Science of the Total Environment 470471 (2014) 9298

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Effect of humic acid source on humic acid adsorption onto titanium dioxide nanoparticles
Mohamed Erhayem, Mary Sohn
Department of Chemistry, Florida Institute of Technology, 150 West University Boulevard, Melbourne, FL 32901 USA

H I G H L I G H T S Extent of humic acid adsorption onto nano-TiO2 depending on HA source Preferential HA adsorption onto nano-TiO2 depending on aromaticity Humic acid effects on sedimentation, aggregation and stability of nano-TiO2

a r t i c l e

i n f o

a b s t r a c t
In many studies, different humic acid (HA) sources are used interchangeably to evaluate the effect of organic matter on geochemical processes in the environment. This research looks more specically at the effect of HA source on HA adsorption onto nano-TiO2 and how HA adsorption affects the fate and transport of nano-TiO2. In this study, six humic acids (HAs) were studied which were derived from soils (SLHA), or from sediments (SDHA) all originating from the state of Florida. Humic acid adsorption onto titanium dioxide nanoparticles (nano-TiO2) and the sedimentation of HA-coated and uncoated nano-TiO2 were monitored by Ultravioletvisible (UVvis) spectroscopy. Synchronous scan uorescence (SSF) spectroscopy was used to complement the study of HA adsorption onto nano-TiO2. Phosphate buffer was found to reduce the amount of HA adsorbed onto nano-TiO2 relative to solutions of NaCl of the same pH and ionic strength. Adsorption constant values (Kads) for HAs varied in the order SLHA N FSDHA (freshwater sedimentary HA) N ESDHA (estuarine sedimentary HA). SSF results suggested that the more highly conjugated fractions of HA, which are more prevalent in SLHAs versus SDHAs, were preferentially adsorbed. In order to better understand the relationship between adsorption and aggregation, sedimentation studies were conducted and it was found that the percentage of nano-TiO2 sedimentation was preferentially enhanced in the order of the presence of SLHA N FSDHA N ESDHA. The extent of nano-TiO2 sedimentation was decreased with increasing HA concentration. TEM imaging of nano-TiO2 conrmed that nano-TiO2 was aggregated in the presence of HAs. The ndings in this study suggest that HAs from different sources inuence the fate and transport of nano-TiO2 in the environment differently. 2013 Elsevier B.V. All rights reserved.

Article history: Received 29 July 2013 Received in revised form 19 September 2013 Accepted 22 September 2013 Available online xxxx Editor: Damia Barcelo Keywords: Soil and sedimentary humic acid Titanium dioxide nanoparticles Adsorption constants Sedimentation

1. Introduction The distribution of metal oxide nanoparticles (NPs) such as CuO, ZnO, Al2O3 and TiO2 in commercial products such as industrial lters, ceramics, coatings, catalysts, semiconductors, and microelectronics has rapidly increased in recent years (Robichaud et al., 2009). In particular, titanium dioxide nanoparticles (nano-TiO2) are one of the most widely produced nanoparticles because of their excellent dielectric and chemical properties (Schmidt and Vogelsberger, 2009). Some uses include stabilizing agents in sunscreens, cosmetics and paints (Aitken et al., 2006; Chen and Mao, 2007; Robichaud et al., 2009; Domingos et al.,

Corresponding author. Tel.: +1 321 674 7379; fax: +1 321 674 8951. E-mail address: msohn@t.edu (M. Sohn). 0048-9697/$ see front matter 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.scitotenv.2013.09.063

2010) and in water treatment plants and pollution remediation due to its high specic surface area and sorption capacity for ionic and nonionic species (French et al., 2009). This huge demand for nano-TiO2 has increased the extent of its possible entry into the environment and several studies have shown that nano-TiO2 is toxic to plants and animals (Aruoja et al., 2009; Singh et al., 2010; Gao et al., 2012). Yang et al. (2013) described the enhanced toxicity of nano-TiO2 to developing zebra sh (Danio rerio) in the presence of Suwannee River humic acid, while Lin et al. (2012) found that both adsorbed and dissolved humic acid reduced the toxicity of nano-TiO2 to Chlorella species. Therefore, an understanding of the effect of humic acid on the fate and transport of nano-TiO2 in environments is crucial. The environmental partitioning of nano-TiO2 in the environment depends on processes such as dispersion and agglomeration which are affected by environmental factors such as the presence of natural

