L155, 1997 September 10
A mixture of the polycyclic aromatic
hydrocarbons (PAHs), acenaphthylene and acenaphthene, when subjected to the
energetic environment of a hydrogen plasma, is transformed into a material
that exhibits an infrared absorption profile in the 3
m region
that is an excellent match of the protoplanetary nebula IRAS 05341+0852
emission profile in the same wavelength region. Acenaphthylene and
acenaphthene were chosen as precursors in the experiment because these
molecules have a structure that can be described as a keystone in a process
in which carbon atoms in a stellar wind condense into PAH species. The
spectral match between experiment and observations appears to validate
that scenario.
Subject headings: H
II regions
infrared:
ISM: lines
and bandsISM: moleculesmolecular processesplanetary
nebulae: general
The object IRAS 05341+0852 (IRAS 05341) is
probably a star that has left the asymptotic giant branch and is becoming a
planetary nebula. It possesses a unique infrared spectrum in the 3
m
(3000 cm-1) region that, while appearing to be related to
what has been called the unidentified infrared (UIR) bands discovered
by Gillett, Forrest, & Merrill (1973),
shows a reversal of strengths of the features in the 3
m region
from what is usually observed (Geballe & van
der Veen 1990; Joblin et al.
1996). Geballe & van der Veen (1990)
recognized that the spectral emission profile of IRAS 05341 at 3
m could
modify interpretations of the UIR bands. Experiments conducted in our
laboratory have resulted in the synthesis of a material that may assist in
that matter. Most 3
m UIR
emission profiles exhibit a dominant feature at about 3.3
m
(
3000
cm-1) that arises from
a CH
stretching vibration along the periphery of a polycyclic
aromatic hydrocarbon (PAH) molecular structure, accompanied by weaker bands
in the region between 3.3
m and 3.6
m. These
bands probably arise from aliphatic
CH
stretching vibrations possibly in attached components, or harmonics of
CC
skeletal vibrations of the aromatic
structure (de Muizon et al.
1986; Joblin et al.
1996; Bernstein, Sanford, &
Allamandola 1996).
For IRAS 05341 there is a band at 3.294
m
(3036 cm-1) that likely arises from an aromatic
CH stretch,
but the 3.419
m (2925
cm-1) feature is about 1.8 times stronger. Furthermore, this
stronger feature along with a structure prominent in the 3.475 to 3.5
m range
has a character that suggests it is best attributed to an aliphatic
CH
stretch in the methylene
(CH2)
functional group (Colthup, Daly, & Wilberley
1990, p. 215). There also appears to be a shoulder at about 3.38
m
(2959 cm-1) in the observed spectrum presented
by Joblin et al. (1996), which could be a
weaker CH
stretch from the methyl
(CH3)
functional group (Colthup et al.
1990). Sloan et al. (1997) showed that
in the Orion Bar, the 3.40
m feature
had two components. They suggested that the 3.405
m
component might arise from attached methyl sidegroups, while the 3.395
m
component might arise from methylene groups in hydrogenated PAHs
(Hn-PAHs), which would reverse the traditional
assignments.
Previous experiments involving the
processing of the PAH naphthalene (C10H8) in the
energetic environment of a plasma have yielded a material with ultraviolet
spectral characteristics of interest in understanding the carrier of the
2175 Å interstellar extinction feature
(Beegle
1997; Beegle et al. 1997). Infrared
spectroscopy in the 3
m region
has also revealed that the relative strengths of the bands due to the
aromatic and aliphatic components were similar to that of IRAS 05341.
However, the profile of the aromatic
CH stretching
feature of the synthesized material from those experiments displayed some
differences from that of the observed band. Mass spectroscopy performed on
this synthesized substance indicates that the precursor, naphthalene, was
transformed into an aggregate of larger PAH structures, some of which may
be in the form of polymers. Of particular interest as components of the
aggregate are the molecules acenaphthylene (C12H8),
which has been suggested by Keller (1987)
to be the keystone molecule after the formation of naphthalene in the
condensation of carbon in stellar winds, and its hydrogenated form
acenaphthene (C12H10). In the stellar wind model the
linear molecule acetylene (C2H2) is transformed into
the likewise linear molecule diacetylene
(C4H2), which is followed by further transformations
in sequence to the aromatic ring structures of benzene
(C6H6), naphthalene (C10H8),
and acenaphthylene (C12H8). Because there are
multiple paths for further transformations from acenaphthylene (and
presumably acenaphthene) to a variety of larger PAH structures in the
Keller model, the acenaphthylene/acenaphthene structures can be labeled
keystones. Hydrogenated polycyclic aromatic hydrocarbons
(Hn-PAHs) are becoming attractive candidates for being
interstellar constituents because of the match of spectral characteristics
with both the UIR emission bands (Bernstein et al. 1996)
and the 2175 Å interstellar extinction feature (Beegle
et al. 1997).
