1 Introduction

The environmental impact of building materials is a topic that became more and more crucial in recent years, both on a national as well as an international level (European Parliament 2011). In this regard, especially cement-based materials like concrete, play a key role due to the vast annual consumption of over 10 Gt (Plank 2004). Superplasticisers and other additives are used in order to adjust the properties of concrete in 80 to 90% of the concreting work (Lou et al. 2013; Albayrak et al. 2015). The use of especially superplasticisers is a crucial point in cement processing as fresh concrete properties such as workability or setting are controlled by adsorption of these so-called admixtures which adsorb on the cement particle surface and disperse the cement slurry by electrostatic and/or steric repulsive effects (Uchikawa et al. 1997; Mollah et al. 2000). Although superplasticisers are added to the cement slurry in little dosages (0.15–0.3% by cement weight), this sums up to a consumption of over 6 million tons of organic additives due to the extensive amounts of concrete produced (Schröter and Fischer 2010).

Concrete is in permanent contact to the environment, e.g., ground water or soil, which generates conditions for possible leaching of inorganic and organic species. Although concrete is used for a very long time, its composition changed in the last decades from a ternary system (cement, aggregate and water) to a multicomponent system by adding concrete supplements like fly ash, ground granulated blast furnace slag and various organic additives like superplasticisers, retarders, accelerators, etc. While some investigations tailor the leaching behaviour of inorganic compounds from cement-based building materials (Kamali et al. 2008; Barbir et al. 2012; Müllauer et al. 2012), there is a lack of knowledge on the behaviour of organic compounds (Guerandel et al. 2011; Märkl and Stephan 2015). To this purpose, phytotoxicity may be used as an indicator.

Testing phytotoxicity with different plant species is a common method to investigate industrial effluents (Charles et al. 2011), sewage sludges (Fuentes et al. 2004, 2006; Silva et al. 2011) and soil stabilisers like lime or fly ash (Papadimitriou et al. 2008; Baderna et al. 2015). These studies report that, on the one hand, organic matter may positively influence soil and plants. On the other hand, heavy metals may be mobilised and act in a toxic way. Both aspects have to be considered in the case of cement. Direct exposure of different plant species to cement dust at high dosages lead to a reduction of pigment content, dry matter production, plant growth, chlorophyll content, transpiration rate and productivity, whereas low dosage have been shown to promote these parameters (Czaja 1960; Singh and Rao 1981; Mishra and Shukla 1986; Shukla et al. 1990). Furthermore, increased metal pollution could be found in soils near cement plants (Al-Khashman and Shawabkeh 2006; Bermudez et al. 2010), which may influence plant growth. However, and to the authors’ best knowledge, no studies report on the influence of leachates from hardened cement on environmental aspects such as the growth behaviour of plants.

In this study, hardened cement paste samples with and without superplasticiser were used in a white mustard (S. alba) and cress (L. sativum) bioassay, a simple and rapid method to indicate phytotoxicity. The effects on seed germination and primary root growth were determined to investigate the environmental impact of cement-based construction materials and the containing superplasticiser on the environment.

2 Experimental

2.1 Materials and Mix Design

Experiments were performed with cement pastes based on microfine cement from the type ordinary portland cement (OPC; CEM I). This special type of microfine cement is normally used for soil stabilisation. The chemical composition measured by means of X-ray fluorescence spectrometer (XRF) is given in Table 1. As superplasticiser, a commercially available polycarboxylate ether (PCE) was used in this investigation. Further information on the CEM I and the PCE based on the supplier technical or safety data sheet can be found in Table 2. The eluent was either demineralised water (DI) or Berlin tap water (BTW). The mean electrolyte composition of the eluents as determined using inductively coupled plasma optical emission spectroscopy (ICP-OES) and total organic carbon (TOC) can be found in Table 3.

