Subscriber access provided by University of Newcastle, Australia
Article
Quantitative Effects of Biochar Oxidation and Pyrolysis Temperature on the Transport of Pathogenic and Nonpathogenic E. coli in Biochar-Amended Sand Columns Waled Omar Suliman, James B. Harsh, Ann-Marie Fortuna, Manuel Garcia-Perez, and Nehal I Abu-Lail Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b04535 • Publication Date (Web): 30 Mar 2017 Downloaded from http://pubs.acs.org on March 31, 2017
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 36
Environmental Science & Technology
1
Quantitative Effects of Biochar Oxidation and Pyrolysis
2
Temperature on the Transport of Pathogenic and Nonpathogenic E.
3
coli in Biochar-Amended Sand Columns
4 5
Waled Suliman1,2, James B. Harsh2, Ann-Marie Fortuna4, Manuel Garcia-Pérez3, and Nehal I.
6
Abu-Lail5,* 1
7 8
2
Department of Crop and Soil Sciences, Washington State University, Pullman, WA 99164, USA 3
9
Department of Microbiology, Benghazi University, Benghazi, LYB2186, Libya
Biological Systems Engineering Department, Washington State University, Pullman, WA, 99164, USA
10 4
11 12
5
Soil Science Department, North Dakota State University, Fargo, ND, 58108, USA
The Gene and Linda Voiland School of Chemical Engineering and Bioengineering, Washington State University, Pullman, WA, 99164, USA
13 14 15 16 17
Corresponding author
18
*
19
1505 Stadium Way
20
313 Wegner Hall
21
Gene and Linda Voiland School of Chemical Engineering and Bioengineering
22
Washington State University
23
Pullman, WA 99164-6515
24
Tel: 509-335-4961
25
[email protected] Nehal I. Abu-Lail
1 ACS Paragon Plus Environment
Environmental Science & Technology
26
Abstract
27
The present study quantifies the transport of Escherichia coli pathogenic O157:H7
28
and nonpathogenic K12 strains in water-saturated Quincy sand (QS) columns amended with
29
oxidized (OX) or unoxidized (UO) pine wood (PW) or pine bark (PB) biochar produced at either
30
350oC or 600oC. Our results showed that: (1) the addition of oxidized biochar into QS columns
31
enhanced the transport of E. coli O157:H7 by 3.1 fold compared to the unoxidized counterparts
32
likely due to an increase in the repulsive forces due to their higher negative charge densities;
33
(2) the retention of E. coli O157:H7 was 3.3 fold higher than that of E. coli K12 in all biochar-
34
amended sand columns; and (3) increased application rates of unoxidized PW600 biochar from 0
35
to 20wt.% led to a reduction in the transport of E. coli O157:H7 and K12 from 98 to 10% and
36
from 95 to 70%, respectively. Our data showed that mixing sand with PW350-UO at a 20 wt.%
37
application rate almost completely retained the pathogenic E. coli in the subsurface; suggesting
38
that utilizing sand mixed with biochar can act as a promising biofilter capable of protecting
39
natural aquafers from pathogens.
40 41 42 43 44 45 46 47 48
2 ACS Paragon Plus Environment
Page 2 of 36
Page 3 of 36
49
Environmental Science & Technology
1. Introduction
50
The transport of pathogenic bacteria through soil represents a scientific and public concern due
51
to their potential to cause groundwater contamination and outbreaks of waterborne diseases1. E.
52
coli O157:H7, a Gram-negative, facultative anaerobic, enteric bacterium normally excreted in
53
human and animal feces2,3, is a recognized pathogen capable of causing a wide range of diseases
54
in humans3,4. In the US annually, E. coli O157:H7 causes 75,000 infections and 61 deaths5,6. In
55
addition, E. coli O157:H7 is responsible for massive disease outbreaks such as the tragedy that
56
occurred in Walkerton, Canada and that resulted in infections to nearly half of the 5,000 people
57
living there due to contamination of drinking groundwater with E. coli O157:H77. Due to their
58
abundance, exploring means to reduce the transport of E. coli O157:H7 in the subsurface is
59
needed.
60
The modification of the natural filtration capacity of sand is one of the means by which
61
the risk of groundwater contamination by pathogenic bacteria can be reduced8. This can be
62
achieved by the incorporation of organic amendments such as compost and/or biochar, a carbon-
63
rich porous material, into sand. Similarly, but relying on the indigenous nonpathogenic bacterial
64
strains in soil, researchers have suggested that biofilms of these strains can clog pores in sand
65
and potentially be used to reduce the hydraulic conductivity in soil9. Although this is an
66
environmental friendly approach, keeping bacterial viability under subsurface soil stresses such
67
as reduced oxygen can be challenging10. Finally, planting and vegetative soil covering have been
68
widely used as means to modify the filtration capacity of soil11. This option is viable only in
69
locations where plants can grow12. In comparison, in addition to its potential role as a pathogen
70
biofilter, biochar can enhance soil fertility and sequester carbon13. The ability of biochar to
71
attenuate microbes depends largely on its physiochemical properties including surface area, pore
3 ACS Paragon Plus Environment
Environmental Science & Technology
72
volume, pH, volatile matter, oxygen content, and hydrophobic properties14. Biochar
73
physiochemical properties are dependent on biochar production conditions as well as on the
74
feedstock used to produce the biochar. Specifically, biochar charge and hydrophobicity can be
75
largely controlled by manipulating biochar’s production temperature15 as well as post-oxidation
76
processes16. In comparison, biochar’s ash content, total carbon, fixed carbon, and mineral
77
concentrations are mainly affected by the feedstock properties17,18.
