Electronic spectroscopy of FUV-irradiated diamondoids: A combined experimental and theoretical study
M. Steglich, F. Huisken, J. E. Dahl, R. M. K. Carlson, Th. Henning
aa r X i v : . [ phy s i c s . c h e m - ph ] A p r published in The Astrophysical Journal 729 (2011) 91, Received 2010 November 25; accepted 2011 January 5 Preprint typeset using L A TEX style emulateapj v. 11/10/09
ELECTRONIC SPECTROSCOPY OF FUV-IRRADIATED DIAMONDOIDS:A COMBINED EXPERIMENTAL AND THEORETICAL STUDY
M. Steglich & F. Huisken Laboratory Astrophysics Group of the Max Planck Institute for Astronomy at the Friedrich Schiller University Jena,Institute of Solid State Physics, Helmholtzweg 3, D-07743 Jena, Germany
J. E. Dahl
Geballe Lab for Advanced Materials, Stanford University, 476 Lomita Mall, Stanford, CA 94305, USA
R. M. K. Carlson
MolecularDiamond Technologies, Chevron Technology Ventures, 100 Chevron Way, Richmond, CA 94802, USA
Th. Henning
Max Planck Institute for Astronomy, K¨onigstuhl 17, D-69117 Heidelberg, Germany published in The Astrophysical Journal 729 (2011) 91, Received 2010 November 25; accepted 2011 January 5
ABSTRACTIrradiation with high energy photons (10.2 − σ − σ ∗ spectrum displays surprisingly strong similarities tometeoritic nanodiamonds containing 50 times more C atoms. Subject headings: astrochemistry — dust, extinction — ISM: molecules — methods: laboratory —molecular data — techniques: spectroscopic INTRODUCTION
Diamond-like material is expected to be abundantin the interstellar medium (Henning & Salama 1998).Small nanodiamonds (2 − hybridized with hydrogens saturat-ing the dangling bonds on the surface. The diamond-likestructure results in a remarkable rigidity, strength, andthermodynamic stability, especially compared to otherhydrocarbons. The smallest of these species, adaman-tane C H , consists of only one diamond cage, fol-lowed by diamantane C H with two, and triaman-tane C H with three faced-fused cages. Diamondoidscan be found in some natural gas reservoirs, and espe-cially diamantane is one of the deposits in gas pipelines [email protected] Corresponding author (Reiser et al. 1996). Lately, diamondoids consisting ofup to 11 diamond cages were isolated from petroleumby Dahl et al. (2003) making these species accessible tolaboratory investigations.Oomens et al. (2006) measured the infrared spectro-scopic properties of powders of higher diamondoids (upto hexamantane) at room temperature. In accordancewith calculations applying density functional theory(DFT), the by far strongest features in the IR spec-tra of neutral diamondoids were found to be the C-Hstretching bands between 3.4 and 3.6 µ m arising fromthe hydrogen-terminated surfaces. Pirali et al. (2007)additionally measured the IR emission spectra of hot(500 K) gas phase adamantane, diamantane, and tria-mantane in the wavelength region of the C-H stretch-ing modes revealing a small redshift of the bands in thesolid-state spectra. Based on the IR spectra obtained byOomens et al. (2006) along with additional DFT calcula-tions, Pirali et al. (2007) also made assignments for twoIR features of two different classes of astronomical ob-jects. The first one is the unusual IR emission at 3.43 and3.53 µ m originating from two objects whose spectra areelsewhere dominated by the well-known infrared emis-sion bands of polycyclic aromatic hydrocarbons (PAHs),Elias 1 and the inner region of the circumstellar diskaround HD 97048 (Habart et al. 2004). Although these Steglich et al.bands were already assigned to nanodiamonds of at least50 nm diameter (Guillois et al. 1999), it was shown thatalso tetrahedral diamondoid molecules containing around130 C atoms (close to the size of the smallest meteoriticnanodiamonds) exhibit the 3.43 and 3.53 µ m bands withthe proper intensity ratio (Pirali et al. 2007). The secondfeature mentioned is the broad (FWHM 0.09 µ m) absorp-tion band centered at 3.47 µ m which is observed in theabsorption spectra of various dense clouds in lines of sighttoward young stellar objects (Allamandola et al. 1992,1993). By co-adding all diamondoid solid-state spec-tra obtained by Oomens et al. (2006), Pirali et al. (2007)showed that the different C-H stretching modes mergeinto a broad structure centered around 3.47 µ m, and theyargued that small diamond-like molecules may there-fore be a major contributor to the interstellar absorptionband which is only observed in or behind dense molec-ular clouds, but not in the diffuse interstellar medium.As the intensity of the interstellar band could be corre-lated with the intensity of the 3.08 µ m band of waterice, it was speculated that its carriers may be formedon icy grains in the shielded environment of molecularclouds (Brooke et al. 1996) which is furthermore sup-ported by recent laboratory experiments where nanodi-amond crystallites were created upon UV irradiation ofinterstellar ice analogs (Kouchi et al. 2005). Based onthe computed intensities of the diamondoid C-H stretch-ing bands, Bauschlicher et al. (2007) deduced that only1 −
