The relationship between humic acid (HA) adsorption on and stabilizing multiwalled carbon nanotubes (MWNTs) in water: Effects of HA, MWNT and solution properties

Journal of Hazardous Materials 241–242 (2012) 404–410

Contents lists available at SciVerse ScienceDirect

Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat

The relationship between humic acid (HA) adsorption on and stabilizing multiwalled carbon nanotubes (MWNTs) in water: Effects of HA, MWNT and solution properties Daohui Lin a,b,∗ , Tingting Li a , Kun Yang a,b , Fengchang Wu c,∗∗ a

Department of Environmental Science, Zhejiang University, Hangzhou 310058, China Zhejiang Provincial Key Laboratory of Organic Pollution Process and Control, Zhejiang University, Hangzhou 310058, China c State Key Laboratory of Environmental Criteria and Risk Assessment, Chinese Research Academy of Environmental Sciences, Beijing 100012, China b

h i g h l i g h t s    

MWNT-surface area-normalized HA adsorption increased with increasing MWNT diameter. HA with higher polarity was adsorbed less by but stabilized more amount of MWNTs. MWNT stabilization increased and then leveled off with increasing HA adsorption. Ionic strength and pH affect relationship between HA adsorption and MWNT suspension.

a r t i c l e

i n f o

Article history: Received 10 April 2012 Received in revised form 2 September 2012 Accepted 26 September 2012 Available online 4 October 2012 Keywords: Humic acid (HA) Carbon nanotube (CNT) Suspension Adsorption

a b s t r a c t This study was aimed to explore the relationship between humic acid (HA, as a model NOM) adsorption on and stabilizing multiwalled carbon nanotubes (MWNTs) in water with a focus on the effects of HA, MWNT and solution properties. It was found that MWNT-surface area-normalized adsorption of HAs (QSA ) increased with increasing outer-diameter of the MWNTs and decreasing polarity of the HAs. However, at low pH values (ca. <4) or high ionic strengths (ca. >1 mmol L−1 Ca2+ ), the HA adsorption decreased with decreasing polarity of the HAs. The MWNT stabilization increased with increasing QSA , but the increase leveled off when QSA exceeded a threshold value markedly lower than the maximum QSA , especially for the MWNTs with relative large outer-diameters. On the whole, the QSA -normalized MWNT stabilization, presenting the capability of the MWNT-adsorbed HAs for the MWNT stabilization, increased with increasing HA polarity and solution pH, but with decreasing Ca2+ concentration. However, the stabilized MWNTs by the HAs with greater polarity could be more subject to destabilization by Ca2+ . The results of this study are believed to shed light on predictive understanding the interaction between MWNTs and NOM and the environmental behavior of MWNTs. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Owing to their excellent properties, carbon nanotubes (CNTs) are increasingly used in a variety of areas such as electronic, biomedical, pharmaceutical, cosmetic, catalytic and environmental applications [1,2]. Therefore, CNTs have the potential to be increasingly released into the aquatic environment during their life-cycles [3]. Once released into the aqueous environment, CNTs

∗ Corresponding author at: Department of Environmental Science, Zhejiang University, Hangzhou 310058, China. Tel.: +86 571 88982582; fax: +86 571 88982590. ∗∗ Corresponding author. E-mail addresses: [email protected] (D. Lin), [email protected] (F. Wu). 0304-3894/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jhazmat.2012.09.060

are prone to agglomeration and deposition due to their huge aspect ratio and hydrophobic nature [4]. However, ubiquitous natural organic matter (NOM) in the aquatic environment may sorb on the CNTs and enhance their suspension [4–10], which would effect on their bioavailability and potential toxicity and attracts increasing research interests [3]. The sorption of NOM [5,6,11] and its analogues (such as humic acids (HAs) [12] and tannic acid [7,8]) on CNTs have been well studied. The effects of CNT outer-diameter [7,8,12], NOM property [6,11,12] and water quality parameters [6,8] on the sorption have also been examined. The pore size between the layers of CNTs is not large enough to accommodate macromolecules and the tube ends of CNTs are generally closed. Therefore, the macromolecular NOM and its analogues can only adsorb on the outer surfaces of CNTs, and CNTs with smaller outer-diameter and larger specific surface

