Conferring materials with antibacterial properties 2

CHAPTER 4 ANTIBACTERIAL AND ADSORPTION CHARACTERISTICS OF ACTIVATED CARBON FUNCTIONALIZED WITH QUATERNARY AMMONIUM MOIETIES... At the same time, we and other groups have reported that a

CHAPTER 4

ANTIBACTERIAL AND ADSORPTION CHARACTERISTICS OF ACTIVATED CARBON FUNCTIONALIZED WITH QUATERNARY AMMONIUM

MOIETIES

4.1 Introduction

Activated carbon (AC) has been widely used in air pollution control and wastewater treatment to remove various pollutants because of its large surface area and high adsorption capacity (Bansal and Goyal, 2005; Quinlivan et al., 2005) It has diverse applications ranging from filters in gas masks and big ventilation systems (Soares et al., 1995) to water treatment systems of hospital renal haemodialysis care units (Morin, 2000) It has been reported that bacteria attached to carbon particles are highly resistant to disinfection processes due to biofilm formation, which causes the carbon itself to become a source of bacterial contamination (Morin, 2000; Stewart et al., 1990) Hence, it would be advantageous if the AC also possesses antibacterial activity to kill air- or water-borne bacteria The preparation of antibacterial ACs has been attempted and much effort has been devoted to impregnation with silver or metal oxides (Oya et al., 1993; Park and Jang, 2003; Tamai et al., 2001; Wang et al., 1998; Zhang et al., 2004) in the AC Though silver and metal oxides have attractive antibacterial activities, their primary shortcoming is that the particles are easily washed out since they are just deposited on the surface of the AC (Wang et al., 1998; Zhang et al., 2004) Furthermore, with increasing silver content the specific surface area of the carbon decreases greatly, resulting in reduced adsorption capability (Wang et al., 1998)

A number of quaternary ammonium compounds are known to exhibit good bactericidal properties (Kenawy et al., 1998; Thorsteinsson et al., 2003) At the same time, we and other groups have reported that antibacterial properties can be confered

on surfaces of substrates by the covalent attachment of quaternary ammonium with resultant bactericidal activities similar to that of free quaternary ammoniums (Cen et al., 2003; Gottenbos et al., 2002; Hu et al., 2005; Tiller et al., 2001) In this approach,

the antibacterial agents will not leach out from the surface, hence providing long term effectiveness In this part of the thesis, we describe how two types of quaternary ammonium compounds can be covalently attached to the surface of the AC The surface chemical compositions of the modified ACs were analyzed by XPS and the characteristics and morphologies of the AC were investigated using surface area and pore size analysis, and SEM The antibacterial properties of the functionalized AC

against Gram-negative E coli and Gram-positive Staphylococcus aureus (S aureus)

were evaluated Since the surface functionalization process employed entails changing the nature of the surface of the AC, the issue of whether this results in compromising the adsorption capacity of the AC needs to be addressed As such, phenol adsorption

by the AC before and after the functionalization process was also assessed

4.2 Experimental

Materials and reagents

A granular AC (20-40 mesh) purchased from Aldrich was used as the starting material 3-(trimethoxysilyl)-propyldimethyloctadecylammonium chloride (QAS) was obtained from Dow Corning Poly(4-vinyl pyridine) (PVP) (Mw of 160,000 g/mol), hexylbromide, phenol and all other chemicals and solvents were from Aldrich

Preparation of functionalized ACs

The AC was thoroughly washed with water until the washing liquid attained a constant

pH of 6.2 It was then dried at 120 oC for 12h before being subjected to the functionalization process An overall scheme of the functionalization process is shown

in Figure 4.1 In a typical experiment, 5.0 g of AC was added to concentrated HNO3(69 wt% HNO3) at a ratio of 1 g/10 ml to achieve oxidation of its surface The mixture was stirred for 10 h under reflux The oxidized carbon was washed with water until no further change in pH could be detected and then dried under vacuum for 12 h at 60 oC, resulting in 4.4 g of oxidized AC, AC-COOH In the next step, the surface carboxylic acid groups were allowed to react with thionyl chloride to generate acid chloride groups In a typical experiment, 3 g of dry AC-COOH was dispersed in 20 ml of

thionyl chloride and the mixture was stirred at room temperature for 4 h The solid was then separated by filtration and washed with anhydrous CHCl3 Subsequently it was dried under vacuum at room temperature for 6 h, obtaining AC-COCl (2.8 g)

