Influence of Post-Deposition Annealing on the Chemical States Of

Home , Tantalum pentoxide

arXiv:1804.09538v2 [cond-mat.mtrl-sci] 22 Nov 2018 ramns o neln eprtrsblw83K, 873 below heat temperatures annealing post-deposition Ta For synthesis in the temperatures treatments. on the depending and polymorphs methods several in lizes n 53Ktocytliepae hwu,namely: up, show phases crystalline two K 1523 and K ieotclbn a of gap band optical wide ihrfatv ne ( index refractive high storage and catalyst, capacitor insulator, applica- detector, gas technological as for such candidate tions good a as considered nvria uńm eCua uae,A.dlCar 450 Charro del Juárez, Av. Ciudad Autónoma de Universidad nuneo otdpsto neln nteceia stat chemical the on annealing post-deposition of Influence ntels er atlmpnoie(Ta pentoxide tantalum years last the In 2 O 5 saoposbtfrtmeaue ewe 873 between temperatures for but amorphous is 1–3 ainlCuclo cec n ehooy(CONACYT)-Inst Technology and Science of Council National hsi u oishg ilcrcconstant, dielectric high its to due is This . eel htteices nanaigtmeaueincreas temperature annealing in increase the that reveals eprtr.Uo nlss ti bevdteT 4 Ta the observed o is properties it chemical analysis, the Upon evaluate to temperature. performed were levels h omto fT-xd hsswt xdto ttsTa states oxidation with phases Ta-oxide of formation the rqec anto pteig h tmcarneet a arrangement, ( atomic hexagonal The predominately sputtering. magnetron frequency rhrobc( orthorhombic )o h hmclpoete fcytlieTa crystalline of properties chemical the on K) nelda eprtrsblw98K -a hteeto s photoelectron X-ray K. 948 below temperatures at annealed euto fteohr niaigta ihrtemperature higher that indicating others the of reduction efudta t17 h ai lgtyrdcssuggesting reduces slightly ratio the K 1273 temperature at annealing that optimal an found We be to seems there however, eivsiaeteeeto otdpsto neln (for annealing post-deposition of effect the investigate We .INTRODUCTION I. v e hro40Cl oeoPrio ..330 Juárez, 32310, C.P. Partido, Romero Col. 450 Charro del Av. v e hro40Cl oeoPrio ..330 Juárez, 32310, C.P. Partido, Romero Col. 450 Charro del Av. Juárez, 32310, C.P. Partido, Romero Col. 450 Charro del Av. v e hro40Cl oeoPrio ..330 Juárez, 32310, C.P. Partido, Romero Col. 450 Charro del Av. nttt fEgneigadTcnlg,UiesddAut´ Universidad Technology, and Engineering of Institute nttt fEgneigadTcnlg,UiesddAut´ Universidad Technology, and Engineering of Institute Aut´ Universidad Technology, and Engineering of Institute nttt fEgneigadTcnlg,UiesddAut´ Universidad Technology, and Engineering of Institute m6At artr rgeo ..7,CP 71 Mérida, Y 97310 C.P. 73, A.P. Progreso, a Carretera Ant. 6 km Mérida, Y 97310 C.P. 73, A.P. Progreso, a Carretera Ant. 6 km n 2 = β ∼ -Ta . eV 4.0 . 8at 18 ple hsc eatet IVSA nddMérida, Unidad CINVESTAV Department, Physics Applied Mérida, Unidad CINVESTAV Department, Physics Applied 2 O 5 o ape nelda 18Kad17 ;adaoposfrs for amorphous and K; 1273 and K 1148 at annealed samples for ) λ 4–7 δ 5 m,ada and nm), 550 = -Ta Ta . 2 ireGoan aiGonzález Mani Giovanni Pierre o´ rnddEiad Galindo Elizalde Trinidad José O 2 O 5 2 o h lsepsdt ettetet t98Kad14 K; 1048 and K 948 at treatments heat to exposed films the for ) O o´ .ErıuzCarrejo Enr´ıquez L. José Dtd oebr2,2018) 27, November (Dated: 5 a been has ) 5 etxd films pentoxide crystal- ie Gamboa Fidel salPerez Israel ui Farias Rurik ´ co Sosa V´ıctor 2 o.Rmr atd,CP 21,J´rz hhau,Méxi Juárez, Chihuahua, 32310, C.P. Partido, Romero Col. O euto hs infiatcagso h hscland Despite physical expected. the are on samples the changes of significant properties this, chemical of result a Ta oesr uloyeaino h ls post-deposition out films, carried the is of annealing oxygenation employed. full methods chemical ensure the the To on and strongly stoichiometric depends not composition is ratio tantalum to sdt rwTa been grow have EBE to and PLD, used as deposition, such assisted methods ion of sputtering, number rf great a years, last the In eaoa ( hexagonal fcytliepaetastosfo eaoa oor- to hexagonal increases temperature from annealing the transitions as thorhombic phase crystalline of 5 f h hmcl hscl n tutrlpoete of properties structural and physical, chemical, The lsgono i10 usrtsb radio by substrates Si(100) on grown films 2 ∗ O pcrmcaatrsi fT nTa in Ta of characteristic spectrum stepretg ftesaeTa state the of percentage the es 5 r oedsrbet rdc Ta produce to desirable more are s 1+ lsdel eedo h arcto processes. fabrication the on depend deeply films l lsa ucino annealing of function a as films all f etocp o a4 Ta for pectroscopy htmxmzsteO oT%ratio. Ta% to O% the maximizes that Ta , eemndb -a ircin is diffraction, X-ray by determined s xgndepletion. oxygen tt fEgneigadTechnology, and Engineering of itute eprtrsfo 4 o1273 to K 848 from temperatures δ nm eCua Juárez, Ciudad de onoma nm eCua Juárez, Ciudad de onoma Juárez, Ciudad de onoma nm eCua Juárez, Ciudad de onoma 2+ hs)adotohmi ( orthorhombic and phase) Ta , 2 O 3+ hhau,MéxicoChihuahua, hhau,MéxicoChihuahua, MéxicoChihuahua, hhau,MéxicoChihuahua, 5 so rsaln tantalum crystalline of es n Ta and , films ucatán, México ucatán, México 18–20 12–17 f hr a enevidence been has There . 