Novel thermoplastic microvalves based on an elastomeric cyclic olefin copolymer

Microfluidic systems combine multiple processing steps and components to perform complex assays in an autonomous fashion. To enable the integration of several bio-analytical processing steps into a single system, valving is used as a component that directs fluids and controls introduction of sample and reagents. While elastomer polydimethylsiloxane has been the material of choice for valving, it does not scale well to accommodate disposable integrated systems where inexpensive and fast production is needed. As an alternative to polydimethylsiloxane, we introduce a membrane made of thermoplastic elastomeric cyclic olefin copolymer (eCOC), that displays unique attributes for the fabrication of reliable valving. The eCOC membrane can be extruded or injection molded to allow for high scale production of inexpensive valves. Normally hydrophobic, eCOC can be activated with UV/ozone to produce a stable hydrophilic monolayer. Valves are assembled following in situ UV/ozone activation of eCOC membrane and thermoplastic valve seat and bonded by lamination at room temperature. eCOC formed strong bonding with polycarbonate (PC) and polyethylene terephthalate glycol (PETG) able to hold high fluidic pressures of 75 kPa and 350 kPa, respectively. We characterized the eCOC valves with mechanical and pneumatic actuation and found the valves could be reproducibly actuated >50 times without failure. Finally, an integrated system with eCOC valves was employed to detect minimal residual disease (MRD) from a blood sample of a pediatric acute lymphoblastic leukemia (ALL) patient. The two module integrated system evaluated MRD by affinity-selecting CD19(+) cells and enumerating leukemia cells via immunophenotyping with ALL-specific markers.


Milling Parameters
Individual test valve structures were milled using a desktop CNC milling machine (Nomad3, Carbine 3D, Torrance, CA).Three milling bits (1/16", 1/32", and 500 µm) were utilized to mill the various microscale features for both the pneumatic and mechanical actuation test valves.Microchannels were milled using the 1/32" bit at a plunge rate of 200 mm/min, a feed rate of 250 mm/min, and 3,500 RPM.The inlet and outlet for PEEK tubing connection were milled with the 1/32" bit using the same milling conditions.The 1/32" bit was used so that the inlet/outlet would be of equal diameter to that of the PEEK tubing with an outer diameter of 1/32".The epoxy used to attach the tubing was thoroughly and vigorously mixed by hand and allowed to sit for 2 -3 min to increase in viscosity so that the epoxy would not flow into the fluidic microchannel.A small bead of epoxy was applied to the end of the tubing before being inserted into the inlet and outlet hole.The 500 µm bit was used to make the through hole in the center of the mechanical actuation test valves at a plunge rate of 150 mm/min, a feed rate of 250 mm/min, and bit spin speed of 9,000 RPM.The solenoid attachment holes (diameter = 2 mm) were milled using the 1/16" bit at a plunge rate of 250 mm/min, a feed rate of 350 mm/min, and 9,000 RPM.Finally, the 1/16" bit was used to mill out the final test valves using a plunge rate of 450 mm/min, a feed rate of 1,500 mm/min, and 15,000 RPM.

Cell Isolation Module and Imaging Module Preparation
The PC linker and mAb immobilization to a thermoplastic surface has been previously reported by our group. 1Briefly, COP cell selection modules were UV/O 3 activated (22 mW/cm 2 , 13 min) and incubated with an EDC (20 mg/mL) / NHS (2 mg/mL) solution for 25 min at room temperature in dry ACN.The solution was removed from the module by introducing air, and then the PC linker (1 mM) dissolved in dry ACN and TEA (2 mM) was infused into the module.After 2 h of incubation at room temperature, the solution was replaced by passing air through the module.Then, 100 mM Tris (pH 7.4) was introduced into the module and incubated for 30 min at room temperature.
After incubation, the EDC/NHS solution was replaced with the mAb (anti-CD7 or anti-CD19, 1 mg/mL) dissolved in PBS (pH 7.4) and incubated overnight at 4°C.Prior to sample infusion, the module was rinsed with 2 mL of 0.5% BSA/PBS at 50 µL/min to remove unbound mAbs and to minimize non-specific binding.
Imaging modules made from PDMS (Sylgard-184, Dow Corning) were cast against a SU-8 positive relief with the resulting PDMS structures consisting of traps and microchannels. 2 A 10:1 ratio of base:curing agent of PDMS was used and was mixed in a large weigh-boat for 5 min prior to pouring onto the relief.After pouring the PDMS onto the relief structures, it was placed into a vacuum chamber for 1 hr.The PDMS was then baked in a convection oven at 75°C for 5 hr.After curing, individual devices were cut from the relief and inlet/outlet holes were punched using a 1.25 mm biopsy punch, then subsequently washed with 10% Micro-90 solution with light abrasion (toothbrush under running water), and finally rinsed with IPA then dried using compressed air.
Devices were then cleaned and dried in a 60°C oven for 20 min prior to bonding.Bonding was performed using glass cover plates that were first cleaned with IPA.To bond, both the cleaned device and cover plate were placed into an O 2 plasma chamber (90 W, 25 sccm O 2 ) for 1 min (Diener Electronic, GmbH & Co KG; Germany) -after 1 min exposure, the cover plate was lightly pressed against the device and irreversible bonding was complete.

