1: \documentclass[]{emulateapj}
2:
3: %\documentclass[12pt,preprint]{aastex}
4:
5: %% manuscript produces a one-column, double-spaced document:
6:
7: % \documentclass[manuscript]{aastex}
8:
9: %% preprint2 produces a double-column, single-spaced document:
10:
11: % \documentclass[preprint2]{aastex}
12:
13: %% Sometimes a paper's abstract is too long to fit on the
14: %% title page in preprint2 mode. When that is the case,
15: %% use the longabstract style option.
16:
17: %% \documentclass[preprint2,longabstract]{aastex}
18:
19: \usepackage{verbatim}
20: \usepackage{multirow}
21: \bibpunct[, ]{(}{)}{;}{a}{}{,}
22: \bibliographystyle{apj}
23:
24: \newcommand{\bey}{\begin{eqnarray}}
25: \newcommand{\eey}{\end{eqnarray}}
26: \newcommand{\beq}{\begin{equation}}
27: \newcommand{\eeq}{\end{equation}}
28: \newcommand{\grb}{GRB~070802}
29: \newcommand{\bump}{$2175$~{\AA}}
30:
31:
32: \shorttitle{Detection of Milky Way type dust}
33: \shortauthors{El{\'i}asd{\'o}ttir et al.}
34: \begin{document}
35:
36: \submitted{Matches version accepted by the ApJ.}
37:
38: \title{Dust Extinction in High z Galaxies with GRB Afterglow Spectroscopy - The 2175 {\AA} Feature at $z=2.45$
39: \thanks{Based on observations
40: collected under progs. ID 079.D-0429(B) and 177.D-0591(P,Q), using the
41: FORS2 instrument installed at the Cassegrain focus of the Very Large Telescope
42: (VLT), Unit 1 - Antu, operated by the European Southern Observatory (ESO)
43: on Cerro Paranal, Chile.}}
44:
45: \author{
46: \'A. El{\'i}asd{\'o}ttir\altaffilmark{2,3},
47: J.~P.~U. Fynbo\altaffilmark{3},
48: J. Hjorth\altaffilmark{3},
49: C. Ledoux\altaffilmark{4},
50: D.~J. Watson\altaffilmark{3},
51: A.~C. Andersen\altaffilmark{3},
52: D. Malesani\altaffilmark{3},
53: P.~M. Vreeswijk\altaffilmark{3},
54: J.~X. Prochaska\altaffilmark{5},
55: J. Sollerman\altaffilmark{3,6},
56: A.~O. Jaunsen\altaffilmark{7}
57: }
58:
59: \altaffiltext{2}{Department of Astrophysical Sciences, Princeton University, Princeton, NJ 08544-1001, USA; ardis@astro.princeton.edu }
60: \altaffiltext{3}{Dark Cosmology Centre, Niels Bohr Institute, University of
61: Copenhagen, Juliane Maries Vej 30, DK-2100 Copenhagen \O, Denmark}
62: \altaffiltext{4}{European Southern Observatory, Avenida Alonso de C{\'o}rdova 3107, Casilla 19001, Vitacura, Santiago, Chile}
63: \altaffiltext{5}{Department of Astronomy and Astrophysics,
64: UCO/Lick Observatory; University of California, 1156 High Street, Santa Cruz,
65: CA 95064; xavier@ucolick.org}
66: \altaffiltext{6}{Stockholm Observatory, Department of Astronomy, Stockholm University, AlbaNova University Center, 106 91, Stockholm, Sweden}
67: \altaffiltext{7}{Institute of Theoretical Astrophysics, University of Oslo, PO Box 1029 Blindern, N-0315 Oslo, Norway}
68:
69: \begin{abstract}
70: We report the clear detection of the 2175~\AA~dust absorption feature in the optical afterglow spectrum of the
71: gamma-ray burst (GRB) GRB\,070802 at a redshift of $z=2.45$. This is the
72: highest redshift for a detected 2175~\AA~dust bump to date, and it is the first clear
73: detection of the 2175~\AA~bump in a GRB host galaxy, while several tens
74: of optical afterglow spectra without the bump have been recorded in the past
75: decade.
76: The derived extinction curve gives $A_V=0.8$--$1.5$ depending on the assumed intrinsic slope. Of the three local extinction laws, an LMC type extinction gives the best fit to the extinction curve of the host of \grb.
77: Besides the 2175~\AA~bump we
78: find that the spectrum of GRB\,070802 is characterized by unusually strong
79: low-ionization metal lines and possibly a high metallicity for a
80: GRB sightline ([Si/H$]=-0.46\pm 0.38$, [Zn/H$]=-0.50\pm 0.68$). In particular, the spectrum of \grb\ is unique for a GRB spectrum in that it shows clear \ion{C}{1} absorption features, leading us to propose a correlation between the presence of the bump and \ion{C}{1}. The gas to dust ratio for the host galaxy is found to be significantly lower than that of other GRB hosts with N(\ion{H}{1})$/A_V=(2.4\pm1.0)\times$~$10^{21}$~cm$^{-2}$~mag$^{-1}$, which lies between typical MW and LMC values. Our results are in agreement with the tentative conclusion reached by
81: \citet{gordon2003} that the shape of the extinction curve, in particular the
82: presence of the bump, is affected by the UV flux density in the environment of
83: the dust.
84: \end{abstract}
85:
86: \keywords{dust, extinction -- galaxies: ISM -- gamma rays: bursts -- galaxies: abundances -- galaxies: distances and redshifts}
87: \section{Introduction}
88:
89: Dust extinction curves quantify as a function of
90: wavelength the amount of light `lost' due to scattering and absorption of the light by dust particles
91: along the line of sight from an object to the observer. In the Milky Way (MW), the
92: extinction curve has been extensively mapped and has been shown to follow an
93: empirical single-parameter function for almost all lines of sight \citep{cardelli1989}.
94: The most characteristic feature of this function is a broad `bump' (i.e.,
95: excess extinction) centered at $2175$~\AA\ first discovered by \citet{stecher}.
96: Today, more than 40 years after its discovery, the origin of the feature remains unknown although several candidates have been suggested (see \S~\ref{sec:2175}).
97:
98: As outlined below (in \S~\ref{sec:grb_after}), a promising method for studying extinction curves
99: in distant galaxies is to use the spectra of the afterglows of gamma-ray bursts
100: (GRBs) as the backlight against which the extinction curve can be inferred. The
101: intrinsic spectral energy distributions (SEDs) of GRB afterglows are very
102: simple, consisting of a single or a broken power-law from the X-ray band to the
103: infrared (IR) and hence it is relatively easy to infer the shape of the
104: extinction curve by measuring the curvature of the optical spectrum, the
105: deviation of the afterglow spectrum extending from the X-ray extrapolation, or from the
106: X-ray--IR interpolation.
107:
108: In this paper we present the detection of the $2175$~\AA~bump in a GRB
109: absorption system at $z=2.45$. The detection was briefly reported by
110: \citet{2007Msngr.130...43F}, but this paper gives the detailed analysis of our
111: detection. The detection of the bump has also subsequently been confirmed based
112: on photometry alone \citep{kruhler08}. This is the first time that a clear signature of the 2175\ \AA\ feature has been observed in a GRB host galaxy and the highest redshift detection of the bump to date.
113:
114: The paper is organized as
115: follows: In the remainder of this section we discuss the nature and previous
116: detections of the 2175~{\AA} feature (\S~\ref{sec:2175}) and the use of GRBs to
117: infer extinction curves (\S~\ref{sec:grb_after}) in more detail. In \S~\ref{sec:obs} we present
118: the observations and data analysis. We discuss the properties of the derived
119: extinction curve in \S~\ref{sec:ext}. The results and possible tracers of the
120: bump are presented in \S~\ref{sec:dis} and finally we summarize our conclusions
121: in \S~\ref{sec:con}.
122:
123:
124: \subsection{Extragalactic extinction curves}
125: \label{sec:ext_ext}
126: The study of extinction curves outside the Milky Way is a challenging task and has mostly been limited to galaxies in the Local Group. While the M31 extinction curve resembles that of the Milky Way \citep{bianchi}, the extinction curves of the Large and Small Magellanic Clouds (LMC and SMC) show significant variation. Lines of sight in the LMC can be broadly put into two categories, one having extinction curves similar to that of the Milky Way ("LMC average") and the other showing extinction curves with a much less prominent bump and a steeper rise into the UV \citep["LMC2 supershell"][]{1981MNRAS.196..955N,1999ApJ...515..128M,gordon2003}. While the phrase "LMC type of extinction" refers to extinction curves of the second type (i.e. less prominent bump and a steep UV rise), it should be kept in mind that Milky Way sightlines also exist in the LMC. Similarly, for the SMC, the canonical SMC extinction curve shows no evidence for a bump and an even steeper
127: rise into the UV \citep{1984A&A...132..389P,gordon2003} although one out of five lines of sight shows the 2175\,\AA\ bump \citep{1982A&A...113L..15L}. It is also worth noting that there are a few known lines of sight in the Milky Way with LMC type of extinction \citep{clayton00} and a single line of sight with a measured SMC type of extinction \citep{2003ApJ...598..369V}.
128:
129:
130: Relatively little is known about dust extinction outside the Local Group as the method
131: of measuring extinction curves by comparing the spectra of single stars is not applicable at
132: larger distances.
133: Several methods have been attempted, including studying
134: statistical samples of reddened and standard quasars \citep{1999ApJ...522..604P,vladilo08},
135: studying individual supernova (SN) Ia lightcurves and using gravitationally lensed quasars.
136: A recent overview of these studies and their results can be found in
137: \citet{eliasdottir2006}. The SEDs of GRBs may also be used, as discussed in \S~\ref{sec:grb_after}.
138: These studies have found varying types of extinction which can deviate significantly from MW-type extinction, leaving open the question of how common MW type dust is.
139: In particular, prior to the present study there
140: have only been three robust detections of the
141: $2175$~\AA~bump in individual extragalactic systems beyond the local group.
142: The first is in a lensing galaxy at a redshift of $z=0.83$ \citep{motta2002}.
143: The second is from a damped Ly$\alpha$ system at
144: $z=0.524$ \citep{junkkarinen2004}. The third is in an intervening absorber at $z=1.11$ toward GRB~060418 \citep{2006MNRAS.372L..38E}.
