Spectroscopy of neat and doped hydrogen clusters

(in collaboration with K.K. Lehmann)



Why are we doing it?

There are two big questions we are trying to answer with our research. The first is rather fundamental, and has been intriguing scientists for almost 30 years [1].
The second is of great technological interest for the next generation of space missions:

  1. Can we make para H2 to be superfluid?
  2. Can we improve the properties of solid hydrogen as a rocket propellant?

 Superfluid H2

Helium has been for for a long time (and as far as condensable gases are concerned, still is) the only system known to exhibit the property of superfluidity. Scientists like to see if their conclusions are of general validity, so some of those trying to understand superfluid helium started looking for another system that could possibly be superfluid. It is only recently that dilute Bose-Einstein condensated of alkali atoms could be studied in a laboratory, and we will not deal with them here (the interested reader can find more in a recent review [2], or at this homepage), so the search was indeed restricted to condensable gases.
Superfluidity is inherently a manifestation of quantum behavior (occurring for a collection of indistinguishable bosons), a convenient measure of which is the De Boer quantum parameter l [3]. After He, H2 has the largest value of l, and is therefore the most natural candidate to look for superfluidity; already 30 years ago it was predicted that H2 should be superfluid below 6 K [1]. Unfortunately, H2 solidifies at 13.8 K, so one needs a high degree of supercooling; despite repeated attempts with rather clever setups, superfluidity in H2 still has to be observed. Clusters are the best candidate to observe superfluidity: they have an estimated temperature of 4.7 K, which is maintained by evaporative cooling [4]. Because of their small size, rapid cooling and absence of nucleation sites, they are most likely liquid.

H2 as a rocket propellant

Part of our research program is linked to the efforts of the Air Force Office for Scientific Research (AFOSR) aimed at the discovery of high-performance rocket propellants. The performance of a propellant is measured by its specific impulse (Isp, defined here, or here) which has the dimensions of a time.
Currently, the most efficient propellant is a mixture of liquid oxygen and liquid hydrogen, and has an Isp of about 450 s. The goal of the High Energy Density Matter (HEDM) program is to reach an Isp of 500 s or greater. If this increase seems small, consider that the fuel mass is a substantial fraction of the mass of a rocket (or a missile, or a space shuttle); conversely, the payload is 2-5% of the total mass. Obviously a modest increase in Isp can result in a 100% increase in payload.

The AFOSR effort on higher-performing fuels is currently focused on:

Seeded solid hydrogen is denser than the liquid (requiring less storage space) and, if seeded with the appropriate material (reactive atoms or unstable molecules) has a higher Isp. Incidentally, if solid hydrogen is to be used as a fuel, a fast melting rate is essential; indeed one of the sought properties of the dopants is that it must accelerates melting, e.g. by release of dopant-dopant recombination energy.

Currently, the key issues are:

  1. To develop techniques for large-scale production of seeded cryogenic solids.
  2. To develop analytical and theoretical methods for characterization of seeded cryogenic solids.
  3. To identify the best dopants and their maximum attainable concentration.

Admittedly, the typical space shuttle "gas station" will look a little bit different than our experimental apparatus, but as far as items (2) and (3), doped hydrogen clusters are possibly the best system for experimental studies. Each cluster is an isolated system that can be loaded with a controlled number of different atoms/molecules, and interrogated with a variety of spectroscopic techniques.

How do we do it?

We produce H2 clusters by expansion in vacuum of high pressure (30 atm) hydrogen gas through a small orifice, or nozzle, (10 µm diameter) kept at a temperature of 60-70 K. Under these conditions, a supersonic beam of H2 is produced; rapid cooling of the gas occurs in the first hundred microns downstream the nozzle, and H2 molecules aggregate into clusters of a few thousands molecules each.
The supersonic beam has a speed of 1200 m/s, so the clusters survive for less than a millisecond (the time it takes to travel from the source to the detector). During this time we (1) load them, (2) excite the doped clusters with our favorite tunable radiation source, (3) detect their interaction with radiation.

