A Method for Redox Mapping by Confocal Micro-X-ray Fluorescence

Subscriber access provided by Drexel University Libraries

Article

A method for redox mapping by confocal micro-X-ray fluorescence imaging: using chromium species in a biochar particle as an example Peng Liu, Carol J. Ptacek, David W. Blowes, Y. Zou Finfrock, Mark Steinepreis, and Filip Budimir Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b05718 • Publication Date (Web): 20 Mar 2019 Downloaded from http://pubs.acs.org on March 22, 2019

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 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 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.

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 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

1 2 3 4

A method for redox mapping by confocal micro-X-ray fluorescence imaging: using chromium

5

species in a biochar particle as an example

6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Peng Liua,b, Carol J. Ptacek,b, David W. Blowesb, Y. Zou Finfrockc,d, Mark Steinepreise, Filip Budimirb

aSchool

of Environmental Studies, China University of Geosciences, 388 Lumo Rd., Wuhan, Hubei, 430074, P. R. China bDepartment of Earth and Environmental Sciences, University of Waterloo, 200 University Ave. W., Waterloo, ON, N2L 3G1, Canada cCLS@APS sector 20, Advanced Photon Source, Argonne National Laboratory, 9700 S. Cass Avenue, Argonne, IL 60439, USA dScience Division, Canadian Light Source Inc., 44 Innovation Boulevard, Saskatoon, SK, S7N 2V3, Canada eStantec Consulting Ltd., 100-300 Hagey Blvd., Waterloo, ON, N2L 0A4, Canada

26

* Corresponding

author: Department of Earth and Environmental Sciences, University of Waterloo, 200 University Ave. W., Waterloo, ON, Canada N2L 3G1. Tel: +01 (519) 888 4567, ext. 32230 E-mail: [email protected]

1 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

28

Graphical abstract

29 30


Page 2 of 27

Page 3 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


31

Abstract

32

Redox mapping of solid-phase particles has been used for speciation mapping of near-surface

33

materials or within grains through the use of thin-sections without depth information. Here, a

34

procedure is presented for data collection and processing of depth-dependent redox mapping

35

within solid particles using confocal micro-X-ray fluorescence imaging (CMXRFI). The

36

procedure was applied to a biochar particle that was reacted with Cr(VI)-spiked water. The total

37

Cr distribution was first obtained at an above-edge energy of the K edge, and showed that Cr was

38

primarily distributed near the surface of the particle. Redox mapping was conducted at 33

39

representative energies and linear combination fitting (LCF) was performed for the 33 data

40

points from each pixel. The results indicate Cr(III) is the primary species with fractions ranging

41

from 0.6 to 1 and that this fraction is greater in the interior pixels of the particle than at the

42

surface; in contrast, the Cr(VI) fraction is greater at the surface than for interior pixels. The

43

results likely indicate Cr(VI) was first adsorbed and diffused into the biochar, and then reduced

44

to Cr(III). With more Cr(VI) adsorption and the exceedance of the reduction potential of the

45

biochar, remaining Cr(VI) was accumulated on the surface. The redox mapping method was

46

validated by micro-XANES (X-ray absorption near-edge structure) and XPS (X-ray

47

photoelectron spectroscopy) results. This demonstration indicates the developed method

48

combined with CMXRFI can be used to delineate the distribution of different oxidation states of

49

an element within an intact particle or layer.

50

Key words: redox/chemical mapping; confocal micro-X-ray fluorescence imaging; linear

51

combination fitting; biochar; chromium



52

Introduction

53

X-ray fluorescence (XRF) mapping is a well-established synchrotron-based method employed

54

for obtaining quantitative distribution of multiple elements by mapping intensities of a selected

55

region of interest (ROI) from a full XRF spectra. Chemical mapping is an extension of the XRF

56

mapping to produce images of a certain chemical species by exploiting chemically sensitive

57

differences in X-ray absorption near-edge structure (XANES) spectra1. Redox mapping, a

58

specific type of chemical mapping, maps the same element in different oxidation states. The

59

application of chemical mapping is limited compared to micro-XRF mapping and micro-

60

XANES, i.e., micro-XRF mapping is used to delineate the spatial distribution of a specific

61

element without speciation information.

62

Micro-XANES spectra are related to the chemical environment and the oxidation state of

63

the element of interest, and provide spatial information of different chemical moieties 2. Micro-

64

XANES has been applied to delineate the speciation of various elements after micro-XRF

65

mapping, including S3, V4, Fe5, As6, Se7, Sr8, Hg9,10, and Pb8. Different chemical species can be

66

shown by micro-XANES spectra at selected locations. One limitation of micro-XANES method

67

is that spectra can only be collected from a limited number of locations based on the XRF maps,

68

so key information can be missed1. Second, the sample might undergo significant chemical

69

changes due to the longer data collection time (>5 min) for micro-XANES spectra as well as the

70

high intensity of the micro-size incident beam. Because dwelling time for each step in chemical

71

mapping is between 100-300 µs, chemical mapping can potentially avoid these limitations.

