Dust and Ionized Gas Association in E/S0 Galaxies with Dust Lanes: Clues to their Origin

Abstract

We present results from an on-going programme to study the dust and ionized
gas in E/S0 galaxies with dust lanes. Our data, together with results from
previous studies of E/S0 galaxies, are used to demonstrate the tight
relationship between these two components. This relationship is discussed in
light of our current understanding of the nature and origin of the interstellar
medium (ISM), and in particular in the context of the interplay between the
different multi-temperature components. We show that focusing on dust obscured
regions as tracers of the ISM, and on their properties, serves as independent
evidence for the external origin of the dust and ionized gas.

Full text

Page 1

arXiv:1202.2119v1 [astro-ph.CO] 9 Feb 2012
Mon. Not. R. Astron. Soc. 000, 1–?? (0000)Printed 13 February 2012(MN LATEX style file v2.2)
Dust and Ionized Gas Association in E/S0 Galaxies with
Dust Lanes: Clues to their Origin
Ido Finkelman1⋆, Noah Brosch1, Jos´ e G. Funes S.J.2, Sudhanshu Barway3,
Alexei Kniazev3,4, Petri V¨ ais¨ anen3,4
1The Wise Observatory and the School of Physics and Astronomy, the Raymond and Beverly Sackler Faculty of Exact Sciences,
Tel Aviv University, Tel Aviv 69978, Israel
2Vatican Observatory, V-00120 Vatican City State, Italy
3South African Astronomical Observatory, PO Box 9, 7935 Observatory, Cape Town, South Africa
4Southern African Large Telescope Foundation, PO Box 9, 7935 Observatory, Cape Town, South Africa
Accepted 2012 February 7. Received 2012 February 6; in original form 2011 September 18
ABSTRACT
We present results from an on-going programme to study the dust and ionized gas in
E/S0 galaxies with dust lanes. Our data, together with results from previous studies
of E/S0 galaxies, are used to demonstrate the tight relationship between these two
components. This relationship is discussed in light of our current understanding of the
nature and origin of the interstellar medium (ISM), and in particular in the context
of the interplay between the different multi-temperature components. We show that
focusing on dust obscured regions as tracers of the ISM, and on their properties, serves
as independent evidence for the external origin of the dust and ionized gas.
Key words: galaxies: elliptical and lenticular, cD; galaxies: ISM; dust, extinction;
HII regions.
1 INTRODUCTION
How do E/S0 galaxies acquire their ISM? What are the pro-
cesses that regulate the evolution of the ISM? Posing such
questions, and considering their answers, modifies our tra-
ditional view of elliptical galaxies as simple and dull objects
while challenging our understanding of the classical Hubble
classification scheme.
The growing amount of observational data accumulated
by recent multi-wavelength surveys implies that there is no
simple answer to the longstanding origin question (de Zeeuw
et al. 2002; Rampazzo et al. 2005; Cappellari et al. 2011).
The general picture that emerges from these studies is that
both accretion and merger events and internal production
could have contributed to some extent to the overall ISM
content in E/S0 galaxies. The relative contribution of exter-
nal versus internal sources does not seem to follow a general
rule but probably varies between objects.
To delve into the properties of the ISM in E/S0 galax-
ies much attention has been paid to spectroscopic analysis
of the dynamics and chemical abundance of the gas and
stars (Sarzi et al. 2006; Annibali et al. 2010; Young et al.
2011). Although highly informative, the limited spectral cov-
erage of these investigations, together with our current lim-
⋆E-mail:ido@wise.tau.ac.il (IF)
ited knowledge of the interplay between different gas phases,
tend to lead to an indecisive conclusion regarding the true
nature of the ISM. Moreover, although it was found that
the motion and orientation of the gas in many systems are
decoupled from the stellar rotation, the overall distribution
of the kinematical misalignment between stars and gas in
E/S0 galaxies is inconsistent with a purely external origin
(Sarzi et al. 2006).
The detection of unusual optical signatures in such sys-
tems could provide more definitive and direct indications to
the dominant source of the ISM. This is particularly true
for a subclass of galaxies where dust is easily recognized by
its light obscuration which produces well-defined dark lanes.
A significant fraction of these galaxies exhibit morphologi-
cal disturbances, such as shells and tidal features, indicative
of recent mergers or accretion events (van Dokkum 2005;
Tal et al. 2009; Kaviraj et al. 2011). In fact, when a single
stellar population evolves the ejected material is expected
to spread evenly throughout the galaxy, whereas the spatial
distribution and orientation of the optical dust features are
frequently misaligned with those of the stars (van Dokkum &
Franx 1995; Tran et al. 2001; Krips et al. 2003; Martel et al.
2004; Quillen 2006; Finkelman et al. 2010a). Furthermore,
the dust content of these galaxies, estimated by measuring
total optical extinction, is usually found to be several times

