IMPACT OF ORGANIC LOAD ON SANITIZER EFFICACY AGAINST ESCHERICHIA
COLI O157:H7 DURING PILOT-PLANT PRODUCTION OF FRESH-CUT LETTUCE
By
Gordon Ray Davidson

A DISSERTATION
Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of
Food Science – Doctor of Philosophy
2013

ABSTRACT
IMPACT OF ORGANIC LOAD ON SANITIZER EFFICACY AGAINST ESCHERICHIA
COLI O157:H7 DURING PILOT-PLANT PRODUCTION OF FRESH-CUT LETTUCE
By
Gordon Ray Davidson
Addition of chemical sanitizers during commercial flume washing of leafy greens
remains the sole microbial mitigation strategy. However, continued Escherichia coli O157:H7
outbreaks have raised concerns regarding recirculation of wash water, with the accumulation of
organic load during processing leading to decreased efficacy of chlorine-based sanitizers.
Reliable methods to quantify the impact of organic load on sanitizing efficacy do not yet exist.
Initially, the efficacy of six different wash treatments (water alone, 50 ppm peroxyacetic
acid, 50 ppm mixed peracid, or 50 ppm available chlorine either alone or acidified to pH 6.5 with
6

citric acid (CA) or T-128) was assessed using 5.4 kg of iceberg lettuce inoculated to contain 10

CFU/g of a 4-strain non-toxigenic, GFP-labeled, ampicillin-resistant cocktail of E. coli O157:H7
in a pilot-scale leafy green processing line consisting of a commercial shredder, conveyor, flume
tank, shaker table, and centrifugal dryer. Without an organic load in the water, none of the
sanitizers were more effective (P ≤ 0.05) than water alone at reducing E. coli O157:H7
populations on lettuce, with reductions ranging from 0.8 to 1.4 log CFU/g. However, chlorine,
chlorine + CA, and chlorine + T-128 were generally more effective (P ≤ 0.05) than the other
treatments against E. coli O157:H7 in the flume water, with reductions of 3.8, 5.5, and 5.4 log
CFU/ml after 90 s of processing, respectively.
Thereafter, a novel and cost-effective carboy system was developed to assess the efficacy
of the same five sanitizing agents against E. coli O157:H7 in wash water containing an organic

load of 0 to 10% (w/v) blended lettuce. After iceberg lettuce previously inoculated to contain E.
6

coli O157:H7 at 10 CFU/g was washed for 90 s, E. coli O157:H7 persistence was subsequently
correlated to various physicochemical parameters of the wash water. Organic load negatively
impacted the efficacy of chlorine, chlorine + CA, and chlorine + T-128 (P ≤ 0.05), with typical
E. coli O157:H7 reductions of < 1 log CFU/ml after 10 min of exposure. However, the efficacy
of peroxyacetic acid and mixed peracid was unaffected by organic load (P > 0.05), with average
E. coli O157:H7 reductions of ~4.8 and ~5.5 log CFU/ml, respectively, after 10 min of exposure.
Finally, efficacy of the same five sanitizer treatments was assessed against E. coli
O157:H7 on iceberg lettuce, in wash water, and on surfaces of a pilot-scale processing line using
flume water containing an organic load of 0 to 10% (w/v) blended lettuce. Organic load
negatively impacted the efficacy of all three chlorine treatments (P ≤ 0.05), with typical E. coli
O157:H7 reductions of > 5 log CFU/ml by the end of processing with no organic load in the
wash water and 0.9 – 3.7 log CFU/ml with a 10% organic load. Organic load rarely impacted (P
> 0.05) the efficacy of either peroxyacetic acid or mixed peracid, with typical reductions of > 5
log CFU/ml in wash water throughout processing for all organic loads. Sanitizer efficacy against
E. coli O157:H7on lettuce was seldom impacted by organic load. In both the carboy system and
the pilot-scale processing line, reduced sanitizer efficacy generally correlated to increases (P ≤
0.05) in total solids, chemical oxygen demand and turbidity, and decreases (P ≤ 0.05) in
maximum filterable volume, indicating that these tests may be effective alternatives to the
industry standard of oxygen/reduction potential. These findings demonstrate that monitoring of
both sanitizer concentration and wash water quality is critical to minimizing the likelihood of
amplifying a previously isolated contamination event.

To my parents, Brad and Janet Davidson

iv

ACKNOWLEDGEMENTS

First and foremost, I would like to thank my advisor, Dr. Elliot Ryser. I will be forever
grateful for the amount of guidance, trust, understanding, and responsibility he gave me over our
5+ years together. He is an excellent role model and I can’t thank him enough for what he has
taught me. I would also like to thank my committee members- Dr. John Linz, Dr. Bradley Marks,
and Dr. Ewen Todd. All of them have been exceedingly constructive and helpful. I am very
thankful for their guidance and for challenging me.
I would not be where I am today without Dr. Zhinong Yan- he taught me the importance
of a collaborative and passionate work environment. I cannot thank my undergraduate assistant,
Chelsea Kaminski, enough. A lot of this work would not have been possible without her
assistance and I am so grateful that she put up with me over the years. I would also like to thank
my old lab mates Annemarie Buchholz, Scott Moosekian, and Haiqiang Wang- I am proud to
refer to all of them as good friends. I am very grateful to Paul Sirmeyer, Rudy Sloup, Matt
Steele, Wenting Zeng, Lin Ren, Dr. Lei Zhang, Dr. Yinfa Zhang, and Dr. Yanyang Xu for their
assistance.
None of this work would have been possible without the financial support of the United
States Department of Agriculture and the Center for Produce Safety at UC Davis.
Lastly, I would like to thank my family for all their love and support. My parents taught
me the importance of a strong work ethic and have always been there to assure me that I was
doing the right thing. I am very proud of the accomplishments of my sister and brother, Gretchen
and Connor, even though they chose to go to the “other school” in Ann Arbor. I know they are
destined for great things.
iv

TABLE OF CONTENTS

LIST OF TABLES

vii

LIST OF FIGURES

xi

KEY SYMBOLS AND ABBREVIATIONS

xvii

INTRODUCTION

1

CHAPTER 1: Review of Pertinent Literature
1.1 The leafy greens industry
1.2 Leafy green associated recalls and outbreaks
1.3 Escherichia coli O157:H7
1.4 Pre-harvest contamination and control
1.5 Post-harvest processing
1.6 Purpose of sanitizer addition to wash water
1.7 Sodium hypochlorite use in wash water
1.8 Peroxyacetic acid use in wash water
1.9 Organic load accumulation in wash water
1.10 Strategies for monitoring wash water efficacy
1.11 Current challenges

4
5
6
13
14
16
18
21
24
26
27
30

CHAPTER 2: Efficacy of Commercial Produce Sanitizers against Escherichia coli
O157:H7 during Processing of Iceberg Lettuce in a Pilot-Scale Leafy Green
Processing Line
2.1 ABSTRACT
2.2 INTRODUCTION
2.3 MATERIALS AND METHODS
2.3.1 Experimental design
2.3.2 Leafy greens
2.3.3 Bacterial strains
2.3.4 Lettuce inoculation
2.3.5 Lettuce processing line
2.3.6 Wash water
2.3.7 Leafy green processing
2.3.8 Sample collection
2.3.9 Microbiological analyses
2.3.10 Sanitizer neutralization confirmation
2.3.11 Statistical analysis
2.4 RESULTS
2.4.1 Lettuce
2.4.2 Flume water
2.4.3 Centrifugation water

32
33
35
38
38
38
38
39
39
40
40
41
45
45
46
47
47
49
51

v

2.4.4 Processing equipment surfaces
2.5 DISCUSSION

53
55

CHAPTER 3: Impact of Organic Load on Sanitizer Efficacy against Escherichia
coli O157:H7 in Simulated Leafy Green Processing Water
3.1 ABSTRACT
3.2 INTRODUCTION
3.3 MATERIALS AND METHODS
3.3.1 Experimental design
3.3.2 Leafy greens
3.3.3 Bacterial strains
3.3.4 Lettuce inoculation
3.3.5 Processing equipment
3.3.6 Wash water
3.3.7 Leafy green washing and sample collection
3.3.8 Physicochemical parameters
3.3.9 Microbiological analyses
3.3.10 Sanitizer neutralization confirmation
3.3.11 Statistical analysis
3.4 RESULTS
3.4.1 Lettuce
3.4.2 Wash water
3.4.3 Impact of temperature on sanitizer efficacy
3.4.4 Linear regression of E. coli O157:H7 reduction in wash water
3.4.5 Physicochemical parameters of wash water
3.5 DISCUSSION

59
60
61
62
62
62
62
62
63
65
65
66
66
67
67
69
69
71
75
77
86
91

CHAPTER 4: Impact of Organic Load on Escherichia coli O157:H7 Survival
during Pilot-Scale Processing of Iceberg Lettuce with Commercial Produce
Sanitizers
4.1 ABSTRACT
4.2 INTRODUCTION
4.3 MATERIALS AND METHODS
4.3.1 Experimental design
4.3.2 Leafy greens
4.3.3 Bacterial strains
4.3.4 Lettuce inoculation
4.3.5 Lettuce processing line
4.3.6 Wash water
4.3.7 Lettuce processing
4.3.8 Sample collection
4.3.9 Physicochemical parameters of flume water
4.3.10 Microbiological analyses
4.3.11 Sanitizer neutralization confirmation
4.3.12 Physicochemical parameters of commercial flume water

97
98
100
101
101
101
101
101
102
103
103
104
104
105
106
106

vi

4.3.13 Statistical analyses
4.4 RESULTS
4.4.1 Lettuce
4.4.2 Flume water
4.4.3 Centrifugation water
4.4.4 Processing equipment surfaces
4.4.5 Physicochemical parameters of flume water
4.4.6 Physicochemical parameters of commercial flume water
4.5 DISCUSSION

106
108
108
115
137
144
151
157
160

CONCLUSIONS

169

FUTURE RESEARCH RECOMMENDATIONS

173

APPENDICES
APPENDIX I: Impact of Organic Load and Sanitizer Concentration on
Inactivation of a Four-Strain Cocktail of Escherichia coli O157:H7 in Simulated
Leafy Green Processing Water
APPENDIX II: Efficacy of Multiple Chlorine Concentrations Acidified with T128 against Escherichia coli O157:H7 during Pilot-Scale Processing of Iceberg
Lettuce Using Water Containing an Organic Load
APPENDIX III: Impact of In-Line Equipment Sanitation Process on
Escherichia coli O157:H7 Persistence during Pilot-Scale Processing of Iceberg
Lettuce Using Flume Water Containing Sodium Hypochlorite and Organic
Load

178

REFERENCES

242

vii

179

205

223

LIST OF TABLES

Table 1.1:

Select leafy greens associated outbreaks from 1995 through 2012

10

Table 1.2:

Commercial produce sanitizers used during flume washing of leafy
greens

20

Table 3.1:

E. coli O157:H7 reduction on iceberg lettuce inoculated at ~6 log
CFU/g after processing

70

Table 3.2:

E. coli O157:H7 reduction in the wash water at ~14°C

73

Table 3.3:

E. coli O157:H7 reduction in the wash water containing chlorine +
CA at two temperatures

76

Table 3.4:

E. coli inactivation trend lines in wash water containing various
organic loads

84

Table 3.5:

Physicochemical parameters of wash water containing various organic
loads and 50 ppm peroxyacetic acid or 50 ppm mixed peracid

87

Table 3.6:

Physicochemical parameters of wash water containing various organic
loads and 50 ppm available chlorine

88

Table 3.7:

Physicochemical parameters of wash water containing various organic
loads and 50 ppm chlorine + CA or 50 ppm chlorine + T-128, both
used to acidify to pH 6.5

89

Table 3.8:

Physicochemical parameters of wash water containing various organic
loads and 50 ppm available chlorine + CA at ~5°C

90

Table 4.1:

E. coli O157:H7 reductions in flume water containing chlorine,
chlorine + CA and chlorine + T-128

122

Table 4.2:

E. coli O157:H7 reduction in flume water containing peroxyacetic
acid and mixed peracid

123

Table 4.3:

E. coli O157:H7 inactivation trend lines for flume water containing
various organic loads and chlorine, chlorine + CA and chlorine + T128

124

viii

Table 4.4:

E. coli O157:H7 inactivation trend lines for flume water containing
various organic loads and chlorine, chlorine + CA and chlorine + T128 (0 - 10 min)

126

Table 4.5:

E. coli O157:H7 inactivation trend lines for flume water containing
various organic loads and chlorine, chlorine + CA and chlorine + T128 (11.5 - 21 min)

128

Table 4.6:

E. coli O157:H7 inactivation trend lines for flume water containing
various organic loads and chlorine, chlorine + CA and chlorine + T128 (23 - 32.5 min)

130

Table 4.7:

E. coli O157:H7 inactivation trend lines for flume water containing
various organic loads and chlorine, chlorine + CA and chlorine + T128 (34.5 - 44 min)

132

Table 4.8:

E. coli O157:H7 inactivation trend lines in flume water containing
organic loads and peroxyacetic acid and mixed peracid

134

Table 4.9:

Physicochemical parameters of flume water containing chlorine and
various organic loads

153

Table 4.10:

Physicochemical parameters of flume water containing chlorine + CA
or chlorine + T-128 with various organic loads

154

Table 4.11

Physicochemical parameters of flume water containing organic load
and 50 ppm peroxyacetic acid or 50 ppm mixed peracid

155

Table 4.12:

Physicochemical parameters of commercial flume water containing
chlorine + CA

158

Table 4.13:

Physicochemical parameters of commercial flume water containing
peroxyacetic acid

159

Table AI.1:

E. coli O157:H7 populations in wash water controls

181

Table AI.2:

E. coli O157:H7 populations in wash water containing peroxyacetic
acid

182

Table AI.3:

E. coli O157:H7 populations in wash water containing mixed peracid

185

Table AI.4:

E. coli O157:H7 populations in wash water containing chlorine

188

Table AI.5:

Physicochemical parameters of wash water controls containing
various organic loads

191

ix

Table AI.6:

Physicochemical parameters of wash water containing various organic
loads and peroxyacetic acid concentrations

192

Table AI.7:

Physicochemical parameters of wash water containing various organic
loads and mixed peracid concentrations

194

Table AI.8:

Physicochemical parameters of wash water containing various organic
loads and chlorine concentrations

196

Table AII.1:

E. coli O157:H7 reductions in flume water

211

Table AII.2:

E. coli O157:H7 inactivation trend lines for flume water containing
various chlorine concentrations

212

Table AII.3:

Physicochemical parameters of flume water containing organic load
and various chlorine concentrations acidified with T-128

213

Table AIII.1: E. coli O157:H7 reductions in flume water

229

Table AIII.2: E. coli O157:H7 inactivation trend lines for flume water after
different processing conditions

230

Table AIII.3: Physicochemical parameters of flume water 2.5% containing organic
load and 50 ppm chlorine during normal operation and with a
sanitized shredder and conveyor

231

x

LIST OF FIGURES

Figure 1.1:

Produce outbreaks attributed to E. coli O157:H7 from 1998 – 2006

7

Figure 1.2:

Produce-linked outbreaks between 1998 and 2006, ranked according to
commodity

8

Figure 2.1:

Flume tank sampling locations

42

Figure 2.2:

Shaker table sampling locations

43

Figure 2.3:

Dewatering centrifuge sampling locations

44

Figure 2.4:

Mean (± SD) E. coli O157:H7 populations on the iceberg lettuce
inoculated at ~6 log CFU/g during and after processing (n=3). Means of
the same wash water treatment with different letters are significantly
different (P ≤ 0.05).

48

Figure 2.5:

Mean (± SD) E. coli O157:H7 populations in flume water during
processing iceberg lettuce inoculated at ~6 log CFU/g (n=3). Half the limit
of detection was used when a sample did not yield any colonies by direct
plating. Means of the same product type with different letters are
significantly different (P ≤ 0.05).

50

Figure 2.6:

Mean (± SD) E. coli O157:H7 populations in spent centrifugation water
from iceberg lettuce inoculated at ~6 log CFU/g (n=3). Half the limit of
detection was used when a sample did not yield any colonies by direct
plating. Means of the same product type with different letters are
significantly different (P ≤ 0.05).

52

Figure 2.7:

Mean (± SD) E. coli O157:H7 populations on equipment surfaces after
processing iceberg lettuce inoculated at ~6 log CFU/g (n=3). Half the limit
of detection was used when a sample did not yield any colonies by direct
plating. Means of the same product type with different letters are
significantly different (P ≤ 0.05).

54

Figure 3.1:

Bench-top system developed to simulate commercial flume washing of
iceberg lettuce.

64

Figure 3.2:

Mean (± SD) E. coli O157:H7 populations (log CFU/ml) in water
containing various organic loads (n=3 per organic load) after washing
iceberg lettuce inoculated at ~6 log CFU/g. Means of the same product
type with different letters are significantly different (P ≤ 0.05).

72

xi

Figure 3.3:

E. coli O157:H7 inactivation (log CFU/ml) in water containing various
organic loads and 50 ppm peroxyacetic acid (n=3 per organic load) after
washing iceberg lettuce inoculated at ~6 log CFU/g. Half the limit of
detection was used when a sample did not yield any colonies by direct
plating.

78

Figure 3.4:

E. coli O157:H7 inactivation (log CFU/ml) in water containing various
organic loads and 50 ppm mixed peracid (n=3 per organic load) after
washing iceberg lettuce inoculated at ~6 log CFU/g. Half the limit of
detection was used when a sample did not yield any colonies by direct
plating.

79

Figure 3.5:

E. coli O157:H7 inactivation (log CFU/ml) in water containing various
organic loads and 50 ppm available chlorine (n=3 per organic load) after
washing iceberg lettuce inoculated at ~6 log CFU/g. Half the limit of
detection was used when a sample did not yield any colonies by direct
plating.

80

Figure 3.6:

E. coli O157:H7 inactivation (log CFU/ml) in water containing various
organic loads and 50 ppm available chlorine + CA (n=3 per organic load)
after washing iceberg lettuce inoculated at ~6 log CFU/g. Half the limit of
detection was used when a sample did not yield any colonies by direct
plating.

81

Figure 3.7:

E. coli O157:H7 inactivation (log CFU/ml) in water containing various
organic loads and 50 ppm available chlorine + T-128 (n=3 per organic
load) after washing iceberg lettuce inoculated at ~6 log CFU/g. Half the
limit of detection was used when a sample did not yield any colonies by
direct plating.

82

Figure 3.8:

E. coli O157:H7 inactivation (log CFU/ml) in 5°C water containing
various organic loads and 50 ppm available chlorine + CA (n=3 per
organic load) after washing iceberg lettuce inoculated at ~6 log CFU/g.
Half the limit of detection was used when a sample did not yield any
colonies by direct plating.

83

Figure 4.1:

Mean (± SD) E. coli O157:H7 populations on lettuce after flume washing
in water containing 50 ppm chlorine and various organic loads (n=3 per
organic load). Means within the same batch with different letters are
significantly different (P ≤ 0.05).

110

Figure 4.2:

Mean (± SD) E. coli O157:H7 populations on lettuce after flume washing
in water containing 50 ppm chlorine + CA and various organic loads (n=3
per organic load). Means within the same batch with different letters are
significantly different (P ≤ 0.05).

111

xii

Figure 4.3:

Mean (± SD) E. coli O157:H7 populations on lettuce after flume washing
in water containing 50 ppm chlorine + T-128 and various organic loads
(n=3 per organic load). Means within the same batch with different letters
are significantly different (P ≤ 0.05).

112

Figure 4.4:

Mean (± SD) E. coli O157:H7 populations on lettuce after processing in
flume water containing various organic loads and 50 ppm peroxyacetic
acid (n=3 per organic load). Means within the same batch with different
letters are significantly different (P ≤ 0.05).

113

Figure 4.5

Mean (± SD) E. coli O157:H7 populations on lettuce after processing in
flume water containing various organic loads and 50 ppm mixed peracid
(n=3 per organic load). Means within the same batch with different letters
are significantly different (P ≤ 0.05).

114

Figure 4.6:

Mean (± SD) E. coli O157:H7 populations in recirculating sanitizer-free
flume water containing various organic loads (n=3 per organic load).

118

Figure 4.7:

E. coli O157:H7 inactivation in recirculating flume water containing
various organic loads and chlorine compared to the sanitizer-free control
(n=3 per organic load).

119

Figure 4.8:

E. coli O157:H7 reductions in recirculating flume water containing
various organic loads and chlorine + CA compared to the sanitizer-free
control (n=3 per organic load).

120

Figure 4.9:

E. coli O157:H7 reductions in recirculating flume water containing
various organic loads and chlorine + T-128 compared to the sanitizer-free
control (n=3 per organic load).

121

Figure 4.10:

E. coli O157:H7 inactivation in recirculating flume water containing
various organic loads and peroxyacetic acid compared to the sanitizer-free
control (n=3 per organic load).

135

Figure 4.11:

E. coli O157:H7 inactivation in recirculating flume water containing
various organic loads and mixed peracid compared to the sanitizer-free
control (n=3 per organic load).

136

Figure 4.12:

Mean (± SD) E. coli O157:H7 populations in spent centrifugation water
from lettuce after washing in flume water containing chlorine and various
organic loads (n=3 per organic load). Means within the same batch with
different letters are significantly different (P ≤ 0.05). Limit of detection =
0.02 CFU/ml.

139

xiii

Figure 4.13:

Mean (± SD) E. coli O157:H7 populations in spent centrifugation water
from lettuce after washing in flume water containing chlorine + CA and
various organic loads (n=3 per organic load). Means within the same batch
with different letters are significantly different (P ≤ 0.05). Limit of
detection = 0.02 CFU/ml.

140

Figure 4.14:

Mean (± SD) E. coli O157:H7 populations in spent centrifugation water
from lettuce after washing in flume water containing chlorine + T-128 and
various organic loads (n=3 per organic load). Means within the same batch
with different letters are significantly different (P ≤ 0.05). Limit of
detection = 0.02 CFU/ml.

141

Figure 4.15:

Mean (± SD) E. coli O157:H7 populations in spent centrifugation water
from lettuce after washing in flume water containing peroxyacetic acid
and various organic loads (n=3 per organic load). Means within the same
batch with different letters are significantly different (P ≤ 0.05). Limit of
detection = 0.02 CFU/ml.

142

Figure 4.16:

Mean (± SD) E. coli O157:H7 populations in spent centrifugation water
from lettuce after washing in flume water containing mixed peracid and
various organic loads (n=3 per organic load). Means within the same batch
with different letters are significantly different (P ≤ 0.05). Limit of
detection = 0.02 CFU/ml.

143

Figure 4.17:

Mean (± SD) E. coli O157:H7 populations on equipment surfaces after
processing lettuce in flume water containing chlorine and various organic
loads (n=3 per organic load). Means within the same equipment with
different letters are significantly different (P ≤ 0.05). Limit of detection =
2
1 CFU/100 cm .

146

Figure 4.18:

Mean (± SD) E. coli O157:H7 populations on equipment surfaces after
processing lettuce in flume water containing chlorine + CA and various
organic loads (n=3 per organic load). Means within the same equipment
with different letters are significantly different (P ≤ 0.05). Limit of
2
detection = 1 CFU/100 cm .

147

Figure 4.19:

Mean (± SD) E. coli O157:H7 populations on equipment surfaces after
processing lettuce in flume water containing chlorine + T-128 and various
organic loads (n=3 per organic load). Means within the same equipment
with different letters are significantly different (P ≤ 0.05). Limit of
2
detection = 1 CFU/100 cm .

148

xiv

Figure 4.20:

Mean (± SD) E. coli O157:H7 populations on equipment surfaces after
processing lettuce in flume water containing peroxyacetic acid and various
organic loads (n=3 per organic load). Means within the same equipment
with different letters are significantly different (P ≤ 0.05). Limit of
2
detection = 1 CFU/100 cm .

149

Figure 4.21:

Mean (± SD) E. coli O157:H7 populations on equipment surfaces after
processing lettuce in flume water containing mixed peracid and various
organic loads (n=3 per organic load). Means within the same equipment
with different letters are significantly different (P ≤ 0.05). Limit of
2
detection = 1 CFU/100 cm .

150

Figure AII.1:

Mean (± SD) E. coli O157:H7 populations on lettuce after flume washing
in water containing 5% organic load (w/v blended lettuce) and chlorine
concentrations of 10, 30 and 100 ppm + T-128 (n=3 per chlorine
concentration). Means within the same batch with different letters are
significantly different (P ≤ 0.05).

207

Figure AII.2:

E. coli O157:H7 reductions in recirculating flume water containing 5%
organic load (w/v blended lettuce) and chlorine concentrations of 10, 30
and 100 ppm + T-128 compared to the sanitizer-free control (n=3 per
chlorine concentration).

208

Figure AII.3:

Mean (± SD) E. coli O157:H7 populations in spent centrifugation water
from lettuce after washing in flume water containing 5% organic load (w/v
blended lettuce) and chlorine concentrations of 10, 30 and 100 ppm + T128 (n=3 per chlorine concentration). Means within the same batch with
different letters are significantly different (P ≤ 0.05). Limit of detection =
0.02 CFU/ml.

209

Figure AII.4:

Mean (± SD) E. coli O157:H7 populations on equipment surfaces after
processing lettuce in flume water containing 5% organic load (w/v
blended lettuce) and chlorine concentrations of 10, 30 and 100 ppm + T128 (n=3 per chlorine concentration). Means within the same equipment
with different letters are significantly different (P ≤ 0.05). Limit of
2
detection = 1 CFU/100 cm .

210

Figure AIII.1: Mean (± SD) E. coli O157:H7 populations on lettuce after flume washing
in water containing 2.5% organic load (w/v blended lettuce) and 50 ppm
chlorine, with or without a sanitized shredder and conveyor (n=3 per
operational condition). Means within the same batch with different letters
are significantly different (P ≤ 0.05).

xv

225

Figure AIII.2: E. coli O157:H7 reductions in recirculating flume water containing 2.5%
organic load (w/v blended lettuce) and 50 ppm chlorine, with or without a
sanitized shredder and conveyor compared to the sanitizer-free control
(n=3 per operational condition).

226

Figure AIII.3: Mean (± SD) E. coli O157:H7 populations in spent centrifugation water
from lettuce after washing in flume water containing 2.5% organic load
(w/v blended lettuce) and 50 ppm chlorine, with or without a sanitized
shredder and conveyor (n=3 per operational condition). Means within the
same batch with different letters are significantly different (P ≤ 0.05).
Limit of detection = 0.02 CFU/ml.

227

Figure AIII.4: Mean (± SD) E. coli O157:H7 populations on equipment surfaces after
processing lettuce in flume water containing 2.5% organic load (w/v
blended lettuce) and 50 ppm chlorine, with or without a sanitized shredder
and conveyor (n=3 per operational condition). Means within the same
equipment with different letters are significantly different (P ≤ 0.05).
2
Limit of detection = 1 CFU/100 cm .

228

xvi

KEY TO SYMBOLS AND ABBREVIATIONS

ANOVA

Analysis of Variance

CDC

Centers for Disease Control and Prevention

CFU

Colony forming unit(s)

COD

Chemical oxygen demand

CPS

Center for Produce Safety

CSPI

Center for Science in the Public Interest

d

days(s)

EHEC

Enterohemorrhagic E. coli

FDA

Food and Drug Administration

FSMA

Food Safety Modernization Act

g

grams(s)

GFP

Green fluorescent protein

GRAS

Generally recognized as safe

h

hour(s)

HACCP

Hazard Analysis and Critical Control Point

HUS

Hemolytic uremic syndrome

min

minute(s)

MFV

Maximum filterable volume

ml

milliliter(s)

ORP

Oxidation/reduction potential

PBS

Phosphate Buffered Saline
xvii

PFGE

Pulsed-field Gel Electrophoresis

PMACS

Portable Multi-use Automated Concentration System

ppm

Parts per million

RMSE

Root-Mean-Square Error

s

second(s)

SAS

Statistical Analysis Systems

Stx

Shiga toxin

TSA-YE

Trypticase Soy Agar with 0.6 % Yeast Extract

TSB-YE

Trypticase Soy Broth with 0.6 % Yeast Extract

US

United States of America

USDA

United States Department of Agriculture

μl

microliter(s)

μm

micron(s)

xviii

INTRODUCTION

The Centers for Disease Control and Prevention (CDC) ranked leafy vegetables as the
leading food commodity for foodborne illnesses in the United States (22%) and the second most
frequent cause of hospitalizations (14%) between 1998 and 2008 (89). In 2009, leafy greens were
ranked as the riskiest food regulated by the Food and Drug Administration (FDA), accounting
for 363 outbreaks and 13,568 reported cases of illness (36). Between 1995 and 2006, leafy greenassociated outbreaks increased by 39% while consumption increased by only 9% (63). Leafy
greens were responsible for 363 separate outbreaks involving 13,568 individual cases of illness
through 2009 (36), with most of the E. coli outbreaks attributed to lettuce.
The widely-publicized nationwide outbreak of E. coli O157:H7 that was traced to baby
spinach in 2006 resulted in 205 confirmed infections, 103 hospitalizations, and three deaths (28,
38, 45). Two additional E. coli O157:H7 outbreaks in November and December of 2006 were
linked to shredded iceberg lettuce that was served at two different Mexican chain restaurants
(123, 124). These two outbreaks resulted in 150 illnesses (35, 123, 124).
Bacterial pathogens can contaminate fresh produce at any point during the farm-to-fork
continuum (83). Major on-farm areas of concern now recognized by the FDA include
agricultural water, biological soil amendments (e.g., manure), domesticated and wild animals,
field worker health and hygiene, and the cleanliness of harvesting equipment, tools and buildings
(127). Processing of leafy greens involves a series of steps that can either decrease or promote
the growth of pathogenic microorganisms. These steps include harvesting, cold storage,
trimming, shredding, washing and rinsing, draining, packaging, cold storage and distribution

1

(118). The sole microbial mitigation strategy during production of fresh-cut leafy greens is the
use of sanitizers during flume washing.
Commercial processing facilities recirculate flume water in order to reduce operational
costs (80). Sanitizing agents are routinely added to recirculating flume water as a means to
prevent the water from becoming a source of microbial contamination, thereby preventing the
spread of pathogens in large, centralized processing facilities. However, the efficacy of various
sanitizing agents has been questioned as product recalls and outbreaks have continued to occur.
Numerous bench-top studies have shown that produce sanitizers reduce pathogen populations
only 1 to 3 logs on lettuce (17, 47, 55, 91, 100), with water alone decreasing E. coli O157:H7
levels about 1 log on lettuce during pilot-scale processing (22).
Chlorine-based sanitizers are most commonly used by commercial processors to
minimize cross-contamination during processing, due to their relatively low cost compared to
other sanitizers, and minimal negative impact on end-product quality (31, 62, 81, 87, 91).
However, chlorine use has raised concerns regarding potentially hazardous by-products, worker
safety, environmental damage, and, most importantly, decreased efficacy in the presence of an
increasing organic load in recirculating flume water, which has heightened interest in other
alternatives, such as peroxyacetic acid-based sanitizers (100, 113).
An organic load, consisting of plant tissues and cellular fluids released during cutting, in
addition to soil, insects, and microbes (62), will accumulate in recirculating flume water as produce
is washed, decreasing the ability of sanitizers to minimize cross-contamination from the water during
processing (57, 110, 139). While organic load impacts the efficacy of sanitizing agents, most

notably chlorine, the means to quantify sanitizer efficacy against pathogens such as E. coli
O157:H7 have not yet been determined. The Center for Produce Safety currently ranks the
2

identification of methods to validate the efficacy of flume water used to wash fruits and
vegetables as one of their top priorities (33).
It is hypothesized that 1) organic load decreases the efficacy of sanitizers used
during simulated commercial processing of leafy greens and 2) E. coli O157:H7 persistence
can be correlated to various physicochemical parameters of the wash water.
In the absence of any major improvements in the methods of growing, harvesting,
processing, transporting, and displaying leafy greens, outbreaks of illness associated with the
consumption of fresh-cut leafy greens have continued to occur in the United States. Given the
need for a safe end-product, the overall objective of this research was to develop methods that
can be employed by processors to determine the efficacy of various sanitizers during commercial
lettuce processing in order to minimize the transfer of pathogens during washing.
The research reported in this dissertation had four primary objectives: 1) assess the
efficacy of five commercial sanitizer treatments against E. coli O157:H7 during simulated
commercial processing of iceberg lettuce in a pilot-scale leafy green processing line; 2)
determine the persistence of E. coli O157:H7 in wash water containing various sanitizers and
organic loads in a novel and cost-effective model bench-top carboy system; 3) determine the
impact of organic load on sodium hypochlorite and peroxyacetic acid efficacy, against E. coli
O157:H7 during simulated commercial processing; and 4) assess the relationship between
various physicochemical parameters and organic load of the wash water on sanitizer efficacy
against E. coli O157:H7 in both the carboy model and pilot-scale leafy green processing line.

3

CHAPTER 1:

Review of Pertinent Literature

4

1.1 The leafy greens industry
Consumption of vegetables has increased dramatically in recent years as Americans have
moved to towards healthier eating habits (27). Vegetables are rich sources of fiber, vitamins,
minerals and essential nutrients, especially carbohydrates (27). Processing methods such as
canning, drying and freezing lead to significant nutrient losses; however, these treatments are
necessary to improve product shelf life and quality, enhance palatability, and inactivate
nutritional inhibitors (27). Additional advancements in modified atmosphere storage and
minimal processing technologies have led to the commercial production of a wide for a variety
of fresh, convenient, and ready-to-eat products including leafy greens (2).
Commercially processed leafy greens are “value added” products and represent a multibillion dollar industry. California and Arizona are the main producers of iceberg lettuce and baby
spinach, with California’s peak production season in May and June, while Arizona’s is in
December through February, with production occurring nearly year-round (121). Prior to the
series of leafy green outbreaks in the fall of 2006, the pre-washed salad market brought in $2.6
billion annually, with the spinach industry worth $286 million (119). Earthbound Farms, under
its parent company Natural Selection Farms, had grown from selling spinach at a roadside stand
in 1986 into a $360 million industry (119).
The leafy green industry as a whole was hit particularly hard after the 2006 outbreaks,
with sales rapidly declining due to the extensive media coverage during the outbreaks and the
slow rate of the outbreak investigation. By mid-October of 2006, Natural Selection’s sales of
conventional salads were down by 70% and down by 10% for products sold under the
Earthbound Farms name (119). The company was forced to lay off 164 employees, and it has
been estimated that Earthbound Farms and Dole may have to pay up to $110 million to settle
5

cases with the victims of the outbreaks (119). Spinach losses alone totaled $205.8 million
following the outbreaks in 2006 (8).

1.2 Leafy green associated recalls and outbreaks
It is estimated that approximately 37.2 million illnesses are caused by 31 different
pathogens annually in the United States (103). As Americans attempt to improve their eating
habits, outbreaks of illness attributed to fresh produce are increasingly being reported. Between
1998 and 2006, produce was responsible for ~40 illnesses per outbreak, while poultry, beef, and
seafood were responsible for ~25, 23 and 9 illnesses per outbreak, respectively (34). The Centers
for Disease Control and Prevention recently ranked leafy vegetables as the leading cause of
foodborne illnesses in the United States (22%) and the second most frequent cause of
hospitalizations (14%) between 1998 and 2008 (89). In 2009, leafy greens were ranked as the
riskiest food regulated by the Food and Drug Administration (FDA), accounting for 363
outbreaks and 13,568 reported cases of illness (36). Between 1995 and 2006, leafy greenassociated outbreaks increased by 38.6% while consumption increased by only 9% (63). During
this same period, E. coli O157:H7 outbreaks increased notably in 2005 and 2006 compared to
1998 – 2004, as seen in Figure 1.1 (34).

6

8
7

Outbreaks

6
5
4
3
2
1
0
1998

1999

2000

2001

2002
Year

2003

2004

2005

2006

Figure 1.1: Produce outbreaks attributed to E. coli O157:H7 from 1998 – 2006 (34).
As seen in Figure 1.2, the Center for Science in the Public Interest also ranked the
specific commodities for produce-linked outbreaks between 1998 and 2006 (34). Salads were
responsible for the majority of outbreaks, followed by potatoes, tomatoes, melons and sprouts.
Salads alone, which include various types of leafy greens, were responsible for over one-third of
the outbreaks during that time period, with lettuce responsible for an additional 8%.
Bacterial pathogens can contaminate fresh produce at any point during the farm-to-fork
continuum (83). Major on-farm areas of concern now recognized by the FDA include
agricultural water, biological soil amendments (e.g., manure), domesticated and wild animals,
field worker health and hygiene, and the cleanliness of harvesting equipment, tools and buildings
(127). However, leafy greens are also prone to contamination during commercial processing,
packing (24), distribution, marketing (138), and in-home preparation (90).

7

Figure 1.2: Produce-linked outbreaks between 1998 and 2006, ranked according to commodity
(34).
The term “leafy greens” is broadly used to describe arugula, baby leaf lettuce, butter
lettuce, cabbage, chard, endive, escarole, green leaf lettuce, iceberg lettuce, kale, red leaf lettuce,
romaine lettuce, spinach and spring mix (29). Leafy greens were responsible for 363 separate
outbreaks involving 13,568 individual cases of illness through 2009 (36), with most of the E. coli
outbreaks attributed to lettuce, much of which is grown in the Salinas Valley in California. A
selection of E. coli outbreaks since 1995 can be seen in Table 1.1, (adopted from Mandrell (84)),
which has been updated to include more recent outbreaks.
The nationwide outbreak of E. coli O157:H7 that was traced to baby spinach in 2006
resulted in 205 confirmed infections, 103 hospitalizations, and three deaths (28, 38, 45). The
outbreak strain was isolated from numerous feral pigs on a ranch approximately one mile from
8

the spinach field implicated in the outbreak. Additionally, 33.8% of the cattle tested on the same
ranch tested positive for the outbreak strain. Evidence of feral pig intrusion, including fecal
droppings in the field and adjacent vineyards, signs of rooting, and tracks were present on the
same ranch (67). Molecular typing by pulsed-field gel electrophoresis (PFGE) and multilocus
variable number tandem repeat analysis were used to confirm that the strain isolated from the
pigs was in fact the outbreak strain. It is hypothesized that the pigs accessed the spinach field and
contaminated the product.
Two additional E. coli O157:H7 outbreaks in November and December of 2006 were
linked to shredded iceberg lettuce that was served at two different Mexican chain restaurants
(123, 124). These two outbreaks resulted in 150 illnesses in the Midwest and northeastern United
States (35, 123, 124). One of the E. coli O157:H7 outbreak strains matched isolates from
samples from a dairy farm near the growing region in central California (124), however the
source of the other strain could not be confirmed (123).

9

Table 1.1: Select leafy greens associated outbreaks from 1995 through 2012
Date

Pathogen

Location

Reported Illnesses

Product

Source Region

Jul. 95

E. coli O157:H7

MT

74

Lettuce, Romaine

MT, WA

Sept. 95

E. coli O157:H7

ME

30

Lettuce, Iceberg

Unknown

Sept. 95

E. coli O157:H7

ID

20

Lettuce, Romaine

Unknown

Oct. 95

E. coli O157:H7

OH

11

Lettuce

Unknown

May 96

E. coli O157:H7

IL, CT

61

Lettuce, Mesclun mix

CA

Jun. 96

E. coli O157:H7

NY

7

Lettuce, Mesclun

Unknown

May 98

E. coli O157:H7

CA

2

Lettuce, salad

Unknown

Sep. 98

E. coli O157:H7

MD

4

Lettuce

Unknown

Feb. 99

E. coli O157:H7

NE

65

Lettuce, salad

Unknown

Sep. 99

E. coli O157:H7

CA

8

Lettuce, Romaine

CA

Sep. 99

E. coli O157:H7

WA

6

Lettuce, Romaine

CA

Oct. 99

E. coli O157:H7

OH, IN

47

Lettuce, salad

Unknown

Oct. 99

E. coli O157:H7

OR

3

Lettuce, Romaine hearts

CA

Oct. 99

E. coli O157:H7

PA

41

Lettuce, Romaine

CA

10

Table 1.1 (cont’d)
Jul. 02

E. coli O157:H7

WA

29

Lettuce, Romaine

CA

Nov. 02

E. coli O157:H7

IL, WI, MN, SD, UT

24

Lettuce

CA

Sep. 03

E. coli O157:H7

CA

57

Lettuce, Iceberg/Romaine

CA

Sep. 03

E. coli O157:H7

ND

5

Lettuce, mixed w/ Romaine

Unknown

Oct. 03

E. coli O157:H7

CA

16

Spinach

CA

Nov. 04

E. coli O157:H7

NJ

6

Lettuce

CA

26 states

>200

Spinach, baby, bagged

CA

NJ, NY, PA, DE

71

Lettuce, Iceberg

CA

MN, IA, WI

81

Lettuce, Iceberg

CA

WA

10

Lettuce, Romaine

CA

Aug./Sep. 06 E. coli O157:H7
Nov. 06

E. coli O157:H7

Nov./Dec. 06 E. coli O157:H7
May 08

E. coli O157:H7

Sep. 08

E. coli O157:H7 MI, IL, NY, OR, OH, Ontario

74

Lettuce, Iceberg

CA

Apr. 09

E. coli O157:H7

MN

16

Lettuce, prepackaged

Unknown

Sep. 09

E. coli O157:H7

NY, WI, UT, NC, CO, SD

29

Lettuce, Iceberg/Romaine

CA

May 10

E. coli O145

MI, OH, NY, PA, TN

26

Lettuce, Romaine

AZ

10 States

58

Lettuce, Romaine

Unknown

Oct./Nov. 11 E. coli O157:H7

11

Table 1.1 (cont’d)
Apr. 12

E. coli O157:H7

Oct./Nov. 12 E. coli O157:H7

CA, Quebec

28

Lettuce, Romaine

CA

CN, MA, NY, PA, VA

33

Spinach and lettuce mix

MA

Information for the September 2008 outbreak based on Foodborne Illness Outbreak Database (49), the April 2009 outbreak is based on
Foodborne Illness Outbreak Database (52), the September 2009 outbreak is based on Foodborne Illness Outbreak Database (50), the
May 2010 outbreak is based on CDC (39), The October/November 2011 outbreak is based on Foodborne Illness Outbreak Database
and the CDC (40, 53), the April 2012 outbreak is based on Foodborne Illness Outbreak Database (51), and the October/November
2012 outbreak is based on CDC (41), all other outbreaks were summarized by Mandrell (84).

