Solution Processed Silver Nanoparticles in Dye-Sensitized Solar Cells

A plasmonic effect of silver nanoparticles (Ag NPs) in dye-sensitized solar cells (DSSCs) is studied. The solutions of silver nitrate in isopropanol, ethylene glycol, or in TiO 2 sol were examined as possible precursors for Ag NPs formation.The solutions were dipcoated on the top of the porous TiO 2 layer.The results of optical measurements confirmed the formation of Ag NPs throughout the porous TiO 2 layer after the heat treatment of the layers above 100C. Heat treatment at 220C was found to be optimal regarding the formation of the Ag NPs.The porous TiO 2 layers with Ag NPs have been evaluated also in DSSC by measuring current-voltage characteristics and the external quantum efficiency of the cells. In addition, the amount of adsorbed dye has been determined to prove the plasmonic effect in the cells. The I-V characterization of the DSSCs revealed an increase of the short circuit current in the presence of Ag NPs although the amount of the attached dye molecules decreased. These results confirm that the performance enhancement is related to the plasmonic effect. However, neither a thin sol-gel TiO 2 layer nor poly(4-vinylpyridine) shells provide effective protection for the long term stability of the Ag NPs against the corrosion of I 3 − /I based electrolyte.


Introduction
In the last decade the processes and new materials for dye-sensitized solar cells (DSSCs) have been extensively researched in order to enhance the performance and the stability of the cells [1][2][3][4].In order to broaden and/or increase the spectral sensitivity of the solar cell several concepts have been considered, for example, the quest for new sensitizers [5][6][7] or cosensitizers [8][9][10][11], different approaches for engineering active layer/electrolyte interface [12], optimizing the thickness of the active layer, or increasing the concentration of the attached dye molecules [13][14][15].Absorption has also been improved by increasing the optical path using larger scattering particles or voids in the active layer [14,[16][17][18].Furthermore, an undyed layer placed on top of the active layer [13,14,19,20] or a mirror-like Pt counter electrode [21][22][23][24][25][26][27][28][29][30] was studied as an internal back reflector.In addition, different materials have also been proposed as external back reflectors placed at the rear side of the DSSCs [31,32].
In recent years a surface plasmon resonance using metal nanoparticles (NPs) has also attracted much attention and has been studied for various solar cell technologies [33][34][35][36][37].The surface plasmon resonance is a light induced coherent oscillation of conduction band electrons on the surface of the metal NPs.The wavelength of resonance peak due to the presence of plasmonic particles greatly depends on factors such as size, shape, and dielectric environment.Two types of enhanced absorption are possible: (1) near-field: a strong electromagnetic field in the vicinity of the NPs around the plasmonic resonance wavelengths greatly enhances the absorption in the active layer and (2) far-field enhancement of the light absorption in the photoactive layer due to the scattering effect.
The proof of the concept was firstly presented on two simplified flat structures.The first one contained dyed Ag NPs on compact TiO 2 layer placed over the front transparent conductive oxide (TCO) [38,41].In the studies, a less harmful Co 2+/3+ based redox electrolyte was used in order to preserve unprotected Ag NPs.The second structure contained Ag NPs deposited on the front TCO that were protected with a thin compact TiO 2 layer, varying in thickness between 2 and 9 nm, covered with a monolayer of dye molecules [39,40].The results confirmed that the electromagnetic field related to the near-field plasmonic effect decreases with a power of two by increasing the thickness of the TiO 2 shell around NPs, that is, by increasing the distance from the dye molecules and NPs [39,40].If the shell was thicker than 2 nm the plasmonic effect was hindered.On the other hand, the shell should protect the NPs from aggressive electrolyte; therefore the shell should not be too thin either [43,55,57,58].The study showed that to fully protect the Ag NPs at least 7.7 nm thick compact crystalline TiO 2 layer is needed [39].The near-field plasmonic enhancement in DSSC was confirmed but the performance of the cells was very low (>0.05%)due to the very thin active layers used in this study.
So far several different locations of NPs and deposition techniques have been proposed to prepare DSSC with plasmonic enhancement.
(i) Location 1: NPs deposited on the front TCO by soaking the front TCO in suspensions containing NPs [49] or by sputtering the NPs [44].
(ii) Location 2: the NPs are formed throughout porous TiO 2 layer.
(iii) Location 3: NPs located at a certain position within the porous TiO 2 layer.A thin metal layer is sputtered on the porous TiO 2 layer which created NPs after heat treatment.Afterwards, additional porous TiO 2 overlayers have been deposited [46].
