7/30/2019 Desinfección baja corriente http://slidepdf.com/reader/full/desinfeccion-baja-corriente 1/15 Bioelectrochemistty and Bioenergetics, 9 (1982) 29-243 A section of J. Electroanal. Chem., and constituting Vol. 141 (1982) Elsevier Sequoia S.A., Lausanne-Printed in The Netherlands 229 457-THE MECHANISM OF LOW FREQUENCY ax. ELECTROCHEMICAL DISINFECTION G.E. STONER and G.L. CAHEN, Jr. Department of Materials Science, School of Engineering and Applied Science, University of Virginia, Charlottesvil le, Va (U.S.A.) M. SACHYANI and E. GILBADI Department of Chemistry, Tel-Aviv University, Ramat Aviv (Israel) (Manuscript received April 23rd 1981) SUMMARY The mechanism of low frequency a.c. electrochemical disinfection was studied. A random oriented graphite fiber-epoxy matrix composite material was employed to fabricate the electrodes. This material was found to have very good electrochemical stability when cycled anodically and cathodically in solutions containing NaCl or NaBr. The lethal species was identified as HCIO and HBrO in solutions containing Cl- and Br-, respectively, although the free halogen molecule may contribute to the observed disinfection activity under certain experimental conditions. The efficiency of disinfection increases with increasing concentration of Cl- (or Br- ) and with decreasing pH. Deaerating the solution decreases the efficiency of disinfection significantly. NaBr was found to be effective at much lower concentrations than NaCl at all pH values but the effect is enhanced at pH >7 due to ionization of HClO. The unique advantage of the method described in this paper is that a high transient concentration (in time and space) of the lethal species is generated, enough to destroy the most resistant microorganisms, while the average concentration of active halogen in the effluent liquid remains well below the objectionable level. INTRODUCTION Disinfection of a fluid can be achieved by different means, some based on chemical reactions, e.g., the use of a halogen or of ozone and others based on physical methods such as application of ultrasonic or electromagnetic radiation such as X-rays or y radiation. The choice of method to be used depends on the fluid treated as well as on cost and safety considerations. In the final analysis, disinfection is caused by the input of energy either directly, as in the power-supply of an ultrasonic transducer or U.V. lamp, or indirectly, as the energy required to produce 0302-4598/82/0000-O/$02.75 0 1982 Elsevier Sequoia S.A.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
(and in certain cases also to store and transport) the chemicals employed. Thus,
when large volumes of fluid are to be treated, as in municipal drinking water or
sewage plants, the cost of energy may become a major consideration. In smaller
systems initial cost, safety, reliability and convenience of operation are usually the
predominant considerations in the choice of the preferred disinfection method.Production of the chemical in situ, for example, by mixing sewage with sea-water
and electrolyzing to produce chlorine, has an obvious appeal in that the dangerous
chemical is not transported and stored. The low frequency U.C. electrochemical
disinfection method presented below is one such in situ method. It is unique in that
the active species is produced by a slow a.~. signal, at a frequency of 1 Hz or less,
instead of the commonly used d.c. or line frequency techniques. In a patent issued in
1973 to G.E. Stoner, it was shown [l] that application of a low frequency voltage
signal of 5- 15 V amplitude between two graphite electrodes could effectively destroy
a large variety of microorganisms contained in the fluid between the electrodes.
Approximately 40 species of microorganisms, ranging in size from viruses throughbacteria and algea to relatively large species such as Euglena were successfully
treated.
The lethal species was not identified but it was established [2] that it was formed
during the anodic half cycle on each electrode. Reversing the polarity helps to
maintain a clean electrode surface, but it also serves an additional purpose, as will be
discussed below. The exact waveform had little effect on the results as long as the
frequency was kept below ca. 10 Hz and preferably not above 1 Hz.
Graphite electrodes were employed because of their low cost and relatively high
stability during alternating voltage conditions in chloride-containing solutions. Much
better performance was later observed [3] with composite electrodes consisting ofgraphite fibers in an epoxy matrix. This type of electrode material was used in all
further experiments *.
In the present work, the lethal species formed in the anodic half-cycle was
identified and its activity under different experimental conditions was determined in
relation to situations which may occur in practical applications.
EXPERIMENTAL
The ceil and electrodes
Two types of cells were employed in this work and are shown in Fig. 1. The flow
cell consists of two compartments in a bipolar configuration. The liquid was allowed
to flow from one compartment to the other through 24 holes in the central electrode
(diameter of each hole 0.1 cm). The electrodes (surface area 54 cm*) were composed
l It was recently shown [4] that the epoxy matrix-graphite fiber composite electrodes maintained their
stability under conditions similar to those employed in the present work after more than 450 hours of
All measurements were performed employing Escherichia co/i B as the test
organism. This bacterium was chosen because it is not pathogenic, is relatively
insensitive to changes in pH (it can grow in the range of 4.5-9.0 and survive in therange of 3.0-9.5), can withstand osmotic pressure changes (from distilled water to
sea water) and is relatively stable thermally (it can grow at temperatures up to
45’C). In addition, E. coli B can grow in the presence of oxygen or anaerobically.
A fresh overnight culture was used in each set of experiments. The cultures
contained lo8 cell/cm3 and were obtained by inoculating 5 cm3 of nutrient broth
(Difco) and incubating at 37°C without shaking.
Since this electrochemical disinfection method was tested earlier [l] with many
different organisms, it is reasonable to assume that the behavior reported below is
representative for many other microorganisms. It should be remembered, though,
that quantitatively these results are valid only for the specific organism tested.
Method of counting
For viable count 10 mm3 (p-11) amples were dispended on the surface of nutrient
agar plates with an Eppendorf micropipette. Different dilutions were used so that
the number of colonies counted would be large enough to be statistically meaningful,
yet small enough to be countable. The use of small samples limits the sensitivity, but
is very convenient and enables counting of a large number of samples easily and
inexpensively. This was considered advantageous since small changes in the viable
count, i.e. the number of living microorganisms per unit volume, are of little
practical significance. The number of colonies counted in a typical experiment was
lo-20 and each dilution was counted twice. This leads to a statistical error of ca.
20-30%, which amounts to half a unit or less on the logarithmic scale shown in Figs.
3-7.
The results are presented in the form of log( No/N), where N,, is the concentration
of bacteria in each experiment before treatment (usually (l-5) X lo5 cells/cm3) and
N the concentration after treatment in the electrochemical cell. Since one colony
counted in 10 mm3 solution represents lo2 cells/cm3, the limit of detection was
N,/N = lo3 or 99.9% destruction of the cells. Thus, the limiting value of log
N,/N = 3 shown in several of the figures below represents the detection limit of this
technique and not necessarily a saturation effect in the disinfection process. While
this limit of detection could be readily increased (e.g. by taking 100 mm3 samples), it
was not considered necessary; since all the significant features required to under-
stand the behavior of this system were already apparent with the method employed.
Determination of active chlorine or bromine
The concentration of active chlorine or bromine was determined by its color
reaction with o-tolidine. Comparison with commercial color standards yields a quick
potassium sulfate solution: 0.15% K,SO, (8.6 X 1O-3 M);
potassium dihydrogen phosphate solution: 0.68% KH,PO, (0.050 M).
Buffer solutions were prepared by standard methods [6,7]. The solutions used,each in its usual pH range, were acetate buffer, phosphate buffer, tris buffer and
carbonate buffer.
The buffer capacity B, defined as the number of moles of strong base which has
to be added to 1.0 dm3 of the buffer solution to change its pH by one unit, was
calculated from the approximate formula [8]
p = 2.3K,cH+c
(K,+c,+jz +‘H+++OH-I
(1)
where K, is the acid dissociation constant of the weak acid and c the total molar
concentration of the acid and the salt.
