Selective recovery of Cu²⁺ and Ni²⁺ from wastewater using bioelectrochemical system

Abstract

As the bioelectrochemical system, the microbial fuel cell (MFC) and the microbial electrolysis cell (MEC) were developed to selectively recover Cu²⁺ and Ni²⁺ ions from wastewater. The wastewater was treated in the cathode chambers of the system, in which Cu²⁺ and Ni²⁺ ions were removed by using the MFC and the MEC, respectively. At an initial Cu²⁺ concentration of 500 mg·L^-1, removal efficiencies of Cu²⁺ increased from 97.0%±1.8% to 99.0%±0.3% with the initial Ni²⁺ concentrations from 250 to 1000 mg·L^-1, and maximum power densities increased from 3.1±0.5 to 5.4±0.6 W·m^-3. The Ni²⁺ removal mass in the MEC increased from 6.8±0.2 to 20.5±1.5 mg with the increase of Ni²⁺ concentrations. At an initial Ni²⁺ concentration of 500 mg·L^-1, Cu²⁺ removal efficiencies decreased from 99.1%±0.3% to 74.2%±3.8% with the initial Cu²⁺ concentrations from 250 to 1000 mg·L^-1, and maximum power densities increased from 3.0±0.1 to 6.3±1.2 W·m^-3. Subsequently, the Ni²⁺ removal efficiencies decreased from 96.9%±3.1% to 73.3%±5.4%. The results clearly demonstrated the feasibility of selective recovery of Cu²⁺ and Ni²⁺ from the wastewater using the bioelectrochemical system.

Keywords

bioelectrochemical system / Cu²⁺ / Ni²⁺ / selective recovery

1 1 Introduction

The escalating global demand for water, driven by population growth and industrial activities, has resulted in severe water pollution and gradually environmental degradation. This degradation is primarily attributed to the release of industrial waste and the discharge of untreated industrial wastewater containing heavy metals (HMs) (Li et al., 2022a). Consequently, addressing water pollution has emerged as a significant global concern. Various human activities, including agricultural practices, domestic wastewater discharge, and industrial effluents, contribute to the release of untreated waste into diverse water bodies, including rivers, lakes, soils, and oceans. Moreover, approximately 34% of the world’s rivers contain factory waste concentrations exceeding 1%, with 6% exceeding 10%, underscoring the severe pollution of local water sources (Jiang et al., 2024).

The significant biological toxicity of HMs has garnered considerable attention in water pollution research due to their widespread presence in rivers. It is now understood that the dispersion, movement, and ultimate fate of HMs in river ecosystems affect the ecological balance, posing continuous risks to human health as they accumulate in food chains over time (Setia et al., 2021). HMs are readily absorbed by aquatic organisms, accumulating in their tissues and entering the food chain, posing risks to aquatic ecosystems and human health (Pan et al., 2024). Studies have elucidated the harmful effects of HMs on aquatic organisms, including growth inhibition, gill necrosis, and other adverse outcomes. Moreover, even at low concentrations, most HMs are toxic to humans, raising significant concerns in ecotoxicology due to their persistent presence, tendency to accumulate over time, and transfer within food chains (Xue et al., 2023; Yi et al., 2024).

For instance, prolonged exposure to arsenic has been associated with the onset of cardiovascular diseases, visceral cancers, and neurological disorders (Chen et al., 2023). Similarly, the continuous consumption of water contaminated with cadmium has been linked to kidney ailments, bone deterioration, anemia, and gastrointestinal cancers (Rajakumar et al., 2020; Guo et al., 2023). Research indicates that up to 90% of cancer cases may be attributed to exposure to chemical carcinogens, and the health hazards posed by these substances are intricately correlated with the level of HM contamination in water sources (Rai et al., 2019; Shokri et al., 2022). Hence, the comprehensive assessment and evaluation of HMs pollution and their concomitant health risks are imperative for elucidating the intricacies of human health and aquatic environments. Such an evaluation represents a prominent area of current research focus within the field of HM geochemistry.

Natural processes like soil erosion and rock weathering minimally influence the presence of HMs in rivers. Instead, anthropogenic activities, including industrial discharges, agricultural fertilizer use, and urban emissions, are the primary sources of HM contamination in these waterways (Chen et al., 2024). Variations exist in the bioavailability of HMs originating from different sources. For instance, stable components found in soil typically stem from natural origins, displaying minimal migration and bioavailability. Conversely, active components, predominantly sourced from anthropogenic activities, exhibit heightened bioavailability, posing increased health risks to organisms (Xia et al., 2020). Furthermore, the levels and distribution patterns of HMs across environmental media, including soil, atmosphere, and water, vary significantly based on regional geological conditions and human interventions. Consequently, a comprehensive understanding of the spatial distribution, pollution assessment, and source analysis of HMs is imperative for elucidating the ecological landscape (Liu et al., 2023).

