Effects of Different Coatings on Concrete Elements Due to Chloride Ion Penetration

Abstract

Reinforced concrete structures are susceptible to chloride ion attack under different conditions, such as water reservoirs, coastal regions, and industrial locations. The physical and mechanical properties of concrete are known to considerably affect the ion penetration velocity. However, studies addressing the effect of coatings on the chloride ion penetration of reinforced concrete are limited. Thus, the objective of this paper is to evaluate the effects of different surface coatings on chloride ion penetration in concrete elements. Acrylic, polyurethane, and epoxy resin coatings were applied in two layers as recommended by the manufacturers. Natural environment chloride ion exposure was conducted in loco in the city of Torres, Brazil, at two marine locations with different geographical characteristics and distances from the sea. In addition, laboratory tests consisting of salt spray and penetration-by-immersion tests were conducted. The concrete’s characteristics, including its compressive strength, water absorption, and void index, were evaluated. The results indicate higher efficiency with the polyurethane coating, while the acrylic resin had the worst results, with a difference of up to 4.5 mm between them.

1. Introduction

In Brazil, studies estimate that CO₂ emissions from the cement industry will reach 66 million tons by 2050 [1]. Considering this projected environmental impact, increasing the durability of reinforced concrete structures would be beneficial, as it would decrease the consumption of cementitious materials over their lifespan for activities such as maintenance or recovery from early structural degradation. The use of more durable concretes is becoming common in most civil construction projects. Concretes are being produced with lower cement consumption and/or the use of supplementary materials to increase their sustainability, extend their durability, and decrease their CO₂ emissions [2,3].

According to standards [4], the durability of a material depends on its capacity to resist effects exerted by the environment. These effects must be predicted during planning so that the safety, stability, and performance of the material are maintained. It was noted that the loss of durability in concrete could be the result of external factors, for example, attacks by aggressive agents, such as chloride or sulfate ions or carbon dioxide [5]. Another durability factor to be analyzed is the quality of the material compared to environmental aggressiveness [6], which is influenced by factors such as its reinforcement cover, water/cement ratio, compressive strength, and porosity.

Aggressive agents, such as chloride ions, create conditions favorable to reinforcement corrosion, which decreases the load-bearing capacity of a structure [7]. The internal alkalinity of concrete creates a protective passive layer over the reinforcement, and corrosion starts from the steel depassivation of this layer [8]. This process starts from a critical chloride concentration in the concrete, which produces a thinning of the protective layer, breaching and eventually depassivating the layer [9]. This process is caused by an increase in the electrical conduction from high concentrations of chloride ions in the concrete [10].

Reinforcement corrosion is electrochemical in nature and occurs in the presence of water through oxidation and reduction reactions. Some researchers [11] found that an electrochemical battery required an anode, cathode, difference in potential, metallic connection between the anode and cathode, and an external closed electrolyte bridge. Severe corrosion damage occurred in the form of iron oxide–hydroxides, which occupied a volume up to 7× greater than the uncorroded metal. It was noted that this volumetric expansion could induce internal tensile stresses of over 15 MPa in concrete [9]. Additionally, corrosion resulted in a considerable loss in the reinforcement cross-section and compromised the load-bearing capacity of the structure or element, increasing the risk of partial or total collapse. Corrosion in reinforced concrete structures, caused by chloride action and carbonation, can cause serious structural damage, a lack of safety, and financial losses [12,13].

Marine environments present a high risk of structural deterioration due to salt spray. It was noted that sea water contained 3.5% (35 g/L) of soluble salts by mass, mostly in ionic concentrations of Na⁺ and Cl⁻ [14]. But salinity levels vary depending on the ocean, with 3.3% in the North Sea, 3.9% in the Indian Ocean, 4% in the Red Sea, and up to 31.5% in the Dead Sea [14]. A salt spray containing these and other elements can be transported by the wind for kilometers beyond the shore, with a decreasing concentration depending on the prevailing wind speed and direction [15]. Therefore, some researchers proposed the following environmental aggressiveness levels for concrete structures, based on their distance from the sea, over a 50-year lifespan: high for distances less than 100 m; moderate for distances between 100 m and 750 m; and negligible for distances over 750 m [16].

It was reported that corrosion from a marine environment was 30× to 40× faster than that in rural environments [17]. Since Brazil has over 7000 km of coastal regions, it is of utmost interest to address the issue of the durability of structures under chloride ion attack. This issue has an additional challenge in that pathological effects of reinforcement corrosion have been more frequently identified in modern structures over the past 40 years.

Chloride ions are found in concrete from internal and external sources. Examples of internal sources are from the use of setting accelerator additives containing CaCl₂ or the contamination of aggregates and mixing water. Examples of external factors are deicing salts, industrial processes, marine environments, or direct contact with seawater [12,18,19,20]. In the case of external sources, it was noted that chloride ingress was mainly from contact with seawater or surface deposition from salt spray through wind action [5]. Chloride ion penetration occurs through pores in the microstructure of concrete and, if the pores are interconnected, it can penetrate deep enough to reach reinforcements [21]. In the exposed environment near the Atlantic Ocean, the time scale for chloride ion propagation through reinforcement covers was much shorter than the projected useful lifespan of structures, with detrimental effects to their durability [22]. Therefore, it is an undeniable fact that chloride ions are the most notable degradation agent in concretes [23].

Chloride ion transport in concrete requires the presence of water [20]. The amount of water present depends on the relative humidity, which, by extension, affects the degradation reactions and the rate of corrosion of reinforcements. The transport mechanisms through pores in concrete are capillary absorption, diffusion, permeability, and ion migration [24]. Capillary absorption is the transport of liquids through capillary pores in the concrete, and it is affected by the pore diameter and connectivity. It was found that the absorptivity contributed significantly to the ingress of harmful substances in aggressive environments [25]. Permeability corresponds to the fluid displacement through pores due to the pressure gradient [24]. However, it was also found that the permeability can be used to contain liquid infiltration and chemical attacks [5]. Finally, ion migration can be understood as the transport of ions in an electrolyte due to forces induced by an electrical field [16]. Migration is the movement of negatively charged substances due to a difference in potential [26].

It was noted that transport mechanisms of aggressive agents in concrete depended on physical and chemical factors such as flowrate, concentration, environmental conditions, structural characteristics, pore size, saturation level, etc. [13]. In the case of chloride ions, concentration also affected ingress and attack profile. It was noted that there was a critical concentration which degraded the thickness of the passive layer, leading to breaching and eventual depassivation of reinforcements [9]. However, there is no consensus in studies and standards with respect to the exact concentration limit of chloride ions necessary to trigger depassivation. For example, standard ABNT NBR 12655 [27] cited a content between 0.05% and 0.40% with respect to the mass of cement depending on the type of structure and exposure environment. On the other hand, another standard, ACI 318 [28] cited contents between 0.06% and 1% with respect to the mass of agglomerate.

It related the critical chloride ion content with respect to the mass of cement necessary to start the corrosion process to ambient relative humidity (RH) and concrete quality [29]. For an RH lower or equal to 50% and good quality concrete, the risk of corrosion was low and an elevated chloride ion concentration would be needed. On the other hand, low-quality concrete in a higher RH environment (above 50%, for example) would allow electrochemical corrosion to start with a relatively low chloride ion concentration. The optimum RH for reinforcement degradation should be between 70 and 80%, since, at higher humidity, the concentration and diffusion of oxygen is lower [5]. Therefore, controlling the quality of the concrete by controlling the chloride content in the raw materials and the w/c ratio, in addition to ensuring a sufficiently thick covering layer by current standards and the environmental aggressiveness class in which the structure is inserted, can help delay corrosion caused by chloride ions [12].

Considering the presence of a pore network and probability of surface cracking, some protective measures could be taken such as the application of an impermeable adhesive coating to ensure a level of concrete durability [30]. This added protective layer would constitute a real defense against aggressive agents and its effect on durability [31]. Different surface-protective measures affect transport mechanisms, durability, substrate properties, economic considerations, and performance [32]. Researchers observed that surface coatings delayed the onset of corrosion from chlorides and reduced the propagation speed of corrosion [33]. This was achieved by reducing water content within the structure of concrete. Other researchers stated that surface coatings could reduce capillary water absorption by 98% and polyurethane coating was the most effective protection coating, with a decrease of up to 86% in the diffusion coefficient of chlorides [34].

Researchers stated that the diffusion rate of chlorides was related to the time needed for the ions to reach a critical concentration that would initiate the corrosion process [9]. Thus, it was concluded that the diffusion coefficient was directly related to the thickness of the cover over reinforcements. This corrosion onset time was used to evaluate several surface protection measures for a 50 mm cover and a polyurethane coating was deemed to have the potential to delay the onset of corrosion. In addition, applying coatings to protect the concrete structure is a more economically viable alternative than repairing or rebuilding a new structure [35].

Thus, the objective of this study was to evaluate the effect of different surface coatings on concrete with respect to chloride ion penetration under practical and laboratory conditions. Practical conditions were in loco with distinct geographical characteristics and distance from the shore at the city of Torres on the north coast of Rio Grande do Sul, Brazil. Simulated laboratory conditions were conducted in a salt spray chamber and an immersion tank. Test samples were surface-coated with acrylic, epoxy, and polyurethane-based resin paints.

2. Materials and Methods

The experimental study was organized in 5 stages, as shown in the flowchart of Figure 1.

Stage I consisted of the production of 36 cylindrical test samples 100 mm in diameter and 200 mm in height and 72 prismatic test samples measuring 40 mm × 40 mm × 160 mm. Stage II applied surface coatings with acrylic-, epoxy-, and polyurethane-based resin paints as instructed by the manufacturers. Stage III conducted chloride ion penetrations under laboratory conditions with salt spray and immersion and practical in loco with distinct geological conditions and distance from shore. Stage IV measured physical and mechanical properties such as water absorptivity, void index, and compressive strength in samples aged 7 days and 28 days. Finally Stage V, analyzed the results.

2.1. Materials

The cement used was CP IV-32, from Cauê Intercement Brasil (Nova Santa Rita/RS), containing up to 50% of added pozzolan as specified in standard NBR 16697 [36] and ASTM C150 [37]. Pozzolanic additions contribute to the gain in compressive strength due to the refinement of pores caused by pozzolanic reactions. Fine aggregate was natural in origin, with a fineness modulus of 1.53 and specific mass of 1.61 g/cm³. Coarse aggregate was basaltic in origin with a fineness modulus of 2.13 and specific mass of 1.36 g/cm³. A fixed water/cement ratio of 0.50 was used and consistency was checked in slump tests and adjusted with the use of a superplasticizer additive to achieve 50 mm to 100 mm in the slump test. The additive was polycarboxylic-ether-based Glenium 51, from BASF (Guaratinguetá/SP). The acrylic coating, from Coral (Recife/PE), was blue paint with 20% water dilution. Epoxy coating, from Brasilux (Guarulhos/SP), was white epoxy with no dilution. Polyurethane coating, from Brasilux, was green paint diluted with a 4:1 paint/catalyst ratio. All coatings were prepared, diluted, and applied in accordance with the manufacturer’s instructions. The coatings used have properties in accordance with NBR 11702 [38], including resistance to wet abrasion, with a minimum abrasive paste of 40 cycles, a minimum dry paint covering power of 5 m²/L, a maximum volatile compound content of 100 g/L, and resistance to fungal growth. Regarding the thickness of the coatings, authors [39,40] claim that surface-protective coatings must have a minimum penetration depth of 2 mm.

2.2. Mix Ratio

Samples were prepared with Portland cement CP IV-32 in a 1:5 ratio, 57% mortar content, and 0.50 water/cement ratio. Superplasticizer additive was used to ensure a slump test result of 10 cm ± 5 cm. The mix ratio is presented in

Table 1. Mix ratio of concrete.

Cement (kg/m3)	Sand (kg/m3)	Gravel (kg/m3)	Water (kg/m3)	w/c	Mortar Content
366.41	886.72	945.34	201.53	0.55	0.57

.

2.3. Test Body Preparation

Test samples were prepared, molded, and cured in accordance with the procedures in standard NBR 5738 [41] and ASTM C192 [42]. Cylindrical test samples, 100 mm in diameter and 200 mm in height, were used to evaluate physical and mechanical properties. After demolding, cylindrical test samples were stored in a climate-controlled chamber kept at 95% ± 5% relative humidity and 23 °C ± 2 °C. Prismatic test samples measuring 40 mm × 40 mm × 160 mm were used for chloride ion penetration tests. After demolding, prismatic test samples were also stored in a climate-controlled chamber kept at 65% ± 5% relative humidity and 23 °C ± 2 °C until the coatings were applied. Surface protection measures were applied to prismatic test samples at 7 days of age with a 2 in (50.8 cm) brush in a room kept at 50% ± 5% relative humidity and 25 °C ± 3 °C. All coatings were applied twice with time intervals and additional procedures in accordance with manufacturer instructions. Eight test specimens were prepared for each type of coating, totaling 24 samples. Figure 2 illustrates samples with the coating already finished.

2.4. Testing Procedures

2.4.1. Physical and Mechanical Properties

Water absorptivity and void index were evaluated at 28 days of age in accordance with standards ASTM C1585-20, ASTM C138, and NBR 9778 [43,44,45]. Compressive strength was evaluated in accordance with standard NBR 5739 [46] at 7 days and 28 days of age, considering three samples for each age. The hydraulic press used was Solotest-brand (Sao Paulo, Brazil), serial number 25,117, with 60 Hz frequency and an applied load of 0.5 MPa/s. Three specimens were evaluated by age. Considering that, even for cement with pozzolanic additions, the compressive strength at 28 days is almost the total value obtained for that mix, the latest age tested was 28 days.

2.4.2. Chloride Ion Penetration by Chamber Testing

Salt spray testing was conducted by standard ASTM B117 [47] in a chamber with a 5% m/m sodium chloride (NaCl) concentration spray in a humidity-saturated environment for 480 h. Chamber-controlled conditions were a flowrate of 1.5 mL/h (±0.5 mL/h), a temperature of 25 °C, pH between 6.5 and 7.2, and density between 1.029 g/cm³ and 1.036 g/cm³. The images were processed and analyzed using Autocad^TM 2022 software. The specimens were measured on the lower horizontal side so that the pictures could be correctly scaled in the software using the Align command. It was decided not to analyze the chloride attack front on the densification face and its opposite side—the flattened face—to rule out possible different behaviors due to the densification of the specimens. Therefore, a 20% distance was considered on each of these sides in the analysis. Using the Polyline command, the regions of chloride ion penetration were outlined, measuring the points of greatest attack and verifying the point that obtained maximum penetration, in mm, using the Line and Distance tools, as shown in Figure 3. Three prismatic samples were subjected to this test for each type of coating.

2.4.3. Chloride Ion Immersion Penetration Test (CIPT)

Prismatic test samples were immersed in a 5% NaCl solution at a volume 7× greater than the total volume of the samples. Samples were subjected to wetting and drying cycles, alternating 7 days immersed with 7 days stored at 23 ± 2 °C and relative humidity of 65 ± 5%. Thus, 2 wetting/drying cycles were applied over 28 days. Three prismatic samples were subjected to this test for each type of coating.

2.4.4. Practical in Loco Testing

Test samples were exposed to natural conditions at the city of Torres/RS for 84 days. Two prismatic samples were subjected to this for each type of coating. According to do Instituto Nacional de Meterologia (INMET), average weather data over this period were a temperature of 23.5 °C, relative humidity of 81.2%, wind speed of 1.63 m/s, and accumulated precipitation of 280 mm. Test samples were placed on an upside-down perforated pallet to prevent water accumulation.

Two locations were selected by the shore. Location A was 50 m from the Southern Atlantic shore at co-ordinates −29.34792 and −49.73008 of latitude and longitude, respectively. There was no interference from buildings, topography, or vegetation so that samples were exposed directly to sea spray. Location B was 550 m from the shore at co-ordinates −29.353481 and −49.735493 of latitude and longitude, respectively. This location offered some geographical protection in the form of rock cliffs and medium-height vegetation. Both locations are shown in Figure 4. The salinity of the Atlantic Ocean is around 37 per mil (g/kg), higher than the average of 35 per mil for the oceans [48]. Other studies have also been carried out in the marine environment [12,19,49,50,51] but under different conditions: different mix, distances from the sea, and surface treatments.

3. Results and Discussion

3.1. Compressive Mechanical Strength

Average compressive strength results of the reference concrete are shown in Figure 5.

Figure 5 shows that compressive strength at 28 days reached Group I—C35 as defined in standard NBR 8953 [48]. According to some authors, this category of concrete was resistant to chloride ion deterioration and recommended for environments of high environmental aggressiveness class (EAC) [52]. As expected, strength increased over time, with the final value at 28 days being 58.4% higher than the initial value at 7 days. It was noted by researchers [53] that the increase in strength over time was related to the hydration level of components in cement, which was responsible for the setting reaction and hardening of the cement/water paste, as pointed out by previous research [51], among others. For the CP IV-32 cement used at later ages, standard NBR 16697 [36] determines a minimum compressive strength of 40 MPa at 91 days.

3.2. Physical Indices

Average water absorption and void index of the reference concrete at 28 days are shown in Figure 6.

The water absorption and void index results of Figure 6 point to a medium-quality concrete, since it could potentially absorb 4.40% of water from the environment and this very water could potentially contain a high concentration of chloride ions [52]. It should be stressed that standard NBR 12655 [27] defined a maximum allowable chloride ion content of 0.15% with respect to the mass of cement in the concrete under exposure conditions. It was observed in a study that concretes with lower water absorptivity were more resistant to chloride migration [50]. This confirmed that water from the environment could contribute to the ingress and transport of harmful agents inside the concrete, with negative effects on useful lifespan. Researchers [52] evaluated concrete mixtures, whose void indices for these mixtures varied between 12.39% and 9.98% for EAC I and IV, respectively. In the case of concrete with the same w/c ratio of 0.50 used in this study, the authors obtained a void index of 9.62% [51], similar to the result of Figure 6.

3.3. Chloride Ion Penetration in Salt Spray Chamber

Test samples with protective coatings were kept in the laboratory salt spray chamber for 44 days. Results of average penetration depth for each type of coating are shown in Figure 7.

Results of Figure 7 show that chloride penetration depth decreased as the coating progressed from acrylic to epoxy and polyurethane. Acrylic paint had a penetration depth approximately 3× greater than polyurethane, while epoxy had an intermediate result being approximately 2.4× greater than polyurethane. Similar results were found, in which greater penetration resistance was also observed in epoxy and polyurethane coatings [54]. Less chloride transport occurred with decreased surface porosity, which, in turn, was related to factors such as solid volume and dry coating thickness. Some researchers highlighted that a good protective performance could be expected from polyurethane, since it tended to form a high-strength film due to polymer crystallization [55].

3.4. Chloride Ion Penetration from Immersion in Saline Solution

Average chloride ion penetration depths in coated test samples subjected to immersion in saline solution are presented in Figure 8.

Results of Figure 8 show the same trend as the salt spray test but with overall slightly deeper penetrations. Acrylic coating was the most susceptible to chloride ion ingress, while polyurethane was the most resistant. Chloride penetration was 1 mm and 1.8 mm deeper for epoxy and acrylic, respectively, with respect to polyurethane. A study evaluated the performance of acrylic and polyurethane coatings with respect to chloride penetration and it was noted that polyurethane had a higher content of solids compared to acrylic so that, when applied with the same number of layers, it produced a thicker film [56]. This was reflected in the laboratory results, with polyurethane-coated concretes acquiring lower chloride content than acrylic ones. However, it should be noted that coating efficiency over time always depended on maintenance, cleaning and reapplication of coatings in the intervals recommended in standards.

3.5. Chloride Ion Penetration from Natural Exposure in Marine Environment

Average chloride ion penetration in samples exposed to an in loco marine environment for 84 days at Locations A and B is shown in Figure 9.

Figure 9 shows that acrylic allowed the most chloride ion penetration when compared to the other coatings. Considering Location A, 50 m from shore, average penetration depths with the acrylic coating were 27% and 140% greater than epoxy and polyurethane, respectively. Polyurethane performance was overall superior, with a reduction in chloride penetrations of up to 58% with respect to acrylic. The overall trends were the same as laboratory tests, albeit with lower numerical differences in between coating types; acrylic coating allowed the most ingress of chloride ions, while polyurethane presented the most resistance to these aggressive agents.

Considering Location B, 550 m from shore, the same trend as laboratory tests and Location A was observed. Acrylic coating allowed the greatest depth of penetration by chloride ions, while polyurethane presented the shallowest penetration. Polyurethane presented reductions in chloride penetration of 45% and 32% with respect to acrylic and epoxy, respectively. However, it is worth noting that these types of polymer-based coatings can release toxic vapors during application, which can be harmful to health. Therefore, given the potential for protecting concrete surface coatings, it is important to develop research into new coating materials with biological and sustainable bases [57].

Comparing both marine locations, the effect of distance from the shore becomes clear in Figure 10 after 84 days of exposure for all coated samples.

Figure 10 shows that chloride ion penetration depths decreased with increasing distance from the shore for the acrylic and epoxy coatings, namely, 19% and 17%, respectively. On the other hand, the polyurethane cover presented a slight increase in penetration of 0.2 mm with increasing distance from the shore. This variation was deemed small enough to be within the margin of error.

A study compared concrete exposure at 50 m and 200 m from the shore and obtained a 41% decrease in peak chloride concentration at the longer distance over a period of 6 months [58]. For 650 m, the decrease was 53.78% with respect to 50 m over the same time period. Another investigation determined an average aggressiveness potential of marine salt spray approximately 8× greater for distances up to 500 m from shore when compared to longer distances [59]. A study evaluated marine environment aggressiveness at distances of 10 m, 65 m, 240 m, and 520 m from shore and determined gradual decreases in chloride surface deposition with increasing distance [60]. Similarly, chloride penetration was compared in concretes with distinct types of cement over distances of 0 m, 500 m, 1000 m, 1500 m, and 2000 m from shore and this confirmed decreases in chloride concentration with respect to the mass of cement for all mixtures [22].

3.6. Comparative Analysis

Figure 11 presents a comparison of laboratory and in loco test results. For in loco tests, exposition periods of 49 days and 84 days were available and the lowest performing results were selected for the comparison as a factor of safety.

Results show that the polyurethane coating outperformed all others in both sets of tests and ensured the lowest chloride ion penetrations. In contrast, acrylic coating was the most vulnerable to chloride ion ingress and allowed the deepest penetration in all tests. These results agreed with other studies [34,61], which linked concrete performance with surface coating and condition. Another investigation also reached a similar conclusion that epoxy resin could seal and protect concrete due to its stability under exposure to varying environmental conditions [62]. The superior performance of polyurethane coating was also in agreement with another study [54], which reported decreased permeation and diffusion of chlorides. Decreases in the diffusion coefficient of chloride ions were also obtained with this coating [9,34,63,64].

Comparing laboratory tests with practical tests, laboratory immersion tests with two cycles of wetting and drying achieved the most chloride ion penetration across all analyses, be they laboratory salt spray or in loco practical tests. Laboratory salt spray tests conducted over 480 h presented deeper penetration for acrylic and epoxy coatings than both practical in loco tests. On the other hand, polyurethane coating under salt spray testing had decreases in penetration of 19% and 24% with respect to in loco tests at Locations A and B, respectively. This result could be attributed to polyurethane degradation due exposure to solar radiation in the in loco tests. Other researchers evaluated polyurethane degradation under ultraviolet (UV) radiation and determined that this type of exposure could lead to weakened adhesion between the coating and concrete substrate [65].

It was expected that laboratory conditions would provide accelerated tests and much deeper chloride penetration. However, that was not the case and laboratory penetrations were not overly greater than in loco tests. Average results over all tests produced penetration depths between 2.2 mm and 7.0 mm, with the deepest penetrations of 6.7 mm and 7.0 mm in the accelerated laboratory tests of salt spray and immersion, respectively.

4. Conclusions

Based on the results, the following conclusions could be drawn for this study:

• Surface protection techniques applied to concrete surfaces could prevent early degradation of structures, increase performance and useful lifespan, and also decrease expenditures in repairs and restoration.
• Laboratory tests indicated that polyurethane offered the most protection against chloride ion penetration, with 3 mm to 4.5 mm lower penetration depth when compared to acrylic and epoxy coatings, respectively. On the other hand, acrylic coating was the most susceptible to chloride penetration.
• Practical in loco tests also demonstrated that polyurethane coating was the most efficient, with decreases of 3.4 mm and 2.1 mm in penetration compared to acrylic and epoxy, respectively. Also, acrylic coating offered the least protection, with increases of up to 140% in chloride ingress.
• Overall, chloride ion penetration decreased with respect to distance from the shore. However, the decreases in this study were not substantial for the tested distances of 50 m and 550 m. It should be noted that wind direction was highly important at locations distant from the shore and this factor was not considered in this study.
• Results were similar between laboratory and in loco tests, albeit with deeper penetrations observed in laboratory tests.