Supporting Information Cu2ZnGeS4 Nanocrystals from Air-Stable ...
Supporting Information
Cu2ZnGeS4 Nanocrystals from Air-Stable Precursors for Sintered Thin Film Alloys
Anthony S. R. Chesman,*† Joel van Embden,† Enrico Della Gaspera, † Noel W. Duffy,† Nathan A. S. Webster, ‡, Jacek J. Jasieniak*†
† ‡
CSIRO Manufacturing Flagship, Bayview Avenue, Clayton, Victoria 3168, Australia. CSIRO Mineral Resources Flagship, Bayview Avenue, Clayton, Victoria 3168, Australia.
*E-mail: [email protected] and [email protected].
(a)
(b)
Figure S1. (a) Colorless plate crystals of [Ge(gly)2(H2O)2]; (b) Non-corrosive, air stable crystals of [Ge(gly)2(H2O)2] on a glove.
100 90
[Ge(gly)2(H2O)2] [Ge(gly)2] (Calc. 86 %, Found 86 %)
Mass (%)
80 70 60 50
GeO2 (Calc. 41 %, Found 42 %)
40 30 20 50
100
150
200
250
300
350
Temperature (°C)
Figure S2. TGA thermogram of decomposition of [Ge(gly)2(H2O)2] to GeO2.
Intensity (a.u., offset)
[Ge(gly)2(H2O)2]
C=O
Ge-OC
[Ge(gly)2(OLA)2]
Oleylamine 3600
3100
2600
2100
1600
1100
600
Wavenumber (cm−1) Figure S3. IR spectra of [Ge(gly)2(H2O)2], [Ge(gly)2(OLA)2] in OLA and neat OLA, highlighting the similar spectral features of the glycolate ligand. 1
8
7
6
4.11
4.79 9
5
4
3
2
1
ppm
2.72 2.69 2.65 2.01 1.98 1.45 1.42 1.28 1.26 1.25 0.90 0.87
3.98
5.35 5.33 5.30 3.98
7.29
Figure S4. 1H-NMR spectrum of [Ge(gly)2(H2O)2] in D2O.
4.4 4.3 4.2 4.1 4.0 3.9 3.8 3.7
8.5
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
ppm
3.5
3.0
2.5
2.0
1.5
1.0
ppm
Figure S5. 1H-NMR spectrum of Ge precursor in OLA solution after degassing for 30 min at 120 °C in CDCl3.
60
177 200
180
160
140
Figure S6.
Figure S7.
120
100
80
60
40
20
ppm
13
13
C-NMR spectrum of [Ge(gly)2(H2O)2] in D2O.
C-NMR spectrum of Ge precursor in OLA solution after degassing for
30 min at 120 °C in CDCl3.
7.5
7.0
6.5
6.0
5.5
5.0
2.73 2.69 2.66 2.04 2.01 1.48 1.45 1.41 1.31 1.29 1.27
5.40 5.38 5.36 5.34
7.29 8.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
ppm
1
200
180
160
140
120
42.13 33.80 31.78 29.62 29.58 29.53 29.40 29.38 29.19 29.11 27.04 26.77
129.66 129.57
Figure S8. H-NMR spectrum of degassed OLA in CDCl3.
100
80
60
40
Figure S9. 13C-NMR spectrum of degassed OLA in CDCl3.
20
ppm
30
35
(200)
(111)
(110)
25
(102)
(101) (100)
Instensity (a.u.)
20
40
45
50
55
60
2θ (°) Figure S10. PXRD pattern of GeO2, which was formed during the heating of [Ge(gly)2(H2O)2] to 250 °C in ODE, with key diffraction peaks labelled as designated in ICDD 43-1016.
16000 14000
Counts
12000 10000 8000 6000 4000 2000 0 200
250
300
350
400
450
500
Wavenumber (cm−1) Figure S11. Raman spectra of thin films of as desposited CZGeS NCs.
Intensity (a.u.)
(a)
(b) 25
30
35
40
45
50
55
60
65
2θ (°) Figure S12. PXRD patterns of CZGeS NCs synthesised (a) by a heat-up method with DDT as a sulphide source and (b) by a hot-injection method with S/OLA as a sulphide source. Black calculated patterns for orthorhombic (top) and tetragonal (bottom) added as guides to the eye. Patterns collected using Cu Kα radiation
Intensity (counts)
source.
10
15
20
25
30 2θ (°)
35
40
45
50
Figure S13. Fit of coaddition of calculated orthorhombic and tetragonal phases of CZGeS (red) to the in situ XRD measurement of mixed phase CZGeS NC (blue) and the difference between the calculated and experimental diffraction patterns (black). Pattern collected using a synchrotron radiation source (λ = 0.7286 Å).
Table S1. XPS measurements of CZTGeSSe thin films with different CZGeS / CZTS NC ink compositions.† 100 % Ge /
75 % Ge /
50 % Ge /
25 % Ge /
0 % Ge/
0 % Sn
25 % Sn
50 % Sn
75 % Sn
100 % Sn
Cu 2p
1.57
1.59
1.52
1.83
1.86
Zn 2p
1.00
1.00
1.00
1.00
1.00
Sn 4d
0.00
0.20
0.31
0.57
0.98
Ge 3d
0.62
0.39
0.18
0.10
0.00
Sn / (Sn + Ge)
0.00
0.33
0.63
0.85
1.00
† Compositions normalized against zinc composition
Discussion of CZGeS NC and CZTGeSSe thin film Raman spectra analysis While XRD can be used for the phase determination of quaternary chalcogenide NCs, it does not allow for differentiation between the target product and a number of binary and ternary nanocrystalline impurities. This problem is exemplified by the CZTS system, in which the presence of impurities such as ZnS and CuSnS3 cannot be confirmed due to the coincidence of their peaks with CZTS in PXRD patterns. 2, 3 Therefore, in order to characterize these systems adequately, Raman spectroscopy must be employed as it has the ability to discriminate between the different metal chalcogenides. The Raman spectrum of the CZGeS NCs reveals a main peak at 358 cm−1, which corresponds well with the main peak observed in bulk CZGeS (Figure S11).4,
5
Examination of the weaker peaks is complicated by the Raman peak
broadening in nanocrystalline materials,6-8 which has been noted to also occur in the CZTS system.9,
10
This phenomenon can also be observed when comparing the
Raman spectra of nanocrystalline and highly crystalline co-evaporated Cu2ZnGeSe4 (CZGeSe).11 The Raman spectrum of the synthesized CZGeS NCs is comparable to the measured spectra of CZGeS thin films formed by chemical spray pyrolysis.12 The wide shoulder present at ~300 cm−1 may be due to the broadening of the 272 and 291 cm−1 present in monocrystalline CZGeS,5 while the two shoulders at 385 and 407 cm−1 are coincident with distinct peaks observed in the Raman spectrum of monocrystalline CZGeS.5 As CuS, Cu2S and GeS have major peaks in their Raman spectra at 474, 472 and 238 cm−1, respectively, their absence can be confirmed by this technique.13, 14 Although GeS2 has a major peak in its Raman spectrum at 342 cm−1, it has peaks present in its PXRD pattern that are not coincident with those of CuZnGeS4, so its presence could be excluded on the basis of prior analysis. 15 It should be noted that Raman peaks for Cu2GeS3 (334, 387 cm−1) and ZnS (351 cm−1) occur at similar wavenumbers to CZGeS, thereby precluding the use of this technique from excluding their presence.5, 16, 17 The three primary peaks in the Raman spectra of the CZTSSe thin film (174, 197 and 238 cm−1) correspond with the peak values previously reported for pure CZTSe thin films, with the remaining peak at 331 cm−1 attributed to the two-mode behaviour observed in the Raman spectrum of single phase CZTSSe.18,
19
Similarly, the two
primary peaks at 175 and 199 cm−1 in the spectra of the CZTGeSSe (50:50 Sn/Ge) thin film match those observed in similar CZTGeSSe thin films reported in the
literature,20 with a shift to higher wavenumbers corresponding to a proportional increase in the germanium content (Figure S14).21 The peaks at 331 and 352 cm−1 in the spectra of CZTSSe and CZGeSSe, respectively, or the presence of both peaks in mixed Sn/Ge thin films, can be attributed to a two-mode behavior. A similar phenomenon was observed in a prior report, in which a minor, broad peak (~327 cm−1) in the Raman spectrum of a Cu2ZnSnS0.25Se0.75 solid was present at a lower wavenumber than that of the pure CZTS A1 peak (338 cm-1).[11a] While the two main peaks in the Raman spectra of the CZGeSSe thin film (180 and 206 cm−1) correlate well with the reported values for CZGeSe thin films, 11 there remain two distinct features of the spectra that cannot be as readily explained; a shoulder at ~224 cm−1 and a broad, minor peak at 275 cm −1. The latter peak has previously been observed in Cu2ZnGeSe4 thin films formed by spray pyrolysis,12 and was reported to be due to the presence of Cu2Se (260 cm−1),13 CuSe (263 cm−1)22 and CuSe2 (260 m−1).23 The presence of a small shoulder on the 112 peak in the diffraction pattern of CZGeSSe supports that a minor impurity phase may form during the selenization of this sample. The source of the shoulder on the primary peak could not be identified as amorphous selenium (250 cm−1),24 ZnSe (205, 250 cm−1),25 or Cu2GeSe3 (189, 266, 300 cm−1).26 However, GeSe2 has Raman bands at 211 and 216 cm−1 that could contributing to the formation of this shoulder. 27,28 This attribution is supported by the fact that the intensity of this shoulder increases with elevated levels of germanium present in the deposition ink. As the XRD patterns of the alloyed thin films do not include contributions from copper and germanium selenide species, it would suggest the contaminants may be localized at the surface and could be removed with additional etching treatments. Furthermore, the use of indepth resolved Raman spectroscopy could allow for the identification of secondary phases that are localised to the surface of the thin film.
197
Normalized Intensity (a.u.)
