Home

Construction of NMR equipment to be used in the - Physik

1. fidd T Figure 6 3 La field scans from 9 T to 8T v 51 2 MHz 6 2For the 8 7 T to 8 3 T measurement sw 300 was used instead of 1000 6 2 SUMMARY OF NMR EXPERIMENTS 59 Discussion All three field scans are shown together in figure 6 3 The figure focuses on the peaks The tales of the flat peak B and the noise field range below 8T is not shown In the figure the scans from the three temperatures are scaled to match each other using the integrated intensities of the spectra from 8 to 9T If we compare with the frequency scan data we observe the large narrow peak A and the broad small peak in the measurements at 300K and 40K In the measurement at 4K peak A appears to be broadened and its intensity is decreased while the shape of peak B seems to be unaffected Comparing the spectra at 4K and 40K a slight shift of the room temperature peak A to lower magnetic fields can be observed In addition a downward structure at the tran sition from peak A to peak B on both sides of peak A at T 300 K could indicate satellites The A s of the three measurements calculated with the same fit function 6 1 as be fore are listed in table6 2 in section6 3 The calculated A data show a slight increase from 300K to 40 K and a doubling of the width from 40K to 4K for peak A The width of peak B seems to slightly increase with temperature 6 2 3 13 field scan at 30 2 MHz For comparison field scans at 30 20015 MHz w
2. Ix_max find ppms_data ppms_temp str2num ppms data Ix 2 1 1x 3 1 field value str2num ppms 1 3 1 1 4 1 73 cernox_R str2num ppms data Ix 4 1 Ix max 1 t2seqi4 scratch sweeps tau RR load scratch signal pp pcdif 1c9410_messung echoint echoeval signal bc window i left i right echointens echointens field value echoint temperature ppms temperature ppms temp cernox Resist cernox Resist cernox R signal 1 2 field value if f f begin mes dat signal else mes dat append mes_dat signal end eval save name echointens ppms temperature cernox Resist mes_dat figure fig_h plot echointens 1 abs echointens 2 o elTim toc disp elapsed time in min num2str elTim 60 end 74 APPENDIX A CODE OF FLDSCANI FUNCTION SCRIPT Appendix B Calibration of the CERNOX Temperature Sensor For the 4 terminal resistivity measurement the sample is mounted on the resistivity sample puck and connected to the ports This sample puck is conceived to measure 3 samples at the same time With a special tool the puck is inserted into the PPMS Through the user bridge the different channels are read out The temperature dependant resistivity is now easily measured by creating an automated sequence B IThe data is deposited in the PPMS computer folder C QdPpms Data alex cernoxCal 75 APPENDIX B CALIB
3. 15 us The FWHM A of the measured La resonance peak in LaBg was 16 7 kHz To have an overview at the spin spectrum in LaBaNiO 5 a frequency scan was done manually at a magnetic field of 8 5 T9 The following parameters where used BUT TIK 8 torso us ms r us Attr dB AttrldBl sw C 8 5 300 2 3 5 100 400 9 22 500 4 Discussion The thus obtained data are plotted in figure 6 2 One can see a sharp peak with its maximum at 51 25 MHz together with a broad smaller peak To distinguish this two peaks we will define the narrow peak as peak A and the broad peak as peakB see figure 6 2 upper left panel 6 1 The resonance circuit was matched for each frequency 6 2 SUMMARY OF NMR EXPERIMENTS 57 irtesity abu irtesty abu frequency MH Figure 6 2 Room temperature La frequency scan at 8 5 T A narrow distinct peak A can be seen with an underlying flat peak B Upper left panel Fit function 6 1 for Peaks A B and A 4 B Following the theory of first order quadrupole interaction in section 2 4 1 6 satellites and a central line peak A are expected see figure2 3 on page11 There are no clear satellites visible in the data Imperfections in the crystal grains presumably related to the oxygen deficiency in the structure may be responsible for a strong broadening of the satellites producing one broad line like peak B section 2 4 The possibility that peak A also con tains some satell
4. gt The results of vg for the field and frequency scans are also listed in table 6 2 By comparing the vo s of the frequency scan 300K and the field scans at 300 K and 40K at 51 2 MHz and 30 2 MHz we find a good agreement of the values which is a strong indication of having second order quadrupole effects The corresponding mean values are VQ 40 K 300 K 3 4 3 MHz and vg 4 K 4 7 7 MHz We note here that the Lorentz fit approach equation 6 1 does not exactly match the the oretical expectations Taking peak B as 6 broadened satellites a reduced intensity towards the center position would be expected which is not the case in the Lorentz fit function see figure6 2 on the upper left panel Nevertheless equation 6 1 still gives an accurate result for calculating the FWHM of the peaks An additional indication for second order quadrupole effects 1 Bo dependence is the asymmetry of the field scan data in comparison with the symmetrical Lorentz fit that is 66 CHAPTER 6 LABANIO 5 INVESTIGATION frequency field T K 6 kHz As kHz vo MHz 8 5T 300 419 80 260 50 4869 900 2933 600 3 1 6 512MHZ 300 599 50 361 30 8222 800 4952 500 3 6 3 51 2 MHZ 40 660 50 397 30 6720 600 4048 400 3 8 3 51 2 MHZ 4 1123 350 676 200 5883 900 3544 600 4 909 30 2MHZ 300 919 90 554 60 7573 700 4562 500 3 5 4 30 2 MHZ 40 750 70 452 40 7196 700 4335 500 3 1 3 30 2 MHZ 4 1446 15
5. 5 2 70 MHZ SPECTROMETER 51 5 2 3 Summary The 70 MHz spectrometer has shown a good performance Nevertheless we have noticed several disadvantages e g the overload problem that could be relevant for measurements of materials with a fast FID We have also to conclude that the spectrometer amplification is too weak but this sit uation could be improved by using an additional amplifier Mini Circuits ZFL 500 after the dual amplifier Wenzel Associates LNFDA 70 46 10 see section4 2 1 and ap pendix C 2 in the receiver path Another limitation of the spectrometer is that it has no phase modulation possibilities as the 4 phase modulators in the Lupotto series that is used in the t2seq14 a Matlab function described in section 3 2 to clean the echos However this limitation can be eliminated by using a fast phase switchable frequency generator 52 CHAPTER 5 TESTING PROCEDURE Chapter 6 LaBaNiO _5 Investigation LaBaNiO4 s is considered to be a promising candidate to represent a parent compound of Ni based superconductors It is believed that like it is the case in LaaCuO spin J 1 2 antiferromagnetism and Mott Hubbard insulation are important ingredients for the occur rence of superconductivity on subsequent doping of the compound with charge carriers 9 The main questions to solve for this project are a Is LaBaNiO really an insulator at T 0 K b Is Ni in a low spin s 1 2 state and if so has it an
6. La spectrum Both scenarios can create in extreme cases even a possible wipe out phenomenon analo gous to Ref 13 To conclude this discussion we state that it is difficult to interpret the NMR measure ments on LaBaNiO due to its oxygen deficiency We have observed a second order quadrupolar field dependence 1 Bo with a vg around 4 MHz and a decrease of the spin signal with decreasing temperature specially of peak A This might be attributed either to 6 3 DISCUSSION 67 8 intensity abu amp B 4 45 5 55 6 raydic fidd T Figure 6 9 Representative second order quadrupole asymmetry between the data points and the symmetrical Lorenz fit Data of the field scan at 4K 31 2 MHz a freezing out of the oxygen diffusion or and to a magnetic Ni spin ordering Nevertheless NMR can still be a very powerful tool to examine LaBaNiO 5 provided that we can improve the sample quality by minimizing the oxygen deficiency To ensure full oxygenation high oxygen pressure annealing techniques will have to be applied 68 CHAPTER 6 LABANIO 5 INVESTIGATION Chapter 7 Summary and Outlook For this diploma thesis a new NMR probe head was constructed to be used in the PPMS Quantum Design which offers the possibility to vary the magnetic field between OT and 9 T and to cool a sample down below 2K For the design of the NMR probe head the construction materials had to be chosen according to spacial therm
7. 1 cm i 109 3 em HEIGHT OF ACMS COIL 5 1 em CENTER OF MAGNETIC FIELD 1482 m I i SURFACE OF __ SAMPLE PUCK THERMOMETERS AND HEATERS IMPEDANCE HEATER AND THERMOMETER IMPEDANCE TUBE BELLOWS MAGNET BOTTOM OF SAMPLE TUBE WITH SAMPLE PUCK CONTACTS SECOND IMPEDANCE PROTECTIVE CAP z TO HELIUM ECTION Figure 3 5 Schematical overview of the PPMS insert head and sample space 4 3 4 COMBINING THE NMR APPARATUS WITH THE PPMS 25 3 4 Combining the NMR apparatus with the PPMS In this chapter we will discuss how the NMR was combined with the PPMS Because of the automated characteristics of the PPMS it would be plausible to use the MultiVu software for this purpose However there are several arguments against this option Although the PPMS software offers many possibilities for control and calibration purposes it is rather difficult to integrate external hardware and software The most common way would be to use LabVIEW To control the NMR apparatus however we would have to rewrite and adapt many parameters Moreover we would have to expect severe compatibility problems because the NMR pulse card is running in an old Windows 3 1 environment So we decided to drive the PPMS with the PC
8. 2 70MHZ NMR SPECTROMETER 37 4 2 70 MHz NMR spectrometer In this section the construction of the 70 MHz spectrometer is described The motivation of building a new spectrometer was that the previously used spectrometers mainly the ones of the Lupotto series are quite old and very complex Moreover RF elements are nowadays commercially available with good performance and at quite low costs We copied the design of a similar spectrometer which is used at the Solid State Physics Laboratory ETHZ 8 and adapted it to our needs Pictures of the 70 MHz spectrometer and its power supply are shown in figure 4 8 4 2 1 Schematical description The following description of the operating mode of the spectrometer refers to figure 4 6 The list of used elements can be found in appendix 2 For a comparison with the Lupotto IV spectrometer we refer to section 3 2 1 The main difference between these two spectrom eters besides their internal working frequency is that the new 70 spectrometer has cannot handle RF phases and is less powerful in amplification In standard operating mode i e no pulse is to be sent the spectrometer is in receive mode Once a pulse is sent it switches to transmit mode How the TTL signal used to switch between this modes is generated is explained in the subsection below The transmitter signal is a mix of two incoming signals The first signal from the fre quency generator s 10 MHz reference is multiplied by
9. Metallic Schifts in NMR Part I Pergamon Press USA Quantum Design Physical Property Measurement Manual Hardware Manual USA 2000 Quantum Design Physical Property Measurement Manual Software Manual USA 2000 Quantum Design www qdusa com USA Meinke F W Grundlach Taschenbuch der Hochfrequenztechnik Springer Verlag Germany 1956 M Weller http cryo nmr ethz ch mweller Spectrometer Solid State Physics Laboratory ETHZ Switzerland A Schilling Synthesis and characterization of LaBaNiO4 MaNEP Year5 Report Project 4 University of Ziirich 2006 G Demazeau et al Mat Res Bull 17 37 1982 J A Alonso et al Solid State Commun 76 1327 1990 R J Cava et al Phys Rev B 42 1229 1991 1 M Abu Shiekah et al Phys Rev Lett 83 3309 1999 I M Abu Shiekah et al Phys Rev Lett 87 237201 2001 A Mehta P J Heaney Phys Ref B 49 563 1994 M Bankay Kernresonanz an Y Ba Cu O Supraleitern Dissertation Universitat Z rich 1995 81
10. anti ferromagnetic ordering at a finite temperature The contribution of this diploma thesis to these LaBaNiO 5 investigations was to perform a NMR study on this compound Previous investigations done on this polycrystalline com pound are rare As a starting guideline for this research we used the results of Refs 9 10 and 11 6 1 Sample characterization The compound LaBaNiO is a structural analogue to the compounds La2NiO LaSrNiO and the well known cuprate LagCuO 11 10 and 12 Stoichiometric LaBaNiO4_5 be longs to the tetragonal space group I4 mmm No 139 with cell parameters a 3 8693 Aand 12 8803 A Its structure and the X ray Guinier spectra are shown in figure 6 1 LaBaNiO has a molar weight of 399 0085 g mol It has been reported that LaBaNiO4_5 is semiconducting 10 and it has been claimed that it shows a crossover form a high spin to a low spin Ni state upon cooling down below a temperature of 100 K 53 54 Intensity CHAPTER 6 LABANIO 5 INVESTIGATION LaBaNio 99 Gee Structure 14 mmm er a 3 8693 APIGA c 12 8803 Q O basal A b ya 40 50 60 m 2Theta Figure 6 1 Top 14 mmm structure of LaBaNiO In between two basal NiO planes are two apical La BaO layers Bottom Calculated X ray Guinier spectra in agree ment with the X ray film pattern The presence of impurities above 5 sample volume can be excluded 6 1 SAMPLE CHARACT
