Corrosion Science 82 (2014) 437–441

Contents lists available at ScienceDirect

Corrosion Science

journal homepage: www.elsevier .com/ locate /corsc i

Short Communication

Improved corrosion resistance in simulated concrete pore solutionof surface nanocrystallized rebar fabricated by wire-brushing

http://dx.doi.org/10.1016/j.corsci.2014.01.0340010-938X/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author.E-mail address: aibin-ma@hhu.edu.cn (A. Ma).

Dan Song a,b,c, Aibin Ma a,⇑, Wei Sun b, Jinghua Jiang a,c, Jinyang Jiang b, Donghui Yang a, Guanghui Guo a

a College of Mechanics and Materials, Hohai University, Nanjing 210098, PR Chinab School of Materials Science and Engineering, Southeast University, Nanjing 210096, PR Chinac Nantong Ocean and Coastal Engineering Research Institute, Hohai University, Nantong 226000, PR China

a r t i c l e i n f o a b s t r a c t

Article history:Received 17 September 2013Accepted 24 January 2014Available online 31 January 2014

Keywords:A. Steel reinforced concreteB. EISB. PolarizationC. PassivityC. Pitting corrosion

Continuous surface nanocrystallization (SNC) of rebar was achieved through wire-brushing process. Auniform NC layer with thickness of 25 lm and average grain size of 50 nm was formed on the rebar sur-face. Due to the enhanced passivation performance of the NC layer, corrosion resistance of the SNC rebarwas significantly improved in Cl�-containing saturated Ca(OH)2 solution. High-energetic crystal defectsof the nano-grains leads to the faster passivation and enhanced stability of the passive film of the SNCrebar.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Rebar corrosion is the key factor leading to the deterioration ofreinforced concrete structure (RCS), therefore, improving corrosionresistance of rebar is very important to extend the life of RCS [1–3].Recent researches have confirmed that the rebar corrosion is mainlyinduced by chloride ion (Cl�) erosion and alkalinity reduction of theconcrete [4,5]. As the corrosion of the rebar initiates from the sur-face and gradually extends into the matrix, it is believed that thecorrosion resistance of the surface material will significantly affectthe service life of rebar, and anti-corrosion modification of the rebarsurface is an effective method to enhance corrosion resistance ofrebar. Surface nanocrystallization (SNC) can lead to formation ofspecial surface microstructures of the material, which exhibits goodmechanical properties, and its benefit to corrosion resistance of thepassive materials through enhancing passive film performance canalso be expected [6–9]. Until now, lots of methods have been devel-oped to achieve NC layer of the materials, but most of them can onlybe used in small workpieces with regular surface shape [10,11].Differently, the wire-brushing method has good shape adaptability[12], and can be used to achieve continuous SNC processing of arebar with irregular surface shape.

The present work focused on investigating the influence ofsurface nanocrystallization on the surface-microstructure and

corrosion resistance of a rebar processed by wire-brushing.Electrochemical experiments on the SNC rebar were carried outin the simulated concrete pore solutions. The results should behelpful for preparing corrosion-resistant rebar with enhancedpassivation performance and corrosion resistance in concretesubjected to Cl� contamination.

2. Experimental

The schematic illustration of SNC processing of a rebar throughwire-brushing is shown in Fig. 1. The rebar was placed betweentwo symmetrically arranged wire brushes, and the rotation of thewire brushes was driven by electric motor with adjustable rotationspeed. The strength was loaded on the wire brushes to keep thewires contacting with the rebar surface appropriately during theprocessing. The feeding wheels were used to provide enough stiff-ness to the rebar and convey the rebar with appropriate speed.Rebar samples with diameter of 20 mm and length of 500 mm usedfor wire-brushing process were cut form a commercial hot-rolledcarbon rebar, whose chemical composition was of Fe–0.22 wt%C–0.22 wt% Si–1.44 wt% Mn–0.02 wt% Cr–0.022 wt% S–0.025 wt%P–0.02 wt% Ni–0.01 wt% Cu–0.038 wt% V. All the rebar sampleswere SNC processed at room temperature with the feeding speedof 5 mm s�1 and the wire brush rotation speed of 8000 r min�1.In order to obtain a uniform NC layer, each rebar was processedfor two passes, and the rebar was rotated by 90� before the sec-ond-pass processing. After SNC processing, the microstructure

Fig. 1. Illustration of SNC processing of rebar through wire-brushing.

