Abstract

Dissolved organic matter (DOM) has an important impact on the water treatment and reuse of petroleum refinery wastewater. In order to improve the treatment efficiency, it is necessary to understand the chemical composition of the DOM in the treatment processes. In this paper, the molecular composition of DOM in wastewater samples from a representative refinery were characterized. The transformation of various compounds along the wastewater treatment processes was investigated. A total of 61 heteroatomic class species were detected from the DOM extracts, in which CHO (molecules composed of carbon, hydrogen, and oxygen atoms) and CHOS (CHO molecules that also contained sulfur) class species were the most abundant and account for 78.43% in relative mass peak abundance. The solid phase extraction DOM from the dichloromethane unextractable fraction exhibited a more complex molecular composition and contained more oxygen atoms than in the dichloromethane extract. During wastewater treatment processes, the chemical oxygen demand (COD) and ammonia-nitrogen were reduced by more than 90%. Volatile organic compounds (VOCs) accounted for about 30% of the total COD, in which benzene and toluene were dominant. After biochemical treatment, the VOCs were effectively removed but the molecular diversity of the DOM was increased and new compounds were generated. Sulfur-containing class species were more recalcitrant to biodegradation, so the origin and transformation of these compounds should be the subject of further research.

HIGHLIGHTS

  • Chemical composition of DOM in a petroleum refinery wastewater was comprehensively characterized by multiple techniques.

  • Molecular composition was characterized by high-resolution mass spectrometry and GC-MS.

  • Molecular transformation along the water processing stream was investigated.

Graphical Abstract

Graphical Abstract
Graphical Abstract

INTRODUCTION

There is a growing interest in water reuse in industry because of the increasing water scarcity, droughts, and population numbers. Petroleum-refining operations generate a significant volume of treated wastewater. The wastewater produced by a petroleum refinery contains a complex composition of organic pollutants and is refractory in terms of degradation (Rahman & Al-Malack 2006; Abdelwahab et al. 2009; Yang et al. 2013; Yan et al. 2014). The dissolved organic matter (DOM) in the wastewater causes the coloration and odor of the water and leads to membrane fouling during reclamation treatment. It also affects the removing of metal ions by complexation. Furthermore, DOM can be a major precursor of carcinogenic disinfection by-products (Chon & Cho 2016; Wang & Chen 2018; Komatsu et al. 2020).

Refineries generally have many problems relevant to wastewater treatment, such as frequent fluctuations of the operation, low treatment efficiency, and difficulties in recycling water resources (Michael-Kordatou et al. 2015). The water quality can be characterized by using machine learning models (Alizadeh et al. 2018; Shamshirband et al. 2018; Asadi et al. 2019), but they cannot provide molecular composition information for pollutants in wastewater.

The removal of DOM depends on the type and amount of organic compounds present (Michael-Kordatou et al. 2015). DOM in refinery wastewater includes natural organic matter (NOM) and organic matter produced in petroleum processing (Nebbioso & Piccolo 2013; Li et al. 2015b). Some humic-like substances in the DOM are recalcitrant during the wastewater treatment processes (Wang & Chen 2018). Therefore, a better understanding of the chemical composition of DOM in treated wastewater effluents is needed. The analysis of the migration and transformation path of pollutants in the refinery wastewater provides a basis for improving the efficiency of wastewater treatment, optimizing the treatment process, and saving energy.

Bulk properties, such as elemental composition, chemical oxygen demand (COD), total organic carbon (TOC), and biological oxygen demand (BOD), can be used to describe the macroscopic properties of DOM (Diya'uddeen et al. 2011). Mass spectrometry techniques, such as gas chromatography-mass spectrometry (GC-MS) (Templier et al. 2005; Parsi et al. 2007) and liquid chromatography-mass spectrometry (LC-MS) (Reemtsma & These 2005; Mawhinney et al. 2009), have been used to analyze the molecular composition and structure of DOM (Janoš 2003; Matilainen et al. 2011). With ultra-high mass resolution and mass accuracy, Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) can accurately determine various combinations of the elements C, H, N, O, S and their main isotopes, making it possible to identify the molecular composition of complex compounds based on accurate molecular weight. FT-ICR MS has been used in the research of DOM in seawater (Hansen et al. 2019), river water, lake water (Liu et al. 2020), groundwater (Valle et al. 2020), glacial water (Smith et al. 2018; Chen et al. 2019a, 2019b) and so on. FT-ICR MS has been widely used to reveal the molecular composition of DOM in natural aquatic systems (Koch et al. 2007; Sleighter & Hatcher 2007; Hertkorn et al. 2008), but has received little attention for the characterization of DOM in refinery wastewater treatment systems.

In our previous studies, (Li et al. 2015a, 2015b; Fang et al. 2017) electrospray ionization (ESI) coupled with FT-ICR MS has been successfully applied to the characterization of the DOM extracted from the refinery wastewater. The refinery was processing heavy crude oil, which produces wastewater containing large amounts of heteroatom oil components and leads to challenges for wastewater treatment.

In this study, a more representative refinery wastewater treatment plant was selected. In addition, a more comprehensive characterization of the molecular composition was conducted. By comparing the COD values of the samples, the proportion of volatile organic compounds (VOCs) in the pollutants can be judged macroscopically. The species and content of VOCs were directly characterized by purge and trap gas chromatography-mass spectrometry (P&T-GC-MS). In addition, the wastewater samples were separated into the organic phase and the aqueous phase and then characterized by negative-ion ESI FT-ICR MS. The purpose of this study was to investigate the molecular transformation of DOM in wastewater from a refinery to provide an overall assessment and guidance for the design and improvement of wastewater treatment process.

EXPERIMENTAL SECTION

Samples

Wastewater samples were obtained from a PetroChina refinery, which is one of the largest refineries in China. In contrast to the refineries that we have previously studied (Li et al. 2015a, 2015b), this refinery processes light crude oil. Figure 1 shows the flow chart of the refinery wastewater treatment processes. The whole treatment consists of oil separation, dissolved air flotation, biochemical degradation and sedimentation, which represent the typical refinery wastewater treatment process. The oil separations include a DCI (Degremont® circular interceptor) separator and an API (America Petroleum Institute) separator. The two step flotations are both dissolved air flotation. The biodegradation is an anaerobic-anoxic-oxic (A2/O) system.

Samples #1–8 were sampled from the plant in April 2018 and filtered in laboratory with a 0.45 μm Pall Supor membrane, then kept in dark at 4 °C before analysis. Samples #1: raw wastewater; #2: effluent from oil separation; #3: effluent from the dissolved air flotation; #4: effluent after the secondary de-oiling; #5: effluent from the secondary flotation; #6: effluent after the biochemical degradation and sedimentation; #7: effluent from the secondary sedimentation; and #8: final discharge export.

Figure 1

Flowchart of the refinery wastewater treatment processes and sampling sites of this study.

Figure 1

Flowchart of the refinery wastewater treatment processes and sampling sites of this study.

Sample preparation

Figure 2 shows the scheme of sample preparation and characterization. The refinery wastewater (300 mL) was extracted with 50 mL of dichloromethane (DCM) which had been purified by distillation with a spinning-band distillation apparatus (B/R 9600, B/R Instrument Corp, USA). The DCM extracts were combined together and the organic phase DOM was obtained by removing DCM in a rotary evaporator. DOM in the aqueous phase was further extracted with an SPE cartridge (Li et al. 2015a). Briefly, the SPE cartridge (Agilent, Bond Elut PPL, 500 mg, 6 mL, USA) was rinsed with 20 mL methanol (purified by distillation before use) followed by 20 mL acidified ultrapure water (pH = 2, LC-MS grade). Then, the wastewater was acidified to pH = 2 with HCl (ACS, Merk, Germany) and then injected into the cartridge. Next, the cartridge was washed with 20 mL acidified ultrapure water to remove the salt and then dried with N2 gas flow. Finally, the adsorbed DOM was eluted out of the cartridge with 10 mL methanol and stored at −18 °C in the dark.

Figure 2

Schematic diagram for DOM analyses of the refinery wastewater.

Figure 2

Schematic diagram for DOM analyses of the refinery wastewater.

COD and NH3-N analysis

The COD and NH3-N of all eight samples were analyzed with the U.S. Environmental Protection Agency digestion colorimetry of HACH® standard method. This method has the advantages of being quick and easy, giving accurate measurements, and having low reagent consumption.

After sufficient aeration, each sample was tested for three sets of parallel samples to obtain an average value. The average values of eight samples before and after aeration are shown in Table S1 (see Supplementary Information).

P&T GC-MS analysis

The P&T process was conducted using an Atomx 15-0000-200 automatic purge-capture sampling device (Teledyne Tekmar, USA). Operating conditions were as follows: valve oven temperature: 140 °C; transfer line temperature: 140 °C; sample mount temperature: 90 °C; water heater temperature: 90 °C; purge time: 11.0 min; purge flow 40 mL/min.

The VOCs in the refinery wastewater samples were analyzed with an Agilent 6890N gas chromatograph coupled with a 5975C mass spectrometer. A DB-5 MS UI column (60 m × 0.25mm × 1.0 μm, Agilent, USA) was used. The inlet temperature was 260 °C. The initial column temperature of the GC oven was programmed to rise from 40 °C to 260 °C at a rate of 5 °C/min, then held constant at 260 °C for 3 min. Helium (purity >99.999) was used as the carrier gas at a flow rate of 1 mL/min with a split ratio of 20:1. The 70 eV electron impact ion source was operated at a temperature of 235 °C; the quadrupole was running in full scan mode with a mass range of m/z 30–300 in a 1 second scan period.

Negative-ion ESI FT-ICR MS analysis

The DOM was analyzed using a Bruker Apex Ultra FT-ICR mass spectrometer equipped with a 9.4 T superconducting magnet. The Apollo II electrospray ion source was running in negative mode. The separated organic and aqueous phase SPE-DOM dissolved in methanol (50 mg/mL) were introduced by a syringe pump through the electrospray source at a rate of 250 μL/h. The typical operating conditions were: spray shield voltage, 3.8 kV; capillary column introduced voltage, 4.3 kV; capillary column end voltage, −300 V. The data size was set to 2 M words over a mass range of m/z 200–700. In order to enhance the signal-to-noise ratio and dynamic range, a data set of 128 FT-ICR scans were accumulated.

The mass calibration and data acquisition of FT-ICR MS have been described elsewhere (Shi et al. 2010; Pan et al. 2020). Data analysis was performed by using in-house software. Mass peaks with a signal-to-noise ratio ≥6 were considered for molecular formula assignment. The average mass error of all assigned molecular formulas was less than 1 ppm. To obtain semi-quantitative results, 5 × 10−6 mol/L equal deuterated stearic acid (C18D35H1O2) was added to each DOM sample as an internal standard. The peak intensity of each identified molecular formula was normalized to the internal standard in each sample. The ratio of intensity of each peak to the internal standard was defined as the relative abundance. The elemental combinations of molecular formulas were limited to those containing 12C0-100, 1H2-200, 14N0-3, 16O0-30, and 32S0-4 atoms.

