Alberto Tosato Thesis

UNIVERSITÀ DEGLI STUDI DI PADOVA

DIPARTIMENTO DI INGEGNERIA INDUSTRIALE

CORSO DI LAUREA IN INGEGNERIA CHIMICA E DEI MATERIALI

Tesi di Laurea in

Ingegneria Chimica e dei Materiali (Laurea triennale DM 270/04 – indirizzo Materiali)

SYNTHESIS CHARACTERIZATION AND STUDY OF

PHOTOINDUCED ELECTRON TRANSFER IN Ag29

NANOCLUSTERS FOR SOLAR APPLICATIONS

Relatore: Prof. Giovanna Brusatin

Correlatori: Prof. Osman M. Bakr, Prof. Omar F. Mohammed Abdelsaboor

Laureando: ALBERTO TOSATO

ANNO ACCADEMICO 2014 – 2015

Summary

This thesis presents the work developed in a five month internship in collaboration with two

labs: Ultrafast laser spectroscopy lab, and Functional Nanomaterial lab at King Abdullah

University of Science and Technology (Saudi Arabia).

This work entails the synthesis and characterization of a new fluorescent nanocluster:

[Ag29(BDT)12(TTP)4]-3; and the photoinduced electron transfer from this nanocluster to a

cationic and neutral fullerene derivative.

The results obtained, besides showing that the synthesized silver nanoclusters are good electron

donors, point out the importance of electrostatic interaction, at the donor-acceptor interfaces, in

electron transfer: the process at the basis of a solar cells.

Table of Contents

1 Nanoclusters ...................................................................................................................... 3

1.1 Metal colloids .............................................................................................................. 3

1.2 Nanoclusters and properties ......................................................................................... 3

1.3 Synthetic methods of noble metal fluorescent nanoclusters ........................................ 4

1.4 Applications ................................................................................................................. 6

2 Photoluminescence ........................................................................................................... 7

2.1 Excited state phenomena ............................................................................................. 7

2.1.1 Fluorescence ......................................................................................................... 8

2.1.2 Phosphorescence ................................................................................................ 10

2.2 Fluorescence quenching............................................................................................. 11

2.2.1 Forster resonance energy transfer....................................................................... 11

2.2.2 Dexter energy transfer ........................................................................................ 12

2.2.3 Photoinduced electron transfer ........................................................................... 12

3 Ultrafast Transient Absorption Spectroscopy ............................................................. 17

3.1 Principle of transient absorption spectroscopy .......................................................... 17

3.2 Experimental setup .................................................................................................... 19

4 Synthesis and characterization of Ag29(BDT)12(TTP)4 ............................................... 21

4.1 Synthesis .................................................................................................................... 21

4.2 Characterization ......................................................................................................... 22

4.2.1 Mass spectroscopy and analytical ultracentrifugation........................................ 22

4.2.2 X-ray diffraction ................................................................................................. 23

4.2.3 Optical properties ............................................................................................... 26

5 Photoinduced electron transfer ..................................................................................... 27

5.1 Steady-state absorption and fluorescence .................................................................. 27

5.2 Femtosecond transient absorption ............................................................................. 31

5.2.1 Instrumentation for time-resolved measurement ................................................ 35

Introduction

In a period where energy consumption is tremendously increasing, and urge the necessity of a

sustainable current production; solar energy, the most abundant form of energy available on our

planet, could be a useful alternative to the conventional energy production methods. Thus, it is

essential to explore new ways for harvesting solar energy in order to develop cost-effective and

highly-efficient solar cells.

Fluorescence noble metal nanoclusters are a new class of nano-materials that have just stepped

into photovoltaic research. Metal nanoclusters are nanoparticles composed of 2 to roughly 150

metal atoms protected by a shell of ligands: organic molecules that bind to the surface atoms of

the cluster. They exhibit quantum-confinement effects, which result in several unique properties

including discrete electronic structure, defined HOMO-LUMO electronic transitions,

photoluminescence emission, size dependent catalytic activity, and magnetism. Research on

nanoclusters is showing their effectiveness in boosting the efficiency of dye-sensitize and

polymer solar cells and in developing new types of solar cells. The first Metal-nanocluster-

sensitized solar cell, “Au-nanoclusters-sensitized solar cell”, was recently assembled, showing

an efficiency greater than 2%. This is a promising success for this field that has just begun.

In the first part of this thesis, is presented in the synthesis, optical properties and structure of a

new atomically precise nanocluster: Ag29(BDT)12(TTP)4. While, in the second part is presented

the study of the excited state interactions of the bimolecular system [Ag29(BDT)12(TTP)4]-3 as

electron donor, and C60-(N,N-dimethylpyrrolidinium iodide)+n as electron acceptor, using

steady state absorption-fluorescence spectroscopy, and femtosecond time resolved absorption

spectroscopy. Femtosecond transient absorption spectroscopy is a powerful technique that

allows to study ultrafast processes such as electron transfer. Being able to determine whether

electron transfer occurs or not between two molecules, is a great advantage that permits to

forecast, without building any device, the current of the donor-acceptor system that is at the

basis of a solar cell. The system studied was than extended to a neutral fullerene, C60-(malonic

acid)n, to verify which role columbic-interaction plays in electron transfer and thus in the active

layer of a possible metal-nanoclusters-based solar cell. To prove definitely the role of

Coulumbic interaction in electron transfer, another system was then considered, substituting the

anionic nanoclusters with anionic carboxyl-capped CdTe quantum dots.

1 Chapter 1 3

Nanoclusters

Recent advance in nanotechnologies have given rise to a new class of nanomaterials:

fluorescent metal nanoclusters. These nanoparticles are of great importance because their sizes,

from few to one hundred or so atoms, are comparable to the Fermi wavelength of electrons,

resulting in molecule-like properties, such as discrete electronic states and size-dependent

fluorescence.

To understand the synthetic methods and properties of nanoclusters is useful to consider another

structure that lies between bulk metals and nanoclusters: metal colloids.

1.1 Metal colloids

Metal colloids historically were the first nanoparticles to be investigated. Metal colloid is

intended to be a solution of metal particles, dispersed in a medium, that have at least one

direction roughly between 1 nm and 1 µm. The agglomeration of the nanoparticles to bulk

material is prevented by the buildup of charges or ligands (molecules of a certain dimension

that can bound to the surface metal atoms of the nanoparticles) on the surfaces. These charges

can be due to different oxidation states of the metal atoms, or to the presence of the ligand itself.

The interaction of metal colloids with light give rise to interesting phenomena. The oscillating

field of light interacts with the free electrons of nanoparticles, causing a concerted oscillation

of electron charge that is in resonance with the frequency of the visible light. These resonant

oscillations are known as surface plasmon resonance and result in the absorption of photons of

the resonance’s wavelength. As the particle size increases, the wavelength of the surface

plasmon resonance, shifts to higher wavelengths.

Although, when the size reduces under few nanometers, the metal nanoparticles are too small

to support plasmons, and another phenomenon takes place: quantum confinement.

1.2 Nanoclusters and properties

Metal nanoclusters typically have diameters below two nanometers (2-150 metal atoms), and

have properties that place them in between isolated atoms and bulk material. When a

nanoparticle size is of the order of the Fermi wavelength of the electron, the continuous density

of state (a characteristic of bulk metals), brakes up into discrete energy levels (quantum

4 1. Nanoclusters

confinement), leading to a dramatic change of optical, electrical and chemical properties, as

compared to larger size nanoparticles or bulk materials.

A nanocluster defines a group of metal atoms bound together through metal-metal bonds or

metal-ligand bonds. To guarantee the clusters stability and prevent their aggregation,

nanoclusters are, in fact, protected by a shell of ligand, that for noble metals usually consist of

a thiol or a phosphine, since S and P give rise to good interaction with noble metal atoms.

The syntheses of nanoclusters are generally made in solution. In the last few years many

reproducible methods were developed to obtain a single sized nanocluster with well-defined

structure and chemical composition.

Nanoclusters are generally soluble in organic solvents, giving clear and strongly colored

solutions, but the ligand can be functionalized to make the clusters water soluble.

