`
`
`APPLIED
`
`PHYSICS .
`
`
`
`
`
`NUMBER 1
`
`
`3
`
`
`
`
`995051"!
`
`
`VOLUMEII
`
`
`
`main.
`
`ti:
`wlu
`
`Md
`TIP
`“1)
`SKID" \.
`
`'uhl‘
`\ Tum-
`m"
`mull
`
`
`
`
`
`
`
`TCL 1025, Page 1
`’FC141025,Page1
`
`
`
`Volume 11, Number 2
`
`APPLIED PHYSICS LETTERS
`
`15 July 1967
`
`mately 2. Experimentally we have found that R
`has the value of unity or slightly greater, within a
`maximum experimental error of ±O.2. There is no
`significant variation in this value between focused
`and unfocused input beams, so spatial non-overlap
`cannot be a factor. Pulse envelope traces (Tektronix
`519 oscilloscope) of the laser output show regular
`modulation indicating some locking of the multi(cid:173)
`modes. While this would increase the efficiency of
`S.H.G.p it would require for the mixing process
`a more critical synchronization of the resulting
`modulation of the two fundamental frequencies.
`Streak interferograms (2 nsec resolution) show that
`such complete synchronization does not exist.
`After this work had been completed, the mixing
`of ruby and neodymium Q-switched lasers operat(cid:173)
`ing in a common resonator to produce a summed
`output at 4192 A was reported13 but the relative
`efficiencies of S.H.G. and sum generation were not
`compared.
`
`I M. Bass, P. A. Franken, A. E. Hill, C. W. Peters, and G. Wein-
`reich,Phys. Rev. Letters 8,18 (1962).
`2 R. C. Miller and A. Savage, Phys. Rev. 128, 2175 (1962).
`aN. 1. Adams and P. B. Schoefer, Proc. IEEE 51, 1366 (1963).
`4 D. H. McMahon and A. R. Franklin, I Appl. Phys. 36, 2073
`(1965).
`'D. H. McMahon and A. R. Franklin,I Appl. Phys. 36, 2807
`(1965).
`6 A. W. Smith and N. Braslau, IBM I of Res. and Dev. 6, 361
`(1962).
`7 M. D. Martin, E. L. Thomas, and J. K. Wright, Phys. Letters
`15,136 (1965).
`8R. Madhavan, M. K. Dheer, and J. S. Jaseja, Appl. Opt. 5,
`1823 (1966).
`9D. J. Bradley, G. Magyar, and M. C. Richardson, Proc. VII
`Int. Conf. on Ion. Phen. in Gases, Belgrade (1965), in press.
`IOD. J. Bradley, G. Magyar, and M. C. Richardson, to be pub(cid:173)
`lished.
`II N. Bloembergen, Nonlinear Optics 133, (Benjamin, New York,
`1965), p. 133.
`12R. L. Kohn and R. H. Pantell, Appl. Phys. Letters 8, 231 (1966).
`l3yU. A. Gol'din, V. G. Dmitriev, V. K. Tarasov, and N. V.
`Shkunov,jETP Letters 4, 297 (1966).
`
`A NEW PHOSPHOR FOR FLYING-SPOT CATHODE-RAY TUBES FOR
`COLOR TELEVISION: YELLOW-EMITTING Y3Als012-Ce3+
`
`The fluorescence of YaAlsO I2-Ce3+ under cathode-ray excitation consists of an emission band peaking at 550 nm.
`The decay time is 0.07 -0.08 fJ.sec. In view of these properties this phosphor is very suitable for flying-spot cathode(cid:173)
`ray tubes for color television.
`
`C. Blasse and A. Brit
`Philips Research Laboratories
`N. V. Philips' GloeiIampenfabrieken
`Eindhoven, The Netherlands
`(Received 29 May 1967)
`
`Ce3+ -activated phosphors are usually char(cid:173)
`acterized by an emission band with a maximum in
`or near the uv region and a very short decay time.1
`In this Letter we report on a Ce3+ -activated phos(cid:173)
`phor with an emission at considerably longer Wave(cid:173)
`lengths, viz. Y3AIs012-Ce.
