News & Views | Published:

Chemical physics

Quantum control of light-induced reactions

Nature volume 535, pages 4244 (07 July 2016) | Download Citation

An investigation of how ultracold molecules are broken apart by light reveals surprising, previously unobserved quantum effects. The work opens up avenues of research in quantum optics. See Letter p.122

The rupture of molecular bonds by the absorption of light drives chemistry in the atmosphere, causes DNA damage and the associated repair response, and provides a superb tool to study how molecules absorb light and then distribute and dispose of its energy. On page 122, McDonald et al.1 report their study of the light-induced break-up (photodissociation) of ultracold strontium molecules, Sr2. Their work provides insight into how molecules behave in the quantum regime of ultralow-energy dynamics that occurs just above energy thresholds for photodissociation.

Early photodissociation studies focused on the energetics2 of the products formed from diatomic molecules, and of the products' angular distribution3 — the distribution of angles at which they recoil relative to the direction of polarization (the polarization axis) of the light that excited them. If the energy of the photon absorbed by the diatomic molecule and the velocities of the resulting atomic fragments were known, then the bond energy of the molecule could be directly determined. The accuracy of these determinations depended on how cold the molecule was initially, and on how accurately one could measure the velocities of the products.

In the early experiments4, diatomic molecules were irradiated with laser light, and if the fragments were found to fly predominantly parallel to the laser polarization axis, then the transition dipole moment responsible for the light absorption was said to be parallel; similarly, perpendicular transitions were named after the associated perpendicular recoil. The transition dipole moment describes coupling between the two electronic states responsible for light absorption, and this classification was helpful in understanding its nature. For polyatomic molecules, the transition dipole moment does not have to align with a particular molecular axis, and many factors affect the measured angular distribution of the fragments. Measurements of the velocities of fragments provide information about the dynamics of the energy deposited within molecules as it evolves into the kinetic energy of the fragments.

Hundreds of photodissociation studies have been performed because of the fundamental information that can be obtained. With the advent of laser-based imaging techniques5,6,7,8 in the late 1980s, it became possible to measure velocities at high resolution (approximately a few metres per second) for particular electronic states of the products, by projecting the ionized products onto position-sensitive ion detectors. However, these experiments typically used pulsed-dye lasers (which produce light at a low frequency resolution of about 3,000 megahertz) to dissociate molecules and detect the products. This precludes experiments such as those performed by McDonald and colleagues, in which molecules are dissociated by photons that have a much higher, 1 MHz frequency resolution and energies just above the dissociation threshold of the molecule (that is, at light frequencies between 5 and 400 MHz greater than the dissociation-threshold frequency).

Moreover, these experiments typically used supersonic molecular beams as a source of cool molecules. When a high-pressure gas is expanded into a vacuum to form a molecular beam, the flow is directed forward supersonically at the expense of the kinetic energy associated with the other directions of flight and with the gas's internal degrees of freedom (the rotational and vibrational motion of its molecules). This allows molecules to be cooled to temperatures of a few kelvin even though they fly at close to velocities of 1,000 m s−1, with a spread of about 50 m s−1. McDonald and co-workers, however, wanted to study photodissociation fragments moving at only about 1 m s−1 (extremely slowly for a molecule, and correlating with a temperature of tens of millikelvin). To see such slow fragments, the authors held their molecules in a stationary laser trap, photodissociated them using a light pulse and then imaged the fragments after they had flown for about a hundred microseconds.

Molecules can interact with light through either the light's oscillating electric field (which causes electric dipole transitions) or its oscillating magnetic field (magnetic dipole transitions). For most covalently bound molecules, the light intensity required to produce electric dipole transitions is a million times less than that required for magnetic dipole transitions. McDonald et al. are the first to have excited a pure magnetic transition and observed the fragments. This was possible because the Sr2 molecules in this study are formed in the highest vibrational energy levels of the molecule's ground state, and therefore have a very long bond length, which increases the magnetic transition dipole moment by approximately 1,000-fold9.

Another groundbreaking feature of McDonald and colleagues' work is that the Sr2 molecules were prepared in a single rotational and vibrational quantum state by the laser-induced association of ultracold atoms, in the presence of an oriented magnetic field. Each state represents the projection (M) of a molecule's angular momentum vector (J) onto a quantization axis (in this case, the quantization axis aligns with the magnetic field). Several M states exist for each J value, and in the absence of a magnetic field they have the same energy (they are said to be degenerate); the number of M states is defined by the formula 2J + 1. When J is zero, it has no magnitude and no alignment in space. In a magnetic field, the M states do not have the same energy, because rotating electrons create a magnetic field that can be either aligned or counteraligned with the external magnetic-field quantization axis.

The authors' experiments started from a single (J,M) quantum state formed in the laser-association process. All of the quantum states reached during photodissociation were dictated by the starting state, and by the laser frequency and polarization relative to the magnetic field's axis. When the researchers obtained a single excited quantum state, they observed fragments recoiling predominantly parallel or perpendicular to the laser polarization. But if several degenerate quantum states were excited and interfered with each other, then the observed velocity distribution deviated spectacularly from purely parallel or perpendicular. These unexpected and previously unobserved angular distributions can be described only by a full quantum-mechanical treatment of the light-absorption process (Fig. 1).

Figure 1: Quantum effects in photodissociation.
Figure 1

McDonald et al.1 studied the light-induced fragmentation (photodissociation) of diatomic strontium molecules, Sr2, and observed surprising angular distributions of the resulting products. The left-hand panel shows a two-dimensional representation of the angular distribution of fragments obtained from Sr2 in a particular rotational quantum state, as predicted by quasiclassical theory; hot colours indicate higher distributions of fragments. The right-hand panel indicates the experimentally observed pattern, which can be explained only by using a full quantum-mechanical description of photodissociation.

At present, this sort of experiment is limited to a few diatomic molecules — some of which, like Sr2, are not covalently bound — that can be generated by cold-atom techniques. However, there is much to be learnt from these studies, and as scientists learn to cool and trap a larger array of covalently bound molecules, the techniques developed and knowledge gained will provide the foundation for future research —for example, in polyatomic molecules. The photo-physics of polyatomic molecules is more complex than that of diatomic molecules, because multiple mechanisms couple their electronic states to each other, and several fragmentation pathways are possible. In the meantime, I personally found this article a joy to absorb.



  1. 1.

    et al. Nature 535, 122–126 (2016).

  2. 2.

    , , & J. Chem. Phys. 51, 449–450 (1969).

  3. 3.

    J. Chem. Phys. 47, 889–895 (1967).

  4. 4.

    & Appl. Opt. (Suppl.) 4, 193–200 (1965).

  5. 5.

    & J. Chem. Phys. 87, 1445–1447 (1987).

  6. 6.

    & Rev. Sci. Instrum. 68, 3477–3484 (1997).

  7. 7.

    & Annu. Rev. Phys. Chem. 46, 335–372 (1995).

  8. 8.

    et al. Phys. Chem. Chem. Phys. 8, 26–53 (2006).

  9. 9.

    et al. Nature Phys. 11, 32–36 (2015).

Download references

Author information


  1. David W. Chandler is at the Combustion Research Facility, Sandia National Laboratories, Livermore, California 94550, USA.

    • David W. Chandler


  1. Search for David W. Chandler in:

Corresponding author

Correspondence to David W. Chandler.

About this article

Publication history




By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing