The First Positive Detection of Molecular Gas in a GRB Host Galaxy

0.29 MB·English

by  J. X. Prochaska

Accepted to ApJL,December3 2008
PreprinttypesetusingLATEXstyleemulateapjv.11/26/04
THE FIRST POSITIVE DETECTION OF MOLECULAR GAS IN A GRB HOST GALAXY
J. X. Prochaska1,2, Y. Sheffer3, D.A. Perley4, J. S. Bloom4,5, L. A. Lopez2, M. Dessauges-Zavadsky6, H.-W.
Chen7, A. V. Filippenko4, M. Ganeshalingam4, W. Li4, A. A. Miller4, and D. Starr4,8
Accepted to ApJL, December3 2008
9 ABSTRACT
0
We report on strong H and CO absorption from gas within the host galaxy of gamma-ray burst
0 2
(GRB)080607. AnalysisofourKeck/LRISafterglowspectrumrevealsaverylargeHIcolumndensity
2
(N = 1022.70±0.15cm−2) and strong metal-line absorption at z = 3.0363 with a roughly solar
H I GRB
n metallicity. We detect a series of A−X bandheads from CO and estimate N(CO)=1016.5±0.3cm−2
a and TCO > 100 K. We argue that the high excitation temperature results from UV pumping of the
J ex
CO gas by the GRB afterglow. Similarly, we observe H2 absorption via the Lyman-Werner bands
5 and estimate N = 1021.2±0.2cm−2 with TH2 = 10–300 K. The afterglow photometry suggests an
H2 ex
extinction law with R ≈ 4 and A ≈ 3.2 mag and requires the presence of a modest 2175 ˚A
] V V
E bump. Additionally, modeling of the Swift/XRT X-ray spectrum confirms a large column density
H with NH = 1022.58±0.04cm−2. Remarkably, this molecular gas has extinction properties, metallicity,
andaCO/H ratiocomparableto those oftranslucentmolecularclouds ofthe Milky Way,suggesting
. 2
h that star formation at high z proceeds in similar environments as today. However, the integrated
p dust-to-metals ratio is sub-Galactic, suggesting the dust is primarily associated with the molecular
- phase while the atomic gas has a much lower dust-to-gas ratio. Sightlines like GRB 080607 serve as
o
powerful probes of nucleosynthesis and star-forming regions in the young universe and contribute to
r
t the population of “dark” GRB afterglows.
s
a Subject headings: gamma rays: bursts — ISM: molecules
[

1 1. INTRODUCTION son et al. 2006), metal-enriched material (Savaglio et al.
v
2003), and large depletion factors indicating significant
6 Most long-duration, soft-spectrum gamma-ray bursts
dustformation(e.g.,Watsonetal.2006;Prochaskaetal.
5 (GRBs) are thought to arise in the death of massive
2007). Indirectly, one also infers significant extinction
5 stars (see Woosley & Bloom 2006, for a review). The
fromthereddeningofsomeafterglows(e.g.,Watsonetal.
0 evidence for this is both direct, such as the observation
2006; El´ıasd´ottir et al. 2008) and from the large metal
. of a contemporaneous supernova associated with some
1 column densities implied by absorption of the X-ray af-
events (e.g., Stanek et al. 2003; Hjorth et al. 2003), and
0 terglow (e.g., Galama & Wijers 2001; Campana et al.
circumstantial, based on physical associations with star
9
2006;Butler & Kocevski 2007).
0 forming galaxies (see Fruchter et al. 2006). Still, some
Evidenceformoleculargas,acrucialingredientofstar
: telltale signatures of massive-star progenitors are rarely,
v formation, has proved elusive thus far. Tumlinson et al.
if ever, observed. The absence of wind-blown density
i (2007)presentedasurveyforH Lyman-Wernerabsorp-
X profilesinthecircumburstenvironment(Chevalier2004) 2
tion in the afterglow spectra of four GRBs and reported
is thought to be due to the pre-burst interactions of the
r no positive detections. Fynbo et al. (2006) tentatively
a wind with the interstellar medium (ISM; Ramirez-Ruiz
interpreted an absorption feature in their spectrum of
et al. 2005). The absence of Wolf-Rayet signatures in
GRB 060206as H , but this may instead be interpreted
afterglow spectra (Chen et al. 2007) is explained by the 2
as an unrelated, Lyα absorption line. The absence of
influence of the afterglow light on gas and dust within
∼100 pc of the GRB event. moleculeslocaltotheGRB progenitormaybeexplained
asanaturalconsequenceofphotodissociationbythestar
Outside the overwhelmingly destructive influence
prior to its death (Whalen et al. 2008). The absence of
within the inner 100pc, there are generic signatures of
the ISM: H I gas (ΣH I ≈ 1 − 100M⊙pc−2; Jakobs- molecules along the full sightline, meanwhile, suggests
thatanelevatedUVradiationfieldissuppressingH for-
2
1UniversityofCaliforniaObservatories/LickObservatory, Uni- mation throughout the host galaxy (Chen et al. 2008).
versityofCalifornia,SantaCruz,CA95064. Nevertheless, these galaxies are actively forming stars
2DepartmentofAstronomyandAstrophysics,UniversityofCal- (Christensen et al. 2004) and GRB sightlines should oc-
ifornia,SantaCruz,CA95064.
