transport properties of porous membranes based on electrospun nanofibers

Colloids and Surfaces

A: Physicochemical and Engineering Aspects 187–188 (2001) 469–481

Transport properties of porous membranes based onelectrospun nanofibers

Phillip Gibson *, Heidi Schreuder-Gibson, Donald RivinMaterials Science, AMSSB-RSS-MS, US Army Soldier Systems Center, Natick, MA 01760-5020, USA

Abstract

Electrospinning is a process by which high voltages are used to produce an interconnected membrane-like web ofsmall fibers (10–500 nm in diameter). This novel fiber spinning technique provides the capacity to lace together avariety of types of polymers, fibers, and particles to produce ultrathin layers. Of particular interest are electrospunmembranes composed of elastomeric fibers, which are under development for several protective clothing applications.The various factors influencing electrospun nonwoven fibrous membrane structure and transport properties arediscussed. Performance measurements on experimental electrospun fiber mats compare favorably with transportproperties of textiles and membranes currently used in protective clothing systems. Electrospun layers present minimalimpedance to moisture vapor diffusion required for evaporative cooling. There may be special considerations in theapplication of elastomeric membranes for protective clothing. Effects of membrane distortion upon transportbehavior of the structure might be significant. Preliminary measurements have found that changes in elastomericmembrane structure under different states of biaxial strain were reflected in measurements of air flow through themembrane. Changes in membrane structure are also evident in environmental scanning electron microscope (SEM)images of the pore/fiber rearrangement as the membrane is stretched. Experimental measurements and theoreticalcalculations show electrospun fiber mats to be extremely efficient at trapping airborne particles. The high filtrationefficiency is a direct result of the submicron-size fibers generated by the electrospinning process. Electrospun nanofibercoatings were applied directly to an open cell polyurethane foam. The air flow resistance and aerosol filtrationproperties correlate with the electrospun coating add-on weight. Particle penetration through the foam layer, whichis normally very high, was eliminated by extremely thin layers of electrospun nanofibers sprayed on to the surface ofthe foam. Electrospun fiber coatings produce an exceptionally lightweight multifunctional membrane for protectiveclothing applications, which exhibits high breathability, elasticity, and filtration efficiency. © 2001 Elsevier ScienceB.V. All rights reserved.

Keywords: Electrospinning; Electrospun; Porous; Membrane; Fiber

www.elsevier.nl/locate/colsurfa

1. Introduction

Electrospinning is a process by which a sus-pended droplet of polymer solution or melt ischarged to high voltage to produce fibers [1–3].

* Corresponding author. Tel.: +1-508-2334273; fax: +1-508-2335521.

E-mail address: [email protected] (P. Gibson).

0927-7757/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.

PII: S0927 -7757 (01 )00616 -1

P. Gibson et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 187–188 (2001) 469–481470

At a voltage sufficient to overcome surface ten-sion forces, fine jets of liquid shoot out toward agrounded target. The jet is stretched and elongatesbefore it reaches the target, dries and is collectedas an interconnected web of small fibers. Thetechnique provides the capacity to lace together avariety of types of polymers and fibers to produceultrathin layers, which are useful for chemicalprotective clothing. Depending on the specificpolymer being used, a range of fabric properties,such as strength, weight and porosity, can beachieved. Fig. 1 shows that much smaller fiberdiameters and increased surface areas are accessi-ble through the electrospinning technique as com-pared with currently available textile fibers. Fibersizes of 40 nm and smaller have been reported [4],although we normally produce fibers in the 200–500 nm range in our apparatus. Fig. 2 showsmicrographs of typical electrospun fibers.

Electrospinning results in submicrometer sizefibers laid down in a layer that has high porositybut very small pore size. For fibers spun frompolymer solutions, the presence of residual solventin the electrospun fibers facilitates bonding ofintersecting fibers, creating a strong cohesiveporous structure. The nanofibers assemble into amembrane-like structure that exhibits good tensilestrength, excellent moisture vapor transport, ex-tremely low air permeability and good aerosolparticle protection.

Fig. 2. (a) SEM image of Nylon 6,6 electrospun fiber mat; (b)SEM image of nylon 6,6 nanofibers spun across surface ofhuman hair fiber.