FT-ICR MS data related parameters integration were as follows: A modified aromaticity index (AImod) and double bond equivalents (DBE) were calculated for each assigned molecular formula according to Koch and Dittmar (Koch & Dittmar 2006). The intensity-weighted average of molecular formulas (CHO, CHON, CHOS, CHONS, i.e. molecules composed with carbon, hydrogen, and oxygen atoms, plus nitrogen, sulfur, or nitrogen and sulfur), and other parameters (hydrogen/carbon ratio (H/C), oxygen/carbon ratio (O/C), DBE, and AImod) were calculated for each SPE-DOM sample (Fang et al. 2019). Compound classification were defined by aromaticity index (AI) and elemental ratios (O/C and H/C) (Dittmar & Koch 2006; Kellerman et al. 2014): Saturated compounds (H/C > 2), aliphatic compounds (2.0 ≥ H/C > 1.5) (Li et al. 2018), highly unsaturated and phenolic compounds (AI < 0.50, H/C < 1.5), polyphenol-like compounds (0.66 ≥ AI ≥ 0.50), and polycyclic aromatic-like compounds (AI ≥ 0.66) (Liu et al. 2010).

RESULTS AND DISCUSSION

Bulk properties

The COD and NH3-N values of the samples are listed in Table 1. TOC is not listed because it was hard to obtain a credible value due to the high inorganic carbon present in the raw water and the loss of volatile compounds in the analysis. The COD values are lower than in refinery wastewater (Xia et al. 2000; Chao et al. 2019) and coal gasification wastewater (Fang et al. 2019; Wang et al. 2019), but are similar to chemical wastewater (Guan et al. 2019; Xia & Xia 2019; Zhang et al. 2019a, 2019b, 2019c) and domestic sewage (Lu et al. 2019).

The organic matter content in the wastewater is low compared to other refinery wastewater because the refinery processes light crude oil. The raw wastewater COD was controlled below 500 mg/L. In the whole process, the most effective way to remove COD was biochemical treatment. After a series of treatment processes, the COD and NH3-N of effluent were about 30 mg/L and 0.01 mg/L. During the wastewater treatment processes, the COD and NH3-N reduced by more than 90%. The effluent met the requirement of the Emission Standard of Pollutants for the Petroleum Chemistry Industry GB31571-2015 (China), in which the COD is less than 60 mg/L.

Table 1

COD and NH3-N of the wastewater (mg/L)

Samples#1#2#3#4#5#6#7#8
COD 442 482 413 376 361 51 44 30 
NH3-N 37.8 53.0 40.5 28.6 26.3 2.9 0.33 0.01 
Samples#1#2#3#4#5#6#7#8
COD 442 482 413 376 361 51 44 30 
NH3-N 37.8 53.0 40.5 28.6 26.3 2.9 0.33 0.01 

Characterization of VOCs by P&T GC-MS

Numerous VOCs have been confirmed as human carcinogens. Owing to the collection and treatment of wastewater, wastewater treatment plants have been identified as a primary VOC emission source (Cheng & Chou 2003). VOCs are defined as compounds having a boiling point of 50–250 °C and a relatively low molecular weight. Table S1 (see Supplementary Information) shows the changes in COD before and after aeration. The COD value decreased by about 1/3 in the processing and VOCs in the refinery wastewater accounted for about 30% of the total COD contribution.

Figure 3 shows the total ion chromatograms of the VOCs of the wastewater. A total of 215 individual compounds, including alkanes, alkenes, alkynes, aromatic hydrocarbons, polycyclic aromatic hydrocarbons, organic phenols, alcohols, ketones, esters, thioethers, and thiophenes (see Supplementary Information Table S2) were tentatively identified. The most abundant compounds were benzene and toluene, followed by acetone and methyl ethyl ketone. Relative contents of benzene, toluene, 2-butanone, and acetone were 23.03%, 21.23%, 12.16%, and 9.55%, respectively. Alkanes were removed after air flotation. No VOCs were detected after biochemical treatment. It shows that most of the VOCs can be removed by sufficient aeration, biodegradation, and sludge interception (Malakar et al. 2017; Zhang et al. 2019a, 2019b, 2019c). The VOC concentration distribution was very similar to that assessed in the literature (Cheng & Chou 2003).

Figure 3

Total ion GC-MS chromatograms of the VOCs in the wastewater.

Figure 3

Total ion GC-MS chromatograms of the VOCs in the wastewater.

Molecular characterization of DCM-extracted DOM by FT-ICR MS

DCM-extracted DOM was characterized by negative-ion ESI, and a molecular weight distribution range of 200–600 was observed in the mass spectra (see Figure S1 in the Supplementary Information). Weakly polar and non-polar compounds are mainly present in the DCM phase. There are very few mass spectral peaks in the DCM phase compared to the aqueous phase. There were 349 molecular formulas in the raw wastewater, but 3,645 molecular formulas were detected after biochemical treatment. An increase was shown in the number of peaks after biochemical treatment, indicating the increase in molecular diversity after biochemical treatment (Chen et al. 2014a, 2014b, 2015).

Figure 4 shows the semi-quantitative relative abundance of class species identified in the DCM phase DOM, including O2-O9, N1O2-N1O8, N2O2-N2O7, O3S1-O7S1 and N1O3S1-N1O7S1. The vertical coordinate represents the sum of the intensity of different DBEs of a compound. The O2 and O3S1 class species were dominant in all wastewater samples. The O2 compounds had the highest abundance in biochemical processes, presumably due to hydrocarbon oxidation in biochemical processes (Wang et al. 2014; Zhang et al. 2019a, 2019b, 2019c; Zhao et al. 2019). The O3S1 class species with four DBEs could be an alkylbenzene sulfonic acid (Li et al. 2015a). It was likely to be derived from an alkylbenzene sulfonate plasma surfactant (Gonsior et al. 2011; Geng et al. 2018). The relative abundance of O3S1 class species increased significantly in oil separation, air flotation and biochemical treatment, but decreased after the second sedimentation plant. It is speculated that the flocculants partially adsorb the surfactants in the secondary sedimentation tank (Alizadeh et al. 2017; Duggan et al. 2019; Shende & Chau 2019). The reason for the increase is the easily ionized selectivity of ESI for surfactants or microbial production during biodegradation (Mohamed et al. 2015; Phungsai et al. 2016; Song et al. 2018).

