Environ. Sci. Technol. 2007, 41, 1339-1344
Quantifying the Dimensions of Nanoscale Organic Surface Layers in Natural Waters C . T . G I B S O N , †,‡ I A N J . T U R N E R , § C L I V E J . R O B E R T S , | A N D J . R . L E A D * ,† School of Geography, Earth and Environmental Sciences, University of Birmingham, Birmingham B15 2TT, Biology Sciences Research Group, School of Education, Health and Science, University of Derby, Derby DE22 1GB, and Laboratory of Biophysics and Surface Analysis, School of Pharmacy, The University of Nottingham, University Park NG7 2RD, United Kingdom
Nanoscale surface films are known to develop on surfaces exposed to natural waters and have potential impacts on many environmental processes. A new method using atomic force microscopy is presented which physically removes the developed film in a defined area and then quantifies the difference in height between the film and the area where the film has been removed. The difference gives the absolute thickness of the surface film, which has not previously been measured. Suwannee River humic acid was exposed to substrates, and the surface film thickness as a function of pH and exposure time was measured. Discrete and very small colloids in the range 1-5 nm were observed as expected, and these sat on a coherent surface film, not the original mica substrate. Low pH values of 2 gave rise to relatively thick surface films of about 3 nm, although these films were not continuous at higher pH values. At pH 4.8, the film thickness increased with exposure time up to about 5 h and did not subsequently increase. The maximum film thickness measured was about 1 nm at that pH. The method is applicable to the measurement of many environmental surfaces, although resolution will depend on the substrate and film roughness.
suitable fractionations and analysis. These techniques include field flow fractionation (11), atomic force microscopy (12), laser-induced breakdown detection (13), and electron microscopy (14). Nevertheless, quantifying these solid phases, and especially the surfaces where initial binding occurs, is essential (8). Interactions between surfaces and colloidal and nanoparticulate material have been known for many years, and films are usually thought to be composed of organic acids, primarily humic substances (HS), although inorganic material such as manganese or iron oxides may also bind to solid surfaces. These surface phases have long been thought to be important in the solid-solution interface and have a controlling influence on the uptake of trace pollutants to the solid phase as well as processes of aggregation (1, 2) and biofilm formation. Direct evidence, primarily from electrophoretic measurements, has been available for over 30 years showing that small organic colloids (typically humic substances) sorb to surfaces and control surface charge (15-17) in marine waters, estuarine waters, and freshwaters. More recently, atomic force microscopy (AFM) has been used in force mode to assess the change in the surface characteristics of mica and iron oxide surfaces in the presence and absence of humic substances (18, 19) and relatively unperturbed natural water samples (20). AFM has also been used in imaging mode to estimate the variability in the thickness of these naturally derived surface films (but not the absolute thickness) (21, 22). In addition, the AFM tips have been used as mass sensors to quantify the amount of material deposited on a silicon nitride surface, but without a direct measurement of the surface layer (8). Much of the previous work (8, 12, 23) has primarily focused on quantifying discrete nanocolloids. In this paper, we primarily focus on the underlying surface film developed. In this paper, we present two related aspects of work on such nanoscale films. First, we show a methodology new to the environmental sciences, based on AFM nanomachining and subsequent imaging, for the determination of the absolute thickness of organic nanoscale surface layers. Second, we present the first data showing the absolute thickness of these surface layers as a function of solution conditions and deposition time.
Methodology Introduction Trace pollutant speciation and the subsequent fate and behavior of pollutants in the aquatic environment are often dominated and controlled by their interactions with naturally occurring solid-phase material such as nanoparticles, colloids, and particles (1-5). Nanoparticles (1 µm) are complex materials composed of organic, inorganic, and biological phases (6-8). They are typically poorly characterized structurally due to their complexity, the ease with which they are altered by sampling and treatment processes, and their low mass concentrations (9, 10) and also because techniques are only now becoming available which are capable of providing * Corresponding author phone: 44 121 414 8147; fax: 44 121 414 5528; e-mail:
[email protected]. † University of Birmingham. ‡ Current address: Laboratory of Biophysics and Surface Analysis, School of Pharmacy, The University of Nottingham, University Park NG7 2RD, U.K. § University of Derby. | The University of Nottingham. 10.1021/es061726j CCC: $37.00 Published on Web 01/13/2007
2007 American Chemical Society
Standard Suwanee River humic acid (SRHA) supplied by the International Humic Substances Society (IHSS) was used to prepare water samples. The samples were prepared in demineralized water (R ) 18 mΩ cm) at a concentration of 10 mg L-1 prepared from the freeze-dried solid, without further pretreatment. The water samples were studied at a range of pH values (2-12) and exposure times (2-24 h) in batch mode; i.e., all samples were independent. For all water samples, the salt concentration was 5 mM NaCl and the pH was adjusted by the addition of dilute acid (HNO3) or base (KOH). No buffers were used to ensure no interferences with measurements. Surfaces for AFM analysis were prepared in a manner similar to that described elsewhere (23). Approximately 1 cm2 of muscovite mica was cleaved and then immersed in the water sample of interest. For the pH dependence experiments the immersion time was constant at 30 min, while for the exposure time experiments the immersion time was between 30 min and 24 h, at a constant pH of 4.8. After immersion the mica was removed and then suspended in water (R ) 18 mΩ cm) for 30 s to remove any material not adsorbed to the surface. The mica was then placed in a clean VOL. 41, NO. 4, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1339
