How to Achieve Efficiencies beyond 22.1% for CdTe-Based Thin-Film Solar Cells

Round 1

Reviewer 1 Report

 

In this review, the authors reported the main issues related to CdTe and CdS/CdTe based solar cells by electrodeposition and closed-space sublimation methods with efficiencies achieved around 15-16% for lab-scale devices. They highlighted the lack of details on commercially available CdTe-based solar cells that achieve higher power conversion efficiencies (22.1%) and provide additional information in this manuscript that will unveil the major limitations of these devices to the research community. The highlights compared well with other similar studies and somewhat appealing. The study may be of interest to researchers working in similar fields. However, the following points should be addressed before further consideration:

 

1.     In the abstract, the authors highlighted their experience on CdTe-based solar cells, I would suggest to present more the main limitations of CdTe solar cell and the different ways of optimization of this technology.

2.     The authors describe very well the difficulties related to the nature of the CdTe material (rich or poor-Cd), the architecture and the electrical contact of the cell, the annealing post-processing of the growth film at very low temperature which would improve the crystallinity of the film and the defects formed, the impact of the grain boundaries. The authors briefly described the paper that conducted a simulation study in the NREL group combined with actual experimental results in 2015. The simulation study based on (SCAPS, wxAMPS, PC1D) is a powerful tool to complement these results and provide additional information on the formation of defects at the interface that inhibit charge separation which could limit the output performance of the device. More importantly, it can provide significant information on the highest theoretical efficiency of this particular device. Has there been a thorough simulation study on this aspect in the literature and would the authors consider adding a subsection on this aspect to the review to make it even more comprehensive and complete for the research community?

3.     The n-CdTe/metal interface has been discussed in this review. I suggest adding additional information about the interface between actual stack layer and in particular (nature and concentration of defects, limitation of the process used to grow the top layer without inducing degradation of the back layer, etc.).

4.     What about the stability of these devices and their long lifetime once they are manufactured? I suggest adding more information on the long-term stability of the thin film used for this particular application or the solar cell as a whole.

5.     The authors mentioned that both techniques (electrodeposition and closed-space sublimation methods) are more suitable for the fabrication of CdTe thin films. I suggest that the authors add more information on why this particular technique was used and whether or not other techniques, such as ALD for example, would be suitable for developing this layer.

Author Response

1. This paper provides authors’ experience on CdTe based solar cells, and ways forward to improve the device performance beyond reported record efficiency of 22.1%. Reviewer is asking for more, but we have summarised all the main limitations we have come across and different ways of optimizing this technology. We believe these main points are adequate to achieve high performing devices.
2. Reviewer’s comments are complimentary and support the content of this manuscript. Authors are experimentalists, and all ideas presented in the manuscript are based on experimental results. We agree with the reviewer that a good simulation work on these new devices will further support this research programme. We haven’t come across any good simulation work on this and we hope this manuscript will encourage such simulation work by experienced and knowledgeable theoretical scientists in the future.
3. Electrical contact work to CdTe is a complex. important and deep subject, where the first author spent his few decades of research. n-CdTe/metal interface has been studied thoroughly and results are published widely. Making simple metal contacts on n-CdTe is heavily affected by Fermi level pinning due to defects and this is summarised in Figure 6 in the revised version. Using intermediate layers to enhance rectifying contacts via MIS structures have also been studied and the results are in the public domain. Adding dopants like Cu and Sb are discussed in the manuscript but going deeper into discussion of this subject is away from the main aim of this paper.
4. Stability of these devices are excellent. The same materials and device structures used by First Solar company is now manufacturing GW of solar panels for more than 9 years. We have added some text to the manuscript to indicate the stability. The main issue is scientific understanding of the materials & device issues, and hence improving the efficiency beyond highest reported 22.1%. Industrial processing and understanding are not available to the academic and research community and this paper is providing that information to the open research community.
5. Manuscript clearly explains that there are 14 different growth techniques for growing CdTe material, but only two methods; electroplating and CSS methods have entered successful scaling up and commercialisation. Therefore, the discussions are limited only to these two growth methods, with first author’s deep experience with low temperature electrodeposition method. This accumulated wealth of knowledge can be used for more suitable high temperature growth by CSS method. The manuscript also indicate that this knowledge can be used for any other CdTe growth technique.

Reviewer 2 Report

1. Some figures showing the preparation strategies of CdTe solar cells should be given.

2. Device structure classification should be given as a Figure.

3. The energy band structure should be linked with electron/hole effective mass when talking about the tandem device.

4. Outlook about the device structure design is encouraged to develop. 

 

Author Response

1. Very good suggestion specially to help the new researchers entering into the field. We have added Figure 1 and relevant text in the Introduction section to show the main preparation stages.
2. Although the device structure is shown in the paper (Figure 3c and d) we have added a device structure with classifications in Figure 1.
3. When talking about tandem devices, there are different semiconductors used in these structures. In different materials, electron/hole effective masses are different, and therefore charge carrier drift velocities are different. For this reason, photo-generated charge carriers could travel fast in one device and travel slowly in the other. These speeds could affect the R&G rates and impact ionisation during charge carrier collection. But the direction of charge carrier flow does not change. What we measure is the resultant of all these effects.
4. We do not really understand this question. In this manuscript, we indeed encourage and propose methods to develop the new device structure.

Reviewer 3 Report

Please see the attached file.

