January 2017

SPEXCertiPrep.com/ECS

Diagnosing and Resolving

GC Problems
by Mark A. Ferry
GC/MS Technical Consultant
[email protected]

A publication by
Environmental Consulting & Supplies
a Division of
Optimize
January 2017

A publication by ECS, a division of SPEX CertiPrep®

203 Norcross Ave.
Metuchen, NJ 08840
Phone: 732-549-7144
Email: [email protected]
Web: www.spexcertiprep.com/ECS
Copyright © 2017 SPEX CertiPrep®
 Diagnosing and Resolving
GC Problems

For all discussions, we will assume you are running QQ It’s a very good idea to purchase a hand-held
Volatiles by GC/MS (524, 624, 8260) and/or Semivolatiles Helium leak detector. It’s a lot to spend I
by GC/MS (525, 625, 8270). Other non-Mass Spec know but it’s an invaluable tool when trying
to find small leaks.
applications are not addressed here. Since it is rare for
these units to “break”, the majority of problems with QQ Never try to detect leaks by squirting a fitting
with soapy water to see if it bubbles. You can
them ends up being with conditions being set incorrectly get some of the soap into the lines and that
or for the system to develop leaks and/or active sites. would be disastrous.
Detecting and Resolving Leaks: If a leak does occur in your inlet, check the following:
Here are some good ideas to follow in order to QQ Carefully inspect the glass liner. If it is
chipped, that would cause a leak.
detect and fix leaks:
QQ Check the Viton O-ring that seals the liner. It
QQ The MSD can detect leaks where air is being
should feel elastic. If in doubt, change it.
sucked into the ion source (by virtue of the
presence of large amounts of ions 28 and QQ Replace the septum if it has been punctured
32), but it has no ability whatsoever to detect many times or has been crushed out of shape.
leaks where Helium and analytes are being Be careful not to over tighten the septum
lost OUT of the system. To detect these leaks nut as that can cause the septum to become
you need to monitor flows and/or use a leak crushed. 1/4 turn past finger tight should do
detector. it on the septum nut.
QQ Once per month, when the system is running QQ Replace the inlet seal if the face seems worn
optimally, measure the split vent flow and out or warped. This sometimes happens if it
record it in a log book. Monitoring changes has been used and reused a lot or has had the
in split vent flow is the best way to detect face cleaned to remove debris.
leaks. This applies to 5890 GCs. 6890 GCs QQ Re-seal the fitting from the column to the
will automatically add more flow if a leak inlet. Be careful not to over tighten the fitting,
exists, so in this case if you notice reduced as this can cause the ferrule to get cracked.
areas of your analytes, you will need to check
the inlet with a Helium leak detector. Coping with Active Sites
QQ Whenever you dismantle the injector to do Another common problem with inlets is the
maintenance, ALWAYS measure the split vent
flow immediately before and immediately development of active sites. An active site is
afterwards. If you have done it correctly, they any buildup of contamination that would cause
will be identical. If not, you need to check a compound to react with or adsorb to, thereby
everything you did prior to heating up the compromising the integrity of the analysis.
inlet and running the system. As above, this
applies to 5890 GCs. For 6890 GCs you will For Volatiles, active sites in the inlet generally only
need to check the inlet with a Helium leak occur if a sample foams over. Otherwise, you would
detector. typically only do your injector maintenance once a
QQ Monitor the areas of your first Internal year on average.
Standard. If the areas drop off, yet the For Semivolatile analysis, fighting active sites is
sensitivity of the MSD (as measured by the
raw abundance of ion 69 in Manual Tune) is nearly an every day battle because of the beating the
constant, check for leaks. For Semivolatile inlet takes by having numerous Methylene Chloride
analysis the leak is likely to be in the inlet. extracts injected into it. Active sites in the inlet can
For Volatiles, it’s more likely to be in the cause many compounds (especially Benzidine and the
purge and trap than the GC so you should Phenolics) to show drastically reduced response.
leak check the purge and trap.