The molecular structure of naphthalene, acenaphthylene, and acenaphthene is represented by diagrams in Figure 1. Because our previous experiments using naphthalene resulted in the formation of acenaphthylene and acenaphthene in comparable abundance, the use of them as precursor materials instead of naphthalene seemed to be of obvious interest, and such an experiment was performed.
Fig. 1
To study the situation when a PAH is
subjected to an energetic environment, a plasma reactor system was designed
and constructed around a glass plasma tube reactor. The plasma tube reactor
component (Fig. 2) was designed for
easy extraction of synthesized products and for the ability to visually
monitor the emission from various places in the reactor using fiber optic
light guides. The manifold and pumping components of the system permitted
greater control and monitoring of the composition of the gases utilized to
create the plasma than was possible in earlier experiments carried out in
the University of Alabama at Birmingham Astrophysics Laboratory
(Lee & Wdowiak 1993). Gas enters and
exits the plasma reactor via two glass side arms. A liquid nitrogen trapped
diffusion pump is utilized to reduce pressure to
1 ×
10-5 torr (as measured with an ionization gauge) prior to
introduction of reagents. The quadrupole mass spectrometer monitors the
composition of gases entering or exiting the reactor, and there is a
provision for continuous electronic pressure monitoring in the range used
during the discharge. Stainless steel was employed in the construction of
the manifold to insure minimum contamination in the mixing of gaseous
reagents.
Fig. 2
For these experiments, approximately 800
mg of a commercial PAH mixture (Aldrich Chemical Company product A80-5) is
coated on the inner wall of a sapphire tube by inserting it into the tube
and melting it with a hot plate while rotating the tube by hand so that
there is coverage of the entire interior surface when the mixture
solidifies. The prepared tube is then inserted in the center section of the
plasma tube, where the plasma will transit through it. The PAH mixture has
a composition of acenaphthylene
(75%),
acenaphthene
(20%),
and an uncharacterized component
(5%). It
should be pointed out that acenaphthylene and acenaphthene are less
volatile than the naphthalene used in previous experiments
(Lee & Wdowiak 1993; Beegle et al.
1997). For acenaphthylene the melting point is 91°C and the
boiling point is 280°C, while that for acenaphthene is 94°C and
279°C.
Hydrogen gas is first introduced into the reactor system at several tens of torr pressure and then pumped down to operational pressure. This purging is repeated several times to ensure that residual atmospheric contaminants are negligible, as confirmed by mass spectroscopy. The pressure inside the system was monitored continuously. Gas is introduced through a cold trap that is cooled with liquid nitrogen (LN2). The pressure at which the experiment took place was 600 mtorr. Once the desired pressure is established, the discharge is initiated with an electrical current from a 9400 VAC transformer. That transformer is then controlled with a variable transformer operating off line voltage.
When the discharge is terminated, the residue, which exhibits a red color under reflected room lighting, is harvested from within the sapphire tube that held the PAH mixture precursor. The downstream sapphire disk, which also has a deposit on it, is removed from the plasma reactor and placed directly in the sample component of a Mattson Polaris Fourier transform infrared (FTIR) spectrometer for spectral measurements in the 4000 to 2000 cm-1 sapphire transparency range at a resolution of 4 cm-1. Residue samples in the amount of 1 mg harvested from the inside of the sapphire tube are pressed with 100 mg of spectroscopic-grade KBr into a pellet for infrared spectroscopy measurements in the 4000 to 400 cm-1 range at the same resolution. The sample spectrum was obtained in ratio against that of a blank KBr pellet. For most of the measurements, an LN2 boiler heated by a ceramic resistor was used to produce a dry nitrogen purge of the spectrometer.
Figure 3
shows the spectrum of IRAS 05341 in the range 3.2 to 3.65
m
(Joblin et al. 1996) along with an overlay (solid
line) of the spectrum of the laboratory synthesized product. The IRAS
05341 spectrum includes two features not due to PAHs: a contribution from
the H II region of
Pf
emission of hydrogen in the form of a spike atop the 3.294
m feature
and an enhanced error in the measurement at about 3.32
m due to
telluric methane absorption. Overall, the profiles of the two spectra
correlate very well. The most striking aspects are the match in relative
strengths of the 3.294
m and
3.419 m
bands, and the long-wavelength skew or asymmetry of the 3.294
m band.