Figure 4

Relative abundance of heteroatom class species of the DOM in the DCM phase for eight refinery wastewaters obtained by negative-ion ESI FT-ICR MS.

Figure 4

Relative abundance of heteroatom class species of the DOM in the DCM phase for eight refinery wastewaters obtained by negative-ion ESI FT-ICR MS.

Molecular characterization of aqueous phase DOM by FT-ICR MS

DOM is rich in oxygen-containing polar compounds, which is suitable for characterization by negative-ion ESI (Pan et al. 2020). Figure S2 (see Supplementary Information) shows the broad-band ESI FT-ICR mass spectra of the aqueous phase DOM. The abundant peaks with m/z 297, 311, and 325 in the mass spectra of #1 to #5 were assigned as O3S1 classes with four DBEs, which is presumed to be an alkyl-benzene sulfonic acid surfactant (Bahri et al. 2018; Geng et al. 2018). In addition, the dominant peaks with m/z values greater than 440 in the mass spectrum of #6 were identified as polymers. The molecular weights of these polymer molecules differ by 44 Da, which corresponds to polyoxyethylene ether (Pei et al. 2016; Cadar et al. 2017; Fedeila et al. 2018; Chen et al. 2019a, 2019b). They were flocculated by flocculants after the second sedimentation plant.

Figure 5 shows the relative abundance of compounds assigned in the mass spectra of the aqueous phase DOM. A total of 61 compound classes were detected in the negative-ion ESI FT-ICR mass spectra for each wastewater sample, including O2-15, N1O2-13, N2O3-12, O3-13S1 and N1O3-12S1 class species. The type of polar compounds was consistent with other refineries (Li et al. 2015a, 2015b; Fang et al. 2019). The relative abundance of all class species in petrochemical refinery wastewater decreases with the overall treatment processes. But traditional biological treatment of refractory wastewater has a low efficiency (Chen et al. 2014a). It can effectively remove small molecular hydrocarbons (Tran et al. 2015), whereas some organic contaminants with high molecular weight and complex molecular structure are found to be difficult to remove (Dai et al. 2016). The compounds with many oxygen atoms may include hydrophilic chemical groups such as carboxyl and phenol groups, which are commonly identified under negative-ion ESI mode (Shon et al. 2006; Shi et al. 2010). For CHOS class species, the O3S1 class species was dominant and almost removed by the biochemical treatment.

Figure 5

Relative abundance of assigned classes of DOM in the aqueous phase from the refinery wastewater.

Figure 5

Relative abundance of assigned classes of DOM in the aqueous phase from the refinery wastewater.

The relative abundance of the CHON class and the CHONS class of all eight wastewater samples are shown in Figures S3 and S4 (see Supplementary Information). The relative abundance of OxS1, N1OxS1 class species decreased gradually with the treatment process, implying each step of the treatment was effective for sulfur-containing compounds (Geng et al. 2018; Fang et al. 2019). The relative abundance of Ox and N1Ox class species decreased after de-oiling and flotation but increased after biochemical treatment.

To compare the composition of DOM in the aqueous phase of all eight samples, the percentage contributions of individual compound classes normalized to the total ion signals were calculated and are presented in Table 2. In the raw wastewater, CHO and CHOS classes were the most abundant and accounted for 78.43% in terms of relative abundance. Molecular formulas totaling 3,501 to 5,012 were found in the eight refinery samples (Table 2). CHOS class species were found in all samples, which accounted for 21.94%–45.90% of all assigned molecular formulas and showed a decreasing trend during the water treatment processes. From #1 to #5, CHOS species accounted for a higher proportion than CHO species. After biochemical treatment, CHO species accounted for the highest proportion (29.67%–46.73%) but CHOS class species were reduced. These compounds may come from the oil feed itself and can be effectively removed (Fang et al. 2019). The AIs of CHOS and CHO species are 0.20–0.29 and 0.28–0.43, respectively.

Table 2

Molecular formulas assigned from ESI FT-ICR mass spectra of DOM in the aqueous phase and their average H/C, O/C, DBE, AI and percentage