3Department of Physics and Astronomy, University of Toledo, casionallypenetrateneighboringmolecularcloudsonthe
Toledo,OH43606. path to Earth.
4Department of Astronomy, University of California, Berkeley, In this Letter, we present the first such unambiguous
CA94720-3411. detectionofH andCOabsorptioninaGRBhostgalaxy
5SloanResearchFellow. 2
6ObservatoiredeGen`eve,51Ch. desMaillettes,1290Sauverny, andprovideinitialanalysisofthegasanddustalongthe
Switzerland. sightline to GRB 080607. Previous detections of inter-
7DepartmentofAstronomy&Astrophysics,andKavliInstitute vening CO absorption toward quasars have only shown
forCosmologicalPhysics,5640S.EllisAve,Chicago, IL,60637. trace quantities of molecular gas (Srianand et al. 2008).
8LasCumbresObservatory,6740CortonaDr. Suite102,Santa
Emission-line studies of CO and other molecules, mean-
Barbara,CA93117.
while, lack sufficient spatial resolution to resolve the gas
2
2.0
)
2
m
c 1.5
g/
n
A
s/
g/
er 1.0
(
x
u
Fl
e
v
ati 0.5
el
R
0.0
4000 4500 5000 5500
Wavelength (Ang)
)
2
m
c
g/
n 6
A
s/
g/ 4
r
e
(
x
u
Fl 2
e
v
ati 0
el 6000 7000 8000 9000
R
Wavelength (Ang)
Fig. 1.—Keck/LRISspectraoftheoptical afterglowofGRB080607takenwith(a)theLRISbcameraandtheB600grismand(b)the
LRISrcamerawiththeR400grating. Thereddashedlinesindicateourmodeloftheintrinsicafterglowspectrumreddenedbydustinthe
hostgalaxy. Atλ≈4900˚AoneidentifiesaDLAprofileassociatedwithHIgasneartheGRB.Theshadedregionoverplottedonthedata
corresponds toanHIcolumndensity NHI=1022.7±0.15cm−2. Themodel(cyan solidline)includesabsorptionfromH2 Lyman-Werner
transitions. Thelineopacity atλ>5500 ˚Aisdominatedbymetal-linetransitionsfromgas inthehost galaxyandincludes bandheads of
theCOmolecule. AbsorptionfromEarth’stheA-bandismarkedbya⊕.
beyondthelocaluniverse(e.g.,Tacconietal.2006). Our We then pausedto obtaincalibrationframes before con-
dataconstrainthephysicalconditionsoftheISMinclud- tinuing with a sequence of two exposures (1500 s each)
ing its metallicity, dust properties, and molecular frac- with the B600 grism and the R1200 grating (FWHM ≈
tions. We discuss these results and the prospects for fu- 110 kms−1, λ ≈ 6300 ˚A). All data were obtained
center
ture studies of molecular clouds using GRB afterglows9. under photometric conditions and a seeing better than
1′′. The data were reduced and calibrated using the
2. OBSERVATIONS LowRedux10 software package developed by S. Burles,
J.Hennawi,andJ.X.P.usingarcimagesandaspectrum
We (J.S.B. and D.P.) initiated spectroscopic observa-
of HZ44.
tions of the GRB 080607 afterglow with Keck I/LRIS
(Oke et al. 1995) at 06:27:39 UT on 7 June 2008, ∼ The Peters Automated Infrared Imaging Telescope
(PAIRITEL;Bloometal.2006)respondedautomatically
20min following the Swift triggerof the GRB (Mangano
totheSwifttriggerandbegansimultaneousobservations
etal.2008). The Keck/LRISspectrographhas twocam-
in the J,H,K bands at 06:08:44 UT. A total of 387 in-
erasoptimized for blue (LRISb; λ≈3000−5000˚A) and s
dividual7.8sexposuresweretakenundermoderatecon-
red (LRISr; λ ≈ 5000− 9000 ˚A) optical wavelengths. ditions,seeing∼3”,foratotaltimeontargetof∼3000s.
