Development of neutron depth profiling at CMRR

Development of neutron depth profiling at CMRR

Nuclear Instruments and Methods in Physics Research A 788 (2015) 1–4 Contents lists available at ScienceDirect Nuclear Instruments and Methods in Ph...

1015KB Sizes 76 Downloads 323 Views

Nuclear Instruments and Methods in Physics Research A 788 (2015) 1–4

Contents lists available at ScienceDirect

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

Development of neutron depth profiling at CMRR Run-dong Li, Xin Yang n, Guan-bo Wang, Hai-feng Dou, Da-zhi Qian, Shu-yu Wang Institute of Nuclear Physics and Chemistry, China Academy of Engineering Physics, Mianyang 621999, China

art ic l e i nf o

a b s t r a c t

Article history: Received 13 February 2015 Received in revised form 19 March 2015 Accepted 19 March 2015 Available online 27 March 2015

A neutron depth profiling (NDP) system has been developed at China Mianyang Research Reactor (CMRR) at Institute of Nuclear Physics and Chemistry (INPC), CAEP. The INPC-NDP system utilizes cold neutrons which are transported along the C1 neutron guide from the cold neutron source. It consists of a beam entrance, a target chamber, a beam stopper, and data acquisition electronics for charged particle pulse-height analysis. A 90 cm in diameter stainless steel target chamber was designed to control the positions of the sample and detector. The neutron beam intensity of 2.1  108 n cm  2 s  1 was calibrated by the Au foil activation method at the sample position. The INPC-NDP system was tested by using a Standard Reference Materials SRM-2137. The measured results agreed well with the reference values. & 2015 Elsevier B.V. All rights reserved.

Keywords: Neutron depth profiling Cold neutron CMRR SRM-2137

1. Introduction Neutron Depth Profiling (NDP) is a non-destructive technique for measuring the depth profiles of some light elements (e.g., 3He, 10B, 6 Li, etc.) in the near-surface region of a sample. The sample is illuminated by a thermal or cold neutron beam, and the concentration profile of the isotope of interest is inferred from the residual energy of ions produced by neutron reactions with the isotope of interest. The NDP technique was first proposed by Ziegler et al. [1]. The capabilities of NDP have been thoroughly investigated by Fink et al. Because of the simplicity and high sensitivity of the NDP technique, it has been further developed and extensively applied in surface analysis, solid state research, lithium ion batteries, and other material sciences [2–10]. Downing from NIST builds a web site to collect and introduce the information of NDP facilities and literature around the world [11]. Nowadays, there are several NDP facilities and most of them are in the U.S.A and Europe. A new Cold NDP was developed at HANARO in 2014 [10]. This paper presents a new neutron depth profiling facility developed at China Mianyang Research Reactor (CMRR) at Institute of Nuclear Physics and Chemistry (INPC), CAEP. The SRM-2137 sample was measured to validate the functionality of the INPC-NDP system.

2. INPC-NDP description The NDP facility is built at a cold neutron beam port of CMRR, which is shared with a cold neutron radiograph system. The beam n

Corresponding author. E-mail address: [email protected] (X. Yang).

http://dx.doi.org/10.1016/j.nima.2015.03.058 0168-9002/& 2015 Elsevier B.V. All rights reserved.

size is a square up to 4  4 cm. However, it is able to be adjusted. The eigen-wavelength of the neutron beam is 2 Å, and the neutron intensity at the sample surface is about 2.1  108 n cm  2 s  1 (thermal neutron equivalent), which is calibrated by using Au foils. The INPC-NDP consists of a beam entrance of 1.5 cm in diameter in a boron carbide block, a target chamber fixed on a supporting bracket, a beam stopper, and three separate charged particle detector systems with various distances and angles to the sample. The structure is illustrated in Fig. 1. 2.1. Target chamber The target chamber is composed of a body and a hemisphere top cover both made of 6 mm thick stainless steel. There is a plexiglass viewing window on the top cover, which will be covered during operation due to the fact that the visible light saturates the detector signals. The body and top cover are joined by an O-ring to provide vacuum seals. On the side of the chamber, there are two 1 mm thickness Aluminum windows for neutron beam in and out, and the diameters are respectively 1.5 cm and 8 cm. A small door is constructed to replace beam collimator and samples without opening the top cover. Three flanges are designed for the molecular vacuum pump and gauges. The air pressure is less than 1.0  10  3 Pa during operation. Two flanges are designed for detector cables. The chamber has additional flanges available for possible future applications. Photographic views of inside the chamber and of several individual components are shown in Fig. 2. The inner diameter of the chamber is 90 cm and the height is 60 cm. Three brackets are respectively for detectors, neutron beam collimator, and sample mounting. There are three separate detector systems, which enable them to independently acquire energy spectrum

2

R.-d. Li et al. / Nuclear Instruments and Methods in Physics Research A 788 (2015) 1–4

Fig. 1. Layout of the INPC-NDP structure.