4+ n 1 O and ngnrl h oxygen the general, In . h study The . 5+ n the and 5+ amples s core- 2 β and O 5 phase) , 8–11 21 As . . co 2 that there is wide number of works studying the chemical Tann properties of amorphous Ta O films17,22–25, to the best TABLE I. Annealing temperature and determined crys- 2 5 talline phase for the films. of our knowledge, there a few works focused on studying Tann these properties for the crystalline phases. Film K Phase In this work we investigate the annealing tempera- F298 298 Amorphous ture dependence of the chemical properties in crystalline F848 848 Amorphous Ta2O5 films deposited on Si substrates by RF magnetron F948 948 δ sputtering. For this purpose we prepared several Ta films F1048 1048 δ and exposed them to annealing temperatures ranging F1148 1148 β from 848 K to 1273 K. Their crystalline structure was F1273 1273 β evaluated with XRD (X-ray diffraction) and the chemical properties such as atomic concentration, chemical states, and atomic bonding were studied by X-ray photoelectron spectroscopy (XPS). chemical properties of the films. For the XPS surveys and scans we used steps of 1 eV and 0.1 eV, respectively; the beam spot had a diameter of 400 µm and made an angle ◦ II. EXPERIMENTAL relative to the sample of 30 . Chemical properties were assessed measuring the Ta 4f and O 1s core-levels. The atomic composition, chemical states, and atomic bonding A. Film growth and annealing were determined by deconvolution of the Ta 4f and O 1s spectra using Shirley-Sherwood background and Voigt Six amorphous Ta films were deposited at room tem- functions [Gaussian σ = (1.43, 0.9) eV and Lorentzian perature on Si(100) substrates by the RF magnetron γ =0.02 eV] as implemented in the AAnalizer software26. sputtering technique. The deposition took place in a vacuum chamber with a base pressure of 6.6 × 10−5 mbar. Argon gas (99.9% purity) was flushed into the chamber to obtain a working pressure of (2.6 ± 0.1) × 10−2 mbar. III. RESULTS AND DISCUSSION Before deposition the substrates were cleaned up by several baths of distilled water, acetone, and ethanol. In A. Crystalline Structure order to eliminate the native oxide layer on the target a 5 min-presputtering was conducted before deposition. To evaluate the crystalline structure of our samples we The deposition was carried out using a 2.5 inch-Ta tar- performed XRD measurements. Samples F298 and F848 get with 99.95% purity and a sputtering power of 120 W; showed no diffraction patterns, indicating a disordered −1 resulting in a deposition rate of 2.8 A˚·s and a thick- atomic structure (see Fig. 1). The rest of the samples ness for all films of 2.4 µm. The target-to-substrate dis- exhibit a crystalline structure. For the sake of phase tance was about 12 cm and during film growth no oxy- indexation we compared our measurements with sev- gen was flushed into the chamber. To induce the desired eral references from the Powder Diffraction File (PDF) crystalline phase, i.e., Ta2O5, five films were exposed to dababase under the International Centre for Diffrac- post-deposition heat treatments in air for 1 h at different tion Data (ICDD) implemented in the software of the annealing temperatures, namely, (848, 948, 1048, 1148, diffractometer. From a list of at least six PDFs (even 1273) K using a Thermo Scientific Thermolyne cylindri- among these references we spotted slight discrepancies) cal furnace (model F21135). To associate the annealing the films F948 and F1048 can be indexed to the hexag- temperature to the samples, the films were labeled F848, onal phase δ−Ta2O5 with PDF 00-019-1299; lattice pa- F948, F1048, F1148, and F1273. The remaining film, rameters a = b = 3.6240 A,˚ c = 3.8800 A,˚ α = β = 90◦, labeled F298, was kept for future reference and was not and γ = 120◦; spatial group P 6/mmm27,28; whereas the subjected to any heat treatment. Table I gives the an- films F1148 and F1273 can be indexed to the orthorhom- nealing temperature Tann and the crystalline phase of the bic phase β−Ta2O5 with either PDF 00-025-0922; lattice films. parameters a = 6.1980 A,˚ b = 40.2900 A,˚ c = 3.8880 A,˚ ◦ and α = β = γ = 90 ; spatial group P 21212; or PDF 01- 089-2843; lattice parameters a = 6.2000 A,˚ b = 3.6600 B. Characterization A,˚ c = 3.8900 A,˚ and α = β = γ = 90◦; spatial group Amm2. In Table I the phases for all films are summa-