In aquatic environments, sediments play a vital role as the main reservoir for HMs, forming an integral part of river ecosystems. HMs discharged into river systems often bind to suspended solid particulate matter (SPM) before settling in the bed sediment, adversely impacting aquatic organisms (Haghnazar et al., 2021). Approximately 90% of metal loads in aquatic ecosystems are associated with SPM and sediments, underscoring sediment's pivotal role as a significant sink for HMs (Islam et al., 2018). The distribution and interactions of metals within sediment are influenced by various physicochemical factors, including water properties such as pH, salinity, temperature, and sediment characteristics like texture composition, mineralogy, and organic matter content (Islam et al., 2015; 2018). Changes in temperature, pH, and redox potential can trigger the release of sorbed metals from sediment into the water column, potentially acting as a secondary source of pollution to the overlying water (Haghnazar et al., 2021; Miranda et al., 2021). Consequently, bottom sediment is a sensitive indicator for monitoring pollutant levels in aquatic environments. Previous studies have consistently highlighted the significant risks of HM-contaminated river sediment to benthic populations, emphasizing the critical need to assess and manage sediment pollution to preserve aquatic ecosystems (Das et al., 2023).

Several methodologies have been utilized in the investigation of HMs sourcing, such as the implementation of the random forest approach (Jia et al., 2020), multi-statistical methods (Xiao et al., 2021), isotopic tracer (Cloquet et al., 2006), and positive matrix factorization (PMF) (Xu et al., 2023). Among these methods, PMF analysis, as endorsed by the EPA, has emerged as a powerful technique for discerning and quantifying pollution sources, particularly HM contamination in soil (Kumar et al., 2024). By breaking down datasets into contribution matrices and summarized groups, PMF enables the qualitative identification and accurate quantification of the contribution of various pollution sources (Zhang et al., 2020). With its constraint on nonnegative values and incorporation of error estimates, PMF modeling has become instrumental in pinpointing sources of pollution across diverse environmental media, from particulate matter to sediments (Huston et al., 2012). This method’s ability to provide insights into the number of sources, significant HM elements per source, and percentage contributions further underscores its importance for formulating effective pollution control policies (Zhang et al., 2020; Dai et al., 2024; Liu et al., 2024).

This study leveraged PMF analysis to assess HM levels in surface water and sediments within the Ningxia section of the Yellow River Basin. The Ningxia region, situated along the Yellow River in Northwest China, is a vital energy and grain-producing hub. Despite its critical role in Chinese industry and agriculture, the area faces longstanding challenges stemming from low investment, high consumption, excessive water use, and severe pollution. Discharging untreated wastewater and significant volumes failing disposal standards directly into the Yellow River have intensified the problem. Consequently, annual sewage discharge into the river’s main channel doubled by the early 1990s, reaching 4.2 × 10⁹ m³. Presently, with only 48.6% of the main channel and primary tributaries meeting water quality standards, urgent action is imperative to address environmental degradation, emphasizing the critical need for comprehensive studies on sediment dynamics in the Yellow River (Guan et al., 2016).

This study focused on 49 sampling points across Ningxia to achieve several objectives. Firstly, it examined the spatial and temporal distribution of HMs (Cr, Hg, Pb, Mn, As, Co, Tl, Sb, Zn, Ni, Cd, Cu, Sr) in surface water and sediments, particularly examining their concentration variations during wet and dry seasons. Secondly, it employed the Single Factor (P_i), and Nemerow pollution indexes (P_N) to assess water HMs pollution levels. It also evaluated metal contamination in sediments using indicators like the Geo-accumulation Index (I_geo), contamination Factor (CF), and Pollution Load Index (PLI). Thirdly, the present study sought to identify potential sources of HMs in Ningxia’s surface water through multivariate statistical methods and the PMF model. Taken together, these findings significantly contribute to filling the existing data gaps regarding HM distribution, pollution status, and plausible sources in Ningxia's surface water.

2 2 Materials and methods

2.1 2.1 Study area and sample collection

In this study, we assessed heavy metal levels in surface water and sediments within the Ningxia section of the Yellow River Basin. Ningxia is located in the upper to middle reaches of the Yellow River in Northwest China, spanning from 104°17′ E to 107°39′ E and 35°14′ N to 39°23′ N (Fig.1). It shares borders with Shaanxi to the east, Inner Mongolia to the west and north, and Gansu Province to the south. Covering an area of 66400 km², Ningxia has a population of approximately 7.28 × 10⁶. Ningxia spans semi-humid, semi-arid, and arid zones within the mid-temperate region. Agricultural land comprises 32.97% of Ningxia, followed by grassland (16.4%) and forest (8.05%). The region is characterized by scattered dunes and sandy areas, surrounded by sand on three sides, leading to a fragile ecological environment and severe desertification. The desertified area is 2.9 × 10⁶ ha, accounting for 43.7% of Ningxia’s land area, much higher than the national average of 27.3% (Ma et al., 2016). Additionally, Ningxia faces a severe water shortage, with per capita water resources at only 30% of the national average (Wang et al., 2022a).

Fig.1 Location of sampling sites.

Full size|PPT slide

Surface water and sediment were sampled from 49 representative sites across Ningxia, selected based on national and provincial control sections to ensure comprehensive coverage of key hydrological and pollution areas (Fig.1 and Table S1). Surface water samples were collected during the wet season (August–September 2022) and the dry season (March–April 2023), with each sample comprising 500 mL. Sediment samples were collected during the dry season (March–April 2023). Surface water samples were acidified, collected using clean samplers, and stored in acids-washed polyethylene flasks. Sediment samples were taken from the top 0–10 cm layer using a grab bucket, and 1 kg of sediments were stored in polyethylene bags. To ensure representativeness, samples from each site taking multiple points were combined into one large sample.