Two types of quaternary ammonium compounds, QAS and poly(vinyl-N- hexylpyridinium bromide), respectively were used to functionalize the as-prepared carbon surface For the attachment of QAS, 2 g of dry AC-COCl were allowed to react

with 20 ml glycol in the presence of triethylamine The suspension was stirred at room temperature for 10 h, and the resulting solid was separated and washed with dry THF and CH2Cl2 After repeated washings, the solid was dried overnight under vacuum, obtaining AC-OH (1.8g) Subsequently 1.5 g of AC-OH was reacted with 40 ml of 10 wt% QAS in water at 80 oC for 24 h QAS possesses a silyl group, which can be covalently bound to AC-OH surface (Gottenbos et al., 2002) The resulting solid was washed with water and dried in vacuum This functionalized AC will be denoted as Q-AC in the subsequent discussion

For the binding of hexyl-PVP, 2 g of dry AC-COCl and 1 g of 3-bromopropylamine hydrobromide were reacted in 20 ml dry CH2Cl2 in the presence of triethylamine The resulting solid (AC-Br) was washed and dried in the same manner as mentioned above Subsequently, 1.5 g of dry AC-Br was placed in a solution of 6 g of PVP in 60 ml of nitromethane/hexylbromide (10:1, v/v) The reaction mixture was stirred at 75 oC for

24 h, and the resulting solid was washed with acetone and methanol followed by drying under vacuum Under the conditions used, only a few pyridine groups of the PVP chain are alkylated by the Br of 3-bromopropylamine, with the majority being alkylated by hexylbromide (Tiller et al., 2002) This functionalized AC will be denoted

as P-AC in the subsequent discussion

Materials and surface characterizations

FTIR and XPS analysis was carried out as described in Section 3.1.2 The surface morphology of the carbons was visualized using a FE-SEM (JEOL, JSM-6700F) The

specific surface area (SBET) was determined using a NOVA 3000 BET Analyzer (Quanta-Chrome) according to the Brunauer-Emmett-Teller (BET) method Before the adsorption isotherms were obtained, the activated carbon samples were purged with pure nitrogen gas overnight at a temperature of 80 °C to remove any contaminant and

moisture that may have been present in the samples The pore volumes (V) were

Figure 4.1 Schematic representation of the two routes for the

functionalization of AC with quaternary ammonium groups

the saturation pressure) of 0.99 The pore size distribution was obtained by applying the density-functional-theory (DFT) method

The amount of N+ immobilized on the ACs was determined by the modified dye interaction method as described in Section 3.1.2 The long-term stability of the quaternary groups on Q-AC and P-AC was assessed using the water abrasion test 1g

of the functionalized AC was stirred in 30 ml of water at 300 rpm and at room temperature After specific time intervals, the AC was filtered About 0.1g of the AC was removed each time, dried overnight under vacuum and the N+ concentration was determined as described above The water abrasion experiment was continued using the remaining AC under the same conditions

Determination of antibacterial activity of functionalized ACs

E coli and S aureus were cultivated as described in Section 3.1.2 and assays of

antibacterial activities were carried out as described in Section 3.2.2

Antibacterial efficacy in repeated applications was investigated using the same 100 mg

of functionalized ACs in consecutive antibacterial assays against E coli After one

batch, the particles were recovered using a filter The particles were shaken in water for 30 min and then used for the next batch of assay

Adsorption of phenol in aqueous solution

Experiments for the determination of the adsorption isotherms of phenol were carried out as follows A known mass of the AC was mixed with aqueous solutions of phenol

of different initial concentrations without adding any chemical to control the pH The

suspensions were shaken at 30 oC for 3 days and the phenol solution was then filtered (Nevskaia et al., 2004) Preliminary kinetic experiments indicated that adsorption equilibrium was reached in less than 3 days for the ACs The concentration of phenol was analyzed using a UV-visible spectrophotometer (UV-1601, Shimadzu, Japan) at a wavelength of 270 nm In order to reduce measurement errors, the UV absorption intensity of each equilibrium solution sample was measured three times and the average value was used to calculate the equilibrium concentration based on a standard calibration curve The adsorbed quantity was determined as the difference between the initial amount of phenol in the solution and the remaining amount after equilibration