Electrospun nanofiber membranes may be pro-duced over a wide range of porosity values, fromnearly nonporous polymer coatings, to veryporous and delicate fibrous structures. Our cur-rent interest is in consolidated membrane struc-tures with porosities ranging from 30 to 60%.Typical capillary flow liquid expulsion porometrymeasurements indicate that pore throat diametersrange from 0.1 to 0.8 �m in size.

One implication of this type of membrane ar-chitecture is that these materials would providegood resistance to the penetration of chemical andbiological warfare agents in aerosol form, whilestill allowing significant water vapor transport topromote evaporative cooling of the body. Electro-spun nonwoven fiber mats may be thought of as amicroporous material that behaves like a mem-brane, as opposed to a more porous, air-perme-able fabric. Because of the small fiber/pore sizes inthese electrospun membranes, the resistance toFig. 1. Electrospun nanofibers have high surface area.

P. Gibson et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 187–188 (2001) 469–481 471

convective gas flow is quite large and is comparableto the flow resistance shown by microporous poly-tetrafluoroethylene (PTFE) membranes, which areoften used as a component of protective clothingsystems.

Our main interest in electrospun nanofiber mem-branes has been for chemical/biological protectiveapplications. Recently, we have directly incorpo-rated activated carbon, carbon nanotubes, andsubmicron enzyme particles into electrospun mem-branes. Besides chemically protective clothing ap-plications, electrospun nanofibers are underconsideration for air and water filtration applica-tions, polymer composite reinforcement, and elec-trically conductive polymeric networks [5].

A significant advantage of electrospun mem-branes is that they may be sprayed directly ontothree-dimensional (3-D) forms. This is analogousto the electrostatic spray painting technique usedfor painting metal objects such as car bodies andmetal parts. The ability to produce coated objectswith a microporous fibrous layer has advantages interms of eliminating seams, and in the ability tovary the thickness of the nanofiber coating atvarious locations over the surface of the object.Certain military applications such as boot andglove liners, chemical/biological protective cloth-ing, and flexible gas mask hoods suffer fromseam-sealing problems; 3-D net shape electrospin-ning offers one possible avenue for a solution.

This paper’s overall objective is to summarizeinformation related to the transport properties ofelectrospun membranes both as single layers, andas coatings applied to existing military chemical/bi-ological protective clothing systems. Results arepresented for water vapor diffusion and air perme-ability (important for comfort) and aerosol particlepenetration (important for protection from chemi-cal and biological warfare agents).

2. Experimental method

2.1. Water �apor diffusion and gas con�ection

Water vapor diffusion and gas convection prop-erties of electrospun fiber mats were determinedwith an automated dynamic moisture permeation

cell [5–7]. Gas flows of known temperature andwater vapor concentration enter the test cell; bymeasuring the temperature, water vapor concentra-tion, and flow rates of the gas leaving the cell, onemay determine the fluxes of gas and water vaportransported through the test sample. With nopressure difference across the sample, transport ofwater vapor proceeds by pure diffusion, driven byvapor concentration differences. If a pressure dif-ference across the sample is present, transport ofvapor and gas includes convective transport, wherethe gas flow through the sample carries water vaporwith it, which may add to or subtract from thediffusive flux, depending on the direction of theconvective gas flow.

Permeability to convective gas flow is given byDarcy’s Law [8] such that:

kD=��Q

A��x

�p�

(1)

where kD is the permeability constant (m2), � thegas viscosity (17.85×10−6 kg m−1 s for N2 at20°C), Q the total volumetric flow rate (m3 s−1),A the area of test sample (m2), �x the thickness (m)and �p is the pressure drop across sample (N m−2

or Pa).For textiles, although thickness measurements

seem simple, they are often problematic, and canbe a large source of error if they are incorporatedinto reported measurements of Darcy permeability.It is preferable to present the pressure-drop/flowrate results in terms of an apparent flow resistancedefined as:

RD=�A�p

�Q�

(2)

where RD is the apparent Darcy flow resistance(m−1).

We define the resistance to mass transfer bydiffusion as the simple addition of an intrinsicdiffusion resistance due to the sample (Ri) and thediffusion resistance of the boundary air layers (Rbl):

(Ri+Rbl)=� �C�

(m� /A)n

(3)

where Ri is the intrinsic diffusion resistance ofsample (s m−1), Rbl the diffusion resistance ofboundary air layers (s m−1), m� the mass flux ofwater vapor across the sample (kg s−1), A the


area of test sample (m2), �C� is the log mean watervapor concentration difference between top andbottom nitrogen streams (kg m−3).

