- Email: [email protected]
Adjustable temperature cryostat for optical studies
ADJUSTABLE TEMPERATURE CRYOSTAT FOR OPTICAL STUDIES M. BENARROCHE, J.-P. DANOY, and P. PESTEIL Laboratoire de Physique P.C.B.I, Facultd des Sciences, Marseille, France Received 8 Jul}' 1966
T H E importance of the temperature parameter for the geometry and position of spectral bands is well known. While the spectra of gases or diluted vapours consist of fine bands, this is no longer true for condensed media in which environmental effects multiply the levels. In the case of substances crystallized at relatively high temperatures the molecules have no rigidly fixed position or orientation and the resulting crystal field fluctuations cause broadening of the levels observed; the bands corresponding to the different levels are broad and overlap so that resolution is very low. The resolution can be improved by lowering the temperature as the molecules then become increasingly fixed in the lattice, the crystal field fluctuates less, and the environmental factor becomes more clearly defined. The temperature is generally lowered by submerging the sample in a cryogenic liquid which is transparent to light in the spectral region studied. 1 This leaves usually three temperatures available, viz.: helium, hydrogen, and nitrogen boiling points. These temperatures can be slightly varied by reducing the pressure above the liquid, but crystallization of hydrogen and nitrogen sets a limit in their case. However, while lowering the temperature improves resolution it also affects the position of the energy levels as well as the geometry of the spectral lines. Hence a study of the variation of these two parameters as a function of temperature is of considerable interest. As regards aromatic crystals, spectra obtained at 20"4 ° K have generally good resolution and are very similar to those obtained at 4.2 ° K. 2 If, however, the samples are submerged in liquid nitrogen (77 ° K), resolution will often be hardly better than at room temperature. It is accordingly preferable, in this case, to study the effect of temperature on spectra in the range 20-80 ° K. A cryostat for optical studies is required for this purpose, capable of producing temperatures between 20 and 80 ° K and maintaining them constant for at least 30 min, the time required for an absorption spectrum photograph. 330
Principle An oblong enclosure is filled with a gas which is not liquefiable within the temperature range used. Two heat sources, one cold, one hot, are installed at the two ends of the enclosure. Owing to the resulting temperature gradient the sample mounted between the two sources assumes a temperature intermediate between that of the cold and the hot source. The cold source consists of a hollow cylinder containing the cryogenic liquid. The hot source is an electrically heated plate. The temperature gradient (and hence the temperature of the sample) can thus be varied by modifying the supply of electric energy to the hot plate.
Description of the cryostat Figure 1 is a distorted diagram of the cryostat. For greater clarity the horizontal dimensions are shown about 3 times larger than the vertical. The cryostat PI
\
/
/
1
t v/
I
I
/,'."-",'.'l.'//.[/l('l
I Figure
1111
1
CRYOGENICS
•
DECEMBER
1966
unit is a cylinder of 140 mm diameter and 80 cm height. A tank of about 1-6 I. capacity is lagged by the method of Pesteil and Philip: 1 three nickel silver cylinders are nested to form three annular spaces at, a2, a3.Annuli at and a3 intercommunicate and are empty (preliminary exhaustion by valve V); a2 is filled by liquid nitrogen (protective refrigerant). The liquid hydrogen is m the central copper cylinder. This cylinder can be detached from the rest of the cryostat and is rendered tight during operations by a ring joint J fastened between two brass plates Pt and P2. Three silica windows F~, F2, F3, 30 mm across, are mounted by similar ring joints on parts welded to the outside cylinder to permit optical absorption and luminescence studies (observation at right-angles to the exciting light). A tube T passing through the central container puts the enclosure E into communication with the upper part of that container; the enclosure is thus kept under hydrogen atmosphere because the liquid hydrogen vaporizes continuously. Tube T also contains several electric wires for the heating and for temperature measurements. A continuation of it, tube T' passes through the central container and plate P1 and brings the wires up to the tight intake M. RI is an Allan Bradley carbon resistor of 100 at room temperature. Its value rises suddenly on contact with the liquid hydrogen level when the cryostat is being refilled; this indicates that the cryostat is full. The hollow cylinder C containing liquid hydrogen is the cold source of enclosure E. The enclosure is shown in more detail in Figure 2. It consists of a cylinder of about 16 cm height and 39 mm outside diameter. The bottom (up to the level of sample S and windows F) is of silica. The remaining parts of the cylinder are Pyrex-silica and Pyrex-metal joints; the upper portion is of machined brass screwed to the bottom portion of the liquid hydrogen container; it is rendered tight by a Wood's fusible alloy weld W which permits easy access to the sample S. A hollowed-out plate 1 of insulating material is shown in profile in Figure 1, full face in Figure 2. It comprises an object slide consisting of two hollow cylinders, screwed one into the other and clamping two silica discs between which the crystal sample S is held. The heater plate Z is an insulating disc, of 26 mm diameter, on which six carbon resistors, in symmetrical star arrangement, are connected in series. K is another insulating disc, of 31 mm diameter, acting as a heat screen for regulating the temperature at the level of the sample by channelling the convection gas currents. Temperatures are measured by a P-N germanium junction G, mounted on the horizontal diameter of sample S. It is known that when a given current I flows through such junctions the potential difference across their terminals varies almost linearly with CRYOGENICS.