Bacteriological preparations
The test organism was E. coli B obtained from Professor E.Z. Ron of the
Department of Microbiology, Faculty of Life Sciences, T.A.U. The bacteria were
grown in a nutrient broth (Difco). The colonies for viable count were grown on
nutrient agar plates. All solutions and test tubes were sterilized (15 min, 1.5 atm) in
an automatic vertical autoclave (Tuttnauer, model 13000 L). Pipettes and larger glass
containers were sterilized in an oven at 18O’C for 2 h. The bacteria on the agarplates were incubated overnight at 37OC in a water jacketed incubator (Tuttnauer,
model 12000).
RESULTS
Flow cell
The relative disinfection (N,,/N) is plotted in Fig. 3 on a semi-logarithmic scale a:
a function of the peak current density, for two concentrations of NaCl. Changing the
salt concentration from 0.05% to 0.85% had little effect. Complete disinfection (i.e.log N,/N 2 3) was observed when the current density reached about j = 7.5 mA
cm-*. The peak voltage across the cell at this current density was 7.OV in artificial
tap water and 5.5 V in normal’ saline. The difference in cell voltage is due to the
change in solution resistance and’has no effect on the electrode reaction proper. The
pH in both solutions was close to 5.
All experiments were performed at 25’C. Experiments at 10°C and 40°C showed
no significant change in the efficiency of disinfection with temperature.
The pH of the solution has a very significant effect on the disinfection efficiency
as seen in Fig. 4. Tap water was used in this experiment and the relative disinfection
Considering Fig. 2b, it is evident that a great part of the current measured is
consumed to charge the double layer capacitors, and the potential across the cell
does not reach steady-state during the voltage pulse. Under these conditions, thecritical value of the peak current density (ca. 3.5 mA cm-*) required for the
disinfection process to commence (cf. Fig. 3) is related to the buildup of the potential
across the double layer to a value required for the faradaic reactions to occur. The
concentration of NaCl could effect the amount of chlorine formed during the anodic
half-cycle in the following ways:
(i) increasing the concentration lowers the reversible potential for chlorine evolu-
tion and increases the exchange current density, both leading to better disinfection;
(ii) more of the faradaic current is consumed for the formation of Cl, as the
concentration of NaCl increases and less for the two possible side reactions. While
this may at first seem advantageous it has a drawback in that the local decreases inpH near the anode due to these side reactions [equations (3) and (4)] will occur to a
lesser extent. This may decrease the efficiency for disinfection, since molecular
chlorine is more active than HClO and ClO- has very little activity compared to
either. Thus, lowering the pH favors the formation of the more active species, as seen
in Fig. 8a.
(iii) A change in capacitance due to the change in concentration of NaCl may
affect the way in which the potential across the double layer is built up during the
anodic half-cycle. This is expected to be a rather small effect, since saturation
adsorption of chloride ions probably occurs already in artificial tap water, and
increasing the concentration of NaCl further would hence have little effect on thedouble-layer capacitance.
The net effect of increasing the concentration of NaCl from 0.05% to 0.85% is
seen to increase the disinfection efficiency to a small extent.
The effect ofpH
The effect of pH on the efficiency of disinfection in tap water is shown in Fig. 4.
Comparison with Fig. 8a shows that the ratio &,/IV decreases as more and more of
the hypochlorous acid is ionized, showing that the anion ClO- is much less effective
in the destruction of the test organism.
A more detailed test of the pH effect is shown in Fig. 6a, where the efficiency of
disinfection is plotted as a function of the concentration of NaCl in well buffered
solutions at three pH values. The value of pK, for HClO is 7.4. Thus one would
expect that the concentration of NaCl required to reach a certain degree of
disinfection would increase somewhat as the pH is increased from 5.2 to 7.0. A
further increase of pH to 8.6 leaves only a few per cent undissociated acid.
