You are on page 1of 12

Received: 15 July 2019 | Revised: 2 November 2019 | Accepted: 4 November 2019

DOI: 10.1111/ina.12619

ORIGINAL ARTICLE

Comparison of disinfection performance of UVC-LED and


conventional upper-room UVGI systems

Sunday S. Nunayon | Huihui Zhang | Alvin C. K. Lai

Department of Architecture and Civil


Engineering, City University of Hong Kong, Abstract
Hong Kong, China We developed a novel, compact upper-room ultraviolet germicidal irradiation sys-
Correspondence tem with light-emitting diode sources (UR-UVGI-LED) to enhance the disinfection
Alvin C. K. Lai, Department of Architecture of bioaerosols in an enclosed room space. Its effectiveness was evaluated and com-
and Civil Engineering, City University of
Hong Kong, Tat Chee Avenue, Kowloon, pared with the conventional upper-room ultraviolet germicidal irradiation system
Hong Kong, China. with mercury vapor sources (UR-UVGI-MV). Escherichia coli, Serratia marcescens, and
Email: alvinlai@cityu.edu.hk
Staphylococcus epidermidis were atomized under the well-mixed condition and ex-
Funding information posed to UR-UVGI-LED (or UR-UVGI-MV) device. The intensity output of the UR-
Research Grants Council of the Hong Kong
Special Administrative Region of China, UVGI-LED was also varied from 0% (no LED), 25%, 50% to 100% to further evaluate
Grant/Award Number: CityU 11273116 and the UR-UVGI-LED disinfection effectiveness under different power levels. The decay
CityU 11204217.
rates for UR-UVGI-LED ranged from −0.1420 ± 0.04 min−1 to −0.3331 ± 0.07 min−1
for Escherichia coli, −0.1288 ± 0.01 min−1 to −0.3583 ± 0.02 min−1 for Serratia marc-
escens, and −0.0330 ± 0.01 min−1 to −0.0487 ± 0.01 min−1 for Staphylococcus epi-
dermidis. It was noticed that the intensity level had a non-linear influence on the
UR-UVGI-LED’s performance. The decay rates achieved by the UR-UVGI-MV sys-
tem were −0.3867 ± 0.08 min−1, −0.4745 ± 0.002 min−1, and −0.1624 ± 0.02 min−1
for Escherichia coli, Serratia marcescens, and Staphylococcus epidermidis, respectively.
Hence, the disinfection performance of both UR-UVGI-LED and UR-UVGI-MV sys-
tems was comparable for Escherichia coli and Serratia marcescens. These results dem-
onstrate that the UR-UVGI-LED system has a high potential to be used as a safe and
effective irradiated light source to disinfect indoor airborne pathogens.

KEYWORDS

airborne pathogens, disinfection bioaerosols, IAQ, light-emitting diode, ultraviolet C, upper-


room UVGI

1 | I NTRO D U C TI O N the HVAC system can remove particles from both outdoor and in-
door sources. The minimum efficiency reporting value (MERV) is a
Airborne transmitted pathogenic infections can cause various dis- standard index, with ratings from 1 to 20, used to rate the effective-
eases and threaten human health and lives severely. Indoor path- ness of air filters for particles in the range of ≤0.3 µm to ≥ 10 µm
ogens can be transmitted through the heating, ventilating and air in size.3 Medium-grade filters—MERV ratings from 5 to 9,4,5 are
conditioning (HVAC) systems.1,2 To reduce infection risk, engineer- not effective for removing small bacteria and viruses.6 For many
ing control strategies should be implemented. A classical solution is general premises, it is not economically feasible to apply high ven-
filtration or dilution in which clean ventilation air is introduced into tilation rates or use high-efficiency particulate air (HEPA) filters.1
the room to replace any contaminated air. Filtration techniques in Considering the adverse implication of exposure to indoor airborne

180 | © 2019 John Wiley & Sons A/S. wileyonlinelibrary.com/journal/ina Indoor Air. 2020;30:180–191.
Published by John Wiley & Sons Ltd
NUNAYON et al. | 181

pathogens on human health,7 there is a huge demand for efficient


indoor air disinfection technologies.8
Practical Implications
Air disinfection is an imperative part of air treatment processes to
remove pathogenic microorganisms from the air. Different technolo- • High power UVC-LED has been recently developed but
gies have been developed and applied for effective air disinfection. its disinfection efficiency for airborne pathogens has
Among them, ultraviolet (UV) radiation is one of the demonstrated not been thoughtfully tested and understood.
technologies. Ultraviolet germicidal irradiation (UVGI), particularly in • We described an innovative, safe and compact UR-
the ultraviolet C (UVC) wavelengths band (180-280 nm), has been UVGI-LED device and compared its disinfection efficacy
recognized as having strong germicidal effects.9 UVC irradiation is with the conventional UR-UVGI-MV system.
known to kill microorganisms by disrupting the bonds that hold the • The uniquely attractive aspect of the UR-UVGI-LED
microorganisms’ deoxyribonucleic acid (DNA) or ribonucleic acid system is the ability to vary its output intensity, which
(RNA) bases and strands together. 10 could not be realized with the traditional UR-UVGI-MV
Ultraviolet C irradiation can be applied in duct systems or lo- system.
cated indoors. As short-term overexposure to UVC irradiation • Both devices were challenged by three bacteria, and
can cause red skin (erythema) and painful eye inflammation (pho- comparable disinfection efficacies were found.
tokeratitis) in humans, precautions must be implemented if such • The UR-UVGI-LED has a high potential for the next gen-
systems are installed indoors.11 Several national and international eration indoor air disinfection device, and even applica-
groups have established human exposure limits (ELs) for UVGI ex- ble for very small environments.
posure. For example, the American Conference of Governmental
Industrial Hygienists (ACGIH) and the International Commission on
Non-Ionizing Radiation Protection (ICNIRP) have recommended a It has also been speculated that other transmitted diseases like
2
threshold limit value (TLV) of 6.0 mJ/cm for a person's exposure the common cold, 23 influenza, 24,25 SARS, 26 and smallpox can poten-
12,13
to UV dose within an 8-hour period per day. Therefore, care- tially be controlled by UVC irradiation.
ful and intelligent application of stipulated international guidelines However, there is no significant improvement in the design and
for human exposure to UVC is very important both in the design construction of the conventional UR-UVGI-MV system since it was
and installation of UVGI fixtures to ensure that UVGI exposure level introduced to the commercial market. Low or medium pressure mer-
of room occupants are below current safe exposure levels. This in- cury vapor lamps are often used to generate UVC irradiation for
cludes measurements of UVGI irradiance in the lower room, espe- UR-UVGI-MV installation. These mercury lamps have many disad-
cially at the eye level, to determine compliance with occupational vantages: the short lifespan of approximately 4000-10 000 hours,
14
safety and health guidelines. Also, a UVC system should be fixed requiring frequent replacement, and the component (mercury) is
to the upper portion of a wall or under the ceiling. Thus, indoor UVC also a major known environmental contaminant. 27 Moreover, the
irradiation is often referred to as “upper-room” UVGI, or UR-UVGI irradiation source is housed in a fixture with long and closely sep-
for short. The disinfection principle of upper-room systems is that arated louvers that horizontally collimate UV rays. The reflection
the airborne pathogens are elevated to reach the irradiated zone and at the louvers’ surfaces leads to energy loss in the current design
then become disinfected and it has been widely tested.15,16 Since its of the UR-UVGI-MV system, reflecting its energy inefficiency. It is
introduction in the early 20th century,9 the UR-UVGI system with reported that only 2% of its total electrical wattage is converted
mercury vapor source (UR-UVGI-MV) has dramatically curtailed the to the irradiation that exits at the louvers. 28 The more critical con-
morbidity and mortality of tuberculosis (TB) pandemic in healthcare cern is the health issue as lamps that do not filter ozone-producing
facilities.17 Reports from more recent studies have also lent cre- wavelengths are sometimes misapplied despite the availability of
dence to the effectiveness of UR-UVGI-MV against TB.18,19 Besides commercial mercury vapor lamps with low/no ozone. 29,30 Another
TB, researchers have extended their studies to cover some other shortcoming of the conventional UR-UVGI-MV system is the bulky
popular bacteria for full-sized room settings. In two operating rooms size. Many units may be required for a large room, which may affect
where ceiling-mounted UR-UVGI-MV lamps were operated, Goldner its esthetics.
et al20 measured approximately 50% reduction in the concentration As an emerging technology, the use of ultraviolet light-emit-
of culturable airborne bacteria. Macher et al21 evaluated the effec- ting diode (UVC-LEDs) to produce UV radiation has many advan-
tiveness of four-numbers 15 W UR-UVGI-MV fixtures on a mixture tages. It is a semiconductor p-n junction device that emits light
of bacteria in a hospital waiting room (90 m3) and observed an effi- in a narrow spectrum. UVC-LEDs require no warm-up time and
ciency ranging between 14% and 19%. In another related study in relatively little energy to operate and have a very long operational
which one 15 W wall-mounted UR-UVGI-MV fixture was installed lifespan, of up to 100 000 hours. 31 UVC-LEDs are also environ-
3 22
and operated in a 36 m room under 2 ACH, Miller and Macher mentally friendly; they do not contain toxic materials or emit pol-
found that the effectiveness of the UR-UVGI-MV system was about lutants indicating that their production, use, and disposal have
50% for Bacillus Subtilis spores and Micrococcus luteus, and approxi- no potential negative impact on the environment and on human
mately 100% for Escherichia coli. health.
182 | NUNAYON et al.

For sterilization or disinfection applications, high power rating an economically reasonable decision on the UV device to adopt for
UVC-LEDs are required, and they are only available commercially conducting the inactivation of bioaerosols is the energy efficiency
just a few years ago. Several studies have proven that UVC-LEDs of such a system. One of the commonly used indices for characteriz-
can effectively inactivate pathogen in water.32,33 Currently, one ing the energy efficiency of an UV disinfection device is the cost of
major disadvantage of UVC-LEDs is that, to disinfect the same vol- electrical energy consumption per year, which has been previously
ume of room air that the conventional mercury vapor lamp would used by Yang et al36 to estimate the electricity efficiency of in-duct
sterilize, a lot of UVC-LEDs are required due to their characteristic UVGI for bacteria disinfection. Thus, the same index for the cost
low wattages. Another key concern is that UVC-LEDs dissipate lots of electrical energy consumption per year can be applied for both
of wasted heat. Nonetheless, with the advancement of optical engi- conventional UR-UVGI-MV and UR-UVGI-LED disinfection devices
neering, it is anticipated that the thermal efficiency of UVC-LEDs will to determine their electrical energy efficiencies. However, studies
be improved substantially in the coming few years. For example, as considering both disinfection performance and energy efficiency of
at 3 years ago, the highest wattage of commercially available UVC- conventional UR-UVGI-MV and UR-UVGI-LED systems have barely
LEDs was less than 10 mW, but now, it has been increased to 50 mW. been reported in the literature.
To the best of our knowledge, this study is the first to consider UR- In this study, we designed and compared the disinfection ef-
UVGI-LEDs for the disinfection of room air. Seeing the considerable ficacy of a novel UR-UVGI-LED system with the conventional
progress achieved and projected trends expected on the develop- UR-UVGI-MV system. The performance of the two systems was
ment of UVC-LEDs, this new UV source is believed to be a promising quantified by log-reduction and single-stage decay rate. Different
alternative for air disinfection. input current levels to the developed UR-UVGI-LED system were
The inactivation efficacy of UV against airborne bacteria differs, investigated. Also, the energy consumption of both disinfection de-
which may technically depend on the UV fluence required for de- vices during UV irradiation was analyzed. The results from this study
creasing the number of bacteria by an order of logarithmic magni- may offer an important understanding of the control of indoor bio-
tude. The term fluence is calculated as the product of UV irradiance aerosols by using UV irradiation and help abate the environmental
(W/m2) and exposure time (s).34 Apart from the UV fluence, the spe- impacts of airborne microbes.
cies of bacteria is also an important parameter determining the UV
inactivation efficacy of bacteria aerosol. Almost all bacteria, except
for spores, are sensitive to moderate levels of UV exposure, but the 2 | M ATE R I A L S A N D M E TH O DS
magnitude of the effect can be tremendously species dependent.35
The UV sensitivity of bacteria depends strongly on the type of cell 2.1 | Study design
membrane with which bacteria are being wired, and bacteria with
thick membranes are significantly less sensitive to UV inactivation. Escherichia coli (E. coli) (ATCC 10536), Serratia marcescens (S. marc-
Maximum effectiveness is usually observed for the UR-UVGI-MV escens) (ATCC 6911), and Staphylococcus epidermidis (S. epidermidis)
system against gram-negative than gram-positive bacteria spe- (ATCC 12228) were used as the challenge bioaerosols. These bacteria
cies.36 However, the inactivation performance of the UR-UVGI-LED are categorized as biosafety level 1. The bacteria cells were atomized
system against different airborne bacteria species has rarely been by a 24-jet Collison nebulizer (BGI, Inc., Waltham, MA, USA) that was
reported. Another important factor for consideration when making connected to a portable compressor and operated at a pressure of 45

F I G U R E 1 Experimental setup (A)