The asymmetric character of the band of the laboratory synthesized material
not only matches that of the 3.294
m band
for IRAS 05341 emission, but also the profile of the 3.29
m
emission band of the Orion Bar (Sloan et al. 1997).
However, for the Orion source the longer wavelength components (3.4
m, etc.)
are at a fraction of the strength of the 3.29
m
features.
Fig. 3
Geballe & van der Veen
(1990) have also commented on the similarity between the 3
m
emission of IRAS 05341 and of Nova Centauri 1986 (V842 Cen) some 300 days
following the outburst and persisting for several months
(Hyland & McGregor 1989). A 650 K dust
shell appeared some 55 days after the outburst, but the 3
m
emission did not occur until later, when the shell temperature reached 800
K. In the Nova spectrum the 3.29
m feature
of the shell is stronger than the feature at 3.42
m, and
the 3.42
m feature
has a longer wavelength than observed for the
3.4 m
feature in the Orion Bar (see Sloan et al. 1997).
However, the 3.42
m feature
in the Nova is coincident in wavelength with the strongest band of both
IRAS 05341 and the laboratory synthesized material.
An ultraviolet/visible absorption spectrum
of a film prepared by evaporating the red residue onto a quartz substrate
in a vacuum shows a character similar to that obtained in experiments in
which naphthalene is a precursor (Beegle et al. 1997).
The absorption maxima is at 230 nm, which places it within the envelope of
the 2175 Å interstellar extinction feature, which ranges between 2500
and 2000 Å. This is consistent with the synthesized product also
being a candidate component of the carrier of the 2175
Å 
bump.
In summary, the experiment has resulted in
a very good match to the IRAS 05341 spectrum in the 3
m range
and appears to validate the modeling effort of Keller
(1987). The formation scheme worked out by Keller
(1987) for carbon atoms condensing into PAH species could be a
representation of what occurs in a stellar wind issuing from objects such
as the protoplanetary nebula IRAS 05341 or Nova Cen 1986. We find it
interesting that the 3
m
emission from Nova Cen 1986 appeared when the dust shell temperature was
measured to be 800 K, as that temperature was utilized by Keller in his
calculation.
The 3.419
m (2925
cm-1) band exhibited by the laboratory material and the emission
band of IRAS 05341 at the same spectral position can by inference be due to
an aliphatic
CH
stretch for the methylene
(CH2)
functional groups. Acenaphthene has this group, since it is the
hydrogenated form, but acenaphthylene does not (Fig. 1).
The methylene
(CH2)
functional group must also be incorporated in other structures present in
the synthesized material that have come about in the energetic environment
of the plasma. This is the case for material synthesized from naphthalene
as well. Our use of the acenaphthylene and acenaphthene mixture results in
a better spectral match to the IRAS 05341 spectrum than what was obtained
when naphthalene was used. While the plasma does possess some of the
qualities probably present in the stellar wind situation, it can not be
considered to be a simulation. Going a step beyond naphthalene to
acenaphthylene and acenaphthene, which have the keystone molecular
structure in Keller's model, appears to have made up for that deficiency,
which is present in all laboratory experiments that attempt to mimic nature
on the cosmic scale.
We thank the referees, Greg Sloan and Max Bernstein, for helpful comments. This work was supported by NASA Exobiology (NAGW-4079), Origins of Solar Systems (NAGW-4158), and Ultraviolet Astronomy (NAGW-4032) programs.
Full image (18kb) | Discussion in text
FIG.
1.Diagrams
that represent the molecular structure of naphthalene, acenaphthylene, and
acenaphthene along with the molecular weight (MW).
Full image (21kb) | Discussion in text
FIG.
2.The
plasma reactor in which the plasma transits the tube joining the electrode
end chambers. The PAH precursor is coated on the interior of a sapphire
tube that is inserted into the plasma reactor tube.
Full image (23kb) | Discussion in text
FIG.
3.The
3 micron spectral region of material synthesized by subjecting a mixture of
acenaphthylene and acenaphthene to the energetic hydrogen
plasma environment (solid line) superposed on the observed
spectrum of IRAS 05341+0852 (Joblin et al. 1996).