January 2017 Optimize 3

Here are some ways to minimize and/or cope with vial, dry the liner in a drying oven (120 °C) and keep
active sites for Semivolatiles: the dried liners air-tight. Be careful not to chip the
glass liners when handling them as the slightest chip
QQ Use common sense before injecting a dirty
looking extract. If an extract is very brown, will cause a leak when inserted into the injection port.
or very viscous, or came from a sample After running “dirty extracts”, cool the injection
you know is loaded, dilute the extract 1:10 port, remove the plug, rinse the liner (do not use
and then run it. Based on this run you can
determine if it’s safe to run a full strength. It’s detergents) and repeat the process. Note: HCl gas is
a good rule of thumb to “dilute the extract given off as a by-product of the reactions so handle
first, ask questions later”. the vials with care and store them in a fume hood:
QQ Having an inert inlet sleeve is essential. ●● Clip off 1/5 meter of column (8 inches)
Unless you are running VERY clean samples, daily or when needed. If peak shape or
you would probably benefit from using responses are poor, clip off 2 loops.
silanized fused silica wool/inlet sleeve
combination. ●● Change autosampler rinse solvents
daily. It’s Ok to use Methanol if the BNA
QQ Buying silanized liners is a very costly
GC/MS is in the same room with the
procedure. I suggest you follow the
procedure listed below and do it yourself. VOA GC/MS. If not, use Methylene
All you need to do is buy a 5-pack of 4 mm Chloride as rinse solvent #1 and Acetone
single gooseneck liners and a tin of untreated as rinse solvent #2.
fused silica wool. Then, make 4 solutions in
4 vials as listed below and do it yourself. It’s ●● Avoid injecting more than 1.0 uL into
a very simple and inexpensive way to always H-P split/splitless injectors. Injecting
have deactivated liners and plugs on-hand. 2.0 uL generally only increases areas by
You can re-use the liners; simply remove the about 30% but peak shape suffers from
plugs each day (or as needed after running a
batch of dirty extracts), rinse with Methanol, overloading the liner. If you have a
dry them, insert a new plug and repeat the 5 uL syringe you can inject 0.5 or 1.5 uL,
entire silanizing process. If you want an so this may be a good solution if you’re
explanation as to the chemistry involved in having sensitivity problems.
the deactivation, simply e-mail me and I’ll
send it to you. GC Problem: Injector pneumatics
Procedure to Silanize Injection Port Liners OK, let’s say you’re still having GC problems even
though you’ve installed a new column and properly
for Semivolatiles done all the injector maintenance. The problem could
Step 1: You need to prepare 4 vials (40 mL VOA vials very well be the injector pneumatics rather than the
are ideal) as follows: injector consumables. Injection ports pneumatics
Vial 1: 10% DCDMS (Dichlorodimethyl Silane) in periodically need to be replaced because the lines and
Toluene. 36 mLs Toluene + 4 mLs DCDMS. You can valves become contaminated with non-semivolatile
get DCDMS from Sigma-Aldrich. The percentage residue.
need not be exact. Here are the components of the split/splitless
Vial 2: Toluene (as a rinse) pneumatics for the 5890 GC:
Vial 3: Initial Methanol ●● 5890 shell weldment
Vial 4: Final Methanol ●● 5890 splitter tube
Step 2: You need to place a fused silica wool plug in ●● 5890 split/splitless solenoid valve
the liner BEFORE silanizing. Use a 4 mm ID liner ●● 5890 split weldment
with a small, loosely packed fused silica wool plug
●● 5890 EPC Proportional control valve
about 1/4” long
It’s a good idea to periodically clean the shell
Steps 3-7: Simply soak the liner with plug in each
weldment. To do this, first cool the inlet. Then
vial for two hours or so-from Vial 1 to Vial 4 (if time
remove the column. Then remove the liner, O-ring
permits allow them to soak overnight). After the 4th
and inlet disc. Then take a bristle test-tube brush and

