Blade Trap Progress

Welcome, my friend! This page is used to report newest progress of our blade trap and share ideas with each other.


Jan 18, 2020  Reload Yb 171 ion and gaining exp for loading it. Winter Vacation Start!

Jan 17, 2020  Another PMT installed.

Jan 16, 2020  Adjust the DC null point to align the RF point horizontally.

Jan 14, 2020  Add a scanning Fabry Perot cavity for locking 370nm laser's frequency, but need improvement before using it.

Jan 11, 2020  Move the second 370nm laser beam to another window and reduce the waist.

Jan 9, 2020  PMT installed.

Dec 30, 2019  SAFE!! Succeed to load 171 ions.

Dec 29, 2019  Add a second path for 370nm laser, ions seems stable with the protection beam.

Dec 23, 2019  Third try to load Yb 174 ion, succeed.

Dec 22, 2019  Second try to load Yb 174 ion, failed. (The order of DAC connection is inversed...)

Dec 21, 2019  First try to load Yb 174 ion, failed. (The postion may be wrong...)

Dec 19, 2019  All the three light are guided into the chamber, need fine tuning further...

Dec 18, 2019  Made a new helical resonator. The Q factor is 160, good enough?

Dec 11, 2019  The chamber and the imaging system is fixed to the optical table, optical system is ready to go...

Dec 10, 2019  You get a nice 370nm Laser, the engineer said.

Dec 4, 2019  Air bake-out finish finally, succeed to get a good vacuum condition.

Nov 27, 2019  Change the dirty angle valve to a new one, air bake-out restart again...

Nov 22, 2019  Failed to get a good vacuum condition, air bake-out restart...

Nov 16, 2019  Air bake-out start!

Nov 15, 2019  Optical small table set up!

Nov 14, 2019  Chamber assembled.

Nov 12, 2019  Optical table set up!

Nov 6, 2019  Oven 171 passed the test?

Nov 5, 2019  Oven 174 passed the test. Oven 171 failed the test, reassembled.

Oct 30, 2019  Oven assembled.

Oct 25, 2019  Gold wire connected with Cu wire! Holder assembling DONE.

Oct 22, 2019  Blade alignment done?

Oct 19, 2019  Air bake-out finished!

Oct 18, 2019  Insulated Cu wire attached to the holder.

Oct 9, 2019  Capacitors and gold wire attached to the blade!

Oct 9, 2019  Some parts start prebake.

Sept 22, 2019  Page created!

Active Job


370, 399, 935的光路已经初步搭好了,之后更换物镜细调。


370,399合束,注意用beam profile观察焦点处光斑的重合情况。











Explore and Buy

由国际知名专家Jarry Ma倾情撰写的《装腔宝典》 现在在本站首发。Never too late to read it!

Maverick, big brother in our lab, gave a lecture "What to do after loading ion?" on Jan 4, 2020. Worth a read!

To explore what we need to do, here is a phD thesis "Remote Entanglement of Trapped Atomic Ions" which can gives us some background for ion photon entanglement.

Click here to look at the parts list for assembling the chamber!

Here is incompelte parts list that we need...

Name Price Info Qt. Status Comment
{{}} {{part.price}} {{}} {{part.qt}} {{part.status}} {{part.comment}}

Do I miss something? Leave messages!

Todo List

There always a lot things to do:

  • Make Yb171 ion stable
  • install another PMT
  • install 935 AOM
  • Impletement a sequencer
  • lock 370's frequency
  • Increase fidelity
  • Applying 12.6GHz microwave
  • Increase qubit coherence
  • ...and much more!

Archived Work

Air Bake-out

To achieve a high vacuum condition , it is critical that materials are free of gasses, water vapor, and other contaminants. A vacuum bake-out is a process used to remove such contaminants. You may like to search it online to learn more.


Some components of the chamber can be baked-out in the air. Reducing flange, octagon vacuum chamber, conical reducer nipples and 5-way cross are baked before assembling.

Our oven can set different phases with different temperatures. Heating is divided to eight phases and each lasts 1080 minutes(the longest time we can set.). The first phase is set to 200℃ while others are set to 300℃.

