Neutrons, a universal toy for physicis

Prof. Dr. Winfried Petry

Technische Universität München

D-85747 Garching

In spring 1956 Hans Meier-Leibnitz, Professor for Technische Physik at the Technische Hochschule München was asked by the "Atom Minister" of the young Federal Republic of Germany, Dr. Franz Joseph Strauß whether he was interested to build the first nuclear facility in Germany, a research reactor. Only fourteen month later in October 1957 the first neutron source of Germany, the Forschungsreaktor München FRM went critically at Garching. From the beginning this neutron source was operated by the Faculty of Physics. Being the first research facility on the potatoe fields north of Garching and lovingly called atom-egg by local people it turned out to be the seed of what is nowadays Germany's largest research campus.

Since then physics with neutrons played a crucial role for research and teaching within the Physics Department. Meier-Leibnitz and his numerous students revolutionized the way how to use neutrons: Guiding neutrons by total reflection at mirrors like visible light has been invented, the first back scattering spectrometer with the highest energy resolution at that time was build, the principle of scattering at small angles to make objects visible as large as 1000 times the wavelength of neutrons was tested for the first time, the flux of ultra cold neutrons has been increased by orders of magnitude by Doppler shifting their energy ... to name a few of the pioneering inventions. All this knowledge accumulated in the construction of the world strongest neutron source, the high flux neutron reactor at the Institute Laue Langevin (Grenoble) by Maier Leibnitz in the years 1967 - 1972.

Today the Technische Universität is about to take into operation a new high flux neutron source, the FRM II. With only 20 MWatt thermal power it is expected to provide German scientists and industry with a powerful, scientifically attractive neutron source for the beginning of the new millennium.

Neutrons what for?

many of you may ask. Let us repeat shortly: neutrons and protons are the constituents of the nuclei. The positive charge of the protons is counterbalanced by the negative charge of the electrons, whereas the neutrons are neutral in charge. Protons and neutrons have about the same mass, each of them 1838 times heavier than an electron. Roughly spoken half of the surrounding matter consists of neutrons, however stably bound in the atoms. As free particles neutrons are much more seldom - we come back to this later on - and miniscule with a diameter roughly 1/100.000 of that of an atom. Therefore and due to its neutrality neutrons penetrate easily massive materials. For instance a beam of thermal neutrons is weakened by only 60 % when penetrating a block of 10 cm thick Aluminium. It remains to explain what thermal neutrons are: The fission reaction of ²³⁵Uranium deliberates in average 2.4 neutrons, however some 10.000.000 °C hot. Only cooling by collision with the protons of surrounding water makes them useful for us. In average the speed of these 21 °C warm neutrons has been reduced to 1800 m/s.

: Fig. 1: Neutron tomography of a turbine blade of a jet engine. Clearly visible are the cooling channels [1].

A simple but extremely useful application of neutron beams for engineers uses the transparency of metallic objects for thermal neutrons. Fig. 1 shows the three dimensional cut through a turbine blade of a jet engine. Here neutron radiographies of about 200 different positions of the turbine blade have been put together in a powerful computer to a three dimensional pixel volume of different attenuation coefficients of the object. Different colours for different absorption visualize then the three dimensional inner of all kind of multi component work peaces. A spatial resolution of 100 µm is achievable - without ever having put into parts these objects. This technique of neutron tomography is known from computer tomography with x-rays. Different to x-rays neutrons penetrate easily several cm thick metallic work peaces with considerable different absorption coefficients for different elements and even different isotopes of the same chemical element. So it is a particular gift of nature, that the scattering cross section of hydrogen is more than 10 times stronger than that of its isotope Deuterium and that of any other element. All this leads to contrast-rich visualisations. For the purpose of material science it is of great importance that this material testing is non-destructive. With the new FRM II in operation it is planed to make visible the oil lubrification of a crankshaft of a running motor or observing in situ the combustion process of an engine - possibly in time steps as small as 1 millisecond.

