Inactivation of microorganisms in untreated water by a continuous flow system with supercritical CO2 bubbling

ABSTRACT The effects of supercritical CO2 bubbling (SC-CO2) treatment on the inactivation of microorganisms in water prior to treatment at a municipal water filtering plant (untreated water) were investigated as a way to produce safe drinking water. The coliform bacterial count decreased concomitantly with increasing CO2/sample flow rate in the SC-CO2 treatment. In particular, coliform bacteria could not be detected at a CO2/sample flow rate greater than 55%. Also, the total bacterial count dropped rapidly at first stage and slowly at second stage in the SC-CO2 treatment. Upon observation of Escherichia coli before and after the SC-CO2 treatment with scanning electron microscopy and transmission electron microscopy, it was observed that the cells treated with SC-CO2 were shorter than untreated cells and that cytoplasm with low electronic density in the treated cells disappeared. In addition, four types of metabolic enzyme in E. coli cells were effectively inactivated by the SC-CO2 treatment. These results suggested that SC-CO2 treatment could effectively inactivate microorganisms in untreated water, and induce morphological changes and inactivate metabolic enzymes of E. coli cells

Inactivation of microorganisms in untreated water by a

Fumiyuki KOBAYASHI*, Futoshi YAZAMA**, Hiromi IKEURA*, Yasuyoshi HAYATA* Norio MUTO** and Yutaka OSAJIMA***

* School of Agriculture, Meiji University, 1-1-1, Higashimita, Tama-ku, Kawasaki, Kanagawa, 214-8571, Japan

** Faculty of Life and Environmental Sciences, Prefectural University of Hiroshima, 562, Nanatsuka, Shobara, Hiroshima 727-0023, Japan

*** Faculty of Agriculture, Kyusyu University, 6-10-1, Hakozaki, Higashi-ku, Fukuoka, 807-8586, Japan

ABSTRACT

The effects of supercritical CO 2 bubbling (SC-CO 2 ) treatment on the inactivation of microorganisms in water prior to treatment at a municipal water filtering plant (untreated water) were investigated as a way to produce safe drinking water The coliform bacterial count decreased concomitantly with increasing CO 2 /sample flow rate in the SC-CO 2 treatment In particular, coliform bacteria could not be detected at a CO 2 /sample flow rate greater than 55% Also, the total bacterial count dropped rapidly at first stage and slowly at second stage in the SC-CO 2 treatment Upon observation of Escherichia coli before and after the SC-CO2 treatment with scanning electron microscopy and transmission electron microscopy, it was observed that the cells treated with SC-CO 2 were shorter than untreated cells and that cytoplasm with low electronic density in the treated cells disappeared In addition, four types of metabolic enzyme in

E coli cells were effectively inactivated by the SC-CO2 treatment These results suggested that SC-CO 2 treatment could effectively inactivate microorganisms in untreated water, and induce

morphological changes and inactivate metabolic enzymes of E coli cells

Keywords: supercritical carbon dioxide bubbling, total and coliform bacteria, electron microscopy

INTRODUCTION

Chlorine inactivation has generally been used to inactivate microorganisms in tap water However, it has been suggested that chlorine is toxic to the human body even at low concentrations In particular, there is a possibility that trihalomethanes, which are carcinogenic agents, are produced by the chemical reaction between chlorine and organic compounds in water (Graham et al., 1989) Thus, bottled water consumption has rapidly increased, as many consumers prefer to drink non-chlorinated water However, imported bottled water might not be sanitary, because it is not subjected to the water quality standards used for drinking water (Satsuta et al., 2001; Warburton, 1993) Therefore, many scientists have proposed alternative methods to chlorine inactivation to inactivate microorganisms in drinking water, such as titanium dioxide photocatalytic reaction (Wist et al., 2002), iodine (Backer and Hollowell, 2000) and ozone (Cho et al., 2003) treatments

Pasteurization using pressurized carbon dioxide has been widely studied in the field of food science (Lin et al., 1992, 1993; Ballestra et al., 1996; Louka et al., 1999; Hong and Pyun, 1999, 2001; Gunes et al., 2006; Garcia-Gonzalez et al., 2009) Ishikawa et al (1995a) developed supercritical carbon dioxide bubbling (SC-CO2 treatment), which

can increase the dissolved CO2 concentration in a solution with the use of a filter, and Shimoda et al (1998, 2001, 2002) reported that some bacteria could be inactivated by SC-CO2 treatment In our previous papers, we reported that certain compounds that give tap water a musty odor could be removed by a continuous flow treatment with SC-CO2

