Modern state of physics and technology for obtaining beams of polarized particles. A source of atomic hydrogen and deuterium with nuclear polarization for experiments on internal beams of accelerators Polarimeter data acquisition systems

-- [ Page 1 ] --

THE RUSSIAN ACADEMY OF SCIENCES

PETERSBURG INSTITUTE OF NUCLEAR PHYSICS

them. B.P. KONSTANTINOV

As a manuscript

Mikirtychyants Maxim Sergeevich

UDC 539.128, 539.188

Development and research of a source of atomic hydrogen and deuterium with nuclear polarization for experiments on internal beams of accelerators 01.04.01 – devices and methods of experimental physics

Scientific supervisors:

Candidate of Physical and Mathematical Sciences V.P. Koptev Candidate of Physical and Mathematical Sciences A.A. Vasiliev Gatchina Contents Introduction............................................... ................................................. ..........................- nine Chapter 1.

Methods for obtaining atomic beams.................................................................... ...................................... 13 1.1 Introduction .......................... ................................................. ..................................- 13 1.2 The mechanism of dissociation in a gas discharge .......... ...............................................- 14 1.3 Theoretical consideration of the formation gas jet ..............................- 17 1.3.1 Molecular regime (outflow) .............. ...............................................- 17 1.3.2 Formation beam with a long channel .............................................. - 18 1.3.3 Hydrodynamic flow regime. Supersonic jet...................................- 20 1.3.4 Estimation of source intensity ............................. ...................................- 24 Chapter 2

Methods for creating polarization in atomic beams.................................................................. .- 27 2.1 Introduction............................................... ................................................. ............- 27 2.2 Sources using the Lamb shift (LSS) .................................. ..........- 31 2.3 Optically pumped sources (OPPIS) .................................................. ........................................... 33 2.4 Sources of polarized atomic beams (PABS) .................................. ..............- 35 Chapter 3

Source of polarized atomic hydrogen and deuterium for the internal gas target of the ANKE spectrometer........................................................... ..................................- 38 3.1 Brief description of the structure .............................. ................................................. .........- 38 3.2 Vacuum system .................................. ................................................. .....- 42 3.2.1 Construction of the vacuum chamber ..................

42 3.2.2 Differential pumping system .............................................................. .........- 44 3.3 Dissociator ..................................... ................................................. ..............- 47 3.3.1 Mechanical design .............................. ...............................................- 48 3.3.2 RF system.... ................................................. ...................- 51 3.3.3 Nozzle cooling system .............................. ...............................................- 52 3.4 Forming system gas jet ............................................... ............- 54 3.4.1 Construction .................................. ................................................. .........- 54 3.5 Spin-separating magnet system .................................. ..............................- 56 3.5.1 Basic principles ............................... ................................................. .........- 56 3.5.2 Spin-separating sextupole magnets ANKE ABS...................- 57 .................................................- 59 3.6.1 Principles of operation............................................... ...................................- 60 -2 3.6.2 ANKE ABS.. .........................................- 62 Chapter 4

Optimizing Source Characteristics .................................................................. ..................- 66 4.1 Intensity of the atomic beam .............................. .........................................- 66 4.1.1 Instruments and measurement technique ................................................. ..........- 66 4.1.2 Absolute calibration method .................................. ...............................- 69 4.1.3 Device for measuring the intensity of an atomic beam .......... ...- 74 4.1.4 Obtained results .......................................... ..................................- 78 4.1.5 Conclusions............. ................................................. ...............................................- 81 4.2 Spatial distribution of the beam density.... ...................................- 82 4.2.1 Instruments and measurement technique .............. ................................................. ....- 82 4.2.2 Nozzle adjustment ....................................... ................................................ ........- 86 4.2.3 Findings .............................................. ...............................- 88 4.2.4 Conclusions....... ................................................. ...............................................- 89 4.3 The degree of dissociation of the atomic beam ................................................. .......- 90 4.3.1 Instruments and measurement technique .................................. .........................- 90 4.3.2 Degree of dissociation of a free atomic jet ............... .................................................. 92 4.3.3 Spatial distribution of the degree of dissociation in a polarized beam .............................. ................................................. .................- 95 4.3.4 Conclusions .............................. ................................................. .......................- 97 4.4 Polarization ............................... ................................................. ...............................- 98 4.4.1 Instruments and measurement technique .............................. ................... .........................- 98 4.4.2 Findings .............................. ................................................. .... - 100 4.4.3 Conclusions............................................... ................................................. ........ - 102 Chapter 5

Perspectives of use .................................................................. ................................... - 104 5.1 Jet targets .......... ................................................. ............................... - 104 5.2 Polarized gas targets. Accumulation box .......................... - 106 Conclusion .................. ................................................. ............................................... - 110 Literature... ................................................. ................................................. .......... - 115 -3 List of illustrations i Fig. 1. Cross sections s in of inelastic processes 16 as a function of the electron energy . - 15 Fig. 2. Scheme of splitting the nozzle into elementary tubes .......................................................... ...- 24 Fig. 3: Diagram energy levels of a hydrogen atom in a magnetic field B. For the ground state, Bc = 507 G; for the 2S1/2 state, Bc = 63.4 G. The energy W is measured in units of DW = h1420.4 MHz (= 5.9 10-6 eV) ............................... ..............- 28 Fig. 4: Energy level diagram of a deuterium atom in a magnetic field B. For the ground state Bc = 117 G, for the 2S1/2 state Bc = 14.6 G. The energy W is measured in units of DW = h327.4 MHz (= 1.4 10-6 eV) ............................... ...............................- 28 Fig. 5. Nuclear polarization of the levels of hyperfine splitting of the hydrogen atom as a function of the external magnetic field................................................. ......................- 30 Fig. 6. Nuclear polarization of the levels of hyperfine splitting of the deuterium atom as a function of the external magnetic field .................................................................. .........................................- 30 Fig. Fig. 7. Diagram of energy levels of hyperfine splitting for 2S1 / 2 and 2P1 / 2 states of the hydrogen atom .................................................. ................................................. ...- 31 Fig. Fig. 8. The main elements of a polarized source at the Lamb shift......- 32 Fig. 9. Operating principle of the source with optical pumping ..............................................- 34 Fig. Fig. 10. Energy levels of the hyperfine splitting of the hydrogen atom in the 2S1 / 2 state as a function of the external magnetic field.................................................... ...............................- 34 Fig. 11: Block diagram of a source of polarized atomic hydrogen/deuterium.

1 - gas flow regulator;

4 - the first group of spin-separating magnets;

6 - the second group of spin-separating magnets;

8 - storage cell (target) .............................................. ................- 35 Fig. 12. ANKE ABS and a special vacuum chamber for mounting various types of targets on the COZY storage ring. The source of polarized atomic hydrogen and deuterium is located between the deflecting magnet D1 and the central magnet of the spectrometer D2. Direction of the COZY beam from left to right .......................... 38 Fig. 13. ANKE ABS drawing. Explanations are given in the text............................................... .- 40 Fig. 14. Photo of ANKE ABS in the laboratory. Height of the upper vacuum chamber – 80 cm ....................................... ................................................. .................................- 41 Fig. 15. Upper movable baffle .............................................. .........................- 43 Fig. 16. Scheme of the vacuum system of the ANKE ABS source. A complete list of vacuum equipment is given in Table 1.................................................... ...................................- 44 Fig. 17. Various schemes for pumping chamber I .................................................. ...................- 45 Fig. 18. ANKE ABS RF Dissociator.................................................................. ..........- 47 Fig. 19. Sectional view of the ANKE ABS dissociator. 1: gas supply flange, 2: coolant inlet, 3: HF entry, 4: HF sliding connection, 5: inductor, 6: capacitors, 7: lower cooling circuit seal, 8:

nozzle, 9: part of the nozzle cooling system (copper thermal bridge).................................- 49 -4 fig. 20. The lower end of the dissociator and the gas jet formation system. one:

discharge tube and cooling tubes, 2: lower cooling circuit seal, 3: Teflon heat flow restrictor, 4: sliding joint, 5:

nozzle and cooling system support, 6: heater, 7: copper thermal bridge, 8:

nozzle mount, 9: nozzle, 10: window in upper vacuum baffle, 11: skimmer, 12: collimator12, 13: first sextupole magnet, 14: lower vacuum baffle ................................................. ................................................. - 50 Fig. 21. Structural diagram of the radio frequency system .................................................... ...- 51 Fig. Fig. 22. Characteristic dependence of the nozzle temperature on time during stabilization using a PID controller .................................................................. ................................................. .......- 53 Fig. 23. Losses in the gas jet formation system .......................................................... ...- 55 Fig. 24. Sextupole magnet used in ABS. An atom flying into a magnet with r = 0 at an angle a0 is shown on the left;

on the right, several lines of force are shown.....- 56 Pic. 25. Dependence of the effective magnetic moment of an atom on the external magnetic field for four levels of hyperfine splitting .................................................................. ..............- 57 Fig. 26. Part of a cylindrical permanent sextupole magnet, consisting of segments .............................................................. ................................................. ...............................- 58 Fig. 27. Scheme of the block of high-frequency transitions .............................................. ..........- 60 Fig. Fig. 28. Structure of the ANKE ABS ultra-thin transition block...................................- 62 29. Scheme of winding a coil of a gradient field (Bgrad) .............................................. - 63 Fig. Fig. 30. Simplified wiring diagram for switching on the WFT and MFT blocks ...............- 64 Fig. 30. 31. Photograph of the MFT ultrafine transition unit (center) installed in the ANKE ABS polarized source. One of the three spin-separating magnets of the first group is seen from above. ..............- 65 Fig. Fig. 32: Device for absolute measurements of beam intensity - compression tube .............................................................. ................................................. ........- 67 Fig. 33: Electron impact ionization cross sections for atomic () and molecular () hydrogen .................................................... ................................................. .......................- 71 Fig. 34: Experimental data of PSV and PCV pressures versus time...- 74 Pic. 35. Assembly drawing of a device based on a compression tube. one:

guide support .................................................................. ...............................- 76 Fig. 36. Scheme of the non-polarized gas supply system .............................................. .- 77 Fig. Fig. 37. Photograph of the ABS lower vacuum chamber with devices for measuring the absolute beam intensity (below) and the degree of dissociation (left).................................................- 78 Fig. 37. Fig. 38: Dependence of the intensity of the atomic beam on the input flow of molecular hydrogen at nozzle temperature Tnozzle = 62 K, dissociator RF power Wdisso = 350 W and additional oxygen flow q(O2) = 1 10-3 mbar l/s..... ................................................. ...............................- 79 Fig. Fig. 39: Dependence of the intensity of the atomic beam on the RF power supplied to the dissociator at the nozzle temperature Tnozzle = 62 K, the input flow of molecular hydrogen q(H2) = 1.2 mbar l/s and the additional flow of oxygen q(O2) = 1 10-3 mbar l/s ............................................... ................................................. ...- 80 -5 Fig. Fig. 40: Dependence of the atomic beam intensity on the nozzle temperature for different nozzle diameters (D = 2.0, 2.3, 2.5 mm). The radio frequency power supplied to the dissociator is Wdisso = 350 W, the input molecular hydrogen flow q(H2) = 1.2 mbar l/s, and the additional oxygen flow q(O2) = 1 10-3 mbar l/s. For comparison, the results of measuring the intensity of the sources HERMES (), PINTEX () and the source of polarized ions of the University of Munich ()...- 81 are shown. Fig. 41. Scheme of the installation for measuring the profile of an atomic beam ..................................- 83 42. Block diagram of a quadrupole mass spectrometer. Solid lines are stable, dash-dotted lines are unstable ion trajectories ..........................- 84 Fig. 43. Simplified scheme of the mass filter .............................................. .........................- 84 Fig. 44. System of control and data collection used in the measurements of the degree of dissociation .................................................................. ................................................. ...............................- 86 Fig. 45. Density distribution of atomic hydrogen in a beam. The shaded area corresponds to the geometrical dimensions of the vertical tube of the storage cell.................................................................. ................................................. .................................- 86 Fig. 46. ​​Atomic hydrogen beam profiles in the X and Y planes corresponding to the distribution maximum in Figs. Fig. 45. The shaded area corresponds to the geometrical dimensions of the vertical tube of the storage cell...............- 87 Fig. 45. Fig. 47. Dependence of the signal of the quadrupole mass spectrometer on the position of the adjusting screw N1 .................................................................. ................................................. ...- 88 Fig. 48. Density distribution of atomic hydrogen in the beam after nozzle adjustment. The shaded area corresponds to the geometrical dimensions of the vertical tube of the storage cell.................................................................. ...............................................- 88 Fig. 49. Atomic hydrogen beam profiles in the X and Y planes corresponding to the distribution maximum in fig. Fig. 48. The shaded area corresponds to the geometric dimensions of the vertical tube of the accumulative cell...............- 89 Fig. 48. Fig. 50. Dependence of the degree of dissociation (a) on the inlet gas flow for various nozzle temperatures and radio frequency power W = 300 W.................................................... 93 Fig. 51. Dependence of the degree of dissociation (a) on the RF power at low input flows and nozzle temperature T = 70 K................................................. .................- 93 Fig. Fig. 52. Dependence of the degree of dissociation (a) on the RF power at high input fluxes and nozzle temperature T = 70 K....................................................... ................- 94 Fig. Fig. 53. Dependence of the degree of dissociation (a) on the temperature of the nozzle at various input flows and RF power W = 300 W............................................ .............- 94 Fig. 54. Degree of dissociation as a function of time for typical operating conditions of ANKE ABS .................................................................. ................................................. ..............- 95 Fig. 55: Distribution of the degree of dissociation in the beam in the plane of the compression tube. The shaded area corresponds to the geometrical dimensions of the compression tube.................................................................. ................................................. ...- 96 Fig. 56: Density distribution of molecular hydrogen in the beam in the plane of the compression tube. The shaded area corresponds to the geometrical dimensions of the compression tube.................................................................. ...................................- 96 Fig. 57: Profiles of the degree of beam dissociation in the X and Y planes along the center of the compression tube. The shaded area corresponds to the geometric dimensions of the compression tube....... ................................................. .......................- 97 -6 Fig. Fig. 58. Scheme of setup for measuring beam polarization...............................................- 99 Fig. 59. Dependence of the number of Ly-a photons on the magnetic field in the spin filter...... - 100 60. Dependence of the number Ly-a of photons on the magnetic field in the spin filter in the case of a polarized hydrogen beam. The left peak corresponds to atoms with mI = +1/2, the right one with mI = –1/2 .............................. ................................................. ......................... - 101 Fig. Fig. 61. Dependence of the number Ly-a of photons on the magnetic field in the spin filter in the case of a polarized deuterium beam: (a) and (b) - vector polarization, (c) and (d) - tensor polarization. The left peak corresponds to atoms with mI = +1, the middle peak corresponds to mI = 0, and the right peak corresponds to mI = –1...................... ................................................. ............................... - 101 Fig. Fig. 62. Distribution of the magnetic and HF fields in the blocks of radio frequency transitions MFT (a), WFT (b) and SFT (c)............................................... ................................................. ............... - 102 Fig. Fig. 63. Principal diagram of a jet target (jet target) .................................................. - 105 Fig. 64. Storage cell for a polarized source............................................... - 106 65. The idea of ​​a storage gas cell and the distribution of pressure in it ....... - 108 -7 List of tables Table 1. List of ANKE ABS vacuum equipment .................. .......................- 46 Table 2. Parameters of the original and optimized beamforming systems and the obtained maximum intensities. Dimensions are in mm..............- 55 Table 3. Dimensions of sextupole magnets and magnetic field on the surface.....- 59 Table 4. Main characteristics of radio frequency transition blocks..... ...................... 61 Table 5. High-frequency equipment of hyperfine junction blocks ..............- 64 -8 Introduction Despite the great success of modern nuclear physics in explaining various properties of of nuclear matter, the question of the high-momentum component of the nuclear wave function, or, in other words, of the structure of nuclear matter at distances of the order of or less than the nucleon radius, is still open. At present, the main problem is the experimental detection of this structure and the determination of the interval of the internal moment of the relative motion of nucleons in the nucleus, in which the traditional description of the nucleus as a collection of nucleons is valid.

