Ultraviolet Germicidal Irradiation Handbook

Abstract

Ultraviolet Germicidal Irradiation (UVGI) is electromagnetic radiation that can destroy the ability of microorganisms to reproduce by causing photochemical changes in nucleic acids. Wavelengths in the UVC range are especially damaging to cells because they are absorbed by nucleic acids. The germicidal effectiveness of UVC peaks at about 260–265 nm. This peak corresponds to the peak of UV absorption by bacterial DNA. The germicidal effectiveness of UVC radiation can vary between species and the broader range wavelengths that include UVB also make a small contribution to inactivation (Webb and Tuveson 1982). Although the methods and details of disinfection with ultraviolet light are fairly well understood, to the point that effective disinfection systems can be designed and installed with predictable effects, the exact nature of the effect of ultraviolet light on microorganisms at the molecular level is still a matter of intensive research. This chapter examines the fundamentals of the complex interaction between UV irradiation and cell DNA at the molecular level and provides detailed background information to aid in the understanding of the various biophysical processes that are involved in microbial inactivation.

Full text

Chapter 2
UVGI Disinfection
Theory
2.1 Introduction
Ultraviolet Germicidal Irradiation (UVGI) is electromagnetic radiation that can
destroy the ability of microorganisms to reproduce by causing photochemical
changes in nucleic acids. Wavelengths in the UVC range are especially damaging
to cells because they are absorbed by nucleic acids. The germicidal effectiveness of
UVC peaks at about 260–265 nm. This peak corresponds to the peak of UV absorp-
tion by bacterial DNA. The germicidal effectiveness of UVC radiation can vary
between species and the broader range wavelengths that include UVB also make a
small contribution to inactivation (Webb and Tuveson 1982). Although the meth-
ods and details of disinfection with ultraviolet light are fairly well understood, to
the point that effective disinfection systems can be designed and installed with pre-
dictable effects, the exact nature of the effect of ultraviolet light on microorganisms
at the molecular level is still a matter of intensive research. This chapter examines
the fundamentals of the complex interaction between UV irradiation and cell DNA
at the molecular level and provides detailed background information to aid in the
understanding of the various biophysical processes that are involved in microbial
inactivation.
2.2 UV Inactivation
The spectrum of ultraviolet light extends from wavelengths of about 100–400 nm.
The subdivisions of most interest include UVC (200–280 nm), and UVB
(280–320 nm). Although all UV wavelengths cause some photochemical effects,
wavelengths in the UVC range are particularly damaging to cells because they are
absorbed by proteins, RNA, and DNA (Bolton and Cotton 2008, Rauth 1965). The
germicidal effectiveness of UVC is illustrated in Fig. 2.1, where it can be observed
that germicidal efficiency reaches a peak at about 260–265 nm. This corresponds to
the peak of UV absorption by bacterial DNA (Harm 1980). The germicidal effec-
tiveness of UVC and UVB wavelengths can vary between species. Low pressure
17
W. Kowa l ski , Ultraviolet Germicidal Irradiation Handbook,
DOI 10.1007/978-3-642-01999-9_2, C
Springer-Verlag Berlin Heidelberg 2009

18 2 UVGI Disinfection Theory
0
20
40
60
80
100
200 220 240 260 280 300 320 340 360
Wavelength, nm
Relative Intensity (Effectiveness), %
Germicidal Effectiveness, E. coli
Medium Pressure
UV Lamp
Low Pressure
UV Lamp
Fig. 2.1 Germicidal efficiency of UV wavelengths, comparing High (or medium) and Low pres-
sure UV lamps with germicidal effectiveness for E. coli. Based on data from Luckiesh (1946) and
IESNA (2000)
mercury vapor lamps radiate about 95% of their energy at a wavelength of 253.7 nm,
which is coincidentally so close to the DNA absorption peak (260–265 nm) that it
has a high germicidal effectiveness (IESNA 2000).
If we assume the LP and MP lamps in Fig. 2.1 produce the same total UV
wattage, and multiply spectrum by the germicidal efficiency at each wavelength,
we find the LP lamp has a net germicidal efficiency of 84% vs. 79% for the MP
lamp. The optimum wavelength for inactivating E. coli, about 265 nm, is about
15% more effective than the UVC peak of 254 nm. The optimum wavelength for
inactivating Bacillus subtilis is 270 nm, and this is about 40% more effective than
254 nm (Waites et al. 1988). The optimum wavelength for destroying Cryptosporid-
ium parvum oocysts is 271 nm and this is about 15% more effective than 254 nm
(Linden 2001). Although UVC is responsible for the bulk of the germicidal effects
of broad-spectrum UV, the effects of UVB wavelengths cannot be discounted alto-
gether. In a study by Elasri and Miller (1999) it was found that UVB had about 15%
of the effect of UVC on Pseudomonas aeruginosa.

2.2 UV Inactivation 19
THE STRUCTURE OF DNA
Deoxyribonucleic acid (DNA) is a large, high molecular weight macro-
molecule composed of subunits called nucleotides. Each nucleotide subunit
has three parts: deoxyribose, phosphate, and one of four nitrogenous bases
(nucleic acid bases). The four bases are thymine (T), adenine (A), cytosine
(C), and guanine (G). These four bases form base pairs of either thymine
bonded to adenine or cytosine bonded to guanine. Since thymine always pairs
with adenine, there will be equal amounts of thymine and adenine. Likewise,
cytosine will always exist in amounts equal to guanine. The specific sequences
formed by these base pairs make up the genetic code that forms the chemical
basis for heredity (Atlas 1995). Nucleotides are the basic repeating unit of
DNA and they are composed of nitrogenous bases called purines and pyrim-
idines. These bases are linked to pentoses to make nucleosides. The nucleo-
sides are linked by phosphate groups to make the DNA chain.
DNA forms a double helix, as shown in the figure above, in which two
complementary strands of nucleotides coil around each other. The two out-
side helices of DNA form a backbone that is held together by strong covalent
bonds, locking in the stability of the hereditary macromolecule. Each helix
terminates in a free hydroxyl group at one end, and a free phosphate group at
the other, conferring directionality. The two halves of the DNA molecule run
in opposite directions and coil around each other. Supercoiling may also occur
as long chains of DNA fold and pack into the available space (i.e. in a cell or
viral capsid). Several million nucleotides may be held together in sequence
and they establish the genetic code for each species.
The two complementary chains of the DNA double helix are held
together by hydrogen bonding between the chains. Two of the nitrogenous
bases(C and T) are single-ring structures called pyrimidines and the other
two (A and G) are double-ring structures called purines. The internal hydro-
gen bonds between the base pairs, which hold the entire structure together,
have only about 5% of the strength of the covalent bonds in the outer helix.
Thymine forms two hydrogen bonds with adenine, while cytosine forms three
hydrogen bonds with guanine. The thymine/adenine bond, therefore, repre-
sents the weakest link in the structure.

20 2 UVGI Disinfection Theory
Hydrogen bonds between complementary bases are not the primary stabi-
lizing force of DNA since the energy of a hydrogen bond (2–4 kcal/bond) is
insufficient to account for the observed stability of DNA. Ionic bonds between
the negative phosphate groups and positive cations reduce the electrostatic
repulsion between the negative charges of the sugar-phosphate backbone. The
stability of DNA is also accounted for by the hydrophobic forces associated
with stacking of the bases, which is due to mutual interactions of the bases and
geometrical considerations (Guschlbauer 1976). In polynucleotide chains, this
interaction results in a compact stack of bases that is restricted by the sugar-
phosphate backbone and results in a narrow range of possible overlap angles
between the bases (36◦in DNA). The stacked bases form a hydrophobic core
which favors hydrogen bonding between the complementary strands (see Fig.
2.2). The stacking and twisting of base pairs creates channels in which water
may bond or be excluded depending of DNA conformation (Neidle 1999).
UVA was also found to have a lethal effect on P. aeruginosa although
considerably more lamp power was needed (Fernandez and Pizarro 1996). The UVA
inactivation effect, however, is relatively insignificant and may involve non-actinic
effects (no photochemical changes). Throughout the following discussion only the
actinic bands of UVC and UVB are considered to be operative.
Two general types of nucleic acids exist, ribonucleic acid (RNA) and deoxyri-
bonucleic acid (DNA). Viruses contain DNA or RNA, but not both. During UV
irradiation and inactivation, the most sensitive target of microorganisms is the DNA
of bacteria, the DNA of DNA viruses, the RNA of RNA viruses, and the DNA of
Fig. 2.2 The two helical backbones of DNA are connected by hydrogen bonds between the
nucleotides

2.2 UV Inactivation 21
fungi. RNA has D-ribose as its main constituent and adenine, cytosine, guanine,
and uracil as bases. DNA has 2-deoxy-D-ribose as its main constituent and adenine,
cytosine, guanine, and thymine as bases. Hydrogen bonds link the bases. UV radi-
ation can cause a crosslink between two thymine bases that is more stable than a
hydrogen bond (Casarett 1968). Bacteria and fungi have DNA, while viruses may
have either DNA or RNA. DNA and RNA are responsible for microbial replication
and protein synthesis and damage to these nucleic acids results in inactivation or the
failure to reproduce.
UV wavelengths inactivate microorganisms by causing cross-links between
constituent nucleic acids. The absorption of UV can result in the formation of
intrastrand cyclobutyl-pyrimidine dimers in DNA, which can lead to mutations or
cell death (Harm 1980, Koller 1952, Kuluncsics et al. 1999). Pyrimidines are molec-
ular components in the biosynthesis process and include thymine and cytosine (see
Fig.2.2). Thymine and cytosine are two of the base pair components of DNA, the
others being adenine and guanine. The primary dimers formed in DNA by UV expo-
sure are known as thymine dimers (see the page on the Structure of DNA). The
lethal effect of UV radiation is primarily due to the structural defects caused when
thymine dimers form but secondary damage is also produced by cytosine dimers
(David 1973, Masschelein 2002). Various other types of photoproducts are also
formed that can contribute to cell death. Photohydration reactions can occur under
UV irradiation in which the pyrimidines cytosine and uracil bond with elements of
water molecules. The same reaction does not occur with thymine. The photohydra-
tion yield is independent of wavelength.
Double stranded RNA has a higher resistance to UV irradiation than single
stranded RNA, and this may be due to various factors, including more structural
stability (Becker and Wang 1989) and the redundancy of information in the com-
plementary strands (Bishop et al. 1967). Ultraviolet light also causes photochemical
reactions in proteins in the cell other than DNA and UV absorption in proteins peaks
at about 280 nm. There is also some absorption in the peptide bonds within proteins
at wavelengths below 240 nm.
Figure 2.3 illustrates how UV absorption can lead to cross-linking between adja-
cent thymine molecules and the formation of thymine dimers. When thymine bases
happen to sit next to each other, the pair is called a doublet. The dimerization of
thymine doublets by UV can lead to inactivation of the DNA, or RNA, with the
result that cell may be unable to reproduce effectively.
Fig. 2.3 Thymine dimers
causedbyUVabsorptionin
adjacent nucleotides (thymine
doublets)