2.2 2.2 Chemical analysis

The water samples were stabilized with HNO₃ to achieve a pH of less than 2 and sealed with a submembrane. Surface water samples were filtered using a 0.45-μm polyether sulfone membrane. Sediment samples underwent lyophilization, grinding, sieving, weighing, and digestion in a microwave digestion system with a concentrated acid (HCl–HNO₃–HF) mixture (Bettinelli et al., 2000). Chengdu Cologne Chemical Co., Ltd., Chengdu, China, supplied HNO₃ (65%–68%), HCl (36%–38%), and HF (≥ 40%). Following pretreatment, Sr, Co, Cu, As, Ni, Cd, Sb, Zn, Pb, and Cr were analyzed using inductively coupled plasma mass spectrometry (ICP-MS; Thermo, USA) at 1200 W and 0.9 L/min argon flow, with indium as an internal standard (Yang et al., 2017). Mn and Tl were quantified by inductively coupled plasma optical emission spectrometry (ICP-OES; IRIS, USA), utilizing a multi-channel detector for enhanced sensitivity. Mercury concentrations were determined using atomic fluorescence spectrometry (AFS; AFS-8500, Haiguang, China) with cold vapor technology to ensure efficient capture and accurate measurement.

2.3 2.3 Assessment of HMs in surface water and sediment

The assessment of HM pollution in surface water and sediment involved several key indices and analytical methods. Levels of HMs pollution in surface water were assessed using the P_i and the P_N, which compared measured concentrations to environmental standards. The I_geo was used for sediments to assess contamination levels relative to background values. Additionally, the CF and PLI provided insights into the enrichment of individual metals and overall pollution extent. Detailed calculations and classifications are described in the supporting information (Text S1 and Table S2).

2.4 2.4 Pollution sources of HMs

The sources of contamination for HMs were defined using PMF and principal component analysis (PCA). PMF, a receptor model based on factor analysis, quantitatively identifies primary pollution sources using sample data (Xu et al., 2023). It employs the least squares method to determine the principal pollution sources and their respective contribution ratios, ensuring accurate error weights in the chemical composition of the receptor. PCA reduces the dimension of the data set and enhances explainability while minimizing information loss (Xiao et al., 2021). It generates new orthogonal variables that maximize variance and uses principal factor loadings derived from chemical elements to discern source types. Detailed methodologies, including calculation equations and evaluation criteria, are displayed in the supporting information (Text S2 and Text S3).

3 3 Results and discussion

3.1 3.1 HM pollution in surface water

3.1.1 3.1.1 Concentration of HMs in surface water

Tab.1 shows the results of soluble HMs for surface water in Ningxia. We found that the mean concentrations of HMs during the wet season followed a decreasing order of Sr >> Mn > Zn > Cu > As > Ni > Cr > Co > Sb > Pb > Tl > Cd > Hg. Conversely, during the dry season, the order shifted to Sr >> Mn > Zn > Cu > Cr > Ni > As > Tl > Co > Pb > Sb > Hg > Cd. The coefficient of variation (CV), a standardized measure indicating dispersion within a probability distribution, revealed that 13 HMs exhibited moderate to high variability (CV > 0.36), suggesting significant fluctuations in surface water metal concentrations across the study area, likely influenced by human activity (Karaouzas et al., 2021).

Tab.1 Statistical analysis of HMs in surface water during wet and dry seasons (unit: μg/L)

Elements	Wet season						Dry season					Class IIIc)	WHOd)
	Mean	Max	Min	SDa)	CVb)		Mean	Max	Min	SD	CV
Cr	5.49	48.01	0.43	11.08	2.02		3.89	16.87	0	3.51	0.90	50	50
Mn	59.88	476.08	6.93	78.98	1.32		48.36	205.61	2.36	42.54	0.88	–e)	400
Co	1.09	2.77	0.26	0.70	0.64		0.48	1.24	0.08	0.27	0.57	1000	–
Ni	6.00	18.08	1.60	3.77	0.63		2.87	11.68	0.50	2.52	0.88	20	20
Cu	16.05	65.07	2.80	14.75	0.92		10.41	47.78	2.07	9.66	0.93	1000	2000
Zn	27.06	115.71	6.45	21.05	0.78		12.62	33.94	3.92	6.34	0.50	1000	–
As	6.20	55.32	1.79	7.77	1.25		1.80	4.52	0.42	1.06	0.59	50	10
Cd	0.06	0.22	0.01	0.05	0.77		0.04	0.08	0.01	0.02	0.55	5	3
Sb	0.68	1.93	0.12	0.39	0.58		0.27	0.70	0.05	0.18	0.65	5	–
Hg	0.03	0.51	0	0.09	2.53		0.09	1.36	0	0.20	2.30	0.1	6
Tl	0.23	8.84	0.01	1.27	5.51		1.50	71.22	0	10.17	6.76	0.1	–
Pb	0.58	3.07	0.12	0.61	1.05		0.47	2.21	0.10	0.42	0.90	50	10
Sr	1569.56	5181.19	184.86	1169.93	0.75		1199.07	3093.10	339.88	757.06	0.63	–	–

Note: a) SD: Standard deviation; b) CV: Coefficient of variation; c) Class III: The threshold values of Class III of the EQSSW; d) WHO: The World Health Organization (WHO); e) –: Denotes no relevant data.

