Dissociative Photoionization of Polycyclic Aromatic Hydrocarbon

arXiv:1508.00259v1 [astro-ph.GA] 2 Aug 2015 ihsasa eprtrsi h ag 900– range the carbon- in of temperatures envelope medium at the interstellar in stars be- the formed rich molecules, into are PAH blown (ISM), that 1985). being proposed al. fore been et Allamandola has bands 1984; It emission Puget infrared & of (Leger inferred observation been the has from molecules (PAH) hydrocarbon Introduction 1. iscaiePoooiaino oyylcAoai Hydr Aromatic Polycyclic of Photoionization Dissociative ycrto OEL ’redsMrses an-ui P4 BP Saint-Aubin Merisiers, des L’Orme SOLEIL, Synchrotron h xsec fitrtla oyylcaromatic polycyclic interstellar of existence The aoaoyAtohsc ru fteMxPac nttt f Institute Planck Max the of Group Astrophysics Laboratory rsoytcnqe h daai oiaineeg a enfoun been coin has photoion energy app ionization photoelectron the adiabatic the The and with measured energy technique. been troscopy ionization have adiabatic ions Their ment 1-ethynylpyrene. species, and spec PAH individual ethynyl-substituted two the of of photoionization photostability sociative the on partly depends molecules tyypeatrn n t7.41 at and ethynylphenanthrene ld htteePHdrvtvsaea htsal stenon-su the as photostable upon as molecules are derivatives PAH th PAH ethynyl-substituted for these of that energy clude stability critical non- the the its estimate and and to cations molecule used a PAH H are ethynyl-substituted a measurements either of The loss for the similar of also indicative ion is fragment the of molecules energy PAH appearance non-substituted corresponding the for determined ujc headings: Subject role important an play may they molecules. medium, PAH interstellar the in present If nvriyJn,Isiueo oi tt hsc,Helmhol Physics, State Solid of Institute Jena, University h iecceo h ouaino neselrplcci rmtchydr aromatic polycyclic interstellar of population the of cycle life The a lnkIsiuefrAtooy ¨ngth 7 D-6911 Königstuhl 17, Astronomy, for Institute Planck Max .Rul´,S .Kanktk,D uvo n .Jäger C. and Fulvio, D. Krasnokutski, A. Rouillé, S. G. srceity—IM oeue oeua aa—molecula — data molecular — molecules ISM: — astrochemistry oeue ariga tyy Group Ethynyl an Carrying Molecules .A aca .F ag n .Nahon L. and Tang, X.-F. Garcia, A. G. [email protected] ± .2e o -tyyprn.Teevle r iia othose to similar are values These 1-ethynylpyrene. for eV 0.02 ABSTRACT h Henning Th. and 1 csit A oeue,peual through presumably molecules, PAH prod- into pyrolysis the ucts of required condensation are efficient K an shown for 1800 actually than have lower phase temperatures that gas the hy- small in of drocarbons pyrolysis Cau the with & Experiments Cherchneff 1999). 1994; Frenklach & al. al. Wang et et 1992; Cherchneff (Frenklach 1989; Feigelson & mechanism Frenklach 1985; (HACA) addition 70Ktruhtehydrogen-abstraction-C the through K 1700 rAtooya h rerc Schiller Friedrich the at Astronomy or ,F912GfsrYet ee,France Cedex, Gif-sur-Yvette F-91192 8, ze ,D073Jn,Germany Jena, D-07743 3, tzweg siue pce nH in species bstituted aey 9-ethynylphenanthrene namely, oso tmb h PAH the by atom H a of loss e hnnheeadprn.The pyrene. and phenanthrene o olwn photoionization following tom edleg Germany Heidelberg, 7 e.W aesuidtedis- the studied have We ies. t7.84 at d ntegot finterstellar of growth the in htinzto.W con- We photoionization. iec PPC)spec- (PEPICO) cidence usiue counterpart. substituted aac nryo frag- of energy earance processes r ± cro (PAH) ocarbon .2e o 9- for eV 0.02 I regions. ocarbon 2 H 2 - HACA (Jäger et al. 2007, 2009). In the various with that of regular ones. The latter comprise regions of the ISM, where temperatures are lower exclusively fused, six-membered aromatic carbon than 100 K, free-flying PAH molecules are thought cycles. They found that while a regular PAH to be formed through the destruction of carbona- would survive in H I regions, its substituted coun- ceous grains in shocks (see Chiar et al. 2013, and terparts would undergo fragmentation in the same references therein). Additionally, despite the low UV radiation field. This could suggest that the temperatures, PAH species could grow through ethynyl derivatives would not survive in H I re- a series of chemical reactions that require little gions, in contrast with our expectation that they or no activation energy. Accordingly, Mebel et al. are photostable (Rouilléet al. 2012, 2013). (2008) have introduced the ethynyl addition mech- The possible presence of ethynyl-substituted anism (EAM), which is barrierless with regard PAH molecules in the ISM and the lack of to activation energy, as a means to form PAH data concerning their photostability called for molecules in cold gas-phase environments, includ- an experimental study. We present here lab- ing the conditions of the ISM. oratory measurements on the dissociative pho- Ethynyl-substituted PAH species appear as in- toionization of two ethynyl-substituted PAH termediates in the HACA mechanism and the species, 9-ethynylphenanthrene (C16H10) and 1- EAM as well. Thus they may be present in ethynylpyrene (C18H10). The photostability in- the ISM, either as species ejected from stellar dex R as defined by Jochims et al. (1999) can be envelopes before their transformation into reg- derived from few experimental data, namely, the ular PAH molecules or as intermediates formed adiabatic ionization energy, the appearance en- locally. The HACA mechanism has been stud- ergy of the singly dehydrogenated fragment ion, ied as a formation process of PAH species by and the internal thermal energy of the parent taking into account the conditions that prevail ion. Their values have been determined in pho- in stellar envelopes (Frenklach & Feigelson 1989; toelectron photoion coincidence (PEPICO) spec- Cherchneff et al. 1992; Cherchneff & Cau 1999). troscopy experiments and, after deriving the value The EAM, however, has yet to be evaluated as of R for both ethynyl-substituted PAH species, we an interstellar growth process, i.e., with regard conclude that they are as photostable as the non- to the abundance of the ethynyl radical (C2H) substituted molecules. and considering the competition with destruction mechanisms, especially photofragmentation. Rel- 2. Experimental details atively high amounts of C2H have been found in the ISM (Tucker et al. 1974), in diffuse clouds The experiments were carried out at the beam- 1 (Lucas & Liszt 2000; Gerin et al. 2011; Liszt et al. line DESIRS (Nahon et al. 2012) of the syn- 2 2012), in translucent clouds (Turner et al. 2000), chrotron SOLEIL . This beamline delivers pho- in certain dark clouds (Wootten et al. 1980; tons with an energy in the range 5–40 eV, Ohishi et al. 1992), in circumstellar envelopes thus covering the VUV wavelength domain. (Huggins et al. 1984), in protoplanetary disks Photons at chosen energies were used to pho- (Henning et al. 2010), and in massive star-forming toionize PAH species in a molecular beam pro- 3 regions (Beuther et al. 2008). On the other hand, duced with the vacuum apparatus SAPHIRS the photostability of ethynyl-substituted PAH (Richard-Viard et al. 1996). The photoelectrons molecules has yet to be experimentally charac- and the photoions – parents and fragments – were terized. analyzed using the double-imaging PEPICO spectrometer DELICIOUS III (Garcia et al. 2013). Jochims et al. (1999) carried out laboratory Briefly, it consists of a velocity-map-imaging measurements on the photostability of several PAH molecules. While the set of substances in- 1Dichro¨ısme Et Spectroscopie par Interaction avec le Ray- cluded methyl-, vinyl-, and phenyl-substituted onnement Synchrotron species, it did not contain ethynyl derivatives. 2Source Optimisée de Lumière d’Energie Intermédiaire du Nevertheless, they defined a structure-dependent LURE index of photostability that allowed them to com- 3Spectroscopie d’Agrégats PHotoIonisés par le Rayon- pare the behavior of substituted PAH molecules nement Synchrotron