The test material’s intrinsic diffusion resistancemay be obtained by subtracting the apparentboundary layer diffusion resistance [6], but inmany cases, it is sufficient to leave the results inthe form of a total resistance to water vapordiffusion which includes the boundary layereffects.

2.2. Aerosol filtration

The aerosol particle filtration test system isbased on the apparatus described by Park andRivin [9]. The majority of aerosol particles gener-ated by this system are between 1 and 5 �m. Adilute solution of aqueous potassium iodide isatomized with an ultrasonic nebulizer; the result-ing mist of liquid salt solution is diluted with drynitrogen and flows to a heated drying chamber.The liquid evaporates in the drying chamber tocreate a solid salt aerosol, which is passed throughthe filter test material. Aerosol particle size andconcentration are monitored by an aerosol ana-lyzer (API Aerosizer, Amherst Instruments, Inc.)which measures the time of flight of individualparticles as they pass through two laser beams.The aerosol analyzer is capable of particle sizemeasurement over the range of 0.5–200 �m. Fur-ther details of the test system are available in[9,10].

The sample test area clamped in the holderexposed a circular area of 1.34×10−3 m2 to theaerosol stream. The approximate test temperatureat the sample holder was 40°C, and the equivalentnitrogen volumetric flow rate at this temperaturewas 2.1×10−5 m3 s−1. Thus the nominal flowvelocity through the sample (u0) was 0.0157 ms−1. Mean aerosol particle size was 2–3 �m [10].

A useful measure for comparing the filtrationperformance of different materials is the totalamount of aerosols removed from the gas stream.One may also be interested in the relative filtra-tion performance for different particle sizes ofaerosols. For this report, we compare perfor-mance based on the experimental filtration effi-ciency given by:

E=n2−n1

n2

(4)

where n2 is the aerosol particle concentration up-stream, and n1 is the concentration downstream ofthe filter test sample. Alternatively, the penetra-tion ratio P may be used, which is defined byP=1−E.

2.3. Factors affecting selection of test conditions

The filter velocity used for testing the materialsis low compared with normal filter test conditionsfor industrial filters. The difference arises from thefact that the use of clothing materials as aerosolbarriers is quite different from porous materialsused as filters. Filters are commonly placed insystems, which have a well-defined flow rate orpressure drop across the filter material. In cloth-ing systems, however, the aerosol barrier is incor-porated into a clothing system covering thehuman body, usually with air gaps present be-tween the body and the clothing. An external airflow due to wind or body motion impinges on theclothed human, and some air flows around thebody, while some penetrates through the clothingsystem and into the air gap between the clothingand body, where aerosol deposition on the skinmay occur. For a given external air velocity, theamount of air which flows around the body, andthe amount which penetrates through the aerosolbarrier layer is determined by the air flow resis-tance (air permeability) of the barrier layer. Mate-rials with a low air flow resistance allow arelatively high flow rate through the fabric, with acorrespondingly low pressure drop. Materialswith a high air flow resistance allow less flowthrough the fabric, and have a higher pressuredrop across the fabric layer (up to the limit of thestagnation pressure for the particular environmen-tal flow conditions). For a truly valid comparisonbetween aerosol barrier materials which differ intheir air permeability properties, it would be de-sirable to test at a unique volumetric flow rate/filter velocity and pressure drop whichcorresponds to that produced by a given externalair velocity on a typical clothing system.

Some methodologies have been developed forrational aerosol testing of protective clothing ma-


terials which takes these factors into account [11–13]. However, to minimize the number of tests forpreliminary characterization efforts, one flow ve-locity test condition was used for this study. Thefilter flow velocity (u0) of 0.0157 m s−1 selectedfor the aerosol test system is a compromise —this flow rate across the porous membranes andmost of the electrospun layers results in very highpressure drops that would be unrealistic in anyclothing system. Thus, the aerosol test conditionsfor the materials with a high flow resistance arevery much a ‘worst case’ situation. However, forthe materials with a low flow resistance, the flowrates through the fabric would be higher than therelatively low flows and low pressure drops usedin the aerosol test system. As will be seen later,the fact that the electrospun layers exhibited highfiltration efficiencies compared with common fab-ric materials, even though the test conditions wererelatively unfavorable, indicates the usefulness ofthese materials for protective clothingapplications.