22 2 UVGI Disinfection Theory
The exact mechanism by which UV causes thymine dimers is not com-
pletely understood. Bhattacharjee and Sharan (2005) demonstrated that exposing
E. coli DNA to UVC irradiation induced sparsely placed, dose-dependent, single
strand breaks, and proposed that the conformational relaxation generates negative
super-coiling strain on the DNA backbone.
It has been repeatedly demonstrated that thymine dimers produced by UV
exposure result in the inactivation of bacteria and DNA viruses. A dose of 4.5 J/m2
is reported to cause 50,000 pyrimidine dimers per cell (Rothman and Setlow 1979).
It has been reported that 100 J/m2induces approximately seven pyrimidine dimers
per viral genome in SV40, which is sufficient to strongly inhibit viral DNA synthe-
sis (Sarasin and Hanawalt 1980). Thymine dimers form within 1 picosecond of UV
excitation provided the bases are properly oriented at the instant of light absorp-
tion (Schreier et al. 2007). Only a few percent of the thymine doublets are likely
to be favorably positioned for reaction and dimerization at the time of UV excita-
tion. Figure 2.4 illustrates the dimerization process for a thymine doublet with the
appropriate orientation.
The two most common conformations of DNA are called A-DNA and B-DNA.
Molecular orientations can vary due to A and B conformation and vibrational or
other movement in the DNA molecule. The average twist angle between succes-
sive base pairs differs between the A conformation and the B conformation is
only a few degrees. The smaller amount of conformational variation in A-DNA vs.
B-DNA explains the greater resistance of A-DNA to cyclobutane pyrimidine dimer
Fig. 2.4 The photodynamics of dimerization. A single strand of the DNA sugar-phosphate
backbone is shown with thymine nucleotides. UV excitation populates the singlet state ππ∗,which
decays into the singlet nπ∗state (left). All energy is converted internally to form a thymine double
hydrogen bond (right)

2.2 UV Inactivation 23
Fig. 2.5 Cross-linking between thymine nucleotides (or uracil nucleotides in the case of RNA
viruses) can occur between adjacent strands of DNA (or RNA). It can also occur between the DNA
(RNA) and the proteins of the capsid, for viruses
formation (Schreier et al. 2007). That is to say, the more dense packing of bases
and lower flexibility in the B-DNA form ensures a higher probability that thymine
doublets will be available for dimerization.
The exact sequences of thymines, cytosines, adenines, and guanines in DNA can
directly impact the probability of dimerization. Adjacent pyrimidines (thymine and
cytosine) are considerably more photoreactive than adjacent purines (adenine and
guanine). Becker and Wang (1989) found that 80% of pyrimidines and 45% or
purines form UV photoproducts in double-stranded DNA.
In addition to cross-links between adjacent thymines, UV may also induce
cross-links between non-adjacent thymines, as illustrated in Fig.2.5. Cross-linking
can also occur between the nucleotides and the proteins in the capsid of viruses,
damaging the capsid of DNA viruses.
Cross-linking can also occur with cytosine and guanine, but the energy required is
higher due to their having three hydrogen bonds instead of two for thymine/adenine
bonds, and so thymine dimers predominate. Besides cross-links with adjacent
thymine nucleotides and with thymine in adjacent strands of DNA, thymine may
also form links with proteins, including proteins in the capsid (in the case of viruses)
as shown in Fig. 2.6. Other biological molecules with unsaturated bonds like coen-
zymes, hormones, and electron carriers may be susceptible to UV damage.
In RNA, whether in prokaryotic cells, eukaryotic cells, or viruses, uracil takes
the place of thymine. Inactivation of RNA viruses involves cross-linking between
the uracil nucleotides and the creation of uracil dimers (Miller and Plageman 1974).
Uracil dimers may also damage the capsid of RNA viruses. Some limited quan-
titative data is available on the specific nature of DNA damage produced by UV
absorption. Miller and Plageman (1974) demonstrated that ultraviolet exposure of
Mengovirus caused rapid formation of uracil dimers and that this appeared to be the

24 2 UVGI Disinfection Theory
Fig. 2.6 Thymine dimerization can also occur between DNA/RNA and adjacent protein
molecules, such as in cell cytoplasm or the capsid of a virus
primary cause of virus inactivation (i.e. loss of infectivity). A maximum of about 9%
of the total uracil bases of the viral DNA formed dimers within 10 min of UV irradia-
tion. Results also indicated that viral RNA became covalently linked to viral protein
as a result of irradiation. A slower process of capsid destruction also occurred in
which capsid proteins were modified and photoproducts were formed. UV irradia-
tion of the virus also caused covalent linkage of viral RNA to viral polypeptides,
apparently due to close proximity between the RNA and proteins in the capsid. The
amount of protein covalently linked to the RNA represented not more than 1.5% of
the total protein capsid. Smirnov et al. (1992) studied Venezuelan equine encephali-
tis (VEE) under UV irradiation and found evidence suggesting that the formation of
uracil dimers led to extensive contacts of the RNA with protein in the nucleocapsid.
Viruses containing many thymine dimers may still be capable of plaque forma-
tion (Rainbow and Mak 1973). An E. coli chromosome exposed to UVB produced
pyrimidine photoproducts in the following proportions: 59% thymine dimers, 34%
thymine-cytosine dimers, and 7% cytosine-cytosine dimers (Palmeira et al. 2006).
Figure 2.7 shows the rate at which uracil dimers form under irradiation, shown in
terms of the uracil bonds remaining intact in RNA. This plot is shown alongside the
decay rate for Mengovirus. It can be observed that the virus is rapidly inactivated
while the formation of uracil dimers proceeds relatively slowly. The scale of the
chart is limited, but the virus goes through six logs of reduction before 9% of the
uracil is cross-linked. Clearly, it takes but little cross-linking to inactivate a virus.
The ratio of the microbial inactivation rate to the dimer production rate should be a
constant for any given species. Theoretically, each species should have a character-
istic inactivation rate that is a function of the dimerization probability.

2.3 UV Absorption Spectra 25
10
100
0 1000 2000 3000 4000 5000 6000 7000
UV Exposure Dose, J/m2
Virus Survival / Uracil Bonds %
Virus Infectivity
Uracil Bonds
D37 = 70 J/m2
Fig. 2.7 Survival of Mengovirus under UV irradiation, plotted along with the percentage of intact
uracil bonds in viral RNA. Based on data from Miller and Plageman (1974)
2.3 UV Absorption Spectra
An absorption spectrum is a quantitative description of the absorptive capacity of a
molecule over some specified range of electromagnetic frequencies. The absorption
of ultraviolet light by a molecule will result in altered electronic configuration, con-
version into radiant energy, rotation, and vibration. When these energy levels are at
a minimum the molecule is in a ground state. The energy imparted to a molecule
by UV absorption produces an excited state. The capacity of a molecule to absorb
UV energy over a band of wavelengths is described in terms of an absorption spec-
trum. The intensity of absorption is generally expressed in absorbance or optical
density. The intensity of an absorption band is directly related to the probability that
the particular transition will take place when a photon of the right energy comes
along. Figure 2.8 represents the absorption spectra for the four DNA bases, which
have peaks in the UVC band, and also below 220 nm, which is in the VUV range.
Thymine and cytosine both have strong peaks near 265 nm.
Pyrimidines (thymine, cytosine, and uracil) absorb about ten times more UV
than purines. The quantum yield at 254 nm is φ~10–3 for pyrimidines and for
purines φ~10–4. The capacity of a molecule to absorb light of a particular wave-
length depends on both the electronic configuration of the molecule and on its
available higher energy states (Smith and Hanawalt 1969). An absorption spec-
trum may be regarded as a summation of a series of individual absorption bands,
each corresponding to a transition between two particular electronic configurations
(Hollaender 1955). This transition typically occurs when an orbital electron is raised
from the normal ground state to an excited state. These transitions occur only in

26 2 UVGI Disinfection Theory
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
200 220 240 260 280 300 320 340
Wavelength, nm
Absorbance, (L /mol-cm)x10−3
THYMINE
Adenine
Guanine
Cytosine
Fig. 2.8 Comparison of response spectra for the four main nucleotides
discrete jumps, and therefore only a specific quantum of energy can be absorbed.
The electronic configuration of an excited molecules can be a very transitory event,
and only those excitations of sufficient duration will have a high probability of
absorption.
An absorption band may be described by the width of the band (or the range
of wavelengths) and the degree of absorption (absorptivity). The width of the band
is defined as the spectral separation between the points of half-maximum (50%)
absorption. The width of a band is inversely dependent upon the duration of the
excited electronic state. Integration of the absorption band over the width determines
the probability that the particular transition will occur when a photon of the right
energy comes along.
The absorption spectrum is typically measured by beaming light through a trans-
parent solution containing microbes or molecules and comparing it against the pure
solution. The transmittance (or transmissivity), T, of a solution is defined as:
T=I
I0(2.1)
where
I=irradiance of light exiting the solution, W/m2
I0=irradiance of light entering the solution, W/m2
Beer’s Law establishes a relationship between the transmittance in Eq. (2.1) and
the absorbance, A, as follows:

2.3 UV Absorption Spectra 27
T=I
I0=10−A(2.2)
The absorbance (or absorptivity), also called the optical density, OD, is defined
as:
A=εlc (2.3)
where
ε=molar absorptivity, liters/mole-cm
l=thickness of the solution, cm
c=concentration of solute, moles/liter
An alternate form of Beer’s Law is:
I
I0=e−nsl (2.4)
where
n=number of molecules per unit volume
s=absorption cross-section, m2or μm2
The absorption cross-section represents the product of the average cross-
sectional area of the molecule and the probability that a photon will be absorbed.
The absorption cross-section is related to the molar absorptivity by the following
relation:
s=3.8x10−21ε(2.5)
Strong absorption bands in the ultraviolet region correlate with molecular struc-
tures containing conjugated double bonds. Ring structures, such as the pyrimidines
and purines, exhibit particularly strong absorption and can define the overall absorp-
tion spectrum of DNA. Von Sonntag (1986) reports that DNA has a peak of UV
absorption not only at 265 nm but also at 200 nm. Most of the absorption at 200 nm
occurs in the DNA backbone molecules of ribose and phosphate. At 265 nm, most
of the absorption occurs at the nucleotide bases, thymine and adenine, and cytosine
and guanine, but dimers of thymine are by far the most common UV photoprod-
ucts. In RNA-based microbes, uracil is also involved in UV absorption in place of
thymine but not necessarily to the same degree. Figure 2.9 compares the absorption
spectrum of uracil with that of thymine. It can be observed that not only are the
absorption spectra very similar for these nucleotides, but that the mercury emission
line at 254 nm is more nearly aligned with the peak absorption of uracil.
Carbohydrates make up about 41% by weight of nucleic acids, but they show
essentially no UV absorption above about 230 nm and would not be expected to

28 2 UVGI Disinfection Theory
0
1
2
3
4
5
6
7
8
9
10
200 220 240 260 280 300 320
Wavelength, nm
Absorbance, (L/mol-cm)x10−3
THYMINE
URACIL (RNA Viruses)
Low Pressure UV Lamp Emission
at 253.7 nm
Fig. 2.9 Comparison of thymine UV absorption spectra with uracil, the nucleotide that takes the
place of thymine in RNA viruses
participate in photochemical reactions at around 254 nm. However, certain photo-
chemical processes that produce uracil radicals can result in chemical alterations to
the carbohydrates of nucleic acid (Smith and Hanawalt 1969).
The ultraviolet absorption spectrum of a polymer is not necessarily the linear
sum of its constituents. This nonadditivity is referred to as hyperchromicity.Ifthe
absorbance of a given oligonucleotide is higher than its constituents molecules, it is
hyperchromic. Hyperchromicity is largely explained by the coulombic interaction
of the ordered bases in the polymer. Hyperchromicity is a kind of resonant elec-
tronic effect in which the partial alignment of transition moments by base stacking
results in coupled oscillation. A relatively small number of bases in a DNA strand
are required for such coupling and about 8–10 base pairs can exhibit roughly 80%
of the hyperchromism of an infinite helix. Becker and Wang (1989) present data that
indicates that hyperchromicity may add about twice the number of photoproducts
when strings of eight or more thymines occur sequentially.
2.4 UV Photoprotection
Microbes have various mechanisms by which they can protect themselves from
UV exposure, including nucleocapsids and cytoplasm which may contain UV
absorbing proteins (i.e. dark proteins). The absorption of UV in any surrounding
complex of proteins will reduce the density of photons reaching the nucleic acid
and thereby provide photoprotection. Comparisons of virus inactivation with

2.4 UV Photoprotection 29
inactivation of purified DNA show the absorption spectrums are not identical, the
implication being that UV is absorbed in the envelope, the nucleocapsid, or other
protein-laden constituents of the viroid, although in some cases the nucleic acid is
more resistant in isolation (Zavadova et al. 1968, Furuse and Watanabe 1971, Bishop
et al. 1967). In bacteria, the cytoplasm may offer photoprotection due to its UV
absorptivity. Unrau et al. (1973) have suggested that there is a in vivo shielding effect
in Bacillus subtilis since dimer formation is doubled when its DNA is irradiated
separately, although they do not attribute this effect to the cytoplasm. Fungal spores
are among the most resistant microbes and they often have melanin-containing dark
pigmented conidia. The photoprotective component melanin increases the survival
and longevity of fungal spores (Bell and Wheeler 1986). Aspergillus niger conidia
are more resistant to UV due to the high UV absorbance of their melanin pigments
(Anderson et al. 2000). Various studies on fungi have suggested that lighter-
pigmented thin-walled conidia are more susceptible to UV than thicker-walled
dark-pigmented conidia (Boyd-Wilson et al. 1998, Durrell and Shields 1960, Valero
et al. 2007). UV scattering (addressed later) can also contribute to photoprotection.
Figure 2.10 illustrates photoprotection mechanisms in viruses – UV scatter-
ing by the envelope, UV absorption by the envelope, and UV absorption by the
nucleocapsid the latter being mostly negligible. UV scattering, occurs when the par-
ticle is in the Mie scattering size range, and the effective scattering cross-section my
be much larger than the actual physical cross-section of the particle (see Sect. 2.13).
Chromophores are chemical groups in molecules that are capable of absorb-
ing photons. Polyatomic molecules have fairly broad absorption bands. In proteins,
the molecular groupings which give rise to absorption are principally amino acids,
Fig. 2.10 Schematic illustration of the levels of photoprotection of an enveloped virus, including
the UV scattering cross-section, the envelope, and the nucleocapsid

30 2 UVGI Disinfection Theory
which have absorption peaks at about 280 nm (Webb 1965). For the nucleic acids,
the chromophores responsible for the absorption peak around 265 nm are the bases,
the purines and pyrimidines, the dimers of which are considered primary products
of the biocidal action of UV. The chromophores that are likely to confer protection
to these bases are those in the cytoplasm or cell wall of bacteria, in the capsid or
protein coat, if any, of the virus, and the spore coat and cortex of a spore.
Some amino acids are optically transparent above UV wavelengths of about
240 nm, while others, the chromophores, have high molar absorptivities at or
near 253.7 nm. The molar absorptivity of a compound is the probability that the
wavelength will be absorbed. Figure 2.11 shows the molar absorptivity of amino
acids at 253.7. These might be called ‘dark proteins’ because of their relatively
high absorbance for UV wavelengths. Microbes that contain higher proportions of
chromophores will likely absorb UV photons that would otherwise be absorbed by
DNA and cause dimerization. However, degradation of the proteins in a cell may
also contribute to cell death.
Enzymes are proteins that function as efficient biological catalysts that increase
the rate of a reaction. Biological systems depend on enzymes to lower the activation
energy of a chemical reaction and thereby facilitate processes of growth, and repair
(Atlas 1995). Enzymes consist of various proportions of the amino acids including
those in Fig. 2.11 and their quantum yield will vary accordingly. The quantum yield
indicates the probability that absorbed UV light will induce a chemical change.
Table 2.1 lists several enzymes, their chromophore constituents, and their measured
quantum yields, based on data from Webb (1965).
0.1
1
10
100
1000
10000
Tryptophan Tyrosine Cystine Phenylalanine Histidine Acetylalanine
Molar Absorptivity at 253.7 nm
Fig. 2.11 Molar absorptivity of ‘dark proteins’ or amino acids with a relatively high probability
of absorbing light at 253.7 nm. Based on data from Webb (1965)

2.5 Covalent Bonding and Photon Interaction 31
Table 2.1 Quantum yields for enzyme inactivation by UV at 253.7 nm
Protein Chromophores Quantum Yield
Ribonuclease Cys4·His4·Phe3·Tyr 16 0.027
Lyzsozyme Cys5·His1·Phe3·Tyr8·Tyr20.024
Trypsin Cys6·His1·Phe3·Tyr4·Tyr 40.015
Insulin Cys18 ·His12 ·Phe18 ·Tyr 24 0.015
Subtilisin A His5·Phe4·Try1·Ty r13 0.007
Japaneses Nagarse His6·Phe3·Try4·Ty r10 0.007
Subtilisin B His6·Phe3·Try4·Tyr 10 0.006
Chymotrypsin Cys5·His2·Phe6·Tyr 7·Tyr40.005
Pepsin Cys3·His2·Phe9·Tyr 4·Tyr16 0.002
Carboxypeptidase Cys2·His8·Phe15 ·Tyr 6·Tyr20 0.001
Since enzymes are catalysts, they are not consumed during normal biological
processes and are relatively few in number, and may therefore contribute little pro-
tective effect. However, their destruction will inhibit repair processes during or after
UV exposure and may limit the effective UV rate constant. Inactivation of enzymes
can be higher at wavelengths other than 253.7 and broadband UV irradiation is
reported to be more effective at eliminating repair enzymes than narrow band UVC
(Zimmer and Slawson 2002, Hu et al. 2005). Enzymes are associated with bacterial
cells and not with viruses, although some viruses (i.e. bacteriophages) may employ
enzymes for self-repair from the cells they parasitize.
Powell (1959) used optical density measurements to estimate the reduction of UV
absorption at 265 nm in Herpes Simplex virus due to shielding by the host cell. The
cells had a radius of 6 μm and this thickness was estimated via the Beer-Lambert
law to result in a transmission, including corrections for scattering, of 40%, which
is an attenuation of 60%. Such levels of photoprotection may be possible for other
bacterial cells in this size range. Viruses, however, have such relatively thin coats
that it seems unlikely that any chromophores present would provide any significant
photoprotection.
2.5 Covalent Bonding and Photon Interaction
Chemical bonding between atoms occurs when a single electron is shared between
more than one atomic nucleus. The wave function between the two atomic orbitals is
called a molecular orbital. Two kinds of bonding molecular orbitals may be involved
in complex molecules, σ(sigma) orbitals, and π(Pi) orbitals. The σorbitals are
localized around two nuclei and the πorbitals are nonlocalized and may involve
two or more nuclei. The larger the nonlocalized orbital the more spread out is
the electron probability distribution, and the longer the wavelength for electrons
in that orbital. The longer wavelength means lower energy and more stability. Con-
jugated ring structures like the pyrimidines and the purines have large nonlocalized
πorbitals and stable structures (Smith and Hanawalt 1969). A single bond is usually

