University of Groningen

Femtosecond vibrational dynamics in water nano-dropletsCringus, Gheorghe Dan

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2008

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Cringus, G. D. (2008). Femtosecond vibrational dynamics in water nano-droplets. [s.n.].

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 27-01-2021

1

Chapter 1

General Introduction

“Water!” is the most comprehensive single-word answer to the philosophical question: “What is life?” Water and life are so closely tightened together that they can hardly be imagined separately. Water is the main component of all living organisms on Earth and any indication of water on another planet generates hope for finding extraterrestrial life. Thus, it is not surprising that water is probably the most researched liquid. Especially the properties of water confined on a nanometer scale, as is the case for many biologically relevant situations, have raised considerable interest. Only in the last decade, the ultrafast dynamics and energy transfer processes became experimentally accessible, after impressive advances in the development of ultrashort laser technology. In this chapter we present a general overview of water properties and outline some peculiarities of water confined on a nanometer scale. Subsequently, the reader is introduced into the modern techniques of ultrafast optical spectroscopy. The chapter is concluded with an overview of the most appropriate model systems to investigate the ultrafast dynamics of confined water.

Chapter 1

2

1.1 Water in Nature Water is probably the most intriguing substance in the universe. Although water has a rather simple chemical formula – really dull compared to any macromolecule, it has always been perceived as a mixture of banality and mystery. On one hand, water is the most widespread liquid on Earth; on the other hand, from ancient times on, water is believed to have magical powers. Amazingly enough, this perception has not changed much with the evolution of science. Modern research has proven that water is indeed filled with magic, and life as we know it, could not exist without liquid water. This idea was beautifully expressed by Albert von Szent-Györgyi, Nobel Prize laureate in Physiology and Medicine in 1937, who referred to water as “the mother of all life” [1]. This intimate relation between life and water occurs at all levels, from macroscopic down to molecular. Life arose from water, and all living organisms consist mostly of water. “The matrix of life”, as water is sometimes called, is the medium in which life’s processes occur [2,3]. For instance, it insures the functionality of enzymes, stabilizes proteins and acts as an intermediate reactant for many bio-chemical reactions [3].

But why is water so exceptional, why can it not be replaced by another liquid, with a similar molecular structure? It may sound surprising, but at ambient temperature all the substances that resemble the chemical composition of water, are gases [4]. For instance, hydrogen sulfide (H2S), which is the closest relative of the H2O, becomes liquid only when cooled below -60O C, while water stays liquid up to +100O C [4]. Notice that the complex life-sustaining processes require a dynamic interaction with their surroundings, which can only occur in the liquid phase [2], and water is thus inimitable in this respect. However, the uniqueness of water does not end here as water exhibits a wide range of anomalous properties, many of which are also believed to be critical in sustaining life processes. More information about these can be found for instance in Ref. [5], where more than sixty unusual properties of water are listed and explained.

Most of the unique properties of water are related to the ability of the water molecules to form strong hydrogen bonds amongst each other. The hydrogen bond is an attractive interaction between an atom with pronounced electronegative character, like oxygen, and a hydrogen atom [6]. In liquid water, a fluctuating three-dimensional network of hydrogen bonds is formed, with each molecule transiently linked to up to four neighboring molecules.

In many circumstances that are relevant to life processes, the water hydrogen bond network is interrupted by biological boundaries and therefore its extent is reduced to a nanometer scale. This “nano-water” is expected to differ substantially from bulk water and

General Introduction

3

as such has been an object of extensive research [7]. The effects of confinement on the water structure and dynamics can be twofold. On one hand, the water properties near the interface with a biological object are mainly determined by the interaction between the two constituents and these interfacial effects are expected to affect only a few water layers situated in the immediate vicinity of the boundary. On the other hand, the truncation of the water hydrogen bond network might yield longer range effects, due to reduced cooperativity. These truly nanoconfinement effects would also be present beyond the interfacial layer.

On a microscopic level, many biochemical reactions involve a limited number of water molecules, for instance, a thin hydration layer or a nano-pocket of water. Typical examples of such processes are protein stabilization, in which the interfacial water plays a decisive role [8], membrane penetration by individual molecules [9], or bio-chemical reactions that occur in intracellular water [10]. Water can accomplish various functions, in particular it can act as an excess energy acceptor and it can supply the energy fluctuations needed to trigger a certain reaction [11]. These processes are intimately related to the ability of water to restructure and to accept, donate or transfer energy. Therefore, we can foresee that understanding the microscopic hydrogen bond dynamics and energy relaxation processes in nanoconfined water, might lead to a better comprehension of water-mediated biochemical reactions and to the discovery of the novel ways of controlling them. Although these prospects are quite tantalizing, it should be noted that this line of research is still in an incipient phase and there is a long way to go until present results will have practical applications.

The difficulties in accessing the water dynamics experimentally are related to the characteristic spatial and temporal scales: 1) spatial because understanding the physical interactions and dynamics on a molecular level requires sub-nanometer resolution, and 2) temporal because the typical time scales for the hydrogen bond dynamics, vibrational energy relaxation, and exchange are in the order of picoseconds (1 ps = 10-12 sec) or even femtoseconds (1 ps = 10-15 sec). Thus, the challenge is to measure nanometer-sized objects that evolve on femtosecond time scales [12].