The clusters have well defined HOMO and LUMO energy levels, and therefore characteristic

peaks are shown in the absorption spectra due to the electron transition from occupied to empty

delocalized molecular orbitals, an example is shown in Figure 1.1, left panel. The HOMO-

LUMO gap generally becomes smaller with the increasing of the cluster size, thus the

absorption peaks are blue shifted as the size decreases (Figure 1.1, right panel). Thanks to the

HOMO and LUMO gap, the clusters are often fluorescent and a wide Stoke shift is usually

observed between absorption and emission spectra.

1.3 Synthetic methods of noble metal fluorescent nanoclusters

There are mainly two ways to synthetize noble metal nanoclusters in solution, a bottom-up

approach, that consist in the reduction and aggregation of metal ions coming from a certain salt.

And a top-down approach that allows to obtain nanoclusters from larger nanoparticles.

Figure 1.1 On the left panel structure and absorption spectra of thiolated-protected Au23 (yellow

atoms: S; other atoms: Au).

On the right panel trends of bandgap energies (Eg) with size (n) of Aun(SR)m nanoclusters.

2. Nanoclusters 5

Thiol and phosphine-containing small molecules are the most commonly adopted ligands in

gold and silver nanoparticle synthesis, owing to the strong interaction of both thiol and

phosphine with these two noble metals.

The bottom up synthesis proceeds as follows in Figure 1.2. A salt of the noble metal is dissolved

in a solvent and is then reduced in the presence of the ligand. The reduction can be either by

chemical reductant (e.g. sodium borohydride) or by light (visible or ultraviolet).

Figure 1.2 Schematic description of the synthesis of DPA-stabilized fluorescent Au nanoclusters

using THPC as a reductant.

Using the top down approach, fluorescent nanoclusters are produced by etching large noble

metal nanoparticles with the ligand (Figure 1.3). This reaction can follow two possible routes:

few atoms from the surface of a nanoparticle are extracted by the ligand, making metal particles

form a cluster. Or, the metal atoms are gradually removed from the surface of the nanoparticles

by the ligand, forming a complex with the ligand itself, and so that the nanoparticles become a

nanocluster.

Many other methods were developed for synthetizing fluorescence nanoclusters, using

polymers, proteins and DNA, however the two most common are the ones just presented.

Figure 1.3 Schematic illustration of two possible route for the formation of fluorescent Au

nanoclusters via etching of preformed MSA-protected Au nanoparticles.

6 1. Nanoclusters

1.4 Applications

The precise number of atoms that nanoclusters are composed of, their selectivity to bind with

other molecules/biomolecules, the possibility to excite them in the visible range, their

fluorescence and their low toxicity are interesting features that make them suitable for many

applications.

Thanks to their luminescence, nanoclusters are widely used for detection of metal atoms, small

biomolecules, proteins, and nucleic acids. In biology, they are used for imaging, labeling and

drug delivery. Many studies are being carried out on their catalytic and photocatalytic

properties, and in the last few years, noble metal nanoclusters started to be used in photovoltaic

devices. In particular, successful results were obtained in boosting dye sensitized solar cells

with gold nanoclusters, and recently the first metal-clusters-sensitized solar cells were

developed, showing cell performance comparable to those of quantum-dot-based solar cells,

but with the advantage of a lower toxicity.

2 Chapter 2 3

Photoluminescence

Photoluminescence is the emission of light from an atom or a molecule, which occurs after

excitation. It can be divided in two different processes, fluorescence and phosphorescence.

These two processes differ in the deactivation mechanism, from the excited state of the

molecule to its ground state.

2.1 Excited state phenomena

Many processes can occur after excitation of a molecule, all of which are presented in the

following paragraph with reference to Figure 2.2.

After excitation of a molecule i.e. through electromagnetic irradiation, the ground state of a

molecule is perturbed, and one of the electrons in the ground state S0 may have enough energy

to be promoted on one of the vibrational levels of a higher energy level Sn. Immediately after

excitation, the following processes might occur.

Through collision of the excited molecule with the solvents, the electrons lose energy going to

lower vibrational levels, this process is called vibrational relaxation, and is so rapid that the

lifetime of a vibrational excited molecule (<10-12 s) is lower than the lifetime of the

electronically excited state. For this reason, fluorescence of a solution always involves the

transition from the lowest vibrational level of the excited state.

If there’s an overlap between vibrational levels of two consecutive energy levels internal

conversion may occur: the electron can go from one excited energy level to the lower one,

through a non-radiative path. This transition, although less likely, is possible from level S1 to

the ground state S0.

Triplet excited state

Singlet excited state

Figure 2.2 Jablonski diagram

8 2. Photoluminescence

According to the Pauli Exclusion Principle, the spins of two electrons in the ground state must

be opposite. When a molecule is excited, the electron that reaches the higher energy level may

swap its spin and be oriented as the one in the ground state.

Thus, two different excited states are possible: the singlet excited state, in which the electrons

have antiparallel spins and the triplet excited state where the spins are parallel (Figure 2.1).

The transition from the singlet to the triplet excited state is called intersystem crossing. It is a

non-radiative transition between electronic states of different multiplicity (singlet to triplet).

This transition is enhanced by the overlapping of vibrational levels of the triplet and singlet

state. Intersystem crossing is most commonly observed with molecules that contains heavy

atoms such as Bromine or Iodine.

Once the electron of the excited molecule reaches the lowest vibrational level of the S1 energy

level, by undergoing the processes mentioned above, it can follow two emissive paths to go

back to the ground state: Fluorescence and phosphorescence.

2.1.1 Fluorescence

Fluorescence is the emission of a photon in the transition from the energy level S1 to one of the

vibrational level of the ground state S0 . This transition is slow (10-8 s), if compared to the ones

mentioned above, that’s why fluorescence always occurs from the lowest vibrational level of

the energy level S1 (Kasha’s rule).

Examination of the Jablonski diagram points out that the photons emitted have a lower energy

than the photon absorbed, because of the energy losses of the processes that occur before

fluorescence. In terms of wavelength this means that emission is always red shifted (shifted to

a higher wavelength) with respect to absorption (Figure 2.2). This phenomena is called Stoke

shift.

An interesting feature of fluorescence is that the emission spectra is often a mirror image of the

absorption spectra. The spacing of vibrational level of the ground state S0 and first excited state

S1 are in fact similar, this means that the energy of the photons absorbed by the transition from

S0 to one of the vibrational levels of S1, is similar to the energy of the emitted photons in

Figure 2.1 Schematic representation of singlet and triplet excited state

2. Photoluminescence 9

transition from S1 to one of the vibrational level of S0. Therefore every peak of absorption can

be considered as a particular transition from S0 to one of the vibrational levels of S1, as well as

every peak of emission can be considered as a particular transition from S1 to one of the

vibrational level of S0 (Figure 2.2 right).

The two most important measurable characteristic of fluorescence are quantum yield and

lifetime.

The quantum yield determines the efficiency of emission after excitation, the highest theoretical

quantum yield is 1. This means that for each photon absorbed, one photon is emitted, or in other

words, fluorescence is the only possible path for the electron from 𝑆1 to 𝑆0. However, there

are few cases where the quantum yield can be greater than 1. This happened when exciting the

molecule with very high energy radiation. This could cause the excitation of more than one

electron.

The quantum yield is therefore the ratio between the number of photons emitted and number of

photon absorbed, which equal the ratio of emission intensity 𝐼𝑒𝑚 against the absorption intensity

𝐼𝑎𝑏𝑠.

Another way to define the fluorescence quantum yield is by the ratio excited state decay rates.

𝜙 =𝐼𝑒𝑚𝐼𝑎𝑏𝑠

=𝑘𝑓[𝐹

∗]

∑ 𝑘𝑖[𝐹∗]𝑖=𝑘𝑓∑ 𝑘𝑖𝑖

(2.1)

Figure 2.2 Stoke shift (left panel). Mirror-image rule for anthracene (right panel), the numbers 0, 1,

and 2 refer to vibrational energy levels.


Where [𝐹∗] is the concentration of the excited molecule, 𝑘𝑓 is the rate of fluorescence and 𝑘𝑖

are the rates of all the reactions that bring the molecule back to a ground state (fluorescence,

internal conversion, quenching, energy transfer, etc).