174
238
198
(a)
331
174 240 331
199
(b) 175
219 202
(c)
331 220
176 206
(d)
352 224
180
275
354
(e)
100
150
200
250
300
350
400
Wavenumber (cm−1) Figure S14. Raman spectra of CZTGeSSe thin films: (a) 100/0 Sn:Ge; (b) 75/25 Sn:Ge; (c) 50/50 Sn:Ge; (d) 25/75 Sn:Ge; (e) 0/100 Sn:Ge.
Discussion of formation mechanism of CZGeS NCs Examination of the evolution of CZGeS NCs was divided into two components; analysis of nuclei at lower temperatures (100 – 250 °C) by UV-vis spectroscopy (Figures S15 and S16) and analysis of the change in the nanocrystal composition once at the growth temperature by XRF (Table S2). It was not possible to analyse the nuclei formed at temperatures below 250 °C by XRF as they could not be satisfactorily isolated from the reaction solution and purified due to the similarity in solubility between the small nuclei and the precursor/monomer. Inspection of the UV-vis absorption spectra of CZGeS during the heat up process shows that at 100 °C the nascent nuclei exhibit a spectrum with a broad shoulder at ~480 nm and an onset ~570 nm. During heating the spectra gradually shift to lower energies. At 250 °C the spectrum shows almost no shoulder and an absorption onset at ~800 nm. We attribute this red-shift to the growth of the NCs. The most plausible interpretation of this data is that small CZGeS nuclei form (from the decomposition of the highly reactive dithiocarbamate-based precursors), which red-shift as they grow as a consequence of quantum confinement.29 Notably, a similar process is observed in quantum confined CZTS NCs. 30-32 Unfortunately, it was not possible for us to deconvolute this process with the effect a possible change in elemental composition would have on the absorption profile, which is anticipated to occur concurrently, as the spectroscopic effects of the latter process are yet to be comprehensively studied. During this synthesis of CZGeS NCs it is possible to exclude the formation of Cu2-xS nuclei, followed by the later incorporation of Ge and Zn into the nascent NCs, as sub-stoichiometric Cu2-xS NCs display a strong plasmon,29 which was not observed in the spectra of CZGeS during heat-up. Furthermore, pure Cu2S (chalcocite) NCs show a absorption onset at lower energies, even with quantum confinement.33 The only remaining possibility is that nucleation primarily involves the formation of quantum confined CGeS NCs. However, the absorbance profiles observed at early reaction times for the pure CGeS system (Figure S16) appear radically different from those of CZGeS, suggesting an entirely different process of formation in the absence of zinc precursors. Raman spectroscopy was not able to provide insight into the elemental evolution of CZGeS NCs as the material isolated at lower temperatures did not give a strong signal (Figure S17).
Aborption (a.u.)
2
1.8
100 C
1.6
140 C
1.4
180 C
1.2
220 C
1
250 C
0.8 0.6 0.4 0.2 0 350
450
550 650 Wavelength (nm)
750
850
Figure S15. UV-vis spectra of aliquots of the reaction solution of CZGeS NCs in CHCl3 at various temperatures; (inset) Solutions of reaction solution in CHCl3 collected at (A) 100 °C, (B) 140 °C, (C) 180 °C and (D) 250 °C.
Absorption (normalised)
100 °C 140 °C 180 °C 220 °C 250 °C 0 min 250 °C 10 min 250 °C 20 min 250 °C 30 min
400
600
800
1000 1200 Wavelength (nm)
1400
1600
Figure S16. UV-vis spectra of aliquots of the reaction solution of Cu2GeS3 NCs in CHCl3 at various temperatures.
Intensity (normalised)
100 °C 150 °C 200 °C 250 °C 0 min 250 °C 15 min 250 °C 30 min 250 °C 60 min
200
250
300
350
400
450
500
550
Wavenumber (cm−1)
Figure S17. Raman spectra of aliquots of CZGeS NCs isolated at various stages of the reaction.
XRF measurements of isolated CZGeS NCs show that during the growth stage of the reaction (250 °C), copper is incorporated into the NCs faster than germanium and zinc, and germanium incorporates into the NCs faster than zinc (Table S2). Determination of the exact mechanism by which the different metal sources react during the nucleation and growth stages of the reaction is beyond the scope of this current study.
Table S2. Composition of CZGeS NCs at various times when heated at 250 °C as measured by XRF (normalised to Zn content).
Time
Cu
Zn
Ge
S
15 min
1.42
1.00
0.76
3.39
30 min
1.51
1.00
0.83
3.71
60 min
1.73
1.00
0.88
3.75
References (1)
Chiang, H.-C.; Wang, M.-H.; Ueng, C.-H. Acta Crystallogr., Sect. C 1993, 49,
244-246. (2)
Fontane, X.; Calvo-Barrio, L.; Izquierdo-Roca, V.; Saucedo, E.; Perez-
Rodriguez, A.; Morante, J. R.; Berg, D. M.; Dale, P. J.; Siebentritt, S. Appl. Phys. Lett. 2011, 98, 181905. (3)
Wang, K. J.; Shin, B.; Reuter, K. B.; Todorov, T.; Mitzi, D. B.; Guha, S. Appl.
