Identifying specific interstellar PAHs

1

• Identifying specific interstellar PAHs

Giacomo Mulas1 Giuliano Malloci1 Ignazio Porceddu1
1INAF Osservatorio Astronomico di Cagliari
2
PAHs who are they?

• A Very Small Grain component of transiently
  heated carbonaceous particles was hypothesized
  since the first IRAS results, which showed excess
  emission from the diffuse ISM in the 12µm band.
• If thermal, this emission implied temperatures
  too high for normal dust particles in thermal
  equilibrium.
• On the other hand, a VSG (less than a few nm
  across) particle, due to its very small heat
  capacity, might be heated to very high
  temperatures upon absorption of a single UV
  photon.
• When more detailed spectra became available, this
  excess mid-IR emission appeared strikingly
  similar to the spectra of coals, or kerogen. Both
  the groups of Salama and Allamandola and of
  Leger, Puget and dHendecourt independently
  proposed large, free flying Polycyclic Aromatic
  Hydrocarbon molecules as VSGs

3
UIBs and PAHs
4
PAHs as the panacea?

• PAHs, provided they are large enough (NC30),
  ought to be extremely stable against
  photodissociation, and have been invoked to
• account for the far UV rise of the extinction
  curve
• contribute to the UV extinction bump at 220nm
• account for UIBs
• account for most DIBs
• scavenge and/or provide electrons in the diffuse
  ISM
• help traditional dust grains in promoting H2
  formation
• produce the Extended Red Emission
• produce blue fluorescence in the Red Rectangle...

5
... but despite all this
No single, specific PAH molecule, in any form
whatsoever, has been firmly identified in the ISM
yet, despite very intensive search!
6
Fingerprinting a PAHthe difficult...

• PAHs in the ISM are expected to be mostly ionised
  and in a distribution of (de)hydrogenation states
• PAH cations and radicals show strong electronic
  transitions in the optical they must show up
  among DIBs if present
• but
• matrix isolation techniques are not accurate
  enough for an identification, due to DIB crowding
• measuring precise wavelengths of gas phase
  electronic spectra of such extremely reactive
  species in the laboratory is a highly nontrivial,
  demanding task
• no PAH identified among DIBs yet

7
DIBs spectrum
8
Fingerprinting a PAHthe worse...

• near and mid IR bands are not very sensitive to
  the specific PAH molecule, they mostly depend on
  the kind of chemical bond and on the functional
  group implied
• the near and mid IR bands due to a large
  population of different molecules are expected to
  stack neatly on one another, making them bright
  and easy to detect, but washing out spectral
  details (i.e. isotopic shifts)
• no identification is possible based on near and
  mid IR bands

9
UIBs and a single PAH emission spectrum
10
Fingerprinting a PAHthe worse...

• on a population of PAHs a fraction of them are
  expected to have a permanent electric dipole
  moment (i. e. partially dehydrogenated ones),
  hence should display a pure rotational emission
  spectrum
• free flying PAHs ought to be rotating
  suprathermally, yielding a very large rotational
  partition function
• each dipolar PAH would produce a veritable forest
  of weak rotational lines around 50GHz
• the pure rotational spectra of even as few as ten
  slightly different PAHs would merge in a
  quasi-continuum, which might contribute to the
  observed excess microwave emission but in which
  single molecules would be impossible to identify

11
Fingerprinting a PAHthe good...

• far-IR bands of PAHs involve collective, skeletal
  vibrations and are very molecular-specific
• the lowest energy vibration of a middle-sized PAH
  falls around 140µm, the second lowest around
  80µm, with predictable relative and absolute
  intensities
• we can have some hope to barely detect some of
  these bands in the ISO LWS database and things
  look very promising with the forthcoming Herschel
  mission

12
Fingerprinting a PAHthe good...

• the rotation of a free-flying interstellar PAH is
  highly non thermal
• it is driven by the interaction of the molecule
  with the radiation field and should be
  predictable
• many DIBs show apparent rotational structure