A versatile solution to the first issue is to make use of local chemical probes which can be interrogated spectroscopically. In this case, even if the experiment is performed on a macroscopic level, a careful choice of the probe allows distinguishing the properties of a subensemble of molecules from a measurement on a larger ensemble. A good choice of probe in this respect is the hydroxyl (-OH) group, which is very sensitive to its environment [13]. For instance, it shows a large jump in its characteristic vibrational frequency when establishing a hydrogen bond with a neighboring molecule [14]. Thus, in order to determine whether we look at hydrogen bonded or non-hydrogen bonded molecules, it is

Chapter 1

4

not necessary to analyze the sample with a real sub-nm resolution, as these subspecies can be singled out by their absorption spectra. The issue of the time resolution, however, does not have a straightforward solution and it can be tackled only by a “brute-force” approach. The only technique which satisfies both requirements is time resolved spectroscopy [15,16]. In this method, laser pulses with duration in the femtosecond range are used to obtain snapshots of the sample dynamics. The experimental temporal resolution is roughly given by the duration of the pulses applied, and thus slower processes can be monitored as they develop in real time.

Because the principles of time resolved spectroscopy will be discussed in more detail in Section 1.3, in the following Section, we now focus on water itself and present a brief summary of general data regarding this liquid, with an emphasize on the spectroscopic perspective. 1.2 The Water Puzzle

In a sense, water resembles a jigsaw puzzle: single scattered pieces have nothing special, but once they are fitted together a new and fascinating picture emerges. How can we visualize the water metamorphosis from isolated molecules (gas) to a liquid state? One of the most accurate and facile techniques which provides a great deal of information about chemical substances, is infrared spectroscopy [13]. This technique uses the fact that each chemical group has specific vibrational frequencies in the mid-infrared spectral region. Moreover, some of these spectral signatures are very sensitive to the environment in which the group is placed. When light traverses a certain material, photons at these particular frequencies are absorbed. Thus, from the infrared absorption spectrum of a sample we can in principle identify the structural groups and learn about their surroundings [13].

This specificity can be easily understood if we visualize the diatomic groups from a chemical structure as harmonic oscillators. Let us consider an (isolated) OH group. Since the oxygen atom is much heavier than the hydrogen, this system resembles a harmonic oscillator with hydrogen’s mass m and force constant kO–H given by the covalent O-H bond. The characteristic frequency will then be given by ( ) mk HO /2/10 −= πυ . It is now clear that another group, for instance C-H, will have a different characteristic frequency, because this covalent bond has a different force constant. Notice also that the isotopically substituted O-D group, where D stands for deuterium, will vibrate with a lower frequency than the O-H group, due to the difference in mass (mD = 2 * mH). Last but not least, the frequency will be also affected by the hydrogen bonding to neighboring molecules, which will act as additional “springs” connected to the O-H oscillator. Although the harmonic

General Introduction

5

approximation relates the microscopic physical picture to the spectroscopic observables in a simple and informative way, it should be noted here that real systems are far from ideal harmonic oscillators. In fact, most of the experiments presented in this thesis rely on the anharmonic character of the oscillators under investigation (see Section 1.3 and Ref. [17]). Thus, a more realistic potential, such as the Morse potential [17], should be considered for a better theoretical modeling.

Let us now construct the water jigsaw puzzle and see how the image of the liquid comes into view. Figure 1.1(a) shows schematically the smallest water entity, the H2O molecule. It consists of two hydrogen atoms connected to same oxygen atom through strong covalent bonds (492 kJ/mol). On average, the two bonds are 0.096 nm long and form an angle of 104.3O. However, the water molecule is not a rigid structure and can vibrate in a number of ways, as the OH bonds can be stretched or bent. As the oxygen atom connects the two OH oscillators, considering water as two independent oscillators represents a very crude approximation. A more appropriate description involves the terminology of normal modes [17]. Water is a triatomic molecule with C2v symmetry for which three normal vibrational modes can be identified: bending (ν2), symmetric stretch (ν1) and asymmetric stretch (ν3). [Fig. 1.1(c)-(e)] These modes are delocalized over the water molecule as they involve simultaneous movements of all three atoms. An isolated water molecule, i.e., gas phase water, absorbs infrared light at the three frequencies associated with the normal modes. Consequently, the infrared absorption spectrum of an isolated water molecule, or gas phase water, contains three vibrational transitions, at ν2 = 1593 cm-1, ν1 = 3555 cm-1 and ν3 = 3750 cm-1, which are marked as dashed vertical lines in Fig. 1.2.

Although the water molecule itself is neutral, due to the pronounced electronegative character of the oxygen the electric charge is distributed highly unevenly over the molecule. (Fig. 1.1 b) The oxygen atom caries a negative charge and the hydrogen atoms the positive counterpart, resulting in a permanent dipole moment of 1.85 Debye [18].

Figure 1.1 Geometrical data and normal vibrations of the isolated water molecule.

Chapter 1

6

Consequently, when two water molecules are in close proximity, the hydrogen of one molecule is attracted by the oxygen of the other, and a hydrogen bond of the type O-H · · · O is formed [18]. In this way, the pieces of the jigsaw puzzle begin to connect.