D. Lin et al. / Journal of Hazardous Materials 241–242 (2012) 404–410

area have higher sorption capacity [7,8,12]. The main mechanisms underlying the NOM adsorption by CNTs can be hydrophobic interactions [12] and/or ␲–␲ interactions [6,7,13]; hydrogen bonding and electrostatic interactions may also be involved in the adsorption [8,13]. NOM with low polarity [12] and high aromaticity [6] is therefore favorable for the adsorption by CNTs. Generally, the NOM adsorption increases with decreasing solution pH and increasing ionic strength [6,8,11,14]. However, different NOM may react differently with solution pH and ionic strength, and therefore the effects of solution pH and ionic strength on the NOM adsorption by CNTs may also be NOM-property dependent, which has not been specifically investigated. Surface adsorption of NOM can facilitate dispersion and stabilization of hydrophobic CNTs in water by electrostatic repulsion, steric repulsion and/or solvation [7]. CNTs with small outerdiameters are relatively difficult to be dispersed and stabilized due to their potentially tight aggregation and entanglement [7]. Hyung and Kim [6] observed the influence of the type of NOM on the CNT stabilization; but, they did not found the relationship between CNT stabilization and NOM properties (carbon functional groups, elemental composition, and distribution of acidic functional group) and ascribed this to the collective contribution of other NOM properties such as electrostatic and configurational characteristics. Different order of the suspending capabilities of CNTs with an increasing or decreasing order of their outer-diameters by different HAs was observed by Zhou et al. [10], but they also did not establish the relationship between CNT stabilization and HA properties. We believe CNT stabilization can be correlated with some NOM properties such as polarity/hydrophilicity if the effect of other NOM properties can be largely ignored. This warrants more specific studies. CNT stabilization in NOM solutions generally increased as pH increased and ionic strength decreased [6,8]. But, the effects of the water quality parameters on the CNT stabilization may also be NOM-property dependent, which is also waiting to be explored. So far, very limited studies [6–8] discussed the relationship between NOM adsorption and CNT stabilization. Hyung and Kim [6] linearly correlated the CNT stabilization with the NOM adsorption, while our previously studies [7,8] observed that the CNT stabilization rapidly increased to a plateau and then leveled off with the continual increase of the adsorption of tannic acid by the CNTs. We assume the relationship between NOM adsorption on and stabilizing CNTs could be dominated by the properties of NOM, CNTs and solution. This study was therefore aimed to investigate the effects of NOM, CNT and solution properties on the relationship between NOM adsorption on and stabilizing CNTs in water. The specific objectives were to (1) investigate the effects of solution pH and ionic strength on the NOM adsorption and CNT stabilization as influenced by the NOM property; (2) identify the NOM property that correlates with the CNT stabilization, and (3) establish the relationship between the NOM adsorption and the CNT stabilization, and discuss its influencing factors. Four sequentially extracted HAs from a peat soil were used in order to shield the potential effect of the HA source on their distinct physicochemical properties. Three multi-walled carbon nanotubes (MWNTs) with different outer-diameters were compared.

2. Materials and methods

700 ◦ C and were then purified by mixed HNO3 and H2 SO4 solution to remove the metal catalyst and amorphous carbon. Detailed properties of the MWNTs such as elemental composition, morphology, specific surface area, and water and ash contents had been characterized in our previous paper [7] and summarized in Table S1 in the Supporting information. Four HAs utilized in this study were sequentially extracted from a peat soil (Amherst, MA, USA). The mixture of the first, second to third, fourth to fifth, and sixth to ninth extracted HA fractions were named as HAA, HAB, HAC, and HAD, respectively. Detailed extraction and purification procedures and properties such as elemental compositions of the four HAs were reported by Yang and Xing [15]. Total acidity of the four HAs was determined by the Ba(OH)2 titration method [16] in this study. The carboxyl groups were determined by the calcium acetate method [17]. Phenolic alcohol group content was calculated by subtracting the carboxyl content from the total acidity. 2.2. Adsorption and suspension experiments The HA particles were dissolved into 1 mol L−1 NaOH solution. The obtained HA solutions were filtered through 0.2 ␮m filter membrane and were adjusted to pH 7.0 using 0.1 mol L−1 HCl. The final HA solutions (about 500 mg TOC L−1 ) with 200 mg L−1 NaN3 were used as the stock HA solutions for the adsorption and suspension experiments. Batch adsorption and suspension experiments were performed to study the interactions between the HAs and the MWNTs. All the experiments were performed at room temperature. Four milligram of the MWNTs were added into 20 mL of the HA solutions with initial HA concentrations of 0, 5, 10, 20, 40, 60, 80 and 100 mg TOC L−1 . Each concentration point, including blanks (i.e., without MWNTs), was run in triplicate. The mixtures were sonicated (240 W, 40 kHz) for 20 min, and subsequently were equilibrated in a thermostat shaker (140 rpm) for 4 days. Preliminary experiments indicated that apparent equilibrium of the HA adsorption was reached within 24 h, and the effect of equilibration time on the MWNT stabilization was insignificant (Fig. S1 in the Supporting information). After equilibration, the mixtures were centrifuged at 3000 × g for 30 min. Concentration of the suspended MWNTs in the resultant supernatants was measured with a UV–vis Spectrometer (Shimadzu, UV-2540) at 800 nm. The suspended MWNTs in the supernatants of the MWNT suspensions after settling for 2–4 days were generally thought to be stabilized [5,6]. Sedimentation of the MWNT suspension by centrifugation at 3000 × g for 30 min was more than that by settling for 4 days (Fig. S2). The suspended MWNTs in the supernatants after centrifugation at 3000 × g for 30 min were therefore regarded as being stabilized in this study. The HA solutions had no absorbance at 800 nm and there was a good correlation (Fig. S3) between the absorbance at 800 nm and the MWNT concentration measured by a total organic carbon analyzer (TOC analyzer, Shimadzu, TOC-VCPH ), therefore the absorbance at 800 nm calibrated with the TOC data was used to quantify the stabilized MWNTs. The dissolved HAs in the MWNT suspensions were determined by the TOC analyzer after separating the MWNTs from the bulk using 0.2 ␮m filter membrane. Electrophoretic mobilities (EPMs) and hydrodynamic sizes of the MWNT suspensions (pH 7.0) were measured by a zetasizer (Nano ZS90, Malvern Instrument, UK). Langmuir model was employed to fit the adsorption data. Its general form [7,12] is:

2.1. Materials qe = Three MWNTs with outer diameters of <10 (MWNT10), 20–40 (MWNT40) and 60–100 nm (MWNT100) were purchased from Shenzhen Nanotech Port Co., China. The MWNTs were synthesized by chemical vapor deposition from the CH4 /H2 mixture at

405

q0 bCe 1 + bCe

where qe (mmol kg−1 ) is the amount of adsorbate adsorbed per unit mass of adsorbent, Ce (mmol L−1 ) is the adsorbate concentration at equilibrium, q0 (mg g−1 ) is the adsorption capacity of

406

D. Lin et al. / Journal of Hazardous Materials 241–242 (2012) 404–410

Table 1 Elemental composition and acidity of the humic acids (HAs). HAs

N (%)

C (%)

H (%)

Ash (%)

O (%)

Acidity (mmol g−1 )

(O + N)/C

Total acidity HAA HAB HAC HAD a

3.18 3.19 3.10 2.78

51.4 50.6 53.6 55.0

5.08 5.33 6.06 6.82

2.60 5.82 3.54 3.85

37.7 35.0 33.7 31.6

0.60 0.57 0.52 0.47

6.19 5.40 4.72 4.08

± ± ± ±

Carboxylic group

0.10 0.09 0.01 0.11

2.88 2.46 2.02 1.70

± ± ± ±

0.09 0.03 0.01 0.04

Phenolic OHa 3.31 2.95 2.70 2.38

Phenolic OH = Total acidity − carboxylic groups. The elemental composition data were from Ref. [12].

adsorbent, and b (L mmol−1 ) is the constant related to the molar heat of adsorption. 2.3. Effect of pH on the interaction between MWNTs and HAs Forty milliliters of 40 mg TOC L−1 HA solutions were added into 40 mL vials with 8.0 mg of MWNT40. The pH values of the HA solutions were adjusted to 2.0–12.0 with a few drops of 0.1 mol L−1 NaOH or HCl solutions. The mixtures were sonicated (240 W, 40 kHz) for 20 min and equilibrated (140 rpm) for 4 days. The stabilized MWNT40 and residual HAs were determined as above. EPMs of MWNT40 at various pH values and the final pH values of the mixtures were also measured. Each pH point, including blanks (i.e., without MWNT40), was run in triplicate. 2.4. Effect of ionic strength on the interaction between MWNTs and HAs Four milligram of MWNT40, 10 mL of 80 mg L−1 HA solutions and 10 mL of Ca2+ (from anhydrous CaCl2 ) solutions were mixed into 20 mL vials. Ca2+ concentrations in the mixtures were 0, 0.1, 0.5, 0.8, 1, 2, 4, 8 and 10 mmol L−1 . The suspension pH was adjusted to 7.0. The mixtures were sonicated and equilibrated, and the stabilized MWNT40 and residual HAs were quantified as above. 3. Results and discussion 3.1. Properties of the HAs Elemental composition and acidity of the HAs were listed in Table 1. The elemental ratio of (O + N)/C is a commonly used indicator of material polarity [18]. (O + N)/C of the later-extracted HAs was lower than that of the previously extracted ones, indicating that polarity of the HAs decreased with increasing extraction sequence, with an order of HAA > HAB > HAC > HAD. Total acidity, carboxyl and phenolic hydroxyl contents of the HAs all decreased with increasing extraction sequence, with the same order as the HA polarity. 3.2. HA adsorption by the MWNTs Adsorption isotherms of the HAs by the MWNTs are presented in Fig. 1. All the isotherms are apparently nonlinear and fitted well by the Langmuir model with r2 ranging from 0.934 to 0.995 (Table 2). Sorption capacities of the MWNTs for the HAs were dependent on the properties of both of the HAs and the MWNTs. For a given MWNTs, the sorption capacity increased with decreasing HA polarity, with an order of HAA < HAB < HAC < HAD (Fig. S4), suggesting that the HA adsorption could be regulated mainly by the hydrophobic interactions. Wang et al. [12] also attributed the HA adsorption by MWNTs to hydrophobic interactions. For a given HA, the sorption capacity largely decreased with increasing outer-diameter of the MWNTs with an order of MWNT10 > MWNT40 ≈ MWNT100 (Table 2), in accordance with the order of their specific surface areas (357, 86 and 58 m2 g−1 , respectively). When normalized by the specific surface area, the sorption capacity however increased