Process of aerosolization of test bacteria
(B) Process of air sampling
NUNAYON et al. | 183

psi (310.26 kPa) with filtered compressed air. Both the nebulizer and and the proper mixing of the aerosolized bacteria within the room, one
the portable compressor were positioned in a chamber next to the test 23 cm wide oscillating fan was mounted on a stool at the left corner of
chamber. The bioaerosols were delivered into the test chamber using a the wall through which the copper tubes were inserted. The fan was
thoroughly disinfected copper tube (labeled as ixa in Figure 1) that was operated with the dual oscillation of vertical and horizontal motions
12 mm in diameter and 1 m long. To improve the delivery efficiency (representing up/down & left/right, respectively) at constant rotation
and minimize possible bacteria loss, it [the copper tube] was attached speed during the experiments. The vertical airspeed of the fan was
to the end of a short buffer tube, between the nebulizer and copper 1.58 m/s at 56 cm away from the fan. On the sidewall of the chamber,
tube, instead of directly to the nozzle of the nebulizer. To achieve the one component of the UR-UVGI-LED device and one component of
aerosolization of the test samples at breathing level, the other end the UR-UVGI-MV device were installed to perform the disinfection,
of the copper tube was inserted at a height of 1.5 m above the floor separately. Both the control and the treatment experiments for the
through the south wall of the chamber. After the aerosolization pro- UR-UVGI-LED (or UR-UVGI-MV) irradiation system were conducted
cess, the copper tube was removed and replaced with another copper under the same test chamber conditions. The test chamber we used to
tube (labeled as ixb in Figure 1) in like manner to the copper tube de- study the efficacy of the UR-UVGI-LED and the UR-UVGI-MV devices
scribed above. To perform the air sampling, the replacement tube was separately is schematically illustrated in Figure 1.
also inserted into the chamber at one end (at the same position) and
connected to an impactor through a cast acrylic sheet squared box
(151 mm × 151 mm × 151 mm) at the other end. The relative positions 2.3 | The UR-UVGI-LED device
of the copper sampling tubes inlet were 54 cm away from the fan face
and 50 cm above the fan top; about 42 and 150 cm away from the UR- We developed a novel UR-UVGI-LED device (left on Figure 2). It
UVGI-LED and conventional UR-UVGI-MV devices, respectively, and consisted of 10 individual UVC-LEDs (Klaran series, Crystal IS Inc).
51 cm below both UR-UVGI-LED and conventional UR-UVGI-MV de- For each UCV-LED, the rated optical output was >20 mW under the
vices. The single-stage viable Andersen cascade impactor (N6, Thermo rated voltage of 9 V and the rated current was 350 mA. Hence, the
Scientific) was used to sample the airborne microbial concentrations. total rated optical output for 10 UVC-LEDs was >200 mW, and the
The sampler was connected to a vacuum pump, and the sampling vol- total input current was 3.5A. The array of the UVC-LEDs was fixed
ume was adjusted to a standard flow rate of 28.3 liters per minute onto a heat sink.
(LPM). The setup of the experimental design is shown in Figure 1. These 10 individual UVC-LEDs were arranged in the form of a
well-packed rectangular array (5 by 2 matrix). The complete module
of the UR-UVGI-LED device was hung midway from one of the walls
2.2 | Experimental chamber of the test room at a height of 2.01 m (see Figure 1).

We conducted the experiments in a full-scale test room with a dimen-


sion of 2.30 m (L) × 2.25 m (W) × 2.30 m (H). The chamber has its 2.4 | The conventional UR-UVGI-MV device
external walls insulated with fiberglass with one door but no windows.
The temperature and relative humidity of the chamber were controlled A 21 W modern ultraviolet germicidal lamp (TB-12-W/50, American
and monitored during the experiments. A detailed description of the Ultraviolet) was installed on the north wall (right on Figure 2), di-
test chamber has been given elsewhere.37 To achieve air movement rectly opposite the door in the test room (Figure 1). It consists of

FIGURE 2 Disinfection devices (A) UR-UVGI-LED (B) UR-UVGI-MV


184 | NUNAYON et al.

a glass envelope, a pair of electrodes, and a mercury amalgam. Once the 5 minutes bioaerosol aerosolization was complete,
Also, it had 14 angled panes of black-painted steel louvers. The UR- the collection of airborne bacteria into the bioaerosol sampler then
UVGI-MV system was also hung at a height of 2.01 m. commenced. Petri dishes containing 15 mL of nutrient agar medium
were interchangeably placed between the base and jet classification
stage of the impactor, and samples were obtained. The total sam-
2.5 | Selection and preparation of tested bacteria pling time was 13 minutes for 8 test times set at t = 0, 1, 3, 5, 7, 9,
11, and 13 minutes. This sampling duration was adopted because
Table 1 lists the bacteria strains and growth media. They are com- samples collected beyond this period were found to be nearly zero.
monly found in a variety of environments, and they are typical For each test time, the air samples were taken for 1 minute. We ap-
airborne microorganisms in bioaerosol research. These bacteria spe- plied a 1 minute interval between successive test times so that the
cies have also been used in previous UVGI studies for both upper surfaces of the impactor could be disinfected, and the agar plate re-
rooms and induct applications because they are vulnerable to inac- placed. For example, at test time t = 0, the air sample was taken im-
tivation by UVC.15,38 They are common bacteria in humans and usu- mediately after aerosolization without turning on the UR-UVGI-LED
ally cause serious infections. The selection criterion and preparation (or UR-UVGI-MV) device. After this moment, the UR-UVGI-LED (or
of the tested bacteria have previously been reported elsewhere.37 UR-UVGI-MV) system was powered up, and the air samples were
Concisely, for each experiment, a colony of each tested bacterium collected after a UV exposure of 1 minute. The same procedure
was separately obtained from frozen stocks, inoculated onto a nutri- was repeated for the other test times. Each experiment was inde-
ent agar (NA, BD) plate, and then incubated in a temperature con- pendently performed at least three times for each bacteria species.
trolled incubator at 37°C. In order to keep the concentration of each For every condition tested, the natural decay was determined by
tested bacterium at an approximately constant value for each test, performing the experiment without operating the UR-UVGI-LED (or
the suspension for each type of bacterium was successively diluted UR-UVGI-MV) system.
to different ratios and this provided a base suspension containing Upon sample collection, the inoculated nutrient agar plates were
populations of approximately 10 8 CFU/mL. The harvested cells in then incubated at 37°C for 24 hours. The number of colony-form-
50 mL of sterilized deionized water were transferred into a 24-jet ing units (CFUs) on each nutrient agar plate was manually counted.
collision nebulizer (BGI) for aerosolization. Occasionally, the bacteria To determine the performance of our UR-UVGI-LED (and the UR-
suspensions were stored at dry ice temperature and used for the UVGI-MV) devices, the rate of reduction in the bioaerosol concen-
aerosol experiments within less than 24 hours after preparation. trations with and without UVGI were determined and compared.
Any possible bacteria loss in the copper pipe from the nebulizer to
the chamber, and in the pipe and Perspex box between the chamber
2.6 | Experimental procedure and the impactor was considered negligible because the control ex-
periment, indirectly, had taken such losses into account.
Before performing the experiment, the impactor, air sampling box, For the UR-UVGI-MV system, the UV irradiance was measured
and copper tubes were sterilized by spraying them with 70% etha- using a UV/VIS fiber-optic spectrometer (ULS3648, Avantes). The
nol. About 1 hour before each experiment, the test chamber was measurement procedure has been reported elsewhere,40 but with
also decontaminated with a 21 W modern UV germicidal lamp (TB- the radiometer (Model ILT1400-A, International Light Technologies)
12-W/50, American Ultraviolet). The UV germicidal lamp was turned equipped with a 254 nm narrow-band detector (SEL/NS254/W).
off after disinfecting the room air. The sensor of the spectrometer, usually fixed to a fiber-optic cable,
For every test, a 50 mL solution of each type of test bacterium was mounted on a tripod. As the intensity is very sensitive to the
from the same stock was independently transferred into the nebu- source orientation, particularly at the vicinity of the fixture, it was
lizer, and filtered compressed air was then added into the nebulizer very crucial to ensure that the detector faced perpendicularly to the
to aerosolize and deliver the specific airborne bacterium into the UR-UVGI-MV fixture.
center of the test chamber. The oscillating fan was operated before The UV irradiance of the UR-UVGI-LED lights was also measured
aerosolization and ran continuously during air sampling to achieve a with the same spectrometer, which was positioned perpendicular to
well-mixed indoor condition. the UR-UVGI-LED device. The distance between the spectrometer's

