Formation of single-crystalline CuS at the organic–aqueous interface

To prepare nano-films of CuS at the liquid–liquid interface, copper cupferronate, Cu(C₆H₅N₂O₂)₂ or [Cu(cup)₂], was used as the copper source in the toluene medium and Na₂S as the sulfur source in the aqueous medium. In a typical preparation for carrying out x-ray scattering measurements, 60 ml of 0.24 mM Cu(cup)₂ solution in toluene was slowly added to 14.5 ml of 3.74 mM Na₂S in water taken in a sample cell (7 cm × 7 cm), as shown in figure 1. The CuS film forms only at the toluene–water interface and gradually turns green, while the two liquid phases remain colorless. In order to prevent the formation of Cu₂S, an excess of Na₂S was required. The molar ratio of Cu(cup)₂ and Na₂S is chosen as 1:4 for this experiment. The cell was mounted on an anti-vibration table to stabilize the interface for carrying out the x-ray measurements during the interfacial reaction at room temperature. For microscopy measurements this interfacial reaction was carried out in a glass beaker of similar volume and the reaction was stopped at different times by removing the upper toluene layer gently without disturbing the formed CuS nano-films. The nano-films could be lifted onto various substrates by means of forceps to enable electron microscopy and atomic force microscopy measurements.

The in situ high-energy x-ray scattering study was carried out at the ID10B beamline of the European Synchrotron Radiation Facility (ESRF). The energy was set to 22 keV to allow the x-ray beam to pass through the upper toluene layer and probe the buried liquid–liquid interface. In figure 1 we have shown the schematic of the sample cell used for these measurements. The Teflon cell has two circular windows on diametrically opposite walls and two Kapton sheets of 25 μm thickness are fixed on both windows. It allows x-rays to go in and out of the cell and prevents liquid leakage during the scattering measurements. Two thin Si-wafers of equal heights were used near the entry and exit x-ray windows of the cell to anchor the water–toluene interface flat (refer figure 1). All the measurements were performed at room temperature (25 ° C) with varying incident beam sizes: 0.002 × 1.0 mm² (V × H) from 0° to 0.2°, 0.02 × 1.0 mm² from 0.2° to 0.5° and 0.1 × 1.0 mm² up to 1.5° grazing angle of incidence, defined using conventional slits.

Zoom In Zoom Out Reset image size

In the x-ray reflectivity technique, the scattered x-ray intensity normalized to the incident intensity is measured at very small grazing angles as a function of the vertical component of the wavevector q_z while keeping the horizontal components q_xy equal to zero. These measurements were carried out as a function of the reaction time by varying the grazing angles θ_i and θ_f (with θ_i = θ_f), while keeping the in-plane angle φ = 0 (refer figure 1). The reflectivity measurements provide us information about the electron density of the interfacial nano-films as a function of depth during the formation process. To obtain in-plane information of the interfacial nano-films, we carried out diffuse scattering measurements. A position sensitive detector (PSD) was used to collect the data as a function of the in-plane angle (φ), keeping the incident grazing angle (θ_i) fixed at 20 millidegrees. We also collected longitudinal diffuse scattering data in the same geometry with an offset of 0.2° in θ_f to find the background arising of reflectivity data. Scattering data in the same geometry from the toluene bulk were also collected by moving down the interface by 0.2 mm to obtain background data arising from bulk scattering. It should be mentioned here that for such measurements of the liquid surface and liquid–liquid interface, the x-ray beam is tilted using a crystal reflector and the sample cell is positioned by moving a goniometer vertically and horizontally to achieve appropriate scattering conditions, as shown in figure 1, as a liquid cell cannot be tilted like a solid surface.

We have carried out transmission electron microscope (TEM) measurements using a FEI (Technai S-twin) microscope operating with an accelerating voltage of 200 kV. For TEM studies the film was lifted from the interface onto a holey carbon coated copper grid at various reaction times and high resolution (HRTEM) images, selected area electron diffraction (SAED) patterns and energy dispersive x-ray spectroscopy (EDS) data were taken for each sample. We used Gatan Digital Micrograph software for the analysis of the TEM images and SAED patterns. Scanning electron microscopy (SEM) images were also recorded with a Quanta 200 FEG microscope. Tapping mode atomic force microscope (AFM) experiments were carried out on samples lifted onto RCA cleaned Si-wafer surfaces using a Digital Instruments Nanoscope-IV, employing standard Si₃N₄ cantilevers. The height profile and particle distributions were obtained from atomic force microscopy images using WS × M software.

Isothermal titration calorimetry (ITC) measurements [24] were carried out using a VP-ITC micro-calorimeter to determine the nature of the reaction. Two identical coin shaped cells are enclosed in an adiabatic outer shield. Both cells were filled with the liquids during operation. The working volume of the cell is approximately 1.5 ml. The temperature differences between the reference cell and a sample cell are measured and it is calibrated to differential power (DP) units between the cells. This DP signal is used to maintain temperature equilibrium as the feedback power. Calibration of this DP signal is obtained by monitoring a known quantity of power through a resistive heater located on the cell.

