- Email: [email protected]
Ellipsometric characterization of MoSe2 thin layers obtained by thermal treatment of molybdenum in selenium vapor
Accepted Manuscript Title: Ellipsometric characterization of MoSe2 thin layers obtained by thermal treatment of molybdenum in selenium vapor Authors: Ayaz Bayramov, Yegana Aliyeva, Gurban Eyyubov, Eldar Mammadov, Zakir Jahangirli, Daniel Lincot, Nazim Mamedov PII: DOI: Reference:
S0169-4332(17)30165-4 http://dx.doi.org/doi:10.1016/j.apsusc.2017.01.153 APSUSC 34948
To appear in:
APSUSC
Received date: Revised date: Accepted date:
31-7-2016 18-12-2016 16-1-2017
Please cite this article as: Ayaz Bayramov, Yegana Aliyeva, Gurban Eyyubov, Eldar Mammadov, Zakir Jahangirli, Daniel Lincot, Nazim Mamedov, Ellipsometric characterization of MoSe2 thin layers obtained by thermal treatment of molybdenum in selenium vapor, Applied Surface Science http://dx.doi.org/10.1016/j.apsusc.2017.01.153 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Ellipsometric characterization of MoSe2 thin layers obtained by thermal treatment of molybdenum in selenium vapor.
Ayaz Bayramova, Yegana Aliyevaa, Gurban Eyyubova, Eldar Mammadova, Zakir Jahangirlia, Daniel Lincotb, Nazim Mamedova a
Institute of Physics, Azerbaijan National Academy of Sciences, H. Javid ave.131, AZ1143 Baku, Azerbaijan b
Institut de Recherche et Développement sur l'Energie Photovoltaïque (IRDEP), 6 quai Watier, 78401 CHATOU Cedex, Paris, France
Highlights
MoSe2 layers are obtained by thermal treatment of molybdenum in selenium vapor. Spectroscopic ellipsometry is applied to the obtained layers and MoSe 2 target. Electronic band structure of MoSe2 is calculated and dielectric function is derived. Calculated and ellipsometry-based data on dielectric function agree fairly well. Excitonic transitions are assumed to form the dielectric function at around 1 eV.
Abstract Submicron MoSe2 layers were prepared by thermal treatment of thick Mo layers on glass substrate in saturated selenium vapor. Spectroscopic ellipsometry was then applied to the obtained MoSe2/Mo/Glass structures and MoSe2 target sample at room temperature. Dielectric function for both the MoSe2 layer and MoSe2 target was retrieved in the spectral range 1901700 nm by using the Kramers-Kronig consistent B-spline dispersion model. The obtained data were similar in both cases. Despite apparent red shift of the dielectric function spectra of the layer in high energy region the peculiarity at around 1 eV is manifested at the same energy for both, layer and target. Comparison of the ellipsometry-based dielectric function of the target and the one, obtained within calculated band structure of MoSe2 for room temperature lattice parameters, has shown that the former is a broadened counterpart of the latter. Abovementioned peculiar feature is not reproduced in the calculated dielectric function and is assumed to have excitonic nature.
Keywords: MoSe2 layer, spectroscopic ellipsometry, dielectric function
1. Introduction MX2 (where M = Mo, X = S, Se, and Te), indirect bandgap layered materials have attracted significant interest for various optoelectronic applications such as phototransistors, light-emitting and photovoltaic devices [1-4]. Particularly, MoSe2 is considered as a good low cost candidate for stable Cu-free back contact in CdTe-based thin film solar cells. The valence band offset of CdTe/MoSe2 interface is about 0.03 eV, which is much suitable for the holes transport between p-CdTe and MoSe2 [5]. Depending on the stacking manner of the layers in the unit cell, MoSe2 crystallizes in 2H or 3R polyform [6, 7]. The former has two layers stacked in hexagonal symmetry while the latter has three layers in the unit cell and
rhombohedral symmetry. The band gap of MoSe2 has been reported in the 1.1-1.42 eV range depending on preparation techniques and conditions [3, 8-10]. Several techniques, such as pulsed electrodeposition, chemical transport and chemical vapor deposition methods have been used for MoSe2 polycrystalline thin film growth [3, 9, 11]. However, a simple and inexpensive technique compatible with the process of substrate-configured CdTe-based thin film solar cell fabrication is in demand. In this regard, thermal annealing of molybdenum film on a glass substrate under selenium vapor is of interest. On the other hand, there is a lack of systematic studies of properties of the films prepared by this method. In this work, we present the results of ellipsometric studies of MoSe2 thin layers prepared by thermal treatment of molybdenum film in selenium vapor along with calculated band structure and dielectric function of MoSe2.
