Role of Air Bubbles Overlooked in the Adsorption of

Article pubs.acs.org/est

Role of Air Bubbles Overlooked in the Adsorption of Perfluorooctanesulfonate on Hydrophobic Carbonaceous Adsorbents Pingping Meng,† Shubo Deng,*,† Xinyu Lu,† Ziwen Du,† Bin Wang,† Jun Huang,† Yujue Wang,† Gang Yu,† and Baoshan Xing‡ †

School of Environment, Beijing Key Laboratory for Emerging Organic Contaminants Control, State Key Joint Laboratory of Environment Simulation and Pollution Control (SKLESPC), Tsinghua University, Beijing 100084, China ‡ Stockbridge School of Agriculture, University of Massachusetts, Amherst, Massachusetts 01003, United States S Supporting Information *

ABSTRACT: Hydrophobic interaction has been considered to be responsible for adsorption of perfluorooctanesulfonate (PFOS) on the surface of hydrophobic adsorbents, but the long C−F chain in PFOS is not only hydrophobic but also oleophobic. In this study, for the first time we propose that air bubbles on the surface of hydrophobic carbonaceous adsorbents play an important role in the adsorption of PFOS. The level of adsorption of PFOS on carbon nanotubes (CNTs), graphite (GI), graphene (GE), and powdered activated carbon (PAC) decreases after vacuum degassing. Vacuum degassing time and pressure significantly affect the removal of PFOS by these adsorbents. After vacuum degassing at 0.01 atm for 36 h, the extent of removal of PFOS by the pristine CNTs and GI decreases 79% and 74%, respectively, indicating the main contribution of air bubbles to PFOS adsorption. When the degassed solution is recontacted with air during the adsorption process, the removal of PFOS recovers to the value obtained without vacuum degassing, further verifying the key role of air bubbles in PFOS adsorption. By theoretical calculation, the distribution of PFOS in air bubbles on the adsorbent surfaces is discussed, and a new schematic sorption model of PFOS on carbonaceous adsorbents in the presence of air bubbles is proposed. The accumulation of PFOS at the interface of air bubbles on the adsorbents is primarily responsible for its adsorption, providing a new mechanistic insight into the transport, fate, and removal of PFOS.

■

INTRODUCTION Perfluorinated compounds (PFCs) have attracted a great deal of attention recently because of their global distribution in aquatic environments and potential toxicity to animals and human beings.1 Perfluorooctanesulfonate (PFOS) is a typical PFC and has been widely produced and applied in many industries.2−4 The occurrence of PFOS in aquatic environments is a long-term environmental problem. Adsorption is not only an effective technology for removing PFOS from water or wastewater but also an important process affecting the distribution of PFOS at solid−liquid interfaces and its fate in aquatic environments. Many adsorbents such as activated carbon and carbon nanotubes (CNTs) have been used to adsorb PFOS.5−10 Some interactions, including electrostatic interaction, the hydrophobic effect, and hydrogen bonding, have been reported to be possibly involved in the adsorption of PFOS on different adsorbents.6,9,10 Among these interactions, the hydrophobic effect is often considered to be responsible for adsorption of PFOS on activated carbon, carbon nanotubes, and nonionic resin, even the organic components in sediments.5,6,11 The hydrophobic C−F chain of PFOS is © XXXX American Chemical Society

thought to directly adsorb on the hydrophobic regions of adsorbents via hydrophobic interaction.6,11−14 PFOS is a perfluorinated surfactant with a nonpolar long C− F chain (hydrophobic and oleophobic) and a polar sulfonic group (hydrophilic).11,15 Because the long C−F chain possesses nonpolar and aprotic character, it turns out to be resistant to not only water but also most hydrocarbon solvents.16 Therefore, the long C−F chain of PFOS with both hydrophobicity and oleophobicitity may not directly adsorb on the hydrophobic adsorbents through hydrophobic interaction. PFOS prefers to exist at the air−water interfaces to significantly decrease the surface tension,17 making it effective in some applications. Acoustic cavitation driven by ultrasound is shown to be effective for PFOS degradation because PFOS molecules can concentrate at bubble−water interfaces and be pyrolyzed into inorganic products.18 Received: August 21, 2014 Revised: October 26, 2014 Accepted: November 3, 2014

A

dx.doi.org/10.1021/es504108u | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

