878 =@ IBJ :H=O FDJ=>IHFJE B C=I FD=IA FHJ=JA@ FAFJE@AI

(1)

University of Groningen

Master Thesis

VUV and soft X-ray photoabsorption of gas phase protonated peptides

Author:

Geert Reitsma

Supervisors:

Dr. T. Schlath¨olter Prof. Dr. R. Hoekstra Referee:

Dr. J.P.M. Beijers

November 4, 2010

Atomic and

Molecular

Physics

(2)

Abstract

The transport of biomolecules from space to early Earth could have had important consequences for the development of early life. Therefore, the stability of peptides upon VUV photon and soft X-ray interactions is of fundamental interest. In order to study these interactions, protonated peptides are produced by an electrospray ionization source and trapped in a three dimensional quadrupole in which they are exposed to synchrotron radiation.

The obtained VUV photofragmentation spectra show diﬀerent regimes of dissociation processes. Below ionization potential, the photon energy is mainly absorbed by the peptide bond, which leads to slow IVR governed processes. Above ionization threshold both fast dissociation through repulsive states and slow IVR dissociation processes can occur. The strongest observed peak is related to the fast loss of the tyrosine sidechain.

Estimations of the internal energy show that the fast tyrosine sidechain loss eﬃciently cools the remaining peptide. It is conceivable that such cooling processes facilitate the survival of peptide substructures after photon absorption in space.

Preliminary soft X-ray absorption induced fragment mass spectra recorded around the oxygen K-edge are presented. These spectra show the feasibility of a systematic study on the photostability of peptides under soft X-ray irradiation.

(3)

Introduction

A lot of key tasks of cells in living organisms are carried out by proteins. For example, enzymes act as catalysts to speed up chemical reactions, which are crucial for the or- ganism. Many other proteins are involved in ligand binding and signal transduction. A well known example of a ligand binding protein is hemoglobin, which transports oxygen molecules from lungs to other organs. A third group, structural proteins, confer stiffness and rigidity in fluid biological compounds, such as keratin in nails and hair. Millions of different proteins can be built using twenty different amino acids in an enormous variety of arrangements. For that reason these amino acids can be considered as building blocks of life.

Since amino acids have been found in the interstellar medium [1], meteoretic materials and recently in comets, it is very likely that they exist in space in larger quantities [2, 3]. Transport of these molecules from space to early Earth could have had important concequences for the development of early life. Therefore the stability of amino acids and small peptides under vacuum ultra violet (VUV) and soft X-ray photon radiation is of fundamental interest.

UV light is electromagnetic radiation with wavelengths between 10 nm and 400 nm.

VUV is the part of this wavelength region which is absorbed by air (100 nm - 200 nm) and therefore can be used only in vacuum. Stars like the Sun are natural sources of UV/VUV photons. Today, the atmosphere of the Earth blocks 97 - 99 % of the UV/VUV intensity.

The early Earth did not have such an atmosphere and for that reason amino acids were naturally exposed to UV/VUV radiation.

X-rays are photons in the wavelength region between 0.01 and 10 nm. The region between 0.1 and 10 nm is called the soft X-ray band. X-rays are expected from very compact objects like neutron stars and material falling into a black hole. The atmosphere of the Earth has a very high opacity over the whole X-ray region.

A lot of research has been done on photoinduced ionization and fragmentation of amino acids in the condenced phase and the gas phase. Synchrotron radiation has been used for near edge X-ray absorption ﬁne sructure (NEXAFS) spectroscopy. X-rays are absorbed by core level electrons, leading to their subsequent emission. The core hole is ﬁlled by an electron from a higher shell, leading to emission of an Auger electron or a

(6)

CHAPTER 1. INTRODUCTION

Figure 1.1: The spectrum of electromagnetic radiation. VUV photons have wavelengths between 100 and 200 nm and soft X-rays have wavelengths between 0.1 and 10 nm.

(7)

photon. The initially emitted electron and the Auger electron or the ﬂuorescent photon can be detected. Experimental NEXAFS spectra of amino acids have been reported in diﬀerent phases, such as adsorbed on a surface [4, 5], dissolved in a liquid [6] and in the gas phase [7–9]. The gas phase absorption and fragmentation spectra have been reported for alanine, proline, glycine, methionine and threorine. Also theoretical calculations for the carbon, oxygen and nitrogen K-edges in amino acids have been reported [10, 11].

Recently an X-ray absorption study of the dipeptide glycine-glycine in the gas phase has been reported [12].

Also ionization and fragmentation of amino acids upon VUV photon irradiation has been studied [13]. Even peptide bond formation was observed under VUV irradiation of ices [14].

Until now, VUV and soft X-ray induced photofragmentation of free complex biomolecules such as peptides, short sequences of amino acids with the same chemical structure as proteins, have not been reported because of the difficulty of getting sufficient densities of intact molecules. Photofragmentation spectra of such peptides in the gas phase could give new insights into their photostability during transport to and during presence on the early Earth. In this report the first systematic study on VUV photon induced fragmentation will be described. Also a feasibility experiment for soft X-ray induced fragmentation will be presented. The influence of the electronic structure of peptides on the ionization and fragmentation will be discussed.

A well known peptide is the neurotransmitter leucine enkephalin (ﬁgure 1.2). This relatively small pentapeptide consists of a sequence of the amino acids tyrosine, glycine, glycine, phenylalanine, leucine (YGGFL) and has been studied by a large range of dissociation techniques. A lot of spectra have been reported, for example by collission induced dissociation (CID) [15], surface induced dissociation (SID) [16], laser induced dissociation (LID) [17] and keV-ion induced dissociation (KID) [18]. To some extent these spectra can be used to interpret photofragmentation spectra.

Electrospray ionization (ESI) is used in order to produce protonated molecules in the gas phase. The beam of these molecules is focussed, guided and mass selected by an ion funnel, an RF-only quadrupole and an RF quadrupole mass ﬁlter. In order to accumulate a suﬃcient dense of molecular target and to be able to irradiate the target over a longer period of time, the protonated molecules are trapped into a Paul trap.

The ion cloud in the trap is irradiated with VUV photons and soft X-rays produced by the third generation synchrotron BESSY II in Berlin. The trap content is analyzed by a time-of-ﬂight mass spectrometer.

The relevant fragmentation processes after photoexcitation and photoionization will be described in chapter 2 of this report. The experimental setup will be described in chapter 3 and the results will be presented and discussed in chapter 4.

(8)

Tyrosine (Y) Glycine (G) Glycine (G) Phenylalanine (F) Leucine (L)

b1(164) b2(221) b3 (278) b4 (425)

y4(393) y3(336) y2(279) y1(132)

a1(136) a2(193) a3(259)

a4(397)

Figure 1.2: Schematic view of leucine-enkephalin which consists of ﬁve amino acids. Also the most important fragments and their masses are indicated.

(9)

Chapter 2

Theory

When a molecule absorbs a photon by undergoing a transition ∣𝑖⟩ → ∣𝑓 ⟩, the photon energy ℎ𝜈 is related to the energies 𝐸_𝑖 and 𝐸_𝑓 by:

ℎ𝜈 = 𝐸_𝑖− 𝐸_𝑓, (2.1)

where 𝐸𝑖is the initial state energy and 𝐸_𝑓 is the ﬁnal state energy. If the photon energy exceeds the ionization potential IP of the molecule, the molecule can be photoionized.