M. Erhayem, M. Sohn / Science of the Total Environment 470471 (2014) 9298

93

organic matter (NOM), including humic acid (HA) and solution chemistry (Wang et al., 2012; Chowdhury et al., 2012). Humic acid has been found to inhibit the aggregation of NPs associated with its adsorption onto NPs (Kilduff et al., 1996; Vermeer and Koopal, 1998; Vermeer et al., 1998; Thio et al., 2011). Chen et al. (2012) reported enhanced mobility of nano-TiO2 in sand columns at an acidic pH but lower mobility at a basic pH due to variation in humic acid adsorption onto nano-TiO2, as well as signicant variation in nano-TiO2 deposition as a factor of humic acid concentration and pH. Humic acids are formed from the microbial decomposition of animal and plant residues and are widely distributed on the planet in soils, sediments, and natural waters (Rezacova and Gryndler, 2006; Pospisilova and Fasurova, 2009). The variability of HA structure is based on its source and environmental exposure. Therefore, studies on how humic acids from different sources affect nano-TiO2 fate and transport are extremely important. The relative abundance of HA functional groups such as carboxylic acid and phenol, as well as aromatic rings and free radicals has a direct effect on HA properties (Chen et al., 2000; Pospisilova and Fasurova, 2009; Kwiatkowska-Malina, 2011; Lishtvan et al., 2012). Some studies have shown that slow microbial decomposition of lignin, which contains benzene carboxylic and phenolic components, is a major contributor to soil humic acids (SLHAs), while sedimentary humic acids (SDHAs) are derived primarily from algal or bacterial residues (Malcolm, 1990). HAs extracted from different environments have different characteristics which can be correlated to the abundance of various functional groups (McDonald et al., 2004; Rezacova and Gryndler, 2006). These differences in HA characteristics may have a signicant effect on the extent of HA adsorption onto NPs and ultimately affect the extent of NP aggregation and deposition. Humic acids are frequently used in studies that look at the effect of naturally occurring organic matter on the fate and transport of NPs. The effect of HA characteristics and sources on HA adsorption onto nano-TiO2 has not been evaluated when studying the fate and transport of NPs. The main objective of this work was to expand our understanding of the effect of HA source on the environmental fate and transport of nano-TiO2.

2.2. Preparation of HA stock solutions Humic acid stock solutions were prepared by dissolving the dry powders (80 to 100 mg) of HA samples in either 100 mmol L 1 NaH2PO4 or 100 mmol L 1 NaCl. The solution pH was adjusted using concentrated HCl or NaOH as required. To dissolve the HAs, the solutions were gently stirred overnight at room temperature and any undissolved HA was removed by a previously weighed 0.45 m lter (MF Membrane Filter, Cat No.: HAWP04700); HA concentrations were adjusted for the weight of the undissolved HA. The HA stock solutions were stored in the dark and their concentrations were routinely checked by measuring the absorbance at 254 nm. The solution pH was also checked routinely. For each experiment, solutions of HAs were prepared by diluting the prepared stock solution in either 100 mmol L 1 NaH2PO4 or 100 mmol L 1 NaCl to produce 10 mg L 1 HA (nal solution pH adjusted to 7.8 0.2). 2.3. Instruments An Agilent Technology 8453 spectrophotometer was used to perform all UVvis absorbance measurements using a 1 cm optical path length quartz cuvette. A FluoroMax-3 (J-Y/Horiba SPEX Fluorolog spectrophotometer) was used for synchronous scan uorescence (SSF) measurements. The HA solutions were placed in a single 1 cm (4 mL) quartz cuvette and analyzed by SSF spectrophotometry. As was reported in a previous study in this laboratory (Horst et al., 2013), the synchronous scan uorescence spectra were measured over a range of 240 to 650 nm at a constant wavelength difference (em ex) of 55 nm. The slit widths for both the excitation and emission monochromators were held at 7 nm band pass and increments of 1.0 nm and an integration time of 0.3 s were used. The TEM images for selected HA-coated and uncoated nano-TiO2 were captured using a low resolution (0.5 nm) Zeiss EM-900 instrument with an accelerating voltage of 80 kV using 400-mesh Carbon/Formvar TEM grids (Electron Microscopy Sciences). 2.4. Time study In order to determine the time at which the equilibrium of HA-nanoTiO2 suspensions was attained, 5 mg of nano-TiO2 was added to 10 mL of 50 mg L 1 of each HA solution (pH adjusted to 7.8 0.2 and ionic strength adjusted to 100 mmol L 1). The HA-nano-TiO2 suspensions were placed on a stirring plate for 96 h. During the experimental time period, 2 mL aliquots of the suspensions were ltered at 0, 4, 8, 12, 24, 48, 72, and 96 h using a 0.45 m syringe lter (Nylon Syringe Filters, Pore: 0.45 m, diameter: 13 mm). Filtered solutions were analyzed by UVvis absorbance measurements at 254 nm. HA concentrations were determined from calibration curves. 2.5. Adsorption study In order to evaluate adsorption isotherms, nano-TiO2 was added in increments of 5 mg for a total of 025 mg to 10 mL of 10 mg L 1 of the various HA solutions. The HA-nano-TiO2 suspensions were placed on a stirring plate for 48 h (based on time study). The suspensions were then ltered using 0.45 m syringe lters. Samples were analyzed using UVvis absorbance measurements at 254 nm. The mass of HA that was adsorbed onto nano-TiO2 was determined from the difference in the absorbance before and after equilibration with nano-TiO2. After completion of the adsorption measurements, synchronous scan uorescence spectroscopy was used as a complementary technique, to provide more information about HA adsorption onto nano-TiO2. Finally, the nano-TiO2 suspended in HA solutions and their controls were subsequently utilized for further studies such as sedimentation measurements and TEM imaging.