Modern materials science tries to explain the function of materials by its microscopic origin. The wavelength of visible light is much to long to make these origins visible. The light's wavelength is roughly 5000 times that of the distance of atoms in solid state. Thermal neutrons however have a wavelength of the order of that interatomic distance. Like visible light by a prisma neutrons of different "colour" are diffracted into different directions by any kind of material. Mathematically this is formulated in the so-called Bragg equation: n λ= 2 d sinΘ/2.

$Fig. 2: Mounting of a part of a crankshaft in a materials diffractometer. Neutrons are coming from the back and scattered neutrons are measured to the left [2, 3].$: Fig. 2: Mounting of a part of a crankshaft in a materials diffractometer. Neutrons are coming from the back and scattered neutrons are measured to the left [2, 3].

The diffraction angle Θ of neutrons of wavelength λ depends on the atomic distance d. The integer n describes the possibility that a wavelength smaller by an integer will be diffracted under the same angle. Measuring the angular position of these diffracted beams then allows to determine the interatomic distances. Fatigue or application of strong forces on any material cause changes of the interatomic distances and thereby overburdening of the interatomic binding. This is the beginning of crack on an atomistic scale. Tiny changes of the Bragg angle ΔΘ due to variations of the atomic spacing Δd are then a measure of the internal stress σ according to the relation σ= E Δd/d (E = Youngs or elasticity modulus). Fig. 2 shows the installation of a part of a crankshaft on a neutron diffractometer in order to measure the internal stresses at different depth. Apertures along the incoming and scattered beam define a small measuring volume within the object. As a result the forging and cold rolling process during manufacturing of the crankshaft could be optimised and the bending stiffness of its main axis could be increased by a factor of two. At FRM II we expect to measure volumes much less than 0.1 mm³. Again this method of visualisation of internal stresses is completely non-destructive.

: Fig. 3: Neutron counts as a function of scattering angle. Inserting the angular positions of the peak intensities into the Bragg equation the positions of the supraconducting YBa2Cu3O7 crystal could be determined [4]. Unusual is that Oxygen atoms contribute to the supraconducting properties.

The location of the atoms in the steel of a crankshaft is known since long, we where only interested by relative changes due to internal stresses. For new materials with eventually completely new functional properties the answer to the question "Where are the atoms?" is one of the most urgent ones. Fig. 3 shows the Bragg peaks of a diffractogramm of YBa₂Cu₃O_7-x, a ceramic which becomes supraconducting at the temperature of liquid nitrogen. Due to the appropriate wavelength of neutrons and due to the excellent scattering contrast of neutrons with oxygen the determination of the oxygen atoms is of particular precision, some 0.01 Angstroem (1 Angstroem = 0.1 nm). This precision is necessary in order to test competing theories to explain supraconductivity. Here one realizes how microscopy in terms of a scattering experiment is meant: the angular position of a Bragg peak is according to the Bragg equation extremely sensitive to distances on Angstroem scale, provided the wavelength is of the order of the atomic distances. Physicists call this a picture in reciprocal space: the larger the Bragg angle, the smaller the distance one measures.

: Fig. 4: Laue picture of a predeuterated Lysozyme crystal. Partly water molecules and Hydrogen atoms are substituted by heavy water D2O and Deuterium D to make their positions visible in a neutron scattering experiment [5, 6].

Fig. 4 gives an excellent example of this "microcopy in reciprocal space": the almost thousand intense spots on the neutron-sensitive film are the Bragg positions of a thermal neutron beam scattered elastically by a tiny single crystal of the protein Lysozyme. This protein has the capability to break the cell walls of bacteria. As a matter of fact almost all protein structures are measured by scattering of x- or synchrotron radiation. However due to the dominating scattering of the Carbon Atoms and the relatively weak scattering of x-rays by Hydrogen theses structures mainly reflect the Carbon backbone of the polypeptide. Positions of Hydrogen are only vaguely known. In favourable cases some of the hydrogen atoms can be exchanged by the chemically equivalent Deuterium. This allows to "colour" the hydrogen atoms; neutron diffraction then selectively sees the hydration shell around the polypeptide, which is important for the function of the protein. In the case of Lysozyme 157 positions of bound D₂O molecules have been determined by that method.

