Quantifying the effects of surface quality on texture measurements of tantalum

Page 1

Qt

HC

a

ARRAA

KTSLX

1

rdlpltstetcr[f[sr[i

h0

Applied Surface Science 339 (2015) 15–21

Contents lists available at ScienceDirect

Applied Surface Science

jou rn al h om ep age: www.elsev ier .com/ locate /apsusc

uantifying the effects of surface quality on texture measurements ofantalum

.Y. Fan, S.F. Liu ∗, Y. Guo, C. Deng, Q. Liuollege of Materials Science and Engineering, Chongqing University, No. 174 Shazheng Street, Shapingba District, Chongqing 400044, China

r t i c l e i n f o

rticle history:eceived 20 October 2014eceived in revised form 24 January 2015ccepted 28 January 2015vailable online 7 February 2015

eywords:

a b s t r a c t

Surface quality plays an important role in texture measurements of the materials with high linear absorp-tion coefficient. In this article, three different surface preparation methods were used and the effects ofsurface quality on texture measurements were quantitatively investigated. A bulk annealed tantalumwas selected as a model material, whose surface was subjected to grinding, electropolishing and etching,respectively. In the case of the three surface states, macrotexture analyses were performed on an X-raydiffraction instrument and features of the surfaces were examined by optical microscope and surface pro-

exture measurementurface qualityinear absorption coefficient-ray diffraction (XRD)

filer. Comparisons were made based on matching the percentages of two primary textures ({1 0 0}<uvw>and {1 1 1}<uvw>) obtained by X-ray diffraction (XRD) and electron back-scattering diffraction (EBSD)methods. It was revealed that states of sample surface, like roughness and damaged region caused bygrinding, have a significant impact on texture results. Electropolishing was proved to be suitable formetals with high linear absorption coefficient for XRD texture measurements.

© 2015 Elsevier B.V. All rights reserved.

. Introduction

Texture governs many properties of materials, such as the cor-osion behavior, sputtering characteristics of targets and deeprawability [1–4]. With increasing demand on quantitatively ana-

yzing the relationships between texture and some properties ofolycrystalline materials, an accurate characterization of crystal-

ographic orientations is extremely important. X-ray diffractionechnique is known as a universal method for macrotexture analy-is. Global texture of a flat sample can be readily detected based onhe Bragg’s law. Different from aluminum and magnesium, how-ver, some metals with high atomic number Z and density �, e.g.antalum (Z = 73, � = 16.69 g cm−3), show a high linear absorptionoefficient for X-rays. Such metals absorb large quantities of X-ays and thereby greatly reduce the penetration depth of X-rays5]. Thus, information of grain orientations can only be detectedrom a limited region of sample. Proceeding from PCapková’s work6], texture measurements are detrimentally influenced by surfacetate for materials with such a coefficient. For example, surface

oughness can cause an appreciable loss in intensity of X-rays7–9]. Obtaining superior results of X-ray texture analysis, luckily,s possible with proper sample preparation methods [10]. In terms

∗ Corresponding author. Tel.: +86 23 65106024; fax: +86 23 65106407.E-mail addresses: [email protected], [email protected] (S.F. Liu).

ttp://dx.doi.org/10.1016/j.apsusc.2015.01.216169-4332/© 2015 Elsevier B.V. All rights reserved.

of tantalum, surface treatments have been attracted much atten-tion [11–13]. Also, various sample preparation methods has beenadopted in previous studies [12,14,15]. Unfortunately, the effectsof surface finishing on texture measurements by XRD have yet tobe estimated, let alone the quantitative analysis.

Based on the above consideration, there is no doubt that thesample preparation of metals with high linear absorption coef-ficient should be taken seriously. In this article, tantalum wasselected as a representative metal and three common preparationmethods, grinding, electropolishing and chemical etching, wereadopted. Combined with EBSD technique, it is expected that sometheories for texture measurement of tantalum can be establishedand extended to some other metals with high linear coefficientabsorption.