It is expected that at distances RNN 0.5 fm there is some transition region between the meson-nucleon and quark-gluon degrees of freedom in the nucleus. One of the confirmations of the existence of such a region at high transferred momenta can be a violation of the traditional picture based on the phenomenological potential of the NN interaction corresponding to the NN phase shift. In this sense, the problem of the high-momentum component of the nuclear wave function is closely related to the problem of choosing the nucleon-nucleon interaction potential at close distances.

A special role in the study of these problems is played by polarization experiments, which make it possible to establish the spin dependence of nuclear forces.

Carrying out such experiments requires the use of both a high-intensity beam of polarized protons and a high-density polarized target.

Traditionally, such targets were solid-state polarized targets. However, during the last decade, a new type of polarized targets, gas polarized targets, has been rapidly developed, which make it possible to avoid the problems of radiation damage and the presence of unpolarized impurities (for example, N in NH3) that are typical for solid targets. The most common r r r polarized gas targets are H -, D - and 3 He targets that do not contain impurities. Since the spatial density of such targets is low, they have found wide application in accelerating storage rings. In this case, it is possible to achieve a sufficiently high value of the lifetime of the accelerator beam, and the high luminosity of the experiment is ensured due to the repeated passage of the beam through the target.

9 Currently, several experiments are being carried out using both a polarized accelerator beam and a polarized target consisting of a polarized atomic beam source (PABS1) and a cryogenic storage cell in which the interaction under study occurs.

For the first time, a gaseous polarized deuterium target was used in Novosibirsk on the VEPP-3 electron storage ring.

The HERMES experiment at DESY (Hamburg, Germany) studies the spin structure of the nucleon. For this purpose, inclusive and semi-inclusive reactions of deep inelastic scattering of a longitudinally polarized 27.5 GeV r r HERA positron beam on polarized H, D, and 3 He gas targets are studied.

Hydrogen and deuterium targets are a source of a polarized atomic beam and a storage cell. Such setups make it possible to create an atomic beam with a sufficiently high (close to 100%) nuclear polarization, and the use of an open storage cell does not destroy the accelerator beam.

On the polarized beam of the IUCF storage ring (Bloomington, USA), experiments were carried out to study nucleon-nucleon interactions, also using an internal polarized gas target. Their goal was to improve modern ideas on the potential of the nucleon-nucleon interaction. For this purpose, the spin-correlation coefficients were measured and pion production near the threshold was studied.

A special role in the study of issues related to the study of NN interactions at close distances is played by the deuteron, as the simplest nuclear system. Despite the fact that the deuteron is a rather loosely bound system, it has become the main object of study in both theoretical and experimental nuclear physics.

One of the experiments aimed at studying the pd interaction at a moment of relative motion of nucleons inside the nucleus q = 0.3 0.5 GeV/c is the COSY2-Jlich storage ring experiment dedicated to the breakup of the deuteron. Of particular interest is the polarization experiment rr (pd ® ppn), aimed at determining the dependence of five polarization (A yp, Ay, A yy, C yy, C yyy) d d observed on the internal moment of relative motion of nucleons in the deuteron breakup reaction. This will make it possible to obtain information about the structure of the deuteron wave function with the new Polarized Atomic Beam Source COoler SYnchrotron - 10, since the polarization observables depend on the ratio of the S- and D-components of the wave function. Taking into account the features of the ANKE3 spectrometer, the experiment can be carried out in conditions of collinear geometry: protons emitted backwards close to 180 will be registered in coincidence with protons emitted forward at small angles (close to 0). In such a geometry, the S- and D-wave functions of the deuteron can be studied up to an internal moment of 0.5 GeV/c.

This experiment will require the use of both a polarized accelerator beam and a polarized target.

At present, a beam intensity of 5·1016 particles/s for unpolarized and 5·1015 particles/s for polarized protons has been achieved on the COZY storage ring. However, modernization of the source of polarized ions, beam transport path and injection system should lead to an increase in the intensity of the beam of polarized protons up to 1 1016 particles/s. In addition, it is planned to inject unpolarized, and later polarized deuterium.

In the experiment, it is planned to use an internal gas target, which is a cryogenic storage cell. The polarized gas, hydrogen or deuterium, enters the target from a source of polarized atomic hydrogen and deuterium (ANKE ABS).

Since one of the main factors determining the efficiency of the experiment at the accelerator is the statistics collection time, which is proportional to the target density, determined by the source atomic beam intensity, and has a quadratic dependence on the target polarization. Therefore, these parameters are subject to special requirements:

· high nuclear polarization of the atomic beam (more than 80%);

· fast change of polarization sign (positive/negative) and, in the case of a deuterium beam, polarization type (vector/tensor);

high intensity of the atomic beam (more than 61016 atoms/s).

· In addition to physical parameters, the source must meet the high requirements for experimental setups on modern storage rings (vacuum conditions, limited space, rapid integration into an existing experimental setup, etc.).

Apparatus for studies of Nucleon and Kaon Ejectiles - 11 Achieving high values ​​of source parameters is impossible without studying the characteristics of atomic beams. The latter implies the need to develop methods and create a number of instruments for measuring and optimizing source parameters.

This work is devoted to the creation of a source of polarized atomic hydrogen and deuterium, as well as the development of instruments for studying and optimizing the parameters of an atomic beam, such as the intensity of the atomic beam, the degree of polarization, and the spatial distribution of the beam density.

The work presents various methods, making it possible to create atomic beams with nuclear polarization. A detailed description of both the principles of operation and the design of the structural elements of the source of polarized atomic hydrogen and deuterium is given. The results of studies of the properties of an atomic hydrogen beam are presented. The prospects for using a source of polarized atomic hydrogen and deuterium as a source for gas targets used in experiments on storage rings are considered.

Chapter 12

Methods for producing atomic beams 1.1 Introduction For many years, experiments with molecular and atomic beams have been a source of valuable information about the properties of molecules, atoms, and nuclei. The first experiments with molecular beams were carried out at the beginning of the 20th century by Dunoyer. In the 1920s, Stern and Gerlach, in their experiments on the deflection of atomic beams in inhomogeneous magnetic fields, showed the presence of spatial quantization. A little later, in the 1950s, Lamb and Riserford discovered a shift in the 2S1/2 and 2P1/2 levels relative to each other. This phenomenon is called the Lamb shift. A decade later, a method was proposed for creating polarized atomic beams, which has found wide application in modern nuclear physics. In this, far from complete, list, the main objects of study were beams of neutral atoms and molecules.

Quite often there is a need to obtain beams of atoms, such as H, D, Cl, etc., despite the fact that under normal conditions these atoms form molecules (H2, D2, Cl2, etc.). If the creation of molecular beams is not particularly difficult, then the methods for obtaining beams of similar atoms are in themselves a separate physical problem for the dissociation of molecules into atoms.

Traditionally, the most commonly used methods for dissociating molecules into atoms are:

· Dissociation under the influence of high temperatures, as, for example, in the work where molecular hydrogen was fed into a tungsten furnace, which was heated to a temperature of 2500 K. At a pressure in the furnace of the order of 1 mbar, the degree of dissociation was ~64%.

· Dissociation in a strong electric field, as, for example, in the work , where a Wood tube was used to dissociate hydrogen. The degree of dissociation was ~7080% at a pressure in the tube of about 1 mbar.

13 · Dissociation under the action of a high-frequency field (see, for example, , where at a pressure in the discharge tube of ~0.25 mbar, the degree of dissociation was ~60%).

In modern installations, the latter method is most widely used. To create and maintain a gas discharge, standard high-frequency or microwave industrial generators are used. At characteristic pressures inside the discharge tube at a level of 12 mbar, the degree of dissociation in such devices reaches 90%.

In addition to the dissociation of molecules, the task of creating atomic beams also includes issues of beam formation. The conditions necessary to create a beam of atomic hydrogen can be very different for each specific case under study. The need to keep the recombination rate low requires that the beamforming systems operate at low density (1017 atoms/cm3) and, moreover, at fairly large nozzle openings. Therefore, the parameters of the formation system cannot be chosen a priori, but rather must be a compromise solution, taking into account the limitations imposed by other parameters of the installation.

1.2 Mechanism of dissociation in a gas discharge The degree of dissociation in a gas discharge is determined by the density of the created atomic component and various recombination mechanisms. The mechanism of these processes is determined by the macroscopic parameters of the discharge, such as the gas pressure in the discharge tube, the power of the radio-frequency field scattered by the plasma, the properties of the material of the discharge tube, and so on. Usually, an oscillating circuit is used to obtain and maintain a discharge, powered by a radio frequency generator and dissipating power electromagnetic field on plasma by inductive coupling with a dielectric discharge tube. The degree of ionization, defined as the ratio of the electron or ion density to the density of neutral particles (atoms and molecules), is rather low and lies in the range of 10-510-3.

The mobility of electrons is much greater than the mobility of ions, and this leads, given the low degree of ionization, to the fact that the temperature of the electron gas is much higher than the temperature of neutral particles and ions. The characteristic temperature range is 14 neutral and ionic components 5002000 K, which corresponds to energies in the range of 0.080.35 eV, while the average electron energy lies within 210 eV. That.

the properties of the discharge are determined by the kinetics of electrons: being in a high-frequency electromagnetic field, free electrons acquire energy and dissipate it on neutral particles through elastic and inelastic collisions.

The following inelastic interactions (with cross section s iin) of free electrons with neutral particles are dominant:

1) Excitation of vibrational levels of molecules (s 1in) e- + H 2 ® H 2 + e-.

ex 2) Dissociation of molecules (s 2) in e- + H 2 ® H + H + e-.

3) Ionization of molecules (s 3) in e - + H 2 ® H 2+ + 2e -.

4) Ionization of atoms (s 4) in e - + H ® H + + 2e -.

5) Excitation of 2p states of atoms (s 5) in e - + H ® H (2 p) + e -.

6) Excitation of 2s states of atoms (s 6) in e - + H ® H (2s) + e -.

0. s2in s in 0. s1in s5in siin 10-15 cm s4in 0. 0. 0. s in 0. 10 20 30 40 Electron energy, eV i 1. Cross sections s in of inelastic processes 16 as a function of the electron energy .

15 As can be seen from fig. 1, the dissociation process (threshold electron energy 8.8 eV) is dominant in the electron energy range 1020 eV.

Taking into account the energy dependence of the cross sections and the Maxwellian spectrum of electron energies, it was shown in the work that, at an average electron energy of less than 5 eV, in addition to the dominant process (1), the intensity of the dissociation process (2) is an order of magnitude higher than the intensity of ionization processes (3) and ( 4).

This leads to the conclusion that, under the typical discharge conditions described above, dissociation rates of up to 90% can be expected. For given atomic and molecular densities, the degree of dissociation is defined as na a= (1) na + 2n m or nm a =1+ (2) off nm off where nm is the molecular density in the absence of discharge and then (3) na = 2(nm - n m) off In addition to the main processes, for charged particles formed as a result of ionization processes (3 and 4), in the above reasoning, diffusion losses, two- and three-particle recombination were taken into account. The calculations presented in show that, in the range of a from 0 to 100% and a dissipated power density of 125 W/cm3, the average electron energy lies below 5 eV. This also confirms the possibility of obtaining a high degree of dissociation.

The density of the atomic fraction created as a result of the dissociation of molecules decreases due to the recombination 2H + M ® H 2 + M + E where M is the third body required to fulfill the conservation laws and E0 4.5 eV is the binding energy of the hydrogen molecule. In the work, an estimate was made of the recombination coefficient (the probability of recombination in a collision with a wall) and it was shown that under typical discharge conditions, i.e. gas pressure, temperature and plasma volume, the predominant process is surface recombination.

Traditionally, borosilicate or quartz glass is used as the material of the discharge tube; these materials are suitable for high temperature applications and have a low surface recombination coefficient. However, existing experimental data show that - 16 the recombination coefficient for hydrogen on borosilicate and quartz glass increases rapidly with increasing temperature. Thus, during operation, it is necessary to cool the discharge tube. Additionally, to reduce the recombination coefficient, a special treatment of the inner surface of the discharge tube, described in the works, is used, as well as a small additional (~ 0.10.5% of the main) oxygen flow.

