Low temperature environmental cell for JEOL 200A

Home

Search

Collections

Journals

About

Contact us

My IOPscience

An electron-microscope helium stage for use with a side-entry window-type environmental cell

This content has been downloaded from IOPscience. Please scroll down to see the full text. 1984 J. Phys. E: Sci. Instrum. 17 228 (http://iopscience.iop.org/0022-3735/17/3/015) View the table of contents for this issue, or go to the journal homepage for more

Download details: IP Address: 128.180.54.183 This content was downloaded on 18/02/2015 at 13:41

Please note that terms and conditions apply.

J. Phys. E: Sci. Instrum., Vol. 17. 1984. Printed in Great Britain

An electron-microscope helium stage for use with a side-entry window-type environmental cell R J Keyset, G Raynerd and J A Venables School of Mathematical and Physical Sciences, University of Sussex, Brighton B N l 9QH, Sussex, UK

higher pressures. and to retain the greater flexibility of side-entry holders. including the tilt about the rod axis, and the ease of sample exchange. A preliminary account of this stage prior to making incremental improvements and extensive operating experience, is given by Tatlock el a1 (1980). The stage has now been in continuous use for more than two years, and studies of solid Xe, Kr (Keyse and Venables 1984), SF6 (Raynerd et a / 1982) and Nz have been made. This paper is organised as follows. In 0 2 the cooling system and environmental cell are described. The performance of the cell and the windows are assessed in 0 3, and in 5 4 some examples of work carried out with the present system are shown. Section 5 concludes with a discussion of the flexibility of the system and future possibilities.

Received 10 August 1983, in final form 7 October 1983

Abstract. A liquid-helium stage with a window-type environmental stage has been developed for use in a 200 kV transmission electron microscope (JEOL 200A). The gas cell is a single tilt device which operates at relatively high pressure (G200 Torr, i.e. -26 kPa) which can be cooled to G20 K or heated to 100 OC. A radiation shield at 50 K provides contamination protection, and cools the specimen holder to 140 K. Specimen exchange at low temperatures and tilts of t 3 0 " are possible. The stage and its performance are described, and some results on condensed gases are presented. The possibility of including side-entry holders other than the environmental cell is discussed.

1. Introduction The condensation of gases into crystals, within a transmission electron microscope, and the study of crystal structure, defect structure and phase changes of these crystals. has been undertaken in our group over a number of years (Venables et a / 1974, Venables and English 1974. Kramer 1976, Venables and Smith 1977. Raynerd et a1 1982). The main experimental need is to grow the crystals over a range of temperatures. at pressures much above the operating pressure of the microscope. This last requirement arises because simple molecular solids sublime at microscope pressures at temperatures much lower than their melting point. Consequently. unless an environmental cell is used. highly defective crystals are produced. which cannot be annealed in a metallurgical sense. Some of these requirements are discussed by Venables and English (1971) and Tatlock (1983). Previous designs of environmental cells for use at low temperatures to study condensed gases have all been of the window type. This allows a fixed volume of gas to be used, which is isolated from the surrounding cold shields. The disadvantages of the window system include the need for a periodic renewal of the windows due to rupture or severe buildup of contamination. For a description of the development of low-temperature electron microscopy and the use of environmental cells, we refer the reader to the relevant sections of reviews by Butler (1979) or Butler and Hale (198 1). This paper describes a stage and environmental cell designed for a 200 kV transmission electron microscope (JEOL ZOOA). The design evolved for the 200 kV machine to obtain greater sample penetration than in the previous 100 kV designs (Venables et a/ 1968, English and Venables 1972). to allow

? Present Address: British Steel Corporation, Sheffield Laboratoires, HVEM Building, Moorgate, Rotherham S60 3AR, South Yorkshire, UK 0022-3735/84/030228 + 06 $02.25 0 1984 The Institute of Physics

2. Detailed description of the cooling system and environmental cell The cooling system is a helium-based circuit, centred on a cryostat of liquid, based on the ideas of Klipping and coworkers (Heide and Urban 1972). The arrangement is shown as a schematic diagram in figure 1. The cryostat is shielded by the exhaust gases and the liquid-He level is servo-controlled and fed by a double-walled transfer tube from a 301 dewar. Figure 1 shows the independent pumping arrangements for the specimen cooling, shield cooling, and boil-off circuits. The cryostat is located behind the objective pumping port of the microscope. The specimen cooling pipes enter through the upper polepiece (which was bored out to 25 mm diameter) and hold the sample cooling block in position.

Trans% tu3e

.._..

Sample cooling blOCK

33 I dekar Retuhn line

Figure 1. Schematic diagram of the liquid-helium cooling system.