3% of the cosmic C has to be locked in diamondoids inorder to explain the observed intensity of the interstellarband. Furthermore, they calculated ionization potentials(IPs) of neutral diamondoids, as well as IR spectra andelectronic transition energies of neutral and cationic dia-mondoids, concluding that cations may also contribute tothe 3.47 µ m absorption band. However, the bands of thecations are somewhat weaker than those of their neutralcounterparts. Furthermore, the cations feature furtherbands with comparable strengths at longer wavelengths(6 − µ m), e.g. due to C-H bending vibrations, whichcould be used in the future to trace ionized diamondoids.Compared to the strong IR emission of PAHs triggeredby the absorption of UV-vis photons, the IR emission ofneutral diamondoids is rather inefficient because theirelectronic absorption onset lies far in the UV between 6and 7 eV (Landt et al. 2009a,b). Calculations imply thatdiamondoid cations would feature absorption bands inthe visible and near-IR due to their open-shell structure(Bauschlicher et al. 2007). However, these transitionsare very weak and, as for the neutrals, efficient IR emis-sion can only be expected in regions of space experiencinghigh-energy and high-flux UV radiation fields, usually inclose proximity of the exciting stars (Bauschlicher et al.2007). Therefore the 3.43 and 3.53 µ m emission featuresare so rarely observed, to date only in HD 97048 andElias 1.The IP of diamondoids (8 − α emission. Hence,ionization of neutral diamondoids in strongly irradiatedregions of space (HD 97048 and Elias 1) may be veryefficient. Consequently, one should address the follow-ing questions: are these cationic species stable and, if so, what are their spectroscopic fingerprints? As alreadystated, one can expect weak absorption bands in thevisible and near-UV wavelength range for the cationsdue to low-energy transitions to semi-occupied molec-ular orbitals, which might be detectable by electronicspectroscopy methods. Unlike in PAHs, there is no de-localized electron cloud. Removing one electron due tophotoionization weakens the bonds between the atoms,and fragmentation (C − H bond breaking) may occur. In-deed, such behavior was observed for the three smallestdiamondoids. By comparing the IR absorption bands ofpositively charged, gas phase adamantane (Polfer et al.2004), diamantane, and triamantane (Pirali et al. 2010)with theoretical spectra of dehydrogenated diamondoidcations, it was shown that these molecules easily lose ahydrogen atom upon ionization to form stable closed-shell species. The loss preferentially happens on a ter-tiary carbon (CH group) rather than on a secondary car-bon (CH group). Since in these experiments, the ion-ization was performed via charge transfer using cationicagents with high IPs, it is not obvious whether this de-hydrogenation will also occur under astrophysical irradi-ation conditions.In this work, we investigated the electronic transitionsof the four smallest diamondoids, namely adamantaneC H , diamantane C H , triamantane C H , andtetramantane C H (three isomers), and their photo-products. For this purpose, we used matrix isolationspectroscopy (MIS). Cationic species were formed via UVirradiation using a hydrogen-flow discharge lamp to sim-ulate the interstellar UV photon field. Our experimentalfindings are supported by theoretical calculations apply-ing DFT and time-dependent DFT (TD-DFT). These re-sults will help astronomers to search for the spectroscopicfingerprints of diamond-like molecular species. Further-more, the UV absorption cross sections of neutral andcationic diamondoids can be used to accurately predictIR emission processes caused by stochastic heating dueto the absorption of UV photons in strongly irradiatedregions of space. METHODS
Theoretical Calculations of Electronic Spectra
For the purpose of identification and comparison, weperformed DFT calculations for differently charged andde-hydrogenated diamondoids. The molecular struc-tures were first optimized with the Gaussian09 soft-ware (Frisch et al. 2009) using the B3LYP functional(Stephens et al. 1994; Becke 1993) in conjunction withthe 6-311++G(2d,p) basis set for adamantane and itsderivatives, and the 6-311+G(d) basis set for the largerdiamondoids, respectively. We chose the smaller ba-sis set for the larger molecules to reduce the com-putational effort since no remarkable differences be-tween 6-311++G(2d,p) and 6-311+G(d) were obtainedfor the ground-state structure and electronic spectrumof adamantane. The vibrational modes were calculatedafterward to check whether the optimized structures werereally at their respective minima of the potential en-ergy surfaces and to determine the zero-point correctionsfor the ground-state energies. The dipole moments pre-sented later have been corrected to correspond to thecenter-of-mass coordinate system. The values obtainedhotoproducts of Diamondoids 3by the Gaussian09 software refer to molecular orienta-tions with the center of nuclear charge as origin. For theinvestigated species, both values differ by less than 5%.Two different approaches were used to evaluate the ener-gies of the excited electronic states. The first one involvesthe TD-DFT implementation of Gaussian09, using thesame functional and basis set as for the ground-state cal-culations. The computational effort scales steeply withthe size of the system under consideration and the num-ber of excited states to be calculated. This TD-DFT ap-proach was used to accurately predict the first few elec-tronic transitions of the diamondoids and their deriva-tives. An error of about 0.3 eV regarding the energeticpositions of the transitions can commonly be expectedat this level of theory. It should be mentioned thatpurely vertical electronic transitions are calculated. Vi-brational excitations in excited electronic states