Differences

This shows you the differences between two versions of the page.

--- uhv_system [2023/03/22 10:34] – rog
+++ uhv_system [2025/10/29 16:07] (current) – doorn
@@ Line 1: / Line 1: @@
 ====== Description ======
 The UHV system is a magnetron sputtering system with a background pressure around 10e-10 mbar and a deposition pressure of 1e-3 to 1e-2 mbar. It is equipped with a loadlock that holds a sample garage with 7 sample spaces. One sample space can hold a 15x15mm wafer. The substrate is transferred to and from the chamber using two linear manipulators and a wobble stick. Most targets in the chamber are at an angle with respect to the sample, and the sample can be tilted towards 3 out of 4 magnetron sources. The system can be used with both DC and RF power supplies (DC being standard). The sputtering gas is Ar; the loadlock is also connected to N<sub>2</sub> and O<sub>2</sub> for oxidation. The substrate stays relatively cool and cannot be cooled or heated away from room temperature.
+The UHV system is maintained by the nanolab and the Lahabi Lab. Contact the Lahabi Lab if you are interested in using the system. Target changes are only performed sporadically, to maintain excellent vacuum conditions.
 ======The Vacuum System======
@@ Line 16: / Line 18: @@
 ===== Venting the loadlock and loading the substrates: =====
-  **- Never touch anything inside the machine without wearing clean, powder free gloves.**
+  - **Never touch anything inside the machine without wearing clean, powder free gloves.**
   - Sampleholders are kept on the workbench in the sputterlab, inside the petri dishes. Always clean your sampleholder with acetone and IPA before starting.
   - Glue your sample onto the sampleholder with a small amount of silver paste. Let it dry (approximately 10 minutes). You should be able to hold your precious sample upside down!
@@ Line 23: / Line 25: @@
   - Turn off the loadlock turbo on the turbo controller panel.
   - Close the valve of the rotary pump. Clockwise = close. This closes off the rotary pump.
-  - Let the turbo spin down for 1 minute. Then let in short bursts of N<sub>2</sub> using the N<sub>2</sub> valve on the bottom panel (max. 5 mbar). Listen to the turbo pump to fully spin it down. Take your time! Spinning down the turbo should take about 5 minutes.
+  - Let the turbo spin down for 1 minute. Then very slowly and carefully let in a bit of N<sub>2</sub> using the N<sub>2</sub> valve on the bottom panel (max. 25 mbar on the gauge *behind* the turbo at this stage). Listen to the turbo pump to fully spin it down. Take your time! Spinning down the turbo should take about 5 minutes.
   - When the turbo has stopped, vent the system completely with N<sub>2</sub>. The loadlock door will burst open, so stay clear. The chamber pressure can sometimes increase to 1e-8 mbar range because of a leaky gate valve. This means work quickly!
-  - Put your sampleholder inside the garage with the long side towards you. Always wear gloves, and use long metal tweezers.
+  - Put your sampleholder inside the garage with the long side towards you. Always wear gloves, and use long metal tweezers. Take care to maintain N<sub>2</sub> flow, such that water vapours cannot enter the loadlock.
-  - Close the loadlock. The door needs to be in perfect alignment, so tilt it a little if uncertain. You should see absolutely no gap between the loadlock and the door.
+  - Close the loadlock. The door needs to be in perfect alignment, so tilt it a little if uncertain. You should see absolutely no gap between the loadlock and the door. As you close the door, also close the N<sub>2</sub> valve.
-  - Close the N<sub>2</sub> valve.
 ===== Pumping down the loadlock: =====
   - Open the valve to the rotary pump.
   - The pressure should start decreasing almost immediately. If it does not decrease, close the rotary pump valve and check if the door is properly shut.
-  - When the pressure on A (the loadlock intermediate pressure gauge) reaches 1e-1 mbar, switch the loadlock turbo pump back on.
+  - When the pressure on A (the loadlock intermediate pressure gauge) goes below 5 mbar, switch on the loadlock turbo.
-  - When the pressure on A reaches 5e-2 mbar, turn on IG2. The heat of IG2 will help pump the system down.
+  - When the turbo reaches its max rate, turn on IG2. The turbo can take up to 5 minutes to spin up fully. The heat of IG2 will help pump the system down.
-  - The turbo can take up to 10 minutes to spin up fully. If all the LEDs are on, it is spinning at maximum speed. Check this!
+  - Pump until base pressure (<5e-8 mbar) in the loadlock has been reached. This could take up to 12 hours.
-  - Pump until base pressure (<5e-8 mbar) in the loadlock has been reached. This could take up to 24 hours.
 ===== Sputtering: =====
@@ Line 41: / Line 41: @@
 **Pre-sputtering**
   - Fill in logbook
-  - Turn on cooling water. There are two valves at the back of the instrument (blue labels on the left). There is no interlock.
+  - Turn on cooling water. There are two valves at the back of the instrument (blue labels on the left).
   - Take care that the sources are connected to the DC switch box. They could be disconnected, especially after a bake-out.
   - Turn off all IGs on the black panel.
   - Open the shutter that you are going to use. Make sure all other shutters are still closed.
-  - F**ully close the NV.** The needle valve is the black needle valve behind the black gas panel. CW = close. **Otherwise you will flood the chamber with Argon!** Don’t force the NV.
