1: \documentclass[12pt,preprint]{emulateapj}

2: 

3: %% manuscript produces a one-column, double-spaced document:

4: \usepackage{amsmath}

5: %\usepackage{emulateapj5}

6: % \documentclass[manuscript]{aastex}

7: 

8: %% preprint2 produces a double-column, single-spaced document:

9: 

10: %\documentclass[preprint2]{book}

11: 

12: \slugcomment{Accepted by the Astrophysical Journal}

13: 

14: \shorttitle{Temperature and abundance profiles in hot Galaxy Clusters}

15: \shortauthors{Baldi et al.}

16: 

17: %% This is the end of the preamble.  Indicate the beginning of the

18: %% paper itself with \begin{document}.

19: 

20: \begin{document}

21: 

22: \title{A Chandra archival study of the temperature and metal abundance profiles

23: in hot Galaxy Clusters at $0.1\la z\la0.3$}

24: 

25: %% Use \author, \affil, and the \and command to format

26: %% author and affiliation information.

27: %% Note that \email has replaced the old \authoremail command

28: %% from AASTeX v4.0. You can use \email to mark an email address

29: %% anywhere in the paper, not just in the front matter.

30: %% As in the title, you can use \\ to force line breaks.

31: 

32: \author{A. Baldi}

33: \affil{Harvard-Smithsonian Center for Astrophysics}

34: %\email{@@@}

35: 

36: \author{S. Ettori}

37: \affil{INAF - Osservatorio Astronomico di Bologna}

38: %\email{@@@}

39: 

40: \author{P. Mazzotta\altaffilmark{1}}

41: \affil{Universit\`a di Roma "Tor Vergata", Dip. di Fisica}

42: %\email{@@@}

43: 

44: \author{P. Tozzi\altaffilmark{2}}

45: \affil{INAF - Osservatorio Astronomico di Trieste}

46: %\email{@@@}

47: 

48: \author{S. Borgani\altaffilmark{2,3}}%\altaffilmark{3}}

49: \affil{Dipartimento di Astronomia dell'Universit\`a di Trieste}

50: %\email{@@@}

51: 

52: %\author{@@@{2}}

53: %\affil{@@@}

54: %\email{@@@}

55: 

56: %\altaffiltext{1}{Universita' di Roma "Tor Vergata", Dip. di Fisica}

57: \altaffiltext{1}{Harvard-Smithsonian Center for Astrophysics}

58: \altaffiltext{2}{INFN - Sezione di Trieste}

59: \altaffiltext{3}{INAF - Osservatorio Astronomico di Trieste}

60: 

61: \begin{abstract}

62: We present the analysis of the temperature and metallicity profiles

63: of 12 galaxy clusters in the redshift range 0.1--0.3 selected from

64: the Chandra archive with at least $\sim20,000$ net ACIS counts and

65: $kT>6$ keV. We divide the sample between 7 Cooling-Core (CC) and 5 Non-Cooling-Core

66: (NCC) clusters according to their central cooling time. 

67: We find that single power-laws can describe properly both the 

68: temperature and metallicity profiles at radii larger than $0.1 r_{180}$ in 

69: both CC and NCC systems, showing the NCC objects steeper 

70: profiles outwards.

71: A significant deviation is only present in the inner $0.1 r_{180}$.

72: We perform a comparison of our sample with the De Grandi \& Molendi

73: BeppoSAX sample of local CC and NCC clusters, finding a complete agreement in the

74: CC cluster profile and a marginally higher value (at $\sim1\sigma$) in the

75: inner regions of the NCC clusters.

76: The slope of the power-law describing $kT(r)$ within $0.1 r_{180}$

77: correlates strongly with the ratio between the cooling time and the age

78: of the Universe at the cluster redshift, being the slope $>0$ and

79: $\tau_c/\tau_{age}\la0.6$ in CC systems.

80: \end{abstract}

81: 

82: %% Keywords should appear after the \end{abstract} command. The uncommented

83: %% example has been keyed in ApJ style. See the instructions to authors

84: %% for the journal to which you are submitting your paper to determine

85: %% what keyword punctuation is appropriate.

86: 

87: \keywords{ galaxies: clusters: general --- (galaxies:) intergalactic medium ---

88: X-rays: galaxies: clusters}

89: 

90: \section{Introduction}

91: Clusters of galaxies represent unique signposts in the Universe, where

92: the physical properties of the cosmic diffuse baryons can be

93: studied in great details and used to trace the past history of cosmic

94: structure formation (e.g. Rosati et al. 2002; Voit 2005, for reviews).

95: As a result of adiabatic compression and shocks generated by

96: supersonic motion during shell crossing and virialization, a hot thin

97: gas permeating the cluster gravitational potential well is

98: formed. Typically this gas, which is enriched with metals ejected form

99: supernovae (SNe) explosions through subsequent episodes of star

100: formation (e.g. Matteucci \& Vettolani 1988; Renzini 1997), reaches

101: temperatures of several $10^7$ K and therefore emits mainly via

102: thermal bremsstrahlung in the X-rays. At such temperatures most of the

103: elements are either fully ionized or in a high ionization

104: state.\\

105: Particularly evident in X--ray spectra of galaxy clusters are the

106: strong transitions to the $n=1$ level (K--shell) of the H--like and

107: He--like ions of Iron in the energy range 6.7--6.9 keV.  Below 2 keV,

108: $n=2$ level (L--shell) transition of Iron and $\alpha$ elements can be

109: detected, especially in the low temperature region in the centers of

110: the so--called cool core clusters, that are characterized by a

111: strong peak in the surface brightness distribution, and therefore

112: short cooling times.

113: Spatially resolved CC clusters show a peak in the metal distribution

114: associated to the low temperature core region

115: (e.g. De Grandi \& Molendi 2001 and 2002, hereafter DM01 and DM02).

116: Although the amount of energy supplied to the

117: intra-cluster medium (ICM) by SNe explosions depends on several

118: factors (e.g. the physical condition of the ICM at the epoch of the

119: enrichment) and cannot be obtained directly from X--ray observations,

120: the radial distribution of metals, as well as their abundance as a

121: function of time, are crucial information to shed light on the cosmic

122: star formation history and to trace the effect of SN feedback on the ICM.\\

123: Several analyses have been presented in the literature with the aim to

124: study the radial distribution of metals in clusters of galaxies.

125: Finoguenov, David \& Ponman (2000) performed a spatially resolved

126: X-ray spectroscopic analysis of 11 relaxed clusters observed by ROSAT

127: and ASCA, deriving a radial distribution of single heavy elements such

128: as Fe, Si, Ne and S.  They found that the total Fe abundance decreases

129: significantly with radius in all clusters, while the Si, Ne and S

130: abundances are either flat or decrease less rapidly. DM01 derived

131: radial metallicity profiles (mainly driven by Fe) of 17 nearby

132: clusters observed by BeppoSAX. They found a strong enhancement in the

133: abundance in the central regions of the CC clusters. A flatter

134: metallicity profile was observed instead for the non-cool core 

135: clusters in their sample. Since all the NCC clusters show signs of

136: recent merger activity, they suggested that the merger events may have

137: redistributed efficiently the metal content of the intracluster

138: medium.  Irwin \& Bregman (2001) derived iron-abundance profiles for

139: 12 clusters with $0.03\le z\le0.2$ observed by Beppo SAX. Although

140: they investigated the differences between CC and NCC clusters in a

141: less systematic way than DM01, they found a negative gradient in the

142: abundance profiles of all the CC clusters and to a lesser significance

143: also in the NCC clusters. Similarly to DM01, they found that CC

144: clusters have higher metallicity than NCC clusters at every radius. It

145: is worth to say that the aforementioned papers investigated the

146: metallicity trends only within $r_{500}$.

147: Spatially resolved measures of the metal abundance in galaxy clusters

148: were performed also with XMM-Newton. In particular Tamura et al. (2004)

149: analyzed a sample of 19 X-ray bright relaxed clusters, obtaining elemental

150: abundances of Fe, Si, S and O. They found that while the distribution of

151: Fe, Si and S is generally peaked toward the center, the O abundances are

152: uniform throughout the cluster, pointing out to a different origin among

153: these metals, most likely in SNe Ia and II.

154: More recently Vikhlinin et

155: al. (2005) have derived temperature and metallicity profiles for 11

156: low-redshift clusters observed by Chandra. The clusters in their

157: sample however are all CC clusters, presenting a very regular overall

158: X-ray morphology and showing only very weak signs of dynamical

159: activity. Although they have not analyzed the metallicity profiles

160: rescaled to the virial radius of the cluster (as they did for the

161: temperature profiles) a negative gradient of $Z$ is

162: present in all the objects in their sample.

163: However, almost all the spatially-resolved metallicity profiles are measured

164: in local clusters ($z<0.1$).

165: 

166: On the other hand, measurements of the metal content of the intracluster medium

167: at high$-z$ has been obtained with single emission-weighted estimates from Chandra

168: and XMM-Newton exposures of 56 clusters at $0.3\la z\la 1.3$ in Balestra et al. (2007).

169: They measured the Iron abundance within (0.15-0.3)$R_{vir}$ and found

170: a negative evolution of $Z_{Fe}$ with the redsfhift, with clusters at $z\ga 0.5$ having

171: a constant average Fe abundance of $\approx0.25Z_\odot$, while objects

172: in the redshift range $0.3\la z \la0.5$ show $Z_{Fe}$ significantly higher

173: ($\approx0.4Z_\odot$). Such evolution is not driven entirely by the presence 

174: of the cool cores.

175: This result has been recently confirmed by Maughan et al. (2007).