3. Results

3.1. Water �apor diffusion and gas flow properties

The magnitude of the convective flow resistancefor two typical electrospun membranes composedof polyacrylonitrile and polybenzimidazole [14] is

shown in Fig. 3, and compared with other com-mon fiber-based materials. Fig. 3 shows the gasflow resistance as a function of relative humidity,which is of most importance in the case of ahydrophilic fiber, such as cotton [15].

The convective flow resistance of the electro-spun fiber mats is quite large compared withnormal clothing materials, yet, as will be shownlater, the high resistance to air flow does notimpede the diffusion of water vapor through thepore structure of the electrospun nonwoven mate-rial. In general, materials with high rates of watervapor diffusion and low air permeability arepromising candidates for protective clothingapplications.

An interesting complicating factor in the analy-sis of gas flow through porous materials with suchsmall fiber diameters is that the mean free path ofgas molecules becomes comparable to the fibersize [16]. There is gas slip at the fiber surface, andthe normal linear dependence of flow rate withpressure drop, valid in low Reynolds numberlaminar flows, becomes less applicable to the flowfield in the electrospun nonwoven pores. It hasbeen estimated that the pressure drop predictedby assuming continuum flow through beds offibers is reduced by a factor of about 1/3 for afiber diameter of 0.1 �m [17].

Water vapor transport properties of electrospunnonwovens are excellent and indicate that layersbased on electrospun fiber technology will be thin,lightweight, and very ‘breathable’ with respect toallowing evaporative cooling to take placethrough the protective clothing system. Water va-por diffusion measurements for the same materi-als given in Fig. 3 are shown in Fig. 4. Theelectrospun nonwoven resistance to water vapordiffusion is much lower than the commerciallyavailable membrane laminates presented in Fig. 4,which are often said to be highly ‘breathable’. Fig.4 compares the performance of the materials as afunction of the ‘mean relative humidity’, which isuseful as an indication of concentration-depen-dent transport behavior in polymer membranesand membrane laminates [6,7].

Electrospinning also lends itself very easily tospraying a fiber coating directly onto other mate-rials. For example, an extremely thin electrospun

Fig. 3. Microporous membranes and electrospun nonwovenshave high convective gas flow resistance.


Fig. 4. Electrospun nonwovens have excellent water vaportransport properties.

coated foam; (2) the uncoated foam tested incombination with a microporous PTFE mem-brane; (3) the microporous PTFE membrane byitself.

Water vapor diffusion and air flow resistanceproperties were determined under simultaneousdiffusion/convection conditions in the DMPC, asshown in Fig. 6.

Gas enters the DMPC at a relative humidity of0.90 (90% r.h.) on the top portion of the cell, and0.0 (0% r.h.) on the bottom of the cell. Theautomated valves are used to restrict the flow onone or the other sides of the cell. This causes thepressure in one side of the cell to be higher than inthe other, producing convective gas flow throughair-permeable samples, in addition to the diffu-sion flux taking place due to the concentrationgradients.

Fig. 5. (a) Porous polyurethane foam containing activatedcarbon; (b) electrospun nylon 6,6 nanofiber coating sprayedonto surface of the foam.

fiber coating may be sprayed onto the surface ofchemical/biological protective garments to im-prove aerosol particle filtration and capture, or toincorporate additional functionality to the protec-tive garment such as agent-specific enzymes orcatalysts. These reactive components would beexposed to reaction conditions in either liquid orvapor phases under varying flow environments.

Water vapor diffusion resistance and air flowresistance properties were determined for an elec-trospun nonwoven (nylon 6,6) deposited onto anactivated-carbon-loaded polyurethane foam,which was an experimental variant of the materialused as a chemical protective layer in the USArmy Battle Dress Overgarment (BDO). Thisfoam had a nominal thickness of 8.7×10−4 mwith an areal density of 0.19 kg m−2; the openfoam structure results in very little resistance toconvective air flow (flow resistance is 3.2×106

m−1). Foam thickness and areal density may varyas much as 10% from the mean value. Scanningelectron microscope (SEM) images of the porouscarbon-loaded polyurethane foam, and the nylonnanofiber coating on the surface and over thepores of the foam, are shown in Fig. 5.

