Nuclear reaction analysis and quantitative hydrogen analysis

---

Method

Nuclear reactions can occur if the speed of an incoming ion is high enough to penetrate the nucleus of a sample atom. Detecting the reaction particles or emitted γ-rays can be used for analysis and this method is called Nuclear reaction analyis (NRA).

In the reaction, an excess of energy can be produced if the total mass of the reaction products (M₁ + M₂) is less then the total mass of the original nuclei (M_p + M_i). The energy difference is called the Q value of a reaction and a positive Q-value means excess energy.

Table with some useful reactions, the reaction Q value and the energy of the outgoing particle (p or α) for a typical ion beam energy of 1.2 MeV.

Element	Reaction	Q (MeV)	Eout (MeV)
6Li	6Li(p,α)3He	4.02	1.36
	6Li(d,α)4He	22.37	9.21
7Li	7Li(p,α)4He	17.35	7.66
12C	12C(d,p)13C	2.719	2.95
16O	16O(d,p0)17O	1.919	2.36
	16O(d,p1)17O	1.048	1.58
	16O(d,α0)14N	3.116	2.61
	16O(d,α1)14N	0.804	0.97

---

Hydrogen analyis

A nuclear reaction that is very useful for the detection of hydrogen is

¹⁵N + ¹H → ¹²C + ⁴He + γ (4.43 MeV)

This reaktion has a sharp resonance at the ¹⁵N energy of 6.385 MeV and the method is therefore called Resonant Nuclear Reaction Analysis (RNRA). Because of the narrow resonance in the cross-sections, the depth in the material where the reaction takes place can be increased by increasing the incidence energy. By steadily increasing the energy, a depth profile of the hydrogen concentration with a depth resolution of less then 10 nm near the surface can be obtained.

Advantages

• Quantitative determination of hydrogen concentrations without the need for (matrix-matched) standards
• High depth resolution of several nm (glancing angle of incidence) at the surface
• Large analysis depth, up to 5 µm depending on the density
• High sensitivity

Limitations

• Hydrogen loss during measurent can occur but can also be monitored and corrected for
• Measurements in vacuum

---

Hydrogen analysis setup at HZDR

This RNRA setup is connected to the 6MV Tandetron accelerator of the Ion Beam Center and a ¹⁵N ion beam is used in the energy range of 6.385 - 12 MeV. The gamma-rays are detected with a large and efficient detector made of Bismuth germanium oxide (BGO).

Properties of the setup

• Incident beam: ¹⁵N²⁺
• Energy: 6.3 - 12 MeV
• Beam current: 10 - 50 nA
• Beam spot: 1 - 100 mm²
• Detection limit: 0.05 at%
• Analyses depth: up to 5 µm (depending on material)
• Depth resolution: about 8 nm (Si), minimum 1 nm (grazing incidence angle)
• Measured quantity: yield of the 4.43 MeV γ-rays
• Detector: 4" x 4 " BGO

Sample requirements

• Sample size: minimum 5x5 mm², maximum 25 mm, optimum 10-15 mm
• Sample thickness: max. 7-8 mm, optimum ≤ 1 mm (total thickness for mounting)
• Low roughness (in nm range)

---

Applications

• Crystalline silicon surface passivation by thick PECVD Al₂O₃ for photovoltaics

Aim: The electrical passivation of crystalline silicon surfaces can be achieved by a layer of aluminium oxide (Al₂O₃) on top. For the photovoltaic industry, such a layer has to be deposited with a high-throughput and low cost technique. In this study plasma-enhanced chemical-vapor-deposition (PECVD) was used to deposit the layer. Because hydrogen plays an important role in the passivation, resonant-NRA has been used to determine the H concentration as a function of depth.

Procedure: An Al₂O₃ of 20 to 50 nm was deposited by PECVD and annealed at 450°C. The stability of the layer was tested by a short high-temperature process (firing). The addition of two types of capping layer (hydrogenated SiN_x and non-hydrogenated SiO_x) on the passivation of the silicon and the stabiltiy of the layers was also studied. H depth profiles have been measured with RNRA.

Note the high depth resolution of RNRA (many points over a range of 120 nm).

Conclusion: The hydrogen concentration at the Al₂O₃ / c-Si interface is similar for the samples with and without capping layers and slightly lower after annealing, leading to the conclusion that the Al₂O₃ layer in itself provides enough H to passivate the interface.

P. Saint-Cast, D. Kania, R. Heller, S. Kuehnhold, M. Hofmann, J. Rentsch, R. Preu
High-temperature stability of c-Si surface passivation by thick PECVD Al2O3 with and without hydrogenated capping layers
Applied Surface Science 258 (2012) 8371–8376

---

• Hydrogen accumulation in tungsten for fusion research

Aim: Tungsten is a prime candidate for plasma facing materials in fusion reactors. However, coarse grained tungsten (CGW) has a high capacity for the retention of light elements that can even lead to blistering and exfoliation. In this study, the role of grain bounderies on the H accumulation has been studied for nanostructured tungsten (NW) samples in comparison to CGW.

Procedure: CGW and NW samples have been implanted with H, with C and then H sequentially, with C and H simultaneously and with C and H sequentially at a temperature of 673 K. The implantation energies were 170 keV and 665 keV for H and C, respectively. RNRA has been used to determine H concentration depth profiles.

The RNRA results evidence that:

1. H retention in NW samples is larger than in CGW,
2. Synergistic effects of the implantation have a large influence on the H retention in CGW samples but do not in NW samples
3. None of the studied samples present H retention when the hydrogen implantation is carried out at 673 K.

R. Gonzalez-Arrabal, M. Panizo-Laiz, N. Gordillo, E. Tejado, F. Munnik, A. Rivera, J.M. Perlado
Hydrogen accumulation in nanostructured as compared to the coarse-grained tungsten
Journal of Nuclear Materials 453 (2014) 287–295

---