Once a dimer is formed, another important effect emerges: the cooperativity of the hydrogen bonds. The formation of one hydrogen bond changes the distribution of the electrons such that the formation of additional hydrogen bonds is facilitated. As a result, the hydrogen bonds become significantly stronger in larger clusters [19].

The force field that acts on each oscillator is different in the presence of hydrogen bonds, and consequently the frequencies of the normal modes change significantly. It has been shown that there is a simple phenomenological connection between the hydrogen bond strength and the OH vibrational frequencies: OH groups involved in stronger hydrogen bonds have higher bend and lower stretch frequencies [14]. This can be seen from the absorption spectrum of liquid water, shown in Fig. 1.2 as a solid curve. The stretching bands are shifted to lower frequencies by more than 300 cm-1 with respect to the gas phase values, while the bend is shifted to the blue by about 50 cm-1. Another important observation is the very pronounced spectral broadening in the liquid phase. This reflects a large distribution of hydrogen bonding environments around each water molecule. In bulk liquid water, strongly hydrogen bonded OH groups, which absorb in the red part of the stretching mode spectrum, coexist with weakly and even non-bonded OH groups, whose spectral signatures appear in the high frequency (blue) side of the absorption band. For the

1000 1500 2000 2500 3000 3500 40000.0

0.2

0.4

0.6

0.8

1.0

ν2

ν1 ν3

Wavenumber (cm-1)

liquid

gas phase

Optical density

Figure 1.2 Infrared absorption spectrum of liquid water (solid curve). The dash lines

show the frequencies of the three normal modes in the gas phase.

General Introduction

7

bending mode, the situation is just opposite: stronger hydrogen bonded H2O molecules absorb in the blue part of the bending mode spectrum.

As mentioned in the previous section, “nano-water” behaves significantly different from bulk water. In this case, we look at an incomplete jigsaw puzzle, which yields a different image than isolated pieces (free molecules), but also different from the fully assembled system (bulk water). This restricted picture is still not very well understood and many questions regarding the nano-confined water are still unanswered. Some of these questions are: how many molecules are needed to yield a particular self-consistent picture, i.e. to accomplish a certain biochemical function? How stable is this image? How fast do the puzzle pieces reorganize to yield a new perspective? How many pieces do we have to put together in order to reveal the overall image or, in other words, how large does a cluster need to be to become bulk water? Is the image influenced by the background, i.e. does the surrounding influence the water nano-cluster properties? These issues are addressed in this thesis using a spectroscopic approach which is outlined in the following section. 1.3 Principles of Femtosecond Spectroscopy In general, an experimental investigation aims to provide a data set with sufficient accuracy to ensure that all the relevant features of the system under investigation have been captured. In particular, time-resolved techniques focus at obtaining a high temporal resolution, which allows monitoring some of the fastest processes that occur in nature [20]. The limiting factor in this respect is the duration of the interaction between the investigated system and the measuring tool. For instance, observing a flying bullet that changes its position in less than a millisecond requires a photographic camera which can take a picture on this time scale. The trajectory of the bullet can be subsequently reconstructed from a series of successive snapshots. Typical motions on the molecular scale occur with speeds of about 1 kilometer per second, which implies that following movements of atoms over the length of a chemical bond (~10-10 m) requires a shutter time of about one tenth of a picosecond (1 ps = 10-12 sec) [12].

Similarly to photography, in spectroscopy a system is investigated by analyzing its interaction with light. When time resolution is desired, or in other words when the goal is to obtain a “snapshot” of the system’s evolution, the light-matter interaction must be reduced to a very short period of time. The most efficient way to realize this is by using a flash of light instead of a continuous source. The system is thus observed only for the duration of the light pulse. From a succession of such “snapshots”, the system’s dynamics that occur slower than the duration of the light burst can be reconstructed.

Chapter 1

8

The limits of the time resolution have nowadays been pushed down to femtosecond [21] (1 fs = 10-15 sec) and even attosecond [22] (1 as = 10-18 sec) time scales. This has become possible only in the last decades, after breathtaking achievements in laser technology, which allowed the generation and characterization of ultrashort laser pulses. The significance of this newly emerging area of science, called femtochemistry, has been recognized by the 1999 Nobel Prize in chemistry awarded to Professor Ahmed H. Zewail (California Institute of Technology), for “His studies of the transition states of chemical reactions using femtosecond spectroscopy" [23].

In general, for a large statistical ensemble in equilibrium, the average properties are not time dependent. For instance, the chemical exchange between two species occurs both ways with equal probabilities and thus no net change can be observed [24,25]. This holds also for spectroscopic experiments performed over a large ensemble, as is usually the case. Therefore, in order to observe dynamical processes, a nonequilibrium situation must be created initially. In other words, a sub-ensemble should be instantaneously “marked” so that its time evolution can be observed separately from the rest of the system. Fortunately, ultrashort laser pulses provide also a convenient means for an impulsive excitation of a (sub)system. In this way, a subensemble having certain initial conditions can be optically labeled and consequently its evolution can be followed separately from that of the rest of the system.

This is exactly how the optical pump – optical probe experiment works (Fig. 1.3), which represents one of the most utilized techniques in time-resolved spectroscopy. The system under investigation is initially prepared in a nonequilibrium state by an interaction with a first laser pulse, called “pump”. After a certain delay, a second laser pulse, usually termed “probe”, investigates the current status of the system. By varying the delay between the pump and the probe, the evolution towards equilibrium is obtained in the form of a succession of snapshots. It is worth mentioning here that usually the system dynamics are not monitored in real time, as they occur too fast to allow varying the time delay and

Figure 1.3 Schematic representation of the pump-probe setup.

