Final Report: Nanostructured Membranes for Filtration, Disinfection, and Remediation of Aqueous and Gaseous Systems

EPA Grant Number: GR832372
Title: Nanostructured Membranes for Filtration, Disinfection, and Remediation of Aqueous and Gaseous Systems
Investigators: Kit, Kevin , Davidson, P. Michael , Weiss, Jochen , Zivanovic, Svetlana
Institution: University of Tennessee - Knoxville , University of Massachusetts - Amherst
EPA Project Officer: Carleton, James N
Project Period: August 1, 2005 through July 31, 2008 (Extended to July 31, 2009)
Project Amount: $349,200
RFA: Greater Research Opportunities: Research in Nanoscale Science Engineering and Technology (2004) RFA Text |  Recipients Lists
Research Category: Hazardous Waste/Remediation , Nanotechnology , Safer Chemicals

Objective:

The objective of this project was to develop electrospun nanofiber chitosan membranes that will have the ability to treat aqueous and gaseous environments by actions of filtration, disinfection, and metal binding.  Chitosan is nontoxic and biodegradable, and has been shown to have beneficial antimicrobial and metal binding properties.  These beneficial properties will be optimized in a nanofiber structure in which the surface area per mass is very high.  The central hypothesis for the proposed research is that the degree to which these nanofiber chitosan membranes effectively filter contaminants, kill microbes, and bind harmful metals will be optimized by minimizing the size of the electrospun fibers and maximizing the available chitosan surface area.  The project was originally divided into the following four tasks:

  1. Determine processing-structure relationships for electrospun chitosan
  2. Measure filtration properties of electrospun chitosan
  3. Measure metal binding and physicochemical properties of electrospun chitosan
  4. Measure antimicrobial properties of electrospun chitosan

However, we found it more logical to divide the project accordingly:

  1. Determine processing-structure relationships for electrospun chitosan
  2. Measure metal binding and antimicrobial properties in static conditions
  3. Measure metal binding, antimicrobial, and filtration properties under dynamic flow conditions

In this way, we were able to determine the most effective fiber structures for metal binding and antimicrobial performance under near equilibrium conditions and then select the most effective for additional dynamic filtration studies.

Summary/Accomplishments (Outputs/Outcomes):

Project Activities
 
Processing-Structure Relationships
The goal of this task was to understand the relationship between electro­spinning process variables and the structure of spun chitosan nanofibers.  The processing conditions (chitosan molecular weight, polymer concentrations) that produced fibrous mats with a low frequency of defects was determined.  Also the effects of these conditions on fiber diameter and actual fiber composition have been studied. 
 
Electrospinning of pure chitosan was hindered by its high degree of inter and intra-chain hydrogen bonding and high solution viscosity. Addition of other synthetic polymers like PEO and PAAm greatly improved the spinnability of chitosan. We were able to obtain fairly uniform sized electrospun fibers by making blend solutions with small amounts of PEO (5%) in the blend solution using acetic acid as the solvent. Fiber formation and size was influenced by blend ratio of the two polymers, polymer concentration, polymer molecular weight and solvent. Heating of the polymer solution with hot air helped improve spinnability by reducing the bead like defects in the formed fibers and enabling the formation of fibers with as low as 5% PEO in the blend solution. Uniform bead-less fibers (as fine as 90 μm) were formed with 10% PEO in blend solution by spinning the solution at 70°C.
 
The advantage of heating the polymer solution during the spinning process was more pronounced for obtaining bead-less fibers of chitosan/PAAm blends. Room temperature spinning solutions were able to form beaded fibers at 75% chitosan in blend solution, however heating the solution to 70°C enabled us to spin HMWchitosan:PAAm (90:10) blend solutions with ~ 300 nm fiber diameter and < 2% bead density.
 
XPS analysis showed that the surface of PEO/chitosan fibers were somewhat depleted in chitosan content compared to the bulk composition.  However, it was also found PAAm/chitosan fibers were enriched in chitosan content compared to the bulk composition
 
Metal binding and antimicrobial properties in static conditions
Chitosan blend nanofibers were highly effective in binding Cr(VI) metal ions and binding efficiency was a function of % chitosan in blend solution, molecular of weight, % DDA of chitosan and synthetic polymer used in blend solutions. The metal binding capacity in chitosan blend fibers is significantly higher than that observed for similar blend ratio chitosan/PEO blend films.  A 93 µm thick Chitosan/PEO (90:10) blend film showed binding capacity of 0.44 mg chromium/g chitosan whereas the same blend ratio fibers showed 16 mg chromium/g chitosan binding capacity. Electrospun fibers exhibit greater binding capacity due to the high surface area to mass offered by the fibers compared to films.
 
Chitosan blend nanofibers showed 2-3 log reduction in E-coli K-12 micro-organism and reduction efficiency was a function of % chitosan in blend solution. This value is similar to ones obtained for 35 µm thick films of chitosan:PEO blends with similar blend ratios, but the mass of chitosan in films was up to 10 times higher than that in the fibers.  It was found that binding capacity increased with a reduction in fiber size and with increased degree of deacetylation of the chitosan.
 