2. Experimental methods 2.1. Materials Titanium dioxide nanoparticles (nano-TiO2) (CAS-Reg. No: 1346367-7) were purchased from Degussa (nominal primary diameter of about 21 nm reported by manufacturer). In order to study the effect of HA extracted from different sources on HA adsorption onto nano-TiO2, HAs which had previously been extracted from soils, freshwater sediments, and estuarine sediments from the state of Florida (Fig. S-1 and Table S-1) were utilized in this study. Soil humic acid samples included Florida Institute of Technology Jungle (SL1HA) and West Melbourne (SL2HA). A freshwater sedimentary humic acid (FSDHA) sample was derived from the Sebastian River (SD1HA) while estuarine sedimentary humic acids (ESDHAs) were derived from Kings Bay (SD2HA and SD3HA) and Mullet Creek (SD4HA). The extraction and purication procedures for HA samples were identical to those described in Cameron and Sohn (1992), however, an additional pre-extraction acidication step was undertaken to dissolve CaCO3 and then wash Ca2+ from the carbonate-rich Florida samples, because not doing so would have hindered the extraction processes by forming insoluble HA-calcium salts in alkaline solutions. These samples were stored in desiccators until use. All chemicals were purchased from Sigma-Aldrich or Fischer Scientic and used without further purication. All aqueous solutions were prepared using Milli-Q water puried by an 18.2 M Milli-Q purication system (Billerica, MA, USA). The pH of solutions was adjusted with NaOH and HCl using an Orion pH/ISEmeter (model 710A, Thermo Fisher Scientic, Waltham, MA). All experiments were performed at room temperature (22 1 C).

94

M. Erhayem, M. Sohn / Science of the Total Environment 470471 (2014) 9298

2.6. Sedimentation study In order to evaluate the effect of HA source on the fate and transport of nano-TiO2, sedimentation of nano-TiO2 was monitored by optical absorbance as a function of time. The HA-nano-TiO2 suspensions were prepared as described above in the adsorption study section. After the suspension attained equilibrium, a 2.5 mL aliquot was then dispensed into a 3 mL UVvis cuvette and placed in the UVvis spectrophotometer for sedimentation measurements recorded at 378 nm every 150 s over a 2 h time period. Calibration curves of nano-TiO2 concentration versus absorbance at 378 nm were initially constructed and the wavelength choice was based on lack of interference from HA absorbance spectra. 2.7. Transmission electron microscopy (TEM) TEM was used in this study to conrm the nano-TiO2 aggregation after HA adsorption onto nano-TiO2 surfaces. In order to obtain TEM images of HA-coated nano-TiO2, the HA and nano-TiO2 suspensions were prepared as described above in the adsorption section. For imaging, 1 mL of the nano-TiO2 suspension was diluted with Milli-Q water to a total volume of 10 mL. After SLHAking to ensure complete mixing, two drops of the nano-TiO2 suspensions were placed onto carbon copper grids and allowed to sit overnight. Images were collected the next day. 3. Results and discussion 3.1. Effect of HA source on HA adsorption onto nano-TiO2 In order to determine the time needed to attain equilibrium adsorption, the absorbance at 254 nm was measured over time for HA solutions before and after addition of nano-TiO2. Generally, the trend was a rapid decrease in absorbance with time after addition of nanoTiO2, followed by a leveling off towards the end of the experiment (Fig. S-2). The UV data clearly showed that no signicant additional adsorption of HA occurred after 48 h. This result was in good agreement with previous research by Yang et al. (2009) in which it was determined that the equilibrium time for HA adsorption onto nano-oxides (nano-TiO2, nano-Al2O3 and nano-ZnO) was reached before 48 h at an initial HA concentration range of 10200 mg L 1, at a pH of 5.0 and an ionic strength of 100 mmol L 1. Thus, 48 h was chosen as the time at which HA adsorption onto nano-TiO2 was evaluated. Isotherms for the adsorption of HAs onto nano-TiO2 were constructed by plotting the equilibrium concentration of the solute in the solid phase versus the equilibrium concentration of solute in the liquid phase (Fig. S-3). When plotting solution versus solid phase equilibrium concentrations of adsorbate, all isotherms most closely resembled Freundlich isotherms due to their linearity. Alternatively, Langmuir isotherms were also constructed for adsorption studies using phosphate buffer and adsorption constants evaluated from these plots were not signicantly different in value from those derived from Freundlich isotherms. Since HA is known to consist of a complex mixture, adsorption equations which describe the adsorption of a single adsorbate such as the Langmuir isotherm may not be appropriate for interpretation of the data (Stumm, 1992). The Freundlich equation that is used in this study to evaluate the HA-nano-TiO2 adsorption constant, as also used by Chen et al. (2011) to describe HA adsorption onto nano-TiO2, can be mathematically viewed as the following equation: S Kads C
1=n