General Introduction

9

recording the data. Instead, a single snapshot is collected at a certain time delay, and then the system is allowed to reach equilibrium, so that a new interaction with the pump would yield the same effect. After that, a new snapshot is obtained by repeating the experiment for a different delay between pump and probe.

To exemplify these ideas, let us consider a practical example, namely an ensemble of OH oscillators in a liquid matrix. As mentioned in the previous section, the OH group is often used as a chemical probe due to its sensitivity to its surroundings. Provided that no interaction occurs between the OH groups, the oscillator can be approximately described by a Morse potential (Fig. 1.4). At room temperature, we can safely assume that all the oscillators are in the ground state |0>. At time zero, an ultrashort pump pulse (thick arrow in Fig. 1.4), resonant with the fundamental OH stretch frequency (~3 µm or 3300 cm-1), promotes a certain amount of oscillators into the first excited state |1>. In time, relaxation will take place and the system will return to equilibrium situation, with all the molecules in the vibrational ground state. During this time-window, the initially excited oscillators can be interrogated by a second (probe) pulse (thin arrows in Fig. 1.4). The observed spectroscopic signal is the difference between the optical density of the sample in equilibrium (no pump) and after the excitation. At the frequency corresponding to the 0-1 transition, the sample becomes more transparent. First, there are fewer molecules on the ground state that can absorb a photon at the ν01 frequency, and second, the excited

Figure 1.4 Schematic representation of a resonant pump-probe experiment.

Chapter 1

10

molecules can emit a photon at the same frequency through stimulated emission. At the frequency ν12, corresponding to the 1→2 transition, the sample is totally transparent in equilibrium. However, upon optical pumping it becomes less transparent as the excited molecules can absorb a photon at the frequency of the 1→2 transition. The changes in the optical density of the samples around the ν01 and ν12 frequencies are called bleaching and induced absorption, respectively. It is interesting to note that no pump-probe signal can be observed for a perfectly harmonic system, for without anharmonicity the two contributions exactly cancel each other [17].

What kind of information can be obtained from this experiment? First of all, the lifetime of the excited state is retrieved directly from the evolution of the pump-probe signal amplitude (bleach and induced absorption) [26]. Another possibility is to excite one of the subspecies of a two-component system and probe the other one, by taking advantage of the fact that they absorb at different frequencies. This would allow us to measure how fast the two species interchange. For instance, in this way it would be possible to asses the time needed for an initially free OH group to form a hydrogen bond or, alternatively, how long a hydrogen bond survives [27]. Additional information can be obtained by employing the polarization dependence of the light-matter interaction. Using linearly polarized light for excitation, it is possible to preferentially label those oscillators which dipole moments are aligned along the direction of the applied electric field. Subsequently, from two measurements performed with the polarization of the probe beam oriented parallel and orthogonal to the pump polarization, one can separate the isotropic (also called rotation-free) and the anisotropic signals [26]. In general, the first one reflects the relaxation of the vibrational population while the anisotropy decay is related to the reorientational diffusion of the initially excited dipoles.

Next to this simplest, essentially a two-pulse scheme, more elaborated pulse configurations can be employed. For instance, a tree pulse geometry known as photon echo [28,29] allows obtaining information on temporal fluctuations of the transition frequency. From this, information on the dynamics of the surrounding becomes directly accessible. Nonetheless, the idea remains the same: the first pulse(s) excite(s) a certain sub-ensemble which evolution is monitored by the probe pulse. 1.4 Generation of Ultrashort Laser Pulses The first optical laser ever built was described by Theodore H. Maiman in his Nature paper in 1960 [30]. Since then, the laser technology developed at an astonishing speed, with three main directions being pursued to address the scientist’s need for higher intensities,

General Introduction

11

narrower bandwidth, and shorter pulses. Nowadays, lasers can produce peak intensities in the order of petawatts (1015 W) [31], generate highly monochromatic light with sub-Hz bandwidth [32], or produce ultrashort pulses with duration in the attosecond range [22].

Time-resolved spectroscopy represents one of the most important applications for pulsed lasers and has promoted the development of ultrafast light sources covering a large spectral range, from deep ultraviolet to far infrared. Q-switching and modelocking are the most common techniques used for obtaining a stable output in pulsed regime [33,34]. Currently, the titanium-doped sapphire (Ti3+: Al2O3) laser forms the primary source of ultrashort laser pulses [35]. This is a solid state laser which can produce visible to near-infrared pulses of only a few femtoseconds in duration [21,35].

As discussed above, accessing the mid infrared spectral region around 3 µm with ultrashort laser pulses is highly desired, due to the excellent specificity of the mid-infrared absorption bands in materials. Unfortunately, there are no laser media which can directly provide an output with such characteristics. This limitation can be overcome by using frequency conversion schemes, which allow shifting the central wavelength of a laser beam while maintaining the pulse duration well below 100 fs. These methods rely on so-called nonlinear optical phenomena, which become relevant only in intense radiation fields [15,16,36]. Therefore, quite often the pulses obtained directly from the laser oscillator are amplified before being sent to frequency converters. The most common laser amplifiers are the multipass and the regenerative ones, which commonly provide an amplification factor of 106 or more [37,38].