1.3 Theoretical consideration of gas jet formation In order to correctly estimate the intensity of an atomic beam, as well as to explain the measurement results, it is necessary to answer the questions that arise when considering beam formation. The theory, unfortunately, has not yet reached a consensus on the formation of a gas jet in the hydrodynamic regime. Therefore, for the time being, we have to talk not about intensity calculations, but only about its assessment.

1.3.1 Molecular regime (outflow) Simple outflow prevails over other modes if the density of the gas behind the hole is low enough, i.e. Knudsen coefficient Kn = l / d 1 , where l is the mean free path, d is the hole diameter. In this case, there is no interaction between the particles both during the outflow and after it4. The differential beam intensity I(q) per unit solid angle dW at angle q (relative to the normal to the hole plane) is given by the cosine distribution :

I (q) = n0 A0 v f (v) cos(q)dWdv, (4) where n0 is the gas density in the source, A0 is the hole area f(v) is the Maxwell-Boltzmann velocity distribution, which is written as:

v v f (v)dv = p exp - dv, (5) z z3 with a viscosity coefficient h having a velocity v inside a pipe with a diameter d.

At R 2200, the flow regime becomes turbulent.

17 z = (2kT0 / m)1 / 2 corresponds to where is the most probable particle velocity at the source temperature T0.

The total flow f 0 through the hole is obtained by integrating over velocities and solid angle 2p:

f0 = n0 A0 v 1/s, (6) where v = (8kT0 / pm)1 / 2 is the average particle velocity in the source at temperature T0.

The beam intensity in the direction of the normal to the hole plane I(q = 0) is maximum and is given by:

f I (0) = 1/s ster. (7) p The fundamental disadvantages of a simple hole as a beam source are the low peak intensity proportional to the density n0, as well as the weak beam directivity.

1.3.2 Beam shaping by a long channel The weak directivity of a beam formed by a simple hole can be significantly improved by replacing the hole with a long channel, usually of a cylindrical cross section. The requirement for a molecular regime of gas flow in a long channel entails a loss in beam intensity. Therefore, usually, when considering the formation of a beam by a long channel, only partial fulfillment of the molecular flow conditions is required. The assumptions for such a model can be formulated as follows:

· Even for a sufficiently high pressure in the gas source, there is a section of the channel in which the conditions of the molecular flow regime are satisfied. Usually, the existence of such a section at the outlet of the channel is implied, while at the beginning of the channel, the gas is in the conditions of a hydrodynamic or intermediate flow regime with (Kn 1).

· In the section of the channel with a molecular flow regime, the density as a function of the distance z along the channel decreases linearly and reaches zero at the exit from it.

Two processes contribute to the peak intensity of the beam (in the forward direction). The first contribution is due to particles that pass through the channel without experiencing - 18 collisions. The second contribution comes from particles that have been scattered by other gas particles but have reached the end of the channel.

The described model has two specific modes depending on the ratio of the particle free path l at the gas density in the source n 0 to the channel length L:

1. Transparent channel: l L/2. For a sufficiently low gas pressure in the source, only the first process contributes to the beam intensity.

Therefore, the peak intensity is:

I (0) = n0 A0 v 1/s ster. (8) 4p It can be expressed in terms of the total gas flow f t using the Clausing formula ft = K n0 A0 v, (9) where K = 4d/3L is the geometric factor, d and L are the channel diameter and length, f t 1/ s ster.

I (0) = (10) pK Expression (10) is a formula for calculating the peak intensity of a beam formed by a long channel. It should be noted that the gas flow in the forward direction in relation to the total flow is greater compared to a simple outflow from the hole (7).

2. Opaque channel: l L/2. This regime corresponds to the case when particles have negligible probability to pass through the channel without collisions.

The criterion for opacity is the condition L/l 12 . In this case, the peak intensity is lower than with a transparent channel, and is given by :

1/ v d I (0) = 0.065 f t1 / 2 1/s ster, (11) s () - where s = 2ln is the particle collision cross section. It can be seen that in the above expression, the peak intensity does not depend on the channel length.

Based on the cases considered, it can be concluded that at a sufficiently low gas density in the source, the peak intensity I(0) is proportional to f t, and at a high density, to f t1 / 2.

19 I (0) µ f t1 / 2, The analysis of the described model shows that the dependence arising as a consequence of the linear law of change in the gas density in the channel, in fact, does not depend on this assumption. Therefore, this relationship can be extended to the case when the "opacity" continues beyond the channel, forming a cloud between the nozzle and the skimmer. In this case, the gas density decreases linearly along the channel axis z at distances less than two nozzle diameters;

and then drops to very small values ​​at a distance of several l. This allows such a model to be used despite the unrealistic assumption of n = 0 at the channel output. As a consequence, expression (11) is a reliable approximation, even if the conditions of the molecular flow regime are not satisfied. A contradiction is expected only in the case when the transition of the gas to the molecular regime occurs at a large distance from the exit from the channel or the gas jet acquires hydrodynamic features as a result of formation.

1.3.3 Hydrodynamic flow regime. Supersonic jet As soon as the gas density in the source becomes so high that the mean free path l becomes small compared to the nozzle diameter, the gas passes into an intermediate flow regime close to laminar. After exiting the nozzle, the gas undergoes adiabatic expansion. Assuming thermalization of gas particles on the nozzle surface and setting a typical nozzle temperature of about 100 K, its diameter of 2 mm, and a pressure in the dissociator discharge tube of about 1 mbar, l » 0.04 mm and Kn » 0.02. Here l is defined as kT l=, (12) 4p 2 pr where k is the Boltzmann constant, T is the gas temperature, p is the source pressure and r = 1.87·10-8 cm is the kinetic radius .

A simple orifice or a long channel used to produce a gas jet in the molecular flow regime is replaced by a conical nozzle in the case of hydrodynamic jet formation. The second "hole", usually called a skimmer and located behind the nozzle, theoretically allows the formation of a supersonic particle beam.

Under the conditions described, the gas jet emitted by the nozzle moves towards the skimmer at a hydrodynamic velocity that can greatly exceed the simple thermal velocity of the gas. This shaping method is very interesting from the point of view of obtaining beams with high intensity, as well as monochromatic beams.

Under ideal conditions of steady gas flow, it leaves the tank through a small opening and experiences adiabatic expansion. The initial enthalpy H 0 of the particles is converted into the kinetic energy of the directed flow mu 2 and the residual enthalpy H = U + PV, where U is internal energy, P – mass pressure, V – volume. The law of conservation of energy gives:

H0 = H + mu 2. (13) With the specific heat capacity of the gas at constant pressure c p, the temperature of the initial reservoir T0 and the local gas temperature T on the axis of the expanding beam, we obtain :

c p T0 = c p T + mu 2, (14) whence - 1 T = T0 1 + (g - 1) M 2, (15) 2 cp where g =. The Mach number M is the ratio of the jet velocity u to the local sound velocity cV in the gas c = (gkT / m)1 / 2. M2 is the fraction of the energy of directed motion from the thermal energy of the gas.

If the residual gas pressure p1 in the chamber directly behind the nozzle satisfies the condition:

g p1 2 g +1 (16), p 0 g - then the Mach number reaches the value M = 1 in the narrowest part of the nozzle. Here p 0 is the gas pressure in the original reservoir. Under such conditions, the gas flow reaches its maximum value.

The specific heat capacity at constant pressure c p for a monatomic gas is equal to kT. Then - 21 5 1 kT0 = mu 2 + kT + kT. (17) 2 2 The first term on the right side is the kinetic energy of the directed mass flow, the second term is the kinetic energy of thermal motion5. The third term is related to the energy contained in an ideal gas at temperature T, which forces the gas to expand.

All three terms can be written as functions of the Mach number, in particular, the kinetic energy is written as:

1 mu = c p (T0 - T) = c pT0 1 - 1 + (g - 1) M.

(18) 2 2 The Mach number increases with increasing distance l from the nozzle, since with the expansion of the gas, not only the gas temperature, but also the density (and, accordingly, the number of particle collisions per unit time) decreases with increasing distance. Finally, at a certain distance l m, the beam cooling stops and the Mach number reaches its maximum value M T and remains `frozen'.

For the distance l m, the empirical relation is given in the work:

p l m = 0.67d t 1, (19) p where d t is the nozzle throat diameter. Further, the maximum Mach number can be approximated by the expression:

g - (20) g M T = 1.2 Kn 0, where Kn 0 is the Knudsen coefficient determined by the nozzle conditions. As already mentioned, for a nozzle with a diameter of 2 mm at a temperature of 100 K, Kn 0 » 0.02, and thus the maximum Mach number is 6.

Approximately the same value under these conditions is given in the work, in which the dependence of the Mach number on the distance, measured in units of nozzle diameter, is derived.

At a distance l m there is a transition from hydrodynamic to molecular flow. Not earlier than this transition has taken place, the beam must pass the skimmer into the region of better vacuum. However, the transition to the region of better vacuum can be done in a frame of reference moving at a speed u relative to the laboratory one.

22 to influence the shock waves arising due to the interaction of the gas jet and the residual gas. Therefore, the location of the skimmer must be chosen with extreme care in order to avoid strong impact from the shock waves that form the Mach disk.

Let us assume that the beam structure is not disturbed by the presence of a skimmer, and, in addition, after the skimmer, the beam is in the molecular flow regime. Then the peak intensity on the beam axis per unit solid angle is given by the following expression:

I (v) = n s As v f (v)dWdv, (21)

(v - Ws) v f (v)dv = p exp - dv, (22) z s3 z s where z s = (2kTs / m)1 / 2 and Ws = M s (gkTs / m)1 / 2 . The index s of the variables reflects their calculation when the beam enters the neck of the skimmer. From expression (22) it is easy to find the most probable velocity of the supersonic beam:

v0 = (23) 2 gMs

gM s2 + f c = As Ac n sWs 1/s. (24) 2pl sc Here l sc is the distance between the skimmer and the collimator. According to the paper, the gain in beam intensity provided by the formation of a beam with hydrodynamic properties compared to simple gas outflow is:

1/ cp G @ p k gM s2. (25) However, the presented gain in intensity is overestimated for real beams. This is due to the fact that there is scattering of beam particles on the residual gas of the vacuum chamber.

23 1.3.4 Evaluation of the source intensity Knowing the density of particles exiting the nozzle is essential for estimating the intensity of a source of polarized atoms. A method that allows in turn to estimate the density is described in. Its idea is that the conical nozzle is replaced by a series of short tubes with a diameter of d in - d out di = i + d out. (26) N and length ln l = li =. (27) N Here l n is the nozzle depth, N is the number of tubes, d in and d out are the nozzle inlet and throat diameters, respectively.

PN N N- N- Pi i Pi P 2. Scheme of splitting the nozzle into elementary tubes.

The flow of gas from one volume to another through an elementary tube (see Fig. 2) is defined as:

Q = C i (Pi 2 - Pi1) (28) where Pi 2 and Pi1 are the pressures on both sides of the tube, moreover, to maintain pressure continuity, Pi 2 = Pi +1.1, C i is the conductivity of the tube, for which there is a universal formula for any gas flow regime:

d (P + Pi1) d i3 Tb 1.96 10 - 2 i i 2 l/s.

C i = 1.25 10 -6 + 3.04 10 4 (29) 2h l M - 24 In this formula, all linear dimensions are expressed in mm, pressures are in mbar, h = 8.58 10 -8 mbar s M - molar mass, viscosities at room temperature.

The room temperature viscosity coefficient can be converted to any other temperature using Sutherland's constant C, which for hydrogen is 73:

T 1 + C / T h T = hT0. (30) T0 1 + C / T Tn = 100 K the gas temperature will be For a nozzle with a temperature equal to Tb = 0.290 Tn = 29 K . Therefore, the viscosity coefficient will be h = 9.59 10 -9 mbar·s. (31) The particle density in the nozzle throat is defined as P n=, (32) kTb where k = 1.38·10-19 mbar·cm3/K is the Boltzmann constant. The expression for Pi 2 following from equations (28) and (29) is written as:

Pe Q e e + + Pi1 + i1 + Pi 2 = - (33), 2z 2z u xz 3.04 10 4.

M To estimate the beam density at the exit from the nozzle, it is necessary to make an assumption about the gas pressure distribution inside the dissociator discharge tube. During ABS operation at the accelerator, the inlet flow of molecular hydrogen is planned to be maintained at 1.7 mbar l/s. For such a flow, the pressure measured at the entrance to the discharge tube of the dissociator is 1.53 mbar.

Let us assume that the discharge region is approximately in the middle of the dissociator.

After the process of dissociation and recombination on the surface of the discharge tube, the degree of dissociation at its outlet end is 90%, and, consequently, the number of particles, 25, is 1.9 times greater than at the entrance to the dissociator. It follows from this that the pressure created by the gas at the entrance to the nozzle is PN 2 = 2.81 mbar.

Having carried out the calculation according to the above procedure with the number of partitions N = 90, the pressure in the nozzle throat turns out to be P12 = 2.78 mbar. Then, from expression (32) the particle density is n = 6.95 1017 cm -3. (35) Relation (24) can be rewritten as As n0 a 0 M Ac fc = (3 + gM 2) 2pl sc 3 (36) 1 1 + (g - 1)M 2 where n0 is the particle density in the nozzle throat, a0 is the speed of sound in the nozzle and M = M s.

Since a0 = 4.1376 10 4 cm/s at temperature Tb, the gas flow through the collimator is f c = 2.24 1018 1/s. (37) As already noted, such an estimate gives an overestimated value for the intensity. The reason for this is the process of intensity attenuation due to the scattering of beam particles by the residual gas.

In practice, to determine the parameters of the beam formation system at which its intensity is maximum, empirical methods are used. For this purpose, the intensity of the atomic beam is measured as a function of the geometry of the nozzle, skimmer, and collimator, the inlet gas flow, and the temperature of the nozzle.

In this regard, when creating installations for working with atomic and molecular beams, it is necessary to ensure the possibility of adjusting the fullest possible set of parameters of the beam formation system.