TA B L E 1 Microorganisms, associated growth media and available UV rate constants

UV rate constants
Gram status Microorganism Collection number Culture medium (m2/J)39

Gram-negative bacteria E. coli ATCC 10536 Nutrient agar 0.21400-0.50628


S. marcescens ATCC 6911 Nutrient agar 0.06167-0.70657
Gram-positive bacteria S. epidermidis ATCC 12228 Nutrient agar 0.00800-0.12670

Abbreviation: ATCC, American Type Culture Collection.


NUNAYON et al. | 185

sensor and the UR-UVGI-LED device was 0.014 m. One distinctive the proportion at which the bioaerosol concentration in the indoor
feature of the UR-UVGI-LED device is the ease of changing its ir- space decays with respect to time under the environmental conditions
radiation by adjusting the input current while the intensity of the tested. The relationship between S and k is expressed in Equation (2).
UR-UVGI-MV system could not be adjusted. In this regard, three dif-
ferent output intensities (output power level) were selected for all Ct
= e−kt (2)
measurements (more details on Results). C0
Figure 3 shows the emission spectra for both the UR-UVGI-LED
and the UR-UVGI-MV systems. The spectrum for the UR-UVGI-LED As no mechanical ventilation was operated, removal due to
system was taken under the 3.5 A rated current with a correspond- mechanical ventilation was zero. Therefore, from Equation (2),
ing voltage of 8 V. The UR-UVGI-LED and the UR-UVGI-MV systems the decay rate of indoor bacteria during UR-UVGI-LED (or UR-
exhibited peak emission wavelengths at 270.8 and 254.96 nm, with UVGI-MV) device operation can be written as Equations (3a,
full widths at half maximum (FWHM) spectral bandwidth of 8 and 3b). 35,41
15 nm, respectively.
CLED−on (t)
= e−kt = e−(kn +kLED )t (3a)
C0
2.7 | Determination of inactivation level and
decay rates
Cmv−on (t)
= e−kt = e−(kn +kmv )t (3b)
C0
The inactivation level was determined by counting the CFUs with
and without the UR-UVGI-LED (or UR-UVGI-MV) device turned on.
A dimensionless parameter called the survival fraction (S) was de- where kn is the natural decay rate, kLED and kmv are decay rates due
termined based on Equation (1). It is the ratio of the concentration to UV irradiations from UR-UVGI-LED and conventional UR-UVGI-MV
of bioaerosols at any specific time, t, to the initial concentration of devices, respectively; the sum of kn and kLED (or kmv) is termed as total
bioaerosols measured at t = 0. decay rate. The slope of the best-fitted straight line of S against t gives
the k. The kLED (or kmv) presented in the results is termed as “effective
Ct
S= (1) k,” which was obtained by subtracting the natural die-off rate from the
C0
total decay rate.

In Equation (1), Ct is the bacteria concentration at a specified


elapsed time (or exposure time), t, and C0 is the initial bacteria con- 3 | R E S U LT S A N D D I S CU S S I O N
centration. We also determined the decay rate for the UR-UVGI ON
condition (kLED or kmv as the case may be) and the natural decay rate 3.1 | UR-UVGI-LED Intensity
(kn), that is, UR-UVGI OFF condition. The decay rate, k, is defined as
We measured the quantity of output light intensity emitted by the
10 UVC-LEDs at different input currents, and the results are shown
in Figure 4. The intensity increased with the input current up to 1.5
210
UR-UVGI-LED A, and then the intensity reached a plateau at 40.77 µW/cm2 until
180 UR-UVGI-MV 3.5 A (referred to as 100%). As the input currents just reached 0.6
Intensity (µW/cm2)

A and 0.3 A, the irradiation intensities were already 50% and 25%,
150
respectively, of the maximum value. The finding is very informative
120 and suggests that there is no need to operate the UR-UVGI-LED sys-
tem at the rated current (3.5 A in this case) to obtain 100% output as
90 the manufacturer suggested.
60
30 3.2 | UR-UVGI-MV irradiance
0
Unlike the UR-UV-GI-LED device, only one UV irradiance level

250 260 270 280 290 300 was achievable for the UR-UVGI-MV system because it was not
feasible to vary the intensity of the UV lamp. The spectrometer
Wavelength (nm)
measurements when the UR-UVGI-MV system was operating re-
F I G U R E 3 Emission spectra for UR-UVGI-LED and UR-UVGI- sulted in an average UV irradiance of 18.95 µW/cm2 at the irradi-
MV systems ated zone.
186 | NUNAYON et al.

50 reduction difference was observed for S. epidermidis (4.2-log10 re-


duction, see Figure 7E).
The decay rate is another parameter that quantifies the disinfec-
40 tion performance of the UR-UVGI system. The comparison of the
Intensity (µW/cm2)

decay rates for E coli, S. marcescens, and S. epidermidis, together with


the natural decay, are depicted in Figures 5, 6, and 7, respectively
30 3.5 A (100 %) for both the UR-UVGI-LED and the conventional UR-UVGI-MV sys-
40.77 µW/cm2 tems. We observed different decay rates among the three bacteria
species. Decay 1, 2, and 3 are results of three sets of the experi-
20 ments for each bacteria species.
0.6 A (50 %)
21.53 µW/cm2 For E. coli, the disinfection was significant compared with the
natural decay. At UR-UVGI-LED system's irradiation intensity levels
10 of 25%, 50%, and 100%, the mean decay rates increased system-
0.3 A (25 %)
atically and were −0.1673 ± 0.04 min−1, −0.2257 ± 0.04 min−1, and
11.35 µW/cm2
−0.3331 ± 0.07 min−1, respectively. The natural decay rate of E. coli was
0 −0.1420 ± 0.04 min−1. For S. marcescens, the mean decay rates were
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 −0.1640 ± 0.02 min−1, −0.2785 ± 0.05 min−1, −0.3583 ± 0.02 min−1,

