Thermal Behavior of Methylammonium Lead-trihalide Perovskite ...
Thermal Behavior of Methylammonium Lead-trihalide Perovskite Photovoltaic Light Harvesters. Amalie Dualeh,§ Peng Gao,§ Sang Il Seok,†‡ Mohammad Khaja Nazeeruddin,§ Michael Grätzel§* § Laboratory of Photonics and Interfaces, Institute of Chemical Sciences and Engineering, École Polytechnique Fédérale de Lausanne (EPFL), Station 6, 1015 Lausanne, Switzerland † Division of Advanced Materials, Korea Research Institute of Chemical Technology, 141 Gajeong-Ro, Yuseong-Gu, Deajeon 305-600, Republic of Korea ‡ Department of Energy Science, Sungkyunkwan University, Suwon 440-746, Korea
*[email protected] SUPPORTING INFORMATION Calculations The Clausius-Clapeyron relation relates the vapor pressure p and the temperature T of a solid with its enthalpy of sublimation ∆Hsub, where R is the gas constant (8.3145 JK1
mol-1) according to equation S1;2 ௗ ୪୬ ௗ௧
=
∆ுೞೠ್ ோ் మ
(S1)
The first derivative of the TGA heat curve gives a direct measure of the instantaneous rate of mass loss msub at temperature T; ௗ ௗ௧
=
௧
= ݉௦௨
(S2)
In equilibrium conditions the rates of vapor condensation and evaporation are assumed to be equal. Hence the rate of mass loss by sublimation msub can be related to the vapor pressure by equation S3 according to Langmuir3 where A is the exposed sublimation surface area (here we take the area calculated from the diameter of the crucible containing the samples during the measurements) and Mw is the molecular mass of the material. ଵ ଶగோ்
=ቀ
ெೢ
ቁ
ଵൗ ଶ
݉௦௨
(S3)
Integrating equation S1 yields equation S4,4 which allows the determination ∆Hsub and temperature of sublimation Tsub from the slope and y-intercept of the linear plot of lnp vs 1/T (see Figure S1):
ln = −
ுೞೠ್ ଵ ோ
ቀ − ்
ଵ
்ೞೠ್
ቁ
(S4)
The linear trend of lnp vs 1/T confirms that sublimation is the only process leading to mass loss and there is no additional thermal decomposition taking place. By substituting equation S3 into equation S4 the following expression is obtained:4
ln(݉௦௨ ܶ
ଵൗ ଶ)
=−
ுೞೠ್ ଵ ோ
ቀ − ்
ଵ
்ೞೠ್
ଵ
ቁ − ln ቀ ଶ
ଶగோ
మ ெೢ
ቁ
(S5)
TGA of Perovskite precursor components Table S1. Quantitative TGA for the precursor materials used, indicating the temperature of the onset of weight loss, and at which 20% and 100% weight loss has occurred. Samples were heated at 10°C min-1 from 30 to 800°C under a 20 ml min-1 N2 gas flow. Precursors
Onset (°C)
weight
loss 20% weight loss (°C)
100% (°C)
CH3NH3Cl
185
267
315
CH3NH3I
234
300
345
PbCl2
500
612
714
PbI2
444
546
646
weight
loss
Table S2. Sublimation enthalpy and temperature of CH3NH3Cl determined from a linear least-squares fitting of the thermogravimetry data obtained at using different heating rates. Heat rate (°C min-1)
∆Hsub (kJmol-1)
Tsub (°C)
10
84.8±0.3
172±1
5
80.2±0.4
170±1
2.5
79.6±0.4
171±1
Figure S1. (a) TGA heating curves of the CH3NH3Cl precursor and (b) corresponding 1st derivatives measured at different heating rates; 10 (blue), 5 (green) and 2.5°C min-1 (red). (c) Calculated lnp vs 1/T.
Figure S2. Program (dashed) and sample (solid) temperature profiles for determination of the isothermal mass loss rate for CH3NH3Cl and CH3NH3I.