the isotherm, and the term (1/n) is the Freundlich parameter indicating the degree of nonlinearity (Hyung and Kim, 2008). When adsorption experiments were run in aqueous 100 mmol L 1 NaCl solutions, adsorption constants were signicantly larger than those determined in phosphate buffer, suggesting that the phosphate may compete with HA on the surface of nano-TiO2 (Fig. S-4) (Weng et al., 2008). Another possible explanation is that the phosphate may be interacting with the HA in the solution and affecting HA adsorption. Overall, the phosphate has a signicant effect on the adsorption process. Therefore, adsorption isotherms were determined in 100 mmol L 1 NaCl, as described in the experimental section. In Table 1, the Kads values of HAs (pH of 7.8 0.2 and an ionic strength of 100 mmol L 1 NaCl) are reported. The Kads values of SLHAs were consistently higher than those of FSDHA and ESDHAs which is consistent with characterization data of HAs. 13C NMR data indicated that aromatic content was higher in SLHAs than that in ESDHAs and the ratios of aromatic to aliphatic carbon for SLHAs were higher than those for ESDHAs as expected since HAs are largely derived from plant materials which contain aromatic lignin in terrestrial ecosystems (Horst et al., 2013). The aromatic content and the ratio of aromatic to aliphatic carbon content of the FSDHA (SD1HA) were similar to those of SLHAs. This is likely due to a signicant terrestrial input to the FSDHA sample which is the only riverine sample in this study. With a higher aromatic content of HA, the hydrophobicity of the HA is also higher and a larger fraction of HA is adsorbed onto the surface of nano-TiO2 resulting in a larger Kads value. Therefore, HA source has a signicant effect on HA adsorption onto the surface of nano-TiO2, which can be related to the HA aromaticity. However, aromaticity alone cannot be the only factor affecting the extent of HA adsorption onto nano-TiO2 in this study since Kads values for SL1HA and SD1HA are signicantly different while aromaticities are nearly identical (Table 1). Functional group content must also be an important factor. Carboxylic carbon makes up 14.9% of the carbon in SL1HA while for SD1HA, the value is 11.0% (Table 1). Additionally, SLHAs, derived from terrestrial lignin-rich sources are known to have higher phenolic contents than SDHAs (Ishiwatari, 1969; Sohn and Hughes, 1981). Thus the higher carboxylic and phenolic content of SLHAs relative to SDHAs in these samples is likely a contributing factor to the higher Kads values. 3.2. Fluorescence studies Initially, HA samples were studied using SSF without nano-TiO2 additions (Fig. 1a, b (0 mg nano-TiO2 added) and Fig. S-5). While SSF spectra showed that all HAs exhibit peaks at similar wavelengths, for the ESDHAs the peak around 280 nm is more prominent than in SLHAs. Meanwhile, SLHAs exhibit a very broad uorescence peak at long wavelengths (450 nm) as also noted in previous studies (Kalbitz et al., 1999; Milori et al., 2002; Alberts and Takacs, 2004; Horst et al., 2013) which is not evident in ESDHAs. The presence of the SSF peak at 450 nm indicates the presence of more highly condensed aromatic uorophores (Senesi et al., 1991; Horst et al., 2013). Also, the SSF spectrum of sample SD1HA is very similar to that of the SLHAs, although it was derived from freshwater sediment. This may be attributed to the origin of SD1HA which is an FSDHA sample with the largest expected terrestrial content of any of the sedimentary samples because it is extracted from riverine sediment which was subject to terrestrial runoff. In a previous study in our laboratory (Horst et al., 2013), uorescence intensity ratios at 450 and 280 nm were found to be useful for differentiating between HA sources (I450/I280). The ratio of these two peak intensities was also found to be useful in this study when considering the differences between soil and sedimentary HAs (Table S-2). The I450/I280 ratios of all the SLHAs were close to 1, while ESDHAs had values less than 1. Also, the areas under the peak at 280 nm (A1 = area under peak from 220 to 300 nm) and the 450 nm peak (A3 = Area under peak from 420 to 520 nm) were studied. The A3/A1 values for SLHAs were higher than those observed for ESDHAs. This is in good agreement with the results from 13C-NMR for HAs (Table 1). These results conrmed that SLHA

where S (mg g 1) is the concentration of the solute adsorbed onto the solid phase and C (mg L 1) is the concentration of the solute in solution at equilibrium, Kads is the adsorption constant indicating the adsorption capacity of the surface for the solute and is determined from the slope of

M. Erhayem, M. Sohn / Science of the Total Environment 470471 (2014) 9298 Table 1 Adsorption constants and carbon-13 NMR data percentages of various carbon types in each humic acid sample according to chemical shift range. Type HA samples Chemical shift range (ppm) Aliphatic (050) Soil Soil Freshwater sediment Estuarine Sediment Estuarine sediment Estuarine sediment
a b

95

KKadsb Acetal (100110) 3.20 3.40 5.00 3.20 4.30 3.00 Aromatic (110160) 38.7 25.9 38.0 19.4 23.5 20.0 Carboxyl (160200) 14.9 24.1 11.0 7.1 7.1 10.0 Aromatic/aliphatic carbona 1.85 1.05 1.72 0.49 0.94 0.50 10.3 4.33 3.89 2.29 2.24 2.54

