Elemental Mercury Adsorption by Cupric Chloride-Modified Mesoporous Carbon Aerogel

Abstract

Mesoporous carbon aerogels (MCA) synthesized via aqueous polymerization of resorcinol and formaldehyde were modified by copper chloride and applied for adsorption removal of elemental mercury (Hg⁰) at a low temperature. The sorbents were characterized by field emission scanning electron microscopy (FESEM), X-ray diffraction (XRD), nitrogen adsorption/desorption isotherms, and X-ray photoelectron spectroscopy (XPS) techniques. The results show that CuCl₂-modified MCA exhibited good ability for Hg⁰ adsorption in the 40–160 °C temperature range, with Hg⁰ removal efficiency all above 95.0%. The Hg⁰ removal efficiency first increased and then decreased with the elevation of reaction temperature. It performed optimally at 80 °C with the highest Hg⁰ removal efficiency of 98.7%. XPS results indicate that covalent chlorines (C-Cl groups) play an important role in elemental mercury adsorption process. Hg⁰ is firstly captured in the form of oxidized mercury (Hg²⁺) and then reacts with C-Cl groups to form HgCl₂.

1. Introduction

Mercury is one of the most toxic heavy metal elements in the environment due to its high volatility, long persistence, and strong bioaccumulation property [1]. Coal-fired power plants are the dominant anthropogenic mercury emission sources [2]. Elemental mercury, oxidized mercury, and particle-bound mercury are three types of mercury species in flue gas [3]. On account of its low aqueous insolubility and high vaporability, elemental mercury (Hg⁰) capture from flue gas is considered the hardest and most urgent work in the mercury emission control of coal-fired power plants [4].

Adsorption removal of elemental mercury by using solid sorbents is a viable method with respect to mercury emission control of coal-fired power plants [5]. Enormous carbon-based or noncarbon-based materials, e.g., biomass-derived chars [6], activated carbon [7], fly ash [8], as well as natural minerals [9], had been extensively studied for elemental mercury removal from flue gas. However, most of them exhibit limited mercury adsorption capacity and poorer Hg⁰ removal efficiency at low temperatures under flue gas free of HCl [10]. Carbon aerogels consisting of interconnected and uniform particles with well-defined pore architectures are promising sorbents for elemental mercury capture [11]. The large surface area and high porosity of carbon aerogels could offer abundant active sites for Hg⁰ adsorption and benefits a mass transfer process (e.g., Hg⁰ vapor diffusion) [12]. CuCl₂-modification is found to be effective for mercury adsorption, and the chlorine ions contained in CuCl₂ are favorable for Hg⁰ oxidation in chlorine-free flue gas [13]. In addition, CuCl₂ possesses the advantage of lower capital cost compared with the commonly used metal oxide additives. Yang et al. [14] developed a CuCl₂-modified magnetosphere for elemental mercury removal. It performed optimally at 150 °C with the highest Hg⁰ removal efficiency of 90.6%. Zhou et al. [15] synthesized a CuCl₂-impregnated CeO₂-ZrO₂ sorbent for Hg⁰ removal in chlorine-free coal combustion flue gas. They revealed that the interaction between Cl and chemical adsorbed oxygen contributed to its superior and stable mercury capture performance with a Hg⁰ removal efficiency of 89.6–97.1%. Therefore, CuCl₂ is employed as the additive for carbon aerogels for the sake of combining the prominent oxidation ability of CuCl₂ and the good mass transfer ability of carbon aerogels toward elemental mercury adsorption.

In this work, mesoporous carbon aerogels (MCA) with large pore volume (>1 cm³/g) and large specific surface area (>500 m²/g) were synthesized via the sol-gel method. The as-prepared MCA was subsequently modified with CuCl₂ by using the pore volume impregnation approach for Hg⁰ adsorption at an upflow bench-scale fixed-bed reactor in the temperature range of 40–160 °C. The samples were characterized by using field emission scanning electron microscopy (FESEM), X-ray diffraction (XRD), nitrogen adsorption/desorption, and X-ray photoelectron energy spectroscopy (XPS) techniques. The effects of CuCl₂ concentration and reaction temperature on Hg⁰ adsorption performances of CuCl₂-modified MCA were investigated. Furthermore, the reaction mechanism of Hg⁰ capture over CuCl₂-modified MCA is also addressed according to the XPS analysis results. Cu06MCA, Cu18MCA, and Cu30MCA denote the MCA modified by CuCl₂ with molar concentrations of 0.06, 0.18, and 0.30 mol/L, respectively.

2. Results and Discussions

2.1. Characterization Analysis

The FESEM images of pristine and CuCl₂-modified MCA are shown in Figure 1. The as-prepared MCA was formed by interconnected irregular spherical particles (Figure 1a), and many pores were produced among the particles which aggregated together. After modification with cupric chloride, the surface of the MCA became a little bit compact, and the MCA particles cohered together (Figure 1b), probably leading to the reduction of the Brunauer-Emmett-Teller (BET) surface area (

Table 1. Textural property of pristine and CuCl₂-modified MCA.