438 D. Song et al. / Corrosion Science 82 (2014) 437–441

change of the rebar was observed by the optical microscope (OM,Olympus BX51M, Japan) and the scanning electric microscope(SEM, Hitachi S3400N, Japan). The microstructure characterizationof the NC layer was also observed by the transmission electronmicroscopy (TEM, Tecnai G2, Holland). The passivation perfor-mance and corrosion resistance of the SNC rebar in simulated con-crete pore solutions were studied via the Parstat 2273 advancedpotentiostat (USA), comparing to the as-received rebar and scale-removed rebar (polished by emery papers up to 1000 level).Three-electrodes system was used, which included a saturated cal-omel reference electrode, a Pt counter electrode and a workingelectrode. Rebar samples with length of 10 mm were used as elec-trochemical testing samples, that the cylindrical surface of therebar was exposed to the tested solutions and two cross-sectionsof the rebar sample were sealed by epoxy. Copper wire was linkedto one of the cross-sections, and the rebar sample was connectedas working electrode. For a good reproducibility, all the sampleswere cleaned with acetone and dired in warm air, and at least threereplicates were run for each sample. Three kinds of electrochemicalmeasurements, namely open circuit potential (OCP), electrochem-ical impendence spectroscopy (EIS) and potentiodynamic polariza-tion (PDP), were applied. The rebar samples were immersed in thesaturated Ca(OH)2 solution with pH value of 12.6 for 5 days toobtain stable passivation, and then they were immersed in0.05 M NaCl added saturated Ca(OH)2 solution to be corroded byCl�. The frequency of EIS tests ranged from 10 kHz to 10 mHz,and the amplitude of the sinusoidal potential signal was 5 mV withrespect to the open circuit potential.

3. Results and discussion

Fig. 2 shows the macro-morphology and microstructure of theas-received and SNC rebars, which demonstrates that the rebarscale has been removed completely after brushing repeatedly byhigh-speed rotating wires. The treated rebars presented brightmetallic luster, and obvious plastic-processing waves can also beobserved on the rebar surface. The cross-sectional microstructureof the treated rebar was observed by OM and SEM. Gradient micro-structure was found according to the OM image of the treatedrebar, where the rebar matrix was composed of initial coarsegrains with ferrite and pearlite, and the surface grains of the rebarhad been modified severely. The total thickness of the microstruc-ture-modified layer was about 40 lm, presenting obvious gradient

plastic flows. According to the intensity of the plastic flows, themodified surface layer can be divided into two parts, where theouter part of the layer was severely-deformed zone with thicknessabout 25 lm, and the inner part of the layer was lightly-deformedzone with thickness about 15 lm. According to the SEM image ofthe severely-deformed zone, the grains had been completelyrefined and deformed. Obvious plastic flows were formed in theseverely-deformed zone, where it was difficult to distinguish theferrite and pearlite. TEM was applied to observe the detailedmicrostructure characteristics of the severely-deformed zone andthe results are shown in Fig. 3. Fig. 3a is the TEM image of theseverely-deformed ferrite grains with low magnification, whichclearly shows that the ferrite grains has been extremely refinedinto equiaxed grains with average grain size of 50 nm. Meanwhile,the selected electron diffraction pattern of the deformed ferritegrains shows the distinct ring, which infers to high-angle grainboundaries of the deformed ferrite nano-grains [13]. The high-angle grain boundary is non-equilibrium grain boundary, whichstores a lot of internal energies. The severe strain also created lotsof intragranular dislocations. As shown in Fig. 3b, mass of disloca-tions are stacked in the deformed ferrite nano-grains. So, accordingto the TEM observation, the severely-deformed zone of the modi-fied layer can be defined as nanocrystalline (NC) layer. With thegradual reduced strains from surface to matrix, the deformationof the inner part of the modified layer was much lighter than thatof the NC layer. The grains of ferrite and pearlite of this part werelightly deformed and elongated along with the plastic flows. Thisinner part of the modified surface layer can be regarded as a tran-sition layer between the NC layer and matrix. Based on the system-atical microstructure observation, the wire-brushing treated rebarcan be defined as the SNC rebar.

