Advances in agronomy, volume 133

VOLUME ONE HUNDRED AND THIRTY THREE

ADVANCES IN
AGRONOMY


ADVANCES IN AGRONOMY
Advisory Board

PAUL M. BERTSCH

RONALD L. PHILLIPS

KATE M. SCOW

LARRY P. WILDING

University of Kentucky

University of California, Davis

University of Minnesota
Texas A&M University

Emeritus Advisory Board Members

JOHN S. BOYER

University of Delaware

EUGENE J. KAMPRATH

North Carolina State University

MARTIN ALEXANDER
Cornell University


VOLUME ONE HUNDRED AND THIRTY THREE

ADVANCES IN
AGRONOMY

Edited by

DONALD L. SPARKS
Department of Plant and Soil Sciences
University of Delaware
Newark, Delaware, USA

AMSTERDAM • BOSTON • HEIDELBERG • LONDON
NEW YORK • OXFORD • PARIS • SAN DIEGO
SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO
Academic Press is an imprint of Elsevier


Academic Press is an imprint of Elsevier
225 Wyman Street, Waltham, MA 02451, USA
525 B Street, Suite 1800, San Diego, CA 92101-4495, USA
125 London Wall, London EC2Y 5AS, UK
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK

First edition 2015
Copyright Ó 2015 Elsevier Inc. All Rights Reserved.
No part of this publication may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, recording, or any information storage and
retrieval system, without permission in writing from the publisher. Details on how to seek
permission, further information about the Publisher’s permissions policies and our
arrangements with organizations such as the Copyright Clearance Center and the Copyright
Licensing Agency, can be found at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by
the Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and
experience broaden our understanding, changes in research methods, professional practices,
or medical treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in
evaluating and using any information, methods, compounds, or experiments described
herein. In using such information or methods they should be mindful of their own safety and
the safety of others, including parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors,
assume any liability for any injury and/or damage to persons or property as a matter of
products liability, negligence or otherwise, or from any use or operation of any methods,
products, instructions, or ideas contained in the material herein.
ISBN: 978-0-12-803052-3
ISSN: 0065-2113
For information on all Academic Press publications
visit our website at />

CONTRIBUTORS
Silvia M. Alfieri
Institute for Mediterranean Agricultural and Forest Systems (CNR-ISAFOM),

Ercolano, Italy
Dionisio And
ujar
Institute of Agricultural Sciences- CSIC, Madrid, Spain
Johan Arvidsson
Department of Soil & Environment, Swedish University of Agricultural Sciences,
Uppsala, Sweden
Fernando Auat Cheein
Autonomous and Industrial Robotics Research Group (GRAI), Advanced Center of
Electrical and Electronic Engineering (AC3E), Department of Electronic Engineering,
Universidad Técnica Federico Santa María, Valparaíso, Chile
Angelo Basile
Institute for Mediterranean Agricultural and Forest Systems (CNR-ISAFOM),
Ercolano, Italy
Antonello Bonfante
Institute for Mediterranean Agricultural and Forest Systems (CNR-ISAFOM),
Ercolano, Italy
Johan Bouma
Soils Department, Wageningen University, The Netherlands
Henrik Breuning-Madsen
Department of Geography and Geology, University of Copenhagen, Copenhagen,
Denmark
David Chevrier
Canadian Light Source Inc., Saskatoon, SK, Canada
Francesca De Lorenzi
Institute for Mediterranean Agricultural and Forest Systems (CNR-ISAFOM),
Ercolano, Italy
James J. Dynes
Canadian Light Source Inc., Saskatoon, SK, Canada
Alexandre Escola

Research Group on AgroICT & Precision Agriculture – Universitat de Lleida, Lleida, Spain
Adam W. Gillespie
Canadian Light Source Inc., Saskatoon, SK, Canada; Department of Soil Science,
University of Saskatchewan, Saskatoon, SK, Canada
Kodigal A. Gopinath
ICAR-Central Research Institute for Dryland Agriculture, Hyderabad, India

vii

j


viii

Contributors

Eduard Gregorio
Research Group on AgroICT & Precision Agriculture – Universitat de Lleida, Lleida, Spain
Mogens H. Greve
Department of Agroecology, Aarhus University, Tjele, Denmark
Vijay S. Jakkula
ICAR-Central Research Institute for Dryland Agriculture, Hyderabad, India
Thomas Keller
Department of Natural Resources & Agriculture, Agroscope, Z€
urich, Switzerland;
Department of Soil & Environment, Swedish University of Agricultural Sciences,
Uppsala, Sweden
Rattan Lal
Carbon Management and Sequestration Center, The Ohio State University, Columbus,
OH, USA