(Kobayashi et al., 2006) and suggested that inactivation of Escherichia coli in sterile

distilled water and coliform bacteria in water prior to treatment at a municipal water filtering plant (untreated water) by the SC-CO2 treatment depended on the dissolved

CO2 concentration (Kobayashi et al., 2007a) In this study, we aim to determine the effects of the CO2/sample flow rate and exposure time in the SC-CO2 treatment on the inactivation of coliform and total bacteria in untreated water and to propose the SC-CO2

treatment as a novel method for inactivating microorganisms in drinking water In addition, we aim to investigate the effect of the SC-CO2 treatment with regard to

structural damage and enzyme inactivation in E coli cells, which we previously

demonstrated could be reduced to nondetectable levels by the SC-CO2 treatment (Kobayashi et al., 2007a)

MATERIALS AND METHODS

Untreated water

As described in our previous paper (Kobayashi et al., 2007a), water was taken from a municipal water filtering plant (Miyoshi, Hiroshima, Japan) prior to being treated for use as tap water The coliform bacteria and total bacterial counts in this untreated water were approximately 4.0×102 MPN/100ml and 9.0×102～1.2×103 CFU/ml, respectively

Preparation of E coli suspensions

Strains of E coli (ATCC 11775) were inoculated to each test tube containing 10 ml of

nutrient broth (Nihon Pharmaceutical Co., Ltd., Tokyo, Japan) and incubated at 37°C for 16 hr The cultures were then transferred to 300 ml flasks containing 190 ml of nutrient broth and incubated at 37°C for 24 hr Microorganisms were collected by centrifuge (4°C, 8000 rpm, 10 min) and suspended in sterile distilled water at approximately 5.0×108 CFU/ml

Continuous flow treatment with SC-CO 2 bubbling

The schematic diagram of the device used for continuous flow treatment with SC-CO2

bubbling is shown in Figure 1 CO2 compressed by a pump was bubbled through a filter (with 10 μm pore size) in a mixing vessel (about 10 ml in volume) and mixed with the sample water The mixed sample and SC-CO2 were fully agitated in a coil (about 30 ml

in volume) and delivered to a gas-liquid separation vessel (about 160 ml in volume), and the foul-odor components in the sample were transferred to SC-CO2, discharged from valve I and collected in a knockout drum When the surface of the sample in the gas-liquid separation vessel touches thermocouple I, the sample is discharged from valve II and decompressed to atmospheric pressure When the surface of the sample loses contact with thermocouple II, the discharge of the sample is stopped Before this device was used, a 0.1% sodium hypochlorite solution was made to flow into the vessel for 30 min, and the device was sufficiently washed with sterile distilled water The SC-CO2 treatment conditions were as follows: temperatures were 35, 45 and 55°C, pressure was 10 MPa, sample flow rates were 2.5, 5.0 and 15 ml/min, CO2 flow rates

CO2/sample flow rate was calculated as follows:

CO2/sample flow rate = CO2 flow rate/sample flow rate ×100

valve I

pump heating coil

pump heater cooler

agitation coil

knockout drum

treatment sample microfilter (pore size 10 μm)

thermocouple II

mixing vessel

valve II thermocouple I

gas-liquid separation vessel

CO2

CO2

sample

CO2

drain

Figure 1 - Schematic diagram of the experimental apparatus for a continuous flow

treatment with SC-CO2 bubbling

Measurement of surviving bacteria

The coliform and total bacterial counts in the untreated water were determined using the method of Japan Water Works Association (JWWA, 2001) For this measurement, 1 ml

of each sample diluted with physiological saline solution was added to each of five test tubes (φ18 mm × 180 mm), each of which contained 10 ml of lactose bile (LB) broth (Nihon Pharmaceutical Co., Ltd., Tokyo, Japan) and a small glass tube (φ8 mm ×

