21. - 22. 11.

2012, Plze, Czech Republic, EU

INFLUENCE OF TUNGSTEN AND CARBON ADDITION ON ELECTROCHEMICAL BEHAVIOUR

OF 9 % Cr CREEP-RESISTANT STEEL

Ji RAPOUCH, Jaroslav BYSTRIANSK

Department of Metals and Corrosion Engineering, Institute of Chemical Technology, Prague, Technick 5,
166 28 Prague 6, Czech Republic, rapouchj@vscht.cz

Abstract
Martensitic 9 % Cr steel belong to modern creep-resistant materials. Structural instability at higher service
temperatures can lead to coarsening of particles and precipitation of new phases. These changes cause
decrease of mechanical properties at high temperatures and deterioration of resistance against steam
oxidation. Tungsten and carbon substantially contribute to forming of phases in the structure, most
significantly of carbides and Laves phases. In this work the influence of addition of these elements on
electrochemical behaviour of 9 % Cr steel with impact on active-to-passive area was studied. Results of
polarization measurement were compared with hardness measurements and microstructure observation.
Keywords: 9 % Cr martensitic steel, electrochemical behaviour, tungsten, polarization curves

1. INTRODUCTION
Improvement of the thermal efficiency of modern power plants could be achieved by increase of steam
parameters. The increase of steam temperature above 600 C also raises the requirements on creep
strength of structural material. Group of materials that fulfil these demands are heat-resistant martensitic
9-12% Cr steels. The sufficient creep resistance of these steels is achieved by two steps [1, 2]:
1. Molybdenum and tungsten alloying which strengthen the solid solution,
2. alloying with vanadium, niobium, carbon and nitrogen, which contribute to precipitation hardening.
The structure of 9-12% Cr steels is composed of tempered martensite, carbides M23C6 and MC (M = Fe, Cr,
Mo, W), nitrides and carbonitrides V(C,N) and Nb(C,N). The dislocation density in tempered state
14 -2
is > 10 m [2]. During service heating some structural changes take place: coarsening of carbides,
precipitation of carbonitride phases enriched with alloying elements, formation of Laves phase
(Fe,Cr)2(W,Mo) and possible precipitation of nitride Z phase with general formula Cr(V,Nb)N [1-3]. These
changes lead to general decrease of creep strength, ductility and also the corrosion resistance of the
material.
This paper is focused on the influence of tungsten and carbon on the structure and electrochemical
behaviour of 9-12% Cr steels. Carbon in these steels improves their heat-resistance by formation of carbides
and carbonitrides of alloying elements. On the other hand high carbon content causes deterioration of
technological properties (formability, weldability). The amount of carbon in modern steels ranged between
0.08 0.20 wt. % [2].
Tungsten is ferrite- and carbide-forming element with higher atomic radius than iron, therefore its addition
results in solid solution strengthening. In amount of 1.6 % W significantly slows down the rate of climb of
dislocations and practically suppresses the precipitation of Cr 23C6 at grain boundaries [1]. Moreover,
tungsten lowers Ms temperature and makes a fine subgrain structure [4]. If its maximal solubility in steel is
exceeded, tungsten precipitates forming Laves phase (Fe,Cr) 2W and the coarsening rate of this phase is
much higher than that of Cr23C6 [5]. In addition, precipitation of Laves phase can be connected with formation
of creep cavities and lead to intergranular fracture [6]. Tungsten together with molybdenum dissolves into
M23C6 carbides and thus increases their thermal stability. However, these elements reduce roughness by
 21. - 22. 11. 2012, Plze, Czech Republic, EU

forming -ferrite and intermetallic compounds [7]. The combined effect of W and Mo can be described by an
equivalent content of molybdenum Mo(eq.) = wt. % Mo + 0.5 wt. % W. The optimal value of Mo(eq.) should
range between 1.2 and 1.5 [2].
From the point of electrochemical behaviour, tungsten significantly influences the active-to-passive transition
of stainless steels. The addition of W up to 8 wt. % to ferritic FeCr29 steel leads to decrease of critical anodic
current density and passivation potential [8]. Moreover, it lowers the current density (and the corrosion rate)
in passive state [9]. Similarly to molybdenum, tungsten enhances the resistance to pitting corrosion, its effect
is approximately half compared to Mo [10, 11]. On the other hand, negative influence of W is observed in
transpassive area and its presence in steel lowers the transpassivation potential of the steel. Tungsten is
2-
expected to dissolve in oxidizing environments (for example in supercritical steam) as WO4 in strong
2+
alkalies and as WO2 in strong acids [12].

