Layers of CH3NH3PbI3 are investigated by modulated surface photovoltage spectroscopy (SPV) during heating in vacuum. As prepared CH3NH3PbI3 layers behave as a p-type doped semiconductor in depletion with a band gap of 1.5 eV. After heating to 140 °C the sign of the SPV signals of CH3NH3PbI3 changed concomitant with the appearance of a second band gap at 2.36 eV ascribed to PbI2, and SPV signals related to charge separation from defect states were reduced after moderate heating.

Rather high solar energy conversion efficiencies of almost 12 (Ref. 1) or 15% (Ref. 2) have been reached with solar cells based on CH3NH3PbI3 absorber layers within a comparably short time of research and development. The band gap of CH3NH3PbI3 layers is about 1.5 eV, and high open circuit voltages (VOC) of 0.83 (Ref. 3) and 0.993 V (Ref. 2) have been measured on related solar cells by different groups. High values of VOC of solar cells based on only 300–400 nm thin CH3NH3PbI3 absorber layers2 point to efficient surface passivation occurring during processing at low temperatures. This work is aimed to get more information about surface passivation of CH3NH3PbI3 absorber layers during low temperature processing. For this purpose CH3NH3PbI3 layers were heated in vacuum and investigated with the very sensitive method of modulated surface photovoltage spectroscopy (SPV).4 

Layers of CH3NH3PbI3 were prepared following a slightly modified procedure of Snaith.1 First, CH3NH3I has been produced by mixing methylamine and HI (sigma Aldrich). The resulting precipitate CH3NH3I was washed three times using diethylether and re-crystallized in ethanol by adding ether (CH3OCH3). Second, PbI2 was dissolved in γ-butyrolactone (sigma Aldrich), and CH3NH3I was added in stoichiometric ratio. Third, the solution was stirred overnight at 60 °C. Fourth, layers of CH3NH3PbI3 were deposited onto infrared pre-heated molybdenum coated soda lime glass substrates by spin coating (2000 rpm for 3 min) using a hot solution of CH3NH3PbI3 in γ-butyrolactone.

The phase composition near the surface of CH3NH3PbI3 layers deposited on molybdenum substrates was investigated by grazing incidence X-ray diffraction (GIXRD, Bruker AXS D8 Advance) before and after annealing in vacuum for 30 min at temperatures of 100, 140, and 160 °C (Figures 1(b)–1(e), respectively). For comparison, the GIXRD pattern is also shown for PbI2 powder (Figure 1(a)).

FIG. 1.

Grazing incidence x-ray diffraction pattern of PbI2 powder and of CH3NH3PbI3 layers before and after annealing at 100, 140, and 160 °C ((a)–(e), respectively). The dashed lines mark diffraction peaks of PbI2.

FIG. 1.

Grazing incidence x-ray diffraction pattern of PbI2 powder and of CH3NH3PbI3 layers before and after annealing at 100, 140, and 160 °C ((a)–(e), respectively). The dashed lines mark diffraction peaks of PbI2.

Close modal

The GIXRD results reveal CH3NH3PbI3 characteristic peaks at 14.04, 28.42, and 43.08 corresponding to the (002), (220), and (330) planes of CH3NH3PbI3, respectively, as shown in Figure 1(b). Our calculated lattice parameters for CH3NH3PbI3 with a tetragonal unit cell are a = 8.881 ± 0.005 Å and c = 12.560 ± 0.005 Å which is in agreement with previous reports.2 The size of the CH3NH3PbI3 crystallites has been obtained by using the Scherrer equation and amounted to about 30–40 nm. There are no PbI2 characteristic diffraction peaks for the CH3NH3PbI3 layer annealed at 100 °C. In contrast, characteristic diffraction peaks of PbI2 appeared after annealing of the CH3NH3PbI3 layer at 140 and 160 °C. The appearance of the PbI2 phase correlated with the appearance of a signature at about 510 nm in optical reflection measurements performed with an integrating sphere.