OCP is the mixed electrode potential in the given corrosionenvironment, which is mainly determined by the surface state ofthe electrode. When rebar is immersed in saturated Ca(OH)2 solu-tion, continuous passivation will be processed. With the effect ofalkalinity and dissolved oxygen of saturated Ca(OH)2 solution, pas-sive film will be formed on the rebar surface [14,15]. OCP has closerelationship with the integrity and density of the rebar’s passivefilm, the nobler OCP value, the better passivation state of the rebarwith more complete and denser passive film. Fig. 4 shows contin-uous OCP evolution of the rebars passivated in saturated Ca(OH)2

solution for 5 days. All kinds of rebars were obtained effectivepassivation during the whole immersion period in saturated

Fig. 2. Macro-morphology and microstructure of the as-received and SNC rebars.

D. Song et al. / Corrosion Science 82 (2014) 437–441 439

Ca(OH)2 solution, presenting gradually elevated OCP values. All theOCP values were elevated rapidly during the initial passivationperiod for 1 day, and were slowly elevated during the followingpassivation period for 2–5 days. However, the SNC rebar presentednobler OCP values during the whole passivation period than thoseof the scale-removed rebar and as-received rebar. After the SNCrebar was passivated for 5 days, the OCP values was about�200 mVSCE, 40 mV and 60 mV nobler than those of the scale-removed rebar and as-received rebar, respectively. Furthermore,the rate of OCP increase of the SNC rebar during the first 1 dayimmersion period was much larger than those of the scale-removed and as-received rebars, which indicated the faster passiv-ation of the SNC rebar.

Recently, EIS spectra has been widely used to evaluate thepassivation performance of rebar in concrete [16,17]. Fig. 5 showsEIS Nyquist plots (Fig. 5a) and Bode plots (Fig. 5b) of the rebars pas-sivated in saturated Ca(OH)2 solution for 2 h and 5 days. As seen inFig. 5, each Nyquist plot of the rebars was dominated by a capaci-tative arc. The arc diameters of the SNC rebar were larger than bothof the scale-removed and as-received rebars after passivated for2 h and 5 days. Meanwhile, seen from the Bode plots, the phaseangles of the SNC rebar tested in middle and low frequency werealso larger than both of the other two kinds of rebars. Larger capa-citative arc diameters and phase angles represented a better pas-sivation of the SNC rebar intuitively. For a better explanation ofthe passivation performance of the tested samples, the experimen-tal impedance data were simulated by the Rs(QdlRp) equivalent cir-cuit via ZSimpwin commercial software (USA), where Rs, Rp and Qdl

represent the electrolyte resistance, the polarization resistance ofthe passive film and the double-layer capacitance of the interfaceof solution/rebar, respectively. After passivated in saturatedCa(OH)2 solution for 5 days, the simulated Rp value of the SNCrebar is about 3.5 � 105 X cm2, 40% and 50% higher than those ofthe scale-removed rebar (about 2.5 � 105 X cm2) and the as-received rebar (about 2.3 � 105 X cm2), respectively. Thus, theOCP and EIS experiment results indicate that the SNC rebar has

rapider and elevated passivation performance in saturated Ca(OH)2

solution than the other two kinds of rebars.Cl� contamination is a key factor leading to the corrosion dam-

age of a rebar in the concrete. According to the literature [2], therebar corrosion in concrete can be roughly divided into two peri-ods, where the first period is the maintaining and breakdown ofpassivation, which will take the longest time in the whole servicelife of a rebar; the second period is the corrosion propagation ofa rebar and corrosion failure of RCS, which will take less time thanthe first period. During the first period, thinning and breaking ofthe passive film of the rebar will occur under the continuous attackof Cl�. Thus, the stability of the passive film subjected to Cl� con-tamination is critical to the corrosion resistant of rebar in theconcrete.

