We demonstrated a simple method to photo-ionize barium atoms using 791 nm diode laser together with 310 nm UV LED. It solved the bottle-neck problem of previous method using 791 nm diode laser and 337 nm N 2 laser, whose ionization rate was limited by the repetition rate of N 2 laser. Compared with previous method, it has advantages of high efficiency together with simple and cheap setups. By tuning the frequency of 791 nm laser to be resonant with the desired isotope, isotope selective photo-ionization has been realized.
© 2011 OSA
Ion traps have become important tools for many areas of physics, including precision frequency standard [1, 2], precision measurement of fundamental physical constants [3, 4], mass spectroscopy , and quantum information science [6, 7]. As an important, practical step of operating an ion trap, a simple ionization method will greatly simplify the experimental setup; a highly efficient and isotope selective method will supply a pure ionic medium and a clean trap environment. There are several methods to ionize atoms, including electron beam bombardment, two-photon ionization [8–18], and photoelectric ionization . Electron beam bombardment method can be applied to ionize any atomic or molecular species, but it will cause charge buildup on insulating surfaces of the trap which may degrade the trap performance over time and enhance micromotion heating of the trapped ions. In addition, this method can not realize isotope selective ionization of the target atoms. Photoelectric ionization process was observed by A.V. Steele et al when they using UV lamp to ionize the barium atoms , and it have the same disadvantages as electron beam bombardment method. With the virtue of isotope selection, high efficiency and no charge buildup, two-photon ionization method has been adopted by many groups to load the particular ions, especially for applications of atomic clocks and quantum computing. However, it also has the disadvantage of using a complicated and expensive experimental setup compared with the other two methods. Normally, these two-step excitation transitions always correspond to two laser systems, and sometimes need expensive laser system such as the dye or diode pumped second-harmonic-generation (SHG) laser system [15, 17, 18], and even the frequency-quadrupled mode-locked Ti:sapphire laser . Recently, some groups [11, 16, 18] used high power light emitting diode (LED) for the second excitation of photo-ionization process of calcium, and isotope-selectively trapped a small amount of 48 Ca +, 43 Ca + and 40 Ca + ions, respectively.
For the case of barium, as shown in Fig. 1, A. V. Steele et al.  first demonstrated photo-ionization loading of a barium ion trap with 6s 2 1 S 0 → 6s6p 3 P 1 transition at 791 nm followed by the excitation into continuum by 337 nm pulsed laser, and shown it is more efficient than electron bombardment ionization. In this scheme, the 791 nm laser was supplied by an extended cavity diode laser (ECDL), and the 337 nm pulsed laser was supplied by a commercial N 2 laser (SRS NL100) which provided 170μJ pulses with a width of 3.5 ns and maximum repetition rate of 20 Hz. We have also employed this photo-ionization scheme (scheme 1) in our experiment, and found that the ionization rate was limited by the repetition rate of N 2 laser, which will be analyzed in detail in this paper. The ionization rate or loading rate is an important parameter for ion traps, because a high loading rate is essential to trap a large number of particular ions. High loading rate can reduce the running time of the atomic oven, which can greatly reduce the quantity of materials sputtered onto the trap electrodes. Ref.  has shown that clean electrode surfaces will reduce the heating rate of trapped ions from the motional ground state. For isotope selective loading of rare isotopes, high loading rate is necessary to overcome the charge exchange processes with the abundant isotope from the atomic beam  (for example, 138 Ba in barium atomic beam). In this paper, we analyze the bottleneck of the Ba ions’ loading rate using 791 nm laser and 337 nm N 2 laser, and demonstrate that using a cheap UV LED operating at 310 nm to replace the N 2 laser, a higher loading rate can be achieved. In our experiment, using a 310 nm LED (SET, Inc., UVTOP310-TO39HS) with driving current of 19 mA (corresponding to the optical power density about 250μW/cm 2 at the trap center), a 434 ions/sec average loading rate for 138Ba+ can be obtained. This loading rate is proportional to the power of the 310 nm LED, and can be increased by using more UV LEDs, using high power UV LED, and optimizing the imaging system of the 310 nm beam. With this photo-ionization scheme (scheme 2), different barium isotopes can be loaded and cooled to crystallization.
