B
I O
P
R O C E S S
TECHNICAL
Chlorine Dioxide, Part 1
A Versatile, High-Value Sterilant
for the Biopharmaceutical Industry
Barry Wintner, Anthony Contino, Gary O’Neill
H
istorically, chlorine dioxide
(CD) became important
in sanitation because of
municipal water treatment
concerns about halomethanes and
chloramines generated during industrial
chlorine-based water treatment. The
American Water Works Association
(1, 2) details the chemical properties
of CD, with gas generator designs and
the history and applications of CD in
water treatment. ERCO Worldwide
(www.ercoworldwide.com) provides
extensive background material
including recent literature, patents,
and microbiology on a dedicated
website: www.clo2.com.
To date, limitations in CD gas
generation technology have kept
this attractive product from many
applications for which its properties
would be advantageous. Several novel
technologies may bring it into the
mainstream of biopharmaceutical
manufacturing and maintenance
operations.
In its aqueous phase, the same basic
CD supply system can be used as a
starting point for the entire range of
biopharmaceutical applications:
sanitization, sterilization, and routine
or emergency disinfection. CD is as
useful as a sanitizer for utility water
systems and surface decontamination
as for process applications. Few
technologies are as easy and
convenient to use while providing
value for such a wide range of
applications. CD has been studied
in-depth for many years. For example,
Young and Setlow (3) compare CD
and bleach, focusing on sporicidal
aspects. Mittelman’s series (4–6)
discusses growth and destruction of
biofilms in purified water systems. As
the industry becomes more familiar
with CD, it could become the choice
for most if not all operational
sanitization, disinfection, and
sterilization applications in
biopharmaceutical manufacturing
facilities.
PRODUCT FOCUS: ����H�����E�������
properties of oxidizing biocides to
consider in choosing a sanitizing/
sterilizing agent. As shown, CD is not
as aggressive an oxidizer (oxidation
potential data) as chlorine, ozone,
peracetic acid, peroxide, or bleach —
and it should be noncorrosive to
common materials of construction.
A high oxidation capacity (seeking
five electrons rather than two),
however, suggests that CD is a most
efficient reagent when oxidation
proceeds to completion.
PROCESS FOCUS: ������������� ���
�E�E���H
WHO SHOULD READ: ����E��
�E�E����E��, �������������, ���������Y
�����E��, ��� Q�����Y �������
KEYWORDS: ��E����� ����������,
������E����������������, ����������,
�������E�� ����E�����, ���������E�
LEVEL: ���E��E����E
4�
BioProcess International
D ECEMBER 2005
Comparing CD with Other
Sterilants: Table 1 summarizes key
CAMBREX BIOPHARMACEUTICAL MANUFACTURING SERVICES
(WWW.CAMBREX.COM/CONTENT/BIOPHARM/BIOPHARM.ASP)
Choosing a sanitizing agent
depends on the philosophy of an
organization as well as particular
process requirements. Clean steam is
the best known sterilant for process
systems. However, it is expensive
because of the necessary specialized
generation equipment and the high
cost of water-for-injection (WFI).
An important, sometimes overlooked
feature of CD is that it exists as a
neutrally charged gas in aqueous
solution, which allows it to penetrate
pores, cracks, and crevices to reach
microbial contaminants. Also, most
plastics and polymers will not
absorb it.
Table 2 compares CD with other
well-known sanitization agents and
sterilants used in gaseous form for
space-fumigation applications. Among
these, only CD is demonstrated to
sterilize as both a liquid and a vapor.
Only the vapor-phase attributes are
compared. In the table, “+” symbols
indicate that an agent is generally
favorable for a given criterion;
“–” symbols mean it is unfavorable.
The unfavorable rating of CD for
the cost criterion assumes that an
equipment-based generator produces
CD gas. Using membrane-sachet
technology with a sparging technique
to generate the gas involves a relatively
small capital investment and lower
operating costs. Thus, CD generated
that way would receive a “+” entry for
cost.
