Data Acquisition&Processing Theory Guide

Read the instruction manual thoroughly before you use the product.
Keep this instruction manual for future reference.

223-60090
July 2008
LabSolutions
Data Acquisition &
Processing Theory Guide
This page is intentionally left blank.
Data Acquisition & Processing Theory Guide i
Introduction
Read this Instruction Manual
thoroughly before using the
product.
Thank you for purchasing Shimadzu analytical instrument
workstation "LabSolutions" (hereafter referred to as "the
software" or "LabSolutions").
This manual describes the procedures for operating this
product. Read this manual thoroughly before using the
product and operate the product in accordance with the
instructions in this manual.
Also, keep this manual for future reference.
This manual assumes that the reader is knowledgeable of
basic operations of Windows. For the operation of Windows,
refer to the instruction manual that comes with that product.
Important
If the user or usage location changes, ensure that this
Instruction Manual is always kept together with the
product.
If this manual is lost or damaged, immediately contact your
Shimadzu representative to request a replacement.
To ensure safe operation, contact your Shimadzu
representative if product installation, adjustment, or re-
installation (after the product is moved) is required.
2008 Shimadzu Corporation All rights reserved.
ii Data Acquisition & Processing Theory Guide
Notice
Information in this manual is subject to change without
notice and does not represent a commitment on the part of
the vendor.
Any errors or omissions which may have occurred in this
manual despite the utmost care taken in its production will
be corrected as soon as possible, although not necessarily
immediately after detection.
All rights are reserved, including those to reproduce this
manual or parts thereof in any form without permission in
writing from Shimadzu Corporation.
Microsoft, Windows and Windows Vista are registered
trademarks of Microsoft Corporation in the United States
and/or other countries.
Adobe, Adobe logo and Adobe Reader are trademarks or
registered trademarks of Adobe Systems Incorporated in
the United States and/or other countries.
Other company names and product names mentioned in
this manual are trademarks or registered trademarks of
their respective companies. The TM and symbols are
omitted in this manual.
Microsoft
Windows