The emitted electron has the kinetic energy

𝐸_𝑘 = ℎ𝜈 − 𝐼𝑃. (2.2)

Both photoexcitation and photoionization can lead to dissociation of the molecule.

The diﬀerent absorption processes and the mechanisms that cause the dissociation of the molecules are shortly discussed in this chapter. Soft X-rays cause transitions from the inner shell electrons, which can be followed by an Auger decay and VUV photons cause valence transitions. A photoabsorption process can be followed by either direct dissociation on femtosecond timescale or IVR governed dissocitation which occurs on longer timescales.

2.1 Molecular orbitals

In atoms, the possible energy states are only determined by different arrangements of the electron cloud. However, molecules have more internal degrees of freedom. Their states are not only determined by the electron cloud, but also by the position of the nuclei and their movements. This makes the derivation of the molecular orbitals very complicated, only the simpliest cases can be solved analytically. For macromolecules, like leucine enkephalin, density funcional theory (DFT) is often used to calculate important properties like the valence molecular orbital density, the location of the highest occupied molecular orbital (HOMO) and the ionization potential (IP) [18]. In DFT the Hohenberg-Kohn energy functional 𝐹_𝐻𝐾(𝑛(𝑟)) is introduced, which is equal to the internal energy of the system. Different systems are uniquely labeled by different electronic densities 𝑛(𝑟). 𝐹𝐻𝐾(𝑛(𝑟)) can be approximated by the sum of the total kinetic

(10)

CHAPTER 2. THEORY

Figure 2.1: In (a),(b) and (c) the HOMO, HOMO-1 and HOMO-2 respectively are indicated as shaded areas; (d) shows the distribution of the valence electrons in Leucine- Enkephalin. [18]

energy, the Hartree term and an exchange correlation term. Taking the derivative of this functional gives a set of Kohn-Sham equations. These equations can be solved numeri- cally with an accuracy of typically 1 percent. A detailed description of DFT is given in reference [19].

The calculated ionization potential of leucine enkephalin is 8.87 eV. Figure 2.1 shows the three highest occupied molecular orbitals and the distribution of the valence electrons. The HOMO and the HOMO-1 are both 𝜋-orbitals and are located at the phenylalanine sidechain ring. The HOMO-2 is also a 𝜋-orbital and is located at the tyrosine sidechain ring.

(11)

CHAPTER 2. THEORY

Figure 2.2: The molecular orbitals formed by 2p atomic orbitals. Each bonding orbital has a corresponding anti-bonding orbital. A molecule can be excited by transfering an electron from a bonding into an antibonding state.

2.2 Molecular electronic transitions

In a simple picture, molecular bond is formed by overlap of two atomic orbitals, which both contain one unpaired electron. Two types of overlapping atomic orbitals are 𝜎 and 𝜋 bonds. A 𝜎 bond occurs when the overlapping orbitals are head to head, a 𝜋 bond when the overlapping orbitals are parallel (see figure 2.2). This figure also shows that all bonding orbitals have corresponding antibonding orbitals. A molecule can be excited by making a transition from the 𝜎 bond to the 𝜎 anti bonding orbital, 𝜎 −→ 𝜎^∗. Also transitions are possible from the 𝜋 bond to its anti bonding orbital and from lone pairs to anti bonding 𝜎^∗ and 𝜋^∗ orbitals, 𝜋 −→ 𝜋^∗, 𝑛 −→ 𝜎^∗ and 𝑛 −→ 𝜋^∗ respectively. Usually 𝜎 bonds are stronger than 𝜋 bonds and bonding orbitals are stronger than corresponding antibonding orbitals, as shown in figure 2.3.

It is expected that in leucine enkephalin the peptide bond contributes to the absorption around the ionization potential. The peptide bond can be described as a four orbital model with two doubly occupied 𝜋 orbitals 𝜋1 and 𝜋2, a lone pair orbital 𝑛0

and an antibonding 𝜋₃ orbital. Photon absorption causes the transitions: 𝑛₀ −→ 𝜋^∗₃, 𝜋2 −→ 𝜋₃^∗ and 𝜋1 −→ 𝜋₃^∗ [20].

(12)

CHAPTER 2. THEORY

Figure 2.3: A qualitative energy scaling of the diﬀerent (anti) bonding orbitals. The molecular orbitals are stronger if they are formed by 2s atomic orbitals than if they are formed by 2p atomic orbitals. In general a bonding orbital has lower energy than its corresponding anti-bonding orbital.

2.3 Photoabsorption around the oxygen K-edge

Leucine enkephalin contains ﬁve (C=O), two (O-H) and two (C-O) bonds (ﬁgure 1.2).

These bonds have slightly diﬀerent resonances on transitions from the oxygen inner shell 1s state to the lowest unoccupied molecular orbital. In a (C=O) double bond a 1s electron can make a transition to the vacant 𝜋^∗ orbital. In the (O-H) bond there is a resonance to the 𝜎^∗ orbital and in the (C-O) bond to 𝜋^∗and 𝜎^∗orbitals. Also transitions are possible to higher unoccupied molecular orbitals as well as direct ionization. Figure 2.4 shows two typical spectra around the oxygen K-edge of L-alanine and L-proline measured by Marinho et al [7]. Resonances are observed for transitions from the oxygen 1s orbital to the 𝜋^∗ orbital in the C=O bond and in the C-O bond and to the 𝜎^∗ orbital of the C-O and O-H bond. The continuous tail can be explained by the fact that many transitions are possible to higher unoccupied molecular orbitals. When leucine enkephalin is irradiated with soft X-rays in the 525 - 545 eV energy range it is expected that the photofragmentation spectra have a similar structure.

When a protonated molecule undergoes a transition due to inner shell excitation, a vacancy is induced in the inner shell 1s orbital:

[M + H]⁺+ h𝜈 ^innershell−→ [[M + H]⁺]^∗ (2.3) This can be followed immediately by an Auger decay. Due to the vacancy, a higher level electron de-exites into the core level, which causes a release of energy. This energy can be transfered to another electron, which is ejected from the system:

[M + H]⁺+ h𝜈 ^innershell−→ [[M + H]⁺]^∗ ^Auger−→ [[M + H]⁺]⁺+ e⁻ (2.4)

(13)

CHAPTER 2. THEORY

Figure 2.4: A typical absorption spectrum around the oxygen K-egde. These spectra are obtained by estimating the ion yield of L-alanine and L-proline under soft X-ray irradiation. [7]

The inner shell electron can also be ejected from the molecule by absorption of a photon with energy above ionization threshold. After the ionization an Auger process occurs, which ejects another electron:

[M + H]⁺+ h𝜈 ^innershell−→ [[M + H]⁺]^∗+ e⁻ ^Auger−→ [[M + H]⁺]²⁺+ 2e⁻ (2.5) If the excitation (equation 2.3) transfers the inner shell electron into a dissociative state, it can be directly followed by a fragmentation process, (direct dissociation):

[M + H]⁺+ h𝜈 ^innershell−→ [[M + H]⁺]^∗ −→ direct dissociation (2.6) If after the Auger relaxation process (equations 2.4 and 2.5) the molecule is in a dissociative state, it can also be followed by direct dissociation. If the ﬁnal states in equations 2.4 and 2.5 are not directly dissociative, a product with high internal energy is left, which tends to dissociate via intramolecular vibrational energy redistribution (IVR), a process which will be described in more detail in section 2.5 [21].

