Mass Spectrometer

Troubleshooting
What to Check When Your Results
Go Wrong

Angela Smith Henry, Ph.D.

Applications Chemist
CSD Supplies Division

1 November 29, 2018

How do we know something is wrong?

Recent Chromatogram
I have a “gut
feeling.” Too vague

My results
Compared to what? don’t look
right.

Known (early in sequence) Chromatogram

My current We can do
data doesn’t
look like my something about
typical data. that!

2 November 29, 2018

What to Check When Your Results Go Wrong: High backgrounds?
Common ailments of Mass Spec

• High background
• Noisy background
• Too many peaks!
Noisy backgrounds or too many peaks?
• Diminished peaks
• Few or no peaks
• Poor data quality (peak
shape, etc.)

Lower response? Few or no peaks?

3 November 29, 2018

My GC/MS has a high background!? What could be the culprit?

Questions to ask:
• Is it really a high background?
• Did I just pump down the instrument?
• What gas quality do I have?
10x increase in background! • Does the system have gas filters?
• Have I checked them recently?
• What do my nitrogen, oxygen and water
backgrounds look like?
• What do they normally look like?
• Do I have any leaks?

4 November 29, 2018

 Decreased retention
Let’s talk about Gas Quality and Filters

Reduced response
• Oxygen in carrier gas is detrimental to GC/MS
• Reduced response
• Elevated background
• Irreversible column damage
• Impaired electron multiplier function
• Premature filament, liner lifetime

• Use UHP carrier gases

• 99.9995% or greater
• Use GasClean carrier gas filters ~5% O2

GC/MS filter
Agilent P/N
CP17973

Page 5
Let’s talk about Gas Quality and Filters:
If you used lower quality gases, how much O2 could the filter clean up?

Extracted Ion Chromatogram (EIC) for O2

He

Installing (and properly purging) the GasClean carrier

gas filter lowered the O2 signal by a factor of 5!

Page 6
Let’s talk about Gas Quality and Filters:
If you used lower quality gases, how much H2O could the filter clean up?

Extracted Ion Chromatogram (EIC) for H2O

He
GasClean filter lowered the H2O signal by a factor >10!
Further increase GasClean and column lifetime with
universal filter installed before GasClean filter

Page 7
Let’s talk about Gas Quality and Filters:
If lower quality gases were used, how much background could the filter clean up?
GC/MS filter
Agilent P/N
CP17973

Big Universal Trap He

RMSH-2

Mounting bracket
UMC-5-2

• Install the Universal trap vertically – use the mounting bracket(s)

• Extend the lifetime of your GasClean (indicating) filter AND (most importantly) your column!

Page 8
I have gas filters and high quality gas, and I still have a high
background!
Use manual tune (before any experiments!) to check for
leaks/background
• Ions 18, 28, 32 m/z

9 November 29, 2018

I have gas filters and high quality gas, and I still have a high
background!
Use manual tune (before any experiments!) to check for
leaks/background
• Ions 18, 28, 32 m/z
60,000 counts for N2 is definitely high.

What to check:
1. Verify gas fittings are leak-free.
Water/methanol mixture (external to GC/MS ONLY!)
Leak detector

Agilent G3388B
Leak Detector

10 November 29, 2018

I have gas filters and high quality gas, and I still have a high
background!
Use manual tune (before any experiments!) to check for
leaks/background
• Ions 18, 28, 32 m/z
60,000 counts for N2 is definitely high.

What to check:
1. Verify gas fittings are leak-free.
My fittings outside of the GC/MS are leak-free, what
else is there to check?

11 November 29, 2018

I have gas filters and high quality gas, and I still have a high
background!
Use manual tune (before any experiments!) to check for
leaks/background
• Ions 18, 28, 32 m/z
60,000 counts for N2 is definitely high.
What to check:
1. Gas line fittings (Done)
2. Check the vent valve, MSD transfer line nut and side
door.
How?

12 November 29, 2018

Use leak detector and/or electronics duster to find your leaks!
Why use a leak detector? Typical Electronic Duster
Components and Ions
• High sensitivity
1,1- m/z

• Recommended for leak detection in gas difluoroethane 51,65

plumbing and fittings 1,1,1- m/z

69
trifluoroethane

1,1,1,2- m/z
tetrafluoroethane 69,83

Use electronics duster

• Hold can upright (don’t spray liquid!)
• Spray short bursts around possible leak points
Agilent G3388B Leak • “Live” tune profiling for ions to pinpoint leak
Detector

link

Page 13
Using electronics duster to find system leaks: Manual Tune

Navigate to MSD Manual Tune in the Data

Acquisition
• Instrument > Edit Tune Parameters
Or
• View > Tune Vacuum Control
• Parameters > Manual Tune

