Significance Photons can generate radiation pressure on objects due to optical momentum transfer. For light sources existing in nature, the mechanical effects are very weak, and thus difficult to be directly observed and utilized. Until the 1960s, the emergence of laser has provided strong collimating light sources for the study of optical radiation pressure, which finally resulted in the birth and development of optical tweezers technology.

As a pioneer of optical tweezers, Arthur Ashkin from Bell Laboratories successfully captured SiO₂ particles in water with double laser beams in 1970. Later, he successfully captured SiO₂ microspheres in air environments, and oil droplets in the environments of low gas pressures. In 1986, Ashkin used a highly converged single laser beam to trap the particles, which is well known as optical tweezers nowadays. The technology of optical tweezers is subsequently applied in biology, and Ashkin received the Nobel Prize in Physics in 2018.

In the past three decades, optical tweezers have been diversified into many schemes, including holographic optical tweezers, time-modulated optical tweezers, femtosecond optical tweezers, optical tweezers in vacuum, etc. Due to the features of non-contact and low damages, this manipulation technology has reached outstanding achievements in biology, nanotechnology, fundamental physics, quantum science, precision measurement, and so on. Especially in the last decade, the rapid development of optical tweezers in vacuum has attracted many top research teams. In 2010, the instantaneous velocity of Brownian particles have been measured in the experimental system of optical tweezers for the first time since Einstein's conclusions in 1907. Later, particles in micron-sizes are stably trapped in dual-beam optical tweezers in high vacuum, and the equivalent temperatures of the mass center motions can be cooled into several mK by the scheme of optical momentum modulations. At almost the same time, nanospheres are also stably trapped in single-beam optical tweezers in high vacuum, and a quite different scheme known as parametric feedback cooling is proposed and employed to decrease the center-of-mass temperatures into sub-Kelvin. For the particles, both in micron-scales and nano-scales, trapped in vacuum optical tweezers, the absence of collisions from the fluid molecules will provide robust decoupling from the heat bath of the fluid environments and no longer need cryogenic precooling. Thus center-of-mass temperatures close to the quantum ground state can be within reach in relatively miniaturized systems in room temperatures. Meanwhile, the trapped particles can be approximated as ideal harmonic oscillators, which is considered as the “ideal platform” for many physical quantity precise measurements, including weak force detection with resolutions in almost 1 aN/Hz^1/2, acceleration sensing with resolutions of about 100 ng/ Hz^1/2, milli-charge measurements, torque detection with a new sensitivity of 4.2×10^-27Nm/Hz^1/2, measurements of high frequency gravitational waves, and so on.

Progress Here we will introduce the fundamental theories, the main experimental setups, and typical applications of optical tweezers in vacuum. First, the introduction of the theories are described in two parts: the calculation models of optical forces and the principles of thermodynamics in optical tweezers. Later, the main experimental setups are introduced including optimized optical structures, efficient particle loading, precise position detection, reasonable stiffness calibrations and effective cooling schemes for center-of-mass temperatures. At last, recent applications in precision measurement are summarized.

Conclusion and Prospect In the past ten years, the quality factors and the trapping duration of the oscillators in optical tweezers in vacuum increase with the advance of experimental technologies. Great potentials have been shown in the following applications including sensing extremely weak forces and accelerations, rapid spinning control, fractional charge calibration, micro torque detection, and high frequency gravitational waves detection, etc. Thus the harmonic oscillators can be used to test the laws of thermodynamics, or to search the evidence of dark matter, or to explore macroscopic quantum effects. The realization of the macroscopic quantum states can further enhance the sensitivity of the precision measurement. In addition, optical tweezers in vacuum can work at room temperatures, without the need for additional refrigeration equipment. Recently, it is predicted that feedback cooling schemes might be eliminated to achieve stably trapped particles when the noise is low enough in ultra-high vacuum environments. This will make the whole system more concise and efficient to have more broad application prospects.

Currently, there are two main ideas for the future development of optical tweezers in vacuum: one is the system consisting of common optical components (named common system) and the other is the system to be built based on optical fibers and integrated optics (named integrated system). The common system will continue to pursue sensitivity breakthroughs in high precision measurement, and find possible applications in the exploration of multidisciplinary frontiers and cutting-edge technologies. Several issues are still waiting for better solutions, including further cooling of center-of-mass temperatures, more precise position detection of the particles, longer trapping duration, fewer laser trapping powers, and so on. A main focus could be how to realize a possibly simpler and more efficient cooling scheme for almost all particles, and thus reduce the system noises in a near future. At the same time, with the rapid development of fiber communications and the etching technologies in integrated optics, the integrated system can be made towards small volumes, miniaturization and low-power consumption. It is an important technical route for practical applications in the future human life. At present, there are still some unsolved problems, such as repeatable loading of single particles and integratable position detection of particles in chip-scale systems.