Journal of Minerals & Materials Characterization & Engineering, Vol. 9, No.7, pp.593-606, 2010
jmmce.org Printed in the USA. All rights reserved

593
High Temperature Corrosion Behaviour of T-91 and T-22 Bare Steel in
75wt.%Na2SO4+25wt.%NaCl Molten Salt Environment at 900°C
Dinesh Gonda*, Vikas Chawlab, D. Puria, S. Prakasha
aMetallurgical & Materials Engineering Department, Indian Institute of Technology Roorkee,
Roorkee 247667, India
bMechanical Engineering Department, L.L.R.I.E.T., Moga, Punjab-152002, India
* Corresponding author:
ABSTRACT
The oxidation behaviour of T-91 steel and T-22 steel in salt of 75wt.% Na2SO4 + 25wt.%
NaCl has been studied under isothermal conditions at a temperature of 900°C in a cyclic
manner. Oxidation kinetics was established for the T-91 steel and T-22 steel in salt at 900°C
under cyclic conditions for 50 cycles by thermogravimetric technique. Each cycle consisted
of 1 hour heating at 900°C followed by 20 min of cooling in air. Both the samples nearly
followed the parabolic rate law of oxidation. X-ray diffraction (XRD) and scanning electron
microscopy/energy dispersive X-ray (SEM/EDAX) techniques were used to characterise the
oxide scales. T-91 steel was found to be more corrosion resistant than T-22 steel under salt
oxidation.
Keywords: Hot corrosion; T-91; T-22; 75wt.%Na2SO4+25wt.%NaCl at 900°C
1. INTRODUCTION
Metals and alloys sometimes experience accelerated oxidation when their surfaces are
covered with a thin film of fused salt in an oxidising gas atmosphere at elevated temperatures.
This is known as hot corrosion where a porous non-protective oxide scale is formed at the
surfaces and sulphides in the substrate [1]. Alloys that are developed for heat and oxidation
resistance typically form a protective layer of chromia or alumina. The more rapidly this
layer is established, the better protection is offered. As this layer grows or as it reforms over
areas from which the original layer was removed, it must withdraw chromium or aluminium
from the metal in order to provide for further scale growth [2]. Oxide scale is constituted by a
layered structure with compositional and microstructural variations from the substrate to the
outer interface [3–8].

594 Dinesh Gond, Vikas Chawla, D. Puri, S. Prakash Vol.9, No.7
The selection of modified 9Cr–1Mo steel for PFBR (pool type fast breeder reactor) SG
applications is based on important considerations such as high temperature mechanical
properties including creep, low cycle fatigue (LCF) and creep fatigue interaction, resistance
to loss of carbon to liquid sodium and consequent reduction in strength, resistance to wastage
in case of small leaks leading to sodium–water reaction and resistance to stress corrosion
cracking in sodium and water media [9]. Among the ferritic steels, modified 9Cr–1Mo
exhibited the highest 105h rupture strength at all temperatures, while 2.25Cr–1Mo steel
exhibited the lowest rupture strength. The strength values of plain 9Cr–1Mo steel lie in
between of those exhibited by 2.25Cr–1Mo and modified 9Cr–1Mo steels. The creep strength
of modified 9Cr–Mo steel remains higher or equal to that of AISI type 304 austenitic stainless
steel up to 898 K [10].The high creep strength of modified 9Cr–1Mo steel is attributed to
microstructural stability at high temperatures derived from the presence of fine V(C, N) and
Nb(C, N) precipitates, which prevents the strength from dropping to inherent creep strength
and to maintain excellent creep strength at long durations [11].
This paper is intended as a contribution to the knowledge of the oxidation behaviour of the T-
91 and T-22 ferritic steel in salt (75%wt Na2SO 4 + 25%wt NaCl) atmosphere under
isothermal conditions in cyclic manner. In this experimental study emphasis is also given to
oxide scales which were separated and fell down in boat while oxidation process was going
on.
2. EXPERIMENTAL MATERIAL AND PROCEDURE
2.1 Substrate Steels
The experimental work was performed by using samples of T-91 & T-22 steel. The T-91 steel
samples were obtained from Gurunank Dev Power plant, Bhatinda, Punjab, India and T-22
steel was received from Prabhakar Engineering Pvt Ltd, Pune, India.
The spectroscopy was done on samples which were taken for experiment, this showed
chemical composition in wt. % which is given below:
Type
of steel C Mn Si S P Cr Mo Cu Ni V Nb Al Fe
T-22 0.097 0.43 0.35 0.014 0.017 2.25 0.93 0.007 0.093 0.021 0.004 0.01 Balance
T-91 0.0607 0.3874 0.2297 - - 8.078 0.8029 0.1168 - - - - Balance

