- Email: [email protected]
A new method for polarization analysis in the VUV
Journal of Electron Spectroscopy and Related Phenomena 101–103 (1999) 875–878
A new method for polarization analysis in the VUV a, b b C.J. Latimer *, M.A. MacDonald , P. Finetti a
Physics Department, The Queen’ s University of Belfast, Belfast BT7 1 NN, UK b CLRC Daresbury Laboratory, Daresbury, Warrington WA4 4 AD, UK
Abstract A new method for determining the degree of circular and linear polarization of synchrotron radiation in the vacuum ultraviolet is described and has been demonstrated to be experimentally feasible. It consists of transforming short wavelength VUV radiation into the longer visible region using atomic resonance absorption in a gas and observing the polarization characteristics of the fluorescent light. 1999 Elsevier Science B.V. All rights reserved. Keywords: Synchrotron radiation; Polarization analyser; Vacuum ultraviolet; Resonance fluorescence.
PACS: 07.60; 32.00
1. Introduction As is well known, an increasing number of experiments require the use of linearly and circularly polarized synchrotron radiation (SR) for a wide variety of applications. In the VUV above 10 eV, where normal transmission polarizers are impractical, the measurement of polarization is not simple and normally involves the use of reflection polarizers and analysers [1,2]. Such measurements are difficult, time consuming and expensive. In this work we describe and demonstrate the experimental feasibility of a new method, based on a similar although as yet unexplored suggestion of Bobashev and co-workers [3,4], which avoids many of these difficulties. It consists of transforming short wavelength VUV SR radiation to the longer wavelength visible region of the spectrum using atomic resonant absorption (in a *Tel.: 144-1232-273534; fax: 144-1232-438918. E-mail address: [email protected] (C.J. Latimer)
magnetic field) and observing the polarization characteristics of the fluorescent light emitted in different directions. Although the method is most easily described in terms of a specific example (helium), it should be noted that the procedure is in fact quite general and can be used at many different photon energies within the range 15–50 eV [4].
2. Method The relevant singlet energy levels of HeI are shown in Fig. 1(a). Helium does not exhibit any hyperfine structure. Clearly, in the absence of a magnetic field, radiation of wavelength 53.7 nm (23.0 eV) can be resonantly absorbed to excite the 3 1 P state. This state decays by two routes, 98% back to the ground state and 2% to the 2 1 S state with the emission of 501.5 nm radiation [5]. The detection and analysis of this visible light is thus a monitor of
0368-2048 / 99 / $ – see front matter 1999 Elsevier Science B.V. All rights reserved. PII: S0368-2048( 98 )00407-1
876
C. J. Latimer et al. / Journal of Electron Spectroscopy and Related Phenomena 101 – 103 (1999) 875 – 878
components is simply a direct measurement of the circular polarization of the initial synchrotron radiation at 53.7 eV [7] and that the normalized Stokes parameters of the fluorescence radiation are equal to those of the VUV synchrotron radiation [4,8]. It is also worth noting the radiated intensities of Zeeman components depend only on sum rules and are independent of the type of atomic coupling [7] hence the more general applicability of this method to other atoms (and ions) to allow a wide range of wavelengths to be selected [4]. In addition, it should be noted that observations of the fluorescence perpendicular to the magnetic field in a similar manner provides information on the linear polarization of the synchrotron radiation. Fig. 1. (a) Singlet energy levels in helium, (b) the Zeeman level splittings in a magnetic field.
3. Experiment the 3 1 P population. In the presence of a magnetic field the 3 1 P splits into three, m J 511, 0, 21, levels. This Zeeman splitting is extremely small (DE(0.5 cm 21 for B|1T) and is not readily resolvable spectroscopically. However as shown in Fig. 1(b), the circularly polarized s 1 and s 2 radiation at |53.7 nm is converted with 2% efficiency into s 1 and s 2 fluorescence radiation at |501.5 nm, which can readily be analysed by observing the green fluorescence along the magnetic field (and photon beam) direction and separating the two components using a 1 / 4 wave plate and polaroid analyser [6]. It is easy to show that the relative intensities of the two visible
A schematic diagram of the apparatus is shown as Fig. 2. The experiments were carried out at the CLRC (UK) Daresbury laboratory synchrotron on a bending magnet beamline (3.3) using a toroidal grating monochromator (TGM) provided with a low energy grating of 710 lines / mm. The photon beam had a flux of approximately 3?10 11 photons / s with a band pass of 100 meV. The synchrotron radiation beam was crossed at 908 by a low pressure gas jet of helium atoms at the centre of a Helmholtz type arrangement of permanent magnets (Nd–Fe–Be) which provided a uniform magnetic field of approximately 600 mT. Fluorescence emerged from the
Fig. 2. A schematic diagram of the apparatus.