2. Experiments

2.1. Material and sample preparation

A bulk annealed tantalum (1300 ◦C for 1 h, vacuum environ-ment) of 99.95% purity served as the sample. The sample size was12L × 10W × 3T mm3 and the average grain size was 100 �m, with

a standard deviation of 60 �m. For the sake of obtaining textureinformation as much as possible from the same grain layer, thethree polishing processes should be carried out successively. Ofcourse, sample preparation in practice for some metals does not

Page 2

1 rface S

ia

fpdab

Ft

6 H.Y. Fan et al. / Applied Su

nvolve this consideration. Specific experimental procedures weres follows.

The sample was ground using progressively finer SiC papersrom 1000, 3000 grits and proceeding to 5000 grits. Grinding waserformed by hand and water was continuously used to rinse the

ebris. The polishing processes were gently carried out to ensures few damages caused by friction. Finally, sample was cleanedy absolute ethanol and dried with a jet of dry air. The cleaning

ig. 1. Optical micrographs of sample surfaces obtained under different treatment conditihem, (b, d) were taken under bright-field (BF) condition; (c, e) were taken under differen

cience 339 (2015) 15–21

and drying processes were the same in electrolytic polishing andetching.

The ground Ta sample was polished for about 9 min in a 10 mlHF/90 ml H2SO4 electrolyte [16]. Sample and graphite were usedas the anode and cathode, respectively. The current density was

0.1 A/cm2 and the temperature of the electrolyte was maintainedat ambient temperature. In addition, slowly stirring was essentialto promote the ionic exchange.

ons. (a) Mechanical grinding; (b, c) Chemical etching; (d, e) Electropolishing. Amongtial interference contrast (DIC) condition.

Page 3

rface Science 339 (2015) 15–21 17

5dmb

2

OtrmI4

2

m(wtotm

shpadi

oepsoasp5q{r

3

3

vbwsca(r[wmi(

H.Y. Fan et al. / Applied Su

Etching was carried out after electropolishing. Etchant was0 ml H2SO4/20 ml HNO3/20 ml HF [17]. The specimen wasrenched in this etchant for 20 s at ambient temperature. Thisethod could not only eliminate the line-shaped surface scratches,

ut also present better defined grain boundaries.

.2. Metallography and surface roughness

Optical microscopes (Zeiss Axiovert 40MAT microscope andlympus DSX 500 microscope) were used to reveal the surface fea-

ures, such as scratches, pits, impurities and surface flatness. Theoughness values of the above three surfaces were finally deter-ined by a Dektak 150 surface profilometer (Veeco Instruments,

nc., USA). The scan length on the sample surface was set to be000 �m, using a stylus radius of 2.5 �m.

.3. Texture measurements

Macrotexture was analyzed by an X-ray diffraction instru-ent (D/max 2500PC, Rigaku Co., Japan) with Cu K� radiation

40 kV/150 mA). Pole figures of {1 1 0}, {2 0 0}, {2 1 1} and {2 2 2}ere recorded by the Schultz back reflection technique and the

ilt angle ran from 20◦ to 90◦ with a step size of 5◦. With the aidf software LaboTex version 3.0, the orientation distribution func-ions (ODFs) were calculated with Arbitrarily Defined Cells (ADC)

ethod [18].EBSD technique is able to readily determine texture based on the

tatistics of the orientations of individual grains [19]. Metals withigh atomic number (e.g. Tantalum) can generate higher qualityatterns than some light metals (e.g. Aluminum) [20]. Taking thesedvantages into account, EBSD should be the most accurate tool foretecting the crystallographic orientations of tantalum. Therefore,

t was selected as a reference standard for the XRD texture analysis.The grain orientations were identified by HKL-EBSD equipped

n a Focused Ion Beam (Auriga FIB/SEM, Carl Zeiss, Germany). Thelectropolished surface was scanned at a step size of 8 �m. Kikuchiatterns were acquired at an acceleration voltage of 20 kV with aample title 70◦. An area (9 × 4 mm2) consisting of eight regionsf the same size containing at least 3500 grains was scanned. Therea was large enough for a sharp textured tantalum to achievetatistical reliability [21,22]. The hit rate (fraction of indexed EBSDatterns) was more than 95% before image noise reduction. Channel

from HKL Technology/Oxford Instruments was used for subse-uent data analysis. Then the area fractions of {1 0 0}<uvw> and1 1 1}<uvw> were calculated (deviation angle 15◦) from EBSD data,espectively.