11. capacitor in series Parallel to both is the matching capacitor Cu The total serial impedance is 1 serial DEE EN W T serial In order to achieve the resonance frequency w wp 229 has to be minimal 1 1 uU RL 0 gt 5 4 2 ENG WR TO 4 2 In resonance 22 e Ro 10 While the resonance frequency is manipulated by the tuning capacitor the parallel matching capacitor Cu adjusts the general impedance to 502 Cm is a commercial capac itor with a range of 92 pF up to 330 pF Useful additional experiment dependent options are an additional capacitor Cu parallel to to achieve the 50 impedance and a resis tor R in between the coil and C7 to get a lower Q factor and broaden the frequency range of the resonance cirquit 32 CHAPTER 4 CONSTRUCTION PART The tuning capacitor is a coaxial capacitor shown in figure 4 4 It consists of a Berilco tube soldered to the ground a rounded brass pin in the inside connected to the coil and a dielectric tube made out of quartz glass that can be moved in and out of the capacitor The range of capacitance can be calculated by using the formula for a coaxial capacitance l In R2 R er 2760 with the permittivity constant 8 85 10712 Cb Vm the dielectric constant e the length of the capacitor 10cm the outer wall of the brass Ry 5 4mm and the inner wall of the Berilco tube R 10mm The res
12. different pulse schemes were developed Generally used simple methods are the single pulse FID analysis for spectral investigations the inversion pulse and pulse saturation method for spin lattice relaxation measurements and the spin echo method for spin spin relaxation studies or spectral analysis In this diploma thesis mainly the spin echo method was used Spin Echo method The spin echo method makes use of the refocusing of the spins A 90 pulse produces a rotating magnetization in the xy plane which loses the coherence of its magnetic moments with time After a spacing time r a 180 pulse is applied producing an inversion of the spins The spins refocus after another time ie roughly 27 after the 90 pulse In solid state materials the FID usually decays very fast parts of a ms in some materials the decay of the signal is even so fast that technical factors impede a proper detection In this case the spin echo method has to be applied It is clear that 7 has to be shorter than in order to assure a transverse magnetization component perpendicular to the external static field 2 4 FREQUENCY BROADENING 9 2 4 Frequency broadening Besides the Zeeman effect there is the quadrupole interaction for J gt 1 2 The quadrupole interaction is described by the third term of the multi pole Taylor expansion of the elec trostatic energy potential Details can be found in Refs 1 and 2 We write down the Hamiltonian consisting of the
13. pulse times were set accordingly The following parameters were used vs MHz thoy 15 toss 5 trep 7 ys Atty AB Atte dB sw CH 51 20157 1 9 3 7 100 30 9 38 1000 4 6 2 SUMMARY OF NMR EXPERIMENTS 63 2000 intensity arb u magnetic field T 1500 a 1000 intens x temp arb u 0 100 200 300 temperature K Figure 6 7 13 temperature scan at 51 2 MHz Upper right panel Measurement at 30 2 MHz and 40K to demonstrate the field positions of Ac 8 5 T 8 425 T and ABg 8 575 T 25 intesity 0 10 20 30 teperatue K Figure 6 8 Temperature dependent intensity ratio of 19 64 CHAPTER 6 LABANIO 5 INVESTIGATION Discussion In figure 6 7 the product of the measured intensities and the temperature I T is plotted vs temperature for all three field values Following the temperature de pendent Boltzmann distribution of the spins 9 see sections2 1 and 2 2 we would expect I T to remain constant as a function of temperature however we see a decrease of I T below 80 K having a minimum at 20 Below 20K has again an increasing tendency In figure 6 8 the ratio between and the mean value of AB and is shown as a function of temperature We can see a approximately flat ratio between 300 K and 140 K an approximately 20 lower ratio between 100 K and 20 K and a fast decrease below 20 K PT 6 3 DISCUSSION 65 6 3 Discussion Follo
14. space The probe head is shown in figure4 3 The probe head is mechanically reinforced by the 50 Q coaxial line and an additional tube The thinner two shafts manipulate the capacitors and are therefore freely movable Along the support five horizontal plate assist thermal insulation and support The upper four plates are made out of stainless steel the lowest one is made out of Berilco Below this plate there are no more stainless steel parts 43 Avoids electrical breakdown 44 quartzglass 3 7 and He 1 00007 45To reduce the thermal conductance this rod is actually a 2 x 1mm tube 4 1 NMR NQR PROBE HEAD 33 capacitor driving mechanism 50Q line Lemo connectors 50Q line Figure 4 2 Top of the NMR probe head The capacitors are located at the bottom part figure4 4 The matching capacitor is screwed to the Berilco plate which is soldered to the tuning capacitor In the sample space the inner coax line ends at the lower Berilco disk on which the three puck holding Berilco bars are attached to The brass pin of the tuning capacitor is screwed into the Torlon plate Three pins the coax line pin the tuning capacitor soldering pin and the ground pin stick out below the Torlon disk There are two open holes through the plates for the additional wires form the Lemo connectors Three thin Berilco bars hold the puck plate where the PPMS sample puck is attached to The sample plateau figure 4 5
15. the deuterium larmor frequency vz 22 915 MHz at a magnetic field of 3 5 T a capacitor 220pF parallel to the matching capacitor Cm was added and to broaden the resonance a resistance R 5 6 was placed between the coil and the tuning capacitor position 0 Figure 5 1 Positioning of the spherical glass container with D gt O in the sample coil holder The probe was installed as seen in figure5 1 with the sphere centered at 5 1 cm above the puck surface which corresponds to the center of the magnetic homogeneity region following the PPMS manual 4 For the measurement the signal was optimized to get a 90 pulse length of 19 us with a transmitter attenuation of 24dB and 180 pulse of 39 us The Fourier transformed signal of the FID is shown in figure 5 2 The Fourier transformed signal shows a narrow peak on the left side of the main peak and a plateau to the right with approximately 2 3 of the slim peak s intensity The full width at half maximum FWHM is about A 1 9 kHz which corresponds to a magnetic field inhomogeneity of Ag 2 7 x 107 T We obtain Ag B 0 0083 which compares reasonably with the 0 01 reported in the PPMS specifications for a cylindrical volume of 5 5 cm height with 1 cm diameter 4 To perform a measurement with a smaller symmetrical probe a smaller test tube was taken and filled with heavy water sealed with rubber inserted into a PVC tube that fitted into
16. voltage which is the response of the spinning nuclear magnetic moments In order to avoid reflections the electric impedance of the resonance circuit has to match the impedance of the amplifiers which have internal impedances of 502 First of all a coaxial 50 down to the circuit is crucial It is also the main sup port of the probe head We made the line of two stainless steal tubes with a lower part of brass The outer tube has diameter dimensions of 6 and 5 5mm and the inner tube which is kept centered with centering Teflon rings has diameter dimensions of 2 and 1 mm To calculate the characteristic wave impedance 21 0 in of the coaxial line we used a 4 1Torlon is a high performance amorphous polyamide imide polymer with exceptional mechanical and thermal properties 42Berilco is a non magnetic thermal stable copper beryllium alloy 4 1 NMR NQR PROBE HEAD 31 common RF technical formula 7 60 D D is the inner diameter of the outer tube and d the outer diameter of the inner tube By setting the dielectric constant for helium He 1 00007 we get a characteristic wave impedance of 21 0 60 7 Q which is reduced by the Teflon centering rings Teflon 2 close to our expected 50 impedance vir L R Figure 4 1 Schematical drawing of the resonance circuit Next we will discuss the resonance circuit shown in figure4 1 The circuit consists of the coil L with an Ohmic resistance Ro and the tuning
17. 0 871 80 6924 700 4171 500 4 4 5 Table 6 2 The A s and vg summary from frequency and field scans clearly visible in the example shown in figure 6 9 The same asymmetry is observed in all the field scans but not in the frequency scan at room temperature Even if second order quadrupole effects are present the possibility of magnetic effects can not be excluded With magnetic effects we would expect a constant or higher FWHM for higher fields Up to now the focus of our discussion was on the field dependence of the data We now want to briefly discuss the temperature dependence We see a clear loss of intensity below 80 K with an intensity minima at 20 K see figure 6 7 The ratio between the top of peak A and the crossover points to peak B average of AB and ABg also indicates a loss of intensity of the central peak A below 20 K see figures 6 3 6 4 and 6 8 Possible explanations for this increase in the signal width and the loss of intensity at low temperatures are e The possible oxygen diffusion in the lattice due to the oxygen deficiency in LaBaNiO 5 slows down or even freezes out at lower temperatures producing a broadening of the 139T a signal due to differing EFG s at different 13 lattice sites At higher temperatures this oxygen diffusion is too fast to be seen on NMR times scale faster than 1 A e A magnetic Ni spin ordering occurs at low temperatures that leads to a magnetic broadening of the
18. 0 MHz D signal is split in the quadrature detec tor into two by 90 phase shifted parts which can be interpreted as the real and imaginary part of a complex signal 0 MHz D This splitting is achieved through a mixing with two 20 MHz signals from the quadrature generator which differ by 90 in phase The quadrature generator receives the phase dependent 20 MHz signal from the 4 phase modulator This phase is also adjustable The resulting signal 0 MHz D is passed to the oscilloscope 3 4Connector OUT 3 5 Rz IN 22 CHAPTER 3 EXPERIMENTAL APPARATUS 3 3 PPMS A short description The physical properties measurement system PPMS a device manufactured by Quan tum Design is an open architecture variable temperature field system which allows auto mated measurements for physical properties such as heat capacity magnetization magnetic torque Hall effect DC resistivity and other thermal and electronic properties of a sam ple The magnetic field can be set up to 9T and the temperature covers a range from 1 8 400 K or down to 0 4K with a Helium 3 refrigerator The possibility of easily con trolling the field and the temperature makes the PPMS an interesting option for doing NMR measurements We restrict this PPMS description to the used features and specifi cations such as the field and temperature controlling the automation of the system and the sample space dimensions For all other features and spec
19. 1074 T as compared to a value 1 07 kHz 1 64 x 10 T in the oscillating mode indicating that the oscillating approach mode produces a slightly more homogeneous field 5 1 5 Summary The NMR probe head has shown good performance handling and stability in high magnetic fields and also at low temperatures The magnetic field homogeneity is in the range of 0 01 as specified by Quantum Design 4 The measurements of the position dependent magnetic field homogeneity show a slightly better homogeneity below the specified center of magnetic homogeneity region The faster linear approach mode shows a clear drift of the magnetic field while the more time consuming oscillating approach mode results in a stable magnetic field with a very small drift and a slightly better homogeneity 5 2 70 MHZ SPECTROMETER 49 5 2 70MHz spectrometer In the 70 MHz spectrometer tests an overload problem was discovered in the receiver path which is described in the next subsection 5 2 1 A signal to noise ratio comparison of the 70 MHz spectrometer with the Lupotto spec trometer running with 95 MHz was also made see subsection 5 2 2 For a description of the Lupotto II spectrometer which is an analogue to the Lupotto IV spectrometer we refer to section 3 2 1 To test the spectrometer we made nuclear quadrupole resonance NQR measurements on the CI isotope in natrium chlorate 1 powder diluted in paraffin oil a method to prevent any piezo e