Vacuum Degassing Experiments. A vacuum dryer was used for vacuum degassing (Figure S1 of the Supporting Information). A certain amount of carbon materials (Table S2 of the Supporting Information) was placed into 60 mL vessels with screw caps, and then 2 mL of ultrapure water was added to immerse the adsorbent. Thereafter, vessels with carbon materials, 100 mL of a 260 mg/L PFOS stock solution in a 250 mL beaker as well as 800 mL of ultrapure water in a 1 L beaker, were placed in a vacuum dryer and degassed by a vacuum pump. The pressure in the vacuum dryer was kept at 0.01 atm for 12 h to obtain the degassed adsorbent solution, PFOS stock solution, and ultrapure water, which were used to prepare the different solutions in the following sorption experiments. After degassing, the PFOS concentration in the degassed solution was measured to be 261 mg/L, very close to the initial value before vacuum degassing. This result indicates that PFOS hardly volatilized during the degassing process. Sorption Experiments. All adsorption experiments were conducted in a back and forth shaker with a speed of 180 rpm at 28 °C for 48 h. All sorption experiments were conducted twice, and the average values were adopted. The control sorption experiments were conducted under the same condition except for the vacuum degassing. In the investigation of the effect of vacuum degassing, only adsorbents on the PFOS sorption, 3 mL of a 260 mg/L PFOS solution, and 55 mL of ultrapure water were added to the 60 mL vessels with the degassed adsorbents (except PAC). Because PAC had an adsorption capacity for PFOS much higher than those of other adsorbents, a higher PFOS concentration was adopted to yield moderate removal percents in the adsorption experiments. For PAC, 20 mL of a 260 mg/L PFOS solution and 38 mL of ultrapure water were added. In the case of the degassing solution only, the adsorbents were directly added to 60 mL vessels, and then 3 mL of a 260 mg/L degassed PFOS solution and 57 mL of degassed ultrapure water were also added (for PAC, 20 mL of a 260 mg/L degassed PFOS solution and 40 mL of degassed ultrapure water). In the study of degassing, both adsorbent and solution, 3 mL of a 260 mg/L degassed PFOS solution and 55 mL of degassed ultrapure water (for PAC, 20 mL of a 260 mg/L degassed PFOS solution and 38 mL of degassed ultrapure water), were added to the 60 mL vessels with 2 mL of a degassed adsorbent solution. All solutions mentioned above were prepared within 3 min to avoid the redissolution of air bubbles in the system, and the vessels were sealed by screw caps and put into a shaker for adsorption experiments. Air contact experiments during PFOS adsorption were conducted in 100 mL flasks. Ten milligrams of CNTs-Pri was added to the flasks, followed by the addition of 10 mL of ultrapure water to immerse the adsorbent. These flasks were placed in the degassing system and kept at 0.01 atm for 24 h. Then, the flasks were taken out immediately, and 3 mL of a 260 mg/L degassed PFOS solution was added to the flasks, followed by the addition of the degassed ultrapure water to reach a volume of 60 mL. These flasks were sealed with Parafilm M laboratory film, and then several pores were punched using a needle. Finally, the flasks were shaken in a shaker at 180 rpm for 48 h. Dissolved Oxygen Measurement. The amount of dissolved oxygen (DO) in solution was determined by the iodometric method. The effect of degassing time on DO was determined with Milli-Q water degassed at 0.01 atm for different times. The effect of degassing pressure on DO was

In recent years, air nanobubbles have been found on hydrophobic surfaces via atomic force microscopy.19−21 Zhang et al. observed that the nanobubbles are affected by environmental conditions, and their height is normally less than 100 nm and their diameter less than 2 μm;22,23 some nanobubbles can exist stably for 4 days.24 Bram et al. reported that the nanobubbles on polyamide and hydrophobized silicon surfaces not only are stable under ambient conditions but also exhibit superstability under reduced pressure in cavitation experiments.25 Also, the presence of bubbles on the material surfaces affects the hydrophobic interaction between mica and polystyrene.26 Ehrenhauser et al. demonstrated that bubble bursting as an aerosol generation mechanism during an oil spill in the deep-sea environment was a pathway for the transport of alkanes into the atmosphere.27 Adsorption of some chemicals (especially surfactants) on air bubbles has been widely reported,28,29 but the effect of air bubbles on pollutant adsorption on adsorbents has hardly been studied to date. According to the discussion given above, hydrophobic adsorbents used for PFOS adsorption likely have air bubbles on their surfaces, affecting the adsorption of PFOS on the adsorbents. In this case, the mechanism for the adsorption of PFOS on hydrophobic carbonaceous adsorbents should be closely related to the air bubbles on the surfaces, which is completely different from the hydrophobic interaction reported. Therefore, we hypothesized that air bubbles are mainly responsible for the adsorption of PFOS by hydrophobic carbonaceous adsorbents, instead of the often believed hydrophobic interaction. The objectives of this study are (1) to test this hypothesis and examine the importance of air bubbles in adsorbing PFOS by different carbonaceous adsorbents via vacuum degassing, (2) to calculate the distribution of air bubbles on the adsorbent surface and the contribution of air bubbles to PFOS adsorption, and (3) to propose the mechanism of adsorption of PFOS on different adsorbents in the presence of air bubbles.

■

MATERIALS AND METHODS Materials. Pristine CNTs (CNTs-Pri, outer diameter of 10−20 nm, length of 10−20 μm, purity of >99.9%) and hydroxylated CNTs (CNTs-OH, outer diameter of 10−20 nm, length of 10−20 μm, purity of >99.9%, hydroxyl content of 1.53 wt %) were purchased from Chengdu Institute of Organic Chemistry, Chinese Academy of Sciences. Graphene (GE, purity of >99.8%) and carboxylated graphene (GE-COOH, purity of >99.8%, carboxyl group content of 5.0 wt %) were obtained from Nanjing Xianfeng Nanomaterials Co. Powdered activated carbon (PAC) (99.9%) were purchased from Chengde Pengcheng Activated Carbon Co. and Tianjin Dengke Chemical Reagent Co., respectively. Characterization of Carbon Adsorbents. The carbonaceous adsorbents were characterized in terms of specific surface area, elemental composition, and ζ potential. Their Brunauer− Emmett−Teller (BET) surface areas were measured by N2 adsorption at 77 K in a gas adsorption instrument (Autosorb iQ, Quantachrome Corp.). Dry-weight-based elemental contents were measured by an elemental analyzer (PE-2400II, PerkinElmer). The results of surface area and elemental composition are listed in Table S1 of the Supporting Information. The ζ potentials of the adsorbents were measured by a ζ potential instrument (Delsa Nano C, Beckman Coulter). B