Fig. 3. Photograph of the inside of the installed chamber.

2.2. Data acquisition electronics

Fig. 2. Inside of the target chamber.

data at the same time. Detector assembles are positioned into 5 mm thick Pb shield to attenuate γ-ray or other background particles and then mounted onto the detector bracket. The distance between the detectors and sample can be adjusted from 7 cm to 20 cm and the detector angles varied relative to the sample surface. The neutron beam approaches the sample at an angle of 45 degrees from the sample normal. A 30 cm long beam collimator is positioned at the beam inlet inside the chamber to define the size of the beam and decrease the neutron scattering in the chamber. It’s composed of boron carbide and lead. The boron carbide collimator is cyclindrical and has a tapered bore aligned with the sample. The collimator is wrapped with 5 mm of lead sheet to attenuate γ and ions produced by nuclear reactions. There are several collimators with various outlet diameters, 0.5 cm, 0.75 cm, and 1 cm. The sample bracket of 18 cm in diameter is made of 3.5 mm thick Aluminum, and apertures are designed to decrease the background. It is rotated by a stepper motor, and has four positions for sample mountings, which means that four samples can be mounted at the same time. Mountings of various shapes and sizes are designed for different samples. Thus rectangular or circular shaped samples varying in the size from 5 mm to 20 mm can be mounted. In addition, it’s able to perform a scan measurement for some larger samples by rotating the bracket. As shown in Fig. 3, the inside of the installed chamber can be viewed via the window on the top cover.

The electronics for the INPC-NDP consist of three detecting modules for the charged particle pulse-height analysis. Fig. 4 shows a block diagram of the data acquisition and analysis electronics. The CANBERRA Passivated Implanted Planar Silicon (PIPS) detectors are used to measure the energy spectrum of charged particles. The detectors employ a 150 mm2 or 300 mm2 active area with a depletion layer thickness of 100 μm. The dead layer thickness of the Si detector is less than 50 nm. The energy shifting of detectors is tested by measuring an 241Am α source. The α source is measured every half an hour, and the channel shifting of the energy peak is less than 0.05% for all detectors in 8 h, which indicates that the detectors have acceptable reliabilities and stabilities. The recommended bias for detectors is from 20 V to 60 V, whereas, the bias is often set to 20 V to decrease the depletion layer thickness, which further decreases the background, and the depletion layer is still thick enough to fully collect the energy of the charged particles.

3. Validation of the INPC-NDP 3.1. Spectra of boron samples Two boron samples were measured by INPC-NDP. One was manufactured by depositing 10 nm thick natural boron on 2  2 cm square and 0.5 mm thick silicon substrate, which was used for the channel-energy calibration. The other was a standard reference material 2137 (SRM-2137) from NIST, which was manufactured by implanting 10B ions into a 1  1 cm square and 1 mm thick silicon substrate. When 10B absorbs a thermal or cold neutron, two distinct energies are produced for each of the 4He and 7Li ions since the 7Li nucleus is produced either in the ground state or in an excited state as shown below. ( 10 1 0 n þ 5 B-

7 n 4 3 Li þ 2 Heð93:7%Þ 7 4 3 Li þ 2 Heð6:3%Þ:

R.-d. Li et al. / Nuclear Instruments and Methods in Physics Research A 788 (2015) 1–4

Canberra PIPS Detector

Preamp Canberra 2003BT

Linear Amp Canberra 2026

Preamp Canberra 2003BT

Linear Amp Canberra 2026

Preamp Canberra 2003BT

Linear Amp Canberra 2026

Preamp Canberra 2003BT

3

Bias Canberra 3102D Canberra PIPS Detector

Preamp Canberra 2003BT

ADC/MCA Canberra MP2-3E

PC

Bias Canberra 3102D Canberra PIPS Detector

Preamp Canberra 2003BT Bias Canberra 3102D

Fig. 4. A schematic block diagram of the data acquisition electronics.

3.2. Deconvolution of the spectrum

5000

1

Li

Counts

1

3000 2000 1000

Li

2

2

0 1000

1500

2000 2500 Channel

3000

3500

Fig. 5. Spectrum of the 10 nm thick Boron sample.

