Hou, Linlin Et Al - Artificial Microfluidic Skin for in vitro perspiration simulation and testing

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
 103 views
of 9

Please download to get full document.

View again

Description
To expedite development of any skin wearable material, product, or device, an artificial perspiration (sweat) simulator can provide improved ease, cost, control, flexibility, and reproducibility in comparison to human or animal tests. Reported here is a human perspiration mimicking device including microreplicated skin-texture. A bottom 0.2 μm track etched polycarbonate membrane layer provides flow-rate control while a top photo-curable layer provides skin-like features such as sweat pore density, hydrophobicity, and wetting hysteresis. Key capabilities of this sweat simulator include: constant ‘sweat’ rate density without bubble-point variation even down to [similar]1 L h−1 m−2; replication of the 2 pores mm−2 pore-density and the [similar]50 μm texture of human skin; simple gravity-fed flow control; low-cost and disposable construction. Lab Chip, 2013,13, 1868-1875 DOI: 10.1039/C3LC41231H
Share
Tags
Transcript
  OFC COVERSCANTO FIT INTOTHIS BOX www.rsc.org/loc Registered Charity Number 207890 Featuring work ro m the groups o Pro. JasonHeikeneld and Pro. Ian Papautsky o the OhioCenter or Microfuidic Innovation at the University o Cincinnati, and Dr. Rajesh Naik and Dr. Joshua Hagen o the U.S. Air Force Research Laboratory. Title: Articial Microfuidic Skin or In Vitro Perspiration Si m ulationand Testing  To expedite development o any skin wearable material, product,or device, an artifcial perspiration (sweat) simulator can provideimproved ease, cost, control, exibility, and reproducibility incomparison to human or animal tests. An artifcial skin is presentedor in-vitro perspiration tests, capable o skin-like texture, wetting,sweat pore-density, and sweat rates. As eatured in: See Jason Heikeneld et al.,   Lab Chip, 2013, 13 , 1868.  Cite this: Lab Chip , 2013, 13 , 1868 Artificial microfluidic skin for  in vitro perspirationsimulation and testing Received 6th November 2012,Accepted 18th March 2013 DOI: 10.1039/c3lc41231h www.rsc.org/loc Linlin Hou, a Joshua Hagen, b Xiao Wang, a Ian Papautsky, a Rajesh Naik, c Nancy Kelley-Loughnane b and Jason Heikenfeld* a To expedite development of any skin wearable material, product, or device, an artificial perspiration(sweat) simulator can provide improved ease, cost, control, flexibility, and reproducibility in comparison tohuman or animal tests. Reported here is a human perspiration mimicking device including microreplicatedskin-texture. A bottom 0.2 m m track etched polycarbonate membrane layer provides flow-rate controlwhile a top photo-curable layer provides skin-like features such as sweat pore density, hydrophobicity, andwetting hysteresis. Key capabilities of this sweat simulator include: constant ‘sweat’ rate density withoutbubble-point variation even down to y 1 L h 2 1 m 2 2 ; replication of the 2 pores mm 2 2 pore-density and the y 50 m m texture of human skin; simple gravity-fed flow control; low-cost and disposable construction. Introduction Skin is the largest organ in humans and in addition toproviding protection, sensation, and insulation; it is also ableto regulate body temperature through perspiration (sweat).Sweat is dominated by eccrine glands which number severalmillion over most of the body surface, each gland having dimensions that are microfluidic in nature (several mm long,and 10 9 s of  m m in diameter). 1 Interest continues to grow in thephysiology of sweat, for a very diverse set of applications: (1)bioactivity and clogging related to cosmetics and other topicalproducts/medicines; 2–4 (2) textile, clothing, and personal careproduct design, 5 particularly those requiring a high sweat  venting rate; 6 (3) drug testing; 7–10 (4) a newly found and richsource of disease and health biomarkers; 11,12 (5) wearablehealth sensors. 13–16 However, human subject testing for thedevelopment of these applications is time and cost intensive.Furthermore, in such testing there is huge variance in thesweat rate density and chemical/molecular composition. 