Article

the unobvious decrease in the level of PFOS removal. For CNTs-OH, GE-COOH, and PAC, the degassing solution alone has an influence on PFOS removal greater than that of only the degassing adsorbent, while other materials show an opposite trend. The reason for this phenomenon could be that air bubbles are more likely to exist on a hydrophobic surface. The CNTs-OH contain 1.53 wt % OH group, and O, H, and N elements are present on its surface; no such elements exist on the CNTs-Pri (Table S1 of the Supporting Information), making CNTs-OH more hydrophilic than CNTs-Pri. The same reason is also applicable for GE-COOH and PAC. Because air bubbles are less likely to stay on their hydrophilic surface, the contribution of this part may be smaller, while the air in the solution becomes more significant for the hydrophobic adsorbents. For CNTs-Pri and GE, their surfaces are more hydrophobic because of their less functional groups, making the contribution of air bubbles greater, and degassing the material makes a larger difference. Although the GI contains a relatively high oxygen content, degassing this material leads to a pronounced decrease in the level of PFOS removal, probably because of the large dose (150 mg/60 mL) used, making more air bubbles on its surface. Effect of Degassing Time and Pressure on PFOS Removal. Degassing time and pressure have significant effects on the dissolved air in solution and air bubbles on adsorbent surfaces, which finally influence the adsorption of PFOS on carbonaceous adsorbents. The effect of degassing time on PFOS removal is illustrated in Figure 2a. With an increasing degassing time, the level of PFOS removal decreased rapidly in the initial 12 h and then approached equilibrium within 24 h. For CNTs-Pri and GI, after vacuum degassing at 0.01 atm for 36 h, the level of PFOS removal decreases 79% and 74%, respectively, indicating the main contribution of air bubbles to PFOS adsorption. The remaining PFOS removal may be attributed to the residual air bubbles in the system because air may redissolve into the solution during the preparation of the solution in the open atmosphere. For the four other materials, the degree of the decrease in the level of PFOS removal is smaller to different extents. The CNTs-OH have a structure similar to that of the CNTs-Pri, and CNTs-OH still achieve a high level of PFOS removal after degassing for 24 h. The difference in surface hydrophobicity between these two CNTs may account for their discrepancy after degassing. For PAC, the level of PFOS removal only has a small decrease of ∼10% after degassing for 48 h. Although PAC is well-known to be

determined with Milli-Q water degassed at 0.01, 0.3, and 0.6 atm for 24 h. The DO of Milli-Q without any treatment was also determined as a control. In recontacting air experiments, Milli-Q water was first degassed and then contacted with air for different times, and finally the samples were taken to measure DO. PFOS Concentration Measurement. After adsorption experiments, the supernatant was filtered with a 0.22 μm nylon membrane, and ∼2 mL of the sample was collected after 20 mL of filtrate to eliminate the effect of membrane sorption. PFOS concentrations in solution were determined by a LC-10ADvp high-performance liquid chromatography instrument with a CDD-6A conductivity detector from Shimadzu, and the detailed procedure was described previously.8,30

■

RESULTS AND DISCUSSION Effect of Vacuum Degassing on the Removal of PFOS by Different Adsorbents. Figure 1 shows the effect of

Figure 1. Effect of vacuum degassing at 0.01 atm for 12 h on the removal of PFOS by different carbon adsorbents.

vacuum degassing at 0.01 atm for 12 h on the removal of PFOS by CNTs-Pri, CNTs-OH, GI, GE, GE-COOH, and PAC. The removal of PFOS by all these adsorbents decreases after degassing. When both adsorbents and the solution are degassed, the decrease in the level of removal of PFOS by all adsorbents except PAC is more obvious than that in the case of only the adsorbent or solution being degassed. In the case of a degassed sorbent only, the number of air bubbles would decrease after degassing. When the degassed adsorbent was mixed with the solution without degassing, the dissolved gas would aggregate and form bubbles on the surface,31 resulting in

Figure 2. Effect of (a) degassing time and (b) degassing pressure in the vacuum degassing experiments on the removal of PFOS by different carbonaceous adsorbents in the subsequent adsorption experiments. C



Article

hydrophobic, the PAC used in this study contains some oxygen-containing functional groups (Table S1 of the Supporting Information), and these hydrophilic sites make it hard for bubbles to stay on the surface. More importantly, the porous structure of PAC could also be an important hindrance for vacuum degassing to remove the bubbles inside the pores. For GE and GE-COOH, the level of PFOS removal decreases ∼50% and ∼40%, respectively, after degassing. According to their elemental analysis (Table S1 of the Supporting Information), both of them have plenty of H and O, especially O on the GE-COOH, suggesting some functional groups on the adsorbent surfaces. These oxygen-containing groups might decrease the influence of bubbles and increase the number of other possible interactions between PFOS and the material, especially for GE-COOH. Figure 2b presents the effect of degassing pressure on PFOS removal on the six adsorbents. With the decrease in degassing pressure, the level of PFOS removal continues to decrease. All adsorbents are sensitive to degassing pressure for PFOS removal, and an obvious decrease in the level of PFOS removal is observed at 0.01 atm. According to Henry’s law, gas solubility in a solvent is related to the gas pressure. The lower the pressure, the less the gas dissolved into solution. Additionally, air bubbles at solid−liquid interfaces are likely to exist on the nanoscale and turn out to be abnormally stable.19,23,24 It has been proven that these nanobubbles remain stable under an intermittent negative pressure of −6 MPa when the material is put into ultrasound.25,32,33 Several researchers have found that only high-vacuum degassing (∼0.1 atm) has a satisfactory result for degassing the nanobubbles.25,34 Atomic force microscopy pictures have indicated that there is a significant decrease in the number of nanobubbles at water−graphite interfaces after degassing at 0.01 atm; 0.4 atm seems to be less effective, and many bubbles remain stable after degassing.34 Evidently, the removal of air bubbles from the solution and at the interface is all related to degassing pressure, and the lower the pressure, the better the degassing effect. Because air bubbles make an important contribution to PFOS adsorption, the level of PFOS removal also decreases with a decreasing number of air bubbles. For comparison, the level of removal of phenanthrene by the CNTs-Pri decreases only insignificantly (61.0% at 1 atm and 58.5% at 0.01 atm) with vacuum degassing (Figure S2 of the Supporting Information, and experimental conditions are described in the Supporting Information). Phenanthrene may adsorb on the hydrophobic CNT surfaces without air bubbles via hydrophobic interaction. The insignificant decrease in the level of phenanthrene removal after degassing is negligible and within experimental uncertainty. Evidently, the adsorption of PFOS on CNTs-Pri is clearly different from that of polycyclic aromatic hydrocarbons (PAHs) such as phenanthrene in terms of adsorption mechanism, and air bubbles substantially contribute to PFOS adsorption, but not for PAHs. Effect of Degassing Time and Pressure on Dissolved Oxygen in Water. Because it is difficult to quantify the effect of degassing on the number of air bubbles on the surface of the adsorbents, dissolved oxygen in solution before and after degassing is detected and used to represent the air in the system. The effect of degassing time and pressure on dissolved oxygen is shown in Figure 3. With the increase in degassing time and the decrease in pressure, the level of dissolved oxygen decreases, indicating less air in solution. The changing trend of PFOS removal is consistent with that of the dissolved oxygen in solution. With the increase in degassing time, the levels of both