  1. Loading: is accomplished by passing the clusters through a pick-up cell containing a low-pressure vapor of the atoms/molecules of interest. Each cluster picks up one or more atoms/molecules, much the same way as flies on the windshield of a car; therefore the probability of having a chosen number of atoms/molecules per cluster follows a Poisson distribution which depends on the pressure in, and length of, the pick-up cell. Since the typical pressure needed to maximize singly-doped clusters is very low (some 10-4 torr), clusters can be doped with any molecule that either is a gas at room temperature, or can withstand moderate heating without decomposing.
  2. Excitation: can be done with any source of tunable radiation: microwaves, infrared, visible, or ultraviolet, as long as the dopant absorbs at some frequency within the tuning range.With powerful radiation sources, or when sensitivity is not an issue, the radiation is made to cross the supersonic beam orthogonally. Otherwise, to maximize the interaction region, either the radiation is propagated collinearly with the beam or multi pass arrangements are used.
  3. Detection: when the excited complexes do emit fluorescence, photon counting (with a photomultiplier) gives the best sensitivity; in addition, the fluorescence can be dispersed with a monochromator and frequency-resolved emission spectra can be collected. This is the standard choice when studying electronic excitations of atoms/molecules in He clusters. With H2 clusters, fluorescence may be quenched, and beam-depletion techniques may be more suitable. The latter are also used in all the cases when the molecule's fluorescence is too slow (infrared, microwaves) and its collection is not an option. Beam depletion detection is based on the concept that the energy of the absorbed radiation causes part of the beam to evaporate in random directions. Detection is accomplished by sensing the kinetic energy of the beam (with a bolometer), or with any other detector that can probe the intensity of the on-axis helium flux.

What we have done

So far we have collected the visible fluorescence and beam depletion spectra of alkali-doped hydrogen clusters. We have found that hydrogen molecules are very effective at quenching the fluorescence of alkali atoms, to the extent that a beam depletion approach is necessary to measure the true absorption spectrum of these complexes [6].
In a different series of experiments, we have measured the rotationally resolved spectra of the first overtone of several molecules containing the ºCH stretch chromophore, solvated in He clusters, with signal-to-noise ratios of up to 103/Hz1/2. It is currently believed that preservation of rotational coherence is a key signature of superfluidity [7]. If we will be able to observe narrow-line rotationally-resolved spectra of molecules captured by H2 clusters, we will have a strong evidence that they are indeed superfluid.

What plans we have for the near future?

We are adapting our infrared He cluster machine so that hydrogen clusters can be produced and investigated as well. Our estimates tell us that we should be able to detect the absorption spectrum of HCN in the first overtone of the CH stretch. If the cluster is superfluid, we expect to be able to resolve the rotational structure. Should the cluster be solid instead, we might see a series of very sharp lines (similar to those observed for CH4 in solid H2 matrices [8]), whose splitting pattern would tell us about the structure and energetics of the trapping site. Since our spectrometer has an instrumental resolution of about 15 MHz, this circumstance would be very advantageous in high-resolution studies of exotic species trapped in a solid matrix.
In all fairness, it is very likely that the HCN or any othe similar molecule will act as a condensation seed and cause the cluster to solidify. A pure cluster is a better candidate to search for superfluidity, if only a suitable spectroscopic transition of the H2 molecule can be found. Moreover, the density of H2 molecules in the beam is obviously much higher than that attainable for dopant molecules. Our plan is to use a Ti:Al2O3 laser and a YAG laser to perform stimulated Raman spectroscopy of pure H2 clusters.

[1] V. Ginzburg and A. Sobyanin, Can liquid molecular hydrogen be superfluid?, JETP Lett. 15, 242 (1972).
[2] F. Dalfovo, S. Giorgini, L. P. Pitaevskii, and S. Stringari, Theory of Bose-Einstein condensation in trapped gases, Rev. Mod. Phys.71, 463 (1999).
[3] For a liquid/solid of atoms/molecules of mass m interacting through a van der Waals potential parametrized by a core radius s and a well depth e, one has l2=h2/(s2me), hence large quantum effects are observed for light, weakly interacting species.
[4] E. Knuth, F. Schünemann and J. P. Toennies, Supercooling of H2 clusters produced in free-jet expansions from supercritical states, J. Chem. Phys. 102, 6258 (1995).
[5] J. A. Sheehy, in Proceedings of the 1999 HEDM Contractors Conference.
[6] C. Callegari, J. Higgins, F. Stienkemeier, and G. Scoles, Beam Depletion Spectroscopy of Alkali Atoms (Li, Na, K) Attached to Highly Quantum
Clusters, J. Phys. Chem. A 102, 95 (1998).
[7] S. Grebenev, J. P. Toennies, and A. F. Vilesov, Superfluidity Within a Small Helium-4 Cluster: The Microscopic Andronikashvili Experiment,Science 279, 2083 (1998).
[8] T. Momose, M. Miki, T. Wakabayashi, T. Shida, M.-C. Chan,S. S. Lee, and T. Oka, Infrared spectroscopic study of rovibrational states of methane trapped in parahydrogen crystal, J. Chem. Phys. 107, 7707 (1997).

Return to the Scoles Group Research Page

Return to the Scoles Group Homepage