2. Experimental procedure, data fitting and DFT-based calculations 2.1. MoSe2 layer preparation conditions and characterization The MoSe2 layers were prepared by thermal treatment of Mo film on glass substrate in saturated selenium vapor at 500˚C during 30 min. The treatment was carried out in evacuated (up to 10 Pa) and sealed quartz ampoules. Mo thin films were deposited on 1.0 mm thick soda lime glass substrates by magnetron sputtering method using EVOVAC deposition system, Angstrom Engineering Inc. Deposition rate was 1 Å/sec and thickness of 900 nm was obtained. Prior to the deposition all of the substrates were chemically cleaned by acid. After chemical cleaning the substrates were rinsed in deionized water in ultrasonic bath and then dried under pure nitrogen flow. The obtained samples were examined by Scanning Electron Microscopy (SEM) for surface quality and by standard four-probe Hall measurements for conductivity. The films exhibited p-type conductivity. Concentration and mobility of holes were found as 5∙1018 cm-3 and 15.3 cm2/V∙s, respectively. All of the measurements were
carried out on the “as-prepared” films which were not subjected to thermal annealing, etching and other treatments. X-Ray diffraction (XRD) analyses of the films were carried out using Bruker D2 Phaser diffractometer in θ-2θ scan mode with Ni-filtered CuKα radiation (λ=1.54060 Å) source. Degree of crystallization was estimated from XRD patterns of the films using DIFFRAC.EVA.V2.1 PC program. 2.2 Ellipsometric measurements and data fitting The spectrometric ellipsometry measurements in 190-1700 nm spectral range were performed using Woollam M2000 rotating compensator instrument. Incident light angles were varied between 55 ° and 65 ° with 5 ° step. The CompleteEase computer program was used for the ellipsometric data fitting procedure. Experimental data were fitted (employing the Levenberg–Marquardt algorithm) to optical model simultaneously for all the data points measured in UV/VIS/NIR ranges. First, Mo film deposited onto glass substrate was measured and fitted to obtain its optical parameters. Further, the obtained data were used as substrate optical constants for fitting of MoSe2 layer ellipsometric data. Sample of commercially available MoSe2 rf sputtering target was also measured and fitted for comparison with layered samples. All measurements were performed at room temperature. 2.3 DFT calculations Full potential linearized augmented plane wave (FP LAPW) method, based on density functional theory (DFT), and WEIN2k code was used [12]. Exchange–correlation interactions were described within the local density approximation. The convergence parameter Rmt Kmax, where Rmt is the muffin-tin sphere radius and Kmax the size of the basis sets, was set to 7.0. Inside atomic spheres the partial waves were expanded up to lmax = 10. Integrations in reciprocal space were performed using the tetrahedron method with 60 points in the
irreducible part of the first Brillouin zone (BZ). The values of 2.48 and 2.2 a.u. were used for Rmt of Mo and Se, respectively. The cut-off energy for separation of the core and valence states was set to -6.0 Ry. Imaginary part of dielectric function in the range of optical frequencies was obtained by calculating the joint density-of-states for optical transitions between the valence and conduction bands, using Monchorst-Pack technique for integration over the BZ [13]. Real part was obtained from the imaginary part by Kramers-Kronig transformation. 3. Results and Discussion 3.1 Structure of MoSe2 layers Typical XRD pattern of molybdenum diselenide layer formed on Mo substrate is shown in Figure 1. The XRD patterns reveal predominance of rhombohedral structure of the obtained films with admixture of hexagonal phase. The wider peaks (003) and (101) represent superposition of those for rhombohedral and hexagonal phases that are found for both phases at very close angles. The narrow peak (104) represents rhombohedral phase only. No peaks indicating the presence of MoSe2-xOx phase has been identified. The average grain sizes and crystallinities of the layers derived from XRD patterns were ~ 45 nm and ~ 41 %, respectively. 3.2 Dielectric function Ellipsometric data were fitted within a two phase optical model with Mo optical constants used for modelling of substrate parameters. B-spline dispersion model available in CompleteEase database has been used to model MoSe2 dielectric function for the entire range of wavelengths. However, the best fit was obtained when MoSe2 was intermixed with pure Se constants within Bruggemann effective medium approximation (EMA) layer revealing around 16 percent of pure selenium content. Presence of pure selenium is obvious since the
measurements were performed on as-prepared samples without annealing them in vacuum. No surface roughness was found from fit, probably, due to selenium residual on the surface. The fit was performed by forcing the Kramers-Kronig consistency of optical constants. Mean square error (MSE) factors of the fit for all three angles simultaneously were as low as 11.4. Figure 2 represents typical experimental data along with the model fit for MoSe2 layer. Thickness of the layer and B-spline parameters were used as the fitting parameters. Apparently, the fit is quite good over the entire spectral range. MoSe2 layer with 76 nm thickness was formed on top of molybdenum film as was found form the fit and verified by SEM. Examination of MoSe2 layers of different batches has given essentially the same results on dielectric function. Complex dielectric function (real and imaginary parts, ε1 and ε2) of MoSe2 layers obtained from the optical model is presented in Figure 3a. Dielectric function of the polycrystalline MoSe2 target sample is also presented for comparison (Fig. 3b). The dielectric function of the layer seems to be red-shifted in the region above 1 eV as compared to that of target. At the same time, the only difference between our data in Fig. 3b and that provided by Demircioglu et. al. [14] concerns with a rather well developed structure at around 2eV in the latter. On the other hand, this structure is also evident on the imaginary part of dielectric function of MoSe2 layer (Fig. 3a). In all the other aspects, ours and available data for MoSe2 targets are very similar and the imaginary part of these two data sets include two structures in the energy range 2.5 - 4.5eV (Fig. 3b). Overall, the presence or absence of one or another of the above-mentioned structures in dielectric function spectra can be misleading and be related to polycrystalline nature of the studied MoSe2 samples. As shown by Comsa et. al. [15], electronic structure and exciton binding energies in MoS2, which is also the transition metal dichalcogenide, are strongly environment dependent. This is likely to hold for MoSe2 as well, where deformations strongly affect the electronic structure [16]. Therefore seemingly
