Influence of N2 Flows on Sputtered Ta(N) films: Electrical,

By reactive DC magnetron sputtering from a pure Ta target onto silicon sub13 strates, Ta(N) films were prepared with a different N2 flow rate of 0, 12, 17, 25, 38, 58 sccm. 14 The effects of N2 flow rate on the electrical properties, crystal structure, elemental com15 position and optical properties of Ta(N) were studied. These properties were character16 ized by the four-probe method, X-ray diffraction (XRD), X-ray photoelectron spectros17 copy (XPS), and spectroscopic ellipsometry (SE). Results show that the deposition rate 18 decreases with an increase of the N2 flows. On the other hand, the resistivity increases, 19 the crystal size decreases, and the crystal structure transitions from β-Ta to TaN(111), and 20 finally becomes the N-rich phase Ta3N5 (130,040). Studying the optical properties, it is 21 found that there are differences in the refractive index (n) and extinction coefficient (k) of 22 Ta(N) with different thicknesses and different N2 flow rates, and dependent on the crys23 tal size and crystal phase structure. 24


Introduction
Transition metal nitrides, especially tantalum nitride (TaN) are highly demanded in 29 a wide range of applications due to their high melting point, hardness, excellent wear 30 and corrosion resistance, refractory character, mechanical and high-temperature stability, 31 chemical inertness, and histocompatibility [1][2][3][4][5][6]. Some prominent examples of such ap- 32 plications as a protective coating material for protection against oxidation and corrosion 33 [7], as a diffusion barrier for Al and Cu metallization in advanced microelectronics [8][9][10][11], 34 in phosphide and nitride optoelectronics as ohmic contact [3,4], in artificial heart valves 35 as histocompatibility materials [12], thin film resistor [13], as ceramic pressure sensors 36 [14], and also different mechanical applications [5,6]. The large interest for TaN arises 37 since it is considered recently as a high thermal conductive material in microelectronic 38 chips for the θ-TaN phase [15].
Numerous reports have been published to characterized sputtered TaN films based on 46 various sputtering parameters such as nitrogen (N2) partial pressure ratio [18,19], N2/Ar 47 flow rate ratio [20][21][22], sputtering power [23], and substrate temperature [24] during 48 deposition. By controlling these different parameters, the influence on the structural, 49 chemical, electrical, and optical properties of the TaN film have been investigated. 50 Among them, there are quite a few studies on how the N2 flow or N2/Ar flow rate 51 ratio and the N2/(N2+Ar) partial pressure ratio affect the properties of the TaN film. Chen 52 et al. [25] used a magnetron sputtering low-power radio frequency deposition method 53 with variable nitrogen flow rate to deposit TaNx barrier layers on silicon. They found that 54 as the N2 flow rate increases, the surface roughness of the deposited TaNx film was 55 slightly increased, and the amorphous structure of TaNx was formed with good thermal 56 stability. Zaman et al. [26] prepared a TaN film with a 3% to 25% N2/(N2+Ar) ratio on Si 57 substrate by reactive magnetron sputtering and studied the effect of N2 partial pressure 58 on the crystal structure and hardness of the TaN film. It was found that the deposited 59 films with 5% and 3% N2 content showed the highest hardness (33 Pa). 60 Although sputtered TaN films have been widely investigated for their different 61 properties, their optical properties have not been analyzed yet much. Recent studies have 62 shown that spectroscopic ellipsometry (SE) can be used to characterize and measure the 63 thin film thickness because of its fast and non-destructive nature [2,[27][28][29]. Aouadi et al. 64 [2] have studied the effects of varying N2 flow rates from 1 to 4 sccm on the structural and 65 optical properties of TaN thin films. They report that the optical constants (ε1, ε2) will be 66 used in conjunction with real-time SE to monitor and control the growth of tantalum ni-67 tride films. Cherfi et al. [30] deposited TaN films with an N2 flow rate of 0-12 sccm on Si 68 (100) and glass substrates by DC magnetron sputtering, and shown that the influence of 69 nitrogen flow on the crystal structure and optical properties of TaN. It was shown that for 70 low N2 flow (1 and 2 sccm), the TaN films show good conductor performance; a further 71 increase of N2 flow shows non-metallic behavior. At the same time, samples with similar 72 structural properties have similar behaviors in terms of optical properties. Waechtler et al. 73 