Experiments on plasma immersion ion implantation inside conducting tubes embedded in an external magnetic field

Applied Surface Science 357 (2015) 1438–1443

Contents lists available at ScienceDirect

Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc

Experiments on plasma immersion ion implantation inside conducting tubes embedded in an external magnetic field E.J.D.M. Pillaca a,∗ , M. Ueda a , H. Reuther b , L. Pichon c , C.M. Lepienski d a

Associated Laboratory of Plasma, National Institute for Space Research, S. J. Campos, S. Paulo, Brazil Institute of Ion Beam Physics and Materials Research, P. O. Box 510109, 01314, Dresden, Germany Institut Pprime-UPR 3346 CNRS-Université de Poitiers-ENSMA, France d Departamento de Física, UFPR, CP 19044, 81531-990, Curitiba, PR, Brazil b c

a r t i c l e

i n f o

Article history: Received 6 April 2015 Received in revised form 21 September 2015 Accepted 25 September 2015 Available online 30 September 2015 Keywords: PIII with magnetic field PIII inside conducting tubes Plasma immersion ion implantation

a b s t r a c t Tubes of stainless steel (SS) embedded in external magnetic field were used to study the effects of plasma immersion ion implantation (PIII) as a function of their diameter. The study was complemented with and without a grounded auxiliary electrode (AE) placed at the axis of the tube. During the discharge tests in tubes of larger diameter (D = 11 cm), with and without AE, nitrogen gas breakdown was established inside the tube at pressures near 2.0 × 10−2 mbar. Under the same operation conditions, stable plasmas with similar PIII current densities were obtained for both arrangements. Reducing the diameter of the tube (D = 1.5 cm) turned the plasma unstable and made it inappropriate for ion implantation. This situation was solved by supplying gas at higher pressure or using higher magnetic field, without the presence of an AE. Under these conditions, nitrogen PIII treatments of these small diameter tubes were performed but gave not the best implantation results yet. Our results have also shown higher ion implantation current density (16 mA/cm2 ) in tube of intermediate diameter (D = 4 cm) using AE, compared to largest diameter tube used. In this case, a thick nitrogen layer of about 9 ␮m was obtained in the SS sample placed inside the tube. As a consequence of this, its structural and mechanical properties were enhanced. These results are attributed to the thermal diffusion promoted by ions hitting the inner wall in a large number due to the presence of the AE and the magnetic field. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Treatments inside metallic tubes are being widely studied nowadays due to its broad utility and practical importance for fluid transports, in general. Actually, several methods of surface treatment are being used to enhance the resistance against corrosion and wear in tubes. Among the several techniques used for this goal, treatment based on plasma such as plasma immersion ion implantation stands out. Plasma immersion ion implantation (PIII) is a well-established method, used mainly for the three-dimensional surface modification of materials [1–3]. However, in the treatments of workpieces for industrial applications with concave geometry, such as piston rings, tubes, pipes, etc., poor implantation has been obtained so far [4]. It was pointed out in the literature that a lower ion energy than expected occurs during ion implantation. Sheridan has attributed

∗ Corresponding author. Tel.: +55 (12) 32086698. E-mail addresses: [email protected], [email protected] (E.J.D.M. Pillaca). http://dx.doi.org/10.1016/j.apsusc.2015.09.210 0169-4332/© 2015 Elsevier B.V. All rights reserved.

this behavior to the decrease of the applied electric potential which is related with  the important scale length called the ion-matrix overlap, d =



−4 ∈0 /en0



and the bore radius, R [5,6]. It has

also been shown that by inserting a grounded conductive auxiliary electrode (AE) along the axis, the average ion impact energy can be recovered [7]. However, this is not heartening when AE is tried inside tubes with small R due to problems of electric insulation. Consequently, generation of plasma inside tubes with short diameters is difficult to take place and new methods as electron cyclotron resonance (ECR) microwave discharge were proposed [8]. Another important aspect studied in tubes was the dependence between the tube diameter (D) and its length (L), on plasma density [9]. The studies demonstrated that plasmas with critical densities occur at L » D. In the same sense, investigations about dose implanted inside tube were related to aspect ratio D/L. Numerical investigation revealed that ions coming from outside tube contribute to the implantation [10,11]. In this study, it was shown that ions pass through the middle of the tube and arrive at the end of it when short length, L ∼ D, is used. That behavior does not occur if tubes have long enough length, L » D. Therefore, effects of the

