We report on the first Doppler-free spectroscopy investigation of an atomic species, xenon, performed in the mid-infrared using difference-frequency radiation. The absorption saturated spectrum of the xenon 6p[3/2]2→5d[5/2]3 transition (2p6→3d’1 in Paschen notation) at 3.1076 μm was investigated using about 60 microwatts of cw narrowband radiation (Δv=50 kHz) generated by difference-frequency mixing in a periodically-poled Lithium Niobate crystal. A single frequency Ti: Sapphire laser (power 800 mW) and a monolithic diode-pumped Nd:YAG laser (300 mW) were used as pump and signal waves respectively. We used natural enriched xenon, which contains nine stable isotopes, two of which, 129Xe and 131Xe, exhibit a hyperfine structure owing to their nuclear spin. The small isotope displacements expected for this atom and the complex hyperfine structure of the odd isotopes make it difficult to fully resolve the recorded saturated-absorption spectra. In spite of this, we have been able to analyze the isolated 129Xe F’’=5/2→F’=7/2 hyperfine component by means of first-derivative FM spectroscopy.
©2005 Optical Society of America
The mid-infrared (mid-IR) spectral region is particularly interesting for several spectroscopic features. For instance, in many applications related to trace-gas analysis, the mid-IR is very attractive since strong absorption bands connected to fundamental vibrations of most molecular species lie just in this spectral region. Moreover, high-resolution laser spectroscopy in the mid-IR is interesting to get detailed information about atomic and molecular structures and to provide new frequency standards, since a limited number of absolute frequency measurements have been made in this region. For many decades the only laser sources available in the mid-IR were represented by lead-salt diode lasers. The drawbacks of these laser sources (cryogenic temperatures, low spectral purity, etc.) have stimulated the recent development of new tunable coherent sources, such as Optical Parametric Oscillators (OPO) , Difference-Frequency Generators (DFG) , and Quantum Cascade Diode Lasers (QCDL) . All these sources have been extensively used in many high sensitivity detection studies of molecular species [4–5] and in few Doppler-free investigations of molecular lines [6–8]. In this work we report, according to our knowledge, the first Doppler-free spectroscopy of an atomic species (xenon atom) by using low-power difference-frequency radiation. Similar experiments, applied to molecular lines, were recently performed by de Natale’s group which investigated a CO2 line at 4.5 μm using a single-pass optical scheme  or enhanced cavity configuration . The spectra of noble gases are among the most widely observed spectra of all chemical elements because they occur in a wide variety of phenomena, such as plasma physics, astrophysics, and metrology. It is well known that the transitions of these elements from the ground state 1S0 are difficult to investigate, since they lie in the vacuum ultraviolet range where tunable and narrow band laser sources are not easily available. In contrast, visible and near-infrared transitions of these elements connect the metastable 6s[3/2]2 and 6s’[1/2]0 states and the radiative 6s’[1/2]1 and 6s[3/2]1 states with energetically higher levels. These transitions give rise to the ns→np series while the np→nd series usually lie in the mid-infrared spectral region. Among the noble gases the xenon atom (Z=54) is of particular interest for Doppler-free analysis: it has nine stable isotopes in natural abundances accessible to high sensitivity laser spectroscopy techniques (124Xe: 0.0096%, 126Xe: 0.0090%, 128Xe: 1.92%, 129Xe: 26.4%, 130Xe: 4.1%, 131Xe: 21.1%, 132Xe: 26.9%, 134Xe: 10.4%, and 136Xe: 8.9%), and, hence, is suitable for isotope shifts analysis. In addition, the nuclei of two of these isotopes have a spin (129Xe:I=1/2 and 131Xe:I=3/2) which gives rise to a hyperfine splitting of the levels. Finally, the series of Xe isotopes crosses the closed neutron shell N=82 (136Xe is a magic nucleus), and this can result in interesting anomalies in the isotope shifts. Moreover, since for atoms with atomic number Z between 30 and 60 the mass shift and field shift contribution are of the same order of magnitude and can also exhibit different sign , expected isotope shifts for xenon (Z=54) are quite small, and. therefore, quite hard to investigate. The isotope shift (IS) and hyperfine structure (HFS) in xenon have been measured in many transitions by using Doppler-limited