In this study, we titrated 286 μl Cu(cup)₂ toluene solution of different concentrations into 1.4 ml aqueous Na₂S at a fixed molar ratio (1:4) to get the heat evolution or absorption during the reaction. We also titrated 286 μl pure toluene into 1.4 ml aqueous Na₂S to measure the heat of dilution (control). In both cases we used the single injection mode (SIM) technique without stirring the solutions at room temperature (25 ° C). Here 'reaction' indicates the sum of the heat of dilution during mixing of two liquids and the heat during formation of chemical products (CuS) and 'control' means the heat of dilution of two liquids without any chemical reaction. The heat evolved during formation of CuS (enthalpy of CuS formation) is thus obtained from the difference of 'reaction' and 'control'.

In figure 2(a), we have presented measured and fitted reflectivity data. The reflectivity data of the water–toluene interface was measured before initiating the interfacial reaction and this data (R_f) was used to normalize all the other reflectivity data (R) measured as a function of reaction time. In figure 2(b), we show three representative R/R_f curves collected after the initiation of the reaction and the time of data collection are also indicated. The reflectivity curves were initially fitted with a model-independent procedure [25] and the obtained electron density profile (EDP) was then approximated with a box model having a minimum number of layers (two here) and interfacial roughness of the layers. In figure 2(c), we show the two layers as dashed lines and the roughness convoluted EDP as solid lines that closely represent EDPs obtained from the model-independent analysis. The solid fitted lines (red) shown in figures 2(a) and (b) are obtained with this two-box model [26]. The higher density box is assigned to the CuS layer and the lower density box just above water is assigned to the cupferronate layer that gets precipitated during the reaction. The thicknesses of the CuS and organic layers obtained are 6.5 nm, 5.3 nm, 5.6 nm and 8.4 nm, 7.9 nm, 9.1 nm after 18 min, 1 h 40 min and 2 h 57 min, respectively. The interfacial roughness values are found to be constant over the reaction time and we obtain roughness values to be 1.1 nm, 2.8 nm and 3.8 nm at the toluene–CuS, CuS–organic and organic–water interfaces respectively. Precipitation of the organic film between the CuS layer and water stops the reaction progressing further and the total film thickness between the toluene and water layers reaches a nearly constant value of 27 nm (see figure 2(c)).

Zoom In Zoom Out Reset image size

In the inset of figure 2(a), we show a typical diffuse scattering intensity profile after subtracting the bulk toluene data as a function of q_x by integrating the PSD data. We could determine [27] the effective interfacial tension and effective roughness of the toluene–water interface from the decay of intensity as ${q}_{x}^{-(2-\eta )}$ (where η is equal to ${k}_{\mathrm{B}}{T q}_{z}^{2}/(2 \pi \gamma )$, with γ as the interfacial tension) by fitting the straight line (blue line) shown in the log–log plot, giving η = 0.19. We get the interfacial tension value as 0.87 mN m⁻¹, giving an effective roughness of 3 nm. It is to be noted here that this roughness value is similar to the value of organic–water roughness obtained from reflectivity analysis. The value of interfacial tension is much lower than that expected (27.8 mN m⁻¹) for a pure toluene–water interface due to presence of the cupferronate. It is to be noted that the observed diffuse scattering profile is more complicated than the predicted straight line in the log–log plot for capillary waves, and the interfacial tension obtained is only an estimate.

In figure 3, we show a representative result from our tapping mode atomic force microscopy (AFM) measurements. The height profile and histogram show the signature of two thicknesses, one at 12 nm and the other at 27 nm, measured from the top of the silicon substrate. The second peak at 27 nm gives the total film thickness and is consistent with the obtained x-ray results. The composite first peak indicates that the 27 nm film is generated by stacking of CuS and organic layers having various thicknesses.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In figures 4(a), (b) and (e) we show typical transmission electron microscope (TEM) images and the corresponding selected area diffraction patterns (SAED) of CuS films obtained from the toluene–water interface at 18 min, 1 h 40 min and 3 h 10 min after initiation of the reaction. While the SAED patterns of the films obtained after 18 min and 1 h 40 min show diffuse rings, those obtained after 3 h 15 min exhibit sharp spots. The SAED patterns clearly reveal the gradual formation of the crystalline CuS film. The films are made of single-crystalline CuS and these crystals are essentially defect-free. The diffraction spots are indexed on the basis of the hexagonal structure (P63/mmc,a = 3.792 Å and c = 16.34 Å, JCPDF no. 06-0464). The crystallinity of the films increases with increasing reaction time.