Page 2

2
Ido Finkelman et al.
larger than expected to be supplied by internal processes,
such as stellar mass loss (Patil et al. 2007).
Studying the interplay between the various multi-
temperature components of the ISM in E/S0 galaxies with
dust lanes provides also vital clues to the external origin
of the dust. Large-scale dust structures are virtually always
accompanied by excited or ionized gas, whose angular mo-
menta are typically relatively high and often orthogonal to
those of the stars (Bertola 1987; Kormendy & Djorgovski
1989; Caon, Pastoriza & Macchetto 2001; Krajnovie et al.
2008; Davies et al. 2011). E/S0 galaxies with dust lanes
show also high detection rates of cold molecular gas (Sage
& Galletta 1993; Wang, Kenney & Ishizuki 1992; Combes,
Young & Bureau 2007; Krips et al. 2010; Young et al. 2011),
whereas large-scale dust features are also often associated
with HI gas in discs rotating at random orientations with
respect to the optical major axis (Kormendy & Djorgov-
ski 1989; Oosterloo et al. 2002; Leeuw et al. 2007). Perhaps
more intriguing is the detection of significant dust reservoirs
in X-ray bright E/S0 galaxies, where sputtering by hot par-
ticles is expected to erode and destroy the dust grains on
short time scales of ∼100 Myr.
To tie these pieces of information Sparks et al. (1989)
and de Jong et al. (1990) independently proposed the ‘evap-
oration flow’ scenario, in which cold dust and gas are exter-
nally acquired by an elliptical galaxy during a gravitational
interaction with a gas-rich galaxy, and then heated due to
the thermal interaction with the ambient hot plasma. As a
result, the interacting hot gas locally cools to its warm (ion-
ized) phase, whereas the dust is partially shielded within
the gradually evaporating dense molecular clouds. This sce-
nario accounts also for the tight spatial correlation between
dust and ionized gas that is often observed along filaments,
or cooling flows, associated with the presence of hot X-ray
emitting coronal gas (Macchetto et al. 1996; Trinchieri &
Goudfrooij 2002; Fabian et al. 2003; sparks et al. 2004). Un-
like cold dust and molecular gas, atomic hydrogen is not
expected to survive in the presence of hot X-ray gas, which
explains the HI deficiency in the central parts of E/S0 galax-
ies with dust lanes (Oosterloo et al. 2002; Leeuw et al. 2008).
This paper summarizes the results of an ongoing study
of a sample of E/S0 galaxies where dust lanes have been
reported in the literature. Our results, combined with those
from similar studies, are used to demonstrate that the prop-
erties of the observed dust strongly support the ‘evaporation
flow’ scenario sketched above.
The paper is organized as follows: Section 2 gives a
short description of the sample, observations and data re-
duction; Section 3 discusses the data analysis and results;
the results are discussed in Section 4 and our conclusions
are summarized in Section 5. We shall assume throughout
the paper standard cosmology with H0 = 73 km s−1Mpc−1,
Ωm = 0.27 and ΩΛ = 0.73.
2 OBSERVATIONS AND DATA REDUCTION
This is the second paper in our survey of the ISM in a
distinct morphological subclass of early-type galaxies (see
Finkelman et al. 2010a). Our sample consists primarily of
galaxies identified by early photographic catalogues as ellip-
tical stellar bodies crossed by extended dust lanes (Hawar-
den et al. 1981, Ebneter & Balick 1985; Bertola 1987; V´ eron-
Cetty & V´ eron 1988). Since the dust lanes presumably rep-
resent highly-inclined structures, the optically-obscured el-
liptical galaxies were often confused with lenticular galaxies,
although (almost) no trace of a stellar disc was later identi-
fied in their CCD images. However, it seems at present, in
many aspects, less important whether to call such objects
elliptical or lenticular galaxies; we therefore refer to them as
transitional dusty ‘E/S0’ galaxies.
Other objects are selected from more recent studies of
E/S0 galaxies with dust lanes, including the search for cold
gas (Gregorini, Messina & Vettolani 1989; M¨ ollenhoff, Hum-
mel & Bender 1992; Wang, Kenney & Ishizuki 1992) and
the determination of the extragalactic dust extinction law
in the dark lanes (Goudfrooij et al. 1994b; Patil et al. 2007;
Finkelman et al. 2008; Finkelman et al. 2010b). The sample
galaxies are listed in Table 1 with their coordinates, morpho-
logical classification, integrated blue luminosity, heliocentric
velocity and optical size taken from NED.
As part of our on-going programme we observed 20
galaxies with broad-band and Hα narrow-band filters. The
observations were conducted during 2009 and 2010 with
the 1.8-m Vatican Advanced Technology Telescope at the
Mt. Graham International Observatory (MGIO), the 1.9-m
telescope at the South African Astronomical Observatory
(SAAO) and the Wise Observatory 1-m telescope. Optical
and near-IR images of NGC 5128 were taken with the CTIO
0.9-m telescope in May 2001 and with the infrared survey
facility 1.4-m telescope in SAAO in April 2010, respectively.
Typical exposure times used for the galaxies in our sample
were 10 min for the optical and near-IR filters and 20 min
for the Hα narrow-band filters.
Detailed descriptions of data reduction steps, flux cali-
bration and error propagation are available in Finkelman et
al. (2010a). Below we briefly summarize these steps.
Image reduction was performed with standard tasks
within IRAF1. These include bias subtraction, overscan sub-
traction and flatfield correction. Cosmic rays events were re-
moved from single CCD exposures by using the L.A.Cosmic
task in IRAF (van Dokkum 2001) while CCD hot pixels were
removed with the FIXPIX task in IRAF using an appropri-
ate mask.
The reduced images are background-subtracted and ge-
ometrically aligned by measuring centroids of several com-
mon stars in the galaxy frames. This alignment procedure
involves IRAF tasks for scaling, translation and rotation of
the images, so that a small amount of blurring is introduced
affecting the accuracy to be better than a few tenths of an
arcsec. We observed each galaxy using two filters, a narrow-
band filter which covers the rest-frame Hα+[NII] emission,
and a broad R-band filter. The Hα images include photons
from the Hα and the [NII] lines and from the continuum.
Since we are interested in the Hα line, the R-band image
is scaled to match the intensity of the stellar continuum in
the narrow-band image and is subtracted from the narrow-
band image. Finally, the measured Hα+[NII] counts are con-
1IRAF is distributed by the National Optical Astronomy Obser-
vatories (NOAO), which is operated by the Association of Uni-
versities, Inc. (AURA) under co-operative agreement with the
National Science Foundation

Page 3

Dust and Ionized Gas Association in E/S0 Galaxies with Dust Lanes: Clues to their Origin
3
Table 1. Global parameters for galaxies in our sample.
ObjectRA DEC
(J2000.0)
Morph.
(NED)
B0
T
vHelio
(km/s)
SizeObservatory
(J2000.0)(mag) (arcmin)
NGC 524
NGC 984
NGC 1172
NGC 1439
NGC 1947
ESO 159-G019
NGC 2076
ESO 087-G028
NGC 2907
NGC 2911
NGC 3302
NGC 3489
NGC 3497
NGC 4753
NGC 5128
NGC 5266
IC 4320
ESO 384-G012
NGC 5525
NGC 7722
01:24:48
02:34:43
03:01:36
03:44:50
05:26:48
05:33:11
05:46:47
06:33:19
09:31:37
09:33:46
10:35:47
11:00:19
11:07:18
12:52:22
13:25:28
13:43:02
13:44:04
13:55:34
14:15:39
23:38:41
+09:32:20
+23:23:47
-14:50:12
-21:55:12
-63:45:36
-52:38:31
-16:46:57
-62:59:39
-16:44:05
+10:09:09
-32:21:31
+13:54:04
-19:28:18
-01:11:59
-43:01:09
-48:10:10
-27:13:54
-33:54:01
+14:16:57
+15:57:17
SA0
SA0
E
E
S0 pec
S0/a
S0
S0
SAa LINER?
SA0 Sy LINER
SA0
SAB0 Sy2
SA0
I0
S0 pec Sy2
SA0 LINER
S0
S0 pec
S0
S0/a
11.3
13.8
13.3
13.0
11.7
14.4
14.0
14.5
12.7
12.5
13.5
11.1
13.0
10.9
7.8
12.1
14.2
14.3
13.8
13.5
2379
4352
1669
1670
1100
4338
2142
8444
2090
3183
4075
677
3672
1239
547
3002
6805
4577
5553
4026
2.8x2.8
3.0x2.0
2.3x1.8
2.5x2.3
3.0x2.6
1.5x0.8
2.2x1.3
1.1x0.7
1.8x1.1
4.1x3.2
1.7x1.2
3.5x2.0
2.6x1.4
6.0x2.8
25.7x20.0
3.2x2.1
1.0x1.0
1.1x0.9
1.4x0.9
1.7x1.2
MGIO
MGIO
MGIO
WO
WO
SAAO
SAAO
SAAO
SAAO
WO
SAAO
WO
SAAO
WO
CTIO, SAAO
SAAO
SAAO
SAAO
WO
WO
verted into physical units by observing standard stars (Lan-
dolt 1992).
Our analysis also include archival images taken from
the Two Micron All Sky Survey (2MASS), the Wide-field
Infrared Survey Explorer (WISE) and the Infrared Astro-
nomical Satellite (IRAS). All images were reduced and flux-
calibrated using the standard pipelines. We note that the
WISE Preliminary Release covers only about half of the sky
and therefore does not include many of our objects.
3 ANALYSIS AND RESULTS
3.1Dust grain properties
Dark lanes are produced by the absorption and scattering
of optical light by dust grains in the dusty structure. The
amount of extinction and its effect on observed colours de-
pends strongly on the size distribution, structure and chem-
ical composition of the grains. Fitting the unextinguished
parts of an E/S0 galaxy with an underlying smooth light dis-
tribution allows the estimation the dust extinction in these
regions. This is done by fitting the galaxies with elliptical
isophotes using the ISOPHOTE package in IRAF and sub-
tracting the observed galaxy image from the dust-free model
to measure the extinction. Applying this method is useful for
characterizing the dust properties and estimating the dust
mass.
The wavelength dependence of extinction in almost
all E/S0 galaxies with dust lanes follows closely the stan-
dard Galactic extinction law, showing only small departures
(Goudfrooij et al. 1994; Patil et al. 2007; Finkelman et al.
2008, 2010a, 2010b). For those galaxies in our sample not
included in previous similar studies we validated this result
by measuring the extinction of light in the optical bands,
with the exception of NGC 5128 which was observed in the
near-IR (see also Kainulainen et al. 2009). Considering the
well-studied properties of Galactic dust grains and their sim-
ilarity with dust grains in dust lanes we conclude that the
typical size of the (large form of) dust grains in our sample
galaxies is about ∼0.1 µm (see Draine & Lee 1984; Casuso &
Beckman 2010). The energy absorbed by these dust grains
is re-emitted in far-IR and sub-mm wavelengths.
3.2Dust mass
3.2.1Optical extinction
The dust mass is calculated by integrating the dust column
density Σd over the image areas S occupied by dust lanes.
The dust column density can be estimated from the total
extinction values by assuming a chemical composition of the
extragalactic dust grains similar to that of the dust in the
Galaxy. While the details of these calculations are described
in Finkelman et al. (2008), a useful approximation of the
dust mass can be given as
Md= S×Σd= S×ld×nH×
a+
?
a−
4
3πa3ρdf (a)da.(1)
where a is the grain radius; f (a) represents the size distribu-
tion of dust grains; a−and a+represent the lower and upper
cutoffs of the size distribution, respectively; ρd is the spe-
cific grain densities taken to be ∼ 3 gr cm−3(Draine & Lee
1984); ld represents the dust column length along the line
of sight; and nH is the hydrogen number density. Note that
we make no assumption about the gas-to-dust relationship
to estimate the hydrogen column density. Instead, the value
of the product ld× nH is inferred by measuring the total
extinction and calculating the extinction efficiency (Finkel-
man et al. 2008). This measurement provides only a lower
limit to the true dust content of the host galaxies since our
calculations assume that the dust forms a foreground screen
for the galaxy.

