Supplementary Information

Uncovering the role of cathode buffer layer in organic solar cells

Boyuan Qi, Zhi-Guo Zhang, and Jizheng Wang*

Photographs, reflectance and external quantum efficiency (EQE) spectra of real cells

Figure S1 | Photographs, reflectance and EQE spectra of devices. (a) Photographs and (b) reflectance spectra of real OSCs with different treatment: none, methanol and CBL; (c) EQE spectra of devices with varying thickness of CBL (0~2.0 mg/ml).

From Fig. S1a, the color of the blend film taken by camera becomes darker after spin-coating, indicating that more lights are absorbed by the active layer. As a result, fewer lights come out from the device. Therefore the color is darker to the naked eyes. In order to verify it, the reflectance spectra measurement is performed on the real devices. It can be seen that with methanol and CBL treatment, the reflectance is evidently reduced from 450-800 nm. Reduced reflectance stands for enhanced absorption, this could also be proved by the EQE spectra (Fig. S1c).

Cyclic voltammograms of NMPAC

Figure S2 | Cyclic voltammograms of the NMPAC.Film is measured on glassy carbon electrodes in a 0.1 mol/L n-Bu4NPF6 acetonitrile solution at a sweep rate of 100 mV/s.

Calculation of the dielectric constantεi of the active layer

is determined by the CV measurement in dark at 1 MHz frequency and 0 V bias. The thickness of the depletion region can be estimated from the following equation:

(S1)

where L is the thickness of the depletion region, A the device area and C the measured capacitance. At high frequency the carriers can’t catch up with AC signal and show no contribution to the total capacitance. At this time, the depletion region could be regarded to extend to the whole device, so L is the thickness of the active layer (90 nm in this work, measured by the Surface Profilometer). Substituting L, A and C back into equation S1, could be calculated.

Performance of OSCs with Ca/Al cathode

Figure S3 | Performance of OSCs with Ca/Al cathode. J-V characteristics of device using Ca/Al as the cathode both under illumination and in dark.

(a)

(b)

(c)

Figure S4 | Typical atomic force microscopy (AFM) topography images of the devices.(a) control device; (b) device treated with methanol; (c) device treated with CBL. The root-mean-square (RMS) roughness of the three films are 0.39 nm for (a), 1.78 nm for (b) and 2.50 nm for (c), respectively.

Electron transporting property

Figure S5 | Electron transporting property of CBL and BHJ materials. J-V characteristics of two devices: (1) Al/CBL/Al; (2) ITO/PEDOT:PSS/PTB7:PC71BM/Al.

NMPAC is spin coated onto Al film at 1000 rpm with solution of 20 mg/ml. The thickness of CBL film is about 90 nm, which is similar to that of PTB7:PC71BM. Therefore device (1) and device (2) have almost the same volume with each other. From Figure S5, it can be seen that the current density of device (1) is much larger that that of device (2). This means CBL has a better electron transporting property than the blend.

Table S1 | Parameters of device using Ca/Al as the cathode. Parameters are measured and extracted from the J-V curves under illumination and in dark.
Device / VOC
[V] / JSC
[mA/cm2] / FF
[%] / PCE
[%] / Rs
[Ω cm2] / Rsh
[kΩ cm2] / n
[n.a.] / J0
[mA/cm2]
Ca/Al / 0.708 / 15.11 / 64.5 / 7.04 / 5.39 / 668.3 / 1.51 / 2.08×10-7

1