Simulation of interstellar aromatic hydrocarbons using ion cyclotron resonance. preliminary results.

RAPID COMMUNICATIONS IN MASS SPECTROMETRY, VOL. 11, 1619-1623 (1997)
Simulation of Interstellar Aromatic
Hydrocarbons Using Ion Cyclotron Resonance.
Preliminary Results +
Christine Joblin1*, Christophe Masselon 2, Pierre Boissel3, Patrick de Parseval3, Suzana
Martinovic2 and Jean-françois Muller2
1CESR, Laboratoire du CNRS, BP 4346, 9 Av. du Colonel Roche, 31028 Toulouse Cedex, France 2LSMCL, Université de Metz, Technopole 2000, 1 Bd Arago, 57078 Metz Cedex 3, France 3LPPM, Laboratoire du CNRS, Université Paris Sud, Bât. 213, 91405 Orsay Cedex, France Using laser microprobe and Fourier transform ion cyclotron resonance (FTICR) mass spectrometry, we have
produced and trapped aromatic species which might well be laboratory analogues of interstellar polycyclic
aromatic hydrocarbons. These species are produced by laser ablation of a sample of pyrolysed coronene. They
are large (up to two hundred carbon atoms) and can be highly dehydrogenated upon irradiation. More or less
condensed forms have also been identified. Although still preliminary, these results open a new field of
investigation: the study of the photophysics and chemistry of these large reactive species in isolation conditions
close to those found in interstellar space. This will be the first objective of the PIRENEA experiment, a
FTICRMS set-up devoted to astrophysics. 1997 by John Wiley & Sons, Ltd.
Received 13 June 1997; Accepted 18 August 1997
Rapid. Commun. Mass Spectrom. 11, 1619-1623 (1997)
No. of Figures: 6 No. of Tables: 0 No. of Refs: 39
atoms. They might be large planar structures, i.e.
graphitic layers with surrounding hydrogen atoms The European satellite, Infrared Space Observatory (more or less dehydrogenated PAHs of large size), or (ISO), launched in November 1995 is currently provid- 3D graphitic islands such as the so-called ‘basic ing a large amount of spectral information regarding structural units' found in coals.14 The inte resting point interstellar matter. One exciting result concerns a now is how to produce and study such species in the family of infrared emission bands at 3.3,6.2,7.7,8.6 and 11.3 ¹m which are easily observed in the regions of high Dramatic changes were observed at about 600 °C UV field. The observations of the balloon experiment when measuring the IR spectra of various gas-phase AROME1,2 have suggested that these features are PAHs at high temperatures.15 At this temperature, present everywhere in our galaxy and in other galaxies, making their carriers an important and widespread 24H12) was pyrolysed into a pitch-like brown residue mixed with red needles. Interestingly, component of interstellar matter. This is beautifully the IR spectrum of the pyrolysed coronene appeared to confirmed by the results of the ISO misson. The IR provide a better match to the astronomical spectra than features are now well observed in regions where the did the spectrum of the precursor PAH (Fig. 1).
radiation field intensity differs from the average by Although it is clear that the sample contains species orders of magnitude.3-5 This strongly supports an which might well be analogues of the interstellar excitation mechanism by transient heating (up to about species, there are also other products which are of less 1000 K) after absorption of single UV photons.6,7 This interest. In particular, we arbitrarily subtracted a mechanism implies that the carriers are small (typically power-law (¸-l) continuum from the laboratory spec- 1-2 nm in size according to Sellgren7). Certainly, trum displayed in Fig. 1. This continuum is most likely classical interstellar grains (0.1 ¹m size) are not able to dominated by the solid phase but a contribution from account for this mid-IR emission since they are at an large aromatic molecules is not excluded and has to be equilibrium temperature of 10-20 K in most places in the interstellar medium.
The observed emission bands at 3.3, 6.2, 7.7, 8.6 and 11.3 ¹m are characteristic of hydrogenated aromatic Analysis of the pyrolysed coronene sample
species. Polycyclic aromatic hydrocarbons PAHs) havenaturally been proposed as their carriers.8,9 However, Samples of pyrolysed coronene (C24H12) were pre- no PAH studied to date in the laboratory has provided pared by sealing under vacuum a quartz tube a convincing spectral match with the astrophysical (V = 9 cm3) containing 25 mg of coronene (Aldrich).
spectrum (Fig. 1). Other candidates such as coal grains The tube was heated in an oven up to 660 °C and then seem to give a better spectral agreement12,13 but they cooled down. The heating rate was typically 3°C/min are too large to account for the emission mechanism.
between 500 and 660 °C.