12

1.3 Escherichia coli O157:H7
E. coli O157:H7, a facultative anaerobic, Gram-negative, rod-shaped bacterium, is an
unusually virulent enterohemorrhagic strain of E. coli (EHEC). It is a cause of enteric disease
resulting in bloody diarrhea and severe abdominal pain and in some cases, hemolytic uremic
syndrome (HUS). Characterized by hemolytic anemia, thrombocytopenia and hemolytic uremic
syndrome (acute kidney failure) (69), HUS appears mainly in children, as first described by
Karmali and others (70). E. coli O157:H7 is a common inhabitant of the bovine gastrointestinal
tract and has historically been a major source of concern in the ground beef industry (58). This
organism first became recognized as a foodborne pathogen in 1982 after an outbreak tied it to the
consumption of undercooked hamburgers, which ultimately sickened 47 individuals (94). Since
its discovery, E. coli O157:H7 has been most often linked to outbreaks associated with ground
beef (15, 94), produce (67, 123, 124), fruit juices (37, 44), unpasteurized milk (58, 71), and
contaminated drinking water (114).
E. coli O157:H7 is the most well-known EHEC strain, but non-O157 strains now cause
an estimated 36,000 illnesses, 1,000 hospitalizations and 30 deaths annually in the United States.
The six most common serotypes of non-O157 EHEC are O26, O111, O103, O121, O45, and
O145 (20). E. coli is characterized by a specific combination of the O (somatic), H (flagellar)
and sometimes K (capsular) antigens which define the various serotypes (69). Strains of
pathogenic E. coli use multistep schemes of pathogenesis starting with colonization of the
intestinal mucosa, evading the hosts immune system, multiplication and damage to the host (69).
Four classes (virotypes) of E. coli in addition to EHEC that cause diarrheal diseases are currently
recognized: enterotoxigenic E. coli, enteroinvasive E. coli, enteropathogenic E. coli and
enteroaggregative E. coli.
13

E. coli O157:H7 has an extremely low oral infectious dose of less than 100 cells (85).
Enterohemorrhagic strains including E. coli O157:H7, destroy the normal microvillar structure of
the colon by expressing the virulence factor protein, intimin, which allows intimate attachment
of the bacteria and inducement of an attaching and effacing lesion. Shiga toxin (Stx) is the major
virulence factor of EHEC and also commonly known as verocytotoxin. Stx is a family of
structurally similar cytotoxins, with related biological activity, including two main groups, Stx1
and Stx2, which share about 40% amino acid similarity. Stx is composed of five identical B
subunits that bind the holotoxin to the glycolipid globotriaosylceramide on the host cell surface
and an A subunit, which cleaves ribosomal RNA to disrupt protein synthesis, ultimately killing
the epithelial and endothelial cells. Stx causes local damage to the colon resulting in bloody
diarrhea, hemorrhagic colitis and intestinal perforation. HUS results when Stx, produced in the
colon, travels through the bloodstream to the renal endothelial cells in the kidney, causing
inflammation by obstructing the microvasculature through both direct toxicity and inducing local
cytokine and chemokine production (30, 69, 92, 116).

1.4 Pre-harvest contamination and control
The increasing frequency of reported foodborne outbreaks associated with fresh fruits
and vegetables is of major concern in the United States. These outbreaks are the driving force for
changes needed by the produce industry in the way products are grown, harvested, and
processed. Contamination can occur at any point during the farm-to-fork continuum. Produce can
become contaminated on the farm through fecal contamination from domestic and wild animals,
water runoff from nearby livestock operations, improperly composted manure, dust/air, insects,
contaminated irrigation water, or poor handling practices during the harvesting process (19).
14

The mechanisms that E. coli O157:H7 uses to attach to the surfaces of leafy greens are
receiving increased interest. One group has recently demonstrated that EHEC strains of E. coli
O157:H7 and non-O157:H7 both attach to the leaf surfaces of spinach, arugula and lettuce
through the EspA filamentous type III secretion system that has also been shown to play an
important role in the organisms colonization of bovine and human hosts. This finding indicates
that EHEC uses plant leaves solely as a transmission vector rather than acting like a plant
pathogen (105). Additionally, flagella play a role in attachment because fliC gene mutants have
been shown to be significantly impaired in their ability to adhere to the surfaces of spinach and
lettuce compared to the parental strains (136). Adhesion has also been shown to be temperature
and time dependent. While intimin adhesion is vital for attachment to the host organism, it is not
required for colonization of leafy greens. Based on this, colonization of leafy greens by EHEC
strains of E. coli O157:H7 is likely a means to survive the processing environment and to allow
transmission to the human host (136).
Pathogens are most likely to attach to stomata, irregularities on intact surfaces, cut
surfaces, or cracks on the external surfaces (55, 91, 100, 102, 111) and can be protected from
sanitizers by biofilms (104). Biofilms are defined as an aggregate of microbes that are embedded
in an extracellular matrix, which encapsulates the cells and provides a physical barrier against
environmental stresses such as sanitizing agents, temperature changes, desiccation or ultraviolet
rays (7, 99, 117). Bacterial attachment and biofilm formation on the surfaces of leafy greens
greatly hinder the removal of microbial contaminants during washing (32, 101). When Lang and
others (75) assessed the impact of post-inoculation drying times on E. coli O157:H7 survival on
parsley and iceberg lettuce, equal or greater populations were recovered from the leafy greens
after 2 h of drying at 22°C compared to 2 h drying at 22°C followed by 22 h at 4°C. Hence, the
15

earlier in the farm-fork continuum the product is contaminated, the less effective bacterial
mitigation strategies are likely to be.
Following passage of the Food Safety Modernization Act (FSMA) into law in 2011, the
FDA released a set of proposed rules in January of 2013 for growing, harvesting, packing, and
holding produce for human consumption. According to the FDA, 16.8% of produce-related
outbreaks are attributed to fresh-cut fruits and vegetables, with the original contamination likely
occurring during growing, harvesting, packing, or holding (128). The major provisions and
regulatory actions of the proposed rule would set new standards for the following areas: worker
training, health and hygiene; sanitary agricultural water; biological soil amendments; proximity
and monitoring procedures for domestic and wild animals; sanitary conditions for equipment,
tools, and buildings; and sprout growth, harvesting, and handling (128). The proposed regulation
is estimated to prevent 1.75 million cases of foodborne illness annually, with an associated
financial benefit of $1.04 billion and an estimated cost of $460 million in total for domestic
farms (128). Even when contamination occurs at the growing or harvesting stages, conditions
during post-harvest processing can intensify and spread what would normally be just a small,
contained contamination event, resulting in a widespread outbreak (19, 26, 128).

1.5 Post-harvest processing
Minimal processing of leafy greens involves a series of steps that can either amplify or
promote the growth of pathogenic microorganisms. These steps include: harvesting, cold storage,
trimming, shredding, washing and rinsing, draining, packaging, cold storage and distribution
(118). After harvesting, microbial contamination can come from many different sources such as
the water used for cooling and washing, equipment surfaces and workers. The equipment used in
16

leafy green processing facilities varies, but most processors use the same general types of
equipment- including a shredder, conveyer belt, flume tank, shaker table, and dewatering
centrifugal dryer. Microbial contaminants can spread to multiple batches of product during
processing (e.g., washing, peeling, shredding, slicing, drying and sorting) which can lead to a
potential outbreak of illness.
Human foodborne pathogens, including E. coli O157:H7, Salmonella, Listeria and
Cryptosporidium that may inadvertently contaminate fresh fruits and vegetables in the field or at
harvest can be readily transferred to much larger quantities during subsequent product handing
and processing. These same foodborne pathogens can also contaminate the product during
processing. In one outbreak of salmonellosis traced to shredded lettuce, Stafford and others (108)
recovered Salmonella Bovismorbificans from the cutting wheel of a mechanical shredder during
an environmental audit, with insufficient cleaning and sanitizing of the shredder cited as a key
factor in this outbreak. In a study examining the microbial changes of lettuce during processing
and storage, a 1 log CFU/g increase was observed after shredding (3), indicating that the
shredder may be a critical in-plant vehicle for amplifying contamination of leafy greens during
processing (54).
After shredding, leafy greens are washed to remove soil and debris, which decreases the
microbial load, improves quality and appearance, and enhances product shelf life and safety (62).
Flume washing during leafy green processing is a common but highly variable practice.
Commercial washing of leafy greens can vary in terms of the number of wash steps (62), product
contact time with the flume water (30 s to 2 min) (62), mechanical agitation of the product in the
water, sanitizing agents and concentrations used, rate of sanitizer and water replenishment,

17

amount and kinds of product washed per shift, and capacity of the flume system. Flume washing
is typically the sole microbial mitigation strategy during processing.

1.6 Purpose of sanitizer addition to wash water
Numerous bench-top studies have shown that produce sanitizers reduce pathogen
populations only 1 to 3 logs on lettuce (47, 55, 91, 100), with water alone decreasing E. coli
O157:H7 levels about 1 log on lettuce during pilot-scale processing (22). Table 1.2 describes a
selection of sanitizing agents typically used by commercial produce processors. While washing
is the only intervention step against microbial populations on product, the primary goal of
sanitizing agents is to reduce the microbial populations in flume water to prevent crosscontamination during washing (137). Commercial processing facilities recirculate flume water in
order to reduce operational costs (80). Sanitizing agents are routinely added to recirculating
flume water as a means to prevent the water from becoming a microbial carrier, thereby
preventing the spread of pathogens in large, centralized processing facilities. However, the
efficacy of various sanitizing agents has been questioned as product recalls and outbreaks have
continued to occur.
Previous work done by our group involved quantification of E. coli O157:H7 transfer
during pilot-scale leafy green processing using sanitizer-free wash water. A major finding from
one of the studies was that ~90% of E. coli O157:H7 inoculum transferred from 22.7 kg of
6

iceberg lettuce inoculated at 10 log CFU/g to the 890 L of flume water used during processing
(22). These results emphasize that effective sanitizing agents in flume water are critical to
prevent the wash water from becoming a carrier of microbial contaminants.

18

Chlorine-based sanitizers are most commonly used by commercial processors to
minimize cross-contamination during processing, due to their relatively low cost compared to
other sanitizers, and minimal negative impact on end-product quality (31, 62, 81, 87, 91).
However, chlorine use has raised concerns regarding potentially hazardous by-products, worker
safety, environmental damage and, most importantly, decreased efficacy in the presence of an
increasing organic load in recirculating flume water, which has heightened interest in other
alternatives, such as peroxyacetic acid-based sanitizers (100, 113).

19

Table 1.2: Commercial produce sanitizers used during flume washing of leafy greens

a

Active Ingredients

Designated Use

Chlorine

Sodium hypochlorite

Sanitizing food contact
equipment, potable
water treatment, fruit
and vegetable washing

Peroxyacetic
acid

Peroxyacetic acid
Hydrogen peroxide

Advantages

Disadvantages

< 200

Cost effective,
effective against
all microbial
forms

pH dependent, high
reactivity with
organic solids,
corrosive to metals

< 80

Sanitizing Agent

Maximum
Allowed
Concentration
(ppm)

Low reactivity
with organic
solids, no
hazardous
breakdown
products

Strong oxidant,
concentrated
solutions may be
hazardous, costly

Pathogen reduction in
fruit and vegetable
processing water

Mixed Peracid

Peroxyacetic acid
Organic acids
Hydrogen peroxide
Acetic acid

Reduction of yeasts,
molds and bacteria in
water and on fruit and
vegetable surfaces

a

Adapted from text in Hedt and Feng (62).

20

1.7 Sodium hypochlorite use in wash water
Sodium hypochlorite (i.e., chlorine) - the most commonly employed sanitizing agent in
the fresh produce industry (31), has a long history of use and is relatively inexpensive when
compared to other sanitizing agents. The following reactions occur when sodium hypochlorite is
combined with water:
NaOCl + H2O ↔ HClO + NaOH
+

HClO ↔ H + OCl

-

HClO + HCl ↔ H2O + Cl2
Hypochlorous acid (HClO) is the main active ingredient formed and is the form of free
available chlorine that has the highest bactericidal activity of all active components. The
-

dissociation of HClO into OCl is pH dependent, with the concentration of HClO increasing
with decreasing pH (62, 91), predominating at pH 6.5 (16, 109). Solutions containing sodium
hypochlorite below a pH of 4 will cause toxic off-gassing and corrosion to processing equipment
(16, 109). Due to the potential of harmful disinfection byproducts, several European countries
have banned the use of chlorine in fresh-cut operations (61). Hypochlorous acid functions by
oxidation, with HClO allowing oxygen to combine with components of the cell wall, resulting in
cell death (62). More specifically, Rosen and others (96) proposed that hypochlorous acid
inactivates membrane proteins involved in DNA replication.
Free chlorine concentrations in clean commercial flume water generally range from 10 to
200 ppm (55, 91, 122). Chlorine is highly reactive with organic compounds, which is a major
issue considering that chlorine is the most commonly used sanitizer in the fresh produce
industry. Chlorine exists as two forms, free chlorine (i.e., available chlorine) and total chlorine.
21

-

Free chlorine refers to the hypochlorite ion (OCl ) and hypochlorous acid (HClO) concentration
in solution (62). Chloromines, such as monochloramine, dichloramine, or trichloramine, form
when chlorine is exposed to ammonia or organic nitrogen and are referred to as combined
chlorine (60). Additionally, the pH of a solution has an impact on combined chlorine formation,
with the reaction of chlorine with ammonia promoted above a pH of 7 (60, 62). Total chlorine is
the sum of combined chlorine and free chlorine. In systems with no organic load present in the
water, free chlorine is equivalent to total chlorine; however, as organic load increases in the
water, free chlorine levels drop (60).
Modification of pH is often achieved through the use of a weak acid, such as citric acid,
although newly-developed products, such as T-128 (SmartWash Solutions, Salinas, CA) are now
being marketed as effective alternatives. T-128 is a generally recognized as safe (GRAS)
acidifying agent comprised of phosphoric acid and propylene glycol, designed to stabilize free
chlorine (76). Previous research by Nou and others (87) showed that chlorine acidified with T128 (P < 0.001) decreased the rate of free chlorine degradation in the presence of organic load.
Additionally, T-128 significantly reduced cross-contamination when lettuce inoculated with E.
coli O157:H7 was washed with uninoculated lettuce. An additional study by the same group
showed that chlorine acidified to pH 5.0 with T-128 significantly reduced E. coli O157:H7
cross-contamination during pilot-scale processing of baby spinach when compared to the control,
where chlorine was acidified to pH 6.5 with citric acid (81). While exhibiting weak bactericidal
activity when used alone, T-128 does not play a significant role in pathogen inactivation (87),
which emphasizes the products role as means to enhance the efficacy of chlorine.
Numerous small-scale laboratory studies have assessed sodium hypochlorite efficacy
under a variety of conditions. More specifically, the impact of organic load in simulated produce
22

wash water on sanitizer efficacy has been previously examined (57, 61, 78, 80, 87, 107, 139).
Gonzalez and others (57) noted than an organic load correlating to a chemical oxygen demand
(COD) value of 3,500 mg O2/L resulted in significant reduction of sodium hypochlorite efficacy
against E. coli O157:H7 on carrots and in water. Haute and others (61) showed that maintaining
a free chlorine concentration of 1 ppm could reduce E. coli O157:H7 populations below the
lower limit of detection in process water containing COD values of 500 and 1,000 mg O2/L.
Additionally, they constructed regression models that incorporated chlorine dose and COD to
predict E. coli O157:H7 inactivation. Lopez-Galvez and others (78) found that wash water
containing an organic load with a COD value of 500 – 700 mg O2/L reduced of E. coli
populations by 2 logs on lettuce after 1 min of exposure to 40 ppm chlorine, with populations in
the wash water falling below the lower limit of detection. Luo (80) showed that washing
Romaine lettuce in water containing 100 ppm chlorine and a COD value of 1861 mg O2/L
resulted in lactic acid bacteria populations that were 0.8 – 1.6 log CFU/g higher than those
washed in clean water (COD value 9.8 mg O2/L) at the end of storage under modified
atmosphere packaging. More recently, Shen and others (107) showed that inactivation of
Salmonella, E. coli O157:H7, and non-O157 shiga toxin-producing E. coli in chlorinated wash
water significantly (P < 0.0001) depended on initial free chlorine concentration and contact time.
Additionally, they showed that pathogen inactivation was specifically dependent on the residual
free chlorine concentration when an organic load (COD values 186 – 460 mg O2/L) was present
in the wash water. According to Zhang and others (139), wash water containing 30 ppm chlorine
did not significantly reduce E. coli O157:H7 in wash water containing a 10% organic load when
23

compared to the sanitizer-free control. In addition to the research completed on leafy greens,
studies have also assessed chlorine efficacy on carrots (57, 97), apples (1, 21), nuts (18) and
melons (6) with generally similar findings.
Sodium hypochlorite has a long history of use in the fresh produce industry. It has
become increasingly evident that commercial produce processors must take steps to monitor, at
least, chlorine concentration and pH in order to minimize the dangers associated with toxic offgassing. Additionally, increasing organic loads are well known to negatively impact the efficacy
of chlorine, which is increasing the interest in alternative sanitizers, such as peroxyacetic acid.
However, since chlorine is still widely used throughout the fresh-produce industry, improved
monitoring methods must be developed for chlorine efficacy to minimize the spread of microbial
contaminants during processing.

1.8 Peroxyacetic acid use in wash water
Peroxyacetic acid (i.e., peracetic acid or POAA) is receiving increasing interest as an
alternative to sodium hypochlorite, due to its non-hazardous breakdown products and resistance
to organic material in wash water. Peroxyacetic acid (C2H4O3) is produced when acetic acid
(CH2CO2H) reacts with hydrogen peroxide (H2O2) (62) as follows:
CH3CO2H + H2O2 ↔ CH3CO3H + H2O
Peroxyacetic acid functions by oxidation, with the suggested mechanism of action being
the oxidation of sulfhydryl bonds in proteins, enzymes and other metabolites, resulting in
increased cell wall permeability (62, 64). It has also been proposed that peroxyacetic acid is
involved in enzymatic blockage of enzymatic and transport systems within microorganisms (74).
24

Peroxyacetic acid breaks down into water, oxygen, and acetic acid (9, 46, 74). The FDA has
established, under the Code of Federal Regulations (43), a peroxyacetic acid limit of 80 ppm in
wash water to be used for fruit and vegetable washing.
While chlorine efficacy has been studied extensively, only a few studies have assessed
the efficacy of peroxyacetic acid in simulated processing water (1, 17, 65, 78, 95, 129) and even
fewer have examined the impact of organic load on peroxyacetic acid efficacy (57, 64, 97, 139).
Gonzalez and others showed that the efficacy of 80 ppm peroxyacetic acid in wash water was not
significantly impacted by an organic load (COD 3,500 mg/L), resulting in E. coli O157:H7
populations below the limit of detection after washing inoculated shredded carrots. Hilgren and
others (64) showed that the efficacy of peroxyacetic acid against Bacillus anthracis spores was
rarely impacted by the presence of whole milk, egg yolk or flour paste. Ruiz-Cruz and others
(97) found that the antimicrobial activity of 40 ppm peroxyacetic acid was not impacted by
organic load (COD 3,500 mg/L), resulting in Salmonella, E. coli O157:H7 and L. monocytogenes
reductions of 2.1, 1.24 and 0.83 log CFU/g, respectively, on carrot shreds after 2 min of washing.
Zhang and others (139) showed that 20 ppm peroxyacetic acid and 20 ppm peroxyacetic acid
mixed with octanoic acid (i.e., mixed peracid) significantly reduced (P < 0.05) E. coli O157:H7
transfer from inoculated lettuce leaves to uninoculated leaves in wash water containing a 10%
organic load comprised of blended lettuce solids.
The addition of octanoic acid, a fatty acid, to peroxyacetic acid has been examined as a
means to increase the fungicidal activity of peroxyacetic acid while also increasing the efficacy
against coliform bacteria. Hilgren and Salverda (65) found that peroxyacetic acid in combination
with octanoic acid (80 ppm) was significantly more effective (P < 0.05) than peroxyacetic acid
alone (80 ppm) at reducing numbers of yeasts and molds in potato wash water and on the
25

surfaces of potatoes. However, no significant difference (P > 0.05) in coliform reductions on
celery, cabbage, and potatoes was seen between peroxyacetic acid and peroxyactiec acid in
combination with octanoic acid (65).
The major advantages of peroxyacetic acid are its non-hazardous byproducts and
increased resistance to organic material in wash water. However, no studies have yet explained
the phenomenon of peroxyacetic acid resistance to organic load. Major disadvantages of
peroxyacetic acid-based sanitizers include the increased cost and relatively short history of use
when compared to sodium hypochlorite. Addition of peroxyacetic acid will increase the organic
load in the wash water due to the acetic acid content (73). However, this increase in organic load
will not impact peroxyacetic acid efficacy, since acetic acid is native to the sanitizer (73, 78), but
the cost of the waste water disposal may be higher. The foremost disadvantage of peroxyacetic
acid is the cost, which is four to five times higher as compared to sodium hypochlorite in the
United States (73). Kits (73) hypothesized than an increased demand of peroxyacetic acid would
reduce this cost, making peroxyacetic acid-based sanitizers more economically feasible for
processors.

1.9 Organic load accumulation in wash water
An organic load, consisting of plant tissues and cellular fluids released during cutting, in
addition to soil, insects, and microbes (62), will accumulate in recirculating flume water as produce
is washed, decreasing the ability of sanitizers to minimize cross-contamination from the water during
processing (57, 110, 139). Based on a series of personal site visits to various commercial leafy-

green processing facilities, numerous factors were observed that will affect the rate of
accumulation and maximum organic load in flume water, such as the type and amount of product
26

being processed, product shred size, the rate of processing, and the means by which the organic
load is decreased in the water during processing (e.g., separating screens, foam removal, filters,
clarifiers, precipitators).
Numerous small-scale laboratory experiments have previously assessed the impact of
organic load on sodium hypochlorite (57, 78, 80, 87, 107, 139), chlorine dioxide (10),
electrolyzed water (56, 79), and peroxyacetic acid (11, 57, 64, 97, 139). In general, sodium
hypochlorite efficacy dramatically decreases, but peroxyacetic acid maintains some activity in
the presence of an organic load. Given the various commercial sanitizers available and the
variable rates at which organic material accumulates in flume water during processing,
commercial leafy green processors are clearly in need of better means to both predict and
monitor the efficacy of sanitizers during processing.

1.10 Strategies for monitoring wash water efficacy
A variety of methods have been developed to either directly or indirectly monitor the
efficacy of sanitizing agents in wash water, with these ranging from titration kits to digital probes
or even fully automated systems that monitor and adjust sanitizing agent concentrations as
needed. Titration test kits are often used by smaller operations to rapidly quantify the level of
total chlorine or peroxyacetic acid content in wash water or as a validation procedure alongside
high-tech automated systems. The principle of the chlorine titration method (i.e., iodometric
method) is that chlorine liberates iodine in the acidified test solution containing potassium
iodide, with the liberated iodine titrated with a standard solution of sodium thiosulfate to a starch
endpoint. Free chlorine concentrations also can be quantified by using the DPD (N, N diethyl-pphenylenediamine) method, which is based on the oxidation of an amine by chlorine into
27

Würster dye, which is then read by a colorimeter (60). Electronic probes - either hand-held
models or as part of an automated system, are also available. Considering that both free and total
chlorine concentrations are variable and do not necessarily correlate to the efficacy of chlorine,
methods to monitor the oxidizing capacity of the water have been employed.
As part of the proposed FSMA regulation, the FDA recommends, at minimum,
monitoring and adjusting as necessary the concentration of the active component of the
sanitizing agent and the pH of wash water, especially if sodium hypochlorite is used (128). The
active component of chlorine, HClO, is most abundant around pH 6.5, resulting in a
recommended range of 6.5 – 7.5 (16, 62, 82, 111, 125). Some commercial processors that use
chlorinated wash water for leafy greens have set their critical limit at a pH of < 7 (82). However,
peroxyacetic acid-based sanitizers remain effective over a wider pH range (62, 129), generally
making pH adjustments unnecessary.
The temperature at which washing occurs varies widely throughout industry, with some
processors opting for ambient water temperatures followed by air-cooling subsequent to
washing. The maximum solubility of chlorine occurs at 4°C (91), allowing the retention of
antimicrobial activity for longer periods (62). It is often reported that the overall bactericidal
activity of water increases with temperature (140); however, so does chlorine volatility (109). In
contrast to chlorine, peroxyacetic acid efficacy is maintained over a wider temperature range (64,
129).
Oxidation/reduction potential (ORP) is the potential (voltage) at which oxidation occurs
at the anode and reduction occurs at the cathode of an electrochemical cell (112). It is used by
processors to monitor the oxidizing capacity in millivolts (mV) of sanitizing flume water as an
alternative or in addition to monitoring sanitizer concentration. An advantage of monitoring ORP
28

is that it is a direct measurement of the ability of the wash water to oxidize microbial
contaminants, thought to be independent of the water quality. Commodity-specific guidelines by
the Florida Department of Agriculture initially required 150 ppm of free chlorine at a pH of 6.5 –
7.5 for flume water processing of tomatoes (125). However, the option to maintain a minimum
ORP of 650 mV was later included due to uncertainty of processors’ ability to maintain such a
chlorine concentration over time (125). No such regulation has been established for leafy greens;
however, some commercial processors have also adopted this same ORP of 650 mV as a critical
limit in their HACCP programs. More recently, the relationship between ORP and sanitizer
strength was shown to be nonlinear, raising questions regarding the ability of ORP to predict
sanitizer efficacy (120).
Increased interest in wash water safety has shifted the focus from direct monitoring of
sanitizer strength or efficacy to methods that can quantify and predict sanitizer efficacy based on
the organic load and other physicochemical parameters of the wash water. Chemical oxygen
demand (COD), which measures the amount of dissolved oxygen (O2) per L of solution, is an
indirect but quantitative means to determine organic load. In this method involves the reaction
of organic matter with a strong acid solution in the presence of a known excess of potassium
dichromate (K2Cr2O7). After 2 h of digestion at 150°C, unreduced K2Cr2O7 is quantified by
titrating with ferrous ammonium sulfate. The amount of oxidizable matter is calculated in terms
of oxygen equivalent (12). While by no means a rapid method, COD has shown to be an
effective means of quantifying organic load in wash water, with increasing COD correlating to
increased organic matter in both produce processing and waste water (10, 78, 87).

29

Additional methods to quantify organic load in wash water include direct methods (e.g.,
total solids and maximum filterable volume) and indirect methods (e.g., turbidity). Measurement
of total solids in water is a direct method to quantify the amount of solids in suspension or
dissolved in a set volume of liquid. The procedure involves determining the mass of dried solids
remainingin a crucible after evaporation of the liquid in an oven (14). Maximum filterable
volume (MFV) is a direct method to quantify organic load that was developed by our group. This
simple and rapid procedure is based on the volume of liquid pulled through a 0.45 µm filter after
1 min of vacuum filtration at -80 kPa. However, this test is not as sensitive as that for total solids.
Measuring the turbidity (i.e., absorbance) of a water sample provides another rapid and indirect
means to quantify organic load, which was also developed by our group. This procedure first
involves a pre-filtration step to remove suspended solids from the water sample to reduce
interference in the spectrophotometer. The absorbance of the filtrate is then measured at 663 nm,
a wavelength previously shown to correlate to chlorophyll content (133). However, these results
are likely sanitizer-specific, due to the different oxidizing strengths impacting the measured
chlorophyll content. While the methods above can be used to directly or indirectly assess
sanitizer efficacy and organic load in wash water, knowledge gaps still remain as to what
physicochemical parameters best correlate to E. coli O157:H7 persistence during simulated
commercial processing.

1.11 Current challenges
Flume water is recirculated during leafy green processing to reduce water waste and cost.
Previous work by Buchholz and others (22) reported that approximately 90% of the E. coli
O157:H7 populations on dip-inoculated leafy greens transferred to sanitizer-free water during
30

washing in a pilot-scale processing line. This reinforces the importance of adding an effective
sanitizer to the wash water to minimize cross-contamination via the water to subsequently
processed product. Oxidation/reduction potential (ORP) is most commonly used commercially to
monitor sanitizer efficacy is (112), however, with multi-state outbreaks and recalls still
occurring, it is clear that isolated contamination events are being amplified during processing of
leafy greens. Organic load, which contains produce tissue, cellular fluids released during cutting,
soil, insects, and microbes (62), will continually accumulate in recirculating flume water as produce
is washed (86), decreasing the ability of sanitizers to minimize cross-contamination from the water
during processing (57, 110, 139). While organic load impacts the efficacy of sanitizing agents,

most notably chlorine, the means to quantify sanitizer efficacy against pathogens such as E. coli
O157:H7 have not yet been determined. The Center for Produce Safety (CPS) currently ranks the
identification of methods to validate the efficacy of flume water used to wash fruits and
vegetables as one of their top priorities (33). Physicochemical parameters that either directly or
indirectly quantify the organic load in flume water may be better predictors of sanitizer efficacy
than ORP. Intervention steps (e.g., dumping of flume water or use of technologies to remove
organic load) are also needed to ensure that the target physicochemical parameters are being
maintained in the flume water, so as to not amplify possible microbial contamination events due
to sanitizer failure.

31

CHAPTER 2:

Efficacy of Commercial Produce Sanitizers against Escherichia coli O157:H7 during
Processing of Iceberg Lettuce in a Pilot-Scale Leafy Green Processing Line

32

2.1 ABSTRACT
Chemical sanitizers are routinely used during commercial flume washing of fresh-cut
leafy greens to minimize cross-contamination from the water. This study assessed the efficacy of
five commercial sanitizer treatments against E. coli O157:H7 on iceberg lettuce, in wash water,
and on equipment during simulated commercial production in a pilot-scale processing line.
6

Iceberg lettuce (5.4 kg) was inoculated to contain 10 CFU/g of a 4-strain cocktail of nontoxigenic, GFP-labeled, ampicillin-resistant E. coli O157:H7 and processed after 1 h of draining
at ~22oC. Lettuce was shredded using a Urschel TransSlicer, step-conveyed to a flume tank,
washed for 90 s using six different treatments (water alone, 50 ppm peroxyacetic acid, 50 ppm
mixed peracid, or 50 ppm available chlorine either alone or acidified to pH 6.5 with citric acid
(CA) or T-128), and then dried using a shaker table and centrifugal dryer. Various product (25 g)
and water (50 ml) samples collected during processing, and equipment surface samples (100
2

cm ) from the flume tank, shaker table and centrifugal dryer, were homogenized in neutralizing
buffer and plated on tryptic soy agar containing 0.6% yeast extract and 100 ppm ampicillin with
or without prior 0.45 μm membrane filtration to quantify E. coli O157:H7 under UV light.
During and after iceberg lettuce processing, none of the sanitizers were significantly more
effective (P ≤ 0.05) than water alone at reducing E. coli O157:H7 populations on lettuce, with
reductions ranging from 0.8 to 1.4 log CFU/g. Regardless of the sanitizer treatment used, the
2

centrifugal dryer surfaces yielded E. coli O157:H7 populations of 3.5 to 5 log CFU/100 cm . In
terms of the flume water, chlorine, chlorine + CA, and chlorine + T-128 were generally more
effective (P ≤ 0.05) than the other treatments, with reductions of 3.8, 5.5, and 5.4 log CFU/ml
after 90 s of processing, respectively. This indicates that chlorine-based sanitizers will likely
33

prevent wash water containing low organic loads from becoming a vehicle for crosscontamination.

34

2.2 INTRODUCTION
In 2009, leafy greens were ranked as the riskiest food category regulated by the Food and
Drug Administration, accounting for 363 outbreaks and 13,568 reported cases of illness (36).
Between 1995 and 2006, leafy green-associated outbreaks increased by 38.6% while
consumption increased by only 9% (63). The nationwide outbreak of E. coli O157:H7 that was
traced to baby spinach in 2006 resulted in 205 confirmed infections, 103 hospitalizations, and
three deaths (28, 45). Following two additional E. coli O157:H7 outbreaks in 2006 linked to
shredded iceberg lettuce resulting in 150 illnesses (35), at least 9 more outbreaks responsible for
nearly 300 cases of E. coli O157:H7 infection have been documented in the United States
through 2012 (40), heightening continued safety concerns surrounding fresh-cut leafy greens.
Bacterial pathogens can contaminate leafy greens at any point during the farm-to-fork
continuum (83). Major on-farm areas of concern now recognized by the Food and Drug
Administration include agricultural water, biological soil amendments (e.g., manure),
domesticated and wild animals, field worker health and hygiene, and the cleanliness of
harvesting equipment, tools and buildings (127). However, leafy greens are also prone to
contamination during commercial processing, packing (24), distribution, marketing (138), and
in-home preparation (90). Regarding leafy greens, pathogens are most likely to attach to stomata,
irregularities on intact surfaces, cut surfaces, or cracks on the external surfaces (55, 91, 100, 102,
111) and can be protected from sanitizers by biofilms (104). Since sanitizers in the wash water
cannot be relied upon to inactivate attached or internalized pathogens during processing, it is
imperative that growers and harvesters follow Good Agricultural Practices (GAPs) and Good
Handling Practices (GHPs) to reduce the likelihood of contamination (48).

35

Washing of leafy greens remains important for removing soil and debris, decreasing the
microbial load, improving quality and appearance, and enhancing product shelf life and safety
(62). Numerous small-scale laboratory studies have shown that produce sanitizers reduce
pathogen populations only 1 to 3 log CFU on lettuce (17, 47, 55, 91, 100) with water alone
decreasing E. coli O157:H7 levels about 1 log CFU on lettuce during pilot-scale processing
(22). Recirculation of this wash water during processing can further magnify the spread of
contaminants at large, centralized processing facilities (62, 80). Hence, the addition of sanitizers
to processing water is imperative to minimize cross contamination during commercial production
of fresh-cut leafy greens (5, 81, 100, 126).
Chlorine-based sanitizers are preferred for commercial flume washing systems due to
their relatively low cost compared to other sanitizers and minimal negative impact on end
product quality (31, 62, 81, 87, 91). Since the active component of chlorine, hypochlorous acid
(HClO), is most abundant at pH 6.5 – 7.0 (16), the pH of the wash water typically needs to be
lowered by adding a weak acid, most commonly citric acid (62). A new, generally recognized as
safe (GRAS) acidifying agent comprised of phosphoric acid and propylene glycol, known as T128 (SmartWash Solutions, Salinas, CA), has been developed to improve the stability of chlorine
(76, 81, 87, 106). However, chlorine use has raised concerns regarding potentially hazardous biproducts, worker safety, environmental damage and, most importantly, decreased efficacy in the
presence of an increasing organic load in recirculating flume water, which has heightened
interest in other alternatives such as peroxyacetic acid-based sanitizers (100, 113).
Numerous small-scale laboratory studies have assessed sanitizer efficacy against
pathogens on leafy greens (4, 17, 66, 72, 78, 82, 88, 115, 139, 140). However, these findings are
difficult to extrapolate to large-scale commercial production facilities. Previous work completed
36

by our group was performed without chemical sanitizers to quantify E. coli O157:H7 transfer
during pilot-plant production of fresh-cut leafy greens (22, 23). Since chemical sanitizers remain
the sole intervention strategy to prevent cross-contamination during commercial production of
fresh-cut leafy greens, it is imperative that these sanitizers be re-evaluated under conditions that
more closely resemble commercial operations. Consequently, the objective of this study was to
assess the efficacy of five commercial sanitizer treatments against E. coli O157:H7 during
processing of iceberg lettuce in a pilot-scale leafy green processing line.

37

2.3 MATERIALS AND METHODS
2.3.1 Experimental design. The efficacy of five different sanitizing treatments was
assessed in triplicate against E. coli O157:H7 by processing a 5.4 kg batch of iceberg lettuce
6

inoculated at 10 CFU/g, with sanitizer-free water serving as the control. All lettuce was
processed by shredding, conveying, fluming, shaker table dewatering, and/or centrifugal drying,
during and/or after which various product, water, and equipment surface samples were collected
and quantitatively examined for E. coli O157:H7.
2.3.2 Leafy greens. Individually wrapped heads of iceberg lettuce (Lactuca sativa L.)
(24 heads per case) were obtained from a local wholesaler (Stan Setas Produce Co., Lansing,
MI), with the product originating from California or Arizona depending on the growing season.
All lettuce was stored in a 4°C walk-in-cooler and used within 5 days of delivery.
-

-

2.3.3 Bacterial strains. Four non-toxigenic (stx1 and stx2 ) strains of E. coli O157:H7
(ATCC 43888, CV2b7, 6980-2, and 6982-2) were obtained from Dr. Michael Doyle at the
Center for Food Safety, University of Georgia, Griffin, GA. All strains had previously been
transformed with a pGFPuv plasmid containing a GFP gene and an amplicillin-resistance gene.
All four strains were stored at -80°C in tryptic soy broth (Difco, BD, Sparks, MD) containing
0.6% (w/v) yeast extract (TSBYE; Difco, BD), and 10% (v/v) glycerol (Sigma Chemical Co., St.
Louis, MO) until needed. Working cultures were prepared by streaking each stock culture on
Tryptic soy agar plates (Difco, BD) containing 0.6% (w/v) yeast extract and 100 ppm ampicillin
(ampicillin sodium salt, Sigma Life Science, St. Louis, MO) (TSAYE plus amp). After 18 – 24 h
of incubation at 37°C, a single colony was transferred to 9 ml of TSBYE containing 100 ppm
ampicillin (TSBYE plus amp) and similarly incubated.
38

2.3.4 Lettuce inoculation. A 0.2 ml aliquot of each non-toxigenic E. coli O157:H7
strain was transferred to 200 ml of TSBYE with amp and incubated for 18 – 20 h at 37°C.
Assuming similar growth rates, as determined previously (22), the four strains were combined in
equal volumes to obtain an 800-ml cocktail, which was added to 80 L of municipal tap water
7

(~15oC) in a 121 L plastic container (Rubbermaid, Wooster, OH) to a achieve a level of ~10
CFU/ml. Hand-cored heads of iceberg lettuce (~12 heads) were immersed in the E. coli
suspension for 15 min and then drained/air-dried for 1 h at 22°C before being spun in a

dewatering centrifuge (described below) to remove residual inoculum from the interior of the
heads. Duplicate 25-g samples were then aseptically collected to determine the initial inoculation
level at the time of processing.
2.3.5 Lettuce processing line. A small-scale commercial leafy green processing line
was assembled that consisted of a lettuce shredder, step conveyer, flume tank, shaker table and
dewatering centrifuge. The commercial lettuce shredder (model TRS 2500 Urschel TranSlicer,
Valparaiso, IN) was operated at a feed belt speed of 198 m/min and a slicing wheel speed of 905
RPM to obtain a shred size of approximately 5 x 5 cm. The polyurethane step conveyer belt
(ThermoDrive, Mol Industries, Grand Rapids, MI) was mounted on a motorized conveyor
(Dorner model 736018 mc series, Dorner Manufacturing, Hartland, WI) that operated at 0.11
m/s. The stainless steel water recirculation tank (~1000 L capacity) containing 890 L of tap
water (~15oC) was connected to a 3.6 m long stainless steel flume tank (Heinzen Manufacturing,
Inc., Gilroy, CA) - equipped with two overhead spray jets (1 m from the start), by a 4.14 m long,
10 cm-diameter hard plastic discharge hose and a centrifugal pump (Model XB754FHA, Sterling
Electric, Inc., Irvine, CA) that circulated the water at a rate of ~10 L/s. A custom-made stainless
steel screen containing pores with a diameter of 1.25 cm spaced 0.65 cm apart (Heinzen
39

Manufacturing, Inc.) was affixed to the end of the flume tank to retain the product during
washing. The stainless steel shaker table for partial dewatering was operated by a 1 HP Baldor
washdown duty motor (Baldor Electric Co., Ft. Smith, AR) at 1760 RPM. Water removed from
the leafy greens during mechanical shaking passed through a fine mesh screen and was fed into
the water holding tank by a water recirculation spout underneath the shaker table. A 110 kg (50
lb) capacity centrifugal Spin Dryer (Model SD50-LT, Heinzen Manufacturing, Inc.) with three
internally timed spin cycles totaling 80 s was used for centrifugal drying.
2.3.6 Wash water. Iceberg lettuce (0.5 kg) was homogenized in 500 ml of Michigan
State University (MSU) tap water using a mechanical blender (Model BLC10650MB, Black &
Decker, New Britain, CT) and then added to 890 L of processing water at 12 to 15oC to achieve a
low-level organic load of ~0.0006% (w/v lettuce solids). The following five commercial produce
sanitizer treatments were assessed: 30 ppm peroxyacetic acid (Tsunami 100, Ecolab, St. Paul,
MN), 30 ppm mixed peracid (Tsunami 200, Ecolab), 30 ppm available chlorine (XY-12, Ecolab)
at pH 7.85, 30 ppm available chlorine (XY-12) acidified to pH 6.50 with citric acid (SigmaAldrich, St. Louis, MO), and 30 ppm available chlorine (XY-12) acidified to pH 6.50 with T-128
(SmartWash Solutions) as measured with a pH probe (pHTestr 30, Oakton, Vernon Hills, IL).
Peroxyacetic acid test kit 311 (Ecolab) was used to confirm the peroxyacetic acid and mixed
peracid sanitizer concentrations while chlorine test kit 321 (Ecolab) was used to measure
available chlorine. Sanitizer-free MSU tap water (< 0.05 ppm free chlorine) served as the
control.
2.3.7 Leafy green processing. Inoculated heads of cored iceberg lettuce (5.4 kg) were
hand-fed into the shredder at a rate of ~0.5 kg per second, with the shredded product then stepconveyed to the top of the conveyor. Processing was then halted for ~10 min to aseptically
40

collect and bag five 25-g lettuce samples in red mesh produce bags (5 lb Header Bag, Pacon Inc.,
Baldwin Park, CA) for subsequent sampling. Thereafter, processing was resumed with the
iceberg lettuce conveyed to the flume tank, washed in 890 L of recirculating wash water with or
without a sanitizer for 90 s, partially dewatered on the shaker table, collected in a single
centrifugation basket and centrifugally dried.
2.3.8 Sample collection. During the 90 s of flume washing, three pre-bagged iceberg
lettuce samples (25 g each) were retrieved at the flume gate at 30 s intervals, and immediately
added to 100 ml of sterile Difco Neutralizing Buffer (BD, Franklin Lakes, NJ) in a Whirl-Pak™
filter bag (Nasco, Fort Atkinson, WI). In addition, nine 50-ml water samples were also collected
at 10-s intervals in 50 ml centrifuge tubes containing 38x concentrated Difco Neutralizing Buffer
(BD). After shaker table dewatering, product in the basket was dried in the pre-set 110 kg (50 lb)
capacity Spin Dryer (Model SD50-LT, Heinzen Manufacturing). During centrifugal drying, four
water samples (50 ml each) were similarly collected from the centrifuge drain at 10 s intervals
for the first 40 s of the 80 s cycle. After centrifugation, two bagged lettuce samples (25 g each)
were also retrieved from the centrifugation basket. Nine product contact areas on the equipment
(3 flume tank (Figure 2.1), 3 shaker table (Figure 2.2) and 3 dewatering centrifuge (Figure 2.3)),
2

previously described in detail by Buchholz et al. (22), measuring 100 cm as previously
identified using Glo Germ™ were sampled immediately after processing as described by Vorst et
al. (130) using 1-ply composite tissues moistened with 1 ml of sterile Difco Neutralizing Buffer
(BD).