Page 4

4
Ido Finkelman et al.
Table 2. Dust mass and ionized gas mass derived from optical images. Masses are given in Solar units.
ObjectLog
?Mdust,opt
M⊙
?
Log
?Mdust,IRAS
M⊙
?
Log
?
MHII
M⊙
?
NGC 524ab
NGC 984
NGC 1172ac
NGC 1439a
NGC 1947d
ESO 159-G019
NGC 2076e
ESO 087-G028
NGC 2907
NGC 2911
NGC 3302d
NGC 3489
NGC 3497f
NGC 4753dg
NGC 5128
NGC 5266d
IC 4320
ESO 384-G012
NGC 5525
NGC 7722
4.58 ± 0.14
5.46 ± 0.01
3.83 ± 0.18
5.10 ± 0.19
5.01 ± 0.01
5.62 ± 0.01
6.18
6.26 ± 0.01
6.13 ± 0.01
5.80 ± 0.02
5.55 ± 0.02
4.00 ± 0.01
6.03 ± 0.03
5.20 ± 0.01
6.58 ± 0.08
5.04 ± 0.03
6.15 ± 0.01
5.64 ± 0.02
6.23 ± 0.01
6.42 ± 0.01
5.58 ± 0.21
4.25 ± 0.47
< 5.13*
< 5.77*
5.46 ± 0.06
6.52 ± 0.07
6.59 ± 0.08
< 7.02*
5.31 ± 0.22
5.76 ± 0.26
6.00 ± 0.23
-
6.34 ± 0.29
5.96 ± 0.09
6.03 ± 0.07
6.16 ± 0.02
< 6.80*
4.48 ± 0.88
-
6.64 ± 0.15
3.60
5.28 ± 0.25
3.42
< 3.48
3.47 ± 0.14
5.41 ± 0.05
4.53 ± 0.09
< 4.69
< 3.83
4.67 ± 0.22
< 4.65
3.71 ± 0.18
5.13 ± 0.11
5.43
4.99 ± 0.01
5.29 ± 0.13
< 4.85
< 4.37
< 4.75
5.07 ± 0.19
*Upper limits assuming T = 20K
References: a - Patil et al. (2007); b - Sarzi et al. (2006); c - Macchetto e tal. (1996); d - Finkelman et al. (2010b); e - Sahu et al.
(1999); f - Finkelman et al. (2008); g - Dewangan et al. (1999).
3.2.2 Thermal emission
The average temperature of the dust in each galaxy can
be estimated by fitting modified blackbody functions to the
IR data. In the wavelength range λ ? 20µm, the emission of
isothermal dust grains can be modeled with a dust emissivity
law given by Fν ∝ Bν(Td)νβ, where Fν and Bν(Td) are the
flux density and the Planck function for the temperature Td
at wavelength ν, respectively. The dust emissivity index β
is generally taken to be between 1.0 and 2.0 (Draine & Lee
1984; Genzel & Cesarsky 2000).
Given a β ≈ 1.0 emissivity law, the dust grain tempera-
ture can be calculated from the IRAS flux densities at 60 µm
?
cold dust was already known to reside in normal, inactive
spiral galaxies, IRAS observations established the presence
of ‘warm’ (∼30-40 K) dust which dominates the 60 µm and
100 µm bands. However, IRAS did not cover the spectral
range beyond 100 µm and was therefore insensitive to dust
colder than ∼20 K.
The far-IR emission of E/S0 galaxies can be used to
estimate their dust content independently of the optical ex-
tinction and reddening measurements. Following the analy-
sis of Hildebrand (1983), the dust mass is given by
and 100 µm using Td=
S60
S100
?0.4
(Young et al. 1989). While
Mdust,IRAS=4
3aρdD2
Fν
QνBν(Td)
(2)
where D is the distance of the galaxy in Mpc and Qν is the
grain emissivity taken from Draine (1985). Table 2 lists the
estimated dust mass from the total optical extinction and
the estimated dust mass based on IRAS flux densities taken
from the catalog of Knapp et al. (1989) for bright early-type
galaxies.
Recent observations of early-type galaxies have detected
also a significant emission in the near- and mid-IR wave-
15 1050 100
10
41
10
42
10
43
10
44
Rest−frame wavelength (µm)
λLλ (erg s−1)
200K
3750K
35K
Figure 1. Spectral energy distribution (SED) of NGC 5128. Tri-
angles, squares and circles represent 2MASS, WISE and IRAS
data, respectively. The SED is well reproduced by the thermal
emission of a dominant stellar population of spectral type M0,
warm dust at 35 K and hot dust at 200 K.
lengths. The emission in these wavelengths is believed to
be produced by very small dust grains and PAH molecules
(Ferrari et al. 2002; Xilouris, Madden & Vigroux 2004). In
particular, the emission at ∼25-60 µm is probably produced
by the transient heating of ∼0.001 µm dust grains to high
temperatures (Li & Draine 2001). While the exact treat-
ment of this emission is complex, it can be approximated
as continuous thermal emission of ‘hot dust’ of ∼200 K (see
Ferrari et al. 2002).
To study the hot dust we use publicly available near-IR
data taken with 2MASS and mid-IR data taken with WISE
to build the spectral energy distribution (SED) and then