To determine where the atoms are is the first step. The answer to the question "How do atoms move?" is of similar importance for the microscopic understanding of the functionality of any material.

: Fig. 5: Density fluctuations in dry and hydrated Myoglobin. Only in the hydrated sample the density fluctuations decay on a time scale of nanoseconds. Similar relaxations are observed in viscous liquids [7].

The structure of Myoglobin, the protein which storages the oxygen in muscles, certainly tells us how the oxygen is stored in its heme pocket, but it does not tell us how oxygen can penetrate through the densely packed protein structure to the heme pocket. Inelastic scattering of neutrons, that means measuring the energy change of the neutrons during scattering provides the answer. Hits a neutron the molecule or atom, it absorbs parts of the kinetic energy of that particle - or alternatively gives part of its own energy to the molecule or atom. Quantum mechanics tells us, that this energy transfer can only happen in units of the internal vibration frequencies of the bound atom. By nature thermal neutrons have about the same speed or kinetic energy like the vibrations in the solid state. The succient remark "by nature" is easily understandable: the mass of a neutron is of the same order of magnitude like that of atoms. Consequently the neutron's speed alters drastically when scattered inelasticly and is therefore an excellent probe to measure amplitude and direction of inner vibrations. Fig. 5 shows the interpretation of such a measurement. Fluctuations in the range of nanoseconds cause a liquid like relaxation of the polypeptides substructure. Only because subunits of the highly ordered protein structure fluctuate between different structural variants the oxygen succeeds to penetrate to the heme pocket. The computer simulation shown in the figure extends to 1 nanoseconds, now a days computer can calculate to at least a factor of 10 slower times. This progress happened also on the experimental side, with the FRM II into operation we hope to access to times as slow as 0.1 mikroseconds.

: Fig. 6: With an energy transfer of 43 meV ( corresponds to 510 K) inelastic scattered neutron intensity from a crystal of Bi2SrCaCu2O8+x at the border of the reciprocal unit cell [8].

A similar importance for the relation between structure and internal movement is discernable for the high temperature supraconductors. Whereas Fig. 3 shows one of the most precise measurements of the oxygen's position in such a material we did not yet argue about the mechanism of the resistance free transport of electrons through matter. Most surprisingly it has been found, that vibrations of the magnetic moments of the atoms play an important role in that context. Once again neutrons are ideally suited to detect this motion. In spite of its neutrality every neutron has a small magnetic moment, so to say a small "compass needle" pointing in a discrete direction. This "compass needle" interacts with "that" of the atoms during scattering and therefore tells us about the magnitude and direction of magnetism in matter. Fig. 6 shows the finding of inelastic neutron scattering for the supraconductor Bi₂Sr₂CaCu_xO_8+x which belongs to the same material class as YBa₂Cu₃O₇. Due to the neutron's ability to identify unambiguously the magnetic origin of the scattered intensity the message is clear: magnetic fluctuations are related to the supraconducting state and theory pretends that these fluctuations are important to couple electrons to pairs passing through matter without resistance. A final answer certainly asks for further experiments - also with intense neutron beams.

: Fig. 7: Resistivity distribution of conventionally and by neutron activation doped Silicon [9].

A last example demonstrates the importance of neutrons for industry. Pure Silicon becomes its desired electronic properties by doping with small amounts of impurities which tune its conductivity. Doping by transmutation of Silicon to Phosphor - ³⁰Si transforms by neutron capture to ³¹P - results in the most homogenous distribution of the dopand - see Fig. 7. Huge single crystalline rods of Silicon are put into a homogenous field of neutrons for typical times of hours to days. After a decay time of a few days all radioactivity has disappeared. Typical applications are transistors or tyristors for high current applications. At FRM II it will be possible to irradiate huge crystalline rods of 50 cm in heights and 20 cm in diameter. Due to the particular homogenous distribution of thermal neutrons and the perfect suppression of fast neutrons doping should be possible in a quality better than ever before.