Comments for author File: Comments.pdf

Author Response

Reviewer has identified the excellent motive of this review paper. However, reviewer has not gathered the main reason why we couldn’t achieve 22.1% ourselves. Our growth technique is a low-temperature electroplating, and by optimising we have achieved only 15.3% efficiency [44]. Low-temperature (~85℃) growth produce less crystalline material with high concentration of defects. The best growth technique for CdTe is high temperature (500 ~ 650℃), CSS method. It produces best crystalline and low-defect material. Scientists cannot have all growth techniques in their labs. Since we do not have CSS method in our labs, we summarise what we have learned from electroplating, and pass this valuable knowledge in a selfless and generous way to CSS researchers to move forward for the benefit of the mankind.

All other suggested minor changes have been corrected as listed below and highlighted in the revised version.

1. “Main Author” changed to “First author (IMD).
2. “The main author’s” in the abstract removed.
3. “horizontal axis” is absolutely correct but we have changed to “x-axis” as preferred by the reviewer.
4. (CdTe + Te) and (CdTe + Cd) are included in the figure now.
5. Vg values are now marked on the x-axis.

Reviewer 4 Report

I found the review "How to achieve efficiencies beyond 22.1% for CdTe-based thin film solar cells" well written and organized and, therefore, I recommend its publication in Energies after some minor revisions. 

- Page 1, line 20: semicolon instead of comma after 'clean energy'

- Page 1, line 25: please, remove the term "reduction", because it is in repetition with the term "reduce". 

- Page 1, line 25-26: please, provide references. Other PV technologies are emerging too, such as perovskite solar cells, quantum dot solar cells... Please, name them.

- Page 1, line 40: eliminate "round the world"

- Page 2, line 52: please, provide references. 

- Page 2, line 68: put comma instead of semicolon between 'electrodeposition' and 'with'. Please, provide references for this consideration.

- Page 2, line 75: please, remove the word 'cadmium'

- Page 2, line 76: please, add "respectively" after 'water'.

- Page 3, line 101-102: maybe the correct sentence is "In order to keep high crystallinity and moderate doping, that are needed for device fabrication, the layers..."

- Page 3, line 105: PL has not been introduced previously as acronym. 

- Page 7, line 207: please, remove the comma between 'also' and 'the drastic reduction', or put a comma after 'Note'.

- Pages 10-11, lines 287-318: Please, provide references.

- Page 11, lines 327-344: Please, provide references.

- Page 14, line 417: a closing brace is missing. A closing round bracket should be removed.

- Page 15, lines 456-467: Please, provide references.

- Page 17, lines 543-588: Please, provide references.

- Paragraph 3.2: Please, provide references for a better clarity.

- Page 20, line 644: The number of the Summary paragraph should be 3.4 and not 3.0.

- In conclusion, many strategies for enhancing the efficiency of CdTe-based thin film solar cells have been reported. However, the recorde efficiency is dated back to 2016. Six years is a long time in hot topic research. Please, add a comment and a possible explanation on the reason why the record efficiency has not been outperformed in such long time. 

Author Response

This reviewer has seen the value of this review article and thank you for very careful reading and indicating required minor corrections and missing references. All these have been corrected, and it has improved the scientific writing further.

Reviewer’s final comments are really important and thank you for requesting a possible explanation for slow progress. The followings are First author’s observations in this research field. CdS/CdTe solar cell was fabricated in early 1980s achieving ~8% efficiency. Since the CdS is always n-type in electrical conduction, and the CdS/CdTe device was a good rectifying diode, CdTe partner was assumed to be a p-type material. Therefore, all research carried out later considered this as a simple p-n junction diode, and results were analysed based on this assumption. Unfortunately, no research group experimentally tested this assumption, and followed the previously reported explanation.

As a result of many material improvements, and trial & error methods, the device efficiency gradually increased to ~15.9% in ~1992 (Ferakides et al). It took another decade to report 16.5% in ~2002 (Wu et al). During this decade, only 0.6% increased and then until 2013, no improvement was visible at all. So, this device development stagnated for ~20 years until 2013. This is mainly due to the lack of understanding of the science behind this device.

The first author (IMD), after working on this subject for a long time in both academia and in industry, recognised a different device structure than the one conventionally assumed. He observed the CdTe can be grown with both p-type and n-type electrical conduction depending on the composition of the material. Cd-richness produce n-CdTe and Te-richness produce p-CdTe. In fact, the best devices are coming from n-CdTe, and therefore he published a paper in 2002 (Ref.31) claiming that the high efficiency devices are NOT a simple p-n junction but a combination of an n-n hetero-junction and a large Schottky barrier at the back electrical contact. This was a highly debated paper, but IMD solidified this work producing 30 PhDs, publishing 2 books (ref. 6,7) and over 250 papers.

First Solar accelerated their research to improve efficiency and between 2013 & 2016, within a very short period, their team increased the efficiency from 16.5% to current record value of 22.1% (a huge achievement).  Understandably due to commercial confidentiality, this knowledge is not available to the PV community working in universities and research institutes. Six years has passed but there is no further improvement. This is why this review was written to present the most important evidence observed on this device. Authors of this review did not have CSS facility in their laboratories but have explored science behind this device structure using available electroplating technique. Right evidences are there from mid-1980s but these are all buried in the scientific literature. The whole idea of this review is for CSS researchers to move forward rapidly with all these filtered out relevant information and break this 6-years of stagnation. We anticipate a breakthrough in the near future surpassing the reported highest value of 22.1%.

A most recent paper (Ref. 50) published in 2015 by NREL is now quoting “Combining theoretical calculations and experimental measurements, we find that for both intrinsic CdTe and CdTe solar cell devices, longer minority-carrier lifetimes can be achieved under Cd-rich conditions, in contrast to the previous belief that Te-rich conditions are more beneficial”.

Now at least after four long decades, the PV community will believe that Cd-rich CdTe are more suitable for development of solar cells. Cd-rich CdTe is very clearly n-type in electrical conduction, and by working along the suggested lines, this stagnation could be ended.