The amount adsorbed at equilibrium, q e (mg/g), was calculated using the following equation:

where C 0 and C e are the initial and equilibrium concentrations of phenol (mg/L),

respectively, V is the volume of the solution and W is the mass of the carbon, q e (mg/g)

is the amount adsorbed at equilibrium concentration C e (mg/L) The adsorption equilibrium data were fitted to the Langmuir equation:

where qmax is the maximum adsorption capacity and b is the Langmuir constant qmax

can be obtained by linear regression of (C e /qe) against C e

W

V C C

q e= ( 0− e)

e

e e

bC

bC q q

+

=1

max

4.3 Results and Discussion

cm-1 indicates an abundance of carboxyl-carbonate structures (Moreno-Castilla et al., 1995; Biniak et al., 1997; Shim et al., 2001; Tamai et al., 2006) These features clearly show that the oxidation of the AC by HNO3 generates a large number of carboxyl groups The carboxyl groups were then activated by thionyl chloride for the subsequent reaction to attach the quaternary ammonium groups

The success of the covalent attachment of the two types of quaternary ammonium compounds on AC can be ascertained by comparing the XPS spectra before and after

the functionalization Figure 4.3(a-c) shows the XPS wide scan of AC, Q-AC and P-AC, respectively The corresponding N 1s core-level spectra of these ACs are shown

in Figure 4.3(d-f) The N 1s peak component is not discernible in the wide scan and core-level spectra of AC (Figure 4.3(a) and (d)) In the case of Q-AC (Figure 4.3(b)) the appearance of a strong N 1s signal at a binding energy of 400 eV, Si 2p signal at

100 eV and Cl 2p signal at 200 eV is consistent with the presence of the QAS (structure given in Figure 4.1) The N 1s core level spectrum of Q-AC (Figure 4.3(e)) has a predominant peak at 401.7 eV attributable to the positively charged nitrogen (N+)

In the case of P-AC, the wide scan indicates the presence of N and Br (at 70 eV) However, the wide scan does not offer conclusive evidence for the presence of the desired poly(vinyl-N-pyridinium bromide) since the reaction of AC-COCl with 3-bromopropylamine would also result in the presence of N and Br on the surface of the AC (see Figure 4.1) This issue was further clarified from the N 1s and Br 3d core-level spectra The N 1s core-level spectrum (Figure 4.3(f)) can be deconvoluted into a predominant N+ peak at 401.7 eV and an additional peak at a binding energy of 398.5 eV due to =NH- of the pyridine groups which were not quaternized As the probing depth of the XPS technique is < 10nm, the peak attributed to -NH- of 3-bromopropylamine is not readily discernible due to coverage by the PVP The degree

of N-alkylation of the pyridine groups on the P-AC, as calculated from the [N+]/[N] ratio, is 85%, ie 15% of N introduced on the AC by PVP was not quaternized during reaction with hexylbromide In the Br 3d core-level spectrum (Figure not shown), the presence of a doublet (Br 3d5/2 and Br 3d3/2) at 67.5 and 68.6 eV attributable to Br-

species confirms the reaction between the PVP and 3-bromopropylamine and hexylbromide which converts the -Br component to Br-

Figure 4.4 shows surface morphologies of the ACs before and after functionalization The SEM images of Q-AC and P-AC were similar in appearance to the AC image and hence there is little difference in the surface morphology of the ACs This implies that HNO3 oxidation and post-treatments of the AC did not cause any apparent change in the surface morphology

Figure 4.3 XPS wide scan of AC (a), Q-AC (b) and P-AC (c); N 1s core-level

spectra of AC (d), Q-AC (e) and P-AC (f)

Binding Energy (eV)

(b)

(c) (a)

Figure 4.4 FE-SEM micrographs of AC (a), Q-AC (b) and P-AC (c)

Antibacterial effect of functionalized ACs

The antibacterial activities of ACs before and after functionalization were investigated

by contacting the ACs with bacterial cells in suspension In each antibacterial experiment, 100 mg of ACs was incubated with 30 ml of the bacterial suspension The results of the experiments conducted with suspensions containing 107 cells/ml of the

Gram-negative bacterium, E coli are shown in Figure 4.5(a) With AC, the number of

viable cells in the suspension decreased by 20% or less after 180 min compared to the 10% decrease observed in the control experiment without any AC In the presence of Q-AC, the viable cell number decreased by two orders of magnitude after 1 h, and this reduction was further enhanced with time After 3 h, the viable cell number decreased

by more than five orders of magnitude With P-AC, the antibacterial efficacy is even higher The number of the viable cells decreased by three orders of magnitude after 1 h, and after 3 h, no cells remain viable