Substituted aliphatic (50100) 22.3 21.9 24.0 30.4 40.2 29.0

SL1HA SL2HA SD1HA SD2HA SD3HA SD4HA

20.9 24.7 22.0 39.9 24.9 40.0

Ratio of aromatic to aliphatic carbon = (percentage peak area of 13C-NMR spectrum 110160 ppm) / (percentage peak area of 13C-NMR spectrum 050 ppm). HA samples dissolved in 100 mmol L1 NaCl. 5, 10, 15, 20 and 25 mg of nano-TiO2 were added to 10 mL of 10 mg L1 HA solutions (pH = 7.8 0.2, ionic strength = 100 mmol L1, T = 22 1 C). The average of standard deviation was 2.57 103.

percentage of SL1HA removal

samples contained a greater fraction of condensed aromatic uorophores than did ESDHAs. In order to obtain more information about HA adsorption onto nanoTiO2, various HA samples were studied using SSF spectroscopy before and after the addition of nan-TiO2 to HA solutions. SLHA samples exhibited a signicantly different trend in adsorption onto nano-TiO2 than did ESDHA samples as can be seen in Fig. 1. For SL1HA adsorption (Fig. 1a), the peak intensity around 450 nm decreased extensively with the rst addition of nano-TiO2. This suggests that the functional groups in SLHAs which uoresce at longer wavelengths (450 nm) were initially adsorbed on the surface of nano-TiO2 more extensively than those which uoresce at shorter wavelengths. For SD4HA (an estuarine sedimentary HA (ESDHA)) adsorption (Fig. 1b), the peak intensity around 280 nm decreased most extensively as nano-TiO2 was initially added, likely due to the preferential adsorption of the functional groups in ESDHAs which uoresce at shorter wavelengths. As can be seen in Fig. 1, adsorption of functional groups which uoresce at long wavelengths is more extensive for soil samples, in which these groups are much more prevalent. Fig. 2 shows the differences between SLHA and ESDHA adsorption patterns more clearly. In SL1HA, representative of SLHAs, the peak intensity at 450 nm continues to decrease preferentially

upon consecutive additions of nano-TiO2 (Fig. 2a) and there is a signicant difference between the extents of adsorption at 450 nm versus 280 nm. A similar trend was observed for all SLHAs used in this study. In SLHAs, when the condensed aromatic structures are prevalent, they preferentially adsorb onto the nano-TiO2 surface relative to the less condensed aromatic structures. SLHA has a greater non-polar character and a higher aromaticity that cause the SLHAs to be less soluble in polar solvents and adsorb rst from solution onto the surface of nano-TiO2. On the other hand, ESDHA samples showed peak intensity at 280 nm was decreased extensively after the addition of nano-TiO2 (Fig. 2b). In ESDHAs, where the condensed aromatic structures are limited, there is less signicant difference between the adsorption at 450 nm compared to 280 nm. This is possibly due to the ESDHA having fewer condensed, or less conjugated aromatic uorophores than is found in the SLHAs, suggesting a more homogeneous mixture. Thus quenching of HA uorescence is believed to be due to HA adsorption onto nano-TiO2, since quenching increases as the concentration of suspended nano-TiO2 increases. Preferential adsorption of highly condensed aromatic

120 100 80 60 40 20 0 0 5 10 15 20 25 30
450 nm 280 nm

a
2e+6
SL1HA

Intensity (cps)

2e+6 1e+6 5e+5 0 300 400 500 600

0 mg 5 mg 10 mg 15 mg 20 mg 25 mg

Amount of nano-TiO2 added (mg) percentage of SD4HA removal Wavelength (nm)

120 100 80 60 40 20 0 0 5 10 15 20 25 30
450 nm 280 nm

b
Intensity (cps)

2e+6 2e+6 1e+6 5e+5 0 300 400 500 600

SD4HA
0 mg 5 mg 10 mg 15 mg 20 mg 25 mg

Amount of nano-TiO2 added (mg)

Fig. 2. Effect of nano-TiO2 concentration on the percentage removal of HA by nano-TiO2 at 280 and 450 nm. (a) SL1HA and (b) SD4HA. 5, 10, 15, 20 and 25 mg nano-TiO2 were added to 10 mL of 10 mg L 1 HA. Solutions pH = 7.8 0.2; ionic strength = 100 mmol L 1 NaCl; T = 22 1 C.

Wavelength (nm)
Fig. 1. Effect of nano-TiO2 concentration on SSF spectra of HA with = 55 nm. (a) SL1HA and (b) SD4HA. 5, 10, 15, 20 and 25 mg nano-TiO2 were added to 10 mL of 10 mg L1 HA. Solution pH = 7.8 0.2; ionic strength = 100 mmol L1 NaCl; T = 22 1 C.