All these examples are a small fraction of the possible applications of neutrons in material science. At FRM II neutrons will be used in a much wider sense. By neutron capture the world most intense source of thermal positrons - the anti-matter to electrons - will be available. The fission reaction of ²³⁵Uranium will be used to produce an intense beam of neutron rich radioisotopes of medium mass. Cooling down thermal neutrons to mK temperatures creates an intense beam of ultra slow neutrons. A beam of unmoderated fast neutrons allows therapy of various kind of tumors on a clinical level.

All this assumes the existence of an intense source of free neutrons. Let us report a little bit about the huge technical effort to get this.

Neutrons, from where?

Free neutrons can be produced in high intensity eather by spallation or fission. In case of spallation an intense beam of high energy protons hits a heavy target, thereby evaporating very fast neutrons. For fission ²³⁵Uranium has to capture a thermal neutron. The exited nuclei ²³⁶Uranium decays then in elements of medium mass and on the average 2.4 neutrons. Spallation sources are best suited for pulsed neutron beams whereas nuclear reactors are best for intense continuous flux of neutrons. The most intense spallation source ISIS at Abington has a factor of 10 less average flux than the reactor based source at the Institut Laue Langevin, Grenoble. FRM II is a consequent further development of that latter source. For intense neutron beams a point like concentration of the fission material is needed. This is best done by using ²³⁵Uranium enriched to 93% and packing it as dense as possible. The single fuel element at FRM II contains about 8 kg of highly enriched Uranium and is 70 cm in heights and 24 cm in diameter. At a maximum thermal power of 20 MWatt one fuel element can be used for 52 days. 20 MWatt is of moderate power, it corresponds to the thermal power needed to provide the current for one ICE, the German high speed train. Modern nuclear power plants have a thermal power of up to 4500MWatt. That free neutrons for science are scare may be guessed from what has been told at the beginning. FRM II delivers in its D₂O moderator 8x10¹⁴ thermal neutrons per cm² and sec. This quantity put into an empty container simply means perfect vacuum. The ability to cool the fuel element at any instance imposes the technical limit. The more important it becomes to extract the neutrons in an optimised way to the instrument and neutron detector. It is in this optimised use of the neutrons where FRM II represents a major progress.

Firstly an optimised use demands tailored wavelength of the neutrons for their particular purpose. The moderate thermal power of 20 MWatt allows to place a containment of about 30 ltr. liquid D₂ at a temperature of 20 K in the maximum of the thermal flux, thereby producing a high flux of cold neutrons comparable to that of the strongest neutron source, the high flux reactor at the ILL. A small volume of solid D₂ further cools these cold neutrons to temperatures in the mK range or wavelength around 1000 Angstroem, providing the strongest source of ultra cold neutrons. For other purposes short wavelength are needed. This is achieved by moderating neutrons in a 2400 °C hot block of graphite.

: Fig. 8: Cut through the reactor core, D2O moderator vessel and the concrete shielding. Cold and hot source are located directly beside the unique fuel element. The converter for fast neutrons is located at the outer diameter of the moderator. The source for fission products is located in the though-going beam hole. Not shown are the inclined beam holes, one of which houses the position source.

By neutron capture and subsequent nuclear reaction sources of different kind of radiation are created of up to now unkown intensity. Fission neutrons of 10 million °C are produced at the outer side of the moderator vessel by a converter plate of ²³⁵Uranium. The are used for neutron tomography and cancer therapy. In an other beam hole ?-radiation generated by Cadmium converts to electron and positron pairs. The anti-matter positron is guided by electromagnetic fields to the experiments providing the most intense beam of thermal positrons for the sake of material science. A further beam hole passing from both sides through the reactor protection serves for the production of fission products which will serve to create super heavy elements and carrier free radioisotopes for radiopharmaca. Fig. 8 gives an overview about the different sources of moderated neutrons.

