Home > Research > Single-photon Nonlinear Optics
Atom Photonics and Slow Light on a Chip

   Home       /       /       /       /       /       /       Electrical Engineering       Baskin School of Engineering       UC Santa Cruz

   © 2010 H. Schmidt.  All rights reserved.                                                                                                        Contact

Atom Photonics and Slow Light on a Chip

We are pioneering the development of atom photonics – self-contained chip-scale devices for integrated atomic spectroscopy and the observation and utilization of quantum interference effects in atomic vapors on a chip [1].

Integration of atomic vapors with semiconductor chips has many advantages (creation of compact atomic spectroscopy devices, large intensities over long distances, creation of complex optical layouts, potential for novel photonic devices, integration with other photonic elements on the same chip) and applications (frequency standards, slow and stopped light devices, metrology).



Our approach is to introduce atomic vapors into hollow-core optical waveguides on a silicon chip. This approach results in fully self-contained miniature atomic vapor cells and enables flexible photonic layouts and convenient coupling with standard optical fiber. Fig. 1 shows this concept schematically along with a completed atom photonic chip the size of a quarter coin.

Fig. 1: Conceptual view of atomic vapor cell on a chip. Hollow-core (orange) and solid-core (blue) ARROW waveguides are combined on a silicon chip to create compact spectroscopy cells of mm length. Right: photograph of completed device.

We are pursuing a number of open questions and research directions in this field:

  • Fabrication and optimization of hollow-core optical waveguides
  • Integration of atomic vapors on a chip
  • Atomic spectroscopy on a chip
  • Quantum interference on a chip: Slow light, stopped light, induced transparency, single photon nonlinearities
  • Optical manipulation of atoms on a chip
  • Novel atom photonic devices

Some of our recent accomplishments include:

  • Demonstration of low-loss hollow-core ARROW waveguides [2,3]
  • Fabrication of chip-scale Rb cells with long lifetimes [4,5]
  • First atomic spectroscopy chip [6]
  • Saturated absorption spectroscopy in Rb atoms on a chip [6]
  • World record slow light on a chip [7]
  • EIT in hot Rb atom on a chip [7]
  • Novel approach to optical slowing of atoms using AC Stark shift [8]

Here are some examples of our work on atom photonics:

Integrated atomic spectroscopy cells

In collaboration with the Hawkins group at BYU, we have developed hollow-core optical waveguides that form the ideal basis for atom photonic chips. Fig. 2 shows cross-section and side-view SEM images of these micron-scale channels that confine light using the thin film dielectric layers [1,2]. The thick top oxide of the hollow core channel simultaneously acts as the core for a solid ridge waveguide that allows for coupling light into and out of the hollow channel. This way, the channel can be routed to a different spot on the chip where a reservoir containing an alkali metal such as Rb is placed (see right image). Separate optical and atomic access to the waveguide channel greatly simplifies the chip design. In addition, the Hawkins group has optimized the approaches to sealing the reservoirs onto the chip without reacting with the alkali vapor. Heating of the chip then produces an optical vapor inside the hollow-core waveguide for carrying out spectroscopy.

Fig. 2: SEM images of hollow-core ARROW waveguides.

Fig. 3 shows the results of atomic spectroscopy on the hyperfine transitions of Rb (Fig. 3b). These can be used for applications in atomic clocks as well as to implement quantum coherence effects (see below). Fig. 3a shows how the linear absorption spectrum compares nicely with that observed in a glass bulk cell while Fig. 3c shows a saturated absorption spectrum where a counterpropagating beam depletes the atomic ground states and leads to characteristic Lamb dips in the spectrum that can be used for frequency calibration and stabilization. The waveguide geometry is ideal for this purpose since counterpropagating beams are automatically aligned via the waveguide modes.


Fig. 3: Atomic spectroscopy on a chip [6]. (a) linear absorption in waveguide chip and glass bulk cell; (b) corresponding Rb hyperfine spectrum; (c) SAS spectrum exhibiting Lamb dips in presence of counterpropagating beam.

Slow light on a chip

One of the most intriguing possibilities for using an integrated atom photonic chip is to create quantum coherence effects at or above room temperature [9]. Many spectacular and well-publicized results have been produced in alkali vapors. This type of atomic quantum state control can be used to reduce the speed of light down to a few meters per second, temporarily store the quantum state of a photon in an atomic medium, render an opaque medium transparent, and realize phase modulation and optical switching with few or single photons [1]. Fig. 4a shows the generic four-level scheme that exhibits giant Kerr nonlinearities [2]. A coherent coupling field applied between levels 2 and 3 creates a quantum interference that leads to a constructive interference for the Kerr nonlinearity while eliminating linear absorption. As a result, the presence of a signal field denoted by WS imparts a large phase shift on the probe field WS through cross-phase modulation. The effect can be so strong that it may enable quantum non-demolition detection with single-photon signals.