145:
146: Extinction curves are
147: the prime diagnostic tool available to study dust in the optical/UV regime.
148: They depend sensitively on the composition of the dust
149: \citep{henning_etal03}, allowing estimates to be made of the nature and the
150: origin of dust and its dependence on cosmic time, metallicity, etc.
151: Furthermore, in studies of distant objects, e.g., using SNe Ia as standard candles or determining flux ratios of strongly lensed
152: QSOs, it has become increasingly important to accurately determine the
153: amount and wavelength dependence of the dust extinction.
154:
155:
156: \subsection{The nature and origin of the 2175~{\AA} extinction feature}
157: \label{sec:2175}
158: \citet{fitzpatrick+massa86} studied the 2175~{\AA} interstellar extinction bump in the
159: direction of 45 reddened stars in the Milky Way, and found that it displays a peak whose central
160: wavelength $\lambda_{0}$ is remarkably constant with extreme deviation of only
161: $\sim \pm 17$~{\AA} from the mean position of $\lambda_{0} = 2174.4$~{\AA}.
162: As this is significantly larger than the measurement uncertainty it seems
163: to indicate a real, but small, variation of the peak position. The full
164: width at half maximum (FWHM) of the bump has a larger range of intrinsic
165: values, from $360$ to $600$~{\AA}. There seems to be no correlation between
166: the width and the central wavelength of the bump.
167:
168:
169: The fact that the 2175~{\AA} feature becomes progressively fainter in the
170: MW, LMC and SMC has been attributed to the progressively lower metal
171: abundances of the LMC and SMC \citep{fitzpatrick04}
172: or to differences in the radiative environment \citep{gordon97,clayton00,mattsson}. \citet{gordon2003} studied the differences in the extinction of the MW, SMC and LMC finding tentative evidence that the bump strength correlates with dust-to-gas ratio.
173:
174:
175: \citet{noll07} found in their study of 108 massive, UV-luminous galaxies at
176: $1 < z < 2.5$
177: that there is a
178: correlation between heavy reddening and the presence of a 2175 {\AA} feature.
179: The least reddened objects have SEDs consistent with the average featureless and
180: steep extinction curve of the SMC.
181: For their objects at $1 < z < 1.5$ a significant UV bump
182: is present in galaxies that appear disk-like and which host a rather large
183: fraction of intermediate-age
184: stars (i.e., from 0.2 to 1--2 Gyr old). More clumpy and irregular objects
185: seem to have extinction curves
186: resembling those of nearby starburst galaxies.
187:
188: Several candidates have been proposed as the dust particles responsible for
189: the bump ranging from iron-poor silicate grains \citep{steel+duley} to
190: carbonaceous materials such as carbon-onions \citep{henrard}, graphite grains
191: \citep{stecher+donn,draine89} or the polycyclic aromatic hydrocarbons (PAHs)
192: believed also to be responsible for the prominent broad emission features found in
193: the mid-infrared \citep{duley+seahra,duley,cecchi-pestellini}.
194:
195: Graphite has
196: been considered a very promising, though contentious, candidate for
197: explaining the 2175~{\AA} bump \citep[e.g.,
198: ][]{stecher+donn,fitzpatrick+massa86,mathis94,nicolai90,sorell,draine+malhotra,rouleau,will+aannestad,andersen03,2003ApJ...592..947C}.
199: In an extensive investigation,
200: \citet{draine+malhotra} conclude that if graphite particles are the carriers of
201: the 2175~{\AA} peak, a variation in their optical properties must be
202: present, e.g., as a result of varying amounts of impurities, variations in
203: crystallinity, or changes in its electronic structure due to surface effects.
204: The observed lack of correlation between the central wavelength and the FWHM of
205: the peak is therefore a challenge for the hypothesis that graphite
206: particles are the source of this peak.
207:
208: Large PAH molecules are expected to at least contribute to the
209: 2175\,{\AA} feature as the interior carbon atoms have electronic orbitals
210: closely resembling those in graphene and an oscillator strength for each C
211: expected to be close to the value for graphite \citep{draine03}.
212: \citet{joblin} have suggested that PAHs could be entirely responsible for the
213: feature, which has been questioned by \citet{mathis94} due to the lack of an absorption
214: feature at about 3000~{\AA} in the spectra of PAHs. Another argument
215: against the PAHs as the main contributers is that the constancy in wavelength
216: along with a rather wide variation in the width, seems surprising if the bump
217: is caused by a mixture of widely varying materials. Such properties are more likely
218: to occur as a result of coating (or surface hydrogenation) of carbon grains
219: \citep{mathis94,draine+li,cecchi-pestellini}.
220:
221: Finally, iron-poor silicates in the form of partially hydrogenated amorphous
222: Mg$_{2}$SiO$_{4}$ particles have been suggested as the carriers of the
223: 2175\,{\AA} peak \citep{steel+duley}. The absorption is observed in silicates
224: of greatly differing composition and under a wide range of different sample
225: preparations and outgasing conditions. However,
226: \citet{mathis96} indicated that the silicates may be problematic as carriers.
227:
228: It seems most plausible that the carrier of 2175\,{\AA} peak is a composition of graphitic
229: and amorphous carbon phases, PAHs and glassy iron poor silicates. Various models have been
230: proposed with different compositions \citep[see e.g.,][and references therein]{mathis96, draine03}.
231:
232: \subsection{GRB afterglows and extinction}
233: \label{sec:grb_after}
234: GRB afterglows have very simple intrinsic synchrotron spectra which are
235: essentially featureless power laws ($f_\nu \propto \nu^{-\beta}$) across
236: the wavelength range of interest for measuring extinction curves
237: \citep[e.g., ][]{1998ApJ...500L..97G,2002ApJ...568..820G,1998ApJ...497L..17S}. This is a significant
238: advantage over using QSOs or galaxies as background objects. In cases where
239: the GRB physics places the cooling frequency between the optical and X-ray
240: regime (at the time of observation) there will be a cooling break in the spectrum
241: leading to a change in spectral slope of $\Delta\beta=0.5$ (softer spectrum in the
242: X-rays). The X-ray slope is usually around 1.0--1.2 \citep[see e.g.,][]{2007ApJ...670..565L} and so the intrinsic
243: optical slope is typically between 0.5 and 1.2. Moreover, the optical-NIR range
244: is typically well-described by a single power-law segment.
245:
246: \citet{1998ApJ...495L..99R} first attempted to constrain the extinction
247: towards GRB 970508 \citep[see also][]{2001ApJ...553..235R}. Several subsequent
248: studies confirmed that a simple SMC-type/linear extinction model provided
249: good fits to observations of several GRB afterglows
250: \citep{1998ApJ...508L..21B,1999ApJ...511L..85C,2001A&A...370..909J,2001A&A...373..796F,2001ApJ...549L...7P,2004ApJ...614..293S, 2004A&A...427..785J,2006ApJ...641..993K}.
251: Findings of a low dust-to-gas ratios in GRB afterglows exhibiting damped
252: Ly$\alpha$ absorption in the GRB host galaxy supports the apparent
253: prevalence of SMC-type dust in GRB environments
254: \citep{2001A&A...370..909J,2003ApJ...597..699H,2004A&A...419..927V}.
255:
256: Gray (wavelength-independent) extinction laws
257: \citep{2003ApJ...587..135G,2007ApJ...661L...9S,2008ApJ...672..449P,2008MNRAS.386L..87N}
258: or MW dust \citep{2006ApJ...641..993K,2007MNRAS.377..273S} have been
259: suggested in rare instances for GRB environments
260: although it should be noted that these claims are somewhat model dependent
261: (gray dust) or statistically insignificant (MW-type dust). The 2175\,\AA\
262: feature itself has been searched for and ruled out in
263: GRB~050401 \citep{2006ApJ...652.1011W}.
264: In GRB~000926 \citep{2001A&A...373..796F}
265: and GRB~020124 \citep{2003ApJ...597..699H}
266: SMC-type extinction was preferred although the existence of a 2175 \AA\ bump
267: was not sampled by the broad-band observations.
268: \citet{2006A&A...447..145V} found a bump in the spectrum of GRB~991216 but at
269: 2360 \AA\ if at the redshift of the GRB ($z=1.02$). To be
270: consistent with restframe 2175 \AA\ it would have to be due to an even
271: more distant absorber at $z=1.19$ which was however not detected as
272: an absorption system in the afterglow spectrum. The first (and thus far,
273: only) clear detection of the 2175 \AA\ bump in the foreground of a GRB,
274: consistent with a detected intervening \ion{Mg}{2} absorber ($z=1.1$), was
275: reported by \citet{2006MNRAS.372L..38E}.
276:
277: A significant fraction of GRB afterglows are optically `dark'
278: \citep{2001A&A...369..373F,2004ApJ...617L..21J,2005ApJ...624..868R}.
279: It is likely that a large fraction of these afterglows go undetected
280: in the optical because of extinction
281: \citep[e.g.,][]{2001ApJ...562..654D}.
282: Recently \citet{2008ApJ...681..453J} and \citet{2008arXiv0803.4100T}
283: demonstrated the validity of this dust obscuration hypothesis by detecting
284: two highly reddened systems. These studies highlight the difficulty in
285: probing systems with significant extinction at high redshift: GRBs with
286: high extinction are generally not accurately localized (because of the lack
287: of an optical afterglow position) and there are no detailed constraints on
288: their extinction laws, particularly in the UV, for systems with of order
289: $A_V\sim 5$. Such selection effects must be kept in mind before making
290: statistical inferences from extinction laws of GRB afterglows.
291:
292: Studies using the X-ray
293: extrapolation to set limits on the total extinction
294: \citep{1999ApJ...523..171V} or using the optical spectral curvature to
295: determine the reddening \citep{1998Natur.393...43R}
296: have now been carried out for nearly a decade. These studies may be further improved if one obtains the largely extinction-free IR flux, typically needing observations
297: in the mid-IR, and is the subject of our ongoing \emph{Spitzer} campaign
298: \citep[see also][]{2008arXiv0803.2879H}.
