You are on page 1of 10

Micron 75 (2015) 110

Contents lists available at ScienceDirect

Micron
journal homepage: www.elsevier.com/locate/micron

Using transmission Kikuchi diffraction to study intergranular stress


corrosion cracking in type 316 stainless steels
Martina Meisnar a, , Arantxa Vilalta-Clemente a , Ali Gholinia b , Michael Moody a ,
Angus J. Wilkinson a , Nicolas Huin c , Sergio Lozano-Perez a
a
University of Oxford, Department of Materials, Parks Road, OX1 3PH Oxford, UK
b
University of Manchester, School of Materials, Grosvenor Street, M1 7HS Manchester, UK
c
AREVA, Dpartement Mtallurgie et Corrosion, Section Corrosion et Tenue des Matriaux, Centre Technique France, France

a r t i c l e i n f o a b s t r a c t

Article history: Transmission Kikuchi diffraction (TKD), also known as transmission-electron backscatter diffraction
Received 13 March 2015 (t-EBSD) is a novel method for orientation mapping of electron transparent transmission electron
Received in revised form 20 April 2015 microscopy specimen in the scanning electron microscope and has been utilized for stress corrosion
Accepted 20 April 2015
cracking characterization of type 316 stainless steels. The main advantage of TKD is a signicantly higher
Available online 27 April 2015
spatial resolution compared to the conventional EBSD due to the smaller interaction volume of the
incident beam with the specimen.
Keywords:
Two 316 stainless steel specimen, tested for stress corrosion cracking in hydrogenated and oxygenated
Transmission Kikuchi diffraction
TKD
pressurized water reactor chemistry, were characterized via TKD. The results include inverse pole gure
Intergranular stress corrosion cracking (IPFZ) maps, image quality maps and misorientation maps, all acquired in very short time (<60 min)
SCC and with remarkable spatial resolution (up to 5 nm step size possible). They have been used in order to
316 stainless steel determine the location of the open crack with respect to the grain boundary, deformation bands, twinning
Scanning electron microscope and slip. Furthermore, TKD has been used to measure the grain boundary misorientation and establish a
SEM gauge for quantifying plastic deformation at the crack tip and other regions in the surrounding matrix.
Transmission electron backscatter Both grain boundary migration and slip transfer have been detected as well.
diffraction
2015 Elsevier Ltd. All rights reserved.
t-EBSD
Grain boundary migration
Grain boundary misorientation

1. Introduction Kain, 2011; Lynch, 2011; Shoji et al., 2011; Staehle, 2010; Das
et al., 2009; Lozano-Perez et al., 2009a; Couvant et al., 2007;
Stainless steel alloys such as SUS316 are widely used for Terachi et al., 2007). However, there is still only limited under-
applications in nuclear power plants because of their excellent per- standing of the mechanisms underlying SCC and there is no general
formance in high-temperature and corrosive environments (Arioka model for the initiation and propagation of SCC yet (Bruemmer
et al., 2006a). For decades, they have been commonly used as the and Thomas, 2010; Guerre et al., 2007; Vaillant et al., 2004; Scott
main constituents of structural components in the primary circuit and Le Calvar, 1993). Most studies are complicated by the fact
of Pressurized Water Reactors (PWRs) and Boiling Water Reac- that there are several parameters potentially simultaneously inu-
tors (BWRs) (Karlsen et al., 2010). However, despite their corrosion encing the occurring crack initiation and propagation processes,
resistance, these materials have shown susceptibility to intergra- including water chemistry, temperature and matrix composition
nular stress corrosion cracking (IGSCC) after long service periods (Terachi et al., 2007; Lozano-Perez et al., 2009b; Arioka et al., 2006b;
(Terachi et al., 2008; Arioka et al., 2007). Terachi and Arioka, 2006; Langevoort et al., 1984).
Hence, for at least two decades stress corrosion cracking (SCC) These previous SCC propagation studies have focused on the
in stainless steels has been the focus of increasing research efforts oxide structure and chemistry, the crack morphology and the
(Andresen, 2013; Arioka, 2013; Dietzel and Bala Srinivasan, 2011; matrix surrounding the crack tip, such as changes in the local matrix
composition and deformations. These features have been studied
using a variety of electron microscopy (EM) methods, including
Corresponding author. Tel.: +44 01865 273700; fax: +44 01865 273789. scanning electron microscopy (SEM), auger electron spectroscopy
E-mail address: martina.meisnar@materials.ox.ac.uk (M. Meisnar). (AES), transmission electron microscopy (TEM) (Kruska et al., 2011;

http://dx.doi.org/10.1016/j.micron.2015.04.011
0968-4328/ 2015 Elsevier Ltd. All rights reserved.
2 M. Meisnar et al. / Micron 75 (2015) 110

Lozano-Perez and Gontard, 2008; Lozano-Perez and Titchmarsh, EELS spectra were acquired with a JEOL ARM200F (cold-FEG gun,
2003; Lozano-Perez et al., 2004; Olszta et al., 2011) as well as 200 kV).
NanoSIMS (Lozano-Perez et al., 2008a,b). To complement these
mechanical and chemical studies of SCC, higher resolution charac-
terization from a crystallographic point of view (such as orientation 3. TKD methodology
mapping) is also necessary.
There is limited information on the relationship between SCC For decades, electron backscatter diffraction (EBSD) has been
crack propagation and crystallographic features in the sample sur- a very important and commonly used tool for the gathering
rounding the crack tip. This includes the misorientation of the of crystallographic data for many materials scientists (Dingley,
grains on either side of the crack, the plastic deformation before, 2004; Schwartz et al., 2009). Like many other techniques, recent
around and ahead of the crack tip as well as the existence of defor- progress in nano-science has required orientation mapping to
mation structures such as twinning, slip and deformation bands. improve in terms of spatial resolution. However, reports sug-
Comparing these features in different types of SCC specimen (with gest that EBSD appears to have reached its threshold at around
regard to water chemistry, temperature or matrix composition) 50100 nm, depending on the analyzed material (Keller and Geiss,
may provide signicant new insights with respect to the propa- 2012; Trimby, 2012; Zaefferer, 2011).
gation of SCC. In 2012, Keller et al. rst reported of a new technique called
This work represents a novel study of crystallographic features transmission-electron backscatter diffraction (t-EBSD) or later also
in two different types of samples, tested in hydrogenated and oxy- known as transmission Kikuchi diffraction (TKD) which uses the
genated PWR primary water chemistry, utilizing a new method of scanning electron microscope (SEM) with a conventional EBSD
high-resolution orientation mapping: transmission Kikuchi diffrac- detector to perform orientation mapping on electron transparent
tion (TKD) or sometimes also known as transmission-electron TEM foils (Keller and Geiss, 2012). While the exact same hard-
backscatter diffraction (t-EBSD). The samples were chosen because ware as for EBSD was used, the TKD setup changed slightly with
they manifest two different types of crack propagation. It is shown regard to the specimen position. The EBSD software accounts for
that, with this novel technique, high-resolution crystallographic this by adapting the pattern recognition algorithm. The TKD geom-
data can be easily acquired which may contribute to solving etry is derived from the basic EBSD setup with the main difference
some remaining questions about the SCC crack propagation pro- being the specimen position and its orientation with respect to
cess (such as the impact of plastic deformation around the crack the SEM column. Instead of the 70 tilt of the sample toward the
tip). EBSD camera, the TEM foil (mounted on a standard TEM Cu grid) is
almost horizontal (max. 10 tilt) with respect to the incident elec-
tron beam. Due to this setup and the thin TEM foil, the interaction
2. Material volume of the electron beam with the sample is minimized in TKD.
This is because of the sample thickness and the fact that the cone
The material used for this study was type 316 (reactor grade) of the incident electron beam in EBSD is much larger due to the tilt
stainless steel. Two different types of specimen were characterized: angle of 70 which increases the interaction volume [as shown via
type 316 tested under oxygenated PWR water conditions, provided Monte Carlo simulations in Keller and Geiss (2012)]. Hence, the lat-
by AREVA; and type 316 tested under hydrogenated PWR water eral resolution is therefore signicantly improved (Trimby, 2012).
conditions, provided by INSS. Both steels exhibited SCC after auto- In addition, the almost-horizontal mounting of the sample reduces
clave testing. The compositions of each alloy are listed in Table 1. the need for dynamical focus or tilt correction.
Specimen 316INSS underwent solution treatment followed by The ideal thickness of the TEM foil for TKD depends on the mate-
water quenching and uni-directional cold-rolling to a thickness rial composition; good quality results have been achieved at under
reduction of 20%. The SCC test (CGR) was performed at the INSS 100 nm thickness (Rice et al., 2014). In the rst years of application
(Japan) laboratories using a pre-cracked 1/2 CT specimen in the of this technique, it has become apparent that the quality of TKD
TS direction in an autoclave under constant load (30 MPa m1/2 ). patterns is very much dependent on the thickness of the specimen,
The sample was exposed to a testing environment of simulated which also has a strong effect on the achieved spatial resolution.
PWR water chemistry (hydrogenated water: 500 ppm B + 2 ppm Li, Therefore, TEM sample preparation is key for the successful and
+30 cm3 -STP/kg-H2 O DH2 ) for 700 h at 360 C. More information reliable data acquisition via TKD (Suzuki, 2013).
about the standard methods for stress corrosion cracking testing The TKD patterns originate from the volume very close to the
can be found on the American Society of Testing and Materials bottom surface of the sample (Trimby, 2012). No additions to the
website (www.astm.org). common EBSD system are necessary: the Kikuchi patterns are
Specimen 316AREVA was a reverse U-bend specimen with a recorded with the EBSD camera and the (for TKD slightly adapted)
shot-peened surface and was tested at the AREVA Laboratories EBSD software is responsible, as usual, for orientation mapping.
in France. The specimen was exposed to a PWR primary water There is a general consensus from researchers applying TKD to
chemistry in shutdown conditions with increased oxygen content the characterization of a variety of materials that this technique
(oxygenated conditions: 10 ppm O + 1200 B + 2 ppm Li) for 1500 h offers a signicant improvement in spatial resolution (Keller and
at 345 C. Geiss, 2012; Trimby, 2012; Rice et al., 2014; Brodusch et al., 2013;
The cross-sectioned surface of both samples was ground with Trimby and Cairney, 2014; Trimby et al., 2014; Babinsky et al.,
SiC paper and polished with 1-m diamond suspension. Mirror- 2014). The authors concur that in most cases that the advantages
nish was achieved by nal treatment (15 min) with colloidal silica. offered by this improved resolution outweighs the extra effort of
Subsequently, the bulk specimens were screened with an opti- having to prepare electron transparent TEM foils.
cal microscope and the SEM to locate the crack tips. These crack In this study, all TKD maps were collected using a Zeiss Merlin
tips have then been lifted out in situ in plan-view orientation FEG-SEM and an eFlashHR Bruker EBSD detector system. An accel-
in a focused ion beam (FIB) instrument. The lift-outs were then erating voltage of 20 kV and a probe current of 3 nA were used.
mounted on TEM Cu grids and thinned to electron transparency Reports (Trimby, 2012) suggest that in thicker sample regions 30 kV
(<100 nm). Before the actual TKD acquisition, the TEM foils were works best, but the used specimens were very thin (6080 nm) and
plasma cleaned for 5 min (Fishione 1020 plasma cleaner). Initial 20 kV produced sufciently good pattern quality. The TEM foil was
TEM observations were made with a JEOL 2100 LaB6 TEM and mounted on a special Bruker TKD sample holder with an intrinsic
M. Meisnar et al. / Micron 75 (2015) 110 3

Table 1
Chemical content of the alloys used in this study (wt%), provided by INSS Japan and AREVA France.

Alloy Fe Cr Ni C Si Mo Cu Mn P S

316AREVA Bal. 17.52 12.31 0.023 0.488 2.45 0.132 1.73 0.025 0.008
316INSS Bal. 16.54 11.00 0.047 0.045 2.07 n.a. 1.42 0.024 0.001

software and misorientation (MO) line proles have been extracted


manually.
Uncertainty estimation has been provided for all MO proles.
The utilized TEM lamellas may vary slightly in thickness across
the entire area and due to heavy deformation in the sample noise
and statistical errors have arisen, especially visible in the MO pro-
les. The standard deviation has therefore been calculated for every
prole around the plateau area, based on the average uctuation
around the extracted plateau value.

