshift quasar candidates selected from X-ray sources in the 3XMM-DR4 survey, using multi-wavelength photometry and spectral energy distribution fitting to identify and estimate their redshifts.

## Contribution

The study introduces a new method combining X-ray data with SDSS, 2MASS, and WISE photometry to efficiently select high-redshift quasar candidates with high completeness.

## Key findings

- Catalog of 903 quasar candidates at 3<z<5.5 compiled.
- Selection completeness for known quasars at z>3 is about 80%.
- Sample exceeds SDSS spectroscopic quasar counts by 50% at similar redshifts.

## Abstract

We have compiled a catalog of 903 candidates for type 1 quasars at redshifts 3<z<5.5 selected among the X-ray sources of the serendipitous XMM-Newton survey presented in the 3XMM-DR4 catalog (the median X-ray flux is 5x10^{-15} erg/s/cm^2 the 0.5-2 keV energy band) and located at high Galactic latitudes >20 deg in Sloan Digital Sky Survey (SDSS) fields with a total area of about 300 deg^2. Photometric SDSS data as well infrared 2MASS and WISE data were used to select the objects. We selected the point sources from the photometric SDSS catalog with a magnitude error Delta z<0.2 and a color i-z<0.6 (to first eliminate the M-type stars). For the selected sources, we have calculated the dependences chi^2(z) for various spectral templates from the library that we compiled for these purposes using the EAZY software. Based on these data, we have rejected the objects whose spectral energy distributions are better described by the templates of stars at z=0 and obtained a sample of quasars with photometric redshift estimates 2.75<zphot<5.5. The selection completeness of known quasars at z>3 in the investigated fields is shown to be about 80%. The normalized median absolute deviation is 0.07, while the outlier fraction is eta= 9. The number of objects per unit area in our sample exceeds the number of quasars in the spectroscopic SDSS sample at the same redshifts approximately by a factor of 1.5. The subsequent spectroscopic testing of the redshifts of our selected candidates for quasars at 3<z<5.5 will allow the purity of this sample to be estimated more accurately.