Regarding seasonal variations, in the wet season, the mean concentrations of Co, Ni, Zn, As, and Sb were 1.09, 1.00, 27.06, 6.20, and 0.68 μg/L, respectively, roughly double those observed during the dry season. Conversely, the mean concentrations of Hg and Tl during the dry season were 0.09 and 1.50 μg/L, respectively, approximately three and five times higher than those recorded during the wet season. This variability might be attributed to runoff and wet and dry deposition (Wang et al., 2022b). It is highly conceivable that during the dry season, the diminished surface water flow rate led to the accumulation of HM dry deposition on the surface. In contrast, during the wet season, heightened water flow facilitated the ingress of HMs into the water through surface runoff and atmospheric sedimentation (Haghnazar et al., 2021).

According to the WHO and the Environment Standards for Surface Water (EQSSW), the mean levels of HMs were not above the standard thresholds, except for Tl (Tab.1). As displayed in Table S3, except for Tl and Sr, the mean concentrations of HMs in surface water in this study were below those previously reported in the literature (Zuo et al., 2016). Additionally, compared to the surface water in other countries or regions such as Greek surface waters (Karaouzas et al., 2021), river basins in India (Krishna et al., 2009), Northeast region of China (Cui et al., 2022), Indonesia (Wulan et al., 2020), and Chinese lakes (Qin and Tao, 2022), the mean levels of most of the measured metals (especially Pb, Mn, Cd, and Cr) were significantly lower in the surface waters of the study area.

3.1.2 3.1.2 Distribution of HMs in surface water

The spatial distribution of HMs from the surface water in Ningxia is shown in Fig.2. We found that Mn, Cu, Zn, Ni, Cd, and Pb exhibited a comparable distribution pattern, characterized by elevated values at sampling points 35 and 36. Conversely, Hg and Tl displayed lower concentrations across most locations, except for sampling points 1, 23, and 48 (Fig.2). Moreover, our analysis revealed considerable CV for Hg and Tl, ranging from 2.30 to 6.67 (Tab.1). Correlation analysis further demonstrated a robust positive association between Hg and Tl concentrations during wet and dry seasons (Fig. S1), suggesting inherent similarities between these elements (Li et al., 2024). Additionally, Co and Ni exhibited analogous spatial distribution patterns with a slight overall variation. Notably, the highest levels of HMs were observed in the irrigation area of northern Ningxia, which corresponded to the region’s densely populated and industrially and agriculturally developed areas.

Fig.2 Distribution of HMs in surface water.

Full size|PPT slide

3.1.3 3.1.3 Pollution assessment of HMs in surface water

Unfortunately, this study omitted a water quality assessment for two elements due to the absence of specified standard limits for the HM elements Mn and Sr in the national water quality standards for surface water environmental quality. As depicted in Fig.3, the mean values of$P_i$ were in the following order: Tl > Hg > Ni > Sb > Cr > As > Zn > Cu > Pb > Cd > Co, indicating that Tl, with higher$P_i$ values, may pose a greater risk compared to HMs with lower $P_i$ values. Notably, the mean $P_i$ values of all HMs except Tl were less than 1, whereas the $P_i$ values of Hg and Tl were greater than 1 at sampling points of 1 to 5.

Fig.3 The $P_i$ index of HMs in surface water.

Full size|PPT slide

The $P_N$ was used to assess the total pollution status of HMs in surface water (Fig.4). The results showed that 18% and 20% of the surface water samples were polluted during the wet and dry seasons, respectively. Specifically, elevated pollution levels were observed in the wet season at sampling points 22 and 23, with moderate pollution detected only at point 25. Low contamination was found at points 9, 24, and 31 (Fig.4(a)). Conversely, in the dry season, high pollution was detected at points 1 and 2, while low contamination was observed at points 3, 4, 5, 9, 14, 25, 27, and 29 (Fig.4(b)). This pattern was attributed to higher concentrations of Tl, Hg, and As. Given that other areas had relatively low pollutant levels, the poor water quality at points 1 and 22 might be linked to anthropogenic activities (Karaouzas et al., 2021). Overall, these findings suggest that most areas in Ningxia are not significantly affected by HM pollution in surface water.

Fig.4 Spatial distribution of $P_N$ for HMs in surface water of Ningxia: (a) wet season, (b) dry season.

Full size|PPT slide

3.2 3.2 HM distribution in sediments

3.2.1 3.2.1 Concentration of HMs in sediments

Tab.2 presents the basic statistics regarding HM concentrations in Ningxia sediments alongside the mean background values of stream sediments in China (BBS) (Xie et al., 2022). The mean concentrations of HMs in sediments decreased in the following order: Mn > Sr > Cr > Zn > Ni > Cu > As > Pb > Co > Sb > Tl > Cd > Hg. Additionally, except for the Hong River (HR), the mean concentrations of different HMs in various water environments exhibited similar contributions (Fig. S2). Notably, Cd, Hg, and Tl displayed higher variability among sites, with CV reaching 139%, 122%, and 110%, respectively. In contrast, the CV values for other HMs generally ranged from 20% to 62%.