26 Chapter 2

Methods for creating polarization in atomic beams 2.1 Introduction The technique for creating beams of polarized protons and (or) deuterons consists in preparing a beam of neutral atoms in such a way that the spins of nuclei (protons or deuterons) in an atom are oriented predominantly along the direction of the external magnetic field. Upon subsequent ionization of atoms, for example, by electron impact, the nuclear polarization is preserved, which makes it possible to obtain beams of polarized protons (or deuterons).

The technique for creating polarized beams of neutral hydrogen or deuterium atoms is based on the fact that the nuclear and electron spins are interconnected. Therefore, by acting on the magnetic moment of the electron, one can also act on the nuclear spin.

The hydrogen atom in the ground state has an electron spin S = 1/2 with the projection mj = ±1/2 and a proton spin I = 1/2 with the projection mI = ±1/2. That. the total rrr spin of the atom F = S + I (F = 0, 1) has the projections mF = 0 and mF = 0, ±1, respectively.

Energy difference between the levels with F = 0 and F = 1 in the absence of an external magnetic field DW = h1420.4 MHz . As a result of the interaction of the magnetic moment of an atom with an external magnetic field, the level with F = 1 splits, according to the Zeeman effect. The strength of the external field is determined by the relation to the so-called. "critical" field Bc, which is defined as DW Bc = (for hydrogen 507 G), (38) (g I - g j)m B electron in units of the Bohr magneton m B = -0.927 10 -20 erg/G. That. the field strength is defined as c = B/Bc.

Energy splitting is determined by the Breit-Rabi formula:

27 DW DW 4m F + g I m B m F Bc c + (-1) F +1 c + c2.

W =- 1+ (39) 2(2 I + 1) 2I + 3.

2 mF F=1 + 1: mj=+1/2 mI=+1/ W/DW 0 2: mj=+1/2 mI=-1/ 0 - 3: mj=-1/2 mI=-1 / 4: mj=-1/2 mI=+1/ F= - - 0 2 4 6 c=B/Bc 3: Energy level diagram of a hydrogen atom in a magnetic field B. For the ground state Bc = 507 G, for the 2S1/2 state Bc = 63.4 G. The energy W is measured in units of DW = h1420.4 MHz (= 5.9 10-6 eV).

In the region c 1 the slope of the curves is determined by the magnetic moment of the electron.

r r When c 1, S and I are no longer independent vectors, therefore, in the region of weak fields, the curves are denoted by the projections mF of the total spin F.

2 mF 1: mj=+1/2 mI=+ +3/ F=3/2 2: mj=+1/2 mI= +1/ 3: mj=+1/2 mI=- W/DW -1 / 0 -3/2 4: mj=-1/2 mI=- -1/2 5: mj=-1/2 mI= 6: mj=-1/2 mI=- F=1/2 +1/ - -4 0 2 4 6 c=B/Bc 4: Energy level diagram of a deuterium atom in a magnetic field B. For the ground state Bc = 117 G, for the 2S1/2 state Bc = 14.6 G. The energy W is measured in units of DW = h327.4 MHz (= 1.4 10-6 eV).

28 For deuterium, which has a nuclear spin I = 1 (F = 1/2 and F = 3/2), the critical field is Bc = 117 G. The dependence of the energy of hyperfine levels of deuterium on the strength of the external magnetic field is shown in fig. 4.

The polarization of a proton with spin I = 1/2 (mI = ±1/2) is defined as the vector polarization N m I = +1/2 - N mI = -1 / Pz (I = 1/2) =, (40 ) N m I = +1 / 2 + N m I = -1 / where, N mI = +1/ 2 and N mI = -1 / 2, respectively, the number of atoms with a spin parallel and antiparallel to the applied external field.

To describe the polarization of the deuteron, which has a nuclear spin I = 1 (mI = –1, 0, +1), in addition to the vector polarization N m I = +1 - N m I = - Pz (I = 1) = (41) N m I = +1 + N m I = 0 + N m I = -1, the tensor polarization is also used, defined as 1 - 3 N mI = Pzz (I = 1) = (42) N m I = +1 + N m I \u003d 0 + N m I \u003d -1.

On fig. Figure 5 shows the dependences of the vector (I = 1/2 and I = 1) and tensor (I = 1) polarizations of the hyperfine levels of hydrogen and deuterium as a function of the external magnetic field. States 1, 3 for hydrogen and 1, 4 for deuterium are pure and completely polarized regardless of the magnitude of the external magnetic field.

In a strong magnetic field in the hydrogen states 2 and 4, the proton and electron are polarized in opposite directions. As the field decreases, the magnetic moments of the proton and the electron begin to precess with respect to each other, as a result of which the proton polarization decreases, as shown in Fig. 5. In the absence of an external field, the proton and electron polarizations change sinusoidally with time (with the Larmor frequency), on average creating a zero polarization. Similar reasoning can be carried out for the deuterium states 2, 3, 5, and 6.

29 Hydrogen Vector polarization Pz 0. -0. - 0.01 0.1 c = B/Bc 5. Nuclear polarization of the levels of hyperfine splitting of the hydrogen atom as a function of the external magnetic field.

Deuterium Vector polarization Tensor polarization Pz Pzz 1 + 1 0. 0. 2 0 -0. thirty. -one. -1 - 0.01 0.1 1 10 0.01 0.1 1 c = B/Bc c = B/Bc 6. Nuclear polarization of the levels of hyperfine splitting of the deuterium atom as a function of the external magnetic field.

In a strong field of a separating magnet, atoms with mj = +1/2 have zero nuclear polarization. For equally populated hydrogen levels 1, 2, i.e.

N mI \u003d + 1 / 2 \u003d N mI \u003d -1 / 2 and deuterium 1, 2 and 3, i.e. N mI = +1 = N mI = 0 = N mI = -1, from fig. It can be seen from Fig. 6 that Pz = 0 and Pzz = 0 at c 1. During the adiabatic transition to the c 1 region, Pz = +1/2 for hydrogen, Pz = +1/3 for deuterium, and Pzz = –1/3.

Thus, to create a polarization of an atomic beam, it is necessary to select atoms in one or several, in the case of tensor polarization of deuterium, hyperfine states.

30 At present, installations for creating polarized beams are widely used in various physical experiments both as sources of polarized protons and deuterons for charged particle accelerators and as polarized gas targets. The most common types of such installations today are:

sources using the Lamb shift (LSS6);

sources with optical pumping (OPPIS7);

· polarized sources of atomic beams (PABS8).

2.2 Sources using the Lamb shift (LSS) As already noted, in order to create polarization in a beam, it is necessary to select the “required” hyperfine states. This is achieved both by spatial separation of the beam components (Stern-Gerlach-type sources) and by using the technique of discharging the 2S1/2-metastable state into the short-lived 2P1/2-state.

E mS +1/ 2S1/2 mI +1/2 -1/ (t = 0.14 s) F=1 1609 MHz 0 -1/ 4.410-6 eV DELamb +1/ DEhyperfine -1/ 2P1/2 1.610-7 eV (t ~ 10-9 s) External magnetic field, G 535 Fig. 7. Diagram of energy levels of hyperfine splitting for 2S1 / 2 and 2P1 / 2 states of the hydrogen atom.

Lamb-Shift Source Optically Pumped Polarized Ion Source Polarized Atomic Beam Source - 31 Figure 7 shows the energy level diagram of the hyperfine splitting for the 2S1 / 2 and 2P1 / 2 states of the hydrogen atom. In the absence of an external magnetic field, the energy difference between these levels (the Lamb shift) is 1058 MHz. The main characteristic of the 2S1 / 2 level is that it is a metastable level with a lifetime t 2 S1 / 2 » 0.1 s. Level 2P1 / 2, in turn, is short-lived, with a lifetime t 2 P1 / 2 ~ 10 -9 s.

a Following the notation adopted in the work, we denote through the components of the 2S states with mj = +1/2, through b, respectively, the components with mj = –1/2. The corresponding components of the 2P state will be denoted by e and f. As can be seen from fig. 7, when the magnetic field reaches 535 and 605 G, the b states mix with the e states in the presence of an external electric field due to the Stark effect. This process is often referred to as the discharge of the 2S1 / 2 metastable state. Thus, only the a states with mj = +1/2 and mI = ±1/2 remain in the beam.

The difference between the energies of the levels a+ and e+, and a– and e–, where the signs “+” and “–” denote the sign of the projection of the proton spin (mI), is ~1600 MHz at a magnetic field of 535 and 605 G, respectively. Thus, by applying, in addition to the magnetic and electric fields, a high-frequency field with a frequency of ~1600 MHz and setting one or another magnetic field value, it is possible to discharge the a+ or a– level. Those.

create positive or negative vector beam polarization.

1 2 Alkali metal vapor chamber Proton source for formation Spin filter (ionizer) of metastable atoms in (2S1/2) 8. Main elements of a polarized source at the Lamb shift.

On fig. Figure 8 shows the main elements of a polarized source at the Lamb shift. To create metastable atoms, a proton beam from an ionizer (1) is passed through a chamber filled with alkali metal vapors (2), where, as a result of charge-exchange interaction, atoms are formed in the 2S1 / metastable state. Next, the beam of metastable atoms enters the spin filter (3), where the b-states are discharged and one of the two a - 32 states is selected. Thus, at the output of the source, a beam with one or another vector polarization is obtained.

Sources using the Lamb shift are mainly used as sources of polarized protons for charged particle accelerators.

However, the principle of operation of sources of this kind has found wide application in the field of polarization measurements. Such devices are called polarimeters using the Lamb shift, and make it possible to measure the polarization of beams of neutral atoms and ions with an energy of several hundred electron volts with an accuracy of about 1%. The setup described in , in particular, was used in the course of measurements of the polarization of the hydrogen and deuterium beams of ANKE ABS.

The advantages of sources using the Lamb shift include a relatively simple design, reliability, and low cost. They also make it possible to obtain proton and deuteron beams with a fairly high (7080%) degree of polarization. However, the main disadvantage of sources of this type is the low intensity rarely exceeding 0.5 mA. It is the low beam intensity that limits the use of LSSs as polarized targets, since the effective density of such a target is ~105 atoms/cm2.

2.3 Optical pumped sources (OPPIS) Optical pumped sources operate as follows.

The proton beam from the ECR9 ionizer (1, Fig. 9) is accelerated to an energy of several kiloelectronvolts and enters the neutralizer chamber filled with alkali metal vapors (2). The use of circularly polarized laser radiation makes it possible to create polarization along the electron spin in alkali metal atoms (optical pumping). Further, as a result of the charge-exchange reaction, the beam protons capture the polarized electrons of the alkali metal atoms and form neutral atoms in the metastable 2S1/2 state.

To maintain the electronic polarization of metastable atoms, chamber (2) is placed in a strong longitudinal magnetic field. Thus, a beam of neutral atoms polarized along the electron spin is formed at the output of the neutralizer.

Electron Cyclotron Resonance - 33 Laser for optical pumping of alkali metal vapors Polarization transfer Capture of polarized Proton source from electron to proton Alkali atom electron ionizer (ECR ionizer) by means of Sohn metal effect 1 2 3 Fig. 9. Operating principle of an optically pumped source.

The creation of nuclear polarization is based on the transfer of electron polarization to a proton, or the so-called Son effect. Its essence is as follows.

Since the beam of metastable atoms is polarized along the electron spin, the beam contains atoms only in states 1 and 2, which have an antiparallel nuclear spin. With an adiabatic decrease in the external magnetic field, states 1 and 2 will go over to states 1' and 2' (see Fig. 10), the nuclear spins of which are parallel. Thus, at the exit of the chamber (3, Fig. 9), the beam of metastable atoms acquires nuclear polarization.

Next, the beam of polarized metastable atoms enters the ionizer (4) or the second chamber containing alkali metal vapors, where H - ions are formed as a result of charge exchange interaction. The nuclear polarization possessed by a beam of neutral metastable atoms is preserved in this case.

E 3’ 2’ 1’ 4’ B negative B positive 10. Energy levels of the hyperfine splitting of the hydrogen atom in the 2S1 / 2 state as a function of the external magnetic field.

Thus, at the output of OPPIS there is a beam of polarized protons or H - -ions with an energy of several kiloelectronvolts.

34 Optically pumped sources have found their main application as sources of polarized protons for various accelerators. Typical, today, OPPIS parameters are: beam current ~1 mA (for DC sources) and polarization ~75%. However, despite the rather high intensity and polarization of the beam, sources of this type are rarely used as polarized targets, because the beam they create has a rather high velocity (~105 m/s), which leads to a decrease in the effective target density to 109 atoms/cm2.

2.4 Polarized Atomic Beam Sources (PABS) The idea of ​​creating a source of polarized atoms was put forward by Norman F. Ramsey in . Its essence lies in the spatial separation of hyperfine beam components in a nonuniform magnetic field and subsequent induction of transitions between hyperfine splitting levels.

vacuum pumps vacuum pumps H2, D2 4 5 6 accelerator beam vacuum pumps vacuum pumps Pic. 11: Block diagram of a source of polarized atomic hydrogen/deuterium. 1 gas flow regulator;

2 - radio frequency dissociator;

3 - gas jet formation system (nozzle, skimmer, collimator);

4 - the first group of spin-separating magnets;

5 - the first group of blocks of hyperfine transitions;

6 - the second group of spin-separating magnets;

7 - the second group of blocks of hyperfine transitions;

8 storage cell (target).

The main elements of PABS are (see Figure 11):

· a device for supplying and controlling the flow of molecular hydrogen (H2) or deuterium (D2);

dissociator, where H2 or D2 molecules dissociate into neutral atoms;

· gas jet formation system (nozzle, skimmer, collimator);

· a system for creating polarization (spin-separating magnets and blocks of hyperfine transitions).

35 An RF dissociator is usually used to create a hydrogen or deuterium atomic beam. Free electrons are accelerated in a high frequency electromagnetic field and excite vibrational levels of hydrogen molecules. This process can be represented as follows:

H 2 + e - + DE ® H 2 + e - ® H + H + e -, * where DE = 8.8 eV is the excitation energy of the vibrational levels of a hydrogen and deuterium molecule.

