Supplementary Figure 7: Mass accumulation rate resolution of the large-channel devices.

Measuring a mixture of plastic microparticles to determine mass accumulation rate resolution on a large-channel serial mass sensor array. (A) We measured a mixture of 4, 6, 7, 8, 9, 10 and 12 μm polystyrene beads (Duke Standards, NIST traceable, Thermo Scientific) at 37 C in 0.01% Tween-20 in water. Sensors were calibrated by linearly rescaling their raw frequency signals such that the 7 μm bead modal mass is the expected buoyant mass (10.15 pg). (B) Across all sizes and sensors, particle buoyant masses match the expected buoyant masses (dashed lines), verifying that the sensors are linear over this size range. (C) 4 μm particles have the lowest size variability (in pg) of these beads according to manufacturer’s datasheet, and therefore their distribution’s width is a reasonable upper bound on the sensor error. Typical sensor root-mean-square-error is on the order of 0.05 pg. (D) Histogram of mass accumulation rates (errors, as particles are not growing) of 85 single particles for which at least 10 sensors could be linked together. Mass accumulation rates were calculated excluding data from the first sensor, which displayed much higher noise than the other sensors. Dashed line shows estimated mass accumulation rate distribution assuming t = 1.4 minutes, k = 11, and s = 0.05 pg, showing good agreement between this approximation for mass accumulation rate error and the observed mass accumulation rate error distribution.