32 2 UVGI Disinfection Theory
aσbond while double bonds may involve both σand πorbitals. There can be no
free rotation about a double bond.
An incoming UV photon will promote an electron to an orbital of higher energy,
which may be a new set of antibonding orbitals called σ∗or π∗orbitals, which may
result in weakening of the bonds. A ππ∗transition involves the excitation of a π
electron into a π∗state. The ππ∗transitions are responsible for the most intense
absorption bands in molecular spectra.
An ultraviolet photon at 253.7 nm has an energy of 4.9 eV and if this were totally
converted to vibrational energy it would be sufficient to break chemical bonds, but
the energy becomes distributed over many possible vibrational modes. Upon absorp-
tion of a UV photon, which may take 10–15 s, molecules may briefly exist in an
excited state before the energy is dissipated, either by re-emission or by vibration
and rotational modes. Differential quantized modes of vibration can be represented
as levels of potential energy, the first of which is the singlet state. The triplet state
may also be stimulated and it is one in which the system has two electrons with
unpaired spins. The triplet state may persist for 10–3 s. The triplet state does not, as
a rule, degrade directly back to the ground state, but it allows more time for photo-
chemistry to occur and the probability of a chemical reaction is briefly increased.
2.6 UV Photoproducts
Thymine dimers are formed when two thymine molecules are cross-linked between
their respective 5 and 6 carbon atoms, forming a cyclobutane ring. There are six
possible isomers of the thymine dimer. Dimers can both be formed by UV exposure
and separated or monomerized by UV. At longer UV wavelengths (about 280 nm)
the formation of the dimer is favored while at shorter wavelengths (about 240 nm)
monomerization occurs due to differences in the absorption spectra of thymine and
its dimer and in the quantum yields for the formation and splitting of the dimer.
The maximum yield of cyclobutane dimers is dependent on equilibrium between
the formation and splitting of dimers. The reversal of dimerization by wavelengths
of UV or visible light is known as photoreactivation, as is the repair of dimers by
enzymes.
Thymine dimers can be formed by wavelengths of light that are not directly
absorbed if the thymine molecule is in close proximity to other molecules that
absorb these wavelengths. This process is called molecular photosensitization and it
requires that the triplet state of the absorbing species (the photosensitizer) be slightly
higher in energy than the triplet state of the thymine. Upon absorption, the triplet
energy of the photosensitizer is transferred to the thymine molecule where it may
induce dimerization. There are at least five other dimers of the natural pyrimidines,
including cytosine dimers, cytosine-thymine dimers, uracil dimers, uracil-thymine
dimers, and uracil-cytosine dimers. Cytosine dimers are formed at lower rates than
thymine dimers and are readily converted to uracil dimers.
Cytosine hydrate, a water addition photoproduct, can be formed in RNA and
single-stranded DNA but is not commonly found in irradiated double-stranded DNA
(Smith and Hanawalt 1969). Uracil hydrates can be formed from the excited singlet

2.7 DNA Conformation 33
state. Uracil can be derived from the monomerization of cytosine dimers. The for-
mation of hydrates is greatly favored in single-stranded DNA.
Many other pyrimidine photoproducts, besides hydrates and cyclobutane dimers,
can be produced and may be at least partly responsible for damage to nucleic acids
or to a cell. Chief among these is the spore photoproduct, also called the azetane
thymine dimer. The spore photoproduct, so named because it was first noted in
spores, can be formed from as much as 30% of the thymine. The spore photoproduct
is a type of thymine dimer that cannot be photoreversed (although it can be repaired)
and the yield of this product can approach the maximum determined from the num-
ber of thymines that are nearest neighbors in DNA. In the normal B conformation
of DNA the planes of the bases are parallel to each other and perpendicular to the
helical backbone, and the cyclobutane dimers are favored. In the dehydrated A con-
formation, which is the form in which DNA is held by spores, the planes of the
bases are parallel but they are inclined at an angle of 70◦to the axis of the helix, a
conformation which favors the spore photoproduct.
DNA cross-linking can occur under UV irradiation and this apparently involves
cyclobutane dimers. Cross-linking can be highly fatal to DNA but such lesions do
not appear to play a major role in UV inactivation since the cyclobutane dimers and
other photoproducts are largely responsible for the inactivation effect. Per Eden-
berg (1983), the hypothesis that DNA replication forks are halted upon encounter-
ing thymine dimers in the template strand is consistent with data on inhibition of
Simian virus replication by ultraviolet light. Per Stacks et al. (1983), the percent-
age of repaired and completed molecules containing dimers increases with time
after irradiation ceases, and they postulate that the cellular replication machinery
can accommodate limited amounts of UV-induced damage and that the progressive
decrease in simian virus 40 DNA synthesis after UV irradiation is due to the accu-
mulation in the replication pool of blocked molecules containing levels of damage
greater than that which can be tolerated.
DNA may also cross-link to proteins in the cell wall, nucleocapsid, or cytoplasm,
forming potentially fatal lesions. Amino acids that may contribute to photoreactivity
in DNA and that may impact cross-linking include cysteine, cystine, tyrosine, serine,
methionine, lysine, arginine, histidine, tryptophan, and phenylalanine. Under dry
conditions (A DNA) the yield of thymine dimers is greatly decreased but there is an
increase in the amount of DNA cross-linked to protein (Smith and Hanawalt 1969).
Per Becker and Wang (1989), the ability of UV to damage a given base in DNA by
inducing dimers or photoproducts is determined by two factors, the DNA sequence
and the flexibility of DNA. Upon absorption of a UV photon, only those bases that
are in a geometry capable of easily forming a photodimer can photoreact.
2.7 DNA Conformation
DNA molecules can exist in two conformations, A or B (Eyster and Prohofsky
1977). The UV susceptibility differs between the conformations. In the A confor-
mation the bases are tilted with respect to the helix axis. In the B conformation the

34 2 UVGI Disinfection Theory
bases are roughly perpendicular to the double helix axis. The interaction of electro-
static and van der Waals forces at the molecular level are influenced by the presence
of water. The B conformation is fully hydrated (i.e. in solution or even in air at 100%
relative humidity) and the A conformation could therefore be considered to be the
dehydrated state. The dry A conformation shrinks in length in comparison with the
wet DNA, and transitions through a phase when the population is mixed with cells
in both A and B conformations. In general, microbes in high relative humidity or
in water (B-DNA), have a higher resistance to UV (Peccia et al. 2001). Microbes
transition from A to B when humidity or moisture increases and it is possible that
the more compact A conformation (see Fig. 2.12 ) lends itself to more cross-links,
but it is also possible that the presence of moisture or bound water provides extra
protection or improved self-repair mechanisms, or that a combination of these
factors is responsible for the difference in UV susceptibility between the A and
B conformation.
DNA undergoes conformational transition from the B form to a disordered form
as the relative humidity is lowered from about 75% to 55%. At the lower relative
humidity, the dry condition, the bases are no longer stacked one above the other but
are slightly angled with reference to the helix and DNA films equilibriated between
75 and 100% RH show no conformational changes and are assumed to be entirely in
the B-form (Rahn and Hosszu 1969). The yield of thymine dimers remains constant
Fig. 2.12 DNA can exist in two states, the hydrated B conformation (left), and the dehydrated A
conformation (right) with tighter packing of the nucleotides

2.8 Photon Density and Single-Hit Concepts 35
in this range and is the same as that found in solution. Although most air-based UV
rate constants in the range of 75–100% RH tend to converge towards water-based
UV rate constants, they do not appear to become equivalent. One possible reason
for this nonequality is that the refractive index of UV in air is different from that in
water, causing differences in the photoprotective effect due to UV scattering.
At lower relative humidities, DNA transitions to the A form with more order and
less probability of contact between thymine bases during irradiation, and there is a
reduction in the rate of thymine dimer formation. The bases have different affinities
for water and can trap available water molecules. The purines have two principal
hydration sites in each of the major and minor DNA grooves, while the pyrimidines
have only one hydration site in each groove (Neidle 1999). The individual hydration
sites for bases in the A and B conformations are much the same, the major difference
being that in B-DNA water is found in both grooves equally while in A-DNA more
water is found in the major groove than in the minor groove.
The B form of DNA contains more bound water molecules, including those that
attach to the internal grooves of DNA. The A form of DNA leads to the exposure
of more hydrophobic portions of the sugar units of the backbone compared to the
B form (Neidle 1999). In A-DNA, water molecules are displaced from the shallow
groove, creating a local environment of low water content that favors and stabilizes
the A form. The mere presence of water is insufficient to induce a conversion from
A-DNA to B-DNA, instead, the water molecules must be able to contact the DNA
directly over its entire length. In spores, the DNA is typically maintained in a tightly
packed hydrophobic environment which prevents the DNA from going into the B-
form even under high humidity or in solution which partly explains their higher UV
resistance. The interaction of water molecules with DNA indicates that water forms
an integral part of DNA structure and stability, and can impact UV inactivation rates.
2.8 Photon Density and Single-Hit Concepts
It can be informative to consider UV energy incident upon a microbe in terms of the
number of photons, or the photon density per unit surface area. Each photon carries
an amount of energy called a quantum, , determined from quantum mechanics as
(Modest 1993):
ε=hv (2.6)
where
h=Planck’s constant, 6.626×10–34 Js
v=frequency, cycles per sec or Hz
The energy of a mole of photons is called an Einstein. It is defined as:
Einstein =Nhv (2.7)
where N =Avogadro’s number, 6.022×1023