cannotbe accounted for. However, for higher energies the life-times of the excited states are expected to decrease sub-stantially resulting in broad bands without discerniblevibrational pattern even for cold gas phase molecules.For the high-energy states, the TD-DFT formal-ism as implemented in the Octopus software package(Castro et al. 2006; Marques & Gross 2004) was applied.Unlike the frequency-space implementation of TD-DFTin Gaussian09, Octopus uses real-space numerical gridsto propagate the Kohn Sham orbitals in real time. Fol-lowed by an initial very short electric pulse, exciting allfrequencies of the system, the time-dependent dipole mo-ment is calculated from which the linear optical absorp-tion spectrum can be derived. Relying on numericalmeshes, the code works without basis sets. However,we used again the B3LYP functional. The volume ofthe box in which the desired molecule is represented waschosen such that each atom was at least 4 ˚A away fromthe edges. In all calculations, the grid spacing was 0.2˚A, the time integration length was 10 ~ eV − , and thetime step was 0.002 ~ eV − . This approach does notyield any information on the symmetry of the excitedstates. The widths of the absorption bands are purely ar-tificial and depend solely on the integration length usedin the calculation. However, the area of each band isdirectly related to the oscillator strength of the corre-sponding transition and can easily be converted into ab-sorption cross section values. Optical spectra derivedfrom the Octopus code already gave reasonable agree-ment between calculated and measured absorption spec-tra and cross sections of PAHs in the energy range abovethe IP (10 −
30 eV) where electronic transitions involving σ electrons dominate (Malloci et al. 2004). It should bestressed that only bound-bound transitions are predictedby this approach, whereas transitions leading to directionization cannot be accounted for. However, in view ofthe good agreement between experiment and theory forPAHs, it was argued that superexcited states account formost of the absorption in the vacuum ultraviolet (VUV)and that they are coupled to the ionization continuum(Malloci et al. 2004). Whether the same situation alsoapplies for the σ − σ ∗ transitions of diamondoids andtheir ionic derivatives can only be clarified when dedi-cated laboratory measurements are available. Matrix Isolation Spectroscopy and FUV irradiation
To measure low-temperature spectra of the diamon-doids and their photoproducts, we used a setup for MIS.A transparent CaF window mounted inside of a vac-uum chamber at the bottom of an expander was cooleddown to temperatures < window by 90 ◦ , transmis-sion spectroscopy down to 190 nm was performed with aspectrophotometer (JASCO V-670 EX), calibrated withan accuracy of 0.3 nm. The chosen resolution was typi-cally 0.5 nm, which is much smaller than the widths ofthe measured absorption bands. Prior to the incorpo-ration of the diamondoids into the matrix, they wereevaporated in an oven kept at suitable temperatures.Due to the high vapor pressures of the smallest diamon-doids, the oven had to be cooled down to temperaturesas low as 0 ◦ C in the case of adamantane and diaman-tane. For adamantane, a special polyvinylidene fluoridefilter further decreased the gas flow to the CaF win-dow. Triamantane was heated up to 20 ◦ C − ◦ C andthe tetramantanes to 90 ◦ C. The deposition rates of theneutral diamondoid precursors and the matrix material,both transparent in the investigated wavelength range,were determined in separate experiments by depositionof the pure substance on the low-temperature windowand measuring the transmission of the as-prepared filmswith the spectrometer. A typical interference patterncorresponding to the transmission of an etalon could beobserved for sufficiently thin films from which the thick-nesses and, therefore, the column densities could be de-rived (see Figure 1).After preparation of the transparent matrix dopedwith the neutral precursor molecules, an absorptionspectrum, usually displaying no discrete features, wasmeasured and used as a new baseline. The sam-ples were then photolyzed with the far-UV emissionfrom a microwave-driven hydrogen-flow discharge lamp(Warneck 1962). The spectra recorded afterward fea-ture absorption bands from species created during theirradiation. For comparison, we also irradiated a cleanNe matrix and films of pure diamondoids which weredeposited on the 7 K cold CaF window without simul-taneous Ne flow. Besides trace amounts of dissociatedwater in the case of the inert gas, we did not observe anyfeatures. This confirms that the measured bands are in-deed due to photoproducts of isolated diamondoids. Thehydrogen lamp itself was operated with a gas mixtureof 10% H and 90% Ar, usually at a pressure of 0.6mbar. The purpose of the inert gas is to suppress the160 nm molecular emission and enhance the 121.6 nm(10.2 eV) Ly α emission, but it also introduces additionallines from the Ar at 106.7 nm (11.6 eV) and 104.8 nm(11.8 eV). The lamp was separated from the vacuum ofthe MIS chamber by a LiF window. Its total photon fluxof 10 − photons s − was determined by measuringthe photocurrent emanating from a Pt plate under FUVlight exposure. During the experiments, the intensity onthe sample surface was typically 10 − photons m − All experimental spectra presented here were obtained by usingthe spectra recorded before irradiation as a baseline.