have been shown that SE can be used to examine the optical properties of Ta and TaN 74 thin films from 75 nm to 380 nm thickness. They have found a good agreement of optical 75 properties with narrow-band data available for similar thin films. It was also shown that 76 the optical properties of the films strongly depends with both substrate and film thick-77 ness [27]. Ma et al. have been studied the temperature-dependent dielectric function of 78 TiN films by SE [28]. Recently, Xu et al. [31] used the method of comparing the measured 79 refractive index of the low-k film under the Ta(N) diffusion barrier with the refractive 80 index of the blank low-k film to study the integrity of the Ta(N) diffusion barrier using 81 the approach developed by Shamiryan et al. [32]. 82 However, there has been limited study of the optical behavior of the different 83 stoichiometric of thicker TaN films by SE with varying N2 flow rates in conjunction with 84 electronic, structural and chemical composition. The study of the optical properties of 85 TaN can provide us with more information about TaN films and some potential possi-86 bilities for the development of new applications. So, the systematic study of TaN films for 87 understanding the electrical, structural, chemical composition and optical properties are 88 required. 89 Therefore, we first focus our attention on the optical properties of TaN films by SE, a 90 non-destructive testing method. By examining the optical properties of the sample, it can 91 provide some guidance for the deposited sample. At the same time, from the existing 92 research on the optical properties of TaN films, people ignore the influence of refractive 93 index (n) and extinction coefficient (k), let alone explore the influence of process condi-94 tions on them, but they are also important optical parameters. Because the refractive in-95 dex (n) and extinction coefficient (k) are regarded as "fingerprints of thin film materials." 96 Then, the effects of deposition rate and N2 flow on Ta(N) films deposition on the electrical, 97 structure, elemental composition and optical properties (n & k) of TaN films were stud-98 ied by using the four-probe method, X-ray diffraction (XRD), X-ray photoelectron spec-99 3 of 17 troscopy. The observation of different phases and chemical composition evaluation ob-100 served by XRD, XPS are correlated with optical properties. 101

102
Ta(N) films were synthesized using a standard magnetron sputtering (JS35-80G) 103 system with sputtering non-uniformity is ≤ ±5%. The Ta(N) films were deposited on Si 104 (100) wafers using magnetron sputtering of a Ta target (8.0 cm in diameter and 6.0 mm 105 thick) of 99.95% purity. The substrate holder (located in the center of the chamber) was a 106 25 cm diameter plate, with rotation set to 10 rpm without heating the substrate for all of 107 the depositions to improve the uniformity of the films. The target to substrate distance 108 was 13.0 cm, and a negative bias was applied to the Ta target. After placing the Si(100) 109 substrate in the deposition chamber, the chamber was evacuated to 9.6×10 -4 Pa (by a tur-110 bo-molecular pump), the background vacuum was sufficient to ensure the vacuum re-111 quired for Ta(N) film sputtering. Ar (99.999% purity) and N2 (99.999% purity) were in-112 troduced into the reaction chamber through a mass flow controller and used as sputter-113 ing and reaction gases, respectively. The Ta target and the Si-substrate were sputter 114 cleaned with Ar plasma prior to the Ta(N) films deposition for 5 min. Following cleaning, 115 Ta(N) film was deposited at 9.56×10 -4 Pa background pressure and 200 W DC applied 116 power in a mixture of Ar and N2. 117 To study the influence of different N2 flow rates on the properties of the sputtered 118 Ta(N) film, in all sputtering processes, the flow rate of Ar was fixed at 58 sccm with var-119 ious N2 flow rates from 0 to 58 sccm. The ratio of the reactive gas (N2, 0-58 sccm) to the 120 sputter gas (argon, 58 sccm) were varied from 0, 0.20, 0.29, 0.43, 0.65, 1.00. The deposition 121 time was also varied from 10 to 30 min to study whether the thickness of the film affects 122 the optical and other properties of Ta(N) under each N2 flow rate. Among them, when the 123 N2 flow is 12 sccm and 17 sccm, we only sputtered for 10 min respectively. 124 The thickness of the film was measured by a German Bruker Dektak step meter. 125 The thickness was determined from the step height between the film and a masked sub-126 strate area. The resistivity of the sample was measured by a double-electric four-point 127 resistance resistivity tester (FT-341) and was obtained from the current between two ex-128 ternal probes and measuring the voltage through the internal probes. 