Copper borate as a photocathode in p-type dye-sensitized solar cells

Abstract

p-Type dye-sensitized solar cells (p-DSSCs) have recently become a major research focus because coupling with n-type DSSCs yields highly efficient tandem DSSCs. Indeed many delafossite-like compounds appear as promising candidates for p-DSSCs due to their deep valence band position and high hole mobility. In this paper, the synthesis of CuBO₂ was attempted by a facile sol–gel methodology. Then, the as-obtained particles were used to prepare photocathodes for p-DSSCs with DPP-NDI dye as sensitizer and tris(4,4′-di-tert-butyl-2,2′-bipyridine)cobalt(III/II) as redox mediator. Due to the deeper valence band position compared with classical NiO photocathode, the “CuBO₂” based p-DSSC presents an open-circuit photovoltage (V_oc) of 453 mV, which is 150 mV higher than that of NiO in the same conditions. The results show that “CuBO₂” is a potential alternative for NiO in p-DSSCs.

---

In traditional n-type semiconductor (SC) based dye-sensitized solar cells (DSSCs), the photocurrent results from electron injection of the photoexcited dye into the conduction band (CB) of the n-type SCs.^1,2 Conversely, in p-type semiconductor based DSSCs, the inverse principle is at work, i.e. holes are injected from the photo-excited dye into the valence band (VB) of the p-type SC. To achieve higher energy conversion efficiency, an attractive strategy is to construct tandem DSSCs, in which an n-type photoanode is coupled with a p-type photocathode.³ Until now, the energy conversion efficiency of p-DSSCs is much lower than that of n-DSSCs (generally a few tenths of percent vs. about 15%), and this mismatch limits the final energy conversion of tandem cells.⁴ This drawback has prompted the development of better performing photocathodes with new p-SCs, in particular to reach higher photopotentials.^4–6 Very recently, Perera et al.⁷ have investigated NiO based p-DSSCs with an energy conversion efficiency of 2.51%, which is the highest reported efficiency for a p-DSSC so far. This breakthrough clearly results from the use of a new redox mediator ([Fe(acac)₃]^0/1−) with a very negative redox potential rather than a p-type semiconductor with a deeper valence band. Unfortunately, such an approach would not be effective in the context of the tandem cell concept where the open-circuit photovoltage (V_oc) is related to the position of the quasi-Fermi levels of the photoanode and the photocathode, i.e. the difference in energy between the bottom of the CB of n-SC and the top of the VB of p-SC.^3,4

The valence band position of NiO is as low as +0.50 V vs. NHE (at pH 7),⁸ which limits the V_oc of p-DSSCs and consequently of the tandem DSSCs. Besides, the photocurrent of NiO based p-DSSCs is restricted by its low hole mobility.⁹ Accordingly, the development of new p-type semiconductors with deep VB has become a priority. Delafossite oxides are promising candidates owing to their high transparency and conductivity.¹⁰ More importantly, in Cu(I)-based delafossite materials, Cu 3d orbitals strongly hybridize with the O 2p orbitals, making the hole charge carriers (related to a slight off-stoichiometry) to be delocalized both on oxygen and copper atoms.¹¹ This facilitates the hole mobility and subsequently, should minimize electron–hole recombinations at the dye/p-SC interfaces due to a regular charge flow from the surface to the interior of the nanoparticles. Several works on CuGaO₂, CuCrO₂, and CuAlO₂ based p-DSSCs were reported recently.^9,10 Under AM 1.5 illumination, V_oc of 375 mV,¹² 145 mV¹³ and 333 mV¹⁴ were reached for the gallium, chromium and aluminum derivatives, with respectively the following dye and redox mediator couples: PMI-NDI and Co³⁺/Co²⁺, coumarin 343 and I₂/I⁻, and PMI-6T-TPA and I₂/I⁻. These values are much higher than the V_oc of 285 mV, 90 mV, and 218 mV measured for NiO based DSSCs prepared in similar conditions.