30 mm), and samples were incubated at 37°C for 48 hr The cultured LB broth was then transferred to test tubes (φ18 mm × 180 mm) that contained 10 ml of brilliant green lactose bile (BGLB) broth (Nihon Pharmaceutical Co., Ltd., Tokyo, Japan) and a small glass tube (φ8 mm × 30 mm), with a platinum loop, and incubated at 37°C for 48

hr Based on the presence of gas in the small glass tube, the most probable number (MPN) was calculated and the coliform bacterial count was determined The detection limit of the coliform bacterial count by the detection method was 0 MPN/100ml (JWWA, 2001) We also determined the total bacterial count using the standard plate counting method One ml of appropriate dilutions of samples was inoculated in standard plate count agar (Nihon Pharmaceutical Co., Ltd., Tokyo, Japan) and aerobically incubated at 37°C for 48 hr, and the colony-forming units (CFU) were counted The coliform and total bacterial counts were expressed in MPN/100ml and CFU/ml, respectively (JWWA, 2001) The data presented are the means with standard errors of the results of experiments performed in triplicate

Electron microscopy

paraformaldehyde and 2.5% glutaraldehyde in 0.1 M phosphate buffer (PB; pH 7.2) at room temperature for 2 hr For scanning electron microscopy (SEM), the fixed samples were then rinsed with PB and post-fixed with 1% osmium in 0.1M PB (pH 7.2) at 4°C for 2 hr After dehydration in a graded concentration of ethanol, immersion in t-butyl alcohol and drying using a freeze drier (ES-2030, Hitachi, Tokyo, Japan), the samples were sputter-coated with gold-palladium and examined in a Hitachi S-2460N scanning electron microscope operated at 10 kV For transmission electron microscopy (TEM), the pre-fixed samples were fixed with 1% osmium in 0.1M PB (pH 7.2) at 4°C for 2 hr After being rinsed with distilled water, the pellets of the fixed cells were embedded in 1% agar, dehydrated using an ethanol series and embedded in Epon 812 Ultra-thin sections were doubly stained with uranyl acetate and lead nitrate and then examined with a JEM-1200 EX transmission electron microscope (JEOL, Tokyo, Japan) operated

at 80 kV Five SEM photographs were taken for each E coli cells before and after

SC-CO2 treatment, ten cells were randomly selected from SEM photographs, and the length and width of cells were measured The data presented was the means with standard deviation of five replications and subjected to the least significant differences

(LSD) test (P<0.01)

Measurement of enzyme activity

Enzyme activities in E coli cells before and after the SC-CO2 treatment were measured with an APIZYM kit (BioMerieux, Marcy-l’Etoile, France) This kit permits monitoring

of 20 different constitutive enzyme activities from a complex sample that had not been

purified (Ballestra et al., 1996) The number of E coli cells to measure the enzyme

activities was approximately 5.0×108 CFU/ml All enzyme activities were visually measured based on color changes, which fit with the increasing quality of hydrolyzed substrate (0-40 nmoles) The data presented were expressed as activity marks (0-5) which corresponded to hydrolyzed substrate of 0-40 nmoles and are the means with

standard errors of the results of triplicate experiments

RESULTS AND DISCUSSION

Inactivation of microorganisms in untreated water by SC-CO 2 treatment

The effect of CO2/sample flow rate in SC-CO2 treatment on the inactivation of coliform bacteria in untreated water is shown in Figure 2 The coliform bacterial count in the untreated water decreased almost linearly with increasing CO2/sample flow rate in the SC-CO2 treatment In particular, coliform bacteria could not be detected at a

CO2/sample flow rate greater than 53% In our previous papers, we reported that the ability of inactivating microorganisms by SC-CO2 treatment depended on the dissolved

CO2 concentration, which increased concomitantly with increasing CO2/sample flow rate (Kobayashi et al., 2007a, b) Dissolved CO2 can easily diffuse into the bacterial cell due to increased membrane permeability and accumulates in the cytoplasmic interior, decreasing the intracellular pH Cell viability will seriously be impaired owing to a drop

of intracellular pH (Garcia-Gonzales et al., 2009) Therefore, it was considered that the increase of the CO2/sample flow rate induced more cellular penetration of CO2 These results suggested that the SC-CO2 treatment was very effective for inactivating coliform