Tab.2 Statistical analysis of HMs in Ningxia sediments (unit: mg/kg)

Elements	Sediments					BBS	CF				Igeo
	Mean	Max	Min	SDa)	CVb)		Mean	Max	Min		Mean	Max	Min
Cr	150.57	358.02	92.78	40.51	0.27	54.00	2.86	6.63	1.72		0.88	2.14	0.20
Mn	776.03	1319.49	437.26	183.77	0.24	653.00	1.19	2.02	0.67		−0.37	0.43	−1.16
Co	18.04	32.89	9.00	4.18	0.23	12.00	1.49	2.74	0.75		−0.04	0.87	−1.00
Ni	48.81	72.11	25.73	10.10	0.21	23.00	2.11	3.14	1.12		0.46	1.06	−0.42
Cu	32.89	61.16	15.05	9.47	0.29	20.00	1.63	3.06	0.75		0.07	1.03	−1.00
Zn	116.25	411.44	58.72	58.22	0.50	67.00	1.73	6.14	0.88		0.09	2.03	−0.78
As	25.88	62.72	10.20	8.40	0.32	9.00	2.87	6.97	1.13		0.87	2.22	−0.40
Cd	0.32	2.50	0.13	0.44	1.39	0.13	2.44	19.20	0.99		0.24	3.68	−0.60
Sb	2.06	6.76	0.40	0.98	0.48	0.70	2.97	9.66	0.58		0.84	2.69	−1.38
Hg	0.15	1.10	0.02	0.19	1.22	0.68	4.57	32.40	0.70		1.08	4.43	−1.11
Tl	0.84	7.04	0.53	0.92	1.10	0.45	1.87	15.75	1.18		0.11	3.39	-0.35
Pb	22.98	52.68	18.01	5.26	0.23	23.00	1.00	2.29	0.78		−0.61	0.61	−0.94
Sr	447.98	2032.88	296.6	279.11	0.62	146.00	3.09	13.92	2.03		0.91	3.21	0.44

Note: a) SD: Standard deviation; b) CV: Coefficient of variation.

The mean concentrations of Cr (105.27 mg/kg), Mn (776.03 mg/kg), Co (18.04 mg/kg), Ni (48.81 mg/kg), Cu (32.89 mg/kg), Zn (116.25 mg/kg), As (25.88 mg/kg), Cd (0.32 mg/kg), Sb (2.06 mg/kg), Hg (0.15 mg/kg), Tl (0.84 mg/kg), and Sr (447.98 mg/kg) exceeded the BBS (Tab.2). Meanwhile, the mean concentration of Pb (22.98 mg/kg) was lower than the BBS.

Furthermore, a comparison of the mean concentrations of HMs with those in other regions (Table S4) revealed that the levels of Hg, Cd, Zn, and Pb in Ningxia sediments were generally lower than reported elsewhere (Santos Bermejo et al., 2003; Karbassi et al., 2008; Islam et al., 2015; Karaouzas et al., 2021; Kumar et al., 2021; Xiao et al., 2021). Conversely, the mean values of Cu, Co, Mn, Ni, and Cr metal concentrations exceeded those recorded in the Lijiang River of China (Xiao et al., 2021), the Nakuvadra-Rakiraki River in Fiji (Kumar et al., 2021), and Bangladesh (Khan et al., 2020). Notably, the mean concentration of Tl (0.84 mg/kg) in Ningxia surpassed that of the water system in Bangladesh (Khan et al., 2020). These variations might be related to the strong impact of anthropogenic activities on HM pollution in Ningxia.

3.2.2 3.2.2 Spatial distribution of HMs in sediments

Next, Inverse Distance Weighting (IDW) interpolation was conducted to obtain a refined representation of HMs distribution in Ningxia, as depicted in Fig.5. The spatial distribution patterns of Mn, Co, Ni, Cu, As, and Sb exhibited similarities with regions of high level predominantly located in the southern part of Yinchuan Plain. Notably, both Cd and Tl demonstrated significantly higher concentrations at site 15, adjacent to Xingqing District. Moreover, Cd concentrations were elevated at site 14 compared to other sampling sites. This observation suggested that point source contamination of these metals exhibited homogeneity, offering valuable insights into the localized distribution patterns of HMs in Ningxia. Furthermore, Pb and Zn exhibited fluctuations in the Drainage Ditch (Fig.5), suggesting that the sedimentary accumulation of these metals may be ascribed to agricultural and anthropogenic activities (Zhang et al., 2020). The zones of elevated Cr concentration were predominantly located in the central region of Ningxia, with comparatively lower values observed in the southern region.

Fig.5 Spatial distribution of HMs in Ningxia sediments.

Full size|PPT slide

3.2.3 3.2.3 Pollution assessment of HMs in sediments

The contamination of HMs in sediments within the study area was assessed using CF, I_geo and PLI. The mean CF values for all water types are illustrated in Fig. S3(a). The mean CF values for HMs in sediments decreased in the following order: Hg > Sr > Sb > As > Cr > Cd > Ni > Tl > Zn > Cu > Co > Mn > Pb, with corresponding values of 4.57, 3.09, 2.97, 2.87, 2.86, 2.44, 2.11, 1.87, 1.73, 1.63, 1.49, 1.19, and 1.00, respectively (Tab.2). The mean CF values for Sb, As, Cr, Cd, Ni, Tl, Zn, Cu, Co, Mn, and Pb indicated moderate contamination levels (1 ≤ CF < 3), while the mean CF values for Sr and Hg signified significant contamination levels (3 ≤ CF < 6).