Dynamic flow studies
Nanofibrous filter media using chitosan based electrospun nanofibers were successfully fabricated by electrospinning onto spunbonded PP non-woven substrates. Dynamic metal binding efficiencies using as little 0.5 gsm of nanofibrous filter media showed promising results (binding capacity up to 35 mg chromium/ g chitosan) for commercial applicability of these filters. The nanofibrous filter media however was unable to achieve desired anti-microbial effectiveness because of the slow reaction between the protonated amine in chitosan and negative components of the bacterial cell wall. Air and water filtration efficiencies of the nanofibrous filter media measured using aerosol and PS beads suspended in water respectively showed high efficiencies which correlated with the fibrous media size and shape. However the nanofibrous layer lacked mechanical strength to with stand pressure applied during the PS bead filtration which affected the results.
 
Technical Details
 
Processing-Structure Relationships
 

Materials

Chitosan with two different molecular weights was used. Chitosan of molecular weight Mv = 1400 kDa (HMW) with varying degree of deactylation (DDA) i.e. 80% DDA, 70% DDA, and 67% DDA was used as received from Primex Inc. Chitosan of lower molecular weight Mw = 100 kDa (LMW) and 83% degree of deactylation was used as received from Sigma.  Polyethylene oxide (PEO, 900 kDa) and polyacrylamide (PAA, 5000 kDa) were used as received from Scientific Polymer Inc.
 

Electrospinning

The electrospinning apparatus consisted of a metered flow pump (Harvard Apparatus Pump II), a high D.C voltage supply (Gamma High Voltage Research, Inc. Model HV ES 30P/DAM), and aluminum foils as targets for fiber collection. Processing conditions were as follows; solution flow rate of 0.08 ml/min, applied voltage of 30 kV, and tip-target distance of 10 cm.  An air assisted heating unit was designed similar to one described by Chun and coworkers1 to heat the polymer solution while it was being ejected through the needle by passing hot air around the needle at flow rates up to 75 ft3/hr, and temperatures ranging from room temperature (25°C) to 70°C.
 

Effect of Solvents and Spinning Solution temperature

Electrospinning of neither HMW nor LMW chitosan at varying concentrations in varying strengths of acetic acid (10% - 90%), hydrochloric acid (0.03N – 0.5N) and trifluroacetic acid (50%) resulted in fiber formation, even when spun at higher temperatures (up to 70°C) as shown in Figure 1. Amongst the three solvents tried initially, further studies were carried out with acetic acid as it was seen as the most promising candidate based on the shape of particles along with appearance of fibrils and the desire to stay away from more toxic solvents like TFA.
 
Figure 1     SEM images of the pure electrospun chitosan samples 1.4 wt% HMW chitosan spun from 50% acetic acid (left), 5 wt% LMW chitosan spun from 90% Acetic Acid (right).
 
Table 1 Summary of processing condition for electrospinning of pure chitosan
 
Type of Chitosan (Molecular weight)
Solvent
Spinning Solution Temperature (°C)
Polymer Concentration
HMW Chitosan
(Mv - 1400 kDa)
0.03N HCl
25, 40, 70
0.6 – 1.5 wt%
0.1N HCl
25, 40, 70
0.1 - 2 wt%
0.5N HCl
25, 40, 70
1.5 wt%
50% TFA
25, 40, 70
1.5 wt%
90% AA
25, 40, 70
1.2 wt% + 1.5 wt% Urea
90% AA
25, 40, 70
1.5 wt%
LMW Chitosan
(Mv - 100 kDa)
0.1N HCl
25, 40, 70
1.7 wt%
90% AA
25, 40, 70
5 wt% with addition of salt
30% AA
25, 40, 70
6 wt%  with addition of salt
Hydrolyzed Chitosan
(Mv – 300kDa)
80% AA
25
5 wt %
Hydrolyzed Chitosan
(Mv – 80kDa)
90% AA
25
4 wt%
Hydrolyzed Chitosan
(Mv – 20kDa)
80% AA
25
5 wt %, 6 wt%

 


Acid hydrolysis of chitosan was done to further reduce the molecular weight of HMW chitosan following a procedure similar to that of Liu et.al2, and chitosan with varying molecular weights (300 kDa, 80 kDa, 20 kDa) was obtained. Electrospinning of these different molecular weights of chitosan also did not result in fiber formation. In order to reduce the amount of inter- and intra-chain hydrogen bonding in chitosan, urea was added which has been shown to disrupt hydrogen bonding in other polysaccharides3. Salt (NaCl) was also added to the electrospinning solution, as it is known that addition of salt helps increase solution conductivity and improve spinnability of polymer solutions4,5.   Table 1 presents a summary of processing conditions that were studied, none of which resulted in fiber formation, instead forming electrosprayed particles.
 

Electrospinning of Chitosan/PEO blends

 

Chitosan, when blended with as little as 10% PEO, resulted in the formation of non-woven mats of fibers. Figure 2 shows SEM images of HMW chitosan blended with HMW PEO with increasing % PEO in the blend. It can be seen with increasing % PEO, fiber diameter increases and number of bead defects is reduced.
 