Page 5

Dust and Ionized Gas Association in E/S0 Galaxies with Dust Lanes: Clues to their Origin
5
34567
3
4
5
6
7
log(Mdust,opt)
log(Mdust,IRAS)
Figure 2. Dust mass derived using IRAS data versus dust mass
derived using optical extinction and reddening. Our sample galax-
ies are represented by filled circles, open circles are from Goud-
frooij et al. (1994a, 1994b) and filled squares are from Patil et al.
(2007). Dust masses are given in Solar units. The drawn solid line
is for equal masses.
follow the analysis by Ferrari et al. (2002). Since the hot
dust thermal emission typically peaks at around 10 µm, and
since up to ∼15 µm it includes a strong contribution of late-
type stars and possibly of silicate or PAH bands (Boselli et
al. 1998; Madden et al. 1999), measurements at longer wave-
lengths would serve as a more sensitive probe of the hot dust.
We therefore measure the near- and mid-IR light within an
elliptical aperture enclosing the apparent emission at 22 µm
(W4 - the reddest channel of WISE). The aperture limits
are set to include resolution elements with emission ∼1-σ
above the background level. For galaxies with no apparent
extended emission structure, a circular aperture with a ra-
dius twice the FWHM seeing size around the galactic centre
is examined in an attempt to detect faint dust emission.
Measuring the hot dust emission requires first to esti-
mate the relative contribution of the Rayleigh-Jeans tail of
the old stellar population. To this purpose we match the
SED with blackbody curves of three different temperatures
(T=3750 K, T=4000 K and T=4600 K) that correspond to
stellar populations of spectral types M0, K7 and K4. Note
that optical images of our sample galaxies suffer significant
internal extinction and are therefore not used to construct
the SED. To model the hot dust emission and determine its
temperature we add a second component with a modified
blackbody curve of Tdin the range 40 K to 500 K. As a rep-
resentative case, we show in Fig. 1 the best-fit SED of NGC
5128. Fig. 1 includes also a third component of a modified
blackbody curve representing the warm dust component de-
tected by IRAS. Although the warm dust component seems
to contribute to some extent to the overall 22 µm emission
of NGC 5128, we note that the IRAS spatial resolution is
typically several times larger than the size of our galaxies,
thus not accounted for by our modeling.
We use the flux density at 22 µm, corrected by subtract-
ing the stellar component, to estimate the hot dust mass
Mdust,WISEfor each galaxy. This is done by using eq. 2
under the assumption that the quantity aρd/Qν is indepen-
dent of a for λ ≫ a (see Hildebrand 1983; Draine & Lee
1984). The emissivity of the small grains was adjusted to be
consistent with β ≈ 1, as assumed above (see also Bianchi,
Davis & Alton 1999; Bendo et al. 2003; da Cunha, Charlot
& Elbaz 2008).
The dust masses estimated based on the mid-IR data
are listed in Table 3 along with the total luminosity at 22
µm and the best-fit hot dust temperature. In addition, we
compare the dust masses derived using reddening and opti-
cal extinction with those using IRAS flux densities and with
those derived using WISE flux densities and plot our results
in Figs. 2 and 3, respectively. To improve the coverage of
these diagrams, we include also data from previous studies
of similar objects (Goudfrooij et al. 1994; Ferrari et al. 1999;
Martel et al. 2004; Patil et al. 2007; Finkelman et al. 2008,
2010a, 2010b).
3.3 Ionized gas mass
The Hα+[NII] flux of each galaxy is measured following the
procedure detailed in Finkelman et al. (2010a). Assuming
case B recombination (Osterbrock 1989), the Hα luminosity
LHαcan be used to roughly estimate the mass of ionized
hydrogen. For a given electron temperature and density this
mass can be written as:
MHII=?LHαmH/ne
where mH is the mass of the hydrogen atom; ne and np are
the number of electrons and protons per cm3and jHαis
the emission coefficient of the Hα line. Assuming an elec-
tron temperature of ∼ 104K the expression above can be
simplified as:
?/?4πjHα/nenp
?
(3)
MHII= 2.33×103
?
LHα
1039erg s−1
??103cm−3
ne
?
M⊙. (4)
We estimate the masses assuming a typical electron density
of ∼ 103cm−3(see Goudfrooij et al. 1994a) and list the
results in Table 2.
To properly derive LHαrequires measuring also the
ratios of the [NII] and Hα lines, which are not available
here. The ionized hydrogen masses are therefore upper lim-
its, while the true values are likely lower by a factor of ∼2-3
(see Goudfrooij et al. 1994a; Macchetto et al. 1996; Finkel-
man et al. 2010a). We compare the ionized gas masses with
the dust masses derived using optical extinction and with
those derived using WISE flux densities and plot our results
in Figs. 3 and 4, respectively. Note that the Hα luminosity
values are not corrected for extinction.
4 DISCUSSION
This is our second paper presenting new results from an
on-going programme to study the ionized gas and dust in
E/S0 galaxies. We are motivated by previous ISM surveys
revealing ionized gas in nearly every galaxy where dust is
found (e.g., Goudfrooij et al. 1994a, 1994b; Macchetto et
al. 1996; Martel et al. 2004; Sarzi et al. 2010). A general

Page 6

6
Ido Finkelman et al.
01234
3
4
5
6
7
log(Mdust,WISE)
log(Mdust,opt)
log(MHII)
3
4
6
7
5
Figure 3. The relation between dust mass derived using optical
extinction and dust mass derived using WISE data (open sym-
bols) and the relation between ionized gas mass and dust mass
derived using WISE data (filled symbols). Circles represent our
data; squares, triangles and ‘+’ signs represent data from Goud-
frooij et al. (1994a, 1994b), Ferrari et al. (1999) and Dewangan
et al. (1999), respectively. Masses are given in Solar units. The
hot dust content tends to be more massive with increasing optical
dust mass; a Spearman rank test shows that the null hypothesis
of no correlation has a probability of only about 1%.
34567
3
4
5
6
7
2
log(Mdust,opt)
log(MHII)
Figure 4. HII mass versus dust mass. Our sample galaxies are
represented by filled circles, open circles are from Goudfrooij et al.
(1994a, 1994b), filled triangles are from Macchetto et al. (1996)
and ‘+’ signs are from Martel et al. (2004). Masses are given in
Solar units. The dashed line is the fitted linear trend with a slope
of ∼ 0.8.
Table 3. Hot dust properties including the best-fit temperature,
total luminosity in the WISE 22µm (W4) band and derived hot
dust mass. The typical errors on luminosity and dust mass are
estimated to be less than 5%. The number in parenthesis represent
the relative contribution of dust emission to the 22µm flux. The
table includes galaxies from our sample and similar studies for
which WISE data are currently available.
Object
Td
(◦K)
L(22µm)
(1030erg s−1)
Mdust,WISE
(M⊙)
NGC 662
NGC 984
NGC 1199
NGC 1407a
NGC 1439
NGC 1600a
ESO 159-G012
NGC 2076b
NGC 2534
NGC 3136a
IC 3370a
NGC 4589a
NGC 4696a
NGC 5018a
NGC 5044a
NGC 5128
NGC 5266
IC 4320
ESO 384-G012
NGC 5525
NGC 5576a
AM 1444-302
AM 1459-722
NGC 5799
NGC 5812c
NGC 5813a
NGC 5903c
NGC 6251
NGC 6314
NGC 6482a
NGC 6483c
IC 4797c
NGC 6758c
IC 4889c
NGC 6909c
180
220
160
200
180
200
220
200
240
200
200
200
200
180
200
200
220
220
240
240
180
200
220
240
180
180
180
160
220
220
160
180
200
200
180
14.27 (1.00)
0.40 (0.84)
0.22 (0.78)
0.16 (0.66)
0.05 (0.77)
0.28 (0.55)
4.50 (0.99)
4.75 (0.99)
0.45 (0.93)
0.13 (0.74)
0.50 (0.85)
0.12 (0.68)
0.28 (0.69)
0.90 (0.86)
0.19 (0.70)
8.52 (0.99)
1.49 (0.89)
0.46 (0.82)
0.23 (0.88)
0.81 (0.88)
0.03 (0.55)
3.38 (0.97)
0.14 (0.80)
0.41 (0.88)
0.10 (0.67)
0.09 (0.63)
0.09 (0.58)
5.53 (0.91)
4.92 (0.96)
0.34 (0.66)
0.21 (0.64)
0.41 (0.72)
0.14 (0.62)
0.18 (0.73)
0.08 (0.70)
2930
41
73
22
11
39
480
670
36
18
70
17
39
180
27
1160
150
48
18
64
6
480
14
33
21
18
18
1810
510
35
69
85
20
25
16
Note: Ionized gas and optical dust mass estimates used for
Fig. 3 are taken from: a - Goudfrooij et al. (1994a, 1994b); b
- Dewangan et al. (1999); c - Ferrari et al. (1999).
result, which holds also for our sample galaxies, is that the
dust and ionized gas almost always strongly correlate while
spread over scales from few hundred parsecs to several kpc.
Hence, any theory of the origin and evolution of the ISM
needs to account also for the co-existence and co-evolution
of these two components with their various phases.
The surprisingly high detection rate of ISM in E/S0
galaxies requires a clarification. In late-type galaxies the
dust is formed primarily in cold, dusty clouds where it is de-
posited into the ISM by condensation of thermally-unstable
material from the coronal phase or is injected by mass-loss
from evolved red giant stars (Dwek & Scalo 1980). Although
this is also an obvious source of ISM in elliptical galaxies,
their overall dust content could be reduced depending on the
grain destruction mechanisms in action. On the other hand,