spectroscopic techniques [12–14]. More recently, high resolution studies have been facilitated by well consolidated Doppler-free laser spectroscopy. The IS and HFS constants of xenon isotopes far from stability were investigated for the 6s[3/2]2→6p[3/2]2 transition (λ=823.3 nm) by collisional ionization laser spectroscopy . Geisen et al.  investigated the transitions 6s’[1/2]0→6p[1/2]1 (λ=626.5 nm) and 6s[3/2]2→5d[5/2]3 (650.7 nm) by means of the laser-induced fluorescence technique in atomic beams. Doppler-free two-photon laser spectroscopy was used by Plimmer et al.  in order to investigate the IS in a two-photon transition at 249 nm from the 1S0 ground state to the 6p[1/2]0 level. Two-photon transition between the two metastable states 6s[3/2]2 and 6s’[1/2]0 involving photons a 2.19 μm, has been proposed as a future optical frequency standard. Sterr et al.  determined the wavenumber of this clock transition (2 Hz natural linewidth) by interferometrically comparing the wavelengths of the 6s[3/2]2→6p’[1/2]1 (λ=450 nm) and 6s’[1/2]0→6p’[1/2]1 (λ=764 nm) transitions with and iodine-stabilized 633 nm He-Ne laser. Walhout has recently  performed precise IS measurements on cold atoms of the 6s[3/2]2→6p[5/2]3 transition (λ= 882.2 nm). In our previous work  we measured the IS and hyperfine structure coupling constants of four 1si→2pj transitions (Paschen notation) in the near-IR using a semiconductor diode laser mounted in an extended cavity configuration and tunable between 820 and 830 nm. The transition investigated in the present work is the 6p[3/2]2→5d[5/2]3 (2p6→3d’1 in Paschen notation) which is the strongest atomic line in the mid-IR region accessible to the tunability of our apparatus. As far as we known, this is the first sub-Doppler investigation of an atomic transition in the mid-IR.
2. Experimental set-up
The experiment was realized with the apparatus schematically shown in Fig. 1. Mid-IR radiation was generated through difference-frequency mixing in a periodically poled lithium niobate (PPLN) optical crystal. A Ti: Sapphire laser (Coherent, Mod. MBR110) emitting about 1.6 W of narrowband (50kHz) tunable (0.7-1.0 μm) radiation (single scan up to 40 GHz) was used as pump beam. A diode-pumped Nd- YAG laser (Innolight, Model Mephisto 500) consisting of a monolithic ring cavity which emits a maximum power of 500 mW (Δv=10 kHz) was used as signal beam. The two laser beams were combined by a dichroic mirror and focused into a 19-mm-long PPLN (Crystal Technology). Within the Ti:Sapphire tunability it was possible to produce 60 μW of mid-IR radiation (idler) from 2.8 to 3.2 μm, with pump and signal powers of 800 mW and 300 mW respectively. The mid-IR radiation wavelength was determined by measuring the Ti: Sapphire wavelength with a traveling Michelson interferometer (accuracy of one part in 107) while the Nd-YAG wavelength was obtained from calibration curves provided by the manufacturer. Finally, DFG frequency scans were accomplished by tuning the Ti-Sa laser frequency and calibrated by using a 300 MHz free-spectral range (FSR) confocal Fabry-Perot interferometer. Excited xenon atoms were produced by means of a radio-frequency discharge (power 50 Watt, frequency 60 MHz) whose details have been described elsewhere . With that discharge, typical excited atom densities in the lower level 2p6 were of the order of 107 atoms/cm3. Sub-Doppler saturation spectra were observed by retroreflecting the laser beam back through the discharge cell ; in this simple geometry, the incoming beam acts as pump and the reflected beam as probe. The mid-IR beam was focused by a 20 cm focal-length lens to a waist of about 800 μm in the region of maximum intensity of the plasma discharge near one end of a 20-cm long Pyrex discharge tube (internal diameter= 7 mm). The focal plane of the focusing lens was almost coincident with the plane of the reflecting mirror. Finally the reflected beam, separated from the incoming beam by a 10% beam-splitter, was collected onto the active area of a liquid-nitrogen InSb detector (Hamamatsu, Model PN5968). Since the available DFG power was too low to observe a saturation dip in direct absorption, we used first-derivative Frequency Modulation (FM) saturation spectroscopy. In this case, phase sensitive detection was performed by modulating the Nd:YAG laser frequency by few MHz by means of a piezoelectric transducer applied to its cavity length. This modulation occurred at a frequency of 10 kHz which was the same frequency as the one used as reference signal for a lock-in amplifier.