A Siemens diffractometer model D-5000 with Cu Kα1 rized. Figure 2 shows the best matches; there we can see radiation (λ =1.5406 A)˚ was used to evaluate the atomic that the pattern of F1273 shows a much richer pattern structure of the samples. Steps of 0.02◦ with a time per than the pattern due to F948. It is evident that F1273 step of 3 s and operating parameters of 34 kV and 25 exhibits a series of small reflections along the whole pat- µA were used to obtain the XRD patterns. A Thermo tern that match the reference pattern PDF 025-0922. It Scientific K-Alpha XPS spectrometer with an Al Kα X- is important to underline that so far there has not been ray source set to 12 kV and 40 W was used to analyze the a consensus regarding the spatial symmetry of these two 3

both the crystalline structure and the spatial symmetry are beyond the scope of the present investigation. The crystallite size D was estimated using the so-called Scherrer relation

F1273 Kλ D = , (1) Γcos θ where we used K = 0.9, λ = 1.5406 A˚ and Γ is the full F1148 width at half maximum. For our calculations we used ◦

the peak at 2θ = 28.3 and found that Γ =(0.79, 0.69, ◦ F1048 0.52, 0.27) . With these values we found the size to be

Intensity/a.u. D =(10, 12, 16, 30) nm, respectively.