Input current (A) and −0.1288 ± 0.01 min−1 for the 25%, 50%, 100% UR-UVGI-LED
output intensities and the natural decay rate, respectively. At the UR-
FIGURE 4 LED intensity against input current UVGI-LED system's irradiation intensity levels of 25%, 50%, and 100%,
the mean decay rates for S. epidermidis were −0.0207 ± 0.01 min−1,
−0.0113 ± 0.03 min−1, and −0.0487 ± 0.01 min−1, respectively.
3.3 | Comparison of disinfection under UR-UVGI- Meanwhile, the natural decay rate was −0.0330 ± 0.01 min−1. Almost
LED and UR-UVGI-MV irradiations all results showed an increase in decay rate with intensity.
Correspondingly, the conventional UR-UVGI-MV system pro-
The disinfection of various bacteria in the test chamber was con- vided an average −0.3867 ± 0.08 min−1, −0.4745 ± 0.002 min−1, and
ducted with the UR-UVGI-LED and the conventional UR-UVGI-MV −0.1624 ± 0.02 min−1, as decay rates for E. coli, S. marcescens, and
devices independently. The quantity ln(Ct/C0) was plotted against S. epidermidis, respectively.
the exposure time. The illustrations of the control groups of three The current finding agrees with the previous discussion of the
bacteria in Figures 5, 6, and 7 demonstrate that the populations that log-reductions in CFU counts, which showed that the gram-positive
were not exposed to UVC radiation (either UR-UVGI-LED or UR- bacterium, S. epidermidis, due to its thicker cell wall42 was less sus-
UVGI-MV system) showed little reduction throughout the duration ceptible to irradiation by both the UR-UVGI-LED and the conven-
of the experiments. tional UR-UVGI-MV systems than the gram-negative bacteria, E. coli,
For the UR-UVGI-LED system operated at 25%, 50%, and 100% and S. marcescens.43,44 Previous studies have suggested that S. epi-
irradiation intensity levels, the highest levels of inactivation for the dermidis has similar responses to UV irradiation.36,37
E. coli bacterium following exposures of 13 minutes were approx- The effective disinfection potential of the UR-UVGI-LED system
imate CFU count reductions of 4.3-log10, 5.3-log10, and 7.4-log10, is observed in this study through the significant reductions and/or
respectively. For the S. marcescens bacterium, the highest levels of decay rates. It is, however, noticed that the UR-UVGI-MV system
inactivation following exposures of 13 minutes to the UR-UVGI-LED was 1.16, 1.32, and 3.33 times higher than the highest decay rates
at 25%, 50%, and 100% irradiation intensity levels were approximate by the UR-UVGI-LED system for E. coli, and S. marcescens, and S. epi-
CFU count reductions of 3.7-log10, 5-log10 , and 6.4-log10, respec- dermidis, respectively. A thoughtful discussion between the two sys-
tively. However, the exposure of the S. epidermidis to the same ir- tems is given.
radiation intensities of the UR-UVGI-LED system at 25%, 50%, and The design flexibility of the UR-UVGI-LED system surpasses that
100% amounted to insignificant inactivation compared with the con- of the mercury vapor lamp system, offering the opportunity of innova-
trol. Approximate reductions of 1.0-log10 in the CFU counts of S. epi- tive disinfection systems with a high level of integration. Several other
dermidis were observed after an exposure duration of 13 minutes for potential advantages of using the UR-UVGI-LED system over the
all the UR-UVGI-LED system's intensities. lamp-based—UR-UVGI-MV system that must be considered for prac-
Similarly, we observed in our data that the concentration reduc- tical applications include compact size, no warm-up time, less mainte-
tions for E. coli (7.1-log10 reduction, see Figure 5E), and S. marcescens nance cost, and the potential for very long operational lifetimes of the
(6.8-log10 reduction, see Figure 6E) induced by the conventional UR- LEDs before replacement. Longer-lifetime is a very important benefit
UVGI-MV system were not significantly higher than the peak reduc- especially for applications in rural areas, where maintenance and re-
tion reported for the UR-UVGI-LED system above, but substantial pairs are usually difficult. For manufacturing LEDs, no environmental
NUNAYON et al. | 187

F I G U R E 5 The inactivation decay (A) Exposure time (min) (B) Exposure time (min)
rates of E. coli under (A) natural die-
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
off condition (no UVC irradiation), (B)
0.0 0.0

Concentration, ln (C)
Concentration, ln (C)
UR-UVGI-LED at a 25% irradiation kLED = –0.1673 ± 0.04 min–1
intensity level (C) UR-UVGI-LED at a 50% –1.5 –1.5
irradiation intensity level, (D) UR-UVGI- –3.0 –3.0
LED at a 100% irradiation intensity level,
–4.5 Decay 1: R 2 = 0.993 –4.5
and (E) UR-UVGI-MV irradiation Decay 2: R 2 = 0.967
–6.0 Decay 3: R 2 = 0.964 –6.0 Decay 1: R 2 = 0.988
–7.5 –7.5 Decay 2: R 2 = 0.992
kn = –0.1420 ± 0.04 min–1 Decay 3: R 2 = 0.943
–9.0 –9.0
Natural die-off condition UR-UVGI-LED at 25%
(no UVC irradiation) irradiation intensity level

(C) Exposure time (min) (D) Exposure time (min)


0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
0.0 0.0
Concentration, ln (C)

Concentration, ln (C)
kLED = –0.2257 ± 0.04 min –1
kLED = –0.3331 ± 0.07 min–1
–1.5 –1.5
–3.0 –3.0
–4.5 –4.5
–6.0 –6.0
Decay 1: R 2 = 0.942 Decay 1: R 2 = 0.995
–7.5 Decay 2: R 2 = 0.991 –7.5 Decay 2: R 2 = 0.994
–9.0 Decay 3: R 2 = 0.993 –9.0 Decay 3: R 2 = 0.989

UR-UVGI-LED at 50% UR-UVGI-LED at 100%


irradiation intensity level irradiation intensity level

(E) Exposure time (min)


0 2 4 6 8 10 12 14
0.0
Concentration, ln(C)

kMV = –0.3867 ± 0.08 min–1


–1.5
–3.0
–4.5
–6.0
Decay 1: R 2 = 0.95
–7.5 Decay 2: R 2 = 0.98
–9.0 Decay 3: R 2 = 0.96

UR-UVGI-MV irradiation

hazardous materials are needed or generated, and no harmful ozone 3.4 | Comparison of energy consumption
is generated during the operation. Moreover, the wavelength diversity between UR-UVGI-LED and UR-UVGI-MV systems
of UV-LEDs creates the opportunity to select a certain wavelength
for a specific purpose. This can target specific pathogens to enhance Energy consumption was adopted as a common index to evaluate
disinfection efficacy. and compare the energy utilization efficiency between the devel-
In view of the clear potential benefits that we have high- oped UR-UVGI-LED and the conventional UR-UVGI-MV systems. In
lighted for LED-based air disinfection systems, there is a need to this study, only the energy consumed by each of these disinfection
strengthen developmental efforts to expressly enhance or improve devices was considered whereas other ancillary consumption was
the UR-UVGI-LED system's performance for air disinfection. This not accounted for. To analyze the energy implication of using the de-
is possible by considering both the typical challenges and various veloped UR-UVGI-LED and the conventional UR-UVGI-MV systems,
advantages of the UR-UVGI-LED system to maintain a highly reli- the annual energy consumed by both UR-UVGI devices was deter-
able, compact and robust system and ensure the preservation of mined at comparable effectiveness, and the annual energy costs (C)
the most beneficial qualities of LED-based systems compared to were further estimated. The annual energy cost was calculated from
lamp-based systems. Equation (4).
188 | NUNAYON et al.

