Materials Science & Engineering A 675 (2016) 92–98

Contents lists available at ScienceDirect

Materials Science & Engineering A

journal homepage: www.elsevier.com/locate/msea

Effect of oxygen and nitrogen on microstructure and mechanical

properties of vanadium
Min-Gu Jo a,b, Phaniraj P. Madakashira b,n, Jin-Yu Suh a, Heung Nam Han b,n
a
High Temperature Energy Materials Research Center, Korea Institute of Science and Technology, Seoul 136-791, Republic of Korea
b
Department of Materials Science and Engineering and Research Institute of Advanced Materials, Seoul National University, Seoul 151-744, Republic of Korea

art ic l e i nf o a b s t r a c t

Article history: The effect of oxygen and nitrogen content on the microstructure and mechanical behavior of vanadium
Received 11 July 2016 was investigated. Vanadium specimens containing 40–4536 ppm nitrogen and 624–9092 ppm oxygen,
Received in revised form respectively were prepared using diffusion heat treatments. The specimens were characterized with
9 August 2016
respect to chemical composition, microstructure and mechanical properties. Both V-O specimens and
Accepted 10 August 2016
V-N specimens had single phase microstructure with no precipitates. Increase in oxygen and nitrogen
Available online 11 August 2016
content increased hardness and tensile strength and decreased ductility. The specimens were char-
Keywords: acterized for grain boundary segregation using scanning transmission electron microscopy (STEM)
Vanadium equipped with super-X EDS and high resolution ion microprobe (NanoSIMS). The mechanical properties
Oxygen
were discussed in view of the measurements of composition and microstructure.
Nitrogen
& 2016 Elsevier B.V. All rights reserved.
Mechanical Properties
Segregation

1. Introduction
Vanadium based alloys have been identified as potential can-
didates as membrane materials in hydrogen separation and pur-
ification technologies [1,2], hydrogen storage applications [3] be-
sides fusion reactor structural applications [4]. However, vanadium
has a high affinity for oxygen and nitrogen and can absorb con-
siderable amounts from the environment during fabrication, heat
treatment processing and while in use. Both oxygen and nitrogen
interstitials strengthen vanadium through solid solution
strengthening but can adversely affect ductility leading to failure
at high concentrations [5–8]. It is therefore important to under-
stand the influence of high concentration of interstitials on me-
chanical properties of vanadium.
Typically both oxygen and nitrogen are present in vanadium
specimens making it difficult to determine their individual effects.
Nevertheless the increase in hardness independently due to oxy-
gen and nitrogen (Hv/wppm) has been reported [6,7,9]. Typically,
nitrogen interstitials increases the hardness of vanadium much
more than oxygen. Effect of nitrogen and oxygen on the tensile
properties of vanadium has also been reported [5,6,8]. Loria [8]
reported that at the same impurity level nitrogen produced a two
fold increase in strength when compared with oxygen. Kainuma
n concentrations are higher than those reported in published lit-
Corresponding authors.
E-mail addresses: mphaniraj@gmail.com (P.P. Madakashira),
hnhan@snu.ac.kr (H.N. Han).
et al. [6] reported that the yield strength increased as the oxygen
or nitrogen content increased, with nitrogen being more effective.
The elongation correspondingly decreased with increasing oxygen
or nitrogen content. At concentrations above
0.48 wt% O and

0.24 wt% N in V-O and V-N alloys the specimens fractured
without plastic deformation.
Interstitials can promote brittle behavior by segregating to
grain boundaries or forming precipitates. They can also stay in
solid solution and inhibit dislocation movement and reduce re-
sistance to crack propagation [10]. Most research (for e.g. Ref.
[11,12]) points to the drop in ductility with increasing nitrogen
and oxygen content but few investigations examine their role in
embrittling vanadium. In V-Cr-Ti alloys the intergranular failure of
the specimens is speculated to be due to segregation of oxygen to
the grain boundary [13–15]. In V-N alloys the loss in ductility was
reported [11] to be due to cracking initiated at grain boundaries. In
Nb [16] and Ta [17], which along with vanadium are group Va
elements, first principles calculations showed that oxygen and
nitrogen interstitials segregate to grain boundary. However, there
are no reports, to the authors’ knowledge, experimental or theo-
retical on the segregation of interstitials in vanadium.
In the present study, microstructure and tensile behavior of
vanadium specimens containing 40–4536 ppm nitrogen and 624–
9092 ppm oxygen, respectively were investigated. These solute
erature. The change in hardness with oxygen and nitrogen (Hv/
wppm) was contrasted with those reported in published literature.