Steglich et al. pure Ne film t r an s m i ss i on [ % ] wavenumber [(cid:181)m -1 ]pure adamantane film Fig. 1.—
Transmission spectra of pure Ne (left) and adaman-tane (right) on a 6.8 K cold CaF window used for the determina-tion of the deposition rates and hence the sample to matrix ratio.The thickness d of the deposited film follows from the wavenum-ber difference ∆ k between two consecutive transmission maximavia ∆ k = (2 nd ) − , where n is the film’s refractive index. Note theincreased scattering toward shorter wavelengths. The transmissionis actually higher than 100% because it is compared to the trans-mission of the clean CaF window which has a higher reflectancethan the deposited films. s − . Usually, the irradiation was performed for 15 − RESULTS AND DISCUSSION
Adamantane
In Figure 2, the structure and calculated ground-state energy of adamantane, as well as the energies ofthe adamantane and adamantyl cations are displayed.Adamantane possesses two structurally different H sitesand therefore two different adamantyl isomers. The num-bers on the adamantane structure shown in Figure 2 indi-cate the C atoms from which the H atoms are removed toform the corresponding adamantyl structures. Accordingto the calculations, only minor distortions of the carbonframework are expected upon H removal. This also ap-plies to the larger species we have investigated. For theground states of the adamantyl cations, different spinstates are principally possible. However, the triplet stateof the 1-adamantyl cation is about 3 eV higher in energythan the singlet state. Because of this rather high energydifference, we expect all other dehydrogenated diamon-doid cations to be closed-shell singlet species as well andthe calculations were performed accordingly. For com-pleteness, we also calculated the structures and ground-state energies of the neutral 1- and 2-adamantyl radicals.In both cases, an energy of 4.2 eV would be necessary toremove one H atom. However, direct photodissociationis unlikely since the absorption onset is further in theUV. Basically, the photons delivered by the hydrogenlamp provide enough energy to cause photoionization,as well as the removal of one H atom from the ionizedmolecule. Further dissociation cannot be accomplishedwith a single photon. Considering the low FUV doses ap-plied during the experiments, comprising typically 15 − Fig. 2.—
Zero-point-corrected ground-state energiesof adamantane and its derivatives calculated at theB3LYP/6 − The calculated spectra of neutral andcationic adamantane and adamantyl, using theB3LYP/6 − − ). The chosen bandwidth is rather arbitrary,but the so-computed spectrum indicates in which energyrange strong electronic transitions of the real moleculecan be expected.The measured spectrum of FUV-processed adaman-tane in solid Ne (6.8 K) can be found in the bottom panelof Figure 3. The ratio of Ne to adamantane in termsof absolute numbers of atoms or molecules was deter-mined as described in Section 2.2. The densities of solidNe and adamantane were taken to be 45 atoms nm − (Timms et al. 1996) and 1.2 g cm − (Yashonath & Rao1986), respectively. Using these values, the isolation ra-tio (Ne to adamantane) was estimated to be better than190 n Ada n − . The factor n Ada n − is the ratio of therefractive indices of the pure solid materials in the visi-ble and should be between 1 and 2. Nevertheless, we alsoperformed measurements at lower isolation ratios and didnot notice appreciable spectral differences. Since it is un-known what fraction of the neutral precursor moleculesis actually transformed into cations we cannot provide Usually, the conversion rate is lower than 10% under theseexperimental conditions. hotoproducts of Diamondoids 5
200 300 400 500 600 700 8000.00.10.20.30.40.50.60.70.00.10.20.30.00.10.20.30.4 8 7 6 5 4 3 2 sum of calculated spectra 50 % A1 + 50 % A2 x2 photoproducts of adamantane in Ne as measured background-corrected wavelength [nm] +0.1 ab s o r p t i on c r o ss s e c t i on [ ¯ ] ab s o r ban c e adamantane neutral adamantane cation energy [eV] Fig. 3.—
Calculated (B3LYP/6 − − to indicate therange where strong electronic transitions can be expected. experimental values for the absorption cross section.As indicated in Figure 3, an additional baseline correc-tion was applied to remove the strong scattering back-ground in the UV. Because of its wavelength dependence( ∼ λ − ), we attribute this background mainly to an in-creased Rayleigh scattering of the charged species com-pared to their neutral counterparts. The derived absorp-tion spectrum consists of four broad features above 200nm. The broad band between 280 and 350 nm with max-imum around 308 nm is not an artifact of the measure-ment. It has been confirmed by repeatedly performedexperiments. The same applies to the two somewhat nar-rower features at 252 and 261 nm. The strongest band in the accessible