4UHPLC Injection Valves - A benchmark for quality, performance and lifetime
Table of contents
Table of contents
1. Overview
2. Injection valves in autosamplers
A. Construction & operation
B. Pressure range
C. Materials & coatings
D. Injection valve performance
3. Benchmark of injection valves
A. Methodology
B. Experimental conditions
4. Performance (results)
A. Initial & long-term performance
• repeatability/reproducibility
• carryover
B. Wear
• Spark
• Vendor A
• Vendor B
5. Comparison & conclusion
4UHPLC Injection Valves - A benchmark for quality, performance and lifetime
1. Overview
1. Overview
The injection valve is one of the most important parts of the
autosampler. The choice of an injection valve for an autosampler
is primarily determined by performance and lifetime. Companies
that supply autosamplers to laboratories usually equip them with
a self-produced injection valve or a valve from a reputable brand.
Several UHPLC valve manufacturers are considered as high-end
suppliers. Spark is an OEM provider of UHPLC valves and therefore
often compared with the well-known vendors.
Is there an alternative for UHPLC valves of high-end suppliers that
performs at least as well and lasts as long? Until now no systematic
research has been conducted to satisfactorily answer that question.
To answer this question, a benchmark in collaboration with NHL Stenden
University has been conducted comparing high-end injection valves
from different brands. This study is conducted to demonstrate the
quality, performance and lifetime of Spark UHPLC valves compared to
other well known, high-end UHPLC valve manufacturers. On the basis
of the benchmark the conclusion is justified that the Spark UHPLC
inspection valve is an equally good, if not better, alternative for other
available high-end valves.
This white paper provides insight into the approach and results of the
benchmark. It helps to make a more considered choice of injection
valves for your UHPLC autosampler in order to optimally advise and
support laboratories.
4UHPLC Injection Valves - A benchmark for quality, performance and lifetime
2. Injection valves in autosamplers
2. Injection valves in autosamplers
The injection valve is the heart of the autosampler in an UHPLC
system. In this publication, the valves with 6 ports and 2 positions
will be discussed, because these valves are the most common to
inject a sample.
A. Construction & operation
An injection valve consists of a valve body with a stator and a rotor seal,
see Figure 1. To get a leak-free connection between the rotor and stator,
the rotor is pushed against the stator by the spring in the valve body.
The front of the valve, where the tubing can be connected, is the
stator. The stator consists of 6 ports, 2 ports in the stator are
connected through the grooves in the rotor seal.
The rotor has 3 arched flow passages (grooves) to connect the stator
ports. The rotor seal rotates in the valve when the valve switches from
position (INJECT / LOAD). The default position is the INJECT position.
To introduce the sample into the flow path, the valve switches from
INJECT to LOAD, see Figure 2. In the LOAD position, the syringe aspirates
the sample into the loop. After the sample is loaded into the loop, the
valve switches back to INJECT. The sample in the loop will be transferred
by the eluent into the column.
Figure 1, Injection valve schematic
Figure 2, Positions and flow direction of an
injection valve
4UHPLC Injection Valves - A benchmark for quality, performance and lifetime
2. Injection valves in autosamplers
2. Injection valves in autosamplers
B. Pressure range
For more than four decades, reducing stationary phase particle size
has been investigated to improve chromatographic separation
efficiency. Until the early 2000’s, Liquid Chromatography (LC) technology
had reached a plateau, in which the benefits of reduced particle size
could not be realized due to limited pressure range of the HPLC system.
In 2004, separation science was revolutionized when Waters introduced
the first Ultra-Performance Liquid Chromatography system. Advances
in instrumentation and column technology were made to achieve
dramatic increases in resolution, speed and sensitivity. Not long after,
vendors (e.g. Spark Holland) entered the LC market with similar systems
which could handle pressures up to 18.000 psi.
As the injection valve is part of the LC flow path, this led to new designs
of the injection valves. Not only the maximum pressure range was
increased, but also internal volumes were decreased. The internal
volumes are kept as small as possible, so this will not affect the
increased separation performance due to the smaller column particles.
C. Materials & coatings
The rotor seal determines the performance of the valve (carryover
takes place within the rotor seal and can occur due to a damaged
stator). The rotor seal rotates continuously and has to be replaced
after a certain number of rotations (part of yearly maintenance).
Rotor seal wear can cause the performance to deteriorate.
Rotor seals predominantly consist of softer materials than stators
to get a leak-free connection, like Polyimide, PTFE or PEEK.
Stators usually consist of materials such as stainless steel (SS), PEEK
or titanium. To receive a robust and tight connection the stator is
often coated. The coating also increases the lifetime of the valve.