We have demonstrated the underlying EIT effect on a chip for the first time. Fig. 4b shows the probe beam absorption profile in presence of the coupling beam on the |2>-|3> transition, exhibiting a restoration of 44% of the vapor transmission with only 1microwatt optical control power. Concomitant with this absorption change is a steep dispersive feature in the real part of the linear susceptibility that is responsible for dramatically reducing the group velocity of a light pulse propagating through this medium. Fig. 4c shows this slow light effect manifest in the delay of a probe pulse by almost a full pulse width compared to a reference pulse. The data are in excellent agreement with a theoretical model (lines). The dependence of the group velocity reduction on the coupling laser power is shown in Fig. 4d and shows a slow-down factor of up to 1,200 – a world record for slowing down light on a chip.



Fig. 4: Quantum interference and coherence on a chip [7]. (a) 4-level scheme for creating giant Kerr nonlinearities. (b) probe absorption in presence of coupling beam, exhibiting 44% transparency at the two-photon resonance; (c) slow light propagation of nanosecond light pulse on the chip (red) compared with free space reference beam (blue); (d); dependence of slow light effect on coupling laser power.

Atomic cooling with AC Stark shift

It is well known that the motion of atoms can be affected by light via radiation pressure and dipole (gradient) forces. This has enabled slowing and stopping of atoms. This effective cooling has formed the basis of magneto-optical trapping and Bose-Einstein condensation. One can cool an atom with a counterpropagating optical beam as the photons are absorbed and isotropically re-emitted. This redistribution of momentum results in a slowing force and effective cooling along the beam propagation direction. To be effective, the laser beam has to be resonant with the atomic transition in the reference frame of the atom. In order to compensate for the continually changing Doppler effect, the atomic resonance or the laser frequency has to be adjusted as a function of space. The former is typically done by using an inhomogeneous magnetic field via the Zeeman effect. We proposed an alternative method for keeping atoms and cooling beams in resonance that is ideal for implementation in (lossy) on-chip waveguides [7]. Fig. 5a shows the principle, in which a second light beam creates a spatially dependent AC Stark shift of the lower atomic level used by the cooling beam. If the power of the AC Stark beam is tailored correctly, the atom will experience a level shift that just compensates for the change in Doppler shift (both Doppler and AC Stark shift get smaller as the atom propagates). This AC Stark power can be matched with the waveguide loss in a hollow-core waveguide, for example by changing the width of the waveguide along the beam direction (Fig.5b).



Fig. 5: AC Stark cooling of atomic beam [8]. (a) Principle and implementation in Rb atoms; (b) tailored hollow-core ARROW width to slow atomic beam from room temperature to 40m/s.

This work is funded by DARPA/AFOSR under the Slow Light Program, and the National Science Foundation (NSF).


[1] H. Schmidt and A.R. Hawkins, “Atomic Spectroscopy and Quantum Optics in Hollow-core Waveguides”, Laser and Photonics Reviews 4, 720 (2010).
[2] D. Yin, J.P. Barber, A.R. Hawkins, and H. Schmidt, "Integrated ARROW waveguides with hollow cores", Optics Express, 12, 2710, (2004).
[3] E.J. Lunt, B. Wu, J.M. Keeley, P. Measor, H. Schmidt, and A.R. Hawkins, “Improving Hollow ARROW Waveguides on Self-Aligned Pedestals for Improved Geometry and Transmission”, IEEE Phot. Tech. Lett. 22, 1041 (2010).
[4] J.F. Hulbert, K. Hurd, B. Wu, H. Schmidt, and A.R. Hawkins, "A versatile approach to Rb vapor cell construction", Journal of Vacuum Science and Technology A, Vol.29, Issue 3, DOI: 10.1116/1.3568954 (2011).
[5] J.F. Hulbert, M.Giruad-Carrier, T. Wall, A.R. Hawkins, S. Bergeson, J. Black, and H. Schmidt, "Versatile Rb vapor cells with long lifetimes", J. Vac. Sci. Technol. A 31, 033001 (2013).
[6] W. Yang, D.B. Conkey, B. Wu, D. Yin, A.R. Hawkins, and H. Schmidt, "Atomic spectroscopy on a chip", Nature Photonics 1, 331 (2007).
[7] B. Wu, J.F. Hulbert, K. Hurd, E.J. Lunt, A.R. Hawkins, and H. Schmidt, “Slow light on a chip via atomic quantum state control.”, Nature Photonics 4, 776 (2010).
[8] J.A. Black and H. Schmidt, "Atomic cooling via AC Stark shift", Optics Letters, 39, 536-539 (2014).
[9] H. Schmidt, and A.R. Hawkins, "Electromagnetically induced transparency in alkali atoms integrated on a semiconductor chip", Applied Physics Letters, 86, 032106, (2005).