Technical Overview
Authors
Charles Nichols,
Behrooz Zekavat,
and Patrick Batoon
Agilent Technologies, Inc.
Abstract
This Technical Overview presents the measurement of the instrument detection
limit (IDL) at two different conditions of MRM dwell times: 1) with sufficient ion
sampling time, and 2) with minimum allowed ion sampling time. The workflow
used an Agilent 1290 Infinity LC and an Agilent 6495C triple quadrupole LC/MS.
Results show that even with short MRM dwell times, low IDLs were still confidently
achievable with this experimental setup.
Instrument Detection Limit at
Ultrashort Dwell Times Demonstrated
on the Agilent 6495C Triple
Quadrupole LC/MS
2
Introduction
To demonstrate suitability for robustness
and reliability in routine applications,
instead of “one-shot” metrics such
as signal-to-noise (S/N), instrument
sensitivity is primarily characterized
using the IDL.1 This specification is more
rigorous and is determined using the
ion statistics of replicate injections.2 IDL
is based on the method detection limit
(MDL), thoroughly defined by the US EPA
and other governing bodies.3 “The MDL
is the minimum measured concentration
of a substance that can be reported
with 99% confidence that the measured
concentration is distinguishable from
method blank results.” A key difference
between the IDL and MDL is that the IDL
is separate from the “analysis method”
of an application but still determines
the absolute lowest level of analyte
the instrument can detect without the
interference of sample matrix.
The IDL is the minimum level of analyte
that results in a statistically differentiable
signal from the instrument's overall
noise baseline. Instead of using a
single injection at high concentration
(such as is the case with a S/N
measurement), the IDL is determined
using replicate measurements at
low levels, typically with a %RSD of
10 to 20%. The RSD of response output
(chromatographic peak area) is used in
the IDL calculation.
A confidence limit is predefined,
typically at 99% confidence, for which a
t-statistic is applied as a multiplication
factor to determine the IDL on-column
amount. This confidence arises from
the statistical variation of a series of
injections around the instrument’s ability
to reliably produce replicable data.
The result of the IDL measurement is
a statistically meaningful instrument
performance specification, rather
than the extrapolation of a single
measurement at a high injection amount.
The IDL is calculated using the equation:
IDL = (t-statistic) × (amount on column) × (%RSD/100)
The calculated IDL states the minimum
amount injected on column that can be
unambiguously distinguished from the
baseline noise of the chromatograph
with 99% confidence, not attributable to
random spikes in noise, or the minimum
point at which the instrument can reliably
replicate data.
Effects of ion sampling
rate on signal variability:
Why would the IDL change
with various dwell times?
An important parameter in the
discussion of instrument sensitivity is
the “MRM dwell time”, which influences
the stochastic sampling of the ion beam.
Generally, the IDL is characterized by a
considerable amount of time dedicated
to sampling the ion beam, allowing a
stable and consistent flux of ions to
hit the detector. At lowered instrument
dwell times (typically <5 ms), ion beam
sampling becomes less precise due to
insufficient number of of ions involved in
the measurement.
Number of ions # ion flux × dwell time
Deficiencies in inadequate sampling
time (that is, reduced MRM dwell times)
manifest themselves as decreased
signal stability (increased %RSD), even
if the analyte is introduced at a constant
rate and high concentration. Figure 1
presents this observation, where the
%RSD of the ion signal increases at
decreasing dwell times.
This Application Note characterizes IDL
performance when the instrument is
operated at extremely short dwell times
(1 and 0.5 ms).
The experiments carried out for this
paper were acquired on and apply to the
6495C LC/TQ. Although the performance
metrics may vary between instrument
type and model, the overall concept may
be applied to any other LC/MS or GC/MS
mass spectrometer.
0%
5%
1%
2%
3%
4%
6%
7%
8%
9%
10%
0
1
2
3
4
5
6
7
8
9
10
RSD
Dwell time (ms)
Figure 1. ESI-L tune mix infusion acquired at various MRM dwell times. Signal was acquired in MRM mode
using the transition 622.0 & 622.0 with CE = 0 V.
3
Experimental
Instrumentation and reagents
•
Agilent 1290 Infinity binary pump
(G4220A)
•
Agilent 1290 Infinity autosampler
(G4226A)
•
Agilent 1200 Series autosampler
thermostat (G1330B)
•
Agilent ZORBAX Eclipse Plus
RRHD C18, 2.5 × 50 mm, 1.8 µm
(p/n 959757-902)
•
Agilent 6495C triple quadrupole
LC/MS (G6495C)
•
ESI/APCI positive ion performance
standard (G1946-85004)
MRM dwell times for IDL
measurements
Measurements using the primary
MRM transition of reserpine
(m/z 609.3 & 195.1) were acquired at
200, 1, and 0.5 ms dwell time. However,
to ensure that the same number of
chromatographic peak data points
were collected, a dummy transition of
m/z 610.3 & 196.1 was included so that
the overall duty cycle of the instrument
was 201 ms (shown in Table 2).