Figure 3. Effect of (a) degassing time and (b) pressure on the level of dissolved oxygen in water.

dissolved oxygen and PFOS removal decrease quickly in the first 12 h, and then the curve becomes flat and almost becomes constant after 24 h (Figures 2a and 3a). With the decrease in degassing pressure, the level of dissolved oxygen decreases linearly, similar to the decrease in the level of PFOS removal (Figures 2b and 3b). The level of dissolved oxygen in solution is ∼5.5 mg/L after degassing at 0.6 atm for 24 h; a similar level of dissolved oxygen can be achieved at 0.01 atm for ∼1 h on the dissolved oxygen−time curve (Figure 3a). It can be found that PFOS removal after degassing at 0.6 atm for 24 h and at 0.01 atm for 1 h is also very similar. It can be concluded that degassing influences the amount of air in the system, which can be partly reflected by the level of dissolved oxygen, and the decrease in the amount of air in the system decreases the number of air bubbles on the adsorbent surfaces, resulting in the decrease in the level of PFOS removal. Relationship between the Levels of Dissolved Oxygen in Solution and PFOS Removal. To further investigate the effect of air bubbles on PFOS adsorption, the relationship between the levels of dissolved oxygen in solution and PFOS removal at different pressures by six carbonaceous adsorbents is correlated and shown in Figure S3 of the Supporting Information. These experimental data exhibit good linearity (R2 > 0.86), and highly linear relationships for the CNTs-Pri, CNTs-OH, GE, and GE-COOH are obtained (R2 > 0.97). The relatively low linearity between the level of PFOS removal on PAC versus the level of dissolved oxygen in solution may be attributed to the difficult removal of air bubbles from the porous material. This result indicates that the level of PFOS removal is linearly related to the level of dissolved oxygen in solution. If there is no air in the solution (x is equal to 0 in the equations), the level of PFOS removal is 2.2% for CNTs-Pri, 13.0% for CNTs-OH, 5.0% for GI, 7.1% for GE, 19.7% for GECOOH, and 44.8% for PAC. Thus, air bubbles play an important role in PFOS adsorption, and the CNTs-Pri have minimal adsorption for PFOS after degassing, indicating that air bubbles almost completely dominate the adsorption of PFOS. Other adsorbents still have some PFOS adsorption after degassing, especially for the CNTs-OH, GE-COOH, and PAC, indicating that other interactions are involved in the adsorption in addition to air bubbles. Electrostatic interaction has been proven to play an important role in the adsorption of PFOS on many adsorbents.5−7,35 To check the involvement of electrostatic interaction for PFOS adsorption, the ζ potentials of six adsorbents at different pH values are measured (Figure S4 of the Supporting Information), and solution pH values before and after adsorption are also measured (Table S3 of the Supporting Information). It is evident that six adsorbents are all negatively charged during the adsorption process. Because the pKa of PFOS is −3.27,5 it exists as an anion at the pH studied. D



Article

and this result does prove that it is air bubbles that participate in PFOS adsorption. Because the shaking speed affects the dissolution of air into solution, it may also influence the adsorption of PFOS on the adsorbents. Figure S5 of the Supporting Information presents the effect of shaking speed on the adsorption of PFOS on the CNTs-Pri in the air contact flask (both solution and adsorbent were degassed before adsorption experiments). The level of PFOS removal increases with the increase in shaking speed, and a sharp increase in the level of PFOS removal is observed when the shaking speed increases from 120 to 150 rpm. The level of PFOS removal increases from 21.3% (RSD = 3.0%) to 25.1% (RSD = 2.4%) when the shaking speed increases from 40 to 240 rpm. The intensified mixing may accelerate the dissolution of air in solution and create more air bubbles on the adsorbent surfaces, thus resulting in a higher level of PFOS removal. The adsorption kinetics of PFOS on CNTs-Pri in the air contact flask was also studied, and the adsorption equilibrium of PFOS is almost achieved after 3 h when the degassed solution is recontacted with air in the adsorption process (Figure S6 of the Supporting Information). The level of dissolved oxygen in the degassed solution after air contact is also measured, and it takes only 1 h to recover almost 90% of the dissolved oxygen, the level then gradually returning to the value obtained without degassing (Figure S7 of the Supporting Information). Because it needs time for air in the solution to spread to the adsorbent surfaces and form bubbles for PFOS adsorption, the level of adsorption of PFOS on the adsorbents lags behind the level of dissolved oxygen in solution. In comparison with the adsorption kinetics in the degassing solution and no degassing solution, the PFOS removal rate in the degassed solution after air recontact is higher than that in the degassed solution and lower than that in the normal solution without degassing at the beginning of the adsorption process. After 3 h, the level of PFOS removal in the degassed solution after air contact reaches the highest and stable values, indicating that air bubbles on the adsorbent surfaces reach the equilibrium. Theoretical Calculation of the Distribution of PFOS in Air Bubbles on the Adsorbent. To better clarify the distribution of PFOS on air bubbles and the coverage of air bubbles on the adsorbent before and after degassing, the following theoretical calculation is conducted. For PFOS, an amendatory Gibbs adsorption equation is used to describe the relationship among the surface tension (γ), the surfactant concentration (C), and the amount of surfactant adsorbed at the interface (Γ), as shown below.37