Mathieu Lamandé
Department of Agroecology, Aarhus University, Tjele, Denmark
Piero Manna
Institute for Mediterranean Agricultural and Forest Systems (CNR-ISAFOM), Ercolano,
Italy
Joan Masip
Research Group on AgroICT & Precision Agriculture – Universitat de Lleida, Lleida, Spain
Eugenia Monaco
Institute for Mediterranean Agricultural and Forest Systems (CNR-ISAFOM),
Ercolano, Italy
Derek Peak
Department of Soil Science, University of Saskatchewan, Saskatoon, SK, Canada
Courtney L. Phillips
Department of Soil Science, University of Saskatchewan, Saskatoon, SK, Canada
Jasti V.N.S. Prasad
ICAR-Central Research Institute for Dryland Agriculture, Hyderabad, India
Lluís Puigdomenech
Research Group on AgroICT & Precision Agriculture – Universitat de Lleida, Lleida, Spain
Thomas Z. Regier
Canadian Light Source Inc., Saskatoon, SK, Canada
Joan R. Rosell-Polo
Research Group on AgroICT & Precision Agriculture – Universitat de Lleida, Lleida, Spain
Kanwar L. Sahrawat
International Crops Research Institute for the Semi Arid Tropics, Hyderabad, India
Per Schjønning
Department of Agroecology, Aarhus University, Tjele, Denmark


Contributors

ix

Alok K. Sikka
Indian Council of Agricultural Research, New Delhi, India
Asko Simojoki
Department of Food and Environmental Sciences, University of Helsinki, Helsinki, Finland
Rajbir Singh
Indian Council of Agricultural Research, New Delhi, India
Cherukumalli Srinivasa Rao
ICAR-Central Research Institute for Dryland Agriculture, Hyderabad, India
Matthias Stettler
Bern University of Applied Sciences, School of Agricultural, Forest & Food Sciences HAFL,
Zollikofen, Switzerland
Jan J.H. van den Akker
Alterra, Wageningen University and Research, Wageningen, The Netherlands
Bandi Venkateswarlu
Vasantrao Naik Marathwada Krishi Vidyapeeth, Maharashtra, India
Surinder M. Virmani
National Academy of Agricultural Sciences, New Delhi, India


PREFACE
Volume 133 contains five first-rate reviews dealing with contemporary
topics important in the crop and soil sciences. Chapter 1 is a comprehensive
review on the advances that have occurred in the use of synchrotron-based
soft X-ray spectroscopy to study biogeochemical processes of important
light elements such as carbon, nitrogen, and phosphorus in soils. Chapter 2
introduces a new hybrid land evaluation system to assess climate change
effects on the suitability of an agricultural area for maize production.
Chapter 3 is a timely review on progress in using structured light sensors

in precision agriculture and livestock farming. Chapter 4 covers the potential
and challenges of rainfed farming in India including features of rainfed
ecosystems and rainfed crops and cropping systems. Chapter 5 presents a
Driver-Pressure-State-Impact-Response (DPSIR) analysis and risk assessment for soil compaction from a European perspective.
I am grateful for the authors’ fine contributions.
Donald L. Sparks
Newark, Delaware, USA

xi

j


CHAPTER ONE

Advances in Using Soft X-Ray
Spectroscopy for Measurement
of Soil Biogeochemical Processes
Adam W. Gillespie*, x, 1, Courtney L. Phillipsx, James J. Dynes*,
David Chevrier*, Thomas Z. Regier* and Derek Peakx
*Canadian Light Source Inc., Saskatoon, SK, Canada
x
Department of Soil Science, University of Saskatchewan, Saskatoon, SK, Canada
1
Corresponding author: E-mail:

Contents
1. Introduction
2. Detector Advancements
3. Slew Scanning of Radiation-Sensitive Solids

3.1 Carbon K-Edge Analysis
4. Soft X-ray Liquid Cells
4.1 Current Applications and Future Prospects
References

2
7
12
16
22
29
30

Abstract
Light elements are particularly important in biogeochemical processes. These include
organic matter components and macronutrients (C, N, O, S, P), micronutrients (Na,
Mg, K, Mg), mineral elements (Si, Al), and transition metals. Determining the chemical
speciation of these light elements in environmental samples is important for understanding bioavailability, decomposition, contamination mobility, and nutrient cycling.
Soft X-ray absorption spectroscopy is a useful tool available to probe the chemistry
of atoms important in biogeochemical processes. X-ray absorption spectroscopy (XAS)
probes the local bonding and coordination environment of these elements in whole
samples. Bulk XAS techniques permit for high throughput, the study of whole soils,
and high sampling density. These analyses are complementary to X-ray transmission
microscopy techniques which are limited by low throughput, thin particles
(<100 nm), and low sampling density. In many projects, these bulk XAS measurements
may be essential to understanding large-scale processes in soils such as the global
C cycle.
Despite these important applications, bulk soft XAS has not been extensively
applied to environmental samples until recently. The primary reasons for this gap is
the lack of beamline endstations that are suitable for “dirty” samples and the technical

challenges related to acquiring and normalizing spectra from dilute samples. Many of
these technical challenges have now been overcome through the development of
Advances in Agronomy, Volume 133
ISSN 0065-2113
/>
© 2015 Elsevier Inc.
All rights reserved.