78
Recent laboratory investigations of E. coli transport in biochar-amended soils
79
documented positive effects of biochar’s ability to increase the attenuation of E. coli through
80
modifying soil surface area, hydrophobicity, charge, organic carbon content, porosity, or solution
81
chemistry14,19,20. However, most of biochar studies on bacterial transport were conducted with
82
raw-heterogeneous biochar grains disregarding the role of biochar surface functionality on
83
transport. Knowledge acquired with only bulk properties of fresh raw-biochar cannot be
84
transferred easily to natural soil environments, where biochar gets exposed to biotic and abiotic
85
oxidation processes. Better fundamental understanding of how biochar physiochemical
86
properties affect bacterial transport in biochar-amended sand will enable better design of
87
production processes to yield biochar with desired properties for enhanced bacterial attenuation.
88
Here, we compared and quantified the transport of a pathogenic and a nonpathogenic E.
89
coli strain in sand amended with oxidized and unoxidized biochars produced at two pyrolysis
90
temperatures (350 oC and 600 oC) from two commonly used feedstock sources (pine wood and
91
pine bark).
92
characterized air-oxidized biochar simulating subsurface natural conditions was also
93
investigated.
94
2. Materials and Methods
The role of biochar’s surface functionality on E. coli transport using well-
4 ACS Paragon Plus Environment
Page 4 of 36
Page 5 of 36
Environmental Science & Technology
95
2.1 Bacterial Cultures
96
E. coli strains investigated were obtained from Douglas Call, a professor of the Department of
97
Veterinary Microbiology and Pathology at Washington State University (WSU). The strains
98
were activated by growing for twelve hours at 37˚C on a shaker rotating at 150 rpm in Luria
99
Bertani (LB) medium adjusted to a pH value of 7.0 using 2N HCl. Following incubation, 1%
100
volume of the activated culture was transferred into 20 ml of LB medium and grown at the
101
conditions above until the late stationary phase of growth. Cells were harvested when the
102
absorbance at 600 nm (read using a Lambda 25 Perkin Elmer spectrophotometer) reached a value
103
of 0.56 ± 0.02. The specific growth constants were found to be 0.018 and 0.014 h-1 for the
104
pathogenic and nonpathogenic strains, respectively. After growth, bacterial suspensions were
105
centrifuged at 5,100g for 10 minutes. Cell pellets were collected and washed three times with
106
deionized water (DIW). Bacterial pellets were then resuspended in DIW to a concentration of
107
0.05 optical density at 600 nm (equivalent to 1.2 x 108 CFU/mL). Ionic strengths of solutions in
108
environmental systems can vary from < 0.01 M in groundwater (such as that in DIW), to about
109
0.1 M (can vary between 0.03 and 0.3 M) in brackish water, to 0.57 M in seawater21. All studies
110
done here were performed in DIW as the extreme low end of ionic strengths present in soil.
111
2.2 Electrophoretic Mobility Measurements
112
Bacterial cells grown as described above were washed twice by centrifugation at 5,100g for 10
113
min each round. Collected bacterial pellets were then diluted with 0.2 µm filtered DIW adjusted
114
to the pH of interest using either 0.1 M HCl or NaOH until an optical density of ~ 0.05 was
115
obtained at λ = 600 nm. Electrophoretic mobilities of E. coli cells were then measured in pH
116
values of 1.8, 2.8, 4, 7.7, and 8.5 for E. coli K12 and in pH values of 1.7, 2.1, 3.6, 5.1, 7 and 8
117
for E. coli O157:H7. To measure the mobilities, 2 ml of the bacterial cell suspension was
5 ACS Paragon Plus Environment
Environmental Science & Technology
118
injected using a syringe into a nano-Zetasizer 3000 HSA (Malvern Instruments Ltd., Malvern,
119
UK). All electrophoretic mobility measurements were performed five times at room temperature.
120
The five times’ replicates were taken from different cultures grown on the same day.
121
2.3 Biochar Preparation and Characterization
122
Pine Wood and Pine Bark biochars were produced and characterized at the biomass
123
thermochemical conversion laboratory at WSU according to details reported elsewhere15,22 and
124
briefly summarized in Table (1). Here, biochars are denoted as PW for the pine wood feedstock
125
and PW-350, and PW-600 for the resulting unoxidized (UO) biochars created at 350 oC and 600
126
o
127
was applied for the pine bark (PB) feedstock batches. For brief descriptions of how biochar was
128
prepared and characterized, please refer to the supporting information sections 1 and 2.
