The characteristics of acid migration through epoxy mortars were examined. Diffusion coefficients of typical sewer bio-metabolised acids: sulphuric, nitric, citric and oxalic acids were determined by gravimetric sorption method and fitted to the multi-phase Jacob–Jones model. Acid permeation was characterised by hindered pore diffusion with the extent being determined by the polarity of the acid and epoxy, and by the microstructure of the epoxy. Epoxy with higher polarity was able to reduce the diffusion coefficients by 49, while dense phases of the coating reduced the diffusion coefficient by 5,100. These results reflect the relative influence of epoxy polarity and microstructure on their performance as protective liners in sewers.

INTRODUCTION

Polymeric resins mortars are widely used in mitigating the corrosive effects in wastewater structures because of their good mechanical properties and acid permeation resistance. As coatings, they prolong the life of the wastewater structures, by providing a barrier between the corrosive environment and the concrete substrate. The effectiveness of a lining-based protective system relies on appropriate selection of the coating material that best fits the needs of the environment. Currently, this is challenged by insufficient knowledge in predicting their performance and durability in real sewer environments.

Biogenic attack of concrete sewer pipes occurs as a result of the actions of various organisms. These organisms grow successively to generate acidic metabolites that corrode the concrete substrate (Okabe et al. 2007; Lamberet et al. 2008). Fungi, such as Aspergillus niger, produces various organic acids including citric and oxalic acids (Gu et al. 1997; Lamberet et al. 2008); chemoautotrophic nitrifying bacteria, such as Nitrobacter, converts amine species to nitric acid; and acidophiles, such as Acidithiobacillus thiooxidans, oxidise H2S and sulphur compounds to form sulphuric acid (Sand & Bock 1991). To understand the performance of lining materials in sewer environments, this study examined the migration of biogenic acids through various epoxy mortars. Although numerous studies of water transportation through polymeric liners have been conducted (Gu et al. 1997; Maggana & Pissis 1999; Feng et al. 2004), little has been conducted in the study of acid permeates. In our investigation, acid permeation through epoxy coatings was performed by gravimetric method. The measured mass transport parameters were correlated to the polarity and microstructure of the epoxy and polarity of the acids.

EXPERIMENTAL

Coatings

Two commercially available epoxy mortars were used in this study. They consisted of bisphenol A and a mixture of bisphenol A and F resins cured with amine curing reagents using instructions provided by the manufacturers. The chemical properties of the coatings are summarised in Table 1. The loss of ignition (LOI), which reflects the organic content of the coatings, was determined using a Philips PW2400 XRF with an Rh end-window tube at 1,050°C (Philips, Almelo, The Netherlands). The filler size and concentration were as provided by the manufacturers.

Table 1

Chemical properties of the epoxy mortars

Coating Epoxy type LOI (wt.%) Filler (wt. %) Filler size (Dmean, μm) 
58.7 41.3 198.9 
Mixture of A & F 12.1 87.9 350 
Coating Epoxy type LOI (wt.%) Filler (wt. %) Filler size (Dmean, μm) 
58.7 41.3 198.9 
Mixture of A & F 12.1 87.9 350 

The dimensions of the epoxy specimen are 5 cm × 5 cm × 5 mm.

Fourier transform infrared analysis

Fourier transform infrared (FTIR) spectra of raw and acid immersed coatings were obtained in KBr pellets using a Thermos Nicolet 6700 FTIR spectrometer with attenuated total reflection (Thermo Fisher Scientific, Waltham, MA, USA).

Gravimetric sorption method

Testing of acid permeation was carried out by the gravimetric method (Fiore et al. 2013). Dry square epoxy coupons were immersed in 300 ml reagent grade acids. The list of acids used and their wetting properties are shown in Table 2. The pKa values reflected acid strength while the acid solubility in water reflected relative polarity. All acids are polar, with oxalic acid having the least polarity. Acidic solutions with concentrations of 1, 5, 10 and 20 (g/ml)% were prepared with deionised water and reagent grade sulphuric, nitric, citric and oxalic acids. The immersed coupons were maintained at 25 °C in a temperature controlled chamber. At appropriate time intervals, samples were removed from the acid bath, blotted dry, and weighed on an analytical balance. Epoxy acid uptake with time, Mt, was calculated as follows: 
formula
1
where wt and wd are the weight of the specimen at time t and of the dry specimen, respectively. The samples were re-immersed after weighing. The pH of the acid baths was monitored and adjusted to maintain the solution pH.
Table 2

Acidity and solubility of the acids in water

Permeate Formula pKa (Haynes 2014Solubility in water 
Citric acid C6H8O7 3.14, 4.77, 6.39 147.76 g/L (20°C) 
Oxalic acid C2H2O4 1.23, 4.19 14.3 g/L (20 °C) 
Nitric acid HNO3 −1.4 miscible 
Sulphuric acid H2SO4 −3, 1.99 miscible 
Permeate Formula pKa (Haynes 2014Solubility in water 
Citric acid C6H8O7 3.14, 4.77, 6.39 147.76 g/L (20°C) 
Oxalic acid C2H2O4 1.23, 4.19 14.3 g/L (20 °C) 
Nitric acid HNO3 −1.4 miscible 
Sulphuric acid H2SO4 −3, 1.99 miscible 

RESULTS AND DISCUSSION

FTIR analysis of the epoxy mortars

The FTIR surface functional groups analysis of coatings A and B (Figure 1) are summarised in Table 3. The relative quantities of these functional groups were determined by normalising their peak areas with the area under the Si-O-Si peak. As shown, various polar groups that could form hydrogen bonding, including hydroxyls, carboxylate, amines and sulphide groups, are present in both coatings. It is apparent that the proportion of polar groups in coating A is greater than in B, suggesting its greater polarity. This could also be attributed the higher epoxy content of coating A (see Table 1).

Table 3

Characteristic FTIR bands of coatings A and B

  Coating A Coating B 
Wavenumber (cm−1Vibrations % Normalised relative areas 
3,390 O—H stretch 87.2 7.4 
2,919–2,887 C–H stretch methylene 24.3  
2,854 Stretching C—H of CH2 and CH aromatic and aliphatic 9.6  
1,608–1,508 C=O—O carboxylate ion, aromatic band stretching 1.6  
1,596 NH deformation primary amine 0.4 4.3 
1,509 Stretching C—C of aromatic 5.5  
1,463 CH3 symmetric deformation of Si-CH3 6.1  
1,168 C=S 10.5 3.4 
1,060–1,039 Si—O—Si stretching vibration 100.0 100.0 
830 C—O—C oxirane 0.9 3.5 
636 C—H out of plane bending for aromatics 0.9 0.9 
  Coating A Coating B 
Wavenumber (cm−1Vibrations % Normalised relative areas 
3,390 O—H stretch 87.2 7.4 
2,919–2,887 C–H stretch methylene 24.3  
2,854 Stretching C—H of CH2 and CH aromatic and aliphatic 9.6  
1,608–1,508 C=O—O carboxylate ion, aromatic band stretching 1.6  
1,596 NH deformation primary amine 0.4 4.3 
1,509 Stretching C—C of aromatic 5.5  
1,463 CH3 symmetric deformation of Si-CH3 6.1  
1,168 C=S 10.5 3.4 
1,060–1,039 Si—O—Si stretching vibration 100.0 100.0 
830 C—O—C oxirane 0.9 3.5 
636 C—H out of plane bending for aromatics 0.9 0.9 
Figure 1

FTIR spectra of (a) coating A and (b) coating B.

Figure 1

FTIR spectra of (a) coating A and (b) coating B.

The polar interaction between the citric acid and the epoxy coatings was examined by FTIR in Figure 2. Coating A was immersed in 5% citric acid for up to 18 months. To examine the interaction more fully, the multicomponent bands were resolved into their individual peaks by using a curve resolving algorithm based on the Levenberg–Marquardt method (Marquardt 1963). Figure 2(a) shows the full FTIR spectra and Figure 2(b) the resolved peaks. The region from 3,700 to 3,100 cm−1 was de-convoluted into three peaks. The partially resolved peak at 3,627–3,563 cm−1 was attributed to free permeate. The band at 3,450 cm−1 was assigned to polar permeate hydrogen-bonded to hydroxyl groups, and 3,220 cm−1 to polar permeate hydrogen-bonded to amine groups (Bellenger et al. 1989; Musto et al. 2000; Feng et al. 2004).

Figure 2

(a) FTIR spectra of raw coating A and samples immersed in 5% citric acid for up to 18 months and (b) resolved FTIR peaks.

Figure 2

(a) FTIR spectra of raw coating A and samples immersed in 5% citric acid for up to 18 months and (b) resolved FTIR peaks.

Figure 2(b) shows the absence of free acid in the raw coating, which increased with immersion to 12 months then declined from 12 to 18 months. The permeation of water does not disrupt the hydrogen bonding of water already sorbed in the epoxy (Musto et al. 2000); our results however, show acids can displace sorbed water. Permeates bonded to the OH showed a slight shift in wavelength from 3,415 to 3,390 cm−1 with acid immersion. A similar shift from 3,250 to 3,210 cm−1 was noted for permeate bonded to the amine group after acid immersion. Both shifts were attributed to the replacement of the bound water with the acid (Omoike & Chorover 2004). The peak area at 3,390 cm−1 increased in 6 months, but little increase occurred from 6 to 12 months. This was attributed to size exclusion of the acids as the amount of acid adsorbed to OH may have promoted pore blockage. However, between 12 and 18 months, a further rise in the peak area was observed, suggesting that additional OH sites, perhaps in the more dense phase, have been accessed. The area at 3,220 cm−1 increased in 6 months and declined thereafter. This reflects the reversibility of citrate adsorption on amine groups consistent with the reversible sorption of water on this group (Bellenger et al. 1989; Musto et al. 2000). These results suggest that acid transportation through epoxy was dictated by the rate of acid adsorption and desorption, which in turn was controlled by the strength of the polarity interactions between acid and epoxy. The relatively higher areas associated with acids' hydrogen bonded to both the OH and NH group show the fraction of acid bound by polarity interaction is greater than acids present in free state in the epoxy. The effects of polarity interactions on acid transport will be further examined below.

Effect of acid and epoxy polarity on the mass transport of acids

The migration of acids through the coatings was assessed by fitting gravimetric immersion data to the multiphase Jacobs and Jones model (Jacobs & Jones 1989). The formation of different phase structures in epoxies is well established and provides a suitable explanation for the observed permeation behaviour of acids (Vanlandingham et al. 1999; Feng et al. 2004). To take into account the presence of the two phases, the normalised permeate content M(t) = Mt/M as a function of time was described by the following morphology dependent equation (Jacobs & Jones 1989): 
formula
2
where Mt and M are the percentage permeate content with time and at equilibrium, Dd and Dl are the diffusion coefficients in the dense and less dense phases, respectively, and Vd is the volume fraction of the dense phase. The diffusion coefficients were estimated from the equilibrium permeate concentrations of the dense and less dense phase, Md and Ml. The value of Ml was estimated by extrapolating the slope of the plot of M(t) as a function of t1/2 back to the M(t) axis.

The diffusion coefficients of the various acids at different concentration in both dense and less dense phases of coatings A and B are reported in Table 4. As shown, the diffusion coefficients in the less dense phase were generally faster than the dense phase in both coatings, by a ratio of up to 5,100. The presence of two phases has been proposed to occur via initial formation of microgels, which deplete their immediate neighbourhood of reactants. At the later stages of cure, the regions between the microgel particles cross-link to form the soft phase (Zhou & Lucas 1999). The high cross-linking density is responsible for the greater packing effect resulting in reduced free volume, thus providing a higher resistance to permeation (Feng et al. 2004).

Table 4

Acid diffusion coefficients in the less dense and more dense phase of coating A and coating B

  Dl × 1010cm2/s
 
Dd × 1012cm2/s
 
Acids Coating A Coating B Coating A Coating B 
1% citric acid 8.61 5.64 2.12 2.56 
5% citric acid 8.66 14.88 2.03 2.90 
10% citric acid 8.26 17.48 1.69 2.74 
1% oxalic acid 96.72 77.41 2.32 1.68 
5% oxalic acid 90.62 148.40 2.05 2.88 
10% oxalic acid 98.55 128.43 2.01 2.39 
1% nitric acid 2.56 6.43 1.61 2.19 
5% nitric acid 6.71 8.42 1.44 2.66 
10% nitric acid 9.80 10.41 1.44 1.47 
5% sulphuric acid 4.94 8.92 1.57 2.12 
10% sulphuric acid 2.50 3.04 1.54 2.09 
20% sulphuric acid 4.95 5.89 1.80 1.87 
  Dl × 1010cm2/s
 
Dd × 1012cm2/s
 
Acids Coating A Coating B Coating A Coating B 
1% citric acid 8.61 5.64 2.12 2.56 
5% citric acid 8.66 14.88 2.03 2.90 
10% citric acid 8.26 17.48 1.69 2.74 
1% oxalic acid 96.72 77.41 2.32 1.68 
5% oxalic acid 90.62 148.40 2.05 2.88 
10% oxalic acid 98.55 128.43 2.01 2.39 
1% nitric acid 2.56 6.43 1.61 2.19 
5% nitric acid 6.71 8.42 1.44 2.66 
10% nitric acid 9.80 10.41 1.44 1.47 
5% sulphuric acid 4.94 8.92 1.57 2.12 
10% sulphuric acid 2.50 3.04 1.54 2.09 
20% sulphuric acid 4.95 5.89 1.80 1.87 

The role of acid polarity in acid transport is exemplified by the faster diffusion coefficients exhibited by oxalic acid in the less dense phase of the epoxies. Its diffusion coefficients are 7–49 times faster compared to the other acids. Because oxalic acid is the least polar among the acids tested (see Table 2), its polarity interaction with other polar species in the epoxy resin was also the least. This permitted oxalic acid to diffuse faster compared to the other acids tested.

The role of epoxy polarity was demonstrated by the lower diffusion coefficients of the acids in the less dense phases of coating A. Coating A, which was relatively richer in polar sites (see Table 3), appears to promote greater hydrogen bonding with the acids that in turn restricted acid diffusion. The diffusion coefficients of the acids in the dense phase of the epoxies, as shown in Table 4, were not significantly influenced by the type of acids and the epoxy polarity.

These results suggest that the acid transport through the epoxy occurs by hindered diffusion (Yang & Guin 1996). In the less dense phase, hindered diffusion occurred as a result of polarity interaction between acids and by the polar groups in the epoxy (Soles et al. 2000). Highly polar acids exhibited lower diffusion coefficients compared to the less polar acids, such as oxalic acid; while in the more dense phase, hindered diffusion appeared to occur primarily as a result of steric exclusion because of the greater packing effect in these crystalline phases (Deen 1987). The eventual outcome of acid permeation will be the failure of the protective coating system manifested by its delamination. As the acids reach the coating-concrete interface, it will begin corroding the concrete. The resulting loss of concrete integrity will promote coating delamination by either the cohesive failure of the concrete and/or adhesion failure. The rate of delamination, in the absence of interfacial flaws and osmotic pressures from the concrete substrate, will thus be primarily dictated by the permeability of the coating material (Raupach & Wolff 2009). These results show that coatings containing higher concentrations of polar groups and higher density phases, as demonstrated by coating A, will provide greater resistance to acid permeation and in turn will be able to prolong coating adhesion.

CONCLUSIONS

The transport of sewer metabolised acids through two-phased epoxy mortar linings was examined by the Jacobs-Jones model (Jacob & Jones 1989). Acid migration was characterised by hindered diffusion. Hindered diffusion in the less dense phase occurred as a result of polarity interactions between the epoxy and acids by hydrogen bonding. This reduced the diffusion coefficient by a ratio of up to 49. In the more dense phase, hindered diffusion occurred as a result of size exclusions from the greater packing or reduced pore volume, resulting in reduction of the diffusion coefficient by a ratio of up to 5,100. These results demonstrate the roles of the lining polarity and polymer density relative to their resistance to acid permeation.

ACKNOWLEDGMENTS

The author acknowledges the financial support provided by the Australian Research Council and many members of the Australian water industry through LP0882016 the Sewer Corrosion and Odour Research (SCORe) Project (www.score.org.au).

REFERENCES

REFERENCES
Feng
J. X.
Berger
K. R.
Douglas
E. P.
2004
Water vapor transport in liquid crystalline and non-liquid crystalline epoxies
.
Journal of Materials Science
39
(
10
),
3413
3423
.
Gu
J. D.
Lu
C.
Mitchell
R.
Thorp
K.
Crasto
A.
1997
Fungal degradation of fiber-reinforced composite materials
.
Materials Performance
36
(
3
),
37
42
.
Haynes
W.
2014
CRC–Handbook of Physics and Chemistry
.
95th edn, CRC Press, Boca Raton, FL, USA
.
Lamberet
S.
Guinot
S.
Lempereur
E.
Talley
J.
Alt
C.
2008
Field investigations of high performance calcium aluminate mortar for wastewater applications
. In:
Calcium Aluminate Cements, Proceedings of the Centenary Conference 2008
,
IHS BRE Press
,
Palais des Papes, Avignon, France
.
Maggana
C.
Pissis
P.
1999
Water sorption and diffusion studies in an epoxy resin system
.
Journal of Polymer Science Part B-Polymer Physics
37
(
11
),
1165
1182
.
Marquardt
D. W.
1963
An algorithm for least-squares estimation of nonlinear parameters
.
Journal of the Society for Industrial and Applied Mathematics
11
(
2
),
431
441
.
Musto
P.
Ragosta
G.
Mascia
L.
2000
Vibrational spectroscopy evidence for the dual nature of water sorbed into epoxy resins
.
Chemistry of Materials
12
(
5
),
1331
1341
.
Okabe
S.
Odagiri
M.
Ito
T.
Satoh
H.
2007
Succession of sulfur-oxidizing bacteria in the microbial community on corroding concrete in sewer systems
.
Applied and Environmental Microbiology
73
(
3
),
971
980
.
Raupach
M.
Wolff
L.
2009
Durability of adhesion of epoxy coatings on concrete; causes of delamination and blistering
.
Soles
C. L.
Chang
F. T.
Gidley
D. W.
Yee
A. F.
2000
Contributions of the nanovoid structure to the kinetics of moisture transport in epoxy resins
.
Journal of Polymer Science Part B-Polymer Physics
38
(
5
),
776
791
.
Vanlandingham
M. R.
Eduljee
R. F.
Gillespie
J. W.
1999
Relationships between stoichiometry, microstructure and properties for amine-cured epoxies
.
Journal of Applied Polymer Science
71
(
5
),
699
712
.
Yang
X. F.
Guin
J. A.
1996
Effects of solute adsorption on hindered diffusion uptake rates in finite bath experiments
.
Chemical Engineering Communications
154
,
101
118
.