Abstract
Nanoparticle suspensions represent a promising route toward low cost, large area solution deposition of functional thin films for applications in energy conversion, flexible electronics, and sensors. However, parameters such size, stoichiometry, and electronic properties must be controlled to achieve best results for the target application. In this report, we demonstrate that such control can be achieved via in situ chemical oxidation of nanoparticles in suspensions. Starting from a microwave-synthesized suspension of ultrasmall ( nm) nanoparticles in n-butanol, we added H2O2 at room temperature to chemically oxidize the nanoparticles. We systematically varied H2O2 concentration and reaction time and found that they significantly affected oxidation state and work function of nanoparticle films. In particular, we achieved a continuous tuning of work function from 4.4 to 5.0 eV, corresponding to oxidation of as-synthesized nanoparticle (20% Mo6+) to essentially pure MoO3. This was achieved without significantly modifying nanoparticle size or stability. Such precise control of stoichiometry and work function is critical for the optimization of nanoparticles for applications in organic optoelectronics. Moreover, the simplicity of the chemical oxidation procedure should be applicable for the development of other transition oxide nanomaterials with tunable composition and properties.
1. Introduction
Metal oxide nanoparticles represent a large class of materials with applications in areas such as energy conversion and storage, [1, 2] catalysis, [3, 4] sensing, [5] and biomedicine [6]. Major advantages of metal oxide nanoparticle suspensions include compatibility with low-temperature, large-area solution processing, versatile surface functionalization, and formation of complex architectures via self-assembly. Molybdenum oxide () has attracted much interest as a hole transport layer (HTL) material in organic light-emitting diodes and solar cells because of its high work function (Φ). By matching Φ of the HTL to the highest occupied molecular orbital (HOMO) of the organic electron donor material, [7] films inserted between the active layer and the anode have been shown to improve performance of organic photovoltaic (OPV) devices [8–15]. Most work in this area use thermally evaporated films [8–11]. While solution deposition of HTL films has been reported, all approaches thus far have required a postdeposition processing step that must be performed either at high temperature (≥160°C) [12–14] and/or in an O2-containing ambient [15] to obtain films with high Φ. In addition, the solvents currently used for the solution deposition of do not wet the organic layer to form a uniform thin film as required in an inverted OPV architecture, which shows superior stability in air compared to conventional devices [16, 17]. Recently, our group developed a microwave-assisted synthesis of nanoparticles (np) in n-butanol suspension, demonstrated formation of uniform thin films on poly(3-hexylthiophene): -phenyl-C61-butyric acid methyl ester (P3HT : PCBM) blends using room temperature solution processing and examined performance of inverted OPV devices using np films as HTLs [18]. Here we focus on in situ chemical oxidation to optimize the np properties for HTL application. Specifically, we examined the effect of chemical oxidation conditions on the size and stability of the np suspension to evaluate its suitability for the formation of uniform and pinhole-free thin films. We also quantified the impact of chemical oxidation on Φ and stoichiometry thin film and demonstrated continuous tuning of Φ from 4.4 eV to 5.0 eV through precise control of the chemical oxidation conditions and Mo oxidation state. We show that in situ chemical oxidation of nanoparticle is a versatile technique to synthesize stable suspensions of ultrasmall nanoparticles with desired stoichiometry and Φ for the formation of thin HTL on top of organic active layer without postprocessing.
2. Experiments
nanoparticle (np) suspensions were synthesized using nonhydrolytic sol-gel conversion of molybdenum dioxide bis(acetylacetonate) (MoAcAc) in anhydrous n-butanol as shown schematically in Figure 1, following a microwave-assisted synthesis procedure modified from Bilecka and colleagues [19]. We selected this approach because it has been shown to yield various metal oxide nanoparticles with low size polydispersity, good stability, and short reaction times down to 3 min. Furthermore, the procedure does not require additional ligands for nanoparticle stabilization, and thus should allow the deposition of nanoparticle films with good carrier transport properties without the need to remove electrically insulating ligands through a postsynthesis step. We used n-butanol as the solvent because of the good solubility for MoAcAc, and because it wets, but does not dissolve or swell hydrophobic organic films such as P3HT : PCBM. For the microwave synthesis, a test tube of MoAcAc solution in n-butanol was mixed in N2 and placed in a microwave reactor (CEM Discovery) containing a single-mode 2.54 GHz microwave cavity. A MoAcAc concentration of 0.1 M was selected to balance complete dissolution and strong absorption microwave to enable rapid heating to 200°C within three minutes. The MoAcAc solution was maintained at 200°C for 15 min to synthesize the brown np suspension (Figure 1). We chose to chemically oxidize np in suspension using H2O2 for three reasons: H2O2 is a strong oxidizing agent, H2O2 is miscible with n-butanol, and byproducts of the reaction, for example, H2O and O2, can be easily removed during subsequent processing. For chemical oxidation, a small amount of 30 wt% H2O2 in H2O (Fisher) was added to the np suspension to achieve the desired concentration (e.g., 10.2 μL of 30 wt% H2O2 in H2O per mL of suspension for 0.1 M H2O2). The mixture was then stirred at room temperature for times ranging from 20 min to 24 hr for oxidation to occur.
To determine the nanoparticle size and distribution, an aliquot of each np suspension was diluted by a factor of 10 with n-butanol, agitated in an ultrasonic bath (Branson) for 5 min and passed through a 0.2 μm PTFE syringe filter. The diluted suspension was analyzed by dynamic light scattering (DLS) under backscattering conditions using a Malvern Zetasizer. Φ was measured using the Kelvin probe technique, and oxidation state of the Mo cation was determined using X-ray photoelectron spectroscopy (XPS). To form thin films for these measurements, the np suspension without dilution was passed through a 0.2 μm PTFE syringe filter and spin coated at 1000 rpm on a cleaned ITO coated glass (20 Ω/sq, thin film devices). Φ of the np film was measured in air using an isoprobe electrostatic voltmeter (model 244, Monroe Electronics), with Au as the reference material (Φ = 5.1 eV). XPS of the np film was carried out using a Perkin-Elmer 5600 ESCA system with monochromated Al KR source (1486.7 eV). All spectra were collected at an angle of 45° to the sample normal, with a pass energy of 58.7 eV and energy step of 0.125 eV. All XPS spectra were fitted using commercial software (MultiPak, PHI) and aligned to the C 1s reference at 284.8 eV. A reference sample of 20 nm MoO3 on ITO was made by thermal evaporation at a rate of 0.1 A/s. A FWHM of 1.2 eV for each peak was used for peak fitting.
3. Results and Discussions
We found that microwave heating of MoAcAc resulted in a brown suspension of np (Figure 1). Even without added ligands, the suspension showed excellent stability and remained dispersed for more than 90 days. DLS of the as-synthesized np suspension found a volume weighted mean diameter of nm (Figure 2(a), dashed line), in good agreement with the size determined using small-angle X-ray scattering (SAXS) [18]. Storage of the as-synthesized np in air for up to 120 hrs led to no change in size, as confirmed by DLS (data not shown). In contrast, we found that chemical oxidation with H2O2 modified the size of the resulting nanoparticles in a complex way (Figure 2(a)). For example, 0.1 M (Figure 2(a), blue) and 0.3 M (Figure 2(a), red) H2O2 at short reaction times ranging from 20 min to 3 hr decreased the np diameter to ~1 nm. In comparison, a further increase in reaction time to 24 hr caused the nanoparticle size to increase back to 2 nm. The increase in size with longer reaction time is consistent with our previous observation that the average diameter of chemically oxidized np from SAXS had increased to 4 nm after ~15 days of chemical oxidation [18] and indicates that Ostwald ripening for these nanoparticles only occurs with the addition of H2O2. Nevertheless, the nanoparticles are still small enough to form multilayer films with thickness of ~10 nm. Indeed, we previously showed using atomic force microscopy that a chemically oxidized suspension spin coated on top of P3HT:PCBM formed a pinhole-free film that planarized the roughness of the underlying P3HT:PCBM layer [18]. Thus, while np size increases after chemical oxidation with H2O2, it remains sufficiently small to form uniform films on top of organic layers at thickness values that are relevant for HTL in OPV devices.
(a)
(b)
Chemical oxidation of the nanoparticles significantly altered the electronic properties of np films, as measured by their Φ values in air. The as-synthesized np exhibited a low Φ of eV. Addition of 0.05 M H2O2 for 20 min caused Φ to increase to eV, and longer reaction times did not significantly increase Φ (Figure 2(b), purple). Increasing H2O2 concentration to 0.1 M and above caused Φ to increase to eV after 1 hr, and the value remained unchanged when increasing the reaction time to 3 hr (Figure 2(b)). However, a further increase in reaction time to 24 hr caused Φ to decrease to eV. We note that the decrease in Φ corresponded to the change of the nanoparticle suspension color from brown to blue (Figure 1).
