gencoa optix optix a new instrument for robust gas analysis in vacuum november 2017 1

Gencoa OptixA new instrument for robust gas analysis in vacuum

November 2017 www.gencoa.com 1

Why Monitor Your Vacuum Process?

• Identification of vacuum or process problems before they have a financial impact

• Optix maps the process environment to ensure reliable production

• Improves quality of products and repeatability

• Outputs for better process control


Optix vs RGA


Quadrupole Residual Gas Analyzers (RGAs)

Philiphofmann.net

Optix – remote plasma gas analysis (RPGA)

Optical method

Remote plasma

Spectrometer

Light

Low ppm detection Low ppm detection

Optix vs RGA


Optix – remote plasma gas analysis (RPGA) Optical method

Quadrupole Residual Gas Analyzers (RGAs)

Robust– detector separated from chemicals by optical window

Detector in contact with chemicals – easy to contaminate, hard to clean

No filaments –simple electrode geometry Filaments and ionizers are consumables

Operates 0.5 to 10-6 mbar Only operates reliably up to 10-4 mbar

Direct chamber monitoring – no need for differential pumping unless atmospheric sampling

Higher than 10-4 mbar pressure needs differential pumping – loss of sensitivity

FAST – ‘speed of light’, 10-50 msec response Typically 0.5 to several seconds range

Tolerates volatiles in the vacuum – hydrocarbons, solvents, long chain polymers

Only small amounts of contamination beforesensor failure

Wide range of useful software applications available –gas tracking, leak detection, pump-down monitoring, water tracker, end-point detection, multi-mode process tracking

Typically gas tracking & leak detection

Sensor degassing mode – avoid false reading Yes, but degas can affect filament lifetime

Optix remote plasma gas analysis


Cr`

Vacuum process 0.5 to 10-6 mbaror with a rotary pump to support atmospheric sensing

Wide pressure range remote plasma generator

High intensity plasma

Wide range spectrometer 200-850nm

Spectrum analysisgives species composition

Operating pressure

• Optix operates in the typical plasma processing pressure range

• Easy to use and robust device


Plasma has some non-linearity

Plasma light intensity too low

1x10-6 mbar – 0.5 mbar

Vacuum-based applications


1x10-6 1x10-5 1x10-4 1x10-3 1x10-2 1x10-1

Pressure (mbar)

Evaporation

Sputtering

MF/RF Plasma surface treatment

CVD / PECVD

Optix & RGA Optix and differentially pumped RGA

Optix plasma generation

• Unlike RGA’s, the Optix detector is separated from the chemicals by an optical window, eliminating detector contamination

• Purpose designed and patented plasma generation source

• Very wide range of operation, with plasma present from 0.5 to 10-6 mbar

• Fast current feedback control

• Constant current – constant excitation source

• DC mode as standard, pulsed DC for highly contaminating atmospheres (ALD, OLED & CVD)

• Atmospheric sampling via simple mechanical pump – no turbo required


Current set-point (fixed) Power Supply fast current feedback control

Voltage regulation (up to 3kV)

++++

+ +++

+

Voltage and current feedback

Software interface


1. Connect / disconnect Optix sensor2. Optix device settings3. Power supply settings. 4. View spectrum plot.5. Sensor settings. Species setup screen. 6. Clear triggers.

7. Trigger settings8. Species table. 9. Species doughnut.10. Trigger settings.11. Species Table.

12. Species doughnut. 13. Species bar chart.14. The current total pressure reading.15. Species trend view. 16. Trigger status.

Optix software applications

• Quantitative gas analysis down to 2 ppm

• Gas mixture balance – up to 8 gases – can be selected

• Process gas tracking with trigger/alarm outputs

• Full spectrum view 200-850nm, control of integration time for sensitivity adjustments

• Tuneable spectrum view – more focussed range

• Automatic gas peak detection – gas auto identification database can be adjusted to incorporate additional un-common peaks of specific interest

• Leak detection mode for any gas


Optix software applications (continued)

• Process moisture monitor with triggers/alarm outputs

• Chamber pump-down tracker with triggers/alarm outputs

• Option of vacuum switch to prevent accidental operation at atmosphere

• In-built vacuum pressure reading

• Control of plasma power supply to tune power parameters for non-standard applications

• Multiple sensor monitoring

• Multiple language display options – English, Japanese, Chinese, French, Spanish


Software spectrum view


The spectrum displays the constituent species of the plasma. Each gas peak is automatically identified – as only common gases present in a vacuum are identified, this is a much more accurate feature than other spectral identification methods. The auto identification database can be user-defined.

