0% found this document useful (0 votes)
4 views7 pages

Leak Checking Sources Service Note V002

Download as pdf or txt
Download as pdf or txt
Download as pdf or txt
You are on page 1/ 7

This service note describes how vacuum leaks are recognised and

located on the sources of EI/CI instruments.


Introduction
The necessity for a leak tight vacuum system on any mass spectrometer can not be over
stressed, in certain applications if a leak is present it may become impossible to obtain the
desired results. On systems fitted with EI/CI sources vacuum leaks are easily recognised by
a few simple checks and may be located with the help of the mass spectrometer itself.
General
Dry air is made up mainly of Nitrogen, 78·08% and Oxygen, 20·95 with other gasses making
up the rest 0·97%. Water vapour is an additional significant constituent of the normal
atmosphere we are exposed to and which varies in concentration widely due to a number of
factors, not the least being simply geographical location. Water is a very difficult molecule to
pump in a high vacuum as it adsorbs strongly to all surfaces within the vacuum system.
Nitrogen on the other hand is very efficiently pumped, 700 l/sec with the Edwards 160 Diffstak
pump. On a leak free system Water is the single most abundant material found in the high
vacuum system.
Test
By monitoring the ratio between Water and Nitrogen a very simple and accurate status of
vacuum may be gained. More so that by simply reading the source vacuum pressure, the
pumping system is very efficient and leaks tend to have little effect on the actual pressure
until being very large. This is particularly so when a GC column is fitted. A second
complimentary check would be to monitor the level of Argon present and compare it to it’s
neighbouring background peaks. Both these tests are valid, in fact become more important,
when a GC column is fitted as we are able to measure the leak tightness of the vacuum
system when a large amount of gas, Helium but sometimes Hydrogen, is entering the source
via the GC.
As mentioned above, on a leak tight vacuum system Water, 18·0106 Da, is the most
abundant compound found. By comparing it with Nitrogen, 28·0061 Da, the most abundant
constituent of the normal atmosphere, a good measure of the vacuum system may be made.
This may be summarised in the table below or graphically in the following diagrams.
Leak Water Nitrogen
18 Da 28 Da
None 3–5 1
Small 1 1
Troubling 1 3–5
Major 1 >10

Leak Organic Argon Organic


39 Da 40 Da 41 Da
None 1 <1 3
Small 1 1-3 3
Troubling 1 5 3
Major 1 >10 3

Document: Leak_Checking_Sources_Service_Note_V002.doc Page 1 of 7


Author : Max Letcher Date: 27/01/05
Graphical representations of the Nitrogen/Water and the Argon/“Background”
Ratio Vacuum Checks.

Leak Tight Good Max. Acceptable Leak Major Leak

2.5 1.2 3.5 12

3
1 10
2

2.5
0.8 8
1.5
2
0.6 6
1.5
1
0.4 4
1

0.5
0.2 2
0.5

0 0 0 0
H20 N2 H20 N2 H20 N2 H20 N2
Compound Com pound Com pound Com pound

Leak Tight Good Max. Acceptable Leak Major Leak

3.5 3.5 6 12

3 3
5 10

2.5 2.5

4 8

2 2

3 6

1.5 1.5

2 4
1 1

1 2
0.5 0.5

0 0 0 0
39 40 41 39 40 41 39 40 41 39 40 41

It should be noted that depending upon the instrument under investigation and the resolution
at which it is operating the actual appearance would vary. If the resolution in a sector field
instrument is greater than about 2000 Resolving Power then at mass 40, two distinct peaks
will be observed, that of Argon at lower mass and a higher mass "background" peak. At lower
resolutions a single poorly shaped peak will be observed. Similarly at mass 28 Da, with
increasing resolving power more substances may be separated out from the Nitrogen peak.
At 10,000 resolution one typically may distinguish 4 to 5 separate peaks depending upon
instrument usage.