Samples	Surface Area (m2/g)	Total Pore Volume (cm3/g)	Micropore Volume (cm3/g)	Mesopore Volume (cm3/g)	Pore Diameter (nm)
MCA	523	1.21	0.15	1.06	9.2
Cu06MCA	641	1.38	0.06	1.32	8.6
Cu18MCA	474	1.12	0.12	1.00	9.5
Cu30MCA	413	1.09	0.11	0.98	10.6

).

The XRD patterns of pristine and CuCl₂-modified MCA are presented in Figure 2. The bigger broad diffraction peaks at ~23.9° were assigned to the reflection of the (002) lattice plane of activated carbon and graphite. The smaller broad diffraction bands at ~43.6° were attributed to the reflection of the (100) lattice plane of graphite [16,17]. The signals of CuCl₂ were not detected in CuCl₂-modified MCA, probably owing to the smaller content of CuCl₂ or the thorough dispersion of amorphous CuCl₂ on the MCA surface.

The nitrogen adsorption–desorption isotherms and pore size distributions of pristine as well as CuCl₂-modified MCA are displayed in Figure 3. The MCA exhibited type IV isotherm (Figure 3a). The position of the hysteresis loop at the higher relative pressure of above 0.8 implies the presence of large mesopores on the MCA surface [18], which is consistent with the broad pore size distribution centered between 10 and 30 nm. The BET surface area, total pore volume, and pore diameter of the MCA were 523 m²/g, 1.21 cm³/g, and 9.2 nm, respectively (Table 1). After modifying with CuCl₂, the mesoporous structures of the MCA were still preserved. However, the pore size distribution became narrow, with pore diameter centered at ~16 nm (Figure 3b).

The XPS spectra of fresh and spent Cu18MCA are displayed in Figure 4 to determine the surface species on sorbent surfaces. The C 1s spectra can be deconvoluted into five components in which chemical shifts are assigned to carboxyl groups (~288.6 eV), C=O bonds (~287.0 and ~287.2 eV), C-OH groups (~285.4, ~285.6, and 286.2 eV), and graphitic carbon (sp² configuration: ~284.2 and ~284.8 eV) [18]. In addition, the broad band at ~290.0 eV is related to shake-up satellite peaks assigned to π–π* transitions in aromatic rings [19]. The Cu 2p peak of fresh Cu18MCA at ~935.8 eV is assigned to the reflection of Cu²⁺ ions in CuCl₂ [20]. It shifts slightly toward higher values of 936.2 eV after Hg⁰ adsorption, possibly attributed to the formation of an oxychloride phase Cu₂OCl₂ [21], which is an important intermediate for mercury oxidation. As for spent Cu18MCA, the signal at ~105.4 eV is due to the reflection of Hg 4f 5/2 orbit of oxidized mercury, implying that Hg⁰ transformed into Hg²⁺ during the mercury adsorption process [22]. The peaks at ~200.2 and ~202.4 eV belong to covalent chlorine (C-Cl groups), while the bands at ~198.2 and ~199.8 eV are ascribed to ionic chlorine (Cl⁻ ions) [23]. The content of the covalent chlorine (C-Cl groups) decreased from 81.3% to 47.8% for fresh and spent Cu18MCA, respectively, indicating that some C-Cl groups transformed into Cl⁻ ions during the mercury adsorption process [24]. Since the mercury on the surface of spent Cu18MCA exists in an oxidation state (Hg²⁺), the adsorbed mercury on the sorbent surface may react with C-Cl groups to form HgCl₂, and therefore the reaction mechanism of elemental mercury reacting with covalent chlorine (C-Cl groups) can be summarized as follows [25]: (i) Cl⁻ ions combined with carbon atoms to form C-Cl groups; (ii) gaseous Hg⁰ was oxidized by Cu₂OCl₂ and captured on the sorbent surface in the form of oxidized mercury (Hg²⁺); and (iii) finally, the adsorbed mercury reacted with the C-Cl groups to generate HgCl₂.

2.2. Effect of CuCl₂ Concentration

The impact of CuCl₂ concentration on Hg⁰ adsorption performance at 120 °C is presented in Figure 5. Pristine MCA showed weaker Hg⁰ adsorption ability. Though the Hg⁰ removal efficiency could reach as high as ~88.0% within 2 min, it dropped rapidly with the time elapsed and maintained at only ~27.3% after mercury adsorption over 80 min. The probable reason is that Hg⁰ adsorption on MCA is a physisorption process, and the number of the active sites of MCA will be consumed during mercury adsorption process, leading to the rapid decrease in Hg⁰ removal efficiency. However, the Hg⁰ capture ability of MCA can be effectively improved by loading with CuCl₂. The CuCl₂-modified MCA performed well toward Hg⁰ adsorption, with Hg⁰ removal efficiency all above 96.0% due to the addition of new chemisorption sites (Cu₂OCl₂ and C-Cl group) [26]. The Hg⁰ removal efficiency of MCA significantly rose to ~97.3% and ~97.6% when CuCl₂ concentrations were 0.06 mol/L and 0.18 mol/L, respectively. However, it dropped slightly to ~96.3% when CuCl₂ concentration further increased from 0.18 to 0.30 mol/L. According to the nitrogen adsorption–desorption data, when CuCl₂ concentration was 0.30 mol/L, BET surface area and pore volume greatly decreased thereby weakening the mercury adsorption process because excessive CuCl₂ will deteriorate the pore structures of the MCA. Thus, the optimal CuCl₂ concentration is 0.18 mol/L.