D E C E M B E R 1966
temperature over a wide range. 3 The junctions used are those of the Philips OC 58 transistors with only the common-emitter or common-collector connection. Between 20 and 143° K, V varies linearly with temperature within the margin of error. For this temperature range the overall variation is of the order of 220 mV when a constant 500 laA current °
_
_
.
.~
m
~~~-~~~------_2 __~
.
-
1~ ~,}r-EL--~ ~J--Bro.ss | ~,-Lr~" ~-~Kovo.r-Pyrex connection
- Pyrex
-- Silica-Pyrex connection :~ L
t
I#~
G
j --
Silica
t Iz Figure 2 flows through the junction; the sensitivity is thus about 0"56° K, i.e. satisfactory accuracy (to within about 0.25 deg) between 20 and 200 ° K. Application
The crystal sample is pressed between two silica discs supported by the object slide S. Enclosure E is screwed to the lower part of the internal container and welded with Wood's metal. The cryostat is reassembled and evacuated. The annular protective space is then filled with liquid nitrogen and hydrogen is siphoned into the central container. If the central container is filled merely until resistor R~ above T is in contact with the liquid hydrogen (double the resistance read off the ohmmeter), the temperature in the enclosure at the level of S will progressively drop and become stable at about 27 ° K after 1 h or so if no heat is supplied from Z. However, the siphoning of hydrogen can be continued after RL has become submerged in the liquid; enclosure E will then be partially or completely filled with liquid hydrogen through tube T and the temperature of S will be 20.4 ° K, of course. After vaporization of the hydrogen 331
in the enclosure E the temperature becomes quickly stable at about 23 ° K ; in this way, if desired, an intermediate point can be obtained between 20 and 27 ° K. The heater plate Z (total resistance variable from 1 200 to 600 Q between 20 ° K and room temperature) is fed by a supply stabilized between 0 and 100 V. The power consumption is measured by. a voltammeter. ~c
/
/
80
60
I.O
/
/
/
/ I
1 I
i
1
2 (Watts)
P Figure
3
The temperature rises quickly .as soon as Z is heated, and then varies almost linearly with the power. Thermal equilibrium is attained in 3 or 4 min, the temperature then becoming constant to within better than 0.25 deg up to about 60 ° K. Beyond this, fluctuations occur owing to gas flows set up by the steep
332
thermal gradient. Towards 80 ° K the temperature is stable only to within _+ 1 deg and towards 100 ° K within + 2 deg. Figure 3 shows the temperature variations as a function of the power output from plate Z. It will be seen that for falling temperatures the function T = f ( P ) r e m a i n s linear down to about P = 0(broken curve section). The fact that the curve is almost linear indicates greater thermal conduction losses as compared with radiation losses. Notes For obtaining very constant temperatures above 80 ° K it is only necessary to replace the hydrogen in the central container by liquid nitrogen. In the same way the device described above can be used between 4 and 20 ° K if the hydrogen is replaced by liquid helium. In this ease it is advisable to use for temperature measurements an Allen Bradley carbon resistor of 100 f~ and 0.25 W at 300 ° K, with resistance variable from 200 to 1 600 fZ between 20 and 4 ° K (R2, Figure 2) instead of the P - N junction, a Power consumption by the P - N junction and carbon resistor is quite negligible (maximum 0-3 mW for the P - N junction, 0.2 laW for the carbon resistor); hence temperature measurements do not in any way affect the temperature gradient in the enclosure. Conclusion The apparatus described is of simple design and reasonable first cost. The sample to be optically examined can be kept at constant temperatures within the range 20-100 ° K. The range of application can be easily extended to 4-150 ° K by exchanging the cryogenic liquid. REFERENCES 1. PESTEIL, P., and PHILIP, R. Cryogenics 1, 53 (1960) 2. BENARROCHE, M. Actaphys. polon. 26, 355 (1964) 3. DONNINI, J. M. Rev. Phys. appI. Paris. To be published
CRYOGENICS.
D E C E M B E R t966
T H E importance of the temperature parameter for the geometry and position of spectral bands is well known. While the spectra of gases or diluted vapours consist of fine bands, this is no longer true for condensed media in which environmental effects multiply the levels. In the case of substances crystallized at relatively high temperatures the molecules have no rigidly fixed position or orientation and the resulting crystal field fluctuations cause broadening of the levels observed; the bands corresponding to the different levels are broad and overlap so that resolution is very low. The resolution can be improved by lowering the temperature as the molecules then become increasingly fixed in the lattice, the crystal field fluctuates less, and the environmental factor becomes more clearly defined. The temperature is generally lowered by submerging the sample in a cryogenic liquid which is transparent to light in the spectral region studied. 1 This leaves usually three temperatures available, viz.: helium, hydrogen, and nitrogen boiling points. These temperatures can be slightly varied by reducing the pressure above the liquid, but crystallization of hydrogen and nitrogen sets a limit in their case. However, while lowering the temperature improves resolution it also affects the position of the energy levels as well as the geometry of the spectral lines. Hence a study of the variation of these two parameters as a function of temperature is of considerable interest. As regards aromatic crystals, spectra obtained at 20"4 ° K have generally good resolution and are very similar to those obtained at 4.2 ° K. 2 If, however, the samples are submerged in liquid nitrogen (77 ° K), resolution will often be hardly better than at room temperature. It is accordingly preferable, in this case, to study the effect of temperature on spectra in the range 20-80 ° K. A cryostat for optical studies is required for this purpose, capable of producing temperatures between 20 and 80 ° K and maintaining them constant for at least 30 min, the time required for an absorption spectrum photograph. 330
Principle An oblong enclosure is filled with a gas which is not liquefiable within the temperature range used. Two heat sources, one cold, one hot, are installed at the two ends of the enclosure. Owing to the resulting temperature gradient the sample mounted between the two sources assumes a temperature intermediate between that of the cold and the hot source. The cold source consists of a hollow cylinder containing the cryogenic liquid. The hot source is an electrically heated plate. The temperature gradient (and hence the temperature of the sample) can thus be varied by modifying the supply of electric energy to the hot plate.
Description of the cryostat Figure 1 is a distorted diagram of the cryostat. For greater clarity the horizontal dimensions are shown about 3 times larger than the vertical. The cryostat PI
\
/
/
1
t v/
I
I
/,'."-",'.'l.'//.[/l('l
I Figure
1111
1
CRYOGENICS
•
DECEMBER
1966
unit is a cylinder of 140 mm diameter and 80 cm height. A tank of about 1-6 I. capacity is lagged by the method of Pesteil and Philip: 1 three nickel silver cylinders are nested to form three annular spaces at, a2, a3.Annuli at and a3 intercommunicate and are empty (preliminary exhaustion by valve V); a2 is filled by liquid nitrogen (protective refrigerant). The liquid hydrogen is m the central copper cylinder. This cylinder can be detached from the rest of the cryostat and is rendered tight during operations by a ring joint J fastened between two brass plates Pt and P2. Three silica windows F~, F2, F3, 30 mm across, are mounted by similar ring joints on parts welded to the outside cylinder to permit optical absorption and luminescence studies (observation at right-angles to the exciting light). A tube T passing through the central container puts the enclosure E into communication with the upper part of that container; the enclosure is thus kept under hydrogen atmosphere because the liquid hydrogen vaporizes continuously. Tube T also contains several electric wires for the heating and for temperature measurements. A continuation of it, tube T' passes through the central container and plate P1 and brings the wires up to the tight intake M. RI is an Allan Bradley carbon resistor of 100 at room temperature. Its value rises suddenly on contact with the liquid hydrogen level when the cryostat is being refilled; this indicates that the cryostat is full. The hollow cylinder C containing liquid hydrogen is the cold source of enclosure E. The enclosure is shown in more detail in Figure 2. It consists of a cylinder of about 16 cm height and 39 mm outside diameter. The bottom (up to the level of sample S and windows F) is of silica. The remaining parts of the cylinder are Pyrex-silica and Pyrex-metal joints; the upper portion is of machined brass screwed to the bottom portion of the liquid hydrogen container; it is rendered tight by a Wood's fusible alloy weld W which permits easy access to the sample S. A hollowed-out plate 1 of insulating material is shown in profile in Figure 1, full face in Figure 2. It comprises an object slide consisting of two hollow cylinders, screwed one into the other and clamping two silica discs between which the crystal sample S is held. The heater plate Z is an insulating disc, of 26 mm diameter, on which six carbon resistors, in symmetrical star arrangement, are connected in series. K is another insulating disc, of 31 mm diameter, acting as a heat screen for regulating the temperature at the level of the sample by channelling the convection gas currents. Temperatures are measured by a P-N germanium junction G, mounted on the horizontal diameter of sample S. It is known that when a given current I flows through such junctions the potential difference across their terminals varies almost linearly with CRYOGENICS.
D E C E M B E R 1966
temperature over a wide range. 3 The junctions used are those of the Philips OC 58 transistors with only the common-emitter or common-collector connection. Between 20 and 143° K, V varies linearly with temperature within the margin of error. For this temperature range the overall variation is of the order of 220 mV when a constant 500 laA current °
_
_
.
.~
m
~~~-~~~------_2 __~
.
-
1~ ~,}r-EL--~ ~J--Bro.ss | ~,-Lr~" ~-~Kovo.r-Pyrex connection
- Pyrex
-- Silica-Pyrex connection :~ L
t
I#~
G
j --
Silica
t Iz Figure 2 flows through the junction; the sensitivity is thus about 0"56° K, i.e. satisfactory accuracy (to within about 0.25 deg) between 20 and 200 ° K. Application
The crystal sample is pressed between two silica discs supported by the object slide S. Enclosure E is screwed to the lower part of the internal container and welded with Wood's metal. The cryostat is reassembled and evacuated. The annular protective space is then filled with liquid nitrogen and hydrogen is siphoned into the central container. If the central container is filled merely until resistor R~ above T is in contact with the liquid hydrogen (double the resistance read off the ohmmeter), the temperature in the enclosure at the level of S will progressively drop and become stable at about 27 ° K after 1 h or so if no heat is supplied from Z. However, the siphoning of hydrogen can be continued after RL has become submerged in the liquid; enclosure E will then be partially or completely filled with liquid hydrogen through tube T and the temperature of S will be 20.4 ° K, of course. After vaporization of the hydrogen 331
in the enclosure E the temperature becomes quickly stable at about 23 ° K ; in this way, if desired, an intermediate point can be obtained between 20 and 27 ° K. The heater plate Z (total resistance variable from 1 200 to 600 Q between 20 ° K and room temperature) is fed by a supply stabilized between 0 and 100 V. The power consumption is measured by. a voltammeter. ~c
/
/
80
60
I.O
/
/
/
/ I
1 I
i
1
2 (Watts)
P Figure
3
The temperature rises quickly .as soon as Z is heated, and then varies almost linearly with the power. Thermal equilibrium is attained in 3 or 4 min, the temperature then becoming constant to within better than 0.25 deg up to about 60 ° K. Beyond this, fluctuations occur owing to gas flows set up by the steep
332
thermal gradient. Towards 80 ° K the temperature is stable only to within _+ 1 deg and towards 100 ° K within + 2 deg. Figure 3 shows the temperature variations as a function of the power output from plate Z. It will be seen that for falling temperatures the function T = f ( P ) r e m a i n s linear down to about P = 0(broken curve section). The fact that the curve is almost linear indicates greater thermal conduction losses as compared with radiation losses. Notes For obtaining very constant temperatures above 80 ° K it is only necessary to replace the hydrogen in the central container by liquid nitrogen. In the same way the device described above can be used between 4 and 20 ° K if the hydrogen is replaced by liquid helium. In this ease it is advisable to use for temperature measurements an Allen Bradley carbon resistor of 100 f~ and 0.25 W at 300 ° K, with resistance variable from 200 to 1 600 fZ between 20 and 4 ° K (R2, Figure 2) instead of the P - N junction, a Power consumption by the P - N junction and carbon resistor is quite negligible (maximum 0-3 mW for the P - N junction, 0.2 laW for the carbon resistor); hence temperature measurements do not in any way affect the temperature gradient in the enclosure. Conclusion The apparatus described is of simple design and reasonable first cost. The sample to be optically examined can be kept at constant temperatures within the range 20-100 ° K. The range of application can be easily extended to 4-150 ° K by exchanging the cryogenic liquid. REFERENCES 1. PESTEIL, P., and PHILIP, R. Cryogenics 1, 53 (1960) 2. BENARROCHE, M. Actaphys. polon. 26, 355 (1964) 3. DONNINI, J. M. Rev. Phys. appI. Paris. To be published
CRYOGENICS.
D E C E M B E R t966