Metal Nanoparticle Enhanced Organic Solar Cell and Its True Efficiency
Il Gu Han*
Fall 2009
University of California, Merced
Organic solar cell is an attractive substitute for silicon based solar cells due to its cheap price and higher coverage, yet low efficiency is one of the problems. There have been claims that the presence of metal nanoparticles(NP) such as gold or silver, can enhance the efficiency of organic solar cells, but it is not generally accepted. In this research, A simpler form of an organic solar cell with metal NP was used to see if enhancement of efficiency is true or not. Unlike the previous design, dip coating was used for metal nanoparticle layer and PEDOT:PSS layer was removed for convenience. Rest of research procedures followed the previous design. At the end of the research, we could conclude that there is no enhancement of the organic solar cell with metal nanoparticle. The result was reliable since the major source of error, amount of nanoparticle applied and the even distribution over the surface could be controlled by using dip coating technique.
Introduction
Photovoltanic cells based on organic semiconductor have been attractive substitute for silicon based solar cells due to its cheap price and large coverage. However, its low efficiency is a disadvantage over silicon solar cells. There have been claims that the presence of metal NP such as gold or silver can enhance the efficiency of organic solar cells. The idea is that the presence of NP amplifies the electromagnetic field of the light by surface plasmon resonance. However, it is not generally accepted as its exact mechanism is not clearly explained yet. Thus, in this research, we will see if the presence of NP can actually enhance the efficiency of organic solar cell especially on its absorption and fluorescence by using a simple form of organic solar cell model.2
In the first part of the research, the metal NPs was applied with drop coating technique. However, due to uncertainty over the amount and distribution of the NPs on the slide, the result was inconclusive.
In order to resolve those major uncertainties, drop coating technique was substituted with dip coating technique. The dip coating technique that was proposed in one of the literatures is based on the following theory:1
By modifying the slides with 3-mercaptopropyltrimethoxysilane (MPTMS) sol-gel which acts linker between the slide and the Ag NPs, the layer will self-assemble as we are dipping it in the Ag NPs solution. The thiolexchange reactions and layer-by-layer deposition through electrostatic interactions are the major factors of Ag immobilization on to the slides. The general pictorial representation is shown on Figure 1.
Figure 1. The representation of chemical interaction during the process of dip coating. Each step is explained in detail in the methods section.1
Originally, this method was proposed to prepare a substrate for SERS. However, since it provides the condition of what we needed, there was no significant modification to it and just followed their procedure in this research for the Ag NPs coating part.
Methodsf
- Chemical preparation
In this part of then experiment, only Ag NPs solution was used despite the fact that gold was used with the previous method.
9 mg of AgNO3 was dissolved in 50mL of hot water around at 40°C. While stirring it with fairly high speed, 2 mL of 2% sodium citrate solution was added rapidly. The solution was refluxed for about 60~90 minutes near at 100°C.2
3-mercaptopropyl-trimethoxysilane(MPTMS) sol-gel was prepared by adding 600µL of MPTMS and 500 µL of 0.1 HCl into 50 mL of water. The solution was stirred for at least 1 hour. As it is instructed on the literature, the sol-gel solution was used only on the same day it was made.1
MPTMS methanolic solution, 92.6 µL of MPTMS was added to 50 mL of methanol in order to have 10 mM. Calculation is shown on the supplementary data section.1
The organic polymer, nile red was prepared by dissolving nile red in toluene and mix it with polystyrene solution which also use toluene as solvent.3 The concentration of the nile red : polystyrene (NR:PS) used was calculated to have its absorption to be around 0.1 when it is spin coated onto a slide. Calculation will be added on the supplementary data section.2
- Slide preparation
In each run, a slide holder that is made out of a cardboard cut in a circular shape that fits a beaker used. The slide holder could hold up to 5 slides (Microscope Cover Glass, Fisherbrand) at a time so far, but in this report, we are considering the run with only 4 slides.
