Ultrasonics Sonochemistry 10 (2003) 315–318

www.elsevier.com/locate/ultsonch

The development and evaluation of ultrasound for the treatment

of bacterial suspensions. A study of frequency, power and
sonication time on cultured Bacillus species q
E. Joyce, S.S. Phull, J.P. Lorimer, T.J. Mason *

Sonochemistry Centre, School of Science and the Environment, Coventry University, Coventry CV1 5FB, UK
Received 17 March 2003; accepted 17 March 2003

Abstract

Some species of bacteria produce colonies and spores which agglomerate in spherical clusters (Bacillus subtilis) and this serves as
a protection for the organisms inside against biocidal attack. Flocs of fine particles e.g. clay can entrap bacteria which can also
protect them against the biocides. It is because of problems such as these that alternative methods of disinfecting water are under
active investigation. One such method is the use of power ultrasound, either alone or in combination with other methods. Ultra-
sound is able to inactivate bacteria and deagglomerate bacterial clusters or flocs through a number of physical, mechanical and
chemical effects arising from acoustic cavitation. The aim of this study was to investigate the effect of power ultrasound at different
powers and frequencies on Bacillus subtilis. Viable plate count techniques were used as a measure of microbial activity. Results
showed a significant increase in percent kill for Bacillus species with increasing duration of exposure and intensity of ultrasound in
the low-kilohertz range (20 and 38 kHz). Results obtained at two higher frequencies (512 and 850 kHz) indicated a significant
increase in bacteria count suggesting declumping. In assessing the bacterial kill with time under different sonication regimes three
types of behaviour were characterized:

• High power ultrasound (lower frequencies) in low volumes of bacterial suspension results in a continuous reduction in bacterial
cell numbers i.e. the kill rate predominates.
• High power ultrasound (lower frequencies) in larger volumes results in an initial rise in cell numbers suggesting declumping of the
bacteria but this initial rise then falls as the declumping finishes and the kill rate becomes more important.
• Low intensity ultrasound (higher frequencies) gives an initial rise in cell numbers as a result of declumping. The kill rate is low
and so there is no significant subsequent decrease in bacterial cell numbers.

2003 Elsevier B.V. All rights reserved.

Keywords: Water disinfection; Biocidal action; Ultrasound

1. Introduction problems alternative methods of disinfecting water are

Some species of bacteria produce colonies and spores,
which agglomerate in spherical clusters (e.g. Bacillus
subtilis). Chlorination destroys the microorganisms on
the surface but often leaves the innermost intact. Flocs
of fine particles e.g. clay can entrap bacteria which also
protects them against chlorination [1]. Due to these
q biological cells via a number of processes [2–5].
This paper was originally presented at Applications of Power
Ultrasound in Physical and Chemical Processing (USound3) Paris
December 2001.
*
Corresponding author.
being investigated, and amongst these is the application
of ultrasound, which is proving to be of considerable
interest. Ultrasound is able to inactivate bacteria and
deagglomerate bacterial clusters or flocs through a
number of physical, mechanical and chemical effects
arising from acoustic cavitation.
On collapse, cavitation bubbles produce enough
energy to mechanically weaken or disrupt bacteria or
• Forces due to surface resonance of the bacterial
cell are induced by cavitation. Pressures and pressure