Exposure time (min) Exposure time (min) F I G U R E 6 The inactivation decay


(A) (B)
rates of S. marcescens under (A) natural
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
die-off condition (no UVC irradiation),
0.0 0.0

Concentration, ln (C)
Concentration, ln (C)

kLED = –0.1640 ± 0.02 min–1 (B) UR-UVGI-LED at a 25% irradiation


–1.5 –1.5 intensity level (C) UR-UVGI-LED at a 50%
–3.0 –3.0 irradiation intensity level, (D) UR-UVGI-
Decay 1: R 2 = 0.940 LED at a 100% irradiation intensity level,
–4.5 Decay 2: R 2 = 0.940 –4.5 and (E) UR-UVGI-MV irradiation
–6.0 Decay 3: R 2 = 0.910 –6.0 Decay 1: R 2 = 0.980
–7.5 –7.5 Decay 2: R 2 = 0.950
kn = –0.1288 ± 0.01 min–1 Decay 3: R 2 = 0.970
–9.0 –9.0
Natural die-off condition UR-UVGI-LED at 25%
(no UVC irradiation) irradiation intensity level

(C) Exposure time (min) (D) Exposure time (min)


0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
0.0 0.0
Concentration, ln (C)
Concentration, ln (C)

kLED = –0.2785 ± 0.05 min–1 kLED = –0.3583 ± 0.02 min–1


–1.5 –1.5
–3.0 –3.0
–4.5 –4.5
–6.0 –6.0
Decay 1: R 2 = 0.960 Decay 1: R 2 = 0.999
–7.5 Decay 2: R 2 = 0.997 –7.5 Decay 2: R 2 = 0.980
Decay 3: R 2 = 0.960 –9.0 Decay 3: R 2 = 0.996
–9.0
UR-UVGI-LED at 50% UR-UVGI-LED at 100%
irradiation intensity level irradiation intensity level

(E) Exposure time (min)


0 2 4 6 8 10 12 14
0.0
Concentration, ln(C)

Decay 1: R2 = 0.94
–1.5 Decay 2: R2 = 0.92
Decay 3: R2 = 0.90
–3.0
–4.5
–6.0
–7.5
kMV = –0.4745 ± 0.002 min–1
–9.0
UR-UVGI-MV irradiation

current did not show any tangible difference in the overall output in-
C = P ⋅ hw ⋅ Re (4) tensity (Figure 4). It is reasonable not to consider the manufacturer's
rated current of 3.5 A in our calculation. Therefore, we adopted 2.0
where C is the cost of the annual energy consumption, US$/year; hw A as the input current to determine the power of the UR-UVGI-LED
is the total working hour of the UR-UVGI-LED or the conventional system. In addition, the rated voltage of the UR-UVGI-LED system
UR-UVGI-MV system in 1 year, hw = 2920 hours per year; Re is the when operated at a full gear was 9 V, but for safety reasons and to
45
rate of one Kilowatt-hour of electricity, Re = 0.14 $/kWh ; P is the avoid damaging the LEDs, we operated the UR-UVGI-LED system at
power of the UR-UVGI-LED or the conventional UR-UVGI-MV sys- 8 V. Thus, by substituting the values of I and V into the power equa-
tem at full operation intensity. In our calculation of the annual en- tion, the power of the UR-UVGI-LED system at full operation inten-
ergy cost, for the developed UR-UVGI-LED device, the power (P) in sity was 16 W. Whereas, for the UR-UVGI-MV system, the power (P)
Equation (4) was the product of the input current at full output inten- was measured using an AC clamp power meter (CM3286-01) and it
sity and the corresponding voltage (ie, P = I·V) where current, I = 2A was found to be 21 W.
and voltage, V = 8V. As mentioned earlier, the UR-UVGI-LED system By our calculation, it was shown that the annual fee consumed
could attain almost full output intensity at 1.5 A and increasing the by the UR-UVGI-LED system (6.54 $/year) was lower than that
NUNAYON et al. | 189

F I G U R E 7 The inactivation decay (A) Exposure time (min) (B) Exposure time (min)
rates of S. epidermidis under (A) natural
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
die-off condition (no UVC irradiation),
0.0 0.0

Concentration, ln (C)
Concentration, ln (C)
(B) UR-UVGI-LED at a 25% irradiation
intensity level (C) UR-UVGI-LED at a 50%
irradiation intensity level, (D) UR-UVGI-
–1.5 –1.5
Decay 1: R 2 = 0.810
LED at a 100% irradiation intensity level, Decay 1: R 2 = 0.860
–3.0 –3.0 Decay 2: R 2 = 0.995
and (E) UR-UVGI-MV irradiation Decay 2: R 2 = 0.960
Decay 3: R 2 = 0.950
Decay 3: R 2 = 0.890
–4.5 –4.5
kn = –0.0330 ± 0.01 min –1 kLED = –0.0207 ± 0.01 min–1
–6.0 –6.0
Natural die-off condition UR-UVGI-LED at 25%
(no UVC irradiation) irradiation intensity level

(C) Exposure time (min) (D) Exposure time (min)


0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
0.0 0.0
Concentration, ln (C)

Concentration, ln (C)
–1.5 –1.5
Decay 1: R 2 = 0.950 Decay 1: R 2 = 0.990
–3.0 Decay 2: R 2 = 0.990 –3.0 Decay 2: R 2 = 0.980
Decay 3: R 2 = 0.980 Decay 3: R 2 = 0.995
–4.5 –4.5
kLED = –0.0113 ± 0.03 min –1 kLED = –0.0487 ± 0.01 min–1
–6.0 –6.0
UR-UVGI-LED at 50% UR-UVGI-LED at 100%
irradiation intensity level irradiation intensity level

(E) Exposure time (min)