The flux of molecular hydrogen (deuterium) usually ranges from 0.5 to 2 mbar l/s. The upper limit is due to a decrease in the degree of dissociation at higher fluxes. Thus, it is necessary to find the optimal operating conditions under which both the degree of dissociation and the gas flow are maximum.

The beam dissociated into atoms passes through the gas jet formation system, namely, a nozzle, a skimmer, and a collimator. The nozzle temperature stabilizes around 80 K, which makes it possible to obtain the velocity distribution of the beam atomic component necessary for the maximum throughput of spin-separating magnets.

After the beam has been formed, it enters the system of spin-separating sextupole magnets, where the atomic component is separated by the orientation of the electron spin. Thus, states with electron spin projections mj = +1/2 and mj = –1/2 are spatially separated in a strong inhomogeneous magnetic field. As a result, the atomic component with mj = –1/2 leaves the beam and is removed by pumps that provide pumping out of the vacuum chamber.

To create a given vector or tensor polarization, i.e. creating a certain population of levels of hyperfine splitting, the technique of excitation of transitions in radio frequency fields is used.

The essence of this technique is as follows. With the passage of an atomic beam through the region of the magnetic field B and the radio frequency field with a frequency corresponding to the energy difference between the hyperfine splitting levels for a given B, transitions between the given levels are excited. Since the transitions between hyperfine splitting levels are bidirectional, it is necessary to exclude the possibility of reverse transitions leading to beam depolarization. This goal is achieved by inducing transitions in a gradient magnetic field. In this case, for an atom moving in such a field, the conditions for the transition are formed only in a limited region of space, where the frequency corresponds to the magnitude of the field. It is important that when the atom moved in this region, the interaction with the photon was single. This is achieved by choosing the HF field amplitude, which determines the photon density.

Sources of polarized atomic beams are widely used both for injection of polarized protons and deuterons into accelerators and as an internal gas target. Typical parameters of PABS today are: beam intensity ~5·1016 atoms/s and polarization 8595%. When using PABS as a jet target, the effective density of such a target will be ~5·1011 atoms/cm2. In the case of injection of a beam of polarized atoms into the storage cell, this makes it possible to increase the target density by one or two orders of magnitude compared to a simple gas jet.

Thus, when creating a polarized gas target for nuclear physics experiments, the source of a polarized atomic beam is the most optimal choice, among those considered above, since it provides both high density and high polarization of the target.

37 Chapter 3

Source of polarized atomic hydrogen and deuterium for the internal gas target of the ANKE spectrometer 3.1 Brief description of the design In fig. 12 shows the position of ANKE ABS on the COZY storage ring between the deflecting magnet D1 and the central magnet of the ANKE spectrometer D , and also shows a special vacuum chamber designed to install various types of targets (accumulation cell, solid-state, cluster and pellet targets). Since the space in the storage ring tunnel is limited, the source will be installed vertically. Such setup scheme is also Source of polarized ANKE central magnet (D2) atomic hydrogen and deuterium ANKE ABS Deflecting magnet (D1) Target chamber Fig. 12. ANKE ABS and a special vacuum chamber for mounting various types of targets on the COZY storage ring. The source of polarized atomic hydrogen and deuterium is located between the deflecting magnet D1 and the central magnet of the spectrometer D2. COZY beam direction from left to right.

38 will make it possible to place the source closest to the central magnet of the spectrometer, which, in turn, is one of the main factors determining the angular capture of the spectrometer.

A detailed drawing of ABS is shown in fig. 13. The created design takes into account the experience of creating and operating similar sources in IUCF and HERMES / DESY, however, it has a number of advantages over them.

The intermediate flange (7) between the upper and lower vacuum chambers is attached to the bearing bridge connecting the yoke of magnets D1 and D2 (see Fig. 12). This fastening ensures the movement of the ABS and the vacuum chamber of the storage cell as a whole, when the central magnet D2 is shifted, and also allows for quick dismantling of the source and the carrier bridge as a whole.

To create a hydrogen or deuterium atomic beam, a radio frequency dissociator is used (1, Fig. 13). RF power is supplied to the parallel LC circuit from a 13.56 MHz generator. Cooling of the discharge tube is provided by the flow of an alcohol-water mixture between two outer coaxial tubes of larger diameter. To stabilize the nozzle temperature in the range of 40100 K, a cryogenerator (2) was used, connected to the nozzle via a flexible copper thermal bridge (3). The upper vacuum chamber is divided by two movable aluminum partitions (4) into three stages of differential pumping (I, II, III). The skimmer, which serves to form the gas jet, is fixed on the partition separating chambers I and II. The design of the upper flange allows movement of the nozzle axis relative to the skimmer axis in all directions. The use of a flexible vacuum connection between the dissociator flange and the upper flange of the vacuum chamber makes it possible to vary the distance between the nozzle and the skimmer without disturbing the vacuum. A collimator is installed on the partition separating chambers II and III, which finally forms the gas jet.

The first group of spin-separating sextupole magnets (5) provides, as in the classical Stern-Gerlach experiment, the spatial separation of the beam along the electron spin. In this case, the component with mj = +1/2 is focused in the strong inhomogeneous magnetic field of the sextupole and enters the hyperfine transition block (6), while the component with mj = –1/2 is defocused and removed by pumps that provide pumping out of the vacuum chamber. The block of ultrafine transitions (6), like the magnets (5), is rigidly fixed on the ABS central flange (7), which determines the entire geometry of the source.

39 I II III IV p Fig. 13. ANKE ABS drawing. Explanations are given in the text.

40 Chamber IV houses the second group of spin-separating sextupole magnets (8) and additional blocks of hyperfine transitions (9), which are responsible for creating the tensor polarization of the deuterium beam.

Finally, at the very bottom is shown a prototype storage cell (10) planned for use on the COZY storage ring.

On fig. 14 is a photograph of ANKE ABS and a Lamb shift polarimeter in the IKP10 laboratory.

Rice. 14. Photo of ANKE ABS in the laboratory. The height of the upper vacuum chamber is 80 cm.

Institut fr Kernphysik, Forschungszentrum Jlich, D-52428 Jlich, Germany - 41 Summing up, we can say that the specifics of the design, dictated by the use of the source under the conditions of the experiment at the accelerator (limited access for maintenance, severe limitations in volumes for experimental equipment, etc. .) composed:

· in compactness, which allows to install the source in the limited space of the COZY storage ring tunnel and at the same time provide the necessary space for the detector system of the ANKE spectrometer.

· in the mobility of the source for quick mounting and dismounting on the storage ring, which makes it possible to drastically reduce the loss of acceleration time when replacing the source with one of the non-polarized targets (solid-state, cluster, pellet target) used in other physical experiments on the ANKE spectrometer.

3.2 Vacuum system One of the main factors determining the intensity of the atomic beam, and hence the density of the target, is the pumping speed of scattered atoms and molecules in the first and second chambers of the source (I, II see Fig. 13). The interaction of the residual gas with beam particles destroys the directed flow of atoms and, ultimately, leads to a decrease in the target density. To minimize the effects of scattering and attenuation of the beam on the residual gas in the sources of atomic beams, a powerful differential pumping system is used, which provides a vacuum in the first and second chambers at a level of 10-410-5 mbar.

3.2.1 Construction of the vacuum chamber The ABS vacuum chamber consists of two cylindrical vacuum chambers fixed above and below the central bearing flange (7, Fig. 13), with dimensions of 40050050 mm3. The wall thicknesses of the upper and lower vacuum chambers made of stainless steel are 8 and 2.5 mm, respectively. To ensure differential pumping, the upper vacuum chamber, having an inner diameter of 390 mm, is divided into three parts by two dividing partitions. Unlike other sources, the dividing baffles are made movable, which greatly simplifies the procedure for optimizing the gas jet formation system.

The complex shape of the baffles is caused by the desire to improve the vacuum conditions near the nozzle, skimmer and collimator and to provide the maximum open space for turbomolecular pumps pumping out the first and second vacuum chambers. The upper partition separating chambers I and II has a diagnostic glass window for observation and nozzle change through a special flange in chamber II. Both baffles with a diameter of 389 mm and a height of 200 mm are made of aluminum by precision casting. Despite the fact that the aluminum casting has a porous surface, during operation there were no problems associated with the deterioration of the vacuum in the upper vacuum chamber. The partitions are processed in such a way that the conductivity of the gap, which is less than 0.5 mm, between the inner surface of the vacuum chamber and the surface of the partition is negligible. This made it possible to avoid additional compaction and significantly simplified the design of the upper vacuum chamber.

Rice. 15. Upper movable partition.

Ball guides fixed on the baffle and sliding along the inner surface of the vacuum chamber make it easy to move the baffles along the beam axis. The position of the lower baffle, on which the collimator is fixed, can be changed with the help of two micrometer vacuum inlets fixed on the central bearing flange without breaking the vacuum.

Thus, it should be noted that the use of movable partitions of complex shape made it possible to:

· for the first time it was possible to combine three stages of the source in one vacuum chamber, which significantly reduced its linear dimensions and minimized the number of seals;

reduce the distance from the gas source to vacuum pump and the ratio of the "passive" surface of the chambers to the "pumping out", which led to a significant improvement in pumping conditions;

Physicists have a habit of taking the simplest example some phenomenon and call it "physics", and it is more difficult to give examples to be torn to pieces by other sciences, say, applied mathematics, electrical engineering, chemistry or crystallography. Even physics solid body for them, only "semi-physics", because she cares too much special issues. For this reason, we will omit many interesting things in our lectures. For example, one of the most important properties of crystals and most substances in general is that their electrical polarizability is different in different directions. If you apply an electric field in any direction, then the atomic charges will shift slightly and a dipole moment will arise; the magnitude of this moment depends very strongly on the direction of the applied field. And this, of course, is a complication. To make life easier for themselves, physicists start the conversation with the special case where the polarizability is the same in all directions. And we leave other cases to other sciences. Therefore, for our further considerations, we will not need at all what we are going to talk about in this chapter.

The mathematics of tensors are especially useful for describing the properties of substances that change with direction, although this is just one example of its use. Since most of you are not going to become physicists, but intend to work in the real world, where the dependence on direction is very strong, sooner or later you will need to use a tensor. So, so that you don't have a gap here, I'm going to tell you about tensors, although not in great detail. I want your understanding of physics to be as complete as possible. Electrodynamics, for example, we have a completely finished course; it is as complete as any course in electricity and magnetism, even an institute one. But mechanics is not finished with us, because when we studied it, you were not yet so firm in mathematics and we could not discuss such sections as the principle of least action, Lagrangians, Hamiltonians, etc., which represent the most elegant way descriptions of mechanics. However, we still have a complete set of laws of mechanics, with the exception of the theory of relativity. To the same extent as electricity and magnetism, we have many sections completed. But here we will not finish quantum mechanics; However, you need to leave something for the future! And yet, what is a tensor, you still should know now.

In ch. 30 we emphasized that the properties of a crystalline substance are different in different directions - we say that it is anisotropic. The change in the induced dipole moment with a change in the direction of the applied electric field is only one example, but that is what we will take as an example of a tensor. We assume that for a given direction of the electric field, the induced dipole moment per unit volume is proportional to the strength of the applied field . (For many substances, at not too large, this is a very good approximation.) Let the constant of proportionality be . Now we want to consider substances that depend on the direction of the applied field, such as the tourmaline crystal you know, which gives a double image when you look through it.

Suppose we have found that for a certain selected crystal, an electric field directed along the axis gives a polarization directed along the same axis, and an electric field of the same magnitude with it, directed along the axis, leads to some other polarization, also directed along axes . What happens if an electric field is applied at an angle of 45°? Well, since it will be just a superposition of two fields directed along the axes and , then the polarization is equal to the sum of the vectors and , as shown in Fig. 31.1, a. The polarization is no longer parallel to the direction of the electric field. It is not difficult to understand why this happens. There are charges in the crystal that are easy to move up and down, but which are very difficult to move sideways. If the force is applied at an angle of 45 °, then these charges are more likely to move up than to the side. As a result of such asymmetry of the internal elastic forces, the displacement does not proceed in the direction of the external force.

Fig. 31.1. Addition of polarization vectors in an anisotropic crystal.

Of course, the 45° angle is not highlighted. The fact that the induced polarization is not directed along the electric field is also true in the general case. Before that, we were simply “lucky” to choose such axes and for which the polarization was directed along the field . If the crystal were rotated with respect to the coordinate axes, then an electric field directed along the axis would cause polarization both along the axis and along the axis. In a similar way, the polarization caused by a field directed along the axis would also have both - and -components. So instead of Fig. 31.1, and we would get something similar to Fig. 31.1b. But despite all this complication, the magnitude of the polarization for any field is still proportional to its magnitude.

Let us now consider the general case of an arbitrary orientation of the crystal with respect to the coordinate axes. An electric field directed along the axis gives a polarization with components along all three axes, so we can write

By this I mean only that an electric field directed along the axis creates polarization not only in this direction, it leads to three polarization components , and , each of which is proportional to . We called the proportionality coefficients , and (the first icon indicates which component we are talking about, and the second refers to the direction of the electric field).

Similarly, for a field directed along the axis, we can write

and for the field in -direction

Further we say that the polarization depends linearly on the field; therefore, if we have an electric field with components and , then the polarization component will be the sum of two defined by equations (31.1) and (31.2), but if it has components in all three directions , and , then the polarization components should be the sum of the corresponding terms in equations (31.1), (31.2) and (31.3). In other words, it is written as

Physicists have a habit of taking the simplest example of a phenomenon and calling it “physics,” and leaving more difficult examples to other sciences, such as applied mathematics, electrical engineering, chemistry, or crystallography. Even solid state physics for them is only "semiphysics", because it is concerned with too many special issues. For this reason, we will omit many interesting things in our lectures. For example, one of the most important properties of crystals and most substances in general is that their electrical,
polarizability is different in different directions. If you apply an electric field in any direction, then the atomic charges will shift slightly and a dipole moment will arise; the magnitude of this moment depends very strongly on the direction of the applied field. And this, of course, is a complication. To make life easier for themselves, physicists start the conversation with the special case where the polarizability is the same in all directions. And we leave other cases to other sciences. Therefore, for our further considerations, we will not need at all what we are going to talk about in this chapter.