Figure 3 shows a plot of fiber diameter of electrospun fibers vs. % PEO in the blend solutions. From the fiber diameter data it can be seen that with increasing % PEO, fiber diameter increases which could be due to higher concentration of polymer in solution with increased PEO content. The concentration of polymers in solution was determined by studying solubility of polymer blends at different blend ratios and optimizing them so as to be able to form solutions which could form a stable jet which would lead to formation of fibers. At constant polymer concentration, viscosity of solution decreases with increased PEO content and reduced strength of acetic acid. Fibers formed using high molecular weight chitosan are thinner compared to those obtained using low molecular weight chitosan; this can be attributed to higher solution concentration of low molecular chitosan blends.
 
Figure 2     SEM images of HMW chitosan: HMW PEO blend fibers (a) 1.33 wt% HMW chitosan: HMW PEO (90:10) (b) 1.6 wt% HMW chitosan: HMW PEO (75:25) (c) 2.00 wt% HMW chitosan: HMW PEO (50:50).
 
Figure 3     Fiber Diameter v/s % PEO (Error bars represent standard deviation (n=60), letters indicate difference at p<0.05, wt% in parentheses indicate total polymer wt% in solution)
 

Effect of Spinning Solution Temperature

Solutions were electrospun at higher temperatures by blowing hot air around the feed needle. Figure 4 shows SEM micrographs of electrospun fibers of 1.33 wt% HMW Chitosan: HMW PEO (95:5) electrospun by blowing air at 25 ft3/hr at room temperature (25°C), 40°C and 71°C. It can be seen that as temperature increases, less defective fibers are obtained as indicated by lower bead density values at higher temperatures as shown in Figure 5. The increased temperature reduces the solution viscosity and the flowing air helps increase the spin-draw ratio, both of which together lead to further stretching of the unstable polymer jet during the whipping motion and aiding in the formation of beadless fibers. As air flow and temperature (71°C) are increased, there would be faster evaporation of solvent which would lead to increase in concentration of polymer solution exiting the syringe and formation of slightly thicker fibers. The effect of air flow rate on fiber formation was also studied.  Increasing air flow from 25 ft3/hr to 75 ft3/hr did not have a significant effect on bead density compared to air temperature but increased air flow at higher temperatures led to a slight increase in fiber diameter.
 
Electrospinning of Chitosan/PAAm blends
Figure 6 summarizes the effect of blend ratios and spinning solution temperature on fiber formation. It can be seen that with increasing temperature, the fiber diameter increases slightly and the bead density decreases. SEM images of electrospun solutions containing 95% chitosan (Figure 6a) shows poor fiber formation at room temperature, and very few fibers are collected on the target. As the temperature is increased (Figures 6bc) fiber formation is improved with bead less fibers formed at 70°C. When chitosan
 
 
Figure 4     SEM images of 1.33 wt% HMW chitosan: HMW PEO (95:5) blend fibers at different spinning solution temperature and constant air flow rate of 25 ft3/hr
 
 
 
Figure 5     Bead density of various HMW chitosan: HMW PEO blend fibers at different spinning solution temperature and constant air flow of 25 ft3/hr (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05)
  
content was reduced to 90% (Figures 6def) with increasing spinning temperature, the transformation from beaded fibers to uniform bead free fibers is seen. Further reduction to 75% chitosan in the blend leads to formation of bead free fibers at room temperature (Figure 6g). As spinning temperature is increased (Figures 6hi), an increase in fiber diameter is seen. At higher temperature there is a faster evaporation of solvent leading to faster drying of the charged jet.  Renekar et.al have observed that spinning highly volatile polymer solutions leads to formation of polymer skin on the outside of the jet, subsequently leading to formation of a flat ribbon like structure.6 SEM images of the Chitosan/PAAm fibers formed at high temperature also show some flat ribbon shaped fibers which would contribute to the apparent increase in fiber diameter at higher temperatures.
 
Figures 7 and 8 show quantitatively the fiber diameter and bead density data, respectively, of electrospun chitosan/PAAm solutions at different blend ratios and increasing spinning solution temperatures.
 
Figure 6   SEM images of 1.4 wt% HMWChitosan: PAAm blends at different blend ratios and hot air blown at 25ft3/hr at different temperatures (fig 6a,6b,6c are HMWChitosan: PAAm (95:05) blend ratio, 6d,6e,6f are HMWChitosan: PAAm (90:10) blend ratio, and 6g,6h,6i are HMWChitosan: PAAm (75:25) blend ratio fibers)
 
Figure 7        Fiber diameter of 1.4 wt% HMW chitosan: HMW PAAm blend fibers at different air temperature. (Error bars represent standard deviation (n=60), letters indicate significant difference at p<0.05)
 
Figure 8        Bead density of 1.4 wt% HMW chitosan: HMW PAAm blend fibers at different air temperature. (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05)
 

XPS Results

The surface chemistry of the electrospun fibers was characterized using a Thermo Scientific K-alpha X-ray photoelectron spectrophotometer (XPS) with an Al K-alpha (1480 eV) monochromatic source and 400 micron spot size.  Surface atomic composition (as atomic %) of C, N, and O were used to determine chitosan surface content is terms of weight fraction.  Due to the measured atomic fractions of pure chitosan film differing from the atomic percentages determined from the molecular structure, our results can only be reported as upper and lower bounds.  Calculations based on the atomic fractions in the molecular stucture give a lower bound while those based on the measured atomic fractions of pure chitosan film give an upper bound on chitosan weight fraction.
 