Vol.9, No.7 High Temperature Corrosion Behaviour 595
595

2.2. Optical Microscopy for Surface Microstructure
The microstructure of the T-91 and T-22 steel samples, after polishing and etching with
marble’s reagent (Marble's Reagent = Distilled Water 50 ml, HCl 50 ml & Copper sulphate
(CuSO4) 10 grams immersion or swab, etch for a few seconds) is shown in Fig. 1.
The microstructure of T-91 steel revealed ferritic structure i.e. the white spot which is seen in
microstructure is ferrite and the rest is other phase. After etching the T-22 sample revealed
the microstructure which was found to consist of white ferrite and the rest is other phase.
2.3. Sample Preparation
The experiment was performed on samples which were made to specified dimensions of
approximately 20 x 15 x 3.5 mm from tubular sections. The specimens were polished on SiC
emery paper down to the 1200 from 120 grades. Polishing was carried out on all six faces.
The specimens were degreased (by ultrasonic cleaning in ethanol) and dried, then they were
accurately weighed and measured to determine the total surface area exposed to the oxidative
environment.
(a) Microstructure of T-91 steel (b) Microstructure of T-22 steel
Fig. 1. Microstructure of T-91 & T-22 steel at 20x magnification.
2.4. High Temperature Oxidation Study in Air
Hot corrosion studies were conducted at 900°C in laboratory using silicon carbide tube
furnace having PID temperature controller (make Digitech, India). The samples were
subjected to mirror polishing which include cloth polishing which will provide uniformity of
reaction while oxidation process. Then dimensions were accurately measured by digital
vernier (make Mototoyo, Japan) so as to calculate area which will be required for plotting of
graph of weight gain per unit area verses number of cycle. Finally specimens were cleaned
i.e. degreased by ethanol and kept in alumina boat. This alumina boat prior to performing of
experiment was kept in oven for 5hr at 250°C in oven and then kept in furnace at 900°C for

596 Dinesh Gond, Vikas Chawla, D. Puri, S. Prakash Vol.9, No.7
2hr so that moisture is totally removed from boat. After this the sample was kept in boat and
weight was taken initially and then slowly inserted in tubular furnace.
Samples of T-91 and T-22 steel were kept in alumina boat and heated in an oven along with
alumina boat up to 150°C and the salt mixture of 75wt.% Na2SO4 + 25wt.% NaCl dissolved
in distilled water was coated on the warm polished samples with the help of a camel hair
brush. The amount of the salt coating varies in the range 3.0–5.0mg/ cm2. The salt coated
samples were then dried at 250°C for 2 and ½ hrs in an oven to remove the moisture and then
weighed, after this sample of T-91 and T-22 were inserted in different tubular furnace .These
samples were kept in furnace for 1 hr at a temperature of 900°C and then they were removed
and cooled further for 20 minutes at room temperature and their weight were taken by
electronic balance (make Contech, India) having sensitivity of 0.001 gms. This cycle was
repeated for 50 times i.e. 50 cycles were made for each sample. The weight of samples was
measured at the end of each cycle and spalled scale was also taken into consideration which
used to fall into the boat i.e. the weight was taken along with the boat.
Corroded samples from salt oxidation were analysed by XRD (BRUKER-binary V3) and
SEM/EDAX and the oxide scale which fell into the boat were also analysed by XRD. Cu
radiation was used in XRD at a step of 2°/min and the range of angle was 5-100°.
3. EXPERIMENTAL RESULTS
3.1. Behaviour in Salt at Elevated Temperature
The oxidation of sample which occurred in salt at a temperature of 900°C is shown by
plotting a graph Fig 2. On x-axis “number of cycles” and on y-axis “weight gain/area
(mg/cm2)” was taken.
Fig.2. Weight gain plot for T-91bare steel exposed to cyclic study in salt at 900°C for 50
cycles.
0
50
100
150
200
250
300
350
400
450
159 13172125293337414549 We ight gain/area (mg/cm
2
)
Number of cycle
T91Steel
T22Steel