C, respectively. The oxidized samples are labeled with an OX. Similar abbreviation procedure
129 130 131 132 133 134 135 136 137 138 139 140
6 ACS Paragon Plus Environment
Page 6 of 36
Page 7 of 36
Environmental Science & Technology
141 142
Table 1. General physicochemical properties of biochars used in this study15,22 PW350
PW600
PB350
PB600
Property UO
OX
UO
OX
UO
OX
UO
OX
Biochar yield (wt. %)
36.0
37.22
16.5
14.96
46.7
41.85
30.9
26.16
Carbon content (wt. %)
70.50
65.37
87.80
85.08
66.07
63.85
78.11
78.07
Oxygen content (wt. %)
23.15
29.17
7.06
10.27
23.67
27.11
8.98
10.21
Nitrogen content (wt. %)
0.26
0.26
0.38
0.32
0.67
0.60
0.75
0.73
Hydrogen content (wt. %)
5.50
4.79
3.63
3.44
4.88
4.49
3.32
3.29
Fixed carbon content (wt. %)
49.58
48.33
83.15
79.27
51.50
48.71
73.89
70.14
Volatile matter content (wt. %)
49.82
51.24
15.72
20.84
42.15
47.33
17.81
21.56
Ash content (wt. %)
0.58
0.42
1.13
0.89
4.72
3.95
8.85
7.70
Mg (mg kg )
0.11
0.17
0.21
0.20
0.16
0.19
0.76
0.19
K (mg kg-1)
0.41
0.40
0.72
0.74
0.61
0.71
2.01
2.20
-1
0.65
0.77
1.18
1.24
1.15
1.44
4.36
4.33
0.16
0.28
0.12
0.23
0.14
0.11
0.17
0.16
Bulk density (g cm )
0.14
0.14
0.24
0.24
0.61
0.61
0.21
0.21
Particle density (g cm-3)
0.56
0.56
0.7
0.7
0.45
0.45
0.48
0.48
71
71
78
78
64
64
68
68
Moisture content (%)
1.69
2.5
1.71
2.6
3.13
3.4
2.22
2.12
pH H2O
8.20
7.8
8.79
8.10
7.88
7.5
10.19
9.00
0.01
0.02
0.07
0.05
0.07
0.06
0.19
0.08
53.96
66.59
52.20
66.90
36.28
56.06
30.37
30.83
3.0
1.2
5.6
3.2
1.6
1.3
4.4
3.7
Total OFGs (atom %)
45.59
49.95
16.79
29.64
57.31
75.54
9.22
23.46
Total AFGs (mmol g-1)
0.15
0.61
0.03
0.13
0.16
0.34
0.02
0.12
0.01
0.00
0.03
0.02
0.04
0.03
0.11
0.11
146
190
500
570
172
187
424
550
0.14
0.23
0.57
0.75
0.20
0.24
0.43
0.81
-1
Ca (mg kg ) -1
Na (mg kg ) -3
Porosity (%)
EC H2O (ds m-1) -1
CEC (C-molc kg ) pHpzc
-1
Total BFGs (mmol g ) 2
-1
SACO2 (m g ) 3
-1
Total PVCO2 (cm g )
143 144 145
PW and PB: pine wood and pine park, 350 oC and 600 oC: pyrolysis temperatures, UO: unoxidized, OX: oxidized, EC: electric conductivity, CEC: cation exchange capacity, OFG: oxygen-containing functional groups, AFG: acidic functional groups, BFG: basic functional groups, SA: surface area and PV: pore volume.
146
2.4 Quincy Sand
147
Quincy sand (QS) was used in this study because it is present on nearly 700,000 acres in
148
Washington, Oregon, and Idaho, and is an agriculturally important sand in the Pacific Northwest 7 ACS Paragon Plus Environment
Environmental Science & Technology
149
region of the US (NRCS, 2013). The sand was collected, air-dried, and sieved through a 2 mm
150
mesh23. Some selected physical and chemical characteristics of the sand are shown in Table S1 in
151
the supporting information. To measure the zeta potential of sand, a 5 g of sand was added to a
152
100 ml of deionized water and agitated on an orbital shaker for 6 hours at 25oC. The aliquot of
153
the supernatant was then decanted from the container and analyzed by a Nano-Zetasizer 3000
154
(Malvern Instruments Ltd., Malvern, UK).
155
2.5 Bacterial Transport in Porous Media
156
72 experiments were conducted to evaluate the transport of pathogenic and nonpathogenic E. coli
157
in biochar-amended sand columns. Three types of porous media were used. These were Quincy
158
sand (QS) and mixtures of QS and 8 types of either unoxidized or oxidized biochars. Prior to
159
measurements, the QS was cleaned and washed with DIW five times in order to remove any
160
residual fine particles that could interfere with measuring the bacterial concentration or clog the
161
column tubing. The sand and biochar samples were sieved to a uniform size with an average
162
diameter of 500 µm. Prior to each experiment, 20 wt.% of a biochar was mixed thoroughly with
163
the sand and packed dry in columns. Packed columns were acclimated to DIW for 10 pore
164
volumes (1 PV = 11 mL) using a peristaltic pump (Vera Manostat/Barnant). The absorbance of
165
the effluent was measured after equilibration and values of zero further ensured that residuals of
166
fine particles were removed. The pH of the effluent was continuously monitored using Hanna HI
167
9813-6 pH/EC/TDS meters (HANNA instruments Inc., US) to ensure a constant pH throughout
168
experimentation. After that, a bacterial suspension was pumped upward at 16 ml/min through
169
sand-biochar columns. Upward flow was used in order to eliminate the gravity force’ effects on
170
transport and to displace air from most pores between column filling material.