Figure 9 shows the calculated surface chitosan wt%  vs. the wt % of chitosan in blend solution.   The location of the upper and lower bounds indicate that the spun blend fiber surfaces are depleted somewhat in chitosan compared to composition of the spinning solutions.
 
 
Figure 9           Calculated surface chitosan wt% vs. chitosan wt% in solution.
 
Similar analysis of the chitosan/PAAm blend fibers is shown in Figure 10. The surface chitosan content decreases with increasing fiber diameter for all blend ratios; similar to what has been observed for chitosan/PEO blends. However the surface chitosan content is in closer agreement with original chitosan concentration in solution then what was observed for chitosan/PEO blends.  The location of the upper and lower bounds indicate that the surface chisosan content is at least equal to (and possibly greater than) the chitosan composition in the spinning solutions.  It was also found that changing degree of deacytelation of chitosan (67, 70, 80 %) did not have any effect on surface chitosan content.
 
Figure 10      Surface chitosan wt % for chitosan:PAAm blend solutions with increasing % chitosan in solution.
 
Metal Binding and Antimicrobial Properties in Static Conditions

Metal binding - Chitosan/PEO Blends

Figure 11 shows the amount of chromium bound (mg Cr per g chitosan) for blend fibers with different chitosan: PEO blend ratios. HMW chitosan: PEO (90:10) blends show the highest amount of Cr bound per g chitosan. It can be observed that metal binding is strongly related to the % chitosan in the blend solution, and molecular weight of chitosan. With decreasing % chitosan in blend fiber, the binding capacity of the fibers decreased.
 
The binding capacity of blends made using high molecular weight chitosan was seen to be higher. High molecular weight chitosan would offer higher number of available –NH3+ sites for metal binding for similar surface area fibers  due to longer chain lengths and higher degree of protonation7. (XPS results in have also shown higher surface nitrogen concentration for HMW blends compared to LMW blends at same blend ratio). The high binding capacity observed for HMW chitosan fibers could also be result of thinner fibers providing higher surface area compared to LMW chitosan blends. To rule out the possibility of association of PEO with the metal ions an electrospun PEO fiber mat was also tested for metal binding and the results showed no binding.
 
The effect of degree of deacetylation on the metal binding capacity of chitosan/PEO (90:10) blend fibers was also studied. Solutions of 1.33 wt% HMW chitosan of varying degrees of deacetylation (80% DDA, 70% DDA and 67%DDA) with PEO in 75% acetic acid were electrospun to form non-woven mats. The metal binding capacity as shown in Figure 12 was highest at 80% DDA as expected because of the increase in number of available –NH3+ sites for metal binding, and there was no significant statistical difference between the 67% (7.35 mg chromium/g chitosan) and
 
Figure 11     Metal binding of chitosan/PEO blend fibers at different % of chitosan in solution (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05)
 
70% DDA (4.44 mg chromium/ g chitosan) chitosan/PEO blend fibers. The slight increase at 67% DDA (45.5 nm) could be due to thinner fibers formed compared to 70% DDA (62.4 nm) resulting in increased surface area. The metal binding capacity in chitosan blend fibers is significantly higher than that observed for similar blend ratio chitosan/PEO blend films8. A 93 µm thick LMW Chitosan/PEO (90:10) blend film showed binding capacity of 0.44 mg chromium/g chitosan. Electrospun fibers exhibit greater binding capacity due to the high surface area to mass offered by the fibers compared to films.
 

Metal Binding - Chitosan/PAAm Blends

Figure 13 shows the Cr (VI) binding capacity vs. % chitosan in blend of chitosan/PAAm blend nanofibers. The results once again show that the blends containing higher molecular weight chitosan and higher % chitosan in blend fiber show greater binding capacity. Figure 14 shows the Cr (VI) binding capacity vs. fiber diameter for 1.4 wt% HMW Chitosan:PAAm (90:10) blend fibers formed by spinning the solution at varying temperatures. Also plotted on the secondary axis is the surface nitrogen atom % vs the fiber diameter. It can be seen that with increasing fiber diameter the surface nitrogen content does not change, hence the binding capacity remain unaffected by change in diameter.
 
Figure 12   Metal binding of HMW chitosan: HMW PEO (90:10) blend fibers at different DDA (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05).
 
Figure 13    Metal binding of chitosan/PAAm blend fibers at varying % chitosan in blend fiber (Error bars represent standard deviation (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05, n=3)
 
Figure 14    Cr (VI) binding capacity (primary axis), surface nitrogen atom % (secondary axis) vs. fiber diameter for 1.4 wt% HMW Chitosan:PAAm (90:10) blend fibers formed by spinning the solution at varying temperatures. (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05, n=3)
 