Using the 1′′-wide long slit, the atmospheric dispersion
We have reduced these data with the standard PAIRI-
corrector, and the D560 dichroic, we obtained a series
TEL pipeline. The Katzman Automatic Imaging Tele-
of three two-dimensional spectra of increasing exposure
scope (KAIT; Filippenko et al. 2001;Li et al. 2003)also
time totaling 2100 s using the B600grism in LRISb
responded to the Swift alert of GRB 080607 automati-
(FWHM ≈270 kms−1) and the R400 grating in LRISr
cally, and began the first image at 06:09:25 UT, 118 s
(FWHM ≈ 300 kms−1, centered at λ ≈ 7310 ˚A).
center
10 http://www.ucolick.org/∼xavier/LowRedux/index.html .
9 http://www.graasp.org
3
λ (Å)
4000 eff,rest2000 1000
β = 0.67 ± 0.05 0.1
14.0 A = 3.21 10000
K V
R = 4.01
V 5000
15.0 C3 = 1.34 −1V
Magnitude (AB) 1176..00 H J Cχ24/ d=o 0f .=2 88.36 / 4 51200000000µF (Jy)ν −1alized Counts s ke 0. 01 2.4
m
18.0 t = 300s I cleVar 200 Nor 1 0−3 Photon Index2.12.22.3 +
Corrected for Galactic A = 0.07
19.0 Host redshift z = 3.036 V 100 10−4 2 3 3.5NH (10224 cm−2) 4.5 5
20000 10000 8000 6000 4000 0 .5 1 2 5 10
λeff (Å) Energy (keV)
Fig. 2.—(a)Optical/IRphotometryfortheafterglowofGRB080607,asobservedbyPAIRITELandKAIT(seeUpdikeetal. inprep.).
Overplotted on the data isour best-fit model forandust-extinguished power-law spectrum, fν ∝ν−β. Themodel, whichhas amodified
Galacticextinctionlaw,mustincludethe2175˚Afeaturetomatchtheobservations. (b)Swift/XRTX-rayspectrum(0.3–10.0keV)obtained
0.6–376 ks after the trigger, with the best-fit model overlaid in red. The data were modeled as a power law, with a Galactic absorption
component aswellasaredshiftedintrinsicabsorptioncomponentNH. Thebest-fitmodelhasanintrinsicNH=1022.58±0.04cm−2 anda
photon index2.26±0.07, usingtheGalactic absorption1.9×1020cm−2 (Dickey &Lockman 1990). The68%, 95%, and99% confidence
contours ofthephoton indexandintrinsicabsorptionareshown.
after the BAT trigger. A series of V, I, and unfiltered
imageswereobtainedforthefirst∼1800saftertheburst TABLE 1
EW Measurements
under good conditions.
3. ANALYSIS Region[˚A] λac [˚A] Wobs [˚A] ID zabs
In Figure 1 we present the combined LRIS spectra of
B600
the GRB afterglow. A very strong absorption feature
is evident at λobs ≈ 4900 ˚A corresponding to damped [[55005455..73,,55006553..05]] 76..5783±±00..2287 CSrIIII12205506 13..406327
Lyα (DLA) absorptionfrom gas within the galaxy host- SII1253 3.037
ing GRB 080607. The coincidence of strong metal-line [5081.2,5093.6] 10.66±0.33 CrII2066 1.462
transitions including fine-structure lines of Fe II and SII1259 3.037
Si II confirms this association (Vreeswijk et al. 2004;
Note. —[Thecompleteversionofthistableisintheelectronic
Prochaska et al. 2006). A Gaussian fit to the strongest
editionoftheJournal. Theprintededitioncontainsonlyasample.]
low-ion transitions gives a heliocentric redshift z =
3.0363±0.0003. GRB aCentralwavelengthdeterminedfromaGaussianfit. Wobsvalues
determinedfromboxcarintegrations havenovaluelisted.
Thedashedcurvesrepresentourmodelfortheintrinsic
afterglow spectrum, reddened by dust along the sight-
line. Assuming an intrinsic optical-infrared power-law
spectral slope f ∝ν−βO, we first fit the JHK photom- galaxy (Kru¨hler et al. 2008;El´ıasd´ottir et al. 2008); and
ν
etry for the single parameterE(B−V)=0.8±0.1 mag. (iii) the data favor a low c4 value suggesting few small
dustgrains. Thesecharacteristicsaredistinguishedfrom
In this analysis we accounted for the uncertainty in the
the extinction laws commonly inferred for GRB sight-
unknownintrinsicspectralindex,whichforafterglowsat
early times is found to be in the range β = 0.1−1.1 lineswhichgenerallyhaveSMC-likeor“grey”extinction
O curves and lack evidence for a 2175 ˚A bump (e.g., Kann
(Kann et al. 2007). We then fitted a Fitzpatrick &
etal.2006;Butleretal.2006;Watsonetal.2006;Perley
Massa (1990) dust model to unabsorbed regions of the
et al. 2008).
