Appl Phys A (2009) 94: 943–948
DOI 10.1007/s00339-008-4870-y
Microstructural characterization of ultrafine-grain
interstitial-free steel by X-ray diffraction line profile analysis
Apu Sarkar · Ayan Bhowmik · Satyam Suwas
Received: 30 April 2008 / Accepted: 6 August 2008 / Published online: 11 September 2008
© Springer-Verlag 2008
Abstract This paper highlights the microstructural features
of commercially available interstitial free (IF) steel speci-
mens deformed by equal channel angular pressing (ECAP)
up to four passes following the route A. The microstructure
of the samples was studied by different techniques of X-ray
diffraction peak profile analysis as a function of strain (ε).
It was found that the crystallite size is reduced substantially
already at ε = 2.3 and it does not change significantly dur-
ing further deformation. At the same time, the dislocation
density increases gradually up to ε = 4.6. The dislocation
densities estimated from X-ray diffraction study are found
to correlate very well with the experimentally obtained yield
strength of the samples.
PACS 61.05.Cp · 61.72.-y · 61.72.Dd
1 Introduction
Nanostructured/ultrafine grain materials have been the sub-
ject of intensive investigations in the recent years due to
their superior mechanical and physical properties. The past
two decades have witnessed remarkable advancement in
processing and characterization of materials with ultra-
fine grains in nanometer and submicrometer range. Several
methods like inert gas condensation [1], ball milling or me-
chanical alloying [2] are being extensively used to prepare
nanomaterials. These techniques allow producing ultrafine
grains with size ∼10 nm in powder form. However, high
A. Sarkar () · A. Bhowmik · S. Suwas
Department of Materials Engineering, Indian Institute of Science,
Bangalore 560012, India
e-mail: apu@platinum.materials.iisc.ernet.in
residual porosity, contamination, and small sample dimen-
sions have been the major barrier in successful utilization of
these nanomaterials in industrial applications [3].
Severe plastic deformation (SPD) methods, in contrast to
powder processing based methods, are the methods to pro-
duce bulk nanostructured materials. The basic principle of
SPD is the heavy straining of materials under high pres-
sure [3]. Nanomaterials produced by SPD are fully dense
and hence there are no chances of contamination, as often
seen in the materials processed through powder based tech-
niques. SPD also allows producing nanomaterials with large
geometrical dimensions [4]. Equal channel angular pressing
(ECAP) is one of the major SPD methods widely used to
fabricate bulk nanostructured materials, almost for a decade
[4–6]. A grain size of the order of few hundred nanometers
is easily achieved in this process. The mechanism of grain
refinement via ECAP has been extensively investigated in
numerous materials [7–12]. It is now well accepted that the
basic underlying mechanism of grain refinement involves an
increase in the dislocation density by heavily deforming the
materials. Dislocations form an ordered arrangement giving
rise to dislocation walls. As the deformation continues, these
dislocation walls finally transform into grain boundaries. In
order to achieve the desired goal of the ECAP, it is manda-
tory to ensure that no restoration process, i.e., recovery of
dislocation substructure, occur during the process [13, 14].
Thus, it is very interesting to investigate the evolution of dis-
location substructure at different stages during ECAP.
The characterization of the microstructural deformation
of materials is a challenging task. In case of SPD processed
materials, this becomes rather complicated due to the pres-
ence of a large fraction of low angle boundaries plus some
special boundaries. Electron backscatter diffraction (EBSD)
is one of the most popular techniques used to investigate the
944
microstructure of the materials. However, EBSD has limi-
tations in both spatial and angular resolutions and therefore
cannot be used to reveal the low-angle boundaries and dis-
locations. Transmission electron microscopy (TEM) is the
most powerful technique used for studying the dislocation
substructure within the material. It provides a direct way to
observe materials’ microstructure within the material. How-
ever, it has some serious drawbacks. Only a miniscule vol-
ume of the sample is investigated under TEM which may
not be representative of the bulk, particularly, when hetero-
geneities are reported to be present at all levels in ECAP
processed material [15, 16]. On the other hand, TEM sam-
ple preparation is a tedious task. In this respect, X-ray dif-
fraction (XRD) is a superior technique to investigate the mi-
crostructure of plastically deformed samples. XRD gives in-
formation of the bulk properties of a polycrystalline solid,
averaged over the sample volume. The sample preparation
in this case is easy and safe.