41

Figure 2.1: Flume tank sampling locations

42

Figure 2.2: Shaker table sampling locations

43

Figure 2.3: Dewatering centrifuge sampling locations

44

2.3.9 Microbiological analyses. All lettuce samples (25 g) were homogenized in a
stomacher (Stomacher 400 Circulator, Seward, Worthington, UK) for 1 min at 260 rpm and then
either appropriately diluted in sterile 1% (w/v) phosphate buffer (8.5 g/L NaCl, 1.44 g/L
Na2HPO4, and 0.24 g/L KH2PO4, J.T. Baker, Mallinckrodt Baker Inc., Phillipsburg, NJ) and
plated on TSAYE with amp (calculated minimum detection limit of 40 CFU/g) or processed
using 0.45 μm membrane filters (Millipore, Millipore Corporation, Billerica, MA) (calculated
minimum detection limit of 0.04 CFU/g), which were placed on 60-mm dia. petri plates
containing TSAYE with amp to quantify E. coli O157:H7. The 1-ply composite tissue samples
were added to 15 ml of sterile Difco Neutralizing Buffer in a Whirl-Pak™ bag, homogenized for
1 min at 260 rpm, and then plated identically to the lettuce samples, giving a calculated lower
2

detection limit of 1 CFU/100 cm . The 50 ml water samples were either appropriately diluted in
sterile 1% phosphate buffer and plated on TSAYE with amp or processed by membrane
filtration, which gave a calculated minimum detection limit of 0.02 CFU/ml. Following 20 – 24 h
of incubation at 37°C, all green fluorescing colonies as seen under ultraviolet light (365 nm,
Blak-Ray, Ultra-violet Product Inc. San Gabriel, CA) were counted as E. coli O157:H7.
2.3.10 Sanitizer neutralization confirmation. Triplicate 1 L water samples containing
30 ppm available chlorine (XY-12, Ecolab), 30 ppm peroxyacetic acid (Tsunami 100, Ecolab) or
30 ppm mixed peracid (Tsunami 200 ppm, Ecolab) were prepared and confirmed with Chlorine
test kit 321 (Ecolab) or Peroxyacetic acid test kit 311 (Ecolab). Citric acid (Sigma-Aldrich) and
T-128 (SmartWash Solutions) were used to acidify the chlorine-based sanitizer solution to pH
6.5. A 50 ml centrifuge tube containing 3 ml of 38x concentrated Neutralizing Buffer (BD) was
filled with the water sample containing sanitizer, agitated for 5 s and then immediately assessed
45

for neutralization of the sanitizer as previously described using the appropriate test kit.
Preliminary experiments found a 38x concentration would neutralize various concentrations of
the active component of each sanitizing agent used in this study without impacting E. coli
O157:H7 counts.
2.3.11 Statistical analysis. E. coli O157:H7 counts were converted to log CFU per g, ml
2

or 100 cm and subjected to an ANOVA using JMP 9.0 (SAS Institute Inc., Cary, NC). Values
equaling half the limit of detection were used for samples without E. coli O157:H7 counts. The
three equipment surface samples from each respective piece of equipment were averaged. A P
value of ≤ 0.05 was considered significant for all tests. The Tukey-Kramer HSD test was used to
identify significant differences in E. coli O157:H7 populations for individual lettuce, water, and
equipment surface samples.

46

2.4 RESULTS
2.4.1 Lettuce. Iceberg lettuce contained an average E. coli O157:H7 inoculum of 5.93
log CFU/g at the time of processing (Figure 2.4). After shredding, conveying, 90 s of washing,
shaker table dewatering and centrifugal drying, no significant difference (P > 0.05) was seen in
populations of E. coli O157:H7 recovered from the finished product, regardless of sanitizer
treatment. Using mixed peracid, E. coli O157:H7 populations decreased 1.4 log CFU/g;
however, this decrease was not significantly different (P > 0.05) than the 0.8 log CFU/g
reduction seen for water alone. Processing significantly reduced (P ≤ 0.05) E. coli O157:H7
populations on lettuce when mixed peracid, chlorine, or chlorine + CA were used, with
reductions of 1.4, 0.8, and 0.9 log CFU/g, respectively. The reductions of 0.8, 0.9, and 1 log
CFU/g seen for water alone, peroxyacetic acid, and chlorine + T-128, respectively, were not
significant (P > 0.05) (Figure 2.4).

47

Initial
A

7.0
A
E. coli O157:H7 (log CFU/g)

6.0

A
A

5.0

A

30 sec

60 sec

A
A

AAA

A

B BAB
B

90 sec
A

After Centrifugation

BBBB

A

A
AA

AA

BBBB

4.0
3.0

2.0
1.0
0.0
Water

Peroxyacetic
Acid

Mixed
Peracid

Chlorine

Chlorine +
T-128

Chlorine +
CA

Flume Water Treatment
Figure 2.4: Mean (± SD) E. coli O157:H7 populations on the iceberg lettuce inoculated at ~6 log
CFU/g during and after processing (n=3). Means of the same wash water treatment with
different letters are significantly different (P ≤ 0.05).

48

2.4.2 Flume water. E. coli O157:H7 populations in the flume water were homogenous
after 10 s of processing, with no significant difference (P > 0.05) seen between 10 – 90 s in the
water control at the flow rate of 9.3 L/s. Wash water containing chlorine, chlorine + T-128, and
chlorine + CA had significantly lower (P ≤ 0.05) E. coli O157:H7 populations at all sampling
times (maximum of 1 log CFU/ml) compared to 4.6 log CFU/ml in water alone. Using chlorine +
CA and chlorine + T-128, E. coli O157:H7 levels were below the limit of detection of 0.02 log
CFU/ml by the end of processing. E. coli O157:H7 populations were similar (P > 0.05) using
water alone and peroxyacetic acid, with respective populations of 3.5 and 3.0 log CFU/ml
recovered after 90 s of processing. Similar E. coli O157:H7 populations were obtained using
mixed peracid (P > 0.05) and peroxyacetic acid, with these populations rarely lower (P ≤ 0.05)
than those in water alone (Figure 2.5).

49

Water
Chlorine

6
E. coli O157:H7 (log CFU/ml)

5

Peroxyacetic Acid
Chlorine + T-128

4
3

A
A
A

A
A
A

A
A
A

Mixed Peracid
Chlorine + CA

A
A
AB
B

2

B

1

C

0

B
B

-2

A
A

AB

A
AB
ABC

A
A
AB

B

B

-1

A

A
A

AB

C
C

C
C

B

BC
C

BC
C
C
C

BCD
CD

B
B
B

C
C

BC

C
C

D

-3
0

10

20

30

40
50
Time (sec)

60

70

80

90

Figure 2.5: Mean (± SD) E. coli O157:H7 populations in flume water during processing iceberg
lettuce inoculated at ~6 log CFU/g (n=3). Half the limit of detection was used to calculate the
mean log value when a sample did not yield any colonies by direct plating. Means of the same
product type with different letters are significantly different (P ≤ 0.05).

50

2.4.3 Centrifugation water. Using peroxyacetic acid, mixed peracid, or chlorine, wash
water exiting the centrifuge drain after spin-drying yielded maximum E. coli O157:H7
populations of 4.5, 4.4, and 5.5 log CFU/ml, respectively, which were not significantly different
(P > 0.05) from those in water alone (maximum population of 5.6 log CFU/ml) during the 40 s
sampling period. However, chlorine + CA and chlorine + T-128 resulted in E. coli O157:H7
populations that were lower than those in water alone (P ≤ 0.05) during the first 20 s of
centrifugation. Water samples collected after 40 s of centrifugation yielded E. coli O157:H7
populations that were not significantly different among any of the treatments (Figure 2.6).

51

Water
Chlorine

7

Peroxyacetic Acid
Chlorine + T-128

Mixed Peracid
Chlorine + CA

E. coli O157:H7 (log CFU/ml)

6
5

A
A
A
A

A

4

AB
ABC

A
A
A
A
A
A

A
A
AB
AB
AB
B

3
ABC
2
1
0

BC
C

B

-1

B

-2
0

10

20
Time (sec)

30

40

Figure 2.6: Mean (± SD) E. coli O157:H7 populations in spent centrifugation water from iceberg
lettuce inoculated at ~6 log CFU/g (n=3). Half the limit of detection was used to calculate the
mean log value when a sample did not yield any colonies by direct plating. Means of the same
product type with different letters are significantly different (P ≤ 0.05).

52

2.4.4 Processing equipment surfaces. After processing iceberg lettuce, all five sanitizer
treatments yielded significantly lower (P ≤ 0.05) E. coli O157:H7 populations remaining on the
flume tank and shaker table as compared to the water control. Significantly lower (P ≤ 0.05) E.
coli O157:H7 populations were recovered on the centrifugal dryer using peroxyacetic acid (3.6
2

2

log CFU/100 cm ) and mixed peracid (3.5 log CFU/100cm ) compared to the other treatments,
2

with the highest level (5 log CFU/100 cm ) seen when water alone was used for washing (Figure
2.7).

53

Water
Chlorine

E. coli O157:H7 (log CFU/100 cm2)

6

Peroxyacetic acid
Chlorine + T-128

Mixed Peracid
Chlorine + CA
A

5

A

A

A AB

BC C

4
A

3
B

B

2
B
B

1
0

B

B

B

B

B

B

-1
Flume Tank

Shaker Table

Dewatering Centrifuge

Equipment
Figure 2.7: Mean (± SD) E. coli O157:H7 populations on equipment surfaces after processing
iceberg lettuce inoculated at ~6 log CFU/g (n=3). Half the limit of detection was used to
calculate the mean log value when a sample did not yield any colonies by direct plating. Means
of the same product type with different letters are significantly different (P ≤ 0.05).

54

2.5 DISCUSSION
Due to the potential production of infectious aerosols during lettuce processing, the same
four non-toxigenic strains of E. coli O157:H7 were used as in earlier transfer studies (22, 23).
The growth and adherence rates for these four non-toxigenic strains were previously shown to be
similar to three strains from the 2006 leafy green outbreaks (22). As previously reported, GFPlabeling also allowed for easy differentiation of the inoculum from background bacteria (22, 23,
131).
Dip-inoculation of the lettuce to contain 6 log CFU/g was crucial to ensure uniform
distribution of E. coli O157:H7 throughout the heads as well as quantifiable results for
subsequent mathematical modeling, with this work to be reported elsewhere. While this
inoculation level clearly exceeds levels thought to occur on field-grown lettuce, feces from
“super-shedding” cows can potentially contain E. coli O157:H7 at levels of 6 log CFU/g (42),
with such fecal material potentially coming in contact with lettuce through irrigation water.
Preliminary experiments completed using a mixture of Glo-Germ™ and water showed uniform
fluorescence in dipped heads of iceberg lettuce. Additionally, Buchholz and others (22) found
that E. coli O157:H7 populations were statistically similar prior to processing and after shredding
(before washing), indicating that the inoculation was homogenous throughout the heads of
iceberg lettuce. Dip-inoculation of cored heads of lettuce may have allowed for the
internalization of the E. coli O157:H7 into lettuce through the damaged tissues, where it would
remain protected from the sanitizing agents (98). Since all lettuce samples were processed by
stomaching, any internalized cells would have gone undetected, with only the cells on the surface
of the leaves recovered.

55

Commercial producers of fresh-cut leafy greens use different sanitizers, sanitizer
concentrations, and contact times, depending on the design of the processing line. In this study,
six different wash treatments were assessed during 90 s of flume washing. Processing inoculated
iceberg lettuce resulted in E. coli O157:H7 reductions of 0.8 to 1.4 log CFU/g on the finished
product. Both during and after processing, no significant differences in sanitizer efficacy (P >
0.05) were seen against E. coli O157:H7 on iceberg lettuce for any of the treatments, including
water alone. However, three wash treatments- mixed peracid, chlorine and chlorine + CAsignificantly reduced (P ≤ 0.05) E. coli O157:H7 populations after washing. Numerous smallscale laboratory studies have shown similar pathogen reductions (~1 log CFU/g) during washing
of various fruits and vegetables with or without sanitizers (17, 21, 25, 134). Using a pilot-scale
leafy green processing line, Luo et al. (81) also reported an E. coli O157:H7 reduction of < 1 log
after processing inoculated baby spinach (81). Consequently, produce sanitizers cannot be relied
upon to ensure end-product safety. Chemical sanitizers are routinely added to recirculating wash
water to minimize the spread of microbial contaminants during flume washing (78). Regarding
their use, peroxyacetic acid-based sanitizers are limited to a maximum of 80 ppm peroxyacetic
acid (43, 62), while free chlorine concentrations typically range from 10 to a maximum of 200
ppm (55, 91, 122). However, soil, debris, and vegetable latexes released during shredding of
leafy greens will accumulate in the flume water over time (86), decreasing the efficacy of many
sanitizers, most notably chlorine (5, 77, 100, 139). The wash water used in this study contained
an organic load of ~0.0006% blended iceberg lettuce (w/v) to simulate wash water quality during
the early stages of processing. Hence, higher E. coli O157:H7 populations would have been
expected after 90 s of processing if the organic load in the wash water had been higher,
especially for the chlorine-based sanitizer. E. coli O157:H7 populations recovered from the wash
56

water were consistently lower (P ≤ 0.05) using chlorine, chlorine + CA and chlorine + T-128, as
compared to water alone, peroxyacetic acid and mixed peracid. Both chlorine + CA and chlorine
+ T-128 treatments yielded E. coli O157:H7 levels that were below the limit of detection, which
is similar to the findings of Lopez-Galvez et al (78) using 40 ppm chlorine.
This study was designed to assess the efficacy of sanitizers during processing - not longterm pathogen persistence in the wash water. Produce sanitizers are primarily used to minimize
cross-contamination during flume washing, with their effectiveness dependent on the type of
sanitizer, concentration, temperature, and organic load in the wash water. The pilot-scale
processing line used in this study was not equipped with a chiller. Therefore, all processing
needed to be conducted at our incoming tap water temperature of 12 to 15°C rather than at the
targeted commercial temperature of 4°C. Since sanitizer efficacy against E. coli O157:H7 is
enhanced at temperatures above 4oC (140), our E. coli O157:H7 reductions likely exceed those
that would be expected in commercial operations.
Levels of E. coli O157:H7 recovered from spent centrifugation water containing
sanitizers were rarely lower than those seen in sanitizer-free water. Similar E. coli O157:H7
populations were recovered from centrifugation water containing peroxyacetic acid, mixed
peracid, chlorine or no sanitizer at all four sampling times. The combination of chlorine and
citric acid or T-128 was significantly more effective than the other sanitizers (P ≤ 0.05) against
E. coli O157:H7 in centrifugation water collected during the first 20 s; however, after 40 s no
significant difference was seen compared to the water control (P > 0.05). These results indicate
that while populations of E. coli O157:H7 may be close to or below the limit of detection in
flume water, populations in the centrifugation water were not significantly different than the

57

water control by the end of sample collection. Therefore, spent centrifugation water would be
best suited for pathogen testing.
E. coli O157:H7 cells recovered from equipment surfaces after processing reflect those
that were present in the film of water on the equipment surface. During processing, the flume
tank was in continuous contact with the recirculating wash water, with water contact decreasing
during shaker table dewatering and centrifugal drying. Numbers of E. coli O157:H7 recovered
from surfaces in the centrifugal dryer were not significantly different from the water control
when any of the three chlorine-based sanitizer treatments were used, indicating that those
surfaces may also be well suited for pathogen testing, depending on the particular sanitizer used.
This study was done to assess the efficacy of commercial produce sanitizers against E.
coli O157:H7 on lettuce, in wash water, and on equipment surfaces during small-scale
processing of iceberg lettuce. While none of the sanitizers were more effective than water alone
against E. coli O157:H7 on leafy greens at any point during or after processing, it is important to
reiterate that sanitizers are designed to reduce the microbial load in wash water rather than on the
product. Overall, the populations of E. coli O157:H7 recovered in wash water containing
peroxyacetic acid or mixed peracid were rarely significantly different from those seen in water
alone. However, the three chlorine-based treatments were significantly more effective than water
alone at reducing E. coli O157:H7 populations in wash water during processing. The wash water
used in this study replicated a “best-case” scenario for processors due to the extremely low
organic load and freshly added sanitizers. Similar studies using higher organic loads will be
needed to assess sanitizer efficacy against E. coli O157:H7 under conditions that more closely
simulate commercial processing (Chapters 3 – 5).

58

CHAPTER 3:

Impact of Organic Load on Sanitizer Efficacy against Escherichia coli O157:H7 in
Simulated Leafy Green Processing Water

59

3.1 ABSTRACT
Chemical sanitizers are routinely used during commercial flume washing of fresh-cut
leafy greens to minimize cross-contamination from the water. This study assessed the efficacy of
five commercial sanitizer treatments against E. coli O157:H7 in wash water containing various
organic loads in a novel and cost-effective bench-top system. Iceberg lettuce (25 g) was
inoculated with a 4-strain cocktail of non-toxigenic, GFP-labeled, ampicillin-resistant E. coli
6

O157:H7 at10 CFU/g and stored for ~24 h at 4°C to simulate commercial storage conditions
before processing. The lettuce was then placed in a red mesh produce bag and washed for 90 s
using six different treatments (water alone, 50 ppm peroxyacetic acid, 50 ppm mixed peracid, or
50 ppm available chlorine either alone or acidified to pH 6.5 with citric acid (CA) or T-128) in 4
L of water containing organic loads of 0, 2.5, 5 or 10% (w/v blended iceberg lettuce) and
immediately removed from the water. Water (50 ml) samples were collected and neutralized at 2
min intervals for 10 min, diluted in phosphate buffer and plated on tryptic soy agar containing
0.6% yeast extract and 100 ppm ampicillin with or without prior 0.45 μm membrane filtration to
assess persistence of E. coli O157:H7. Various physicochemical parameters of the wash water
were correlated to E. coli O157:H7 inactivation at each organic load. Organic load negatively
impacted the efficacy of chlorine, chlorine + CA, and chlorine + T-128 (P ≤ 0.05), with typical
E. coli O157:H7 reductions of < 1 log CFU/ml after 10 min of exposure. However, the efficacy
of peroxyacetic acid and mixed peracid was unaffected by organic load (P > 0.05) with average
E. coli O157:H7 reductions of ~4.8 and ~5.5 log CFU/ml, respectively, after 10 min of exposure.
Reduced sanitizer efficacy generally correlated to increased total solids, chemical oxygen
demand, turbidity, and decreased maximum filterable volume, indicating that these tests may be
effective alternatives to the industry standard of oxygen/reduction potential.
60

3.2 INTRODUCTION
Numerous small-scale laboratory studies have been conducted to assess sanitizer efficacy
against pathogens on leafy greens (4, 17, 66, 72, 78, 82, 88, 115, 139, 140). However, the
findings from these studies have not provided a means by which commercial processors can
predict the efficacy of their wash water based upon water quality. Previous work completed by
our group was performed without chemical sanitizers to quantify E. coli O157:H7 transfer during
pilot-plant production of fresh-cut leafy greens (22, 23), or to assess the efficacy of commercial
produce sanitizers against E. coli O157:H7 during pilot-plant production of iceberg lettuce using
water without an organic load (Chapter 2). Since sanitizers are designed to minimize crosscontamination during washing, it is necessary to determine the extent of E. coli O157:H7
persistence in recirculating wash water containing various levels of organic solids that are
generated during production. Consequently, this study aimed to 1) determine the efficacy of five
commercial sanitizer treatments against E. coli O157:H7 in simulated processing water
containing various organic loads in a novel bench-top model and 2) assess the relationship
between various physicochemical parameters and organic load of the wash water on E. coli
O157:H7 inactivation.

61

3.3 MATERIALS AND METHODS
3.3.1 Experimental design. The impact of four different organic loads on five different
sanitizing treatments against E. coli O157:H7 was assessed in triplicate by washing 25 g of
6

iceberg lettuce inoculated at 10 CFU/g in a 4 L glass carboy, with sanitizer-free water serving
as the control for each organic load. Thereafter, various water samples were collected and
quantitatively examined for E. coli O157:H7 to determine persistence. Additional trials were
done to assess the impact of water temperature (~5°C vs. 14oC) on acidified sodium hypochlorite
efficacy using the same experimental design. Log linear inactivation trend lines for E. coli
O157:H7 inactivation were correlated to seven physicochemical parameters of the wash water temperature, pH, oxidation/reduction potential, chemical oxygen demand, total solids, maximum
filterable volume and turbidity.
3.3.2 Leafy greens. Identical to 2.3.2
3.3.3 Bacterial strains. Identical to 2.3.3
3.3.4 Lettuce inoculation. A 0.2 ml aliquot of each non-toxigenic E. coli O157:H7
culture was transferred to 9 ml of TSBYE with amp and incubated for 18 – 20 h at 37°C. Based
on similar growth rates as determined previously (22), the four strains were combined in equal
volumes to obtain a 20-ml cocktail which was added to 2 L of municipal tap water (~15oC, <
0.05 ppm free chlorine) in a sterile 2 L polypropylene container (Nalgene, Rochester, NY) to
7

achieve a level of ~10 CFU/ml. Hand-torn pieces of iceberg lettuce (150 g) were immersed in
the E. coli O157:H7 suspension for 15 min in a 28-cm long red mesh produce bag (5 lb header
bag, Pacon Inc., Baldwin Park, CA). The lettuce was then drained/air-dried for 15 min at 22°C
before being spun in a salad spinner (Model 32480V2, OXO, Chambersburg, PA) by hand62

pumping 5 times to remove residual inoculum, then transferred from the mesh bag into a WhirlPak™ bag (Nasco, Fort Atkinson, WI) and stored for 20 – 24 h at 4°C before use. A 25 g
sample was then aseptically collected to determine the initial inoculation level at the time of
processing.
3.3.5 Processing equipment. A 4 L Kimax glass carboy with a 5 cm diameter opening
at the top and a 0.64 cm diameter discharge spout at the bottom (Kimble Chase, Vineland, NJ)
plugged by a cork was used. Before the start of each experiment, a 8-cm long magnetic stir bar
was inserted into the carboy, which was placed on a stirring hotplate (Model LMS-100, Daihan
Labtech Co., Ltd., Korea) at a stir speed of 6 (Figure 3.1).

63

Figure 3.1: Bench-top system developed to simulate commercial flume washing of iceberg
lettuce.

64

3.3.6 Wash water. Iceberg lettuce (0, 100 , 200, or 400 g) was blended for 2 min in 250
– 500 ml of tap water using a household blender (Model BLC10650MB, Black & Decker, New
Britain, CT) and added to the wash water at ~14°C to achieve organic load levels of 0, 2.5, 5, or
10% (w/v), respectively, in 4 L. The following five commercial produce sanitizer treatments
were assessed: 50 ppm peroxyacetic acid (Tsunami 100, Ecolab, St. Paul, MN), 50 ppm mixed
peracid (Tsunami 200, Ecolab), 50 ppm available chlorine (XY-12, Ecolab) (pH 7.09 – 8.03), 50
ppm available chlorine (XY-12) acidified to pH 6.5 with citric acid (Sigma-Aldrich, St. Louis,
MO), and 50 ppm available chlorine (XY-12) acidified to pH 6.5 with T-128 (SmartWash
Solutions) as measured with a pH probe (pHTestr 30, Oakton, Vernon Hills, IL). Peroxyacetic
acid test kit 311 (Ecolab) was used to confirm the peroxyacetic acid and mixed peracid sanitizer
concentrations while chlorine test kit 321 (Ecolab) was used to measure available chlorine. Free
chlorine was measured immediately before exposing iceberg lettuce to the water using a Pocket
Colorimeter (Model 46770-00, Hach Company, Loveland, CO). Sanitizer-free MSU tap water (<
0.05 ppm free chlorine) served as the control for each for each organic load. Chlorine + CA
efficacy was also assessed at ~5°C by refrigerating MSU tap water ~12 h before use, with all
other aspects of the experimental design remaining the same.
3.3.7 Leafy green washing and sample collection. Inoculated iceberg lettuce (25 g) was
aseptically transferred to a 28-cm long red mesh produce bag that was knotted at one end. The
lettuce was inserted into the carboy through the top after which the cork was removed from the
spout. Following 90 s of exposure to the wash water with mechanical stirring, the spout was
closed, after which the bag of lettuce was removed from the carboy and added to 100 ml of
sterile Difco Neutralizing Buffer (Becton Dickson, Franklin Lakes, NJ) in a Whirl-Pak™ filter
bag (Nasco, Fort Atkinson, WI). After removing the lettuce, magnetic stirring of the wash water
65

continued for an additional 10 min, during which time water samples (50 ml) were collected at 2
min intervals for 10 min in 50 ml centrifuge tubes containing 38x concentrated Difco
Neutralizing Buffer (Becton Dickinson).
3.3.8 Physiochemical parameters. Seven physicochemical parameters of the wash water
were monitored before iceberg lettuce washing. Temperature and pH were measured with a
probe (pHTestr 30, Oakton, Vernon Hills, IL). Oxidation/reduction potential (ORP) was
monitored with a probe (ORPTestr 10, Oakton). Chemical oxygen demand (COD) was measured
using Hach Digestion Reactor Method 8000 (Hach, Loveland, CO) (59). Total solids was
measured after drying 10 ml of wash water in a pre-heated and pre-weighed crucible in a drying
oven (Model 625-A, Precision Scientific Inc, Chicago, IL) at 103 ± 2°C for 2 h (14). Turbidity
was assessed by pouring a 50 ml water sample through a 24-cm diameter Whatman Filter (Grade
113, Piscataway, NJ) to remove suspended solids followed by measuring absorbance of a 1 ml
aliquot of the filtrate at 663 nm in a spectrophotometer (Model SB-100XR, Spectronics
Corporation, Westbury, NY) (133). Maximum filterable volume (MFV) was quantified as the
volume of a 50 ml water sample pulled through a 0.45 µm membrane in 1 min at-80 kPa using a
vacuum pump (Model WP6211560, Millipore, Billerica, MA). Free chlorine was measured using
a pocket colorimeter (Model 46770, Hach).
3.3.9 Microbiological analyses. All lettuce samples (25 g) were homogenized in a
stomacher (Stomacher 400 Circulator, Seward, Worthington, UK) for 1 min at 260 rpm,
appropriately diluted in sterile 1% (w/v) phosphate buffer (8.5 g/L NaCl, 1.44 g/L Na2HPO4,
and 0.24 g/L KH2PO4, J.T. Baker, Mallinckrodt Baker Inc., Phillipsburg, NJ) and plated on
TSAYE with amp (calculated minimum detection limit of 40 CFU/g). The 50 ml water samples
66

were either appropriately diluted in sterile 1% phosphate buffer and plated on TSAYE with amp
(calculated minimum detection limit of 10 CFU/ml) or processed using 0.45 μm membrane
filters (Millipore, Millipore Corporation, Billerica, MA) (calculated minimum detection limits of
0.02 CFU/ml, 0.033 CFU/ml, 0.1 CFU/ml or 0.1 CFU/ml for organic loads of 0, 2.5, 5 and 10%
(w/v), respectively), which were placed on 60-mm dia. petri plates containing TSAYE with amp.
Following 40 – 48 h of incubation at 37°C, all green fluorescing colonies as seen under
ultraviolet light (365 nm, Blak-Ray, Ultra-violet Product Inc. San Gabriel, CA) were counted as
E. coli O157:H7.
3.3.10 Sanitizer neutralization confirmation. One liter aliquots of tap water containing
50 ppm chlorine (XY-12, Ecolab), 50 ppm chlorine + CA, 50 ppm chlorine + T-128, 50 ppm
peroxyacetic acid (Tsunami 100, Ecolab) or 50 ppm mixed peracid (Tsunami 200 ppm, Ecolab)
with organic loads of 0, 2.5, 5 or 10% were prepared using a mechanical blender (Model
BLC10650MB, Black & Decker, New Britain, CT), after which the concentration of available
chlorine or peroxyacetic acid was confirmed using a commercial test kit (Chlorine test kit 321,
Ecolab and Peroxyacetic acid test kit 311, Ecolab). A 50 ml centrifuge tube containing 3 ml of
38x concentrated neutralizing buffer (BD) was filled with these prepared solutions, agitated for 5
s and then similarly assessed for chlorine or peroxyacetic acid neutralization. In preliminary
experiments, a 38x concentration neutralized all sanitizer concentrations without impacting E.
coli O157:H7 recovery.
3.3.11 Statistical analysis. E. coli O157:H7 counts were converted to log CFU per g or
ml, and individual physicochemical parameter values were subjected to an ANOVA using JMP
9.0 (SAS Institute Inc., Cary, NC). For samples without E. coli O157:H7 counts, values equaling
half the limit of detection were used. A P value of ≤ 0.05 was considered significant for all tests.
67

The Tukey-Kramer HSD test was used to identify significant differences in E. coli O157:H7
populations for individual lettuce and water samples as well as individual physicochemical
parameter values. Microsoft Excel software was also used to conduct a log linear regression
analysis and calculate 95% confidence intervals for the E. coli O157:H7 reductions in water over
time. The following log linear equation was used:
Log(N/N0) = -k*t
Where Log(N/N0) is the microbial reduction, t is the time in minutes, and k is the inactivation
2

rate. Excel was used to determine the P and R values, and JMP 9.0 software was used to
determine the RMSE (log CFU/ml) by running the curves for each organic load used and
sanitizer combination.

68

3.4 RESULTS
3.4.1 Lettuce. In general, E. coli O157:H7 population decreased 0.5 to 1.0 log CFU/g
regardless of the sanitizer treatment or organic load. Organic load did not impact the efficacy of
peroxyacetic acid or mixed peracid, resulting in E. coli O157:H7 reductions that were not
significantly different (P > 0.05) on iceberg lettuce regardless of organic load (Table 3.1).
Organic load significantly decreased (P ≤ 0.05) the efficacy of chlorine, chlorine + CA, and
chlorine + T-128, with reductions of ~1 log CFU/g achieved without an organic load while
organic loads of 2.5, 5 and 10% yielded reductions of ~0.5 log CFU/g. No significant difference
in E. coli O157:H7 reduction was observed for any of the organic loads using chlorine + CA at
5°C.

69

Table 3.1: E. coli O157:H7 reduction on iceberg lettuce inoculated at ~6 log CFU/g after
processing

a

Mean ± SD E. coli O157:H7 reduction (log CFU/g)

0

1.1 ± 0.1 A

Chlorine + Chlorine +
CA (pH
T-128 (pH
6.5)
6.5)
1.2 ± 0.2 A 1.0 ± 0.1 A 1.0 ± 0.1 A 0.9 ± 0.1 A

2.5

1.0 ± 0.0 A

0.9 ± 0.3 A 0.5 ± 0.1 B 0.4 ± 0.1 B

0.5 ± 0.0 B

0.5 ± 0.2 A

5

0.9 ± 0.2 A

0.9 ± 0.3 A 0.5 ± 0.3 B 0.6 ± 0.2 B

0.5 ± 0.1 B

0.5 ± 0.2 A

10

1.0 ± 0.1 A

1.0 ± 0.2 A 0.4 ± 0.1 B 0.5 ± 0.0 B

0.5 ± 0.1 B

0.5 ± 0.1 A

% Organic
load (w/v)

a

Peroxyacetic
acid

Mixed
Peracid

Chlorine

Chlorine +
CA (5°C)
(pH 6.5)
0.9 ± 0.3 A

Means with different capital letters designate E. coli O157:H7 reductions that differ

significantly in terms of sanitizing treatment (P ≤ 0.05).

70

3.4.2 Wash water. Average E. coli O157:H7 populations in sanitizer-free wash water
were ~3.53 log CFU/ml at all time points (Figure 3.2). Minimal E. coli O157:H7 persistence was
seen in wash water for both peroxyacetic acid and mixed peracid, regardless of organic load,
with most populations recovered after 4 min being at or below the limit of detection (0.02 to 0.1
log CFU/ml) (Table 3.2). Efficacy of peroxyacetic acid was not significantly influenced by
organic load (P > 0.05), with an average reduction of 4.8 log CFU/ml seen at the end of
sampling. Organic load negatively impacted the efficacy of mixed peracid, with greater (P ≤
0.05) E. coli O157:H7 reductions seen after 10 min at organic loads of 0 and 2.5% (5.2 log
CFU/ml) as compared to 5 and 10% (5.8 log CFU/ml). Organic load also decreased the efficacy
of chlorine, with greater E. coli O157:H7 reductions (P ≤ 0.05) seen with a 0% organic load than
all other organic loads after 0 (5.2 log CFU/ml), 4 (5.4 log CFU/ml), and 10 min (5.1 log
CFU/ml) (Table 3.2). Similar results were obtained for chlorine + CA and chlorine + T-128, with
significantly greater E. coli O157:H7 reductions (4.9 – 5.4 log CFU/ml) seen at all times points
using a 0% organic load.

71

Figure 3.2: Mean (± SD) E. coli O157:H7 populations (log CFU/ml) in water containing various
organic loads (n=3 per organic load) after washing iceberg lettuce inoculated at ~6 log CFU/g.
Means of the same product type with different letters are significantly different (P ≤ 0.05).

72

Table 3.2: E. coli O157:H7 reduction in the wash water at ~14°C

a

Mean ± SD E. coli O157:H7 reduction in wash water
b
(log CFU/ml)
% Organic
Load (w/v)
0

0 min

4 min

10 min

2.1 ± 0.2 A

5.2 ± 0.4 A

5.4 ± 0.0 A

2.5

1.5 ± 0.7 A

4.2 ± 1.2 A

4.2 ± 1.1 A

5

1.9 ± 0.5 A

4.6 ± 0.2 A

4.5 ± 0.1 A

10

2.3 ± 0.3 A

4.9 ± 0.2 A

4.9 ± 0.2 A

0

4.3 ± 0.9 A

5.1 ± 0.5 A

5.3 ± 0.2 A

2.5

3.7 ± 0.3 A

5.1 ± 0.1 A

5.2 ± 0.1 A

5

3.2 ± 0.2 A

4.8 ± 0.0 A

4.7 ± 0.1 B

10

3.2 ± 0.6 A

4.8 ± 0.1 A

4.8 ± 0.1 B

0

5.2 ± 0.6 A

5.4 ± 0.1 A

5.1 ± 0.5 A

2.5

-0.2 ± 0.1 B

-0.1 ± 0.1 B

0.8 ± 0.3 B

5

-0.2 ± 0.2 B

-0.2 ± 0.2 B

-0.2 ± 0.1 C

10

0.0 ± 0.1 B

0.1 ± 0.2 B

0.1 ± 0.1 BC

0

4.9 ± 0.8 A

5.2 ± 0.3 A

5.4 ± 0.0 A

2.5

Sanitizing Treatment

-0.1 ± 0.2 B

0.3 ± 0.3 B

1.7 ± 1.0 B

5

-0.2 ± 0.2 B

-0.1 ± 0.0 B

-0.0 ± 0.2 C

10

0.0 ± 0.1 B

0.2 ± 0.1 B

0.2 ± 0.1 C

Peroxyacetic Acid

Mixed Peracid

Chlorine

Chlorine + CA

73

Table 3.2 (cont’d)
0

5.4 ± 0.2 A

5.1 ± 0.4 A

5.0 ± 0.6 A

2.5

-0.3 ± 0.2 B

0.1 ± 0.2 B

1.2 ± 0.5 B

5

-0.2 ± 0.2 B

-0.2 ± 0.1 B

-0.3 ± 0.3 C

10

0.1 ± 0.1 B

0.1 ± 0.1 B

0.2 ± 0.1 BC

Chlorine + T-128

a

Means with different capital letters designate E. coli O157:H7 reductions that differ

significantly in terms of sanitizing treatment and time (P ≤ 0.05).
b

Time in minutes of E. coli exposure to the wash water after removal of the inoculated lettuce

from the system.

74

3.4.3 Impact of temperature on sanitizer efficacy. Temperature rarely impacted
chlorine + CA efficacy against E. coli O157:H7 in wash water, regardless of exposure time or
organic load. Regardless of temperature, consistently greater E. coli O157:H7 reductions were
seen for an organic load of 0% (4.9 – 5.4 log CFU/ml) as opposed to 2.5, 5 and 10% (Table 3.3).
The only instance in which temperature influenced chlorine + CA efficacy against E. coli
O157:H7 occurred with a 10% organic load after 10 min of exposure, where a greater reduction
(P ≤ 0.05) was observed using water at 5°C (0.4 log CFU/ml) as opposed to 14°C (0.2 log
CFU/ml). However, neither of these reductions was significantly different from those seen for
both temperatures at 2.5 and 5% organic loads at the same time point.

75

Table 3.3: E. coli O157:H7 reduction in the wash water containing chlorine + CA at two
temperatures

a

Mean ± SD E. coli O157:H7 reduction in wash water (log
b
CFU/ml)
% Organic
Load (w/v)

Temperature (°C)
13

0 min
4.9 ± 0.8 A a

4 min

10 min

5.2 ± 0.3 A a

5.4 ± 0.0 A a

0
5

5.3 ± 0.1 A a

5.4 ± 0.2 A a

5.3 ± 0.1 A a

14

-0.1 ± 0.2 B a

0.3 ± 0.3 B a

1.7 ± 1.0 B a

5

0.4 ± 0.4 B a

0.9 ± 1.9 B a

1.5 ± 1.8 B a

14

-0.2 ± 0.2 B a

-0.1 ± 0.0 B a

0.0 ± 0.2 B a

5

0.0 ± 0.1 B a

-0.1 ± 0.2 B a

0.0 ± 0.1 B a

14

0.0 ± 0.1 B a

0.2 ± 0.1 B a

0.2 ± 0.1 B b

5

0.3 ± 0.1 B a

0.4 ± 0.2 B a

0.4 ± 0.1 B a

2.5

5

10
a

Means with different same capital letters designate E. coli O157:H7 reductions that differ

significantly in terms of time, while means with different lowercase letters differ significantly in
terms of temperature for each organic load (P ≤ 0.05).
b

Time in minutes of E. coli exposure to the wash water after removal of the inoculated lettuce

from the system.

76

3.4.4 Linear regression of E. coli O157:H7 reduction in wash water. E. coli O157:H7
populations were typically below the lower limit of detection for all organic loads when
peroxyacetic-acid based sanitizers were used, resulting in intercepts ranging from -2.8 to -3.7 log
CFU/ml for peroxyacetic acid and -3.9 to -4.8 log CFU/ml for mixed peracid. Slopes were
similar for the different organic loads. Root-mean-square errors were ~1 log CFU/ml for
2

peroxyacetic acid and ~0.5 log CFU/ml for mixed peracid. R values for both peroxyacetic acid
and mixed peracid were typically between 0.4 – 0.5 (Figures 3.3 and 3.4, Table 3.4). E. coli
O157:H7 populations were typically near, at, or below the lower limit of detection for all
chlorine variations with a 0% organic load in the wash water, resulting in intercepts of -5.4, -5.1,
-5.5 and -5.5 for chlorine, chlorine + CA, chlorine + T-128 and chlorine + CA at 5°C,
respectively. Using organic loads of 2.5, 5 and 10%, intercepts were close to 0, with inactivation
rates at organic loads of 5 and 10% being close to 0, while a 2.5% organic load yielded
inactivation rates of -0.1, -0.2, -0.1 and -0.1 log CFU/min for chlorine, chlorine + CA, chlorine +
T-128, and chlorine + CA at 5°C, respectively. Root-mean-square errors ranged from 0 to 0.3 log
2

CFU/ml for all chlorine treatments. R values for all chlorine treatments ranged from 0 to 1, with
the higher values attributed to the 2.5% organic load. (Figures 3.5 – 3.8, Table 3.4).

77

Figure 3.3: E. coli O157:H7 inactivation (log CFU/ml) in water containing various organic
loads and 50 ppm peroxyacetic acid (n=3 per organic load) after washing iceberg lettuce
inoculated at ~6 log CFU/g. Half the limit of detection was used when a sample did not yield any
colonies by direct plating.

78

Figure 3.4: E. coli O157:H7 inactivation (log CFU/ml) in water containing various organic
loads and 50 ppm mixed peracid (n=3 per organic load) after washing iceberg lettuce
inoculated at ~6 log CFU/g. Half the limit of detection was used when a sample did not yield any
colonies by direct plating.

79

Figure 3.5: E. coli O157:H7 inactivation (log CFU/ml) in water containing various organic
loads and 50 ppm available chlorine (n=3 per organic load) after washing iceberg lettuce
inoculated at ~6 log CFU/g. Half the limit of detection was used when a sample did not yield any
colonies by direct plating.

80

Figure 3.6: E. coli O157:H7 inactivation (log CFU/ml) in water containing various organic
loads and 50 ppm available chlorine + CA (n=3 per organic load) after washing iceberg lettuce
inoculated at ~6 log CFU/g. Half the limit of detection was used when a sample did not yield any
colonies by direct plating.

81

Figure 3.7: E. coli O157:H7 inactivation (log CFU/ml) in water containing various organic
loads and 50 ppm available chlorine + T-128 (n=3 per organic load) after washing iceberg
lettuce inoculated at ~6 log CFU/g. Half the limit of detection was used when a sample did not
yield any colonies by direct plating.

82

Figure 3.8: E. coli O157:H7 inactivation (log CFU/ml) in 5°C water containing various organic
loads and 50 ppm available chlorine + CA (n=3 per organic load) after washing iceberg lettuce
inoculated at ~6 log CFU/g. Half the limit of detection was used when a sample did not yield any
colonies by direct plating.