Phys. Lett. 2011, 98, 051912. (4)
Levcenko, S.; Guc, M.; Merschjann, C.; Gurieva, G.; Schorr, S.; Lux-Steiner,
M.; Arushanov, E. Phys. Status Solidi C 2013, 10, 1079-1081. (5)
Guc, M.; Izquierdo-Roca, V.; Pérez Rodríguez, A.; Gurieva, G.; Levcenko, S.;
Schorr, S.; Arushanov, E. Phys. Status Solidi C 2013, 10, 1075-1078. (6)
Bottani, C. E.; Mantini, C.; Milani, P.; Manfredini, M.; Stella, A.; Tognini, P.;
Cheyssac, P.; Kofman, R. Appl. Phys. Lett. 1996, 69, 2409-2411. (7)
Soni, R. K.; Fonseca, L. F.; Resto, O.; Buzaianu, M.; Weisz, S. Z. J. Lumin.
1999, 83–84, 187-191. (8)
Khoo, K. H.; Zayak, A. T.; Kwak, H.; Chelikowsky, J. R. Phys. Rev. Lett. 2010,
105, 115504. (9)
Sarswat, P. K.; Free, M. L. J. Cryst. Growth 2013, 372, 87-94.
(10)
Liu, W. C.; Guo, B. L.; Wu, X. S.; Zhang, F. M.; Mak, C. L.; Wong, K. H. J.
Mater. Chem. A 2013, 1, 3182-3186. (11)
Uday Bhaskar, P.; Suresh Babu, G.; Kishore Kumar, Y. B.; Sundara Raja, V.
Thin Solid Films 2013, 534, 249-254. (12)
Khadka, D. B.; Kim, J. CrystEngComm 2013, 15, 10500-10509.
(13)
Minceva-Sukarova, B.; Najdoski, M.; Grozdanov, I.; Chunnilall, C. J. J. Mol.
Struct. 1997, 410, 267-270. (14)
Wiley, J. D.; Buckel, W. J.; Schmidt, R. L. Phys. Rev. B 1976, 13, 2489-2496.
(15)
Málek, J. i. J. Non-Cryst. Solids 1989, 107, 323-327.
(16)
Neumann, H.; Riede, V.; Sharif, N.; Sobotta, H.; Omar, M. S. J. Mater. Sci.
Lett. 1989, 8, 1360-1361. (17)
Brafman, O.; Mitra, S. S. Phys. Rev. 1968, 171, 931-934.
(18)
Grossberg, M.; Krustok, J.; Raudoja, J.; Timmo, K.; Altosaar, M.; Raadik, T.
Thin Solid Films 2011, 519, 7403-7406. (19)
van Embden, J.; Chesman, A. S. R.; Della Gaspera, E.; Duffy, N. W.; Watkins,
S. E.; Jasieniak, J. J. J. Am. Chem. Soc. 2014, 136, 5237-5240. (20)
Hages, C. J.; Levcenco, S.; Miskin, C. K.; Alsmeier, J. H.; Abou-Ras, D.;
Wilks, R. G.; Bär, M.; Unold, T.; Agrawal, R. Prog. Photovolt: Res. Appl. 2013, DOI: 10.1002/pip.2442. (21)
Zong, K.; Lu, S.; Wang, H.; Sun, Y.; Zheng, H.; Liu, J.; Yan, H.
CrystEngComm 2013, 15, 6942. (22)
Ishii, M.; Shibata, K.; Nozaki, H. J. Solid State Chem. 1993, 105, 504-511.
(23)
Anastassakis, E. Solid State Commun. 1973, 13, 1297-1301.
(24)
Carroll, P. J.; Lannin, J. S. Solid State Commun. 1981, 40, 81-84.
(25)
Taylor, W. Phys. Lett. A 1967, 24, 556-558.
(26)
Marcano, G.; Rincón, C.; Marín, G.; Delgado, G. E.; Mora, A. J.; Herrera-
Pérez, J. L.; Mendoza-Alvarez, J. G.; Rodríguez, P. Solid State Commun. 2008, 146, 65-68. (27)
Matsuda, O.; Inoue, K.; Murase, K. Solid State Commun. 1990, 75, 303-308.
(28)
While the 211 cm−1 Raman band is more intense than the 216 cm−1 band in
crystalline GeSe2, the relative intensity of the latter band increases as disorder in the crystal increases. See Matsuda et al. for further discussion. (29)
Zhao, Y.; Pan, H.; Lou, Y.; Qiu, X.; Zhu, J.; Burda, C. J. Am. Chem. Soc.
2009, 131, 4253-4261. (30)
Khare, A.; Wills, A. W.; Ammerman, L. M.; Norris, D. J.; Aydil, E. S. Chem.
Commun. 2011, 47, 11721-11723. (31)
Chernomordik, B. D.; Beland, A. E.; Trejo, N. D.; Gunawan, A. A.; Deng, D.
D.; Mkhoyan, K. A.; Aydil, E. S. J. Mater. Chem. A 2014, 2, 10389-10395.
(32)
Arul, N. S.; Yun, D. Y.; Lee, D. U.; Kim, T. W. Nanoscale 2013, 5, 11940-
11943. (33) 366.