Therefore, electrostatic repulsion occurs between anionic PFOS and negatively charged adsorbents, preventing the adsorption of PFOS on the six adsorbents. However, because the CNTsOH contain some nitrogen atoms, these nitrogen-containing groups may be protonated at pH ∼6 (Table S3 of the Supporting Information) and adsorb some anionic PFOS via electrostatic attraction on some sites. Speltini et al. reported that an anion-exchange mechanism between the anionic PFC head and the protonated amino group on MWCNT-R-NH2 was involved in the solid-phase extraction of PFOS and PFOA,10 the same as the electrostatic attraction in our study. It should be noted that there are some H and O elements on the adsorbents except CNTs-Pri (Table S1 of the Supporting Information), and some oxygen-containing groups such as hydroxyl groups may be present on adsorbent surfaces. Thus, hydrogen bonding may be involved in the adsorption of PFOS via the formation of this bond between the sulfonic group of PFOS and the hydroxyl group on the adsorbents. Besides these specific interactions, aspecific interactions such as van der Waals forces between PFOS and adsorbents tested may be also involved in the adsorption,12,36 but their contribution should be unimportant because of the presence of air bubbles on the adsorbent surface. Effect of Air Contact on PFOS Removal. To further confirm the role of air bubbles in the adsorption of PFOS, the degassed solution is contacted with air during the adsorption of PFOS on the CNTs-Pri. As shown in Figure 4, when PFOS

Figure 4. Effect of air contact on PFOS removal in the degassed system by the CNTs-Pri. (A) PFOS removal in a 60 mL vessel containing the degassed solution, with almost no space left inside. (B) PFOS removal in a sealed 100 mL flask containing 60 mL of the degassed solution. (C) PFOS removal in a sealed 100 mL flask containing 60 mL of the normal solution without degassing. (D) PFOS removal in an unsealed 100 mL flask containing 60 mL of the degassed solution.

Γ=−

adsorption is conducted in a sealed flask that has some air above the solution but is not open to the atmosphere (column B in Figure 4), the level of PFOS removal increases to 23.3%, much higher than the value of 6.0% in the degassed system (column A in Figure 4) and close to the value of 24.3% in the same sealed flask containing a normal solution without degassing (column C in Figure 4), suggesting that the residual air in the flask is enough to recover the air bubbles on the adsorbent surface for PFOS adsorption. When PFOS adsorption is conducted in the same unsealed flask containing a degassed solution (open to the atmosphere), the level of PFOS removal increases to 24.7% (column D in Figure 4), a bit higher than that obtained in the sealed system. Recontact with air in the adsorption process can bring air back into the system,

C ⎛ dγ ⎞ ⎜ ⎟ 2RT ⎝ dC ⎠

(1)

where T is the temperature and R is an empirical constant. Because dγ/dC is unknown, we need the derivation to do the final calculation. Some studies have discovered that the γ of a dilute solution conforms to the Szyszkowski empirical formula.38 γ0 − γ C = b ln +1 0 a γ

(2)

where γ and γ are the surface tensions at surfactant concentrations of 0 and C, respectively, and a and b are empirical constants. When the surfactant concentration is quite low, eq 2 can be changed into eq 3 (K is a constant). By 0

E



Article

air bubbles, the schematic adsorption diagram of PFOS on six carbonaceous adsorbents is proposed (Figure 5). Air bubbles

deriving eqs 2 and 3, we can get an approximate value of dγ/dC = Δγ/ΔC and finally determine the amount of surfactant adsorbed on the surface (Γ) in eq 1.

γ = γ 0 − KC

(3)