F948 It is worth mentioning that Ta2O5 is the most sta- ble oxide of Ta; other oxides such as TaO, Ta2O, TaO2,

F848 Ta2O3 are diﬃcult to synthesize as pure phases (ex- 10 cept for TaOx) and usually appear as contamination . F298 Furthermore, most suboxides are crystalline, except for

20 30 40 50 60 70 Ta2O3 and TaOx that are amorphous. This issue will

2 / degrees become important for our forthcoming discussion.

B. Chemical properties FIG. 1. X-ray diffraction patterns for the crystalline films. The patterns for the samples F948 and F1048 are indexed to the δ phase and the patterns for F1148 and F1273 are The chemical properties of our samples were studied by indexed to the β phase of Ta2O5. Samples F298 and F848 XPS analysis. The binding energy of all spectra for Ta are amorphous. 4f and O 1s core-levels was calibrated at 532 eV using as reference the oxygen peak. During the analysis, we also spotted traces of carbon contamination in all films as revealed by the surveys in Figure 3. The results of the spectra for the Ta 4f core-level are

P DF 025-0922 shown in Fig. 4(a). All spectra exhibit the typical spin-

F1273 (001)

(0 22 1) orbit doublet corresponding to the levels 4f and 4f (1 11 0) (0 22 0) 7/2 5/2 (1 11 1) (002) located at 27.8 eV and 29.6 eV with peak splitting of 1.9

(1 10 0) eV. The binding energies and peak splitting are char- (0 22 2) (0 14 0) (2 15 1) (2 22 1) (2 22 0)

(340) 5+ (342) acteristic of Ta in stoichiometric amorphous Ta2O5

ﬁlms37–40. A closer look at the black spectrum for the as-

P DF 0019-1299 deposited film (F298) reveals a satellite on the low energy F948 Intensity/a.u. region. This feature has been reported to be caused by screening of 5d electrons in a-TaOx (x =1.86, 2.00), thus 23 suggesting the presence of Ta suboxides such as TaOx . Since this film was not exposed to a heat treatment, one would expect, besides amorphous Ta2O5, the appearance 20 30 40 50 60 70 of other Ta suboxides that were mainly generated dur- 2 / degrees ing deposition as there still exists residual oxygen in the deposition chamber. FIG. 2. Phase indexation of our samples. Samples F1148 and As the annealing temperature is increased, the shoul- F1273 can be indexed to PDF 025-0922 whereas films F948 der vanishes indicating that the amount of suboxides is and F1048 can be indexed to PDF 0019-1299. reduced. The O 1s core-level gives additional information on the chemical states of our samples; the corresponding spectra are shown in Fig. 4 (b). We first observe a peak at phases, and, due to the overlapping of several peaks, one 532 eV associated to Ta-O bonds39,41. At ∼534 eV there 21,29–36 cannot rule out the coexistence of both phases . is a satellite whose intensity reduces as Tann decreases. To the best of our knowledge, at least 12 spatial groups This feature is attributed to residual oxygen and surface has been proposed and we do think it is not worth com- contamination; mainly carbon compounds (see below). puting the lattice parameters or conducting a Rietveld We also notice that for the sample F1273, the shoulder refinement, since these depend on the particular choice reappears. To verify the source of this shoulder we sput- of the group symmetry. The precise determination of tered the sample for 9 min with an argon ion beam with 4

F298

Ta-O F298 F298

O 1s Ta

F848

Ta4f O1s

F948

3+ F1048

Ta 4f 2+

Ta 1+

F1148

Ta 4d

F2173

Ar2p

F848 F848

3/2 Ta4f

O1s

Ta 4p

1/2 Intensity/a.u.

Ta 4p

C 1s

O KLL

F948 F948

Ta4f

O1s

F1048 F1048 Ta4f

1000 800 600 400 200 0

O1s

Binding energy/eV

Intensity/a.u.

F1148 FIG. 3. XPS surveys for all samples. Some degree of carbon Ta4f F1148

O1s

contamination in all samples is observed.

F1273 F1273

9 min

Ta4f

O1s

536 534 532 530 528 34 32 30 28 26 24 22 Ta4f

Binding energy/eV Binding energy/eV

Intensity/a.u.