Vista
Operating System is referred

to as "Windows Vista".
Microsoft
Windows
XP Professional Edition is referred

to as "Windows XP".
Replacement parts for this product will be available for a
period of seven (7) years after the product is discontinued.
Thereafter, such parts may cease to be available. Note,
however, that the availability of parts not manufactured by
Shimadzu shall be determined by the relevant
manufacturers.
Data Acquisition & Processing Theory Guide iii
List of Instruction Manuals
Indications Used in Instruction Manuals
Cautions and Notes are indicated using the following conventions, and the following symbols are
used in this manual:
Instruction Manuals
Name Media Content
Getting Started
Guide
Booklet/
CD-ROM
This manual follows an actual data acquisition procedure to
describe basic methods of operation for first-time users.
Read this manual to learn basic operations of the software.
Operators Guide Booklet/
CD-ROM
This manual describes overall operations and handy func-
tions in more details, such as the software's system configu-
ration, data analysis, batch processing, confirmation of data
acquisition results, and report functions.
System Users
Guide
Booklet/
CD-ROM
This manual describes system administration and data man-
agement of the software. Refer to this manual as necessary.
Installation &
Maintenance
Guide
Booklet/
CD-ROM
This manual describes installation and maintenance of the
software.
DataAcquisition
& Processing
Theory Guide
Booklet/
CD-ROM
This manual describes peak detection and quantitation of
sample components.
Refer to this manual as necessary.
Help Online help Clicking the on-screen [Help] button or pressing the [F1] key
displays a description of on-screen parameters, answers to
specific questions or solutions to various problems. Also,
clicking the [Help] button on the error message window dis-
plays the details of the error or solutions to the error. Be sure
to refer to Help before contacting us.
Indication Meaning
! CAUTION Indicates a potentially hazardous situation which, if not
avoided, may result in minor to moderate injury or equipment
damage.
@ NOTE Emphasizes additional information that is provided to ensure
the proper use of this product.
^ Reference Indicates the location of related reference information.
[ ] Indicates the names of buttons, menu options, setting
options, windows/sub-windows, and icons that are displayed
in a window.
Example: Click [OK].
iv Data Acquisition & Processing Theory Guide
Warranty
Shimadzu provides the following warranty for this product.
1. Period:
Please contact your Shimadzu representative for information
about the period of this warranty.
2. Description:
If a product/part failure occurs for reasons attributable to
Shimadzu during the warranty period, Shimadzu will repair
or replace the product/part free of charge (including USB
dongles). However, in the case of products which are usually
available on the market only for a short time, such as
personal computers and their peripherals/parts, Shimadzu
may not be able to provide identical replacement products.
3. Exceptions:
Failures caused by the following are excluded from the
warranty, even if they occur during the warranty period.
1) Improper product handling
2) Repairs or modifications performed by parties other than
Shimadzu or Shimadzu designated companies
3) Product use in combination with hardware or software other
than that designated by Shimadzu
4) Computer viruses leading to device failures and damage to
data and software, including the product's basic software
5) Power failures, including power outages and sudden voltage
drops, leading to device failures and damage to data and
software, including the product's basic software
6) Turning OFF the product without following the proper shutdown
procedure leading to device failures and damage to data and
software, including the product's basic software
7) Reasons unrelated to the product itself
8) Product use in harsh environments, such as those subject to
high temperatures or humidity levels, corrosive gases, or strong
vibrations
9) Fires, earthquakes, or any other act of nature, contamination by
radioactive or hazardous substances, or any other force
majeure event, including wars, riots, and crimes
10)Product movement or transportation after installation
11) Consumable items
Note: Recording media such as floppy disks and CD-ROMs are
considered consumable items.
* If there is a document such as a warranty provided with the product, or there
is a separate contract agreed upon that includes warranty conditions, the
provisions of those documents shall apply.
* Warranty periods for products with special specifications and systems are
provided separately.
* The license cannot be reissued if you lose the USB dongle provided with the
product.
Data Acquisition & Processing Theory Guide v
Contents
1 Data Processing Parameters
1.1 Peak Integration Parameters...............................................................................2
1.1.1 Peak Integration Parameters and Operation Flow............................................. 3
1.1.2 [Width] (The Minimum Half-Height Width of Peaks)........................................... 5
1.1.3 [Slope] (Peak Detection Sensitivity Determining Peak Start/End)...................... 6
1.1.4 Slope Test .......................................................................................................... 7
1.1.5 [Drift] (Slope to Define Baseline) ...................................................................... 11
1.1.6 Integrating Unresolved Peaks .......................................................................... 13
1.1.7 [T.DBL] (Peak Detection Parameter Change Time) ......................................... 15
1.2 Identification Parameters...................................................................................17
1.2.1 Window Method and Band Method.................................................................. 17
1.2.2 Absolute/Relative Retention Time Methods ..................................................... 18
1.2.3 Identification of Adjacent Peaks ....................................................................... 21
1.2.4 Grouping .......................................................................................................... 23
1.3 Quantitative Methods.........................................................................................24
1.3.1 Quantitative Methods and Equations (with Dilution Factor) ............................. 25
1.3.2 Quantitative Methods and Equations (without Dilution Factor) ........................ 27
1.3.3 Rounding.......................................................................................................... 29
1.4 Concentration Deviations between Compound Table and Calibration Curve....30
1.4.1 Accuracy [%] .................................................................................................... 30
1.4.2 %Deviation....................................................................................................... 30
2 Calibration Curve
2.1 Calibration Curve Type......................................................................................31
2.1.1 Linear ............................................................................................................... 31
2.1.2 Point to Point .................................................................................................... 32
2.1.3 Quadratic and Cubic ........................................................................................ 32
2.1.4 Mean RF .......................................................................................................... 33
2.1.5 Exponential ...................................................................................................... 33
2.1.6 Manual RF (Linear, Exponential)...................................................................... 34
2.1.7 Quantitation Using Other Component Calibration Curve ................................. 34
2.1.8 Least Square Method and Weighted Least Square Method ............................ 36
2.2 Calibration Curve Correction .............................................................................39
2.2.1 Calibration Curves Using Standard Concentration Factors.............................. 39
2.2.2 Calibration Curves When Target Components Refer to Multiple Standards
Separately ........................................................................................................ 40
Contents
vi Data Acquisition & Processing Theory Guide
3 Equations
3.1 Noise/Drift Calculation Parameters ................................................................... 43
3.1.1 Noise Calculation Methods .............................................................................. 43
3.1.2 Detection/Quantitative Limit Coefficients ......................................................... 44
3.1.3 Drift Settings..................................................................................................... 44
3.2 QA/QC Parameters ........................................................................................... 45
3.2.1 Common Items................................................................................................. 45
3.2.2 Calibration........................................................................................................ 45
3.2.3 Quality Control ................................................................................................. 47
3.2.4 Recovery .......................................................................................................... 47
3.2.5 Degradation Check .......................................................................................... 47
3.2.6 Noise/Drift Check ............................................................................................. 47
3.3 Column Performance Equations........................................................................ 48
3.4 Background Correction Processing................................................................... 55
4 Peak Purity Algorithms
4.1 Basic Principles of Peak Purity Analysis ........................................................... 57
4.1.1 Calculating the Similarity.................................................................................. 57
4.1.2 Calculating the Threshold ................................................................................ 58
4.1.3 Calculating the Peak Purity .............................................................................. 59
4.1.4 Example of Peak Purity Analysis Result .......................................................... 62
4.2 Method Development for Peak Purity Analysis ................................................. 63
4.2.1 Specifying the Multi-Chromatogram Wavelength............................................. 63
4.2.2 Calculating the Noise Spectrum....................................................................... 64
4.2.3 Defining the Wavelength Range ...................................................................... 64
4.2.4 Specifying the Compensation Coefficient ........................................................ 65
4.2.5 Spectrum Background Compensation ............................................................. 65
4.2.6 Setting the [Compute Purity] Options............................................................... 66
4.3 Method Optimization.......................................................................................... 67
4.3.1 Method Development Procedure ..................................................................... 67
4.3.2 Validating the Method ...................................................................................... 69
4.3.3 Interpreting the Method Validation Result ........................................................ 69
4.3.4 Application Examples....................................................................................... 70
5 File Structure
5.1 Method File Structure ........................................................................................ 73
5.2 Data File Structure............................................................................................. 75
Data Acquisition & Processing Theory Guide 1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1 Data Processing
Parameters
This chapter describes data processing parameters and equations used for analyzing data in this software.
Data processing parameters are displayed in the [Method View] of the Data Analysis window.
This chapter consists of the following sections:
"1.1 Peak Integration Parameters" P.2
"1.2 Identification Parameters" P.17
"1.3 Quantitative Methods" P.24
"1.4 Concentration Deviations between Compound Table and Calibration Curve" P.30
2 Data Acquisition & Processing Theory Guide
1.1 Peak Integration Parameters
Peak integration processes such as target peak detection, baseline fluctuation compensation, unresolved
peak separation, and unnecessary peak rejection, are executed according to peak integration parameters
stored in method files.
This section describes the peak integration parameters and how peaks are integrated using these
parameters.
The following describes the three approaches for setting peak integration parameters:
^ Reference
For details on how to set the peak integration parameters, refer to Operator's Guide.
Name Description
Peak Integration
Parameters
Specifies the basic parameters for integrating peaks across the entire section of
chromatograms. These parameters can be set in [Method View] - [Integration] tab page.
Integration time
program
Used for changing peak integration parameters at specific times on chromatograms. To set
this, click the [Program] button on the [Peak Integration] tab page in the [Method view], to
open the [Integration Time Program] sub-window.
Manual Peak
Integration
Manually specifies peak integration parameters for each individual peak. The set parameters
are only applied to the set data file.
To set this, right click on the [Chromatogram View] in the Data Analysis window to display the
[Manual Integration Bar].
1
1.1.1 Peak Integration Parameters and Operation Flow
The user can execute peak integration by the following steps. The table below describes how specific peak
integration parameters relate to each integration process.
Peak Integration Flow Parameter
Function
< >: Default Value
[ ]: Setting Range
Description
Step 1: Detecting peaks
Peak start, peak end, and
peak top are first detected
on chromatograms
according to the [Width]
and [Slope] values.
Setting these parameters
properly detects only
necessary peaks and
excludes noise.
[Width] The minimum half-
height width of peaks
<3> (Unit: sec)
[0.04 to 200]
Specifies the minimum half-height width of the
peaks detected during analysis. Peaks with
smaller half-height width are regarded as noise.
[Slope] Peak detection
sensitivity
(which defines peak
start/end)
<1000>
(Unit: V/min)
[0 to 4 10
11
]
Specifies the slope used as a reference for
detecting peaks.
The point where a graph slope exceeds (falls
under) the set value is recognized as peak start
(end).
By executing the Slope test, the user can obtain
a suitable [Slope] value from baseline noise
signals.
^ Reference
"1.1.4 Slope Test" P.7
[T.DBL] Time to change peak
detection parameters
<1000>
(Unit: min)
[0 to 1000]
Increases the minimum half-height width
(=[Width]) by two times, and decreases peak
detection sensitivity (=[Slope]) to 1/2 at each set
value 2
n
(n=0 to 14). This parameter
facilitates the integration of chromatogram
peaks becoming broader with time.
If "0" is set, these parameters are adjusted
automatically.
^ Reference
"1.1.7 [T.DBL] (Peak Detection
Parameter Change Time)" P.15
Half-height
width
Step 2: Processing the
baseline (Determining how
to integrate overlapped
peaks)
To calculate (= integrate)
peak area, a baseline is
assumed between the first
peak start and the last
peak end of overlapped
peaks.
Set the [Drift] value as
appropriate when the
baseline fluctuation is
significant.
[Drift] Slope to define
baseline
<0> = Auto-process
(Unit: V/min)
[-10
7
to 10
7
]
Determines how to integrate overlapped peaks.
As shown in the left figure below, when a valley
of overlapped peaks exists above the slope
(= set value) starting from the first peak start,
the peaks are divided by a vertical line. When
the valley exists below the slope, the baseline is
drawn through the valley, as shown in the figure
at right.
If "0" is set, the software automatically
determines how to process the baseline.
A negative value can also be set.
^ Reference
"1.1.5 [Drift] (Slope to Define Baseline)"
P.11
Step 3: Integrating
unresolved peaks
In principle, unresolved
peaks are automatically
divided by a vertical line.
According to conditions,
peaks are recognized as
one peak appearing on the
other peak's tail.
None Auto-process -
Step 4: Calculating peak
area
Peak area is calculated
(integrated) for each
identified peak.
None Auto-process -
Function
< >: Default Value
[ ]: Setting Range
Description
Baseline
Vertical
division
Tailing
1
1.1.2 [Width] (The Minimum Half-Height Width of Peaks)
The [Width] is the most fundamental parameter among all peak integration parameters. The software
detects peaks according to the [Width] value, and performs peak integration under the conditions most
suitable for the peaks.
In [Width], specify the minimum half-height width (width at 50% of peak height) of peaks detected during
analysis. Specifically, set the value equal to or slightly smaller than the half-height width of the sharpest
peak in obtained chromatogram.
The unit of the [Width] is sec.
Fig.1-1 Half-Height Width
@ NOTE
Noise width is normally smaller than peak width. By setting the half-height width of the target peak to [Width],
peaks with smaller half-height width can be excluded as noise.
Example of [Width] setting
The following shows the comparison between the [Width] settings of 30 and 10 for the same
chromatogram.
[Width] = 30: Only one peak is detected.
Step 5: Removing
unnecessary, small peaks
Peaks with areas smaller
or heights lower than the
set value can be hidden
from the screen and peak
reports.
[Min Area/
Height]
The minimum peak
area/height to be
displayed
<1000>
(Unit: counts)
[0 to 10
7
]
Selecting "Area" or "Height" as a reference to
define the small peaks to be hidden (from the
screen, peak report, etc.) can be set in
[Calculated by] on the [Integration] tab page.
The unit of counts:
Area (Vsec), Height (V)
Function
< >: Default Value
[ ]: Setting Range
Description
Half-height
width
[Width] = 10: Two peaks are detected.
1.1.3 [Slope] (Peak Detection Sensitivity Determining Peak Start/End)
This software detects peaks (i.e., determines peak start/end) according to the graph slope. When is set to
[Slope] as shown in Fig.1-2, the peak start is detected when the graph slope exceeds the angle .
Conversely, the peak end is detected when the negative graph slope falls under the angle .
If a large value is set to [Slope], the software detects only sharp peaks with a large graph inclination. When
a smaller value is set, the software detects broader peaks, as well as a lot of noises in some case.
Fig.1-2 Peak Detection and [Slope]
Example of [Slope] setting
The following shows the comparison between the [Slope] settings of 1000 and 100000 for the same
chromatogram.
[Slope] = 1000: All noises are detected as minute peaks.
[Slope] = 100000: Only peaks with inclination lager than the [Slope] value are detected.
1
1.1.4 Slope Test
The Slope test is the feature to automatically calculate the [Slope] value from baseline noise and drift on
chromatograms.
There are two methods for obtaining the slope value: (1) using the baseline before analysis when analytical
equipment signals are displayed, and (2) using the baseline on chromatograms already obtained.
Slope test before analysis
1
Right-click on the chromatogram in the [Data Acquisition] window. Click [Slope Test] on
the displayed menu.
The Slope test begins. The test is executed for either 10 times the [Width] value or 10 sec., whichever is
longer.
2
The [Setting Slope] sub-window is displayed. After verifying the value in the [Slope] box,
click [Set to Parameter].
The Slope test result (= slope value displayed in the [Slope] box in the [Setting Slope] sub-window) is
saved to the peak integration parameter [Slope]. Close the [Setting Slope] sub-window. Click [Cancel] to
close the [Setting Slope] sub-window without saving.
@ NOTE
The Slope test often obtains a different slope value for each execution. Therefore, before saving the
result to the [Slope] parameter, it is recommended to slightly increase the value in the [Slope] box in
the [Setting Slope] window.
@ NOTE
The value obtained in the Slope test is based on the baseline before analysis, and does not reflect the
baseline drift which occurs during analysis. Therefore, the following must be observed:
DO NOT use the Slope test result for gradient LC and programmed temperature GC analyses where a
large baseline drift is expected. In such cases, manually set the [Slope] value larger than the drift value so
that the baseline will not be detected as an instance of peaks.
^ Reference
"Operator's Mnual "2.3.4 Calculating the Slope Value from Baseline (Slope Test)""
Slope test after analysis
1
In the [Postrun Analysis] window, open the data file in which chromatogram data is
already acquired.
2
Click (Edit Mode) in the [Method View].
The software goes into the edit mode.
3
Click the [Integration] tab, and click .
1
4
Set the [Integration Time Program] sub-window, and click [OK].
1 Click (Slope Test).
Moving the cursor onto the chromatogram displays a vertical line.
2 Move the cursor to the desired starting point for Slope test on the baseline, and click the mouse
button. Choose a section where no peak, no temporary noise, and no drift is observed.
A vertical line appears at the starting point.
3 Move the cursor to the desired test ending point, and click the mouse button.
A vertical line appears at the ending point. Specify the ending point so that the test period becomes
either 10 times the [Width] value or 10 sec., whichever is longer.
A suitable slope value is automatically calculated according to the noise within the specified baseline
section. The [Slope Test] sub-window is then displayed.
4 After verifying the displayed value, click [Set to Parameter].
The [Slope Test] sub-window is closed. The Slope test result will be set to the [Slope] parameter in
the [Method View] - [Integration] tab page. Click [Cancel] to close the [Slope Test] sub-window
without setting.
2
1
3
5
Click (View Mode) in the [Method View].
Once the [Slope] value is changed, the software re-executes peak integration according to the changed
value (postrun analysis).
1
1.1.5 [Drift] (Slope to Define Baseline)
Specify the slope to define baseline in the [Drift] parameter. If [Drift] is set to "0", the baseline is
automatically processed according to the predetermined rules. If a small value is set to [Drift], adjacent
peaks are regarded as unresolved peaks, and divided by a vertical line at the peak valley.
Automatic baseline correction ([Drift] = 0)
When [Drift] is set to "0", baseline is corrected automatically by the following rules: (1) As shown in Fig.1-3,
when the time width of the valley (T2: interval between the previous peak end and next peak start defined
by [Slope]) is smaller than the estimated half-height width of the previous peak (T1), the peaks are divided
by a vertical line as unresolved peaks. (2) When T2 is larger than T1, T2 is processed as baseline.
T1 > T2: Regarded as unresolved peaks and divided vertically.
T1 < T2: Regarded as separated peaks.
Fig.1-3 Automatic Baseline Correction ([Drift] = 0)
Baseline correction by specifying [Drift]
To correct baseline by user's definition, set a value other than "0" for [Drift]. By doing so, a valley even with
a small width (T2) can be processed as a baseline point.
Fig.1-4 Baseline Correction by Setting [Drift] Other than "0"
When [Drift] is set to a value other than 0, a slope defined by the [Drift] value (a dashed-dot line) is virtually
drawn from the peak start S, as shown in Fig.1-4. When the peak end E is positioned below the slope, the
baseline is drawn from S to E diagonally.
@ NOTE
When setting [Drift] to values other than "0", be sure to set a value larger than actual baseline drift observed
during analysis. If the value is smaller, the peak end will never fall below the assumed slope, resulting that all
peaks are recognized as unresolved peaks.
Example of [Drift] setting
The following shows the comparison between the [Drift] setting of 100 and 5000 for the same
chromatogram.
[Drift] = 100: Baseline is drawn with peaks vertically divided.
[Drift] = 5000: Baseline is drawn with peaks completely separated.
1
1.1.6 Integrating Unresolved Peaks
When peaks are detected according to [Width] and [Slope] and the baseline is corrected with [Drift],
unresolved peaks may appear.
In this case, the software automatically judges whether to divide the peaks vertically as overlapped peaks
(Vertical Division), or to process them as a small peak appearing on the other peak's tail (Tailing).
Vertical division
In principle all unresolved peaks are divided by a vertical line at the peak valley.
Fig.1-5 Vertical Division
Tailing
The software determines whether or not the accompanying peak is located on a tailing peak, using two
peak height ratio, width ratio, ratio of valley height and peak height, etc.
Fig.1-6 Conditions for Tailing
When all conditions below are satisfied, the accompanying peak is regarded as being on the tail.
When any of the following conditions is met, the accompanying peak is no longer regarded as being on the
tail.
Valley height H
3
is positioned above the Drift line (including when Drift = 0).
H
1
/ H
2
> 10 The accompanying peak is sufficiently smaller than the main peak.
W
1
/ W
2
> 3 The accompanying peak is sufficiently narrower than the main peak.
H
2
/ H
3
< 100 The accompanying peak start is positioned relatively higher than the peak height.
The valley height H
3
fell below the Drift line.
The software detected the end of unresolved peaks (when [Drift] is set to 0).
H
1
/ H
2
< 10 There is not much height difference between the main and accompanying peaks.
Drift
Example of unresolved peak integration
Fig.1-7 shows an example of automatic unresolved peak Integration.
Peak A is processed as a main tailing peak. Peaks B and C are processed as being located on the main
peak A's tail. Note that unresolved peaks on the tail such as Peak C are always divided vertically. Whether
or not one peak is on the other's tail is no longer examined.
Peaks E, F, and G are unresolved, but not located on the main peak's tail. Therefore, they are just divided
vertically.
Fig.1-7 Unresolved Peak Integration
@ NOTE
To indicate various integration statuses, the following peak marks can be displayed on the peak table and
chromatogram peak top on screen and reports:
S: Main tailing/leading peak
T: Peak on a tail
L: Peak on a head
V: The second or later peaks of unresolved peaks (divided vertically)
H: Peak whose baseline is corrected as horizontal
M: Peak integrated manually
E: Error peak (due to over flow or under flow)
Vertical Division
Tailing
1
1.1.7 [T.DBL] (Peak Detection Parameter Change Time)
When [T.DBL] is set, the software detects peaks by increasing [Width] by 2 times and decreasing [Slope] to
1/2, at each [T.DBL] value 2
n
(n = 0 to 14). If "0" is set, [Width] and [Slope] are automatically adjusted
according to predetermined ratio. To not change these parameters during analysis, specify a time longer
than the duration of the analysis. Normally, leave this parameter as default (= 1000).
@ NOTE
It is recommended to NOT change these parameters during analysis. Set the time longer than analysis
duration.
When NOT changing [Width] and [Slope] automatically
In gradient LC and programmed temperature GC analyses, peaks do not become broader with time. In
such case, set the [T.DBL] value longer than analysis duration to not change [Width] and [Slope]
automatically.
When automatically changing [Width] and [Slope] by setting [T.DBL]
When [T.DBL] is set to values other than "0", peaks are detected by automatically increasing [Width] by 2
times and decreasing [Slope] to 1/2 at the time specified in [T.DBL] to increase peak detection sensitivity
(see Fig.1-8).
This parameter change will be continuously executed at the time defined by [T.DBL] as shown in Fig.1-8.
Fig.1-8 Changing [Width] and [Slope] Automatically at the Time in [T.DBL]
@ NOTE
[Width] can be changed up to 15 times per analysis. This number includes cases of changing the value using
[Width =] or [T.DBL =] command in the integration time program.
For the [T.DBL] value, set the time when the peak width becomes two times that of the first peak width.
When no such peak (whose width is two times the first peak width) is observed on the chromatogram,
obtain [T.DBL] by the following calculation:
Example:
Fig.1-9 T.DBL Calculation
Sharpest peak eluted at the beginning of
chromatogram: half-height width 2 sec
Small peak at the end of chromatogram: half-
height width 30 sec, retention time 20 min
1
Set the half-height width of the sharpest peak eluted at the beginning of chromatogram
to the [Width] box (normal procedure).
Here, assume 2 sec.
2
Measure the retention time and half-height width of a small peak eluted at the end of
chromatogram.
Assume the retention time and half-height width as 20 min and 30 sec, respectively.
Since the peak width has increased by 15 times in 20 minutes,
The time required to double the peak width = 20 min 15 times 2 = Approx. 2.7 min
The general equation will be,
When adjusting [Width] and [Slope] automatically ([T.DBL] = 0)
If [T.DBL] is set to 0, peaks are detected by adjusting the [Slope] and [Width] values automatically as peaks
become broader. In isocratic LC and isothermal GC analyses, etc., the analysis peak width increases with
time. Therefore, high peak detection sensitivity (= low [Slope] value) is not necessary in the beginning,
since early-eluting peaks have a large slope. However, since the width of later peak becomes broader, the
software automatically decreases the [Slope] value, increasing the peak detection sensitivity. Similarly, the
minimum half-height width of peaks (= [Width]) is low at the beginning, but is automatically increased as
peaks become broader.
Fig.1-10 Chromatogram Suitable for Auto-setting by [T.DBL]
@ NOTE
In the following cases, set [T.DBL] longer than analysis, instead of setting "0":
(1) Peaks do not become broader with time as in gradient LC analyses or temperature-programmed GC.
(2) As shown in Fig.1-11, a sharp peak appears first, a broad peak then appears after a while, and again
another sharp peak appears.
Fig.1-11 Chromatogram NOT Suitable for Auto-setting by [T.DBL]
T.DBL = [Width] set value
Later peak retention time
Later peak half-height width
2
Sharp peak
Sharp peak
Large, broad peak
(such as solvent peaks)
1.2 Identification Parameters
1
The software identifies detected peaks according to the values set in the compound table, as well as in
identification parameters.
This section describes the identification parameters, and how detected peaks are identified using these
parameters.
^ Reference
For actual procedures for setting identification parameters, refer to the "Operator's Guide "6.3 Setting the
Peak Identification Parameters"".
1.2.1 Window Method and Band Method
The user can define the time allowance for peak identification. Even though a peak's measured retention
time deviates from the retention time set in the compound table, the software identifies the peak as such,
as long as the deviation falls within the time allowance specified.
There are two methods for setting the time allowance: [Window] (= Time Window) and [Band] (= Time
Band). Click a radio button to select either one of the methods.
Window method
In the Window method, the time allowance is defined in proportion to each peak retention time. The longer
the retention time of each peak, the larger the absolute time allowance for the peak becomes, as shown in
Fig.1-12. Using this method, the user can specify the time allowance for all peaks at once. However,
multiple peaks may be detected within the same time allowance depending on the set value.
The Window method is useful for isocratic LC and isothermal GC analyses, where peak width and
retention time fluctuation increase in later peaks, and it is therefore necessary to increase the time
allowance.
Fig.1-12 Window Method
Peak 1 Peak 2 Peak 3
Time allowance (min) =
Standard retention time (min) Window width (%)
100
+0.02
Window width (%)
Standard retention time for each peak (min)
Band method
In the Band method, the time allowance is defined for each peak. Using this method, the user can
individually specify the time allowance optimum for each peak, as shown in Fig.1-13.
The Band method is useful for gradient LC and temperature programmed GC analyses, where peak width
and retention time remain comparatively unchanged, and peaks appear adjacently.
Fig.1-13 Band Method
@ NOTE
For setting time allowance for each peak, make the [Band] column specifiable beforehand by setting in the
[Table Style] sub-window for the [Compound] tab page (= Compound Table).
The [Default Band Time] value set on the [Identification] tab page becomes the default band time for all
peaks.
1.2.2 Absolute/Relative Retention Time Methods
This software identifies peaks based on the retention time. The user can specify whether the retention time
is to be counted from analysis start (absolute retention time method), or to be corrected using a reference
peak (relative retention time method).
Select [Absolute Rt] or [Relative Rt] from the [Identification Method] drop-down menu.
Absolute retention time method
In this method, the software identifies target peaks according to the equation below, using the preset
standard retention time and time allowance for each peak. It is not necessary to specify a reference peak.
This method is generally used for peak identification.
Peak 1 Peak 2 Peak 3
Time allowance (min) = Bandwidth (min)
Bandwidth for each peak (min)
Standard retention time for
each peak (min)
: Standard retention time of target peak
:
Measured retention time of target peak
:
Time allowance of target peak
T t W < T
t
W
1
Relative retention time method
In this method, the software identifies peaks after relatively correcting the retention time deviation due to
analysis condition change. A preset reference peak is first identified using the absolute retention time
method. Target peaks are then identified according to the equation below.
This method is useful for consecutive analysis, where the retention time varies with use frequency, as well
as for analysis where the time error is significant at sample injection (i.e., analysis start).
To correct the retention time more accurately, set multiple reference peaks. When a target peak exists
within the section between two reference peaks as shown in Fig.1-14, the peak is identified using both
reference peaks by the equation below.
Fig.1-14 Relative Retention Time Method
@ NOTE
The relative retention time is obtained by the following equation:
: Standard retention time of reference peak
: Measured retention time of target peak
: Measured retention time of reference peak
: Time allowance of target peak
:
Standard retention time of reference peak 1
:
Standard retention time of reference peak 2
:
Measured retention time of target peak
:
Measured retention time of reference peak 1
:
Measured retention time of reference peak 2
: Time allowance of target peak
: Relative retention time
: Retention time of target peak
: Retention time of reference peak
: Unretained peak time
T
T
1
t
1
------ t W <
T
T
1
t
t
1
W
T
t t
1
t
2
t
1
-------------- T
2
T
1
( ) T
1
+
)
`