Paraformaldehyde will not be
widely used in the future because of
concerns about its toxicity, residues,
and unpredictability. The National
Research Council (7) has reported on
formaldehyde’s need for neutralization
with ammonium carbonate, as well as
the need for careful venting with this
Group B1 carcinogen. Over time,
vapor-phase peroxide (VHP) has
found a niche in the bioprocessing
industry. But VHP is of limited use
because of careful preconditioning
required, long aeration times for
removal, and its aggressiveness toward
rubbers and some polymers. The
aeration time requirements have been
a nagging issue with VHP — in some
cases requiring four to eight hours to
reduce it to a safe level in real-world
systems.
Actual aeration times for CD in
isolators and similar closed systems
are very close to the theoretical airexchange period expected (8, 9). Both
gas and aqueous-phase treatments
benefit from CD’s remarkable ability
to penetrate into dead areas and
porous materials. It can thus penetrate
and disrupt the plaque buildups
associated with many microorganisms.
For effective vapor-phase cycles, CD
introduction must be accompanied by
humidification of the air to about 70%
relative humidity (RH).
PROVEN APPLICATIONS
Decontamination of Isolators: Eylath
et al. (8) documented use of gaseous
CD to sterilize a large (240 ft3), hardwall isolator made of grade 316
stainless steel (SS), Lexan brand
polycarbonate resin (GE Plastics), and
other polymers. The unit contained
two half-suits, which are known to
present a sterilization challenge. The
isolator was humidified and sterilized
for 15–60 min with CD for a total
exposure time of less than two hours,
and excellent results were indicated by
biological indicator (BI) analysis (8).
Czarneski and Lorheim (9)
reported on gaseous CD
decontamination testing of isolators in
several different configurations. They
also compared the effectiveness and
repeatability of their results with those
obtained in other testing using VHP.
The authors concluded that because
CD is a true gas, it produced superior
performance over vaporous agents that
can condense during the
decontamination process. CD gas can
be evacuated more quickly as well, and
it produces more repeatable,
reproducible results.
Tests were conducted in a transfer
isolator fully packed with media or
components and in a train including
two isolator systems and an autoclave.
For three configurations (isolator with
media load, isolator with component
load, and isolator train) total cycle
times of 83 min (both loaded
scenarios) and 115 min (isolator train)
gave conclusive decontamination
results. Cycle times were better than
for VHP, for which three- to fivehour cycle times were observed. Total
cycle times included 30 min for
conditioning, 30–35 min for exposure
to CD, and 15 min for aeration down
to OSHA-acceptable levels. Only 12–
15 air changes were required to meet
regulatory standards.
Sterilization of Process Vessels:
Eylath et al. (10) then used CD gas to
sterilize two conventional
biopharmaceutical 316 SS vessels with
normal connections and agitators.
Those process vessels were relatively
small (100 L and 500 L), but the
reported technique could easily be
used for larger vessels such as those
typical in media and buffer
preparation. The authors claim
sterilization with CD treatment cycles
of 10–30 min, similar to the isolator
study.
In evaluating those results, capital
and operating costs should also be
considered. Increased capital cost for
clean steam (the current industry
standard) comes from required vessel
pressure ratings, so it is modest for
small vessels but substantial for large
����� �: Summarizing key properties of
oxidizing biocides to consider in choosing an
agent to sanitize or sterilize a system —
compiled data from several sources (SELECTIVE
MICRO TECHNOLOGIES, WWW.SELECTIVEMICRO.COM)
���������
���������
(v����)
���������
�������y
(���������)
O3 (ozone)
2.07
2e–
CH3COOOH
(peracetic acid)
1.81
2e–
H2O2 (peroxide)
1.78
2e–
1.49
2e–
0.95
5e–
�������� �����
NaOCl (sodium
hypochlorite
bleach)
ClO2 (chlorine
dioxide)
ones. Savings can be substantial when
using CD for sterilization in typically
large media and buffer tanks.
Operating costs for steam primarily
came from generating clean steam and
the WFI used as feedstock. The
operating cost of using CD for the
same purpose can be as little as one
fifth of those for clean steam (11).
Additionally, Bioprocess Associates
has shown that sterile water and clean
steam prepared using CD are
substantially less costly than those
prepared by conventional means (12).