different overall appearance of the dielectric function spectra is, most probably, due to different degree of crystallinity and grain size of the studied samples. According to a recent work [17], two peaks attributed to the A and B direct exciton transitions at the K-point of the BZ, as well as a broad structure from higher-lying inter-band optical transitions are observed at room temperature on the imaginary part of the dielectric function of the bulk single crystalline MoSe2 [17]. The energy regions for excitons and broad structure are 1.50-1.75eV and 2.5-3eV, respectively. Neither targets (both ours and the one studied earlier [14]) nor thin layers exhibit discrete exciton peaks in the just-mentioned energy range. On the other hand, the broad structure at above 2.5eV can be easily identified on the imaginary part of dielectric function of the target (Fig.3b). Strong broadening of inter-band optical transitions is likely to take place when going from single crystal to polycrystalline material leading to suppression of excitonic effects. Encircled in Figure 3a for MoSe2 layer, the ~ 200 meV wide feature at around 1eV of the incident photon energy is practically the same as the one observed by Demircioglu et. al. [14] for MoSe2 target. This feature transforms into a well-developed discrete line on the imaginary part of dielectric function of our target material, shown by the circle in Figure 3b. In other words, the observed peculiar feature is inherent in dielectric function of all MoSe2 polycrystalline samples. It would have been natural to account the feature for indirect optical transitions in indirect-gap semiconductor MoSe2. However, absorption coefficient at around 1 eV exceeds 104 cm-1, which is too high for indirect optical transitions. 3.3 Band structure, calculated and ellipsometry-based dielectric function The obtained band structure of the bulk single crystalline MoSe2, shown in Fig. 4 is in agreement with the results of the augmented-spherical-wave calculations performed for MoSe2 long time ago [18] and reproduces very well results of the recent studies [16, 19]. The
top of the valence band is in the center of the BZ, the bottom of the conduction band is situated on the symmetry line halfway between Г and K points. Fig. 5 displays the complex dielectric function derived from the obtained band structure of MoSe2 by integration of the direct optical transitions over the entire BZ. Indirect transitions have not been considered. Besides, excitonic effects are not included in the above dielectric function since our band structure calculations have been performed without taking electron-hole interaction into account. So far, as we have used room temperature lattice parameters in our band structure calculations and have not considered phonon influence, the obtained dielectric function is a non-broadened one with reasonable energy position of non-excitonic peculiarities. In Fig. 6 we have plotted the calculated and ellipsometry-based dielectric function. Similar sections on curves for calculated and experimental dielectric functions can be identified in the photon energy region above 1 eV. Comparison with other available data [14] on the target leads to even more favourable results. Thus, the calculated dielectric function is in good agreement with experimentally obtained dielectric function of the polycrystalline targets. The feature at around 1eV sustainably exhibited by polycrystalline targets and thin layers and located in the region of indirect optical transitions in single crystalline MoSe2 is not reproduced on the calculated dielectric function. This behavior can be attributed to the absence of excitonic affects that were not considered in the calculations. The discrete broad line emerged in the imaginary part of dielectric function of our target sample corresponds to phonon- free indirect exciton transitions that may encounter in indirect band gap materials due to impurities or structural imperfections. On the other hand, direct exciton transitions have not been observed for the studied polycrystalline samples. Thus, to reveal the origin of the peculiarity further detailed studies are required.
4. Conclusions MoSe2 layers with thicknesses below 100 nm have been prepared by annealing molybdenum films in saturated selenium vapor and studied along with polycrystalline target MoSe2 using spectroscopic ellipsometry in UV/VIS/NIR spectral range. The experimental data have been fitted to optical model data using B-spline dispersion model and the dielectric function has been retrieved for both the thin layer and target. Band structure calculations have been performed to derive dielectric function of the bulk single crystalline MoSe2. The obtained function has been found in good agreement with the ellipsometry-based dielectric function of the target in the energy region above 1eV. The observed peculiarity on the imaginary part of dielectric function around 1 eV is assumed to be due to indirect excitonic transitions. Further studies are necessary to completely clarify the origin of the peculiarity.
Acknowledgements This work is partly supported by research grant of Azerbaijan National Academy of Sciences.