129 The crystallographic structure of the sputtered Ta(N) films was measured by an 130 x-ray diffractometer (Rigaku Ultima IV, Japan) using a θ-2θ scan with a 1.54 Å wave-131 length Cu Kα radiation, at room temperature, working at 40 kV and 30 mA, and record 132 the diffraction intensity in the scattering angle range of 20-60°. 133 X-ray photoelectron spectroscopy (XPS) was used to investigate the elemental 134 composition and chemical states in Ta(N) films using PHI 5300 (PerkinElmer, USA) with 135 an Mg Kα (1253.6 eV) excitation source. This source was operated using a voltage of 12.5 136 kV and an emission current of 20 mA. The films were sputter cleaned in an Ar + envi-137 ronment with 89.45 eV pass energy for 5 min prior to measurement. Survey scans were 138 conducted in the 0-1100 eV range. 139 The optical characterization of the films was carried out using a SENpro spectro-140 scopic ellipsometer (SENTECH, Germany) to measure ellipse deflection phase (Δ) and 141 amplitude (Ψ) angles at an incident angle of 70° in the spectral range from 400 nm to 1050 142 nm at 5 nm increments. In all cases, the ellipsometric data were processed using Spec-143 traRay/3 software for the data analysis. By treating the Ψ−∆ spectra, the refractive index 144 (n) and extinction coefficient (k) of the corresponding Ta(N) films were extracted by us-145 ing E. Kondoh ELLIPSHEET [33] for an infinitely thick film. The validity of this approach 146 will be proved below in the ellipsometry part of this paper. The specific details and film 147 thickness are given in Table 1. The deposition rate of Ta(N) films, sheet resistance and resistivity depend on the 154 flow of nitrogen. The film thickness decreases with increasing the flow of nitrogen, as 155 shown in Figure 1a and the increase of sputtering time from 10 min to 30 min makes the 156 film thicker. However, the deposition rate does not depend on the sputtering time (Fig-157 ure 1b). With the increase of nitrogen flow in the sputtering atmosphere of ionized Ar+, 158 the intensity of ion bombardment of Ta target decreases due to the reduction of mean free 159 path length and the number of sputtered Ta atoms also decreases leading to a gradual 160 reduction in deposition rate [34]. In addition, since there are a large number of active N 161 atoms in the sputtering atmosphere (an increase of the N2 flow rate), the number of active 162 N atoms in the atmosphere gradually increases, which increases the chemical reaction 163 between active N atoms and the surface of the Ta target. The probability of TaN com-164 pound causes slight poisoning of the target [35,36], thereby reducing the sputtering rate. The electrical resistivity of pure Ta films sputtered for 10, 20 and 30 min are nearly 168 same (193.5, 197.4, and 193.9 µΩ-cm, respectively (Table 1)). It is interesting to notice that 169 the measured resistivity for pure Ta films are similar to the values reported for tetragonal 170 crystalline Ta (β-Ta) films (165 µΩ-cm, Schauer et al. [37]; 210 µΩ-cm, Cuong et al. [38]) 171 and 242 µΩ-cm, Arshi et al. [39]). Therefore, our pure Ta films are most likely β-Ta. The 172 electrical resistivity of Ta(N) films deposited with different nitrogen flows is shown in 173 Figure 2. It can be seen that the resistivity of TaN is higher than the resistivity of pure Ta 174 observed at zero nitrogen flow. Introduction of nitrogen increases resistivity: first, it 175 changes linearly (embedded graph) and thinner film has lower resistivity. The trend of 176 increased resistivity can be attributed to the decreasing of the low resistivity Ta phase in 177 the deposited Ta-N films and to the increasing of the low resistivity N-rich TaN phase. A 178 much more dramatic change of resistivity is observed in the films deposited with 58 sccm 179 N2 flow, especially after 30 min deposition (relatively thick films, see Table 1). Further increase of N2 flow increases the resistivity of Ta(N) film and might depend on 184 the formation of different surface topography, grain size, changes of composition, 185 amorphous structure formation and defects/imperfection (scattering from grain bound-186 ary) [40]. The resistivity of TaN films in the current work is similar to the values observed 187 in the literature [40,41]. When the N2 flow rate is 12 sccm, the resistivity of the sputtered 188 TaN film is increased to 524 µΩ-cm and close to FCC TaN [42][43][44][45], or cubic TaN(111) [39][40][41][42][43][44][45][46][47][48]. After a further increase in the nitrogen flow to 25 sccm, the re-190 sistivity of the sputtered TaN film is close to the resistivity of Ta3N5 (1126 µΩ-cm, [39]). 