Ruthenium polypyridyl photochemistry

Ruthenium polypyridyl complexes certainly rank amongst the most researched family of compounds in inorganic photochemistry. They are interesting complexes to study, having relatively long (100’s ns) emission lifetimes and a range of applications. It was the oil crisis of the 1970’s that sparked interest in these compounds, as potential hydrogen fuel generators by the photochemical splitting of water, and as seen in other posts, they are currently at the forefront in terms of efficiency in dye-sensitised solar cells. In addition, they have been used as DNA probes and oxygen sensors. The photochemistry of these complexes is discussed below. Readers are recommended to be familiar with the concepts in the “Light Absorption and Fate of the Excited State” article before studying this material.

Like so many aspects of modern photochemistry, Ireland has some key researchers in ruthenium photochemistry and the article below draws from a recent perspective by John Kelly (TCD) and Han Vos (DCU). The fundamentals are discussed here with applications discussed in a forthcoming article.

1. Introduction to Inorganic Photochemistry

We have looked elsewhere at Jablonski diagrams for organic molecules. Inorganic molecules, or more specifically d-block complexes, add an extra layer of molecular orbitals to this Jablonski diagram, between the ground state (HOMO) of the organic compound (which is now the ligand) and the excited state (LUMO). This opens up a range of new transitions, aside from the HOMO-LUMO transition observed in organic chromophores. This latter transition in inorganic photochemistry is called a ligand-field or ligand-ligand transition, as in the excited state the electron is located on the ligand.  As well as this, because of the presence of the metal’s molecular orbitals, three other transitions are available – a d-d transition, where an electron is excited from a metal orbital to an unoccupied metal orbital (this is usually referred to as a metal centred (MC) transition as well as transitions between the metal and the ligand. These can be either an electron excited from the ligand to the metal, called Ligand to Metal Charge Transfer (LMCT) or from the metal to the ligand (MLCT). Because of the energy differences between the various types of transitions, ligand field transitions are usually in the near-UV region (analogous to where we would expect organic molecules to absorb light), charge transfer transitions are in the visible region. The resulting emission from charge-transfer states is often highly coloured.

Light absorption in d-block (octahedral) complexes resulting in from left: metal centred (MC), ligand to metal charge transfer (LMCT), metal to ligand charge transfer (MLCT) and ligand-ligand transition (L-L)

Light absorption in d-block (octahedral) complexes resulting in from left: metal centred transition (MC), ligand to metal charge transfer (LMCT), metal to ligand charge transfer (MLCT) and ligand-ligand transition (L-L)

In order to discuss these transitions in context, we will focus on the, that is, the, inorganic photochemistry complex: Ru(II)(bpy)32+.

2. Fundamentals of ruthenium polypyridyl photochemistry

2.1 Absorption and Emission

Because of the incorporation of metal orbitals, the Jablonski diagram needs to incorporate the notation discussed above. Ruthenium in oxidation state II is d6, and so as an octahedral complex its electrons are in the low-spin t2g6 configuration. Incident light at about 450 nm promotes one of these electrons to a ligand anti-bonding orbital, a metal to ligand charge transfer. (We’ll discuss this, but you might consider how this was established.) Therefore we modify the S0 – S1 notation used in the Jablonski diagrams of organic molecules to one which denotes the type of excited state in inorganic ones – in this case 1MLCT. Transfer to 3MLCT is efficient (heavy atom effect) and so ruthenium complex’s photochemistry generally happens from here. [Remember intersystem crossing is effectively an electron flip, from a situation where electrons are paired to one where they are unpaired.]

Jablonski diagram for ruthenium polypyridyl complexes.

Jablonski diagram for ruthenium polypyridyl complexes. Solid lines and dashed lines are radiative and non-radiative processes respectively.