2 (VMI) electron analyzer and a modified Wiley- ber. This chamber is the first of the three differ- McLaren ion imaging spectrometer operated in entially pumped vacuum chambers of SAPHIRS, coincidence. The coincidence treatment provides which communicate by means of skimmers 1 mm photoelectron images that can be tagged by their in diameter. The jet and the skimmers separat- corresponding ion mass and ion translational en- ing the chambers were aligned so as to produce a ergy. From these images, photoelectron spectra molecular beam of He atoms and PAH molecules. can be obtained from all the masses in the molecu- In the third chamber, the molecular beam crossed lar beam simultaneously. The energy of the VUV the synchrotron radiation beam in the ionization radiation was calibrated using the photoioniza- volume of the PEPICO spectrometer. 1 + + 2 + tion of N2(X Σg ) into N2 (B Σu ) as a reference Adiabatic ionization energies were obtained (Innocenti et al. 2013). from the mass-selected photoelectron images We used samples of phenanthrene (C14H10, recorded at a fixed photon energy of 9 eV. The Aldrich, purity ≥99.5%), 9-ethynylphenanthrene proximity of this energy to the ionization thresh- (C16H10, Aldrich, purity 97%), pyrene (C16H10, old of the molecules led to the production of slow Aldrich, purity 99%), and 1-ethynylpyrene (C18H10, electrons and the possibility to lower the extrac- abcr, purity 96%), as received, without undertak- tion field to 90 V cm−1 with the consequent gain in ing a further purification. kinetic energy resolution (Garcia et al. 2013). The In each experiment, the sample powder was corresponding mass-selected photoelectron spec- heated in an oven placed in the source chamber tra were obtained from the images applying the of SAPHIRS, so as to increase its vapor pressure pBasex algorithm for Abel inversion (Garcia et al. up to a useful level. The oven temperature was 2004). A gas filter upstream the monochromator 383 K for phenanthrene, 493 K for pyrene, up was filled with Kr to effectively cut off the high to 443 K for 9-ethynylphenanthrene (413 K for harmonics of the undulator (Mercier et al. 2000). energy-dependent breakdown measurements and For these measurements, the photoelectrons and 443 K for photoelectron spectroscopy measure- ions were counted during 6483 s for phenanthrene, ments; see below in this Section), and 403 K for 3627 s for 9-ethynylphenanthrene, and 18034 s for 1-ethynylpyrene. In order to prevent its obstruc- 1-ethynylpyrene. tion by condensing sample vapor, the nozzle was The appearance energy of the ion fragments in- heated to a temperature 10 K higher than that dicative of dissociative photoionization was mea- of the oven. The phenanthrene and pyrene pow- sured by varying the photon energy from 15 to ders were poured in a stainless-steel boat that was 20 eV with a step of 0.1 eV. The gas filter was inserted into the oven in a free fit. A sufficient not used for these measurements as the high har- amount of the 9-ethynylphenanthrene sample was monics were not transmitted by the optics. At transferred into the gas phase by following the each energy the ion signal was integrated for 180 s same procedure, although most of the substance for all masses simultaneously. In the case of 1- polymerized and remained in the oven. In the ethynylpyrene, two sets of measurements were av- case of 1-ethynylpyrene, however, it was necessary eraged to improve the signal-to-noise ratio. In to finely ground the sample in a mortar; the fine this kind of experiments, the extraction field of powder was dispersed in a ball of ultrafine glass the spectrometer was set to provide full transmis- wool, which was placed in the boat in the oven. sion for all the electrons and ions present in the Increasing the surface area of the sample improved sample. All the data were corrected for false co- the transfer of molecules into the gas phase, allow- incidences and normalized for the incident photon ing us to work with a lower oven temperature and flux as measured by an IRD AXUV100 photodi- for a longer time with the same sample load. ode. In the oven, the PAH molecules diffused in an atmosphere of helium gas (Air Liquide, purity 3. Results ≥99.999%) provided with a backing pressure of 0.5 3.1. Adiabatic ionization energies to 1 bar. The mixed vapor leaked out of the oven through a nozzle 70 µm in diameter, producing The adiabatic ionization energies Ei of phenan- a continuous, supersonic jet in the source cham- threne, 9-ethynylphenanthrene, and 1-ethynylpyrene