20. 3A technique that will not be described any further It was used for hardware testing only and has no relevance for the present work 50 CHAPTER 5 TESTING PROCEDURE 0 03 0 02 0 01 noise with spectrometer overloaded 0 01 intensity arb u normal noise 0 02 20 30 40 50 60 70 80 time 10 s Figure 5 6 Spectrometer overload The noise shrinks to about a third of the nor mal noise level for 23 us The measurement was done with a 180 pulse and no attenuation and the MITEQ AU 1448 broadband preamplifier The signal to noise ratio was calculated by taking the real part of the root mean square RMS of the respective signal as shown in the following formula R RMS signal R RMS noise ratio 5 1 The ratios that where calculated from the FID are frequency 70MHz spectrometer Lupotto II spectrometer 29 934 kHz 7 3552 6 0938 29 937 kHz 7 3704 6 2433 The signal to noise ratio of the 70 MHz spectrometer is comparable slightly better to the one of LupottoII spectrometer although it must be pointed out that the Lupottoll spectrometer was adapted to have an equal FID amplitude of about 200 mV peak to peak as on the 70 MHz spectrometer To do this the amplification of the Lupotto II spectrometer had to be decreased to a minimum with a negative influence on its signal to noise ratio
21. 3U Figure 2 3 NMR quadrupole spectra for a Spin J 7 2 nuclei Six peaks can be seen The central peak 3 4 is represented by the dashed line v is the position of the central line and can be shifted compared to vr 3 An additional first order broadening occurs if there are imperfections in the lattice such as dislocations strains vacancies interstitial foreign atoms etc The EFG varying not only in orientation but also in magnitude from site to site has a considerable influence on the shape of the resonance lines In some cases the satellites will be completely wiped out by first order quadrupole broadening whereas the central line 3 will be practically unaffected A single resonance line will then be observed and the quadrupole broadening will manifest itself only through the loss in intensity 12 CHAPTER 2 THEORY 2 4 2 Second order quadrupole effects For axial symmetry the second order transition frequency for a single crystal is given by 1 Ve VL suq 3 1 m 1 2 su p VL 102m m 1 1871 1 39 u m m 1 21 1 1 3 2 28 v m m 1 x This induces a shift of the central line By defining Vi as the frequency of the central line which is also called the center of gravity the shift can be described in the case of a powder sample by taking the difference 3 MN MES 21 1 21 3 2 29 The shift of the central line to lower frequencies is prop
22. 4 Frequency broadening 2 4 1 First order quadrupole 2 4 2 Second order quadrupole 3 Experimental Apparatus 3 1 Schematical do NMR components i s s m G dg Bee oe X q ele EEE E 3 3 2 1 Lupotto VI spectrometer 3 3 PPMS A short 3 4 Combining the NMR apparatus with the PPMS 4 Construction Part 41 NMR NQR probe T Motivations ee yes ae 3 4 1 2 Requirements for the probe head 4 1 3 Electrical properties of the probe head and the resonance circuit Construction eu y sar erba es iae uH Ok de en e EU 4 1 5 Temperature 42 70 MHz NMR spectrometer ADA Schematical 4 2 2 Power supply and the TTL control circuit 4 2 3 Comments on the construction of the 70 MHz Spectrometer 5 Testing Procedure 5 1 NMR probe head and magnetic field homogeneity of the PPMS 5 1 1 Cool down behavior of the probe head NOOO OA OW Ww m eR on N 29 29 29 30 30 32 36 37 37 39 Al I a gt 5 1 2 Magnetic field homogeneity of the 5 5 1 8 P
23. 40 2 MHz and 40K To study the barium nuclei a calibration measurement was performed with BaCls The low natural abundance of barium isotopes Ba 11 32 and 6 59 made the search for a spin signal very difficult At a field of 8 5 T and a frequency of 40 308 MHz 1 Ba spin response was found The echo optimization gave a 90 pulse length of 8 5 us and a 180 pulse length of 17 us having a transmitter attenuation of 9 dB A field scan from 9 T to 7 T with 0 02 T steps was done at a temperature of 40K The set parameters are given in the following table Vs MHz to torso Hs trep ms r us Attr dB Atta dB sw C 40 2081 40 4 8 50 30 9 18 1000 3 Discussion The measured data are presented in figure6 5 As it can be seen the signal to noise ratio is low making the interpretation of the data very difficult The FWHM Az 0 5 1 T A 2 3 5 MHz of the peak was evaluated from the graph The increasing intensity for field values below 8 T stems form the high field tail at the 13 spin signal 6 2 SUMMARY OF NMR EXPERIMENTS 61 irtesity abu 75 8 85 9 naysiic fidd T Figure 6 5 Ba field scan from 9 T down to v 40 2 MHz at 40 K 6 2 5 Ba field scan at 33 8 MHz and 40K The field scan for Ba was performed at a frequency of 33 8 MHz and a temperature of 40 K The field was scanned from 8 7 T to 7 3 T in 0 1 T steps with the following parameters vs MHz lus tor
24. 7 in the multiplier 1 cleaned by the band pass 2 and switched 3 with a logical TTL 0 to the transmitter path In mixer 5 this 70 MHz frequency is mixed with the second signal that is generated with the frequency generator fed into the spectrometer through Lo IN and switched on the trans mitter path with switch 12 This second signal is the chosen spectrometer frequency v for the experiment plus the 70 MHz internal working frequency of the spectrometer The resulting signal from mixer 5 is at the chosen frequency vs This signal is amplified in 8 passed through by switch 9 attenuated 10 with up to 65 dB and sent through Tr OUT to the power amplifier Similar to the mixers used in the Lupotto IV spectrometer section 3 2 1 we obtain two output signals where the unwanted one is cleaned out with an external low pass filter 11 In the transmit mode all switches have a logical 0 which means that the switches 3 and 12 are on transmit switch 9 is feeding through and switch 21 is 50 Q termi nated This state is the one shown in figure4 6 For receive mode all switches have a logical 1 Now the switches 3 and 12 are switched to the receiver path switch 9 is 50 terminated and switch 21 feeds the signal through CHAPTER 4 CONSTRUCTION PART 38 TRANSMITTER Att 16 amp 18 15 RECEIVER Figure 4 6 Schematical description of the 70 MHz spectrometer Switches are on transmi
25. 916 MHz and using the same resonance circuit as with the previous 3 5 T measurements The sample plateau was attached at different heights for each measurement Our posi tioning scale was set as shown in figure 5 3 where position 0 corresponds to the center of magnetic homogeneity region at 5 1 cm above the puck surface 4 At this position several measurements were taken with and without Berilco screws holding the sample plateau yielding almost identical results The result of FWHM A vs position is presented in figure 5 4 FWHM kHz x 5 22 0 e 2 0 2 4 6 position cm Figure 5 4 Position dependent field spread We noticed a decrease of the FWHM when moving the sample down towards the puck The FWHM measurement at 6cm is about 10 narrower than the one at Ocm This could indicate a slight magnetic inhomogeneity produced by the probe head around the position of the tuning and matching capacitor 5 1 NMR PROBE HEAD AND MAGNETIC FIELD HOMOGENEITY OF THE PPMS47 5 1 4 Drift of the magnetic field The drift of the magnetic field of the PPMS was tested in two ways The first test was to check fluctuations of the magnetic field on a short time scale For this purpose echos were produced with different pulse spacing times 7 during which the echos stability was observed No fluctuations were detected With a spacing time of about 70 ms the echo started to become smaller to finally disappear at a spacing time of 100ms T
26. Construction of NMR equipment to be used in the Physical Properties Measurement System PPMS Quantum Design Alexander Gafner Diplomarbeit an der Universitat Zurich Mathematisch naturwissenschaftliche Fakultat Begutachtet von Prof Dr A Schilling Ausgef hrt am Physik Institut der Universitat Z rich unter der Leitung von Dr J Roos Juli 2006 Acknowledgments First I would like to thank Prof Dr Andreas Schilling for the opportunity of absolving this Diploma thesis in his group Special thanks go to Dr Joseph Roos for his advisory guiding support and for the numerous discussions through this work A special thank goes to the members of the institutes workshop for the help on the construction of the NMR probe head and to Peter Soland for the advisory help on the building of the 70 MHz spectrometer To the Ph D student Simon Str ssle friend and office mate I m thankful for the inspiring discussions Y sobre todo le agradezco a mi queridisima familia por su apoyo su paciencia y darme la oportunidad de estudiar esta carrera Tambi n a mis amigos as y coinquilinos as con que he convivido en todo este tiempo les agradezco Contents 1 Introduction 2 Theory 2L NMR S d uc nr ser a Aa a tu aem a 22 Hlochequatiolisas Cu aa ait ee thn EN THEE EL PU TEN 2 3 Free induction decay and measuring methods 2 32 Pulseschenie kta a era ele BAe ee aid etr 2
27. ERIZATION 59 The nuclear data relevant for NMR are listed in table 6 1 nucleus spin nat abundance 7 T7 s ia 12 99 91 6 0036 Da 3 2 11 32 4 7314 135Ba 3 2 6 59 4 2294 Table 6 1 Relevant nuclear data for NMR research on LaBaNiO s The preparation of the polycrystalline sample was done through a standard wet chemical procedure In a solution of distilled water 50 ml and nitric acid 10 ml the cor responding metal nitrates La NiO3 3 Ba NiO3 2 and Ni NiO3 2 were mixed The liquid was then warmed up and evaporated The remaining powder of nitrates was ground in a mortar and pre reacted at 900 C for 24 hours air The color of the powder has turned from light green to black at this stage This powder is ground in a mortar again pressed into a pellet and sintered at 1100 C for days in air The X ray diffraction done with a Guinier camera Cu Ka radiation showed no impurities above 5 sample volume see figure6 1 At the same time the I4 mmm structure was confirmed Neutron diffraction experiments chemical analysis electrical resistivity magnetic suscep tibility and specific heat measurements were done on the sample and are presented in Ref 9 The chemical analysis showed an oxygen deficiency of 6 0 15 1 A Torlon capsule was filled with LaBaNiO4 powder The cylindrical volume of the powder sample was 4mm in diameter at 1 5cm length with 845 5 mg of sample mass 56 CHAPTER 6 LABANIO 5 INVESTI
28. Frequency mixer Lowpass filter Attenuator Amplifier TTL switch 50 termination Variable attenuator Lowpass filter TTL switch I amp Q Attenuator Bandpass filter Frequency multiplier Amplifier Bandpass filter Frequency mixer High isolation TTL switch Variable attenuator Attenuator Attenuator Attenuator model number LNOM 10 7 10 17 SBP 70 ZYSWA 2 50DR VAT ZEM 2B SLP 300 VAT ZFL 500 ZYSWA 2 50DR ANNE 50TA 006 SMA F JFW BLP ZYSWA 2 50DR ZFMIQ 70D SBP 70 LNOM 10 7 10 17 ZFL 500LN SBP 70 ZEM 2B ZYSWA 2 50DR 50TA 007 SMA F JFW VAT VAT VAT Total power consumption for RF elements supplier Wenzel Ass Mini Circuits Mini Circuits Mini Circuits Mini Circuits Mini Circuits Mini Circuits Mini Circuits Mini Circuits Mini Circuits Emitec Mini Circuits Mini Circuits Mini Circuits Mini Circuits Wenzel Ass Mini Circuits Mini Circuits Mini Circuits Mini Circuits Emitec Mini Circuits Mini Circuits Mini Circuits 5V mA 20 20 20 60 120 5V mA 15V mA 410 20 80 20 20 410 60 60 120 960 SEZNAWATA A8 LSIT GO 62 80 APPENDIX C 70 MHZ SPECTROMETER CONNECTORS AND RF ELEMENTS Bibliography 10 11 12 13 14 15 16 A Abragam The Principles of Nuclear Magnetism Oxford University Press UK 1961 C P Slichter Principles of Magnetic Resonance Springer Verlag Germany 1990 G C Carter L H Bennet D J Kahan