Anti-microbial Properties of Chitosan/PEO Blends

In anti-microbial studies usually the reduction in microbial activity is reported on log basis however since all our analysis so far has been based on weight of chitosan we have also plotted data based on reduction in bacteria divided by weight of film. Figure 15 shows a plot of effect of % chitosan in blend fiber and molecular weight of chitosan in blend fiber on the anti-microbial effectiveness of chitosan/PEO blend fibers. As the weight of all the fiber mats studied was not same, the weight of the fiber mats is being plotted on the secondary y-axis to better correlate the anti-microbial properties with fiber structure and composition. We see a 2.5-3.0 log reduction in cfu (colony forming unit) indicating a bacteriostatic effect, this value is similar to ones obtained for 35 µm thick films of chitosan:PEO blends with similar blend ratios, but the mass of chitosan in films was up to 10 times higher than that in the fibers.8  There is no statistical difference in log reduction with increasing PEO content. To better understand the anti-microbial data # of cfu (colony forming units) reduced per g of chitosan was calculated and plotted as shown in Figure 15. When the anti-microbial test data is normalized to weight (Figure 16) then we can observe a trend that with increasing % PEO in blend fiber and decreasing molecular weight of chitosan leads to reduction in anti-microbial properties.
 
The effect of molecular weight on anti-bacterial activity of chitosan is not fully understood, some groups have suggested there is a threshold molecular weight ~ 220 kDa until which the anti-microbial activity increases with increasing chitosan molecular weight. However upon exceeding this threshold molecular weight the anti-microbial activity decreases because they believe the molecules pack more densely leading to increased inter and intra-molecular hydrogen bonding utilizing some of the available protonated amine sites.9 Figure 16 shows a plot of effect of increasing chitosan % DDA for 1.33 wt% HMW chitosan: PEO (90:10) blend fibers. Figure 17 shows the same data normalized to the weight of the fibers i.e. # cfu reduced per gram of chitosan is plotted against DDA. Although one would expect an increase in anti-microbial activity with increasing % DDA because of the increase of # of available protonated amine sites, results from Figures 17 and 18 show the contrary. The slight decrease in anti-microbial activity with increasing % DDA could be because fibers formed at 80% DDA have larger fiber diameter (118 nm) compared to the fibers formed using 70 and 67% DDA (62 and 45 nm respectively). This increase in fiber diameter would lead to greater reduction in # available of protonated –NH3+ amine sites then would be increased by increasing % DDA.
 
Figure 15        Effect of % chitosan in blend fiber and molecular weight of chitosan in blend fiber on the anti-microbial effectiveness of chitosan/PEO blend fibers. (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05, n=3)
 
Figure 16    Effect of % chitosan in blend fiber and molecular weight of chitosan in blend fiber on the anti-microbial effectiveness of chitosan/PEO blend fibers. (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05, n=3)
 
 
Figure 17 Anti-microbial activity of HMW Chitosan:PEO (90:10) blend fibers as function of chitosan % DDA. (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05, n=3)
 
 
Figure 18      Reduction in # of cfu/g of chitosan for HMW Chitosan:PEO (90:10) blend fibers made with increasing chitosan % DDA. (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05, n=3)

 

Anti-microbial Properties of Chitosan/PAAm Blends

 
Table 2 summarizes the anti-microbial efficiencies of chitosan/PAAm fibers. From the data it can be observed that for all samples there was ~ 3 log reduction in bacteria after 6 hrs. The # of cfu reduced per g of chitosan has also been tabulated in Table 2. It can be seen that at same blend ratio with increasing fiber diameter and decreasing chitosan molecular weight there was decrease in anti-microbial efficiency. The physical structure of the fiber mats also after 6 hrs of testing had disintegrated compared to the chitosan/PEO fibers as PAAm is highly hydrophilic which could have affected the test and the results.
 
Table 2 Anti-microbial properties of Chitosan/PAAm blend nanofibers
Sample
log reduction (cfu/ml)
Fiber Diameter (nm)
cfu /
(g chitosan)
Average
Std.Dev
1.4 wt% HMW Chitosan:PAAm (90:10) espun @ 70°C
3.34
0.12
305
2.14E+13
1.4 wt% HMW Chitosan:PAAm (75:25) espun @ 25°C
3.11
0.35
132
2.61E+13
1.4 wt% HMW Chitosan:PAAm (75:25) espun @ 70°C
3.17
0.19
328
2.47E+13
2.85wt%LMW Chitosan:PAAm (75:25) espun @ 70°C
3.15
0.04
421
1.96E+13
 
Dynamic Flow Studies
 
The goal of this research was to develop chitosan based nanofibrous filtration media which possess enhanced filtration efficiencies owing to the positive charge on filter fiber surface and size effect of nanofibers. As seen previously, nanofibers with higher chitosan % in blend solution (90%), higher molecular weight (HMW Chitosan) and higher degree of deacetylation (80% DDA) exhibited the highest metal binding and anti-microbial efficiencies for both chitosan/PEO and chitosan/PAAm blends.
 

Fabrication of Chitosan Blends Nanofibrous Filter Media

A nanofibrous filter media comprising of a top layer of chitosan blend nanofibers electrospun on a spunbonded non-woven polypropylene (PP) fiber substrate was fabricated. Spunbonded PP was used a substrate to provide mechanical and structural support to the thin layered electrospun nanofibers. Initially melt-blown nonwoven PP mats were chosen as substrate material as melt-blown mats have thinner fibers and lower pore size compared to spunbonded nonwovens (Fiber diameter and pore size of melt blown PP = 3.2 µm (±1.17 µm), 13.83 µm respectively, fiber diameter and pore size of spun bonded PP = 19.6 µm (±1.33 µm), 47.46 µm respectively). However, it was not possible to electrospin a continuous layer of chitosan fibers on melt-blown PP webs, possibly due to the dense nature of the PP mat acting as an insulator and repelling the charged electrospun fibers away to the surrounding metallic aluminum plate on which it was laid.
 