E.J.D.M. Pillaca et al. / Applied Surface Science 357 (2015) 1438–1443

ions implanted nearby the end of the tube promote a non-uniform dose along the inside of the tube. This situation is critical, thus additional setup were proposed to improve it [8,9,12]. By use of a grounded cylindrical grid electrode inside tube, reasonably uniform dose was obtained [9]; however, its use is limited to short diameter. This restriction is surpassed by the combined application of ECR and magnetic field [8,12]. In this method, the plasma is made to move along the axis direction during treatment to obtain a uniform dose; however a complicated system and intense magnetic field are required. Knowledge of these issues is of great importance to try achieving implantation with great uniformity but it still remains a challenge to date. In particular, the use of a simpler external magnetic field can be an option in order to overcome part of these limitations. Recently, application of magnetic bottle configuration in PIII in materials proved to be more suitable for the improvements of the surface properties in relation to the conventional PIII [13–15]. In this process, a system of crossed E × B fields is produced around the target by application of transverse magnetic and electric fields during the PIII. The use of this configuration allows obtaining high plasma density due to intense background gas ionization by the trapped electrons [16]. Thus, a high ion flux hitting the target is obtained making it possible to rise the temperature on it. As a consequence, the mechanism of thermal diffusion is set-up, allowing thick nitrogen layer, at a relatively short treatment time compared to other conventional methods used for the same goal. In fact, this advantage is now being used for the formation of DLC film [17] and for the treatments inside conducting tubes [18]. Taking advantage of these recently obtained improvements, in this work, PIII process using E × B fields was tested in tubes of Stainless Steel with different diameters.

2. Experimental details The experimental arrangement of PIII with tubes embedded in external magnetic field is shown in Fig. 1. The experiment was carried out in a cylindrical vacuum vessel with 38-cm length and 26-cm diameter. For generation of the axial magnetic field, four magnetic coils were mounted on the outer side of the PIII cylindrical chamber. The coils were wound to form a magnetic bottle type configuration, that is to say, a region with one magnetic field minimum in the center of the chamber and maximums near the coils. The point of minimum field was chosen to coincide with the place where the tubes were positioned. Three tubes made of stainless steel (SS) of 15 cm-length and diameters of 11 cm (larger), 4 cm (medium) and 1.5 cm (smaller) were prepared for the experiments, including the possibility of positioning an AE of 0.4 cm-diameter inside. The tubes were connected electrically to the high voltage feed through (see Fig. 1) whereas the AE was grounded. To monitor the effects of ion implantation on the inner wall of the tube, Silicon and SS samples were prepared. SS304 samples were chosen due to its similar composition of the tubes whereas Silicon samples with (1 0 0) orientation were used for comparison purpose with SS samples. SS disks with 0.3-cm thickness and diameters of 1.0 cm and 1.5 cm were sand papered using silicon carbide paper in a series of 350, 500 and 1200 grit and subsequently a 0.5 ␮m alumina in liquid solution was used for final polishing. We used 0.5 mm thick and 1.5 cm × 1.5 cm pieces of commercial Si wafer which was subsequently chemically cleaned. Afterwards, SS and Si samples were cleaned in ultrasound with acetone and deionized water before treatment. After this, seven SS and Si samples were mounted on Sample Holder (SH) of 15-cm length and then placed on the tube inner wall for the tubes with medium and large diameter. For the tube with smaller diameter,

1439

Fig. 1. Schematic diagram of the PIII experimental setup.