1

j


2

Adam W. Gillespie et al.

energy-resolving detectors, proper detector positioning, and development of liquid cell
applications. Technical developments and recent applications will be presented,
showing how bulk soft X-ray XAS is now positioned to contribute significantly to
advancing the characterization of soils and environmental samples.

1. INTRODUCTION
The field of biogeochemistry describes how the biotic and abiotic
worlds interact. More specifically, it studies the chemical and biochemical
processes in which elements are transformed and moved within and between living and nonliving parts of the ecosystem. On a global scale, the
cycles of major elements including carbon, nitrogen, oxygen, and iron are
linked with global climate change, food security, water security, and economic development (Godfray et al., 2010; Lal, 2009; McBratney et al.,
2014). Understanding the forms, pool sizes, and transformation rates of these
elements is critical to assessing their behavior on a global scale. The availability and cycling of nutrients is determined by the interaction of physical,

chemical, and biological processes in an ecosystem. This interaction of processes is important as it determines the forms, transformations, and ultimate
fate of nutrients in a given system. Therefore, an assessment of their chemistry is a key in determining how these elements will react, cycle, and persist
in the environment.
Research that targets biogeochemical cycling in terrestrial ecosystems is a
key component to understanding global systems. Examples of biogeochemical processes include global biological processes of photosynthesis, nitrogen
fixation, and soil organic matter decomposition, as well as critical redox processes of nitrification, denitrification, iron reduction, sulfate reduction, and
the myriad of transformations associated with phosphorus cycling. Indeed,
soil carbon pools are greater than both the atmospheric and biomass carbon
pools combined (Lal, 2007; Stevenson and Cole, 1999), and soils contain the
majority of reactive N as organic forms in soil organic matter (Olk, 2008;
Stevenson, 1996). Soils also harbor an enormously diverse and dense microbial population which is ultimately responsible for organic matter transformations. The composition of soil organic matter and the composition of
the soil microbial community are tightly coupled. They participate in a
dynamic system where organic matter composition is constantly changing
in response to biological processes and, in turn, microbial community
composition changes in response to available substrates and nutrients. In


Advances in Using Soft X-Ray Spectroscopy for Measurement of Soil Biogeochemical Processes

3

addition, there is growing evidence describing the role of soil minerals as
reactive surfaces that act as reaction catalysts, microbial niches, and adsorbents/stabilizers of organic molecules.
Light or low atomic number (Z) elements are particularly important in
biogeochemical processes. Light elements (C, N, O, P, and S) are the primary atomic building blocks of all biological systems, and are implicated
in global climate change, acidification, and eutrophication. Light basic cations (Na, Mg, K, Ca) are essential nutrients for plant growth, and minerals
are primarily composed of Al, Si, O, and Fe. Finally, third row transition
metals comprise essential micronutrients and contaminants of interest. An
ideal analytical approach for studying biogeochemical processes in environmental samples would therefore target the chemistry of these low Z elements and 3rd row transition metals.
Among the tools available to probe biogeochemical processes, synchrotron-based X-ray absorption spectroscopy (XAS) is an especially promising

choice. XAS is a technique that uses X-rays to probe the local electron structure, thus elucidating the speciation and bonding environment of an
element. Since the core electron energies of different elements are unique,
specific elements can be probed in a complex sample by scanning the X-ray
beam energy across the binding energies of the core electrons in the
elements of interest. Promotion of these core level electrons into unoccupied (i.e., valence) electron orbitals occurs when the incident X-ray energy
coincides with an accessible transition energy, resulting in the resonant
absorption of the X-ray photons. This resonant absorption results in a reduction in the penetration depth of the incident beam. Observations of subtle
changes in the resonant absorption energies can be used to differentiate
oxidation states, bonding environments, and neighboring atoms.
X-rays in the “soft” X-ray region (250e2500 eV, or wavelength 5 to
0.6 nm) access the chemistry of lighter elements of biogeochemical interest.
These include the K-edges (i.e., excitation of the 1s electrons) of elements
ranging from C to Si and the L2,3 edges (i.e., excitation of the 2p electrons)
from K to Se (Figure 1).
Soft X-rays provide a unique way to probe the electronic structure of
C, N, and O at their K-edges in environmental samples, and is a powerful
complement to established and widely used NMR and IR spectroscopy.
XAS offers advantages because it is sensitive to all isotopes of a particular
element, and is not limited by 13C or 15N abundances. Indeed, for analysis
of N chemistry in environmental samples, XAS currently offers the only tools
which are sensitive to most N-species and largely free of technique-induced