3 have been determined by analyzing photoelectron differences not smaller than 0.15 eV. spectra. These spectra were extracted from the In addition to ionization energies, the photoelectron images obtained in the PEPICO measure- electron spectra of Figure 1 reveal low-lying elec- ments carried out with a fixed photon energy equal tronic states of the cations. The peak at 8.07 eV to 9 eV. The spectra are shown in Figure 1 and in the spectrum of phenanthrene likely marks the the experimental ionization energies are given in origin of a transition to such a state. This ion- Table 1. As detailed below, the ionization energy ization was measured at 8.10 eV by Boschi et al. of pyrene was taken from the literature. (1972) and semi-empirical calculations actually We have measured the adiabatic ionization en- predicted the first excited doublet state of the ergy of phenanthrene to demonstrate the cali- cation to lie about 0.5 eV above its ground state bration of the spectrometer as described in Sec- (Khan 1992; Parisel et al. 1992). Since the pho- tion 2. We found this energy at 7.88 ± 0.02 eV, toelectron spectra of 9-ethynylphenanthrene and in agreement with the various values reported in phenanthrene are essentially similar, we conclude the literature, which were determined with differ- that the cation of the derivative possesses an ent techniques: 7.86 ± 0.10 eV (Boschi et al. 1972, equivalent state. The spectrum measured with photoelectron spectroscopy), 7.90 ± 0.01–0.03 eV 1-ethynylpyrene clearly shows a feature with an (Jochims et al. 1994, photoion mass spectrometry, origin at 8.3 eV, indicating that the cation has an fully deuterated phenanthrene), 7.903 ± 0.005 eV electronic doublet state 0.9 eV above its ground (Thantu & Weber 1993, resonant two-photon ion- state. The pyrene cation also exhibits such a low- ization), 7.8914 eV (Hager & Wallace 1988, two- lying electronic state (Boschi & Schmidt 1972). laser ionization mass spectrometry), and 7.87 ± 0.02 eV (Gotkis et al. 1993, time-resolved pho- 3.2. Appearance energies toionization mass spectrometry, TPIMS). Breakdown graphs describe the fraction of all As the adiabatic ionization energy we have ob- detected ions, parent and fragments, as a function tained for phenanthrene is in agreement with the of the energy of the incident photons. values reported in the literature, we have adopted In Figure 2, the breakdown graph for the disso- for the ionization energy of pyrene the most re- ciative photoionization of phenanthrene (C H ) cent value of 7.415 ± 0.010 eV that was deter- 14 10 between 15 and 20 eV shows the appearance of mined in a PEPICO experiment (Mayer et al. three fragment ions revealing the loss by the 2011). It is consistent with other reports: 7.41 eV parent ion of one H atom (C H+), of two H (Boschi & Schmidt 1972, photoelectron spec- 14 9 atoms (C H+), and of two C and two H atoms troscopy), 7.45 ± 0.01–0.03 eV (Jochims et al. 14 8 (C H+). The same figure reports the break- 1994, photoion mass spectrometry), 7.4251 ± 12 8 down graph obtained in the experiment on 9- 0.0009 eV (Zhang et al. 2010, coupled resonance- ethynylphenanthrene (C H ), during which only enhanced multiphoton ionization and zero kinetic 16 10 the two fragment ions indicating the loss of one energy photoelectron spectroscopies, REMPI- H atom (C H+) and the loss of two H atoms ZEKE, value obtained by conversion of 59888 ± 16 9 (C H+) were detected. 7 cm−1), and 7.4256 eV (Hager & Wallace 1988, 16 8 two-laser ionization mass spectrometry). The breakdown graphs derived from the measurements on pyrene (C H ) and 1-ethynylpyrene For 9-ethynylphenanthrene and 1-ethynylpyrene, 16 10 (C H ) are displayed in Figure 3. In either case, the adiabatic ionization energies were found at 18 10 only the fragment ion corresponding to the loss 7.84 ± 0.02 eV and 7.41 ± 0.02 eV, respec- of a single H atom was detected in the 15–20 eV tively. Thus the ionization energy of the ethynyl- interval. This is consistent with the breakdown substituted PAH molecules is equal to that of graph in the recent PEPICO study of the disso- the corresponding non-substituted species within ciative photoionization of pyrene by West et al. the experimental error bars. This is in con- (2014b). In that study, the next fragment ion, trast to the lower ionization potentials reported which corresponds to the loss of two H atoms, ap- by Jochims et al. (1999) for methyl-, vinyl-, and peared at ≈20 eV, the upper limit of the energy phenyl-substituted PAH species with respect to domain we have scanned. the non-substituted molecules. They measured

4 The appearance energy Ea of a fragment ion, minimum photon energy for H atom loss upon i.e., the lowest photon energy for which it is ob- photoionization was also at least 2 eV lower in 1- served, is an effective value since it depends on and 2-methylnaphthalene than in naphthalene, for the sensitivity of the experiment. Indeed, a slower any time scale. fragmentation rate requires a longer time interval The different behaviors of the diverse PAH to produce a number of fragment ions that is suf- derivatives with respect to dissociation are ex- ficient for a sensible counting. Thus the time scale plained by the various degrees of conjugation of of the measurement affects the determination of their π electronic system (Jochims et al. 1999). the appearance energy – an effect known as the Presently, the CC triple bond of an ethynyl side kinetic shift – as illustrated with TPIMS experi- group yields a stronger conjugation of the π elec- ments (see, for instance, Lifshitz 1991). With the tronic orbitals over the whole molecule in compar- photoionization of phenanthrene in a trap, the ap- ison with the CC double bond of a vinyl group, pearance energy of the singly dehydrogenated ion the aromatic ring of a phenyl group, and the satu- Ea,−H was found to be 14.9 ± 0.1 eV when mea- rated C atom of a methyl group. This is valid for sured 24 µs after ionization and 14.0 ± 0.2 eV after the neutral species and the cations as well. As a 5 ms (Ling & Lifshitz 1998). Similarly, TPIMS ex- result, ethynyl-substituted PAH molecules exhibit periments on pyrene yielded 16.2 ± 0.2 eV at 24 µs a strengthened structure. and 15.2 ± 0.2 eV at 5 ms for the appearance en- With regard to the origin of the H atom lost by ergy of the singly dehydrogenated ion (Ling et al. the ethynyl-substituted PAH molecules upon pho- 1995). toionization, we propose that it is one of the H In our PEPICO spectroscopy experiment, the atoms attached to the aromatic rings rather than detected photofragmentations occurred within mi- the H atom of the ethynyl side chain. This propo- croseconds, the time interval necessary for the ion- sition is first suggested by the conjunction of simi- ized molecules to leave the acceleration zone. Fig- lar ionization energies and similar appearance en- ures 4 and 5 are expanded views of the breakdown ergies of the singly dehydrogenated ions for the graphs showing the onset of H loss by the cations regular and ethynyl-substituted PAH molecules. we have studied. Considering the microsecond The proposition in also supported by the fact that time scale of our measurements, the appearance the dissociation energy of a CH bond is lower energies determined for phenanthrene and pyrene, for PAH molecules (469 ± 6 kJ mol−1 for naph- respectively 15.4 ± 0.1 eV and 16.3 ± 0.1 eV, are thalene; see Reed & Kass 2000) than for acety- consistent with those derived previously in TPIMS lene (549 ± 4 kJ mol−1, Ervin et al. 1990) or di- experiments and reminded above. Moreover, in acetylene (539 ± 12 kJ mol−1, Shi & Ervin 2000). the case of pyrene, the value is the same as that This difference holds for the cations (4.48 eV reported by Jochims et al. (1994), i.e., 16.30 ± for the naphthalene cation and 5.92 eV for the 0.15 eV, showing that the two setups have equal acetylene cation; see, respectively, Ho et al. 1995; sensitivities in terms of fragmentation rate. van der Meij et al. 1988). Comparing the dissociative photoionization of Interestingly, the fragmentation behavior of the each ethynyl-substituted PAH molecule with the present ethynyl-susbtituted PAH cations differs corresponding non-substituted species, the ap- from that of ethynylbenzene (also called pheny- pearance energies of the singly dehydrogenated lacetylene). In an imaging PEPICO experiment, ions stand out by their proximity, with a differ- West et al. (2014a) did not found any H loss chan- ence of 0.5 eV for the phenanthrenes and 0.1 eV nel while they measured the loss of a C2H2 unit. for the pyrenes. In contrast, Jochims et al. (1999) Our observations of the dissociative photoioniza- had observed in their experiments that the mini- tion of the ethynyl-susbtituted PAH species are mum photon energy causing the loss of a H atom opposite since the loss of a H atom is clearly de- by a photoionized PAH molecule was at least tected while that of a C2H2 unit is not. 2 eV lower when the molecule carried a methyl, a vinyl, or a phenyl group. Prior to the latter 3.3. Photostability study, Gotkis & Lifshitz (1993) and Gotkis et al. (1993) had found in TPIMS experiments that the Jochims et al. (1999) designed the photosta-