additional arrangements were mounted outside it to place the SH. The base pressure set by mechanical and diffusion pumping was about 4.0 × 10−5 mbar, while nitrogen working pressure of 2–4 × 10−2 mbar was used throughout the treatment. Negative voltage pulses of 6 kV amplitude, 20 ␮s duration and 500 Hz frequency were applied at the tube for ion implantation. The total current on the tube was measured with a Rogowsky coil. Depending on the diameter of the tube, the magnetic field was set between 45 G and 90 G. Finally, the PIII treatments were performed for 60 min. Experiments in smaller tube using AE had caused some difficulties (instability of the plasma turn on) during the test of electrical discharge. Due to this, results for D = 1.5 cm with AE are not exhibited in this work. Treated and untreated SS samples were analyzed using different characterization techniques. These were performed in samples placed in the middle of the inner surface of the tube. Nitrogen concentration profiles versus depth were measured according to thickness. For thicker layers: Glow Discharge Optical Emission Spectroscopy (GDOES) was performed using a Jobin–Yvon–Horiba GD-Profiler at a sputtering rate of 0.07 ␮m s−1 . For thinner layer: high resolution Auger Emission Spectroscopy (AES) was performed using an equipment from FISONS Instruments Surface Science, model MICROLAB 310-F, at a sputtering rate of 0.3 nm s−1 . The structural changes in the surface layer were investigated by X-ray diffraction (XRD) in a Bragg–Brentano geometry with CuK˛ radiation. The surface hardness of the samples was obtained using a Nanoindenter XP (MTS Instruments). An infrared thermometer Micron model M90 with range between 250 ◦ C and 2000 ◦ C was used to monitoring the temperature on the tube, during PIII. The characteristic voltage and current wave-forms were recorded using a digital oscilloscope Tektronix model TDS360 for further analysis. 3. Results and discussions Prior to ion implantation inside the tubes of large, medium and small diameters, some aspects related to the breakdown and maintenance of the discharge for PIII embedded in magnetic field were analyzed, for cases with and without AE. The first test was conducted in tube of D = 11 cm without AE. Here, the voltage and the magnetic field were kept at 6 kV and B = 45 G, respectively, to find the gas breakdown point by gradually increasing the gas pressure from 8 × 10−3 mbar to 6 × 10−2 mbar. For p = 8 × 10−3 mbar, formation of the plasma occurred outside the tube, preferentially. This effect can be explained by the efficient ionization of the residual gas promoted by the special distribution of the E × B fields surrounding the tube [16,19]. By increasing the gas pressure to up to near 10−2 mbar, an unstable plasma was obtained

1440

E.J.D.M. Pillaca et al. / Applied Surface Science 357 (2015) 1438–1443

Table 1 Results of measurements of the PIII current density and temperature in the tubes at the end of their treatment time, with and without AE. Quantities

j (mA/cm2 ) T (◦ C)