29. GATION 6 2 Summary of NMR experiments This section is a resume of the experimental part on LaBaNiO A short discussion for each measurement is given while the general discussion combining these measurements is presented in the next section Because of better amplification and phase modulation possibilities the Lupotto IV spec trometer was chosen for these NMR experiments The spin echo method described in the theory section 2 3 1 was used To get a value for the intensity of the spin response at a given magnetic field value the absolute value of the integral over the echo was calculated For this purpose the fldscan1 Matlab function script described in section 3 4 was used The used parameters are col lected in tables listing coil setup C 90 pulse duration tp 180 pulse duration t so repetition time trep the spacing time between the pulses transmitter attenuation Attr the receiver attenuation Atta amount of sweeps sw set temperature T magnetic field value and the chosen spectrometer frequency vs 6 2 1 frequency scan at 8 5 T The calibration of the pulse lenghts of La was done with a powder of the metallic cubic compound LaBg To prevent electrical conduction between the powder grains paraffin oil was added and mixed with the powder Apart form the pulse lengths the parameters were the same as for the subsequent La measurement in LaBaNiO 5 The optimized reference pulses were tpo 7 5 us and tpg
30. Manual in Ref 5 The GPIB string command to set the field looks as follows FIELD Field Rate ApproachMode MagnetMode The Field has to be set in Oe the Rate defines the field setting speed in Oe s Then the optional parameters follow The ApproachMode is 0 for default Linear Approach mode 1 is for No Overshoot Approach mode and 2 for Oscillating Approach mode which is used by us because of the smaller drift of the magnetic field as will be described in section 5 1 4 The MagnetMode is set to 0 for Persistent Mode default also in all our experiments and 1 for Driven Mode Each string command for the Model 6000 Controller has to end with a semicolon Next a while loop had to be inserted to wait for the PPMS to set the field and turn into the Persistent Mode To do so the PPMS is asked periodically for the status This is done with the readgpib command containing the string command GETDAT DataFlag NoUpdateFlag In DataFlag the data strings to be returned have to be chosen This is done by setting a 32 bit integer which is chosen according to the list in appendix A of the PPMS GPIB Commands Manual 5 To read out the status we set 1 which gives us first the time stamp of the PPMS ending with a comma followed by the PPMS status and finally a semicolon The status is a four digit hexadecimal number The first digit is the tempera ture status the second the magnet status the third the chamber status and the last
31. RATION OF THE CERNOX TEMPERATURE SENSOR 76 50 100 150 200 250 teper we H Figure B 1 Calibration measurements for the CERNOX temperature sensor Appendix C 70 MHz Spectrometer Connectors and RF Elements The connectors pin lineup for the spectrometer and the power supply and the list of RF elements are presented in this appendix C 1 Connectors lineup The given switch numbers are the ones described in section 4 2 1 and the schematic seen in figure 4 6 in page38 The spectrometer has 2 ports the spectrometer TTL signal and the power lines The power supply is directly connected to this two ports and receives the TTL signal from the Pulse Controller PC TT 78 APPENDIX C 70 MHZ SPECTROMETER CONNECTORS AND RF ELEMENTS C 1 1 70 MHz spectrometer 131211 sie a DE 4 e 0000 e LAM NR e 25 242322 212019 7161514 5 1 Figure C 1 Figure C 2 Steer Bridge Port Power Connector o C 1 2 Power supply 1234567 8 9101112 13 542321 d 5 000000000000 2 e 14 15 16 17 18 19 20 21 22 23 24 25 5 A Figure C 3 Figure C 4 Figure C 5 Steer Bridge Port PC Steer Port Power Connector 6 S GND 9 GND v Dev I WNW m 11 12 14 15 16 amp 18 19 20 21 22 23 24 25 List of RF elements device type Frequency multiplier Bandpass filter TTL switch Attenuator
32. Zeeman term H and the quadrupolar term Ho eQV Hoa hl 31 P Sm I2 2 25 The first term contains also a magnetic shift K which arises from the magnetic coupling of the electrons with the nucleus This shift either originates form the motion of the electrons or from the magnetic moment associated with the electron spin The former gives rise to the so called chemical shifts the latter to the Knight shift in metals and to a coupling between nuclear spins In the quadrupolar term J J L is the raising and J_ I I the lowering operator eQ represents the quadrupole moment of the nucleus and Q the quadrupole tensor which describes the deviation of the charge distribution of the nucleus form the spherical symme try The EFG tensor V can be described by choosing the principal axis system with the convention V lt Vyy lt and using the asymmetry parameter 7 ulm zz The effect of the Zeeman and the quadrupole interaction for a spin J 3 2 nucleus is illustrated in figure 2 2 treating the quadrupole Hamiltonian as a perturbation the values of the energy levels can be computed analytically see Refs 1 and 3 2 4 1 First order quadrupole effect For axial 7 0 symmetry of the the transition frequency is given by v m m 1 ur 2 34 1 m 1 2 2 26 with cos 0 and 0 the angle between the z axis and the applied external magnetic fi
33. al magnetic and electric criteria The functionality of the probe head was tested in the PPMS giving us information about the behavior of the system upon cooling the magnetic stability and homogeneity The probe head showed good performance easy handling and mechanical stability at high magnetic fields and at low temperatures Moreover the NMR probe head can be used for different NMR and low temperature NQR experiments The magnetic homogeneity of the PPMS corresponds to the specifications given by the manufacturer A slight drift of the magnetic field was noticed using the oscillating approach mode to change fields but this was not of concern for the present investigation A 70 MHz spectrometer was built consisting out of industrial RF devices The spectrometer unit was successfully tested but showed fairly weak amplification and overload problems after a pulse The amplification could be increased by using an additional amplifier The overload problem was solved by attenuation in the transmitter path The spectrometer has no capability of handling phases as the spectrometers from the Lupotto series However phase modulations can be done by using a fast phase switchable frequency generator Finally an NMR investigation on the polycrystalline compound LaBaNiO 5 has been carried out Frequency magnetic field and temperature scans on the La nuclei and magnetic field scans on the Ba and Ba nuclei were performed Possibly due to a
34. certain oxygen deficiency in the sample the spectral lines are broadened which makes an interpretation difficult The measurements on showed the presence of second order quadrupole effects A broadening and decrease of intensity of the central peak at low temperatures is observed Slowing down or even freezing out of oxygen diffusion due to the oxygen deficiency and or 69 70 CHAPTER 7 SUMMARY AND OUTLOOK a magnetic Ni spin ordering may explain this broadening at low temperatures The spin signals of the Ba and the Ba nuclei are visible but not conclusive within the limited time range of this diploma thesis A reduction of the oxygen deficiency in LaBaNiO 5 by high oxygen pressure annealing techniques in order to obtain more conclusive NMR data is highly desirable Appendix A Code of fldscan1 Function Script In this appendix section the code of our fieldscan measurements is presented The expla nations for this code are given in section 3 4 function dummy fldscani name f begin f step f end query function lecory fldscani name f begin f step f end query name file name for saving of data f begin in Oersted f step in Oersted f end in Oersted Aquery 1 for param input by query 0 for param by direct editing see below output of LeCroy data in matrix lecroy and output of integrated intensity in matrix echointens 4 loop trigger E cycle phase correction if query 1 input of param
35. e elements For the spectrometer we made a baseplate out of copper to have a grounded base both for heat transport and electrical ground The electrical part done in the electronics laboratory consisted of ordering cables soldering constructing and adapting 50 semi rigid lines to connect the different RF elements In view of the high working frequencies of the spectrometer we had to assure that the setup was electrically well insulated and that there were no parasitic leaks The connectors lineup between the spectrometer and its power supply and between pulse controller and power supply respectively are listed in the appendixC In section 5 2 the spectrometer tests will be presented 42 Figure 4 8 CHAPTER 4 CONSTRUCTION PART place for TTL steering Top 70 MHz spectrometer Bottom Power supply for the spectrometer Chapter 5 Testing Procedure 5 1 NMR probe head and magnetic field homogeneity of the PPMS In this section the behavior of the probe head in the PPMS will be discussed After a first cool down performance check the magnetic field homogeneity was analyzed and drift properties of the PPMS were examined For this whole testing series we used our home made 70 MHz spectrometer As sample D gt O of 99 9 purity was used The Deuterium nucleus was examined It consists of a proton and a neutron and has spin J 1 The gyromagnetic factor z of Deuterium is 6 5349 MHz T 5 1 1 Cool down behavio
36. ecause of problems with the signal to noise ratio we used also the fldscan2 function where the cernox resistance measurement is not included The main goal of this step was to reduce the antenna effect of the connection leads to the cernox and prevent the PPMS to feed in extra noise through the user bridge To prevent quenching of the superconducting magnet due to a low helium level below 6096 and reduce the helium consumption a function script called ppmsdown was made ppmsdown consists of a sequence setting the field to 0 Oe waiting for the persist mode status on the magnet and setting the temperature on standby modus with the sendgpib command SHUTDOWN 5 Using fldscani revealed a problem mainly a Matlab problem which concerns the filling of the virtual memory trough the storing of all variables during the whole scan especially the variable mes dat At about 150 or 200 field points the computer is likely crash For that reason it is recommendable to restart the computer in order to clear the virtual mem ory before starting a field scan If there is a need to perform scans with more field points we suggest to not use the mes dat variable 28 CHAPTER 3 EXPERIMENTAL APPARATUS Chapter 4 Construction Part The construction part of the NMR probe head and the 70 MHz spectrometer are the main subjects of this diploma thesis and its construction will be presented in this chapter 41 NMR NQR probe head This section will motivate a
37. eld 3 The quadrupole frequency vg is defined as 3eV Q UQ HCI Dh 2 27 We observe that the 3 o 3 transition is not affected by the first order quadrupole effect All other transitions are shifted in frequency resulting in satellites in the Fourier spectrum see figure 2 2 This way we get 21 different lines In the case of a powder sample consisting of many small crystallites oriented randomly the resonance lines broaden To calculate the line shape the contributions from all spatial 10 CHAPTER 2 THEORY I 3 2 Hz H Ha m 3 2 77 lt X 3 2 2 1 2 NN z 7 2 I Sa EN 41 2 d NN 1 2 nt ae 3 2 ana 1 I 4 I yo 4 4 i 4 Va 1 Na d 1 1 1 d 4 7 0C M i i i i 1 1 i v v VL vo Figure 2 2 Result of the combination of the Zeeman and the quadrupole interaction for a Spin J 3 2 nuclei in an external magnetic field 2 4 FREQUENCY BROADENING 11 orientations have to be included via powder averaging The theoretical frequency pattern of a powder sample for a spin J 7 2 nuclei is displayed in figure 2 3 In powder samples Vg represents the splitting of the first pair of peaks half the splitting between the following pairs of peaks etc see figure 2 3 I 3 Vo Tr e2qQ h M 3VqQ Vo o VotUQ Ugt
38. equency generator This frequency is transformed by the harmonic generator into a 390 MHz and a 80 MHz signal The 80 MHz signal is split by the 4 phase modulator into four 20 MHz signals with four different phases 0 90 180 and 270 The 20 MHz signal with the chosen phase is mixed in conversion 2 transmitting with the 390 MHz one to continue with a 410 MHz signal with the chosen phase 33Connector Lo IN CHAPTER 3 EXPERIMENTAL APPARATUS 20 probe Frequency circuit synthesizer signal with frequencies F D signal F D preamplifier conversion 1 RX a receiving conversion 1 transmitting amplitude modu broad band transmit amplifier with gate Tx out 1 Vpp lt 410MHz gt 410MHz TX out to power amplifier NMR NOR spectrometer with double conversion intermediate frequencies 410MHz 20MHz F 410 MHz very sedlom used lation 80MHz 410 MHz 0 410MHz 20 MHz 20MHz 0 4 phase gt quadratur detector quadrature generator modulator 20 MHz complex time signal harmonic 390 MHz generator 80 MHz main amplifier conversion 2 intermediate frequency 1 intermadiate frequency 2 410 MHz 5MHz receiving signal 20 MHz D conversion 2 transmitting 20 MHz 90 imaginary signal 0 low frequency amplifi