MassHunter Data Acquisition

14 November 29, 2018

Using electronics duster to find system leaks: Manual Tune

• Use Profile tab to watch the main ions (69 and 83 m/z for my electronics duster)
• Spray short bursts at vent valve, transfer line, and side door

15 November 29, 2018

Forgetting to retighten - or overtightening - the vent valve
• How does this happen?
• Easily overlooked
• Urge to overtighten fittings
• Why is it bad?
• Overtightening compresses o-ring
• Prevents leak-free seal formation
• How is it fixed?
• After venting, retighten vent valve before maintenance
activity
Agilent MSD

• Tighten gently to avoid cross-threading Vent Valve o-ring

P/N 0905-1014
link
• If leaks are persistent, replace vent valve o-ring

Page 16
Check the Transfer Line Nut for Leaks
• Using a graphite/vespel ferrule, right?
• Installed with the flat end of ferrule facing the MSD?

• Nut “loosens”* with heat cycles

• If you find a leak:
• Tighten in small increments and then check again until no leak
• Try to not overtighten the nut
– If you have to apply a lot of pressure, vent and check the ferrule/threads

Crushed end of column

Be careful with the transfer line nut!
• Over-tightening damages transfer line threads, column
• Audible squeaking → overtightening

Page 17
I always forget to check my transfer line nut the morning after a
pump down….
• Switch to self-tightening MS column nuts
• Spring-driven piston continuously presses
against ferrule
• Easily installed by hand
• Longer column lifetime, lower column bleed
• ONLY use the short graphite vespel ferrules
– Install with flat edge TOWARDS MSD

Self-tightening Nut for Agilent MS transfer line: P/N 5190-5233

0.25mm column short graphite vespel ferrules: P/N 5181-3323

Page 18
We’ve checked the vent valve and MSD transfer line, is there
anywhere else?

Don’t forget the side door!

…if leaks are found at side door, you will need to vent.

19 November 29, 2018

Good Habit Tip # 1: Check the side panel BEFORE you close the door

Side door o-ring was dislodged; how do I prevent an

unnecessary vent?
• Using gloves, run finger along o-ring in groove
• Visually inspect seal integrity prior to pump down
• Wipe o-ring perimeter with lint-free tissue
• Replace worn or damaged o-rings

Other good habits

• Don’t use the side panel thumbscrews
• Only use during shipment/relocation
Agilent GCMS Side
• Close vent valve before opening side panel door plate o-ring
P/N 0905-1442
• Periodically vacuum beneath outer covers link

• Proper cooling for turbo pump and boards

Notch for o-ring removal

20 November 29, 2018

Good Habit Tip #2: Check other connection points in the GC

• Check the inlet

• Installed with a quarter turn with a wrench?
• Inlet nut may need slight tightening after
heating cycles
• Or, pre-swage flexi-metal ferrules
• Or, use Self-tightening column nut for inlet
– Only use graphite/vespel ferrules with self-
tightening nuts!
• Check CFT connections
For inlet or detector
P/N 5190-6194
• Use leak detector or electronics duster
• Short bursts of electronics duster at each
connection point

21 November 29, 2018

If my system is leak-free, what should my air ion abundances be?

• These are just estimates!

• H2O: ~2,000 counts (less is ok)
• N2: ~10,000 counts(less is ok)*
• O2: ~3,000 counts (less is ok)
*Make sure to purge your GasClean filter

High vacuum gauge pressure (for SQ):

~1 x10-5 torr†
† dependent on flow rate

22 November 29, 2018

One Last High Background discussion: Troubleshooting Column Bleed

Have you installed and/or conditioned the

Column Bleed column?

Are you exceeding the column’s upper

temperature limit?

Is your column’s film size too thick?

Could leaks be present in your flow path, or

Original Chromatogram are your carrier gases contaminated with air?

Do you need to change your split vent trap?

Good Habit Tip #3: Save a chromatogram from immediately after the column was installed. Overlay the
problematic chromatogram with your reference chromatogram to determine whether column bleed may be a
problem.
23 November 29, 2018
“Noisy” Background and Too Many Peaks
How do you know it’s noisy?

24 November 29, 2018

Good habit tip #4: Have system “baselines”

Known standard: Agilent Semivolatiles Checkout Standard 5190-0473

What kind of system baselines?