Life time is the average time that the excited molecule, in a specific environment, remains in

the excited state before emitting a photon. This is really important because it determines the

time available for the fluorophore (the specie that undergo to fluorescence) to interact with the

environment. For a first order decay (equation 2.2) life time is represented by 𝜏

[𝐹∗] = [𝐹∗]0𝑒−𝑡/𝜏 (2.2)

The lifetime is

𝜏 =1

∑ 𝑘𝑖𝑖 (2.3)

Note that since [𝐹∗] is directly proportional to 𝐼𝑒𝑚 (𝐼 ∝ 𝑑[𝐹∗]𝑓𝑙𝑢𝑜/𝑑𝑡 = 𝑘𝑓[𝐹∗]), equation 2.2

can be written as

𝐼𝑒𝑚 = 𝐼𝑒𝑚,0𝑒−𝑡/𝜏 (2.4)

So the lifetime can be inferred through time-resolved fluorescence measurements.

2.1.2 Phosphorescence

Molecules in the excited state 𝑆1 can undergo to a spin conversion from the excited singlet state

to the excited triplet state 𝑇1, as explained before. Emission from 𝑇1 to the ground state 𝑆0 is

called phosphorescence. This transition is forbidden (Pauli Exclusion Principle), and as a result

the rate constants for triplet emission are several order of magnitude smaller than those from

fluorescence.


In the following table all the rates of the processes described above are reported for comparison.

Process Rate (seconds) Comments

Photon Absorption 10-14 to 10-15

vibrational relaxation <10-12

Internal conversion ~10-10

Fluorescence emission 10-5 to 10-10 singlet to singlet transition

Phosphorescence Emission 10-4 to 10 Slow: forbidden transition

2.2 Fluorescence quenching

Fluorescence quenching refers to any process that decreases the fluorescence intensity of a

sample. A variety of molecular interactions can result in quenching. These include excited-state

reactions, energy transfer, ground-state complex formation, and collisional quenching. In the

following paragraphs these different processes are presented.

2.2.1 Forster resonance energy transfer

The Förster energy transfer (FRET) is the phenomenon that an excited donor transfers energy

to an acceptor group through a non-radiative process. This non-radiative transfer mechanism is

schematically represented in Figure 2.2.3. Donor group (D) is excited by a photon and then

relaxes to the lowest excited singlet state, S1 (by Kasha’s rule). If the acceptor group is not too

far, the energy released when the electron returns to the ground state S0 may simultaneously

excite the donor group. This non-radiative process is referred to as “resonance”. After

excitation, the excited acceptor returns to the ground state either through a radiative or non-

radiative decay.

For the FRET to happen, overlap of the emission spectrum of the donor and absorption spectrum

of the acceptor (Figure 2.2.3 right), is important. This means that the energy lost from excited

donor to ground state could excite the acceptor group.


2.2.2 Dexter energy transfer

This interaction occurs between a donor D and an acceptor A. The excited donor has an electron

in the LUMO, and this electron is transferred to the acceptor. The acceptor then transfers an

electron from the HOMO back to the HOMO of the donor, so the acceptor is left in an excited

state (Figure 2.2.4). Electron exchange is similar to FRET because energy is transferred to an

acceptor.

Dexter energy transfer is therefore a process where the donor and the acceptor exchange their

electrons. Hence, besides the overlap of emission spectra of D and absorption spectra of A, the

exchange in energy transfer needs the overlap of wavefunctions. In the other words, it needs

the overlap of the electron cloud. The overlap of wavefunctions also implies that the excited

donor and ground-state acceptor should be close enough so the exchange could happen.

2.2.3 Photoinduced electron transfer

Photoinduced electron transfer has been extensively studied to understand quenching and to

develop photovoltaic devices. In photoinduced electron transfer (PET), after excitation of the

fluorophore, the electron can jump from the LUMO of the fluorophore to the LUMO of the

quencher, or from the HOMO of the quencher to the HOMO of the fluorophore. In both cases

a complex is formed between the electron donor D and the electron acceptor A, yielding the

Figure 2.2.3 FRET Jablonski diagram representation on the left, and overlapping requirement on the

right.

Figure 2.2.4 Schematic representation of Dexter energy transfer.


charge-transfer complex [D+ A– ]*. This complex may emit as an exciplex or be quenched and

return to the ground state (Figure 2.2.5).

The energy change for PET is given by the Rehm-Weller equation:

𝛥𝐺 = 𝐸𝑟𝑒𝑑(𝐷+ 𝐷⁄ ) − 𝐸𝑟𝑒𝑑(𝐴 𝐴−⁄ ) − 𝛥𝐺00 − 𝑒2

𝜀𝑑 (2.5)

In this equation the reduction potential Ered (D+/D) describes the process

𝐷+ + 𝑒 → 𝐷 (2.6)

and the reduction potential Ered (A/A–) describes the process

𝐴 + 𝑒 → 𝐴− (2.7)

ΔG00 is the energy of the S0 → S1 transition of the fluorophore, which can be either D or A.

The last term on the right is the coulombic attraction energy experienced by the ion pair

following the electron transfer reaction, ε is the dielectric constant of the solvent, and d is the

distance between the charges. This term is taken into account only in case that the radical anion

and cation separate.

For the charge-transfer complex [D+ A– ]* to happen, there are two possible ways, either the

quencher diffuses to the fluorophore during the lifetime of the excited state, or a complex is

formed between the fluorophore and the quencher, and this complex is non fluorescent. The

first process is called collisional quenching, because the quenching occurs only if quencher and

fluorophore collide, and the second static quenching, because no diffusion is needed, given that

Figure 2.2.5 Energy diagram for photoinduced electron transfer. The excited molecule is assumed to be the

electron donor. νF and νE are emission from the fluorophore and exciplex, respectively


a complex between fluorophore and quencher is formed. Nevertheless, in both collisional and

static quenching, fluorophore and quencher must be in contact.

2.2.3.1 Theory of collisional quenching

The processes that occur when exciting a fluorophore are the following:

(2.8)

(2.9)

The number of photons absorbed and emitted by a fluorophore depends respectively on the

concentration of the fluorophore [𝐹] , and of the concentration of excited state of the

fluorophore [𝐹∗]

𝐼𝑎𝑏𝑠 = 𝑘𝑎𝑏𝑠[𝐹] (2.10)

𝐼𝑒𝑚 = 𝑘𝑓[𝐹∗] (2.11)

In the presence of a quencher Q, the following process is added

(2.12)

The experimentally observed rate constant for the quenching reaction kq, also called

bimolecular quenching constant, is equal to γ kd, where γ is the efficiency of quenching:

𝑘𝑐/(𝑘𝑐 + 𝑘−𝑑). When the quenching reaction is completely diffusion-limited (𝑘−𝑑 = 0),

then kq= kd.

The quantum yield in the absence of a quencher 𝜙0 becomes

𝜙0 =𝐼𝑒𝑚0

𝐼𝑎𝑏𝑠0 =

𝑘𝑓[𝐴∗]

𝑘𝑓[𝐴∗] + 𝑘𝑖[𝐴

∗]=

𝑘𝑓

𝑘𝑓 + 𝑘𝑖 (2.13)

In the presence of a quencher the quantum yield results

𝐹∗

𝑘𝑓

𝑘𝑖

𝐹 + ℎ𝜈

𝐹 + ℎ𝑒𝑎𝑡

Radiative process

Internal radiationless process

𝐹∗ + 𝑄

𝑘𝑑

𝑘−𝑑

[𝐹 𝑄]∗ 𝑘𝑐 → 𝐹 + 𝑄 + ℎ𝑒𝑎𝑡

𝐹 + ℎ𝜐 𝐾𝑎𝑏𝑠 → 𝐹∗ Absorption of a photon


𝜙 =𝐼𝑒𝑚𝐼𝑎𝑏𝑠

=𝑘𝑓[𝐴

∗]

𝑘𝑓[𝐴∗] + 𝑘𝑖[𝐴∗] + 𝑘𝑞[𝑄][𝐴∗]=

𝑘𝑓

𝑘𝑓 + 𝑘𝑖+𝑘𝑞[𝑄] (2.14)

Combining equation 2.13 and 2.14 the Stern-Volmer equation is obtained

𝜙0𝜙=𝐼𝑒𝑚0

𝐼𝑒𝑚= 1 + 𝑘𝑞𝜏0[𝑄] (2.15)

Where 𝜏0 is the lifetime of the fluorophore in the absence of the quencher 𝜏0 = 1/(𝑘𝑓 + 𝑘𝑖),

and 𝐼𝑒𝑚0 and 𝐼𝑒𝑚 are the fluorescence intensity in the absence and in the presence of a quencher

respectively.

