Other parts
YOU WILL NEED SOME OTHER PARTS
O rings
We used standard silicone-type O-rings, which are widely available. We found these to be stronger and more flexible than the 3-D printed ones. Ours were supplied by ‘The O-Ring Store LLC’ 1847 Wilma Dr, Clarkston, WA 99403. www.theoringstore.com These are 2x60mm metric buna-N 70 O-rings model N2.00x060, and 2x75 mm, model N2.00x075.
Clock motor
There are dozens of different clock motors available on the web. They come in several sizes as well as in several formats (12 and 24 hour and 7 day).
Most are of similar dimension, (52x52x16 mm) for the body, which is what our sampler is designed around. HOwever they often differ in the length of the shaft (see more below).
The prices range from about $3 to $10 each. We are not aware of any differences in quality between these. Regrettably we could not find a higher precision clock motor for a higher price.
We purchased ours from banggood.com, see https://www.banggood.com/Black-Hands-Quartz-Clock-Movement-Kit-DIY-Clock-Kit-p-979315.html?cur_warehouse=CN. The package is labelled SKU197305-ZSVO PCK. On the website, when searching for this it comes up with model SKU 979315, (as it does on our order). The website image of ours appears to be the same although ours did not contain the hanging bracket.
Note, the back of the clock bears the imprint: “Sangtai 6168S”, however there are many different models within this brand with different shaft configurations, different shaft lengths, and types.
If clocks are being sourced from elsewhere, it is important to note a couple important variables if these differ from ours.
(1) The length of the threaded shaft; that is from the base of the clock to where the hour hand emerges. From the website it is described in imperial as 3/16 inch which is 4.76mm. The width is given as 5/16th inch = 7.93mm. There is a choice of style of hour hands that are supplied. This hand fits into the underside of the timer wheel. The 3-D design of the timer wheel is for the hour hand as shown above and on the website. If you end up with a different hour hand, then the design of the timer wheel may need to be changed accordingly.
(2) The overall height of the unit, particularly the threaded shaft and the drive mechanisms for the hands. Ours is 11.9 mm.
For our earlier model 1 (see development), which used a different motor with a more protruding minute and second hand shaft, the clock was disassembled (by undoing small clips) and some of the relevant cogs and shafts relating to these hands removed and the clock reassembled; not for those with shaky hands, but it is do-able.
It would be preferable to use a 'high-torque' motor. However, we could not find any that were high torque and had a short shaft.
For some applications a 24 hour clock or even 7 day (a cheap replacement for Burkard samplers?) could be used. Again, those we have seen have a slightly longer shaft. One example, on the banggood web site, there is a clock which can either do a single rotation in 24 hours or in 7 days, see The length of the brass shaft here is given as 5/16 inch, and the total length of this plus the shaft of the hand is ½ inch (12.7 mm). This would require a small modification of our sampler body.
We purchased half dozen different clocks for comparison. They all had a bit of play in the hour hand shaft. This play is in two dimensions: first the hour hand shaft loosely fits outside of the minute hand shaft, and secondly, there is some play as to exactly where the hour hand is pointing, that is it could point to say 1200 or to 1202 if forced, but in use it does not wobble between these positions. Presumably they are all manufactured with these tolerances. It does not affect their performance as a timepiece, it just provides a bit less certainty for our application.
2 mm threaded screw.
The sampler was designed to the the inlet jet could be moved within the glide, either so the distance of the outlet jet and the sampling surface could be adjusted or so that modified inlet jets could be fitted. The inlet is held firmly in the glide by a small XX mm grub screw, this is a standard workshop item, and if not available, can be purchased here (link). It is advised is possible to run a XXmm tap through the hole in the glide section prior to inserting the screw. For aesthetics, we have used a screw with an allen head.
The sample collection surface
We have used two types of quite different surfaces; these are electrostatic cloth and adhesive film. (the film is actually two sorts depending on applications, more later). When compared the cloth with the film in the 2016 PLoS paper, they appeared to give very similar results.
Adhesive film (for press-blotting). For press-blotting we used the adhesive side of the removable film that is used to protect the screens of iPhones/iPads.
The actual adhesive is not clear from the pictures below, as they are in Japanese.
This is produced for Daiso of Japan, (C008, display protective screen no. 59); and is made in China.
If someone has Japanese language skills, please tell us what is understood by these two labels.
O rings
We used standard silicone-type O-rings, which are widely available. We found these to be stronger and more flexible than the 3-D printed ones. Ours were supplied by ‘The O-Ring Store LLC’ 1847 Wilma Dr, Clarkston, WA 99403. www.theoringstore.com These are 2x60mm metric buna-N 70 O-rings model N2.00x060, and 2x75 mm, model N2.00x075.
Clock motor
There are dozens of different clock motors available on the web. They come in several sizes as well as in several formats (12 and 24 hour and 7 day).
Most are of similar dimension, (52x52x16 mm) for the body, which is what our sampler is designed around. HOwever they often differ in the length of the shaft (see more below).
The prices range from about $3 to $10 each. We are not aware of any differences in quality between these. Regrettably we could not find a higher precision clock motor for a higher price.
We purchased ours from banggood.com, see https://www.banggood.com/Black-Hands-Quartz-Clock-Movement-Kit-DIY-Clock-Kit-p-979315.html?cur_warehouse=CN. The package is labelled SKU197305-ZSVO PCK. On the website, when searching for this it comes up with model SKU 979315, (as it does on our order). The website image of ours appears to be the same although ours did not contain the hanging bracket.
Note, the back of the clock bears the imprint: “Sangtai 6168S”, however there are many different models within this brand with different shaft configurations, different shaft lengths, and types.
If clocks are being sourced from elsewhere, it is important to note a couple important variables if these differ from ours.
(1) The length of the threaded shaft; that is from the base of the clock to where the hour hand emerges. From the website it is described in imperial as 3/16 inch which is 4.76mm. The width is given as 5/16th inch = 7.93mm. There is a choice of style of hour hands that are supplied. This hand fits into the underside of the timer wheel. The 3-D design of the timer wheel is for the hour hand as shown above and on the website. If you end up with a different hour hand, then the design of the timer wheel may need to be changed accordingly.
(2) The overall height of the unit, particularly the threaded shaft and the drive mechanisms for the hands. Ours is 11.9 mm.
For our earlier model 1 (see development), which used a different motor with a more protruding minute and second hand shaft, the clock was disassembled (by undoing small clips) and some of the relevant cogs and shafts relating to these hands removed and the clock reassembled; not for those with shaky hands, but it is do-able.
It would be preferable to use a 'high-torque' motor. However, we could not find any that were high torque and had a short shaft.
For some applications a 24 hour clock or even 7 day (a cheap replacement for Burkard samplers?) could be used. Again, those we have seen have a slightly longer shaft. One example, on the banggood web site, there is a clock which can either do a single rotation in 24 hours or in 7 days, see The length of the brass shaft here is given as 5/16 inch, and the total length of this plus the shaft of the hand is ½ inch (12.7 mm). This would require a small modification of our sampler body.
We purchased half dozen different clocks for comparison. They all had a bit of play in the hour hand shaft. This play is in two dimensions: first the hour hand shaft loosely fits outside of the minute hand shaft, and secondly, there is some play as to exactly where the hour hand is pointing, that is it could point to say 1200 or to 1202 if forced, but in use it does not wobble between these positions. Presumably they are all manufactured with these tolerances. It does not affect their performance as a timepiece, it just provides a bit less certainty for our application.
2 mm threaded screw.
The sampler was designed to the the inlet jet could be moved within the glide, either so the distance of the outlet jet and the sampling surface could be adjusted or so that modified inlet jets could be fitted. The inlet is held firmly in the glide by a small XX mm grub screw, this is a standard workshop item, and if not available, can be purchased here (link). It is advised is possible to run a XXmm tap through the hole in the glide section prior to inserting the screw. For aesthetics, we have used a screw with an allen head.
The sample collection surface
We have used two types of quite different surfaces; these are electrostatic cloth and adhesive film. (the film is actually two sorts depending on applications, more later). When compared the cloth with the film in the 2016 PLoS paper, they appeared to give very similar results.