3. Results and discussion
Typical Doppler-limited and Doppler-free absorption spectra of the 6p[3/2]2→5d[5/2]3 transition are shown in Fig. 2.
The Doppler-limited spectrum (Fig. 2(b)) represents the envelope of the numerous even isotopes and of the complex hyperfine structure concerning the odd isotopes. The first-derivative spectrum (Fig. 2(a)) was obtained with a modulation depth of 8.3 MHz, and a lock-in time constant of 10 ms. The xenon discharge pressure of 0.08 mTorr was properly chosen to optimize the contrast of the saturation dip, represented by the sharp resonances on the broader Doppler profiles (see arrows in Fig. 2(a)). The expected hyperfine patterns of the two odd isotopes, 129Xe and 131Xe, are shown in Fig. 3 where are depicted even the most significant xenon levels. According to our knowledge, no information exists about the hyperfine structure of the transition 6p[3/2]2→5d[5/2]3. In our previous study  of the 6s[3/2]2→6p[3/2]2 (1s5→2p6 in Paschen notation) transition at λ=0.8233 μm (which shares the 2p6 level with the mid-IR transition investigated in this work), the hyperfine splitting of the 2p6 level was estimated for both odd isotopes. These values are reported in Fig. 3. For 129Xe (I=1/2), the hyperfine structure is only due to the magnetic dipole interaction (as the electric quadrupole constant B=0 for nuclei with I<1) and the hyperfine multiplets result inverted. So far, if we look at the hyperfine structure of the 2p6 and 3d’1 levels of 129Xe, the whole hyperfine pattern is expected in a frequency range corresponding to |a-b|, where a and b represent the hyperfine splitting of the 2p6 and 3d’1 levels respectively. Since a=2.25 GHz and considering that the whole recorded Doppler-free spectrum has a range of about 1.5 GHz, we can conclude that the hyperfine splitting b of the upper level 3d’1 is comparable to that of the lower level 2p6. That explains why the hyperfine structure patterns of the odd isotopes (three lines for 129Xe and nine for 131Xe according to Fig. 3) are only partially resolved. In addition, since the IS of Xe is expected to be of the order of few tens of MHz (see ref. ), the even isotope peaks collapse into a single unresolved peak (see the largest peak of Fig. 2(a)). The natural linewidth of the transition 6p[3/2]2→5d[5/2]3 was estimated by focusing our attention on the isolated peak occurring in the lower frequency region in the spectrum of Fig. 2.
On the basis of similarity of this spectrum with that obtained for near-IR Xe lines (ref. ), we suggest that this peak corresponds to the most intense F’’=5/2→F’=7/2 hyperfine component of 129Xe. Figure 4 shows the experimental spectrum of this peak (modulation amplitude =3 MHz): the wide and the sharper resonances correspond to the first-derivative of the Doppler and homogeneous line-profile respectively. The experimental line-shape was fitted with the derivative of the sum of a Gaussian and a Lorentzian curve as shown in Fig. 4. The agreement is rather good, although some little deviations could reflect the presence of some weak line blended in the profile. The best-fit parameters give a Doppler width of 128±3 MHz FWHM (which corresponds to a discharge temperature of 450 K) and a homogeneous width of 30±1 MHz FWHM. Therefore, the agreement between the experimental and fitted parameters confirms our hypothesis that the analyzed peak is effectively a single hyperfine component. Its width of 30 MHz can be considered to be the natural width of the investigated transition since, pressure broadening and power broadening can be neglected in our experimental conditions of low pressure and low radiation power. Moreover, even time-of-flight broadening, which in our case is of the order of 1.8 MHz, can be assumed ineffective on the observed homogeneous width. Knowing the lifetime of the lower level (τ(2p6)=40 ns) , we can give an estimation of the radiative lifetime of the upper level 3d’1 of 6.0±0.2 ns.
In order to estimate the saturation intensity Isat of the investigated transition, we have measured the contrast H referred to the ratio between the sub-Doppler signal and the Doppler profile by varying the laser intensity I. This analysis was performed taken into account the different modulation index for the two line-profiles, Gaussian and Lorentzian, when a given modulation amplitude was fixed. The linear behavior of H versus the laser intensity (see Fig. 5) means that the intensities used in our experiment are quite smaller than Isat. Therefore the contrast, which is given by , becomes , and from a linear fit of the data shown in Fig. 5 results in Isat = 9.2(5) mW/cm2, which is typical for atomic transitions. Moreover, from the expression of , where is the sub-Doppler HWHM, the dipole moment μ of the investigated transition can be inferred: μ=3.7(1)∙10-29 C∙m.