5 bility index R in the framework of the Rice- i.e., the difference between the internal energy cor- Ramsperger-Kassel theory for unimolecular reac- responding to the slowest fragmentation rate that tion rates (RRK theory; Rice & Ramsperger 1927; can be detected in our experiment and the inter- Kassel 1928). The index was introduced to exam- nal energy giving the slowest rate for a fragmenta- ine the behavior of PAH derivatives with regard tion faster than a radiative deexcitation. We take to fragmentation relative to that of the regular the value 104 s−1 determined by Jochims et al. species. It compares Ec,−H, the critical internal (1994) for the slowest fragmentation rate that can energy for the loss of a H atom by a PAH cation, be detected because the appearance energy we 2 with E−H 10 , the internal energy causing the have measured in the case of pyrene is the same loss of a H atom by a regular PAH cation at the as theirs. The computed kinetic shifts are given 2 −1 rate of 10 s . While Ec,−H is derived from mea- in Table 1. They are consistent with the results 2 surements, E−H 10 is the result of a calculation of the TPIMS experiments that showed a kinetic using a model of the fragmentation of regular PAH shift of ≈1 eV for dissociations measured at the cations. The rate of 102 s−1 represents the limit microsecond time scale, taking the measurements under which fragmentation is not efficient as it at the millisecond time scale as references for infi- is slower than relaxation through emission of IR nite time (Ling et al. 1995; Ling & Lifshitz 1998). photons. Thus Like Jochims et al. (1994), we take the average internal thermal energy hE i to represent the ther- E t c,−H mal contribution E to the internal energy. The R = 2 . (1) t E−H (10 ) internal thermal energy is assumed to be purely vibrational by nature. Consequently the average The value of E 102 is computed using the −H internal thermal energy can be computed from the RRK theory. The reaction rate for the loss of a H vibrational partition function. Thus atom k−H is then defined with s s−1 hνi E0,−H hEti = , (6) k = ν 1 − , (2) exp(hνi/kT ) − 1 −H E Xi=1 where h is Planck’s constant, νi is the frequency where E is the internal energy of the PAH ion, s of mode number i, k is Boltzmann’s constant, and is the number of its vibrational modes, E0,−H = T is the vibrational temperature. For the calcula- 2.8 eV is the average activation energy for the loss tion, we use scaled theoretical frequencies. The vi- 16 −1 of a H atom by a PAH species, and ν = 10 s is brational modes of the neutral ethynyl derivatives a frequency factor (Jochims et al. 1994). It follows were previously computed in an application of that E−H (k−H), the internal energy correspond- the density functional theory (Rouilléet al. 2012). ing to this rate, is obtained with Collective scaling factors for frequencies below and above 2000 cm−1 were derived to fit the theoret- E 0,−H ical harmonic frequencies to band positions ob- E−H (k−H)= 1/(s−1) . (3) 1 − (k−H/ν) served for the substances imbedded in CsI pel- lets (Rouilléet al. 2012). We have calculated the The critical internal energy for fragmentation is vibrational modes of neutral phenanthrene and derived from the experimental appearance energy neutral pyrene at the same level of theory us- of the singly dehydrogenated ion E , corrected a,−H ing the same Gaussian 09 software (Frisch et al. for the kinetic shift ∆E, from which is subtracted 2009). Frequency scaling factors were derived for the ionization energy E . Moreover, the thermal i these two regular PAH molecules simultaneously contribution E to the internal energy has to be t using the spectra measured in argon matrix by taken into account. Thus Hudgins & Sandford (1998) as references. The scaling factor is 0.979 for all frequencies except Ec,−H = Ea,−H − ∆E − Ei + Et. (4) those of the CH stretching modes, for which it is The kinetic shift ∆E is evaluated according to 0.964. The vibrational temperature of the molecules 4 2 ∆E = E−H 10 − E−H 10 , (5) would be comprised between their translational 6 temperature and the temperature of the oven, the stances we have studied. As a consequence, they latter being given in Section 2. The translational would be photostable in H I regions. As seen temperature has been derived for each molecular at the beginning of the section, the stability in- species from its velocity along the molecular beam dex used throughout relies on the H loss occur- coordinates, which has been determined by ana- ring within a hot ground state, after internal con- lyzing the ion image. Each translational temper- version. Non-statistical processes such as direct ature has been rounded to the tens of K because dissociation from an excited repulsive state might unknown factors such as the exact composition of happen although previous works have favoured the the molecular beam give uncertainties that are of statistical path considered here (see Jochims et al. this order. Thus a temperature interval has been 1994; Allain et al. 1996a; Le Page et al. 2001). defined for each species using its translational tem- Figure 6 presents for comparison some of perature as the lower limit and the oven temper- the photostability index values determined by ature as the upper one. All temperatures and the Jochims et al. (1999) for various regular PAH internal thermal energies derived from them are molecules and derivatives along with those we reported in Table 1. Table 1 also contains the re- have derived by considering that the internal ther- sulting critical energies Ec,−H and photostability mal energy of the molecules reflected their transla- ∗ indexes R. Although the translational tempera- tional temperature. It also reports the Ec,−H −E ture differs largely from the oven temperature, the differences. The figure shows that the ethynyl presently modest contribution of Et (or hEti) to derivatives of PAH molecules are as photostable as Ec,−H results in a small dependence of R on T . the corresponding regular species while the other The first result of this study is the consistency derivatives are clearly less photostable and are ac- of the R values presently derived for phenanthrene tually expected to fragment upon photoionization and pyrene, (0.929–0.968) ± 0.017 and (1.057– in H I regions (Jochims et al. 1999). 1.114) ± 0.015, respectively, with those given by Jochims et al. (1999), 0.89 and 1.05, respectively. 4. Discussion One may note that the latter values, the accuracy of which was not given, lie at the lower end of In a theoretical study, Allain et al. (1996a) pro- the intervals we have obtained through our anal- posed a criterion for the survival of interstellar ysis. This may indicate that the vibrational tem- PAH molecules with respect to photofragmenta- perature of the molecules in the helium beam is tion. Defining the molecules by their carbon con- closer to their translational temperature than to tent, the survival requires that the rate for the the oven temperature. Nevertheless, the values loss of a C2H2 group be not faster than the rate are close to 1, which is the ideal index value for a for the addition of carbon atoms, either neutral regular PAH. The simplicity of the model used to or ionized. After computing and comparing the corresponding loss and addition rates for PAH determine E−H (k−H) leads to the observed devia- tions. The second result is the similarity between molecules, they reached the conclusion that only the index values obtained for an ethynyl derivative those containing more than 50 C atoms survive and for the corresponding regular PAH molecule. in the ISM. Of relevance to the present work, From this similarity we conclude that the ethynyl after Allain et al. (1996a) noted the production derivatives exhibit the same photostability as the of ethynyl derivatives through the HACA mech- regular species. anism, hence their possible presence in the ISM, they computed the loss rate of a C H group for In H I regions, molecules may interact with pho- 2 2 ethynyl-substituted PAH molecules. The values tons carrying a maximum energy of 13.6 eV. The they reported did not indicate a substantial dif- stability of a neutral PAH molecule upon pho- ference with regular PAH species, which is consis- toionization is evaluated by comparing the criti- tent with our experimental breakdown graphs for cal energy for the loss of a H atom by a cation, phenanthrene and ethynylphenanthrene. These E , with the maximum internal energy that c,−H graphs would even suggest a faster rate for the can be transferred to this cation, E∗, where E∗ = loss of a C2H2 group by non-substituted species 13.6 − Ei (Jochims et al. 1999). As shown in Ta- ∗ since the corresponding fragment was not observed ble 1, Ec,−H is larger than E for the four sub-

