Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stsci.edu/~kgordon/papers/PS_files/ic435.ps.gz Äàòà èçìåíåíèÿ: Thu Apr 24 02:49:35 2003 Äàòà èíäåêñèðîâàíèÿ: Tue May 27 07:52:34 2008 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: reflection nebula

Scattering Properties of the Dust in the Reflection Nebula IC 435
Daniela Calzetti, Ralph C. Bohlin
Space Telescope Science Institute, 3700 San Martin Dr., Baltimore, MD 21218, USA
Karl D. Gordon, Adolf N. Witt
Ritter Astrophysical Research Center, The University of Toledo,
Toledo, OH 43606, USA
and Luciana Bianchi
Space Telescope Science Institute
ABSTRACT
New IUE observations of the reflection nebula IC 435 are presented, and the
scattering properties of the nebular dust grains derived. Previous studies of this nebula
suggested the presence of enhanced scattering in the 2175 š A interstellar extinction
bump, which the new data do not confirm. The UV observations plus optical data
from the literature determine the most likely star­nebula geometry, the nebular dust
density distribution, and the values of the albedo and the phase function asymmetry
for the dust grains. The results show a broad minimum for the albedo around 2300 š A
followed by rising values in the far­UV, up to a ' 0:75 \Gamma 0:8. The phase function for
the nebular grains tends to be forward­scattering in the UV, with values for the phase
function asymmetry rising from g ' 0:60 at 3000 š A to g ' 0:75 around 1500 š A.
Subject headings: interstellar: matter ­ nebulae: individual (IC 435) ­ nebulae:
reflection ­ ultraviolet: spectra
1. Introduction
Reflection nebulae represent important astrophysical laboratories for studying the scattering
properties of the interstellar dust grains as a function of wavelength. At ultraviolet (UV)
wavelengths the illumination of the nebula often arises from a single hot star, implying that the
geometrical distribution of the dust around the star can be constrained. Therefore, reflection
nebulae are particularly suitable for investigations of the optical properties of the dust grains in
the UV.
The study of the UV characteristics of dust scattering in the Galaxy is essential to our
understanding of the characteristics of dust obscuration in external galaxies. In particular, one
of the unsolved puzzles is the apparent absence of the 2175 š A dust bump in the UV spectra of

-- 2 --
reddened starburst galaxies (Calzetti, Kinney, & Storchi­Bergmann 1994). The interpretation
of this finding is complicated by the many factors contributing to the extragalactic obscuration
curve, which include the geometrical effects due to the dust­stars­gas distribution, and, possibly,
the optical properties and the composition of dust. Therefore, the properties of the 2175 š A bump
need to be fully investigated as a function of the environment, in order to understand the observed
absence in external galaxies.
Since its discovery (Stecher 1965), the 2175 š A bump has been attributed to absorption by
dust grains which are sufficiently small to be operating in the Rayleigh limit (Stecher & Donn
1965, Lillie & Witt 1976, Morgan, Nandy, & Thompson 1976, Mathis 1994). However, in specific
environments or under specific physical and geometrical conditions, part of the bump may be due
to enhanced scattering around 2200 š A.
Evidence of enhanced scattering in the bump was reported for IC 435 and CED 201 by
Witt, Bohlin & Stecher (1986a). In the case of IC 435, the enhanced scattering, the low far­UV
extinction curve towards the illuminating star HD 38087 (Witt, Bohlin, & Stecher 1984), and the
displacement of the 2175 š A bump to a peak wavelength of 2193 š A (Fitzpatrick & Massa 1986)
suggest depletion of small grains, which is typical of dense regions (Witt et al. 1984). Finally, the
bump seen toward HD 38087 is unusually strong compared to the average interstellar medium.
We report the results from recent, high S/N IUE observations of two regions in IC 435 at
different radial distances from the central illuminating star. Optical data on the nebula from
Witt & Schild (1986) define a model for the dust density distribution in the nebula. The two UV
spectra at different offsets allow us to calculate independently the albedo and the phase function
asymmetry of the dust as a function of wavelength.
2. Data
IC 435 (ff[1950]=05 h 40 m 29 s .5, ffi[1950]=\Gamma02 o 20 0 05 00 ) is a bright, symmetric reflection nebula,
about 2 0 .5 in radius, situated at a distance of about 550 pc from the Sun in the northern Orion
molecular cloud L1630 (Witt & Schild 1986). The nebula is illuminated by the B5 V star HD 38087
(E(B\GammaV)=0.29, Guetter 1979, Schild & Chaffee 1971), which is located approximately in the
center of the projected nebular region.
Two regions in the nebula, IC435\Gamma1 and IC435\Gamma2, were observed with IUE, at 23 00 .4 and
43 00 .3 North­East from the central illuminating star, yielding ¸6 š A resolution spectra in the
wavelength range 1200­3000 š A. The observations were obtained via service observing at both the
NASA/GSFC and the ESA/VILSPA stations. The observational log is given in Table 1. The
spectra for the two positions are reduced and co­added using the optimal extraction routine by
Kinney, Bohlin & Neill (1991). The nebular spectra are corrected for the instrumentally scattered
stellar light (Witt et al. 1982): the contamination is ¸10% for IC435\Gamma1 and ¸4% for IC435\Gamma2.

-- 3 --
TABLE I
Log of the IUE Observations
Image Exposure
Name Position Number (minutes) Date
IC 435­1...... 16: 00 2 E, 16: 00 9 N SWP 52769 300 1994 Nov 9
LWP29481 337 1994 Nov 9
LWP29492 370 1994 Nov 11
IC 435­2...... 30: 00 3 E, 31: 00 0 N SWP52781 320 1994 Nov 11
LWP29486 360 1994 Nov 10
LWP29557 407 1994 Nov 24
A total of 6 short wavelength and 5 long wavelength UV spectra for HD 38087 were retrieved
from the IUE archive, and are calibrated and co­added using the same routine as for the nebular
spectra. The ratio of the nebular surface brightness to the stellar flux, log(S=F \Lambda ), is shown in
Figure 1 for the two nebular regions. The solid angle subtended by the IUE slit is 200 square
arcsec or 4.8\Theta10 \Gamma9 ster.
One SWP and one LWR spectrum of a region in IC 435 ¸2 00 NE of our IC435\Gamma1 exist in the
IUE archive (see Witt et al. 1986a, who report the position of the region to be in the NW sector
of the nebula, instead of the true position in the NE sector). The total exposure time for IC435\Gamma1
is about a factor 10 larger than the old spectrum in each of the two IUE cameras.
For the 10 00 \Theta20 00 IUE aperture, the IC435\Gamma1 region and the archival region are nearly
coincident and should have identical UV spectra. Differences in the aperture orientation of 189 o for
our images and 145 o for the archival images (N=0 o and E=90 o ) can cause different contamination
of the nebular spectra from instrumentally scattered stellar light; however, this contamination is at
the level of about 10% of the observed flux and the uncertainty is Ÿ 5%. The ratio of the nebular
surface brightness to the stellar flux for the archival spectrum is reported also in Figure 1.
The LWR archival spectrum agrees with the LWP IC435\Gamma1 spectrum within the uncertainties.
The small excess in the 1994 spectrum longward of 2800 š A can be attributed to the recently
developed problem of solar contamination in the LWP Camera. On the other hand, the SWP
archival spectrum is about a factor two below the IC435\Gamma1 spectrum. As a net effect, the
(S=F \Lambda ) ratio of the archival nebular intensity to the stellar flux shows an excess around 2300 š A,
which led Witt et al. (1986a) to conclude that there is enhanced scattering in the bump of the
nebula. The reason for the strong difference between the archival and the new spectra is unclear.
Some instrumental background problem, coupled with the low signal­to­noise ratio of the old
observations, may be at the origin of the low net SWP archival spectrum.
There is no evidence for enhanced scattering at the bump wavelengths in the (S=F \Lambda ) ratio of

-- 4 --
Fig. 1.--- The ratio of the nebular intensity to the stellar flux, log(S=F \Lambda ) is shown for the two
regions observed in IC 435, namely IC435\Gamma1 (solid squares and continuous line) and IC435\Gamma2
(solid triangles and continuous line), and for the archival spectrum (open squares and dashed line).
The spectra are binned over 10 š A. The position of the archival spectrum is coincident with IC435\Gamma1,
while the two spectra differ markedly at short wavelengths.