I_geo values were employed to evaluate sediment quality, providing a means of comparison for the same HM across different regions unaffected by geological factors (Karbassi et al., 2008). The calculated I_geo for metals (excluding Hg) ranged from −0.61 to 1 (Tab.2), indicating that sediments ranged from unpolluted to moderately contaminated by HMs (refer to Table S2 for classification). Notably, 61% of sampling sites exhibited I_geo indices suggesting unpolluted to moderately contaminated levels of HMs. Specifically, sites JR (1.10), MYR (1.28), HR (1.37), DNR (1.42), and DD (1.48) were classified as moderately contaminated with Hg, while site DR (3.15) was identified as heavily contaminated with Hg (Fig. S3(b)).

To characterize the overall pollution of sediments by HMs, PLI was computed (Fig.6). The PLI for HMs ranged from 1.45 to 3.40 across all samples, indicating the mild to moderate pollution of sediment samples in Ningxia (Islam et al., 2015). A spatial analysis of PLI values revealed elevated levels in the northern part of Ningxia, particularly at sampling points 11, 14, 15, 17, 18, and 21, situated near the industrialized cities of Shizuishan and Yinchuan. These regions exhibited relatively darker shades of red in the PLI map. Pollution factor analysis indicated that Cd, Pb, Tl, Co, and Hg were the primary contributors to metal pollution in these areas. Additionally, sampling sites at 38 and 39 exhibited relatively high PLI values, primarily due to Co, Cu, Pb and Zn.

Fig.6 PLI spatial distribution of HMs in sediments.

Full size|PPT slide

3.3 3.3 Source identification of HMs in surface water

3.3.1 3.3.1 Principal component analysis

The PCA findings are depicted in Fig.7 and Table S5. During the wet season, four principal components were identified from the surface water data, accounting for 77.27% of the total variance. PC1 had an Eigenvalue of 5.258 and explained 37.56% of the variance, with Co, Mn, and Ni showing the highest loadings. PC2 had an Eigenvalue of 2.531 and contributed 18.08% to the variance, with Cd, Hg, and Tl also having the highest loadings. PC3, with an Eigenvalue of 1.75 and explaining 12.51% of the variance, was characterized by high loadings of Zn, Pb, Sb, and As. PC4 had an Eigenvalue of 1.28 and contributed 9.12% to the variance, with Cr, Sr, and Cu being the most significant.

Fig.7 PCA load map: (a) wet season, (b) dry season.

Full size|PPT slide

During the dry season, four primary components were similarly extracted, explaining 78.38% of the total variance. PC1 had an Eigenvalue of 4.949 and accounted for 35.35% of the variance, with the highest loadings observed for Co, Cr, Cd, Mn, Sr, and Cu. PC2 had an Eigenvalue of 2.965 and explained 21.18% of the variance, with Hg, Zn, Pb, and Tl showing the highest loadings. PC3, with an Eigenvalue of 1.694 and explaining 12.10% of the variance, was dominated by Sb and Ni. Finally, PC4, with an Eigenvalue of 1.364 and contributing 9.74% of the variance, had As as the most significant loading.

3.3.2 3.3.2 Positive matrix factorization

The contributions of various factors to the observed variability in the dataset are illustrated in Fig.8. During the wet season (Fig.8(a)), Factor 1 was primarily influenced by Cu (74.77%), Cr (71.62%), Sr (65.82%), Co (52.35%), and Ni (45.73%). This factor was associated with anthropogenic sources, particularly agricultural activities. It is well-established that Cu is extensively utilized in chemical fertilizers and pesticides. Furthermore, reports have linked Ni accumulation to agricultural activities (Liu et al., 2024). Additionally, several surveys have documented that Cr originates from the parent material (Xie et al., 2022). Consequently, Factor 1 was inferred to originate predominantly from agricultural sources.

Fig.8 Profiles and contributions of HM sources based on the PMF model: (a) wet season, (b) dry season.

Full size|PPT slide

Factor 2 displayed higher Tl and Hg loadings, contributing 95.91% and 36.40%, respectively, making Tl and Hg distinctive identifiers for Factor 2. Consistently, PCA identified the same source for Tl and Hg (Fig.7(a)). Spatial distribution analysis of surface water revealed that sites with elevated Tl and Hg levels were predominantly situated in Pingluo and Huinong County (Fig.2 and Tab.1). Pingluo County, located in the northwestern interior of China’s Ningxia region, is historically known for its industrial activities. Prolonged industrial operations have released pollutants, particularly HMs, into the soil (Luo et al., 2022). Additionally, coal combustion, plastic manufacturing, and metal extraction have been identified as potential sources of Hg emissions (Xie et al., 2022). Thus, Factor 2 was primarily inferred to originate from industrial sources.

Factor 3 was primarily influenced by Pb (52.08%) and As (51.37%). Pb is commonly regarded as a representative element for transportation-related sources in source apportionment studies (Liu et al., 2024). A recent study (Li et al., 2022b) reported a strong association between Pb and transportation activities in the Ningxia Weining Plain. Additionally, a comparison with Fig.2 revealed that sites exhibiting elevated Pb and As concentrations (4, 7, 9, and 39) were located close to roadways. Based on these observations, Factor 3 was attributed to traffic-related pollution.