191 Remarkably, when the increase of N2 flow was increased to 38 and 58 sccm, the resistivity 192 of the sputtered TaN films drastically increased, with only exception of 10 min deposition 193 TaN films. Normally, it can be explained by increasing electron scattering from intersti-194 tial N atoms. [39]. However, this model does not explain so strong difference between the 195 films deposited with 38 and 58 sccm [49]. It is also well-known that an excess of N2 flow 196 rate will decrease the mean free path of ionized Ta atoms and disturb the formation of 197 TaN structures [40,50] and also increased electron scattering from interstitial N atoms. 198 Therefore, the phase of the TaN film generated under our N2 flow rate will also be dif-199 ferent. The existence of N-rich phases in the TaN films at higher nitrogen flows is con-200 sistent with both the XRD patterns and the XPS analysis. 201

Structural Properties (XRD Analysis)
202 Figure 3 shows the X-ray diffraction (XRD) patterns of the Ta(N) films deposited for 203 10 min, 20 min, and 30 min with different N2 flows in the gas mixture varying from 0 to 204 58 sccm. The XRD spectra of the Ta films (N2 flow rate is 0) shows a mixed phase of β-Ta 205 (221), β-Ta (002), β-Ta (330), and Ta (110), among which β-Ta The diffraction intensity of 206 (002) is the highest, and the peak area of the diffraction peak is the largest, which indi-207 cates that the Ta film we sputtered is mainly β-Ta (002) preferred orientation (PDF#: 208 04-0788). Peaks at 35.40°, 37.04° and 41.20° are indexed to be the TaN (111), TaN (111) [39], 209 and TaN (200) (PDF#: 49-1283) structure respectively (Figure 3a). The peaks at 31.86°, and 210 35.10°corresponds to Ta3N5 (123) and Ta3N5 (130) or (040) compounds respectively [51]. 211 When the N2 flow rate is 12 and 17 sccm, a mixed phase of TaN (111) and TaN (200) 212 appear, and the diffraction peak of Ta3N5 (023) also appears when the N2 flow rate is 17 213 sccm. However, the diffraction peak of TaN (111) is the highest, and the diffraction peak 214 area is also the largest, which indicates that the preferred orientation of the TaN film 215 under these two flow rates is TaN (111). Similarly, when the N2 flow rate is 38 and 58 216 sccm, the TaN film has a preferred orientation of Ta3N5 (130,040). At the same time, when 217 the N2 flow rate increases, the diffraction peaks gradually widen, which indicates that the 218 grain size is gradually decreasing that leads to high resistivity Table 1 and correspond-219 ingly it could be attributed to the mixture of fcc TaN and amorphous structure [46,52,53]. 220 These phenomena observed by XRD are in line with our previous conjectures in the sec-221 tion on resistivity and similar to those reported in an earlier study [43,44]. 222 When the nitrogen flow rate is increased, the phase of the film evolved gradually 223 from TaN(111) to Ta3N5 (130) or Ta3N5 (040) (35.40° to 35.03°) [51]. On the other hand, the 224 TaN (200) peaks are gradually decreasing. A broad peak corresponding to Ta3N5 ap-225 peared for the 58 sccm sample and significant broadening of the peaks could be due to 226 the formation of a two-phase nanocomposite structure. This can be attributed to the high 227 nitrogen fraction, which is known to inhibit the crystallization of nitrogen rich TaNx 228 sputtered films. At the same time, the XRD patterns of TaN films with a nitrogen flow 229 rate of 25-58 sccm and a sputtering time of 10-30 min were compared (Figure 3b and 3c), 230 and it was found that the XRD patterns of TaN films with the same nitrogen flow rate but 231 different sputtering times/different thicknesses did not change much, which shows that 232 the thickness of the film does not affect the formation of the crystal structure of the film. 233 Generally, the films deposited with 58 sccm of nitrogen do not have pronounced patterns 234 and this suggests that the films are losing their crystalline structure and becoming more 235 amorphous (

241