C. J. Latimer et al. / Journal of Electron Spectroscopy and Related Phenomena 101 – 103 (1999) 875 – 878
877
vacuum system through a MgF 2 window whose visible wavelength fluorescence following the absorption of VUV synchrotron radiation is negligible. The green fluorescence arising from the interaction region was selected using a confocal optical arrangement coupled with a 501.5 nm interference filter and photomultiplier.
4. Results and discussion The green fluorescence signal observed as a function of the synchrotron radiation wavelength is shown in Fig. 3. In accord with expectation it shows a clear peak at approximately 23.00 eV corresponding to the resonant absorption transition 1 S 0 – 1 P1 in helium. Fig. 4 shows the result of analysing this green resonance fluorescence with a l / 4 plate and polaroid filter as a function of filter angle, u, when the grating was separately illuminated with light from both above and below the plane of the storage ring which is expected to produce a degree of left or right polarized light respectively [9–11]. The amplitude and phase difference between the two
Fig. 3. Fluorescence at 501.1 nm observed in helium as a function of synchrotron radiation photon energy.
Fig. 4. Fluorescence at 501.1 nm observed in helium excited with 53.7 nm VUV radiation along a magnetic field and analysed with a l / 4 plate and polaroid filter as a function of filter angle. The amplitude and phase difference between measurements above and below the plane of the storage ring illustrate the differing amounts and type of circular radiation obtained. The dashed lines represent the theoretical fits to the data (see text).
measurements indicates the differing amounts and type of circular radiation obtained [6]. The observed 908 phase shift clearly shows the different sign of the circular polarisation in these two cases. The data was readily fitted to the theoretically expected forms: IB 1I0 sin 2 (u 1u0 ) and IB 1I0 cos 2 (u 1u0 ), respectively, where IB is a constant background and u0 5 22.58638 is an apparatus based angular shift. The difference in the amplitudes, I0 , (2980 cps, 650 cps) is known to be a characteristic of the beamline alignment. Additional measurements with the grating mask at intermediate positions also confirmed these conclusions. Unfortunately, beamtime restrictions did not permit us to make the relatively straightforward additional measurements required for a complete determination of the degree of linear and circular polarization in the beams. However, it should be noted that measurements of the linear polarization of the fluorescence, made perpendicular to the photon beam direction both horizontally and vertically, were in accord with the degree of horizontal linear polari-
878
C. J. Latimer et al. / Journal of Electron Spectroscopy and Related Phenomena 101 – 103 (1999) 875 – 878
zation being 0.7860.02 as determined in previous experiments on this beamline [11].
field was used in the present work to simplify the analysis and understanding, it may well be that it is not a necessary requirement [8].
5. Conclusions References Following a theoretical proposal of Bobashev et al. [3], the present work has demonstrated the experimental feasibility of a method for determining the polarization characteristics of VUV radiation, based on atomic resonance fluorescence. The degree of fluorescence polarization is determined uniquely by the degree of the VUV light polarization and by the angular momenta of the states involved in the transition [12]. Thus an analysis of the Stokes parameters for the fluorescent light allows the recovery of those of the original VUV radiation by well known and established formulae [3,12]. For simplicity it is necessary to observe the fluorescence emitted along the propagation direction of the VUV light. It has been shown that the experimental difficulties associated with such observations can be alleviated by using an atomic beam coupled with a confocal optical arrangement. Although the experiments were performed with a helium atom beam, other atoms, and even molecules, can be used at selected energies within the range 15–50 eV [4]. Finally it should be noted that, although a magnetic
[1] W. Gudat, C. Kunz, in: Synchrotron Radiation – Techniques and Applications, Springer, Berlin, 1979. [2] T. Koide et al., Nucl. Instrum. Methods A308 (1991) 635. [3] S.V. Bobashev, O.S. Vasyutinskii, Sov. Tech. Phys. Lett. 11 (1985) 599. [4] V. Bobashev, O.S. Vasyutinskii, Rev. Sci. Instrum. 63 (1992) 1509. [5] W. L. Weise, in: Atomic Transition Probabilities, NBS (1966) [6] R.S. Longhurst, Geometrical and Physical Optics, Longmans, London, 1967. [7] H.E. White, Introduction To Atomic Spectra, McGraw Hill, New York, 1934. [8] V.Yu. Backkman, S.V. Bobashev, O.S. Vasyutinskii, Solar Physics 164 (1996) 277–290. [9] J. Schwinger, Phys. Rev. 75 (1949) 1912. [10] U. Heinzmann, B. Osterheld, F. Schafers, Nucl. Instrum. Methods 195 (1982) 395. [11] J.B. West, K. Ueda, N.M. Kabachnik, K.J. Ross, H.J. Beyer, H. Kleinpopppen, Phys. Rev. A 53 (1996) R9. [12] A.C.G. Mitchell, H.W. Zemansky, Resonance Radiation and Excited Atoms, Cambridge University Press, Cambridge, 1934.