7 for the ethynyl derivative. This single instance e.g., C2H (see Introduction). In the EAM scheme, of comparison, however, may not be representa- a PAH molecule would undergo a series of reac- tive. We can still state, taking the survival crite- tions with C2H leading to the formation of new rion defined by Allain et al. (1996a), that ethynyl- aromatic cycles and, as a consequence, the growth substituted PAH molecules containing more than of the molecule (Mebel et al. 2008; Jones et al. 50 C atoms survive in the ISM as far as photofrag- 2010). As an experimental argument favorable mentation is concerned. to the EAM, it was found that C2H reacted Rather than defining the destruction of a PAH with benzene (C6H6) in the gas phase in a bar- molecule by the loss of C atoms, one may regard rierless mechanism that produces ethynylben- the loss of a H atom by the PAH molecule as the zene (phenylacetylene, C6H5C2H) plus a H atom decisive step toward further destruction. Indeed, (Goulay & Leone 2006; Jones et al. 2011). Note the loss of a H atom is the unimolecular dissocia- that C2H reacts also with acetylene (C2H2), giving tion mechanism that requires the least internal diacetylene (C4H2) and a H atom. This process is energy (Jochims et al. 1999) and the dehydro- also free of energy barrier at the entrance channel genation results in the weakening of the molec- as shown by Lee et al. (2000) with another exper- ular structure (Allain et al. 1996b; Jochims et al. iment in the gas phase. Thus C2H may react with 1999). Using Jochims et al.’s photostability in- ethynyl-substituted PAH molecules under inter- dex R, which is built on the loss of a H atom, we stellar conditions, at an aromatic cycle or at the have determined that 9-ethynylphenanthrene and side chain, as suggested by the different paths de- 1-ethynylpyrene were as photostable as phenan- scribed by Mebel et al. (2008) for the reaction of threne and pyrene, respectively. These four C2H with ethynylbenzene. molecules were also found to be stable upon pho- The C2H radical, however, is not the only abun- toionization in H I regions, unlike every PAH dant interstellar species that may be involved in molecule carrying a methyl, vinyl, or phenyl group the growth of PAH molecules. Neutral atomic studied by Jochims et al. (1999). Note that the carbon (C I) is also abundant in the diffuse same study reported the dihydrogenated PAH ISM (Jenkins & Tripp 2011) and in denser re- species as undergoing dissociative photoionization gions (see, for instance, Phillips & Huggins 1981; in H I regions, too, as shown in Figure 6. As Schilke et al. 1995; Stark et al. 1996; Beuther et al. mentioned in Section 3.2, the degree of conjuga- 2014). This induced Allain et al. (1996a) to view tion of the π orbitals affects the stability of the the addition of carbon atoms as the growth mech- molecular structures. Concerning dihydrogenated anism that would compete effectively with the derivatives, the addition of H atoms lowers the destruction of interstellar PAH molecules. They conjugation by creating adjacent single CC bonds, assumed that C atoms would react with PAH weakening the structure. As a conclusion, by ex- molecules under the conditions of the ISM and amining the photostability of PAH molecules using that the reactions would contribute to the growth R, we find that the ethynyl derivatives of regular of the molecules rather than to their destruc- PAH molecules that survive in the UV radiation tion. It has actually been demonstrated exper- field of H I regions would also survive in this envi- imentally that C atoms react with benzene in ronment, i.e., in most of the ISM. In terms of size, a barrierless addition that gives a radical with Jochims et al. (1994) found that survival required a seven-membered carbon ring plus a H atom of the regular PAH molecules that they contained (Hahndorf et al. 2002). Therefore barrierless re- at least 30 C atoms. While this limit was derived actions between C atoms and PAH molecules can from observing the fragmentation of the cations, be considered. Such reactions, however, have yet it also applies to the neutrals, as stated at the end to be studied in the laboratory. Finally, experi- of Section 3.3. The same requirement with respect ments have shown that C atoms react with C2H2 to size can be extended to the ethynyl derivatives molecules without energy barrier, where C + C2H2 in view of their similar fragmentation pattern. → C3H + H (Kaiser et al. 1996; Chastaing et al. As photostable interstellar molecules, ethynyl- 2001). Thus it is worthwhile to examine whether substituted PAH species would survive long carbon atoms react with ethynyl-substituted PAH enough to react with other interstellar substances, species, either at the polycyclic aromatic moiety