4. Results

The two samples selected for this study have been initially char-
acterized by TEM (JEOL 2100, LaB6 lament, 200 kV) in order to
detect interesting features and to determine the location of the
crack tip as well as oxides in the crack opening. STEM images of both
Fig. 1. Image of SEM chamber in measurement ready position with TKD sample FIB-prepared TEM lamellae are illustrated in Fig. 2 (a: 316AREVA,
holder (left); EBSD detector (right) and electron pole piece (top).
b: 316INSS).
Noticeably, the crack opening in 316AREVA appears much larger
than in 316INSS. At the crack tip, the opening measures 600 nm for
specimen tilt of 45 . The stage has been tilted in the opposite sense 316AREVA and 150 nm for 316INSS (as illustrated in Fig. 2).
to 35 in order to achieve a net tilt of 10 . The oxygenated water chemistry is known for the enhanced
Fig. 1 shows the sample-detector geometry inside the SEM dissolution of matrix material at crack anks and the crack tip com-
chamber. The sample is mounted at a net angle of 10 with ref- pared to the hydrogenated PWR typical water chemistry (Arioka
erence to the horizontal and the working distance is 7.5 mm. The et al., 2007, 2008; Yamada et al., 2009). This behavior could be a
EBSD detector has been inserted from the right-hand side and is fac- potential explanation for the dimensional difference of the crack
ing upwards slightly (4 ) to collect the incoming electrons exiting openings observed in 316AREVA (oxygenated, Fig. 2a) and 316INSS
the bottom sample surface. The detector distance to the specimen (hydrogenated, Fig. 2b), although it is also expected that stresses
is 12 mm. All these parameters were recommended by Bruker, the are higher in the reverse U-bend specimen.
manufacturer of the EBSD/TKD instrumentation and software. After the initial screening of the samples via TEM, TKD mea-
All TKD maps have been collected in 1 h with a step size surement has been performed as described in the previous section.
of 11 nm (note the increased resolution with respect to conven- Fig. 3 shows the IPFZ-, image quality- and the average misorienta-
tional EBSD) and TKD pattern resolution of 160 120 pixels. A total tion (misorientation will be henceforth abbreviated MO) maps of
area of 3.2 m 2.4 m was recorded. The data were subsequently the crack tip area in both samples.
analyzed with Esprit 1.9.4 version, the Bruker orientation map- As in conventional EBSD, inverse pole gure (z direction) maps
ping software. Inverse pole gure (IPFZ) maps, image quality maps (IPFZ) in Fig. 3a and d shows the existence of different grains with
and misorientation maps have been created automatically by the different crystallographic orientations by appearing in different

Fig. 2. STEM dark eld images acquired with JEOL 2100 at 200 kV; (a) 316AREVA TEM lamella, SCC test under (oxygenated) BWR conditions, crack tip opening 600 nm; (b)
316INSS TEM lamella, SCC test under (hydrogenated) PWR conditions, crack tip opening 150 nm; dashed lines indicate the location of the grain boundary (GB).
4 M. Meisnar et al. / Micron 75 (2015) 110

Fig. 3. TKD maps: (a) 316AREVA Inverse Pole Figure in z direction (IPFZ) map: different color represents a difference in crystallographic orientation (legend above); (b)
316AREVA image quality map: contrast arises form crystallographic defects or differences in crystallographic orientation; (c) 316AREVA average MO map: the brighter the
region the higher is the average MO; (d) 316INSS IPFZ map (legend above); (e) 316INSS image quality map; (f) 316INSS average MO map.

colors (colors according to the orientation triangle for fcc austen- to all (averaged) neighboring pixels. A color temperature scale (dark
ite; a legend for both IPFZ maps is displayed above Fig. 3a and d, blue to red) has been applied. Ultimately, regions with high MO
respectively). In addition, large oxide particles in the open crack average appear brighter (green, yellow, red) than those areas in
in 316INSS (in Fig. 3d) have also been indexed due to the excel- the sample with low MO average (light and dark blue).
lent spatial resolution of this method. Fig. 3a shows a similar grain It can be observed in 316INSS (Fig. 3f) that the region ahead
pattern along the crack anks with a lateral translation along the of the crack tip with a high density of defects such as deformation-
grain boundary (GB) of 200 nm. This indicates a crack opening and slip bands appears as region with high MO average, specically
mechanism that may have involved shear strain. where the slip bands intersect with the grain boundary. As will be
Similarly, based on low contrast in the band identication, seen next, deformation in the crack tip region is very localized in
image quality maps were generated (Fig. 3b and e) and show some this sample. Fig. 3c shows high average MO in the bottom grain
darker areas as regions of high defect density in the material. This close to the crack ank just before the crack tip. This zone of ele-
includes deformation bands, twinning and slip bands as well as the vated lattice rotation indicates a high level of plastic deformation
open crack advancing between the two grains and the grain bound- and its presence reinforces the previous suggestion of a crack open-
ary itself. The deformation history of the samples (cold rolling or ing mechanism that involved the sliding of the two grains against
reverse u-bending) indicates that the bands intersecting the grain each other (which would result in elevated lattice rotation at the
boundary are deformation bands. crack tip).
Finally, Fig. 3c and f illustrates the average misorientation (MO) In addition, misorientation-to-reference (MOTR) maps have
where each pixel represents the MO value in this pixel with regard been generated (an option available in the software). Here, each
M. Meisnar et al. / Micron 75 (2015) 110 5

Fig. 4. Misorientation to reference MOTR maps and misorientation proles along the GB ahead of the crack tip; (a) MOTR map in 316AREVA, white line: location of MO
prole, 0: origin of MO prole; (b) MO prole in 316AREVA, GB MO 4 1.2 ; (c) MOTR map in 316INSS; (d) MO prole in 316INSS, GB MO 29 1.5 .