wavelength region extends from 200 to 240nm and peaks at 223.5 nm. Considering the theoreticalresults, we exclude the presence of open-shell adaman-tane cations in the photo-processed matrix due to theabsence of bands in the visible. Instead, the observedspectrum points toward the creation of the more stable,closed-shell, singly de-hydrogenated cation. Therefore,we assign the strongest band at 223.5 nm to the S → S transition of the 1-adamantyl cation (point group C )which is the isomer with lower ground-state energy. Thecalculated oscillator strength of this band is f = 0 . → S transition of the 1-adamantylcation ( f = 0 . s ) with calculated oscillator strengths below f = 0 . − H + hν → C H +15 + e − + H . (1)This process has been described for a few smallermolecules, e.g., H O (Cairns et al. 1971) or CH (Samson et al. 1989). After ionization, some excess en-ergy is stored in the vibrational degrees of freedom ofthe cationic molecule which subsequently leads to thedestruction of one terminal C − H bond. During this pro-cess, electrons and neutral H atoms are released whichare usually trapped on defects or impurities in the ma-trix. Because of recombination reactions between posi-tively charged molecules and released electrons (and Hatoms), the ion yield saturates when a certain irradia-tion dose is reached. The formation of negatively chargedadamantane or neutral adamantyl molecules due to elec-tron attachment can be ruled out for the following rea-sons. First, these species possess an open-shell electronicstructure and, like the adamantane cation, would fea-ture absorption bands in the visible part of the spectrum.And second, as far as neutral adamantane is concerned, Steglich et al.the negative electron affinity (Drummond 2007) hampersfurther electron attachment.Notably, we observed neither sharp absorption bandsnor a clear vibrational pattern. We want to remark thatthe measured bands are much broader than what is ex-pected for typical matrix-induced broadening (at least for7 K neon matrices) which is mainly due to site effects. We tentatively attribute this to an intrinsic property ofthe molecule, i.e., a very short lifetime of the excitedstate which is not entirely caused by the interaction withthe rare gas atoms. This would lead to the conclusionthat the main difference to astrophysically more relevantspectra of cold gas phase adamantyl cations is a smallmatrix-induced redshift, but not a broadening of the ab-sorption bands. By coincidence, the spectral shape is insurprisingly good agreement with the computed spectraof the purely electronic (vertical) transitions. Therefore,the absolute values of the absorption cross sections, asthey appear in the calculated spectra, may be regardedas representative for real gas phase molecules. Finally, we want to highlight another important prop-erty of the adamantyl cations. Unlike their neutral pre-cursor adamantane, these species possess rather strongpermanent dipole moments pointing from the molecularcenter toward the C atom from which the H atom hasbeen removed. The calculated dipole moments of the 1-and 2-adamantyl cations amount to 0.96 and 2.57 Debye,even stronger than the dipole moment of the open-shellcation (0.61 Debye). This opens the possibility to de-tect and identify molecular diamond-like species in spaceusing their rotational spectra.
Diamantane
The calculated ground-state energies of diamantaneand its singly dehydrogenated cationic derivatives aredisplayed in Figure 4. Again, the numbers on the dia-mantane structure shown indicate the positions fromwhere the hydrogen atoms are removed to form the cor-responding closed-shell cations. There are three possibleisomers for the diamantyl cation, possessing quite strongpermanent dipole moments of 1.76 (D1), 3.09 (D4), and3.71 Debye (D3) due to the localized charge at the edgeof the molecule.The electronic spectra of diamantane and its relatedspecies are presented in Figure 5. An additional back-ground correction has been applied on the red side of themeasured spectrum to remove non-reproducible bumpsand fringes due to baseline variations. Using 1.2 g cm − (Karle & Karle 1965) as mass density of the solid dia-mantane deposit, the isolation ratio (Ne to diamantane)varied in different experiments between 450 n Dia n − and 750 n Dia n − , where n Dia is the refractive index ofthe diamantane film. Compared to the photoprocessedadamantane, the spectrum of irradiated diamantane re-veals a slightly broader peak (7800 cm − versus 5900cm − ), positioned further to the red at 255 nm. As isevident upon inspection of the calculated and measuredspectra, this feature cannot be explained by the presence Molecules in different sites of the matrix exhibit different red-shifts of their absorption bands, effectively resulting in a broaden-ing. Otherwise, only the integrated cross section or the oscillatorstrength could be taken.