39. er A Filter bandwidth 10kHz 5MHz this sign stands for control from pulsgenerator CH2 LeCroy double channel oszilloscope tor similar Figure 3 3 Schematical description of the Lupotto IV spectrometer 3 2 NMR COMPONENTS 21 The conversion units are mixers working as multipliers with two output signals where the unwanted ones are filtered out For more clarity we resume the conversion units in table 3 1 conversion element input 1 input 2 wanted output unwanted output conv 1 transmitting F 410 MHz 410MHz F F 820 MHz conv 2 transmitting 390 MHz 20MHz 410 MHz 370 MHz conv 1 receiving F 410 MHz F D 410MHz D 410MHz 2 F D conv 2 receiving 410MHz D 390MHz 20MHz D 800 MHz D Table 3 1 Summary of conversion elements used in the Lupotto IV spectrometer This 410 MHz signal is mixed in conversion 1 transmitting with the input signal coming from the frequency generator What remains is the chosen frequency F that is passed to the power amplifier during the pulse time which is set by the pulse controller The spin signal returning from the experiment is the chosen frequency F with an added frequency window D In conversion 1 receiving the spin signal is mixed with the fre quency generator signal yielding the signal at 410 MHz D The main amplifier amplifies a window of 410 MHz 5 MHz In conversion 2 receiving the main amplifier output is mixed with the 390 MHz signal The resulting 2
40. et from the pulse generator is a logical 1 for transmitting and a logical 0 for receiving It has to be pointed out that the main subject of the control circuit is to keep the spectrometer on receive mode in any case and switch to transmit mode just for sending a pulse To prevent the spectrometer to switch to the transmit mode in case of e a broken or floating connection to the pulse controller the input line of the control circuit is grounded to pull down the line to a logical 0 e a broken or floating connection between control circuit and spectrometer each TTL line is pulled up to 5 V with 1 resistor logical 1 inside the spectrometer Time lag fuses are used 40 TTL from Pulse controller Power Supply Power Supply 5V 74 00 Poa SS st a 2 CHAPTER 4 CONSTRUCTION PART 70 MHz Spectrometer Switch 9 R Switch 3 Switch 12 NAND Switch 21 R 1kQ Figure 4 7 TTL control circuit of the switches in the spectrometer 4 2 70MHZ NMR SPECTROMETER 41 4 2 3 Comments on the construction of the 70 MHz Spectrome ter The construction of the home made spectrometer contained different stages of handcraft The hardware part in the mechanical workshop consisted mainly of preparing the front back and base plates of both spectrometer and power supply The baseplate had to be prepared to hold all th
41. eters by query tau input tau in ms RR input repetition time in ms sweeps input how many sweeps bc window input how large is bc window in points i left input integration interval left edge in points 9 i right input integration interval right edge in points else 71 72 APPENDIX A CODE OF FLDSCAN1 FUNCTION SCRIPT OR input of parameters by direct editing right here tau 0 02 RR 50 sweeps 1000 bc_window 200 i_left 270 i_right 775 end fig_h figure Position 5 220 320 320 plot 1 mes dat echointens ppms_temperature cernox Resist initgpib Apause 10 for f f begin f step f end tic 9 xokekekeloleleeloleleleleleleleleleleekse t 1 flag 0 f str int2str f sendgpib 15 FIELD f str 100 2 pause 20 while flag 0 status readgpib 15 GETDAT 1 Ix find status Ix_max find status HEX dec2hex str2num status Ix 2 1 Ix_max 1 if str2num HEX 3 1 flag 1 pause 5 end Zdisp sch end 4disp 0K field value and ppms_data readgpib 15 GETDAT 2220 find ppms_data
42. fect of this oscillation on the magnetic moment is described in 2 12 This temporary disturbance rotates the magnetization with the Larmor frequency over a cone surface around the z axis observed from the laboratory system producing a macro scopic magnetization component which rotates in the xy plane With an ideal 90 pulse the full magnetization would completely rotate in the ry plane This xy component rotating magnetization induces an electromotive force a measurable response an external detection coil Through inhomogeneities of the external and or internal local magnetic fields the rotation frequencies of the spins are distributed which makes the spins run out of phase This fanning out of the spins generates a loss of the transverse magnetization within the time T Finally the net transverse magnetization disappears This decay is called the free induction decay FID An example of a FID shown in the left panel of figure 2 1 This FID can be Fourier transformed FT from the time space into the frequency space where peaks will appear at the responding frequencies visible in the right panel of figure 2 1 This is a very common NMR approach called Fourier transformed NMR 8 CHAPTER 2 THEORY Ufw Figure 2 1 The FID shown in the left panel frequency space produces after Fourier transformation a peak spectrum in the frequency space right panel 2 3 1 Pulse scheme Through NMR history many
43. ffect related response of the powder grains The natural abundance of CI is of 75 78 and it has a nuclear spin I 3 2 5 2 1 Overload of the spectrometer By using the MITEQ AU 1448 broadband preamplifier with a gain of 53dB we discov ered that after the pulse the receiver path of the spectrometer was not working well for a short time due to an overload caused by extrusion of the transmitting pulse As an FID was clearly seen the possibility an electrical sparking in the resonance circuit could be excluded By changing the preamplifier to a MITEQ AU 1313 with a gain of about 43 db this paralyzing of the spectrometer could be eliminated For a safe operation of the spectrometer a standard receiver attenuation of 8dB was chosen with which no sign of overload could be seen In figure 5 6 the overload lasting for 23 us after a 180 pulse is shown 5 2 2 Comparison 70 Mhz spectrometer vs LupottoII To make a signal to noise ratio comparison of the spectrometers we optimized the 70 MHz spectrometer for a 90 pulse at resonance frequency of 29 931 MHz Then the ampli tude of the FID and its noise was measured at the slightly off resonance frequencies of 29 934 MHz and 29 937 MHz The same procedure was applied with the Lupotto II spec trometer Because of the strong temperature dependence of the NQR frequency of the C13 nuclei in NaClOs the repetition time between two pulses was extended to 2s For each measurement we used 50 sweeps 5
44. field behavior of the quadrupole effects chapter 2 and to clarify if there is any magnetic Ni spin ordering CHAPTER 1 INTRODUCTION Chapter 2 Theory Nuclear magnetic resonance NMR makes use of the interaction of atomic nuclei with its surroundings which are of electrical and magnetic nature using the nucleus as a highly sensitive sensor at an atomic level Already small samples 10 to 10 spins liquid or solid can give conclusive experimental results about its magnetic properties microscopic buildup and phase transitions The continuing NMR evolution began in the 1940 s and has today meaningful applications in many scientific subjects for instance in chemistry biology geology medicine archaeology and physics In this NMR theory section we will restrain to introduce the NMR basics and present shortly the used methods and expected effects For deeper consideration we cite the NMR text books of A Abragam Ref 1 and C P Slichter Ref 2 2 1 NMR basics For a nucleus we have a relation between the total angular momentum J and the magnetic dipole moment 2 1 y is the gyromagnetic ratio a scalar nucleus dependent number J can be written as RI with T being the nuclear spin operator An applied static magnetic field Bo interacts with the magnetic moment of the nucleus described by the Hamiltonian operator 8 2 2 Assuming is aligned parallel to the z axis we get H u hB
45. he PP2000 parallel processor pulse card and is controlled by the software Oscilloscope LeCroy Model LeCroy 9410 Dual 150 MHz Oscilloscope The received complex signal is shown on the screen temporarily saved in the memory 3 1For more information check Ref 16 18 CHAPTER 3 EXPERIMENTAL APPARATUS and its average is calculated Up to one million entries can be summed Over the IEEE Port the complex data is passed to the computer Attenuator Trilithic Model BMA 580 Is used to attenuate the transmitted signal before the power amplifier In the 70 MHz spectrometer this is done internally see section 4 2 1 With the LupottoIV spec trometer we use the following device which is able to attenuate up to 80 dB Wideband power amplifier Kalmus Model Wideband Pulse Amplifier LP1000LRF 1000 Watt 5 160 MHz 61 dB Gain The pulse package is strongly amplified and is given forward through the duplexer to the probe head Because of the amplifier s slow response the gate delay is set before the amplifier Gate Delay NMR Team Physics Institute University of Z rich When the gate receives a pulse from the pulse generator it sends a TTL signal to the power amplifier to get it activated After an adjustable delay the pulse is sent to the spectrometer Duplexer NMR Team Physics Institute University of Z rich Model 30 76 MHz amp NMR Team Physics Institute University of Z rich Model 15 30 MHz The signal fro
46. he line the central line width varies inversely with the Larmor frequency An additional second order broadening arises from by imperfections of the crystallites similar to the ones discussed in subsec tion 2 4 1 14 CHAPTER 2 THEORY Chapter 3 Experimental Apparatus In this chapter the experimental apparatus is introduced At first there will be a schemat ical overview of the instrumental part of NMR and then the different components are described Continuing with a presentation of the Physical Properties Measurement System PPMS manufactured by Quantum Design we will advance to the combination of this two systems The home made NMR components the 70 MHz spectrometer and NMR probe head will be discussed in separate chapters for deeper consideration 3 1 Schematical description The following description refers to figure3 1 The pulse generator creates rectangular pulses whose lengths t can be set to a desired valued Several pulses in a row can be generated whose time spaces are controlled by the pulse card In the software the repeti tion time trep of a pulse sequence and the spacing time between the pulses of a package are set Typical timescales are us for a pulse some ms for the spacing time and frac tions of a second to several seconds for the repetition time The spectrometer allows a high frequency AC signal pass coming from the frequency generator during a chosen pulse length This signal has an amplitude that is t
47. here made at the same three temperatures 300 K 40K and At 4K and 40K the field scans were made starting from 6T down to 4T with steps of 0 02 T At 300 K two field scans were made and combined one from 5 13 T to 4 83 T and another one from 6 T to 4 7 T both with 0 02 steps The parameters of the three measurements are listed in the following table vs MHz thon us torso is trep lms ys Attr 48 Atte dB sw C4 300 30 20015 2 3 5 50 30 9 17 10000 1 40 30 20015 2 3 5 50 30 9 20 50000 1 4 30 20015 2 3 5 1000 30 9 33 400 1 Discussion The corresponding data are presented in figure 6 4 Again to scale the data plots the integrated intensities were calculated between 4 7T to 6T A temperature be havior analogous to the La measurement at 51 2 MHz can be observed Peak has a similar amplitude and width above at 300 and 40 At 4K we see again the broadening and the decrease in amplitude of the peak The peak width summary in table 6 2 shows a slight decrease of A for peak A from 300 K to 40 and again a doubling of the width from 40 K down to AK Peak B shows a width increase with temperature although less than in the measurement at 51 2 MHz A slight shift of peak A at room temperature is also visible 60 CHAPTER 6 LABANIO 5 INVESTIGATION irtesity abu 46 48 5 52 54 fidd T Figure 6 4 La field scans from AT to 6 T v 30 2 MHz 6 2 4 Ba field scan at
