The role of surface passivation for efficient and photostable PbS quantum dot solar cells

Transcription

1 ARTICLE NUMBER: DOI: /NENERGY The role of surface passivation for efficient and photostable PbS quantum dot solar cells Yiming Cao 1,+, Alexandros Stavrinadis 1,+, Tania Lasanta 1, David So 1, Gerasimos Konstantatos 1,2 1 ICFO-Institut de Ciencies Fotoniques, The Barcelona Institute of Science and Technology, Castelldefels (Barcelona), Spain 2 ICREA-Institució Catalana de Recerca i Estudis Avançats, Lluis Companys 23, Barcelona, Spain + These authors contribute equally to this work. gerasimos.konstantatos@icfo.es. NATURE ENERGY 1

2 DOI: /NENERGY Supplementary Figures Absorption (%) nm 220 nm 300 nm ,100 Wavelength (nm) Supplementary Figure 1. Optical simulation of absorption of PbS QD solar cells with different thickness of TBAI-processed QD layers. The model solar cell structure is ITO (95nm)/ZnO (120 nm)/ TBAI-PbS (130, 200, 280nm)/EDT-PbS (20nm)/MoO3 (10nm)/Ag (40 nm)/au (150 nm). 2 NATURE ENERGY

3 DOI: /NENERGY SUPPLEMENTARY INFORMATION 100 Absorption (%) non-annealed 80 C-annealed ,100 Wavelength (nm) Supplementary Figure 2. Absorption of non- and 80 ºC-annealed TBAI-PbS /EDT-PbS layers on ZnO/glass substrates. NATURE ENERGY 3

4 DOI: /NENERGY (a) normalized intensity bound thiolate unbound thiol S2p PbS (b) COO, CO 2 Pb-OH O1s Pb-O Binding energy (ev) Binding energy (ev) Supplementary Figure 3. (a) S2p and (b) O1s spectra of 120 ºC-annealed PbS QD layer processed with EDT. The fitting peak of S2p3/2 from PbS is normalized to unity. The fitting peak from Pb-OH in O1s is normalized to that of non-annealed PbS QD layer processed with EDT in Fig. 2a. 4 NATURE ENERGY

5 DOI: /NENERGY SUPPLEMENTARY INFORMATION (a) non-annealed I3d 5/2 (b) 80 C-annealed I3d 5/2 Intensity (a.u.) I3d 3/2 I3d 3/ Binding energy (ev) Binding energy (ev) Supplementary Figure 4. I3d spectra of (a) non- and (b) 80 ºC-annealed PbS QD layer processed with TBAI. The raw spectra can be best deconvoluted to three components which, from higher to lower energy (those binding energies for the case of I3d5/2 are: ev, ev and ev, all constrained to FWHM of <1.1 ev and within <0.2 ev deviation between non- and 80 ºC annealed cases), should correspond to complexes with increasing charge transfer to the iodide S1. Thus while the lower energy species correspond to iodide atoms that are strongly bound to Pb atoms on the QD surface, the high energy component corresponds to iodide ions loosely bound to organic cations or weakly attached on the QD surface. The percentage of those three species before annealing (from higher to lower energy) are 6.3%, 89.5% and 4.2% and upon annealing they are 3.2%, 91.4% and 5.4%. The 80 C-annealing results to a decrease of the relative intensity of the highest energy peak and a corresponding increase of the lower energy peaks. The amount of increased iodide bound to Pb atoms, which is calculated by 0.60x(91.4%+5.4%)-0.54x(89.5%+4.2%)=0.07, matches closely to the amount of decreased Pb-OH in Supplementary Table 6. We propose that this effect is attributed to an increase of iodide atoms that are strongly bound to the surface of the QDs upon annealing replacing hydroxides. NATURE ENERGY 5

