A lithographically-etched chip containing 1254 microscopic electrodes at a\ncenter-to-center spacing of 40 micron is used to produce an array of electric\nfield strength minima 25 micron above the surface with a spacing between\nneighboring minima of 120 micron. These minima are produced by applying a\nseries of six potentials to the electrodes. For polar molecules in low field\nseeking states, the electric field minima are potential energy minima and thus\nact as traps. By varying the potentials in a controlled manner, the field\nminima can be translated with respect to the electrodes at a constant distance\nfrom the surface. Experiments were performed with CO in the metastable a\n3-Pi-1, v = 0, J = 1 state, which has a strong, linear Stark shift with low\nfield seeking components, a 6 eV internal energy that enables easy detection\nthrough an Auger process, and a radiative lifetime of 2.6 ms, which is long\nenough for the experiments here. CO molecules in a pulsed molecular beam are\ndirectly excited to the metastable state with a 5 ns laser pulse. The\nmetastable CO molecules pass through a skimmer and arrive to the chip\ntraveling parallel to the surface and perpendicular to the electrodes. While\nabove the chip, the molecules are manipulated by the electric fields and\nsubsequently ejected from the other side toward the detector. Metastable 12-CO\nmolecules have been guided over the chip at constant velocities and\ndecelerated. As the final velocity decreases (i.e. the deceleration\nincreases), the decelerated molecules arrive at the detector successively\nlater, and the number of decelerated molecules reaching the detector also\ndecreases. While this effect is expected, the fraction of molecules lost is\nmuch greater than trajectory simulations predict. Most of this excess loss at\nhigher decelerations results from non-adiabatic transitions to untrapped\nstates. At zero electric field, the low field seeking level of the a 3-Pi-1, v\n= 0, J = 1 state of 12-C-16-O is exactly degenerate with a level that is only\nweakly affected by the fields. As the molecules pass near the zero field\nregion at the center of a microtrap, they can transition to this level, after\nwhich they are lost from the trap. One way to avoid this loss channel is to\nuse 13-C-16-O instead of 12-C-16-O, which results in a breaking of this\ndegeneracy by 50 MHz due to hyperfine structure. Because of this splitting,\nensembles of metastable 13-C-16-O molecules can and have been decelerated to\nzero velocity above the chip and held for variable durations ranging from 0 ms\nto 2.5 ms. It is observed that fewer molecules are detected after longer\ntrapping times on the chip, but the rate of this loss is consistent with the\n2.6 ms radiative lifetime of the metastable CO molecules. Further experiments\ndemonstrate that the Zeeman splitting due to an external magnetic field\nperpendicular to the electric field is a valid and more general way to prevent\nnon-adiabatic losses in 12-CO.
A lithographically-etched chip containing 1254 microscopic electrodes at a\ncenter-to-center spacing of 40 micron is used to produce an array of electric\nfield strength minima 25 micron above the surface with a spacing between\nneighboring minima of 120 micron. These minima are produced by applying a\nseries of six potentials to the electrodes. For polar molecules in low field\nseeking states, the electric field minima are potential energy minima and thus\nact as traps. By varying the potentials in a controlled manner, the field\nminima can be translated with respect to the electrodes at a constant distance\nfrom the surface. Experiments were performed with CO in the metastable a\n3-Pi-1, v = 0, J = 1 state, which has a strong, linear Stark shift with low\nfield seeking components, a 6 eV internal energy that enables easy detection\nthrough an Auger process, and a radiative lifetime of 2.6 ms, which is long\nenough for the experiments here. CO molecules in a pulsed molecular beam are\ndirectly excited to the metastable state with a 5 ns laser pulse. The\nmetastable CO molecules pass through a skimmer and arrive to the chip\ntraveling parallel to the surface and perpendicular to the electrodes. While\nabove the chip, the molecules are manipulated by the electric fields and\nsubsequently ejected from the other side toward the detector. Metastable 12-CO\nmolecules have been guided over the chip at constant velocities and\ndecelerated. As the final velocity decreases (i.e. the deceleration\nincreases), the decelerated molecules arrive at the detector successively\nlater, and the number of decelerated molecules reaching the detector also\ndecreases. While this effect is expected, the fraction of molecules lost is\nmuch greater than trajectory simulations predict. Most of this excess loss at\nhigher decelerations results from non-adiabatic transitions to untrapped\nstates. At zero electric field, the low field seeking level of the a 3-Pi-1, v\n= 0, J = 1 state of 12-C-16-O is exactly degenerate with a level that is only\nweakly affected by the fields. As the molecules pass near the zero field\nregion at the center of a microtrap, they can transition to this level, after\nwhich they are lost from the trap. One way to avoid this loss channel is to\nuse 13-C-16-O instead of 12-C-16-O, which results in a breaking of this\ndegeneracy by 50 MHz due to hyperfine structure. Because of this splitting,\nensembles of metastable 13-C-16-O molecules can and have been decelerated to\nzero velocity above the chip and held for variable durations ranging from 0 ms\nto 2.5 ms. It is observed that fewer molecules are detected after longer\ntrapping times on the chip, but the rate of this loss is consistent with the\n2.6 ms radiative lifetime of the metastable CO molecules. Further experiments\ndemonstrate that the Zeeman splitting due to an external magnetic field\nperpendicular to the electric field is a valid and more general way to prevent\nnon-adiabatic losses in 12-CO.