Gas trending/tracking view


Hardware configuration options


Standard Optix package, plasma generator with power supply (DC as standard, pulsed DC as an option) with Spectrometer head and Optix software package/cables

Plasma generator with DC power supply (pulsed DC optional) and cables. Generates an intense plasma over a wide pressure range – can link to Speedflo or other control platforms

Standard Optix package, with optional optical fibrelink between sensor and spectrometer – increases flexibility of the package – use items separately

Spectrometer head with Optix software package – take advantage of the Optix software suite to manage your plasma monitoring and take advantage of the communication and trigger facilities

Dimensions and I/O options


95 mm

17

0 m

m

300 mm

Optix sensor: 300mm x 170mm x 95mmPSU: 165mm x 105mm x 55mmSensor weight: 2.2kgVacuum connection: KF25 flangeMounting orientation: Any

Communication interfaces

USB, RS232, Ethernet Digital relay output x 4

Optional PLC communication interface

Software compatible with Windows 7, 8, 10

Optix software

• An unlimited number of gas species can be monitored using the trend-line feature


Technical information

• Electrical Input voltage: 15V

• Input power: 20W typical

• Output voltage: 3kV max

• Output current: 1.5 mA max

• Total pressure operating range: 1x10-6 mbar – 0.5 mbar

• Sensitivity: 50 ppm air in argon at 1.6x10-2 mbar total pressure

• Spectral range: 200nm – 900nm

• Update speed: 5ms – 5s (depending on sensitivity selected)

• Total pressure measurement: Integrated (1x10-6 mbar –1x10-2 mbar measurement range)

• Electronics maximum operating temperature: 40°C


Optix hardware


• Much simpler and smaller hardware than a differentially pumped RGA arrangement

• Comprises the sensor head, small power supply and a remote PC or machine control

• Software for loading onto PC is provided

Remote Plasma Emission MonitoringHighly Sensitive Optical Method

Which species can be observed?

• Atomic emissions and molecular emissions (relative concentrations)

• Larger molecules are observed as fragments – due to disassociation in the sensor’s plasma


0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

200 300 400 500 600 700 800 900

OHCO

CH

H

CH

O

𝐴2Σ+ − 𝑋2Π

𝐵2Σ − 𝑋2Π

𝐴2Δ − 𝑋2Π𝐵1Σ − 𝐴1Π

Caveats and considerations when using RPEM Disassociation

• “Fingerprints” of the original molecule


0

2000

4000

6000

8000

10000

12000

14000

200 300 400 500 600 700 800 900

Acetone (CH3COCH3)

Isopropanol (CH3CHOHCH3)

Remote Plasma Emission Monitoring

• Spectrum interpretation

• A single leak can emit multiple emission lines showing the exact composition

• Automated identification of the spectrum makes interpretation easier


N2+

H

OCO2+

N2+

Sensitivity quantification

• Experiments were repeated at lower pressures (reduced Arflow) and increased sensitivity was observed

• Based on this PPM limits at lower pressures can be predicted


1

10

100

1.00E-03 1.00E-02 1.00E-01

PP

M li

mit

Pressure (mbar)

Projected PPM limit

Mea

sure

d P

PM

lim

it

-1500

-1000

-500

0

500

1000

1500

10:48:52 10:49:35 10:50:18 10:51:01 10:51:45

531 PPM

Leak closed

Leak opened

Leak open

Application example 1: Hard coating

Optix gas measurement during heating phase of an Arc based hard coating cycle, comparison of different tools (Pressure c. 1x10-5 mbar)

• Outgassing ofspecies during heating phase prior to coating

• Comparison of outgassing between different tools


Application example 1: Hard coating

• Observation of gas levels during coating phase of an Arc based hard coating cycle

• Consumption of nitrogen and water vapour

• Outgassing of hydrogen


Example 2: reactive sputtering

• Optix gas measurement during reactive sputtering, comparison with high pressure RGA (process pressure 4x10-3 mbar)


Target clean (4 kW)

Poisoned (1 kW)

Cle

an (

4 k

W) Reactive mode (4 kW)

H outgassing from magnetron Shutter opened


• Carbon sputtered coating• Deposited on particle accelerator inner

surface to reduce secondary electron yield• Deposition pressure of 1.1x10-1 mbar• Performance of coating is sensitive to the

presence of H outgassing from the magnetron environment

• Objective to monitor H outgassing during the deposition

Courtesy of CERN Vacuum Surfaces and Coatings Group

Case study 1 – Measurement of outgassing during carbon coating


• Comparison between Optix and a differentially pumped RGA• Hydrogen - 656 nm, 2 AMU