References [1] A. Sanne, R. Ghosh, A. Rai, H. C. P. Movva, A. Sharma, R. Rao, L. Mathew, S. K. Banerjee, Top-gated chemical vapor deposited MoS2 field-effect transistors on Si3N4 substrates, Appl. Phys. Lett. 106 (2015) 062101. [2] J. Park, N. Choudhary, J. Smith, G. Lee, M. Kim, W. Choi, Thickness modulated MoS2 grown by chemical vapor deposition for transparent and flexible electronic devices Appl. Phys. Lett. 106 (2015) 012104. [3] S. Chandra, D. P. Singh, P. C. Srivastava, S. N. Sahu, Electrodeposited semiconducting molybdenum selenide films. II. Optical, electrical, electrochemical and photoelectrochemical solar cell studies, J. Phys D: Appl. Phys. 17 (1984) 2125-2130. [4] A. Pospischil, M. M. Furchi, T. Mueller, Solar-energy conversion and light emission in an atomic monolayer p–n diode, Nat. Nanotechnol. 9 (2014) 257–261. [5] T.Löher, Y.Tomm, C.Pettenkofer, A.Klein, W.Jaegermann,, Structural dipoles at interfaces between polar II-VI semiconductors CdS and CdTe and non-polar layered transition metal dichalcogenide semiconductors MoTe2 and MoSe2, Semiconductor Science and Technology 15 (2000) 514-522. [6] P.B. James, M.T Lavik, The crystal structure of MoSe2, Acta Crystallographica 16 (1963) 1183. [7] L.Towle, V. Oberbeck, R. Stajdohar, B. Brown, Molibdenum diselenide: rhombohedral high pressure-high temperature polymorphScience 154 (1966) 895-896. [8] S.M. Delphinec, M. Jayachandranb, C. Sanjeevirajaa, Pulsed electrodeposition and characterization of molybdenum diselenide thin film, Materials Research Bulletin 40 (2005) 135–147.
[9] H. Patel, J. Rathod, K. Patel, V. Pathak, R. Srivastava, Optical absorption Study of Molybdenum Diselenide and Polianiline and their Use in Hybrid Solar Cells, Advanced Materials Research 665 (2013) 239-253. [10] B. Evans and R. Hazelwood, Optical and structural properties of MoSe2, Phys. Stat. Sol. (a) 4 (1971) 181-192. [11] N. D. Boscher, C. J. Carmalt, R.G. Palgrave, J. J. Gil-Tomas, I. P. Parkin, Atmospheric Pressure CVD of Molybdenum Diselenide Films on Glass, Chem. Vap. Deposition 12 (2006) 692–698. [12]. P. Blaha, K. Schwarz, G. K. H. Madsen, D. Kvasnicka, and J. Luitz, WIEN2k, An Augmented Plane Waves + Local Orbitals Program for Calculating Crystal Properties, rev. ed. (Vienna University of Technology, Vienna, 2008). [13] H. Monkhorst and J. Pack, Special points for Brillouin-zone integrations, Phys. Rev. B 13 (1976) 5188-5192. [14] O. Demircioqlu, M. Mousel, A. Redinger, G.Rey, T. Weiss, S. Siebentritt, I. Riedel, L. Gutay, Detection of a MoSe2 secondary phase layer in CZTSe by spectroscopic ellipsometry, J. Appl. Phys. 118 (2015) 185302. [15] H. P. Komsa and A. V. Krasheninnikov, Effects of confinement and environment on the electronic structure and exciton binding energy of MoS2 from first principles, Phys. Rev. B 86 (2012) 241201(R). [16] B. Ghosh and N. Kishor, An ab initio study of strained two-dimensional MoSe2, J. Semicond. 36 (2015) 043001. [17] Y. Li, A. Chernikov, X. Zhang, A. Rigosi, H. M. Hill, A. M. van der Zande, D.A. Chenet, E. M. Shih, J. Hone, and T. F. Heinz, Measurement of the optical dielectric function of monolayer transition-metal dichalcogenides: MoS2, MoSe2, WS2 and WSe2 Phys. Rev. B 90 (2014) 205422.
[18] R. Coehoorn, C. Haas, and R.A. de Groot, Electronic structure of MoSe2, MoS2 and WSe2. II. The nature of the optical band gaps, Phys. Rev. B 35 (1987) 6203-6206. [19] Th. Böker, R. Severin, A. Müller, C. Janowitz, R. Manzke, D. Voß, P. Krüger, A. Mazur, and J. Pollmann, Band structure of MoS2 , MoSe2 , and α−MoTe2 : Angle-resolved photoelectron spectroscopy and ab initio calculations, Phys. Rev. B 64 (2001) 235305.
Captions Fig. 1. XRD pattern of MoSe2 layer. Range of angles with only MoSe2 peaks is shown for clarity. Fig. 2. Experimental ellipsometric data and model fit for 76 nm thick MoSe2 layer. Fig. 3. Dielectric function of MoSe2 (a) – for layer (insert shows index of refraction n and coefficient of extinction k); (b) – for target sample. Fig. 4. Calculated electronic band structure of MoSe2. Arrows indicate indirect and direct transitions. Fig. 5. Calculated dielectric function of MoSe2. Optical constants are inserted for convenience. Fig. 6. Real (a) and imaginary (b) parts of MoSe2 dielectric function. Dielectric function (scaled) obtained from ellipsometry data for target material is compared with that obtained from DFT-based calculations.
