A standard tool in optics labs, titanium-doped sapphire lasers are valued for their ability to be precisely tuned across a wide wavelength range. In Stanford University’s Nanoscale and Quantum Photonics Lab, led by Jelena Vučković, researchers use tabletop Ti:sapphire lasers to excite artificial atoms in solid-state quantum optics experiments.
But the lasers typically require a bulky, expensive, high-power pump laser. And the Vučković group, like many others, require only a fraction of the Ti:sapphire’s output power: The researchers in the Nanoscale and Quantum Photonics Lab often end up attenuating the laser from watts to microwatts.
Dissatisfied with the standard laser setup’s wasted power, high cost, and other shortcomings, the Stanford researchers saw an opportunity to miniaturize. Vučković is no stranger to shrinking lab components: She and members of her team had previously collaborated on the construction of a particle accelerator on a chip.1
Taking advantage of recent improvements in integrated photonics, the team has now scaled down the laser to an on-chip platform in which the light passes through a waveguide amplifier fabricated from a Ti:sapphire film.2 Although it is only a first iteration, the smaller laser is a better match to the power, stability, and tunability needs of Vučković’s lab and presumably those of many others.
Why Ti:sapphire?
Developed in the 1980s, Ti:sapphire lasers are solid-state lasers that can operated in continuous or pulsed mode. A high-power pump laser—the 10 kg one in Vučković’s lab costs around $100 000—is directed through a crystal of Ti:sapphire, which serves as the gain medium. A pump laser of sufficient power excites the Ti atoms to generate lasing at a different, longer wavelength.
Sapphire doped with Ti3+ ions has the largest gain bandwidth of any laser crystal, which allows for a broad wavelength-tuning range of 650–1100 nm and the generation of pulses as short as 5 fs. In their pulsed mode, Ti:sapphire lasers reach powers that are 100 times as great as those of dye lasers. The laser’s combination of flexibility and power has been crucial for advancements in two-photon microscopy and optical frequency combs.
Previous attempts to scale down Ti:sapphire lasers have been only partially successful. In 2023, researchers at Yale University bonded Ti:sapphire material onto a silicon nitride photonic chip, but only a small portion of the optical field was inside the gain material. That partial overlap led to higher lasing thresholds and milliwatt output powers that are not well matched for many common experimental needs. Moreover, the lasers were not designed to be tunable, which severely limited their potential applications.3
Integrating photonics
The device developed by Vučković’s group is built on a thin-film monocrystalline sapphire-on-insulator photonics platform. High-quality crystalline Ti:sapphire isn’t easily grown in a thin, uniform layer atop anything other than sapphire. Growth on a different material results in a lattice mismatch, and the imperfections in the crystalline structure prevent the attainment of the desired laser properties. So instead, the Stanford researchers start with a bulk material: a uniform wafer of approximately 0.5-mm-thick Ti:sapphire. The wafer is bonded to the sapphire (the translucent rod seen in figure 1) using silica as a glue before the wafer is ground and polished down to 450 nm.
Once the Ti:sapphire layer is thinned, it needs to be etched to create a waveguide, illustrated in figure 2, that serves as the laser cavity. The spiral waveguide, seen in figure 3, keeps the light inside the lasing medium for longer, thereby amplifying the light, than if the light had taken a straight path. Sapphire’s hardness makes waveguide etching a difficult and often long process. So the researchers developed a pattern-transfer method that uses an electron-beam photoresist and a chromium mask that can withstand the lithography process.
In tabletop Ti:sapphire lasers, the pump and lasing modes only partially overlap, which limits the pumping efficiency and increases the pump-power requirements. But the on-chip waveguide achieves a near-perfect overlap between the pump and lasing modes, which increases the efficiency of the system and allows the laser threshold to be reached with a lower-power pump laser.
The laser developed by the Vučković group has a spiraling waveguide with a footprint of only 0.15 mm2, and it can be pumped with an off-the-shelf 110 mW green laser diode. Tuning the laser across the whole 650–1100 nm range is achieved via an integrated microheater that changes the index of refraction and, hence, the resonance wavelength.
Laser technology for all
For now, the new laser operates only in continuous-wave, nonpulsed mode. It uses a basic $37 pump laser—three orders of magnitude cheaper than full-sized pump lasers—that only weighs a few hundred milligrams and delivers the few milliwatt output power needed. Vučković and her team have demonstrated that the laser is good enough to start using in their research experiments.
The inventors of the original Ti:sapphire laser didn’t file a patent. The open technology and lack of legal complications allowed Vučković and her lab to improve on the design. Now Joshua Yang, one of Vučković’s students and the first author of the paper, is starting a company to develop and distribute the tiny laser and amplifier technology.
Meanwhile, Vučković and her group continue innovating on the design of their laser. Priorities include increasing the coupling efficiency from the current 16% and developing a pulsed version. Other plans include achieving better thermal stability based on the placement of the heater and achieving increased laser frequency stabilization.4