W <
T
T
1
T
2
t
t
1
t
2
W
Reference peak 1
Reference peak 2
Section
Target peak
RRT
t
1
t
0
t
2
t
0
-------------- =
RRT
t
1
t
2
t
0
Reference peak and ISTD peak
When identifying peaks with the relative retention time method, set reference peaks. When quantitating
peaks with the internal standard method, set ISTD peaks.
Note that for identification of the reference and ISTD peak, the peak with the largest area (or height) within
the time allowance is identified as such, in case of either the Window or the Band method.
Therefore, if there is a peak larger than the intended peak within the time allowance, the peak may be
mistakenly identified as the reference (or ISTD) peak, which leads to incorrect identification and
quantitation of other peaks.
To avoid this, be sure to (1) select the largest peak (or height) for the reference or ISTD peak, or (2) set a
small time allowance for such peaks so that peaks larger than the intended peak do not fall within the time
allowance.
Fig.1-15 Reference Peak Identification
Time allowance of
reference peak ID #
To be identified as the
reference peak.
1
1.2.3 Identification of Adjacent Peaks
When multiple peaks exist within the same time allowance
Fig.1-16 Multiple Peaks within the Same Time Allowance
When multiple peaks exist within the same time allowance, the user can select how to identify those peaks,
from the [Peak Selection] drop-down menu: [All Peaks], [Closest Peak], [Largest Peak], and [Similarity]
(PDA).
@ NOTE
For identification of ISTD and reference peaks, the software always picks up the largest (highest) peak within
the time allowance, even though [Peak Selection] is set to [All Peaks], [Closest Peak], or [Similarity] (PDA).
Fig.1-17 Identifying ISTD Peak and Reference Peak
Parameter Description
All Peaks Identifies all peaks.
Closest Peak Identifies only the peak closest to the retention time set in the compound table.
Largest Peak Identifies only the peak with the largest area (or height).
Similarity
(PDA)
Identifies only the peak with the highest similarity between the spectrum at the peak retention time
and the standard spectrum set in the compound table.
Time allowance
Peak 2 Peak 1
The largest (highest) peak
Time allowance
When one peak exists within multiple time allowances with different ID #s
When multiple time allowances with different ID numbers overlap, and a peak falls within the overlapped
section, the software identifies the peak in a different manner depending on the Window or Band method.
Fig.1-18 One Peak Included in Two Component Time Allowances
Window method
The peak is identified as the ID No. whose standard retention time is the closest to the peak retention time.
In the example in Fig.1-18, the peak is identified as ID No. 2.
Band method
The peak is identified as the smallest ID No.
In the example in Fig.1-18, the peak is identified as ID No. 1.
When identifying adjacent peaks as different ID #s
When peaks are very close to each other as shown in Fig.1-19, it may be difficult to specify the time
allowance for each peak. To identify and quantitate those peaks, use the Window method and set each
peak retention time to the same value.
By doing so, the peaks can be identified in the order of resolution. Note that, however, the value in [Peak
Selection] becomes invalid when using the Window method. Therefore, if an unnecessary peak is
detected, the peaks and ID numbers may become inconsistent, resulting in incorrect identification.
Fig.1-19 Adjacent Peaks
Peak 1
Peak 2
1
1.2.4 Grouping
Grouping is the process of grouping peaks by types, such as homologues and isomers, to perform
calculations for each group.
There are two types of grouping: [Group Calibration] and [Conc. Summation].
Group calibration
In the Group Calibration, the software first obtains the sum of area (or height) of the grouped compound
peaks. Then the calibration curve is created and quantitation is executed per group.
Area/height per group can be obtained by the sum of area/height of the compounds in the same group.
Conc. summation
In the Conc. Summation, the software first creates the calibration curve for each grouped compound, and
quantitates the peak individually. Then the concentrations of all grouped compounds are summed to obtain
the group concentration.
1.3 Quantitative Methods
This software can utilize the following six types of quantitative methods.
This section explains the equations and rounding method applied to each quantitative method.
Quantitative
Method
Description
Area Normalization The ratio of each peak area (height) in relation to all peak area (height) is obtained.
Corrected Area
Normalization
Using standard samples (or from literature), the sensitivity correction factor is first obtained for
all components to be detected. Using the factor, the peak area (height) measured from
unknown sample is corrected. Then the ratio of the corrected peak area (height) in relation to
all peak area (height) is obtained.
Internal Standard Target peaks are quantitated by adding an ISTD substance to both of the standard and
unknown sample, as below. This method can obtain stable results with injection errors
eliminated.
(1) A certain amount of ISTD substance is added to the standard containing the target
component with known quantity.
(2) The standard is measured, and the calibration curve of area (height) ratio and
concentration (component amount) ratio (between target and ISTD) is created.
(3) The area ratio (between target and ISTD) is obtained from the measured unknown sample
with the same ISTD substance added.
(4) Using the concentration (component amount) ratio acquired from the calibration curve
above, the target component is thus quantitated from the sample and ISTD amounts.
External Standard
(Absolute
Calibration Curve)
By creating a calibration curve expressing the relation between standard concentration
(component amount) and peak area (height), target component is quantitated by applying
unknown sample peak area (height) to the calibration curve.
Corrected Area
Normalization with
Scale Factor
The content of each component is obtained by calculating the total quantitation value as the
dilution factor (or sample volume, if no dilution factor is set) instead of 100.
Standard Addition Unknown sample with a known amount of objective component and unknown sample without
addition are measured in the same conditions. The objective component is then quantitated
using the difference of two obtained peak areas (heights). This method is useful for analyses
where the target component relative sensitivity varies with solvent composition or coexisting
components (as in head space GC).
1
1.3.1 Quantitative Methods and Equations (with Dilution Factor)
@ NOTE
Set [Dilution Factor] to either [Apply] or [Not Used] in the [Data Processing Setting] sub-window in the
[System Settings] sub-window.
When X-axis is area/height
Quantitative
Method
One Point Two Points Content
Area Normalization - -
Content (%) =
Corrected Area
Normalization Content (%) =
Internal Standard
Content =
External Standard
(Absolute
Calibration Curve)
Content =
Corrected Area
Normalization with
Scale Factor
Content (%) =
Standard Addition -
Content =
A
i
A
i
------------- 100
F1
C
1
A
1
------ = F1
C
1
C
2
A
1
A
2
------------------- =
F2 C
2
F1 A
2
=
F1
i
A
i
F2
i
F1
i
A
i
F2 +
i
( )
----------------------------------------- 100
F1
C
1
Cisn
1
--------------
A
1
Aisn
1
--------------- = F1
C
1
Cisn
1
--------------
C
2
Ci sn
2
--------------
\ .
| |
A
1
Aisn
1
---------------
A
2
Aisn
2
---------------
\ .
| |
=
F2
C
2
Cisn
2
-------------- F1
A
2
Aisn
2
--------------- =
F1
i
A
i
Aisn
-------------- F2
i
+
\ .
| |
Wisn
Wspl
-------------- DFACT
F1
C
1
A
1
------ = F1
C
1
C
2
A
1
A
2
------------------- =
F2 C
2
F1 A
2
=
F1
i
A
i
F2
i
+
Wspl
----------------------------- DFACT
F1
C
1
A
1
------ = F1
C
1
C
2
A
1
A
2
------------------- =
F2 C
2
F1 A
2
=
F1
i
A
i
F2
i
F1
i
A
i
F2 +
i
( )
----------------------------------------- DFACT
F1
C
1
C
2
A
1
A
2
------------------- =
F2 C
2
F1 A
2
=
F2
i
Wspl
-------------- DFACT
When X-axis is concentration (component amount)
Quantitative
Method
Content (%) =
Corrected Area
Normalization
Content (%) =
Internal Standard Content =
External Standard
(Absolute
Calibration Curve)
Content =
Corrected Area
Normalization with
Scale Factor
Content (%) =
Standard Addition -
Content =
: Peak area (height) of Standard 1
: Peak area (height) of Standard 2
: Concentration (component amount) of Standard 1
: Concentration (component amount) of Standard 2
: Peak area (height) of Nth ISTD peak in Standard 1
: Peak area (height) of Nth ISTD peak in Standard 2
: Concentration (component amount) of Nth ISTD peak in Standard 1
: Concentration (component amount) of Nth ISTD peak in Standard 2
: Sample amount of Standard 1
: Sample amount of Standard 2
: Slope compensation factor
: Constant compensation factor
: Peak area (height)
: Area (height) of Nth ISTD peak
: Sample amount
: Amount of the Nth ISTD
A
i
A
i
------------- 100
F1
A
1
C
1
------ = F1
A
1
A
2
C
1
C
2
------------------- =
F2 A
2
F1 C
2
=
A
i
F2
i
( ) F1
i
A
i
F2 ( )
i
F1
i
--------------------------------------------- 100
F1
A
1
Aisn
1
C
1
C i sn
1
------------------------- = F1
A
1
Aisn
1
---------------
A
2
Aisn
2
---------------
\ .
| |
C
1
Cisn
1
--------------
C
2
Cisn
2
--------------
\ .
| |
=
F2
A
2
Aisn
2
--------------- F1
C
2
Cisn
2
-------------- =
A
i
Aisn F2
i
( )
F1
i
-----------------------------------------
Wisn
Wspl
-------------- DFACT
F1
A
1
C
1
------ = F1
A
1
A
2
C
1
C
2
------------------- =
F2 A
2
F1 C
2
=
A
i
F2
i
F1
i
-------------------
DFACT
Wspl
---------------------
F1
A
1
C
1
------ = F1
A
1
A
2
C
1
C
2
------------------- =
F2 A
2
F1 C
2
=
A
i
F2 ( ) F1
i
A
i
F 2
i
( ) F1
i
------------------------------------------ DFACT
F1
A
1
A
2
C
1
C
2
------------------- =
F2 A
2
F1 C
2
=
F2
i
F1
i
--------
DFACT
Wspl
---------------------
A
1
A
2
C
1
C
2
Aisn
1
Aisn
2
Cisn
1
Cisn
2
Wspl
1
Wspl
2
F1
i
F2
i
Ai
Aisn
Wspl
Wisn
1
1.3.2 Quantitative Methods and Equations (without Dilution Factor)
Quantitative
Method
Content (%) =
Corrected Area
Normalization
Content (%) =
Internal Standard
Content =
External Standard
(Absolute
Calibration Curve)
Content =
Corrected Area
Normalization with
Scale Factor
Content (%) =
Standard Addition -
Content =
A
i
A
i
------------- 100
F1
C
1
A
1
------ = F1
C
1
C
2
A
1
A
2
------------------- =
F2 C
2
F1 A
2
=
F1
i
A
i
F2
i
F1
i
A
i
F2 +
i
( )
----------------------------------------- 100
F1
C
1
Cisn
1
--------------
A
1
Aisn
1
--------------- = F1
C
1
Cisn
1
--------------
C
2
Ci sn
2
--------------
\ .
| |
A
1
Aisn
1
---------------
A
2
Aisn
2
---------------
\ .
| |
=
F2
C
2
Cisn
2
-------------- F1
A
2
Aisn
2
--------------- =
F1
i
A
i
Aisn
-------------- F2
i
+
\ .
| |
Wisn
Wspl
-------------- 100
F1
Wspl C
1
100 A
1
-------------------------- = F1
C
1
Wspl
1
C
2
Wspl
2

100 A
1
A
2
( )
---------------------------------------------------------------- =
F2
C
2
Wspl
2
100
----------------------------- F1 A
2
=
F1
i
A
i
F2
i
+
Wspl
----------------------------- 100
F1
C
1
A
1
------ = F1
C
1
C
2
A
1
A
2
------------------- =
F2 C
2
F1 A
2
=
F1
i
A
i
F2
i
F1
i
A
i
F2 +
i
( )
----------------------------------------- Wspl
F1
C
1
Wspl
1
C
2
Wspl
2

100 A
1
A
2
( )
---------------------------------------------------------------- =
F2
C
2
Wspl
2
100
----------------------------- F1 A
2
=
F2
i
Wspl
-------------- 100
Quantitative
Method
Content (%) =
Corrected Area
Normalization
Content (%) =
Internal Standard
Content =
External Standard
(Absolute
Calibration Curve)
Content =
Corrected Area
Normalization with
Scale Factor
Content (%) =
Standard Addition -
Content =
A
i
A
i
------------- 100
F1
A
1
C
1
------ = F1
A
1
A
2
C
1
C
2
------------------- =
F2 A
2
F1 C
2
=
A
i
F2
i
( ) F1
i
A
i
F2 ( )
i
F1
i
--------------------------------------------- 100
F1
A
1
Aisn
1
C
1
C i sn
1
------------------------- = F1
A
1
Aisn
1
---------------
A
2
Aisn
2
---------------
\ .
| |
C
1
Cisn
1
--------------
C
2
Cisn
2
--------------
\ .
| |
=
F2
A
2
Aisn
2
--------------- F1
C
2
Cisn
2
-------------- =
A
i
Aisn F2
i
( )
F1
i
-----------------------------------------
Wisn
Wspl
-------------- 100
F1
A
1
100
C
1
Wspl
-------------------------- = F1
100 A
1
A
2
( )
C
1
Wspl
1
C
2
Wspl
2

------------------------------------------------------------- =
F2 A
2
F1 C
2
Wspl
2
100 =
A
i
F2
i
F1
i
-------------------
100
Wspl
--------------
F1
A
1
C
1
------ = F1
A
1
A
2
C
1
C
2
------------------- =
F2 A
2
F1 C
2
=
A
i
F2 ( ) F1
i
A
i
F 2
i
( ) F1
i
------------------------------------------ Wspl
F1
100 A
1
A
2
( )
C
1
Wspl
1
C
2
Wspl
2