• System blank – Should only see a rise in baseline with temperature (column bleed)
• Solvent blank – May contain contaminant peaks (e.g. phthalates, siloxanes, etc.)
• Best practice: Use the same bottle of solvent that was used for any dilutions/extractions
• Known Standard – GC/MS checkout standard, DFTPP tuning mix, or known calibration
standard at easily detectable level for your system (e.g. 1-10ppm)

25 November 29, 2018

Good Habit Tip #5: Run a manual scan to review background peaks
Understand what the normal scan range background looks like
• Look at your normal scan range and an
extended scan range
• 35-500 and 10-700
• Use your normal threshold
• Perform the scan after initial set-up and
after major changes (new column, source
clean, etc).
• Track the results in notebook or on computer

• Normal # of peaks in manual scan: ~100-

250 peaks

26 November 29, 2018

Does the noisy background negatively affect my results?

• Do compounds fall the

quantitative criteria? 380000
Run 1
• Can you distinguish target 340000 Run 102

compounds from matrix? 300000

• Can you integrate the peaks 260000

from EICs? 220000

180000

140000

• If your results fail criteria, 100000

something(s) is/are dirty 60000

20000
13 14 15 16 17 18 19 20 min

Does this mean I have to

clean my source?

27 November 29, 2018

Check these before you vent for the source….

• Do the easy steps first! Does your liner look clean or dirty?

• Liner and inlet septum

• Change these first
• Did the chromatography improve on your check
sample?
• Trim head of column or install new guard
column
• Trim ~0.5m from front of column
• Use a guard column
• Did the results improve on the check sample?
• If chromatography hasn’t improved, it may be
time to check and clean the source….

28 November 29, 2018

Too Many Peaks Case Study: What are these repeating peaks?

Is it column bleed?
Are these peaks from my solvent?

29 11/29/2018
Septum maintenance: TIC of an inlet septum

Septa contamination in wash vials or inlet liners can be diagnosed by looking for siloxane polymers in your total ion
Common ions for
siloxane molecules: chromatogram. Each peak in the chromatogram corresponds to a cyclized (ring structure) siloxane molecule. These molecules
fragment with very similar patterns.
73
Example spectrum:
147
207
281
355

30 November 29, 2018

My peaks are much smaller than I expected; what’s going on?
Diminished peaks

32 November 29, 2018

 Remember the Matrix: SIM or MRM can hide a lot
When you only look at EICs, you can miss a large picture

500ppb ethofenprox 1.7

+ EIC(93.1) Scan Extract_00003.D

in black tea 1.6

1.5
1.4
1.3
1.2 7.210
1.1
1
0.9
0.8
0.7
0.6
x10 1 +EIC(93.1) Scan Extract_00003.D 0.5
4.2
0.4
4
3.8 0.3
3.6
3.4
10ppb linalool in 0.2
6.978
0.1
3.2
3
2.8
shampoo 0
-0.1
2.6
-0.2
2.4
2.2 -0.3
2 -0.4
1.8
6.9 6.95 7 7.05 7.1 7.15 7.2 7.25 7.3 7.35 7.4 7.45 7.5 7.55 7.6 7.65
1.6
Counts (%) vs. Acquisition Time (min)
1.4
1.2
1
0.8
0.6
0.4
0.2
0 …(or improve your sample cleanup)
-0.2
-0.4
-0.6
-0.8
-1

3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 10.5 11 11.5 12 12.5 13 13.5 14 14.5 15 15.5
Counts (%) vs. Acquisition Tim e (min)

33 11/29/2018
What should I check when my peak response is diminished?

Run 1
Run 42
Where’d my response go?

• Check for leaks

• Liner and inlet septum
• Trim head of column
• If column bleed has significantly increased or matrix interferences cannot be
removed from column, replace the column
• If chromatography hasn’t improved, it may be time to check and clean the source
• While the system is open, take a look at your filaments….

34 November 29, 2018

Good habit tip #5: Filament care: Have an extra pair (or two) on hand

• Have (at least) 2 extra filaments on hand Filament may fail soon
• More than 1 GC/MS system? Keep >2 on hand, depending on the
number of systems.
• Check filaments when you clean the source
• Look for discoloration behind the filament and unraveling of the coil
• Replace them as a pair
• End-of-life filaments may cause diminished response or odd
artifacts in TIC
• Keep them, just in case the problem is not the filaments

Careful! High Efficiency source (5977B HES single quad MS and Agilent 5977 InertPlus,
7010 HES tandem quad MS/MS) have different filament designs from Extractor, & 5975 Filament
5977B InertPlus, extractor source and older MSD designs! Assemblies: G7005-6001

35 November 29, 2018

Fewer peaks than expected or no peaks: What should we check first?
• Easy stuff first!
• Check syringe, inlet septum and liner
• Check the vacuum reading
– Is it normal?
• Verify column is intact
• Verify all cables, etc. are connected and method parameters are normal
• Do the filaments turn on?
• Run the known standard
– Does it look normal?
• If these steps don’t work, then it’s probably time to call for the lab manager and/or Online Support

36 November 29, 2018

Data quality
How do I know if my flow conditions, gain, scan speed, etc. is right?