2. Experimental Procedure
The experiment was performed in a linear quadruple trap which was composed by 12 electrodes with gold coating. As shown in Fig. 2(a), all of the even numbered electrodes are applied with a rf voltage of the form Vrf cos(Ωt). A typical rf driving frequency is Ω = 2π × 1.96 MHz, with amplitude Vrf = 100V. All odd numbered electrodes are ac grounded. On the eight end electrodes (electrodes 1–4 and 9–12), a positive dc voltage (Uz = 120V) is added to the rf voltage (or ac ground) through the LC coupling to provide axial confinement. These eight end electrodes serve as the end-caps. The length of the center electrodes and end-caps are 25 mm and 17 mm, respectively. The diameter of all electrodes are 2R=8 mm, and the distance between two diagonal electrodes’ inner surface is 2r 0 = 7 mm. The barium oven with 3 mm diameter points the trap center along ẑ – x̂ direction, and the distance between the oven exit and the trap center is 40 mm. The 791 nm laser beam with 5 mm diameter is directed into the trap along –x̂ – ẑ direction, which interacts with the transition 6s 2 1 S 0 → 6s6p 3 P 1 of neutral barium. The 337 nm pulse laser from N 2 laser and 310 nm light from the UV LED can be switched alternately by a flip mirror. They are directed into the ion trap along the x̂ + ẑ direction which is perpendicular to the barium atomic beam to minimize the Doppler shifts. The 337 nm pulse beam is focused down to a waist of 1 × 3 mm by a cylindrical lens with a focal length of 300 mm. The UV LED has a hemispherical output coupling lens, therefore its output beam has a relatively small emission angle of 7°. After focusing by a lens of 75 mm focal length, it is focused down to spot of 4 × 4 mm, which is the image of the LED’s chip. At the typical operating current of 19 mA, its emission spectrum, centered at 310 nm, has a width of ΔλFWHM = 30nm. Considering the coupling loss of imaging system and the delectric coating of the vacuum window (AR coating for 493 nm+650 nm+791 nm +337 nm; reflectivity at 310 nm > 10% ), the optical power of 310 nm light at the trap center is about 40 μW (power density is 250μW/cm 2). The cooling laser at 493 nm with power P 493 = 4.5mW and repumping laser at 650 nm with power P 650 = 4mW are coupled into a single mode fiber and directed into the trap along the –ẑ direction. The diameters of these two beams are 2 mm inside the trap.
We choose 138 Ba as an example to measure the loading rate of two photo-ionization schemes. During the loading process, we first heat the barium oven at 7 A, which corresponding to the oven temperature of approximately 550 °C. After 30 s heating, which ensures a uniform atomic beam, we turn on all of the light beams simultaneously (493 nm + 650 nm + 791 nm + 337 nm, or 493 nm +650 nm +791 nm +310 nm ). The 493 nm cooling laser is about −100 MHz detuning to the transition 138 Ba + 6S 1/2 → 6P 1/2; the 650 nm repumping laser is resonant with the transition 138 Ba + 5D 3/2 → 6P 1/2; and the 791 nm laser is resonant with the transition 138 Ba 6s 2 1 S 0 → 6s6p 3 P 1. After illuminating the barium atomic beam for a duration of τ =10 s, we turn off the photo-ionization beams (791 nm + 337 nm or 791 nm + 310 nm) and the barium oven heating current, simultaneously. Through slowly decreasing the frequency detuning of cooling laser, we can obtain the 138 Ba + ion crystals in the trap center. The ion crystals are imaged onto a charge coupled device (CCD) camera (Olympus CC-12) with a 3× imaging system. The entire imaging system is mounted on a three-axis transition stage with a motional resolution of 0.1μm. By moving the imaging system with the transition stage, we can measure the size of whole ion crystals, and therefore the distance between two neighboring ions, a. Figure 2(b) shows the ion crystals we obtained using photo-ionization scheme 2, when the 791 nm laser power is P 791 = 22mW and the LED current is 19 mA. The trapped ion crystals have a shape of ellipsoid with the long axis length of 2.18 mm and short axis length of 0.23 mm, which corresponds to a spatial volume of about Vcry = 6.15 × 10−5 cm 3. The measured distance a is about 30μm, which corresponds to the unit ion volume of about Vunit = 1.41 × 10−8 cm 3. Using the formula Vcry/(Vunit τ), we obtain a loading rate for this ionization process of ∼434 ions/s. This ionization process is repeated several times, and the average loading rate can be obtained.
