Rydberg carbon clusters prepared by pulsed field recombination

The technique of pulsed field recombination is used to recombine a free electron with carbon cation clusters, ranging from C to C. It is shown that after recombination the electron is left in a highly excited state. The recombination method 60 50 Ž . produces a Rydberg state n;180 in a large carbon cluster, such as C , with an electronic energy spread of ;1 K. We 60 measured a recombination efficiency of ;6%. q 2000 Elsevier Science B.V. All rights reserved. The study of core excited cations has gained interest with the introduction of zero kinetic energy Ž . w x ZEKE 1,2 , mass analyzed threshold ionization Ž . w x MATI 3 , and photoinduced Rydberg ionization Ž . w x PIRI 4 spectroscopy. In PIRI spectroscopy, a core electronic transition is photoexcited in a Rydberg molecule, which then decays by rapid autoionization. Because Rydberg molecules only weakly interact with optical radiation, the absorption wavelengths for a Rydberg molecule are very similar to those of the corresponding cation. A similar technique developed w x by Fujii et al. is called IRrPIRI 5,6 in which the OH stretching vibration of cations is investigated by means of exciting Rydberg molecules with infrared Ž . IR radiation. There, the coupling of the vibrational with the electronic energy leads to autoionization. Not all molecular systems are accessible for these studies because the preparation of a Rydberg molecule is not always trivial, due to the fact that at non-zero temperatures of the molecules many vibra) Corresponding author: Fax: q31-20-668-4106; e-mail: [email protected] tional degrees of freedom in the system are occupied. Therefore, until now, it has not been possible to w x produce Rydberg states of large carbon clusters 7,8 Ž . such as C , nor has the IRphotoelectron cation 60 Ž . spectrum with the technique of IRr PIRI spectroscopy been measured. This would be a major w x achievement in the IR-spectroscopy of fullerenes 9 . In this Letter, we report on the observation of RydŽ berg carbon clusters in a Rydberg state ns180, . Dnf20 , by means of pulsed field recombination Ž . w x PFR 10,11 . Previously we have experimentally and theoretiŽ . cally shown that an atomic rubidium ion can be recombined with a free electron by means of the PFR scheme, the recombined atom being left in a highly excited state with a mean principal quantum Ž . w x number of n;180 "20 10,11 . Classical calculations were in excellent agreement with the experimental observations. These calculations showed that the dynamics of the PFR scheme was unaffected by the non-hydrogenic core of the rubidium atoms, thus leading to the prospect that the PFR scheme is a universal method to recombine a free electron with any sort of ion. It was proposed that the PFR scheme 0009-2614r00r$ see front matter q 2000 Elsevier Science B.V. All rights reserved. Ž . PII: S0009-2614 00 00516-9 ( ) C. Wesdorp et al.rChemical Physics Letters 323 2000 192–197 193 can be used to produce 100–1000 cold anti-hydrogen w x atoms in a single experiment 10,11 in the already realized combined trap for particles of opposite w x charge of Gabrielse et al. 12 . In this Letter, we show that the PFR scheme can also be used to recombine a free electron with an ionic carbon clusŽ q q . ter C –C . An experimental method is presented 50 60 Ž . where excited states in ‘hot’ clusters ;750 K are prepared with an electronically excited state energy spread on the order of 1 K. This observation opens Ž the way for photoelectron spectroscopy PIRI, IR. PIRI on large ionic carbon clusters. The PFR scheme can best be described as the time inverse of pulsed field ionization of Rydberg w x atoms 13 and is schematically sketched in the upper panel of Fig. 1. An ion is situated in a static electric field. The static electric field modifies the Coulomb Ž . potential such that a saddle point is created Fig. 1a . If an electron passes over the saddle point in the modified Coulomb potential, it will take a small, but Ž . Fig. 1. Schematic representation of the PFR scheme. a An electron travels toward a positive ion in a static field; the turning Ž . point for the electron is within the ion cloud. b When the electrons are at the turning point, the electric field is quickly Ž . turned off. c If the field is turned off while the electron is near the ion, the electron remains bound. A small electric field bias of 200 mVrcm removes the electrons that did not recombine with an Ž . Ž ion. d The efficiency number of recombined carbon clusters . divided by the number of free ionic carbon clusters as a function of the delay of the quick turnoff. not negligible, amount of time to return to the saddle Ž . point and escape from the ion Fig. 1a . If the static Ž . field is turned off Fig. 1b before the electron returns to the saddle point, it will remain bound in a Ž . highly excited state Fig. 1c . The time required for the electron to travel from the saddle point, to the nucleus, and back to the saddle point is roughly 