Fig.1
Fig.2
Fig.3 (a)
(b)
Fig.4
Fig.5
Fig.6 (a)
(b)
S0169-4332(17)30165-4 http://dx.doi.org/doi:10.1016/j.apsusc.2017.01.153 APSUSC 34948
To appear in:
APSUSC
Received date: Revised date: Accepted date:
31-7-2016 18-12-2016 16-1-2017
Please cite this article as: Ayaz Bayramov, Yegana Aliyeva, Gurban Eyyubov, Eldar Mammadov, Zakir Jahangirli, Daniel Lincot, Nazim Mamedov, Ellipsometric characterization of MoSe2 thin layers obtained by thermal treatment of molybdenum in selenium vapor, Applied Surface Science http://dx.doi.org/10.1016/j.apsusc.2017.01.153 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Ellipsometric characterization of MoSe2 thin layers obtained by thermal treatment of molybdenum in selenium vapor.
Ayaz Bayramova, Yegana Aliyevaa, Gurban Eyyubova, Eldar Mammadova, Zakir Jahangirlia, Daniel Lincotb, Nazim Mamedova a
Institute of Physics, Azerbaijan National Academy of Sciences, H. Javid ave.131, AZ1143 Baku, Azerbaijan b
Institut de Recherche et Développement sur l'Energie Photovoltaïque (IRDEP), 6 quai Watier, 78401 CHATOU Cedex, Paris, France
Highlights
MoSe2 layers are obtained by thermal treatment of molybdenum in selenium vapor. Spectroscopic ellipsometry is applied to the obtained layers and MoSe 2 target. Electronic band structure of MoSe2 is calculated and dielectric function is derived. Calculated and ellipsometry-based data on dielectric function agree fairly well. Excitonic transitions are assumed to form the dielectric function at around 1 eV.
Abstract Submicron MoSe2 layers were prepared by thermal treatment of thick Mo layers on glass substrate in saturated selenium vapor. Spectroscopic ellipsometry was then applied to the obtained MoSe2/Mo/Glass structures and MoSe2 target sample at room temperature. Dielectric function for both the MoSe2 layer and MoSe2 target was retrieved in the spectral range 1901700 nm by using the Kramers-Kronig consistent B-spline dispersion model. The obtained data were similar in both cases. Despite apparent red shift of the dielectric function spectra of the layer in high energy region the peculiarity at around 1 eV is manifested at the same energy for both, layer and target. Comparison of the ellipsometry-based dielectric function of the target and the one, obtained within calculated band structure of MoSe2 for room temperature lattice parameters, has shown that the former is a broadened counterpart of the latter. Abovementioned peculiar feature is not reproduced in the calculated dielectric function and is assumed to have excitonic nature.
Keywords: MoSe2 layer, spectroscopic ellipsometry, dielectric function
1. Introduction MX2 (where M = Mo, X = S, Se, and Te), indirect bandgap layered materials have attracted significant interest for various optoelectronic applications such as phototransistors, light-emitting and photovoltaic devices [1-4]. Particularly, MoSe2 is considered as a good low cost candidate for stable Cu-free back contact in CdTe-based thin film solar cells. The valence band offset of CdTe/MoSe2 interface is about 0.03 eV, which is much suitable for the holes transport between p-CdTe and MoSe2 [5]. Depending on the stacking manner of the layers in the unit cell, MoSe2 crystallizes in 2H or 3R polyform [6, 7]. The former has two layers stacked in hexagonal symmetry while the latter has three layers in the unit cell and
rhombohedral symmetry. The band gap of MoSe2 has been reported in the 1.1-1.42 eV range depending on preparation techniques and conditions [3, 8-10]. Several techniques, such as pulsed electrodeposition, chemical transport and chemical vapor deposition methods have been used for MoSe2 polycrystalline thin film growth [3, 9, 11]. However, a simple and inexpensive technique compatible with the process of substrate-configured CdTe-based thin film solar cell fabrication is in demand. In this regard, thermal annealing of molybdenum film on a glass substrate under selenium vapor is of interest. On the other hand, there is a lack of systematic studies of properties of the films prepared by this method. In this work, we present the results of ellipsometric studies of MoSe2 thin layers prepared by thermal treatment of molybdenum film in selenium vapor along with calculated band structure and dielectric function of MoSe2.