17 There are many forms of  in vitro 18 or artificial substitutes fortesting with other types of body fluids, but remarkably there isa lack of a device which can accurately mimic humansweating. We speculate that this lack of a sweat simulationdevice can be traced to three technical challenges: (1) sweat rate densities are incredibly slow ( e.g. 0.75 L h 2 1 m 2 2 forrunners 17 ) compared to the flow rate densities typically generated by commercial porous membranes ( e.g. 100s to1000s of L h 2 1 m 2 2 for track etched membranes 19,20 ); (2)commercially available membranes do not provide the surfaceenergy, pore density, nor texture representative of wetting onhuman skin; (3) membranes suffer from bubble-point thresh-holding of flow which at low flow rates causes a very smallpercentage of pores to dominate the entire flow of fluid. We have created the first artificial microfluidic skin for invitro sweat simulation and testing (Fig. 1). In addition to utility for cosmetic, textile, medical, and other applications, ourmotivation also stems from the need for rapid and repeatabletest protocols for our own burgeoning work in wearable sweat biomarker sensors. 21–23 Furthermore, several of the sweat simulators described herein have been provided to an industry partner and successfully used for in vitro testing of a new hydration monitoring device. The sweat simulator employs asimple bi-layer membrane design to resolve all drawbacksassociated with use of commercial membranes. As shown inFig. 1b, a first layer creates a pressure drop and therefore aconstant sweat flow (comparable to the secretory portion of asweat gland shown in Fig. 1a), and second top layer providesthe requisite pore density and skin-replicated wetting surface.Characterization of the sweat simulator also shows agreement  with a simple theoretical model, and therefore the sweat simulator can be rapidly and predictably redesigned/adaptedfor multiple body surface areas, sweat rate densities, andsolute compositions. Device design and theoretical model  As stated in the introduction, our key design requirementsinclude replication of skin texture, sweat pore density, skin-like surface wettability, and the most challenging requirement  a School of Electronic and Computing Systems, University of Cincinnati, OH 45221,USA. E-mail: heikenjc@ucmail.uc.edu b  Air Force Research Laboratory, 711th Human Performance Wing, Wright Patterson AFB, OH 45433, USA c  Air Force Research Laboratory, Materials and Manufacturing Directorate, Wright  Patterson AFB, OH 45433, USA Lab on a Chip PAPER  1868 | Lab Chip , 2013, 13 , 1868–1875 This journal is ß The Royal Society of Chemistry 2013    D  o  w  n   l  o  a   d  e   d   b  y   U   N   I   V   E   R   S   I   T   Y   O   F   C   I   N   C   I   N   N   A   T   I  o  n   2   6   /   0   4   /   2   0   1   3   2   3  :   2   1  :   5   7 .   P  u   b   l   i  s   h  e   d  o  n   1   8   M  a  r  c   h   2   0   1   3  o  n   h   t   t  p  :   /   /  p  u   b  s .  r  s  c .  o  r  g   |   d  o   i  :   1   0 .   1   0   3   9   /   C   3   L   C   4   1   2   3   1   H  of uniform flow rate among all pores even at the very low flow rates associated with human sweating. The incorrect choice of a single-layer design (Fig. 2a) is first discussed. Any set of pores with different diameters (  D S1 . D S2 ) will exhibit different Laplace pressures (  P  L1 = 4 c cos( h )/  D 1 ,  P  L2 = 4 c cos( h )/  D 2 ). As aresult, the larger diameter pore will have less Laplace pressureand will first activate with fluid flow ( Q 1 ). Under low flow rateconditions flow resistance is small and the larger diameterpore will continue to provide 100% of the fluid flow as flow from the smaller diameter pore is prevented ( Q 2 y 0) by itshigher Laplace pressure. Conversely, under high flow rateconditions, different Laplace pressures will be less consequen-tial as the high pressure associated with high-flow rate willdominate over Laplace pressure, and activate all poressimultaneously regardless of