2.3. Effect of Reaction Temperature

The impact of reaction temperature on Hg⁰ adsorption performance of Cu18MCA is presented in Figure 6. Cu18MCA performed well toward Hg⁰ adsorption in the 40–160 °C temperature range, with Hg⁰ removal efficiency all above 95.0%, indicating that Hg⁰ adsorption on CuCl₂-modifed MCA was mainly dominated by the chemisorption process. The Hg⁰ sorption rate at the initial reaction stage and the equilibrium of Hg⁰ removal efficiency first increased and then decreased with the reaction temperature rising from 40 to 160 °C. Cu18MCA performed the best at 80 °C with the fastest Hg⁰ sorption rate and the highest Hg⁰ removal efficiency of ~98.7%. Though higher temperatures can enhance chemisorption processes [27], the product mercury chloride will decompose at elevated temperatures, resulting in the decline of Hg⁰ removal efficiency. Thus, the optimal reaction temperature for Cu18MCA is 80 °C. In addition, Cu18MCA performance was fairly stable over 600 min on stream, with Hg⁰ removal efficiency staying above 98.0% all times (Figure 7).

3. Experimental

3.1. Catalyst Synthesis and Characterization

Mesoporous carbon aerogels (MCA) were synthesized via the aqueous polymerization of resorcinol and formaldehyde [28]. CuCl₂-modified MCA was fabricated via the incipient wetness impregnation method. First, certain amounts of CuCl₂·2H₂O were dissolved in 3.0 g of deionized water to obtain solutions with CuCl₂ molar concentrations of 0.06, 0.18, and 0.30 mol/L, respectively. Then, about 1.0 g of as-prepared MCA was added into the aforementioned solution and subsequently dried at 70 °C overnight. The resultant sorbents were denoted as Cu06MCA, Cu18MCA, and Cu30MCA, respectively. The FESEM images were detected on a Phillips XL-30 FEG/NEW instrument (Eindhoven, The Netherlands). The XRD patterns were measured on Bruker D8 Advance equipment (Karlsruhe, Germany). The nitrogen isotherms were examined at 77 K on a Beishide 3H-2000PS4 apparatus (Norcross, GA, USA). The pore size distributions were estimated with the Barrett-Joyner-Halenda (BJH) method using the desorption branch of N₂ isotherm, which is expressed by the Kelvin equation as follows:

$$\ln (p^{*}/p_0)=-\left(\frac{2\gamma \nu \cos \theta }{RT{r}_m}\right)$$

(1)

where p* is the condensing pressure, p₀ is the saturated vapor pressure, γ is the liquid surface tension, ν is the adsorbate molar volume, θ is the angle between the liquid and solid surface, and r_m is the mean radius of curvature of the liquid meniscus. XPS analyses were carried out on a RBD upgraded PHI-5000C ESCA system (Perkin Elmer, Waltham, MA, USA) with Mg Kα radiation (hν = 1253.6 eV) or Al Kα radiation (hν = 1486.6 eV).

3.2. Mercury Adsorption

The mercury adsorption testing system has been described in detail elsewhere [29,30]. About 50 mg of sorbents were loaded into a fixed-bed reactor and heated up to the desired temperatures. Then, the nitrogen stream containing Hg⁰ vapor produced by a mercury generator at 50 °C was consecutively charged into the reactor. Hg⁰ concentration was measured by an on-line mercury analyzer (Lumex, RA-915-M, St. Petersburg, Russia). The mercury conversion can be attained by

η = (1 − C_out/C_in) × 100%

(2)

where η denotes the mercury removal efficiency (%), and C_in and C_out denote the inlet and outlet mercury concentration (µg/m³), respectively.

4. Conclusions

CuCl₂-modified MCA exhibited excellent Hg⁰ adsorption performance in the temperature range of 40–160 °C. The Hg⁰ removal performance of Cu18MCA first ascended and then descended with rising reaction temperatures. It possessed the highest Hg⁰ removal efficiency of ~98.7% and the fastest Hg⁰ sorption rate at 80 °C. Chemisorption dominated the mercury adsorption process over CuCl₂-modified MCA. Covalent chlorine (C-Cl groups) played a crucial role in elemental mercury adsorption process. Hg⁰ was first oxidized by Cu₂OCl₂ and captured in the form of oxidized mercury (Hg²⁺) and then reacted with C-Cl groups to form HgCl₂.