2.4 VUV photoabsorption

Whereas soft X-rays access strongly bound inner shell electrons, VUV photons cause transitions from the valence orbitals of a molecule. When the energy of the incident

(14)

CHAPTER 2. THEORY

photons is smaller than the ionization potential, only excitation of the molecule can occur. Obviously, when the energy is higher, ionization processes dominate. The excitation processes are called valence transitions:

[M + H]⁺+ h𝜈 valenceband

−→ [[M + H]⁺]^∗ (2.7)

In leucine enkephalin four peptide bonds, two aromatic sidechains, one amine group and a carboxyl group are available to contribute to valence transitions (ﬁgure 1.2).

If ℎ𝜈 > IP an ionization proces can occur which can leave a leucine enkephalin radical.

If the energy of the incident photon exceeds the ionization potential only slightly, an ionized molecule will be left which has relatively low internal energy:

[M + H]⁺+ h𝜈 valenceband

−→ [[M + H]⁺]⁺_cold+ e⁻ (2.8) If energy excess is small, a lot of dissociation pathways are energetically not accessible.

When the excess of energy is bigger, higher bound electrons can be emitted from the molecule and an excited internally hot molecule will be left:

[M + H]⁺+ h𝜈 ^innershell−→ [[M + H]⁺]^+∗_hot+ e⁻ (2.9) Since the molecule has more internal energy, many more dissociation pathways are expected to contribute to the spectra.

After photoexcitation many diﬀerent processes can occur, such as de-excitation, direct dissociation and IVR.

2.5 Intramolecular Vibrational energy Redistribution

To explain the IVR process it has to be concidered that for each electronic state there exist a set of vibrational modes. Transitions from one vibrational mode to another can take place through absorption of a photon. The vibrational mode can be described as a superposition of all possible normal modes. Vibrational modes are thus coupled. Due to this interaction vibrational energy can be transfered from an originally excited mode to other vibrational modes. This way energy is eventually distributed over all energetically accesible modes.

If the energy excess in the molecule is relatively low (equation 2.8), relatively little vibrational energy is transfered from the excited mode to other vibrational modes in the molecules. If the excited molecule is internally hot, more vibrational modes are accessible. For that reason it is expected that molecules which are ionized by photons with energies slightly above ionization threshold are much less prone to dissociation compared to molecules which are irradiated with photons of a few electronvolts above ionization threshold.

Direct dissociation can occur on a femtosecond timescale, while IVR governed dissociation occurs on picosecond timescales. [22]

(15)

Chapter 3

Experimental Setup

Gas-phase protonated peptides are generated by an electrospray ionization source (ESI).

The beam of these positively charged ions is focussed by an RF ion funnel. The ions are then guided to a Paul trap by an RF quadrupole ion guide and an RF quadrupole mass ﬁlter. After being trapped in the Paul trap, the cloud of peptide ions can be exposed to keV ions from an electron cyclotron resonance ion source (ECRIS). Alternatively irradiation with UV photons from the U125/2-10m-NIM beamline and soft X-rays from the U49/2-PGM1 beamline at the third generation synchrotron BESSY II has been performed. The trapped ions and fragments are extracted from the Paul trap into a time-of-ﬂight (TOF) mass spectrometer. The particles are detected by a micro channel plate (MCP) detector. Figure 3.1 shows a schematic view of the setup.

In this chapter the most important components of the setup are described in more detail.

3.1 Electrospray Ionization

Electrospray ionization is a technique that allows a large range of ion types to be transfered from solution to the gas phase and subjected to mass spectrometric analysis [23].

needle capillary

ion funnel

mass filter Paul trap time-of-flight MCP quadrupole

lenses

10-1 mbar 10-4 mbar 10-7 mbar 10-9 mbar

1 bar

ion / photon beam

Figure 3.1: A schematic view of the setup.

(16)

CHAPTER 3. EXPERIMENTAL SETUP

The technique is based on the theories of John William Strutt (3rd Baron Rayleigh), who studied what happens during the evaporation of a solvent from a droplet which has a net electric charge. It was predicted that the density of charges in such a droplet will increase to a critical value at which the Coulomb repulsion will overcome the surface tension. This instability will cause the droplet to break up into fragments. The described critical value is also known as the ’Rayleigh Limit’ [24]. The mechanisms underlying the electrospray process have signiﬁcant implications for the properties of the produced ions [23, 25].

3.1.1 The electrospray process

It is convenient to divide the electrospray ionization process into three stages: charged droplet formation at the electrospray capillary tip, droplet shrinkage and gas-phase ion formation (Figure 3.2).

A voltage of 3 - 4 kV is applied to a metal needle, which has typically an inner diameter of 0.1 mm and is located roughly 1 cm from a counter electrode. Because the needle tip is very narrow, the electric field is very high (𝐸 ≈ 10⁶ V/m). The solution, which is passed through the needle, consists of a dipolar solvent, a low concentration of an electrolyte (10⁻⁵− 10⁻³ M) and the analyte itself (typically 40𝜇M). Assuming that the potential is positive, positive ions in the solution will accumulate at the surface, which is thus drawn out in a downfield direction to establish a ’Taylor cone’. At a sufficient high potential this cone becomes instable and a filament, whose surface is enriched with positive ions, is emitted from the cone tip. At some distance downfield, the filament breaks into seperate droplets.

The radius of the droplets is approximately 𝑅₀ ≈ 1.5 𝜇m and they carry a charge of 𝑄0 ≈ 10⁻¹⁴ C. The Rayleigh limit is deﬁned by the Ralyeigh equation: 𝑄²_𝑅 = 64𝜋²𝜖0𝛾𝑅³_𝑅, in which 𝛾 is the surface tension. Due to evaporation of the solvent 𝑅𝑅

decreases. When the Rayleigh limit is reached, the droplet breaks apart. A succesion of Rayleigh instabilities will cause shrinkage of the droplets.

There are two models which describe the third stage, formation of fully desolvated ions. The ﬁrst model is the charge residue model (CRM) which assumes that the proces of evaporation and droplet shrinkage is continued until only one ion is left. The second model assumes that the ions are directly emitted if a certain suitable radius (< 10 nm) has been reached. This model is known as the ion evaporation mechanism (IEM). A solid concensus has been emerged that the CRM is applicable for analyte molecules of masses above at least 3300 Da. For smaller molecules such a concensus has not been reached. It is likely that for these molecules the mechanism is some combination of both CRM and IEM.

3.1.2 Properties of electrosprayed ions

The mechanisms underlying the electrospray process have several implications for the properties of the produced ions. Three important issues are the charge state of the

(17)

Figure 3.2: The three stages in the electrospray process: droplet formation, droplet shrinkage and gas-phase ion production. In the paultje setup, a capillary acts as counter electrode, through which the protonated molecules are injected into the funnel region [25].

ions, the internal energy of the ions and the possiblity for preservation of noncovalent bondings.

During the analysis of ions with large molecular mass a broad range of charge states can be observed. This multiplicity of protonation is related to the number of sites with signiﬁcant basicity in solution. Figure 3.3 shows the electrospray peak of singly protonated leucine-enkephalin. The proton is probably located at the most basic site, the N-terminal primary amino group. Figure 3.4 shows the spectrum of the triply protonated oligonucleotide GTAC; It is expected that these molecules can be more eﬃciently electrosprayed in the negative ion mode.