The antibacterial activity against Gram-positive S aureus was also evaluated in the

same manner with cultures containing 107 cells/ml and the results are shown in Figure 4.5(b) AC again showed no significant antibacterial activity but the functionalized

ACs have a higher antibacterial activity against S aureus as compared to E coli For

Q-AC, no cells remain viable after 3 h while for P-AC no cells remain viable after 2 h This result is consistent with a recent report by Tiller et al (2002) that an immobilized quaternary ammonium compound is less bactericidal against Gram-negative bacteria

such as E coli than Gram-positive S aureus This difference in antibacterial efficacy is postulated to be the result of the differences in cell membrane structures between the E coli and S aureus bacteria The multilayered cell envelope structure of Gram-negative

bacteria would be more resistant to access by the bactericidal moieties to the inner

0 1 2 3 0

2 4 6 8

Figure 4.5 Viable E coli cell number (a) and viable S aureus cell number (b) as a

function of time in contact with the different substrates: Control (■), AC (●), Q-AC

(Œ), and P-AC (▲) The cell number was determined by the spread plate method

membrane of the organism Based on the above results which show that the functionalized ACs are effective against both Gram-negative and Gram-positive bacteria, it is reasonable to expect that many other bacterial species will also be susceptible

It should be mentioned that the killing potency depends on both the type of the quaternary ammonium compound as well as the structure of the bacteria Although the mechanism of the antibacterial activity of immobilized quaternary ammonium groups

is not entirely clear, it has been hypothesized that these immobilized moieties disrupt the integrity of the cytoplasmic membrane to cause cell death, similar to the mechanism of free biocides The role played by the hydrophobic chain containing the quaternary ammonium group in the antibacterial effect has been studied Tiller et al found that pyridine groups N-alkylated with C6 showed the highest killing efficacy for

S aureus and E coli, followed by those with C3 and C4 chains while the C8-C16

chains are significantly less effective (Lin et al., 2002; Tiller et al., 2001) However, it has been reported that for soluble quaternary ammonium compounds such as QAS, the antibacterial effects increase with the length of the alkyl moieties attached to the nitrogen atom (Ahlstrom et al., 1995; Ahlstrom et al., 1999; Calas et al., 1997), with an optimum chain length of 16-18 carbon atoms (Lindstedt et al., 1990) In earlier work,

we have shown that the concentration of quaternary ammonium groups on the surface has a very significant effect on antibacterial efficacy (Cen et al., 2003)

The quantitative amounts of N+ present on the Q-AC and P-AC as obtained from the titration method using fluorescein (Na salt) are shown in Table 4.1 It can be seen that the amount of N+ on the Q-AC is almost double that on P-AC, but, as indicated by

Figure 4.5, the antibacterial activity of Q-AC is lower than that of P-AC This apparent

discrepancy may be due to the different types of quaternary ammonium groups present

on Q-AC and P-AC The contributing action of the hydrophobic polymer chains inherent in the polymeric structures present on P-AC enhances the cell membrane

disruption mechanism as compared to the smaller amphiphilic molecules of QAS on

Q-AC (Kuroda and DeGrado, 2005) Furthermore, QAS, being a smaller molecule,

may be present in the micropores (< 2nm) to a greater extent than the larger poly(vinyl-N-hexylpyridinium) molecules It will be shown in the next section that a

significant fraction of micropores are indeed filled by the QAS Since bacteria are

micron-sized, the presence of antibacterial groups in micropores would not contribute

to the killing of the bacteria

Table 4.1 Characteristics of ACs before and after functionalization

The antibacterial efficacy of Q-AC and P-AC in repeated applications against E coli

was also investigated The results for P-AC in Figure 4.6 show a slight decrease in

efficacy after 10 repeats of antibacterial experiments Similar results were obtained

with Q-AC It has been proposed that bacterial cells can adsorb on solid surfaces by

electrostatic or hydrophobic interaction, or both (An and Friedman, 1998; Ista et al.,