36 2 UVGI Disinfection Theory
The frequency of UVC light at a wavelength of 253.7 nm is 1.18×1015 Hz, and
the energy of UVC is computed to be 7.819×10–19 J/photon. Inverting this value
gives us 1.279×1018 photons/Joule.AUVdoseof10J/m
2produces 1.279×1019
photons per m2. A virus of 0.1 micron diameter has a cross-sectional area of
3.14×10–14 m and will be subject to the passage of about 401,000 photons when
exposed to 10 J/m2, which is sufficient to highly inactivate most viruses.
Despite of the vast number of photons passing through a virus, only an extremely
small number are absorbed. Klein et al. (1994) report that Vaccinia virus experi-
enced some 15 dimers per genome after a dose of 8 J/m2. Miller and Plageman
(1974) found that 1.7 uracil dimers were formed per PFU of inactivated Mengovirus.
Based on data from Rainbow and Mak (1973), 100 J/m2produced about 102 dimers
in Adenovirus Type 1, and a lethal hit (D37) involved 30 thymine dimers and one sin-
gle strand break. Per Ryan and Rainbow (1977), 0.3 dimers and 3.5 uridine hydrates
were formed per three lethal hits in herpes simplex virus. Cornelis et al. (1981)
reports that UV dosing of Parvovirus H-1 produced 10 dimers per genome, and that
80 dimers were formed in Simian virus SV40. Peak and Peak (1978) report that a
frequency of 0.3 single-strand breaks occurs per lethal hit in phage T7. Sarasin and
Hanawalt (1978) report that a 100 J/m2dose results in 7 pyrimidine dimers per the
SV40 genome. Studies with phage T7 DNA suggest a rate of damage of 0.21 sites
per 10,000 base pairs per 10 J/m2(Hanawalt et al. 1978). Clearly, very few photons
out of the total interact photochemically with the nucleotides, implying that virions
are virtually transparent to UV.
The First Law of Photochemistry (Grotthus-Draper Law) states that light must
be absorbed by a molecule before any photochemical reactions can occur. The Sec-
ond Law of Photochemistry (Stark-Einstein Law) states that absorbed light may not
necessarily result in a photochemical reaction but if it does, then only one pho-
ton is required for each molecule affected (Smith and Hanawalt 1969). Since not
every quantum of incident energy is absorbed by a molecule, there is an absorption
efficiency that describes photochemical absorptivity. This efficiency is called the
quantum yield, φ, and it is defined as:
=Nc
Np(2.8)
where
Nc=Number of molecules reacting chemically
Np=Number of photons absorbed
The number of photons absorbed is sometimes specified in Einsteins, or moles of
photons as defined in Eq. (2.7). Quantum yields may be extremely low for macro-
molecules of low absorptivity. Since most of these photochemical excitations of
molecules do not lead to chemical reactions, energy is dissipated by various means.
Light may be re-emitted at a different wavelength, and energy absorption may result
in molecular vibrations that translate into heat.
The inactivation of nucleic acids involves quantum yields on the order of 10–3 to
10–4 and at the UV doses typically applied, it is clear that a relatively small number

2.8 Photon Density and Single-Hit Concepts 37
of photons photoreact, and that they are absorbed at discrete genomic sites – nor-
mally those bases that produce dimers. Based on UV inactivation studies of E. coli,
only about 0.025% of the DNA molecule is photochemically altered at the point that
99% of E. coli are killed (Smith and Hanawalt 1969). A UV dose of 180 J/m2dimer-
izes only 0.1% of the total thymine. This dose represents (180 J/m2)(1.279×1018
photons/J) =2.3×1020 photons/m2. With a diameter of about 0.5 microns, and a
DNA size of 5490 kb, this would imply that 1.8×108photons impinged upon the
bacterial cell to produce about 1372 photochemical reactions. Even if we ignore
the cell space that is not occupied by the bacterial DNA, it is clear that it takes a
relatively large number of photons to induce a relatively limited number of pho-
toreactions sufficient to inactivate a cell. Figure 2.13 shows a comparison of the
typical number of photons necessary for certain processes to provide some further
perspective.
Rauth (1965) measured the inactivation and absorption cross-sections of sev-
eral viruses and computed the quantum yields to be in the range of 5–65×10–4.
The quantum yield is computed from the inactivation cross-section divided by the
absorption cross-section as follows:
=σ
S(2.9)
where
σ=inactivation cross-section, m2/photon
S=absorption cross-section, m2/photon
1
10
100
1000
10000
100000
1000000
Photons
Vision Photosynthesis Enzyme Virus Bacteria
Fig. 2.13 Typical numbers of absorbed photons necessary for certain processes or to inactivate
enzymes, bacteria, and viruses. Based on Setlow and Pollard, 1962

38 2 UVGI Disinfection Theory
The absorption cross-section can be determined by various means but the mea-
surement typically involves putting specific densities of cells (or virions) in solution
and measuring the difference in irradiance that passes through the solution con-
taining cells versus a solution containing no cells. The inactivation cross-section is
equivalent to the UV rate constant, which is normally given in units of m2/J.
2.9 Photochemistry of RNA Viruses
RNA viruses are characteristically in the A conformation and this partly dictates the
type of photoproducts produced under UV irradiation. The main photoproducts pro-
duced are pyrimidine hydrates and cyclobutadipyrimidines, while other photoprod-
ucts, like altered purines and pyrimidine dimers, occur at much lower rates, if at all
(Fraenkel-Conrat and Wagner 1981). Remsen et al. (1970) found that inactivation
of R17 phage at 280 nm was a log-linear function of the number of uridine hydrates
formed and that no cyclobutapyrimidines were formed. RNA animal viruses and
phages demonstrate little or no photoreactivation, and the photoreactive effects that
have been observed are attributed to host cells (Fraenkel-Conrat and Wagner 1981).
RNA viruses have no repair enzymes and any photoreactive effects may be due
strictly to thermal, visible light, or near-UV light effects.
Much information on the photochemistry of RNA viruses has come from
studies of Tobacco Mosaic Virus (TMV), and although plant viruses are not the
subject of this text some useful information can be garnered from UV studies regard-
ing the protective effect of protein coats. The quantum yield for inactivation of the
whole TMV virus is barely 1% that of RNA at longer wavelengths of the absorp-
tion spectrum (Fraenkel-Conrat and Wagner 1981). The relative insensitivity to UV
of the whole virus is a property of the coat protein, which modifies the UV photo-
products formed in RNA. Protein inhibits the formation of pyrimidine photoprod-
ucts (cyclobutadipyrimidines) and inhibits the formation of other photoproducts by
reducing the quantum yield for photoreactions. This could occur through shielding
of the RNA, through quenching of the excited states of RNA, and by surrounding
the bases with a hydrophobic environment and limiting the mobility of the individ-
ual bases. Although the protein coat reduces the overall rate of photoreactions, it
allows the formation of noncyclobutane-type dipyrimidines and of uridine hydrates.
In irradiated TMV, the number of uridine hydrates formed was about two per lethal
hit (D37 or 37% survival), while only about one dimeric photoproduct was formed.
RNA bacteriophages, which are viruses that infect bacteria, typically consist of
a capsid composed principally of one protein, with small numbers of one or two
other proteins. The action spectra of several RNA phages has been studied by Rauth
(1965), who showed that the quantum yield for virus inactivation was approximately
the same as the quantum yield for RNA inactivation, which suggests that not only is
the RNA the primary target but that the protein coats of the phages studied contribute
little protective effect. Under UV irradiation, a lethal hit for mengovirus (at 70 J/m2)
produced 1.7 uracil dimers but no apparent structural damage to the RNA (Miller
and Plageman 1974).

2.11 Photochemistry of Bacteria 39
Most RNA virus coats consist of one major protein and sometimes one or two
other proteins. There are five known viral proteins, M, G, L, N, and NS. The pro-
tein of the common strain of TMV consists of three tryptophan and four tyrosine
residues, both of which have high molar absorptivities (see Fig. 2.11). The coat
protein itself may suffer UV photodamage and may become cross-linked to RNA,
but the extent to which this contributes to overall inactivation may be of limited
significance (Fraenkel-Conrat and Wagner 1981).
Double stranded RNA viruses tend to be much more resistant to UV exposure
than single stranded RNA viruses, by almost an order of magnitude. Zavadova
(1971) showed that the D90 for double stranded encephalomyocarditis virus RNA
was about six times that for the single stranded version.
2.10 Photochemistry of DNA Viruses
UV irradiation of DNA produces photoproducts called pyrimidine dimers as
well as non-dimer photoproducts. Dimers produced in DNA can consist of
thymine:thymine, thymine:cytosine, and cytosine:cytosine. Thymine dimers, the
most common photoproducts, were the focus of much early investigation in UV
irradiation experiments. The number of thymine dimers produced per lethal hit in
the DNA of phage φX174 is about 0.3 (David 1964). For coliphage lambda, about
two dimers were produced per lethal hit (Radman et al. 1970). For phage T4, 10.2
dimers were formed per lethal hit (Meistrich 1972). For a given dose, more dimers
are produced when AT-rich DNA is irradiated than GC-rich DNA, but this difference
is not more than twofold. The relative efficiency of dimer formation in DNA is in
the order TT > CT > CC (Setlow and Carrier 1966). However, this can vary with the
thymine content since, for viruses with high GC content, the number of CC dimers
produced under UV exposure can exceed the number of TT dimers (Matallana-
Surget et al. 2008). Photoproducts other than dimers are also produced, including
pyrimidine adducts, which occur at about a tenth of the frequency of dimers, and
others which occur at even lower frequencies (Wang 1976). Cytidine-derived pho-
toproducts include cytidine hydrate (or the deamination product, uridine hydrate)
and cytosine dimer (deamination product, uracil dimer).
2.11 Photochemistry of Bacteria
The kinetics of bacterial inactivation by ultraviolet light are much the same as in
DNA viruses since they contain DNA, except that many bacteria have more pho-
toprotection and often have the ability to photorecover or photoreactivate. About
65 dimers are produced per every 107nucleotides in the DNA of E. coli, for every
J/m2of UV irradiation (Fraenkel-Conrat and Wagner 1981). The number of dimers
formed varies from one species to another but the ultraviolet sensitivities of bac-
teria with varying GC content are not directly proportional to the TT frequency,