Microstructural studies of deformed polycrystalline sam-
ple from X-ray diffraction line profile analysis (XRDLPA)
have been a topic of renewed interest for the last two
decades. Many new methods of XRDLPA have been pro-
posed to extract microstructural information from the XRD
peaks [17–19]. XRDLPA has been recently used in several
studies to investigate the microstructural evolution during
ECAP of different materials [20–23]. However, most of the
previous studies were focused on the face-centered cubic
(fcc) and hexagonal close-packed (hcp) materials. In this
study, we extend the application of XRDLPA to characterize
the microstructure of SPD processed body-centered cubic
(bcc) interstitial free (IF) steel material.
2 Experimental details
2.1 Material and method
Interstitial free (IF) steel, with chemical composition given
in Table 1, was processed through equal channel angular
pressing up to 4 passes without rotation around the axis of
the billet (route A) in a specially designed die [24]. Details
of the pressing parameters are given in the reference [25].
For a die with interchannel angle ϕ without any rounding
of a corner, the total strain, εn imparted to the specimen after
N passes can be estimated as [4]:
ϕ
εn = 1.15N cot
2
Table 1 Chemical composition of IF steel
Fe C Mn S P Si Al Ti N hydraulic DARTEC machine. From the experimental stress
Bal. 0.0027 0.07 0.007 0.007 0.006 0.05 0.056 0.003
A. Sarkar et al.
In the present study ϕ was 90◦ . Therefore, the strains after
one, two, three, and four passes were 1.15, 2.30, 3.45, and
4.60 respectively.
2.2 X-ray diffraction
For X-ray diffraction study samples were cut from the mid-
plane section (ED-ND plane) and subjected to metallo-
graphic preparation up to electropolishing. The X-ray dif-
fraction patterns were recorded from the polished surface
using a PANalytical X’PertPro diffractometer with Cu Kα
radiation. All the diffraction profiles were obtained by vary-
ing 2θ from 40◦ to 120◦ in a continuous scan mode with a
step size of 0.017◦ . The time spent for collecting the data
per step was 400 s. Figures 1a and 1b show typical XRD
patterns for the starting and the 4 pass deformed samples,
respectively.
The observed line profile is the convolution of instrumen-
tal and specimen (microstructural) broadening. Therefore
the first step before any attempt to analyze diffraction line
broadening is to correct the observed line profile for instru-
mental effects. A careful scan of suitable standard sample,
showing minimal physical broadening, defines the instru-
mental contribution to the broadening [26]. We have used
standard silicon sample to correct the instrumental broaden-
ing. The peaks of the silicon sample were fitted with pseudo-
Voigt (pV ) function. A pV function is basically a linear
combination of Loretzian (L) and Gaussian (G) functions
and is expressed as:
pV = ηL + (1 − η)G (2)
where η is the mixing parameter. From the fit the full width
at half maximum (FWHM) and η are calculated for all the
peaks. The Caglioti relation [27]
(FWHM)2 = U tan2 θ + V tan θ + W and (3)
(1 − η) = a + bθ (4)
are used to characterize the instrumental broadening. Five
parameters U , V , W , a, and b are used to synthesize the
instrumental broadening profile at any particular angle cor-
responding to the peak position of the sample. Finally the
instrumental broadening is deconvoluted from the observed
profile using the Stokes deconvolution method [28].
(1) 2.3 Mechanical testing
Samples, with dimensions of 3 mm in diameter and 4 mm in
length, were subjected to the compression test using a servo-
strain curve, the yield strength (σmeasured ) was determined
for all the ECAP-processed specimens.