The mathematics of tensors are especially useful for describing the properties of substances that change with direction, although this is just one example of its use. Since most of you are not going to become physicists, but intend to study real in a world where the dependence on direction is very strong, sooner or later you will need to use a tensor. So, so that you don't have a gap here, I'm going to tell you about tensors, although not in great detail. I want your understanding of physics to be as complete as possible. Electrodynamics, for example, we have a completely finished course; it is as complete as any course in electricity and magnetism, even an institute one. But mechanics is not finished with us, because when we studied it, you were not yet so firm in mathematics and we could not discuss such sections as the principle of least action, Lagrangians, Hamiltonians, etc., which represent mostelegant way descriptions of mechanics. However, the full set laws mechanics, with the exception of the theory of relativity, we still have. To the same extent as electricity and magnetism, we have many sections completed. But here we will not finish quantum mechanics; However, you need to leave something for the future! And yet, what is a tensor, you still should know now.

In ch. 30 we emphasized that the properties of crystalline matter are different in different directions - we say that it anisotropically. The change in the induced dipole moment with a change in the direction of the applied electric field is just one example, but that is what we will take as an example of a tensor. We will assume that for a given direction of the electric field, the induced dipole moment per unit volume P is proportional to the strength of the applied field E. (For many substances at not too large E, this is a very good approximation.) Let the constant of proportionality be α . Now we want to consider substances that have α depends on the direction of the applied field, for example the tourmaline crystal you know, which gives a double image when you look through it.

Suppose we have found that for some selected crystal, the electric field E 1 directed along the axis X, gives polarization P 1 directed along the same axis, and alonethe same size as him electric field E 2 directed along the axis y, leads to some other polarization P 2 , also directed along the axis y. What happens if an electric field is applied at an angle of 45°? Well, since it will just be a superposition of two fields directed along the axes X and y, then the polarization P is equal to the sum of the vectors P 1 and P 2 , as shown in FIG. 31.1, a. The polarization is no longer parallel to the direction of the electric field. It is not difficult to understand why this happens. There are charges in the crystal that are easy to move up and down, but which are very difficult to move sideways. If the force is applied at an angle of 45 °, then these charges are more likely to move up than to the side. As a result of such asymmetry of the internal elastic forces, the displacement does not proceed in the direction of the external force. Of course, the 45° angle is not highlighted. That induced polarization not directed along the electric field, just and in general. Before that, we were just “lucky” to choose such axes X and y, for which the polarization P was directed along the field E. If the crystal were rotated with respect to the coordinate axes, then the electric field E 2 directed along the y axis would cause polarization both along the axis y, as well as along the axis X. Similarly, the polarization P caused by a field directed along the axis X, would also have X-, so are the y-components. So instead of Fig. 31.1, a we would get something similar to Fig. 31.1b. But despite all this complication, magnitude polarization P for any field E is still proportional to its magnitude.

Let us now consider the general case of an arbitrary orientation of the crystal with respect to the coordinate axes. Electric field directed along the axis X, gives a polarization P with components in all three axes, so we can write

By this I only want to say that the electric field directed along the axis X, creates polarization not only in this direction, it leads to three components of polarization R x,RU and Pz, each of which is proportional E x. We called the coefficients of proportionality a xx, a yx and a zx(the first icon indicates which component is involved, and the second refers to the direction of the electric field).

Similarly, for a field directed along the axis y, we can write

and for the field in the z-direction

Further we say that the polarization depends linearly on the field; so if we have an electric field E with components X and y, then the x-component of the polarization P will be the sum of two R x, defined by equations (31.1) and (31.2), but if E has components in all three directions x, y and z, then the polarization components P must be the sum of the corresponding terms in equations (31.1), (31.2) and (31.3). In other words, P is written as

The dielectric properties of a crystal are thus completely described by nine quantities (α xx, α xy,α xz, α yz , ...), which can be written as a symbol α ¡j . (Indices i and j replace one of three letters: x, y or z.) An arbitrary electric field E can be decomposed into components E x, E y and Ez. Knowing them, you can use the coefficients α ¡j and find P x, P y and P z , which together give the complete polarization of P. A set of nine coefficients α ¡j called tensor- in this example polarizability tensor. Just like the three quantities (E x, E y,Ez) "form a vector E", and we say that nine quantities (α xx, α hu,...) "form a tensor α ¡j ».

480 rub. | 150 UAH | $7.5 ", MOUSEOFF, FGCOLOR, "#FFFFCC",BGCOLOR, "#393939");" onMouseOut="return nd();"> Thesis - 480 rubles, shipping 10 minutes 24 hours a day, seven days a week and holidays

Isupov Alexander Yurievich. Measurements of the tensor analyzing ability T20 in the reaction of deuteron fragmentation into pions at zero angle and software development for data acquisition systems for installations on polarized beams: dissertation ... Candidate of Physical and Mathematical Sciences: 01.04.16, 01.04.01 .- Dubna, 2005. - 142 p.: ill. RSL OD, 61 06-1/101

Introduction

I Setting up the experiment 18

1.1 Motivation 18

1.2 Experimental setup 20

1.3 Methodological measurements and modeling 24

1.4 Organization and principle of operation of the trigger 33

II Software 40

II.1 Introductory remarks 40

II.2 Data collection and processing system qdpb 42

II.3 Configurable views of data and hardware 56

II.4 Session-dependent means of data representation. 70

II.5 DAQ system SPHERE 74

II. 6 Polarimeter Data Acquisition Systems 92

III. Experimental results and discussion 116

III.1 Analysis of sources of systematic errors 116

III.2 Experimental data 120

Sh.3. Discussion of experimental data 127

Conclusion 132

Literature 134

Introduction to work

B.1 Introduction

The dissertation paper presents the experimental results of measurements of the tensor analyzing power of Ggo in the reaction of fragmentation of tensor polarized deuterons into cumulative (sub-threshold) pions. The measurements were carried out by the SPHERE collaboration on a beam of tensor polarized deuterons at the accelerator complex of the High Energy Laboratory of the Joint Institute Nuclear Research(LHE JINR, Dubna, Russia). The study of polarization observables provides more detailed, compared to reactions with non-polarized particles, information about the interaction Hamiltonian, reaction mechanisms, and the structure of the particles involved in the reaction. To date, the question of the properties of nuclei at distances smaller than or comparable to the size of a nucleon has not been adequately studied both from the experimental and theoretical points of view. Of all the nuclei, the deuteron is of particular interest: firstly, it is the most studied nucleus from both experimental and theoretical points of view. Secondly, for the deuteron, as for the simplest nucleus, it is easier to understand the reaction mechanisms. Third, the deuteron has a nontrivial spin structure (spin equal to 1 and a nonzero quadrupole moment), which provides wide experimental possibilities for studying spin observables. The measurement program, within the framework of which the experimental data presented in the dissertation work were obtained, is a natural continuation of studies of the structure of atomic nuclei in reactions with the production of cumulative particles in the collision of unpolarized nuclei, as well as polarization observables in the deuteron decay reaction. The experimental data presented in the dissertation work make it possible to advance in understanding the spin structure of the deuteron at small internucleon distances and supplement the information on the structure of the deuteron obtained in experiments with a lepton probe and in the study of the breakup reaction of tensor polarized deuterons, and therefore seem to be relevant. To date, the data presented in the dissertation work are the only ones, since such studies require beams of polarized deuterons with an energy of several GeV, which at present and in the next few

years will be available only at the JINR LHE accelerator complex, where it is natural to continue research in this direction. The mentioned data were obtained as part of an international collaboration, were reported at a number of international conferences, and also published in peer-reviewed journals.

Further in this chapter, we give the information about cumulative particles necessary for further presentation, the definitions used in the description of polarization observables, and also give short review known in the literature results on the deuteron decay reaction.

B.2 Cumulative particles

Studies of the regularities of the birth of cumulative particles have been carried out since the beginning of the seventies of the XX century, , , , , , , , , , , , . The study of reactions with the production of cumulative particles is interesting in that it provides information about the behavior of the high-momentum (> 0.2 GeV/c) component in fragmenting nuclei. These large internal momenta correspond to small ones (xx > 1, where the cross sections become very small.

First of all, let us define what will be further understood by the term "cumulative particle" (see, for example, the references therein). Particle with, born in reaction:

Ag + AP -Ї- c + x, (1)

is called "cumulative" if the following two conditions are met:

the particle c is produced in a kinematic region inaccessible in the collision of free nucleons having the same momentum per nucleon as the nuclei A/ and Ats in reaction (1);

particle with belongs to the fragmentation region of one of the colliding particles, i.e. must be done either

\YAt-Yc\^\YAn-Yc\., (2)

where Yi is the speed of the corresponding particle z. It follows from the first condition that at least one of the colliding particles must be a nucleus. It can be seen from the second condition that the colliding particles enter this definition asymmetrically. In this case, the particle that lies closer to the cumulative one in terms of speed will be called the fragmenting particle, and the other of the colliding particles will be called the particle on which fragmentation occurs. Usually, experiments with the production of cumulative particles are set up in such a way that the detected particle lies outside the rapidity interval [Vpn, )%]. In this case, the second condition reduces to the requirement of a sufficiently large collision energy:

\UAp - Atwith\ « \YAl~ Yc\ = |U L// - Yc\ + \YAn-YAl\ . (4)

It follows from experimental data (see, for example, , , , , , , , ) that for experiments on a fixed target, the shape of the spectrum of cumulative particles weakly depends on the collision energy, starting from the energies of incident particles Th > 3-4 GeV. This statement is illustrated in Fig. 1, reproduced from , which shows the dependences on the energy of the incident proton: (b) the ratio of the outputs of pions of different signs 7r~/tr + and (a) the parameter of the inverse slope of the spectrum T 0 for approximation Edcr/dp= Sehr(- T^/Tq) cross sections for the production of cumulative pions measured at an angle of 180. This means that the independence of the shape of the spectra from the primary energy begins with the difference in the speeds of the colliding particles \Yau-YAl\ > 2.

Another established pattern is the independence of the spectra of cumulative particles from the type of particle on which fragmentation occurs (see Fig. 2).

Since the dissertation paper considers experimental data on the fragmentation of polarized deuterons into cumulative pions, the regularities established in reactions with the production of cumulative particles (dependence on the atomic mass of the fragmenting nucleus, dependence on the type of detected particle, etc.) will not be discussed in more detail. If necessary, they can be found in the reviews: , , , .

- h

h 40 ZO

M і-

present experiment

About 7G*1TG "I

+ -

Present experiment v Reference 6

Rice. 1: Dependence on the energy of the incident proton (TR) (a) the inverse slope parameter T 0 and (b) the ratio of the outputs tt~/tg + , integrated starting from a pion energy of 100 MeV. Figure and data marked with circles are taken from . Data marked with triangles are cited from .

B.3 Description of polarized states of particles with spin 1

For the convenience of further presentation, we give a brief overview of the concepts , , which are used in describing the reactions of particles with spin 1.

Under ordinary experimental conditions, an ensemble of spin particles (beam or target) is described by the density matrix R, whose main properties are as follows:

Normalization Sp(jo) = 1.

hermiticity p = p + .

D-H"

.,- Withf

O - Si 4 -Pbsh l

, . f,

" -" -. і.. -|-і-

Cumulative Scale Variable Xwith

Rice. 2: Dependence of the cross section for the production of cumulative particles on the cumulative scaling variable Xwith (57) (see paragraph III.2) for the fragmentation of a deuteron beam on various targets into pions at zero angle. Picture taken from work.

3. Average from the operator About calculated as (O) = Sp(Op).

The polarization of an ensemble (for definiteness, a beam) of particles with spin 1/2 is characterized by the direction and average value of the spin. As regards particles with spin 1, one should distinguish between vector and tensor polarizations. The term "tensor polarization" means that the description of particles with spin 1 uses a tensor of the second rank. In general, particles with spin / are described by the rank tensor 21, so that for / > 1 one should distinguish between the polarization parameters of the 2nd and 3rd ranks, and so on.

In 1970, at the 3rd International Symposium on Polarization Phenomena, the so-called Madison Convention was adopted, which, in particular, regulates the notation and terminology for polarization experiments. When recording a nuclear reaction L(a, b)B Arrows are placed over particles that react in a polarized state or whose polarization state is observed. For example, the notation 3 H(rf,n) 4 He means that the unpolarized target 3 H is bombarded by polarized deuterons d and that polarization of the resulting neutrons is observed.

When talking about measuring the polarization of a particle b in a nuclear reaction, we mean the process L(a, b) B, those. in this case, the beam and the target are not polarized. The parameters describing the changes in the reaction cross section when either the beam or the target (but not both) are polarized are called the analyzing powers of the reaction of the form A(a, b)B. Thus, apart from special cases, polarizations and analytical abilities must be clearly distinguished, since they characterize different reactions.

Type reactions A(a, b)B, A(a, b)B etc. are called polarization transfer reactions. Parameters relating the spin moments of a particle b and particles a, are called polarization transfer coefficients.

The term "spin correlations" is applied to experiments on the study of reactions of the form A(a, b)B and A(a, b)B, moreover, in the latter case, the polarization of both resulting particles must be measured in the same event.

In experiments with a beam of polarized particles (measurements of analyzing abilities) in accordance with the Madison Convention, the axis z guided by the momentum of the beam particle kjn, axis y - on to(P X kout(i.e. perpendicular to the reaction plane), and the axis X must be directed so that the resulting coordinate system is right-handed.

Polarization state of a system of particles with spin I can be fully described by (2/+1) 2 -1 parameters. Thus, for particles with spin 1/2, three parameters pi form a vector R, called the polarization vector. Expression in terms of the spin 1/2 operator, denoted a, following:

Pi =yy,Z, (5)

where angle brackets mean averaging over all particles of the ensemble (in our case, the beam). Absolute value R limited \p\ 1. If we incoherently mix n + particles in a pure spin state, i.e. completely polarized in some given direction, and n_ particles completely polarized in the opposite direction, the polarization will be p =" + ^~ , or

+ p = N + ~N_, (6)

if under N + = PP+ P _ and JV_ = ~jf^- understand the fraction of particles in each of the two states.