Electrospinning of chitosan blend solutions on spunbonded PP substrates led to successful fabrication of chitosan based nanofibrous filter media. Filter media of both HMW chitosan/PEO and HMW chitosan/PAAm blends were fabricated with 90% chitosan in blend solution.  Variables included web density (achieved by spinning for different time intervals), fiber diameter and different DDA chitosan (only for chitosan:PEO blends).
 
To obtain varying fiber diameter HMW chitosan:PEO blends the strength of the acid solution was varied and a non-ionic surfactant Brij-35 (polyoxyethyleneglycol dodecyl ether) was used. We10 have shown that increasing strength of acid reduces solution surface tension with an  increase in solution viscosity and addition of 2mM brij-35 leads to increase in solution viscosity with slight increase in solution conductivity and surface tension. Thicker fibers are formed by spinning 1.33 wt% HMW chitosan:PEO (90:10) blends with increasing strength of acetic acid from 75% to 90% and addition of 2mM brij-35 as shown in Figure 19. To obtain HMW chitosan:PAAm (90:10) fibers of varying fiber diameter, solutions were made and electrospun as at different solution temperatures.
 
Figure 19       Increase in fiber diameter with strength of acid in solvent and addition of surfactant (Error bars represent standard deviation (n=60), letters indicate significant difference at p<0.05)
 
 
Figure 20        mg chromium bound/g chitosan for HMW chitosan:PEO blends after each pass for 10 passes. (Error bars represent standard deviation (n=3))
 

Metal Binding Efficiency of Chitosan Blends Nanofibrous Filter Media

The dynamic metal binding properties of chitosan blend fiber mats were measured by passing 100 ml of a 5 mg/l K2CrO4 solution through chitosan nanofibrous filter media ten consecutive times.  Samples were taken after each pass to determine reduction in solution chromium concentration. Figure 20 shows the binding capacity achieved after each pass using 1.5 gsm 1.33 wt% HMW chitosan: PEO (90:10) blend fiber mats. It can be seen that with increasing pass number binding efficiency increased. Therefore, for all further tests binding efficiency was only determined after the 5th and 10th passes.
 
Figure 21 shows the Cr (VI) binding capacity of HMW chitosan: PEO blend fibers as function of fiber diameter using 0.5 gsm and 1 gsm chitosan nanofibers. It can be seen with increasing fiber diameter binding capacity decreases or remains statistically unchanged. The binding capacity is statistically similar but shows a decreasing trend with increasing fiber diameter. We estimated that there should be a 30 % drop in binding capacity between the fiber diameters studied which is well within the standard deviation of the obtained results.
 
Figure 21    Effect of fiber diameter on binding capacity of different gsm HMW chitosan:PEO (90:10) nanofibrous filter media. (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05)
 
Figure 22 shows the binding capacity as function of fiber diameter for 1gsm HMW chitosan: PAAm (90:10) blend fibers and there is no statistical difference in binding capacity with increasing fiber diameter.
 
 
Figure 22          Effect of fiber diameter on binding capacity of different gsm HMW chitosan:PAAm (90:10) nanofibrous filter media. (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05)
 
Figures 23 and 24 show the binding capacity of HMW chitosan:PEO (90:10) and HMW chitosan: PAAm nanofibrous filter media of increasing basis weight (gsm) respectively (constant fiber diameter). It is observed that with increasing gsm for both blend fibers a slight decrease in binding capacity is observed which is statistically mostly insignificant. The % chromium bound does not change with increasing fiber mat gsm (% chromium bound after 10 passes for 0.5 gsm web = 5.67%, 1 gsm web = 6.77% and 1.5 gsm web = 5.4%) which means that the binding efficiency of the fibers is constant irrespective of the basis weight of the mat. The binding activity could be restricted only to the top layers of the espun fiber mat.  
 
Figure 23       Effect of gsm on binding capacity of HMW chitosan:PEO (90:10) nanofibrous filter media. (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05)
 
 
Figure 24       Effect of gsm on binding capacity of HMW chitosan:PAAm (90:10) nanofibrous filter media. (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05)
 
Figure 25      SEM images of 1.33 wt% HMW chitosan:PEO (90:10) (left) and 1.4 wt% HMW chitosan: PAAm (90:10) (right) nanofibrous filter media after washing with water.
 
As shown in Figure 25, upon drying of these samples, a film has formed on the surface of the fibers. The formation of this film could restrict the binding of chromium to the top few layers and prevent the solution effectively wetting the entire depth of the fiber mat.  Characterizing the nanofibrous filter media after repeated washing with water using SEM and XPS it is seen that although a fibrous structure is distinctly visible it is covered by a layer of polymer film. XPS results show that this film-like layer seen is rich in chitosan.  It is possible that during the binding experiments in aqueous medium, chitosan may partially dissolved or swell in water.  This dissolution of material leads to swelling or partial dissolution of fibrous structure allowing the diffusion of chromium ions to the bulk of the fiber and opening up additional NH3+ sites for binding. Upon drying of this wet nanofibrous filter media, formation of polymer film layer appears on top of the electrospun fiber. This may explain why we were not able to trace Cr content after metal binding experiments using XPS as the Cr bound fibers may be shielded by this film layer. The swelling of the fiber surface could lead to opening of the fiber structure formation of film like layer and diffusion of chromium ions to the bulk of the fiber reducing the impact of surface chitosan on binding.
 