Table 1. LC and MS parameters.
LC Parameters
Flow Rate
0.4 mL/min
Solvent A
H
2O w/ 0.1% FA
Solvent B
ACN w/ 0.1% FA
Gradient
Time (minutes) %B
0.00 10
0.20 10
1.50 100
2.50 100
2.51 10
Stop Time
3 minutes
Post Time
0.30 minutes
MS Parameters
AJS parameters
Sheath Gas Temperature
400 ˚C
Sheath Gas Flow
11 L/min
Drying Gas Temperature
325 ˚C
Drying Gas Flow
11 L/min
Capillary Voltage
4,000 V
Nozzle Voltage
0 V
iFunnel parameters
High-Pressure RF
200 Vpp
Low-Pressure RF
110 Vpp
MRM Transition
609.3 & 195.1
Dummy Transition
610.3 & 196.1
Fragmentor
166 V
Collision Energy
42 V
Duty Cycle
201 ms
Dwell Time
200 , 1, or 0.5 ms
Method parameters
Table 2. Primary and dummy MRM dwell times
used equating to the same instrument duty cycle
for all experiments.
MRM Dwell
609.3 & 195.1
Dummy MRM Dwell
610.3 & 196.1
Total Cycle
Time
200.0 ms
1.0 ms
201.0 ms
1.0 ms
200.0 ms
201.0 ms
0.5 ms
200.5 ms
201.0 ms
4
Results and discussion
Demonstration of IDL using sufficient
ion sampling
To demonstrate IDL characterizations
using the ideal ion sampling scenario, an
acquisition using an MRM dwell time of
200 ms was used.
In this case, 1 fg reserpine was injected
on-column and repeated 10 times
in MRM mode of acquisition. The
%RSD of the chromatographic peak
area was determined to be 17.62%.
For n = 10 injections, the single-tailed
t-statistic at 99% confidence was
found to be 2.821. Putting these
parameters into the equation, the
IDL at 200 ms dwell time = 0.49 fg.
Given that the ion beam has been
sufficiently sampled for 200 ms,
uncertainty in the measurement arises
due to the physical absence of ions,
producing a chromatographic peak
in close proximity to baseline noise.
Inspecting the overlaid chromatograms
in Figure 2, it may appear that injecting
anything less than 0.49 fg on-column
would not provide enough ion current.
With injections less than the IDL, signal
produced will become statistically
confounded with the chromatographic
baseline, and it cannot be determined
with ≥99% confidence that the signal is
distinct from the noise.
Demonstration of IDL at extremely
short dwell times
Referring to Figures 3A and 4A, sufficient
ion current is produced, demonstrating
that the instrument can detect these
on-column amounts with adequate
“visual” S/N. However, due to insufficient
ion sampling time, uncertainty in the
measurement arises from the stochastic
variation of number of ions hitting the
detector, thus affecting chromatographic
peak area variability.
In line with guidance for characterizing
IDL, the injected on-column amount of
reserpine was targeted to produce a
%RSD between 10 and 20%. For 1 ms
dwell time, 100 fg reserpine was used,
resulting in IDL
1 ms = 52.4 fg, while for
0.5 ms dwell time, 250 fg reserpine was
used, resulting in IDL
0.5 ms = 119.5 fg.
Both cases produced a chromatographic
peak far above a reasonable S/N
threshold, however, the variation
in chromatographic peak area fell
between ~40 to 50%, which is deemed
unacceptable for confident quantitation
(Figures 3B and 4B).
When examining additional
measurements at short dwell times
far below the characterized IDL
(10 fg at 1 ms; 50 fg at 0.5 ms),
some injection replicates produced
sufficient chromatographic S/N
(Figures 3C and 4C). However, when
replicated over a series of injections,
the chromatographic peak variability
was obviously unacceptable at
%RSD≈146.07% and %RSD≈99.35%,
rendering the data unsuitable for reliable
quantitative analysis.
Figure 2. Replicate (n = 10) injections of 1 fg reserpine on-column using a 200 ms MRM dwell time.
Injection No.
Area
1
22
2
20
3
28
4
21
5
25
6
24
7
20
8
14
9
23
10
19
Average
21.6
St. dev.
3.8
%RSD
17.62%
IDL = (2.821) × (1 fg) × (0.1762)
IDL = 0.49 fg on-column
×101
4.8
4.9
5
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
5.9
6
6.1
6.2
6.3
Acquisition time (min)
2.2
2.3
2.4
2.5
2.6
Counts
5
Figure 3. Replicate (n = 10) injections of various amounts of reserpine on-column using 1 ms MRM dwell
times.