Table 1 Chemical composition of the cement used determined by means of XRF. Data given in [wt. %]
Full size table
Table 2 Information on the used materials
Full size table
Table 3 Mean electrolyte content of the eluents. Data given in [mmol/l]
Full size table

The mixtures were prepared with a water-to-cement ratio (w/c) of 0.5. The superplasticiser dosage was 0.5 wt. % of active component with respect to the amount of cement. Preparation was done using an Ultra Turrax Mixer T50 from IKA, Germany, in order to avoid the presence of agglomerates. First, demineralised water was mixed with the superplasticiser for 30 s at 7000 min−1, followed by the addition of the cement and mixing for 3 min at the same speed. Different mixtures and eluents were used for this study. The considered matrix under investigation is listed in Table 4. The samples were either prepared with or without the addition of superplasticiser (REF vs. PCE). As eluent, either DI or BTW was used.

Table 4 Mixtures used in the experiments
Full size table

2.2 Leaching Test

Leaching tests were performed on paste cubes 20 · 20 · 20 mm3 based on the European tank leach test according to the guidelines of the European Committee for Standardisation (DIN CEN/TS 16637–2 2014). Each cube was immersed in the eluent with a liquid-to-solid-ratio of 80 l/m2 in an amber glass bottle and stored at 20 °C/99% RH. A full exchange of the eluent took place after 6 h, 1 day, 2 days 6 h, 4 days, 9 days, 16 days, 36 days and 64 days. All eluate fractions were kept frozen until measurement and were taken for phytotoxicological tests.

2.3 Test Methods

2.3.1 Primary Root Growth Test with White Mustard S. alba Cress and L. sativum

For the germination test, white mustard seeds (S. alba) and cress seeds (L. sativum) were used. The test was based on DIN EN ISO 11269–2 2012. Twenty-five seeds were placed on four layers of tissue (Carl Roth, Germany) in 9 cm petri dishes. The petri dishes were equipped with a lid to prevent evaporation of the eluate. The tissues were taken to exclude possible influence of soil. Different proportions of leachate of 100, 50, 25 and 12.5% were adjusted from the pure eluates using distilled water. To each petri dish, 5 ml of the appropriate concentration was added. Distilled water was used as a control and each experiment was conducted with five replicates. The seeds were incubated at 25 °C with a light intensity of 4–5 lm/m2. After 4 days, the primary root growth was measured. The percentage of relative seed growth (RSG), the relative root growth (RRG) and germination index (GI) were calculated as follows:

$$ \mathrm{R}\mathrm{S}\mathrm{G}\ \left[\%\right]=\frac{\mathrm{number}\ \mathrm{of}\ \mathrm{seeds}\ \mathrm{germinated}\ \mathrm{in}\ \mathrm{eluate}}{\mathrm{number}\ \mathrm{of}\ \mathrm{seeds}\ \mathrm{germinated}\ \mathrm{in}\ \mathrm{control}} \times 100 $$
(1)
$$ \mathrm{R}\mathrm{R}\mathrm{G}\ \left[\%\right]=\frac{\mathrm{mean}\ \mathrm{root}\ \mathrm{length}\ \mathrm{in}\ \mathrm{eluate}}{\mathrm{mean}\ \mathrm{root}\ \mathrm{length}\ \mathrm{in}\ \mathrm{control}}\times 100 $$
(2)
$$ \mathrm{G}\mathrm{I}\ \left[\%\right]=\frac{\mathrm{RSG}\times \mathrm{RRG}}{100} $$
(3)

2.3.2 Statistical analysis

Experimental results were statistically analysed using a two-tailed t – test to compare paired means between different treatments. One-way analysis of variance (ANOVA) was performed to investigate the generated leachates on RSG, RRG and GI. The analysis was performed with GraphPad Prism version 7 for Windows, GraphPad Software, La Jolla California, USA.

3 Results and discussion

3.1 Relative Root Growth

As simple bioassays provide a method to test stabilised solidified waste and sludge leachates with regard to phytotoxicity (Walter et al. 2006), they were included in the test scheme. In Fig. 1, the root elongation is depicted for S. alba and L. sativum after the incubation time. It can be seen that in general S. alba shows a diminished root length compared to L. sativum. However, both test species exhibit a suitable root length after the chosen test period. Anyhow, S. alba shows a higher distribution of root elongation, which can be seen in Fig. 1.

Fig. 1
figure 1

Observed root growth of S. alba (top) and L. sativum (bottom) after exposure time of 4 days

Full size image
Fig. 2
figure 2figure 2

Mean RRG of S. alba and L. sativum in water eluates depicted for all four concentrations after 0.25 days (a, b), 1 day (c, d) and 2.25 days (e, f). Distilled water was used as control for which the RRG was 100% throughout. Asterisks indicate significant differences from the control (Bonferroni test, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001). Mean RRG of S. alba and L. sativum in water eluates depicted for all four concentrations after 4 days (g, h), 9 days (i, j) and 16 days (k, l). Distilled water was used as control for which the RRG was 100% throughout. Asterisks indicate significant differences from the control (Bonferroni test, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001). Mean RRG of S. alba and L. sativum in water eluates depicted for all four concentrations after 36 days (m, n) and 64 days (o, p). Distilled water was used as control for which the RRG was 100% throughout. Asterisks indicate significant differences from the control (Bonferroni test, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001)

Full size image

According to other studies (Grattan and Grieve 1999; Hu and Schmidhalter 2005), saline conditions with extreme ionic ratios, cause reduced plant growth due to ion toxicities and ionic imbalances influencing biophysical and/or metabolic components of plants. Due to the high pH – values and high amounts of leached ionic species (Märkl and Stephan 2015) in cement leachates, reduced root growth was expected to be observed. However, hardly any of the leachates lead to significantly reduced growth rates for both test species (Fig. 2). Only for 1 day leachate REF-BTW tested with S. alba and PCE-DI tested with L. sativum showed significantly reduced root elongation at a 5% level. It has been reported on a beneficial effect of Ca2+ on the root growth while also alleviation of the adverse effect of salt stress on germination and vegetative growth might happen (Awada et al. 1995; Hu and Schmidhalter 2005). This could be an explanation of the promoted RRG during the bioassays as Ca2+ is constantly leached from the cement paste samples (for further explanation see Appendix Fig. 3). In fact, for most of the leachates, beneficial effects on the root growth could be observed although in most cases without statistical significance (p < 0.05).

S. alba shows reduced RRG towards late-age leachates, which means that the leached organic matter of the first eluates promotes the root elongation of S. alba beneficially. In addition, the counter ionFootnote 1 (potassium) of the superplasticiser is also leached which has to be taken into account as another potential source influencing RRG. However, a possible release of inorganic species from hardened cement paste (e.g. aluminium and chromium) towards late-age leachates may influence S. alba in a more toxic way (Mengel et al. 2001), although no significant difference at a 5% level compared to the control can be found. The root growth of L. sativum tended to show an RRG comparable to the control for the early-age leachates (0.25 and 1 day) whereas towards late-age leachates the RRG is often significantly promoted (16 and 64 days). This might be caused by L. sativum being very sensitive to the leached organic superplasticiser and thus showing RRG around 100% as the superplasticiser is washed from the sample during the first few eluate exchanges.

In fact, some of the chemical elements contained in cement such as calcium, silicon and iron are essential for plants. The possible counter ions of superplasticisers typically sodium and potassium act in the same way as macronutrients and micronutrients that have to be available for healthy plant growth (Mengel et al. 2001; Hu and Schmidhalter 2005). However, cement also contains heavy metals such as nickel, (zinc which is a micronutrient) and lead. These elements may be harmful to plants (Arambašić et al. 1995).

In general, it was not possible to calculate effect-concentration values, e.g. EC50, out of the obtained data from the different dilutions. Down to a dilution of 12.5%, the RRG was promoted, which led to even higher values compared to the control. Either effect-concentration values could only be calculated at higher dilution factors or at higher concentrated leachates. The latter could be achieved through choosing a lower liquid-to-solid – ratio in the leaching test.

3.2 Relative Seed Germination and Germination Index

An ANOVA followed by a Bonferroni post – hoc test evidenced that there are no significant differences (p < 0.05) between dilution steps of the samples and the control when they are evaluated with S. alba and L. sativum seeds in the germination bioassay (RSG). However, the GI, which is the mathematical product of RRG and RSG, shows beneficial effects on both test species often with significant differences at a 5% level compared to the control (Table 5). The order in means of S. alba seeds exposed to the 0.25 day eluates was PCE-BTW < REF-BTW < PCE-DI < REF-DI and that for L. sativum seeds was REF-DI < PCE-BTW < PCE-DI < REF-BTW. All of them promoted the GI, which may be explained by the leached organic matter and the leached ions from the cement at the beginning of the leaching experiment. The order in means of S. alba seeds exposed to the 64-day eluates was REF-BTW < REF-DI < PCE-BTW < PCE-DI and for L. sativum was REF-DI < PCE-BTW < REF-BTW < PCE-DI. This indicates that the samples without superplasticiser exhibit slight toxic effects, with a significance at a 5% level for S. alba, which can be related to the different discharge of organic and inorganic species, the pH – value and the electric conductivity of the samples. Also, a nearly full coverage of the cement particles with superplasticiser can give rise to a protective layer which in turn prevents the release of components from the cement matrix, e.g. heavy metal ions, which may negatively influence the germination. However, the exact mechanisms underlying this process remain undetermined. In-depth investigations on the composition of the eluates and the corresponding effects on relevant parameters like seed germination need to be carried out to get better insight to this behaviour.

Table 5 RSG and GI of S. alba seeds and L. sativum seeds as affected by leachates of cement samples in four different concentrations after 0.25 days and 64 days. Entries marked with the same letter within each type of sample are not significantly different (p > 0.05)
Full size table

Other studies (Fuentes et al. 2004; Alvarenga et al. 2007) investigated different sewage sludge concentrations in soils and water extracts thereof in phytotoxicity tests with species like L. sativum and barley. It was found that L. sativum is more sensitive to toxic effects. In contrast, both species as used in this study were equal in sensitivity and confirmed that the tested cement leachates show only minor phytotoxicity. While Kapustka and Reporter (2009) stated that seed germination is a critical stage in the plant life cycle and could therefore be representative. Dorn et al. (1998) expound that seed germination is limited in estimating the toxicity of substances, due to two reasons. First, various chemicals may not penetrate into the seed and second, the germinating seed receives its nutritional requirements from the internal seed storage materials.

In general, more research needs to be carried out using different methods including statistical correlation methods to get better insight on the effects of cement paste eluates and the included superplasticisers on the environment.

4 Conclusions

The present study shows the effect of cement leachates with and without superplasticiser on white mustard S. alba and cress L. sativum.

In general, it can be concluded that cement leachates promote RRG, RSG and GI of both test species although the promoting effect is only significant in few cases. The cement leachates exert different influences in both test species. The root growth of S. alba is promoted by leachates from the early stage (0.25 and 1 days) while for L. sativum, a promotion is observed towards later ages of the leachates (16 and 64 days). The RSG was not significantly influenced throughout.

From the RRG, RSG and GI data, it was not possible to calculate effect-concentration – values, as all the diluted eluates (from pure eluate 100% down to 12.5%) influence the plant growth in a positive way. As a consequence, no or only slight harmful effects could be observed during this study. In the case of samples containing superplasticiser, a promotion of the GI is indicated. However, this effect cannot be explained from the data obtained during this study.

Further research needs to be done including correlation analysis and multivariate statistics. This applies also concerning the leaching of inorganic and organic constituents and their impact on phytotoxicity which is beyond the scope of this study. To gain better insight towards this aspect, the buffer function of soil needs to be investigated in detail.