2. EXPERIMENTAL
Four different laboratory melts (2 kg) of Fe-9Cr steels with various carbon and tungsten content were used
for experiments. The chemical compositions of the steels are given in Tab. 1. One set of specimens was
quenched (1050 C, 20 min, air), the second was afterwards tempered at 780 C for 10 min.
Tab. 1 Chemical composition samples
chemical composition [wt. %]
steel
C Mn Si S Cr Ni Mo W

C1W1.5 0.01 0.47 0.53 0.02 9.3 0.13 0.04 1.5

C15W2.3 0.15 0.54 0.45 0.15 9.3 0.13 0.03 2.3

C53W1.9 0.53 0.52 0.43 0.02 9.9 0.14 0.04 1.9

C63W4.0 0.63 0.56 0.44 0.02 9.7 0.15 0.03 4.0

All samples were grinded, polished and etched in Nital and Villela-Bain solution mixture before study of
structure. Scanning electron microscope Tescan Vega 3 equipped with EDS analyser was used for
microstructure observation. Vickers hardness was measured at room temperature under the load of 294 N.
A common three-electrode cell with a platinum wire as counter-electrode and a silver/silver chloride electrode
(ACLE) as the referent electrode (197 mV compared to the standard hydrogen electrode) was used for
electrochemical measurements. All potentials reported in this paper are related to the ACLE. Aerated H2SO4
solution (0.5 mol/l) at ambient temperature was used for all measurements. Potentiodynamic measurements
were carried out by the PC3 system and Gamry electrochemical software. The corrosion potential was
stabilized before the polarization measurement. The scan rate was 3 mV/s.

3. RESULTS AND DISCUSSION

The structure of alloys in as quenched state was composed of martensitic plates, moreover, XRD analyses
of these specimens showed small amounts of retained austenite (up to 6 %). Precipitation of phases
preferentially on martensitic plates boundaries occurred during tempering. Because laboratory alloys were
used, their structures differ from those of standard martensitic Fe-9Cr steels. Structures of specimens
C53W1.9 and C63W4.0 in as quenched and as tempered state are shown on Fig. 1.
 21. - 22. 11. 2012, Plze, Czech Republic, EU

Fig. 1 Structures of steels a) C53W1.9 as quenched, b) C63W4.0 as quenched, c) C63W4.0 as tempered

The influence of steel composition and structure on hardness is plotted on Fig. 2. Quenched and tempered
steels alike show increasing tendency of hardness with carbon content. In as quenched state C increases
hardness of martensitic plates. In addition, W contributes to formation of fine grain structure. Carbides M 23C6
enriched with tungsten are formed in the structure during tempering. Moreover, increasing carbon content in
the matrix leads to formation of higher amount of carbides during tempering which contributes to increase of
hardness.

.
Fig. 2 Influence of hardness on carbon content in Fe-9Cr alloy
Different structural states of alloys were evaluated using electrochemical methods. Potentiodynamic curves
of steels C1W1.5 and C63W4.0 are shown on Fig. 3. All measured samples in all metallurgical states hold
corrosion potential around -450 mV. The active-to-passive transition is connected with development of two
anodic peaks on polarization curve at potentials -250 and 0 mV. At potential 600 mV transpassive dissolution
of material caused by transition of alloying elements into higher oxidation states takes place. Local peak at
potential 1200 mV is expected to be connected with secondary passivation of surface. This phenomenon is
induced by relative enrichment of surface with iron, which is passive in this area and causes decrease of
current density [13]. The potential position of this peak is similar for samples with different content of carbon
and tungsten, therefore, its occurrence is probably connected with transition of chromium from 3+ to 6+
state. The shape of this peak for as quenched and as tempered steel shows an interesting dissimilarity. This
local maximum was separated into 2-3 smaller peaks in the case of tempered samples (Fig. 3, red lines).
This separation could be caused by preferential dissolution of phases enriched with Cr (and other alloying
elements) formed during tempering. The area of secondary passivity is not studied closer in this paper.
 21. - 22. 11. 2012, Plze, Czech Republic, EU