4 Optimize January 2017 SPEXCertiPrep.com/ECS

scrub the inside of the shell weldment with Acetone. GC Problem: Column and conditions
This will often help remove any dried-on debris that
has accumulated on the inside surface. If it still looks Another common GC problem would be defective
grimy to you after cleaning, it’s worth the money to column, improper installation, or non-optimized
replace it. conditions. At this point I want to discuss the
flows and pressures of the split/splitless injector for
The splitter tube and split weldment are next to semivolatiles analyses.
impossible to clean and are inexpensive enough to
justify replacement. Although the flow path with the 6890 is slightly
different than the 5890, for this discussion the points I
The EPC valve and split/splitless solenoid valve cannot want to make apply to both.
be cleaned. You would suspect that the EPC valve
needs replacement if the GC cannot maintain the It’s important to understand the flow path of the split/
set head pressure (this sometimes makes the system splitless injection port in order to optimize conditions
“beep”). If you don’t have a leak, then a defective EPC and troubleshoot problems.
valve is probably the culprit with this problem. Let’s assume the TOTAL FLOW into the injector is
Here are the components of the split/splitless 43 mL/min. We can’t measure that directly, but can
pneumatics for the 6890 GC: easily compute it by adding the 3 flows into which the
total flow splits. The flow comes into the split/splitless
1. 6890 shell weldment
injection port and then is routed in 3 directions as
2. 6890 splitter tube (1/8” copper) follows:
3. 6890 split gas filter
1. About 2.0 mL/min hovers under the septum
4. 6890 split weldment and exists via the septum purge vent. For
5. 6890 split/splitless flow module- (integrated EPC systems it is set to about 2.0 mL/min;
flow/pressure control assembly which slides for non-EPC systems it has to be set manually
in slot on top of the GC) and we recommend about 2 mL/min. This
flow can be measured via a flow meter.
Part #5 above is very expensive, so try a “mini”
injector rebuild by replacing parts #1-4 above. If that 2. About 1.0 mL/min enters the column. The
fails to restore performance, or if the system cannot exact flow is a combination of column ID,
hold head pressure, then replace part #5 as well. head pressure applied and temperature
programming used. Typically, a 0.25 mm ID
Another way to know you have contaminated
column is used with head pressures ranging
pneumatics is by monitoring Benzidine. If your
from 10-25 psi, the column flow ends up
Benzidine response is very low when you run a
being about 1.0 mL/minute on the average, so
50 ppm full 625/8270 standard, run the a 50 ppm
we’ll use that figure in our calculation. We can
Benzidines standard alone. If the area is much higher
measure this flow by doing a linear velocity
in the Benzidines standard alone, then that’s a good
check, but that’s usually not necessary as we
hint your injector pneumatics are contaminated. If
can estimate the flow to be about 1.0 mL/
you have EPC, you probably won’t need a new EPC
minute.
valve as long as the system is holding head pressure.
3. The rest of the flow, called the split vent flow,
Final note: Replacing injector pneumatics can be
passes through the split/splitless injection
a costly expense, but one that’s well worth it. If you
solenoid and exists via the split/splitless vent.
have been running your GC for methods 625/8270
For semivolatiles we typically set it to 40 mL/
for years, chances are the pneumatics are somewhat
min. This flow can be measured via a flow
contaminated. At some point they will become
meter. Here’s where the split/splitless part
so bad that the problems are insurmountable
comes in.
without overhauling the inlet. Installing new injector
pneumatics into a GC basically gives it a “new lease ●● In the SPLIT MODE (also called the Purge
on life” and should restore the GC to “like-new” “on” mode), the flow sweeps through the
inlet liner sleeve, then through the split/
condition. splitless injection solenoid and then out
the vent.