Microstructural characterization of ultrafine-grain interstitial-free steel by X-ray diffraction line profile 945
Fig. 1 X-ray diffraction
patterns of (a) the starting
material and (b) sample after
4 ECAP passes

3 Methods of analysis broadening creates different asymptotic behavior of the kth

order restricted moments defined as:
In this study we have adopted the following techniques  q k
of line profile analysis to obtain microstructural informa-  −q  q I (q) dq
M k (q ) =  ∞ (6)
tion from the broadened X-ray diffraction profiles of ECAP −∞ I (q) dq
processed IF steel samples.
in which I (q) is the intensity distribution as a function of
q = λ2 [sin(θ ) − sin(θ0 )], where λ is the wavelength of the
3.1 Williamson–Hall technique X-ray, θ is the diffraction angle, and θ0 is the Bragg angle.
Accordingly, the second and the fourth order restricted mo-
Williamson and Hall plot is a classical method to obtain ments have the following asymptotic forms:
a qualitative information of anisotropy in broadening [29].
The basic assumption of this method is that the both size 1 L Λρ ln(q/q0 )
M2 (q) = q− + (7)
and strain broadened profiles are of Lorentzian shape. Based π 2 Ds 4π 2 K 2 Ds2 2π 2
on this assumption, a mathematical relation was established and
between the integral breadth (β), volume weighted average
crystallite size (Dv ), and the microstrain (e) as follows. M4 (q) 1 Λρ 3Λ2 ρ 2  2
= q + + ln (q/q1 ) (8)
  q2 3π 2 Ds 4π 2 4π 2 q 2
β cos θ 1 2 sin θ
= + 2e (5) where K is the Scherrer constant, L is the so-called taper
λ Dv λ
parameter depending on the rate of decrease of the cross-
The plot of ( β cos θ 2 sin θ sectional area of the crystallites. Ds is the average column
λ ) versus ( λ ) gives the value of the
microstrain from the slope and the domain size from the or- length or surface-weighted crystallite size measured in the
direction of diffraction vector. ρ is the average disloca-
dinate intercept. If the points in the WH plot are scattered,
tion density and ρ 2  is the average of square of the dis-
i.e., if ( β cos θ 2 sin θ
λ ) is not a monotonic function of ( λ ), the location density. q0 and q1 are fitting parameters not inter-
broadening is termed as anisotropic.
preted physically. Λ is a geometrical constant describing the
strength of dislocation contrast and is calculated using the
3.2 Variance method anisotropic elastic constants of the material following the
prescription described in the reference [32]. It is seen that
The variance method has been applied to determine the crys- the leading two terms of M2 (q) and the first term of M4 (q)
tallite size and dislocation density from the asymmetric in- originate from the finite crystallite size (Ds ). The third term
tensity profile. The method is based on the analysis of the of M2 (q) and the second term of M4 (q) are due to density of
moments of the intensity profile. Wilson [30] and Groma dislocations ρ. Finally, the last term of M4 (q) corresponds
[19, 31] have shown that the particle size and strain induced to an average of the square of the dislocation density ρ 2 .
946
4 Results and discussions
Figure 2 shows typical WH plots for the starting and the
2 pass deformed sample. The values of integral breadth
as obtained from the XRD patterns clearly indicate an
anisotropic broadening for the 2 pass deformed sample
(Fig. 2). This anisotropic broadening of the XRD profile
is attributed to the presence of dislocations in the mater-
ial. Figure 3 shows the (110) peak of the starting mater-
ial and the 3 pass deformed sample. It is evident that for
the 3 pass deformed sample the XRD peak is broadened
and asymmetric, the latter gives the indication regarding the
formation of subgrains or dislocation cell structure in the
material [33]. The variance method has been proved to be
very effective to account this kind of anisotropic broadening
and asymmetric peak shape of the XRD profile. We have
applied this method to the (110) peak of all the samples.
Fig. 2 Williamson–Hall plot for the starting and 2 pass deformed sam-
ples
Fig. 3 (110) peak of the starting and the 3 pass deformed samples
A. Sarkar et al.
It is to be mentioned here that for bcc materials the (110)
peak is the highest intensity diffraction peak. From the ana-
lyzed data, the typical fourth order moment divided by q 2
is presented in Fig. 4a for the 1 pass and the 4 pass de-
formed samples. Figure 4b shows the representative plot of
M2 for the typical (110) peak for the 1 pass and the 4 pass
samples. The asymptotic regions of the curves of M2 and
M4 /q 2 are fitted with (7) and (8), respectively. From the
above analysis, the regime was chosen for which both M2
and M4 yielded the same surface weighted mean crystallite
size (Ds ) values. The dislocation densities ρ in the differ-
ent samples are determined from the M4 /q 2 vs. q curve.