TGA of Perovskites from solution Table S3. Possible combinations of perovskite precursors and products formed. The molar ratio in the mixtures was determined to maintain the relationship between the Pb atom and the halide atom in the perovskite material at 1 to 3. The perovskite solutions contain 0.7 M of the inorganic component and use DMF or DMSO as solvent. DMSO was used for the perovskite mixtures 1 and 4, expected to form CH3NH3PbCl3, due to the poor solubility of CH3NH3Cl in DMF. Solution
Precursor
Molar ratio
Expected Perovskite
Excess
Inorganic
Organic
Inorganic:Organic
1
PbI2
CH3NH3Cl
1:3
CH3NH3PbCl3 2CH3NH3I
2
PbI2
CH3NH3I
1:1
CH3NH3PbI3
-
3
PbCl2
CH3NH3I
1:3
CH3NH3PbI3
2CH3NH3Cl
4
PbCl2
CH3NH3Cl
1:1
CH3NH3PbCl3 -
Figure S3 presents the TGA measurements for the four solutions. The initial large decrease corresponding to over 60% mass loss observed for all samples arises due to the loss of solvent. Once this is achieved it is believed that the material formed corresponds to the CH3NH3PbX3 perovskite and, in the case of solutions 1 and 3, any excess organic CH3NH3Y. The calculated contributions of the various possible components to the overall weight % of the perovskite are presented in Table S4, displaying the weight % of the solid perovskite products and in a 35 wt% solution.
Figure S3. TGA heating curves and corresponding 1st derivatives for the perovskite solutions heated at 10°C min-1 under a constant gas flow of N2 (20 ml min-1) from 30 to 800°C; solution 1 – 1:3 molar ratio of PbI2 and CH3NH3Cl in DMSO (blue), solution 2 – 1:1 molar ratio of PbI2 and CH3NH3I in DMF (red), solution 3 – 1:3 molar ratio of PbCl2 and CH3NH3I in DMF (black), solution 4 – 1:1 molar ratio of PbCl2 and CH3NH3Cl in DMSO (green).
Table S4. Calculated weight % contributions of expected components to total mass for the different perovskite mixtures. Component (%)
Perovskite Mixtures 1
2
3
4
1 PbI2
1 PbI2
1 PbCl2
1 PbCl2
3 CH3NH3Cl
1 CH3NH3I
3 CH3NH3I
1 CH3NH3Cl
Solid
Solution
Solid
Solution
Solid
Solution
Solid
Solution
Total materiala
100
35b
-
-
100
35
-
-
CH3NH3PbI3
-
-
100
35
82.11
28.74
-
-
CH3NH3PbCl3
52.09
18.23
-
-
-
-
100
35
PbI2
-
-
74.36
26.03
61.06
21.37
-
-
PbCl2
41.91
14.67
-
-
-
-
80.46
28.16
CH3NH3I
47.91c
16.77c
25.64
8.97
21.06
7.37
-
-
CH3NH3Cl
10.18
3.56
-
-
17.89c
6.26c
19.53
6.84
CH3NH2
4.68
1.64
5.01
1.75
4.11
1.44
8.99
3.15
HI
-
-
20.63
7.22
16.94
5.93
-
-
HCl
5.50
1.92
-
-
-
-
10.50
3.69
a
Total material weight % when excess organic material is formed in addition to the perovskite. b For ease of comparison all solutions were calculated based on a total of 35 wt%. c Excess organic component formed. Following this initial mass loss from the solvent vaporization, the solutions containing only one type of halide and a 1 to 1 molar ratio of inorganic to organic component (solutions 2 and 4) undergo two further distinct steps of weight loss. Solution 2, the pure iodide system, containing 35% by weight, displays a 9–10% weight loss between 300 and 350°C and a 26–30% mass loss between 450 and 550°C. The first step, corresponding to 10% mass lost originates due to the liberation of the organic component of the perovskite material formed (CH3NH3I), which has been
suggested to be in the form of the free amine, CH3NH2, and HI.5 The total amount of mass lost in this step corresponds to approximately 25%6 of the perovskite material mass, which is in good agreement with the calculated amount of the organic component (Table S4). Furthermore the temperature range of this mass loss step corresponds to that determined for the CH3NH3I precursor in Table S1. The remaining material can be deduced to be mainly PbI2 which undergoes thermal decomposition at 500°C as seen previously in Figure 1. Any remaining mass is a metallic residue and/or impurity. The pure chloride mixture (solution 4) when prepared to contain 0.7 M of the inorganic component, PbCl2, correlates to weight % of 17.5%. This solution displayed a mass loss of 3–3.5% between 220 and 300°C, which is in agreement with the calculated mass loss of 6.84% from a 35 wt% solution (Table S4) corresponding to the loss of the organic CH3NH3Cl component. The temperature range of this mass loss step is slightly lower than in the case for solution 2, which corresponds well to the shift in the 100% weight loss step observed between the organic precursors, CH3NH3Cl at 315°C and CH3NH3I at 345°C (Table S1). After the liberation of its organic component, solution 4 undergoes a mass loss of 13% between 500 and 650°C, corresponding to the decomposition of the inorganic PbCl2 constituent. This mass loss step is shifted by approximately 50–100°C relative to the inorganic decomposition step of the pure iodide system, reflecting the shift observed in the weight loss step of the thermal decomposition of the pure inorganic precursors (cf. Figure 1), hence confirming the nature of the inorganic component as PbI2 and PbCl2 for solutions 2 and 4 respectively. This is further supported by the magnitude of the mass loss step, which is within error of the calculated 28% PbCl2 contribution in a 35 wt% solution (Table S4).