40 2 UVGI Disinfection Theory
indicating that thymine dimers are not the sole cause of lethality. David (1973)
inferred that for a constant G+C content (or T+A content), the sensitivity to UV
radiation is a reciprocal function of the molecular weight of the genome, suggesting
that the smaller the DNA molecule, the higher the probability that a hit would be
lethal.
Bacteria invariably contain enzymes and other repair mechanisms that may allow
for photoreactivation and photorecovery from UV exposure (Atlas 1995). The quan-
tum yield for inactivation of an enzyme is approximately proportional to its cystine
content and is roughly inversely proportional to its molecular weight (Smith and
Hanawalt 1969). The latter is explained by the fact that the cystine content of pro-
teins is inversely proportional to their molecular weight. The action spectrum for an
enzyme can be resolved into the contributions from its constituent chromophores.
2.12 Photoreactivation
Photoreactivation is a natural process in which bacterial cells can partially recover
from ultraviolet damage when visible and UV wavelengths of light reverse DNA
damage by monomerizing cyclobutane pyrimidine dimers. It was first identified
in E. coli by Prat (1936) and later demonstrated by Kelner (1949), and has since
been noted to occur in many other bacteria. Photoreactivation is an effect that pri-
marily operates on bacteria and spores. Viruses and certain bacteria seem to have
very limited capability to self-repair or photoreactivate, including Haemophilus
influenzae, Diplococcus pneumoniae, Bacillus subtilis, and Deinococcus radiodu-
rans (Masschelein 2002). David et al. (1971) reports photoreactivation rates of
40–56% in mycobacteria. Little evidence exists for the photoreactivation of animal
viruses since they lack enzymes although they may be photoreactivated by host-cell
repair mechanisms (Samad et al. 1987). Photoreactivation has never been observed
in animal virus RNA (Bishop et al. 1967). The photoreactivation effect may be
dependent on RH, with the effect possibly absent when RH is less than approxi-
mately 65%. Evidence suggests that the conformation change in DNA that occurs at
higher RH may allow microbes to experience photoreactivation (Rahn and Hosszu
1969, Munakata and Rupert 1974).
Many bacterial cells possess repair enzymes that can repair gaps and defects
in the DNA. Thymine dimers formed by UV irradiation of DNA are hydrolyzed
by specific DNases and are replaced with correct sequences by repair enzymes
(Guschlbauer 1976). The maximum yield of thymine dimers in irradiated bacterial
cells depends on the wavelength as well as the conditions (i.e. RH%). After a suffi-
cient UV dose has been imparted to the bacteria a steady state is reached in which
the relative numbers of dimers do not change (Smith and Hanawalt 1969). Dimer
formation is a reversible process and thymine dimers may revert to free thymines via
the absorption of UV and visible light. Photoreactivation cannot completely reverse
damage to DNA since UV may cause other types of photoproducts but it can effec-
tively limit UV damage.

2.12 Photoreactivation 41
Fig. 2.14 After inactivation by UV irradiance, exposure to visible light for 2–3 h may produce
photoreactivation, in which many broken thymine links (thymine dimers) are repaired by enzymes
Thymine dimers absorb light in the visible range (blue light) and this leads to
self-repair of the nucleotide bonds, as illustrated in Fig. 2.14 . Reactivation can also
occur under conditions of no visible light, or what is called dark repair. The ability
to self-repair can depend on the biological organization of the microorganism, as
well as the amount of UV damage inflicted on the cell.
Photoreactivation can be catalyzed by enzymes, which are commonly present in
bacterial cells. The process occurs in two stages; the first involves the production
of an enzyme-substrate complex at the DNA lesion site in the absence of light,
and the second is a photolytic reaction in which light energy is absorbed and the
lesion is repaired (Fletcher et al. 2003). Enzymatic photoreactivation is facilitated
by visible light and results in the splitting of pyrimidine dimers, called monomer-
ization. Thymine dimers are more efficiently eliminated than other types. Only
polynucleotide strands containing adjacent pyrimidines are photoreactivable and
a minimum length of about nine bases appears to be necessary for the enzyme to
attach and excise dimers.
In photoreactivation, repair is due to an enzyme called photolyase. Photolyase
reverses UV-induced damage in DNA. In dark repair the damage is reversed by the
action of a number of different enzymes. All of these enzymes must initially be acti-
vated by an energy source, which may be visible light (300–500 nm) or nutrients that
exist within the cell. Masschelein (2002) suggests that the enzymatic repair mech-
anism requires at least two enzyme systems: an exonuclease systems (i.e. to dis-
rupt the thymine-thymine linkage), and a polymerase system to reinsert the thymine
bases on the adenosine sites of the complementary strain of DNA.
In DNA repair mechanisms, the damaged strand is excised by the enzyme and
then the complementary strand of DNA is used as a template for inserting the cor-
rect nucleotides. Failure to repair UV damaged DNA can result in errors during the
replication process during which base substitutions can occur and result in the devel-
opment of mutants. The most favored sites for base substitutions leading to mutants
involve transitions from GC to AT at sites with adjacent pyrimidines (Miller 1985).

42 2 UVGI Disinfection Theory
Enzymes can be damaged by broadband UV wavelengths other than 253.7 nm, and
it has been reported that the use of medium pressure UV lamps inhibits photoreac-
tivation due to the fact that broadband wavelengths inflict damage on photorepair
enzymes (Kalisvaart 2004, Quek and Hu 2008).
No types of DNA damage other than that produced by UV can be photoreacti-
vated. In cell systems with efficient dark repair mechanisms, like D. radiodurans,
little or no photoreactivation occurs. UV damage produced at 253.7 nm can be atten-
uated by exposure to wavelengths between 330 and 480 nm (Hollaender 1955). The
enzymatic monomerization of pyrimidine dimers operates when pyrimidine dimers
are the primary type of photodamage, and when photodamage is due to other pho-
toproducts, like the spore photoproduct, photoreactivation appears to be absent.
2.13 UV Scattering
Another kind of photoprotection, other than shielding or photoreactivation, occurs
when light is scattered from microbes. Scattering of UV light from microbes is a
phenomenon that is routinely observed during the measurement of optical density
of microbes in solution to obtain absorption spectra, and during which corrections
for scattering must often be made (Holler et al. 1998). Luria et al. (1951) used
corrections of 10–20% for scattering at 260 nm while Zelle and Hollaender (1954)
found the absorbance corrections for phages T2 and T7 were somewhat greater than
20%. In studies on phage T2 (0.065–0.095 μm), Dulbecco (1950) found that for
wavelengths longer than 320 nm the absorption closely followed Rayleigh’s law
of scattering, and that the photosensitive pigments were part of the phage (but not
the DNA) and tended to darken after exposure. Rauth (1965) found that for small
viruses like MS2 and φX174, the corrections for scattering are almost negligible and
only become appreciable above about 280 nm, where they can approach 20–25%
depending on virus size. Powell (1959) found that UV scattering effects accounted
for no more than 25% attenuation in water. According to Jagger (1967) the UV
transmission through an E. coli cell is only 70% at 254 nm, leaving a maximum of
30% to be scattered or absorbed.
Scattering may cause appreciable loss of light when the exposed microbes have
dimensions comparable with UV wavelengths (Hollaender 1955). The scattering
effect is reduced as the index of refraction of the microbe approaches the index of
refraction of the medium (i.e. air or water). The scattering effect increases, however,
when the size parameter (a function of the diameter) approaches the wavelength of
the ultraviolet light (van de Hulst 1957).
Mie scattering is the dominant form of light scattering in the micron-size range
of viruses and small airborne bacteria (Bohren and Huffman 1983). Scattering can
have significant impact on the amount of UV that actually reaches the nucleocapsid
or DNA of a microbe in air and the effect appears to become significant at diameters
of about 0.03 microns and greater. Scattering is a protective effect and not dependent
on the protein content of nucleocapsids or cell walls, since most microbes appear to
have similar indices of refraction (i.e. about 1.05–1.08).

2.13 UV Scattering 43
Absorption of photons takes place as ultraviolet radiation penetrates a particle.
Light that is not absorbed may be scattered from a particle in the virus and bacteria
size range (0.02–20 μm) by three different mechanisms: (1) reflection of photons
from the particle, (2) refraction of photons that pass unabsorbed through the parti-
cle, and (3) diffraction of photons that pass through or near a particle. Diffraction
may alter the path of photons even though they are not in the direct path of the
particle. This latter phenomena can result in a particle scattering more light than it
would actually intercept due to its physical size alone (Modest 1993). The interac-
tion between ultraviolet wavelengths and the particle is a function of the relative
size of the particle compared with the wavelength, as defined by the size parameter:
x=2πa
λ(2.10)
where
a=the effective radius of the particle
λ=wavelength
If the size parameter, x << 1, then Rayleigh scattering dominates and for simple
spherical particles of diameters less than λ/10 the scattering will approximately vary
with the inverse of the wavelength raised to the fourth power (1/λ4). If x>>1, the
principles of normal geometric optics may be applied. If x≈1, Mie scattering
dominates, and this is the case for small viruses and bacteria. For Mie scattering in
air, the size parameter can be written as follows (Chen et al. 2003):
x=πdnm
λ(2.11)
where
nm=refractive index of the medium (air)
d=particle diameter (typically nanometers)
For nonspherical microbes where the length is significantly greater than the diam-
eter (i.e. aspect ratio > 5), the size parameter for rods may be used. In such cases
the length is merely substituted for the diameter (from Stacey 1956). The scatter-
ing of light is due to differences in the refractive indices between the medium and
the particle (Modest 1993, Garcia-Lopez et al. 2006). The scattering properties of a
spherical particle in any medium are defined by the complex index of refraction:
m=n−iκ(2.12)
where
n=real refractive index
κ=imaginary refractive index (absorptive index or absorption coefficient)