D = 11 cm

D = 4 cm

D = 1.5 cm

Without AE

With AE

Without AE

With AE

Without AE

With AE

7.33 < 250

7.72 <250

10.6 340

15.9 475

57 440

– –

inside the tube. At p > 4 × 10−2 mbar it became very stable and with higher brightness. When an AE was inserted inside the tube, under the same conditions, the plasma was also turned on. By simple visual inspection, it could not be distinguished from the case without AE. At this point, one important characteristics related to the discharge current was noticed which is reported in Table 1. The PIII current density measured for both cases was about 7 mA/cm2 . This value is very high compared with the 0.97 mA/cm2 measured for the case of the presence of plasma outside the tube. These results suggest no significant contribution of AE in PIII current, in tubes of D = 11 cm. In order to evaluate the behavior of the electric discharge in tube of D = 4 cm, the voltage and the magnetic field were kept at 4 kV (due to the limitation of the current from the high voltage pulser) and B = 65 G, respectively. At p = 6 × 10−3 mbar, the plasma is formed preferably outside the tube whereas an unstable plasma was obtained inside it. It became stable when the pressure was increased to p > 4.0 × 10−2 mbar. Other is the situation when AE is placed inside the tube. To obtain the gas breakdown inside the tube, a low magnetic field of 45 G and a low gas pressure of 6 × 10−3 mbar were necessary. An intense brightness was observed when the pressure was increased up to 5 × 10−2 mbar. In this case, plasma filling all the chamber volume was also seen but with much lower brightness. In this condition, a high current, of about 19 mA/cm2 , was measured and reported in Table 1. It means 1.5 times more than that obtained without AE. In comparison, this high current is twice more than for the case of D = 11 cm. The discharge conditions used above do not work for the gas breakdown in tubes of D = 1.5 cm without AE. In this case, a high magnetic field of 85 G and a high gas pressure of 6 × 10−2 mbar were necessary to obtain a stable discharge inside the tube. In this condition, a current density of about 57 mA/cm2 was measured. So far, our results indicate that the plasma formation inside the tube with and without AE is very sensitive to the gas pressure and to the magnetic field. This behavior is in agreement with investigations of the Paschen curve for a standard PIII inside tubes [20]. This research indicated that plasma generation ocurred outside the tube instead of inside it, at low pressure. However, with the application of the magnetic field, as in our case, a significant generation of gas ionization can be caused by magnetized drifting electrons. This mechanism is interesting because it is the main responsible for the gas breakdown and the maintenance of the PIII current by simple increase of gas pressure or by the increase of the intensity of the magnetic field, especially in small tube. Another important feature in our system is the aspect ratio D/L of the tube, such as suggested by Liu et al. [9] for conventional PIII. Their studies demonstrated that plasma density inside the cylinder, with or without AE, increases with the aspect ratio D/L. When D/L approaches unity, the plasma density exhibits similar values, with or without the electrode. This feature is interesting because it coincides with our results, since the discharge current is proportional to plasma density, for the larger tube (D/L = 0.7 cm) a current of about 7 mA/cm2 was measured for the cases with and without AE. On the other hand, even though a situation of very low current density was expected in medium diameter tubes (D/L = 0.3) [9], it did not happen. On the contrary, a moderate difference of about

1.5 times was noticed by inserting the AE (see Table 1). This value is not very high in relation to the case without AE but it is enough to cause significant changes in the surface properties of the samples, as is discussed below. According to these results, the role of AE inside the tube in presence of magnetic field is very clear now. We believe that in the case of the tube with D = 11 cm, the AE is somehow shielded by the plasma. Then an AE with larger diameter is necessary to show some positive effect, such as suggested in [21]. Continuing the studies in tubes of D = 11 cm, in the presence of an AE: in the absence of magnetic field, a strong heating of the AE rod (which turned red hot) was noticed during the experiment. This singular behavior can be attributed to the emission of a great number of secondary electrons towards the center, coming from the tube inner surface during the ion bombardment. In fact, it is a direct manifestation of change in the electric potential configuration when an AE is placed inside the tube [22]. This phenomenon could help to provide heating of the interior of the tube, for auxiliary heating of the samples or workpieces placed there. However, the continuous bombarding by secondary electrons on the AE can be dangerous during the experiment. Depending on the material of the center rod, it could result in the melting of AE. By application of the magnetic field (few tens of gauss), the heating of the AE was reverted. Of course, in this case, the flux of secondary electrons was radially suppressed limiting their motion to reach the AE. They are rapidly expelled to outside the tube by the presence of Bz and Ez fields. Other effect noticed in this experiment with magnetic field was the presence of the Hull cutoff magnetic field. Above the cutoff, the plasma was not observed inside the tube for a given intensity of magnetic field. This mechanism was also observed by other researchers in other type of applications [23,24]. Since the cutoff magnetic field is inversely proportional to the tube diameter [23], higher B is expected to be required for tubes with minor diameters. In our experiments, the cutoff happened at about B = 55 G for D = 11 cm with and without AE. For D < 11 cm, the Hull cutoff was not possible to be observed due to the limitation of our magnetic field source able to supply 130 G, at most. By simple calculations using reference [24], for D = 4 cm, about 264 G is required for the cutoff, whilst for D = 1.5 cm, about 749 G is necessary. As part of the present study, PIII treatments inside the tube with and without AE were carried out to evaluate the retained doses for tubes of different diameters embedded in magnetic field. Results of GDOES and AES characterizations in SS samples are shown in Fig. 2. AES nitrogen profiles for treated SS samples in tubes of D = 11 cm is shown in Fig. 2(a). The AES profile for treatment without AE shows a nitrogen concentration of about 5 at.% at the surface. This value falls lightly down to 3 at.%, remaining constant until reaching 120 nm in depth. After this, the nitrogen concentration falls abruptly to zero. This depth is four times larger than that one with AE, as can be confirmed in the same figure. It was not an expected result because the energy of ion inside the tube should have increased in the presence of AE [7]. This apparent discrepancy was solved by using AES results in silicon samples treated together with SS samples. Here (AES profiles are not shown), a large depth of about 60 nm was obtained with AE, meaning twice as deep as the case without AE. This result indicates that the thermal diffusion is larger in tube treated without AE. We believe that the bombardment of ions inside the tube without AE happens preferentially on the tube bore by the edge effect. Comparing these results with the one reported in [25] for the standard PIII, we can notice some interesting differences. For example, in such treatment, a similar depth of about 40 nm for treatment with and without AE was obtained by using 15 kV. However, in our case, this depth is largely overcome (about four times) when magnetic field is used and in low voltage. The reason for these differences is the working pressure used during the treatment, as was discussed above.