Vol.9, No.7 High Temperature Corrosion Behaviour 597
597
The hot corrosion behaviour of T-91 and T-22 steel in salt was parabolic as shown above, but
in T-22 steel after 35th cycle there was more weight gain. The above two graph reveals that T-
91 steel is better than T-22 steel in an environment of salt oxidation (for 50 cycles). Every
line or curve in graph is having its approximate equation which is given below.
For T-91 salt oxidation the approximated curve is
Y = -13.78649 + 52.32262*X - 6.51182*X^2 + 0.40256*X^3 - 0.01266*X^4 + 1.94808E-
4*X^5 -1.16293E-6*X^6
For T-22 salt oxidation the approximated curve is
Y = -13.78649 + 52.32262*X - 6.51182*X^2 + 0.40256*X^3 -0.01266*X^4 + 1.94808E-
4*X^5 -1.16293E-6*X^6
(Where X is number of cycle and Y is weight gain/area & this equation is calculated by using
analysis mode of Origin software)
As shown in macrograph T-91 salt oxidised sample did not showed much more crack or
extrusion of material from beneath but T-22 steel had more corrosion effect as shown in
macrograph, its layer got separated from its substrate as is evident from Fig 3 and the oxide
layer formed on T-22 sample is also thick as shown in Fig 8.
(a) (b)
Fig. 3. Macrograph of (a) T-91 salt oxidised (b) T-22 salt oxidised samples.
3.2. X-Ray Diffraction Analysis
The samples after oxidation were removed from boat and their oxide scales which were
separated from surface were also removed. Then they were analysed separately by XRD and
after that only oxidised sample were analysed by SEM / EDAX. The re sults of XRD an alysis
contained graph indicating peak values (i.e. d values) which were used to identify various
phases with the help of inorganic X-ray Diffraction data card from Po wder diffracti on file of
JCPDS. Help of software named Philips X’pert High score and Eva was also taken for
finding out compounds at respective peaks.

598 Dinesh Gond, Vikas Chawla, D. Puri, S. Prakash Vol.9, No.7
3.2.1. XRD result for T-91 and T22 sample
From the X-Ray Diffraction analysis it is found that ferrous oxide (Fe2O3), chromium ferrous
oxide (Cr, Fe)2O3 are mainly formed along with Cr2O3, in T-91 salt oxidised sample but in
T-22 sample NiO is also formed in addition to Fe2O3 , (Cr, Fe)2O3, Cr2O3.
Fe2O3, Cr2O3 form a protective oxide layer at surface due to which further oxidation is
prevented as it acts as barrier for further corroding media to interact with substrate but as in
case of T-22 steel, Nickel oxide (NiO) is formed which are not protective, rather than that
NiO is said to be loose structured and thus lead to more pore size and thus results in more
corrosion [12-14].
Fig 4. XRD graph for T-91, T-22 sample and its oxide scales in salt oxidised environment at
900°C.
3.3. Energy Dispersive X-ray (EDAX) Studies
3.3.1. Surface scale
The SEM/EDAX analysis for T-91 & T-22 sample after oxidation in salt for 50 cycle at
900°C is shown in Fig 5 & 6. Surface morphology of T-91 salt oxidised sample Fig 5(a & b)
reveals that the oxide formed is layer wise but at higher magnification it revealed that the
layer contain granules. Analysis of these granules revealed that they were formed when the
amount of chromium oxide was less in it i.e. at about 1-2%.and it showed some cracking in
20 30 40 50 60 70
T-22 Sputtered Ox ide Scale
T-22 Salt Oxidised
T-91 Sputtered Oxide Scale
T-91 Salt Oxidised
a
a a
a,c
a,b
a,c
da
a
a
a
c
a,b,c
c
c
c c
c
c
c
c
a,b
a,b,c a,b, c
a= Fe
2
O
3
b= Cr
2
O
3
c= (Cr,Fe)
2
O
3
d= NiO
In te n s ity (Ar b itr a r y Con sta nt)
Diffraction Angle (2 Theta)