8 ACS Paragon Plus Environment
Page 8 of 36
Page 9 of 36
Environmental Science & Technology
171
The bacterial suspension was allowed to flow through the column for at least 7 PV
172
followed by 5 PV of bacteria-free solution to elute all cells that may not have been retained on
173
the porous media. Samples of the column effluent were collected in 3.25-mL fractions using an
174
automated fraction collector (Model CF1, Spectrum Chromatography). Subsequently, the
175
bacterial concentration in the effluent was quantified using a spectrophotometer at 600 nm.
176
Bacterial breakthrough curves (BTCs) were presented in a dimensionless form by dividing the
177
outlet concentration (C) by the inlet concentration (C0) and the C/C0 fractions were plotted
178
against the PV values. Each column experiment was repeated at least three times on separate
179
days using different cultures taken from different inoculations to ensure a true replication.
180
2.6 Mathematical Modeling of the Filtration Efficiency
181
In porous media, the rate at which bacteria attach to sand can follow different forms of kinetics.
182
However, a second order blocking kinetics’ model is commonly used to account for the
183
reduction in the bacterial attachment rate as a function of time due to the saturation of the
184
favorable sand available sites for bacterial attachment24-26. The lack of a steady state bacterial
185
concentration region in the majority of our collected breakthrough curves suggests that blocking
186
is occurring in the sand column in the presence and absence of biochar. To check if blocking
187
indeed is affecting bacterial retention in our porous media (sand and biochar), an advection-
188
dispersion model with a second order bacterial attachment kinetics was employed to the data24–28.
189
According to Tan et al. 1994 and Saiers et al. 1994, the continuity equation for the one-
190
dimensional transport of bacteria in the water filling the pores of the sand column can be written
191
as 24,25:
192
∂C ρ b ∂S q ∂ 2C q ∂C + = α1 − ∂t θ ∂t θ ∂x 2 θ ∂x 9 ACS Paragon Plus Environment
(1)
Environmental Science & Technology
193
Where C is the bacterial concentration in the pore water (mg dry cell weight/ cm3 of pore water),
194
t is time, ρb is the fluid bulk density (mg dry sand/ cm3 of bed), θ is bed porosity (cm3 pore
195
volume/cm3 of bed), S is the concentration of bacteria attached to the sand (mg dry cell weight/
196
mg dry sand weight), x is the dimension coordinate of flow (in the length direction of the
197
column), q is Darcy’s velocity (cm/s), and α1 is the longitudinal dispersivity (cm). Equation 1
198
assumes no growth of bacteria occurs in the pore water during experimentation time. Because the
199
columns we used were relatively short, bacteria were assumed to be well dispersed in the
200
longitudinal direction of the column and as such, the concentration of bacteria was assumed to be
201
constant as a function of column depth. As such, the right hand side of the equation was set to
202
zero.
203
When blocking is a part of the retention process, then a second order kinetics’ model is used to
204
describe the attachment rate and a first order kinetics’ model is used for detachment rate. In
205
equation 1, the attachment rate now depends both on the bacterial concentration in the pore water
206
(C) as well as the concentration of bacteria attached to the sand (S) and is described by:
ρ b ∂S ρ = k cψC − b k y S θ ∂t θ
207
208
(2)
ψ =
209
S max − S S max
210
(3)
211
Where kc and ky are the attachment and detachment coefficients (1/s), Smax represents the
212
maximum mass of bacterial cells attached to the sand divided by the dry weight of sand and can
213
be obtained experimentally through performing a mass balance on the bacterial cells in the 10 ACS Paragon Plus Environment
Page 10 of 36
Page 11 of 36
Environmental Science & Technology
214
column. ψ describes a fraction of the maximum bacterial attachment capacity as shown by
215
equation 3.
216
In equations 1-3, θ is measured experimentally and kc and ky are considered as fitting
217
parameters that can be tailored to allow the solutions of equations 1-3 to fit the experimental
218
data. Equations 1-3 were solved simultaneously using the initial conditions (at t=0, C=0, S=0).
219
The equations were solved numerically. The solutions of the equations were fit to the
220
experimental data of C/C0 as a function of time. The quality of the fit was judged by maximizing
221
the value of the coefficient of determination (R2) which is used to indicate the model’s ability to
222
fit the experimental data. Values closer to 1 indicate a better regression model. On average, R2
223
values were 0.94±0.03 (n=9). We choose to fit only the pseudo-steady state portion of the data
224
because we are mainly interested in comparing the strengths of the attachment for the different
225
biochar treatments investigated. Figure 1 shows a representative example of how the second
226
order kinetics’ model fit our data. For all the conversions and the parameters used in our
227
calculations, please refer to supporting materials section 4. For modeling results, please refer to
228
section 5 and Tables S3 and S4.
11 ACS Paragon Plus Environment
Environmental Science & Technology
Page 12 of 36
1.0
C/C0
0.8 0.6 0.4 0.2 0.0 0
2
4
6
8
10
12
14
Pore Volume
229 230 231 232 233 234
Figure 1. A representative example of the ability of the second order kinetics’ model of bacterial attachment (equations 1-3) to fit the experimental data (kc=2.88x10-4 sec-1, ky =1.8x10-3 sec-1, and R2= 0.9315). Black circles show breakthrough data obtained for the transfer of E. coli O157:H7 in sand mixed with 5 wt.% PW-600 biochar. The red solid line is the blocking model fit to the data.
