Compton reflection and iron fluorescence in BeppoSAX observations of Seyfert type 1 galaxies

¹ Dipartimento di Fisica, Università degli Studi “Roma Tre", Via della Vasca Navale 84, 00146 Roma, Italy
² Istituto di Astrofisica Spaziale e Fisica Cosmica, CNR, Via Gobetti 101, 40129 Bologna, Italy
³ Osservatorio Astronomico di Roma, Via dell'Osservatorio, 00044 Monteporzio Catone, Italy
⁴ XMM-Newton SOC, VILSPA–ESA, Apartado 50727, 28080 Madrid, Spain
⁵ Osservatorio Astronomico di Brera, Via Brera 28, 20121 Milano, Italy
⁶ Laboratoire d'Astrophysique de l'Observatoire de Grenoble, BP 53X, 38041 Grenoble Cedex, France
⁷ Istituto di Astrofisica Spaziale e Fisica Cosmica, CNR, Via Fosso del Cavaliere, 00133 Roma, Italy

Corresponding author: G. C. Perola, perola@fis.uniroma3.it

Received: 18 September 2001
Accepted: 25 April 2002

Abstract

A sample of nine bright Seyfert 1 and NELG type galaxies, observed with BeppoSAX, is analyzed to assess on a truly broad band basis (0.1–200 keV) the issue of the spectral contributions of Compton reflection and iron line fluorescence from circumnuclear gas. The empirical description adopted for the direct continuum is the commonly used power law with an exponential cut–off. The most direct test of the theoretical predictions, namely that the equivalent width of the line, , and the strength R of the reflection relative to the direct continuum are closely related to each other, gives a substantially positive result, that is their mean ratio is very close to expectation, and only a modest spread in the iron abundance seems implied. The existence of a steep correlation between R and the slope Γ of the power law is not confirmed. A weak evidence is found that the existence of a very shallow trend to increase on average with Γ cannot be altogether excluded in both R and , but needs to be tested with a larger sample. The energy E_f in the exponential cut–off spans a range from about 80 to more than 300 keV. A possible correlation is found, with E_f increasing on average with Γ: if ignored, for instance by keeping E_f at a fixed value in a sample study, it could be cause of artificial steepening in a correlation between R and Γ.