Adhesive film (for press-blotting). For press-blotting we used the adhesive side of the removable film that is used to protect the screens of iPhones/iPads.
The actual adhesive is not clear from the pictures below, as they are in Japanese.
This is produced for Daiso of Japan, (C008, display protective screen no. 59); and is made in China.
If someone has Japanese language skills, please tell us what is understood by these two labels.
|
|
According to the packaging description of another screen protector we used (but not for obtaining data for publication, this protective film has a silicone adhesive,
As far as we know there are no precedents for the use of this protective film as an adhesive surface for sample collection.
It arose as an intuitive leap by one of us (Damien Liu-Brennan) after we had tried unsuccessfully perform press-blotting with several other adhesives (including those cited in the literature for this purpose). This film proved suitable for both press blotting and for eluting the allergens from for ELISA assay. The film has almost no stickiness or tackiness to touch, but possess a thin soft almost gelatinous feel when closely examined by scratching with a sharp object. Obviously it binds tightly but also reversibly to the glass screens of devices.
Adhesive film for halogen assays:
Most of our previous experience has been using acrylic adhesives for collection for halogen assays. We have not done halogen with this sampler, but there is no reason not to, indeed it has some advantages over press-blotting. I will review these under the section assays.
In particular we have used Duro-tak adhesive (it now looks as though it is supplied by Henkel). Various duro-tak tapes seem to be available for other applications, but we cannot advise on any specifically suitable for halogen.
In the late 1990s we commissioned some research into a suitable adhesive to use both as an impaction surfaces and also suitable for Halogen assay. Duro-tak was identified and we had a large batch of Duro-tak adhesive tape prepared for us. We have been using these ever since.
Certainly there were other close contenders. Some of this early research is described in Razmovski, V, O'Meara, T, Hjelmroos, M, Marks, G Tovey E, Grana 37: 305-310, 1998, Adhesive tapes as capturing surfaces in Burkard sampling.
Another researcher in the USA, Dr Felix Rivera-Mariani has published several papers using the halogen assay to identify different allergenic fungi. He used ABgene PCR sealing film (Fisher Scientific, New Hampton, N.H., USA) as the adhesive to laminate to protein binding membranes. (It would appear to be part number AB 1170). See, Rivera-Mariani FE, Nazario-Jimenez S, Lopez-Malpica F, Bolanos-Rosero B. Sensitization to Airborne Ascospores, Basidiospores, and Fungal Fragments in Allergic Rhinitis and Asthmatic Subjects in San Juan, Puerto Rico. International Archives of Allergy and Immunology. 2011;155(4):322-34.
We can supply the liquid duro-tak, (have a couple of litres left) but there are restrictions on posting as it is a fire hazard. We can also supply some of the old tape. Both are free until we run out.
Electret electrostatic cloth
Our PLoS 2016 paper used electret EWE50 from Vylene, Japan as a collection surface. This is highly electrostatic and has a relatively fine texture, about the same as a thin soft bed-sheet.
We identified this particular electret around 2005 in a small research project to build nasal filters against inhaled virus (H5N1 at the time). We were looking for electrets which had very good air-flow and particle capture. We obtained samples of and tested about 100 brands/types; some of these were rescued out of N95 masks, where they are the fundamental filtering element. There were other close contenders at the time.
We purchased a 1000M roll of EWE50 and have widely used it to sample exhaled virus aerosols (using two layers), (or have used this or H&V Technostat 70+ to collect airborne allergen by air filtration) – we have several papers on this.
If we were starting this again today we would probably end up with a nano-fibre electret. These have superior capture and flow properties and it would be exciting to try these. This presumes I could get the cooperation of the companies producing them, (Hollingsworth and Vose, and 3M in USA; NAMI in Hong Kong, and others) failing that I would rescue the material from their N95 masks or their other filter applications. I know from correspondence that other scientists have not been able to get any collaboration with these companies; I guess we were lucky.
Using an electret as a collection surface for impaction has not been done before (as far as known) and yet it presents interesting properties. It is non-standard, but possibly intuitive.