In conclusion, we have demonstrated the possibility to perform high resolution spectroscopy of atomic species using difference frequency radiation at low power. This opens interesting possibilities for absolute frequency measurements to provide new secondary frequency standards in the mid-IR. Indeed, the frequencies of the two lasers used for frequency mixing can be precisely measured: the Hall’s group  has provided very precise absolute frequency measurements of a frequency-doubled Nd:YAG laser locked to iodine lines, while, nowadays, frequency comb technique  allows direct measurements of frequency within the Ti:Sapphire laser emission spectrum. Of course, using enriched xenon samples, simplified spectra are available (single odd isotope line or isolated HFS component) which make easier such kind of measurements. Moreover, the use of enriched samples can make feasible even the HFS and IS studies of the highly xenon excited levels investigated in this work.
References and links
1 . M. M. J. Van Herpen , S. Li , S. E. Bisson , S. Te Lintel Hekkert , and F. J. M. Harren , “ Tuning and stability of a continuous-wave mid-infrared high-power single resonant optical parametric oscillator ,” Appl. Phys. B 75 , 329 – 333 ( 2002 ) [CrossRef]
2 . D. Richter , A. Fried , B. P. Wert , J. G. Walega , and F. K. Tittel , “ Development of a tunable mid-IR difference frequency laser source for highly sensitive airborne trace gas detection ,” Appl. Phys. B 75 , 281 – 288 ( 2002 ). [CrossRef]
3 . K. Namjou , S. Cai , E. A. Whittaker , J. Faist , C. Gmachl , F. Capasso , D. L. Sivco , and A. Y. Cho , “ Sensitive absorption spectroscopy with a room-temperature distributed-feedback quantum-cascade laser ,” Opt. Lett. 23 , 219 – 221 ( 1998 ) [CrossRef]
4 . D. G. Lancaster , D. Richter , R. F. Curl , and F. K. Tittel , “ Real-time measurements of trace gases using a compact difference-frequency-based sensor operating at 3.5 μm ,” Appl. Phys. B 67 , 339 – 345 ( 1998 ) [CrossRef]
5 . S. Stry , P. Hering , and M. Murtz , “ Portable difference-frequency laser-based cavity leak-out spectrometer for trace-gas analysis ,” Appl. Phys. B 75 , 297 – 303 ( 2002 ) [CrossRef]
6 . A. Hecker , M. Havenith , C. Braxmaier , U. Stroner , and A. Peters , “ High resolution Doppler-free spectroscopy of molecular iodine using a continuous wave optical parametric oscillator ,” Opt. Commun. 218 , 131 – 134 ( 2003 ) [CrossRef]
7 . D. Mazzotti , P. De Natale , G. Gagliardi , C. Fort , J. A. Mitchell , and L. Hollberg , “ Saturated-absorption spectroscopy with low-power difference-frequency radiation ” Opt. Lett. 25 , 350 – 352 ( 2000 ) [CrossRef]
8 . J. T. Remillard , D. Uy , W. H. Weber , F. Capasso , C. Gmachl , A. L. Hutchinson , D. L. Sivco , J. N. Baillargeon , and A. Y. Cho , “ Sub-Doppler resolution limited Lamb-dip spectroscopy of NO with a quantum cascade distributed feedback laser ” Opt. Express 7 , 243 – 248 ( 2000 ) http://www.opticsexpress.org/abstract.cfm?URI=OPEX-7-7-243 [CrossRef] [PubMed]
9 . P. Maddaloni , G. Gagliardi , P. Malara , and P. De Natale , “ A 3.5-mW continuous-wave difference-frequency source around 3 mu m for sub-Doppler molecular spectroscopy ” Appl. Phys. B 20 , 141 – 145 ( 2005 ). [CrossRef]
