The different mode modulated LIFT experimental setup is illustrated in Figure 2. A 532-nanometer single-pulse laser with a pulse duration of 2 nanoseconds (FWHM) was chosen as the light source. The energy modulation module consists of a half-wave plate (HP) and a polarizing beam splitter (PBS). To expand the laser beam, laser expansion mirrors (Lens 1, Lens 2) were employed, and the influence of stray light was minimized through pinhole filtering. The pure-phase reflective liquid-crystal-on-silicon spatial light modulator (LCOS-SLM) has a resolution of 1,920 × 1,080, operates in the wavelength range of 400–850 nanometers, and provides at least 2π phase delay for the 532-nanometer wavelength laser. Lens 3–6 constitute two 4F systems, where mirrors (mirrors 1–5) adjust the laser’s direction, and the modulated beam passes through two sets of 4F systems into the ejection objective. Two 4F systems were designed to meet spatial requirements and minimize optical intensity loss. The bottom imaging system captures cell images and is composed of a microscopic objective and a CMOS camera. During fluorescence imaging, the BS is switched to a fluorescence cube. The high-speed imaging system, comprised of a long-focal-length objective and a high-speed camera, ensures high temporal and spatial resolution, to record the whole process of microbial cells ejection.

The algorithm for generating vortex beams employs a fork grating method. By calculating the laser wavelength, the phase of the fork-shaped grating is obtained and then loaded onto the SLM to generate the vortex beam. Adjusting the order of the grating allows the generation of vortex beams with different diameters (Haotong et al., 2013; Chen et al., 2016). The generation of flat-top beams utilizes a hybrid holographic method, combining binary gratings and geometric masks (Liu et al., 2018). This method loads a hybrid holographic pattern onto the SLM, achieving the shaping of high-quality flat-top beam. The horizontal polarization laser beam, reflected by the spatial light modulator, diffracts into multiple beams, with energy mainly distributed in the ±1 orders and 0 order. These ±1 order beams are blocked by a slit, significantly reducing diffraction effects around the shaped beam. By loading corresponding holographic patterns onto the spatial light modulator, we can conveniently obtain flat-top beam profiles of any shape.

The LIFT chip is coated with a 25-nanometer-thick aluminum film as an absorption layer. Under the action of laser pulses, the aluminum film is heated, generating powerful ejection force to achieve precise cell sorting. We directly attached the E. coli samples onto the LFIT chip, and then turned over the chip and mounted it on an XY displacement stage. And then observed and selected the samples under a microscope. A 0.17 mm thick cover glass was employed as the receiver to show the distribution of received cells (as shown in Figure 3A). By adjusting the imaging objective to focus on different planes, we could clearly observe the distribution of cells on both the chip and the receiver, with a distance of approximately 0.5 mm between them.

To provide a clearer characterization of the ejection process of cells under different mode beam conditions, we established a side-view high speed microscopy imaging system, to observe the entire process of microbial cells from ejection to reception from a lateral perspective (as shown in Figure 4A). The side-view imaging objective was mounted on a XYZ three-dimensional displacement stage. By adjusting the displacement stage, the objective could be aligned to the targeted cell. The optical path behind the objective guided the image to the high-speed camera, which owns 1,920 × 1,080 pixels and the frame rate of up to 1,088 fps.

In the process of LIFT single-cell sorting and cultivation, we implemented a series of steps to ensure the accuracy and recovery rate. Initially, to ensure uniform spreading of the bacterial suspension on the aluminum film during LIFT, three layers of phytic acid PA-Al³⁺ film were attached to provide hydrophilicity (Wang et al., 2018). Precise control of the XY electrical displacement stage allowed us to select the target cells and position them at the location of the laser pulse. Pulsed laser energy used for sorting is 2.55 μJ. To prevent contamination, the entire experimental setup, chip, and receiver were placed in a cleanroom and subjected to 20 min of UV irradiation before the experiment. For the receiver, a 35 mm culture dish with agar medium was used, and the distance between the receiver and the LIFT chip was approximately 2 mm. We developed a program to control the displacement stage to stop at 9 receiving place on the culture dish. While observing bacteria or cells on the LIFT chip, the receiver was moved out of the optical path. Once the target cells were identified, the receiver was moved to the correct position to capture the ejected cells. The singly sorted cells were then cultured in the agar dish: for Saccharomyces cerevisiae, the dish was cultured at 30°C for 36 h, and for E. coli, it was cultured at 37°C for 24 h.