13
Fingerprinting a PAHthe best...

• It must all match together, at the same time, for
  a molecule identification!
• the electronic spectrum of many PAH ions and
  radicals is dominated by one strong transition,
  hence even a precise DIB wavelength
  identification, per se, might be a coincidence
• the same holds for a match in wavelength of a
  far-IR band
• if we get an approximate DIB wavelength
  identification, a good match of the observed DIB
  profile and a match in position and intensity of
  the far-IR bands, three hints make (almost) a
  proof

14
PAH physics in the ISM

• A typical middle sized PAH in the diffuse ISM
  will
• absorb an UV-Vis photon every 10 hours
• emit a pure rotational photon once a year
• have a collision once in 20 days
• change its ionisation state once a week (poorly
  determined)
• change its hydrogenation state in much longer
  time scales

15
PAH photophysics

• upon absorption of an UV-Vis photon, a PAH very
  quickly transfers most or all excitation energy
  to vibrations (transient heating) with a
  radiationless and essentially irreversible
  transition
• it then cascades down (cools) the ladder of
  vibrational states to the ground level by the
  emission of IR photons (including UIBs)
• between photon emissions, quick intramolecular
  transitions redistribute the excitation energy
  among vibrational (IVR) and rotational (IVRET)
  degrees of freedom, strictly conserving total
  angular momentum and energy

16
Photophysics of an isolated PAH
17
IC, IVR and IVRET

• Internal Conversion, Internal Vibrational
  Redistribution and Internal Vibration-Rotation
  Energy transfer all stem from terms in the
  Hamiltonian which are neglected in the
  Born-Oppenheimer approximation
• the Hamiltonian terms causing IC and IVR scale
  essentially with the density of available
  vibrational states
• the Hamiltonian terms causing IVRET scale with
  the density of vibrational states and with either
  J or J2

18
IVR and IR emission

• between IR emissions, the excited molecule can be
  considered a closed system in thermodynamical
  equilibrium
• the probability to have N quanta of vibrational
  energy in a given vibrational mode is given by
  the ratio of the density of all the other
  vibrational states at the remaining energy and
  the total density of vibrational states
• the probability of emission of IR photons can be
  readily computed given the population of
  vibrational levels and the (measurable or
  calculable) activities of the IR bands, i.e. the
  IR absorption spectrum of the molecule
• we can model a series of absorptions and their
  following emission cascades

19
IVRET and rotation

• IVRET couples rotational levels at a given total
  angular momentum J with the vastly larger energy
  reservoir of vibration
• it keeps energy levels within a given J in
  statistical equilibrium, their relative
  populations proportional to the density of
  vibrational levels at the energy Etot-Erot,
  ?(Etot-Erot)
• given the population of rotational sublevels and
  the transition dipole moments of the vibrational
  transitions, we can calculate the probability of
  emission of each single rovibrational line
• we can thoroughly model the exchange of angular
  momentum via IR emission

20
Jumping among ionisation states

• we are interested in achieving a statistical
  equilibrium condition of the rotation of the
  molecule under study
• if the molecule can change its ionisation or
  hydrogenation state, or there are other channels
  of formation and destruction on time scales
  comparable or shorter than the time it takes to
  reach equilibrium, we will need to model a
  network of interconnected species, following the
  evolution of angular momentum through its random
  walk among them
• it turns out that, as far as compact, saturated
  PAHs are concerned, this seems utterly unimportant

21
PAH model ingredients
ab initio optimised geometry ab initio
vibrational analysis
22
PAH model ingredients from ab initio calculations
NWChem (TD)DFT module with gaussian basis
Very good accuracy/computational cost ratio for

• optimised geometry
• vibrational analysis

but inefficient if many excited states are
required
23
PAH model ingredients from ab initio calculations
24
PAH model ingredients from ab initio calculations
Octopus TD-DFT code in real space
Very good accuracy/computational cost ratio for

• complete UV-Visible spectra of PAHs

25
PAH model ingredients from ab initio calculations
Yabana Bertsch, 1999, Int. J. Q. Chem. Marques
et al., 2003, Comp. Phys. Comm.
26
PAH model ingredients from ab initio calculations
27
PAH model ingredients from ab initio calculations
28
PAH model ingredients from ab initio calculations
29
PAH model ingredients from ab initio calculations
30
PAH model ingredients from ab initio calculations
Comparison with previously available PAH cation
data
31
PAH model ingredients from ab initio
calculations does it work?
Yes!
32
Cascading towards rotational equilibrium
33
What do we get out of the whole machine?