In field testing performed by
Selective Micro Technologies, CD
solution was generated in a partially
filled water storage tank. After 60
min total CD generation and soak,
swab samples showed zero cfu
remaining
at three locations tested, one of which
was the top surface of the tank (in the
vapor space above the level of the
liquid contents). Before treatment,
levels of 1.01 × 103 to 7.26 × 103 cfu
were recorded. So the liquid does not
need to directly contact all surfaces to
be effective.
Ultrafiltration (UF) Membrane
Sanitization: Selective Micro
Technologies and NCSRT (www.
ncsrt.com) (13) have applied aqueous
CD to the sterilization of a 5-m 2
polysulfone UF membrane system
in testing at Wageningen University
Research in The Netherlands. Their
membrane module was used to process
Pichia pastoris fermentation broth.
Dilute CD was circulated through
the system while both retentate and
filtrate streams were recycled for about
ADVANTAGES OF CD
CD benefits for the biopharmaceutical
industry include
• Broad range of biocidal and
sporicidal properties
• Rapid acting, effective at ambient
temperature and atmospheric pressure
• Nontoxic, nonhazardous,
environmentally friendly, and non–
skin-sensitizing at normal use
concentrations in water
• Effective as aqueous solution or gas
• Easily and quickly inactivated
(purging/aeration, ultraviolet light, or
chemical inactivation) and removed
from process areas and equipment
• Residue free, easily detectable and
measurable
• Noncorrosive to construction
materials commonly used in the
biopharmaceutical industry
• Less costly (based on efficacy) than
other broad-spectrum, highperformance sterilants (e.g.,
vaporous hydrogen peroxide)
• Versatile: can be used in many
applications, minimizing the number
of agents that must be stored.
an hour at room temperature. Two
separate tests were conducted, with
targeted final CD concentrations at
100 ppm and 50 ppm. Concentrations
were monitored using CD test strips
and spectrophotometry.
Microbial inactivation in the
crossflow module was achieved after
one hour of exposure at either CD
concentration. Samples were cultured
using standard plating techniques,
with all colonies identified. Following
treatment, no growth was detected in
samples taken at all UF module
openings. No changes in membrane
performance or expected membrane
life were detected through integrity
testing (forward-air diffusion rates
at 5 psig). When compared with a
sanitization regimen originally used
in Wageningen for the same system,
significant improvements in total cycle
times (from 24 hours to two hours)
and completeness of sanitization
were observed.
Water System Sanitization: Wise
(14) used CD for sanitization of
reverse-osmosis (RO) membranes,
which are widely used in WFI water
preparation. The most common
material of construction is cellulose
acetate (CA), although sophisticated
multilayer membranes may displace
that in the future. For CA
membranes, chlorine cannot be
used as a sanitizing agent; in many
industrial systems, microflora can
grow to unacceptable levels. RO units
must be taken off-line for extended
cleaning. In using CD to sanitize the
system, Wise was careful to show that
at low levels it does not damage the
membranes to cause unwanted salt
breakthrough. Even at a 1 mg/L CD
level with a two-hour treatment cycle
(93 ppm-minutes), he saw reductions
����� �: Comparing attributes of three biocidal agents — formaldehyde (CH2O), hydrogen peroxide
(H2O2), and chlorine dioxide (ClO2) (HENRY S. LOFTMAN, PHD, MICRO-CLEAN, INC., WWW.MICROCLN.COM )
�����
�H�� ���
H��� ���
���� ���
+
+
–
0.75
–
+
?
+
1.0
–
+
+
+
0.1
+
60–90%
65–90%
–
+
30% (Steris) or
ambient (Bioquell)
+
+ (dry), ? (condensed)
+
+ (– with Cl2)
Method of removal
Neutralizer
Catalytic breakdown
Scrubbing
Development effort
Low cost
+
+
–
–
+
–
Sporocidal effectiveness
Effective through HEPA filters
Noncarcinogenic
Toxicity (TWA PEL, ppm)
Nonexplosive (at normal
use concentrations)
Relative humidity requirement
No residue
Noncorrosive
44 BioProcess International
D ECEMBER 2005
of 77% (permeate) and 96%
(concentrate) of the mixed flora
typical in such systems.