The frequency conversion itself is based on an optical parametric processes, such as difference-frequency mixing, used to obtain lower frequencies, or sum-frequency mixing, for generating higher frequencies [36]. When an intense laser pulse mixes with another (possibly weaker) pulse in an appropriate medium, a new pulse having a different central frequency is produced (Fig. 1.5, left). The frequency and propagation direction of the newly emerging radiation are related to those of the input beams by energy (ω3 = ω1 + ω2 , see Fig. 1.5, center) and momentum (k3 = k1 + k2 , see Fig. 1.5, right) conservation laws. The input beams can in principle be any of the three waves listed in these equations, and consequently the output frequency might be either the sum or the difference of the initial frequencies. The beam with the highest frequency is usually called pump, the one with lowest frequency is called idler and the third one is called signal. The coupling between the input fields is determined by the medium in which it occurs. Therefore, nonlinear crystals are specially designed to enhance the desired parametric process through their nonlinear susceptibilities and phase-matching possibilities (which is another name for the momentum conservation law). For instance, difference frequency mixing between the Ti:Sapphire output at 800 nm (pump) and a signal wave at 1.1 µm occurs very efficiently in KTiOPO4

Chapter 1

12

[39,40] or KNbO3 [41] crystals, yielding an idler wave around 3 µm. In this way, mid-infrared pulses as short as 45 fs can be generated [41].

The duration of femtosecond pulses cannot be measured directly, because the available detectors are far too slow to follow the pulse envelope. Even the fastest photodiodes available today cannot resolve a burst of light shorter than few picoseconds. Therefore, the characterization of sub-ps pulses represents an important aspect of ultrafast spectroscopy. The pulse duration is often measured by using the same parametric processes described above. The most common one is the second harmonic generation, which is a particular case of sum-frequency generation. In this experiment, the pulse to be characterized is split into two replicas which are then spatially overlapped in a nonlinear crystal. When the two pulses also overlap in time, i.e. when they travel equivalent optical paths from the beam splitter to the doubling crystal, they mix and produce a new pulse having twice their frequency (i.e. the second harmonic beam). The intensity of the second harmonic decreases as one of the pulses is slightly delayed, such that only its leading edge overlaps with the trailing edge of the other one. For even larger delays, when the pulses are temporally separated, no second harmonic is generated. The time delay is accurately scanned by changing the optical path of one of the beams and subsequently the duration of the initial pulses can be determined from the second harmonic intensity as function of delay. The advantage of this method is that it does not require a detector faster than the pulse duration, as the experimental issue is shifted from the time to the space domain. Note that a delay of 1 fs in air is equivalent to changing the light pathway by 300 nm, which does not represent a considerable technical difficulty.

Figure 1.5 Schematic representation of optical parametric generation. Two input beams mix in an appropriate medium to produce a third one (left). The frequency and propagation direction of the emerging beam are determined by the energy (center) and momentum (right) conservation laws, respectively.

General Introduction

13

1.5 Simplifying Models in Femtosecond Mid-IR

Spectroscopy on OH Vibrations To date, infrared femtosecond spectroscopy is certainly the most suitable method to investigate the ultrafast dynamics of water. However, although the technical limitations have been overcome and the proper experimental tools are now available (Section 1.4), the interpretation of the experimental data is not always straightforward. This is due to the fact that several processes influence the spectroscopic observables in a similar fashion and as a consequence it is quite difficult to discern their individual contributions [42,43]. For instance, the origin of the 100 fs anisotropy decay observed in neat H2O is still under debate [44-47]. In this system, the anisotropy decays due to intramolecular [46,47] and intermolecular [44,45] energy transfer, librational movements [46,48], and rotational diffusion [39,48,49], but, as shown in Ref. [47], it is practically impossible to reveal the partition of each of these processes.

Consequently, over the years several model systems have been used, targeted to understanding specific phenomena. One of these makes use of probe molecules dissolved in water [50,51]. However, this is of course a very indirect way for studying the water dynamics, as it relies heavily on the assumption that the presence of the probe molecule does not significantly influence the water properties. Fortunately, one does not necessarily have to introduce an alien probe, as water and its isotopes themselves can be used as such. Figure 1.6 presents a schematic representation of the advantages and shortcomings of the most often used model systems. An excellent probe molecule is the water isotopomer – HDO, where D stands for deuterium [39,44,49,52,53] (Fig. 1.6, top). When dissolved in either H2O or D2O, the HDO probe provides two essential simplifications without disturbing the local structure or the water hydrogen bond network. First, the difference in mass of the hydrogen and deuterium atoms shifts the two oscillators (OH and OD) out of resonance, thus preventing an intramolecular coupling. Second, by studying the vibrations of the OH or OD oscillators of HDO dissolved in D2O or H2O, respectively, one can selectively access the appropriate mode of the probe molecule without significant contributions from its coupling to the surrounding. In this case, the system, i.e., the OH or OD oscillator, and the bath, i.e. D2O or H2O, respectively, are effectively separated, which greatly simplifies the theoretical modeling. Of course, a correct interpretation of the experiments on HDO in either H2O or D2O requires a profound understanding of the probe molecule itself. Therefore, separated HDO molecules dissolved in various liquid matrices have constituted the object of study for several research groups [49,54] (Fig. 1.6, left).

Another convenient model system, which provides complementary information to the studies on probe molecules dissolved in water, consists of isolated H2O or D2O molecules

Chapter 1

14

Figure 1.6 Different model systems can be used to reveal various aspects of the

complex physics of liquid water. Isolated HDO molecules in various liquids (left)

provide invaluable information about the OH oscillator, which allows using it as a

probe for the D2O dynamics (top). Information about the intramolecular processes can

be obtained by studying separated H2O molecules in other liquids (right). All these

model systems reveal different information about pure water (center). However, in

some biological settings where water is confined to nanodroplets or nanolayers (lower

right), not all of the bulk water properties are maintained,. Therefore, other systems,

such as reverse micelles (lower left), are used to reveal specific properties of water

confined on a nanometer scale.

General Introduction

15

in different liquid environments [46,54,55] (Fig. 1.6, right). In this case, a direct access to intramolecular processes is obtained, without any complications related to the water-water intermolecular resonant interactions that occur in the bulk phase. In addition, by using solvents which form hydrogen bonds of different intensities with water molecules, it is possible to get a comprehensive understanding on how the surrounding hydrogen bond network influences the intramolecular dynamics and energy flow [46,55].

Although these studies have revealed a tremendous amount of information about the ultrafast dynamics of water, the truth is that even very delicate detunings are sufficient to change “the matrix of life” into an ordinary solvent. Therefore, by using a simplified version of water one can hardly hope to reveal the complete physical picture of this miraculous liquid. Consequently, several groups have recently focused on pure water (H2O or D2O) (Fig. 1.6, center), aiming at filling of the remaining gaps. Once again, water did not fail in surprising the researchers with unexpected properties, often regarded as anomalous. For instance, the amazingly fast energy transfer between the OH oscillators of the same [46] or different [44,45] water molecules could not have been revealed by studying a probe molecule, such as the azide ion [51] or even HDO.

As was mentioned above, in many biological systems water is not found in its bulk phase, i.e. with an infinitely extended hydrogen bond network (Fig. 1.6, center), but rather in geometrically restricted environments, where the hydrogen bond network is truncated. As a result, water confined to nanodroplets or nanolayers (Fig. 1.6, bottom right) has markedly different properties than bulk water. Consequently, over the years a significant amount of research has aimed to understanding the differences between bulk water and micro- or nanopools of water. For instance, one of the most fascinating topics in this field concerns the effects of nanoconfinement on vibrational energy relaxation and dynamics of the hydrogen bond network [47,49,54,56-60].

An ideal system to model water nanodroplets would consist of water immersed in non-hydrogen-bonding solvents and would allow for progressively increasing the complexity of the hydrogen bond network from monomers to bulk water. Unfortunately, the utility of nonpolar solvents is limited because they can dissolve only minute quantities of water [55]. This conflict can be partially resolved by employing weakly bonding solvents, such as acetonitrile. This solvent presents several important advantages [49,61]. Water and acetonitrile form hydrogen bonds of the type O-H · · · N that increase the miscibility of the two liquids, allowing them to mix in any ration. Although the formation of these bonds makes it possible to mix the two substances, they are much weaker than the water-water hydrogen bonds (O-H · · · O). Therefore, the water molecules bind to each other more efficiently then to acetonitrile, forming aqueous nanopools surrounded by the solvent [61].

Chapter 1

16

The dimension of the water clusters can be varied by changing the relative concentration [49], but unfortunately their shape is not very well defined.

In this respect, the reverse micelles [47,56-60] provide a better geometrically-defined model system for water in geometrically restricted environments (Fig.1.6, lower left). The reverse micelles consist of nearly spherical nanometer-sized water droplets covered by a monolayer of amphiphilic surfactant, such as the bis-2-ethylhexyl sulfosuccinate (AOT). They are formed in the ternary oil-surfactant-water mixtures and offer the advantage of rather well defined size and shape [7]. However, the reverse micelles have their limitations. First, reverse micelles smaller than one nanometer can hardly be synthesized and therefore the smallest encapsulated water droplet still contains a few tens of water molecules [7]. This amount already suffices to provide the vibrational relaxation times of core water that are similar to bulk water [47,58]. Second, water dynamics in small reverse micelles are inherently influenced by the significant interaction between the polar side of the membrane and the encapsulated water [47,56]. Therefore, it might not be clear whether the parameters obtained are inherent to truncated hydrogen bond network or specific for the water-ion interactions. 1.6 Outline of this Thesis In this thesis the ultrafast dynamics and energy transfer phenomena in water confined to nanometers-sized droplets are investigated. Several model systems (Fig. 1.6) are used to mimic biologically relevant situations. The high temporal resolution of the described experiments allows disentangling site-specific subpicosecond dynamics and reveals significant differences in the hydrogen-bond network structure and dynamics in the immediate vicinity of a biological membrane. Whereas the results of measurements using a HDO probe molecule suggest the presence of longer range effects of the nanoconfinement, for H2O clusters the membrane appears to influence only an interfacial water layer of one or two molecules in thickness. Beyond this layer the structure and dynamics are identical to the bulk phase. One of the most interesting results is that the intermolecular vibrational energy exchange is obstructed in the interfacial layer, which puts a new perspective onto the commonly accepted picture of water as a very efficient channel for vibrational energy transfer.

