Scribd
Upload
52 views29 pages

Database Search Procedures and Electronic Data Manipulation

Spectral database searching is a powerful tool for identification but requires careful use to obtain meaningful results. The search should use a library with similar data types to the unknown and always examine the potential matches instead of assuming the first result is correct. Data manipulation can alter the original data and produce misleading results, so it is best to collect high quality initial data and only manipulate when necessary, retaining the original for comparison.

Uploaded by

maopacific
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
52 views29 pages

Database Search Procedures and Electronic Data Manipulation

Spectral database searching is a powerful tool for identification but requires careful use to obtain meaningful results. The search should use a library with similar data types to the unknown and always examine the potential matches instead of assuming the first result is correct. Data manipulation can alter the original data and produce misleading results, so it is best to collect high quality initial data and only manipulate when necessary, retaining the original for comparison.

Uploaded by

maopacific
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 29

Database Search Procedures and

Electronic Data Manipulation


When and Why
Spectral Database (Library) Searching

• Library Search procedures can be a


powerful tool for identification of infrared
spectra.
• Improperly used, library search results can
be very misleading.
• Library Searching should be approached
carefully to obtain meaningful results.
Spectral Database (Library) Searching

• Library Search procedures should be performed


whenever possible with data similar to the
unknown, i.e. ATR spectra vs. an ATR database,
transmission spectra vs. a transmission database.
• NEVER assume that the first spectrum listed in
the results is the correct match.
• ALWAYS examine the spectra in the search results
versus the unknown.
• If the unknown is suspected to be a proprietary
compound, most probably it WON’T be in the
Search library database.
Spectral Database (Library) Searching

• Most Search software provides multiple pattern


match algorithms or a peak match algorithm.
• Pattern matching algorithms try to match spectra
based on the relative peak intensity and position of
the unknown spectral absorptions.
• Peak matching algorithms specifically compare
the peak table for the unknown with peak tables
developed from the library data.
• When using a Search software for the first time, it
is helpful to Search a known compound vs. the
library to evaluate the various pattern matching
algorithms.
Example Search Result

HI #12135; 2-P ropanol; 2-P ropano


Absorbance

0
T ransmission Dat a - Isopropanol

-1

3000 2000 1000


W avenumbers
Search Hit List Result
Example Search Result
Percent Transmittance

DEMO #83; ACET ONIT RILE


0

T ransmission Dat a - Acet onit rile

3000 2000 1000


W avenumbers
Search Hit List Result
Incorrect Search Result
Percent Transmittance

QF #793; 4-(METHYLTHIO)-2-PHENYLBU
0

Poly sty rene

3000 2000 1000


W avenumbers
Search Result Hit List
Search with Proper Library
Percent Transmittance

HD #7289; POLYSTYRENE
0

Poly sty rene

3000 2000 1000


W avenumbers
Proper Library Hit List
ATR Data Search
ATR Search Result Hit List

1 SEA426 314 0.959 Poly(methyl methacrylate) (nom mw:


400,000)
2 SEA426 219 0.953 Poly(methyl methacrylate) (nom mw: 12,000)
3 SEA426 313 0.950 Poly(methyl methacrylate) (nom mw: 90,000)
4 SEA426 276 0.950 Poly(2-acrylamido-2-methylpropanesulfonic
acid:methyl methacrylate) (95% methy
5 SEA426 40 0.947 Poly(methyl methacrylate) (mw: 33,000)
6 SEA426 178 0.937 Poly(ethylene adipate)
7 SEA426 410 0.935 Polyurethane resin
8 SEA426 207 0.933 Poly(isobornyl methacrylate)
9 SEA426 290 0.932 Poly(cyclohexyl methacrylate)
10 SEA426 408 0.927 Polyurethane resin (n=650/100C)
Electronic Data Manipulation

• Data manipulation CAN alter the original


data and produce meaningless results.
• ALWAYS attempt to collect good data,
manipulating data ONLY when necessary.
• ALWAYS retain the original data for
comparison with manipulated data.
• Some data manipulation is necessary for
interpretation, others are useful for display
purposes only.
Types of Data Manipulation for Interpretation

• Spectral subtraction – subtraction of one spectrum from


another; i.e. subtracting solvent from a solution
• Derivative – used to highlight shoulders or small peaks
• Normalize – multiplication of the spectrum and baseline
reset - increasing the intensity of spectral peaks
• Kramer’s Kronig – conversion for specular reflectance data
• Kubelka Munk – conversion for diffuse reflectance data
• Data conversion - %T to Absorbance, cm-1 to microns, etc.
• Arithmetic – adding several spectra and dividing to
produce an average, or obtaining a ratioed spectrum from
single beam spectra
Types of Data Manipulation for Display

• Smoothing – removing noise: loss of spectral resolution


• Deconvolution or curve fitting – increasing spectral
resolution or band separation: can increase spectral noise
• Baseline correction – correcting baseline slope: can distort
spectral peaks
• Interpolation – changing data point interval: can produce
noise or decrease spectral resolution
• Line out or ‘Blank’ – removal of known contaminant
peaks; e.g. CO2 or H2O
• Truncation and/or concatenation – cutting out an undesired
spectral range or ‘pasting’ together two spectral ranges
Subtraction
Subtraction of a solvent from a solution can provide a spectrum that is
easier to interpret, and, possibly search versus a library database.

3 0 .7

0 .6

0 .4

A b s A b s

0 .2

0 -0 .1

4 0 0 0 3 0 0 0 2 0 0 0 1 0 0 0 7 0 0 4 0 0 0 3 0 0 0 2 0 0 0 1 0 0 0 7 0 0

W a v e n u m b e r [ c m -1 ] W a v e n u m b e r [c m -1 ]
Derivative
1st Order Derivative Result
0 . 1 2

0 . 1

0
A b s

- 0 . 1

- 0 . 1 2

3 5 0 0 3 0 0 0 2 0 0 0 1 0 0 0 4 0 0

W a v e n u m b e r [ c m - 1 ]
Normalize
1 . 5

A b s

0 . 5

- 0 . 1

4 0 0 0 3 0 0 0 2 0 0 0 1 0 0 0 7 0 0

W a v e n u m b e r [ c m - 1 ]
Kramer’s Kronig Transform
Kubelka-Munk Transform
1.7

1.5

K/M

0.5

0
4000 3000 2000 1000 400
Wavenumber [cm-1]
Smoothing
80

50

%T

-30
4000 3000 2000 1000 500
Wavenumber [cm-1]
Deconvolution
4

Abs 2

0
1650 1500 1000 500 400
Wavenumber [cm-1]
Baseline Correction
100

50

%T 0

-50

-90
4000 3000 2000 1000 400
Wavenumber [cm-1]
Data Interpolation
101

90

AU

80

70
2250 2200 2100 2020
Wavenumber [cm-1]
Line Out – ‘Blank’
100

90

80

%T

70

60

50
4000 3000 2000 1000 400
Wavenumber [cm-1]
Search and Data Manipulation

• Spectral Searching should always be approached


with caution, examining the results carefully.
• The best library is built from spectra collected on
your instrument with your samples.
• Data manipulation should be used sparingly as the
results may distort the spectrum, producing
artifacts.
• Always retain the original data.
• The most useful data is collected with care and
does not require extensive modification.

You might also like