Gencoa - Dermot Monaghan Using a standard Penning Gauge as a powerful means of monitoring and feedback control Victor Bellido-González, Sarah Powell, Benoit Daniel, John Counsell, Dermot Monaghan
Structure of the presentation Plasma Spectroscopy In-situ & Ex-situ Monitoring Feedback Control Penning Gauge and PEM Examples of spectra and uses Examples of plasma instabilities / drifts Feedback modes from Penning Gauge Conclusions
Plasma or Optical Emission Monitoring (PEM) Optic fibre monitoring
Optical Emission Spectroscopy - The Plasma Spectrum Magnetron Sputtering Ti/Ar P la sm a e m issio n, a.u. 1500 1000 500 0 350 450 550 650 750 850 Wavelength, nm
Process Control by Optical Emission Spectroscopy Select a high intensity peak Can be done by filter, monochromator or CCD array Use the change in peak intensity as the feedback control for the reactive gas PEM provides valuable process information for feedback control or condition monitoring 450 550 65
Reactive gas input alters intensity due to target poisoning 500 550
PEM picks up the changes in the plasma environment PACVD - DLC deposition P la sm a Em issio n, a.u. 4000 3000 2000 1000 0 DIRECT mode: More REACTIVE GAS More signal e.g. DLC- PACVD process 325 375 425 IPA-01 IPA-03 IPA-04 IPA-06 Wavelength, nm
Control of the reactive gas input in a INVERSE mode P la sm a Em ission, a.u. 1200 1000 800 600 400 200 0 Magnetron Sputtering - TiO x deposition 400 450 500 550 600 Wavelength, nm INVERSE mode: More REACTIVE GAS Less signal e.g. Ti sputtering (Ar + O 2 ) O2= 0 O2= 14 O2= 18 O2= 22
In-situ plasma monitoring - conventional
Examples of Reactive Sputtering Exhaust Part. Press. Target Reactive gas input Control Voltage, Freq. Plasma emmission Substrate Transp.,σ,n
Ideal Controller Handles ALL I/O s Inputs PEM Metal line / multiple lines Gas line Argon line Plasma spectrum Target voltage Gas partial pressure C o n t r o l l e r Outputs (Actuator) Reactive Gas Flow Target Voltage Target Power Output Gas partial pressure (Ar / throttle valve) Why is Multi-function important? Some processes are hard to control and drifts / fluctuations can occur
Stepped control by PEM at 520nm St.St. Control at different % of PEM Signals, (%, a.u.) 100 %PEM-Control setpoint 90% %O2 MFC 90 %Target Voltage 80 70% 70 60 50% 50% 50% 50 40 30 30% 20 10 20% 10% 5% 0 0 100 200 300 400 500 600 700 Time, s The controller can control at any value of PEM setpoint (520nm- Cr peak). Different setpoints correspond to different target status.
Ex-situ monitoring inside a Penning Gauge A plasma is generated inside a Penning type pressure gauge. The ionisation in the gas creates a current between the anode and cathode which represents the pressure. The information from the plasma spectrum can be extracted.