Page 7

Dust and Ionized Gas Association in E/S0 Galaxies with Dust Lanes: Clues to their Origin
7
some material could be accreted onto an elliptical galaxy
by tidally capturing a gas-rich spiral or dwarf, or through a
collision of two gas-rich galaxies of comparable mass.
Todistinguishbetween
externally-acquired ISM considerable efforts were inves-
tigated in studying the multiphase gas in E/S0 galaxies.
Here, however, we rely primarily on the properties of the
dust to draw conclusions on the origin of the ISM. While
modern multi-wavelength surveys show that the presence of
dust in E/S0 galaxies is the rule rather than the exception,
note that the dust distribution is usually concentrated
on the nucleus and extends out to radii of a few hundred
parsecs, rather than spread throughout the galaxy as in
most of our sample galaxies (Sarzi et al. 2006; Finkelman
et al. 2010a) . Such nuclear dust structures are usually
associated with well-defined gaseous rings or discs that are
often oriented parallel to the major axis and trace separate
stellar components (Krajnovi´ c et al. 2008; Krajnovi´ c et
al. 2011). The distribution of mean misalignment between
the stellar and gaseous angular momenta of these inner
structures is inconsistent with a purely external origin,
implying a good balance between gas that has been accreted
and internally recycled (Sarzi et al. 2006).
Our sample does not pretend to be representative of
E/S0 galaxies, but rather of a distinct subclass of E/S0
galaxies that exhibit prominent, large-scale dust features in
their optical images. However, since we aim to test whether
the tight relationship between gas and dust holds over vari-
ous spatial scales, we refer in our study also to E/S0 galaxies
with inner, sub-kpc dust components that were previously
suggested of being externally accreted (see e.g. Martel et al.
2004).
Many of the galaxies in our sample show signs of recent
interaction, with NGC 5128 being the most famous example
(see Finkelman et al. 2010a). Searching NED and known cat-
alogues of groups of galaxies (Huchra & Geller 1982; Geller
& Huchra 1983; Garcia 1993; Tago et al. 2010) we find no
correlation between the amount of dust and the environment
of the galaxies (see also Tran et al. 2001; Patil et al. 2007).
However, Kaviraj et al. (2011) did find that the environ-
ments of dusty early-type galaxies are typically less dense
than those of systems with no sign of optical dust obscura-
tion. Kaviraj et al. found also that dusty early-type galaxies
are much more likely to host star formation and nuclear
activity than early-type galaxies with no prominent dust
lanes. Furthermore, both the active galactic nucleus (AGN)
and starburst ages were found to be younger in dust-lane
early-type galaxies than in other early-type galaxies, im-
plying these were triggered by gas-rich galaxy interactions
(Shabala et al. 2011). In this view, relaxed gaseous discs and
chaotic filamentary structures, and different rates of star for-
mation and black hole accretion rates, could represent dif-
ferent states of the same type of event. We show below that
the grain size distribution, content and spatial distribution
of dust in our galaxies support this picture.
internally-producedand
4.1 Grain size distribution
While the presence of dust in a cold ISM environment of
late-type galaxies is natural, hot coronal gas in E/S0 galax-
ies will erode the dust grains via thermal sputtering, or even
destroy them, depending on the grain size. This will change
the dust size distribution and with it the wavelength depen-
dence of the extinction. Other mechanisms, such as grain-
grain collisions, are less efficient in destroying dust in galax-
ies embedded in hot gas (see Goudfrooij 1999).
Studying the wavelength dependence of extinction in
E/S0 galaxies implies that the dust grain properties are re-
markably similar to those of dust grains in the Galaxy with
an average grain size of ∼0.1 µm (Finkelman et al. 2008;
2010a, 2010b). If grains condense out of a cooling flow they
would be expected to be destroyed within less than ∼100
Myr. However, such a destruction rate cannot account for
the grain size or for the total build-up of dust mass observed
in this class of galaxies (Patil et al. 2007; Casuso & Beckman
2010; Clemens et al. 2010; Kaneda et al. 2011). While several
studies point that the mean grain size is somewhat smaller
for galaxies with well-ordered dust lanes (e.g., Goudfrooij et
al. 1994; Patil et al 2007), note that sputtering destroys the
smallest grains first, hence actually increasing the average
grain size. Furthermore, as demonstrated for NGC 5128 in
Figs. 1 and 5, we detected in our sample the presence of
very small (a ? 0.01µm) grains by their mid-IR emission
and found that their distribution is associated with the op-
tical dust. In fact, Fig. 3 shows that the masses of these
two dust components also correlate; the Spearman rank test
shows that the null hypothesis of no correlation has a prob-
ability of only about 1%. Previous near- and mid-IR stud-
ies detected also PAH features in a large fraction of (X-ray
emitting) E/S0 galaxies where considerable amounts of dust
are clearly present (Xilouris et al. 2004; Pahre et al. 2004;
Kaneda, Onaka & Sakon 2005; Kaneda et al. 2010; Panuzzo
et al. 2011). According to these findings, the dust grain dis-
tribution seems to be indifferent to the presence of hot gas.
4.2 Dust content
A large part of our sample galaxies are not listed in
catalogues of X-ray sources, probably due to the rela-
tively low sensitivity of large X-ray surveys of hot gas
in elliptical galaxies. While these shallow surveys paid
considerable attention to X-ray bright ‘giant’ elliptical
galaxies, deep Chandra and XMM-Newton observations
revealed substantial amounts of hot gas in ‘normal’ ellip-
tical galaxies as well (see e.g., Mathews & Brighenti 2003;
Diehl & Statler 2007; Boroson, Kim & Fabbiano 2011).
As a general rule, less massive galaxies are expected to
be less able to retain the hot gas, which could escape
the gravitational potential of the galaxy. However, the
hot gas content probably depends to some extent also on
the evolutionary history of a galaxy. For instance, E/S0
galaxies that have undergone a recent merger have low
X-ray luminosities while relaxed E/S0 galaxies are stronger
X-ray emitters (Sansom, Hibbard & Schweizer 2000). This
can be explained if hot gas halos build up through mass
loss from stars over a timescale of several Gyr after a merger.
Our sample galaxies lie in diverse environments, cover
a wide range of size and have probably experienced differ-
ent merger histories. It is therefore reasonable to assume
that the hot gas content varies significantly between these
objects. On one hand, observations of ionized gas in E/S0
galaxies where x-ray halos were not detected show that
at least some of the ‘warm’ ionized gas does stay within