Vol.9, No.7 High Temperature Corrosion Behaviour 599
599

oxide layer. In this there is more amount of Mo2O3 and this is also one reason that the oxide
layer formed in this case are strong as when they react with salt.
Surface morphology of T-22 salt oxidised sample Fig 6 (a & b) revealed some fibrous
structure at higher magnification which mainly comprised of Fe2O3 and the rest other
compounds were Mo2O3, Cr2O3 and NiO and at very middle nodule of aluminium oxide was
also found. As compared to T-91 steel T-22 steel formed very less amount chromium oxide
which is mainly responsible for corrosion protection and in this some NiO oxide is also
formed and as sa i d before in description of XRD analysis NiO is loose structured, hence there
are two main reason for less corrosion protection for T-22 steel is that less amount of Cr2O3
formed and presence of NiO.
(a) T-91 salt oxidised sample at scale of 500 µm
(b) T-91 salt oxidised sample at scale of 100µm
Fig 5. Surface scale morphology and EDAX analysis (wt.%) for T-91 steel sample subjected
to the cyclic oxidation at 900°C for 50 cycles in salt environment.

600 Dinesh Gond, Vikas Chawla, D. Puri, S. Prakash Vol.9, No.7
(a) T-22 salt oxidised sample at scale of 500 µm
(b) T-22 salt oxidised sample at scale of 100µm
Fig 6 (a-b). Surface scale morphology and EDAX analysis (wt.%) for T-22 steel sample
subjected to the cyclic oxidation at 900°C for 50 cycles in salt environment
3.3.2. Cross-sectional scale
The BSEI micrograph shown in Fig. 8 reveals the condition of scale of T-91 and T-22 steel
sample exposed to the cyclic oxidation for 50 cycles at 900°C.Elemental variation for
corroded cross-section of T-91 and T-22 is also shown along with in form of point wise
analysis. Fig.7 shows the macrograph of cross section which were cut from the samples. In T-
91 salt oxidised sample Fig 8(a) at point 9 i.e. at outer edge, ferrous oxide is formed along
with molybdenum oxide but at point 5 where there is void ferrous oxide seems to be
decreasing as compare to other points and rate of molybdenum oxide is nearly constant and
its direct effect is seen at point 6 where ferrous oxide is very less. In this two distinct layer
can be seen in which layer near the substrate is rich in ferrous and the second layer is rich in
chromium as compared to first layer.

Vol.9, No.7 High Temperature Corrosion Behaviour 601
601
(a) (b)
Fig.7. Cross-sectional macrograph of (a) T-91 salt oxidised (b) T-22 salt oxidised samples.
In T-22 salt oxidised sample Fig.8(b) after point 3 where the oxide layer has got separated
from substrate shows decrease in elemental composition but at outer side i.e. at point 9 there
is more ferrous oxide and at point 1 there is no oxygen but high amount of ferrous as it is
substrate material and at point 4 where there is void less amount of ferrous oxide has formed
as compared to other points. In this outer oxide layer seems to be thick and in T-91 salt
oxidised sample middle layer seems to be more protective than outer one.
(a) T-91 salt oxidised sample (b) T-22 salt oxidised sample
Fig.8. Oxide scale morphology and elemental composition variation across the cross section of T-91 steel exposed
to salt environment at 900°C for 50 cycles.
Substrate Oxide Substrate Oxide