Fig. 4.—
Zero-point-corrected ground-state energies of diaman-tane and its derivatives calculated at the B3LYP/6 − of open-shell cation radicals. Regarding the formation ofnegatively charged diamantane or neutral diamantyl rad-icals, the same reasoning applies as in Section 3.1. Princi-pally, the photons from the hydrogen lamp carry enoughenergy to induce dissociative photoionization and cre-ate all three diamantyl isomers. However, the conclusionmay be drawn that the main photoproduct in the matrixexperiment is the 4-diamantyl cation (point group C ).Its first strong transitions S → S , are predicted at 284nm ( f = 0 . f = 0 . s ) for which several close-lying ab-sorptions at 289 nm (S → S , f = 0 . → S , f = 0 . → S , f = 0 . ) in the matrix, i.e. the H removal from aCH group, can rather be excluded. Because of the ap-parent absence of fine structure, it seems difficult to drawfurther conclusions. Triamantane
Figure 6 displays the calculated ground-state energiesof triamantane and its seven isomers of singly dehy-drogenated cations. The necessary energies to removeone electron and one H atom from the parent moleculeare comparable to the previously discussed diamondoids.Basically, the FUV lamp delivers photons with energieshigh enough to create all seven isomers. Their dipolemoments, again quite strong, range between 1.13 and5.72 Debye. The dipole moment of the open-shell cationamounts to 0.51 Debye.The corresponding calculated electronic spectra, aswell as the measured spectrum of FUV-irradiated tria-mantane in Ne are displayed in Figure 7. The isola-tion ratio in the matrix experiment was on the orderhotoproducts of Diamondoids 7
200 300 400 500 600 700 8000.000.050.100.150.200.00.10.20.30.00.10.20.30.40.5 8 7 6 5 4 3 2 photoproducts of diamantane in Ne as measured background-corrected wavelength [nm] +0.15+0.07 ab s o r p t i on c r o ss s e c t i on [ ¯ ] diamantane neutral diamantane cation ab s o r ban c e energy [eV] Fig. 5.—
Calculated (B3LYP/6 − of 500 − n Tria n − with unknown refractive indexof the pure triamantane ( n Tria ) film. The photodisso-ciation of trace amounts of water in the matrix is re-sponsible for the narrow bands from the OH radical at308 and 283 nm (Tinti 1968). Compared to the pre-vious measurements of irradiated adamantane and dia-mantane, the lowest-energy spectral feature shifts furtherto the red, extending roughly from 300 to 500 nm. Itpeaks at 368 nm and has a shoulder around 450 nm. Astrong FUV rise also slides into the accessible wavelengthregion. An assignment to a certain isomer of the tria-mantyl cation is rather difficult. Oddly, the best matchseems to be possible with the calculated spectrum of the5-triamantyl cation which is 10.56 eV higher in energythan the triamantane neutral. (In the previously dis-cussed measurements, the strongest bands seemed to becaused by species which were 10.35 eV (adamantane) and10.27 eV (diamantane) away from the parent molecule.)Besides dehydrogenated triamantyl cations, an alterna-tive explanation for the measured broad band would bethe formation of the open-shell triamantane cation as canbe seen from the comparison with the corresponding cal-culated spectrum. Possibly, triamantane is already bigenough, and the energy stored in the molecule upon ab-sorption of an FUV photon is distributed over sufficient
Fig. 6.—
Zero-point-corrected ground-state energies of triaman-tane and its derivatives calculated at the B3LYP/6 − vibrational modes to avoid H abstraction. Nevertheless,a clear identification of the created species on the ba-sis of electronic absorption spectroscopy is not possibleand one should take into account the possibility that theapplied quantum chemical model deviates more stronglyfrom reality than expected. Tetramantane
In contrast to the smaller diamondoids, there are al-ready three different isomers of neutral tetramantaneC H , while one of them has actually two enantiomers(P and M [123]-tetramantane). Their structures, thecalculated spectra of their cations, as well as the mea-sured spectra of their photoproducts, are displayed inFigure 8. We did not calculate the structures and spec-tra of the tetramantyl cations because of the increas-ing number of isomers and therefore escalating compu-tational effort. Regarding the necessary energies for Habstraction, we do not expect large deviations from thesmaller diamondoids. With almost no difference amongthe three species, the measured spectra are very simi-lar to what has been measured for triamantane. Besides[121]-tetramantane, the broad feature extending roughlyfrom 300 to 500 nm could be assigned to the open-shellcations as is obvious upon comparison with the calcula-tions. The [121]-tetramantane cation, however, shouldhave stronger bands at longer wavelengths suggestingthat the measured spectrum is actually caused by thecorresponding singly dehydrogenated cation. Whetherthe other two tetramantanes ([123] and [1(2)3]) lost a On the other hand, it should be taken into account that therise in the baseline beyond 510 nm could indicate a very broad and,compared to the calculated spectrum, rather weak band around 570nm.
Steglich et al.
200 300 400 500 600 700 8000.000.050.100.00.10.20.30.00.10.20.30.00.10.20.30.40.5 8 7 6 5 4 3 2 x7OH photoproducts of triamantane in Ne as measured background-corrected wavelength [nm] +0.1+0.05 ab s o r p t i on c r o ss s e c t i on [ ¯ ] +0.15+0.1+0.05 ab s o r ban c e triamantane neutral triamantane cation energy [eV] Fig. 7.—
Calculated (B3LYP/6 − peripheral H atom upon ionization or not, cannot com-pletely be clarified, as there are too many tetramantylisomers and, as is the case with triamantane, a compari-son with TD-DFT theory would not provide unambigu-ous insights. Complete σ − σ ∗ Absorption Spectra
The electronic spectra of the neutral and cationic di-amondoids resulting from (all possible) bound-bound( σ − σ ∗ ) transitions, as calculated with the Octopuscode, are displayed in Figure 9. These spectra may beused to model photophysical interactions of diamond-like molecules in the interstellar medium. For a briefdiscussion about their physical relevance refer to Sec-tion 2.1. Regarding the energy range below 8.5 eV,more detailed, measured gas phase absorption spectra
200 300 400 500 600 700 8000.000.040.080.120.00.10.20.30.40.5 8 7 6 5 4 3 2
FUV irradiated [1(2)3]-tetramantane FUV irradiated [123]-tetramantane FUV irradiated [121]-tetramantane +0.01+0.02OH x5x5x5 ab s o r ban c e wavelength [nm] +0.2+0.1 [1(2)3]-tetramantane cation [123]-tetramantane cation [121]-tetramantane cation ab s o r p t i on c r o ss s e c t i on [ ¯ ] energy [eV] Fig. 8.—