Seven different electrospun coating add-on lev-els were produced over the range of 1.6×10−4–1.1×10−2 kg m−2. Three other materials werealso tested for comparison purposes, (1) the un-


Fig. 6. Convection/diffusion experiment in the DMPC.

Fig. 8. Convective gas flow resistance related to electrospuncoating level.

The plot of convective flow as a function ofpressure drop in Fig. 7 shows the expected trends.In this test method, materials with a low slope inFig. 7 have a high flow resistance; the numericalvalue of the slope can be converted to a flowresistance as given in Eq. (2).

The electrospun membranes’ air flow resistancecorrelates well with the electrospun layer coatinglevel, as shown in Fig. 8. Fig. 8 also shows theresults for the microporous PTFE membrane,which has an areal density comparable to thehighest electrospun coating level.

Measurements are taken as a function of pres-sure drop across the sample, where the convectiveflow and pressure drop are gradually increased instepwise increments. In addition to the pressuredrop, an electronic mass flow meter connected tothe lower outlet of the cell is used to record themass flow rate of gas through the test material.More details on this test method, and correlationswith other standard methods, are available in[18,19].

Fig. 7. Convective flow through samples as a function of pressure drop.


The water vapor diffusion resistance as a func-tion of pressure drop (Fig. 9) primarily reflects thedifferent air permeability properties of each coat-ing level. Water vapor diffusion resistance isdefined as the intersection of each curve with thepoint where the pressure difference across thesample is equal to zero. Fig. 9 contains informa-tion about both convective and diffusive transportacross the test sample. Those materials withhigher flow resistances are affected less by thepressure difference across the sample. Fig. 9 in-cludes the boundary layer resistance due to lami-nar air flow over the sample, which in this case isabout 105 s m−1. It is known from previous work[6] that the intrinsic diffusion resistance of thePTFE membrane is 6–8 s m−1. All the electro-spun coated foam samples, the foam sampletested in combination with the PTFE membrane,and the uncoated foam sample converge to ap-proximately 200 s m−1 at the zero pressure differ-ence point. This translates to an intrinsic diffusionresistance for the samples of approximately 95 sm−1 after subtraction of the boundary layerresistance.

The small variations in intrinsic diffusion resis-tance of the coated and uncoated samples arenegligible in the practical sense, since they arewithin the experimental variation of diffusion re-

sistance caused by the large thickness variations inthe foam substrate. It is likely that none of theelectrospun layers added more than 10 s m−1 tothe total diffusion resistance of the foam layer. Itis worthwhile to refer to Fig. 4 where the watervapor diffusion resistances of commercial breath-able laminates are shown to be much higher(200–1500 s m−1). Other factors in clothing sys-tems, such as stagnant air trapped underneath/be-tween layers, or boundary layer resistances due toair flow over the body surface, also result in watervapor diffusion resistance factors which far out-weigh the small contribution of the electrospunmembrane layer.

3.2. Aerosol filtration

Aerosol particle filtration properties were deter-mined for an electrospun nonwoven (nylon 6,6)deposited onto an activated-carbon-loadedpolyurethane foam. The materials produced werenominally the same as those used for the vapordiffusion and air flow testing presented previ-ously, but the aerosol testing used a separate setof samples.

Eight different electrospun coating add-on lev-els were produced over the range of 1.2×10−4–1.9×10−3 kg m−2. Nominal coating times of 30

Fig. 9. Water vapor diffusion resistance as a function of pressure drop.


Fig. 10. Comparison of relative aerosol penetration through various porous membranes and textiles. Coating weight listed inparentheses in units of kg m−2.

s, 1, 2, 3 and 4 min were used for the first fivecoating levels. The last three coating levels werenot timed, but were monitored visually to produceincreasing coating levels. Unlike the previous setof electrospun coated layers, the coating level wasnot measured directly. Measuring the coatinglevel is destructive to the sample, since the layermust be peeled or scraped off to measure theweight. Instead, the air flow resistance propertiesof the electrospun-coated foam samples was mea-sured. The relationship between coating level (kgm−2) and the air flow resistance (1 m−1) given inFig. 8 was then used to estimate the actual coat-ing level.