34 32 30 28 26 24 22 FIG. 5. Deconvolution of the Ta 4f (left column) and O 1s

Binding energy/eV (right column) core-levels for the six ﬁlms. The inset in the O

F298 1s spectrum of F1273 shows the O 1s spectrum after 9 min of b)

O1s F848 sputtering. One can see that the satellite almost disappears F948 and it is therefore caused by contamination.

F1048

F1148 Intensity/a.u.

F1273

538 536 534 532 530 528 526

Binding energy/eV show no traces of crystalline phases of Ta suboxides we believe that, due to the unstable nature of these phases, during annealing the surface atoms readsorbed O to form FIG. 4. Ta 4f (a) and O 1s (b) core-levels for the six films. Ta2O5 and the other phases are trapped in a few top- most layers42. On the other hand, the deconvolution of the O 1s spectra is shown on the right column of the same figure. There we can see that the oxygen spectrum a voltage of 3 kV, and an electric current of 10 µA — is well fitted with three peaks. The low energy peak is the incidence angle between the sample and the ion gun attributed to Ta-O bonding, the peak at 533 eV is due ◦ was 90 . Then the O 1s core-level was measured and the to residual oxygen and the peak at 534 eV is due to car- shoulder considerably diminished (as seen in the inset of bon contamination, most probably carbon dioxide which the corresponding film in Fig. 5). We thus believed that is quite ubiquitous in all samples22,23,43. As discussed this feature is caused by contamination. above the sample F1273 exhibits a pronounced satellite The deconvolution analysis performed on the Ta 4f whose intensity was reduced after sputtering the sample spectra reveals that there are five contributions from five for 9 min. The deconvolution confirms that this satellite Ta oxidation states (see left column in Fig. 5). The is due to contamination. five doublets are located, with respect to Ta 4f7/2, at In order to assess the effect of the annealing temper- binding energies of (23.0, 24.1, 25.4, 26.7, 27.8) eV with ature on the chemical properties, we used the doublets spin-orbit splittings of 1.9 eV. These binding energies are of the Ta 4f spectra to compute the percentage of oxi- 1+ 2+ 3+ 4+ 5+ attributed to Ta , Ta , Ta , Ta , and Ta states, dation state as function of Tann. Figure 6(a) shows the respectively22,44,45. These results strongly suggest the results. We can observe a monotonic increase of the per- formation of metastable phases of Ta oxides such as TaO, centage of Ta in the state Ta5+ as a function of annealing 5+ Ta2O, TaO2, TaOx, and Ta2O3. However, since the XRD temperature. The increase for the state Ta goes from 5

IV. CONCLUSIONS

100

a) 80 The atomic structure of Ta2O5 ﬁlms was studied by +1

60 +2 X-ray diﬀraction. We found evidence for the presence of

+3 40 the hexagonal phase of Ta2O5 for samples annealed below +4 1048 K and the orthorhombic phase of Ta O for samples 20 2 5

+5 % Oxidation state Oxidation %

annealed above 1048 K. In all these cases there are no

1.6 traces of crystalline Ta suboxides. The ﬁlms exposed 5+ b) to heat treatments below 848 K showed no diﬀraction 1.2 patters and were found to be amorphous.

0.8 XPS studies were realized in order to characterize the chemical properties of the ﬁlms. According to the anal- 0.4 O%/Ta% of Ta of O%/Ta% ysis carried out on the spectra for Ta 4f and O 1s core- 0.0 levels, it can be concluded that Ta suboxides show up in 400 600 800 1000 1200

T /K small amounts as revealed by the appearance of satellites ann in both the Ta 4f and O 1s core-levels. The deconvolution of the XPS spectra strongly indicates the existence of FIG. 6. Annealing temperature dependence of the oxidation several chemical states such as Ta1+, Ta2+, Ta3+, Ta4+, state (a) and of the oxygen to tantalum ratio (b). and Ta5+. We conclude that as the annealing temperature is increased the presence of the state Ta5+ increases, indicating that high annealing temperatures are desired to favor the generation of tantalum pentoxide. However, there seems to be a limit in the value of the annealing temperature if one wishes to avoid oxygen depletion. This is deduced by observing a reduction in the oxygen to tantalum ratio for the sample F1273.