• along a given line of sight, the same molecules,
  if present, must be producing both DIBs, with
  equivalent witdh proportional to column density,
  and far-IR bands, again proportional to column
  density, hence
• the model can predict the ratios between the
  equivalent width of the DIB and the intensities
  of the far-IR emission bands, yielding yet one
  more independent criterion for PAH identification
• assuming sensible numbers for the oscillator
  strength of a permitted electronic transition in
  a PAH, we derive far-IR band intensities which
  might be barely detectable on ISO LWS database
  data and will be easily measured by Herschel

34
Equilibrium DIB rotational profiles
35
Full UIBs spectrum of a single PAH...
36
...including low energy modes, comparable with
ISO LWS data which are becoming available
37
What if we do identify PAHs?

• provide a firm, starting point to begin to
  understand quantitatively the chemistry of PAHs
• direct measure of some of the carbon which is not
  in the atomic state in the gas, nor in simple,
  observable molecules such as CO
• estimate and subtract their contribution from the
  extinction curve, constrain the other dust
  components
• use them as direct, ubiquitous probes of the
  radiation field in the diffuse ISM

38
What if we dont find any PAHs?

• you cannot turn off selected properties of
  molecules at will
• if PAHs dont produce at least some DIBs and/or
  their expected far-IR bands, they cannot produce
  any of the features commonly ascribed to them,
  such as the non linear far-UV rise of the
  extinction curve or the UIBs

Forget PAHs, explore alternative solutions
39
Current status NWChem in mass production...
40
Current status Octopus in mass production...
41
Current status Octopus in mass production...
42
Current status M-C model in mass production...
43
Current status M-C model in mass production...
44
Forthcoming (more) NWChem in mass production...

• Accurate individual electron affinities and
  polarisabilities, to be used to more accurately
  calculate ionisation equilibria for individual
  molecules
• Geometries and vibrational analyses for stable
  PAH anions
• Geometries of low-energy excited states

45
Forthcoming (more) Octopus in mass production...

• Individual photoabsorption spectra of stable PAH
  anions, to be used for accurate modelling of
  their individual ionisation equilibria

46
Forthcoming (more) M-C model in mass
production...

• Predicted interstellar rotational profiles of
  electronic transitions in the visible for PAHs in
  our database (to be compared with astronomical
  DIBs)
• Vibronic structure of electronic transitions in
  the visible for PAHs in our database (to be
  compared with astronomical DIBs). Follow up on
  the work by e. g. Dierksen Grimme, 2004, J.
  Chem. Phys.
• Predicted interstellar rotational profiles of IR
  emission for PAHs in our database (to be compared
  with astronomical observations of AIBs made by
  herschel)
• Extend the database to larger molecules, more
  charge states, PAHs substituted with heteroatoms,
  linear carbon chains, fullerenes...

47
Desiderata (quickly becoming feasible)

• Improve the accuracy of the calculated
  intensities of IR transitions (better xc
  functionals, larger basis sets...)
• Calculate photoionisation yields (Octopus can do
  it to some extent, TD-CDFT will do it better)
• De-excitation branching ratios in collision-free
  environments (Car-Parrinello-like simulations
  including diabatic treatment of level crossings)
• Theoretical estimate of the IVRET and/or ITE
  thresholds (multi-component TD-DFT)

48
featuring
Silvia Casu Giacomo Mulas Giuliano Malloci Cesare
Cecchi Pestellini (new group member!) Ignazio
Porceddu