E.J.D.M. Pillaca et al. / Applied Surface Science 357 (2015) 1438–1443

Fig. 2. AES and GDOES profiles: (a) AES for D = 11 cm; (b) GDOES for D = 4 cm; (c) AES for D = 1.5 cm.

Fig. 2(b) displays the GDOES profiles of SS samples implanted in tube of D = 4 cm. Here, a high N concentration of about 20 at.% was reached at the surface, in the presence of AE. This extends up to about 9 ␮m into the sample which finally falls down to zero concentration. A remarkable characteristic in this treatment was the thick layer achieved by the presence of high current density of about 16 mA/cm2 during 60 min of treatment (see Table 1). During this time, the temperature reached 475 ◦ C, causing strong thermal diffusion. This high PIII current obtained can be associated with the high nitrogen ion flux hitting the inner tube surface which was

1441

enhanced by the magnetic field and the use of AE. On the other hand, in the same Fig. 2(b), a magnification of the GDOES profile near the surface of SS sample treated without AE is shown. A nitrogen concentration of about 30 at.% was possible to be reached at the surface. Compared to the case with AE, it shows a thin layer of about 0.05 ␮m. In this treatment, the highest temperature measured was 340 ◦ C. It means 1.4 times less that the case with AE. Consequently, it was not enough to induce large thermal diffusion in SS samples. In conclusion, PIII treatment in tubes of D = 4 cm, with E × B fields, the presence of AE is mandatory to reach large retained dose, ensuring thick layer by high thermal diffusion during the PIII treatment. Nextly, AES results of SS samples treated in tubes of D = 1.5 cm, without AE, is displayed in Fig. 2(c). In this case, an N concentration with maximum peak of 16 at.% was obtained at the surface. This concentration quickly decays down to 8 at.% and the N penetration reached 20 nm from the surface. Continuing in Fig. 2(c), N concentration decreases slowly from 8 at.% down to zero. In total, a nitrogen penetration depth of about 160 nm was achieved. Here, a temperature of 440 ◦ C was measured after 60 min of treatment. This high temperature can be attributed to the very high current of 57 mA/cm2 measured during the enhanced PIII process in the presence of the magnetic field. Due to this fact, the penetration depth was larger than for the case of D = 11 cm (with and without AE) and D = 4 cm (without AE). However, this penetration remains smaller than the one observed for D = 4 cm with AE, yet. We believe that ions bombarding the inside wall of the tube had lower energy, such as suggested in [5]. The mechanisms involved in this case are not yet clear and must be studied in more detail. On the other hand, as was discussed previously above, PIII treatment in tubes with this dimension is hard to be performed due to the fact that plasma inside small tube is difficult to be obtained. Other method as coaxial electron cyclotron resonance (ECR) discharge could be used for this goal [8,12,26]. In our case, the results obtained so far, demonstrated that PIII treatment in tubes with D = 1.5 cm worked reasonably well, if no AE was used. XRD profiles of reference and implanted samples for tubes of D = 11 cm, D = 4.0 cm and D = 1.5 cm are shown in Fig. 3. XRD results in Fig. 3(a) for samples treated in tubes of D = 11 cm with and without AE show similar diffractograms in comparison to the untreated one. However, a notable difference related to intensities of the peaks from samples treated inside the tube of D = 4 cm with AE can be seen in Fig. 3(b). There, peaks of  N , CrN and ␣ phase are present next to the principal peak ( 111 ). These peaks are usually related to the changes of the crystal lattice of SS samples due to intense nitrogen ion implantation [27]. The high intensity of  N peak is in agreement with the large implanted dose of nitrogen shown by GDOES results. Formation of CrN resulted from precipitation of Cr due to the high temperature of 475 ◦ C reached at 60 min of treatment (see Table 1), as predicted in [28]. In this case, the excessive temperature rise can be avoided by control of the magnetic field intensity to lower values [19]. For the treatment of PIII in the same diameter tube, without AE, a thinner nitrogen layer of 0.05 ␮m was not enough to be detected by XRD. On the other hand, XRD results for D = 1.5 cm tube (Fig. 3(c)) show a minimal change near the principal peak  111 . This is in agreement with the AES result where nitrogen penetrated only about 160 nm. However, the temperature achieved of 440 ◦ C in this case did not promote the chromium nitride precipitation. Having completed a detailed study of the ion implantation into SS samples with respect to the retained doses and modified structures for different diameter tubes, PIII effects on hardness were studied for the best implantation case. The hardness profile as a function of depth in the SS treated samples for tubes of D = 4 cm, with AE, is shown in Fig. 4. Presence of very thick nitrogen layer has enhanced the surface hardness in about fivefold (16 GPa) in relation to the untreated one. This extraordinary result in hardness