pixel has a global MO value with regard to a single reference at different locations in the sample by utilizing the (indirect) rela-
point within the entire map. Each value within the map represents tionship between the measured MO and plastic deformation of the
the MO from the reference point to the respective pixel. Therefore, crystal lattice. This relationship stems from the fact that the misori-
the MO from one point in the map to another can be measured, entation relates to plastic strain gradients originating from lattice
revealing for instance the grain boundary misorientation. curvature (Lehockey et al., 2000). Such strain gradients are very
In the rst instance, the MOTR maps (316AREVA: Fig. 4a, likely to occur near crack tips in a constrained volume. It should be
316INSS: Fig. 4c) were used to measure the MO between the stressed at this point that all misorientation values are relative and
two grains adjacent to the open crack by acquiring MO proles plastic strain has not been quantied.
(316AREVA: Fig. 4b, 316INSS: Fig. 4d) from one grain to the other. Fig. 5a shows a MOTR map of the 316INSS sample and the loca-
The origin (0) of the MO prole is the point of reference for each tions of a number of MO proles acquired in the top and the bottom
pixel along the prole as well as for the entire MOTR map. Since grain at different distances from the crack tip. While the misorien-
there is only little change in MO in one grain, but presumably high tation between the two grains has been previously determined as
change from one grain to the other, the MO proles exhibit a surge 29 1.5 (Fig. 4d), the current MO proles are used to determine
at the location of the grain boundary. the MO within a single grain from the crack tip outwards into the
The MO prole for 316AREVA (Fig. 4b) indicates that there is surrounding matrix.
only 4 1.2 MO between the two grains (the low MO is also the MO proles 1 and 2 have been acquired at the crack tip in the
reason for an apparently noisier prole), whereas the MO prole sample and are shown in Fig. 5b and c. After exhibiting a sudden rise
in 316INSS (Fig. 4d) jumps from 0 in the top grain to a plateau of starting from the origin (close to the crack tip), the MO remains at
29 1.5 in the bottom grain. a certain value for the remaining distance of the prole. A gradient
MO prole measurements have been repeated for both samples of MO is apparent, indicated by the rising MO values over a certain
at different locations across the grain boundary leading away from distance, and suggests substantial lattice deformation in that region
the crack tip with very similar results each time. (which henceforth will be called the plastic zone).
The nal step in this characterization of SCC specimen via TKD The two parameters that have been extracted from the MO pro-
was the establishment of a method to measure the plastic strain les are the distance from the origin at which the MO reaches

Table 2
MO prole values for Fig.5a. The plastic zone refers to the distance from the origin of the misorientation prole within which there is a misorientation gradient present. The
II component refers to the plastic zone component perpendicular to the crack.

No. Grain Plastic zone (nm) Angle ( ) II plastic zone (nm) Plateau ( )

1 Bottom 100 40 76.6 2.8 0.8


1a Bottom 200 40 153.2 2.9 0.75
1b Bottom 200 40 153.2 2.9 0.8
1c Bottom 25 40 19.15 2 0.9
1d Bottom 25 40 19.15 2 0.85
1e Bottom 20 40 15.32 1.2 0.5
1f Bottom 100 40 76.6 3 0.9
2 Top 120 30 60 1.4 0.6
2a Top 200 30 100 1.4 0.6
2b Top 170 30 85 2.4 0.95
2c Top 400 30 200 4.5 1.2
2d Top 550 30 275 7 1.2
2e Top 50 30 25 4.2 1.15
2f Top 100 30 50 4 1.0
6 M. Meisnar et al. / Micron 75 (2015) 110

as well as the top grain ahead of the crack tip at different distances
from the crack tip.
Notably, the MO plateau values decrease the further away the
proles were acquired from the crack tip. MO prole 3 (Fig. 6e) was
taken at the crack tip and exhibits a plateau at 3.5 1.2 , whereas
MO prole 6, acquired at a distance of 4 m from the crack tip,
plateaus at 1.5 0.5 (Fig. 6h). As in 316INSS (Fig. 5b and c), the
region of plastic deformation can be extracted from the MO proles
by determining the length over which the MO gradient extends,
although here, MO proles 46 are already perpendicular to the
grain boundary. Due to that, only the II-value of MO proles 13
had to be calculated. While the plastic region in MO proles 3 and
6 extends over the rst 200 nm, MO prole 4 shows a shorter region
of plastic deformation (100 nm). In contrast, MO prole 5 reaches
the rst plateau after 200 nm (at 1.8 0.6 ) but exhibits another
MO gradient between 900 and 1100 nm (due to the bending of
the deformation bands), ending at a plateau of 2.5 0.7 .
In addition, the plastic zones at both crack anks of 316AREVA
have been determined. Fig. 6a shows the location of MO prole
1 (top grain) and MO prole 2 (bottom grain) in 316AREVA. Both
proles were acquired with the same angle to the horizontal and
therefore the distances are directly comparable. These measure-
ments show, that the plastic zone in the top grain extends much
further (500 nm) than in the bottom grain (20 nm).
Finally, Fig. 7 shows the comparison of a STEM bright eld image
with the MOTR map of the crack tip in the 316INSS sample and the
respective elemental EELS SI maps.
The STEM image in Fig. 7a shows the boundary between the
upper and the lower grain very clearly due to diffraction contrast.
The grain boundary extends from the left edge of the image in a
straight line until reaching close vicinity to the crack tip. Instead of
continuing straight however, the grain boundary advances into the
bottom grain on a path along the interface of the matrix with the
open crack, revealing that grain boundary migration has occurred.
This is clearly visible in the MOTR map (Fig. 7b) due to the con-
trast resulting from the misorientation between the two grains. In
Fig. 5. MO proles in 316INSS sample: (a) local MO map, location of MO proles 1
(white line, origin: white square), 1a1f (white dashed lines), MO proles 2 (black
instances, where the diffraction contrast is not as perceptible as in
line, origin: black dot) and 2a2f (black dashed lines); (b) MO prole1 (solid line: the present STEM image (due to specimen tilt for example), TKD
raw data, dashed line: tted data): MO reaches plateau of 2.8 0.8 at 100 nm is perhaps the most suitable method for studying grain boundary
after the origin; (c) MO prole 2: MO reaches plateau of 1.4 0.6 at 120 nm migration ahead of SCC crack tips.
after the origin; note: MO proles measured at different angles from GB; II-value
Furthermore, Fig. 7d (EELS SI map of OK edge) shows that grain
has been calculated for each prole and corresponds to the actual distance from the
GB (see Table). boundary migration has taken place ahead of the oxidation front
of the crack tip. The region through which the GB has passed is
enriched in Ni (Fig. 7c, EELS SI map of Ni L edge), but Fe and Cr
the plateau (the length of the plastic zone; bottom grain: 100 nm, depleted (Fig. 7e and f, EELS SI maps of Fe and Cr L edge). While
top grain: 120 nm) and the relative matrix rotation (in degrees) at most of this region is also depleted in O, there is a thin nger of Cr
which the plateau is reached (bottom grain: 2.8 0.8 , top grain: oxide (see O and Cr map) protruding into the Ni-rich region.
1.4 0.6 ). While in both grains the MO reaches a plateau after
similar distances from the origin, the MO plateau values are very
different (twice as high in the bottom grain compared to the top 5. Discussion
grain). However, the MO proles have been acquired at different
angles from the horizontal (grain boundary orientation), in order to In this study, two specimen exhibiting SCC have been analyzed
avoid intersection with deformation bands, and therefore, II-values via the novel TKD technique which allows orientation mapping
(normal to the GB, in order to get an actual comparable distance with remarkable spatial resolution.
from the crack tip, grain boundary or open crack) have been calcu- First of all, the occurrence of IGSCC can easily be veried via TKD
lated via basic trigonometry in order to be comparable. The results IPFZ maps (Fig. 3a and d), as they highlight the different crystal-
of these calculations for all MO proles are shown in Table 2. lographic orientation of the two grains. In addition, intergranular
MO proles 2c2f in the top grain show elevated plateau values oxide particles can be easily indexed via TKD. It has been previ-
compared to the MO values around the crack tip, probably because ously established that crystalline oxide particles in open SCC cracks
of the high density of deformation bands in that region. Remark- are usually Fe-rich magnetite or spinel oxides (Bruemmer and
ably, MO proles 1a and 1b (taken at the crack anks of the bottom Thomas, 2010; Lozano-Perez et al., 2009b, 2012; Olszta et al., 2012;
grain) show a much larger plastic zone than the proles taken at or Staehle, 2007). In TKD, they appear as particles of certain crystal-
ahead of the crack tip. This leads to the conclusion that the strain lographic orientation in the IPFZ maps and can be easily located
might be much more localized at the crack tip and ahead of it than and distinguished from each other. The misorientation of the oxide
at the crack anks. Fig. 6 shows a MOTR map in 316AREVA, indi- particles and the surrounding matrix can also be easily measured
cating the location of 6 MO proles extracted from the crack anks with this method. This is particularly important in measuring the
M. Meisnar et al. / Micron 75 (2015) 110 7