44 2 UVGI Disinfection Theory
Since the refractive index of ultraviolet light approaches 1 in air (or about
1.00029 for visible light), Eqs. (2.10) and (2.11) are virtually identical. If scattering
is not affected by the presence of other surrounding particles, and this is generally
the case for airborne microbes since concentrations will never be so high as to even
be visible, the process is known as independent scattering. The process of indepen-
dent Mie scattering is also governed by the relative refractive index, defined with
the same symbol (m) as follows:
m=ns
nm(2.13)
where ns=refractive index of the particle (microbe)
Water has a refractive index of approximately nm=1.4 in the ultraviolet range
and about 1.33 in the visible range. The refractive index of microbes in visible light
has been studied by several researchers. Balch et al. (2000) found the median refrac-
tive index of four viruses to be 1.06, with a range of 1.03–1.26. Stramski and Keifer
(1991) assumed viruses to have a refractive index of 1.05. Biological cells were
assumed by Mullaney and Dean (1970) to have relative refractive indices of about
1.05 in visible light. Klenin (1965) found S. aureus to have a refractive index in
the range 1.05–1.12. Petukhov (1964) gives the refractive index of certain bacteria
in the limits of 1.37–1.4. There are no studies that address the real refractive index
of bacteria or viruses at UV wavelengths except Hoyle and Wickramasinghe (1983)
who suggest ns=1.43 as a reasonable choice for coliform bacteria. Garcia-Lopez et
al. (2006) state that for soft-bodied biological particles n is between 1.04 and 1.45.
For the imaginary refractive index (the absorptive index or absorption coefficient) in
the UV range no information is currently available. Per Garcia-Lopez et al. (2006),
hemoglobin has a κof 0.01–0.15, while polystyrene has a κof 0.01–0.82.
The mathematical solution of Mie scattering is so complicated as to generally
require the use of advanced computational methods. For details of these solutions
see van de Hulst (1957), Bohren and Huffman (1983), and Modest (1993). A vari-
ety of software packages are freely available for solving the scattering problem for
small particles in air, such as DDSCAT, and tables have also been published for use
(Draine and Flatau 2004).
Per Eq. (2.13) the refractive index for microbes in air would be about
(1.05)(1.33) =1.4. Figure 2.15 shows two examples of scattering effects in microbes
of 0.2 μm (small virus) and 1 μm (large virus or small bacteria) when light is inci-
dent from the left passing to the right. The scattering was evaluated for a 253 nm
wavelength, a real refractive index of 1.4, an imaginary refractive index of –1.4, a
medium refractive index of 1.0003 (air), and wide dispersion of particles (negligible
concentration). Computations were performed using the Mie Scattering Calculator
(Prahl 2009).
The amount of scattering and absorption by a particle is defined by the scattering
cross-section, Csca, and the absorption cross-section, Cabs. The scattering cross sec-
tion is defined as the area which when multiplied by the incident irradiance gives
the total power scattered by the particle. The absorption cross section is the area
which when multiplied by the incident irradiance gives the total power absorbed.

2.13 UV Scattering 45
Fig. 2.15 Angular UV light scattering functions for spherical microbes in air, with diameters as
indicated. Plots reprinted courtesy of Scott Prahl, Oregon Medical Laser Center
The total amount of absorption and scattering is the extinction cross-section, Cext,
defined as the area which when multiplied by the incident irradiance gives the total
power removed from the incident wave by scattering and absorption.
Cext =Cabs +Csca (2.14)
The fraction of UV that is scattered from the total incident irradiation, Suv, can
be computed as follows (Kowalski et al. 2009):
Suv =Csca
Cabs +Csca =Csca
Cext (2.15)
Equation (2.15) effectively defines the correction factor (as a complement) for
UV incident on a particle that scatters UV. Efficiency factors used in Mie scattering
are the cross-sections divided by the area, and include the absorption efficiency
factor, Qabs, the scattering efficiency factor, Qsca, and the extinction efficiency factor,
Qext, defined as follows:
Qabs =Cabs
πa2(2.16)
Qsca =Csca
πa2(2.17)
Qext =Cext
πa2(2.18)
The extinction efficiency factor is equal to the sum of the other two factors:
Qext =Qabs +Qsca (2.19)

46 2 UVGI Disinfection Theory
0
0.5
1
1.5
2
2.5
10
1
0.10.01
Particle Diameter, µm
Efficiency or Fraction
Qabs
Qsca
Viruses Bacteria
RNA DNA
UV Scatter Fraction, Suv
Fig. 2.16 UV Scattering Efficiency, Absorption Efficiency, and Scatter Fraction for spherical par-
ticles in air. Real refractive index =1.4, imaginary absorptive index =–1.4, nm=1.0003. Dots
show approximate logmean diameters of RNA and DNA viruses
The degree of scattering increases from small viruses up to small bacteria.
Figure 2.16 shows an example of the scattering efficiency, absorption efficiency, and
scatter fraction of spherical viruses and bacteria in the 0.02 – 8 μm size range in air,
based on computations performed using Mie scattering software (Prahl 2009). The
refractive index (real component) used for this example is n=1.4, while the absorp-
tive index (the imaginary component) is assumed the same as water, κ=1.4. Note
that the scattering efficiency increases dramatically above about 0.06 μm. The range
of sizes for viruses and bacteria are also shown, and it can be seen that DNA viruses
and bacteria would be most impacted by scattering effects. This graph is revisited
for water in Chap. 4, where it will be seen that although the medium changes the
efficiency factors, it has little effect on the scattering fraction.
The refractive index and the absorptive index of individual species of microbes
may be somewhat different from the values assumed above, but testing of alternate
values indicates that the general pattern of behavior observed in Fig. 2.16 remains
essentially unchanged for all possible microbial values previously cited. The UV
scatter fraction confers some significant UV photoprotection, especially to larger
viruses and to bacteria. Such photoprotection translates into an effective UV dose
lower than that to which the microbe is exposed. The subject of UV scattering as
a mechanism of photoprotection, and how it relates to predicting UV susceptibility
will be revisited in Chap. 4.

References 47
References
Anderson J, Rowan N, Macgregor S, Fouracre R, Farish O. 2000. Inactivation of food borne
enteropathogenic bacteria and spoilage fungi using pulsed-light. IEEE Trans Plasma Sci
28:83–88.
Atlas RM. 1995. Microorganisms in Our World. St. Louis: Mosby.
Balch WM, Vaughn J, Novotny J, Drapeau D, Vaillancourt R, Lapierre J, Ashe A. 2000. Light
scattering by viral suspensions. Limnol Oceanogr 45(2):492–498.
Becker MM, Wang Z. 1989. Origin of ultraviolet damage in DNA. J Mol Biol 210:429–438.
Bell A, Wheeler M. 1986. Biosynthesis and function of melanin. Ann Rev Phytopathology
24:411–451.
Bhattacharjee C, Sharan RN. 2005. UV-C radiation induced conformational relaxation of pMTa4
DNA in Escherichia coli may be the cause of single strand breaks. Int J Radiat Biol
81(12):919–927.
Bishop JM, Quintrell N, Koch G. 1967. Poliovirus double-stranded RNA: Inactivation by ultravio-
let light. J Mol Biol 24:125–128.
Bohren C, Huffman D. 1983. Absorption and Scattering of Light by Small Particles. New York:
Wiley & Sons.
Bolton JR, Cotton CA. 2008. Ultraviolet Disinfection Handbook.American Water Works
Association. Denver, CO:
Boyd-Wilson K, Perry J, Walter M. 1998. Persistence and survival of saprophytic fungi antagonis-
tic to Botrytis cinerea on kiwifruit leaves. Proc of the 51st Conf of the New Zealand Plant Prot
Soc, Inc.:96–101.
Casarett AP. 1968. Radiation Biology. Englewood: Prentice-Hall.
Chen K, Kromin A, Ulmer M, Wessels B, Backman V. 2003. Nanoparticle sizing with a resolution
beyond the diffraction limit using UV light scattering spectroscopy. Optics Commun 228:1–7.
Cornelis JJ, Su ZZ, Ward DC, Rommelaere J. 1981. Indirect induction of mutagenesis of intact
parvovirus H-1 in mammalian cells treated with UV light or with UV-irradiated H-1 or simian
virus 40. Proc Natl Acad Sci 78(7):4480–4484.
David, CN. 1964. UV inactivation and thymine dimerization in bacteriophage phiX. Z. Verber-
bungsl. 95:318–325.
David HL, Jones WD, Newman CM. 1971 Ultraviolet light inactivation and photoreactivation in
the mycobacteria. Infect and Immun 4:318–319.
David HL. 1973. Response of mycobacteria to ultraviolet radiation. Am Rev Resp Dis
108:1175–1184.
Draine B, Flatau P. 2004. User guide for the discrete dipole approximation code DDSCAT 6.1.
Dulbecco R. 1950. Experiments on photoreaction of bacteriophages inactivated with ultraviolet
radiation. J Bact 59:329–347.
Durrell L, Shields L. 1960. Fungi isolated in culture from soil of the Nevada test site. Mycologia
52:636–641.
Edenberg HJ. 1983. Inhibition of simian virus 40 DNA replication by ultraviolet light. Virology
128(2):298–309.
Elasri MO, Miller RV. 1999. Response of a biofilm bacterial community to UV Radiation. Appl
Environ Microbiol 65(5):2025–2031.
Eyster JM, Prohofsky EW. 1977. On the B to A Conformation Change of the Double Helix.
Biopolymers 16:965–982.
Fernandez RO, Pizarro RA. 1996. Lethal effect induced in Pseudomonas aeruginosa exposed to
ultraviolet-A radiation. Photochem Photobiol 64(2):334–339.
Fletcher LA, Noakes CJ, Beggs CB, Sleigh PA, Kerr KG. The ultraviolet susceptibility of
aerosolised microorganisms and the role of photoreactivation; 2003; Vienna. IUVA.
Fraenkel-Conrat H, Wagner RR. 1981. Comprehensive Virology. New York: Plenum Press.
Furuse K, Watanabe I. 1971. Effects of ultraviolet light (UV) irradiation on RNA phage in H2O
and in D2O. Virol 46:171–172.