0 2 4 6 8 10 12 14
0.0
Concentration, ln (C)

kMV = –0.1624 ± 0.02 min–1

–1.5

–3.0

–4.5 Decay 1: R 2 = 0.98


Decay 2: R 2 = 0.99
–6.0 Decay 3: R 2 = 0.99

UR-UVGI-MV irradiation

by the conventional UR-UVGI-MV system (8.58 $/year) over the 4 | CO N C LU S I O N S


course of a year. This implies that the conventional UR-UVGI-MV
system is 31.19% more expensive to operate. Notwithstanding the The Minamata Convention on Mercury has a serious impact on the
heat dissipation of LEDs, our (developed) UR-UVGI-LED device is production of mercury-added products. A new UV source is needed
still marginally energy efficient compared with the conventional to substitute the conventional low-pressure mercury vapor lamps.
UR-UVGI-MV system. This indicates that useful energy is gener- In this work, a novel upper room UR-UVGI-LED device was
ally minimal compared with the amount wasted as heat. Unlike tested in a full-size experimental chamber and the performance
the conventional UR-UVGI-MV system, the application of the was compared with the conventional UR-UVGI-MV system. Three
UR-UVGI-LED system for room air disinfection is new and has a different bacteria were selected: two gram-negative bacilli—E. coli
high potential for future development. It is expected that with the and S. marcescens, and one gram-positive coccus—S. epidermidis.
recent advancements in electro-optical technologies, the energy The performance metrics employed in this study to compare
efficiency of the UR-UVGI-LED system will likely increase more the effectiveness of the two UR-UVGI systems in reducing in-
substantially. door bioaerosols were log reduction and decay rate. Overall, the
190 | NUNAYON et al.

disinfection performance (in terms of the decay rate) of the UR- 5. Stephens B, Novoselac A, Siegel JA. The effects of filtration on
pressure drop and energy consumption in residential HVAC sys-
UVGI-LED system at maximum intensity was 13.86%, 24.49%, and
tems (RP-1299). HVAC&R Res. 2010b;16(3):273-294.
70.01% less than the disinfection performance of the conventional 6. Lu Z, Lu WZ, Zhang JL, Sun DX. Microorganisms and parti-
UR-UVGI-MV system for E. coli, S. marcescens, and S. epidermidis, cles in AHU systems: Measurement and analysis. Build Environ.
respectively. 2009;44:694-698.
The gram-negative species showed significant decay rate and 7. Liu Z, Ma S, Cao G, Meng C, He BJ. Distribution characteristics,
growth, reproduction and transmission modes and control strate-
sensitivity upon exposure to separate UV irradiance from both
gies for microbial contamination in HVAC systems: A literature re-
UR-UVGI-LED and conventional UR-UVGI-MV systems while the view. Energy Build. 2018;177:77-95.
gram-positive species showed reduced decay rate and sensitivity. 8. Huang R, Agranovski I, Pyankov O, Grinshpun S. Removal of viable
This demonstrates that the magnitude of reduction in indoor micro- bioaerosol particles with low-efficiency HVAC filter enhanced by con-
tinuous emission of unipolar air ions. Indoor Air. 2008;18:106-112.
bial concentration by the UR-UVGI-LED and the conventional UR-
9. Reed NG. The history of ultraviolet germicidal irradiation for air dis-
UVGI systems varies widely between species of bacteria. infection. Public Health Rep. 2010;125:15-27.
The UR-UVGI-LED system has largely shown that it is a prom- 10. Santos AL, Moreirinha C, Lopes D, Esteves AC, Henriques I, Almeida
ising tool for future applications in airborne pathogen disinfection. A, Domingues MR, Delgadillo I, Correia A, Cunha A. Effects of UV
radiation on the lipids and proteins of bacteria studied by mid-infra-
The high disinfection potential at this emerging stage indicates that
red spectroscopy. Environ Sci Technol. 2013;47:6306-6315.
it stands the chance to outperform the conventional UR-UVGI-MV 11. NIOSH. Criteria for a recommended standard: Occupational ex-
system in the near future. In addition, the UR-UVGI-LED system can posure to ultraviolet radiation. Publication No. 73-11009 (1972)
be slightly more energy cost-effective at comparable disinfection ef- 1-100.
12. American Conference of Governmental Industrial Hygienists
fectiveness than the conventional UR-UVGI-MV system for room air
(ACGIH). Ultraviolet Radiation: TLV(R) Physical Agents 7th Edition
microbial contaminants. Documentation (2010).
We also recommend that the performance of the UR-UVGI-LED 13. International Commission on Non-Ionizing Radiation Protection
system under different environmental conditions should be explored (ICNIRP). Guidelines on limits of exposure to ultraviolet radiation
of wavelengths between 180 and 400 nm (incoherent optical radi-
in future investigations. Also, airborne bacteria have significantly
ation). Health Phys. 2004;87(2):171-186.
different biological characteristics compared with airborne viruses, 14. Centers for Disease Control and Prevention-National Institute
therefore, the disinfection efficacy of the UR-UVGI-LED system for for Occupational Safety and Health (CDC-NIOSH). Environmental
disinfection of room airborne viruses is worth investigating in future Control for Tuberculosis: Basic Upper-Room Ultraviolet Germicidal
studies. Irradiation Guidelines for Healthcare Settings.
15. Ko G, First MW, Burge HA. The characterization of upper-room ul-
traviolet germicidal irradiation in inactivating airborne microorgan-
AC K N OW L E D G E M E N T isms. Environ Health Perspect. 2002;110(1):95-101.
This research was fully supported by General Research Funds 16. Escombe AR, Moore DA, Gilman RH, Navincopa M, Ticona E,
from the Research Grants Council of the Hong Kong Special Mitchell B, Noakes C, Martínez C, Sheen P, Ramirez R, Quino W,
Gonzalez A, Friedland JS, Evans CA. Upper-room ultraviolet light
Administrative Region of China [CityU 11273116 and CityU
and negative air ionization to prevent tuberculosis transmission.
11204217]. PLoS Medicine. 2009;6(3):0001-0012.
17. Riley RL, Knight M, Middlebrook G. Ultraviolet susceptibility of BCG
ORCID and virulent tubercle bacilli. Am Rev of Respir Dis. 1976;113:413-418.
18. Mphaphlele M, Dharmadhikari AS, Jensen PA, Rudnick SN, van
Sunday S. Nunayon https://orcid.org/0000-0001-6962-0484
Reenen TH, Pagano MA, Leuschner W, Sears TA, Milonova SP, van
Huihui Zhang https://orcid.org/0000-0002-5316-5533 der Walt M, Stoltz AC, Weyer K, Nardell EA. Institutional tuber-
Alvin C. K. Lai https://orcid.org/0000-0002-6202-1988 culosis transmission. Controlled trial of upper room ultraviolet air
disinfection: a basis for new dosing guidelines. Am J Respir Crit Care
Med. 2015;192:477-484.
REFERENCES
19. Nardell EA. Indoor environmental control of tuberculosis and other
1. Azimi P, Stephens B. HVAC filtration for controlling infec- airborne infections. Indoor Air. 2015;26:79-87.
tious airborne disease transmission in indoor environments: 20. Goldner JL, Moggio M, Beissinger SF, McCollum DE. Ultraviolet
Predicting risk reductions and operational costs. Build Environ. light for the control of airborne bacteria in the operating room. Ann
2013;70:150-160. N Y Acad Sci. 1980;353:271-284.
2. Piekarska K, Trusz A, Szczęśniak S. Bacteria and fungi in two 21. Macher JM, Alevantis LE, Chang YL, Liu KS. Effect of ultraviolet
air handling units with air recirculating module. Energy Build. germicidal lamps on airborne microorganisms in an outpatient
2018;178:154-164. waiting room. Applied Occupational and Environmental Hygiene.
3. ANSI/ASHRAE Standard 52.2-1999. Method of Testing General 1992;7(8):505-513.
Ventilation Air-Cleaning Devices for Removal Efficiency by Particle 22. Miller SL, Macher JM. Evaluation of a methodology for quantifying
Size. the effect of room air ultraviolet germicidal irradiation on airborne
4. Stephens B, Siegel J, Novoselac A. Energy implications of filtration bacteria. Aerosol Sci Technol. 2000;33(3):274-295.
in residential and light-commercial buildings, ASHRAE Research 23. Dick EC, Jennings LC, Mink KA, Wartgow CD, Inhorn SL. Aerosol
Project (RP-1299) Final Report 2010. Available at http://built- transmission of rhinovirus colds. J Infect Dis. 1987;156:442-448.
envi.com/wp-conte​nt/uploa​ds/2014/06/ASHRAE-D-RP-1299.pdf 24. Musher DM. How contagious are common respiratory tract infec-
(Accessed 27/09/19). tions? N Engl J Med. 2003;348:1256-1266.
NUNAYON et al. | 191