Figure 26 shows that the binding capacity of these nanofibrous filter media increased with increasing % DDA.  For the 67% and 70% DDA chitosan blend fibrous media there is no increase in binding between 5 and 10 passes as can be seen in the 80% DDA sample. Fibrous media fabricated with lower % DDA chitosan could become saturated with chromate ions before or after pass # 5 compared to the 80% DDA chitosan fibrous media.
 
Figure 26       Effect of chitosan % DDA on binding capacity of varying DDA HMW chitosan:PEO (90:10) nanofibrous filter media. (Error bars represent standard deviation (n=3), letters indicate significant difference at p<0.05)
 

Anti-microbial Properties of Chitosan Blends Nanofibrous Filter Media

The concentration of Escherichia coli K-12  test microorganism had to be reduced (104 cfu/ml) compared to the static tests (107 cfu/ml) because the higher concentration of bacteria overwhelmed the nanofibrous mat and no solution passed through the filter membrane, even after 3 hrs. Figure 27 shows the effect of fiber diameter, nanofiber gsm and chitosan DDA on anti-microbial activity of HMW chitosan:PEO (90:10) blend fibers after 1 pass of 104 cfu/ml of E-coli K-12. It can be seen that < 0.5 log reduction in bacteria is observed for all samples after ~ 2 mins of contact of fiber with bacterial solution. To understand the kinetics of the anti-microbial activity of chitosan we did a time dependant test wherein bacterial survival after 2mins, 15 mins, 30 mins, 1hr, 2 hr, 4 hr and 6 hr was measured for 1 gsm HMW chitosan:PEO (90:10) blend fibers soaked in 107 cfu/ml bacteria solution and results are shown in Figure 28. It can be seen that up to 2 hrs there is < 1 log reduction in bacteria and increased activity > 2 log really happens after 4 hrs.  Whatever reduction in bacteria was seen in the dynamic filtration test is due to the size effect of the nanofiber which can trap the approximately 0.5 micron sized E-coli bacteria.
 
Figure 27  Log reduction in E-coli test micro-organism after 1 pass of 100 ml bacteria solution through different gsm, diameter and %DDA chitosan/PEO nanofibrous filter media. (Error bars represent standard deviation (n=3))
 
Figure 28       Log reduction in E-coli test micro-organism after soaking 1gsm HMW chitosan/PEO (90:10) nanofibrous filter media for different times in bacteria solution. (Error bars represent standard deviation (n=3))
 

Latex PS Bead Filtration Efficiency of Chitosan Blends Nanofibrous Media

The applicability of chitosan based nanofibrous filter media to effectively filter out heavy metal ions and micro-organism from pollutant water streams based on the polycationic nature of chitosan has been demonstrated. The particle filtration efficiency of chitosan based nanofibrous filter media was characterized by passing 10 ml of 3 micron sized 200 ppm polystyrene beads through filter media of varying fiber diameter and fiber gsm.
 
Figure 29 shows the SEM images of HMW chitosan/PEO (90:10) blend nanofibrous filter media before and after passing PS beads. It can be seen that the fiber mats appear to be torn after filtration. The mechanical integrity of the mat could have been affected by the pressure exerted by the applied vacuum (~ 2 mm Hg) on the filter membrane during the experiment. Figure 30 shows the filtration efficiency of HMW chitosan/PEO (90:10) blend nanofibrous filter media of varying fiber gsm and fiber diameter. With increasing fiber diameter, the PS bead filtration efficiency decreased. This could be due to higher maximum pore size observed with increasing fiber diameter (measured max. pore size of 1 gsm 65 nm diameter fiber = 1.95 µm, measured max. pore size of 1 gsm 110 nm diameter fiber = 2.5 µm).  In order to better characterize the filtration efficiency, experiments were conducted without applying vacuum to the filter media for filtration and varying the fiber media gsm. A 1 gsm nanofibrous filter media of 92 nm fiber diameter and 1.562 microns maximum pore size achieved a 50% filtration efficiency and a similar 3 gsm nanofibrous filter media exhibited 70% filtration efficiency.
 
Figure 29      SEM images of 1 gsm HMW chitosan:PEO nanofibrous filter media before and after passing 10 ml of 200 ppm 3 µm PS beads.
 
Figure 30      PS bead removal efficiency of varying fiber diameter and fiber gsm HMW chitosan:PEO nanofibrous filter media.
 