. Results

.1. Surface morphology and roughness

As evident from Fig. 1, different surface treatments give rise toarious surface features. The micrographs (b) to (e) were takeny an optical microscope (Opto-digital DSX500, Olympus, Japan),hich can explain details better than some conventional micro-

copes. Numerous scratches and inclusions distributing randomlyan be revealed from the ground surface (Fig. 1a). Etching leads to

scratch-free surface and general metallographic microstructureFig. 1b). However, the rates of etchant attack are different at theegions with different grain orientations, crystalline imperfections23,24]. Surface roughness appears in this case (Fig. 1c). Compared

ith the above two methods, electropolishing is comparativelyild and creates a mirror-like surface (Fig. 1d), whose differential

nterference contrast (DIC) image reveals a considerably flat surfaceFig. 1e).

Fig. 2. XRD patterns from different surfaces. Rq is the root-mean-square deviationof roughness.

Roughness values of different surfaces were measured by Dek-tak 150 and quantitatively described by “Rq”, respectively. Rq isthe root-mean-square deviation from profile mean over samplingarea, which is often referred to be better than Ra (average rough-ness) to reflect surface topography [25]. Rq values are 146, 345 and67 nm for the surfaces of grinding, etching and electropolishing,respectively.

3.2. X-ray diffraction patterns

Fig. 2 shows the XRD patterns corresponding to different prepa-ration methods. Theoretically, diffraction peaks of pure Ta appearat 2� of 38.472◦, 55.549◦, 69.581◦, 82.461◦, 94.936◦ and 107.640◦

which are assigned (h k l) values as (1 1 0), (2 0 0), (2 1 1), (2 2 0),(3 1 0) and (2 2 2) planes, respectively [26]. The positions and shapesof diffraction peaks from ground surface are changed when com-pared with the other two spectra. Peaks, such as those of (2 0 0) and(2 2 2) planes, exhibit broadening effect and shift slightly from theideal angles of diffraction. There are normal numbers of peaks in theangle range (35–110◦), whereas some peaks vanish after etchingand electrolytic polishing. Also, the significant discrepancy of peakintensity needs to be focused on, especially the (2 0 0) diffractionpeak.

3.3. Texture

Quantitative analysis of texture requires the calculation of theorientation distribution function (ODF) from pole figures. The Ori-entation Distribution Function (ODF) is defined as follows [17]:

dV

V= f (g)dg (1)

where f (g) is the value of ODF, i.e. orientation density; each point{ϕ1, ˚, ϕ2

}within the Euler space represents an orientation g to

which many grains belong and dg is the element of the orienta-tion; dV defines the volume of all crystals which have orientation gwithin the element of orientation space dg; V is the sample volume.Random distribution of such points in the Euler space, i.e. f (g) ≡ 1,indicates that the material is textureless. On the contrary, a cluster

region of the points means presence of a texture and the value oforientation density f (g) increases with the increase of the clusterlevel of similar orientations. Therefore, the value of f (g)max signifiesthe largest cluster degree of the primary orientations.

Page 4

18 H.Y. Fan et al. / Applied Surface Science 339 (2015) 15–21

Fig. 3. ODFs (XRD) of (a) Ground surface, (b) Etched surface and (c) Electropolished surface. CONST ANGLE: ϕ2 = 0–90◦ , � = 5◦ . f (g) represents the orientation density levels.(d) Orientation map of the electropolished sample surface.

Fig. 4. Volume fractions of {1 0 0}<uvw> and {1 1 1}<uvw> grains calculated from the XRD and EBSD data.

Page 5

H.Y. Fan et al. / Applied Surface Science 339 (2015) 15–21 19

Table 1The evolution of Gx with depth x (�m) for different diffraction planes in tantalum.