Taking the CNTs-Pri as an example, we can determine the distribution of PFOS at the interface of air and water. The initial concentration of PFOS is ∼0.024 mmol/L, and it comes to 0.023 and 0.018 mmol/L after adsorption in the degassing and no degassing experiments, respectively. The concentration of PFOS in our experiments is low enough to meet the criterion (nearly 0.04 mmol/L) for a linear relation.39 The surface tensions of the PFOS solution at 0.018 and 0.023 mmol/L at 28 °C are 70.6 and 70.2 mN/m, respectively.39 The surface tension of pure water is 72.0 mN/m.39 When these data are put into eqs 3 and 1, Γdegassing and Γno degassing are calculated to be 3.62 × 10−7 and 2.81 × 10−7 mol/m2, respectively. When the aggregation of PFOS at the interface reaches saturation, the Γ is reported to be 3.7 × 10−6 mol/m2,37 which is 9.2 and 12.2 times higher than the calculated Γdegassing and Γno degassing values, respectively. It is clear that PFOS was distributed sparsely on the surface of air bubbles on the CNTs-Pri under the experimental conditions. On the basis of our finding, only air bubbles are responsible for the adsorption of PFOS on the CNTs-Pri. The adsorbed amounts of PFOS on the CNTs-Pri before and after degassing are 19.4 and 3.3 mg/g, respectively. According to the obtained Γdegassing and Γno degassing values, the total surface area of air bubbles that are needed for PFOS adsorption is calculated to be 17 m2 on 1 g of CNTs-Pri after degassing and 128 m2 on 1 g of CNTs-Pri without degassing. The BET surface area of the CNTs-Pri used in this experiment is 131 m2/g, and thus, the air bubbles on the adsorbent with and without degassing account for 6.5 and 49.2% of the total surface area of CNTs-Pri, respectively, if the air bubbles are adsorbed on CNTs-Pri in the form of a hemisphere. Solubility of PFOS in Different Hydrocarbon Solvents. To illustrate that the C−F chain of PFOS cannot directly adsorb on the hydrophobic adsorbents via hydrophobic interaction, the solubility of PFOS in different organic solvents is measured. Figure S8 of the Supporting Information shows the solubility of PFOS in methanol, acetone, tetrahydrofuran, tetrachloromethane, toluene, and hexane at 25 °C and the dielectric constants of these solvents. With the decrease in the dielectric constants, the polarity of organic solvents decreases, and PFOS solubility also decreases, indicating that it is difficult for PFOS to dissolve in nonpolar solvents. For conventional hydrocarbon compounds, the solubility increases with the decreasing polarity of organic solvents.40 This contrary finding indicates that PFOS is different from hydrocarbon compounds. PFOS hardly dissolves in toluene and hexane (nonpolar hydrocarbon solvents), indicating that it is extremely difficult for the C−F chain of PFOS to directly adsorb on the hydrophobic adsorbent surfaces via hydrophobic interaction, and the oleophobic property of the C−F chain prevents it from adsorbing on the hydrophobic surface. Mechanism of Adsorption of PFOS on Typical Carbonaceous Adsorbents. The existing theories stress hydrophobic interaction between the C−F chain of PFOS and hydrophobic adsorbents more,5,6,11 but the long C−F chain in PFOS is not only hydrophobic but also oleophobic, giving it a tendency to repel hydrophobic hydrocarbon materials. According to our experimental results based on the effect of

Figure 5. Schematic diagram for the adsorption of PFOS on the carbonaceous adsorbents in the presence of air bubbles: (a) CNTs-Pri and CNTs-OH, (b) PAC, and (c) GI, GE, and GE-COOH. A stands for the contribution of air bubbles. B represents the hydrogen bond. C denotes the electrostatic attraction on some surface sites.

are surely present on the hydrophobic parts on the carbonaceous adsorbents, and PFOS preference at the interface of air bubbles on the adsorbents is responsible for its adsorption. The hydrophobic and oleophobic C−F chains of PFOS can stretch into the bubbles, while the sulfonic head of PFOS exists in the aqueous solution near the interface. For the CNTs-Pri, only air bubbles are responsible for PFOS adsorption (Figure 5a), which is verified by a significant decrease in the level of PFOS removal after vacuum degassing. Because there are some oxygen- and nitrogen-containing groups on the CNTs-OH, electrostatic attraction and hydrogen bonding may be involved in the adsorption besides the adsorptive role of air bubbles (Figure 5a), and thus, vacuum degassing only decreases the level of removal of PFOS to some extent. The level of removal of PFOS by the CNTs-OH decreased with an increasing solution pH (Figure S9 of the Supporting Information), indicating the involvement of electrostatic interaction. The F



■

PAC contains oxygen-containing groups (Table S1 of the Supporting Information), such as hydroxyl and phenolic hydroxyl groups, and they may form hydrogen bonds with the sulfonic group in PFOS. PFOS are believed to adsorb on air bubbles on the hydrophobic sites of PAC (Figure 5b), and the insignificant decrease in the level of PFOS removal after degassing may be attributed to the difficult removal of air bubbles from the micropores of PAC and specific adsorption of PFOS on the surface functional groups. The GI has a low content of oxygen and hydrogen (Table S1 of the Supporting Information); thus, the contribution of possible hydrogen bonding is limited, and air bubbles play a dominant role in PFOS removal (Figure 5c), which is verified by the significant decrease in the level of PFOS removal after degassing (Figure 2a). Because the GE and especially GE-COOH contain high levels of oxygen (Table S1 of the Supporting Information), hydrogen bonds may easily form as shown in Figure 5c, and the contribution to PFOS adsorption is obvious. However, the contribution of air bubbles to the adsorption of PFOS on the GE and GE-COOH is still above 40% (Figure 2a). Therefore, the important role of air bubbles in the adsorption PFOS on carbonaceous adsorbents warrants research attention, and their contribution is dependent on the adsorbent surface polarity. In addition, the effect of vacuum degassing on PFOS solubility and micelle or semimicelle formation should be negligible because of the low concentration of PFOS in our experiments. Although the adsorption of PFOS on air bubbles looks like micelles of hemimicelles of PFOS (Figure 5), PFOS molecules are separated and C−F chains do not aggregate together. To the best of our knowledge, this is the first proposal that air bubbles on the surface of hydrophobic carbonaceous adsorbents are mainly involved in PFOS adsorption. The contribution of air bubbles to PFOS adsorption is dependent on the surface polarity of carbonaceous adsorbents, and air bubbles on the adsorbent surfaces are completely responsible for the adsorption of PFOS on the hydrophobic CNTs-Pri (without functional groups). The PFOS surfactant prefers to exist at the air−water interface, and the long C−F chain of PFOS can stretch into air bubbles with the polar head in aqueous solution. This new finding clarifies the fuzzy understanding of adsorption of PFOS on hydrophobic adsorbents, and also the adsorption difference of PFOS from hydrocarbons. Because air bubbles are ubiquitous, the involvement of air bubbles may be applicable for the adsorption of PFCs on different environmental media, such as sediment, soil, activated sludge, and natural organic matters in soil and water, as well as biological materials for the investigation of the interfacial process, transport, and fate of PFCs in aquatic environments.