FIGURE 1. Diagram showing how AFM was used to measure the thickness of humic layers. In (a), a tapping mode image of a humic layer is obtained (image size approximately 5 µm). In (b), a contact mode image is acquired which penetrates and removes the humic layer (image size approximately 1 µm). Then in (c), the area is imaged again in tapping mode. dry container, open to the atmosphere to allow air drying (relative humidity (RH) ca. 50%), but covered to prevent deposition of air particles. The surfaces were left overnight to dry under ambient conditions before AFM analysis. AFM images were obtained using a Multimode Nanoscope IIIa with a type E scanner (Veeco, Cambridge, U.K.). The tips used were Si tapping mode probes with a spring constant of 34-75 N m-1 (Olympus, OMCL-AC160TS). All experiments were performed in air at room temperature. Samples were thus air-dried at an RH value of approximately 60-70%. It is likely that these samples retained sufficient hydration water to maintain their structure (23). The thickness of the humic layers was determined using a three-stage process (refer to Figure 1). Stage 1. An initial image of approximately 2-5 µm2 was acquired in “tapping” mode (24). All tapping mode images were recorded in topography mode with a pixel size of 512 × 512 at a scan rate of 1-2 Hz. Tapping mode is a popular technique for studying soft materials because of the reduction in lateral and vertical forces which may cause sample movement and deformation. In tapping mode the surface topography is measured with an oscillating probe that “taps” the surface. The oscillation amplitude is used as the control signal. This technique is used for different applications, especially for imaging delicate and fragile samples such as weakly adsorbed DNA molecules (25), colloidal particles (12), or surfaces of polymer materials (26). Stage 2. The tip was retracted and the AFM configuration changed to contact mode. The tip was then brought back into feedback with the surface and scanned at 10-20 Hz. The contact mode scans were 0.5-1 µm in size, and the force offset was typically 0.15-0.2 V. It was found that, during control experiments, force offsets greater than 0.2 V damaged the mica substrates underneath the humic layer. Complete removal of the HS layer was ensured by comparing the roughness values of the scratched surface with pristine mica and the HS layer. Stage 3. After the contact mode image was acquired, the tip was retracted and the AFM configuration returned to tapping mode. The tip was then brought back into feedback and the “scratched” section reimaged. The image was set to the same size as the initial tapping mode image in stage 1. From the AFM images line traces (cross-sections) were obtained which allowed the thickness of the humic layer to be determined. Typically, 5-15 line traces were used for each image. This “scratch” method was performed two to three times on each surface, and an average of all the line traces for each surface was then used to determine the thickness of the humic layer. Since the tapping mode tips came into sustained contact with the surfaces during the scratch procedure, contamination and eventual damage to the tips were inevitable. Therefore, changing old tips for new tips was crucial, and up to three tips could be used to analyze each surface. The tips were considered too contaminated once image quality and resolution became poor. This would 1340
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 4, 2007
FIGURE 2. A typical tapping mode image of a layer developed on the mica after sorption of SRHA. The line across the image shows the point at which the transect was measured. Ra and Rrms (( standard errors on the measurements) for this section were 0.49 ( 0.02 and 0.42 ( 0.02 nm, respectively. also often show up as double or triple peaks in the resonance profiles for the tips. Potentially, this AFM preparation method, especially airdrying and imaging in tapping mode, can lead to some artifacts, which we have thoroughly discussed in a previous paper (8). Nevertheless, air-drying may retain hydration water, which maintains the sample shape (23), and imaging in liquid introduces other potential artifacts such as tip contamination. Tapping mode is an improvement over contact mode imaging of deformable material and is currently the recommended method. Other groups have used similar AFM nanomachining techniques to determine the thickness of dendrimer monolayers (27) and spin-cast polymer thin films (28) and measure the wear resistance of thin carbon films (29), nanopatterning (30), and the plowing friction of thin polymer films (31). This is, to our knowledge, the first use for environmental samples.
Results and Discussion In this paper, we report a new method of applying AFM to quantify the absolute thickness of nanoscale films on environmental aquatic surfaces. Previously, we have measured only the surface roughness of the developed film (32) or performed indirect measurements (8). The method shows similarities to focused ion beam (FIB) methods (33) associated with electron microscopy, but has the advantage of lower spatial resolution in the z direction and can be performed in air and potentially in liquid rather than in ultrahigh vacuum (UHV) as with FIB. This difference is key to environmental applications as nanoparticles can remain hydrated, and relatively unmodified, throughout the measurement. Mica was used as a substrate as it is flat at an atomic scale and is a relevant surface in environmental aquatic systems. Likewise, HS is a reasonable analogue for unperturbed colloids and nanoparticles, and SRHA is a type of HS widely used as a surrogate for unperturbed organic macromolecules. Figure 2 shows a typical AFM image of the mica after sorption of HS collected at pH 4.8. In addition, a line transect along the image, with heights above the surface, is displayed. The images are typical of those collected previously of extracted HS, and the line transects are similar to those we have collected from natural waters (8). The heights of
TABLE 1. Thickness Measurements from Lines a-c in Figure 3a line
thickness, t (nm)
a b c
1.1 0.8 1.0
a pH 4.8, exposure time 5.5 h. The average thickness from these measurements is 1 nm.
FIGURE 3. A tapping mode image of a humic layer that has a 1 × 1 um2 area scratched away in contact mode. Lines a-c that cut across the image are where the cross-sections below the image were taken. individual peaks are all less than 5 nm and most substantially lower. These results are in good agreement with data on HS size from AFM and other techniques (23). Nevertheless, we have focused on the developed background film and not the discrete colloids in this work. The surface roughness is low with Ra (average roughness) and Rrms (root-mean-square roughness) values of 0.49 ( 0.02 and 0.42 ( 0.02 nm, respectively, where Ra is the arithmetic average of the absolute values of the surface height deviations measured from the mean plane of the image. Rrms is the root-mean-square average of height deviations taken from the mean data plane. Comparable roughness values for blank mica (exposed to pure water only) are