SampleTypeNumberH/CavgO/CavgDBEavgaAIavgbPercentage
#1 CHO 1,340 1.21 0.37 8.37 0.36 38.27 
 CHON 341 1.13 0.38 9.24 0.44 9.74 
 CHOS 1,406 1.46 0.29 5.71 0.22 40.16 
 CHONS 414 1.19 0.38 8.27 0.36 11.83 
 Total 3,501 1.37 0.32 6.65 0.27 100 
#2 CHO 1,692 1.21 0.35 8.41 0.37 39.87 
 CHON 477 1.13 0.36 9.54 0.44 11.24 
 CHOS 1,504 1.52 0.30 5.31 0.20 35.44 
 CHONS 571 1.22 0.36 8.34 0.35 13.45 
 Total 4,244 1.39 0.32 6.56 0.27 100 
#3 CHO 1,126 1.09 0.39 9.33 0.43 29.67 
 CHON 268 1.02 0.39 10.21 0.51 7.06 
 CHOS 1,742 1.32 0.35 7.00 0.29 45.90 
 CHONS 659 1.11 0.40 9.22 0.41 17.36 
 Total 3,795 1.25 0.36 7.69 0.33 100 
#4 CHO 1,405 1.18 0.36 8.59 0.38 37.81 
 CHON 374 1.11 0.37 9.53 0.45 10.06 
 CHOS 1,453 1.43 0.29 5.96 0.24 39.10 
 CHONS 484 1.22 0.37 8.15 0.35 13.02 
 Total 3,716 1.35 0.32 6.89 0.29 100 
#5 CHO 1,385 1.23 0.36 8.12 0.35 37.77 
 CHON 346 1.16 0.36 8.90 0.43 9.44 
 CHOS 1,469 1.41 0.32 6.15 0.24 40.06 
 CHONS 467 1.21 0.38 8.12 0.35 12.74 
 Total 3,667 1.34 0.33 6.92 0.29 100 
#6 CHO 2,317 1.33 0.37 7.46 0.28 46.73 
 CHON 1,137 1.24 0.41 8.47 0.35 22.93 
 CHOS 1,088 1.33 0.38 6.46 0.25 21.94 
 CHONS 416 1.31 0.38 6.89 0.28 8.39 
 Total 4,958 1.32 0.38 7.36 0.28 100 
#7 CHO 2,146 1.29 0.35 7.87 0.31 42.82 
 CHON 1,047 1.24 0.40 8.62 0.35 20.89 
 CHOS 1,293 1.33 0.34 6.82 0.26 25.80 
 CHONS 526 1.23 0.38 7.94 0.33 10.49 
 Total 5,012 1.30 0.35 7.61 0.30 100 
#8 CHO 1,992 1.30 0.33 7.92 0.31 41.23 
 CHON 1,036 1.18 0.39 9.07 0.41 21.44 
 CHOS 1,253 1.30 0.32 7.29 0.29 25.94 
 CHONS 550 1.23 0.40 8.03 0.33 11.38 
 Total 4,831 1.29 0.34 7.85 0.31 100 
SampleTypeNumberH/CavgO/CavgDBEavgaAIavgbPercentage
#1 CHO 1,340 1.21 0.37 8.37 0.36 38.27 
 CHON 341 1.13 0.38 9.24 0.44 9.74 
 CHOS 1,406 1.46 0.29 5.71 0.22 40.16 
 CHONS 414 1.19 0.38 8.27 0.36 11.83 
 Total 3,501 1.37 0.32 6.65 0.27 100 
#2 CHO 1,692 1.21 0.35 8.41 0.37 39.87 
 CHON 477 1.13 0.36 9.54 0.44 11.24 
 CHOS 1,504 1.52 0.30 5.31 0.20 35.44 
 CHONS 571 1.22 0.36 8.34 0.35 13.45 
 Total 4,244 1.39 0.32 6.56 0.27 100 
#3 CHO 1,126 1.09 0.39 9.33 0.43 29.67 
 CHON 268 1.02 0.39 10.21 0.51 7.06 
 CHOS 1,742 1.32 0.35 7.00 0.29 45.90 
 CHONS 659 1.11 0.40 9.22 0.41 17.36 
 Total 3,795 1.25 0.36 7.69 0.33 100 
#4 CHO 1,405 1.18 0.36 8.59 0.38 37.81 
 CHON 374 1.11 0.37 9.53 0.45 10.06 
 CHOS 1,453 1.43 0.29 5.96 0.24 39.10 
 CHONS 484 1.22 0.37 8.15 0.35 13.02 
 Total 3,716 1.35 0.32 6.89 0.29 100 
#5 CHO 1,385 1.23 0.36 8.12 0.35 37.77 
 CHON 346 1.16 0.36 8.90 0.43 9.44 
 CHOS 1,469 1.41 0.32 6.15 0.24 40.06 
 CHONS 467 1.21 0.38 8.12 0.35 12.74 
 Total 3,667 1.34 0.33 6.92 0.29 100 
#6 CHO 2,317 1.33 0.37 7.46 0.28 46.73 
 CHON 1,137 1.24 0.41 8.47 0.35 22.93 
 CHOS 1,088 1.33 0.38 6.46 0.25 21.94 
 CHONS 416 1.31 0.38 6.89 0.28 8.39 
 Total 4,958 1.32 0.38 7.36 0.28 100 
#7 CHO 2,146 1.29 0.35 7.87 0.31 42.82 
 CHON 1,047 1.24 0.40 8.62 0.35 20.89 
 CHOS 1,293 1.33 0.34 6.82 0.26 25.80 
 CHONS 526 1.23 0.38 7.94 0.33 10.49 
 Total 5,012 1.30 0.35 7.61 0.30 100 
#8 CHO 1,992 1.30 0.33 7.92 0.31 41.23 
 CHON 1,036 1.18 0.39 9.07 0.41 21.44 
 CHOS 1,253 1.30 0.32 7.29 0.29 25.94 
 CHONS 550 1.23 0.40 8.03 0.33 11.38 
 Total 4,831 1.29 0.34 7.85 0.31 100 

aDBE values were defined as: DBE = C-1/2H + 1/2N + 1.

bAI was the aromaticity index values and determined as: AI = [1 + C-1/2O-S-1/2H]/[C-1/2O-S-N-P] (Kim et al. 2003; Stenson et al. 2003; Koch & Dittmar 2016).