Diamond-like carbon (DLC) coatings have emerged as the ideal
solution for demanding tribological applications where components
are under high loads or subject to extreme friction, wear and contact
with other parts. In these types of environments, only the high hardness
of a DLC coating – along with a corresponding low coefficient of
friction – can prevent parts from pitting, galling, seizing and ultimately
failing in the field. DLC coating reduces friction and increases the
llfetime of a stator.
4UHPLC Injection Valves - A benchmark for quality, performance and lifetime
2. Injection valves in autosamplers
2. Injection valves in autosamplers
D. Pressure range
In UHPLC chromatography it is important to maintain a constant
pressure to increase the lifetime of the column, and to avoid
sample dilution at the moment the valve is switched. To optimize
the performance, valve manufactures have optimized the design
of the injection valve the recent years. Currently, some improvements
that may be found in an injection valve are Make-Before-Break (MBB)
and Pre-pressurize loop. Another inno vation, which is particularly
noteworthy, is Intermediate Loop Decompression, patented by
Spark Holland (ILDâ„¢).
At the start of analysis, prior to sample aspiration into the sample loop, the injection
valve is in the inject position. In this position, the sample loop is under ultra-high
pressure as mobile phase is flowing from the LC pump through the sample loop to
the LC column. For full loop and partial loop fill injection modes, when the injection
cycle starts, the syringe fills the needle and some of the buffer tubing with sample
to prepare to load the sample loop. The injection valve switches to the load position
causing the compressed liquid in the sample loop to expand at the outlet, into the
sample needle. This results in an uncontrolled dilution of the sample at the inlet of
the sample loop, and although the syringe draws a correct sample volume into the
loop, part of the sample has been diluted owing to the change from high pressure
to ambient pressure as the valve switches. Introducing a sample volume into the
UHPLC system that has been partially diluted results in peaks that represent an
inaccurate analyte concentration.
The ILD injection valve, patented by Spark Holland, is a 7-port valve with a
strate gically placed radial groove in the rotor seal (Figure 3). This radial groove is
connec ted to an additional waste outlet port in the center of the valve. This allows
for decompression of the sample loop in milliseconds when the valve switches
position from ultrahigh pressure (inject) to ambient pressure (load). When the
injection valve switches into the load position, the radial groove passes by one of
the sample loop ports (position 2 on the valve). The liquid in the sample loop is then
allowed to expand via the central outlet port to waste, releasing pressure. Therefore,
the sample loop will be at ambient pressure as the valve continues to switch and
arrives at the load position.
Thus, the ILD injection valve ensures that an accurate sample volume will be
injected onto the UHPLC system every time. The result is excellent reproducibility
and longer column lifetime.
Figure 3, ILD injection valve rotor seal, arrow
pointing to the additional radial groove in the
center of the rotor seal
4UHPLC Injection Valves - A benchmark for quality, performance and lifetime
3. Benchmark of injection valves
3. Benchmark of injection valves
This benchmark is a long term analytical performance comparison of
the Spark injection valve with two other high-end UHPLC valves (vendor
A, and vendor B). All three valves are well known OEM injection valves in
the UHPLC market, and have a good reputation. All three valves have a
diamond-like carbon (DLC) coating or some variant thereof but use a
different rotor seal material.
A. Methodology
All three valves have been installed in a common UHPLC system as
the only part to be exchanged for these experiments. Each valve is
used up to 30,000 injections under UHPLC conditions at 75 – 85% of the
maximum pressure of the valve. After each 5,000 injections, the valve is
inspected for analytical performance and after each 30,000 injections
a visual inspection of wear was carried out. Every injection valve has
been tested using three rotor seals to test the lifetime of the stator.
This makes a total of 90,000 switching cycles per injection valve.
The performance of the injection valves is determined by measuring
injection repeatability of the peak area and carryover of caffeine. In
doing so a maximum reproducibility (< 0.3) and minimal carryover
(≤ 0.005%) was expected.
1. Injection repeatability
The repeatability (or reproducibility) is the precision of an instrument
or method under the same circumstances. When the injections of a
valve are not repeatable, for example caused by intensive wear, the
results are not precise. To determine the repeatability of the injection
valves one sample was injected 10 times from the same vial. The
repeatability is expressed in the relative standard deviation, RSD.
2. Carryover
Carryover is a type of sample contamination in the UHPLC system.
When injecting a sample a small part may remain behind. When
injecting the next sample, the sample will take this small remaining
to the column. Resulting in a sample peak of a component that is
not present in the sample. Besides wear can aggravate carryover.
The carryover is expressed in a percentage of the concentration
of the previous injection. In this benchmark carryover of caffeine
was investigated.
