BEST KNOWN METHODS
Transpector® XPR3 Gas Analysis System
DESCRIPTION
The Transpector XPR3 is a third-generation, quadrupole-bad residual gas analyzer that operates at PVD process pressures and is the first process monitor with an Electron Multiplier (EM) that can operate at 10 mTorr operating pressures. The XPR3 does not require the large differential pumping system normally required for PVD process monitoring. The XPR3 can operate up to 20 mTorr, and it is linear at pressures up to 10 mTorr. The XPR3 measures major components and impurities common in a process with a 10 ppm detection limit.
Using the recommended Best Known Methods will provide you with a reliable Transpector XPR3 for process monitoring a high pressure application.
XPR3 APPLICATIONS
The XPR3 utilizes a High Pressure MicroChannel Plate Electron Multiplier (HPEM). The HPEM can be ud at lower pressures, such as background pressures, and it can also be ud at higher pressures, such as process pressures.
The XPR3 is typically ud for process monitoring of PVD applications. The applications normally operate in the mTorr range with backgrounds from 1e-6 to 1e-9 Torr. While the XPR3 can be ud for other applications where the process pressure is less than 10 mTorr, precautions should be ud. Applications that have high levels of hydrocarbon contamination or a significant amount of fluorines, chlorines or halogens are inappropriate for the XPR3.
PHYSICAL INSTALLATION
The XPR3 package includes an interlock weldment approximately 3.6" (91.44 mm) long with a VCR connection tube for the Pirani gauge. The Pirani gauge is ud for turning the XPR3 filament off above 20 mTorr and optionally turning the filament back on at pressures below the turn off point.
The XPR3 nsor mounts within the interlock weldment. The weldment must be mounted to the proc
ess chamber via a 90° valve (or mitered elbow). This prevents any line of sight plasma from reaching the ion source plate, which prevents any material from depositing on the XPR3 nsor. A heating jacket is provided with the XPR3 package and should be installed over the interlock weldment such that the cable is oriented as en in Figure 1.
Figure 1
Once the XPR3 nsor, electronics module, valve, and Pirani gauge are installed, the valve should be opened to allow the XPR3 to obtain high vacuum. It is strongly recommended that the XPR3 be kept under high vacuum conditions for at least eight hours before the filament is turned on. It is also recommended that the XPR3 be baked out with the supplied heating jacket (which operates at 150 °C) for a period of at least eight hours. This eight hour minimum bake out is required to reduce residual water vapor levels that may be higher due to local surface outgassing effects. The recommendations should be followed whenever the XPR3 nsor is expod to atmosphere for long periods of time and will rve to increa nsor life.
PIRANI SET-UP
Clicking first on the XPR3 icon from the TWare32™ main screen to reach the Sensor Properties screen, and then lecting the TSP Ur Settings tab will display the Pirani gauge t points. The Pressure Interlock Functions dialog is shown in Figure 2. The Emission OFF Pirani Interlock function is automatically enabled and cannot be disabled. The default (and maximum) value for emission off is 20 mTorr. The Pirani Auto Emission ON is
disabled by default, but can be enabled by checking the box and assigning a value less than or equal to 3.00e-3 Torr.
Figure 2
USING THE XPR3
Once the nsor has been conditioned, by baking it out and then keeping it under vacuum, the emission can be safely turned on. At
this point, typical us for the XPR3 would be leak detection, background monitoring, and process monitoring. The following are recommended parameters when operating the XPR3 in any of
the applications.
The ttings are reached from the Recipe Editor by choosing
Recipe Editor >> Sensor State >> Advanced Functions (e Figure 3).
Baline should be enabled by checking the box Baline Subtract On and choosing Spectra as the Baline Type . U the default Subtraction Mass of 9, 23, 33, and 47.
Linearization should always be ON, by checking the box. U the Factory determined values that have been programmed into the Transpector firmware.
Peak Lock should always be OFF , by not checking the box.
Figure 3
LEAK DETECTION
Using TWare32, there is no recipe required for operating in Leak Mode. Select the Leak Mode icon as pictured above to default to sampling Helium (Mass 4) over time. When leak checking a vacuum system that has a pressure of 1x10-5 Torr or lower, the HPEM should be ud. The HPEM voltage that is necessary is bad on the level of the leak that you are arching for. Adjust the HPEM voltage so that the Helium (Mass 4) signal can be obrved, but do not exceed an intensity of 1e-7 amps.
