Page 20 - manuscript_ijb05583

intensity, the peak intensity value is validated. Depending on whether it’s below or above a

threshold, the event gets flagged for sorting (green flag) or discarded (red flag).

The values for classifying microscaffolds as either intact or debris were derived from Figure 5
c, whereas the minimum distance between microscaffolds was calculated as the sum of events

within the sorting cascade without the time required to move the microscope stage between the
wells. For a liquid pressure of 20 mbar, this time was determined to be 900 ms, therefore a time

threshold of 1 s between events was chosen.

Once microscaffolds were flagged sortable (green), a sorting sequence was started. This

sequence consisted of an initial waiting period that would allow the BB to arrive at the

bifurcation (values chosen from Figure 4 l). Just before actuating the valve, the flag of the BB
event was checked one last time. This double-flagging system was introduced to abort sorting

events in the last moment, due to a subsequent BB following too closely. This double-flagging
greatly reduced the amount of microscaffolds stuck in the PDMS valve. After a BB was

secluded in the ejection channel, it was dispensed into a cell culture well plate beneath using

pressurized air. Repeatable droplet ejection was possible at 60 mbar for 500 ms. Shorter
ejection times did not fully clear the channel, whereas higher pressures resulted in sideways

spraying of the droplet.

The sorting efficiency of the system was determined as percentage of wells containing exactly

one microscaffold and amounted to 95%, measured on a total of 250 microscaffolds. The

system’s capability to effectively distinguish between intact microscaffolds and debris was
assessed by sorting a mixed sample containing intact BBs and buckyball halves simulating

debris. Next, a sample size of 96 sorting events emulating the production of a fully loaded 96-
well microtiter plate was analyzed with only intact microscaffolds being dispensed, leading to

100% sorting efficiency. Once the sorting efficiency was determined for 96-well plates, 384-
well plates and 1536-well plates were sorted to increase the amount of sorted microscaffolds

while keeping the same plate form factor (Figure 7 a). The process from event detection to

scaffold ejection and moving to the next microwell took approx. 2.5 s. Depending on the
microscaffold concentration in the reservoir, sorting a full 96-well plate typically took between

12 and 15 minutes, whereas 384-well plates required between 50 – 60 minutes. Microscaffold
reservoirs were loaded at 400 BB per 50 mL of liquid. An even distribution of the BBs in the

reservoir was crucial to keep the concentration stable. This was especially important when

19