Document: Leak_Checking_Sources_Service_Note_V002.doc Page 2 of 7


Author : Max Letcher Date: 27/01/05
Sources of leaks.
Any connection that has recently been opened is to be suspected as the source of the leak. If
no connections have been opened and a leak is suspected the following should be checked.
The reference gas inlet.
Check if it is open or closed. If the leak is present only when open, pump-out the
reservoir and monitor further. Check the condition of the septum.
GC. i.
Check a single air peak, normally that of Nitrogen, when the GC injector state is toggled
from Splitless to Split modes or vice versa and wait for “the retention time” of the column.
A 30m column will require about 1·5 to 2 minutes for air to pass through the column, a
60m column on the other hand between 3 to 5 minutes. If the injector is leaking then a
significant reduction in the Nitrogen peak and improvement in the Air ratio will be
observed when the injector is being purged, (in split mode for a Split/Splitless injector).
Here a small, “constant” flow of Helium passes from the injector into the column. The
rest, split, flow is vented out from the injector into the lab. This means that only a small
proportion of any material entering the injector will find it’s way in to the column and on
into the Mass Spectrometer’s source. When the injector however is in Splitless mode the
purge flow is stopped. The only flow through the injector is onto the column. Any air
entering the injector through the leaking position will enter onto the column thus
increasing and changing the ratios of the observed air peaks in the system. It is
completely normal even on leak tight injector systems that after switching from Split to
Splitless modes and waiting for the column’s retention time, a slow increase in the air
peak intensities over the period of two or so minutes will be observed. A maximum
allowable increase would be between double and treble the intensity. Should there be a
sudden, large change in intensity a major leak is to be suspected in or before the injector
needing further investigation. This is important for two reasons: 1. Injected samples will
be “split” out of the injector uncontrollably via the leak just as if we were making a split
injection therefore reducing sensitivity and reproducibility; 2. Air, more importantly Oxygen
and Water will enter the carrier gas flow and pass through the column at high temperature
significantly reducing the column’s lifetime/chromatographic performance.
GC. ii.
The column is interfaced into the vacuum system via the GC interface and
secured and sealed with a nut and Vespule ferrule. It is very often seen that after
ramping the GC’s temperature program, particularly when columns have been recently
fitted, that this connection may loosen and leak: New ferrules tend initially to distort
slightly and then settle down.

Leak finding:
If a leak in the source is suspected by monitoring either, or both, of the tests described above
it must be localised and corrected. Should a GC column be fitted then there will be a large
flow of generally Helium present in the source preventing it’s use when leak checking unless
the column is removed. To find very small leaks however this must be done but to check the
complete system including the GC we should arrange a controllable supply of Argon. This is
often present in the lab but if not, a small bottle, regulator and a length of preferably plastic
tubing terminated in an “Eppendorf” pipette tip will be required. A small gas flow is best for
accurate location: by dipping the tip/tube end into a small solvent (Ethanol) beaker of a good
idea of the flow will be gained. Select mass 40, 39·9624 Da, on the mass spectrometer at
any standard operating resolution and set the peak display such that normalise is off, a span
such that the peak nearly fills the scope and is about ¾ of the height of the display. It is
always a good idea to check that you really are monitoring the peak of interest, the easiest
way to do this is to inject a few microlitres of Argon into the reference inlet and ensure that the
peak you are monitoring increases accordingly. This should be done initially with a somewhat
larger span and then reducing it such that the Argon peak is clearly visible. Naturally, before
testing the Argon must be fully pumped away. To check: spray around each vacuum
connection/seal with Argon for a second or two, withdraw, wait and monitor the peak intensity
on the display. On large surfaces, view-ports, lids etc. the tip should be moved all around the
sealing surfaces. If a leak is present then at a certain point an increase in the Argon signal
will be observed. An idea of the size of the leak is made with two observations: 1. by how