The variation of the Ds with strain is shown in Fig. 5a.
It can be seen that Ds is reduced to 48 ± 5 nm already
after the second pass, i.e., after a strain of 2.30. There-
after, the crystallite size does not change significantly for
higher strains. It is worth mentioning that for severely de-
formed materials, the mean crystallite size determined by
X-ray line profile analysis is usually lower than the grain
size observed in TEM. The grains in severely deformed ma-
terials are divided into subgrains or dislocation cells which
Fig. 4 Variation of (a) forth order restricted moment (M4 /q 2 ) and
(b) second order restricted moment (M2 ) with q for the (110) peak
Microstructural characterization of ultrafine-grain interstitial-free steel by X-ray diffraction line profile
Fig. 5 Variation of (a) crystallite size and (b) dislocation density with
strain
are separated from each other by low-angle grain bound-
aries. The crystallite size obtained from XRD is equivalent
to the mean size of domains which scatter X-rays coher-
ently. Consequently, X-ray diffraction makes a difference
between dislocation cells or subgrains even if the misori-
entations are very small. The variation of dislocation den-
sity with strain is shown in Fig. 5b. It can be seen from
the figure that the dislocation density increases to 3.6 ×
1014 m−2 after the first pass, i.e., at a strain of 1.15. Dur-
ing further deformation, dislocation density increases rather
slowly.
To correlate the yield strength of the samples with the dis-
location density obtained from the XRDLPA, we have cal-
culated the yield strength using Taylor’s equation
σTaylor = σ0 + αMGbρ1/2
where σ0 is the friction stress, α a constant (α is taken
as 0.33), G the shear modulus, b the length of the Burg-
ers vector, and M is the Taylor factor (M = 3 for untex-
tured polycrystalline materials). For pure Iron, G = 82 GPa,
947
Fig. 6 Values of yield strength measured by compression test
(σmeasured ) vs. values calculated from the dislocation density accord-
ing to (6) (σTaylor )
b = 0.25 nm, and σ0 = 30 MPa. These values are used
to calculate the yield strength of the sample after differ-
ent ECAP passes. The values of σmeasured versus σTaylor are
plotted in Fig. 6. It is apparent that the values of the yield
strength measured by mechanical tests are in good agree-
ment with the values calculated according to Taylor’s equa-
tion.
5 Conclusions
Microstructural evolution during equal channel angular
pressing of interstitial free steel has been reliably as-
sessed by X-ray diffraction line profile analysis using the
Williamson–Hall plot technique and the variance method.
The anisotropic broadening of the line profile of the de-
formed sample provided an estimate of dislocations, while
the asymmetric peak shape confirmed the formation of sub-
grain or cell structure in the deformed samples. Crystallite
size decreased with strain up to a strain level of 2.3 beyond
which it reached saturation. The dislocation density consis-
tently increased with the increasing strain, indicating that no
dynamic restoration processes occurred during equal chan-
nel angular pressing. Yield strength of the samples measured
by mechanical tests are found to correlate very well with the
dislocation density obtained from the X-ray diffraction line
profile analysis.
(9)
Acknowledgements The authors are thankful to Dr. D. Bhattachar-
jee, Dr. N. Gope, and Prof. R.K. Ray, R & D Division, Tata Steel for
their encouragement and support to this work. The financial support
from Tata Steel through a collaborative project “Investigation on Ultra-
fine Grain Steels” is duly acknowledged.
