2.

1 Introduction to Spectroscopy

What is spectroscopy?

Studying the properties of matter through its interaction with different frequency
components of the electromagnetic spectrum.

Latin: “spectron”—ghost or spirit

Greek: “  ”—to see

With light, you aren’t looking directly at the molecule—the matter—but its
“ghost.” You observe the light’s interaction with different degrees of freedom of
the molecule. Each type of spectroscopy—different light frequency—gives a
different picture → the spectrum.

Spectroscopy is a general methodology that can be adapted in many ways to

extract the information you need (energies of electronic, vibrational, rotational
states, structure and symmetry of molecules, dynamic information).

Goals:

• Understand how light interacts with matter and how you can use this to
quantitatively understand your sample.

• Understand spectroscopy the way you understand other common tools of

measurement like the watch or the ruler.

• See that spectroscopy is a set of tools that you can put together in different ways
to understand systems → solve chemical problems.

The immediate questions that we want to address are:

• What does light do to sample?

• How do you produce a spectrum?
• What EXACTLY is a spectrum a measurement of?
What does a spectrum measure?

Interaction of light with a sample can influence the sample and/or the light.

Method involves: (1) excitation and (2) detection.

The basic idea: Characterize light after

2a sample.
1
Light (E.M. wave) Sample

Characterize change in
sample. (photoacoustic spec.)
Other excitation source 2b This borders on photochemistry

In most spectroscopies, we characterize how a sample modifies light

entering it.

1) Absorption: Change in intensity I of incident light

Sample attenuates light → transmission T=I/I0

I0 I

sample detector

We measure the absorption of light at different frequency or wavelength.

// (characteristic frequency/wavelength of

light entering sample)

2.1 Introduction to Spectroscopy Page 2

2) Emission: Excitation induces emission of light from the sample (usually of
different frequency).

(Emitted in all directions)

in sample

 out detector

Includes: Fluorescence (emission from excited electronic singlet states)

Phosphorescence (emission from excited electronic triplet states)
Raman Scattering (light scattering involving vibrational transition)

3) Optical Rotation: Change of phase of light incident on sample (rotation of

polarization)

Let’s work on describing absorption.

E
Let’s look at a typical absorption spectrum.

2.1 Introduction to Spectroscopy Page 3

What are the axes?

X- axis: Characterizes the input light in terms of frequency-wavelength-energy

Wavelength  (nm, m, Å),
 = c
Frequency  (cycles/sec or s-1 or Hz) =
2 
 =  (rad/sec) (angular frequency)
 = c = 1/ expressed in units of cm-1 (wavenumbers)
Energy E = h (expressed as eV or as cm-1 using E/hc = /c)

Conversions
 ( cm-1 ) =   ( nm )
 ( eV ) =   ( nm )

y-axis:
Absorption

I
A (  ) = − log = () c L (Beer’s Law)
I0

I0 = light intensity incident on the sample

I = light intensity after the sample
 = molar decadic extinction coefficient (M−1cm−1 ) – the molecular quantity
c = concentration (M)
L = sample length (cm)

This comes from assuming that the fraction of light absorbed as you propagate
through the sample is proportional to the distance traversed: dI I = − dx

2.1 Introduction to Spectroscopy Page 4

 How do your measure absorption spectra?

Measure the change of intensity of light at different frequencies as it

passesthrough a sample.

Two types of spectrometers:

1) Dispersive
2) Fourier transform

Dispersive spectrometer: Separate different frequency components

red
slit

prism

blue
sample
detector

 

Well talk about Fourier transform spectrometers later.

Fourier transform spectroscopy (FTIR)
Instead of separating frequency components, could measure the time-dependent field

Simple example: Gaussian distribution of frequencies, peaked at p with bandwidth .

E (t ) = 

−(− ) cost  e −(t  cos(pt )
−
 2 )2
d E e  p

0 
Ee
2.1 Introduction to Spectroscopy Page 5
with  = 1/(2).

Light field temporal profile is a pulse with a Gaussian time dependence and a duration, ,
that is given by the inverse of the spectral bandwidth. Let’s look at such a pulse.

E(t)

Pulse is short because many frequency components go out of phase and cancel at long
times. Broader frequency range  shorter pulse duration

2.1 Introduction to Spectroscopy Page 6

Integral above is a Fourier integral. In general, f(t) and F(w) are Fourier transforms ofeach
other if they are related through the following integrals.

1 f (t )cos(t ) dt
F ( ) =
2 −



1 
f (t ) = F (  ) cos ( t ) d
2  −

General result: We can measure the time-dependence of the field – whatever it is,
Gaussian or otherwise – and determine the frequency-dependence taking the inverse
Fourier transform as shown above.

The field can be fully characterized through measurement, and mathematical description,
in the frequency domain (measure the amplitude of each frequency component, and if
needed the phase also) or in the time domain (measure the time-dependent profile of the
field).

How can we determine the time-dependent profile of a light field? It oscillates too fast
to measure electronically.

Can measure the field optically with an interferometer.

Beamsplitter (dashed line) transmits and reflects equal parts of the light which return to
the beamsplitter after reflection off the mirrors. Depending on mirror positions, they

2.1 Introduction to Spectroscopy Page 7

interfere constructively or destructively. Detector measures oscillating intensity as
function of mirror position!

2.1 Introduction to Spectroscopy Page 8

 I (t) = I0 1+ cost

where t ’ is the temporal delay between the two fields. In terms of the spatial delay x’

I ( x) = I0 1+ cos kx

“Position” domain measurement of the field

Fourier transform of


I ( k ) =  − dx I ( x ) cos kx

gives output as a function of wavelength through relation  = 2/k

or as a function of frequency through relation /k = c

As before, make measurement with & without sample in place, and compare results to
determine absorption spectrum.

Works the same way in FTIR even with an incoherent (or partially coherent) light source

This is a way of processing all wavelength/frequencies simultaneously →

IR/NMR

2.1 Introduction to Spectroscopy Page 9