In these next two paragraphs the letters refer to figure 2 . which shows the arrangement inside the objective lens schematically. Photographs of the shields and stage are shown in figure 3. Helium is drawn from the cryostat through thinwalled stainless steel tubing to the top of the objective lens. The shields are cooled by a copper pipe (Z) soldered around the end of the pipe duct (R). The shields consist of a complete cylinder of tellurium-copper passing down the upper polepiece (U) coupled to a specially shaped can (B) which is held fairly rigidly against the objective lens spacer (V) with nylon screws. The pipe duct (R) is supported and insulated by three quartz rods (Q) which are held in a flange (W) screwed into the upper polepiece. The specimen cooling block (D) is supported by pipes which pass inside the cylindrical cold shield; the pipes at this stage are

Electron microscope liquid-helium stage

Figure 2. Cross section of cooling block and specimen clamping mechanism. See text for discussion. Scale in mm.

narrow bore (0.5 mm inside diameter). Heat transfer takes place through the walls of a copper tube soldered to the cooling block. All stainless steel to copper joints were hard-soldered to prevent cracking on thermal cycling. Figure 2 does not show the cooling pipes explicitly, but they were coiled around the block to allow for rotation and translation as seen (0)in figure 3(c). The environmental cell enters through a hole in the cold shield (B) and is clamped into the cooling block by a screw (S). This is operated from the outside of the microscope by a cam-gear arrangement which engages with the gear wheel plus tube (I). This is housed in a PTFE bushing (X) which is springloaded (Y) against a top-plate (M). Some of these features can be seen in figure 3. The environmental cell is shown schematically in figure 4; the holder shown in figure 4(a) fits into a rod which is compatible with the standard JEOL goniometer. Two gas pipes (G) and several copper wires are brought down the centre of the rod and clamped at room temperature. The wires are connected to chromel-constantan thermocouples, a Si-diode thermometer (T) and two constantan 1 W heaters (F), one on the sample, and one on a clamp (C) at an intermediate temperature. The two clamps are separated by a thin stainless steel band (H) which is soldered to the tellurium-copper holder. The holder is thermally insulated from the specimen translation rod by a quartz rod (Q) at the holder tip. The specimen holder is loosely coupled to the cold shields by two copper straps and a self-sprung clamp seen (P) in figure 3(b). The intermediate temperature reached has been measured to be 5140 K using a chromekonstantan thermocouple on the clamp (C). The environmental cell body (A), shown in cross section in figure 4(b), is supported by two stainless steel pipes (E) (1.5 x 1.O mm diameter) soft-soldered to the sides of the body, whose other end is supported by a quartz tube (Q) under compression. The gas carrying pipes (G) (0.8 x 0.5 mm

Figure 3. View of the cooling stage. See text for discussion.

diameter) are brought through the support pipes, and are arranged so as not to come into thermal contact with them. A short length of these pipes at the coldest temperature is needed to avoid blockages, to ensure that crystal growth takes place inside the cell and that pressure measurements are meaningful. The Si diode (T) is soft-soldered to the cell body with Wood’s metal, and is as close as possible to the cell itself. The cell body is made from tellurium-copper, the cell diameter being 3 mm. The top and bottom window are held 0.75 mm apart with the bottom window on the central (tilt) axis. The volume of the cell plus gas pipes is small (35 mm3), but the pressure is measured with an external capacitance gauge (MKS Baratron), with the internal volume of 1000 mm3 acting as a reservoir. This gas supply line is connected to a conventional gas handling and pumping system.

3. Cell and window performance The cell described above obtains temperatures TKE

sleeges ( N :

Irdium rings I L ) ?

Gss pi?e ( G J

' Cocling blsck ( D )

Figure 4. Environmental cell holder: ( a ) holder assembly; ( b ) cross section of the cell. See text for discussion.

The preparation of the windows requires the greatest attention to detail. The supporting plates are 5.5 mm wide by 21 mm long, with a top plate (J) of stainless steel containing a 4 j 3 cone which locates the cell body in the cooling block (D) using the screw (S). The bottom plate (K) is of tellurium-copper which has a 30° cone angle. to allow for 130" tilt, as does the screw (S). A standard 3.05 mm diameter copper grid is used to support the top film. while a 2.3 mm diameter R2OM grid is used for the bottom. This grid has a set of 20 pm holes set on 35 ,um centres over an area of -1 mm diameter, and is made of 5 ,um thick copper. The grids were carefully soldered onto the rims of the respective cones using a water-soluble flux, so that the plates could be readily cleaned. The cell windows have to combine strength with good transparency, and in the case of the bottom film on which the sample is grown, good thermal conductivity also. The windows form the vacuum seal to the microscope, in conjunction with the top and bottom plates, which are sealed to the cell body by squashing small indium rings (L in figure 4(6)). The basic material used was pioloform? since this combines good electron transmission with sufficient strength to withstand pressure differences of several hundred Torr when laid across a 150 mesh grid. A problem with such plastics is their tendency to charge-up and rupture at low temperatures under the electron beam. This problem was eased by evaporating a thin aluminium layer on top of the plastic. The pioloform was prepared in the same manner as formvar (Kay 1967); 5 mm squares were floated onto distilled water, and picked up with the top and bottom plates so as to cover the relevant grid completely. After drying and visual inspection, around 10 nm of aluminium was evaporated. the indium rings inserted and the plates screwed into place. The bottom window forms the substrate in which the

t Pioloform is a polyvinyl formaldehyde resin. similar to. but stronger than, formvar. 230

condensed gas crystals will grow. It was found that crystals grown on low-conductivity substrates were difficult to grow as a thin film: the crystals were facetted, but were very thick (a micrometer or more) at the edges. It was realised that increased thermal conductivity was required of the substrate, the shape of the isotherms within the crystal-substrate combination being indentified as a critical factor. Experiments with graphite flakes laid onto the grid before the pioloform film were very successful, changing the crystal form in the desired manner. The cleaved graphite. in pieces up to 1 mm, 30-70 nm thick, with a large grain size, also added appreciably to the strength of the bottom window. Cells have been produced which were used daily for several days before window renewal was required. After preliminary assembly and testing. the environmental cell was inserted into the cooling block using the standard JEOL airlock, fitted with a reducing value to slow pump-down and so prevent windows rupturing. Positioning of the cell was monitored with a mirror on the axis of the tube (I in figure 2 ) ; it was clamped in place using the cam-gear-screwdriver arangements described earlier. Pumped double 0 ring systems have been installed on all the moving parts to ensure good vacuum during these operations (Takayanagi et a1 1978). After pressure testing of the cell. cooling can begin. The cryostat and shields are cooled first. Within 20 min the column vacuum is about I O - ' Torr and liquid has been detected in the cryostat. The sample cooling circuit is activated and the cooling rate regulated by the helium flow rate: the complete cooling sequence can be achieved in 30 min? although usually an intermediate temperature is used for crystal growth. The cooling system consumes approx. 1.5 1 of liquid He during rapid cooling from room temperature to 20 K. At a steady 40 K the rate is about 1 1 h-'. The selection of the working temperature is usually made by adjusting the flow rate, but fine control is available with the electrical heater. The cell temperature is stable to within 1 K over periods of an hour, thanks to the servo-controlled valve which maintains the liquid level in the cryostat (figure 1).

Electron microscope liquid-helium stage Some examples of the experiments performed with this lowtemperature stage/environmental cell combination are presented in the next section. 4. Low-temperature condensed gas experiments The environmental cell/cold stage combination has been used to grow and study crystals of Xe, Kr. N2 and SFb. The main advantage of the system was the maintenance of high pressures and stable temperatures during crystal growth and annealing procedures. Dynamic experiments were a major feature and an image intensifier/-rv camera/vTR system was regularly used. The following are examples.

4.1. Crwtal growth of krypton

The cell was imaged at low magnification so that the overall picture of events occurring inside the cell could be seen. Figure 5 shows a sequence, taken directly from the VTR, illustrating the growth of Kr at 95 K. The pressure was initially 52 Torr, and was increased by approximately 1 Torr s-I; the remaining pictures are 0.5 s apart. The growth rate of the single nucleus is rather rapid in this example (-50pm s-'), although the nucleation density (-1 mm-*) is typical of crystal growth at these high temperatures. The crystal thickness is difficult to estimate, but is too thick for 200 kV electrons to penetrate. Similar growth features were exhibited when the cell temperature was reduced through the sublimation curve at a particular reservoir pressure but the technique was not controllable. In both cases the crystals finally covered large areas of the field of view and were annealed to relatively high temperatures (typically 0.7 of the melting point). These crystals were thinned by reducing the pressure and temperature simultaneously so that the system stayed around the sublimation curve, but that net sublimation occurred at a slow rate. 4.2. Xenon ctystals and crystal defects Crystals were thinned in the way described above until around ten of the R 2 0 M grid holes (the small holes in figure 5) were

Figure 5. Videotape pictures of growth of a Kr crystal in the cell at 95 K and around 53 Torr. Times ~ ( sand ) pressures P(Torr) are (a) 0,52; (b)0.5,52.5;(c) 1.0,53;( d ) 1.5,53.5;(e) 2.0, 54;(f)2.5, 54.5.The videotape was signal-averaged for 0.4s before photographing, with an Intellect- 100 digital picture analysis system.

visible on the TV monitor. Final cooling was then made to