602 Dinesh Gond, Vikas Chawla, D. Puri, S. Prakash Vol.9, No.7
4. DISCUSSION
The results which were seen till now resemble that corrosion resistance property of T-91 steel
is better than T-22 steel as weight gain of T-22 steel is more. Internal oxidation further led to
the cracking of the scale due to the different thermal expansion coefficients of oxides in the
scale from that of coating as suggested by P. Niranatlumpong [13]. As there are various
elements and each have different thermal coefficient of expansion hence there will be more
stress generated which will lead to more cracking. Through these cracks, corrosive gases can
penetrate to the base material and will thus allow significant grain boundary corrosion attack
[15–17]. Comparatively higher weight gain values in case of T-22 steels might be attributed
to the presence of molybdenum in this substrate steel.
The better corrosion resistance of T-91 steels may also be attributed to the absence of a NiO
layer in the scale. During investigation, the NiO layer was observed in the oxide scale of T-22
steel. This layer has been suggested to be loose- structured by X. Wu [12], which may not be
able to provide effective protection. P. Niranatlumpong [13] have also suggested increase in
the pore size of Ni and Cr scale with increase in exposure time, which allows the degrading
species to penetrate through the coating thereby resulting in the oxidation of substrate steels.
It could be seen from Fig. 2 that the over-all weight gains was conceived only during the
initial cycles of the oxidation study in T-91 salt oxidation case. This may be due to the rapid
formation of the oxides at the grain boundaries and within open pores due to the penetration
of the oxidizing species along the grain boundaries/open pores during the early cycles of the
study, as has been proposed by Choi [18] and Niranatlumpong [13].
However, once the oxides are formed at the places of the porosity and the grain boundaries,
the oxide layer becomes dense and the diffusion of the oxidizing species to the internal
portions of the substrate gets slowed down and the growth of the oxides becomes limited
mainly to the surface of the specimens.
When the oxide layer was removed by using orbital shaking machine from the substrate in
case of salt applied sample it revealed that pitting action occurred due salt on the surface Fig
11(a & b). In case of T-91 steel Fig 11(a) the pits formed were very small in size and were
uniform all over the surface whereas in case of T-22 steel the pits formed were not uniformly
distributed all over the surface but they occurred in specific pattern as shown in figure Fig
11(b) and were very large and deep as compared to T-91 steel pits.
SubstrateOxide

Vol.9, No.7 High Temperature Corrosion Behaviour 603
603
Fig. 9 BSEI and elemental X-ray mapping of the cross-section of T-91 sample exposed to
cyclic hot corrosion at 900°C for 50 cycles in salt of 75%wt Na2SO4 + 25%wt NaCl
Substrate Oxide
Fe
O
Cr Mn
Mo
BSEI

604 Dinesh Gond, Vikas Chawla, D. Puri, S. Prakash Vol.9, No.7
Fig. 10. BSEI and elemental X-ray mapping of the cross-section of T-22 sample exposed to
cyclic hot corrosion in at 900°C for 50 cycles in salt of 75%wt Na2SO4 + 25%wt NaCl.
Substrate Oxide
BSEI
Cr Ni
Fe
O
Mn Mo