Accordingly, no disinfection was observed up to a NaCl concentration of 1 X lo4
ppm.An obvious test of this interpretation of the experimental results is to replace
NaCl by NaBr. Since the acid dissociation constant of HBrO is pK, = 9 one would
not expect a drastic decrease in disinfection efficiency in NaBr solutions up to a pH
of 9. This is seen to be the case in Fig. 6b. Although the concentration of NaBr
required to reach a given value of (No/N) increases steadily with pH, there is no
drastic change between pH = 7.0 and 9.2 as in the case of NaCl. One might haveexpected that in a pH = 10 buffer the efficiency of disinfection would decrease
drastically. However, at this high pH the test organism can no longer survive. The
viable count was found [ 1 l] to decrease by one order of magnitude in 3 h and by at
least three orders of magnitude in 24 h without application of a current. The cells
which are not immediately destroyed are probably rendered more sensitive at this
high pH. Thus the disinfection efficiency is not very much decreased from pH 9.2 to
10.0, even though the concentration of undissociated HBrO is decreased about
fivefold.
The steady decrease in efficiency of disinfection with increasing pH seen in Fig.
6b may be caused by two factors:(i) it is generally accepted that the free halogen is more active than the hypoha-
lous acid formed by its hydrolysis. Considering Fig. 8b, it is seen that the concentra-
tion of molecular bromine is not negligible in the pH range shown in Fig. 6b. Thus,
some loss of activity may be expected as the pH is increased;
( i i ) the reversible potential for oxygen evolution decreases with increasing pH,
thus the relative rate of oxygen evolution would be expected to increase, leaving less
current for the production of bromine.
A further test of the model was performed by studying the effect of the buffer
capacity at a fixed pH of 7.0 on the efficiency of disinfection in NaCl solution. It
was shown above that this is not a very favorable pH value, since almost half of the
active material (HCIO) is lost by dissociation. On the other hand, the two plausible
side reactions [equations (3) and (4)] both lead to the formation of H30f ions and a
local decrease in pH, which coincides in t ime and space with the formation of Cl,
and HClO. Thus, one would expect to have a relatively high disinfection efficiency
in unbuffered solutions at pH = 7, decreasing as the buffer capacity increases. This
is seen to be the case in Fig. 7. The concentration of NaCl needed to reach 99.9%
disinfection (log N,/N = 3) increases steadily from ca. 250 ppm in an unbuffered
solution to about 2500 ppm in a well buffered solution at the same pH.
The eff ect of aerat i on
The effect of deaerating the solution on the efficiency of disinfection is shown in
Fig. 5a and 5b for NaCl and NaBr, respectively. A very substantial decrease in
efficiency of disinfection is caused by deaerating the solution in both cases. In
addition, the cell voltage is found to increase and the current density at which
formation of the active halogen is observed (positive o-tolidine) is roughly doubled.
On the other hand, experiments in which NaClO was added to a solution containing
the test organism revealed no effect of oxygen content on the rate of deactivation of
the cells. Thus; the possibility that deaeration influenced the stability of the
The effect of deaeration on the disinfection efficiency in the electrochemical cell
can be understood in terms of the side reactions which take place during both the
anodic and the cathodic half-cycle.
Consider the electrochemical processes taking place in a deaerated solution,
starting with the cathodic cycle. Since there is no halogen or oxygen in the solution,the only reaction which can take place is hydrogen evolution. When the polarity is
reversed the first anodic process will be oxidation of molecular hydrogen [equation
(4)] followed by halogen formation [equation (2)] and possibly some oxygen evolu-
tion [equation (3)]. The reversible potential for chlorine and bromine evolution in
neutral solutions is about 1.8 V and 1.5 V respectively, more anodic than that for the
oxidation of molecular hydrogen. Thus, the latter may be expected to be essentially
completed before the former starts. In the next cathodic half-cycle, the first reaction
will be reduction of the halogen, followed by a shift of potential by ca. 1.8-2.0 V to
hydrogen evolution *.