48. his can be explained with the diffusion of the molecules in the D20 liquid The second measurement was made to observe a drift over a larger time scale and to see the difference in drifting between using the linear approach mode or the oscillating approach mode in setting the field section 3 3 Both of the measurements were done with a magnetic field of 3 5 T at a spectrometer frequency of 22 921 MHz a 90 pulse of 15 us a transmitter attenuation of 22dB and with the resonance circuit setup as used before The resulting data is shown in figure 5 5 4 5 4 e linear approachemo amp oscillatingapproache mode fit linear ap mode fit oscillatingap mode magndic dift G 0 100 200 300 400 500 timestamp min Figure 5 5 Magnetic drift of the PPMS With the linear approach mode the magnetic field drifts at a rate of 46 49 G h while whith the oscillating approach the drift rate is 0 26 G h We measured a fast drift using the linear approach mode and no clearly detectable drift with the oscillating approach mode Making a linear fit for both approaching modes we obtain 0 26 G h for the oscillating approach and 46 49 G h for the linear approach By observing the resulting FWHM of the resonance lines of these measurements we can 48 CHAPTER 5 TESTING PROCEDURE also observe a difference between the two approaching modes For the linear mode we get a mean value A 1 20 kHz Ag 1 84
49. ifications we refer to Refs 4 and 5 A picture of the PPMS is shown in figure3 4 The system can be completely controlled and monitored through the model 6000 controller or the MultiVu software running on a PC Easy sequence script can be written with the MultiVu software to let the PPMS work fully automated Unfortunately this MultiVu software is not as easy to use for communication with external devices since the standard IEEE communication protocols are not directly supported Figure 3 4 PPMS and Model 6000 Controller 3 3 PPMS A SHORT DESCRIPTION 23 The high capacity nitrogen jacked dewar 4 prevents fast liquid helium boil off The PPMS insert figure 3 5 with the superconducting magnet is immersed in liquid helium The outer layer of the insert is a superconducting magnet which is made of NbTi Nb3Sn The su perconducting coil can be charged and discharged using the persistent switch After charging the magnet is set to persistent mode i e disconnected from its power supply There are three modes for setting the magnetic field The oscillating mode the no overshooting mode and the linear mode As shown in section 5 1 4 the oscillating mode is almost stable while the other modes result in a drift of the field 7 The field uniformity is claimed to 0 01 96 over a 5 5cm 1 cylindrical volume with the center at 5 1 cm above the puck surface This has been verified in the testing phase of the NMR probe head sectio
50. is out of and can be fixed with three small Berilco screws onto the bars The coil sample holder and the holder mounting are fastened on top of it with nylon screws All these parts are made of Torlon Two coil sample holders of different sizes were made The available standart Torlon sample capsules fit into the smaller of these coil sample holders 34 CHAPTER 4 CONSTRUCTION PART E Bottom Figure 4 3 Full NMR probe head 4 1 NMR NQR PROBE HEAD 35 Figure 4 4 Bottom of the NMR probe head sample plateau noise blocking capacitor holder mount Dy thermal sensor coil sample holder Figure 4 5 Coil sample holder plateau thermal sensor and sample puck 36 CHAPTER 4 CONSTRUCTION PART 4 1 5 Temperature readout To have better control of the sample temperature a CERNOX temperature sensor was installed close to the coil as seen in figure4 5 The CERNOX sensor is read out through the User Bridge 1 port by a 4 terminal measurement The used connector pins on the puck are chosen according to a user Bridge resistance measurement using Channel 1 see the PPMS Hardware Manual 4 The calibration measurements are carried out at 0T 3T 6T 9T and are presented in the appendix B46 Because of strong disturbing interference noise two 1000 pF blocking capacitors were added 46The data is saved on the PPMS computer in folder C QdPpms Data alex cernoxCal 4
51. ites can not be excluded The line widths were calculated by fitting the data points with the sum of two Lorentz functions using the Levenberg Marquardt algorithm 2 AA 2 Tea 55 peak A peak B yo is the intensity offset vp is the frequency at peak maxima 24 7 and 2B the respec tive peak amplitudes A and Ag the respective FWHM The results are presented later in table 6 2 in the discussion section 6 3 We note that a spin echo response was still detectable at the limiting frequencies of this measurement 44 0 MHz and 57 2MHz not shown in figure 6 2 58 CHAPTER 6 LABANIO 5 INVESTIGATION 6 2 2 11 field scan at 51 2 MHz Three field scans at the temperatures 300 K 40 K and 4K where done starting form 9 T and decreasing magnetic field To get the peak maxima at 8 5 T we chose v 51 20157 MHz Three different field intervals were measured at room temperature 300 K and combined afterwards Scans from 9 T to 8 7 T and form 8 3 T to 6 T were done in 0 02 T steps and from 8 7 T to 8 3 T in 0 01 T steps At T 40K and field scans were made from 9 T to and 9 T to 7 7 respectively both with 0 02 T steps The used parameters are listed below vs MHz tpo us Gus us trep lms ys Attr dB Atte dB sw C 300 51 20157 2 3 5 50 20 9 22 1000 4 40 51 20157 2 3 5 50 30 9 22 1000 4 4 51 20157 2 3 5 2000 30 9 37 28 4 8 irtesity aby B
52. ivalent to dn no n 2 16 where W 2 17 P 3 a 1 I 2 18 i W W To solve 2 16 we take the basic exponential approach n no 2 19 where no is the population difference in thermal equilibrium and T the characteristic time which the system needs to reach the thermal equilibrium Analogous to equation 2 16 we get the phenomenological equation for the magnetization M yhn 2 which is assumed to be parallel in the z axis dM M M 4 q i with Mo the magnetization in thermal equilibrium By combining 2 20 with the equation 2 20 of motion of the magnetization M we get dM p ME M B 2 21 dt 6242 2 21 Since the magnetization in thermal equilibrium is aligned with the external static magnetic field Bo the components of the magnetization have to dissappear along the x axis and the y axis after the time T 5 M x B 2 22 n M x B E 2 23 Combining equations 2 21 2 23 and 2 22 we get the Bloch equation dM A 2 24 7 M56 Mo M _ sai x B ENG Ji To The solution of the Bloch equation if a radio frequency RF field is applied can be found in Refs 1 and 2 2 3 FREE INDUCTION DECAY AND MEASURING METHODS 7 We resume the relaxation times T The spin lattice relaxation time is a measure of the time for the energy exchange between the spin system and the
53. lattice being the time it takes for the magnetization to return to thermal equilibrium in the external magnetic field direction Typical time values for solid state materials are parts of ms until hours T gt The spin spin relaxation time is the time the magnetic moments of the nuclear spins need to get out of phase in the plane perpendicu lar to the external magnetic field causing a deacay of the magnetization in this plane Typical values for T gt are about 100 pus It is clear that 75 cannot be larger than 2 3 Free induction decay and measuring methods To resume a spin system with no preferred direction shows a completely isotropic spin orientation pattern By applying an external magnetic field in the z direction the spins order to be in an energetically favorable orientation For the nuclear moments with a positive y an antiparallel alignment is favorable The presence of the external magnetic field splits the degenerated levels so that the Zeeman effect can be observed The popu lation of the m levels in thermal equilibrium follows the Boltzmann distribution shown in 2 7 This distribution describes the spin polarization 2 9 resulting in a magnetization M along the magnetic field 2 10 As there is no phase coherence of the spins there is no transverse magnetization Transitions are induced through a RF excitation pulse if an oscillating magnetic field B is applied perpendicular to Bo via a coil during the time t The ef
54. m the power amplifier is guided to the probe head The returning spin signal is directed to the preamplifier To prevent the power amplifier to feed noise through the duplexer when no pulse is to be sent antiparallel diodes in series are set before the duplexer Only AC voltages higher than 0 6 V can pass Preamplifier MITEQ Model AU 1448 Amplifies the weak spin signal and sends it to the spectrometer The amplification is of 53 5 dB according to the constructors data sheet Antiparallel diodes are set to the ground before the preamplifier to prevent damage on it through extruding remnants of the pulse This way voltages over 0 6 V are absorbed Software The software runs on Matlab on a Windows 3 1 operating system We used mainly 3 types of Matlab function scripts M Files pppc stands for pulse programming for personal computer It is an interface to con trol the repetition time trep the spacing time the number of sweeps the phase changes and the toggle mode It is very useful to get started and for optimization of the pulses 32The reason of using this older computer is the incompatibility of the pulse card with newer computers and operating systems 3 2 NMR COMPONENTS 19 t2seq14 is a Matlab function script to make 75 echo measurements and it is a useful tool to obtain clean echo signals In the command the name of the output file the amount of sweeps the spacing time and the repetition time trep are set t2seq14 c
55. n 5 1 2 For temperature control the PPMS offers the high temperature control and the continuous low temperature control techniques 4 which allow a fast cooling and heating of the PPMS sample space a smooth transition through the 4 2 K helium boiling point and maintain indefinitely a temperature below 4 2 K Thermal control and monitoring is situated below the PPMS sample puck To monitor and control thermal gradients in the PPMS sample space a neck thermometer and heater is wrapped around the sample at about 14 cm above the puck surface For additional information we refer to the PPMS User s Manual see Refs 4 5 and 6 3 6This with a maximal charging rate of 196 3 Oe s 3 7We tested linear mode only 38Typical warming and cooling rates are around 6 K min although they can be extended and set up to 20 K min CHAPTER 3 EXPERIMENTAL APPARATUS 24 BAFFLES MAGNET ee HELIUM LEVEL METER NECK THERMOMETER 2 5 cm SAMPLE SPACE COOLING ANNULUS SUPER INSULATION 4 gt VACUUM SAMPLE TUBE INNER ee VACUUM TUBE HEAT gt SHIELD Psa OUTER VACUUM TUBE BE REGION OF ACMS DETECTION COILS gt 001 UNIFORMITY 1 HORIZONTAL ROTATOR 3 8 cm PLATFORM AXIS OF ROTATION lt i 7 14
56. n of the magnetic dipole moment under the influence of the static field Bo and the oscillating field B can be described in a system rotating with angular frequency w see Ref 2 by 3 Des ae x Bert with Bert Bo De 2 12 This means that observed in the rotating system the magnetic moment feels an effective static magnetic field Ber f The moment precesses along a cone surface with a constant 2 2 BLOCH EQUATIONS 5 angle around the axis of B having an angular frequency of In resonance i e with yBo the effective field is equal to B A moment being aligned along the z axis will rotate in the yz plane of the rotating system If B is applied for the time span tp the moment turns by an angle a yB tp By choosing B and t so that a 7 the magnetic moment inverts its orientation The corresponding pulse is called a 180 pulse If we set the field and the time span to get a 7 2 a 90 pulse is gen erated and the magnetic moment will have rotated from the z direction to the y direction observed in the rotating system As soon as the 90 pulse is over the magnetic moment will precess in the laboratory system around the direction of the static magnetic field Bo along the z axis Applying a pulse on a sample results in an energy and temperature gain of the spin system Due to relaxation processes energy is transfered between the spin system and its surround ings Since the spi
57. n system is coupled to the lattice the latter is capable to absorb this energy which is described in more detail in Ref 2 The possible relaxation channels are of magnetic and electrical type The perturbed spin system returns to thermal equilibrium due to fluctuating local magnetic fields which is the so called magnetic relaxation In simple metals the relaxation is dominated by the magnetic Fermi contact interaction of the nuclear spins with the spins of the conduction electrons If the relaxation is produced by fluctuations of the electric field gradient EFG at the nuclear sites it is a quadrupolar relaxation see section 2 4 2 2 Bloch equations For two energy levels only one transition is possible Let us define N and N as the population numbers in thermal equilibrium then n N N denotes the population difference between the levels Assuming Maxwell Boltzmann statistics we can write the population ratio as follows N AE exp 313 ar em 2 13 If there is a mechanism which induces transitions between the two energy levels a rate can be defined as dN N W NW 2 14 W and W denote the probability of inducing a spin transition upward and downward in energy respectively By defining the total population number as N N N_ the 6 CHAPTER 2 THEORY difference of the population number N N_ can be written as d N W Wi n W W 2 15 which is equ