(iv) Location 4: NPs located on the top of the porous TiO 2 .
A thin metal layer is sputtered at the back side of the dyed/undyed TiO 2 layer which created NPs at the rear side after heat treatment [44,46,47].
The researchers confirmed a significant effect of the NPs location, concentration, and size on the performance enhancement.Therefore, they have to be carefully optimized according to the DSSCs structures (thickness of the active layer, electrolyte, dye, etc.) in order to achieve the highest plasmonic enhancement.Generally, the highest enhancement is found when the NPs are spread throughout the active layer [42,53].Furthermore, larger NPs shift the plasmonic resonance toward longer wavelengths, but the magnitude of the shift strongly depends on the materials used for NPs and the dielectric environment (surrounding material), for example, the resonance peak of the Au NPs in a suspension shifts from 519 nm to 579 nm, when the diameter of the NPs increases from 20 to 80 nm [49].On the other hand, a parasitic absorption within NPs could significantly hamper the plasmonic effect [45,59] if the NPs exceed optimal size or agglomeration takes place.The simulations showed that the absorption of metal nanoparticles could take up about 20-30% of the total absorption of DSSC at visible wavelengths even under the optimized conditions [59].
Some researchers studied the structures containing unprotected NPs [44,46,47,49,54,56,58,59] but the long term stability remains questionable, since the I 3 − /I − based electrolyte is very aggressive even to Au NPs [39,52].Additionally, the unprotected NPs could act as effective recombination centres [55]; therefore many research focused also on core/shell or even core/shell/shell NPs with Ag or Au cores and TiO 2 [42,43,45,48,50,53], SiO 2 [50,53,55], or organic shells [51,52].In addition to careful design regarding the location, concentration, and size of the NPs it has been stressed on that the shell should be pinhole free to ensure the protection against electrolyte, but on the other hand the shell should be as thin as possible to utilize the plasmonic effect [39,40,45,52,57] and to reduce parasitic absorption within a shell.A red shift in plasmonic resonance by increasing the thickness of the shell was also observed [39].In addition to the plasmonic effect the NPs present in the cell structure could induce different effects influencing the solar cell parameters.The NPs with the TiO 2 shell can upward shift the Fermi level of the TiO 2 , reduce the dark current, and increase the electron lifetime which all increase the  OC [43,53].Additionally, the NPs can simultaneously increase the internal surface available for dye loading which additionally increases the  SC [45].Therefore, the performance enhancement due to the plasmonic effect should be always very carefully examined.
Many researchers significantly improved the conversion efficiency when NPs were incorporated (from 30% up to 140%, relatively) [47,50,52,55,60], but high relative improvements were generally not obtained for the solar cells achieving state-of-the-art efficiency.Nevertheless, some researchers found remarkable improvements also in absolute terms.Dang et al. [42] found an improvement of the performance from 8.3 to 10.8% when TiO 2 /Au/TiO 2 core/shell/shell NPs were included throughout the active layer of the DSSC.Furthermore, Choi et al. [53] increased the efficiency of DSSC from 9.3% to 10.2% after incorporating Au/SiO 2 core/shell NPs, while the  increased only to 9.8% when the TiO 2 shell was applied.
Herein, the DSSCs containing Ag NPs throughout the active layer are examined in Figure 2. Ethylene glycol and isopropanol were used to prepare silver nitrate solutions that have been used for the impregnation of the porous TiO 2 layer.The formation of Ag NPs formed at five different temperatures was examined using optical measurements of the samples deposited on microscope glass.The optimal temperature determined to be 220 ∘ C was used for the preparation of the Ag NPs within porous TiO 2 layers for DSSCs.The prepared Ag ISO NPs were afterwards protected with either a thin TiO 2 layer or organic shell based on poly(4vinylpyridine), PVP.In addition, the feasibility of the formation of Ag/TiO 2 core/shell NPs in a single step was studied by incorporation of the silver nitrate to the Ti sol.
The porous TiO 2 layers with Ag NPs have been tested in the DSSC.To evaluate the plasmonic enhancement external quantum efficiency (EQE), current-voltage characteristics (I-V) at standard test conditions and the amount of the adsorbed dye were measured and compared to our reference DSSCs without NPs.Finally, a stability of the Ag NPs against electrolyte in DSSCs is presented by remeasuring the I-V characteristics of DSSCs three months after manufacturing and by exposing the Ag ISO /TiO 2 and Ag ISO /PVP core shell NPs to I 3 − /I − redox electrolyte and measuring the difference in absorption.