Using this method, we have experimentally investigated the dependence of loading rate on the optical power of 791 nm laser for these two different photo-ionization schemes, respectively. The results are shown in Fig. 3. We can see that the loading rates of 138 Ba + ions tend to be saturated with the increase of 791 nm laser power for both ionization schemes. This means that for our current experimental system, further increasing the 791 nm laser power will not dramatically increase the loading rate of 138 Ba + ions any more. We have also investigated the dependence of loading rates of 138 Ba + ions on the repetition rate of 337 nm N 2 laser for ionization scheme 1 and on driving current of UV LED for ionization scheme 2, for which, the results are shown in Fig. 4. For ionization scheme 1 (Fig. 4(a)), the loading rate of 138 Ba + ions increases linearly with the repetition rate of 337 nm pulse laser. While the problems is, the maximum repetition rate of 337 nm N 2 laser is 20 Hz, which means the 427 ions/s’ loading rate (corresponding to 5.6 × 10−21 ionization probability per photon of 337 nm laser) is the maximum loading rate for our current experimental system using ionization scheme 1. For ionization scheme 2 (Fig. 4(b)), the loading rate of 138 Ba + ions also increases linearly with the driving current of UV LED. The 434 ions/s’ average loading rate (corresponding to 7.1×10− 12 ionization probability per photon of 310 nm light) for 138Ba+ has been achieved when the driving current of UV LED is 19 mA, which is comparable to the maximum loading rate using ionization scheme 1. As the UV LED is very small and easy to operate compared with the N 2 laser, the ionization scheme 2 can simplify the experimental setup for the photo-ionization process; as UV LED is much cheaper than the N 2 laser, we can easily add more UV LEDs to irradiate the barium atomic beam from different directions or use high power UV LED, thus further increasing the loading rate of barium ions.
3. Theoretical Analysis
For the photon-ionization scheme 2, we can use rate equations  to describe each atom’s ionization process.22] gives the photon-ionization cross-section value of σ 310 = 3.7 × 10−17 cm 2. Γ21 is the spontaneous transition rate from state |2〉 to state |1〉, which is about Γ21 = 2π × 47.6kHz; β is the spontaneous decay rate from state |2〉 to two lower D states (6s5d 3 D 1 and 6s5d 3 D 2) of barium atoms (as shown in Fig. 1), which is about β = 2π × 19.6kHz + 2π × 50.6kHz = 2π × 70.2kHz . From the perspective of atom, the photon-ionization beam can be seen as a pulse, and the pulse duration time τL is the mean transition time of atoms through the light beam(τL = 6.4μs at the barium oven temperature of 550°C). With this assumption and the initial condition of n 1(0) = 1, n 2(0) = nI 310(0) = 0, the population probability nI 310 can be solved. During the loading process, we heat the barium oven at 550°C, which corresponds to the atomic beam flux rate of about N 0 = 1.96 × 1012 s −1 at the interaction region interacting with photon-ionization beam. Thus, we can evaluate the loading rate through the formula NL 310 = AηN 0 nI 310, where η is the trapping and cooling efficiency of our ion trap system, A is the abundance of 138 Ba atom (Table 1).