with increasing outer-diameter of the MWNTs (Table 2). This could be because the MWNTs with smaller outer-diameter had greater potential to form agglomerates and the surface areas inside the agglomerates were partly inaccessible to the macromolecular HAs. 3.3. HA stabilization of the MWNTs The MWNTs quickly agglomerated and precipitated in ultrapure water but could be stabilized in the HA solutions (Fig. 2). The stabilized MWNTs sharply increased with increasing HA concentrations and then leveled off at the concentrations higher than about 10 mg TOC L−1 . Stabilization of a given MWNTs in the HA solutions decreased with decreasing polarity of the HAs, with the order of HAA > HAB > HAC > HAD. The measured hydrodynamic sizes of the MWNTs in ultra-pure water were all larger than 1 ␮m (data not shown), while they were all lower than 400 nm (Fig. 3A) in the HA solutions. This indicated that the HAs dispersed the MWNT agglomerates. The dispersion could mainly be caused by the increased electrostatic repulsion and/or steric repulsion as the result of the HA adsorption on the MWNTs. The lower EPMs of the MWNTs in the HA solutions than that in ultra-pure water (Fig. 3B) indicated the higher electrostatic repulsion between the MWNTs in the HA solutions. However, there was no significant difference in the EPMs of a given MWNTs in the four HA solutions (Fig. 3B), suggesting that the electrostatic repulsion could be ruled out as a factor dominating the difference of the MWNT stabilizations in the four HA solutions. The later-extracted HAs were adsorbed more by the MWNTs but stabilized less amount of the MWNTs than the previously extracted ones, indicating that the steric repulsion produced by the MWNT-adsorbed HAs could not account for the difference of the MWNT stabilizations in the four HA solutions either. Solvation of O-containing hydrophilic moieties of the MWNT-adsorbed HAs could also contribute to the MWNT stabilizations by making the hydrophobic surfaces more or less hydrophilic, though it is generally assumed to make a marginal contribution [7]. The difference of the MWNT stabilizations in the four HA solutions may be attributable to the difference in the solvation of the HAs. The Table 2 Parameters of Langmuir model fitting sorption data of the HAs by the MWNTs. MWNTs

HAs

q0 (mg g−1 )

r2

qSA a (mg m−2 )

MWNT10

HAA HAB HAC HAD

181 180 186 196

± ± ± ±

15 14 7 6

0.070 0.072 0.101 0.113

± ± ± ±

0.019 0.019 0.013 0.012

0.975 0.977 0.994 0.995

0.507 0.504 0.521 0.549

MWNT40

HAA HAB HAC HAD

44.9 60.2 63.2 71.1

± ± ± ±

1.6 1.5 1.3 2.0

8.73 13.6 3.09 1.01

± ± ± ±

2.18 2.22 0.37 0.13

0.963 0.986 0.994 0.992

0.522 0.700 0.735 0.827

HAA HAB MWNT100 HAC HAD

45.0 65.1 66.1 62.3

± ± ± ±

2.7 5.8 5.6 7.1

0.087 0.042 0.054 0.064

± ± ± ±

0.020 0.011 0.015 0.025

0.972 0.975 0.968 0.934

0.776 1.122 1.140 1.074

b

a qSA stands for the specific surface area-normalized sorption capacity of the MWNTs.

D. Lin et al. / Journal of Hazardous Materials 241–242 (2012) 404–410

Adsorbed HAs (mg g-1)

200

100

100

MWNT40

MWNT10

MWNT100

160

80

80

120

60

60

80

40

40

40

20

20

0 0

20

40

60

80

100

407

0 0

20

40

60

80

0

100

0

20

40

60

80

100

Equilibrium HA conc. (mg TOC L-1) Fig. 1. Adsorption isotherms of the HAs by the MWNTs. The solid lines were Langmuir model-fitting results. The error bars represent the standard deviations (n = 3). ( HAA; (

) HAB; (

) HAC; and (

4

Stabilized MWNTs(mg L-1)

)

) HAD.

30

30

25

25

20

20

15

15

10

10

3

2

1 5

MWNT10

5

MWNT40

0

0 0

20

40

60

80

100

MWNT100

0

20

0

40

60

80

100

0

20

40

60

80

100

Initial HA conc. (mg TOCL-1) Fig. 2. Changes of the stabilized MWNTs with the initial HA concentrations. The error bars represent the standard deviations (n = 3). ( HAD.