The effect of temperature during the SC-CO2 treatment on the inactivation of total bacteria in untreated water is shown in Figure 3 The total bacterial count in the untreated water dropped rapidly at first stage and slowly at second stage in the SC-CO2

treatment and the inactivation rate increased as the temperature increased The survival curves described by a rapid-to-slow two stage kinetics were in agreement with the findings of some earlier studies (Ishikawa et al., 1995b; Louka et al., 1999; Liao et al., 2007) However, there were some previous reports in which the survival curves were described by a slow-to-rapid two stage kinetics (Ballestra et al., 1996; Oulé et al., 2006) The reported differences in inactivation kinetics may be due to the difference in the efficiency of contact between CO2 and the microorganisms (Zhang et al., 2006) Therefore, it was considered that our present result was obtained by the difference of the penetration of CO2 into the bacterial cells Since it is stated by law that the number of coliform and total bacteria must be kept less than 0 and 100 CFU/ml, respectively (JWWA, 2001), SC-CO2 treatment at the CO2/sample flow rate higher than 67% and exposure time longer than 13 min might become a sterilization method to produce drinking water

3

1

2

0

20 30

10

* *

Figure 2 - Effect of CO2/sample flow rate in SC-CO2 treatment on the inactivation of coliform bacteria in untreated water SC-CO2 treatment conditions: temperature 35°C, pressure 10 MPa, CO2 flow rate 4.0, 6.0, 8.0 and 10 g/min, sample flow rate 15 ml/min,

CO2/sample flow rate 27, 40, 53 and 67%, exposure time 13 min * not detected

1

3 2

0 Exposure time (min)

Figure 3 - Effect of temperature during the SC-CO2 treatment on the inactivation of total bacteria in untreated water SC-CO2 treatment conditions: temperature 35(◆), 45(■) and 55°C(▲), pressure 10 MPa, CO2 flow rate 10 g/min, sample flow rate 2.5, 5.0 and

15 ml/min, CO2/sample flow rate 67, 200 and 400%, exposure time 13, 40 and 80 min

Electron micrographs of E coli cells before and after SC-CO2 treatment

SEM photographs of E coli cells before and after SC-CO2 treatment are shown in

Figure 4 In an earlier study we confirmed that surviving E coli cells could not be

detected after SC-CO2 treatment (Kobayashi et al., 2007a) The damage in E coli cells

by SC-CO2 treatment could not be observed by SEM Ballestra et al., (1996) found that

25% of E coli cells treated by pressurized CO2 were intact, but the cell viability was

only 1% Since their reports agreed with our present result, we suspect that the E coli

cells treated with SC-CO2 were not necessarily killed by cellular membrane damages

However, the lengths of E coli cells treated with SC-CO2 were shortened by 18% (from 1.067 to 0.876 μm), while the widths were hardly changed from 0.583 to 0.561 μm (Figure 5)

TEM photographs of E coli cells before and after SC-CO2 treatment are shown in Figure 6 On the observation with TEM, the cell membranes might be slightly damaged

by the SC-CO2 treatment We also found during TEM observation that cytoplasm with

low electronic density in E coli cells subjected to the SC-CO2 treatment had disappeared Hong and Pyun (1999, 2001) found that a possible leakage of cytoplasm

was caused by the modification of the cell membrane of L plantarum by pressurized

CO2 In our present results, the leakage of cytoplasm could not be confirmed, although the leakage is somehow conveyed by our observations Thus, it was considered that the

shortening of E coli cell indicated by SEM photograph might show the leakage of cytoplasm caused by damaging the cell membrane, when E coli was killed by the

SC-CO2 treatment

Figure 4 - SEM photographs of E coli cells before (A) and after (B) SC-CO2 treatment SC-CO2 treatment conditions: temperature 35°C, pressure 10 MPa, CO2 flow rate 10 g/min, sample flow rate 15 ml/min, CO2/sample flow rate 67%, exposure time 13 min

Length Width

0 0.2 0.4 0.6 0.8 1.0

1.2

*

Figure 5 - The size of E coli cells before (■) and after (■) SC-CO2 treatment SC-CO2

treatment conditions: temperature 35°C, pressure 10 MPa, CO2 flow rate 10 g/min, sample flow rate 15 ml/min, CO2/sample flow rate 67%, exposure time 13 min Five

SEM photographs were taken for each E coli cells before and after SC-CO2 treatment, ten cells were randomly selected from SEM photographs, and the length and width of cells were measured The data presented was the means with standard deviation of five

replications Asterisk (*) indicated significant difference between the size of E coli cells

before and after SC-CO2 treatment by the least significant differences (LSD) test (P<0.01)