1442

E.J.D.M. Pillaca et al. / Applied Surface Science 357 (2015) 1438–1443

Fig. 4. Hardness profile for stainless steel samples after 60 min of treatment at tube of D = 4 cm.

value is twice as superior compared with the untreated one. In comparison to the case with AE, the hardness decreases slowly from surface to 500 nm in depth, reaching about 6 GPa. After this, the hardness remains constant throughout the measurements. On the other hand, observing again the Fig. 4, we can notice that the hardness for the case with AE begins with a low value contrary to the case without AE. Perhaps some coating was formed during the PIII treatment with AE. Nevertheless, this reduced value of hardness near the very surface is insignificant compared to 16 GPa seen for as deep as 2500 nm obtained by this treatment. According to our results, high hardness is associated with the large retained dose of nitrogen. Then, for D = 11 cm and D = 1.5 cm, hardness is expected to be lower due to their very thin nitrogen layer. 4. Conclusions

Fig. 3. XRD diffractograms: (a) for D = 11 cm; (b) for D = 4 cm; (c) for D = 1.5 cm.

is extended for more than 2500 nm from the sample surface. This is as consequence of the presence of the large nitrogen layer of about 9 ␮m. In fact, these important results are attributed to the presence of the thermal diffusion caused by the high flux of ions with high bombarding energy of the inner wall of tube in presence of both the AE and the magnetic field [15]. For the case without AE, a hardness of about 9 GPa on the surface was measured. This