4UHPLC Injection Valves - A benchmark for quality, performance and lifetime
3. Benchmark of injection valves
3. Benchmark of injection valves
B. Experimental conditions
The repeatability and carryover were determined using a stepwise
protocol, using the Spark UHPLC system:
1. Switch on the pump and program a flow rate of 0.5 ml/min.
2. Auto zero the UV-detector.
3. Start the sequence as indicated in Table 1.
4. Calculate the RSD of the peak areas of injection 5 to 14.
5. Calculate the carryover of injections 16, 17, and 18.
6. Check if the RSD and carryover meet the following targets:
a. RSD ≤ 0.3%.
b. Carryover ≤ 0.005%.
A schematic overview of the system is shown in Figure 4.
In Table 1 the experimental setup is listed.
The sequence for measuring the repeatability and carryover
is shown in Table 2.
Figure 4, Schematic drawing of the
analytical system
Table 1, Experimental setup (solvents & methods)
Wash solvent autosampler:
Eluent:
Flow
Sample - blank:
- low:
- medium:
- high:
Injection:
Preflush volume:
Air segment:
Analysis time:
Wash volume inside:
Wash volume outside:
20/80 methanol/water
20/80 methanol/water
0.5 mL/min
Water
0.5 ppm caffeine in water
50 ppm caffeine in water
10,000 ppm caffeine in water
Full loop fill
30.0 µL
Off
2 minutes (blank, low, medium), 6 minutes (high)
250 µL
30.0 µL
Solvents
Method
Table 2, Sequence for repeatability and carryover
1
2
3
4
5
6
7
8
9
10
11
Blank
Blank
Low
Low
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Injection InjectionSample Sample
A. Initial & long-term
perform ance
To make a comparison of the initial
and long-term performance two
indicators have been applied:
repeatability and carryover.
Repeatability/reproducibility
As a target, a repeatability of ≤ 0.30%
was set. This target is based on the
specifications of the integrity auto
samplers with the Spark injection
valves. Figure 5 shows that the
injection valves in general meet
the target and perform equally.
The repeatability of the injection
valves is mainly around 0.1%.
The repeatability of the injection
valves does not decrease for 30,000
injections. This is for the new injection
valve, the second rotor seal and the
third rotor seal almost identical.
This indicates that the repeatability
of the injection valves does not
decrease due to wear on the stator
within 90,000 injections.
Carryover
As a target, a carryover of ≤ 0.005%
is set. This target is based on the
specifications of the integrity
autosampler with the Spark injection
valves. Figure 6 shows that the
injection valves in general meet the
target. The carryover of the injection
valves does not increase for 30,000
injections. This is for the new
injection valves, the second rotor
seal, and the third rotor seal almost
identical. This indicates that the
carryover of the injection valves
does not increase due to wear on
the stator within 90,000 injections.
Figure 5, All injection valves meet the repeatability target of ≤ 0.30%
Figure 6, All injection valves meet the carryover target of ≤ 0.005%
7UHPLC Injection Valves - A benchmark for quality, performance and lifetime
4. Performance (results)
4. Performance (results)
Spark
Vend or A
Vend or B
Tar g et
Spark
Vend or A
Vend or B
Tar g et
Spark
Vend or A
Vend or B
Specication
Spark
Vend or A
Vend or B
Specication
B. Wear
Wear of the rotor seal or stator may
result in leaking and may impact
the analytical performance of the
injection valve. After every 30,000
injections, the rotor seal and stator
were examined using a microscope.
Spark
Rotor seal
The used rotor seals of the Spark
injection valve all show some disco-
lorations and superficial wear after
30,000 injections, see figure 7.
These light traces however have no
consequences for the analytical
performance of the injection valve.
Stator
After disconnecting the stator of
the rotor seal and the body of the
injection valve the stator seemed
affected by the injections. It turned
out to be rotor seal material on the
stator. After cleaning the stator with
tissue and acetone there were no
traces left of the rotor seal material.
From microscope pictures and the
analytical performance may be
concluded that the Spark stator can
be used for 90,000 injections without
any problems, under comparable
circumstances.
Figure 7, The Spark rotor seal shows some superficial wear after 30,000 injections. The stator shows no wear after 90,000 injections.
There are no consequences for the analytical performance
7UHPLC Injection Valves - A benchmark for quality, performance and lifetime
4. Performance (results)
4. Performance (results)
Vendor A
Rotor seal
The used rotor seals of the Vendor A
injection valve show some disco lo-
rations and deformations. The rotor
seal for 0 to 30,000 injections shows
less wear than the rotor seal for
60,000 to 90,000 injections. This
might conclude that the state of the
stator and/or valve body decreased,
causing the rotor seal to wear faster.
The used Vendor A rotor seals have
a circle of wear around the grooves,
while the new rotor seals were clean.
Another observation is the wear
between groove 2 and groove 3.
