Desalination 433 (2018) 164–171
Contents lists available at ScienceDirect
Desalination journal homepage: www.elsevier.com/locate/desal
Preparation of high-performance graphene nanoplate incorporated polyether block amide membrane and application for seawater desalination
T
Filiz Ugur Nigiz Kocaeli University, Chemical Engineering Department, Kocaeli 41380, Turkey
A R T I C L E I N F O
A B S T R A C T
Keywords: Graphene nanoplate Nanohybrid membrane Pervaporative seawater desalination Seawater purification Polyether block amide
In this study, a novel graphene nanoplates (GNPs) incorporated polyether block amide (PEBA) membrane was successfully prepared to be used for seawater desalination. Pervaporative desalination performances were performed in a temperature range of 35–65 °C. Effects of GNPs content in PEBA, membrane thickness, and temperature were evaluated in terms of the flux and total salt rejection. The long-term stabilities of the membranes were experimented. Incorporating graphene nanoplate into PEBA matrix enhanced flux and ion rejections simultaneously. Experimental stability of the membrane was improved by GNPs incorporation. Graphene incorporated membranes showed excellent seawater desalination performance with the salt rejection of > 99.89% and flux of > 2.58 kg/m2·h. Based on the flux and rejection results, optimum GNPs contents in PEBA matrix were observed as 2 wt% GNPs and 3 wt% GNPs. Increasing temperatures improved the water flux and did not significantly affect the salt rejection. The highest salt rejection was obtained as 99.94% with a flux of 2.58 kg/ m2·h at 35 °C when the 3 wt% GNPs incorporated membrane was used. The nanohybrid membrane preserved 99.8% of its performance during 60 h, while the rejection performance of the pristine membrane decreased to 96.8%.
1. Introduction Demand for freshwater shows a rapid increase due to the non-stable population, industrialization and urbanization growth. When the impact of climate change and global warming on depleting of the freshwater source are taken into account, it will not be surprising that the water scarcity will cause global problems in a very close future. Water production from the natural water source is one of the best solutions to meet the future's water need. Desalination is a process that produces freshwater from the natural water reservoir by removing hydrated ions, salts, heavy metals, and the other constituents [1–3]. Conventional desalination technologies are classified into two main groups as thermal and membrane-based systems. Owing to the lower energy consumption and fewer equipment requirements, membrane-based technologies such as reverse osmosis (RO) seems to overcome both economic and environmental problems of desalination technologies [4–7]. Besides the well-known and commercialized membrane desalination technologies, relatively new and emerging processes such as membrane distillation (MD) [8,9] and pervaporation (PV) have attracted researcher's interest for deep purification of saline water. Recently, pervaporative desalination has been considered to be a promising and emerging technology depending on its energy-saving process property [10–13]. Pervaporation is a chemical potential driven process. The separation
E-mail address: fi
[email protected]. http://dx.doi.org/10.1016/j.desal.2017.08.025 Received 22 June 2017; Received in revised form 28 August 2017; Accepted 30 August 2017 Available online 05 September 2017 0011-9164/ © 2017 Elsevier B.V. All rights reserved.