4

Figure 1 Biogeochemically significant elements and core electron shells accessible by soft X-ray spectroscopy. Black lines are K-edges (1s),
red (light grey in print versions) lines are L-edges (2p), and blue (grey in print versions) line is N-edge (4f).
Adam W. Gillespie et al.

Advances in Using Soft X-Ray Spectroscopy for Measurement of Soil Biogeochemical Processes

5

sample modifications. It is also not necessary, to enrich organic matter concentrations by separating it from the mineral fraction or to digest samples in
HF to remove minerals, paramagnetic materials, or other components for
the analysis to proceed.
The soft X-ray region also accesses the L-edges of the transition metals.
For first row transition metals, soft X-ray absorption can result in the excitation of a 2p electron into unoccupied 3d orbitals. In transition metals, the
electrons in the 3d orbital participate directly in ligandemetal complexation
(de Groot and Kotani, 2008). Probing the L-edge of organometal complexes therefore provides direct information on the bonding and coordination environment of those complexes. The energy of the absorption
resonance can be used to determine the energy of the lowest unoccupied
molecular orbitals (LUMOs) of the complex. From ligand field theory, all
ligands behave as either sigma donors, pi acceptors, or pi donors, and their
complexation with a metal center will change the relative (splitting) and
absolute (ELUMO) positions of the LUMOs. These effects can be measured
directly by the intensity, position, and number of peaks in the L-edge XAS
spectrum of the metal-ligand complex.
Soil biogeochemical processes occur at all levels of spatial variability,
from the global scale to the nanoscale. Indeed, the quality of biogeochemical
research depends not only on the techniques available, but also upon the
sampling strategy, study location, and spatial resolution in many dimensions,
including time. XAS offers the flexibility to provide analytical information at
many spatial scales, and this depends on a particular beamline design. For
example, most ecological soil sampling is conducted to understand the
role of spatial information such as landscape position/topography (soil
catenas) or vertical soil horizons (cores/sampling from pits). Bulk soft Xray XAS measurements can probe a w1 Â 0.1 mm area in 10e100 subsamples from such a sample set and provide quantitative speciation of an element
that fits nicely into field research. However, there may not always be a
unique solution to bulk speciation in heterogeneous samples such as soils;
having mm- or nm-scale spectromicroscopy capabilities with a soft X-ray

microprobe (10 mm pixel) or scanning transmission X-ray microscope
(STXM, 20e50 nm pixel) can provide additional information into the variability of the sample and improve confidence in quantitative fitting
approaches. In the example of Figure 2, the bulk Fe L-edge X-ray measurements on the right were performed with a beam roughly 1 mm  0.1 mm,
the microprobe image in the lower portion shows the distribution of Fe over
the same sample at 0.5 mm  0.5 mm scale with 5 mm resolution, and the


6

Adam W. Gillespie et al.

Field-scale
Bulk sample speciation
(mm spot size)

Speciation-based
Management Decision

Fe L-edge

STXM Imaging
(25 nm spot size)
720

Microprobe Imaging
(5 μm spot size)

Figure 2 Soft X-ray synchrotron techniques are relevant at many scales. Many field
samples can be analyzed on bulk beamlines, microscale distribution can be accessed
at microprobe beamlines, and nanoscale distribution and speciation are accessed at

X-ray microscopy beamlines. In this experiment, the bulk Fe L-edge X-ray measurements on the right were performed with a beam roughly 1 mm  0.1 mm. The microprobe (bottom) shows the distribution of Fe over the same sample at 0.5 mm  0.5 mm
total image size with 5 mm resolution. The scanning transmission X-ray microscope
(STXM) image (left) covers a 5 mm  5 mm spot (roughly 1 pixel of the microprobe) in
detail sufficient to observe microbial colonization of defect sites on Fe (III)-rich particles.

STXM map covers a 5 mm  5 mm spot (roughly 1 pixel of the microprobe)
in detail sufficient to observe microbial colonization of defect sites on Fe
(III)-rich particles.
Synchrotron-based studies are advantageous because they require little
sample preparation before measurement. There are, however, several challenges that have limited the adoption of this technique for soil and environmental samples. First of all, chemical speciation on environmental samples
should ideally be conducted at ambient moisture and gas pressure/composition, and at dilute concentrations. The current widespread practice of
bulk soft X-ray spectroscopy requires samples to be under vacuum, which
require them to be dry and pulverized. Next, elements in environmental
samples are typically present at concentrations nearing detection limits and
model systems with higher concentrations suffer from saturation effects.