Acquisition time (minutes)
×103
1.3
1.4
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.5
Counts
Acquisition time (minutes)
×103
1.3
1.4
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.5
Counts
Acquisition time (minutes)
1.3
1.4
×103
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.5
Counts
100 fg reserpine
50 fg reserpine
10 fg reserpine
A
B
C
Table 3. Replicate (n = 10) injections of 100, 50, and
10 fg reserpine on-column using 1 ms MRM dwell
times.
Injection No. Area (100 fg) Area (50 fg) Area (10 fg)
1
1,791
935
13
2
1,650
448
169
3
1,211
516
202
4
1,902
609
18
5
1,387
189
1
6
1,727
626
11
7
1,159
917
111
8
1,657
836
6
9
2,130
531
0
10
1,747
430
0
Average
1,636.1
603.7
53.1
St. dev.
303.7
236.1
77.6
%RSD
18.56%
39.11%
146.07%
IDL = (2.821) × (100 fg) × (18.56%/100) = 52.4 fg on-column
1 ms MRM dwell time
Figure 4. Replicate (n = 10) injections of various amounts of reserpine on-column using 0.5 ms MRM dwell
times.
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
1.3 1.35 1.4 1.45 1.5
Acquisition time (minutes)
×103
Counts
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
1.3 1.35 1.4 1.45 1.5
Acquisition time (minutes)
×103
Counts
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
1.3 1.35 1.4 1.45 1.5
Acquisition time (minutes)
×103
Counts
250 fg reserpine
100 fg reserpine
50 fg reserpine
A
B
C
Table 4. Replicate (n = 10) injections of 250, 100,
and 50 fg reserpine on-column using 0.5 ms MRM
dwell times.
Injection No. Area (250 fg) Area (100 fg) Area (50 fg)
1
4,868
1,722
511
2
3,118
1,548
8
3
2,825
665
227
4
4,476
477
139
5
4,041
1,292
13
6
4,002
2,072
126
7
3,380
709
82
8
3,769
966
135
9
4,234
2,105
239
10
4,643
2,251
31
Average
3,935.6
1,380.7
151.1
St. dev.
667.1
655.3
150.1
%RSD
16.95%
47.46%
99.35%
IDL = (2.821) × (250 fg) × (16.95%/100) = 119.5 fg on-column
0.5 ms dwell time
www.agilent.com/chem
For Research Use Only. Not for use in diagnostic procedures.
This information is subject to change without notice.
© Agilent Technologies, Inc. 2020
Printed in the USA, January 29, 2020
5994-1368EN
Conclusion
This Technical Overview summarizes
the characterization of the IDL at two
different MRM dwell time regimes:
(1) at 200 ms providing sufficient ion
sampling time; or (2) at 1 and 0.5 ms
with inadequate sampling time of the ion
beam. Table 5 shows the results.
When operated at extremely short
dwell times such as 1 or 0.5 ms,
IDL characterization relies on the
instrument’s ability to reproducibly
sample and stabilize the ion beam within
the restricted time frame. Although
the MRM dwell times are reduced,
components along the ion optics rail
were designed to transmit as many ions
as possible. This results in reasonable
performance (low IDLs), even at the
shortest MRM dwell times.
IDL characterization was carried out to
demonstrate the changes in sensitivity
with extremely short dwell times
(demonstrated on the 6495C LC/TQ).
This was done to provide the customer
with an expectation of sensitivity
changes when running extremely
challenging methods for high-throughput
and routine applications.
Although using extremely short dwell
times is not a generally recommended
practice, Agilent recognizes that
customers are facing various challenges,
demanding scientific accuracy and
confident results while maximizing
sample throughput to keep up with the
cost of running the lab.
References
1. Sheehan, T.; Yost, R. What’s the
Most Meaningful Standard for Mass
Spectrometry: Instrument Detection
Limit or Signal-to-Noise Ratio?
Spectroscopy 2018, 13, 16–22.
2. Wells, G.; et al. Signal, Noise,
and Detection Limits in Mass
Spectrometry. Agilent Technologies
Application Note, publication number
5990-7651EN, 2011.
3. Definition and Procedure for the
Determination of the Method
Detection Limit, Revision 2.
Environmental Protection Agency.
(2016). doi:EPA 821-R-16-006.
Table 5. Replicate (n = 10) injections of 100, 50, and
10 fg reserpine on-column.
Dwell Time
Injection
Amount
%RSD
IDL
200 ms
1 fg
17.62%
0.49 fg
1 ms
100 fg
18.56%
52.4 fg
0.5 ms
250 fg
16.95%
119.5 fg