------------------------------------------------------------- =
F2 A
2
F1 C
2
Wspl
2
100 =
F2
i
F1
i
--------
100
Wspl
--------------
1
1.3.3 Rounding
The following table shows the number of significant digits and rounding method applied to CLASS-LC10,
CLASS-VP and LabSolutions.
*1 For CLASS-VP, only the quantitative calculation factors in the original method used for analysis are saved in data files.
@ NOTE
The parameters internally calculated for output (marked with "-") are not saved in data files.
[Reference]
Single precision floating
Floating-point numbers use the IEEE format. Single-precision values consist of an 8-bit binary exponent,
and a 24-bit mantissa. Significant digits are 6 to 7 for decimal format. Available range of exponent part is
10
38
.
Double precision floating
Double precision values consist of an 11-bits exponent and a 53-bit mantissa. Significant digits are 15 to 16
for decimal format. Available range of exponent part is 10
308
.
^ Reference
For LabSolutions, the user can set the rounding method and digits of values displayed on the screen and
reports.
For detailed procedures, refer to the System User's Guide.
CLASS-LC10 CLASS-VP LabSolutions
Parameter
Internal
Calculation
Stored
Format
Internal
Calculation
Stored
Format
Internal
Calculation
Stored
Format
Peak
Integration
Area 48-bit integer
(unit: 0.1)
48-bit integer
(unit: 0.1)
Double
precision
Double
precision
Double
precision
Double
precision
Sum of Areas 48-bit integer
(unit: 0.1)
- Double
precision
- Double
precision
-
Height 32-bit integer
(unit: 1/16)
32-bit integer
(unit: 1/16)
Double
precision
Double
precision
Double
precision
Double
precision
Sum of heights 32-bit integer
(unit: 1/16)
- Double
precision
- Double
precision
-
Quantitation Concentration Single
precision
Single
precision
Single
precision
Single
precision
Double
precision
Double
precision
Sum of
Concentrations
Single
precision
- Single
precision
- Double
precision
-
Quantitative
Calculation
Factor
Single
precision
- Single
precision
Single
precision
(*1)
Double
precision
Double
precision
Column
Performance
Separation
Factor
Single
precision
- Single
precision
Single
precision
Double
precision
Double
precision
Number of
Theoretical
Plate
Single
precision
- Single
precision
Single
precision
Double
precision
Double
precision
Number of
Theoretical
Plate/meter
Single
precision
- Single
precision
Single
precision
Double
precision
Double
precision
Capacity
Factor(k')
Single
precision
- Single
precision
Single
precision
Double
precision
Double
precision
Resolution Single
precision
- Single
precision
Single
precision
Double
precision
Double
precision
Tailing Factor Single
precision
- Single
precision
Single
precision
Double
precision
Double
precision
Tailing Factor
(10%)
None None Single
precision
Single
precision
Double
precision
Double
precision
1.4 Concentration Deviations between Compound Table
and Calibration Curve
In the [Quantitative Results View] of the [Quant Browser] window, the user can verify the accuracy
(Accuracy [%]) and deviation (%Deviation) between the standard concentration in the compound table and
that obtained from the calibration curve.
This section describes the equation for obtaining Accuracy [%] and %Deviation.
^ Reference
For procedures for setting Accuracy [%] and %Deviation, refer to the help files.
1.4.1 Accuracy [%]
[Accuracy [%]] can be obtained by the following equation:
When [Sample Type] is [Standard], [Control], or [Unknown(QA/QC)]
@ NOTE
Concentration obtained from calibration curve = (Concentration displayed in the compound result table)
(Sample volume set in the [Single Run] sub-window or batch table at data acquisition) / Dilution factor
To verify the current sample amount and dilution factor, navigate from the data file's [Properties] sub-window
to the [Sample Info.] tab page.
When [Sample Type] is [Spiked]
1.4.2 %Deviation
[%Deviation] can be obtained by the equation described below. The software executes the statistic
calculation based on the absolute value of [%Deviation].
When [Sample Type] is [Standard], [Control], or [Unknown(QA/QC)]
When [Sample Type] is [Spiked]
Accuracy [%] : Concentration obtained from calibration curve
: Concentration set to the relevant level in the compound table
C
r
C
c
100 = C
r
C
c
Accuracy [%] : Concentration obtained from calibration curve
: Spiked amount set in the compound table
C
r
C
s
100 = C
r
C
s
%Deviation : Concentration obtained from calibration curve
: Concentration set to the relevant level in the compound table
C
r
C
c
( ) C
c
100 = C
r
C
c
%Deviation : Concentration obtained from calibration curve
: Spiked amount set in the compound table
C
r
C
s
( ) C
s
100 = C
r
C
s
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2 Calibration Curve
This chapter describes types of calibration curves, equations used in calibration, the standard
concentration correction factor, as well as calibration curve creation with multiple standards.
"2.1 Calibration Curve Type" P.31
"2.2 Calibration Curve Correction" P.39
2.1 Calibration Curve Type
This software can utilizes the following seven types of calibration curves: [Linear], [Point to point],
[Quadratic], [Cubic], [Mean RF], [Exponential], and [Manual RF].
2.1.1 Linear
Standards with different concentrations up to the [# of Calib. Levels] are individually analyzed, and a linear
calibration curve is drawn by the least square method.
When using a standard at one concentration level, the calibration curve is drawn as a simple linear line
passing through the obtained point and the origin.
When using the standard at two concentration levels, and when the calibration curve does not go through
the origin, the calibration curve is drawn as a simple linear line passing through two obtained points.
In other cases, the calibration curve is drawn as a linear line using the least square method.
The user can set the [# of Calib. Levels] up to 64, and obtain each level's calibration point by averaging up
to ten measurements.
2 Calibration Curve
2.1.2 Point to Point
Standards with different concentrations up to the [# of Calib. Levels] are individually analyzed, and the
calibration curve is drawn by connecting the obtained points.
When using a standard at one concentration level, the calibration curve is drawn as a simple linear line
passing through the obtained point and the origin.
When using the standard at two or more concentration levels, the calibration curve is drawn by connecting
those obtained points. The calibration point at the smallest concentration level is always connected to the
origin.
2.1.3 Quadratic and Cubic
Standards with different concentrations up to the [# of Calib. Levels] are individually analyzed, and a
quadratic or cubic calibration curve is drawn by the least square method.
To draw a quadratic calibration curve, it is required to set calibration points at three or more levels.
To draw a cubic calibration curve, it is required to set calibration points at four or more levels.
2
2.1.4 Mean RF
Standards with different concentrations up to the [# of Calib. Levels] are individually analyzed, and each of
the obtained points is connected to the origin. The slope coefficients of all linear lines are then simply
averaged.
2.1.5 Exponential
Standards with different concentrations up to the [# of Calib. Levels] are individually analyzed, and a linear
calibration curve is drawn in a natural log-natural log graph.
When using standards at two concentration levels, the calibration curve is a simple linear line passing
through two obtained points.
In other cases, the calibration curve is drawn as a linear line using the least square method.
@ NOTE
Standards at 2 or more levels are required since a natural log-natural log graph is drawn.
[Zero] can not be set to [Force Through] since there is no origin.
Set [Curve Fit Type] to [Linear], normally.
Averaged
calibration curve
2 Calibration Curve
2.1.6 Manual RF (Linear, Exponential)
The user can manually define an arbitrary calibration curve by entering the slope (= linear coefficient) and
intercept in the compound table. The software then quantitates the peak by the specified calibration curve.
In the Manual RF method, the calibration curve graph displays only the calculated curve with no calibration
point.
Particularly for the Manual RF (Exponential), the calibration curve is created regarding X and Y as natural
log axes.
2.1.7 Quantitation Using Other Component Calibration Curve
When it is difficult to prepare a standard for a component (such as impurity), the user can use a calibration
curve of a standard with the same or proportional relative sensitivity to the detector, in order to quantitate
the component concentration (content).
Fig.2-1 Impurity Content Calculation
The following are the parameters used for the quantitation with the other component calibration curve:
Reference Standard ID
(Ref STD ID)
Specify the ID No. of the peak with standard prepared. Based on the standard calibration
curve, the concentration (content) of a component without standard can be quantitated.
Set "-1 (negative value)" to hide the ID No. component row from the quantitative result
table (i.e., The standard peak with known concentration is excluded from the quantitative
results). In this case, the ID No. row displays "Only reference".
@ NOTE
Be careful NOT to refer two reference standard ID #s to each other. (e.g. ID # 1 = 2,
ID # 2 = 1)
Correction Factor Specify the factor to compensate the sensitivity difference between the relevant peak and
reference standard ID peak.
Correction Factor
= (Target component peak area (height) per weight) / (Related substance peak area
(height) per weight)
2
@ NOTE
[Ref STD ID] and [Correction factor] are not displayed as default in the [Compound] tab page of the [Method
View]. To calculate impurity concentration (content), navigate to [Method View] - [Compound] tab page -
[Table Style] sub-window, and add [Ref STD ID] and [Correction factor] to [Display Items].
Equations
@ NOTE
Correction factor is applied to the peak area (height, area ratio, or height ratio), instead of to the
quantitative result (concentration).
This equation is available for all quantitative methods using a calibration curve. However, the [Correction
factor] value is not used for internal standards (whose [Type] is [ISTD] or [ISTD & Ref.]).
: Calibration curve equation for a compound set to reference standard ID
: Peak concentration (content)
: Peak area (height, area ratio, or height ratio)
: Correction factor
C
i
f A
i
Corr ( ) =
f
C
i
A
i
Corr
2 Calibration Curve
2.1.8 Least Square Method and Weighted Least Square Method
This section describes the least square method specifically for collinear approximations (linear expression:
Y = F1 X + F2). For collinear approximations, contribution rate and correlation coefficient can be obtained
by the equations described below, even in conditions where F1 and F2 are not determined.
The least square method for quadratic and cubic approximations can be expressed by matrix operations,
even though the explanation will be omitted here.
Least square method
The least square method for collinear approximations can be expressed by the following equations:
and values vary according to quantitative methods as follows:
With dilution factor
Without dilution factor
Contribution rate:
Correlation coefficient:
Quantitative Method
Corrected Area Normalization
(including that with scale factor)
Internal Standard
External standard (Absolute
Calibration Curve)
Standard Addition
Quantitative Method
Internal Standard
Calibration Curve)
Standard Addition
F1
N XY
N X
2
---------------------------------------------------------------- =
F2
X
2
XY
N X
2
-------------------------------------------------------------------------- =
r
2
XY
1
N
---- X

\ .
| |
2
X
2
X
\ .
| |
2
N
--------------------
)

`

Y
2
Y
\ .
| |
2
N
--------------------
)

`

------------------------------------------------------------------------------------------------ =
r
XY
1
N
---- X

\ .
| |
X
2
X
\ .
| |
2
N
--------------------
)

`

Y
2
Y
\ .
| |
2
N
--------------------
)

`

---------------------------------------------------------------------------------------------------- =
X Y
X Y
A
i
C
i
A
i
Ais
i
C
i
Cis
i
A
i
C
i
X Y
A
i
C
i
A
i
Ais
i
C
i
Cis
i
A
i
C
i
Wspl
i
100
2
With dilution factor
Without dilution factor
Weighted least square method
The least square method usually assumes the same degree of error at each concentration level, and
obtains the calibration curve satisfying that the sum of the squares of the deviation from each level is the
least possible value. This method, however, can create relatively large errors at the levels with low
concentrations (component amounts), which do not reflect the actual case. To solve this problem, the
calibration curve is calculated so that the sum of the squares of deviation of the ratio between each level
and its concentration becomes the least possible value, equating the relative error at each level in relation
to the concentration. This is called weighting.
Quantitative Method
Internal Standard
Calibration Curve)
Standard Addition
Quantitative Method
Internal Standard
Calibration Curve)
Standard Addition
: The number of calibration points : Internal standard peak area (height)
: Peak area (height) : Internal standard peak concentration
(component amount)
: Peak concentration (component amount) : Sample amount
(Subscript i indicates the i th standard.)
X Y
C
i
A
i
C
i
Cis
i
A
i
Ais
i
C
i
A
i
X Y
C
i
A
i
C
i
Cis
i
A
i
Ais
i
C
i
Wspl
i
100 A
i
N Ais
i
A
i
Cis
i
C
i
Wspl
i
2 Calibration Curve
The weighted least square method for collinear approximations can be expressed by the following
equations.
where W signifies "weight". The four types of [Weighting Method] are available: [1/C], [1/C^2], [1/A], and [1/
A^2].
Set [1/C^2] or [1/A^2] to weight in proportion to the concentration (component amount). The effect of
weighting becomes smaller when [1/C] or [1/A] is set. Normally, use [1/C] or [1/C^2] when the Y-axis is
concentration, and use [1/A] and [1/A^2] when the Y-axis is area/height.
@ NOTE
X and Y values vary with quantitative methods. For the definitions of X and Y, refer to " When X-axis is area/
height" P.36, and " When X-axis is concentration (component amount)" P.37 in "Least square method".
Contribution rate:
Correlation coefficient:
1/C Weights as much as 1/C. "C" signifies concentration (component amount).
1/C^2
Weights as much as 1/C
2
. "C" signifies concentration (component amount).
1/A Weights as much as 1/A. "A" signifies [Area] or [Height] set in [Calculated by] in the [Quantitative] tab
page.
1/A^2
Weights as much as 1/A
2
. "A" signifies [Area] or [Height] set in [Calculated by] in the [Quantitative] tab
page.
F1 WXY
WX
WY
----------------------------------------------
\ .
|
|
| |
WX
2
WX
\ .
| |
2
W
--------------------------
\ .
|
|
|
| |
=
F2
W
i
Y
i
W
i
--------------------- F1
W
i
X
i
W
i
--------------------- =
r
2
WXY
WX
WY
-------------------------------------------
\ .
|
|
| |
2
WX
2
WX
\ .
| |
2
W
--------------------------
)

`

WY
2
WY
\ .
| |
2
W
--------------------------
)

`

------------------------------------------------------------------------------------------------------------------------ =
r
WXY
WX
WY
-------------------------------------------
\ .
|
|
| |
WX
2
WX
\ .
| |
2
W
--------------------------
)

`

WY
2
WY
\ .
| |
2
W
--------------------------
)