299:
300:
301:
302: \section{Observations and data analysis}
303: \label{sec:obs}
304:
305: \subsection{VLT imaging, spectroscopy and redshift of GRB 070802}
306:
307: GRB\,070802 \citep{Barthelmy2007GCN} was detected by \textit{Swift} \citep{gehrels2004} on 2007 Aug 2.29682 UT. The ESO VLT started observations about one hour after the burst in Rapid Response Mode, starting with a photometric sequence spanning from the $B$-band through $V$, $R$, and $I$ to the $z$-band in excellent seeing conditions ($\lesssim0.5$ arcsec) and at low airmass ($\lesssim1.4$). We detected the source previously detected by \citet{2007GCN..6695....1B} in all bands (see Table~\ref{tab:photometry}). The source was very red and classifies as a dark burst according to the definition of \citet{2004ApJ...617L..21J}. Based on 5 R-band points taken from about 1 to 3 hours after the trigger we infer a decay slope of 0.62$\pm$0.05 consistent with the measurement in \citet{kruhler08} at the same time. We infer a celestial position of $\mathrm{R.A. (J2000)} = 02^{\rm h} 26^{\rm m} 35\fs77$, $\mathrm{decl. (J2000)} = -55\arcdeg 31\arcmin 39\farcs2$ (calibrated against USNO-B1.0). Following the afterglow detection, spectroscopy was immediately started, using the FORS2 instrument equipped with the grism 300V ($\sim 10$~{\AA} FWHM spectral resolution). A total of 5400~s of exposure were acquired, split into three exposures, with mean epoch of Aug 2.38935 UT (2.2 hours after the burst). The spectra were obtained in excellent seeing conditions
308: (about $0\farcs5$) and a relatively small airmass of 1.2. We used a $1\farcs0$
309: slit and therefore slitloss should be neglicible. The spectra were flux calibrated
310: using a spectrum of the spectrophotometric standard star LTT1020 obtained on the
311: same night.
312: \input{tab1}
313:
314: The optical spectrum of the afterglow is shown in Fig.~\ref{fig:spectrum}.
315: The afterglow spectrum is characterized by a red continuum
316: with a broad dip centered at $\lambda_{\rm obs}\approx 7500$ \AA. There is
317: a break in the spectrum at $\lambda_{\rm obs}\approx 4250$ \AA\ blueward of
318: which little residual flux, if any, is detected. We identify the latter
319: feature as the imprint of the Ly$\alpha$ forest of absorption lines at
320: $z_{\rm abs}<2.5$. Numerous narrow absorption lines are detected
321: throughout the spectrum (see Table~\ref{tab:lines}). The highest redshift absorber
322: is a strong metal line system at $z_{\rm abs}=2.4549$, which we adopt as the GRB
323: host galaxy redshift. In addition we detect two intervening \ion{Mg}{2} absorption
324: systems at $z=2.078$ and $z=2.292$. The system at $z=2.078$ is very strong
325: with a restframe equivalent width of the 2796 \AA\ line of $W_r=3.9$~\AA.
326: \begin{figure}
327: \includegraphics[angle=0,width=\columnwidth,clip=]{f1.eps}
328: \caption{The flux-calibrated spectrum of the GRB\,070802 afterglow vs. observed wavelength.
329: Metal lines at the host redshift are marked with solid lines whereas the lines
330: from the two intervening systems are marked with dotted lines. The broad
331: depression centered around 7500 \AA\ is caused by the 2175 \AA\ extinction
332: bump in the host system at $z_{\rm abs}=2.4549$. The cleaned spectrum used for
333: the extinction curve analysis in \S~\ref{sec:ext} is overplotted in red. Telluric
334: features are indicated with $\earth$.}
335: \label{fig:spectrum}
336: \end{figure}
337: \input{tab2}
338:
339: \subsubsection{\ion{H}{1} content and metal lines}
340: We measured the total neutral hydrogen column density (in units of cm$^{-2}$
341: throughout) of the absorber at
342: the GRB redshift by fitting a damped Ly$\alpha$ absorption line
343: profile fixed at the redshift of the detected metal lines.
344: Both the saturated core and the red damped wing of the
345: Ly$\alpha$ line are constrained by the residual flux at $\lambda_{\rm obs}
346: \ge 4250$ \AA.
347: We derived $\log N($\ion{H}{1}$)=21.5\pm 0.2$ (see Fig.~\ref{fig:HI}).
348:
349: \begin{figure}
350: \includegraphics[angle=0,width=\columnwidth,clip=]{f2.eps}
351: \caption{A section of the spectrum centered on the position of Ly$\alpha$
352: at $z_{\rm abs}=2.4549$. Overlaid is the best fitting DLA profile and associated 1~$\sigma$ uncertainty,
353: corresponding to $\log{N}$(\ion{H}{1})$=21.5\pm0.2$.}
354: \label{fig:HI}
355: \end{figure}
356:
357: As listed in Table~\ref{tab:lines},
358: a large number of metal species are identified at the GRB
359: redshift.
360: The detected lines from singly ionized
361: elements are very strong compared to other GRB- or QSO-DLA
362: sightlines. For instance, $W_r=2.9$~\AA\ for \ion{Si}{2}\,$\lambda1526$ is larger than for any LBG/GRB/QSO-DLA observed to date, hinting at a
363: high metallicity \citep{prochaska2008}. The restframe equivalent width of \ion{Si}{2}\,$\lambda1808$, $W_r=1.13$ \AA\, is
364: also the largest ever detected in any extragalactic
365: sightline. Assuming the latter line is located on the linear part of
366: the curve of growth implies that $\log N($\ion{Si}{2}$)=16.3$. This
367: is a very conservative lower limit to the actual column density and
368: translates into a conservative lower limit on the metallicity of
369: [Si/H$]>-0.8$. Note also that Si could be depleted onto dust grains
370: \citep{petitjean2002}.
371:
372: We further performed Voigt-profile fitting (see Table~\ref{tab:voigt}) using
373: some of the weakest (least saturated) metal lines (see Fig.~\ref{fig:metals}).
374: The absorption line at $\lambda _{\rm obs}\approx 7000$ \AA\ is likely to be a
375: blend of \ion{Zn}{2} and \ion{Mg}{1}\,$\lambda$2026 with possibly similar
376: strengths. It is therefore important to fit both transitions simultaneously.
377: The other \ion{Zn}{2} ($\lambda$2062) line is also blended (with one of the
378: \ion{Cr}{2} triplet lines). Therefore, the best line to estimate the
379: metallicity remains \ion{Si}{2}\,$\lambda$1808. All lines were simultaneously
380: fitted with a single component constraining their redshifts as well as their
381: broadening parameter (assumed to be purely turbulent) to have identical values.
382: We find [Si/H$]=-0.46\pm 0.38$ and also consistently get [Zn/H$]=-0.50\pm 0.68$.
383: Assuming the \ion{Zn}{2}\,$\lambda$2026 line is located on the linear part of the
384: curve-of-growth implies that [Zn/H]>-0.35 but the actual metallicity could
385: be lower than that due to the unknown contribution of the \ion{Mg}{1}\,$\lambda$2026 line
386: to the observed feature (see above).
387: \input{tab3}
388: \begin{figure}
389: \includegraphics[angle=0,width=\columnwidth,clip=]{f3.eps}
390: \caption{Our Voigt-profile fits to low-ionization metal lines.
391: Best-fitting parameter values are given in Table~\ref{tab:voigt}.}
392: \label{fig:metals}
393: \end{figure}
394:
395:
396:
397: The large error bars on the measured column densities are mainly a
398: consequence of the large uncertainty on the broadening parameter $b$
399: which we find to be: $b=123\pm 74$ km s$^{-1}$.
400: One might argue that $b$ could be significantly smaller
401: (e.g., $b<30$ km s$^{-1}$) in which case the column densities would be
402: much larger than found in our best fit. However, it has been shown
403: that the width of metal line profiles in DLAs is larger at higher
404: metallicities \citep{2006A&A...457...71L,prochaska2008}. Therefore,
405: a large $b$ value is consistent with the relatively high
406: metallicity inferred, which is among the highest measured for
407: any GRB \citep{2006A&A...451L..47F}.
408:
409: The measured abundance ratios are consistent with what is seen in
410: other GRB- or QSO-DLAs. We find that the mean depletion of Fe-peak
411: elements (Fe, Cr and Mn) compared to Si is $\sim 0.5$ dex. This is
412: not particularly high, but yet significant implying that
413: dust is present in this
414: system. According to our best fit, the Si metallicity could be as
415: large as solar. This in fact is quite likely taking into account that
416: Si is affected by dust depletion \citep{petitjean2002}.
417:
418: \subsubsection{Neutral species}
419: In the afterglow spectrum, we clearly identify neutral carbon lines,
420: namely \ion{C}{1}\,$\lambda\lambda$1560,1656 (blended with
421: \ion{C}{1}$^\star$ and \ion{C}{1}$^{\star\star}$; see
422: Fig.~\ref{fig:metals}). These are the first ever clear detections of \ion{C}{1}
423: in a GRB-DLA which typically have very low \ion{C}{1}$/$\ion{C}{2} ratios \citep{2007ApJ...666..267P}. The two features are very strong. We measured
424: $W_r=1.4$ \AA\ at $\lambda_{\rm obs}\approx 5720$ \AA\ which is dominated
425: by \ion{C}{1}\,$\lambda$1656 absorption. Using the above fit results,
426: we get log N$($\ion{C}{1}$)/N($\ion{Si}{2}$)=-1.65$. This can be
427: compared to QSO-DLAs from Fig. 10 of \citet{srianand2005} showing
428: that H$_2$ is most likely present as well in this system, although this can not be verified as it falls too much in the blue into the Ly$\alpha$ forest, and because of low signal to noise and spectral resolution.
429: On the other hand, the strength of the detected \ion{C}{4} doublet
430: line is only modest and similar to most other GRB sight-lines.