In the case of the mixed halide systems (solution 1 and 3) the TGA curves become more complicated, in particular regarding the loss of the organic component. The formation of perovskite from the mixed halide precursors results in the additional byproduct of 2 molar equivalents of organic material. Where the pure halide systems undergo a simple, single step for the loss of the organic component, the mixed halide systems show two distinct stages in this weight loss step, which is clearly seen as two minima in the region between 200 and 300°C in the 1st derivative of the TGA curve. As previously mentioned these mixed halide systems lead to the formation of excess organic material which undergoes sublimation as the perovskite material is heated, manifesting as an additional mass loss step in the TGA curve. Solution 3, a 1 to 3 molar mixture of PbCl2 and CH3NH3I at 35 wt% in DMF, displays sequential mass losses of approximately 6–7%, 5% and 1–2% between 200 and 400°C. From Table S4 it is possible to identify these processes and correlated them with the initial sublimation of the 2 molar equivalents of CH3NH3Cl (6.26% weight loss), followed by the liberation of the organic component of the perovskite in the form of HI and CH3NH2 (5.93 and 1.44% weight loss respectively). This implies that the loss of the organic component of the perovskite follows the initial liberation of HI followed subsequently by the CH3NH2. Finally the mass loss of 22% between 450 and 550°C matches the decomposition of the inorganic PbI2 component. While similar features are observed for the other mixed halide system (solution 1) the temperature range of the mass loss step corresponding to the inorganic component decomposition is attributed to the decomposition of PbI2 and not PbCl2. As a result it is possible to infer that the perovskite formed from this mixture is not the pure CH3NH3PbCl3, but likely is a mixture of the trichloride perovskite, the triiodide perovskite and the PbI2 precursor. Additionally the concerned mass loss step
corresponds to 20%, which is equal to the mass of the PbI2 in the solution mixture, consisting of a 29 wt% mixture with 0.7 M of PbI2 and 1 to 3 molar ratio with CH3NH3Cl. The mass loss between 200 and 300°C associated with the loss of the organic component corresponds to 8–9% which coincides with the loss of 3 molar equivalents of CH3NH3Cl. This might suggest that this mixture does not form any type of perovskite and heating the mixture simply dries the precursor components. However the presence of the features in the organic mass loss step imply that the organic component is not lost in a single sublimation step, as would be the case if the materials were simply dried. This indicates that the organic component is somehow incorporated into the material matrix and subsequently stepwise liberated in the form of any excess organic material, HCl and CH3NH2.
References (1) (2) (3) (4) (5) (6)
Lee, M. M.; Teuscher, J.; Miyasaka, T.; Murakami, T. N.; Snaith, H. Science 2012, 338, 643. Vieyra-Eusebio, M. T.; Rojas, A. J. Chem. Engineer. Data 2011, 56, 5008. Langmuir, I. Phys. Rev. 1913, 2, 329. Fahlman, B. D.; Barron, A. R. Adv. Funct. Mater. 2000, 10, 223. Stoumpos, C. C.; Malliakas, C. D.; Kanatzidis, M. G. Inorg. Chem 2013, 52, 9019. Heo, J. H.; Im, S. H.; Noh, J. H.; Mandal, T. N.; Lim, C.-S.; Chang, J. A.; Lee, Y. H.; Kim, H.-J.; Sarkar, A.; Nazeeruddin, M. K.; Grätzel, M.; Seok, S. I. Nature Photon. 2013, 7, 486.