Hydrophobic Recovery of PDMS
Sessile water contact angle (WCA) measurements were performed on PDMS (Standard Rubber Products Co., HT6240) and eCOC (TOPAS Polyplastics) after various times following O 2 plasma activation (1 min, 20 sscm O 2 gas flow, 50 W).Within 72 h, PDMS had a WCA of 96° ±5.4°, but the eCOC remained stable with an average of 73° ±2.5° over the same time period as well as

WCA of Thermoplastics Exposed to UV/O 3 Radiation
Sessile WCA measurements were performed on pristine COC, COP, PC, PETG, and PMMA.
UV/O 3 activation was performed through eCOC and quartz plates (1 mm thickness) placed between the eCOC and the appropriate plastic (see Fig. S2a), as well as exposing the plastic surface directly (Fig. S2b).The graphs in Fig. S2c shows the resulting WCAs before and after UV/O 3 exposure and with/without the eCOC/quartz plate.Interestingly, PETG showed a significant reduction in its WCA whether the eCOC/quartz cover was in place or not.A more thorough description of this data can be found in the main text.

Surface Characterization of eCOC films
All eCOC films became more hydrophilic after UV/O 3 exposure as noted by decreases in their water contact angles.There was no significant difference in the water contact angles between the supplier films (TOPAS Polyplastics and Roehm) or between the extruded films (all were in the range of 60 -70°).In addition, ATR-FTIR measurements did not show any noticeable difference in FTIR bands following 10 min UV/O 3 activation between different films.

UV-Vis Spectra of Pristine eCOC
UV-Vis spectra were taken for pristine eCOC at different thicknesses (130, 150, 280, and 350 µm) as compared to Fig. 3(a) in the main manuscript, which were taken following UV/O 3 activation.As the thickness of the eCOC increased, the UV transmittance decreased when monitored at 254 nm, which is due to an increase in the effective pathlength.This was substantiated by the fact that careful inspection of the spectra shown in Fig. S4 did not show any new bands in the spectra as the plastic thickness was changed.

ATR-FTIR of PETG
ATR-FTIR results did not show any significant differences between pristine and 10 min UV/O 3 activated PETG, because PETG contains many functional groups in its pristine state.

Photodegradation Mechanism of PETG
Grossetete et al. studied the photodegradation of PETG under both polychromatic irradiation and monochromatic irradiation. 6The PETG samples were artificially weathered for 20 -150 h and changes to the surface chemistry were observed via UV, emission, and IR spectra.They found that the aromatic chromophores in PETG strongly absorb radiation <350 nm inducing a photolytic process as proposed in Fig. S6.Polychromatic and monochromatic radiation resulted in the creation of similar photoproducts, but with a difference in the distribution profile of the resulting photoproducts.Photoproducts produced with 254 nm light remain in the first 5 µm of the substrate surface with longer wavelengths reaching 50 µm into the substrate.We hypothesize that it is these photoproducts, including a number of different highly reactive radicals, that can react with eCOC polymer chains that also generate photolytically induced radicals albeit at a lower abundancy generating cross linked polymer chains that facilitate strong bonding without requiring heat.The heating induces polymer chain motion that allows cover plate/substrate bonding in other cases.

SUP-B15 Cell Isolation using SMART-Chip
Before processing the clinical blood sample, we evaluated the SMART-Chip using SUP-B15 cell line, which is a model for B-ALL circulating leukemia cells.The recovery of these cells was found to be ~57 ±12% at volume flow rate of whole blood equal to 25 µL/min with 87% release efficiency from the cell isolation chip.Spiked SUP-B15 into 500 μL PBS were retained on the imaging chip at an efficiency of 85.4%.The staining protocol for the SUP-B15 cell line followed these steps: (i) Staining for surface antigens (CD10, CD34; 30 min); (ii) fixation with 2% buffered PFA (10 min); (iii) permeabilization with 0.1% Triton X-100 (2 min); and (iv) staining for TdT and DAPI (30 min).
We successfully implemented the staining protocol to immunophenotype SUP-B15 cells trapped within the imaging module as shown in

5 Figure S1 .
Figure S1.WCA measurements of PDMS and eCOC before and after surface activation (1 min O 2 plasma at 20 sscm O 2 gas flow and 50 W) over the course of 72 h.

Figure S2 .
Figure S2.(a) Diagram of how the tested plastic was exposed to UV/O 3 by first passing through a layer of eCOC and quartz.(b) Diagram of a particular plastic directly exposed to UV/O 3 .(c) WCA measurements for pristine COC, COP, PC, PETG, and PMMA and after each plastic was exposed to UV/O 3 using the arrangement shown in (a) or (b).

Figure S6 .
Figure S6.Scheme of chemical reactions of PETG caused by UV/O 3 exposure [adapted from Grossetête et al.].6 Fig. S7.SUP-B15 cells are nucleated and should be positive for TdT and CD10 as shown in Fig. S7.

Figure S7 .
Figure S7.SUP-B15 cell staining performed on the SMART-Chip.DAPI, TdT, and CD10 signals of SUP-B15 cells trapped at the containment pores of the imaging chip.