The electrospray process is a subtle process. In fact, the molecules are cooled by the desolvation process. That is why macromolecules, like oligonucleotides and peptides, do not necessarily break apart during the ionization process. Under the right conditions very large ions with relatively low internal energy can be introduced into the mass spectrometer.

Another useful property is that during the electrospray process the noncovalent interactions between the solvent and the ions are disrupted. It is possible to intercept this process in order to conserve relatively strong non-covalent interactions, which are of analytical signiﬁcance. In this way it is for example possible to produce clusters of two or more molecules.

(18)

200 400 600 800

0 1

m/z

intensity

556.6

Figure 3.3: The spectrum of singly protonated leucine enkepahlin. The proton is located at the most basic site of the molecule in solution, the N-terminal primary amino group.

200 400

0

1 392.2

m/z

intensity

Figure 3.4: The spectrum of the triply protonated oligonucleotide GTAC, which has mass 1174. The mass-to-charge ratio of multiple protonated molecules can be calculated using ^𝑚_𝑧 = ^{𝑀 +𝑛}_𝑛 .

(19)

3.1.3 Electrospray in the Paultje setup

The ﬂow rate of the solution through the electrospray needle is controlled by a sy- ringepump. A capillary (length 160 mm, diameter 1mm) is used as a counter electrode.

In order to stimulate the desolvation process, the capillary can be held on a temperature up to 200 degrees Celcius. After passing through the capillary, the ions are injected into the funnel region.

3.2 The RF ion funnel

The ion cloud, which is injected from atmospheric pressure to the first differential pumping region, is diffuse as a result of both Coulomb repulsion and sonic expansion. An ion funnel can effectively focus the ion cloud and transmit them through a relatively small exit aperture. This exit aperture is compatible to the acceptence aperture of the RF quadrupole in the next differential pumping stage [26, 27].

3.2.1 Working principle

An ion funnel consists of a series of ring electrodes with decreasing inner diameter. An RF voltage with a 180 degree phase shift between adjacent electrodes is applied. A DC voltage is applied across the funnel. Figure 3.5 shows the ion funnel cross section.

The funnel region can be approximately seperated into an RF strong field region and an RF field free region. The approximation is shown in figure 3.6 for a simple ring electrode system. Shown is that the ratio between the inner diameter of the electrodes and the spacing between them defines the field free region. An electrodynamic ’funnel’

appears when the inner diameters are slightly decreasing. The shape of this ’funnel’ is shown by the dashed lines in ﬁgure 3.5. The ion cloud is guided through the funnel region by a DC voltage.

The ion funnel operates in a region with relatively high pressure (0.1 mbar). The presence of this neutral background gas causes multiple collissions. Due to these collisions both kinetic and internal energy of the ions are reduced. The phase space is thus compressed. This compression is beneﬁcial for further guiding and trapping. The internal cooling is important because it leaves the ions intact. [28]

3.2.2 The funnel design for Paultje

The ion funnel for Paultje is based on the design of Julian et al [27] and is shown in ﬁgure 3.5. It consists of 26 ring electrodes of 0.51 mm thickness which are seperated 5.1 mm from each other. The ﬁrst 10 electrodes are constructed of copper and have an inner diameter of 38.1 mm. The remaining electrodes are constructed of stainless steel and have inner diameters starting with 36.3 mm and reducing linearly to 7.9 mm. The diaphragm between the funnel region and the RF quadrupole region has a diameter of 1 mm. The capillary electrode extends 1.5 cm into the funnel.

(20)

Figure 3.5: The funnel interface: 26 electrodes, having RF and DC voltages; The dashed line represents the shape of the ﬁeld free region; The cloud is a result of SIMION simu- lations by Julian et al [27].

Figure 3.6: Approximation of the ﬁeld free region in a cross sectional view of the ring electrode system. The ﬁeld free region begins at approximately 60% of the distance between the electrodes, following the formula 𝑎 = 3^−0.5𝑐 [27].

(21)

Figure 3.7: The electrode structure of the quadrupole (b) and the equipotential lines of the quadrupole ﬁeld (a). [29]

The RF voltages are applied by way of 1000 pF capacitors to adjacent electrodes. A voltage divider, consisting of 27 180 kΩ resistors, is used to provide a linear decreasing or increasing DC voltage across the funnel. The RF voltage is generated by a synthesizer and an RF ampliﬁer. Typical parameters for the RF voltage are 500-1000 kHz and 50-100 W. A balun is used to produce two RF voltages with a phase diﬀerence of 180 degrees.

3.3 Quadrupole ion guides and mass ﬁlters

After being focussed and cooled by the RF funnel device, the ions are injected through an electrostatic diaphragm into the entrance aperture of an RF quadrupole ion guide.

The ion guide consists of four rods as approximation of hyperbolically shaped bars on which voltages are applied such that particle trajectories are stable in the electrodynamic field. In the next stage the ions enter a quadrupole mass filter, which has the additional property that the voltage on the rod pairs can be biased in order to select one specific mass. The mass selected beam is guided through an electrostatic Einzel lens into the Paul trap [29].

(22)

3.3.1 Two- and three-dimensional quadrupole ﬁelds

In the electric quadrupole ﬁeld the potential is quadratic in Cartesian coordinates:

Φ = 𝜙₀ 2𝑟₀²

(𝛼𝑥²+ 𝛽𝑦²+ 𝛾𝑧²⁾ (3.1)

Two conﬁgurations satisfy the condition 𝛼 + 𝛽 + 𝛾 = 0, which is imposed by the Laplace condition ΔΦ = 0. Assumption of the condition 𝛼 = 1 = −𝛾 and 𝛽 = 0, results in the following two-dimensional solution:

Φ = 𝜙0

2𝑟₀²

(𝑥²− 𝑧²⁾ (3.2)

The three-dimensional solution is obtained by assuming 𝛼 = 𝛽 = 1 and 𝛾 = −2:

Φ = 𝜙0(

𝑟²− 2𝑧²⁾

𝑟₀²+ 2𝑧²₀ (3.3)

3.3.2 The two dimensional-quadrupole ﬁeld

The conditions for equation 3.2 can be generated by an electrode conﬁguration which is shown in ﬁgure 3.7b. Four hyperbolically shaped linear electrodes are put on potential

±^Φ₂⁰. The ﬁeld strength is given by:

𝐸𝑥= −Φ0

𝑟₀²𝑥, 𝐸𝑧 = Φ0

𝑟₀²𝑧, 𝐸𝑦 = 0 (3.4)

A charged particle which is injected into the quadrupole ﬁeld oscillates harmonically in the x-y plane. Due to the opposite sign in the equation for 𝐸_𝑧 the amplitude in the z-direction increases exponentially. Therefore the trajectory is unstable and the ions will hit the electrodes. If the applied voltage is periodic in time, this behaviour can be avoided. The ions are periodically focussed and defocussed in both x- and z-direction, which causes a stable motion around the y-axis.

If a DC voltage 𝑈 and an RF amplitude 𝑉 are applied with an angular frequency 𝜔.

