Fig. 4: Positional accuracy.
a, b Assembled 3 × 3 grids of 1 µm PS spheres. The red circles coincide with the centroid locations of the observed spheres, while the yellow circles correspond to an idealized grid with uniform spacings. The grid in (a) has σMAE = 43 nm, while the grid in (b) has σMAE = 132 nm. The grid in (a) was fabricated with application of the calibration procedure, while grid (b) was fabricated in the absence of calibration. c Placement success rate as a function of silane concentration, using the silane molecular weight (MW) of 3.4 kDa. d–f Mean absolute positional error as a function of (d) silane concentration (MW = 3.4 kDa), (e) silane molecular weight (concentration = 25 mg/mL), and (f) trapping laser power (MW = 3.4 kDa, concentration = 20 mg/mL). Error bars correspond to the standard error of the mean. d Lower silane concentrations lead to increased positional errors, as the reduced surface density of binding groups causes objects to shift upon placement. Data shown in blue are obtained using the calibration process, while the red × data point is obtained without calibration, demonstrating the importance of recalibrating the stage before object placement. e Molecular weights between 600 Da and 5000 Da do not show any appreciable effect on positional errors. f Reduced laser powers correspond to large positional errors, as the object has a larger mean displacement due to Brownian motion in the weak optical trap.