The emitting species are, therefore, probably inter- The red needles were identified with an IR micro- mediate compounds containing hundreds of carbon scope as crystals of dicoronene (C48H20) suggesting an efficient polymerization process.15 Lewis16 and Lewis and Singer17 have studied the pyrolysis products in the Presented at the Fourth European Workshop on Fourier liquid phase around 500 °C of small PAHs, anthracene Transform Mass Spectrometry. Pont-à-Mousson, France,28-30 April 1997 (C14H10) and naphthalene (C10H8). They pr oposed that * Correspondence to: Christine Joblin the pyrolysis process is dominated by polymerization CCC 0951-4198/97/141619-05 $17.50 1997 by John Wiley & Sons, Ltd.
SIMULATION OF INTERSTELLAR AROMATIC HYDROCARBONS which produces mixtures of polymers with varyingdegrees of condensation.
Analysis of the sample using standard chemical separation methods18 led to the extraction of coroneneand dicoronene but not of larger oligomers. Tricor-onene and tetracoronene were first observed usingtime-of-flight secondary-ionization mass spectrome-try.19 In these experiments,the species were desorbedwith an infrared CO2 laser and further ionized by a pulsed UV laser beam. 20 The next goal was clearly to beable to isolate these large molecules and to study theirphysico-chemical properties and spectroscopy in condi-tions close to those found in astronomical environ- ments, in particular in high vacuum conditions.
In the present paper, we report some experiments Figure 1. The emission spectrum of the interstellar PAHs observed in
performed with the microprobe laser Fourier transform the Ml7 object by the European ISO satellite. The spectrum wasmeasured at the interface between a dense molecular cloud and a ion cyclotron resonance (FTICR) mass spectrometer highly UV-excited medium produced by surrounding young stars (see available at the LSMCL (Metz University). This tech- Ref. 5). The astronomical spectrum is compared with the laboratory nique is widely used for analytical purposes but spectrum of: (i) gas-phase coronene measured at 820 K10 and (ii) the presents further advantages for simulation experiments solid residue (pitch) obtained by pyrolysis of coronene at 660 °C. The on interstellar matter. In particular, very reactive spectrum of pyrolysed coronene was measured in absorption at roomtemperature through a pellet of the solid residue mixed with CsI salt species can be trapped in the ICR cell for long and the contribution of large dust particles ( ¸-l continuum) was durations, which enables further studies of their subtracted. An emission temperature of 820 K and 500 K was assumed for coronene and pyrolysed coronene respectively. Thisemission temperature depends on the size of the emitters (cf. Léger etal.11 for details). Note that the 3.3 ¹m band emitted by coronene at 820 K is much more intense than the astronomical feature.
The experimental set-up consists of a modified, differ-entially pumped dual cell FTMS 2000 (Finnigan, SanJose, CA, USA) which is operated in a 3 T magnetic Laser 266
field and is coupled with a reflection laser interface and special sample manipulation hardware. Both 266 nm(Nd-YAG laser) and 193 nm (excimer laser chargedwith ArF) wavelengths were used throughout this work.
A focusing telescope allowed the adjustment of thelaser spot diameter on the sample from 5 ¹m to about1 mm, corresponding to a power density ranging from 105 to 109 W cm-2.
All experiments were performed in the source cell at a 2 V trapping voltage. After a relaxation time of100 ms following the ionization step, the ions wereexcited by a radiofrequency chirp (110 Vp-p). The excitation and detection bandwidths were typtcally inthe range 1.0 to 2.6 MHz.
Laser 193
2.5 10' W cm
RESULTS AND DISCUSSION
At low irradiance (between 106 and 107 W cm-2), themass spectrum of the pyrolysed coronene sample is dominated by coronene and its oligomers (Figs 2(a), 3and 4). At higher irradiance new species generatedduring the laser ablation process are observed (Fig.
2(b), 5 and 6).
Building of large PAHs
Whereas dicoronene, tricoronene and tetracoroneneare easily observed, higher mass oligomers (up to theoctamer) have only been clearly observed in some Figure 2. Positive ion mass spectra obtained from laser ablation of the
experiments using the 266 nm Nd-YAG laser at an products of pyrolysis of coronene (C24H12) at 600 °C using (a) an irradiance of 5 x 106 W cm-2 (Fig. 2(a)). Condensed irradiance of 5 x 106 W cm-2 at 266 nm and (b) an irradiance of forms of the trimer and tetramer were identified at m/z 2.5 x 107 W cm-2 at 193 nm. In (a) the mass peaks correspond to the 888 and 1180 (M = 888 and 1180) respectively (Fig. 3).
oligomers of coronene (C24pH8p+4, with p=0,.,8). In (b) even These species result from the condensation of the most carbon-number clusters dominate the high-mass range. The close-upview in the C bent forms (Fig. 4).