■

Article

AUTHOR INFORMATION

Corresponding Author

*Telephone: +86-10-62792165. Fax: +86-10-62794006. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank the National Natural Science Foundation of China (Project 21177070), the National High-Tech Research and Development Program of China (Project 2013AA06A3), and the Collaborative Innovation Center for Regional Environmental Quality for financial support. Additionally, the analytical work was supported by the Laboratory Fund of Tsinghua University.

■

REFERENCES

(1) Route, W. T.; Key, R. L.; Russell, R. E.; Lindstrom, A. B.; Strynar, M. J. Spatial and temporal patterns in concentrations of perfluorinated compounds in bald eagle nestlings in the upper midwestern United States. Environ. Sci. Technol. 2014, 48 (12), 6653−6660. (2) Paul, A. G.; Jones, K. C.; Sweetman, A. J. A first global production, emission, and environmental inventory for perfluorooctane sulfonate. Environ. Sci. Technol. 2009, 43 (2), 386−392. (3) Wang, T.; Wang, Y.; Liao, C.; Cai, Y.; Jiang, G. Perspectives on the inclusion of perfluorooctane sulfonate into the Stockholm Convention on Persistent Organic Pollutants. Environ. Sci. Technol. 2009, 43 (14), 5171−5175. (4) Rattanaoudom, R.; Visvanathan, C.; Boontanon, S. K. Removal of concentrated PFOS and PFOA in synthetic industrial wastewater by powder activated carbon and hydrotalcite. J. Water Sustainability 2012, 2 (4), 245−258. (5) Yu, Q.; Zhang, R.; Deng, S.; Huang, J.; Yu, G. Sorption of perfluorooctane sulfonate and perfluorooctanoate on activated carbons and resin: Kinetic and isotherm study. Water Res. 2009, 43 (4), 1150− 1158. (6) Du, Z.; Deng, S.; Bei, Y.; Huang, Q.; Wang, B.; Huang, J.; Yu, G. Adsorption behavior and mechanism of perfluorinated compounds on various adsorbents: A review. J. Hazard. Mater. 2014, 274, 443−454. (7) Zhang, Q.; Deng, S.; Yu, G.; Huang, J. Removal of perfluorooctane sulfonate from aqueous solution by crosslinked chitosan beads: Sorption kinetics and uptake mechanism. Bioresour. Technol. 2011, 102 (3), 2265−2271. (8) Deng, S.; Yu, Q.; Huang, J.; Yu, G. Removal of perfluorooctane sulfonate from wastewater by anion exchange resins: Effects of resin properties and solution chemistry. Water Res. 2010, 44 (18), 5188− 5195. (9) Li, X. N.; Zhao, H. M.; Quan, X.; Chen, S.; Zhang, Y. B.; Yu, H. T. Adsorption of ionizable organic contaminants on multi-walled carbon nanotubes with different oxygen contents. J. Hazard. Mater. 2011, 186, 407−415. (10) Speltini, A.; Maiocchi, M.; Cucca, L.; Merli, D.; Profumo, A. Solid-phase extraction of PFOA and PFOS from surface waters on functionalized multiwalled carbon nanotubes followed by UPLC-ESIMS. Anal. Bioanal. Chem. 2014, 406, 3657−3665. (11) Higgins, C. P.; Luthy, R. G. Sorption of perfluorinated surfactants on sediments. Environ. Sci. Technol. 2006, 40 (23), 7251−7256. (12) Deng, S.; Zhang, Q.; Nie, Y.; Wei, H.; Wang, B.; Huang, J.; Yu, G.; Xing, B. Sorption mechanisms of perfluorinated compounds on carbon nanotubes. Environ. Pollut. 2012, 168, 138−144. (13) Senevirathna, S. T. M. L.; Tanaka, S.; Fujii, S.; Kunacheva, C.; Harada, H.; Ariyadasa, B. H. A. K.; Shivakoti, B. R. Adsorption of perfluorooctane sulfonate (n-PFOS) onto non ion-exchange polymers and granular activated carbon: Batch and column test. Desalination 2010, 260 (1−3), 29−33.

ASSOCIATED CONTENT

S Supporting Information *

Setup for vacuum degassing, properties of carbonaceous adsorbents, adsorbent doses and PFOS concentrations, phenanthrene adsorption, relationship between dissolved oxygen and PFOS removal ζ potentials of adsorbents, changes in solution pH, effect of shaking speed on PFOS removal by CNTs-Pri, PFOS adsorption kinetics, PFOS solubility in different organic solvents, and levels of dissolved oxygen in the degassed water after air contact. This material is available free of charge via the Internet at http://pubs.acs.org. G