Nitrogen PIII processing inside SS tubes of different diameters, with and without AE, embedded in an external magnetic field was studied. One feature found in this work was that stable gas breakdown inside tube was preferably established above pressures of 2.0 × 10−2 mbar, for all cases studied. In addition to this result, the importance of using the magnetic field as a method for the control of secondary electrons during the treatment of tubes with AE was demonstrated. In relation to treatment of tubes of different diameters, our result in SS samples placed inside them showed that the presence of AE in tubes of larger diameter is not beneficial for PIII treatment because smaller doses were obtained when compared with the case without AE. However, presence of AE inside the tube with medium diameter tube was essential to obtain a large nitrogen layer of about 9 ␮m in a short time of treatment. This was attributed to both mechanisms: implantation and thermal diffusion which was enhanced by the presence of the E x B configuration. As a result, the surface hardness was improved by about five times in this case due to the structural change suffered. For the case of tubes of smaller diameter without AE, a similar nitrogen layer thickness compared to the case of larger diameter was obtained. We conclude that PIII treatment inside the tube was enhanced by the presence of the magnetic field for all three size tubes, as well as, by the introduction of an AE for medium diameter one. References [1] J.R. Conrad, J.L. Radtke, R.A. Dodd, F.J. Worzala, N.C. Tran, Plasma source ion-implantation technique for surface modification of materials, J. Appl. Phys. 62 (1987) 4591–4596, http://dx.doi.org/10.1063/1.339055. [2] A. Anders, Handbook of Plasma Immersion Ion Implantation and Deposition, Jhon Wiley & Sons, Inc, 2000.

E.J.D.M. Pillaca et al. / Applied Surface Science 357 (2015) 1438–1443 [3] M.A. Lieberman, Model of plasma immersion ion implantation, J. Appl. Phys. 66 (1989) 2926–2929, http://dx.doi.org/10.1063/1.344172. [4] T.E. Sheridan, Ion-matrix sheath in a cylindrical bore, J. Appl. Phys. 74 (1993) 4903–4906, http://dx.doi.org/10.1063/1.354323. [5] T.E. Sheridan, Analytic theory of sheath expansion into a cylindrical bore, Phys. Plasmas. 3 (1996) 3507, http://dx.doi.org/10.1063/1.871501. [6] T.E. Sheridan, Sheath expansion into a large bore, J. Appl. Phys. 80 (1996) 66–69. [7] X. Zeng, B. Tang, P.K. Chu, Improving the plasma immersion ion implantation impact energy inside a cylindrical bore by using an auxiliary electrode, Appl. Phys. Lett. 69 (1996) 3815–3817, http://dx.doi.org/10.1063/1.117114. [8] K. Baba, R. Hatada, Ion implantation into inner wall surface of a 1-m-long steel tube by plasma source ion implantation, Surf. Coatings Technol. 128–129 (2000) 112–115, http://dx.doi.org/10.1016/S0257-8972(00)00584-3. [9] B. Liu, G. Zhang, D. Cheng, C. Liu, R. He, S.-Z. Yang, Inner surface coating of TiN by the grid-enhanced plasma source ion implantation technique, J. Vac. Sci. Technol. A Vacuum, Surfaces, Film 19 (2001) 2958, http://dx.doi.org/10.1116/ 1.1415356. [10] D.T.K. Kwok, X. Zeng, Q. Chen, P.K. Chu, T.E. Sheridan, Effects of tube length and radius for inner surface plasma immersion ion implantation using an auxiliary electrode, IEEE Trans. Plasma Sci. 27 (1999) 225–238, http://dx.doi. org/10.1109/27.763123. [11] T.E. Sheridan, T.K. Kwok, P.K. Chu, Kinetic model for plasma-based ion implantation of a short, cylindrical tube with auxiliary electrode, Appl. Phys. Lett. 72 (1998) 1826, http://dx.doi.org/10.1063/1.121188. [12] K. Baba, R. Hatada, Ion implantation into the interior surface of a steel tube by plasma source ion implantation, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 148 (1999) 69–73. [13] E.J.D.M. Pillaca, M. Ueda, K.G. Kostov, H. Reuther, Study of plasma immersion ion implantation into silicon substrate using magnetic mirror geometry, Appl. Surf. Sci. 258 (2012) 9564–9569, http://dx.doi.org/10.1016/j.apsusc.2012.05. 132. [14] E.J.D.M. Pillaca, M. Ueda, H. Reuther, C.M. Lepienski, Study of the effects of plasma immersion ion implantation on austenitic stainless steel using E × B fields, Surf. Coatings Technol. 246 (2014) 1–5, http://dx.doi.org/10.1016/j. surfcoat.2014.02.044. [15] E.J.D.M. Pillaca, M. Ueda, R.M. Oliveira, L. Pichon, Applied surface science, study of the effects of E × B fields as mechanism to carbon – nitrogen plasma immersion ion implantation on stainless steel samples, Appl. Surf. Sci. (2014), http://dx.doi.org/10.1016/j.apsusc.2014.04.003. [16] K.G. Kostov, M.A. Algatti, E.J.D.M. Pillaca, M.E. Kayama, R.P. Mota, R.Y. Honda, Effect of electron magnetic trapping in a plasma immersion ion implantation system, Eur. Phys. J. D. 54 (2009) 205–209, http://dx.doi.org/10.1140/epjd/ e2009-00036-6.