6 DOI: /NENERGY (a) Intensity (x10 5 counts) (b) Intensity (x10 5 counts) (c) 0.12 Intensity (x10 5 counts) processed with EDT non-annealed 80 C 120 C Binding energy (ev) 1 non-annealed 80 C 120 C Binding energy (ev) non-annealed 80 C 120 C Binding energy (ev) (d) Intensity (x10 5 counts) (e) Intensity (x10 5 counts) (f) Intensity (x10 5 counts) processed with TBAI non-annealed 80 C 120 C Binding energy (ev) non-annealed 80 C 120 C Binding energy (ev) non-annealed 80 C 120 C Binding energy (ev) Supplementary Figure 5. UPS of PbS QD layer processed with EDT (a) and TBAI (d) with varying annealing conditions. Detailed UPS with high binding energies of secondary electron cutoff (b,e) and low binding energies of the onset regions (c,f), which are obtained by the intercept of yellow lines, are presented. The ionization energy (valence band edge, Ev) was 6 NATURE ENERGY

7 DOI: /NENERGY SUPPLEMENTARY INFORMATION deduced by subtracting the energy of the incident beam with the difference in energy between the secondary electron cutoff and onset S2. The conduction band edge (Ec) was inferred by taking the Ev together with the bandgap of the QD layers non-annealed 80 C 120 C Energy levels (ev) E c E v Supplementary Figure 6. Band edge levels of EDT-processed PbS QD layers with varying annealing derived from the UPS spectra of Supplementary Figure 5. Energy levels (ev) non-annealed C C E c E v Supplementary Figure 7. Band edge levels of TBAI-processed PbS QD layers with varying annealing derived from the UPS spectra of Supplementary Figure 5. NATURE ENERGY 7

8 DOI: /NENERGY C (cm 2 V 1 s 1 ) non annealed 80 C Supplementary Figure 8. Carrier mobility in EDT-processed PbS QD layers with varying annealing. The average (symbols) and stand deviation (error bars) were calculated from a sample of four to seven devices. 120 C (cm 2 V 1 s 1 ) 10-4 non annealed 80 C Supplementary Figure 9. Carrier mobility in TBAI-processed PbS QD layers with varying annealing. The average (symbols) and stand deviation (error bars) were calculated from a sample of four to seven devices. 8 NATURE ENERGY

9 DOI: /NENERGY SUPPLEMENTARY INFORMATION (a) Lifetime (s) non-annealed 80 C-annealed (b) R (cm 3 s 1 ) non-annealed 80 C-annealed V oc (V) V oc (V) Supplementary Figure 10. (a) Carrier lifetime and (b) recombination rate (R) as a function of the open-circuit potential for the non- and 80 ºC-annealed solar cell devices processed with TBAI EMII Count PCE (%) Supplementary Figure 11. Histogram of PCE of 80 ºC-annealed PbS QD solar cells with Au anode processed with TBAI and EMII. NATURE ENERGY 9

10 DOI: /NENERGY Supplementary Figure 12. Certificated PbS QD solar cell with first exciton peak at 930 nm of colloidal QDs. 10 NATURE ENERGY

11 DOI: /NENERGY SUPPLEMENTARY INFORMATION (a) 25 (b) J (ma cm 2 ) Count V (V) PCE (%) (c) EQE (%) (nm) Supplementary Figure 13. (a) J-V curves, (b) histogram of PCE and (c) EQE spectrum of EMII-processed PbS QD solar cells fabricated by colloidal QDs with first exciton absorption peak at 850 nm. NATURE ENERGY 11

12 DOI: /NENERGY Supplementary Figure 14. Certificated PbS QD solar cells with first exciton peak at 850 nm of colloidal QDs. 12 NATURE ENERGY

13 DOI: /NENERGY SUPPLEMENTARY INFORMATION Carrier density (x10 18 cm 3 ) processed with TBAI EMII V oc (V) Supplementary Figure 15. Voc-dependent carrier density of EMII and TBAI-processed PbS QD solar cells annealed at 80 ºC. NATURE ENERGY 13

