- Email: [email protected]
The effect of chloride and sulfate ions on reinforcement corrosion
CEMENT and CONCRETE RESEARCH. Vol. 23, pp. 139-146, 1993. Printed In the USA. 0008-8846/93. $6.00+.00. Copyright © 1993 Pergamon Press Ltd.
THE EFFECT OF CHLORIDE AND SULFATE IONS ON REINFORCEMENT CORROSION Omar Saeed Baghabra AI-Amondi Mohammed Maslehuddin
King Fahd University of Petroleum and Minerals Dhahran 31261, Saudi Arabia (Communicated by C.D. Pomeroy) (Received Jan. 2, 1992)
ABSTRACT The effect of chloride, sulfate and chloride-sulfate solutions on corrosion of steel embedded in cement paste has been investigated. The reinforcement corrosion was evaluated by measuring corrosion potentials and corrosion current density using D.C. linear polarization resistance technique. Results indicate that the corrosion activity was very minimal in specimens immersed in pure sulfate solution. The reinforcement corrosion activity was found to be higher in specimens immersed in chloride-sulfate solutions as compared to those immersed in pure chloride solution. The corrosion rate was observed :o be doubled when the sulfate concentration in 15.7% C1- solution was raised from 0.55 to 2.1%.
INTRODUCTION Concrete provides both physical and chemical protection to reinforcing steel. The physical protection is provided by its impermeable structure which retards the ingress of aggressive agents like moisture, oxygen, carbon dioxide and other fluids into its interior. The chemical protection is provided by the high pH of the pore solution which forms a submicroscopically thin protective film of gamma ferric oxide on the steel surface. The integrity and protective quality of this film depends on the alkalinity (pH) of the environment. Well-hydrated portland cement containing from 15 to 30 percent calcium hydroxide and other alkalies usually maintains the pore solution pH in the range of 13-13.5. The protective film prevents iron cations (Fe++) from entering into solution in the electrolyte and acts as a barrier to prevent oxygen anions (O-) from contacting the steel surface. However, this protective film can be disrupted either by a reduction in the pH of the pore solution due to carbonation, or by the penetration of aggressive ions like chlorides to the steel-concrete interface. The chlorides may be introduced into concrete through an accelerator, contaminated aggregates and/or mixing water. Alternatively, chlorides may penetrate the hardened concrete from external environment. Corrosion of reinforcement, due to deicer salts, has resulted in a vast amount of resources to be diverted to the repair and rehabilitation of bridge decks in the US. Deterioration of concrete structures, within a short span of 10 to 15 years, in the coastal areas of the Arabian Gulf is another example where concrete has failed to come up to its expectations of maintenance-free performance. 139
140
O.S.B. AI-Amoudi and M. Maslehuddin
VoL 23. No. 1
In the former case, reinforcement corrosion is caused primarily due to the chloride ions. The aggressive environmental conditions, characterized by high temperature and humidity variations, atmosphere charged with chloride and sulfate salts, are understood to be the primary factors contributing to the deterioration of concrete structures in the latter case. The elevated temperature in the Arabian Gulf countries accelerates the chemical reactions involved in the corrosion process. High temperature causes a great difference in the degree of concrete deterioration and reinforcement corrosion. The penetration of aggressive species such as chloride ions and carbon dioxide proceeds more rapidly at elevated temperatures. The rate of corrosion appears to be accentuated by an increase in temperature in the range of 20 to 40° C, especially at high humidity. The concomitant presence of sulfate ions with chlorides is believed to be another cause of accelerated corrosion observed in the Arabian Gulf countries. Studies carried out by Holden et al [1] on the pore solution composition of pastes made with irucedquantities of chloride and sulfates, indicated an increase in the OH- ion concentration due to inclusion of sulfates as compared to the alkalinity of pore solution of cement pastes contaminated with similar quantities of chlorides salts. Their results also showed a substantial decrease in the chloride binding capacity of cements to which sodium chloride and sodium sulfate were added. These results reflect the tendency of sulfate ions to react preferentially with the C3A phase, thus inhibiting the formation of calcium chloroaluminate (Friedel's salt). Thus, corrosion risk is likely to be significantly increased in circumstances where concrete is exposed to both chloride and sulfate salts. AI-Tayyib et al [2] report seven fold increase in corrosion activity in mild steel exposed to sulfate containing calcium hydroxide solutions over those containing chloride salts. Cheng et al [3] have investigated the effects of chloride and sulfate ions on the electrochemical properties of reinforcing steel using AC impedance technique. Their results indicated that presence of sulfate ions could modify the reinforcement surface characteristics. While, the studies cited above have indicated a varying role of sulfate ions, vis-a-vis reinforcement corrosion, there is a need to investigate the mechanisms of reinforcement corrosion in sulfate, chloride and chloride-sulfate environments. The authors feel, such studies should be carried out on steel in cement paste or concrete as the kinetics of reinforcement corrosion in such media is expected to be different from that in simulated pore solution. EXPERIMENTAL PROGRAM Soecimens Steel bars, 6 mm in diameter, were cast in ordinary portland cement paste specimens of 31 x 31 x 152 ran-. A cover of 12.5 nun was provided on all the sides. The bars were degreased with acetone before casting in cement paste. The ends of the bars were coated with an epoxy coating to prevent crevice corrosion effect. The specimens were cured in water for 14 days and then exposed to the test solutions. Table 1 shows the chemical composition of the cement used. Test Solutions The specimens were placed in four test solutions. The composition of the test solutions is as detailed below: Solution 1: Solution 2: Solution 3: Solution4:
15.7% C12.1% S04-" 0.55% S04"" + 15.7% CI 2.1% SO4"'+ 15.7% Cl
Test solutions 3 and 4 represent the concentrations of chloride and sulfate ions commonly found in the continental sabkha brines [4] and coastal sabkha brines [5] respectively in the Arabian Gulf countries. Test solutions 1 and 2 represent the maximum chloride and sulfate concentrations in these solutions.
Vol. 23. No. 1
CHLORIDES, SULFATES,CORROSION, CEMENT PASTE
141
Sodium chloride was used to provide the chloride ions, while sodium sulfate and magnesium sulfate were used to provide the sulfate ions. The latter two salts were so proportioned as to provide 50% of the sulfate concentration from each of them. TABLE 1 Chemical Composition of Cement Constituent
Weight(%)
Silicon Dioxide Aluminum Oxide Ferric Oxide Calcium Oxide Magnesium Oxide Sulfur Trioxide Loss on Ignition Potassium Oxide Sodium Oxide
20.5 5.6 3.8 64.4 2.1 2.1 0.7 0.3 0.2
C3S C2S
56.7 16.1 8.5 11.6
C3A C4AF
Corrosion Monitorin~ Reinforcement corrosion was monitored by evaluating time to initiation of corrosion and measuring corrosion current density. The corrosion activity was monitored using a high impedance Voltmeter and recording the potentials with respect to a saturated calomel electrode (SCE). Linear polarization resistance technique was used to obtain quantitative information on corrosion of reinforcement. The polarization resistance (Rp) was determined by conducting a polarization scan in the range o f + 10 mV o f the corrosion potential. A microprocessor-based Potentiostat/Galvanostat was used for polarizing the steel. A stainless steel frame placed outside the specimen was used as a counter electrode, while a saturated calomel electrode was used as a reference electrode. Figure 1 is a schematic representation of the test set-up. A scan rate of 0.1 mV/sec was used. Positive feed back technique was used to compensate for the ohmic drop (IR) between the reference electrode and the reinforcing bar. The corrosion current density, Icorr, was determined using Stern and Geary formula [6]:
Where:
Icorr = B/Rp Icorr = corrosion current density, gA/crn 2 Rp = polarization resistance, Ohms.cm 2 B = (Ba.flc)/2.3(Ba + lhz)
where Ba and 13c arc the anodic and cathodic Tafel constants. For steel in aqueous media, values of Ba and Bc equal to 100 mV are normally used. However, in the absence of sufficient data on Ba and Bc for steel in concrete, a value of B equal to 52 mV for steel in the passive condition and equal to 26 mV for steel in the active condition arc used. Gonzalez and Andradc have demonstrated a good correlation between corrosion rates determined by linear polarization resistance technique and weight loss measurements for active and passive states of corrosion of steel in concrete [7,8].
142
O.S.B.AI-Amoudiand M. Maslehuddin
DATA
I
POTENTIOSTAT/ fiALVANOSTAT
I Ar.auISITIONI
I SYsT
Vol. 23, No. I
I
I
] / /
_v____
•
I
I
/
EPOXYCOATINfi
STAINLESS STEEL COURTER ELECTRODE SATURATED CALOHFJ. ELECTRODE
i
71
111
~
- PLASTIC TANK
i
3 h 1S2mmmCEMENT MORTAR SPECIMEN
TEST SOLUTION
6ram STEEL BAR
FIGURE 1 Experimental Set-up for Measurement of Corrosion Current Density.
0r,llmKLmal. After 500 days of exposure to the test solutions the specimens were broken along the reinforcing bar and the paste surrounding the steel bar was collected. The paste sample was crushed to pass 150 I~m sieve. This material was used to determine the water-soluble chlorides and sulfates. The water soluble chloride and sulfate ions were extracted by dispersing 5 grams of powdered sample in 100 ml of distilled water, stirring for 5 minutes, then standing for 24 hours. The paste-water mixture was f'dtered and made to 200 ml. The chlorides were analyzed by titrating against standard mercuric nitrate solution. The water-soluble sulfates were analyzed using spectrophotometric technique [9]. RESULTS The time-corrosion potential record for the specimens immersed in the four test solutions is shown in Figure 2. The corrosion potentials (Figure 2) represent average values of measurements carried out on three specimens representing similar mix composition and placed in a similar solution. These data indicate active corrosion of reinforcement, determined on the basis of A S T M C 876 criteriaof -270 m V SCE, in paste specimens imnm'sed in solutions I, 3 and 4. These solutions are all characterized by the presence of chloride ions either alone or associated with sulfate salts. The corrosion potentials of steel in specimens immersed in the pure sulfate solution was much lower than -270 m V SCE even after 500 days of immersion. The data on time to initiationof cohesion and the corrosion potentials of steel after 500 days of exposure arc shown in Table 2. The time to initiationof corrosion was observed to be about 20 days in specimens immersed in chloride solution (solution I), where as itwas 55 to 57 days in specimens immersed in solution 3 and 4 (chloride-sulfatesolutions). The corrosion potentialsof steelin paste specimens immersed
Vol. 23, No. 1
CHLORIDES, SULFATES, CORROSION, CEMENT PASTE
143
900 m,l
un800 -
700. . . .
~TE
_.~'600I-z
500--
I,m O
~-.N'
o_ tOO-
•
--O-- 15.7% EHLORIDE
.--J --,I
2.1% SULFATE .
"L.J ' 300-
~
=<-
0% EHLORIDE 0.55% SULFATE +
THRESHOLD POTENTIAL, -270 mV (SEE)
° •--R---
00- o o.
15.7% EHLORIDE 2.1% SULFATE * 15.7% EHLORIDE
100 -
0 . , , ~ - ~ - . w - - ¢ . . . ¢I . . ~ 0
.-.-
~ ~ I . I ~00 300 TIME OF IMMERSION, DAYS
~_~: I
100
200
I
500
600
FIGURE 2 Time-Corrosion Potential Record for Specimens Immersed in the Test Solutions TABLE 2 Time to Initiation of Corrosion and Corrosion Potentials of Steel in Cement Paste Specimens. Solution Composition 15.47% CI" 2.1% SO4-15.7% CI" + 0.55 SO4"15.7% CI" + 2.1% SO4--
Time to Initiation of Corrosion (Days) 20 still passive 57 55
Corrosion Poten (mV) -800 -20 -705 -743
in solution 1 (El- 15.7%) after 500 days was -800 mV SCE compared to corrosion potentials of 20, -705 and -743 mV measured on specimens immersed in solutions 2, 3, and 4 respectively. The data on corrosion current de~ity on steel in cement paste specimens, immersed in the four test solutions for 500 days, are shown in Table 3. DISCUSSION
The corrosion data developed in this investigation indicate that the concomitant presence of chloride and sulfate salts increases the reinforcement corrosion activity. The corrosion potentials of steel in cement paste specimens immersed in chloride and chloride-sulfate solutions were in the range of705 to -800 mV SCE. The corrosion potential of steel in specimens immersed in the sulfate solution was -20 mV SEE, indicating no corrosion activity. The corrosion current data also
144
O.S.B. A1-Amoudiand M. Maslehuddin
Vol. 23, No. 1
TABLE 3 Corrosion Current Density on Bars in Cement Paste Specimens Solution Composition
15.7% CI" 2.1% SO4-15.7% CI" + 0.55 SO4"" 15.7% CI" + 2.1% SO4""
Corrosion Current Density O~A/cm2)
4.30 0.018 4.82 9.05
indicate a similar trend. The corrosion current density on steel in specimens placed in the sulfate solution was less than 0.2 gA/cm 2 generally used as a criterion for corrosion activation. The corrosion current density on steel in specimens placed in 15.7% C1- + 2.1% SO4" was 2.11 times the corrosion current density on bars in specimens placed in the 15.7% C1- solution. The corrosion current density on bars in specimens placed in the pure chloride solution (solution 1) was 70 times that of specimens placed in 2.1% SO4"" solution. Visual inspection of the specimens after 500 days of immersion in the test solutions indicated cracking of the paste specimens along the steel in specimens placed in solution 4 (2.1% SO4"" + 15.7% (21). Deterioration akin to eating away of the surface skin was observed in specimens placed in solution 2 (2.1% SO4"-). No deterioration of any type was observed in specimens placed in solution 1 (15.7% CI-) and solution 3 (0.55% SO4"- + 15.7% C1-). The low corrosion of bars in paste specimens placed in solution 2 (2.1% SO4"-) indicates that sulfate ions do not influence reinforcement corrosion activation in cement paste specimens. The presence of chloride ions is necessary for the initiation and propagation of reinforcement corrosion. The precise role of these two ions in influencing reinforcement corrosion is not very well understood. However, it is envisaged that chloride ions are primarily responsible for depassivation of the passive film. This provides a direct access of the metal surface to both the chloride and sulfate ions. At the anode, iron dissolves into the electrolyte. Fe = Fe ++ + 2e-
(1)
After the formation of ferrous ions, both chloride and sulfate ions compete with each other for the formation of ferrous chloride and ferrous sulfate respectively. Fe ++ + 2C1- = FeC12 F e ++ +
(SO4)" = FeSO4
(2) (3)
Once ferrous chloride and ferrous sulfate are formed, the presence of moisture and oxygen near the corrosion product/metal interface promote other types of reactions. Ferrous chloride is gradually converted into iron oxyhydroxide (FeOOH) and hydrochloric acid (HCI) [10]. 2FeC12 + 3H20 + 02 = 2FeOOH + 4HCI
(4)
Ferrous sulfate undergoes similar oxidative hydrolysis reaction with the generation of iron oxyhydroxide and sulfuric acid.
Vol. 23, No. 1
CHLORIDES, SULFATES, CORROSION, CEMENT PASTE
145
(5)
4FeSO4+ 6H20 + 02 = 4FcOOH + 4H2SO4
Both sulfuric acid and hydrochloric acid continue to react with the metal, thus producing more FeCI2 and FeSO4 respectively, and generating H2SO4 and HCI (acid generation cycle). In the presence of chlorides and sulfates the corrosion of steel is compounded due to the reaction of these ions with the metal as compared to corrosion of steel in the presence of chloride ions only. Table 4 shows the concentration of water-soluble chloride and sulfate ions in the paste near the steel-concrete interface. The concentration of water soluble chlorides was approximately 3.6% by weight of cement in the specimens immersed in solutions 1, 3, and 4. The concentration of watersoluble sulfate ions was observed to increase with the increasing concentration of these salts in the test solutions. Figure 3 shows the relationship between corrosion current density and the water soluble sulfate ions associated with approximately similar (3.6%) water-soluble chloride concentration. These data indicate that reinforcement corrosion due to chlorides is not influenced by water-soluble sulfate concentration below 0.18% by weight of cement. For sulfate concentrations above this value (0.18%), the corrosion activity was significantly enhanced. The very minimal corrosion activity evidenced by lower corrosion current density measured on TABLE 4 Concentrations of Water Soluble Chloride and Sulfate Ions Solution Composition
Chloride (%) 3.64 0.18 3.58 3.56
15.7% C12.1% SO4"" 15.7% CI + 0.55 SO4"" 15.7% (21" + 2.1% SO4""
Sulfate (%) 0.05 0.12 0.18 0.30
CHL
90 >- 8.0 I'-- . Z
'-" 7.0 Z eY
~ 6.0 Z
~ 5.0 o¢"
0.0
I
I
I
0.1 0.2 0.3 WATER SOLUBLE SULFATE (% WT. CEMENT)
FIGURE 3 Influence of Sulfate Ions Associated with 3.6% Water Soluble Chlorides on Reinforcement Corrosion.
146
O.S.B. AI-Amoudi and M. Maslehuddin
Vol. 23, No. 1
reinforcing steel in specimens placed in pure sulfate solution indicates that, within the duration of this study, sulfate ions do not influence the corrosion process. The concomitant presence of chloride and sulfate ions is necessary for the initiation and propagation of reinforcement corrosion in cement concrete. The quantum of C3 A in plain cements andthe type of the blending material in blended cements may also influence reinforcement corrosion in concrete. As such, more detailed studies are needed to elucidate reinforcement corrosion mechanisms in plain and blended cements placed in chloride and chloride-sulfate solutions. CONCLUSIONS The influence of sulfate, chloride and sulfate-chloride solutions on reinforcement corrosion in cement paste specimens was investigated. It was observed that the corrosion activity measured as time to initiation of corrosion and corrosion current density was minimum in specimens placed in the pure sulfate solution. The time to initiation of corrosion was found to be approximately 20 days in specimens placed in 15.7% c r solution as compared to 56 days in specimens placed in chloride-sulfate solutions. The corrosion current density was slightly higher in 15.7% c r + 0.55% S04- solution compared to 15.7% Cl- solution. The corrosion current density on bars in specimens placed in 15.7% c r + 2.1% s04-- solution was more than double the corrosion current density on bars in specimens placed in 15.7% c r solution. The data developed in this study indicate that, within the duration of this investigation, sulfate ion alone does not influence the reinforcement corrosion process. However, the concomitant presence of sulfate and chloride may significantly affect the rate of corrosion. More detailed studies axe required to elucidate the reinforcement corrosion mechanisms involving cements of varying C3A, blended cements and differing exposure conditions. ACKNOWLEDGEMENT The support of King Fahd University of Petroleum and Minerals in the conduct of this research is gratefully acknowleged. REFERENCES .
.
3. 4. 5.
.
7. 8. 9. 10.
Holden, W.R., Page, C.L., and Short, N.R., Corrosion of Reinforcement in Concrete Construction, Alan P. Crane Ed., Society of Chemical Industry, London, 1983, pp. 143149. AI-Tayyib, A.J., et al. Cement and Concrete Research, 18 (5), 1988, pp. 774-782. Cheng, T., Lee, J., and Tsai, W., Cement and Concrete Research, Vol. 20 (2), 1990, pp. 243-252. Smith, C.L., U.S. Department of the Interior, Geologic Survey, Saudi Arabia Mission, TRI, 1980, p. 26. Johnson, H., Kamal, M.R., Pierson, G.O., and Ramsay, J.B., Quaternary Period in Saudi Arabia, AI-Sayyari, S.S., and Zotl, J.G. (Eds), Springler-Verlag, Austria, 1978, pp. 84-93. Stem, M., and Geary, A.L., Journal, Electrochemical Society, No. 104, 1957, p. 56. Gonzalez, J.A., Algaba, S., and Andrade, C., British Corrosion Journal, Vol. 15, No. 3, 1980, pp. 135-139. Gonzalez, J.A., and Andrade, C., British Corrosion Journal, Vol. 17, No. 1, 1982, pp. 21-28. APHA-AWWA-WPCF, Standard Methodsfor the Examination of Water and Wastewater, 15th Edition, pp. 439-440. Evans, U.R., The Corrosion and Oxidation of Metals, First Supplementary Volume, 1968, Edward Arnold, London.
THE EFFECT OF CHLORIDE AND SULFATE IONS ON REINFORCEMENT CORROSION Omar Saeed Baghabra AI-Amondi Mohammed Maslehuddin
King Fahd University of Petroleum and Minerals Dhahran 31261, Saudi Arabia (Communicated by C.D. Pomeroy) (Received Jan. 2, 1992)
ABSTRACT The effect of chloride, sulfate and chloride-sulfate solutions on corrosion of steel embedded in cement paste has been investigated. The reinforcement corrosion was evaluated by measuring corrosion potentials and corrosion current density using D.C. linear polarization resistance technique. Results indicate that the corrosion activity was very minimal in specimens immersed in pure sulfate solution. The reinforcement corrosion activity was found to be higher in specimens immersed in chloride-sulfate solutions as compared to those immersed in pure chloride solution. The corrosion rate was observed :o be doubled when the sulfate concentration in 15.7% C1- solution was raised from 0.55 to 2.1%.
INTRODUCTION Concrete provides both physical and chemical protection to reinforcing steel. The physical protection is provided by its impermeable structure which retards the ingress of aggressive agents like moisture, oxygen, carbon dioxide and other fluids into its interior. The chemical protection is provided by the high pH of the pore solution which forms a submicroscopically thin protective film of gamma ferric oxide on the steel surface. The integrity and protective quality of this film depends on the alkalinity (pH) of the environment. Well-hydrated portland cement containing from 15 to 30 percent calcium hydroxide and other alkalies usually maintains the pore solution pH in the range of 13-13.5. The protective film prevents iron cations (Fe++) from entering into solution in the electrolyte and acts as a barrier to prevent oxygen anions (O-) from contacting the steel surface. However, this protective film can be disrupted either by a reduction in the pH of the pore solution due to carbonation, or by the penetration of aggressive ions like chlorides to the steel-concrete interface. The chlorides may be introduced into concrete through an accelerator, contaminated aggregates and/or mixing water. Alternatively, chlorides may penetrate the hardened concrete from external environment. Corrosion of reinforcement, due to deicer salts, has resulted in a vast amount of resources to be diverted to the repair and rehabilitation of bridge decks in the US. Deterioration of concrete structures, within a short span of 10 to 15 years, in the coastal areas of the Arabian Gulf is another example where concrete has failed to come up to its expectations of maintenance-free performance. 139
140
O.S.B. AI-Amoudi and M. Maslehuddin
VoL 23. No. 1
In the former case, reinforcement corrosion is caused primarily due to the chloride ions. The aggressive environmental conditions, characterized by high temperature and humidity variations, atmosphere charged with chloride and sulfate salts, are understood to be the primary factors contributing to the deterioration of concrete structures in the latter case. The elevated temperature in the Arabian Gulf countries accelerates the chemical reactions involved in the corrosion process. High temperature causes a great difference in the degree of concrete deterioration and reinforcement corrosion. The penetration of aggressive species such as chloride ions and carbon dioxide proceeds more rapidly at elevated temperatures. The rate of corrosion appears to be accentuated by an increase in temperature in the range of 20 to 40° C, especially at high humidity. The concomitant presence of sulfate ions with chlorides is believed to be another cause of accelerated corrosion observed in the Arabian Gulf countries. Studies carried out by Holden et al [1] on the pore solution composition of pastes made with irucedquantities of chloride and sulfates, indicated an increase in the OH- ion concentration due to inclusion of sulfates as compared to the alkalinity of pore solution of cement pastes contaminated with similar quantities of chlorides salts. Their results also showed a substantial decrease in the chloride binding capacity of cements to which sodium chloride and sodium sulfate were added. These results reflect the tendency of sulfate ions to react preferentially with the C3A phase, thus inhibiting the formation of calcium chloroaluminate (Friedel's salt). Thus, corrosion risk is likely to be significantly increased in circumstances where concrete is exposed to both chloride and sulfate salts. AI-Tayyib et al [2] report seven fold increase in corrosion activity in mild steel exposed to sulfate containing calcium hydroxide solutions over those containing chloride salts. Cheng et al [3] have investigated the effects of chloride and sulfate ions on the electrochemical properties of reinforcing steel using AC impedance technique. Their results indicated that presence of sulfate ions could modify the reinforcement surface characteristics. While, the studies cited above have indicated a varying role of sulfate ions, vis-a-vis reinforcement corrosion, there is a need to investigate the mechanisms of reinforcement corrosion in sulfate, chloride and chloride-sulfate environments. The authors feel, such studies should be carried out on steel in cement paste or concrete as the kinetics of reinforcement corrosion in such media is expected to be different from that in simulated pore solution. EXPERIMENTAL PROGRAM Soecimens Steel bars, 6 mm in diameter, were cast in ordinary portland cement paste specimens of 31 x 31 x 152 ran-. A cover of 12.5 nun was provided on all the sides. The bars were degreased with acetone before casting in cement paste. The ends of the bars were coated with an epoxy coating to prevent crevice corrosion effect. The specimens were cured in water for 14 days and then exposed to the test solutions. Table 1 shows the chemical composition of the cement used. Test Solutions The specimens were placed in four test solutions. The composition of the test solutions is as detailed below: Solution 1: Solution 2: Solution 3: Solution4:
15.7% C12.1% S04-" 0.55% S04"" + 15.7% CI 2.1% SO4"'+ 15.7% Cl
Test solutions 3 and 4 represent the concentrations of chloride and sulfate ions commonly found in the continental sabkha brines [4] and coastal sabkha brines [5] respectively in the Arabian Gulf countries. Test solutions 1 and 2 represent the maximum chloride and sulfate concentrations in these solutions.
Vol. 23. No. 1
CHLORIDES, SULFATES,CORROSION, CEMENT PASTE
141
Sodium chloride was used to provide the chloride ions, while sodium sulfate and magnesium sulfate were used to provide the sulfate ions. The latter two salts were so proportioned as to provide 50% of the sulfate concentration from each of them. TABLE 1 Chemical Composition of Cement Constituent
Weight(%)
Silicon Dioxide Aluminum Oxide Ferric Oxide Calcium Oxide Magnesium Oxide Sulfur Trioxide Loss on Ignition Potassium Oxide Sodium Oxide
20.5 5.6 3.8 64.4 2.1 2.1 0.7 0.3 0.2
C3S C2S
56.7 16.1 8.5 11.6
C3A C4AF
Corrosion Monitorin~ Reinforcement corrosion was monitored by evaluating time to initiation of corrosion and measuring corrosion current density. The corrosion activity was monitored using a high impedance Voltmeter and recording the potentials with respect to a saturated calomel electrode (SCE). Linear polarization resistance technique was used to obtain quantitative information on corrosion of reinforcement. The polarization resistance (Rp) was determined by conducting a polarization scan in the range o f + 10 mV o f the corrosion potential. A microprocessor-based Potentiostat/Galvanostat was used for polarizing the steel. A stainless steel frame placed outside the specimen was used as a counter electrode, while a saturated calomel electrode was used as a reference electrode. Figure 1 is a schematic representation of the test set-up. A scan rate of 0.1 mV/sec was used. Positive feed back technique was used to compensate for the ohmic drop (IR) between the reference electrode and the reinforcing bar. The corrosion current density, Icorr, was determined using Stern and Geary formula [6]:
Where:
Icorr = B/Rp Icorr = corrosion current density, gA/crn 2 Rp = polarization resistance, Ohms.cm 2 B = (Ba.flc)/2.3(Ba + lhz)
where Ba and 13c arc the anodic and cathodic Tafel constants. For steel in aqueous media, values of Ba and Bc equal to 100 mV are normally used. However, in the absence of sufficient data on Ba and Bc for steel in concrete, a value of B equal to 52 mV for steel in the passive condition and equal to 26 mV for steel in the active condition arc used. Gonzalez and Andradc have demonstrated a good correlation between corrosion rates determined by linear polarization resistance technique and weight loss measurements for active and passive states of corrosion of steel in concrete [7,8].
142
O.S.B.AI-Amoudiand M. Maslehuddin
DATA
I
POTENTIOSTAT/ fiALVANOSTAT
I Ar.auISITIONI
I SYsT
Vol. 23, No. I
I
I
] / /
_v____
•
I
I
/
EPOXYCOATINfi
STAINLESS STEEL COURTER ELECTRODE SATURATED CALOHFJ. ELECTRODE
i
71
111
~
- PLASTIC TANK
i
3 h 1S2mmmCEMENT MORTAR SPECIMEN
TEST SOLUTION
6ram STEEL BAR
FIGURE 1 Experimental Set-up for Measurement of Corrosion Current Density.
0r,llmKLmal. After 500 days of exposure to the test solutions the specimens were broken along the reinforcing bar and the paste surrounding the steel bar was collected. The paste sample was crushed to pass 150 I~m sieve. This material was used to determine the water-soluble chlorides and sulfates. The water soluble chloride and sulfate ions were extracted by dispersing 5 grams of powdered sample in 100 ml of distilled water, stirring for 5 minutes, then standing for 24 hours. The paste-water mixture was f'dtered and made to 200 ml. The chlorides were analyzed by titrating against standard mercuric nitrate solution. The water-soluble sulfates were analyzed using spectrophotometric technique [9]. RESULTS The time-corrosion potential record for the specimens immersed in the four test solutions is shown in Figure 2. The corrosion potentials (Figure 2) represent average values of measurements carried out on three specimens representing similar mix composition and placed in a similar solution. These data indicate active corrosion of reinforcement, determined on the basis of A S T M C 876 criteriaof -270 m V SCE, in paste specimens imnm'sed in solutions I, 3 and 4. These solutions are all characterized by the presence of chloride ions either alone or associated with sulfate salts. The corrosion potentials of steel in specimens immersed in the pure sulfate solution was much lower than -270 m V SCE even after 500 days of immersion. The data on time to initiationof cohesion and the corrosion potentials of steel after 500 days of exposure arc shown in Table 2. The time to initiationof corrosion was observed to be about 20 days in specimens immersed in chloride solution (solution I), where as itwas 55 to 57 days in specimens immersed in solution 3 and 4 (chloride-sulfatesolutions). The corrosion potentialsof steelin paste specimens immersed
Vol. 23, No. 1
CHLORIDES, SULFATES, CORROSION, CEMENT PASTE
143
900 m,l
un800 -
700. . . .
~TE
_.~'600I-z
500--
I,m O
~-.N'
o_ tOO-
•
--O-- 15.7% EHLORIDE
.--J --,I
2.1% SULFATE .
"L.J ' 300-
~
=<-
0% EHLORIDE 0.55% SULFATE +
THRESHOLD POTENTIAL, -270 mV (SEE)
° •--R---
00- o o.
15.7% EHLORIDE 2.1% SULFATE * 15.7% EHLORIDE
100 -
0 . , , ~ - ~ - . w - - ¢ . . . ¢I . . ~ 0
.-.-
~ ~ I . I ~00 300 TIME OF IMMERSION, DAYS
~_~: I
100
200
I
500
600
FIGURE 2 Time-Corrosion Potential Record for Specimens Immersed in the Test Solutions TABLE 2 Time to Initiation of Corrosion and Corrosion Potentials of Steel in Cement Paste Specimens. Solution Composition 15.47% CI" 2.1% SO4-15.7% CI" + 0.55 SO4"15.7% CI" + 2.1% SO4--
Time to Initiation of Corrosion (Days) 20 still passive 57 55
Corrosion Poten (mV) -800 -20 -705 -743
in solution 1 (El- 15.7%) after 500 days was -800 mV SCE compared to corrosion potentials of 20, -705 and -743 mV measured on specimens immersed in solutions 2, 3, and 4 respectively. The data on corrosion current de~ity on steel in cement paste specimens, immersed in the four test solutions for 500 days, are shown in Table 3. DISCUSSION
The corrosion data developed in this investigation indicate that the concomitant presence of chloride and sulfate salts increases the reinforcement corrosion activity. The corrosion potentials of steel in cement paste specimens immersed in chloride and chloride-sulfate solutions were in the range of705 to -800 mV SCE. The corrosion potential of steel in specimens immersed in the sulfate solution was -20 mV SEE, indicating no corrosion activity. The corrosion current data also
144
O.S.B. A1-Amoudiand M. Maslehuddin
Vol. 23, No. 1
TABLE 3 Corrosion Current Density on Bars in Cement Paste Specimens Solution Composition
15.7% CI" 2.1% SO4-15.7% CI" + 0.55 SO4"" 15.7% CI" + 2.1% SO4""
Corrosion Current Density O~A/cm2)
4.30 0.018 4.82 9.05
indicate a similar trend. The corrosion current density on steel in specimens placed in the sulfate solution was less than 0.2 gA/cm 2 generally used as a criterion for corrosion activation. The corrosion current density on steel in specimens placed in 15.7% C1- + 2.1% SO4" was 2.11 times the corrosion current density on bars in specimens placed in the 15.7% C1- solution. The corrosion current density on bars in specimens placed in the pure chloride solution (solution 1) was 70 times that of specimens placed in 2.1% SO4"" solution. Visual inspection of the specimens after 500 days of immersion in the test solutions indicated cracking of the paste specimens along the steel in specimens placed in solution 4 (2.1% SO4"" + 15.7% (21). Deterioration akin to eating away of the surface skin was observed in specimens placed in solution 2 (2.1% SO4"-). No deterioration of any type was observed in specimens placed in solution 1 (15.7% CI-) and solution 3 (0.55% SO4"- + 15.7% C1-). The low corrosion of bars in paste specimens placed in solution 2 (2.1% SO4"-) indicates that sulfate ions do not influence reinforcement corrosion activation in cement paste specimens. The presence of chloride ions is necessary for the initiation and propagation of reinforcement corrosion. The precise role of these two ions in influencing reinforcement corrosion is not very well understood. However, it is envisaged that chloride ions are primarily responsible for depassivation of the passive film. This provides a direct access of the metal surface to both the chloride and sulfate ions. At the anode, iron dissolves into the electrolyte. Fe = Fe ++ + 2e-
(1)
After the formation of ferrous ions, both chloride and sulfate ions compete with each other for the formation of ferrous chloride and ferrous sulfate respectively. Fe ++ + 2C1- = FeC12 F e ++ +
(SO4)" = FeSO4
(2) (3)
Once ferrous chloride and ferrous sulfate are formed, the presence of moisture and oxygen near the corrosion product/metal interface promote other types of reactions. Ferrous chloride is gradually converted into iron oxyhydroxide (FeOOH) and hydrochloric acid (HCI) [10]. 2FeC12 + 3H20 + 02 = 2FeOOH + 4HCI
(4)
Ferrous sulfate undergoes similar oxidative hydrolysis reaction with the generation of iron oxyhydroxide and sulfuric acid.
Vol. 23, No. 1
CHLORIDES, SULFATES, CORROSION, CEMENT PASTE
145
(5)
4FeSO4+ 6H20 + 02 = 4FcOOH + 4H2SO4
Both sulfuric acid and hydrochloric acid continue to react with the metal, thus producing more FeCI2 and FeSO4 respectively, and generating H2SO4 and HCI (acid generation cycle). In the presence of chlorides and sulfates the corrosion of steel is compounded due to the reaction of these ions with the metal as compared to corrosion of steel in the presence of chloride ions only. Table 4 shows the concentration of water-soluble chloride and sulfate ions in the paste near the steel-concrete interface. The concentration of water soluble chlorides was approximately 3.6% by weight of cement in the specimens immersed in solutions 1, 3, and 4. The concentration of watersoluble sulfate ions was observed to increase with the increasing concentration of these salts in the test solutions. Figure 3 shows the relationship between corrosion current density and the water soluble sulfate ions associated with approximately similar (3.6%) water-soluble chloride concentration. These data indicate that reinforcement corrosion due to chlorides is not influenced by water-soluble sulfate concentration below 0.18% by weight of cement. For sulfate concentrations above this value (0.18%), the corrosion activity was significantly enhanced. The very minimal corrosion activity evidenced by lower corrosion current density measured on TABLE 4 Concentrations of Water Soluble Chloride and Sulfate Ions Solution Composition
Chloride (%) 3.64 0.18 3.58 3.56
15.7% C12.1% SO4"" 15.7% CI + 0.55 SO4"" 15.7% (21" + 2.1% SO4""
Sulfate (%) 0.05 0.12 0.18 0.30
CHL
90 >- 8.0 I'-- . Z
'-" 7.0 Z eY
~ 6.0 Z
~ 5.0 o¢"
0.0
I
I
I
0.1 0.2 0.3 WATER SOLUBLE SULFATE (% WT. CEMENT)
FIGURE 3 Influence of Sulfate Ions Associated with 3.6% Water Soluble Chlorides on Reinforcement Corrosion.
146
O.S.B. AI-Amoudi and M. Maslehuddin
Vol. 23, No. 1
reinforcing steel in specimens placed in pure sulfate solution indicates that, within the duration of this study, sulfate ions do not influence the corrosion process. The concomitant presence of chloride and sulfate ions is necessary for the initiation and propagation of reinforcement corrosion in cement concrete. The quantum of C3 A in plain cements andthe type of the blending material in blended cements may also influence reinforcement corrosion in concrete. As such, more detailed studies are needed to elucidate reinforcement corrosion mechanisms in plain and blended cements placed in chloride and chloride-sulfate solutions. CONCLUSIONS The influence of sulfate, chloride and sulfate-chloride solutions on reinforcement corrosion in cement paste specimens was investigated. It was observed that the corrosion activity measured as time to initiation of corrosion and corrosion current density was minimum in specimens placed in the pure sulfate solution. The time to initiation of corrosion was found to be approximately 20 days in specimens placed in 15.7% c r solution as compared to 56 days in specimens placed in chloride-sulfate solutions. The corrosion current density was slightly higher in 15.7% c r + 0.55% S04- solution compared to 15.7% Cl- solution. The corrosion current density on bars in specimens placed in 15.7% c r + 2.1% s04-- solution was more than double the corrosion current density on bars in specimens placed in 15.7% c r solution. The data developed in this study indicate that, within the duration of this investigation, sulfate ion alone does not influence the reinforcement corrosion process. However, the concomitant presence of sulfate and chloride may significantly affect the rate of corrosion. More detailed studies axe required to elucidate the reinforcement corrosion mechanisms involving cements of varying C3A, blended cements and differing exposure conditions. ACKNOWLEDGEMENT The support of King Fahd University of Petroleum and Minerals in the conduct of this research is gratefully acknowleged. REFERENCES .
.
3. 4. 5.
.
7. 8. 9. 10.
Holden, W.R., Page, C.L., and Short, N.R., Corrosion of Reinforcement in Concrete Construction, Alan P. Crane Ed., Society of Chemical Industry, London, 1983, pp. 143149. AI-Tayyib, A.J., et al. Cement and Concrete Research, 18 (5), 1988, pp. 774-782. Cheng, T., Lee, J., and Tsai, W., Cement and Concrete Research, Vol. 20 (2), 1990, pp. 243-252. Smith, C.L., U.S. Department of the Interior, Geologic Survey, Saudi Arabia Mission, TRI, 1980, p. 26. Johnson, H., Kamal, M.R., Pierson, G.O., and Ramsay, J.B., Quaternary Period in Saudi Arabia, AI-Sayyari, S.S., and Zotl, J.G. (Eds), Springler-Verlag, Austria, 1978, pp. 84-93. Stem, M., and Geary, A.L., Journal, Electrochemical Society, No. 104, 1957, p. 56. Gonzalez, J.A., Algaba, S., and Andrade, C., British Corrosion Journal, Vol. 15, No. 3, 1980, pp. 135-139. Gonzalez, J.A., and Andrade, C., British Corrosion Journal, Vol. 17, No. 1, 1982, pp. 21-28. APHA-AWWA-WPCF, Standard Methodsfor the Examination of Water and Wastewater, 15th Edition, pp. 439-440. Evans, U.R., The Corrosion and Oxidation of Metals, First Supplementary Volume, 1968, Edward Arnold, London.