Nanometer-scale helium droplets [1, 2, 3] provide a unique solvent in which to perform molecular spectroscopy and are likely to become widely used by the molecular spectroscopic community in the coming years. Currently, this new technique is in its early stage of development and is being practiced in only a few laboratories, but the early results firmly support our view that this type of spectroscopy combines many of the most attractive features of more traditional molecular beam and matrix isolation spectroscopies. In particular, unstable and sometime exotic molecules can be synthesized in the condensed fluid nano-matrices without giving up the possibility of obtaining rotational resolution and therefore of obtaining direct structural information[4]. A review describing the early results has been published [5].
Large He clusters, 
, are formed upon the
expansion of low-temperature gas ( 
K) or liquid He
through a 
m nozzle into a vacuum. Such clusters cool by
evaporation, reaching temperatures of 0.4 K for 
He and 0.15 K for
He [6]. Uniquely, they remain liquid, as He will not
freeze even at zero K, except above 25 atm. In order to
emphasize the liquid nature of these clusters, and also their
characteristic size, they are referred to as nanodroplets. The
He nanodroplets are believed to be superfluid [7],
while the colder and less dense He clusters are believed to be
well above their superfluid transition temperature.
As for their heavier noble-gas counterparts, helium nanodroplets can
be doped with one or more atoms or molecules using the pick-up
technique [8], in which a beam of droplets passes
undeflected through a scattering chamber containing a small pressure
( 
mtorr) of the dopant species. Spectroscopic study of
doped He nanodroplets began with the work of Goyal et
al. [9], where line-tunable CO 
lasers were used to
probe the spectra of SF 
and its dimer. This work demonstrated
that the SF spectrum in He clusters was qualitatively different
from that observed in other inert-gas nanoclusters [10] in
that the lines were much narrower and the matrix-induced shifts smaller.
However, the line-tunable lasers did not provide enough spectral
coverage to resolve the spectrum rotationally. This was achieved at
the Max Planck Institute in Göttingen using a tunable diode laser
counterpropagating with the droplet beam, which allowed for a
continuous scan of the spectrum at high resolution [11].
In that investigation it was discovered that the spectrum had the
characteristic rovibrational structure of a gas-phase spherical top
molecule, including the first-order Coriolis splitting [12].
This work provided the first measurement of the cluster temperature,
which was found to be in excellent agreement with that predicted by both
classical [6] and quantum [13] evaporation
models. Rovibrational spectra of a range of other molecules
have now confirmed that He nanodroplets have a
temperature of 0.38K, essentially independent of the dopant, but
somewhat dependent upon cluster size. The SF spectrum of
ref. [11] also showed that the effective rotational
constant measured in He droplets is 1/3 of the gas phase value,
while the centrifugal distortion constant is four orders of
magnitude larger than that of the isolated dopant. These trends have
been confirmed by the recent experiments in Göttingen, Chapel
Hill, and our own laboratory, where the rotationally resolved spectra for
several other molecules have been
obtained [14, 15, 16]
Many of the spectroscopic experiments performed to date on doped He
nanodroplets have focussed on electronic spectroscopy. This is largely
due to the fact that the highly sensitive method of laser-induced
fluorescence could be used to detect absorption and possible subsequent
dynamics. We mention in particular only the work of our laboratory on
alkali-doped droplets, where spectra of Li, Na, and K atoms, dimers
and trimers have been extensively studied, including the Jahn-Teller
coupling and time-resolved photochemistry of the quartet state of
Na 
, which had never previously been
observed [17, 18, 19]. One problem with
electronic excitation, however, is that the transition-induced changes
in He solvation can result in most of the intensity being found in the
`phonon side band', though `zero phonon lines' have been observed in
several systems. This makes it difficult to extract unambiguous
rotational information in such experiments, particularly since the
low-energy excitation region of the spectra, in the neighborhood of the
zero phonon line, is not yet understood [17, 18].
For transitions that substantially change the size of the electronic
cloud of the atom or molecule, very broad structures are observed and
have been explained by the `bubble' model originally developed to
treat electrons solvated in bulk liquid He [5]. In
contrast, vibrational transitions induce only small changes in the
solvation of He around the dopant and, as a result, infrared spectra
have almost all their intensity in the `zero phonon' line, where no
direct optical excitation of the internal degrees of the He droplet
occurs. In fact, to date no `phonon side band' has been observed in
the IR.
IR spectra are normally observed by the method of beam depletion. In
the time between IR absorption and detection of the droplet beam
(typically 
s), vibrational relaxation of the dopant
leads to He evaporation. Each departing He atom reduces the internal
energy of the droplet by 
cm 
, so near-IR excitation
reduces the size of the droplets by about one thousand atoms. This
photoinduced beam depletion has been detected either optothermally
with a bolometer [20] or by reduction in the electron impact
ionization yield combined with detection of all ions above a certain
mass threshold [11].