Tungsten contamination in ion implantation

Nuclear Instruments and Methods in Physics Research B 377 (2016) 99–104

Contents lists available at ScienceDirect

Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb

Tungsten contamination in ion implantation M.L. Polignano ⇑, F. Barbarossa, A. Galbiati, D. Magni, I. Mica STMicroelectronics, Via Olivetti, 2, 20864 Agrate Brianza (MB), Italy

a r t i c l e

i n f o

Article history: Received 29 February 2016 Received in revised form 11 April 2016 Accepted 11 April 2016 Available online 19 April 2016 Keywords: Implantation Contamination DLTS Tungsten

a b s t r a c t In this paper the tungsten contamination in ion implantation processes is studied by DLTS analysis both in typical operating conditions and after contamination of the implanter by implantation of wafers with an exposed tungsten layer. Of course the contaminant concentration is orders of magnitude higher after contamination of the implanter, but in addition our data show that different mechanisms are active in a not contaminated and in a contaminated implanter. A moderate tungsten contamination is observed also in a not contaminated implanter, however in that case contamination is completely not energetic and can be effectively screened by a very thin oxide. On the contrary, the contamination due to an implantation in a previously contaminated implanter is reduced but not suppressed even by a relatively thick screen oxide. The comparison with SRIM calculations confirms that the observed deep penetration of the contaminant cannot be explained by a plain sputtering mechanism. Ó 2016 Published by Elsevier B.V.

1. Introduction

2. Experimental details

Sputtering is a very common contamination mechanism in ion implantation [1], because material sputtered by the ion beam from parts close to the wafer can easily reach the wafer surface. Examples of this sort are the dopant cross-contamination [2] and the iron contamination in implantations with heavy ions [3,4]. Implantations of wafers with exposed metal layers are suspected to be responsible for contamination of the implantation disk and as a consequence for contamination of wafers in subsequent implantations. In a recent paper [5] we showed that the implantation of wafers with an exposed tungsten layer is responsible for tungsten contamination of wafers implanted later. We set up a procedure to quantify tungsten contamination in ion implantation processes by DLTS (Deep Level Transient Spectroscopy), and used this procedure to evaluate the tungsten contamination and the efficiency of a decontamination process by implantation of dummy wafers. In the present paper we investigate more deeply this contamination, specifically the effect of the implantation energy, of the screen oxide and of the equipment setting. In a comparison among various techniques for contamination monitoring it was shown that DLTS is the best choice for slow diffusers [6], and for this reason we used DLTS in our study.

2.1. Sample preparation

⇑ Corresponding author. E-mail address: [email protected] (M.L. Polignano). http://dx.doi.org/10.1016/j.nimb.2016.04.026 0168-583X/Ó 2016 Published by Elsevier B.V.

P-type, (100), 200 mm diameter, 725 lm thick 10 X cm resistivity Magnetic Czochralski (MCZ) wafers were used in this study. Arsenic is known to be responsible for a relevant contamination by sputtering [3,4], and therefore arsenic implantations (1015 cm
2, 60 keV) were used to monitor the contamination of the equipment after previous implantations of wafers with an exposed metal layer (tungsten). The implantations were performed in an Axcelis NV-GSD200EE/80 High Current ion implanter. The implanter is equipped with an Extended Life Source (ELS) source type, a 2D-SOLID silicon coated disk, and a Plasma Electron Flood (PEF) electron shower. PEF is used to avoid charging effects during ion implantation. This device is placed close to the process chamber and generates low energy electrons that are drawn into the beam. Low energy electrons are generated in a molybdenum arc chamber by ionization of a xenon gas by electrons coming from a tungsten filament. Electrons are extracted toward the beam (plasma bridge) with an energy 62.1 eV, depending on the implant application. The amount of emitted electrons depends on the beam potential (selfextraction) and assists in reducing the beam space-charge, the beam divergence and the beam potential before reaching the wafer. The contamination induced by implantation was analyzed both under ordinary not-contaminated operation conditions and after contamination of the implanter by implantations of wafers with an exposed tungsten layer. The efficiency of a screen oxide layer (up to 150 Å thick) in reducing contamination was tested in two

100

M.L. Polignano et al. / Nuclear Instruments and Methods in Physics Research B 377 (2016) 99–104

Table 1 Process flow used to prepare the samples for contamination monitoring experiments.

different process flows, i.e. by etching the oxide off before or after the thermal treatment. The ion energy is expected to be a relevant parameter in contamination by sputtering, and for this reason the impact of energy reduction (from 60 to 10 keV) was studied. The role of the PEF device was studied by varying the PEF arc voltage and the extraction voltage. The flow used for the sample preparation is schematically reported in Table 1. After implantation, the metal surface concentration was measured by TXRF. Then, the wafers were cleaned by a conventional SC1–SC2 cleaning (Standard Cleaning 1 and Standard Cleaning2, [7]) and thermally treated by a Rapid Thermal Process (RTP) at 1100 °C for 3 min in an inert environment. This thermal treatment had the purpose to allow the contaminant diffusion in silicon. Then, the wafers received a Reactive Ion Etching (RIE) of 1.2–1.4 lm silicon with the aim to remove the doped layer. To obtain Schottky diodes for DLTS measurements, the native oxide was etched off and 1000 Å titanium layer was deposited on the silicon surface, masked and etched.

2.2. Experimental techniques TXRF measurements were obtained by a RigakuTXRF300 instrument in direct mode, with the high energy beam at 0.05° angle and 500 s acquisition time for each point. The tool has a W filament working at 30 kV and 300 mA. 1 mm2 area Schottky diodes were measured by DLTS. A Semilab DLS-83D instrument was used. In this instrument, lock-in integration is used for averaging capacitance transients, and temperature can be scanned in the range 30–300 K. Alternatively, constant temperature spectra can be obtained as a function of the frequency of

excitation pulses in the range 0.5–2 kHz [8]. Both methods were used in this work. The differential DLTS method was used. In this method, the Schottky diode (or the p–n junction) is reverse biased at a voltage Vr and two filling pulses are applied: the first pulse V1 is applied at the beginning of the lock-in integration period, and a second pulse V2 is applied a half period later. In the lock-in integration, the difference DC is obtained between the integrals of the capacitance transients caused by the first pulse and by the second pulse. The differential DLTS method can also be used to obtain the in-depth trap concentration profile. Indeed, this method yields the trap concentration in the interval [xd(V1), xd(V2)], where xd is the depletion region edge at a given reverse voltage. So by selecting appropriate values for V1, V2, and Vr, the trap concentration can be measured as a function of depth. In our measurements, samples were reverse biased at
5 V, and filling pulses with amplitudes of
0.5 V and
4.5 V were applied with a pulse width of 20 ls during each integration period. Under these conditions, a region ranging from 0.8 lm to 2 lm was analyzed. The spectra shown in this paper were obtained with a 23 Hz filling pulse frequency. When measuring the trap concentration profile, V1
V2 was set at 0.5 V, with 6 V reverse bias. In any case, the region close to the surface cannot be analyzed by DLTS. In concentration profile measurements, V1 can be positive to shrink the depletion region and acquire the trap concentration as close as possible to the surface; however, under our operating conditions reliable concentration data are obtained starting about 0.5 lm from the surface, down to 2.4 lm. Due to the sample preparation procedure, these depths correspond to 1.8–3.7 lm from the original wafer surface. The reverse voltage cannot be further increased because of Schottky diode leakage current issues, so the analysis cannot go deeper into silicon. The profile analysis is possible when the total trap concentration is high enough, and for this reason this analysis was carried out when the concentration estimated from the DLTS peak was larger than 1012 cm
3.