RECIPE GENERATION
Using the XPR3 for background monitoring or process monitoring is accomplished by creating and running a recipe. The XPR3 ur can generate the recipes, or sample recipes can be obtained from INFICON. The recipe file sizes are rather small (about 1 Kb) and can easily be e-mailed if desired. Plea contact INFICON by phone at (315) 434-1128 or by e-mail at
BACKGROUND MONITORING
BEST PRACTICAL DETECTION LIMITS BKG-BEST.RCP
For acquiring a spectrum where the full mass range is desired, the recipe parameters should be: Spectrum Scanning Mass Range 0-50 1 point per AMU Dwell time = 128 ms EM ON
EM voltage t for 300 gain The EM voltage is t for 300 gain at the Factory, and may require periodic adjustment depending on the current levels delivered to the EM.
Electron Energy = 40 eV
The above recipe will provide the best results for background monitoring, but will take approximately 11 conds for one scan. See Figure 4 for typical background results.
Figure 4
BACKGROUND MONITORING
FAST RESULTS - BKG-FAST.RCP
Faster scanning can be accomplished, but detection limit and accuracy will be sacrificed. For faster scanning, the dwell time can be reduced to 32 ms. This will lower the scan time to about 4 conds.
PVD PROCESS MONITORING WITH
ARGON PROCESS GAS:
BEST RESULTS - PRO-BEST.RCP
Since it is assumed that speed is very important in monitoring various gas during the process, Selected Peaks mode should be ud instead of Spectrum Scanning. The recipe parameters should be:
Electron Energy = 40 eV
EM ON with a gain t to 300. The EM voltage is t for 300 gain at the Factory, and may require periodic adjustment
depending on the current levels delivered to the EM.
While the mass to be sampled are customer lectable, the following mass and dwell times are recommended for a typical PVD application involving Argon process gas: (e Table 1). Table 1 Recommended Mass and Dwell Times for
Typical PVD Applications
that produces ion currents in excess of
1e-7 amps. For example, do not scan over
Mass 40 Argon in a PVD process involving
Argon gas since this mass will produce
large peaks. U Mass 36 Argon isotope as
a safe alternative. Scanning over any large
ion currents for extended periods of time
will damage the Electron Multiplier and
substantially shorten its lifetime.
This recipe will take about 3 conds per scan and will produce results similar to tho shown in Figure 5. Peaks for hydrocarbons can be added at mass 15 and/or 42.
Figure 5
PVD PROCESS MONITORING WITH
ARGON-NITROGEN PROCESS GAS
For metal-nitride process, the prence of nitrogen can also generate high ion currents at Mass 28. For the process, alternate recipes are recommended with peaks listed in Table 2. Table 2 Recommended Peaks for Metal-Nitride PVD Process Using Argon-Nitrogen Gas
Mass Species Dwell Time Multiplier
2H232 ms1
18H2O128ms1
28N2 / CO128 ms1
32O2128 ms1
36Ar3632 ms297*
44CO2128 ms1 Optional:
15 or 42
Hydrocarbons128 ms1
* Multiplier is derived from the natural abundance of Ar36 in
Argon gas: 100/.0337 = 297
Mass Species Dwell Multiplier 2H232 ms1
14N2 / CO128 ms25 **
18H20128ms1
32O2128 ms1
36Ar3632 ms297
44CO2128 ms1 Optional:
15 or 42
Hydrocarbons128 ms1
** Default value. Sensor specific value can be found by FC measurement of N2: Multiplier = I(28)/I(14)
that produces ion currents in excess of
1e-7 amps. For example, do not scan over
Mass 28 Nitrogen in a PVD process
involving Argon-Nitrogen gas since this
mass will produce large peaks. U Mass
14 Nitrogen isotope as a safe alternative.
Scanning over any large ion currents for
extended periods of time will damage the
Electron Multiplier and substantially
shorten its lifetime.
PVD PROCESS MONITORING WITH
ARGON-OXYGEN PROCESS GAS
For metal-oxide process, the prence of oxygen can also generate high ion currents at Mass 32. For the process, alternate recipes are recommended with peaks listed inTable 3. Table 3 Recommended Peaks for Metal-Oxide PVD Process Using Argon-Oxygen Gas
that produces ion currents in excess of
1e-7 amps. For example, do not scan over
Mass 32 Oxygen in a PVD process
involving Argon-Oxygen gas since this
mass will produce large peaks. U Mass
16 Oxygen isotope as a safe alternative.
Scanning over any large ion currents for
extended periods of time will damage the
Electron Multiplier and substantially
shorten its lifetime.PREVENTIVE MAINTENANCE
The XPR3 nsor has yttria-coated iridium filaments with a defined lifetime as well as a high-pressure electron multiplier that may degrade over time.