2004; Nakagawa et al., 1984) The decrease in bactericidal activity may be due to the

accumulation of dead cells or cell debris on the surface of the ACs, which would then

reduce the interaction of the quaternary ammonium groups with bacterial cells If the P-AC and Q-AC after 10 repeats of antibacterial experiments were washed using

ethanol and then exposed to a new bacterial culture, the antimicrobial activities of Q-AC and P-AC will be restored to its original level It has been reported that quaternary ammonium on porous glass in the range of 19-72 μmol/g can completely recover its antibacterial ability on being washed with ethanol However, if the quaternary ammonium is in the range of 880-1170 μmol/g, its antibacterial activity can only be recovered upon treatment of the functionalized glass with alkaline solutions due to the stronger electrostatic interaction with the dead cells (Nakagawa et al., 1984)

In this work, the N+ concentration of Q-AC and P-AC is 32 and 18 μmol/g, respectively, and washing with ethanol is sufficient to reduce the electrostatic

0 2 4 6

8

1st experiment 10th experiment

Contact Time (h)

Figure 4.6 Repeated antibacterial assay using 100 mg of P-AC in contact with 30

ml E coli suspension (107 cells/ml)

interaction between the dead cells and ACs The long term stability of the Q-AC and P-AC was also assessed using the water abrasion test The N+ concentrations on both the Q-AC and P-AC surface showed only a slight decrease with time, and after 7 days

of the water abrasion test, the N+ concentrations are > 91% of the value of the respective freshly functionalized sample This indicates that the antibacterial groups are stable on the surface of the functionalized AC, and hence long term antibacterial effectiveness can be expected

Phenol adsorption

ACs have been widely used for the adsorptive removal of hydrocarbons such as phenols and their derivatives which are frequently present in many wastewater streams (Bertoncini et al., 2003) Since phenol has the adsorption characteristics of small polar aromatic compounds, the adsorption of phenol by the different ACs was investigated

to assess how the surface functionalization process has affected their adsorption capacity Figure 4.7 shows the adsorption isotherms of phenol on AC, Q-AC and P-AC The maximum adsorption capacity for the three samples as calculated from these

isotherms, qmax, are given in Table 4.1 The results show that the functionalization process for preparing Q-AC significantly reduces its phenol adsorption capacity whereas the phenol adsorption capacity of P-AC is only slightly decreased from that of the original AC

Figure 4.7 Adsorption isotherms of phenol on AC (●), Q-AC (Œ) and P-AC (▲)

The phenol adsorption capacity depends on various factors such as surface area, pore size distributions and surface chemistry (Villacanas et al., 2006) In general, the adsorption capacity depends, to a large extent, on the accessibility of the organic molecules to the inner surface of the carbon adsorbent Since phenol is a small molecule (molecular size < 1nm), the adsorption occurs mainly in the micropores The micro volume and BET surface area (Table 4.1) of Q-AC are significantly lower than those of AC as a result of the functionalization process From the pore distribution results given in Figure 4.8, changes in the internal structure of Q-AC are clearly indicated It is possible that many of the pores of Q-AC have been filled by QAS, which is a small molecule In addition, the quaternary ammonium hydrophilic groups

of QAS can increase the polarity of the surface and enhance adsorption of water, which

0 20 40 60 80

Ce (mg/l)

gives rise to the formation of water clusters that may block the access of phenol For P-AC, the polymeric chains of PVP cannot enter into the micropores and hence will have a much less significant effect on the pore volume and pore size distribution, which results in only slight reduction of the phenol adsorption capacity

Figure 4.8 Pore size distributions of AC, Q-AC and P-AC

0.000 0.005 0.010 0.015 0.020 0.025

0.030

AC Q-AC P-AC

4.4 Conclusions

Two different surface functionalization routes for attaching quaternary ammonium groups to AC to achieve antibacterial properties were demonstrated The AC was first treated with HNO3 and then activated by thionyl chloride followed by the reaction with the appropriate N-containing reagents,QAS or PVP.Assays with E coli and S aureus

showed that the functionalized ACs possessed highly effective bactericidal activity The functionalized ACs can be used in repeated applications and the antibacterial efficiency can be restored via a simple ethanol washing process The preparation of the P-AC containing poly(vinyl-N-hexylpyridinium bromide) is considered to be a better technique than that of attaching QAS groups since the former has higher antibacterial activity and the adsorption capacity of the AC is preserved to a larger extent The present work provides an alternative method to silver or metal oxide impregnation for imparting antibacterial properties to AC applied in water treatment Such antibacterial ACs may also have potential applications for eliminating air-borne bacteria in hospitals

as well as for combating bio-terrorism