`

---------------------------------------------------------------------------------------------------------------------------- =
2.2 Calibration Curve Correction
2
This section describes procedures for creating calibration curves (1) by applying a compensation factor to
standard concentrations in the compound table, and (2) by using multiple standards.
2.2.1 Calibration Curves Using Standard Concentration Factors
To create calibration curves, the software uses true concentrations automatically calculated by multiplying
each level concentration by [Std Conc Factor]. In the conventional method, when using the standards with
concentrations of 1, 0.5, 0.25, and 0.125 prepared from a stock solution with concentration of 0.998, the
user needed to multiply each concentration by 0.998, and enter the values into the compound table one by
one. Moreover, all concentration values needed to be replaced, each time preparing a new stock solution.
If [Std Conc Factor] is used in such case, it is only required to change the [Std Conc Factor] value without
changing any concentrations set in the compound table.
Fig.2-2 Example of Using [Standard Concentration Factor]
Conventional compound table setting
Using standard concentration factors
@ NOTE
To use the standard concentration factor, navigate to the [Compound] tab page - [Table Style] sub-window,
and add [Std Conc Factor] to [Display Items]. By doing so, the [Std Conc Factor] column becomes
specifiable in the compound table.
Application example of [Std Conc Factor]
Commercially-available standard solutions with traceability come with not only the concentration value, but
also the f value (factor) to correct the concentration. When creating calibration curves for them, enter the
concentrations of the prepared standard into the compound table [Conc.] columns, and enter the f value
into the [Std Conc Factor] column. This setting allows obtaining true concentrations automatically, without
time-consuming manual calculations (See Fig.2-2).
Std Conc Factor Calibration curves are created based on concentrations (component amounts) derived by
multiplying each level concentration by the set value.
ID # Compound Name Conc. (1) Conc. (2) Conc. (3) Conc. (4)
1 Something 0.12475 0.2495 0.499 0.998
ID # Compound Name Conc. (1) Conc. (2) Conc. (3) Conc. (4) Std Conc Factor
1 Something 0.125 0.25 0.5 1 0.998
Diluted to 1/2 Diluted to 1/2 Diluted to 1/2
True concentration 0.998 0.499 0.2495 0.12475 (Stock solution
concentration = 0.998)
Target concentration 1.00 0.500 0.250 0.125
2 Calibration Curve
2.2.2 Calibration Curves When Target Components Refer to Multiple
Standards Separately
To quantitate multiple components in an unknown sample, it is required to create a calibration curve for
each component. For example, when having acquired the standard data Std-A1 and Std-A2 for Component
A, and Std-B1 and Std-B2 for Component B, (1) create the calibration curves for Components A and B in
the same method file. (2) Then, using the method file, quantitate the concentrations (contents) of
components A and B contained in unknown sample data (Unk) already acquired. The detailed procedure
will be given in the following. (Note that the user can also set this operation from the [Realtime Batch]
window, if the retention time of target component is known.)
1
Double click the unknown sample data file (Unk), which contains the retention time data
of the target component peaks A and B, in the [Data Explorer] sub-window of the
[Postrun Analysis] program.
The Data Analysis window is displayed, and the data file (Unk) is loaded to the software.
1 Click (Edit Mode) on the [Method View].
The software goes into the edit mode.
2 Click the [Quantitative] tab.
Set [# of Calibration Levels] to 4.
Std-A1
Std-A2
Std-B1
Std-B2
Component A Component B
Unk
2
3 Click the [Compound] tab.
Specify the concentrations (1) to (4) of Components A and B in the compound table.
To set [Not Used], press the [Delete] key.
4 Click (View Mode) on the [Method View].
The software goes into the view mode.
5 Click the (Apply to Method) icon in the [Data Analysis] assistant bar.
The [Save Method As] sub-window is displayed.
Verify the file name, and follow the screen instructions. The set parameters are then applied to the
method file.
2
Click the title labeled as [Main] on the assistant bar.
The [Main] assistant bar for the [Postrun Analysis] program is displayed.
3
Click the (Postrun Batch) icon on the [Main] assistant bar.
The [Postrun Batch] window is displayed.
1 Create the batch table by specifying the method file saved in the procedure above.
2 Specify the sample types.
Set Std-A1 in the first row to "Standard-Initialize Calibration Curve (I)". Set Std-A2, Std-B1, and Std-
B2 in the subsequent rows to "Standard-Add Calibration Level".
3 Specify the level numbers.
Set Std-A1, Std-A2, Std-B1, and Std-B2 to level number 1, 2, 3, and 4, respectively.
4
Click the (Start Postrun Batch) icon on the [Postrun Batch] assistant bar.
The postrun batch process begins.
The calibration curve for Component A is created by the level 1 and 2 data. The calibration curve for
Component B is created by the level 3 and 4 data. By applying the method file to the unknown sample
data, the software quantitate the concentrations (component amounts) of Components A and B, based on
the created calibration curves.
2 Calibration Curve
This page is intentionally left blank.
3
33
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3 Equations
This chapter describes various calculation equations used in this software.
"3.1 Noise/Drift Calculation Parameters" P.43
"3.2 QA/QC Parameters" P.45
"3.3 Column Performance Equations" P.48
"3.4 Background Correction Processing" P.55
3.1 Noise/Drift Calculation Parameters
This section describes parameters and equations to obtain the detection/quantitative limit from the noise
values calculated from measured chromatograms.
To verify the current noise/drift parameters, navigate to [Data Analysis window] - [Method View] -
[Integration] tab page - [Noise/Drift Calculation Settings] sub-window.
3.1.1 Noise Calculation Methods
There are three methods for calculating noise: ASTM, rms, and EP.
ASTM
A 15-minute baseline period is first separated into 30 second sections. Noise width of each section is
individually calculated as described below. Then the noise widths of all sections are averaged to obtain the
ASTM noise.
[Calculation procedure]
(1) The slope of the approximated linear line for all data points within each section is calculated by the least
square method. (2) Two parallel lines are assumed by shifting the slope upward and downward, so that all
data points within the section fall between two lines. (3) The distance between the two parallel lines in the
intensity axis direction is defined as the noise width for the section.
rms
The rms noise refers to the standard deviation of the measurement point signals detected within a
specified section.
EP
A time range of 20 times the half-height width of the target solution peak is first defined, placing the peak
retention time in its center. The blank solution is analyzed to obtain its chromatogram. The time range
above is applied to the blank chromatogram. The EP noise is then obtained as follows.
[Calculation procedure]
(1) The slope of the approximated linear line for all blank data points within the time range is calculated by
the least square method. (2) Two parallel lines are assumed by shifting the slope upward and downward,
so that all data points within the section fall between the two lines. (3) The distance between the two
parallel lines in the intensity axis direction is defined as the noise width of the time range, and half the
distance is defined as the EP noise.
3 Equations
3.1.2 Detection/Quantitative Limit Coefficients
Detection limit refers to the minimum amount of a target component that can be detected as such. In this
case, it is not necessary to be able to quantitate the target component. This means that the obtained value
does not need to have the accuracy and precision required for quantitation. Namely, the detection limit is
the minimum amount of a component to be recognized as a peak on LC/GC chromatograms.
Quantitative limit refers to the minimum amount of a target component that can be quantitated as such with
required accuracy and precision.
To obtain detection/quantitative limit, there are three methods roughly described below. The method to be
applied varies according to analysis methods (e.g. whether or not the method is instrumental analysis):
Method based on visual judgment
Method referring to signal-noise relation
Method referring to response standard deviation and calibration curve slope
In LC and GC analyses, the second and third methods are normally adopted. (i.e., methods referring to
signal-noise relation/ response standard deviation and calibration curve slope).
This software obtains detection/quantitative limits using the second method, the one referring to signal-
noise relation. To obtain those values, specify detection/quantitative limit coefficients in the [Noise/Drift
Calculation Settings] sub-window.
@ NOTE
Detection/quantitative limits are expressed in concentrations. Therefore, to obtain those values, it is
necessary to create the calibration curve and determine each peak concentration beforehand.
^ Reference
Use QA/QC functions for obtaining detection/quantitative limits by the method referring to the response
standard deviation and calibration curve slope. For details on the QA/QC functions, refer to the help files.
3.1.3 Drift Settings
To obtain drift values, first define the calculation section by specifying its start and end times in the [Noise/
Drift Calculation Settings] sub-window. The software calculates the slope of the approximated linear line for
data points within the section, using the least square method. The variation per hour is then calculated
from the slope, to obtain the drift value.
: Detection limit
: Coefficient (normally 3 to 3.3)
: Peak concentration
: Noise
: Peak height
: Quantitative limit
: Coefficient (normally 10)
: Peak concentration
: Noise
: Peak height
DL Conc N S = DL
Conc
N
S
QL Conc N S = QL
Conc
N
S
3.2 QA/QC Parameters
3
The user can set various QA/QC parameters and criteria to evaluate the validity of the analytical system.
This section explains the content of the QA/QC parameters and equations used.
To verify the current QA/QC parameters, select [QA/QC Parameters] from the [Method] menu of the [Data
Acquisition] window, to display [QA/QC] sub-window.
3.2.1 Common Items
3.2.2 Calibration
Item Equation
Averaging
: Mean
: Data
: Number of the data
Standard Deviation : Standard deviation
: Data
: Mean data
Relative Standard
Deviation (%)
Relative standard deviation (%)
: Standard deviation
: Mean data
x
X
i
i 1 =
N
N
-------------- =
x
x
i
N
s
x
i
x ( )
2
i 1 =
N
N 1
----------------------------- =
s
x
i
x
N
s 100
x
----------------- =
s
x
Item Equation
Correlation
Coefficient
: Correlation coefficient
: Concentration (Concentration ratio)
: Mean
: Area (Area ratio) or height (height ratio)
: Mean
Residual Sum of
Squares Residual sum of squares
: Calibration curve slope
*1
Residual SD of
Regression Line
: Residual SD of regression line
: Slope of calibration curve
*1
Residual SD of Y
Intercepts
: Residual standard deviation of y intercept
: Residual SD of regression line
: X-axis data of calibration curve
: Mean
r
x
i
x
( ) y
i
y
( ) { }
i 1 =
N
x
i
x
( )
2
i 1 =
N
)

`

y
i
y
( )
2
i 1 =
N
)

`

------------------------------------------------------------------------------- =
r
x
i
x x
i
y
i
y y
i
N
y
i
y
( )
2
i 1 =
N
=
y
i
y
N
S
y/x
y
i
y
( )
2
i 1 =
N
N 2
------------------------------ =
S
y/x
y
i
y
N
S
y
S
y/x
x
i
2
i 1 =
N
N x
i
x
( )
2
i 1 =
N
----------------------------------- =
S
y
S
y/x
x
i
x x
i
N
3 Equations
*1: Use a linear calibration curve not passing through the origin.
Response Factor : Response factor
: Signal
: Concentration
When the internal standard method is used:
: Response factor
: Signal
: Concentration
: Signal of the ISTD
: Concentration of the ISTD
S/N S/N refers to the ratio between signal and noise level. This software does not obtain the noise
level directly from equations, but from repeated calculations using parallel lines as below:
The figure above illustrates a case obtaining the noise level in 0.5 min steps within 33.5 to
36.5 min. First, the target chromatogram is divided into 0.5 min interval sections. Two parallel
lines satisfying the conditions below are then assumed for each section.
-The upper line stays above all data points within the section.
-The lower line stays under all data points within the section.
-The distance between two lines in the intensity axis direction should be the minimum possible
value.
The parallel-line distances in all sections are then averaged to obtain the noise level.
The user can specify the time interval in the [S/N Detail Settings] sub-window. If [Automatically
calculate the baseline outside detected peaks] is selected, chromatogram data is automatically
divided into 0.5 minute intervals and calculated.
Detection Limit The detection limit is obtained by the
following equation.
: Detection limit
: Standard deviation of signal for zero
concentration
: Coefficient (normally 3 to 3.3)
*1
According to the method, select the standard deviation of signals for zero concentration from
the next: Residual SD of regression line, Residual SD of Y intercepts, and S/N. When using an
internal standard, the detection limit is calculated from signal ratio and concentration ratio, and
the result is expressed as concentration ratio.
To obtain detection limits based on noise, set the Quantitative Parameters [Calculated by] to
[Height], and [X Axis of Calib. Curve] to [Conc.].
Quantitative Limit The quantitative limit is obtained by the
following equation.
: Quantitative limit
: Standard deviation of signal for zero
concentration
: Coefficient (normally 10)
*1
Deviation (%) Signifies the deviation of the calibration
point at each level from the created
calibration curve. [Deviation (%)] is obtained
by the following equation:
: Deviation (%) at level N
: Concentration function based on
calibration point signal intensity S at
level N
: Specified concentration at level N
Item Equation
RF
S
Conc
-------------- =
RF
S
Conc
RF
S Conc
ISTD
Conc S
ISTD
---------------------------------- =
RF
S
Conc
S
ISTD
Conc
ISTD
DL S
B
f =
DL
S
B
f
QL S
B
f =
DL
S
B
f
Deviat ion
N
f S
N
( ) Conc
N
Conc
N
------------------------------------- 100 =
Deviat ion
N
f S
N
( )
Conc
N
3
3.2.3 Quality Control
3.2.4 Recovery
3.2.5 Degradation Check
3.2.6 Noise/Drift Check
Item Equation
Mean Deviation :Mean Deviation
: Mean concentration
: True concentration value
Method
Detection Limit
: Method detection limit
: Standard deviation of the concentration
: 100% confidence interval with N-1 degree of freedom
Deviat ion
Mean
Conc
Mean
Conc
True
---------------------- =
Deviat ion
Mean
Conc
Mean
Conc
True
MDL s t n 1 ( , ) = MDL
s
t n 1 ( , )
Item Equation
Recovery When calculating from the spike amount
: Recovery
: Blank
: Spike amount
When calculating with standard (ISTD recovery)
: Recovery of the compound
: Area of the compound
: Mean area of the compound in sample specified as
[Standard (ISTD Recovery)] sample type
: Recovery standard
R
Conc Blank
SpikeAmount
--------------------------------- =
R
Blank
SpikeAmount
R
i
A
i
A
i0
Ai Ai0
----------------- =
R
i
i
A
i
i
Ai i
i0
Item Equation
Breakdown : Breakdown rate (%)
: Area of analyte
: Total area in the same group except analyte
breakdown
Area
rest
Area
analyt e
Area
rest
+
------------------------------------------------------ = 100
breakdown
Area
analyt e
Area
rest
Item Equation
Noise
(ASTM)
(1) A 15 minute baseline period (specifiable by the user) is first separated into 30 second sections. (2) The
slope of the approximated linear line for all data points within each section is calculated by the least square
method. (3) Two parallel lines are assumed by shifting the slope upward and downward, so that all data
points fall between the two lines. (4) The distance between the two parallel lines in the intensity axis
direction is defined as the noise width for the section. (5) The mean value of all section noise widths is
defined as the ASTM noise.
Noise
(rms)
The rms noise refers to the standard deviation of the measurement point signals detected within a
specified section.
Noise
(EP)
(1) The time range of 20 times the half-height width of the target solution peak is defined, placing the peak
retention time in its center. (2) The blank solution is analyzed to obtain its chromatogram. (3) The time
range above is applied to the blank chromatogram. (4) The slope of the approximated linear line for all
blank data points within the time range is calculated by the least square method. (5) Two parallel lines are
assumed by shifting the slope upward and downward, so that all data points fall between the two lines. (6)
The distance between the two parallel lines in the intensity axis direction is defined as the noise width of
the time range, and half of the distance is defined as the EP noise.
Drift (1) The slope of the approximated linear line for all data points within a specified section is obtained by the
least square method. (2) The variation per hour is calculated from the slope to obtain the drift value.
3 Equations
3.3 Column Performance Equations
USP
@ NOTE
"Unretained Peak Time" refers to the time when a compound with no adsorption to the stationary phase is
eluted. To use the first-eluted peak as the reference, set [Unretained Peak Time] in the [Performance] tab-
page to [1st Peak Time].
Item Equation
Number of Theoretical
Plate
: Number of theoretical plate
: Retention time
: Peak width. The peak width is the time width between two
points at the intersections between straight lines through the points
of inflection to the left and right of a peak and the baseline.
HETP
(m)
: HETP
: Column length (mm)
Tailing Factor : Tailing factor
: Peak width at 5% height of peak
: Width of the front half of the peak (from the start point to the
apex) at 5% height of the peak
Resolution : Resolution
: Retention time
: Retention time of the previous peak
: Peak width. The peak width is the time width between two
points at the intersections between straight lines through the points
of inflection to the left and right of a peak and the baseline.
: Peak width for previous peak
Capacity Factor (k') : Retention time
Separation Factor : Separation factor
: Capacity factor (k') of the previous peak (Peak 1)
: Capacity factor (k') of the test peak (Peak 2)
: Retention time of Peak 1
N 16
t
R
W
-----
\ .
| |
2
=
N
t
R
W
H
L 1000
N
--------------------- =
H
L
N
S
W
0.05
2 a
0.05
------------------- =
S
W
0.05
a
0.05
R 2
t
R
t
Rp
W Wp +
--------------------- =
R
t
R
t
Rp
W
Wp
k
t
t
0
---- 1 =
t
t
0
a
k
2
k
1
-------
t
2
t
0
t
1
t
0
-------------- = =
a
k
1
k
2
t
1
t
2
t
0
3
JP, EP, BP, and DAB
Item Equation
Plate
: Retention time
HETP
(m)
: HETP
: Width of the front half of the peak (from the start point
to the apex) at 5% height of the peak
: Retention time
: The previous peak width at 50% height of the peak
N 5.54
t
R
W
0.5
-----------
\ .
| |
2
=
N
t
R
W
0.5
H
L 1000
N
--------------------- =
H
L
N
S
W
0.05
2 a
0.05
------------------- =
S
W
0.05
a
0.05
R 1.18
t
R
t
Rp
W
0.5
Wp
0.5
+
-------------------------------- =
R
t
R
t
Rp
W
0.5
Wp
0.5
k
t
t
0
---- 1 =
t
t
0
a
k
2
k
1
-------
t
2
t
0
t
1
t
0
-------------- = =
a
k
1
k
2
t
1
t
2
t
0
3 Equations
JP2
In accordance with revisions authorized by the Japan Pharmacopoeia in April 1996, the number of
theoretical plates is calculated with 5.55 for JP2.
Item Equation
Plate
: Retention time
HETP
(m)
: HETP
: Width of the front half of the peak (from the start point
to the apex) at 5% height of the peak
: Retention time
N 5.55
t
R
W
0.5
-----------
\ .
| |
2
=
N
t
R
W
0.5
H
L 1000
N
--------------------- =
H
L
N
S
W
0.05
2 a
0.05
------------------- =
S
W
0.05
a
0.05
R 1.18
t
R
t
Rp
W
0.5
Wp
0.5
+
-------------------------------- =
R
t
R
t
Rp
W
0.5
Wp
0.5
k
t
t
0
---- 1 =
t
t
0
a
k
2
k
1
-------
t
2
t
0
t
1
t
0
-------------- = =
a
k
1
k
2
t
1
t
2
t
0
3
EMG
Item Equation
Plate
: Retention time
: Width of the back half of the peak (from the apex to the end
point) at 10% height of the peak
HETP
(m)
: HETP
: Retention time
N 41.7
t
R
W
0.1
-----------
\ .
| |
2
b
0.1
a
0.1
-------- 1.25 +
------------------------- =
N
t
R
W
0.1
a
0.1
b
0.1
H
L 1000
N
--------------------- =
H
L
N
S
W
0.05
2 a
0.05
------------------- =
S
a
0.05
W
0.05
R 2.15
t
R
t
Rp
W
0.1
Wp
0.1
+
-------------------------------- =
R
t
R
t
Rp
W
0.1
Wp
0.1
k
t
t
0
---- 1 =
t
t
0
a
k
2
k
1
-------
t
2
t
0
t
1
t
0
-------------- = =
a
k
1
k
2
t
1
t
2
t
0
3 Equations
EMG (50%)
Item Equation
Plate
: Retention time
: Width of the back half of the peak (from the apex to the end
point) at 10% height of the peak
HETP
(m)
: HETP
: Retention time
N 41.7
t
R
W
0.1
-----------
\ .
| |
2
b
0.1
a
0.1
-------- 1.25 +
------------------------- =
N
t
R
W
0.1
a
0.1
b
0.1
H
L 1000
N
--------------------- =
H
L
N
S
W
0.05
2 a
0.05
------------------- =
S
a
0.05
W
0.05
R 1.18
t
R
t
Rp
W
0.5
Wp
0.5
+
-------------------------------- =
R
t
R
t
Rp
W
0.5
Wp
0.5
k
t
t
0
---- 1 =
t
t
0
a
k
2
k
1
-------
t
2
t
0
t
1
t
0
-------------- = =
a
k
1
k
2
t
1
t
2
t
0
3
Area/Height
Item Equation
Plate
: Retention time
: Peak width (where )
@ NOTE
2
, signifying peak shape (smoothness), is a dispersion term for
the probability density function when assuming peak shapes
can be approximated by normal distribution curves. Peak width
is defined as peak top 2 (= 4), which includes almost the
entire peak area (= 95.44%).
By transforming the normal distribution probability density
function formula, can be obtained by the following equations:
: Area, : Height