For the ionization scheme 1, the ionization process is different with the ionization scheme 2. First, the maximum interaction time of an atom with the 337 nm laser pulse will not exceed the pulse width of τ 337 = 3.5ns, which is far smaller than the mean interaction time of an atom with 791 nm laser beam. Second, not all barium atoms transmitted through the interaction region will interact with the 337 nm laser pulse. And the interaction time window for an atom to see the 337 nm laser pulse is only ρ = f τ 337 ≪ 1 at repetition rate of f. Consequently, the interaction processes of atom with 791 nm laser beam and 337 nm laser pulse can be approximately seen as two independent processes, and the rate equation of the process interacting with 791 nm laser beam can be written as:22], F 337 is the photon flux of each 337 nm laser pulse at the trap center, in our case, it is about 2.5 × 1024 cm −2 s −1. Other parameters are same as those in Eq. (1).
Using Eqs. (1)–(5), we theoretically calculated the loading rates, and gave the relationship of loading rates with the optical power of 791 nm laser (the gray lines in Fig. 3) for the two ionization schemes, respectively. During the calculation, the only fitting parameter is η. Our fitting results is η = 2.27%, and it is same for two ionization schemes. We also calculated the dependence of loading rate on the repetition rate of 337 nm laser (gray line in Fig. 4(a)) and on the driving current of UV LED (gray line in Fig. 4(b)), respectively. The theoretical results fit the experimental results very well. From these results, we clearly see that the short laser pulse and low repetition rate of the N 2 laser are the limiting factors for ionization scheme one. Hence, the UV LED with cw output light is an appropriate replacement to N 2 pulse laser. As it is very cheap and easy to use, one can easily use more LEDs to realize a higher loading rate of barium ions.
4. Isotope Selection
As the 310 nm light from the UV LED only need to supply enough energy to excite the Ba atoms from the 6s6p 3 P 1 state to continuum, isotope-selective loading can be achieved by tuning the 791 nm laser to be resonant with the desired isotope in both ionization schemes. The abundances of naturally occurring barium isotopes and their isotope shifts respecting to 138 Ba atom  and ion [25, 26] are shown in Table 1. For loading 136 Ba +, and 134 Ba +, we tune the 791 nm laser frequency to 109.2 MHz and 122.3 MHz detuning to the 6s 2 1 S 0 → 6s6p 3 P 1 transition of 138Ba atom, tune the 493 nm cooling laser frequency to 179.4 MHz and 222.6 MHz detuning to the cooling transition of 138 Ba + ion, and tune the 650 nm repumping laser frequency to −68.0 MHz and −174.5 MHz detuning to the repumping transition of 138 Ba +. Other processes are same with those of loading 138 Ba ions. Owing to a special properity of barium ion, namely, the less abundant barium isotopes generally have higher cooling transition frequencies than 138 Ba +. When we load and cool down the rare abundant barium isotopic ions, for example, 134 Ba +, the 493 nm cooling laser will cool the 134 Ba + and simultaneously heat the more abundant isotopes (136 Ba + and 138 Ba +). Thus we can trap the pure barium isotope easily, even though its abundance is very small compare to that of 138 Ba. For the odd isotope 137 Ba, it has a nuclear spin I = 3/2, giving several hyperfine levels. To ionize the 137 Ba atoms, we tune the 791 nm laser to resonance with one of the hyperfine transitions of 6s 2 1 S 0 → 6s6p 3 P 1. The cooling scheme of odd isotopes is different to that of the even isotopes. Using a technique similar to that of Ref. , we modulated the frequencies of cooling and repumping lasers to generate sidebands corresponding to the different hyperfine cooling and repumping transitions, respectively. To overcome the “coherent population trapping” (CPT) effect and the “dark states” between two hyperfine levels, we make the two-photon detuning of these two hyperfine cooling transitions non-zero. This effectively eliminates the CPT effect and the details will be shown elsewhere. Figure 5 shows the different barium isotope crystals which were loaded and cooled down using ionization scheme 2. From top down, they show crystals of 138 Ba +, 137 Ba +, 136 Ba +, and 134 Ba +, respectively. From Fig. 5, we estimate that the number of ions in different crystals are approximately, 4340, 780, 488, and 149, respectively. This means the loading rate ratio of 138 Ba +, 136 Ba +, and 134 Ba + are proportional to their natural abundance, yielding that their ionization cross-sections are very similiar. In Fig. 5, we did not give the ionic crystal of 135 Ba + because that it has different hyperfine splitting comparing with 137 Ba + and needs different experimental setups for cooling.
We have demonstrated a highly efficient and isotope selective photo-ionization method for barium atoms using 791 nm diode laser and 310 nm UV LED. Compared with the method using 791 nm diode laser and 337 nm N 2 laser, it has advantages of high loading efficiency, simple and cheap setups. With a mere power density of 250μW/cm 2 of 310 nm light at the trap center, ∼434 ions/s average loading rate for 138Ba+ has been achieved, and the loading rate is proportional to the power of 310 nm LED, which can be increased by using more LEDs, using high power UV LED, and optimizing the imaging system of 310 nm beam. With this method, different barium isotopes can be loaded and cooled to crystallization.
We acknowledge funding supports from the Major State Basic Research Development Program of China (973 Program) (No. 2010CB922901) and the Tsinghua University Initiative Scientific Research Program (No. 20091081474).
References and links
1. J. D. Prestage, R. L. Tjoelker, and L. Maleki, “Higher pole linear traps for atomic clock applications,” in Proceedings of the IEEE International Frequency Control Symposium, 13–16 April 1999, Besancon, France (1999), Vol. 1, pp. 121–124.
2. P. T. H. Fisk, M. J. Sellars, M. A. Lawn, and C. Coles, “Performance of a prototype microwave frequency standard based on laser-detected trapped 171Yb+ ions,” Appl. Phys. B 60, 519–527 (1995). [CrossRef]
4. V. A. Dzuba and V. V. Flambaum, “Atomic optical clocks and search for variation of the fine-structure constant,” Phys. Rev. A 61, 034502 (2000). [CrossRef]
5. W. Paul, “Electromagnetic traps for charged and neutral particles,” Rev. Mod. Phys. 62, 531–540 (1990). [CrossRef]
6. H. Häffner, C. F. Roos, and R. Blatt, “Quantum computing with trapped ions,” Phys. Rep. 469, 155–203 (2008). [CrossRef]
7. P. Herskind, A. Dantan, J. P. Marler, M. Albert, and M. Drewsen, “Realization of collective strong coupling with ion Coulomb crystals in an optical cavity,” Nat. Phys. 5, 494–498 (2009). [CrossRef]
8. N. Kjærgaard, L. Hornekaer, A. M. Thommesen, Z. Videsen, and M. Drewsen, “Isotope selective loading of an ion trap using resonance-enhanced two-photon ionization,” Appl. Phys. B: Lasers Opt. 71, 207–210 (2000). [CrossRef]
9. S. Gulde, D. Rotter, P. Barton, F. Schmidt-Kaler, R. Blatt, and W. Hogervorst, “Simple and efficient photo-ionization of ions for precision ion-trapping experiments,” Appl. Phys. B: Lasers Opt. 73, 861–863 (2001). [CrossRef]
10. D. M. Lucas, A. Ramos, J. P. Home, M. J. McDonnell, S. Nakayama, J. P. Stacey, S. C. Webster, D. N. Stacey, and A. M. Steane, “Isotope-selective photoionization for calcium ion trapping,” Phys. Rev. A 69, 012711 (2004). [CrossRef]
11. U. Tanaka, H Matsunishi, I. Morita, and S. Urabe, “Isotope-selective trapping of rare calcium ions using high-power incoherent light sources for the second step of photo-ionization,” Appl. Phys. B 81, 795–799 (2005). [CrossRef]
12. A. V. Steele, L. R. Churchill, P. F. Griffin, and M. S. Chapman, “Photoionization and photoelectric loading of barium ion traps,” Phys. Rev. A 75, 053404 (2005). [CrossRef]