*[email protected] SUPPORTING INFORMATION Calculations The Clausius-Clapeyron relation relates the vapor pressure p and the temperature T of a solid with its enthalpy of sublimation ∆Hsub, where R is the gas constant (8.3145 JK1
mol-1) according to equation S1;2 ௗ ୪୬ ௗ௧
=
∆ுೞೠ್ ோ் మ
(S1)
The first derivative of the TGA heat curve gives a direct measure of the instantaneous rate of mass loss msub at temperature T; ௗ ௗ௧
=
௧
= ݉௦௨
(S2)
In equilibrium conditions the rates of vapor condensation and evaporation are assumed to be equal. Hence the rate of mass loss by sublimation msub can be related to the vapor pressure by equation S3 according to Langmuir3 where A is the exposed sublimation surface area (here we take the area calculated from the diameter of the crucible containing the samples during the measurements) and Mw is the molecular mass of the material. ଵ ଶగோ்
=ቀ
ெೢ
ቁ
ଵൗ ଶ
݉௦௨
(S3)
Integrating equation S1 yields equation S4,4 which allows the determination ∆Hsub and temperature of sublimation Tsub from the slope and y-intercept of the linear plot of lnp vs 1/T (see Figure S1):
ln = −
ுೞೠ್ ଵ ோ
ቀ − ்
ଵ
்ೞೠ್
ቁ
(S4)
The linear trend of lnp vs 1/T confirms that sublimation is the only process leading to mass loss and there is no additional thermal decomposition taking place. By substituting equation S3 into equation S4 the following expression is obtained:4
ln(݉௦௨ ܶ
ଵൗ ଶ)
=−
ுೞೠ್ ଵ ோ
ቀ − ்
ଵ
்ೞೠ್
ଵ
ቁ − ln ቀ ଶ
ଶగோ
మ ெೢ
ቁ
(S5)
TGA of Perovskite precursor components Table S1. Quantitative TGA for the precursor materials used, indicating the temperature of the onset of weight loss, and at which 20% and 100% weight loss has occurred. Samples were heated at 10°C min-1 from 30 to 800°C under a 20 ml min-1 N2 gas flow. Precursors
Onset (°C)
weight
loss 20% weight loss (°C)
100% (°C)
CH3NH3Cl
185
267
315
CH3NH3I
234
300
345
PbCl2
500
612
714
PbI2
444
546
646
weight
loss
Table S2. Sublimation enthalpy and temperature of CH3NH3Cl determined from a linear least-squares fitting of the thermogravimetry data obtained at using different heating rates. Heat rate (°C min-1)
∆Hsub (kJmol-1)
Tsub (°C)
10
84.8±0.3
172±1
5
80.2±0.4
170±1
2.5
79.6±0.4
171±1
Figure S1. (a) TGA heating curves of the CH3NH3Cl precursor and (b) corresponding 1st derivatives measured at different heating rates; 10 (blue), 5 (green) and 2.5°C min-1 (red). (c) Calculated lnp vs 1/T.
Figure S2. Program (dashed) and sample (solid) temperature profiles for determination of the isothermal mass loss rate for CH3NH3Cl and CH3NH3I.