2. Experimental procedure, data fitting and DFT-based calculations 2.1. MoSe2 layer preparation conditions and characterization The MoSe2 layers were prepared by thermal treatment of Mo film on glass substrate in saturated selenium vapor at 500˚C during 30 min. The treatment was carried out in evacuated (up to 10 Pa) and sealed quartz ampoules. Mo thin films were deposited on 1.0 mm thick soda lime glass substrates by magnetron sputtering method using EVOVAC deposition system, Angstrom Engineering Inc. Deposition rate was 1 Å/sec and thickness of 900 nm was obtained. Prior to the deposition all of the substrates were chemically cleaned by acid. After chemical cleaning the substrates were rinsed in deionized water in ultrasonic bath and then dried under pure nitrogen flow. The obtained samples were examined by Scanning Electron Microscopy (SEM) for surface quality and by standard four-probe Hall measurements for conductivity. The films exhibited p-type conductivity. Concentration and mobility of holes were found as 5∙1018 cm-3 and 15.3 cm2/V∙s, respectively. All of the measurements were
carried out on the “as-prepared” films which were not subjected to thermal annealing, etching and other treatments. X-Ray diffraction (XRD) analyses of the films were carried out using Bruker D2 Phaser diffractometer in θ-2θ scan mode with Ni-filtered CuKα radiation (λ=1.54060 Å) source. Degree of crystallization was estimated from XRD patterns of the films using DIFFRAC.EVA.V2.1 PC program. 2.2 Ellipsometric measurements and data fitting The spectrometric ellipsometry measurements in 190-1700 nm spectral range were performed using Woollam M2000 rotating compensator instrument. Incident light angles were varied between 55 ° and 65 ° with 5 ° step. The CompleteEase computer program was used for the ellipsometric data fitting procedure. Experimental data were fitted (employing the Levenberg–Marquardt algorithm) to optical model simultaneously for all the data points measured in UV/VIS/NIR ranges. First, Mo film deposited onto glass substrate was measured and fitted to obtain its optical parameters. Further, the obtained data were used as substrate optical constants for fitting of MoSe2 layer ellipsometric data. Sample of commercially available MoSe2 rf sputtering target was also measured and fitted for comparison with layered samples. All measurements were performed at room temperature. 2.3 DFT calculations Full potential linearized augmented plane wave (FP LAPW) method, based on density functional theory (DFT), and WEIN2k code was used [12]. Exchange–correlation interactions were described within the local density approximation. The convergence parameter Rmt Kmax, where Rmt is the muffin-tin sphere radius and Kmax the size of the basis sets, was set to 7.0. Inside atomic spheres the partial waves were expanded up to lmax = 10. Integrations in reciprocal space were performed using the tetrahedron method with 60 points in the
irreducible part of the first Brillouin zone (BZ). The values of 2.48 and 2.2 a.u. were used for Rmt of Mo and Se, respectively. The cut-off energy for separation of the core and valence states was set to -6.0 Ry. Imaginary part of dielectric function in the range of optical frequencies was obtained by calculating the joint density-of-states for optical transitions between the valence and conduction bands, using Monchorst-Pack technique for integration over the BZ [13]. Real part was obtained from the imaginary part by Kramers-Kronig transformation. 3. Results and Discussion 3.1 Structure of MoSe2 layers Typical XRD pattern of molybdenum diselenide layer formed on Mo substrate is shown in Figure 1. The XRD patterns reveal predominance of rhombohedral structure of the obtained films with admixture of hexagonal phase. The wider peaks (003) and (101) represent superposition of those for rhombohedral and hexagonal phases that are found for both phases at very close angles. The narrow peak (104) represents rhombohedral phase only. No peaks indicating the presence of MoSe2-xOx phase has been identified. The average grain sizes and crystallinities of the layers derived from XRD patterns were ~ 45 nm and ~ 41 %, respectively. 3.2 Dielectric function Ellipsometric data were fitted within a two phase optical model with Mo optical constants used for modelling of substrate parameters. B-spline dispersion model available in CompleteEase database has been used to model MoSe2 dielectric function for the entire range of wavelengths. However, the best fit was obtained when MoSe2 was intermixed with pure Se constants within Bruggemann effective medium approximation (EMA) layer revealing around 16 percent of pure selenium content. Presence of pure selenium is obvious since the
measurements were performed on as-prepared samples without annealing them in vacuum. No surface roughness was found from fit, probably, due to selenium residual on the surface. The fit was performed by forcing the Kramers-Kronig consistency of optical constants. Mean square error (MSE) factors of the fit for all three angles simultaneously were as low as 11.4. Figure 2 represents typical experimental data along with the model fit for MoSe2 layer. Thickness of the layer and B-spline parameters were used as the fitting parameters. Apparently, the fit is quite good over the entire spectral range. MoSe2 layer with 76 nm thickness was formed on top of molybdenum film as was found form the fit and verified by SEM. Examination of MoSe2 layers of different batches has given essentially the same results on dielectric function. Complex dielectric function (real and imaginary parts, ε1 and ε2) of MoSe2 layers obtained from the optical model is presented in Figure 3a. Dielectric function of the polycrystalline MoSe2 target sample is also presented for comparison (Fig. 3b). The dielectric function of the layer seems to be red-shifted in the region above 1 eV as compared to that of target. At the same time, the only difference between our data in Fig. 3b and that provided by Demircioglu et. al. [14] concerns with a rather well developed structure at around 2eV in the latter. On the other hand, this structure is also evident on the imaginary part of dielectric function of MoSe2 layer (Fig. 3a). In all the other aspects, ours and available data for MoSe2 targets are very similar and the imaginary part of these two data sets include two structures in the energy range 2.5 - 4.5eV (Fig. 3b). Overall, the presence or absence of one or another of the above-mentioned structures in dielectric function spectra can be misleading and be related to polycrystalline nature of the studied MoSe2 samples. As shown by Comsa et. al. [15], electronic structure and exciton binding energies in MoS2, which is also the transition metal dichalcogenide, are strongly environment dependent. This is likely to hold for MoSe2 as well, where deformations strongly affect the electronic structure [16]. Therefore seemingly