The potential 𝜙₀ can be written as Φ₀ = 𝑈 + 𝑉 cos 𝜔𝑡, the equations of motion are:

¨ 𝑥 + 𝑒

𝑚𝑟₀²(𝑈 + 𝑉 cos 𝜔𝑡) = 0 (3.5)

and

¨ 𝑧 − 𝑒

𝑚𝑟²₀ (𝑈 + 𝑉 cos 𝜔𝑡) = 0. (3.6)

These equations of motion can be rewritten in dimensionless parameters:

𝑑²𝑥

𝑑𝜏² + (𝑎 + 2𝑞 cos 2𝜏 ) = 0 (3.7)

(23)

(a) (b)

Figure 3.8: The stability diagram of the two-dimensional quadrupole ﬁeld. Only the regions of both x and z stability give a stable motion along the y-axis. (a). The most relevant stability region is the one in which 𝑎 < 0 and 0 < 𝑞 < 1 (b). [29]

and

𝑑²𝑧

𝑑𝜏² − (𝑎 + 2𝑞 cos 2𝜏 ) = 0 (3.8)

with

𝑎 = 4𝑒𝑈

𝑚𝑟²₀𝜔², 𝑞 = 2𝑒𝑉

𝑚𝑟₀²𝜔², 𝜏 = 𝜔𝑡

2 . (3.9)

These equations are known as Mathieu equations, which have two types of solutions:

stable motion and unstable motion. Whether stability exists only depends on a and q. Therefore, the regions of stability and instability in both x and z direction can be mapped on the a-q plane as shown in ﬁgure 3.8a. For a stable motion along the y-axis the solutions have to lie in the overlapping regions for x and z stability. The most relevant region of overlap, 𝑎 < 0 and 0 < 𝑞 < 1, is shown in ﬁgure 3.8b.

3.3.3 Practical use of the stability diagram for the two-dimensional quadrupole ﬁeld

Usually a quadrupole is operated in a manner such that the values of parameters a and q are related by a simple ratio. That is, the ratio 2𝑈/𝑉 is constant regardless of the actual magnitude of either 𝑈 or 𝑉 . In the a-q diagram, this is equivalent to a straight line through the zero intercept. Since the ratio 2𝑈/𝑉 is independent of the mass, all masses are lying on this line. Because 𝑎 and 𝑞 are inversely proportional to mass (equation 3.9), the lighter particles will appear at the top right and the heavier particles near the lower left-hand corner. In ﬁgure 3.8b the operating line is drawn in the stability diagram. In this conﬁguration 𝑚₁ and 𝑚₂ are in the stable region and 𝑚₃ will be lost.

If one wants to use the quadrupole only for ion guiding and not for mass ﬁltering, it can be operated in the RF-only mode. In this mode the DC potential between the

(24)

rods is zero. This means that the ratio 2𝑈/𝑉 is zero and the operating line will lie on the a=0 axis. A broad band of ion masses is guided through the quadrupole. The ﬁrst quadrupole in Paultje is operated in this mode.

The quadrupole can be operated as a mass ﬁlter by modiﬁcation of the operation line. If the magnitudes of 𝑈 and 𝑉 are increased, with keeping the ratio 2𝑈/𝑉 constant, also the mass corresponding to a certain point in a-q space will increase, due to equation 3.9. This way the transmitted mass can be varied along the operating line.

The slope of the operating line can be increased by increasing the ratio 2𝑈/𝑉 . If the slope is appropriately adjusted, the operating line ideally crosses the very narrow tip of the stability region. This results in a very high mass resolution.

The mass ﬁlter which is used in Paultje can be operated in the RF-only mode and in the resolving mode. In the RF-only mode only an RF voltage is applied, together with an oﬀset bias. In the resolving mode the DC voltage is switched on and the slope of the operating line can be adjusted by change of the DC voltage with respect to the RF voltage.

3.4 The Paul trap

In order to accumulate a sufficiently dense molecular target and be able to irradiate it for a longer period, the ions need to be interfaced with the irradiating beam over a longer period of time. For this reason the ions need to be stored in a three dimensional quadrupole ion trap, a Paul trap. The properties of the ion cloud in this Paul trap significantly influence the shape of the observed TOF-mass spectra.

The potential configuration in the ion trap is given by equation 3.3. This configuration can be generated by an hyperbolically shaped ring electrode and two rotationally symmetric endcap electrodes, as shown in figure 3.9. Similar to the two-dimensional case a voltage Φ₀ = 𝑈 + 𝑉 cos 𝜔𝑡 is applied, in this case between the endcap electrodes and the ring electrode. Because the field in z-direction is now a factor two stronger than the field in the r-direction, the shape of the stability region is different, as shown in figure 3.10. Also in this case the stable region can be chosen by changing the ratio 2𝑈/𝑉 [29].

3.4.1 Properties of the ion cloud

The time resolution, the sensitivity and the signal-to-noise ratio of the experiment are highly inﬂuenced by the properties of the ion cloud in the Paul trap. Mean kinetic energy, position and background gas play an important role.

The ions are trapped in a pseudopotential well of a few eV depth [30], which means that ions with a higher kinetic energy have unstable trajectories. The kinetic energy of the ions can be reduced by using buﬀer gas cooling, i.e. by elastic collissions with a non-reactive neutral background gas [31]. In Paultje Helium is used at an estimated pressure of 10⁻³ mbar.

Due to buﬀer gas cooling more ions can be trapped and their orbitals collapse to- wards the center of the trap, enlarging the ion density at the center and reducing the

(25)

Figure 3.9: A schematic view of the Paul Trap. A potential is applied between the endcap electrodes and the ring electrodes. [29]

(26)

Figure 3.10: The lowest stability region of the ion trap. [29]

(27)

size of the cloud. The larger ion density in the center increases the sensitivity of the experiment, because the extraction diaphragm in the endcap and the entrance diafragm of the irradiating beam in the ring eletrode are both in the center of the trap. A smaller size of the ion cloud also improves the time resolution of the experiment. The difference in time-of-flight of particles of equal mass, at different positions inside the ion cloud, influence the width of the observed signals.

Contributions from background gas, neutral particles from the ESI-source and con- taminations in the buﬀer gas can cause noise in the mass spectra. In order to decrease this noise, a nitrogen-cooled cryotrap is used to condense background gases to a cold surface close to the trap. Also secondary electrons caused by an irradiating beam hitting a surface can cause noisy contributions to the mass spectra. For that reason a correct alignment of the beam to the trap center is important.

3.5 Particle detection and data acquisition

Protonated peptides and their positive fragments are extracted by an extraction pulse between the endcap electrodes of the Paul Trap (Δ𝑈 = 400V, 𝜏 = 5𝜇s), via two correc- tion lenses, into a linear time-of-ﬂight (TOF) mass spectrometer [32] and detected by a micro channel plate (MCP) detector [33]. The ﬂight time between the extraction pulse and the signal of the detector is recorded by a 1 GHz digitizer. The data is acquired in three-scan cycles.

3.5.1 Time-of-ﬂight mass spectrometry

When a charged particle is accelerated in a time-of-ﬂight tube, the relation between ﬂight time 𝑡 and mass-to-charge ratio ^𝑚_𝑞 is the following:

𝑡 = 𝑘

√𝑚

𝑞 + 𝑐 (3.10)

in which k is a proportionality constant containing design characteristics and settings and c is a small oﬀset. The values of k and c can be easily calibrated by reproducing a known spectrum. Usually the well known spectra of keV ions on leucine enkephalin are used [18].