Preparation of the Active
Layer.An optimized Pechini sol-gel TiO 2 paste (based on P25, Degussa) [61] was applied to the glass substrate using a "doctor blading" technique.For optical measurements a microscope glass was used as substrate while a fluorine-doped SnO 2 coated glass (TCO) with a sheet resistance of 8 Ω/square was used for the DSSCs' substrate.After deposition, the TiO 2 layers were sintered at 450 ∘ C for 1 hour.
Three different solutions containing Ag NPs precursors were studied.Solution 1 (Ag ISO ) was 0.004 M AgNO 3 in isopropanol, solution 2 (Ag EG ) was 0.004 M AgNO 3 in ethylene glycol, and solution 3 (Ag/TiO 2 ) was Ag ISO mixed with TiO 2 sol in a molar ration 0.4 M/0.004 M. The TiO 2 sol was prepared by mixing the Titanium (IV) isopropoxide (Aldrich) with acetyl acetone (Fluka) in molar ratio 1 : 1.5  in isopropanol to obtain the sol with 0.4 M of titanium.To deposit the Ag NPs on porous TiO 2 layers the dip-coating method was used (speed 10 cm/min).The layers were heattreated for 1 h at different temperatures from 100 to 300 ∘ C.After heat treatment the bare Ag NPs (Ag ISO and Ag EG ) or Ag/ TiO 2 core/shell NPs within a porous TiO 2 layer were formed (Ag/TiO 2 ).
For the stability tests (Section 3.3) the bare Ag ISO NPs were coated with either a thin TiO 2 or PVP shell.To make a TiO 2 shell the Ag-TiO 2 layers were additionally dipcoated using the TiO 2 sol (speed 10 cm/min) and heat treated at 220 ∘ C for 1 h.To form a PVP shell around Ag ISO NPs we soaked the Ag-TiO 2 layers for 24 h into a solution containing 0.005 M poly(4vinylpyridine) in acetonitrile.

DSSC Preparation.
The TiO 2 layers without/with Ag NPs have been immersed in an ethanol solution of a Ruthenium complex based dye (Ru(2,2  bipyridyl-4,4  dicarboxylate) 2 (NCS) 2 , Solaronix) for 12 hours.For the counter electrode, platinum (thickness ∼ 5 nm) was sputtered onto a TCO glass substrate.Both electrodes were then sealed using a 25 m thick polymer foil (Surlyn, DuPont) which also acts as a spacer between the electrodes.Electrolyte was injected through two predrilled holes in the counter electrode.The electrolyte was a binary ionic liquid mixture in a volume ratio 13 : 7 of 1-propyl-3-methyl-imidazolium iodide (Iolitec) mixed with 1-ethyl-3-methyl-imidazolium tetracyanoborate (Merck), 0.5 M methyl benzimidazol, and 0.1 M guanidinium thiocyanate.The concentration of I 2 was 0.2 M. Three identical DSSCs were assembled for each material combination, each with an active area ranging from 0.58 to 0.72 cm 2 .Before characterization the cells were stored in the dark for 24 hours to allow the electrolyte to penetrate into the TiO 2 pores.