A new method for polarization analysis in the VUV a, b b C.J. Latimer *, M.A. MacDonald , P. Finetti a
Physics Department, The Queen’ s University of Belfast, Belfast BT7 1 NN, UK b CLRC Daresbury Laboratory, Daresbury, Warrington WA4 4 AD, UK
Abstract A new method for determining the degree of circular and linear polarization of synchrotron radiation in the vacuum ultraviolet is described and has been demonstrated to be experimentally feasible. It consists of transforming short wavelength VUV radiation into the longer visible region using atomic resonance absorption in a gas and observing the polarization characteristics of the fluorescent light. 1999 Elsevier Science B.V. All rights reserved. Keywords: Synchrotron radiation; Polarization analyser; Vacuum ultraviolet; Resonance fluorescence.
PACS: 07.60; 32.00
1. Introduction As is well known, an increasing number of experiments require the use of linearly and circularly polarized synchrotron radiation (SR) for a wide variety of applications. In the VUV above 10 eV, where normal transmission polarizers are impractical, the measurement of polarization is not simple and normally involves the use of reflection polarizers and analysers [1,2]. Such measurements are difficult, time consuming and expensive. In this work we describe and demonstrate the experimental feasibility of a new method, based on a similar although as yet unexplored suggestion of Bobashev and co-workers [3,4], which avoids many of these difficulties. It consists of transforming short wavelength VUV SR radiation to the longer wavelength visible region of the spectrum using atomic resonant absorption (in a *Tel.: 144-1232-273534; fax: 144-1232-438918. E-mail address: [email protected] (C.J. Latimer)
magnetic field) and observing the polarization characteristics of the fluorescent light emitted in different directions. Although the method is most easily described in terms of a specific example (helium), it should be noted that the procedure is in fact quite general and can be used at many different photon energies within the range 15–50 eV [4].
2. Method The relevant singlet energy levels of HeI are shown in Fig. 1(a). Helium does not exhibit any hyperfine structure. Clearly, in the absence of a magnetic field, radiation of wavelength 53.7 nm (23.0 eV) can be resonantly absorbed to excite the 3 1 P state. This state decays by two routes, 98% back to the ground state and 2% to the 2 1 S state with the emission of 501.5 nm radiation [5]. The detection and analysis of this visible light is thus a monitor of
0368-2048 / 99 / $ – see front matter 1999 Elsevier Science B.V. All rights reserved. PII: S0368-2048( 98 )00407-1
876
C. J. Latimer et al. / Journal of Electron Spectroscopy and Related Phenomena 101 – 103 (1999) 875 – 878
components is simply a direct measurement of the circular polarization of the initial synchrotron radiation at 53.7 eV [7] and that the normalized Stokes parameters of the fluorescence radiation are equal to those of the VUV synchrotron radiation [4,8]. It is also worth noting the radiated intensities of Zeeman components depend only on sum rules and are independent of the type of atomic coupling [7] hence the more general applicability of this method to other atoms (and ions) to allow a wide range of wavelengths to be selected [4]. In addition, it should be noted that observations of the fluorescence perpendicular to the magnetic field in a similar manner provides information on the linear polarization of the synchrotron radiation. Fig. 1. (a) Singlet energy levels in helium, (b) the Zeeman level splittings in a magnetic field.
3. Experiment the 3 1 P population. In the presence of a magnetic field the 3 1 P splits into three, m J 511, 0, 21, levels. This Zeeman splitting is extremely small (DE(0.5 cm 21 for B|1T) and is not readily resolvable spectroscopically. However as shown in Fig. 1(b), the circularly polarized s 1 and s 2 radiation at |53.7 nm is converted with 2% efficiency into s 1 and s 2 fluorescence radiation at |501.5 nm, which can readily be analysed by observing the green fluorescence along the magnetic field (and photon beam) direction and separating the two components using a 1 / 4 wave plate and polaroid analyser [6]. It is easy to show that the relative intensities of the two visible
A schematic diagram of the apparatus is shown as Fig. 2. The experiments were carried out at the CLRC (UK) Daresbury laboratory synchrotron on a bending magnet beamline (3.3) using a toroidal grating monochromator (TGM) provided with a low energy grating of 710 lines / mm. The photon beam had a flux of approximately 3?10 11 photons / s with a band pass of 100 meV. The synchrotron radiation beam was crossed at 908 by a low pressure gas jet of helium atoms at the centre of a Helmholtz type arrangement of permanent magnets (Nd–Fe–Be) which provided a uniform magnetic field of approximately 600 mT. Fluorescence emerged from the
Fig. 2. A schematic diagram of the apparatus.