67% to 93% whereas the rest of the states remain be- ACKNOWLEDGEMENTS low 10%. This tendency clearly demonstrates that the crystalline phases of Ta2O5 are favoured at high tem- We are grateful to Wilian Cauich and Daniel Aguilar peratures. We also computed the oxygen to tantalum for their technical support during the XPS and XRD ses- ratio as a function of Tann. The results are given in Fig. sions. Dr. Israel Perez is indebted to Dr. Alberto Her- 6(b). As expected, the ﬁlm F298 exhibits an O% to Ta% rera for helpful discussions and technical support in the ratio close to 1; nevertheless as the temperature is in- XPS analysis. We also thank the anonymous reviewer creased from 298 K to 1148 K, the ratio remains around and one of the editors of this journal for their comments 1.5 (non-stoichiometric) and decreases to 1.2 at 1273 K. that greatly improved the quality of this work. The au- This behaviour suggests that there is a range of temper- thors gratefully acknowledge the support from the Na- ature that favours the generation of tantalum pentoxide tional Council of Science and Technology (CONACYT) with the highest oxygen content. Accordingly, however, Mexico and the program C´atedras CONACYT through from 1148 K to 1273 K the oxygen seems to deplete. project 3035.

∗ Contact Author: [email protected] cursor, Electrochem. J. Appl. Phys. 83, 4823-4829 (1998) 1 T. Kaga, H. Shinriki, F. Murai, Y. Kawamoto, Y. Nak- 5 R.H. Dennard, F.H. Gaensslen, H. Yu, V.L. Rideout, E. agone, F. Takeda, K. Itoh, DRAM Manufacturing in the Bassous, A.R. LeBlanc, Design of ion-implanted MOS- ’90s -Part 3: A Case Study, Semicond. Int. 6, 98-101 (1991) FET’s with very small physical dimensions, IEEE J. Solid- 2 K. W. Kwon, C.S. Kang, S.O. Park, H.K. Kang, S.T. State Circ. 9, 256-268 (1974) 6 Ann, Thermally robust Ta2O5 capacitor for the 256-Mbit S. Shibata, Dielectric constants of Ta2O5 thin films de- DRAM, IEEE Trans. Electron. Devices. 43, 919-923 (1996) posited by r.f. sputtering, Thin Solid Films 277, 1-4 (1996) 3 7 C. Chaneliere, J.L. Autran, R.A.B. Devine, B. Baland, E. Atanassova, Thin RF sputtered and thermal Ta2O5 on Tantalum pentoxide (Ta2O5) thin films for advanced di- Si for high density DRAM application, Microelectron. Re- electric applications, Mater. Sci. Eng. R-rep 22, 269-322 liab. 39, 1185-1217 (1999) (1998) 8 J. D. T. Kruschwitz and W. T. Pawlewicz, Optical and 4 C. Chaneliere, S. Four, J. L. Autran, R. A. B. Devine, durability properties of infrared transmitting thin films, and N. P. Sandler, Properties of amorphous and crystalline Appl. Opt. 36, 2157-2159 (1997) Ta2O5 thin films deposited on Si from a Ta(OC2H5)5 pre- 6