Page 8

8
Ido Finkelman et al.
∆α (arcmin)
∆δ (arcmin)
420 −2 −4
−4
−2
0
2
4
∆α (arcmin)
∆δ (arcmin)
42 0 −2−4
−4
−2
0
2
4
Figure 5. Left panel: 22µm map of NGC 5128, shown as contours superposed on our R-band image of the galaxy. Right panel: Hα
continuum-subtracted image of NGC 5128. Both images are displayed as negative.
the galaxy potential well and is often accompanied by dust
(Macchetto et al. 1996). On the other hand, massive, X-ray
bright galaxies are able to retain their warm gas content,
while the internally produced dust is subject to rapid ero-
sion by sputtering. Therefore, internally-produced ionized
gas is expected to anti-correlate with the dust component.
To test this, we plot in Fig. 4 the ionized gas mass ver-
sus the dust mass, including also data from previous stud-
ies (Goudfrooij et al. 1994a, 1994b; Macchetto et al. 1996;
Martel et al. 2004). The two components have comparable
masses over about 4 orders of magnitude, spreading over
spatial scales from the nuclear regions to extended struc-
tures several kpc away from the centre. The ionized gas spa-
tially correlates not only with the optically obscured regions,
but also with the hot dust emission, which suggests that the
source of ionization and dust heating is the same (see Figs. 3
and 5 here, and figs. A1-8 in Finkelman et al. 2010b). These
findings, therefore, provide important evidence against an
internal origin of these components. In fact, since the ob-
served mass relation seems to be independent of the hot gas
content of each galaxy, there must be a mechanism to protect
dust grains embedded in hot ambient gas from destruction.
The dust grain lifetime could be much longer for mate-
rial acquired via an interaction as part of a cooler medium
that is (partly) isolated from the hot medium. Inside dense
molecular clouds, dust would be self-shielded against the
diffuse galactic UV radiation and the hot thermal electrons
in the adjacent hot gas (Temi, Brighenti & Mathews 2007).
On the other hand, other ionizing sources, such as AGN
or star formation activity, could disperse and heat some of
the dusty gas. These conditions could possibly explain why
dust is heated but not destroyed, even in the presence of hot
plasma. Shocks triggered by cloud-cloud interactions during
accretion of cold gas could produce also a multiphase dusty
gas composed of a dust-free, X-ray emitting plasma and a
cold component of mixed dust and molecular gas (Guillard
et al. 2009). However, whether shock-heating can account
for the observed line-emission characteristics is still unclear
(Tang et al. 2011). If the gas and dust in the ionized regions
are not the product of ISM cooling processes, they could be
part of the original warm phase of the accreted ISM. Pos-
sible evidence for such a process was given by Kauffmann,
Li & Hickman (2010) who suggested that satellite galaxies
are likely to trace an underlying reservoir of ionized gas that
can be accreted onto the host galaxy.
Although E/S0 galaxies were once thought to be almost
entirely devoid of cold gas, we know by now that the molec-
ular gas in E/S0 galaxies in invariably associated with dust
lanes (Wang et al. 1992; Sage & Galletta 1993; Combes et
al. 2007; Krips et al. 2010; Young et al. 2011). A correlation
between the cold molecular gas reservoir and the cold dust
content was also established for far-IR selected early-type
galaxies (see Kaviraj et al. 2011 and references therein).
However, we caution that a significant part of the far-IR
emission originates probably from dust distributed through-
out the galaxy and heated by the ambient stellar radiation
field (see Section 4.3). This diffuse dust component is asso-
ciated with the diffuse atomic gas phase rather than with
the dense molecular gas.
The molecular gas in E/S0 galaxies does not correlate
with the underlying stellar population implying it was most
likely externally accreted (Wiklind, Combes & Henkel 1995).
Several E/S0 galaxies with dust lanes were also found to con-
tain significant amounts of cold neutral hydrogen, generally
distributed in an extended disc-like structure of low surface
density (Oosterloo et al. 2002; Serra et al. 2008). While the
latter is believed to be a result of an accretion event, the
HI-to-dust ratios measured are considerably lower than the
gas-to-dust ratio typical of spiral galaxies. It is therefore
possible that much of the gas in these galaxies could be in
molecular rather than atomic form. Since molecular gas nor-
mally has a higher surface density than atomic gas, it falls