The Stern-Volmer equation, can be expressed in terms of lifetime. If we consider the lifetime

of the fluorophore alone 𝜏0 = 1/(𝑘𝑓 + 𝑘𝑖), against the lifetime of the fluorophore in the

presence of the quencher 𝜏 = 1/(𝑘𝑓 + 𝑘𝑖 + 𝑘𝑞[𝑄])

𝐼𝑒𝑚0

𝐼𝑒𝑚=𝜏0𝜏

(2.16)

2.2.3.2 Theory of static quenching

In the case of static quenching, the formation of a non-fluorescent ground-state complex

between the fluorophore and the quencher occurs (equation 2.16 and 2.17).

𝐹 + 𝑄

𝐾𝑠

[𝐹𝑄] (2.17)

𝐾𝑠 =[𝐹𝑄]

[𝐹][𝑄] (2.18)

When this complex absorbs light, immediately returns to the ground state without emission of

a photon (equation 2.18).

[𝐹 𝑄]∗ 𝑘𝑐 → [𝐹 𝑄] + ℎ𝑒𝑎𝑡 (2.19)

As for dynamic quenching, the Stern-Volmer can be derived and results in

𝐼𝑒𝑚0

𝐼𝑒𝑚= 1 + 𝐾𝑠[𝑄] (2.20)


Since the effect of the formation of a complex simply reduces the concentration of the

fluorophore, the lifetime of the excited fluorophore doesn’t change in the presence or absence

of a quencher: 𝜏 = 𝜏0.

The measurement of the lifetime is a definitive method to distinguish static and dynamic

quenching.

2.2.3.3 Combined static and dynamic quenching

In real experiments, often happened that collisional and static quenching occur at the same time.

In this case the quenching effect will be greater of both the mechanisms described above, indeed

the fraction of quencher that is not bounded to the fluorophore can still quench the fluorophore

by dynamic quenching.

The modified Stern-Volmer can be obtained by multiplying: the ratio of fluorescents in the case

of static quenching, by the ratio of fluorescents in the case of collisional quenching. This yield

a second order equation in [Q].

𝐹0𝐹= (1 + 𝑘𝑞𝜏0[𝑄])(1 + 𝐾𝑠[𝑄]) (2.21)

3 Chapter 3 3

Ultrafast Transient Absorption

Spectroscopy

Ultrafast transient absorption (TA) spectroscopy is a wide spread technique that permits to

investigate ultrafast processes. This technique provides a large amount of information regarding

the dynamics and nature of photo-induced processes, such as chemical reaction, conformational

change, energy and electron transfer and the like, both in solution and in solid state.

3.1 Principle of transient absorption spectroscopy

Ultrafast TA experiments involve two femtosecond laser pulses, a pump and a probe. The

monochromatic pump pulse, which goes through a certain volume of the sample, is resonant

with a transition of the photosystem of interest, and is used to trigger the studied photoreaction.

Thus is induced a vertical Franck Condon transition to the excited state, of a certain amount of

molecules (usually a few percent, depending on the pump power and absorption cross section

of the molecule). The probe is a weak femtosecond white laser pulse of variable wavelength,

which reaches the same volume of the sample hit by the pump after a certain delay. The light

of the probe not absorbed by the sample is diffracted on a grating and collected by a detector

(photodiode array detector or CCD), thus the absorption spectra at a certain delay is obtained

(Figure 3.1).

Figure 3.1 Schematic depiction of transient absorption spectroscopy principle. On the right,

transient absorption spectroscopy applied to a simple reaction.

18 3. Ultrafast Transient Absorption Spectroscopy

For each time delay, absorption spectra will before and after the pump, be registered. In this

way, the difference of absorption spectra before and after excitation is calculated, to emphasize

the signal variation (equation 1.1).

∆𝐴 = 𝐴𝑡 − 𝐴0 (3.1)

The measurement 𝛥𝐴(𝜆, 𝑡) is the sum of four single contributions from different physical

phenomena, as shown in Figure 3.2.

The first contribution is given by ground-state bleach. As a consequence of the excitation of

the molecule, from the ground state to the excited state by the probe, the number of molecules

in the ground state is decreased. Hence the absorption of the ground state after excitation is

lower than the one before excitation. Consequentially, a negative 𝛥𝐴(𝜆, 𝑡) contribution is given

in the wavelength region where the ground state absorbs.

The second contribution is by stimulated emission. A photon from the probe pulse can induce

the emission of a photon, with the same phase and direction of the incident photon, from an

excited molecule, which returns to the ground state. This phenomena will result in an increase

in light intensity on the detector, thus the 𝛥𝐴(𝜆, 𝑡) contribution for stimulated emission will be

negative. The stimulated emission’s spectral profile will follow more or less, the same emission

spectra of the fluorophore. The peak of stimulated emission, thus, will be Stoke-shifted with

respect to the ground state bleach.

The third contribution is provided by excited-state absorption. Upon excitation with the probe

beam, optically allowed transitions from the excited state of the Chromophore, to higher excited

states may be possible in a certain wavelength region, and absorption of the probe beam in this

wavelength region will occur. In this case a positive 𝛥𝐴(𝜆, 𝑡) contribution is given in the spectra

region where the excited state absorbs.

A fourth contribution is possible if the pump triggers a reaction, therefore the 𝛥𝐴(𝜆, 𝑡) is given

by product absorption. The absorption of the product results in a positive contribution in the

wavelength region where the product is absorbed. As a consequence of this phenomena, a

further ground state bleach will be observed.

Note that the intensity of the probe beam is so weak that the excited-state population is not

appreciably affected by excited-state absorption and stimulated emission.

As shown in the figure below, the spectra of the single contributions are most likely overlapped.

3. Ultrafast Transient Absorption Spectroscopy 19

3.2 Experimental setup

In Figure 3.3 is shown a typical scheme of ultrafast absorption spectroscopy setup.

A laser pulse is generated by an oscillator, and then amplified by a regenerative amplifier

(USP). The output from the laser system, in the considered setup, is a 40 fs pulse at an energy

of 2.5 mJ, centered on the 800 nm wavelength, with a bandwidth of 30 nm, and a repetition rate

of 1 KHz. In order to be able to shift the wavelength, is used an optical parametric amplifier or

generator coupled with non-linear mixing processes such as frequency-doubling, sum-

frequency generation and difference frequency generation. Thus a broad range pulse can be

now obtained, from the UV to mid-IR. This pulse, that is the pump pulse, is sent through an

optical delay line, which consist of a retroreflector mounted on a high precision motorized

computer-controlled translation stage. A 1 μm shift of the retroreflector correspond to 6.7 fs of

delay. The pump beam is focused in the sample to a diameter of 130-200 μm and blocked after

the sample.

Regarding the probe beam, a part of the pump beam is deflected after the 800 nm beam

amplification. The deflected beam is focused on a Calcium Fluoride plate (Magnesium

Fluoride, quartz, water and ethylene glycol can be used as well) to generate the white light

probe pulse ranging from ~400 to ~1100 nm . This beam is therefore focused in the sample to

a diameter slightly smaller than the pump, and overlapped with it. It worth noting again that the

intensity of the incident probe beam is so weak that doesn’t change appreciably the population

of the excited state. The light that pass through the sample is than collimated on a grating, where

is diffracted toward an array diode or a CCD detector. The diode or the CCD is than read by a

computer on a shot-to-shot basis, thus a whole absorption spectrum is measured with each shot.

Frequently a reference beam is used to increase the signal-to-noise ratio taking in account the

probe light intensity fluctuation. In such a case, the white beam is split in two, the probe beam

and the reference one. The reference beam is than collected by a detector either after passing

through the sample or not.

Figure 3.2 Contributions to ΔA spectrum. On the right panel: ground state bleach (dashed line),

stimulated emission (dotted line) excited state absorption (thin solid line), sum of these contribution

(thick solid line).

20 3. Ultrafast Transient Absorption Spectroscopy

For measurement in solution at room temperature the sample is placed in a 1-2 mm quartz

cuvette which is either stirred or connected to a flow system, to prevent exposure of the same

excited volume of sample to consecutive excitation.