84 to C124 range shows that the maxima observed in the mass distribution of Cn clusters fall close to the masses of fully Millon et al.21 showed that dicoronene and tricor- dehydrogenated coronene oligomers.
Rapid Communications in Mass Spectrometry, Vol. 11, 1619-1623 (1997) 1997 by John Wiley & Sons, Ltd. SIMULATION OF INTERSTELLAR AROMATIC HYDROCARBONS 1621 Laser 193 nm lrradiance
Laser 193 nm lrradiance 2.5 10 W
890.00 9 0 0 . 0 0
1140.0 1160.0 1160.0
Figure 3. A close-up view of the mass peaks corresponding to the coronene oligomers C48H20, C72H28, and C96H36 observed in positive-ion
mode using 193 nm laser ablation. Increasing the irradiance (b) clearly induces molecular dissociation, essentially by loss of pairs ofhydrogen atoms (H2?) but also by C2 loss Note that the condensed forms of the oligomers (Fig. 4) are observed even at low onene can be formed directly by UV laser ablation of generated plasma can induce saturation of carbon pure coronene at low irradiance (106 W cm-2 at bonds. In particular, the dominant mass 670 is probably 266 nm). However, the observation of higher mass a hydrogenated derivative of mass 668 obtained by oligomers as well as the results of the previous studies addition of three C2 to dicoronene (Fig. 6).
summarized above strongly support the idea that mostof the observed oligomers are present in the initial Building of large carbon clusters
pyrolysed coronene sample.
On the contrary, at high irradiance (typically At high irradiance, the high-mass range (m/z 1000 to 2 x 107 W cm-2 at 193 nm), new species appear in the 3000) is dominated by pure carbon clusters. The mass spectrum which are very probably products of the observed 24 u spacing is typical of fullerene-like species chemistry induced by the laser ablation process. In the which are easily produced by laser ablation of carbona- range m/z 600 to 700 (Fig. 5), there is strong evidence of ceous materials (see for instance Refs 21-25). These accretion of C2 by dicoronene leading to larger PAHs species were not observed in negative-ion mode, which (Fig. 6). Starting from dicoronene (M = 596), new supports the proposal that they are not present in the hexagonal cycles can be built by addition of C2 (Fig. 6).
initial pyrolysed coronene sample but produced during The accretion of C2 is best observed in experiments on the laser ablation process.21 In order to obtain the pure coronene at high irradiance (108 W cm-2) as curvature necessary for the formation of fullerene shown in Fig. 5. In these experiments a high concentra- cages, five-membered rings are required.26 The forma- tion of C2 is provided by the dissociation of coronene, tion of these rings might be related to a growth e.g. m/z 276 is observed instead of m/z 300 which mechanism by ion/molecule reactions in the laser- corresponds to the molecular mass of coronene.
generated plasma27,28 as suggested above (Fig. 6).
The formation of five-membered rings could also be Interestingly, the size distribution of the carbon involved. In particular, the series of peaks between m/z clusters that are produced by ablation of the pyrolysed 602 and 608 (Fig. 5), which differ from dicoronene by coronene sample (Fig. 2(b)) presents some maxima only one C atom, can be explained by the formation of which often fall very close to the carbon content of the five-membered rings as shown in Fig. 6(d). The dom- coronene oligomers (72, 96,120, 144, 168, 192 and 216 inance of odd-mass species such as m/z 607,605 and 603 carbon atoms for the trimer up to the nonamer). This is further indicates that the precursors are not fully different from the rather flat distribution which has aromatic. Indeed, for PAHs, only pairs of hydrogen been previously reported for clusters larger than C70, in atoms are involved (Fig. 3).
particular in the laser ablation of graphite and of Finally, hydrogen which is present in the laser- various small PAHs, including anthracene, chrysene, 1997 by John Wiley & Sons, Ltd. Rapid Communications in Mass Spectrometry, Vol. 11 , 1619-1623 (1997)
1622 SIMULATION OF INTERSTELLAR AROMATIC HYDROCARBONS pyrene29,30 and coronene21 (see also the present work).
This suggests that the large coronene oligomers couldcontribute to the formation of the observed carbonclusters provided that they are able to lose all theirhydrogen atoms to form PA networks, as was pointedout by Creasy and Brenna.23 The formation of full-erenes by curling up graphitic sheets in a collisionalenvironment was initially proposed by Kroto et al.31 butthis mechanism has not received much experimentalevidence.23 Still, Campbell et al.32 suggested thatfullerene-like species might be formed by rearrange-ment of polyimide ‘pieces'.
