The Color Purple: It Takes Three Photons

Breck Hitz

Radiation in the violet and ultraviolet wavelength ranges is important for photolithography and fluorescence imaging, for future methods of three-dimensional optical storage and for other applications. There are several with which to generate this short-wavelength radiation, including excimer lasers and complex combinations of nonlinear techniques such as harmonic generation and parametric oscillation.

Recently, investigators at the Chinese Academy of Sciences and at Shanghai Jiao-Tong University, both in Shanghai, demonstrated violet up-conversion luminescence precipitated by the simultaneous absorption of three infrared photons in a femtosecond-scale pulse. Although the conversion efficiency from the infrared to the violet was relatively low, the mechanism offers the possibility of new sources of short-wavelength radiation, including the enticing possibility of all-solid-state violet and ultraviolet lasers.

Figure 1. The violet emission is bright enough to be visible to the naked eye.

The researchers generated violet light by focusing 800-nm, 120-fs, 1-kHz pulses from a regenerative-amplified Ti:sapphire mode-locked laser onto a Ce³⁺:Lu₂Si₂O₇ crystal. They grew the crystal using the Czochralski technique and sliced the resulting boule into 2-mm-thick slices. The cerium impurity replaced the lutetium within the pyrosilicate crystal and acted as the luminescence center for the violet light. The high peak power generated in the femtosecond pulses facilitated nonlinear three-photon absorption in the crystal.

The violet light was bright enough to be observed easily with the naked eye (Figure 1). Its emission spectrum was very nearly the same whether it was excited by 800- or 267-nm radiation (800/3), lending credence to the scientists’ hypothesis that the three-photon absorption was behind the emission (Figure 2). They believe that transitions in the cerium ion from the lowest 5d excited states to the 4f ground states were responsible for the violet light.

Figure 2. The emission spectra Ce³⁺:Lu₂Si₂O₇under 800-nm excitation is nearly identical to that under 267-nm excitation, indicating that three-photon absorption is a likely explanation for the emission under 800-nm excitation. ©OSA.

To bolster their hypothesis, they observed that the intensity of the violet output varied quantifiably with the cube of the infrared intensity, which is consistent with three-photon absorption.

The cubic dependence, however, also could be explained by absorption of a single photon followed by the simultaneous absorption of two photons, or by the inverse of that process (i.e., two-photon absorption followed by absorption of a single photon).

To eliminate these factors, the investigators measured the absorption of the Ce³⁺:Lu₂Si₂O₇crystal at wavelengths corresponding to the energies of one photon and of two photons (Figure 3). They found that there was virtually no absorption at either wavelength (800 and 400 nm, respectively), but that there was significant absorption at the three-photon wavelength.

Figure 3. The absorption spectrum of Ce³⁺:Lu₂Si₂O₇shows no absorption at wavelengths corresponding to the energies of one or two 800-nm photons, but a significant absorption at 367 nm, corresponding to the simultaneous absorption of three 800-nm photons. ©OSA.

Similarly, the scientists considered and eliminated other possible explanations, including excited-state absorption, energy transfer up-conversion, cooperative up-conversion and photon avalanche.

Finally, they performed transmission measurements through the crystal at 800 nm and calculated from the resulting data that the cross section for three-photon absorption was 2.44 × 10^–77 cm⁶ · s².

Optics Letters, July 15, 2006, pp. 2175-2177.