Fig. 6. MO proles at crack anks and ahead of the crack tip in 316AREVA; (a) MOTR map indicating the locations of MO proles 1 (top grain) and 2 (bottom grain), the crack
tip region and MO prole 3 at the crack tip (top grain); (b) GB region ahead of crack tip, location of MO proles 46 (top grain); (c) MO prole 1 (plastic zone extends over
500 nm); (c) MO prole 2 (plastic zone extends over 20 nm); (eh) MO proles 36, decreasing MO plateau value with increasing distance from crack tip.

crystallographic relationship between the individual oxide parti- that there is a connection between intergranular SCC crack propa-
cles to the grains on either side of the crack and identifying trends gation and the MO of the grain boundaries the crack propagates
in the oxide particles misorientation and proximity to the parent along, is a prominent theory within the SCC research commu-
grain. In this study, no epitaxial relationships were found, which nity (Bruemmer et al., 2012, 2013; Baik et al., 2012; Britton and
suggests an outer oxide growth enhanced by precipitation from Wilkinson, 2012; Gertsman and Bruemmer, 2001; Lynch, 2013;
ions in solution. Lozano-Perez, 2008a,b; Raja and Shoji, 2011). In particular, West
In addition, TKD has shown to be very effective for the identi- et al. have established that there is a connection between slip trans-
cation of defects in the material. Image quality maps (Fig. 3b and e) fer and other grain boundary parameters (e.g. Schmid and Taylor
are constructed based on the index-ability of the Kikuchi pattern in factor) and SCC susceptibility of certain GBs (West et al., 2012).
each pixel. Regions of higher defect density are harder for the soft- Increasing grain boundary inclination and decreasing Schmid fac-
ware to index and therefore appear as darker regions in the image tor lead to the highest normal stress on the grain boundary, making
quality map (Wright and Nowell, 2006). Hence, it is possible with it more susceptible to cracking. In addition, their ndings suggest
TKD to acquire high-resolution maps of defects and locate the SCC that cracking is more likely along grain boundaries, which exhibit
crack tip, deformation bands, twinning and slip in the sample. discontinuous slip. It should be noted that although slip trans-
For example slip transfer (Guo et al., 2014), is a phenomenon fer through the cracked grain boundary was observed in 316INSS
frequently observed in IGSCC research and is straight-forward to (Fig. 3e), there was enough slip incompatibility to create areas
detect in the TKD image quality maps. The slip bands ahead of the of localized deformation at the intersection of deformation bands
open SCC crack in 316INSS (Fig. 3e) intersect the grain boundary and with the grain boundary (Fig. 3f).
protrude into the neighboring grain, however at a slightly different Within this study, TKD has been discovered as a method where
angle. It was also highlighted that deformation bands intersect- the determination of the grain boundary MO is particularly easy
ing grain boundaries can create regions of high local deformation to achieve. It has been determined, that the grain boundary MO in
(Fig. 3f). TKD image quality maps are very well suited for this 316AREVA is 4 1.2 and 29 1.5 in 316INSS. Low-angle grain
type of research and are as such extremely effective for an initial boundaries as in 316AREVA are unusual for IGSCC, but there are
high-resolution study of defects surrounding SCC crack tips before many other factors inuencing SCC crack propagation that could
commencing with methods such as TEM. be the reason for propagation along this particular grain boundary.
Moreover, grain boundary misorientation ahead of the crack For example, due to the high level of deformation this sample was
tip has also been measured in both samples (Fig. 4). The notion exposed to previous to the SCC test (reverse U-bend and surface
8 M. Meisnar et al. / Micron 75 (2015) 110

Fig. 7. Illustration of grain boundary migration in 316INSS; (a) STEM bright eld image recorded with JEOL 2100 at 200 kV; diffraction contrast between the two grains (b)
local misorientation map recorded via TKD: the contrast between grain 1 and grain 2 results from the different crystallographic orientations of the two grains; (cf) EELS SI
maps acquired via JEOL ARM200 at 200 kV (0.7 eV energy resolution, 5 nm pixel size).