To the best of our knowledge, there is no report on CuBO₂ based p-DSSCs. Yet, in the ABO₂ series, boron is a valuable alternative to Al, Cr or Ga. Based on the investigations of Snure and Tiwari¹⁵ on thin films, the hole mobility of CuBO₂ is 100 cm² V⁻¹ s⁻¹ which is much higher than that of CuAlO₂ (10 cm² V⁻¹ s⁻¹),¹¹ CuGaO₂ (0.2 cm² V⁻¹ s⁻¹),¹⁶ and CuCrO₂ (0.2 cm² V⁻¹ s⁻¹).¹⁷ Then CuBO₂ appears at first sight as a very good candidate for p-DSSCs.

Herein, we report on the synthesis and characterization of CuBO₂ particles and their subsequent use to fabricate p-DSSC exhibiting a higher V_oc than conventional NiO-based DSSCs with the same dye and electrolyte.

The CuBO₂ particles were synthesized by the sol–gel method of Santra et al.¹⁸ Typically, soluble copper(II) nitrate trihydrate and boric acid were used as precursors, which were dissolved in water and nitric acid, respectively. Then citric acid was added in the mixed solution of the precursors to generate the gel. After sintering the gel under air at 550 °C, dark green CuBO₂ particles were obtained. More details are given in the ESI.†

Powdered CuBO₂ samples were first analyzed by XRD. The XRD pattern, displayed in Fig. 1, gathers the characteristic peaks of CuBO₂, i.e. contributions at 32.4, 35.4, 38.6, 48.7, 53.3, and 58.1° in 2θ assigned to (006), (100), (012), (106), (107), and (018) planes by Snure and Tiwari,¹⁵ Santra et al.^19,20 and Ruttanapun²¹ for a rhombohedral structure with a and c parameters of about 2.84 Å and 16.52 Å, respectively. Unfortunately, any attempt to refine the XRD pattern via the Rietveld analysis based on CuBO₂ material with the delafossite structure type, i.e. a structure built upon edge-sharing (BO₆) octahedra defining infinite 2/∞[BO₂] layers linked to each other by Cu cations in dumbbell coordination²² (space group (SG): R3̄m) was not successful. Thus, even if reported in many articles, the real existence of CuBO₂ with a delafossite structure type is highly questionable. This issue was initially raised by Cerqueira et al.²³ who first witnessed the improbability of the reported structure type for CuBO₂. From ab initio calculations, they allege that the ground state of the aforementioned material would correspond to a layered structure with triangular coordinated boron in 2/∞ [BO₂] layers separated by Cu in linear coordination forming infinite 1/∞ [CuO] zigzag chains (SG: Cc, a = 3.63 Å, b = 12.45 Å, c = 4.42 Å, β = 86.78°). Two other metastable structures, a tetragonal one (SG: I4̄m2, a = 2.54 Å, c = 10.82 Å) and an orthorhombic one (SG: Cmc2₁, a = 2.57 Å, b = 11.67 Å, c = 4.51 Å), both with tetrahedral coordinated boron, were predicted as more stable than the common P6₃/mmc or R3̄m delafossite structures. Attempts to refine the collected XRD pattern with all these structures were not successful, as well as trials with the wurtzite-derived β-NaFeO₂-structure type, recently stabilized via soft chemistry route by Omata et al.²⁴ for CuGaO₂. Thus, at this stage the CuBO₂ structure has to be viewed as non-identified, in contradiction with many reports in the literature. A more accurate examination of the XRD pattern, followed by a Rietveld refinement, even led to the assignment of the aforementioned well defined peaks to CuO, i.e. a monoclinic system with a = 4.6741(19) Å, b = 3.4307(15) Å, c = 5.1295(23) Å and β = 99.33(3). This clearly means that the XRD peaks commonly attributed to CuBO₂ with a delafossite structure type belong in fact to the CuO phase. On the other hand, chemical analyses by Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES) clearly attest the presence of boron with a Cu : B ratio of 1.15 : 1. This perfectly agrees with the ratio reported by Santra et al. determined from EDX investigations^19,20 that ranges from 1.03 to 1.35. Moreover, measurements by X-ray photoelectron spectroscopy (XPS, Fig. S1†) reveal the presence of Cu^I and Cu^II cations at the surfaces. All these measurements suggest therefore that our product is composed of a “CuBO₂” majority phase with an amorphous character which is mixed with a small amount of cupric oxide.