72

Chemical mapping was pioneered by Kinney et al.11 and has been applied to various

73

samples2,12-18. Pickering et al.12 and Sutton et al.13 present procedures for data collection and

74

subsequent chemical mapping, and apply this approach to map Se speciation in water-saturated


Page 4 of 27

Page 5 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


75

sediments, respectively. Linear combination fitting (LCF) was applied, and the intensity of

76

references and chemical mapping were normalized to the intensity at an above-edge energy in

77

Sutton et al.13; furthermore, the number of energies for intensity collection was the same as the

78

number of references in Pickering et al.12 and Sutton et al.13. Oram et al.19 created Se oxidation

79

state distribution maps for rhizosphere soils by mapping at the energy of maximum absorbance

80

of each Se species. Mayhew et al.16 mapped Fe speciation and distribution in complex samples

81

using principal component analysis of redox maps at four energies. Dauphin et al.2 report the

82

chemical mapping of S in P. nobilis at four energies. The method provides key information on

83

the spatial speciation of different elements and will be important in the future study, but the

84

method is suitable for species for which the energy of maximum absorbance is distinctive for

85

each and the number of species is limited.

86

Full-field XANES, a combination of the advantages of micro-XANES and micro-XRF,

87

employs the full spectra with a large number of incident energies and pixels18,20-23. Full-field

88

XANES has been applied to the characterization of two-dimensional redox and speciation of Fe

89

in thin-sections of complex geological materials and ceramics18,20,21 and Ca speciation in human

90

bone23. An algorithm was developed for the alignment of hyperspectral images from full-field

91

XANES analysis22. However, chemical mapping with depth information for intact objects has

92

not been performed. Confocal micro-X-ray fluorescence imaging (CMXRFI) analysis can be

93

used to obtain elemental distributions along the cross section within various intact samples

94

through the addition of a confocal optics in front of the detector24, and CMXRFI has the potential

95

to provide chemical speciation with depth information25,26. Compared to conventional micro-

96

XRF mapping, advantages of CMXRFI include non-destructive characterization of samples27,

97

high signal-to-noise ratios (low detection limit)28,29, and minimal sample preparation24.



98 99

The most widely used approach for quantitative XANES analysis is LCF. The underlying principle of LCF is the additive nature of the spectra of each species in a sample to create the

100

overall spectrum for that sample. LCF is more suitable and objective for speciation calculation of

101

diluted samples compared with spectral deconvolution method30. Limited studies related to

102

chemical mapping employed LCF due to the limited number of energies used in map collection.

103

This study presents procedures for data collection and processing to construct redox maps

104

using CMXRFI. A biochar particle loaded with chromium (Cr) is used as an example to show the

105

process of data collection and energy selection for redox mapping. Data are processed in

106

MATLAB® using LCF, and the fraction and intensity of each Cr species plotted. The aim is to

107

facilitate the application of combining CMXRFI and redox mapping to delineate element-

108

specific spatial distributions of different species.

109

Methods

110

Batch experiment

111

An oak-wood biochar (rejects from industrial products) pyrolyzed at ~700 °C was obtained from

112

Cowboy Charcoal Inc., USA. The rejects were crushed and sieved, and the 0.5-2 mm particles

113

were washed 3 times using Milli-Q (Millipore) water to detach impurities and fine particles and

114

used for batch experiments. The experiment was performed by adding 0.4 g of biochar to 40 mL

115

of K2Cr2O7-spiked water (50 mg L-1 as Cr(VI)) at an initial pH of 2 for 24 h in an anaerobic

116

chamber (Coy Laboratory Products, USA). The pH was adjusted to 2 by adding hydrochloric

117

acid. Triplicate experiments were performed and mean values were reported.

118

The aqueous phase was filtered through 0.20-µm Supor membrane filters (Acrodisc, UK)

119

after 24 hours for analysis of pH, Cr(VI), and total Cr (tCr). The pH was determined using a

120

combination electrode (Orion Ross 815600). Cr(VI) concentrations were measured using a Hach


Page 6 of 27

Page 7 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


121

DR2800 spectrometer using the 1,5-diphenylcarbohydrazide method (HACH Method 8023). tCr

122

concentration was measured using inductively coupled plasma-optical emission spectrometry

123

(Thermo iCAP 6500). Cr(III) concentrations were calculated as the difference between

124

concentrations of tCr and Cr(VI). The content of Cr loaded to biochar was calculated based on

125

the mass balance. The biochar particles were filtered out and immediately rinsed three times

126

using ultra-pure water, freeze-dried for 24 h, and stored under anaerobic conditions. The particles

127

were transferred to the Advanced Photon Source (APS) in a portable anaerobic container (BD

128

BBL™ GasPak™) for CMXRFI analysis.

129

CMXRFI, micro-XANES, and redox mapping

130

CMXRFI analysis of a biochar particle loaded with Cr was performed at Beamline Sector 20ID

131

of APS, Argonne National Laboratory, IL, USA. The biochar particle was in a rectangle-like

132

shape (~1.0×1.8×0.8 mm, Fig. S-1) and a particle density of ~1.34 g cm-3. Epoxy (Devcon

133

14250, USA) was used to mount the particle on a quartz slide. The slide was attached to a stage

134

oriented at 34° to the incident beam.