Three other classes of materials were also testedfor comparison purposes,1. porous membrane filters consisting of a mi-

croporous PTFE membrane and a commer-cially available polyvinylidene fluoride(PVDF) microporous membrane;

2. stand-alone electrospun membrane layers, oneof which was composed of polybenzimidazolefiber, and another two samples spun frompolyurethane polymer;

3. five clothing/fabric materials used in militaryor protective clothing.

The fabric materials included the nylon/cottonfabric used as the outer shell component in theUS Army’s BDO chemical protective suit. Alsotested was the ‘Saratoga’ chemical protective linerused in the US Marine Corps chemical protectivesuit, as well as the woven cotton fabric used as theouter shell for this suit. Two tightly-woven nylonfabrics were tested, the nylon fabric used as theouter shell of the US Army’s Extreme ColdWeather Clothing System (ECWCS) and a ‘mi-crofiber’ nylon fabric used as a water-resistantbreathable shell in outdoor clothing.

The relative aerosol penetration through thevarious types of materials is shown in Fig. 10. Noaerosol particle penetration was detected for thetwo microporous membrane filter materials or forthe three electrospun membrane samples. For thegroup of electrospun-coated foam samples, thereis a definite relationship between the coating leveland the relative aerosol penetration through thecoated samples. All the fabric samples show sig-nificant aerosol penetration levels, with the nylonmicrofiber fabric having the best aerosol particle


filtration properties and the Saratoga Liner mate-rial the worst filtration properties.

Fig. 11 combines the aerosol filtration proper-ties and the air flow resistance properties of thenylon electrospun-coated polyurethane foam sam-ples. Fig. 11 shows that even the lightest coatingof electrospun fibers significantly reduces aerosolpenetration through the carbon-loadedpolyurethane foam.

Particle size distributions of penetrating parti-cles and SEM micrographs of aerosol particlescaught in the electrospun fiber coating are givenin [10]. [10] also calculates single fiber overall filtercollection efficiencies of electrospun fiber matsaccording to various standard models [17,20–22].

3.3. Effect of biaxial strain on elastomeric porousnanofiber membranes

In the course of studying microporous mem-branes for protective clothing systems, we aredeveloping new elastomeric membranes comprisedof nonwoven nanofibers. By using the process ofelectrospinning, highly deformable membranestructures have been made that exhibit strain ca-pacities of over 200% with full elastic recovery.These membranes are soft, flexible, and possessmean pore diameters of �1 �m.

Measurement of porosity, air flow, and aerosolfiltration for highly deformable membranes mustbe performed carefully on these soft membranes.Furthermore, the ability of these membranes tostretch at flexure points is critical to their perfor-mance. These membranes can be laminated toknitted fabrics to impart wind and water resis-tance to various types of rainwear and sportclothing. However, the permeability of these lami-nated structures, for example at elbows and knees,would be expected to change upon flexing. Thereis a need to fully characterize the penetrationcharacteristics of these membranes if they are tobe used to protect personnel from biological oraerosol threats.

Electrospun thermoplastic elastomer (TPE)membranes are produced using Pellethane (DowChemical) and Estane (B.F. Goodrich) commer-cial-grade pellets. Electrospinning of these TPEsin solvents such as dimethyl formamide and/ortetrahydrofuran result in a tough elastomericmembrane which looks like a dense nonporousrubber sheet to the naked eye. Samples of theseelectrospun membranes were stretched in two di-rections to various strain levels, and tested forwater vapor diffusion and air flow properties inthe DMPC apparatus [15,23]. Samples under bi-axial strain were also examined using an Environ-mental Scanning Electron Microscope (ESEM)

Fig. 11. Aerosol filtration and air flow resistance as function of coating level.


Fig. 12. Effect of strain level on water vapor diffusion resis-tance and air flow resistance of a porous elastomeric mem-brane (example shown is pellethane thermoplasticpolyurethane).

straighten and interfiber spacings increase, sug-gesting a possible increase in porosity. The densityof fibers that we see in the image diminishes asstrain increases. The fibers in the unstrained elas-tomeric membrane are much thicker, on average,than they are in the strained membrane. Theseelastomeric membranes behave much like nonwo-ven fibrous structures which are widely used inprotective clothing, albeit at a much smaller scaleof fiber and pore sizes. The advantages inherent inthe increased surface area and smaller pore sizes,as well as in the very thin and lightweight natureof membranes based on elastic nanofiberpolyurethanes, have many applications in militaryand civilian protective clothing systems.