Figure 3 illustrates that, at high irradiance, series of peaks separated by 2 u are indeed observed which arecharacteristic of successively dehydrogenated PAHs.
For example, the peak at m/z 586 (Fig. 3(b)) corre-sponds to dicoronene which has lost ten hydrogenatoms. Larger oligomers also appear to be highlydehydrogenated, although the most dehydrogenatedspecies cannot be identified, because their mass peaksfall within the distribution corresponding to C2 loss. De Parseval33 has reported the production of C24 by irradiation of the coronene cation isolated in an ICRtrap; note that cyclo[24]carbon has been shown to beone precursor of fullerenes.34,35 Coronene oligomersmight then be fully dehydrogenated, provided that theyabsorb enough energy. The UV spectra of various PAHmixtures present a broad absorption band at 6 eV.36 The 6.44 eV photons from the excimer laser are Figure 5. A close-up view of the m/z 600 to 730 range corresponding
to the laser ablation at 193 nm of (a) pyrolysed coronene and (b) pure
therefore probably very efficiently absorbed by these coronene from Aldrich at an irradiance of 2.5 x 107 W cm-2 and molecules. After full dehydrogenation, the PA net- 108 W cm-2 respectively . The observed mass peaks are interpreted in works obtained may rearrange to fullerene cage-like terms of carbon and hydrogen addition to dicoronene (see Fig. 6).
clusters to saturate their dangling bonds.26,31,32 The This effect is strengthened in the coronene experiment where the excited clusters might then relax by evaporation of C plasma is certainly denser in hydrogen atoms and small carbon fragments. The spectrum between m/z 717 and 725 in (b) is fragments37,38 and further grow by addition of C2.39.
contaminated by electronic noise and should not be considered.
Figure 6. Building of larger PAHs from dicoronene. The accretion of
C2 leads to the growing of the graphitic plane (a) and (b) but could
also induce the formation of five-membered rings (c). These rings arelikely to account for the peaks around m/z 607 which can be derivedfrom dicoronene by addition of one carbon (d).
FUTURE WORK
Large aromatic species which might be laboratoryanalogues of interstellar PAHs have been isolated andtrapped in an ICR cell. The experimental apparatus ofthe LSMCL is not equipped with diagnostic tools which would enable further studies on the properties of the Figure 4. Coronene and its oligomers. There are three possible forms
trapped species and their IR signatures. Still, the first of tricoronene, the most bent form giving the condensed form at encouraging insights into the physical-chemistry of M = 888. The dotted bonds illustrate the condensation process which these species have been obtained, viz. facile dehydroge- is accompanied by hydrogen toss. For tetracoronene, one isomer ofmass 1188 and the condensed form at M = 1180 have been nation (with possible rearrangement into fullerene-like clusters), addition of carbon (C2, C) and hydrogen Rapid Communications in Mass Spectrometry, Vol. 11, 1619-1623 (1997)
1997 by John Wiley & Sons, Ltd.
SIMULATION OF INTERSTELLAR AROMATIC HYDROCARBONS atoms. Further studies on these compounds are the first 16. L. C. Lewis, Carbon 18, 191 (1980).
objectives of the PIRENEA experiment, a FTICRMS 17. L. C. Lewis and L. S. Singer, in Polynuclear aromatic compounds. L. B. Ebert, (Ed.), Advances in Chemistry Series 217. p. 269 apparatus devoted to astrophysics, which is now under construction at the CESR-CNRS in Toulouse.
18. J. Fetzer (unpublished work).
19. S. J. Clemett, et al. (in preparation).
20. S. J. Clemett, C. R. Maechling, R. N. Zare, P. D. Swan, R. M.
Walker, Science 262, 721 (1993).
21. E. Millon, J. V. Weber, B. Kubler. J. Theobald and J.F. Muller, The first author thanks Dr. L. d'Hendecourt, Dr. A. Léger, Dr. L.
Analusis 21, 319 (1993).
Allamandola, and Dr. F. Salama, for having given her the opportunity 22. W. R. Creasy and J. T. Brenna, Chem. Phys. 126, 453 (1988).
to begin and pursue these studies. We thank Dr. A. Klotz for useful 23. W. R. Creasy and J. T. Brenna, J. Chem. Phys. 92, 2269 (1990).