1350-4177/$ - see front matter
2003 Elsevier B.V. All rights reserved.
doi:10.1016/S1350-4177(03)00101-9
316 E. Joyce et al. / Ultrasonics Sonochemistry 10 (2003) 315–318
gradients resulting from the collapse of gas bubbles
which enter the bacterial solution on or near the bac-
terial cell wall. Bacterial cell damage results from me-
chanical fatigue, over a period of time, which depends
on frequency.
• Shear forces induced by microstreaming occurs with-
in bacterial cells.
• Chemical attack due to the formation of radicals (H
and OH ) during cavitation in the aqueous medium
[6]. These radicals attack the chemical structure of
the bacterial cell wall and weaken the cell wall to
the point of disintegration.
• Amongst the final products of this sonochemical de-
gradation of water is hydrogen peroxide (H2 O2 ),
which is a strong bactericide.
Sonication alone can provide powerful disinfection.
However, to achieve 100% kill rates using only ultra-
sound it is necessary to use high ultrasonic intensities.
Unfortunately this makes the technique expensive to use
for general large-scale microbiological decontamination
[7]. Nevertheless there is a drive towards the use of ul-
trasound in decontamination as an adjunct to other
techniques (e.g. chlorination [8]) because some micro-
organisms are becoming resistant to existing disinfection
techniques involving biocides, ultraviolet light and heat
treatment. This paper addresses the disinfection of water
using sonication alone in order to determine the fun-
damental effects of changes in frequency, power and
sonication time on bacterial kill.
2. Materials and methods
A stock solution of Bacillus subtilis was prepared by
removing a loop of bacteria from a mother dish. The
loop of bacteria was used to inoculate 50 cm3 of nutrient
broth (Oxoid, 28 g/l). This was placed in a shaking in-
cubator at a temperature of 37 C for 12 h. After 12 h,
40 cm3 of the resultant inoculum solution was added to
4 l of nutrient broth (Difco 8 g/l), (10 cm3 /l, 4
1 l broth
in 2 l conical flasks). The flasks were placed in the
shaking incubator at a temperature of 37 C for another
12 h. The bacterial suspensions were now ready for use
in the experiments.
Since the bacterial cell concentration present in a
given sample can be determined by measurement of the
optical density (OD) (turbidity) of the suspension, OD
was the chosen method of bacterial calibration (k ¼ 440
nm). For reproducibility between experiments, bacterial
cell suspensions were prepared to an OD of 0.3 by either
diluting a concentrated bacterial cell suspension or the
addition of more bacteria. Calorimetry was used to de-
termine the actual ultrasonic power and intensity en-
tering a sonicated system.
The bacterial suspensions were sonicated using dif-
ferent intensities and frequencies. Samples were taken
after 1, 2, 5, 10, and 15 min and analysed by two
methods; counting of colony forming units (CFUs) via
plate counting and determination of the fluorescence of
fluoroscein diacetate absorbed into the bacteria cells
(FDA analysis) [9]. Each determination is the mean of at
least three experiments and for the graphical represen-
tation the readings were normalised to an initial value of
100% CFU.
Low frequency ultrasound: 200 cm3 bacterial suspen-
sion was placed in the reaction vessel (a 250 cm3 beaker)
and sonicated either with an ultrasonic cleaning bath
(Langford SO 3757 Sonomatic, 38 kHz) or a 20 kHz
probe system (VC-600 Sonics and Materials, USA, horn
tip diameter 13 mm). The powers entering the system
were determined using calorimetry, 0.18 and 0.24
W cm3 respectively, the temperature was maintained at
room temperature and readings were taken periodically
over a total time of 15 min. The effect of sonicating
different volumes (100 and 150 cm3 ) was measured for
the same time intervals.
High frequency ultrasound: 200 cm3 bacterial sus-
pension was placed directly into an ultrasonic bath, one
operating at 512 kHz (Undatim Ultrasonics, Belgium)
and the other at 850 kHz (Meinhart Ultrachalltechnik,
Germany, K80-5). Sonications were performed at in-
tensities determined using calorimetry of 0.071 or 0.064
W cm3 respectively for 15 min. In each case the sus-
pensions were kept at room temperature using the
water-cooled jackets on the ultrasonic devices.
3. Results and discussion
3.1. Plate count monitoring of the bio-effects of sonication
at different frequencies
The results of sonicating 200 cm3 of bacterial sus-
pension at four different frequencies are shown (Fig. 1).
200
180
160
140
120
% CFU's
5 1 2 kHz
100
8 5 0 kHz
80
V C 6 0 0 2 0 kHz
3 8 kHz b a t h
60
40
20
0
0 5 10 15 20
Time [min]
Fig. 1. Plate count monitoring of the bio-effects of sonication at dif-
ferent frequencies.
 E. Joyce et al. / Ultrasonics Sonochemistry 10 (2003) 315–318 317

The results indicate that at both 20 and 38 kHz there
appears to be no dramatic effect on the viability of the
bacteria except that there is a small but detectable drop
at 15 min sonication. In contrast the higher frequencies
produce an immediate rise in the CFU over the first 5
min followed by a steady fall, but the level remains
above the original concentration even after 15 min so-
nication. These results suggest that the major effect of
high frequency ultrasound is the declumping of bacterial
agglomerates with little deactivation. This might also
indicate that the declumping effect at the lower fre-
quencies (to produce a greater number of CFUÕs) masks
the actual deactivation.
3.2. Plate count monitoring of the bio-effects of sonication
at lower frequencies at different volumes
When the same source of ultrasound is used to so-
nicate a smaller volume of bacterial suspension there is a
resultant increase in the intensity of ultrasound entering
the system. A series of experiments were undertaken in
which both 38 and 20 kHz were used to sonicate vol-
umes of 100, 150 and 200 cm3 contained in a 250 cm3
beaker. Under temperature controlled conditions the
resulting bacterial kill for each volume are shown (Fig.
2). As might be expected, for both frequencies, sonica-
tion of smaller volumes produced a more rapid kill. The
results were more marked for the more powerful 20 kHz
probe system but in both cases using 150 and 100 cm3
samples a reduction in CFU was observed from the
outset of sonication.
20kHz Probe
120
100
80
% CFU's
3.3. Comparison of plate count and FDA monitoring of
the bio-effects of sonication at 512 and 850 kHz
High frequency ultrasonic baths deliver low power
ultrasonic waves and consequently low intensities enter
the system (0.064 and 0.071 W cm3 respectively). These
devices were not efficient for bacterial kill using ultra-
sound alone. After a 15-min treatment of 200 ml of
suspension with both the 512 and 850 kHz ultrasonic
baths the plate count assay showed an increase in the
activity of the bacteria (Fig. 1). The assay results using
FDA are consistent with those obtained using viable
plate counts but with a much more striking increase in
activity for the 850 kHz system (Fig. 3).
Plate counting and FDA analysis both show the re-
sults of sonication on the bacteria in suspension but they
do not reflect the same effects. Measurement of CFU
reflects the viability of the cells after sonication although
it is important to note that a CFU may be a single cell or
a clump of cells. Logically therefore if ultrasound de-
clumps the cells then more CFUÕs will be formed. On the
other hand FDA will reflect the absorption of a dye into
the cell and reflects a change in transport across the cell
membrane––not necessarily resulting in cell death.
Nevertheless both techniques definitely show how ul-
trasound can influence living bacteria cells.
When particle size measurements were made on the
sonication of E. coli a surprising result was obtained [8].
During the initial stages the size of the particles ap-
peared to increase. It was suggested that ultrasound
killed some of the E. coli cells as expected but then ag-
gregated the dead cells into larger clumps. Initial studies
with this current system appear to corroborate this ob-
servation but it is possible this may be the result of an
anomaly. For example individual cells are very small
and difficult to count on the system we employed i.e. a
Galai particle size analyser with a 1l lower limit. It is