Historically, the surfaces for impaction samplers were films of waxes, grease, thick oils etc. Adhesive films have also been used, see our Razmovski et al paper above. Fibre-based materials such as paper have been tested and found not very useful. Also metal and polyurethane sponges have been used (look up papers by Petros Koutrakis), with the observation that these collect smaller sized particles that conventional surfaces. As noted by Fengxia, Li et al Dec 2015 Environ Sci Pollut Res. “…Normally, the porous substrate will shift the cutoff diameter towards smaller diameters and also reduce the sharpness of the cutoff curve (Demokritou et al. 2004; Kavouras and Koutrakis 2001; Lee et al. 2005; Rao and Whitby 1978; Saarikoski et al. 2008; Sillanpaa et al. 2004). However, sometimes, the usage of porous substrates such as quartz-fiber filter (QFF) or polyurethane foam (PUF) is preferable to mitigate problems like overload, particle bounce off, and re-entrainment (Demokritou et al. 2004; Lee et al 2005; Rao and Whitby 1978 ). For the intended 24-h ambient PM sampling , especially the overload was a major concern. ...”
I guess our thoughts are summarised in our 2016 PLoS paper “…... EWE50 was chosen based on our prior selection for collecting airborne virus for PCRs [19] and because it is thin with a fine porous texture. Electret may have advantages as an impaction surface due to both its porous surface [20] and its electrostatic attraction of particles [21] however this remains to be explored. ..”
It is notable that when you use EWE50, when viewed post-collection under the microscope you can see very fine particles coating the fibres within the matrix of the material. Presumably with a fine open fibre material there is really no single planar impaction surface – so the jet to surface distance is not defined and it can be quite short, as the air will continue to flow into the matrix and circulate before exiting out again. Add to this the electrostatic properties.
Anyway, we can all speculate, and all this needs proper quantitative research.
Happy to supply EWE50 for free, cannot guarantee any properties, its 10 years old, has been kept rolled up in plastic, and would not be ‘sterile’ (never was).
Other things you may need
An airpump, capable of at least 2 L min. I have used a Casella TUFF pump, As of 17th March 2017, this is now a ‘legacy support product’, and is replaced by the Casella Apex2. In choosing, it is useful to know backpressure so maximum sampling time can be calculated. Sorry I don’t know the actual back pressure of the current model 1 or 3 samplers, have estimated it to be around 0.6 inches water (see design considerations), but don’t have the gear anymore to measure this. Using a higher flow, eg 4L/min, will of course reduce the size of the cut point, increase pressure and sample twice the volume of air as well as reduce the total time the pump will run for on a single charge.
Means of logging the location: Depending on the application, a time-based sampler may require some way of knowing the location of the wearer. We used a life-blogging Blynk LyfeShot camera running at 4 frames per minute. This produced a time-stamped video for later resolution of activities. There are several other cameras on the market: iON, Narrative, SnapCam, Vicon, Autographer, Parashoot … take your pick.
There are privacy considerations with using cameras (and people these days are much more concerned about their data and privacy). While the person wearing the device will know their location is monitored, others in the house for example may not want their images recorded. The concern also applies to people being filmed who may be unaware of this. For example, if the person doing the sampling inadvertently puts a bag containing sampler and camera on the floor of a bus and so starts up-skirt filming without realising it. Not a good look. Of course you can always just film the ground and still work out where a person is, or degrade the image by shooting through some semi opaque medium.
Alternatively or additionally would be possible to use GPS tracker, although this may struggle with location within a house. Examples include TileMate, TrackR, NutTagFind3 – there are many others.
It would be useful to also add motion detection, 3D accelerometer to know when the subject is immobile or active (and dust disturbing)
There is smart phone software that can perform most of these functions, that is take serial pictures, monitor GPS and do 3D acceleration. i am not sure whether can run multiple apps on Apple phones (unless jail broken), perhaps Android. (?) - see Figure 2 in Tovey and Ferro, Curr Allergy Asthma Rep. 2012 Oct;12(5):465-77 for an example combining 3D-movement, particles (using a laser sampler) and serial images of locations.