Characterization.
The - characterization was performed using an Oriel Class ABA solar simulator equipped with a 1.5 G air mass filter, the spectrum of which closely matches the required AM1.5 spectrum.In accordance with the IEC 60904-3 standard the short-circuit current mismatch parameter was calculated in conjunction with a calibrated c-Si reference solar cell, covered with a KG5 glass filter; the level of standard illumination (1 sun, 100 mW/m 2 ) was determined.The cells were also masked to leave only the active area of the cell exposed, which is stated as the most rigorous condition regarding the cell's efficiency.- characteristics of the cells were then measured using a Keithley 238 source meter by applying a voltage and measuring the current.The cells were scanned stepwise (10 mV) from 0 V (short circuit regime) to 0.9 V (beyond  OC ).
The UV-vis optical measurements (total transmittance  TOT and reflectance  TOT ) were performed on a Lambda 950 Perkin Elmer spectrophotometer equipped with a 150 mm Ulbricht integration sphere coated with Spectralon.The spectra were recorded from 300 to 800 nm in decrements of 5 nm.Absorption was calculated accordingly to the  = 100% −  TOT −  TOT .
External quantum efficiencies (EQEs) were measured without bias light using an Xe lamp and a monochromator (70525 Arc Lamp Apex Illuminator and Cornerstone 260 Monochromator, Oriel, Newport) at increments of 5 nm from 300 to 800 nm.The monochromatic light was focused so that the beam was smaller than the active areas of both the reference photodetector and the DSSCs.A stable reading was obtained after a 3 s delay between setting the wavelength and taking a current measurement.
The amount of the dye adsorbed on the TiO 2 layer was determined with UV-vis spectroscopy.The Ruthenium dye was firstly detached from the TiO 2 layers using 0.02 M NaOH (Merck); the dye solution has been diluted with 0.02 M NaOH to 50 mL before obtaining UV-vis spectra.In order to determine the amount of dye adsorbed on the TiO 2 layers a calibration curve was made.For the calibration curve and the evaluation of the samples concentration the absorbance peak of the dye at the 500 nm has been considered.

Results and Discussion
Herein we present the results of the very simplified integration of the Ag NPs within the porous TiO 2 layer.The Ag NPs within the TiO 2 layers were prepared using three different solutions, that is, Ag ISO , Ag EG , and Ag/TiO 2 (see Section 2).In Section 3.1 the influence of the heat treatment temperature on the formation of Ag NPs by measuring the absorption spectra is presented.The TiO 2 layers with Ag NPs were tested also in the DSSCs and the results were compared with the cells without Ag NPs.We have evaluated the DSSCs by measuring the I-V characteristic, EQEs, and the amounts of the adsorbed dye (Section 3.2).Section 3.3 deals with the stability of the DSSCs containing Ag NPs.The performance of the DSSCs has been remeasured three months after the manufacturing.
Additionally, the degradation of the TiO 2 layers containing Ag ISO /TiO 2 or Ag ISO /PVP core/shell NPs due to the exposure to the I 3 − /I − electrolyte is presented.The Ag ISO NPs within the porous TiO 2 layer are formed already at 100 ∘ C visible as a homogeneous dark red layer, shown in Figure 3(a).The absorption enhancement (compared to the reference TiO 2 layer without NPs) is noticeable above 365 nm reaching the enhancement peak (Δ =  Ag −   ) around 410 nm.A further increase in the heat treatment temperature up to 220 ∘ C gradually increases the absorption without changing the shape of the curve, while a further increase in the temperature to 300 ∘ C noticeably decreases the absorption and changes the shape of the curve, that is, the layers become grey.In a further study the heat treatment at 220 ∘ C was used for the preparation of a TiO 2 layer with Ag ISO NPs.The optimal heat treatment temperature where the enhancement is most pronounced is 220 ∘ C.
Similar absorption enhancement and optimal heat treatment temperature at 220 ∘ C was found for Ag EG NPs (without shell), as shown in Figure 3(b).This is in accordance with our expectation since only the ethylene glycol instead of isopropanol was used in solution 2. In both cases the organic components decompose and reduce the Ag + to Ag 0 while leaving the Ag NPs on the surface of the TiO 2 .However, the Ag EG samples are less homogeneous which is probably related to the different surface tension and viscosity of the ethylene glycol.
The absorption of the Ag/TiO 2 core/shell NPs are presented in Figure 3(c).In this case the plasmonic enhancement cannot be compared with the bare TiO 2 layer on a microscope glass substrate (black full line in Figure 3(c)) since solution 3 (which includes the TiO 2 sol) is expected to form not only Ag/TiO 2 core/shell NPs but also a thin overlayer on the porous TiO 2 layer.Therefore, a new reference having a thin "shell" TiO 2 overlayer has been used; that is, the porous TiO 2 layer has been dip-coated into a TiO 2 sol (speed 10 cm/min) and heat treated afterwards at 220 ∘ C for 1 hour.For Ag/TiO 2 NPs the colour of the samples was greyish (Figure 3(c)) showing significantly smaller absorption enhancement.The absorption enhancement can be only found at 220 ∘ C while heat treatment at lower or higher temperatures shows even deterioration in absorbance compared to the reference.The minor enhancement might be observed if each studied heat treatment temperature would have its own reference produced at the same temperature.Nevertheless, the 220 ∘ C seem to be the optimal heat treatment also for the formation of Ag/TiO 2 core/shell NPs.FF, and conversion efficiency ) of three identical cells are gathered in Table 1.