Luther, J. M.; Jain, P. K.; Ewers, T.; Alivisatos, A. P. Nat Mater 2011, 10, 361-
Cu2ZnGeS4 Nanocrystals from Air-Stable Precursors for Sintered Thin Film Alloys
Anthony S. R. Chesman,*† Joel van Embden,† Enrico Della Gaspera, † Noel W. Duffy,† Nathan A. S. Webster, ‡, Jacek J. Jasieniak*†
† ‡
CSIRO Manufacturing Flagship, Bayview Avenue, Clayton, Victoria 3168, Australia. CSIRO Mineral Resources Flagship, Bayview Avenue, Clayton, Victoria 3168, Australia.
*E-mail: [email protected] and [email protected].
(a)
(b)
Figure S1. (a) Colorless plate crystals of [Ge(gly)2(H2O)2]; (b) Non-corrosive, air stable crystals of [Ge(gly)2(H2O)2] on a glove.
100 90
[Ge(gly)2(H2O)2] [Ge(gly)2] (Calc. 86 %, Found 86 %)
Mass (%)
80 70 60 50
GeO2 (Calc. 41 %, Found 42 %)
40 30 20 50
100
150
200
250
300
350
Temperature (°C)
Figure S2. TGA thermogram of decomposition of [Ge(gly)2(H2O)2] to GeO2.
Intensity (a.u., offset)
[Ge(gly)2(H2O)2]
C=O
Ge-OC
[Ge(gly)2(OLA)2]
Oleylamine 3600
3100
2600
2100
1600
1100
600
Wavenumber (cm−1) Figure S3. IR spectra of [Ge(gly)2(H2O)2], [Ge(gly)2(OLA)2] in OLA and neat OLA, highlighting the similar spectral features of the glycolate ligand. 1
8
7
6
4.11
4.79 9
5
4
3
2
1
ppm
2.72 2.69 2.65 2.01 1.98 1.45 1.42 1.28 1.26 1.25 0.90 0.87
3.98
5.35 5.33 5.30 3.98
7.29
Figure S4. 1H-NMR spectrum of [Ge(gly)2(H2O)2] in D2O.
4.4 4.3 4.2 4.1 4.0 3.9 3.8 3.7
8.5
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
ppm
3.5
3.0
2.5
2.0
1.5
1.0
ppm
Figure S5. 1H-NMR spectrum of Ge precursor in OLA solution after degassing for 30 min at 120 °C in CDCl3.
60
177 200
180
160
140
Figure S6.
Figure S7.
120
100
80
60
40
20
ppm
13
13
C-NMR spectrum of [Ge(gly)2(H2O)2] in D2O.
C-NMR spectrum of Ge precursor in OLA solution after degassing for
30 min at 120 °C in CDCl3.
7.5
7.0
6.5
6.0
5.5
5.0
2.73 2.69 2.66 2.04 2.01 1.48 1.45 1.41 1.31 1.29 1.27
5.40 5.38 5.36 5.34
7.29 8.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
ppm
1
200
180
160
140
120
42.13 33.80 31.78 29.62 29.58 29.53 29.40 29.38 29.19 29.11 27.04 26.77
129.66 129.57
Figure S8. H-NMR spectrum of degassed OLA in CDCl3.
100
80
60
40
Figure S9. 13C-NMR spectrum of degassed OLA in CDCl3.
20
ppm
30
35
(200)
(111)
(110)
25
(102)
(101) (100)
Instensity (a.u.)
20
40
45
50
55
60
2θ (°) Figure S10. PXRD pattern of GeO2, which was formed during the heating of [Ge(gly)2(H2O)2] to 250 °C in ODE, with key diffraction peaks labelled as designated in ICDD 43-1016.
16000 14000
Counts
12000 10000 8000 6000 4000 2000 0 200
250
300
350
400
450
500
Wavenumber (cm−1) Figure S11. Raman spectra of thin films of as desposited CZGeS NCs.
Intensity (a.u.)
(a)
(b) 25
30
35
40
45
50
55
60
65
2θ (°) Figure S12. PXRD patterns of CZGeS NCs synthesised (a) by a heat-up method with DDT as a sulphide source and (b) by a hot-injection method with S/OLA as a sulphide source. Black calculated patterns for orthorhombic (top) and tetragonal (bottom) added as guides to the eye. Patterns collected using Cu Kα radiation
Intensity (counts)
source.
10
15
20
25
30 2θ (°)
35
40
45
50
Figure S13. Fit of coaddition of calculated orthorhombic and tetragonal phases of CZGeS (red) to the in situ XRD measurement of mixed phase CZGeS NC (blue) and the difference between the calculated and experimental diffraction patterns (black). Pattern collected using a synchrotron radiation source (λ = 0.7286 Å).