in PV is directly related to the pressure difference across the membrane. The major difference between PV and RO is the vapor phase presents in the permeate stream of pervaporation. Since a non-porous selective membrane is used in pervaporation, the membrane is able to reject a wide variety of contaminants including heavy metals and hydrated ions having small kinetic diameters. By enabling the rejection of almost all impurities in seawater, > 99.5% salt rejections have been achieved using PV [14–20]. Recent studies have shown that PV is effective to remove many of chemicals including chloride, bromide, and boron where RO cannot purify in a single stage [16]. The main drawback of using pervaporation in commercial scale is insufficient flux obtained during the separation. Since the membrane is the main constituent of pervaporation system, scientific studies have been performed to develop water-selective membranes having excellent separation capability. Organic polymeric membranes are suitable for desalination applications which pose good film forming ability and ease of modification; however, they still suffer from structural problems that lead to a decrease in desalination performance. Although many of inorganic membranes (zeolite, graphene or clay) have been reported in the literature, the higher production cost of inorganic membrane limits the commercial application. Over the past decade, inorganic particle incorporated organic membranes, in other words, mixed matrix, composite or hybrid membranes are gaining attendance based on their
Desalination 433 (2018) 164–171
F.U. Nigiz
2. Materials and methods
higher permeability, and long-term thermal-mechanical stability. The major factors that directly affect the separation performance of the membranes are the problems related to the particle dispersing within the membrane, structural defect in the membrane, poor polymer-inorganic interactions, and the surface properties of the particles. In the most of the reported pervaporative desalination studies, hydrophilic non-porous membranes have been used to enhance water flux and rejection factor simultaneously. Xie and co-workers [18,19], reported two studies by preparing a polyvinyl alcohol/maleic acid/silica (PVA/MA/Si) membranes for the pervaporative desalination of the NaCl-water solution. In the first study, a flux of 11.7 kg/m2·h and rejection of > 99.9% were obtained. Differently from the first study, the influence of membrane thickness was investigated. In the second study, water flux and the rejection were achieved as 6.93 kg/m2·h and 99.5% respectively. An et al. [21] synthesized a clinoptilolite-phosphate membrane and performed pervaporation experiments using different concentrations of NaCl-water solutions. As a result, they obtained a flux of 15 kg/m2·h accompanied with final rejection above 95% at 95 °C. Cho et al. reported a desalination experiment using pervaporation technique by preparing NaA zeolite membrane [15]. They pointed out that they found a rejection of 99.9% with a flux of 1.9 kg/m2·h. Zhu and co-workers [22,23] fabricated MFI zeolite-based membranes and they achieved salt rejection as 80% and 93% in different two studies respectively. Most recently, nano-sized inorganic materials or fillers have also been studied for desalination purposes. Besides them, graphenebased materials are more attractive. Graphene is a two-dimensional carbon-based material. Due to its unique properties, it is used as functional filler in a wide variety of industries from electric industry to automotive. Graphene-based hybrid materials such as nanohybrid polymers provide good mechanical strength, electrical conductivity, and thermal properties. Owing to its precise sieving properties and small thickness, graphene layer restricts the diffusion of the small molecules through its structure [24–26]. Depending on its selective separation capability, it is considered a good candidate to be used for filler within a selective membrane. Graphene and graphene oxide membranes have been used in the many of membrane processes including gas separation, nanofiltration, and reverse osmosis [27–30]. In the earlier literature studies, graphene nanosheet derivative pervaporation membranes were fabricated to enhance both flux and separation factor of the target component [31–33]. Recently, graphene has been used for water desalination [24,26,34–36]. It is reported that the separation performance and lifespan of graphene incorporated polymeric membranes significantly increase due to the anti-microbial properties of graphene [37]. Investigations also show that graphenebased membranes may be more effective to retain salt ions compared to the commercial reverse osmosis membranes [38]. In this study, graphene nanoplates (GNPs) incorporated polyether block amide (PEBA) nanohybrid membranes were fabricated to desalinate seawater. Differently from the literature, a hydrophobic matrix (PEBA) matrix was used and graphene was incorporated to enhance the selective separation capability of the membrane. In the literature, researchers have mostly fabricated mixed matrix membrane by incorporating hydrophilic fillers into a hydrophilic polymer. Hydrophilic polymers may exhibit unstable desalination performance due to the high swelling degree. Therefore, using a hydrophobic polymer as the matrix may improve the stability of the membrane during desalination. According to the author's knowledge, this is the first study on using of graphene nano-plate loaded PEBA membrane for desalination application. Due to the high aspect ratio and high surface area of GNPs, they exhibit good compatibility with polymers [39]. Based on its anti-fouling property, it resists the biological fouling of the membrane in seawater media. To take these advantages of graphene, different amounts of graphene particles were incorporated into PEBA matrix. Regarding GNPs incorporation, it was expected to improve flux, rejection results and long-term stability of the PEBA based membranes.
2.1. Materials Polyether-block-amide (Pebax 2533) which is a block copolymer contains 80 wt% poly (ethylene oxide) and 20 wt% polyamide was kindly obtained from the distributor of Arkema, Turkey. Graphene nanoplates with a surface area of 750 m2/g and acetic acid (analytical grade) were purchased from Aldrich Chemicals, Turkey. The seawater was taken from the North coast of Marmara Seawater in Turkey. 2.2. Membrane preparation PEBA-graphene nanoplate membranes with different ratios of GNPs were fabricated using phase inversion technique. Firstly, a PEBA-acetic acid solution with the polymer concentration of 10 wt% was prepared and stirred at 60 °C for 5 h until a homogeneous polymer solution was obtained. Separately, GNPs were dispersed in acetic acid for 15 min using a sonic mixer. The polymer solution was then added to the acetic acid-GNPs solution and further stirred at room temperature for 2 h at a stirring rate of 500 rpm. Following the homogenization, the solution was cast on to a Teflon plate and allowed to evaporate all acid at the room conditions. After the evaporation, flat-sheet membranes formed with an average thickness of 120 μm ± 5. 2.3. Membrane characterization Morphological properties and the homogeneous phase continuity of the pristine and nanohybrid membranes were analyzed by means of scanning electron microscope (JEOL JSM-6335 F). Before the analysis, membrane samples were covered with platinum. For the cross-sectional analysis, samples were fractured using liquid nitrogen. The influence of GNPs addition on the surface's hydrophilicity of PEBA matrix was investigated by means of contact angle measurements (Attension KVS) using the Sessile Drop Method. For the measurement accuracy, each test was repeated for three times. The effects of temperature on weight losses of the pristine and nanohybrid membranes were analyzed using thermogravimetric analysis (TGA) (Mettler Toledo) in the temperature range from 25 °C to 600 °C at a heating rate of 10 °C/min. The influence of GNPs addition on the degradation temperatures of the polymeric matrix was examined by means of differential scanning calorimetry (DSC) (Mettler Toledo). For DSC analysis, membrane samples were cut into a constant diameter (3 mm). Then, samples were weighed and encapsulated in aluminum pans. DSC experiment was performed with the temperature rising from − 60 to 200 °C at a heating rate of 20 °C/min under nitrogen atmosphere [40]. 2.4. Analysis of permeate water After the pervaporative desalination experiment, the ion concentration of the desalinated water was determined using inductively coupled plasma mass spectrometer (Perkin Elmer Elan DRC-e ICP-MS). The chloride (Cl−) amount in the desalinated water was measured using Hach Dr. 5000 UV–Vis spectrophotometer. The total conductivity of the desalinated water was measured using a Mettler Toledo Seven Compact S230 model conductivity meter with an accuracy of ± 2 μs/ cm. Resistances of membranes before and after desalination were recorded using VersaStat 3 Potentiostat Galvanostat. 2.5. Seawater desalination experiment Desalination experiments were performed using laboratory-scale pervaporation unit as seen in Fig. 1. The experimental desalination unit consists of a stainless steel membrane cell equipped with a mechanical stirrer and thermocouple. 165
Desalination 433 (2018) 164–171
F.U. Nigiz
PEBA is a block copolymer that consists of soft polyether (PE) and hard polyamide (PA) segments. Therefore, different melting points are seen in Fig. 3. Whereas the peaks around 10 °C and 58 °C are attributed to melting transition of PE, the peak in the region of 139 °C corresponds to PA segment. The first clear observation that is found from the spectra of the nanohybrid membrane is the decrease in the first peak's area (around 10 °C). Also, a disappearance in the second peak's area (around 58 °C) of PE segment is observed. Although a significant shift in the first crystallization temperature of PEBA is not seen (from 11.84 °C to 11.25 °C), the melting enthalpy of the nanohybrid membrane is found as 150.98 mJ. The melting enthalpy of the pristine membrane is also seen as 164.24 mJ. This result indicates that the total crystalline structure distribution within the hybrid membrane reduced because of GNPs addition. During the membrane forming process in preparation, GNPs would spontaneously embed into PEBA and lead to a slight decrease in the total crystalline structure. The similar observation was also reported by Choudhari et al. [41]. TGA spectra of membranes are indicated in Fig. 4. The figure confirmed that GNPs addition improved the thermal stability of PEBA membrane. Both the first thermal degradation temperature and remain amount of membrane after the experiments increased because of GNPs addition. A similar observation was also reported in the literature [41]. Fig. 5 shows the contact angles between the water and surface of the pristine, 2 wt% GNPs, and 5 wt% GNPs incorporated PEBA membranes. The highest contact angle was observed as 86° on the surface of the pristine membrane. A slight decrease from 86° to 75° was seen when GNPs incorporation increased. 80° and 75° water contact angles were measured on the surface of the 2 wt% GNPs and 5 wt% GNPs incorporated PEBA nanohybrid membranes respectively. These results indicated that the surface hydrophilicity of the membranes enhanced with GNPs dispersion within PEBA. During the evaporation process of the membrane, graphene particles would migrate towards to the top surface of the membrane depending on the differences in the density of PEBA and GNPs; therefore, the surface hydrophilicity enhanced [42]. In accordance with these results, an improvement in the water flux of nanohybrid membranes would be expected depending on GNPs incorporation. Owing to the increased surface hydrophilicity of the membrane, it was also expected to obtain more consistent rejection and flux value by nanohybrid membrane usage.
Fig. 1. Pervaporative desalination scheme.
The cell was put inside an oven to provide different temperatures. The effective separation area of the membrane was 19.625 cm2. During the experiments, the feed was kept at atmospheric pressure and 5 mbar vacuum was applied on the downstream side of the membrane. Desalination performance was evaluated as a function of the water flux (J) and salt rejection (R) using the following equations;
J = Wp A⋅t
(1)
R = ((Cf − Cp) Cf ) ∗100
(2)
where Wp represents the total weight of the permeate sample, A is the effective area of the flat sheet membrane and t is the operating time. Cf and Cp are conductances of the feed and permeate stream respectively. Seawater desalination experiments were carried out using different amounts of GNPs incorporated nanohybrid membranes. Firstly, the influence of GNPs concentration (a concentration of 1, 2, 3, 4, 5 wt% with respect to the dry polymer's weight) on water flux and rejection were investigated at 65 °C. Then, the influence of the feed temperature (35, 45, 55, 65 °C) was evaluated when GNPs concentration in the polymer was 3 wt%. Effect of membrane thickness on desalination performance was also evaluated using the 3 wt% GNPs incorporated membrane when the temperature was 65 °C. In the present study, experiments were repeated three times and average data were given with experimental errors in figures. The experimental errors for rejections were < ± 1%.
3.2. Influence of GNPs content on desalination performance In the present study, a series of PEBA nanohybrid membranes with different concentrations of GNPs range from 0 wt% to 5 wt% (respect to the dry polymer's weight) were prepared and performed to desalinate Marmara Seawater. The influences of GNPs concentrations on the flux and total salt rejection are given in Fig. 6. In this study, graphene was used to improved flux, rejection and long-term stability of PEBA based membranes. It was observed from the figure that both flux and rejection results of all loaded membranes were higher than that of the pristine membrane. The slight increment in flux should be attributed to the increasing hydrophilicity of nanohybrid membrane by GNPs addition which was also confirmed by contact angle measurements. It was previously reported that graphene plate has polar chemical groups in its structure and these groups are responsible for the increase in water uptake capacity of the membrane. Graphene acts as a water-selective channel when it is homogeneously dispersed within the polymer matrix [31,32,43]. Besides of its inherent chemical property, graphene plate provides functionality to the polymer matrix. This feature contributes to enhancing water selective-productive property of the membrane. Graphene incorporation provided an experimental stability to membranes. The highest experimental variation in rejection obtained when the unfilled PEBA membranes were used. Another observation from Fig. 6, although only a slight improvement was able to obtain in the flux value, the increase in rejection was
3. Results and discussion 3.1. Characterization of membranes Fig. 2 shows the cross-sectional and surface images the pristine, 2 wt % and 5 wt% GPNs incorporated PEBA nanohybrid membranes. The average thickness of the membrane was seen as 121.6 μm in Fig. 2a. GNPs and their adoption within PEBA matrix were indicated in Fig. 2b. Due to the sonication procedure during the membrane preparation stage, graphene plates were intercalated within the polymer. Also, it was obvious that there was no interfacial void between GNPs and PEBA. However, agglomerated graphene particles were found in the cross structure of PEBA when GNP content was 5 wt% as shown in Fig. 2c and d. Excessive loading of GNPs would cause a particle agglomeration owing to the force interaction between graphene particles. Fig. 2e shows the surface of the 2 wt% GNPs loaded membrane after desalination. As seen in the figure, Na+ and Cl− ions penetrated on the surface of the membrane (on the feed side) and it was confirmed by EDX results (Fig. 2f). DSC analysis that represents the cooling-heating scans of PEBA based membranes is seen in Fig. 3. 166
Desalination 433 (2018) 164–171
F.U. Nigiz
Fig. 2. Cross-sectional SEM micrographs of the (a) pristine, (b) 2 wt% GNPs-PEBA (c, d) 5 wt% GNPs-PEBA (e) surface micrograph of 2 wt% GNPs-PEBA (after desalination) (f) EDX result of 2 wt% GNPs-PEBA membranes.
by increasing tortuous pathway through the membrane [44]. The rejection decrement was also related to the particle agglomeration. Surface defects that were created by agglomeration would allow uncontrolled ion passage and lead to decrease in the rejection which was referred to the size exclusion theory [10,15]. Based on this observation, saturation limit of GNP in PEBA matrix was determined as 3 wt%.
more remarkable. When GNPs contents in PEBA increased from 0 wt% to 3 wt%, total salt rejection enhanced from 99.12% to 99.92% at 65 °C. This improvement could be related to the increase in the tortuous pathway and decrease in free volume between the molecules of the polymer depending on GNPs addition [31,32,43]. However, both flux and rejection increments were hindered by the excessive incorporation of GNPs. As it was indicated in SEM results, GNPs showed agglomeration tendency and the local agglomerations were observed. Therefore, this situation would decrease the passage of water molecules and lead to decrease in flux. The decrement was related to the pore blocking effect of nano particles. Actually, GNP does not have any pores in its structure and it is impermeable to molecules or ions. By creating the nano pores inside GNP can allow the selective passage of the ions, otherwise; it can only act as a filler to ban the uncontrolled ion passage
3.3. Influence of temperature on desalination performance To determine the influence of the operating temperature on desalination performance, experiments were carried out at the temperature range from 35 °C to 65 °C using 3 wt% GNPs loaded membrane. Effect of temperature on flux and total salt rejection is illustrated in Fig. 7. 167
Desalination 433 (2018) 164–171
F.U. Nigiz
Fig. 6. Effect of GNPs content on flux and total rejection (65 °C).
Fig. 3. DCS curves of the pristine and GNPs filled PEBA membranes.
Fig. 7. Effect of temperature on flux and total rejection (3 wt% GNP incorporated nanohybrid membrane).
Fig. 4. TGA curves of the pristine and GNPs filled PEBA membranes.
As seen in the figure, when the temperature was changed from 35 °C to 65 °C, flux enhanced from 2.58 kg/m2·h to 3.61 kg/m2·h and 39.9% flux improvement was achieved. The increase in flux was not an unexpected result depending on the temperature. The relationships between the feed temperature and membrane separation results have been previously reported and explained in the literature. Increasing driving force between the sides of the membrane created by the increasing vapor pressure in the feed side is one of the main factors for flux enhancement. The change in physicochemical properties in the polymeric membrane is also responsible for flux increment. The relaxation in the structure of the polymer increased the transition of water molecules. As can be seen in Fig. 7, the highest rejection was achieved as 99.94% when the temperature was 35 °C and the rejection plot exhibited almost stable trend even the temperature increased to 65 °C. Since the driving force across the membrane was also maintained by the phase change and the salt was in solid state, it was not surprising
Fig. 8. Effect of temperature on total rejection of all membranes.
to obtain a stable trend in rejection. Nevertheless, a very slight decrease in rejection was seen in Fig. 8, when the pristine membrane was used. This was due to the change in free volume capacity of the pristine Fig. 5. Contact angle results of pristine (86°), 2 wt % GNPs filled (80°), and 5 wt% GNPs filled PEBA membrane.
168
Desalination 433 (2018) 164–171
F.U. Nigiz
Table 1 Resistance of the membrane. Membrane
R (ohm)
Pristine membrane 3 wt% GNP loaded 3 wt% GNP loaded 5 wt% GNP loaded 5 wt% GNP loaded
(before desalination) (before desalination) (after desalination) (before desalination) (after desalination)
791 480 450 529 470
with GNPs addition. Meanwhile, GNPs changed the charge density and conductivity of the membrane. When GNPs concentration increased from 0 wt% to 3 wt%, the resistance decreased from 791 Ω to 480 Ω before desalination. However, excessive GNPs addition increased the resistance from 480 Ω to 529 Ω due to the agglomeration. As also seen in Table 1, the charge load on GNPs incorporated nanohybrid membranes changed before and after desalination. The resistance of the membrane decreased after desalination; therefore, the conductivity and the charge load on membrane increased during desalination. This was due to the penetration of the hydrated ions into membrane structure (on the feed side) which was also confirmed by SEM-EDX analysis (Fig. 2e and f). The decrease in resistance of the 5 wt % GNPs filled membrane (− 12.5%) was more significant than that of the 3 wt% GNPs (− 6.6%) filled membrane. Meanwhile, the numbers of cumulated ions on the surface of the 5 wt% GNPs filled membrane were more than that of the 3 wt% GNPs filled membrane. Therefore, it could be predicted that the increasing concentration of GNP in the membrane would possibly cause a concentration gradient. Table 2 shows the drinking water guidelines and the ion concentration of permeate streams that were purified using the pristine and 3 wt% GNPs filled nanohybrid membranes [45,46]. As shown in the table, ICP-MS analysis confirmed that PEBA based membranes exhibited superior performance to retain many ions, mineral and heavy metal on the feed side. Both the pristine and GNPs incorporated membranes gave excellent purification results to meet the drinking water standards of World Health Organization and Europe. Table 2 indicates that the incorporating of GNPs into PEBA was improved the rejection of all hydrated ions and heavy metals. Especially in the case of hydrated salt ions such as Na+, Cl−, K+, Mg+ 2, at least ten times lower ion concentrations were observed in the permeate stream of the nanohybrid membrane. Moreover, heavy metal concentrations in the permeate stream of the nanohybrid membrane were significantly lower than that of the pristine PEBA membrane. Therefore, the graphene incorporated membrane can be considered to be a promising device that can be used in other hazardous chemical purification
Fig. 9. Effect of membrane thickness flux and total rejection (3 wt% GNP incorporated nanohybrid membrane, 65 °C).
polymer matrix. It appeared that GNPs addition overcame the trade-off trend between the flux and rejection owing to the temperature change. All GNPs filled nanohybrid membranes exhibited more stable rejection results compared to the unfilled membrane. Another observation obtained from Fig. 8 that the experimental repeatability of the membranes. GNPs incorporated membranes exhibited more consistent rejection results at the end of the three different experiments. Errors in nanohybrid membranes were smaller compared to pristine PEBA membrane. 3.4. Influence of membrane thickness on desalination performance A series of membranes having different thickness ranging from 72 μm to 181 μm were prepared with constant GNPs incorporation (3 wt%) and the effects of their thickness on desalination performance were evaluated at the constant temperature of 65 °C. It is clear from Fig. 9 that the flux change depending on the thickness was more remarkable than rejection variation. Flux increment depending on the membrane thickness is a predictable result in terms of the Fick's law. Owing to the reduction in hydraulic resistance of the membrane [10,18], flux decreased from 5.12 kg/m2·h to 2.73 kg/m2·h when the membrane thickness increased from 72 μm to 181 μm. Thickness variation did not remarkably affect salt rejections and stable trends in rejection (> 99.89) were observed in the figure. Xie et al. also reported a very similar relationship between flux and salt rejection [18]. 3.5. Determination the concentrations of hydrated ions in desalinated water
Table 2 Comparison of the experimental results with related guidelines.
Indeed, water permeation through the membrane is explained by the solution-diffusion model in a pervaporative desalination system. In addition to this model, the water passage or salt rejection is directly related to the size exclusion and/or charge exclusion phenomena [10,15]. In the first period of pervaporative desalination, water passage through the uncharged membrane is only related to the diffusion process where the size of the molecules – to be separated – is responsible for the transition. After the first period, the surface of the membrane is charged by the ions in seawater. Therefore, an electrostatic interaction occurs between the membrane surface and seawater. Consequently, the water passage is affected by the charge exclusion and size exclusion cases simultaneously. In the present study, potentiostat experiments were performed to determine the effect of GNPs surface resistance of the membrane depending on the GNPs addition. Additionally, resistances of the membranes before and after desalination experiments were analyzed to evaluate the charge variation depending on the ion separation. As seen in Table 1, the resistance of the PEBA membrane decreased 169
Ions
WHO drinking water standard
Europe/ Turkish drinking water standard
Permeate content (PEBA)
Permeate content (3 wt % GNP-PEBA)
Conductivity (μs/cm) Na (mg/L) Mg (mg/L) Cl (mg/L) K (mg/L) Li (μg/L) Al (μg/L) Mn (μg/L) Fe (μg/L) Br (μg/L) As (μg/L) Cd (μg/L) Hg (μg/L)
–
2500
110
27
200 – 250 – – 200 100 300 – – – –
200 – 250 – – 200 50 200 – 10 5 1
74.7 1.8 1.9 2.8