Advances in Using Soft X-Ray Spectroscopy for Measurement of Soil Biogeochemical Processes

7

Finally, radiation damage is a significant and well-documented problem at
soft X-ray energies (Beetz and Jacobsen, 2003; Wang et al., 2009;
Zubavichus et al., 2004). In this document, we describe solutions to these
challenges which have been implemented at the bulk soft X-ray beamline,
11ID-1 SGM (Spherical Grating Monochromator) at the Canadian Light
Source (CLS) (Regier et al., 2007a,b). What follows describes the use of
energy-resolving detectors at soft X-ray energies, the development of a
flow-through liquid cell, the application of fast (slew) scanning, and the
spectral acquisition and normalization procedures specific to carbon

measurements.

2. DETECTOR ADVANCEMENTS
XAS is most often measured using transmission, total electron yield
(TEY) by measuring the drain current through the sample (Gudat and
Kunz, 1972), or using total fluorescence yield (TFY) using either microchannel plates or solid state diode detectors (Jaklevic et al., 1977). Transmission methods are not generally suitable for bulk measurements in the soft
X-ray region because the attenuation of the photons at these energies
requires samples to be uniform and less than 100 nm thick for measurements
at the C K-edge, and ca. 2 mm at the Si K-edge. TEY and TFY are more
often used because the analyst is not restricted by sample thickness. TEY
and TFY are measurements of secondary decay processes which operate
on the principle that, after an electron core hole is produced, the core
hole will decay with the emission of either an Auger electron or a fluorescence photon. As the incident X-ray energy is scanned through an absorption resonance, the penetration depth is reduced and more of the secondary
electrons or fluorescence photons are produced nearer to the surface of the
sample and will exit the sample and make it to the detector. Thus, under
ideal conditions, the measured intensity of the secondary emission will be
proportional to the linear attenuation coefficient.
Electron yield is a surface sensitive detection method, whereas TFY is
considered a bulk-sensitive detection method. Electrons are much larger
than photons, and have an escape depth of 10 nm before they are reabsorbed
by surrounding atoms. Fluorescent photons however, have an escape depth
on the order of 100 nm (Frazer et al., 2003; Katsikini et al., 1997). Electron
yield is also sensitive to the conductivity of the sample material. Many environmental materials, including soil, contain some insulating components,


8

Adam W. Gillespie et al.

which will restrict the movement of charge through a sample, causing

charge buildup and release (sample charging). The measured current will
be dependent on exactly how the sample was mounted and what components of the sample happen to rest on a conducting substrate material. Fluorescence yield does not suffer from charging effects, making it more suitable
for the measurement of insulating sample materials.
TFY measurements, however, have several limitations. These measurements are commonly obtained using either microchannel plates or solid state
diode detectors, both of which output a signal proportional to the power
deposited in a detection medium. As the TFY is comprised of both the resonant and nonresonant emission from the sample, it has two major drawbacks. The first is that, for a dilute system (where saturation effects are
small) the resonant emission from the target element is very weak compared
to the nonresonant emission (background) from the matrix. This can
dramatically reduce the concentration limits of the detection method. The
second drawback is that the background emission from the matrix can
exhibit an energy dependence at resonance due to a change in the penetration depth of the incident photons (Achkar et al., 2011). This effect can
result in a severe reduction in fluorescence intensity that can create a subbackground in the pre-edge of some samples that depends on the concentration of the sample. Finally, in grating-based beamlines, higher order light
is transmitted, and the TFY will contain spectral contributions from
elements excited by this light. As a specific example, this is especially problematic in C measurements of soil and environmental samples in which
oxygen-containing clay minerals are present. Measurement at the C Kedge (at 285 eV) in natural systems will always have artifacts in TFY arising
from the C K-edge coincidence with the second order O K-edge
(at 260 eV).
Another limitation of fluorescence yield-based absorption spectroscopy
is saturation of the fluorescence signal due to overabsorption, commonly
referred to as self-absorption (note that this term is somewhat misleading
as it suggests that the saturation arises due to the reabsorption of the fluorescence signal from the sample when, in fact, this is not true). This type of
saturation introduces large distortions on the measured fluorescence spectra
when the attenuation coefficient of the element of interest becomes comparable to the total attenuation coefficient of the sample. In these conditions,
the resonant absorption of the incident beam happens so close to the surface
of the sample that all of the fluorescence photons are able to escape from the
surface of the sample, resulting in a loss of contrast in the measured spectrum.