4. Conclusions

Performance measurements on experimentalelectrospun fiber mats compared favorably withtransport properties of textiles and membranes

[24], to observe the changes in pore/fiber structureat different strain levels.

Typical results for an elastomeric nanofibermembrane are shown in Fig. 12.

As might be expected, stretching the membraneuniformly opens up the pores and decreases themeasured air flow resistance. However, the changein air flow resistance would still be within the‘windproof’ category in terms of the relative prop-erties of protective clothing. It is interesting thatthe water vapor diffusion properties remain un-changed over the strain levels used for this mem-brane. This membrane would be categorized asvery ‘breathable’, and the measured resistance towater vapor diffusion is due primarily to theboundary layer flow over the sample surface pro-duced within the DMPC apparatus. Any changesin water vapor diffusion properties caused bypore/fiber rearrangement are insignificant com-pared with the other effects present in a protectiveclothing system, such as stagnant air layers be-tween the skin and clothing layers, or theboundary layer present in the external flow over aclothed human.

A visual indication of the physical appearanceof an electrospun elastomeric polyurethane mem-brane at two different strain levels is shown inFig. 13.

It can be seen that as biaxial strain increasesfrom 0 to 100%, the electrospun fibers begin to

Fig. 13. ESEM images of elastomeric nanofiber membranesunder two different levels of biaxial strain.


currently used in protective clothing systems. Theelectrospun layers presented minimal impedanceto moisture vapor diffusion required for evapora-tive cooling.

Experimental measurements and theoretical cal-culations showed electrospun fiber mats to beextremely efficient at trapping aerosol particles.The high filtration efficiency is a direct result ofthe submicron-size fibers generated by the electro-spinning process. Because the electrospun fiberdiameters are comparable to the mean free pathof air molecules, the high pressure drop penaltyincurred by small-fiber filter materials which oper-ate in the continuum flow regime is muchreduced.

Electrospun nanofiber coatings were applied di-rectly to a polyurethane foam containing acti-vated carbon, which is used as a component inmilitary chemical protective clothing systems. Theair flow resistance and aerosol filtration propertiescorrelated with the electrospun coating add-onweight. Aerosol particle penetration through thefoam layer, which is normally very high, waseliminated by minuscule levels of nylon electro-spun nanofibers sprayed on to the surface of thefoam.

Porous elastomeric nanofiber membranes werestretched in biaxial tension to strain levels of100%. Convective gas flow properties were signifi-cantly affected as the interfiber pores opened up.Water vapor diffusion properties remained un-changed within the limits of the test device usedfor the evaluation. ESEM images of the deformedelastomeric nanofiber membranes confirmed thatthe elastic fibers were under an increasing state oftension while interfiber pore space increased.

Potential future applications of electrospun lay-ers include direct application of elastomeric mem-branes to garment systems, eliminating suchcostly manufacturing steps as laminating and cur-ing. It may be possible to electrospin fibers di-rectly onto 3-D screen forms obtained by 3-Dbody scanning. Scientists are currently using alaser-based optical digitizing system to record thebody surface coordinates of test subjects [25]. Thisinformation could be integrated with computeraided design and manufacturing (CAD/CAM)processes to allow electrospun garments to be

sprayed onto the digitized form, resulting in cus-tom-fit, seamless clothing.

References

[1] P.K. Baumgarten, Electrostatic spinning of acrylic mi-crofibers, J. Colloid Interf. Sci. 36 (1971) 71.

[2] L. Larronda, R.S. John Manley, Electrostatic fiber spin-ning from polymer melts. I. Experimental observations onfiber formation and properties, J. Polym. Sci.: Polym.Phys. 19 (1981) 909.

[3] J. Doshi, J. Reneker, D.H. Reneker, Electrospinning pro-cess and applications of electrospun fibers, J. Electrostat.35 (1996) 151.

[4] D. Reneker, I. Chun, Nanometre diameter fibres of poly-mer, produced by electrospinning, Nanotechnology 7(1996) 215.

[5] H. Schreuder-Gibson, K. Senecal, M. Sennett, Z. Huang,J. Wen, W. Li, Y. Tu, D. Wang, S. Yang, Z. Ren, C.Sung, Characteristics of electrospun fibers containing car-bon nanotubes, in: Proceedings of 197th ElectrochemicalSociety Conference, Toronto, Canada, May 14–19, 2000.