431:
432:
433: \subsubsection{Non detection of \ion{Fe}{2}$^\star$ in \grb.}
434: Over the past few years, excited lines of \ion{Fe}{2} have been detected along
435: several GRB sightlines \citep[e.g.,
436: ][]{2004ApJ...614..293S,2005ApJ...634L..25C}, which have been shown to be due
437: to indirect excitation by UV photons produced by the GRB afterglow
438: \citep{2006ApJ...648...95P,2006ApJ...648L..89D,2007A&A...468...83V}. For two
439: specific GRBs, modeling of the afterglow flux exciting a cloud of \ion{Fe}{2}
440: and/or \ion{Ni}{2} atoms has led to an estimate of the distance between the GRB
441: and the bulk of the neutral absorption material (responsible for low-ionization
442: lines from ions such as \ion{C}{1}, \ion{O}{1}, \ion{Fe}{2}, \ion{Cr}{2},
443: \ion{Mn}{2}, \ion{Si}{2}, \ion{Zn}{2}, \ion{Ni}{2}): ~2 kpc for GRB 060418
444: (Vreeswijk et al. 2007), and ~0.5 kpc for GRB 050730 (Ledoux et al. 2009, submitted). For the GRB 070802 sightline we detect consistent features at the expected positions of the \ion{Fe}{2}$^*\lambda$2396 and
445: \ion{Fe}{2}$^{**}\lambda$2405 lines at z=2.455 but, since the lines are likely to be at the same time
446: slightly saturated and blended, we cannot infer from them secure column densities.
447: Based on the distance estimates in previous GRBs we
448: note that the bulk of the neutral material, which likely corresponds to the
449: region where the extinction bump is originating, is probably at least 0.5 kpc
450: away from GRB 070802.
451:
452: \subsubsection{Constraints on the $z=2.078$ foreground absorber extinction}
453: \label{sec:foreground}
454: The foreground absorption system at $z=2.078$ exhibits
455: strong \ion{Mg}{2} ($W_r^{\lambda 2796}=3.9$ \AA) and \ion{Fe}{2} lines with
456: equivalent width ratio $W_r^{\lambda 2796}/W_r^{\lambda 2600}\la 2$.
457: Additionally the detected \ion{Mg}{1}\,$\lambda$2852 line has $W_r=1.0$ \AA. These
458: characteristics make this absorber very likely a DLA
459: \citep[see, e.g., ][their fig.~11]{2006ApJ...636..610R}. We note that such a strong \ion{Mg}{2} absorber is very rare for QSO sightlines, but it has been suggested that GRB sightlines may have stronger absorbers although the reason for this is still unknown \citep{2006ApJ...648L..93P} .
460:
461: To constrain $A_V$ from the foreground absorber we look at the metal lines as metallicities and dust depletion factors (and hence the dust content) correlate in DLAs \citep{2003MNRAS.346..209L}.
462: The non-detection of \ion{Zn}{2}\,$\lambda$2026 leads to an upper limit on
463: the zinc column density.
464: Unfortunately, the constraint we can derive from the spectrum,
465: $\log N($\ion{Zn}{2}$)<13.5$ at the $3\sigma$ level, is fairly weak
466: because there is an absorption feature immediately bluewards of the
467: expected position of the \ion{Zn}{2} line.
468: Moreover, we cannot use \ion{Si}{2} instead of \ion{Zn}{2} because the expected
469: position of the \ion{Si}{2}\,$\lambda$1808 line at $z=2.078$ is blended with
470: \ion{Fe}{2}\,$\lambda$1608 from the GRB host galaxy absorber at $z=2.455$.
471: The best constraint on $A_V$ therefore comes from the observed \ion{Fe}{2} lines
472: at $z=2.078$. The ratio of the weakest detected \ion{Fe}{2} line, i.e.,
473: \ion{Fe}{2}\,$\lambda$2374, at $z=2.455$ to that at $z=2.078$ is 4.3. Because
474: of saturation effects, this value is a lower limit on the ratio of the
475: \ion{Fe}{2} column densities. Altogether, the weaker \ion{Fe}{2} lines
476: at $z=2.078$ and the non-detection of \ion{Zn}{2}\,$\lambda$2026 at $z=2.078$ are an indication of lower metallicity because in DLAs the dust depletion
477: factor is only a slowly decreasing function of metallicity \citep[e.g., ][]{2003MNRAS.346..209L,2008A&A...481..327N}.
478: A factor of $>4.3$ leads to $\log N($\ion{Zn}{2}$)<13$ and therefore $A_V<0.25$
479: \citep[][their fig.~1]{2005A&A...444..461V}.
480:
481: An alternative way to estimate the contribution of the foreground absorber is to us the relation reported by \citet[][their eq. (18)]{menard2008} which uses the equivalent width of the \ion{Mg}{2} line to estimate $E(B-V)$. Assuming a standard MW extinction law, with $R_V=3.1$, this translates into $A(V)\approx 0.06$ which is well below the limit derived above.
482:
483:
484: \subsection{X-ray observations}
485: The X-ray telescope onboard \emph{Swift} observed the afterglow of
486: GRB\,070802 beginning observations in photon counting mode at 147\,s
487: after the trigger time ($t_0 =$ 07:07:26 on 2 August 2007). The data
488: were extracted and reduced in a standard way using the HEAsoft
489: software (version 6.2) and the most recent calibration files. The
490: X-ray spectrum (with a total exposure time of 3419\,s in the interval
491: $t_0+147$--$t_0+16800$\,s) was extracted and fit with a power-law with
492: absorption fixed at the Galactic level ($2.9\times10^{20}$\,cm$^{-2}$) \citep{1990ARA&A..28..215D}
493: and freely variable absorption at the redshift of the host galaxy. The
494: best fit resulted in a power-law with a photon spectral index
495: $\Gamma = 2.02^{+0.17}_{-0.15}$ ($\beta = 1.02^{+0.17}_{-0.15}$, $68\%$ confidence level)
496: and an equivalent
497: absorbing hydrogen column density
498: of $9.7_{-0.4}^{+0.6}\times10^{21}$\,cm$^{-2}$ at $z=2.45$ assuming solar abundances. To
499: produce the X-ray segment of the near-infrared to X-ray spectral
500: energy distribution (SED), this X-ray spectrum was normalized to the
501: flux level obtained from the X-ray lightcurve at the SED time
502: ($t_0+3517$\,s) and corrected for both absorptions. No significant evidence
503: (above the $1\,\sigma$ level) was found for spectral variations over the period
504: of the spectrum, by comparing the first 400\,s and the remainder of the observation.
505:
506:
507: \subsection{The host galaxy}
508: $K$- and $R$-band observations of the host galaxy of GRB 070802 were obtained with the ESO VLT during the nights of 27 September 2007 and 5 October 2007
509: (57 and 65 days after the GRB). At the afterglow position, a source is clearly
510: detected with $R = 25.03 \pm 0.10$ (Fig.~\ref{fig:finding}).
511: A faint source is also visible in $K$, with magnitude $K = 21.70 \pm 0.25$.
512: The centroid of the host is offset by $0.15 \pm 0.04$ arcsec with respect to
513: the afterglow position (1.2~kpc at $z = 2.455$). From the observed $K$
514: magnitude (rest frame 6270~\AA), and assuming a power-law spectrum consistent
515: with the $R-K$ color ($f_\nu \propto \nu^{-1.3}$), we can estimate the
516: absolute $B$-band magnitude of the host to be $M_B = -21.0$ (uncorrected
517: for extinction) which is fairly luminous for a GRB host \citep{2006Natur.441..463F,2008arXiv0803.2718S}.
518:
519: \begin{figure}
520: \includegraphics[angle=0,width=\columnwidth,clip=]{f4.eps}
521: \caption{The field of \grb. The left panel shows the afterglow as detected about 1~hr after the burst. The right panel shows a deep late-time exposure of the same field (65 days after the burst), taken when the afterglow had faded away, and revealing a host galaxy at the GRB position. Each panel is $30\arcsec \times 30\arcsec$ wide.}
522: \label{fig:finding}
523: \end{figure}
524:
525:
526:
527: \section{The extinction curve}
528: \label{sec:ext}
529: The spectrum shows strong extinction features, with a clear detection of a bump
530: near $\lambda=2175$~\AA\ in the restframe of the host (Fig.~\ref{fig:spectrum}). To extract the continuum part of the spectrum to use for the extinction curve fitting we start by selecting the wavelength range $4500$--$9000$\AA. We then remove emission and absorption lines by recursively fitting the spectrum to an 8th order polynomial removing parts which deviate by more than $2.2\sigma$ (see Fig.~\ref{fig:spectrum}).
531:
532:
533: The intrinsic flux of the afterglow is modeled as
534: \begin{equation}
535: f_\nu^{\mathrm{intr}}=f_{\nu,0} \left(\frac{\nu}{\nu_0}\right)^{-\beta}
536: \end{equation}
537: (where $\nu$ is frequency, $\beta$ the intrinsic slope and $f_{\nu,0}$ is the flux at $\nu_0$) but is partially extinguished due to dust along the line of sight.
538: The observed flux is therefore given by
539: \begin{eqnarray}
540: f_\nu^{\mathrm{obs}}&=&f_\nu^{\mathrm{intr}} 10^{-0.4A_\lambda^{\mathrm{total}}}\\
541: &= &f_{\nu,0} \left(\frac{\nu}{\nu_0}\right)^{-\beta} 10^{-0.4\left(A_\lambda+A_\lambda^{\mathrm{Gal}}\right)}
542: \end{eqnarray}
543: where the total extinction is $A_\lambda^{\mathrm{total}}=A_\lambda+A_\lambda^{\mathrm{Gal}}$,
544: $A_\lambda$ is the extragalactic extinction along the line of sight as a function of the wavelength $\lambda$ and $A_\lambda^{\mathrm{Gal}}$ is the extinction in the Milky Way. The extragalactic extinction can be caused by both extinction in the host of the GRB and in foreground objects, i.e. $A_\lambda=A_\lambda^{\mathrm{host}}+A_\lambda^{\mathrm{fore}}$. The observed flux is corrected for extinction in the Milky Way using the maps of \citet[][$E(B-V)=0.025$, $R_V=3.1$]{schlegel1998}, leaving
545: \begin{equation}
546: f_\nu=f_{\nu,0} \left(\frac{\nu}{\nu_0}\right)^{-\beta} 10^{-0.4 A_\lambda}
547: \end{equation}
548: which is the intrinsic flux of the burst affected only by extragalactic extinction along the line of sight. Solving for the extinction, we find
549: \begin{eqnarray}
550: A_\lambda&=&-2.5\log_{10}\left(\frac{f_\nu}{f_{\nu,0}}\left(\frac{\nu}{\nu_0}\right)^\beta\right)\\
551: &=&-2.5\log_{10}\left(\frac{f_\nu}{f_{\nu,0}}\left(\frac{x}{x_{0}}\right)^\beta\right)
552: \end{eqnarray}
553: where $x\equiv1/\lambda=\nu/c$ is the wavenumber.