Absorption (top, source unknown) and emission (bottom, author's results) of Ru(bpy)3 complex in water

Absorption (top, source unknown) and emission (bottom, author's results) spectra of Ru(bpy)3 (2+) complex in water

The absorption and emission data are shown. Ruthenium absorbs at 450 nm (2.8 eV) and emits strongly at ~620 nm (~2.0 eV) in water. This emission is caused by radiative process from the 3MLCT state to the ground state. Emission lifetimes are approximately 200 ns in water in aerated solution and 600 ns in deaerated water. The oxygen in water is a very efficient quencher, and quenches emission with a rate of ~ 109 M-1 s-1. It is possible to map out the various deactivation processes of the excited state to investigate its kinetics:

Deactivation processes of an excited state M* in the presence of a quencher (oxygen)

Deactivation processes of an excited state M* in the presence of a quencher (oxygen)

The quantum yield of emission is therefore affected by how efficient the rate of emission is compared to the rates of deactivation and quenching. This is quantified by the Stern-Volmer relationship (oxygen quenches according to the dynamic quenching model) as discussed in the Quenching section, according to the equation below:

Stern Volmer Equation for Quenching with oxygen as quencher

Stern Volmer Equation for Quenching with oxygen as quencher

The rate constants, in particular the rate constant for deactivation, are dependent on how close the ground and excited states are. The excited state of this complex is a charge-transfer state (charge has moved from one region of the molecule to another), and therefore is very sensitive to solvent polarity – it will be stabilised in more polar solvents. Therefore, changing solvent polarity will affect the energy of the emitting state. It is found that on changing the solvent from water to acetonitrile, the emission lifetime increases from 635 ns to 870 ns, and the quantum yield of emission increases by 50% from 0.o4 to 0.o6. The emission maximum increases in energy from 627 nm to 615 nm.

These results can be explained as follows: on decreasing polarity of the solvent, the emitting state is destabilised by about 12 nm. This increase in energy difference between ground and excited state means that there is poorer overlap of the vibrational levels of the ground and excited state, so the deactivation process is not as efficient. Therefore the deactivation rate constant term is lower in the expression for the emission quantum yield in the presence of quencher, above, indicating a larger emission quantum yield. All of this is based on the assumption that the radiative rate constant remains unchanged, which is found to be true in practice. This observation is generally summarised as the Energy Gap Law – the larger the gap between ground and excited state, the less efficient deactivation processes are.

2.2 Nature of the Excited State

Absorption and emission spectra give initial information on the excited state, and are the photochemist’s initial tools to probe the excited state chemistry of molecules. To delve further, flash photolysis/transient spectroscopy give more detailed information. Flash photolysis, as mentioned elsewhere on this site, allows us to study the excited state by obtaining its lifetime and absorption spectrum. An experimental set-up is outlined below (more details onthe general details of flash photolysis in the Experimental article on Flash Photolysis). Excitation using, for example a Nd:YAG laser at 355 nm, generates the excited state which quickly equilibrates to the 3MLCT state. At this stage, a Xe or Hg/Xe obtains an absorption spectrum of the excited state. This was traditionally acquired point by point (i.e. measuring the change in absorption at 400, then 410, then 420 nm, etc) but iCCD (intensified charge coupled device) detectors are now the norm – these acquire information across a broad spectral range (~600 nm) at once. As well as providing structural information on the nature of the excited state by generating its absorption spectrum, flash photolysis also allows for the lifetime of this state to be measured, by acquiring a spectrum at intervals after the laser flash, therefore monitoring the decay of the excited state.

Schematic of Transient Absorption Spectroscopy Experiment: Laser excites sample and change in absorption is monitored by a xenon lamp. The simulated transient spectrum (top right) is the difference in absorption after laser flash, showing negative (dissapearance of ground state) and positive (formation of transients) absorbances. The absorption spectrum is shown on the bottom for comparison. Inset shows a kinetic trace of any of the transient peaks from which lifetime information can be gleaned.

Schematic of Transient Absorption Spectroscopy Experiment: Laser excites sample and change in absorption is monitored by a xenon lamp. The simulated transient spectrum (top right) is the difference in absorption after laser flash, showing negative (disapearance of ground state) and positive (formation of transients) absorbances. The absorption spectrum is shown on the bottom for comparison. Inset shows a kinetic trace of any of the transient peaks from which lifetime information can be gleaned.

