Model 1 was comprised of two 3-D printed components: a top section modelled to fit onto the top of the body of an alarm clock replacing the cover glass and a sampling wheel designed to fit onto the hour hand of the clock. The inlet orifice was later shimmed with ~0.1 mm aluminium sheet to reduce its width to ~0.5 mm and to smooth the inlet surface. The clock body was sealed airtight with adhesive tape.
The device was used in the 2016 PLoS paper, see here.
The files can be downloaded here.
Some minor modifications have been made to these, to reflect changes made to the sampler wheel and the inlet slot.
The sampler was worn in the day across the chest, along with a small life-blogging camera, with the pump carried in a small backpack. When a 12 hour clock was used, the subjects changed the sampler wheel at this time, this was not required when the 24 hour clock was used. The start and stop positions of sampling were indicated by puffing a small quantity of green and red (respectively) powder into the orifice.
Figure (left) from the PLoS paper, showing (a) the sampler as worn, along with camera (arrow), (b) closeup of the sampler showing inlet (arrow), (c) closeup of the sampling surface (iPad film) showing a heavily loaded with dust samples. The green (start) and red (stop) makers are shown with heavy arrows, and the bands of dust at different intervals shown with smaller arrows. The numbers are to help with performing the sampling, so that the approximate time of starting and current time when sampling can be viewed through the small window in surface of the unit.
A couple of samples collected onto electret during journeys in the New York Underground. The dark bars correspond to episodes of high exposure to particulates in underground stations.
Model 2 was originally designed as a successor to Model 1, but only a couple of these were actually built. Initially we could not decide whether samples should be collected on the top of the sampler wheel, (as in model 1) or should be collected on the edge of the wheel, providing the collection surface as a long strip.
In the couple of prototypes we had printed, there were numerous printing problems, including poor resolution of the inlet orifices (it depends in which orientation the build occurs) and some warping of the body, so that the parts did not meet and seal properly.
In the end it seemed that we should make a decision about which orifice orientation should be used, and we chose, impaction onto the edge of a wheel, providing a strip, and we realised a circular body would be easier to seal, hence model 3.
As there may be useful aspects to this design, the files for model 2 are provided here as a combination showing it all together, here as STEP files, and here as STL files.