Page 9

Dust and Ionized Gas Association in E/S0 Galaxies with Dust Lanes: Clues to their Origin
9
deeper in the galaxy potential well and is more difficult to
remove by interactions with neighboring galaxies (Young et
al. 2011). Part of the atomic gas could also be heated by
hot plasma and later cool to the cold neutral medium tem-
peratures, condense and become molecular. Consequently,
the cold gas in the central regions could be dominated by
molecular gas (Oosterloo et al. 2010).
4.3Dust distribution
Studying NGC 5128 in the far-IR to sub-mm wavelength
range Leeuw et al. (2002) derived a total dust mass of
2.2 × 106M⊙ within 225 arcsec of the nucleus. This value
is consistent with our estimate of the dust mass from opti-
cal extinction. However, far-IR observations of more distant,
less-resolved galaxies indicate that the entire dust mass in
E/S0 galaxies is generally at least an order of magnitude
higher than estimated by optical extinction, as illustrated
in Fig. 2. The discrepancy can be explained by the presence
of a diffuse, massive dust component which light attenuation
is difficult to detect (Goudfrooij & de Jong 1995). Signifi-
cant reservoirs of dust might also be located in the central
parts of these galaxies (Temi et al. 2007; Leeuw et al. 2008).
The distribution of the optical dust is difficult to ex-
plain purely by internal processes. If dust is produced by
stellar mass loss, it is expected to follow the stellar light dis-
tribution, rather than lie in well-organized structures (see
for instance Athey et al. 2002; Bregman & Athey 2004).
AGN outflows disrupting the dense central dusty structure
in the galactic core could generate morphologically chaotic
or asymmetric obscuring dust features away from the cen-
tre, but not well-settled dust discs (Temi et al. 2007). Fur-
thermore, Bregman et al. (1998) argued that the observed
mass discrepancy between the two dust components can-
not be accounted for in hot gas environments if the ISM
was internally-generated, unless the grain sizes are typically
much higher than for the Galaxy. The authors concluded
also that the ratio between the masses derived by IRAS data
and optical extinction would appear to grow after a merger.
We note that there seems to be no clear correlation be-
tween far-IR emission or dust content and blue luminosity
in early-type galaxies (e.g., Forbes 1991; Goudfrooij & de
Jong 1995; Trinchieri & Goudfrooij 2002; Temi et al. 2007;
Kaviraj et al. 2011; Smith et al. 2011). However, since pre-
vious far-IR surveys (e.g., IRAS) were not sensitive to dust
colder than ∼ 20 K, testing this correlation and the presence
of different dust components requires detecting the thermal
emission of the very cold dust phase at sub-mm wavelengths.
To establish what is the dominant source of dust in E/S0
galaxies requires also mapping its emission for a represen-
tative sample of galaxies (Temi et al. 2004; Xilouris et al.
2004). The details of dust grinding processes that could de-
stroy such a correlation should also be accounted for.
Finally, although we showed throughout this paper that
the association of gas and dust argues strongly against their
internal origin, explaining why externally-acquired dust and
ionized gas have comparable masses is far from trivial. Con-
sidering the various ionization mechanisms in action over
different spatial scales makes it difficult to account for such
a relation in any scenario (see Sarzi et al. 2010). Moreover,
if the ISM is acquired from a gas-rich dwarf or through a
merger between spiral galaxies, then the overall gas mass is
expected to be about 2-3 order of magnitude higher than the
overall dust content (e.g., Walter et al. 2007; Mu˜ noz-Mateos
et al. 2009). Where is the rest of the gas and, assuming that
the ionization sources should not be confined to the dust
lane, why do we not see significant amounts of dust-free ion-
ized gas? Solving these issues seems to require making a
full inventory of the multiphase ISM ingredients, which is
beyond the scope of our study.
5CONCLUSIONS
We present broad-band and narrow-band images to study
the content of dust and ionized gas in E/S0 galaxies with
dust lanes. We find that:
(1) the grain size distribution typically follows that of the
Galaxy.
(2) the dust mass, measured from the optical extinction by
‘large’ dust grains, ranges from 103to 107M⊙;
(3) the hot dust mass, measured from the 22µm emission
from ‘small’ dust grains, ranges from 1 to 3000 M⊙;
(4) the cold dust mass, measured from IRAS data, is typi-
cally much higher than the optical dust mass.
(5) the ionized gas content strongly correlates with that of
the optical dust;
(6) the ionized gas is morphologically associated with the
optical dust structure and the hot gas distribution;
We argue that these observed relations indicate that the
ionized gas and the obscuring material have the same ori-
gin, are heated by the same sources and are well mixed. We
conclude that an internal origin of the dust and ionized gas
in E/S0 galaxies with dust lanes is highly unlikely; the hot
gas content of E/S0 galaxies is quite heterogeneous and ex-
pected to affect differently the grain size distribution, mass
content and dust distribution of individual galaxies, whereas
our findings are independent of the hot gas content of each
galaxy. We argue also that our results are consistent with
the ‘evaporation flow’ hypothesis, albeit with some uncer-
tainty as to the exact details of the process. If the dusty gas
that we observe in the optical is part of the dense material
arriving from outside during an accretion or merger event,
than it could survive destruction even in hot and extreme
environments. Relaxed gaseous discs and chaotic filamentary
structures represent in this picture different states of similar
events. The frequent detection of tidal features, atomic and
molecular gas and kinematically decoupled gas components
in E/S0 galaxies with dust lanes support this proposed view.
6ACKNOWLEDGEMENT
We thank Robert C. Kennicutt Jr. and Janice C. Lee for
allowing us to publish their Hα image of NGC 5128. IF
wishes to thank Rivay Mor for useful discussions. SB, AK
and PV acknowledge support from the National Research
Foundation of South Africa.
Based on observations with the VATT: the Alice P.
Lennon Telescope and the Thomas J. Bannan Astrophysics
Facility.
This research has made use of the NASA/IPAC Ex-
tragalactic Database (NED) which is operated by the Jet
Propulsion Laboratory, California Institute of Technology,

Page 10

10
Ido Finkelman et al.
under contract with the National Aeronautics and Space
Administration.
This publication makes use of data products from the
Wide-field Infrared Survey Explorer, which is a joint project
of the University of California, Los Angeles, and the Jet
Propulsion Laboratory/California Institute of Technology,
funded by the National Aeronautics and Space Administra-
tion.
This publication makes use of data products from the
Two Micron All Sky Survey, which is a joint project of the
University of Massachusetts and the Infrared Processing and
Analysis Center/California Institute of Technology, funded
by the National Aeronautics and Space Administration and
the National Science Foundation.
REFERENCES
Athey A., Bregman J., Bregman J., Temi P., Sauvage M.,
2002, ApJ, 571, 272
Bendo G.J. et al., 2003, AJ, 125, 2361
Bertola F., 1987, in IAU Symp. 127, ed. P.T. de Zeeuw
(Dordrecht: Reidel), 135
Bertola F., Buson L.M., Zeilinger W.W., 1992, ApJ, 401,
79
Bianchi S., Davies J.I., Alton P.B., 1999, A&A, 344, 1
Boroson B., Kim D.-W., Fabbiano G., 2011, ApJ, 729, 12
Boselli A., Lequeux J., Sauvage M., Boulade O., Boulanger
F., Cesarsky D., Dupraz C., Madden S., Viallefond F.,
Vigroux L., 1998, A&A, 335, 53
Bregman J.N., Snider B.A., Grego L., Cox C.V.,1998, ApJ,
499, 670
Bregman J.N., Athey A.E., 2004, in the Proceedings of
The Riddle of Cooling Flows in Galaxies and Clusters of
Galaxies, held in Charlottesville, VA, May 31 - June 4,
2003, Eds. T. Reiprich, J. Kempner, and N. Soker.
Cappellari M. et al., 2011, MNRAS, 413, 813
Casuso E., Beckman J.E., 2010, AJ, 139, 1406
Caon N., Pastoriza M., Macchetto D., 2001, ApSSS, 277,
409
Clemens M.S. et al., 2010, A&A, 518, 50
Combes F., Young, L.M., Bureau, M., 2007, MNRAS, 377,
1795
da Cunha E., Charlot S., Elbaz D., 2008, MNRAS, 388,
1595
de Jong T., Norgaard-Nielsen H.U., Jorgensen H.E.,
Hansen L., 1990, A&A, 232, 317
de Zeeuw P.T. et al., 2002, MNRAS, 329, 513
Dewangan G.C., Singh K.P., Bhat P.N., 1999, AJ, 118, 785
Diehl S., Statler T.S., 2007, ApJ, 668, 150
Draine B.T. & Lee H.M, 1984, ApJ, 285, 89
Dwek E. & Scalo J.M., 1980, ApJ, 239, 193
Ebneter K. & Balick B., 1985, AJ, 90, 183
Emsellem E. et al., MNRAS, 352, 721
Fabian A.C., Sanders J.S., Crawford C.S., Conselice C.J.,
Gallagher J.S., Wyse R.F.G., 2003, MNRAS, 344, 48
Ferrari F., Pastoriza M.G., Macchetto F., Caon N., 1999,
A&ASS 136, 269
Ferrari F., Pastoriza M.G., Macchetto F.D., Bonatto C.,
Panagia N., Sparks W.B., 2002, A&A, 389, 355
Finkelman I., et al., 2008, MNRAS, 390, 969
Finkelman I., Brosch N., Funes J.G., Kniazev A.Y.,
V¨ ais¨ anen P., 2010a, MNRAS, 407, 2475
Finkelman I. et al., 2010b, MNRAS, 409, 727
Forbes D.A., 1991, MNRAS, 249, 779
Garcia A.M., 1993, A&AS, 100, 47
Geller M.J. & Huchra J.P., 1983, ApJS, 52, 61G
Genzel R. & Cesarsky C.J., 2000, ARA&A, 38, 761
Goudfrooij P., 1999, ASPC, 163, 55
Goudfrooij P., Hansen L., Jorgensen H.E., Norgaard-
Nielsen H.U., 1994a, A&AS, 105, 341
Goudfrooij P., de Jong, T., Hansen, L., Nørgaard-Nielsen,
H. U., 1994b, MNRAS, 271, 833
Goudfrooij P., de Jong T., 1995, A&A 298, 784
Guillard P., Boulanger F., Pineau Des Forˆ ets G., Appleton
P.N., 2009, A&A, 502, 515
Hau G.K.T., Bower R.G., Kilborn V., Forbes D.A., Balogh
M.L., Oosterloo T., 2008, MNRAS, 385, 1965
Hawarden T.G., Longmore A.J., Tritton S.B., Elson
R.A.W., Corwin H.G. Jr., 1981, MNRAS, 196, 747
Hildebrand R.H., 1983, QJRAS, 24, 267
Huchra J.P., Geller M.J., 1982, ApJ, 257, 423
Kainulainen J.T. et al., 2009, A&A, 502, 5
Kaneda H., Onaka T., Sakon I., 2005, ApJ, 632, 83
Kaneda H., Onaka T., Sakon I., Kitayama T., Okada Y.,
Suzuki T., Ishihara D., Yamagishi M., 2010, ApJ, 716, 161
Kaneda H., Onaka T., Sakon I., Ishihara D., Mouri A.,
Yamagishi M., Yasuda A., 2011, EAS, 46, 157
Kavirag S. et al., 2011, arXiv1107.5306
Kauffmann G., Li C., Heckman T.M., 2010, MNRAS, 409,
491
Knapp G.R., Guhathakurta P., Kim D.-W., Jura M.A.,
1989, ApJS, 70, 329
Kormendy J., Djorgovski S., 1989, ARA&A, 27, 235.
Krips M., Pott J.-U., Eckart A., Leon S., Straubmeier C.,
2003, Ap&SS, 284, 511
Krajnovi´ c D. et al., 2008, MNRAS, 390, 93
Krajnovi´ c D. et al., 2011, MNRAS, 414, 2923
Krips M., Crocker A.F., Bureau M., Combes F., Young
L.M., 2010, MNRAS, 407, 2261
Landolt, A.U., 1992, AJ, 104, 340
Leeuw L.L., Davidson J., Dowell C.D., Hildebrand R.H.,
Matthews H.E., 2007, in Island Universes, Astrophysics
and Space Science Proceedings, Springer, p. 383
Leeuw L.L., Hawarden T.G., Matthews H.E., Robson E.I.,
Eckart A., 2002, ApJ, 565, 131
Leeuw L.L., Davidson J., Dowell C.D., Matthews H.E.,
2008, ApJ, 677, 249
Li A, & Draine B.T., 2001, ApJ, 554, 778
Macchetto F., Pastoriza M., Caon N., Sparks W. B., Gi-
avalisco M., Bender R., Capaccioli M., 1996, A&AS, 120,
463
Madden S.C., Vigroux L. & Sauvage M., 1999, ESASP, 427,
933
Martel A.R., Ford H.C., Bradley L.D., Tran H.D., Menan-
teau F., Tsvetanov Z.I., Illingworth G.D., Hartig G.F.,
Clampin M., 2004, AJ, 128, 2858
Mathews W.G., Brighenti F., 2003, ARA&A, 41, 191
McDermid R.M. et al., 2006, MNRAS, 373, 906
Micelotta E.R., Jones A.P., Tielens A.G.G.M., 2010, A&A,
510, 37
Michielsen D., de Rijcke S., Zeilinger W.W., Prugniel P.,
Dejonghe H., Roberts S., 2004, MNRAS, 353, 1293