The van Krevelen (VK) diagram is usually used to illustrate the difference in molecular composition between wastewaters (Koch & Dittmar 2006; Sleighter & Hatcher 2007; Seidel et al. 2014; Yuan et al. 2017). Figure S5 (see Supplementary Information) shows the VK diagrams of CHO class species for the aqueous phase DOM of all eight samples. Colors represent different relative intensities. The darker the color, the greater the intensity. The peak intensity of each identified molecular formula was normalized to the peak intensity of the internal standard in each sample. The result can be drawn from the distribution of the different dots in the VK diagrams: (1) The range of O/C ratio of CHO compounds was 0.1–0.8 and the H/C ratio was 0.4–2.0. Distribution ranges of CHON, CHOS, and CHONS compounds were 0.1–0.8 O/C with 0.4–1.8 H/C, 0.1–0.8 O/C with 0.5–2.3 H/C, and 0.15–0.75 O/C with 0.5–2.1 H/C, respectively (Figures S6, S7, S8). CHOS compounds were widely distributed in the VK diagram (Figure S7). (2) The O/C ratio of CHO after biochemical treatment (#6) showed the widest distribution (0.1–0.8), implying that biochemical treatment increased the molecular composition diversity (Chen et al. 2014a, 2014b, 2015). (3) The H/C > 1.5 species, which correspond to lipids, proteins, and carbohydrates in natural DOM, decreased after biochemistry (Herzsprung et al. 2012; Šantl-Temkiv et al. 2013; Antony et al. 2014; Feng et al. 2016).

The molecular transformation during specific water treatment processes was further assessed by comparing which molecular formulas were formed or removed. The molecular changes during oil separation, air flotation and biochemical treatment were plotted in VK diagrams, as shown in Figure 6. For CHO class species, the area of O/C lower than 0.5 were removed but the area of H/C higher than about 1.5 were produced after biochemical treatment. The newly formed compounds included aliphatic compounds (2.0 ≥ H/C > 1.5) and highly unsaturated and phenolic compounds (AI < 0.50, H/C < 1.5). In the meantime, polyphenol-like compounds (0.66 ≥ AI ≥ 0.50) and polycyclic-aromatic-like compounds (AI ≥ 0.66) were removed. This shows that the biochemical process converts some CHO compounds with high AI values into phenols and aliphatic compounds (Dai et al. 2016). In contrast, for CHOS species, the biochemically removed molecules mainly had an O/C lower than 0.5 and the molecules produced were mainly in the range of O/C higher than about 0.5. The compounds removed after the biochemical process included aliphatic compounds (2.0 ≥ H/C > 1.5), highly unsaturated and phenolic compounds (AI < 0.50, H/C < 1.5), polyphenol-like compounds (0.66 ≥ AI ≥ 0.50) with lower O/C. But aliphatic compounds (2.0 ≥ H/C > 1.5), highly unsaturated and phenolic compounds (AI < 0.50, H/C < 1.5) with higher O/C were formed. This shows that the biochemical process removed some CHOS compounds with low oxygen numbers.

Figure 6

VK diagrams of removed and newly formed CHO (a) and CHOS (b) molecular formulas in the aqueous phase of the biochemical treatment process.

Figure 6

VK diagrams of removed and newly formed CHO (a) and CHOS (b) molecular formulas in the aqueous phase of the biochemical treatment process.

CONCLUSIONS

The DOM in refinery wastewater had a complex molecular composition, in which the small molecular VOCs accounted for about 30% of the total dissolved organic carbon. In the long processing stream, biochemical treatment was the most effective, removing more than 90% of COD and NH3-N. A total of 61 heteroatomic class species were detected in the DOM extracts, in which CHO and CHOS class species were the most abundant and accounted for 78.43% in relative abundance. The SPE-DOM from the DCM unextractable fraction exhibited a more complex molecular composition and contained more oxygen atoms than that in the DCM extract. After biochemical treatment, the VOC was effectively removed but the molecular diversity of the DOM was increased and new compounds were generated. Sulfur-containing class species were more recalcitrant to biodegradation. To develop and optimize the technique for wastewater processing, molecular selectivity of various compounds for the process should be considered. Further quantitative analysis and structure characterization of the DOM molecules in refinery wastewater is needed.

ACKNOWLEDGEMENTS

This study was supported by the Open Project Program of the State Key Laboratory of Petroleum Pollution Control (Grant No. PPC2018012) and under the institute's basic science research and strategic reserve technology research fund (2017D-5008), CNPC Research Institute of Safety and Environmental Technology.