1443

[17] S. de Fátima Magalhães Mariano, E.J. de Dios Mitma Pillaca, M. Ueda, R. de Moraes Oliveira, Influence of the magnetic field on DLC coatings grown by plasma immersion ion implantation and deposition in crossed fields, Surf. Coatings Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.01. 012. [18] E.J. de, D.M. Pillaca, M. Ueda, S. de, F.M. Mariano, R. de Moraes Oliveira, Study of plasma immersion ion implantation inside a conducting tube using an E × B field configuration, Surf. Coatings Technol. 249 (2014) 104–108, http://dx.doi. org/10.1016/j.surfcoat.2014.03.042. [19] E.J.D.D.M. Pillaca, M. Ueda, K.G. Kostov, Investigation of plasma immersion ion implantation process in magnetic mirror geometry, IEEE Trans. Plasma Sci. 39 (2011) 3049–3055, http://dx.doi.org/10.1109/TPS. 2011.2160209. [20] Y. Miyagawa, M. Ikeyama, S. Miyagawa, M. Tanaka, H. Nakadate, Plasma analysis for the plasma immersion ion implantation processing by a PIC-MCC simulation, Comput. Phys. Commun. 177 (2007) 84–87, http://dx.doi.org/10. 1016/j.cpc.2007.02.034. [21] X. Zeng, T.K. Kwok, A. Liu, P.K. Chu, B. Tang, T.E. Sheridan, Plasma-immersion ion implantation of the interior surface of a small cylindrical bore using an auxiliary electrode for finite rise-time voltage pulses, IEEE Trans. Plasma Sci. 26 (1998) 175–180, http://dx.doi.org/10.1109/27.669623. [22] B.Y. Zeng, X.C. Kwok, T.K. Liu, A.G. Chu, P.K. Tang, Plasma immersion ion implantation of the interior surface of a large cylindrical bore using an auxiliary electrode, J. Appl. Phys. 83 (1998) 44–49. [23] Y.Y. Lau, Y.Y. Lau, P.J. Christenson, P.J. Christenson, D. Chernin, D. Chernin, Limiting current in a crossed-field gap, Phys. Fluids B Plasma Phys. 5 (1993) 4486, http://dx.doi.org/10.1063/1.860563. [24] M.T. Ngo, K.H. Schoenbach, G.A. Gerdin, J.H. Lee, Temporal development of hollow cathode discharges, IEEE Trans. Plasma Sci. 18 (1990) 669–676, http:// dx.doi.org/10.1109/27.55942. [25] A.G. Liu, X.F. Wang, B.Y. Tang, P.K. Chu, Dose and energy uniformity over inner surface in plasma immersion ion implantation, J. Appl. Phys. 84 (4) (1998) 1859–1862, http://dx.doi.org/10.1063/1.368344. [26] K. Baba, R. Hatada, Deposition of diamond-like carbon films on inner wall of sub-millimeter diameter steel tube by plasma source ion implantation, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. Mater. Atoms 206 (2003) 704–707, http://dx.doi.org/10.1016/S0168-583X(03)008383. [27] S. Mändl, B. Rauschenbach, Concentration dependent nitrogen diffusion coefficient in expanded austenite formed by ion implantation, J. Appl. Phys. 91 (2002) 9737–9742, http://dx.doi.org/10.1063/1.1479749. [28] D.I. Samandi, M. Shedden, B.A. Smith, Microstructure, corrosion and tribological behaviour of plasma immersion ion-implanted austenitic stainless steel, Surf. Coatings Technol. 59 (1993) 261–266.