Table S1. XPS measurements of CZTGeSSe thin films with different CZGeS / CZTS NC ink compositions.† 100 % Ge /
75 % Ge /
50 % Ge /
25 % Ge /
0 % Ge/
0 % Sn
25 % Sn
50 % Sn
75 % Sn
100 % Sn
Cu 2p
1.57
1.59
1.52
1.83
1.86
Zn 2p
1.00
1.00
1.00
1.00
1.00
Sn 4d
0.00
0.20
0.31
0.57
0.98
Ge 3d
0.62
0.39
0.18
0.10
0.00
Sn / (Sn + Ge)
0.00
0.33
0.63
0.85
1.00
† Compositions normalized against zinc composition
Discussion of CZGeS NC and CZTGeSSe thin film Raman spectra analysis While XRD can be used for the phase determination of quaternary chalcogenide NCs, it does not allow for differentiation between the target product and a number of binary and ternary nanocrystalline impurities. This problem is exemplified by the CZTS system, in which the presence of impurities such as ZnS and CuSnS3 cannot be confirmed due to the coincidence of their peaks with CZTS in PXRD patterns. 2, 3 Therefore, in order to characterize these systems adequately, Raman spectroscopy must be employed as it has the ability to discriminate between the different metal chalcogenides. The Raman spectrum of the CZGeS NCs reveals a main peak at 358 cm−1, which corresponds well with the main peak observed in bulk CZGeS (Figure S11).4,
5
Examination of the weaker peaks is complicated by the Raman peak
broadening in nanocrystalline materials,6-8 which has been noted to also occur in the CZTS system.9,
10
This phenomenon can also be observed when comparing the
Raman spectra of nanocrystalline and highly crystalline co-evaporated Cu2ZnGeSe4 (CZGeSe).11 The Raman spectrum of the synthesized CZGeS NCs is comparable to the measured spectra of CZGeS thin films formed by chemical spray pyrolysis.12 The wide shoulder present at ~300 cm−1 may be due to the broadening of the 272 and 291 cm−1 present in monocrystalline CZGeS,5 while the two shoulders at 385 and 407 cm−1 are coincident with distinct peaks observed in the Raman spectrum of monocrystalline CZGeS.5 As CuS, Cu2S and GeS have major peaks in their Raman spectra at 474, 472 and 238 cm−1, respectively, their absence can be confirmed by this technique.13, 14 Although GeS2 has a major peak in its Raman spectrum at 342 cm−1, it has peaks present in its PXRD pattern that are not coincident with those of CuZnGeS4, so its presence could be excluded on the basis of prior analysis. 15 It should be noted that Raman peaks for Cu2GeS3 (334, 387 cm−1) and ZnS (351 cm−1) occur at similar wavenumbers to CZGeS, thereby precluding the use of this technique from excluding their presence.5, 16, 17 The three primary peaks in the Raman spectra of the CZTSSe thin film (174, 197 and 238 cm−1) correspond with the peak values previously reported for pure CZTSe thin films, with the remaining peak at 331 cm−1 attributed to the two-mode behaviour observed in the Raman spectrum of single phase CZTSSe.18,
19
Similarly, the two
primary peaks at 175 and 199 cm−1 in the spectra of the CZTGeSSe (50:50 Sn/Ge) thin film match those observed in similar CZTGeSSe thin films reported in the
literature,20 with a shift to higher wavenumbers corresponding to a proportional increase in the germanium content (Figure S14).21 The peaks at 331 and 352 cm−1 in the spectra of CZTSSe and CZGeSSe, respectively, or the presence of both peaks in mixed Sn/Ge thin films, can be attributed to a two-mode behavior. A similar phenomenon was observed in a prior report, in which a minor, broad peak (~327 cm−1) in the Raman spectrum of a Cu2ZnSnS0.25Se0.75 solid was present at a lower wavenumber than that of the pure CZTS A1 peak (338 cm-1).[11a] While the two main peaks in the Raman spectra of the CZGeSSe thin film (180 and 206 cm−1) correlate well with the reported values for CZGeSe thin films, 11 there remain two distinct features of the spectra that cannot be as readily explained; a shoulder at ~224 cm−1 and a broad, minor peak at 275 cm −1. The latter peak has previously been observed in Cu2ZnGeSe4 thin films formed by spray pyrolysis,12 and was reported to be due to the presence of Cu2Se (260 cm−1),13 CuSe (263 cm−1)22 and CuSe2 (260 m−1).23 The presence of a small shoulder on the 112 peak in the diffraction pattern of CZGeSSe supports that a minor impurity phase may form during the selenization of this sample. The source of the shoulder on the primary peak could not be identified as amorphous selenium (250 cm−1),24 ZnSe (205, 250 cm−1),25 or Cu2GeSe3 (189, 266, 300 cm−1).26 However, GeSe2 has Raman bands at 211 and 216 cm−1 that could contributing to the formation of this shoulder. 27,28 This attribution is supported by the fact that the intensity of this shoulder increases with elevated levels of germanium present in the deposition ink. As the XRD patterns of the alloyed thin films do not include contributions from copper and germanium selenide species, it would suggest the contaminants may be localized at the surface and could be removed with additional etching treatments. Furthermore, the use of indepth resolved Raman spectroscopy could allow for the identification of secondary phases that are localised to the surface of the thin film.
197
Normalized Intensity (a.u.)
174
238
198
(a)
331
174 240 331
199
(b) 175
219 202
(c)
331 220
176 206
(d)
352 224
180
275
354
(e)
100
150
200
250
300
350
400
Wavenumber (cm−1) Figure S14. Raman spectra of CZTGeSSe thin films: (a) 100/0 Sn:Ge; (b) 75/25 Sn:Ge; (c) 50/50 Sn:Ge; (d) 25/75 Sn:Ge; (e) 0/100 Sn:Ge.