DSSCs with Ag
The DSSC with Ag ISO NPs is the only structure where a small increase in  SC was observed compared to the reference structure.The increase can be related to the plasmonic effect since the DSSCs with Ag ISO NPs have a smaller amount of dye molecules attached to the TiO 2 than the reference DSSCs (see Figure 4).
The DSSCs with Ag/TiO 2 have smaller  SC than the reference (which has no additional "shell" TiO 2 overlayer), as shown in Table 1.This is according to our expectation since a thin "shell" TiO 2 overlayer is formed together with the formation of Ag/TiO 2 NPs.It should be mentioned that the formation of a TiO 2 overlayer on the top of the porous TiO 2 layer significantly decreases the dye loading reaching a half value compared to the reference porous layer (Figure 4).Therefore, the performance of the DSSCs with Ag/TiO 2 cannot be directly compared with the reference DSSCs which has no TiO 2 overlayer (ref., Table 1).However, we can see that the  SC of Ag/TiO 2 is only 7% lower than the reference (9.0 mA/cm 2 for the ref.versus 8.4 mA/cm 2 for the Ag/TiO 2 , Table 1) although the amount of the adsorbed dye is halved (Figure 4).This is another indirect proof of the plasmonic effect in the DSSC.The results are well in accordance also with the optical measurements presented in Figure 3(d) showing a higher absorption of a dyed porous TiO 2 layer containing Ag/ TiO 2 core/shell NPs (Figure 3(d), red full line) than a dyed TiO 2 layer covered with a thin "shell" TiO 2 (black dashed line) although the amount of the attached dye molecules is in both cases similar (1.7 × 10 −8 mol/cm 2 for the reference with a "shell" TiO 2 overlayer and 1.6 × 10 −8 mol/cm 2 for the Ag/TiO 2 , Figure 4).
The DSSCs with Ag EG have significantly smaller  SC than the reference although a comparable  SC to DSSC with Ag ISO NPs would be expected due to similar absorption enhancement (Figures 3(a) and 3(b)) and a similar amount of the attached dye (Figure 4).The origin for the decrease is unknown.It might be related to the very inhomogeneous deposition of the Ag NPs due to different surface tension of the ethylene glycol compared to the isopropanol and/or poor attachment of the Ag NPs to the porous TiO 2 layer.This creates a very low concentration of the Ag EG NPs at the edges of the DSSCs and a very high concentration in the centre which increases the parasitic absorption within the NPs and significantly decreases the  SC .
The normalized EQEs of the DSSCs are presented in Figure 6 (the EQEs were normalized to 0.8 at 345 nm).According to the expectation, the EQEs show the same trends as observed for the  SC determined from I-V measurements; that is, the only enhancement is found for DSSCs with Ag ISO while smaller values were observed for DSSCs with Ag EG and Ag/TiO 2 .The increase or decrease in EQEs is observed above 365 nm which is in agreement with the absorption measurements (Figure 3).
The  OC of the DSSCs are gathered in Table 1.The DSSCs with Ag ISO NPs have on average 14 mV higher  OC than the reference DSSC which might be related to the upward shift of the Fermi level induced by the Ag NPs [43,53].The DSSCs with Ag/TiO 2 NPs have 38 mV smaller  OC than the reference.This could be explained by a smaller dye loading which reduces the number of injected electrons in a conduction band of the TiO 2 .Additionally, the smaller dye loading might leave a larger part of the TiO 2 surface exposed to the electrolyte which increases the recombinations.Both effects shift the Fermi level downward and decrease the  OC .A significantly smaller  OC for the DSSCs with Ag EG is difficult to explain but might be due to the poor attachment of Ag NPs that act as strong recombination centres.
The FF is practically unaffected by the NPs; therefore the  is the outcome of the  SC and  OC .The increase in  was found only for the DSSCs with Ag ISO while the DSSCs with Ag EG and Ag/TiO 2 NPs show a lower  than the reference DSSCs.