Fig. 3 Potentiodynamic curves of steels a) C1W1.5, b) C63W4.0

The active-to-passive area with double-peak was the main object of this work. Tempering is obvious to
slightly increase current density in whole measured area, the most significant increase was observed in
areas of active dissolution and secondary passivation (Fig. 3). While the first anodic peak (-250 mV) is
increased only slightly, the second anodic peak (0 mV) is shifted very significantly into higher current
densities. Fig. 4 compares this area for samples in quenched as well as in tempered state. The behaviour of
samples in as quenched state is very similar and does not differ from that of P92 [14]. On the contrary, the
current density of the first and even more significantly of the second anodic peak tend to increase with rising
carbon content in the alloy, see Fig. 5.

Fig. 4 Potentiodynamic curves of samples a) as quenched, b) as tempered,

The occurrence of second anodic peak is assumed to be caused by several phenomena surface nickel
enrichment, oxidation of absorbed hydrogen, preferential attack along phase and ferrite/martensite grains
 21. - 22. 11. 2012, Plze, Czech Republic, EU

boundaries, presence of the precipitates of Cu or Mo or dissolution of Cr-depleted areas [15, 16]. Because of
high amount of carbon and chromium in used alloys Cr-depletion theory is preferred. Therefore, formation
and following preferential dissolution of areas depleted with Cr surrounding carbides is considered to be the
main cause of the second anodic peak on polarization curve. Some carbide particles seem to be present in
the quenched matrix and their amount and size rise during tempering. This assumption is in accordance with
the tendency plotted on Fig. 5. The structure was potentiostatically etched at anodic peaks potentials (-250
and 0 mV) and observed on scanning electron microscope.

Fig. 5 Influence of anodic current densities at potentials -250 and 0 mV of tempered samples on C content
Structure of sample C63W4.0 after potentiostatic etching is shown on Fig. 6. In as quenched state, at
potential -250 mV dissolution of the matrix takes place, at 0 mV (potential of second anodic peak) significant
etching of grain boundaries was observed. Carbide particles precipitate predominantly on grain boundaries
and surrounded depleted zones are probably preferentially dissolved/etched at potential 0 mV. Etching of
carbides at potential -250 mV in the case of tempered sample was observed. Slight enlargement of etched
areas in the vicinity of these carbides was observed after etching at 0 mV. Approximate chemical
composition of etched phases was detected using EDS analyzer. These phases composed of 20 24 % W,
31 37 % Cr and 40 48 % Fe (all in wt. %). Determination of accurate content of carbon is not possible
using this method, but higher amount of C was detected compared to matrix. Therefore, phases M 23C6
(M = Cr, W, Fe) are supposed to be present after tempering and zones depleted with chromium are expected
to occurs in their vicinity. Increasing second anodic current density caused by higher amount of precipitates
in the structure is in accordance with hardness tendency on Fig. 2.

Fig. 6 Structure of sample C63W4.0 a) quenched, etched at -250 mV, b) quenched, etched at 0 mV, c)
tempered, etched at -250 mV, d) tempered, etched at 0 mV
 21. - 22. 11. 2012, Plze, Czech Republic, EU

4. CONCLUSIONS

Electrochemical behaviour of four Fe-9Cr alloys with various amounts of carbon and tungsten is studied in
this paper. Polarization curves of these steels in 0.5 mol/l H2SO4 solution in active-to-passive area are
characterized by two anodic peaks observed also in the case of steel P92. The second anodic peak is
assumed to be caused by dissolution of Cr-depleted zones in the vicinity of precipitates. Formed precipitates
were determined as carbides M23C6 with higher content of chromium and tungsten. Second anodic current
density shows increasing tendency with carbon content, which could be attributed to increasing amount and
size of carbides. Moreover, the increasing carbides precipitation could explain rising of hardness with content
of C.