83

Table 3.4: E. coli inactivation trend lines in wash water containing various organic loads

Log Reduction

R

0

Log(N/N0) = -0.2t - 3.7

2.5

RMSE (log
CFU/ml)

P Value

0.42

1.1

0.16

Log(N/N0) = -0.2t - 2.8

0.44

1.1

0.15

5

Log(N/N0) = -0.2t - 3.1

0.42

1.1

0.16

10

Log(N/N0) = -0.2t - 3.5

0.44

1.1

0.15

0

Log(N/N0) = -0.1t - 4.8

0.29

0.5

0.27

2.5

Log(N/N0) = -0.1t - 4.4

0.44

0.5

0.15

5

Log(N/N0) = -0.1t – 4.0 0.41

0.6

0.17

10

Sanitizing Agent

% Org.
Load (w/v)

a

Log(N/N0) = -0.1t - 3.9

0.45

0.6

0.15

0.09

0.2

0.56

Log(N/N0) = -0.1t + 0.4 0.89

0.2

< 0.001

0.32

0.0

0.24

0.80

0.0

0.02

2

Peroxyacetic acid

Mixed Peracid

0
2.5

b

Log(N/N0) = -5.4

Chlorine
5
10

Log(N/N0) = 0.2

b

Log(N/N0) = 0.0

84

Table 3.4 (cont’d)
0
2.5

Log(N/N0) = -5.1

0.33

0.3

0.24

Log(N/N0) = -0.2t + 0.3 0.98

0.1

< 0.001

Chlorine + CA
5

Log(N/N0) = +0.2

0.82

0.0

0.01

10

Log(N/N0) = -0.1

0.68

0.1

0.04

0

Log(N/N0) = -5.5

0.22

0.21

0.35

Log(N/N0) = -0.1t + 0.3 0.96

0.12

< 0.001

2.5
Chlorine + T-128
5

Log(N/N0) = 0.2

0.00

0.07

0.95

10

Log(N/N0) = 0.0

0.56

0.0

0.09

0

Log(N/N0) = -5.45

0.37

0.1

0.20

2.5

Log(N/N0) = -0.1t - 0.4

0.95

0.1

< 0.001

5

Log(N/N0) = +0.1

0.53

0.1

0.10

10

Log(N/N0) = -0.2

0.50

0.1

0.12

Chlorine + CA (5°C)

a

Inactivation rates were determined by using Microsoft Excel to create the linear regression

trend lines for E. coli O157:H7 log reduction over time (t) in minutes.
b

The slope (k) was < 0.1 and was therefore not included.

85

3.4.5 Physicochemical parameters of wash water. Organic load had no significant
effect (P > 0.05) on the average pH values for peroxyacetic acid (6.79) mixed peracid (5.32)
(Table 3.5). A significant reduction (P ≤ 0.05) in pH was seen for each increase in organic load
when chlorine was not acidified, with pH values decreasing from 8.03 (0%) to 7.09 (10%) (Table
3.6). MSU tap water was ~14°C, with some significant differences (P ≤ 0.05) observed between
different organic loads for the same sanitizing treatment, with the lower temperature often
observed at the lower organic loads. For all three chlorine treatments, free chlorine was
significantly higher (P ≤ 0.05) at a 0% organic load (50 – 53 ppm) compared to the other organic
loads (0.77 – 1.9 ppm), with available chlorine having stabilized at 50 ppm before the inoculated
lettuce was exposed to the water (Tables 3.6 – 3.8). Total solids, COD and turbidity significantly
increased (P ≤ 0.05) with increasing organic load, whereas MFV decreased significantly (P ≤
0.05) as organic load increased for each sanitizing treatment. As the organic load increased with
peroxyacetic acid or mixed peracid, ORP increased significantly (P ≤ 0.05), likely due to the
increase in peroxyacetic acid sanitizer added to the water to achieve a concentration of 50 ppm.
Conversely, ORP decreased significantly (P ≤ 0.05) as organic load increased for the chlorine
treatments.

86

Table 3.5: Physicochemical parameters of wash water containing various organic loads and 50 ppm peroxyacetic acid or 50 ppm
mixed peracid

a

Controlled
Physicochemical
Parameters
%
Sanitizing
Org.
agent
Load
(w/v)
0
Peroxyacetic
acid

2.5
5
10

0
Mixed
Peracid

2.5
5
10

a

Mean ± SD Dependent Physicochemical Parameters

pH
6.89 ±
0.08 A
6.79 ±
0.08 A
6.79 ±
0.01 A
6.69 ±
0.16 A
5.40 ±
0.05 A
5.35 ±
0.06 A
5.33 ±
0.12 A
5.19 ±
0.12 A

T (°C)

Total Solids (g)

COD (mg
O2/L)

Turbidity (abs
@ 663 nm)

MFV
(ml)

ORP (mV)

13±0 B

0.0061±0.0012 D

282±18 D

0±0 C

50±0 A

355±5 B

14±0 A

0.0190±0.0034 C

1703±41 C

0.082±0.033 BC

24±7 B

362±4 B

15±0 A

0.0288±0.0020 B

3787±160 B

0.142±0.029 B

15±4 BC

371±5 B

15±0 A

0.0502±0.0009 A

6890±301 A

0.275±0.050 A

10±2 C

404±18 A

14±0 A

0.0087±0.0006 D

874±21 D

0.000±0.000 D

50±0 A

479±4 B

15±1 A

0.0180±0.0009 C

2917±513 C

0.100±0.005 C

27±6 B

488±4 B

15±1 A

0.0312±0.0010 B

5100±304 B

0.199±0.013 B

17±5 BC

497±5 AB

15±1 A

0.0522±0.0041 A

7783±455 A

0.396±0.014 A

14±5 C

515±19 A

Means with different capital letters designate physicochemical parameters that differ significantly in terms of organic load (P ≤

0.05).
87

Table 3.6: Physicochemical parameters of wash water containing various organic loads and 50 ppm available chlorine
Controlled
Physicochemical
Parameters
% Org. Load
(w/v)

a

Mean ± SD Dependent Physicochemical Parameters

pH

T (°C)

Free
chlorine
(ppm)
53.33±2.08
A

Total solids (g)

COD (mg
O2/L)

Turbidity (abs
@ 663 nm)

MFV
(ml)

ORP
(mV)

8.03 ±
0.0061±0.0002
13±0 C
15±1 D
0.001±0.006 C 50±0 A 788±9 A
0.04 A
D
7.52 ±
0.0173±0.0026
0.081±0.018
437±29
2.5
15±1 AB
1.17±0.64 B
1208±286 C
18±7 B
0.02 B
C
BC
B
7.41 ±
0.0257±0.0010
14±2
5
14±0 BC
1.27±0.47 B
2340±60 B 0.133±0.005 B
413±7 B
0.01 C
B
BC
7.09 ±
0.0492±0.0017
360±70
10
15±1 A
1.00±0.44 B
5465±420 A 0.300±0.065 A 8±1 C
0.04 D
A
B
a
Means with different capital letter designate physicochemical parameters that differ significantly in terms of organic load (P ≤ 0.05).
0

88

Table 3.7: Physicochemical parameters of wash water containing various organic loads and 50 ppm chlorine + CA or 50 ppm
a

chlorine + T-128, both used to acidify to pH 6.5
Controlled
Physicochemical
Mean ± SD Dependent Physicochemical Parameters
Parameters
%
Org.
Free chlorine
COD (mg
Turbidity (abs
Supplement
T (°C)
Total solids (g)
Load
(ppm)
O2/L)
@ 663 nm)
(w/v)
0
13±0 B
50.67±1.53 A 0.0078±0.0003 D
244±32 B
0.001±0.001 C

MFV
(ml)

ORP
(mV)

50±0 A

903±12 A

2.5

1426±318 B

0.079±0.023 B

17±1 B

492±14 B

5

14±0 A

0.90±0.20 B

0.0296±0.0040 B

2243±91 B

0.123±0.005 B

14±5 BC

491±3 B

14±0 A

1.27±0.57 B

0.0492±0.0031 A

5815±306 A

0.255±0.032 A

10±1 C

425±32 C

13±0 B

51.67±1.53 A

0.0086±0.0002 D

278±17 D

0.001±0.001 D

50±0 A

902±6 A

2.5

14±1 AB

0.27±0.12 B

0.0181±0.0014 C

1323±160 C

0.081±0.010 C

17±4 B

482±1 B

5

14±0 AB

0.47±0.12 B

0.0275±0.0010 B

2073±64 B

0.128±0.013 B

17±6 B

495±23 B

10
a

0.0188±0.0026 C

0

T-128

0.77±0.38 B

10

CA

14±1 AB

15±1 A

0.67±0.15 B

0.0454±0.0060 A

5733±329 A

0.282±0.026 A

9±1 B

427±60 B

Means with different capital letters designate physicochemical parameters that differ significantly in terms of organic load (P ≤

0.05).

89

Table 3.8: Physicochemical parameters of wash water containing various organic loads and 50 ppm available chlorine + CA at ~5°C
a

Controlled
Physicochemical
Parameters

Mean ± SD Dependent Physicochemical Parameters

% Org. Load
(w/v)

Free chlorine
(ppm)

Total solids (g)

COD (mg O2/L)

Turbidity (abs @
663 nm)

MFV
(ml)

ORP (mV)

0

5±0A

50 ± 1.5 A

0.0090 ± 0.0075 D

296 ± 13 C

0±0C

50 ± 0 A

881 ± 10 A

2.5

6±0A

0.97 ± 0.57 B

0.0194 ± 0.0009 C

1518 ± 64 C

0.077 ± 0.060 BC

18 ± 5 B

563 ± 51 B

5

6±0A

0.77 ± 0.25 B

0.0306 ± 0.0010 B

3717 ± 990 B

0.178 ± 0.050 AB

19 ± 2 B

488 ± 72 B

10
a

T (°C)

5±0A

1.9 ± 0.1 B

0.0494 ± 0.0006 A

6557 ± 549 A

0.236 ± 0.041 A

17 ± 1 B

482 ± 97 B

Means with different capital letters designate physicochemical parameters that differ significantly in terms of organic load (P ≤

0.05).

90

3.5 DISCUSSION
The impact of organic load on sanitizer efficacy has been assessed previously in other
small-scale laboratory studies (57, 87, 107, 139). However, this is the first study to correlate
wash water physicochemical parameters to E. coli O157:H7 inactivation trend lines for a variety
of commercial produce sanitizers, organic loads and temperatures, with such efforts identified as
a priority by the Center for Produce Safety (33). This novel bench-top system was designed to
simulate the wash phase of a pilot-scale or commercial leafy green processing line. The simple
and economical carboy system presented here could be used to easily assess a range of sanitizers,
pathogens, processing conditions, and produce products.
The same four non-toxigenic strains of E. coli O157:H7 were used as in our earlier
transfer studies (22, 23) as well as in our pilot-scale studies (Chapters 2, and 4). Shen and others
(107) found that inactivation of E. coli O157:H7, Salmonella, and non-O157:H7 Shiga toxinproducing E. coli in chlorinated water was a function of different pathogenic strains. However,
most strains decreased 4.5 log CFU/ml (below the limit of detection) after 30 s of exposure when
free chlorine concentrations exceeded 0.5 ppm, indicating that while there may be some
differences in how strains of the same pathogen respond to sanitizing agents under various
conditions, these differences are likely to be noticeable only initially and are likely negligible in
experiments of longer duration. Preliminary work was completed to assess the extent of injury
for each individual strain in 30 ppm peroxyacetic acid after 90 s of exposure. Results indicated
immense variability after incubation at 37°C for 24 h, with far more consistent results obtained
after 48 h of incubation at the same temperature. Extending the incubation time no doubt allowed
for recovery of additional injured cells as previously reported (68, 135).

91

Dip inoculating lettuce to contain 6 log CFU/g was crucial to ensure a consistent
inoculation level. While this inoculation level clearly exceeds expected levels on field-grown
lettuce, feces from “super-shedding” cows can contain E. coli O157:H7 at levels greater than 6
log CFU/g (42), with such fecal material potentially coming in contact with lettuce through
irrigation water.
Lettuce was stored for 24 h at 4°C after inoculation to simulate industrial storage
conditions that occur before processing, which can promote further attachment of E. coli
O157:H7 to the leaf surface. A previous study completed by Buchholz and others (22) found that
42.2% of the original inoculum remained on product that was held for 24 h at 4oC before
washing and centrifugal drying, significantly higher than the 8% retention rate seen for product
that had been identically processed 1 h after inoculation.
The inoculated lettuce was in a mesh produce when exposed to 4 L of water to allow for
quick removal from the water, simulating lettuce leaving a commercial flume system. E. coli
O157:H7 reductions on the lettuce ranged from 0.4 – 1.2 log CFU/g. The fact that these
reductions were on the lower end of the spectrum when compared to other small-scale laboratory
studies (1 – 3 log CFU/g) (17, 47, 55, 91, 100)can be attributed to the lettuce being bagged,
which likely hindered exposure to higher shear forces as would have been encountered by loose
pieces. (62, 132, 141). Organic load had a significant impact (P ≤ 0.05) on the efficacy of
chlorine, chlorine + CA and chlorine + T-128, with greater E. coli O157:H7 reductions seen with
an organic load of 0% (~ 1 log CFU/g) as opposed to 2.5, 5 or 10% (~0.5 log CFU/g). However,
both peroxyacetic acid and mixed peracid reduced E. coli O157:H7 populations ~1 log, with no
significant difference (P > 0.05) seen between any of the organic loads. This emphasizes the

92

importance of monitoring water quality, especially if a chlorine-based sanitizer is being used, in
order to maximize the removal/inactivation of pathogens from the leaf surface.
Chemical sanitizers are routinely added to recirculating wash water to minimize the
spread of microbial contaminants during flume washing (78). Regarding their use, peroxyacetic
acid-based sanitizers are limited to a maximum of 80 ppm peroxyacetic acid (43, 62), while free
chlorine concentrations typically range from 10 to a maximum of 200 ppm (55, 91, 122).
However, soil, debris, and vegetable latexes released during shredding of leafy greens will
accumulate in the flume water over time (86), decreasing the efficacy of many sanitizers, most
notably chlorine (5, 77, 100, 139). The wash water used in this study contained organic loads of
0, 2.5, 5 and 10% blended iceberg lettuce (w/v) to simulate wash water quality at various stages
of processing. Hence, E. coli O157:H7 reductions in wash water would likely decrease with
increasing organic loads especially when using chlorine. E. coli O157:H7 reductions were
consistently higher (P ≤ 0.05) in water without an organic load for chlorine, chlorine + CA and
chlorine + T-128. In contrast, organic load rarely impacted the efficacy of peroxyacetic acid and
mixed peracid, with similar results seen by Zhang and others (139).
This study was designed to assess the efficacy of sanitizers against E. coli O157:H7 in
wash water after lettuce has been washed, with the assumption that pathogen persistence in the
wash water will result in contamination of subsequently washed product. Produce sanitizers are
primarily used to minimize cross-contamination during flume washing, with their effectiveness
dependent on the type of sanitizer, concentration, temperature, and organic load in the wash
water. The pilot-scale processing line used in the studies described in Chapters 2, 4 was not
equipped with a chiller. The work described here was intended to be a model for the pilot-scale
processing line, which is why most of this work was conducted at a water temperature of ~14°C.
93

However, subsequent analysis showed significant differences (P < 0.05) in E. coli O157:H7
reduction using similar water conditions and exposure times to those in Chapters 4 and 5. While
the results were not always similar to the pilot-scale processing line, this inexpensive bench-top
model is still a good indicator of the results that can be expected in larger pilot plant-scale
processing studies. Additionally, the bench-top model allows easy and rapid testing of far more
diverse water conditions, such as the work completed at a lower temperature (~5°C), in order to
simulate targeted commercial processing conditions. Zhang and Farber (140) reported a decrease
in sanitizer efficacy at a water temperature of 4°C as compared to of 22°C. In this study, no
significant differences (P > 0.05) in E. coli O157:H7 reduction were seen between 5 and 14°C
when chlorine + CA was used, indicating that temperature may not affect sanitizer efficacy to the
extent previously reported (62).
Chlorine, chlorine + CA and chlorine + T-128 were all significantly impacted (P ≤ 0.05)
by organic load in the wash water. In the absence of an organic load, the E. coli O157:H7
inactivation trend lines illustrate a > 5 log reduction in wash water for all three chlorine
treatments. However, organic loads of 2.5, 5 and 10% resulted in intercepts that were close to 0,
with organic loads of 5 and 10% having predicted slopes at or near 0 for the log linear trend
lines. The dependent physicochemical parameters of free chlorine, MFV, and ORP were all
significantly higher (P ≤ 0.05) at a 0% organic load compared to organic loads of 2.5, 5, or 10%,
while turbidity, COD and total solids significantly increased (P ≤ 0.05) with organic load. The
dependent physicochemical parameters for each chlorine treatment were correlated to the E. coli
2

O157:H7 inactivation rates with the RMSE, R and P values for each linear trend line shown in
Table 3.4. Considering that E. coli O157:H7 populations decreased > 5 logs for all chlorine
treatments without an organic load, commercial lettuce processors should maintain their wash
94

water physicochemical parameters within the standard deviations for an organic load of 0%
when using chlorine-based sanitizers to minimize the spread of E. coli O157:H7 and other
contaminants from the wash water to the product. Most of the dependent physicochemical
parameters of the wash water are not critical to the prediction of peroxyacetic acid or mixed
peracid efficacy, because water quality rarely had a significant impact on their respective
performance; however, pH and ORP could be used to predict efficacy because both are
dependent on sanitizer concentration more than organic load. Regardless of the physicochemical
wash water parameters measured, the slopes from the inactivation rate equations were similar for
each respective sanitizer, with minor differences predicted for the intercepts. The efficacy of
both peroxyacetic acid-based sanitizers can best be predicted based on the controlled parameters
of sanitizing agent and sanitizer concentration. The numerous physicochemical parameters
monitored in this study offer alternatives to ORP which is the most used to assess sanitizer
efficacy (112, 125). Inactivation trend lines with P values > 0.05 are likely not ideal models for
the data.
Commercial produce sanitizers are primarily designed to minimize cross-contamination
from the wash water to the product during processing. This study was done to assess the efficacy
of commercial produce sanitizers against E. coli O157:H7 in wash water using a new and novel
bench-top system that is both customizable and a cost-effective alternative to pilot-scale
processing. While E. coli O157:H7 reductions on the lettuce surface ranged from 0.4 – 1.2 log
CFU/g, this lettuce was only used as a means of inoculating the water in order to assess pathogen
persistence at different organic loads with different sanitizing agents. Those physicochemical
parameters dependent on organic load for chlorine, chlorine + CA, or chlorine + T-128 are
excellent predictors of E. coli O157:H7 persistence in wash water. Independent parameters, such
95

as sanitizer concentration or sanitizing agent, are the best predictors of peroxyacetic acid or
mixed peracid efficacy, considering that sanitizer efficacy against E. coli O157:H7 was rarely
dependent on organic load. While unable to accurately model the pilot-scale processing work
presented in Chapter 4, this bench-top model is a good indicator of the E. coli O157:H7
reductions that can be expected at different organic loads with different sanitizing agents.
Additionally, the bench-top system described here is a simple and cost-effective model that can
be customized to assess various sanitizers and organic loads at different temperatures.

96

CHAPTER 4:

Impact of Organic Load on Escherichia coli O157:H7 Survival during Pilot-Scale
Processing of Iceberg Lettuce with Commercial Produce Sanitizers

97

4.1 ABSTRACT
Chemical sanitizers are routinely used during commercial flume washing of fresh-cut
leafy greens to minimize cross-contamination from the water. This study assessed the efficacy of
five commercial sanitizer treatments against E. coli O157:H7 on iceberg lettuce, in wash water,
and on surfaces of a pilot-scale processing line using flume water containing various organic
6

loads. Iceberg lettuce (5.4 kg) was inoculated to contain 10 CFU/g of a 4-strain cocktail of nontoxigenic, GFP-labeled, ampicillin-resistant E. coli O157:H7 and held for ~24 h at 4oC to
simulate commercial storage conditions before processing. Lettuce was shredded using a Urschel
TransSlicer, step-conveyed to a flume tank, washed for 90 s using water alone or five different
sanitizing treatments (50 ppm peroxyacetic acid, 50 ppm mixed peracid, or 50 ppm available
chlorine either alone or acidified to pH 6.5 with citric acid or T-128) in water containing organic
loads of 0, 2.5, 5 or 10% (w/v) blended iceberg lettuce, and then dried using a shaker table and
centrifugal dryer. Thereafter, three 5.4-kg batches of uninoculated iceberg lettuce were
2

identically processed. Various product (25 g), water (50 ml) and equipment samples (100 cm )
were homogenized in neutralizing buffer, serially diluted and plated on tryptic soy agar
containing 0.6% (w/v) yeast extract and 100 ppm ampicillin with or without prior 0.45 μm
membrane filtration for quantification of E. coli O157:H7. Various physicochemical parameters
of the wash water were correlated to E. coli O157:H7 inactivation trend lines at each organic
load. Organic load negatively impacted the efficacy of all three chlorine treatments (P ≤ 0.05),
with typical E. coli O157:H7 reductions of > 5 log CFU/ml by the end of processing with no
organic load in the wash water and 0.9 – 3.7 log CFU/ml with a 10% organic load. Conversely,
organic load rarely impacted peroxyacetic or mixed peracid efficacy against E. coli O157:H7,
98

with typical reductions of > 5 log CFU/ml in flume water throughout processing for all organic
loads. Organic load rarely impacted sanitizer efficacy against E. coli O157:H7 on lettuce.
Reduced chlorine efficacy generally correlated to (P ≤ 0.05) to increased total solids, chemical
oxygen demand, and turbidity, indicating that these tests may be effective alternatives to the
industry standard of oxygen/reduction potential. However, E. coli O157:H7 persisted on all
previously uninoculated lettuce following the inoculated batch, emphasizing the need for
improved intervention strategies that can better ensure end-product safety.

99

4.2 INTRODUCTION
Numerous small-scale laboratory studies have assessed sanitizer efficacy against
pathogens on leafy greens (4, 66, 72, 78, 82, 88, 115, 139, 140). However, much work remains in
predicting sanitizer efficacy in commercial flume systems that contain various organic loads.
Previous sanitizer work completed by our group was performed without an organic load in order
to determine the immediate effects of sanitizers in flume water during processing of E. coli
O157:H7-inoculated iceberg lettuce (Chapter 2), or with an organic load in a 4 L bench-top
model in an attempt to predict sanitizer efficacy during pilot-plant processing of leafy greens
(Chapter 3). Since chemical sanitizers remain the sole intervention strategy to prevent crosscontamination during commercial production of fresh-cut leafy greens, it is imperative that
sanitizing agents be evaluated under conditions that more closely resemble commercial
operations. Consequently, the two aims of this study were to 1) determine the efficacy of five
commercial sanitizer treatments against E. coli O157:H7 during pilot-scale processing of iceberg
lettuce using water containing various organic loads and 2) assess the relationship between
various physicochemical parameters and organic load of the wash water on E. coli O157:H7
inactivation.

100

4.3 MATERIALS AND METHODS
4.3.1 Experimental design. The efficacy of five different sanitizing treatments against E.
coli O157:H7 was assessed in triplicate by processing 5.4 kg of iceberg lettuce inoculated at 10

6

CFU/g followed by three consecutive 5.4 kg batches of uninoculated iceberg lettuce in flume
water containing various organic loads. All lettuce was processed by shredding, conveying,
fluming, shaker table dewatering and/or centrifugal drying, during and/or after which various
product, water and equipment surface samples were collected and quantitatively examined for E.
coli O157:H7. Pathogen persistence in wash water was then correlated to six physicochemical
parameters of thewater: pH, oxidation/reduction potential, chemical oxygen demand, total solids,
maximum filterable volume and turbidity.
4.3.2 Leafy greens. Identical to 2.3.2.
4.3.3 Bacterial strains. Identical to 2.3.3.
4.3.4 Lettuce inoculation. A 0.2 ml aliquot of each non-toxigenic E. coli O157:H7
strain was transferred to 200 ml of TSBYE with amp and incubated for 18 – 20 h at 37°C. Based
on similar growth rates as determined previously (22), the four strains were combined in equal
volumes to obtain an 800-ml cocktail which was added to 80 L of municipal tap water (~15oC, <
0.05 ppm free chlorine) in a 121 L plastic container (Rubbermaid, Wooster, OH) to achieve a
7

level of ~10 CFU/ml. Hand-cored heads of iceberg lettuce (~12 heads) were immersed in the E.
coli O157:H7 suspension for 15 min and then drained/air-dried for 1 h at 22°C before being spun
in a dewatering centrifuge (described below) to remove residual inoculum from the interior of
the heads and stored at 4°C for 24 h before processing. Duplicate 25-g samples were then
aseptically collected to determine the initial inoculation level at the time of processing.
101

4.3.5 Lettuce processing line. A small-scale commercial leafy green processing line
was assembled that consisted of a lettuce shredder, step conveyer, flume tank, shaker table and
dewatering centrifuge. The commercial lettuce shredder (Model TRS 2500 Urschel TranSlicer,
Valparaiso, IN) was operated at a feed belt/slicing wheel speed of 198 m/min and 905 RPM,
respectively, to obtain a shred size of approximately 5 x 5 cm. The polyurethane step conveyer
belt (ThermoDrive, Mol Industries, Grand Rapids, MI) on the conveyor (Dorner model 736018
mc series, Dorner Manufacturing, Hartland, WI) was operated at 0.11 m/s. A stainless steel
water recirculation tank (~1000 L capacity) was filled with 890 L of tap water. This water tank
was connected to a 3.6-m long stainless steel flume tank (Heinzen Manufacturing, Inc., Gilroy,
CA) - equipped with two overhead spray jets (1 m from the start), by a 4.14 m-long, 10 cmdiameter hard plastic discharge hose and a centrifugal pump (Model XB754FHA, Sterling
Electric, Inc., Irvine, CA) that circulated the water at a rate of ~10 L/s. A custom-made stainless
steel gate with 1.25 cm-diameter holes spaced 0.65 cm apart (Heinzen Manufacturing, Inc.) was
secured to the end of the flume tank to prevent product flow while allowing continuous
circulation of the wash water. The stainless steel shaker table for partial dewatering was operated
by a 1 HP Baldor washdown duty motor (Baldor Electric Co., Ft. Smith, AR) at 1760 RPM.
Water removed from the leafy greens during mechanical shaking passed through a mesh screen
and was fed into the water holding tank by a water recirculation spout underneath the shaker
table. A 50-lb capacity centrifugal Spin Dryer (Model SD50-LT, Heinzen Manufacturing, Inc.)
with three internally timed spin cycles totaling 80 s was used for centrifugal drying. The pilotscale processing line was not equipped with a chiller to maintain the processing water at a target
commercial temperature of 4°C (62). However, the average flume water temperature of 13.5°C
did not likely have a significant effect on chlorine efficacy based on results from the temperature
102

study discussed in Chapter 3 and a previous study showing that peroxyacetic acid maintained its
efficacy over a range of temperatures (64).
4.3.6 Wash water. Iceberg lettuce (0, 22.5, 45, or 90 kg) was blended in 11 – 18 kg
batches in a 55 L cylindrical polypropylene tank (Nalgene, Rochester, NY) containing 14 L of
tap water for 10 min using a Rotostat blender (Model XP-02, Admix, Manchester, NH). All
batches were combined in the recirculation tank with additional water at 12 to 15oC to achieve
organic load levels of 0, 2.5, 5, and 10% (w/v), respectively, in a total volume of 890 L. The
following five commercial produce sanitizer treatments were assessed: 50 ppm peroxyacetic acid
(Tsunami 100, Ecolab, St. Paul, MN), 50 ppm mixed peracid (Tsunami 200, Ecolab), 50 ppm
available chlorine (XY-12, Ecolab) at pH 7.85, 30 ppm available chlorine (XY-12) acidified to
pH 6.5 with citric acid (Sigma-Aldrich, St. Louis, MO), and 50 ppm available chlorine (XY-12)
acidified to pH 6.5 with T-128 (SmartWash Solutions) as measured with a pH probe (pHTestr
30, Oakton, Vernon Hills, IL). Peroxyacetic acid test kit 311 (Ecolab) was used to confirm the
peroxyacetic acid and mixed peracid sanitizer concentrations while chlorine test kit 321 (Ecolab)
was used to measure available chlorine. Sanitizer-free MSU tap water (< 0.05 ppm free chlorine)
served as the control. The sanitizer concentration and pH were maintained throughout
processing, with necessary adjustments made during the 10 min pauses between lettuce
processing.
4.3.7 Lettuce processing. To prime the system, 5.4 kg of uninoculated iceberg lettuce
was hand-fed into the shredder at a rate of about 0.5 kg per second, with the shredded product
then step-conveyed to the flume tank and washed in 890 L of recirculating sanitizing wash water
for 90 s. The product was then released from the flume tank and partially dewatered on the
shaker table, collected in a single centrifugation basket and centrifugally dried. Following a 10
103

min pause in processing to adjust the sanitizer concentration, the 5.4 kg batch of inoculated
lettuce was identically processed followed by three uninoculated 5.4 kg batches.
4.3.8 Sample collection. For each batch of lettuce processed after the original
uninoculated priming batch, a 50 ml water sample was collected in a 50 ml centrifuge tube
containing 38x concentrated Difco Neutralizing Buffer (Becton Dickinson) at the fixed flume
screen after 90 s of flume washing. Additional 50-ml water samples were collected every 2 min
during the 10 min pause in processing between each batch of lettuce. During centrifugal drying,
a 50 ml water sample was collected from the centrifuge drain after 15 s of the 90 s cycle. After
centrifugation, a 25 g lettuce sample was collected from the centrifugation basket. After
processing all five batches of lettuce, nine product contact areas on the equipment (3 flume tank
(Figure 2.1), 3 shaker table (Figure 2.2) and 3 dewatering centrifuge (Figure 2.3)) measuring 100
2

cm as previously identified using Glo Germ™ were sampled as described by Vorst and others
(130) using 1-ply composite tissues moistened with 1 ml of sterile Difco Neutralizing Buffer
(Becton Dickinson).
4.3.9 Physiochemical parameters of flume water. Six physicochemical parameters of
the wash water were monitored before processing each of the five batches of iceberg lettuce. The
pH and oxidation/reduction potential (ORP) (mV) were monitored with probes pHTestr 30 and
ORPTestr 10, respectively (Oakton, Vernon Hills, IL). Chemical oxygen demand (COD) (mg O2
per L of solution) was quantified using Hach Digestion Reactor Method 8000 (Hach, Loveland,
CO) (59). Total solids was measured by drying 10 ml of wash water in a pre-heated and preweighed crucible in a drying oven (Model 625-A, Precision Scientific Inc, Chicago, IL) set at
103°C ± 2°C for 2 h to determine the mass of solids in suspension (14). Turbidity was assessed
104

by pouring a 50 ml water sample through a 24-cm diameter Whatman Filter (Grade 113,
Piscataway, NJ) to remove suspended solids, followed by measuring absorbance of a 1 ml
aliquot of the filtrate at 663 nm in a spectrophotometer (Model SB-100XR, Spectronics
Corporation, Westbury, NY) (133). Maximum filterable volume (MFV) was quantified as the
volume of a 50 ml water sample pulled through a 0.45 µm membrane in 1 min at-80 kPa using a
vacuum pump (Model WP6211560, Millipore, Billerica, MA).
4.3.10 Microbiological analyses. All lettuce samples (25 g) were homogenized in a
stomacher (Stomacher 400 Circulator, Seward, Worthington, UK) for 1 min at 260 rpm and then
either appropriately diluted in sterile 1% (w/v) phosphate buffer (8.5 g/L NaCl, 1.44 g/L
Na2HPO4, and 0.24 g/L KH2PO4, J.T. Baker, Mallinckrodt Baker Inc., Phillipsburg, NJ) and
plated on TSAYE with amp (calculated minimum detection limit of 40 CFU/g) or processed
using 0.45 μm membrane filters (Millipore, Millipore Corporation, Billerica, MA) (calculated
minimum detection limit of 0.04 CFU/g) which were placed on 60-mm dia. petri plates
containing TSAYE with amp to quantify E. coli O157:H7. The 1-ply composite tissue samples
were added to 15 ml of sterile Difco Neutralizing Buffer in a Whirl-Pak™ bag, homogenized
for 1 min at 260 rpm and then plated identically to the lettuce samples, giving a lower detection
2

limit of 1 CFU/100 cm . The 50 ml water samples were either appropriately diluted in sterile 1%
phosphate buffer and plated on TSAYE with amp or processed by membrane filtration, which
gave a calculated minimum detection limit of 0.02, .033, 0.1 and 0.1 CFU/ml for organic loads of
0, 2.5, 5 and 10% (w/v), respectively. Following 40 – 48 h of incubation at 37°C, all green
fluorescing colonies as seen under ultraviolet light (365 nm, Blak-Ray, Ultra-violet Product Inc.
San Gabriel, CA) were counted as E. coli O157:H7.
105

4.3.11 Sanitizer neutralization confirmation. Identical to 3.3.10.
4.3.12 Physicochemical parameters of commercial flume water. Samples of
commercial flume water containing chlorine acidified with citric acid were collected at a
commercial iceberg lettuce processing facility on three separate occasions just before processing
commenced and then at 30 min intervals over the next 5 hours. Water was assessed for
temperature, pH, oxidation/reduction potential, available chlorine, and free chlorine as measured
by a Pocket Colorimeter (Model 46770, Hach). Additional samples collected in 50 ml centrifuge
tubes at the same 30 min intervals were packed in ice and shipped overnight to Michigan State
University and assessed for chemical oxygen demand, maximum filterable volume, and total
solids within 30 h of collection using the previously described methods.
Samples of flume water containing peroxyacetic acid were collected at a commercial
iceberg lettuce processing facility on three separate occasions just before processing commenced
and then at 30 min intervals over the next 5 hours. Water was assessed for temperature, pH,
oxidation/reduction potential, and peroxyacetic acid content. Additional samples collected in 50
ml centrifuge tubes at the same 30 min intervals were packed in ice, transported to Michigan
State University and then assessed for chemical oxygen demand, maximum filterable volume,
turbidity and total solids within 8 h of their collection using the previously described methods.
4.3.13 Statistical analysis. E. coli O157:H7 counts were converted to log CFU per g,
2

ml or 100 cm and subjected to an ANOVA using JMP 9.0 (SAS Institute Inc., Cary, NC).
Values equaling half the limit of detection were used for samples without E. coli O157:H7
counts. The three equipment surface samples from each respective piece of equipment were
averaged. Physicochemical parameter data from the wash water were averaged for each
processing replicate before averaging between the three replicates. A P value of ≤ 0.05 was
106

considered significant for all tests. The Tukey-Kramer HSD test was used to identify significant
differences in E. coli O157:H7 populations for individual lettuce, water, and equipment surface
samples as well as individual physicochemical parameter values. Microsoft Excel software was
also used to conduct a linear regression analysis and calculate the 95% confidence intervals for
the E. coli O157:H7 reductions in water over time. The following log linear equation was used:
Log(N/N0) = -k*t
Where Log(N/N0) is the microbial reduction, t is the time in minutes, and k is the inactivation
2

rate. Excel was used to determine the P and R values, and JMP 9.0 software was used to
determine the RMSE (log CFU/ml) by running the curves for each organic load and sanitizing
agent combination.

107

4.4 RESULTS
4.4.1 Lettuce. In general, none of the organic loads significantly impacted sanitizer
efficacy against E. coli O157:H7 on lettuce. Average E. coli O157:H7 reductions on inoculated
lettuce were 0.9, 1.0 and 1.2 log CFU/g for chlorine (Figure 4.1), chlorine + CA (Figure 4.2) and
chlorine + T-128 (Figure 4.3), respectively. For each of the sanitizer treatments, populations of
E. coli O157:H7 were rarely significantly lower (P ≤ 0.05) in the three uninoculated 5.4 kg
batches that followed the inoculated batch of lettuce, regardless of the organic load in the wash
water. None of the organic loads had a significant impact on chlorine alone for batches 1 and 2,
with average E. coli O157:H7 populations of 2.6 and 2.0 log CFU/g recovered, respectively
(Figure 4.1). In the third batch, populations were significantly higher (P ≤ 0.05) for lettuce
washed in water containing a 10% (2.4 log CFU/g) as compared to 0% (1.1 log CFU/g) organic
load. Similar results were seen for chlorine + CA, with none of the organic load treatments
having a significant impact (P > 0.05) on sanitizer efficacy (Figure 4.2). Average E. coli
O157:H7 populations for batches 1 and 2 were 2.6 and 1.5 log CFU/g, respectively, while the
only significant difference (P ≤ 0.05) in batch 3 was seen at organic loads of 2.5% (0.6 CFU/g)
and 5% (1.4 log CFU/g). None of the organic loads had a significant impact (P > 0.05) on
chlorine + T-128 efficacy (Figure 4.3), with average E. coli O157:H7 populations of 2.4, 1.6 and
1.1 log CFU/g for batches 1, 2 and 3, respectively.
None of the organic loads significantly impacted peroxyacetic acid or mixed peracid
efficacy against E. coli O157:H7 on lettuce. E. coli O157:H7 reductions on inoculated lettuce
ranged from 1.0 to 1.7 log CFU/g as seen with peroxyacetic acid (Figure 4.4), with an average of
1.4 log CFU/g for mixed peracid (Figure 4.5). Regardless of the organic load in the wash water,
neither sanitizer treatment significantly affected E. coli O157:H7 populations in the three
108

uninoculated 5.4 kg batches that were processed after the inoculated lettuce (P > 0.05). Using
peroxyacetic acid, uninoculated batches 1, 2 and 3 yielded average E. coli O157:H7 counts of
1.8, 0.9 and 0.2 log CFU/g, respectively. Similar results were seen using mixed peracid, with
average E. coli O157:H7 populations for batches 1, 2 and 3 of 1.7, 1.2 and 0.2 log CFU/g,
respectively.

109

E. coli O157:H7 (log CFU/g)

7
6

0%
A A

2.5%

5%

10%

A A
A A A A

5
4

A

A

A

A

3
A

A

2

A

A

A

AB
B AB

1
0
InoculatedPrewash

InoculatedPostwash

1

2

3

Uninoculated Batches
Figure 4.1: Mean (± SD) E. coli O157:H7 populations on lettuce after flume washing in water
containing 50 ppm chlorine and various organic loads (n=3 per organic load). Means within
the same batch with different letters are significantly different (P ≤ 0.05).

110

E. coli O157:H7 (log CFU/g)

7
6

0%

2.5%

5%

10%

A A A A
A
A A A

5
4

A

A

A

3
A

A

2

A A

A

A

AB

1

AB

B

0
InoculatedPrewash

InoculatedPostwash

1

2

3

Uninoculated Batches

Figure 4.2: Mean (± SD) E. coli O157:H7 populations on lettuce after flume washing in water
containing 50 ppm chlorine + CA and various organic loads (n=3 per organic load). Means
within the same batch with different letters are significantly different (P ≤ 0.05).

111

7

E. coli O157:H7 (log CFU/g)

6

0%
A

AB A

2.5%

5%

10%

B

5

A

A A

A

4

A

3

A

A A

A

A A
A

2

A A

A
A

1
0
-1

InoculatedPrewash

InoculatedPostwash

1

2

3

Uninoculated Batches

Figure 4.3: Mean (± SD) E. coli O157:H7 populations on lettuce after flume washing in water
containing 50 ppm chlorine + T-128 and various organic loads (n=3 per organic load). Means
within the same batch with different letters are significantly different (P ≤ 0.05).

112

7
A
E. coli O157:H7 (log CFU/g)

6

0%

ABBC
C
A A

5

A

2.5%

5%

10%

A

4

A
A

3

A
2

A

A

A

A
A
A

1

A A A

0
-1
-2

InoculatedPrewash

InoculatedPostwash

1

2

3

Uninoculated Batches

Figure 4.4: Mean (± SD) E. coli O157:H7 populations on lettuce after processing in flume water
containing various organic loads and 50 ppm peroxyacetic acid (n=3 per organic load). Means
within the same batch with different letters are significantly different (P ≤ 0.05).

113

7

E. coli O157:H7 (log CFU/g)

6

0%

2.5%

5%

10%

A A A A

5

A

A A

A

4
A A

3
A

2

A

A

A

A
A

A

A

A

1

A

0

-1
-2

InoculatedPrewash

InoculatedPostwash

1

2

3

Uninoculated Batches

-3

Figure 4.5: Mean (± SD) E. coli O157:H7 populations on lettuce after processing in flume water
containing various organic loads and 50 ppm mixed peracid (n=3 per organic load). Means
within the same batch with different letters are significantly different (P ≤ 0.05).

114

4.4.2 Flume water. Average E. coli O157:H7 populations in sanitizer-free wash water
were statistically similar (P > 0.05) for all organic loads, with an average of 3.9 log CFU/ml at
all time points (Figure 4.6). Using organic loads of 5 and 10%, similar population reductions (P
> 0.05) were seen for all three chlorine treatments. Average E. coli O157:H7 reductions at the
start of processing (0 min) with organic loads of 5 and 10% were 0.2, 0.1 and 0.2 log CFU/ml for
chlorine (Figure 4.7), chlorine + CA (Figure 4.8) and chlorine + T-128 (Figure 4.9), respectively.
By the end of processing (44 min), average E. coli O157:H7 reductions were 0.9, 4.0 and 3.9 log
CFU/ml for chlorine, chlorine + CA and chlorine + T-128, respectively, at a 10% organic load,
with no significant difference (P > 0.05) seen at a 5% organic load (Table 4.1). Using chlorine
alone, E. coli O157:H7 reductions were significantly higher (P ≤ 0.05) after 21 min of processing
(2.3 log CFU/ml) with an organic load of 2.5 than at 5 and 10% with the average final population
being 3.9 log lower than the water control. Using chlorine + CA with a 2.5% organic load, the E.
coli O157:H7 population initially decreased 0.2 log CFU/ml at the start of processing, with a
final reduction of 5.2 log CFU/ml after 44 min, which was not significantly different (P > 0.05)
from the E. coli O157:H7 reductions seen for the 0, 5 and 10% organic loads. Using chlorine +
T-128 with a 2.5% organic load, E. coli O157:H7 populations decreased 0.0 and 4.8 log CFU/ml
after 0 min and 44 min of processing, respectively, with these reductions statistically similar (P
> 0.05) to those seen at organic loads of 5 and 10%. As expected, E. coli O157:H7 populations
were typically at or below the lower limit of detection for all chlorine treatments containing a 0%
organic load, with reductions of 5.2, 5.8 and 5.6 log CFU/ml seen for chlorine, chlorine + CA
and chlorine + T-128, respectively, at the start of processing.
At the start of sample collection, as soon as the inoculated lettuce exited the flume tank (0
min), no significant difference (P > 0.05) in E. coli O157:H7 reductions was seen among the
115

different organic loads, with average reductions ranging from 2.9 – 3.6 log CFU/ml using
peroxyacetic acid (Table 4.2). In addition, statistically similar (P > 0.05) E. coli O157:H7
reductions were observed at organic loads of 0 and 2.5% for peroxyacetic acid , with average
reductions of 5.9 and 5.8 log CFU/ml after 21 min and 5.9 and 5.8 log CFU/ml after 44 min of
processing, respectively. E. coli O157:H7 reductions using a 10% organic load with peroxyacetic
acid after 21 (5.3 log CFU/ml) and 44 min (5.2 log CFU/ml) of processing were statistically
similar (P > 0.05) to those seen at a 5% organic load. At the start of sample collection, organic
load did not significantly impact (P > 0.05) mixed peracid efficacy, resulting in average E. coli
O157:H7 reductions ranging from 3.9 – 4.9 log CFU/ml (Table 4.2). After 21 min of processing
with mixed peracid, statistically similar E. coli O157:H7 reductions (P > 0.05) were observed at
organic loads of 0 (5.9 log CFU/ml) and 2.5% (5.6 log CFU/ml) with no significant differences
(P > 0.05) observed between organic loads of 5 (5.3 log CFU/ml) 10% (5.0 log CFU/ml). After
44 min of processing with mixed peracid, E. coli O157:H7 reductions were significantly
different (P ≤ 0.05) for each organic load, with reductions of 5.9, 5.6, 5.3 and 4.9 at organic
loads of 0, 2.5, 5 and 10%, respectively.
E. coli O157:H7 populations were typically below the limit of detection in flume water
containing a 0% organic load for all three chlorine treatments, resulting in y- intercepts of -5.4, 5.7 and -5.8 log CFU/ml and root-mean-square errors (RMSEs) of 0.2, 0.2 and 0.1 log CFU/ml
for chlorine, chlorine + CA and chlorine + T-128, respectively (Table 4.3, Figures 4.7 – 4.9).
Trend lines for chlorine alone with organic loads of 2.5, 5 and 10% had similar intercepts while
the rate of inactivation differed, with 2.5% being 2.5 and 5 times greater than 5 and 10%,
respectively. The y-intercept values increased with increasing organic load from -1.8, -0.7 and 0.5 log CFU/ml for 2.5, 5 and 10% organic loads, respectively, for chlorine + CA, while the
116

inactivation rates were similar for each of the three organic loads. Similar inactivation rates and
intercepts were seen for chlorine + T-128, -2.0, -0.3 and 0.0 log CFU/ml seen for organic loads
of 2.5, 5 and 10%, respectively. RMSE values decreased as organic load increased from 2.5 to
2

10% for each of the three chlorine treatments, with the inverse seen for the R values.
Regional trend lines for chlorine, chlorine + CA and chlorine + T-128 developed for 0 –
2

10, 11.5 – 21, 23 – 32.5, and 34.5 – 44 min of processing showed higher R values overall than
the trend lines developed from 0 – 44 min of processing (Tables 4.4 – 4.7). However, the P
values were typically higher in the regional trend lines than the 0 – 44 min trend lines.
E. coli O157:H7 populations were typically near the lower limit of detection in flume
water for all organic loads, resulting in intercepts ranging from -4.6 to -5.4 log CFU/ml and -4.8
to -5.5 log CFU/ml for peroxyacetic acid mixed peracid, respectively (Table 4.8, Figures 4.10
2

and 4.11). Not surprisingly, the R values for each linear regression were < 0.2 due to the nearhorizontal slopes of the lines, indicating that exposure time is not a good predictor of E. coli
O157:H7 inactivation in flume water containing peroxyacetic acid or mixed peracid.