100ml
60 150ml possible that since the larger (dead) clumps are more
40
200ml
easily detected this might have distorted the results.
20

0
0 2 4 6 8 10 12 14 16

(a) Time [min] 600

500
Normalised measurement

38kHz Bath
120
400
100
5 1 2 kHz - C F U
8 5 0 kHz - C F U
80 300
% CFU's

100ml 5 1 2 kHz - F D A
8 5 0 kHz - F D A
60 150ml
200ml 200
40

20 100

0
0
0 2 4 6 8 10 12 14 16 0 5 10 15 20
(b) Time [min] Time [min]

Fig. 2. Plate count monitoring of the bio-effects of sonication at lower Fig. 3. Comparison of plate count and FDA monitoring of the bio-
frequencies at different volumes. effects of sonication at 512 and 850 kHz (error less than 1%).
318 E. Joyce et al. / Ultrasonics Sonochemistry 10 (2003) 315–318
4. Conclusions
Sonication has two effects on Bacillus subtilis sus-
pensions. The first is bacterial declumping which breaks
up bacterial clumps into a greater number of individual
bacteria in a suspension, and the second bacterial killing
which results in less individual viable bacteria present in
a suspension. The overall effect of applying ultrasound is
thus the result of a competition between killing and
declumping bacteria in solution. The net effects of such a
competition can be classified into three apparently dif-
ferent shapes for curves representing bacterial survival
against time.
Type 1 Using high power ultrasound in low volumes of
bacterial suspension results in a continuous re-
duction in bacterial cell numbers i.e. the kill rate
predominates (e.g. Fig. 2 results at 20 kHz using
100 cm3 suspension).
Type 2 Using high power ultrasound in larger volumes
results in effective declumping giving an initial
rise in bacterial cell numbers but this initial rise
then falls as the declumping finishes and the kill
rate becomes more important (e.g. Fig. 2 results
at 20 kHz using 150 cm3 suspension).
Type 3 Using low intensity ultrasound an initial rise in
cell numbers occurs as a result of declumping.
The kill rate is low and so there is only a small
subsequent decrease in bacterial cell numbers
(e.g. Fig. 1 results at 512 and 850 kHz).
High frequency low power ultrasound (512 and 850
kHz) used alone is not particularly effective for disin-
fection but produces effective declumping of bacteria.
Low frequency power ultrasound (20 and 38 kHz) at
higher powers also produces declumping but the kill rate
is substantially higher.
References
[1] J. Mir, J. Morato, F. Ribas, J. Appl. Microbiol. 82 (1997) 7.
[2] C.H. Johnson, J. Physiol. 67 (1929) 354.
[3] F. Schmitt, B. Uhlemeyer, Proc. Soc. Exp. Bio. Med. 27 (1930)
626.
[4] P. Hughes, N. Nyborg, Science 138 (1962) 108.
[5] H. Allinger, Amer. Lab. 10 (1975) 75.
[6] P. Reisz, Free radical generation by ultrasound in aqueous solu-
tions of volatile and non-volatile solutes, in: T.J. Mason (Ed.),
Advances in Sonochemistry, vol. 2, JAI Press, 1991, pp. 23–64,
ISBN 1-55938-267-8.
[7] FWR––Research Report, Applications of Ultrasonics in the Water
Industry, 1995.
[8] S.S. Phull, T.J. Mason, J.P. Lorimer, A.P. Newman, B. Pollet,
Ultrasonics 4 (1997) 157.
[9] T.J. Mason, E. Joyce, S.S. Phull, J.P. Lorimer, Potential uses of
ultrasound in the biological decontamination of water, Third
Conference on Applications of Power Ultrasound in Physical and
Chemical Processing Paris, 2001.