Aerosol Filtration Efficiency of Chitosan Blends Nanofibrous Media

The aerosol filtration efficiency was measured using a TSI Corp. model 8130 automated filtration testing unit at UTNRL. NaCl aerosol particles of 0.26 µm mean diameter, 0.075 µm count median diameter and concentration of 15 to 20 mg/m3 were used. The penetration and pressure drop across the 7 in x 7 in chitosan based nanofibrous filter media was measured.  The aerosol filtration efficiency of 1 gsm HMW chitosan: PEO (90:10) fibers of varying fiber diameters is shown in Figure 31.  It can be seen that with increasing fiber diameter, the filtration efficiency decreased because the maximum pore size and air permeability increased (Table 3). SEM images of the fiber sample before and after filtration showed no damage to the electrospun layer.
 
Table 3 Air permeability data of 1 gsm HMW chitosan:PEO blend fibers
Fiber Diameter (nm)
Air Permeability (cfm)
64.87
1.29 (± 0.39)
91.04
1.21(± 0.35)
109.84
4.23(± 0.65)
 
Figure 31      Aerosol filtration efficiency and maximum pore size of 1 gsm HMW chitosan:PEO (90:10) nanofibrous filter media. (Error bars represent standard deviation (n=3))
 
Effectiveness and Feasibility
 
This work has demonstrated that non-toxic, biodegradable chitosan nanofibers can be produced quite easily by electrospinning when blended with small amounts of flexible, water soluble polymers such as poly(ethylene oxide) and polyacrylamide.  Filters produced from these fibers can effectively and quickly remove heavy metals and fine contaminants from aqueous environments.   In addition, these materials also possess strong anti-microbial properties when contact with the environment is on the order of hours.  Some additional work is necessary before the application of these filter materials is feasible.  First, the tendency for these fibers to form film-like coatings in an aqueous environment needs to be eliminated in order to increase filtration and metal binding efficiencies.  Second, the mechanical strength of the fiber mats must be increased so that tearing does not occur during use.  Partial crosslinking of the chitosan and/or selection of alternative blend polymers are likely solutions to these challenges.

References:

1.     Um, I. C.; Fang, D. F.; Hsiao, B. S.; Okamoto, A.; Chu, B., Electro-spinning and electro-blowing of hyaluronic acid. Biomacromolecules 2004, 5, (4), 1428-1436.
2.     Liu, N.; Chen, X. G.; Park, H. J.; Liu, C. G.; Liu, C. S.; Meng, X. H.; Yu, L. J., Effect of MW and concentration of chitosan on antibacterial activity of Escherichia coli. Carbohydrate Polymers 2006, 64, (1), 60-65.
3.     Cho, J.; Heuzey, M. C.; Begin, A.; Carreau, P. J., Effect of urea on solution behavior and heat-induced gelation of chitosan-beta-glycerophosphate. Carbohydrate Polymers 2006, 63, (4), 507-518.
4.     Fong, H.; Chun, I.; Reneker, D. H., Beaded nanofibers formed during electrospinning. Polymer 1999, 40, (16), 4585-4592.
5.     McKee, M. G.; Hunley, M. T.; Layman, J. M.; Long, T. E., Solution rheological behavior and electrospinning of cationic polyelectrolytes. Macromolecules 2006, 39, (2), 575-583.
6.     Sureeporn Koombhongse; Liu, W.; Reneker, D. H., Flat polymer ribbons and other shapes by electrospinning. Journal of Polymer Science Part B: Polymer Physics 2001, 39, (21), 2598-2606.
7.     Wang, Q. Z.; Chen, X. G.; Liu, N.; Wang, S. X.; Liu, C. S.; Meng, X. H.; Liu, C. G., Protonation constants of chitosan with different molecular weight and degree of deacetylation. Carbohydrate Polymers 2006, 65, (2), 194-201.
8.     Li, J.; Zivanovic, S.; Davidson, M.; Kit, K., Surface properties of chitosan/PEO blend films as affected by film preparation method. Abstracts of Papers of the American Chemical Society 2007, 234
9.     Shimojoh, M.; Masaki, K.; Kurita, K.; Fukushima, K., Bactericidal effects of chitosan from squid pens on oral streptococci. Nippon Nogeikagaku Kaishi-Journal of the Japan Society for Bioscience Biotechnology and Agrochemistry 1996, 70, (7), 787-792.
10.   Kriegel, C.; Weiss, J.; Kit, K., Influence of surfactants on electrospinning of chitosan-poly(ethylene oxide) blend nanofibers. Abstracts of Papers of the American Chemical Society 2008, (042).

Journal Articles:

No journal articles submitted with this report: View all 13 publications for this project

Supplemental Keywords:

RFA, Scientific Discipline, Waste, Water, Sustainable Industry/Business, Physics, Remediation, Sustainable Environment, Technology for Sustainable Environment, New/Innovative technologies, Chemistry and Materials Science, Engineering, Chemistry, & Physics, Environmental Engineering, detoxification, industrial wastewater, in situ remediation, membranes, remediation technologies, nanotechnology, antimicrobial nanostructured membranes, environmental sustainability, metal binding, nanocatalysts, catalytic studies, environmental chemistry, environmentally applicable nanoparticles, aquifer remediation design, groundwater remediation, electrospun nanofiber chitosan membranes, groundwater contamination, innovative technologies, ultrafiltration, membrane-based nanostructured metals, pollution prevention, contaminated aquifers, membrane technology, disinfection, transition metal carbides

Progress and Final Reports:

Original Abstract
  • 2006 Progress Report
  • 2007
  • 2008 Progress Report