ACKNOWLEDGEMENTS
Financial support from specific university research (MSMT No 21/2012) and MPO FR-TI1/086.

REFERENCES
[1] KAYBYSHEV, R., SKOROBOGATYKH, V., SHCHENKOVA, I. New martensitic steels for fossil power plant:
Creep resistance. The Physics of Metals and Metallography, 2010, 109 (2), 186-200.
[2] VODREK, V. Fyzikln metalurgie modifikovanch ocel (9-12)% Cr ocel. Vysok kola bsk - technick
univerzita Ostrava, 2003.
[3] KOMAZAKI, S.-I., KISHI, S., SHOJI, T., et al. Influence of pre-aging on creep rupture strength of tungsten alloyed
9%Cr ferritic steel and creep damage evaluation by electrochemical method. JSME International Journal, 2002,
45, 30-38.
[4] MARUYAMA, K., SAWADA, K., KOIKE, J. -I. Strengthening mechanisms of creep resistant tempered martensitic
steel. ISIJ International, 2001, 41 (6), 641-653.
[5] ABE, F. Creep rates and strengthening mechanisms in tungsten-strengthened 9Cr steels. Materials Science &
Engineering, A: Structural Materials: Properties, Microstructure and Processing, 2001, A319321, 770-773.
[6] LEE, J. S., ARMAKI, H. G., MARUYAMA, K., et al. Causes of breakdown of creep strength in 9Cr1.8W0.5Mo
VNb steel. Materials Science and Engineering A, 2006, 428 (12), 270-275.
[7] SHEN, Y. Z., KIM, S. H., CHO, H., et al. Precipitate phases of a ferritic/martensitic 9% Cr steel for nuclear power
reactors. Nuclear Engineering and Design, 2009, 239 (4), 648-654.
[8] AHN, M. K., KWON, H. S., LEE, H. M. Quantitative comparison of the influences of tungsten and molybdenum on
the passivity of Fe-29Cr ferritic stainless steels. Corrosion Science, 1998, 40 (2/3), 307-322.
[9] EDSTRM, J. O., CARLN, J. C., KMPINGE, S. Anforderungen an Sthle fr die chemische Industrie.
Werkstoffe und Korrosion, 1970, 21 (10), 812-821.
[10] KIM, J. -S., KWON, H. -S. Effects of tungsten on corrosion and kinetics of sigma phase formation of 25%
chromium duplex stainless steels. Corrosion, 1999, 55 (5), 512-521.
[11] Park, C. -J., KWON, H. -S. Effects of aging at 475 C on corrosion properties of tungsten-containing duplex
stainless steels. Corrosion Science, 2002, 44 (12), 2817-2830.
[12] JOHNSON, J. W., WU, C. L. The Anodic Dissolution of Tungsten. Journal of The Electrochemical Society 1971,
118 (12), 1909-1912.
[13] BOJINOV, M., BETOVA, I., FABRICIUS, G., et al. The stability of the passive state of iron-chromium alloys in
sulphuric acid solution. Corrosion Science, 1999, 41 (8), 1557-1584.
[14] RAPOUCH, J., BYSTRIANSK, J. Chovn ropevnch ocel v transpasivn oblasti. Koroze a ochrana materilu
2012, 56 (2), 31-37.
[15] WALZAK, T. L., SHEASBY, J. S. The Effect of Heat Treatment Parameters on the Anodic Polarization Behavior of
440C Stainless Steel, Corrosion 1983, 39 (12), 502-507.
[16] FELLONI, L., et al. Investigation on the second anodic current maximum on the polarization curves of commercial
stainless steels in sulphuric acid. Corrosion Science 1973, 13 (10), 773-789.