XPR3 FILAMENT
The XPR3 filaments should last a minimum of 4000 hours when following the Best Known Methods. It is strongly recommended that the filaments be replaced after 4000 hours of operation (a
pproximately six months of continuous operation).
allowed to burn out, coating from the
filament could contaminate the ion source
plate and create electrical shorts
preventing operation with a new t of
filaments.
The yttria-coated filament kit (INFICON part number 914-022-G2) is field replaceable. Replacement instructions are included in the filament kit and in the Transpector XPR3 Operating Manual (INFICON part number 074-378).
To determine how many hours the filaments have been operational, click on the XPR3 icon from the TWare32 main screen to reach the Sensor Properties screen and then lect the Maintenance tab. T
he information displayed is shown in Figure 6.
Figure 6
HIGH PRESSURE ELECTRON MULTIPLIER Since the HPEM is ud at background and process pressures, the EM hours will mirror tho of the emission hours. The HPEM gain may degrade over time and it is recommended to replace the EM when the EM voltage can no longer be adjusted to achieve a 300 gain. It is expected that the HPEM will last longer than one year, when ud continuously.
The HPEM degrades from monitoring high ion currents. Figure 7 shows estimated age for the EM as a function of process pressure for common gas mixtures and for monitoring different peaks. The plot highlights the incread lifetime advantage of measuring the recommended mass peaks. [The are
estimates bad on test data. Individual HPEMs may have different lifetimes depending on usage history.]
Mass Species Dwell Multiplier 2H232 ms1
16O2128 ms15 ***
18H20128ms1
28N2 / CO128 ms1
36Ar3632 ms297
44CO2128 ms1 Optional:
15 or 42
Hydrocarbons128 ms1
*** Default value. Sensor specific value can be found by FC measurement of O2: Multiplier = I(32)/I(16)
Figure 7
MASS SCALE TUNING
Another part of preventative maintenance is checking the
functional operation of the XPR3. This includes the mass position and mass resolution of the instrument. While this mass scale
tuning is accomplished in a similar fashion to any other
Transpector, the XPR3 does have some slightly different values for peak width adjustment.
The grid on the right of the screen shown in Figure 8 shows the typical mass from a factory calibration gas mixture. The TWare32 Operating Manual (INFICON part number 074-334) provides details on how to tune the resolution and mass position for any Transpector and the following ctions are to be ud for mass scale tuning in the field.
Figure 8
MASS SCALE TUNING AT BASE PRESSURE
Mass scale tuning can be done at ba pressure using the background peaks of water vapor (18 AMU) and Nitrogen (28 AMU). The following procedure should be ud to check peak location and peak widths at mass 18 and mass 28 and to make adjustments as needed.
86 AMU from the Tune Table or adjust the resolution at the mass.
1.Open the Tune window and t the points per AMU to 25 for all Tune mass.
2.If necessary, enable the Low energy tting: 40 eV (200 µA emission).
3.
Delete mass 40 from the Tune Table and inrt mass 18 AMU into the Tune Table.
4.
Turn on the Electron Multiplier so that the mass 18 and 28 peaks are visible in the 5e-11 amp range (or greater). It might be necessary to increa the dwell time so that the amount of noi on the peaks is reduced.
5.
Adjust the peak width and peak position of mass 18 and/or 28 AMU as needed. Set the peak width of the mass to 1.00 +/- 0.04 AMU wide at 10% of the peak height. Also t the peak position to nominal mass.
6.
Save the mass calibration upon exiting Tune mode.
NOTE:Do not attempt to add or delete Tune mass prior to
exiting Tune mode.
MASS SCALE TUNING WITH PROCESS GAS
For mass scale tuning at process pressures, the following
procedure should be ud to adjust the Argon (40 AMU) and/or Nitrogen (28 AMU). This tuning procedure can be ud for Argon, Argon-Nitrogen, or Argon-Oxygen process.
NOTE:For this mass scale tuning procedure, the Tune Mass
Table should be the default Tune list, which is mass 1, 2, 4, 28, 40, and 86. If this list is not prent when the Tune window is opened, modify the Tune Mass Table as necessary to show only mass 1, 2, 4, 28, 40, and 86.
86 AMU from the Tune Table or adjust the resolution at the mass.
1.
Open the Tune window and t the points per AMU to 25 for all Tune mass.