3. Experimental results 3.1. Not contaminated equipment Fig. 1(a) shows the DLTS spectra of a wafer implanted in a not contaminated implanter with 1015 cm
2 arsenic and 60 keV energy on the bare silicon surface. In these implantations the PEF arc voltage was 23 V and the extraction voltage was 2 kV. Two low concentration peaks are observed, and the corresponding Arrhenius plots of ep/T2 (where ep is the hole emission rate and T is the absolute temperature measurement) are shown in Fig. 1(b). These peaks can be identified with molybdenum and tungsten by comparison with literature data [9,10], and are therefore labeled ‘‘H1–Mo” and ‘‘H1–W”, respectively. Fig. 2 reports the DLTS spectra of samples implanted with PEF arc voltage in the range 15–40 V and 2 kV PEF extraction voltage, and of samples implanted with PEF off. The samples implanted with arc voltage up to 23 V essentially have the same DLTS spectrum as the samples implanted with PEF off. Vice versa, an implantation with 40 V arc voltage results in a significant increase of the tungsten concentration, while the molybdenum concentration is unaffected. In addition, in the sample implanted with 40 V arc voltage one more peak is observed, located between H1, Mo and H1, W. This peak was previously observed [5] in tungsten-contaminated samples, and is here labeled H2, W. The peak H2, W will be better discussed in Section 3.2, where samples with a higher concentration of tungsten are analyzed, so that the tungsten peaks can be more reliably separated and identified. According to the data in [5], the concentration of H2, W is expected to be about 10–20%

101

M.L. Polignano et al. / Nuclear Instruments and Methods in Physics Research B 377 (2016) 99–104

1.5

T (K) 10

250

200

150

W, Fujisaki et al. Mo, Benton et al.

H1,W

1.0

H1,W H1,Mo

H1,Mo

10

-2

10

-3

10

-4

-1

11

-2

ep/T (s K )

-3

Nt(10 cm )

300

-1

2

0.5

0.0 100

150

200

250

3

4

5

-1

6

7

1000/T (K )

T (K)

(a)

(b) 15

Fig. 1. DLTS spectrum (a) and Arrhenius plot of the observed peaks (b) in a sample implanted with 10

cm
2 arsenic dose in a not contaminated implanter.

12 10

PEF OFF 15 V arc voltage 23 V arc voltage 40 V arc voltage

6

no screen oxide 24 A screen oxide

H1,W

8

11

-3

As, 1015cm-2, 60 keV, 40 V arc voltage

H1,W

8

Nt(10 cm )

-2

Nt(1011cm-3)

10

15

As, 10 cm , 60 keV

4 2

6 4 2

H2,W

H2,W 0

H1,Mo 100

150

0 200

150

100

250

200

of the concentration of H1, W, hence H2, W can only be observed when the total tungsten concentration is high enough, for instance in the 40 V arc voltage sample in Fig. 2. The data in Fig. 2 show that increasing the PEF arc voltage to 40 V enhances the tungsten contamination, and for this reason this setting is used in this section to analyze the tungsten concentration as a function of the screen oxide thickness and of the PEF extraction voltage. Fig. 3 compares the DLTS spectra of 1015 cm
2 arsenic dose implanted samples with 40 V PEF arc voltage and 2 kV extraction voltage, with no screen oxide and with 24 Å screen oxide. The screen oxide was etched off before the thermal treatment. These data show that a very thin screen oxide is enough to suppress the tungsten contamination. Finally, Fig. 4 reports the DLTS spectra of 1015 cm
2 arsenic dose implanted samples with 40 V PEF arc voltage and the PEF extraction voltage in the range 0–10 kV. These data show that the tungsten contamination is reduced (but not suppressed) by increasing the PEF extraction voltage. These data show that arsenic implantation is responsible for a moderate but detectable molybdenum and tungsten contamination. Tungsten contamination increases when the PEF arc voltage is increased to 40 V, suggesting that this contamination may be due to sputtering from the PEF emitter by the xenon ions. Vice versa, molybdenum contamination is not affected by the PEF parameters, and may come from the ion source. The tungsten

Fig. 3. DLTS spectra of 1015 cm
2 arsenic dose implanted samples with 40 V PEF arc voltage with and without a screen oxide layer.

Nt(1011cm-3)

Fig. 2. DLTS spectra of 1015 cm
2 arsenic dose implanted samples with 2 kV PEF extraction voltage and arc voltage in the range 15–40 V.