HAs with higher polarity and acidity could have greater potential of solvation, and consequently had higher capability for stabilizing MWNTs, which merits further specific research. The stabilization of the three MWNTs in a given HA solution apparently increased with increasing outer-diameter of the MWNTs with the order of MWNT10 < MWNT40 ≈ MWNT100 (Fig. 2). Hydrodynamic sizes of a given MWNTs in the four HA solutions were similar, while the MWNTs with smaller outer-diameter had larger hydrodynamic size (Fig. 3A). The MWNTs with smaller outer-diameter could be more difficult to be dispersed and stabilized due to their stronger agglomeration and entanglement [8].

A

) HAC; and (

)

It is no doubt that the MWNT stabilization was caused by the HA adsorption. Therefore, it seems reasonable to assume that the MWNT stabilization would increase with increasing HA adsorption on the MWNT surfaces [6]. Fig. 4 shows the relationship between the MWNT stabilization and the MWNT-surface area-normalized HA adsorption (QSA ). QSA but not the adsorption amount was used in order to minimize the possible effect of the difference in the specific surface areas of the three MWNTs. From Fig. 4, it can be seen

500

0

-1

EPMs (μm s-1V-1cm)

Hydrodynamic size (nm)

) HAB; (

3.4. Relationship between the HA adsorption and the MWNT stabilization

B

600

) HAA; (

400 300 200

-2

-3

-4

100

-5

0 0

30

60

90

Initial HA conc. (mg TOCL-1)

120

0

30

60

90

120

Initial HA conc. (mg TOCL-1)

Fig. 3. Hydrodynamic sizes (A) and EPMs (B) of the MWNTs suspended in the HA solutions. (♦) HAA; () HAB; () HAC; and (×) HAD. Red for MWNT10; Black for MWNT40; Blue for MWNT100.

D. Lin et al. / Journal of Hazardous Materials 241–242 (2012) 404–410

Stabilized MWNTs(mg L-1)

408

4

35

3 2 1 0

35

MWNT40

MWNT10 30

30

25

25

20

20

15

15

10

10

5

5 0

0

0.0

0.2

0.4

0.6

0.8

1.0

MWNT100

0.0

0.2

0.4

0.6

0.8

1.0

0.0

0.2

0.4

0.6

0.8

1.0

1.2

QSA (mgm-2) Fig. 4. The relationship between the MWNT stabilization and the MWNT-surface area-normalized HA adsorption (QSA ). (

that the MWNT stabilization did increase with increasing QSA , but the increase leveled off when QSA exceeded a threshold value (QSA-T , mg m2 ). To quantitatively compare the relationships between the MWNT stabilization and QSA for the different MWNTs and HAs, we arbitrary divided the data in Fig. 4 into two groups showing the rapid increase and the equilibrium of the MWNT stabilization with increasing QSA . No obvious equilibrium of MWNT10 stabilization with increasing QSA was reached (Figs. 2 and 4), and consequently all the stabilization data of MWNT10 were distributed into the rapid increase group. The linear equation A = K · QSA was used to fit the data in the rapid increase group, where A (mg L−1 ) stood for the amount of the stabilized MWNTs and the constant K (m2 L−1 ) was the slop indicating capability of the MWNT-adsorbed HAs for stabilizing MWNTs. The average value (AE ) of the data in the equilibrium group was calculated to present the capacity of the MWNT-adsorbed HAs for stabilizing MWNTs. QSA-T can be obtained at the point of intersection of the two lines fitted by the linear equations A = K · QSA and A = AE , respectively. The obtained K, QSA-T and AE (Table 3), especially of K and QSA-T , may be not very accurate, as indicated by the relative low r2 (0.650–0.999, Table 3) of the fitness of the equation to the data of the rapid increase group. However, they did tell some important information and could be used to analyze the interaction between the MWNTs and the HAs as influenced by the properties of both of the MWNTs and the HAs. For a given MWNTs, K and AE apparently increased with increasing polarity of the HAs; HAA had obviously higher K and AE than HAD. This indicates that HAs with higher polarity had greater capability and capacity in stabilizing MWNTs. The lower QSA-T of HAA than HAD further suggests that HAs with higher polarity were more efficient in stabilizing MWNTs. Namely, less HAs with higher polarity were needed to adsorb on

Table 3 Parameters describing the relationship between the MWNT stabilization and the MWNT-surface area-normalized HA adsorption. K (m2 L−1 )

r2

AE (mg L−1 )

QSA-T (mg m−2 )

MWNTs

HAs

MWNT10

HAA HAB HAC HAD

8.0 7.2 6.7 5.6

0.650 0.657 0.719 0.879

– – – –

MWNT40

HAA HAB HAC HAD

68.8 37.8 48.8 40.3

0.780 0.916 0.819 0.964

26.6 24.0 23.9 22.2

± ± ± ±

1.8 1.6 1.8 2.2

0.387 0.635 0.489 0.552

MWNT100

HAA HAB HAC HAD

74.0 79.3 65.1 53.1

0.963 0.999 0.997 0.922

24.9 22.9 23.2 21.4

± ± ± ±

1.5 1.1 0.3 1.1

0.336 0.288 0.356 0.402

– – – –

) HAA; (

) HAB; (

) HAC; and (

) HAD.