January 2017 Optimize 5

 ●● In the SPLITLESS MODE (also called 1.0 mins and achieving good results, stay with it. If
the Purge “off ” mode), most of the flow not, try 0.4, 0.5, 0.75 and 1.0 mins and see what gives
bypasses the inlet liner sleeve, then you the best overall results.
passes through the split/splitless injection
solenoid and then out the vent. There
is enough flow through the liner to Troubleshooting the Split/Splitless Injector
maintain head pressure. for Semivolatiles
When setting our method, we want the inlet in the Here are recommended inlet conditions for running
SPLITLESS MODE (also called the Purge “off ” mode) Method 625/8270
at time of injection and for however long it takes for Column: 30 m/0.25 mm ID/0.25 or
the solvent cloud to pass through the liner and enter 0.5 u Film thickness.
the column. Once that happens, we cycle the split/
Injection volume: 1.0 uL
splitless injection solenoid so as to have the inlet in
the SPLIT MODE (also called the Purge “on” mode), Liner size: 4 mm single gooseneck ID
thereby sweeping the inlet of any residual solvent, Injector temperature: 250 °C (be sure to use the
contamination, air, or other material that we DON’T insulation)
want in the column.
Head pressure: 12 psi for Non-EPC systems,
So, by using the split/splitless injection port in the for EPC systems I like to
SPLITLESS mode, we essentially allow virtually all pressure pulse or program
that is injected to enter the column. This helps us it; 16 psi for 30 seconds-
maximize the signal. Once the analytes are into the variable pressure to achieve
column, we cycle it to SPLIT mode, which keeps the 1.0 mL/min for balance of
inlet clean and reduces the noise we would otherwise run for 5970 and 1.2 mL/min
get. It’s a simple, yet elegant and effective system for for 5971-5.
obtaining good signal-to-noise chromatograms.
Assuming you have proper inlet conditions, the 2
So, the total flow into the injector is the sum of all 3 biggest problems you will face are leaks and active
sub-flows, namely 2.0 mL/min (septum purge vent) sites. As mentioned earlier, leaks can often be detected
+ 1.0 mL/min (carrier flow) + 40 mL/min (split vent by changes in split vent flow, so once per month, when
flow)= 43.0 mL/min TOTAL flow into the injector. the system is running optimally, measure the split
An important variable is Purge “On” Time. This is vent flow and record it in a log book.
the time that the system stays in the Splitless (Purge I like to monitor the response of Benzidine,
“Off ”) mode for at time of injection before cycling 2,4-Dinitrophenol and 4-Nitroaniline to determine
to the Split (Purge “On” Mode). Typically this is if active sites are present. If so, these 3 compounds
set anywhere between 0.4-1.0 mins. The way it is are among the first to suffer. I also monitor the
calculated is as follows: peak shape of 4-Nitrophenol as this is often the first
- Injection volume 1.0 uL compound to look “ragged” the column is in need of
replacement.
- Liner size 2 or 4 mm ID
Here’s a good idea whenever you install a new
- Injector temperature 250 °C
Semivolaties column. Before running your DFTPP
A 1 uL injection of Methylene Chloride will vaporize and curve, it would be wise to be sure the column
to about 500 uL (0.5 mL) of gas upon injection in the you just installed isn’t defective and has been installed
hot inlet. If our flow is 1.0 mL/minute, then it should correctly.
take 0.5 mins for that solvent cloud to pass through
After I install a new column, I prepare a solution
the liner and into the column. That is why 0.5 mins
of the 6 Internal Standards in Methylene Chloride
is typically used, although there’s a practical range of
at 40 ppm. You should get 6 sharp peaks with no
0.4 and 1.0 mins from which to choose. The optimum
extraneous peaks and a flat baseline throughout the
time is a function of liner size and type, head pressure
run.
at time of injection, solvent used, and injector
temperature. If you’re using a setting between 0.4 and

6 Optimize January 2017 SPEXCertiPrep.com/ECS

Here are some things to look for. ●● column not inserted properly into inlet (it