25. McDevitt JJ, Rudnick SN, Radonovich LJ. Aerosol Susceptibility 37. Lai ACK, Nunayon SS, Tan TF, Li WS. A pilot study on the disinfec-
of Influenza Virus to UV-C Light. Appl Environ Microbiol. tion efficacy of localized UV on the flushing-generated spread of
2012;78:1666-1669. pathogens. J Hazard Mat. 2018;358:389-396.
26. Olsen SJ, Chang HL, Cheung TYY, Tang AF, Fisk TL, Ooi SP, Kuo 38. Green CF, Scarpino PV. The use of ultraviolet germicidal irradia-
HW, Jiang DD, Chen KT, Lando J, Hsu KH, Chen TJ, Dowell SF. tion (UVGI) in disinfection of airborne bacteria. Environ Eng Policy.
Transmission of the severe acute respiratory syndrome on aircraft. 2001;3(1):101-107.
N Engl J Med. 2003;349:2416-2422. 39. Kowalski W. Ultraviolet Germicidal Irradiation Handbook: UVGI for Air
27. Sarigiannis DA, Karakitsios SP, Antonakopoulou MP, Gotti A. and Surface Disinfection. London, New York: Springer Heidelberg
Exposure analysis of accidental release of mercury from compact Dordrecht; 2009.
fluorescent lamps (CFLs). Sci Total Environ. 2012;435–436:306-315. 40. Yang Y, Lai ACK, Kong RYC, Deng Q. Experimental and numerical
28. Rudnick SN, First MW, Sears T, Vincent RL, Brickner PW, Ngai PY, study of the performance of upper-room ultraviolet germicidal ir-
Zhang J, Levin RE, Chin K, Rahn RO, Miller SL, Nardell EA. Spatial radiation with the effective Z-value of airborne bacteria. Aerosol Sci
distribution of fluence rate from upper-room ultraviolet germicidal Technol. 2017;51:1123-1134.
irradiation: Experimental validation of a computer-aided design 41. Yang Y, Lai ACK, Wu C. Study on the disinfection efficiency of mul-
tool. HVAC&R Res. 2012;8:774-794. tiple upper-room ultraviolet germicidal fixtures system on airborne
29. Morrison GC, Shaughnessy R, Shu S. Setting maximum emission microorganisms. Build Environ. 2016;103:99-110.
rates from ozone emitting consumer appliances in the United States 42. Beauchamp S, Lacroix M. Resistance of the genome of Escherichia
and Canada. Atmos Environ. 2011;45:2009-2016. coli and Listeria monocytogenes to irradiation evaluated by the in-
30. Schalk S, Adam V, Arnold E, Brieden K, Voronovand A, Witzke H. duction of cyclobutane pyrimidine dimers and 6-4 photoprod-
UV-Lamps for Disinfection and Advanced Oxidation - Lamp Types. ucts using gamma and UV-C radiations. Radiat Physics Chem.
Technologies and Applications. IUVA News. 2006;8(1):32-37. 2012;81:1193-1197.
31. Crawford MH, Banas MA, Ross MP, Ruby DS, Nelson JS, Boucher R, 43. Gayan E, Serrano MJ, Raso J, Alvarez I, Condon S. Inactivation of
Allerman AA. Final LRDR report: Ultraviolet water purification sys- Salmonella enterica by UV-C light alone and in combination with
tems for rural environments and mobile applications. Sandia Report. mild temperatures. Appl Environ Microbiol. 2012;78:8353-8361.
2005. 44. Kim SJ, Kim DK, Kang DH. Using UVC light-emitting diodes at
32. Li GQ, Wang WL, Huo ZY, Lu Y, Hu HY. Comparison of UV-LED and wavelengths of 266 to 279 nanometers to inactivate foodborne
low pressure UV for water disinfection: Photoreactivation and dark pathogens and pasteurize sliced cheese. Appl Environ Microbiol.
repair of Escherichia coli. Water Res. 2017;126:134-143. 2016;82:11-17.
33. Zou X, Lin Y, Xu B, Cao T, Tang Y, Pan Y, Gao Z, Gao N. Enhanced 45. HK Electric Investment. Estimated electricity cost of common ap-
inactivation of E. coli by pulsed UV-LED irradiation during water dis- pliances, 2017. https​://www.hkele​c tric.com/en/custo​mer-servi​
infection. Sci Total Environ. 2019;650:210-215. ces/billi​ng-payme​nt-elect​ricity-tarif​fs/know-more-about-elect​rici-
34. Bowker C, Sain A, Shatalov M, Ducoste J. Microbial UV fluence-re- ty-consu​mptio​n/estim​ated-elect​ricity-cost-of-common-appli​ances​
sponse assessment using a novel UV-LED collimated beam system. Accessed March 16, 2019.
Water Res. 2011;45(5):2011-2019.
35. Xu P, Peccia J, Fabian P, Martyny JW, Fennelly KP, Hernandez
M, Miller SL. Efficacy of ultraviolet germicidal irradiation of up-
How to cite this article: Nunayon SS, Zhang HH, Lai ACK.
per-room air in inactivating airborne bacterial spores and mycobac-
teria in full-scale studies. Atmos Environ. 2003;37:405-419.
Comparison of disinfection performance of UVC-LED and
36. Yang Y, Zhang H, Nunayon SS, Chan V, Lai ACK. Disinfection ef- conventional upper-room UVGI systems. Indoor Air.
ficacy of ultraviolet germicidal irradiation on airborne bacteria in 2020;30:180–191. https​://doi.org/10.1111/ina.12619​
ventilation ducts. Indoor Air. 2018;28(6):806-817.

You might also like