A microstructured array of 1254 electrodes on a substrate has been configured to generate an array of local minima of electric field strength with a periodicity of 120 μm about 25 μm above the substrate. By applying sinusoidally varying potentials to the electrodes, these minima can be made to move smoothly along the array. Polar molecules in low-field seeking quantum states can be trapped in these traveling potential wells. Recently, we experimentally demonstrated this by transporting metastable CO molecules at constant velocities above the substrate (Meek et al 2008 Phys. Rev. Lett. 100 153003). Here, we outline and experimentally demonstrate how this microstructured array can be used to decelerate polar molecules directly from a molecular beam. For this, the sinusoidally varying potentials need to be switched on when the molecules arrive above the chip, their frequency needs to be chirped down in time and they need to be switched off before the molecules leave the chip again. Deceleration of metastable CO molecules from an initial velocity of 360 m s -1 to a final velocity as low as 240 m s -1 is demonstrated in the 15–35 mK deep potential wells above the 5 cm long array of electrodes. This corresponds to a deceleration of almost 10 5 g , and about 85 cm -1 of kinetic energy is removed from the metastable CO molecules in this process.
The subject of this thesis’s work is research of cold molecules on a microchip. On the one hand, cold and ultracold molecules might allow high- precision measurements of molecular properties and the determination of their behavior at the quantum level. On the other hand, one could try to study highly controlled chemical reactions or to build a quantum computer with their help. In the experiments presented here, electrically neutral, polar molecules in low-field-seeking states are trapped and decelerated on a microchip with the help of electric fields. This exploits the Stark effect: the energy level shift of atoms and molecules in an electric field. After manipulation on the chip the molecules can be detected quantum-state specifically via the REMPI method. At first, CO molecules were prepared in the upper Λ-doublet component of the a3Π1, v = 0, J = 1 state, trapped on the chip and get vibrationally excited by an infrared laser in the middle of the chip to the J = 1 or J = 2 rotational state of the vibrational state v = 1. This could be shown experimentally and was analyzed theoretically for molecules which were trapped at all times. Secondly, CO molecules were detected on the chip and the structure of previously trapped molecule clouds of several of the only 120 μm spaced traps was spatially resolved. For this purpose the chip was extended with an ion lens setup, which enabled the determination of the temperature of the clouds by following the free ballistic expansion of the molecules in time; either by comparing the data to trajectory simulations (T = 11 ± 1 mK) or by applying a ballistic expansion model (T = 13 ± 3.5 mK), comparable to the method used by the ultracold atom community. The spatial imaging enabled the observation of a phasespace manipulation scheme, which adiabatically expanded the molecules in the trap and significantly reduces their temperature to approximately one third of their initial value. The third goal was to trap ground state ammonia molecules, which should extend the trapping time and would show that also other molecules could be trapped on the chip. The ammonia isotopologues 14NH3, 14ND3, and 15ND3 in the |vInv = 1, J = 1,K = 1> state of the electronic ground state X were used. No signal could be measured from trapped molecules, which is attributed to large nonadiabatic transitions to untrappable quantum states, leading to severe trap loss. At last, a new chip- based Stark decelerator is presented, which is a miniaturized version of the so-called “traveling-wave” ring decelerator. This chip-based decelerator should be able to trap more molecules compared to the chip that was used so far, and due to improved electronics it should be able to decelerate molecules from molecular beam velocities of approximately 360 m/s down to a standstill, in contrast to large-scale traveling-wave decelerators so far. The chip is still in development to achieve operation with the full specification (voltage). Performance simulations show the usefulness of this design, which needs to be tested in experiments.
A microstructured array of over 1200 electrodes on a substrate has been configured to generate an array of local minima of electric field strength with a periodicity of 120 microm about 25 microm above the substrate. By applying sinusoidally varying potentials to the electrodes, these minima can be made to move smoothly along the array. Polar molecules in low field seeking quantum states can be trapped in these traveling potential wells. This is experimentally demonstrated by transporting metastable CO molecules in 30 mK deep wells that move at constant velocities above the substrate.