By the nature of the white light generation the “blue” wavelengths are generated later than the

“white” one. Hence the white light beam is generated with an intrinsic group-velocity

dispersion, which also increases passing through optical dense material such as lenses and

cuvette. This velocity dispersion must be taken in account during data analysis, and will result

as a shift to higher delay time of the “blue” wavelengths, or can be fixed by compression of the

white-beam trough a grating pair or a prism pair.

A transient absorption experiment proceed as follows. The pump beam before reaching the

sample pass through a mechanical chopper, which block the pulse every other time. In such a

way for every shot is measured the absorption spectrum before 𝐴0 and after excitation 𝐴 𝑡 at a

certain delay time. A number of shot that is sufficient for an acceptable signal-to-noise ratio is

measured for the fixed delay time, usually 103 - 104.

The average difference ∆𝐴 = 𝐴(𝑡) − 𝐴0 is thus calculated at this delay time, and then the delay

is increased and the above procedure is repeated. In this way an entire dataset 𝛥𝐴(𝜆, 𝑡) is

collected.

Figure 3.3 Schematic representation of an experimental transient absorption setup

4 Chapter 4 3 Synthesis and Characterization of

Ag29(BDT)12(TTP)4

4.1 Synthesis

Ag29(BDT)12(TPP)4 nanoclusters (NCs) were prepared by dissolving silver nitrate in a solvent

solution of methanol and dichloromethane prior to the addition of 1,3-Benzenedithiol (BDT)

ligands. The solution turned turbid with insoluble yellow flakes, indicating the formation of a

Ag-S complex. Triphenylphosphine (TPP) was dissolved in dichloromethane and introduced

to the reaction vial immediately after mixing the silver salt with BDT. The yellow flakes

disappeared immediately and the solution turned clear. The reaction mixture was then reduced

with an aqueous solution of NaBH4, and the resulting dark brown solution turned dark orange

during 5-7 h of stirring. Figure 4.1 shows detailed pictures of the reaction vials throughout the

reaction. To purify the product, the solution was centrifuged at 9000 rpm: the product consisted

of a dark brown pellet which was then dried under vacuum. The so obtained clusters where than

dissolved and filtered. The purified NCs showed high solubility in various aprotic polar

solvents, including DMF and DMSO, and fair solubility in less polar solvents such as

acetonitrile and dichloromethane.

Figure 4.1 Synthesis of [Ag29(BDT)12(TTP)4]-3 nanoclusters.

22 4. Synthesis and Characterization of Ag29(BDT)12(TTP)4

4.2 Characterization

4.2.1 Mass spectroscopy and analytical ultracentrifugation

Negative ion mode electrospray ionization mass spectroscopy (ESI-MS) of the NCs in

acetonitrile was performed, and analyzing the spectra (Figure 4.2) was postulate that this NC

has a full molecular formula of [Ag29(BDT)12(TTP)4]-3, it follows the electron count rule of the

superatom theory, with an electron count of n = 29-24 + 3 = 8, corresponding to a stable

superatom with the Aufbau shell filling 1S2|1P6|

To rule out the existence of any other species, and to confirm the purity of the samples used for

the measurement of the optical properties, was used analytical ultracentrifugation (AUC), a

potent technique to determine the homogeneity of macromolecules and nanoclusters in

solutions. The sedimentation and diffusion distributions of the synthesized NCs in acetonitrile

are shown in Figure 4.3. The distributions show that the NCs are highly homogeneous; at least

97% of the sample is composed of one species whose sedimentation coefficient is 2.9 × 10-13 s.

The molecular weight corresponding to this most abundant species is 5381.49 Da which is in

very good agreement with the mass spec assignment of Ag29(BDT)12(TPP)4.

Figure 4.2 Negative ion mode ESI MS of [Ag29(BDT)12TPP4] indicating the presence of one species

only with a charged state of -3. Phosphines are lost during ionization (top panel). Exact match of

experimental (one of the five sets of peaks) and simulated mass spectra (bottom panel) of

[Ag29(BDT)12TPP4]3- confirms the cluster composition to be [Ag29(BDT)12TPP4]3-.

4. Synthesis and Characterization of Ag29(BDT)12(TTP)4 23

4.2.2 X-ray diffraction

For crystallization, the centrifuged NCs pellet was dispersed in DMF, filtered using a syringe

filter and left to evaporate slowly in a dark box inside a ventilated fuming hood. Within 1- 2

days, self-assembled supramolecular structures had formed, as shown in Figure 4.4. They were

obtained by drop casting onto a glass microscope slide from a concentrated stock solution;

fluorescent crystals suitable for X-ray diffraction were harvested. DMF was used as a dispersing

solvent because of its high boiling point and slow evaporation time, which increased the

tendency of the NCs to assemble into a large solid with a long-range order.

Figure 4.3 2D plot of sedimentation and diffusion coefficients of Ag29(BDT)12TPP4 .


Single crystal X-ray diffraction analysis revealed a core-shell NC with an overall composition

of Ag29(BDT)12(TPP)4, which crystallizes in a cubic Pa3-space group. The structure was refined

to a resolution of 1.1 Å and to an R1 value of 8.9%. Ag29(BDT)12(TPP)4 features a centered

icosahedral metal core (Figure 4.5a), similar to the well-known Au25 and the most recently

discovered Au133. An exterior shell (Figure 4.5Figure 4.5b) composed of the remaining 16 Ag

atoms caps the core. The crystal structure reveals two types of silver atoms in the shell. Twelve

silver atoms cap all the 12 atoms of the icosahedron, giving rise to four tetrahedrally oriented

trigonal prisms as shown in Figure 4.5 c. The remaining four Ag atoms face-cap the core at four

tetrahedral positions (Figure 4.5d). Starting from the center of the icosahedron, the radial bond

lengths give rise to an average of 2.77 ± 0.01 Å per Ag-Ag bond. The average length of the

peripheral Ag-Ag bonds is 2.92 ± 0.06 Å, comparable to the 2.88 Å bond length in bulk silver,

indicating a strong interaction between the atoms of the core.

Figure 4.5 Anatomy of the structure of Ag29(BDT)12(TPP)4 showing the core−shell configuration

and the position of the Ag atoms: (a)Ag13 centered icosahedral core; (b) Ag16S24P4 shell; (c)

arrangement of 12 Ag 1atoms of the shell forming 4 trigonal prisms tetrahedrally oriented; (d)

tetravalent sites of the NC. Color labels: Ag, blue and navy blue; S, red; P, green; all C and H atoms

are omitted for clarity.

Figure 4.4 Optical microscopy image of self-assembled Ag29(BDT)12TPP4 NCs. Inset shows separate

rhombohedral single crystals.

4. Synthesis and Characterization of Ag29(BDT)12(TTP)4 25

The shell is composed of two motifs unique to Ag29(BDT)12(TPP)4: (i) a Ag3S6 crown motif

(Figure 4.6b) where three S atoms connect the three Ag atoms of the crown in such a way that

they form an alternating chair configuration and the remaining three S atoms encapsulate the

underlying icosahedron face (Figure 4.6c); (ii) a Ag1S3P1 motif where the S atoms connect the

Ag atoms to the nearest Ag atoms and the P binds on the top site of Ag atoms (Figure 4.6d).

Figure 4.6e shows that the shell composed of four Ag3S6 and four Ag1S3P1 motifs provides

complete passivation of the NC.

Figure 4.7 shows how the shell is formed around the Ag13 core. Starting from the core (Figure

4.7a) outward, one S moiety of the BDT ligand is attached to each of the 12 Ag atoms of the

icosahedron (Figure 4.7b). These S atoms bridge the core atoms to the Ag atoms in the shell.

The second S moiety bridges Ag atoms in the shell (Figure 4.7c). The overall core-shell

structure is then shown in Figure 4.7d highlighting two pairs of sulfurs to show which pair of

sulfurs originates from a single BDT molecule.

Figure 4.6 X-ray crystal structure of Ag29(BDT)12(TPP)4 highlighting the two motifs present in the

shell: (a) Ag13 centered icosahedral ore; (b) Ag12S24 shell made of 4 Ag3S6 crowns; (c) Ag25S24 motif,

where the four Ag3S6 crowns capping the core; (d) 4 Ag1S3P1 motifs; (e) total structure of

Ag29(BDT)12(TPP)4. Color labels: Ag, blue; S, red; P, green; all C and H atoms are omitted for

clarity.