Document: Leak_Checking_Sources_Service_Note_V002.doc Page 3 of 7


Author : Max Letcher Date: 27/01/05
much the Argon peak increases; 2. how quickly the Argon peak returns to the initial level.
Large leaks result in a very strong Argon signal but it will return back to the initial level quickly
as the leak is continually flushed with Air. It is always good to work from the bottom up with
Argon as being heavier than Air it will tend to sink down. Helium on the other hand will rise.
When working in this way it is important to note that the flow of gas does not limit itself to the
tip of the tube. It will tend to disperse in the air and we may see an increase in the Argon
peak when testing one particular connection but the actual leak is located somewhere near
by. Care and time should be taken when leak testing.
Checking the GC
When checking for leaks on the GC, the injector’s purge flow should always be stopped either
by setting it to Splitless mode or by setting a purge flow of 0 m/min. If the split (purge) is open
then the Argon entering the system will be flushed out of the injector before it can enter the
column and will therefore have a much reduced effect on the level of Argon entering the
source. This is particularly necessary when checking the injectors and the associated
tubing/control systems before the injector. Before attempting to check however, the Argon
peak must be allowed to stabilise under the Splitless conditions: i.e. wait for the retention time
and any further time required until the Argon peak no longer increases in intensity. Spray
over the top of the injector for several seconds, then withdraw the Argon flow and wait for the
retention time. If there is a leak, the Argon peak will be seen to rise. The bottom of the
injector can also be tested in this way. To remove the Argon quickly between tests the purge
can be opened for a short time, 10 to 20 seconds, and then closed again. Once more, after
the retention time of the column, the Argon peak will be seen to drop dramatically and then to
return to the elevated intensity before the previous test.
Leaks on the column itself do occur and may be found by passing Argon over the coiled
column several times, withdrawing and then waiting for a reaction. Such leaks are not
affected be the purge state of the injector and the Argon peak tends to respond quicker than
for injector leaks as depending upon the leak’s position on the column the Argon will simply
have to travel a shorter distance to reach the source.
Helium Leak Checking
As in the factory, the “absolute” vacuum tightness of the system is determined by Helium leak
checking. Obviously this must be carried out with the GC column removed and blanked off.
The mass to monitor is 4 Da, (4·0026 Da), it is imperative that the Electron Energy is set to 70
eV or the Helium will not be ionised efficiently. As the mass of Helium is so low it is important
to tune the source specifically for Helium and helpful to use low resolution. With Quad
systems it is sometimes difficult to monitor Helium successfully as the RF levels are very low
indeed; here Argon tends to be used exclusively. A very simple source of Helium with which
to leak check with is the GC itself: arrange a piece of tubing on the injector’s split purge outlet
and then program a split flow of some 20 to 30 ml/min. The flow can be checked by dipping
the tube end into solvent as before. Helium is very mobile in air so great caution needs to be
taken particularly when several connections are located close together to find the leak
correctly. Helium being lighter than Air rises so leak checking should work from the top of the
instrument down. Helium being so small and mobile will tend eventually to pass through a
“leak tight” graphite ferrule: it is normal that small “leaks” will be found on Graphite and to
some extent Vespule sealed connections.
Leak Checking CI Lines
For correct CI operation it is very important that the whole path the CI reactant gas must pass,
from bottle to source, is completely leak tight. If any Oxygen or Water enters the reactant gas
supply the CI conditions could be altered dramatically and inconsistently resulting in wildly
varying ionization processes leading in the worst case to no ionization at all. The cleanliness
of the reactant gas may also be adversely affected resulting in the source becoming dirty
much quicker. Additionally as a direct result of a leaking CI line the running costs will be
higher due to the need to replace the bottler more frequently and also very important to
consider: Health and Safety. The leaking reactant gas would escape into the lab/inside of the
instrument with it’s associated dangers.

Document: Leak_Checking_Sources_Service_Note_V002.doc Page 4 of 7


Author : Max Letcher Date: 27/01/05
A basic schematic showing how the CI gas is fed from the regulated supply bottle into the
mass spectrometer, regulated further and switched into the source and/or Inlet pump is
shown below.