To quantitatively understand the origin of the change in Φ with chemical oxidation, we examined the stoichiometry of np films with different H2O2 reaction conditions using XPS. As a reference, we measured evaporated MoO3 and found that the Mo 3d XPS spectra can be fitted to doublet peaks at 232.3 eV and 235.4 eV, corresponding to Mo 3d5/2 and Mo 3d3/2 peaks for Mo6+, plus very weak doublet peaks at 231.0 eV and 233.8 eV corresponding to Mo 3d5/2 and Mo 3d3/2 peaks for Mo5+ [18, 20]. The Mo6+ fraction, defined as the area under the Mo6+ peaks divided by the total peak area, was 0.95, indicating that the reference sample was almost pure MoO3. The FWHM of the fitted peaks was 1.2 eV, and this value was used for all peaks when fitting XPS spectra of np films where there are significant contributions to the overall signal from multiple oxidation states, in order to quantify the atomic fraction from each oxidation state. For example, the Mo 3d XPS spectra of as-synthesized np contain peaks at 232.4 eV and 235.5 eV from Mo6+, 231.4 eV and 234.5 eV from Mo5+, and 229.8 eV and 232.9 eV from Mo4+ (Figure 3(a)). The Mo6+ fraction was 0.21, showing that as-synthesized np is mostly reduced. Reaction with 0.05 M H2O2 for 1 hr caused the Mo6+ peaks to increase in intensity at the expense of the Mo5+ and Mo4+ peaks, (Figure 3(b)), so that the Mo6+ fraction increased to 0.60. Nevertheless, the XPS spectra clearly show that 0.05 M H2O2 only partially oxidized the as-synthesized nanoparticles to MoO3. Increasing the H2O2 concentration to 0.1 M resulted in Mo 3d XPS spectra consisting almost entirely of peaks at 232.6 eV and 235.7 eV from Mo6+, with a miniscule contribution from Mo5+ peaks at 230.9 eV and 234.1 eV (Figure 3(c)). The corresponding Mo6+ fraction of 0.97 indicates that 0.1 M H2O2 for 1 hr oxidized np to the extent similar to the evaporated MoO3 film. Increasing the reaction time to 24 hr at 0.1 M H2O2 caused the Mo5+ and Mo4+ peaks to reappear in the Mo 3d XPS spectra (Figure 3(d)), decreasing the Mo6+ fraction to 0.55. We believe the partial reduction of np with increased H2O2 reaction time can be explained by the electrochromism of MoO3. MoO3 is known to change to a blue coloration upon the insertion of small cations such as H+, following the reaction [21] In our case, H+ is supplied by the H2O in the 30 wt% H2O2 solution, and e- may be supplied by oxidation of neighboring Mo4+ and Mo5+ atoms. This reaction scheme is consistent with the observation that the change in np coloration from brown to blue was only observed after 24 hr reaction with 0.1 M and 0.3 M H2O2. The onset of color change, and by reference the proton insertion, also coincides with the increase in nanoparticle size and reduction in Φ. Thus, analysis of XPS spectra reveals that chemical oxidation with H2O2 consists of two concurrent processes, a fast oxidation that is complete at 0.1 M and higher H2O2 concentrations within ~1 hr, and a slow reduction caused by H+ insertion that occurs after ~24 hr, highlighting the importance of the chemical oxidation conditions on the stoichiometry of the resulting np.
(a)
(b)
(c)
(d)
Figure 4 depicts Φ and Mo6+ fraction that can be achieved using the in situ H2O2 chemical oxidation approach. For np films that were chemically oxidized at different H2O2 concentrations for 1 hr (Figure 4, circles) and at 0.1 M H2O2 for 24 hr (Figure 4, triangle), the data show a strong correlation between the two parameters, suggesting that a clear relationship exists between nanoparticle stoichiometry and electronic properties. A continuous tuning of the Mo6+ fraction between 0.2 and 1.0, and Φ from 4.4 eV to 5.0 eV was achieved. Such control should allow us to systematically study the effect of composition on device performance in various applications.
4. Conclusions
In conclusion, through a systematic study of the effect of H2O2 concentration and reaction time on nanoparticle size, work function, and stoichiometry, we demonstrate that in situ chemical oxidation of nanoparticle suspensions with H2O2 is a simple but versatile method to control their stoichiometry and electronic properties. Starting from suspensions of ultrasmall ( nm) nanoparticles in n-butanol synthesized by a one-step microwave heating procedure, we found that short time (≤3 hr) reactions at room temperature with sufficiently high (≥0.1 M) concentration of H2O2 result in ≤1 nm nanoparticles with high work function and almost entirely MoO3 properties which are desirable for HTL material in OPV devices. However, long chemical oxidation times (≥24 hr) increase nanoparticle size and reduce Mo oxidation state and film’s work function. By comparing Kelvin probe results and XPS spectra of np films that have been chemically oxidized under various conditions, we established a clear correlation between the work function and the Mo6+ fraction of the np and achieved a continuous tuning of work function values from 4.4 to 5.0 eV and Mo6+ fraction from 0.2 to 1.0. Such precise control of stoichiometry and properties is crucial for the optimization of np as a solution processible material for various applications. Moreover, the simplicity of the chemical oxidation procedure should be generally applicable in synthesizing other transition oxide nanomaterials with tunable stoichiometry and properties.
Acknowledgments
This project is supported by the University of Texas at Dallas. D. Barrera acknowledges financial support from Consejo nacional de Ciencia y Tecnología and Centro de Investigación en Materiales Avanzados, Unidad Monterrey. J. W. P. Hsu acknowledges the support of Texas Instrument Distinguished Chair in nanoelectronics.