Some means of cutting electret and adhesive film. To use the sampler, you will need to be able to prepare strips of the sampling surface, 5mm wide (more like 4.75mm) and around 180 mm long (for model 3), which fit into the groove. This is trickier than it sounds. This will be covered more in the section on analysis, but we find you do need a selection of adhesive tapes and also a guillotine that will cut clean strips; you can try to use a scalpel blade and a ruler but its not easy.
As far as we know there are no precedents for the use of this protective film as an adhesive surface for sample collection.
It arose as an intuitive leap by one of us (Damien Liu-Brennan) after we had tried unsuccessfully perform press-blotting with several other adhesives (including those cited in the literature for this purpose). This film proved suitable for both press blotting and for eluting the allergens from for ELISA assay. The film has almost no stickiness or tackiness to touch, but possess a thin soft almost gelatinous feel when closely examined by scratching with a sharp object. Obviously it binds tightly but also reversibly to the glass screens of devices.
Adhesive film for halogen assays:
Most of our previous experience has been using acrylic adhesives for collection for halogen assays. We have not done halogen with this sampler, but there is no reason not to, indeed it has some advantages over press-blotting. I will review these under the section assays.
In particular we have used Duro-tak adhesive (it now looks as though it is supplied by Henkel). Various duro-tak tapes seem to be available for other applications, but we cannot advise on any specifically suitable for halogen.
In the late 1990s we commissioned some research into a suitable adhesive to use both as an impaction surfaces and also suitable for Halogen assay. Duro-tak was identified and we had a large batch of Duro-tak adhesive tape prepared for us. We have been using these ever since.
Certainly there were other close contenders. Some of this early research is described in Razmovski, V, O'Meara, T, Hjelmroos, M, Marks, G Tovey E, Grana 37: 305-310, 1998, Adhesive tapes as capturing surfaces in Burkard sampling.
Another researcher in the USA, Dr Felix Rivera-Mariani has published several papers using the halogen assay to identify different allergenic fungi. He used ABgene PCR sealing film (Fisher Scientific, New Hampton, N.H., USA) as the adhesive to laminate to protein binding membranes. (It would appear to be part number AB 1170). See, Rivera-Mariani FE, Nazario-Jimenez S, Lopez-Malpica F, Bolanos-Rosero B. Sensitization to Airborne Ascospores, Basidiospores, and Fungal Fragments in Allergic Rhinitis and Asthmatic Subjects in San Juan, Puerto Rico. International Archives of Allergy and Immunology. 2011;155(4):322-34.
We can supply the liquid duro-tak, (have a couple of litres left) but there are restrictions on posting as it is a fire hazard. We can also supply some of the old tape. Both are free until we run out.
Electret electrostatic cloth
Our PLoS 2016 paper used electret EWE50 from Vylene, Japan as a collection surface. This is highly electrostatic and has a relatively fine texture, about the same as a thin soft bed-sheet.
We identified this particular electret around 2005 in a small research project to build nasal filters against inhaled virus (H5N1 at the time). We were looking for electrets which had very good air-flow and particle capture. We obtained samples of and tested about 100 brands/types; some of these were rescued out of N95 masks, where they are the fundamental filtering element. There were other close contenders at the time.
We purchased a 1000M roll of EWE50 and have widely used it to sample exhaled virus aerosols (using two layers), (or have used this or H&V Technostat 70+ to collect airborne allergen by air filtration) – we have several papers on this.
If we were starting this again today we would probably end up with a nano-fibre electret. These have superior capture and flow properties and it would be exciting to try these. This presumes I could get the cooperation of the companies producing them, (Hollingsworth and Vose, and 3M in USA; NAMI in Hong Kong, and others) failing that I would rescue the material from their N95 masks or their other filter applications. I know from correspondence that other scientists have not been able to get any collaboration with these companies; I guess we were lucky.
Using an electret as a collection surface for impaction has not been done before (as far as known) and yet it presents interesting properties. It is non-standard, but possibly intuitive.