8 or at the ethynyl side group. measured the UV/vis absorption bands of C2H-

In addition to reactions involving C2H and and C4H-substituted species isolated in Ne matri- C I, one could also think of a possible role ces. The spectra led us to propose photostable of reactions with methylidyne (the CH radical, substituted PAH molecules, or similar species, also carbyne) in the growth of PAH molecules. as valid candidates for the carriers of the DIBs Methylidyne is abundant in the ISM (see, for in- (Rouilléet al. 2012, 2013). It was indeed ob- stance, Danks et al. 1984; Gerin et al. 2010) and served that the addition of an acetylenic side is known to react at low temperatures with C2H2 chain to a regular PAH lowers the peak inten- (Canosa et al. 1997; Maksyutenko et al. 2011) as sity and broadens the profile of the strong absorp- well as with anthracene (Goulay et al. 2006). tion bands that these species typically show at UV Thus several types of neutral-neutral reactions wavelengths (the β-bands), where DIBs are not may participate to the chemical network that con- observed (Clayton et al. 2003; Gredel et al. 2011; stitutes the growth process of PAH species in the Salama et al. 2011; Bhatt & Cami 2015). Conse- ISM. When considering barrierless reactions only, quently regular PAH molecules have become less the growth rate depends essentially on the fre- plausible as potential DIB carriers. Nonetheless, quency of collisions between the reactants, i.e., they may contribute with other PAH species to the on the abundance of the latter and on the tem- galactic extinction curve as a rich mixture of PAH perature of the environment. Any environment, substances gives a smooth absorption spectrum however, has a limited lifetime even if it is of (Steglich et al. 2011, 2012). The population of a the order of 107 yr as in the case of molecular DIB carrier would be larger than that of any other clouds (see, for instance, Larson 1981). It has PAH substance due to a favorable photoselective yet to be examined whether the abundance of the mechanism. The present measurements on the dis- relevant atomic and molecular species in the di- sociative photoionization of ethynyl-substituted verse interstellar environments allows the growth PAH molecules reveal that such species can sur- of PAH molecules within the lifetime of these en- vive in the UV radiation field of H I regions until vironments. Moreover, we have only considered they react with another substance. As the time neutral-neutral chemical reactions that may lead interval between the formation and the chemical to the addition of C atoms to PAH molecules, thus destruction of the molecules would be long enough contributing to their growth. The complete chem- to allow them to absorb photons and relax numer- istry of PAH species in the ISM, however, com- ous times, the molecules would leave their mark prises other mechanisms, e.g., ion-neutral reac- in the absorption spectrum of the ISM. The cor- tions, electron attachment, double ionization, and responding bands would peak out of the galactic multiphoton processes, which have been evaluated extinction curve provided that large enough popu- with various models (Le Page et al. 2001, 2003; lations were formed. Thus the present experimen- Montillaud et al. 2013). In support to double ion- tal results give support to our proposal. izazion, a recent experimental study by Zhen et al. (2015) concluded that, in the ISM, photoioniza- 5. Conclusions tion prevailed over fragmentation in PAH species The study of the dissociative photoionization of containing about 50 C atoms and more, leading to two ethynyl-substituted PAH species has yielded the formation of doubly and even triply charged their photoionization energy and the appearance cations. As a consequence, these large species energy of some photofragments. While the adia- were found to be photostable in interstellar condi- batic ionization energies are not affected by the tions, in agreement with previous theoretical stud- addition of the ethynyl chain, the appearance en- ies (Allain et al. 1996a; Montillaud et al. 2013). ergies for the H-loss channel are slightly increased Beside their place in the life cycle of inter- in the ethynyl derivatives, especially in the case of stellar PAH molecules, ethynyl derivatives are in- phenanthrene where an inhibition of the C2H2-loss teresting for their spectroscopic properties that channel is also seen. suggest a link with the still unidentified carriers We have determined the photostability index R of the diffuse interstellar bands (DIBs; see, for of each ethynyl derivative and found that it is sim- instance, Herbig 1995). In previous works, we

9 ilar to the value derived for the corresponding non- sistance in using beamline DESIRS. The financial substituted PAH molecule. Thus these derivatives support of the CALIPSO Consortium for Transna- are as photostable as the regular PAH species. tional Access funded by the European Commission Consistently, having found through the determina- under its Seventh Framework Program is grate- ∗ tion of Ec,−H −E that the regular PAH molecules fully acknowledged by GR and SK. phenanthrene and pyrene are stable upon photoionization by the photons typical of H I regions, REFERENCES we have found that their ethynyl derivatives would Allain, T., Leach, S., & Sedlmayr, E. 1996, A&A, also survive this process. 305, 602 Our measurements on phenanthrene and its ethynyl derivative suggest that the presence of the Allain, T., Leach, S., & Sedlmayr, E. 1996, A&A, side chain does not increase the unimolecular re- 305, 616 action rate for the loss of a C2H2 unit. This supports the theoretical results of Allain et al. Allamandola, L. J., Tielens, A. G. G. M., & (1996a) who did not found a significant variation Barker, J. R. 1985, ApJ, 290, L25 of this rate when examining ethynyl-substituted Bhatt, N. H. & Cami, J. 2015, ApJS, 216, 22 and non-substituted PAH species. Thus ethynyl-substituted PAH molecules ap- Beuther, H., Ragan, S. E., Ossenkopf, V., et al. pear to be as photostable as the non-substituted 2014, A&A, 571, A53 species whether one considers the photostabil- Beuther, H., Semenov, D., Henning, Th., & Linz, ity index R defined by Jochims et al. (1999) or H. 2008, ApJ, 675, L33 the survival criterion proposed by Allain et al. (1996a). Size matters for survival, however. As Boschi, R., Murrell, J. N., & Schmidt, W. 1972, a consequence, a conservative conclusion is that Faraday Discuss. Chem. Soc., 54, 116 ethynyl-substituted PAH molecules containing at least 50 C atoms would survive the interstellar UV Boschi, R. & Schmidt, W. 1972, Tetrahedron radiation field. Whether they are injected into the Lett., 13, 2577 ISM or formed there, the molecules would grow Brand, W. A. & Baer, T. 1983, Int. J. Mass Spec- through EAM at a rate that depends on the abun- trometry Ion Phys., 49, 103 dance of C2H. Reactions with C I and CH may also play a role in the growth of the PAH species Canosa, A., Sims, I. R., Travers, D., Smith, I. W. and ought to be studied experimentally. We have M., & Rowe, B. R. 1997, A&A, 323, 644 mentioned neutral-neutral reactions because experimental results are reported in the literature, Chastaing, D., Le Picard, S. D., Sims, I. R., & yet processes involving reactions at low tempera- Smith I. W. M. 2001, A&A, 365, 241 ture between neutral PAH molecules and ions such Cherchneff, I. & Cau, P. 1999, in IAU Symp. II as the abundant C should be investigated as 191, Asymptotic Giant Branch Stars, ed. T. Le well. Bertre, A. Lèbre, & C. Waelkens (San Fran- Finally, the present results support our pro- cisco, CA: ASP), 251 posal that DIB carriers could be PAH molecules whose photostability is enhanced by the presence Cherchneff, I., Barker, J. R., & Tielens, A. G. G. of side groups. M. 1992, ApJ, 401, 269 Chiar, J. E., Tielens, A. G. G. M., Adamson, A. This work was carried out within the frame- J., & Ricca, A. 2013, ApJ, 770, 78 work of a cooperation between the Max-Planck- Institut für Astronomie, Heidelberg, and the Clayton, G. C., Gordon, K. D., Salama, F., et al. Friedrich-Schiller-Universität Jena. We acknowl- 2003, ApJ, 592, 947 edge SOLEIL for provision of synchrotron radiation facilities under project No. 20140337 and Danks, A. C., Federman, S. R., & Lambert, D. L. we would like to thank Jean-Fran¸cois Gil for as- 1984, A&A, 130, 62