different overall appearance of the dielectric function spectra is, most probably, due to different degree of crystallinity and grain size of the studied samples. According to a recent work [17], two peaks attributed to the A and B direct exciton transitions at the K-point of the BZ, as well as a broad structure from higher-lying inter-band optical transitions are observed at room temperature on the imaginary part of the dielectric function of the bulk single crystalline MoSe2 [17]. The energy regions for excitons and broad structure are 1.50-1.75eV and 2.5-3eV, respectively. Neither targets (both ours and the one studied earlier [14]) nor thin layers exhibit discrete exciton peaks in the just-mentioned energy range. On the other hand, the broad structure at above 2.5eV can be easily identified on the imaginary part of dielectric function of the target (Fig.3b). Strong broadening of inter-band optical transitions is likely to take place when going from single crystal to polycrystalline material leading to suppression of excitonic effects. Encircled in Figure 3a for MoSe2 layer, the ~ 200 meV wide feature at around 1eV of the incident photon energy is practically the same as the one observed by Demircioglu et. al. [14] for MoSe2 target. This feature transforms into a well-developed discrete line on the imaginary part of dielectric function of our target material, shown by the circle in Figure 3b. In other words, the observed peculiar feature is inherent in dielectric function of all MoSe2 polycrystalline samples. It would have been natural to account the feature for indirect optical transitions in indirect-gap semiconductor MoSe2. However, absorption coefficient at around 1 eV exceeds 104 cm-1, which is too high for indirect optical transitions. 3.3 Band structure, calculated and ellipsometry-based dielectric function The obtained band structure of the bulk single crystalline MoSe2, shown in Fig. 4 is in agreement with the results of the augmented-spherical-wave calculations performed for MoSe2 long time ago [18] and reproduces very well results of the recent studies [16, 19]. The
top of the valence band is in the center of the BZ, the bottom of the conduction band is situated on the symmetry line halfway between Г and K points. Fig. 5 displays the complex dielectric function derived from the obtained band structure of MoSe2 by integration of the direct optical transitions over the entire BZ. Indirect transitions have not been considered. Besides, excitonic effects are not included in the above dielectric function since our band structure calculations have been performed without taking electron-hole interaction into account. So far, as we have used room temperature lattice parameters in our band structure calculations and have not considered phonon influence, the obtained dielectric function is a non-broadened one with reasonable energy position of non-excitonic peculiarities. In Fig. 6 we have plotted the calculated and ellipsometry-based dielectric function. Similar sections on curves for calculated and experimental dielectric functions can be identified in the photon energy region above 1 eV. Comparison with other available data [14] on the target leads to even more favourable results. Thus, the calculated dielectric function is in good agreement with experimentally obtained dielectric function of the polycrystalline targets. The feature at around 1eV sustainably exhibited by polycrystalline targets and thin layers and located in the region of indirect optical transitions in single crystalline MoSe2 is not reproduced on the calculated dielectric function. This behavior can be attributed to the absence of excitonic affects that were not considered in the calculations. The discrete broad line emerged in the imaginary part of dielectric function of our target sample corresponds to phonon- free indirect exciton transitions that may encounter in indirect band gap materials due to impurities or structural imperfections. On the other hand, direct exciton transitions have not been observed for the studied polycrystalline samples. Thus, to reveal the origin of the peculiarity further detailed studies are required.
4. Conclusions MoSe2 layers with thicknesses below 100 nm have been prepared by annealing molybdenum films in saturated selenium vapor and studied along with polycrystalline target MoSe2 using spectroscopic ellipsometry in UV/VIS/NIR spectral range. The experimental data have been fitted to optical model data using B-spline dispersion model and the dielectric function has been retrieved for both the thin layer and target. Band structure calculations have been performed to derive dielectric function of the bulk single crystalline MoSe2. The obtained function has been found in good agreement with the ellipsometry-based dielectric function of the target in the energy region above 1eV. The observed peculiarity on the imaginary part of dielectric function around 1 eV is assumed to be due to indirect excitonic transitions. Further studies are necessary to completely clarify the origin of the peculiarity.
Acknowledgements This work is partly supported by research grant of Azerbaijan National Academy of Sciences.