3.5.2 Micro channel plate detectors

Figure 3.11a shows a schematic picture of an MCP detector. An MCP is an array of miniature electron multipliers oriented parallel to each other. The channels are usually fabricated in such a way that they are optimized for secondary particle emission and charge replenishment from an external source. In a typical MCP detector two MCP’s with angled channels are rotated 180 degrees with respect to each other. An electric ﬁeld is applied over the channels, typically 1000 V accros one MCP. If an incident particle hits the MCP surface, secondary electrons are produced. Due to the electric ﬁeld in the

(28)

MCP 1 MCP 2 Grid

Anode

Signal -3 kV -5 kV -3 kV

e- e-

(a)

Semiconductor layer

Electroding

Primary radiation Secondary electrons Glas channel wall

Output electrons -HV+

(b)

Figure 3.11: A schematic picture of a MCP detector (a) and a single channel (b). [33]

MCP, the secondary electron signal is amplified by multiple collisions with the channel walls (figure 3.11b). Therefore the detector can have a gain of 10⁷ (figure 3.12a). An anode is used to collect the produced electron cloud and deliver the output signal. The detection efficiency of an MCP detector depends strongly on the type of radiation and the energy of the incident particles, as shown in figure 3.12. For positive ion detection, the highest detection efficiency has been reached if the particles have a few keV of kinetic energy. For that reason the TOF tube is biased at typically 3 kV.

3.5.3 The three-scan cycle data-acquistion process

In order to obtain the net eﬀect of the ion irradiation, a three scan cycle data acquisition procedure is used. During the ﬁrst scan the irradiating beam is blocked (by an optical shutter for photons or a high voltage chopper for keV ions) to obtain an electrospray-only signal. In the second scan the mass spectrum of irradiated ions is taken. In the third scan the electrospray is blocked by a high voltage on the diaphragm between the ion funnel

(29)

600 800 1000

Chevron

Single MCP 107

105

103

Gain

Voltage (per plate) (a)

Type of radiation Detection efficiency (%) Electrons

Positive ions

UV

0.2 - 2 keV 50 - 85

2 - 50 keV 10 - 60

0.5 - 2 keV 5 - 85

2- 50 keV 60 - 85

50 - 200 keV 4 - 60

30 - 110 nm 5 - 15

110 - 150 nm 1 - 5

Soft X-rays 0.2 - 5 nm 5 - 13

Diagnostic X-rays 0.012 - 0.02 nm 1

(b)

Figure 3.12: The gain versus voltage per plate for both the single MCP and the Chevron (a) and a the detection eﬃciency of diﬀerent particle types and energies b). [33]

(30)

Buffer gas

Extraction

Buffer gas + ESI Buffer gas + ESI

Irradiation

Irradiation Buffer gas pumped down

Buffer gas pumped down Buffer gas pumped down Scan 1

Scan 2

Scan 3

Time Extraction Extraction

Figure 3.13: Schematic view of the three-scan cycle.

and the ﬁrst quadrupole, to obtain the spectrum from irradiation of the background.

The ﬁrst and third scan are substracted from the second scan in order to obtain the net eﬀect of ion irradiation. Figure 3.13 shows a schematic overview of the three scans.

The width of the ESI and beam pulses and the length of the time delays between them depends strongly on the type of experiment that has been performed.

3.6 VUV and soft X-ray beamlines at the third generation synchrotron BESSY II

The ions are exposed to photons in the vacuum ultra violet (VUV) and the soft X-ray region. The VUV photons are produced by the quasiperiodic undulator U125/2 at the third generation synchrotron facility BESSY II. Photons in the energy range 8 - 40 eV are obtained by a 10 meter focal length normal incidence monochromator (10m NIM) [34].

The soft X-rays are produced by the undulator U49/2 and the tunable photons beams in the range 85 - 1600 eV are obtained by a plane grating monochromator (PGM) [35].

3.6.1 Synchrotron radiation produced by undulator devices

An undulator consists of a periodic structure of dipole magnets, which causes an alter- nating magnetic ﬁeld with wavelength 𝜆𝑢. An electron moving trough the undulator is forced to oscillate, and therefore to emit electromagnetic radiation due to the periodic acceleration, following the Maxwell equations. If the electron has ultrarelativistic ve-

(31)

locity it emits a well-collimated beam of synchrotron radiation in the forward direction.

Because the oscillation amplitude of the motion is small, the radiation of the diﬀerent periods of the magnetic ﬁeld exhibits an interference pattern. Therefore the synchrotron radiation is very intense within a narrow energy band in the electromagnetic spectrum.

The ﬁeld parameters of undulator U125 and U49 are 𝜆 = 125mm and 𝜆 = 49mm with 32 and 84 periods, respectively.

3.6.2 Normal incidence monochromators and plane grating monochromators

A monochromator is an optical device which is able to transmit a narrow band of wavelengths chosen from an incident beam with a wider spectral range. The 10m NIM consists of an entrance slit, an exit slit and a spherical grating. When a collimated beam of light with a wavelength 𝜆 is incident on a reﬂection grating under a normal angle it is diﬀracted following

𝑑 sin Θ𝑚= 𝑚𝜆, (3.11)

with 𝑑 the spacing between the grating lines, Θ𝑚the diffraction angle and m the order of diffraction. When the grating is illuminated with light containing several wavelengths, the different wavelengths are diffracted in different angles. To tune the wavelength which is passed through the exit slit, the grating can be moved both rotationally and translationally. In order to control the wavelength resolution, the size of the exit slit can be adjusted. Figure 3.14 shows a schematic picture of the different components of the U125/2 10m NIM beamline. The PGM consists of a plane grating and a plane mirror.

The ratio between the angle of incidence and the angle of diffraction (the c_ff value) can be freely chosen. A schematic overview of the U49/2 PGM1 beamline is given in figure 3.15.

(32)

Undulator U125 U125

U125 M1

M1

M2 M3 M2

M3 M4

M1: toroïdal mirror 172 º M2: elliptical mirror 178 º M3: toroïdal mirror 155 º M4: toroïdal mirror 155 º G: spherical grating ES: entrance slit AS: exit slit

ES AS ES/AS

G

G M4

M4 Top View

Side View

Figure 3.14: Top view and side view of the U125/2 10m NIM beamline at BESSY II.

M1 is a toro¨ıdal mirror and M2 an eliptical mirror. The slit size of the entrance slit (ES) can be varied between 0 and 2000 𝜇m. A spherical grating (G) has been used and the exit slit is rotatable and variable between 0 and 2000 𝜇m. Two toro¨ıdal mirrors are used to guide the photon beam to the setup [35].

(33)

Figure 3.15: Top and side view of the U49/2 PGM1 beamline at BESSY II. A toro¨ıdal mirror (M1) and a plane mirror (M2) have been used. Diﬀerent plane gratings can be used in the plane grating monochromator (G). A cylindrical and a toro¨ıdal mirror are used to guide the photon beam to the setup. The exit slit can be translated and its size can be adjusted between 0 and 2000 𝜇m [34].

(34)

Chapter 4

Results and Discussion

4.1 VUV absorption

The mass spectra which are obtained after VUV photon irradiation are normalized to the relative number of photons and the target density. The relative photon number is calculated from the photodiode current, the quantum eﬃciency of the diode and the irradiation time. The target density is obtained from the area of the leucine enkephalin peak in the electrospray-only spectrum.

4.1.1 Results

Figure 4.2 shows the normalized mass spectra in the photon energy range between 8 and 40 eV. The relative intensity scales on the y-axis of these spectra are signiﬁcantly diﬀerent.

At photon energies around the ionization threshold (8 and 9 eV) the peaks at 278/279 and 120 are the dominating peaks. Figure 4.1 suggest that these masses are the b₃/y₂ fragment and the phenylalanine fragment (F) respectively. This ﬁgure also shows the a4

fragment at 397, the a1 fragment at 136 and the b2 fragment at 221. The fragments at 380, 323, 233 and 𝑚/𝑧 = 217 represent the a₄ fragment with losses of NH₃, NH₃+ G, NH3 + F and NH3 + Y respectively. Other internal fragments are fragments at 262 (GGF), 205 (GF) and 177 (GF − CO). The peak at 449 appears due to the loss of the tyrosine sidechain. The peak at 318 represents the b₄ fragment with loss of the tyrosine sidechain.

The spectra at energies between 10 - 16 eV show dramatic diﬀerences with respect to those at 8 and 9 eV. Firstly, the relative intensity increases by a factor of 15, probably because the molecules can be photoionized in this region. Furthermore, fragments are observed which are not present in the 8 and 9 eV spectra. Peaks appear at 86, 91 and 107 which are identiﬁed as leucine (L), phenylalanine (F) and tyrosine (Y) sidechain residues. Other new fragments are visible at 295 and 146/147. The most important fragments in this region are the phenylalanine residue at 120, the tyrosine sidechain residue at 136 and the internal glycine-phenylalanine (GF) fragment at 205.

(35)

CHAPTER 4. RESULTS AND DISCUSSION

Tyrosine (Y) Glycine (G) Glycine (G) Phenylalanine (F) Leucine (L)

b1(164) b2(221) b3 (278) b4 (425)

y4(393) y3(336) y2(279) y1(132)

a1(136) a2(193) a3(259)

a4(397)

Figure 4.1: Schematic view of the Leu-Enkephalin molecule with its ﬁve amino acid building blocks indicated. Also the most important fragments and their masses are shown.

The spectra at 20, 25 and 30 eV belong to a different dissociation region. New fragments are observed at 113, 131, 171, 187, 291 and 318. Interesting is the fact that these peaks could be identified as known a,b and c fragments (figure 4.1) with loss of a tyrosine sidechain residue. At the same time this tyrosine sidechain appears as the strongest peak in the spectra at 𝑚/𝑧 = 107. The mentioned fragments are b2 − 107, c2− 107, b₃− 107, c₃− 107, a₄− 107 and b₄− 107 respectively.

The spectra at 36 and 40 eV show that the cross sections at higher photon energies are signiﬁcantly lower. This is in line with the expectations. Valence shell photoionization cross sections are known to decrease brieﬂy with increasing photon energy [36]. The peak at 107 is less dominating and the a, b and c fragments with loss of 107 are less intense than in the spectra at 20, 25 and 30 eV.

In ﬁgure 4.3 the loss of leucine enkephalin and the yield of a few fragment ions are plotted as a function of photon energy. The plots have a local maximum at 15 eV and an absolute maximum at 20 eV, except for the phenylalanine fragment with mass 120.

This fragment has an absolute maximum at 15 eV.

4.1.2 Discussion

As mentioned in section 2.4 valence transitions are the dominating excitation process below ionization threshold. In section 2.2 was already mentioned that the peptide bond is expected to contribute to the absorption around the ionization threshold. The peptide bond can be described as a four orbital model with two doubly occupied 𝜋 orbitals (𝜋₁

(36)

0 1,4

0 5

200 400

0 2 4 6 8 10

200 400

0 2 4 6 8 10 0

10

0 20

400 0 20

0 10

0 5

200

100 300 400 500

8 eV

m/z

Relative Intensity

9 eV

10 eV

13 eV

14 eV

15 eV

16 eV

20 eV

25 eV

30 eV

36 eV

40 eV

120 (F)136

(Y) 177

205 (GF)

217

221 233 262

(GGF)

278/279 (b3/y2)

323

318

380 397(a4)

449

107 86 (Y) (L)

146/147

91 113 131

161 171187

193 (a2)

295

291

Figure 4.2: Spectra of Leu-Enkephalin irradiated with 8 - 40 eV photons.

(37)

Normalized ion loss / yield

Energy (eV)

Loss leu-enkephalin Yield 107 (Y)

Yield 120 (F) Yield 136 (Y)

Yield 205 (GF) Yield 278/279 (b3/y2)

5 10 15 20 25 30 35 40 45

0 40 80 120 160 200 240

5 10 15 20 25 30 35 40 45

0 5 10 15 20 25 30

5 10 15 20 25 30 35 40 45

0 5 10 15 20 25 30

5 10 15 20 25 30 35 40 45

0 5 10 15 20 25 30

5 10 15 20 25 30 35 40 45

0 5 10 15 20 25 30

5 10 15 20 25 30 35 40 45

0 5 10 15 20 25 30

Figure 4.3: Loss of protonated leucine enkephalin and yield of fragment ions as a function of energy.

(38)

and 𝜋₂), lone pair orbital (𝑛₀), and antibonding 𝜋^∗₃ orbital. The electronic absorption spectrum of the peptide bond is due to the transitions: 𝑛₀ −→ 𝜋₃^∗, 𝜋₂ −→ 𝜋₃^∗ and 𝜋1 −→ 𝜋₃^∗. The latter is expected to contribute to the absorption around 9 eV [20].

If IVR would be expected to be the main dissociation pathway, the spectra should be directly comparable to surface induced dissociation (SID) studies in which fast higher energy dissociation pathways are eliminated. The SID spectra obtained by Laskin et al [16] show the same fragments as our VUV absorption mass spectrum at 8 eV. How- ever, the fragment yields are significantly different. The main fragments in the VUV spectra are those at 120 and 278/279, the heavier 𝑎4 and 𝑏4 fragments are the strongest contributions in the SID study. This difference suggests that after excitation not only IVR occurs, but also fast dissociation processes take place.

When the photon energy exceeds the ionization threshold, the molecule can be photoionized. An electron from the HOMO can be ejected from the system and a cold ion is formed. The fact that the molecule is cold gives rise to relative low fragment yields in the spectra at 9 and 10 eV. If the photon energy is further increased, deeper lying molecular orbitals can be reached for ionization, which leaves an electronically excited leucine enkephalin ion. The amount of excitation energy which is available for IVR induced dissociation determines the fragment yields. In ﬁgure 4.3 the photofragmentation spectra show a strong increase between 13 eV and a local maximum at 15 eV. This suggests that the required excess of energy above ionization threshold lies around 4 - 6 eV.

The photofragmentation spectra in ﬁgure 4.3 show that most of the fragments have an absolute maximum at 20 eV with a full width half maximum of 10 eV. At the same time, calculations and experiments show a high density of molecular states with binding energies between 10 and 20 eV.

As mentioned before a lot of fragments in the 20 - 30 eV spectra could be identiﬁed as known fragments with an additional loss of the tyrosine sidechain residue 𝑚 = 107.

Furthermore, this fragment itself is observed as the strongest contribution to the afore- mentioned spectra. The fact that this fragment is not observed in pure IVR induced studies like SID, suggests that the loss of this fragment is a result of fast dissociation.

The fact that the 𝑏4 fragment is not stable in the 8 - 9 eV spectra while it is in the 20 eV spectra, in which it has the same intensity as the 𝑎₄− 107 fragment, implies that the tyrosine sidechain loss reduces the internal energy of the system.

4.2 Soft X-ray absorption

The first experiments on soft X-ray absorption have been done using the U49/2 PGM-1 beamline at BESSY II in july 2010. A first mass spectrum is shown in figure 4.4. Small peaks are observed at masses 120, 177, 205, 278/279, 397 and 425. The spectrum was obtained by using a relatively large slit width, which results in a small energy resolution (Δ𝐸 ≈ 3eV).

(39)

0 1

100 200 300 400 500

537 eV

120

425 397

177 205 278/279

m/z

Relative Intensity

Figure 4.4: A spectrum of leucine-enkephalin irradiated with 537 eV photons.

4.2.1 Discussion

The determination of experimental cross sections is a useful tool to indicate the likelihood of interaction between the target molecule and the irradiating photon. The ratios between the theoretical cross sections of VUV irradiation and soft X-ray irradiation will be compared to the ratio between the relative cross sections obtained from the experiments.

Leach et al [37] measured a maximum cross section of = 7.22.10⁻¹⁷cm² for 16.83 eV photons on acetatic acid, which has a molecular mass of 60 Da. Scully et al [38] found a cross section of = 6.10⁻¹⁶ cm² when C60was irradiated by 20 eV photons. If we assume that the cross section is roughly proportional to the molecular mass, the cross section of leucine enkephalin with VUV photons is in the order of 6.10⁻¹⁶ cm².

The photoabsorption cross sections for X-ray irradiation could be estimated in a similar way. A cross section of 3.9.10⁻¹⁸cm² has been found for Thymine (𝑚 = 126 Da) around the oxygen K-edge by Akamatsu et al [39]. Therefore, the cross section of leucine enkephalin around the K edge could be estimated in the order of 2.10⁻¹⁷cm². The ratio between the cross sections for VUV and soft X-rays is:

𝜎_{𝑉 𝑈 𝑉} 𝜎𝑆𝑋𝑅

≈ 30 (4.1)

If we want to compare the cross sections of the experiments, the ion yields have to be normalized to the relative target density, the relative number of photons and the relative detection eﬃciency of the MCP detectors.

The cross sections which were measured in the experiments are much lower than the estimated value in equation 4.1. This means that a lot of improvement in the soft X- ray experiments is possible. Probably, the alignment between molecule cloud and beam focus has to be optimezed.

(40)

Chapter 5

Conclusion

For the first time photo fragmentation spectra were obtained by interfacing gas phase aminopeptides with VUV and soft X-ray photons. In the VUV region between 8 and 40 eV different absoption regimes were identified. Below ionization threshold only photoexcitation can occur, which leads to IVR governed dissociation processes on a picosecond timescale and fast, femtosecond timescale dissociation processes through dissociative states. Slightly above ionization threshold also photoionization is possible, which pro- duces an internally cold ionized molecule; In this region relatively low fragment yields are observed. 4-6 eV above ionization threshold, a local maximum in photofragmentation appears. In this regime deeper lying molecular orbitals can be ionized, which leaves an highly excited ion with internal energies of several eV available for the IVR process. An absolute maximum appears at 20 eV, which confirms the expectation that up to 20 eV additional molecular orbitals become available for ionization. For higher energies, 30 - 40 eV, the cross sections decrease.

At photon energies between 20 and 30 eV, fragments are observed which could be identiﬁed as known leucine enkephalin fragments accompanied with the loss of a tyrosine sidechain residue (𝑚/𝑧 = 107). Internal energy estimations show that a fast tyrosine sydechain loss eﬃciently cools the remaining peptide. This observation has interesting implications for the survival of peptides during transport to or survival on the early Earth.

Promising experiments have been done using soft X-ray irradiation. Several known fragments of leucine enkephalin are visible in the spectra. On the other hand, the observed cross sections for soft X-ray absorption appeared to be much lower than were predicted. Probably, the alignment between the focus of the beam and the target was not fully optimized. Possibly, penetrating electrical fields from extraction lenses cause the ion cloud to be slightly off-center. The soft X-ray experiments will be further optimized during a beamtime at Max-lab (Lund, Sweden) in the end of October 2010. Pulsed voltages on the extraction lenses should prevent electrical field disturbance inside the trap.

In Max-lab also soft X-ray absorption around the C, K and O edges of gas phase tetranucleotides will be studied. The soft X-ray induced dissociation of these DNA

(41)

CHAPTER 5. CONCLUSION

fragments are of fundamental interest in the context of clinical radiation therapy and radiation damage in living cells. The results can be compared to studies in which nucle- obases and deoxyribose are interfaced with soft X-rays to get new insighs in the relevance of mechanisms in relative small building blocks to more complex biomolecules.

The data can also be compared with keV induced dissociation of tetranucleotides, on which a systematic study will be performed with the Paultje setup, in the context of heavy ion radiation therapy. Tetranucleotides and larger DNA fragments will be positively as well as negatively electrosprayed and interfaced with ions from the Electron Cyclotron Resonance Ion Source (ECRIS) in the Atomic and Molecular Physics group in Groningen. For this purpose H⁺, He²⁺ and He²⁺ of energies varying 5 to 40 eV will be used.

878 =@ IBJ :H=O FDJ=>IHFJE B C=I FD=IA FHJ=JA@ FAFJE@AI

University of Groningen

Master Thesis

VUV and soft X-ray photoabsorption of gas phase protonated peptides

Atomic and

Molecular

Physics

Contents

Chapter 1

Introduction

Chapter 2

Theory

2.1 Molecular orbitals

2.2 Molecular electronic transitions

2.3 Photoabsorption around the oxygen K-edge

2.4 VUV photoabsorption

2.5 Intramolecular Vibrational energy Redistribution

Chapter 3

Experimental Setup

3.1 Electrospray Ionization

3.2 The RF ion funnel

3.3 Quadrupole ion guides and mass ﬁlters

3.4 The Paul trap

3.5 Particle detection and data acquisition

3.6 VUV and soft X-ray beamlines at the third generation synchrotron BESSY II

Chapter 4

Results and Discussion

4.1 VUV absorption

4.2 Soft X-ray absorption

Chapter 5

Conclusion

878 =@ IBJ :H=O FDJ=>IHFJE B C=I FD=IA FHJ=JA@ FAFJE@AI

University of Groningen

Master Thesis

VUV and soft X-ray photoabsorption of gas phase protonated peptides

Atomic and

Molecular

Physics

Contents

Chapter 1

Introduction

Chapter 2

Theory

2.1 Molecular orbitals

2.2 Molecular electronic transitions

2.3 Photoabsorption around the oxygen K-edge

2.4 VUV photoabsorption

2.5 Intramolecular Vibrational energy Redistribution

Chapter 3

Experimental Setup

3.1 Electrospray Ionization

3.2 The RF ion funnel

3.3 Quadrupole ion guides and mass ﬁlters

3.4 The Paul trap

3.5 Particle detection and data acquisition

3.6 VUV and soft X-ray beamlines at the third generation synchrotron BESSY II

Chapter 4

Results and Discussion

4.1 VUV absorption

4.2 Soft X-ray absorption

Chapter 5

Conclusion

878 =@ IBJ :H=O FDJ=>IHFJE B C=I FD=IA FHJ=JA@ FAFJE@AI