Modulated SPV spectra were measured4 in vacuum at a modulation frequency of 8 Hz. Illumination was performed with a halogen lamp and a quartz prism monochromator. In-phase and phase-shifted by 90° SPV signals were detected with a high-impedance buffer and a double phase lock-in amplifier. To form the measurement capacitor a clean mica sheet (thickness about 20–30 μm) was gently pressed between the sample surface and the cylindrical electrode coated with SnO2:F. The samples were heated in vacuum to temperatures up to 220 °C and characterized in situ after cooling down to 30 °C. As remark, the mica sheets showed a haze after heating due to partial mass transfer from the substrate onto the mica sheets. Therefore a fresh mica sheet has been taken for each new measurement.

A CH3NH3PbI3 layer was successively heated and cooled down to 30 °C. Figure 2 depicts spectra of the in-phase and phase-shifted by 90° SPV signals measured at the as-deposited layer and after successive heating to 60, 100, and 140 °C. The in-phase SPV signal of the as-prepared CH3NH3PbI3 layer was negative and set on at photon energy of 1.5 eV corresponding to the band gap of CH3NH3PbI3. The SPV spectra of the as prepared CH3NH3PbI3 layer correspond to a p-type doped semiconductor with a depletion region at the surface. After heating to 60 °C the shapes of the SPV spectra remained unchanged while the maximum in-phase and phase-shifted by 90° SPV signals increased from −272 to −376 μV and from 110 to 183 μV, respectively. After heating to 100 °C signs of the in-phase and phase-shifted by 90° SPV signals changed between 1.56 and 2.04 eV and between 1.58 and 1.80 eV, respectively. The reduction of the signal heights and the change of the signs give evidence for the onset on surface chemical reactions changing the surface electronic properties. After heating to 140 °C the in-phase SPV signals were positive up to photon energies of 2.42 eV and negative for higher photon energies while the phase-shifted by 90° SPV signals were negative up to photon energies of 2.52 eV and positive for higher photon energies, i.e., the direction of modulated charge separation changed in comparison to the as prepared and heated to 60 °C samples. Further, after heating to 140 °C a strong change of the SPV spectra set on at photon energy of 2.36 eV which corresponds to the band gap of a new phase appearing during heating and which is in very good agreement with the band gap of PbI2 (Ref. 5) and with our results obtained by GIXRD. As a remark, there is now straight forward interpretation of the change of the sign of the modulated SPV signals due to the complexity of transport phenomena and charging/discharging of defect states (see, for example, Refs. 6 and 7).

FIG. 2.

Modulated surface photovoltage spectra of the in-phase (filled circles) and phase-shifted by 90° (open circles) signals measured at 30 °C on an as-prepared CH3NH3PbI3 layer (a) and on CH3NH3PbI3 layers after heating to maximum temperatures of 60 °C (b), 100 °C (c), and 140 °C (d).

FIG. 2.

Modulated surface photovoltage spectra of the in-phase (filled circles) and phase-shifted by 90° (open circles) signals measured at 30 °C on an as-prepared CH3NH3PbI3 layer (a) and on CH3NH3PbI3 layers after heating to maximum temperatures of 60 °C (b), 100 °C (c), and 140 °C (d).

Close modal

In contrast to conventional semiconductors such as silicon, there was a mass transfer from the CH3NH3PbI3 layer onto the mica sheet during heating. In order to eliminate the influence of material transferred onto the mica sheet during heating, ex situ SPV measurements were performed with a fresh mica sheet replacing the mica sheet which has been used during heating experiments. Figure 3 shows the spectra of the photovoltage amplitude, which is the square root of the sum of the squared in-phase and phase-shifted by 90° SPV signals, after heating to 60, 100, 140, and 160 °C. After heating to 100 °C, the PV amplitude decreased for photon energies below 1.5 eV by a factor of 4–5 in comparison to the sample heated up to 60 °C whereas the PV amplitude increased by a factor of about 1.5 for photon energies above 1.6 eV. This behavior is a strong indication for passivation of surface defects at CH3NH3PbI3 surface during moderate heating. On the other hand, after heating to 140 °C, the PV amplitude increased by a factor of about 5 for photon energies below 1.4 eV and by a factor of about 11 or 7 at 1.6 or 1.7 eV, respectively, in comparison to the sample heated up to 60 °C. After heating to 160 °C, the PV amplitude increased by a factor of about 20–22 for photon energies below 1.5 eV and by a factor of about 15–16 at photon energies above 1.6 eV in comparison to the sample heated up to 60 °C. The stronger increase of the PV amplitude related to charge separation from defect states after heating at 160 °C shows that defect formation at the CH3NH3PbI3/PbI2 interface can become significant.