Figure 6 - TEM photographs of E coli cells before (A) and after (B) SC-CO2 treatment SC-CO2 treatment conditions: temperature 35°C, pressure 10 MPa, CO2 flow rate 10 g/min, sample flow rate 15 ml/min, CO2/sample flow rate 67%, exposure time 13 min

Enzyme activities in E coli cells before and after the SC-CO2 treatment

To investigate the effect of the SC-CO2 treatment on the cytoplasm in E coli cells (Figure 7), we measured seven types of metabolic enzyme activities in E coli cells

before and after the SC-CO2 treatment with an APIZYM kit Four types of metabolic enzymes, Leucine arylamidase, β-galactosidase, β-glucuronidase and α-glucosidase, were completely inactivated by the SC-CO2 treatment, whereas phosphatase (alkaline and acid) or naphthol-AS-BI-phosphohydrolase was slightly or a little inactivated Ishikawa et al (1996) reported that enzyme inactivation by the SC-CO2 treatment was caused by irreversible destruction of the α-helix structure of protein enzyme Ballestra et

al (1996) suggested that selective inactivation of enzymes could result from a drop of internal pH of the microorganisms during the pressurized CO2 treatment The dissociation of dissolved CO2, present in large amounts in the cells, would lower the internal pH of the bacteria to a critical level which would induce inhibition of essential metabolic systems Thus, we considered that these phenomena observed by SEM and TEM might have been caused by the denaturation of intracellular protein

0

1

2

3

4

5

A B C D E F G

Figure 7 - Enzyme activities in E coli cells before (■) and after (■) SC-CO2 treatment Alkaline phosphatase (A), Leucine arylamidase (B), Acid phosphatase (C), Naphthol-AS-BI-phosphhydrolase (D), β-galactosidase (E), β-glucuronidase (F), α-glucosidase (G), SC-CO2 treatment conditions: temperature 35°C, pressure 10 MPa,

CO2 flow rate 10 g/min, sample flow rate 15 ml/min, CO2/sample flow rate 67%,

CONCLUSION

In the present study, SC-CO2 treatment was effective for inactivating microorganisms in untreated water In particular, coliform bacteria were reduced to nondetectable levels by the SC-CO2 treatment We therefore propose that SC-CO2 treatment might be a feasible

way to produce safe drinking water Also, although SEM and TEM observations of E

observed that SC-CO2 treatment induced cell shortening and the disappearance of cytoplasm with low electronic density These phenomena might have been caused by the leakage of cytoplasm from cells or the denaturation of intracellular protein

ACKNOWLEDGMENT

We wish to thank Kunio Tatewaki and the staff of the Public Health Center (Kure,

Hiroshima, Japan) for their technical assistance

REFERENCES

Backer, H and Hollowell, J (2000) Use of iodine for water disinfection: Iodine toxicity

and maximum recommended dose Environ Health Perspectives 108(8), 679-684

Ballestra, P., A.Abreu Da Silva and Cuq, J.L (1996) Inactivation of Escherichia coli by

carbon dioxide under pressure J Food Sci., 61(4), 829-831, 836

Cho, M., Chung, H and Yoon, J (2003) Disinfection of water containing natural

organic matter by using ozone-initiated radical reactions Appl Environ Microbiol.,

69, 2284-2291

Garcia-Gonzalez, L., Geeraerd, A H., Elst, K., Van Ginneken, L., Van Impe, J F and Devlieghere, F (2009) Influence of type of microorganism, food ingredients and

food properties on high-pressure carbon dioxide inactivation of microorganisms Int

J Food Microbiol 129, 253-263

Graham, N J D., Reynolds, G., Buckley, D., Perry, R and Croll, B (1989) Laboratory

simulation of disinfection regimes for trihalomethane control J IWEM., 3, 604-611

Gunes, G., Blum, L.K and Hotchkiss, J.H (2006) Inactivation of Escherichia coli (ATCC 4157) in diluted apple cider by dence-phase carbon dioxide J Food Prot.,

69(1), 12-16

Hong, S I and Pyun, Y R (1999) Inactivation kinetics of Lactobacillus plantarum by

high pressure carbon dioxide J Food Sci., 64(4), 728-733

Hong, S I and Pyun, Y R (2001) Membrane damage and enzyme inactivation of

Lactobacillus plantarum by high pressure CO2 treatment Int J Food Microbiol., 63,

19-28

Ishikawa, H., Shimoda, M., Kawano, T and Osajima, Y (1995a) Inactivation of enzymes in an aqueous solution by micro-bubbles of supercritical carbon dioxide

Biosci Biotechnol Biochem., 59(4), 628-631

Ishikawa, H., Shimoda, M., Shiratsuchi, H and Osajima, Y (1995b) Sterilization of

microorganisms by the supercritical carbon dioxide micro-bubble method Biosci

Biotechnol Biochem., 59(10), 1949-1950

Ishikawa, H., Shimoda, M., Yonekura, A and Osajima, Y (1996) Inactivation of enzymes and decomposition of α-helix structure by supercritical carbon dioxide

microbubble method J Agric Food Chem., 44, 2646-2649

Japan Water Works Association (JWWA) (2001) The method of Japan Water Works Association p.573-613 (in Japanese)

Kobayashi, F., Hayata, Y., Kohara, K., Muto, N., Miyake, M and Osajima, Y (2006) Application of supercritical CO2 bubbling to deodorizing of drinking water Food Sci

Technol Res., 12(2), 119-124

Kobayashi, F., Hayata, Y., Kohara, K., Muto, N and Osajima, Y (2007a) Application of supercritical CO2 bubbling to inactivate E.coli and coliform bacteria in drinking water

Food Sci Technol Res., 13(1), 20-22

Kobayashi, F., Hayata, Y., Muto, N and Osajima Y (2007b) Effect of the pore size of microfilters in supercritical CO2 bubbling on the dissolved CO2 concentration Food

Sci Technol Res 13(2), 118-120

Liao, H., Hu, X., Liao, X., Chen, F and Wu, J (2007) Inactivation of Escherichia coli inoculated into cloudy apple juice exposed to dense phase carbon dioxide Int J

Food Microbiol 118, 126-131

Lin, H M., Yang, Z and Chen, L F (1992) Inactivation of Saccharomyces cerevisiae

by supercritical carbon dioxide Biotechnol Prog., 8, 458-461

Lin, H M., Yang, Z and Chen, L F (1993) Inactivation of Leuconostoc dextranicum

with carbon dioxide under pressure Chem Engin J., 52, B29-B34

Louka, E D., Louka, N., Abraham, G., Chabot, V and Allaf, K (1999) Effect of

compressed carbon dioxide on microbial cell viability Appl Environ Microbiol., 65,

626-631

Oulé, M K., K Tano, A M Bernier, and J Arul (2006) Escherichia coli inactivation

mechanism by pressurized CO2 Can J Microbiol 52, 1208-1217

Satsuta, K., Miyazaki, M and Yoshimi, R (2001) A bacteriological study on the safety

of food and drink Bull Tokyo kasei-gakuin univ., 41, 15-20 (in Japanese)

Shimoda, M., Yamamoto, Y., Cocunubo-Castellanos, J., Tonoike, H., Kawano, T., Ishikawa, H and osajima, Y (1998) Antimicrobial effects of pressured carbon

dioxide in a continuous flow system J Food Sci., 63(4), 709-712

Shimoda, M., Cocunubo-Castellanos, J., Kago,H., Miyake, M., Osajima, Y and Hayakawa, I (2001) The influence of dissolved CO2 concentration on the death

kinetics of Saccharomyces cerevisiae J Appl Microbiol., 91, 306-311

Shimoda, M., Kago, H., Kojima, N., Miyake, M., Osajima, Y and Hayakawa, I (2002)

Accelerated death kinetics of Aspergillus niger spores under high-pressure

carbonation Appl Environ Microbiol., 68(8), 4162-4167

Warburton, D W (1993) A reviews of microbiological quality of bottled water sold in

Canada Part 2 The need for more stringest standards and regulations Can J

Microbiol., 39, 158-168

Wist, J., Sanabria, J., Dierolf, C., Torres, W and Pulgarin, C (2002) Evaluation of

photocatalytic disinfection of crude water for drinking-water production J

Photochem Photobiol., A: Chemistry 147, 241-246

Zhang, J., David, T.A., Matthews, M.A., Drewsb, M.J., LaBerge, M., An, Y.H (2006)

Sterilization using high-pressure carbon dioxide J Supercritical Fluids, 38, 354-372