176: 

177: In this paper, we present measurements of the radial temperature and

178: metallicity profiles of a sample of 12 clusters with temperatures larger than

179: 6 keV observed with Chandra at intermediate redshift, $0.11\le z\le0.32$.

180: We take advantage of the ACIS superior spatial and spectral resolution to

181: investigate in a systematic fashion the differences that may exist

182: between CC and NCC clusters.

183: The spectroscopic measurements of the ICM temperature and metallicity

184: allow to characterize statistically the radial profiles

185: and to quantify their gradients in this unexplored

186: redshift region.

187: 

188: All the uncertainties are quoted at 1$\sigma$ (68\%) for one

189: interesting parameter. The abundance estimates are relative to the

190: compilation of cosmic values given in Anders \& Grevesse (1989)

191: (hereafter AG89), unless otherwise stated.  Indeed, these values for

192: the solar metallicities have more recently been superseeded by the new

193: values by Grevesse \& Sauval (1998) and Asplund et al. (2005)

194: (hereafter A05), who introduced a 0.676 and 0.60 times lower Iron

195: solar abundance, respectively (photospheric value), while the other

196: elements do not change significantly.  Our measures of metallicity are

197: expected to be driven mainly by Iron, however, for clarity, we also

198: performed the fits using solar abundances by A05.

199: \\

200: Throughout this paper we assume $H_0=100\:h$ km s$^{-1}$ Mpc$^{-1}$,

201: $h=0.7$, $\Omega_m=0.3$ and $\Omega_\Lambda=0.7$.

202: 

203: 

204: \section{Sample Definition and Data Analysis} \label{dataprep} From

205: Chandra archival data we select a sample of twelve 'intermediate'

206: redshift clusters ($0.11\le z\le0.32$). We also require the clusters

207: to have at least $\sim20,000$ ACIS-S or ACIS-I counts in order to

208: study their properties in at least 3 circular annuli.  The sample is

209: presented in Table~\ref{sample}, where the name of the cluster and the

210: Chandra observing logs are listed.\\

211: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

212: \begin{deluxetable*}{lccccccc}

213: \tabletypesize{\footnotesize}

214: \tablecaption{Observation Log for the Chandra Cluster Archive Sample. 

215: Column (2) shows the redshift $z$ of the clusters. The columns

216: (3), (4) and (5) show the instrument used, the observation date and

217: the observation ID, respectively. The columns (6) and (7) show

218: the observing time before ($t_{exp}$) and after ($t_{clean}$)

219: the removal of high background intervals. Column (9) is the Galactic

220: column density $N_H$ in the line of sight of the observation.

221: \label{sample}}

222: \tablewidth{0pt}

223: \tablehead{

224: \colhead{Name} &

225: \colhead{$z$} &

226: %\colhead{$\alpha$ (J2000)} &

227: %\colhead{$\delta$ (J2000)} &

228: \colhead{Instrument} &

229: \colhead{Obs. Date} &

230: \colhead{Obs. ID} &

231: \colhead{\begin{tabular}{c}

232: $t_{exp}$\\

233: (ksec)

234: \end{tabular}} &

235: \colhead{\begin{tabular}{c}

236: $t_{clean}$\\

237: (ksec)

238: \end{tabular}} &

239: \colhead{\begin{tabular}{c}

240: $N_H$\\

241: (10$^{20}$ cm$^{-2}$)

242: \end{tabular}}

243: }

244: \startdata

245: A2034    & 0.113 & ACIS-I & 2001 May 05 & 2204 & 53.9 & 53.9 & 1.6 \\

246: A1413    & 0.143 & ACIS-I & 2001 May 16 & 1661 &  9.7 &  9.7 & 2.2 \\

247:          &       & ACIS-I & 2004 Mar 06 & 5003 & 76.1 & 75.0 &     \\

248:          &       & ACIS-I & 2005 Feb 03 & 5002 & 37.2 & 36.5 &     \\

249: A907     & 0.153 & ACIS-I & 2000 Jun 29 &  535 & 11.0 & 10.9 & 5.4 \\

250:          &       & ACIS-I & 2002 Jun 14 & 3185 & 48.7 & 47.9 &     \\

251:          &       & ACIS-I & 2002 Oct 30 & 3205 & 47.7 & 40.5 &     \\

252: A2104    & 0.155 & ACIS-S & 2000 May 25 &  895 & 49.8 & 48.9 & 8.7 \\

253: A1914    & 0.171 & ACIS-I & 2003 Sep 03 & 3593 & 18.9 & 18.8 & 0.9 \\

254: A2218    & 0.176 & ACIS-I & 2001 Aug 30 & 1666 & 49.2 & 20.2 & 3.2 \\

255: A963     & 0.206 & ACIS-S & 2000 Oct 11 &  903 & 36.8 & 35.8 & 1.4 \\

256: A2261    & 0.224 & ACIS-I & 2004 Jan 14 & 5007 & 24.6 & 24.3 & 3.3 \\

257: A2390    & 0.228 & ACIS-S & 2000 Oct 08 &  500 &  9.8 &  9.8 & 6.8 \\

258:          &       & ACIS-S & 2003 Sep 11 & 4193 & 96.3 & 91.0 &     \\

259: A1835    & 0.253 & ACIS-S & 2000 Apr 29 &  496 & 10.8 & 10.3 & 2.3 \\

260: ZwCl3146 & 0.291 & ACIS-I & 2000 May 10 &  909 & 46.6 & 45.6 & 3.0 \\

261: A1995    & 0.319 & ACIS-S & 2000 May 08 &  906 & 57.5 & 53.8 & 1.4 \\

262: \enddata

263: 

264: \end{deluxetable*}

265: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

266: The Chandra data analysis has been performed using the latest version

267: of CIAO (v3.3.0.1).  All of our datasets are processed by a version of

268: the Standard Data Processing (SDP) pipeline prior to version DS 7.4.0,

269: which uses the tool $acis\_detect\_afterglow$ to flag possible cosmic

270: ray events in the level 1 event file; it has been determined that a

271: significant fraction of the X-ray events from a source in imaging mode

272: might be removed using this tool.  Therefore we reset the correction

273: performed by $acis\_detect\_afterglow$ on the Level=1 event file, so

274: that the hot pixels and the afterglow events may be properly removed

275: by the improved CIAO tool $acis\_run\_hotpix$, (introduced after SDP

276: version DS 7.4.0). A new Level=1 event file is then created (through

277: the CIAO tool $acis\_process\_events$) to apply the latest calibration

278: files to the data (e.g.  apply the newest ACIS gain maps, apply the

279: time-dependent ACIS gain correction, apply the ACIS Charge Transfer

280: Inefficiency correction, etc.). Moreover in the case of observations

281: telemetered in VFAINT mode it is possible to reduce the background

282: using the additional screening of the events with significantly

283: positive pixels at the border of the $5\times5$ event island.  Two

284: further filtering steps are then required to obtain the Level=2 event

285: files, i.e. filter for bad grades (using ASCA grades) and for a

286: "clean" status column and apply the Good Time Intervals (GTIs)

287: supplied by the pipeline.  The final step is to examine background

288: light curves during each observation to detect and remove the periods

289: of high background, due to flaring episodes. We perform the flare

290: detection and removal following the recommendations suggested in

291: Markevitch et al. (2003); both the total and the clean exposure times

292: are listed in Table~\ref{sample}. Most of the observations are

293: slightly affected by background flares thus we were able to use

294: practically all the exposure time. The only exceptions are ObsID 3205,

295: 4193, 906 and especially ObsID 1666, where $\sim29$ ks of the exposure

296: were lost due to high background.\\

297: 

298: \subsection{Background Subtraction}

299: An accurate subtraction of the background is crucial to perform a

300: correct study of the spectral properties of the clusters in our

301: sample, especially in their outskirts. Since we are dealing with

302: extended objects, occupying most of the ACIS field of view we need to

303: use a compilation of the blank-field observations, processed

304: identically to the cluster observation (i.e. as described above) and

305: reprojected onto the sky using the aspect information from the cluster

306: pointing. It is worth noticing that the synthetic backgrounds

307: correspond to longer exposure times ($\sim0.5$ Msec) than any of our

308: observations, giving us a very good sampling in the estimate of the

309: background to subtract.  Moreover, in order to 'tailor' the background

310: to our data we follow the recommendations given in the CIAO

311: web-pages\footnote{{\tiny http://cxc.harvard.edu/cal/Acis/Cal\_prods/bkgrnd/acisbg/COOKBOOK}}.

312: In particular we renormalize the blank-fields to the background in

313: each observation, considering a region of the ACIS field of view

314: practically free from cluster emission (mainly ACIS-S1 for ACIS-S

315: observations, and ACIS-S2 for ACIS-I observations) and a spectral band

316: (9.5-12 keV) where the Chandra effective area is nearly zero,

317: therefore all the observed flux is due to the particle background.

318: 

319: In addition to the particle-induced background we check also if the

320: diffuse soft X-ray background could be an important factor

321: in our observations 

322: and if appropriate adjustments are needed. For each observation, 

323: we follow the procedure of Vikhlinin et al. (2005), extracting a 

324: spectra in the source-free regions of the detector, subtracting the

325: renormalized blank-field background and fitting the residuals in

326: XSPEC v11.3.2p (in the 0.4-1 keV band) with an unabsorbed {\tt mekal}

327: model, whose normalization was allowed to be negative.

328: The best-fit model obtained is therefore included as an additional 

329: component in the spectral fits (with its normalization scaled by the area).