Consider now the effect of aerating the solution. During the cathodic half-cycle,
oxygen reduction will take place first l, and the amount of hydrogen formed will be
significantly reduced. In the following anodic half-cycle, most of the faradaic current
will be consumed for the production of the halogen and oxygen, since there is little
or no molecular hydrogen near the electrode. Thus the formation of active halogen
(positive o-tolidine) and the consequent disinfection may be expected to occur in
aerated solutions at lower current densities than in deaerated solution, as shown in
Fig. 5a and 5b. It may be noted that the system is self regulating with respect to its
oxygen content. Thus, if one starts with a deaerated solution, the hydrogen formed
during the cathodic half-cycle will retard the formation of oxygen in the following
anodic half-cycle, leading again to formation of molecular hydrogen in the next
cathodic half-cycle and so on. If, .on the other hand, the solution contains oxygen to
start with, little or no hydrogen will be formed during the cathodic half-cycle, and
this will lead to the formation of more oxygen at the electrode surface during the
anodic half-cycle, preventing the formation of hydrogen in the next cathodic
half-cycle even more efficiently.
Aerating the solution has an effect of reducing the potential during the cathodic
half-cycle since oxygen reduction can occur instead of hydrogen evolution. During
the anodic half-cycle, it has no effect on the potential since chlorine evolution occurs
at the same potential irrespective of the presence of molecular hydrogen.
* The magnitude of the shift depends on the kinetics of the two reactions, which are not known on this
type of electrodes. However, on graphite the reduction of chlorine has a higher exchange current density
than hydrogen evolution, hence the shift in potential may be expected to be larger than the difference in
reversible potentials.
l * This may lead to the formation of hydrogen peroxide which could enhance the disinfection efficiency,
but we shall ignore this added effect for lack of direct experimental evidence.
Literature data on the relative activity of chlorine and bromine are rather
scattered. It seems to be generally accepted that the free halogen is more active than
the hypohalous acid. Studies with different organisms seem to indicate that Cl, and
HClO are more active than Br, and HBrO [ 12,131, although this is not universally
accepted [14]. When the pH of the solution is increased above 7 or 8, the activity of
bromine is found to be much higher than that of chlorine, due to ionization of
HClO, while HBrO is still undissociated.
In the present work, the disinfection efficiency was found to be much higher in
solutions of NaBr than solutions of NaCl, under otherwise identical conditions.
Comparison of Fig. 6a and 6b shows that in relatively acid solutions (pH = 5) the
concentration of NaCl required to reach 99.9% disinfection was about 17 mM while
the concentration of NaBr required to achieve the same effect was only about 0.5mM. Even at pH = 10, a concentration of 2.5 mM NaBr was sufficient to reach the
same extent of disinfection.
The enhanced efficiency found in solutions of NaBr is due to the lower reversible
potential of the BrJBr- couple. Thus, bromine evolution can start earlier in the
anodic half-cycle than chlorine evolution and the amount of bromine (and HBrO)
formed is larger than the amount of chlorine (and HClO) formed at the same current
density. Moreover, the part of the current consumed for side reactions is also smaller
and the beneficiary effect of the side reaction of reducing the local pH is of no
importance in NaBr solutions, due to the higher value of pK, for HBrO.
CONCLUSIONS
The lethal species in the electrochemical disinfection method was identified as
HClO or HBrO in NaCl and NaBr solutions. In acid solutions (pH 4-5) some of the
activity may also be due to the presence of free halogen. The acid dissociation
constant of HClO is pK , = 7.4 and since the anion is much less active as a
bactericide, the disinfection efficiency decreases rapidly with increasing pH beyond
the pK, value, particularly in buffered solutions. In unbuffered solution the genera-
tion of H,O+ ions associated with oxygen evolution or hydrogen oxidation, which
may occur as side reactions, can cause a decrease in local pH near the anode and
thus enhance the disinfection efficiency in solutions of NaCl.