Valve Description
1 CI Pump-out
2 Flow regulator
3 CI Gas On/Off

Analyser Source
Reactant Gas
Bottle/Regulator

CI Line to 1 2 3
Instrument

Inlets Pump

Basic checks should, and must, be carried out each time a bottle is changed or CI is to be
operated again after a long time interval.
Setting up:
• Ensure the gas bottle is valved off at the bottle itself.
• Select CI mode on the instrument and pump out the CI line. If necessary, open the
CI regulating valve/mass flow controller “fully” too allow its evacuation.
• Monitor the Inlets line pressure. On a leak tight line the Inlets pressure should
recover to approximately that before pumping out the CI line.
• Open the output of the bottle’s regulator valve, NOT the valve on the Bottle itself to
evacuate the regulator valve. Again the Inlets pressure should return to it’s initial
pressure. Note: Some regulator valves are not capable of holding a vacuum and will
leak through the valve when evacuated. In this case close off the regulator’s output
valve.
• Close the CI Pump-out Valve and monitor the Inlets pressure. No large improvement
should be seen for a leak tight system. On systems with a mass flow controller fitted
one can monitor the flow read-back with the line closed off, evacuated and with the
flow controller set to fully open. Should a leak be present a flow read-back will be
seen.
• Again open the CI Pump-out Valve and open the CI valve to connect the CI line
through into the source. When the Inlets line recovers, close the CI Pump-out valve.
• Monitor the source high vacuum reading. It should remain at the same pressure as
when the CI Pump-out valve was open. If not, a leak is to be suspected.
The complete CI line from the bottle/regulator connection to the ion source can be leak
checked when the system is in operate mode, set to monitor Argon or Helium, masses
40 Da or 4 Da respectively. Even though the instrument is in CI mode, possibly with a
dedicated CI source fitted, Argon or Helium may be monitored under EI conditions: until
we actually fill the CI source with reactant gas the ionisation processes present in fact EI.
Tuning is as normal, either start tuning on mass 28, 40, 2 or 4 Da depending upon signal
levels, set an emission current at 70eV Electron Energy of about 1000uA to ensure that a
reasonable electron beam passes through the inner source and then tune as normal. For

Document: Leak_Checking_Sources_Service_Note_V002.doc Page 5 of 7


Author : Max Letcher Date: 27/01/05
a CI only source there will be no Repeller to tune so the Ion Extraction / Source Focus 1
lens may tune very differently than compared to “true” CI or EI. When tuning is complete,
set the mass to the corresponding mass of Argon or Helium. Ensure that the CI pump-out
valve is closed, the CI valve and CI gas flow regulating valves are open and begin by
testing the bottle connections, waiting some minutes looking for a response on the peak
intensity. In this way we can work our way up the CI line testing all connections right up
to the source itself. When concluding the leak test, ensure that all the CI valves and the
(mass) flow controller valve have been closed before opening the bottle’s valves. If this
were the case, then the source and the Inlets pump would be exposed to an unusually
large amount of gas possibly causing the vacuum system to trip off.

Document: Leak_Checking_Sources_Service_Note_V002.doc Page 6 of 7


Author : Max Letcher Date: 27/01/05
Summarising.
The state of the source vacuum may be easily checked by monitoring the peak intensities
of the Water, Nitrogen and Argon peaks.
The two displays below show the expected ratios between firstly Water / Nitrogen and
secondly organic background peaks and Argon for a leak tight system.

Leak Tight Leak Tight

2.5 3.5

2.5

1.5

1
1.5

0.5 1

0.5

0
H20 N2
Compound 0
39 40 41

The following two displays show the same substances but when a major leak is present:

Major Leak Major Leak

12
12

10 10

8 8

6
6

4
4

0
H20 N2
0
Com pound 39 40 41

Quick Tests.
Are the intensities of “air” peaks affected by the state of the GC’s injector purge valves:
Yes Leak in or before the injector including the column connection to the injector.
No Leak not associated with the Injector system /column connection.
Do the air levels/ratios improve when the Reference Gas is valved off:
Trapped air/Leaking Septum/Leak in the Reference gas system. Pump out try
again/investigate further
None of the above:
Leak associated with the remaining system requiring further investigation.
How strong a response / quickly does it return to initial conditions is observed:
Large signal quick to return to the initial level. Big Leak
Weak signal slow to return to the initial level. Small Leak
Useful Substances / Masses.
Hydrogen H2 2 Da 2·0157 Da
Helium He 4 Da 4·0026 Da
Water H2O 18 Da 18·0106 Da
Nitrogen N2 28 Da 28·0061 Da
Argon Ar 40 Da 39·9624 Da

Document: Leak_Checking_Sources_Service_Note_V002.doc Page 7 of 7


Author : Max Letcher Date: 27/01/05

You might also like