Bragg angle � (◦) Diffraction plane Penetration depth x (�m)into tantalum, Cu K�

Gx = 95% Gx = 50%

19.236 (1 1 0) 1.8 0.427.774 (2 0 0) 2.6 0.634.791 (2 1 1) 3.1 0.753.821 (2 2 2) 4.4 1.0

aetgftv

tmeaupppfemot

4

4

�ttutdiif

G

waci

oaFd5

Fig. 5. Schematic representation of three possible cases of the reflected beam.

mainly consists of plastic deformation and residual stress extend-ing several microns into the sample [25,29], as shown in Fig. 6. Thenumber of diffraction peaks corresponding to the grinding surface

90 – 5.5 1.3

Three ODFs corresponding to different surface treatments and combined EBSD map are presented in Fig. 3. The biggest differ-nce between the ODFs is the value of f (g)max. The max intensity ofhe electropolishing surface is considerably larger than that of therinding surface and almost the same as that of the etching sur-ace. Also considering the EBSD map, it is seen that the dominatingextures in the sample are {1 0 0} and {1 1 1} components, whoseolume fractions are shown in Fig. 4.

The histogram intuitively presents the volume fractions of thewo textures under different conditions. There is a rough agree-

ent between the EBSD and XRD results corresponding to thelectropolishing surface. The maximum percentages of both {1 0 0}nd {1 1 1} grains are measured by EBSD, and the minimum val-es belong to the grinding surface. Even more than ten percentageoints distinguish the two sets of results from each other. Com-aring the results of etching and electropolishing, an interestinghenomenon worth thinking about is the difference of the volumeractions corresponding to {1 0 0} and {1 1 1} textures between thetched and electropolished surfaces. The fraction of {1 0 0} textureeasured from etched surface is about five percent less than that

f the electropolished one while there is hardly no discrepancy forhe {1 1 1} texture.

. Discussion

.1. Working depth of X-rays

In the case of materials with high atomic number Z and density, the ability of XRD to collect surface information is impeded by

he limited penetration depth of X-rays. This phenomenon leadso an inaccurate diffraction pattern and thereby causes a faultynderstanding of crystal orientations. Blanton et al. [5] reportedhat factors such as sample thickness and density, Bragg angle ofiffraction, and the mass absorption coefficient all affect the work-

ng depth of X-rays. The relationship between the total diffractedntensity and the corresponding working depth can be estimatedrom the following Eq. (2) [27]:

x = 1 − exp(−2�x

sin �

)(2)

here Gx is the fraction of total diffracted intensity originated from depth of x (cm). The linear absorption coefficient � (cm−1) islosely connected with the specific gravity � (g cm−3) thereby yield-ng the mass absorption coefficient �/� (cm2 g−1).

Thus, for Cu K� radiation (� = 0.15184 nm), texture informationf Ta (� = 16.69 g cm−3, �/� = 164 cm2 g−1 [28]) is only derived from

shallow zone within a few microns depth as calculated in Table 1.urther, 95% of the diffracted intensity of (2 0 0) and (2 2 2) planes

erives from regions within 2.6 and 4.4 �m, respectively, whereas0% of that comes from the depth of 0.6 and 1.0 �m.

Effects of surface quality on X-ray diffraction. (a) Idealized X-ray diffraction on aperfectly smooth surface. (b) Effects of the surface quality. (I) Reflected beam isshielded; (II) Normal diffraction; (III) The diffuse scattering of reflected beams.

4.2. Effects of surface quality

As analyzed above, measurement must be particularly sensi-tive to the surface quality for such a shallow working depth ofX-rays. Effects of the surface quality can be summarized in Fig. 5.Fig. 5a shows a perfect X-ray specular reflection process, whileFig. 5b describes three cases of X-ray acting on a rough surface.Two anomalous ways, absorption and diffuse scattering, are sev-erally presented as (I) and (III). This two modes of interaction candefinitely decrease the diffraction intensity by absorbing and redi-recting beam into other directions, respectively.