10 Ervin, K. M., Gronert, S., Barlow, S. E., et al. Hudgins, D. M. & Sandford, S. A. 1998, J. Phys. 1990, J. Am. Chem. Soc., 112, 5750 Chem. A, 102, 329 Frenklach, M., Clary, D. W., Gardiner, W. C., & Huggins, P. J., Carlson, W. J., & Kinney, A. L. Stein, S. E. 1985, Symposium (International) 1984, A&A, 133, 347 on Combustion, 20, 887 Innocenti, F., Eypper, M., Stranges, S., et al. Frenklach, M. & Feigelson, E. D. 1989, ApJ, 341, 2013, J. Phys. B: At. Mol. Opt. Phys., 46, 372 045002 Frisch, M. J., Trucks, G. W., Schlegel, H. B., et al. Jäger, C., Huisken, F., Mutschke, H., Llamas- 2009, Gaussian 09, Revision A.02 (Wallingford Jansa, I., & Henning, Th. 2009, ApJ, 696, 706 CT: Gaussian, Inc.) Jäger, C., Huisken, F., Mutschke, H., et al. 2007, Garcia, G. A., Cunha De Miranda, B. K., Tia, Carbon, 45, 2981 M., Daly, S., & Nahon, L. 2013, Rev. Sci. In- Jenkins, E. B. & Tripp, T. M. 2011, ApJ, 734, 65 strum., 84, 053112; erratum, Rev. Sci. Instrum., 84, 069902 Jochims, H. W., Baumgärtel, H., & Leach, S. 1999, ApJ, 512, 500 Garcia, G. A., Nahon, L., & Powis, I. 2004, Rev. Sci. Instrum., 75, 4989 Jochims, H. W., Rühl, E., Baumgärtel, H., Tobita, S., & Leach, S. 1994, ApJ, 420, 307 Gerin, M., De Luca, M., Goicoechea, J. R., et al. 2010, A&A, 521, L16 Jochims, H. W., Rühl, E., Baumgärtel, H., Tobita, S., & Leach, S. 1997, Int. J. Mass Spectrom. Ion Gerin, M., Ka´zmierczak, M., Jastrzebska, M., et Process., 167/168, 35 al. 2011, A&A, 525, A116 Jones, B., Zhang, F., Maksyutenko, P., Mebel, A. Gotkis, I. & Lifshitz, C. 1993, Org. Mass Spec- M., & Kaiser, R. I. 2010, J. Phys. Chem. A., trom., 28, 372 114, 5256 Gotkis, Y., Oleinikova, M., Naor, M., & Lifshitz, Jones, B. M., Zhang, F., Kaiser, R. I., et al. 2011, C. 1993, J. Phys. Chem., 97, 12282 Proc. Natl. Acad. Sci. U.S.A., 108, 452 Goulay, F. & Leone, S. R. 2006, J. Phys. Chem. Kaiser, R. I., Ochsenfeld, C., Head-Gordon, M., A, 110, 1875 Lee, Y. T., & Suits, A. G. 1996, Science, 274, 1508 Goulay, F., Rebrion-Rowe, C., Biennier, L., et al. 2006, J. Phys. Chem. A, 110, 3132 Kassel, L. S. 1928, J. Phys. Chem., 32, 225 Gredel, R., Carpentier, Y., Rouillé, G., et al. 2011, Khan, Z. H. 1992, Acta Phys. Pol. A, 82, 937 A&A, 530, A26 Larson, R. B. 1981, MNRAS, 194, 809 Hager, J. W. & Wallace, S. C. 1988, Anal. Chem., 60, 5 Lee, S., Samuels, D. A., Hoobler, R. J., & Leone, S. R. 2000, J. Geophys. Res., 105, 15085 Hahndorf, I., Lee, Y. T., Kaiser, R. I., et al. 2002, Leger, A. & Puget, J. L. 1984, A&A, 137, L5 J. Chem. Phys., 116, 3248 Le Page, V., Snow, T. P., & Bierbaum, V. M. 2001, Henning, Th., Semenov, D., Guilloteau, St., et al. ApJS, 132, 233 2010, ApJ, 714, 1511 Le Page, V., Snow, T. P., & Bierbaum, V. M. 2003, Herbig, G. H. 1995, ARA&A, 33, 19 ApJ, 584, 316 Ho, Y.-P., Dunbar, R. C., & Lifshitz, C. 1995, J. Lifshitz, C. 1991, Int. J. Mass Spectrom. Ion Pro- Am. Chem. Soc., 117, 6504 cess., 106, 159

11 Ling, Y., Gotkis, Y., & Lifshitz, C. 1995, Eur. J. Salama, F., Galazutdinov, G. A., Krelowski, J., et Mass Spectrom., 1, 41 al. 2011, ApJ, 728, 154 Ling, Y. & Lifshitz, C. 1998, J. Phys. Chem. A, Schilke, P., Keene, J., Le Bourlot, J., Pineau des 102, 708 Forêts, G., & Roueff, E. 1995, A&A, 294, L17 Liszt, H., Sonnentrucker, P., Cordiner, M., & Shi, Y. & Ervin, K. M. 2000, Chem. Phys. Lett., Gerin, M. 2012, ApJ, 753, L28 318, 149 Lucas, R. & Liszt, H. S. 2000, A&A, 358, 1069 Stark, R., Wesselius, P. R., van Dishoeck, E. F., & Laureijs, R. J. 1996, A&A, 311, 282 Maksyutenko, P., Zhang, F., Gu, X., & Kaiser, R. I. 2011, Phys. Chem. Chem. Phys., 13, 240 Steglich, M., Bouwman, J., Huisken, F., & Hen- ning, Th. 2011, ApJ, 742, 2 Mayer, P. M., Blanchet, V., & Joblin, C. 2011, J. Chem. Phys., 134, 244312 Steglich, M., Carpentier, Y., Jger, C., et al. 2012, A&A, 540, A110 Mebel, A. M., Kislov, V. V., & Kaiser, R. I. 2008, J. Am. Chem. Soc., 130, 13618 Thantu, N. & Weber, P. M. 1993, Z. Phys. D, 28, 191 Mercier, B., Compin, M., Prevost, C., et al. 2000, J. Vac. Sci. Technol. A, 18, 2533 Tucker, K. D., Kutner, M. L., & Thaddeus, P. 1974, ApJ, 193, L115 Montillaud, J., Joblin, C., & Toublanc, D. 2013, A&A, 552, A15 Turner, B. E., Herbst, E., & Terzieva, R. 2000, ApJS, 126, 427 Nahon, L., De Oliveira, N., Garcia, G. A., et al. 2012, J. Synchrotron Rad., 19, 508 van der Meij, C. E., van Eck, J., & Niehaus, A. 1988, Chem. Phys., 119, 135 Ohishi, M., Irvine, W. M., & Kaifu, N. 1992, in IAU Symp. 150, Astrochemistry of Cosmic Phe- Wang, H. & Frenklach, M. 1994, J. Phys. Chem., nomena, ed. P. D. Singh (Dordrecht: Kluwer), 98, 11465 171 West, B., Sit, A., Bodi, A., Hemberger, P., & Parisel, O., Berthier, G., & Ellinger, Y. 1992, Mayer, P. M. 2014a, J. Phys. Chem. A, 118, A&A, 266, L1 11226 Phillips, T. G. & Huggins, P. J. 1981, ApJ, 251, West, B., Useli-Bacchitta, F., Sabbah, H., et al. 533 2014b, J. Phys. Chem. A, 118, 7824 Reed, D. R. & Kass, S. R. 2000, J. Mass Spec- Wootten, A., Bozyan, E. P., Garrett, D. B., Loren, trom., 35, 534 R. B., & Snell, R. L. 1980, ApJ, 239, 844 Richard-Viard, M., Delboulbé, A., & Vervloet, M. Zhang, J., Han, F., & Kong, W. 2010, J. Phys. 1996, Chem. Phys., 209, 159 Chem. A, 114, 11117 Rice, O. K. & Ramsperger, H. C. 1927, J. Am. Zhen, J., Castellanos, P., Paardekooper, D. M., et Chem. Soc., 49, 1617 al. 2015, ApJ, 804, L7 Rouillé, G., Steglich, M., Carpentier, Y., et al. 2012, ApJ, 752, 25 Rouillé, G., Jäger, C., Huisken, F., & Henning, Th. 2013, Proc. Int. Astron. Union 9 (S297), 276 This 2-column preprint was prepared with the AAS LATEX macros v5.2.