235
The values of kc obtained from fitting equations 1-3 to the experimental data can be
236
related to collision efficiency values (α) via the following equation:
kc =
237
3 q 1−θ ηα 2 θ dP
238
(4)
239
Where dP is the particle diameter and η is collector efficiency, taken as the sum of the physical
240
forces affecting collisions. These forces include diffusion, effects of neighboring particles,
241
London-van der Waals forces, interception, and gravitational settling. Using the model of
12 ACS Paragon Plus Environment
Page 13 of 36
Environmental Science & Technology
242
Rajagopalan and Tien, (1976), as modified by Logan et al., (1995), the single collector efficiency
243
was calculated as: −2 / 3 1/8 /8 η = 4 As1 / 3 N Pe + As N Lo N 15 + 0 .0338 As N G1.2 N R−0.4 R
244 245
(5)
246
where As is dimensionless Happel correction factor, Npe is dimensionless Peclet number, NLo is
247
dimensionless number that incorporates London-van der Waals attractive forces, NR is
248
dimensionless interception number, and NG is dimensionless gravitation number. Hereafter, the
249
single soil/biochar grain is termed as the collector while a bacterial cell is called particle. Table
250
S3 in supporting information summarizes the values of the parameters used in the calculations of
251
η and α.
252
For comparison purposes between treatments, the average value of C/C0 in the pseudo-
253
steady state region (approximately present between pore volumes 2 and 10 for most treatments)
254
was used to calculate the retention rates. Table S4 in the supporting information provides all the
255
pseudo-steady-state C/C0-ss values and the calculated collision efficiencies using the blocking
256
model for the pathogenic E. coli O157:H7 strain and the nonpathogenic E. coli K12 strain.
257
3. Results
258
3.1 Zeta Potentials of Bacteria, Sand, and Biochars
259
The ζ-potentials of the bacteria, sand, and biochars under different pH conditions are shown in
260
Figure 2. Our data show that the ζ-potentials for all the biochars investigated decreased linearly
261
with increase in environmental pH with the oxidized biochars being more negatively charged
262
than the un-oxidized ones (Figures 2A-2D). The ζ-potential of PB-600 whether oxidized or not
263
(Fig. 2D) decreased linearly with increase in pH. Similarly, the ζ-potential for QS decreased 13 ACS Paragon Plus Environment
Environmental Science & Technology
Page 14 of 36
264
linearly with increasing the pH (Fig. 2E). In the range of pH between 5.0 and 8.0, ζ-potentials of
265
both bacterial strains were negative and generally became more negative as the pH increased
266
(Fig. 2F). In a dramatic contrast, the negative ζ-potential of E. coli K12 increased sharply in the
267
pH range investigated while the negative ζ-potentials E. coli O157:H7 increased slightly over the
268
same range; indicating that the ζ-potentials of the nonpathogenic E. coli were more sensitive to
269
pH changes compared to those measured for the pathogenic strain. Our results also indicate that
270
the pathogenic E. coli was less negatively charged than the nonpathogenic E. coli. This finding
271
suggests that the repulsive electrostatic forces between the bacterium and sand or sand/biochar
272
mixture will be weaker for the pathogenic E. coli compared to the nonpathogenic E. coli.
273 274 275 276 277 278 279
40
A)
Zeta Potential (mV)
Zeta Potential (mV)
280
PW-350
20
UO
OX
0 -20 -40 -60 0
2
4
6
8
40
PW-600
B)
20
UO
OX
6
8
0 -20 -40 -60
10
0
pH
14 ACS Paragon Plus Environment
2
4
pH
10
Environmental Science & Technology
40
PB-350
C)
UO
20
Zeta Potential (mV)
Zeta Potential (mV)
Page 15 of 36
OX
0 -20 -40 -60
D)
40
PB-600 UO
20 0 -20 -40 -60
0
2
4
6
8
10
0
2
4
6
E)
QS sand
Zeta Potential (mV)
40
8
10
pH
pH
Zeta Potential (mV)
OX
20 0 -20 -40 -60
40
F)
E. coli
20 0 -20 -40
K12
O157:H7
-60 0
2
4
6
8
10
0
pH
2
4
6
8
10
pH
281 282 283 284 285
Figure 2. Average values of zeta potentials for unoxidized and oxidized biochars [(A) PW-350, (B) PW-600, (C) PB-350, and (D) PB-600), (E) sand, and (F) pathogenic and nonpathogenic bacteria (n=3). Error bars represent the standard deviation and are included for all samples but sometimes are too small to be seen on the graph.
286
Figure 3 shows BTCs collected for E. coli O157:H7 and K12 transport using PW-600-UO at
287
concentrations (0, 1, 5, 10, and 20 wt.%). As shown in Figure 3, all BTCs displayed a relatively
288
sharp breakthrough front and a high degree tailing, except in the case of E. coli O157:H7 with 20
289
wt.% biochar, where the BTC was rising slowly. The pathogenic strain was more susceptible to
290
biochar addition compared to the nonpathogenic strain (Figure 3). At 20% biochar application,
291
~89% and 26% of the pathogenic and nonpathogenic bacteria, respectively were retained in the
3.2 Effect of Biochar Concentration on E. coli Transport
15 ACS Paragon Plus Environment
Environmental Science & Technology
292
column. This is an indication of possible stronger interactions present between the surface of the
293
biochar/sand mixture and that of the pathogenic bacteria compared to the nonpathogenic one.