Advances in Using Soft X-Ray Spectroscopy for Measurement of Soil Biogeochemical Processes

9

These main complications with TFY measurements can be overcome
with the use of an energy dispersive detector, capable of resolving the resonant, nonresonant, and higher order emission from the sample. This allows
for the measurement of the partial fluorescence yields (PFY), which are
free from background distortions and much more sensitive than total yield
methods (Figure 3). Silicon drift detectors (SDD) provide such energyresolved detection, and have been used routinely in the hard X-ray region
for many years. More recently, improvements in thin-window manufacturing
has resulted in the extension of their use to soft X-ray detection.
On the SGM beamline at the CLS, the SDD’s are configured for two
operating modes: pulse counting mode and full spectrum acquisition. In
pulse counting mode, a region of interest is specified in the detector hardware to electronically restrict the energy range of reported photons at the
time of data acquisition. In this scenario, the detector reports only photons
arising from fluorescence of a particular atom. This setup can be optimized
for faster counting because it does not require the complete readout of the
detector electronics at each step of the scan. Figure 3 shows the X-ray fluorescence spectrum of a soil when excited by 2000 eV photons. The fluorescence produced by single elements in the mixture can be isolated during
analysis, offering the ability for rapid data acquisition and lowered detection
limits.
In full spectrum acquisition, the entire X-ray fluorescence spectrum is
read out and recorded for each excitation energy across the absorption
edge being probed. In these experiments, the fluorescence associated with

Figure 3 Fluorescence spectrum of a soil excited by 2000 eV photons as measured with
a silicon drift detector. Detector output can be set to an emission energy range to
isolate the fluorescence of a single element, allowing for rapid data acquisition with
minimal background interference.


10

Adam W. Gillespie et al.

resonant atoms can be isolated to lower detection limits of dilute materials.
In the example illustrated in Figure 4, copper nitrate is present at mM concentrations in a liquid sample, and TFY is unable to distinguish resonances
from the strong oxygen background (note that liquid cell construction
and spectroscopy will be detailed later in this work). By using an energyresolved detector, the fluorescence from Cu can easily be extracted from
an excitation/emission map by applying a software filter to the SDD output.
In addition, using an energy discriminating detector, the inverse partial
fluorescence yield (IPFY) method can be applied (Achkar et al., 2011).
IPFY is a method of detection which is bulk sensitive and does not suffer
from saturation effects. This method consists of scanning across the absorption edge of the element of interest, but monitors the change in fluorescence
intensity of another (spectator) element which is present in the mixture. One
important requirement for this method to work is that the spectator element
must have a core hole binding energy below that of the element of interest.
IPFY operates on the principle that the nonresonant emission will be
inversely proportional to the total attenuation coefficient of the sample.
For a full description and mathematical proof, see Achkar et al. (2011).
This method resolves the problem of saturation because it is based on the

Figure 4 Silicon drift detector output from a 100 mM copper (II) nitrate solution sample
across the Cu L-edge. Both the O and Cu emissions are labeled in the excitationemission matrix (above). Analysis across the Cu edge includes strong emission line
from O, which dominates the total fluorescence, shown in blue in the lower panel.
Copper spectrum (in red (black in print versions)) is accessible by isolating only the
Cu emission line.


Advances in Using Soft X-Ray Spectroscopy for Measurement of Soil Biogeochemical Processes 11

nonresonant scattering intensity from the sample, it is directly proportional
to the linear attenuation coefficient as it does not contain distortions from

the modulation of the resonant scattering cross section.
IPFY is applicable for the measurement of iron and other transition metal
oxides because oxygen is always present in environmental samples and can
act as a suitable spectator. This technique is especially suited to exploring
the chemistry of iron oxides and oxyhydroxides. Peak and Regier (2012)
used IPFY to describe the structure of ferrihydrite (FeOOH), ultimately
reporting a model in which ferrihydrite contained some tetrahedral iron
within a structure of predominantly octahedral iron. This experiment was
conducted at the Fe L2,3 edge, on solid state powdered samples, and the
use of energy-resolving detectors was critical to the success of the study.
Figure 5 shows results from the analysis of ferrihydrite at the Fe L-edge.

Figure 5 Structure of ferrihydrite shown to contain some tetrahedral Fe as determined
using inverse partial fluorescence yield. Reprinted with permission from Peak and Regier
(2012). Copyright 2012 American Chemical Society.


12

Adam W. Gillespie et al.

The Fe L-edge is excited above ca. 705 eV. These energies are above the
binding energy for O, and so strong fluorescence from O in the sample is
detected. These two emission phenomena can be resolved with an
energy-resolving SDD, as seen in Figure 5. These correspond to the Fe
La and Lb emission (625À725 eV) and O Ka emission (450À550 eV) lines.
The importance of this emission resolution is evident when looking at the
TFY in Figure 5. In this case, this plot is the sum of all Fe and O fluorescence
at a particular excitation energy. This is analogous to TFY as reported by
microchannel plate or diode-type detectors. The ability to separate the fluorescence from Fe and O shows that the total fluorescence signal is a convolution of increasing Fe fluorescence and decreasing O fluorescence. Key to

the interpretation of this data is the principle that fluorescence from Fe is
distorted by saturation, whereas fluorescence from O is not. The inverse
of the PFY of O in the sample is proportional to the total attenuation
coefficient, and is thus a high quality Fe spectrum that is free from saturation
effects.

3. SLEW SCANNING OF RADIATION-SENSITIVE SOLIDS
Once purported as a nondestructive method (Vairavamurthy and
Wang, 2002), XANES studies have shown degradation and transformation
of some reference compounds through beam exposure over time,
particularly when using high photon flux undulator beamlines (Cody
et al., 2009; Leinweber et al., 2007; Zubavichus et al., 2004). These
problems associated with radiation-induced decomposition also are
compounded at soft X-ray energies because X-ray absorption cross sections
are higher compared to hard X-ray cross sections. This increased cross
section manifests in a smaller radiation absorption volume concentrated
at the sample surface and results in a higher absorbed radiation dose.
Second, for lower Z elements, the probability that a core hole will be filled
through a nonradiative process (i.e., Auger electron emission) is much
higher than the probability of radiative emission (Hubbell et al., 1994).
Since the escape depth of electrons is very shallow, the energy of Auger
electrons is mostly deposited within the sample, whereas fluorescence photons, have a longer escape depth, and have a lower probability of interacting with the sample before leaving. Together, the increased X-ray
absorption cross section, strong probability of Auger electron processes
over fluorescence, and the decreased likelihood that electrons will leave


Advances in Using Soft X-Ray Spectroscopy for Measurement of Soil Biogeochemical Processes 13

a sample before depositing kinetic energy, all contribute to the problem of
radiation damage in samples at soft X-ray energies.

Radiation damage manifests itself in mass loss, chemical changes in the
sample through the formation of free radicals, and reordering of bonding arrangements and changes to overall structural orders in crystalline and semiconductor systems (Beetz and Jacobsen, 2003; Leontowich et al., 2012;
Wang et al., 2009; Wilks et al., 2009; Zubavichus et al., 2004). Cryogenics
have been successfully employed to reduce radiation damage due to mass
loss in hard X-ray systems and are routinely employed for protein crystallography. Mass loss, however, is not necessarily the main source of radiation
damage at soft X-ray energies, and studies evaluating the use of cryogenic
cooling have not shown appreciable improvements in radiation-induced
damage artifacts over ambient temperature (Beetz and Jacobsen, 2003;
Terzano et al., 2013).
A more viable approach in addressing problems of radiation-induced
damage is to reduce the dose incident on the sample. The total absorbed
dose rate (DR) for the top 10 nm of sample is estimated as mega Gray per
sec (MGy$sÀ1, or 106 J$kgÀ1sÀ1) using:
DR ¼ ðF$EÞ=ðr$d$AÞ
where F is the incident photon flux, E is the photon energy, r is the
material density, d is the depth, and A is the beam spot area. One solution
to reducing the dose is to enlarge the beam spot size via defocusing, thus
distributing the photon flux throughout a larger area. A second would be
to reduce the overall photon flux by attenuating the incident beam using
filters or slits; this, however, negates the advantages afforded from the
higher flux of an undulator-based beamline. A third way therefore to
reduce the overall photon flux is to reduce the time required to record a
single scan. Reducing the scan time on a sample can be accomplished by
increasing the photon use efficiency of a beamline. This requires several
modifications to the methods by which spectral data are normally
collected.
Spectra can be acquired by operating the beamline in a slew scanning
mode, which continuously scans the energy of the monochromator and adjusts the gap of the beamline undulator while acquiring data. These scans can
vary in length, and typically are 5e20 s in duration. This is in contrast to step
scan mode where an absorption spectrum is obtained across an absorption

edge by a stepwise movement of the monochromator and undulator to positions which optimize the target photon energies before data collection


14

Adam W. Gillespie et al.