Then we need to cool down the oven to room temperature to get back our chamber. Cooling is divided to four phases and each lasts 1080 minutes. The first phase is set to 40℃ while others are set to 20℃. When the temperature is near 80~90℃, we can open the oven door slightly to speed up the cooling process.

Fig 1: The chamber components are wrapped with foil. We've put some bricks around it to make it more like baking!
Fig 2: Our hot oven
Bake the Vaccum Chamber

After assembling the vacuum chamber, we should start the air bake-out. The contaminants will be pumped out by the both molecular pump and getter-ion pump outside through the angle valve during baking. (Note: StarCell ION pump can have a high pumping speed only to some specific contaminants, thus it's important the chamber has get an high vacuum condition before finish baking.)

To start the air bake-out, you should first connect the angle valve to the outside pump. Then connect the controller wire and power wire of NEG pump. You can also connect the oven pin to the outside current source in order to do degassing during the bake-out.

Slowly increase the oven temperature from room temperature to 145℃, and keep this temperature during bake-out. (Rapidly temperature change and too high temperature may break the chamber window.) The NEG pump should keep in conditioning mode (250℃ by default) during bake-out process too. At the same time, start mechanical pump, wait until it to be stable, then start molecular pump. When the chamber reaches a reasonable vacuum condition (E-8 Torr magnitude), you can start the getter-ion pump. As a rule of thumb, you can start ion pump when the oven reaches 145℃, as fig 3 does.

With the contaminants slowly pumping out, the pressure should gradually reach the pump's limit (about 1~3 E-9 Torr). It's time to activate the NEG pump. Since we will open the oven and close the angle valve the time the activation is finished, we will cool the oven temperature to 100℃ first, in case the oven temperature change too rapidly. (We forgot do it during the first activation in fig 3, thus we stop the activation and do it again after cooling down to 100℃.) You can consider to perform degassing for ytterbium oven when waiting for the hot oven cooling down. When activation finished, close the angle valve and start StarCell ION pump. Stop the NEG pump and wait the oven cooling to room temperature. The chamber pressure should reach E-11 Torr region thus finishes the baking process.

The pressure was slowly growing up from 3.4E-11 Torr to 2.2E-10 Torr after we taking out the chamber from the oven. Fortunately, after about a week, the pressure went back to 1.3E-10 Torr and became stable, which is an acceptable vacuum condition.

Fig 3: The first try of air bake-out. After start the ion pump, the pressure dropped slowly in the first day, and seemed not changed in the next few days. Typcally, the ion pump pressure can reach E-9 Torr region. Thus we decide to start cooling in advance. After the activation, the NEG current is finally stuck in 850nA, which is far from satisfied.
Fig 4: After tighting some bolts, we start the second try of air bake-out. Although we had a better start than before, but the ion pump pressure is still stuck at 2E-8 Torr. After a few days I realised I made a mistake before that I forget to clean the angle valve before assembling the chamber. Therefore we stop this bake-out and change the angle valve to a clean one.
Fig 5: The third try of air bake-out looks normal in the end. (The current data is recorded as the operation voltage is set to -5000kV.)

Wire Bonding

Task Description

The blade electrodes are connect with externel voltage source via wires. The blade is bonded with gold bonding wires, which will extend by insulated Cu wire to connect ouside voltage source eventually. We also need to bond the capacitors to the blade via gold wires to filter the noise.


Blades, wires and capacitors, together with device or tools that will contact with them are all cleaned by pure ethanol to avoid organic matters influence the vacuum condition. The insulated wire is not recommended to clean ultrasonically since insulating material may take off from the wire during the cleaning.


The first step is to bond gold wires on the bonder platform. You would like blades, gold bonding wires, capacitors, a wire bonder, some other tools like tweezers to start. (Parameters for the wire bonder should be proper choosed so that bonder joint is tight while the blade or capacitors won't be broken.) The procesure to bond gold wires is quite trivial:

  • bond a short wire to one side of a capacitor, then bond a long wire to the other side of the capacitor (each DC blade need 5 capacitors)
  • bond the short wire to the blade segment while the long wire to the blade ground
  • bond another gold wires to every DC, RF and ground electrodes. Make sure they are long enough to connect with insulated wires (14 wires in total)
Note: we just confirmed that if you wind the gold wire around the copper wire, it's possible to use the bonder to bond those two wires, which may be a bit difficult if using a spot welder. Thus you can avoid the tedious work later by connecting those wires first, though it may not easy to wind the insulated wire to the holder. After all, troubles conserve.