58. nd present the construction of the NMR probe head describe the environment to which it will be exposed present the electrical system and the tem perature readout 4 1 1 Motivation There is a rich history of NMR NQR studies on solid state materials in the Physics In stitute of the University of Z rich In recent times this was especially driven forward by J Roos and M Mali members of the Keller Group The used experimental method having a static field of about 9 T is designed to perform frequency scans The magnetic field of the superconducting magnet produced by Cryomagnetics Inc is very homogeneous and stable over years of usage A sample can be cooled down to z 7K With the PPMS magnet we can not reach such a high stability and homogeneity but the possibility to have variable magnetic fields up to 9 T and to be able to cool the sam ple down to 1 8K makes the PPMS a very interesting complementary option Having a NMR NQR probe for the PPMS would open various possibilities such as performing magnetic field scans in comparison to the used frequency scans and to low temperature NMR or NQR 29 30 CHAPTER 4 CONSTRUCTION PART 4 1 2 Requirements for the probe head Concerning the environment to which the probe head will be exposed the choice of the material and the construction design had to be done with special care Because of the high magnetic fields the probe head had to be manufactured with low magnetic response material
59. of the NMR apparatus To do so the MultiVu software was not used at all and the controlling of the PPMS was done by direct connection through the general purpose interface bus GPIB port to the Model 6000 Controller As the NMR apparatus is controlled with Matlab there is also the possibility to drive the PPMS system with basic Matlab GPIB commands For this purpose the fldscan1 m function was written which we will overview in the fol lowing We will point out the most important script commands and GPIB commands The whole program code is presented in appendix A The function to run a field scan is defined as follows fldscan1 f_begin f_step f_end query The parameters to be set are name The name of the data file must be chosen and set as a string variable The variables mes_dat echointens cernox_Resist and ppms_temperature will be saved into a MAT File with this name f_begin Initial magnetic field value in Oersted f_step Magnetic field stepping size in Oersted f_end Final magnetic field value in Oersted query Setting the NMR parameters By setting query 1 the user will be asked to set the values for the NMR parameters If the query is set 0 or else the values have to be set directly in the f1dscan1 m file The NMR parameters are The spacing time tau the repetition time RR the number of sweeps sweeps the size of the baseline correction window bc window and the integration limits over the echo i left and i
60. ol 2 3 I describes the projection of I to the quantization axis Bo causes a splitting of the degenerated states into 2 1 equidistant energy eigenvalues of the form En yhBom with m lI I 1 I 2 4 3 4 CHAPTER 2 THEORY This energy splitting induced by an external static magnetic field is called the nuclear Zeeman effect Building the energy eigenvalue Em m H m we get the energy relation AE is the energy difference between the split energy levels Am 1 With this we obtain the angular Larmor frequency wr yBo 2 6 From thermodynamics we know that the occupation of the energy levels in thermal equi librium follows the Boltzmann distribution P Em x exp Em keT 2 7 This unequal population of the energy levels produces a polarization of the nuclear spin system By summing all the magnetic moments per volume unit and writing out the nuclear spin polarization 1 we get the magnetization M T iz M 2 8 yhI I 1 B h Ly 2 9 with L HIT 2 9 gt YRII 1 gt M 2 10 3kpT where N is the nuclear spin density The split energy multiplets can be detected when transitions between these levels occur induced by alternating electromagnetic fields If we apply an oscillating magnetic field B along the x axis which is perpendicular to equation 2 3 reads H yhBol yh D L cos t 2 11 The motio
61. onsists of two measurement series making a different pulse phase in terplay to reduce instrumental noise produced by the pulses Afterwards the difference between these two measurements is taken with the Matlab function script pp_pcdif to eliminate an overall offset Each of these measurement series carries out the given amount of sweeps t2seq14 is also used in the fldscani function fldscani isa function to combine NMR apparatus with the PPMS This function is described in more detail in section 3 4 Spectrometer Two spectrometers were used the self made 70 MHz spectrometer and the LupottoIV spectrometer The 70MHz spectrometer is described in section 4 2 1 For comparison the LupottoIV spectrometer is described in the following subsection 3 2 1 Lupotto VI spectrometer The LupottoIV spectrometer was built for the Institute of Physics University of Z rich by P Lupotto It is the fourth and last one of the Lupotto series The spectrometer was complemented later on documented and repaired on various occasions by P Soland at the same institute To describe the spectrometer we will follow the sketch shown in figure3 3 on page 20 The spectrometer s internal working frequency is 410 MHz The incoming frequency gen erator frequency synthesizer signal is the sum of the working frequency and the chosen frequency F v used for the experiment Parallel to it the spectrometer receives the ref erence frequency of 10 MHz from the fr
62. oo large for the broad band power amplifier and is therefore attenuated Due to the slow response of the power amplifier a gate delay is set in between It ac tivates the power amplifier and delays the pulse command to the spectrometer to switch into the transmitting mode for sending a pulse This amplified package is guided through the duplexer to the probe head There in the resonance circuit which consists mainly of two capacitors and a coil the pulse excites an oscillating magnetic field in the coil The resonance circuit will be explained in more detail in the NMR probe head construction section 4 1 3 After the excitation the whole system switches to the receiving mode The weak out coming signal is driven by the duplexer into the receiver path beginning with the pream 15 16 CHAPTER 3 EXPERIMENTAL APPARATUS Broadband Power Amplifier n 1000 W Temperature Control Pre Field Control Amplifier Attenuator Computer Data Analysis with Matlab Gate Delay F zx A Attenuator Pulse Generator Spectrometer Fourier Transformation Frequency Generator lose Network Oscilloscope Network Figure 3 1 Schematical description of the NMR apparatus plifier for intensification The receiving attenuation is used to prevent an overload of the spectrometer The spectrometer transforms the complex signal into two parts with a phase difference of 90 and sends them to the digital memo
63. ortional to 1 v and becomes neg ligible for high magnetic fields The satellites are also shifted with respect to vz With weak second order quadrupole interaction the satellites are not noticeably affected but when the second order effects become stronger the satellites shift by different amounts from their first order positions In extreme cases this causes the satellites to change to different relative positions 3 Pairwise 2 3 2 8 etc their separation remains the same The shape of the central line of a powder sample is also influenced it becomes asymmetric and a splitting can be observed figure 2 4 Figure 2 4 Spectrum for the central line 3 gt 4 with second order quadrupole effects present The dashed line represents the theoretical function while the solid line is a symmetrically broadened shape The frequencies v and v correspond to values of 0 and 4 5 9 respectively 3 The shape is determined by different shifts due to the different angels 0 of the randomly oriented powder crystallites Considering figure 2 4 the right maxima occurs at u 0 2 4 FREQUENCY BROADENING 0 90 with the frequency 2 ee Tee 16V The left maxima is at u 4 5 9 0 41 8 with the frequency 2 p eT a A Ov The difference between the two maxima is 25 2 Api pit Q d d TAG T 3 4 13 2 30 2 31 2 32 If the second order quadrupole interaction dominates the shape of t
64. osition dependent magnetic field homogeneity 5 1 4 Drift of the magnetic field EI a se Se ra E Ge ae Peds ee 5 2 70MHz spectrometer 5 2 1 Overload of the spectrometer 5 2 2 Comparison 70 Mhz spectrometer vs Lupottoll 2242 UES Q t lead Ms MEO Hehe ee UGG LaBaNiO Investigation 6 1 Sample characterization 6 2 Summary of NMR experiments 6 21 19La frequency scan at 8 5T 6 2 2 139La field scan at DISZMBS 5 a 62 3 Seba field scan at a 6 24 137Ba field scan at 40 2 MHz and 40 6 2 5 79 Ba field scan at 33 8 MHz and 40K 4 2 626 La temperature run at 51 2 MHz cxx en 6 3 Discussion Summary and Outlook Code of fldscan1 Function Script Calibration of the CERNOX Temperature Sensor 70 MHz Spectrometer Connectors and RF Elements C 1 Connectorslineup _ I eS C 1 1 70 gt C 1 2 Power supply C7 RE elements ce JU q Seino ds SE ey Vie ern ee ean References 71 75 77 TT 78 78 79 81 Chapter 1 Introduction In the first half of the 20 century with the increasing knowledge and unders
65. r of the probe head As the probe head was expected to work well at low temperatures the first check focused on the mechanical behavior where we wanted to make sure that the tuning and matching of the resonance circuit could still be done without the capacitors getting stuck due to ma terial dilatation or freezing out of gases At the same time the behavior of the CERNOX thermometer placed next to the sample could be monitored The probe head was mounted into the PPMS and the resonance circuit was tuned and matched properly After a night stable at room temperature 300 K the cooling process started matching and tuning was frequently checked and adjusted The capacitors were indeed still movable down to 1 8 the final puck temperature During cool down the CERNOX thermometer at the sample position was tightly following the temperature measured by the PPMS puck thermometer Concerning temperature control below 10K best stability was obtained by interrupting a cool down from room temperature at 10 K for 30 45 minutes before going lower in temper atures gt With equation 2 6 vr 7Bo 43 44 CHAPTER 5 TESTING PROCEDURE 5 1 2 Magnetic field homogeneity of the PPMS For this homogeneity test a spherical glass container with an outer diameter of 9mm was filled with D5O One end of the container closed by melting had an additional 8 mm long tail which was also containing a small amount of D20 To tune the resonance circuit at
66. right respectively The output variables ppms temperature and cernox Resist are vectors with one readout per magnetic field point echointens is a 2 column matrix containing in the first column the magnetic field points and in the second one the integral over the echo section 2 3 1 which represents the intensity of the signal at a given field as a complex number The 26 CHAPTER 3 EXPERIMENTAL APPARATUS values are drawn in a figure window which allows online inspection while the experiment is running mes dat is a so called format 1 matrix see NMR Matlab description In the following we describe the sequence of commands of the script fldscani m ap pendix A First the function is defined After some comments the relevant NMR param eters are specified Then the figure is plotted and the variables and the GPIB port are initialized At that point the main for loop which runs the experiment starts At first the magnetic field is set with the following command sendgpib 15 FIELD f str 100 2 The sendgpib like the readgpib command are Matlab commands in order to connect to other machines through the GPIB port The first value has to be the GPIB port number of the machine which is in our case 15 The second expression is expected to be the string command to manipulate the machine by applying the language of the machines For this purpose Quantum Design offers a chapter in the software manual called PPMS GPIB Commands
67. ry oscilloscope At this point the cho sen amount of sweeps is saved in the buffer summed up and the mean value is displayed This way it is possible to considerably improve the signal to noise ratio The oscilloscope is triggered by the pulse controller The processed complex signal is taken out through the IEEE interface into the computer and it is converted into a Matlab variable in matrix form and saved for further analysis 3 2 NMR COMPONENTS 17 3 2 NMR components In this section the different used NMR components will be described A picture of the NMR apparatus is shown in figure 3 2 Figure 3 2 Wrack holding units of the NMR Apparatus Frequency generator Rohde amp Schwarz Model Signal Generator 100 kHz 1000 MHz SMG amp PTS Model PTS 310 Frequency Synthesizer 1 810 MHz It creates the frequency signal for the spectrometer The radio frequency RF to be set is the sum of the spectrometer s internal working frequency and the chosen frequency v for the experiment It also produces the reference frequency of 10 MHz for the spectrometer Pulse generator University of Z rich H Zimmermann L Pauli Model PP2000 Pulse Controller The pulse lengths are set with the pulse generator Pulses of 0 5 us up to 10ms are available Five pulses can be created whereas the order of the pulses must be set One of those pulses is used as a trigger for the oscilloscope The timing of these pulses is controlled through t
68. s so that no mechanical forces are created and no inhomogeneity of the magnetic field is induced Also thermal conductance from the top of the sample space being at room temperature to the bottom with temperatures below the temperature of liquid helium at 4 2 over a distance of 89 3cm had to be minimized At the same time the probe head has to be mechanically stable and fit into the narrow tube of the sample space having a diameter of 2 68 cm In the following an overview of the used materials is given The upper part of the probe head is mainly made of stainless steal to achieve mechanical and thermal stability with the exception of the capacitor driving mechanism that is made partly out of aluminum see figure 4 2 The used non magnetic materials in lower part extending 38 cm above the sample puck surface are brass Torlon and Berilco see figure4 4 The sample coil holder and the sample plateau are made of Torlon 4 1 3 Electrical properties of the probe head and the resonance circuit This subsection has two parts in the first part we describe the coaxial line down to the resonance circuit in the second one we explain the resonance circuit that is used with special focus on the tuning and matching capacitor The main purpose of the electrical system of the probe head is to deposit the major part of electrical energy as an oscillating magnetic field in the coil where the sample is placed Moreover it is supposed to pick up the
69. so us te ms r us dB Attg dB sw CH 33 8352 40 4 8 50 30 9 11 40000 2 Discussion The data are shown in figure6 6 The graphically evaluated FWHM at the peak is Ag 0 55 15 T A 2 3 7 MHz As with to the Ba scan an interpretation of the data is difficult although the peak at 8T is clearly visible In order to improve the signal to noise ratio longer signal averaging at each data point would be needed implicating time consuming scans of up to one week 62 CHAPTER 6 LABANIO INVESTIGATION N irtesity abu k 7 6 7 8 8 a2 84 naydicfidd T Figure 6 6 Ba field scan from 8 7 T to 7 3 T v 33 8 MHz at 40K 6 2 6 La temperature run at 51 2 MHz As we have seen in the above La measurements above peak A loses intensity and broad ens by cooling down the sample To analyze this behavior a temperature run was done manually from 300 K down to 2 K At each temperature a 3 point field scan was done with measurements at 8 575 T 8 5 T and 8 425 T This three field values represent the maxi mum of peak A and the crossover points between peak A and B at each side of peak A respectively We will define this points as 8 5 for the maximum of peak AB 8 425 T for the left crossover and ABg 8 575 T for the right crossover see upper right panel of figure 6 7 To prevent electrical breakthrough in the resonance circuit and spectrometer overload at the lowest temperature the attenuation and
70. t mode with logical 0 4 2 70MHZ NMR SPECTROMETER 39 The input for the receiver path is Rx IN The signal from the preamplifier has the chosen frequency v with a band width window of A The preamplifier amplifies the spin signal coming from the probe head and can be attenuated 22 with up to 45 5dB After switch 21 this signal is mixed 20 with the signal from switch 12 to get the spectrometer working frequency range of 70 MHz A The overall band width is reduced by a 70 MHz band pass 19 to 7 MHz In the next stage 16 amp 18 this signal is amplified After another 70 MHz band pass filter 15 the signal enters to the 1620 demodulator 12 to be mixed with the 70 MHz frequency from switch 3 The output signals of the spectrometer are tA separated in two 90 phase differing signals X and Y which are connected to channel 1 and 2 of the digital oscilloscope Both signals pass a low pass filter 26 and 27 of 1 9MHz 4 2 2 Power supply and the TTL control circuit The power supply is equipped with two POWER ONE HA5 1 5 OVP A power units de livering 0 125 for 5 V and 5 V each and one POWER ONE HC15 3 A unit with 0 5 A at 15 For the TTL control signal we equipped a board with a FAIRCHILD 7 00 Quad 2 Input NAND Gate IC containing 4 NAND gates one for each switch of the spectrometer As explained before we need a logical 1 for receiving and a logical 0 for transmitting However what we g
71. tanding of quantum mechanics resonance experiments became more and more important Ex perimental techniques using electrons and the nuclei where developed Nuclear magnetic resonance NMR makes use of the interaction of the nuclei with its electric and magnetic surroundings chapter 2 Using the nuclei as a local probe being positioned directly in the sample opens various possibilities to study materials and compounds in different areas of natural science and medicine In material and solid state physics research the solid state NMR is widely applied to crystals and powder samples The goal of this diploma thesis is to implement an NMR system to be used with the Phys ical Properties Measurement System PPMS manufactured by Quantum Design with the possibility of driving the magnetic field up to 9T and the temperature down to 1 8 K Such a system then opens the opportunity of performing field scans and extending the cooling range for NMR experiments at this institute The NMR apparatus chapter3 was ex tended by the construction of anew NMR probe head and a new spectrometer chapter 4 which were tested on functionality and performance chapter 5 Finally the constructed elements where applied to do investigation on the La Ba and Ba nuclei of LaBaNiO _ chapter 6 which is thought to be a candidate to repre sent a parent compound of Ni based superconductors The main goal of our NMR study is to investigate the temperature and
72. the sample position This is also described in the appendix of the PPMS GPIB Commands Manual 5 In our case we are looking for a magnet status 1 The default optional 3 2Not described in Ref 5 3 4 COMBINING THE NMR APPARATUS WITH THE PPMS 2r NoUpdateFlag is by default O for updating all readings before returning the data and 1 for returning the most current values immediately We use the default To read out these data some skills had to be applied in order to turn strings into integers to select the right output values to be treated The field value the temperature and the resistance of the cernox thermometer are read out For this the same GETDAT GPIB command is used but with a DataFlag value of 22 to read out the desired value The NMR measurement starts at the set magnetic field For this we use the scripts t2seq14 and pp pcdif which are explained in section 3 2 The echo intensity integral is evaluated with the echoeval command which makes the baseline correction and the integration Then the values are written into the variables which are saved and the scan data are drawn in the figure window We also inserted in the for loop the tic and toc command to have an idea how long the measurement at each field takes We have to emphasize that the NMR computer is unaccessible during the whole scan the only way to have access to the data is by downloading the MAT file containing the measured values to another computer B
73. the sample coil holder and placed with its center axis at the height of 5 1 cm above the puck surface see figure 5 3 With this setup measurements at magnetic fields of 3 5 T 22 916 MHz and 9 T 58 954 MHz were made for purpose of comparison The results are shown in the following table 5 2Cylindrical volume of 3mm diameter and 4 7 mm height 5 1 NMR PROBE HEAD AND MAGNETIC FIELD HOMOGENEITY OF THE PPMS45 0 003 0 002 0 001 intensity arb u 5 0 requency kHz Figure 5 2 Fourier transformed of the deuterium spin signal The FWHM is about A 1 9 kHz BIT A kHz Az 10 T 3 5 1 2 1 8 0 0052 9 2 3 1 0 0034 To prove that there was no field inhomogeneity produced by the probe head itself a test probe head was built having a plexiglass tube to guide a coax cable down to the sample space with a resonance circuit containing the same coil sample holder as before The signal was similar to the one with our NMR probe head and no further narrowing of the FWHM could be detected position 0 Figure 5 3 Positioning of the small D2O test tube sample in the sample coil holder 46 CHAPTER 5 TESTING PROCEDURE 5 1 3 Position dependent magnetic field homogeneity To check the position dependence of the magnetic field homogeneity we used the same sam ple as shown in figure 5 3 This was done at a field of 3 5 T with a spectrometer frequency of 22
74. ulting range of capacitance is of 7 8 pF with the quartz glass maximally pulled out and 28 8 pF with the quartz glass pushed in The coil itself has to be made to match the desired parameters of the planned experiment 4 1 4 Construction While the existing NMR NQR probe heads in the Physics Institute have a diameter of 5cm and more our new probe head had to be reduced to a diameter of 2 6cm which presented the main difficulty of the probe head design The parts were designed using the construction environment software AutoCAD and then converted to the more powerful construction software CATIA for easier adaptation to the workshop machines Two analogous probe heads were produced In the following the different parts of the probe head will be explained On the top of the probe head figure 4 2 there is the driving mechanism for the capacitors The quartz glass position for the tuning capacitor can be set by turning the screw head This screwing results in a vertical up and down motion of the stainless steel which is attached to the quartz glass The matching capacitor is directly attached through a shaft stainless steel rod with its screwing head The top also contains a BNC connector to the coaxial line described in the chapter above and two 4 pin Lemo connectors prepared for electrical connections to the sample space The flange stainless steel had to be carefully sealed so that sufficient vacuum can be produced inside the PPMS sample
75. wing the descriptions the theory section 2 4 1 figure 2 3 we expect 6 satellites for a spin 7 2 nuclei like La As there are no clear satellites detectable the calculation of the quadrupole frequency vg from first order quadrupole effects is difficult By taking the field scan at 300K and 9T figure6 3 assuming that the satellites are located symmetrically around the central peak and examining the different possibilities of satellite pairs outside of the crossover points between peak A and B we get a possible range for vg between 0 8 MHz and 4 5 MHz as a first guideline A more promising approach is to observe the second order quadrupole effects theory sec tion 2 4 2 on the central peak A as a function of the magnetic field There is an inverse proportionality between the split central line s maxima Av and the Larmor frequency vz 2 32 while vz is proportional to the magnetic field 2 6 Equation 2 32 can therefore be rewritten as 25 9 I I 1 3 4 E 6 2 45 94 1 Ba Because no splitting can be seen compare to figure 2 4 we set the measured FWHM values A Ay Av The proportionality to the inverse field can be checked by comparing the 4 s at each temperature of the 51 2 MHz field scan measurements with the corresponding 30 2 MHz data table 6 2 For peak this inverse field dependency is less clear We calculate the quadrupole frequency _ 7B Aa la TI 1

Construction of NMR equipment to be used in the - Physik

Contents

Download Pdf Manuals

Related Search

Related Contents