The efficiency of disinfection is much higher in solutions of NaBr than in
solutions of NaCl, because of the lower anodic potential required for bromine
evolution. The difference is even larger at pH > 7 since HClO is ionized and the
anion ClO- is much less active, while HBrO remains mostly undissociated up to
pH = 9. This higher efficiency of disinfection may not be of practical importance in
many instances, since most natural fluids (e.g. drinking water, sewage etc.) contain
already enough NaCl for this electrochemical disinfection technique to operate
efficiently. However, where the addition of a salt is undesirable and the pH should
not be changed, the higher activity of NaBr will be invaluable.
One of the most important features of this electrochemical disinfection method is
that the local concentration of the active halogen during the anodic half-cycle can
exceed the average concentration found in the fluid emerging from the cell by one or
two orders of magnitude. A detailed calculation for the specific pulse shape
employed in this study may be rather involved, but an order-of-magnitude typeestimate can serve to elucidate this point. Assume that a current density of 6
mA/cm* is applied (cf. Fig. 3) for 0.3 s and the efficiency of chlorine evolution is
20% l The total amount of Cl, formed in a single pulse will then be 3.6 X lop9
es/cm*. If this amount will be in the diffusion layer of thickness 6 = (rDt)‘/* = 3 X
lo3 cm, the average concentration in this region will be 1.2 X 10e3 eq/dm3, or 42
ppm. If we assume that half of this will be reduced in the cathodic half-cycle and the
rest will be diluted in the cell, which has a separation of 0.1 cm between the
electrodes, the concentration of chlorine in the treated fluid will be about 016 ppm.
Thus, the local concentration could be high enough to destroy the hardiest micro-
organisms, yet the concentration of active halogen in the treated liquid would bebelow the objectionable level.
ACKNOWLEDGEMENTS
The authors wish to thank Professor E.Z. Ron of the Department of Microbiol-
ogy, Faculty of Life Sciences, Tel-Aviv University, for her most valuable assistance
in the microbiological aspects of this study.
Partial support of this work by the U.S.-Israel Binational Foundation is grate-
fully acknowledged.
REFERENCES
1
2
3
4
5
6
7
8
9
10
11
12
13
14
G.E. Stoner, U.S. Patent No. 3,725,266 (1973).
G.L. Cahen Jr., Ph.D. Dissertation, Univ. of Virginia (1976).
G.E. Stoner, G.L. Caben Jr. and G. Stafford, Report No. U.VA/525lOl/MS79/101 (1979).
G. Stafford, Ph.D. Dissertation, Univ. of Virginia (1980).
Standard Methods for the Examination of Water and Wastewater, 12th edition, Amer. Public Health
Assoc., New York, 1965, p. 93.
R.M.C. Dawson, D.C. Elliot, W.H. Elliot and K.M. Jones (Editors), Data for Biochemical Research,
2nd edition, Clarendon Press, Oxford, 1969.
S.P. Colowick (Editor), Methods in Enzymology, Vol. 1, Academic Press, New York, 1955, pp.
138-146.
R.G. Bates, Determination of pH. Theory and Practice, John Wiley and Sons, New York, 1964, p. 99.
G.C. White, Handbook of Chlorination, Van Nostrand-Reinhold Co., Princeton, 1972.
G.M. Fair, J.C. Morris, S.L. Chang, J. Am. Water Works Assoc., 40 (1948) 1051.
M. Sachyani, M.Sc. Dissertation, Tel-Aviv University (1979).
H.C. Marks and F.B. Standskov, Ann. N.Y. Acad. Sci., (1950) 163.
T.A. Koski, L.S. Stuart and L.F. Ortenzio, Appl. Microbial., 14 (1966) 276.
J. Kamlet, U.S. Patent No. 2,662,855 (1950).
l This low efficiency is assumed because the current densities shown in Fig. 3 are peak current densities
(the average is much smaller) and because the faradaic current does not start as soon as the anodic voltageis applied.