135

CMXRFI data were obtained for Cr intensity (If) at an incident beam energy (E) of 7080

136

eV, which is greater than its binding energy of 5.989 keV at the K edge (Fig. 1). The

137

monochromatic incident beam (intensity as I0) was focused with KB mirrors down to ~2×2 μm2.

138

Full spectra of X-ray fluorescence were collected with a single element Si-drift detector. The

139

confocal geometry was completed by installing a germanium collimating channel array optic unit

140

in front of the detector. The depth resolution of the optic was ~2 μm, the confocal volume was 8

141

μm3, and the working distance was ~1.5 mm. More details of the optics are provided by Liu et

142

al.29. Full XRF spectra were acquired by rastering the particle using a high precision stage

143

(Aerotech Inc.) in the horizontal direction (xy plane) with a step size of 5 μm and a dwelling time



144

of 0.2 s. The mapping area was 250×400 (depth×width) µm2 and the data acquisition time was

145

~17 min. Cr Kα fluorescence line absorption length was ~1400 µm for the biochar particle

146

calculated by Hephaestus 31, which indicates the mapping depth (250 µm) did not result in

147

complete absorption of Cr Kα edge XRF. The imaging area was located in the cenral area of the

148

particle in the horizontal plane. The intensity of each pixel was corrected by considering the

149

geometry of the CMXRFI setup and the heterogeneity of the particle using the elemental

150

composition of the biochar (Table S-1) and a method presented by Liu et al.29.

151

Micro-XANES spectra were collected at points of interest determined from the full

152

spectra maps (Fig. 1), including Cr-enriched and less enriched areas as well as the surface and

153

interior of the particle. Confocal micro-XANES spectra were recorded at the Cr K-edge using the

154

same CMXRFI setup. The scan range was -150 to 528 eV. The monochromater energy was

155

calibrated with Cr foil. Five to seven repeated scans were collected for each selected point of

156

interest area with 2-µm displacement (Fig. S-2) between scans to decrease radiation exposure

157

and no obvious changes were observed. Spectra for calibration standards for Cr foil, Cr(OH)3,

158

Cr2O3, Cr(acetate)3, CrO3, K2Cr2O7, and K2CrO4 were obtained from previous studies32,33 at

159

Sector 13 bending magnet beamline of the APS and were used for LCF.

160

Redox mapping data were collected at selected energies (q=33 in total), which covered

161

the key features of the micro-XANES spectra, and additional energies were selected at positions

162

with dramatic changes in the absorbance (for example: peaks, valleys, and edges). The selected

163

energies were determined using Monte Carlo simulations in MATLAB described in the

164

Supporting Information (SI). Briefly, a small number of energy points were randomly selected in

165

combination with the energies from key features; a cubic spline interpolation method was

166

applied to obtain the absorbance for the non-selected energy points in the full spectra; Monte


Page 8 of 27

Page 9 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


167

Carlo simulation was applied to obtain the minimum R factor for the specific number of energy

168

points; the number of points increased one by one until the R factor decreased to a threshold

169

(2*10-4 for this study). The calculated minimum number of energy points was 23 (Table S-2; Fig.

170

S-3), whereas 33 points were selected in this study to include more information of the spectra.

171

Redox mapping data were collected with a step size of 2 μm for a 140×80 µm2 area. The

172

dwelling time was 0.2 s and the collection time at each energy was ~10 min with a total time of

173

~5 hrs for the 33 energies. The repositioning was accurate from map to map based on the

174

potassium (K) distribution maps at the 33 incident energies. The intensities at 33 energies were

175

corrected in the same manner with the intensities at 7080 eV.

176

Data processing and plotting

177

The confocal micro-XANES spectra at each location were corrected for sample attenuation along

178

the beam path. The intensities of incident beam (I0) and emitted XRF (If) at each energy point of

179

a scan from the specific pixel were corrected using information obtained from the intensity

180

correction for the whole imaging area at 7.08 keV. The information included the path length,

181

density distribution, and total cross section calculated from the elemental distribution along the

182

beam path. The process was embedded into the program developed by Liu et al.29. The corrected

183

spectra were then merged and normalized (µ(E)) with the normalization order set to 1 to simplify

184

the data analysis process. The normalization was performed in Athena31 and MATLAB.

185

LCF fitting was performed near the XANES (full spectra from selected points) region

186

from -20 to 20 eV relative to 6003 eV for µ(E) and the derivative of µ(E). The coefficient (α)

187

was forced between 0 and 1, and the sum forced to be 1. The LCF fitting strategy followed the

188

methods delineated by Voegelin et al.34 and Jamieson-Hanes et al.35. Sample spectra were first

189

fitted using all of the available reference spectra, then the number of references decreased until



190

the goodness-of-fit parameters (χ2, χv2, and R; SI) increased by at least 30%. The match between

191

characteristic peak positions in the sample and references was also considered in reference

192

selection. The spurious inclusion of references was likely avoided through these procedures. The

193

error of the LCF procedure is expected to be

A Method for Redox Mapping by Confocal Micro-X-ray Fluorescence

Recommend Documents