9 C. Chaneliere, S. Foura, J. L. Autran, R. A. B. Devine, Process. 13, 245-251 (2010) Comparison between the properties of amorphous and 26 A. Herrera-Gomez, M. Bravo-Sanchez, O. Ceballos- crystalline Ta2O5 thin films deposited on Si, Microelec- Sanchez, M.O. Vazquez-Lepe, Practical Methods for Back- tron. Reliab. 39, 261-268 (1999) ground Subtraction in Photoemission Spectra, Surf. Inter- 10 S. P. Garg, N. Krishnamurthy, A. Awashi, and M. Venka- face Anal. 46, 897-905 (2014) 27 traman, The O-Ta (Oxygen-Tantalum) system, J. Phase A. Fukumoto and K. Miwa, Prediction of hexagonal Ta2O5 Equilib. 17, 63-77 (1996) structure by first-principles calculations, Phys. Rev. B 55, 11 K.T. Jacob, C. Shekhar, Y. Waseda, An update on the 11155 (1997) 28 thermodynamics of Ta2O5, J. Chem. Thermodyn. 41, 748- Y.-N. Wu, L. Li, and H.-P. Cheng, First-principles studies 753 (2009) of Ta2O5 polymorphs, Phys. Rev. B 83, 144105 (2011) 12 E. Atanassova, T. Dimitrova, J. Koprinarova, AES and 29 S. Pérez-Walton, C. Valencia-Balv´ın, A.C.M. Padilha, XPS study of thin RF-sputtered Ta2O5 layers, Appl. Surf. G.M. Dalpian, J.M. Osorio-Guillén, A search for the Sci. 84, 193-202 (1995) ground state structure and the phase stability of tantalum 13 H. Shinriki, M. Nakata, UV-O3 and Dry-O2: Two-step An- pentoxide, J. Phys. Condens. Matt. 28, 035801-11 (2016) 30 nealed Chemical Vapor-Deposited Ta2O5 Films for Storage J.Y. Kim, B. Magyari-Köpe, K.-J. Lee, H.-S. Kim, S.-H. Dielectrics of 64-Mb DRAMs, IEEE Trans. Electron De- Lee, and Y. Nishi, Electronic structure and stability of low vices ED. 38, 455-462 (1991) symmetry Ta2O5 polymorphs, Phys. Status Solidi RRL 8, 14 S. Kamiyama, P.-Y. Lesaicherre, H. Suzuki, A. Sakai, I. 560-565 (2014) Nishiyama, A. Ishitani, J. Electrochem. Soc. 140, 1617 31 S.-H. Lee, J. Kim, S.-J. Kim, S. Kim, G.-S. Park, Hidden (1993) Structural Order in Orthorhombic Ta2O5, Phys. Rev. Lett. 15 G. Q. Lo, D. L. Kwong, S. Lee, Appl. Phys. Lett. 62, 973 110, 235502-5 (2013) (1993) 32 J. Lee, W. Lu, and E. Kioupakis, Electronic properties of 16 Y. Kuo, Reactive Ion Etching of Sputtered Deposited Tan- tantalum pentoxide polymorphs from first-principles cal- talum Oxide and its Etch Selectivity to Tantalum, J. Elec- culations, Appl. Phys. Lett. 105, 202108-5 (2014) trochem. Soc. 139, 579-583 (1992) 33 Z. Helali, M. Calatayud, and C. Minot, Novel Delta- 17 N. Donkov, A. Zykova, V. Safonov, R. Rogowska, J. Smo- Ta2O5 Structure Obtained from DFT Calculations, J. lik, Tantalum Pentoxide Ceramic Coatings Deposition on Phys. Chem. C. 118, 13652-13658 (2014) Ti4A16V Substrates for Biomedical Applications, Prob- 34 Y. Guo, J. Robertson, Comparison of oxygen vacancy de- lems At. Sci. Technol. Plasma Physics Series 17, 131-133 fects in crystalline and amorphous Ta2O5, Microelectron. (2011) Engin. 147, 254-259 (2015) 18 T. Dimitrova,U, K. Arshak, and E. Atanassova, Crystal- 35 J.-Y. Kim, B. Magyari-Köpe, Y. Nishi, J.-H. Ahn, First- lization effects in oxygen annealed Ta2O5 thin films on Si, principles study of carbon impurity effects in the pseudo- Thin Solid Films 38, 31-38 (2001) hexagonal Ta2O5, Current Appl. Phys. 16, 638-643 (2016) 19 S.-J. J. Wu, B. Houng, B. S. Huang, Effect of growth 36 Y. Yang Y. Kawazoe, Prediction of a new ground-state and annealing temperatures on crystallization of tantalum crystal structure of Ta2O5, Phys. Rev. Mat. 2, 034602 pentoxide thin film prepared by RF magnetron sputtering (2018) method, J. Alloys Compd. 