DATA AVAILABILITY STATEMENT

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

REFERENCES

Abdelwahab
O.
Amin
N. K.
El-Ashtoukhy
E. S.
2009
Electrochemical removal of phenol from oil refinery wastewater
.
Journal of Hazardous materials
163
(
2–3
),
711
716
.
Alizadeh
M. J.
Jafari Nodoushan
E.
Kalarestaghi
N.
Chau
K. W.
2017
Toward multi-day-ahead forecasting of suspended sediment concentration using ensemble models
.
Environmental Science and Pollution Research International
24
(
36
),
28017
28025
.
Alizadeh
M. J.
Kavianpour
M. R.
Danesh
M.
Adolf
J.
Shamshirband
S.
Chau
K.-W.
2018
Effect of river flow on the quality of estuarine and coastal waters using machine learning models
.
Engineering Applications of Computational Fluid Mechanics
12
(
1
),
810
823
.
Antony
R.
Grannas
A. M.
Willoughby
A. S.
Sleighter
R. L.
Thamban
M.
Hatcher
P. G.
2014
Origin and sources of dissolved organic matter in snow on the East Antarctic ice sheet
.
Environmental Science and Technology
48
(
11
),
6151
6159
.
Asadi
E. I. M.
Samadianfard
S.
Ramli
M. F.
Mosavi
A.
Nabipour
N.
Shamshirband
S.
Hajnal
E.
Chau
K.-W.
2019
Groundwater quality assessment for sustainable drinking and irrigation
.
Sustainability
12
(
1
),
177
.
Bahri
M.
Mahdavi
A.
Mirzaei
A.
Mansouri
A.
Haghighat
F.
2018
Integrated oxidation process and biological treatment for highly concentrated petrochemical effluents: a review
.
Chemical Engineering and Processing – Process Intensification
125
,
183
196
.
Cadar
D.
Olteanu
N. L.
Rogozea
E. A.
Petcu
A. R.
Meghea
A.
Mihaly
M.
2017
Recovery of targeted hydrophilic compounds from simulated wastewaters using nonionic microemulsion systems
.
Process Safety and Environmental Protection
109
,
648
658
.
Chao
L.
Meng
J.
Wang
H.
Li
Y.
2019
Comparative study of three filler modified A_O processes for refinery wastewater treatment
.
Thchnology of Water Treatment
45
(
8
),
103
107
.
Chen
C.
Chen
H.
Guo
X.
Guo
S.
Yan
G.
2014a
Advanced ozone treatment of heavy oil refining wastewater by activated carbon supported iron oxide
.
Journal of Industrial and Engineering Chemistry
20
(
5
),
2782
2791
.
Cheng
W. H.
Chou
M. S.
2003
VOC emission characteristics of petrochemical wastewater treatment facilities in southern Taiwan
.
Journal of Environmental Science and Health, Part A: Toxic/Hazardous Substances & Environmental Engineering
38
(
11
),
2521
2535
.
Dittmar
T.
Koch
B. P.
2006
Thermogenic organic matter dissolved in the abyssal ocean
.
Marine Chemistry
102
(
3–4
),
208
217
.
Diya'uddeen
B. H.
Wan Daud
W. M. A.
Abdul Aziz
A. R.
2011
Treatment technologies for petroleum refinery effluents: a review
.
Process Safety and Environmental Protection
89
(
2
),
95
105
.
Duggan
K. L.
Morris
M.
Bhatia
S. K.
Khachan
M. M.
Lewis
K. E.
2019
Effects of cationic polyacrylamide and cationic starch on aquatic life
.
Journal of Hazardous, Toxic, and Radioactive Waste
23
(
4
), 04019022.
Fang
Z.
Li
L.
Jiang
B.
He
C.
Li
Y.
Xu
C.
Shi
Q.
2019
Molecular composition and transformation of dissolved organic matter (DOM) in coal gasification wastewater
.
Energy & Fuels
33
(
4
),
3003
3011
.
Fedeila
M.
Hachaichi-Sadouk
Z.
Bautista
L. F.
Simarro
R.
Nateche
F.
2018
Biodegradation of anionic surfactants by Alcaligenes faecalis, Enterobacter cloacae and Serratia marcescens strains isolated from industrial wastewater
.
Ecotoxicology and Environment Safety
163
,
629
635
.
Geng
C.
Cao
N.
Xu
W.
He
C.
Yuan
Z.
Liu
J.
Shi
Q.
Xu
C.
Liu
S.
Zhao
H.
2018
Molecular characterization of organics removed by a covalently bound inorganic-organic hybrid coagulant for advanced treatment of municipal sewage
.
Environmental Science & Technology
52
(
21
),
12642
8
.
Gonsior
M.
Zwartjes
M.
Cooper
W. J.
Song
W.
Ishida
K. P.
Tseng
L. Y.
Jeung
M. K.
Rosso
D.
Hertkorn
N.
Schmitt-Kopplin
P.
2011
Molecular characterization of effluent organic matter identified by ultrahigh resolution mass spectrometry
.
Water Research
45
(
9
),
2943
2953
.
Hansen
C. T.
Niggemann
J.
Giebel
H. A.
Simon
M.
Bach
W.
Dittmar
T.
2019
Biodegradability of hydrothermally altered deep-sea dissolved organic matter
.
Marine Chemistry
217
, 103706.
Hertkorn
N.
Frommberger
M.
Witt
M.
Koch
B. P.
Schmitt-Kopplin
P.
Perdue
E. M.
2008
Natural organic matter and the event horizon of mass spectrometry
.
Analytical Chemistry
80
,
8908
8919
.
Herzsprung
P.
von Tümpling
W.
Hertkorn
N.
Harir
M.
Büttner
O.
Bravidor
J.
Friese
K.
Schmitt-Kopplin
P.
2012
Variations of DOM quality in inflows of a drinking water reservoir: linking of van Krevelen diagrams with EEMF spectra by rank correlation
.
Environmental Science & Technology
46
(
10
),
5511
5518
.
Kellerman
A. M.
Dittmar
T.
Kothawala
D. N.
Tranvik
L. J.
2014
Chemodiversity of dissolved organic matter in lakes driven by climate and hydrology
.
Nature Communications
5
,
3804
.
Koch
B. P.
Dittmar
T.
2006
From mass to structure: an aromaticity index for high-resolution mass data of natural organic matter
.