24. H. Y. So and C. L. Wilkins, J. Phys. Chem. 93, 1184 (1989).
25. P. F. Greenwood. M. G. Strachan. G. D. Willett and M. A. Wilson.
Or. Mass Spectrom. 25, 353 (1990).
26. Q.L. Zhang, S. C. O'Brien, J. R. Heath, Y. Liu, R. F. Curl, H. W.
Kroto and R. E. Smallev, J. Phys. Chem. 90, 525 (1986).
1. M. Giard. E Pajot, J. M. Lamarre, et al., Astron. and Astrophys. 27. A. O'Keefe, M. M. Ross and A.P. Baronavski, Chem. phys. Lett. 201, Ll (1988).
130, 17 (1986).
2. I. Ristorcelli. M. Giard, C. Meny, G. Serra, J. M. Lamarre, C. Le 28. S. W. McElvany, H. H. Nelson, A. P. Baronavski, C. H. Watson Naour, J. Leotin and F. Pajot, Atron. and Astrophys. 286, L23
and J. R. Eyler, Chem. Phys. Lett. 134, 214 (1987).
29. C. E. Brown, P. Kovacic, K. J. Welch, R. B. Cody, R. E. Hein and 3. F. Boulanger. W. T. Reach, A. Abergel, et al., Astron. and J. A. Kinsinger, J. Polym. Sci.: Part A: Polym. Chem. 26, 131
Astrophys. 315, L325 (1996).
4. K. Mattila. D. Lemke, L. K. Haikala, et al., Astron. and Astrophys. 30. D. N. Lineman, K. V. Somayajula, A. G. Sharkey and D. M.
315, L353 (1996).
Hercules, J. Phys. Chem, 93, 5025 (1989).
5. L. Verstraete, J. L. Puget, E. Falgarone, S. Drapatz, C. Wright and 31. H. W. Kroto, J. R. Heath, S. C. O'Brien, R. F. Curl and R. E.
R. Timmermann. Astron. and Astrophys. 315, L337 (1996).
Smalley, Nature 318, 162 (1985).
6. C. D. Andriesse, Astron. and Astrophys. 66, 169 (1978).
32. E. E. B. Camobell. G. Ulmer. B. Hasselberner. H. G. Busmann 7. K. Selleren. Astrophys. J. 277, 623 (1984).
and I. V. Heriel, J. Chem. Phys. 93, 6900 (1990).
H. A. Léger and J.L. Puget, Astron. and Astrophys. 137, L5
33. P. de Parseval, PhD thesis, Université Paris Sud (1997).
34. Y. Rubin, M. Kahr, C. B. Knobler, F. Diederich and C. L. Wilkins, 9. L. J. Allamandola. A. G. G. M. Tielens and J. R. Barker, J. Am. Chem. Soc. 113, 405 (1991).
Astrophys. J. 290, L25 (1985).
35. S. W. McElvany, M. M. Ross, N. S. Goroff and F. Diederich.
10. C. Joblin, L. d'Hendecourt, A. Léger and D. Défourneau, Astron. Science 259, 1594 (1993).
and Astrophys. 281, 923 (1994).
36. C. Joblin, A. Léger and P. Martin, Astrophys. J. 393, L79
11. A. Léger, L. d'Hendecourt and D. Défourneau, Astron. and Aslrophys., 216, 148 (1989).
37. S. C. O'Brien, J. R. Heath, R. F. Curl and R. E. Smalley, J. Chem. 12. 0. Guillois, I. Nenner, R. Papoular and C. Reynaud, Astron. and Phys. 88, 220 (1988).
Astrophys. 285, 1003 (1994).
38. S. Maryuma, L. R. Anderson and R. E. Smalley, Rev. Sci. Insrrum. 13. 0. Guillois, I. Nenner, R. Papoular and C. Reynaud, Astrophys. J. 61, 3686 (1990).
463, 810 (1996).
39. T. Pradeep and R. G. Cooks, Int. J. Mass Spectrom. Ion Processes 14. J. N. Rouzaud and A. Oberlin, Thin Solid Fibers 105, 75 (1983).
l35, 243 (1994).
15. C. Joblin, PhD thesis, Université Paris 7 (1992).
1997 by John Wiley & Sons, Ltd. Rapid Communications in Mass Spectrometry, Vol. 11,1619-I623 (1997)

Source: http://parseval.perso.neuf.fr/CV/Articles/Article6.pdf