330: However, in every observation the adjustments required are minimal 

331: and do not affect significantly the determination of $kT$ and $Z$, even at large radii. 

332: This is also due to the properties of the clusters in our sample, 

333: whose high $kT$ values ($>4$ keV) even in the resolved outer regions are not

334: affected significantly from the method applied for the subtraction of the diffuse

335: soft background.

336: 

337: \subsection{Cash statistics vs. $\chi^2$ statistics}

338: The $\chi^2$ statistics require grouping of the spectra, having at least

339: 20 counts per bin, in order to be able to approximate the Poissonian

340: distribution of counts with a Gaussian. On the contrary, Cash

341: statistics do not require any grouping and represent a more reliable

342: (and less biased) approach to fit the data.  Indeed, it is well known

343: from the literature (see e.g. Nousek \& Shue 1989; Balestra et

344: al. 2007)

345: that the $\chi^2$ statistics systematically ``sees'' the observed

346: spectra softer than the real ones.  This usually leads to an

347: overestimate of the slope of the observed spectra in the case of a

348: simple power-law fit, while in the case that a thermal model is fitted

349: to the data, the temperature measured is usually underestimated.

350: As a test to see whether this systematics is present also in our data,

351: we have decided to apply both these fit statistics.

352: We find that for all the clusters of the sample, a systematically

353: lower temperature is measured with the $\chi^2$ (on average $7\%-17\%$

354: lower, depending on the cluster). On the other hand no obvious

355: systematic trend is observed in the determination of $Z$, being the

356: variation in the best--fitting value of the metallicity in each

357: cluster ranging between $\Delta Z\sim0.01$ and $\sim0.07$

358: with no preferential direction.

359: To avoid the dependence on the grouping method, and the bias in the

360: best-fit temperature, we decide to use the modified Cash statistics, 

361: as implemented in XSPEC v11.3.2p,

362: to determine the best-fit parameters and their uncertainties.

363: 

364: \subsection{Spectral Analysis}\label{specanal}

365: In order to study the radial properties of the cluster emission, we

366: subdivide each cluster in annuli (circular or elliptical, depending on

367: the morphology of the cluster) centered on the X-ray emission peak. In

368: the more disturbed clusters, where an emission peak is not clearly

369: identifiable, we assume the center of the cluster to correspond with

370: the X-ray centroid at $0.5r_{500}$.  We require each region to have at

371: least $\sim7,000$ net counts, so that it would be possible to estimate

372: the temperature and the metallicity of the annulus with sufficient

373: accuracy.

374: For each cluster the outermost annulus corresponds to an area where

375: the intensity of the source in the 0.8-8 keV band is roughly equal to

376: that of the background.  We extract a spectra from each annulus after

377: excluding the 3$\sigma$ point sources detected by the CIAO tool

378: $wavdetect$. The source list produced is also inspected 'by-eye' in

379: order to remove possible additional sources not detected by

380: $wavdetect$ (especially in the regions where the diffuse emission from

381: the cluster is brighter).  The CIAO script used to perform the

382: spectral extraction is

383: $specextract$, which generates source and background spectra and build

384: the appropriate RMFs and ARFs. The background

385: is taken from the re-normalized blank field observations using the same region of the source.\\

386: The spectra are analyzed with XSPEC v11.3.2p (Arnaud et al. 1996) and

387: fitted by a single-temperature {\tt mekal} model (Kaastra 1992;

388: Liedahl et al. 1995) in which the ratio between the elements is fixed

389: to the solar value as in AG89. However, as explained in \S1, these

390: values for the solar metallicities have more recently been superseeded

391: by the new values by Grevesse \& Sauval (1998) and A05. For clarity

392: and completeness, we also performed the fits using solar abundances by

393: A05.  The free parameters in the model are the temperature $kT$, the

394: metallicity $Z$ of the gas and the normalization.  The spectral band

395: considered in the fit is the 0.6-8 keV. We choose not to consider the

396: data below 0.6 keV because of uncertainties in the ACIS calibration

397: below that energy.

398: The $N_H$ derived from the X-rays is

399: found to be

400: consistent (within 1$\sigma$) with the Galactic value in the line of

401: sight of each observation, as derived from radio data (Stark et

402: al. 1992), except in the cases of A2104 and A2390 (see

403: Table~\ref{sample} and following \S~\ref{text2104}). In these

404: clusters the $N_H$ value measured from X-ray data is significantly

405: different (at more than 2$\sigma$ confidence level) from the radio

406: value, therefore we adopt the X-ray value. The $N_H$ value is fixed to

407: the Galactic value obtained from the radio data (and listed

408: in Table~\ref{sample}) in the rest of the sample.

409: We have measured $N_H$ from the X-ray data in each annulus, finding no

410: evidence of radial variation. Therefore, $N_H$ is fixed to the same value

411: in all the radial annuli. 

412: 

413: We have divided then our sample in Cooling-Core and

414: Non-Cooling-Core clusters according to their central cooling time.

415: The gas temperature and density profiles are recovered from the single-phase 

416: spectral fit done in annular rings by correcting the emissivity in each shell 

417: by the contrubution of the outer shells moving inwards. A detailed description 

418: of the procedure is presented in Ettori et al. (2002). In brief, the 

419: normalization of the thermal component, being proportional to the Emission 

420: Integral, provides the gas density, whereas the deprojected temperature is 

421: provided by weighting for the corrected emissivity the spectral measurement.  

422: The deprojected values in the innermost bin are then used to estimate the 

423: central cooling times $\tau_c =

424: 5/2 (\mu_e/\mu) T_e (n_e / \epsilon)$, where $\mu=0.613$ and

425: $\mu_e=1.174$ are appropriate for a plasma with a metallicity of 0.3

426: times the solar values in AG89, $T_e$, $n_e$ and $\epsilon$

427: are the gas temperature, electron density and emissivity in the

428: innermost bin, respectively.\\

429: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

430: \begin{deluxetable*}{lccccccc}

431: \tabletypesize{\small}

432: \tablecaption{Global cluster properties of the Chandra sample. 

433: Column (2) shows the the total net counts from the inner to the outer annulus 

434: considered in the spectral analysis. The columns (3), (4) and (5) show the 

435: global temperature $\langle kT \rangle$, the global metallicity 

436: $\langle Z \rangle$ and the virial radius $r_{180}$, computed within

437: $0.07-0.4r_{180}$, respectively. Column (6) shows the aperture used to measure 

438: $\langle kT \rangle$ and $\langle Z \rangle$.

439: The columns (7) and (8) show the central cooling times and the ratio with 

440: respect to the age of the Universe at the cluster redshift.

441: \label{xrayprop}}

442: \tablewidth{0pt}

443: \tablehead{

444: \colhead{Name} &

445: \colhead{\begin{tabular}{c}

446: Net cts\\

447: (0.6-8 keV)

448: \end{tabular}} &

449: \colhead{\begin{tabular}{c}

450: $\langle kT \rangle$\\

451: (keV)

452: \end{tabular}} &

453: \colhead{\begin{tabular}{c}

454: $\langle Z \rangle$\\

455: ($Z_\odot$)

456: \end{tabular}} &

457: \colhead{\begin{tabular}{c}

458: $r_{180}$\\

459: (kpc)

460: \end{tabular}} &

461: \colhead{\begin{tabular}{c}

462: Aperture\\

463: ($^{\prime\prime}$)

464: \end{tabular}} &

465: \colhead{$\tau_c$ (Gyr)} &

466: \colhead{$\tau_c/\tau_{age}$}

467: }

468: \startdata

469: A2034    & 78,900  & $6.36\pm0.15$  & $0.30\pm0.04$ & 2,222 & 76-433 

470: & $21.2\pm3.2$ & $1.77\pm0.27$\\

471: A1413    & 181,500 & $7.52_{-0.12}^{+0.20}$  & $0.23\pm0.03$ & 2,416 & 67-385 

472: & $4.2\pm0.3$  & $0.36\pm0.03$\\

473: A907     & 87,500  & $5.82\pm0.12$  & $0.34\pm0.04$ & 2,125 & 56-320 

474: & $2.0\pm0.1$  & $0.17\pm0.01$\\

475: A2104    & 63,100  & $6.76\pm0.19$  & $0.24\pm0.05$ & 2,290 & 60-341 

476: & $18.1\pm2.1$ & $1.58\pm0.18$\\

477: A1914    & 39,100  & $9.20_{-0.37}^{+0.39}$ & $0.27\pm0.07$ & 2,672 & 64-367 

478: & $12.2\pm1.0$ & $1.07\pm0.09$\\

479: A2218    & 18,300  & $6.25\pm0.31$  & $0.24\pm0.07$ & 2,202 & 52-295 

480: & $21.3\pm1.8$ & $1.89\pm0.16$ \\

481: A963     & 41,800  & $6.02_{-0.19}^{+0.28}$  & $0.18\pm0.06$ & 2,161 & 45-256 

482: & $6.5\pm0.4$ & $0.59\pm0.04$ \\

483: A2261    & 21,500  & $7.43_{-0.27}^{+0.49}$  & $0.30_{-0.06}^{+0.07}$ & 2,400 & 47-267

484: & $7.0\pm0.4$  & $0.65\pm0.03$ \\

485: A2390    & 202,600 & $9.35\pm0.15$ & $0.30\pm0.03$ & 2,693 & 52-295

486: & $1.3\pm0.2$  & $0.12\pm0.02$\\

487: A1835    & 23,100  & $8.06\pm0.53$  & $0.31\pm0.09$ & 2,500 & 44-254

488: & $0.9\pm0.1$  & $0.08\pm0.01$\\

489: ZwCl3146 & 40,500  & $8.59\pm0.39$  & $0.24\pm0.06$ & 2,582 & 41-237

490: & $1.0\pm0.1$  & $0.10\pm0.01$\\

491: A1995    & 30,200  & $7.59_{-0.44}^{+0.57}$  & $0.50_{-0.11}^{+0.12}$ & 2,427 & 37-209

492: & $12.7\pm1.3$ & $1.28\pm0.14$ \\

493: \enddata

494: \end{deluxetable*}

495: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

496: The central cooling times are reported in Table~\ref{xrayprop}, as well as

497: the age $\tau_{age}$ of each cluster, and the

498: ratio between the two quantities. The age of the Universe at the $z$ of

499: observation is used as an upper limit to the age of the cluster.