The etching surface is much rougher than the other two surfaces,as shown in Fig. 2. With the increase of roughness, intensity of theprimary diffraction peaks decreases, especially the (2 0 0) plane. Asseen in Fig. 6, due to the limited detection depth of X-rays, thedepth x that supports the reflecting intensity is becoming smallerand smaller with the decrease of the angle �. According to Eq. (2),(50)% of the intensity comes from the thickness of 0.6 �m for the(2 0 0) diffraction peak. However, the roughness corresponding togrinding and etching are about 150 and 350 nm, respectively. Asshown in Fig. 2, the intensity of (2 0 0) diffraction peak is notice-ably volatile under different roughness conditions, while that of(2 2 2) peak remains broadly stable. It shows that diffraction atlower Bragg angle is much more sensitive to the surface roughness.In addition, the interesting discrepancy between etching and elec-tropolishing also proves this point (Section 3.3, Paragraph 2). Thevolume fractions of {1 1 1} texture have no obvious difference butthose of {1 0 0} texture do not. Actually, the prominent differenceswith respect to the above two surfaces are the surface topographyand corresponding roughness (Fig. 1).

Compared with etching and electropolishing, grinding intro-duces simultaneously a rough surface and damaged zone. This layer

Fig. 6. Schematic diagram showing the effects of the surface quality on the X-raytexture measurement.

Page 6

20 H.Y. Fan et al. / Applied Surface Science 339 (2015) 15–21

Table 2Evolution of Gx = 95% with working depth x (�m) analyzed at a Bragg angle � = 20◦ .

Metal (atomic number) � (g cm−3) X-ray Cu K� (� = 0.15418 nm)

�/� (cm2 g−1) � (cm−1) Working depth (�m)

Mg (12) 1.7 38.6 65.6 77.7Al (13) 2.7 48.6 131.2 38.9Hf (72) 13.3 157 2088.1 2.4Ta (73) 16.6 164 2722.4 1.9W (74) 19.4 171 3317.4 1.5

1

mfrrcbavatpstsapcpc

iarasBnot

4

fadf

Xlttamttaawp

ft

Pb (82) 11.3 24

ay be attributed to the role of plastic deformation. The externalorces introduced by grinding may cause subdivision, rotation andeorientation of some grains and therefore a plastically deformedegion forms. Texture of the deformed layer is obviously changedompared to the original state, which gives a more random distri-ution of crystal orientations. In this context, ODFs in Fig. 3 can given explanation. The max intensity of ODF from grinding surface isery weak with f (g)max = 8.058 in comparison of that of etchingnd electropolishing surfaces. It means that there is an increase inhe randomization of the surface texture [30]. Due to the limitedenetration depth of X-rays, it follows that the volume fractions ofome primary orientations decrease greatly. In addition, as men-ioned by Hyatt and Bieler [31], the inhomogeneous distribution oftrain field can induce variation in intensity of pole figure as wells a small shift in Bragg angles during XRD measurements. Latticearameters of the affected crystallites may change slightly, whichauses a slight shifting in angles of the corresponding diffractioneaks [32], as proved by the (2 0 0) and (2 2 2) peaks (Fig. 2). In thisase, some deviations from the exact texture results are inevitable.

The above analyses prove that the surface quality plays anmportant role in texture measurements. The roughness and dam-ges coexist in the grinding sample and thereby lead to the worstesults. However, the influence degree of the roughness and dam-ged zone is different. Roughness reduces the volume fraction ofome textures slightly, especially diffraction that occurs at lowerragg angles, while the damaged zone can introduce significantumber of other random textures. Fig. 4 illustrates that the resultsf etching surface is far more accurate than that of grinding, thoughhe grinding surface is smoother.

.3. Metals with high linear absorption coefficient

Influences of surface quality on the XRD texture measurementor tantalum has been verified. The working depth of X-rays islmost entirely responsible for this trouble [33]. X-ray workingepths of some familiar metals are represented in Table 2. Dataor X-rays is taken from Ref. [28].

Compared with magnesium and aluminum, working depth of-rays for some metals with a high linear absorption coefficient �,

ike Ta, is limited to a few microns. This issue has a strong impact onexture determination by XRD. Although a high-quality EBSD pat-erns can be acquired for this kind of metals, it is a time-consumingnd costly method, after all. XRD is still an irreplaceable approach toeasure macrotexture. In this case, sample quality is what we need

o focus on prior to XRD texture analysis. To avoid the introduc-ion of scratches, internal strains, damaged layer and some otherdverse factors which may mislead texture statistics, grinding isbsolutely unsuitable for the final sample preparing stage. Gentleater grinding followed by electropolishing is supposed to be a

referred way to prepare these samples.