Suppose we want to connect gold wires to insulated wires later, the next step is winding insulated wires to the holder (or blade alignment, discussing later). Before winding, the insulated material on the wire ends should be taken off. You can cut it with a knife or scissors, or even burn it with fire, whatever you like. A small trick is applied here to keep the wires neat: the wire should enclose previous wires when winding around the hole. (See Fig 6)

Since the blade and insulated wire is attached to the holder now, it seems impossible to use the bonder. Therefore we use a spot welder instead. Using stainless steel sheets to assist is optional. Winding the gold wire then welding directly is fine too. (See Fig 8) Again, Parameters for the spot welder should be proper choosed so that joint is tight while the gold wire won't be broken.

Fig 1: Our wire bonder
Fig 2: An RF blade and a capacitor with gold wires
Fig 3: A already bonded DC blade
Fig 4: Our lovely blade in microscope view
Fig 5: Wind the insulated wire around the hole...
Fig 6: Then put blades...
Fig 7: Our spot welder. Press 'PROG' and 'CHG' can set the pulse power; Press 'RUN' to start welding.
Fig 8: Welder leaving wires and finish this job!

Blade alignment and measurement

The bonder cannot be used after blades attached to the holder while the blade alignment may be harder after connecting gold wire with copper wire. Thus we aligned blades right after winding insulated wires to the holder.

The goal of blade alignment is making DC null point and RF null point coinside when applying symmetry voltage. Specifically, the blade eages should form a rectangle on the side view. The segment of DC blades should also aligned so that the null point can located at the center. We used a WiFi digital handheld microscope to help alignment.

For a nearly perfect alignment, you can refer to Jarry Ma's 《装腔宝典》. An absolute coordinate system method is introduced. You may find the operation describe there cumbersome. (I feel the same.) A more feasible way is first align the blades with eye estimation. Then apply greedy algorithm to adjust the blades step by step.

Lacking of experiments, our alignment result is far from satisfied (Fig 3). We just pay too much attention on segments alignment than side edges alignment. I would adjust my tactic and be more patient if I would do it again.

Fig 1: Aligning blades by F100 microscope...
Fig 2: Observing the side view of the blades with another microscope...
Fig 3: This figure is combined with pictures of two side views, whose 4 edges are denoted by blue and orange box respectively. Suppose the blade's edge is straight, the side view of 4 edges at the middle should form the pink box. The RF and DC null for a symmetry voltage is denoted as pink and green point. Using CCD camera's data, we estimate that the height difference of each blade pair is 348.4um and 263.9um.
Fig 4: Buttom view of blades in CCD using LMU-5x-NUV as objective. The reading of stage's vernier micrometer is showed in figure. It was read when each corner (orange circle) was tuned to the middle of CCD screen. We estimate that the gap between downner blades is about 421.5um.

Oven assembling and test

The ytterbium source is indispensable for our ion trap experiments. Here we refer to it as "oven". The oven should have a container to contain the ytterbium, two ends to connect to the current source and a widget to fix its postion. Once we assembled the oven, we would like to put it into a vacuum glass tube and see if it works.


We use a plastic syringe needle as the container. A customized metal bracket (see Fig 1) together with ceramic cement is used to fix the needle to the chamber. A ceramic tube is also needed to seperate the needle and the braket so that the oven will not directly connect to the ground. Again, we use insulated wires (should be abled to withstand 10A current) and stainless steel sheets to connect the oven with external current source. You may also want a steel wire to poke the ytterbium into the needle. Most of things mentioned above can clean ultrasonically. (Be Cautioned! Someone of our neighbor lab just tried to clean the barium with pure ethanol and found it disappeared.)

We need to cut the ytterbium to small pieces (cut it to thin stripes may be helpful) so as to put it to the needle. Yb174 is in the form of ingots while Yb171 is in the form of patches. It's not easy to tell which one is easier to handle. Once you poked the ytterbium to the needle, following below steps:

  • fix the ceramic tube to the bracket with ceramic cement (can speed up by a heat gun)
  • take off the insulated material from the wire ends and weld it with a steel sheet
  • put the needle to the tube with the right direction and connect it with two insulated wire via steel sheets
  • fix the needle to the ceramic tube with ceramic cement
Tips: The resistance of welder joint should be small, otherwise all the power will be absorbed and the ytterbium will not be heated. However, it's not easy even just to connect wires or needles with steel sheets by a spot welder. One reason is that the steel sheet is not well contacted with the wire when welding. Thus using pliers to squeeze the joint before welding is strongly recommended. Using gold wire to reduce the resistance may be feasible too.

Test Method

As Fig 5 shows, we connect a vacuum glass tube to a 4-way cross fitting, another two ways of which connect to a mocular pump and a current source. Be careful and don't let screws or wires touch the chamber, otherwise it'll short to the ground. The copper gasket can be reused in this test, as we don't need a very high vacuum condition.

Oven test is also a process of degassing. When increasing the current, you should see the pressure quickly increases, then slowly drop down. Record the data before the ytterbium spout out if you feel boring.

Test Result

The threshhold current for oven 174 is about 5.5A. (The resistance is large because our outside ground wire here is too thin.) Below is the testing data:

current(A) 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5
voltage(V) 1.2 1.5 1.8 2.2 2.6 3.0 3.5 4.9 4.5 5.0
pressure(e-6 mBar) - 5.0 5.2 5.2 5.2 5.2 5.4 5.6 5.4 drop quickly

The threshhold current for oven 171 is about 6.0A. Below is the testing data:

current(A) 2.5 3 3.5 4 4.5 5 5.5 6.0
voltage(V) 0.482 0.578 0.643 0.740 0.842 0.938 1.038 -
pressure(e-6 mBar) 6.1 6.1 5.8 5.6 5.4 5.2 4.8 drop quickly

Our 171 Oven failed the test several times. Here is some lessons:

  • The steel sheet should welded closely to needle tip, or the front end of the needle could be hardly heated.
  • Make sure the welder joint is tight. Reduce the resistance as much as poosible, or the oven may be heated unevenly.
  • Put the needle tip close to the tube, thus we can see the spouted ytterbium more clearly.

Fig 1: Some parts to assemble the oven. Can you tell the ytterbium 171 from the steel sheet?
Fig 2: Poking the ytterbium to the needle. Quite tedious work...
Fig 3: Assembled ovens
Fig 4: Our molecular pump. Open the valve and turn "PUMPING" button from 0 to 1 to open mechanical pump firstly. Wait the pressure to get down below 0.1 mBar, then open molecular pump by press "Start/Stop" button.
Fig 5: We use a vacuum glass tube to test the oven.
Fig 6: 7.5A current is going through the oven. Though temperature is very high, we still can't observe the ytterbium. This oven is abandoned finally.
Fig 7: The spouted ytterbium 174.
Fig 8: The spouted ytterbium 171. Noticed the spot is a semicircle, which may casued by the last ytterbium sheet we put in.

Assemble the Chamber

Now we have all parts to assemble a chamber. We first connect the vacuum chamber with a 5-way cross through a nipple reducer, as Fig 1 shows. Then we install the blade holder and the oven. The next step is to connect insulated Cu wires of oven, DC and RF electrodes with the feedthrough. Finally, we install all other vacuum parts to the chamber. Click here to see how we assembling the chamber.

Before we put the chamber into the oven, we tested all the electronic connections. The pressure reached 9e-7 mBar after 3 hours' pumping with a molecular pump, which proved no air leakage of the chamber. A good helical resonator was also used for a quick test, which resulted a 285's Q factor and 23MHz's resonant frequency.

It's a simple but tedious work. As Jarry Ma's 《装腔宝典》 indicated, there are 4 things need to be pay attention:

  • Whether the light path is clear.
  • Make sure all the wires is correctly connected, not short or open circuits.
  • All the parts should be vaccum compatible. Avoid to introduce new contaminants.
  • Some parts are fragile. Take care with them when mapulating.
Unfortunately, we almost fell into very trap listed above and pay the price.

Since there are only 5 holes on each edge of the holder, we connected the DC ground electrode via a gold wire to the insulated wire of the other side. After wire bonding, we cannot change the position of this gold wire, for which we take risk that a laser from side window may be blocked by the gold wire.

As metioned on "Oven Assembling" section, it's not easy to connect wires or needles with steel sheets by a spot welder. Thus we tend to leave a relativetely long needle and steel sheet. Such a length should be well selected so as to avoid them touching the chamber. After installing two ovens to the chamber, we found that steel sheets of oven 174 touched the chamber and shorted to the ground. The oven is no longer floating to the ground, thus we should make sure the current go from the correct pin to heat the oven by denoting the ground pin. DC electrodes connection also matters. Due to some mistakes, we connect them with the Sub D pins in a strange order, which partly account for the wrong DC voltage in the second trial to load ions.

How does the vaccum condition perform if I don't clean the vaccum parts and assemble it to the chamber directly? I think many people have such thoughts but dare not give a try. Well, I tried. (I would rather not do it.) Most carelessness is harmless. But an angle valve coming back from a machine factory is definitely dangerous. Because of no cleaning the angle valve, it take me 10 days to find the problem and get a reasonable good vaccum condition.(See "Air Bake-out" section.)

It has been about two months since I installed the holder to the chamber. The first feeling when I saw the holder broken to two parts is not frustrated, but shocked. We finally rescued the holder with ceramic cement. Here's the suggestion: take care of the last 10 minutes. You definitely don't want to repeat what you did a month ago.

Fig 1: Main body of the chamber.
Fig 2: Parts to assemble SubD connector.
Fig 3: Install DC feedthrough is not easy...
Fig 4: Top view of the chamber.
Fig 5: Buttom view od the chamber.
Fig 6: Oven Feedthrough. Connect red circled pin to the current source and green pin to the ground for Yb174 oven. Connect them with pink circled pin for Yb171 oven. (Either can be connected to ground.)
Fig 7: RF Feedthrough. The pins' mark corresponds to RF and ground electrodes showed in Fig 4 and 5.
Fig 8: DC Feedthrough. The ten colored numbers corresponds to denoted electrodes showed in Fig 4 and 5. The other 15 pins is denoted by white dashes as they are open circuits.

Set up Optical Table

The table plate for the optical small table is customized by 大恒光电. Here is the drawing.

While the table legs is from Thorlabs. Here is the list to build a leg:

Modal Description Unit Price
C1515/M Metric Mounting Post Bracket ¥826.83
PSHA/M Metric Adjustable Height Collar ¥534.55
PB4/M 6662-001 Rev C, P-Series Pedestal Base Adpater ¥106.05
P300/M Ø38mm Solid Post 300mm Length ¥674.48
PF175 P-Series Clamping Fork ¥149.50
Fig 1: Optical table set up!
Fig 2: Our optical small table

Set up Chamber and Imaging System

Before we build up our imaging system, we fix the chamber's position. Homemade stands are used to rest the chamber on to a designed height. (The window center is set to 3 inches height, thus most of the mirrors can be mounted directly to pedestal posts.)

Below is the schematic of the imaging system. As fig 1 indicates, the objective will collect the light emitted from the object point at its working distance. The Best Form lens which designed to minimize spherical aberration will then focuses the light to the position of the aperture. We use two lens (F=50) to form a doublets as an eyepiece. The doublets' focus length is 25mm. So to realize magnification, the distance from aperture to the doublets should between 25mm to 50mm.

This is a decoupled imaging system. Therefore, you can tune the telescoping lens tube first to see a clear aperture eage on CCD. Then adjust the stages to observe the blades' structure.

Fig 1: The schematic of our imaging system
Fig 2: A photo of our imaging system

Chatting Room

Hello {{user_name}} [logout], want to leave a comment?

{{message.user_name}}  {{message.create_time}}