C. J. Latimer et al. / Journal of Electron Spectroscopy and Related Phenomena 101 – 103 (1999) 875 – 878
877
vacuum system through a MgF 2 window whose visible wavelength fluorescence following the absorption of VUV synchrotron radiation is negligible. The green fluorescence arising from the interaction region was selected using a confocal optical arrangement coupled with a 501.5 nm interference filter and photomultiplier.
4. Results and discussion The green fluorescence signal observed as a function of the synchrotron radiation wavelength is shown in Fig. 3. In accord with expectation it shows a clear peak at approximately 23.00 eV corresponding to the resonant absorption transition 1 S 0 – 1 P1 in helium. Fig. 4 shows the result of analysing this green resonance fluorescence with a l / 4 plate and polaroid filter as a function of filter angle, u, when the grating was separately illuminated with light from both above and below the plane of the storage ring which is expected to produce a degree of left or right polarized light respectively [9–11]. The amplitude and phase difference between the two
Fig. 3. Fluorescence at 501.1 nm observed in helium as a function of synchrotron radiation photon energy.
Fig. 4. Fluorescence at 501.1 nm observed in helium excited with 53.7 nm VUV radiation along a magnetic field and analysed with a l / 4 plate and polaroid filter as a function of filter angle. The amplitude and phase difference between measurements above and below the plane of the storage ring illustrate the differing amounts and type of circular radiation obtained. The dashed lines represent the theoretical fits to the data (see text).
measurements indicates the differing amounts and type of circular radiation obtained [6]. The observed 908 phase shift clearly shows the different sign of the circular polarisation in these two cases. The data was readily fitted to the theoretically expected forms: IB 1I0 sin 2 (u 1u0 ) and IB 1I0 cos 2 (u 1u0 ), respectively, where IB is a constant background and u0 5 22.58638 is an apparatus based angular shift. The difference in the amplitudes, I0 , (2980 cps, 650 cps) is known to be a characteristic of the beamline alignment. Additional measurements with the grating mask at intermediate positions also confirmed these conclusions. Unfortunately, beamtime restrictions did not permit us to make the relatively straightforward additional measurements required for a complete determination of the degree of linear and circular polarization in the beams. However, it should be noted that measurements of the linear polarization of the fluorescence, made perpendicular to the photon beam direction both horizontally and vertically, were in accord with the degree of horizontal linear polari-
878
C. J. Latimer et al. / Journal of Electron Spectroscopy and Related Phenomena 101 – 103 (1999) 875 – 878
zation being 0.7860.02 as determined in previous experiments on this beamline [11].
field was used in the present work to simplify the analysis and understanding, it may well be that it is not a necessary requirement [8].
5. Conclusions References Following a theoretical proposal of Bobashev et al. [3], the present work has demonstrated the experimental feasibility of a method for determining the polarization characteristics of VUV radiation, based on atomic resonance fluorescence. The degree of fluorescence polarization is determined uniquely by the degree of the VUV light polarization and by the angular momenta of the states involved in the transition [12]. Thus an analysis of the Stokes parameters for the fluorescent light allows the recovery of those of the original VUV radiation by well known and established formulae [3,12]. For simplicity it is necessary to observe the fluorescence emitted along the propagation direction of the VUV light. It has been shown that the experimental difficulties associated with such observations can be alleviated by using an atomic beam coupled with a confocal optical arrangement. Although the experiments were performed with a helium atom beam, other atoms, and even molecules, can be used at selected energies within the range 15–50 eV [4]. Finally it should be noted that, although a magnetic
[1] W. Gudat, C. Kunz, in: Synchrotron Radiation – Techniques and Applications, Springer, Berlin, 1979. [2] T. Koide et al., Nucl. Instrum. Methods A308 (1991) 635. [3] S.V. Bobashev, O.S. Vasyutinskii, Sov. Tech. Phys. Lett. 11 (1985) 599. [4] V. Bobashev, O.S. Vasyutinskii, Rev. Sci. Instrum. 63 (1992) 1509. [5] W. L. Weise, in: Atomic Transition Probabilities, NBS (1966) [6] R.S. Longhurst, Geometrical and Physical Optics, Longmans, London, 1967. [7] H.E. White, Introduction To Atomic Spectra, McGraw Hill, New York, 1934. [8] V.Yu. Backkman, S.V. Bobashev, O.S. Vasyutinskii, Solar Physics 164 (1996) 277–290. [9] J. Schwinger, Phys. Rev. 75 (1949) 1912. [10] U. Heinzmann, B. Osterheld, F. Schafers, Nucl. Instrum. Methods 195 (1982) 395. [11] J.B. West, K. Ueda, N.M. Kabachnik, K.J. Ross, H.J. Beyer, H. Kleinpopppen, Phys. Rev. A 53 (1996) R9. [12] A.C.G. Mitchell, H.W. Zemansky, Resonance Radiation and Excited Atoms, Cambridge University Press, Cambridge, 1934.