Paper No. 671 EFFECT OF TEMPERATURE AND PH ON THE BINDING OF CHLORIDE BY CALCIUM ALUMINATE 320 Benjamin S.E. Benjamin Pakistan Engineering Congress, 70th Annual Session Proceedings 321 322 Benjamin EFFECT OF TEMPERATURE AND pH ON THE BINDING OF CHLORIDE BY CALCIUM ALUMINATE S.E. Benjamin Department of Chemistry, Forman Christian College-University, Lahore ABSTRACT The risk of pitting corrosion attack of the embedded steel reinforcement is determined by the amount of the free chloride present in the cement pore environment. Some of the chloride anion is bound as the Friedel salt i.e., 3CaO.Al2O3. CaCl 2. 10H 2O or it is adsorbed in the cement matrix by CSH gel or other cement phases and is thus rendered harmless. Even the effectively bound chloride may be conditionally released due to a change in the equilibrium parameters and may cause serious damage to the exposed steel. The present investigations focus upon studies conducted on synthesized calcium chloroaluminate (SCCA) as the function of temperature and pH of the solution. The effect of temperature upon the chloride release was also monitored upon small mortar balls. It is observed that the pH of the medium is more important i.e., more chloride is released from the SCCA in at pH 14 solution as compared to the SCCA at a solution pH of 10. However, the SCCA is stable between the temperature range of 20 ºC till 60 ºC. In case of the real mortar balls physical adsorption of the chloride ion is more important and it shows temperature dependence across the temperature range of 20 ºC – 60 ºC. It is deduced that these parameters play a decisive role in the overall solution chemistry regarding bound/free chloride ratio in the cement pore environment. The effect of pH is also reflected for OPC, OPC/fly ash and OPC/slag cement blended mortars of the composition 1:1 and 4:1 cement/sand ratio. Polarisation resistance measurements have shown that higher the cement content higher is the protection offered to the embedded steel during the hydration period. Pakistan Engineering Congress, 70th Annual Session Proceedings 323 INTRODUCTION The chloride anion (Cl −) may be bound in cement in two ways: a) by a chemical bond through reaction with the hy dration products or b) by the physical adsorption on to the surface of the hydration products. However, when the Cl− levels are very high a part may be bound and thus harmless1-2, and a portion of Cl− may be free to cause damage to the embedded steel i.e., the ratio of the free Cl −/bound Cl− is an important controlling factor in determining whether the steel will pit or not. The C 3A i.e., tricalcium aluminate (3CaO.Al 2O3) and tetra alumino ferrite (C4AF) in cement binds Cl − anion to form an insoluble product, called the Friedel salt, with the composition 3CaO.Al 2O3/Fe2O3. CaCl2. 10H2O and reduces the levels of free Cl− in the pore solution3-4. Rasheeduzzaffar and coworkers 5 have shown that the concentration of water soluble chloride, extracted with hot water, falls from 86 % for 2% C 3A content in OPC mortars to 33 % for OPC mortars containing 14 % C3A. A higher chloride complex of the composition, 3CaO.Al 2O3. 3CaCl 2. 30H 2O6, has also been also reported which is formed with a 21-23 % CaCl 2 solution at −10 °C: a condition which is not normally enco untered in hydrating cements. Thus, the corrosion resistance of steel in cements can be elucidated on the basis of the C 3A content of the cement 7. e.g., the poor performance of sulphate resisting cement is generally attributed to low C3A content i.e., low binding capacity resulting in higher corrosion rates8. Several researchers have also reported the physical adsorption of the Cl − anion: Ramchandran suggested that a high proportion of Cl− ions form an interlayer chemisorbed complex within C-S-H gel 9. Midgley and Illston 10 have reported a reduction in the pore size of the hardened cement pastes in presence of Cl − ion, by the mercury intrusion porosimetry technique. Lambert 11, however, did not find any detectable capacity to bind Cl − in C 3S (Jeffery’s alite) from analysis of the pore solution expressed from C3S hardened pastes. Benjamin and Sykes observed that a steel mortar specimen containing 3 % NaCl (by weight of OPC) showed pitting at 40 °C12. The same pitted specimen did not show pitting at 0 °C. This raised an important question that whether the stability of Friedel salt was temperature dependent. The present investigations specifically deal with two important parameters i.e., the stability of synthesised calcium ch loroaluminate (SCCA) as the function of temperature and the pH to understand the role this component of cement in binding chloride. The studies are further extend ed to simple experiments with mortar spheres to correlate the binding mech anism operative in the real cement environment. The studies have been extended to the performance of blended cements i.e., ordinary Portland Cement (OPC) mixed with blast furnace slag (BFS) cement and fly ash (FA). Earlier works on blended cements have been conducted on hardened cement pastes13-14. However, the chemistry changes in the hydrating cement with 324 Benjamin inherently present Cl – ion may start to effect the passivation of steel embedded in concrete at a very early stage. The resu lts are amalgamated to reach a reasonable conclusion regarding the role of C3A in cements. EXPERIMENTAL Calcium chloroaluminate was synthesised by the titration of NaOH solution against a mixed solution of aluminum chloride and calcium chloride in the correct stoichiometric ratio, till pH 11.99 was achi eved. The reaction can be presented as follows: 5CaCl2 + 2AlCl3 + 14NaOH → 3CaO.Al2O3.CaCl2.10H2O + 14NaCl + Ca(OH)2 [1] The detailed procedure is described elsewhere 15. The titration was performed in deaerated atmosphere. The precipitates were filtered in deaerated conditions and dried to constant weight in a desiccator over fused CaCl 2. The SCCA precipitates were identified by XRD on Phillips PW 1050 and compared against standard tables16. The chloride content of SCCA was also determined by chemical analysis by Mohr’s method17. Finally, the SCCA was analyzed by electron probe microanalyses (EPMA) on Cameca Cambex, to confirm the calcium/aluminum and calcium/chloride ratio in the compound. An accurately weighed amount of dr ied SCCA was added to solutions of pH- 10, pH-12.6 and pH-14. The compound was allowed to equilibrate in the solutions for two weeks before analyzing for the chloride content by the spectrophotometric technique on a Perkin Elmer 552. This standard method involves addition of 0.25 M ammonium iron (III) sulphate in 9 M nitr ic acid, followed by the addition of a saturated solution of mercury(II) thiocyanate in alcohol. The absorbency was measured at 460 nm. The same samples were held at different temperatures i.e., 22°, 40°, 50° and 60 °C for two hours and the Cl – released was checked. After tests the suspensions were cooled for one day an d fresh aliquots were again checked for chloride levels to observe whether the reaction was reversible or not. Lastly, XRD patterns of SCCA in the test solutions were taken again to confirm the compound. To monitor the chloride binding mechanism in mortars, small spherical mortar specimens were cast in 1 cm diameter, with 1:1 cement/sand ration and a water/cement ratio of 0.5. NaCl was added to the mix water to ensure even distribution of salt in the cement matrix. The mortar slurry was shaped into balls in plastic molds. The samples were de-molded after 24 hours, sealed inside polythene bags and cured for 40 days at 25 °C. Six mortar spheres of each mix were immersed in 50 ml of distilled water which was just enough to cover the sa mples. Mortar spheres were allowed to equilibrate with the surrounding solution for two weeks at room temperature i.e., 22 ± 2 °C, sealed inside polythene bottle. Periodically, chloride determination tests of the immersion water were carried out by the same method as for SCCA. After 14 days tests were started when the Cl – concentration changed very little especially for 1 % NaCl mortar spheres. Tests were conducted at 22 °, 40 ° and 60 °C. The solutions with mortar samples were cooled and kept at th e specific temperature for Pakistan Engineering Congress, 70th Annual Session Proceedings 325 ten days and again checked for the chloride levels to understand the binding mechanism in mortars. Studies on blended cements were performed on commercial grade Pakistani steel. Its nominal composition was determined through SEM analysis on Philips XL 30, Holland and given in table I. Table-I Element Fe Si P S Ti Mn % composition 98.79 0.27 0.04 0.03 0.10 0.62 Nominal composition of the steel Samples of steel having dimensions 1cm x 1cm were prepared. A steel wire was wrapped around the steel specimen an d these were cold mounted leaving an exposed area of 78.5 mm 2. The specimen were lightly mechanically polished and washed and dried with alcohol. The mounted samples were mechanically embedded in mortars prepared from OPC, OPC/BFS, OPC/FA mixtures. The OPC/BFS and OPC/FA ratios were kept at 1 : 0.54. The composition of OPC, BFS and FA is given in table II. Table II Type of Cement Chemical composition % OPC BFS FA L.O.I SiO2 Al2O3 Fe2O3 CaO MgO SO3 Na2O K2O I.R LSF 1.05 21.12 6.08 3.78 63.86 1.15 2.37 0.97 0.35 0.14 0.90 1.1 24.85 8.08 3.12 57.6 2.5 2.33 – – 0.30 0.69 3.5 59.7 4.94 26.71 1.63 0.5 traces 0.28 0.80 0.28 0.89 Chemical composition of OPC, slag cement and Fly ash Ordinary silica sand after washing an d drying was used for the preparation of mortar specimen. Two mix proportions were studied i.e., OPC/blending cement, in the ratio of 1:1 and 1:3. The behavior was monitored against plain OPC cement mortars. The water/cement ratio was kept at 0.5 for both the mix proportions. The specimens were embedded in the mortar in a way that the exposed face was held in a vertical fashion having a cover of 0.75 inches on all sides. 326 Benjamin The mortars were all owed to se t for 2 days before these were remolded and placed in the immersion solutio n. The me asurements were started afte r 2 days o f immersion in water containing 3 % sodium chloride (NaCl) solution . Linear potentiostatic polarization resistance (LPR) measurements were continuously conducted for 30 days on a home made lo w noise potentiostat. The changes in the potential were reco rded against stand ard Calomel reference elect rode and a platinum wire was used as a counter elec trode. A potential of ±10 mV was applie d around the free corrosion potential (E corr), and the corresponding changes in the current were noted on a voltmeter in series with the potentiostat, after a time span of 60 seconds. This method was adopted for quick assessment and nondestructiv e effects. RESULTS AND DISCUSSION The XRD pattern for SCCA is presented in Fig. 1. The major peak appearing at 2d of 7.8 identified the compound as β-calcium chloroaluminate. Fig. 1: X-Ray diffraction pattern of the synthesized calcium chloroaluminate. Also, the ratios of the intensities of the characteristic peaks were similar. The chemical analysis sh owed that 0.148 gm Cl −/gm of the sample. Theoretically Cl − = 0.123g for the Friedel salt. The EPMA re sults showed that the SCCA, o n average, contained Al=0.42, Cl=0.53 and Ca=1.04. The ratio is satisfied in terms o f Ca/Cl=1.96 but the Al/Cl=1.26, and Ca/Al=2. 3 was higher than the Friedel salt. T he data as weight percent however, confirms that this compound is β-calcium chloroaluminate in the anhyd rous form. Co mplexes with different composition are expected to form in a situation when OH – ions are compe ting with Cl − ions to combine with the aluminate species 18. Lea 6 has reported that the phase i.e., β- calcium chloroaluminate is only stable at temperatures above 28 °C. Therefore, it is expected that this phase ex ists in hot climates. Kuzel 16 found, through precise optical measurements that the rhombohedral β-phase crystals that are high symmetry and uniaxial, can change to monoclinic α-phase, double axial and lower symmetry upon cooling through twinning. α-3CaO.Al2O3.CaCl2.10H2O = β-3CaO. Al2O3.CaCl2.10H2O [2] This change that occurs at a n equilibrium temperature of 28 ±2 °C was reversible and spontaneous without changing the crystal morphology and the water content. Pöllman and Kuzel 19 have also reported that all monoclinic phases in the Pakistan Engineering Congress, 70th Annual Session Proceedings 327 system can be conve rted into rhombohedral by heating or light rubbing. Thus, any of these factors can change α-phase into β-phase and vice versa. Test in SCCA : The results obtained for SCCA i n differen t pH solutions, presented in Fig. 2, show that as th e pH of the medium is increased the concentration of Cl- released into the solution is also increased. 28 26 24 22 20 18 10 20 30 40 50 60 70 pH 14 Temperature – C chloride released – m g pH 12.6 pH 10 Fig. 2: Effect of temperature on the stability of the synthesized calcium chloroaluminate in different pH solutions. It seems that the SCCA for mation took place according to the reaction described in Fig. 3. So reasonably it may be assumed that as the OH− concentration of the me dium increases, the complex would take up OH – ion and Cl − would be released: Ca 4[Al(OH)6Cl]24-.4H2O + xH2O [ 3] ↓↑ Ca 4[Al(OH)6+xCl1-x]24-.4H2O + xCl- [ 4] Though pH has its influence showing that OH – ion takes the place of C l− in this complex, the interchange cannot be explained in terms of simple thermodynamics. It is obvio us that some Cl − will be released in aqueous environments, from systems like the one under study unless sufficient Cl− is already present20. Page 21 suggested that poor performance of silica fume-OPC blend (30%) was due to solubility of Friedel salt at low p H-11.6. Our studies show contrary results i.e., SCCA is stable at this pH and higher alkalinity does not appreciably favors Cl – binding by this compound. Al-Amoudi 22 and co-workers have supported the use of silica fume-OPC blend as compared to plai n OPC as it sho wed the least corrosio n rates for steel reinfo rcement in sulphate environments in presence of chloride ions. It is plausible that Page only presented a speculation and did not actually investigated the Friedel salt’s stability. The amount of Cl – that stayed bound, calculated from Fig. 2, at pH 10 = 0.72, pH12.6 = 0.6 and pH 14 = 0.5, wh ich indicates the stability of mo no chloride complex. 328 Benjamin It is expected that the hydrate f ormation i.e., C4AH13 is favored if the Cl- c ontent of the compound falls to 0.323. Temperature affected the Cl – release from 20 °C to 40 °C, but after this temperature, the Cl- ion concentration in the solution did increased markedly. This shows that Cl- release takes place when the equilibrium conditions are disturbed. Low Cl- release from 40 °C to 60 °C indicates that some stability of the compound is maintained. The Cl – concentration decreased upon cooling in all the p H solutions, falling even below the starting equilibrium values suggesting that Cl – binding is favored at low temperatures and a reversible process is o perating. Roberts 24 observed similar release of Cl- with temperature in solutions of NaOH and Ca(OH)2. He however, described it in terms of an ir reversible process due to the formation of stable C4AH13 complex. We observed a reversible behavior i.e. , cooling leads to the re-absorption of Cl- ion. Results do not seem to fit a simple thermodynamic treatme nt of ion exchange either, but this will be influe nced by the thermodynamics of the solid phase. Re-absorption can, ho wever, be explained in terms of the structural consideration of the complex. Butler and Taylor 25, established the structure of C4AH13 as parallel layers stacked in a complex manner with six OH – groups in octahedral layer and seventh OH- group outside it where it is closely associated with water molecules. Complete replacement of OH- groups outside the octahedral layers occurs in the synthetic compound as CaO.Al2O3.CaX2 aq. where X is an anion. Thus, water and extra interchangeable ions can sit at interfaces between layers and can interchange reversibly with species fr om solution whenever, equilibrium is disturbed. This was confirmed by the X RD patterns after the tests whi ch showed that the compound was still mono chloride complex as shown in Fig. 3. pH = 10 pH = 14 pH = 12.6 Fig. 3: X-Ray diffraction patterns of the synthesized calcium chloroaluminate after tests in solutions of pH 10, 12.6 and 14 at different temperatures. U Tests in 1% NaCl OPC mortar sphe res: The chloride co ncentration in solution increased slightly with an incr ease in temperature from 22-40 °C a nd Pakistan Engineering Congress, 70th Annual Session Proceedings 329 markedly at 60 °C, Fig. 4. Upon cooling the temperature fells to level below the start level . In these experime nts, it is assumed that the outer layer of mortar spheres achieved equilibrium with the solution. Only the outside layer released Cl , more can come into the solution but slowly with time from the interior. Raising the temperature also accelerated the transpo rt outward from the i nner core but the Cl determined was mainly from the outside layer as the steady state conditions will take much longer to achieve. – – 40 35 30 25 20 18 10 20 30 40 50 60 70 Temperature – C chloride released – m g 1:1 4:1 Fig. 4: Concentration of free Cl− in solution from the mortar spheres containing 1% NaCl (by wt of OPC) as the function of temperature having aggregate/cement ratio 1:1 and 4:1. It seems at first that the increase in the Cl− concentration is due to the higher diffusion rate of Cl − into the surrounding water. Page and coworkers 25 observed 9 times increase in th e value of effective diffu sivities of chloride ion into hardened cement pastes for temperature change from 7° − 44 °C. If solely the diffusion of free chloride into the surrounding water was responsib le then even upon lowering the temperature the Cl − concentration in t he solution should remain constant or continue to increase till equilibrium is established. This shows that some absorption into the mortar spheres has taken place, and that there is reversi ble chloride release wi th a strong effect of temperature. The results are similar to the work reported by Luping and Nelson20 with their work on OPC pastes and mortars. They showed th at the relationship between the bound Cl − and the free Cl − can be well described by Cl− binding isotherm which obeys Fr eundlich equation at high free Cl − concentrations of >0.01 mol/l, and Langmu ir equation at low concentrations. T he absorption is mainly the function of the amount of CSH gel regardless of the W/C and the A/C content. They also emphasized that so me percentage of Cl − is always retained as C 3A-complex. The behavior of these mortar spherical balls have als o shown that adsorptio n of Cl – by the mortar matrix is the key operative mechanism involved when the temperature is var ied. The results disagree with Robert’s 24 postulation that the Cl − is released fro m the Friedel salt a t similar temperatures and the process is irreversible. Chatterje27, Ludwig and Darr 28 in separa te studies have shown that Friedel salt with different chloride content are stable and can form at 5 °C, 20 °C and 60 °C. Maslehuddin and co workers29 also observed an increase in th e Cl− ion levels in th e 330 Benjamin pore solution with both OPC and blended cements at higher exposure te mperatures and also when sulphate ions were pre sent. However, the results also showed tha t the OH – levels decreased above 70 °C, which was conside red to be due to the instability of calcium chloroaluminate at this temperature. Chloride compounds formed w ith S CCA w ere s table u p t o 6 0 °C in this study. Differen tial thermogravimetric analysis30 have also shown that the mono chloride peak appears at 350 °C and this again attests to the stability of Friedel salt. It is only the adsorbed chloride in the matrix which is affected by the changes in the equilibrium state. The mortar ball experiments are also consistent wi th potentiodynamic tests for iron e mbedded in mortars of A/C ratio 1:1 and w/c ratio of 0.5 containing 3 % NaCl as an admixture, given in our earlier papers 12. The specimen showed pittin g at 40 °C, but the same specimen when cooled quickly to 0 °C did not pit at all. This also agrees with the postulation that absorption in the matrix plays an i mportant role after stable binding with tr i-calcium aluminate. Parkinson31 also found a lot of temperature dependence of free Cl −/OH- ratio with SRPC c ements, though lower ratio was obtained with OPC. Lambert and Page 11 however, did not find any role of calcium silicates in binding Cl −. This discrepancy could be due the differen ce in the method adopted. These investigations with the synthesized calcium chloroaluminate and simple mortar ball experiments, both attest that, adsorption of chloride in the cement matrix plays a key role in the corrosion of steel reinforcement embedded in concrete. U TESTS ON OP C, OPC/FA and OPC/SLAG BLENDE D CEMENTS The results for mortars with cement/sand ratio 1:1 are presented in Fig. 5-6. Ecorr mVSCE Time – days Fig. 5: Changes in Free Corrosion Potential (ECorr) of the specimen with time embedded in different types of cements (W/C ratio 1:1) Pakistan Engineering Congress, 70th Annual Session Proceedings 331 Time – days Rp K .cm 2 Fig. 6: Changes in Polarisation Resistance (Rp) of the specimen with time embedded in different types of cements (W/C ratio 1:1) The OPC and OPC/PFA mortars showed higher E corr at around −300 mV and the Rp declined from 300 Kohn m.cm2 to 210 Kohm.cm 2 at later stages. T he Rp for the plain OPC mortars stayed around 175 Kohm.cm 2. The OPC/BFS mortars (cement/sand ratio 1:1) failed only after 4 days of immersion in the saline media. The Ecorr values settled to low values of -600 mV in 13 days with the Rp values also declining to 90 Kohm.cm2. Such Rp valu es indicate that still the steel is maintaining some passivity. Plausibly a crevice developed at the steel mortar joint which is expected in case of low oxygen availability. The Ecorr for all types of morta rs with ce ment/sand ratio of 3:1 was low i.e., between -450 to -600 mV with OPC/PFA slightly higher than the OP C plain and OPC/BFS blended cements, Fig. 7-8. Ecorr mVSCE Time – days Fig. 7: Changes in Free Corrosion Potential (ECorr) of the specimen with time embedded in different types of cements (W/C ratio 3:1) 332 Benjamin Time – days Rp K .cm 2 Fig. 8: Changes in Polarisation Resistance (Rp) of the specimen with time embedded in different types of cements (W/C ratio 3:1) It is interesting to note that slowly the Rp values started to recover especially for OPC plain and OPC/BFS . The mai n reason could be the increase in the alkalinity of the pore solution of slag and OPC as the hydration reaction proceeded, see table II. Also, some chloride binding proceeded with time. More alkalinity is required to trigger the hydration reactions in PFA and it is expected that PFA will recover with time. These results show that if the cement content is high, then the cement mortars even if blended, can provide a good protection to the emb edded steel immediately after the setting of cement. CONCLUSIONS 1. The stabil ity of SCCA is affected by the changes in temperature but more markedly by the changes in pH. 2. Synthesized calcium chloroaluminate (SCCA) alwa ys releases some chloride in different alkaline solutions. The highest amount of Cl – ion is released for the synthesized compound at pH- 14 and the least is observed for the pH-10 solution. The amount of Cl – that stayed bound, calculated from Fig. 2, amounts to 0.72 at pH 10, 0.6 at pH 12.6 and 0.5 at pH 14. This Cl – cont6ent slightly increases with an incr ease in temperature and decreases when the solution is cooled. The XRD patterns of th e compound Fig. 3, after the Cl – ion release, in all pH solutions, and subjected to different temperatures, indicates the stability of mono chloride complex. 3. Reversible mechanisms such as exchange of anions OH -↔ Cl- or adsorption is involved. 4. Adsorbed Cl- is most affected by the ch anges in temperature. In mortars, besides C3A, Cl- binding can be associated with adsorption by other cement constituents as CSH gel, which is affe cted by the changes in the equilibrium Pakistan Engineering Congress, 70th Annual Session Proceedings 333 state, e.g., temperature. Fig. 4. The trend is the same for both 1:1 and 4:1 cement/sand mix proportions. The Cl – ion concentration marked increases with an increase in temperature and re-adsorption of Cl – anion takes place when the temperature is lowered down. 5. The blended cements, especially FA/OPC, offer good protection to the embedded steel even during the hydrat ion period, provided the OPC content of the blend is high. REFERENCES 1. Treadaway, K.W.J, “Corrosion of Steel Reinforcement in Concrete Construction”, SCI publishers, 1979, pp1. 2. Berman, H.A., J. of Mater., JMSLA, 7, 1972, p.330. 3. Csizmadia, J., Belazs. G. and Tamas, F.D., Periodica Polytechnica Ser. Civ. Eng., 44, No.2, 2000, p.153. 4. Mohammed, T.U. Cem. and Conc. Res., 33, 33, Issue 9, 2003, p.1487, 5. Rasheeduzzaffar, Saadoun, S.S. Da khill, F.H and Ghatani, “Corrosion of reinforcement in Concrete”, ed. C.L. Page, K.W.J. Treadaway, and P.B. Bamforth, SCI: Elsevier Applied Science, London, ISBN 1-85166-487-4, 1990, p.213. 6. Lea, F.M., “The Chemistry of Ceme nt and Concrete”, 3rd ed., Edward Arnold Ltd., 1970, p.232. 7. Sakr, K.1 Cement & Concrete Res., Issue 9, 2005, p.1820. 8. Holden, W.R., Page, C.L., and Short, N.R., “Corrosion of Steel Reinforcement in Concrete Construction”, ed. A.P. Crane, Ellis Horwood, Chicester, 1983, p.143. 9. Ramchandran, V.S., “Calcium chloride in Concrete”, App. Sci. publishers, London, 1976, p. 57. 10. Midgley, H.G., and Illston, J.M ., Cem. and Conc. Res., 14, 1984, p.546. 11. Lambert, P., Page, C.L., and Shor t N.R., Cem. and Conc. Res., 15, 1985, p.675. 12. Benjamin, S.E., and Sykes, J.M., Proc. of the 3rd Int. Conf. on deterioration and repair of Concrete in Ar abian Gulf, Vol.1, 1989, p.573. 13. Page, C.L., Short, N. R. and El Ta rras, A. Cement and Concrete Research, V.11, No.3, 1981, p. 395. 14. Maslehuddin, M., Shirokoff, J. an d Siddiqui, M.A.B., Advances in Cement Research, V.8, No. 32, 1996, p.167. 15. Benjamin,S.E., D.Phil. Thesis, Oxford University, 1990, p.152. 16. Kuzel, H.J., Neues Jahr b. Min., Monatsh., 1966, p.193. 334 Benjamin 17. Vogel, “Text Book of Quantitative Inorganic analysis”, Longman Group Ltd., 1978. 18. Rosenberg, A.M., J. Am. Con. Inst., Title No. 61-63, 1964, p.1261. 19. P öllman, H., and Kuzel, H.J., Fortschr. Min., 60, 1982, p.165. 20. Luping, T., and Nilsson, L, Cem. and Concrete. Res., 23, 1993, p.247. 21. Page, C.L., and Vennesland, O., Materiaux et Const., 16, 1983, p.19. 22. Al-moudi, OSB., Rasheeduzzaffer, Ma slehuddin, M. and Abduljauwad, SN., Cem. Concrete & Aggregates, 16, Issue 1, 1994. 23. Herbert, P., 8 th I.C. Chem., 3, 1986, p.1. 24. Roberts, M. H., Mag. of Conc. Res., 14, 1962, p.143. 25. Butler, F.G., Dent, S.G., and Taylor, H. F.W., J. of the Am. Cer. Soc., 1959, 42, p.121. 26. Page, C.L., Short, N.R., El-Tarras, A., Cem. & Conc. Res., 11, 1981, p.395. 27. Chatterjee, S., Chem. and Conc. Res., 8, 1978, p.461. 28. Ludwig and Darr, G.M. , and Ludwig, U., Ind. Chem., Belg., 39, 1974, p.687. 29. M. Maslehuddin, C.L. Page, Rasheedu zzafar and A.I. Al-Mana, “Corrosion of reinforcement in concrete construction”, editors: C.L. Page, P.B. Bamforth and J.W. Figg, SCI/RSC Cambridge, Sp ecial Publication 183, 1996, ISBN 0- 85404-731-X, p. 67. 30. Worthington, J.C., Bonner, D.G., an d Nowell, D.V., Mat. Sci. & Tech., 1988, 4, p.305. 31. Parkinson, D. T., 2nd Year Re s. Report, Aston University, 1989. Pakistan Engineering Congress, 70th Annual Session Proceedings 335
Filename:
Paper671.pdf
File Type:
pdf
File Size:
192 KB