`Samples with the general formula Y 3-xCexAls012
`(x < 0.3) were prepared by firing intimate mix(cid:173)
`tures of high-purity Y 203, hydrated alumina and
`Ce20 3 at 1500° in nitrogen. X-ray analysis showed
`them to be pure garnets. According to chemical
`analysis the CeH content of the samples is neg(cid:173)
`ligible. This was also the case if the samples were
`fired in oxygen, air, or nitrogen containing 5%
`of hydrogen. Optical measurements were per(cid:173)
`formed as described previously. 2
`Y 3AIsOlcCe shows a bright yellow emission
`
`under excitation with cathode rays as well as with
`blue radiation. Figure 1 shows the spectral energy
`
`Fig. 1. Spectral energy distribution of Y '.94Ceo.08~0I2
`under cathode ray excitation (full line), uv excitation (broken
`line) and of Y •. 94Ceo.08AI.Gaa0I2 under uv excitation (dotted
`line). The solid and broken lines coincide in the visible region.
`
`53
`TCL 1025, Page 2
`
` Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 198.65.204.101 On: Mon,
`14 Nov 2016 15:02:16
`
`
`
`Volume 11, Number 2
`
`APPLIED PHYSICS LETTERS
`
`15 July 1967
`
`distribution of the emission of Y3AIsO I2-Ce under
`cathode-ray (cr) as well as under ultraviolet (uv)
`excitation. Figure 2 shows the excitation spectrum
`of the visible fluorescence of Y3AIsOI2-Ce and the
`diffuse reflection spectra of Y3AIsOI2-Ce and un(cid:173)
`activated Y 3AIsOI2' The radiant efficiency for cr
`excitation (20 keY) is 3.5% for the emission in the
`visible region and 0.1 % for the emission in the uv
`region.3 These figures were found for a sample
`with composition Y2.94Ceo.06AIsOI2 and vary only
`slightly with the Ce concentration. At very low Ce
`concentration the ratio of the intensities of uv to
`visible emission increases.
`Decay times were measured with cr and uv exci(cid:173)
`tation. With cr excitation the phosphor was irradi(cid:173)
`ated with short electron beam pulses (duration 0.1
`JLsec, repetition rate 104 cps} in a demountable tube.
`The fluorescence was collected on a photomultiplier
`and displayed on an oscilloscope. The decay time
`was found to be shorter than 0.1 JLsec. The phos(cid:173)
`phor was also excited by a short pulse of uv radia(cid:173)
`tion with the aid of a TRW nanosecond spectral
`source system (duration 0.004 JLsec, repetition rate
`5 X 103 cps). The output was again displayed on an
`oscilloscope. The decay time proved to be 0.07-
`0.08 JLsec.
`All Ce3+ -activated phosphors described in the
`literature show the maximum of their emission in
`the uv or blue (up to =410 nm). The emission from
`Y3AIsOI2-Ce under cr excitation, however, lies
`almost entirely in the visible with a broad band
`peaking at 550 nm. There is also a uv emission
`band under short-wave uv excitation. In view of
`this peculiar behavior it may be wondered whether
`this emission is really due to the Ce3+ center. There
`are several reasons for believing that this is the
`case:
`(a) Chemical analysis excludes other oxidation
`states of cerium; (b) unactivated Y3AIsOl2 does not
`
`fOO~----------------~~------~(W
`
`-------------------------------'-
`
`I
`Iso
`j
`t
`
`Fig. 2. Relative excitation spectrum of the visible ftuores-
`cence of Y .... Ceo.oaAIsOI2 (solid line) and dift1ue reftection
`spectra of Y2 ... Ceo.oaAIsO'2 (broken line) and Y,A1s0 12 (dot-
`ted line).
`
`54
`
`emit in this region; (c) the decay time of Ce3+ phos(cid:173)
`phors is always very short, viz. shorter than 0.1
`JLsec, since an allowed transition is involved en ~
`2F). Y3AIsOI2-Ce has such a short decay time; (d)
`the emission band of Ce3+ phosphors should con(cid:173)
`sist of two bands, the ground state being split
`eF7I2 , 2Fsd. The energy difference between these
`two bands is some 2000 cm-I. Particularly at higher
`Ce concentrations this splitting is not always ob(cid:173)
`served. The uv emission band ofY3AlsOl2-Ce does
`two peaks (29.800 and 27.800
`indeed exhibit
`cm-I) but the visible band does not. Y3A12Ga3012-Ce
`with similar but less efficient fluorescence, however,
`shows an indication of the splitting of this emission
`band in the visible.
`Figure 2 shows that Y3AIsOI2-Ce is most effi(cid:173)
`ciently excited by 460 nm radiation. The efficiency
`for long-wave uv excitation is less, and that for
`short-wave uv excitation may even be termed low.
`The excitation bands correspond to absorption in
`the Ce3+ center. This follows from a comparison
`between the diffuse reflection spectra of Y3AIsOl2
`and Y3AIsOI2-Ce. These bands show the complete
`crystal field splitting of the 2n level of Ce3+.