475, 488-493 (2009) 37 X. M. Wu, P. K. Wu, T.M. Lu, and E. J. Rymaszewski, 20 D. Cristea, D. Constantin, A. Crisan, C.S. Abreu, J.R. Reactive sputtering deposition of low temperature tanta- Gomes, N.P. Barradas, E. Alves, C. Moura, F. Vaz, L. lum suboxide thin films. Appl. Phys. Lett. 62, 3264-3266 Cunha, Properties of tantalum oxynitride thin films pro- (1993) duced by magnetron sputtering: The influence of process- 38 K. Chen, G.R. Yang, M. Nielsen, T.M. Lu, E.J. Ry- ing parameters, Vacuum 98, 63-69 (2013) maszewski, X-ray photoelectron spectroscopy study of 21 I. Perez, J. L. Enr´ıquez-Carrejo, V. Sosa, F. Gamboa, Al/Ta2O5 and Ta2O5/Al buried interfaces, Appl. Phys. J. R. Farias-Mancillas, J. T. Elizalde-Galindo, Evidence Lett. 70, 399-401 (1997) for structural transition in crystalline tantalum pentoxide 39 A. Muto, F. Yano, Y. Sugawara and S. Iijima, The Study films grown by RF magnetron sputtering, J. Alloys. Comp. of Ultrathin Tantalum Oxide Films Before and After An- 712, 303-210 (2014) nealing with X-Ray Photoelectron Spectroscopy, Jpn. J. 22 E. Atanassova, D. Spassov, X-ray photoelectron spec- Appl. Phys., 33, 2699-2702 (1994) 40 troscopy of thermal thin Ta2O5 films on Si, Appl. Surf. R. Simpson, R. G. White, J.F. Watts, M. A. Baker, XPS Sci. 135, 71-82 (1998) investigation of monatomic and cluster argon ion sput- 23 Takashi Tsuchiya, Hideto Imai, Shogo Miyoshi, Per- tering of tantalum pentoxide, Appl. Surf. Sci. 405, 79-87 Anders Glans, Jinghua Guo and Shu Yamaguchi, X-Ray (2017) absorption, photoemission spectroscopy, and Raman scat- 41 H. Szymanowski, O. Zabeida, J. E. Klemberg-Sapieha, L. tering analysis of amorphous tantalum oxide with a large Martinu, Optical properties and microstructure of plasma extent of oxygen nonstoichiometry, Phys. Chem. Chem. deposited Ta2O5 and Nb2O5 films, J. Vac. Sci. Technol. A Phys. 13, 17013-17018 (2011) 23, 241-247 (2005) 24 S.-C. Wang, K.-Y. Liu, J.-L. Huang, Tantalum oxide film 42 E. Atanassova, M. Kalitzova, G. Zollo, A. Paskaleva, prepared by reactive magnetron sputtering deposition for A. Peeva, M. Georgieva, G. Vitalib, High temperature- all-solid-state electrochromic device, Thin Solid Films 520, induced crystallization in tantalum pentoxide layers and 1454-1459 (2011) its influence on the electrical properties, Thin Solid Films 25 S.V. Jagadeesh-Chandra, C.-J. Choi, S. Uthanna, G. 426, 191-199 (2003) Mohan-Rao, Structural and electrical properties of radio 43 O. Kerrec, D. Devilliers H. Groult, P. Marcus, Study of dry frequency magnetron sputtered tantalum oxide films: In- and electrogenerated Ta2O5 and Ta/Ta2O5/Pt structures fluence of post-deposition annealing, Mater. Sci. Semicond. by XPS, Mat. Sci. Eng. B55, 134-142 (1998) 7

44 M.V. Ivanov, T.V. Perevalov, V.S. Aliev, V.A. Grit- 45 M.V. Ivanov, T.V. Perevalov, V.Sh. Aliev, V.A. Gritsenko, senko,V.V. Kaichev, Electronic structure of δ−Ta2O5 with V.V. Kaichev, Ab Initio Simulation of the Electronic Struc- oxygen vacancy: ab initio calculations and comparison ture of δ − Ta2O5 with Oxygen Vacancy and Compari- with experiment, J. Appl. Phys. 110, 024115 (2011) son with Experiment, J. Exp. Theo. Phys. 112, 1035-1041 (2011)