250

T (K)

T (K)

15

-2

15

As, 10 cm , 60 keV, 40 V arc voltage

10

PEF extraction voltage: 0 kV 2 kV 10 kV

H1,W

5 H2,W

0 100

150

200

250

T (K) Fig. 4. DLTS spectra of 1015 cm
2 arsenic dose implanted samples with 40 V PEF arc voltage and with PEF extraction voltage in the range 0–10 kV.

contamination is suppressed by implantation through a thin screen oxide, showing that tungsten is brought to the wafer with a very low energy, which is compatible with a sputtering mechanism [3,4]. Finally, the dependence of the tungsten concentration on the PEF extraction voltage suggests that tungsten is sputtered in the form of positive ions, which are repelled by the positive extraction voltage.

102

M.L. Polignano et al. / Nuclear Instruments and Methods in Physics Research B 377 (2016) 99–104

3.2. Implantations in a previously contaminated implanter

10-1

250

200

150

W, Fujisaki et al.

W,H1 W,H2

10-2

3.2.1. The impact of a screen oxide It was previously shown [3,4] that iron contamination by sputtering is effectively suppressed by implanting through a screen oxide. In that case, the screen oxide layer was effective in suppressing contamination even if the oxide was not removed before the thermal treatment that follows the implantation, because the diffusivity of iron in the oxide was low enough to prevent diffusion from the oxide into silicon. Here, we tested the efficiency of a screen oxide layer in reducing the tungsten contamination coming from arsenic implantations in a previously contaminated implanter. Fig. 5 shows the DLTS spectra of wafers implanted after contamination of the implanter with 1015 cm
2 arsenic dose and 60 keV energy with different screen oxide thicknesses. In these wafers the oxide thickness was etched off before RTP. The spectra in Fig. 5 are quite similar to those reported in [5], as well as to those in Figs. 3 and 4, though now the peak intensity is much higher. Fig. 6 shows the Arrhenius plots of ep/T2 (where ep is the hole emission rate and T the absolute temperature) related to the peaks in Fig. 5. Literature data for tungsten [9] are also reported for a comparison, and confirm that the level W, H1 can be identified with the tungsten-related hole trap, while the level W, H2 does not correspond to any tungsten-related level. However it was shown [5] that W, H2 increases in proportion to W, H1, and that W, H2 is formed upon the RIE silicon etching used to remove the doped layer, so we assumed that W, H2 is due to a complex between tungsten and some element introduced by RIE. These data show that the tungsten contamination of samples implanted after the implantation of wafers with an exposed tungsten layer is sensitively increased with respect to the level detected in wafers implanted in a non-contaminated implanter, as expected (compare Figs. 1 and 5). The spectra in Fig. 5 show that the tungsten concentration is significantly reduced by a screen oxide layer, however tungsten contamination is not suppressed, even if 150 Å oxide is used.

As, 1015cm-2, 60 keV before contamination after contamination: no screen oxide 50 A screen oxide 100 A screen oxide 150 A screen oxide

12

-3

Nt(10 cm )

6

10-4

3

4

5

-1

6

7

1000/T (K ) 2

Fig. 6. Arrhenius plot of ep/T obtained from the samples in Fig. 1. Literature data are shown for a comparison.

Fig. 7 compares the tungsten concentration measured by DLTS as a function of the screen oxide thickness in samples where the screen oxide was etched off before or after RTP. Etching the oxide before drive-in results in a moderate reduction of the contaminant concentration in samples with thin oxides only, showing that tungsten diffusion from the oxide into silicon adds a second-order contribution to the contamination of these samples.

2 W,H2 0 150

-3

W,H1

4

100

10

Fig. 7. Tungsten concentration measured by DLTS in samples implanted with arsenic in a tungsten-contaminated implanter as a function of the screen oxide thickness. Data from samples with the screen oxide etching before drive-in (empty dots) or after drive-in (solid dots) are shown.

10

8

2

-1

-2

ep/T (s K )

In this section the contamination resulting from arsenic implantations in a contaminated implanter is investigated. The implanter was previously contaminated by implanting wafers with an exposed tungsten layer. In these tests, 1015 cm
2 arsenic dose was implanted, the PEF arc voltage and extraction voltage were 23 V and 2 kV, respectively.