Figure 4.7 X-ray crystal structure of Ag29(BDT)12TPP4 highlighting the formation of the shell on the

Ag13 core. Color labels: Ag: blue; S: red & yellow; P: green; C: black. Most C and all H atoms are

omitted for clarity. In d, only two ligands are shown for clarity.


Figure 4.8 UV−vis absorbance (solid curves) and emission (dashed) of Ag29(BDT)12(TPP)4 NCs in

acetonitrile (black) and dried (red).

The arrangement of all the Ag atoms in the shell are influenced by the particular spacing

between the two thiol groups of the ligand in addition to the high tendency of S to coordinate

with Ag forcing the benzene rings to bend in such a way that all the S atoms of the bidentate

ligand would coordinate to the Ag. Ag29(BDT)12(TPP)4 is by far the first molecular NC where

the underlying geometry is highly affected by the structure of the ligand. All attempts to make

the NCs with similar bidentate ligands with different spacing between the two thiol groups, for

example, 1,2-benzenedithiol and 1,4- benzenedithiols, failed to produce NCs stable enough for

a period of time to carry any meaningful characterization, which shows how crucial is the

distance between the two thiols in obtaining this tetravalent NC.

4.2.3 Optical properties

Figure 4.8 shows the absorption and emission spectra of Ag29(BDT)12(TPP)4 in solution and

as a crystallized film. Upon crystallization, two main features were observed: an overall

increase and broadening of the long wavelength band of absorption and a red shift of the

emission band by more than 50 nm. The broadening and minute red shift of the absorption band

are explained in terms of electronic coupling between the NCs via interaction between the

transition dipole moment of the individual absorbing Ag29(BDT)12(TPP)4 NC and the induced

dipole moments in the neighboring Ag29(BDT)12(TPP)4 NCs. This interaction is thought to

lower the initial transition energy. The red shift of the emission band is expected to be caused

by a combined effect of the electronic coupling quoted before and of lattice-origin, nonradiative

decay pathways occurring through electron−phonon interaction that lower the emission energy

and also slightly broaden the emission bands. It is important to stress that when

Ag29(BDT)12(TPP)4 NCs are assembled into a crystal, a proper lattice dynamics of the

superstructure, not present in isolated NCs, is generated.

5 Chapter 5 3

Photoinduced electron transfer

A new material that can be excited by light, is interesting for photovoltaic applications if it can

give rise to photo-induced electron transfer, that is, if can donate his excited electron to another

molecule. To achieve high light-to-energy conversion efficiency of solar cell devices, rapid

electron injection at donor-acceptor (D-A) interfaces is a highly desirable dynamical process.

Overall electron transfer efficiency at D-A interfaces is dependent on the distance, energy level

alignment, redox potentials and electrons coupling between electron donor and acceptor

moieties.

For probing the suitability of Ag29 nanoclusters (expected to be electron donors) as a material

for solar applications, the interaction of the clusters with some of the most common electron

acceptors used in photovoltaic was studied.

5.1 Steady-state absorption and fluorescence

A simple way to verify whether electron transfer may occur or not, is to measure the emission

spectra of the clusters in the absence and presence of the electron acceptor. If the peak of

emission is significantly decreased after adding the quencher (which is expected to be the

electron acceptor), the process which competes with radiative emission occurs while an excited

Ag29 nanoclusters (NCs) relaxes to the ground state. This process could be either electron

transfer or energy transfer, and can’t be distinguished only through a fluorescence quenching

study.

To carry out the experiment, a fixed concentration of Ag29 was dissolved in dimethylformamide

(DMF), and absorption and emission spectra where measured consequentially without and with

the addition of a significant amount of quencher (electron acceptor). The concentration of Ag29

being fixed, means that if there is no interaction between the clusters and the added molecules,

then no change in emission peak should be observed. However if the emission peak is found to

be decreased, there is a possibility of an interaction between the clusters and the quencher, or,

that the quencher absorbs part of the light used to excite the cluster. Thus, before drawing any

conclusion, one must consider the absorption change of the solution (before and after adding

the quencher) at the wavelength used to excite the sample.

The absorption spectra of Ag29 shows a shoulder at 510 nm and the peak at 450 nm. Thus the

excitation wavelength was set at 450 nm in order to maximize the excitation.

28 5. Photoinduced electron transfer

With ZnO nanoparticles and TiO2 nanoparticles, no change in the emission peaks were noticed.

With C60-(malonic acid)n, a functionalized fullerene, there was a slight decrease in the emission

peak. Considering that the clusters are -3 charged, was thought to exploit the Coulombic

interaction between negatively charged nanoclusters and positively charged fullerene to bring

them closer and give rise to a better interaction. The positively charged fullerene used is C60-

(N,N-dimethylpyrrolidinium iodide)+n . For both the fullerenes 𝑛 ≅ 3.

Fullerenes are considered to be strong electron acceptors with a relatively high electron affinity

and are generally used in high-performance bulk-heterojunction polymer solar cell devices.

As expected, the interaction between positively charged fullerene and Ag29 leads to a stronger

quenching.

Emission and absorption spectra for Ag29 plus positively charged fullerene were measured for

different concentrations of fullerene (and fixed concentrations of the clusters). Figure 5.2 shows

that the emission peak of nanoclusters at 645 nm is quenched by the fullerene up to 80%. It

cannot reach 100% quenching, because the positively charged fullerene emits, (emission peak

595 nm), thus the emission of Ag29 is summed to the one of the fullerene. The fullerene

obviously absorbs at 450 nm (the excitation wavelength), so the absorbed amount of light by

the nanoclusters is reduced with respect to the clusters alone, however the quenching observed

is strong, and can’t be ascribed just to the absorption of the fullerene.

For the same concentrations of the neutral fullerene the emission peak doesn’t show any

variation. This suggests that the Coulombic interaction between the clusters and fullerene plays

an important role in the process involved.

Figure 5.1 Functionalized fullerenes: on the left the positively charged (cationic) one, on the right the

neutral one.

5. Photoinduced electron transfer 29

To answer the question whether the quenching could be ascribed to electron transfer or to

energy transfer, it must be taken into account that for energy transfer to happen, the overlap of

the emission spectrum of the donor, with the absorption spectrum of the acceptor is required.

As shown in Figure 5.3 the absorption spectrum of C60-(N,N-dimethylpyrrolidinium iodide)+n

and the emission spectrum of Ag29 NCs do not show any significant overlap, so it can be ruled

out that energy transfer happened, therefore, can be claimed with high confidence that electron

transfer occurs between the clusters and the cationic fullerene, and is responsible for quenching.

Figure 5.2 Absorption and emission of the Ag29 clusters with the two fullerenes, the dashed line shows

absorption and emission of each fullerene.


HOMO and LUMO where measured both for the clusters and the positively charged fullerene

using Photoelectron spectroscopy in air (PESA) to measure the HOMO, and the Tauc plot (a

convenient way of displaying the absorption spectrum) for determining the optical bandgap.

The results where then confirmed with cyclic voltammetry for both the fullerene and the

clusters.

As displayed in Figure 5.4 the LUMO of the fullerene results to be lower than the one of the

clusters. This is an essential requirement for electron transfer to happen, because the electrons

in passing from the LUMO of the donor to the LUMO of the acceptor, cannot gain energy.

Looking carefully to the absorption peak of Ag29 in presence of the positively charged fullerene,

it can be noted that, in increasing the concentration of the fullerene, the absorption peak is red-

Figure 5.3 Normalized absorption and emission spectra of respectively the positively charged

fullerene and Ag29 nanoclusters

Figure 5.4 HOMO and LUMO of Ag29 and C60-(N,N-dimethylpyrrolidinium iodide)+n


shifted. This, together with the fact that the Ag29 solution slightly change color (compared the

clusters and the fullerene alone) after adding the fullerene, suggests that between the clusters

and the fullerene a complex is formed.

If the fullerene forms a complex with the clusters, the expected quenching mechanism is static.

From the Stern-Volmer plot (Figure 5.5) the trend of 𝐼0/𝐼 with the concentration looks like

quadratic, which would represent the combined static and dynamic quenching. But this result

is in contradiction with the lifetime measurements, showing no change in lifetime of the clusters

with or without the fullerene, depicting a static quenching. The discrepancy can be explained

considering the fullerene absorbs light at 450 nm (the excitation wavelenght), thus the absorbed

light from the clusters 𝐼0 will be lower and so 1/𝐼 will be greater.