Factor 4 was primarily influenced by Mn (75.43%) and Cd (46.02%). The spatial distribution patterns of Mn mirrored those of Cd HM, with notably elevated concentrations observed at sites 1, 10, 35, 36, 37, and 49 (Fig.2). While some researchers have suggested that Mn primarily originates from crustal sources, the presence of Cd is predominantly attributed to industrial emissions and other anthropogenic inputs (Cloquet et al., 2006). Meanwhile, studies have found that Cd is mainly derived from atmospheric deposition (Han et al., 2017; Wu et al., 2021). Therefore, it is highly conceivable that both Mn and Cd in this study area may have natural origins, such as atmospheric precipitation and chemical weathering.

Factor 5 displayed higher weights attributed to Sb (67.04%), Hg (47.1%), and Zn (37.99%) (Fig.8(a)). This group of elements has been consistently associated with HMs and chemical industry activities in various studies (Wang et al., 2022b), which appears relevant for surface water in Ningxia as well. For instance, factories near sampling sites 22, 31, and 35 could contribute to surface water pollution via atmospheric transport and deposition (Ma et al., 2016).

Similarly, during the dry season (Fig.8(b)), Factor 1 exhibited a heavy load of Mn (66.3%) and Co (38.5%) due to the presence of a large manganese factory in Qingtongxia City and Zhongning City (sites 35 and 30). The primary waste residue from electrolytic manganese production comprises an acid-leaching residue containing cobalt sulfide, among other elements (Xu et al., 2023). Thus, Factor 1 indicated industrial pollution resulting from electrolytic manganese production.

Factor 2 was primarily influenced by Zn (77.1%), Sb (62.7%), Pb (69.2%), As (50.1%), Tl (46.0%), and Ni (38.5%). Atmospheric deposition serves as the primary source of Zn. Pb, with a contribution rate of 69.2%, is primarily attributed to human activities like coal and oil combustion as the primary sources of lead (Huston et al., 2012). Furthermore, farmland soils in Zhongwei City, irrigated from the Yellow River, have formed silted soil, which previous research found carries HM pollution (Xie et al., 2022). Factor 2 likely originated from both fuel burning and natural sources.

Factor 3, mainly Cd (52.2%) and Sr (50.6%), exhibited a significant correlation (p ≤ 0.01) between Cd and Sr during correlation analysis (Fig. S1), indicating similar origins. Cd is commonly associated with automobile and fuel emissions, with tire wear being a notable contributor to environmental Cd release (Das et al., 2023). A study (Zhang et al., 2015) found that traffic was the primary source of Sr in the semi-humid region of Northwest China. In addition, sites 28, 29, 30, and 40, which exhibited elevated Cd and Sr levels, are located near transportation hubs such as highways and bridges (Fig.2). Hence, factor 3 was attributed to traffic-related sources.

Factor 4, dominated by Cr (74.3%) and Cu (60.5%), was likely caused by the accumulation of various domestic waste materials, as indicated by previous studies (Rajeshkumar et al., 2018). Additionally, Cu is often used as a marker for human activities. Factor 4 was associated with emissions originating from domestic waste.

Factor 5 was characterized by Hg, with a contribution rate of 74.0%. Studies have shown higher levels of Hg, primarily in the northern part of Ningxia, where intensive agricultural production occurs. Hg is a major component of pesticides and fertilizers, with notable volatility and transport tendencies (Xiao et al., 2021). Surface water used for irrigation in surrounding fertile lands could facilitate the return of pesticides and fertilizers, making them a significant source for Factor 5.

Overall, the contributions of these sources are depicted in Fig. S4. Compared with global studies, the pollution sources identified in this research aligned with patterns observed in other regions while exhibiting localized characteristics. For instance, agriculture and traffic were a predominant source of HMs pollution in the Ningxia region, contributing significantly to the overall pollutant load (Fig. S4), consistent with studies from the Yellow River (Upstream) (China) (Xie et al., 2022) and West Java (Indonesia) (Wulan et al., 2020), where agriculture plays a critical role in water contamination. Moreover, industrial practices represented a major source of contamination in our study, similar to findings in the Yangtze River Basin (Chen et al., 2024) and the Bengal Basin river system (Khan et al., 2020), where industrial activity also dominates. Our results demonstrated that controlling human activities in Ningxia, like agriculture and traffic, is necessary to reduce the HM content in watersheds.

4 4 Conclusions

This study comprehensively evaluated the distribution, origins, and hazards associated with HMs in surface water and sediments within the Ningxia region. Tl exhibited the highest pollution levels among the HMs studied, while Hg showed comparatively lower pollution levels in surface water. Assessments using CF, I_geo, and PLI indicated that sediment contamination ranged from mild to moderate. Specifically, sediments showed high mean concentrations of Mn (776.03 mg/kg), Sr (447.98 mg/kg), and Cr (150.57 mg/kg). A notable trend of decreasing HM concentrations from northern to southern Ningxia was observed. P_N calculations highlighted that 18% and 20% of surface water samples were polluted during the wet and dry seasons, respectively. Notably, concentrations of Tl and Hg occasionally exceeded WHO guidelines, indicating significant health risks. In sediments, Cd, Hg, and Tl displayed significant variability among sites, with the highest CF values observed for Hg (4.57) and Sr (3.09). PMF analysis identified five primary sources of HM pollution, including agricultural activities (e.g., fertilizers and pesticides), industrial emissions (e.g., metal extraction and coal combustion), traffic-related sources (e.g., vehicle emissions and tire wear), domestic waste, and natural origins (e.g., soil erosion and rock weathering). These findings collectively underscore the urgent need for targeted pollution control measures in industrial and agricultural areas. This study provides crucial insights for developing strategies to manage and mitigate HM pollution, thereby protecting water resources and public health in the Ningxia section of the Yellow River basin. Future research should focus on longitudinal studies to monitor the effectiveness of implemented pollution control measures and explore advanced remediation techniques.