14 DOI: /NENERGY Nomalized intensity I3d 3/2 processed with TBAI EMII I3d 5/ Binding energy (ev) Supplementary Figure 16. I3d spectra of 80 ºC-annealed PbS QD layer processed with TBAI and EMII. The intensity is respectively normalized to the peak of Pb4f7/2. 14 NATURE ENERGY

15 DOI: /NENERGY SUPPLEMENTARY INFORMATION J (ma cm 2 ) initial PCE: 8.5% aged PCE: 8.3% V (V) Supplementary Figure 17. Air stability of 80 ºC-annealed EMII-processed PbS QD solar cells with Au anode aging for 30 days stored under dark in ambient condition. NATURE ENERGY 15

16 DOI: /NENERGY (a) 8 (b) PCE (%) non-annealed 80 C-annealed V oc (V) Time (hours) Time (hours) (c) 26 (d) J sc (ma cm 2 ) FF Time (hours) Time (hours) Supplementary Figure 18. Stability of non- and 80 ºC-annealed EMII-processed PbS QD solar cells with Au anode soaked under simulated AM1.5G 100 mw cm 2 illumination in ambient condition. The average (symbols) and stand deviation (error bars) were calculated from a sample of four to eight solar cells. The PbS solar cells are neither encapsulated nor equipped with ultraviolet filter. 16 NATURE ENERGY

17 DOI: /NENERGY SUPPLEMENTARY INFORMATION V (V) I (ma) x x x x Time (s) 0.14 Time (s) V (V) I (ma) x x x x x10-4 V (V) Time (s) I (ma) Time (s) x x x x x x10-4 Time (s) Time (s) 1.22 V (V) x x10-5 Time (s) I (ma) x x10-4 Time (s) Supplementary Figure 19. Transient photovoltage (left) and photocurrent (right) decay traces of 80 C-annealed PbS QD solar cells processed with EMII ligand under variant white light illumination intensity conditions as indicated by different data colour. NATURE ENERGY 17

18 DOI: /NENERGY D-S current (na) D-S current (na) 15 V ds = 20V processed with EDT Gate voltage (V) 20 V ds = 20V 10 0 processed with TBAI Gate voltage (V) Supplementary Figure 20. Representative traces of the FET characteristics of non-annealed PbS QD FETs processed with EDT and TBAI ligands. Double traces represent forward and reverse scans showing low hysteresis. 18 NATURE ENERGY

19 DOI: /NENERGY SUPPLEMENTARY INFORMATION Supplementary Tables Supplementary Table 1. Impact of thermal annealing on photovoltaic parameters of TBAIprocessed QD solar cells under simulated AM1.5G 100 mw cm 2 illumination conditions. anneal Voc (V) Jsc (ma cm 2 ) FF PCE (%) non ºC ºC ºC ºC NATURE ENERGY 19

20 DOI: /NENERGY Supplementary Table 2. Atomic ratio of non- and 80 ºC-annealed PbS QD layer processed with EDT. anneal Pb S O C non ºC Supplementary Table 3. Fitting parameters and quantitative analysis of S2p spectra of non- and 80 C-annealed PbS QD layer processed with EDT in Fig. 2a and 2d. anneal component peak (ev) FWHM (ev) area (%) component ratio a non PbS bound thiolate unbound thiol ºC PbS bound thiolate a The component ratio is calculated from the fitting area percentage of the corresponding component multiplied by the atomic ratio of S to Pb in Supplementary Table 2. The amount of the ratio of bound thiolate to Pb in the 80 ºC-annealed sample, being 0.53, is close to the total amount of the ratio of bound thiolate and unbound thiol to Pb in the non-annealed counterpart ( =0.56). 20 NATURE ENERGY