XPS spectra was obtained to ascertain the elemental composition of the deposited 242 TaN films. Figure 4 illustrates an evolution of the XPS survey spectrum of the deposited 243 TaN films as a function of N2 flow rate in the binding energy range of 0-1150 eV and 244 show the Ta, O, N and C signals. It is confirmed that the survey does not contain addi-245 tional component arises in the Si2p spectrum because our films were thicker. The unex-246 pected O and C signal in these spectra might come from the ambient atmosphere in the 247 sputtering chamber and/or the presence of background oxygen in the chamber during 248 sputtering, and/or from organic residues during the storage, as already reported in liter-249 atures [35,39,54], also depends on different sputtering instruments, although there is a 250 chemical shift for both O and C in the lower binding energy with increasing the N2 flow 251 rates due to the formation of Ta2O5 (Figure 5a and 5b).  The XPS core-level spectra of Ta4f, Ta4d and N1s for different N2 flow rates (12-58 258 sccm) were shown in Figure 6a, Figure 6b and Figure 6c, respectively. Figure 6a shows 259 the XPS region of Ta4f, revealing that it is composed of three overlaying bonding envi-260 ronments: Ta4f5/2 of Ta-Ox (Ta bonded with O), Ta4f7/2, 5/2 of Ta-N and Ta4f7/2 of Ta-N lo-261 cated at 30.25, 28.15, and 25.97 eV, respectively. As shown in Figure 6a, increasing the 262 flow of N2 from 12 to 58 sccm is likely to chemical shift the Ta4f7/2 peaks from 25.97 to 263 25.78 eV, which were attributed to TaN Ta4f5/2 peaks. On the other hand, the chemical 264 shift from 30.25 to 29.81 revealing that O-rich films composed of Ta2O5 with an increase of 265 N2 flows were observed. Figure 7 shows the deconvoluted spectra of Ta4f for the TaN 266 films with N2 flow of 12 sccm. Compared with the Ta binding energy values of TaN and 267 TaOx in references, the binding energy values in the Ta4f spectrum (Ta4f7/2 = 25.7 eV and 268 Ta4f5/2 = 27.7 eV) accorded with the chemical state of Ta in Ta-O binding [55]. The Ta4f 269 doublet at binding energy Ta4f7/2 = 27.3 eV and Ta4f5/2 = 29.0 eV matched Ta 5+ state in 270 Ta2O5 [56,57], while the corresponding Ta4f doublet peaks are located respectively at 271 Ta4f7/2 = ~25.1 eV and Ta4f5/2 = ~27.3 eV that should be attributed to N-rich TaN phase 272 [58,59]. The existence of the Ta-N bonding in the film is further confirmed by the N1s peak 279 located around 397.0 eV [60]. As the nitrogen flow rate increases, more Ta-N bonds form 280 and the N1s peak also increases. This is consistent with the XRD spectra. It can be seen 281 from Figure 6c [35,39], while the binding energies for the Ta4f doublets 287 also shift toward lower energy values and do not agree with previous studies. 288

289
Spectroscopy ellipsometry (SE) is broadly used as an important tool for thin films 290 thickness, refractive indices and optical properties analysis, and its basic principle is 291 shown elsewhere [28]. For optically thick films, SE spectra can be fitted using a single 292 layer (substrate model) and optical properties of the films can further be analyzed. As the 293 refractive index (n) and extinction coefficient (k) cannot be measured directly and so 294 must be calculated by some quantities that are related to them and can be directly meas-295 ured. In this study, ellipsometry parameters Ψ and Δ were obtained by SE with Spec-296 trumfit Levenberg-Marquardt + thickness scan fitting. Therefore, we measured the Ψ and 297 Δ of all Ta films and TaN films by using only the substrate model, and extracted the n 298 and k of the corresponding film by using E. Kondoh ELLIPSHEET [33]. 299 Figures 8a and 8b depict the complex refractive index (n) and extinction coefficient 300 (k), respectively for Ta films and with a comparison of the literature report. After com-301 paring with the data of Tompkins et al. [63]. and Waechtler et al. [27], it is found that the 302 changing trends of n and k of Ta films with different thicknesses are roughly the same, 303 and the thickness does have an effect on the optical properties of Ta films. It is different 304 from the conclusion that Waechtler et al. reported the same n, k of Ta films with different 305 thicknesses. al. [2] and Waechtler et al. [27]. Also, it is found that when the N2 flow rate is 12 and 313 17 sccm, the n and k of our sputtered TaN film are close to the values in the reference, 314 but as the N2 flow rate increases, the n and k values of our sputtered TaN film are 315 getting less and less close to the n and k values in the references ( Fig. 9a and b). Be-316 sides, different N2 flow rates and different thicknesses of TaN films have significant 317 differences in n and k, which shows that both N2 flow rate and thickness affect the 318 optical properties of TaN. In addition, we can see that there are differences in the n 319 and k values and curve shapes of TaN films with different crystal structures. The in-320 fluence of N2 flow rate on the n and k of TaN film may be caused by the different 321 crystal structures and grain sizes of TaN film deposited with different N2 flow rates. 