Page 16 of 36
294
For both bacterial strains and under all conditions investigated, bacterial retention in the
295
column was dependent on the biochar concentration. For example, at pore volumes between 4
296
and 9, the addition of only 1wt.% of biochar to sand influenced bacterial transport in the column
297
for both strains (Figures 3A-3C). The transport rates of both E. coli O157:H7 and E. coli K12
298
were reduced by 2 fold when 5 wt.% biochar was added to sand compared to the control (0%
299
biochar). Moreover, and as can be seen from Figure 3 A-C, the BTCs obtained for the two
300
bacterial strains were similar when no biochar was applied. The same holds true when only 1%
301
biochar was applied. However, with increase in biochar concentration from 5 to 20 wt.%, BTCs
302
obtained for both strains were quite different (Figures 3A & 3B). For E. coli O157:H7, pseudo-
303
steady state plateaus were significantly different among the BTCs displaying the different
304
biochar concentrations with most reduction in transport observed when the 20 wt.% biochar was
305
used. On the contrary, E. coli K12 pseudo-steady states obtained for 5, 10 and 20% biochar
306
concentrations were statistically similar (p > 0.05). Throughout the washing phase, C/C0 fell
307
sharply to near zero values in all experiments; indicating an irreversible attachment of both
308
strains to QS or QS/biochar mixtures. The pseudo-steady state values of C/C0 decreased linearly
309
with increased biochar concentrations for E. coli O157:H7 (Figure 3C). However, the pseudo-
310
steady state values reached a plateau after 5 wt.% biochar for the E. coli K12 strain. Because
311
using 20 wt.% biochar resulted in almost complete attenuation of pathogenic E. coli in the sand
312
column, only 20 wt.% biochar concentration was investigated in further experiments.
16 ACS Paragon Plus Environment
Page 17 of 36
Environmental Science & Technology
1.2
1.2 1.0
C/C0
0.8 0.6
B ) K1 2
0% 1% 5% 10% 20%
0% 1% 5% 10% 20%
1.0 0.8 C/C0
A) O 157 :H 7
0.6
0.4
0.4
0.2
0.2 0.0
0.0 0
2
4
6
0
8 10 12 14 16 18 20 PV
2
4
6
8 10 12 14 16 18 20 PV
Steady State C/C0
1.2
C)
1 0.8 0.6 0.4 0.2 0 0
313 314 315 316 317 318 319 320 321 322 323
10 20 Biochar concentration (wt.%)
30
Figure 3. Bacterial breakthrough curves (BTCs) obtained for E. coli (A) O157:H7 and (B) K12 in saturated sandy columns amended with PW600-UO biochar at different application rates. Error bars indicate the standard deviation obtained from triplicate experiments. C) A scatter graph that shows the relationship between the biochar concentration (% wt.) and the steady state normalized effluent bacterial concentration (C/C0) for E. coli O157:H7 (black-filled circles) and K12 (yellow-filled circles). The solid line is described by: Y=-0.0395X+0.9276, r2=0.9931. Note that Figures 3A and 3B display some widening in our BTCs, especially for the 20% biochar application treatment. The widening observed was the result of running the experiment for longer times prior to the washing phase to capture more data that can be used to explain the blocking effects observed.
17 ACS Paragon Plus Environment
Environmental Science & Technology
Page 18 of 36
324
3.3 The Effects of Low-Temperature Biochars on Bacterial Transport
325
Figure 4 shows the BTCs of E. coli O157:H7 and K12 transport through QS mixed with either
326
low-temperature (350 oC) PW or PB derived biochars at 20 wt.% application rate. The shapes of
327
the BTCs varied as a function of biochar used as well as microbe tested. For example, Figures
328
4C and 4D showed symmetrical BTCs with relatively sharp breakthrough front and sharp tailing.
329
This shape was the most dominant BTC shape for E. coli O157:H7 transfer in QS alone or in
330
QS-PW350-OX as well as for all cases of E. coli K12 transport. Such BTCs indicate that the
331
rates of adsorption and desorption of bacteria from biochar-sand mixture are similar. In
332
comparison, Figure 4B which represents the transport kinetics of E. coli O157:H7 in PB with or
333
without biochar, was asymmetrical exhibiting a degree of tailing; indicating that the adsorption
334
rates of bacteria are smaller than their desorption rates. Figure 4A showed a horizontal-like shape
335
BTC; indicating strong and irreversible attachment of E. coli O157:H7 bacteria to UO-PW350
336
sand mixture.
337
Our results revealed that feedstock source and oxidation of low-temperature biochar had
338
a relatively minor influence on the transport of K12 with almost no bacterial retention in the sand
339
columns (Figure 4C & 4D). In a dramatic contrast, both the feedstock source and the oxidation
340
status of the biochar significantly altered the transport of E. coli O157:H7. The PW-350-UO had
341
the highest performance with 98.5% of cells retained. Oxidation of PW and PB reduced the
342
retention of E. coli O157:H7 by 9- and 7.5- fold, respectively.