Isomers of tetramantane, calculated spectra of theircations, and measured spectra of their photoproducts isolated insolid Ne (6.8 K). of neutral diamondoids can be found in the publicationof Landt et al. (2009b). The vibrational structure thatcan be seen in these spectra can hardly be predicted withcurrent theoretical methods. The spectra presented herefeature broader bands that are purely artificial. An ad-ditional energy-dependent broadening by convolving thespectra with Lorentzians of increasing bandwidths hasbeen applied to account for an increased lifetime broad-ening which is expected at higher energies. Due to thehigh density of states above ∼
10 eV, changing the band-widths does not substantially alter the absolute cross sec-tion values. While the positions and shapes of resonancesappearing in the spectra may be affected by uncertain-ties of the computational method the general trend ofthe absorption curves and the cross section values maybe regarded as real (at least within the limitations dis-cussed before). Comparing the spectra of the neutraland ionized molecules with each other, it is obvious thatthere are not many differences, especially for the transi-hotoproducts of Diamondoids 9
200 150 100 50 200 150 100 50200 150 100 50 200 150 100 50 +2+1 A2A1 ab s o r p t i on c r o ss s e c t i on [ ¯ ] adamantane neutral adamantyl cations D3D4D1+3+2+1 diamantane neutral diamantyl cations ab s o r p t i on c r o ss s e c t i on [ ¯ ] +2+1 triamantane neutral triamantane cation 5-triamantyl cation ab s o r p t i on c r o ss s e c t i on [ ¯ ] energy [eV] adamantane neutral diamantane neutral triamantane neutral Allende nanodiamonds +0.8+0.5+0.2 m a ss ab s o r p t i on c oe ff i c i en t [ c m g - ] wavelength [nm] wavelength [nm]wavelength [nm] wavelength [nm] Fig. 9.—
Complete calculated electronic σ − σ ∗ absorption spectra of neutral and ionized small diamondoids. For comparison, the bottomright panel contains the IR to VUV spectrum of meteoritic nanodiamonds from the Allende meteorite ( ∼ tions at higher energies, as the electronic structure of theC skeleton is equivalent. For all species, the high-energyabsorption is dominated by a broad hump with a maxi-mum between 15 and 20 eV. Also other features on thered and blue tails of this σ − σ ∗ hump, like peaks at 11,14, and 28.5 eV for the cations or 9, 12.5, and 27.5 eVfor the neutrals are very much comparable. As alreadydiscussed in the previous sections, the absorption onsetof the cations appears further to the red compared totheir neutral precursors which is also evident from Fig-ure 9. By increasing the molecular size, two effects areobvious: rising values for the absolute absorption crosssection and a trend toward a less-structured absorptioncurve due to an increased density of states.We want to point out an interesting aspect of theseresults. Even though laboratory experiments in thediscussed energy range for these molecular species are lacking, experimental data on nanodiamonds, extractedand isolated from the Allende meteorite (Mutschke et al.2004), display surprising resemblances (see Figure 9).These nanodiamonds possess an average size of less than2 nm, corresponding to ≈
500 C atoms, which is muchbigger than the molecular diamonds presented here, thelargest of which contains 22 C atoms. Their electronicabsorption spectra solely consist of the broad σ − σ ∗ bandwith maximum at 17.1 eV. Furthermore, a shoulder canbe seen around 30 eV which may have its equivalent ina band at 28.5 eV in the diamondoid spectra. The peakmass absorption coefficient κ for the meteoritic nanodia-monds was found to be 1 . × cm g − , very closeto the calculated values of κ = 1 . − . × cm g − for the molecular diamond. The main differences of themolecular compared to the nanoscopic material are theredshifted absorption onset and a more structured ab-0 Steglich et al.sorption curve, especially on the red wing of the collec-tive σ − σ ∗ hump. SUMMARY
We have investigated the electronic absorption prop-erties of the smallest diamondoids, a possible molec-ular part of the interstellar carbonaceous dust withdiamond-like structure. For adamantane and diaman-tane, our results confirm the formation of closed-shellsingly dehydrogenated cations upon FUV irradiationwhich, similarly, was already found in previous studiesby Polfer et al. (2004) and Pirali et al. (2010) in the in-frared regime using an indirect ionization method. Fur-ther ionization of the adamantyl and diamantyl cations,even in strongly irradiated regions of space, may be ham-pered by the rather large second IP. For instance, DFTcalculations (B3LYP / 6-311++G(2d,p)) imply a 14.1 eVenergy difference between the cation and dication of 1-adamantyl. A clear identification of the created isomersin the matrix experiments, i.e., information about whichH atom was removed from the edge of the molecule,through comparison with TD-DFT calculations is ratheruncertain. Starting with triamantane, our results leaveopen the possibility that larger diamondoids are just ion-ized upon FUV irradiation, and that the absorbed pho-ton energy is distributed over sufficient vibrational de-grees of freedom to avoid H abstraction. Further spec-troscopic investigations of FUV-processed diamondoidsin the IR range would help to clarify this issue as molec-ular vibrations can be more accurately predicted by cur-rent theoretical models.In our experiments, the dissociation of adamantaneand diamantane was initiated via photoionization. Be-sides recombination reactions due to the close proximityof the molecules in the matrix, we do not expect thatthe interaction with the matrix atoms has a strong influ-ence on the dissociation process itself and suggest that,if present as gas phase molecules, small diamondoids ininterstellar space are subjected to H abstraction uponFUV irradiation. The measured spectra of the createdions isolated in solid Ne matrices display broad bands inthe UV, shifting to longer wavelengths with increasingmolecular size. The widths of these bands were found tobe much larger than expected from typical broadeningeffects caused by the interaction with the rare gas atomsof the matrix. Therefore, we attribute this to an intrinsiceffect of the molecules, i.e., a very short lifetime of theexcited state which is not entirely caused by the inter-action with the Ne matrix. This implies that, besides a small matrix-induced redshift, UV spectra of cationic di-amondoids in the gas phase would feature bands similarin shape and width as in the matrix spectra. Unfortu-nately, the lack of narrow bands also hampers a possibledetection in space via UV observations. The photopro-cessed adamantane displays a broad absorption band at223.5 nm which may be found at slightly shorter wave-lengths in the gas phase. However, a possible contribu-tion to the interstellar 217.5 nm UV bump can be ratherexcluded due to its weakness ( f ≈ .