HETP
(m)
: HETP
Tailing factor : Tailing factor
: Retention time
: Peak width (where )
: The previous peak width
@ NOTE
2
, signifying peak shape (smoothness), is a dispersion term for
the probability density function when assuming peak shapes
can be approximated by normal distribution curves. Peak width
is defined as peak top 2 (= 4), which includes almost the
entire peak area (= 95.44%).
By transforming the normal distribution probability density
function formula, can be obtained by the following equations:
: Area, : Height

N 16
t
R
W
-----
\ .
| |
2
=
N
t
R
W W 4 =
1
2
----------
A
Ht
------
\ .
| |
= A Ht
0.399
A
Ht
------
\ .
| |
H
L 1000
N
--------------------- =
H
L
N
S
W
0.05
2 a
0.05
------------------- =
S
W
0.05
a
0.05
R 2
t
R
t
Rp
W Wp +
--------------------- =
R
t
R
t
Rp
W W 4 =
Wp
1
2
----------
A
Ht
------
\ .
| |
= A Ht
0.399
A
Ht
------
\ .
| |
k
t
t
0
---- 1 =
t
t
0
a
k
2
k
1
-------
t
2
t
0
t
1
t
0
-------------- = =
a
k
1
k
2
t
1
t
2
t
0
3 Equations
User defined
Item Equation
Plate
: Coefficient
: Retention time
: Peak width at (the value set in User Defined)% height of peak
@ NOTE
By setting [Peak Width at] in [Performance] tab page, the coefficient is
automatically defined. When the value set in [Peak Width at] is N, the
coefficient c is calculated by the following equation:
HETP
(m)
: HETP
apex) at (the value set in User Defined)% height of the peak
: Coefficient
: Retention time
: The previous peak width
@ NOTE
By setting [Peak Width at] in [Performance] tab page, the
coefficient is automatically defined. When the value set in [Peak
Width at] is N, the coefficient is calculated by the following
equation:
N c
t
R
W
-----
\ .
| |
2
=
N
c
t
R
W
c 8
100
N
---------
e
log =
H
L 1000
N
--------------------- =
H
L
N
S
W
2 a
----------- =
S
W
a
R c
t
R
t
Rp
W Wp +
--------------------- =
R
c
t
R
t
Rp
W
Wp
c
c 2
100
N
---------
e
log =
k
t
t
0
---- 1 =
t
t
0
a
k
2
k
1
-------
t
2
t
0
t
1
t
0
-------------- = =
a
k
1
k
2
t
1
t
2
t
0
3.4 Background Correction Processing
3
3.4 Background Correction Processing
When start, end, and sampling rate/interval are identical
The background chromatogram is subtracted from the original chromatogram.
When sampling rate/interval differs
The background intensities at the points corresponding to the original chromatogram sampling points are
interpolated from the background sampling points before/next to the interpolated point. The background
chromatogram is then subtracted from the original chromatogram.
Original
Background
After background correction
Original
Background
Interpolated point
3 Equations
When background chromatogram has the same or longer length
The background chromatogram is subtracted from the original chromatogram only within the original
chromatogram range.
When background chromatogram has the shorter length
The starting and ending points of the background chromatogram are extrapolated. The background
chromatogram is then subtracted from the original chromatogram.
Original
Background
Original
Background
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
This chapter describes the peak purity algorithms. Peak purity refers to the index used to evaluate whether
a chromatogram peak, obtained by a photodiode array detector, contains multiple components.
This chapter explains:
-the principles for peak purity analyses,
-parameter setting procedures and actual applications,
-parameter optimization and limitation items (since the applicable range for purity calculation depends on
measurement samples/conditions), and
-validation of the method parameters, allowing more accurate purity calculations.
This chapter consists of the following three sections:
"4.1 Basic Principles of Peak Purity Analysis" P.57
"4.2 Method Development for Peak Purity Analysis" P.63
"4.3 Method Optimization" P.67
4.1 Basic Principles of Peak Purity Analysis
In purity analysis, the software calculates the "similarity" between the spectrum on an objective peak and
reference spectrum. The peak purity is then evaluated by comparing the similarity with a "threshold value"
calculated from noise components. To better understand the purity analysis algorithm, it is first important to
understand the calculation methods for these parameters.
4.1.1 Calculating the Similarity
This section describes the equations used to calculate the similarity - index of how two spectra match.
To understand the similarity, first consider spectra as sets of absorbances at scanning wavelengths. In this
case, absorbance spectra S1 and S2 are expressed by the following vectors:
S1 = (a1 (1) ,a1 (2) , ,a1 (n))
S2 = (a2 (1) ,a2 (2) , ,a2 (n))
where a (1) signifies the absorbance at wavelength (1).
When focusing on two wavelength data, these two vectors can be graphically represented as shown in
Fig.4-1.
Fig.4-1 Spectra Represented as Vectors
If two spectrum shapes are identical, vectors S1 and S2 should point to the same direction, and the angle
in Fig.4-1 becomes "0", even though absorbances are different. Since the smaller the angle , the greater
the similarity between two spectra, the similarity of two spectra (SI) can be obtained by calculating cos.
The equation will be as follows:
or
The closer the similarity (SI) to 1, the more closely two spectra match.
4.1.2 Calculating the Threshold
Generally, when two spectra are obtained from a single component peak, the shapes of those spectra
should be the same within the peak. In actual measurement, however, the angle does not necessarily
become "0", because of the effect of the background noise of the detector, noise from mobile phase
absorption, etc. Fig.4-2 illustrates the case.
Fig.4-2 Noise Components
In Fig.4-2, circles with radius indicate the uncertainty due to noise components of size .
This means that it is possible that the similarity between two peaks is reduced by the amount
due to the noise.
However, if the angle between vectors and exceed ( ), there may exist some factor other than
noise components, which makes the two spectrum shapes different.
The value here is called "threshold ".
The software compares the threshold with obtained similarity, to evaluate whether or not two spectrum
shapes match.
Threshold is calculated by the following equation:
where
Assuming the noise spectrum intensity at wavelength is , radius N is obtained as follows:
SI
S
1
S
2
S
1
S
2
----------------- cos = =
SI
a
1

i
( )
a
2

i
( )
a
1

i
( )
2
a
2

i
( )
2
----------------------------------------------------------------- =
N N
1 2 + ( ) cos
S1 S2 1 2 +
1 2 + ( ) cos t ( )
t ( )
t ( )
t 1 2 + ( ) cos 1
N
2
S
1
2
-----------
\ .
|
| |
1
N
2
S
2
2
-----------
\ .
|
| |
N
2
S
1
S
2
----------------- = =
S
1
a
1