1–2 ns for the fields and energies used in this experiment. In Fig. 1d the experimentally determined efficiency Ždensity of recombined carbon clusters atoms di. vided by the density of free ionic clusters of this scheme is depicted as a function of the delay of the fast field turnoff. This delay is with respect to the time when the free electron has its turning point in the electric field. Clearly, a maximum number of recombination events is recorded at zero delay. Note Ž . y2 that efficiencies up to 6 "4 =10 are obtained. The PFR scheme was experimentally realized as w x follows: a pulsed electron source 14 was created by means of photoionizing lithium atoms in a static Ž electric field of 1.50 Vrcm, with a pulsed 9 ns . Ž y1 . duration narrowband Dl-0.2 cm dye laser Ž . operating at 10 Hz . Typically electron pulses of 9 Ž . 4 ns duration, with about 5"2 =10 electrons in a volume of 0.01 mm were produced. The electric Ž field is created by two parallel capacitor plates sep. aration: 10.0 mm over which a voltage is applied Ž . Fig. 2a . The ionizing laser pulse is focussed, creating a focus of 35 mm diameter. Focussing the laser Ž . in the direction of the electric field z-direction is important since the focus size determines the energy spread of the created electron pulses. After ionization, the free electrons are pushed towards the anode Ž . plate connected to ground through which a small hole is drilled. A grid covering the hole minimizes the distortions on the electric field. After this region, the electrons enter a field-free region of 15.0 mm, after which another set of parallel capacitor plates Ž . with small holes, covered by a grid is situated. Since the final plate is on y1.50 V, the electrons turn around halfway between these plates, separated by 10.0 mm. A cloud of ionic carbon clusters is awaiting the electrons in this region. A magnetic Ž . field was applied B s2.5 mT parallel to the direcz tion of propagation to minimize the perpendicular spread of the electrons during their travel to the ions. The ionic carbon clusters were produced 500 ns before the photoelectrons were created by photoion( ) C. Wesdorp et al.rChemical Physics Letters 323 2000 192–197 194 Fig. 2. Upper panel: Experimental setup. A static electric field of 1.50 Vrcm pushes the electrons towards the ion region, where the ions await the electrons in an opposite electric field. The cathode plate is connected to a fast pulse generator so the field can be turned off within 1.0 ns. The anode plate is connected to a slow pulse generator so that after the quick turnoff the population distribution of the created Rydberg atoms can be probed. Lower panel: Mass spectrum of the produced carbon clusters by photoionization of C molecules with 266 nm. 60 Ž izing and fragmenting gas-phase C molecules T; 60 . 750 K in a static electric field of 1.50 Vrcm. The gas-phase C molecules were produced by a resis60 tively heated oven. We photoionized C molecules 60 Ž . Ž . IPs7.6 eV by the fourth harmonic 266 nm of a Nd:YAG laser. A typical mass spectrum of the created ions is depicted in Fig. 2. The multiple ionization with 266 nm mainly produces the higher ionic carbon clusters. Successful recombination experiments were also performed where ionic carbon clusters were produced by photoionizing C molecules 60 Ž . with the second harmonic 532 nm , resulting in average cluster sizes of C–C . The smaller cluster 7 size, in the 532 nm experiment, can be understood by the fact that the absorption of multiple photons opens up another decay channel of the C molecules: 60 fragmentation. Also the effect of delayed ionization is seen in the C mass peak, which is due to 60 thermionic emission of the highly excited molecular w x core 15,16 . Detailed studies of the fragmentation and ionization dynamics of C are reported in Ref. 60 w x 15–19 . The decay time of this delayed ionization is measured to be ;240 ns. By producing the ionic carbon clusters 500 ns prior to the arrival of the electron pulses assures that the major fraction of the excited C molecules have ionized at the moment of 60 the recombination event. The static electric field of 1.5 Vrcm assures that all the electrons of the ionized carbon clusters have left this interaction region at the time the lithium photoelectrons are created. However, the heavy ionic carbon clusters are virtuŽ ally standing still during our experiment D ls10 mm in 100 ns due to thermal velocity and D ls0.1 . mm in 100 ns due to the static electric field . The ionic carbon cloud resembles a cylinder with a radius of 35 mm in the zand y-direction and a length of 10 mm in the x-direction, with a density of Ž . 