Harmonic acoustics for dynamic and selective particle manipulation

Abstract

Precise and selective manipulation of colloids and biological cells has long been motivated by applications in materials science, physics and the life sciences. Here we introduce our harmonic acoustics for a non-contact, dynamic, selective (HANDS) particle manipulation platform, which enables the reversible assembly of colloidal crystals or cells via the modulation of acoustic trapping positions with subwavelength resolution. We compose Fourier-synthesized harmonic waves to create soft acoustic lattices and colloidal crystals without using surface treatment or modifying their material properties. We have achieved active control of the lattice constant to dynamically modulate the interparticle distance in a high-throughput (>100 pairs), precise, selective and reversible manner. Furthermore, we apply this HANDS platform to quantify the intercellular adhesion forces among various cancer cell lines. Our biocompatible HANDS platform provides a highly versatile particle manipulation method that can handle soft matter and measure the interaction forces between living cells with high sensitivity.

Fig. 1 ∣. Fourier synthesis of harmonic acoustic waves of HANDS to create soft flexible lattices for colloidal crystals or cell–cell pairing and separation.

a, Schematic of HANDS manipulation with Fourier-synthesized harmonic acoustic waves. b, Top, a colloidal crystal with controllable particle numbers and conformation can be assembled. Bottom, large area patterning with a tuneable lattice constant can be generated for colloidal clusters or single cells. c, The modulation of frequency and amplitude of harmonic waves can generate soft lattices and reconfigurable colloidal crystal or tuneable patterning of single particles or cells. d, Fluorescent imaging showing selective pairing of two U937 cells among six cells by localized modulation of the intercellular distance. The positions of the cells are indicated by the fluorescent intensity profile (averaged from five cell groups). Scale bars, 10 μm. e, Comparison of two patterned colloidal clusters with equal trapping spacing and tuneable trapping spacings, as analytically simulated and experimentally generated using HANDS manipulation. Colloidal clusters are formed with 2 μm fluorescent polystyrene particles in each trapping acoustic wells. Scale bars, 20 μm.

Fig. 2 ∣. Creation of colloid crystal monolayers with cluster and spin dynamics studies via HANDS manipulation.

a, Schematic of colloidal clusters being trapped in acoustic wells by applying SAWs in the x and y directions. b, Illustration of the generation of colloidal crystal monolayer by vertical focusing of particles with applied harmonic SAWs. c, Fluorescent imaging of colloidal crystal monolayers using 2 μm polystyrene particles formed by HANDS manipulation. Scale bar, 5 μm. d, Illustration of the spin directions, controlled by the phase difference (Δϕ = ϕ_x − ϕ_y) between X-SAWs and Y-SAWs. Scale bar, 5 μm. e, Characterization of the spin speed as a function of the phase difference at different excitation signal amplitudes ranging from 4 to 8 V. f, Comparison of the spin speed of colloidal crystal monolayers with different particle numbers and configurations.

Fig. 3 ∣. HANDS for manipulation of soft matter and living cells for precision quantitative measurements.

a,b, Analytical simulations and experimental results for single-colloid trapping and pairing with connected (a) and isolated (b) acoustic wells for colloidal particles (9.51 μm). The green and red arrows indicate the directions of the acoustic radiation forces. Scale bar, 10 μm. c, Analytical simulations and experimental results with synthesized harmonic acoustic wells for the generation of customized colloidal patterns of 2 μm polystyrene particles, with the letters ‘O’, ‘D’ and ‘K’. Scale bar, 20 μm. d, Separation forces (averaged from six particles) for the colloidal particles with various excitation signal amplitudes ranging from 4 to 8 V. e, Quantitative characterization of the peak separation forces (F) and the corresponding separation velocities (V) with varied peak-to-peak voltage (U) for the 9.51-μm-diameter polystyrene particles (averaged from six particles). From a quadratic fit of the peak separation forces (e), the force quadratic constant q_F is 0.91 ± 0.05 pN V⁻² (fit value ± s.d.). A quadratic fit of the separation velocities gives a velocity quadratic constant q_V of 0.25 ± 0.04 mm s⁻¹ V⁻² (fit value ± s.d.). f, Reversible pairing of U937 cells along both the x and y directions. The blue arrows indicate the direction of the acoustic radiation forces during reversible pairing. Scale bar, 10 μm.

Fig. 4 ∣. Reversible cell–cell pairings via HANDS for the quantification of intercellular adhesion strength in different cell lines.

a, Illustration of harmonic acoustic wells generated by the HANDS platform for tuning of particle or cell spacings. Single particles or cells can be reversibly paired or separated with HANDS manipulation. b, Representative normalized traces of cell–cell separation distances for M0 and M1 THP-1 cells. c, Dependence of the lifetime adhesion frequency with respect to the contact signal duration for M0 and M1 THP-1 cells. d–g, Histograms of cell adhesion lifetimes for MCF-7 (CytoD⁻) (d), MCF-7 (CytoD⁺) (e), MDA-MB-231 (f) and HEK293T (g) cells. The black lines are the Gaussian fits for each cell line, and the dashed blue lines are the Gaussian fits for MCF-7 CytoD⁻ cells. t_short and t_long represent the average values of short and long lifetimes from the histograms with bimodal distribution, as determined by a double Gaussian kernel function. h, Fraction of long lifetime subpopulations for MCF-7 (CytoD⁻), MCF-7 (CytoD⁺), MDA-MB-231 and HEK293T cells.