TGA of Perovskites from solution Table S3. Possible combinations of perovskite precursors and products formed. The molar ratio in the mixtures was determined to maintain the relationship between the Pb atom and the halide atom in the perovskite material at 1 to 3. The perovskite solutions contain 0.7 M of the inorganic component and use DMF or DMSO as solvent. DMSO was used for the perovskite mixtures 1 and 4, expected to form CH3NH3PbCl3, due to the poor solubility of CH3NH3Cl in DMF. Solution
Precursor
Molar ratio
Expected Perovskite
Excess
Inorganic
Organic
Inorganic:Organic
1
PbI2
CH3NH3Cl
1:3
CH3NH3PbCl3 2CH3NH3I
2
PbI2
CH3NH3I
1:1
CH3NH3PbI3
-
3
PbCl2
CH3NH3I
1:3
CH3NH3PbI3
2CH3NH3Cl
4
PbCl2
CH3NH3Cl
1:1
CH3NH3PbCl3 -
Figure S3 presents the TGA measurements for the four solutions. The initial large decrease corresponding to over 60% mass loss observed for all samples arises due to the loss of solvent. Once this is achieved it is believed that the material formed corresponds to the CH3NH3PbX3 perovskite and, in the case of solutions 1 and 3, any excess organic CH3NH3Y. The calculated contributions of the various possible components to the overall weight % of the perovskite are presented in Table S4, displaying the weight % of the solid perovskite products and in a 35 wt% solution.
Figure S3. TGA heating curves and corresponding 1st derivatives for the perovskite solutions heated at 10°C min-1 under a constant gas flow of N2 (20 ml min-1) from 30 to 800°C; solution 1 – 1:3 molar ratio of PbI2 and CH3NH3Cl in DMSO (blue), solution 2 – 1:1 molar ratio of PbI2 and CH3NH3I in DMF (red), solution 3 – 1:3 molar ratio of PbCl2 and CH3NH3I in DMF (black), solution 4 – 1:1 molar ratio of PbCl2 and CH3NH3Cl in DMSO (green).
Table S4. Calculated weight % contributions of expected components to total mass for the different perovskite mixtures. Component (%)
Perovskite Mixtures 1
2
3
4
1 PbI2
1 PbI2
1 PbCl2
1 PbCl2
3 CH3NH3Cl
1 CH3NH3I
3 CH3NH3I
1 CH3NH3Cl
Solid
Solution
Solid
Solution
Solid
Solution
Solid
Solution
Total materiala
100
35b
-
-
100
35
-
-
CH3NH3PbI3
-
-
100
35
82.11
28.74
-
-
CH3NH3PbCl3
52.09
18.23
-
-
-
-
100
35
PbI2
-
-
74.36
26.03
61.06
21.37
-
-
PbCl2
41.91
14.67
-
-
-
-
80.46
28.16
CH3NH3I
47.91c
16.77c
25.64
8.97
21.06
7.37
-
-
CH3NH3Cl
10.18
3.56
-
-
17.89c
6.26c
19.53
6.84
CH3NH2
4.68
1.64
5.01
1.75
4.11
1.44
8.99
3.15
HI
-
-
20.63
7.22
16.94
5.93
-
-
HCl
5.50
1.92
-
-
-
-
10.50
3.69
a
Total material weight % when excess organic material is formed in addition to the perovskite. b For ease of comparison all solutions were calculated based on a total of 35 wt%. c Excess organic component formed. Following this initial mass loss from the solvent vaporization, the solutions containing only one type of halide and a 1 to 1 molar ratio of inorganic to organic component (solutions 2 and 4) undergo two further distinct steps of weight loss. Solution 2, the pure iodide system, containing 35% by weight, displays a 9–10% weight loss between 300 and 350°C and a 26–30% mass loss between 450 and 550°C. The first step, corresponding to 10% mass lost originates due to the liberation of the organic component of the perovskite material formed (CH3NH3I), which has been
suggested to be in the form of the free amine, CH3NH2, and HI.5 The total amount of mass lost in this step corresponds to approximately 25%6 of the perovskite material mass, which is in good agreement with the calculated amount of the organic component (Table S4). Furthermore the temperature range of this mass loss step corresponds to that determined for the CH3NH3I precursor in Table S1. The remaining material can be deduced to be mainly PbI2 which undergoes thermal decomposition at 500°C as seen previously in Figure 1. Any remaining mass is a metallic residue and/or impurity. The pure chloride mixture (solution 4) when prepared to contain 0.7 M of the inorganic component, PbCl2, correlates to weight % of 17.5%. This solution displayed a mass loss of 3–3.5% between 220 and 300°C, which is in agreement with the calculated mass loss of 6.84% from a 35 wt% solution (Table S4) corresponding to the loss of the organic CH3NH3Cl component. The temperature range of this mass loss step is slightly lower than in the case for solution 2, which corresponds well to the shift in the 100% weight loss step observed between the organic precursors, CH3NH3Cl at 315°C and CH3NH3I at 345°C (Table S1). After the liberation of its organic component, solution 4 undergoes a mass loss of 13% between 500 and 650°C, corresponding to the decomposition of the inorganic PbCl2 constituent. This mass loss step is shifted by approximately 50–100°C relative to the inorganic decomposition step of the pure iodide system, reflecting the shift observed in the weight loss step of the thermal decomposition of the pure inorganic precursors (cf. Figure 1), hence confirming the nature of the inorganic component as PbI2 and PbCl2 for solutions 2 and 4 respectively. This is further supported by the magnitude of the mass loss step, which is within error of the calculated 28% PbCl2 contribution in a 35 wt% solution (Table S4).