Discussion of formation mechanism of CZGeS NCs Examination of the evolution of CZGeS NCs was divided into two components; analysis of nuclei at lower temperatures (100 – 250 °C) by UV-vis spectroscopy (Figures S15 and S16) and analysis of the change in the nanocrystal composition once at the growth temperature by XRF (Table S2). It was not possible to analyse the nuclei formed at temperatures below 250 °C by XRF as they could not be satisfactorily isolated from the reaction solution and purified due to the similarity in solubility between the small nuclei and the precursor/monomer. Inspection of the UV-vis absorption spectra of CZGeS during the heat up process shows that at 100 °C the nascent nuclei exhibit a spectrum with a broad shoulder at ~480 nm and an onset ~570 nm. During heating the spectra gradually shift to lower energies. At 250 °C the spectrum shows almost no shoulder and an absorption onset at ~800 nm. We attribute this red-shift to the growth of the NCs. The most plausible interpretation of this data is that small CZGeS nuclei form (from the decomposition of the highly reactive dithiocarbamate-based precursors), which red-shift as they grow as a consequence of quantum confinement.29 Notably, a similar process is observed in quantum confined CZTS NCs. 30-32 Unfortunately, it was not possible for us to deconvolute this process with the effect a possible change in elemental composition would have on the absorption profile, which is anticipated to occur concurrently, as the spectroscopic effects of the latter process are yet to be comprehensively studied. During this synthesis of CZGeS NCs it is possible to exclude the formation of Cu2-xS nuclei, followed by the later incorporation of Ge and Zn into the nascent NCs, as sub-stoichiometric Cu2-xS NCs display a strong plasmon,29 which was not observed in the spectra of CZGeS during heat-up. Furthermore, pure Cu2S (chalcocite) NCs show a absorption onset at lower energies, even with quantum confinement.33 The only remaining possibility is that nucleation primarily involves the formation of quantum confined CGeS NCs. However, the absorbance profiles observed at early reaction times for the pure CGeS system (Figure S16) appear radically different from those of CZGeS, suggesting an entirely different process of formation in the absence of zinc precursors. Raman spectroscopy was not able to provide insight into the elemental evolution of CZGeS NCs as the material isolated at lower temperatures did not give a strong signal (Figure S17).
Aborption (a.u.)
2
1.8
100 C
1.6
140 C
1.4
180 C
1.2
220 C
1
250 C
0.8 0.6 0.4 0.2 0 350
450
550 650 Wavelength (nm)
750
850
Figure S15. UV-vis spectra of aliquots of the reaction solution of CZGeS NCs in CHCl3 at various temperatures; (inset) Solutions of reaction solution in CHCl3 collected at (A) 100 °C, (B) 140 °C, (C) 180 °C and (D) 250 °C.
Absorption (normalised)
100 °C 140 °C 180 °C 220 °C 250 °C 0 min 250 °C 10 min 250 °C 20 min 250 °C 30 min
400
600
800
1000 1200 Wavelength (nm)
1400
1600
Figure S16. UV-vis spectra of aliquots of the reaction solution of Cu2GeS3 NCs in CHCl3 at various temperatures.
Intensity (normalised)
100 °C 150 °C 200 °C 250 °C 0 min 250 °C 15 min 250 °C 30 min 250 °C 60 min
200
250
300
350
400
450
500
550
Wavenumber (cm−1)
Figure S17. Raman spectra of aliquots of CZGeS NCs isolated at various stages of the reaction.
XRF measurements of isolated CZGeS NCs show that during the growth stage of the reaction (250 °C), copper is incorporated into the NCs faster than germanium and zinc, and germanium incorporates into the NCs faster than zinc (Table S2). Determination of the exact mechanism by which the different metal sources react during the nucleation and growth stages of the reaction is beyond the scope of this current study.
Table S2. Composition of CZGeS NCs at various times when heated at 250 °C as measured by XRF (normalised to Zn content).
Time
Cu
Zn
Ge
S
15 min
1.42
1.00
0.76
3.39
30 min
1.51
1.00
0.83
3.71
60 min
1.73
1.00
0.88
3.75
References (1)
Chiang, H.-C.; Wang, M.-H.; Ueng, C.-H. Acta Crystallogr., Sect. C 1993, 49,
244-246. (2)
Fontane, X.; Calvo-Barrio, L.; Izquierdo-Roca, V.; Saucedo, E.; Perez-
Rodriguez, A.; Morante, J. R.; Berg, D. M.; Dale, P. J.; Siebentritt, S. Appl. Phys. Lett. 2011, 98, 181905. (3)
Wang, K. J.; Shin, B.; Reuter, K. B.; Todorov, T.; Mitzi, D. B.; Guha, S. Appl.