References [1] A. Sanne, R. Ghosh, A. Rai, H. C. P. Movva, A. Sharma, R. Rao, L. Mathew, S. K. Banerjee, Top-gated chemical vapor deposited MoS2 field-effect transistors on Si3N4 substrates, Appl. Phys. Lett. 106 (2015) 062101. [2] J. Park, N. Choudhary, J. Smith, G. Lee, M. Kim, W. Choi, Thickness modulated MoS2 grown by chemical vapor deposition for transparent and flexible electronic devices Appl. Phys. Lett. 106 (2015) 012104. [3] S. Chandra, D. P. Singh, P. C. Srivastava, S. N. Sahu, Electrodeposited semiconducting molybdenum selenide films. II. Optical, electrical, electrochemical and photoelectrochemical solar cell studies, J. Phys D: Appl. Phys. 17 (1984) 2125-2130. [4] A. Pospischil, M. M. Furchi, T. Mueller, Solar-energy conversion and light emission in an atomic monolayer p–n diode, Nat. Nanotechnol. 9 (2014) 257–261. [5] T.Löher, Y.Tomm, C.Pettenkofer, A.Klein, W.Jaegermann,, Structural dipoles at interfaces between polar II-VI semiconductors CdS and CdTe and non-polar layered transition metal dichalcogenide semiconductors MoTe2 and MoSe2, Semiconductor Science and Technology 15 (2000) 514-522. [6] P.B. James, M.T Lavik, The crystal structure of MoSe2, Acta Crystallographica 16 (1963) 1183. [7] L.Towle, V. Oberbeck, R. Stajdohar, B. Brown, Molibdenum diselenide: rhombohedral high pressure-high temperature polymorphScience 154 (1966) 895-896. [8] S.M. Delphinec, M. Jayachandranb, C. Sanjeevirajaa, Pulsed electrodeposition and characterization of molybdenum diselenide thin film, Materials Research Bulletin 40 (2005) 135–147.
[9] H. Patel, J. Rathod, K. Patel, V. Pathak, R. Srivastava, Optical absorption Study of Molybdenum Diselenide and Polianiline and their Use in Hybrid Solar Cells, Advanced Materials Research 665 (2013) 239-253. [10] B. Evans and R. Hazelwood, Optical and structural properties of MoSe2, Phys. Stat. Sol. (a) 4 (1971) 181-192. [11] N. D. Boscher, C. J. Carmalt, R.G. Palgrave, J. J. Gil-Tomas, I. P. Parkin, Atmospheric Pressure CVD of Molybdenum Diselenide Films on Glass, Chem. Vap. Deposition 12 (2006) 692–698. [12]. P. Blaha, K. Schwarz, G. K. H. Madsen, D. Kvasnicka, and J. Luitz, WIEN2k, An Augmented Plane Waves + Local Orbitals Program for Calculating Crystal Properties, rev. ed. (Vienna University of Technology, Vienna, 2008). [13] H. Monkhorst and J. Pack, Special points for Brillouin-zone integrations, Phys. Rev. B 13 (1976) 5188-5192. [14] O. Demircioqlu, M. Mousel, A. Redinger, G.Rey, T. Weiss, S. Siebentritt, I. Riedel, L. Gutay, Detection of a MoSe2 secondary phase layer in CZTSe by spectroscopic ellipsometry, J. Appl. Phys. 118 (2015) 185302. [15] H. P. Komsa and A. V. Krasheninnikov, Effects of confinement and environment on the electronic structure and exciton binding energy of MoS2 from first principles, Phys. Rev. B 86 (2012) 241201(R). [16] B. Ghosh and N. Kishor, An ab initio study of strained two-dimensional MoSe2, J. Semicond. 36 (2015) 043001. [17] Y. Li, A. Chernikov, X. Zhang, A. Rigosi, H. M. Hill, A. M. van der Zande, D.A. Chenet, E. M. Shih, J. Hone, and T. F. Heinz, Measurement of the optical dielectric function of monolayer transition-metal dichalcogenides: MoS2, MoSe2, WS2 and WSe2 Phys. Rev. B 90 (2014) 205422.
[18] R. Coehoorn, C. Haas, and R.A. de Groot, Electronic structure of MoSe2, MoS2 and WSe2. II. The nature of the optical band gaps, Phys. Rev. B 35 (1987) 6203-6206. [19] Th. Böker, R. Severin, A. Müller, C. Janowitz, R. Manzke, D. Voß, P. Krüger, A. Mazur, and J. Pollmann, Band structure of MoS2 , MoSe2 , and α−MoTe2 : Angle-resolved photoelectron spectroscopy and ab initio calculations, Phys. Rev. B 64 (2001) 235305.
Captions Fig. 1. XRD pattern of MoSe2 layer. Range of angles with only MoSe2 peaks is shown for clarity. Fig. 2. Experimental ellipsometric data and model fit for 76 nm thick MoSe2 layer. Fig. 3. Dielectric function of MoSe2 (a) – for layer (insert shows index of refraction n and coefficient of extinction k); (b) – for target sample. Fig. 4. Calculated electronic band structure of MoSe2. Arrows indicate indirect and direct transitions. Fig. 5. Calculated dielectric function of MoSe2. Optical constants are inserted for convenience. Fig. 6. Real (a) and imaginary (b) parts of MoSe2 dielectric function. Dielectric function (scaled) obtained from ellipsometry data for target material is compared with that obtained from DFT-based calculations.
Fig.1
Fig.2
Fig.3 (a)
(b)
Fig.4
Fig.5
Fig.6 (a)
(b)