Stability of Ag NPs.
The DSSCs have been remeasured 3 months after manufacturing, that is, after keeping the cells in the dark under open circuit conditions for 3 months (Table 1).The DSSC with Ag ISO NPs is the only structure where a decrease in  SC was observed (−3%) while the other DSSCs showed an increase.The  SC of the reference DSSC increased for 2% while the DSSCs with Ag EG and Ag/TiO 2 increased for 18% and 9%, respectively.Interestingly, for the  SC of reference, Ag ISO and Ag/TiO 2 DSSCs became practically alike (Figure 5(b)) which might be related to the degradation of   Thanachayanont et al. [58] who also found a decrease in  SC and an increase in  OC and FF during the 10 days stability study.
The changes (especially in  SC ) in remeasured DSSCs suggest that the dye molecules produce a certain degree of protection but they cannot prevent the degradation entirely.Therefore, the stability of the Ag NPs upon exposure to the aggressive I undyed porous TiO 2 layers are presented in Figure 7.The full lines present the Δ after production while the dashed lines present the Δ after exposure of the layer to the electrolyte for 24 hours under ambient conditions.As expected the Δ decrease substantially for the unprotected Ag EG and especially Ag ISO NPs.Surprisingly, a small decrease in Δ of the Ag/TiO 2 core/shell NPs was also observed which reveals that the thin TiO 2 shell does not provide sufficient protection required for long term stability.
Since the Ag/TiO 2 core/shell NPs that were formed in a single step do not provide sufficient protection we used a secondary step to protect the Ag ISO NPs with either a thin sol-gel TiO 2 or PVP shell, which should protect the NPs from the electrolyte and allow the dye attachment [52].The protected NPs were then exposed to the I 3 − /I − electrolyte and the photographs were taken at different time periods, Figure 8.A substantial degradation of the protected Ag NPs was noticed if the sol-gel TiO 2 layer was used.The thin TiO 2 layer prepared via sol-gel route has a certain degree of porosity which allows the penetration of the electrolyte.Moreover, in our study the PVP shell did not protect the Ag ISO NPs although it provided adequate protection for the Au NPs [52].In addition a few different techniques to prepare TiO 2 shells around NPs can be found in the literature.Dang et al. [42] synthesized core shell NPs before being deposited on the porous TiO 2 layer.On the other hand, a compact TiO 2 shell around Ag NPs could be formed afterwards by atomic layer deposition [39,40] or by calcination of the Ag-TiO 2 films refluxed with a Titanium (IV) isoproxide in isopropyl alcohol solution [45].