http://dx.doi.org/10.1016/j.msea.2016.08.040
0921-5093/& 2016 Elsevier B.V. All rights reserved.
 M.-G. Jo et al. / Materials Science & Engineering A 675 (2016) 92–98 93

Fig. 1. Microstructure and tensile stress-strain behavior of annealed vanadium.

Table 1 Sub-size tensile specimens with a gage length of 8 mm were wire

Composition of vanadium specimens (ppm by weight). cut from these strips. The specimens were then annealed for 1h at
1200 °C in high purity argon (99.999%) gas, which was further
Notation Nitrogen Oxygen Vanadium
purified with a gas purifier that reduced the oxygen content to
Vacuum annealed* Annealed 1 463 Balance
10  10 wppm. Additionally the specimens were surrounded by
V-N specimens V-N1 40 1132 titanium pieces to getter argon gas.
V-N2 412 889
The vanadium specimens with different nitrogen and oxygen
V-N3 727 1157
V-N4 1028 735
content were prepared using diffusion heat treatment as follows.
V-N5 4536 1038 Quartz tube containing the annealed specimen was evacuated
V-O specimens V-O1 29 624 and purged with argon repeatedly and then filled with nitrogen at
V-O2 2 1538 pressures between 0.02–0.05 MPa and sealed. The sealed quartz
V-O3 1 2080
tubes were held at 700 °C for 264 h and quenched. Similarly, the
V-O4 6 3008
V-O5 1 9092 vanadium specimens containing different amounts of oxygen were
prepared using evacuated and purged quartz tubes filled with
*
The results of carbon and sulfur analysis were: 270 wppm carbon, 37 wppm oxygen at pressures between 0.005–0.07 MPa, held at 650 °C for
sulfur
144 h and quenched. Another set of specimens were prepared by
heat treating them in evacuated quartz tubes at the same tem-
perature-time conditions that were used for nitrogen and oxygen
treatments but without oxygen or nitrogen. These specimens were
used to compare the effect of oxygen and nitrogen on grain size.
All quartz tubes were sealed when the pressure was r0.013 Pa.
The specimens’ heat treated in nitrogen and oxygen will hence-
forth be referred to as V-N and V-O specimens, respectively.
Eltra ON-900 was used to determine the oxygen and nitrogen
contents of the specimens. LECO CS600 was used to determine
carbon and sulfur contents in the annealed specimen. Gatan Mi-
croTest300N was used to conduct tensile tests at the crosshead
speed of 1.0 mm/min (strain rate
2.1  10  3 s  1). Hardness was
measured using 100 g load on Mitutoyo HM-122 vickers hardness
testing machine. X-ray diffraction (Bruker D8 ADVANCE, Cu target
40 kV, 40 mA) was used to monitor alloying and phase formation
in V-N and V-O specimens using the step size: 2θ ¼0.02° and time
per step of 2 s. The microstructure was characterized using scan-
ning electron microscopy (FEI SEM, Inspect F50) and high resolu-
Fig. 2. Oxygen and nitrogen content of the vanadium specimens. tion scanning transmission electron microscopy (STEM, Talos
F200X) equipped with advanced energy dispersive spectrometer
The high resolution ion-microprobe (NanoSIMS) and STEM-EDS (Super-X EDS). The specimens for STEM were prepared using the
were used to determine segregation at grain boundaries. The focused ion beam technique (FIB Nova 600). High resolution ion-
mechanical properties were discussed in view of the measure- microprobe (Cameca NanoSIMS 50) was used to determine ele-
ments of composition and microstructure. ment composition maps and line scans across grain boundaries. A
focused 16 keV Cs þ ion beam with beam current 0.4pA was used
for all experiments and the following ion species were collected
2. Experimental procedure simultaneously: 16O  , 12C  , 12C 14N  , 32S  and 51V16O  . The
12 14 
C N signal was chosen in preference to the 14N  signal be-
Granules of electro-refined grade of vanadium were arc-melted cause it gives the highest signal intensity. Prior to each measure-
and suction cast into ingots of dimension 10  40  4 mm. These ment the specimen surface was cleaned with a high current pri-
were then cold rolled (90% reduction) into 0.3 mm thick strips. mary ion beam.
94 M.-G. Jo et al. / Materials Science & Engineering A 675 (2016) 92–98