Texture information derives from a thin region of sample sur-ace, which is far less than thickness of the first layer of grains. Sohere may not be sufficient grains being irradiated to contribute

2723.3 1.9

to an effective diffraction data. Hence, keeping sample vibratingalong transverse direction and using sample with a larger size arenecessary for XRD texture measurement.

5. Conclusions

This study reveals the absolute necessity of assessing the effectsof specimen preparation on XRD texture analysis. From the presentquantitative investigation, some major conclusions can be drawn.

1. The mechanical mean is effective to clean sample but its surfaceis damaged at the same time. Etching is good at removing thedamaged region while the considerable roughness is inevitable.In order to obtain a relatively reliable XRD data for metalswith high linear absorption coefficient, measurement should beperformed on a grinded and following electrolytically polishedsample.

2. Effects of surface roughness on XRD texture measurement areeven more significant for planes at lower Bragg angles. Dataof {1 0 0} texture is less reliable with regard to tantalum whensample surface is rough, like etching surface.

3. The damaged region has a far greater impact than roughness onthe texture measurement with respect to the grinding sample.

Acknowledgements

The present work was supported by the National Natural Sci-ence Foundation of China (Grants 51301212), the Major NationalScience and Technology Projects of China (No. 2011ZX02705), theChongqing Science and Technology Commission in China (CSTC,2012jjA50023), and the Fundamental Research Funds for the Cen-tral Universities of China (Project No. CDJZR11130010).

References

[1] Y. Yazawa, Y. Ozaki, Y. Kato, O. Furukimi, Development of ferritic stainlesssteel sheets with excellent deep drawability by {1 1 1} recrystallization texturecontrol, JSAE Rev. 24 (2003) 483–488.

[2] D. Field, J. Yanke, E. Mcgowan, C. Michaluk, Microstructural developmentin asymmetric processing of tantalum plate, J. Electron. Mater. 34 (2005)1521–1525.

[3] L. Margulies, G. Winther, H. Poulsen, In situ measurement of grain rotationduring deformation of polycrystals, Science 291 (2001) 2392–2394.

[4] B. Ravi Kumar, R. Singh, B. Mahato, P. De, N. Bandyopadhyay, D. Bhattacharya,Effect of texture on corrosion behavior of AISI 304L stainless steel, Mater. Char-act. 54 (2005) 141–147.

[5] T. Blanton, C. Barnes, M. Lelental, The effect of X-ray penetration depth onstructural characterization of multiphase Bi–Sr–Ca–Cu–O thin films by X-raydiffraction techniques, Phys. C: Supercond. 173 (1991) 152–158.

[6] P. Capková, V. Valvoda, The surface roughness effect in texture measure-ments by the Schulz X-ray reflection method, Phys. Status Solidi (a) 81 (1984)K113–K115.

[7] W.K. Kuhn, G. Andermann, Effects of surface roughness on surface analysis viasoft and ultrasoft X-ray fluorescence spectroscopy, Appl. Surf. Sci. 185 (2001)84–91.

[8] E. Vlieg, ROD: a program for surface X-ray crystallography, J. Appl. Crystallogr.33 (2000) 401–405.

Page 7

rface S

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[[[

[

[

H.Y. Fan et al. / Applied Su

[9] R. M’Saoubi, T. Larsson, J. Outeiro, Y. Guo, S. Suslov, C. Saldana, S. Chandrasekar,Surface integrity analysis of machined Inconel 718 over multiple length scales,CIRP Ann. Manuf. Technol. 61 (2012) 99–102.

10] G. Kimmel, D. Dayan, G. Frank, A. Landau, X-ray diffraction (XRD) character-ization of microstrain in some iron and uranium alloys, Israel Atomic EnergyCommission, Annual report, 1996, pp. 28–56.

11] A. Kelly, S. Bingert, R. Reiswig, New Metallographic Preparation Techniques forTantalum and Tantalum Alloys, Los Alamos National Lab., NM (United States),1995.