Vol.9, No.7 High Temperature Corrosion Behaviour 605
605

(a) Micrograph of pitting on T-91 salt oxidised sample at scale of 500µm and 200µm.
(b) Micrograph of pitting on T-22 salt oxidised sample at scale of 500µm and 200µm.
Fig.11. Micrograph of pitting occurred on salt oxidised sample (a) T-91 and (b)T-22 steel
5. CONCLUSION
The cyclic oxidation of this T-91 steel in salt follows parabolic law but in T-22 steel after 35th
cycle it showed steep increase in weight gain. In case of T-91 the weight gain (0.617 gms)
was much less as compared to T-22 steel (1.292 gms) because T-91 steel has formed
chromium iron oxide (Cr, Fe)2O3 , chromium oxide Cr2O3 and hematite (Fe2O3) at top
surface while T-22 formed nickel oxide (NiO) which is said to be loose structured and not
able to provide effective protection. It also increases the pore size of Ni and Cr scale, which
allows the degrading species to penetrate through the coating thereby resulting in the
oxidation of substrate [12-14]. Hence T-91 is found to be superior to T-22 steel and as per the
cross section morphology the scale thickness of T-91 is comparatively smaller than T-22
steel.
REFERENCES
[1] Harpreet Singh, D. Puri, S. Prakash, Some studies on hot corrosion performance of
plasma sprayed coatings on a Fe-based superalloy, Surface & Coatings Technology
192, 25 May 2004,pp.27– 38.
[2] R.J. Link, N. Birks, F.S. Pettit, F. Dethorey, The response of alloys to erosion–
corrosion at high temperatures, Oxid. Met. 49 (3–4) (1998) pp.213–236.

606 Dinesh Gond, Vikas Chawla, D. Puri, S. Prakash Vol.9, No.7
[3] I. Saeki, T. Saito, R. Furuichi, M. Itoh, Growth process of protective oxides formed on
type 304 and 430 stainless steels at 1273°K, Corros. Sci. 40 (8) (1998) p.1295.
[4] S. Jianian, Z. Longjiang, L. Tiefan, High temperature oxidation of Fe–Cr alloys in wet
oxygen, Oxid. Met. 48 (3, 4) (1997) p.347.
[5] Z. Tokei, H. Viefhaus, H.J. Grabke, Initial stages of oxidation of a 9CrMoV steel: role
of segregation and martensite laths, Appl. Surf. Sci. 165 (1) (2000) p.23.
[6] A. P. Greeff, C.W. Louw, H.C. Swart, The oxidation of industrial FeCrMo steel,
Corros. Sci. 42 (10) (2000) p.1725.
[7] A. Arztegui, T. Gomez-Acebo, F. Castro, Steam oxidation of ferritic steels: kinetics and
microstructure, Bol. Soc. Esp. Ceram. Vidr. 39 (3) (2000). p.305.
[8] A.S. Khanna, P. Rodriguez, J.B. Gananamoorthy, Oxidation kinetics, breakaway
oxidation, and inversion phenomenon in 9Cr–1Mo steels, Oxid. Met. 26 (3, 4) (1986)
p.171.
[9] Mathew MD, Choudhary BK, Albert SK. Selection of materials for PFBR nuclear
steam supply system components. Report-PFBR/MDG/2002/001, IGCAR Kalpakkam.
[10] Baldev Raj, B.K. Choudhary, R.K. Singh Raman, Pressure Vessels and Piping 81
(2004) 521– 534.
[11] Kimura K, Kushima H, Abe F, Yagi K, Irie H. In: Strang A, Banks WM, Conroy RD,
Goulette MJ, editors. Advances in turbine materials, design and manufacturing.
London: The Institute of Materials; 1997. pp. 257–69.
[12] X. Wu, D. Weng, Z. Chen, L. Xu, Surface Coating Technology, 140 (2001), p.231.
[13] P. Niranatlumpong, C.B. Ponton, H.E. Evans, Oxid. Met, 53 (3–4) (2000), pp. 241-
258.
[14] Buta Singh Sidhu, S. Prakash , Performance of NiCrAlY, Ni–Cr, Stellite-6 and Ni3Al
coatings in Na2SO4–60% V2O5 environment at 900°C under cyclic conditions,
Surface & Coatings Technology 201,17 April 2006, pp.1643–1654.
[15] S. Danyluk, J.Y. Park, Corrosion 35 (12) (1979) p.575.
[16] D. Wang, Surf. Coat. Technol. 36 (1988) p.49.
[17] S.E. Sadique, A.H. Mollah, M.S. Islam, M.M. Ali, M.H.H. Megat, S.Basri, Oxid. Met.
54 (5–6), (2000) p.385.
[18] H. Choi, B. Yoon, H. Kim, C. Lee, Surf. Coat. Technol. 150 (2–3) (2002) pp.297–308.