Fig. 3. X-ray diffractograms of a) nitrogen treated and b) oxygen treated specimens.

Fig. 5. Stress strain curves of a) V-N and b) V-O specimens.

Fig. 4. Microstructure of a) high nitrogen specimen (V-N4) and b) high oxygen
specimen (V-O5).
3. Results
The microstructure after annealing revealed a coarse grain
size of about 160 mm (Fig. 1a). The stress-strain curve (Fig. 1b) of
the annealed specimen showed a distinct upper and lower yield
stress and reasonable ductility. Composition analysis showed that
V-O and V-N specimens contain 1538–9092 wppm oxygen and
412–4536 wppm nitrogen, respectively (Table 1 and Fig. 2). While
the V-O specimens were mildly contaminated by nitrogen (1–
6 wppm) the V-N specimens contained 735–1038 wppm of oxy-
gen. Table 1 shows that the annealed specimen contains
463 wppm of oxygen, which was present in the vanadium gran-
ules used in arc melting.
X-ray diffractograms from V-N and V-O specimens showed that
the peaks shift to lower angles with increase nitrogen and oxygen
contents (Fig. 3). This indicates that the nitrogen and oxygen dis-
solve in the vanadium lattice and expand it. No new peaks ap-
peared in the XRD patterns indicating that second phase or pre-
cipitates did not form in the specimens.
The microstructures of V-N and V-O specimens are shown in
Fig. 4. The grain size of the specimens was
400 mm, and was
insensitive to oxygen and nitrogen contents. Comparing the mi-
crostructures in Fig. 4 with Fig. 1b indicates that grain growth
 M.-G. Jo et al. / Materials Science & Engineering A 675 (2016) 92–98
Fig. 6. Fracture surfaces of annealed specimen: a) Overview b) detail of area
marked by square box showing dimples, V-N5 specimen: c) Overview and d) detail
of area marked X, and V-O4 specimen: e) Overview and f) detail of area marked X.
95
occurred during the diffusion heat treatment.
The stress-strain curves (Fig. 5) of V-N and V-O specimen show
distinct upper and lower yield points. Pronounced yield point
elongation did not occur in any specimen. Increasing the oxygen
and nitrogen content increased the flow stress and decreased
ductility. The complete stress strain curves of V-N4 and V-O5 could
not be recorded because the flow stress exceeded the load limit of
the tensile testing machine. The specimen containing the highest
nitrogen content (V-N5) failed before yielding.
Fig. 6 shows the overview and detail of fracture surface of the
annealed and, V-N and V-O specimens with high interstitial con-
centration. The fracture is transgranular in all the specimens. A
few dimples are visible in the annealed specimen (Fig. 6b).
Fig. 7 shows the hardness of V-O and V-N specimens plotted
against the oxygen and nitrogen contents respectively. Evidently,
hardness of vanadium increased with increase in concentration of
oxygen and nitrogen. The hardness increases linearly with solute
content up to 1028 wppm N and 3008 wppm O. At the highest
solute concentration viz., 4536 wppm N and 9092 wppm O, the
hardness appears to be off the linear trend with composition.
Fig. 8 shows the NanoSIMS elemental maps from a region that
included the grain boundary in annealed, V-O and V-N specimens.
The V-O and V-N specimens chosen for NanoSIMS analysis con-
tained the highest oxygen and nitrogen contents. Evidently, sulfur
was enriched at the grain boundary in all three specimens however
oxygen, nitrogen or carbon did not segregate to the grain boundary.
Fig. 9 shows a high annular dark field image in STEM and
element maps using EDS from an area that contains the grain
boundary in the high nitrogen specimen (V-N5). Evidently, the
grain boundary is enriched in sulfur and to a lesser extent by
phosphorus. However nitrogen, oxygen or carbon were not en-
riched at the grain boundary.
4. Discussion
4.1. Hardness coefficient
The hardness of vanadium increased linearly with the con-
centration of nitrogen and oxygen except at the highest level
(Fig. 7). Assuming that nitrogen and oxygen contributed in-
dependently to hardness the effect of nitrogen and oxygen con-
tents on hardness can be determined as determined as
Hv=H0 + K OXO + KNXN (1)
Where Hv stands for Vickers hardness, H0 is a constant, XO and KO
are the oxygen concentration (wppm) and hardness coefficient,
and XN and KN are the nitrogen concentration and hardness
coefficient respectively. H0 represents the base hardness of the
specimens which includes the effect of other contaminants in
vanadium such as carbon. The hardness coefficients calculated
using Eq. (1) are given in Table 2 along with those reported in
published literature [6,7,9]. Kainuma et al. [6] reported the change
in hardness with nitrogen and oxygen content however they did
not determine the hardness coefficient. The hardness coefficients
from their data were estimated using linear fits to oxygen and
nitrogen content respectively. It can be seen from Table 2 that the
nitrogen hardness coefficient determined in the present study is
comparable to those reported in Ref. [9]. The oxygen hardness
coefficient in the present study is lower than that reported in Ref.
[9]. However it is a better estimate of the intrinsic effect of oxygen
because the V-O specimens had relatively low contamination (1–
29 wppm N) when compared with the V-O specimens in Ref. [9]
(
100–150 wppm N). Similar observations could not be made
regarding Refs. [7and 6] because the composition of the specimens
96 M.-G. Jo et al. / Materials Science & Engineering A 675 (2016) 92–98