should be about 5 mm above the ferrule)
1. No ski slope early in the run. If you see
a high baseline to start, then about 2 or 3 ●● incorrect head pressure at time of
mins into the run the baseline drops back injection (it should be between 10-25 psi)
to normal, you have either an air leak or an ●● dirty inlet disk
injector problem. If you see ions 40 and 44 in
the “ski” slope, you have a leaky fitting inside ●● contaminated injector pneumatics
the GC, either at the inlet or detector. If you 4. No stray peaks. You should see 6 sharp,
see ions 49,84 and 86, then you either have crisp peaks and a flat baseline throughout.
a clog in the column, a defective column or Extraneous peaks indicate either
problems with the pneumatics. These ions contamination in the inlet or contamination
would be indicative of Methylene Chloride. in the solvent or Internal Standard Mix.
Presence of Methylene Chloride early in the
run (i.e. eluting during or after your first At this point I want to discuss the flows and
target compound) means that the solvent is
not being swept out of the system properly. pressures of the split/splitless injector for
Volatiles analyses.
So if you see them during the run (which you
Although the flow path with the 6890 is slightly
shouldn’t), the first thing to do is clip off a
different than the 5890, the points I want to make
loop of column and re-install. If the problem
apply to both.
persists, replace the column. If it still persists,
then cleaning the split/splitless solenoid It’s important to understand the flow path of the split/
with Methanol would be indicated. If it still splitless injection port in order to optimize conditions
persists after all this then an injector rebuild and troubleshoot problems.
would be next. The main distinction between using the split/
2. No rise in the baseline at the end of the run. splitless injection port with Volatiles as opposed to
If you see a noticeable rise in the baseline, Semivolatiles is that it is used only in the Split (Purge
chances are the ions would be 207 and 281. “On”) mode. So, even though the term “Volatiles by
This is column bleed. Run 3 more of these split/splitless injection” is used, what’s really meant is
test runs and see if the baseline is improving. “Volatiles by split injection”. This simplifies things for
If so, continue running these test runs until us.
the baseline is acceptable. If the bleed The flow path for volatiles by split/splitless injection
doesn’t dissipate, you can bake the column at depends on how it was plumbed. Regardless of that
300 °C for 4 hours, but the problem with that aspect, the flow feeding the injection port comes from
is that any baked off bleed will contaminate the transfer line of the purge and trap. The helium
your Mass Spec. That’s why thin-film, low and analytes are transferred from the trap, through
bleed columns (0.25 u film thickness) are the 6-port valve and into the inlet during the Desorb
desirable. A brand new column that has been mode. Typically that flow is set to about 25 mL/min.
conditioned overnight with these blank runs Once the flow enters the injector, it is basically split
should have a good baseline the subsequent 24:1, with 24 mL/min sweeping through the inlet,
day. If not, I suggest replacing the column. through the solenoid valve and out the split vent and 1
mL/min entering the column to serve as carrier flow.
3. The height of ISTD #6 (Perylene-d12)
should be at least half of that of ISTD#1 For 5890 GCs, the septum vent is not needed and is
(1,4-Dichlorobenzene-d4). If you don’t get either shut off or has the vent capped off. For 6890
the proper peak formations and/or your 6th GCs, you need to leave it in the flow path at the usual
Internal Standard is low, check the following: 2-3 mL/min. The head pressure used depends on
●● correct injector temp (should be 250 °C) whether EPC is used or not and the internal diameter
of the column. Assuming you are using a 0.25 mm ID
●● be sure you have enough insulation in the column, you can use the following as a guideline:
lower insulation cover
●● clog in the column