begins. These scans take on the order of 7e10 min due to the time required
to move optical components and operate with a duty cycle (time spent
acquiring data over the time that the sample is exposed to the beam) of
approximately 50%. In slew scan mode, the optics and detectors are run
continuously, and data is streamed to a database buffer during the scan’s
collection time. Real-time monochromator encoder feedback is included
in the exported dataset so that after the scan is complete, the stream of
data can be parsed into energy-resolved bins and output as a data file. For
this technique to be feasible, detector and encoder output must be acquired
rapidly and amplified cleanly. In practice, the signal to noise ratio is low, and
many slew scans must be recorded and averaged for a single sample. To
minimize radiation dose each scan is recorded on a fresh spot using a robotic
sample manipulator. Multiple slew scans are then combined, averaged, and
normalized. Figure 6 shows a comparison between multiple summed and
averaged slew scans (n ¼ 60) and a single step scan of 10% citric acid

Figure 6 Comparison of a single 10 min step scan to the average of 60 slew scans of
20 s on fresh sample spots of ca. 10% citric acid in Al2O3. Radiation-induced artifacts
are evident in the step scan.


Advances in Using Soft X-Ray Spectroscopy for Measurement of Soil Biogeochemical Processes 15

(w/w) in Al2O3. Radiation-induced artifacts are clearly present in the step
scan spectrum, as identified with arrows.
An important question in evaluating the effectiveness of slew scanning is the
need to determine if indeed radiation damage is occurring before the completion of the first scan. This can be evaluated empirically and theoretically.
Empirically, radiation damage can be assessed by measuring a candidate
compound repeatedly at a single spot without moving, and evaluating any
transformation or change in the spectra with respect to time. Figure 7 shows
a time series scan of glycine (an amino acid) and bovine serum albumin (a
protein) at the N K-edge. Proteins and peptides are known to be susceptible
to radiation damage, however, a single 20 s slew scan does not produce any
discernible artifact peaks.
Theoretical calculations using density functional theory of model compounds can be compared to experimental measurements to ascertain
whether features are present experimentally which are not predicted theoretically. Figure 8 shows the calculated X-ray transition intensities of citric
acid obtained using the StoBe density functional theory package along
with the slew scan X-ray absorption measurement of citric acid. Individual
transition intensities for the nonequivalent carbon sites on the citric acid
molecule are plotted to illustrate that there is a set of low energy transitions
(red and blue) arising from core level excitation of the atoms labeled 2 and 3.
(A)

(B)

Figure 7 Radiation damage to glycine and albumin monitored at the N K-edge, as a
function of increasing dose. Each spectrum represents a single slew scan.


16

Adam W. Gillespie et al.

Figure 8 Calculated electronic transition intensities for citric acid shown with the
measured XAS. The inset diagram and legend indicate the specific atoms giving rise
to each transition.

These transitions give rise to a pre-edge shoulder in the citric acid spectrum
that was previously believed to be due beam damage that occurred in the
duration of the first scan. The calculation was able to confirm that this
pre-edge feature is not due to damage but is inherent to the molecule’s
X-ray absorption spectrum.

3.1 Carbon K-Edge Analysis
Measurements at the C K-edge present a special set of challenges to the
analyst. First, samples must be prepared on C-free substrates. Second, carbon
contamination and the transmission of higher order light in beamline monochromators complicate the measurement of the incident flux (Io) during the
scan and can result in normalized spectra that exhibit distortions. The spectrum becomes more sensitive to the normalization procedure at low carbon
concentrations. Therefore, it is extremely important to use proper sample
preparation protocols and incident flux characterization for soils measurement where the C concentrations can be low.
For X-ray analysis of most elements, samples can be simply mounted on
double-sided conductive carbon tape. Indeed, a C-derived substrate is
unsuitable for conducting C analysis on any sample. Instead, samples are
mounted on gold-coated silicon wafers. In this procedure, gold pellets are
heated above 650  C in a vacuum chamber containing the silicon wafers.


Advances in Using Soft X-Ray Spectroscopy for Measurement of Soil Biogeochemical Processes 17

Figure 9 Sample preparation procedure for measurement at the C K-edge. Soil is
weighed into microcentrifuge tube, slurried in water, and evaporated onto a carbonfree substrate. On site preparation of gold-coated silica is a suitable substrate for this
measurement.

Sample material is then drop-coated onto the wafers by dissolving or slurrying in water, placing 8e10 mL of the mixture on the wafer, and allowing
to dry (Figure 9).
The incident flux of X-ray photons on a sample is modulated by C
contamination of the beamline optics. This C contamination, which arises
due to the beam-induced adsorption of residual hydrocarbon gases in the
beamline vacuum, will therefore introduce its own structure into the measurement (Figure 10). Ideally, the adventitious carbon would be removed

Figure 10 Lineshape of the incident beamline flux across the C K-edge. Measurement
is of light scattered by an Au-coated silicon wafer and a silicon drift detector.