the MWNTs to reach the maximum stabilization of the MWNTs, and the maximum MWNT stabilization by high polar HAs was greater than that by low polar HAs. For a given HAs, MWNT40 had relatively lower K and higher QSA-T than MWNT100, which could be due to that MWNTs with smaller outer-diameters had greater potential to agglomerate and needed more adsorbed HAs on their surfaces to disperse them. The fact that MWNT10 had no apparent AE and much lower K than MWNT40 and MWNT100 further indicated the difficulty in dispersing and stabilizing MWNTs with small outer-diameters. 3.5. Effect of pH on the interaction between MWNTs and HAs MWNT40 stabilization steeply increased with increasing solution pH from about 2 to 6 (Fig. 5A) and then leveled off with the further increase of pH, whereas the HA adsorption largely decreased with increasing pH. This indicated that solution pH had a significant effect on the relationship between the HA adsorption and the MWNT stabilization. Similar variations of tannic acid adsorption on and stabilizing MWNTs with solution pH had been observed in our previous study [8]. Self-deposition of the four HAs within the text pH range was insignificant (Fig. S5A) and could not contribute to the apparent HA adsorptions. The dissociation of acidic functional groups on the HAs, especially carboxyl groups, would increase with increasing solution pH, which may be responsible for the variation of the HAs-CNTs interaction with solution pH. It is known that carboxyl groups has a pKa value of about 4, and therefore about 1–99% of the carboxyl groups on the HAs may be dissociated with solution pH increasing from 2 to 6. The dissociation of the acidic functional groups would increase hydrophilicity of the HAs and could thus inhibit the HA adsorption through the hydrophobic interactions. The dissociation could impart negative charges on the HAs, which could also inhibit the HA adsorption on the negatively charged MWNT40 at pH > 4 (Fig. 5B) due to the increased electrostatic repulsion. The dissociation of protons from the polar functional groups could reduce the HA adsorption by inhibiting hydrogen bond and/or other polar interactions between the MWNT-adsorbed and dissolved HA molecules, and between the MWNT surface functional groups and the HA molecules. The hydrogen bond and/or other polar interactions among the MWNTadsorbed HA molecules could enhance the agglomeration and deposition of the MWNTs at low pH values, which could account for the nearly no stabilization of MWNT40 at low pH values (Fig. 5A). The increase of MWNT stabilization with increasing solution pH could mainly be caused by the increased electrostatic repulsion between MWNTs in the HA solutions as indicated by the increased charge negativity (decreased EPM) of MWNT40 in the HA solutions with increasing pH from 2 to 6 (Fig. 5B).

D. Lin et al. / Journal of Hazardous Materials 241–242 (2012) 404–410

25

15

HAA HAB HAC HAD

80

40

10 5

0

0.8

4

6

8

10

12

0.6 0.4

HAA HAB HAC HAD

0.2 0.0

14

0

2

4

6

HAA HAD

3

HAB blank

HAC

B

12

4.0 3.5 3.0

0

-3

2.5 2.0 1.5

-6

y = -9.795x + 8.689 R² = 0.940

0.5

2

4

6

8

10

12

14

0.0 0.4

140

0.5

0.6

0.7

(O+N)/C 120

C

100

HAA HAB HAC HAD

100

80

80

A (m2 L-1)

A(m2 L-1)

10

4.5

1.0

C

8

Ca2+ Conc. (mmol L-1)

6

CCC

EPMs (μm s-1V-1cm)

B

1.0

0 2

200 180 160 140 120 100 80 60 40 20 0

1.2

Apparent adsorption (mg g-1)

20

120

A

C/C0

160

Stabilized MWNTs(mg L-1)

Adsorbed HAs (mg g-1)

A

409

60 HAA HAB HAC HAD

40

20

60 40 20

0 2

4

6

8

10

12

14

pH Fig. 5. Changes of (A) HA adsorption on and stabilizing MWNT40, (B) EPMs of MWNT40 in ultra-pure water and the HA solutions and (C) QSA -normalized MWNT40 stabilization (A) in the HA solutions with solution pH. QSA is the MWNT40-surface area-normalized HA adsorption.