[6] P. Gibson, C. Kendrick, D. Rivin, M. Charmchi, L.Sicuranza, An automated water vapor diffusion testmethod for fabrics, laminates, and films, J. Coated Fab-rics 24 (1995) 322.

[7] P. Gibson, M. Charmchi, The use of volume-averagingtechniques to predict temperature transients due to watervapor sorption in hygroscopic porous polymer materials,J. Appl. Polym. Sci. 64 (1997) 493.

[8] F. Dullien, Porous Media-Fluid Transport and PoreStructure, Academic Press, New York, 1979, p. 157.

[9] H.B. Park, D. Rivin, An aerosol challenge test for perme-able fabrics, US Army Natick Research, Developmentand Engineering Center Technical Report, NATICK/TR-92/039L, July, 1992.

[10] P. Gibson, H. Schreuder-Gibson, Aerosol particle filtra-tion, gas flow, and vapor diffusion properties of electro-spun nanofiber coatings, NATICK/TR-99/016L, USArmy Soldier Systems Center, Natick, MA, February,1999.

[11] P. Fedele, W. Bergman, R. McCallen, S. Sutton, Hydro-dynamically induced aerosol transport through clothing,in: Proceedings of the 1986 Army Science Conference,vol. I, 1986, pp. 279–293.

[12] J. Hanley, P. Fedele, Wind tunnel study of aerosol pene-tration through protective overgarments, in: Proceedingsof the 1987 US Army Chemical Research, Development,and Engineering Center Scientific Conference on Chemi-cal Defense Research, 1988, pp. 444–450.

[13] J. Hanley, D. VanOsdell, P. Fedele, Methodologies fortesting permeable chemical protective fabrics and gar-ments against an aerosol threat, in: Proceedings of the1988 US Army Chemical Research, Development, andEngineering Center Scientific Conference on ChemicalDefense Research, 1989, pp. 347–353.


[14] J.S. Kim, D.H. Reneker, Polybenzimidazole nanofiberproduced by electrospinning, Polym. Eng. Sci. 39 (5)(1999) 849.

[15] P. Gibson, A.E. Elsaiid, C. Kendrick, D. Rivin, M.Charmchi, A test method to determine the relative humid-ity dependence of the air permeability of textile materials,J. Testing Eval. 25 (1997) 416.

[16] A.A. Kirsch, I.B. Stechkina, N.A. Fuchs, Gas flow inaerosol filters made of polydisperse ultrafine fibres, Aero-sol Sci. 5 (1974) 39.

[17] W. Hinds, Aerosol Technology, Wiley, New York, 1982,pp. 183–184.

[18] P. Gibson, M. Charmchi, Modeling convection/diffusionin porous textiles with inclusion of humidity-dependentair permeability, Int. Commun. Heat Mass Transfer 24(1997) 709.

[19] P. Gibson, C. Kendrick, D. Rivin, Convection/diffusiontest method for porous textiles, Int. J. Clothing Sci.Technol. 12 (2) (2000) 96–113.

[20] R.C. Flagan, Fundamentals of Air Pollution Engin-eering, Prentice Hall, Englewood, NJ, 1988, pp. 433–455.

[21] K.W. Lee, B.Y.H. Liu, Theoretical study of aerosol filtra-tion by fibrous filters, Aerosol Sci. Technol. 1 (1982)147.

[22] D. Boulaud, A. Renoux, Stationary and nonstationaryfiltration of liquid aerosols by fibrous filters, in: K.Spurny (Ed.), Advances in Aerosol Filtration, CRC PressLLC, Boca Raton, FL, 1998, p. 54.

[23] P. Gibson, D. Rivin, C. Kendrick, H. Schreuder-Gibson,Humidity-dependent air permeability of textile materials,Textile Res. J. 69 (5) (1999) 311.

[24] G. Danilatos, Introduction to the ESEM instrument,Microsc. Res. Technol. 25 (1993) 354.

[25] S. Paquette, 3-D scanning in apparel design and humanengineering, IEEE Comput. Graphics Applic. 16 (5)(1996) 11.

.

transport properties of porous membranes based on electrospun nanofibers

Documents