37 November 29, 2018

Data quality: What are the best flow conditions to use for MSD?

Use constant flow methods!

5977 Extractor, InertPlus, or older sources

• Suggested flow rate range: 1 - 2 mL/min
• Optimal flow: 1 - 1.2 mL/min

5977B High Efficiency (HES) source

• Suggested flow rate range 1 - 1.5 mL/min
• Optimal flow: 1 - 1.2 mL/min

38 November 29, 2018

Data quality: How do I know if my parameters are right?
What is the proper column installation into MSD?
5977 Extractor or InertPlus, or 5975 MSD 5977B HES
Installation length: 1-2mm beyond end of Installation length: 5 mm beyond end of
transfer line transfer line
5977 HES
5 mm beyond transfer line

1-2 mm beyond transfer line

5977 InertPlus or older MSD

Remember to re-install the ceramic tip, spring, and nut after column installation!

Ceramic tip for 5977 and 70x0 series Transfer line spring Transfer line cap
G3870-20542 G7005-20024 G3870-20543

39 November 29, 2018

Data quality: How do I know if my scan speed is right?

Best acquisition frequency

• ~ 2.5 to 5 Hz

Optimal # of points across peak

• 8-12+ points across a peak

Set scan speed at N=1 or N=2

• Frequency in 2.5-5 Hz range

• Generally most samples will have 8-12+ points
across a peak

What about SIM dwell time?

• Default 100 ms = good start

• Depends on number of ions in time segment
• Watch Frequency (scan/sec) as you add ions and/or
change dwell time

Gain selection: https://www.agilent.com/cs/library/technicaloverviews/public/5991-2105EN.pdf

40 November 29, 2018

Data quality: How do I know if my gain is right?

• Good practice: Start at gain 1 Process for testing gain levels: Do I need to
increase my gain?
• Best Practice: Choose lowest gain factor
for detection of most and least intense • Test gain factor in highest and lowest
ions over target concentration range concentration standards
• Ideal gain could be < 1 • Choose most intense compound/ion
• Lowest suggestion gain: 0.3 • Run known standard at highest concentration
with Gain = 1 (or current gain)
• Avoid high gain for long periods of time
• Is largest ion peak between 3x106 – 6x106
• High gain can shorten life of EM counts?
• If yes, run lowest concentration standard
– Are all compounds detectable?
• If no, try a different drawout lens or
increase/decrease gain (e.g. from gain 1 to
gain 2)
– Repeat process

Gain selection: https://www.agilent.com/cs/library/technicaloverviews/public/5991-2105EN.pdf

41 November 29, 2018
 Thank you for your attention!

START FINISH
Step 1 Step 2 Step 3 Step 4 Step 5

What does a Develop steps

Half-split the Make repairs, as Put the system
working GC/MS to prevent re-
problem necessary back together
look like? occurrence

Use the WebEx Chat functionality

to submit your questions
Column Installation: Self-Tightening Column Nut

For inlet or detector • Spring driven piston continuously presses

p/n 5190-6194 against ferrule

• Automatically retightens when ferrule shrinks

• No leaks! No downtime! No frustration!

• Wing design for finger tightening

• No tools needed!

• No polymer materials for durability

• Compatible with ONLY short Graphite

• Vespel ferrules
For mass spec transfer line
November 29, 2018
p/n 5190-5233

43 Confidentiality Label
What steps should I follow when troubleshooting?

44 November 29, 2018

Follow a logical troubleshooting procedure

START FINISH
Step 1 Step 2 Step 3 Step 4 Step 5

What does a Develop steps

Half-split the Make repairs, as Put the system
working GC/MS to prevent re-
problem necessary back together
look like? occurrence
Troubleshooting starting points
• Do method parameters match what you see in/on the system
Check
parameters

• Are sample/wash vials dirty or have particulates?

• Watch syringe moving/make an injection, or remove syringe and manually pull up solution.
Autosampler • Try a manual injection to rule out the autosampler

• Determine if users have completed normal maintenance recently

Recent • Change liner, inlet septum, syringe, trim column
Maintenance?

• Generate a tune report to see if the current tune is working. This does not change any
Quick MSD parameters.
check

• Inject a checkout sample you know

Sample you • Do you observe the expected peaks in the expected sizes at the expected retention times?
know