117

2.5%
Linear (2.5%)

E. coli O157:H7 (log CFU/ml)

0%
Linear (0%)

5%
Linear (5%)

10%
Linear (10%)

y = -0.0009x + 3.9771
y = -0.0004x + 3.9343
R² = 0.1228
yy= -0.0006x + 3.8529
R² = 0.014
= 0.0008x + 3.9213
R² ==0.0239
R² 0.171

Time (min)
Figure 4.6: Mean (± SD) E. coli O157:H7 populations in recirculating sanitizer-free flume water
containing various organic loads (n=3 per organic load).

118

0%
Linear (0%)

2.5%
Linear (2.5%)

5%
Linear (5%)

10%
Linear (10%)

y = -0.0215x - 0.0112
R² = 0.9567
y = -0.0361x - 0.1063
Log (N/N0)

R² = 0.9574

y = -0.0967x + 0.1339
R² = 0.9379
y = -0.0035x - 5.4318
R² = 0.0561

Time (min)
Figure 4.7: E. coli O157:H7 inactivation in recirculating flume water containing various organic
loads and chlorine compared to the sanitizer-free control (n=3 per organic load).

119

2.5%
Linear (2.5%)

Log (N/N0)

0%
Linear (0%)

5%
Linear (5%)

10%
Linear (10%)

y = -0.0821x - 0.5308
R² = 0.845
y = -0.0985x - 0.7344
R² = 0.7678
y = -0.0049x - 5.6523
-0.0942x 1.7955
R² = 0.1336
0.6885

Time (min)
Figure 4.8: E. coli O157:H7 inactivation in recirculating flume water containing various organic
loads and chlorine + CA compared to the sanitizer-free control (n=3 per organic load).

120

2.5%
Linear (2.5%)

Log (N/N0)

0%
Linear (0%)

5%
Linear (5%)

10%
Linear (10%)

yy = -0.087x - 0.342
= -0.093x - 0.0058
R² = 0.9747
R² = 0.8823

y = -0.0792x - 1.9792
y = -0.0024x - 5.7505
R² = 0.5901
R² = 0.0577
Time (min)

Figure 4.9: E. coli O157:H7 inactivation in recirculating flume water containing various organic
loads and chlorine + T-128 compared to the sanitizer-free control (n=3 per organic load).

121

Table 4.1: E. coli O157:H7 reductions in flume water containing chlorine, chlorine + CA and
chlorine + T-128

a

Mean ± SD E. coli O157:H7 reduction in flume water
(log CFU/ml)
Sanitizing
Treatment

% Organic
Load (w/v)
0

0 min

21 min

44 min

5.2 ± 1.1 A

5.5 ± 0.5 A

5.4 ± 0.5 A

2.5

0.2 ± 0.3 B

2.3 ± 0.6 B

3.9 ± 0.9 A

5

0.3 ± 0.2 B

0.8 ± 0.4 C

1.6 ± 0.7 B

10

0.1 ± 0.2 B

0.5 ± 0.2 C

0.9 ± 0.3 B

0

5.8 ± 0.1 A

5.6 ± 0.6 A

5.9 ± 0.0 A

2.5

0.2 ± 0.0 B

4.9 ± 0.3 AB

5.2 ± 0.1 AB

5

0.2 ± 0.4 B

4.3 ± 0.4 AB

4.7 ± 0.7 AB

10

0.0 ± 0.1 B

3.3 ± 1.4 B

4.0 ± 1.2 B

0

5.6 ± 0.5 A

5.7 ± 0.5 A

5.9 ± 0.0 A

2.5

0.0 ± 0.2 B

4.6 ± 0.7 A

4.8 ± 0.6 AB

5

0.2 ± 0.1 B

2.9 ± 0.4 B

4.2 ± 1.1 AB

10

0.2 ± 0.5 B

2.3 ± 0.2 B

3.9 ± 0.4 B

Chlorine

Chlorine + CA

Chlorine + T-128

a

Means with different capital letters within a given sanitizing treatment designate E. coli

O157:H7 reductions that differ significantly in terms of sanitizing treatment and time (P ≤ 0.05).

122

Table 4.2: E. coli O157:H7 reduction in flume water containing peroxyacetic acid and mixed
peracid

a

Mean ± SD E. coli O157:H7 reduction in flume water
(log CFU/ml)
0 min
3.6 ± 0.2 A

21 min
5.9 ± 0.2 A

44 min
5.9 ± 0.0 A

2.5

3.6 ± 0.4 A

5.8 ± 0.1 A

5.8 ± 0.1 A

5

2.9 ± 0.4 A

5.3 ± 0.0 B

5.3 ± 0.0 B

10

3.5 ± 0.1 A

5.3 ± 0.0 B

5.2 ± 0.1 B

0

4.1 ± 0.8 A

5.9 ± 0.2 A

5.9 ± 0.0 A

2.5

5.0 ± 0.3 A

5.6 ± 0.2 A

5.6 ± 0.2 B

5

3.9 ± 1.5 A

5.3 ± 0.0 B

5.3 ± 0.0 C

10

Sanitizing Treatment

% Organic
Load (w/v)
0

4.0 ± 1.3 A

5.0 ± 0.0 B

4.9 ± 0.1 D

Peroxyacetic Acid

Mixed Peracid

a

Means with different capital letters within a given sanitizing treatment designate E. coli

O157:H7 reductions that differ significantly in terms of sanitizing treatment and time (P ≤ 0.05).

123

Table 4.3: E. coli O157:H7 inactivation trend lines for flume water containing various organic
loads and chlorine, chlorine + CA and chlorine + T-128

a

% Org.
Load
(w/v)

Available
Chlorine
(ppm)

0

50 ± 4

2.5

50 ± 4

5

50 ± 6

Log(N/N0)= -0.1

10

50 ± 3

0

RMSE (log
CFU/ml)

P Value

0.06

0.2

0.26

Log(N/N0) = -0.1t + 0.1 0.94

0.3

< 0.0001

0.96

0.1

< 0.0001

Log(N/N0) = 0.0

0.96

0.1

< 0.0001

50 ± 4

Log(N/N0) = -5.7

0.14

0.2

0.07

2.5

50 ± 3

Log(N/N0) = -0.1t - 1.8

0.70

0.9

< 0.0001

5

50 ± 4

Log(N/N0) = -0.1t - 0.7

0.78

0.7

< 0.0001

10

50 ± 2

Log(N/N0) = -0.1t - 0.5

0.85

0.5

< 0.0001

0

50 ± 6

0.06

0.1

0.25

2.5

Supplement

50 ± 2

Log(N/N0) = -0.1t – 2.0 0.60

0.6

< 0.0001

5

50 ± 4

Log(N/N0) = -0.1t - 0.3

0.89

0.4

< 0.0001

10

50 ± 2

Log(N/N0) = -0.1t

0.98

0.2

< 0.0001

Log Reduction
Log(N/N0) = -5.4

R
b

2

-

CA

Log(N/N0) = -5.8

b

T-128

124

Table 4.3 (cont’d)
a

Inactivation rates were determined by using Microsoft Excel to create linear regression trend

lines for E. coli O157:H7 log reduction over time (t) in minutes.
b

The slope (k) was < 0.1 and was therefore not included.

125

Table 4.4: E. coli O157:H7 inactivation trend lines for flume water containing various organic
loads and chlorine, chlorine + CA, or chlorine + T-128 (0 – 10 min)
Available
Chlorine
(ppm)

0

50 ± 4

2.5

50 ± 4

5

RMSE
(log
CFU/ml)

P Value

0.55

0.2

0.09

Log(N/N0) = -0.1t + 0.1

0.61

0.3

0.07

50 ± 6

Log(N/N0) = -0.2

0.26

0.1

0.31

10

50 ± 3

Log(N/N0) = -0.1

0.24

0.0

0.32

0

50 ± 4

Log(N/N0) = -5.6

0.01

0.2

0.86

2.5

50 ± 3

Log(N/N0) = -0.4t + 0.1

0.96

1.5

< 0.001

5

50 ± 4

Log(N/N0) = -0.2t + 0.3

0.73

0.8

0.03

10

50 ± 2

Log(N/N0) = -0.2t + 0.3

0.91

0.7

< 0.001

0

50 ± 6

Log(N/N0) = -5.8

0.00

0.2

0.93

2.5

Supplement

% Org.
Load
(w/v)

a

50 ± 2

Log(N/N0) = -0.4t + 0.1

0.96

1.7

< 0.0001

5

50 ± 4

Log(N/N0) = -0.1t

0.83

0.3

0.01

10

50 ± 2

Log(N/N0) = -0.1t - 0.2

0.84

0.2

0.01

Log Reduction
Log(N/N0) = -5.2

R

2

b

-

CA

T-128

126

Table 4.4 (cont’d)
a

Inactivation rates were determined by using Microsoft Excel to create linear regression trend

lines for E. coli O157:H7 log reduction over time (t) in minutes.
b

The slope (k) was < 0.1 and was therefore not included.

127

Table 4.5: E. coli O157:H7 inactivation trend lines for flume water containing various organic
loads and chlorine, chlorine + CA, or chlorine + T-128 (11.5 – 21 min)
Available
Chlorine
(ppm)

0

50 ± 4

2.5

50 ± 4

5

RMSE
(log
CFU/ml)

P Value

0.16

0.2

0.43

Log(N/N0) = -0.2t + 1.4

0.92

0.6

< 0.01

50 ± 6

Log(N/N0) = -0.3

0.60

0.1

0.07

10

50 ± 3

Log(N/N0) = -0.1

0.32

0.1

0.24

0

50 ± 4

Log(N/N0) = -6.2

0.38

0.1

0.19

2.5

50 ± 3

Log(N/N0) = -0.2t - 0.3

0.78

1.0

0.02

5

50 ± 4

Log(N/N0) = -0.2t - 0.1

0.99

0.8

< 0.0001

10

50 ± 2

Log(N/N0) = -0.2t - 0.4

0.99

0.6

< 0.0001

0

50 ± 6

Log(N/N0) = -6.0

0.10

0.1

0.54

2.5

Supplement

% Org.
Load
(w/v)

a

50 ± 2

Log(N/N0) = -0.2t - 0.1

0.90

0.9

< 0.01

5

50 ± 4

Log(N/N0) = -0.2t - 0.1

0.97

0.5

< 0.01

10

50 ± 2

Log(N/N0) = -0.1t + 0.6

0.92

0.5

< 0.01

Log Reduction
Log(N/N0) = -5.3

R

2

b

-

CA

T-128

128

Table 4.5 (cont’d)
a

Inactivation rates were determined by using Microsoft Excel to create linear regression trend

lines for E. coli O157:H7 log reduction over time (t) in minutes.
b

The slope (k) was < 0.1 and was therefore not included.

129

Table 4.6: E. coli O157:H7 inactivation trend lines for flume water containing various organic
loads and chlorine, chlorine + CA, or chlorine + T-128 (23 – 32.5 min)
Available
Chlorine
(ppm)

0

50 ± 4

2.5

50 ± 4

5

RMSE
(log
CFU/ml)

P Value

0.05

0.3

0.66

Log(N/N0) = -0.1t + 0.3

0.95

0.4

< 0.001

50 ± 6

Log(N/N0) = -0.4

0.94

0.1

< 0.001

10

50 ± 3

Log(N/N0) = -0.1

0.76

0.1

< 0.001

0

50 ± 4

Log(N/N0) = -4.9

0.34

0.2

0.22

2.5

50 ± 3

Log(N/N0) = -0.1t -1.0

0.64

0.6

0.06

5

50 ± 4

Log(N/N0) = -0.1t + 0.5

0.96

0.5

< 0.001

10

50 ± 2

Log(N/N0) = -0.1t

0.88

0.4

0.01

0

50 ± 6

Log(N/N0) = -6.1

0.39

0.0

0.19

2.5

Supplement

% Org.
Load
(w/v)

a

50 ± 2

Log(N/N0) = -0.2t + 0.3

0.78

0.7

0.02

5

50 ± 4

Log(N/N0) = -0.1t - 0.1

0.88

0.4

0.01

10

50 ± 2

Log(N/N0) = -0.1t + 0.4

0.99

0.4

< 0.0001

Log Reduction
Log(N/N0) = -5.9

R

2

b

-

CA

T-128

130

Table 4.6 (cont’d)
a

Inactivation rates were determined by using Microsoft Excel to create linear regression trend

lines for E. coli O157:H7 log reduction over time (t) in minutes.
b

The slope (k) was < 0.1 and was therefore not included.

131

Table 4.7: E. coli O157:H7 inactivation trend lines for flume water containing various organic
loads and chlorine, chlorine + CA, or chlorine + T-128 (34.5 – 44 min)
Available
Chlorine
(ppm)

0

50 ± 4

2.5

50 ± 4

5

RMSE
(log
CFU/ml)

P Value

0.04

0.1

0.70

Log(N/N0) = -0.3

0.91

0.2

< 0.01

50 ± 6

Log(N/N0) = -0.9

0.66

0.1

0.05

10

50 ± 3

Log(N/N0) = -0.1

0.77

0.1

0.02

0

50 ± 4

Log(N/N0) = -0.48

0.43

0.1

0.16

2.5

50 ± 3

Log(N/N0) = -0.1t - 0.7

0.74

0.5

0.03

5

50 ± 4

Log(N/N0) = -0.2t + 1.6

0.94

0.5

< 0.01

10

50 ± 2

Log(N/N0) = -0.1t + 0.8

0.99

0.4

< 0.0001

0

50 ± 6

Log(N/N0) = -4.4

0.57

0.2

0.08

2.5

Supplement

% Org.
Load
(w/v)

a

50 ± 2

Log(N/N0) = -0.1t + 0.1

0.71

0.5

0.04

5

50 ± 4

Log(N/N0) = -0.1t + 0.7

0.86

0.4

< 0.01

10

50 ± 2

Log(N/N0) = -0.1t - 0.2

0.94

0.3

< 0.01

Log Reduction
Log(N/N0) = -5.9

R

2

b

-

CA

T-128

132

Table 4.7 (cont’d)
a

Inactivation rates were determined by using Microsoft Excel to create linear regression trend

lines for E. coli O157:H7 log reduction over time (t) in minutes.
b

The slope (k) was < 0.1 and was therefore not included.

133

Table 4.8: E. coli O157:H7 inactivation trend lines in flume water containing various organic
loads and peroxyacetic acid and mixed peracid

Sanitizing
Agent

a

% Org.
Load (w/v)

POAA
(ppm)

Log Reduction

RMSE (log
CFU/ml)

P Value

0

48 ± 4

Log(N/N0) = -5.4

0.13

0.5

0.09

Peroxyacetic

2.5

50 ± 4

Log(N/N0) = -5.3

0.16

0.4

0.05

Acid

5

51 ± 4

Log(N/N0) = -4.6

0.19

0.5

0.03

10

49 ± 2

Log(N/N0) = -4.9

0.14

0.4

0.07

0

49 ± 2

Log(N/N0) = -5.5

0.16

0.4

0.05

2.5

49 ± 2

Log(N/N0) = -5.5

0.12

0.1

0.10

5

49 ± 4

Log(N/N0) = - 5.1

0.10

0.3

0.12

10

48 ± 3

Log(N/N0) = -4.8

0.08

0.2

0.18

R
b

2

Mixed Peracid

a

Inactivation rates were determined by using Microsoft Excel to create linear regression trend

lines for E. coli O157:H7 log reduction over time (t) in minutes.
b

The slope (k) was < 0.1 and was therefore not included.

134

0

0%
Linear (0%)

2.5%
Linear (2.5%)

5%
Linear (5%)

10%
Linear (10%)

-1

Log (N/N0)

-2
-3
-4
-5
-6
-7
0

5

10

15

20
25
Time (min)

30

35

40

45

Figure 4.10: E. coli O157:H7 inactivation in recirculating flume water containing various
organic loads and peroxyacetic acid compared to the sanitizer-free control (n=3 per organic
load).

135

0

0%
Linear (0%)

2.5%
Linear (2.5%)

5%
Linear (5%)

10%
Linear (10%)

-1

Log (N/N0)

-2
-3
-4
-5
-6
-7
0

5

10

15

20
25
Time (min)

30

35

40

45

Figure 4.11: E. coli O157:H7 inactivation in recirculating flume water containing various
organic loads and mixed peracid compared to the sanitizer-free control (n=3 per organic load).

136

4.4.3 Centrifugation water. Regardless of organic load, water exiting the centrifuge
drain after spin-drying yielded similar E. coli O157:H7 populations in the absence of sanitizers
(P > 0.05), with average E. coli O157:H7 populations of 4.8 log CFU/ml for the inoculated batch
and 4.0 log CFU/ml for batches 1 – 3. Chlorine with an organic load of 10% resulted in E. coli
O157:H7 populations of 4.6, 3.8, 3.3 and 3.1 log CFU/ml in the inoculated batch and batches 1, 2
and 3, respectively, none of which were significantly different (P > 0.05) than the populations
recovered at a 5% organic load (Figure 4.12). Using chlorine alone E. coli O157:H7 populations
were significantly lower (P ≤ 0.05) at an organic load of 2.5 as compared to 10% for batches 2
(2.2 log CFU/ml) and 3 (1.8 log CFU/ml). At an organic load of 0% E. coli O157:H7 populations
were significantly lower than the other organic loads, with -0.1, -0.7, 0.1, and -0.4 log CFU/ml
seen in the inoculated batch and batches 1, 2 and 3, respectively. Using chlorine + CA, an
organic load of 10% resulted in E. coli O157:H7 populations that were not significantly different
(P > 0.05) from the 2.5 or 5% organic loads at any batch, with 4.8, 3.3, 2.1 and 1.8 log CFU/ml
seen in the inoculated batch and batches 1, 2 and 3, respectively (Figure 4.13). Using chlorine +
CA at an organic load of 0%, E. coli O157:H7 populations were significantly lower (P ≤ 0.05)
compared to the other organic loads, with populations of -1.9, -1.8, -1.8 and -1.2 log CFU/ml
seen for the inoculated batch and batches 1, 2 and 3, respectively. Organic load treatments of 2.5,
5 and 10% yielded statistically similar (P > 0.05) E. coli O157:H7 populations for the inoculated
batch and batches 2 and 3 when chlorine + T-128 was used, with 4.4, 2.1 and 1.6 log CFU/ml,
respectively, recovered at a 10% organic load (Figure 4.14). Using chlorine + T-128 at an
organic load of 0%, E. coli O157:H7 populations were significantly lower (P ≤ 0.05) in all
batches, with -1.1, -1.6, -1.9 and -2.0 log CFU/ml for the inoculated batch and batches 1, 2 and 3,
respectively.
137

Organic load rarely affected peroxyacetic acid (Figure 4.15) or mixed peracid (Figure
4.16) efficacy, with statistically similar E. coli O157:H7 populations seen in centrifugation water
for each batch of iceberg lettuce processed at the various organic loads. After washing the
inoculated batch of lettuce with peroxyacetic acid in the presence of a 2.5% organic load, E. coli
O157:H7 populations in the centrifugation water ranged from 2.3 – 2.9 log CFU/ml, with these
populations significantly lower than those observed at organic loads of 5 and 10%. Regardless of
the organic load, statistically similar E. coli O157:H7 populations (P > 0.05) were shed in the
centrifugation water after peroxyacetic acid washing of lettuce, with average populations of 1.4
and 0.5 log CFU/ml, respectively. After washing in peroxyacetic acid, the third uninoculated
batch of lettuce shed E. coli O157:H7 populations into the centrifugation water that were
significantly lower (P ≤ 0.05) at an organic load of 2.5 (-0.1 CFU/ml) as compared to 5% (0.8
CFU/ml). Organic load did not significantly impact (P > 0.05) the efficacy of mixed peracid,
with average E. coli O157:H7 populations of 2.4, 1.0, 0.1, and -0.1 log CFU/ml for the
inoculated batch and uninoculated batches 1, 2 and 3, respectively.

138

6

0%

5%

A A A

4

AB A

3

AB

B
C

C

1
B

A

B

2

0

10%

A A A

5
E. coli O157:H7 (log CFU/ml)

2.5%

B

-1
-2
-3

Inoculated

1

2

3

Uninoculated Batches
Figure 4.12: Mean (± SD) E. coli O157:H7 populations in spent centrifugation water from
lettuce after washing in flume water containing chlorine and various organic loads (n=3 per
organic load). Means within the same batch with different letters are significantly different (P ≤
0.05). Limit of detection = 0.02 CFU/ml.

139

6

0%

5%

10%

A A A

5
E. coli O157:H7 (log CFU/ml)

2.5%

A A

4
A

3

A A

A

2

A A

A

1
0

B

-1
B

B

B

-2
-3

Inoculated

1

2

3

Uninoculated Batches
Figure 4.13: Mean (± SD) E. coli O157:H7 populations in spent centrifugation water from
lettuce after washing in flume water containing chlorine + CA and various organic loads (n=3
per organic load). Means within the same batch with different letters are significantly different
(P ≤ 0.05). Limit of detection = 0.02 CFU/ml.

140

6
A

E. coli O157:H7 (log CFU/ml)

5

A

0%

2.5%

5%

10%

A
A AB

4
B

3

A A

A
A A

A

2
1
0
-1

B

C
B

B

-2
-3

Inoculated

1

2

3

Uninoculated Batches
Figure 4.14: Mean (± SD) E. coli O157:H7 populations in spent centrifugation water from
lettuce after washing in flume water containing chlorine + T-128 and various organic loads
(n=3 per organic load). Means within the same batch with different letters are significantly
different (P ≤ 0.05). Limit of detection = 0.02 CFU/ml.

141

E. coli O157:H7 (log CFU/ml)

3.5
3

0%

A A

AB

2.5%

5%

10%

B
2.5
2

A

A

A

1.5

A
A

1

A

0.5

A

A

A

AB

AB

B

0
-0.5
-1

Inoculated

1

2

3

Uninoculated Batches
Figure 4.15: Mean (± SD) E. coli O157:H7 populations in spent centrifugation water from
lettuce after washing in flume water containing peroxyacetic acid and various organic loads
(n=3 per organic load). Means within the same batch with different letters are significantly
different (P ≤ 0.05). Limit of detection = 0.02 CFU/ml.

142

E. coli O157:H7 (log CFU/ml)

3.5
3
2.5

0%

A

A

2.5%

5%

10%

A A

2

A

A

1.5

A

A

1

A
A

0.5

A

A

A

A
A

0

A

-0.5
-1

Inoculated

1

2

3

Uninoculated Batches
Figure 4.16: Mean (± SD) E. coli O157:H7 populations in spent centrifugation water from
lettuce after washing in flume water containing mixed peracid and various organic loads (n=3
per organic load). Means within the same batch with different letters are significantly different
(P ≤ 0.05). Limit of detection = 0.02 CFU/ml.

143

4.4.4 Processing equipment surfaces. After iceberg lettuce was processed in chlorinefree water, statistically similar (P > 0.05) E. coli O157:H7 populations of 3.5, 3.6 and 3.3 log
2

CFU/100 cm were recovered from the flume tank, shaker table, and centrifugal dryer,
respectively, regardless of organic load. When lettuce was washed with chlorine containing a
2

10% organic load, E. coli O157:H7 populations of 2.6, 2.9 and 3.0 log CFU/100 cm were
recovered from the flume tank, shaker table, and centrifugal dryer, respectively, with these
populations statistically similar (P > 0.05) to those seen at a 5% organic load (Figure 4.17).
Chlorine efficacy significantly decreased at organic loads above 2.5% (P ≤ 0.05) for water
samples collected from the flume tank and shaker table. However, E. coli O157:H7 populations
recovered from the centrifugal dryer were not significantly different (P > 0.05) than those
2

recovered after washing with a 5 or 10% organic load, with 2.3 log CFU/100 cm . Significantly
lower (P ≤ 0.05) E. coli O157:H7 populations were seen on the flume tank using chlorine
2

without an organic load (-0.1 log CFU/100 cm ), while E. coli O157:H7 populations on the
2

2

shaker table (0.4 log CFU/100 cm ) and centrifugal dryer (1.6 log CFU/100 cm ) were not
significantly different using a 2.5% organic load. Chlorine + CA containing an organic load of
2

10% yielded E. coli O157:H7 populations of 0.6, 1.2, and 2.4 log CFU/100 cm on the flume
tank, shaker table and centrifugal dryer, respectively, all of which were statistically similar to the
populations seen at an organic load of 5% (Figure 4.18). In the absence of an organic load, E.
coli O157:H7 populations were significantly lower (P ≤ 0.05) on the flume tank (-0.1 log
2

2

CFU/100 cm ), shaker table (0.1 log CFU/100 cm ), and centrifugal dryer (0.6 log CFU/100

144

2

cm ) compared to a 10% organic load. Chlorine + T-128 containing a 10% organic load yielded
2

E. coli O157:H7 populations of 0.4, 1.1 and 2.1 log CFU/100 cm on the flume tank , shaker
table and centrifugal dryer, respectively, all of which were statistically similar (P > 0.05) to those
populations seen at organic loads of 2.5 and 5% (Figure 4.19). Significantly lower (P ≤ 0.05) E.
coli O157:H7 populations were also seen on the shaker table and centrifugal dryer using an
organic load of 0 as compared to 10%.
Using peroxyacetic acid, E. coli O157:H7 populations from the flume tank were
2

significantly lower (P ≤ 0.05) at organic loads of 0 (-0.2 log CFU/100 cm ) and 2.5% (-0.3 log
2

2

2

CFU/100 cm ) as compared to 5 (0.1 log CFU/100 cm ) and 10% (0.4 log CFU/100 cm )
(Figure 4.20). Regardless of organic load, statistically similar E. coli O157:H7 populations were
2

recovered from the shaker table (0.1 to 0.6 log CFU/100 cm ) and centrifugal dryer (-0.3 to 0.2
2

log CFU/100 cm ) using peroxyacetic acid. In addition, no significant differences (P > 0.05) in
2

E. coli O157:H7 populations were seen on the flume tank (-0.2 to -0.1 log CFU/100 cm ), shaker
2

2

table (-0.2 to -0.1 log CFU/100 cm ), and centrifugal dryer (-0.3 to -0.1 log CFU/100 cm )
between any of the organic loads when mixed peracid was used (Figure 4.21).

145

E. coli O157:H7 (log CFU/100cm2)

3.5

0%

2.5%

3

5%

10%
A

A

2.5

A
AB

A
B

2
1.5

A

A

B

B

1

B

0.5
0

C

-0.5
Flume Tank

Shaker Table

Dewatering Centrifuge

Equipment
Figure 4.17: Mean (± SD) E. coli O157:H7 populations on equipment surfaces after processing
lettuce in flume water containing chlorine and various organic loads (n=3 per organic load).
Means within the same equipment with different letters are significantly different (P ≤ 0.05).
2

Limit of detection = 1 CFU/100 cm .

146

3
E. coli O157:H7 (log CFU/100cm2)

0%

2.5%

5%

10%

A
A

2.5

A

2
A
A

1.5
A
1

B
B

0.5
0

B

AB
B

B

-0.5
Flume Tank

Shaker Table
Equipment

Dewatering Centrifuge

Figure 4.18: Mean (± SD) E. coli O157:H7 populations on equipment surfaces after processing
lettuce in flume water containing chlorine + CA and various organic loads (n=3 per organic
load). Means within the same equipment with different letters are significantly different (P ≤
2

0.05). Limit of detection = 1 CFU/100 cm .

147

3
E. coli O157:H7 (log CFU/100 cm2)

0%

2.5%

5%

10%

AB

2.5
2

A

AB
A

1.5

A
B

1
A

A

AB
A
B

0.5
0

A

-0.5

Flume Tank

Shaker Table

Dewatering Centrifuge

Equipment
Figure 4.19: Mean (± SD) E. coli O157:H7 populations on equipment surfaces after processing
lettuce in flume water containing chlorine + T-128 and various organic loads (n=3 per organic
load). Means within the same equipment with different letters are significantly different (P ≤
2

0.05). Limit of detection = 1 CFU/100 cm .

148

E. coli O157:H7 (log CFU/100 cm2)

1.4

0%

2.5%

5%

10%

A

1.2
A

1
A

0.8
0.6
0.4

A

A

AB

A

0.2

A

0
B
-0.2

A

A

B

-0.4
Flume Tank

Shaker Table
Equipment

Dewatering Centrifuge

Figure 4.20: Mean (± SD) E. coli O157:H7 populations on equipment surfaces after processing
lettuce in flume water containing peroxyacetic acid and various organic loads (n=3 per organic
load). Means within the same equipment with different letters are significantly different (P ≤
2

0.05). Limit of detection = 1 CFU/100 cm .

149

0.6

0%

2.5%

5%

10%

A
E. coli O157:H7 (log CFU/100 cm2)

0.5
0.4

A

A

0.3
0.2
0.1

A

0

-0.1

A

A

A

A

A

A

A

-0.2
A

-0.3
-0.4
Flume Tank

Shaker Table

Dewatering Centrifuge

Equipment
Figure 4.21: Mean (± SD) E. coli O157:H7 populations on equipment surfaces after processing
lettuce in flume water containing mixed peracid and various organic loads (n=3 per organic
load). Means within the same equipment with different letters are significantly different (P ≤
2

0.05). Limit of detection = 1 CFU/100 cm .

150

4.4.5 Physicochemical parameters of flume water. All wash water physicochemical
parameters were monitored throughout processing, with significant changes rarely seen for the
different sanitizer treatments. Increasing the organic load significantly increased (P ≤ 0.05) total
solids, COD, and turbidity, whereas MFV and ORP significantly decreased (P ≤ 0.05) for all
three chlorine treatments (Tables 4.9 and 4.10). Using chlorine alone, pH significantly decreased
(P ≤ 0.05) from 8.52 to 8.10 as the organic load increased from 0 to 10%. A significant increase
(P ≤ 0.05) in total solids was seen for each increase in organic load for chlorine + CA (0.0112 to
0.0482 g) and chlorine + T-128 (0.0090 to 0.0533 g) (Table 4.10), while no significant difference
(P > 0.05) in total solids was observed for organic loads of 2.5 and 5% when chlorine was used
alone. COD increased significantly (P ≤ 0.05) with organic load from 23 to 4058 mg O2/L using
chlorine alone and was also significantly higher for chlorine + CA and chlorine + T-128
containing a 10% as opposed to a 0 or 2.5% organic load. Turbidity increased (P ≤ 0.05) between
organic loads of 0, 5 and 10% for all chlorine treatments. MFV and ORP decreased as the
organic load increased, with significantly higher (P ≤ 0.05) values seen at 0 as compared to a
10% organic load for all three chlorine treatments (Tables 4.9 and 4.10). While free chlorine was
not quantified during processing, it is likely that the concentrations were similar to those seen in
the carboys discussed in Chapter 3- with ~50 ppm of free chlorine present in water containing a
0% organic load and < 1.5 ppm in water containing organic loads of 2.5 – 10%.
Increasing organic load correlated to significant increases (P ≤ 0.05) in total solids, COD,
ORP and turbidity for both peroxyacetic acid and mixed peracid. However, there was an inverse
correlation between increasing organic load with significant decreases (P ≤ 0.05) in MFV and
pH for both sanitizing treatments (Table 4.11). As the organic load increased from 0 to 10%, a
significant decrease (P ≤ 0.05) in pH was observed for peroxyacetic acid (pH 7.00 to 5.98) and
151

mixed peracid (pH 6.66 to 4.94). When peroxyacetic acid was used, total solids increased
significantly (P ≤ 0.05) at each organic load (from 0.0065 to 0.0533 g/10 ml), while the only
significant difference (P ≤ 0.05) in total solids for mixed peracid was seen between organic loads
of 0 (0.0069 g/10 ml) and 10% (0.0771 g/10ml). When peroxyacetic acid alone was used, COD
increased significantly (P < 0.05) with increasing organic load (from 251 to 4977 mg O2/L),
while the COD at 10% (5444 mg O2/L) was significantly higher (P ≤ 0.05) than the 0 (548 mg
O2/L), 2.5 (1924 mg O2/L) and 5% (2896 mg O2/L) organic loads for mixed peracid. Significant
differences (P ≤ 0.05) in turbidity were seen between organic loads of 0, 5 and 10% for both
peroxyacetic acid and mixed peracid with maximum turbidity values of 0.155 and 0.106,
respectively. ORP increased significantly (P ≤ 0.05) as organic load in the flume water increased
for both peroxyacetic acid (311 – 604 mV) and mixed peracid (348 – 497 mV), likely due to the
increased amount of sanitizer added to achieve 50 ppm peroxyacetic acid at the higher organic
loads. As organic load levels increased, significant decreases in pH were also observed (P ≤
0.05) for both peroxyacetic acid (7.00 – 5.98) and mixed peracid (6.66 – 4.94), which were likely
due to the additional amount of sanitizer needed to achieve the target concentration at the higher
organic loads. As expected, MFV was inversely related to organic load, with significantly higher
(P ≤ 0.05) volumes seen at an organic load of 0 as compared to 10% for both peroxyacetic acid
(50 and 11 ml) and mixed peracid (50 and 16 ml).

152

Table 4.9: Physicochemical parameters of flume water containing chlorine and various organic loads
Controlled
Physicochemical
Parameters
%
Available
Org.
Chlorine
Load
(ppm)
(w/v)
0
50 ± 4

a

Mean ± SD Dependent Physicochemical Parameters

Temp
(°C)

pH

Total Solids (g/10
ml)

COD (mg
O2/L)

Turbidity (abs
@ 663 nm)

MFV
(ml)

ORP (mV)

13±0 B

8.52±0.05 A

0.0070±0.0011 C

23±9 D

0.001±0.001 C

50±0 A

720±13 A

2.5

13±0 B

8.38±0.06 AB

0.0247±0.0123 B

971±72 C

0.021±0.001 C

43±2 A

366±16 B

5

50 ± 6

13±0 B

8.21±0.08 BC

0.0289±0.0020 B

2067±271 B

0.055±0.009 B

23±6 B

391±15 B

10
a

50 ± 4

50 ± 3

14±0 A

8.10±0.12 C

0.0552±0.0049 A

4058±499 A

0.124±0.017 A

11±1 C

368±19 B

Means labeled with different capital letters designate physicochemical parameters that differ significantly in terms of organic load

(P ≤ 0.05).

153

Table 4.10: Physicochemical parameters of flume water containing chlorine + CA or chlorine + T-128 with various organic loads
Controlled Physicochemical
Parameters
%
Available
Org.
Supplement
Chlorine
Load
(ppm)
(w/v)
0
50±4

a

Mean ± SD Dependent Physicochemical Parameters
Temp
(°C)

Total Solids (g/10
ml)

COD (mg
O2/L)

Turbidity (abs @
663 nm)

MFV
(ml)

ORP (mV)

14±0 AB

0.0112±0.0015 D

467±162 C

0.001±0.000 C

50±0 A

887±4 A

2.5

50±3

13±0 B

0.0210±0.0010 C

1330±45 BC

0.016±0.003 BC

45±2 A

545±4 B

5

50±4

14±0 AB

0.0335±0.0040 B

2695±802 AB

0.039±0.012 B

40±10 A

519±6 BC

10

50±2

14±0 A

0.0482±0.0026 A

4527±1446 A

0.103±0.013 A

12±0 B

502±22 C

0

50±6

13±0 B

0.0090±0.0006 D

345±9 B

0.001±0.000 D

50±0 A

887±3 A

2.5

50±2

14±0 C

0.0190±0.0007 C

1230±133 B

0.023±0.003 C

36±9 B

564±22 B

5

50±4

13±0 B

0.0326±0.0018 B

2163±654 AB

0.060±0.003 B

25±3 B

503±8 C

10

50±2

14±0 A

0.0533±0.0019 A

4030±1527 A

0.127±0.006 A

12±2 C

502±8 C

CA

T-128

a

Means with different capital letters designate physicochemical parameters differ significantly in terms of organic load (P ≤ 0.05).

154

Table 4.11: Physicochemical parameters of flume water containing organic load and 50 ppm peroxyacetic acid or 50 ppm mixed
peracid

a

Controlled
Physicochemical
Parameters

Mean ± SD Dependent Physicochemical Parameters

%
Org.
Load
(w/v)
0

Temp
(°C)

pH

Total Solids (g/10
ml)

COD (mg
O2/L)

Turbidity (abs
@ 663 nm)

MFV (ml)

ORP (mV)

11±0 B

7.00±0.01 A

0.0065±0.0002 D

251±32 D

0.001±0.000 C

50±0 A

311±3 B

Peroxyacetic

2.5

12±0 B

6.98±0.08 A

0.0154±0.0006 C

1244±72 C

0.018±0.002 C

50±0 A

538±61 A

Acid

5

11±1 B

6.70±0.06 B

0.0272±0.0018 B

2866±208 B

0.074±0.010 B

30±5 B

619±7 A

10

14±1 A

5.98±0.11 C

0.0533±0.0019 A

4977±298 A

0.155±0.015 A

11±2 C

604±45 A

0

11±0 C

6.66±0.16 A

0.0069±0.0003 B

548±157 C

0.000±0.000 C

50±0 A

348±11 C

Mixed

2.5

13±1 B

5.28±0.11 A

0.0184±0.0008 AB

1924±94 BC

0.033±0.007 B

37±8 AB

418±7 BC

Peracid

5

11±0 C

6.16±0.04 B

0.0237±0.0017 AB

2896±480 B

0.047±0.008 B

32±10 BC

604±84 A

10

15±0 A

4.94±0.11 C

0.0771±0.0502 A

5444±960 A

0.106±0.005 A

16±4 C

497±8 AB

Sanitizing
Agent

155

Table 4.11 (cont’d)
a

Means with different capital letters designate are physicochemical parameters differ significantly in terms of organic load (P ≤

0.05).

156

4.4.6 Physicochemical parameters of commercial flume water. The available chlorine
concentration was 65 ppm with a pH of ~5 and an ORP of 868 mV for the commercial flume
water containing chlorine + CA (Table 4.12). Total solids content averaged 0.0244 g with a
maximum of 0.0329 g after 3 h of processing. COD averaged 2550 mg O2/L, with a maximum of
3647 mg O2/L after 3 h of processing. MFV was ~35 ml.
The peroxyacetic acid concentration was 44 ppm with a pH of 5.56 and an ORP of 459
mV for the commercial flume water (Table 4.13). Total solids content averaged 0.0229 g with a
maximum of 0.0345 g seen after 4 h of processing. COD averaged 2130 mg O2/L, with a
maximum of 3216 mg O2/L seen after 4 h of processing. The average turbidity was 0.033, with a
maximum of 0.056 seen after 4 h of processing. MFV averaged ~45 ml.

157

Table 4.12: Physicochemical parameters of commercial flume water containing chlorine + CA
Mean ± SD Physicochemical Parameters
Available
Chlorine (ppm)

Free Chlorine
(ppm)

Temp
(°C)

Total Solids (g)

COD (mg
O2/L)

MFV
(ml)

ORP
(mV)

Industry
Flume Avg.

65 ± 12

0.31 ± 0.34

3±1

5.06 ± 0.78 0.0244 ± 0.0066

2550 ± 891

35 ± 7

868 ± 74

Industry
Flume Max.

90 ± 0

2.15 ± 1.15

3±1

5.53 ± 0.95 0.0329 ± 0.0051

3647 ± 646

24 ± 2

a

1000 ± 0

pH

a

MFV inversely correlated with organic load. Minimum MFV is shown.

158

Table 4.13: Physicochemical parameters of commercial flume water containing peroxyacetic acid

POAA
(ppm)
Industry
Flume
Avg.

T (°C)

pH

44 ± 6

7±2

5.56 ± 0.22

Mean ± SD Physicochemical Parameters
Turbidity (abs @
Total Solids (g/10 ml) COD (mg O2/L)
663 nm)
0.0229 ± 0.0127

Industry
Flume
58 ± 13 11 ± 4 6.09 ± 0.38
0.0345 ± 0.0181
Max.
a
MFV inversely correlated with organic load. Minimum MFV is shown.