554: The absolute extinction along the line of sight can therefore be determined if the underlying spectral slope $\beta$ and the normalization $f_{\nu,0}$ at $x_0$ are known.
555:
556: The simple shape of the intrinsic spectra of GRB afterglows (see \S~\ref{sec:grb_after}) allows us to constrain both $\beta$ and the normalization. Assuming that there is no cooling break in the intrinsic spectrum, one may use the slope and normalization derived from the X-ray data, i.e. $\beta=\beta_X$. Alternatively, assuming that there is a cooling break in the spectrum, the slope must be $\beta=\beta_X-0.5$ while the normalization depends on the location of the cooling break. These two possibilities will both be addressed here and are depicted in Fig.~\ref{fig:xraynobreak}. In addition, we will start by assuming that all the extinction is caused by dust in the host galaxy of \grb, i.e. that
557: \begin{equation}
558: A_\lambda=A_\lambda^{\mathrm{host}}.
559: \end{equation}
560: The possible contribution from the foreground \ion{Mg}{2} absorption systems (\S~\ref{sec:foreground}) will be addressed in \S~\ref{sec:ext_fore}.
561: \begin{figure}
562: \includegraphics[angle=270,width=\columnwidth,clip=]{f5.eps}
563: \caption{Spectral energy distribution of the afterglow of \grb\ at $t_0+3517$~s. The plot shows X-ray data (triangles, corrected to the time frame of the optical photometric points) and the derived intrinsic powerlaw spectrum (black solid line) and the 68\% confidence levels (black dashed lines). The spectrum (red curve) is scaled to the $R$-band. The squares are the photometric points from this paper, the diamonds are the photometric points of \citet{kruhler08}. Assuming that the intrinsic slope of the GRB can be described as a single powerlaw (i.e. $\beta=\beta_X=1.02_{-0.15}^{+0.17}$), the difference in the observed spectrum and the extrapolated X-ray spectrum is interpreted as being due to extinction. The blue lines correspond to the upper and lower limits of the intrinsic power law given a cooling break in the spectrum. The intrinsic slope in the optical is then $\beta=\beta_X-0.5=0.52_{-0.15}^{+0.17}$ and the normalization is determined by the location of the cooling break, which can be anywhere between the optical and X-ray data sets with the additional constraint that the optical data points can not be brighter than the intrinsic curve. The intrinsic powerlaw spectrum with a cooling break corresponding to the 68\% confidence levels have been omitted for clarity.}
564: \label{fig:xraynobreak}
565: \end{figure}
566:
567: We fit the derived extinction curves to four types of extinction laws, i.e. MW type extinction as parametrized by \citet{cardelli1989}, LMC and SMC type extinction as parametrized by \citet{pei1992} and the modified parametrization for the UV of \citet[][FM]{fitzpatrick2007}. A description of these extinction laws and how the \citet{pei1992} parametrization compares to the \citet{gordon2003} SMC and LMC extinction curves is given in Appendix~\ref{app:ext}. The parametrization of \citet{fitzpatrick2007} does not directly tie the slope to the bump, but keeps the parameters of each independent. It therefore has much greater freedom in fitting the curve and is able to accurately trace the derived extinction. The parametrization however assumes that $A_V$ and $R_V$ have been independently derived from the optical extinction. As our data do not reach into the restframe optical, we have opted to fix $R_V=3.1$ and allow $A_V$ to vary while fitting the UV. We also tested choosing $R_V=2.7$ and $R_V=3.5$ and find that it leads to a change in $A_V$ and $c_1$ (one of the parameters of the FM extinction law, see App.~\ref{app:ext}) of around $10\%$ but does not significantly affect the other parameters.
568:
569: We compare the extrapolation of our fits to the infrared photometric points of \citet{kruhler08}. We used the spectral energy distribution (SED) from GROND created by
570: citet{kruhler08} to derive the near-infrared data we show here simply by
571: extrapolating the mean SED time of \citet{kruhler08} to our mean SED time using
572: our estimate of the optical decay rate. We note that these points are not used in constraining the fits.
573:
574: \subsection{The extinction assuming no cooling break}
575: If there is no cooling break between the X-ray and optical regimes,
576: clearly the X-ray and optical emission lie on the same power-law.
577: Under this assumption, the intrinsic slope and normalisation in the
578: optical can therefore be derived from a simple extrapolation of the
579: X-ray data. Any discrepancy between the observed optical flux and the
580: extrapolation is then interpreted as extinction along the line of
581: sight. A slight complication is introduced by the non-simultaneity of
582: the observations in different bands, since the flux decays with time.
583: However the uncertainty introduced by this correction is not large,
584: since the optical photometry mean times are $I=t_0+3517$\,s,
585: $V=t_0+3627$\,s and $R$ is interpolated between observations at
586: $t_0+3249$\,s and $t_0=3949$\,s, and the X-ray flux is derived from a
587: lightcurve that covers this time period.
588:
589: Fig.~\ref{fig:xraynobreak}
590: shows the X-ray spectrum (scaled to the SED time, $t_0+3517$) and its
591: extrapolation into the optical. The fluxed optical spectrum (mean
592: time $t_0+2.2$\,hours) was scaled to the $R$-band data point. It is
593: worth noting that the scaled optical spectrum agrees with the observed
594: $V$ and $I$ photometric data, indicating that the small temporal
595: extrapolations are correct and that there is no significant spectral
596: change in the optical in this period. The extinction curve, shown in
597: the top panel of Fig.~\ref{fig:extnobreak}, is then derived from
598: the difference between the spectrum and the extrapolated slope.
599: \begin{figure}
600: \includegraphics[angle=0,width=\columnwidth,clip=]{f6.eps}
601: \caption{The derived extinction curve (black line) of the \grb\ host galaxy. The curve has been smoothed for clarity in presentation. The curve has been fit by an SMC (blue long-dashed line), LMC (cyan dotted line), MW with $R_V$ fixed (red solid line) and allowed to vary (red dashed line) and finally using the FM parametrization (yellow dash-dot line). The filled diamonds are the photometric points presented in this paper used to scale the spectrum while the empty diamonds are the photometric points of \citet{kruhler08}. Note that the fits were done using only the spectroscopic data. {\it Top panel:} The derived extinction curve assuming no cooling break between the optical and X-rays in the intrinsic spectrum of the burst, i.e. $\beta=\beta_X=1.02$. The best fit is obtained using the FM parametrization. Of the three local extinction laws (MW, SMC, LMC), the LMC provides the best fit. The photometric points in the infrared are in poor agreement with the extrapolations of the fits and would require a break in the extinction curve at $x\lesssim1.5$~$\mu$m$^{-1}$. {\it Middle panel:} The same as the top panel, but using the 1~$\sigma$ deviation value for $\beta$ towards a shallower intrinsic slope, $\beta=0.87$. In this case, the photometric points are in better agreement with the predicted extinction curve and the sharp break in steepness at around $x\lesssim1.5$~$\mu$m$^{-1}$ is not required. However, it implies $\beta_X<1$. {\it Bottom panel:} The derived extinction curve (black line) assuming a cooling break in the intrinsic spectrum, i.e. $\beta=\beta_X-0.5=0.52$, and shifted by a constant $A_{\infty}$ which depends on the location of the cooling break. The FM parametrization gives the best fit to the data while the LMC provides the best fit amoung the three local extinction laws. The photometric points of \citet{kruhler08} are inconsistent with the extrapolation of the SMC and MW (with a fixed $R_V=3.1$), but are in agreement with the other fits.}
602: \label{fig:extnobreak}
603: \end{figure}
604:
605: To study the properties of the extinction curve, we fit it to the four previously mentioned extinction laws
606: with the additional constraint that the extinction must go to zero for infinite wavelengths. The derived extinction and the fits are shown in the top panel of Fig.~\ref{fig:extnobreak} and the parameters of the fit are given in Tables~\ref{tab:extfits} and \ref{tab:extfitsfm}. The freedom of the FM parametrization allows for a good fit ($\chi^2/\mathrm{d.o.f.}=261/968$), the LMC also provides a reasonable fit ($\chi^2/\mathrm{d.o.f.}=815/975$), while the SMC and MW fail to reproduce the extinction curve. However, the infrared photometry points of \citet{kruhler08} clearly do not agree with any of the fits and require the extinction curve to take a very sharp break in its steepness at $x<1.5$~$\mu$m$^{-1}$ which would suggest that the host of \grb\, has a very strange form of extinction law. However, a more plausible explanation is that the assumed intrinsic slope is wrong, either due to the extrapolation or because of the presence of a cooling break in the spectrum (see \S~\ref{sec:extbreak}).
607: \input{tab4}
608: \input{tab5}
609:
610: From Fig.~\ref{fig:xraynobreak} we see that due to the several order of magnitudes extrapolation, the 1~$\sigma$ difference in $\beta$ translates into a difference of $\sim$1 magnitude in the optical. While going towards steeper slopes will make the apparent 'break' in the extinction curve greater, a shallower slope will act to reduce it as seen in Fig.~\ref{fig:extnobreak}. Redoing the analysis for a shallower slope corresponding to 1~$\sigma$ deviation (i.e. $\beta=0.87$) results in the extinction curve shown in the middle panel of Fig.~\ref{fig:extnobreak}. The parameters of the fit, for both 1~$\sigma$ deviations are given in Tables~\ref{tab:extfits} and \ref{tab:extfitsfm}. The goodness of fit for the MW, SMC and LMC is improved for the shallower slope, while the effect on the FM fit is minimal. The fits are also more consistent with the additional photometric points in the infrared of \citet{kruhler08}. We note however that most GRBs have $\beta_X > 1$ \citep[see e.g.,][ their fig.~3, with $\Gamma=\beta+1$]{2007ApJ...670..565L}.