diameter. This higher-pressure/flow-rate test is similar to the ‘bubble point’ test for amembrane, but is irrelevant to the very low flow rate densitiesassociated with sweating ( , 1 L h 2 1 m 2 2 ). Both the low andhigh flow rate conditions described were experimentally seen with a single-layer design. We evaluated numerous commer-cial membranes and found none which came even close toproviding both: (1) pore density similar to sweat pores in skin;(2) uniform flow from all pores at the low flow rates associated with sweating. For most commercial membranes, flow woulddominate at only very few locations ( , 1%) creating large/talldroplets that would then activate additional pores only as thefew and large growing droplets wetted over them. One might propose to use a membrane with super-wetting (Young’s angleof 0 u ) to eliminate Laplace pressure effects and non-uniformflow from pores, but this is then no-longer representative of  wetting on the surface of skin. A novel approach is needed.Our approach to overcome the Laplace pressure varianceamong the ‘sweat’ pores is a bi-layer membrane design asshown in Fig. 2b. The bottom membrane mimics the flow-rateof the secretory portion of a sweat gland while the topmembrane provides the proper pore-density and surfaceenergy/texture. The bottom membrane dominates the fluidpressure drop and therefore creates a constant flow throughthe top membrane. In addition, the following finer designconsiderations must be implemented. First, the bottommembrane pores should be small enough such that theirflow-induced pressure drop is . 10 6 larger than any pressuredrop induced by the large pores of the top layer (therefore Fig. 1 (a) Diagram of human eccrine sweat gland, (b) diagram and SEM photo of fabricated bi-layer artificial microfluidic ‘skin’ membrane; and (c) diagram of ‘sweat’simulation experimental setup and photo of an operating sweat simulator. Fig. 2 Illustration of flow through ‘sweat pores’ in (a) a single layer of ‘sweat’pores and (b) a bi-layer approach. Provided in (c) is a resistance network modelof the bi-layer approach shown in (b). This journal is ß The Royal Society of Chemistry 2013 Lab Chip , 2013, 13 , 1868–1875 | 1869 Lab on a Chip Paper     D  o  w  n   l  o  a   d  e   d   b  y   U   N   I   V   E   R   S   I   T   Y   O   F   C   I   N   C   I   N   N   A   T   I  o  n   2   6   /   0   4   /   2   0   1   3   2   3  :   2   1  :   5   7 .   P  u   b   l   i  s   h  e   d  o  n   1   8   M  a  r  c   h   2   0   1   3  o  n   h   t   t  p  :   /   /  p  u   b  s .  r  s  c .  o  r  g   |   d  o   i  :   1   0 .   1   0   3   9   /   C   3   L   C   4   1   2   3   1   H  promoting uniform flow rate at all pores). Secondly, the smallpores in the bottom membrane should be of density adequatesuch that each large pore in the top layer should statistically average a similar number of small pores beneath it. Thirdly,the bottom membrane must be adequately hydrophilic that the fluid flow can be activated at low pressure and spread ontop of the bottom membrane immediately to merge andactivate all large pores in the top layer. Commercially availablehydrophilic track etched polycarbonate membranes meet these requirements for the bottom membrane, and areavailable with pore dimensions ranging from 0.01 to 30 m m. As will be discussed in the fabrication section, the top layer will be formed by laser milling and microreplication, andbonded to the bottom membrane.Next, a pumping mechanism was needed. A simple gravity fed approach was chosen (similar to an intravenous drip-bag)to provide a constant fluid pressure at a very low flow rate forcontinuous perspiration simulation, which syringe pumpsgenerally fail to provide. A tank or container of fluid is simply suspended at a height of  h above the sweat simulator device.This allows fine and repeatable adjustment of flow rate by adjusting  h (therefore adjusting hydrostatic pressure) andmore importantly sustains a very low pressure required for low flow rate through the sweat simulator device. Now that thefundamental features of the sweat simulator system aredescribed, a