The Stern-Volmer plot is therefore distort by the fact that 𝐼0 is not constant, therefore the

quenching, according to the lifetimes, can be considered static.

5.2 Femtosecond transient absorption

Ultrafast TA spectroscopy with femtosecond temporal resolution and broadband capabilities

was used to probe further the reaction mechanism of the excited-state interaction.

Figure 5.7 shows TA spectra of a solution of Ag29 in DMF, in the absence and in the presence

of C60-(N,N-dimethylpyrrolidinium iodide)+n and C60-(malonic acid)n . As can be seen,

excitation of the clusters alone immediately results in ground-state bleaching (GSB) at around

445 nm, and two excited-state absorption (ESA) peaks centered at 410 nm and 480 nm

respectively, and no decay is shown in a time window of 200 ps. No change in the dynamics of

either GSB or ESA within a few hundred ps-time delay indicates slow relaxation dynamics,

thus high excited state lifetime.

Figure 5.5 Stern-Volmer plot for different concentration of C60-(N,N-dimethylpyrrolidinium iodide)+n

in a Ag29 nanoclusters solution.


When exciting the cluster in the presence of C60-(N,N-dimethylpyrrolidinium iodide)+n after

few picoseconds, can be noted a slight recover of the ground-state absorption peak, and the

decrease of the two excited-state absorption peaks. A single exponential fit to the data (Figure

5.7 right panel) suggests that the kinetics of the GSB recovery of the Ag29 NCs with cationic

fullerene has a time costant of 3.9 ± 0.4 ps. The relaxation of excited Ag29 to the ground state,

from few picoseconds to 200 ps after excitation, is due to the electron transfer process between

the clusters and the charged fullerene. To confirm the process of electron transfer between NCs

and cationic fullerene, fs-TA spectra of Ag29 in the presence of C60-(N,N-

dimethylpyrrolidinium iodide)+n was measured in the infra-red (IR) range (Figure 5.9). The

appearance of a new peak around 850 nm was assigned to the formation of the fullerene radical,

clearly proving the donor-acceptor relationship between anionic NCs and cationic fullerene.

The TA spectra of Ag29 with C60-(malonic acid)n instead, is exactly the same of the clusters

alone, ruling out, at this concentration, any significant interaction between the clusters and the

neutral fullerene.

Excitation wavelength was set to 530 nm to avoid the fullerene excitation.

Figure 5.6 Example of a 3D time resolved absorption spectra for the solution Ag29 plus C60-(N,N-

dimethylpyrrolidinium iodide)+n .


To fully characterize the dynamics of the clusters after excitation, the time window for the TA

measurements was extended to the nanosecond time scale. Figure 5.8 shows ns-TA spectra of

Ag29 NCs with charged and neutral fullerene, at time delays up to a few hundred ns, revealing

slow GSB recovery and ESA decay. The kinetics of GSB recovery and ESA decay of Ag29 NCs

fit a single exponential function with a time constant of 120±10 ns, similar time constants were

also observed in the absence and presence of electron transfer. These results indicate that

dynamics in the ns time scale are related to excited free-standing or uncomplexed Ag29 NCs

even when fullerene derivatives are present in the solution. This observation confirms that

electron transfer is driven by electrostatic interactions and that it is static in nature.

Figure 5.7 On the left panel: Transient absorption spectra using photoexcitation at 530 nm for (A)

Ag29 alone, (B) Ag29 in the presence of C60-(N,N-dimethylpyrrolidinium iodide)+n 71.4 μM, and (C)

Ag29 in the presence of 71.4 μM C60-(malonic acid)n .

On the right panel: Kinetic traces derived from femtosecond transient absorption spectra for Ag29 in

the absence and in the presence of the two 71.4 μM fullerenes. On top: kinetic of the excited-state peak

(ESA). At the bottom: kinetic of the ground-state bleaching (GSB). Fitting is indicated by red lines.


To rule out any doubt on the fact that the high interaction of the cationic fullerene with the NCs

were due to coulumbic interaction, and not by other interactions occurring between that

particular fullerene and the Ag29 NCs, the experiment presented above was carried out using

anionic carboxyl-capped CdTe quantum dots instead of Ag29 NCs. The result was exactly the

same: for the neutral fullerene no significant interaction is detected, while for the cationic

fullerene it’s shown that TA shows the dynamic of electron transfer.

Figure 5.8 Ns-TA spectra of Ag29 NCs (A) in the absence of fullerene and (B) in the presence of 71.4

μM C60-(N,N-dimethylpyrrolidinium iodide)+n. (C) The kinetic traces of Ag29 NCs in the absence and

presence of C60-(N,N-dimethylpyrrolidinium iodide)+n. Solid lines represent the best kinetic fit of the

data.

Figure 5.9 Femtosecond Transient Absorption spectra in the IR region of Ag29 NCs in the presence

of 71.4 C60-(N,N-dimethylpyrrolidinium iodide)+n µM. The peak at 850 is attributed to the radical of

the fullerene, resulting from the electron transfer. Excitation at 610 nm.


In conclusion, it was demonstrated that interfacial electrostatic interactions has a significant

influence on electron transfer processes, from excited anionic Ag29 NC (and CdTe QD) surface

to cationic fullerene, facilitating the donor-acceptor nanoassembly formation. Steady-state

absorption and fluorescence measurements were used to prove that the system shows strong

electrostatic interactions in the presence of cationic fullerene but not in the presence of neutral

fullerene. Furthermore, fs-TA spectroscopic data clearly demonstrates that the time constant of

electron transfer from excited Ag29 NCs to cationic fullerene occurs within the instrument

temporal resolution of 120 fs, which is much faster than the excited lifetimes of NCs. Thus,

electrostatic interactions at D-A interfaces may be useful for preserving rapid electron transfer

processes upon photoexcitation, which is among the key components in determining the overall

efficiency in both photovoltaic and photo-catalysis applications.

5.2.1 Instrumentation for time-resolved measurement

Ultrafast fs-transient absorption measurements were performed using a Helios UV-NIR

transient absorption spectrometer system provided by ultrafast systems. Helios is equipped with

CMOS VIS and InGaAs NIR spectrometers covering a range of 350-800 nm with a 1.5-nm

resolution at 9500 spectra and a range of 800-1600 nm with a 3.5-nm resolution at 7900 spectra,

respectively. The fundamental output delivers by a Spitfire Pro 35fs-XP regenerative fs

amplifier, which produces 35-fs pulses at 800 nm with 4 mJ/pulse and a repetition rate of 1

kHz. The white-light continuum probe beam is generated by focusing a few µJ pulse energy of

the fundamental beams onto a 2-mm-thick sapphire plate. The spectrally tunable fs pulses with

a few µJ energy are generated in an optical parametric amplifier (Spectra-Physics). Note that

Figure 5.10 Left panel: Absorption and fluorescence spectra after excitation of CdTe QDs at 565 nm

upon the successive addition of (A) cationic fullerene C,F and (B) neutral fullerene NF.

Right panel: Fs-TA spectra of CdTe QDs (A) in the absence of fullerene and (B) in the presence of

27.32 μM cationic fullerene CF.


the solution was stirred constantly to keep a fresh sample volume for each laser shot and to

avoid photodegradation. The pump and probe beams were overlapped spatially and temporally

on the sample solution, and the transmitted probe light from the samples was collected and

focused on the broad-band UV-visible-NIR detectors to record the time-resolved excitation-

induced difference spectra.

To measure the transient spectra from ns to µs time delays, an EOS from an ultrafast system

with a time resolution of 200 ps and a detection limit of up to 400 µs was also used. The EOS

light source is coupled with an fs-TA spectrometer, which is used as probe beam. In both Helios

and EOS, a two-channel probe-reference method was used to split the probe beam in two: one

beam travels through sample and the other is sent directly to the reference spectrometer that

monitors fluctuations in the intensity of the probe beam. The pump was used by introducing the

fundamental beams into an optical parametric amplifier to select a certain wavelength from the

tunable output (240-2600 nm). Samples were measured in DMF solutions (aqueous for the

QDs) at room temperature. Transient absorption spectra were recorded at low-intensity

excitation to prevent Auger recombination of the photogenerated charge carriers.

Conclusions

The aim of this thesis is to investigate if the new synthetized [Ag29(BDT)12(TTP)4]-3 NCs could

act as an electron donor, giving rise to photoinduced electron transfer in the presence of an

electron acceptor.

In this work were accomplished the bottom-up synthesis of a new noble metal nanocluster:

[Ag29(BDT)12(TTP)4]-3. After characterization of this NC trough electron spry ionization mass

spectroscopy and X-ray diffraction, its properties as an electron donor were investigated.

Among the different electron acceptors tried, a fullerene derivative, commonly used as electron

acceptor in solar cells, turns out to show a slight interaction with the NCs that could be ascribed

to electron transfer (evidence from steady state fluorescence measurements). Because the

clusters are negatively charged, it was then probed whether a cationic fullerene derivative would

have led to stronger interaction with the NCs. A strong quenching of fluorescence was indeed

observed using the cationic fullerene, giving the hint that coulumbic interaction might play an

important role in electron transfer.

Femtosecond transient absorption measurements were run on the following systems in

solutions: (i) Ag29 NCs; (ii) Ag29 NCs and cationic fullerene derivative; (iii) Ag29 NCs and

neutral fullerene. In the picosecond time window, for the (i) and (iii) system, no change of the

GSB and ESA peaks were observed. While for the (ii) system, was observed a ps-GSB peak

recovery, and the ESA peak decrease, a typical signature of electron transfer.

However fluorescence quenching, and the ps-GSB recovery, can both be due to energy transfer,

and electron transfer. Nevertheless, the absence of the donor-florescence and acceptor-

absorption spectra overlapping, together with the fact that the peak of the radical fullerene in

TA measurements was observed, ruled out any possibility of energy transfer. Thus the process

involved between Ag29 NCs and cationic fullerene, after excitation, is certainly ascribed to

electron transfer.

To rule out any doubt on the fact that the high interaction of the cationic fullerene with the NCs

were due to coulumbic interaction, and not by other interactions occurring between that

particular fullerene and the Ag29 NCs, steady state and Transient experiments were carried out,

using anionic carboxyl-capped CdTe quantum dots instead of Ag29 NCs. The result was exactly

the same: for the neutral fullerene, no significant interaction was detected, while the cationic

fullerene TA shows the dynamic of the electron transfer.

In conclusion, studying the photo-induced processes of the new synthetized silver nanocluster,

the importance of Coulombic interaction in the electron transfer (process at the basis of a solar

cell), was discovered. This electrostatic interaction at the D-A interfaces, may be useful for

gaining rapid electron transfer process upon photoexcitation, which is among the key

components in determining the overall efficiency of both photovoltaic and photocatalysis

applications.

References

1. Chaar, L. El, A, L. & Zein, N. El. Review of photovoltaic technologies. Renewable

Sustainable Energy Rev. 15, 2165–2175 (2011).

2. Chen, Y., Choi, H. & Kamat, P. V. Metal-Cluster-Sensitized Solar Cells. A New Class

of Thiolated Gold Sensitizers Delivering E ffi ciency Greater Than 2%. (2013).

3. Razykov, T. M., Ferekides, C. S., Morel, D., Stefanakos, E. & Ullal, H. S. Solar

photovoltaic electricity : Current status and future prospects. Sol. Energy 85, 1580–

1608 (2011).

4. Choi, H., Chen, Y., Stamplecoskie, K. G. & Kamat, P. V. Boosting the Photovoltage

of Dye-Sensitized Solar Cells with. (2015). doi:10.1021/jz502485w

5. Kamat, P. V. Meeting the Clean Energy Demand : Nanostructure Architectures for

Solar Energy. 2834–2860 (2007).

6. Andrienko, D. Cyclic Voltammetry (2008).

7. The Oxidation of Ferrocene;  Chem 367-2/Cyclic Voltammetry

8. Cyclic Voltammetry of Ferrocene, [Ru(bpy) 3 ] 2+ , [Co(bpy) 3 ] 2+ and Iodide. 1–6.

Umass Boston Chem 371.

9. AbdulHalim, L. G. et al. Ag 29 (BDT) 12 (TPP) 4 : A Tetravalent Nanocluster. J. Am.

Chem. Soc. 29, 150624042107008 (2015).

10. Cui, M., Zhao, Y. & Song, Q. Synthesis, optical properties and applications of ultra-

small luminescent gold nanoclusters. TrAC - Trends Anal. Chem. 57, 73–82 (2014).

11. Díez, I. & Ras, R. H. a. Fluorescent silver nanoclusters. Nanoscale 3, 1963–1970

(2011).

12. Hou, W. & Cronin, S. B. A review of surface plasmon resonance-enhanced

photocatalysis. Adv. Funct. Mater. 23, 1612–1619 (2013).

13. Shang, L., Dong, S. & Nienhaus, G. U. Ultra-small fluorescent metal nanoclusters:

Synthesis and biological applications. Nano Today 6, 401–418 (2011).

14. Wilcoxon, J. P. & Abrams, B. L. Synthesis, structure and properties of metal

nanoclusters. Chem. Soc. Rev. 35, 1162–1194 (2006).

15. D. Michael, P. Mingos. Gold Clusters, Colloids and Nanoparticles I. Springer.

16. Walter, M. et al. A unified view of ligand-protected gold clusters as superatom

complexes. Proc. Natl. Acad. Sci. U. S. A. 105, 9157–9162 (2008).

17. Joseph R. Lakowicz, Principles of Fluorescence Spectroscopy, Third Edition. Springer.

18. Thallium, I. Charge Effects on the Dynamic Quenching of Fluorescence of. 88, 255–

264 (1980).

19. Berera, R., van Grondelle, R. & Kennis, J. T. M. Ultrafast transient absorption

spectroscopy: Principles and application to photosynthetic systems. Photosynth. Res.

101, 105–118 (2009).

20. Green, T. D. & Knappenberger, K. L. Relaxation dynamics of Au25L18 nanoclusters

studied by femtosecond time-resolved near infrared transient absorption spectroscopy.

Nanoscale 4, 4111 (2012).

21. Ruckebusch, C., Sliwa, M., Pernot, P., Juan, A. De & Tauler, R. Journal of

Photochemistry and Photobiology C : Photochemistry Reviews Comprehensive data

analysis of femtosecond transient absorption spectra : A review. "Journal Photochem.

Photobiol. C Photochem. Rev. 13, 1–27 (2012).

22. Alsam, A. A. et al. Bimolecular Excited-State Electron Transfer with Surprisingly

Long-Lived Radical Ions. Journal of Physics and Chemistry (2015).

Acknowledgments

I would like to thank Professors Omar F. Mohammed Abdelsaboor and Osman Bakr for

supervising my work, and having always led me in the right direction.

Also, thank you, to the all the people with whom I have worked, in particular Lina Abdul Halim

for having welcomed and introduced me to the research world, furthermore, for having taught

me the clusters “recipe”; Silvano Del Gobbo for all the answers to my questions, and all the

time spent together trying to work on defective instruments; Federico Cruciani, for having

taught me the basics of synthetizing the fullerene derivative, and for having let me work in his

hood; Manas Parida for introducing me to the femtosecond timescale, Ghada Ahmed for her

quantum dots and Qana Alsulami for being an helpful colleague.

A special thanks to Ahmed and Amoudi, for the great time spent together during those 5 months,

and for all the “rusty” notes we played.

I have to thank also, Prof. Giovanna Brusatin for making this internship happen, David Yeh for

all his useful suggestions and for having made every effort to make this experience a great one;

Abdulrhaman for being a great roommate, and having provided me the best transportation

means, and his friend Abdulrhaman for having saved my luggage at the airport; Ali for having

processed my visa and for the time spent toghether; Mattia for his relieving poems; and Ilana

for having read and corrected my thesis with passion.

A big thank to Prof. Alberto Petrocelli because he was the first person who sparked my interest

toward science, and made me love physics. Without him I probably wouldn’t be here writing.

The last thank to my parents and my brother, who always supported me with love, and

encouraged me in my choices.

https://sperc.kaust.edu.sa/Pages/profile-Manas-Parida.aspx

Documents

Alberto Tosato Thesis