Table 1 Theoretical full energies and energies modified. Ion type

Full energies (keV)

Energies modified (keV)

Li1 Li2 α1 α2

839.635 1013.126 1471.763 1775.868

837.344 1010.700 1470.729 1774.937

1200

10

Gross Counts

10 Counts

1000 800

Net Counts

10 10 10 10 10

600

0

Li

1000

2000 3000 Channel

4000

400 200 0

The observed counts vector C(N  1) of the spectrum (background subtracted) is generally related to the concentration profile vector S(M  1) and the response matrix P(N  M), PS¼ C. It means that the sample is divided into M slice, and the channel number of the spectrum is N. Pij is the probability of an ion emerging at the jth slice and generating a count at channel i. The response matrix P is efficiently computed by MC-NDP code. For α1, the SRM-2137

Gross Counts Net Counts

4000

Counts

4 He ions are produced with an energy 1471.763 keV (α1) or 1775.868 keV (α2). The corresponding lithium ions have energies of 839.635 keV (Li1) or 1013.126 keV (Li2). Spectrum of the 10 nm thick boron sample is given in Fig. 5, which shows the result of a 300 mm2 detector directly facing and centered with the sample at a distance of 12 cm. The energy resolution of α1 is about 15 keV (FWHM). As Table 1 shows, the full energies of ions need to be modified due to the thickness of the boron film. A Monte Carlo code for NDP called MC-NDP [12] was used to modify the energies. MC-NDP is developed by the INPCNDP group and simulates the behavior of ions transmitted through a matrix such as silicon and generates the energy spectrum output of a detector. Ions are projected within a bounded conical direction within the active area of the detector. Sufficiently energetic ions are tracked through the sample (modeled by the computer program simulating ion transport in matter), to build an energy spectrum according to physical and electrical characteristics of the detector, even the dead layer. The stopping powers are from tables computed by CORTEO [13] using the SRIM [14,15] code. The ion transport module of MC-NDP takes advantages of both TRIM and CORTEO programs and uses the Impulse Approximation method [14] to determine the flight length of ions and the Index Interpolating technique [13,16] for the scattering angles between ions and nucleus. Thus, MC-NDP simulates the ion transport process much more efficiently and enables the efficient use of a Monte Carlo process to extract the detector response function for the NDP technique. The ion tracking method mentioned above is also applied in a specific purposed code, RSMC (Reaction sequence Monte Carlo), which focuses on the coupling problem of the deuteron (or triton) and neutron [17]. The spectrum of SRM-2137 is illustrated in Fig. 6. The lithium peaks are blurred by the high background which is due to radiation from other sources in the chamber. The signal from the boron is just low compared to the competing radiation sources. However, lithium peaks still can be recognized. The a1 spectrum (Fig. 7) is usually used to determine the concentration profile of 10 B.

1000

1500

2000

2500

3000

3500

Channel Fig. 6. Spectrum of the SRM-2137.

sample was divided into 35–10 nm slices, and there were 105 α1 ions which were simulated for each slice, and nevertheless, it required only half a minute to calculate P. Then, the objective of

4

R.-d. Li et al. / Nuclear Instruments and Methods in Physics Research A 788 (2015) 1–4

1000

Counts

800

1

deconvolution of the NDP technique were developed. The SRM2137 sample was measured to validate the INPC-NDP system. The determined depth profile of 10B agrees well with the reference values, which indicates that the measurement result of the INPCNDP system is reliable. The INPC-NDP facility will be further applied to investigate the lithium ion batteries, burnable boron layer on the surface of nuclear fuel elements, etc. However, there are still some issues needed to be improved for this facility. Thinner Al foils will be used as the neutron in/out windows and 6LiF will be used instead of B4C to absorb the scattering neutrons to further reduce the background noise.

of SRM-2137

600 400 200 0 1350

1400

1450

1500

1550

Energy[keV] Fig. 7. Spectrum of α1 of the SRM-2137.

10

B Concentration [atoms/cm ]

B Concentration [atoms/cm3]

10 x 10 8 6

Acknowledgment

19

Reference Measurement Reference

10

Measurement

10 10

0

0.1

0.2

0.3

4

This work was supported by the National Natural Science Foundation of China (11475152) and the Neutron Physics Laboratory Foundation of CAEP (2013BB01). The authors would like to acknowledge the helpful discussions and suggestions of the following individuals: Dr. R.G. Downing from NIST, Dr. J. Vacik, Dr. V. Hnatowicz and Dr. I. Tomandl from Academy of Sciences of the Czech Republic, and Dr. D. Fink from the Universidad Autonoma Metropolitana-Iztapalapa, Mexico.