i
( )
2
=
S
2
a
2

i
( )
2
i
no
i
( )
N no
i
( )
2
=
4
4.1.3 Calculating the Peak Purity
This section describes how this software calculates peak purity based on similarity and threshold values,
and how it displays the evaluation results.
@ NOTE
The presence of impurities is evaluated using similarity (SI), threshold (t), and purity index, as follows:
This software evaluates peak purity using the following three methods: 3-point peak purity method, N-point
peak purity method, and total peak purity method. Note that peaks obtained with a photodiode array
detector have spectrum values at each sampling point.
3-point peak purity method refers to spectrum values at three sampling points for evaluation.
N-point peak purity method is an extended 3-point peak purity method, which can refer to 5, 7, and 9
sampling points.
Total peak purity method refers to spectrum values at all sampling points.
Even though these methods use different sampling point spectra, they all evaluate purity using similarity
and threshold values, calculated by the equation in "4.1.1 Calculating the Similarity" and "4.1.2 Calculating
the Threshold", respectively.
3-point peak purity method
This method uses spectra at three sampling points on the peak. Fig.4-3 shows the positions of the three
sampling points.
Fig.4-3 Sampling Points for 3-point Peak Purity Method
In this method, the objective peak is divided at the peak top for purity evaluation.
For the front half of the peak, the purity is evaluated at the middle of the peak start and peak top (Upslope
point). For the rear half of the peak, considering the tailing characteristics of chromatogram peaks, the
purity is evaluated at 1/3 between the peak top and peak end (Downslope point). In either case, the
reference spectrum is peak top spectrum.
This software displays similarity, threshold, and purity index (difference between similarity and threshold)
for each of the upslope point, downslope point, and 3-points (average of the front and rear halves).
SI t Purity index 0 No impurity is detected
SI < t Purity index < 0 Impurity is detected
Peak top
Upslope point
Downslope point
Peak end
Peak start
1/2 1/3
N-point peak purity method
This purity evaluation method uses spectra at the sampling points shown below. Since spectra near peak
shoulders are also calculated, purity can be evaluated based on a broader region than that used in the 3-
point peak purity method. When peak start and end are not properly determined by peak integration,
however, purity calculation may be impossible due to the lack of spectrum intensities at those sampling
points. Therefore, it is required to more accurately set peak integration parameters and manual peak
integration.
Fig.4-4 Sampling Points for N-point Peak Purity Method
Similarly to the 3-point peak purity method, the software calculates the similarity between spectrum at each
sampling point and peak top spectrum, and evaluates the purity for the front and rear halves of the peak.
Total peak purity method
This method uses spectra at all sampling points on the peak for purity calculations. Even though the
calculation process takes longer than the 3-point/N-point peak purity methods, it allows for more accurate
purity evaluations.
Fig.4-5 shows the position of sampling points used for total peak purity method.
Fig.4-5 Sampling Points for Total Peak Purity Method
In this method, spectrum at each of sampling points (S1, S2, Sn) is compared with spectrum at the
reference point with higher absorbance (Sref), located at a slightly inner area towards the peak center.
Each sampling point on the peak has a corresponding reference point. Even though the position of the
reference point varies with absorbance level, it is set to give a difference of several mAU to several tens of
mAU from the sampling point, based on the maximum absorbance of the spectrum (Max. Plot designated
wavelength range). As shown in Fig.4-5, the reference spectrum for spectrum S1 (S
ref
) is obtained from the
adjacent spectra S
i
and S
j
.
N-point Sampling Point Used
5 3/8 4/5
7 Peak start 4/5 3/8 Peak end
9 9/8 7/8 5/8 3/8
Several mAU to
several tens of mAU
Time
Abs
4
Compared to the previous methods using the peak top spectrum (S
top
) as a reference for all sampling
points, this method can evaluate the purity by eliminating the effect of spectrum shape difference due to
sample concentration variations.
This software calculates the purity at each sampling point using the total peak purity method, and displays
the result with retention time along the graph horizontal axis. The user can refer to two types of graphs: a
similarity curve, which plots the similarity and threshold value, and a purity curve, which plots the purity
index (= similarity-threshold). Fig.4-6 and Fig.4-7 show examples of these curves.
Fig.4-6 Similarity Curve
Fig.4-7 Purity Curve
Two graphs can be switched by setting in [Purity View (Peak Purity) Display Settings] sub-window.
The software searches the sampling point at which the purity index is the minimum, and displays the
similarity, threshold, and purity index obtained at the point. These values are called "Peak Purity Index",
"Single Point Threshold", and "Min Peak Purity Index", respectively. "Min Peak Purity Index" is the
threshold subtracted from the similarity, multiplied a million times for convenience of reference.
When "Min Peak Purity Index" is a negative value, the retention time at the sampling point is also
displayed. Note that this retention time does not indicate the retention time of impurity peak, but the point
where the two spectrum shapes differ the most.
Further, although the positive/negative of "Min Peak Purity Index" indicates whether two spectrum shapes
match, the absolute value of the index does not indicate impurity amount.
Each reference point is defined as the point with absorbance higher (several mAU to several tens of mAU)
than that of the corresponding sampling point, located at a slightly inner area towards the peak center,
based on the max plot of the peak. (The wavelength range for max plot is defined by [From] and [To] of
Similarity curve
Threshold curve
Max plot
Max plot
Purity curve
Zero assistant line
[Purity Calculations], set in the [Purity] tab page [see "4.2 Method Development for Peak Purity Analysis"]
in the [Method View] of the [PDA Data Analysis] window.) The software displays the max plot of the peak
overlapping on similarity and purity curves. In this case, the time axis is defined by the range between the
peak detection start and end on the relevant multi-chromatogram.
4.1.4 Example of Peak Purity Analysis Result
This section shows an example of the result using the 5-point peak purity method.
The figure below illustrates the result of the following peak evaluation:
"Upslope Purity Index" at the graph bottom indicates the mean value of the results from two sampling point
spectra on the front half of the peak. "Downslope Purity Index" indicates the same value for the rear half of
the peak.
@ NOTE
If calculation is impossible at a sampling point due to the lack of spectrum intensity, the sampling point is
disregarded from the result. If calculation is impossible at both sampling points, "Cannot be calculated" is
displayed.
"5-point Purity Index" outputs the mean value of the upslope and downslope indexes. If purity calculation
is impossible for either front or rear half, "Cannot be calculated" is displayed.
If there is a point where spectrum similarity falls under the threshold, the retention time is displayed as
"Impurity: Detected at".
The graph displays the spectrum extracted at each sampling point. The label indicates the sampling point
at which each spectrum is obtained (e.g. peak top, upslope, and downslope), as well as its retention time.
@ NOTE
When the spectral similarity between peak top and any sampling point falls under the threshold, "Impurity:
Detected at" is displayed.
When purity calculation is impossible, "Cannot be calculated" is displayed.
The sampling points at the front half of the peak are: Peak top (Peak top Peak start) 4/5 = 2.49 min
Peak top (Peak top Peak start) 3/8 = 2.54 min
The sampling points at the rear half of the peak are: Peak top + (Peak end Peak top) 3/8 = 2.63 min
Peak top + (Peak end Peak top) 4/5 = 2.68 min
By adding the peak top, a total of 5 points are used for purity calculations.
Peak top: 2.58 min
Peak start: 2.464 min
Peak end: 2.709 min
4.2 Method Development for Peak Purity Analysis
4
This section describes the development of method for the purity analysis of measured chromatogram
peaks. To develop methods, set the purity calculation parameters according to actual measurement
conditions in the [PDA Data Analysis] window - [Method View] - [Purity] tab page. Setting appropriate
parameters allows for more accurate purity evaluation.
^ Reference
For actual procedures for setting method development parameters, refer to "4.3.1 Method Development
Procedure" P.67.
4.2.1 Specifying the Multi-Chromatogram Wavelength
Data obtained with a photodiode array detector are saved as three-dimensional data of wavelength,
retention time, and intensity. By extracting data at a specific wavelength from the 3-D data, a
chromatogram is obtained, and the extraction wavelength can be saved in the method file. This type of
chromatogram is called a "multi-chromatogram".
To conduct purity analysis for a peak, its peak start, peak end, and retention time are required. Therefore, it
is first necessary to isolate the objective peak by extracting a multi-chromatogram and conducting peak
integration on the chromatogram. For multi-chromatogram extraction wavelength, select the wavelength
allowing the objective peak (for purity analysis) to exhibit sufficiently high absorbance.
4.2.2 Calculating the Noise Spectrum
The software calculates noise spectra based on the time data specified in [Retention Time Range] on the
[Purity] tab page.
Noise spectrum intensity at a wavelength [ ] is calculated as three times the standard deviation
of baseline fluctuations on multi-chromatogram at within the set [Retention Time Range]. The equation
is as follows:
The noise spectrum can be obtained by executing the calculation above for each wavelength.
@ NOTE
It is impossible to evaluate peak purity when the intensity of noise spectrum is greater than that of the
objective peak spectrum. Note that in the following cases, calculated noise spectrum intensity may differ
from the noise value contained in the actual objective peak:
The mobile phase composition differs between the period of elution of the objective peak (for
purity analysis) and the retention time range used for noise spectrum calculation.
Contaminations are eluted within the retention time range used for noise spectrum calculation.
Background noise from hardware greatly differs between noise calculation data and objective
peak data, due to lamp replacement, use of different equipment, etc.
To avoid these problems, it is necessary to set the noise spectrum calculation parameters according to
the following:
In [Retention Time Range], always set the range including the objective peak for purity analysis.
Otherwise, analyze a blank, calculate the noise spectrum using the time near the objective peak
on the blank data, and save the noise spectrum as a method.
Plot the max plot of the data used for noise spectrum calculation, and specify the retention time
range not including contaminations.
Select the [Compute noise spectrum from current data for peak purity] checkbox to always
calculate noise spectrum using objective peak data.
In this case, however, obtained noise spectra will not be saved to the method, and noise spectra
in the method becomes unavailable. For each purity calculation, noise spectrum is calculated
from objective peak data using the set retention time range, allowing for more accurate noise
spectrum calculation. However, if a contamination peak appears within the set retention time
range, an extremely large noise level is detected, preventing an accurate purity evaluation.
4.2.3 Defining the Wavelength Range
Peak purity is calculated within the wavelength range defined by [From] and [To] in [Purity Calculations].
Note that correct purity calculation may be obstructed when the target substance has no UV absorbance in
the set range, or when a low-wavelength range is set in which the mobile phase absorbs UV.
where
n: The number of sampling points within the set retention time range
a(t
j
): Absorbance at time t
j
i
no
i
( ) SD ( )
i
no
i
( ) 3 SD =
SD
n a t
j
( )
2
a t
j
( )
\ .
| |
2
n n 1 ( )
--------------------------------------------------------- =
4
4.2.4 Specifying the Compensation Coefficient
Leave this value as "0" normally. However, if a purity evaluation result differs from expected values, the
user can adjust the evaluation threshold by specifying a numerical value to this parameter. Then the
evaluation result will be output after subtracting the set value from the min peak purity index. For example,
consider the case of a measured peak that obviously contains no impurity or contains purities within the
permitted limit. If this peak is evaluated to contain impurities (i.e., the min peak purity index indicates a
negative value) due to, for example, spectrum distortion from indeterminate noise effect or from measuring
a concentrated sample, the user can change the evaluation result to "no impurity" by specifying
[Compensation Coefficient]. (For example, enter the compensation coefficient as -1000, when the min
peak purity index is -1000.)
Conversely, consider that a purity curve is obtained as shown in Fig.4-8, which clearly shows that there is
an impurity near 4.09. If this peak is evaluated to contain no impurity, the user can modify the evaluation by
specifying the corresponding positive value in [Compensation Coefficient].
Fig.4-8 Example of Purity Curve Result Output
@ NOTE
As described above, the purity evaluation can be intentionally modified by specifying [Compensation
Coefficient]. Therefore, always make sure that the value cannot be changed illegally by taking necessary
measures (e.g. printing [Compensation Coefficient] in the purity evaluation report, enabling the audit trail
function, etc.). Additionally, the evaluation result can also be changed when changing the noise spectrum
used for purity evaluation. It is therefore also recommended to always confirm that the noise spectrum
calculation parameters have not been changed illegally.
4.2.5 Spectrum Background Compensation
If the [Background Compensation] checkbox is ON during purity calculation, the background compensation
is performed on the spectrum used for purity evaluation. This setting reduces the effect of baseline drift,
allowing for more accurate purity evaluation. The background compensation is executed by the following
equation:
: Spectrum after background compensation
: Spectrum before background compensation
: Background spectrum
S
corr
S S
b
=
S
corr
S
S
b
The absorbance at wavelength
i
after background compensation is calculated by the following equation:
is obtained by the following equation:
@ NOTE
For 7 or 9 point peak purity method, the software does not execute background compensation even though
[Background Compensation] is set to ON, since the peak start/end and sampling points outside the peak are
used for calculation.
can be obtained by calculating this across the entire wavelength range.
To improve purity calculation accuracy, it is recommended to always select the [Background
Compensation] checkbox, even though it might take slightly more time for the calculation process.
4.2.6 Setting the [Compute Purity] Options
The purity analysis involves various calculations using three-dimensional data. If multi-chromatograms are
specified at multiple channels, for example, and if peaks are detected at each chromatogram, the software
response becomes slower during peak integration. This is because each time the peak integration is
executed, the purity of an extremely large number of peaks is calculated.
To avoid this, set the [Compute Purity] options described below. Select an appropriate parameter in the
[PDA Data Analysis] window - [Method View] - [Purity] tab page - [Compute Purity].
Note that regardless of this option setting, the similarity curves and purity curves indicate the result of purity
calculation for selected peaks.
: Absorbance after background compensation
: Absorbance before background compensation
: Background absorbance
: Absorbance at peak start
: Absorbance at peak end
: Retention time for the background-compensated spectrum
: Retention time of peak start
: Retention time of peak end
a
corr

i
( ) a
i
( ) a
b

i
( ) =
a
corr

i
( )
a
i
( )
a
b

i
( )
a
b

i
( )
a
b

i
( )
a
s

i
( ) t
e
t ( ) a
e

i
( ) t t
s
( ) +
t
e
t
s
--------------------------------------------------------------------- =
a
s