Pro / Cons of In/Ex situ PEM In-Situ Pro s High intensity plasma signal Represents the process environment displays all species Faster response from process Allows local zone control Measures all species In-Situ Con s Subject to drifts and disturbances as process environment changes Requires a process plasma Ex-Situ Pro s Less subject to drifts and substrate disturbances Process can be off remote plasma generation Measures the excess gas Optics don t coat or heat-up and are outside the vacuum Ex-Situ Con s Plasma intensity lower Reaction time slower Generally no local zone control Only for gases
What other things can PEM Achieve? Diagnose the system condition Predict a suitable process start condition Eliminate long term drift condition monitoring Rate measurement Intelligent plasma pre-treatment Improve uniformity of deposition or pre-treatment zone control Leak detection Preventative maintenance
Full optical spectrum from penning gauge 1200 Measured Intensity - plasma and gauge on 1000 Measured Intensity - dark - process plasma off - gauge on - Hydrogen peak from moisture 800 Intensity 600 400 200 0 200 300 400 500 600 700 800-200 Wavelength (nm)
Full optical spectrum from penning gauge with O 2 in chamber 50% O2 8.5E-04 Torr 800 O2 peak 700 600 500 Inensity 400 300 200 100 0 0 100 200 300 400 500 600 700 800 900 1000 Wavelength (nm)
Full optical spectrum from penning gauge with N 2 in chamber 30% N2 2.4E-03 Torr 1800 1600 1400 1200 Intensity 1000 800 600 N2 peak 400 200 0 0 100 200 300 400 500 600 700 800 900 1000 Wavelength (nm)
Examples of water monitoring
Variable setpoint feedback control from O 2 signal in Gauge TiOx Penning gauge sensor control using Gencoa Speedflo 4.5 4 3.5 Gas Feedback (SCCM) Target voltage (V) Sensor (V) Set point (V) 1.8 1.6 1.4 3 1.2 Gas feedback (sccm) 2.5 2 1 0.8 Target / Sensor (v) 1.5 0.6 1 0.4 0.5 0.2 0 0 0 50 100 150 200 250 300 Time (s)
Variable setpoint feedback control from N 2 signal in Gauge Penning gauge sensor control for TiN 90 3.5 80 70 60 Gas Feedback (SCCM) Sensor (V) Set point (V) Target Voltage (V) 3 2.5 Gas feedback (sccm) 50 40 30 2 1.5 Target / Sensor (V) 20 1 10 0 0 50 100 150 200 250 300 350 400 450 0.5-10 Time (s) 0
In chamber plasma disturbance as a result of substrate movement & position TiN penning sensor with Speedflo feedback control - substrate rotation on/off MFC, Ti (500nm) and Target V signals 14 12 10 8 6 4 2 0 Setpoint Ch1 Ch2 Rotation on Rotation off Intensity in Gauge - Setpoint & Control not disturbed samples moved closer to plasma samples moved away from plasma Ch4 MFC1 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 Setpoint and Control signals Ch2 Ti 500 nm from process chamber - monitor only not control Ch4 Target V signal - monitoring only of target voltage as substrate movement varies MFC1 - change in reactive gas input from Speedflo to maintain control Ch1 Penning N2 670nm as feedback control signal setpoint Ch1 - the intensity / gas signal to control around 0 50 100 150 200 250 Time, s
Other considerations in long term stability -Plasma drifts Anode at 149mm width - earthed Anode at 147mm width - floating The plasma changes shown simulate drifting conditions that can happen as the anode or chamber is covered in oxide: Disappearing anode + web damage + outgassing!!! Anode at 147mm width - earthed
Monitoring - increased productivity Things can change with time on a short and / or long term basis and in different ways and for different reasons depending upon the material system and operating parameters. Monitoring can help. A summary of possible advantages: Increase rates with feedback control Detect faults by monitoring water, leaks, outgassing Set plasma pre-treatment parameters to the environment Predict drift into unsuitable process regimes Optimise uniformity of deposition of plasma treatment Predicting maintenance intervals
Where to monitor In/Ex-Situ? Both have their advantages and disadvantages and the optimum solution is probably to combine both. In-Situ is better for zonal control (large area) and where drifts and local plasma disturbances are less likely to occur. Ex-Situ in principle is better for reactive gas control as it measures the excess. Also can sense the chamber environment before any process has begun. There are now a very powerful array of sensors that can be fed into a controller and combined to optimise short and long-term performance. This intelligent monitoring can then be use to control any number of outputs simultaneously MFC (gas or argon), source power (thermal or sputtering), pumping.
The controller needs to very flexible To be able to provide the required short term and long term control the use of multiple inputs and outputs are highly desirable. The software then needs to be able to easily combine the configurations of the different inputs and outputs and allow the most appropriate combination to be used. In addition the integration of the input signal to smooth out some short-term process shifts should be possible
Conclusions Process monitoring and feedback control can improve process yields and reduce product variations. Plasma emission monitoring provides an excellent tool to diagnose the process environment. By choosing or combining In-situ with Ex-situ monitoring a more robust solution is possible.