611:
612:
613: \subsection{The extinction curve assuming a cooling break}
614: \label{sec:extbreak}
615: A cooling break in the intrinsic spectrum occurring between the X-ray and the optical data, i.e. $\beta=\beta_X-0.5$, would result in an intrinsic spectrum at optical wavelengths of the form
616: \begin{equation}
617: f^{\mathrm{intr}}_{\nu}=\tilde{f}_{\nu,0}\left(\frac{x}{x_0}\right)^{-\beta}=\tilde{f}_{\nu,0}\left(\frac{x}{x_0}\right)^{-(\beta_X-0.5)}
618: \end{equation}
619: where the normalization $\tilde{f}_{\nu,0}$ depends on the location of the cooling break, $x_{\mathrm{break}}$, by
620: \begin{equation}
621: x_{\mathrm{break}}=x_0\left(\frac{f_{\nu,0}}{\tilde{f}_{\nu,0}}\right)^{2}.
622: \end{equation}
623: The absolute extinction is then given by
624: \begin{eqnarray}
625: A_{\lambda}&=&-2.5\log_{10}\left(\frac{f_{\nu}}{\tilde{f}_{\nu,0}}\left(\frac{x}{x_{0}}\right)^{\beta}\right)\\
626: &=&-2.5\log_{10}\left(\frac{f_\nu}{f_{\nu,0}}\left(\frac{x}{x_0}\right)^{\beta}\right) - A_{\infty}.
627: \end{eqnarray}
628: The first term of the sum can be determined as before from the X-ray data, while the second part of the sum
629: \begin{equation}
630: A_{\infty}\equiv2.5\log_{10}\frac{f_{\nu,0}}{\tilde{f}_{\nu,0}}=2.5\log_{10}\left(\frac{x_{\mathrm{break}}}{x_0}\right)^{0.5}
631: \end{equation}
632: is an additional unknown to be added to the fits. Its effect is to shift the extinction curve up or down (i.e., the value of $A_\lambda + A_\infty$ at $x=0$ is $A_\infty$). Constraining the break to occur between the optical and X-ray regimes translates into constraints on the allowed values of $A_\infty$ corresponding to setting $x_{\mathrm{break}}$ to be the upper $x$ of the optical data and the lower $x$ of the X-ray data. An additional requirement is that the extinction remains positive.
633:
634:
635: We fit the three local extinction laws (MW, SMC and LMC) to the curve in addition to the FM parametrization. Again we find that of the three local extinction laws the LMC provides the best fit ($\chi^2/\mathrm{d.o.f.}=311/974$), but the FM parametrization still gives the best overall fit ($\chi^2/\mathrm{d.o.f.}=253/967$). In this instance, the infrared photometric points of \citet{kruhler08} agree with the extrapolated LMC and FM fits.
636:
637:
638: \subsection{Possible contribution from a foreground object}
639: \label{sec:ext_fore}
640: The MW extinction law provides a poor fit to the observed extinction, in part because the observed bump is shallower than the predicted MW bump and the rise into the UV is steeper (see e.g. Fig.~\ref{fig:extnobreak}). However, as Fig.~\ref{fig:ext_effect_foreground} shows, such effects can be caused if part of the extinction is due to a foreground object without a bump.
641: \begin{figure}
642: \includegraphics[angle=0,width=\columnwidth,clip=]{f7.eps}
643: \caption{The effect on an "observed" extinction curve assuming that the extinction is caused by a MW type extinction in addition to an SMC foreground object (located at $z=2.45$ and $z=2.08$ respectively). The three lines correspond to $A_V=1.0$, $A_V^{\mathrm{fore}}=0.0$ (solid line), $A_V=0.75$, $A_V^{\mathrm{fore}}=0.25$ (dotted line) and $A_V=0.5$, $A_V^{\mathrm{fore}}=0.5$ (dashed line) where the $A_V$ values are quoted in their restframes.}
644: \label{fig:ext_effect_foreground}
645: \end{figure}
646:
647:
648: The spectrum of \grb~shows two intervening \ion{Mg}{2} absorption systems which might also cause part of the reddening. Neither foreground system shows any signs of a $2175$~\AA\ bump. If the foreground absorber had a $2175$~\AA\ bump in its restframe, it would act to broaden and shift the peak of the observed bump in the restframe of the host, which is inconsistent with the data. Therefore, the most straightforward way to model them is with an SMC type extinction law. As this extinction law is roughly linear, there is no way of separating the effect of two systems without additional constraints on the dust content of the two systems. As the absorber at $z=2.29$ has much weaker absorption lines, and therefore by assumption much weaker extinction, we will only look at the possible contribution from the system at $z=2.078$.
649:
650: We will therefore limit ourselves to studying the possibility of the extinction being of the form
651: \begin{equation}
652: A_\lambda=A_\lambda^{\mathrm{host}}+A_\lambda^{\mathrm{fore}}
653: \end{equation}
654: where $A_\lambda^{\mathrm{host}}$ is the extinction in the host galaxy and $A_\lambda^{\mathrm{fore}}$ is the extinction in the foreground \ion{Mg}{2} absorber. We have already seen that the extinction curve is poorly fit by an SMC type extinction, as it lacks the characteristic bump. We will therefore parametrize the extinction of the host galaxy of \grb\, as either a MW (with fixed $R_V=3.1$) or an LMC type, while the foreground absorber will be parametrized as an SMC type. In addition, we place the limit $A_V^{\mathrm{fore}}<0.25$ for the foreground absorber based on the discussion in \S~\ref{sec:foreground}.
655:
656: The results are shown in Fig.~\ref{fig:extnobreak_double} and tabulated in Tables~\ref{tab:extfits} and \ref{tab:extfitsbreak}. We find that for the LMC, the addition of the foreground contributor does not improve the fits and the best fits are found by setting $A_V^{\mathrm{fore}}$ to be zero. For the MW, the fits are improved and the best fit is found by setting $A_V^{\mathrm{fore}}$ to its maximum value of $0.25$. The MW fits are still worse than the LMC leading us to conclude that the extinction of \grb~is not well fitted by a MW type extinction.
657: \begin{figure}
658: \includegraphics[angle=0,width=\columnwidth,clip=]{f8.eps}
659: \caption{Extinction curve fits taking into account a possible contribution from the foreground absorber at $z=2.078$. The extinction of the host is taken to be a MW (red solid line) or an LMC (cyan dotted line). The extinction of the foreground absorber is taken to be an SMC (as there is no evidence of a bump in the data) with an upper limit of $A_V^{\mathrm{fore}}\le0.25$ (see \S~\ref{sec:foreground}). The filled diamonds are the photometric points presented in this paper used to scale the spectrum while the empty diamonds are the photometric points of \citet{kruhler08}. Note that the fits were done using only the spectroscopic data. The top panel shows the derived extinction assuming no cooling break. The middle panel shows the 1~$\sigma$ deviation towards a shallower intrinsic slope. The bottom panel shows the derived extinction assuming a cooling break. For the LMC fits the contribution of the foreground extinction is not found to improve the fits (setting $A_V^{\mathrm{fore}}=0$). The MW fits are improved by adding an SMC contribution, however, they are still worse than LMC only, and would require $A_V^{\mathrm{fore}}>0.25$.}
660: \label{fig:extnobreak_double}
661: \end{figure}
662:
663: \input{tab6}
664:
665: \subsection{Summary of the extinction curve properties}
666: Fig.~\ref{fig:ext_clear} shows the scaled extinction curve for \grb\ for $\beta=0.87$.
667: We clearly detect the 2175~\AA\ bump in the extinction curve of \grb. The data sample the bump very well and its detection does not depend on whether we assume cooling breaks or not in the spectrum of the GRB. It is one of the most robust detections of the bump in extragalactic environments to date and currently the highest redshift detection at z=2.45 (corresponding to 2.5 Gyr assuming a flat universe with $\Omega_m=0.3$, $\Omega_\Lambda=0.7$ and $H_0=73$~km~s$^{-1}$~Mpc$^{-1}$).
668: \begin{figure*}
669: \begin{center}
670: \includegraphics[angle=0,width=0.6\textwidth,clip=]{f9.eps}
671: \end{center}
672: \caption{An absolute extinction curve for the afterglow \grb\ at $z=2.45$. It is based on the simplest model of the afterglow that is consistent with all the available data (single power-law spectrum, $\beta = 0.87$, corresponding to a $1$~$\sigma$ deviation from the best fit) and with the constraint that the extinction should be zero at $1/\lambda=0$. The extinction curve shows a clear bump at $2175$~\AA. Also shown are the best fits (to the spectroscopic data) for an LMC extinction (dashed red line) and MW extinction (dotted blue line), with the LMC clearly providing the better fit. To obtain the absolute extinction, A$_V$ was derived using a linear interpolation of the two datapoints closest to the restframe $V$-band. The filled diamonds are the photometric points presented in this paper used to scale the spectrum while the empty diamonds are the photometric points of \citet{kruhler08}. }
673: \label{fig:ext_clear}
674: \end{figure*}
675:
676: For the three local extinction laws, the LMC provides the best fit (i.e., it has the lowest $\chi^2$ per degree of freedom (d.o.f.)) to the shape of the derived extinction curve, regardless of whether we assume there is a cooling break or not between the X-rays and the optical, and whether we take into account a possible contribution from the foreground absorbers or not (see Tables~\ref{tab:extfits} and \ref{tab:extfitsbreak}). The SMC clearly provides a poor fit in all cases as the bump is completely missing, while the bump of the MW extinction law is too strong and the rise into the UV not steep enough. A foreground absorber with SMC type extinction could make the bump shallower and the UV rise steeper (see Fig.~\ref{fig:ext_effect_foreground}), however, the absorption would need to be larger than the upper limits placed on $A_V^{\mathrm{fore}}$ in \S~\ref{sec:foreground}.
677:
678: The best fits are obtained using the FM parametrization. This is not surprising, as it has more freedom in tracing the shape of the curve (see Appendix~\ref{app:ext}). The three parameters giving the shape of the bump are only weakly dependent on the assumed $\beta$ and whether we assume a cooling break or not (Table~\ref{tab:extfitsfm}). The bump is found to be centered at $x_c\approx4.6\pm0.1$~$\mu$m$^{-1}$, the width of the bump is found to be $\gamma\approx1.08\pm0.05$~$\mu$m$^{-1}$ while the 'strength' (i.e. its height above the linear extinction, see Appendix~\ref{app:ext}) is $c_3\approx2.7\pm0.1$. Therefore, both the area of the bump $\Delta_\mathrm{bump}\equiv \pi c_3/(2\gamma)$ and its maximum height above the linear extinction $E_{\mathrm{bump}}\equiv c_3/\gamma^2$ are well defined. The relative strength of the bump $A_{\mathrm{bump}}/A_V=c_3/(\gamma^2 R_V)$ \citep[as defined by ][]{gordon2003} is also well defined, although here the uncertainty is dominated by the uncertainty in $A_V$ (or equivalently $R_V$). We note that the value of $c_3$ is the same as the average $c_3=2.7\pm0.1$ that \citet{gordon2003} find for the LMC average sample while the width $\gamma$ is a bit wider compared to their $\gamma=0.93\pm0.02$~$\mu$m$^{-1}$ (although it is still within their scatter). The value of $c_3$ is however higher than \citet{gordon2003} find for the LMC2 sample ($c_3=1.5\pm0.1$) although it is within the scatter.
679:
680: We estimate the amount of dust extinction to be $A_V=1.341\pm0.002$ for $\beta=1.02$ and $A_V=1.259\pm0.002$ for $\beta=0.52$ given the FM parametrization. The LMC fits correspondingly give $A_V=1.474\pm0.002$ for $\beta=1.02$ and $A_V=1.27\pm0.02$ for $\beta=0.52$. Taking into account the possible $1~\sigma$ deviation of the slope \citep[excluding steeper slopes than $\beta=1.02$, as this would be in strong disagreement with the photometric points of][]{kruhler08}, we estimate $A_V=0.8$--$1.5$ in the host along the line of sight to \grb.
681:
682:
683: \section{Discussion}
684: \label{sec:dis}
685: The nature of the interstellar extinction peak at 2175\,{\AA} remains
686: poorly understood more than 40 years after its discovery by \citet{stecher}.
687: As its detection has been limited to the Milky Way with only a few exceptions
688: it has proven hard to search for correlations between the dust environment
689: and the detection or non-detection of the bump. The detection of the bump
690: in the spectrum of \grb~is interesting in itself for two reasons, i.e., as
691: being the highest redshift detection of the bump and as being the first
692: robust detection of the bump in a GRB host galaxy. It shows that the carrier of the 2175\,{\AA} bump, which is characteristic for Milky Way
693: type dust, was in place 2.6 Gyr after the Big Bang (when the Universe was
694: only 20~\% of its current age). It also shows that the conditions for both forming and
695: not destroying the 2175\,{\AA} were satisfied in a GRB host galaxy -- surprising
696: in view of the fact that most GRB host galaxies are faint, blue, young,
697: low-metallicity galaxies
698: \citep{2003A&A...400..499L,2004A&A...425..913C,2006Natur.441..463F} in contrast to the massive, evolved and
699: chemically enriched Milky Way.
700:
701: Moreover, the detection of 2175\,{\AA} as well as a very rich spectrum
702: of redshifted UV metal absorption lines allows us to explore various
703: hypotheses for the origin of the bump -- an experiment that cannot
704: easily be done along lines of sight in the Milky Way. Below we explore
705: whether there are any other properties of the galaxy which are correlated
706: with the presence of the bump. We discuss what separates \grb~from other
707: GRBs with reddening but no detected bump and compare the properties of the
708: host galaxy with other galaxies for which the extinction curve has been
709: determined.
710:
711: \subsection{Metallicity}
712: As mentioned in \S~\ref{sec:2175},
713: based on evidence from the MW, LMC and SMC, the simplest hypothesis would be
714: that the strength of the 2175 \AA\ bump is simply controlled by the
715: metallicity. However,
716: \citet{savaglio2003} infer a metallicity of
717: [Zn/H]$ = -0.13\pm0.25$ for GRB\,000926 (compared to [Zn/H]$=-0.50\pm0.68$ we derive for \grb) and the \ion{H}{1} column density for GRB\,000926 is
718: similar to that of GRB\,070802. This would indicate that the extinction curve of GRB\,000926 should also have a bump if metallicity was its only driver. However, the extinction curve derived for
719: GRB\,000926 is inconsistent with those of the MW and LMC and fully consistent
720: with that of the SMC, i.e., without a 2175 \AA\ bump \citep{2001A&A...373..796F}.
721:
722:
723: \subsection{Gas to dust ratio}
724: \citet{gordon2003} suggested that the average extinction properties of the SMC, LMC and MW were described by the same underlying extinction law, which could be described by the $R_V$ parameter of the MW extinction and one additional parameter characterizing the strength of the bump and the UV slope. They suggested the gas to dust ratio as the second parameter and showed that for the LMC and SMC there is a correlation between the gas to dust ratio and the UV slope (i.e. $c_2/R_V$) and an anti-correlation between the gas to dust ratio and the strength of the bump (i.e. $ A_{\mathrm{bump}}/A_V=c_3/(\gamma^2 R_V)$), albeit with a large scatter.
725:
726: GRBs typically originate in host galaxies with low dust to gas ratios \citep[i.e. high gas to dust ratios][]{2003ApJ...597..699H,jakobsson2004,2004A&A...419..927V,2006ApJ...641..993K,2007ApJ...666..267P,2007MNRAS.377..273S}. While the column density, $N_H$, of the host of GRB\,070802 is not exceptionally large, it does have a very large $A_V$, leading to a low gas to dust. This is shown in Fig.~\ref{fig:n_H}, where we have plotted $N_H/A_V$ vs. the strength of the bump $A_{\mathrm{bump}}/A_V$ for the SMC, LMC, MW and \grb. Also plotted is the region where GRBs for which an extinction curve analysis exists in the literature lie. The plot shows that the anti-correlation between the bump strength and the gas to dust ratio found by \citet{gordon2003} for the SMC and the LMC also holds for the GRBs, and \grb\ in particular, although with large uncertainties in the values.
727: \begin{figure*}
728: \begin{center}
729: \includegraphics[angle=0,width=0.8\textwidth,clip=]{f10.eps}
730: \end{center}
731: \caption{ $N_H/A_V$ vs. $A_{\mathrm{bump}}/A_V$. The black filled circles are the averaged values for the SMC (bar and wing (only one line of sight)) and LMC (shell and average) from \citet{gordon2003}. The MW is marked with a star. \grb~is marked by a square, while other GRBs from the literature fall into the shaded region bounded by the arrows in the upper left corner. The $N_H/A_V$ values for the GRBs are calculated from the values quoted in Table~\ref{tab:n_H}. Their $A_{\mathrm{bump}}/A_V$ values are estimated to be lower than $0.3$ as they are all consistent with SMC type extinction (i.e. no sign of a bump in the SED). The values for \grb~are calculated directly from the parameters of the \citet{fitzpatrick2007} parametrization (see Table~\ref{tab:extfitsfm}). The error bars are dominated by the uncertainty in $A_V$ (and equivalently $R_V$), for which we have taken $A_V=1.3\pm0.5$ (or $R_V=3.1\pm1.2$). The $A_{\mathrm{bump}}/A_V$ for the MW is obtained by fitting an FM curve to a standard MW extinction (i.e. $R_V=3.1$) and $N_H/A_V=4.93/R_V\cdot10^{21}$~cm$^{-2}$~mag$^{-1}$ is taken from \citet{1994ApJ...427..274D}. We find that the trend suggested by \citet{gordon2003} of decreasing bump strength with increasing gas-to-dust ratio holds for the MW, \grb\ and the other GRBs from the literature.}
732: \label{fig:n_H}
733: \end{figure*}
734: \input{tab7}
735:
736: \subsection{Detection of \ion{C}{1} and the UV radiation field}
737: \label{sec:CI}
738: \grb\ is to our knowledge the first GRB to date to show
739: prominent \ion{C}{1} in its afterglow spectrum and also the first GRB to show the
740: 2175\,{\AA} feature, which is suggestive of a link between these two components.
741: Should such a link be established by
742: further observations it constitutes a constraint on the environment in which the
743: 2175\,{\AA} feature occurs. The ionization potential of
744: neutral carbon is lower than that of hydrogen
745: (11.3 vs.\ 13.6 eV) and neutral carbon is therefore only expected in regions
746: without intense UV radiation. This would be consistent with the general lack
747: of the 2175\,{\AA} feature in GRB host galaxies, as GRBs arise predominantly
748: from star-forming regions where the UV radiation is expected to be strong
749: \citep{2006Natur.441..463F, 2008arXiv0809.2608C}.
750:
751: To test whether this suggested correlation holds in general, we have looked at
752: the other three robust detections of the 2175\,{\AA} bump in individual
753: systems beyond the local group. For the lensing galaxy reported by
754: \citet{motta2002} and the intervening absorber towards GRB~060418 reported by \citet{2006MNRAS.372L..38E}, the existing data do not cover the wavelength range where
755: one would see the \ion{C}{1} absorption line if present. However, for the
756: intervening damped Lyman-$\alpha$ system toward AO~0235+164 reported by
757: \citet{junkkarinen2004} we see tentative evidence for the \ion{C}{1}\,$\lambda$1656
758: absorption line. The regions around \ion{C}{1}\,$\lambda$1560,$\lambda$1656 are shown in
759: Fig.~\ref{fig:26vs02} for GRB\,070802 and AO~0235+164. For comparison we also
760: show the same region in the afterglow spectrum of the high metallicity burst
761: GRB\,000926 which shows no significant evidence for the \ion{C}{1} absorption
762: features (and no bump in its extinction curve).
763:
764: It would be of interest to further check whether the sightlines to the MW, LMC and SMC for which extinction curve analysis exist are consistent with this correlation. Such a study is beyond the scope of this current work, but we note that a quick and incomplete search of the literature resulted in two more lines of sight consistent with this correlation. The first is a detection of \ion{C}{1} in the line of sight towards HD 185418 in the MW by \citet{2003ApJ...596..350S}. This line of sight is included in the sample of \citet{fitzpatrick+massa86} and has the 2175~\AA\ bump in its extinction curve. The second is a line of sight towards the SMC, which has \ion{C}{1} detected in the MW but only as an upper limit in the SMC \citep{1997ApJ...489..672W}. Although this sightline does not have an extinction curve analysis, it is consistent with our prediction that most lines of sight in the SMC should not show significant \ion{C}{1} absorption.