7 y3 2.4 "0.4 =10 cm . In the ion source region, the electrons are decelerated by the electric field of 1.50 Vrcm, and have their turning point at the position of the ion cloud. At that time, the field is turned off by dipping the voltage on the cathode plate from y1.50 Ž . V to y0.20 V in 1.0 "0.2 ns. The electric field turnoff is realized by connecting an impedance Ž matched, fast pulse generator Stanford Research . Systems, model DG 535 to the cathode plate. The density of ionic carbon clusters is kept low with respect to the density of free electrons to prevent the effect of ‘trapping’ the electrons in the attractive potential of the ion cloud. For the given volumes and densities, the electrons feel an attractive field of -30 mVrcm at the edge of these volumes. Such a low density plasma is therefore not stable when the ion region is biased with an electric field of 200 mVrcm. The experiments were performed in a vacuum chamber with a background pressure of 5= 10 Torr. Most recombination processes are characterized by cross-sections. In our PFR scheme, it is more appropriate to give the volume in which the electron has to be at the moment of the quick turnoff to recombine with the ion. We define this volume as the interaction volume. The interaction volume is the volume of space for which an electron with an initial Ž . velocity 0,0,Õ will recombine with the ion after z the electric field is ramped down to 200 mVrcm; it is not necessary to use a distribution in velocity due ( ) C. Wesdorp et al.rChemical Physics Letters 323 2000 192–197 195 to the very low effective temperature of the electron source. This interaction volume can be estimated with the following formula: ` X X X N t sr r V f t yt V t d t 1 Ž . Ž . Ž . Ž . H ion e over int y` Ž . where N t is the number of recombinations at a Ž . certain delay t of the quick turnoff, r is the ion ion density, r is the electron density, V is the e over macroscopic overlap volume of the electron and ion clouds, and 2 4ln2 yt 4ln2 f t s exp Ž . ( 2 2 ž / pa a is a function which describes the time profile of the Ž electron pulses a is the duration of the electron . pulses, 9 ns . In the case of ionic carbon cluster Ž q q . production with 266 nm C –C , we have mea60 50 Ž . sured typically 15 "5 recombination events per 5=10 electrons in an overlap volume of 0.01 mm 7 y3 Ž . and an ion density of 2.4=10 cm Using Eq. 1 , Ž this observation yields an interaction volume is 1.3 . y11 3 "1.2 =10 cm . A theoretical estimate of V is int obtained by solving Newton’s equations, with a force Ž . Ž Ž . Ž . . equal to: F t sq E t qz t =Brc . The electric Ž . field E is the superposition of the Coulomb potential of the ions and external field, z is the velocity of the electrons and B is the magnetic field. An interaction volume of 2.5=10 cm is theoretically calculated for a 9 ns electron pulse, which is more than the estimation from the experimental data, the explanation for this will be discussed below. We note that the spatial extent of such an interaction volume corresponds to Rydberg states around nf180. The radius of the interaction volume is two orders of magnitude smaller than the average distance between neighboring ionic carbon clusters, which means perturbations from unrecombined ionic carbon clusters are negligible in this experiment. A detailed picture of the final state distribution of the pulsed field recombined Rydberg clusters was obtained with the technique of selective field ionizaŽ . w x tion SFI 13,20 . A Rydberg state of an atom or molecule can be ionized if an electric field is applied. Each state of an atom or a molecule has an associated field strength at which it ionizes. The relation between the value of the static electric field at which ionization classically occurs and the principal quantum number, n, of a state is in atomic units: Ž .2 4 Fs Er2 s1r16n . When the electric field is slowly ramped in time and the field at which Rydberg molecules decay is monitored, the populated Rydberg states can be approximately deduced; Ž .2 changes from Fs Er2 arise because the energy of the electron changes during the ramp and because the electron is not immediately stripped from the ion when the energy condition is satisfied. In our experiment, we use a field ramp of 1.50 Vrcm in 1.8 ms, 900 ns after the fast field turnoff. This was realized by connecting a voltage ramp to the anode plate of the ion region. The field ramp pushes the field ionized electrons towards the cathode plate through which a hole was drilled covered with a grid. Directly after this plate a set of micro-sphere-plates Ž . MSP is situated that record the ionized electrons. In this way, we record at which times, and hence at what field strengths, electrons are being field ionized. During the 900 ns after the quick turnoff and before the SFI ramp, effects of three body recombination and radiative recombination are ruled out Ž y1 . since the rates of these processes 1 s are far too low at our densities. In Fig. 3, the Rydberg state distribution as a result of the PFR by the 1.0 ns turnoff is depicted for Ž . recombination of the carbon clusters C –C , 50 60 Ž . Fig. 3. Calculated state distribution dotted line of the recombined Rydberg clusters, compared with experimental state distriŽ . bution solid line . The experimentally determined interaction volume is varied within its uncertainty to optimize the comparison