Fig. 1 XRD pattern of dark green “CuBO₂” prepared by sol–gel method annealed at 550 °C for 1 h under air.

In order to verify this assumption our prepared material was examined with a transmission electron microscope (TEM). The TEM micrographs displayed in Fig. 2a and b, clearly evidence the existence of well faceted crystals embedded in a gangue with a granular texture. From the electron diffraction (Fig. 2c), the crystals can be ascribed to CuO. The gangue, which would correspond to more than ∼90% in volume of the sample, turns to be very reactive under the electron beam at room temperature and requires low temperature (−180 °C) to be examined. This gangue reveals an electron diffraction pattern of a poorly crystallized phase (see Fig. 2d) and likely corresponds to an “amorphous” phase of CuBO₂.

Fig. 2 (a) and (b) Room temperature TEM image of our prepared sample with CuO crystal in dark and the “CuBO₂” gangue in bright. (c) and (d) Selected area electron diffraction patterns of CuO and “CuBO₂”.

Room temperature Raman spectrum of “CuBO₂” was collected and compared to that of CuO, Cu₂O and B₂O₃, three binary phases that might be present as sub-products (see Fig. 3). Namely, in the 100–1400 cm⁻¹ range, six well defined vibration modes at 142, 196, 282, 719, 958, and 1248 cm⁻¹ are observed for “CuBO₂”. In addition, two featured broad massifs are located in the 420–520 cm⁻¹ and 1100–1200 cm⁻¹ regions. Peaks at 196 cm⁻¹ and 719 cm⁻¹ exhibit analogies with peaks observed in delafossite materials and could correspond to pseudo A_g (205 cm⁻¹, 213 cm⁻¹ for CuCrO₂ and CuGaO₂) and A_1g (707 cm⁻¹, 738 cm⁻¹ for CuCrO₂ and CuGaO₂) modes, respectively.^25–27 Others peaks are unexplained. Nevertheless, it is clear that the Raman spectrum of “CuBO₂” is significantly different to those of CuO, Cu₂O and B₂O₃. Consequently, if present, we may suppose that these secondary phases have to be viewed as minor impurities in agreement with our TEM analysis.

Fig. 3 Room temperature Raman spectrum of “CuBO₂” CuO, Cu₂O, and B₂O₃ under excitation at 514 nm.

The thermal stability of “CuBO₂” was studied by TGA-DTA as shown in Fig. 4. Below 400 °C, the weight loss originates from the departure of physically and chemically adsorbed water molecules on the surface, which generates two peaks in heat flow curve at 100 and 220 °C. Above 600 °C, the phase Cu₃B₂O₆ (JCPDS 28-0398) is generated as confirmed by XRD (see Fig. S2 in the ESI†). This confirms the presence of a large amount of Boron in our amorphous “CuBO₂” powder. The oxidation of “CuBO₂” may then occur in air according to the reactions.

12CuBO₂ + 3O₂ → 4Cu₃B₂O₆ + 2B₂O₃

4CuBO₂ + 2CuO + O₂ → 2Cu₃B₂O₆

Fig. 4 TGA and DSC curves of CuBO₂ as prepared under air atmosphere.