12 1.0

phenanthrene

C H

14 10

0.8

0.6

0.4

0.2

7.88 eV

0.0

1.0

9-ethynylphenanthrene

C H

16 10

0.8

0.6

0.4

0.2

7.84 eV

0.0 normalized photoelectron signal [arbitrary unit] [arbitrary signal photoelectron normalized

1.0

1-ethynylpyrene

C H

18 10

0.8

0.6

0.4

0.2

7.41 eV

0.0

6.0 6.5 7.0 7.5 8.0 8.5 9.0

electron binding energy [eV]

Fig. 1.— Photoelectron spectra measured with photons of 9 eV in energy. The vertical dotted lines indicate the adiabatic ionization energies. The error bars are obtained assuming a Pois- son distribution on the image pixel intensities and then propagating the incertitudes across all alge- bra operations involving the Abel inversion.

13 100

+ +

C H C H

14 10 16 10

90 100

phenanthrene cation 9-ethynylphenanthrene

fragmentation cation fragmentation

+ + +

C H C H C H

14 9 12 8 14 8 +

10 C H

16 9 species fraction fraction species

C H 5

16 8

15 16 17 18 19 20 15 16 17 18 19 20

photon energy [eV] photon energy [eV]

Fig. 2.— Breakdown graphs of the dissociative photoionization of phenanthrene and 9- ethynylphenanthrene. The error bars result from a Poisson distribution on the integrated ion signal and standard-error propagation formulae.

14 100

+ +

C H C H

16 10 18 10

95 100

pyrene cation 1-ethynylpyrene

fragmentation cation fragmentation

+ +

C H C H 5 species fraction fraction species

16 9 18 9

15 16 17 18 19 20 15 16 17 18 19 20

photon energy [eV] photon energy [eV]

Fig. 3.— Breakdown graphs of the dissociative photoionization of pyrene and 1-ethynylpyrene. The error bars are described in the caption of Fig- ure 2.

15 1.0

+ +

C H fragment of C H fragment

14 9 16 9

the phenanthrene of the

0.8

cation 9-ethynyl-

phenanthrene 100

cation 0.6

0.4

0.2 species fraction fraction species

0.0

15 16 17 15 16 17

photon energy [eV] photon energy [eV]

Fig. 4.— Onset of H loss by the phenanthrene and 9-ethynylphenanthrene cations. The error bars are described in the caption of Figure 2. The horizontal dotted lines frame the background signal, which is centered at the mean value of the points in the interval defined with the horizontal bar and extends on both side by one sample standard devi- ation. The arrow indicates the first of two consec- utive points for which the signal and its error bar are above the background, the energy corresponding to this point being taken as the appearance energy of the fragment ion. The thick gray line represents a centered five-point moving average of the measurements.

16 1.0

+ +

C H fragment C H fragment of the

16 9 18 9

of the pyrene 1-ethynylpyrene

0.8

cation cation 100

0.6

0.4

0.2 species fraction fraction species

0.0

15 16 17 15 16 17

photon energy [eV] photon energy [eV]

Fig. 5.— Onset of H loss by the pyrene and 1- ethynylpyrene cations. The error bars are described in the caption of Figure 2. The other items are deﬁned in the caption of Figure 4.

17 This work

Jochims et al. (1999)

naphthalene

anthracene-d

phenanthrene

pyrene

9-ethynylphenanthrene

1-ethynylpyrene

1-methylnaphthalene

2-methylnaphthalene

1-methylanthracene

2-methylanthracene

9-methylanthracene

2-methylphenanthrene

2-vinylnaphthalene

1-phenylnaphthalene

1,2-dihydronaphthalene

9,10-dihydrophenanthrene

0.4 0.6 0.8 1.0 1.2 -2 -1 0 1 2

R E E * [eV]

c, H

Fig. 6.— Comparison of the R values (left) and ∗ of the Ec,−H − E diﬀerences (right) for several PAH molecules. The vertical dotted line in the right panel indicates the limit above which PAH molecules are stable in the UV radiation ﬁeld of H I regions.

18 Table 1 Photostability Index of PAH Molecules and Related Parameters

∗ a Ei E Ea,−H ∆E T hEti Ec,−H Species (eV) (eV) (eV) (eV) (K) (eV) (eV) R phenanthrene 7.88 ± 0.02 5.72 15.4 ± 0.1 0.925 190 0.060 6.655 ± 0.120 0.929 ± 0.017 383 0.340 6.935 ± 0.120 0.968 ± 0.017 9-ethynylphenanthrene 7.84 ± 0.02 5.76 15.9 ± 0.1 1.013 320 0.290 7.337 ± 0.120 0.956 ± 0.016 413 0.513 7.560 ± 0.120 0.985 ± 0.016 b pyrene 7.415 ± 0.010 6.185 16.3 ± 0.1 1.013 320 0.241 8.112 ± 0.110 1.057 ± 0.015 493 0.672 8.544 ± 0.110 1.114 ± 0.015 1-ethynylpyrene 7.41 ± 0.02 6.19 16.4 ± 0.1 1.101 190 0.089 7.978 ± 0.120 0.975 ± 0.015 403 0.525 8.414 ± 0.120 1.028 ± 0.015

∗ Note.—Ei: ionization energy; E = 13.6 − Ei: maximum internal energy of the parent cation in H I regions; Ea,−H: appearance energy of the singly dehydrogenated fragment ion; ∆E: kinetic shift; T : vibrational temperature; hEti: average thermal energy; Ec,−H: critical energy for the loss of a H atom by the parent cation; and R: photostability index. a Two values are given for each species: the translational temperature and the oven temperature, which are respectively considered as the lowest and highest values possibly taken by the vibrational temperature. b Mayer et al. (2011).