500: Bauer et al. (2005) computed the

501: cooling times for 6 of the clusters in our sample (A1835, A1914,

502: A2218, A2261, A2390 and ZwCl3146) finding a $\tau_c$ in the center of

503: the cluster, or at 50~kpc, consistent with the values computed for the

504: central bin in our spectral analysis (which might extend farther out

505: than 50~kpc from the center in some cases).  Following their

506: criterium, a clear separation between the CC and the NCC in our sample

507: can be located at $\tau_c\sim10$ Gyr (Figure~\ref{histotc};

508: corresponding to $\tau_c/\tau_{age}\sim1$).  We have 4 clusters

509: presenting signs of strong cooling ($\tau_c<2$ Gyr) and 3 clusters

510: exhibiting signs of mild cooling ($\tau_c<10$ Gyr). The remaining 5

511: clusters can be classified as NCC,

512: presenting longer cooling times in the center.

513: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

514: \begin{figure}

515: %\epsscale{0.90}

516: %\plottwo{histo_tc.ps}{histo_tctage.ps}

517: \plotone{f1.eps}

518: \caption{Histograms of the distribution in our sample of the values of 

519: $\tau_c / \tau_{age}$, computed as described in \S~\ref{selftemp} and quoted 

520: in Table~\ref{xrayprop}.}

521: \label{histotc}

522: \end{figure}

523: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

524: The projected temperature and metal abundance profiles for 

525: both CC and NCC objects are shown in Figures~\ref{temperatures} and

526: ~\ref{abundances}.

527: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

528: \begin{figure*}

529: \plotone{f2.eps}

530: \caption{{\it a):} Temperature profiles for the CC clusters in the sample: A907 (green empty triangles), 

531: A963 (blue empty squares), A1413 (red empty circles), 

532: A1835 (magenta filled triangles), A2261 (green stars), 

533: A2390 (cyan filled circles) and ZwCl3146 (red filled squares).

534: {\it b):} Normalized temperature profiles for the CC clusters, plotted

535: against the radii in units of $r_{180}$. The symbols have the same meaning as in panel a).

536: {\it c):} Temperature profiles for the NCC clusters in the sample: A1914 (red filled circles), 

537: A1995 (green filled triangles), A2034 (cyan empty circles), 

538: A2104 (blue filled squares) and A2218 (magenta empty triangles).

539: {\it d):} Normalized temperature profiles for the NCC clusters, plotted

540: against the radii in units of $r_{180}$. The symbols have the same meaning as in panel c).

541: }

542: \label{temperatures}

543: \end{figure*}

544: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

545: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

546: \begin{figure*}

547: \plotone{f3.eps}

548: \caption{{\it a):} Abundance profiles for the CC clusters in the sample: the different symbols correspond

549: to the clusters listed in Figure~\ref{temperatures}a.

550: {\it b):} Normalized abundance profiles for the CC clusters, plotted

551: against the radii in units of $r_{180}$. The symbols have the same meaning as in panel a).

552: {\it c):} Abundance profiles for the NCC clusters in the sample: the different symbols correspond

553: to the clusters listed in Figure~\ref{temperatures}c.

554: {\it d):} Normalized abundance profiles for the NCC clusters, plotted

555: against the radii in units of $r_{180}$. The symbols have the same meaning as in panel c)}

556: \label{abundances}

557: \end{figure*}

558: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

559: 

560: \subsection{Notes on individual clusters} \label{individual}

561: 

562: {\bf A2034}:

563: A2034 (z=0.113) has been observed with Chandra in one ACIS-I pointing 

564: (ObsID: 2204).

565: The temperature profile we derived is quite flat in the central

566: regions of the cluster, where the temperature is $kT\sim8$ keV. It

567: shows however a negative gradient after 400 kpc from the center.  A

568: similar trend is observed in the metallicity, where the average value

569: of $Z=0.4$ within 400 kpc from the center decreases to $Z<0.2$ at

570: larger radii.

571: 

572: {\bf A1413}:

573: A1413 (z=0.143) has been observed four times with ACIS-I. We discard one 

574: observation (ObsID: 537) which is affected almost entirely by a

575: persisting flare.

576: In one of the observations used in our analysis (ObsID: 5003) 

577: the source is placed in a position of the ACIS-I array very close to

578: the S2 chip, therefore S2 is still contaminated by source emission and

579: we can not use it to re-normalize the blank field to the background in

580: the observation. We use instead part of the I1 chip (which is

581: front-illuminated as S2) to re-normalize, since it is more distant

582: from the cluster center than S2 and therefore

583: less contaminated by cluster emission.

584: The resulting temperature profile shows a slight decrease in

585: temperature towards the center ($\Delta kT=1.2_{-0.6}^{+0.7}$ keV within the

586: inner 150 kpc).  A1413 has been also observed by XMM-Newton (Pratt \&

587: Arnaud 2002), representing one of the clusters with the most accurate

588: temperature profile observed by this satellite. This cluster is also

589: part of the sample of Chandra clusters analyzied by Vikhlinin et

590: al. (2005).  The XMM-Newton observation does not find any evidence of

591: a cool core, in contrast with the temperature profiles obtained with

592: Chandra both in our analysis and even more evidently in Vikhlinin et

593: al. (2005). This might be due to the poorer angular resolution of

594: XMM-Newton with respect to Chandra.\\

595: The metallicity profile is decreasing towards larger radii, and

596: consistent within 1$\sigma$ with the measures of Vikhlinin et

597: al. (2005).

598: 

599: {\bf A907}:

600: A907(z=0.153) has been observed with Chandra in three separate ACIS-I

601: pointings (ObsID: 535, 3185 and 3205), all of them used

602: in our analysis.

603: The temperature profile shows an evidence of a cool core in the center

604: of the cluster ($\Delta kT=1.4_{-0.3}^{+0.2}$ keV in the central 100 kpc). The

605: metallicity profile presents a decreasing trend toward larger radii.

606: A907 is also part of the cluster sample analyzed by Vikhlinin et

607: al. (2005). Their results, both for the temperature and the

608: metallicity, are fully consistent with ours within the 1$\sigma$

609: statistical uncertainties.

610:  

611: {\bf A2104}:\label{text2104}

612: A2104 (z=0.155) has been observed with Chandra in one ACIS-S pointing

613: (ObsID: 895).

614: As described in \S~\ref{specanal}, the value of the $N_H$ measured

615: from the X-ray data alone is significantly different from the radio

616: value, thus we have decided to fix the $N_H$ to the best fit

617: value obtained from the fit ($1.55\times10^{21}$ cm$^{-2}$).

618: The cluster does not show any evidence of a cool core in its center,

619: having a temperature profile decreasing towards the outskirts.  The

620: metallicity profile is consistent with being flat, with a value

621: $Z\sim0.3-0.4Z_\odot$, within the 1$\sigma$ uncertainties.

622: 

623: {\bf A1914}:

624: Two ACIS-I pointings of A1914 (z=0.171) are available in Chandra

625: archive. However the oldest (and shortest) observation (ObsID: 542)

626: has been performed in 1999. For observations performed in that year an

627: accurate modeling of the ACIS background is not currently

628: available. Thus to avoid problems in background subtraction we have

629: decided to discard it and keep only the longest observation

630: (ObsID: 3593).

631: A negative gradient in $kT$ is quite clear: the temperature drops from

632: $kT=12.0_{-0.6}^{+0.9}$ keV in the center down to 

633: $kT=8.5\pm0.6$ keV in the outer radial

634: bin. A similar trend is observed also in the abundance profile, where

635: $Z=0.5\pm0.1Z_\odot$ in the center, then decreasing to 

636: $0.2\pm0.1 Z_\odot$ in the

637: two outer radial bins.

638: This is one of the few examples of metallicity peak without a

639: corresponding cool core (or temperature drop) towards the center.

640: 

641: {\bf A2218}:

642: A2218 (z=0.176) has been observed three times with

643: ACIS-S. Unfortunately two of these observations (ObsID: 553 and 1454)

644: were performed in 1999 and for the reason described in the case of

645: A1914 we have decided to discard them. Moreover the remaining

646: observation (ObsID: 1666) has been strongly affected by a flare which

647: reduces the good exposure time to

648: only $\sim20$ ks.

649: With these data we are able to observe the presence of a centrally

650: peaked temperature profile (a hot, instead of a cool core) and a costant

651: metallicity profile.  A temperature profile peaked toward the center

652: has been also seen by Machacek et al. (2002), analyzing the two

653: Chandra observations performed in 1999.  This is consistent with the

654: picture of A2218 being involved in a line-of-sight merger, as

655: suggested by a considerable disturbance of the intracluster gas in the

656: X-rays and by the observed substructure in the optical (e.g. Pratt et

657: al. 2005).

658: 

659: {\bf A963}:

660: A963 (z=0.206) has been observed with Chandra in one ACIS-S pointing (ObsID: 903).