765:
766: The presence of the \ion{C}{1} absorption feature in GRB\,070802 suggests that the UV radiation field is weaker than in typical GRB environments (see \S~\ref{sec:CI}). \citet{gordon2003} reached the tentative conclusion that the shape of the extinction curve is affected by the UV flux density in the environment of the dust. In particular, a weaker UV flux density is found to correlate with the presence of the bump. Continuing our comparison to the bump-less GRB\,000926 host, we find evidence that GRB\,070802 does indeed have a weaker UV radiation field:
767: {\it i)} the GRB\,000926 absorption system has stronger high ionization
768: lines and much weaker lines from neutral species (e.g., \ion{C}{1},
769: see Fig.~\ref{fig:26vs02}) than the GRB\,070802 system.
770: {\it ii)} the host galaxy of GRB\,000926 appears to have a stronger
771: UV flux density as illustrated by the very strong Lyman-$\alpha$ emission
772: line \citep{fynbo2002}.
773: For GRB\,070802 we can exclude the presence
774: of such a strong Lyman-$\alpha$ emission line (Fig.~\ref{fig:spectrum} and Milvang-Jensen
775: et al., in preparation).
776: This supports the conclusion reached by \citet{gordon2003}.
777:
778: It is not immediately clear what the physical significance of the possible
779: correlation of strong \ion{C}{1} and the 2175\,\AA\ bump is. Given that the
780: bump is generally believed to be carried by carbonaceous material, and that
781: carbonaceous grain growth and formation requires free, neutral carbon and
782: molecules \citep{henning+salama}, it would not be surprising to find both observed properties in
783: the same environments. The simultaneous presence of \ion{C}{1} and the
784: 2175\,\AA\ bump as well as the lower overall ionization state of the gas,
785: relatively (though not exceptionally) high metallicity, and large
786: dust-to-gas ratio may be explained in a scenario in which the dust column is
787: strongly enriched by the presence of asymptotic giant branch (AGB) stars.
788:
789: For massive stars to move onto the AGB requires at least 600\,Myr and
790: typically much longer for a large population \citep{1992A&A...264..105M}.
791: The star-forming environment at such an age will be intrinsically relatively
792: benign, with a softer UV field, and one in which a large amount of dust and
793: molecular and free carbon is produced
794: \citep{andersen03,2004A&A...422..289G}. Furthermore the interstellar medium
795: (ISM) is likely to be reasonably metal-rich and dust-rich. These properties
796: are in contrast to the normal environments of GRB hosts which are typically
797: metal-poor. However some GRB hosts may be fairly metal-enriched \citep{2006A&A...451L..47F} but
798: still have hard radiation fields and young stellar populations \citep{2003A&A...400..499L,2004A&A...425..913C,2004ApJ...611..200P} and in particular, low
799: dust-to-gas ratios \citep{2006A&A...451L..47F,2006A&A...460L..13J,2007ApJ...660L.101W}.
800: This is consistent with the above scenario since the metal-enrichment timescale could well be much shorter than the
801: $\gtrsim10^9$\,yr required to have a reasonable number of AGB stars
802: producing dust \citep{2004MNRAS.351.1379S}. Such a scenario is then also
803: consistent with the fact that GRB\,070802 is the only GRB host galaxy so far
804: discovered with a 2175\,\AA\ bump. It should also be noted that the host
805: galaxy of \grb\ is fairly luminous and red for a GRB host
806: \citep{2008arXiv0803.2718S}, suggesting that it is a massive, evolved system, which would be in agreement with the claim of \citet{noll07} that
807: the presence of the bump requires an evolved population.
808: \begin{figure}
809: \includegraphics[angle=0,width=\columnwidth,clip=]{f11.eps}
810: \caption{A comparison of the region around \ion{C}{1} in the normalized spectra
811: of GRB~000926, GRB~070802, and AO~0235+164. The vertical lines show the
812: positions of the \ion{C}{1} lines.
813: The two GRB systems have nearly
814: identical \ion{Si}{2} line strengths, but GRB~000926 has stronger \ion{C}{4}
815: and no significantly detected \ion{C}{1}. \grb\ also has much stronger \ion{Al}{2} and \ion{Fe}{2}
816: than GRB~000926. AO~0235+164, which has a similar \ion{H}{1} column density as
817: the two GRB sightlines, displays both the 2175~{\AA} bump and significant \ion{C}{1}.}
818: \label{fig:26vs02}
819: \end{figure}
820:
821: \section{Conclusions}
822: \label{sec:con}
823: We have presented VLT observations of the afterglow of \grb. In a low resolution spectroscopy of the optical afterglow we detect a large number of strong metal lines from absorption systems at $z=2.078$, $z=2.292$ and $z=2.4549$. The highest redshift system is remarkable in showing very strong metal lines, e.g. with higher $W_r$ for the \ion{Si}{2} lines at 1526 and 1808 \AA\ than for any other known
824: absorption system we are aware of. We also detect strong absorption from \ion{C}{1} implying that the gas is shielded from strong UV radiation. The spectrum shows a red wing of a Ly$\alpha$ line from which we derive a \ion{H}{1} column density of $\log(N($\ion{H}{1}$))=21.50\pm$0.20. Imaging of the field revealed a fairly bright and red host, detected both in $R$ and $K$ bands, suggesting that it is an evolved, massive galaxy.
825:
826: The spectrum is also remarkable in that the extinction curve of the line of sight towards \grb\ shows a clear signature of the so called Milky Way or $2175$~\AA\ bump. At a redshift of $z=2.45$, it is by far the highest redshift detection of the $2175$~\AA\ bump to date. It shows that the conditions for the creation and non-destruction of the carrier of the bump must already have been in place early in the universe. This is the first clear detection of the bump in the host of a GRB, with the SMC being the typical type of extinction for GRB sightlines. This makes \grb\ an ideal candidate to study the environment needed for the creation and/or non-annihilation of bump, by comparing it to other bump-less GRBs.
827:
828: To accurately derive the properties of the extinction curve in the UV and the bump we fit it to the parametrization of \citet{fitzpatrick2007}.
829: The bump is found to be centered at $x_c\approx4.6\pm0.1$~$\mu$m$^{-1}$ (or $\lambda_c\approx2174\pm50$~\AA), the width of the bump is found to be $\gamma\approx1.08\pm0.05$~$\mu$m$^{-1}$ while the 'strength' is $c_3\approx2.7\pm0.1$ (taking into account the uncertainty in the intrinsic spectral slope $\beta$). The value of $c_3$ is the same as the average that \citet{gordon2003} find for the LMC average sample ($c_3=2.7\pm0.1$) but higher than for the LMC2 sample ($c_3=1.5\pm0.1$) although it is within the scatter. The width $\gamma$ is a bit wider compared to their $\gamma=0.93\pm0.02$~$\mu$m$^{-1}$ forethe LMC average and $\gamma=0.95\pm0.03$ for LMC2 (although it is still within the scatter of both samples). The amount of extinction is $A_V\approx1.3$, but when taking into account $1~\sigma$ deviations in $\beta$, $A_V=0.8$--$1.5$ for the FM and LMC fits.
830:
831: In the Local Group, the MW bump is a characteristic feature of the MW extinction curve. It is also observed in the LMC, although it is usually weaker and followed by a steeper rise in the UV, while it is not observed in four out of five curves measured for the SMC (see \ref{sec:ext_ext} for more details and exceptions). Of these three 'local type extinction laws' we find that the extinction of \grb\ most closely resembles that of the LMC. We find that this result is robust, even taking into account a possible contribution to the extinction from the strong foreground \ion{Mg}{2} absorber. It has been suggested, based on the difference in the SMC, LMC and MW, that the strength of the bump correlates with metallicity. However, \citet{gordon97} found that although starburst galaxies can have varying metallicities, their extinction curves all lack the 2175~\AA\ bump. By comparing \grb\ to another high metallicity GRB sightline which does not show any sign of a bump in its extinction curve, we similarly conclude that metallicity is not the only driver of the 2175~\AA\ bump.
832:
833: Another special feature in the spectrum of \grb\ is the detection of a strong \ion{C}{1} absorption. This is to our knowledge the first GRB spectrum to contain \ion{C}{1} absorption and we propose that there may be a correlation between its detection and the presence of the bump. We have checked this suggested correlation for a few other lines of sight, and found all of them to be in agreement. This prediction has also been checked by \citet{2009ApJ...691L..27P} who find a \ion{C}{1} absorption line and a 2175~\AA\ bump (based on photometric data) in GRB~080607.
834: We also find a high dust-to-gas ratio, which is consistent with a proposed correlation by \citet{gordon2003}, suggesting that the strength of the bump is related to the dust-to-gas ratio. Extending their correlation plot to include the MW, \grb\ and other GRBs from the literature with extinction analysis, we find that they all follow the proposed correlation. Finally, the presence of the \ion{C}{1} absorption feature in GRB\,070802 suggests that the UV radiation field is weaker than in typical GRB environments. This is in agreement with the tentative conclusion of \citet{gordon2003} that a weaker UV flux density is found to correlate with the presence of the bump.
835:
836: The simultaneous presence of \ion{C}{1} and the
837: 2175\,\AA\ bump as well as the lower overall ionization state of the gas,
838: relatively (though not exceptionally) high metallicity, and large
839: dust-to-gas ratio may be explained in a scenario in which the dust column is
840: strongly enriched by the presence of asymptotic giant branch (AGB) stars. This would be consistent with the conclusion that the host of \grb\ is a massive evolved galaxy, and supports the conclusions of \citet{noll07} that the presence of the bump requires an evolved population.
841:
842:
843: \acknowledgments
844: We thank Javier Gorosabel for providing us with his code for the Pei parametrization of the SMC, LMC and MW extinction laws. We also thank P\'all Jakobsson (Palli) and the anonymous referee for their comments on the manuscript. We thank the various members of the GRB community for regularly sacrificing their beauty sleep in their chase for GRB afterglows.
845: The Dark Cosmology Centre is supported by the DNRF. \'A.~E. and PMV acknowledge the support of the EU under a Marie Curie International Outgoing Fellowship, contract PIOF-GA-2008-220049, and a Marie Curie Intra-European Fellowship, contract MEIF-CT-2006-041363. J.~X.~P. is partially supported by NASA/Swift grants NNG06GJ07G and NNX07AE94G and an NSF CAREER grant (AST-0548180).
846:
847:
848: %\begin{thebibliography}{}
849: \bibliography{references}
850: %\end{thebibliography}{}
851:
852:
853:
854: \appendix
855: \include{app_ext}
856:
857: \end{document}
858:
859:
860:
861: