Research Article
18 July 2019

Influence of Dissolved-Aluminum Concentration on Sulfur-Oxidizing Bacterial Activity in the Biodeterioration of Concrete


Several studies undertaken on the biodeterioration of concrete sewer infrastructures have highlighted the better durability of aluminate-based materials. The bacteriostatic effect of aluminum has been suggested to explain the increase in durability of these materials. However, no clear demonstration of the negative effect of aluminum on cell growth has been yet provided in the literature. In the present study, we sought to investigate the inhibitory potential of dissolved aluminum on nonsterile microbial cultures containing sulfur-oxidizing microorganisms. Both kinetic (maximum specific growth rate) and stoichiometric (oxygen consumption yield) parameters describing cells activity were accurately determined by using respirometry measurements coupled with modeled data obtained from fed-batch cultures run for several days at pH below 4 and with increasing total aluminum (Altot) concentrations from 0 to 100 mM. Short-term inhibition was observed for cells poorly acclimated to high salinity. However, inhibition was significantly attenuated for cells grown on mortar substrate. Moreover, after a rapid adaptation, and for an Altot concentration up to 100 mM, both kinetic and stoichiometric growth parameters remained similar to those obtained in control culture conditions where no aluminum was added. This argued in favor of the impact of ionic strength change on the growth of sulfur-oxidizing microorganism rather than an inhibitory effect of dissolved aluminum. Other assumptions must therefore be put forward in order to explain the better durability of cement containing aluminate-based materials in sewer networks. Among these assumptions, the influence of physical or chemical properties of the material (phase reactivity, porosity, etc.) might be proposed.
IMPORTANCE Biodeterioration of cement infrastructures represents 5 to 20% of observed deteriorations within the sewer network. Such biodeterioration events are mainly due to microbial sulfur-oxidizing activity which produces sulfuric acid able to dissolve cementitious material. Calcium aluminate cement materials are more resistant to biodeterioration compared to the commonly used Portland cement. Several theories have been suggested to describe this resistance, and the bacteriostatic effect of aluminum seems to be the most plausible explanation. However, results reported by the several studies on this exact topic are highly controversial. This present study provides a comprehensive analysis of the influence of dissolved aluminum on growth parameters of long-term cultures of sulfur-oxidizing bacterial consortia sampled from different origins. Kinetic and stoichiometric parameters estimated by respirometry measurements and modeling showed that total dissolved-aluminum concentrations up to 100 mM were not inhibitory, but it is more likely that a sudden increase in the ionic strength affects cell growth. Therefore, it appears that the bacteriostatic effect of aluminum on microbial growth cannot explain the better durability of aluminate based cementitious materials.


The deterioration of concrete pipelines in the sewer network stands for major environmental and economic issues and this is partly due to destructive biological activity (1). The formation of septic zones within the network is prone to the development of sulfate-reducing bacteria able to convert sulfate into H2S. When the H2S in the wastewater strips into the gas phase of the sewers, it can react on the surface of the concrete pipe and decrease the surface pH below 9. In contact with the concrete, sulfur compounds from H2S are oxidized into sulfuric acid by neutrophilic and then by acidophilic sulfur-oxidizing bacteria (SOB), which decreases the surface pH to approximately 2. The biogenic sulfuric acid induces the leaching (acid-base reaction) of the cementitious matrix, which creates decalcified/deteriorated zones on the surface of the cementitious material. Sulfate penetrates inside the cement pores, reacts with the hydrates (and to a minor extent with the anhydrous phase), and finally creates expansive phases such as gypsum and ettringite precipitates, which are responsible for mortar cracking (24). Cementitious materials with high aluminum content (typically in the range of 50 wt% [weight percent]) have demonstrated a higher resistance to biological attack compared to more conventional cementitious materials (the aluminum content of which does not exceed 15 wt%) but a lower resistance compared to other various abiotic attack (5). The stronger resistance of cement containing high levels of aluminum content when facing biotic attacks can be explained either by physicochemical processes or by biological activities. The resistance due to physicochemical processes is thought to be based on a higher buffer capacity, a higher stability of hydrates and a lower diffusion of aluminum rich cementitious materials (610). The resistance due to biological activity is mainly explained in the literature by an inhibition or a bacteriostatic effect of soluble aluminum on SOB (5, 913).
A review of results from the literature (see Table S1 in the supplemental material) reveals that only very few authors have evaluated this inhibition degree for chemoautotrophic microorganisms, and there are even fewer studies for SOB. Moreover, the reported results are largely controversial since the inhibition degree conventionally measured as a significant decrease in the maximal specific growth rate (μmax) ranged from 0 to 98% for added aluminum concentrations from 13 to 400 mM. It is generally accepted that the degree of inhibition by adding a compound is proportional to its free metal ion concentration which does not stand out from the literature results (Table S1). A thorough analysis of the methodologies used by the authors showed that the microbial culture conditions and the methods used to assess the degree of inhibition were not systematically detailed. However, the reliability of an inhibition test depends on numerous interdependent factors, such as the nature and physiological state of the microorganisms, the efficient control of operating conditions of the culture (pH, temperature, concentrations of CO2, O2, and nutrients, stirring intensity, etc.), the method used to quantify bacterial growth or activity, and the duration of the test. The choice of the strain or microbial population to be used in the test is crucial. Two options can be adopted: (i) the use of (one or various) pure strains in a controlled nutrient medium or (ii) the use of a microbial population selected under environmental conditions that seeks to be representative for those relevant at in situ conditions. Except for the study performed by Herisson et al. (6), no other test has been undertaken on microbial consortia, and none has been undertaken for consortia grown directly on cementitious material. The adaptation capacity of a microbial population to an inhibitory compound during long-term experiments is thus another crucial factor to make an inhibition test valid for practical applications. This aspect was, however, not considered within these studies. Moreover, regarding the large spatial and temporal variations of pH within biofilms which grow at the surface of the cementitious material, the effect of pH on inhibition should also be considered (4). Elsewhere, the actual method used to quantify inhibitory effects of a specific compound (i.e., reduced cell growth rate with increasing compound concentration) must be chosen with care. The complexity of biological, physical, and chemical processes, all three occurring during cell growth, most often make the interpretation of experimental responses a difficult and delicate matter. Responses are based on online respirometry, which gives the required data to calculate oxygen uptake rate (OUR) all along the culture time. This technique is very useful for determining reliable and accurate cell growth parameters, especially when combined with elemental mass balances, such as chemical oxygen demand or nitrogen, and is associated with a cell growth model (14, 15). In the latter case, both kinetic and stoichiometric parameters can be determined with accuracy (16). Recently, this technique has been successfully applied for characterizing SOB activities (17, 18). Nevertheless, some biases may be encountered. When spiking an inhibitory compound, a surprising increase in OUR instead of a decrease was reported, and this was attributed to an increase in energy requirements for cell maintenance (19, 20). Therefore, special attention must be given to differentiate oxygen consumption related to cell growth from other oxygen-demanding endogenous processes and cell maintenance (21).
In this present study, the inhibitory effect of dissolved aluminum was assessed on microbial enrichments presenting sulfur-oxidizing activity and characterized by 16S rRNA gene sequencing. Sequential fed-batch cultures with different concentrations of dissolved aluminum were run on a long-term basis in order to allow microbial acclimation and under controlled conditions to avoid, for instance, any growth limitation by CO2 and O2. Two acidic pH ranges were tested: (i) pH 4 to simulate a slightly deteriorated concrete surface and (ii) pH between 3.20 and 1.65 to simulate a highly deteriorated concrete surface. A respirometry technique developed by Spérandio (15) was used in this study; this method involved two parameters, the maximum specific growth rate (μmax) and the oxygen consumption yield (Yo/s), expressed as a molar ratio O2/S4O62−. These two parameters are highly complementary for evaluating the inhibition effect of compounds. In response to a potential inhibition occurring at increasing dissolved-aluminum concentrations, a change in growth kinetic was identified by comparing μmax values, and a change in the efficiency of growth due to an increase in maintenance energy requirements was highlighted by comparing Yo/s values. Finally, the origin of inhibition, i.e., a direct effect of dissolved aluminum or an effect of the increase in the ionic strength was further discussed.


Bacterial population analysis of the selected inocula.

For the inhibition tests, three inocula were produced from activated sludge samples (see the detailed description of inoculum cultivation in Materials and Methods). The objective was to get microbial enrichments presenting well-calibrated sulfur-oxidizing activity. The first inoculum was grown as a suspended culture (free cells and microaggregates with small diameters of <10 μm). The second and the third inocula were grown as attached biofilms on plates of either aluminum-rich mortar or inert polyethylene plastic. These differences between inocula were chosen in order to test the effect of cell acclimation to high values of ionic strength on cell sensitivity to aluminum. A feeding protocol, based on a specific sulfur substrate (i.e., the tetrathionate), allowed us to enrich the inoculum cultures in SOB and to reproduce a high acid production rate and thus aggressive biodeterioration conditions such as occur in sewer networks. Activities obtained for these inocula can be seen below (see Table 1) and correspond to kinetic (μmax) and stoichiometric (Yo/s) parameters given for controls. These parameters remained rather constant for these controls all along the inhibition test. A 16S rRNA gene sequencing was performed on bacterial populations of each inoculum, and the results are presented in Fig. 1. The diversity of the SOB was very low and is mainly represented by two species of the genus Acidithiobacillus: A. albertensis and A. thiooxidans. The A. thiooxidans species appeared to be largely dominant for all tested inoculum and thus independent of the culture condition. Their proportions were 94, 85, and 97% for suspended cultures, biofilms grown on mortar, and biofilms grown on polyethylene (PE) plastic, respectively. Heterotrophic bacteria were not found in large proportion in the samples (<10% in all). The following main heterotrophic species were detected: Acidomonas methanolica, a facultative methylotrophic Gram-negative bacillus able to grow at between pH 2 and pH 5 (22); Sphingomonas echinoides, a Gram-negative bacillus with polar flagellation (23); Legionella lytica, a Gram-negative, motile rod-shaped bacterium; and Rudaeicoccus suwonensis, a Gram-positive, nonmotile coccoid (24). All of these heterotrophs are aerobic.
FIG 1 Analysis of the bacterial microbiota of each inoculum used for the inhibition tests. The relative abundances of species-level bacterial taxonomies based on PCR amplifications and high-throughput sequencing of 16S rRNA gene fragments are depicted. The “other” group includes minor classes with <0.6% of total abundance.

Effect of dissolved aluminum on SOB grown under slightly deteriorated concrete surface conditions.

In this first experiment, termed experiment I, a microbial consortium enriched in SOB, previously developed as a biofilm structure on a PE plastic plate (no salt effect), was used to inoculate two fed-batch cultures: a control culture without AlCl3 addition and a test culture receiving increasing AlCl3 concentrations. The evolution of the OUR in the test culture is presented in Fig. 2, and a summary of the μmax and Yo/s values obtained during this experiment I is given in the first row of Table 1. The OUR profile in Fig. 2 highlighted five distinct phases. Phase 1, from 0 to 24 h, refers to a control period where SOB grew without aluminum. Together with the use of the OUR data obtained from the control culture, this phase allowed to calculate the values of μmax and Yo/s in the absence of Altot in the medium. Phase 2, from 24 to 60 h, refers to a period with an Altot concentration reaching 52 mM. A change in the OUR profile compared to the control period was rapidly observed during the substrate consumption phase. No value of μmax could be obtained for this period. At 50 h, OUR suddenly decreased with a strong decrease of μmax and an increase of Yo/s. Growth was hence affected by the sudden presence of Altot. Phase 3, from 60 to 122 h, characterized first by an exponential increase in the OUR between 60 and 100 h and second by a drop in the OUR from 122 h onward, until reaching levels of endogenous respiration was observed due to a lack of substrate. Substrate injection temporarily stopped at 90 h. Therefore, between 90 and 100 h, the OUR slightly increased due to the residual substrate introduced at 90 h. Phase 4, from 160 to 194 h, was identified to assess the influence of a new addition of aluminum on bacterial growth. At 168 h, AlCl3 was added in order to reach an Altot concentration of 72 mM. The Yo/s temporally increased but then dropped again to the conventional value, although the Altot concentration in the bulk liquid remained constant at 72 mM. Finally, at 194 h, a new addition led to an Altot concentration of 90 mM. Furthermore, growth increased exponentially showing that the cells acclimated to the high concentrations of Altot.
FIG 2 Evolution of the OUR of the microbial population selected on inert material for increasing concentrations of Altot. This experiment was conducted with pulsed addition of S4O62− (↓) at a controlled pH of 4.
TABLE 1 Maximum specific growth rate (μmax) and oxidation yield (Yo/s) variation with increasing inhibitory compound concentrations of microbial populations selected at pH 2 or 4 with or without preacclimation
ExptInoculum originpHCompoundaC concn
Ionic strength
Value (% variation compared to control)c
(mol O2/mol S4O62−)
IBiofilm developed on inert PE materiald4Control0Low value0.0612.69
 AlCl352312Low value (–100)2.98 (11)
 AlCl3523120.048 (–21)2.67 (–1)
 AlCl3905400.057 (–7)2.98 (11)
IIBiofilm developed on CAC mortar4Control0?0.0382.59
 AlCl352? + 3120.010 (–74)2.91 (12)
 AlCl352? + 3120.034 (–11)2.50 (–3)
IIIFree cells and microaggregates
developed in a reactor
 AlCl352323*0.020 (–51)3.33 (10)
 AlCl3104630*0.035 (–15)3.09 (2)
IVFree cells and microaggregates
developed in a reactor
 NaCl310316*0.044 (7)3.13 (3)
 NaCl937954*0.017 (–59)3.24 (7)
Control, no addition.
?, salts brought with particles sampled with the biofilm and transferred into the batch reactor prevented correct estimation of the ionic strength. *, the ionic strength is given for the higher pH.
Underlined variations are significant compared to the error.
Inert means there was no presence of aluminum and no release of salt, so these substances had no influence on bacteria (in contrast to what happened in the mortar).
Experiment II, was performed under similar conditions compared to experiment I except that the inoculum was the one developed as biofilm on a mortar containing high levels of aluminum (50 wt%). Figure 3 presents the evolution of the OUR under AlCl3 additions. The growth parameters for these new fed-batch cultures are given in the second row of Table 1. Phase 1 corresponds to the reference growth, phase 2 corresponds to the inhibited growth, and phase 3 corresponds to the growth after acclimation. The OUR profile did not show any strong decrease after the addition of AlCl3 in the second phase. This result contrasts to what was observed in experiment I using an inoculum from a biofilm produced on inert plastic material. In addition, the μmax determined for the control culture was significantly less compared to experiment I. The μmax and the Yo/s values decreased immediately after the addition of aluminum but reached values similar to those of the control less than 24 h after injecting AlCl3.
FIG 3 Evolution of the OUR of the microbial population developed on CAC mortar for increasing concentrations of Altot. This experiment was conducted with pulsed addition of S4O62− (↓) at a controlled pH of 4.

Effect of dissolved aluminum or sodium chloride on SOB grown under highly deteriorated concrete surface conditions.

Experiment III was performed in order to evaluate the effect of aluminum on SOB growth, but under lower pH values compared to the previous experiments, which could reproduce highly deteriorated concrete surface conditions. The pH ranged between 3.20 and 1.75. Aluminum addition trials were undertaken under 52 mM Altot (ionic strength, 310 mM) and then 104 mM Altot (ionic strength, 620 mM). The estimated growth parameters are given Table 1, together with parameter variations compared to the control culture. In experiment IV, NaCl was used as ionic compound instead of AlCl3 in order to differentiate a direct inhibitory effect of aluminum from the negative effect of ionic strength. The NaCl concentrations were tested at 310 mM (ionic strength, 310 mM) or 937 mM (ionic strength, 937 mM). The pH ranged from 3.00 to 1.65. The obtained growth parameters are also presented in Table 1. Cultures for experiments III and IV were run in parallel with an inoculum from the same origin (suspended culture). The OUR profiles are shown in Fig. 4. The same OUR profiles were observed between the control and the test cultures for the two Altot tested concentrations and for the lowest concentration of NaCl. However, the highest concentration of NaCl led to a decrease in the OUR and μmax values. These results show that, after a short period of acclimation and compared to values obtained from the control culture, there was no significant effect of the addition of AlCl3 or NaCl at 310 mM. However, the growth was highly affected by the presence of NaCl at 937 mM.
FIG 4 Time courses of oxygen uptake rate for sequential batch cultures of sulfur-oxidizing bacteria selected from activated sludge and run at a controlled pH of 2 for the experiment with addition of salt at t = 0 (red) and the control (blue). Vertical arrows (↓) represent the times when additions of S4O62− were performed.


Representativeness of the inhibition test conditions.

The aim of this present study was to evaluate the inhibitory effect of dissolved aluminum on SOB in general terms but also in the specific case of sewer networks. As shown in Fig. 1, the culture conditions to produce the three inocula all led to the predominance (>85%) of a bacterial species Acidithiobacillus thiooxidans, which is well known to develop at very low pH values in corroded concrete pipes (25). According to Islander et al. (3), A. thiooxidans is the most commonly cultured SOB organism among Acidithiobacillus from sewer concrete corrosion. The dominance of one SOB species, i.e., Acidithiobacillus thiooxidans, is consistent with the findings of Okabe et al. (4), who performed a metagenomic analysis of biofilms grown on sample mortar plates placed 1 year in real sewer networks. A similar dominance of one species of Acidithiobacillus on coupons placed on the corrosion layers of sewers was confirmed by Vincke et al. (26) and recently by Jiang et al. (27). A recent study on acidophilic microorganisms encountered in aerobic biofilms of sewers has underlined the high biodiversity of the corrosion community and its variation from different sewer environmental conditions (28). The bacterial consortia used in the present study showed a low relative abundance of heterotrophs and even in lower proportions compared to previous studies (4, 26, 27). This difference was obviously due to the absence of organic nutrients added to the culture, as opposed to sewer conditions.
The choice of the microorganisms used for inhibition tests is undoubtedly a crucial point. The use of pure strains alone or in a mixture, as well as the use of selected microbial consortia, has always been subject to debate. Literature studies involving inhibition tests of aluminum, in most cases, use pure bacterial cultures. However, looking at the results reviewed from the literature (see Table S1 in the supplemental material), the inhibition or bacteriostatic role of dissolved aluminum appears to be highly variable from one study to another, and no clear effect can be highlighted. Factors such as community interactions, the presence of specific niches, nutrient cycling, the bioavailability of aluminum, etc., should be considered when selecting an inoculum for inhibition tests willing to represent real conditions and notably conditions within sewer pipes. The use of complex microbial consortia enriched in acidophilic SOB should therefore bring consistency to the results. It is obvious that microbial communities associated with corrosion fronts vary in space and time due to the unstable environment. It would be far reaching to represent the whole community structure which should be encountered. However, the use of complex microbial consortia enriched in acidophilic SOB, especially in Acidithiobacillus spp. such as the inoculum used in this study should improve representation of the bacterial community. Finally, the inhibition tests were carried out under two pH ranges but systematically below pH 4. As mentioned in the Materials and Methods, low-pH conditions were chosen to lead to the leaching of aluminum, which occurs in a large majority at high levels of protons (Fig. S1), such levels being only attainable via acidophilic SOB activity (29, 30).

Ability to quantify the inhibitory effect of aluminum.

Several factors are able to impact the growth of SOB in a complex environment. Hence, the level of precision with which the degree of inhibition is determined must be assessed. In this study, oxygen uptake rate measurements were used to estimate kinetic parameters. The μmax values were estimated by using a specific growth model dedicated to SOB, whose mathematical structure was adapted to the experimental operational conditions. As described in Materials and Methods, this model, developed by Peyre-Lavigne (18), only considered tetrathionate as the limiting substrate for SOB growth. No limitation of growth by oxygen, CO2, or other nutrients was included. The experimental setup developed in the present work allowed this hypothesis to be respected, since the continuous bubbling of CO2/O2 prevented the limitation of these substrates, and other nutrients necessary for growth were added in excess before any growth experiment. Moreover, the model considered only one microbial population. Consequently, a change in the selected microbial population could occur over time, with a possible change in the μmax. Because two reactors were systematically run in parallel, one as a control and the other to test the effect of aluminum or salt, a change in the microbial population could not interfere in quantifying the inhibition. Likewise, the presence of heterotrophs (bacteria and fungi), representing realistic conditions, are not likely to modify the assessment of SOB inhibition by aluminum salts because a specific sulfur-based substrate is added for these tests. The OUR profiles and most of all its dynamic change were hence mainly due to the SOB activity. It is also well known that the addition of aluminum can coagulate bacteria, thereby changing their environment and possibly inducing a decrease in the specific growth rate. In this present work, bacteria were grown in the liquid as free cells and microaggregate suspensions (<10 μm). Since no further strong microbial aggregation was observed after the addition of aluminum, the change in biological parameters could not be attributed to a strong change in the physical structure of the microbial culture. The accuracy of the resulting biological parameters obtained by the model may also be questioned. Indeed, the sequential addition of a significant amount of tetrathionate performed for each experiment led to a quantifiable growth value and repeated dynamic signals, which made the estimation of μmax and Yo/s reliable and robust (15, 16). Considering the accuracy of the OUR measurements, error estimations on the determination of μmax only reached 0.003 h−1. This value is equivalent to a difference of 7.5% of the μmax obtained under reference conditions (μmax was ca. 0.04 h−1 for the control). Therefore, a decrease below 15% of the μmax cannot be significant. The oxygen consumption yield Yo/s was determined with an accuracy of 0.11 mol O2/mol S4O62−.
Literature studies in both research areas, whether environmental chemistry (31) or general microbiology (32), indicate that there are significant interactions between the chemical medium and the metal and that the degree of inhibition is linked to the concentration of free metal ion (33). Therefore, metal speciation can affect the bioavailability of the free metal ion, which in general terms is considered to be the toxic metal species (34). In consequence, measuring the degree of inhibition could be affected by the bioavailability of aluminum. However, with the operating conditions used in this study, pH values remained below 4. PHREEQC simulation software was used in this study to evaluate ionic speciation by thermodynamic calculation of the chemical equilibrium. The results of these calculations are shown in Fig. S2, and the conditions for the simulations are given in Table S2. The results indicated that the free Al3+ concentrations exceeded 95% (35). A mass balance on aluminum added to a microbial suspension showed that at pH 2 all the aluminum remained dissolved, proving a good agreement between the theory and the experiment (Fig. S3, Table S3). In conclusion, the method developed here, which is based on a model estimation of growth parameters from oxygen uptake rate measurements undertaken on selected SOB, can be considered an accurate and robust way to quantify an inhibitory effect of dissolved aluminum on SOB and in an acidic environment.

Is dissolved aluminum an inhibitor of SOB growth?

Three aspects must be taken into account in order to answer this question: the bacterial acclimation potential, the type of bacteria used in the test, and the difference between a direct effect of dissolved aluminum compared to the effect of a change in the ionic strength. The ability to acclimate has been clearly demonstrated when comparing the responses obtained by different microbial consortia, either developed on an inert plastic support (no-salt effect) or on a mortar or in a liquid mineral medium (high-salt effect and stringent conditions), when facing AlCl3 additions. The results showed that biofilms previously grown on inert plastic plates were much more sensitive to aluminum additions. However, this sensitive bacterial population can adapt quickly (within a few days) and can grow under high concentrations of Altot (up to 90 mM). This result highlighted the great importance of running long-term cultures to determine the inhibitory character of a compound on a microbial consortium. Moreover, when using the same microbial population, additions of NaCl at a high concentration also showed a strong decrease in μmax, although this decrease was not observed at the lowest NaCl concentration (Fig. 4). These results indicate, as has been previously suggested (36, 37), that cells are sensitive to a first sudden exposure to high ionic strength levels in the media. Therefore, inhibition by addition of AlCl3 can be observed on a short-term basis and for nonacclimated populations. This inhibition is very low or absent when the microbial cultures are previously acclimated to high-ionic-strength media or stringent conditions. Therefore, it was difficult to differentiate bacterial responses obtained due to the effect of dissolved aluminum and those obtained by an increase in the ionic strength. Consequently, a bacteriostatic effect of dissolved aluminum cannot explain the better resistance of aluminum-rich cementitious materials. Other hypotheses, such as a difference in the stabilities of hydrated and anhydrous phases and in the porosity of the materials, have to be further studied in order to address the observed resistance.

Does the effect of AlCl3 or salt vary according to the microbial population selected?

Due to differences in inoculum cultivation conditions, the microorganisms used here experienced different physicochemical environments before they were used for the inhibition tests, and this explains why the μmax and Yo/s values varied between the control cultures. More stringent conditions, such as higher ionic strength or lower pH, led to a decrease in the μmax, whereas only low pH values increased the Yo/s. Control cultures run in parallel to each test culture removed the variability brought by the growth condition of each inoculum and hence gave a fair estimation of the effect of the AlCl3 or NaCl additions alone.
In all cases, the growth parameters of the acclimated populations in the test cultures were identical to those of the control. It can therefore be concluded that, once bacterial populations are acclimated, dissolved aluminum has no inhibitory effect. On the other hand, for microbial populations that did not undergo acclimation, it is the sudden increase in ionic strength that generates the observed transient inhibition. Indeed, a better resistance to salt and to addition of AlCl3 was observed for populations grown under stringent conditions. Typically, these stringent conditions are encountered in aerobic biofilms, which occur on the surfaces of concrete sewer pipes.


The effect of dissolved aluminum on SOB growth was assessed in batch reactors sequentially fed with tetrathionate under controlled environmental conditions at two pH values below 4. From the numerous experiments performed in this study, it can be concluded that dissolved aluminum does not have a significant long-term inhibitory effect on the growth of acidophilic SOB previously developed as biofilm grown either on cement mortar or on a plastic medium. After acclimation, a similar specific growth rate compared to a reference was measured for added concentrations of total dissolved aluminum up to 100 mM. Therefore, the higher resistance of mortar with a high aluminum content in sewer networks cannot be explained by an inhibition of the SOB by dissolved aluminum, and an explanation must be sought elsewhere, such as considering the stability of hydrated and anhydrous phases and a reduced porosity of the materials.


Sulfur-oxidizing bacterium selection.

The influence of dissolved aluminum on sulfur-oxidizing bacteria activity was studied both at pH 4 and at pH ranging from 3.20 to 1.65, two simulated conditions named “slight” and “high” deteriorated concrete surface, respectively. Only low pH values, <4, were used in this study because the leaching of aluminum occurs only when a large number of protons is produced (see Fig. S1 in the supplemental material). These conditions are favorable for the development of acidophilic SOB and hence to pH <4 (29, 30). This is explained by the lower stability of the calcic phases compared to the aluminous phases. For the experiments carried out at a slight level of deteriorated concrete surface, SOB were obtained by precultivated biofilms. After inoculation with an activated sludge sampled at the wastewater treatment plant of the city of Toulouse (France), these biofilms were grown for 2 months, either on a PE material or on a calcium aluminate cement mortar (10 by 4 cm2). The PE did not contain aluminum and, unlike the cement mortar, did not release salts. The biofilms were fed dropwise at 80 ml · h−1 with a nutritive solution containing 2.27 mM potassium tetrathionate (P2926; Aldrich) in order to select SOB (30). After 2 months, the pH reached 3.5, and SOB were manually removed, introduced into 100 ml of deionized water, and vortexed for 15 s in order to disperse the microorganisms without destroying them. The resulting suspended microorganisms were transferred into 250 ml of solution containing the following nutrients: NH4Cl, 2.7 mg liter−1; Na3(PO3)3, 1 mg liter−1; MgCl2⋅6H2O, 1.3 mg liter−1; FeCl3⋅6H2O, 0.07 mg liter−1; MnCl2⋅H2O, 0.08 mg liter−1; and trace elements. This solution was used to inoculate the two reactors run in parallel for studying the inhibitory effect of dissolved aluminum. Suspended cultures were preferred for the experiments in the highly deteriorated concrete surface to avoid the buffering potential of cementitious materials and ensure a high biomass concentration in the inoculum, allowing further activity measurements. Hence, a controlled jacketed open reactor (1.6 liters) was seeded with activated sludge from the wastewater treatment plant of Toulouse (France) and fed with a nutritive solution containing the following: K2S4O6 (P2926; Aldrich), 0.2 mol liter−1; NH4Cl, 2.7 mg liter−1; Na3(PO3)3, 1 mg liter−1; MgCl2⋅6H2O, 1.3 mg liter−1; FeCl3⋅6H2O, 0.07 mg liter−1; MnCl2⋅H2O, 0.08 mg liter−1; and trace elements. The pH was initially 7 and decreased during microbial selection due to the bacterial activity to ca. pH 2.

Experimental setup for measurement of sulfur-oxidizing activity.

In order to quantify inhibition, two reactors were systematically run in parallel: the first one as a reference (no aluminum or salt addition) and the second one to test the inhibitory effect of either AlCl3 or NaCl. The home-made glass reactors were run as batch at a temperature regulated at 20°C and were equipped with online measurements of the dissolved-O2 concentration and temperature (VisiFerm DO Arc 120; Hamilton) and pH (WTW ph296 [WTW, Germany] equipped with a Schott pH probe H8481HD [SI Analytics, Germany]). The reactor volume was adapted to the amount of inoculum produced and was 300 ml for the experiments in the slightly deteriorated concrete surface environment but 1.6 liters for the highly deteriorated concrete surface environment. Sequential bubbling of pressurized air into the reactor was performed to maintain dissolved-oxygen concentration between 4 and 7 mg of O2 liter−1 and to provide the CO2 required for the autotrophic growth of SOB. This sequential aeration was operated by a control-command software (LAC) self-developed by the laboratory which measures the oxygen uptake rate throughout the culture duration. Successive pulses of a concentrated solution of potassium tetrathionate (P2926; Aldrich) at 0.2 M were performed. The volume of each pulse was chosen in order to measure the maximum oxygen uptake rate (OURmax) and was 1.5 ml in the simulated slightly deteriorated concrete surface experiments and 4.5 ml in the simulated highly deteriorated concrete surface conditions. The sulfur mass balance was regularly assessed by measuring the production of sulfate compared to the tetrathionate consumed, and the values were close to 97%. For the experiments in the simulated slight deteriorated concrete surface environment, the pH was buffered at 4 by CO2 enrichment (∼10%) of the pressurized air. For the experiments in the high-acid environment, CO2 enrichment was not performed, and pH was not controlled in order to avoid injecting salts that could affect SOB growth and bias the evaluation of inhibition. The inhibition effect of dissolved aluminum was assessed by injecting a small volume of a concentrated solution of AlCl3 into one reactor, while the other was run as a control without addition. Two Altot concentrations, ∼50 and ∼100 mM, were tested. For the experiments in strongly acidic conditions, the addition of concentrated AlCl3 solution increased the pH of the solution. Hence, the pH in the control culture was systematically adjusted to the pH of the test culture by addition of KOH. It was assumed that the presence of microaggregates (<10 μm) did not influence the effect of the dissolved aluminum. Indeed, considering the small size of the aggregates, all cells were exposed to the same Altot concentration. The walls of the containers and the probes were washed twice a day in order to prevent significant biofilm development.

Evaluation of activity.

Sulfur-oxidizing bacteria growth parameters, i.e., the maximum specific growth rate (μmax) and the oxygen consumption yield (Yo/s) were estimated from the OUR online measurements. The period for the estimation of μmax corresponded to the culture time to obtain a doubling of OUR when possible. The μmax was obtained by fitting a model developed on AQUASIM (EAWAG, Switzerland) (38) with the experimental oxygen uptake rate. Yo/s (mol O2/mol S4O6) was determined for every pulse of substrate by dividing the quantity of dioxygen consumed by the quantity of tetrathionate consumed. The sulfur-oxidizing bacteria growth model was based on equations 1 to 4. Equation 1, proposed by Peyre-Lavigne (18), who worked on tetrathionate conversion by sulfur-oxidizing bacteria, describes the biological oxidation of 1 mol of tetrathionate to produce biomass by reduction of carbon dioxide and ammonium depending on the measured Yo/s. The next equations (equations 2, 3, and 4), respectively, describe the kinetic processes, in a batch reactor, of the biomass growth, the biological oxidation of tetrathionate, and the accumulation of dissolved oxygen resulting from biological consumption and gas-liquid mass transport.
where [C] is the molar concentration of a compound in mol liter−1, [X] is the molar concentration of biomass in mol liter−1, μmax is the maximal specific growth rate in h−1, Yo/s is the oxygen consumption yield in mol O2/mol S4O62–, KLa is the oxygen transfer coefficient, and Ks is the half-saturation constant of S4O62– in mol liter−1. Values associated with the parameters are reported in Table 2.
TABLE 2 Values of parameters used in equations 1 and 4 for the AQUASIM model (38)a
Ks3.5 × 10−4, 3.0 × 10−5, 5.0 × 10−6, and 5.0 × 10−6 mol liter−1I, II, III, IV, respectivelyExperimental estimation and modeling
KLa0.85 liter−1 h−1AllExperimental measurement
Measurements for the parameters Yo/s and μmax may be be found in Table 1.
The model reliability was verified under exponential growth obtained both when substrate pulses were performed and when an excess of sulfur substrate was maintained for a long time. In the latter case, the μmax was directly calculated by fitting the experimental oxygen uptake rate with equation 5:
with OUR (oxygen uptake rate) in mg of O2 liter−1 h−1, A is the constant in mg of O2 liter−1 h−1, μmax is the maximum specific growth rate in h−1, and t is the time in h.
Figure 5 shows the evolution of the experimental and the model oxygen uptake rates. Both methods of μmax estimation gave very similar results (±0.001 h−1). The accuracy of μmax determination was evaluated at ±0.003 h−1 (standard deviation on three identical exponential growths) and that of the oxygen consumption yield was evaluated at ±0.11 mol O2/mol S4O62– (standard deviations on each pulse of two identical growths). To evaluate the effect of dissolved aluminum chloride or sodium chloride on the bacterial growth, the μmax and Yo/s obtained for the control and the test reactors were compared throughout the period of culture.
FIG 5 Time-course of the oxygen uptake rate (OUR) of the selected microbial population for experimental data (orange circles) and for the model data (—). The oxygen uptake rate was fitted using equation 5 and the experiment was conducted in permanent excess of S4O62– (a) or using the complete model (b), and the experiment was conducted under conditions of growth limitation by S4O62– due to the additions of substrate (↓).

Microbial population analysis by 16S rRNA gene sequencing.

The DNA extractions were performed with the PowerSoil DNA Isolation kit (MoBio Laboratories, Inc., Carlsbad, CA) according to the manufacturer’s instructions. The quality of the extracted DNA was evaluated by gel electrophoresis (0.8% agarose), and the nucleic acid concentrations were quantified by Qubit (Thermo Scientific).
The libraries were performed from 3 ng of DNA and amplified using a Thermo Fisher Ion 16S Metagenomics kit. The libraries were created using an Ion Plus Fragment Library kit and Ion Xpress Barcode Adapters according to the manufacturer’s recommendations (Thermo Fisher). The obtained libraries were quantified with Qubit (dsDNA HS assay kit) and a Bioanalyzer 2100 (Agilent DNA1000 kit). Using the Ion 520 and 530 OT2 kits (Thermo Fisher), 8 pg/μl from each library was pooled and sequenced (400 pb) on an Ion 520 chip on an Ion Torrent S5 using according to the manufacturer’s instructions. All the raw reads were processed with Ion Reporter software (V5) using an Ion 16S Metagenomics kit analysis module. This includes a first step of primer trimming and length check. Then, a hash table containing all the unique reads and abundances is created. Finally, the affiliation is performed by multistage BLAST of reads against the Curated Microseq database from Life Technologies and the Curated Green Genes database.

Data availability.

The 16S RNA sequences have been deposited under the Sequence Read Archive (SRA) under SRS4676158, SRS4676162, and SRS4676748 and the reference BioSample accession numbers SAMN11506750, SAMN11506730, and SAMN11506731 .


We gratefully acknowledge the technical staff for their assistance with the experimental setup and Saint-Gobain PAM for funding the project. We also thank the platform GET-Biopuces at the Genopole in Toulouse for the sequencing experiments and the associated bioinformatics analysis. The funders had no role in study design, data collection and interpretation, or the decision to submit this work for publication.

Supplemental Material

File (aem.00302-19-s0001.pdf)
ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.


Jensen HS, Nielsen AH, Hvitved-Jacobsen T, Vollertsen J. 2009. Modeling of hydrogen sulfide oxidation in concrete corrosion products from sewer pipes. Water Environ Res 81:365–373.
Davis JL, Nica D, Shields K, Roberts DJ. 1998. Analysis of concrete from corroded sewer pipe. Int Biodeterior Biodegrad 42:75–84.
Islander R, Devinny J, Mansfeld F, Postyn A, Shih H. 1991. Microbial ecology of crown corrosion in sewers. J Environ Eng 117:751–770.
Okabe S, Odagiri M, Ito T, Satoh H. 2007. Succession of sulfur-oxidizing bacteria in the microbial community on corroding concrete in sewer systems. Appl Environ Microbiol 73:971–980.
Alexander MG, Fourie C. 2011. Performance of sewer pipe concrete mixtures with Portland and calcium aluminate cements subject to mineral and biogenic acid attack. Mater Struct 44:313–330.
Herisson J, Guéguen-Minerbe M, van Hullebusch ED, Chaussadent T. 2014. Behaviour of different cementitious material formulations in sewer networks. Water Sci Technol 69:1502–1508.
Scrivener KL, Cabiron J-L, Letourneux R. 1999. High-performance concretes from calcium aluminate cements. Cem Concr Res 29:1215–1223.
Grandclerc A, Dangla P, Gueguen-Minerbe M, Chaussadent T. 2018. Modeling of the sulfuric acid attack on different types of cementitious materials. Cem Concr Res 105:126–133.
Ehrich S, Helard L, Letourneux R, Willocq J, Bock E. 1999. Biogenic and chemical sulfuric acid corrosion of mortars. J Mater Civ Eng 11:340–344.
Kiliswa MW. 2016. Composition and microstructure of concrete mixtures subjected to biogenic acid corrosion and their role in corrosion prediction of concrete outfall sewers. PhD thesis. University of Cape Town, Cape Town, South Africa.
Herisson J, van Hullebusch ED, Moletta-Denat M, Taquet P, Chaussadent T. 2013. Toward an accelerated biodeterioration test to understand the behavior of Portland and calcium aluminate cementitious materials in sewer networks. Int Biodeterior Biodegrad 84:236–243.
Lors C, Hondjuila Miokono ED, Damidot D. 2017. Interactions between Halothiobacillus neapolitanus and mortars: comparison of the biodeterioration between Portland cement and calcium aluminate cement. Int Biodeterior Biodegrad 121:19–25.
Saucier F, Lamberet S. 2009. Calcium aluminate concrete for sewers: going from qualitative to quantitative evidence of performance, p 398–407. In Alexander MG, Bertron A (ed), Final conference on concrete in aggressive aqueous environments: performance, testing and modeling. Proceedings of the RILEM TC 211-PAE, Toulouse, France.
Spanjers H, Vanrolleghem PA, Olsson G, Dold PL. 1998. Respirometry in control of the activated sludge process: principles. IAWQ, London, United Kingdom.
Spérandio P. 2000. Estimation of wastewater biodegradable COD fractions by combining respirometric experiments in various So/Xo ratios. Water Res 34:1233–1246.
Vanrolleghem PA, Kong Z, Rombouts G, Verstraete W. 1994. An on-line respirographic biosensor for the characterization of load and toxicity of wastewaters. J Chem Technol Biotechnol 59:321–333.
Mora M, López LR, Lafuente J, Pérez J, Kleerebezem R, van Loosdrecht MCM, Gamisans X, Gabriel D. 2016. Respirometric characterization of aerobic sulfide, thiosulfate and elemental sulfur oxidation by S-oxidizing biomass. Water Res 89:282–292.
Peyre-Lavigne M. 2014. Transformations biologiques impliquées dans la dégradation des revêtements cimentaires en réseau d’assainissement: application à la définition d’un test de résistance à la biodétérioration. PhD thesis. INSA Toulouse, Toulouse, France.
Chen G-H, Mo H-K, Saby S, Yip W-K, Liu Y. 2000. Minimization of activated sludge production by chemically stimulated energy spilling. Water Sci Technol 42:189–200.
Low EW, Chase HA, Milner MG, Curtis TP. 2000. Uncoupling of metabolism to reduce biomass production in the activated sludge process. Water Res 34:3204–3212.
Van Bodegom P. 2007. Microbial maintenance: a critical review on its quantification. Microb Ecol 53:513–523.
Urakami T, Tamaoka J, Suzuki K-I, Komagata K. 1989. Acidomonas gen. nov., incorporating Acetobacter methanolicus as Acidomonas methanolica comb. nov. Int J Syst Evol Microbiol 39:50–55.
Denner EBM, Kämpfer P, Busse H-J, Moore E. 1999. Note: Reclassification of Pseudomonas echinoide Heumann 1962, 343AL, in the genus Sphingomonas as Sphingomonas echinoide comb. nov. Int J Syst Evol Microbiol 49:1103–1109.
Kim S-J, Jang Y-H, Ahn J-H, Weon H-Y, Schumann P, Chun S-C, Kwon S-W, Kim W-G. 2013. Rudaeicoccus suwonensis gen. nov., sp. nov., an actinobacterium isolated from the epidermal tissue of a root of a Phalaenopsis orchid. Int J Syst Evol Microbiol 63:1291–1296.
Milde K, Sand W, Wolff W, Bock E. 1983. Thiobacilli of the corroded concrete walls of the Hamburg sewer system. Microbiology 129:1327–1333.
Vincke E, Boon N, Verstraete W. 2001. Analysis of the microbial communities on corroded concrete sewer pipes: a case study. Appl Microbiol Biotechnol 57:776–785.
Jiang G, Zhou M, Chiu TH, Sun X, Keller J, Bond PL. 2016. Wastewater-enhanced microbial corrosion of concrete sewers. Environ Sci Technol 50:8084–8092.
Li X, Kappler U, Jiang G, Bond PL. 2017. The ecology of acidophilic microorganisms in the corroding concrete sewer environment. Front Microbiol 8:683.
Buvignier A. 2018. Caractérization du rôle de l’aluminium dans les interactions entre les microorganismes et les matériaux cimentaires dans le cadre des réseaux d’assainissement. PhD thesis. INSA Toulouse–Université Fédérale Toulouse Midi-Pyrénée, Toulouse, France,
Peyre Lavigne M, Bertron A, Auer L, Hernandez-Raquet G, Foussard J-N, Escadeillas G, Cockx A, Paul E. 2015. An innovative approach to reproduce the biodeterioration of industrial cementitious products in a sewer environment. I. Test design. Cem Concr Res 73:246–256.
Morel FMM, Hering JG. 1993. Principles and applications of aquatic chemistry. John Wiley & Sons, New York, NY.
Hughes MN, Poole RK. 1991. Metal speciation and microbial growth: the hard (and soft) facts. J Gen Microbiol 137:725–734.
Di Toro DM, Allen HE, Bergman HL, Meyer JS, Paquin PR, Santore RC. 2001. Biotic ligand model of the acute toxicity of metals. 1. Technical basis. Environ Toxicol Chem 20:2383–2396.
Angle JS, Chaney RL. 1989. Cadmium resistance screening in nitrilotriacetate-buffered minimal media. Appl Environ Microbiol 55:2101–2104.
Martin RB. 1986. The chemistry of aluminum as related to biology and medicine. Clin Chem 32:1797–1806.
Aston JE, Peyton BM, Lee BD, Apel WA. 2010. Effects of ferrous sulfate, inoculum history, and anionic form on lead, zinc, and copper toxicity to Acidithiobacillus caldus strain BC13. Environ Toxicol Chem 29:2669–2675.
Watling HR, Shiers DW, Zhang GJ. 2012. Microbial behaviour under conditions relevant to heap leaching: Studies using the sulfur-oxidising, moderate thermophile Acidithiobacillus caldus. Hydrometallurgy 127–128:104–111.
Reichert P. 1994. AQUASIM: a tool for simulation and data analysis of aquatic systems. Water Sci Technol 30:21–30.

Information & Contributors


Published In

cover image Applied and Environmental Microbiology
Applied and Environmental Microbiology
Volume 85Number 151 August 2019
eLocator: e00302-19
Editor: Alfons J. M. Stams, Wageningen University
PubMed: 31126946


Received: 4 February 2019
Accepted: 27 April 2019
Published online: 18 July 2019


Request permissions for this article.


  1. concrete biodeterioration
  2. inhibition by aluminum
  3. kinetic parameters
  4. respirometry
  5. sulfur-oxidizing bacteria



Amaury Buvignier
LISBP, Université de Toulouse, CNRS, INRA, INSA, Toulouse, France
LMDC, Université de Toulouse, INSA, UPS, Toulouse, France
Matthieu Peyre-Lavigne
LISBP, Université de Toulouse, CNRS, INRA, INSA, Toulouse, France
Orlane Robin
LISBP, Université de Toulouse, CNRS, INRA, INSA, Toulouse, France
LMDC, Université de Toulouse, INSA, UPS, Toulouse, France
Mansour Bounouba
LISBP, Université de Toulouse, CNRS, INRA, INSA, Toulouse, France
Cédric Patapy
LMDC, Université de Toulouse, INSA, UPS, Toulouse, France
Alexandra Bertron
LMDC, Université de Toulouse, INSA, UPS, Toulouse, France
Etienne Paul
LISBP, Université de Toulouse, CNRS, INRA, INSA, Toulouse, France


Alfons J. M. Stams
Wageningen University


Address correspondence to Etienne Paul, [email protected].

Metrics & Citations



  • For recently published articles, the TOTAL download count will appear as zero until a new month starts.
  • There is a 3- to 4-day delay in article usage, so article usage will not appear immediately after publication.
  • Citation counts come from the Crossref Cited by service.


If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. For an editable text file, please select Medlars format which will download as a .txt file. Simply select your manager software from the list below and click Download.

View Options

Figures and Media






Share the article link

Share with email

Email a colleague

Share on social media

American Society for Microbiology ("ASM") is committed to maintaining your confidence and trust with respect to the information we collect from you on websites owned and operated by ASM ("ASM Web Sites") and other sources. This Privacy Policy sets forth the information we collect about you, how we use this information and the choices you have about how we use such information.
FIND OUT MORE about the privacy policy