661: We found a decreasing trend of $Z$ with the radius, with only a very weak hint for the presence

662: of a lower temperature in the center. 

663: 

664: {\bf A2261}:

665: Two pointings of A2261 (z=0.224) are available in the Chandra archive. One of the observations

666: (ObsID: 550) has been performed in 1999 and therefore we discard it for the reason described 

667: above in the case of A1914.

668: The temperature profile does show only a hint (more than 2$\sigma$

669: however)

670: of a decrease in the center, where 

671: the temperature drops down from $9.0\pm0.4$ keV to $7.7\pm0.4$ keV. 

672: The metallicity profile shows a constant behaviour for the first two bins and a decrease (significant

673: at more than 1$\sigma$) in the outer radial bin.

674: 

675: {\bf A2390}:

676: A2390 (z=0.228) has been observed three times with ACIS-S. One of the observations (ObsID: 501) has

677: been performed

678: in 1999 and therefore we discard it. We concentrate our analysis on the remaining two observations

679: (ObsID: 500 and 4193), yielding a total of $\sim100$ ks of good observing time.

680: The value of the $N_H$ derived from the X-rays ($N_H=1.1\times10^{21}$

681: cm$^{-2}$) is significantly different than the radio value, therefore we adopted 

682: the X-ray value in the spectral fits.

683: A2390 is also part of the sample analyzed in Vikhlinin et al. (2005). Similarly to them we find a cool

684: core ($kT=5.8\pm0.2$ keV) in the center of the cluster with a $kT$ profile getting flatter going towards the

685: outskirts, being fully consistent with their measured temperatures at every radius. 

686: On the other hand, the metallicity profile shows just a hint of a peak in the central part of the 

687: cluster. It is however consistent at 1$\sigma$ with the profile of Vikhlinin et al. (2005) and not sensitive to

688: the choice of the $N_H$.

689: 

690: {\bf A1835}:

691: Two different ACIS-S observations of A1835 (z=0.253) are present in

692: the Chandra archive. We discard the older (and longer) observation

693: (ObsID: 495), performed in 1999 because of the reason described

694: in the case of A1914, keeping only the $\sim10$ ks observation performed in 2000 (ObsID: 496).

695: The temperature profile of A1835 shows a clear evidence of a cool core

696: in its center where $kT$ drops down by a factor of $\sim2$.  Moreover

697: the temperature shows a decline after 300 kpc, going toward larger

698: radii.  Piffaretti et al. (2005) analysed XMM-Newton observations of

699: A1835 and detected a temperature decrease at large radii, as in our

700: data. Majerowicz et al. (2002) also analysed XMM-Newton data and found

701: a decrease in the temperature profile at large radii (at $\sim400$ kpc

702: from the center); however their temperature profile becomes constant

703: after such decrease.  The decrease at large radii has not been

704: observed in the analysis of Chandra data by Voigt \& Fabian (2006),who

705: found a constant temperature outside the central 100 kpc. However it

706: is worth noticing that in their work they analysed the 1999

707: observation (instead of the 2000 observation, as in our analysis) which

708: might have background subtraction problems especially at large

709: radii. This may explain the difference between the

710: two profiles.

711: The metallicity profile shows a decreasing gradient in the first two

712: bins, becoming constant afterwards.  The only comparison with the

713: literature comes from an XMM-Newton observation analyzed by Majerowicz

714: et al.  (2002), where an almost constant metallicity profile at every

715: radius has been observed.

716: 

717: {\bf ZwCl3146}:

718: ZwCl3146 (z=0.291) has been observed with ACIS-I in one pointing (ObsID: 909).

719: Also this cluster clearly shows the presence of a cool core ($kT$

720: dropping down by a factor of almost 2). A decreasing trend in

721: metallicity from $Z=0.50\pm0.05$ in the center, down to $Z=0.17\pm0.12$ in the

722: outer bin, is also observed.

723: 

724: {\bf A1995}:

725: One ACIS-S observation of A1995, the farthest cluster in our sample

726: (z=0.319), is present in the Chandra archive (ObsID: 906). Although

727: this observation is quite long ($\sim50$ ks of good exposure

728: time) the number of counts available allowed us to divide this cluster 

729: only in three radial bins.

730: The temperature profile of A1995 is consistent to be flat within the

731: errors with a temperature around 9 keV.  The abundance profile seems

732: to have a positive gradient in the outer bin, however the errors are

733: large and this increase in $Z$ is not very significant.

734: 

735: 

736: 

737: 

738: \section{Self-similarity of radial profiles} \label{selftemp}

739: 

740: One of the main goals of this paper is to look for a (purely

741: phenomenological) self-similarity in the radial profiles of

742: temperature and metallicity, after they are scaled to the cluster

743: virial radius $r_{180}$. A measure of $r_{180}$ is thus crucial to

744: test for such self-similarity in our cluster sample. This quantity can

745: be approximated by the following relation:

746: \begin{equation}

747: r_{180}=1.95h^{-1}\:Mpc(\langle kT \rangle/10\:keV)^{1/2},

748: \end{equation}

749: as calibrated from the non-radiative hydrodynamical simulations of

750: clusters by Evrard et al. (1996). It is worth noticing that this

751: relation is in agreement with the scaling relations observed

752: (e.g. Ettori et al. 2004) in the X-rays, where the dependency on $kT$ is

753: consistent with Eq.(1) and only the absolute normalization may

754: experience some variations. To compute the global temperature

755: $\langle kT \rangle$ necessary to estimate $r_{180}$ we extract spectra

756: including emission going from 0.07$r_{180}$ to 0.4$r_{180}$ in each

757: cluster.  The central regions of each cluster are therefore excluded

758: from the spectra in order to avoid contamination from a possible cool

759: core. The values of $\langle kT \rangle$ and $r_{180}$ have been

760: evaluated iteratively until a convergence to a stable value of the

761: temperature is obtained ($\Delta kT\leq0.01$ keV between two different

762: iterations).

763: From the fits we are able to determine also a global metallicity $\langle Z \rangle$ in each cluster.

764: In Table~\ref{xrayprop} we list the best-fit values for $\langle kT \rangle$ and $\langle Z \rangle$, 

765: and the value of $r_{180}$ computed using

766: the formula above. The minimum and maximum apertures used to extract the total spectrum are listed as well.\\

767: 

768: \subsection{The temperature profiles}

769: 

770: Figure~\ref{temperatures} shows the normalized temperature profiles

771: for all the CC clusters (panel b) compared with the NCC clusters

772: (panel d). This figure has been obtained by normalizing the temperatures

773: in each cluster to its average temperature $\langle kT \rangle$

774: computed from the total cluster spectrum excluding the central

775: $0.07 r_{180}$.

776: The error-weighted mean and the best-fit results after fitting with

777: single power-laws $Y \propto r^{\mu}$ are presented in Table~\ref{bestfit}.

778: 

779: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

780: \begin{figure}

781: \plotone{f4a.eps}

782: \plotone{f4b.eps}

783: \caption{Error-weighted mean temperature profile of the CC ({\it filled circles})

784: and NCC ({\it open diamonds}) clusters in our sample at intermediate redshift (top panel).

785: (Bottom) Comparison between our results and the local measurements in DM02 (shaded regions).

786: The 1$\sigma$ errors on the means are plotted as solid lines (dark gray region for the local

787: estimates) while the scatter (rms) in each data bin is shown as a dotted line

788: (light gray region for local estimates).}

789: \label{T_CC_vs_noCC_mean}

790: \end{figure}

791: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

792: 

793: Within $0.1 r_{180}$, the temperature profiles in CC objects increases with a slope

794: $\mu = 0.25$. Moving outwards, between $0.1 r_{180}$ and the outer radial limit

795: of our spectral analysis at $\approx 0.5 r_{180}$, these profiles behave as $r^{-0.1}$.

796: Non-cooling-cores systems have, on average, a profile that is almost flat

797: at $r<0.1 r_{180}$ and then decreases rapidly as $r^{-0.3}$.

798: In the outskirts, the temperature profiles of CC and NCC clusters

799: show a significant discrepancy between their slopes, being

800: NCC more deviant from the isothermal case.

801: 

802: Our best fit functional for the CC sample is fully consistent with

803: the best fit functional form found by Vikhlinin et al. (2005) in their sample

804: of CC clusters, at $r\la0.3r_{180}$.

805: The two functionals diverge significantly only above $0.3r_{180}$, where our 

806: profile is flatter (and therefore the value of $kT/\langle kT \rangle$ 

807: is higher) than Vikhlinin et al. (2005) profile.

808: However, only a few of our data points are located beyond that radius, 

809: preventing us from any statistically significant comparison between the two

810: samples at $r\ga0.3r_{180}$.