One of the most facile ways of obtaining water nanodroplets in a liquid environment is mixing water and acetonitrile. In such mixtures, the water molecules tend to cluster and the solution presents microheterogeneity for a large concentration range. In Chapter 2 we present a study on the vibrational relaxation and ultrafast dynamics of water-acetonitrile

General Introduction

17

mixtures over the whole range of concentrations. To simplify the data interpretation and exclude inter- and intramolecular couplings, deuterated water (HDO in D2O) is studied rather then H2O. The formation of clusters is observed with increased concentration of water. Two distinct contributions are observed in the spectro-temporal response of the sample, those associated with the water-bonded OH groups and those related to the acetonitrile-bonded groups. In other words, the water nanodroplets appear to be composed of two regions, a shell of interfacial water and a core which resembles the bulk properties. While the lifetime of the acetonitrile-bonded groups remains unaffected by the concentration, the water-bonded oscillators display a clear acceleration of vibrational relaxation when the water concentration is increased. This observation is discussed in terms of nanoconfinement effects.

The HDO/D2O samples are in many respects different from the real biological situations, where pure H2O is involved. Therefore, limited but nonetheless valuable information is obtained from such studies. An alternative approach is taken in Chapter 3, where H2O molecules dissolved in acetonitrile are studied by means of frequency resolved pump-probe. In contrast to the case of HDO in mixtures with D2O and acetonitrile, in these samples new effects are observed that are related to the intramolecular coupling between the two OH oscillators, such as subpicosecond anisotropy decay.

The underlying physics of this system appears to be quite complex and therefore requiring additional experiments for a better understanding. Chapter 4 presents two-dimensional correlation data on monomeric H2O dissolved in acetonitrile. This technique, which involves 4 laser pulses, is much more powerful than the two-pulse pump-probe. In particular, it directly reveals coherent coupling and intramolecular energy transfer within a single water molecule.

Once the general features regarding the hydrogen bonding and the vibrational relaxation in water nano-clusters have been investigated in Chapter 2 and the specific properties of isolated H2O molecules have been revealed (Chapter 3 and 4), we can extend our investigation to H2O clusters. Chapter 5 presents a comprehensive spectroscopic study of H2O entrapped in nanometer sized reverse micelles. By monitoring the vibrational kinetics we conclude that the H2O nanoclusters contain two components with different relaxation pathways and hydrogen bonding structure. Quite surprisingly, in contrast to bulk water studies, the interfacial layer appears to be composed of molecules that do not transfer vibrational energy among each other nor to the bulk-like core. In addition, the rotational dynamics of the water molecules in the vicinity of the membrane appear to be significantly slower then in bulk water or in other regular solvents.

Finally, in the Summary (Samenvatting) section we revisit the highlights of this thesis’s content in a for a general reader, comprehensible way.

Chapter 1

18

References: [1] A.von Szent-Györgyi, Nobel Lecture, December 11, 1937,

(URL: http://nobelprize.org/nobel_prizes/medicine/laureates/1937/index.html)

[2] P. Ball, Life's matrix: a biography of water, Farrar, Straus, and Giroux, New York, (1999).

[3] M. Chaplin, Nature Reviews - Molecular cell biology 7, 861 (2006). [4] CRC handbook of chemistry and physics, Crc Press Llc1999). [5] M. Chaplin, URL: http://www.lsbu.ac.uk/water/index2.html

[6] G.C. Pimentel, A.L. MacClellan, The hydrogen bond, Freeman, San Francisco, (1960). [7] M.P. Pileni, Structure and reactivity in reverse micelles, Elsevier, Amsterdam, (1989). [8] S.K. Pal, A.H. Zewail, Chem. Rev. 104, 2099 (2004). [9] B.L.d. Groot, H. Grubmuller, Science 294, 2353 (2001). [10] Z. Szolnoki, Biochemical and Biophysical Research Communications 357, 331 (2007). [11] L.D. Barron, L. Hecht, G. Wilson, Biochemistry 36, 13143 (1997). [12] A.H. Zewail, Nature 412, 279 (2001). [13] J. Coates, Interpretation of Infrared Spectra, A Practical Approach, John Wiley &

Sons Ltd., Chichester, (2000). [14] R.E. Miller, Science 240, 447 (1988). [15] Y.R. Shen, The principles of nonlinear optics, Wiley & Sons, New York, (1984). [16] S. Mukamel, Principles of Nonlinear Optical Spectroscopy, Oxford University

Press1999). [17] G. Herzberg, Infrared and raman spectra of polyatomic molecules, Van Nostrand,

New York, (1945). [18] F. Franks, Water: a comprehensive treatise, Plenum Press, New York, (1972). [19] J.J. Dannenberg, J. Molec. Struc. 615, 219 (2002). [20] A. Zewail, The Chemical Bond: Structure and Dynamics, Academic Press, Boston,

(1992). [21] A. Baltuska, Z. Wei, M.S. Pshenichnikov, D.A. Wiersma, Opt. Lett. 22, 102 (1997). [22] N.A. Papadogiannis, B. Witzel, C. Kalpouzos, D. Charalambidis, Phys. Rev. Lett. 83,

4289 (1999). [23] Official web site of the Nobel Foundation: http://nobelprize.org/

[24] S. Woutersen, Y. Mu, G. Stock, P. Hamm, Chem. Phys. 266, 137 (2001). [25] J. Zheng, K. Kwak, J. Asbury, X. Chen, I.R. Piletic, M.D. Fayer, Science 309, 1338

(2005). [26] R.G. Gordon, J. Chem. Phys. 45, 1643 (1966).

General Introduction

19

[27] J.D. Eaves, J.J. Loparo, C.J. Fecko, S.T. Roberts, A. Tokmakoff, P.L. Geissler, PNAS 102, 13019 (2005).

[28] W.P. de_Boeij, M.S. Pshenichnikov, D.A. Wiersma, Chem. Phys. Lett. 253, 53 (1996). [29] M. Khalil, N. Demirdoven, A. Tokmakoff, J. Phys. Chem. A 107, 5258 (2003). [30] T.H. Maiman, Nature 187, 493 (1960). [31] M.D. Perry, G. Mourou, Science 264, 917 (1994). [32] C. Salomon, D. Hils, J.L. Hall, J. Opt. Soc. Am. B 5, 1576 (1988). [33] O. Svelto, Principles of lasers, Plenum press, New York, (1976). [34] A.E. Siegman, Lasers, University Science Books, Mill Valley, CA, (1986). [35] P.F. Moulton, J. Opt. Soc. Am. B 3, 125 (1986). [36] R.W. Boyd, Nonlinear optics, Academic Press, Boston, (1992). [37] F. Salin, J. Squier, G. Mourou, G. Vaillancourt, Opt. Lett. 16, 1964 (1991). [38] S. Backus, R. Bartels, S. Thompson, R. Dollinger, H.C. Kapteyn, M.M. Murnane, Opt.

Lett. 26, 465 (2001). [39] S. Woutersen, U. Emmerichs, H.J. Bakker, Science 278, 658 (1997). [40] S. Yeremenko, A. Baltuska, F. de Haan, M.S. Pshenichnikov, D.A. Wiersma, Opt.

Lett. 27, 1171 (2002). [41] C.J. Fecko, J.J. Loparo, A. Tokmakoff, Opt. Commun. 241, 521–528 (2004). [42] J. Lindner, P. Vöhringer, M.S. Pshenichnikov, D. Cringus, D.A. Wiersma, M.

Mostovoy, Chem. Phys. Lett. 421, 329 (2006). [43] J. Lindner, P. Vöhringer, CORRECT, Chem. Phys. xxx, xx (2007). [44] S. Woutersen, H.J. Bakker, Nature 402, 507 (1999). [45] M.L. Cowan, B.D. Bruner, N. Huse, J.R. Dwyer, B. Chugh, E.T.J. Nibbering, T.

Elsaesser, R.J.D. Miller, Nature 434, 199 (2005). [46] D. Cringus, T.l.C. Jansen, M.S. Pshenichnikov, D.A. Wiersma, J. Chem. Phys. 127,

084507 (2007). [47] D. Cringus, A. Bakulin, J. Lindner, M.S. Pshenichnikov, P. Vöhringer, D.A. Wiersma,

in preparation [48] J.J. Loparo, C.J. Fecko, J.D. Eaves, S.T. Roberts, A. Tokmakoff, Phys. Rev. B 70,

180201 (2004). [49] D. Cringus, S. Yeremenko, M.S. Pshenichnikov, D.A. Wiersma, J. Phys. Chem. B

108, 10376 (2004). [50] R. Jimenez, G.R. Fleming, P.V. Kumar, M. Maroncelli, Nature 369, 471 (1994). [51] P. Hamm, M. Lim, R.M. Hochstrasser, Phys. Rev. Lett. 81, 5326 (1998). [52] C.J. Fecko, J.D. Eaves, J.J. Loparo, A. Tokmakoff, P.L. Geissler, Science 301, 1698

(2003).

Chapter 1

20

[53] J.B. Asbury, T. Steinel, K. Kwak, S.A. Corcelli, C.P. Lawrence, J.L. Skinner, M.D. Fayer, J. Chem. Phys. 121, 12431 (2004).

[54] J.J. Gilijamse, A.J. Lock, H.J. Bakker, PNAS 102, 3202 (2005). [55] H. Graener, G. Seifert, A. Laubereau, Chem. Phys. 175, 193 (1993). [56] J.C. Deak, Y. Pang, T.D. Sechler, Z. Wang, D.D. Dlott, Science 306, 473 (2004). [57] H.S. Tan, I.R. Piletic, R.E. Riter, N.E. Levinger, M.D. Fayer, Phys. Rev. Lett. 94,

057405 (2005). [58] D. Cringus, J. Lindner, M.T.W. Milder, M.S. Pshenichnikov, P. Vöhringer, D.A.

Wiersma, Chem. Phys. Lett. 408, 162 (2005). [59] A.M. Dokter, S. Woutersen, H.J. Bakker, Phys. Rev. Lett. 94, 178301 (2005). [60] A.M. Dokter, S. Woutersen, H.J. Bakker, PNAS 103, 15355 (2006). [61] R.D. Mountain, J. Phys. Chem. A. 103, 10744 (1999).