Baking the System


After the vacuum system is assembled and integrated onto the baking station we can begin the baking process. The system is first pumped down as far as it will go at room temperature. This is accomplished by first using the sorption pump to get the system down to the milliTorr range of temperatures. The sorption pump contains a zeolite which when cooled absorbs molecules from the atmosphere. Conversely, when it is heated back up it releases what it absorbed. We cool the pump with liquid nitrogen until it will not absorb anymore, then close the valve between the pump and the rest of the vacuum system. To reuse the pump we open a small pipe that releases the molecules that were absorbed. Here is a picture of the pump in action.
When the system reaches the milliTorr range we can start the large ion pump. This pump takes a much longer time period to pump the vacuum system down to lower pressures, but it is clean and quiet so we put up with that small flaw. We will leave this pump to work on the system for about a week. Once it is no longer dropping in pressure we will begin to bake the system. Baking the system causes all the crud that is attached to the inside of the vacuum system to come off the walls much faster than if we just left the pump to work on its own. Baking takes another few days to complete.
	In order to bake the system we need to enclose and insulate the part of the system to be baked. We do this by building a brick enclosure. Since the temperature is not going to exceed 200 degrees Celsius we can use a standard firebrick. There are no problems with stability, as a matter of fact they are more stable than high-temperature firebrick because they weigh more. We line the inside of the oven with aluminum foil and seal the gaps with high temperature foil tape. The oven is then covered with many layers of aluminum foil to complete the enclosure.

Inside of the oven we placed strip heaters on little aluminum stands. We chose strip heaters because of the easy placement, the sealed design, and their cost-effectiveness. Power cables made of high temperature wire with high temperature ring terminal connectors were fed through the bricks and attached to the heaters. They were powered by a variable transformer which allows for easy control of the temperature. We also attached a series of thermocouples at various points on the vacuum system body to monitor the temperature and evenness of the heating.

The temperature of the oven was brought up slowly over a period of two days. The final temperature was 200 degrees Celsius. The temperature was brought up slowly to avoid overloading the ion pump. Once I check my lab notebook I will add a few more details on how the baking went.

The Ion Pump - Don't try UHV without it

When the pressure in the vacuum system has been reduced to around 10^-3 Torr by the sorption pump, the ion pump then comes into play. The ion pump, like the sorption pump, has no moving parts and therefore no oils or other lubricants. This pump is therefore as clean as the sorption pump and perfect for UHV. Also, as the pressure drops in the system, so does the current. This has two benefits. One is that the pump draws very little power and is therefore very economical. The other is that since the current drop is related to the pressure it can be used as a pressure gauge, thus eliminating the need for a separate pressure gauge. This is both economical and allows for a simpler system. An additional gauge would require more parts and increase the volume of the vacuum system which we would like to avoid.

Ion Pump

The pump is started by applying high voltage between the tube shaped anode and the cathode of the ion pump. Electrons are accelerated toward the positive anode and are forced to follow a spiral path in the tube shaped anode because of the magnetic field. This has the effect of sweeping out more space and increasing the probability that an electron will collide with a gas molecule. The positive ions that are formed in the collisions strike the chemically active titanium cathode "getter" plate. The ions combine with the cathode material and eject more cathode material which ends up on the surface of the anode. This constantly replenished the film of chemically active cathode material on the anodes which combines with active gas molecules and effectively pumps them from the system. This process of removing chemically active gasses such as Nitrogen, Oxygen, and Hydrogen is called "gettering".

Bisic idea of inside of ion pump spiral path of electrons and ions
Inert gasses are handled a little differently. They are buried in the pump surfaces. This happens when they are ionized and hurled into the cathode. They penetrate a few layers and bury themselves in the cathode lattice structure. They can be re-emitted when other ions strike the surface so they tend to collect where there is little of this "sputtering" going on.


The first step in building a UHV system is to design it...and Dr. Feng took care of that. Here is a rough drawing of the layout of the vacuum system. I left out the detail on the chamber (i.e. viewports and such) because they confuse the drawing.