811: 

812: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

813: \begin{deluxetable*}{lccc}

814: \tabletypesize{\small}

815: \tablecaption{Error-weighted means, with errors on the mean and r.m.s. quoted within round brackets, 

816: and best-fit parameters of the single power-laws $Y = Y_{0.1} (x/0.1)^{\mu}$, with

817: $x = r/r_{180}$. $\langle kT \rangle$ and $\langle Z \rangle$ are measured in the radial range

818: $0.07-0.4 r_{180}$. \label{bestfit} }

819: \tablewidth{0pt}

820: \tablehead{

821: \colhead{} &

822: \colhead{} &

823: \colhead{$kT/\langle kT \rangle$} &

824: \colhead{}

825: }

826: \startdata

827:     &  all $r$ &  $r < 0.1 r_{180}$ & $r > 0.1 r_{180}$ \\

828: All & $0.84\pm0.04 (0.28)$ & $0.79\pm0.03 (0.35)$ & $1.02\pm0.06 (0.10)$ \\

829:  & $Y_{0.1}=1.00\pm0.01$ & $Y_{0.1}=1.20\pm0.01$ & $Y_{0.1}=1.10\pm0.02$ \\

830:  & $\mu=0.15\pm0.01$ & $\mu=0.26\pm0.01$ & $\mu=-0.16\pm0.03$ \\

831:  & $\chi^2$/dof=$1223.4/78$ & $\chi^2$/dof=$720.0/43$ & $\chi^2$/dof=$59.1/33$ \\

832:  &&& \\

833: CC  & $0.80\pm0.04 (0.24)$ & $0.76\pm0.03 (0.26)$ & $1.02\pm0.06 (0.08)$ \\

834:  & $Y_{0.1}=1.00\pm0.01$ & $Y_{0.1}=1.16\pm0.02$ & $Y_{0.1}=1.07\pm0.02$ \\

835:  & $\mu=0.17\pm0.01$ & $\mu=0.25\pm0.01$ & $\mu=-0.10\pm0.03$ \\

836:  & $\chi^2$/dof=$744.3/52$ & $\chi^2$/dof=$496.3/30$ & $\chi^2$/dof=$26.7/20$ \\

837:  &&& \\

838: NCC & $1.13\pm0.07 (0.20)$ & $1.27\pm0.08 (0.14)$ & $1.02\pm0.07 (0.13)$ \\

839:  & $Y_{0.1}=1.12\pm0.01$ & $Y_{0.1}=1.25\pm0.04$ & $Y_{0.1}=1.23\pm0.04$ \\

840:  & $\mu=-0.15\pm0.01$ & $\mu=-0.02\pm0.03$ & $\mu=-0.32\pm0.05$ \\

841:  & $\chi^2$/dof=$76.6/24$ & $\chi^2$/dof=$28.8/11$ & $\chi^2$/dof=$17.1/11$ \\

842: % &&& \\

843: %\hline

844:  &&& \\

845:  & & $Z/\langle Z \rangle$ & \\

846: \hline

847:     &  all $r$ &  $r < 0.1 r_{180}$ & $r > 0.1 r_{180}$ \\

848: All & $1.31\pm0.29 (0.54)$ & $1.41\pm0.26 (0.60)$ & $1.11\pm0.33 (0.32)$ \\

849:  & $Y_{0.1}=1.13\pm0.04$ & $Y_{0.1}=1.03\pm0.06$ & $Y_{0.1}=1.35\pm0.10$ \\

850:  & $\mu=-0.23\pm0.02$ & $\mu=-0.28\pm0.03$ & $\mu=-0.44\pm0.15$ \\

851:  & $\chi^2$/dof=$149.2/78$ & $\chi^2$/dof=$120.7/43$ & $\chi^2$/dof=$20.0/33$ \\

852:  &&& \\

853: CC  & $1.40\pm0.28 (0.58)$ & $1.53\pm0.26 (0.62)$ & $1.12\pm0.32 (0.33)$ \\

854:  & $Y_{0.1}=1.17\pm0.05$ & $Y_{0.1}=1.11\pm0.06$ & $Y_{0.1}=1.37\pm0.11$ \\

855:  & $\mu=-0.24\pm0.03$ & $\mu=-0.27\pm0.03$ & $\mu=-0.52\pm0.18$ \\

856:  & $\chi^2$/dof=$80.2/52$ & $\chi^2$/dof=$64.9/30$ & $\chi^2$/dof=$11.0/20$ \\

857:  &&& \\

858: NCC & $1.06\pm0.31 (0.44)$ & $1.05\pm0.28 (0.56)$ & $1.08\pm0.36 (0.31)$ \\

859:  & $Y_{0.1}=1.06\pm0.07$ & $Y_{0.1}=1.08\pm0.15$ & $Y_{0.1}=1.27\pm0.20$ \\

860:  & $\mu=0.00\pm0.07$ & $\mu=0.03\pm0.14$ & $\mu=-0.29\pm0.27$ \\

861:  & $\chi^2$/dof=$41.7/24$ & $\chi^2$/dof=$31.7/11$ & $\chi^2$/dof=$8.5/11$ \\

862: \enddata

863: \end{deluxetable*}

864: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

865: Adopting the A05 abundances has not changed the best-fit values of $kT$

866: at every radius (always fully consistent within the 1$\sigma$ errors)

867: in the individual clusters, therefore the best fit functionals

868: representing both the CC and the NCC sample

869: have not varied.

870: 

871: A clearer picture can be seen also if we compute an error-weighted

872: average of the $kT$ profiles in several bins of width 0.05 in $r/r_{180}$.

873: The contribution to the single bin is provided from the measurements (and relative

874: error) that fall into that bin, weighted in proportion to the percentage

875: of the spatial coverage of the bin.

876: The error-weighted mean $kT/\langle kT \rangle$ profile is plotted in

877: Figure~\ref{T_CC_vs_noCC_mean} and compared with the local estimates

878: from DM02.

879: CC and NCC objects show well defined opposite gradient in the inner radial

880: and more similar behaviour moving outwards.

881: A good agreement is also observed with the local profiles, apart from two

882: significant deviations: (i) our CC mean profile appears flatter

883: at $r>0.2 r_{180}$, with an error-weighted value of $0.97 \pm 0.06$

884: (r.m.s. 0.07), to be compared with the local value of $0.85 \pm 0.11$

885: (r.m.s. 0.16), (ii) our NCC profile is steeper within $0.1 r_{180}$,

886: with a mean value of $1.27 \pm 0.08$ (r.m.s. 0.14) with respect to

887: the local value of $1.04 \pm 0.05$ (r.m.s. 0.07).

888: 

889: 

890: \subsection{The metal abundance profiles}

891: 

892: The metallicity profiles are plotted against the radius

893: normalized to $r_{180}$ for the CC and NCC sample

894: in Figure~\ref{abundances}b and \ref{abundances}d, respectively.

895: A different behaviour

896: in the very central regions between the two samples is quite clear.

897: To characterize this behaviour, we have fitted

898: the normalized profiles $Z/\langle Z \rangle$ with respect

899: to $r/r_{180}$ with single power-laws $Y \propto r^{\mu}$

900: over different radial ranges.

901: While the NCC clusters presents a flat profile within $\sim0.1 r_{180}$,

902: a sharper negative gradient is observed in the CC

903: cluster sample ($\mu=-0.24$; see Table~\ref{bestfit}).

904: Also at $r > 0.1 r_{180}$, where a model with a single power-law

905: well reproduce the data (reduced $\chi^2$ less than 1),

906: the CC clusters show hints of a

907: steeper profile ($\mu = -0.52 \pm 0.18$ in the CC sample,

908: $\mu = -0.29 \pm 0.27$ in the NCC sample).

909: 

910: We compute an error-weighted average $Z$ profile as done for the

911: temperature profile and compare it with the local measurements in DM01,

912: after scaling the radii to $r_{180}$ (see Figure~\ref{Z_CC_vs_noCC_mean}).

913: The two profiles are in agreement at $r>0.1r_{180}$, 

914: with both subsamples showing an evidence for a negative

915: gradient in metallicity significant at least at 2$\sigma$. A

916: different behaviour between the two subsample is observable in the

917: central bin, where the value in the CC sample is $\sim20\%-30\%$

918: higher than in the NCC sample. It is worth noticing that all the

919: clusters in our sample have $kT>6$ keV and this trend may be different

920: in lower temperature clusters.  DM01 in their analysis observed a

921: clear gradient in the metallicity profiles of their CC clusters, while

922: the profiles of their NCC clusters were almost constant. Moreover the

923: average metallicity observed in the CC clusters was systematically

924: higher than in the NCC sample, at least within $\sim0.3r/r_{180}$

925: from the center.  In our analysis, we do not find a clear difference

926: between the CC and NCC abundance profiles as in DM01. Except for the

927: inner radial bin where the metallicity in CC objects can be higher

928: by 50 per cent than in NCC ones, the CC and NCC profiles are very similar and

929: consistent within the errors.

930: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

931: \begin{figure}

932: \plotone{f5a.eps}

933: \plotone{f5b.eps}

934: \caption{Error-weighted mean metal abundance profile of the CC ({\it filled circles})

935: and NCC ({\it open diamonds}) clusters in our sample at intermediate redshift (top panel).

936: (Bottom) Comparison between our results and the local measurements in DM01 (shaded regions).

937: The 1$\sigma$ errors on the means are plotted as solid lines (dark gray region for the local

938: estimates) while the scatter (rms) in each data bin is shown as a dotted line

939: (light gray region for local estimates).}

940: \label{Z_CC_vs_noCC_mean}

941: \end{figure}

942: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

943: 

944: To compare our mean values at intermediate redshift with the results

945: obtained locally from DM01, we estimate an error-weighted average

946: $Z$ profile for both CC and NCC objects in local and intermediate $z$ samples

947: (Figure~\ref{Z_CC_vs_noCC_mean}).

948: While the slope of the profiles is generally in agreement with DM01 for

949: the both the CC and NCC clusters,

950: the value of $Z/\langle Z \rangle$ is systematically higher in our sample

951: with respect to the DM01 sample, with differences up to 50\%

952: within $0.1 r_{180}$ of CC systems.

953: This might be due to the different method used to compute

954: $\langle Z \rangle$, being in DM01 work estimated as fit with a constant

955: to the radial metallicity profile.

956: However, apart from the inner radial region, the discrepancy between

957: the local profiles and the ones at intermediate redshits are

958: within 1$\sigma$.