shot-peening), the high stress concentration and defect densities the matrix (Fig. 5). At the crack tip in sample 316INSS, the plastic
at the GBs makes them susceptible to intergranular SCC. Additional zones were measured in MO proles 1 and 2 and extend over a
data has to be acquired in order to make a more general statement very similar distance in both grains. However, their nal MO value
about the dependence of SCC crack propagation on grain boundary (the plateau) is very different (top grain: 1.4 0.6 , bottom grain:
MO, which can now be easily accomplished with TKD. 2.8 0.8 ). Therefore, it can be concluded that the lattice deforma-
Furthermore, grain boundary migration, a phenomenon that has tion is approximately twice as high in the bottom grain compared
recently been discussed with respect to intergranular SCC (Lozano- to the top grain.
Perez et al., 2014; Schreiber et al., 2014), can clearly be identied In the top grain, the plateau values were higher in the region of
via TKD. The evolution of a grain boundary, often under stress and high deformation band (DB) density, which may be an indicator for
corrosion-enhanced diffusion, by traveling short distances through high stress concentration. In the bottom grain, the plastic zone was
the crystal and simultaneous growth of one grain at the expense of much more localized at the crack tip and ahead of it, compared to
the other is known as grain boundary migration. This phenomenon the measurements at the crack ank.
is possibly enhanced in the vicinity of dislocations or lattice defor- The acquisition of MO proles in the top grain of sample
mations and has been discovered close to SCC crack tips several 316AREVA (Fig. 6) showed that the plateau value of the MO
times (Lozano-Perez et al., 2014; Schreiber et al., 2014). decreases with increasing distance away from the crack tip along
Finally, it has been shown how MO proles were used to deter- the GB. The extent of the plastic zone at the crack tip region was
mine the plastic zone reaching from the crack tip outwards into also much bigger than in the 316INSS sample. Comparison of the
M. Meisnar et al. / Micron 75 (2015) 110 9

MO proles at the crack anks in 316AREVA yielded a much larger Acknowledgements


plastic zone in the top grain than in the bottom grain. As dis-
cussed previously, metal dissolution plays an important role in The authors are grateful for the support of AREVA (France) and
the oxygenated water chemistry of the oxygenated environment. INSS (Japan) in terms of providing the samples and funding. Finally,
Therefore, we can expect that the crack anks we see in 316AREVA we would like to thank Bruker for supporting the establishment of
(Fig. 6a) are not the original crack anks present at the beginning of TKD at Oxford Materials.
the SCC test and part of the plastic zone might have been dissolved
since then. However, this seems to only apply to the bottom grain,
which exhibits a much smaller plastic zone. A possible conclusion
could be that the bottom grain has been preferentially dissolved. References
In general, comparing the zone of plastic deformation at the
crack tip in both samples, it appears that the MO gradient extends Andresen, P.L., 2013. Basic Elements of Initiation and Propagation of SCC in Water-
cooled Nuclear Power Systems.
over a smaller distance (100 nm) in 316INSS than in 316AREVA
Arioka, K., 2013. Role of Diffusion of Vacancy in Materials Lead to SCC Initiation.
(200 nm). Both specimens are not directly comparable in terms Arioka, K., Yamada, T., Terachi, T., Staehle, R.W., 2006a. Intergranular stress
of the governing SCC mechanisms due to different test con- corrosion cracking behavior of austenitic stainless steels in hydrogenated high-
ditions (water chemistry and form of pre-existing stress), but temperature water. Corrosion 62, 7483.
Arioka, K., Yamada, T., Terachi, T., Chiba, G., 2006b. Inuence of carbide pre-
the differences in crack opening and morphology can be corre- cipitation and rolling direction on intergranular stress corrosion cracking of
lated to a different oxidation and mechanical response of the austenitic stainless steels in hydrogenated high-temperature water. Corrosion
alloys. 62, 568575.
Arioka, K., Yamada, T., Terachi, T., Chiba, G., 2007. Cold work and temper-
Regardless of the form of pre-existing stress (cold-rolling for ature dependence of stress corrosion crack growth of austenitic stainless
316INSS and reverse U-bend for 316AREVA), the plastic strain steels in hydrogenated and oxygenated high-temperature water. Corrosion 63,
around the crack tip seems to be much more localized in 316INSS 11141123.
Arioka, K., Yamada, T., Terachi, T., Miyamoto, T., 2008. Dependence of stress corrosion
than in 316AREVA. In addition, it has been observed that a much cracking for cold-worked stainless steel on temperature and potential, and role
larger portion of the crack ank was dissolved in 316AREVA during of diffusion of vacancies at crack tips. Corrosion 64, 691706.
the SCC test. In contrast, the crack anks in 316INSS appear to have Babinsky, K., De Kloe, R., Clemens, H., Primig, S., 2014. A novel approach for
site-specic atom probe specimen preparation by focused ion beam and trans-
been maintained to a very high degree. This suggests that the inu-
mission electron backscatter diffraction. Ultramicroscopy 144, 918.
ence of oxidation and plastic deformation on crack propagation in Baik, S., Olszta, M.J., Bruemmer, S.M., Seidman, D.N., 2012. Grain-boundary structure
both samples might not be the same. It is known that materials and segregation behavior in a nickel-base stainless alloy. Scr. Mater. 66, 809812.
