Advantages of Fourier Transform Spectroscopy PowerPoint PPT Presentation

presentation player overlay
1 / 39
About This Presentation
Transcript and Presenter's Notes

Title: Advantages of Fourier Transform Spectroscopy


1
(No Transcript)
2
(No Transcript)
3
(No Transcript)
4
(No Transcript)
5
(No Transcript)
6
(No Transcript)
7
(No Transcript)
8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
Advantages of Fourier Transform Spectroscopy
  • Time domain spectroscopy

26
Advantages of Fourier Transform Spectroscopy
  • Fig. 7-40, pg. 185 "Illustrations of time doamin
    plots (a) and (b) frequency domain plots (c),
    (d), and (e)."

27
Advantages of Fourier Transform Spectroscopy
  • Fig. 7-41, pg. 186 "Time domain spectrum of a
    source made up of several wavelengths."

28
Advantages of Fourier Transform Spectroscopy
  • Michelson Interferometer
  • Fig. 7-42, pg. 186 "Schematic of a Michelson
    interferometer illuminated by a monochromatic
    source."

29
Advantages of Fourier Transform Spectroscopy
  • Fig. 7-43, pg. 188 Comparison of interferograms
    and optical spectra.

30
FT-IR
31
Advantages of Fourier Transform Spectroscopy
  • Throughput no slits or gratings. Full power of
    source reaches detector
  • Resolution depends on length of mirror scan,
    can be very high
  • Rapid data collection

32
Scan Times, Frequency Domain
  • Resolution elements spectral width/resolution
  • IR spectrum, 500 cm-1 to 4500 cm-1
  • For 3 cm-1 resolution, 4500/3 1500 elements
  • For 1 cm-1 resolution, 4500/1 4500 elements
  • Scan at 0.5 sec/element,1500 elements,12.5min
  • Scan at 0.5 sec/element,4500 elements,37.5min
  • Signal/Noise S/N
  • S/N improvement n1/2 where n time per element
  • To improve S/N by 3, n 9. 9 X 0.5 s 4.5 sec
  • Scan time goes to 9 x 37.5 min 5.6 hrs for high
    resolution scan.

33
Scan Times, Time Domain
  • About 0.5 sec per scan required
  • Entire spectrum collected with each scan
  • Assume short data collection, 12.5 min
  • 750 sec/0.5 sec per scan 1500 spectra
  • S/N improvement is 15001/2 39
  • Multiplex advantage due to short scan time in
    time domain spectroscopy

34
Measurements in the time domain
  • Light moves fast, c 3 x 108 m/s
  • Consider infrared, ? 3 x 10-6 m
  • 3 x 10-6 m/3 x 108 m/s 10-14 s for one
    wavelength of light to pass by a detector
  • To resolve 3 x 10-6 m from 4 x 10-6 m in
    frequency domain requires time resolution of less
    than 10-14 s - too fast for any known photo
    detector!

35
Measuring time versus intensity
  • P(t) kcos(2p?1t) kcos(2 p?2t)
  • After this is measured, P(?) is given by a
    Fourier transform of P(t).
  • How do you measure P(t) when c3 x 108m/s?
  • Answer Interferometry
  • P(t) P(t)/k, k is called modulation factor
  • f frequency at detector 2 vm/?
  • vm mirror velocity, ? wavelength of light

36
Retardation
Retardation d 2(M-F) dmax 0 dmin ½ ? dmin
dmax ½ ? ? 2(dmin dmax) At M F, d
0 At M F ½ ?, d - ?
37
Optical frequency and detector frequency
  • Detector frequency f 2vm/?
  • c ??photon
  • f 2vm/? (2vm/c)?photon k?photon
  • Since c is large, k is small and f is an easily
    measured frequency for a simple detector

38
Optical frequency and detector frequency
  • detector frequency (observed) is given by f
    2vm/?. This is the frequency for observing
    fringes in the interference pattern.
  • Example vm 3 cm/s, ? 3 x 10-4 cm
  • f104 s-1 and 1/f 10-4 s
  • For ? 3 x 10-4 cm, ?photon c/?
  • ?photon (3 x 1010 cm/s)/(3 x 10-4 cm) 1014
    s-1 and 1/? 10-14 s
  • Conclusion Detector needs to detect signals
    with a time resolution of about 10-4 s to
    reproduce the optical signal that would require
    time resolution of about 10-14 s

39
Resolution with interferometry
  • Principle one complete beat pattern must be
    scanned to fully resolve all components
  • Closely spaced features have smaller differences
    in frequencies, so a long cycle is needed to see
    a full beat
  • To resolve 2 features separated by ??, the length
    of the scan distance is given by
    L 1/2??
  • Example For resolution of 0.1 cm-1,
    L 1/(2)(0.1 cm) 5 cm
Write a Comment
User Comments (0)
About PowerShow.com