96

M. Erhayem, M. Sohn / Science of the Total Environment 470471 (2014) 9298

structures as seen most dramatically in SLHA samples is likely associated with the lower solubility of these fractions. However, since ring structures in HAs are known to be highly substituted with functional groups such as carboxylic acid, phenol, etc., it is likely that these functional groups are also involved in adsorption and the associated uorescence quenching. 3.3. Sedimentation studies Sedimentation removes NPs from the water column and decreases their mobility, which signicantly decreases NP bioavailability to organisms in the water column and increases bioavailability to benthic organisms. During the time and adsorption isotherm studies described above, once stirring was stopped and solutions were allowed to sit, particle sedimentation was visually obvious and believed to be due to aggregation of particles. Since the extent of adsorption varies with HA source, HA origin should affect the fate and transport of nano-TiO2 in the environment. In order to determine the effect of HA source on the sedimentation of HA-coated nano-TiO2, UVvis sedimentation measurements at 378 nm were performed over a 2 h time period as described in the experimental section for all of the HAs used in this study. Fig. 3 shows the effect of HA source on the sedimentation of HA coated-nano-TiO2 at a pH 7.8 0.2 and an ionic strength of 100 mmol L 1 NaCl relative to nano-TiO2 suspended in water and in 100 mmol L 1 NaCl. An initial concentration of 10 mg L 1 of HA in NaCl destabilized the nano-TiO2 relative to water without electrolytes present, however, the destabilization was slightly less extensive for SLHAs than ESDHAs and the most extensive destabilization occurred in NaCl without HA present. The results suggest a decrease in electrostatic repulsive forces between NPs, leading to agglomeration and eventual particle settling. All HAs tend to destabilize nano-TiO2 relative to the Milli-Q water, but tend to stabilize the particles relative to the presence of 100 mmol L 1 NaCl. It is likely that Na+ neutralizes the charge on the surface of nano-TiO2, leading to the agglomeration and eventual particle settling. But the HAs modify that destabilization of particles due to the adsorption of HA onto the surface of nano-TiO2 and because the adsorption of SLHAs is more extensive, they affect the particle surface to a greater extent than do the ESDHAs. An increase in the charge on the nano-TiO2 surface due to HA adsorption leads to an increase in the electrostatic or steric repulsive forces between the particles resulting in particle stabilization. Therefore, the sedimentation of nano-TiO2 in the presence of HA decreased as more HA was adsorbed onto the surface of nano-TiO2. In addition, ESDHAs appeared more effective at neutralizing the charge on nano-TiO2 than did SLHAs, resulting in more sedimentation. Thus, the source of HAs could affect the fate and transport of nano-TiO2 in natural environments differently. In the experiments described above, it was hypothesized that SLHAs stabilized NPs more extensively than do ESDHAs in the presence of NaCl
1.2 1.0

due to more extensive adsorption. In order to evaluate that hypothesis, concentrations of SL1HA (pH of 7.8 0.2), a soil HA, were increased incrementally from 10 to 50 mg L 1 (Fig. 4). For nano-TiO2 suspended in Milli-Q water, a very slight decrease of 378 absorbance over time was observed, suggesting minimal sedimentation. Nano-TiO2 suspended in 10 mg L 1 SL1HA showed the most extensive NP sedimentation, while for nano-TiO2 suspended in 25 and 50 mg L 1 SL1HA, nanoTiO2 aggregation and settling was virtually eliminated. Thus the presence of 10 mg L 1 SL1HA in the nano-TiO2 suspension resulted in increased aggregation and sedimentation likely due to a decrease in the electrostatic repulsive forces between nano-TiO2 particles due to HA adsorption. However at higher HA concentrations, more extensive HA adsorption resulted in increased repulsive forces between the particles, causing NP stabilization. In other words, more extensive HA adsorption onto the nano-TiO2 surfaces at higher dissolved HA concentrations may have resulted in more extensive steric repulsive forces between the nano-TiO2 particles, providing a greater energy barrier to aggregation. Therefore, the fate and transport of nano-TiO2 is strongly dependent on dissolved organic matter concentrations in natural environments. 3.4. TEM study In order to examine the effect of HA on nano-TiO2 aggregation after HA adsorption, TEM images were captured for selected coated nanoTiO2 samples, as shown in Fig. 5. The TEM images showed that nanoTiO2 appeared to be largely agglomerated in the presence of 100 mmol L 1 NaCl (Fig. 5b), 10 mg L 1 SD4HA and 100 mmol L 1 NaCl (c), and 10 mg L 1 SL2HA and 100 mmol L 1 NaCl (d) likely due to charge neutralization of the nano-TiO2 surface as discussed earlier. These results suggest that both HAs at the relatively low concentration of 10 mg L 1 and NaCl increased nano-TiO2 aggregation. 4. Conclusions The adsorption, sedimentation, aggregation and stability of nanoTiO2 are greatly affected in the presence of HA. The anionic background electrolyte (phosphate versus chloride) signicantly affected HA adsorption onto nano-TiO2. A possible competition between HA and phosphate for adsorption onto nano-TiO2 surfaces resulted in a lower amount of HA adsorbed onto the surface of nano-TiO2, which may strongly inuence the fate and mobility of nano-TiO2 in natural environments. SLHAs exhibit signicantly different trends than do ESDHAs with respect to adsorption onto nano-TiO2 over time. The adsorption constants of SLHAs characteristically have higher values than those of ESDHAs in the presence of NaCl. The difference in Kads values may be related to differences in aromaticity of the HA structure. At low concentrations of HA (10 mg L 1), destabilization of the nano-TiO2 particles in solution was observed even in the absence of NaCl. However at HA concentrations 25 mg L 1, stabilization of nano-TiO2 particles was