Conclusions
A plasmonic effect of Ag NP in the dye-sensitized solar cells (DSSCs) has been confirmed.We used three different solutions containing silver nitrate in isopropanol, ethylene glycol, or isopropanol/TiO 2 sol for dip-coating of porous TiO 2 layers.The results confirmed the formation of the Ag NPs without/with shells within a porous TiO 2 layer.The optimal heat treatment temperature of the dip-coated layers, which gave the most intensive plasmonic enhancements, was 220 ∘ C for all three solutions.However, the ethylene glycol solution of silver nitrate gave a very nonhomogeneous distribution of Ag NPs over the TiO 2 layers.On the other hand, the results show that the formation of the Ag NPs from the isopropanol solution leads to a uniform formation of the NPs throughout the layer giving also the highest plasmonic enhancement in DSSCs.The incorporation of the Ag NPs in the porous TiO 2 layer decreases the dye loading but increases the short circuit current which is a direct indication of the plasmonic effect.The plasmonic effect has been confirmed also by EQE measurements.Nevertheless, the degradation of Ag NPs was observed in the cells that have been measured again three months after the manufacturing.In addition, the study confirmed that neither a thin sol-gel TiO 2 overlayer nor a poly(4-vinylpyridine) can provide sufficient protection for the long term stability of the Ag NPs against the corrosion of the I 3 − /I − electrolyte.

Figure 2 :
Figure 2: Cross section of DSSC with integrated Ag/TiO 2 core shell NPs in the active layer.

3. 1 .
Ag NPs Preparation.The influence of the heat treatment temperature on the formation of Ag NPs was evaluated by optical spectroscopy.The absorption spectra ( = 100% −  TOT −  TOT ) of the TiO 2 layers (reference) deposited on microscope glass and further impregnated with Ag ISO , Ag EG , and Ag/TiO 2 are presented in Figures 3(a), 3(b), and 3(c).

Figure 3 :Figure 4 :
Figure 3: (a-c) The absorption ( = 100% −  TOT −  TOT ) of the Ag ISO (a) and Ag EG (b) NPs (without shell) and Ag/TiO 2 core/shell NPs (c) located within the undyed TiO 2 layer on microscope glass substrate heat treated at different temperatures.The absorption of bare TiO 2 layer (black full line) and the bare TiO 2 layer covered with a thin "shell" TiO 2 layer (black dashed line) on a microscope glass substrate are shown for the reference.(d) The same as (c) but dyed porous TiO 2 layers.

Figure 5 :
Figure 5: The current-voltage (I-V) characteristics of typical reference DSSC and typical DSSCs with Ag NPs (Ag ISO , Ag EG , and Ag/TiO 2 ) that were measured after manufacturing (a) and after 3 months (b).

Figure 7 :
Figure 7: The absorption enhancement (Δ =  Ag −  ref ) of the undyed TiO 2 layers with Ag ISO , Ag EG , and Ag/TiO 2 NPs before (full lines) and after being exposed to the I 3 − /I − electrolyte (dashed lines).All NPs were thermally treated at 220 ∘ C.

3 −Figure 8 :
Figure 8: Degradation of the Ag ISO NPs coated with either a "shell" TiO 2 or a PVP shell (a) layer before and (b) immediately after being exposed to the electrolyte, and (c) after 24 hours.The Ag ISO NPs located within undyed porous TiO 2 layer were thermally treated at 220 ∘ C.

Table 1 :
Performance parameters ( SC ,  OC , FF, and ) of reference DSSCs and DSSCs with Ag NPs (Ag ISO , Ag EG , and Ag/TiO 2 ) measured after manufacturing/after 3 months.
[46]re 6: The normalized EQEs of the DSSCs containing Ag ISO or Ag EG NPs or Ag/TiO 2 core/shell NPs.The DSSCs without NPs or "shell" TiO 2 overlayer is given for reference.The EQEs were normalized to 0.8 at 345 nm.the Ag NPs.Additionally all cells showed a 5-8% and 11-12% increase in the  OC and FF, respectively.These changes cannot be associated with the presence of Ag NPs since similar variations were observed also for reference DSSCs.The  increased for 20, 15, 33, and 33%, relatively, for the reference and the DSSCs with Ag ISO , Ag EG , and Ag/TiO 2 NPs.The Ag ISO DSSC has still the highest performance but the differences became less distinctive.The degradation of the Ag NPs in the presence of I 3 − /I − based electrolyte is not surprising and is in accordance with the literature.A quick degradation of the Ag NPs was found by Lin et al.[46]and