Due to the low melting point of B₂O₃ (450 °C), no diffraction peaks of B₂O₃ are identified in Fig. S2.† Thus, CuBO₂ clearly behaves differently with other delafossite materials which commonly transit towards a spinel phase under heating in air.²⁸ The specific surface area of our as-prepared “CuBO₂” material determined by Brunauer–Emmett–Teller (BET) method is 18.5(1) m² g⁻¹ (i.e. a much lower value than the one of Inframat® NiO (151.9(5) m² g⁻¹) used as reference in the following). This value also turns to be lower than that of CuCrO₂ and CuGaO₂ prepared by hydrothermal synthesis.^29,30

The transformed Kubelka–Munk reflectance spectrum of “CuBO₂” and NiO is given in Fig. S3† and agrees with the dark green color of the material since absorptions in the red and blue regions of the visible range take place. UV-vis transmittance spectrum of CuBO₂ film prepared by screen printing and sintered at 350 °C for 60 min for using in p-DSSC (see ESI† for details) is shown in Fig. 5. The film has a good transmittance in visible region and its transparence is not affected by the presence of CuO in small amount.

Fig. 5 Transmittance spectrum of “CuBO₂” film on FTO substrate prepared by screen printing and sintered at 350 °C for 60 min under air. The thickness of the film is ∼1 μm.

To determine the VB position of CuBO₂ the electrochemical impedance spectroscopy (EIS) measurement was carried out (see ESI† for details), in which the Mott–Schottky plot was extracted at 1 kHz as shown in Fig. 6a. The negative slope of the plot confirms the p-type semiconductivity of “CuBO₂”. The flat band potential was determined from the intercept of the axis with potential values, which is 0.31 V (vs. SCE) for CuBO₂ at pH 9.4. Consequently, there is a 0.2 V positive shift compared to that of NiO measured (Fig. 6b) in the same conditions (and a 0.16 V shift compared to CuO).³¹ The flat band potential (FBP) of “CuBO₂” is quite similar to that of other delafossite compounds. Hence, FBP of CuAlO₂ (0.49 V vs. Ag/AgCl at pH 7.2)³² and CuGaO₂ (0.49 V vs. SCE at pH 6.3)¹² are extrapolated to be ∼0.31 V and ∼0.47 V vs. SCE at pH of 9.4, respectively, when taking into account the difference of the reference electrodes and the pH values. Since the V_oc is equal to the difference of the quasi-Fermi level close to the VB and redox potential of electrolyte, a higher V_oc is therefore expected for “CuBO₂” based p-DSSCs compared to NiO based p-DSSCs. We can also notice that the carrier density calculated from the slope of Mott–Schottky plots is 3.07 × 10¹³ cm⁻³ for CuBO₂ and 5.00 × 10¹⁶ cm⁻³ for NiO (see ESI† for details).

Fig. 6 (a) Flat band potential of CuBO₂ and (b) NiO in 1 M LiClO₄ at pH 9.4.

Next, mesoporous films were deposited on FTO glass by screen printing using a paste made of “CuBO₂” particles (see ESI† for details). The film was sintered at 350 °C under air atmosphere for 1 h to eliminate the organic components of the used ink and favour a good adhesion of the mineral on the substrate. The dye DPP-NDI³³ was used as sensitizer and the redox mediator was tris(4,4′-di-tert-butyl-2,2′-bipyridine)cobalt(III/II)³⁴ (Fig. S4 and S5†). The photovoltaic performances of the solar cells made with “CuBO₂” and NiO cathodes in the same conditions are gathered in Table 1. Clearly, the V_oc of 453 mV recorded with “CuBO₂” is 150 mV higher than that of NiO based solar cell (Fig. 7). This is a direct consequence of the higher flat band potential of “CuBO₂” compared to NiO. On the other hand, the short-circuit current (J_sc) is lower than that of NiO. This may be explained by the much lower specific surface area of “CuBO₂” film prepared so far compared to that of NiO leading to a much lower dye loading, and consequently a lower capability to absorb light (see Fig. S6†). We would like to also emphasis that the carrier density of CuBO₂ is much lower than that of NiO limiting thus the J_sc value. In spite of this, the dark current of CuBO₂ based p-DSSC is about 0.7 μA cm⁻² at V = 0 (vs. 1.9 μA cm⁻² for NiO-based DSSC) which confirms that the photocurrent comes from hole injection from DPP-NDI dye into the valence band of CuBO₂.