Selective Micro Technologies used
CD (generated using the company’s
microreactor) to sanitize a complete
USP water loop, including the RO
membrane unit (15). The water system
and distribution loop (Figure 1)
included two 125-gallon storage
tanks plumbed in parallel. Those
tanks store RO or DI water that feeds
a distribution loop. CD was generated
directly in the storage tanks. DI units
were bypassed and UV light turned
off for that portion of the testing.
The loop was charged with 40ppm CD, which circulated overnight
(~16 hours). Storage tanks were then
drained and refilled to 40% with
RO-quality water, which went
through the distribution system with
the return line directed to a drain.
Finally, all valves were flushed with
RO water until their measured CD
concentration was <1 ppm. Total
time required to flush the system
of residual CD was only a matter
of minutes.
At the same time, RO membranes
were also sanitized with a CD solution
of about 50 ppm. This CD was
generated using a single microreactor
sachet in a covered container and
injected into the RO feed with a
dilution pump. Because CD does
not ionize, it can pass through
RO membranes, which allows
simultaneous decontamination of
both the feed and permeate sides of
RO membranes. The RO unit was a
thin-film composite type supplied by
TriSep Corporation (www.trisep.com).
CD was visually detected in the RO
reject water within a minute. After 10
min, CD concentrations on both sides
of the membrane were essentially
equal. The system was then isolated
and allowed to soak with treatment
conditions held for about an hour
before CD was flushed from the
system. After ~10 minutes of flushing,
CD concentrations in both the
product and reject lines were measured
at less than 1 ppm.
The entire USP system was then
returned to service. Before the test it
had been heavily contaminated, with
������ �: Ultrafiltration membrane sanitization
����
�����
����������
������
������
����
��������
���
������
������
��
��
�����
�� ����
����
��
��
���������
�� �����
����
�� �������
����
��
�������
�����
������������
����
������ ���������������
����� ���������������
most samples showing microbiological
counts too numerous to count
(TNTC) and positive counts even in
water sampled directly downstream
of an in-line UV light. No microbial
contamination was detected after 24
hours of normal operation following
the CD treatment cycle.
Hard Surface Disinfection:
����
��
��
Laboratories, especially those
involved in animal testing, need to
be disinfected both routinely and
in periodic emergencies to prevent
potential infections by adventitious
organisms. Apel discusses such
applications for the produce industry
(16). Hard surfaces can be treated
with a CD liquid or foam, but the
foam is more easily applied to ceilings.
Many other successful applications of
CD within the food industry have
been published.
Cleanroom Decontamination: The
use of CD to disinfect entire rooms
and suites has been convincingly
demonstrated by several authors.
Luftman used CD to disinfect a
very large facility (170,000 ft3) at the
Widener ICU Animal Hospital (17).
The treatment cycle used <0.5 mg/L,
(400 ppm) for about an hour, with
additional time for humidification
and venting. All details (e.g., sealing
the room, HVAC circulation, and
training) proved straightforward.
(Anecdotal evidence indicates that
CD does not harm furniture, most
plastics, or computers and electronics
under the usual treatment conditions.)
After the CD cycle, the room was
simply exhausted to the outside air.
����
��
No EPA permits were needed
because CD is not considered an
environmental pollutant.
The results were a 5–6 log kill
of test spores and target bacteria
(Geobacillus stearothermophilus). Those
results would not have been very
different with Bacillus subtilis niger
or its variant Bacillus globiggi. The
extremophile G. stearothermophilus is
a model organism used to test worstcase scenarios for steam sterilization.
B. subtilis is a common spore-former
found in soil. CD’s activity against
spore-formers is an unusual and
valuable property.
CD is economical and effective
in cases of accidental microbial
contamination. Contaminated piping
(especially vents and drains), vessels,
and HVAC systems can benefit from
CD exposure.
APPLICATIONS WITH HIGH POTENTIAL
Below are applications in the
biopharmaceutical industry for which
CD could improve on traditional
methods. Testing is already in
progress for some of these.
Production of Sterilized Water
from USP Grade Water: In the absence
of published data, the term WFI is
purposely avoided here; sterilized
water is used instead, referring to
water free of biological activity and
having endotoxin levels below typical
detection limits. Preliminary tests
indicates that CD at very low
concentrations (<1 ppm) can effectively
inactivate endotoxin in a few minutes.
Depending on microbiological
conditions of feed water, CD oxidation
reactions will produce some level of
salts (mostly chlorides). The quantity
of salts produced may lead to
resistivity values that fall outside the
range of acceptability for classification
as USP or WFI quality. It can be
stated with some level of certainty,
however, that the product water will
be free of microorganisms, which in
and of itself could add significant
value in certain applications currently
using more costly WFI (e.g.,
noncritical and intermediate wash
downs). For feed water with lower
levels of microorganisms present, CD
treatment should lead to WFI quality
levels. In other situations, there may
be other ways to treat water in the
deionization-sterilization sequence
for more favorable economics than
traditional approaches. More work
must be done.
Improved Sanitization of
Chromatography Columns, Resins, and
Membranes: Testing is currently under
way to define protocols and determine
the effectiveness and suitability of CD
for capacity recovery and sanitization
of packed-bed chromatography
columns. Even at 100 ppm CD
solutions appear to have no
detrimental effect on even the most
sensitive of common stationary phases.
The effectiveness of CD for sanitizing
membranes is established. If column
testing is successful, it should be
relatively straightforward to demonstrate
CD’s effectiveness in membrane
chromatography technologies, which
may play a significant role in the
future of bioprocessing.
Biowaste Kill: Warriner (18)
compared CD with ozone and
chlorine as a liquid-phase treatment
for wastewater. Quantitative testing
involved seeded polio virus and typical
coliform bacteria. Of the three agents
tested, CD was most effective at
typical concentrations. Because the
challenges in typical biowaste kill
systems for biopharmaceutical facilities
are similar to those in municipal
systems (except for scale), CD
potentially provides an economically
attractive alternative that is effective
at ambient temperatures and displaces
more dangerous, toxic, and/or
flammable chemicals. Laboratory
testing on specific waste samples from
PROVEN EFFECTIVENESS
Here are some organisms for which
chlorine dioxide’s effectiveness has
been proven. Testing for bacteria,
viruses, and algae/fungi was
performed at an EPA-certified
laboratory. DATA FROM SELECTIVE MICRO
TECHNOLOGIES (WWW.SELECTIVEMICRO.COM)
Bacteria: Staphylococcus aureus,
Methicillin-resistant Staphylococcus
aureus (MRSA), Pseudomonas
aeruginosa, Salmonella choleraesuis,
multiple drug resistant Salmonella
typhimurium (MDRS), tuberculosis,
Escherichia coli 0157:H7 and ATCC
11229, Vancomycin-resistant
Enterococcus faecalis (VRE), Klebsiella
pneumoniae, and Bacillus subtilis (a
spore-forming bacterium)
Viruses: Coronavirus, human
immunodeficiency virus, hepatitis A,
rotovirus, feline calici virus, and
poliovirus
Algae/Fungi: Phormidium boneri,
T-mentag (athlete’s foot fungus),
Penicillium digitatum, Botrytis
species, and Fusarium solani
Yeasts: Saccharomyces cerevisiae and
Pichia pastoris
a particular process or facility is a
good starting point for those
contemplating CD use in this
application.
Sterilization of Disposable
Processing Systems: Disposable
systems offer many advantages, as
the biopharmaceutical industry is
slowly but surely recognizing. These
technologies are expected to be widely
adopted over the next five to 10 years.
Generally, components of such
systems are gamma-irradiated. Once
components are linked together to
form systems, the sterile condition of
that system (if required) is in jeopardy.
Extraordinary measures must often
be taken for tubing connections.
Some processors irradiate their entire
systems. Preliminary testing indicates
that CD treatment could be the
quickest, most economic, and most
effective method available for
presterilization of disposable systems.
Once processing is complete,
disposable materials must be
46 BioProcess International
D ECEMBER 2005
eliminated. Depending on the nature
of that processing, it may be necessary
to decontaminate those materials or
treat them as medical-grade waste,
requiring destruction in a specialized,
costly way (e.g., incineration at a
certified facility). CD could provide
a low-cost decontamination approach
that saves time and eliminates special
handling and destruction challenges.
Testing continues, but preliminary
results indicate a total kill is possible
in under five minutes, with an
additional 10–20 min required for
system evacuation.