959: Using the same method to determine $\langle Z \rangle$ on our data, we find 

960: that the values of $Z/\langle Z \rangle$ are fully consistent with DM01.

961: 

962: 

963: 

964: \subsection{Comparison with the new compilations of solar values}

965: 

966: All the analysis described in the current section has been performed

967: adopting the AG89 compilation of photospheric abundances. This choice

968: has been due mainly by the necessity to have a direct comparison with

969: previous works in the literature (e.g. DM01).  However, as explained

970: above in \S1, the abundance values listed in AG89 have been recently

971: superseeded by the new photospheric values by Grevesse \& Sauval (1998)

972: and A05, who introduced a 0.676 and 0.60 times lower Iron solar abundance,

973: respectively.  Therefore, we have also performed

974: the fits using solar abundances by A05 to check whether adopting

975: the "old" AG89 values might have introduced any bias in our analysis.\\

976: The shape of the $Z$ profile for both the CC sample and the NCC sample

977: resembles very closely that observed in Figure~\ref{Z_CC_vs_noCC_mean} for

978: the AG89 values of $Z$. However to better quantify this comparison we

979: fitted also these new values with a power-law functional.

980: We measure

981: \begin{eqnarray}

982: Z_{\rm CC} / \langle Z \rangle = 0.67\pm0.06\: x^{-0.25\pm0.03}; & \\ 

983: Z_{\rm NCC} / \langle Z \rangle = 1.00_{-0.18}^{+0.20}\: x^{-0.02\pm0.07}, &

984: \end{eqnarray}

985: with $x\equiv r/r_{180}$.

986: If we compare the result of the fit with the last two rows in the 

987: first column in Table~\ref{bestfit},

988: it is quite clear that the slope of the power-law in both the CC and

989: NCC sample is consistent with what we obtained using the AG89 values

990: (well within the 1$\sigma$ uncertainties) and the difference is only

991: in the normalization. As expected, this result might indicate that our

992: metallicities are mostly driven by Iron and that the contribution of

993: the $\alpha$ elements to the determination of the abundances is

994: negligible. Indeed, all our clusters have a temperature larger than 

995: $\sim6$ keV,

996: therefore the abundance measures are dominated by the Fe-K$\alpha$ line.

997: As a further test to this hypothesis we tried to fit the

998: Fe abundance independently from the $\alpha$ elements abundances.  To

999: this aim we used a {\tt vmekal} model where the abundances of O, Mg,

1000: Si and S were tied-up to the same value and fitted as a single free

1001: parameter (in order to reduce the number of free parameters), while the

1002: other elements (apart from Fe) were frozen at solar. While the Fe

1003: abundance remained always consistent with the value of $Z$ measured

1004: considering the metallicity as a single parameter, in almost all the

1005: spectra we could not find any 

1006: statistically significant detection of a contribution from  $\alpha$

1007: elements.  

1008: Their abundances has been measured as upper limits

1009: in most cases (generally $Z_\alpha<0.3$), and as very low values in

1010: the rest of the spectra (generally $Z_\alpha\sim0.1-0.2$, often

1011: consistent with $Z_\alpha\sim0$ at 1$\sigma$).  This is true even in

1012: the inner part of the clusters in the sample, where the

1013: signal-to-noise of the spectra is higher and in principle it may be

1014: easier to detect the presence of elements

1015: other than Iron.\\

1016: These results suggest that the measure of $Z$ in our cluster sample

1017: consists mainly in a measure of Iron metallicity. Therefore, adopting

1018: the AG89 solar abundances instead of the A05 (which have different

1019: $Z_{Fe}/Z_\alpha$ ratios) produces only a difference in the

1020: value of the relative $Z$ measured. This does not introduce 

1021: any bias in the analysis of the radial profiles, since 

1022: the absolute value of $Z$ does not change and only the reference 

1023: value assumed for the solar metallicity experience a variation. 

1024: 

1025: 

1026: \subsection{Gradients and cooling times}

1027: 

1028: We investigate the correlation of the central slopes of the temperature 

1029: and metallicity profiles with the central cooling time in each cluster. 

1030: To this aim, the values of $kT(r) /\langle kT \rangle$,

1031: $Z(r) /\langle Z \rangle$ and $r/r_{180}$ (normalized to the average

1032: temperature, the average metallicity and the virial radius measured in

1033: each cluster, respectively, as described in Section~3) 

1034: with $r<0.1 r_{180}$ have been considered to characterize the cooling cores.

1035: We find that the most robust correlation is present between 

1036: the slope $\mu$ of the temperature profiles,

1037: \begin{equation}

1038: kT/\langle kT \rangle = A \:(r/r_{180})^\mu,

1039: \end{equation}

1040: and $\tau_c/\tau_{age}$, with a Spearman's $\rho$ rank correlation

1041: value of -0.87 that corresponds to a significance of the

1042: non-correlation case of $P=2\times10^{-4}$. In the $kT-Z$ and $Z-r$

1043: relations, the values of the Spearman's $\rho$ are 0.44 and 0.54, 

1044: corresponding to a significance of 0.15 and 0.07, respectively.

1045: The exponent $\mu$ correlates with $\tau_c/\tau_{age}$, 

1046: being higher at lower values of $\tau_c/\tau_{age}$,

1047: with all the CC clusters having $\mu>0$ and $\tau_c/\tau_{age}\la0.6$

1048: (Figure~\ref{alphatc}).

1049: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1050: \begin{figure}

1051: %\epsscale{0.90}

1052: \plotone{f6.eps}

1053: \caption{Distribution of the exponential $\mu$ in the relation 

1054: $kT/\langle kT \rangle=A\:(r/r_{180})^\mu$ as a function of $\tau_c/\tau_{age}$.

1055: }

1056: \label{alphatc}

1057: \end{figure}

1058: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1059: 

1060: 

1061: \section{Conclusions}

1062: 

1063: In the present work we analyzed a sample of 12 galaxy clusters present

1064: in the Chandra archive with at least $\sim20,000$ net ACIS counts and

1065: $kT>6$ keV.  These clusters were chosen in the 0.1-0.3 redshift range,

1066: regardless of their shape. We computed the cooling time of the

1067: clusters, subdividing the sample in 7 cool core clusters and 5

1068: non--cool core clusters. This subdivision allowed us to compare

1069: the two categories in a systematic fashion, following the approach of

1070: DM01.  We performed a spectral analysis in radial bins of each cluster

1071: in the sample requiring each bin to have $\sim7000-8000$ counts,

1072: fitting the spectra with a thermal model with Galactic

1073: absorption. This allowed us to derive temperature and metallicity

1074: profiles for each cluster.  The virial radius $r_{180}$ was computed

1075: in order to renormalize the radii to physically meaningful quantities

1076: and investigate for self-similarities in the radial profiles. To this

1077: aim the global temperature $\langle kT \rangle$ and metallicity

1078: $\langle Z \rangle$ in each cluster were measured as well.

1079: The main results coming from our work can be summarized as follows.\\

1080: \begin{itemize}

1081: 

1082: \item The temperature profiles in the inner $0.1 r_{180}$ have, on

1083:   average, a positive gradient, $kT(r)\propto r^\mu$ with $\mu \simeq

1084:   0.25$ in CC systems, whereas it is almost flat in NCC systems. The

1085:   outer regions are well fitted with a single power-law with slopes

1086:   significantly different, being steeper ($\mu =-0.32 \pm 0.05$) in

1087:   NCC objects.

1088:   The general trend of our CC sample is fully consistent with 

1089:   Vikhlinin et al. (2005) at $r\la0.3r_{180}$. The low number statistics

1090:   above $0.3r_{180}$ prevents us from any statistically significant 

1091:   comparison between the two samples at $r\ga0.3r_{180}$.

1092: 

1093: \item The metallicity profiles in the inner regions is almost constant

1094:   in NCC clusters around the value measured excluding counts

1095:   from $r<0.07 r_{180}$. In the CC sample, a steep negative gradient

1096:   is observed ($\mu =-0.27 \pm 0.03$) in the central regions.

1097:   At $r > 0.1 r_{180}$, a power-law reproduces well the distribution

1098:   of the spectral measurements, with a slope that is marginally

1099:   steeper in CC clusters ($\mu =-0.52 \pm 0.18$) than in NCC

1100:   clusters ($\mu =-0.29 \pm 0.27$). 

1101: 

1102: \item Comparing our averaged metallicity profiles with the ones in DM01,

1103:   we found that our values of $Z/\langle Z \rangle$ are systematically higher,

1104:   with differences up to 50\% within $0.1 r_{180}$ of CC systems.

1105:   This may be explained by the different method adopted in DM01 to estimate 

1106:   $\langle Z \rangle$, as best-fit with a constant over the entire 

1107:   metallicity profile, without any exclusion of the central core.

1108: 

1109: \item Using the solar abundances from Asplund (2005, A05) gives

1110:   consistent results with what we obtain using the values by Anders \&

1111:   Grevesse (1989, AG89), with a discrepancy only in the normalization 

1112:   (as expected, $\sim60-70\%$ higher)

1113:   but not in the slope of the $Z$ radial profiles.  Together with the

1114:   fact that, in most cases, we were able to measure the $\alpha$

1115:   elements only as upper limits, this indicates that our metallicities

1116:   are mostly driven by Iron and that adopting the AG89 solar

1117:   abundances instead of the A05 results in a difference only in the

1118:   absolute values of the $Z$ measured but does not introduce any bias

1119:   in the radial profile analysis.