R400spectrum(correctingforGalacticextinction),hold-
ing E(B−V) fixed but allowing all other parameters to With the reddened afterglow spectrum as our input
vary11. Foranassumedβ =0.5,ourbest-fitparameters continuum, we have fitted the Lyα absorption to the
are R =4.0±0.23,A O=3.2±0.5mag,c =1.3±0.3, data redward of 5000 ˚A with centroid λ = 1215.67(1+
anAdltch4Vo=ug0h.3th±er0e.1is(cFoingVs.id2ear)a.bledegeneracy3amongthese NzGHRBI)=˚A1.0T22h.7e0±be0.s1t5-ficmtv−a2l,uwehfoerrethteheHuInccoelrutmainntdyeinssditoymis-
values, we reach three robust conclusions: (i) the R inatedbythecontinuummodel. Wedetecttheafterglow
valueishigherthanthetypicalGalacticandSMCvalueVs; down to at least λ ≈ 4000 ˚A, primarily in a series of
(ii) the data require a nonzero c3 parameter indicating emission“gaps”(e.g.,atλobs ≈4200, 4275, 4315˚A).Al-
the positive detection of the 2175˚A bump, only the sec- thoughthe absorptionbetween the gaps could be strong
IGM absorption, this would require a much higher inte-
onddetectionofthisfeatureassociatedwithaGRBhost
gratedopacitythanobservedalongquasarsightlines. We
11 Thisallowsforcolorevolutionbetween thetimesofthepho- interpret the majority of this opacity as Lyman-Werner
tometricandspectroscopic observations. bands of molecular hydrogen at z =zGRB.
4
1.0 1.0 1.0 1.0
0.5 0.5 0.5 0.5
SiII 1526
J=1/2 GeII 1602 CO 8−0 CO 4−0
0.0 0.0 0.0 0.0
1.0 1.0 1.0 1.0
0.5 0.5 0.5 0.5
x SiII 1533 FeII 1611 x
u J=3/2 J=9/2 u CO 6−0 CO 2−0
Fl0.0 0.0 Fl0.0 0.0
ed 1.0 1.0 ed 1.0 1.0
z z
ali ali
m m
0.5 0.5 0.5 0.5
r r
o o
N FeII 1618 N
CIV 1550 J=7/2 CO 1−0 CO 1−0
0.0 0.0 0.0 0.0
1.0 1.0 1.0
1.0
0.5 0.5 0.5
0.5
CI 1560 MgI 1683 CO 0−0 CO 0−0
0.0 0.0 0.0
−400 −200 0 200 400 −400 −200 0 200 400 −500 0 500 1000 −500 0 500 1000
Relative Velocity (km s−1) Relative Velocity (km s−1) Relative Velocity (km s−1) Relative Velocity (km s−1)
Fig. 3.—Velocityplotsofasubsetoftheatomicandionictran- Fig. 4.— Velocity plots of the molecular transitions associated
sitionsassociatedwiththeISMofthegalaxyhostingGRB080607. withtheISMofthegalaxyhostingGRB080607. Thefirstcolumn
Inallpanels,v=0kms−1 correspondstozGRB=3.0363andthe showstheCOA−XbandheadstakenwiththeLRISb/B600grism
orangedashedlinesindicateblendsbetweentransitions. (top two) and LRISr/R400 grating (bottom two). All other data
refersto the LRISr/R1200 spectrum. The redlineoverplotted on
The LRISr/R400spectrum (Fig. 1b) coversrestwave- theCOdataindicateourbest-fitmodelcorrespondingtoN(CO)=
lengthsλrest >1300˚Arelativeto the GRBand, assuch, 1o0ff1s6e.t5fcrmom−2z,GRbCBOof=δvC2.O0k=m30s−k1m, sT−eCx1O. In=al3l0p0anKe,lsa,nvd=a0vkemlocsi−ty1
noneoftheabsorptionisrelatedtotheLyαforestdespite correspondstozGRB=3.0363andtheorangedashedlinesindicate
its similar appearance. Instead, the features are metal- blendsbetweentransitions.
line absorption by gas in the GRB host galaxy (Fig. 3)
The optical spectra also reveal the positive detection
and/or intervening gas. We have measured equivalent
of a series of CO A−X bandheads (Fig. 4). It is note-
widths (W ) for these features by fitting Gaussian pro-
λ
worthythatthepeakdepthofthebandheadsisoffsetby
files or, in cases of severe line-blending, by boxcar sum-
δv ≈+150kms−1 from the atomic transitions and that
mation over the spectral window (Table 1). The ab-
the bandheads show absorption to δv ≈ +1000kms−1.
sorption is dominated by resonance transitions and fine-
At first glance, this suggests that the molecular gas has
structure lines ofabundantatomsandions atz =z ,
GRB
but we also identify absorption from two strong Mg II an extremely high velocity dispersion, a highly unlikely
absorbers (W > 1 ˚A) at z = 1.340 and 1.462, an attribute for molecular gas. We interpret these unusual
2796
characteristics, instead, as the result of absorption from
unusuallycommoncharacteristicofGRBafterglowspec-
high rotational levels (J > 10) which have been popu-
troscopy (Prochter et al. 2006).
lated by UV pumping from the GRB afterglow. This
TheGRBabsorptionfeaturesarehighlysaturatedand
is analogous to the excitation of fine-structure levels in
one can only estimate lower limits to the ionic column
densities. Under the very conservative assumption that Fe+, Si+, and O0 gas in other GRB galaxies (Prochaska
N = 1023.05W /(fλ2)cm−2 (i.e., the weak limit of the et al. 2006; Vreeswijk et al. 2007). Such levels would be
λ
easily resolved with an echelle spectrum.
curve of growth, with f the oscillator strength), we in-
fer [S/H] > −1.7, [Si/H] > −2, [Fe/H] > −1.8, and We havefittedthe COA−X bandheadsinthe R1200
[Zn/H] > −1.7. We also detect very weak transitions and B600 data with a series of models and find pre-
of several common elements (e.g., O I 1355, Ni II 1804, ferred values N(CO) = 1016.5±0.3cm−2, bCO = 2.0 ±
Mg I 1683, Co II 1574) and a few transitions from rare 0.3kms−1,TeCxO =300+−27000 K,andavelocityoffsetfrom
elements (e.g., Ge II 1602). The equivalent width of z of δv = 30±15kms−1 (Fig. 4). The preferred
GRB CO
the Ge II 1602 transition alone implies a lower limit values were found by minimizing the residuals between
[Ge/H] > 0 and analysis of Co II transitions, a re- thedataandthemodelforthe1-0,2-0,and4-0CObands
fractory element, yields [Co/H] > −0.7dex. This is simultaneously. Each model included all rotational lev-
consistent with the equivalent width of the absorption els up to J = 25, for a total of 225 CO transitions with
coinciding with the O I 1355 transition which implies transition data taken from fits to empirical data (Mor-
[O/H] > −0.2dex. We conclude that the integrated gas ton & Noreau 1994). We then visually compared the
metallicityissolarorevensuper-solar,aresultthatchal- model prediction with the 6-0 and 8-0 CO bands seen
lenges the prevailing collapsar model (e.g., Woosley & inthe B600spectrum to assureconsistencyovera larger
Heger 2006). range of f-values; these bandheads set an upper limit
5
of N(CO) < 1017.2cm−2. Although a parameterization preferred model of the CO absorption suggests a veloc-
of the gas with an excitation temperature (TeCxO) is not ity offset of δvCO ≈ +30kms−1 from the atomic metal
formally valid for a photon-pumped gas, we achieve a lines. The second key difference is that we infer a much
reasonable model. lower dust-to-metals ratio than Galactic whether that
We have analyzed the H gas in a similar fashion. metal column is estimated from the UV or X-ray spec-
2
Because the Lyman-Werner lines are heavily saturated tra. We interpret this difference as the result of a dust-
andblended,there is little sensitivityto the Doppler pa- poor, atomic phase that is spatially separated from the
rameter b . Adopting b = 3kms−1, we varied TH2 highly extinguished molecular cloud. Both phases of gas
H2 H2 ex
andN andcomparedthesemodels againstthedataat give rise to metal absorption (the profiles span the es-
λ <H21100 ˚A. Under the assumption that H absorp- timated 30kms−1 offset; Fig. 3) but we speculate that
rest 2
tion dominates the opacity at λ < 4200 ˚A, we con- the atomic phase has a much lower dust-to-gas ratio.
obs
strain N = 1021.2±0.2cm−2 and TH2 = 10 to 300 K. Severallines of evidence argue against identifying this
H2 ex
Models with larger/smaller N values conflict with the molecularcloudasthebirthplaceoftheGRBprogenitor.
observations. Our analysis sHet2s an upper limit to the We detect strong absorption from C I and Mg I tran-
integratedmolecularfractionf (H )<0.13,butweex- sitions indicating the gas lies > 100pc from the GRB
int 2
pectthatthesightlineiscomprisedofatleasttwoclouds: (Prochaskaetal.2006). Ifthe moleculargaswerewithin
one predominantly atomic and the other molecular. ∼10pcoftheafterglow,asignificantcolumnofH2would
Assuming solar relative abundances, the implied O be excited to unbound states (Draine & Hao 2002). We
column density is N(O) > 1019.6cm−2 for a Galac- have searched for such transitions at λrest >1400 ˚A and
tic equivalent column density N > 1022.7cm−2. This reportnopositivedetections,settinganequivalentwidth
H
metal column density implies that there should be sig- limitWrest(H∗2)<0.1˚A(95%c.l.). Wehavesearchedfor
nificant absorption of soft X-rays by K-shells of O, Si, variability in the dust opacity, H I absorption, and in
etc. InFig.2b,wepresentthe X-rayspectrumandbest- the equivalent widths of the metal-line transitions but
fit model of the Swift/XRT data obtained 0.6–376 ks detect none. The most plausible scenario, therefore, is
after the burst (Evans et al. 2008). We estimate an that the gas lies beyond 100pc from the afterglow, con-
effective N = 1022.58±0.04cm−2 and a photon index sistent with the estimate for photodissociation regions
H
Γ = 2.26±0.07 for a power-law spectrum. Therefore, surrounding GRB progenitors (Whalen et al. 2008). We
themetalcolumndensitiesfromanalysisofboththe UV concludethatthe molecularcloudidentifiedinthe after-
andX-rayspectraareconsistent. Ontheotherhand,as- glow spectrum of GRB 080607 arises in a neighboring,
suming the Galactic dust-to-metals ratio would give an star-forming region.
extinction(AV >10mag)inconsistentwiththeobserva- This marks the first unambiguous detection of H2 in
tions. We conclude that the gas has a much lower dust- a GRB sightline (among six intensively searched) and
to-metals ratio than Galactic (Galama & Wijers 2001; of CO (among ∼ 20 sightlines). Our results offer in-
Watson et al. 2006). sight into the rarity of these detections. From our
dust model, we estimate a rest-frame UV extinction of
4. DISCUSSION A ≈8mag. Ifthisafterglowwerejust1magfainter,
˚
The previous section describedthe physicalproperties 1100A
then the Lyman-Werner series could not have been de-
of the ISM and the first clear detection of molecular gas
tected. Similarly, the CO bandheads would not have
in a GRB host galaxy. With an A ≈3.2 mag,this rep-
V beenobservableforthemajorityofGRBs. Weconclude,
resentsamolecule-rich“translucent”sightlinewitheffec-
therefore, that one requires an uncommonly luminous
tive UV shielding of CO (van Dishoeck & Black 1988).
afterglow combined with a fortuitous path through an
With N(CO)≈1016.5cm−2 and N ≈1021.2cm−2, the
H2 unrelated molecular cloud.
abundanceratioofCOtoH is∼2×10−5. Thesevalues
2 We contend further that a significant fraction of the
are consistent with detailed models of CO photodissoci-
subluminous optical bursts (also called “dark bursts”;
ationintranslucentenvironments(e.g.,modelT3invan
Djorgovskietal.2001;Jakobssonetal.2005;Cenkoetal.
Dishoeck & Black 1988), and with the transition region
2008) are events like GRB 080607 but with fainter in-
betweendiffuseanddarkcloudsinthecorrelationplotof
trinsic luminosity. An open and intriguing question is
CO versus H (Sheffer et al. 2008). Altogether, the gas
2 whether these dust-extinguished events correspond to
exhibitsroughlysolarmetallicity,thedusthasanextinc-
unique host galaxies (e.g., intensely star forming and
tioncurvesimilartothatobservedforGalacticmolecular
dustysystems),orwhetherthehostshavepropertiessim-
clouds, including a 2175 ˚A bump, and the CO/H ratio
2 ilartothegeneralpopulationandthevariationsalongin-
is consistent with Galactic translucent clouds. We infer
dividual sightlines is geometrical (i.e., molecular clouds
thatatleastsomemolecularcloudsintheyounguniverse
have small cross-section and are rarely intersected). A
have similar properties to those observed locally.
resolution of these two effects bears on the nature of
There are, however, two notable differences from the
molecular gas at high redshift including its distribution
Galactic ISM. First, we observe a very large HI column and relation to star formation. A multi-wavelength sur-
density. ThisattributeiscommontoGRBsyetcontrasts
veyofthehosts,includingsearchesforCOemission,may
with the surface densities of local H I disks. While the
help distinguish between these scenarios.
formation of a molecular cloud requires a large dust col-
Thoughrare,these high-z events enable studies at op-
umn to absorb H -dissociating photons, the analysis of
2 tical wavebands of rest-frame UV features that serve as
Krumholz et al. (2008) implies a 20 times lower column
powerful diagnostics of star-forming regions. This in-
than observed. Perhaps the large N value is a coin-
HI cludes a precise redshift for follow-up observations, for
cidence; these are routinely observed along GRB sight- example studies of CO in emission at millimeter wave-
lines without molecules (Tumlinson et al. 2007) and our
6
lengths. Indeed,the CO3-2and1-0transitionslie atfa- the California Institute of Technology, the University of
vorable observational frequencies. High-resolution spec- California,and the National Aeronautics and Space Ad-
troscopy of systems like GRB 080607 would also per- ministration(NASA).KAIThasbeensupportedprimar-
mit searches for additional molecules in the UV (e.g., ily by donations from AutoScope Corp., Lick Observa-
C ; Sonnentrucker et al. 2007). Finally, such systems tory, the NSF, the Sylvia & Jim Katzman Foundation,
2
easily satisfy the metal-strong DLA criteria of Herbert- andtheTABASGOFoundation. ThePAIRITELproject
Fort et al. (2006) and afford a special opportunity to is operated by the Smithsonian Astrophysical Observa-
study rare elements in the young universe. Assuming tory (SAO) and was made possible by a grant from the
solar abundances, for example, we estimate that the Harvard University Milton fund, the camera loan from
Sn II λ1400, B II λ1362, and Pb II λ1433 transitions the University of Virginia, and the continued support
should all have rest-frame equivalent widths exceeding of the SAO and UC Berkeley; it is further supported by
50 m˚A. NASA/SwiftGuestInvestigatorGrant#NNG06GH50G.
WethankM.SkrutskieandA.Szentgyorgyifortheirsup-
port of PAIRITEL. We are grateful to E. E. Falco and
J. X. P. is partially supported by NASA/Swift grant
the Mt. Hopkins staff (W. Peters and T. Groner) for
NNX07AE94G, and an NSF CAREER grant (AST–
their continued assistance with PAIRITEL. This work
0548180). A.V.F.isgratefulforNSFgrantAST–0607485
madeuseofdatasuppliedbythe UKSwiftScienceData
and the TABASGO Foundation. Most of the data pre-
Centre at the University of Leicester. We acknowledge
sentedhereinwereobtainedattheW.M.KeckObserva-
D. Welty and M. Krumholz for valuable discussions.
tory,whichis operatedasa scientific partnershipamong
REFERENCES
Bloom,J.S.,Starr,D.L.,Blake,C.H.,Skrutskie,M.F.,&Falco, Krumholz, M. R., McKee, C. F., & Tumlinson, J. 2008, ArXiv
E. E. 2006, in Astronomical Society of the Pacific Conference e-prints,805
Series, Vol. 351, Astronomical Data Analysis Software and Li, W., Filippenko, A. V., Chornock, R., & Jha, S. 2003, PASP,
SystemsXV,ed.C.Gabriel,C.Arviset,D.Ponz,&S.Enrique, 115,844
751–+ Mangano, V., Cummings, J. R., Cusumano, G., Gehrels, N., La
Butler,N.R.,&Kocevski,D.2007,ApJ,663,407 Parola,V.,Markwardt,C.B.,Osborne,J.P.,Sbarufatti,B.,&
Butler,N.R.,etal.2006,ApJ,652,1390 vandenBerk,D.E.2008,GRBCoordinatesNetwork,7847,1
Campana,S.,etal.2006, A&A,449,61 Morton,D.C.,&Noreau,L.1994,ApJS,95,301
Cenko, S. B., et al. D. A., & Moon, D. . 2008, ArXiv e-prints Oke, J. B., Cohen, J. G., Carr, M., Cromer, J., Dingizian, A.,
(0808.3983) Harris,F. H.,Labrecque, S.,Lucinio, R.,Schaal, W., Epps, H.,
Chen,H.-W.,Perley,D.A.,Pollack,L.K.,Prochaska,J.X.,Bloom, &Miller,J.1995, PASP,107,375
J.S.,Dessauges-Zavadsky,M.,Pettini,M.,Lopez,S.,Dall’Aglio, Perley,D.A.,etal.2008,ApJ,672,449
A.,&Becker,G.D.2008,ArXive-prints,809 Prochaska,J.X.,Chen,H.-W.,&Bloom,J.S.2006,ApJ,648,95
Chen, H.-W., Prochaska, J. X., Ramirez-Ruiz, E., Bloom, J. S., Prochaska,J.X.,Chen,H.-W.,Dessauges-Zavadsky,M.,&Bloom,
Dessauges-Zavadsky, M.,&Foley,R.J.2007, ApJ,663,420 J.S.2007,ApJ,666,267
Chevalier,R.A.2004,inAmericanInstituteofPhysicsConference Prochter, G. E., Prochaska, J. X., Chen, H.-W., Bloom, J. S.,
Series,Vol.727, Gamma-RayBursts:30YearsofDiscovery,ed. Dessauges-Zavadsky, M., Foley, R. J., Lopez, S., Pettini, M.,
E.Fenimore&M.Galassi,355–360 Dupree,A.K.,&Guhathakurta, P.2006,ApJ,648,L93
Christensen,L.,Hjorth,J.,&Gorosabel,J.2004,aap,425,913 Ramirez-Ruiz, E., Garc´ıa-Segura, G., Salmonson, J. D., & P´erez-
Dickey,J.M.,&Lockman,F.J.1990,ARA&A,28,215 Rendo´n,B.2005,ApJ,631,435
Djorgovski, S. G., Frail, D. A., Kulkarni, S. R., Bloom, J. S., Savaglio,S.,Fall,S.M.,&Fiore,F.2003, ApJ,585,638
Odewahn, S.C.,&Diercks,A.2001,ApJ,562,654 Sheffer, Y.,Rogers, M.,Federman, S. R.,Abel, N.P., Gredel, R.,
Draine,B.T.,&Hao,L.2002,ApJ,569,780 Lambert,D.L.,&Shaw,G.2008, ApJ,687,1075
El´ıasdo´ttir,A´.,etal.2008,ArXive-prints(0810.2897) Sonnentrucker, P., Welty, D. E., Thorburn, J. A., & York, D. G.
Evans,P. A.,Osborne,J.P.,Burrows,D.N.,&Barthelmy,S. D. 2007,ApJS,168,58
2008, GRBCoordinates Network,7955,1 Srianand, R., Noterdaeme, P., Ledoux, C., & Petitjean, P. 2008,
Filippenko,A.V.,Li,W.D.,Treffers,R.R.,&Modjaz,M.2001,in A&A,482,L39
Astronomical Societyof thePacificConferenceSeries,Vol.246, Stanek,K.Z.etal.2003,ApJ,591,L17
IAU Colloq. 183: Small Telescope Astronomy onGlobal Scales, Tacconi, L. J., Neri, R., Chapman, S. C., Genzel, R., Smail, I.,
ed.B.Paczynski,W.-P.Chen,&C.Lemme,121–+ Ivison, R. J., Bertoldi, F., Blain, A., Cox, P., Greve, T., &
Fitzpatrick,E.L.,&Massa,D.1990, ApJS,72,163 Omont,A.2006,ApJ,640,228
Fruchter,A.S.,etal.2006,Nature,441,463 Tumlinson, J., Prochaska, J. X., Chen, H.-W., Dessauges-
Fynbo,J.P.U.,etal.2006, A&A,451,L47 Zavadsky, M.,&Bloom,J.S.2007, ApJ,668,667
Galama,T.J.,&Wijers,R.A.M.J.2001,ApJ,549,L209 vanDishoeck,E.F.,&Black,J.H.1988,ApJ,334,771
Herbert-Fort, S., Prochaska, J. X., Dessauges-Zavadsky, M., Vreeswijk,P.M.,etal.2004,A&A,419,927
Ellison, S. L., Howk, J. C., Wolfe, A. M., & Prochter, G. E. Vreeswijk,P. M.,Ledoux, C., Smette, A., Ellison,S. L., Jaunsen,
2006, PASP,118,1077 A.O.,Andersen,M.I.,Fruchter,A.S.,Fynbo,J.P.U.,Hjorth,
Hjorth,J.,etal.2003,Nature,423,847 J., Kaufer, A., Møller, P., Petitjean, P., Savaglio, S., & Wijers,
Jakobsson, P., Bjo¨rnsson, G., Fynbo, J. P. U., Jo´hannesson, G., R.A.M.J.2007, A&A,468,83
Hjorth, J., Thomsen, B., Møller,P.,Watson, D., Jensen, B. L., Watson,D.etal.2006,ApJ,652,1011
O¨stlin,G.,Gorosabel,J.,&Gudmundsson,E.H.2005,MNRAS, Whalen, D., Prochaska, J. X., Heger, A., & Tumlinson, J. 2008,
362,245 ArXive-prints,802
Jakobsson,P.,etal.2006,A&A,460,L13 Woosley,S.E.,&Bloom,J.S.2006,ARA&A,44,507
Kann,D.A.,Klose,S.,&Zeh,A.2006, ApJ,641,993 Woosley,S.E.,&Heger,A.2006,ApJ,637,914
Kann,D.A.,etal.2007, ArXive-prints(0712.2186)
Kru¨hler,T.,Ku¨pcu¨Yolda¸s,A.,Greiner,J.,Clemens,C.,McBreen,
S.,Primak,N.,Savaglio,S.,Yolda¸s,A.,Szokoly,G.P.,&Klose,
S.2008,ApJ,685,376