Within the examined pH range, the MWNT stabilization in the HA solutions all increased with increasing polarity of the HAs with the order of HAD < HAC < HAB < HAA (Fig. 5A). But interestingly, the order of the adsorptions of the four HAs on the MWNTs changed with increasing solution pH. The HA adsorptions increased at low pH values (ca. <4) but decreased at high pH values (ca. >4) with increasing polarity of the HAs. The higher adsorption of the HAs with lower polarity at the high pH values could mainly be due to the greater hydrophobic interactions between the HAs and the MWNTs as previously reported [12]; whereas the higher adsorption of the HAs with higher polarity at the low pH values may be attributable to the hydrogen bond and/or other polar interactions between the MWNT-adsorbed and the dissolved HA molecules, and between the MWNT surface functional groups and the HA molecules. At pH < 4, majority of the acidic functional groups of the HAs and on the MWNTs were protonated and were prone to form hydrogen bonds. The HAs with higher polarity had higher density of acidic functional groups (Table 1) and could therefore adsorb more on the MWNT surfaces at the low pH values. Fig. 5C shows variation

0 0

2

4

6

8

10

Ca2+ Conc. (mmol L-1) Fig. 6. (A) Variation of the HA adsorption and the relative MWNT stabilization in the HA solutions as a function of Ca2+ concentrations. C and C0 represent the stabilized MWNT40 concentrations in the presence and absence of Ca2+ , respectively. (B) The relationship between (O + N)/C of the HAs and the critical coagulation concentration (CCC) of Ca2+ for the HA-stabilized MWNT40; (C) The effect of Ca2+ concentration on A (the QSA -normalized MWNT40 stabilization in the HA solutions). QSA is MWNT40surface area-normalized HA adsorption.

of the capability of the MWNT-adsorbed HAs in stabilizing MWNTs (similar as the above K value, i.e., the QSA -normalized MWNT stabilization) with solution pH. On the whole, the capability increased with increasing solution pH and HA polarity. 3.6. Effect of ionic strength on the interaction between MWNTs and HAs The HA adsorption increased but the MWNT stabilization decreased with increasing solution Ca2+ concentration (Fig. 6A). Ca2+ could compress the loose structure of the HAs and make them more hydrophobic through bridge interaction among their polar functional groups [6,8,14,19,20], and therefore could facilitate their adsorption on the MWNTs through hydrophobic interactions.

410

D. Lin et al. / Journal of Hazardous Materials 241–242 (2012) 404–410

Ca2+ could also directly precipitate the HAs and increase their apparent adsorption on the MWNTs. Fig. S5B shows the direct precipitation of the HAs in the solutions increased with increasing Ca2+ concentration and was up to 58–76% of the total added HAs (40 mg TOC L−1 ), and the precipitation apparent increased with increasing HA polarity at relatively high Ca2+ concentrations. In the presence of relatively high concentrations of Ca2+ (such as >1 mmol L−1 ), the HA adsorption increased with increasing HA polarity as shown by the obvious lower adsorption of HAD than the other three HAs (Fig. 6A), presenting a reversed order of the HA adsorption in the absence of Ca2+ (Fig. 1). This could be because the higher polar HAs with higher density of acidic functional groups had higher potential to complex with the added Ca2+ and precipitate on the MWNTs. Ca2+ could compress electric double-layer of the HA-stabilized MWNTs, and consequently lead to the aggregation and precipitation of MWNT40. The critical coagulation concentration (CCC), which is defined as the concentration at which the normalized suspended colloids becomes 0.5 [8], of Ca2+ for the stabilized MWNT40 in the HAA, HAB, HAC and HAD solutions was estimated to be 2.7, 3.3, 3.5, 4.1 mmol L−1 , respectively. The CCCs negatively correlated with the polarities presented by (O + N)/C of the HAs (Fig. 6B), suggesting that the stabilized MWNTs by the HAs with higher polarity could be more subject to destabilization by Ca2+ . The QSA normalized MWNT40 stabilization in the HA solutions decreased with increasing Ca2+ concentration, but there was no significant difference among the four HAs (Fig. 6C). The high polar HAs had high density of acidic functional groups which could complex with Ca2+ and consequently bridge and precipitate the stabilized MWNTs. This may counteract their high potential of stabilizing MWNTs in the absence of Ca2+ as addressed in the above sections and lead to the no significant difference in the QSA -normalized MWNT stabilization in the four HA solutions. 4. Conclusions Hydrophobic MWNTs may be suspended in the water environment through the interaction with ubiquitous HAs. The interaction is dependent on the properties of MWNT, HA and water. MWNTs with smaller outer-diameters had higher sorption capacity for the HAs but had lower potential to be dispersed and stabilized. The HAs with higher polarity could be more efficient in stabilizing MWNTs, but the HA-stabilized MWNTs would be more subject to destabilization by Ca2+ . Low pH and high ionic strength were favorable for the apparent HA adsorption but were unfavorable for the MWNT stabilization. The HA-stabilized MWNTs may expand their potential ecological effects through transport in the environmental matrices and interaction with coexisting contaminants and organisms, which warrants more specific studies. Acknowledgments This work was supported by the National Basic Research Program of China (2008CB418204), National Natural Science Foundation of China (21077089), Zhejiang Provincial Natural Science Foundation of China (LR12B07001), Program for New