Rapid Communications in Mass Spectrometry
20
(
5
),
926
932
.
Li
X.
Sun
G.
Chen
S.
Fang
Z.
Yuan
H.
Shi
Q.
Zhu
Y.
2018
Molecular chemodiversity of dissolved organic matter in paddy soils
.
Environmental Science & Technology
52
(
3
),
963
971
.
Liu
Y.
Ye
Q.
Huang
W. L.
Feng
L.
Wang
Y. H.
Xie
Z.
Yong
S. S.
Zhang
S.
Jiang
B.
Zheng
Y.
Wang
J. J.
2020
Spectroscopic and molecular-level characteristics of dissolved organic matter in the Pearl River Estuary, South China
.
Science of the Total Environment
710
,
136307
.
Lu
M.
Zhang
W.
Yu
Y.
Miu
Q.
2019
ABR and ornamental aquatic artificial combined treatment of domestic sewage
.
Guangzhou Chemical Industry
47
(
14
),
85
88
.
Malakar
S.
Saha
P. D.
Baskaran
D.
Rajamanickam
R.
2017
Comparative study of biofiltration process for treatment of VOCs emission from petroleum refinery wastewater – a review
.
Environmental Technology & Innovation
8
,
441
461
.
Matilainen
A.
Gjessing
E. T.
Lahtinen
T.
Hed
L.
Bhatnagar
A.
Sillanpää
M.
2011
An overview of the methods used in the characterisation of natural organic matter (NOM) in relation to drinking water treatment
.
Chemosphere
83
(
11
),
1431
1442
.
Mawhinney
D. B.
Rosario-Ortiz
F. L.
Baik
S.
Vanderford
B. J.
Snyder
S. A.
2009
Characterization of fulvic acids by liquid chromatography-quadrupole time-of-flight mass spectrometry
.
Journal of Chromatography A
1216
(
9
),
1319
1324
.
Michael-Kordatou
I.
Michael
C.
Duan
X.
He
X.
Dionysiou
D. D.
Mills
M. A.
Fatta-Kassinos
D.
2015
Dissolved effluent organic matter: characteristics and potential implications in wastewater treatment and reuse applications
.
Water Research
77
,
213
248
.
Mohamed
M. H.
Wilson
L. D.
Shah
J. R.
Bailey
J.
Peru
K. M.
Headley
J. V.
2015
A novel solid-state fractionation of naphthenic acid fraction components from oil sands process-affected water
.
Chemosphere
136
,
252
258
.
Nebbioso
A.
Piccolo
A.
2013
Molecular characterization of dissolved organic matter (DOM): a critical review
.
Analytical and Bioanalytical Chemistry
405
(
1
),
109
124
.
Parsi
Z.
Hartog
N.
Górecki
T.
Poerschmann
J.
2007
Analytical pyrolysis as a tool for the characterization of natural organic matter – a comparison of different approaches
.
Journal of Analytical and Applied Pyrolysis
79
(
1–2
),
9
15
.
Pei
L.
Wu
P.
Liu
J.
Wang
J.
2016
Effect of nonionic surfactant on the micro-emulsifying water in silicone media
.
Journal of Surfactants and Detergents
20
(
1
),
247
254
.
Šantl-Temkiv
T.
Finster
K.
Dittmar
T.
Hansen
B. M.
Thyrhaug
R.
Nielsen
N. W.
Karlson
U. G.
2013
Hailstones: a window into the microbial and chemical inventory of a storm cloud
.
PLoS One
8
(
1
),
e53550
.
Seidel
M.
Beck
M.
Riedel
T.
Waska
H.
Suryaputra
I. G. N. A.
Schnetger
B.
Niggemann
J.
Simon
M.
Dittmar
T.
2014
Biogeochemistry of dissolved organic matter in an anoxic intertidal creek bank
.
Geochimica et Cosmochimica Acta
140
,
418
434
.
Shamshirband
S.
Jafari Nodoushan
E.
Adolf
J. E.
Abdul Manaf
A.
Mosavi
A.
Chau
K. W.
2018
Ensemble models with uncertainty analysis for multi-day ahead forecasting of chlorophyll a concentration in coastal waters
.
Engineering Applications of Computational Fluid Mechanics
13
(
1
),
91
101
.
Shende
S.
Chau
K. W.
2019
Forecasting safe distance of a pumping well for effective riverbank filtration
.
Journal of Hazardous Toxic and Radioactive Waste
23
(
2
), 04018040.
Shon
H. K.
Vigneswaran
S.
Snyder
S. A.
2006
Effluent organic matter (EfOM) in wastewater: constituents, effects, and treatment
.
Critical Reviews in Environmental Science and Technology
36
(
4
),
327
374
.
Smith
H. J.
Tigges
M.
D'Andrilli
J.
Parker
A.
Bothner
B.
Foreman
C. M.
2018
Dynamic processing of DOM: insight from exometabolomics, fluorescence spectroscopy, and mass spectrometry
.
Limnology and Oceanography Letters
3
,
225
235
.
Valle
J.
Harir
M.
Gonsior
M.
Enrich-Prast
A.
Schmitt-Kopplin
P.
Bastviken
D.
Hertkorn
N.
2020
Molecular differences between water column and sediment pore water SPE-DOM in ten Swedish boreal lakes
.
Water Research
170
,
115320
.
Wang
W.
Gao
J.
Gao
J.
2019
Pretreatment_A_O_Ozone Oxidation_BAF advanced treatment of coal chemical wastewater
.
Industrial Water Treatment
39
(
6
),
103
106
.
Xia
J.
Xia
T.
2019
Commissioning and operation of petrochemical wastewater from seaport sewage treatment plant
.
Water & Waste Water Engineering
45
,
220
224
.
Xia
C.
Huang
N.
He
X.
2000
Treatment of refinery wastewater with poly ferric aluminum silicate coagulant
.
Environmental Protection of Chemical Industry
20
(
4
),
29
32
.
Zhang
M.
Shi
Q.
Song
X.
Wang
H.
Bian
Z.
2019a
Recent electrochemical methods in electrochemical degradation of halogenated organics: a review
.
Environmental Science and Pollution Research International
26
(
11
),
10457
10486
.
Zhang
M.
Zhang
J.
Zhang
S.
2019b
Pilot test_MBR_RO combined process advanced treatment of chemical wastewater
.
Hebei Journal of Industrial Science and Technology
36
(
2
),
135
141
.

Supplementary data