theoretical model can be developed.The theoretical model is presented for the bi-layer design, adesign which produces uniform flow rate at all sweat pores(Fig. 1b, Fig. 2b). A fluid resistance network approach (Fig. 2c) was used to calculate the fluid flow rate ( Q , L h 2 1 ), andtherefore the simulated ‘sweat’ rate densities ( V  , L h 2 1 m 2 2 ).First, equilibrium conditions are assumed where the driving pressure (  P  0 ) will be dropped across the top sweat pore layer(  P  S ) and the bottom membrane layer (  P  m ):  P  0 = P  m + P  S (1)Note that the pressure drop in the hydrophilic tubing connecting the tank and the ‘sweat’ simulator chamber isignored because it has much smaller fluid resistance (id = 1mm tube). The bottom membrane is hydrophilic enough suchthat Laplace pressure does not contribute to P  m (fluid readily  wets through it and onto its exposed surface). Pressure drop inthe bottom membrane (  P  m ) is therefore only due to the fluidresistance of the membrane (  R m ), as given by   P  m = R m 6 Q n (2) where Q n is the fluid flow rate (L h 2 1 ). At this point, the theoretical development will focus on asingle ‘sweat’ pore in the top layer. For one ‘sweat’ pore in thetop layer with a diameter D S , the fluid resistance in the bottommembrane can be calculated as: R m ~ 128 m L m p D 4m | 1 p D 2S a = 4 (3) where L m is the membrane thickness, D m is the membranepore diameter, m is the viscosity of the fluid, and a is themembrane pore density ( p  D 2S a /4 is the amount of pores in themembrane under one ‘sweat’ pore in the top layer).Next, pressure drop in the top ‘sweat’ pore (  P  S ) can becalculated from:  P  S = R S 6 Q n + P  L (4) where R S is the fluid resistance in one ‘sweat’ pore and P  L isthe Laplace pressure, and are calculated respectively as, R S ~ 128 m L S p D 4S (5) P  L ~ 4 c cos h a ð Þ D S (6) where L S is the membrane thickness of the ‘sweat’ pore layer, c is the surface tension of the fluid, and h a is the advancing contact angle for the fluid in the sweat pore. As a reminder,once fluid wets out onto the outer surface of the top sweat porelayer, Laplace pressure becomes small enough that it can beignored.The bottom membrane is selected such that the fluidresistance in the bottom membrane is much larger than theresistance in the top sweat pore layer (  R m & R S ), whichrequires the bottom membrane pore diameter to be D m vv  ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  4 D 2S L m = p a L S ð Þ 4 q  . The fluid flow rate ( Q , L h 2 1 ) in one‘sweat’ pore for the bi-layer design can then be calculated asfollows: Q n ~ p 2 D 4m D S a r  ghD S { 4 c cos h ð Þ½  512 m L m (7)The fluid flow rate density or ‘sweat’ rate density ( V  n , L h 2 1 m 2 2 ) in one ‘sweat’ pore area of  p  D 2S is then calculated as: V  n ~ p D 4m a r  gh { 4 c cos h ð Þ = D S ½  512 m L m (8)For the fabricated artificial microfluidic skin with a ‘sweat’pore density of  v , the ‘sweat’ rate density (L h 2 1 m 2 2 ) for theartificial microfluidic skin is therefore: V  ~ v p 2 D 4m D S a r  ghD S { 4 c cos h ð Þ½  512 m L m (9)This model does not include effects of evaporation vs. droplet size, 25 as our rough calculations determined evapora-tion to be inconsequential with respect to the flow rates andLaplace pressures driving the model. With a theoretical modelfor ‘sweat’ rate density now in-hand, and dimensionalconstraints better understood, we can now discuss thefabrication of the artificial microfluidic skin. 1870 | Lab Chip , 2013, 13 , 1868–1875 This journal is ß The Royal Society of Chemistry 2013 Paper Lab on a Chip    D  o  w  n   l  o  a   d  e   d   b  y   U   N   I   V   E   R   S   I   T   Y   O   F   C   I   N   C   I   N   N   A   T   I  o  n   2   6   /   0   4   /   2   0   1   3   2   3  :   2   1  :   5   7 .   P  u   b   l   i  s   h  e   d  o  n   1   8   M  a  r  c   h   2   0   1   3  o  n   h   t   t  p  :   /   /  p  u   b  s .  r  s  c .  o  r  g   |   d  o   i  :   1   0 .   1   0   3   9   /   C   3   L   C   4   1   2   3   1   H
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks