Photochemistry- Fundamentals

and Applications
Brian Seger
9-10-2014
 Energy
• Thermodynamics- Things will go to a lower energy state.
• Photochemistry- Takes photons and creates high energy
states.

Thermodynamics
Photochemistry
Thermodynamics

Usable
Energy
 Objectives
• Photoabsorption- You should be able to tell how much light
any photocatalyst use to do useful amount of work

• Doping- You should be able to tell the difference between an

n-type and p-type material.

• Band Alignment- You should know why band alignment to a

redox reaction is important.

• Applications- You should be able to know multiple

applications for photocatalyst and the basic approaches to
making them succsessful.
 Breaking photocatalysis down to the
fundamentals.
• A photon forces an electron to a higher energy level.
• The electron needs to get to the surface and react before it
falls back down into it’s initial/lower energy level.

e- e-
𝐵

𝐴
h+ h+ Lost Energy
𝐴+
 How much energy can we get from the sun?

• The x-axis is in wavelength,

but we would like this axis
to be energy.
• How do we do that?

• Speed of light equation:

𝑐=𝑣
Wavelength Frequency http://rredc.nrel.gov/solar/spectra/am1.5/

• Plank-Einstein relation:
1240
𝐸=h 𝑣 𝐸 ( 𝑒𝑉 ) =
 ( nm ) 
Energy Frequency
 200

Photons x10 (m hr eV ) This is the AM1.5 solar

-1
160

spectrum. ~average solar

-1

120
irradiation in the middle of
-2

the USA. See Wikipedia for

23

80

exact details.
40

0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Photon Energy(eV)

• This graph shows us the energy of all the photons coming

from the sun.
• Important note:
• The unit is ‘electron-Volt’, which is a unit of energy.
• ‘Volt’ is a unit of potential.
• 1 eV is the amount of energy it takes 1 electron to change
its potential by 1 volt.
 Molecular Photocatalysts
• Molecular photocatalysts have distinct energy levels.

3 eV 2 eV 1.5 eV
LUMO
200
Photons x10 (m hr eV )
-1

160 Photon Flux at AM1.5

-1

120
-2
23

80

40

HOMO e-
0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Photon Energy(eV)

molecular dye

This photon energy needs to match this energy gap.

• Molecular photocatalyst only absorb efficiently at one

wavelength.
 Semiconductor Photocatalysts
• Semiconductors have bands rather than distinct levels.
3 eV 1.5 eV 1 eV
200
Conduction Band
Photons x10 (m hr eV )
-1

160 Photon Flux at AM1.5

-1

120
-2
23

80

40

e-
0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Photon Energy(eV)
Valence Band

This photon energy needs to roughly this energy gap.

• Semiconductors can absorb a much broader range of light.

Photochemistry vs. Photoelectrochemistry
• Photochemistry- having both the oxidation and reduction
reaction take place on the same material
• Photoelectrochemistry- only having the photoabsorber doing
oxidation (or reduction) and the electron (or hole) then goes
through a wire to a counter electrode to do the other reaction

Photochemistry Photoelectrochemistry
-
Can measure/ alter the
e -
𝐵 current or voltage
h+
𝐴+ 𝐵 e- e-
𝐴 -

h+
𝐴+
𝐴
 Analysis
• It’s hard to analyze what is going on in photochemistry.
• Analysis techniques include:
- Measuring the products
- Spectroscopy (normally expensive equipment)
• Photoelectrochemistry is much easier to analyze
• Common photo-electrochemcial techniques include.
- Standard electrochemical techniques- CV, CA…
- Incident Photon to Current Efficiency (IPCE)
- Faradaic Efficiency (electron to chemical product efficiency)
- Impedance Techniques
 Semiconductor Fundamentals
• Valence band are filled energy states.
• Conduction bands are empty states.
• The Fermi Level is electronic version of chemical potential.
• Undoped materials have a Fermi
Conduction Band
level ½ between VB and CB.
• p-type semiconductors have a
Fermi level near the VB. Fermi Level
• n-type semiconductors have a
Fermi level near the CB. Valence Band

Photocatalyst
 PEC Fundamentals
• PEC is all about the semiconductor-electrolyte interface.
• Much of this work is purely semiconductor electrochemistry.
• Initially I will describe dark conditions.

Dominant redox
reaction in solution
(i.e chemical potential
of solution)
H+/H2

How do these 0.5M H2SO4

interface ?

Photocatalyst Electrolyte
 PEC Fundamentals
• The electrolytes charge dominates the chemical potential of the
system.
• The Fermi level matches the chemical potential of the solution.
• The VB and CB should move linearly with the Fermi level.

• The 1st few layers of water bind

strongly to the semiconductor.
• There are so few carriers, the
semiconductor can’t balance
H+/H2
this charging.
• This results in charging at the 0.5M H2SO4
surface.

Photocatalyst Electrolyte
 PEC Fundamentals
• The semiconductor doesn’t have many electronic states to
conduct.
• Thus the charging must be dissipated over some distance.
• This dissipation results in 1-10,000 nm
band bending.
• The higher the dopant
density, the smaller H+/H2
depletion width.
• Metals basically have no 0.5M H2SO4
band bending.

Photocatalyst Electrolyte
 Advantages of Band Bending
• Band bending is great for photochemistry!!
• Electrons love to go down hill, and holes (i.e. anti-electrons) like
to go uphill.
• Thus band bending
separates electrons and
hole. e-
• Therefore they can not
H+/H2
recombine.
• The holes go to the surface Electrode
h+
to react.
• The electrons go to the
bulk/electrode. Photocatalyst
 The Electrode
• Rather than have the electrolyte produce band bending, we can
have an electrode do this.
• The electrode sets the Fermi level, but does not effect the
solution chemical potential.
• This produces a non-
equilibrium state.
• This also increases our e-
band bending. H+/H2

• Increased band bending Electrode

means better electron-hole h+
separation.

Photocatalyst
 n-type versus p-type
• n-type will almost always force holes to the surface.
• Holes oxidize, thus n-type electrode do oxidation reactions.
• p-type will almost always force electrons to the surface.
• Electrons reduce, thus p-type electrode do reduction reactions.

p-type Semiconductor n-type Semiconductor

e-

e-
H+/H2
H+/H2
Electrode
Electrode

h+

h+
 Counter Electrode
• If the semiconductor does either an oxidative (or reductive)
reaction, where is the corresponding reductive (or oxidation)
reaction?
• This takes place on the metallic counter electrode (in the dark).
• The transfer of e- or h+ through the wire can be measured as
current.

𝐵 e- e-
-

h+
𝐴+
Counter Electrode 𝐴
 What reactions to do?
• p-type reduction reactions:
• H2O + e- H2 + OH-
• CO2 + e- CxHyOz (hydrocarbons) + OH-
• O2 + e- O2* Superoxide/O2 radical – oxidizing agent
This is how photographic films are made.
• Ag+Cl- + e- Ag + Cl-

• n-type oxidation reactions: hole proton

• OH + h
- +
H2O + O2
• Carbon based pollutants + h+ CO2 + H+
• Cl- + h+ Cl2

• Overall reactions (using counter electrode or 2 photoabsorbers):

• H2O H2 + O2
• Carbon based pollutants CO2 + H2O + others (This is not well defined)
• CO2 + H2O CxHyOz (hydrocarbons) + O2
 Issues with Band Levels
• For a p-type material, the electron will go to the electrolyte at
an potential no higher than the conduction band.
• The conduction band is pinned at the semiconductor/water
interface.
• Thus photo-electrons can only reduce species, which can get
reduced at or below the conduction band.
• In the case to the right, A can
not be reduced, whereas B can e- A/ A-
be reduced.
• Even if we bias the electrode to
B/ B-
increase band-bending, we still h+
can’t get the electron to
reduce A.
Energy
 Trick with Band Positions
• The band levels of oxides in water can be modified.
• Changes in pH, modify the oxide surface.
• For all oxides the valence band (and conduction band) vary via
the following equation.
RT  
𝑉𝐵=𝑉𝐵 ( @ 𝑝𝐻=0 ) + ×𝑝𝐻
𝑛𝐹
59 mV
• The VB or CB position are always relative to some reference
such as Ag/AgCl, SCE, or normal hydrogen electrode.
• The general rule is every increase in pH gives you 59 mV of
reducing power.
• However the redox potential of any reaction that has an H + as a
reactant needs 59 mV/pH more of reducing power.
 Band Positions and Redox Reactions
• Below are a list of band positions and redox reactions.
• For the semiconductors, more negative means more reducing
power.
Reduction Reactions
Reaction Potential
H+ + e- H2 -0.0
Semiconductor
Material VB CB CO2 + e- + H2O CH4 + OH- 0.1
Si 0.6 -0.5 CO2 + e- + H2O CO + OH- -0.14
TiO2 3.0 -0.1 O2 + e- O 2* -0.5
WO3 2.8 0.2 Oxidation Reactions
CdS 1.75 -0.5 Reaction Potential
GaP 1.25 -1.0 H2O + h+ O 2 + H+ 1.23
Cu2O 1.0 -1.0 Cl- + h+ Cl2 1.4
Fe2O3 2.4 0.3 HCOOH + h+ CO2 + H2O 0.25

All potentials are vs. NHE All potentials are vs. NHE
 More Issues with Semiconductors
• Just because an electron is in the conduction band, doesn’t
mean that is the chemical potential of the electron.
• Just because an electron is in the conduction band, doesn’t
mean that is the chemical potential of the electron.
• The photo-electron has a quasi Fermi-level, which is
somewhere below the conduction band. Electron quasi
Fermi level
• The quasi-Fermi level is based e-
on light intensity and the ability
to efficiently separate electrons
and holes.
B/ B-
• Ole Hansen will discuss this in h+

detail when he presents.

Energy
 Even More Issues with Semiconductors
• Issue #1- the electron quasi-Fermi level needs to have a higher
energy than the redox couple to reduce it.
• I know of no pure way to determine where your quasi-Fermi
level is located at.
• If you apply bias, you can increase band bending, which may
help increase the Quasi-Fermi level.
• Issue #2- If the redox reaction Overpotential
isn’t easy, you will also have e-
needed for
reaction
catalytic overpotentials.
• If overpotential is too large, B/ B-
you still won’t be able to have h+
reduction.
Energy
 Photoabsorber + Catalyst
• Unless you are extremely lucky, the photoabsorber will not be
the best electro-catalyst for your reaction.
• Thus you need a co-catalyst.
• At this point you are doing coupled photo-absorption and
electrochemistry.
• Ifan Stephens explained the electrochemistry in a previous
lecture. Co-catalyst to
e-
reduce
• If all this becomes
overpotential
electrochemistry, why do we
do photoelectrochemistry? B/ B-
h+ Small overpotential
allows this reaction
to occur.
Energy
 Photovoltage
• PEC lets us do electrochemistry with less/no applied bias.
• The electrode potential matches the Fermi level of the hole
quasi-Fermi level.
• Thus the difference between the hole quasi-Fermi level and
electron quasi-Fermi level is the extra voltage the PEC gives us.

Electrochemistry Photo-electrochemistry
e-

Electrode
B/ B- This reaction takes B/ B-
more energy than
this reaction.
Electrode

h+
 Results
• Below is a cyclic voltammogram of various electrodes for the
H+/H2 reaction.
PEC gain - good catalyst
PEC gain – no catalyst
4

0 Gain from solar energy

-4 p- Si
Current (mA/cm )
2

Pt
-8
Lose from inefficient
-12 catalysis reaction
-16
Pt on p-Si
-20

-24
-0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4
Thermodynamic
V (vs. RHE)
redox potential
 Losses at the Counter Electrode
• Since the photocatalyst has catalytic losses the counter
electrode will also have catalytic losses.
• To eliminate counter electrode losses, we use a 3-electrode cell.
• In a 3 electrode cell the voltage is measured between the
photocatalyst and reference electrode.
• The current still runs between the photocatalyst and counter.

• In a 3 electrode cell, the

system provides extra
voltage to the counter
electrode to mitigate the 𝐵 e- e-
catalytic losses. -

h+
𝐴+
Counter Electrode 𝐴
Applications of Photochemistry and
Photoelectrochemistry

• Water Splitting
• CO2 Reduction
• Pollutant Degradation
• Dye-Sensitized Solar Cells
• Perovskite Solar Cells
• Sensors
 Water Splitting
•
• The main idea behind this reaction is to produce hydrogen,
which can replace fossil fuels as an energy source.
• Water is our reactant, and oxygen is a byproduct.
• 1.2 eV.
• Realistically you need about 2.3 eV.
200
560 Photon Flux at AM1.5
𝐻2
)
-1 s -1)

e-
-2 -1

480
160
hr eV

Photon Flux at AM1.5

-2 m

400 +
-1
(nm

120
x1023 (m

320
1.23 eV
16

240
80
Photons x10

160
h+
Photons

40
80

0
0.5 400
1.0 600
1.5 2.0
800 2.510003.0 1200
3.5 1400
4.0
𝐻 2𝑂
Photon
Wavelength
Energy(eV)
(nm)
 Water Splitting
• H2 has to compete with fossil fuels, which are extremely cheap.
• There are 2 approaches:
- Do it really cheap, and then try to bring up efficiencies.
- Do it really efficient and then bring down costs.
• I will discuss the high efficiency first approach because that is
more of a physics way of doing it.
• To optimize water splitting
you need 2 photocatalysts
• 1 absorbs the blue light
• 1 absorbs the red light
• A 2 photocatalyst design
gives you just enough
voltage to split water.
 Physical
Design

Optical Absorption
Properties

Energetic
Properties
 Energy Levels of Our Processes
• The photoanode will oxidize water to oxygen while the
silicon will reduce the protons to hydrogen.
• The Fermi levels of both photoabsorbers must equilibrate.
• Catalysts will be needed to improve reaction kinetics.

Potentiostat Potentiostat

Walter et al. , Chem Review, 2010

 1-Photon Overall Water Splitting
• This is inherently a cheap method because you can use
nanoparticles dispersed in water rather than aligned films.
• The problem is: Should the photocatalyst be p-type or n-type?
• There is no good answer to this.

• If it is n-type, holes will easily H+

go to the surface, but
electrons can’t. e- 𝐻2

• If it is n-type, holes will easily

go to the surface, but + H+
h+
electrons can’t. 𝐻 2𝑂
 1-Photon Overall Water Splitting
• Approaches to resolving the band bending issue.
• Don’t dope the material, thus mitigating bandbending.
- Creating a non-doped semiconductor is hard.
• Have small particles.
- Very small particles prevent band bending from building up (< 50 nm)
• Have a co-catalyst create an ohmic contact.
- Forming an ohmic contact prevents band bending at localized spots.

Co-Catalyst
H+ e-
𝐻2

+ H+
h+
no band 𝐻 2𝑂
bending
 Water Splitting- Recombination
• Electrons and holes can recombine to waste energy.
• In a 1 photon device it is very easy for the H2 and O2 to
recombine to form water.
• Also the hole could oxidize H2 instead of H2O. And the electron
could reduce O2 rather than H+.
• In a 2-photon/film these back reactions are less prevalent.
1 Photon 2 Photon
 CO2 Reduction
• CO2 reduction to hydrocarbons is not easy electrochemically,
thus it is even harder for PEC.
• If we simply look at CO2 to methane.
Overall Reaction: CO2 + H2O CH4 + O2 Same as water
Anode Reaction: OH- + h+ H2O + O2 splitting
Cathode Reaction: CO2 + H2O + e- CH4 + OH-

8 electron transfer, thus hard reaction

• The biggest impediment to CO2 reduction is this cathode side

reaction.
H2 evolution: H2O + e- H2 + OH-

Most catalysts would rather do this reaction than CO 2 reduction.

 CO2 Reduction-Recombination
• Just like water splitting the hydrocarbons from CO 2 reduction,
can be oxidized back to CO2.

• This recombination will happen more for liquid CO 2 reduction

products rather than gaseous CO2 reduction products.
𝐻 2𝑂+𝐶𝑂 2
e-
OH

𝐻 2𝑂+𝐶𝑂 2
h+
OH

• We can take advantage of this troublesome issue though.

 Pollutant Removal
• Often there is organic pollutants that need to be removed for
drinking or sanitation purposes.
• Most of these are very easy to oxidize with n-type
semiconductors.
• Oxides are especially nice because they can produce OH
radicals from water.
• OH radicals are highly
e-
oxidative and can migrate Redox potentials
into solution. of organics
• Thus either the photo-hole or
the radical can oxidize the h + H2O/OH*+ H+ (2.85)
pollutant. 3.0
TiO2
All potentials are vs. NHE
 Pollutant Removal
• While the hole oxidizes, what does the electron do?
• If there is oxygen in the system, which there always is, the
electron can reduce the oxygen to form a superoxide (or
oxygen radical.)
• Oxides are especially nice because they can produce OH
radicals from water.
e-
(-.5) O2/O2*

How does this Organic

reaction happen h+
H2O + CO2 3.0
TiO2
All potentials are vs. NHE
 Pollutant Removal
• Here is a more detailed mechanism for electron and hole
degradation reactions for TiO2
• Radicals react very fast, thus it is hard to analyze mechanisms.

• .

Hoffman et. al, Chem. Rev. 1995

 Dye Sensitized Solar Cell (DSSC)
• Michael Gratzel popularized this approach to solar cells in 1991.

e- e-
Load

e-

Solar
Irradiation Open Circuit
3I- Voltage
h+
h+
I3-
Dye
TiO2
Transparent Conductive
Electrode (TCE)
 Dye Sensitized Solar Cell (DSSC)
• The key to success in a nanoporous TiO 2.

Ineffective Planar Effective Porous

Approach Approach
Blocking TiO2
layer to prevent
back reactions

• There is no TiO2
surface area for
these dyes.
• This gives low
TiO2
currents
Porous TiO2 allows for much
Only first layer is active more dye adsorption.
 Dye Sensitized Solar Cell- Results
• In 1997 the record was 10% and in 2014 it is 11.9%.
• The biggest issue is the liquid electrolyte leaking/evaporating
over time.
• The major focus is on finding good hole conductors to replace
iodide.
• The annoying thing is a
change in dye many times
means a new hole
conductor is needed.
• To the right is state of the
art performing DSSC

Campbell et. al., JPC-C, 2014

 Perovskite Solar Cell
• These were initially an off-shoot of dye sensitized solar cells.
• The only difference was PbCH2NH3I3 replaced the dye.
• A hole conductor named ‘SPIRO’ is now used for DSSC and
Perovskites instead of Iodide because it is more efficient.
e- e-
Load

e-

Solar
Irradiation Open Circuit
Spiro
+
Voltage
h+
h+
Spiro
Perovskite
TiO2
 Perovskite Solar Cell
• Unlike DSSC, these do not need porous TiO 2.
• Electrons can hop from one perovskite until cell to another
very efficiently.
• Perovskites naturally are undoped (i.e neither n-type nor p-
type).

Efficiencies
2006 - 2.2 %
2009 - 3.8 %
Nov. 2012 - 10.9 %
July 2013 - 12.9 %
July 2014 - 17.9 % *
* Unpublished data
Liu et. al, Nature, 2013
 Solar Cell Efficiencies
• NREL in the US tests and verifies all record solar cells. They
frequently update the graph showing the best efficiencies

Greg Wilson and Keith Emery - National Renewable Energy Laboratory (NREL),
 Sensors
• One approach is just to monitor pollutants in water.
• In this case you use photocatalysts to oxidize a small fraction of
the pollutants.
• To oxidize the pollutants the photocatalyst must use oxygen.
• You then use a detector to monitor changes in oxygen in the
water.
• A decrease in O2 means e-
O2/O2*
an increase in pollutants.
Organic Organic
• Once calibrated you can
h+
measure pollutants in H2O + CO2
H2O + CO2
solution via oxygen in TiO2
solution. All potentials are vs. NHE
 Sensors
• A non-photochemical way to use semiconductors is to use
them as an electrical sensor.
• Hydrogen is known to intercalate/ react with TiO 2.
• Hydrogen can change the resistivity of TiO2 by orders of
magnitude.
• Thus this is a phenomenal H2 sensor.
Platinum
TiO2 Nanotubes

H2

Titanium
Mor et al., Sol. Ener. Mat. & Sol. Cells, 2006
 Summary
Fundamentally
• We kind of understand the general concepts such as band-
bending, quasi-Fermi levels and redox reactions.
• We do not understand defect sites at the interface, and co-
catalyst-semiconductor interactions.

Applications
• The 2 main applications are energy production and
contaminant removal.
• For energy production we need to increase efficiency and
durability, while decreasing costs.
• For contamination removal, we need to understand
degradation mechanisms and by-products.