Before the actual dip coating, the slides were dipped into Piranha solution vertically by using the slide holder. The ratio between sulfuric acid and H2O2 which is the two components of piranha solution was about 3:1 by volume.6 Then, those were rinsed with excess amount of water and dried with air flow. In the literature, nitrogen flow was used, but considering that it wouldn’t influence the data too much, air was used for convenience. The dried slides were rinsed with methanol and immersed into 10 mM MPTMS methanolic solution for 24 hours to complete the modification.1
- Layering the slides
Before the actual dip coating part, the modified slides were taken out of the MPTMS methanolic solution, thoroughly washed with methanol, dried with air flow, rinsed with water, and then it was immersed into Ag NPs solution for another 24 hours.1
For each additional Ag NPs layer, the slides were removed from the Ag NPs solution, rinsed with water, dipped into the MPTMS sol-gel for 20 minutes, and immersed into the Ag NPs solution for 1 hour. The layers were added until desired absorbance is achieved. When the layering is finished, it was washed with toluene in order to prevent aggregation of Ag NPs due to the interaction between methanol and toluene in NR:PS. In this experiment, there were 3 layers of Ag PNs were applied.1
When Ag NPs solution dip coating is done, NR:PS solution was spin coated onto the PEDOT:PSS layer at 2000 rpm for about 30 seconds.3 PEDOT:PSS layer was skipped with this procedure since PEDOT:PSS layer was not stably spin coated onto the Ag NPs layer and just washed away. The significance of PEDOT:PSS was also not as highly expected.
Each time the layer is added, the absorption of the slide was taken by 50 scan UV-Visible spectrophotometer. It was taken over 750nm to 350nm, ave time at 0.1, data interval at 1.0, and scan rate at 600.00 nm/min. The absorption of the NR:PS was calculated by taking the difference of the absorption of Ag + NR:PS layer and Ag layer only. By comparing the calculated absorption of NR:PS from a set with Ag layer and the one without, the enhancement of metal NP should be able to be observed. 2
The same procedure should be done with fluorescence, but it was not done because absorption alone gave a good sense of the result, and the fluorimeter was not easy to access to in the lab.
Result
Although dip coating technique was expected to have higher resemblance in absorbance between each layer compared to drop coating technique, the result didn’t show it much as shown on Graph 1.
Graph 1 Ag NPs layers absorbance comparison. Total 3 layers of Ag NPs are applied. All the slides were coated in a same beaker with the same solution at the same time. During first and second dip coating, the shape of the absorbance did not show a clear peak at ~450 nm. The example of absorbance shape change will be added on the supplementary data section. By comparing with the graph from the literature and the pattern from our graphs, it is expected to have clearer shape of Ag NPs absorbance as we add more layers.
Although we did not have the same absorbance, each spectrum was absorbing at the right wavelength, and the surface of the coated slide seemed to be very even. Since having smooth and even surface of Ag NPs was more major concern than having exact same absorbance, these slides were considered to be acceptable for the further procedure.
The NP:PS solution we used had the absorbance ~0.3, which is three times higher than what I expected to have. However, it showed a good shape and absorbed at the right wavelength. So this solution was also accepted for the further procedure. The result of having too high absorbance will be discussed in the discussion section.
Graph 2 A absorption of NR:PS only. Although it showed higher absorbance than expected, it did show a good shape of nile red absorbance and it also absorbed at the right wavelength which is near 515 nm.
When NP:PS was spin coated on to the Ag NPs layer, it showed much higher degree of resemblance of absorbance between each layer compared to when drop coating was used for Ag NPs.
Graph 3 Ag NPs layers + NR:PS absorbance comparison. Their absorbance are closer to each other than when drop coating was used. At 515 nm, Their standard deviation of the absorbance was 0.04065. If we consider the slide #2(yellow) as an outlier, we have even smaller standard deviation.
The enhancement of the NRPS was calculated by following calculation:
“Ag NPs + NR:PS” – “NR:PS only” = “Ag NPs + NR:PS enhancement”
“Ag NPs + NR:PS enhancement” – “Ag NPs only” = NR:PS enhancement
The result is shown on the Graph 4.
Intermediate trials with no significant result are not included in this report.
Graph 4 Calculated NR:PS absorbance comparison. Except the slide #1, all the slides showed rather degradation of absorbance. Even with slide #1, it showed the weakest enhancement at the wavelength where nile red absorbs.
Discussion
Although the degree of degradation shown on Graph 4 were different from each slide, it is easy to conclude that having Ag NPs on the slide did not cause enhancement on NP:PS absorbance in general.