T (K) 300

200

250

T (K) Fig. 5. DLTS spectra of samples implanted with arsenic before contamination of the implanter and after contamination by implantation of wafers with an exposed tungsten layer, with the screen oxide thickness as the parameter.

3.2.2. Tungsten profiles measured by DLTS Fig. 8 reports the in-depth profile of the tungsten concentration, as measured by DLTS in a sample implanted with arsenic on bare silicon after contamination of the equipment. In this plot, the depth axis 0 was set at the original wafer surface. This profile shows some depletion in the region approaching the surface, suggesting partial segregation of tungsten at the wafer surface. Indeed, such segregation was previously observed by ToF-SIMS measurements when the tungsten dose exceeded 51010 cm
2 [5]. We remark that the

103

M.L. Polignano et al. / Nuclear Instruments and Methods in Physics Research B 377 (2016) 99–104

-2

10

13

10

12

-3

[W] (cm )

W-contaminated implanter

0

1

2

3

4

DEPTH( µm) Fig. 8. Tungsten concentration profile measured by DLTS in a sample implanted with arsenic on bare silicon after contamination of the implanter. The depth axis zero is the original wafer surface.

tungsten fraction diffusing as an interstitial impurity only can be observed by DLTS, because the fraction diffusing as tungsteninterstitial pairs is much slower and cannot reach the region analyzed by DLTS [11]. If a gaussian diffusion profile is assumed for pffiffiffiffiffiffiffiffiffiffiffiffiffi interstitial tungsten, the profile width would be 2DW i t  5 lm; where DWi is the interstitial tungsten diffusivity at the annealing temperature and t the annealing duration. We can analyze a portion of this profile only. 3.2.3. The impact of the implantation energy Table 2 reports the tungsten concentrations measured by DLTS in samples implanted with 60 keV or 10 keV arsenic after contamination of the implanter by implantation of wafers with an exposed tungsten layer. The data obtained in samples implanted on bare silicon or on 150 Å screen oxide are compared. The surface concentration of tungsten measured by TXRF after the implantation is also shown. These data show that DLTS is more effective than TXRF in revealing tungsten contamination in implanted wafers. These data also show that tungsten contamination is reduced by decreasing energy, and it is completely suppressed by low energy implantation through 150 Å screen oxide. However, with such implantation condition about 50% of the arsenic implanted dose remains in the screen oxide. 3.2.4. Experimental data vs. model In a previous work [4] iron contamination by sputtering was successfully modeled by using a binary collision code (TRIM, [12]) and by assuming that the contamination process can be schematically divided into two steps, i.e. sputtering of contaminating material and knock-on of this material into the target matrix. Here, we followed this approach and tried to estimate the depth reached by tungsten after arsenic implantation by simulating the knock-on of tungsten from a thin (10 Å) contaminated surface layer. The most recent version of the code (SRIM, [13]) was used

Table 2 Tungsten concentrations measured by DLTS in samples implanted after contamination of the implanter by implantation of wafers with an exposed tungsten layer, and surface concentration of tungsten measured by TXRF after the implantation. Energy (keV)

10 60

CW,DLTS (cm
3)

NW,TXRF (cm
2)