Since the polarization of particles with spin 1 is described by a tensor, its representation becomes more complicated and less visual. Polarization parameters are some observable quantities

spin operator 1, S. Two different sets of definitions for the corresponding polarization parameters are used - Cartesian tensor moments ri rc and spin tensors tjsq. In Cartesian coordinates, according to the Madison Convention, the polarization parameters are defined as

Pi= (Si)(vector polarization), (7)

pij- -?(SiSj.+ SjSi)- 25ij(tensor polarization), (8)

where S- spin operator 1, i, j= x,y,z. Insofar as

S(S+1).= 2 , (9)

we have a connection

Рхх + Ruy + Pzz = 0 (10)

Thus, the tensor polarization is described by five independent quantities (pzx, Ruu, Rhu, pXz, Pyz)-> which, together with the three components of the polarization vector, gives eight parameters for describing the polarized state of a particle with spin 1. The corresponding density matrix can be written as:

P = \i^ + \is + \vij(SiSj+ SjSi)).. (11)

The description of the polarization state in terms of spin tensors is convenient, since they are easier than Cartesian ones, they are transformed during rotations of the coordinate system. The spin tensors are related to each other by the following relationship (see):

hq~N(fc i9i fc 2&|fcg)4 w ,4 2(ft , (12)

where (kiqik 2 q2\kq) ~ Clebsch-Gordan coefficients, and N- normalization coefficient, chosen so that the condition is fulfilled

Sp.(MU) = (^ + 1)^,^ (13)

The lowest spin moments are:

І 11 \u003d 7 ^ (^ + ^ y) "(14)

t\ -\ = -^(Sx- iSy) .

For spin/index to runs values ​​from 0 to 21, a |e| j. Negative values q can be discarded because there is a connection tk _ q = (-1)41 + $# spin 1 spherical tensor moments are defined as

t\\ ~ ~*-(Sx ) (vector polarization),

tii.= -&((Ss+ iSy)Sg.+ Sx(Sx+ iSy)) ,

hi = 2 ((Sx+ iSy) 2 ) (tensor polarization).

Thus, vector polarization is described by three parameters: real t\o and comprehensive "tu, and tensor polarization - five: real I20 and complex I2b ^22-

Next, consider the situation when the spin system has axial symmetry with respect to the С axis (notation z leave for the coordinate system associated with the reaction under consideration, as described above). This particular case is interesting because beams from sources of polarized ions usually have axial symmetry. Let us imagine such a state as an incoherent mixture containing a fraction N+ particles with spins along, fraction N- particles with spins along - and the fraction JVo of particles with spins uniformly distributed in directions in the plane perpendicular to k. In this case, only two polarization moments of the beam are nonzero, t to (or sch) and t 2 Q(or R#). Let us direct the quantization axis along the symmetry axis C and replace i in notation with t and z to (". It is obvious that (*%) is simply equal to N + - iV_, and according to (15) and (7):

tyu = \-(iV+-JV_) or (17)

p = (N + - i\L) (vector polarization).

From (16) and (8) it follows that

T2o = -^(l-3iVo) or (18)

Ptf= (1 - 3iVo) (tensor polarization or alignment),

where it is used that (JV+ + i\L) = (1 - iV 0).

If all moments of the 2nd rank are absent (N 0 = 1/3) speak of a purely vector beam polarization. The maximum possible values ​​of the polarization of such a beam

tії" = yfifi or C 19)

pmax. _ 2/3 (pure vector polarization).

For the case of purely tensor polarization (tu = 0) from equations (17) and (18) we obtain

-y/2 2 oily (20)

The lower bound corresponds No= 1, top - N+ ~ N_= 1/2.

In general, the axis of symmetry WITH, polarized beam from the source can be arbitrarily oriented with respect to the coordinate system xyz, associated with the reaction in question. Let us express the spin moments in this system. If the axis orientation ( set by angles /3 (between the axes z and C) and f(rotation on - f around the axis z brings the C-axis into a plane yz), as shown in Fig. 3, and in the system WITH, beam polarizations are t\ 0 , m 20 , then the tensor moments in the system xyz are equal:

Vector moments: Tensor moments:

t 20 = y(3cos 2 /?- i) , (21)

itn = ^8 IP0ЄIf. til= " %T2 % Silljgcos/fe**",

y/2 y/2

In the general case, the invariant section a = Edijdp reactions A(a,b)B is written as:

Quantities T)sch are called the analyzing abilities of the reaction. The Madison Convention recommends that tensor analyzing powers be denoted as Tkq (spherical) and AitAts(Cartesian). Four analyzing abilities - vector GTand and tensor T 20 , TG\ and Тії

Rice. 3: Orientation of the axis of symmetry ( polarized beam relative to the coordinate system xyz, associated with the reaction xz- reaction plane, /3 - angle between the axes z(direction of the incident beam) and, rotation on - f around the axis z leads axle; into the plane yz.

- are real due to parity conservation, and 7\ 0 = 0. Taking into account these restrictions, equation (22) takes the form:

a = cro, , , . On the whole, the experimental spectra obtained are well described by the spectra

tator mechanism using conventional WFD, for example, the Reid or Paris WFD.

Rice. 5: Nucleon relative momentum distribution in the deuteron extracted from experimental data for various reactions involving the deuteron. Picture taken from work.

So, from Fig. 5 shows that the momentum distributions of nucleons in the deuteron are in good agreement, extracted from the data for the reactions: inelastic scattering of electrons on the deuteron d(e,e")X, elastic proton-deuteron backward scattering p(d,p)d, and the collapse of the deuteron. Except for internal pulse interval to from 300 to 500 MeV/c, the data are described by the spectator mechanism using the Paris PFD. Additional mechanisms have been invoked to explain the discrepancy in this area. In particular, taking into account the contribution from pion rescattering in the intermediate state , , makes it possible to satisfactorily describe the data. However, the uncertainty in the calculations is about 50 % due to uncertainty in the knowledge of the vertex function irn, which, in addition, in such calculations must be known outside the mass shell. In order to explain the experimental spectra, we took into account the fact that for large internal momenta (i.e., small internucleon distances)

yany Inn- 0,2/"to) non-nuclear degrees of freedom may appear. In particular, in that work, an admixture of the six-quark component \6q), the probability of which was ~-4.%.

Thus, it can be noted that, on the whole, the spectra of protons obtained during the fragmentation of deuterons into protons at zero angle can be described up to internal momenta of ~ 900 MeV/c. In this case, it is necessary either to take into account the diagrams following after the momentum approximation, or to modify the PFD taking into account the possible manifestation of nonnucleon degrees of freedom.

The polarization observables for the deuteron breakup reaction are sensitive to the relative contribution of the PFD components corresponding to different angular momenta, so experiments with polarized deuterons give Additional information about the structure of the deuteron and reaction mechanisms. At present, there are extensive experimental data on the tensor analyzing power T 2 about for the breakup reaction of tensor polarized deuterons. The corresponding expression in the spectator mechanism is given above, see (30). Experimental data for T 2 q, obtained in works , , , , , , , , , are shown in Figs. 6, which shows that starting from internal momenta of the order of 0.2 × 0.25 GeV/c, the data are not described by generally accepted two-component PFDs.

Accounting for interaction in the final state improves agreement with experimental data up to momenta of the order of 0.3 GeV/c. Accounting for the contribution of the six-quark component in the deuteron allows one to describe the data up to internal momenta of the order of 0.7 GeV/c. Behavior T 2 about for momenta of the order of 0.9 - L 1 GeV/c is in best agreement with calculations in the framework of QCD using the method of reduced nuclear amplitudes, , taking into account the antisymmetrization of quarks from different nucleons.

So, summing up the above:

Experimental data for the fragmentation cross section of unpolarized deuterons into protons at zero angle can be described in terms of the nucleon model.

Up to now, the data for T20 have been described only in terms of non-nucleon degrees of freedom.

Methodical measurements and modeling

Measurements of the tensor analyzing capacity G20 of the reaction d + A -(0 - 0) + X fragmentation of relativistic polarized deuterons into cumulative pions were carried out on channel 4V of the slow extraction system of the Synchrophasotron LHE JINR. Channel 4B is located in the main measuring hall of the accelerator complex (the so-called building 205). Polarized deuterons were created by the POLYA-RIS source, which is described in .

The measurements were carried out in following conditions: 1. the stretching value (extraction time) of the beam was 400 500 ms; 2. repetition rate 0.1 Hz; 3. the intensity varied in the range from 1109 to 5109 deuterons per drop; 4. The magnitude of the tensor polarization of the deuteron beam was pzz 0.60-0.77, varying slightly (by no more than 10%, see .25; 5: the quantization axis for polarization was always directed vertically; 6. Three polarization states were provided - "+" (positive sign of polarization), "-" (negative sign of polarization), "0" (absence of polarization), which changed every accelerator cycle, so that in three successive cycles the beam had different polarization states. In the first series of measurements, carried out in March 1995, the magnitude of the vector and tensor polarization was measured at the beginning and end of the full cycle (session) of measurements using a high-energy polarimeter described in the work - the so-called. polarimeter ALPHA.

In the first series of measurements , , , we used the one shown in Fig. 8 is the configuration of the setup with the target located at the focus F3 (we will call it the “first setup” for brevity).

The extracted beam of primary deuterons was focused by a doublet of quadrupole lenses onto a target located at focus F3. The intensity distribution on the target in the plane perpendicular to the beam direction was close to the Gaussian distribution with dispersions mx n 6 mm and y ≈ 9 mm along the horizontal and vertical axes, respectively. Cylindrical carbon targets (50.4 g/cm2 and 23.5 g/cm2) with a diameter of 10 cm were used, which made it possible to assume that the entire primary beam hit the target.

Monitoring of the intensity of the deuteron beam incident on the target was carried out using the ionization chamber 1C (see Fig. 8), located in front of the target at a distance of 1 m from it, and two scintillation telescopes Mi and M2, three counters each, aimed at an aluminum foil 1 mm thick. Monitors have not been completely calibrated. The difference in determining the relative intensity on different monitors reached 5%. This difference was included in the systematic error.

Scintillation counters at foci F4 (F4b F42), F5 (F5i) and F6 (F6i) were used to measure the time of flight at bases of 74 meters (F4-F6) and 42 meters (F5-F6). Scintillation counters Si and Sz, and, if necessary, a Cherenkov counter C (with a refractive index n = 1.033) were used to generate the trigger. Scintillation hodoscopes HOX, HOY, HOU, H0V were used to control the beam profile in F6. The characteristics of the counters are given in Table 1. The first setting of the experiment, due to the presence of six deflecting magnets, made it possible to have a negligibly small (less than 10–4) background/signal ratio for time-of-flight spectra even on positively charged particles. The suppression of protons (by two orders of magnitude) in the trigger using a Cherenkov counter was used to reduce the dead time. The inconvenience of such a setting is associated with the need to reconfigure a large number of magnetic elements. Therefore, experimental data in the first setting were collected at a fixed pion momentum of 4 V (3.0 GeV/c), the increase in the subthreshold degree of which was achieved by reducing the deuteron momentum. In the second series of measurements, carried out in June-July 1997, data were collected in a slightly different configuration of the setup with the target located at the F5 focus (hereinafter referred to as the "second setup"), as shown in Fig. 9. In such a formulation, the loads of head counters increase, especially in measurements on positive particles. To reduce the influence of such loadings, an NT scintillation hodoscope was used in the head part, which consisted of eight plastic scintillators viewed from both sides of the FEU-87. Signals from this hodoscope were used for time-of-flight analysis (based on 30 m), which in this case was carried out for each element independently. The position and profile of the beam (ax 4 mm, ty = 9 mm) on the target were monitored by a wire chamber, the intensity - by a 1C ionization chamber and M and Mg scintillation telescopes. The measurements of the second series were carried out with a hydrogen target (7 g/cm2), a beryllium target (36 g/cm2) in the form of a parallelepiped with a minimum transverse (relative to the beam) size of 8x8 cm2 and a carbon target (55 g/cm2) of a cylindrical shape with a diameter of 10 cm. are shown in table 3.

Configurable data and hardware views

The recommended way to write a working module is that reads and writes are performed as buffered input and output operations on the standard input and output streams of a blocking process; the SIGPIPE signal and the EOF state cause the process to terminate normally. The working module can be implemented both dependent and independent of the composition of the collected data (i.e., the content of the packet bodies) and the serviced equipment (hereinafter referred to as "session-dependent" and "session-independent"4, respectively).

The control module is a process that does not work with a stream of data packets and is intended, as a rule, to control some element (s) of the qdpb system. The implementation of such a module, therefore, does not depend on the contents of the packet stream, nor on the contents of the packet bodies, which ensures its universality (session independence).

In addition, processes that receive source data not through packet flows are also classified here, for example, modules for representing (visualizing) processed data in the current implementation of the SPHERE DAQ system, see paragraph II.5. Such a control module may be implemented in either a session-independent or a session-dependent manner.

A service module is a process that organizes packet flows and does not make changes to them. It can read from the packet stream and/or write to the packet stream, while the contents of the input and output streams of the service module are identical. The implementation of the service module does not depend on the contents of the packet stream, nor on the contents of the packet bodies, which ensures its universality.

A branch point is a start and/or end point for multiple packet streams and is intended to create multiple identical output packet streams from several different input packet streams (generated by different sources). The branch point does not change the content of the packages. The implementation of the branch point is independent of the contents of the packet streams, which makes it universal. The order of the packets from the various input streams in the output stream is arbitrary, but the order of the packets of each of the input streams is preserved: The branch point also implements a packet buffer and provides a means of managing it. It is recommended to implement a branch point as part of the OS kernel (in the form of a loadable module or driver) that provides the appropriate system call (calls) for managing its own state, issuing this state outside, managing the packet buffer, registering input and output streams working with it. Depending on the internal state, the branch point syscall receives (blocks receiving, receives and ignores) packets from any input stream and syscall sends (blocks sending) all(x) received packet(s) to output streams.