948 A. Sarkar et al.

References 16. S. Suwas, R. Arruffat-Massion, L.S. Toth, J.J. Fundenburger,

A. Eberhardt, W. Skrotzki, Metall. Mater. Trans. A 37, 739 (2007)
1. H. Gleiter, Prog. Mater. Sci. 33, 223 (1989) 17. T. Ungar, Scr. Mater. 51, 777 (2004)
2. C.C. Koch, Nanostruct. Mater. 9, 13 (1997) 18. D. Balzar N. Audebrand, M.R. Daymond, A. Fitch, A. Hewat,
3. R. Valiev, Nat. Mater. 3, 511 (2204) J.I. Langford, A. Le Bail, D. Louër, O. Masson, C.N. McCowan,
4. R.Z. Valiev, R.K. Islamgaliev, I.V. Alexandrov, Prog. Mater. Sci. N.C. Popa, P.W. Stephens, B.H. Toby, J. Appl. Crystallogr. 37, 911
45, 103 (2000) (2004)
5. V.M. Segal, Mater. Sci. Eng. A 197, 157 (1995) 19. I. Groma, Phys. Rev. B 57, 7535 (1998)
6. Y. Saito, N. Tsuji, H. Utsunomiya, T. Sakai, R.G. Hong, Scr. 20. J. Gubicza, N.H. Nam, L. Balogh, R.J. Hellmig, V.V. Stolyarov,
Mater. 39, 1221 (1998) Y. Estrin, T. Ungár, J. Alloys Comp. 378, 248 (2004)
7. Y.T. Zhu, T.C. Lowe, Mater. Sci. Eng. A 291, 46 (2000) 21. J.Y. Chang, G.H. Kim, I.G. Moon, Scr. Mater. 44, 331 (2001)
8. D.H. Shin, I. Kim, J. Kim, K.T. Park, Acta Mater. 49, 1285 (2001) 22. J. Gubicza, Gy. Krallics, I. Schiller, D. Malgin, Mater. Sci. Forum
9. T.G. Langdon, Mater. Sci. Eng. A 462, 3 (2007) 473–474, 453 (2005)
10. I.J. Beyerlein, L.S. Tóth, C.N. Tomé, S. Suwas, Philos. Mag. 87, 23. A.P. Zhilyaev J. Gubicza, G. Nurislamova, A. Revesz, S. Surinach,
885 (2007) M.D. Baro, T. Ungar, Phys. Stat. Sol. (a) 198, 263 (2003)
11. Y.H. Zhao, J.F. Bingert, Y.T. Zhu, X.Z. Liao, R.Z. Valiev, 24. J.P. Mathieu, S. Suwas, A. Eberhardt, L.S. Toth, P. Moll, J. Mater.
Z. Horita, T.G. Langdon, Y.Z. Zhou, E.J. Lavernia, Appl. Phys. Process. Technol. 173, 29 (2006)
Lett. 92, 081903 (2008) 25. D. Bhattacharjee, R.K. Ray, S. Suwas, A. Bhowmik, Patent
12. D. Jia, Y.M. Wang, K.T. Ramesh, E. Ma, Y.T. Zhu, R.Z. Valiev, 26. D. Balzar, in Defect and Microstucture Analysis by Diffraction,
Appl. Phys. Lett. 79, 611 (2001) ed. by R.L. Snyder, J. Fiala, H.J. Bunge (1999), p. 94
13. P.B. Pragnell, J.R. Bowen, A. Gholinia, in Proceedings of the 27. G. Caglioti, A. Paoletti, F.P. Ricci, Nucl. Instr. 3, 223 (1958)
22nd 559 Riso International Symposium on Materials Science, 28. A.R. Stokes, Proc. Phys. Soc. Lond. 61, 382 (1948)
Science of Metastable and Naocrystalline Alloys, ed. by A.R. Di- 29. G.K. Williamson, W.H. Hall, Acta Metall. 1, 22 (1953)
nesen M. Eldrup, D. Juul Jensen, S. Linderoth, T.B. Pederson, 30. A.J.C. Wilson, Proc. Phys. Soc. 80, 286 (1962)
N.H. Pryds, A. Schroder Pedersen, J.A. Werst, Roskilde, Denmark 31. A. Borbely, I. Groma, Appl. Phys. Lett. 79, 1772 (2001)
(2001), p. 105 32. T. Ungar, G. Tichy, Phys. Stat. Sol. (a) 171, 425 (1999)
14. J.R. Bowen, P.B. Pragnell, F.J. Humphreys, Mater. Sci. Technol. 33. T. Ungar, H. Mughrabi, D. Rönnpagel, M. Wilkens, Acta Metall.
16, 1246 (2000) 32, 333 (1984)
15. W. Skrotzki, N. Schherbaum, C.-G. Oertel, R. Arruffat-Massion,
S. Suwas, L.S. Toth, Acta Mater. 55, 2013 (2007)