Polar molecules in selected quantum states can be guided, decelerated, and trapped using electric fields created by microstructured electrodes on a chip, as shown on p. 1799 by G. Santambrogio et al. Transitions between two of these quantum states can be induced while the molecules are on the chip. CO (a3Π1, v=0) molecules, prepared at the J=1 rotational level, are trapped and guided to the center of the chip. There, they undergo the J=2←J=1 rotational transition driven by narrow-band sub-THz (mm-wave) radiation and, after being guided off the chip, are state-selectively detected.
This thesis describes the design and construction of a novel molecule-surface scattering apparatus. The setup combines a Stark decelerator with a state-of- the-art surface scattering chamber, and allows background-gas free scattering of packets of velocity-controlled polar molecules in selected quantum states from well-defined surfaces with unprecedented energy resolution. Results for scattering of CO (a3Π) in different vibrational quantum states from a clean Au(111) surface are presented. CO is co-expanded with Xe or Ar to yield a supersonic molecular beam which produces fast, internally cold molecules. Excitation of CO to the long-lived electronically excited a3Π1 state makes it accessible to manipulation with electric fields due to the state’s large dipole moment. The complete removal of both carrier gas and remaining ground- state CO is achieved by focusing and deflecting the metastable CO with a switched electrostatic hexapole state selector. The 131-stage Stark decelerator is used to tune the velocity of a part of the metastable CO pulse, with an initial velocity of 360 m/s, in the range from (512 ± 4) m/s – (50 ± 4) m/s, corresponding to an incident translational energy between (307 ± 5) cm-1 and (3.0 ± 0.5) cm-1, respectively.This range can be extended up to (1300 ± 25) m/s, corresponding to an energy of(1980 ± 80) cm-1, by simple modifications to the setup. Prior to collision, the molecules can be laser- prepared in different vibrational and rotational quantum states. State- selective detection of the scattering products makes it possible to determine the quantum state, the scattering angle, and the kinetic energy of the scattered molecules. Therefore, detailed information about the scattering process can be obtained. The design of the machine allows for flexible laser access near the surface, such that a variety of different preparation and detection schemes can be employed. Ultra-high vacuum conditions in the scattering chamber are achieved by baking the entire apparatus, including the decelerator and the hexapole. Because of the carrier gas removal and a total of three differential pumping stages, no pressure rise is detected in the scattering chamber due to the incident molecular beam. Scattering experiments of CO (a3Π) on Au(111) have been performed. The scattering process leads to electron emission because the metastable excitation energy exceeds the work function of the surface. The electron emission yield, i.e. the number of electrons emitted per incident metastable particle, has been measured to be 0.13 ± 0.05. A laser ionization-depletion method is used which makes it possible to directly compare the intensity of the metastable CO ionization signal to the intensity decrease of the electron emission signal. It was also shown that vibrationally excited metastable CO leads to a higher electron emission yield. By preparing different vibrational distributions of CO in the a3Π electronic state via Franck-Condon pumping, the electron emission yield was determined for groups of vibrational levels. The first three vibrationally excited levels have an increased electron emission yield which on average is a factor of 1.5 larger than that of the vibrational ground state. Averaged over the vibrational levels with quantum number v = 4, 5, 6 , this factor is approximately 2.6 . In the range between 93 m/s and 512 m/s, the electron emission was observed to be independent of the velocity of the incident molecules. When the surface temperature is raised from 250 K to 900 K, the intensity of the electron emission signal increases by 27 %. These results are not in agreement with the models currently used to describe the de-excitation of electronically excited molecules at metal surfaces. This demonstrates the need for more theoretical work to incorporate the role of vibration into the description of the de-excitation mechanism.
A Stark decelerator is an effective tool for controlling motional degrees of freedom of polar molecules. Due to technical limitations, many of the current Stark decelerators focus on molecules in low-field-seeking quantum states and are built based on a fixed electrode size and spacing (type-A architecture). Here, we present an alternative method based on a different architecture, so-called type-B, with microstructured electrodes and a simpler electric field pulse timing with the prospect of producing cold, quasi-CW molecular beams. We demonstrate with simulations the feasibility of this method and show that a 2 cm decelerator consisting of 100 stages can bring molecules from 200 m/s to a near standstill in about 150 microseconds. Subsequently, the molecules can be reaccelerated to 200 m/s in a phase-stable manner. Our simulations focus on the longitudinal transmission leaving the problem of a transverse confinement for a future study. This work points to a possible way of integrating polar molecule accelerators with solid-state electronics.