i
( )
a
e

i
( )
t
t
s
t
e
S
corr
Not Calculated Select this parameter when purity analysis is not necessary. The processing speed does not
become slow since the software does not calculate the purity at each peak integration. In this
case, the purity analysis result is not displayed on the peak table or peak top comments on
chromatograms.
Identified Peaks Select this parameter when the objective peaks for purity analysis are already determined among
all detected peaks. The result of purity analysis is only displayed on the peaks identified by the
compound table (ID table) parameters.
All Peaks The purity of all detected peaks is calculated.
4.3 Method Optimization
4
Due to characteristics of absorption detectors (such as the photodiode array detector), the detector
response to the sample concentration may become non-linear for higher-concentration samples. Since the
purity analysis calculates the similarity between two spectra, this effect becomes significant even in the
concentration region where the linearity of the quantitative calibration curve is not affected. In such
concentration regions, a peak without impurity may be wrongly evaluated to contain impurities, since the
two spectra are judged to be different due to the spectrum distortion.
This effect becomes even more significant in purity calculation where the peak top spectrum is used as a
reference. This is because when using a high concentration sample, the spectrum distortion becomes
greater and the similarity decreases more from peak shoulders towards peak top. Even so, using a unique
method such as "Total peak purity method", this software calculates the purity of the peaks with up to an
absorbance of 1 AU, without specifying any compensation parameter.
However, if purity calculation parameters are set incorrectly, purity analysis may be impossible even when
the peak maximum absorbance does not exceed 1 AU.
To avoid this, follow the procedure described below to develop the optimum method, and to verify the
developed method.
4.3.1 Method Development Procedure
Follow the procedure described below to develop the optimum method.
1
Specify the multi-chromatogram extraction wavelength.
In the multi-chromatogram table, specify the wavelength to extract multi-chromatogram by referring to the
peak top spectrum of the objective peak or taking other measures. Be sure to specify the wavelength that
obtains the highest absorbance for the peak.
^ Reference
When the objective peak has not been completely integrated, set the peak integration parameters
appropriately. For details on how to set the peak integration parameters, refer to "the Operator's Guide "6.2
Setting the Peak Integration Parameters"".
For details on how to set multi-chromatogram extraction wavelengths, refer to "4.2.1 Specifying the Multi-
Chromatogram Wavelength" P.63.
2
Set the parameters on the [Purity] tab page in the [Method View] of the [PDA Data
Analysis] window.
1 Specify the retention time range for noise spectrum calculation.
Click [Open Data File] to open the objective data for purity analysis, or data file obtained in the same
conditions as those for the objective data.
The max plot of the opened data is displayed on the upper graph window. The chromatogram can be
enlarged by dragging the mouse. Verify the retention time range where no peak is observed and the
baseline is flat. Specify the range as that for the noise spectrum calculation. For details, refer to
"4.2.2 Calculating the Noise Spectrum".
@ NOTE
PDA data files do not contain any signals from 0 to 0.2 min. Be sure to NOT specify the value
smaller than 0.2 min. The default setting is from 0.2 min to 1 min.
2 Click [Compute noise spectrum].
The calculated noise spectrum is displayed on the lower graph window. Ensure that the maximum
absorbance of the noise spectrum does not exceed that of the objective peak for purity analysis. If
the maximum absorbance of the noise spectrum is larger, the software cannot analyze the peak
purity using the noise spectrum. In this case, take the following measures:
Redefine the retention time range.
Restrict the wavelength range to perform purity calculation only within the range where the noise
spectrum absorbance becomes smaller.
Increase the sample concentration.
@ NOTE
Three separated retention time ranges can be set. Only the column at the top is valid as
default. To analyze the purity of multiple peaks obtained in gradient analysis, the user can set
multiple areas near the peaks for noise spectrum calculations.
When a method is saved, the current noise spectrum is saved in the method (unless
[Compute noise spectrum from current data for peak purity] is selected). The noise spectrum
is then used for the subsequent purity analyses for the data obtained by the method, and it is
no longer necessary to calculate noise spectrum for each purity analysis. Conversely, to
always calculate noise spectrum using the objective data for purity analysis, select the
[Compute noise spectrum from current data for peak purity] check box. Even so, it is
recommended NOT to select the check box normally, since the purity analysis results can be
more easily compared when obtained in the same conditions.
3 Specify the wavelength range in [From] and [To] on [Purity Calculations].
Define the wavelength range to perform purity calculations. By referring to the spectra contained in
the objective peak for purity analysis, specify the wavelength range in which the peak has sufficient
absorbance. In this case, it is recommended NOT to include a wavelength region shorter than
210 nm, since the wavelength in the region often detects mobile phase absorption.
4 Specify [Step].
Leave the default value as "1" normally. If "2" or greater is set, the uncertainty of purity evaluation
may increase, even though the processing speed increases due to the lesser number of calculation
points.
5 Specify [Compensation Coefficient].
Leave the default value "0" as initial setting. According to the obtained purity analysis result, change
the value as appropriate. For details, refer to "4.2.4 Specifying the Compensation Coefficient".
6 Check/uncheck the [Background Compensation] checkbox.
It is recommended to always check this checkbox. Remove the check only when the processing
speed is extremely slow, and when it is certain that its removal does not affect the purity analysis
result. For details, refer to "4.2.5 Spectrum Background Compensation".
7 Specify [Compute Purity].
The default setting is [Not Calculated]. Select [Identified Peaks] when the objective peak for purity
analysis has already been identified by the compound table parameters. In other cases, select [All
Peaks]. For details, refer to "4.2.6 Setting the [Compute Purity] Options".
4
4.3.2 Validating the Method
This section describes procedures for validating the appropriateness of the method created by "4.3.1
Method Development Procedure".
1
Prepare a standard sample to be evaluated to contain no impurity.
Prepare the sample concentration to exhibit similar absorbance to that of the objective peak for purity
analysis. When the objective peak absorbance is not constant, adjust the sample concentration exhibiting
the highest expected absorbance.
2
Analyze the sample with the created method.
To verify the method more accurately, repeat this analysis five times.
3
Display the obtained result of the purity analysis in the [Purity View].
Double-click the objective peak on the chromatogram in the [PDA Data Analysis] window to display the
analysis result in the [Purity View]. Verify the Min Peak Purity Index obtained.
The Min Peak Purity Index can also be displayed in peak top comments, peak table, and compound result
table, unless [Purity Index Mode] is set to [None] in the [Purity View(Peak Purity) Display Settings] sub-
window.
4.3.3 Interpreting the Method Validation Result
When the Min Peak Purity Index indicates 0 or a positive value for all analyzed
data
Use the method for the subsequent purity analysis, without change.
When the Min Peak Purity Index indicates a negative value for any of the
analyzed data
Peaks analyzed by the method may be wrongly evaluated to contain impurities, because of spectrum
distortion due to high concentration, effect of background noise, effect of chemical purity of standard
samples, etc. Set the min peak purity index with the largest absolute value obtained in the repeated
analyses to the [Compensation Coefficient] for the method.
When calculation is impossible
If "Cannot be calculated" is displayed, it shows that the purity index calculation is impossible since the
maximum absorbance of the objective peak is smaller than the noise spectrum absorbance.
In this case, an accurate purity evaluation is obstructed because the peak obtained at the sample
concentration is overwhelmed by the noise. To solve this problem, increase the sample concentration. As
another possibility, the method parameters may not be appropriate. Review the method parameters.
Specifically, check the noise spectrum calculation parameters and the wavelength range for purity
calculations. For the former, recalculate noise spectrum to decrease noise level. For the latter, narrow the
wavelength range, setting only the range where the noise spectrum absorbance is smaller than the
objective peak spectrum absorbance.
4.3.4 Application Examples
Applicable concentration range
When analyzing the purity of high concentration samples, it is required to set a compensation coefficient.
This is because (as explained in "4.3 Method Optimization"), the detector response to sample
concentration becomes non-linear in higher concentration regions. However, the correct purity evaluation
may be obstructed if the compensation coefficient is set inappropriately.
The following example shows a purity analysis using high concentration samples without setting a
compensation coefficient:
Fig.4-9 A Purity Curve of Ethyl Paraben
Fig.4-9 shows the purity curve of the peak obtained by analyzing Ethyl Paraben. From the result, it is
observed that even the peak at which maximum absorbance exceeds 1 AU is not affected by spectrum
distortion.
Fig.4-10 shows the result of analyzing the same sample after mixing 0.1% impurity. It can be observed that
an existence of impurity was detected near 4.47 min.
Fig.4-10 A Purity Curve of Ethyl Paraben Containing 0.1% p-Ethyl Aminobenzoate Ester
As shown in the examples above, the purity can be evaluated for a sample with absorbance up to
approximately 1 AU, even without setting any compensation coefficient.
4
Detection limit for impurity content
Whether or not an impurity can be detected by purity analysis depends on the content of the impurity
relative to the major component amount. However, the detection limit for the impurity content varies
according to the following factors:
1. Spectral similarity between major component and impurity
The purity analysis is impossible between two components having identical spectral shapes. Even if not
identical, when both spectral shapes have similar patterns, the spectrum difference may fall under the
threshold. To detect the small differences in such cases, a sufficient absorbance level is required. The
detection limit therefore varies according to the similarity between the two components.
2. Retention time difference between major component and impurity
Purity analysis is impossible between two components with identical retention time, since the spectrum
shapes become identical. The detection limit therefore varies according to the retention time difference
between the two components.
3. Absolute impurity content
Even though the spectrum shapes of two components greatly differ and the retention time difference is
sufficiently large, purity analysis is impossible unless the absolute impurity content reaches the detection
limit. For example, consider analyzing the purity of a peak containing 0.1% impurity. When the absorbance
of the major component peak is 1 AU, the impurity absorbance is 1 mAU, which can be detected without
any problem. Meanwhile, when the major component absorbance is 1 mAU, the impurity absorbance
becomes 1 nAU, which may be difficult to detect.
The detection limit of impurity content thus greatly varies according to analysis conditions and measured
sample type. Fig.4-11 shows an example of the purity curve for ethyl paraben containing 0.05% p-ethyl
aminobenzoate ester.
Fig.4-11 A Purity Curve of Ethyl Paraben Containing 0.05% p-Ethyl Aminobenzoate Ester
It was observed from the example that even 0.05% impurity was detected. Next, Fig.4-12 and Fig.4-13
show the UV spectra of p-ethyl aminobenzoate ester and ethyl paraben standard solution, respectively.
Fig.4-12 UV Spectrum of p-Ethyl Aminobenzoate Ester
Fig.4-13 UV Spectrum of Ethyl Paraben
The similarity between these spectra became 0.688. The separation factor was 1.053, obtained by
analyzing the standard solutions in the same analysis conditions.
Under such conditions, the presence of impurities can be sufficiently detected even for impurity with 0.05%
concentration. However, the detection limit is expected to become higher when two component spectra are
more similar, or the retention times closer, than is the case in this example.
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5 File Structure
This software manages chromatograms and method data using specific file formats.
Data files employ the All-In-One structure, which allows the user to refer to chromatograms, sample
information, data acquisition conditions, computing parameters, batch tables used for data
acquisition, result reports, etc. by operating a single data file. Data files are also designed to maintain
data traceability.
This section describes the file structure of the method files and data files.
5.1 Method File Structure
Method files in this software store the following data.
The file extensions are ".lcm".
@ NOTE
When using a photodiode array (PDA) detector, the following data processing parameters are also saved
in the method file: [Multi Chromatogram Table], [UV Spectrum Parameters], [Library Search Parameters],
and [Purity Parameters].
When acquiring data in multiple channels, chromatograms and compound tables are saved for each
channel and detector.
Analytical instrument information
System configuration files (system configuration data)
Instrument parameters
Acquisition settings on each analytical instrument
Baseline check parameters
Data analysis information
Data processing parameters
Peak integration parameters
Integration time program
Noise/drift settings
Identification parameters
Quantitative parameters
Compound table
Grouping table
Calibration curve data (compound)
Calibration curve data (group)
Column performance parameters
Custom calculation parameters
System suitability settings
QA/QC parameters
QC check parameters
Screen display information
X/Y-axis scale settings
Audit trail log
5 File Structure
System configuration files (system configuration data)
System configuration files store the link information between analytical instruments and the computer,
system instrument names, consumable data, etc. By double-clicking the icon for an analytical instrument
from the [LabSolutions Main] window, the software loads the system configuration file corresponding to the
instrument, and connects the instrument to the computer according to the file information. The [Acquisition]
program then starts up.
By dragging & dropping a method file onto the [Instrument Parameter View], the current system
configuration file is copied to the method file to make the current system information identical with that
stored in the method file. In this case, the parameters on newly added instruments are set to the default
values.
Data acquisition parameters for each instrument (= instrument parameters) can be set in the [Instrument
Parameter View] in the [Data Acquisition] window. This software transfers the instrument parameters set in
the method file to the relevant analytical instruments for each data acquisition.
The user can set data processing parameters for peak integration, quantitative processing, compound
table, etc. in the [Method View] in the Data Analysis window. When acquiring data in multiple channels or
detectors, data processing parameters are set for each channel.
The user can set parameters for validating repeatability of area values, retention times, etc.
Specify items to be output, evaluation criteria, and so on in method files, and complete the system
suitability setting in the batch table row for checking repeatability.
QC check parameters
In the [Method View] of the [Quant Browser] window, the user can specify the criteria for checking
"Accuracy [%]" and "%Deviation" between standard concentration set in the compound table and that
obtained from the calibration curve.
The evaluation results can be verified in the [Quantitative Results View] of the [Quant Browser] window or
on reports.
Criteria for "Accuracy[%]" and "%Deviation" can be set to samples whose type is other than standards, and
the quantitative result for each sample can be validated.
@ NOTE
QC check parameters can also be set in the Data Analysis window. In this case, evaluation results are
displayed in the compound result table (= [Compound] tab page in the [Results View]). Modify the table
style setting as necessary.
When creating a calibration curve based on multiple calibration points, each standard data file stores the
concentration value calculated from the current calibration curve. In this case, the concentration values
differ between those obtained when the standard data file is opened from the Data Analysis window and
when it is opened from the [Quant Browser] window. This is because the [Quant Browser] window
indicates the concentration values calculated from the calibration curve based on all calibration points.
5.2 Data File Structure
5
5.2 Data File Structure
Data files of this software employ the All-In-One structure, which stores the following data onto a single file:
instrument parameters for data acquisition, data processing parameters for analyzing obtained
chromatograms, batch tables, report formats, etc.
The file extensions are ".lcd".
Data files in this software store the following data:
Chromatograms
Analysis results
Sample information
System configuration data
Method/batch table/report format data at data acquisition
Method/batch table/report format data at postrun analysis
Report format
The previous system check results
Acquisition logs
Audit trail logs
Data files contain two types of data: (1) unalterable data such as acquired chromatograms and analysis
conditions, and (2) alterable data such as data processing parameters to optimize peak integration
process.
When an alterable data is modified, the change history is recorded if the audit trail function is ON.
Unalterable data
Chromatograms
Sample information (vial No., analysis date, and full path of the file used at data acquisition)
System configuration data
Method data at data acquisition
Batch table data at data acquisition
The previous system check results
Acquisition logs
Audit trail logs
Alterable data
Chromatograms used for background compensation
Sample information (sample name, sample ID, sample amount, dilution factor, data comment, sample
type, level No., and full path of the file used at postrun analysis)
Analysis results
Custom calculation parameters
QA/QC parameters
QC check parameters
Batch table data for postrun analysis
Analysis result report information
Screen display information
5 File Structure
Chromatograms
Signal data (chromatograms) acquired from the detector cannot be altered.
Blank chromatogram (background data) is stored in a separated data file for baseline correction. The
background data can be altered in post-run analysis.
Sample information
Data files store the sample information entered during single and batch analyses, analysis date, analyst
name, etc.
Analysis results
Data files store the analysis results calculated by the latest data processing parameters.
Method data
Separately from method files (.lcm), data files store method data created at data acquisition, and the latest
method data used for post-run analysis.
The method data has the following characteristics:
Even though changing a parameter in the method file created at data acquisition (.lcm), the change is not
reflected in the method data in data files.
Even though a change is made in a data processing parameter in the data file displayed in the Data Analysis
window, the change is not reflected in the method file (.lcm) used for data acquisition.
Even though a method file is lost, the method file can be created from the method data stored in data files.
The method data at data acquisition, which is stored in data files can be resumed.
Batch table data
Separately from batch files (.lcb), data files store batch table data created at data acquisition and used for
post-run analysis. Another batch file can be created from the batch table data stored in data files.
@ NOTE
When creating a data file in single analysis, the batch table data is not included.
Analysis result report information
Navigate from the Data Analysis window's [File] menu to [Data Report], and click [Print]. Data is printed as
a report in the latest report format in the data file. To print the data in the report format at data acquisition,
navigate from the [File] menu to [Data Report], and click [Initialize Format].
System check results
The latest system check result at data acquisition is copied onto the data file.
The system check results stored in the data file can be verified at the [System Check] report item.
Acquisition logs
Errors occurring during data acquisition and direct operation of analytical instruments are automatically
recorded as log data. To verify log data obtained during analysis, select [Acquisition Log] from the Data
Analysis window's [View] menu to display the [Logs] sub-window.
Symbols
Accuracy ..............................................................30
%Deviation...........................................................30
(Absolute Calibration Curve) ................................24
Numerics
3-point peak purity method...................................59
A
Absolute retention time method ...........................18
Adjacent peak identification .................................21
Area Normalization...............................................24
ASTM...................................................................43
B
Background Correction Processing......................55
Band method........................................................18
C
Calculating the Noise Spectrum...........................64
Calculating the Peak Purity..................................59
Calculating the Similarity......................................57
Calculating the Threshold ....................................58
Calibration Curve .................................................31
Calibration Curve Correction................................39
Calibration Curves Using Standard Concentration
Factors .................................................................39
Column Performance ...........................................48
Conc. summation .................................................23
Corrected Area Normalization..............................24
Corrected Area Normalization with Scale Factor .24
Correction Factor .................................................34
D
Data File Structure...............................................75
Defining the Wavelength Range ..........................64
Detecting peaks .................................................... 3
Detection Limit Coefficients................................. 44
Detection limit for impurity content ...................... 71
Detection sensitivity .............................................. 3
Determining peak start/end................................... 6
Drift.................................................................. 4, 11
Drift Settings........................................................ 44
E
EP ....................................................................... 43
Exponential ......................................................... 33
External Standard ............................................... 24
F
File Structure....................................................... 73
G
Group calibration................................................. 23
Grouping ............................................................. 23
H
Half-height width ................................................... 3
I
Identification Parameters .................................... 17
Identified peak selection...................................... 21
Internal Standard................................................. 24
Ion chromatograph.............................................. 39
ISTD peak ........................................................... 20
L
Least square method .......................................... 36
M
Method File Structure.......................................... 73
Index
Index
Min Area/Height .................................................... 5
Multi-chromatogram............................................ 63
N
Noise Calculation Methods ................................. 43
P
Peak Detection Parameter Change Time ........... 15
Peak Integration Flow ........................................... 3
Peak Integration Parameters ................................ 2
Peak Purity Algorithms........................................ 57
Peak Sensitivity..................................................... 6
Q
QA/QC Parameters............................................. 45
Quantitation Using Other Component Calibration
Curve................................................................... 34
Quantitative Limit Coefficients............................. 44
Quantitative Methods .......................................... 24
R
Reference peak................................................... 20
Reference Standard ID ....................................... 34
Relative retention time method ........................... 19
rms ...................................................................... 43
Rounding............................................................. 29
S
Significant Digits.................................................. 29
Slope................................................................. 3, 6
Slope Test ............................................................. 7
Slope to Define Baseline..................................... 11
Slope to define baseline........................................ 4
Specifying the Compensation Coefficient ........... 65
Spectrum Background Compensation ................ 65
Standard Addition................................................ 24
T
T.DBL.............................................................. 3, 15
Tailing.................................................................. 13
The Minimum Half-height Width of Peaks............. 5
Time to change peak detection parameters.......... 3
Total peak purity method..................................... 60
U
Unresolved peaks................................................ 13
V
Vertical division ................................................... 13
W
Wavelength setting.............................................. 63
Weighted least square method............................ 37
which defines peak start/end................................. 3
Width ................................................................. 3, 5
Window method................................................... 17

Data Acquisition&Processing Theory Guide

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Data Acquisition&Processing Theory Guide

Uploaded by

Copyright:

Available Formats

Read the instruction manual thoroughly before you use the product.

Keep this instruction manual for future reference.

Operating System is referred

XP Professional Edition is referred

You might also like