1.2 1.0
H2O SL1HA+NaCl SL2HA+NaCl SD1HA+NaCl SD2HA+NaCl SD3HA+NaCl SD4HA+NaCl H2O+NaCl

A/Ao @ 378 nm

0.8 0.6 0.4 0.2 0.0 0 20 40 60 80 100 120

0.8

A/Ao

0.6 0.4 0.2 0.0 0 20 40 60 80 100 120

H2O

10 mg L-1 SL1HA 25 mg L-1 SL1HA 50 mg L-1 SL1HA

Time (min)
Fig. 3. Effect of HA source on 2 h UVvis sedimentation measurements of nano-TiO2 suspensions. Five mg nano-TiO2 was added to 10 mL of 10 mg L1 HA. Solutions pH =7.8 0.2; ionic strength = 100 mmol L1 NaCl; T = 22 1 C; = 378 nm.

Time (min)
Fig. 4. Effect of HA concentration on 2 h UVvis sedimentation measurements of nanoTiO2. Five mg of nano-TiO2 was added to 10 mL of 10, 25 and 50 mg L1 SL1HA. Solutions pH = 7.8 0.2; ionic strength = 100 mmol L 1 NaCl; T = 22 1 C; = 378 nm.

M. Erhayem, M. Sohn / Science of the Total Environment 470471 (2014) 9298

97

Fig. 5. Selected TEM images of nano-TiO2 suspended in: (a) Milli-Q water, (b) 100 mmol L 1 NaCl, (c) 10 mg L1 SD4HA and 100 mmol L1 NaCl, and (d) 10 mg L1 SL2HA and 100 mmol L1 NaCl. In each case 5 mg of nano-TiO2 was added to 10 mL of HA. Solution = pH of 7.8 0.2; T = 22 1 C. TEM images of samples were captured at a magnication of 85,000.

enhanced. The SSF spectra indicated that there was preferential adsorption of polycondensed structures which are more prevalent in soil HAs than in sedimentary HAs. Overall the results indicate that soil humic acids affect the environmental stability and mobility of nano-TiO2 differently than do sedimentary humic acids because of structural differences. Finally, the TEM images of HA-coated nano-TiO2 indicated that nanoTiO2 aggregation occurred. Acknowledgments Authors acknowledge the nancial support of Libyan High Education. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.scitotenv.2013.09.063. References
Aitken RJ, Chaudhry MQ, Boxall ABA, Hull M. Manufacture and use of nanomaterials: current status in the UK and global trends. Occup Med (Lond) 2006;56:3006. Alberts JJ, Takacs M. Total luminescence spectra of IHSS standard and reference fulvic acids, humic acids and natural organic matter: comparison of aquatic and terrestrial source terms. Org Geochem 2004;35:24356. Aruoja V, Dubourguier H, Kasemets K, Kahru A. Toxicity of nanoparticles of CuO, ZnO and TiO2 to microalgae Pseudokirchneriella subcapitata. Sci Total Environ 2009;407: 14618. Cameron DF, Sohn ML. Functional group content of soil and sedimentary humic acids determined by CP/MAS carbon-13 NMR related to conditional zinc(2+) and cadmium(2+) formation constants. Sci Total Environ 1992;113:12132. Chen X, Mao SS. Titanium dioxide nanomaterials: synthesis, properties, modications, and applications. Chem Rev 2007;107:2891959. [Washington, DC, U. S.]. Chen J, LeBoeuf EJ, Choi S, Gu B. In Spectroscopic characterization of structural and functional properties of natural organic matter fractions; 2000. p. ENVR-322.