Page 11

Dust and Ionized Gas Association in E/S0 Galaxies with Dust Lanes: Clues to their Origin
11
Mu˜ noz-Mateos J.C. et al., 2009, ApJ, 701, 1965
Oosterloo, T.A., Morganti, R., Sadler, E.M., Vergani, D.,
Caldwell, N. 2002, AJ, 123, 729
Oosterloo, T.A. et al., 2010, MNRAS, 409, 500O
Osterbrock, D.E., 1989, Research supported by the Uni-
versity of California, John Simon Guggenheim Memorial
Foundation, University of Minnesota, et al. (Mill Valley,
CA: University Science Books)
Pahre M.A., Ashby M.L.N., Fazio G.G., Willner S.P., 2004,
ApJS, 154, 229
Panuzzo P., Rampazzo R., Bressan A., Vega O., Annibali
F., Buson L.M., Clemens M.S., Zeilinger W.W., 2011,
A&A, 528, 10
Patil M.K., Pandey S.K., Sahu D.K., Kembhavi A.K.,
Singh M., 2007, A&A, 461, 103
Quillen A.C., Brookes M.H., Keene J., Stern D., Lawrence
C.R., Werner M.W., 2006, ApJ, 645, 1092
Rampazzo R., Annibali F., Bressan A., Longhetti M.,
Padoan F., Zeilinger W.W., 2005, A&A, 433, 497
Sage L.J., Galletta G., 1993, ApJ, 419, 544
Sahu D.K., Pandey S.K., Kembhavi A.K., 1999, BASI, 27,
271
Sansom A.E., Hibbard J.E., Schweizer F., 2000, AJ, 120,
1946
Sarzi M. et al., 2006, 366, 1151
Sarzi M. et al. 2010, MNRAS, 402, 218
Serra, P., Trager, S.C., Oosterloo, T.A., Morganti, R., 2008,
A&A, 483, 57
Shabala S.S., 2011, arXiv1107.5310
Smith M.W.L. et al., 2011, arXiv1112.1408
Sparks W.B., Macchetto F., Golombek D., 1989, ApJ, 345,
153
Sparks W.B., Donahue M., Jord´ an A., Ferrarese L., Cˆ ot´ e
P., 2004, ApJ, 607, 294
Tal T., van Dokkum P.G., Nelan J., Bezanson R., 2009,
AJ, 138, 141
Tago E., Saar E., Tempel E., Einasto J., Einasto M., Nurmi
P., Hein¨ am¨ aki P., 2010, A&A, 514, 102
Tang Y., Gu Q., Zhang S., Tang B., 2011, arXiv1101.0965
Temi P., Brighenti F., Mathews W.G., Bregman J.D., 2004,
ApJS, 151, 237
Temi P., Brighenti F., Mathews W.G., 2007, ApJ, 660, 1215
Tran H.D., Tsvetanov Z., Ford H.C., Davies J., Jaffe W.,
van den Bosch F.C., Rest A., 2001, AJ, 121, 2928
Trinchieri G. & Goudfrooij P., 2002, A&A, 386, 472
van Dokkum P.G., 2001, PASP, 113, 1420
van Dokkum P.G., 2005, AJ, 130, 264
V´ eron-Cetty M.-P., V´ eron P., 1988, A&A, 204, 28.
Wang Z., Kenney J.D.P., Ishizuki S., 1992, AJ, 104, 2097
Walter F. et al., 2007, ApJ, 661, 102
Wiklind T., Combes F., Henkel C., 1995, A&A, 297, 643
Xilouris E.M., Madden S.C., Galliano F., Vigroux L.,
Sauvage M., 2004, A&A, 416, 41
Young L.M. et al., 2011, MNRAS, tmp, 688
Young J.S., Xie S., Kenney J.D.P, Rice W.L., 1989, ApJS,
70, 699
Zhu G., Blanton M.R., Moustakas J., 2010, ApJ, 722, 491