Fig. 7. Vickers hardness of vanadium as a function of a) nitrogen content in V-N

specimens and b) oxygen content V-O specimens.

used for hardness measurements was not available.

4.2. Segregation and loss in ductility

Using calculations based on electron theory of chemisorption,

Cottrell [18] suggested that embrittlement due to interstitial seg-
regation depends on the character of the bond viz. ionic or cova-
lent, between segregant and body centered cubic (bcc) iron. He
showed that the bonding character of oxygen, unlike nitrogen, at
the interface is ionic and thereby it promotes interface decohesion.
This is assumed to be generally true for light elements (O,N,B,H) in
transition metals and has been confirmed based on first principles
calculations for nitrogen and oxygen segregants in Nb [16] and
oxygen segregant in Tantalum [17]. Since both Nb and Ta belong to
group Va (V, Nb, Ta) in the periodic table, vanadium may also be
expected to show similar response to segregation of nitrogen and
oxygen. Indeed, ductility loss in V-Cr-Ti alloys [13,15] was specu-
lated to be due to oxygen segregation to grain boundaries.
In the present study it is evident from Figs. 8 and 9 that oxygen
and nitrogen did not segregate to grain boundary in specimens
containing high nitrogen and oxygen concentration. However,
sulfur segregated to grain boundaries in annealed, V-O and V-N
specimens. Sulfur segregation to grain boundaries is well known
Fig. 8. NanoSIMS elemental maps for carbon, sulfur, nitrogen and oxygen in a)
to cause embrittlement in bcc-iron [19] and has also been sus- annealed, b) V-N5 and c) V-O5specimens, from a region containing the grain
pected to be the cause of embrittlement in V-Ti [20,21] and V-Cr-Ti boundary.
alloys [22]. However, the loss of ductility in the OT and NT speci-
mens cannot be attributed to the sulfur segregation in the speci- 4.2.1. Friction stress
mens because none of the specimens failed in an intergranular Yield points are typically observed in bcc transition metals
manner (Fig. 6). Another cause for the loss in ductility with in- because of the significant friction stress [23]. Even in well an-
creasing oxygen and nitrogen content could be the increase in nealed bcc transition metals yield points occur when they contain
friction stress. sufficient impurity atoms. This is confirmed in the present study as
 M.-G. Jo et al. / Materials Science & Engineering A 675 (2016) 92–98 97

Fig. 9. a) HAADF image of the area containing the grain boundary in high nitrogen specimen (V-N5). EDS element maps of b) sulfur, c) phosphorus, d) oxygen, e) carbon and
f) nitrogen from the same area as (a).