Table 1 Photovoltaic performances of p-DSSC fabricated with “CuBO₂” and NiO

Material	Voc (mV)	Jsc (mA cm^−2)	FF (%)	η (%)
CuBO2	453	0.02	41	0.0036
NiO	305	0.31	29	0.0276

---

Fig. 7 Photovoltage–photocurrent curves under AM 1.5 illumination (1000 W m⁻²) and dark current curves of solar cells constructed from CuBO₂ (a) and NiO (b).

Conclusions

In summary, we have synthesized a “CuBO₂” powder as reported in the literature. This material described in many reports as a delafossite corresponds actually to an unknown amorphous “CuBO₂” phase containing a small amount of crystallized CuO impurity. This amorphous “CuBO₂” powder revealed a p-type character and a flat band potential of 0.31 V vs. SCE at a pH of 9.4. By using this powder, DPP-NDI dye as sensitizer and tris(4,4′-di-tert-butyl-2,2′-bipyridine)cobalt(III/II) as redox mediator, we have fabricated CuBO₂ based p-DSSCs. They exhibit a high V_oc of 450 mV, which is 150 mV higher compared to NiO based p-DSSCs. Accordingly, CuBO₂ is a promising candidate to fabricate higher performing p-DSSCs, and the investigation is under away to increase the specific surface area of this material.

Acknowledgements

For financial support, we are grateful to the agency ANR-Progelec via program POSITIF (No. ANR-12-PRGE-0016-01). The authors thank Prof. A. Barnabé from Toulouse for fruitful discussions.