Fig. 6 shows the PDP curves of the rebars tested in the saturatedCa(OH)2 solution containing 0.05 M NaCl. The rebars were pre-pas-sivated in the saturated Ca(OH)2 solution for 5 days, and thenimmersed in 0.05 M NaCl added saturated Ca(OH)2 solution for20 min before the PDP tests. All the PDP curves presented intensi-tive anodic polarization, indicating effective passivation of thethree kinds of rebars in Cl�-containing saturated Ca(OH)2 solution.Pitting potential (Epit) and passivation maintaining current are twoimportant electrochemical parameters in a passivation system,which can characterize corrosion resistance of a material directly.As seen in Fig. 6, the Epit value of the SNC rebar was 640 mVSCE,180 mV and 233 mV nobler than those of the scale-removed rebarand the as-received rebar, respectively. Meanwhile, the lowerpassivation maintaining current had also achieved by the SNCrebar. Nobler Epit value and lower passivation maintaining currentindicate that the SNC rebar has markedly improved pitting resis-tance than scale-removed and as-received rebars.

The above experimental results reveal that the passivation per-formance and corrosion resistance of the rebar in the simulatedconcrete pore solution can be enhanced markedly by surface naon-crystallization processing through wire-brushing. Since the surfacemicrostructure of rebar can affect its corrosion resistance

Fig. 3. TEM observation of NC grains obtained from the NC layer of the rebar. (a)Grains size and SAED pattern with low magnification; (b) Intragranular dislocationsin nano-grains with high magnification.

Fig. 4. Open circuit potential of the rebars passivated in saturated Ca(OH)2 solutionfor 5 days.

Fig. 5. EIS spectra of the rebars passivated in saturated Ca(OH)2 solution for 2 h and5 days. (a) Nyquist plots; (b) Bode plots.

440 D. Song et al. / Corrosion Science 82 (2014) 437–441

significantly, the enhanced passivation performance and corrosionresistance of the SNC rebar should have close relationship with thespecial microstructure characteristic of its NC layer. It has beenwildly believed that rebar can be protected from corrosion by apassive iron oxide film (c-FeOOH) in the alkaline environment pro-vide by concrete [18–20]. So, the performance of the iron oxidefilm is a key factor to the rebar’s corrosion resistance, which isdetermined by both concrete environment and the surface micro-structure of a rebar. For a given concrete environment (includingalkalinity, oxygen concentration and aggressive ions concentra-tion), the performance of the iron oxide film will be mainly deter-mined by its surface microstructure. Many references have

confirmed that the passive film (oxide film and/or hydrated oxidefilm) are prone to nucleate at the surface crystalline defects[21,22]. Due to the severe strains imposed by surface nanocrystal-lization processing through wire-brushing, the NC layer of the SNCrebar possesses special detailed microstructures (such as extre-mely fine grains, large-angle grain boundaries and mass of stackeddislocations), which can be found clearly in the TEM image shownin Fig. 3. The energy stored in the crystalline defects within thenano-grains can provide the SNC rebar more nucleate sites to form

Fig. 6. Potentiodynamic polarization curves of the rebars tested in saturatedCa(OH)2 solution containing 0.05 M NaCl.

D. Song et al. / Corrosion Science 82 (2014) 437–441 441

the passive film. As soon as the SNC rebar surfaces contact withalkaline solution, more intense formation of the iron oxide film willhappen, and the obtained iron oxide film will be thicker, denserand more stable than those of the common rebars. Similar phe-nomena were also reported in the SNC AISI 304 and 409 stainlesssteels fabricated by surface mechanical attrition treatment (SMAT)[9,10].