18 ACS Paragon Plus Environment
Page 19 of 36
Environmental Science & Technology
1.2
1.2 QS/Control
A) O157:H7
B) O157:H7
QS+PW-350-UO
1
QS/Control
1
QS+PB-350-UO
QS+PW-350-OX
0.8
QS+PB-350-OX
C/Co
C/Co
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0
0 0
2
4
6
8 PV
10
12
14
16
0
2
4
6
1.2
1.2 1
12
14
16
QS+PB-350-UO
1
QS+PW-350-UO
10
QS/Control
D) K12
QS/Control
C) K12
8 PV
QS+PB-350-OX
QS+PW-350-OX
0.8
C/Co
C/Co
0.8 0.6
0.6
0.4
0.4
0.2
0.2 0
0 0
2
4
6
8 PV
10
12
14
16
0
2
4
6
8 10 PV
12
14
16
343 344 345 346
Figure 4. Bacterial breakthrough curves (BTCs) for column experiments run with: E. coli O157:H7 in (A) PW350, (B) PB350, and E. coli K12 in (C) PW350, and (D) PB350. Error bars indicate the standard deviation of triplicates.
347
3.4 Effects of High Temperature Biochar on E. coli Transport
348
As can be seen from Figure 5, when unoxidized PW600 was mixed with QS at rate of 20wt.%,
349
the attenuation of E. coli O157:H7 and K12 increased by 82% and 22%, respectively, compared
350
to the control. Similarly, the addition of PB600 to QS at 20wt% increased the attenuation of E.
351
coli O157:H7 by 62% compared to QS control. However, when the transport of K12 was
19 ACS Paragon Plus Environment
Environmental Science & Technology
Page 20 of 36
352
assessed in PW600 or PB600, only slight improvements were observed in the retention of cells.
353
Figure 5 also shows that mixing the oxidized biochars, irrespective of their feedstock type, into
354
QS promoted the transport of both E. coli strains by 5% to 65% compared with their unoxidized
355
counterparts.
1.2
1.2
A) O157:H7
QS/Control
1
QS/Control
B) O157:H7
QS+PW-600-UO
QS+PB-600-UO
1
QS+PW-600-OX
QS+PB-600-OX
0.8 C/Co
C/Co
0.8 0.6
0.6
0.4
0.4
0.2
0.2
0
0 0
2
4
6
8
10 12 14 16 18 PV
0
1.2
10
12
14
16
QS/Control QS+PB-600-UO QS+PB-600-OX
0.8
0.8 C/Co
C/Co
8 PV
1
QS+PW-600-OX
0.6
0.6
0.4
0.4
0.2
0.2
0
0 0
356 357 358 359
6
D) K12
QS+PW-600-UO
1
4
1.2
QS/Control
C) K12
2
2
4
6
8
10 12 14 16 18 PV
0
2
4
6
8
10 12 14 16 18 PV
Figure 5. Bacterial breakthrough curves (BTCs) of column experiments run with E. coli O157:H7 in (A) PW600, (B) PB600 (B), and E. coli K12 in (C) PW600, and (D) PB600. Error bars indicate the standard deviation measured from triplicates.
20 ACS Paragon Plus Environment
Page 21 of 36
Environmental Science & Technology
360
3.5 Collision Efficiencies
361
To quantify the influence of biochar addition to QS on bacterial retention, values of the collision
362
efficiencies of bacterial adsorption on QS or QS/biochar were calculated using the blocking
363
model (equations 1-3) (Table S4). The ability of the blocking model to fit the pseudo-steady state
364
portion of our data was judged by the R2 values which were on average (0.934±0.043, n=17).
365
When the blocking model was considered, the highest collision efficiency was found for E. coli
366
O157:H7 (α=7.4×10-4) when transported through QS mixed with PW600-OX biochar. The
367
lowest collision efficiency (α=8.22 ×10-6) was obtained when QS was used alone indicating that
368
amending sand with biochar is effective for bacterial attenuation. However, the highest collision
369
efficiency (α=4.03×10-4) for E. coli K12 was observed when PB350-UO biochar was added to
370
QS. The collision efficiencies estimated for E. coli O157:H7 were, on average, 1.7 times higher
371
than those estimated for E. coli K12. Observing higher collision efficiencies for the pathogenic
372
strain compared to the nonpathogenic strain in most of the cases can be justified by the fact that
373
E. coli K12 is more negatively charged than E. coli O157:H7 at the experimental conditions
374
(pH>7) (Figure 2). As such, electrostatic repulsion forces will be higher to biochar-amended
375
sand or pure sand and as such a higher energy barrier is to be overcome in order for E. coli K12
376
to attach to the porous media compared to E. coli O157:H7 as our data demonstrates.