References

[1] Foster N S, Noble R D, Koval C A. Reversible photoreductive deposition and oxidative dissolution of copper ions in titanium dioxide aqueous suspensions. Environmental Science & Technology, 1993, 27(2): 350–356

[2] Nriagu J O, Pacyna J M. Quantitative assessment of worldwide contamination of air, water and soils by trace metals. Nature, 1988, 333(6169): 134–139

[3] Kim K, Keller A A, Yang J. Removal of heavy metals from aqueous solution using a novel composite of recycled materials. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 2013, 425: 6–14

[4] İslamoğlu S, Yılmaz L. Effect of ionic strength on the complexation of polyethyleneimine (PEI) with Cd²⁺ and Ni²⁺ in polymer enhanced ultrafiltration (PEUF). Desalination, 2006, 200(1–3): 288–289

[5] Molinari R, Poerio T, Argurio P. Selective separation of copper(II) and nickel(II) from aqueous media using the complexation-ultrafiltration process. Chemosphere, 2008, 70(3): 341–348

[6] Fornari P C A, Abbruzzese C. Copper and nickel selective recovery by electrowinning from electronic and galvanic industrial solutions. Hydrometallurgy, 1999, 52(3): 209–222

[7] Bijmans M F M, van Helvoort P J, Dar S A, Dopson M, Lens P N L, Buisman C J N. Selective recovery of nickel over iron from a nickel-iron solution using microbial sulfate reduction in a gas-lift bioreactor. Water Research, 2009, 43(3): 853–861

[8] Sahinkaya E, Gungor M, Bayrakdar A, Yucesoy Z, Uyanik S. Separate recovery of copper and zinc from acid mine drainage using biogenic sulfide. Journal of Hazardous Materials, 2009, 171(1–3): 901–906

[9] Mu Y, Rozendal R A, Rabaey K, Keller J. Nitrobenzene removal in bioelectrochemical systems. Environmental Science & Technology, 2009, 43(22): 8690–8695

[10] Logan B E, Cheng S, Watson V, Estadt G. Graphite fiber brush anodes for increased power production in air-cathode microbial fuel cells. Environmental Science & Technology, 2007, 41(9): 3341–3346

[11] Hu H Q, Fan Y Z, Liu H. Hydrogen production using single-chamber membrane-free microbial electrolysis cells. Water Research, 2008, 42(15): 4172–4178

[12] Tao H C, Liang M, Li W, Zhang L J, Ni J R, Wu W M. Removal of copper from aqueous solution by electrodeposition in cathode chamber of microbial fuel cell. Journal of Hazardous Materials, 2011, 189(1–2): 186–192

[13] Wang Z J, Lim B, Choi C. Removal of Hg²⁺ as an electron acceptor coupled with power generation using a microbial fuel cell. Bioresource Technology, 2011, 102(10): 6304–6307

[14] Tao H C, Gao Z Y, Ding H, Xu N, Wu W M. Recovery of silver from silver(I)-containing solutions in bioelectrochemical reactors. Bioresource Technology, 2012, 111: 92–97

[15] Qin B, Luo H, Liu G, Zhang R, Chen S, Hou Y, Luo Y. Nickel ion removal from wastewater using the microbial electrolysis cell. Bioresource Technology, 2012, 121: 458–461

[16] Modin O, Wang X, Wu X, Rauch S, Fedje K K. Bioelectrochemical recovery of Cu, Pb, Cd, and Zn from dilute solutions. Journal of Hazardous Materials, 2012, 235–236: 291–297

[17] Cheng S, Xing D, Call D F, Logan B E. Direct biological conversion of electrical current into methane by electromethanogenesis. Environmental Science & Technology, 2009, 43(10): 3953–3958

[18] Ter Heijne A, Liu F, Weijden R V D, Weijma J, Buisman C J N, Hamelers H V M. Copper recovery combined with electricity production in a microbial fuel cell. Environmental Science & Technology, 2010, 44(11): 4376–4381

[19] Liu H, Cheng S, Logan B E. Power generation in fed-batch microbial fuel cells as a function of ionic strength, temperature, and reactor configuration. Environmental Science & Technology, 2005, 39(14): 5488–5493

[20] Luo H, Xu P, Roane T M, Jenkins P E, Ren Z. Microbial desalination cells for improved performance in wastewater treatment, electricity production, and desalination. Bioresource Technology, 2012, 105: 60–66

Acknowledgements

This work was partly supported by the National Natural Science Foundation of China (Grant Nos. 51039007, 51179212, and 51278500), the program of Guangzhou Science & Technology Department (No. 2012J4300115), and the research fund program of Guangdong provincial key laboratory of environmental pollution control and remediation technology (No. 2013K0002); the research fund program of key laboratory of water and air pollution control of Guangdong Province (No. GD2012A01).

Supplementary material is available in the online version of this article at http://dx.doi.org/10.1007/s11783-014-0633-5 and is accessible for authorized users.

RIGHTS & PERMISSIONS

2014 Higher Education Press and Springer-Verlag Berlin Heidelberg