Magnetron turned offMagnetron turned off

Initial H value higher due to water vapour disassociation




• Optix is more sensitive to water vapour than differentially pumped RGA’s• Water vapour - 309.6 nm (OH), 18 AMU

Magnetron turned off and on –moisture increases as gettering is stopped – RGA not sensitive enough to detect the event

Case study 2: AlOx magnetron sputter deposition on roll-to-roll web

• Characterising an AlOx magnetron sputter deposition on roll-to-roll web

• Roll-to-roll deposition of reactively sputtered AlOx onto 125µm PET

• Optix sensor teed with a differentially pumped RGA


Courtesy of Emerson and Renwick


• Pump-down monitoring of different zones within a web coater

• Different zones in the vacuum can have different pump-down characteristics


Data courtesy of Emerson and Renwick, UK

Application: Plasma pre-treatment of plastic web based substrate

• Pre-treatment prior to coating is essential for adhesion in many vacuum coating processes

• The effectiveness of the pre-treatment can be monitored on-line



Application: Plasma pre-treatment of plastic web based substrate

• Process pressure 4x10-2 mbar

• Different materials have different treatment requirements

• The pre-treatment can be controlled on-line for each material




• Very large H outgassing – taking significant time to reach steady state

• Other species also observed initially outgassing – OH, CO2, O

• Subsequent power increases cause increased H outgassing and additional settling time

• Consumption of N2 also observed – small chamber leak



Target cleaningMagnetron on

1 kW 5 kW 10 kW 15 kW 20 kW


• Reactive sputter characterisation



Ar signal interaction with O2 flow

Target poisoned

Target de-poisoned

O2 controller tuning mode

Unstable feedback control

Stable feedback control


• There is a clear increase in CO2 observed when the target becomes poisoned


CO2 increase when target is poisoned


• Web speed varied to determine the source of the CO2

• No influence = cathode

• Influence = web


0.9

m/m

in

0.7

m/m

in

0.5

m/m

in

1 m

/min

1.5

m/m

in

3 m

/min

3 m

/min

5 m

/min

8 m

/min

11

m/m

in

1 m

/min

• Strong influence of web speed observed

• Inverse effect on O2 observed


Problem summary

• Main species observed developing are CO2 and N2

• Only occurs when target is poisoned

• Web speed influences the observed concentration


Hypotheses

• Negative oxygen ion bombardment of the substrate occurs when the target (cathode) becomes poisoned

• In this case, the bombardment can be liberating species from the film surface. This is a mix of hydrocarbons and nitrogen

• The hydrocarbons are reacting with the activated oxygen process gas to form CO2

Simulation of O- ions emanating from poisoned cathode

-400V

- O

- O

CxHy

O2 CO2

Cathode

Substrate (web)



RGA comparison – CO2

RGA comparison – O / O2



RGA comparison – OH / H2O

Start of target cleaning

Optix 309 nm OH


• 656 nm (H) and amu 1 (H) are generally a good match

• The difference at the start of the trace can be attributed to water vapour disassociating inside the Optix sensor into H, increasing the H reading


RGA comparison – HStart of target cleaning

Optix 656nm H

Case study 3 - Characterising a reactive ion etch

• Detection of reactive ion etching effluent in the process backing line


N2 purge

CF4

Ar

Etch head

Processing chamber

Roughing pump

Pressure 4x10-2 mbar

Optix


• Purge detection


N2 purge on


• CF4 detection (no Ar background)


40 sccms

20 sccms

10 sccms

5 sccms


• CF4 detection (Ar background)


40 sccms

20 sccms

10 sccms5 sccms

Case study 4 – Sensing from atmosphere

• Experimental setup


Optix

Pirani pressure gaugeNeedle valve

Rotary pump


• Solvent placed in close proximity to the inlet of the valve allowing detection of solvent vapour

• The spectrum was baselined at the background condition – therefore any changes in the spectrum are due to the detection of the vapour

• Three substances were tested – Isopropyl alcohol (IPA), Acetone and Naphtha (white spirit)




IPA Acetone

IPA Acetone

IPAAcetone

Naphtha Naphtha

Acetone has a higher CO reading due to the presence of a CO bond

IPA Acetone

Summary

• Remote PEM combined with spectroscopy can perform “RGA-like” functions

• Can use this method directly at higher process pressures – no need to differentially pump unless atmospheric sensing

• As case studies have shown, there is a clear benefit of improved monitoring of the vacuum at pressures used for industrial processing

• RPEM is less sensitive to contamination than RGA’s, can be used for ’dirty’ hydrocarbon environments as well as etch, CVD and ALD type processes

• This sensing technique offers a lower cost and complexity solution


gencoa optix optix a new instrument for robust gas analysis in vacuum november 2017 1

Documents