Notes and references

1. A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo and H. Pettersson, Chem. Rev., 2010, 110, 6595–6663.
2. M. Grätzel, J. Photochem. Photobiol., C, 2003, 4, 145–153.
3. A. Nattestad, A. J. Mozer, M. K. R. Fischer, Y.-B. Cheng, A. Mishra, P. Bauerle and U. Bach, Nat. Mater., 2010, 9, 31–35.
4. F. Odobel and Y. Pellegrin, J. Phys. Chem. Lett., 2013, 4, 2551–2564.
5. F. Odobel, Y. Pellegrin, E. A. Gibson, A. Hagfeldt, A. L. Smeigh and L. Hammarström, Coord. Chem. Rev., 2012, 256, 2414–2423.
6. F. Odobel, L. le Pleux, Y. Pellegrin and E. Blart, Acc. Chem. Res., 2010, 43, 1063–1071.
7. I. R. Perera, T. Daeneke, S. Makuta, Z. Yu, Y. Tachibana, A. Mishra, P. Bäuerle, C. A. Ohlin, U. Bach and L. Spiccia, Angew. Chem., Int. Ed., 2015, 54, 3758–3762.
8. J. He, H. Lindström, A. Hagfeldt and S.-E. Lindquist, J. Phys. Chem. B, 1999, 103, 8940–8943.
9. L. D'Amario, L. J. Antila, B. Pettersson Rimgard, G. Boschloo and L. Hammarström, J. Phys. Chem. Lett., 2015, 6, 779–783.
10. M. Yu, T. I. Draskovic and Y. Wu, Phys. Chem. Chem. Phys., 2014, 16, 5026–5033.
11. H. Kawazoe, M. Yasukawa, H. Hyodo, M. Kurita, H. Yanagi and H. Hosono, Nature, 1997, 389, 939–942.
12. A. Renaud, B. Chavillon, L. le Pleux, Y. Pellegrin, E. Blart, M. Boujtita, T. Pauporte, L. Cario, S. Jobic and F. Odobel, J. Mater. Chem., 2012, 22, 14353–14356.
13. D. Xiong, Z. Xu, X. Zeng, W. Zhang, W. Chen, X. Xu, M. Wang and Y.-B. Cheng, J. Mater. Chem., 2012, 22, 24760–24768.
14. A. Nattestad, X. Zhang, U. Bach and Y.-B. Cheng, J. Photonics Energy, 2011, 1, 011103.
15. M. Snure and A. Tiwari, Appl. Phys. Lett., 2007, 91, 092123.
16. T. Mine, H. Yanagi, K. Nomura, T. Kamiya, M. Hirano and H. Hosono, 5th Int. Symp. Transparent Oxide Thin Films Electron. Opt., 2008, vol. 516, pp. 5790–5794.
17. A. Barnabe, Y. Thimont, M. Lalanne, L. Presmanes and P. Tailhades, J. Mater. Chem. C, 2015, 3, 6012–6024.
18. S. Santra, N. S. Das and K. K. Chattopadhyay, Mater. Lett., 2013, 92, 198–201.
19. S. Santra, N. S. Das and K. K. Chattopadhyay, Mater. Res. Bull., 2013, 48, 2669–2677.
20. S. Santra, N. S. Das, B. Das, D. Banerjee and K. K. Chattopadhyay, Cryst. Growth Des., 2015, 15, 1518–1525.
21. C. Ruttanapun, J. Appl. Phys., 2013, 114, 113108.
22. A. Buljan, P. Alemany and E. Ruiz, J. Phys. Chem. B, 1999, 103, 8060–8066.
23. T. F. T. Cerqueira, R. Sarmiento-Pérez, F. Trani, M. Amsler, S. Goedecker, M. A. L. Marques and S. Botti, MRS Commun., 2013, 3, 157–160.
24. T. Omata, H. Nagatani, I. Suzuki, M. Kita, H. Yanagi and N. Ohashi, J. Am. Chem. Soc., 2014, 136, 3378–3381.
25. O. A. Yassin, S. N. Alamri and A. A. Joraid, J. Phys. D: Appl. Phys., 2013, 46, 235301.
26. M. Han, K. Jiang, J. Zhang, W. Yu, Y. Li, Z. Hu and J. Chu, J. Mater. Chem., 2012, 22, 18463–18470.
27. O. Aktas, K. D. Truong, T. Otani, G. Balakrishnan, M. J. Clouter, T. Kimura and G. Quirion, J. Phys.: Condens. Matter, 2012, 24, 036003.
28. A. P. Amrute, Z. Łodziana, C. Mondelli, F. Krumeich and J. Pérez-Ramírez, Chem. Mater., 2013, 25, 4423–4435.
29. D. Xiong, Z. Xu, X. Zeng, W. Zhang, W. Chen, X. Xu, M. Wang and Y.-B. Cheng, J. Mater. Chem., 2012, 22, 24760–24768.
30. R. Srinivasan, B. Chavillon, C. Doussier-Brochard, L. Cario, M. Paris, E. Gautron, P. Deniard, F. Odobel and S. Jobic, J. Mater. Chem., 2008, 18, 5647–5653.
31. K. Nakaoka, J. Ueyama and K. Ogura, J. Electrochem. Soc., 2004, 151, C661–C665.
32. B. Das, A. Renaud, A. M. Volosin, L. Yu, N. Newman and D.-K. Seo, Inorg. Chem., 2015, 54, 1100–1108.
33. L. Favereau, J. Warnan, Y. Pellegrin, E. Blart, M. Boujtita, D. Jacquemin and F. Odobel, Chem. Commun., 2013, 49, 8018–8020.
34. E. A. Gibson, A. L. Smeigh, L. le Pleux, J. Fortage, G. Boschloo, E. Blart, Y. Pellegrin, F. Odobel, A. Hagfeldt and L. Hammarström, Angew. Chem., Int. Ed., 2009, 48, 4402–4405.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details, XPS spectrum, XRD pattern of Cu₃B₂O₆ and NiO, reflectance of CuBO₂, absorption spectra of CuBO₂ and NiO films and the molecular structure of the sensitizer and electrolyte. See DOI: 10.1039/c5ra24397a

---