Bare silicon

150 Å

(3.3 ± 0.2)1012 (6–8)1012

621010 (5–9)1011

Below detection limit (8.1 ± 0.8)1010

in the present work. The total amount of tungsten per unit area was estimated from the tungsten concentration measured by DLTS in the samples with no screen oxide by using the calibration reported in [5]. Fig. 9 shows the tungsten knock-on profiles obtained for 1015 cm
2 arsenic implantation with 60 keV or 10 keV energy on 150 Å screen oxide. By integrating the profiles in Fig. 9, we estimated the tungsten dose that is expected to go through the screen oxide and reach silicon as a function of the screen oxide thickness. These doses were converted into the expected DLTS concentrations by the calibration in [5]. Fig. 10 compares the tungsten concentration measured by DLTS and the calculated concentration as a function of the oxide thickness, with the implantation energy as a parameter. The data measured by etching the screen oxide before the contaminant drive-in are considered here. This comparison shows that the tungsten concentration going through the screen oxide layer is strongly underestimated by the calculations. Therefore, it is suggested that a fraction of the contaminating tungsten receives some energy in the process, so in this case the contamination mechanism is more complex than initially assumed. 3.2.5. The role of the Plasma Electron Flood The role of the Plasma Electron Flood (PEF) device in tungsten contamination was also studied. In some tests PEF was deactivated, in other tests it was activated with 23 V arc voltage, and the extraction voltage was varied in the range 0–10 kV. This experiment was carried out on bare silicon wafers and on wafers with 150 Å screen oxide. When present, the oxide was etched off before drive-in. When comparing implantations carried out in a sequence after contamination of the equipment, it must be taken into account that a high dose arsenic implantation also involves some reduction of the equipment contamination by sputtering [5]. Therefore, contamination data obtained by subsequent implantations refer to different contamination levels of the implanter, and some correction is necessary to compare these data and determine the impact of the implanter settings on the contaminant concentration. Here, subsequent arsenic implantations with 1015 cm
2 dose and 60 keV energy were carried out after contamination of the implanter. Samples implanted with identical settings in the beginning and in the end of the sequence were used to estimate the contamination reduction due to each implantation step. The measured contaminant concentrations were then corrected by assuming that a constant fraction of the contaminant amount on the implanter is removed at each implantation.

Oxide

17

10

W concentration (cm-3)

15

As, 10 cm , 60 keV,

Silicon 15

-2

As, 10 cm W knock-on: 60 keV 10 keV

16

10

15

10

14

10

13

10

-150

-100

-50

0

50

100

DEPTH (A) Fig. 9. Tungsten knock-on profile calculated by SRIM for 1015 cm
2 arsenic implantation by assuming that tungsten contamination is initially deposited in 10 Å surface layer.

104

M.L. Polignano et al. / Nuclear Instruments and Methods in Physics Research B 377 (2016) 99–104

1013

[W]DLTS (1012cm-3)

60 keV 10 keV

1012

11

10

Calculations: 60 keV 10 keV 1010

0

50

100

150

200

tox (A) Fig. 10. Tungsten concentration measured by DLTS in samples implanted with 10 and 60 keV arsenic in a previously contaminated implanter, and concentrations calculated from the profiles in Fig. 3, as a function of the screen oxide thickness.

Table 3 Tungsten concentration measured by DLTS in samples implanted with 60 keV 1015 cm
2 arsenic dose in a previously contaminated implanter. PEF condition

Off On, 0 kV extraction voltage On, 2 kV extraction voltage On, 10 kV extraction voltage

Screen oxide thickness (Å) 0

150

5.01012 cm
3 6.51012 cm
3 7.41012 cm
3 6.91012 cm
3

4.41011 cm
3 4.71011 cm
3 6.01011 cm
3 5.21011 cm
3

½W150A oxide ½WBare silicon

0.09 0.07 0.08 0.08

Table 3 collects the tungsten concentration data measured by DLTS in samples implanted after contamination of the equipment with no screen oxide and with 150 Å screen oxide, when the PEF extraction voltage is varied and when PEF is off. The ratio between the tungsten concentrations measured in samples implanted on the bare silicon surface and with 150 Å screen oxide is also shown. The tungsten concentration is significantly decreased by a screen oxide, as already observed. In addition, in this case no significant trend is observed with the PEF extraction voltage, opposite to what observed before contamination of the equipment (see Fig. 4). The tungsten contamination is found to be lower when the PEF is off, however the fraction of tungsten that goes through the screen oxide is essentially the same irrespective of PEF activation and settings. 4. Conclusions The tungsten contamination in ion implantation processes has been studied both in typical operating conditions and after contamination of the implanter by implantation of wafers with an exposed tungsten layer. Of course the contaminant concentration