159

MFV
(ml)

ORP (mV)
459 ± 16

2130 ± 1088

0.033 ± 0.0098

45 ± 4

3216 ± 2124

0.056 ± 0.0070

35.5 ± 7

a

486 ± 17

4.5 DISCUSSION
The impact of organic load in simulated produce wash water on sodium hypochlorite
efficacy has been previously assessed in numerous small-scale laboratory experiments (57, 78,
80, 87, 107, 139). Additional studies have assessed the efficacy of peroxyacetic acid in simulated
processing water (1, 17, 65, 78, 95, 129) or peroxyacetic acid in the presence of an organic load.
(57, 64, 97, 139). However, relatively few studies have quantified the impact of various
wastewater parameters, such as temperature, free and total chlorine, COD, pH, and ORP on the
efficacy of chlorine dioxide (10) and electrolyzed water (56, 79) against E. coli O157:H7, with
this study being the first to correlate E. coli O157:H7 inactivation to the physicochemical
parameters of leafy green flume water in a processing line.
The same 4-strain cocktail of non-toxigenic, GFP-labeled E. coli O157:H7 from previous
studies (22, 23) as well as in Chapters 2 and 3 was used for this work. As in Chapter 3, lettuce
was held for 24 h at 4°C rather than 1 h at room temperature to better simulate industrial
conditions and promote attachment of E. coli O157:H7. Additionally, recovery of sub-lethally
injured E. coli O157:H7 cells was enhanced by counting the plates after 48 rather than 24 h of
incubation in accordance with other sub-lethal injury studies (68, 135).
Chlorine is the most widely used sanitizer in the fresh produce industry due to its low
cost and minimal impact on product quality (62). Additionally, chlorine has been effective in
previous studies with low organic load, resulting in E. coli O157:H7 populations in the wash
water being below the limit of detection after exposure (78, 139). Increased concern over the
impact of organic load on chlorine efficacy and the potential for noxious or carcinogenic offgassing have lead to a heightened interest in methods to better monitor sanitizer efficacy against
pathogens in wash water (33) and in alternatives to chlorine, such as peroxyacetic acid-based
160

sanitizers (1, 62). Free chlorine concentrations in clean commercial flume water generally range
from 10 to 200 ppm (55, 91, 122). However, as processing continues, soil, debris, and vegetable
latexes will accumulate in the flume water (86), decreasing chlorine efficacy (5, 77, 100, 139).
This study was performed using organic loads of 0, 2.5, 5 and 10% in order to simulate flume
water at different stages of processing. The rate at which the organic load increases in
commercial flume systems will vary depending on the type and amount of product being washed,
product shred size, the rate of processing, and the means by which the organic load is decreased
in the water during processing (e.g., separating screens, foam removal, filters, clarifiers,
precipitators). The maximum COD and total solids of commercial flume water containing
chlorine + CA collected in this study correspond to organic loads of ~9% and 5.6% (w/v blended
iceberg lettuce) based on values obtained from simulated flume water prepared for pilot-scale
processing. These differences in physicochemical parameters between the commercial and pilotscale processing line are likely due to variations in particle size and the nature of the organic load
(latexes, soil, and cellular fluids).
Organic load rarely impacted sanitizer efficacy against E. coli O157:H7 on iceberg
lettuce. E. coli O157:H7 persisted on the lettuce throughout processing, even with an organic
load of 0% - where E. coli O157:H7 populations were typically near or below the lower limit of
detection in flume water. When Buchholz and others (22) used the same leafy green processing
line to quantify E. coli O157:H7 transfer during production, 0.20 and 0.67% of the total E. coli
O157:H7 population was recovered from the shredder and step conveyor, respectively, after
6

processing iceberg lettuce inoculated at 10 CFU/g. Given the contact between uninoculated
lettuce and E. coli O157:H7-contaminated equipment during the processing of batches 1, 2 and
3, the shredder and conveyor also likely contributed to the numbers of E. coli O157:H7
161

recovered from the contaminated lettuce (93). A small-scale laboratory study completed by
Zhang and others (139) found that E. coli O157:H7 transferred from inoculated to uninoculated
lettuce leaves during washing while no E. coli O157:H7 cells were recovered from processing
water containing 30 ppm mixed peracid and a 10% organic load. This phenomena was also
observed in a pilot-scale study by Luo and others (81) while processing both E. coli O157:H7
inoculated and uninoculated spinach in wash water containing chlorine + T-128. However,
inoculated lettuce in the study reported here was used to expose the flume water to E. coli
O157:H7 in order to determine sanitizer efficacy against the pathogen in wash water, not to
determine the effectiveness of washing against E. coli O157:H7 on the lettuce. The persistence of
E. coli O157:H7 on finished product indicates that factors other than sanitizer efficacy of the
wash water contribute to end-product safety.
In this study, after first processing 5.4 kg of inoculated iceberg lettuce, a total of only
16.4 kg of uninoculated lettuce was processed in three 5.4-kg batches at 10 min-intervals during
a total processing period of 44 min. The pilot-scale processing line as used in this study is
capable of processing ~216 kg/h or 158 kg during the 44 min of sample collection. The
inoculated lettuce served as a vehicle to contaminate the flume water for sanitizer efficacy testing
rather than a means to assess effectiveness of washing. The relatively small amount of lettuce
processed in this study offers “snapshots” of what may have been seen had the maximum of 158
kg been processed.
A linear regression analysis was conducted to further assess E. coli O157:H7 reductions
for each organic load and chlorine treatment. Using a 0% organic load, E. coli O157:H7
populations immediately decreased > 5 log CFU/ml for all three chlorine treatments, resulting in
trend lines with minimal or no slope and intercepts ranging from -5.4 to -5.8 log CFU/ml. E. coli
162

O157:H7 inactivation by chlorine was inversely related to organic load. While chlorine + CA
and chlorine + T-128 were also impacted by organic load, the inactivation rates at the higher
organic loads of 5 and 10% (-0.1 and -0.1 log CFU/ml per min, respectively) were similar to
those seen at a 2.5% organic load with chlorine alone (-0.1 log CFU/ml per min). This result
illustrates the importance of acidifying chlorinated wash water to increase the hypochlorous acid
2

content, thereby increasing the efficacy of chlorine at higher organic loads. The R values for the
chlorine + CA and chlorine + T-128 trend lines at an organic load of 2.5% were 0.70 and 0.60,
respectively, which is attributable to the variable nature of the E. coli O157:H7 reduction in wash
water. P values for the inactivation trend lines for chlorine, chlorine + CA and chlorine + T-128
at organic loads of 2.5, 5 and 10% were < 0.0001, indicating that the trend lines were accurate
predictors of E. coli O157:H7 inactivation. As each batch of uninoculated lettuce entered the
flume tank 10, 21.5, and 33 min after the initial batch of inoculated lettuce was processed, some
E. coli O157:H7 cells that were likely transferred to the lettuce during shredding and conveying
2

were released into the water during washing. Unsurprisingly, the R values for the regional trend
lines (0 – 10, 11.5 – 21, 23 – 32.5, and 34.5 – 44 min) were typically higher than the trend lines
developed over the entirety of procession (0 – 44 min). However, the variable nature of a
commercial processing line placed an emphasis on creating trend lines for the entire 44 min of
exposure. Additionally, the regional trend lines typically had higher P values than the trend lines
developed from all 44 min of processing- indicating a weaker fit to the inactivation of E. coli
over shorter periods of time.
As discussed in Chapter 2, E. coli O157:H7 populations were consistently higher in
flume water containing peroxyacetic acid or mixed peracid than in flume water containing
163

chlorine acidified with citric acid or T-128. Hence, E. coli O157:H7 populations were expected
to persist in the flume water during processing, especially at higher organic loads when
peroxyacetic acid-based sanitizers were used. However, using the bench-top system developed in
Chapter 3, E. coli O157:H7 did not persist in wash water containing peroxyacetic acid or mixed
peracid after removal of the inoculated product. These findings are similar to those from the
current processing line study where E. coli O157:H7 populations typically decreased > 5 log
CFU/ml in the wash water with little persistence occurring. While some significant differences in
E. coli O157:H7 reductions were observed in the flume water based on organic load, notably
after 21 and 44 min of processing, populations were often below the limit of detection which
decreased with increasing organic load in the flume water.
The maximum total solids and COD of the commercial flume water containing
peroxyacetic acid corresponded to organic loads of 6.3 and 6.2% (w/v blended iceberg lettuce),
respectively, based on comparisons to the same values from flume water prepared in the pilotscale processing line. At an organic load of ~6.25%, the commercial flume water likely provided
a log linear inactivation of Log(N/N0) = -4.7. However, minimal E. coli O157:H7 persistence
was seen in the flume water during pilot-scale processing, regardless of the dependent
physiochemical parameters of the water. Therefore, as in the bench-top carboy study, the
dependent physicochemical parameters in Table 4.11 are likely not ideal for predicting E. coli
O157:H7 inactivation rates in flume water containing different organic loads for either
peroxyacetic acid or mixed peracid in Table 4.8. Monitoring the sanitizer concentration may be
the most accurate predictor of E. coli O157:H7 inactivation. While the log linear inactivation
trend lines derived from the E. coli O157:H7 reductions in flume water may not correlate well to

164

2

time considering the low R values, the y-intercept can be used to predict the E. coli O157:H7
reductions in flume water for both peroxyacetic acid and mixed peracid.
Regardless of the chlorine treatment, total solids and COD were the best indicators of
organic load in the flume water. The physicochemical parameters of the wash water in Tables 4.9
and 4.10 can be correlated to the inactivation equations in Table 4.3 to predict E. coli O157:H7
persistence. Based upon the persistence of E. coli O157:H7 at organic loads of 2.5, 5 and 10%
with all three chlorine treatments, the physicochemical parameters of flume water should be
maintained within the standard deviations for a 0% organic load to minimize the possibility of
water becoming a carrier for E. coli O157:H7. The numerous physicochemical parameters
monitored in this study offer alternatives to ORP which is most widely used to assess sanitizer
efficacy in commercial flume systems (112, 125). However, since a minimum of 2 h is needed to
determine COD and total solids, both of these tests are impractical for routine testing. In order
for commercial processors to rapidly predict sanitizer efficacy of their flume water, studies
specific to the commercial processing line in question must be completed to determine which
rapid tests (e.g. turbidity or MFV) correlate to the more time-consuming but more accurate tests
of COD and/or total solids.
One additional study in which a pilot-scale leafy green processing line was used to
process inoculated spinach leaves reported E. coli O157:H7 reductions of < 1 log on the product
when the flume water containing chlorine-based sanitizer was acidified to pH 5.0 with T-128.
(81). However, E. coli O157:H7 populations were significantly lower in the wash water (81). In
this study, a pH of 6.5 was used to compare the efficacy of T-128 to the targeted pH when citric
acid is used (62, 81). An additional study by the same group concluded that T-128 on its own had

165

weak bactericidal activity and did not play a significant role in pathogen inactivation (87), which
emphasizes the products role as a means to enhance the efficacy of chlorine.
A survey of commercial leafy greens was conducted by Barrera and others (13) to
determine the impact of different processing parameters on the efficacy of various sanitizing
agents during processing. Peroxyacetic acid concentration, pH, turbidity, and solids content of
the wash water were correlated to aerobic and coliform counts on spinach before and after
washing at one of the three facilities tested. An average peroxyacetic acid concentration of 6.1
ppm was maintained by manual dosing. Average aerobic and coliform counts were not
significantly lower after washing, which is not surprising considering the concentration of
peroxyacetic acid in the wash water. Additionally, average coliform counts in the wash water of
3.12 log CFU/100 ml were negatively correlated with the log reductions, which were attributed
to the lower average sanitizer concentration. There was a positive correlation between the solids
content of the wash water and the reductions of aerobic bacteria, emphasizing that organic load
plays a critical role in sanitizer efficacy, at least with a low peroxyacetic acid concentration. This
finding emphasizes the importance of monitoring sanitizer concentration in the wash water in
addition to organic load.
Regardless of the chlorine treatment, E. coli O157:H7 populations were typically higher
on all wet equipment surfaces when an organic load was present in the wash water, emphasizing
the spread of E. coli O157:H7 by means other than washing. Populations on the flume tank,
shaker table and dewatering centrifuge were rarely significantly different at any of the organic
loads tested for peroxyacetic acid and never significantly different for mixed peracid. The fact
that E. coli O157:H7 was quantifiable on most surfaces after 44 min of processing with either
sanitizing agent likely contributed to persistence of the pathogen on the product.
166

Regardless of the chlorine treatment, higher E. coli O157:H7 populations were always
seen in spent centrifugation water after washing each batch of lettuce in water containing an
organic load. Levels of E. coli O157:H7 recovered from the spent centrifugation water were
rarely significantly different at any of the organic loads for peroxyacetic acid and never
significantly different with mixed peracid. While often below the limit of detection in the flume
water, E. coli O157:H7 was consistently present at levels above the lower limit of detection in
the centrifugation water (0.2 CFU/ml). Therefore, spent centrifugation appears to be an attractive
target for pathogen testing during commercial processing.
This study was conducted to assess the impact of organic load on sodium hypochlorite
and peroxyacetic acid efficacy against E. coli O157:H7 on lettuce, in wash water, and on
equipment surfaces during small-scale commercial leafy green processing, with the pathogen
recovered from all uninoculated batches of lettuce processed after the inoculated batch. Chlorine
efficacy markedly decreased with increasing organic load organic load, resulting in increased E.
coli O157:H7 persistence in the flume water. Total solids, COD and turbidity of the flume water
were inversely related to linear trend lines of E. coli O157:H7 inactivation, with the opposite true
for ORP and MFV. Since organic load rarely impacted the efficacy of either peroxyacetic acid or
mixed peracid, the dependent physicochemical parameters that were assessed for flume water are
not ideal predictors of their efficacy, with sanitizer concentration being a better predictor of E.
coli O157:H7 inactivation. While the minimum effective concentration of peroxyacetic acid or
mixed peracid was not determined in this study, concentrations of 50 ppm should result in
minimal E. coli O157:H7 persistence, regardless of organic load. Based on these findings,
commercial processors should consider maintaining the organic load in chlorinated flume water
as close to 0% as possible to minimize microbial cross-contamination during processing.
167

However, given the ability of E. coli O157:H7 to persist on iceberg lettuce throughout
processing, the incorporation other microbial intervention steps in addition to flume washing
with a sanitizer may be advisable to better enhance end-product safety.

168

CONCLUSIONS

Flume washing remains a critical step during commercial processing of leafy greens in
order to remove soil and debris, which decreases the microbial load, improves quality and
appearance, and enhances product shelf life and safety. Due to the large amount of product that is
regularly processed, flume water is recirculated to reduce waste and operational costs. Various
sanitizing agents are regularly added to the water in order to prevent the water from becoming a
microbial carrier. The first research objective assessed the efficacy of five different commercial
sanitizing treatments against E. coli O157:H7 during pilot-scale processing of iceberg lettuce. E.
coli O157:H7 populations decreased 0.8 – 1.4 log CFU/g on the lettuce, with no sanitizing agent
being significantly more effective than water alone. When compared to flume water containing
peroxyacetic acid or mixed peracid, significantly fewer E. coli O157:H7 were typically
recovered in flume water containing chlorine, chlorine + citric acid, or chlorine + T-128 during
processing of inoculated lettuce, with some populations falling below the lower limit of
detection. However, this lettuce was processed in the absence of an organic load under a “bestcase” scenario. Since produce sanitizers are designed to decrease the persistence of microbial
contaminants during processing, it is critical that they remain effective throughout processing in
order to prevent the amplification of an isolated contamination event.
Organic load, consisting of plant tissues and cellular fluids released during cutting in
addition to soil, insects, and microbes, will accumulate in recirculating flume water as produce is
washed, decreasing the ability of sanitizers to minimize cross-contamination from the water during
processing. Chlorine is the most widely used sanitizing agent by commercial leafy greens processors
due to its low cost. However, organic load and pH are well known to impact the efficacy of

169

chlorine, increasing the interest in chlorine supplements, such as T-128, or chlorine alternatives,
such as peroxyacetic acid. Given the various commercial sanitizers available and the variable
rates at which organic material accumulates in flume water during processing, commercial leafy
green processors are clearly in need of better means to both predict and monitor the efficacy of
sanitizers during processing.
The second research objective focused on the persistence of E. coli O157:H7 in wash
water containing sanitizers and different organic loads and the correlation of various
physicochemical parameters of the wash water to sanitizer efficacy. A novel carboy system was
developed to simulate leafy green washing in a pilot-scale processing line. Organic load
impacted the efficacy of all three chlorine treatments, with typical E. coli O157:H7 reductions of
< 1 log CFU/ml after 10 min of exposure. In contrast, efficacy of peroxyacetic acid and mixed
peracid was largely unaffected by organic load, with average E. coli O157:H7 reductions of ~4.8
and ~5.5 log CFU/ml, respectively, after 10 min of exposure. Reduced sanitizer efficacy
generally correlated to increased total solids, chemical oxygen demand, turbidity, and decreased
maximum filterable volume, indicating that these tests may be effective alternatives to the
industry standard of oxygen/reduction potential.
The research goals of the third study were to 1) quantify the persistence of E. coli
O157:H7 in wash water and on lettuce during pilot-scale processing using flume water
containing five different sanitizers and various organic loads and 2) assess the relationship
between various physicochemical parameters and organic load of the wash water on E. coli
O157:H7 inactivation. E. coli O157:H7 was recovered from all lettuce that was processed after
the inoculated batch of iceberg lettuce, regardless of organic load in the wash water. Organic
load negatively impacted the efficacy of all three chlorine treatments, with typical E. coli
170

O157:H7 reductions of > 5 log by the end of processing with no organic load in the wash water
and 0.9 – 3.7 log CFU/ml with a 10% organic load. Much like the second study, reduced
sanitizer efficacy generally correlated to significant differences in total solids, chemical oxygen
demand, turbidity, and maximum filterable volume, indicating that these tests may be effective
alternatives to the industry standard of oxygen/reduction potential. Additionally, the numerous
physicochemical parameters of the flume water correlated to E. coli O157:H7 inactivation
throughout processing. These findings emphasize the importance of monitoring flume water
quality as a means to predict chlorine efficacy. However, organic load rarely impacted the
efficacy of either peroxyacetic acid or mixed peracid, with typical reductions of > 5 log CFU/ml
in wash water throughout processing for all organic loads. Additionally, E. coli O157:H7
persisted on all previously uninoculated lettuce following the inoculated batch, emphasizing the
need for improved intervention strategies that can better ensure end-product safety.
The two major limitations to this research include the use of 1) only one sanitizer
concentration for each sanitizing agent used in each individual study and 2) and a processing
water temperature that was higher than the targeted industry standard of 4°C. Chemical sanitizer
efficacy is concentration dependent; however, these studies did not assess the extent to which
sanitizer concentration impacted E. coli O157:H7 inactivation. Due to the lack of a chilling
system in the pilot-scale leafy green processing line used in Chapters 2 and 4, the average water
temperature was ~14°C. The impact of water temperature on chlorine + CA efficacy was
described in Chapter 3; however, the impact of temperature on any of the sanitizing agents
during pilot-scale processing was not assessed. Additionally, the organic load concentrations
used in the studies described in Chapter 4 were fixed at specific concentrations of 0 – 10% (w/v
blended lettuce). Organic load levels would naturally accumulate over an entire processing shift
171

in a commercial operation, theoretically resulting in variable sanitizer efficacy throughout
processing.
Overall, this research showed that sanitizing agents in flume water are effective against
E. coli O157:H7 in wash water when no organic load was present. However, chlorine efficacy
was significantly impacted by the presence of organic load in the wash water, allowing E. coli
O157:H7 to persist in the wash water and on finished product. Conversely, both peroxyacetic
acid and mixed peracid remained largely unaffected by the presence of organic load in the wash
water. With sanitizing agents in wash water remaining the sole microbial mitigation strategy
during commercial processing of leafy greens, it is imperative that growers and harvesters follow
GAPs and GHPs in order to minimize product contamination before processing because
conventional flume washing will only provide a 1 – 3 log reduction of microbial contaminants.
Considering that chlorine is the most used sanitizing agent in the leafy green industry and that its
efficacy is impacted by organic load, monitoring the physicochemical parameters of the water is
critical to ensure that the flume water does not promote the amplification of an isolated
contamination event.

172

FUTURE RESEARCH RECOMMENDATIONS

The studies presented in Chapters 3 and 4 showed that both peroxyacetic and mixed
peracid maintained their efficacies in the presence of organic load, resulting in E. coli O157:H7
populations that were typically below the lower limit of detection in the wash water. Chlorinebased sanitizers are more cost-effective than peroxyacetic acid-based sanitizers; however the
results of Chapters 3 and 4 showed that their efficacy is significantly impacted in the presence of
organic load. Considering that a concentration of 50 ppm for both peroxyacetic acid and mixed
peracid consistently provided E. coli O157:H7 reductions of > 5 log CFU/ml, it is likely that
lower concentrations may be just as effective.
A study could be completed using the carboy system described in Chapter 3 in which
various concentrations of peroxyacetic acid and mixed peracid are assessed in wash water
containing organic load. Even though both POAA-based sanitizers maintained their efficacy at
50 ppm in the presence of organic load, it would be critical to examine the impact of organic
load on lower POAA concentrations in order to verify that organic load immunity was not
dependent on sanitizer concentration. Once a minimum effective dose was determined in the
carboy model, a study similar to that presented in Chapter 4 could be conducted in order to
determine if the lower sanitizer concentration prevented E. coli O157:H7 persistence in flume
water throughout processing of numerous batches of leafy greens. Determination of the
minimum effective dose of peroxyacetic-acid based sanitizers would give commercial leafy
green processors a target concentration thereby preventing the over-use of sanitizing agents.
Maintenance of a lower peroxyacetic acid concentration would reduce operational costs and

173

corrosion to processing equipment. It is likely that commercial processors would be more likely
to adopt the use of peroxyacetic acid if it was more cost-effective.
The studies presented in Chapters 3 – 4 focused on the impact of organic load (blended
iceberg lettuce) on sanitizer efficacy against E. coli O157:H7. In a commercial processing
facility, organic load may be comprised of soil and debris in addition to cellular fluids and
latexes and may not have a similar impact on sanitizer efficacy. Additionally, other fresh produce
products (e.g., baby spinach, tomatoes, celery, carrots, and onions) are also frequently subjected
to flume washing after cutting or dicing, resulting in the inevitable accumulation of cellular
fluids in the wash water. Due to the differences between the compositions of products in regards
to pH, latexes, and cellular fluids, no conclusions can be made regarding the impact of other
types of organic load on sanitizer efficacy based upon the series of experiments performed with
iceberg lettuce in Chapters 3 – 4.
The carboy system described in Chapter 3 could be modified to examine the impact of
various types of organic load (e.g., soil, various blended produce products) on both chlorine and
peroxyacetic acid-based sanitizers. Physicochemical parameters of the wash water could be
correlated to the inactivation of native microflora or other pathogens of concern, such as Listeria
monocytogenes or Salmonella. Such research would benefit a much larger group of produce
processors and may help to prevent future product recalls and outbreaks as demand for
convenient, value-added fresh-cut produce increases.
6

The lettuce used in these studies was dip-inoculated to contain 10 CFU/g of E. coli
O157:H7 in order to provide quantifiable results. The levels of E. coli O157:H7 on leafy greens
contaminated during pre- or postharvest operations, while unknown, are thought to be lower than
the inoculation level used in the studies discussed in Chapters 2 – 4. Inoculated lettuce that was
174

used in the studies discussed in Chapters 3 – 4 was held for 24 h at 4°C after dip inoculation in
order to simulate industrial practice as well as promote the attachment of E. coli O157:H7 to the
leaf surface.
A study could be conducted to assess the impact of low-level field inoculation (by
spraying or flooding) of heads of iceberg lettuce on E. coli O157:H7 persistence during
subsequent pilot-scale processing. It is assumed that the greater attachment and potentially even
biofilm formation by E. coli O157:H7 on the leaf surface would occur in the field- grown as
compared to dip-inoculated lettuce as used in the previously described studies. Due to greater
attachment and lower levels of E. coli O157:H7 on the leafy greens, quantification or even
detection would likely be difficult with typical methods. In order to detect the pathogen, a
Portable Multi-use Automated Concentration System (PMACS) would likely need to be used.
The PMACS rapidly and efficiently collects microbial pathogens and other particulates from
large volumes of water and uses a back-flush protocol to recover the pathogens in a concentrated
sample for subsequent analysis. The results of a study conducted with field inoculated lettuce
would determine what difference, if any, inoculation method has on E. coli O157:H7 persistence
in wash water and on finished product during subsequent pilot-scale processing.
This dissertation primarily focused on the efficacy of various chemical sanitizers against
E. coli O157:H7 in wash water, with lettuce essentially used as a means to inoculate the water.
However, inactivation of E. coli O157:H7 on lettuce using wash water containing sanitizing
agents was rarely impacted by organic load. While it was not surprising, it is still a troubling
result of how fresh-cut leafy greens are processed and reinforces countless previous claims that
pathogens cannot be removed from product by conventional washing once it is contaminated.
Considering the results of Chapter 4, where E. coli O157:H7 was recovered from all previously
175

uninoculated lettuce that was processed after the inoculated batch, it is clear that microbial
intervention steps other than having effective sanitizing agents in the wash water are critical to
ensure end-product safety. Considering that the amount of uninoculated lettuce (16.2 kg) that
was processed after the inoculated batch was relatively small in comparison to the throughput of
a commercial processor, the extent to which E. coli O157:H7 can spread to finished product
under more realistic processing conditions remains unclear. A previous study by our group
showed that E. coli O157:H7 transferred to all 907 kg of iceberg lettuce that was processed
subsequent to E. coli O157:H7-inoculated product. However, that study was performed without
sanitizing agents in the wash water. In order to quantify the persistence of E. coli O157:H7 on a
larger amount of product, a similar large-scale study could be conducted with sanitizing agents in
the wash water.
Since it has been established that conventional washing procedures will only reduce
microbial populations 1 – 3 log CFU from the surfaces of leafy greens, it is critical that
contamination of the product is prevented. However, it is important that steps are taken to
prevent the amplification of an isolated contamination event during processing of leafy greens
because contamination of product will undoubtedly continue. Considering that E. coli O157:H7
was found throughout the previously uninoculated product after processing in the study
conducted in Chapter 4, regardless of the efficacy of the sanitization agents, it is clear that
contamination is being amplified in processing steps that occur during washing- namely
shredding and conveying. Visual inspection of the shredder and conveyor after processing
showed that some product remained on the shredder and conveyor, some of it likely the
contaminated product. Modifications to the pilot-scale processing line could be made where
sanitizing sprayers are installed inside the shredder and on the conveyor to minimize the
176

accumulation of product on equipment surfaces. Another could assess the persistence of E. coli
O157:H7 in wash water and on finished product after processing in the modified pilot-scale leafy
green processing line. It is expected that the reductions on the inoculated product would be
similar to the studies conducted previously and the amplification contamination would be
inhibited, thereby improving end-product safety.

177

APPENDICES

178

APPENDIX I:

Impact of Organic Load and Sanitizer Concentration on Inactivation of a Four-Strain
Cocktail of Escherichia coli O157:H7 in Simulated Leafy Green Processing Water

179

AI.1 ABSTRACT
Chemical sanitizers are routinely used during commercial flume washing of fresh-cut
leafy greens to minimize cross-contamination from the water. This study assessed the efficacy of
various concentrations of chlorine, peroxyacetic acid, and mixed peracid against E. coli O157:H7
in simulated wash water containing various organic loads. Four liters of water containing organic
loads of 0, 1, 5 and 10% (w/v blended lettuce) and chlorine (30, 50 and 100 ppm), peroxyacetic
acid (10, 50 and 80 ppm), or mixed peracid (10, 50 and 80 ppm) were used to assess E. coli
O157:H7 survival over 90 sec of exposure. Organic load negatively impacted the efficacy of all
three sanitizing agents at each concentration, resulting in significantly (P ≤ 0.05) higher E. coli
O157:H7 populations. Increasing organic load correlated to significant increases in COD,
turbidity, and total solids for all sanitizing treatments.

180

AI.2 INTRODUCTION
Wash water quality has been a major focus of the produce industry for many years with
these concern heightened in response to three nationwide outbreaks in 2006 traced to fresh-cut
lettuce and baby spinach. Despite the widespread use of chemical sanitizers in produce wash
water, efficacy of these sanitizers remains problematic due to the presence of organic material in
the water that decreases antibacterial activity.
The two aims of this study were to 1) determine the efficacy of three concentrations of
three different sanitizing agents against E. coli O157:H7 in water containing organic loads of 0,
1, 5 and 10% (w/v blended lettuce solids) and 2) assess various physicochemical parameters of
the wash water.

181

AI.3 MATERIALS AND METHODS
AI.3.1 Experimental design. The impact of four different organic loads on three
concentrations of three different sanitizing treatments against a 4-strain cocktail of E. coli
8

O157:H7 at 10 CFU/ml was assessed in triplicate in a 4 L glass carboy, with sanitizer-free water
serving as the control for each organic load. Various water samples were collected and
quantitatively examined for E. coli O157:H7 to determine persistence during 90 sec of exposure.
Temperature, pH, oxidation/reduction potential, chemical oxygen demand, total solids,
maximum filterable volume and turbidity of the wash water were also assessed.
AI.3.2 Leafy greens. Identical to 2.3.2.
AI.3.3 Bacterial strains. Identical to 2.3.3.
AI.3.4 E. coli O157:H7 cocktail. The four E. coli O157:H7 cultures were combined in
equal volumes in a 50 ml centrifuge tube to create a 36 ml cocktail.
AI.3.5 Processing equipment. Identical to 3.3.5.
AI.3.6 Wash water. Iceberg lettuce (0, 40, 200, or 400 g) was blended for 2 min in 250
– 500 ml of tap water using a household blender (Model BLC10650MB, Black & Decker, New
Britain, CT) and added to the wash water at ~14°C to achieve organic loads of 0, 1, 5, or 10%
(w/v), respectively, in 4 L. The following nine sanitizer treatments were assessed: 10, 50 and 80
ppm peroxyacetic acid (Tsunami 100, Ecolab, St. Paul, MN), 10, 50 and 80 ppm mixed peracid
(Tsunami 200, Ecolab), and 30, 50 and 100 ppm available chlorine (XY-12, Ecolab).
Peroxyacetic acid test kit 311 (Ecolab) was used to confirm the peroxyacetic acid and mixed
peracid sanitizer concentrations while chlorine test kit 321 (Ecolab) was used to measure
available chlorine. Sanitizer-free MSU tap water (< 0.05 ppm free chlorine) served as the control
for each for each organic load.
182

AI.3.7 E. coli O157:H7 exposure to wash water and sample collection. The E. coli
O157:H7 cocktail (4 ml) was inoculated directly into the carboy containing 4 L of wash water
during mechanical stirring, immediately after which the cork was removed from the spout. Water
samples (50 ml) were collected at 10 s intervals over 90 s of exposure in 50 ml centrifuge tubes
containing 38x concentrated Difco Neutralizing Buffer (Becton Dickinson).
AI.3.8 Physiochemical parameters. Identical to 3.3.8.
AI.3.9 Microbiological analyses. Identical to 3.3.9.
AI.3.10 Sanitizer neutralization confirmation. One liter aliquots of tap water
containing 30, 50 and 100 ppm chlorine (XY-12, Ecolab), 10, 50 and 80 ppm peroxyacetic acid
(Tsunami 100, Ecolab) or 10, 50 and 80 ppm mixed peracid (Tsunami 200 ppm, Ecolab) with
organic loads of 0, 1, 5 or 10% were prepared using a mechanical blender (Model
BLC10650MB, Black & Decker, New Britain, CT), after which the concentration of available
chlorine or peroxyacetic acid was confirmed using a commercial test kit (Chlorine test kit 321,
Ecolab and Peroxyacetic acid test kit 311, Ecolab). A 50 ml centrifuge tube containing 3 ml of
38x concentrated neutralizing buffer (BD) was filled with these prepared solutions, agitated for 5
s and then similarly assessed for chlorine or peroxyacetic acid neutralization. In preliminary
experiments, a 38x concentration neutralized all sanitizer concentrations without impacting E.
coli O157:H7 recovery.
AI.3.11 Statistical analysis. E. coli O157:H7 counts were converted to log CFU/ml and
individual physicochemical parameter values were subjected to an ANOVA using JMP 9.0 (SAS
Institute Inc., Cary, NC). For samples without E. coli O157:H7 counts, values equaling half the
limit of detection were used. A P value of ≤ 0.05 was considered significant for all tests. The
Tukey-Kramer HSD test was used to identify significant differences in E. coli O157:H7
183

populations for individual lettuce and water samples as well as individual physicochemical
parameter values.

184

AI.4 RESULTS
AI.4.1 Wash water. Average E. coli O157:H7 populations in the water controls ranged
from 5.9 – 6.0 log CFU/ml across all organic loads (Table AI.1). A concentration of 10 ppm
peroxyacetic acid was impacted by organic load, with significantly higher (P ≤ 0.05) E. coli
O157:H7 populations typically recovered in organic loads of 1, 5 and 10% (Table AI.2).
Concentrations of 50 and 80 ppm peroxyacetic acid were also impacted by organic load, with
average E. coli O157:H7 populations recovered after 50 s typically significantly lower with a 0%
organic load (-2.0 to -1.0 log CFU/ml) than at a 10% organic load (0.0 to 2.9 log CFU/ml). All
three concentrations of mixed peracid were impacted by organic load, with significantly fewer (P
≤ 0.05) E. coli O157:H7 consistently recovered with a 0% organic load (-2.0 to 5.7 log CFU/ml)
than at a 10% organic load (-1.0 to 6 log CFU/ml) after 50 s of exposure (Table AI.3). Organic
load had a negative impact on all chlorine concentrations, with significantly (P ≤ 0.05) fewer E.
coli O157:H7 populations recovered at a 0% organic load (-2.0 to -1.8 log CFU/ml) than at a
10% organic load (5.7 – 6.1 log CFU/ml) at all times sampled (Table AI.4).

185

Table AI.1: E. coli O157:H7 populations in wash water controls

a

Mean ± SD E. coli O157:H7 (log CFU/ml)
%
Organic
Load
0

10
5.5±0.2 B

20
5.8±0.1 A

30
5.8±0.1 B

40
5.9±0.1 A

50
5.9±0.1 A

60
6.0±0.1 AB

70
6.0±0.1 AB

80
6.1±0.1 A

90
6.0±0.1 AB

1

5.6±0.1 B

5.9±0.0 A

6.0±0.1 A

6.0±0.0 A

6.0±0.0 A

6.1±0.1 A

6.1±0.1 A

6.1±0.0 A

6.1±0.1 A

5

6.0±0.1 A

5.9±0.1 A

5.9±0.0 AB

5.9±0.1 A

5.9±0.1 A

5.9±0.1 AB

5.9±0.0 B

5.9±0.0 B

5.9±0.0 AB

10

6.0±0.1 A

5.9±0.0 A

5.9±0.1 AB

5.9±0.1 A

5.9±0.1 A

5.9±0.1 B

5.9±0.0 B

5.9±0.1 B

5.8±0.1 B

a

Means with different capital letters designate E. coli O157:H7 populations that differ significantly in terms of time (P ≤ 0.05).

186

Table AI.2: E. coli O157:H7 populations in wash water containing peroxyacetic acid

a

Mean ± SD E. coli O157:H7 (log CFU/ml)
%
POAA Organic
(ppm)
Load
0

10
20
30
4.9±0.8 A 5.6±0.3 A 5.8±0.2 A

40
5.8±0.2 A

50
60
70
5.6±0.1 B 5.5±0.1 B 5.2±0.3 B

80
4.6±0.8 B

90
4.0±1.1 B

1

5.7±0.4 A 6.0±0.2 A 6.0±0.0 A

6.0±0.1 A

6.0±0.1 A 5.9±0.2 A 5.8±0.1 A 5.6±0.1 AB 5.4±0.3 AB

5

6.0±0.2 A 6.1±0.1 A 6.0±0.1 A

5.9±0.1 A

6.0±0.1 A 6.0±0.1 A 6.0±0.1 A

6.0±0.2 A

5.9±0.1 A

10

6.0±0.1 A 6.0±0.1 A 6.0±0.0 A

6.0±0.0 A

5.9±0.1 A 5.9±0.0 A 5.9±0.1 A

5.8±0.2 A

5.8±0.3 A

10

187

Table AI.2 (cont’d)

4.8±0.7
0

3.8±1.0 B

0.8±0.2 B

5.2±0.1
1

B

4.8±0.3

1.8±1.1

AB

AB

AB

-0.4±0.8 C

-1.0±0.6 C

-1.8±0.2 B

-1.9±0.1 B

0.7±1.3

0.3±1.8

0.2±1.9

BC

AB

AB

0.8±1.5 BC

-1.3±1.2 A

-2.0±0.0 B
0.7±2.4

0.2±2.0 A

AB

50
5.7±0.1
5

1.9±0.6
5.4±0.3 A

5.9±0.2
10

A

AB

1.2±1.1

4.7±0.3

A

1.1±0.5

3.5±0.8 A

1.9±0.2 AB

AB

1.6±0.3 A

AB

0.9±1.0 A

AB

3.3±0.3 A

3.1±0.1 A

2.9±0.5 A

2.3±1.2 A

2.0±1.4 A

2.1±0.9 A

2.0±0.6 A

188

Table AI.2 (cont’d)

4.2±0.8
0

A

1.7±0.7 A

1.0±0.6 A

-0.7±1.5 A

-1.9±0.2 B

-2.0±0.0 B

-2.0±0.0 B

4.5±0.4
1

A

-2.0±0.0 C

-2.0±0.0 A

-1.7±0.4
1.3±0.6 A

0.3±0.9 A

-1.0±0.7 A

-1.4±0.6 B

-1.4±0.8 B

-0.8±1.4 B

BC

-1.6±0.5 A

-0.8±0.9 B

-1.2±0.2 B

-0.9±0.5 B

-1.3±0.0 B

-0.2±1.0 A

2.2±0.4 A

2.1±1.0 A

2.3±0.4 A

0.8±0.0 A

0.0±1.1 A

80
5.1±0.3
5

A

0.05±0.13
1.8±0.3 A

-0.0±0.6 A

4.3±0.5
10
a

A

A
1.11±1.08

0.9±0.4 A

0.1±1.0 A

A

Means with different capital letters designate E. coli O157:H7 populations that differ significantly in terms of each sanitizer

concentration and time (P ≤ 0.05).

189

Table AI.3: E. coli O157:H7 populations in wash water containing mixed peracid

a

Mean ± SD E. coli O157:H7 (log CFU/ml)

POAA
(ppm)

%
Orga
nic
Load
0

10
5.7±0.2 A

20
5.9±0.1 A

30
5.9±0.1 A

40
5.9±0.1 A

1

10

AB

AB
5.8±0.1

AB

AB

70
4.9±0.0 C

80
4.2±0.2 C

90
3.3±0.2 C

5.7±0.1

5.8±0.1

5.6±0.2

5

B

5.4±0.1

10

5.3±0.1 B

60
4.8±0.7 B

5.8±0.1

5.6±0.07

50
5.7±0.1 C

5.8±0.1 A

5.9±0.1 A

BC

AB

5.5±0.1 B

5.1±0.1 B

4.8±0.4 B

5.9±0.0 A

5.9±0.0 A

6.0±0.1 A

5.9±0.0 A

5.9±0.1 A

5.9±0.1 A

5.8±0.2 A

6.0±0.1 A

6.0±0.0 A

6.0±0.1 A

6.0±0.0 A

6.0±0.1
6.0±0.1 A

5.9±0.1 A

190

AB

Table AI.3 (cont’d)
0

4.6±0.2 A

2.6±0.3 A

-1.0±1.1 B -1.9±0.1 B -2.0±0.0 B -2.0±0.0 B -2.0±0.0 B -2.0±0.0 B -2.0±0.0 B
0.4±0.2

1

4.9±0.1 A

3.1±0.2 A

-1.5±0.2

5

5.0±0.4 A

3.0±0.8 A

-1.5±0.2 B

0.5±0.4

50

AB

-0.5±0.8

AB

AB

AB

-1.5±0.2
-1.5±0.2 B -1.5±0.2 B

-1.5±0.2

AB

AB

-0.9±0.7
0.1±1.2 A

-0.9±0.6

-0.7±1.0

-1.3±0.0

AB

AB

AB

AB

0.1±0.8 A

0.4±1.5 A

-0.1±0.7
10

5.2±0.5 A

3.8±0.5 A

1.6±1.0 A

0.5±1.1 A

191

0.3±0.8 A

0.1±0.8 A

A

Table AI.3 (cont’d)

-2.0±0.0
0

1.3±1.0 A -1.9±0.1 B -1.9±0.1 C

D

-2.0±0.0 C

-1.6±0.0
1

1.9±0.4 A -1.6±0.0 B

BC

-2.0±0.0

-2.0±0.0

-2.0±0.0

-2.0±0.0

D

D

D

D

-1.50±0.2
-1.6±0.0 C

B

-1.6±0.0 C -1.6±0.0 C -1.6±0.0 C -1.6±0.0 C

80
-1.2±0.2

10
a

AB

AB

-0.3±0.9

5

-1.3±0.0

-0.7±0.5

-1.0±0.0

-1.0±0.0

-1.0±0.0

-1.0±0.0

-1.0±0.0

-1.0±0.0

A

A

A

A

A

A

A

A

1.0±0.7 A

0.8±1.8 A

-1.3±0.0 B -1.3±0.0 B -1.3±0.0 B -1.3±0.0 B -1.3±0.0 B -1.3±0.0 B

Means with different capital letters designate E. coli O157:H7 populations that differ significantly in terms of each sanitizer

concentration and time (P ≤ 0.05).