FIG. 3.

Spectra of the PV amplitude measured with a fresh mica sheet replacing the mica sheets after heating to Tmax = 60, 100, 140, and 160 °C (open circles, filled squares, filled stars, and open triangles, respectively).

FIG. 3.

Spectra of the PV amplitude measured with a fresh mica sheet replacing the mica sheets after heating to Tmax = 60, 100, 140, and 160 °C (open circles, filled squares, filled stars, and open triangles, respectively).

Close modal

For comparison, the open circuit voltage of solar cells based on a CH3NH3PbI3 absorber increased after heating at 100 °C and continued to increase even after heating at 160 °C while the short circuit current density decreased after heating at 160 °C (Ref. 8) due to an increased transformation rate of CH3NH3PbI3 to PbI2. Further, mass transfer from CH3NH3PbI3 layers to mica sheets gives evidence for the high mobility of ionic species in CH3NH3PbI3 layers. This is beneficial for the formation of intimate contact regions in solar cells based on CH3NH3PbI3 absorber layers. As remark, solar cells based on CH3NH3PbI3 absorber layers have a tremendous strategic potential exceeding that of most photovoltaic absorber materials9 and the formation of intrinsic passivating layers on CH3NH3PbI3 layers is a prerequisite for their broad application.

P. Supasai and N. Rujisamphan are grateful to the Faculty of Science of the Kasetsart University and to the Thai Government, respectively, for funding a research stay at the Helmholtz-Centre Berlin for Materials and Energy. The authors are grateful to the reviewer for helpful criticism.

1.
J. M.
Ball
,
M. M.
Lee
,
A.
Hey
, and
H. J.
Snaith
,
Energy Environ. Sci.
6
,
1739
(
2013
).
2.
J.
Burschka
,
N.
Pellet
,
S. J.
Moon
,
R. H.
Baker
,
P.
Gao
,
M. K.
Nazeeruddin
, and
M.
Grätzel
,
Nature
499
,
316
(
2013
).
3.
B.
Cai
,
Y.
Xing
,
Z.
Yang
,
W.-H.
Zhang
, and
J.
Qiu
,
Energy Environ. Sci.
6
,
1480
(
2013
).
4.
Th.
Dittrich
,
A.
Gonzales
,
T.
Rada
,
T.
Rissom
,
E.
Zillner
,
S.
Sadewasser
, and
M.
Lux-Steiner
,
Thin Solid Films
535
,
357
(
2013
).
5.
A.
Ferreira da Silva
,
N.
Veissid
,
C. Y.
An
,
I.
Pepe
,
N.
Barros de Oliveira
, and
A. V.
Batista da Silva
,
Appl. Phys. Lett.
69
,
1930
(
1996
).
6.
D. R.
Frankl
and
E. A.
Ulmer
,
Surf. Sci.
6
,
115
(
1967
).
7.
T. K.
Sharma
,
S.
Kumar
, and
K. C.
Rustagi
,
J. Appl. Phys.
92
,
5959
(
2002
).
8.
J.-H.
Im
,
C.-R.
Lee
,
J.-W.
Lee
,
S.-W.
Park
, and
N.-G.
Park
,
Nanoscale
3
,
4088
(
2011
).
9.
C.
Wadia
,
A. P.
Alivisatos
, and
D. M.
Kammen
,
Environ. Sci. Technol.
43
,
2072
(
2009
).