10 . S. Borri , P. Cancio , P. De Natale , G. Giusfredi , D. Mazzotti , and F. Tamassia “ Power-boosted difference-frequency source for high-resolution infrared spectroscopy ” Appl. Phys. B 76 , 473 ( 2003 ) [CrossRef]
11 . J. Bauche and R. J. Champeau , “ Recent progress in the theory of atomic isotope shifts ” Adv. At. Mol. Phys. 37 , 39 – 86 ( 1976 ). [CrossRef]
12 . D.A. Jackson , F.R. S. , and M.-C. Coulombe , “ Isotope shifts in the arc spectrum of xenon ,” Proc. R. Soc. Lond. A 338 , 277 – 281 ( 1974 ) [CrossRef]
13 . D.A. Jackson , F.R. S. , M.-C. Coulombe , and J. Bauche , “ Isotope shifts in the arc spectrum of xenon II ” Proc. R. Soc. Lond. A 343 , 453 – 459 ( 1975 ). [CrossRef]
14 . W. Fischer , H. Huhnermann , G. Kromer , and H. J. Schafer , “ Isotope shift in the Atomic Spectrum of Xenon and Nuclear Deformation Effects ,” Z. Phys. 270 , 113 – 120 ( 1974 ). [CrossRef]
15 . W. Borchers , E. Arnold , W. Neu , R. Neugart , K. Wendt , and G. Ulm , “ Xenon Isotopes far from Stability studied by Collisional Ionization Laser Spectroscopy ,” Phys. Lett. B 216 , 7 – 10 ( 1989 ). [CrossRef]
16 . H. Geisen , T. Krumpelmann , D. Neuschafer , and Ch. Ottingen , “ Hyperfine splitting measurements on the 6265 Å and 6507 Å lines of seven Xe isotopes by lif on a beam of metastable Xe( 3 P 0,2 ) atoms ,” Phys. Lett. A 130 , 299 – 304 ( 1988 ). [CrossRef]
17 . M. D. Plimmer , P.E.G. Baird , C.J. Foot , D.N. Stacey , J.B. Swan , and G.K. Woodgate , “ Isotope shift in xenon by Doppler-free two-photons laser spectroscopy ” J. Phys. B: At. Mol. Opt. Phys. 22 , L241 – L244 ( 1989 ). [CrossRef]
18 . U. Sterr , A. Bard , C. J. Sansonetti , S. L. Rolston , and J. D. Gilapsy , “ Determination of the xenon 6s[3/2](2)-6s’[1/2](0) clock frequency by interferometric wavelength measurements ,” Opt. Lett. 20 , 1421 – 1423 ( 1995 ). [CrossRef] [PubMed]
19 . M. Walhout , H. J. L. Megens , A. Witte , and S. L. Rolston , “ Magnetooptical trapping of metastable xenon -isotope-shift measurements ,” Phys. Rev. A 48 , R879 – R882 ( 1993 ). [CrossRef] [PubMed]
20 . G. D’Amico , G. Pesce , and A. Sasso , “ Isotope-shift and hyperfine-constant measurements of near-infrared xenon transitions in glow discharges and on a metastable Xe(P-3(2)) beam ,” Phys. Rev. A 60 , 4409 – 4416 ( 1999 ). [CrossRef]
21 . G. M. Tino , L. Hollberg , A. Sasso , M. Inguscio , and M. Barsanti , “ Hyperfine-structure of the metastable S-2(5) state of O-17 using an AlGaAs diode-laser at 777 nm ,” Phys. Rev. Lett. 64 , 2999 – 3002 ( 1990 ). [CrossRef] [PubMed]
22 . A. A. Radzig and B. M. Smirnov in “Reference Data on Atoms, Molecules, and Ions,” Springer Series in Chemical Physics - Berlin ( 1980 ).
23 . V. S. Letokhov and V. P. Chebotayev in Nonlinear Laser Spectroscopy , Springer Verlag Series in Optical Sciences Berlin ( 1977 ).
24 . J. Hall , L. Ma , M. Taubmann , B. Tiemann , F. Hong , O. Pfister , and J. Ye , “ Stabilization and frequency measurement of the I-2-stabilized Nd : YAG laser ,” IEEE Trans. Instrum. Meas. 48 , 583 – 586 ( 1999 ). [CrossRef]
25 . R. Holzwarth , T. Udem , T. Hansch , W. Knight , W. J. Wadsworth , and P. St. J. Russel , ” Optical frequency synthesizer for precision spectroscopy ,” Phys. Rev. Lett. 85 , 2264 – 2267 ( 2000 ). [CrossRef] [PubMed]