192

Table AI.4: E. coli O157:H7 populations in wash water containing chlorine

a

Mean ± SD E. coli O157:H7 (log CFU/ml)
Chlorine
(ppm)

%
Organi
c Load

0

10
-1.8±0.3

20
-1.9±0.1

30
-1.9±0.1

40
-1.8±0.2

50
-1.9±0.1

60

70
-2.0±0.0

80
-2.0±0.0

90
-2.0±0.0

B

B

B

B

B

-2.0±0.0 B

B

B

B

1

5.9±0.2 A 6.0±0.0 A 6.0±0.0 A 5.9±0.1 A 5.9±0.1 A

5.9±0.0 A

5.9±0.1 A 5.9±0.0 A 6.0±0.1 A

5

6.0±0.1 A 6.0±0.1 A 6.0±0.1 A 6.0±0.1 A 6.0±0.1 A

6.0±0.1 A

6.0±0.1 A 6.0±0.1 A 6.0±0.0 A

30

5.94±0.01
10

5.9±0.1 A 5.9±0.0 A 5.9±0.0 A 6.0±0.1 A 6.0±0.1 A

193

A

6.0±0.0 A 6.0±0.1 A 6.0±0.0 A

Table AI.4 (cont’d)

-1.9±0.1

-1.9±0.1

-1.9±0.1

-1.9±0.1

C

C

C

B

B

-0.7±1.0

0

-1.9±0.1

-0.6±0.9

B

B

-2.0±0.0

-2.0±0.0

-2.0±0.0

-2.0±0.0 B

B

B

C

-0.3±1.7 B

0.1±1.9 B

0.4±2.1 B

1.4±2.5 B

50
1

1.8±0.2 B

1.4±1.3 B

0.4±1.3 B

5

6.1±0.1 A 6.0±0.0 A 6.0±0.0 A 6.0±0.1 A 6.0±0.1 A

6.0±0.1 A

6.0±0.1 A 6.0±0.1 A 6.0±0.1 A

10

6.0±0.1 A 6.0±0.1 A 6.0±0.1 A 6.1±0.1 A 6.0±0.0 A

6.1±0.0 A

6.0±0.0 A 6.0±0.0 A 6.0±0.0 A

194

Table AI.4 (cont’d)
-1.9±0.1

-1.9±0.1

-1.9±0.1

-1.9±0.1

B

B

B

B

B

-1.6±0.0

0

-1.9±0.1

-1.6±0.0

-1.6±0.0

-1.6±0.0

-1.6±0.0

B

B

B

B

B

-2.0±0.0

-2.0±0.0

B

B

B

-1.6±0.0

-2.0±0.0 B

-2.0±0.0

-1.6±0.0

-1.6±0.0

B

B

B

100
1

-1.6±0.0 B

5

5.6±0.4 A

5.6±0.5 A 5.7±0.3 A 5.7±0.4 A

10
a

6.0±0.1 A 5.8±0.2 A 5.9±0.2 A 5.7±0.3 A 5.7±0.4 A
5.9±0.2 A 5.8±0.1 A 5.9±0.1 A 5.8±0.2 A 5.7±0.3 A

5.9±0.1 A

5.9±0.2 A 5.8±0.2 A 5.8±0.2 A

Means with different capital letters designate E. coli O157:H7 populations that differ significantly in terms of each sanitizer

concentration and time (P ≤ 0.05).

195

AI.4.2 Physicochemical parameters of wash water. Increasing organic load correlated
to significant increases (P ≤ 0.05) in COD (10 – 5543 mg O2/L), turbidity (0 – 0.293), and total
solids (0.0064 – 0.0471 g/10 ml) in the water control (Table AI.5). Increasing organic load
correlated to significant increases (P ≤ 0.05) in COD (66 – 6020 mg O2/L), turbidity (0 – 0.421),
ORP (282 – 539 mV), and total solids (0.0057 – 0.0471 g/10 ml) for all three peroxyacetic acid
treatments (Table AI.6). Increasing organic load correlated to significant increases (P ≤ 0.05) in
COD (78 – 6267 mg O2/L), turbidity (0 – 0.431), ORP (431 – 590 mV), and total solids (0.0048
– 0.0505 g/10 ml) for all three mixed peracid treatments (Table AI.7). Increasing organic load
correlated to significant increases (P ≤ 0.05) in COD (1 – 5937 mg O2/L), turbidity (0 – 0.289),
and total solids (0.0053 – 0.0492 g/10 ml) for all three chlorine treatments (Table AI.8).

196

Table AI.5: Physicochemical parameters of wash water controls containing various organic loads

a

Mean ± SD Dependent Physicochemical Paramters
% Organic
Load (w/v
blended lettuce)
0

COD (mg
O2/L)
10±3 D

MFV
(ml)
50±0 A

Turbidity (abs.
@ 663 nm)
0±0 C

ORP (mV)
pH
393±21 A 6.82±0.03 A

1

451±6 C

37±2 B

0.040±0.013 C

301±23 B

5

2330±159 B

9±2 C

0.150±0.021 B

10

5543±205 A

6±1 C

0.293±0.046 A

a

Temperature
(°C)
14±1 A

Total Solids (g/10
ml)
0.0064±0.0007 D

6.87±0.04 A

14±1 A

0.0132±0.0041 C

238±2 C

7.05±0.06 A

15±0 A

0.0267±0.0025 B

206±14 C

6.90±0.16 A

15±1 A

0.0471±0.0014 A

Means labeled with different capital letters designate physicochemical parameters that differ significantly in terms of organic load

(P ≤ 0.05).

197

Table AI.6: Physicochemical parameters of wash water containing various organic loads and peroxyacetic acid concentrations

a

Mean ± SD Dependent Physicochemical Parameters

COD (mg
O2/L)
66±13 C

MFV
(ml)
50±0 A

1

697±49 C

35±5 B 0.034±0.009 BC

5

3243±352 B

14±3 C

10

5317±1043 A

0

POAA
(ppm)

% Organic
Load (w/v
blended
lettuce)
0

Turbidity (abs.
@ 663 nm)
0±0 C

Temperature
pH
(°C)
6.00±1.59 A
20±4 A

Total Solids
(g/10 ml)
0.0057±0.0012 C

373±2 A

6.73±0.07 A

14±1 A

0.0103±0.0009 C

0.119±0.012 B

282±46 B

6.54±0.10 A

16±1 A

0.0291±0.0025 B

3±2 D

0.347±0.075 A

376±17 A

6.29±0.11 A

16±1 A

0.0449±0.0069 A

231±28 C

50±0 A

0±0 C

409±47 B

5.57±1.51 A

19±5 A

0.0060±0.0005 C

1

885±52 C

36±6 B

0.040±0.005 C

429±2 B

6.31±0.04 A

14±0 A

0.0109±0.0008 C

5

3477±337 B

15±3 C

0.146±0.026 B

451±35 AB 6.29±0.17 A

15±1 A

0.0280±0.0023 B

10

5820±870 A

5±1 D

0.370±0.069 A

15±1 A

0.0471±0.0066 A

ORP (mV)
410±30 A

10

50

198

531±50 A

5.78±0.04 A

Table AI.6 (cont’d)
0

372±29 C

50±0 A

0±0 C

435±8 C

6.17±0.11 A

14±1 B

0.0066±0.0008 C

1

959±65 C

41±6 A 0.039±0.006 BC

450±5 BC

5.92±0.09 A

14±0 B

0.0105±0.0010 C

5

3827±679 B

13±4 B

0.146±0.022 B

464±4 B

5.88±0.16 A

15±1 AB

0.0269±0.0019 B

10

6020±1054 A

6±1 B

0.421±0.092 A

539±10 A

5.44±0.14 B

16±1 A

0.0457±0.0077 A

80

a

Means labeled with different capital letters designate physicochemical parameters that differ significantly in terms of organic load at

each sanitizer concentration (P ≤ 0.05).

199

Table AI.7: Physicochemical parameters of wash water containing various organic loads and mixed peracid concentrations

a

Mean ± SD Dependent Physicochemical Paramters

COD (mg
O2/L)
78±15 B

MFV Turbidity (abs.
(ml)
@ 663 nm)
50±0 A
0±0 C

ORP (mV)
443±12 A

1

636±54 B

31±6 B 0.033±0.007 C

431±5 A

6.71±0.15 A

14±0 BC

0.0100±0.0017 C

5

6863±3386 A 10±2 C 0.154±0.065 B

433±28 A

6.87±0.09 A

15±0 AB

0.0254±0.0013 B

10

5233±445 A

5±0 C

0.276±0.032 A

356±30 B

6.77±0.04 A

15±0 A

0.0414±0.0018 A

0

359±9 B

50±0 A

0±0 B

482±7 C

6.36±0.06 A

13±0 C

0.0053±0.0009 D

1

Mixed
POAA
(ppm)

% Organic
Load (w/v
blended
lettuce)
0

1039±217 B

24±7 B 0.048±0.017 B

483±8 BC

6.41±0.10 A

14±0 BC

0.0105±0.0006 C

5

6070±3403 A

8±1 C

0.144±0.014 B 564±51 AB 6.25±0.01 A

14±1 AB

0.0266±0.0009 B

10

5880±695 A

5±1 C

0.355±0.113 A

15±1 A

0.0461±0.0023 A

Temperature
Total Solids
pH
(°C)
(g/10 ml)
6.68±0.18 A
14±1 C
0.0048±0.0003 D

10

50

200

570±35 A

5.98±0.13 B

Table AI.7 (cont’d)
0

557±10 C

50±0 A

1

1088±94 C

5
10

0±0 C

475±4 B

5.96±0.11 A

13±0 C

0.0051±0.0010 D

25±7 B 0.044±0.014 C

505±5 B

5.96±0.03 A

14±0 BC

0.0117±0.0015 C

3807±890 B

8±0 C

0.171±0.019 B

558±34 A

5.72±0.08 A

14±0 B

0.0259±0.0004 B

6267±521 A

5±1 C

0.431±0.077 A

590±18 A

5.33±0.23 B

15±1 A

0.0505±0.0046 A

80

a

Means labeled with different capital letters designate physicochemical parameters that differ significantly in terms of organic load at

each sanitizer concentration (P ≤ 0.05).

201

Table AI.8: Physicochemical parameters of wash water containing various organic loads and chlorine concentrations

a

Mean ± SD Dependent Physicochemical Paramters
% Organic
Load (w/v
blended
lettuce)
0

COD (mg
O2/L)
1±1 C

MFV
(ml)
50±0 C

Turbidity (abs.
@ 663 nm)
0±0 C

ORP
(mV)
724±8 A

pH
7.54±0.04 A

1

617±30 C

31±13 C

0.048±0.037 C

437±44 B

7.40±0.08 A

15±0 A

0.0113±0.0006 C

5

3260±1474 B

9±2 B

0.181±0.054 B

403±35 B

7.16±0.13 A

15±1 A

0.0275±0.0027 B

10

5937±900 A

6±1 A

0.283±0.009 A

361±12 B

7.24±0.32 A

15±A A

0.0492±0.0017 A

0

8±8 B

50±0 A

0±0 C

743±30 A

7.68±0.09 A

15±0 A

0.0054±0.0005 D

1

Available
Chlorine
(ppm)

514±85 B

34±7 B

0.030±0.003 C

758 ±42 A 7.57±0.10 A

15±0 A

0.0109±0.0009 C

5

3517±1737 A

10±1 C

0.132±0.016 B

439±3 B

7.28±0.05 B

15±0 A

0.0264±0.0008 B

10

5727±1141 A

7±2 C

0.289±0.052 A

363±13 C

7.10±0.06 B

16±1 A

0.0492±0.0031 A

Temperature
Total Solids
(°C)
(g/10 ml)
14±0 A
0.0053±0.0006 D

30

50

202

Table AI.8 (cont’d)
0

3±3 C

50±0 A

0±0 C

761±16 A

7.84±0.09 A

15±1 A

0.065±0.0005 C

1

649±69 C

20±4 B

O.034±0.005 C

788±16 A

7.91±0.16 A

14±1 A

0.0158±0.0047 C

5

3330±734 B

10±1 C

0.124±0.037 B

454±12 B

7.38±0.15 B

15±1 A

0.0302±0.0027 B

10

5880±630 A

6±1 C

0.281±0.039 A

441±27 B

7.18±0.09 B

16±1 A

0.0454±0.0060 A

100

a

Means labeled with different capital letters designate physicochemical parameters that differ significantly in terms of organic load at

each sanitizer concentration (P ≤ 0.05).

203

AI.5 SUMMARY OF FINDINGS
The two aims of this study were to 1) determine the efficacy of three concentrations of
three different sanitizing agents against E. coli O157:H7 in water containing organic loads of 0,
1, 5 and 10% (w/v blended lettuce solids) and 2) assess various physicochemical parameters of
the wash water. This study was originally meant to serve as a model for the pilot-scale
processing line used in Chapters 2, 4 and 5, however, comparisons of these results with the
results from the processing line found that the methods presented here do not adequately predict
the results from the processing line. Additionally, this study did not assess persistence of E. coli
O157:H7 in wash water, simply the immediate effect of the sanitizing agents during a 90 s
exposure period. The work presented here served as a framework for the study presented in
Chapter 3, where lettuce was used to inoculate the wash water in the carboy and E. coli
persistence was assessed over 10 min. Differences in results seen between the pilot-scale
processing line, the carboy system in Chapter 3, and the carboy system here illustrate that the
methods of exposing E. coli O157:H7 to the wash water in addition to the scale of the system
used for the study are critical factors for consistency between various systems.
Increases in organic load correlated to significant increases (P ≤ 0.05) in COD, turbidity,
and total solids for all sanitizers and sanitizer concentrations. Future work can be done in order
to determine the significance of different inoculation methods (e.g., lettuce vs. cocktail) and
different systems (e.g., centrifuge tubes, 4 L carboy, and a pilot-scale leafy greens processing
line).

204

Appendix II:

Efficacy of Multiple Chlorine Concentrations Acidified with T-128 against Escherichia coli
O157:H7 during Pilot-Scale Processing of Iceberg Lettuce Using Wash Water Containing
an Organic Load

205

AII.1 ABSTRACT
Chemical sanitizers are routinely used during commercial flume washing of fresh-cut
leafy greens to minimize cross-contamination from the water. This study assessed the efficacy of
three chlorine concentrations acidified with T-128 against E. coli O157:H7 on iceberg lettuce, in
wash water, and on surfaces of a pilot-scale processing line using flume water containing an
organic load. Methods were similar to those in Chapter 4 except that a single organic load (5%
w/v blended lettuce) and three different chlorine concentrations (10, 30, and 100 ppm), all
acidified to pH 6.0 with T-128 were used. Increasing chlorine concentrations correlated to
significantly greater E. coli O157:H7 reductions (P ≤ 0.05) in wash water and on lettuce. A
chlorine concentration of 100 ppm prevented the persistence of E. coli O157:H7 on equipment
surfaces and in water spun off during centrifugal drying. ORP correlated with increased chlorine
concentration and antimicrobial activity.

206

AII.2 INTRODUCTION
Wash water quality has been a major focus of the produce industry for many years with
these concern heightened in response to three nationwide outbreaks in 2006 traced to fresh-cut
lettuce and baby spinach. Despite the widespread use of chemical sanitizers in produce wash
water, efficacy of these sanitizers remains problematic due to the presence of organic material in
the water that decreases antibacterial activity. Weak acids, such as citric acid, are routinely added
to chlorinated wash water to acidify to a targeted pH 6.5 – 7.0, where the active component,
HClO predominates. A new, generally recognized as safe (GRAS) acidifying agent comprised of
phosphoric acid and propylene glycol, known as T-128 (SmartWash Solutions, Salinas, CA), has
been developed to improve the stability of chlorine. T-128 was examined in Chapters 2 – 4;
however, only one chlorine concentration was used for each study.
The two aims of this study were to 1) determine the efficacy of three different chlorine
concentrations acidified with T-128 (SmartWash Solutions) against E. coli O157:H7 during
pilot-scale processing of iceberg lettuce using water containing an organic load of 5% (w/v)
blended lettuce solids and 2) assess the relationship between various physicochemical parameters
and chlorine concentration of the wash water on E. coli O157:H7 inactivation.

207

AII.3 MATERIALS AND METHODS
AII.3.1 Experimental design. The efficacy of three chlorine concentrations acidified to
pH 6.0 using T-128 against E. coli O157:H7 was assessed in triplicate by processing a 5.4 kg
6

batch of iceberg lettuce inoculated at 10 CFU/g followed by three consecutive 5.4 kg batches of
uninoculated iceberg lettuce in flume water containing an organic load of 5% (w/v) blended
lettuce solids. All lettuce was processed by shredding, conveying, fluming, shaker table
dewatering and/or centrifugal drying, during and/or after which various product, water and
equipment surface samples were collected and quantitatively examined for E. coli O157:H7. E.
coli O157:H7 persistence was correlated to six physicochemical parameters of the wash water:
temperature, oxidation/reduction potential, chemical oxygen demand, total solids, maximum
filterable volume and turbidity.
AII.3.2 Leafy greens. Identical to 2.3.2.
AII.3.3 Bacterial strains. Identical to 2.3.3.
AII.3.4 Lettuce inoculation. Identical to 4.3.4.
AII.3.5 Lettuce processing line. Identical to 4.3.5.
AII.3.6 Wash water. Iceberg lettuce (45 kg) was blended for 10 min in tap water using a
Rotostat blender (Model XP-02, Admix, Manchester, NH) and added to the recirculation tank
where water was added to achieve an organic load level of 5% (w/v) in a total volume of 890 L
at 12 to 15oC. Three chlorine concentrations were used – 10, 30 and 100 ppm available chlorine
(XY-12, Ecolab, St. Paul, MN) which were acidified with T-128 (SmartWash Solutions) to a pH
of 6.0 as measured with a pH probe (pHTestr 30, Oakton, Vernon Hills, IL). The chlorine
concentration and pH was maintained throughout processing, with necessary adjustments made

208

during the 10 min pauses in lettuce processing. Chlorine test kit 321 (Ecolab) was used to
confirm the available chlorine concentration. Sanitizer-free MSU tap water served as the control.
AII.3.7 Leafy green processing. Identical to 4.3.7.
AII.3.8 Sample collection. Identical to 4.3.8.
AII.3.9 Physiochemical parameters of flume water. Identical to 4.3.9.
AII.3.10 Microbiological analyses. Identical to 4.3.10.
AII.3.11 Sanitizer neutralization confirmation. Chlorine concentrations of 10, 30 and
100 ppm (XY-12, Ecolab) acidified to pH 6.0 using T-128 (SmartWash Solutions) were prepared
and confirmed with Chlorine test kit 321 (Ecolab) in triplicate 1 L water samples containing a
5% organic load prepared a mechanical blender (Model BLC10650MB, Black & Decker, New
Britain, CT). A 50 ml centrifuge tube containing 3 ml of 38x concentrated neutralizing buffer
(BD) was filled with the water sample containing sanitizer, agitated for 5 sec and then
immediately assessed for neutralization of the sanitizer as previously described using the
Chlorine test kit. Preliminary experiments found that a 38x concentration would neutralize
various concentrations of the active component of each sanitizing agent used in this study
without impacting E. coli O157:H7 counts.
AII.3.12 Statistical analysis. Identical to 4.3.13.

209

AII.4 RESULTS
AII.4.1 Lettuce. Average E. coli O157:H7 reductions on inoculated lettuce were 1.2, 1.1
and 1.3 log CFU/g as seen with 10, 30, and 100 ppm available chlorine acidified to pH 6.0 with
T-128, respectively (Figure AII.1). None of the chlorine concentrations were significantly more
effective (P > 0.05) against E. coli O157:H7 on the first uninoculated batch of lettuce, with
populations ranging from 2.2 – 2.9 log CFU/g. One hundred ppm was significantly more
effective (P ≤ 0.05) than 10 and 30 ppm chlorine against E. coli O157:H7on uninoculated lettuce
from batch 2, with populations of 2.3, 2.0, and 0.9 log CFU/g recovered, respectively. Similarly,
100 ppm was also significantly more effective than 10 ppm chlorine against E. coli O157:H7 on
uninoculated lettuce batch 3, with populations of 2.1 and 0.1 log CFU/g recovered, respectively.

210

7

E. coli O157:H7 (log CFU/g)

6

A A

10 ppm

A

30 ppm
A A A

5

100 ppm
A

4

A A

3

A

A
A

2

B

1

B

AB

0
-1

InoculatedPrewash

InoculatedPostwash

1

-2

2

3

Uninoculated Batches

Figure AII.1: Mean (± SD) E. coli O157:H7 populations on lettuce after flume washing in water
containing 5% organic load (w/v blended lettuce) and chlorine concentrations of 10, 30 and 100
ppm + T-128 (n=3 per chlorine concentration). Means within the same batch with different
letters are significantly different (P ≤ 0.05).

211

AII.4.2 Flume water. Average E. coli O157:H7 reductions typically increased with
increasing chlorine concentration in the flume water. At the start of water sampling when the
inoculated lettuce was first removed from the flume system (0 min), average E. coli O157:H7
reductions with 10 (0.3 log CFU/ml) and 30 ppm chlorine (0.2 log CFU/ml) were significantly
lower (P ≤ 0.05) than those seen with 100 ppm (5.3 log CFU/ml) (Table AII.1). After 21 min of
processing, the average E. coli O157:H7 reduction with 10 ppm (0.5 log CFU/ml) was
significantly lower (P ≤ 0.05) than those seen for 30 (3.1 log CFU/ml) or 100 ppm chlorine (5.3
log CFU/ml). By the end of processing (44 min), average E. coli O157:H7 reductions with 10
ppm chlorine (0.8 log CFU/ml) were significantly lower (P ≤ 0.05) than for 30 (4.4 log CFU/ml)
and 100 ppm (5.3 log CFU/ml).
E. coli O157:H7 populations were typically below the limit of detection in flume water
containing 100 ppm available chlorine, resulting in a y-intercept of -5.3 andno slope over time
2

with an R value of 0.12 and a RMSE of 0.03 CFU/ml) (Table AII.2, Figure AII.2). Trend lines
for E. coli O157:H7 inactivation in flume water containing 10 and 30 ppm had similar yintercepts of -0.3 and -0.3 log CFU/ml, however, the slope of the inactivation equation for 30
2

ppm (-0.1t) was 10x greater than the slope seen at 10 ppm chlorine (-0.01t). The R values for 10
(0.92) and 30 ppm (0.93) show that E. coli O157:H7 inactivation is strongly correlated with
chlorine concentration. The RMSE for 10 ppm was 0.2 log CFU/ml, with a higher RMSE seen
for 30 ppm (1.4 log CFU/ml).

212

Table AII.1: E. coli O157:H7 reductions in flume water

a

Mean ± SD E. coli O157:H7 reduction in flume water (log
CFU/ml)
Available Chlorine
(ppm)

21 min

44 min

10

0.3 ± 0.2 B

0.5 ± 01 C

0.8 ± 0.1 B

30

0.2 ± 0.1 B

3.1 ± 1.2 B

4.4 ± 0.8 A

100
a

0 min

5.3 ± 0.1 A

5.3 ± 0.0 A

5.3 ± 0.0 A

Means with different capital letters designate E. coli O157:H7 reductions that differ

significantly in terms of sanitizing treatment and time (P ≤ 0.05).

213

Table AII.2: E. coli O157:H7 inactivation trend lines for flume water containing various
chlorine concentrations

a

E. coli O157:H7 Inactivation
Available Chlorine
(ppm)

2

10

R
Log Reduction
RMSE (log CFU/ml) P Value
Log(N/N0) = -0.01t - 0.3 0.92
0.2
1.38E-13

30

Log(N/N0) = -0.10t - 0.3 0.93

100
a

Log(N/N0) = -5.3

b

0.12

6.08E-14

0.0

0.09

Inactivation rates were determined by using Microsoft Excel to create linear regression trend

lines for E. coli O157:H7 log reduction over time (t) in minutes.
b

1.4

The slope (k) was < 0.1 and was therefore not included.

214

0
-1

Log (N/N0)

-2
10 ppm
30 ppm
100 ppm
Linear (10 ppm)
Linear (30 ppm)
Linear (100 ppm)

-3
-4
-5
-6
0

5

10

15

20
25
Time (min)

30

35

40

45

Figure AII.2: E. coli O157:H7 reductions in recirculating flume water containing 5% organic
load (w/v blended lettuce) and chlorine concentrations of 10, 30 and 100 ppm + T-128
compared to the sanitizer-free control (n=3 per chlorine concentration).

215

AII.4.3 Centrifugation water. A chlorine concentrations of 100 ppm resulted in E. coli
O157:H7 populations that were always significantly lower (P ≤ 0.05) than those recovered at 10
or 30 ppm, with -1.3 log CFU/ml seen for the inoculated batch as well as batches 1 – 3 (Figure
AII.3). Numbers of E. coli O157:H7 recovered from water spun off from the inoculated batch
(4.6 log CFU/ml) and first uninoculated batch of lettuce (3.6 log CFU/ml) after washing in 10
ppm were not significantly different (P > 0.05) from populations recovered after washing in 30
ppm chlorine. E. coli O157:H7 populations recovered from centrifugation water removed from
uninoculated batches 2 (2.1 log CFU/ml) and 3 (1.6 log CFU/ml) after product was washed in 30
ppm chlorine were significantly lower (P ≤ 0.05) than recovered with 10 ppm.

216

E. coli O157:H7 (log CFU/ml)

6
5

A

10 ppm
30 ppm
100 ppm

A
A

4

A

A

A

3
B
B

2
1
0
-1

B

B

C

C

-2

Inoculated

1

2

3

Uninoculated Batches
Figure AII.3: Mean (± SD) E. coli O157:H7 populations in spent centrifugation water from
lettuce after washing in flume water containing 5% organic load (w/v blended lettuce) and
chlorine concentrations of 10, 30 and 100 ppm + T-128 (n=3 per chlorine concentration).
Means within the same batch with different letters are significantly different (P ≤ 0.05). Limit of
detection = 0.02 CFU/ml.

217

AII.4.4 Processing equipment surfaces. E. coli O157:H7 populations recovered from
the flume tank, shaker table, and dewatering centrifuge after processing with 10 ppm chlorine
2

were 2.7, 2.6 and 2.5 log CFU/100 cm , respectively, and were significantly higher (P ≤ 0.05)
than those seen using 30 or 100 ppm chlorine (Figure AII.4). Washing with 100 ppm chlorine
resulted in E. coli populations that were consistently at or below the limit of detection for all
2

equipment surfaces, resulting in -0.1 log CFU/100 cm for all surfaces. E. coli O157:H7
populations recovered after washing with 30 ppm chlorine were significantly lower (P ≤ 0.05)
than those recovered after processing with 10 ppm, resulting in populations of 0.2, 0.6, and 1.6
2

log CFU/100 cm for the flume tank, shaker table, and dewatering centrifuge, respectively.

218

Figure AII.4: Mean (± SD), E. coli O157:H7 populations on equipment surfaces after processing
lettuce in flume water containing 5% organic load (w/v blended lettuce) and chlorine
concentrations of 10, 30 and 100 ppm + T-128 (n=3 per chlorine concentration). Means of the
same equipment with different letters are significantly different (P ≤ 0.05). Limit of detection = 1
2

CFU/100 cm .
3
E. coli O157:H7 (log CFU/100 cm2)

A

10 ppm
30 ppm
100 ppm

A

A

2.5
2

B

1.5
B
1
0.5
0

B
B

C

Flume Tank

Shaker Table
Equipment

C

-0.5

219

Dewatering Centrifuge

AII.4.5 Physicochemical parameters of flume water. Increasing sanitizer concentration
correlated to significant increases (P ≤ 0.05) in ORP, with 416, 477 and 855 mV recorded for
chlorine concentrations of 10, 30 and 100 ppm, respectively (Table AII.3). Considering that the
organic load was 5% (w/v) blended lettuce for all three chlorine treatments, it is not surprising
that no significant difference was seen in total solids, COD or turbidity between any of the
treatments. Total solids ranged from 0.0289 to 0.0336 g/10 ml, COD ranged from 2870 to 3374
mg O2/L, and turbidity ranged from 0.043 to 0.056. MFV was significantly lower (P ≤ 0.05)
with 100 ppm (12 ml) than with 10 ppm (43 ml) or 30 ml (42 ml).

220

Table AII.3: Physicochemical parameters of flume water containing organic load and various chlorine concentrations acidified with
T-128

a

Independent
Physicochemical Parameters

Available Chlorine (ppm)
10

Mean ± SD Dependent Physicochemical Parameters
Turbidity (abs @
MFV
Temp (°C)
COD (mg O2/L)
Total Solids (g)
663 nm)
(ml)
14 ± 0 A 0.0289 ± 0.0016 A 3374 ± 414 A
0.046 ± 0.003 A
43 ± 1 A

ORP (mV)
416 ± 13 C

30

0.0326 ± 0.0035 A

3235 ± 1197 A

0.043 ± 0.011 A

42 ± 8 A

477 ± 17 B

100
a

14 ± 0 A
14 ± 0 A

0.0336 ± 0.0019 A

2870 ± 676 A

0.056 ± 0.006 A

12 ± 0 B

855 ± 15 A

Means labeled with different capital letters designate physicochemical parameters that differ significantly in terms of chlorine

concentration (P ≤ 0.05).

221

AII.5 SUMMARY OF FINDINGS
The two aims of this study were to 1) determine the efficacy of three different chlorine
concentrations acidified with T-128 (SmartWash Solutions) against E. coli O157:H7 during
pilot-scale processing of iceberg lettuce using water containing an organic load of 5% (w/v)
blended lettuce solids and 2) assess the relationship between various physicochemical parameters
and chlorine concentration of the wash water on E. coli O157:H7 inactivation. Previous work
presented in Chapter 4 of this dissertation focused on the impact of various organic load
concentrations (0 – 10% w/v blended lettuce) using a single concentration of chlorine (50 ppm)
either unacidified or acidified with citric acid or T-128. This study was completed in
collaboration with a research group from the University of South Florida, with their study
focusing on determining the efficacy of a water concentrator at quantifying E. coli O157:H7
populations in wash water containing an organic load.
Not surprisingly, E. coli O157:H7 persistence in wash water and on finished product was
dependent on sanitizer concentration of the wash water with significantly fewer (P ≤ 0.05) E.
coli O157:H7 cells recovered from uninoculated batches 2 and 3 when 100 ppm available
chlorine was present in the wash water. Chapter 4 of this dissertation showed that a concentration
of 50 ppm chlorine was significantly impacted by an organic load of 5%. In this study, 100 ppm
free chlorine reduced E. coli O157:H7 populations below the limit of detection in the flume
water. ORP proved to be the best predictor of sanitizer efficacy.
Based on these findings, processors should consider acidifying their wash water using T128 to pH 6.0 when using 100 ppm chlorine. This sanitizing treatment proved to be effective
against E. coli O157:H7 on the product, equipment surfaces, and in wash water. However, proper
ventilation will be needed due to off-gassing of the chlorine.
222

Appendix III:

Impact of In-Line Equipment Sanitation Processes on Escherichia coli O157:H7 Persistence
during Pilot-Scale Processing of Iceberg Lettuce Using Flume Water Containing Sodium
Hypochlorite and Organic Load

223

AIII.1 ABSTRACT
Chemical sanitizers are routinely used during commercial flume washing of fresh-cut
leafy greens to minimize cross-contamination from the water. This study assessed the impact of
sanitizing practices on E. coli O157:H7 persistence on finished product, in wash water, and on
equipment using flume water containing 50 ppm chlorine. Methods were similar to those in
Chapter 4 except that a single organic load (2.5% w/v blended lettuce) and chlorine
concentration (50 ppm) were used. By sanitizing the shredder and conveyor after processing the
inoculated lettuce, significantly greater E. coli O157:H7 reductions (P ≤ 0.05) were seen in the
wash water and on equipment surfaces. However, no significant differences (P > 0.05) in E. coli
O157:H7 populations were observed on lettuce after processing between either operational
condition. Physicochemical parameters of the wash water were rarely significantly different
between either operating condition.

224

AIII.2 INTRODUCTION
Wash water quality has been a major focus of the produce industry for many years with
these concern heightened in response to three nationwide outbreaks in 2006 traced to fresh-cut
lettuce and baby spinach. Despite the widespread use of chemical sanitizers in produce wash
water, efficacy of these sanitizers remains problematic due to the presence of organic material in
the water that decreases antibacterial activity. While the focus of Chapter 4 was the impact of
sanitizing deficiencies on E. coli O157:H7 in flume water, it was noted that E. coli O157:H7
persisted on all lettuce after processing, even if the pathogen was below the lower limit of
detection in the flume water.
Consequently, the two aims of this study were to 1) determine the persistence of E. coli
O157:H7 during pilot-scale processing of iceberg lettuce using a shredder and conveyor that had
been sanitized after processing the inoculated product and before the uninoculated product was
then washed in flume water containing 50 ppm available chlorine and an organic load of 2.5%
(w/v blended lettuce solids) and 2) assess the relationship between various physicochemical
parameters and chlorine concentration of the wash water on E. coli O157:H7 inactivation.

225

AIII.3 MATERIALS AND METHODS
AIII.3.1 Experimental design. A 5.4 kg batch of iceberg lettuce inoculated with E. coli
6

O157:H7 at 10 CFU/g was initially processed, after which the shredder and conveyor were
sanitized before processing three additional 5.4 kg batches of uninoculated lettuce, thereby
leaving the flume water and surfaces of the shaker table as only means to contaminate the
product. All lettuce was processed by shredding, conveying, fluming in water containing 50 ppm
chlorine and an organic load of 2.5% (w/v) blended lettuce solids, shaker table dewatering and/or
centrifugal drying. During and/or after processing various product, water and equipment surface
samples were collected and quantitatively examined for E. coli O157:H7. E. coli O157:H7
persistence was correlated to six physicochemical parameters of the wash water: temperature,
oxidation/reduction potential, chemical oxygen demand, total solids, maximum filterable volume
and turbidity. A water control (containing 2.5% organic load and no chlorine) was processed
without sanitizing the shredder or conveyor.
AIII.3.2 Leafy greens. Identical to 2.3.2.
AIII.3.3 Bacterial strains. Identical to 2.3.3.
AIII.3.4 Lettuce inoculation. Identical to 4.3.4.
AIII.3.5 Lettuce processing line. Identical to 4.3.5.
AIII.3.6 Wash water. Iceberg lettuce (22.5 kg) was blended for 10 min in tap water
using a Rotostat blender (Model XP-02, Admix, Manchester, NH) and added to the recirculation
tank where water was added to achieve an organic load level of 2.5% (w/v) in a total volume of
890 L at 12 to 15oC. A chlorine concentration of 50 ppm available chlorine (XY-12, Ecolab, St.
Paul, MN) was maintained throughout processing, with necessary adjustments made during the

226

10 min pauses in lettuce processing. Chlorine test kit 321 (Ecolab) was used to confirm the
available chlorine concentration. Sanitizer-free MSU tap water served as the control.
AIII.3.7 Leafy green processing. To prime the system, 5.4 kg of uninoculated iceberg
lettuce was hand-fed into the shredder at a rate of about 0.5 kg per s, with the shredded product
then step-conveyed to the flume tank and washed in 890 L of recirculating sanitizing wash water
for 90 s. The product was then released from the flume tank and partially dewatered on the
shaker table, collected in a single centrifugation basket and centrifugally dried. Following a 10
min pause in processing to adjust the sanitizer concentration, the 5.4 kg batch of inoculated
lettuce was identically processed. The shredder and conveyor belt were then cleaned, sanitized,
and rinsed to remove inoculated product and residual inoculum from the equipment surfaces.
Three uninoculated 5.4 kg batches were then processed identically to the primer and inoculated
batches of lettuce.
AIII.3.8 Sample collection. Identical to 4.3.8.
AIII.3.9 Physiochemical parameters of flume water. Identical to 4.3.9.
AIII.3.10 Microbiological analyses. Identical to 4.3.10.
AIII.3.11 Sanitizer neutralization confirmation. Identical to 4.3.11.
AIII.3.12 Statistical analysis. Identical to 4.3.13.

227

AIII.4 RESULTS
AIII.4.1 Lettuce. E. coli O157:H7 persisted on all iceberg lettuce after washing and
centrifugation, regardless of operational conditions. Average E. coli O157:H7 reductions from
inoculated lettuce were 1.0, 0.9 and 1.1 log CFU/g as seen with the water control, 50 ppm
chlorine, and 50 ppm chlorine + clean shredder and conveyor (Figure AIII.1). No significant
difference (P > 0.05) in E. coli O157:H7 populations were seen after processing inoculated
lettuce or the first batch of uninoculated lettuce, with populations ranging from 2.2 – 2.8 log
CFU/g for uninoculated batch 1. By the second uninoculated batch, E. coli O157:H7 populations
on lettuce processed with a sanitized shredder and conveyor had significantly less (P ≤ 0.05) E.
coli O157:H7 than the water control, with 0.7 and 2.7 log CFU/g, respectively. E. coli O157:H7
populations on the third batch of uninoculated lettuce were significantly higher (P ≤ 0.05) on
lettuce washed with the water control (2.4 log CFU/g) than lettuce washed in chlorine (1.4 log
CFU/g) or chlorine + a clean shredder and conveyor (0.6 log CFU/g). Additionally, no
significant difference in E. coli O157:H7 populations recovered from batch 3 were seen for
lettuce processed with chlorine or chlorine + a clean shredder and conveyor.

228

Water (Control)

7
6

Chlorine

A A A

E. coli O157:H7 (log CFU/g)

A
5

A

Chlorine + Clean
Shredder & Conveyor

A

4

A
A

A

3

A
A
AB

2

B

B
B

1
0
-1

InoculatedPrewash

InoculatedPostwash

1

2

3

Batch

Figure AIII.1: Mean (± SD) E. coli O157:H7 populations on lettuce after flume washing in water
containing 2.5% organic load (w/v blended lettuce) and 50 ppm chlorine, with or without a
sanitized shredder and conveyor (n=3 per operational condition). Means within the same batch
with different letters are significantly different (P ≤ 0.05).

229

AIII.4.2 Flume water. Cleaning and sanitizing the shredder and conveyor reduced the
populations of E. coli O157:H7 in the flume water. As soon as the inoculated lettuce left the
flume tank (0 min), E. coli O157:H7 reductions in water with a clean shredder and conveyor
were significantly higher (P ≤ 0.05) (4.1 log CFU/ml) than seen with the normal processing
operation (-0.2 log CFU/ml) (Table AIII.1). About halfway through processing (21 min), E. coli
O157:H7 reductions were still significantly higher (P ≤ 0.05) using the previously cleaned and
sanitized shredder and conveyor (5.2 log CFU/ml) than seen with the normal processing
operation (2.3 log CFU/ml). By the end of processing (44 min), no significant difference (P >
0.05) in E. coli O157:H7 reductions in flume water were observed between the normal
processing (3.9 log CFU/ml) or the cleaned and sanitized shredder and conveyor (5.2 log
CFU/ml).
E. coli O157:H7 populations were typically much closer to the lower limit of detection in
flume water using the cleaned and sanitized shredder and conveyor, resulting in a y-intercept of 2

4.4, slope of -0.02t, R value of 0. 7, and RMSE of 0.2 log CFU/ml (Table AIII.2, Figure AIII.2).
In contrast, un-sanitized shredder and conveyor yielded a y-intercept of 0.1. While this y2

intercept was much higher, the slope was five times higher at -0.1t, with a resulting R value of
0.94 and a RMSE of 0.3 log CFU/ml. P values for both trend lines were < 0.0001.

230

Table AIII.1: E. coli O157:H7 reductions in flume water

Operational Conditions
Chlorine
Chlorine + Clean Shredder & Conveyor
a

a

Mean ± SD E. coli O157:H7 reduction in flume
water (log CFU/ml)
0 min
21 min
44 min
-0.2 ± 0.3 B
2.3 ± 0.6 B
3.9 ± 0.9 A
4.1 ± 1.0 A

5.2 ± 0.1 A

5.2 ± 0.1 A

Means with different capital letters designate E. coli O157:H7 reductions that differ

significantly in terms of operational condition and time (P ≤ 0.05).

231

Table AIII.2: E. coli O157:H7 inactivation trend lines for flume water after different processing
conditions

a

E. coli O157:H7 Inactivation
RMSE (log
CFU/ml)

P Value

0.94

0.3

< 0.0001

Log(N/N0) = -0.02t - 4.4 0.67

0.2

< 0.0001

Operational Conditions

Log Reduction

R

Chlorine

Log(N/N0) = -0.1t + 0.1

2

Chlorine + Clean Shredder &
Conveyor
a

Inactivation rates were determined by using Microsoft Excel to create linear regression trend

lines for E. coli O157:H7 log reduction over time (t) in minutes.

232

1

Chlorine

0

Chlorine + Clean Shredder &
Conveyor
Linear (Chlorine)

-1

Linear (Chlorine + Clean Shredder
& Conveyor)

Log (N/N0)

-2
-3
-4
-5
-6
0

5

10

15

20
25
Time (min)

30

35

40

45

Figure AIII.2: E. coli O157:H7 reductions in recirculating flume water containing 2.5% organic
load (w/v blended lettuce) and 50 ppm chlorine, with or without a sanitized shredder and
conveyor compared to the sanitizer-free control (n=3 per operational condition).

233

AIII.4.3 Centrifugation water. Average E. coli O157:H7 populations in spent
centrifugation from the water control were 4.7 log CFU/ml for the inoculated batch and 3.8 – 4.0
log CFU/ml for the three uninoculated batches (Figure AIII.3). Addition of chlorine resulted in
E. coli O157:H7 populations that were not significantly lower (P ≤ 0.05) than the water control
for the inoculated batch or the first uninoculated batch, with 4.3 and 3.3 log CFU/ml recovered,
respectively. E. coli O157:H7 populations in centrifugation water from uninoculated batches 2
(2.2 log CFU/ml) and 3 (1.8 log CFU/ml) after washing in chlorine were significantly lower than
the water control. E. coli O157:H7 populations recovered from water spun off after processing
with the cleaned and sanitized shredder and conveyor were consistently significantly lower (P ≤
0.05) than the water control from the inoculated batch of lettuce (3.3 log CFU/ml) and 1.4, 0.4
and 0.3 log CFU/ml for uninoculated batches 1, 2 and 3, respectively. Additionally, the E. coli
O157:H7 populations recovered from the centrifugation water from uninoculated batches 1 – 3
after processing with the cleaned and sanitized shredder and conveyor were significantly lower
than the populations recovered after processing under normal conditions using chlorine.

234

6

E. coli O157:H7 (log CFU/ml)

5
4

Water

Chlorine

Chlorine + Clean Shredder & Conveyor

A AB
B

A

A

A

A

3
B
B

2

B
C

1
C
0
Inoculated

1

2

3

-1
Uninoculated Batches
Figure AIII.3: Mean (± SD) E. coli O157:H7 populations in spent centrifugation water from
lettuce after washing in flume water containing 2.5% organic load (w/v blended lettuce) and 50
ppm chlorine, with or without a sanitized shredder and conveyor (n=3 per operational
condition). Means within the same batch with different letters are significantly different (P ≤
0.05). Limit of detection = 0.02 CFU/ml.

235

AIII.4.4 Processing equipment surfaces. E. coli O157:H7 populations recovered from
the surfaces of the flume tank, shaker table, and dewatering centrifuge after processing with
2

water were 3.4, 3.6 and 3.1 log CFU/100 cm , respectively, and were significantly higher (P ≤
0.05) than those seen on the flume tank and shaker table after processing with chlorine or
chlorine + a clean shredder and conveyor (Figure AIII.4). E. coli O157:H7 populations were
2

significantly higher (P ≤ 0.05) on the flume tank (0.2 log CFU/100 cm ) and shaker table (1.0
2

log CFU/100 cm ) after processing with chlorine than seen with chlorine + a cleaned and
2

sanitized shredder and conveyor (-1.3 and -0.6 log CFU/100 cm , respectively). Average E. coli
O157:H7 populations on the centrifugal dryer were significantly lower (P ≤ 0.05) than the water
control after processing with a cleaned and sanitized shredder and conveyor, with 1.1 log
2

CFU/100 cm , but statistically similar to the populations recovered after processing with
2

chlorine (2.3 log CFU/100 cm ).

236

Water

4

Chlorine

Chlorine + Clean Shredder & Conveyor
A

A
E. coli O157:H7 (log CFU/100 cm2)

3.5

A

3
AB
2.5
2
1.5

B

B

B

1
C

0.5
C

0
-0.5
Flume

Shaker

Centrifuge

Equipment
Figure AIII.4: Mean (± SD) E. coli O157:H7 populations on equipment surfaces after processing
lettuce in flume water containing 2.5% organic load (w/v blended lettuce) and 50 ppm chlorine,
with or without a sanitized shredder and conveyor (n=3 per operational condition). Means
within the same equipment with different letters are significantly different (P ≤ 0.05). Limit of
2

detection = 1 CFU/100 cm .

237

AIII.4.5 Physicochemical parameters of flume water. Significant differences (P ≤
0.05) between the physicochemical parameters of the flume water were rarely seen between the
two different operational conditions. Normal operation with chlorine resulted in significantly
lower (P ≤ 0.05) pH (8.38) and ORP (366 mV) than the pH (8.67) and ORP (464 mV) seen for
the cleaned and sanitized shredder and conveyor (Table AIII.3). There were no significant
differences (P > 0.05) between the other physiochemical parameters of the two operational
conditions with total solids ranging from 0.0175 – 0.0178 g/10 ml, COD ranging from 956 – 971
mg O2/L, turbidity ranging from 0.017 – 0.021, and MFV ranging from 38 – 43 ml.

238

Table AIII.3: Physicochemical parameters of flume water 2.5% containing organic load and 50 ppm chlorine during normal
operation and with a sanitized shredder and conveyor

a

Dependent Physicochemical Parameters
Operational
Conditions

MFV
(ml)

0.021±0.001

43±2

B

A

8.67±0.06

Chlorine

Turbidity (abs
@ 663 nm)

0.017±0.004

38±1

B

A

pH

Total Solids (g/10
ml)

COD (mg
O2/L)

B

Temp (°C)

8.38±0.06
13±0 A

0.0178±0.0007 A

971±67 A

ORP
(mV)

366±15 B

Chlorine + Clean
Shredder &

13±0 A

0.0175±0.0013 A
A

956±52 A

464±53 A

Conveyor
a

Means labeled with different capital letters designate physicochemical parameters that differ significantly in terms of operation

condition (P ≤ 0.05).

239

AIII.5 SUMMARY OF FINDINGS
The two aims of this study were to 1) determine the persistence of E. coli O157:H7
during pilot-scale processing of iceberg lettuce using a shredder and conveyor that has been
sanitized after processing the inoculated product and before the uninoculated product then
washed in flume water containing 50 ppm available chlorine and an organic load of 2.5% (w/v
blended lettuce solids) and 2) assess the relationship between various physicochemical
parameters and chlorine concentration of the wash water on E. coli O157:H7 inactivation.
Overall, physicochemical parameters of the flume water were rarely different between the two
operational conditions, emphasizing that those differences in E. coli O157:H7 populations in
wash water, on lettuce and equipment surfaces are attributable to the level of E. coli O157:H7
contamination on the product before washing. The studies in Chapters 4 and 5 assessed E. coli
O157:H7 persistence while processing under a variety of water conditions. This study was
completed in response to the persistence of E. coli O157:H7on all finished product throughout
processing in the earlier studies.
There were no significant differences in E. coli O157:H7 populations on the finished
product between the two operating conditions, emphasizing that cross-contamination of product
during processing is a systemic issue and cannot be pin-pointed to a single processing step.
However, by simulating a sanitizing step during shredding and conveying, it was noticed that
average E. coli O157:H7 populations in the wash water were consistently significantly lower in
the wash water when the shredder and conveyor were sanitized after processing the inoculated
lettuce, resulting in a much lower y-intercept for the predictive E. coli O157:H7 inactivation
trend line. Additional studies would need to be completed in order to determine if E. coli
O157:H7 cross-contamination could be reduced by processing additional lettuce after the
240

uninoculated lettuce. E. coli O157:H7 contamination of the three uninoculated batches of lettuce
processed after the inoculated batch likely occurred in the wash water and on the shaker tablewhere lettuce was seen to remain on equipment between batches.
Based on these results, commercial processors should consider including a sanitizing
spray rinse on equipment that leafy greens contact before washing to reducing the likelihood of
amplifying a previously isolated contamination event. Additional studies would need to be
completed in order to determine if an E. coli O157:H7 contamination event would be more
isolated if more uninoculated lettuce was processed than the 16.2 kg of lettuce in this study.

241

REFERENCES

242

REFERENCES

1.
Abadias, M., I. Alegre, J. Usall, R. Torres, and I. Vinas. 2011. Evaluation of alternative
sanitizers to chlorine disinfection for reducing foodborne pathogens in fresh-cut apple.
Postharvest Biol. Technol. 59:289-297.
2.
Ahmed, J., and T. Alam. 2011. Minimal processing and novel technologies applied to
vegetables. p. 317-333. In N.K. Sinha (ed.). Handbook of Vegetables and Vegetable Vegetable
Processing. Wiley-Blackwell, Ames, IA.
3.
Allende, A., E. Aguayo, and F. Artes. 2003. Microbial and sensory quality of commercial
fresh processed red lettuce throughout the production chain and shelf life. Int. J. Food.
Microbiol. 91:109-117.
4.
Allende, A., J. McEvoy, Y. Tao, and Y. Luo. 2009. Antimicrobial effect of acidified
sodium chlorite, sodium chlorite, sodium hypochlorite, and citric acid on Escherichia coli
O157:H7 and natural microflora of fresh-cut cilantro. Food Control. 20:230-234.
5.
Allende, A., M. V. Selma, F. Lopez-Galvez, R. Villaescusa, and M. I. Gil. 2008. Impact
of wash water quality on sensory and microbial quality, including Escherichia coli crosscontamination, of fresh-cut escarole. J. Food Prot. 71:2514-2518.
6.
Alvardo-Casillas, S., L. S. Ibarra-Sanchez, E. Martinez-Gonzalesn, M. O. RodriguezGarcia, and A. Castillo. 2010. Validation of a washing and sanitizing procedure for cantaloupes
at a Mexican packing facility. J. Food Prot. 73:362–365.
7.
Annous, B. A., P. M. Fratamico, and J. L. Smith. 2009. Quorum sensing in biofilms: why
bacteria behave the way they do. J. Food Sci. 74:R24-R37.
8.
Arnade, C., L. Calvin, and F. Kuchler. 2009. Consumer response to a food safety shock:
the 2006 food-borne illness outbreak of E.coli O157: H7 linked to spinach. Appl. Econ. Perspect.
Policy. 31:734-750.
9.
Artés, F., P. Gómez, F. Artés-Hernández, E. Aguayo, and V. Escalona. 2007. Improved
strategies for keeping overall quality of fresh-cut produce. Acta Hort. 746:245-258.
10.
Ayyildiz, O., B. Ileri, and S. Sanik. 2009. Impacts of water organic load on chlorine
dioxide disinfection efficacy. J. Hazard. Mater. 168:1092-1097.
11.
Baert, L., I. Vandekinderen, F. Devlieghere, E. V. Coillie, J. Debevere, and M.
Uyttendaele. 2009. Efficacy of sodium hypochlorite and peroxyacetic acid to reduce murine
norovirus 1, B40-8, Listeria monocytogenes, and Escherichia coli O157:H7 on shredded iceberg
lettuce and in residual wash water. J. Food Prot. 72:1047-1054.

243

12.
Baird, R. B. 2005. 5220 B. Open Reflux Method. In M.A.H. Franson (ed.). Standard
methods for the examination of water and wastewater. American Public Health Association,
Washington DC.
13.
Barrera, M. J., R. Blenkinsop, and K. Warriner. 2012. The effect of different processing
parameters on the efficacy of commercial post-harvest washing of minimally processed spinach
and shredded lettuce. Food Control. 25:745-751.
14.
Baxter, T. E., and D. J. Rexing. 2005. 2540 B. Total Solids Dried at 103-105°C. In
M.A.H. Franson (ed.). Standard methods for the examination of water and wastewater. American
Public Health Association, Washington DC.
15.
Bell, B. P., M. Goldoft, P. M. Griffin, M. A. Davis, D. C. Gordon, P. I. Tarr, C. A.
Bartleson, J. H. Lewis, T. J. Barrett, and J. G. Wells. 1994. A multistate outbreak of Escherichia
coli O157:H7-associated bloody diarrhea and hemolytic uremic syndrome from hamburgers. The
Washington experience. JAMA. 272:1349-1353.
16.
Beuchat, L. R. 1998. Surface decontamination of fruits and vegetables eaten raw: a
review. Available at:
http://www.who.int/foodsafety/publications/fs_management/en/surface_decon.pdf. Accessed 28
August 2012.
17.
Beuchat, L. R., B. B. Adler, and M. M. Lang. 2004. Efficacy of chlorine and a
peroxyacetic acid sanitizer in killing Listeria monocytogenes on iceberg and Romaine lettuce
using simulated commercial processing conditions. J. Food Prot. 67:1238-1242.
18.
Beuchat, L. R., D. A. Mann, and W. Q. Alali. 2013. Efficacy of sanitizers in reducing
Salmonella on pecan nutmeats during cracking and shelling. J. Food Prot. 76:770-778.
19.
Beuchat, L. R., and J. H. Ryu. 1997. Produce handling and processing practices. Emerg.
Infect. Dis. 3:459-465.
20.
Brooks, J. T., E. G. Sowers, J. G. Wells, K. D. Greene, P. M. Griffin, R. M. Hoekstra,
and N. A. Stockbine. 2005. Non-O157 Shiga toxin-producing Escherichia coli infections in the
United States, 1982-2002. Journal of Infectious Diseases. 192:1422-1429.
21.
Buchanan, R. L., S. G. Edelson, R. L. Miller, and G. M. Sapers. 1999. Contamination of
intact apples after immersion in an aqueous environment containing Escherichia coli O157:H7.
J. Food Prot. 62:444-450.
22.
Buchholz, A. L., G. R. Davidson, B. P. Marks, E. C. D. Todd, and E. T. Ryser. 2012.
Quantitative transfer of Escherichia coli O157:H7 to equipment during small-scale production of
fresh-cut leafy greens. J. Food Prot. 75:1184-1197.
23.
Buchholz, A. L., G. R. Davidson, B. P. Marks, E. C. D. Todd, and E. T. Ryser. 2012.
Transfer of Escherichia coli O157:H7 from equipment surfaces to fresh-cut leafy greens during
244

processing in a model pilot-plant production line with sanitizer-free water. J. Food Prot.
75:1920–1929.
24.
Buchholz, A. L., G. R. Davidson, R. E. Sloup, B. P. Marks, E. C. D. Todd, and E. T.
Ryser. 2010. Tracking an Escherichia coli O157:H7 contaminated batch of leafy greens through
a commercial processing line, Abst. T4-02. In, Abst. Ann. Mtg. Int. Assoc. Food Prot., Anaheim,
CA.
25.
Burnett, A. B., M. H. Iturriaga, E. F. Escartin, C. A. Pettigrew, and L. R. Beuchat. 2004.
Influence of variations in methodology on populations of Listeria monocytogenes recovered
from lettuce treated with sanitizers. J. Food Prot. 67:742–750.
26.
Burnett, S. L., and L. R. Beuchat. 2001. Human pathogens associated with raw produce
and unpasteurized juices, and difficulties in decontamination. J. Indust. Microbiol. Biotechnol.
27:104-110.
27.
Butt, M. S., and M. T. Sultan. 2011. Nutritional profile of vegetables and its significance
to human health. p. 107-123. In N.K. Sinha (ed.). Handbook of Vegetables and Vegetable
Processing. Wiley-Blackwell, Ames, IA.
28.
California Department of Public Health. 2007. Investigation of an Escherichia coli
O157:H7 outbreak associated with Dole pre-packaged spinach. Available at:
http://www.cdph.ca.gov/pubsforms/Documents/fdb%20eru%20Spnch%20EC%20Dole032007w
ph.PDF. Accessed 3 October 2012.
29.
California Leafy Green Products Handler Marketing Agreement. 2012. About us.
Available at: http://www.caleafygreens.ca.gov/about-us. Accessed 12 January 2013.
30.
Caprioli, A., S. Morabito, H. Brugere, and E. Oswald. 2005. Enterohaemorrhagic
Escherichia coli: emerging issues on virulence and modes of transmission. Vet. Res. 36:289-311.
31.
Carlin, F., and C. Nguyen-The. 1994. Fate of Listeria monocytogenes on four types of
minimally processed green salads. Lett. Appl. Microbiol. 18:222-226.
32.
Carmichael, I., I. S. Harper, M. J. Coventry, P. W. J. Taylor, J. Wan, and M. W. Hickey.
1999. Bacterial colonization and biofilm development on minimally processed vegetables. J.
Appl. Microbiol. 85:45S-51S.
33.
Center for Produce Safety. 2013. Grants program: Center for Produce Safety 2013
request for proposals. Available at: https://cps.ucdavis.edu/grant_opportunities.php. Accessed 6
February 2013.
34.
Center for Science in the Public Interest (CSPI). 2008. CSPI Outbreak Alert Data: Info on
Produce Outbreaks. Available at: http://www.cspinet.org/new/pdf/cspi_outbreak_alert.pdf.
Accessed 12 December 2012.

245

35.
Center for Science in the Public Interest (CSPI). 2008. Outbreak Alert! 2008. Available
at: http://www.cspinet.org/new/pdf/outbreak_alert_2008_report_final.pdf. Accessed 13 October
2012.
36.
Center for Science in the Public Interest (CSPI). 2009. The ten riskiest foods regulated by
the US Food and Drug Administration. Available at:
http://cspinet.org/new/pdf/cspi_top_10_fda.pdf. Accessed 2 October 2012.
37.
Centers for Disease Contol and Prevention. 1996. Outbreak of Escherichia coli O157:H7
infections associated with drinking unpasteurized commercial apple juice -- British Columbia,
California, Colorado, and Washington, October 1996. MMWR. 45:975.
38.
Centers for Disease Contol and Prevention. 2006. Ongoing multistate outbreak of
Escherichia coli serotype O157:H7 infections associated with consumption of fresh spinach --United States, September 2006. MMWR. 55:1-2.
39.
Centers for Disease Contol and Prevention. 2010. Investigation update: multistate
outbreak of human E. coli O145 infections linked to shredded Romaine lettuce from a single
processing facility. Available at: http://www.cdc.gov/ecoli/2010/ecoli_o145/. Accessed 12
January 2013.
40.
Centers for Disease Contol and Prevention. 2012. Investigation update: Multistate
outbreak of E. coli O157:H7 infections linked to Romaine lettuce. Available at:
http://www.cdc.gov/ecoli/2011/ecoliO157/romainelettuce/032312/index.html. Accessed 20
February 2013.
41.
Centers for Disease Contol and Prevention. 2012. Multistate outbreak of Shiga toxinproducing Escherichia coli O157:H7 infections linked to organic spinach and spring mix blend
(final update). Available at: http://www.cdc.gov/ecoli/2012/O157H7-11-12/. Accessed 12
January 2013.
42.
Chase-Topping, M. E., I. J. McKendrick, M. C. Pearce, P. MacDonald, L. Matthews, J.
Halliday, L. Allison, D. Fenlon, J. C. Low, G. Gunn, and M. E. Woolhouse. 2007. Risk factors
for the presence of high-level shedders of Escherichia coli O157 on Scottish farms. J. Clin.
Microbiol. 45:1594-603.
43.
Code of Federal Regulations. 2012. Title 21, Part 173.315. Chemicals used in washing or
to assist in the peeling of fruits and vegetables. Available at:
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/cfrsearch.cfm?fr=173.315. Accessed 16
September 2012.
44.
Cody, S. H., M. K. Glynn, J. A. Farrar, K. L. Cairns, P. M. Griffin, J. Kobayashi, M.
Fyfe, R. Hoffman, A. S. King, J. H. Lewis, B. Swaminathan, R. G. Bryant, and D. J. Vugia.
1999. An outbreak of Escherichia coli O157:H7 infection from unpasteurized commercial apple
juice. Annal. Intern. Med. 130:202-209.

246

45.
Cooley, M., D. Carychao, L. Crawford-Miksza, M. T. Jay, C. Myers, C. Rose, C. Keys, J.
Farrar, and R. E. Mandrell. 2007. Incidence and tracking of Escherichia coli O157:H7 in a major
produce production region in California. PLoS ONE. 2:e1159.
46.
Ecolab. 2010. Acid antimicrobial additive: Tsunami 100. Available at:
http://salustarim.com/uploads/Tsunami_orjinal_etiket.pdf. Accessed 15 October 2012.
47.
Foleya, D. M., A. Dufourb, L. Rodriguezb, F. Caporasob, and A. Prakash. 2002.
Reduction of Escherichia coli O157:H7 in shredded iceberg lettuce by chlorination and gamma
irradiation. Rad. Phys. Chem. 63:391–396.
48.
Food and Agriculture Organization and the World Health Organization. 2010. CODEX
Alimentarius: Code of hygenic practice for fresh fruits and vegetables. Available at:
http://www.codexalimentarius.org/standards/list-of-standards/en/. Accessed 20 February 2013.
49.
Foodborne Illness Outbreak Database. 2013. Aunt Mid's Produce Company lettuce 2008.
Available at: http://outbreakdatabase.com/details/aunt-mids-produce-company-lettuce2008/?organism=E.+coli+O157%3AH7&vehicle=lettuce. Accessed 12 January 2013.
50.
Foodborne Illness Outbreak Database. 2013. Church Brothers LLC/ Cafe Rio Restaurant/
Chipotle Restaurant Romaine lettuce 2009. Available at:
http://outbreakdatabase.com/details/church-brothers-llc-cafe-rio-restaurant-chipotle-restaurantromaine-lettuce-2009/?organism=E.+coli+O157%3AH7&vehicle=lettuce. Accessed 12 January
2013.
51.
Foodborne Illness Outbreak Database. 2013. E. coli O157:H7 outbreak linked to
Romaine lettuce - Canada and California, April 2012. Available at:
http://www.outbreakdatabase.com/details/e.-coli-o157h7-outbreak-linked-to-romaine-lettucecanada-and-california-april-2012/. Accessed 12 Janurary 2013.
52.
Foodborne Illness Outbreak Database. 2013. Multistate pre-packaged lettuce 2009.
Available at: http://outbreakdatabase.com/details/multistate-pre-packaged-lettuce2009/?organism=E.+coli+O157%3AH7&vehicle=lettuce. Accessed 12 January 2013.
53.
Foodborne Illness Outbreak Database. 2013. Multistate schnucks salad bars, college
campuses Romaine lettuce 2011. Available at: http://outbreakdatabase.com/details/multistateschnucks-salad-bars-college-campuses-romaine-lettuce2011/?organism=E.+coli+O157%3AH7&vehicle=lettuce. Accessed 12 Janurary 2013.
54.
Garg, N., J. J. Churney, and D. F. Splittstoesser. 1990. Effect of processing conditions on
the microflora of fresh-cut vegetables. J. Food Prot. 53:701-703.
55.
Gil, M. I., M. V. Selma, F. López-Gálvez, and A. Allende. 2009. Fresh-cut product
sanitation and wash water disinfection: Problems and solutions. Int. J. Food. Microbiol. 134:3745.

247

56.
Gomez-Lopez, V. M., J. Gobet, M. V. Selma, M. I. Gil, and A. Allende. 2013. Operating
conditions for the electrolytic disinfection of process wash water from the fresh-cut industry
contaminated with E. coli O157:H7. Food Control. 29:42-48.
57.
Gonzalez, R. J., Y. Luo, S. Ruiz-Cruz, and J. L. McEvoy. 2004. Efficacy of sanitizers to
inactivate Escherichia coli O157:H7 on fresh-cut carrot shreds under simulated process water
conditions. J. Food Prot. 67:2375–2380.
58.
Griffin, P. M., and R. V. Tauxe. 1991. The epidemiology of infections caused by
Escherichia coli O157:H7, other enterohemorrhagic E. coli and the associated hemolytic uremic
syndrome. Epidem. Rev. 13:60-98.
59.
Hach. 2009. Chemical oxygen demand, Method 8000 for water, wastewater, and
seawater. Available at: http://www.hach.com/quick.search-download.search.jsa?keywords=8000.
Accessed 10 November 2012.
60.
Harp, D. L. Date 2002. The current technology of chlorine analysis for water and
wastewater. Technical information series- booklet No. 17 [Type of Medium].
61.
Haute, S. V., I. Sampers, K. Holvoet, and M. Uyttendaele. 2013. Physicochemical quality
and chemical safety of chlorine as a reconditioning agent and wash water disinfectant for freshcut lettuce washing. Appl. Environ. Microbiol. 79:2850-2861.
62.
Herdt, J., and H. Feng. 2009. Aqueous antimicrobial treatments to improve fresh and
fresh-cut produce safety. p. 169-190. In B.A.N. Xuetong Fan, Christopher J. Doona, Florence E.
Feeherry, Robert B. Gravani (ed.). Microbial Safety of Fresh Produce. Wiley-Blackwell, Ames,
Iowa.
63.
Herman, K. M., T. L. Ayers, and M. Lynch. 2008. Foodborne disease outbreaks
associated with leafy greens, 1973–2006. Available at:
http://wwwnc.cdc.gov/eid/pdfs/ICEID2008.pdf. Accessed 10 October 2012.
64.
Hilgren, J., K. M. J. Swanson, F. Diez-Gonzalez, and B. Cords. 2007. Inactivation of
Bacillus anthracis spores by liquid biocides in the presence of food residue. Appl. Environ.
Microbiol. 73:6370-6377.
65.
Hilgren, J. D., and J. A. Salverda. 2000. Antimicrobial efficacy of a
peroxyacetic/octanoic acid mixture in fresh-cut-vegetable process waters. J. Food Sci. 65:13761379.
66.
Huang, Y., and H. Chen. 2011. Effect of organic acids, hydrogen peroxide and mild heat
on inactivation of Escherichia coli O157:H7 on baby spinach. Food Control. 22:1178-1183.
67.
Jay, M. T., M. Cooley, D. Carychao, G. W. Wiscomb, R. A. Sweitzer, L. CrawfordMiksza, J. A. Farrar, D. K. Lau, J. O'Connell, A. Millington, R. V. Asmundson, E. R. Atwill, and
R. E. Mandrell. 2007. Escherichia coli O157:H7 in feral swine near spinach fields and cattle,
central California coast. Emerg. Infect. Dis. 13:1908-1911.
248

68.
Jordan, S. L., C. Pascual, E. Bracey, and B. M. Mackey. 2001. Inactivation and injury of
pressure-resistant strains of Escherichia coli O157 and Listeria monocytogenes in fruit juices. J.
Appl. Microbiol. 91:463-469.
69.
Kaper, J. B., J. P. Nataro, and H. L. T. Mobley. 2004. Pathogenic Escherichia coli. Nat.
Rev. Microbiol. 2:123-140.
70.
Karmali, M. A., M. Petric, B. T. Steele, and C. Lim. 1983. Sporadic cases of haemolyticurameic syndrome associated with faecal cytotoxin and cytotoxin-producing Escherichia coli in
stools. Lancet. 321:619-620.
71.
Keene, W. E., K. Hedberg, D. E. Herriott, D. D. Hancock, R. W. McKay, T. J. Barrett,
and D. W. Fleming. 1997. A prolonged outbreak of Escherichia coli O157:H7 infections caused
by commercially distributed raw milk. J. Infect. Dis. 176:815-818.
72.
Keskinen, L. A., and B. A. Annous. 2011. Efficacy of adding detergents to sanitizer
solutions for inactivation of Escherichia coli O157:H7 on Romaine lettuce. Int. J. Food.
Microbiol. 147:157-61.
73.
Kitis, M. 2004. Disinfection of wastewater with peracetic acid: a review. Environ. Intern.
30:47-55.
74.
Koivunen, J., and H. Heinonen-Tanski. 2005. Inactivation of enteric microorganisms with
chemical disinfectants, UV irradiation and combined chemical/UV treatments. Water Res.
39:1519-1526.
75.
Lang, M. M., L. J. Harris, and L. R. Beuchat. 2004. Survival and recovery of Escherichia
coli O157:H7, Salmonella, and Listeria monocytogenes on lettuce and parsley as affected by
method of inoculation, time between inoculation and analysis, and treatment with chlorinated
water. J. Food Prot. 67:1092-1103.
76.
Lemons, K. E. 2009. Antimicrobial compositions and methods of use thereof. U.S. patent
2009/0192231 A1.
77.
Li, Y., R. E. Brackett, J. Chen, and L. R. Beuchat. 2001. Survival and growth of
Escherichia coli O157:H7 inoculated onto cut lettuce before or after heating in chlorinated water,
followed by storage at 5 or 15 C. J. Food Prot. 64:305-309.
78.
López-Gálvez, F., A. Allende, M. V. Selma, and M. I. Gil. 2009. Prevention of
Escherichia coli cross-contamination by different commercial sanitizers during washing of freshcut lettuce. Int. J. Food. Microbiol. 133:167-171.
79.
Lopez-Galvez, F., G. D. Posada-Izquierdo, M. V. Selma, F. Perez-Rodriguez, J. Gobet,
M. I. Gil, and A. Allende. 2012. Electrochemical disinfection: An efficient treatment to
inactivate Escherichia coli O157:H7 in process wash water containing organic matter. Food
Microbiol. 30:146-156.
249

80.
Luo, Y. 2007. Fresh-cut produce wash water reuse affects water quality and packaged
product quality and microbial growth in Romaine lettuce. Hort. Sci. 42:1413-1419.
81.
Luo, Y., X. Nou, P. Millner, and B. Zhou. 2012. A pilot plant scale evaluation of a new
process aid for enhancing chlorine efficacy against pathogen survival and cross-contamination
during produce wash. Int. J. Food. Microbiol. 158:133–139.
82.
Luo, Y., X. Nou, Y. Yang, I. Alegre, E. Turner, H. Feng, M. Abadias, and W. Conway.
2011. Determination of free chlorine concentrations needed to prevent Escherichia coli O157:H7
cross-contamination during fresh-cut produce wash. J. Food Prot. 74:352-358.
83.
Lynch, M. F., R. V. Tauxe, and C. W. Hedberg. 2009. The growing burden of foodborne
outbreaks due to contaminated fresh produce: risks and opportunities. Epidemiol. Infect.
137:307-315.
84.
Mandrell, R. E. 2009. Enteric human pathogens associated with fresh produce: sources,
transport, and ecology. p. 5-41. In B.A.N. Xuetong Fan, Christopher J. Doona, Florence E.
Feeherry, Robert B. Gravani (ed.). Microbial Safety of Fresh Produce. Wiley-Blackwell, Ames,
IA.
85.
Meng, J., M. P. Doyle, T. Zhao, and S. Zhao. 2007. Enterohemorrhagic Escherichia coli.
p. 260. In M.P. Doyle, and L.R. Beuchat (ed.). Food Microbiology: Fundamentals and Frontiers,
3rd ed. ASM Press, Washington D.C.
86.
Nou, X., and Y. Luo. 2010. Whole-leaf wash improves chlorine efficacy for microbial
reduction and prevents pathogen cross-contamination during fresh-cut lettuce processing. J.
Food Sci. 75:M283-M290.
87.
Nou, X., Y. Luo, L. Hollar, Y. Yang, H. Feng, P. Millner, and D. Shelton. 2011. Chlorine
stabilizer T-128 enhances efficacy of chlorine against cross-contamination by E. coli O157:H7
and Salmonella in fresh-cut lettuce processing. J. Food Sci. 76:218-224.
88.
Ölmez, H. 2010. Effect of different sanitizing methods and incubation time and
temperature on inactivation of Escherichia coli on lettuce. J. Food Safety. 30:288-299.
89.
Painter, J. A., R. M. Hoekstra, T. Ayers, R. V. Tauxe, C. R. Branden, F. J. Angulo, and P.
M. Griffin. 2013. Attribution of foodborne illnesses, hospitalizations, and deaths to food
commodities by using outbreak data, United States, 1998–2008. Emerg. Infect. Dis. 19:In press.
90.
Palumbo, M. S., J. R. Gorny, D. E. Gombas, L. R. Beauchat, C. M. Bruhn, B. Cassens, P.
Delaquis, J. M. Farber, L. J. Harris, K. Ito, M. T. Osterholm, M. Smith, and K. M. J. Swanson.
2007. Recommendations for handling fresh-cut leafy green salads by consumers and retail
foodservice operators. Food Prot. Trends. 27:892-898.
91.
Parish, M. E., L. R. Beuchat, T. V. Suslow, L. J. Harris, E. H. Garrett, J. N. Farber, and F.
F. Busta. 2003. Methods to reduce/eliminate pathogens from fresh and fresh-cut produce.
Compr. Rev. Food Sci. F. 2:161-173.
250

92.
Paton, J. C., and A. W. Paton. 1998. Pathogenesis and diagnosis of Shiga-toxin producing
Escherichia coli infections. Clinical Microbiology Reviews. 11:450-479.
93.
Ren, L., G. D. Posada-Izquierdo, and E. T. Ryser. 2013. Dissemination of Escherichia
coli O157:H7 from a contaminated shredder or conveyor belt to fresh-cut iceberg lettuce during
simulated commercial production. In, Unpublished data.
94.
Riley, L. W., R. S. Remis, S. D. Helgerson, H. B. McGee, J. G. Wells, B. R. Davis, R. J.
Hebert, E. S. Olcott, L. M. Johnson, N. T. Hargrett, P. A. Blake, and M. L. Cohen. 1983.
Hemorrhagic colitis associated with a rare Escherichia coli serotype. New Engl. J. Med.
308:681-685.
95.
Rodgers, S. L., J. N. Cash, M. Siddiq, and E. T. Ryser. 2004. A comparison of different
chemical sanitizers for inactivating Escherichia coli O157:H7 and Listeria monocytogenes in
solution and on apples, lettuce, strawberries, and cantaloupe. J. Food Prot. 67:721–731.
96.
Rosen, H., B. R. Michel, D. R. V. Devanter, and J. P. Hughes. 1998. Differential effects
of myeloperoxidase-derived oxidants on Escherichia coli DNA replication. Infect. Immun.
66:2655–2659.
97.
Ruiz-Cruz, S., E. Acedo-Felix, M. Diaz-Cinco, M. A. Islas-Osuna, and G. A. GonzalezAguilar. 2007. Efficacy of sanitizers in reducing Escherichia coli O157:H7, Salmonella spp. and
Listeria monocytogenes populations on fresh-cut carrots. Food Control. 18:1383-1390.
98.
Ryser, E. T., J. Hao, and Z. Yan. 2009. Internalization of pathogens in produce. p. 55-80.
In B.A.N. Xuetong Fan, Christopher J. Doona, Florence E. Feeherry, Robert B. Gravani (ed.).
Microbial Safety of Fresh Produce. Wiley-Blackwell, Ames, IA.
99.
Ryu, J. H., and L. R. Beuchat. 2005. Biofilm formation by Escherichia coli O157:H7 on
stainless steel: effect of exopolysaccharide and curli production on its resistance to chlorine.
Appl. Environ. Microbiol. 71:247-254.
100. Sapers, G. M. 2001. Efficacy of washing and sanitizing methods for disinfection of fresh
fruit and vegetable products. Food Technol. Biotech. 39:305–311.
101. Sapers, G. M. 2006. Washing and sanitizing treatments for fruits and vegetables. In
J.R.G. G. M. Sapers, and A. E. Yousef (ed.). Microbiology of Fruits and Vegetables. CRC Press,
Boca Raton, FL.
102. Sapers, G. M. 2009. Disinfection of contaminated produce with conventional washing
and sanitizing technology. p. 393-424. In G. Sapers, E. Solomon, and K.R. Matthews (ed.). The
Produce Contamination Problem: Causes and Solutions. Elsevier, Burlington. MA.
103. Scallan, E., R. M. Hoekstra, F. J. Angulo, R. V. Tauxe, M. A. Widdowson, S. L. Roy, J.
L. Jones, and P. M. Griffin. 2011. Foodborne illness acquired in the United States- major
pathogens. Emerg. Infect. Dis. 17:7-15.
251

104. Seo, K. H., and J. F. Frank. 1999. Attachment of Escherichia coli O157:H7 to lettuce leaf
surface and bacterial viability in response to chlorine treatment as demonstrated by using
confocal scanning laser microscopy. J. Food Prot. 62:3-9.
105. Shaw, R. K., C. N. Berger, B. Feys, S. Knutton, M. J. Pallen, and G. Frankel. 2008.
Enterohemorrhagic Escherichia coli exploits EspA filaments for attachment to salad leaves.
Appl. Environ. Microbiol. 74:2908-2914.
106. Shen, C., Y. Luo, X. Nou, G. Bauchan, B. Zhou, Q. Wang, and P. Millner. 2012.
Enhanced inactivation of Salmonella and Pseudomonas biofilms on stainless steel by use of T128, a fresh-produce washing aid, in chlorinated wash solutions. Appl. Environ. Microbiol.
78:6789-6798.
107. Shen, C., Y. Luo, X. Nou, Q. Wang, and P. Millner. 2013. Dynamic effects of free
chlorine concentration, organic load, and exposure time on the inactivation of Salmonella,
Escherichia coli O157:H7, and non-O157 Shiga toxin–producing E. coli. J. Food Prot. 76:386393.
108. Stafford, R. J., B. J. McCall, A. S. Neill, D. S. Leon, G. J. Dorricott, C. D. Towner, and
G. R. Micalizzi. 2002. A statewide outbreak of Salmonella Bovismorbificans phage type 32
infection in Queensland. Commun.Dis. Intell. 26:568-573.
109. Suslow, T. V. 1997. Postharvest chlorination: basic properties and key points for
effective disinfection. Publication 8003. Available at: http://anrcatalog.ucdavis.edu/pdf/8003.pdf.
Accessed 15 September 2012.
110. Suslow, T. V. 2000. Postharvest handling for organic crops. University of California,
Davis, Agriculture and Natural Resources. Publication 7254. Available at:
http://anrcatalog.ucdavis.edu/pdf/7254.pdf. Accessed 4 February 2013.
111. Suslow, T. V. 2001. Water disinfection, a practical approach to calculating dose values
for preharvest and postharvest applications. University of California, Davis, Agriculture and
Natural Resources. Publication 7256. Available at: http://anrcatalog.ucdavis.edu/pdf/7256.pdf.
Accessed 22 September 2012.
112. Suslow, T. V. 2004. Oxidation-reduction potential (ORP) for water disinfection
monitoring, control, and documentation. University of California, Davis, Agriculture and Natural
Resources. Publication 8149. Available at: http://anrcatalog.ucdavis.edu/pdf/8149.pdf. Accessed
13 September 2012.
113. Suslow, T. V. 2006. Making sense of rules governing chlorine contact in postharvest
handling of organic produce. University of California, Davis, Agriculture and Natural Resources.
Publication 8198. Available at: http://anrcatalog.ucdavis.edu/pdf/8198.pdf. Accessed 22
September 2012.
114. Swerdlow, D. L., B. A. Woodruff, R. C. Brady, P. M. Griffin, S. Tippen, H. D. Donnell,
E. Geldreich, B. J. Payne, A. Meyer, J. G. Wells, K. D. Greene, M. Bright, N. H. Bean, and P. A.
252

Blake. 1992. A waterborne outbreak in Missouri of Escherichia coli O157:H7 associated with
bloody diarrhea and death. Annal. Intern. Med. 117:812-819.
115. Szabo, E. A., L. Simons, M. J. Coventry, and M. B. Cole. 2003. Assessment of control
measures to achieve a food safety objective of less than 100 CFU of Listeria monocytogenes per
gram at the point of consumption for fresh precut iceberg lettuce. J. Food Prot. 66:256–264.
116. Tarr, P. I., C. A. Gordon, and W. L. Chandler. 2005. Shiga-toxin-producing Escherichia
coli and haemolytic uraemic syndrome. Lancet. 365:1073-1086.
117.

Tarver, T. 2009. Biofilms: a threat to food safety. Food Technol. 63:46-52.

118. Tirpanalan, Ö., M. Zunabovic, K. J. Domig, and W. Kneifel. 2011. Mini review:
antimicrobial strategies in the production of fresh-cut lettuce products Available at:
http://www.formatex.info/microbiology3/book/176-188.pdf. Accessed 17 Janurary 2013.
119. Todd, E. C. D., C. K. Harris, A. J. Knight, and M. R. Worosz. 2007. Spinach and the
media: how we learn about a major outbreak. Food Prot. Trends. 27:314-322.
120. Tomas-Callejas, A., G. Lopez-Velasco, A. M. Valdez, A. Sbodio, F. Artes-Hernandez,
M. D. Danyluk, and T. V. Suslow. 2012. Evaluation of current operating standards for chlorine
dioxide in disinfection of dump tank and flume for fresh tomatoes. J. Food Prot. 75:304-313.
121. Turini, T., M. Cahn, M. Cantwell, L. Jackson, S. Koike, E. Natwick, R. Smith, K.
Subbarao, and E. Takele. 2011. Iceberg lettuce production in California. University of California,
Davis, Agriculture and Natural Resources. Publication 7215. Available at:
http://anrcatalog.ucdavis.edu/pdf/7215.pdf. Accessed 15 January 2013.
122. U.S. Food and Drug Administration. 1998. Guidance for industry: guide to minimize
microbial food safety hazards for fresh fruits and vegetables. Available at:
http://www.fda.gov/Food/GuidanceComplianceRegulatoryInformation/GuidanceDocuments/Pro
duceandPlanProducts/ucm064574.htm#ii. Accessed 8 September 2012.
123. U.S. Food and Drug Administration. 2006. UPDATE: FDA narrows investigation of E.
coli O157:H7 outbreak at Taco Bell restaurants. Available at:
http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/2006/ucm108803.htm.
Accessed 2 September 2012.
124. U.S. Food and Drug Administration. 2007. FDA and states close to identifying sources of
E. coli contamination associated with illnesses at Taco John's restaurants. Available at:
http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/2007/ucm108827.htm.
Accessed 2 September 2012.
125. U.S. Food and Drug Administration. 2008. Commodity specific food safety guidelines
for the fresh tomato supply. Available at:
http://www.fda.gov/downloads/Food/GuidanceRegulation/UCM171708.pdf. Accessed 18
February 2013.
253

126. U.S. Food and Drug Administration. 2009. Guidance for industry: guide to minimize
microbial food safety hazards of leafy greens; draft guidance. Available at:
http://www.fda.gov/Food/GuidanceComplianceRegulatoryInformation/GuidanceDocuments/Pro
duceandPlanProducts/ucm174200.htm#fresh. Accessed 7 September 2012.
127. U.S. Food and Drug Administration. 2013. Standards for the growing, harvesting,
packing and holding of produce for human consumption, a proposed rule. Available at:
https://www.federalregister.gov/articles/2013/01/16/2013-00123/standards-for-the-growingharvesting-packing-and-holding-of-produce-for-human-consumption. Accessed 20 February
2013.
128. U.S. Food and Drug Administration. 2013. Standards for the growing, harvesting,
packing, and holding of produce for human consumption; current good manufacturing practice
and hazard analysis and risk-based preventive controls for human food; draft qualitative risk
assessment of risk of activity/food combinations for activities (outside the farm definition)
conducted in a facility co-located on a farm; availability; proposed rules. 21 CFR Parts 1, 16,
106. Available at: http://www.gpo.gov/fdsys/pkg/FR-2013-01-16/pdf/2013-00123.pdf. Accessed
14 January 2013.
129. Vandekinderen, I., F. Devlieghere, B. De Meulenaer, P. Ragaert, and J. Van Camp. 2009.
Optimization and evaluation of a decontamination step with peroxyacetic acid for fresh-cut
produce. Food Microbiol. 26:882-888.
130. Vorst, K. L., E. C. D. Todd, and E. T. Ryser. 2004. Improved quantitative recovery of
Listeria monocytogenes from stainless steel surfaces ssing a one-ply composite tissue. J. Food
Prot. 67:2212–2217.
131. Wachtel, M. R., L. C. Whitehead, and R. E. Mandrell. 2002. Association of Escherichia
coli O157:H7 with preharvest leaf lettuce upon exposure to contaminated irrigation water. J.
Food Prot. 65:18–25.
132. Warning, A., and A. K. Datta. 2013. Interdisciplinary engineering approaches to study
how pathogenic bacteria interact with fresh produce. J. Food Eng. 114:426–448.
133. Watanabe, M. M., Y. Takeda, T. Sasa, I. Inouye, S. Suda, T. Sawaguchi, and M. Chihara.
1987. A green dinoflagellate with chlorophylls and morphology, fine structure of the chloroplast
and chlorophyll composition. J. Phycol. 23:382-389.
134. Weissinger, W. R., W. Chantarapanont, and L. R. Beuchat. 2000. Survival and growth of
Salmonella baildon in shredded lettuce and diced tomatoes, and effectiveness of chlorinated
water as a sanitizer. Int. J. Food. Microbiol. 62:123-131.
135. Wuytack, E. Y., L. D. T. Phuong, A. Aertsen, K. M. F. Reyns, D. Marquenie, B. D.
Ketelaere, B. Masschalck, I. V. Opstal, A. M. Diels, and C. W. Michiels. 2003. Comparison of
sublethal injury induced in Salmonella enterica serovar Typhimurium by heat and by different
nonthermal treatments. J. Food Prot. 66:31-37.
254

136. Xicohtencatl-Cortes, J., E. S. Chacon, Z. Saldana, E. Freer, and J. A. Giron. 2009.
Interaction of Escherichia coli O157:H7 with leafy green produce. J. Food Prot. 72:1531-1537.
137. Zagory, D. 1999. Effects of post-processing handling and packaging on microbial
populations. Postharvest Biol. Technol. 15:313-321.
138. Zeng, W., K. Vorst, W. Brown, B. Marks, F. Perez-Rodriguez, and E. Ryser. 2012.
Monte Carlo simulation of Escherichia coli O157:H7 and Listeria monocytogenes growth in
bagged salad greens during commercial transport, retail storage and display, Abst T4-12. In,
Ann. Mtg. Int. Assoc. Food Prot., Providence, Rhode Island.
139. Zhang, G., L. Ma, V. H. Phelan, and M. P. Doyle. 2009. Efficacy of antimicrobial agents
in lettuce leaf processing water for control of Escherichia coli O157:H7. J. Food Prot. 72:1392–
1397.
140. Zhang, S., and J. M. Farber. 1996. The effects of various disinfectants against Listeria
monocytogenes on fresh-cut vegetables. Food Microbiol. 13:311-321.
141. Zhou, B., H. Feng, and A. J. Pearlstein. 2012. Continuous-flow ultrasonic washing
system for fresh produce surface decontamination. Innov. Food Sci. Emerg. 16:427-435.

255