-- 5 --
the new spectra (see Figure 1).
Photometric data for IC 435 in the B and V bandpasses are from Witt & Schild (1986), who
used a thinned CCD detector for their observations. The nebular surface brightness in the two
filters was mapped by these authors at various distances and positions from HD 38087, using
an aperture of 10 00 .24 in diameter. The R and I data of Witt & Schild (1986) were considered
unsuitable for the present study because of the presence of extended red emission (ERE) in these
bandpasses.
Detailed spatial information for the nebular surface brightness up to 140 00 from the illuminating
star allows us to model the nebular dust density distribution. In addition, two of the positions
reported by Witt & Schild coincide with the positions of IC435\Gamma1 and IC435\Gamma2, and provide an
extension of our UV spectra to optical wavelengths.
3. A Model of IC 435
Derivations of dust scattering properties from reflection nebula observations are strongly
model­dependent. Usually, such exercises are only meaningful when observations covering a range
of wavelengths are available; and the aim is to derive the relative variation of scattering properties
with wavelength for a constant geometry. Information needed in this case includes the wavelength
dependence of extinction by the nebular dust and an estimate of the dust fraction which is
involved in the nebula, usually less than the total column density of dust between the illuminating
star and the observer. Extensive surface brightness measurements at one wavelength are helpful
in constraining the details of the nebular geometry. All these conditions are satisfied in the case of
IC 435.
The wavelength dependence of the extinction toward HD 38087 (Witt et al. 1984) is
characterized by a far­UV rise lower than the average, which is typical of dense regions with larger­
than­average dust grains (Cardelli, Clayton, & Mathis 1989, Snow & Witt 1989). Dense regions
are also characterized by large values of the total­to­selective extinction R V = A(V )=E(B \Gamma V ),
and we use the value R V = 5:3 from Cardelli et al. (1989) to extend the UV values of the HD 38087
extinction curve toward optical wavelengths. The results are shown in Figure 2: the HD 38087
extinction curve is fairly well reproduced by the Cardelli et al. curve for R V = 5:3, except for the
unusually strong 2175 š A feature.
The nebular spectra of IC435\Gamma1 and IC435\Gamma2 exhibit strong depressions due to the 2175
š A band, which are both similar to that in the illuminating star and indicate that most of the
dust toward HD 38087 is located at distances from the star greater than the 0.4 pc radius of the
nebula. The scattering in IC 435 must be dominated by single scattering, which is consistent
with a fairly small radial optical depth. If more of the extinction were to occur within IC 435
itself, a greater 2175 š A band strength should appear in the spectrum with the larger offset, which
is not the case. Thus, the column density of dust associated with IC 435 itself is estimated as

-- 6 --
Fig. 2.--- The extinction A(–)=E(B \Gamma V ) towards HD 38087 (crosses) is shown in comparison to
the standard extinction curve of Cardelli et al. (1989) for R V = 5:3 (solid line).
E(B \Gamma V ) = 0:06 or Ü V = 0:29, which is only ¸20% of the total observed extinction. As will be
shown in the next section, this value of the optical depth, when combined with the geometrical
model and the extensive surface brightness measurements of IC 435 in the B and V bands yield
results for the dust albedo and the phase function asymmetry at B and V which are consistent
with results obtained from numerous other studies of interstellar scattering (Gordon et al. 1994
and references therein).
The geometrical distribution of dust with Ü V = 0:29 in front of HD 38087 is derived by
modeling the detailed surface brightness distribution in the B band using a Monte Carlo code
(Witt 1977; Witt et al. 1982; Witt et al. 1992; Gordon et al. 1994)). The radial density profile
for the dust is characterized by a shell­like structure. A low density region of radius about
0.16 pc surrounds the central star; from 0.16 pc to 0.3 pc the density increases up to about 2.7
times the central value; a steady decrease in density down to zero follows out to 0.4 pc. This
density distribution is very close to that derived for IC 435 by Witt & Sturm (1990). A fit of this
spherically symmetric model to the B band data of Witt & Schild (1986) is shown in Figure 3,
using the distance of 550 pc to IC 435.
For the derivation of UV scattering properties, we apply the same geometrical model to the
IUE observations at the two offset distances, varying only the optical depth according to the
extinction law given in Figure 2.