13. C. Balzer, A. Braun, T. Hannemann, C. Paape, M. Ettler, W. Neuhauser, and Chr. Wunderlich, “Electrodynamically trapped Yb+ ions for quantum information processing,” Phys. Rev. A 73, 041407 (2006). [CrossRef]
14. L. Deslauriers, M. Acton, B. B. Blinov, K.-A. Brickman, P. C. Haljan, W. K. Hensinger, D. Hucul, S. Katnik, R. N. Kohn Jr., P. J. Lee, M. J. Madsen, P. Maunz, S. Olmschenk, D. L. Moehring, D. Stick, J. Sterk, M. Yeo, K. C. Younge, and C. Monroe, “Efficient photoionization loading of trapped ions with ultrafast pulses,” Phys. Rev. A 74, 063421 (2006). [CrossRef]
15. M. Brownnutt, V. Letchumanan, G. Wilpers, R. C. Thompson, P. Gill, and A. G. Sinclair, “Controlled photoionization loading of 88Sr+ for precision ion-trap experiments,” Appl. Phys. B 87, 411–415 (2007). [CrossRef]
16. U. Tanaka, I. Morita, and S. Urabe, “Selective loading and laser cooling of rare calcium isotope 43Ca+,” Appl. Phys. B 89, 195–200 (2007). [CrossRef]
17. D. N. Madsen, S. Balslev, M. Drewsen, N. Kjærgaard, Z. Videsen, and J. W. Thomsen, “Measurements on photoionization of 3s3p1P1 magnesium atoms,” J. Phys. B: At. Mol. Opt. Phys. 33, 4981–4988 (2000). [CrossRef]
18. C. Schuck, M. Almendros, F. Rohde, M. Hennrich, and J. Eschner, “Two-color photoionization of calcium using SHG and LED light,” Appl. Phys. B 100, 765–771 (2010). [CrossRef]
19. Q. A. Turchette, D. Kielpinski, B. E. King, D. Leibfried, D. M. Meekhof, C. J. Myatt, M. A. Rowe, C. A. Sackett, C. S. Wood, W. M. Itano, C. Monroe, and D. J. Wineland, “Heating of trapped ions from the quantum ground state,” Phys. Rev. A 61, 063418 (2000). [CrossRef]
20. A. Mortensen, J. J. T. Lindballe, I. S. Jensen, P. Staanum, D. Voigt, and M. Drewsen, “Isotope shifts of the 4s21S0 → 4s5p1P1 transition and hyperfine splitting of the 4s5p1P1 state in calcium,” Phys. Rev. A 69, 042502 (2004). [CrossRef]
21. G. S. Hurst, M. G. Payne, S. D. Kramer, and J. P. Young, “Resonance ionization spectroscopy and one-atom detection,” Rev. Mod. Phys. 51, 767–819 (1979). [CrossRef]
22. M. A. Kalyar, M. Rafiq, Sami-ul-Haq, and M. A. Baig, “Absolute photoionization cross section from the 6s6p1,3P1 excited states of barium,” J. Phys. B: At. Mol. Opt. Phys. 40, 2307–2319 (2007). [CrossRef]
23. D. Kulaga, J. Migdalek, and O. Bar, “Transition probabilities and lifetimes in neutral barium,” J. Phys. B: At. Mol. Opt. Phys. 34, 4775–4784 (2001). [CrossRef]
24. P. Grundevik, M. Gustavsson, G. Olsson, and T. Olsson, “Hyperfine-structure and isotope-shift measurements in the 6s5d ↔ 6p5d transitions of Ba I in the far-red spectral region,” Z. Phys. A: Hadrons Nucl. 312, 1–9 (1983).
25. P. Villemoes, A. Arnesen, F. Heijkenskjöld, and A. Wännström, “Isotope shifts and hyperfine structure of 134–138Ba II by fast ion beam-laser spectroscopy,” J. Phys. B: At. Mol. Opt. Phys. 26, 4289–4299 (1993). [CrossRef]
26. K. Wendt, S. A. Ahmad, F. Buchinger, A. C. Mueller, R. Neugart, and E. W. Otten, “Relativistic J-dependence of the isotope shift in the 6s – 6p doublet of Ba II,” Z. Phys. A 318, 125–129 (1984). [CrossRef]
27. R. G. DeVoe and C. Kurtsiefer, “Experimental study of anomalous heating and trap instabilities in a microscopic 137Ba ion trap,” Phys. Rev. A 65, 063407 (2002). [CrossRef]