- Email: [email protected]
Para-ortho transitions in solid hydrogen
Volume 27A, number 8
PHYSICS
LETTERS
The samples were electropolished immediately before the measurements. The variation of the magnetisation with external magnetic field in the sheath region found experimentally for an indlum - 4.3 at. ‘j&lead sample containing flux trapped in the inner hole is shown in part (B) of fig. 1. The expected peak is clearly seen. The field extent of the peak is about 10 Oe suggesting that the critical current of the inner surface is of the order of 10 A in this field region. The slope of the diamagnetic in the sheath state is of the order of one quarter of that for the sample in the Meissner state. This also is expected since the volume of the central hole is one quarter of the total volume for this sample. The peak was not always seen in the measurements. But when a sample did show the effect then it appeared at all temperatures (and also in subsequent runs if the sample surface was not changed). It is rather difficult to polish the inner surface. The most likely reason why the peak did not appear for some samples is that the inner surface of these was too deformed to carry an appreciable current. It is possibly for this reason that earlier workers failed to see the effect. Another possible reason for this is that the sensitivity of the magnetometers used previously was not high enough since, on the scale of the whole magnetisation curve, the effect is rather small.
PARA-ORTHO
TRANSITIONS
9 September 1968
We conclude from the experimental results that a surface sheath can exist on the inner surface of a superconducting cylinder in the sheath state and that this sheath can carry current. This conclusion supports the suggestion made by Williamson and Furdyna [2] that surface superconductivity can nucleate within a sample at, for example, grain boundaries. The author wishes to acknowledge financial support from the Science Research Council in carrying out the work reported here. Thanks are due to Dr. N. Ridley and Mr. Gil West for help in preparing the samples and to members of the Low Temperature Group at Manchester for a number of useful discussions.
References
1. B.Bertman and D.G. Schweitzer, Phys. Letters 21 (1966) 361; D.G.Schweitzer and M.Garber, Phys. Rev. 160 (1967) 348. 2. S. J. Williamson and J.K. Furdyna, Phys. Letters 21 (1966) 376.
IN SOLID
HYDROGEN
W. SCHOTT Physik-Department,
Technische Hochschule Miinchen, Germany
H. RIETSCHEL and W. GLASER Institut jW Angewandte Kernpkysik,
Kernforschungsaentrum
Karlsruhe,
Germany
Received 30 July 1968
For solid hydrogen at 4.4’K the rotational transition para-ortho (0 - 1) has been measured by inelastic neutron scattering. The linewidth caused by quadrupole-quadrupole interaction has been determined. A good agreement with the theoretical results of Elliott et al. [l] was found.
Above the X-point of the specific heat (- 1.5’K), solid hydrogen is forming a hcp molecular lattice in which the molecules rotate al566
most freely. This is shown by the weak broadening of the 0 - 1 and 1 - 0 transitions essentially caused by the quadrupole-quadrupole interaction of the ortho molecules.
PHYSICS
Volume 27A, number 8
LETTERS
9 September 1968
experiments on solid hydrogen of 4.40K with an incident neutron energy of 21.8 meV. The energy
of the scattered neutrons has been measured by time-of-flight technique. Thereby the linewidth of the 0 - 1 transition could be determined with good accuracy. Fig. 1 a) and b) show (two time-of-flight spectra for ortho concentrations c = 0.68 f 0.06 and c = 0.265 f 0.05 respectively at a scattering angle of 55.2’K. The measured points of the 0 - 1 peak fitted with a gaussian distribution are given in fig. 1 c) and d), after the background was subtracted. The dotted lines are the physical ones where the instrumental resolution has been folded out. For the physical linewidth we got 1.18 f 0.07 meV for c = 0.68 and 0.79 f 0.07 meV for c = 0.265. This agree well with the theoretical results of Elliott et al. [l], who obtained for the mean square energy transfer
200dloo
120 . ,-
140
180 * 180
2bo channel number
.T
220
e)
= c 0.028 Q4 c(R$'), i
(1)
where Q is the quadrupole moment of the molecules and Rij the intermolecular distance. Eq. (1) gives for the linewidth 1.23 meV for c = 0.68 and 043 meV for c = 0.265. According to eq. (l), AE should not depend on the scattering angle. This was also verified by our experiments which were done at three different scattering angles for each concentration. For lower concentrations (c < 0.2), the definition of a linewidth does not make sense since the line shape is no more gaussian. Fig. 1 e) shows the calculated shape of a 0 - 1 peak which may be measured for c = 0.08 with an instrumental resolution of 1% (present resolution: 5%). A corresponding experiment is in preparation. In thjs w_aywe hope to get a good value of the ratio
1 -- 2.0 Y)
meV
Q2/R$
Fig. 1. a, b) Time-of-flight spectrum for c = 0.68 and c = 0.265 respectively. c, d) The fitted experimental points of the 0 - 1 peak of the above spectra. e) Calculated line shape for an experiment with c = 0.08 and AE/E = 0.01. On the rotating crystal spectrometer in Karlsruhe [2]we have performed neutron scattering
References 1. Elliott et al., Proc. Phys. Sot. SO (1967) 671. 2. W.GlBser, Nukleonik 7 (1965) 64.
*****
567
PHYSICS
LETTERS
The samples were electropolished immediately before the measurements. The variation of the magnetisation with external magnetic field in the sheath region found experimentally for an indlum - 4.3 at. ‘j&lead sample containing flux trapped in the inner hole is shown in part (B) of fig. 1. The expected peak is clearly seen. The field extent of the peak is about 10 Oe suggesting that the critical current of the inner surface is of the order of 10 A in this field region. The slope of the diamagnetic in the sheath state is of the order of one quarter of that for the sample in the Meissner state. This also is expected since the volume of the central hole is one quarter of the total volume for this sample. The peak was not always seen in the measurements. But when a sample did show the effect then it appeared at all temperatures (and also in subsequent runs if the sample surface was not changed). It is rather difficult to polish the inner surface. The most likely reason why the peak did not appear for some samples is that the inner surface of these was too deformed to carry an appreciable current. It is possibly for this reason that earlier workers failed to see the effect. Another possible reason for this is that the sensitivity of the magnetometers used previously was not high enough since, on the scale of the whole magnetisation curve, the effect is rather small.
PARA-ORTHO
TRANSITIONS
9 September 1968
We conclude from the experimental results that a surface sheath can exist on the inner surface of a superconducting cylinder in the sheath state and that this sheath can carry current. This conclusion supports the suggestion made by Williamson and Furdyna [2] that surface superconductivity can nucleate within a sample at, for example, grain boundaries. The author wishes to acknowledge financial support from the Science Research Council in carrying out the work reported here. Thanks are due to Dr. N. Ridley and Mr. Gil West for help in preparing the samples and to members of the Low Temperature Group at Manchester for a number of useful discussions.
References
1. B.Bertman and D.G. Schweitzer, Phys. Letters 21 (1966) 361; D.G.Schweitzer and M.Garber, Phys. Rev. 160 (1967) 348. 2. S. J. Williamson and J.K. Furdyna, Phys. Letters 21 (1966) 376.
IN SOLID
HYDROGEN
W. SCHOTT Physik-Department,
Technische Hochschule Miinchen, Germany
H. RIETSCHEL and W. GLASER Institut jW Angewandte Kernpkysik,
Kernforschungsaentrum
Karlsruhe,
Germany
Received 30 July 1968
For solid hydrogen at 4.4’K the rotational transition para-ortho (0 - 1) has been measured by inelastic neutron scattering. The linewidth caused by quadrupole-quadrupole interaction has been determined. A good agreement with the theoretical results of Elliott et al. [l] was found.
Above the X-point of the specific heat (- 1.5’K), solid hydrogen is forming a hcp molecular lattice in which the molecules rotate al566
most freely. This is shown by the weak broadening of the 0 - 1 and 1 - 0 transitions essentially caused by the quadrupole-quadrupole interaction of the ortho molecules.
PHYSICS
Volume 27A, number 8
LETTERS
9 September 1968
experiments on solid hydrogen of 4.40K with an incident neutron energy of 21.8 meV. The energy
of the scattered neutrons has been measured by time-of-flight technique. Thereby the linewidth of the 0 - 1 transition could be determined with good accuracy. Fig. 1 a) and b) show (two time-of-flight spectra for ortho concentrations c = 0.68 f 0.06 and c = 0.265 f 0.05 respectively at a scattering angle of 55.2’K. The measured points of the 0 - 1 peak fitted with a gaussian distribution are given in fig. 1 c) and d), after the background was subtracted. The dotted lines are the physical ones where the instrumental resolution has been folded out. For the physical linewidth we got 1.18 f 0.07 meV for c = 0.68 and 0.79 f 0.07 meV for c = 0.265. This agree well with the theoretical results of Elliott et al. [l], who obtained for the mean square energy transfer
200dloo
120 . ,-
140
180 * 180
2bo channel number
.T
220
e)
= c 0.028 Q4 c(R$'), i
(1)
where Q is the quadrupole moment of the molecules and Rij the intermolecular distance. Eq. (1) gives for the linewidth 1.23 meV for c = 0.68 and 043 meV for c = 0.265. According to eq. (l), AE should not depend on the scattering angle. This was also verified by our experiments which were done at three different scattering angles for each concentration. For lower concentrations (c < 0.2), the definition of a linewidth does not make sense since the line shape is no more gaussian. Fig. 1 e) shows the calculated shape of a 0 - 1 peak which may be measured for c = 0.08 with an instrumental resolution of 1% (present resolution: 5%). A corresponding experiment is in preparation. In thjs w_aywe hope to get a good value of the ratio
1 -- 2.0 Y)
meV
Q2/R$
Fig. 1. a, b) Time-of-flight spectrum for c = 0.68 and c = 0.265 respectively. c, d) The fitted experimental points of the 0 - 1 peak of the above spectra. e) Calculated line shape for an experiment with c = 0.08 and AE/E = 0.01. On the rotating crystal spectrometer in Karlsruhe [2]we have performed neutron scattering
References 1. Elliott et al., Proc. Phys. Sot. SO (1967) 671. 2. W.GlBser, Nukleonik 7 (1965) 64.
*****
567