A Comparative Study of TiO2 Paste Preparation Methods Using Solvothermally Synthesised Anatase Nanoparticles in Dye-Sensitised Solar Cells
Round 1
Reviewer 1 Report
The author reports a new method to increase the conversion efficiency with smaller power windows. Overall, the paper was good writing and clearly described the method, results and discussion. However, the fundamental mechanism to explain why the PCE can be improved with N719, etc.
So, I would suggest authors can give more discuss and description in their "results and discussion" sections before the paper can be considered to be published.
Author Response
The author reports a new method to increase the conversion efficiency with smaller power windows. Overall, the paper was good writing and clearly described the method, results and discussion. However, the fundamental mechanism to explain why the PCE can be improved with N719, etc.
So, I would suggest authors can give more discuss and description in their "results and discussion" sections before the paper can be considered to be published.
Response:
The reasons for the higher PCEs with SANP-SǀN719 and SANP-SǀD149 devices compared to SANP-TǀN719 and SANP-TǀD149, were mainly attributed to the differences in their light scattering capabilities which explained by:
· Diffuse reflectance (Figure 4b) and direct transmittance (Figure S2)
· IPCEs curves (Figure 6b)
· Normalised IPCEs curves (Figure S3, highlighting scattering effects on IPCE responses)
SANP-SǀD149 gave only slightly enhanced performance to SANP-T, while differences for SANPǀN719 were more distinct. This was explained in part due to the differences in N719 and D149 extinction coefficients. (The higher extinction coefficient of D149 enabled similar light harvesting efficiency, while the low extinction coefficient of N719 along with lower light scattering affected the light harvesting efficiency).
On the other hand, it is expected that there may be slightly more amorphous material present in SANP-T as a result of less sintering time compared to SANP-S (see XRD results and Raman analysis). This may affect the charge injection efficiency, especially for N719 devices with its variety of possible binding modes. This is in line with the findings of our previous publication [Chemical Communications 54 pp381-384].
All these explanations were already stated in the discussion and conclusion parts.
From p4. “The amorphous content in SANP-S was quantified in our previous report [26], to be approximately 3%, however the lower crystallinity seen in SANP-T suggests the possibility of a slightly higher amorphous content, which may affect the charge injection efficiency. [18]”
Reviewer 2 Report
Al-Attafi et al. reported a detailed side-by-side comparison of different paste forming techniques, with one yielding scattering and the other non-scattering films. Solar cells using D149 dye delivered a comparable performance using both approaches (6.9% with drying versus 6.4% without). The results are interesting. I would recommend accepting this manuscript for publication after addressing the following comments.
In Figure 3, the authors explained that SANP-S (Figure 2a and b) has a more aggregated structure and non-uniform surface topography compared to SANP-T (Figure 2c and d). It seems to this reviewer that the Figure b showed that SANP-S has larger particle size and bigger pores that that in figure d. This may be confirmed with even higher resolution SEM images.
Figure 4 shows that the light absorption for the N719 dye desorbed from the SANP-S or SANP-T films are very similar; however, Figure 6 shows that Jsc is the main reason for the difference in the DSSC performance. The authors explained that dye loading was different. There seems some inconsistence in these explanations. I suggest the authors to correlated figure 4 and 6 and provide a clearer explanation why Jsc was different using different paste of TiO2.
There are other reported works for TiO2 paste preparation. Here are some examples: ACS Applied Energy Materials 1, no. 11 (2018): 6288-6294; Journal of Molecular Liquids, 244, 97-102, 2017; AIP Advance, 5, 067134, 2015; Journal of Materials Chemistry A, 2014,2, 11448-11453; Nano Research, 2014, 7 (8), 1154-1163; Energy & Environmental Science,3, 1507-1510, 2010. The authors may find these papers useful and read them.
Author Response
Al-Attafi et al. reported a detailed side-by-side comparison of different paste forming techniques, with one yielding scattering and the other non-scattering films. Solar cells using D149 dye delivered a comparable performance using both approaches (6.9% with drying versus 6.4% without). The results are interesting. I would recommend accepting this manuscript for publication after addressing the following comments.
In Figure 3, the authors explained that SANP-S (Figure 2a and b) has a more aggregated structure and non-uniform surface topography compared to SANP-T (Figure 2c and d). It seems to this reviewer that the Figure b showed that SANP-S has larger particle size and bigger pores that that in figure d. This may be confirmed with even higher resolution SEM images.
Response:
We thank the reviewer for the cautious eye. We realise that the particle size of TiO2 nanoparticles generally can be affected by hydrothermal reaction conditions and sintering temperatures. In this case however the difference between the preparations of the two materials was the omission of a drying step and one sintering process. Ultimately both SANP-T and SANP-S were exposed to 500 °C, but for different times. Further to this, these temperatures are too low to cause significant particle growth over the time scales used here.