1120: 

1121: \item Fitting a power--law shape to the temperature profiles,

1122:   $kT/\langle kT \rangle=A\:(r/r_{180})^\mu$, we found that $\mu$

1123:   correlates strongly with the cluster cooling times, being higher at low

1124:   values of $\tau_c/\tau_{age}$, with all the CC clusters having

1125:   $\alpha>0$ and $\tau_c/\tau_{age}\la0.6$. 

1126:   As expected, strong correlation is also observed between the inner slope 

1127:   of the metallicity profile and cluster cooling time.

1128: 

1129: \end{itemize}

1130: 

1131: In general, our results further demonstrate the invaluable role played

1132: by X--ray archival studies of the chemo-- and thermo--dynamical

1133: properties of galaxy clusters. Analyses based on the Chandra archive,

1134: like that presented here (see also Vikhlinin et al. 2005; Balestra et

1135: al. 2007; Maughan et al. 2007), in combination with analogous studies

1136: from the XMM--Newton archive, will constitute an important heritage

1137: from the present generation of X--ray satellites for years to

1138: come. Nowadays, available data on the evolution of the chemical

1139: enrichment of the ICM provide important constraints on models aimed

1140: at explaining the past history of star formation and the dynamical

1141: processes taking place during the cosmological build up of galaxy

1142: clusters. However the study of the thermo-dynamical properties of the 

1143: cooling cores and the evolution of the abundance distributions in clusters

1144: with the redshift are just a part of what can be currently done 

1145: exploiting in full the existing Chandra and XMM--Newton archives.

1146: Archival works like ours have the potential to shed new light on the 

1147: properties of the

1148: stellar populations responsible for the ICM enrichment, and on the

1149: mechanisms which lead to the generation of the cool cores and

1150: determine the transport and diffusion of heavy elements from star

1151: forming regions. 

1152: 

1153: \acknowledgments AB and PM acknowledge financial support from CXO

1154: grant AR6-7015X and from NASA grant GO5-6124X. 

1155: We acknowledge financial contribution from contract

1156: ASI--INAF I/023/05/0. PT and SB acknowledge financial

1157: contribution from the PD51 INFN grant.

1158: We thank A. Vikhlinin for providing us the temperature and abundance 

1159: profiles of some of the clusters in his sample.

1160: We also thank F.Gastaldello for useful discussions.

1161: We thank the anonymous referee for comments and suggestions useful 

1162: to improve the presentation of the paper.

1163: 

1164: \begin{thebibliography}{}

1165: 

1166: \bibitem[{{Anders} \& {Grevesse}(1989)}]{and89}

1167: {Anders}, E., \& {Grevesse}, N. 1989, \gca, 53, 197 (AG89)

1168: 

1169: \bibitem[{{Arnaud}(1996)}]{arn96}

1170: {Arnaud}, K.~A. 1996, in ASP Conf. Ser. 101: Astronomical Data Analysis

1171:   Software and Systems V, ed. G.~H. {Jacoby} \& J.~{Barnes}, 17--+

1172: 

1173: \bibitem[{{Asplund} {et~al.}(2005){Asplund}, {Grevesse}, \& {Sauval}}]{asp05}

1174: {Asplund}, M., {Grevesse}, N., \& {Sauval}, A.~J. 2005, in ASP Conf. Ser. 336:

1175:   Cosmic Abundances as Records of Stellar Evolution and Nucleosynthesis, ed.

1176:   T.~G. {Barnes} \& F.~N. {Bash}, 25--+ (A05)

1177: 

1178: \bibitem[{{Balestra} {et~al.}(2007)}]{bal07}

1179: {Balestra}, I., {Tozzi}, P., {Ettori}, S., {Rosati}, P., {Borgani}, S., {Mainieri}, V., 

1180: \& {Norman}, C.,  2007, \aap, 462, 429

1181: 

1182: \bibitem[{{De Grandi} \& {Molendi}(2001)}]{deg01}

1183: {De Grandi}, S., \& {Molendi}, S. 2001, \apj, 551, 153 (DM01)

1184: 

1185: \bibitem[{{De Grandi} \& {Molendi}(2002)}]{deg02}

1186: {De Grandi}, S., \& {Molendi}, S. 2002, \apj, 567, 163 (DM02)

1187: 

1188: \bibitem[{{Ettori}, {De Grandi} \& {Molendi}(2002)}]{ett02}

1189: {Ettori}, S., {De Grandi}, S., \& {Molendi}, S. 2002, \aap, 391, 841

1190: 

1191: \bibitem[{{Ettori} et al. (2004)}]{ett04}

1192: {Ettori}, S., {Tozzi}, P., {Borgani}, S., \& {Rosati}, P. 2004, \aap, 417, 13

1193: 

1194: \bibitem[{{Evrard} {et~al.}(1996){Evrard}, {Metzler}, \& {Navarro}}]{evr96}

1195: {Evrard}, A.~E., {Metzler}, C.~A., \& {Navarro}, J.~F. 1996, \apj, 469, 494

1196: 

1197: \bibitem[Finoguenov et al.(2000)]{fin00} Finoguenov, A., David, L.~P., \& Ponman, T.~J. 2000, \apj, 544, 188

1198: 

1199: \bibitem[{{Grevesse} \& {Sauval}(1998)}]{gre98}

1200: {Grevesse}, N., \& {Sauval}, A.~J. 1998, Space Science Reviews, 85, 161

1201: 

1202: \bibitem[Irwin \& Bregman (2001)]{irw01} Irwin, J.~A., \& Bregman, J.~N. 2001, \apj, 546, 150

1203: 

1204: \bibitem[{{Kaastra}(1992)}]{kaa92}

1205: {Kaastra}, J.~S. 1992, (Internal SRON--Leiden Report, updated version 2.0)

1206: 

1207: \bibitem[{{Liedahl} {et~al.}(1995){Liedahl}, {Osterheld}, \&

1208:   {Goldstein}}]{lie95}

1209: {Liedahl}, D.~A., {Osterheld}, A.~L., \& {Goldstein}, W.~H. 1995, \apjl, 438,

1210:   L115

1211: 

1212: \bibitem[Machacek et al. (2002)]{mac02} Machacek, M.~E., Bautz, M.~W., Canizares, C., Garmire, G.~P.

1213: 2002, \apj, 567, 188

1214: 

1215: \bibitem[Majerowicz et al. (2002)]{maj02} Majerowicz, ,S. Neumann, D.~M., Reiprich, T.~H. 2002,

1216: \aap, 394, 77

1217: 

1218: \bibitem[Markevitch et al. (2003)]{mar03} Markevitch, M., et al. 2003, \apj, 583, 70

1219: 

1220: \bibitem[{{Matteucci} \& {Vettolani}(1988)}]{mat88}

1221: {Matteucci}, F., \& {Vettolani}, G. 1988, \aap, 202, 21

1222: 

1223: \bibitem[Maughan et al.(2007)]{2007astro.ph..3156M} Maughan, B.~J., Jones, 

1224: C., Forman, W., \& Van Speybroeck, L.\ 2007, \apj, submitted, arXiv:astro-ph/0703156

1225: 

1226: \bibitem[{{Nousek} \& {Shue}(1989)}]{nou89}

1227: {Nousek}, J.~A., \& {Shue}, D.~R. 1989, \apj, 342, 1207

1228: 

1229: \bibitem[Piffaretti et al. (2005)]{pif05} Piffaretti, R., Jetzer, P., Kaastra, J.~S., Tamura, T.

1230: 2005, \aap, 433, 101

1231: 

1232: \bibitem[Pratt \& Arnaud 2002]{pra02} Pratt, G.~W., \& Arnaud, M. 2002,

1233: \aap, 394, 375

1234: 

1235: \bibitem[Pratt et al. (2005)]{pra05} Pratt, G.~W., B\"ohringer, H.,

1236: \& Finoguenov, A. 2005, \aap, 433, 777

1237: 

1238: \bibitem[Renzini(1997)]{1997ApJ...488...35R} Renzini, A.\ 1997, \apj, 488, 

1239: 35 

1240: 

1241: \bibitem[{{Rosati} {et~al.}(2002){Rosati}, {Borgani}, \& {Norman}}]{rbn02}

1242: {Rosati}, P., {Borgani}, S., \& {Norman}, C. 2002, \araa, 40, 539

1243: 

1244: \bibitem[Stark et al. (1992)]{sta92} Stark, A.~A., Gammie, C.~F., Wilson, R.~W., Bally, J., 

1245: Linke, R.~A., Heiles, C., Hurwitz, M. 1992, \apjs, 79, 77

1246: 

1247: \bibitem[Tamura et al. (2004)]{tam04} Tamura, T., Kaastra, J.~S., den Herder,

1248: J.~.W.~A., Bleeker, J.~A.~M., Peterson, J.~R. 2004, \aap, 420, 135

1249: 

1250: \bibitem[{{Vikhlinin} {et~al.}(2005){Vikhlinin}, {Markevitch}, {Murray},

1251:   {Jones}, {Forman}, \& {Van Speybroeck}}]{vik05}

1252: {Vikhlinin}, A., {Markevitch}, M., {Murray}, S.~S., {et~al.} 2005, \apj, 628,

1253:   655

1254: 

1255: \bibitem[Voigt \& Fabian (2006)]{voi06} Voigt, L.~M., \& Fabian, A.~C.

1256: 2006, \mnras, 368, 518

1257: 

1258: \bibitem[{{Voit}(2005)}]{v05}

1259: {Voit}, G.~M. 2005, Reviews of Modern Physics, 77, 207

1260: 

1261: \end{thebibliography}

1262: 

1263: 

1264: \end{document}

1265: 

1266: 

1267: