The effect of using jug-type water filters on selected metals concentration in tap water – a case study

From the perspective of regular consumers, nowadays, the two main sources of drinking water are tap and bottled water. Differences in their properties have been widely discussed. Inorganic components and heavy metals concentrations are comparable in both cases (Cidu et al. 2011; Bakirdere et al. 2013). However, microbiological studies show that different bacterial communities dominate in tap and bottled water (Zamberlan da Silva et al. 2008; Sala-Comorera et al. 2020). It must be noticed that bottled water has several significant disadvantages, e.g. negative environmental impact attributed mainly to plastic waste production but also to manufacturing and transportation processes (Parag et al. 2023). Contamination of bottled water by synthetic polymers (Mason et al. 2018) and bisphenol A (Wang et al. 2020; Parto et al. 2022) was also found to be a problem. Moreover, in developed countries, tap water must meet the standards described precisely in documents published by the World Health Organization (World Health Organization 2022), European Council (European Union 2023) or local governments (The United Kingdom of Great Britain and Northern Ireland 2016; Environmental Protection Agency and of Water 2018; Han et al. 2022). Its quality is being carefully monitored and reports show that it is safe to drink tap water as well as bottled water (European Commission 2021a, 2021b, 2021c). However, sociological studies indicate that despite the relatively good quality of tap water and its ecological aspects, a certain number of people prefer drinking bottled water (Doria 2006; Saylor et al. 2011; Levêque & Burns 2018). The reasons for that are, among others, the smell and taste of tap water.

The possible solution to these problems is using filtering jugs and bottles, which have become very popular in the last few years because of their practicality, low price and wasting less plastic in comparison to using bottled water. Producers claim that filters remove solid pollutants, organic compounds, toxic metals and chlorine compounds from tap water, enhancing its quality. Water filters usually combine various water purification techniques. Solid pollutants are removed from the water by filtration. Undesirable chemical entities are removed by adsorption on activated carbon and/or by ion-exchange resins. Sometimes jug filters also contain additives that enrich water with selected ions, change pH or have antibacterial properties.

In contrast to the high popularity of filtering jugs and bottles, several studies on their effectiveness are rather limited. The efficiency of jug filters has been examined by measuring basic physicochemical parameters (Puszczało et al. 2019), and potential secondary water pollution caused by filtering has been studied (Puszczało et al. 2020). Moreover, the efficiency of removing arsenic(V) compounds from water by jug water filters has been tested (Kalachev et al. 2019). Silver migration from filters containing modified activated carbon has been also investigated (Garboś & Świecicka 2012). The behavior of Enterobacter aerogenes and Pseudomonas aeruginosa in water treated by a jug filter has been described (Briancesco et al. 2020). Furthermore, a simple method based on ascorbic acid degradation for jug filter effectiveness analysis has been proposed (Jezierska et al. 2020). The possibility of adsorbing organic pollutants from water by using filter cartridges has been tested (Szymańska & Nowicki 2023). Nevertheless, many aspects of water quality have been omitted or not tested yet, for example, the impact of the filtration process on the concentration of macro- and microelements in water.

The presence of metal ions in water can significantly affect its quality. For instance, high concentration levels of zinc and iron can negatively change the taste or appearance of drinking water and the excess of these ions should be removed during the filtration (World Health Organization 2022). However, one of the possible risks associated with filtration is removing not only pollutants but also the important and desired species from tap water (e.g. sodium, potassium, calcium and magnesium; Table 1) (World Health Organization 2012a, 2012b; Beto 2015; Al-Fartusie & Mohssan 2017).

The aim of this study is to investigate how the filtration process changes the concentration of Al, Ca, Fe, K, Mg, Na and Zn in jug-type filtered water in comparison with unfiltered tap water and how the filtration effects change, as the filter wears out. The presented test results are analyzed in terms of possible falsifications regarding the information on product characteristics as well as in terms of the occurrence of phenomena not described in the product characteristics. Moreover, the study determined the impact of the filtration process on water quality by comparing the content of tested elements in filtered water with selected bottled water and recommendations regarding the daily intake of tested macroelements.

Four jug water filters manufactured by one of the leading European suppliers were tested. Two of them (filters 3 and 4) were the same type (called ‘standard filters’ by the producer). According to the information from the producer's website, these filters contain activated carbon and ion-exchange resin. Thanks to this, the filter should reduce the content of heavy metals, chlorine compounds and other undesirable substances that may affect the taste of tap water. Filters 1 and 2 are different compared to the previous two. These are called ‘magnesium filters’ by the producer. This type of filter should have the same properties as standard filters but is expected also to enrich the water with magnesium to cover 20% of the daily consumption requirement for this element.

The filtering process takes about 3–4 min per liter and is rather simple: water poured into the upper part of the jug flows through the filter cartridge and goes to the bottom part. Following the producer's instructions, filtering devices were first rinsed with 1 L of tap water (this portion of water was disposed of). After that, 1 L of tap water was taken using a measuring cylinder (1,000 mL) and filtered. The filtrate from the bottom part of the jug was mixed, a 10 mL volumetric pipette was rinsed with the filtrate and the sample was further analyzed using the falcon test tube. Then the jug was emptied and the next liter of water was filtered. After filtering, 3, 5, 10 and then every 10th liter of tap water samples were collected using the procedure described previously. The producer claims that the filtering device will work properly for filtering 200 L of water. The experiment was designed to check the properties of filters in the studied field during the entire recommended period of use and if it is slightly exceeded. Therefore, total volume filtered with each device was 210 L. In addition, after filtering 1, 50, 100, 150 and 200 L of tap water, the cartridge was removed from the jug, 1 L of water was poured into the jug and the samples of water without filtration were taken according to the procedure presented above (control test of the composition of unfiltered water). The only difference between these two groups of samples is the presence of the filtering cartridge. Each sample was then preserved by adding 0.2 mL of 65% high-purity nitric acid (for trace metals analysis) and stored in a dark place.

Selected validation parameters of implemented analytical methods, based on the MIP-OES technique, are presented in Table 3. All reference solutions that were used for the preparation of calibration solutions were inductively coupled plasma spectroscopy (ICP)-standard purity (Merck Certipur^® Single-Element Standards for ICP series). Calibration solutions were characterized by a highly similar matrix composition as tested samples (2% HNO₃ in water). Four measurement repetitions were made for each element in every single sample, both in case of calibration solutions and samples analysis.

Moreover, an attempt was also made to determine Al, Cd, Cr, Cu, Ni and Pb in tap water; however, the concentrations of the mentioned elements were below the limit of quantification of used analytical methods (Al < 0.065 mg/L; Cd < 0.012 mg/L; Cr < 0.0039 mg/L; Cu < 0.023 mg/L; Ni < 0.0039 mg/L; Pb < 0.023 mg/L). Thus, these metals were removed from the research scope.

All results are presented in figures in the following sections with the calculated uncertainty values based on the standard deviation. The values for tap water were calculated as an average of the values obtained for all samples without filtration.

Analyzing the obtained test results in terms of meeting the daily demand for calcium, it can be concluded that drinking water filtered through a filter that has been used for a long time (>20 L of filtered water) allows covering 13–23% of the RDI for this element, and this value is slightly lower than when drinking unfiltered tap water (22–24%). However, in the case of fresh filters, the degree of calcium removal from water is so significant that the content of this element in filtered water drops to 0.2% of RDI at the beginning of filter use. However, this effect is not long-lasting and after just 10 L of filtered water, the calcium content reaches 7–15% of the RDI.

In the studied case, drinking unfiltered tap water results in a similar sodium requirement coverage as drinking water filtered through a filter used for a long time (>20 L of filtered water), i.e. 2.3–3.3% and 2.0–5.5%, respectively. However, the sodium content in water filtered using a fresh, low-consumption filter reaches concentrations of up to 7.1–12.5% of the maximum RDI based on drinking 2 L of water. Considering that the average daily sodium consumption in developed countries is two times higher than the RDI, enriching tap water with Na is not a desired phenomenon (Ginos & Olde Engberink 2020). It is particularly undesirable for very young children to consume high-sodium water. It has been proven that newborns consuming a powdered milk solution prepared with high-sodium water results in an increase in blood pressure (Pomeranz et al. 2002). A similar effect could occur when filtered water was used, which was obtained using the low-consumed filters, although it should be noted that the sodium concentration in the high-sodium water used in the above-mentioned studies was higher (196 mg/L) than the highest sodium concentration recorded in filtered water (125 ± 1 mg/L).

Average zinc and iron concentrations in water before and after filtration are listed in Table 4. Zn concentration value in tap water varies from 0.050 to 1.58 mg/L. After filtration with filters 1 and 2, the concentration of the mentioned element is below the limit of detection (0.016 mg/L), which indicates that most of Zn is removed from tap water. Interestingly, after filtration of 100 L of water with filter 1, Zn concentration is slightly higher (between 0.016 and 0.047 mg/L); thus, the process of removing it from water is less effective. The values obtained for filters 3 and 4 are higher (from 0.03 to 0.11 mg/L). This shows that those filters also remove Zn from water but to a lesser extent. Taste and appearance changes caused by the presence of Zn in drinking water are observed for concentrations higher than 4 and 3 mg/L, respectively (World Health Organization 2022). All the measured concentrations (in tap and filtered water) are significantly lower; nevertheless, it can be stated that removing Zn from tap water positively affects its quality. Fe concentration before the filtration is between 0.070 and 0.16 mg/L. The concentration levels in water filtered with filters 1 and 2 are noticeably lower – after filtering 20 L of water the value was below the limit of detection (LOD) (0.020 mg/L). In the case of filter 1, after filtering 100 L of water it reaches the value between 0.020 and 0.061 mg/L; as observed from Zn, iron is removed to a lesser extent. Average values measured for filters 3 and 4 are higher compared to Mg filters. The presence of Fe in a concentration higher than 0.30 mg/L can negatively change the taste of water (World Health Organization 2022). There are also literature reports that prove that iron concentrations in water lower than 0.30 mg/L may also be perceived as unpleasant by some people (Sain & Dietrich 2015). However, higher concentration of iron in drinking water can be considered to positively influence the health aspect of water quality (e.g. by reducing the number of cases of anemia) (Dutra-de-Oliveira et al. 2011). In the presented case, measured concentrations are lower than the mentioned 0.30 mg/L, but the positive effect of the filtration should be noticed when tap water contains more iron. To conclude, the results discussed above show that filters 1 and 2 (enriched with Mg) and filters 3 and 4 (standard filters) are considered much more efficient in the case of the mentioned heavy metals removal.

The presented research showed that both tested types of filters initially (up to approximately 20 filtered portions of water) significantly increase the concentration of sodium and potassium in filtered water and diminish it in terms of calcium content. This effect is not described by the manufacturer in the product characteristics. It is important to underline that the tested filters increase the sodium concentration in water (up to 300% at the beginning of usage), which is particularly unfavorable for children and people who follow a low-sodium diet every day. The authors recommend that filter manufacturers should inform customers about the phenomenon of sodium enrichment of water if their products have properties similar to those tested. The tested filters whose characteristics describe enriching water with magnesium partly meet this assumption, but the increasing content of this element in filtered water is lower than declared and is of limited importance to the daily requirement for this element. Tap water is enriched with this element to a relatively small extent and still contains less magnesium than the vast majority of bottled mineral waters. Therefore, the use of magnesium filters should not be considered an effective method of magnesium supplementation. The authors draw attention to the need to develop more effective filters that could enrich water with magnesium more effectively and with less loss of enrichment as the filter wears out. The tested standard filters deplete the water in magnesium while a new filter is used (below 20th liter of filtered water). This action results in a significant deterioration of the quality of consumed water. The tested filters with ion-exchange resin enriching the water with magnesium seem to effectively remove undesirable transition metals from water (which include the tested elements such as Zn and Fe). However, standard filters only slightly reduce the content of Zn and Fe in water.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.