377
4. Discussion
378
4.1 Effect of PW600-UO Biochar Concentration on Bacterial Transport in
379
Biochar Amended Sand Columns
380
Our results indicate that the collision efficiencies of both bacterial strain are dependent on
381
biochar concentration with collision efficiencies of the pathogenic strain, on average, being 1.7
21 ACS Paragon Plus Environment
Environmental Science & Technology
382
fold higher than those calculated for the nonpathogenic strain. The collision efficiencies
383
estimated for both bacterial types were linearly increasing with increase in PW600-UO biochar
384
application rates (Figure 6). As more biochar gets added, more sites will be available for
385
bacterial attachment. At low application rates of biochar, the surface area as well as pores
386
available for bacterial attachment are not sufficient. Even though biochar is characterized by a
387
high surface area that can be available for adsorption22, bacterial cells can’t access the area
388
provided by the micro-pores of the biochar due to their large size32. In comparison, when the
389
PW600-UO biochar concentration increased from 5% to 20%, the collision efficiencies increased
390
by 9.4 fold for E. coli O157:H7 and by 4.1 fold for E. coli K12 (Figure 6). This increase in
391
collision efficiencies can be attributed, in part, to additional sites available on the biochar surface
392
for bacterial attachment. However, if surface area was the only attributing factor to enhanced
393
attachment, we should have seen a similar fold increase in the attachment of E. coli K12 to that
394
observed for O157:H7, which was not the case. Attachment is affected by surface area,
395
physiochemical properties of QS, biochar and bacteria as well as by environmental conditions. It
396
has to be noted here that E. coli K12 are characterized by higher negative charge compared to E.
397
coli O157:H7 (Figure 2F). The higher negative charge is expected to result in repulsive forces
398
and lower attachments for E. coli K12 compared to E. coli O157:H7.
22 ACS Paragon Plus Environment
Page 22 of 36
Page 23 of 36
Environmental Science & Technology
0.0006
Collision efficiency (α)
0.0005
K12
0.0004
O157:H7
0.0003 0.0002 0.0001 0 0
5
10
15
20
25
Biochar Concentration (wt.%) 399 400 401 402 403 404 405 406 407 408
Figure 6 The relationship between the collision efficiencies calculated using a blocking model (equations 1-3) and biochar (unoxidized PW-600) concentration for E. coli O157:H7 (blackfilled circles) and E. coli K12 (yellow-filled circles). The dashed lines are the linear fits to the data with the gray and blue lines having slopes of 3×10-5 and 1 ×10-5, respectively and intercepts of zero. The R2 values are 0.986 and 0.934 for the E. coli O157:H7 and E. coli K12 strains, respectively.
409
Bacterial retention was strongly influenced by the biochar type investigated. Biochar preparation
410
conditions including feedstock origin, pyrolysis temperature, and post-pyrolysis oxidation all
411
affected the biochar type produced and its surface physiochemical properties and as such
412
bacterial attachment to it (Figure 2-6, Tables S1-S2). Different biochar preparations led to
413
variations in biochar surface charge, degree of hydrophobicity, surface roughness, hydrogen-
414
bonding capacity, and porous structure (Table 1). Below, the roles of biochar bulk and surface
415
properties on bacterial retention will be discussed.
4.2 Effect of Biochar on Bacterial Retention
416
The accumulation of ash content and mineral elements increased with the increase in
417
pyrolysis temperature and varied in accordance with the feedstock origin (Table 1). Our results 23 ACS Paragon Plus Environment
Environmental Science & Technology
Page 24 of 36
418
indicate that cations, such as Ca2+, Mg2+, K+, and Na+, were much higher in PB biochars
419
compared to PW biochars22. The reduction in E. coli O157:H7 transport through the columns of
420
QS mixed with PB-UO biochars irrespective of their pyrolytic temperature (Figures 4B and 5B)
421
can be attributed to electrostatic repulsive forces being screened by the positive increased metal
422
content of the biochar. Rivera-Utrilla et. al. (2001)33 observed that bacterial adsorption on
423
activated carbon increased in the presence of Ca2+ and Mg2+ and the adsorption was reduced
424
considerably after demineralization of carbon. The authors also suggested that bacterial cell
425
surface hydrophobicity, which is known to have a positive influence on cell-substrate adsorption,
426
is much higher in the presence of soluble metal cations due to the binding of metallic ions to the
427
outer membrane of bacteria. In particular, Ferris et al. (1986)34 explained the linkage between
428
cell hydrophobicity and soluble metals in terms of the diffuse-double layer theory which predicts
429
that water molecules become coordinated to site-bound counter-ions at charged surfaces. They
430
suggested that the increase in the degree of local order of liquid water results in an increase in the
431
hydrophobicity of the surface and a decrease of solvent entropy. Furthermore, several
432
researchers34–38 found that cations, especially divalent cations, neutralized the anionic charge of a
433
bacterial cell and made its planar charge distribution heterogeneous, thus decreasing repulsive
434
electrostatic forces. Note that aqueous cations may neutralize the surface charge of either the
435
bacterial cell or the biochar surface. Our results suggest that the high retention of E.
436
coli O157:H7 in PB-UO biochars amended sand columns compared to the QS control can be
437
linked to their high contents of divalent and monovalent cations.
438
A common source of the inhomogeneity of surface functionality of biochar that has been
439
recognized in the literature is the presence of various functional groups on the surface of biochar,
440
along with metal impurities. Our data indicate that mixing oxidized biochars, regardless of their
24 ACS Paragon Plus Environment
Page 25 of 36
Environmental Science & Technology
441
type and production temperature, into QS increased the collision efficiencies (α) by an average
442
48.9 and 4.6 for E. coli O157:H7 and E. coli K12 compared to QS alone, respectively. Results
443
were statistically significant (P