322 It is interesting that the extinction coefficient of TaN films deposited at high nitrogen 323 flows (38 and 58 sccm) decreases starting from 700 nm and especially from 900 nm. 324 The reduction of extinction coefficient suggests that the films are becoming more like 325 dielectric and explains the drastic increase of their resistivity ( Figure 2). 326 In this work, we used optically thick films to be able to extract optical charac-327 teristics. The thickness of these films are measured by using a Bruker Dektak step 328 meter that also is used only for relatively thick film. However, many applications of 329 Ta and TaN layers need very low thickness. For instance, when they are used as dif-330 fusion barriers in advanced microelectronics. Taking it into account we examined the 331 applicability of ellipsometry to measure the thickness of the thin Ta and TaN layers. 332 For this purpose, we used values of optical characteristics of these layers found in 333 literature and measured in our work. Then we calculated Δ-Ψ trajectories for the 334 films with the deferent thickness ( Figure 10). The presented curves demonstrate that 335 Ta(N) films thickness can be measured by ellipsometry when d ≤ 100 nm. The sensi-336 tivity is reduced with thickness but it can be very good for evaluation of Ta(N) films 337 deposited as diffusion barriers for microelectronics application (d < 10 nm). It is also 338 obvious that ellipsometry will be efficient for evaluation of the d < 100 nm films con-339 tinuity as it was done in the Ref. [31,32]. The solid curves were calculated by using optical characteristics at 633 nm measured in our films deposited with 12 sccm N2. The 342 dashed curves are based on optical characteristics reported in Ref. [27] (633 nm, Ar to N2 ratio = 4:1, the flow rate is not reported).

344
The Ta(N) film with different N2 flow rate (0-58sccm) and sputtering time of 345 10-30 min was deposited by the DC reactive magnetron sputtering method, and it 346 was found that the deposition rate Ta(N) film, electrical, structural, chemical and 347 optical properties depend on N2 flow rate. As the N2 flow rate increases from 0 to 58 348 sccm, the crystal structure of the sputtered film transitions from β-Ta to TaN(111) 349 and finally becomes the N-rich phase Ta3N5 (130) or Ta3N5 (040). When the N2 flow 350 rate increases, the diffraction peaks gradually widen, which indicates that the grain 351 size is decreasing that's leads to higher resistivity (Table 1) and correspondingly it 352 could be attributed to the mixture of fcc TaN and amorphous structure The films 353 deposited with 58 sccm of nitrogen lose specific crystallographic patterns and 354 therefore becoming amorphous. In the part of the optical properties study, we can 355 see that both the thickness and the N2 flow rate will affect the refractive index (n) and 356 extinction coefficient (k) of TaN film, and have a greater impact on k. The curve 357 shapes of n and k of similar crystal structures have a small difference. The influence 358 of N2 flow rate on the refractive index and extinction coefficient of TaN film may be 359 caused by the different crystal structures and grain size of TaN film deposited with 360 different N2 flow rates. The extinction coefficient of the films deposited with 58 sccm of 361 nitrogen decreases that suggests the formation of a more dielectric like nature of the de-362 posited films. This fact explains the drastic increase the resistivity of the films shown in 363 Figure 2. The reason for the transformation to the dielectric state is the incorporation of 364 Ta oxide that can be seen from XPS data. When the nitrogen flow is so large, it reduces 365 the free path length of sputtered atoms, poisons the Ta target and therefore reduces the 366 effective deposition rate. As a result, the role of the residual oxygen is drastically in-367 creased and the deposited film is becoming Tantalum oxynitride with much higher re-368 sistivity. This effect is becoming more pronounced when the deposition time is long and 369 this is the reason of the strong difference in resistivity between the films deposited during 370 10 min and 30 min. If deposition time is short, the target poisoning might be negligible. 371 Finally, the presented results suggest the resistivity of TaN films deposited by 372 magnetron sputtering can be precisely controlled by changing nitrogen concentration 373 during deposition. It is also important that a too high concentration of nitrogen might 374 have a negative effect increasing the impact of gas phase impurities.