Some researchers have also pointed out that surface state of therebar, such as scale, roughness and pre-rust layer, can affect itscorrosion resistance [23]. Generally, the scale will deteriorate cor-rosion resistance of the rebar in concrete under the attack of Cl�

due to its micro-cracks and porosity, while improving surfacefinishing of a rebar will benefit its corrosion resistance [24,25]. Inthis study, electrochemical experiments results indicated that thescale-removed rebar presented better corrosion resistance thanthe as-received rebar. After the SNC processing through wire-brushing, the rebar’s scale was removed completely, and the elim-ination of scale did improve the corrosion resistance of the SNCrebar.

4. Conclusion

Continuous surface nanocrystallization of rebar has beenachieved through wire-brushing. Due to the imposed large strainsduring the processing, a uniform NC layer with thickness of 25 lmand average grain size of 50 nm can be formed on the rebar surface.The SNC processing of the rebar can improve its corrosion resis-tance in Cl�-containing saturated Ca(OH)2 solution significantly.Enhanced passive film of the SNC rebar is the key factor leadingto its improved corrosion resistance. Rapider passivation andenhanced stability of the passive film are benefited from high-energetic crystal defects of the nano-grains, such as large-anglegrain boundaries and mass of intragranular dislocations. Elimina-tion of rebar scale during wire-brushing process also provide activeeffect to corrosion resistance improvement of the SNC rebar.

Acknowledgements

This work was supported by Natural Science Foundation ofChina (51308111 and 51278098), Natural Science Foundation ofJiangsu province of China (BK2011249), China Postdoctoral ScienceFoundation (2013M531249), Postdoctoral Science Foundation of

Jiangsu province of China (1202008C) and Applied Research Foun-dation of Nantong city (BK2013001). Professor Xinmin Fan ofNanjing University of Science and Technology was also acknowl-edged for providing wire-brushing equipment.

References

[1] U. Angst, B. Elsener, C.K. Larsen, Øystein Vennesland, Critical chloride contentin reinforced concrete – a review, Cem. Concr. Res. 39 (2009) 1122–1138.

[2] R.D. Moser, P.M. Shingh, L.F. Kahn, K.E. Kurtis, Chloride-induced corrosionresistance of high-strength stainless steels in simulated alkaline andcarbonated concrete pore solutions, Corros. Sci. 57 (2012) 241–253.

[3] S. Ahmad, Reinforcement corrosion in concrete structures, its monitoring andservice life prediction – a review, Cem. Concr. Compos. 25 (2003) 459–471.

[4] G.K. Glass, N.R. Buenfeld, The presentation of the chloride threshold level forcorrosion of steel in concrete, Corros. Sci. 39 (1997) 1001–1013.

[5] K.Y. Ann, H.W. Song, Chloride threshold level for corrosion of steel in concrete,Corros. Sci. 49 (2007) 4113–4133.

[6] X.H. Chen, J. Lu, L. Lu, K. Lu, Tensile properties of a nanocrystalline 316Laustenitic stainless steel, Scripta Mater. 52 (2005) 1039–1044.

[7] N.R. Taoa, Z.B. Wanga, W.P. Tonga, M.L. Suia, J. Lub, K. Lua, An investigation ofsurface nanocrystallization mechanism in Fe induced by surface mechanicalattrition treatment, Acta Mater. 50 (2002) 4603–4616.

[8] G. Liu, S.C. Wang, X.F. Lou, J. Lu, K. Lu, Low carbon steel with nanostructuredsurface layer induced by high-energy shot peening, Scripta Mater. 44 (2001)1791–1795.

[9] T. Balusamy, S. Kumar, T.S.N. Sankara Narayanan, Effect of surfacenanocrystallization on the corrosion behaviour of AISI 409 stainless steel,Corros. Sci. 52 (2010) 3826–3834.

[10] T. Balusamy, T.S.N. Sankara Narayanan, K. Ravichandran, K. Ravichandran, I.l.Song Park, M.H. Lee, Influence of surface mechanical attrition treatment(SMAT) on the corrosion behaviour of AISI 304 stainless steel, Corros. Sci. 74(2013) 332–344.