By using dip coating technique instead of drop coating, we could eliminate the major source of error that made our result inconclusive last time; amount of chemical applied on the slide and the even distribution of the particles. With drop coating technique, it was not possible to control those factors by hand. Thus, the result with drop coating was not reliable. However, by using dip coating technique, controlling over those factors is possible. Since the slides are modified to bind the Ag NPs chemically, we can expect each slide will have about the same amount of Ag NPs applied at a time. Also, the Ag NPs will be bind throughout the surface evenly. Even if it does not bind evenly, the difference in absorbance over the surface wouldn’t be as big as when drop coating is used.
However, Using a UV-Vis spectrophotometer is still a considerable source of error. 50 Scan UV-Vis spectrophotometer was designed to work with regular 1 cm cuvettes. Because of that, the position of the cover slide was not accurately controlled to have the beam of spectrophotometer to go through the right place every time. According to the result, the reproducibility of a single slide seems to be acceptable as it is shown in graph 5.
Graph 5 Multiple measurements of a bsorbance stectra of the same Ag NPs on a slide. (3 layers applied) The standard deviation of the absorbance at ~457nm is 0.003377, which is less than 10 times smaller than the absorbance. Thus, the reproducibility is acceptable.
Despite of removing some of the critical source of errors, there are some other factors that need to be considered. In this experiment, nile red concentration was 3 times higher than of Ag NPs. Considering that the Ag NPs is coated on both side of the slides, only half of the Ag NPs was able to influence the nile red absorption. This makes nile red to be 6 times more concentrated than the Ag NPs. This big difference in concentration or absorbance might weaken the surface plasmon resonance and did not show proper enhancement. Also, not having PEDOT:PSS could be another considerable factor. Although in theory, having a layer of PEDOT:PSS between the Ag NPs and nile red does not support the idea of enhancing surface plasmon resonance, it is still something to consider and test it with the actual experiment.
In order to confirm the result, we may design an experiment with the same design, but with more slides to compare, higher absorbance for Ag NPs, and addition of PEDOT:PSS condition to compare. If it still does not show significant enhancement at ~515 nm, we may conclude that applying nanoparticles on a organic solar cell does not enhance its efficiency.
Acknowledgements
Thanks to Professor Anne Myer Kelly for supervise during the whole research process. Also, thanks to Miah, Marina and Chenfei for assisting and advising me with general use of lab equipments and chemicals.
Reference
[1] Fan M. Brolo A. G. Silver nanoparticles self assembly as SERS substrates with near single molecule detection limit DOI: 10.139/b904744a
[2] Kelley, A. “Han background” An instructional paper with this research’s background. 2009
[3] Kim H, Nam S, Tae S, Yoon M, Laser scanning confocal microscope (LSCM)-fluorescence spectral properties of Nile Red embedded in polystyrene film of different thickness, In Chemical Physics Letters; 432 (2006) 200-204
[4] Lee and Meisel, J. Phys. Chem. 86. 3391 (1982)
[5] Safety data of polystyrene. <http://msds.chem.ox.ac.uk/PO/polystyrene.html. (6-22-09). under “physical data” section>
[6] Piranha solution, Wikipedia, < http://en.wikipedia.org/wiki/Piranha_solution>
Supplementary Materials
- General contents of intermediate trials
- The concentration of NR:PS was calculated as following:
- Expected absorption: 0.1
- Estimated thickness of the layer: 200nm
- Concentration of Polystyrene in toluene: 47.8 mg/mL
- In 5 mL of PS solution: 239 mg of PS
- Density of PS: 1050 Kg/m3 [4]
- Volume of 239 mg of PS:
- For 5 mL of PS solution, amount of Nile red needed:
- A = ɛlc
- 0.1 = (3.8 x 104 M-1cm-1)(2.00 x 10-5 cm)(x)
- x = 0.1315 M (concentration of NR in NR:PS spin-coated layer)
- =0.1315 M
- Mol of NR = 2.99 x
- (2.99 x 10-5 mol) x (318.369 ) = 0.00953 g or 9.53 mg in 5 mL of NR solution
- Concentration of NR solution: 1.912
- The concentration of MPTMS methanolic solution calculation:
- Expected concentration: 10 mM
- Molecular mass of MPTMS: 196.33994 D
- Solvent used: methanol
- Density of MPTMS: 1.6 g/mL
- Mole of MPTMS needed: 0.01 M = x/0.05 L x = 0.0005 mol
- Mass: 0.0005 mol x 196.33994 D = 0.09817 g
- Volume: 0.09817 g / 1.06 g/ml = 0.0926 mL = 92.6 µL for 50 mL of methanol.