09) and several fur-ther absorption bands in close proximity at shorter wave-lengths which are incompatible with the interstellar ex-tinction curve. However, the specific structures of thedehydrogenated cations open other possibilities for de-tection. Unlike their neutral precursor molecules, suchspecies possess strong permanent dipole moments due tothe missing H atom and localized charge at the molecularperiphery, permitting an identification by means of rota-tional spectroscopy. Potential targets for radio-based ob-servations could be (the edges of) dense molecular cloudswhere small diamondoids can be expected, if the assign-ment of the 3.47 µ m absorption band is valid, or the closeproximity of objects with intense UV radiation fields,such as HD 97048 and Elias 1, where the 3.43 and 3.53 µ m emission features were observed.The main UV absorption feature at higher energies( >
10 eV) of diamond-like material is the collective σ − σ ∗ peak. Also the diamondoids and their cationic deriva-tives exhibit this feature and, interestingly, its position(18 eV) and cross section (max. 1 . − . × cm g − )are very similar to the extinction hump measured for themuch larger meteoritic nanodiamonds (Mutschke et al.2004). However, the molecular material additionallydisplays certain narrow features on the red tail of the σ − σ ∗ hump (see also Landt et al. 2009b). These resultsmay prove useful for the modeling of diamondoid IRemission caused by stochastic heating from UV photonsand, thus, they can contribute to the discussion aboutthe origin of the IR emission features of HD 97048 andElias 1.This work was supported by the Deutsche Forschungs-gemeinschaft (DFG) and (in part) by the Departmentof Energy, Office of Basic Energy Sciences, Division ofMaterials Sciences and Engineering, under contract DE-AC02-76SF00515. M.S. thanks Dr. Harald Mutschke forproviding the nanodiamond spectrum and for measuringthe UV flux of the H discharge lamp together withKamel A. Khalil Gadallah. REFERENCESAllamandola, L. J., Sandford, S. A., Tielens, A. G. G. M., &Herbst, T. M. 1992, ApJ, 399, 134Allamandola, L. J., Sandford, S. A., Tielens, A. G. G. M., &Herbst, T. M. 1993, Science, 260, 64Anders, E., & Zinner, E. 1993, Meteoritics, 28, 490Bauschlicher, C. W., Jr., Liu, Y., Ricca, A., Mattioda, A. L., &Allamandola, l. J. 2007, ApJ, 671, 458Becke, A. D. J. 1993, J. Chem. Phys., 98, 1372Brooke, T. Y., Sellgren, K., & Smith, R. G. 1996, ApJ, 459, 209Cairns, R. B., Harrison, H., & Schoen, R. I. 1971, J. Chem. Phys.,55, 4886Castro, A., et al. 2006, Phys. Stat. Solidi b, 243, 2465Dahl, J. E., Liu, S. G., & Carlson, R. M. K. 2003, Science, 299, 96 Drummond, N. D. 2007, Nature Nanotechnol., 2, 462Frisch, M. J., et al. 2009, Gaussian 09 (Revision A.02;Wallingford, CT: Gaussian Inc.)Guillois, O., Ledoux, G., & Reynaud, C. 1999, ApJ, 521, L133Habart, E., Testi, L., Natta, A., & Carbillet, M. 2004, ApJ, 614,L129Henning, Th. & Salama, F. 1998, Science, 282, 2204Jones, A. P., d’Hendecourt, L. B., Sheu, S.-Y., Chang, H.-C.,Cheng, C.-L., & Hill, H. G. M. 2004, A&A, 416, 235Karle, I. L., & Karle, J. 1965, J. Am. Chem. Soc., 87, 918Kouchi, A., Nakano, H., Kimura, Y., & Kaito, C. 2005, ApJ, 626,L129 hotoproducts of Diamondoids 11hotoproducts of Diamondoids 11