In the case of the mixed halide systems (solution 1 and 3) the TGA curves become more complicated, in particular regarding the loss of the organic component. The formation of perovskite from the mixed halide precursors results in the additional byproduct of 2 molar equivalents of organic material. Where the pure halide systems undergo a simple, single step for the loss of the organic component, the mixed halide systems show two distinct stages in this weight loss step, which is clearly seen as two minima in the region between 200 and 300°C in the 1st derivative of the TGA curve. As previously mentioned these mixed halide systems lead to the formation of excess organic material which undergoes sublimation as the perovskite material is heated, manifesting as an additional mass loss step in the TGA curve. Solution 3, a 1 to 3 molar mixture of PbCl2 and CH3NH3I at 35 wt% in DMF, displays sequential mass losses of approximately 6–7%, 5% and 1–2% between 200 and 400°C. From Table S4 it is possible to identify these processes and correlated them with the initial sublimation of the 2 molar equivalents of CH3NH3Cl (6.26% weight loss), followed by the liberation of the organic component of the perovskite in the form of HI and CH3NH2 (5.93 and 1.44% weight loss respectively). This implies that the loss of the organic component of the perovskite follows the initial liberation of HI followed subsequently by the CH3NH2. Finally the mass loss of 22% between 450 and 550°C matches the decomposition of the inorganic PbI2 component. While similar features are observed for the other mixed halide system (solution 1) the temperature range of the mass loss step corresponding to the inorganic component decomposition is attributed to the decomposition of PbI2 and not PbCl2. As a result it is possible to infer that the perovskite formed from this mixture is not the pure CH3NH3PbCl3, but likely is a mixture of the trichloride perovskite, the triiodide perovskite and the PbI2 precursor. Additionally the concerned mass loss step
corresponds to 20%, which is equal to the mass of the PbI2 in the solution mixture, consisting of a 29 wt% mixture with 0.7 M of PbI2 and 1 to 3 molar ratio with CH3NH3Cl. The mass loss between 200 and 300°C associated with the loss of the organic component corresponds to 8–9% which coincides with the loss of 3 molar equivalents of CH3NH3Cl. This might suggest that this mixture does not form any type of perovskite and heating the mixture simply dries the precursor components. However the presence of the features in the organic mass loss step imply that the organic component is not lost in a single sublimation step, as would be the case if the materials were simply dried. This indicates that the organic component is somehow incorporated into the material matrix and subsequently stepwise liberated in the form of any excess organic material, HCl and CH3NH2.
References (1) (2) (3) (4) (5) (6)
Lee, M. M.; Teuscher, J.; Miyasaka, T.; Murakami, T. N.; Snaith, H. Science 2012, 338, 643. Vieyra-Eusebio, M. T.; Rojas, A. J. Chem. Engineer. Data 2011, 56, 5008. Langmuir, I. Phys. Rev. 1913, 2, 329. Fahlman, B. D.; Barron, A. R. Adv. Funct. Mater. 2000, 10, 223. Stoumpos, C. C.; Malliakas, C. D.; Kanatzidis, M. G. Inorg. Chem 2013, 52, 9019. Heo, J. H.; Im, S. H.; Noh, J. H.; Mandal, T. N.; Lim, C.-S.; Chang, J. A.; Lee, Y. H.; Kim, H.-J.; Sarkar, A.; Nazeeruddin, M. K.; Grätzel, M.; Seok, S. I. Nature Photon. 2013, 7, 486.