Article

(14) Chen, H.; Zhang, C.; Yu, Y.; Han, J. Sorption of perfluorooctane sulfonate (PFOS) on marine sediments. Mar. Pollut. Bull. 2012, 64 (5), 902−906. (15) Davis, F.; Frary, E.; Stirling, C. J. M. Changing surface hydroand oleophobicity with resorcinarene multilayers: A simple water/oil proofing process. Langmuir 2004, 20 (21), 9075−9079. (16) Vincent, J. M. Noncovalent associations in fluorous fluids. J. Fluorine Chem. 2008, 129 (10), 903−909. (17) Vecitis, C. D.; Park, H.; Cheng, J.; Mader, B. T.; Hoffmann, M. R. Enhancement of perfluorooctanoate and perfluorooctanesulfonate activity at acoustic cavitation bubble interfaces. J. Phys. Chem. C 2008, 112 (43), 16850−16857. (18) Vecitis, C. D.; Wang, Y.; Cheng, J.; Park, H.; Mader, B. T.; Hoffmann, M. R. Sonochemical degradation of perfluorooctanesulfonate in aqueous film-forming foams. Environ. Sci. Technol. 2010, 44 (1), 432−438. (19) Attard, P.; Moody, M. P.; Tyrrell, J. Nanobubbles: The big picture. Physica a 2002, 314, 696−705. (20) Hampton, M. A.; Donose, B. C.; Nguyen, A. V. Effect of alcohol-water exchange and surface scanning on nanobubbles and the attraction between hydrophobic surfaces. J. Colloid Interface Sci. 2008, 325 (1), 267−274. (21) Dickinson, E.; Ettelaie, R.; Kostakis, T.; Murray, B. S. Factors controlling the formation and stability of air bubbles stabilized by partially hydrophobic silica nanoparticles. Langmuir 2004, 20 (20), 8517−8525. (22) Zhang, X. H.; Li, G.; Wu, Z. H.; Zhang, X. D.; Hu, J. Effect of temperature on the morphology of nanobubbles at mica/water interface. Chin. Phys. B 2005, 14 (9), 1774−1778. (23) Zhang, L.; Zhang, X.; Zhang, Y.; Hu, J.; Fang, H. The length scales for stable gas nanobubbles at liquid/solid surfaces. Soft Matter 2010, 6 (18), 4515−4519. (24) Zhang, X. H.; Quinn, A.; Ducker, W. A. Nanobubbles at the interface between water and a hydrophobic solid. Langmuir 2008, 24 (9), 4756−4764. (25) Borkent, B. M.; Dammer, S. M.; Schoenherr, H.; Vancso, G. J.; Lohse, D. Superstability of surface nanobubbles. Phys. Rev. Lett. 2007, 98 (20), 204502. (26) Saavedra, J. H.; Acuna, S. M.; Toledo, P. G. AFM forces between mica and polystyrene surfaces in aqueous electrolyte solutions with and without gas bubbles. J. Colloid Interface Sci. 2013, 410, 188− 194. (27) Ehrenhauser, F. S.; Avij, P.; Shu, X.; Dugas, V.; Woodson, I.; Liyana-Arachchi, T.; Zhang, Z.; Hung, F. R.; Valsaraj, K. T. Bubble bursting as an aerosol generation mechanism during an oil spill in the deep-sea environment: Laboratory experimental demonstration of the transport pathway. Environ. Sci.: Processes Impacts 2014, 16 (1), 65−73. (28) Vitasari, D.; Grassia, P.; Martin, P. Simulation of dynamics of adsorption of mixed protein-surfactant on a bubble surface. Colloids Surf., A 2013, 438 (SI), 63−76. (29) Maruyama, H.; Seki, H.; Matsukawa, Y. Adsorption behavior of bisphenol-A and diethyl phthalate onto bubble surface in nonfoaming adsorptive bubble. Chem. Eng. J. 2008, 141 (1−3), 112−118. (30) Deng, S.; Nie, Y.; Du, Z.; Huang, Q.; Meng, P.; Wang, B.; Huang, J.; Yu, G. Enhanced adsorption of perfluorooctane sulfonate and perfluorooctanoate by bamboo-derived granular activated carbon. J. Hazard. Mater. 2014, DOI: 10.1016/j.jhazmat.2014.03.045. (31) Zhang, H. X.; Zhang, X. D.; Sun, J. L.; Zhang, Z. X.; Li, G.; Fang, H. P.; Xiao, X. D.; Zeng, X. C.; Hu, J. Detection of novel gaseous states at the highly oriented pyrolytic graphite-water interface. Langmuir 2007, 23 (4), 1778−1783. (32) Brotchie, A.; Zhang, X. H. Response of interfacial nanobubbles to ultrasound irradiation. Soft Matter 2011, 7 (1), 265−269. (33) Zhang, X.; Lhuissier, H.; Enriquez, O. R.; Sun, C.; Lohse, D. Deactivation of microbubble nucleation sites by alcohol-water exchange. Langmuir 2013, 29 (32), 9979−9984. (34) Zhang, X. H.; Li, G.; Maeda, N.; Hu, J. Removal of induced nanobubbles from water/graphite interfaces by partial degassing. Langmuir 2006, 22 (22), 9238−9243.

(35) Deng, S.; Niu, L.; Bei, Y.; Wang, B.; Huang, J.; Yu, G. Adsorption of perfluorinated compounds on aminated rice husk prepared by atom transfer radical polymerization. Chemosphere 2013, 91 (2), 124−130. (36) He, G. Z.; Pan, G.; Zhang, M. Y. Assembling structures and dynamics properties of perfluorooctane sulfonate (PFOS) at watertitanium oxide interfaces. J. Colloid Interface Sci. 2013, 405, 189−194. (37) Rosen, M. J. Surfactants and interfacial phenomena; Wiley & Sons, Inc.: Hoboken, NJ, 2004. (38) Ross, S.; Morrison, I. D. On the alleged ideality of szyszkowskilangmuir adsorption. J. Colloid Interface Sci. 1983, 91 (1), 244−247. (39) Vecitis, C. D.; Park, H.; Cheng, J.; Mader, B. T.; Hoffmann, M. R. Enhancement of perfluorooctanoate and perfluorooctanesulfonate activity at acoustic cavitation bubble interfaces. J. Phys. Chem. C 2008, 112 (43), 16850−16857. (40) Craft, N. E.; Soarse, J. H. Relative solubility, stability, and absorptivity of lutein and β-carotene in organic-solvents. J. Agric. Food Chem. 1992, 40 (3), 431−434.

H


Role of Air Bubbles Overlooked in the Adsorption of

Recommend Documents