January 2017 Optimize 7

Systems without EPC: set the head pressure to 15 Periodic injector maintenance similar to semivolatiles
psi (clipping a loop, changing the liner and disk) should
Systems with EPC: set the flow to a constant be done only as needed. If no foamovers occur it’s
flow of 1.0 mL/min or possible to go for many months without having to
pressure pulse it so that the service the inlet.
head pressure is high during One common problem we’ve seen with the Volatiles
desorb (25 psi or so) then by split/splitless injection is where the Internal
remains constant at about 1 standard areas increase as the standard level increases.
mL/minute thereafter. Here’s an example of what might happen. The response
It’s a simple flow path- the flow comes in via the of your internal standard fluorobenzene (which is 5.0
Purge and Trap and gets split about 24:1. You might ppb for all points) for your 5 point curve is as follows:
think that this system would kill sensitivity. Indeed 1.0 ppb= 1,656,345 counts
it does hurt sensitivity, which is why most labs tune
their Mass Specs with significantly higher voltages 2.0 ppb= 1,878,620 counts
than Semivolatiles systems. Fortunately, advances 5.0 ppb= 2,014,890 counts
in column and electron multiplier technologies over 10.0 ppb= 2,414,450 counts
the past few years enable labs to meet low level MDLs
20.0 ppb= 2,868,978 counts
even with this seemingly high split ratio.
This inconsistency is causing your curve to be non-
Troubleshooting the Split/Splitless Injector linear and your surrogates to fail. What is going on
for Volatiles here and how would you go about resolving this
Here are recommended inlet conditions for running problem?
Method 624/8240/524.2 Answer: This is a good one. We see it happen often
Column: 60 m/0.25 mm ID/0.25 or 1.4 u so we have a handle on what to do. What’s happening
Film thickness. is that the internal standard area is increasing with
Liner size: 1.0., 2.0 or even 4.0 mm ID the concentration of your analytes. This should
straight (see which works best) NOT happen. The internal standard area should be
reasonably consistent in every run, be it a blank, low
Injector temperature: 
150 °C (be sure to use the level standard, high level standard, or heavily loaded
insulation). It’s counter- sample. So, what causes this?
productive to set the injector to
very hot temperatures What’s happening is that the presence of other
analytes is giving the “analyte slug” coming off the
Head pressure: 15 psi for Non-EPC systems, trap during desorb more mass and hence more inertia
for EPC systems I like to as it travels to the injector. The problem is either in
pressure pulse or program it; the injection port and/or the purge and trap- possibly
25 psi for desorb time-variable a leak, contamination or non-optimized conditions
pressure to achieve 1.0 mL/min somewhere. The first thing to do is to leak check the
for balance of run (1.2 mL/min system. The next thing to do would be to Methanol
for 5971-5). flush the concentrator.
Assuming you have proper inlet conditions, the 2 If no leak or active site exists, then try changing
biggest problems you will face are leaks and active conditions. If fluorobenzene in the 20 ppb standard
sites. As mentioned earlier, leaks can often be detected shows more area than it does in the 1.0 ppb
by changes in split vent flow, so once per month, when standard, then some of the fluorobenzene may
the system is running optimally, measure the split be getting lost out the split vent. Try the following
vent flow and record it in a log book. to improve the consistency of the responses:
Active sites in the inlet can occur from buildup of
acid or from “foamovers” whereby a soapy sample
overflows the purge tube and some of the soapy
residue actually makes it all the way to the inlet.
8 Optimize January 2017 SPEXCertiPrep.com/ECS
 A. Doing the injector maintenance (new liner, “slice and dice” plumbing whereby the purge
gold seal, septum). and trap transfer line is connected to the
B. Adjust the desorb flow to 25-30 mL/min for inlet line on the injector. You’ll need to butt
5890s, 30-35 mL/min for 6890s. Increasing connect a 6” piece of 0.32 mm guard column
the desorb flow usually mitigates this tubing to the purge and trap transfer line
problem but hurts sensitivity so raise it as (with a 1/16” to 0.32 mm union) and insert
much as you can provided you can still see the 0.32 mm guard column through the
your low level standard. septum and into the injector until it is about
1/4” from the analytical column. This reduces
C. 
Increase the Injector temperature by contact of the volatile organics with stainless
20 °C from what you have been using, steel and we have seen numerous times where
then decrease it by 20 °C from what you this improves performance.
have been using and see which direction
improves results. For 6890 GCs, the conventional “slice and dice”
plumbing works better.
D. Do the same as “C” for the heated zones in
your purge and trap. QQ Never use graphite ferrules on detector end.
Always use the longer high vespel content
E. If you have a fritted sparger switch to a graphite/vespel ferrules (85% Vespel, 15%
fritless one. graphite) on the fitting into the mass spec.
F. Clip off 2 loops of the column (replace if you QQ Use the Agilent “bullet” ferrules with shallow
suspect a sample foamover). nuts on injector ports.
If all else fails, you may need an injector rebuild. QQ Use UHP Helium and Indicating Moisture/
Some systems always have lower surrogate recoveries Oxygen and Hydrocarbon traps. It’s usually a
good idea to remove the molecular sieve traps
in samples and blanks than what they obtain when in the GC because they are not indicating
running a curve. Some labs have overcome this by so you have no idea of their condition.
skewing down the surrogate recoveries QC limits Instead, filter the carrier supply externally
(e.g. from 80-120 to 70-110; same window but skewed with indicating Moisture/Oxygen and
about 10% lower). Hydrocarbon traps.
Final thoughts when using a 5890 or 6890 GC Questions or comments on this or any issue of
OPTIMIZE may be emailed to the author, Mark
QQ Use stainless steel nuts on GC to MSD Ferry, at [email protected].
interface. The brass ones can strip, causing
leaks and other bigger problems. Agilent likes
to sell the brass ones with the logic that it’s
better to have the $10 brass nut strip than it
is to have the interface strip. Our logic is that
it’s better to have neither strip.
QQ The splitless flow path on the 6890 is
different than the 5890. As a result, it is often
necessary to start runs at 50 °C or lower with
the 6890. If you have a 6890 and start a run
at >100 °C and are unhappy with the results,
change the temperature programming to start
at 50 °C for 1 minute, then ramp 20 °C to the
temperature you would have started at. This
often results in better chromatography.
QQ For volatiles analysis with a split/splitless
injector and a 5890 GC, consider using a
0.32 mm guard column inserted into the
injection port rather than the conventional

January 2017 Optimize 9

US Address: UK Address:
SPEX CertiPrep, Inc. SPEX Europe
203 Norcross Avenue 2 Dalston Gardens
Metuchen, NJ 08840 Stanmore
Tel: 1-800-LAB-SPEX Middlesex, HA7 1BQ UK SPEX CertiPrep
Fax: 732-603-9647 Tel: +44 (0) 208 204 6656 Go supports saving our
E-mail: [email protected] Fax: +44 (0) 208 204 6654 Green! natural resources.
Web: www.spexcertiprep.com E-mail: [email protected]
Please
Web: www.spexeurope.com Recycle!