2

References 0

0

0.05

0.1

0.15

Fig. 8. Concentration profile of

0.2

0.25

0.3

0.35

10

B in the SRM-2137.

deconvolution is to solve the linear equations PS¼C. Since the choice of N and M is a little arbitrary, the equations may be over-determined or under-determined, or even ill-posed. To deal with this problem, three algorithms are employed, Constrained Linear Regularization (CLR) method [18,19], Singular Value Decomposed solving the Least Square (SVDLS) method [18], Probability Iteration method [20]. The background subtracted spectrum can be deconvoluted directly without any smoothing or fitting, which will best retain the information of the spectrum. The areal density of the 10B can be calculated as 4πN0/0.937ΩfσAs, where N0 is the count rate of the α1 spectrum, Ω is the solid angle of the detector to the sample, f is the intensity of the neutron beam, σ is the micro cross-section of the concerned nuclear reaction, and As is the area of sample irradiated by the neutron beam. The concentration profile of 10B in SRM-2137 is shown in Fig. 8, and the Probability Iteration method was used for deconvolution. The peak depth of the deconvoluted NDP spectrum is 0.189 μm, and the areal density of 10B is (9.910 70.365)  1014 atom cm  2. The SRM reference values are respectively 0.188 μm and (1.018 7 0.035)  1015 atom cm  2. The relative difference of areal density between the measurement and the reference is  2.65%. 4. Conclusion INPC-NDP facility has been developed at CMRR. Besides designing and constructing the facility, codes for physical analysis and

[1] J.F. Ziegler, G.W. Cole, J.E.E. Baglin, Journal of Applied Physics 43 (1972) 3809. [2] R.G. Downing, R.F. Fleming, J.K. Langland, D.H. Vincent, Nuclear Instruments and Methods in Physics Research 218 (1983) 47. [3] D. Fink, J.P. Biersack, H.D. Carstanjen, F. Jahnel, K. Muller, H. Ryssel, A. Osei, Radiation Effects 77 (1983) 11. [4] D. Fink, J.P. Biersack, H. Liebl, In Ion Implantation: Equipment and Techniques, Springer-Verlag, Berlin, 1983. [5] V. Havránek, V. Hnatowicz, J. Kvítek, J. Vacík, J. Hoffmann, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 73 (1993) 523. [6] R.G. Downing, G.P. Lamaze, J.K. Langland, Journal of Research of the National Institute of Standards and Technology 98 (1993) 109. [7] K. Ünlü, B.W. Wehring, Detectors and Associated Equipment 353 (1994) 402. [8] J. Vacik, J. Cervena, V. Hnatowicz, V. Havra, D. Fink, Acta Physica Hungarica 75 (1994) 369. [9] S.M. Çetiner, K. Ünlü, R.G. Downing, Journal of Radioanalytical and Nuclear Chemistry 276 (2008) 623. [10] B. Park, G. Sun, H. Choi, Detectors and Associated Equipment 752 (2014) 20. [11] R.G. Downing, 〈https://sites.google.com/site/nistndp〉. [12] X. Yang, R. Downing, G.-b. Wang, D.-z. Qian, H.-g. Liu, K. Wang, A Monte, Journal of Radioanalytical and Nuclear Chemistry 301 (2014) 213. [13] F. Schiettekatte, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 266 (2008) 1880. [14] J.F. Ziegler, J.P. Biersack, M.D. Ziegler, SRIM The Stopping and Range of Ions in Matter, Lulu, Morrisville, 2008. [15] J.F. Ziegler, M.D. Ziegler, J.P. Biersack, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 268 (2010) 1818. [16] Q. Yang, D.J. O’Connor, Z. Wang, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 61 (1991) 149. [17] G.-b. Wang, H.-g. Liu, K. Wang, X. Yang, Q.-j. Feng, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 287 (2012) 19. [18] W.H. Press, S.A. Teukolsky, W.T. Vetterling, B.P. Flannery, Numerical Recipes, Cambridge, third ed., Cambridge University Press, 2007. [19] J.K. Shultis, Detectors and Associated Equipment 526 (2004) 359. [20] X. Yang, R. Li, H. Liu, G. Wang, K. Wang, Chinese Journal of Computational Physics 29 (2013) 891.