is orders of magnitude higher after contamination of the implanter, but in addition our data show that different mechanisms are active in a not contaminated and in a contaminated implanter. A moderate tungsten contamination is observed also in a not contaminated implanter, however in that case contamination is completely not energetic and can be effectively screened by a very thin oxide (24 Å). On the contrary, the contamination due to an implantation in a previously contaminated implanter is reduced but not suppressed even by a relatively thick screen oxide (150 Å). The comparison with SRIM calculations confirms that the observed deep penetration of the contaminant cannot be explained by a plain sputtering mechanism. Unfortunately, our data do not allow us to formulate a model to explain the observed phenomenology. We also remark that the dependence of the tungsten concentration on the equipment settings is different in the two cases. In a not contaminated implanter, the tungsten contamination decreases with increasing the PEF extraction voltage, indicating that in this case tungsten actually comes from the PEF device and is in the form of positive ions. After contamination of the implanter, no significant trend with the PEF extraction voltage is observed, though contamination is lower if PEF is off. References [1] H. Ryssel, L. Frey, Contamination problems in ion implantation, in: J.F. Ziegler (Ed.), Handbook of Ion Implantation Technology, North-Holland, Amsterdam, 1992, p. 675. [2] J.D. Bernstein, A.W. Alvarez, E.B. Benton,, K.C. Cherukuri, C.M. Otten, Characterization of boron and phosphorus surface contamination in high current ion implantation, Proc. of the XIV International Conference on Ion Implantation Technology, IEEE, 2002, p. 177. [3] M.L. Polignano, C. Bresolin, F. Cazzaniga, A. Sabbadini, G. Queirolo, Investigation of metal contamination by photocurrent measurements: validation and application to ion implantation processes, in: J.K. Lowell, R.T. Chen, J.P. Mathur (Eds.), Optical Characterization Techniques for High Performance Microelectronic Device Manufacturing II, Proc. SPIE 2638, 1995, p. 14. [4] M.L. Polignano, D. Caputo, A. Giussani, V. Soncini, G. Di Toma, Metal contamination reduction in the evolution of ion implantation technology, Proc. of the XIII International Conference on Ion Implantation Technology, IEEE, 2000, p. 686. [5] M.L. Polignano, F. Barbarossa, A. Galbiati, S. Grasso, I. Mica, V. Soncini, Contamination by slow diffusers in ion implantation processes: the examples of molybdenum and tungsten, Nucl. Instr. Meth. B 356–357 (2015) 164–171. [6] M.L. Polignano, D. Codegoni, S. Grasso, I. Mica, G. Borionetti, A. Nutsch, A comparative analysis of different measurement techniques to monitor metal and organic contamination in silicon device processing, Phys. Status Solidi A 212 (2015) 495. [7] W. Kern, Handbook of Semiconductor Wafer Cleaning Technology, Noyes Publications, Park Ridge, NJ, USA, 1993, p. 20. [8] G. Ferenczi, C.A. Londos, T. Pavelka, M. Somogyi, A. Mertend, Correlation of the concentration of the carbon-associated radiation damage levels with the total carbon concentration in silicon, J. Appl. Phys. 63 (1988) 183. [9] Y. Fujisaki, T. Ando, H. Kozuka, Y. Takano, Characterization of tungsten-related deep levels in bulk silicon crystal, J. Appl. Phys. 63 (1988) 2304. [10] J.L. Benton, D.C. Jacobson, B. Jackson, J.A. Johnson, T. Boone, D.J. Eaglesham, F.A. Stevie, J. Becerro, Behavior of molybdenum in silicon evaluated for integrated circuit processing, J. Electrochem. Soc. 146 (1999) 1929. [11] A. De Luca, A. Portavoce, M. Texier, C. Grosjean, N. Burle, V. Oison, B. Pichaud, Tungsten diffusion in silicon, J. Appl. Phys. 115 (2014) 013501. [12] J.P. Biersack, L.G. Haggmark, A Monte Carlo computer program for the transport of energetic ions in amorphous targets, Nucl. Instr. Meth. B 174 (1980) 257. [13] J.F. Ziegler, M.D. Ziegler, J.P. Biersack, SRIM – the stopping and range of ions in matter (2010), Nucl. Instr. Meth. Phys. Res. B 268 (2010) 1818.