48 2 UVGI Disinfection Theory
Garcia-Lopez A, Snider A, Garcia-Rubio L. 2006. Rayleigh-Debye-Gans as a model for continuous
monitoring of biological particles: Part I, assessment of theoretical limits and approximations.
Optics Express 14(19):17.
Guschlbauer W. 1976. Nucleic Acid Structure. New York: Springer-Verlag.
Hanawalt PC, Friedberg EC, Fox CF, editors. 1978. DNA Repair Mechanisms. New York:
Academic Press.
Harm W. 1980. Biological Effects of Ultraviolet Radiation. New York: Cambridge University
Press.
Hollaender A. 1955. Radiation Biology, Volume II: Ultraviolet and Related Radiations. New York:
McGraw-Hill.
Holler S, Pan Y, Bottiger JR, Hill SC, Hillis DB, Chang RK. Two-dimensional angular scattering
measurements of single airborne micro-particles; 1998; Boston, MA. SPIE. pp 64–72.
Hoyle F, Wickramasinghe C. 1983. The ultraviolet absorbance spectrum of coliform bacteria and
its relationship to astronomy. Astrophysics and Space Science 95:227–231.
Hu JY, Chu SN, Quek PH, Feng YY, Tang XL. 2005. Repair and regrowth of Escherichia coli after
low- and medium-pressure ultraviolet disinfection. Wat Sci Technol 5(5):101–108.
IESNA. 2000. Lighting Handbook: Reference & Application IESNA HB-9-2000. New York: Illu-
mination Engineering Society of North America.
Jagger J. 1967. Ultraviolet Photobiology. Englewood Cliffs: Prentice-Hall, Inc.
Kalisvaart BF. 2004. Re-use of wastewater: Preventing the recovery of pathogens by using
medium-pressure UV lamp technology. Wat Sci Technol 33(10):261–269.
Kelner A. 1949. Effect of visible light on the recovery of Streptomyces griseus conidia from ultra-
violet irradiation injury. Proc Nat Acad Sci 35(2):73–79.
Klein B, Filon AR, vanZeeland AA, vanderEb AJ. 1994. Survival of UV-irradiated vaccinia virus
in normal and xeroderma pigmentosum fibroblasts; evidence for repair of UV-damaged viral
DNA. Mutat Res 307(1):25–32.
Klenin V. 1965. The problem concerning the scattering of light by suspensions of bacteria.
Biofizika 10(2):387–388.
Koller LR. 1952. Ultraviolet Radiation. New York: John Wiley & Sons.
Kowalski W, Bahnfleth W, Hernandez M. 2009. A Genomic Model for the Prediction of Ultra-
violet Inactivation Rate Constants for RNA and DNA Viruses; 2009 May 4–5; Boston, MA.
International Ultraviolet Association.
Kuluncsics Z, Perdiz D, Brulay E, Muel B, Sage E. 1999. Wavelength dependence of ultraviolet-
induced DNA damage distribution: Involvement of direct or indirect mechanisms and possible
artefacts. J Photochem Photobiol 49(1):71–80.
Linden KG. 2001. Comparative effects of UV wavelengths for the inactivation of Cryptosporidium
parvum oocysts in water. Wat Sci Technol 34(12):171–174.
Luckiesh M. 1946. Applications of Germicidal, Erythemal and Infrared Energy. New York: D. Van
Nostrand Co.
Luria SE, Williams RC, Backus RC. 1951. Electron micrographic counts bacteriophage particles.
J Bact 61:179–188.
Masschelein WJ. 2002. Ultraviolet Light in Water and Wastewater Sanitation. Boca Raton: Lewis
Publishers.
Matallana-Surget S, Meador J, Joux F, Douki T. 2008. Effect of the GC content of DNA
on the distribution of UVB-induced bipyrimidine photoproducts. Photochem Photobiol Sci
7:794–801.
Meistrich ML. 1972. Contribution of thymine dimers to the ultraviolet light inactivation of mutants
of bacteriophage T4. J Mol Biol 66:97.
Miller RL, Plageman PGW. 1974. Effect of Ultraviolet Light on Mengovirus: Formation of Uracil
Dimers, Instability and Degradation of Capsid, and Covalent Linkage of Protein to Viral RNA.
J Virol 13(3):729–739.
Miller J. 1985. Mutagenic specificity of ultraviolet light. J Mol Biol 182:45–65.
Modest MF. 1993. Radiative Heat Transfer. New York: McGraw-Hill.

References 49
Mullaney P, Dean P. 1970. The small angle light scattering fo biological cells. Biophys J 10:
764–772.
Munakata N, Rupert CS. 1974. Dark Repair of DNA Containing “Spore Photoproduct” in Bacillus
subtilis. Molec Gen Genet 130:239–250.
Neidle S. 1999. Oxford Handbook of Nucleic Acid Structure. New York: Oxford University Press.
Palmeira L, Guegen L, Lobry J. 2006. UV-targeted dinucleotides are not depleted in light-exposed
prokaryotic genomes. Mol Biol Evol 23:2214–2219.
Peak MJ, Peak JG. 1978. Action spectra for the ultraviolet and visible light inactivation of phage
T7: Effect of host-cell reactivation. Radiat Res 76:325–330.
Peccia J, Werth HM, Miller S, Hernandez M. 2001. Effects of relative humidity on the ultraviolet
induced inactivation of airborne bacteria. Aerosol Sci Technol 35:728–740.
Petukhov V. 1964. The feasibility of using the Mie theory for the scattering of light from suspen-
sions of spherical bacteria. Biofizika 10(6):993–999.
Powell WF. 1959. Radiosensitivity as an index of herpes simplex virus development. Virology
9(1–19).
Prahl S. 2009. Mie Scattering Calculator. Portland, OR: Oregon Medical Laser Center.
Prat S. 1936. Strahlung und antagonistische wirkungen. Protoplasma 26:113–149.
Quek PH, Hu J. 2008. Indicators for photoreactivation and dark repair studies following ultraviolet
disinfection. J Ind Microbiol Biotechnol 35(6): 533–541.
Radman M, Cordone L, Krsmanovic-Simic D, Errera M. 1970. Complementary action of recombi-
nation and excision in the repair of ultraviolet irradiation damage to DNA. J Mol Biol 49:203.
Rahn RO, Hosszu JL. 1969. Influence of relative humidity on the photochemistry of DNA films.
Biochim Biophys Acta 190:126–131.
Rainbow AJ, Mak S. 1973. DNA damage and biological function of human adenovirus after U.V.
irradiation. Int J Radiat Biol 24(1):59–72.
Rauth AM. 1965. The physical state of viral nucleic acid and the sensitivity of viruses to ultraviolet
light. Biophys J 5:257–273.
Remsen JF, Miller N, Cerutti PA. 1970. Photohydration of uridine in the RNA of coliphage R17. II.
The relationship between ultraviolet inactivation and uridine photohydration. Proc Natl Acad
Sci USA 65:460–466.
Rothman R, Setlow R. 1979. An action spectrum for cell killing and pyrimidine dimer formation
in Chinese hamster V-79 cells. Photochem Photobiol 29:57–61.
Ryan DKG, Rainbow AJ. 1977. Comparative studies of host-cell reactivation, cellular capacity
and enhanced reactivation of herpes simplex virus in normal, xeroderma pigmentosum and
Cockayne syndrome fibroblasts. Photochem Photobiol 26:263–268.
Samad SA, Bhattacharya SC, Chatterjee SN. 1987. Ultraviolet inactivation and photoreactivation
of the cholera phage ‘Kappa’. Radiat Environ Biophys 26:295–300.
Sarasin AR, Hanawalt PC. 1978. Carcinogens enhance survival of UV-irradiated simian virus 40
in treated monkey kidney cells: Induction of a recovery pathway? Proc Natl Acad Sci 75(1):
356–350.
Sarasin AR, Hanawalt PC. 1980. Replication of ultraviolet-irradiated Simian Virus 40 in monkey
kidney cells. J Mol Biol 138:299–319.
Schreier WJ, Schrader TE, Koller FO, Gilch P, Crespo-Hernandez CE, Swaminathan VN, Carell T,
Zinth W, Kohler B. 2007. Thymine dimerization in DNA is an ultrafast photoreaction. Science
315:625–629.
Setlow RB, Pollard EC. 1962. Molecular Biophysics. Reading, MA: Addison-Wesley.
Setlow RB, Carrier WL. 1966. Pyrimidine dimers in ultraviolet-irradiated DNA’s. J Mol Biol
17:237–254.
Smirnov Y, Kapitulez S, Kaverin N. 1992. Effects of UV-irradiation upon Venezuelan equine
encephalomyelitis virus. Virus Res 22(2):151–158.
Smith KC, Hanawalt PC. 1969. Molecular Photobiology: Inactivation and Recovery. New York:
Academic Press.

50 2 UVGI Disinfection Theory
Stacey K. 1956. Light-Scattering in Physical Chemistry. London: Butterworths Scientific Publica-
tions.
Stacks PC, White JH, Dixon K. 1983. Accomodation of Pyrimidine Dimers During Replication of
UV-Damaged Simian Virus 40 DNA. Molec Cell Biol 3(8):1403–1411.
Stramski D, Keifer D. 1991. Light scattering by microorganisms in the open ocean. Prog Oceanogr
28:343–383.
Unrau P, Wheatcroft R, Cox B, Olive T. 1973. The formation of pyrimidine dimers in the DNA of
fungi and bacteria. Biochim Biophys Acta 312:626–632.
Valero A, Begum M, Leong S, Hocking A, Ramos A, Sanchis V, Marin S. 2007. Fungi isolated
from grapes and raisins as affected by germicidal UVC light. Lett Appl Microbiol 45:238–243.
Van de Hulst H. 1957. Light Scattering by Small Particles. New York: Chapman & Hall, Ltd.
Von Sonntag C. 1986. Disinfection by free radicals and UV-radiation. Water Supply 4:11–18.
Waites WM, Harding SE, Fowler DR, Jones SH, Shaw D, Martin M. 1988. The destruction of
spores of Bacillus subtilis by the combined effects of hydrogen peroxide and ultraviolet light.
Lett Appl Microbiol 7:139–140.
Wang SY. 1976. Photochemistry and Photobiology of Nucleic Acids. New York: Academic Press.
Webb SJ. 1965. Bound Water in Biological Integrity. Springfield, IL: Charles C. Thomas.
Webb RB, Tuveson TW. 1982. Differential sensitivity to inactivation of NUR and NUR∗strains of
Escherichia coli at six selected wavelengths in the UVA, UVB, and UVC ranges. Photochem
Photobiol 36:525–530.
Zavadova Z, Gresland L, Rosenbergova M. 1968. Inactivation of single- and double-stranded
ribonucleic acid of encephalomyocarditis virus by ultraviolet light. Acta Virol 12:515–522.
Zavadova Z. 1971. Host-cell repair of vaccinia virus and of double stranded RNA of encephalomy-
ocarditis virus. Nature (London) New Biol 233:123.
Zelle MR, Hollaender A. 1954. Monochromatic ultraviolet action spectra and quantum yields for
inactivation of T1 and T2 Escherichia coli bacteriophages. J Bact 68:210–215.
Zimmer JL, Slawson RM. 2002. Potential repair of Escherichia coli following exposure to UV
radiation from both medium- and low-pressure UV sources used in drinking water treatment.
Appl Environ Microbiol 68(7):3293–3299.

http://www.springer.com/978-3-642-01998-2