You may notice that this design includes three different pumps. The sorption pump and two different size ion pumps. These pumps were chosen first and foremost for their clean operation. There is no oil to grime up the system. That is of paramount importance in a UHV system. The cleaner it is, the lower the pressure will get. The reason there are three of them is that the system is built on a baking station and will be broken apart later on. The sorption is the roughing pump, taking the system down to the milliTorr range. When the sorption pump has been properly prepared it can easily pump down to 4-5 milliTorr. Once we reach this low pressure we are able to start the large ion pump. The "large" ion pump is a 20 L/s pump, the "small" ion pump is an 11 L/s pump. There are only large and small relative to each other, you can get much larger and much smaller pumps. The large ion pump will take the pressure down to the 10^-8 or 10^-9 Torr pressure range. Once the upper protion of the system is "broken" from the lower portion at the middle valve, the small ion pump will maintain that pressure and hopefully even take it down a little further. But enough about the pumps.
When assembling the UHV system the utmost care needs to be taken. The pieces must be cleaned extremely well. Our procedure for cleaning each part is the following: Hot Oakite bath in an Ultra-Sonic cleaner Hot water rinse Boiling distilled water bath Wipe with Acetone with special clean wipes If necessary, run through Ultra-Sonic cleaner in Acetone bath Wipe with Methanol with special clean wipes Store in enclosed cabinet to keep off dust Cover each item in the cabinet for added protection This procedure has worked well to this point. All new pieces were cleaned. The packages say that they are clean and ready for use right out of the package, but I think you would be surprised at how much came off of those "clean" new pieces in the cleaning process. Before we can assemble the vacuum system we needed to construct a baking station. This is a structure made of angle iron that has an aluminum top. It holds the pump, the pump controller, and has a top that has room enough to build a brick oven on the top. The station will house the sorption pump and the large ion pump permenantly and only needs to be used to get the trapping chamber down to a pressure low enough that the small pump can maintain it. Here is a picture of the complete baking station.

Once that structure is complete the rest of the vacuum system can be assembled. This protion of the system has eight anti-reflection coated windows for the laser beams and two small viewports. Each of these windows and viewports needs to be cleaned very well and then covered so there is no dust, dirt, oil, or any other vile contamination that might get baked to the surface permenantly. The cleaning is done carefully with lens cleaning tissue and acetone and methanol. Here is what the complete system looks like. This is now ready to be baked.

The Sorption Pump - So cool its cryogenic

The Sorption pump is our roughing pump. We use a sorption pump because it has no moving parts and therefore no oils or other lubricants. This pump is therefore very clean and will introduce no contaminants to the system. This is essential when working on a ultra high vacuum system because backstreaming oil vapor from conventional pumps would muck up the system. The sorption pump is fast and economical. It can pump our system from atmospheric pressure to 10^-3 Torr in a matter of hours. It can do all this with no electricity, just about 5 liters of liquid nitrogen.

Sorption pump in action

Sorption pump with condensation

The Sorption pump evacuates gas molecules from the vacuum system by cryosorption, which is adsorbing them on a chilled surface. The surface is provided by the molecular sieve. The molecular sieve has a very high ratio of surface area to volume. The amount of molecular sieve contained in our sorption pump has on the order of one square kilometer of sorbant surface area. Therefore, a relatively small pump can pump a large amount of gas.

The Finishing touches - prop it wrap it and go

Once the system has been build and baked there is very little left to do. All that is left is to add the Rubidium to the system, break it from the baking station, and wrap the Helmholtz coils.

Adding the Rubidium to the system is not a pleasant job. In order to accomplish this task we had to break the vacuum system from the baking station and transport it to a fume hood with flowing Nitrogen. Unfortunately the fume hoods in the Physics department do not have flowing nitrogen, so we had to take it to the Chemistry department which is on the fourth floor. Once we got the system into the hood we used a bag attached to the Rubidium holder filled with nitrogen as a protective environment. We opened the small nipple that was to house the Rubidium to atmosphere. We quickly broke the Rubidium ampule, blew out the Argon with Nitrogen and put it in the nipple. We then quickly sealed the nipple again. We now have the problem of having opened a very small part of our vacuum system to atmosphere. In order to get back down to ultrahigh vacuum we had to reconnect the system to the baking station and repump the system. Fortunatly, the procedure went well and we never lost the integrity of the system.

The Rubidium Nipple

Since the nipple had been opened to the atmosphere there was a lot of crud inside there, and the ion pump could not bring the system down very fast. To solve this problem we baked the nipple with a heat gun for short intervals ranging from 10 minutes to 2 hours until it did not make a difference any more. After we did this the ion pump was able to pump the system back down to the pressure range it was at before we added the Rubidium. Once that was done that step was completed.

magnet wires on the vacuum system

Once the Rubidium has been added we can break the system from the baking station. If all goes well the vacuum system will not return to the baking station for a long time. Once on the table we mount the system so that we can wrap the coils. The holders for the coils are two aluminum grooves which clamp to the vacuum system body. The wire lays inside these grooves. Wrapping the coils is a pain, but it only requires 36 turns because the magnet wire is 18 gauge. Once the coils are wrapped and the system is mounted it is done. All that is left is to position it on the table so the laser can enter the chamber and trap the atoms.

broken system on the optics table
Here we see the vacuum system on the table. Optics man is busy preparing to wrap the coils.

Links:

Jane ; gateway battery ; best detox urine test ; help with essay


When the pressure in the vacuum system has been reduced to around 10^-3 Torr by the sorption pump, the ion pump then comes into play. The ion pump, like the sorption pump, has no moving parts and therefore no oils or other lubricants. This pump is therefore as clean as the sorption pump and perfect for UHV. Also, as the pressure drops in the system, so does the current. This has two benefits. One is that the pump draws very little power and is therefore very economical. The other is that since the current drop is related to the pressure it can be used as a pressure gauge, thus eliminating the need for a separate pressure gauge. This is both economical and allows for a simpler system. An additional gauge would require more parts and increase the volume of the vacuum system which we would like to avoid.

The pump is started by applying high voltage between the tube shaped anode and the cathode of the ion pump. Electrons are accelerated toward the positive anode and are forced to follow a spiral path in the tube shaped anode because of the magnetic field. This has the effect of sweeping out more space and increasing the probability that an electron will collide with a gas molecule. The positive ions that are formed in the collisions strike the chemically active titanium cathode "getter" plate. The ions combine with the cathode material and eject more cathode material which ends up on the surface of the anode. This constantly replenished the film of chemically active cathode material on the anodes which combines with active gas molecules and effectively pumps them from the system. This process of removing chemically active gasses such as Nitrogen, Oxygen, and Hydrogen is called "gettering".
	Inert gasses are handled a little differently. They are buried in the pump surfaces. This happens when they are ionized and hurled into the cathode. They penetrate a few layers and bury themselves in the cathode lattice structure. They can be re-emitted when other ions strike the surface so they tend to collect where there is little of this "sputtering" going on.


The Sorption pump is our roughing pump. We use a sorption pump because it has no moving parts and therefore no oils or other lubricants. This pump is therefore very clean and will introduce no contaminants to the system. This is essential when working on a ultra high vacuum system because backstreaming oil vapor from conventional pumps would muck up the system. The sorption pump is fast and economical. It can pump our system from atmospheric pressure to 10^-3 Torr in a matter of hours. It can do all this with no electricity, just about 5 liters of liquid nitrogen.
	The Sorption pump evacuates gas molecules from the vacuum system by cryosorption, which is adsorbing them on a chilled surface. The surface is provided by the molecular sieve. The molecular sieve has a very high ratio of surface area to volume. The amount of molecular sieve contained in our sorption pump has on the order of one square kilometer of sorbant surface area. Therefore, a relatively small pump can pump a large amount of gas.


Once the system has been build and baked there is very little left to do. All that is left is to add the Rubidium to the system, break it from the baking station, and wrap the Helmholtz coils.
Adding the Rubidium to the system is not a pleasant job. In order to accomplish this task we had to break the vacuum system from the baking station and transport it to a fume hood with flowing Nitrogen. Unfortunately the fume hoods in the Physics department do not have flowing nitrogen, so we had to take it to the Chemistry department which is on the fourth floor. Once we got the system into the hood we used a bag attached to the Rubidium holder filled with nitrogen as a protective environment. We opened the small nipple that was to house the Rubidium to atmosphere. We quickly broke the Rubidium ampule, blew out the Argon with Nitrogen and put it in the nipple. We then quickly sealed the nipple again. We now have the problem of having opened a very small part of our vacuum system to atmosphere. In order to get back down to ultrahigh vacuum we had to reconnect the system to the baking station and repump the system. Fortunatly, the procedure went well and we never lost the integrity of the system.
Since the nipple had been opened to the atmosphere there was a lot of crud inside there, and the ion pump could not bring the system down very fast. To solve this problem we baked the nipple with a heat gun for short intervals ranging from 10 minutes to 2 hours until it did not make a difference any more. After we did this the ion pump was able to pump the system back down to the pressure range it was at before we added the Rubidium. Once that was done that step was completed.
	Once the Rubidium has been added we can break the system from the baking station. If all goes well the vacuum system will not return to the baking station for a long time. Once on the table we mount the system so that we can wrap the coils. The holders for the coils are two aluminum grooves which clamp to the vacuum system body. The wire lays inside these grooves. Wrapping the coils is a pain, but it only requires 36 turns because the magnet wire is 18 gauge. Once the coils are wrapped and the system is mounted it is done. All that is left is to position it on the table so the laser can enter the chamber and trap the atoms.

Here we see the vacuum system on the table. Optics man is busy preparing to wrap the coils.