INSTRUCTION FOR THE
HYDROTHERMAL DIAMOND ANVIL CELL HDAC III

BILL BASSETT
765 BOSTWICK RD.
ITHACA, NY 14850

DEPARTMENT OF GEOLOGICAL SCIENCES, SNEE HALL,
CORNELL UNIVERSITY, ITHACA, NY 14853, USA
TEL. (607) 255 7502 FAX (607) 254 4780
E-mail BASSETT@GEOLOGY.CORNELL.EDU

DIAMOND ANVILS - The anvils that have been provided are 1/8 carat [actually 15 points instead of 12.5 points]. The table face (the base of each anvil) has been enlarged to >2.2 mm for additional support. The diamonds are purchased from Lazare Kaplan, 529 Fifth Avenue, New York, NY 10017 [Tel. 212-972-9700].

HEATERS - Each heater consists of a tungsten carbide core that is also the support for the diamond anvils. This core is wound with molybdenum wire (0.010 inch or 254 microns diam.) separated from the tungsten carbide core by a thin layer of ceramic cement. Ceramic cement on the wound wire holds the wire in place. NEVER RUN THE HEATERS AT TEMPERATURES OVER 300°C WITHOUT SURROUNDING THEM WITH AN INERT GAS SUCH AS ARGON OR BETTER, AN INERT GAS MIXED WITH 1% - 2% REDUCING GAS SUCH AS HYDROGEN. The cement we use is an alumina-based Zircar fibrous ceramic manufactured by Zircar Products, Inc., 110 North Main Street, Florida, NY 10921. For the windows, we use Duco cement. The lower electric connectors are purchased from Omega, P.O. Box 4047, Stamford, CT 06907 USA or P.O. Box 1, Broughton Astley, Leicestershire, LE9 6XR, ENGLAND. We have tapped one of the holes to take the screws.

THERMOCOUPLES - The two themocouples are both chromel-alumel (type K). Each is in contact with one of the diamond anvils and yields a temperature that is within 5°C of the sample temperature. This can be checked with salts of known melting temperatures. We use sodium nitrate (306.8 °C) and sodium chloride (800.5 °C). Alumel, which is magnetic is labelled with an "A" and is closest to the heater connectors. Chromel which is not magnetic is labelled with a "C". The lower connectors are made of the same metals as the wires to minimize unwanted emf at junctions. The wires leading to the upper connectors are also alumel and chromel. Because the thermocouples are at 1 bar pressure, there is no pressure correction needed. Note that there are additional holes for installing other electrical connections if desired.

ELECTRICAL CONNECTORS - The electrical connectors on this new HDAC have been designed to provide reliable electrical contacts during high-temperature runs. In our experience plugs, while more convenient, often become loose leading to instabilities. It is important to tighten the connections on the new design so that instabilities do not occur. For the upper connections two special wrenches are provided, a nut driver and an end wrench, to fit the small hex nuts. IT IS VERY IMPORTANT TO TIGHTEN THESE WITH BOTH WRENCHES AT THE SAME TIME SO THAT THE SCREW IS NOT TORQUED. There are four tabs for the lower electrical connections made of the correct metals. LIKEWISE, THE LOWER SCREWS SHOULD BE TIGHTENED ENOUGH TO MAKE GOOD CONNECTIONS BUT NOT SO MUCH AS TO STRIP THE THREADS IN THE TABS. These tabs are held in place between two sheets of teflon by friction. They can pull out. If this happens, loosen the three slotted screws on the bottom of the cell and insert the tabs being careful to get each tab on the lower side of the wire inside as it is inserted. Then tighten the three slotted screws. This ability to remove the tabs is intentional. There may be situations in which the tabs could interfere with mounting or other operations. Although the tabs are purchased from Omega, our shop has tapped the holes to take the screws.

WINDOWS - Glass or mica windows are necessary above and below the diamond anvils. Their purpose is to prevent oxidation of the diamonds by containing the inert gas. Each window is spaced away from the metal surface to provide a controlled leak. IF THESE WINDOWS ARE REPLACED, THEY MUST BE INSTALLED SO THAT THERE IS STILL A CONTROLLED LEAK. A small channel is cut in the bottom of the rocker and the circular ceramic (fired pyrophyllite) disk to allow the gas to travel around to the back side of each diamond anvil and fill the space between the anvil and the window. Glass windows, made from microscope cover slips, remain clearer than mica for optical observations but absorb more x-rays when x-ray diffraction is used. If you are using high energy x-rays 17 keV or greater, the glass is probably ok. If <17 keV maybe mica or capton should be used. These windows can be replaced with other materials such as LiF if transparency to IR is important. All windows must provide a small leak as described above. There are also two windows on the cylinder. These are easily replaced with glass cover slides if broken. They are there to permit direct observation of the heaters during a run.

SPRINGS - Posts, screws, and springs (these are the bellville type of springs that look like washers) have been color coded and marked so that they can provide equal forces at each of the three points. Marking with permanent color pens is only one way of accomplishing this. After some use, you may decide to devise some other means of calibrating such as scribed disks. In any case, it is important to have some means of preserving alignment and keeping track of the load being applied to the sample. The springs may be omitted in which case the screws will play a more important role in preserving alignment. This is an approach which may be valuable when using smaller anvil faces to higher pressures. However, the user should realize that the pressure increases very rapidly with the turning of the screws when the springs are omitted. See the section below titled "ALTERNATIVE METHOD OF APPLYING PRESSURE" for more detail.

ALIGNMENT TEST - All alignments are made as accurately as possible before shipping. To test the alignments, place a small pile of AgI on the lower anvil face and squeeze it between the anvil faces. If there is an adequate amount of sample and if the alignments are good, the phase transition at about 4 to 5 kbar should appear as a ring that is symmetrical with respect to the anvil faces as you look down through the upper diamond. If your cell has anvil faces 0.6 mm across or smaller, you should also be able to use AgI for aligning up to 100 kbar when the 100 kbar phase appears. It too should be centered. AgI also gets progressively darker with pressure and so the symmetry of the pressure distribution can be observed at pressures between 5 and 100 kbar.

ALIGNMENT PROCEDURES - The alignment procedures are very simple and rapid once they are understood. A GREAT DEAL OF CARE IS CALLED FOR, HOWEVER, BECAUSE MISALIGNMENT OR ROUGH TREATMENT CAN EASILY RESULT IN DAMAGE TO THE DIAMONDS. There are three small allen wrenches with yellow handles. These should all be inserted at the same time into the lower set of set screws for making the anvil faces parallel. With no sample between the anvils and with the upper anvil gently touching the lower one, look for interference fringes or newton’s rings. If you don’t see these, take the cell apart and run your finger over one of the anvils to leave a film of oil. Gently bring the anvils together again; the contact should now collect the oil so you can see just where the contact is. Looking at the anvils from the side may be necessary in order to judge when contact is made. Once you observe the contact point or see interference fringes, you should back off the three set screws. Slowly advance the set screw on the side where the contact or most intense fringes are observed. This should separate the anvils on that side. If necessary back the other set of screws out further. By backing out some screws and advancing others, it should be possible to spread the contact over the entire anvil area. You will know when you have achieved this because one fringe or one color will cover the entire anvil area. Once you have accomplished this, gently tighten the three set screws to hold the lower rocker in place. NEVER OVER TIGHTEN THESE SET SCREWS AS THAT WILL DEFORM THE STAINLESS STEEL ROCKER WHERE THE SCREWS CONTACT IT.

When the lower rocker has been aligned, you should move the three allen wrenches to the upper set of three set screws for translating the upper anvil. Again this should be done while the upper anvil is gently resting on the lower anvil so that the upper anvil assembly doesn’t fall out when the upper set screws are loosened. If this part of the alignment is done with AgI between the anvils, a small amount of pressure can be applied to hold the upper anvil in place. Now adjust the upper set screws until the two anvil faces are just superimposed. An intense light from the side of the diamond cell will often help to show how well matched the two faces are. NEVER OVER TIGHTEN THESE SET SCREWS AS THAT WILL INDENT THE STAINLESS STEEL RIM AND CRACK THE CERAMIC DISK AT THE PRESSURE POINTS.

You should not consider the alignment procedure completed until you have exerted some pressure on an AgI sample, enough to spread the 4-5 kbar ring out to the edge of the anvil area and then checked the alignment again. The reason for this is that some seating may be necessary. If applying pressure results in some misalignment, you will have to go through the alignment procedure again.

ALTERNATIVE METHOD OF APPLYING PRESSURE - The shorter screws can be used without the springs for better alignment during operation. The use of springs increases the sensitivity, but can allow the anvils to misalign under pressure. Because the pressure increases rapidly with turns of the screws, omitting the springs should be done only after the operator has experience and feels comfortable with the procedures. Then more care than usual should be given to advancing the screws by small increments.

COOLING - There is generally no need for cooling during high-temperature runs. If, however, the expansion of parts of the HDAC cause changes in pressure or other undesirable effects, then air cooling may be needed. Three short stainless steel extenders for the hollow posts have been provided. These can be screwed into the threaded holes on the tops of the posts. Flexible tubing with 1/4 inch inside diameter can then be slipped over one of these to allow the user to pass compressed air through the cell for air cooling. There is no need to use much pressure for this purpose. In fact, too much pressure could result in displacing the inert/reducing gas with air causing unwanted oxidation of the hot parts. If there is additional concern about keeping key parts cool, screws with holes can be used so that the air passes through each screw providing even better cooling. Screws with holes are not provided but can be supplied if needed.

For temperatures down to -200°C, liquid nitrogen can be introduced into the base by means of the same extenders described above, or it can be introduced directly into the chamber surrounding the anvils using the tube attached to the container cylinder where the inert/reducing gas is introduced for high-temperature runs. Temperatures as low as -200°C can be achieved with either approach. Reheating with the electric heaters can provide good temperature control.

HEATING - One of the simplest ways to heat the anvils is to connect the two heaters in series and then connect them to a variable transformer. The specs are typically 1-10 volts & 1-10 amps with meltdown at 10 v & 10 A. Hand control of the variable transformer can be very satisfactory. In order to make the full scale of the variable transformer 20 volts, it may be desirable to place a step down transformer ahead of it in the circuit. This can also be a variable transformer. Both transformers need to be rated at 10 amps or better. An isolation step-down transformer is safer. However, a variable (auto transformer) can be used for the step-down transformer if the user is careful to insure proper grounding. Usually this means no more than being sure to use equipment with three prong pugs or measuring voltage against ground at all points along the circuit.

TEMPERATURE CONTROL - There are several approaches to controlling the temperature. If balance between the temperatures of the upper and lower anvils is important, variable resistors with a resistance of a few ohms can be placed across the two heaters and adjusted to bring the heater temperatures into agreement. These variable resistors should be rated for several watts. Another approach is to have a separate temperature controller for each heater using that heater’s thermocouple for input into the controller. A commercial temperature controller is most useful for holding temperatures for long times. It is less necessary for controlled raising and lowering of the temperature. That is best done by hand. Commercial controllers are usually designed for heater of much greater thermal mass and as a result can run away and cause melt-down. IF YOU USE COMMERCIAL TEMPERATURE CONTROLLERS, SET THE SAFETY LIMIT FIRST.

TAKING THE HDAC APART IN CASE OF TOO MUCH FRICTION - Sometimes after a high temperature run the cell binds, and it is very difficult to separate the upper and lower platens. This is especially likely to happen if a lubricant has been used on the posts. If this happens, remove the three driver screws and springs. Install the three longer driver screws. But, after passing them through the holes in the upper platen and before screwing them into the base, screw them through the three nuts and then screw them into the base. Now spin the three nut upward until they press against the upper platen. With an end wrench or crescent wrench turn the nuts in sequence so as to lift the upper platen. It may be necessary to hold each screw with the large allen wrench so that it doesn't turn at the same time. This will generally succeed in lifting the upper platen off the posts