Historically, the surfaces for impaction samplers were films of waxes, grease, thick oils etc. Adhesive films have also been used, see our Razmovski et al paper above. Fibre-based materials such as paper have been tested and found not very useful. Also metal and polyurethane sponges have been used (look up papers by Petros Koutrakis), with the observation that these collect smaller sized particles that conventional surfaces. As noted by Fengxia, Li et al Dec 2015 Environ Sci Pollut Res. “…Normally, the porous substrate will shift the cutoff diameter towards smaller diameters and also reduce the sharpness of the cutoff curve (Demokritou et al. 2004; Kavouras and Koutrakis 2001; Lee et al. 2005; Rao and Whitby 1978; Saarikoski et al. 2008; Sillanpaa et al. 2004). However, sometimes, the usage of porous substrates such as quartz-fiber filter (QFF) or polyurethane foam (PUF) is preferable to mitigate problems like overload, particle bounce off, and re-entrainment (Demokritou et al. 2004; Lee et al 2005; Rao and Whitby 1978 ). For the intended 24-h ambient PM sampling , especially the overload was a major concern. ...”
I guess our thoughts are summarised in our 2016 PLoS paper “…... EWE50 was chosen based on our prior selection for collecting airborne virus for PCRs [19] and because it is thin with a fine porous texture. Electret may have advantages as an impaction surface due to both its porous surface [20] and its electrostatic attraction of particles [21] however this remains to be explored. ..”
It is notable that when you use EWE50, when viewed post-collection under the microscope you can see very fine particles coating the fibres within the matrix of the material. Presumably with a fine open fibre material there is really no single planar impaction surface – so the jet to surface distance is not defined and it can be quite short, as the air will continue to flow into the matrix and circulate before exiting out again. Add to this the electrostatic properties.
Anyway, we can all speculate, and all this needs proper quantitative research.
Happy to supply EWE50 for free, cannot guarantee any properties, its 10 years old, has been kept rolled up in plastic, and would not be ‘sterile’ (never was).
Other things you may need
An airpump, capable of at least 2 L min. I have used a Casella TUFF pump, As of 17th March 2017, this is now a ‘legacy support product’, and is replaced by the Casella Apex2. In choosing, it is useful to know backpressure so maximum sampling time can be calculated. Sorry I don’t know the actual back pressure of the current model 1 or 3 samplers, have estimated it to be around 0.6 inches water (see design considerations), but don’t have the gear anymore to measure this. Using a higher flow, eg 4L/min, will of course reduce the size of the cut point, increase pressure and sample twice the volume of air as well as reduce the total time the pump will run for on a single charge.
Means of logging the location: Depending on the application, a time-based sampler may require some way of knowing the location of the wearer. We used a life-blogging Blynk LyfeShot camera running at 4 frames per minute. This produced a time-stamped video for later resolution of activities. There are several other cameras on the market: iON, Narrative, SnapCam, Vicon, Autographer, Parashoot … take your pick.
There are privacy considerations with using cameras (and people these days are much more concerned about their data and privacy). While the person wearing the device will know their location is monitored, others in the house for example may not want their images recorded. The concern also applies to people being filmed who may be unaware of this. For example, if the person doing the sampling inadvertently puts a bag containing sampler and camera on the floor of a bus and so starts up-skirt filming without realising it. Not a good look. Of course you can always just film the ground and still work out where a person is, or degrade the image by shooting through some semi opaque medium.
Alternatively or additionally would be possible to use GPS tracker, although this may struggle with location within a house. Examples include TileMate, TrackR, NutTagFind3 – there are many others.
It would be useful to also add motion detection, 3D accelerometer to know when the subject is immobile or active (and dust disturbing)
There is smart phone software that can perform most of these functions, that is take serial pictures, monitor GPS and do 3D acceleration. i am not sure whether can run multiple apps on Apple phones (unless jail broken), perhaps Android. (?) - see Figure 2 in Tovey and Ferro, Curr Allergy Asthma Rep. 2012 Oct;12(5):465-77 for an example combining 3D-movement, particles (using a laser sampler) and serial images of locations.
Some means of cutting electret and adhesive film. To use the sampler, you will need to be able to prepare strips of the sampling surface, 5mm wide (more like 4.75mm) and around 180 mm long (for model 3), which fit into the groove. This is trickier than it sounds. This will be covered more in the section on analysis, but we find you do need a selection of adhesive tapes and also a guillotine that will cut clean strips; you can try to use a scalpel blade and a ruler but its not easy.