21 DOI: /NENERGY SUPPLEMENTARY INFORMATION Supplementary Table 4. Fitting parameters and quantitative analysis of O1s spectra of non- and 80 ºC-annealed PbS QD layer processed with EDT in Fig. 2b and 2e. anneal component peak (ev) FWHM (ev) area (%) component ratio a non Pb-O Pb-OH COO, CO ºC Pb-O Pb-OH COO, CO OH a The component ratio is calculated from the fitting area percentage of the corresponding component multiplied by the atomic ratio of O to Pb in Supplementary Table 2. Upon 80 ºCannealing, the amount of decreased ratio of Pb-OH ( =0.2), is comparable to that of increased ratio of bound thiolate ( =0.1) in Supplementary Table 3, indicating that at least half of the native hydroxide ligands being removed from the surface upon mild annealing are replaced by originally unbound thiol moieties yielding OH containing byproducts, which show a new peak at ev. That hypothesis is supported by the fact that the component ratio from OH (0.09) is almost equal to the increase (by 0.1) of the component ratio of the bound thiolate. NATURE ENERGY 21

22 DOI: /NENERGY Supplementary Table 5. Atomic ratio of non- and 80 ºC-annealed QD layer processed with TBAI. anneal Pb S O I C non ºC NATURE ENERGY

23 DOI: /NENERGY SUPPLEMENTARY INFORMATION Supplementary Table 6. Fitting parameters and quantitative analysis of O1s spectra of non- and 80 ºC-annealed PbS QD layer processed with TBAI in Fig. 2c and 2f. anneal component peak (ev) a FWHM (ev) area (%) component ratio b non Pb-O Pb-OH COO, CO OH ºC Pb-O Pb-OH COO, CO OH a Even when correcting the energy axis of the XPS spectra by positioning the centre of the C1s peak at 284.8eV, all other XPS peaks (O1s, S2p, Pb4f) of the TBAI-treated films appear to be shifted by the same energy of 0.4 ev compared to the respective ones of the EDT samples (e.g. compared to the respective O1s peak positions from Pb-OH shown in Supplementary Table 4). We attribute this effect to the different surface dipole of the QD surface caused by the different moieties (iodide and thiol ones). b The component ratio is calculated from the fitting area percentage of the corresponding component multiplied by the atomic ratio of O to Pb in Supplementary Table 5. Although the surface of TBAI-processed QD are passivated by iodide ligands as indicated by the ratio of I to Pb being 0.54 in Supplementary Table 5, the native hydroxide ligands, of which the ratio is 0.14, could have profound effects on efficiency and stability of QD solar cells. After 80 ºCannealing, such ligands can be replaced by iodide ions from residual TBAI molecules found in the QD layer. The amount of the ratio of Pb-OH to Pb reduces by 0.06 ( =0.06), NATURE ENERGY 23

24 DOI: /NENERGY which matches closely to that of increased ratio of iodide ions bound to Pb atoms in the following Supplementary Figure NATURE ENERGY

25 DOI: /NENERGY SUPPLEMENTARY INFORMATION Supplementary Table 7. Light intensity-dependent photovoltaic parameters of 80 ºC-annealed PbS QD solar cells with Au anode processed by EMII and TBAI. processed with irradiation (mw cm 2 ) Voc (V) Jsc (ma cm 2 ) FF PCE (%) EMII TBAI NATURE ENERGY 25

26 DOI: /NENERGY Supplementary References S1. Yang, C.-H., Yau, S.-L., Fan, L.-J. & Yang, Y.-W. Deposition of lead iodide films on Rh(100) electrodes from colloidal solutions-the effect of an iodine adlayer. Surf. Sci. 540, (2003). S2. Schlaf, R., Parkinson, B. A., Lee, P. A., Nebesny, K. W. & Armstrong, N. R. HOMO/LUMO alignment at PTCDA/ZnPc and PTCDA/ClInPc heterointerfaces determined by combined UPS and XPS measurements. J. Phys. Chem. B 103, (1999). 26 NATURE ENERGY