Energy dispersive x-ray spectroscopic (EDS) data in figure 4(d) show the presence of Cu Kα, Lα, L β and S Kα lines. It is interesting that the normalized (with respect to Cu Kα) intensity of Cu Lα and S Kα increase continuously with the reaction time. Our x-ray reflectivity results (see figure 2(c)) clearly show that the CuS structures are embedded in the organic layers; we, therefore, monitored the lower energy Cu Lα and S Kα to probe the formation process. The binding of a large number Cu²⁺ and S²⁻ ions initially form nano-crystals of CuS that remain embedded in organics and cannot exhibit diffraction spots. The intensity enhancement of Cu Lα and S Kα signals indicate that these tiny crystals aggregate to form large crystallites of CuS towards the organic surface through phase separation as the reaction time increases.

In figure 4(c), we show a typical TEM image of 'soft' composite films that form in the early phase of the reaction, showing the homogeneous presence of the cupferronate, (light) and CuS films (dark). These films show the presence of Cu Lα and S Kα signals, but no diffraction data in TEM. After about 3 h 10 min of reaction, the films start showing crystal formation, as indicated by the SAED pattern in figure 4(e) (inset: TEM image from which SAED pattern was taken). A typical high-resolution TEM image in figure 4(f) shows an interplanar spacing (d spacing) of 2.04 Å. This d spacing is close to the (008) interplanar spacing of hexagonal CuS. The moiré fringes (figure 4(g)) formed by interfering two sets of CuS lattice planes having nearly common periodicities. A high-resolution TEM image (figure 4(h)) shows collection of oriented nano-crystals, having a d spacing of 3.28 Å of the (100) plane of a hcp CuS crystal. These results show that small crystallites come together to form bigger crystallites, possibly through a 'contact epitaxy' process [28].

Figures 5(a) and (b) show the large-scale organization of CuS nano-flowers exhibiting hexagonal in-plane ordering. Such films were obtained after about 2 h 15 min of the reaction. The inset in figure 5(a) shows the SAED pattern of these nano-flowers, confirming their polycrystalline nature. It is interesting that the average size crystallites obtained finally (figure 5(c)) are similar to those of nano-flowers. The electron density of CuS from reflectivity analysis remained at 0.63, compared with the expected bulk density of 1.294 electrons Å⁻³, suggesting a fractional occupancy of only 0.5 in a unit volume over the entire reaction time.

Zoom In Zoom Out Reset image size

In figure 5(b), we show a 440 nm sized hexagonal box for one nano-flower, where only 220 nm is dark due to the presence of CuS. In figure 5(d), we show a simple model of such arrangements of the nano-flowers. Out of the cylindrical volume of length 27 nm and diameter 440 nm, the top 5.7 nm holds the nano-flowers with 0.5 fractional occupancy. The rest of the cylinder is occupied by cupferronate with an electron density of 0.16 electrons Å⁻³, about 40% of the value of the crystalline state. The number of CuS and cupferronate molecules in this model cylinder results as 6.0 × 10⁶ and 8.2 × 10⁶, respectively. This model calculation shows that almost all of the cupferronate molecules are near the water surface as a low-density spongy layer containing a large amount of pores and voids.

The obtained electron density profiles from x-ray reflectivity data clearly show that the amount of CuS crystals at the liquid–liquid interface remains constant for more than 3 h. Electron diffraction results shown in figure 4 clearly show that there is a growth in crystal size over this reaction time and the final crystal size (refer figure 5(c)) is similar to the size of the nano-flowers (refer figure 5(a)).

Zoom In Zoom Out Reset image size

As a nano-crystal grows at the oil–water interface, three contributions to the interfacial energy, namely the particle–oil interface (γ_p/o), the particle–water interface (γ_p/w), and the oil–water interface (γ_o/w) change. ΔE due to the assembly of a single particle at the oil–water interface is given by [29]

$$\begin{eqnarray*} &&\Delta E=\frac{\pi {R}^{2}}{{\gamma }_{\mathrm{o}/\mathrm{w}}}\times [{\gamma }_{\mathrm{o}/\mathrm{w}}-({\gamma }_{\mathrm{p}/\mathrm{w}}-{\gamma }_{\mathrm{p}/\mathrm{o}})]^{2} \end{eqnarray*}$$

where R is the effective radius of the nanoparticle. In the present case, as the CuS crystals (size R) grow, we expect to see heat evolution over 3 h time. We carried out calorimetric measurement to probe this growth process. ITC measurements show that for each concentration of the reactants, heat evolution from the reaction during 3 h from the initiation of the reaction is different, and found to be higher for higher concentrations (figure 6(a)). In figure 6(b), we show the control (one of the titrants is missing, thereby no reaction occurs) subtracted heat for different concentrations with time, with the inset providing the enthalpy changes charactering the chemical reaction at three different concentrations of the titrants. Titration of 0.12 mM, 0.18 mM and 0.24 mM 286 μl of Cu(cup)₂ in toluene with 0.10 mM (green), 0.15 mM (blue) and 0.20 mM (red) 1.4 ml of Na₂S aqueous solution gives heats of 31 kcal mol⁻¹,46 kcal mol⁻¹ and 61 kcal mol⁻¹ respectively. Figure 6(c) represents the resolution function (orange) of the ITC setup for a standard binding reaction with a typical heat evolution for CuS formation (wine).