Chen Q, Yin D, Hu X. Inuence of solution conditions on adsorption of humic acid onto titanium dioxide nanoparticles. Huan Jing Ke Xue 2011;30:797802. Chen G, Liu X, Su C. Distinct effects of humic acid on transport and retention of TiO2 rutile nanoparticles in saturated sand columns. Environ Sci Technol 2012;46:714250. Chowdhury I, Cwiertny DM, Walker SL. Combined factors inuencing the aggregation and deposition of nano-TiO2 in the presence of humic acid and bacteria. Environ Sci Technol 2012;46:696876. Domingos RF, Peyrot C, Wilkinson KJ. Aggregation of titanium dioxide nanoparticles: role of calcium and phosphate. Environ Chem 2010;7:616. French RA, Jacobson AR, Kim B, Isley SL, Penn RL, Baveye PC. Inuence of ionic strength, pH, and cation valence on aggregation kinetics of titanium dioxide nanoparticles. Environ Sci Technol 2009;43:13549. Gao J, Powers K, Wang Y, Zhou H, Roberts SM, Moudgil BM, et al. Inuence of Suwannee River humic acid on particle properties and toxicity of silver nanoparticles. Chemosphere 2012;89:96101. Horst C, SLHArma VK, Clayton Baum J, Sohn M. Organic matter source discrimination by humic acid characterization: Synchronous scan uorescence spectroscopy and Ferrate(VI). Chemosphere 2013;90:20139. Hyung H, Kim J. Natural organic matter (NOM) adsorption to multi-walled carbon nanotubes: effect of NOM characteristics and water quality parameters. Environ Sci Technol 2008;42:441621. Ishiwatari R. An estimation of the aromaticity of a lake sediment humic acid by air oxidation and evaluation of it. Soil Sci 1969;107:537. Kalbitz K, Geyer W, Geyer S. Spectroscopic properties of dissolved humic substances a reection of land use history in a fen area. Biogeochemistry 1999;47:21938. Kilduff JE, Karanl T, Weber Jr WJ. Competitive interactions among components of humic acid in granular activated carbon adsorption systems: effects of solution chemistry. Environ Sci Technol 1996;30:134451. Kwiatkowska-Malina J. Properties of soil and elemental compisition of humic acids after treatment with brown coal and cow manure. Pol J Soil Sci 2011;44:4350. Lin D, Ji J, Long Z, Yang K, Wu F. The inuence of dissolved and surface-bound humic acid on the toxicity of TiO2 nanoparticles to Chlorella sp. Water Res 2012;46:447787. Lishtvan II, Kaputskii FN, Abramets AM, Yanuta YG, Monich GS, Aleinikova VN, et al. Fractionation of humic acids as method receiving of standard humic material. Vestn BGU Ser 2012;2:711. Malcolm RL. The uniqueness of humic substances in each of soil, stream and marine environments. Anal Chim Acta 1990;232:1930. McDonald S, Bishop AG, Prenzler PD, Robards K. Analytical chemistry of freshwater humic substances. Anal Chim Acta 2004;527:10524. Milori DMBP, Martin-Neto L, Bayer C, Mielniczuk J, Bagnato VS. Humication degree of soil humic acids determined by uorescence spectroscopy. Soil Sci 2002;167:73949.

98

M. Erhayem, M. Sohn / Science of the Total Environment 470471 (2014) 9298 Stumm W. Chemistry of the Solid-Water Interface: Processes at the Mineral-Water and Particle-Water Interface in Natural Systems; 1992. p. 1124. [June]. Thio BJR, Zhou D, Keller AA. Inuence of natural organic matter on the aggregation and deposition of titanium dioxide nanoparticles. J Hazard Mater 2011;189:55663. Vermeer AWP, Koopal LK. Adsorption of humic acids to mineral particles. 2. Polydispersity effects with polyelectrolyte adsorption. Langmuir 1998;14:42106. Vermeer AWP, van Riemsdijk WH, Koopal LK. Adsorption of humic acid to mineral particles. 1. Specic and electrostatic interactions. Langmuir 1998;14:28109. Wang Y, Gao B, Morales VL, Tian Y, Wu L, Gao J, et al. Transport of titanium dioxide nanoparticles in saturated porous media under various solution chemistry conditions. J Nanopart Res 2012;14:1095/19. Weng L, Van Riemsdijk WH, Hiemstra T. Humic nanoparticles at the oxidewater interface: interactions with phosphate ion adsorption. Environ Sci Technol 2008;42:874752. Yang K, Lin D, Xing B. Interactions of humic acid with nanosized inorganic oxides. Langmuir 2009;25:35716. Yang SP, Bar-Ilan O, Peterson RE, Heideman W, Hamers RJ, Pedersen JA. Inuence of humic acid on titanium dioxide nanoparticle toxicity to developing zebrash. Environ Sci Technol 2013;47:471825.

Pospisilova L, Fasurova N. Spectroscopic characteristics of humic acids originated in soils and lignite. Soil Water Res 2009;4:16875. Rezacova V, Gryndler M. Fluorescence spectroscopy: a tool to characterize humic substances in soil colonized by microorganisms? Folia Microbiol 2006;51:21521. [Prague, Czech Repub.]. Robichaud CO, Uyar AE, Darby MR, Zucker LG, Wiesner MR. Estimates of upper bounds and trends in nano-TiO2 production as a basis for exposure assessment. Environ Sci Technol 2009;43:422733. Schmidt J, Vogelsberger W. Aqueous long-term solubility of titania nanoparticles and titanium(IV) hydrolysis in a sodium chloride system studied by adsorptive stripping voltammetry. J Solution Chem 2009;38:126782. Senesi N, Miano TM, Provenzano MR, Brunetti G. Characterization, differentiation, and classication of humic substances by uorescence spectroscopy. Soil Sci 1991;152: 25971. Singh N, Jenkins GJS, Asadi R, Doak SH. Potential toxicity of superparamagnetic iron oxide nanoparticles (SPION). Nano Rev 2010;1. [No pp. given.]. Sohn ML, Hughes MC. Metal ion complex formation constants of some sedimentary humic acids with Zn(II), Cu(II) and Cd(II). Geochim Cosmochim Acta 1981;45:23939.