PoS(LCDU 2013)064
Using Spinning Dust Emission To Constrain The
Evolution Of Dust Grains In Cold Clumps
C.T. Tibbs
∗
,
a
R. Paladini,
a
K. Cleary,
a
K.J.B. Grainge,
b
S.J.C. Muchovej,
a
T.J.
Pearson,
a
Y.C. Perrott,
c
C. Rumsey,
c
A.M.M. Scaife,
d
M.A. Stevenson
a
and J.
Villadsen
a
a
California Institute of Technology, Pasadena, CA 91125, USA
b
University of Manchester, Manchester, M13 9PL, UK
c
University of Cambridge, Cambridge, CBE 0HE, UK
d
University of Southampton, Southampton, SO17 1BJ, UK
E-mail:
ctibbs@ipac.caltech.edu
,
paladini@ipac.caltech.edu
,
kcleary@astro.caltech.edu
,
keith.grainge@manchester.ac.uk
,
sjcm@astro.caltech.edu
,
tjp@astro.caltech.edu
,
ycp21@mrao.cam.ac.uk
,
cr461@mrao.cam.ac.uk
,
a.scaife@soton.ac.uk
,
mas@astro.caltech.edu
,
jrv@astro.caltech.edu
Within many molecular clouds in our Galaxy there are cold, dense regions known as cold clumps
in which stars form. These dense environments provide a great location in which to study
dust grain evolution. Given the low temperatures (
∼
10–15 K) and high densities (
∼
10
5
cm
−
3
),
these environments are dark at mid-infrared (IR) wavelengths and emit strongly at wavelengths
≥
160
μ
m. The lack of mid-IR emission can be attributed to one of two reasons: i) a deficit of
the small dust grains that emit stochastically at mid-IR wavelengths; or ii) small dust grains are
present, but due to the high densities, the stellar photons cannot penetrate deep enough into the
clumps to excite them. Using mid-IR observations alone it is impossible to distinguish between
these two scenarios. However, by using spinning dust emission at cm wavelengths it is possible to
break this degeneracy, because if small dust grains are present in these clumps, then even though
stellar photons cannot excite them to emit at mid-IR wavelengths, these dust grains will be spun-
up by collisions and hence emit spinning dust radiation. If spinning dust were detected in these
clumps it would prove that there are small dust grains present and that the lack of mid-IR emission
is due to a lack of stellar photons. Conversely, a lack of spinning dust emission would indicate a
deficit of small dust grains in these clumps. Since small dust grains require harsh radiation fields
to be destroyed, a lack of small dust grains is likely a result of dust grain coagulation. With this in
mind, we present preliminary results illustrating our method of using spinning dust observations
to determine the evolution of small dust grains in these environments.
The Life Cycle of Dust in the Universe: Observations, Theory, and Laboratory Experiments
18-22 November, 2013
Taipei, Taiwan
∗
Speaker.
c
©
Copyright owned by the author(s) under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike Licence.
http://pos.sissa.it/
PoS(LCDU 2013)064
Using Spinning Dust Emission To Constrain Dust Grain Evolution
C.T. Tibbs
1. Introduction
Star formation is the result of gravitational instability occurring in cold, dense structures
known as pre-stellar cores. These cores represent one of the earliest phases of star formation and
the investigation of the physical properties of these environments is crucial for understanding the
early stages of star formation. These dense environments also play a vital role in the life cycle of
dust in the Galactic interstellar medium, making them an ideal location in which to study dust grain
evolution. As part of the
Planck
Early Release Compact Source Catalog [1], the Early Cold Cores
Catalog (ECC) containing all of the Galactic cold clumps
1
with temperatures
<
14 K and a signal
to noise ratio
>
15 was published. Recent modelling [2] found that if there are small dust grains
present in these clumps then spinning dust emission should be detected. Spinning dust emission
is a relatively new emission mechanism due to electric dipole emission from small, spinning dust
grains [3]. Observed in the wavelength range 3–0.3 cm, spinning dust emission has been detected
in a variety of Galactic environments (e.g. [4, 5, 6, 7]), and produces a very distinctive peaked
spectrum, peaking at wavelengths of
∼
1 cm. Given this result [2], spinning dust observations can
be used to probe the abundance of the small dust grains, and hence the dust grain evolution, in these
dense environments. To test this hypothesis, we observed 15 of the
Planck
ECC sources with the
CARMA interferometer at 1 cm. In this analysis we simply focus on one of our targets, ECC224,
which is located in the Cepheus molecular cloud complex at
l
= 113.62
◦
,
b
= +15.01
◦
. Displayed
in Figure
1
are maps of ECC224 from mid-infrared (IR) to cm wavelengths. The emission is most
prominent in the far-IR and sub-mm wavelengths, with little or no emission at shorter (70 and
100
μ
m) and longer (1 cm) wavelengths. The lack of emission at 70 and 100
μ
m is due to the fact
that this clump is cold, while the lack of emission at 1 cm, implies that there is no significant spin-
ning dust emission. However, even with a no detectable spinning dust emission, it is still possible
to constrain the evolution of dust grains in this clump.
2. Thermal Dust Emission
To model the thermal dust emission in ECC224 we used
Herschel
data. All of the
Herschel
maps were reprocessed, convolved to a common angular resolution, and to remove any contribution
from foreground/background emission, we subtracted the median value of the flux computed in a
reference position that is devoid of emission in each of the maps. We modelled the far-IR emission
using
I
ν
=
μ
m
H
N
H
κ
ν
B
ν
(
T
d
)
, where
I
ν
is the intensity at frequency
ν
,
μ
is the molecular weight of
hydrogen,
m
H
is the mass of a H atom,
N
H
is the hydrogen column density,
B
ν
(
T
d
)
is the Planck
function for temperature
T
d
, and
κ
ν
is the dust opacity which was assumed to be of the form:
κ
ν
∝
ν
β
, with
β
= 2. Given the lack of emission at 70 and 100
μ
m, we excluded these bands and
fitted the 160, 250, 350, and 500
μ
m maps on a pixel by pixel basis for
N
H
and
T
d
(see Figure
2
).
Herschel
, with its sub-arcmin angular resolution, is able to resolve the structure in the
Planck
ECC sources (see Figure
1
) and for this reason, we used
CLUMPFIND
[8] to identify sub-clumps
in ECC224 (see Figure
2
). Since
CLUMPFIND
computes the effective circular radius of each sub-
clump, we assumed that the sub-clumps are spherical. For each sub-clump we computed the mean
column density,
N
mean
H
, and mean dust temperature,
T
mean
d
. We divided
N
mean
H
by the linear size of
1
We refer to the
Planck
sources as clumps as
Planck
lacks the angular resolution to observe individual cores.
2
PoS(LCDU 2013)064
Using Spinning Dust Emission To Constrain Dust Grain Evolution
C.T. Tibbs
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
Figure 1:
Maps of ECC224: (a)
Herschel
70
μ
m map; (b)
Herschel
100
μ
m map; (c)
Herschel
160
μ
m map;
(d)
Herschel
250
μ
m map; (e)
Herschel
350
μ
m map; (f)
Herschel
500
μ
m map; (g) Planck 850
μ
m map;
(h) CARMA 1 cm map. All maps are 30 arcmin
×
30 arcmin in size and centered on ECC224.
10
20
H/cm
2
0
20
40
60
80
100
5
0
-5
-10
Arc Minutes
-5
0
5
10
Arc Minutes
Center: R.A. 22:21:37.34 Decl. +75:06:33.5
1
2
3
ECC224
K
12
13
14
15
16
17
5
0
-5
-10
Arc Minutes
-5
0
5
10
Arc Minutes
Center: R.A. 22:21:37.34 Decl. +75:06:33.5
1
2
3
ECC224
Figure 2:
N
H
and
T
d
maps of ECC224 derived from the
Herschel
observations with the three identified
sub-clumps labelled.
the sub-clumps to estimate the mean density,
n
H
, and we converted
T
mean
d
into the mean radiation
field,
χ
, using
χ
= (
T
d
/17.5)
(
4
+
β
)
.
3. Spinning Dust Emission
To estimate the expected level of spinning dust emission in ECC224 we used the spinning dust
model
SPDUST
[9] with
n
H
and
χ
(derived from the
Herschel
observations), in combination with
the idealized parameters for dark clouds [3], as input parameters. The dust grain size distribution
incorporated within
SPDUST
is parametrized using the ratio of total to selective extinction,
R
V
, and
the abundance of the small carbonaceous dust grains, b
C
[10]. Using the
R
V
= 5.5 size distribu-
tions, we computed the spinning dust emission for four different values of
b
C
. The final SEDs
are plotted in Figure
3
, with the far-IR emission constrained using the
Herschel
data (asterisks),
and the predicted spinning dust curves plotted at microwave frequencies. It is clear to see how the
amplitude of the spinning dust curves increases as the abundance of the small carbonaceous dust
grains increases. Given the lack of detection, we plotted the conservative 5
σ
upper limit estimated
from the 1 cm CARMA data (solid triangle), which allows us to constrain the abundance of the
small carbonaceous dust grains by ruling out values of b
C
>
2
×
10
−
5
for each of the sub-clumps.
4. Conclusions
The goal of this work is to demonstrate the ability of using spinning dust observations to
determine how dust grains are evolving in cold clumps. We use
Herschel
far-IR observations of
one of the
Planck
cold clumps (ECC224) to derive maps of
N
H
and
T
d
, which we use to estimate
n
H
and
χ
in identified sub-clumps. These derived properties are used as inputs to the spinning
3
PoS(LCDU 2013)064
Using Spinning Dust Emission To Constrain Dust Grain Evolution
C.T. Tibbs
1
10
100
1000
10000
Frequency (GHz)
0.001
0.010
0.100
1.000
10.000
100.000
1000.000
Flux Density (Jy)
ECC224 - Clump 1
b
C
= 0
×
10
-5
b
C
= 1
×
10
-5
b
C
= 2
×
10
-5
b
C
= 3
×
10
-5
1
10
100
1000
10000
Frequency (GHz)
0.001
0.010
0.100
1.000
10.000
100.000
1000.000
Flux Density (Jy)
ECC224 - Clump 2
b
C
= 0
×
10
-5
b
C
= 1
×
10
-5
b
C
= 2
×
10
-5
b
C
= 3
×
10
-5
1
10
100
1000
10000
Frequency (GHz)
0.001
0.010
0.100
1.000
10.000
100.000
1000.000
Flux Density (Jy)
ECC224 - Clump 3
b
C
= 0
×
10
-5
b
C
= 1
×
10
-5
b
C
= 2
×
10
-5
b
C
= 3
×
10
-5
Figure 3:
SEDs of the three sub-clumps identified in ECC224. The spinning dust curves are produced using
SPDUST
based on
n
H
and
χ
derived from the far-IR emission for a range of different values of the abundance
of small carbonaceous dust grains (b
C
). The 1 cm data point (solid triangle) is the 5
σ
upper limit estimated
from the CARMA data. Also plotted are the flux densities of the
Herschel
data (asterisks) at 160, 250, 350,
and 500
μ
m.
dust model
SPDUST
to predict the level of spinning dust emission for each of the sub-clumps, for
a range of abundances of the small carbonaceous dust grains. We compare the predicted spinning
dust emission with CARMA 1 cm observations to constrain the small dust grain abundance. We
find that for the three sub-clumps in ECC224, there is no detectable emission at 1 cm, and that
the three sub-clumps exhibit a deficit of small dust grains, suggesting that dust grain coagulation
is occurring. This is the first time that spinning dust observations have been used to constrain the
evolutionary properties of dust grains.
References
[1] Planck Collaboration,
Planck early results. VII. The Early Release Compact Source Catalogue
, 2011,
A&A, 536, A7
[2] N. Ysard, M. Juvela, & L. Verstraete,
Modelling the spinning dust emission from dense interstellar
clouds
, 2011, A&A, 535, A89
[3] B.T. Draine, & A. Lazarian,
Electric Dipole Radiation from Spinning Dust Grains
, 1998, ApJ, 508,
157
[4] Planck Collaboration,
Planck intermediate results. XV. A study of anomalous microwave emission in
Galactic clouds
, 2013, arXiv:1309.1357
[5] C.T. Tibbs, R.A. Watson, C. Dickinson, et al.,
Very Small Array observations of the anomalous
microwave emission in the Perseus region
, 2010, MNRAS, 402, 1969
[6] C. Dickinson, S. Casassus, R.D. Davies, et al.,
Infrared-correlated 31-GHz radio emission from Orion
East
, 2010, MNRAS, 407, 2223
[7] C.T. Tibbs, R. Paladini, M. Compiègne, et al.,
A Multi-wavelength Investigation of RCW175: An H II
Region Harboring Spinning Dust Emission
, 2012, ApJ, 754, 94
[8] J.P. Williams, E.J. de Geus, & L. Blitz,
Determining structure in molecular clouds
, 1994, ApJ, 428,
693
[9] K. Silsbee, Y. Ali-Haïmoud, & C.M. Hirata,
Spinning dust emission: the effect of rotation around a
non-principal axis
, 2011, MNRAS, 411, 2750
[10] J.C. Weingartner, & B.T. Draine,
Dust Grain-Size Distributions and Extinction in the Milky Way,
Large Magellanic Cloud, and Small Magellanic Cloud
, 2001, ApJ, 548, 296
4