Phys. Lett. 2011, 98, 051912. (4)
Levcenko, S.; Guc, M.; Merschjann, C.; Gurieva, G.; Schorr, S.; Lux-Steiner,
M.; Arushanov, E. Phys. Status Solidi C 2013, 10, 1079-1081. (5)
Guc, M.; Izquierdo-Roca, V.; Pérez Rodríguez, A.; Gurieva, G.; Levcenko, S.;
Schorr, S.; Arushanov, E. Phys. Status Solidi C 2013, 10, 1075-1078. (6)
Bottani, C. E.; Mantini, C.; Milani, P.; Manfredini, M.; Stella, A.; Tognini, P.;
Cheyssac, P.; Kofman, R. Appl. Phys. Lett. 1996, 69, 2409-2411. (7)
Soni, R. K.; Fonseca, L. F.; Resto, O.; Buzaianu, M.; Weisz, S. Z. J. Lumin.
1999, 83–84, 187-191. (8)
Khoo, K. H.; Zayak, A. T.; Kwak, H.; Chelikowsky, J. R. Phys. Rev. Lett. 2010,
105, 115504. (9)
Sarswat, P. K.; Free, M. L. J. Cryst. Growth 2013, 372, 87-94.
(10)
Liu, W. C.; Guo, B. L.; Wu, X. S.; Zhang, F. M.; Mak, C. L.; Wong, K. H. J.
Mater. Chem. A 2013, 1, 3182-3186. (11)
Uday Bhaskar, P.; Suresh Babu, G.; Kishore Kumar, Y. B.; Sundara Raja, V.
Thin Solid Films 2013, 534, 249-254. (12)
Khadka, D. B.; Kim, J. CrystEngComm 2013, 15, 10500-10509.
(13)
Minceva-Sukarova, B.; Najdoski, M.; Grozdanov, I.; Chunnilall, C. J. J. Mol.
Struct. 1997, 410, 267-270. (14)
Wiley, J. D.; Buckel, W. J.; Schmidt, R. L. Phys. Rev. B 1976, 13, 2489-2496.
(15)
Málek, J. i. J. Non-Cryst. Solids 1989, 107, 323-327.
(16)
Neumann, H.; Riede, V.; Sharif, N.; Sobotta, H.; Omar, M. S. J. Mater. Sci.
Lett. 1989, 8, 1360-1361. (17)
Brafman, O.; Mitra, S. S. Phys. Rev. 1968, 171, 931-934.
(18)
Grossberg, M.; Krustok, J.; Raudoja, J.; Timmo, K.; Altosaar, M.; Raadik, T.
Thin Solid Films 2011, 519, 7403-7406. (19)
van Embden, J.; Chesman, A. S. R.; Della Gaspera, E.; Duffy, N. W.; Watkins,
S. E.; Jasieniak, J. J. J. Am. Chem. Soc. 2014, 136, 5237-5240. (20)
Hages, C. J.; Levcenco, S.; Miskin, C. K.; Alsmeier, J. H.; Abou-Ras, D.;
Wilks, R. G.; Bär, M.; Unold, T.; Agrawal, R. Prog. Photovolt: Res. Appl. 2013, DOI: 10.1002/pip.2442. (21)
Zong, K.; Lu, S.; Wang, H.; Sun, Y.; Zheng, H.; Liu, J.; Yan, H.
CrystEngComm 2013, 15, 6942. (22)
Ishii, M.; Shibata, K.; Nozaki, H. J. Solid State Chem. 1993, 105, 504-511.
(23)
Anastassakis, E. Solid State Commun. 1973, 13, 1297-1301.
(24)
Carroll, P. J.; Lannin, J. S. Solid State Commun. 1981, 40, 81-84.
(25)
Taylor, W. Phys. Lett. A 1967, 24, 556-558.
(26)
Marcano, G.; Rincón, C.; Marín, G.; Delgado, G. E.; Mora, A. J.; Herrera-
Pérez, J. L.; Mendoza-Alvarez, J. G.; Rodríguez, P. Solid State Commun. 2008, 146, 65-68. (27)
Matsuda, O.; Inoue, K.; Murase, K. Solid State Commun. 1990, 75, 303-308.
(28)
While the 211 cm−1 Raman band is more intense than the 216 cm−1 band in
crystalline GeSe2, the relative intensity of the latter band increases as disorder in the crystal increases. See Matsuda et al. for further discussion. (29)
Zhao, Y.; Pan, H.; Lou, Y.; Qiu, X.; Zhu, J.; Burda, C. J. Am. Chem. Soc.
2009, 131, 4253-4261. (30)
Khare, A.; Wills, A. W.; Ammerman, L. M.; Norris, D. J.; Aydil, E. S. Chem.
Commun. 2011, 47, 11721-11723. (31)
Chernomordik, B. D.; Beland, A. E.; Trejo, N. D.; Gunawan, A. A.; Deng, D.
D.; Mkhoyan, K. A.; Aydil, E. S. J. Mater. Chem. A 2014, 2, 10389-10395.
(32)
Arul, N. S.; Yun, D. Y.; Lee, D. U.; Kim, T. W. Nanoscale 2013, 5, 11940-
11943. (33) 366.
Luther, J. M.; Jain, P. K.; Ewers, T.; Alivisatos, A. P. Nat Mater 2011, 10, 361-