The concrete shielding around the core is full of holes like a Swiss cheese, altogether neutrons are fed to the experiments through 12 beam holes, the biggest of which feeds 6 neutron guides. All these beam tubes end in the maximum of thermal flux and are tangential oriented towards the fuel element, thereby suppressing to a maximum fast background neutrons. As neutral particles neutrons can not be guided to first order electro-magnetically. The intensity at the experiment depends therefore from the solid angle seen by the point like source.

: Fig. 9: Floor plan of the FRM II with its large neutron-guide hall, eventually extending into the hall of the now-a-days stopped atom-egg.

Thanks to their optical properties neutrons can be guided by simple reflection on perfect surfaces. Due to the relatively small difference in the refraction index between vacuum and matter this is only possible for small incident angles typically below 1o. This principle has been discovered by Maier-Leibnitz and his students at the atom-egg in the sixties. Today neutron guides of much better performance are available. Instead of a single covered glass surface today several hundreds of thin layers with alternating optical properties are evaporated on perfect float glass surfaces, thus adding to the angle of total reflection a broad band of Bragg refraction. At the FRM II neutron guides with a divergence of up to 4 times the glanzing angle are used. Additionally theses super-mirrors enable focussing optics, thus increasing even more the flux at the often tiny samples. An impression of the different instruments available at the FRM II is given in Fig. 9.

With routine operation foreseen for 2003 scientist, engineers, physicians and industry will have available optimised instruments, some of them unique, others the best of their kind, all of them with innovative techniques: several irradiation positions with a flux of thermal neutrons between 5x10¹² to 4x10¹⁴ n/(cm²s) allow activation times between 300 milliseconds and several weeks for smallest volumes up to Silicon rods of 50 cm in length and 20 cm in diameter; irradiation position for clinical tumor therapy; radiography and neutron tomography with thermal and fast neutrons; several instruments for determining structures of novel materials; materials diffractometer for internal stresses and textures; neutron reflectometry for characterising surfaces and multi-layers of soft and magnetic matter; several inelastic spectrometers to follow the internal dynamics of the atoms and molecules; positrons to explore electronic properties of surfaces.

Literature

[1] Schillinger, B., Gebhard, R., Haas, B., Ludwig, W., Rausch, C., Wagner, U.: 3-D Neutron Tomography in Material Testing and Archeology. Proc. Vth World Conf. on Neutron Radiography. Berlin 1996. 688-693. DGZfP (Publ.) Berlin 1997.

[2] Mayer, H.M., Achmus, C., Pyzalla, A., Reimers, W.: Neutron and X-Ray Diffraction Analysis of the Influence of Induction Hardening and Deep-Rolling on the Residual Stresses in Crankshafts. Materials Science Forum, 347-349, 340-345 (2000).

[3] Achmus, Ch., Reimers, W., Wohlfahrt, H.: Materialprüfung 40, 88 (1998).

[4] Cava, R.J., Hewat, A.W., Hewat, E.A., Batlogg, B., Marezio, M., Rabe, K.M., Krajewski, J.J., Peck, W.F., Rupp, L.W.: Physics C 165, 419 (1990).

[5] Oleinik, Ph.: Diploma Thesis, Physics Department, 1995.

[6] Niimura, N., Minezaki, Y., Nonaka, T., Castagna, J.C., Cipriani, F., Hoghoj, P., Lehmann, M.S., Wilkinson, C.: Nature structural biology 4, 909 (1997)

[7] Settles, M.: Die Zeitabhängigkeit und die Geometrie der intramolekularen Dynamik globulärer Proteine bis 100 ps aus Neutronenstreudaten. Doktorarbeit. Technische Universität München (1996).

[8] Fong, H.F., Bourges, P., Sidis, Y., Regnault, L.P., Ivanov, A., Gu, G.P. Koshizuka, N., Keimer, B.: Nature 398, 588 (1999).

[9] Neutrons for Industry and Medicine. Schriftenreihe der Technischen Universität München (2000), erhältlich über Pressereferat FRM II.

: Fig. 10: Entrance of the FRM II some 100 m beside the building of the Physics Department.