-- 7 --
Fig. 3.--- The observed ratio of the nebular intensity to the stellar flux, log(S N =F \Lambda ), in the B band
(crosses) is plotted as a function of the projected distance from HD 38087. The fit to the data
from the geometrical model for the dust density distribution discussed in section 3 is also shown
(continuous line).

-- 8 --
Fig. 4.--- The albedo a and phase function asymmetry g are shown as a function of – \Gamma1 . The
UV values are derived from the IUE spectra of the two offset position in the nebula, IC435\Gamma1 and
IC435\Gamma2. The values at B (– = 4400 š A) and V (– = 5500 š A) are derived ?from the nebular data
of Witt & Schild (1986) at the same offset positions of the IUE spectra. Typical 1 oe error bars for
a and g are also shown.
4. Results
The separate fits to the surface brightness spectra of IC435\Gamma1 and IC435\Gamma2, at 50 š A intervals,
yield formally independent solutions for both the albedo and the phase function asymmetry, with
mutually dependent uncertainties. The results are shown in Figure 4.
The most important findings are a broad minimum in the albedo associated with the 2175 š A
band, followed by rising albedo values in the far­UV. The minimum for the albedo, a ' 0:4, is
reached at – ' 2300 š A. Below 1600 š A, the albedo has values around 0.7 ­ 0.8, which is higher by
about 40% than the corresponding values in the B and V bands, aB = 0:63 and a V = 0:48. The
solutions for the phase function asymmetry, g, are indicative of a constant value of g ú 0:6 from
5500 to 3000 š A, with a subsequent rise to g = 0:75 \Sigma 0:05 near 1500 š A. The scatter in the values
of g and a is due to the independent noise in the original data at the two offset positions, which
may result in a lower formal solution for a, compensated for by a higher g­value, or vice­versa.
Since a = Q sca =Q ext and Q abs = Q ext \Gamma Q sca , the wavelength dependent values of a and of the
extinction efficiency Q ext / A – =EB \GammaV (see Figure 2) are used to derive the scattering efficiency

-- 9 --
Fig. 5.--- The scattering (open circles) and absorption (filled circles) components, Q sca and
Q abs respectively, of the extinction efficiency are shown. The dust extinction efficiency, Q ext , is
proportional to the extinction curve for HD 38087 (crosses).
Q sca and the absorption efficiency Q abs for the nebular dust. The results are shown in Figure 5;
the 2175 š A bump in the extinction curve of HD 38087 is entirely due to absorption, while the
scattering efficiency exhibits a gradual increase toward the UV.
5. Discussion and Conclusions.
Our results show a steadily increasing phase function asymmetry and mirror the findings of
Witt et al. (1992) for the bright reflection NGC 7023. The trend toward larger g­values appears
stronger in IC 435 and may be explained by the larger­than­average size of the dust grains toward
HD 38087, as discussed by Snow & Witt (1989). The albedo values at the shortest UV wavelengths
are higher than found in any previous investigation (e.g. Gordon et al. 1994). However, such
results were anticipated by Witt et al. (1993), who associated the decline of the dust albedo found
in NGC 7023 at wavelengths shorter than 1300 š A with the presence of a steeply rising extinction
curve, caused by small absorbing grains. HD 38087 exhibits an unusually flat extinction curve in
the UV (Witt et al. 1984; see Figure 2) with little sign of the presence of small grain absorption.
For such cases, Witt et al. (1993) had predicted a high far­UV albedo.
The albedo minimum centered near 2300 š A is very similar in width and depth to the albedo

-- 10 --
minimum found from the study of the diffuse galactic light by Lillie & Witt (1976). These results
are also consistent with the IUE spectra of the Pleiades reflection nebulae (Witt, Bohlin, &
Stecher 1986b), which indicated a higher albedo at wavelengths shortward of 2000 š A compared to
the 2200--2700 š A range.
Our values for the albedo and phase function asymmetry are in remarkable agreement with
the predictions of the ``modified astronomical silicate'' model with R V =5.3 by Kim & Martin
(1995). There are, however, two noticeable differences: the albedo minimum is broader and deeper
in our data than in their model; and we do not see the decline in the albedo values for – \Gamma1 ? 6 ¯m.
Both differences can be attributed to the peculiarities of the extinction curve towards IC 435
compared to the ``standard'' interstellar extinction curve (see section 3, cf. Witt et al. 1993).
The present results have implications for the interpretation of the dust obscuration curve in
external starburst galaxies (Calzetti et al. 1994). In agreement with studies on other reflection
nebulae (Gordon et al. 1994), the 2175 š A bump in the interstellar extinction curve towards
HD 38087 is confirmed as entirely due to absorption by dust grains. Therefore, the absence of the
bump in the obscuration curve of starburst galaxies is difficult to explain in terms of geometrical
effects alone, even if complex models are used for the distribution of stars and dust. Indeed, the
lack of the bump is observationally coupled with reddened UV spectra. Geometrical models which
can lower the importance of the 2175 š A feature tend also to predict rather blue UV spectra. The
presence of enhanced scattering in the bump would have helped the geometrical interpretation,
since, under specific conditions, a fraction of the bump strength would have been associated
with excess scattering. However, the presence of enhanced scattering in the bump, suggested by
previous observations, is not confirmed in IC 435.
The authors thank the IUE Observatory staff at NASA/GSFC and at ESA/VILSPA (in
particular VILSPA resident astronomer C. La Dous and GSFC resident astronomers L. Rawley and
T. Teays) for carrying out the observations through Service Observing. D.C. acknowledges support
from the NASA Grant NAGW­3757 during this research. K.D.G. and A.N.W. acknowledge the
support from NASA LTSAP grant NAGW­3168.
REFERENCES
Calzetti, D., Kinney, A. L., & Storchi­Bergmann, T. 1994, ApJ, 429, 582
Cardelli, J. A., Clayton, G. C., & Mathis, J. C. 1989, ApJ, 345, 245
Fitzpatrick, E. L., & Massa, D. 1986, ApJ, 307, 286
Gordon, K. D., Witt, A. N., Carruthers, G. R., Christensen, S. A., & Dohne, B. C. 1994, ApJ,
432, 641

-- 11 --
Guetter, H. H. 1979, AJ, 84, 1846
Kim, S.­H., & Martin, P.G. 1995, ApJ, 442, 172.
Kinney, A. L., Bohlin, R. C., & Neill, J. D. 1991, PASP, 103, 665
Lillie, C. F. & Witt, A. N. 1976, ApJ, 208, 64
Mathis, J. S. 1994, ApJ, 422, 176
Morgan, D. H., Nandy, K., & Thompson, G.I. 1976, MNRAS, 177, 531
Schild, R. E. & Chaffee, F. 1971, ApJ, 169, 529
Snow, T. P. & Witt, A. N. 1989, ApJ, 342, 295
Stecher, T. P. 1965, ApJ, 142, 1683
Stecher, T. P., & Donn, B. 1965, ApJ, 142, 1681
Witt, A. N. 1977, ApJS, 35, 1
Witt, A. N., Bohlin, R. C., & Stecher, T. P. 1984, ApJ, 279, 698
Witt, A. N., Bohlin, R. C., & Stecher, T. P. 1986a, ApJ, 305, L23
Witt, A. N., Bohlin, R. C., & Stecher, T. P. 1986b, ApJ, 302, 421
Witt, A. N., Petersohn, J. K., Bohlin, R. C., O'Connell, R. W., Roberts, M. S., Smith, A. M., &
Stecher, T. P. 1992, ApJ, 395, L5
Witt, A. N., Petersohn, J. K., Holberg, J. B., Murthy, J., Dring. A., & Henry, R. C. 1993, ApJ,
410, 714
Witt, A. N. & Schild, R. E. 1986, ApJS, 62, 839
Witt, A. N. & Sturm, J. C. 1990, BAAS, 22, 839
Witt, A. N., Walker, G. A. M., Bohlin, R. C., & Stecher T. P. 1982, ApJ, 261, 492
This preprint was prepared with the AAS L A T E X macros v3.0.