The event stitcher5 is a variant of the branch point, also designed to create several identical output packets from several different (from different sources) input streams of packets. The event stitcher modifies the contents of the packets in the following way: the header of each of the output packets is obtained by making a new packet header, and the body is obtained by sequentially connecting the bodies of one or more (one from each registered input stream - the so-called input channel) so-called. input packets "corresponding" to it. In the current implementation, in order to match input and output packets, the following is required: - match of types (header.type) of input and output packets declared for each input channel when it is registered, and - match of numbers (header.num) of input packets for candidates for matching in all input channels. The term "event stitcher" was introduced because it more accurately characterizes the proposed (rather simple) functionality, in contrast to rather complex systems called "event builder". Packets with types that do not have a declared match are discarded when they enter the input channels. Packets with numbers that do not match in all input channels are discarded. The implementation of the event stitcher is independent of the contents of the packets. It is recommended to implement the event stitcher as part of the OS kernel (in the form of a loadable module or driver) that provides the appropriate system call (calls) for managing its own state, issuing this state outside, and registering input and output streams working with it. The supervisor is a control (or working, if control packages are implemented) module that at least starts, stops and controls the qdpb system at the commands of the system user (hereinafter referred to as the "operator"). The correspondence of the supervisor's actions to the operator's commands is described in the configuration file of the first sv.conf(S). In the current implementation, the configuration file is a makefile. The elements of the qdpb system are managed through the mechanisms provided by those elements. The managed elements of the qdpb system are: elements of the OS kernel (loadable modules of the hardware maintenance subsystem, branch point(s), event stitcher(s); working modules. Management of other elements of the qdpb system is not provided, as well as the reaction to situations in the system. For remote control, i.e. managing elements of the qdpb system on computers other than the supervisor running the process (hereinafter referred to as "remote computers"), the supervisor launches control modules on them using standard OS tools - rsh(l) / ssh(l), rcmd(3) win rpc(3 ). For the operator's dialogue with the supervisor, the latter can implement an interactive graphical user interface (Graphics User Interface, hereinafter referred to as "GUI") or an interactive command line interface. Some elements of the qdpb system that have their own GUI can be controlled directly by the operator, without the participation of a supervisor (for example, data presentation modules). The above project was largely implemented. Let's consider in more detail the key points of implementation.

Polarimeter Data Acquisition Systems

By default, the sphereconf utility configures the specified loadable module module to work with the "kkO" CAMAC hardware driver. No specific information is passed to the loadable module. When specified on the command line, the sphereconf utility tests the configuration of the specified module load module and prints it to the error output stream. The default behavior of the sphereconf utility is changed by the above command line switches. The sphereconf utility returns code zero on success and positive otherwise. The sphereoper(8) control utility for the CAMAC interrupt handler is called sphereoper and has the following command interface: sphereoper [-v] [-b # ] startstop)statusinitfinishqueclJcntcl line, in the loadable module attached to the 0th branch of CAMAC, and outputs the execution result to the error output stream. Thus, the sphereoper utility can be used to implement some of the actions described in the supervisor's sv.conf(5) configuration file. The default behavior of the sphereoper utility is changed by the above command line switches. The sphereoper utility returns code zero on success and positive otherwise. To measure the speed of execution of CAMAC commands, a custom CAMAC speedtest interrupt handler was also implemented (for more details on testing the DAQ SPHERE system on the bench, see below), which, for each processed interrupt from CAMAC, executes the configured number of times the tested CAMAC command (selected by changing the source file speedtest.c ). The speedtest load module is configured by the stconf(8) utility and controlled by the sphereoper(8) utility (only the start, stop, status, and cntcl values ​​of the first positional argument are supported).

Compared to the sphereconf (8) utility, the stconf(8) configuration utility has an additional optional command line switch -p # for passing specific information to the loadable module, which means the number of repetitions of the tested CAMAC command, which is 10 by default, otherwise similar to the last one.

The SPHERE DAQ system uses (in a non-distributed, i.e., executable entirely on one computer, configuration) at least the working module writer(1), the service module bpget(l) and (optionally) control modules - the supervisor sv(l) and the module a graphical representation of the alarm(1) system log from the session-independent set of plug-ins provided by the qdpb system. Next, consider the software modules specific to the DAQ SPHERE system.

The statistics collector in the current implementation is called statman and is, in terms of the qdpb system, a working module, a packet stream consumer that accumulates data in shared memory in a form convenient for use by data presentation software modules (see below), and has the following command interface: statman [- o] [-b bpemstat [-e] ] [-c(- runcffile )]. [-s(- cellcffile )J [-k(- knobjcffile )] [-i(- cleancffile )] [-p(- pidfile )]

By default, the statman module reads packets from the standard input stream, collects information from the packet.data body of each incoming packet, and accumulates it in shared memory in accordance with the default configuration files. On startup, the statistics collector reads configuration files in the RVN.conf(5), cell.conf(5), knobj.conf(5) and clean.conf(5) formats (see paragraph P.3) and accordingly initializes the internal arrays of structures pdat, cell, knvar, knfun, knobj; runs a creation cycle over all initialized known objects and generates the PR0G_BEG event, after which it reads packets from the standard input stream and for each received packet increases the global counter corresponding to its type of event and performs a cycle of calculating the results for all initialized cells and a filling/clearing cycle for all initialized known objects. Upon receiving an EOF end-of-file condition on standard input or a SIGTERM signal, it generates a PR0G_END event, so a SIGKILL crash is not recommended. The PR0G_BEGIN and PR0G_END events are also used to calculate the results for all initialized cells and the fill/clear cycle for all initialized known objects.

The default behavior of the statman module is changed by the above command line options.

The statman module returns code zero on success and positive otherwise.

The statman module ignores the SIGQUIT signal. The SIGHUP signal is used to reconfigure an already running statman module by rereading the configuration files runcffile , cellcffile , and knobjcffile (however, with the same names as when the module was started), which leads to a complete clearing of all information accumulated at the moment and resetting the results of all computations. cells, i.e. completely equivalent to configuring at startup. The SIGINT signal results in a new reading of the cellcf file configuration file (with the same name as at startup) without resetting the cell results, which can be used to "reprogram" them on the fly. The SIGUSR1 signal clears all accumulated information, including internal global event counters, the SIGUSR2 signal clears accumulated information according to the configuration file cleancffile . Both of these signals also reset the results of all calculation cells. The SIGTERM signal must be used to send a request for a graceful termination to the module.

The configuration file of known objects of the statman module can only contain declarations of types supported by the module, currently the following: "hist", "hist2", "cnt", "coord" and "coord2" (see section II.3 for details). For each line of data in such a file, the first (name), third (type), fifth (fill event), sixth (fill condition), and seventh (fill event) fields have their standard knobj.conf(5) format value. The fields representing the arguments of the create (second), fill (fourth), clear (eighth), and destroy (ninth) functions must conform to the API of the respective families of known functions.

Analysis of sources of systematic errors

The textual data representation module is intended for textual visualization of information accumulated in shared memory by the statistics collector, it is called cntview and has the following command interface: cntview [-k(-I knobjconffile )] [-p(- pidfile )] [ sleeptime.

By default, the cntview module reads the data accumulated in shared memory by the statistics collector statman(l), interprets it according to the default configuration file in the knobj.conf(5) format, and prints its text (ASCII) representation to the error output stream.

The default behavior of the cntview module is changed by the above command line switches. The cntview module returns code zero on success and positive otherwise. The cntview module ignores the SIGQUIT signal. The SIGHUP signal is used to reconfigure an already running cntview module by rereading the configuration file (but with the same name as when the module was started). The SIGUSR1 signal suspends and the SIGUSR2 signal resumes reading information from shared memory and displaying it. The SIGINT signal redirects the next data output to the printer with the compiled name via the Ipr(1) utility. The SIGTERM signal must be used to send a request for a normal termination to a module. The cntview module's known object configuration file can only contain declarations of the "dent" type supported by the module (see Section II.3 for details). For the known object "dent", the first (name), third (type), fifth (fill event), sixth (fill condition) and seventh (fill event) fields of the data string have their standard value for the knobj.conf(S) format, then as the fields representing the arguments of the create (second), fill (fourth), clear (eighth) and destroy (ninth) functions, must conform to the API of the corresponding family of known functions. For example, the declaration of one known object of type "dent" is written as follows: Obj0041 41;shmid;semid dent 41;3;semid;type_ULong;nht,type_String;4;cnt21:cnt22:cnt23 \ DATA_DAT_0 - NEVERMORE gen prescfg(l) utility (see paragraph II.3) generates the declaration of the known object "dent" above from the prototype of the following form: dent 41 1 -1 shmid semid 3 ULong nht 4 cnt%2lN DAT_0 - N The OS kernel load module control utility is called watcher and has the following command interface: watcher [-b # ] [-p(- pidfile )] [ sleeptime ] By default, the watcher utility collects status information at intervals of 60 seconds (by calling oper() with the HANDGETSTAT subfunction) from the KA user interrupt handler -MAK, attached to the 0th branch of CAMAC, analyzes the state of the latter, taking into account previously received similar information, and issues error messages to the error output stream. Thus, the watcher utility can be used in conjunction with the alarm(1) syslog graphical module to report certain errors in the SPHERE DAQ system. The default behavior of the watcher utility is changed by the above command line switches. The watcher utility returns code zero on success and positive otherwise. The watcher utility ignores the SIGHUP, SIGINT, and SIGQUTT signals. The SIGUSR1 signal suspends and the SIGUSR2 signal resumes information collection. The SIGTERM signal must be used to send a request for a graceful termination to the module. The supervisor sv(l) described in paragraph II.2 can be used to control the SPHERE DAQ system. It is also possible to directly, without the help of the supervisor, execute the make (1) utility with the same name to the commands of the target operator (target) from the configuration file of the supervisor sv.conf. Let's describe the purpose of the operator's main commands: load - loading and configuring loadable modules of the OS kernel - branchpoint (4) and the custom CAMAC sphere (4) interrupt handler, launching the bpget(l) service module and attaching it (in the BPRUN state) to the branch point , initialization of CAMAC equipment. unload (command inverse to load) - deinitialization of the CAMAC hardware, termination of the bpget(l) module, unloading of the branch point and CAMAC custom interrupt handler, loadw - launch of the working module writer (1) with a request to enter the necessary parameters and a reminder of the possibility of entering optional ones and attaching it (in the BPSTOP state) to the branch point. unloadw (reverse to loadw command) - end of the writer module (1). loads - Runs a statman(l) worker and attaches it (in BPSTOP state) to a branch point. unloads (reverse to loads command) - completion of the statman (1) module. loadh - launches the histview (1) graphical data representation module using the xterm(l) utility in a separate window of the XII graphic system. unloadh (reverse to loadh command) - end histview module (1). loadc - launches the cntview (1) textual data representation module using the xterm(l) utility in a separate window of the XII graphic system. unloadc (inverse to loadc command) - end of the cntview (1) module. start_all - Change the state of all attachments to the branch point to BPRUN. stop_all (reverse to start_all command) - change the state of all attachments to the branch point to BPSTOP. init - initialization of the CAMAC equipment (it is necessary to execute it, for example, after switching on the power supply of the crates being read, it is also included in the load). finish (reverse to init command) - deinitialization of CAMAC equipment (should be performed, for example, before turning off the power, also included in unload). continue - start processing CAMAC interrupts and start the watcher utility. pause (reverse to continue command) - the end of the watcher utility and the termination of CAMAC interrupt processing. cleanall - cleanup of all information accumulated in shared memory by the statman module (1). clean - cleanup of information accumulated in shared memory by the statman (1) module, in accordance with the configuration file specified when the module was launched in the clean.conf(5) format. pauseh (reverse to conth command) - pause the rendering of data by the histview module (1). pausec (inverse to contc command) - suspending data rendering by the cntview (1) module. conth - continuation of data visualization by the histview module (1). contc - continuation of data visualization by the cntview module (1). status - outputs a summary of the status of the loaded elements of the DAQ SPHERE system to the log files of the syslogd(8) daemon. seelog - start viewing messages from the DAQ SPHERE system entering the log files of the syslogd(8) daemon using the tail(l) utility. confs - pause data visualization by histview (1) and cntview (1) modules, reconfigure statman (1), histview (1) and cntview (1) modules, continue data visualization (used after changing the corresponding configuration files). The DAQ SPHERE system currently uses the following freely distributed third-party software packages (in addition to those "inherited" from the qdpb system): satas package - implementation of the CAMAC service subsystem. ROOT package - used as a histogram graphical visualization API to implement the histview (1) data view module.

Golyshkov, Vladimir Alekseevich

1972

/

June

Modern state of physics and technology for obtaining beams of polarized particles

Contents: Introduction. Spin state of the particle. Principles of obtaining polarized ions. Atomic beam method. Dissociation of hydrogen molecules. Formation of a free atomic beam. Hydrogen and deuterium atoms in a magnetic field. Separating magnet. RF transitions. RF transitions in a weak field. RF transitions in a strong field. Operating installations. Ionization of an atomic beam. Ionizer with a weak magnetic field. Ionizer with a strong magnetic field. Obtaining negative ions by recharging positive polarized ions. Ionization by heavy particles. Lamb method. Energy levels of hydrogen and deuterium atoms with n= 2 in a uniform magnetic field. Times of life. Polarization in the metastable state. recharge processes. Getting negative ions. Getting positive ions. Methods for increasing beam polarization. Source of negative polarized ions. Measurement of ion polarization. fast ions. slow ions. Sources of polarized helium-3 and lithium ions. Polarized singly charged helium-3 ions. Sources of polarized lithium ions. Magnetized single crystal as a polarization donor. Injection of polarized ions into the accelerator. Cockcroft-Walton accelerator and linear accelerator. Van de Graaff accelerator. Tandem accelerator. Cyclotron. Accumulation of polarized ions. Acceleration of polarized ions. Cyclotron. Synchrocyclotron. Phasotron with spatial variation of the magnetic field. Synchrotron. Achievements of individual laboratories. Berkeley, California. Los Alamos. Conclusion. Cited Literature.