Influence of brick and mortar properties on bioreceptivity of masonry – Results from experimental research

Abstract The effect of mortar and brick properties on the growth of ivy-leaved toadflax (Cymbalaria muralis) and yellow corydalis (Pseudofumaria lutea) has been investigated in laboratory. Different mortar compositions were designed and tested in combination with two different bricks. Highly porous bricks and mortars showed good bioreceptivity; mortars with lime-trass and, in lower extent, those with natural hydraulic lime binder, gave the best results in terms of bioreceptivity. The addition of vermiculite to the mortar was beneficial for plant growth. The brick-mortar combinations most favourable for plant growth were those with estimated low compressive and flexural bond strength values. Proposals are advanced for obtaining a compromise between mechanical strength and bioreceptivity.


Introduction
Traditionally, biological growth on buildings and structures is an undesired event, as this can cause damage to building materials, due to pressure generated by the roots and/or to chemical reactions leading to dissolution of materials components (e.g. [1]). Besides, biological growth is often associated to feelings of untidiness and neglect. The terms ''biodeterioration" and ''biodegradation", used to address biological growth on building materials, reflect this negative connotation.
However, in the last years this attitude is slowly changing: urban ecology gained increasing attention and greening the city has become one of the aims of many municipalities. In particular, walls, and among these quay walls, have the potential to be ecologically engineered to encourage a greater diversity and range of spe-cies [2]. Research has therefore recently focused on the study of engineering solutions and materials to favour biological growth on walls.
The capacity of a wall to act as habitat for biological growth depends on several variables, including wall dimension, construction materials, inclination, microclimate, exposure, accessibility, wall age, sediment and humus and moisture [3,2]. The bioreceptivity of construction materials is one of the important variables. Bioreceptivity can be defined ''the aptitude of a material to be colonized by one or several groups of living organisms without necessarily undergoing any biodeterioration" or as ''the totality of material properties that contribute to the establishment, anchorage and development of fauna and/or flora" [4]. Guillitte further differentiates between ''primary bioreceptivity", which indicates the initial potential of colonization, and ''secondary" and ''tertiary" bioreceptivity which refers to the bioreceptivity of a material following changes in its properties due to biological growth or human action respectively. Another difference is made between ''intrinsic bioreceptivity", which depends mainly on the properties of the material and ''extrinsic" bioreceptivity, in which an exogenous deposit (such as soil, dust or organic particle) substantially modifies the material properties and makes biological growth possible.
Different variables contributes to the intrinsic bioreceptivity of a material: these include moisture transport behaviour (including water absorption, moisture retention, permeability and the related characteristics, such as porosity and pore size distribution), pH, roughness and presence of imperfections such as cracks, voids, fractures etc. [2]. As the presence of water is crucial for biological growth, materials allowing for high absorption and high water retention (i.e. material with sufficient open porosity and a bimodal pore size distribution with both coarse and fine pores) can theoretically provide favourable conditions. Studies confirm that the moisture transport properties of a material (e.g. open porosity, permeability, water absorption kinetics, etc.) can significantly affect biocolonization [5]. A study on growth of algae on clay bricks mentions surface roughness as another important factor [6].
An interesting review of the bioreceptivity of different building materials and of the variables affecting it, is given in [5]. Among building materials, masonry seems to have a better bioreceptivity and supports a larger range of species than monolithic concrete. The growth of plants is generally observed in relatively old masonry, more often in cases where the mortar has (partially) lost its original internal cohesion and its adhesion to the brick. This suggests that presence of (fine) cracks and voids, as well as the roughness of the surface of the masonry units and mortar have a positive influence on biological growth on walls. The pH value is another crucial factor affecting plant growth [5,7]. Plants cannot grow at very high pH, like the one present in fresh, not yet carbonated mortar or concrete. The pH can be up to a certain level adjusted by the mortar composition; a mortar with a high porosity and permeability will carbonate faster; the decrease of pH due to carbonation will positively affect plant growth. Based on the literature, lime-based mortars seem to be more prone for biological growth than cement-based ones [8]. This might be due to their faster carbonation, which is in turn related to their generally better permeability in comparison to natural hydraulic lime and cement-based mortars [9,10]. Moreover, the generally higher water absorption and retention of lime-based mortars in comparison to mortars based on hydraulic binders [10], may contribute as well to their better bioreceptivity, by favouring a sufficient and constant moisture content in the material. These considerations suggest that hydrated lime and, in case a hydraulic binder is necessary, hydraulic lime or a mixture of lime and a pozzolanic material, such as trass powder, could be better alternatives than cement. Besides, some cement types seem to show a higher bioreceptivity than others: for example, magnesium phosphate cement [11] or blast furnace slag cement are shown to have a better behaviour than Portland cement (Ottele, personal communication).
Despite the existence of some studies on bioreceptivity of building materials, literature on the engineering of material compositions is very scarce and mostly limited to concrete [12,13]. To the authors' best knowledge, no specific study on the effect of mortar composition on its bioreceptivity exists. As in masonry biological growth mainly occurs in mortar, elucidating the effects of the mortar composition on its intrinsic bioreceptivity is of crucial importance for favouring bioreceptivity of masonry walls.
The research presented in this paper, carried out in the framework of the European UIA project ''Inclusive Quays", aims at improving the bioreceptivity of brick masonry with the final aim to build nature inclusive quay walls in the Dutch city of Breda. First of all, the effects of components (binder, aggregates, additives) on the physical and mechanical properties of the mortar are assessed; then the bioreceptivity of the mortars in masonry made with two different brick types is evaluated. As the requirements posed to masonry for bioreceptivity may be different, if not opposite, to those of sufficient strength and durability, both these aspects are considered. Based on both the bioreceptivity and estimated mechanical strength of the brick-mortar combination, a proposal for promising brick-mortar combinations to be tested in a following phase is made.  Fig. 1). These bricks have been chosen on the basis of some preliminary water absorption measurements carried out on several brick types, differing in clay type, production process and firing temperature. Bricks with different moisture transport properties have been selected in order to assess the effect of the brick properties on the bioreceptivity of the masonry. Bricks with perforations or a frog have been selected, as these cavities can be used for water or soil storage within the wall construction, probably favouring bioreceptivity. Only bricks declared by the producer as frost-resistant, i.e. within class F2 according to EN 771-1 [14], were chosen.

Materials
The selected bricks have been further characterized in terms of density, porosity, capillary water absorption, pore size distribution and mechanical strength.

Mortar
The composition of the mortars has been determined with the aim of investigating the effect of the following variables on bioreceptivity, moisture transport properties and mechanical strength: -Binder/aggregate ratio: two binder/aggregate ratios have been selected: 1:2 and 1:4 in volume. -Grain size distribution of the aggregate: well-graded quartz sand (CEN Standard Sand EN 196-1 [15]) and a gap-graded sand (obtained by sieving the 1 to 2 mm portion in the CEN standard sand) have been used. -Type of aggregate: quartz aggregate has been used. In some of the mortars, the effect of an addition of expanded vermiculite (Voorzaaivermiculiet, Agra F2, Makkelijke Moestuin) has been investigated; part of the sand was replaced by vermiculite, to obtain a sand/vermiculite ratio 1:1 in volume. Vermiculite is a hydrous phyllosilicate mineral, which undergoes significant expansion when heated. The result is a very porous material, which is often used as a substrate for seed germination. -Type of binder: different binders have been used, either alone or in combination. Hydrated lime (Supercalco 90) has been used in combination with trass in the proportion of 1:1 in volume (AT mortars). Natural hydraulic lime (Saint Astier NHL 3,5) has been used as only binder (H mortars) and in combination with blast furnace slag cement CEMIII-b (HC mortars), in the proportion of 1:1 by volume.
Next to self-mixed mortars, two ready-to-use mortars, representative of basic mortar types often used in the Netherlands in masonry walls, have been tested as comparison: Remix Masonry mortar without lime (MMzK) and Remix Masonry mortar with lime (MMmK). According to the information provided by the seller of the product, both mortars are based on Portland cement, with (MMmK) or without (MMzk) the addition of about 5% (as weight % of the dry mix) of hydrated lime; the aggregate is a quartz sand (with a well-graded grain size distribution (max grain size 4 mm) with the addition of a small percentage (4-6%) of limestone powder; an air-entraining agent is used as additive.
In order to speed up the growth of plants, seeds of ivy-leaved toadflax (Cymbalaria muralis) and yellow corydalis (Pseudofumaria lutea) were added to some of the mortars. For each mortar type, specimens with and without seeds were prepared. Table 2 summarizes the composition of the mortars used in this research.
The water content of the fresh mortar was determined as that sufficient to obtain a spreading of 160 ± 5 mm, measured using a flow table, according to the standard NEN EN 1015-3 [16].
The water content of the fresh mortars is reported in Fig. 2. As the water content for mortars of the same composition, with and without seeds, is identical, a single value is reported. It is clear that the presence of vermiculite increases the water demand considerably.
For each mortar type, two different series of specimens were prepared, to be used for different tests: -Slabs with a size of about 210 Â 100 Â 20 mm: For each mortar type, 2 slabs were prepared. These slabs were prepared on bricks of type 2 (B2) and detached from the brick after 7 days of curing. In order to facilitate the detachment of the mortar slab without affecting the moisture transport between brick and fresh mortar, a plastic net was used between the mortar and the brick. These specimens were used for the measurement of the water absorption. -Mortar prisms of the size of 160 Â 40 Â 40 mm: these specimens were prepared in polystyrene moulds. For each mortar type, 3 prisms were prepared. These specimens were used to measure the compressive strength of the mortars and the carbonation depth.
The mortar specimens, both slabs and prisms, were cured according to the standard NEN-EN 1015-11 [17] They were stored for a period of 7 days (of which the first 2 days inside the mould) under plastic; afterwards, they were stored in a room with a RH of 65 ± 5%, until the moment the characterisation tests were carried out.

Brick-mortar combinations
While microorganisms such as algae and mosses can grow on brick and natural stones, plant growth (which is the object of this work) generally occurs in mortar joints. Therefore, assessing the bioreceptivity on mortar specimens could have been an option. However, as moisture content in the substrate is one of the most relevant variables affecting plant growth, and this is affected by the moisture transport properties of the brick/mortar combination [18], the use of brick/mortar specimens has been preferred.
All prepared brick-mortar combinations are reported in Table 3. For each combination, both specimens without and with seeds mixed in the mortar were prepared. One brick-mortar specimen was prepared for each combination. The code of the brick-mortar combination is given as follow: mortar type -brick type -without (ref) or with seeds (seeds), e.g. ''Hst2-B2-seeds" is built by mortar with NHL binder, 1:2 binder-aggregate ratio and well-graded sand with the addition of seeds; the mortar is combined with a brick type B2.
Each specimen consisted of 2 bricks and two mortar joints. Mortars were prepared as described in section 2.1.2; bricks were prewetted prior to the application of the mortar. The surface of the mortar joint was not smoothed, but raked rough, in order to favour bioreceptivity. Once prepared, brick-mortar specimens were stored under plastic for 7 days and then placed outside (autumn 2019). A polystyrene structure was designed to position all stacks with a slope of 20 degrees, lifted from the ground (Fig. 3). In order to provide all specimens similar exposure conditions, they were protected from rain (which could reach the stacks in different amounts depending on their location) but sprayed with water at regular time intervals.

Characterization tests on brick
The water absorption by capillarity of the bricks was measured according to EN 772-11 [19]. Measurements were carried out in threefold. The average water absorption coefficient (WAC) and the initial rate of absorption (IRA) of the bricks have been calculated. The WAC is the slope of the initial, linear part of the absorption curve; the IRA is the amount of water absorbed per unit of area in the first minute of the test. After saturation by capillarity at atmospheric pressure, the density and porosity of the bricks was assessed according to the following procedure [20]: the bricks were immersed in water for one week; then the saturated weight in water and in air was measured and their density (D, kg/dm 3 ) and porosity (P, vol%) calculated as follows: where: m d = mass of the dry brick m a = mass of the saturated brick in air m w = mass of the saturated brick in water and where 2650 kg m À3 is the density of a stone-like material with no porosity [20].
The compressive strength of the bricks, having length l u , height h u and thickness t u , was determined in agreement with EN 772-1. [14] The test was carried out through a displacement-controlled apparatus including a hydraulic jack with 300-ton capacity. The  For both types of bricks tested, the rate of the jack displacement was set to 0.01 mm/s in order to reach the maximum force in not less than 2 min. Tests have been carried out on 6 specimens for each type of brick. The compressive strength of the masonry unit f b * is determined as follows: where F max is the maximum force, l u and t u are the length and thickness of the masonry unit, respectively. Following the Annex A of standard EN 772-1 [14], the normalised compressive strength of the masonry unit f b is determined as: where d is the shape factor determined in agreement with Table A.1 of the standard EN 772-1 [14], that for the studied bricks is equal to 0.755 ± 0.005.

Characterization tests on mortar
The water absorption by capillarity of the mortar was measured according to the same procedure used for the brick. In this case, specimens of the size 50 Â 50 Â 20 mm were cut from the mortar slab and the water absorption was measured through the 50 Â 50 mm surface, originally in contact with the brick. When the mortar specimens reached a constant weight, they were immersed in water, and their porosity and density were determined, according to the same procedure as used for the brick (see section 2.2.10 [20]). For each mortar composition, only mortars without the additions of seeds were measured. All measurements were carried out in threefold.
The compressive strength of the mortar prisms (with and without seeds) was measured according to the procedure described in the standard NEN EN 1015-11 [17]. The rate of loading was selected in such a way to fulfil the requirements established in the standard. For each mortar at least 6 specimens were tested.
The mortar pH is crucial for plant growth. The pH of a mortar is largely determined by the occurrence of carbonation. The pH of a fresh mortar is about 13, too high for plant growth. After carbonation, the mortar pH becomes lower (about 9) and some microorganism and plant species may grow. In order to assess whether the pH of the mortar was low enough to allow for plant growth, a solution of phenolphthalein in ethanol was sprayed on the two halves of freshly broken 4x4x16 cm specimens before the execution of the compressive test. Phenolphthalein is a pH indicator: when the colour of the mortar turns to pink, it means that the pH is higher than 9 and thus that the mortar is not carbonated yet; if the mortar is carbonated, its colour will not change.

Monitoring of plant growth
The plant growth in the brick-mortar specimens was visually and photographically monitored at different time intervals. Specimens were removed from the outdoor exposure to be photographed. A special set-up was developed in order to always ensure the same light conditions. In order to facilitate monitoring and comparison between walls, a qualitative scale was created to evaluate the plant growth.

Moisture transport properties
The water absorption curves of brick 2 and 8 are reported in Fig. 4. Density, porosity, WAC and IRA values are summarized in Table 4.
The results show that brick 8 has a much slower water absorption rate and a lower capillary absorption and open porosity than brick 2. The slow water absorption suggests that brick B8 has smaller pores than brick B2. The differences in moisture transport properties are in agreement with what was expected based on the production process of the bricks. Soft mud moulded bricks, such   as brick B2, are generally made with lean clay mixtures (i.e. low clay content) with generally high water content, in order to facilitate the shaping process; this results in a quite high porosity of the bricks. In the case of extruded bricks, such as brick B8, generally fatter clay mixtures are used, often in combination with a vacuum pump in the extruder; this leads to brick with a lower porosity. Table 4 reports the compressive strength f b * and normalised compressive strength f b for B2 and B8 (average of 6 specimens for each type). The compressive strength of brick B8 is much higher than that of brick B2, as expected based on the physical properties and the production process.

Density, porosity and water absorption
The density and porosity of the mortars are reported in Table 5.
Based on the total porosity, it can be concluded that: -The binder/aggregate ratio and the grain size distribution of the aggregate have a major influence on the total porosity. The use of a lower binder/aggregate ratio and of sieved sand leads to a higher porosity in the mortar. -The binder type has a minor influence on the total porosity of the mortar. -The mortars with an addition of vermiculite show the highest porosity, because of the high porosity of this material. -The ready-to-use mortars have very high porosity, comparable to that of mortars with vermiculite. The high values are most probably due to the presence of an air entraining agents.
The capillary water absorption curves of the mortar specimens are reported in Fig. 5. Based on the water absorption curves of the mortars, the following conclusions can be drawn: -The binder has a strong influence on the water absorption rate: among the self-mixed mortars, the lime-trass mortars show the fastest capillary absorption, the cement-based mortars the slowest. These results are consistent through all the mortar specimens, independently from binder to aggregate ratio and grain size distribution of the aggregate. -A smaller binder/aggregate ratio lead to a higher porosity and slightly faster water absorption rate; this is most probably due to the presence of more intergranular pores, not fully filled by the binder, in mortars with a lower binder content. -The mortars made with gap-graded sand have higher porosity and faster water absorption than those made with wellgraded sand. These results are in agreement with the expectations, as the use of gap-graded sand leads to coarser pores (for an equal binder/aggregate ratio). -The ready-to-use mortars, despite the cement binder, have a very fast rate of absorption; the type containing hydrated lime (MMmk) has the highest rate of absorption of all mortars. Based on the information on the composition of the mortars provided by the producer, a lower rate of absorption would be expected than actually measured. While the high porosity of these mortars can be related to the presence of an air entraining additive, which leads to round, coarse pores in the hardened mortar, the fast rate of capillary absorption can hardly be explained. The high absorption rate measured suggests that a well-connected network of coarse pores is present. However, pores created by air entraining additives are generally only connected through the fine porosity present in the binder (e.g. [21]) and have therefore no positive effect on the capillary absorption rate. Microscopy observations on thin sections could provide more information on the pore network of these mortars and clarify this issue.

Carbonation depth
The pH of the mortar, and thus indirectly their carbonation depth, has been indicatively assessed by means of phenolphthalein. The carbonation depth measured on mortar prisms after 28 days curing (Table 5) varies significantly. There are large and not always consistent differences between mortars with and without seeds, which are hard to explain. In general, the mortars made with air-hardening lime and trass (ATsf2, ATst2, ATst4) show the largest carbonation depth among the self-made mortars; also the ready-to-use mortars show deep carbonation. These results are most probably related to the high porosity and capillary absorption rate (thus, likely a high permeability) of these ready-to-use mortars (see section 3.2.1).

Compressive strength
The average compressive strength of the mortars after 28 days curing are reported in Table 5. Based on these results it can be concluded that the binder type has a strong influence on the mechanical strength of the mortar, for similar values of total porosity (Fig. 6). Self-mixed mortars prepared with hydrated lime and cement (HC mortars) have the highest compressive strength; pure NHL mortar and air-hardening lime-trass mortar have similar compressive strength values. Ready-to-use mortars (MMmk and MMzk) have low compressive strength, despite the presence of a cement binder: these low values can be attributed to the very high porosity of these mortars.
The binder/sand ratio is another parameter to have a substantial effect on the mechanical strength of the mortar (Fig. 7). This result could be expected as a low binder/sand ratio corresponds to the high porosity of the mortar and thus to a lower strength.
The effect of the grain size distribution of the sand on the mechanical strength is less important: in mortar with 1:4 binder/ sand ratio, a gap graded sand (between 1 and 2 mm) leads to a lower strength than a well-graded sand (Fig. 8); however, in the case of mortars with a 1:2 binder/sand ratio, the strength is not significantly affected by the grain size distribution of the sand.
When considering the effect of the additions, vermiculite (half of the volume of the aggregate) leads generally to a decrease of the compressive strength; this is more evident in the case of the mortars with a binder/sand ratio 1:2 than in those with 1:4 ratio (Fig. 9).
The addition of seeds (about 450 seeds, equivalent to 1,7 g) did not cause a significant change in the mortar strength; in most cases, a slightly lower mechanical strength was measured in the presence of the seeds.

Estimation of mechanical strength of brick-mortar combinations
To select brick-mortar combinations suited for the construction of the masonry quay wall, the compressive and flexural bond strength of masonry is estimated and compared with a dataset of mean mechanical properties for Dutch masonry reported in NPR 9998:2018 [22].
Based on the compressive strength of an individual materials (brick and mortar), an indication of the compressive strength of the masonry was obtained making use of the procedure reported in Eurocode 6 [23] and EN 1052-1 [24]. For the type of masonry units and mortar thickness considered, the following formulas have been applied to obtain an estimation of the compressive strength of the masonry f' m : where: f k , f b and f m are, respectively, the characteristic compressive strength of masonry, the mean normalised compressive strength of bricks and the mean compressive strength of mortar expressed in N/mm 2. ; K, and, = constants, equal to 0.55, 0.7 and 0.3 for clay masonry units Group 1 with general purpose mortar.
The flexural bond strength of masonry f w was estimated by adopting the formulation proposed by [25]:  A minimum and a preferred lower limit were defined for the strength of masonry on the basis of, respectively, the properties of Dutch masonry built before and after 1945, as reported in NEN-NPR 9998:2018 [22]. The values reported in the standard were obtained based on an extensive experimental campaign for the characterisation of existing residential masonry buildings [26,27] and on literature data for the Dutch masonry (e.g. [28,29]) Fig. 10 shows the relation between the estimated compressive and the flexural bond strength of masonry, together with the two lower limits, for each mortar-binder combination. For both brick types, the estimated strength value approaches or overcomes the lower limits only in the case of mortar with a binder made of natural hydraulic lime and cement (HC mortars). All the other brickmortar combinations, including the ones built with the ready-touse mortar, show very low estimated strength values. In case of the brick-mortar combinations built with brick B2, the estimated masonry properties are below the limit for both the compressive and the flexural bond strength. On the contrary for masonry with brick B8, a sufficient masonry compressive strength is expected independently of the type of mortar.

Bioreceptivity of brick-mortar combinations
After 3 months exposure, plant sprouts are observed to grow mostly in mortars in which seeds where added during mixing. These are all plant sprouts of ivy-leaved toadflax (Cymbalaria muralis) and yellow corydalis (Pseudofumaria lutea), of which the seeds were added during mortar preparation. In a few cases the growth of plant sprouts has been observed in mortars in which no seeds were added (Hvt4-B2, Hvt4-B8 and HCvt4-B2). In these cases, it is supposed that seeds have been transported by wind. It is interesting to notice that all these mortars have a vermiculite addition.
The specimens made with air lime-trass mortar (with 1:4 binder/sand ratio) in combination with brick B2 shows the most developed plant growth: several plant sprouts have grown on the mortar and they are still alive after more than 3 months (Fig. 11).  The results of the plant growth from seeds after several weeks of exposure are summarized in Table 6. Based on these results, the following conclusions can be drawn: -The moisture transport properties of the brick/mortar combination have an important influence on the growth of the plants in the mortar: in all cases but one, plants have grown in mortar joints of masonry made with brick type B2. This brick has a higher open porosity and faster absorption than brick B8 (suggesting the presence of coarser pores): the brick works therefore as a kind of water reservoir for the mortar, which, thanks to its finer pores can extract water by capillarity from the brick and retain it long enough for the plants to grow. -The binder has a strong influence on plant growth. Air-lime/ trass-based mortars are the best-performing ones; the NHL/ cement-based mortar shows the lowest performance (no growth is observed). -No growth is observed in ready-to-use mortars, despite their high porosity and favourable pH. A possible reasons of this behaviour might be the composition of these mortars, e.g. the presence of additives and/or the hydration products of the cement binder. Another possible explanation might be related to the low water retention of these mortars. Despite this property was not measured in this research, hypotheses can be made based on the water absorption curves. The water absorption curves of these mortars have two clear branches, (an initial fast, linear absorption followed by a flat line) suggesting a pore size distribution with mostly coarse pores. Based on these observations, a fast drying and a low water retention can be expected for these mortars. The fast drain of water would have a negative effect on plant growth. -The aggregate/binder ratio and the grain size of the aggregate seem to have limited influence on plant growth -The addition of vermiculite has a strong positive effect on plant growth.
It should be mentioned that, at this stage, it is likely that the plant sprouts are relying on the seeds for the nutrients. Only in a later stage it will be possible to assess whether sufficient nutrients are provided to the plant by the soil and/or the mortar.

Discussion and conclusions
The research has highlighted that the growth of the selected plant species is possible in relatively fresh mortars, provided the conditions are favourable enough for this. Some plant sprouts were observed to grow on the mortar with mixed-in seeds in the 3 months after preparation.
The effect of brick and mortar properties on bioreceptivity of masonry walls has been elucidated. The moisture transport properties of the brick-mortar combination and the composition of the mortar (mainly binder type and additions) were proven to be the most relevant factors influencing and plant growth.
Despite plant sprouts grow mainly in the mortar joint, the brick properties were shown to be crucial to determine bioreceptivity of the mortar. Brick B2 was proven as most favourable for plant growth. This brick has a higher open porosity and faster absorption than brick B8 (suggesting the presence of coarser pores): the brick works therefore as a kind of water reservoir for the mortar, which, thanks to its finer pores, can extract water by capillarity from the brick and retain it long enough for the plants to grow.
Similarly, mortars with a higher rate of capillary absorption and higher porosity were shown to have better bioreceptivity. Only exception were the ready to-use mortars which, despite their high porosity and rate of capillary absorption, were not prone to biological growth. This might be related to their composition (cement binder and/or additives) and/or to their supposed low water retention. Further investigation of their pore system (e.g. by microscope observations on thin sections) and drying behaviour could clarify this point.
The binder type has an important role in favouring plant growth: mortars with lime-trass and, in lower extent, those with a natural hydraulic lime binder, perform the best in terms of bioreceptivity. The addition of vermiculite, a light-weight aggregate, to the mortar has been shown to be very beneficial for plant growth. This is most probably due to its high porosity, which can work as water reservoir and to its high ion exchange capacity, which provides a buffer for the supply of nutrients to the plants [7].
When considering the mechanical strength, unfortunately, the brick-mortar combinations most favourable for plant growth are those with estimated low compressive and flexural bond strength values. Therefore, in the next phases of the research and for the final practical application, a compromise needs to be found: the masonry should be sufficiently strong to withstand the applied loads, but still able to favour plant growth on its surface.
Based on these considerations, the following choices have been made for the next phases of the research: -In addition to masonry with tooled bedding mortar joint, masonry with joints including bedding and pointing mortars will be tested: the bedding mortar confers the necessary strength to the masonry, while the pointing mortar provides an optimal bioreceptivity. HCst2 and Hst2 are selected as bedding mortars for their higher strength; ATst2 mortar as pointing mortar being favourable for plant growth. -Brick B8 was shown to not be suitable for plant growth and thus not considered in further research. Next to brick B2, a different type of brick, with a higher compressive strength than B2, but still a high porosity and water absorption rate will be considered.
At the moment both above mentioned solutions are tested in the laboratory and on site.
Another point which would need to be considered in the next step of the research is the durability of the mortar with mixed-in plant seeds. When seeds germinate and roots develop, these might damage the mortar, with possible negative effect on the strength of the masonry. To tackle this issue the following solutions can be considered: -as mentioned above, the bio-receptive mortars can be used only as pointing mortars; therefore, damage would have very limited effect on the strength of the masonry. Moreover, the roots of the selected plants are relatively weak and hardly able to generate pressures high enough to affect sound, relatively strong mortars, such as the HCst2 cement-lime bedding mortar. -a dry-stack masonry system can be used, in which bricks are assembled using plastic connection elements, leaving space free in between the bricks for mortar or compost [30]. This way, the mechanical strength of the masonry does not relay on the mortar joint and its interaction with the brick, and the roots growing in the joint would not constitute a problem.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.