LOOKING AHEAD
As the industry becomes more familiar
with CD, it could become an attractive
choice for many operational sanitization,
disinfection, and sterilization applications
in biopharmaceutical manufacturing.
Next month, Part 2 of this article
will discuss validation and economic
issues and examine methods of
making CD for local use. Because
the US Department of Transportation
will not permit manufactured CD to
be transported, generation must be
performed on-site. That is a major
reason why CD has not been widely
used in biopharmaceutical
manufacturing — but new production
methods are changing things.
REFERENCES
1 Water Disinfection with Ozone,
Chloramines, or Chlorine Dioxide (Report
#20152). American Water Works Association:
Denver, CO, December 1980.
2 Gates DJ. The Chlorine Dioxide
Handbook (Water Disinfection Series).
American Water Works Association:
Denver, CO, June 1998.
3 Young SB, Setlow P. Mechanisms
of Killing of Bacillus subtilis Spores By
Hypochlorite and Chlorine Dioxide. J. Appl.
Microbiol. 95(1) 2003: 54–67.
4 Mittelman MW. Biological Fouling
of Purified Water Systems: Part 1, Bacterial
Growth and Replication. Microcontamination
3(10) October 1985: 51–55, 70.
5 Mittelman MW. Biological Fouling of
Purified Water Systems: Part 2, Detection and
Enumeration. Microcontamination 3(11)
November 1985: 42–58.
6 Mittelman MW. Biological Fouling of
Purified Water Systems: Part 3, Treatment.
Microcontamination 4(1) January 1986: 30–40, 770.
7 National Research Council. Reopening
Public Facilities after a Biological Attack: A
Decision Making Framework. The National
Academies Press: Washington, DC, 2005;
147–8; www.nap.edu/openbook/0309096618/
html/147.html.
8 Eylath A, et al. Successful Sterilization
Using Chlorine Dioxide Gas: Part One —
Sanitizing an Aseptic Fill Isolator. BioProcess
International 1(7) 2003: 52–56.
9 Czarneski MA, Lorheim P. Isolator
Decontamination Using Chlorine Dioxide
Gas. Pharm. Technol. April 2005: 124–133;
www.pharmtech.com/pharmtech/data/
articlestandard/pharmtech/172005/156880/
article.pdf.
10 Eylath A, et. al Successful Sterilization
Using Chlorine Dioxide Gas: Part Two —
Cleaning Process Vessels. BioProcess
International 1(8) 2003: 54–56.
11 Wintner B. Net Present Value Analysis
of Chlorine Dioxide Versus Clean Steam
(unpublished report), April 2005.
12 Wintner B. Using CD to Get to Sterile
Water (unpublished report), March 2005.
13 Kopf H, O’Neill G. Chlorine Dioxide
Solution for Microbial Inactivation in
Crossflow Filtration Systems (unpublished
report). Selective Micro Technologies:
Beverly, MA, April 2005.
14 Wise B, et al. Membranes: Effectiveness
of Chlorine Dioxide in Sanitizing Thin Film
Membrane Systems. Ultrapure Water September
2004: 12–16.
15 O’Neill G. Field Trial:
Decontamination of USP Water Purification
System and Distribution Loop (unpublished
report). Selective Micro Technologies: Beverly,
MA, April 2005.
16 Apel G. Chlorine Dioxide. Tree Fruit
Postharvest J. 4(1): 12–13.
17 Luftman HS, Regits MA. Chlorine
Dioxide Gas Decontamination of the
University of Pennsylvania’s George D.
Widener Large-Animal Hospital Intensive
Care Unit. Presented at the 47th Annual
Biological Safety Meeting, San Antonio, TX,
October 2004 (American Biological Safety
Association, Mundelein, IL, www.absa.org).
18 Warriner R, et al. Disinfection of
Advanced Wastewater Treatment Effluent By
Chlorine, Chlorine Dioxide, and Ozone. Water
Res. 19(12) 1985: 1515–1526.
Barry Wintner is senior bioprocess
consultant, and corresponding author
Anthony Contino is president of
Bioprocess Associates, LLC, PO Box 128,
Morton, PA 19070, 1-610-742-2748,
acontino@bioprocessllc.com.
Gary O’Neill, PhD, is director of research
and product development at Selective
Micro Technologies, Beverly, MA.