The Vendor A injection valve met
the target for repeatability and
carryover up to 90,000 injections.
This concludes that the wear
between groove 2 and groove 3
for the Vendor A rotor seals is
superficial and does not impact
the analytical performance of the
injection valve.
Stator
On microscope pictures no wear is
visible. There are some scratches
visible in the middle of the stator,
but these scratches do not seem
to have consequences since the
Vendor A injection valve met the
target for the analytical perfor-
mance. From the microscope
pictures and the analytical
performance of the Vendor A may
be concluded that the Vendor A
stator can be used for 90,000
injections without any problems,
under comparable circumstances.
Figure 8, The Vendor A rotor seal and stator show some traces, but they have no consequences for the analytical performance
7UHPLC Injection Valves - A benchmark for quality, performance and lifetime
4. Performance (results)
4. Performance (results)
Vendor B
Rotor seal
The used rotor seals of the Vendor B
injection valve show some disco-
lorations. The discolorations of the
rotor seals are consistent. The Vendor
B valve show wear between groove 1
and groove 2. Since the eluent did
not leak out of the sample needle
and the injection valve did meet the
targets for the analytical perfor-
mance for the Vendor B up to 30,000
injections, it can be conclu ded that
this wear is superficial and does not
impact the analytical performance
of the injection valve. Also some
wear at the bottom of groove 2 was
observed. It may be concluded that
this wear is super ficial and does not
impact the analytical performance
of the injection valve.
Stator
The Vendor B hardly had some rotor
seal material on the stator. The
Vendor B stator was cleaned with
a tissue and isopropyl alcohol.
Microscope pictures show some
wear on the stator of the Vendor B
injection valve. This wear seems to
increase after more injections are
made. The wear is visible as bright
spots in the microscope pictures.
These bright spots are most likely
created because the coating has
crumbled off the stator. After 30,000
injections the first spots are shown.
But the spots stand out after 50,000
injections. Except from these bright
spots the stator does not show any
other signs of wear.
Figure 9, The Vendor B rotor seal shows some light wear on the low pressure side and the stator shows a damaged coating, but they have
no consequences for the analytical performance
7UHPLC Injection Valves - A benchmark for quality, performance and lifetime
4. Performance (results)
4. Performance (results)
4UHPLC Injection Valves - A benchmark for quality, performance and lifetime
5. Comparison & conclusion
5. Comparison & conclusion
A systematic and thorough research has been conducted
comparing high-end injection valves from different brands
(Spark, Vendor A, and Vendor B). The benchmark was conducted
transparently and the results are reliable and verifiable.
Repeatability
The Spark, Vendor A, and Vendor B
injection valves meet the target for
the repeatability. There is no trend
visible for a decreasing repeatability
at an increasing amount of
injections. The repeatability of the
Spark, Vendor A, and Vendor B are
comparable, under these
circumstances.
Carryover
The Spark, Vendor A, and Vendor B
injection valves in general meet
the target for the carryover. The
carryover remains stable from 0 to
90,000 injections. The Vendor B
injection valves seem to have less
carryover than the Vendor A, and
Spark injection valves. The carryover
of the Vendor A, and Spark injection
valves is between 0.001% and 0.005%,
while the carryover of the Vendor B
injection valves is below 0.001%. This
difference in carryover might be
caused by the lack of
the Make-Before-Break function of
the Vendor B injection valve.
Wear
The rotor seals for the Spark, Vendor
A, and Vendor B injection valves
all show signs of use. The Vendor B
rotor seals seems to have less wear
(and discolorisations) than the
Spark, and Vendor A rotor seals. The
Spark, Vendor A, and Vendor B rotor
seals show the most wear between
the grooves on the ambient
pressure side of the injection valve.
The stators of the Spark, and Vendor
A injection valves show hardly any
wear. The Vendor B injection valve
however does show wear in the
form of crumbled off coating close
to three ports. This damage has
resulted in a leaking injection valve.
Conclusion
On the basis of the benchmark, the
following general conclusions are
justified:
- Spark injection valves last their
performance without any
decrease up to 90,000 injections,
after replacing the rotor seal every
30,000 injections.
- Spark rotor seals show only some
discolorations and superficial
wear, but this does not impact the
analytical performance of the
injection valve.
- Spark stators do hardly show any
wear after 90,000 injections, and
do not impact the analytical
performance of the inspection
valve.
- Spark injection valves, compared
with the Vendor A and Vendor B
valves, do not show substantial
problems, like leakage causing
wear and subsquent decrease in
analytical performance.
The Spark UHPLC inspection valve
is an equally good, if not better,
alternative for other available
high-end valves. There is no
objective reason why companies
supplying autosamplers to
laboratories would limit them-
selves to autosamplers with an
injection valve from a reputable
brand.