The transient spectrum is shown with the accompanying ground state absorption spectrum. In the transient spectrum, it can be seen that some peaks have negative changes in absorbance whereas others have positive changes. The negative changes in absorbance (“bleaching”) occur where the molecule shows absorbance bands in the ground state. Hence, with a transient spectrum, the lash flash results in the formation of the excited state, and the xenon lamp records the loss of ground state chromophores – any absorbance that was present because of these chromophores is now registered as negative changes in absorbance in the transient spectrum. On formation of excited/transient state, new chromophores are present, which are monitored by the xenon lamp, and hence appear as positive changes in absorption (remember ground and excited states are chemically different species). To generate a true transient spectrum, the differences in absorption is subtracted from the absorption spectrum, although this is rarely necessary. The decay curve, in the inset is the rate of decay of one of the peaks – e.g. the transient peak at 390 nm. Fitting this curve to an exponential function allows for the rate constant (and hence lifetime) of the transient state to be easily determined. For example, if the decay was found to be mono-exponential, the curve of intensity (I) versus time (t) would be fitted to the expressionand allow for calculation of k.

emission_decay

The above experiment discusses results from a nanosecond experiment, but if we were to push faster, into the picosecond and femtosecond domain, the processes of intersystem crossing and relaxation in the triplet state would be observed. These kind of experiments are how information such as charge injection rates  in dye-sensitized solar cells can be determined.

The extent of positive absorbances in transient spectroscopy provide information on the nature of the transient species or excited state. Like conventional UV/vis spectroscopy, broad featureless bands very often don’t provide much direct information. However, considering the various types of transitions available, why is the excited state assigned as a MLCT state? This state, as indicated above, results in an extra electron residing on the bipyridyl (bpy) ligand, after an electron was transferred from the metal to it. Therefore, the transient spectrum should show characteristics of this bpy radical (called “bpy dot minus”). How can this be done? Well with the assistance of our electrochemical friends, we can electrochemically generate the bpy radical, and obtain its UV/vis spectrum (this technique is called spectroelectrochemistry). If it has characteristics similar to those in the transient spectrum (which in this case it does, the band at 368 nm), we can conclude that they must be attributed to the same chromophore.

3. Conclusion

In this first of two articles, we have looked at basic photophysical properties of a ruthenium complex and examined how absorption, emission and transient spectroscopic studies provide information on their excited state. In the second article, we will look at how these properties are used in a variety of applications.

4. References and Further Reading

Photochemistry of polypyridine and porphyrin complexes, K. Kalyanasundaram, Academic, London: 2002. Very comprehensive book on the area with excellent introduction covering theory in much more detail than above.

Vos, J. G. and Kelly, J. M., Ruthenium polypyridyl chemistry: from basic research to applications and back again, Dalton. Trans., 2006, 4869 – 4883. Good ooverview of the synthesis of these complexes and their variety of applications, especially looking at the role of Irish researchers in the area

Advertisements

Luminescent PET Sensors

PET sensors are very simple in principle and can signal the presence or absence of a very small amount of analyte in solutions/biological samples by using a readily available instrument – the fluorimeter. This article looks at the design principles of PET sensors, as well as examining how they may be modified to enhance specificity/selectivity. [Aug 2009]

1. Introduction

Luminescent sensors provide for an easy, often visual, method for detection of a wide range of ions, physical properties such as pH and other components such as nanoparticles. Their wide variety of sensing capabilities are already used in commercial devices, and a greater understanding of their mechanism of operation is leading to newer exciting concepts, such as molecular logic gates for computing. This article aims to summarise the background theory to luminescent sensors – especially concentrating on photoinduced electron transfer (PET) which we can consider as the third type of quenching (Type I and Type 2 discussed elsewhere – links will be here) – and examining practical applications in the context of this theory. In summarising the work here, special attention is paid to the work of Prof A. P. de Silva who is based in Queen’s University, Belfast and who is one of the pioneers in the area. Interested students are referred to some of his papers in the references, where even in the formal constraints of academic journals, his passion for the subject is very evident.

Here’s a short podcast outlining what we will cover in this section (with audio)
Vodpod videos no longer available.

2. Background

We’ve seen several times that molecules that have absorbed light may transfer their electron to another system – which is usually called a quencher, although is not constrained to this (see for example dye-sensitized solar cells). Because this electron is transferred after absorption of light by the molecule, the process is called photo-induced electron transfer (PET). It’s one of the great applications of photochemistry in general – excited states are better oxidisers and reducers than their ground state counterparts. Suppose we have donor (reducing agent, itself going to be oxidised) and acceptor (oxidising agent, itself going to be reduced) molecules as shown below, where the energy level of the donor HOMO is lower than the energy of the acceptor LUMO. As it stands, the donor cannot reduce the acceptor, as electron transfer is not energetically feasible. However, if we form a donor excited state, D*, by light excitation, the process becomes energetically feasible once the donor LUMO is higher than the acceptor LUMO. You should consider how excited states are also better oxidising agents.

Excited states are better reducing agents (and oxidising agents) than their corresponding ground states

Excited states are better reducing agents (and oxidising agents) than their corresponding ground states

In applying this general concept to this section, we will be looking at molecules that have both these components in one molecule. In this system, we will term them luminophore and receptor, which are analogous to the discussion of donor and acceptor respectively. The luminophore is the light absorber which results in an excited state and may luminesce, or accept an electron into the vacancy in the ground state, or donate an electron from the excited state. The receptor will accept/donate an electron from/to the luminophore. The two components are linked by a non-conjugated bridge (e.g. two sp3 carbon bridge). You may wish to consider why a non-conjugated bridge is required. The three components – luminophore-spacer-receptor – make up what we will term here a class III type of quenching; where the PET process occurs within one molecule.

Schematic of a Fluorophore-Spacer-Receptor model

Schematic of a Fluorophore-Spacer-Receptor model

The power of the model is that if the receptor’s energy level’s can be tweaked in the presence or absence of an analyte – for example an metal cation – and the energy levels between the receptor and luminophore are close, well the presence of the ion may turn on or shut off the electron transfer process. This concept is now a sensor, as luminescence is affected by the presence or absence of an analyte. If that luminescence can be measured (which it can), we have a very powerful, sensitive analytical tool to measure the extent of analyte present.

3. Details

3.1 Principles

The general principle discussed above is summarised in the diagram below. A luminophore gives off light in the absence of an analyte, but emission is quenched in the presence of an analyte. This is called an ON-OFF system – emission was turned off by the presence of an analyte. The systems become very powerful when the receptor – the docking area for the analyte – is selective to only a particular type of species (e.g. a halide) or even better a specific ion. An example of one of de Silva’s systems is shown below. The three components are obvious: the anthracene molecule is the luminophore (in this case a fluorophore as it emits from the excited singlet state); the bipyridyl component is the receptor and the two carbon-chain acts as a spacer. In the absence of an ion, this system fluoresces strongly. However, in the presence of either zinc ion of a proton the emission is shut off. So how can we explain what happened?

An example of an ON-OFF sensor: in the absence of protons or zinc ions, emission is observed (ON), whereas in the presence of either of these ions, emission is strongly quenched. (Based on de Silva, J. Am. Chem. Soc., 1999, 121, 1393)

An example of an ON-OFF sensor: in the absence of protons or zinc ions, emission is observed (ON), whereas in the presence of either of these ions, emission is strongly quenched. (Based on de Silva, J. Am. Chem. Soc., 1999, 121, 1393)

Examining the emission spectrum, we can see that the anthracene emission is very much reduced in the presence of one of the specific ions, but the shape or position of the bands do not change. Therefore its energy levels remain as they were prior to the presence of the receptor. It can be concluded that the receptor levels change, opening up a PET pathway that was not available prior to the presence of the analyte. The diagram below summarises the schematic, molecular and energetic processes.

Schematic (top); energy level (middle) and molecular (bottom) representations of an ON-OFF sensor. The dashed line in the energy level representation of the acceptor-analyte shows the energy level prior to binding of the analyte

Schematic (top); energy level (middle) and molecular (bottom) representations of an ON-OFF sensor. The dashed line in the energy level representation of the acceptor-analyte shows the energy level prior to binding of the analyte

3.2 Developing the idea further

In a paper discussing the developments of luminescent sensors so far this century, de Silva illustrates a scenario that is already a reality (Tetrahedron, 2005, 61, 8551 – 8588). An ambulance is called for a car accident trauma, and upon arrival paramedics take a blood sample. From this, they can alert the hospital en route of the required electrolyte levels required for a blood bag, having used a small portable instrument to determine concentrations of for example, sodium, potassium, calcium as well as oxygen levels and pH. The key here is that the the device incorporates a range of PET sensors, each of which have receptor components of the moleculethat selectively sense a particular of ion/molecule. Therefore, the question is how is this done?

Selectivity has been achieved by a range of interesting approaches, and we have to be thankful to organic chemists for their synthesis of the diverse range of molecules available. Crown ethers are useful receptors, and are known to bind well to sodium ions. The image below is from de Silva’s homepage. It shows the use of a crown ether in a PET sensor that selectively senses for sodium. The principle is the same as above (except this is now an OFF-ON sensor – you should sketch out the energy levels to see how emission is turned on in the presence of sodium ions)

Fluorescent sensor for sodium, with emission spectra showing an incease in emission on additional increments of sodium concentration (From AP de Silva, used with permission)

Fluorescent sensor for sodium, with emission spectra showing an incease in emission on additional increments of sodium ion concentration (From AP de Silva website - link in text, used with permission)

Now that the general principle is understood, it is easy to get selective across a wide range of ions. By modifying the size of the crown ether, the sensor can lose its selectivity for sodium ions and move on to potassium ions (see image below). Various minor modifications can make it calcium selective. Very quicky, the range of ion selectivities can be acquired.

An example of a sensor that is potassium ion selective (because of the larger crown ether diameter compared to the sodium ion equivalent, above), from de Silva et al, Dalton Transations - used by permission of the Royal Society of Chemistry - see link below)

An example of a sensor that is potassium ion selective (because of the larger crown ether diameter compared to the sodium ion equivalent, above), from de Silva et al, 2003, Dalton Transations - used by permission of the Royal Society of Chemistry - see link below)

The principle can also be applied as a pH monitor, by developing a proton sensitive detector. In this case, amines linked to an anthracene molecule via a spacer make for a very simple but effective pH sensor. An example is shown below – a clever attribute of this design is that the use of two amines mean that much greater sensitivity across the range of pH is “built in” to the molecules design. The emission at various pH values is shown alongside the molecule. You should aim to sketch out the appropriate changes to the molecule on increasing acidity resulting in this change in emission. As well as simple pH monitors, these types of materials have been used in examining the effectiveness of cancer treatment. Cancer cells respond to treatment by developing acidic environments, and by examining the emission of a pH sensor on treating with radiation. The images, available in fig. 3 in the original paper show increased pH, easily detected, indicating that treatment is having an effect.

An amine-anthracene pH sensor, with changes in emission in at different pH's shown. (Source unknown)

An amine-anthracene pH sensor, with changes in emission in at different pH's shown. (Source unknown)

As well as cations, a lot of work has been invested into developing anion sensors. Gunnlaugsson, based in Trinity College Dublin is a leading researcher in this area and a suummary of many of his results can be found in the reference at the end. A video demonstrating the effect will be posted here September 09.

4. Conclusion

The principle of photo-induced electron transfer can be utilised in the development of molecular sensors. The mechanism of PET was examined above, as well as looking how shutting off or turning on this process provides information on the nature of ions interacting with the sensor. Sensors can be designed so that they  selectively sense an ion in the presence of others. Current research in this area is looking at using these principles in molecular computing, and this work will be surveyed in a future article.

[Aug 2009 – updates/amendments will be logged in comments.]

5. References

Newer optical-based molecular devices from older coordination chemistry, de Silva, A. P., McCaughan, B., McKinney, B. O. F. and Querol, M, Dalton Trans., 2003, 1902 – 1913. Really excellent overview of the area with lots of examples.

Luminescent sensors and switches in the early 21st century, Callan, J. F., de Silva, A. P. and Magri, D. C., Tetrahedron, 61, 8551 – 8588. Overview of  developments in 21st Century across range of analyte types (cations, anions, molecules)

Anion recognition and sensing in organic and aqueous media using luminescent and colorimetric sensors, Gunnlaugsson, T., Glynn, M., Tocci (nee Hussey), G. M., Kruger, P. E. and Pfeffer, F. M., Coord. Chem. Rev. , 2006, 250, 3094–3117. Good overview of anion sensors.