Having said that, upon initial inspection of SEM images (Figure 2a-d), we agree that SANP-S do appear larger, however this is due to aggregation (i.e. they look like bigger, but when measured are actually not). The particle sizes for both SANP-S and SANP-T were measured from SEM images to be approximately 25 nm, which coincides with calculations made by applying the Scherrer equation to XRD traces (highlighted in the main text). Figure 3a shows that the pore size distribution is generally similar, although slightly broader for SANP-S. This is attributed to formation some larger pores resulted from the aggregated SANP-S which is further described by Figure 3b (pore volume/pore size distribution).
Further, surface area values of both materials are again similar, indicating similar average particle size and pore size distribution (Figure 3a).
Figure 4 shows that the light absorption for the N719 dye desorbed from the SANP-S or SANP-T films are very similar; however, Figure 6 shows that Jsc is the main reason for the difference in the DSSC performance. The authors explained that dye loading was different. There seems some inconsistence in these explanations. I suggest the authors to correlated figure 4 and 6 and provide a clearer explanation why Jsc was different using different paste of TiO2.
The dye loading, for both dyes, on SANP-T is slightly higher that on SANP-S (but not enough to justify the observed differences in Jsc). SANP-S and SANP-T display similar difference in dye loading when either D149 or N719 was employed (Table 1). The dye absorption spectra of desorbed dyes in Figure 4a were plotted as overlayed, so the absorption spectra of the desorbed N719 appeared similar due to the difference in their intensities in absorption units. Figure 4a has been separated to two panels to better show the difference in the absorption of the desorbed dyes.
The reasons for the higher PCEs with SANP-SǀN719 and SANP-SǀD149 devices compared to SANP-TǀN719 and SANP-TǀD149, were mainly attributed to the differences in their light scattering capabilities which explained by:
· Diffuse reflectance (Figure 4b) and direct transmittance (Figure S2)
· IPCEs curves (Figure 6b)
· Normalised IPCEs curves (Figure S3, highlighting scattering effects on IPCE responses)
SANP-SǀD149 gave only slightly enhanced performance to SANP-T, while differences for SANPǀN719 were more distinct. This was explained in part due to the differences in N719 and D149 extinction coefficients. (The higher extinction coefficient of D149 enabled similar light harvesting efficiency, while the low extinction coefficient of N719 along with lower light scattering affected the light harvesting efficiency).
On the other hand, it is expected that there may be slightly more amorphous material present in SANP-T as a result of less sintering time compared to SANP-S (see XRD results and Raman analysis). This may affect the charge injection efficiency, especially for N719 devices with its variety of possible binding modes. This is in line with the findings of our previous publication [Chemical Communications 54 pp381-384].
All these explanations were already stated in the discussion and conclusion parts.
From p4. “The amorphous content in SANP-S was quantified in our previous report [26], to be approximately 3%, however the lower crystallinity seen in SANP-T suggests the possibility of a slightly higher amorphous content, which may affect the charge injection efficiency. [18]”
There are other reported works for TiO2 paste preparation. Here are some examples: ACS Applied Energy Materials 1, no. 11 (2018): 6288-6294; Journal of Molecular Liquids, 244, 97-102, 2017; AIP Advance, 5, 067134, 2015; Journal of Materials Chemistry A, 2014,2, 11448-11453; Nano Research, 2014, 7 (8), 1154-1163; Energy & Environmental Science,3, 1507-1510, 2010. The authors may find these papers useful and read them.
We thank the reviewer for pointing these articles out, however the direct link to this work is not so clear. On the other hand, there are many, many studies reporting the preparation of photolectrodes for DSCs, as we point out in our introduction
(page 2 - “A large portion of work published on DSC development (~40%, which equates to more than 1200 publications) has been on the synthesis and/or modification of metal oxides, mostly TiO2, as efficient photoanode materials. [17]”).
It is simply not possible to include reference to all of these papers.
Round 2
Reviewer 1 Report
Although authors use their previous papers [26], [18] to support their results, [26] has not been published yet. As one of important factors, please indicate if this paper has been published or not. Otherwise, it will cause trouble to understand their key point "he amorphous content in SANP-S was quantified in our previous report [26], to be approximately 3%".
Please addressed.
Author Response
We again thank the reviewers for their time and effort in the consideration of our manuscript. We have carefully read and considered their feedback, addressing their concerns in responses below and with amendments to our manuscript:
#Reviewer 1: Comments and Suggestions for Authors
Although authors use their previous papers [26], [18] to support their results, [26] has not been published yet. As one of important factors, please indicate if this paper has been published or not. Otherwise, it will cause trouble to understand their key point "The amorphous content in SANP-S was quantified in our previous report [26], to be approximately 3%".
Please addressed.
Response:
Thank you to the reviewer for picking this up. Initially we had expected the other paper to be finalised much earlier, to proceed this one. Unfortunately, there have been other hold-ups. As such we have revised the text on page 4:
“Previously we demonstrated role of amorphous content plays on DSC performance, as a result of decreased charge injection efficiency. [18] Preliminary measurements show the amorphous content to be quite low in these materials (only a few percent at most; data to be published elsewhere [26]), indicating that this will not be a major factor affecting the charge injection efficiency.”