Figure 5 illustrates the process of generating different mode beams. Initially, by calculating the phase of the vortex beam at a wavelength of 532 nm and loading the phase map (Figure 5A) onto the SLM, we successfully realized the formation of vortex beam (Figure 5B). Subsequently, we employed a hybrid holographic approach for shaping the flat-top beam. Setting a geometric mask with the desired shape as zero grayscale level, creating zero-phase delay to act as a reflective mirror, and loading the hybrid phase map (Figure 5D) onto the SLM, we achieved the configuration of flat-top beams (Figure 5E).

Under the influence of pulsed laser, these two beam modes exhibited different material transfer effects, providing ideal optical tools for subsequent experiments. The vortex beam induced a circular material transfer (Figure 5C), while the flat-top beam presented a square material transfer (Figure 5F).

The precise separation and reception of single cells represent crucial steps in studying and experimenting at the single-cell level, holding paramount significance for the isolation and cultivation of individual cells. In this study, we utilized E. coli as a model organism for separation and reception experiments. The results demonstrate that individual E. coli cells can be successfully ejected from the chip by different mode beams, leaving distinct markings and effectively recovering on the receiver (Figures 3C–E). In comparison to Gaussian beam, where energy follows a Gaussian distribution, the system with a circular vortex light spot exerts a force converging towards the center of the cell, while the system with uniformly distributed flat-top beam exerts a downward force, enabling precise separation and collection of individual cells (Figure 3B). To verify the accuracy of the receiving by different modes light beam, 100 single E. coli cells were separately ejected by using GS-LIFT, OV-LIFT, and FT-LIFT. Cells ejected by GS-LIFT showed a distribution diameter of approximately 400 micrometers on the receiver (Figure 3F), while cells ejected by FT-LIFT exhibited a distribution diameter below 50 micrometers (Figure 3G), and OV-LIFT resulted in a distribution diameter below 100 micrometers (Figure 3H). The standardized deviation of cells ejected from FT-LIFT and OV-LIFT was 8.4 and 36.4, respectively. This indicates that the FT-LIFT and OV-LIFT single-cell separation systems can achieve efficient and precise isolation and reception of single cells.

Figure 4 illustrates the ejection and reception process of E. coli under the influence of Gaussian, vortex, and flat-top beams, captured by a high-speed camera. From the ejection of cell from the aluminum film to its landing on the receiving chip, the entire process takes 250 μs, with a distance of 170 μm between the aluminum film and the receiving chip. Through high-speed camera recording, we observe the impact of different beam modes on single-cell sorting clearly. GS-LIFT ejection generates a significant offset, while OV-LIFT ejection produces a smaller offset, and FT-LIFT ejection exhibits almost no offset, achieving highly precise sorting and reception. These results provide visual support for the performance differences among various LIFT systems, further confirm the exceptional capabilities of flat-top and vortex beams sorting systems in achieving high-precision sorting and reception.

We employed S. cerevisiae and E. coli as models for cell sorting and cultivation. First, we ejected the blank area near the target cell as control group, and then the target cell was ejected. As shown in Figure 6, the single-cell ejected vortex and flat-top beams could form individual colonies on the corresponding agar plates, and for the control group, in which ejecting the area near the target, no colonies grow. The results confirms that the system can precisely eject cells at single cell level and the sorted single cells could be cultivated further without harm.

In Figure 7, we present the single-cell cultivation results of ejecting S. cerevisiae and E. coli at nine predetermined receiving positions by gaussian, vortex, and flat-top beams, respectively. Six independent repetitions in each experiment were carried to ensure the reliability of the results. For S. cerevisiae cells, the average recovery rate was 48% for Gaussian beam ejection, 85% for flat-top beam ejection, and 89% for vortex beam ejection. In the case of E. coli cells, the average recovery rate was 19% for Gaussian beam ejection, 57% for flat-top beam ejection, and 78% for vortex beam ejection. This cultivation results demonstrates a significant improvement in recovery rates of S. cerevisiae and E. coli cells ejected by vortex and flat-top beams compared to traditional gaussian beam.