Table 2 interstitials barely segregated to the grain boundary and had no

Concentration range and Nitrogen and Oxygen hardness coefficients. effect on ky. In nitrogen alloyed austenitic steels [29,30]and Ni-S
[31] alloys it was argued that increase in ky was due to segregation
Reported by Concentration range Hardness coefficient, K (Hv/wppm)
of N, P and S to grain boundaries respectively. In view of the
(wppm)
Nitrogen Oxygen aforementioned reports, ky in the present study is not expected to
be affected with nitrogen and oxygen content. As noted earlier, in
Present study 40–4536 N, 624–9092 0.117 0.01 0.038 7 0.005 the present study the grain size of V-N and V-O specimens was not
O
altered by nitrogen and oxygen content. Since there were no
Kainuma et al. 50–
3500 N, 200–
0.069 0.037
[6] 4800 O
precipitates in the specimens and grain size and ky were not af-
Fromm and 600–2800 N, 600– 0.054 0.054 fected by oxygen and nitrogen content the increase in yield
Horz[7] 3200 O strength was mainly due to an increase in friction stress. Thus, the
Nagasaka et al. 10–500 N, 50–1000 O 0.12 0.057 loss in ductility can be mainly attributed to the increase in friction
[9]
stress with interstitial content, which inhibits dislocation move-
ment and reduces the resistance to crack propagation.
yield points were observed in annealed, V-O and V-N specimens.
In the present study, increasing both nitrogen and oxygen contents
increased the yield strength of vanadium. The yield strength can 5. Conclusions
be increased in one of three ways (see Eq. 2) [24,25]: by decreasing
the grain size, increasing grain boundary hardening or Hall-Petch V-O and V-N specimens with different oxygen and nitrogen
coefficient and increasing the friction stress content were prepared using diffusion heat treatment. Their
composition, microstructure, mechanical behavior and grain
σy = σi + k yd−1/2 (2) boundary segregation was characterized.