[11] T.S. Wang, J.K. Yu, B.F. Dong, Surface nanocrystallization induced by shotpeening and its effect on corrosion resistance of 1Cr18Ni9Ti stainless steel,Surf. Coat. Technol. 200 (2006) 4777–4781.

[12] H. Kitahara, T. Yada, F. Hashiguchi, M. Tsushida, S. Ando, Mg alloy sheets with ananocrystalline surface layer fabricated by wire-brushing, Surf. Coat. Technol.,Corrected Proof, 2012 (in Press, http://dx.doi.org/10.1016/j.surfcoat.2012.04.020).

[13] K.Y. Zhua, A. Vasselb, F. Brissetc, K. Lud, J. Lua, Nanostructure formationmechanism of a-titanium using SMAT, Acta Mater. 52 (2004) 4101–4110.

[14] P. Ghods, O.B. Isgor, G.J.C. Carpenter, J. Li, G.A. McRae, G.P. Gu, Nano-scalestudy of passive films and chloride-induced depassivation of carbon steelrebar in simulated concrete pore solutions using FIB/TEM, Cem. Concr. Res. 47(2013) 55–68.

[15] S.M. Abd El Haleem, E.E. Abd El Aal, S. Abd El Wanees, A. Diab, Environmentalfactors affecting the corrosion behaviour of reinforcing steel: I. The early stageof passive film formation in Ca(OH)2 solutions, Corros. Sci. 52 (2010) 3875–3882.

[16] P. Ghods, O.B. Isgor, G.A. McRae, G.P. Gu, Electrochemical investigation ofchloride-induced depassivation of black steel rebar under simulated serviceconditions, Corros. Sci. 52 (2010) 1649–1659.

[17] J.K. Singh, D.D.N. Singh, The nature of rusts and corrosion characteristics of lowalloy and plain carbon steels in three kinds of concrete pore solution withsalinity and different pH, Corros. Sci. 56 (2011) 129–142.

[18] P. Ghods, O.B. Isgor, J.R. Brown, F. Bensebaa, D. Kingston, XPS depth profilingstudy on the passive oxide film of carbon steel in saturated calcium hydroxidesolution and the effect of chloride on the film properties, Appl. Surf. Sci. 257(2011) 4669–4677.

[19] Z.H. Dong, W. Shi, X.P. Guo, Initiation and repassivation of pitting corrosion ofcarbon steel in carbonated pore solution, Corros. Sci. 53 (2011) 1322–1330.

[20] Y.X. Zhao, Y.Y. Wu, W.L. Jin, Distribution of mill scale on corroded steel barsand peneration of steel corrosion products in concrete, Corros. Sci. 66 (2013)160–168.

[21] K.D. Ralston, N. Birbilis, C.H.J. Davies, Revealing the relationship between grainsize and corrosion rate of metals, Scripta Mater. 63 (2010) 1201–1204.

[22] L.Y. Zhang, A.B. Ma, J.H. Jiang, D.H. Yang, D. Song, J.Q. Chen, Sulphuric acidcorrosion of ultrafine-grained mild steel processed by equal-channel angularpressing, Corros. Sci. 75 (2013) 434–442.

[23] Z.Q. Tan, C.M. Hansson, Effect of surface condition on the initial corrosion ofgalvanized reinforcing steel embedded in concrete, Corros. Sci. 50 (2008)2512–2522.

[24] L.T. Mammoliti, L.C. Brown, C.M. Hansson, B.B. Hope, The influence of surfacefinish of reinforcing steel and ph of the test solution on the chloride thresholdconcentration for corrosion initiation in synthetic pore solutions, Cem. Concr.Res. 26 (1996) 545–550.

[25] P. Novak, R. Mala, L. Joska, Influence of pre-rusting on steel corrosion inconcrete, Cem. Concr. Res. 31 (2001) 589–593.