Britton, T.B., Wilkinson, A.J., 2012. High resolution electron backscatter diffraction
exposed to oxygenated water chemistry (as 316AREVA) are much
measurements of elastic strain variations in the presence of larger lattice rota-
more prone to metal dissolution due to oxidation than materials tions. Ultramicroscopy 114, 8295.
exposed to PWR primary water (like 316INSS). Therefore, it could be Brodusch, N., Demers, H., Gauvin, R., 2013. Nanometres-resolution Kikuchi patterns
proposed that the governing factor in SCC propagation in 316AREVA from materials science specimens with transmission electron forward scatter
diffraction in the scanning electron microscope. J. Microsc. 250, 114.
might have been a mixture of dissolution and plastic deformation. Bruemmer, S.M., Thomas, L.E., 2010. Insights into stress corrosion cracking
In contrast, the plastic zone around the crack tip in 316INSS is much mechanisms from high-resolution measurements of crack-tip structures and
more localized and hence plastic deformation appears to have been compositions. Mater. Res. Soc. Symp. Proc. 1264, 159170.
Bruemmer, S.M., Olszta, M.J., Toloczko, M.B., Thomas, L.E., 2012. Linking grain bound-
the major contributor to the SCC process. However, at this stage ary microstructure to stress corrosion cracking of cold rolled alloy 690 in PWR
further study is required to conrm the validity of these proposed primary water. NACE Int. Corrosion Conf. Ser. 7, 59125927.
mechanisms and to propose new insights regarding a general SCC Bruemmer, S.M., Olszta, M.J., Toloczko, M.B., Thomas, L.E., 2013. Linking grain
boundary microstructure to stress corrosion cracking of cold-rolled alloy 690
model. in pressurized water reactor primary water. Corrosion 69, 953963.
Couvant, T., Legras, L., Pokor, C., Vaillant, F., Brechet, Y., Boursier, J.M., et al., 2007.
Investigations on the mechanisms of PWSCC of strain hardened austenitic
stainless steels. Canadian Nuclear Society 13th International Conference on
6. Summary and conclusion
Environmental Degradation of Materials in Nuclear Power Systems, 1., pp.
499514.
A number of signicant microstructural and micromechanical Das, N.K., Suzuki, K., Ogawa, K., Shoji, T., 2009. Early stage SCC initiation analysis of
fcc Fe-Cr-Ni ternary alloy at 288 C: a quantum chemical molecular dynamics
features related to SCC have been successfully analyzed via TKD, a
approach. Corrosion Sci. 51, 908913.
novel high-resolution orientation mapping method based on tradi- Dietzel, W., Bala Srinivasan, P., 2011. Testing and evaluation methods for stress cor-
tional EBSD. TKD uses electron transparent TEM lamellae in an SEM rosion cracking (SCC) in metals. In: Raja, V.S., Shoji, T. (Eds.), Stress Corrosion
and offers higher spatial resolution than EBSD due to the smaller Cracking: Theory and Practice. , 1st edition. Woodhead Publishing, pp. 133166.
Dingley, D., 2004. Progressive steps in the development of electron backscatter
interaction volume of the incident beam with the specimen. Tradi- diffraction and orientation imaging microscopy. J. Microsc. 213, 214224.
tional EBSD hard- and software can be used for this technique with Gertsman, V.Y., Bruemmer, S.M., 2001. Study of grain boundary character along
slight changes in the setup. intergranular stress corrosion crack paths in austenitic alloys. Acta Mater. 49,
15891598.
In this study, TKD has been used to identify and characterize Guerre, C., Raquet, O., Herms, E., Marie, S., Le Calvar, M., 2007. SCC crack growth
grain boundary misorientation, grain boundary migration, oxide rate of cold-worked austenitic stainless steels in PWR primary water conditions.
particles in the crack opening, deformation and slip bands ahead of Canadian Nuclear Society 13th International Conference on Environmental
Degradation of Materials in Nuclear Power Systems, 1., pp. 676699.
the crack tip as well as the size and degree of plastic deformation Guo, Y., Britton, T.B., Wilkinson, A.J., 2014. Slip band-grain boundary interactions in
around and ahead of the crack tip. commercial-purity titanium. Acta Mater. 76, 112.
In conclusion, the application of TKD for characterizing SCC Kain, V., 2011. Stress corrosion cracking (SCC) in stainless steels. In: Raja, V.S., Shoji,
T. (Eds.), Stress Corrosion Cracking: Theory and Practice. , 1st edition. Woodhead
specimen proved to be very successful for the analysis of some
Publishing, pp. 199244.
of the key aspects of SCC and makes it possible to avoid time- Karlsen, W., Diego, G., Devrient, B., 2010. Localized deformation as a key precursor to
consuming diffraction techniques in order to gain crystallographic initiation of intergranular stress corrosion cracking of austenitic stainless steels
employed in nuclear power plants. J. Nucl. Mater. 406, 138151.
information about the sample. In TKD, a map can be acquired
Keller, R.R., Geiss, R.H., 2012. Transmission EBSD from 10 nm domains in a scanning
in 30 min and individual points in the specimen can be ana- electron microscope. J. Microsc. 245, 245251.
lyzed in a few seconds. TKD is highly recommended for future Kruska, K., Lozano-Perez, S., Saxey, D.W., Terachi, T., Yamada, T., Smith, G.D.W., 2011.
SCC research in order to study the inuence of grain orienta- 3D atom-probe characterization of stress and cold-work in stress corrosion
cracking of 304 stainless steel. 15th International Conference on Environmen-
tion, lattice defects and plastic deformation on the governing SCC tal Degradation of Materials in Nuclear Power Systems-Water Reactors, 2., pp.
mechanisms. 891898.
10 M. Meisnar et al. / Micron 75 (2015) 110