12] C.A. Michaluk, M.M. Nowell, R.A. Witt, Quantifying the recrystallization textureof tantalum, JOM 54 (2002) 51–54.

13] O. Piotrowski, C. Madore, D. Landolt, Electropolishing of tantalum in sulfuricacid–methanol electrolytes, Electrochim. Acta 44 (1999) 3389–3399.

14] K.H. Song, H.S. Kim, W.Y. Kim, Texture development in cold rolled and annealedTa, Mater. Sci. Forum, Trans Tech Publ 695 (2011) 153–157.

15] Z. Zhang, D. Chen, H. Zhao, S. Liu, A comparative study of clock rolling and unidi-rectional rolling on deformation/recrystallization microstructure and textureof high purity tantalum plates, Int. J. Refract. Metals Hard Mater. 41 (2013)453–460.

16] ASTM, E 1558-09, Standard guide for electrolytic polishing of metallographicspecimens, American Society and Materials, US, 2009.

17] O. Engler, V. Randle, Introduction to Texture Analysis: Macrotexture, Microtex-ture, and Orientation Mapping, CRC press, Boca Raton, 2009.

18] K. Pawlik, Determination of the orientation distribution function from polefigures in arbitrarily defined cells, Phys. Status Solidi (b) 134 (1986) 477–483.

19] C. Briant, E. MacDonald, R. Balliett, T. Luong, Recrystallization textures in tan-talum sheet and wire, Int. J. Refract. Metals Hard Mater. 18 (2000) 1–8.

20] N. Mateescu, M. Ferry, W. Xu, J. Cairney, Some factors affecting EBSD patternquality of Ga+ ion-milled face centred cubic metal surfaces, Mater. Chem. Phys.106 (2007) 142–148.

[

[

cience 339 (2015) 15–21 21

21] R. Penelle, T. Baudin, P. Paillard, L. Mora, Characterization of recrystallizationtextures in Fe–3% Si sheets by EBSP: comparison with X-ray diffraction, Tex-tures Microstruct. 14 (1991) 18.

22] S.I. Wright, M.M. Nowell, J.F. Bingert, A comparison of textures measured usingx-ray and electron backscatter diffraction, Metall. Mater. Trans. A 38 (2007)1845–1855.

23] H. Baker, H. Okamoto, ASM Handbook, Alloy Phase Diagrams, 3, ASM Interna-tional, Materials Park, OH, 1992, pp. 2.

24] M. Dahlstrøm, Effect of Copper Content on etching Response of Aluminum inAlkaline and Acid Solutions, Norwegian University of Science and Technology,Norway, 2012.

25] X. Wen, X. Wang, N. Zhang, Microrough surface of metallic biomaterials: aliterature review, Bio-Med. Mater. Eng. 6 (1996) 173–189.

26] PDF04-0788, The International Centre for Diffraction Data, JCPDS-ICDD, JCPDS-International Center for Diffraction Data, 12 Campus Boulevard, NewtownSquare, PA 19073-3273, USA, 1998.

27] B.D. Cullity, S.R. Stock, Elements of X-ray Diffraction, Pearson, 2001.28] C.S. Barrett, T.B. Massalski, Structure of metals, McGraw-Hill, New York, 1966.29] P.S. Prevéy, Current applications of X-ray diffraction residual stress mea-

surement, in: Developments in Materials Characterization Technologies, ASMInternational, Materials Park, OH, 1996, pp. 103–110.

30] M. Huh, S. Cho, O. Engler, Randomization of the annealing texture in aluminum5182 sheet by cross-rolling, Mater. Sci. Eng. A 315 (2001) 35–46.

31] S. Hyatt, T. Bieler, The effect of a superimposed strain field on frictional texturemeasurements, Wear 184 (1995) 245–247.

32] B. Fultz, J. Howe, Diffraction and the X-ray powder diffractometer, in: Trans-mission Electron Microscopy and Diffractometry of Materials, Springer, Berlin,2013, pp. 1–57.

33] I. Tomov, Orientation dependence of the effective depth of X-ray penetration,Phys. Status Solidi (a) 98 (1986) 43–50.