AN INVESTIGATION INTO NaOH ATTACK ON CEMENTITIOUS MATERIALS

H. Trinh Cao, L. Bucea and V. Sirivivatnanon

CSIRO – Building, Construction and Engineering

PO Box 310 North Ryde NSW 1670

ABSTRACT

This paper presents the findings obtained from an investigation into the degradation mechanisms of cementitious materials exposed to concentrated NaOH solutions. By monitoring the length change of mortar prisms, it was found that dimension instability can result from exposure to concentrated NaOH solutions. Considerable expansion was observed with portland cement mortars. This appears to be the result of precipitation of coarse calcium hydroxide and Na-rich calcium silicate gel in voids and particularly at aggregate/paste interface. The NaOH attack on blended cements was found to be different to that on portland cement. The implications of the findings in terms of designing concrete for better resistance to NaOH attack are given.

KEY WORDS:

NaOH attack, dimensional instability, blended cements

An investigation into NaOH attack

ON CEMENTITIOUS MATERIALS

H. Trinh Cao, L. Bucea and V. Sirivivatnanon

1.0 Introduction

Hardened cement hydrates are alkaline materials. Alkaline attack (from external source) on concrete is not normally encountered. However, it has been known that high concentration of alkaline materials that come in contact with concrete in industrial processes can cause deterioration (Lea, 1970; Mindess and Young, 1981; Taylor, 1990; Biczok, 1972). In Australia, premature failure of concrete slabs at Bayer process refineries has been identified as a concern for the alumina industry in a recent Concrete Slab Workshop held in Booragoon Western Australia. The major issues of slab failure were noted as potential environmental damage, safety and additional ongoing cost involved in operation, clean up, and maintenance.

Based on existing literature, the following observations can be made with regard to attack on cementitious materials by alkali hydroxides:

  • The extent of attack depends on type of alkali hydroxide, concentration and temperature;
  • The extent of attack depends on composition of the cement type;
  • The attack is probably caused by decomposing hydrated aluminate phases;
  • Surface deterioration and loss of strength are the main modes of deterioration
  • Dense materials corrode at a slower rate.

Based on these observations, it appears that by specifying a cement with low alumina content (particularly with respect to tricalcium aluminate or C3A content) and sufficient curing, the concrete can be made more resistant to attack by alkali hydroxides from external sources.

In this paper, an investigation into the attack of alkali hydroxides on cementitious materials is presented. The objective of this work is to examine the above observation and to obtain information for combating against premature deterioration of concrete slabs in alumina refinery.

2.0 Materials and Experimental works

The compositions of cementitious materials used in this investigation are shown in Table 1. These are commercially available materials.

Table 1

Compositions of portland cement and supplementary cementitious materials

Oxide (%)

Portland Cement

Blast furnace slag

Fly ash

Silica Fume

SiO2

21.3

33.7

63.1

98.5

CaO

64.4

40.8

2.5

0.2

Al2O3

4.9

13.9

24.7

0.2

Fe2O3

2.5

0.6

3.5

0.1

MgO

1.5

6.0

1.0

0.5

TiO2

0.2

0.8

0.9

0.1

K2O

0.9

0.43

2.66

0.2

Na2O

-

0.15

0.8

0.1

SO3

2.6

3.9

0.38

0.1

The sand used in making of mortars was Graded Sand as specified by ASTM C 778 – 91.

The attack of NaOH on cementitious materials was monitored by determination of length change of mortar bars. Unless stated otherwise, the mortar composition, mixing procedures, curing procedures, sample size and length change measurement were carried out in accordance to ASTM C 1012- 95a (Details of the mortar mixes are given in Table 2). This means that the behaviours of mortars were compared based on similar flow and initial compressive strength (~20MPa) prior to immersion in different NaOH solutions.

Table 2

Details of mortar mixes

Mortar Composition

Water to binder ratio

Curing Conditions

(in lime water)

Observations

Cement – General Purpose

0.485

1 day at 350C, 1 day normal conditions

Mortar with a flow of 105 +5%

Cement - General Purpose

0.4

1 day at 350C, 1 day normal conditions

-

Cement - General Purpose

0.55

1 day at 350C, 1 day normal conditions

-

Cement +20% fly ash

0.45

1 day at 350C, 3 days normal conditions

Mortar with a flow of 105 +5%; Cured to achieve same strength as cement mortar with equal flow

Cement +40% fly ash

0.435

1 day at 350C, 6 days normal conditions

As above

Cement +40% slag

0.475

1 day at 350C, 6 days normal conditions

As above

Cement +60% slag

0.48

1 day at 350C, 9 days normal conditions

As above

Cement +80% slag

0.475

1 day at 350C, 12 days normal conditions

As above

Cement +5% silica fume

0.504

1 day at 350C, 1 day normal conditions

As above

Cement +10% silica fume

0.528

1 day at 350C, 1 day normal conditions

As above

This work consisted of the following items:

  • Investigation on the effect of NaOH concentrations ranging from 1M, 2M, 5M and 8M NaOH solutions. Portland cement mortars with W/C = 0.485 were used.
  • Investigation on the effect of temperature. 1M and 2M NaOH solutions kept at 23oC and 50oC were used. Portland cement mortars with W/C = 0.485 were used.
  • Investigation on the effect of W/C. Portland cement mortars with W/C of 0.4, 0.485 and 0.55 were used in 8M NaOH.
  • Investigation on the influence of blended cements. Blended cement mortars (Table 2) were used in 8M NaOH.
  • Microstructure examination of mortars was carried out using Scanning Electron Microscopy.

The length change results reported in this paper are the average results of three samples.

 

3.0 Results and discussion

3.1 Effect of NaOH Concentration on Length Change of Portland Cement Mortar

The effect of NaOH concentration on expansion of mortar bars was investigated by using mortars made with portland cement (PC). Figure 1 shows the expansion of PC mortar bars at 23oC.

Figure 1

Effect of NaOH concentration on length change of portland cement mortar bars

From Figure 1, it can be clearly seen that exposure of mortar to NaOH solution can lead to expansion. The influence of NaOH concentration on mortar expansion is very conclusive. At a concentration of 2M NaOH and lower, the expansions of PC mortar bar was small and relatively comparable to that immersed in water. This indicates that the attack by NaOH on portland cement mortar is either negligible or do not result in expansion at 2M NaOH or less at normal temperature. At concentration of 5M NaOH and higher, considerable expansion was observed after about 4 weeks of immersion and continued to increase with time. The magnitude of mortar expansion is clearly increased with increasing NaOH concentration.

In conjunction with the reported loss of strength (Lea, 1970; Biczok, 1972) due to exposure to high concentration of NaOH, the data presented in Figure 1 suggests that expansion induced cracking of concrete could be an important mode of deterioration even when the concrete is kept saturated.

3.2 Effect of Temperature on Length Change of Portland Cement Mortar

Figure 2 shows the influence of temperature on expansion of portland cement mortars in 1M and 2M NaOH solution. This figure indicates that expansion of portland cement mortar was more than doubled in 2M NaOH solution at 50oC in comparison to that shown at 23oC. Early results of mortar expansion in 5 M and 8 M NaOH solution indicate that this trend is consistent. However, in 1M NaOH solution, the influence of temperature on mortar bar expansion is not apparent. This suggests that the effect of temperature on NaOH attack is considerable only at concentration higher than 2M.

By monitoring the length change of portland cement mortar bars immersed in NaOH solution, the following observation can be made:

  • Exposure to NaOH solution at concentration higher than 5M can result in considerable expansion;
  • The expansion is magnified at elevated temperature.

The implication of these observations is that expansion of portland cement in concentrated NaOH solution could be responsible for the cracking of concrete slabs in alumina refineries especially those observed at early age.

Figure 2

Influence of temperature on length change of portland cement mortar

3.3 Influence of Water-to-cement ratio (W/C) on Length Change of Portland Cement Mortars

The influence of water-to-cement (W/C) ratio on length change of portland cement mortar is shown in Figure 3. The mortars were prepared with W/C of 0.4, 0.485 and 0.55. The curing regime for all mortars was kept similar to that of ASTM C1012 procedures for standard portland cement mortar (W/C = 0.485).

There are two unexpected features observable from Figure 3. The first feature is that for a same curing regime, higher expansion was observed with mortar made with lower W/C at 8 M NaOH. The higher expansion of lower W/C mortar was observed at all ages. The second feature is the "shrinkage" shown by high W/C mortar during the early stage of immersion in 8 M NaOH solution. Expansion was observed afterwards.

The high expansion of low W/C mortar in 8M NaOH solution suggest that the attack of high NaOH concentration on portland cement involves the precipitation of reacted products in capillary pore system. It is known that fine capillary pore system and less total porosity are the result of lowering W/C for a given curing regime (Mindess and Young, 1981). The pressure exerted by precipitation of reacted products is more pronounced in fine and limited pore system. At higher W/C, the pressure and hence expansion would be less due to higher availability of porosity.

It must be noted that as the W/C increases, the "permeability" of the mortar will also increases in terms of penetration of Na+ and OH-. As noted previously, high NaOH solution lead to the decomposition of hydrated aluminate phases to soluble compound such as NaAl(OH)4. It is likely that the initial "shrinkage" observed for high W/C mortar is caused by aluminate decomposition. The decomposition of aluminates will be facilitated by the readily available NaOH in high W/C mortar.

Figure 3

Effect of W/C on length change of portland cement mortar

3.4 Influence of Blended Cements on Length Change of Mortar Exposed to 8M NaOH Solution

Blended cements containing fly ash (FA), ground granulated blast furnace slag (BFS) and silica fume (SF) have been used in concrete for a number of years. Their popularity have been attributed to the high durability of blended cement concrete, particularly in regard to chemical attack and resistance to ingress of external elements.

Their influences on length change of mortar in 8M NaOH solution are shown in Figures 4, 5 and 6 for blended cements containing SF, FA and BFS respectively. Figure 4 shows that silica fume blended cements in the range of 5 to 10% replacement can effectively reduce the expansion in mortar in comparison to that shown by portland cement. There was an observable shrinkage shown by the 5% SF blended cement mortar at early age.

In the cases of blast furnace slag and fly ash blended cement mortars, "shrinkage" was observed in 8M NaOH. The magnitude of shrinkage was clearly related to the percentage of cement replacement by either slag or fly ash. This was not expected.

It should be noted that silica fume, fly ash and slag are amorphous materials. Hence they can be "dissolved" in high alkaline environment. The "reactivity’s" of these materials in blended cement rely mostly on this characteristic. This characteristic is used in making alkali-activated slag cement. It has been known that in NaOH-slag mixture, the initial phases formed are rich in silica-hydrosilicates and silica acid which polymerised into silica gel (Gluchovsky et al, 1983; Iler, 1979; Kutti, 1992). The silica gel has been described as microamorphous, 3-D network of colloidal silica (Lea, 1970). The formation of this type of gel (the process is known as syneresis) can cause shrinkage. This is the likely explanation for the observed shrinkage shown by slag blended cement mortars in 8M NaOH. It would probably applicable for fly ash blended cements and silica fume blended cements.

Figure 4

Influence of silica fume on length change of mortar bars

Figure 5

Influence of fly ash on length change of mortar bars

Figure 6

Influence of ground granulated BFS on length change of mortar bars

The results and discussion given in this section suggests that the attack of concentrated NaOH solution on blended cements is different to that on portland cement. At early age of immersion, syneresis shrinkage is more prevalent for blended cements with high replacement percentages. At later ages, formation of products causing expansion is more prevalent to the NaOH attack. In terms of dimensional stability in the presence of concentrated NaOH, it appears that silica fume blended cement would provide the required performance.

3.5 SEM Examination of Portland Cement Mortar in NaOH Solution

In this section, the microstructures of mortars immersed NaOH solutions are examined by Scanning Electron Microscopy (SEM). The objective of this work was to determine the cause of expansion. Figures 7 and 8 show typical morphologies of NaOH attack on hardened cement mortars.

It can be seen from these figures that the NaOH attack results in precipitation of large calcium hydroxide plates and glassy looking gel in voids and at aggregate-paste interface. These types of microstructures were associated with samples showing large expansion. The glassy looking gel was identified by EDS as calcium silicate gel containing a high amount of sodium. The precipitation of large calcium hydroxide can be explained by the effect of alkali on lowering the concentration of CaO in solution. The formation of Na-rich calcium silicate gel is not easily explained since alkali raises the concentration of SiO2 in solution. It is suggested that these gel’s could be the result of decomposition C-S-H gel and subsequent precipitation in NaOH solution. Further work is needed to clarify this aspect. At this stage, it appears that the precipitation of these materials in the voids and at aggregate-paste interface is the cause of observed expansion.

Figure 7

Precipitation and cracking at aggregate/paste interface

Figure 8

Large calcium hydroxide plates and Na-rich calcium silicate gel

4.0 Conclusions

Based on the results obtained in this investigation, the following conclusion can be made with respect to NaOH attack on hardened cement mortars:

  • Exposure to NaOH solution results in dimensional instability in hardened cement mortars.
  • Considerable expansion was observed with portland cement, while shrinkage was observed with blended cement containing fly ash or slag. A small expansion was observed with silica fume blended cement.
  • Lowering water-to-cement ratio did not lead to better dimensional stability.
  • High concentration and high temperature magnified the extent of NaOH attack.
  • The mechanisms of attack of NaOH are different for portland cement and blended cements. For portland cement, the expansion could be caused by the precipitation of calcium hydroxide and Na-rich calcium silicate gel in voids and at aggregate-paste interface. The attack on blended cements can result in formation of silica gel.
  1. Practical Implications

This investigation has shown that dimensional instability of hardened cement will result from exposure to hot concentrated NaOH solution. This can cause cracking of the concrete. The data suggest that it is possible that the cracking can occur at an early age. This work deals with cementitious materials immersed in NaOH solution. In practice, the exposure to NaOH spillage is a wet-and-dry situation. In such a case, the effect of dimensional instability caused by NaOH will be magnified. Hence cracking and crazing will be more prevalent. This is due to the added effect of recrystallisation and carbonation occurring during the drying stage. The penetration of NaOH into concrete will be assisted by absorption.

In designing concrete for high resistance to NaOH attack, the dimensional stability of the concrete in NaOH should be considered seriously. It appears that increasing concrete grade, ie. lowering W/C, does not always result in higher resistance to NaOH attack. It is known that permeability can be reduced with low W/C and sufficient curing. In practice, "sufficient curing", eg. 7 days, is not always a feasible option. In such a situation, selecting binder type and good concrete mix design are essential.

From the data obtained in this work, it appears that silica fume blended cement could provide improved resistance to NaOH attack. It should be noted that since properties of blended cements can vary widely with cement type and source of mineral additive, specific performance testing is recommended for the selection of concrete ingredients (Moorehead and Cao, 1997). In addition to the appropriate selection of concrete ingredients, the use of fibre reinforcement can provide further protection against cracking of concrete.

References

Biczok, I. (1972), Concrete corrosion Concrete Protection, Akademiai Kiado, Budapest.

Gluchovsky, V., Zaitsev, Y. and Pakhomov, V. (1983), "Slag alkaline cements and concretes", Silicates Industriel, 1983-10.

Iler, R.K. (1979), The chemistry of silica, John Wiley & Sons.

Kutti, T. (1992), "Hydration products of alkali activated slag", 9th Int. Congress on the Chemistry of Cement, New Delhi, India, 1992, Vol IV, 468-474.

Lea, F.M. (1970), The chemistry of cement and concrete, 3rd Edition, Chemical Publishing Co, NY.

Mindess, S. and Young, J.F. (1981), Concrete, Prentice-Hall Inc. Englewood Cliffs, NJ.

Moorehead, D. and Cao, H.T. (1997), "Concrete for aggressive environments – A specifier’s guide", Concrete 97 Proc., Adelaide, 1997.

Taylor, H.F.W. (1990), Cement Chemistry, Academic Press.