Langevoort, J.C., Fransen, T., Geilings, P.J., 1984. On the inuence of cold work on Schreiber, D.K., Olszta, M.J., Bruemmer, S.M., 2014. Grain boundary depletion and
the oxidation behavior of some austenitic stainless steels: High temperature migration during selective oxidation of Cr in a Ni-5Cr binary alloy exposed to
oxidation. Oxidation Met. 21, 271284. high-temperature hydrogenated water. Scr. Mater. 89, 4144.
Lehockey, E., Lin, Y., Lepik, O., 2000. Mapping Residual Plastic Strain in Materials Schwartz, A.J., Kumar, M., Adams, B.L., Field, D., 2009. Electron Backscatter Diffraction
Using Electron Backscatter Diffraction., pp. 247264. in Materials Science, 2nd edition. Springer.
Lozano-Perez, S., 2008a. A guide on FIB preparation of samples containing stress Scott, P.M., Le Calvar, M., 1993. Some Possible Mechanisms of Intergranular Stress
corrosion crack tips for TEM and atom-probe analysis. Micron 39, 320328. Corrosion Cracking of Alloy 600 in PWR Primary Water., pp. 657667.
Lozano-Perez, S., 2008b. Novel characterization of stress corrosion cracks. J. Phys.: Shoji, T., Lu, Z., Peng, Q., 2011. Factors affecting stress corrosion cracking (SCC) and
Conf. Ser. 126. fundamental mechanistic understanding of stainless steels. In: Raja, V.S., Shoji,
Lozano-Perez, S., Gontard, L.C., 2008. Understanding stress corrosion cracking with T. (Eds.), Stress Corrosion Cracking: Theory and Practice. , 1st edition. Woodhead
electron tomography. Microsc. Microanal. 14, 642643. Publishing, pp. 245272.
Lozano-Perez, S., Titchmarsh, J.M., 2003. TEM investigations of intergranular stress Staehle, R.W., 2007. Critical analysis of tight cracks. Canadian Nuclear Society 13th
corrosion cracking in austenitic alloys in PWR environmental conditions. Mater. International Conference on Environmental Degradation of Materials in Nuclear
High Temp. 20, 573579. Power Systems, 3., pp. 18771957.
Lozano-Perez, S., Titchmarsh, J.M., Jenkins, M.L., 2004. TEM characterization of stress Staehle, R.W., 2010. Approach to Predicting SCC of Fe-Cr-Ni Alloys.
corrosion cracks in 304SS. Inst. Phys. Conf. Ser. 179, 233236. Suzuki, S., 2013. Features of transmission EBSD and its application. JOM 65,
Lozano-Perez, S., Kilburn, M.R., Yamada, T., Terachi, T., English, C.A., Grovenor, C.R.M., 12541263.
2008a. High-resolution imaging of complex crack chemistry in reactor steels by Terachi, T., Arioka, K., 2006. Characterization of oxide lm behaviors on 316 stainless
NanoSIMS. J. Nucl. Mater. 374, 6168. steels in high-temperature water inuence of hydrogen and oxygen consider-
Lozano-Perez, S., Schrder, M., Yamada, T., Terachi, T., English, C.A., Grovenor, C.R.M., ations for initiation of SCC. In: NACE International Corrosion Conference Series,
2008b. Using NanoSIMS to map trace elements in stainless steels from nuclear pp. 0660810660810.
reactors. Appl. Surf. Sci. 255, 15411543. Terachi, T., Yamada, T., Chiba, G., Arioka, K., 2007. Inuence of cold work on IGSCC of
Lozano-Perez, S., Saxey, D., Terachi, T., Yamada, T., Cervera-Gontard, L., 2009a. 3-D 316 stainless steel in hydrogenated high-temperature water. In: NACE Inter-
characterization of SCC in cold worked stainless steels. 14th International Con- national Corrosion Conference Series, pp. 076051076059.
ference on Environmental Degradation of Materials in Nuclear Power Systems Terachi, T., Yamada, T., Miyamoto, T., Arioka, K., Fukuya, K., 2008. Corrosion behavior
Water Reactors, 2., pp. 926934. of stainless steels in simulated PWR primary water-effect of chromium content
Lozano-Perez, S., Yamada, T., Terachi, T., Schrder, M., English, C.A., Smith, G.D.W., in alloys and dissolved hydrogen. J. Nucl. Sci. Technol. 45, 975984.
et al., 2009b. Multi-scale characterization of stress corrosion cracking of cold- Trimby, P.W., 2012. Orientation mapping of nanostructured materials using
worked stainless steels and the inuence of Cr content. Acta Mater. 57, transmission Kikuchi diffraction in the scanning electron microscope. Ultrami-
53615381. croscopy 120, 1624.
Lozano-Perez, S., Kruska, K., Iyengar, I., Terachi, T., Yamada, T., 2012. Understand- Trimby, P.W., Cairney, J.M., 2014. Transmission kikuchi diffraction in the scanning
ing surface oxidation in stainless steels through 3D FIB sequential sectioning. J. electron microscope: orientation mapping on the nanoscale. Adv. Mater. Pro-
Phys.: Conf. Ser. 371. cess. 172, 1315.
Lozano-Perez, S., Dohr, J., Meisnar, M., Kruska, K., 2014. SCC in PWRs: learning from Trimby, P.W., Cao, Y., Chen, Z., Han, S., Hemker, K.J., Lian, J., et al., 2014. Characterizing
a bottom-up approach. Metall. Mater. Trans. E 1, 194210. deformed ultrane-grained and nanocrystalline materials using transmission
Lynch, S.P., 2011. Mechanistic and fractographic aspects of stress-corrosion crack- Kikuchi diffraction in a scanning electron microscope. Acta Mater. 62, 6980.
ing (SCC). In: Raja, V.S., Shoji, T. (Eds.), Stress Corrosion Cracking: Theory and Vaillant, F., Boursier, J., Rouillon, Y., Couvant, T., Amzallag, C., Raquet, O., 2004. Stress
Practice. , 1st edition. Woodhead Publishing, pp. 178. Corrosion Cracking of Cold Worked Austenitic Stainless Steels in Laboratory
Lynch, S.P., 2013. Mechanisms and kinetics of environmentally assisted cracking: Primary PWR environment, 490. American Society of Mechanical Engineers,
current status, issues, and suggestions for further work. Metall. Mater. Trans. A: Pressure Vessels and Piping Division (Publication) PVP, pp. 5163.
Phys. Metall. Mater. Sci. 44, 12091229. West, E.A., McMurtrey, M.D., Jiao, Z., Was, G.S., 2012. Role of localized deformation
Olszta, M.J., Schreiber, D.K., Thomas, L.E., Bruemmer, S.M., 2011. Electron microscopy in irradiation-assisted stress corrosion cracking initiation. Metall. Mater. Trans.
characterizations and atom probe tomography of intergranular attack in alloy A: Phys. Metall. Mater. Sci. 43, 136146.
600 exposed to PWR primary water. 15th International Conference on Environ- Wright, S.I., Nowell, M.M., 2006. EBSD image quality mapping. Microsc. Microanal.
mental Degradation of Materials in Nuclear Power Systems-Water Reactors, 2., 12, 7284.
pp. 14221435. Yamada, T., Terachi, T., Miyamoto, T., Arioka, K., 2009. Crack growth behav-
Olszta, M., Schreiber, D., Thomas, L., Bruemmer, S., 2012. High-resolution crack imag- ior of welded and cast stainless steels in hydrogenated and oxygenated
ing reveals degradation processes in nuclear reactor structural materials. Adv. high-temperature water. 14th International Conference on Environmental
Mater. Process. 170, 1721. Degradation of Materials in Nuclear Power Systems Water Reactors, 1., pp.
Raja, V.S., Shoji, T., 2011. Stress Corrosion Cracking Theory and Practice. Woodhead 684689.
Publishing, Cambridge, UK. Zaefferer, S., 2011. A critical review of orientation microscopy in SEM and TEM. Cryst.
Rice, K.P., Keller, R.R., Stoykovich, M.P., 2014. Specimen-thickness effects on trans- Res. Technol. 46, 607628.
mission Kikuchi patterns in the scanning electron microscope. J. Microsc. 254,
129136.