As mentioned in the introduction to this chapter this task was made difficult by the lack of Raman-heterodyne signals. In an attempt to collect enough
3.2 Hyperfine interaction in europium doped Y2SiO5 77 0 20 40 60 80 100 7 6 5 4 3 2 1 0 1
log echo height
Delay (ms)
Figure 3.17: A figure taken from [81] showing spin echo amplitude verses time interval between the driving pulses. The pluses show the echo decay behaviour for the insensitive transition. The crosses and circles show the echo decay behaviour for transitions at the same magnetic field but where the gradient of transition frequency with respect to magnetic field does not vanish. A line corresponding to theT2 = 500µs for zero magnetic field is shown for comparison.
FID
Magnetic Field Laser intensity
Time
Figure 3.18: The sequence events used to take each experimental shot.
Figure 3.19: Europium energy level diagram showing the transitions involved in the holeburning spectra. The splitting of the±5/2 energy levels due to the magnetic field is exaggerated. The hyperfine splittings are of the order of 50 MHz (see Fig. 5.1 on page 112) and the Zeeman splittings were of the order of 500 kHz. The four transitions that might be expected to be seen in the spectra taken are shown. The transitions represented by dotted lines were not observed.
3.2 Hyperfine interaction in europium doped Y2SiO5 79
data to infer the hyperfine parameters, a holeburning technique was used. The sequence of events in each experimental shot is shown in Fig. 3.18. First a long weak pulse (≈ 160 ms) was applied to the sample to burn a narrow hole in the sample. The transition was then split by the application of an applied magnetic field, and a time of 500 ms was allowed for the supplies to reach the appropriate currents. The spectra of the resulting feature was then read out with a short intense pulse (≈0.6 µs). The resulting free induction decay was Fourier transformed to obtain the spectra. The experimental setup used was essentially the same as described in Sec. 5.2.
Just as in the case of Praseodymium dopants, there are four different crys- tallographic positions at which the europium can substitute yttrium. The four positions can be divided into two pairs with the members of each pair related to each other by the crystal’s C2 axis. These two pairs have differ-
ent crystal field splittings and hence different optical transition frequencies. Unlike praseodymium which has only one stable isotope, europium has two stable isotopes of approximately equal abundance, 151Eu and 153Eu. These two isotopes have the same transition optical frequencies but different hy- perfine transition frequencies. The experiments of this thesis were carried out on the 151Eu isotope at “site 1”[102]. In order to simplify the spectra,
the procedure was applied to a broad anti-hole that had been prepared in a manner similar to that described in Sec. 5.3.2. Although other measure- ments were made, all the results presented here were for the±5/2→ ±5/2 transition as shown on Fig. 3.19.
The magnetic field was supplied by a set of custom made superconducting
XY Z coils. These were newer than those used for the praseodymium work and had a larger number of turns. The X and Y coils had sensitivities of 70 G/A and theZ, 50 G/A.
In order to better distinguish between the quadratic Zeeman effect and laser drifts, it was decided to take spectra while increasing the magnetic field along a discrete set of directions.
The crystal was delivered cut with faces perpendicular to the crystallo- graphic axes and the faces marked as such, using an a, b or c. The sample was placed in theXY Z coils with X axis parallel to the b (which is the C2)
axis and the normal of the face markedcpointing along theZ direction. The laser beam was directed along theX axis and polarised in the vertical plane. The optical transition was somewhat polarised with maximum absorption occurring for this vertical polarisation.
3.2.2
Results
From each of the two orientations of “site 1” up to four spectral lines were expected. Thus, in general, eight spectral lines could be expected for an arbitrary field and four when the field was applied along or perpendicular to the C2 axis. Instead, either four lines or two lines were observed.
The spectra as a function of magnetic field for two field directions are shown in Figs. 3.20 and 3.21. Measurements were taken for the field changing along 13 directions. These were approximately the directions3 [abc] where each of a, b and ctook on the values {0,1,¯1}.
For Fig. 3.20, the applied magnetic field was along the C2 axis and two
transitions can be seen, as well as a number of what could be weaker lines. One serious weakness in the method used is that it is not possible to tell if these are due to weakly allowed transitions or to signals from other subgroups of ions, for example, from the other isotope or a packet that is resonant with the laser due to some other transition. The spectral lines split at approxi- mately 1 kHz/G and the line traced out by the peaks of the spectra is gently curved due to the quadratic Zeeman. The magnetic field values used are shown in Fig. 3.22(a).
For Fig. 3.21 the field was ramped along a direction close to the [¯111]. Because this direction was not perpendicular to or parallel with the C2 axis,
four spectral lines can be seen, two for each of the orientations of the site. The quadratic Zeeman splitting was among the highest seen for all the directions measured and was of the order of 4 Hz/G2. The magnetic field values used are shown in Fig. 3.2.4.