Century Excellent Talents in University (NCET-10-0731), Zhejiang Provincial “Qianjiang Talent Program” (2010R10041), Fundamental Research Funds for the Central Universities, and Zhejiang Provincial Innovative Research Team of Water Treatment Functional Materials and their Application. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j. jhazmat.2012.09.060. References [1] R.H. Baughman, A.A. Zakhidov, W.A. Heer, Carbon nanotubes-the route toward applications, Science 297 (2002) 787–792. [2] M.S. Mauter, M. Elimelech, Environmental applications of carbon-based nanomaterials, Environ. Sci. Technol. 42 (2008) 5843–5859. [3] E.J. Petersen, L.W. Zhang, N.T. Mattison, D.M. O’Carroll, A.J. Whelton, N. Uddin, T. Nguyen, Q.G. Huang, T.B. Henry, R.D. Holbrook, K.L. Chen, Potential release pathways, environmental fate, and ecological risks of carbon nanotubes, Environ. Sci. Technol. 45 (2011) 9837–9856. [4] D.H. Lin, N. Liu, K. Yang, B.S. Xing, F.C. Wu, Different stabilities of multiwalled carbon nanotubes in fresh surface water samples, Environ. Pollut. 158 (2010) 1270–1274. [5] H. Hyung, J.D. Fortner, J.B. Hughes, J.H. Kim, Natural organic matter stabilizes carbon nanotubes in the aqueous phase, Environ. Sci. Technol. 41 (2007) 179–184. [6] H. Hyung, J.H. Kim, Natural organic matter (NOM) adsorption to multi-walled carbon nanotubes: effect of NOM characteristics and water quality parameters, Environ. Sci. Technol. 42 (2008) 4416–4421. [7] D.H. Lin, B.S. Xing, Tannic acid adsorption and its role for stabilizing carbon nanotube suspensions, Environ. Sci. Technol. 42 (2008) 5917–5923. [8] D.H. Lin, N. Liu, K. Yang, L.Z. Zhu, Y. Xu, B.S. Xing, The effect of ionic strength and pH on the stability of tannic acid-facilitated carbon nanotube suspensions, Carbon 47 (2009) 2875–2882. [9] N.B. Saleh, L.D. Pfefferle, M. Elimelech, Influence of biomacromolecules and humic acid on the aggregation kinetics of single-walled carbon nanotubes, Environ. Sci. Technol. 44 (2010) 2412–2418. [10] X.Z. Zhou, S. Liang, H.B. Zhao, X.Y. Guo, X.L. Wang, S. Tao, B.S. Xing, Suspending multi-walled carbon nanotubes by humic acids from a peat soil, Environ. Sci. Technol. 46 (2012) 3891–3897. [11] C.Y. Lu, F.S. Su, Adsorption of natural organic matter by carbon nanotubes, Sep. Purif. Technol. 58 (2007) 113–121. [12] X.L. Wang, L. Shu, Y.Q. Wang, B.B. Xu, Y.C. Bai, S. Tao, B.S. Xing, Sorption of peat humic acids to multi-walled carbon nanotubes, Environ. Sci. Technol. 45 (2011) 9276–9283. [13] D.H. Lin, B.S. Xing, Adsorption of phenolic compounds by carbon nanotubes: role of aromaticity and substitution of hydroxyl groups, Environ. Sci. Technol. 42 (2008) 7254–7259. [14] S.B. Yang, J. Hu, C.L. Chen, D.D. Shao, X.K. Wang, Mutual effects of Pb(II) and humic acid adsorption on multiwalled carbon nanotubes/polyacrylamide composites from aqueous solutions, Environ. Sci. Technol. 45 (2011) 3621–3627. [15] K. Yang, B.S. Xing, Sorption of phenanthrene by humic acid-coated nanosized TiO2 and ZnO, Environ. Sci. Technol. 43 (2009) 1845–1851. [16] E.M. Perdue, Acidic functional groups of humic substances, in: G.R. Aiken, et al. (Eds.), Humic Substances in Soil, Sediment and Water: Geochemistry, Isolation and Characterization, John Wiley & Sons, New York, 1985, pp. 439–526. [17] M. Schnitzer, Organic matter characterization, in: A.L. Page, et al. (Eds.), Methods of Soil Analysis, Part 2, Agronomy 9, Wisconsin American Society of Agronomy and Soil Science Society of America, Madison, 1982, pp. 581–594. [18] D.H. Lin, B. Pan, L.Z. Zhu, B.S. Xing, Characterization and phenanthrene sorption of tea leaf powders, J. Agric. Food Chem. 55 (2007) 5718–5724. [19] S. Ghosh, H. Mashayekhi, P. Bhowmik, B.S. Xing, Colloidal stability of Al2 O3 nanoparticles as affected by coating of structurally different humic acids, Langmuir 26 (2010) 873–879. [20] S.J. Zhang, T. Shao, S.S.K. Bekaroglu, T. Karanfil, Adsorption of synthetic organic chemicals by carbon nanotubes: effects of background solution chemistry, Water Res. 44 (2010) 2067–2074.