Where sy is the yield strength, si is friction stress, ky is the grain
boundary hardening parameter and d is grain size. In ferritic steels
[26,27] it was reported that increasing the carbon interstitial
content increased ky. Based on atom probe tomography analysis
they suggested that segregation of carbon to the grain boundary
lead to the higher strength of the grain boundary [28]. Nitrogen
1. Oxygen and nitrogen in solution increase the hardness, flow
stress and decrease ductility of vanadium. The nitrogen hard-
ness coefficient is nearly 3 times higher than that of oxygen.
2. Sulfur segregates to the grain boundaries in annealed, V-O and
V-N specimens. There is no segregation of oxygen or nitrogen to
the grain boundaries.
98 M.-G. Jo et al. / Materials Science & Engineering A 675 (2016) 92–98
3. The ductility loss is mainly because solute oxygen and nitrogen
cause a large increase in friction stress. At high nitrogen con-
centration the specimen fails before yielding.
Acknowledgment
This work was supported by the National Research Foundation
of Korea (NRF) grant funded by the Ministry of Science, ICT &
Future Planning (MSIP) (NO. NRF-2015R1A5A1037627). The work
at Korea Institute of Science and Technology (KIST) (NO. CAP-11-
05-KRISS) was supported by the Convergence Agenda Program
(CAP) of the Korea Research Council of Fundamental Science and
Technology. The authors thank Korea Basic Science Institute, Bu-
san, for NanoSIMS facility and Dr. T. E. Hong for fruitful discussions.
References
[1] K.H. Kim, H.C. Park, J. Lee, E. Cho, S.M. Lee, Scr. Mater. 68 (2013) 905–908.
[2] Y.-S. Lee, C. Ouyang, J.-Y. Suh, E. Fleury, Y.W. Cho, J.-H. Shim, J. Membr. Sci. 423–
424 (2012) 332–341.
[3] M. Tsukahara, T. Kamiya, K. Takahashi, A. Kawabata, S. Sakurai, J. Shi, H.
T. Takeshita, N. Kuriyama, T. Sakai, J. Electrochem. Soc. 147 (2000) 2941–2944.
[4] S.J. Zinkle, H. Matsui, D.L. Smith, A.F. Rowcliffe, E. van Osch, K. Abe, V.
A. Kazakov, J. Nucl. Mater. 258–263, Part 1 (1998) 205–214.
[5] D.L. Harrod, R.E. Gold, Int. Met. Rev. 4 (1980) 163–221.
[6] T. Kainuma, N. Iwao, T. Suzuki, R. Watanabe, J. Nucl. Mater. 80 (1979) 339–347.
[7] E. Fromm, G. Hörz, Int. Met. Rev. 25 (1980) 269–311.
[8] E.A. Loria, J. Nucl. Mater. 61 (1976) 158–168.
[9] T. Nagasaka, H. Takahashi, T. Muroga, T. Tanabe, H. Matsui, J. Nucl. Mater. 283
(2000) 816–821.
[10] G.T. Hahn, A. Gilbert, R.I. Jaffee, Battelle Memorial Institute, Columbus, Ohio,
USA, 1962.
[11] R.W. Thompson, Iowa State University, 1964.
[12] L.M. Wagner, M.J. Strum, International conference on laser and applications,
San Diego, CA, USA, 1991, pp. 10.
[13] K. Natesan, W.K. Soppet, J. Nucl. Mater. 233–237 (Part 1) (1996) 482–487.
[14] K. Natesan, W.K. Soppet, M. Uz, J. Nucl. Mater. 258–263 (Part 2) (1998)
1476–1481.
[15] J.R. DiStefano, J.H. DeVan, J. Nucl. Mater. 249 (1997) 150–158.
[16] R. Janisch, C. Elsasser, Phys. Rev. B 67 (2003) 11.
[17] G.L. Krasko U.S. Army Research Laboratory, 1996, pp. 20.
[18] A.H. Cottrell, Mater. Sci. Technol. 6 (1990) 807–810.
[19] B. H.K.D.H, H. R.W.K, Steels: Microstructure and Properties, 3rd ed., Butter-
worth Heinemann, Great Britain, 2006.
[20] J. Kameda, T.E. Bloomer, A.H. Swanson, D.Y. Lyu, J. Nucl. Mater. 252 (1998)
1–12.
[21] J. Kameda, T.E. Bloomer, D.Y. Lyu, J. Nucl. Mater. 258 (1998) 1482–1486.
[22] R.J. Kurtz, M.L. Hamilton, H. Li, J. Nucl. Mater. 258 (1998) 1375–1379.
[23] R.E. Smallman, Modern Physical Metallurgy, 4th ed., Butterworth & Co., Lon-
don Great Britain, 1985.
[24] E.O. Hall, Proceedings of the Physical Society of London Section B, 64, 1951, pp.
747–753.
[25] N.J. Petch, J. Iron Steel Inst. 174 (1953) 25–28.
[26] K. Takeda, N. Nakada, T. Tsuchiyama, S. Takaki, ISIJ Int. 48 (2008) 1122–1125.
[27] S. Takaki, D. Akama, N. Nakada, T. Tsuchiyama, Mater. Trans. 55 (2014) 28–34.
[28] J. Takahashi, K. Kawakami, K. Ushioda, S. Takaki, N. Nakata, T. Tsuchiyama, Scr.
Mater. 66 (2012) 207–210.
[29] L.Å. Norström, Met. Sci. 11 (1977) 208–212.
[30] W.A. Spitzig, Metall. Trans. 3 (1972) 1183 (&).
[31] S. Floreen, Westbroo Jh, Acta Metall. 17 (1969) 1175–1181.