+  - **Fully close the NV.** The needle valve is the black needle valve behind the black gas panel. CW = close. **Otherwise you will flood the chamber with Argon!** Don’t force the NV.
   - Close the valve conductance to 10%. Otherwise the turbo will just pump all the argon away.
-  - Turn on the argon-measuring ion gauge (Pfeiffer controller on the rack). Give it a minute to warm up. It should report underpressure (ur=under range) or some small number, as its range is limited to 1e-6 mbar. You can turn it on at the back of the rack.
+  - Turn on the argon-measuring gauge (Pfeiffer controller on the rack). Give it a minute to warm up. It should report underpressure (ur=under range) or some small number, as its range is limited to 1e-6 mbar. This dual-gauge is not accurate calibrated below 1e-5 mbar. You can turn it on at the back of the rack.
-  - Slowly open the argon line on the black panel
+  - Slowly open the argon line on the black panel.
-  - Gradually open the NV and let in ~2e-3 mbar of argon pressure. It takes some “human PID” to get this right.
+  - Gradually open the NV and let in ~2e-3 mbar of argon pressure. It takes some “human PID” to get this right. N.B.: even when the NV is fully closed, there is still a substantial flow through the NV. This does not mean that the needle valve is broken; all needle valves have this to some extent.
   - Now use the valve-conductance controller to regulate the pressure for your sputtering process.
   - Select your target on the DC-switching box.
@@ Line 90: / Line 90: @@
   - Close the Ar valve on the gas panel fully.
   - First pump out whatever gas remains, by opening the valve-conductance to 100%. The red “open” light should remain on and the Ar gauge should go to 1e-4 mbar range. This is because there is still gas in between the needle valve and the Ar valve.
-  - Open the NV fully (slowly!). The Ar gauge should go under range. You can now turn it off.
+  - Open the NV fully. Once the dual-gauge goes below 5e-6 mbar, the chamber is empty and you can now turn it off.
   - Turn on IG1 again, and check that the chamber is returning to UHV conditions.
   - Switch off cooling water.
   - Vent the loadlock as described above in order to remove your samples
   - Always pump down the loadlock after use (as described above).
+===== Bake-out: =====
+To achieve a high-quality vacuum after the chamber has been vented, a bake-out is required. A good bake-out should take at least 48 hours to properly heat the whole system.
+  * Turn on the cooling water
+  * Disconnect the HV cables from the targets
+  * Remove the Ar gauge
+  * Cover the viewports with aluminium foil
+  * Place all the panels to fully cover the whole machine
+The blue cables for the vacuum gauges do not need to be covered. To start the bake-out, flip the breaker switch up and turn the black knob to the left. This permanently turns on the heaters (will go between the two setpoints, should be 80 and 100 C). Turning it to the right will use a timer but this was found to be iffy. After the bake-out make sure to reattach all cables and the Ar gauge. Please be aware that there is no interlock preventing you from turning on the heaters without the cooling water. The cooling water must be on during the bake-out!
 ====== Deposition rates ======
+RATES AFTER THE BIG MAINTENANCE OF 2025
+All rates measured by Jibbe Reuver with tapping mode-AFM in June 2025.
+^ Material ^ Process parameters ^ Thickness ^ Rate (nm/min) ^ Rate (nm/s)  ^
+| Co | 4.0e-3 mbar, 100 mA, 10 min          | 31.2 nm   | 3.12  nm/min | 0.052 nm/s  |
+| Cu | 4.0e-3 mbar, 65 mA, 10 min           | 136.86 nm | 13.69 nm/min | 0.228 nm/s  |
+| Nb | 4.0e-3 mbar, 200 mA, 20 min, 45 deg  | 61.58 nm  |  3.08 nm/min | 0.0513 nm/s |
+| Pt | 4.0e-3 mbar, 150 mA, 10 min, 45 deg  | 122.9 nm  | 12.29 nm/min | 0.205 nm/s  |
+| Pt | 4.0e-3 mbar, 100 mA, 10 min          | 111.37 nm | 11.14 nm/min | 0.1856 nm/s |
+| Pt | 4.0e-3 mbar, 120 mA, 10 min          | 138.53 nm | 13.85 nm/min | 0.2309 nm/s |
+| Pt | 4.0e-3 mbar, 100 mA, 10 min, 45 deg  | 87.8 nm   | 8.78 nm/min  | 0.1463 nm/s |
 Definition: ''sputtering angle'' is the angle between substrate and target. If this angle is not zero, it should be noted here. Note that for sources 1 and 3 the substrate needs to be tilted 45 degrees to have a zero ''sputtering angle''.
-RECENT RATES
+**Small note on rate calibrations: it is our great preference that we do not calibrate with profilometer.** In the past, this has given wildly inconsistent results. While the profilometer setup has received some love recently, the working principle of the machine has not changed. Similarly, the "marker drop" technique, that is often used to estimate sputter rates, can give wrong rates even on AFM because it can lead to round ramps over the same length-scale as higher-order non-linearities in AFM piezos. The best method for rate estimation is to make a calibration sample at e.g. the EBPG, then run a liftoff and perform AFM on this sample.
+RATES BEFORE THE BIG MAINTENANCE OF 2025
 ^ Material            ^  Date    ^ Sample ID ^ Process parameters          ^ Measurement ^ Result  ^ Rate          ^