`Elsewhere we will discuss this and other Ce3+
`phosphors more extensively. For the moment we
`conclude that Y3AIsOI2-Ce3+
`is an exceptional
`Ce3+ phosphor with an emission at relatively long
`wavelengths and a high efficiency under cr exci(cid:173)
`tation.
`Therefore this phosphor can be of use for flying(cid:173)
`spot cathode-ray tubes, especially for color tele(cid:173)
`vision. When the screen of such a tube is scanned
`with a normal television raster, a decay time of
`the phosphor shorter than =0.1 JLsec is required.
`Usually ZnO-Zn is used with an emission band at
`500 nm containing only a small amount of red and
`with an efficiency which is about the same as that
`of Ce3+ -activated phosphors. Its decay time, how(cid:173)
`ever, is too long, viz. = 1 JLsec. Therefore a large
`correction is applied to the output signal by elec(cid:173)
`trical compensation with unavoidably a great re(cid:173)
`duction in this signal. YaAIsO I2-Ce has a much
`higher output in the red region and, moreover, a
`much shorter decay time. Therefore the output
`signal of this tube is expected to be an order of
`magnitude higher than for ZnO- Zn. The emission
`of YaAIsOI2-Ce does not contain blue. However,
`efficient blue phosphors with short decay time are
`available (for example, Ca2AI2Si07-Ce (ref. 1), so
`that a mixture of these two phosphors will suffice.
`
`'See, e.g., A. Bril and H. A. Klasens, Philips Res. Repts. 7,421
`(1952).
`
`TCL 1025, Page 3
`
` Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 198.65.204.101 On: Mon,
`14 Nov 2016 15:02:16
`
`
`
`Volume 11, Number 2
`
`APPLIED PHYSICS LETTERS
`
`15 July 1967
`
`2A. Bril and W. L. Wanmaker,J. Electrochem. Soc. 111,1363
`(1964).
`
`3For comparison: the most efficient Ce3+-activated phosphor
`C~AI2Si07 has an efficiency of 4% (ref. 1).
`
`HIGH TEMPERATURE RESONANCE LOSSES IN SILICON-DOPED
`YTTRIUM-IRON GARNET (YIG)*
`
`D. J. Epsteint
`Electric Power Engineering Department
`Technical University of Denmark
`Lyngby, Denmark
`L. Tocci
`Department of Electrical Engineering and
`Center for Materials Science and Engineering
`Massachusetts Institute of Technology
`Cambridge, Massachusetts
`(Received 27 March 1967; in final form 31 May 1967)
`
`The ferrimagnetic resonance linewidth of silicon-doped YIG, measured as a function of temperature at 13.4
`kMHz, is found to show a pronounced peak at 105°C. The anisotropic behavior of this peak is in good agreement
`with the four-level valence-exchange model proposed by Clogston. The model yields for the electron ordering
`energy a value 5 X 10-4 eV which agrees closely with the energy deduced from magnetic anneal studies. The acti(cid:173)
`vation energy for electron transfer (0.25 eV) is virtually identical with values reported in investigations of electrical
`conductivity and acoustic loss.
`
`It is known that the introduction of small con(cid:173)
`centrations of silicon into the YIG lattice leads to a
`significant increase in magnetic loss. There is gen(cid:173)
`eral agreement, however, that silicon itself plays
`only an incidental role and that the losses are actu(cid:173)
`ally due to the presence of divalent iron created in
`response to the addition of quadrivalent silicon.
`In YIG samples containing Fe2+ two peaks have
`been reported in the plot of microwave resonance
`Iinewidth vs temperature. Typically, one peak oc(cid:173)
`curs at low temperature below ca. 100oK, and the
`other above ca. 300°K. It was suggested, at the time
`of its discovery, that the low temperature peak was
`due to valence exchange!,2 i.e., electron transfer
`between Fe2+ and Fe3+ cations. More recently,
`Tchernev3 has. argued that the electron-hopping
`mechanism is, in all likelihood, "frozen out" at
`tern peratures below 100oK, and therefore, it is more
`reasonable to ascribe the peak to the localized Fe2+
`ion acting, more or less,4 as a slow relaxer. The
`high temperature peak has been looked at to a
`
`*The experimental work was performed under contract with
`the Air Force Materials Laboratory, Research and Technology
`Division, Air Force Systems Command, Wright-Patterson Air
`Force Base, Ohio.
`t Presently on leave of absence from the Department of Elec(cid:173)
`trical Engineering, Massachusetts Institute of Technology, Cam(cid:173)
`bridge, Massachusetts.
`
`lesser extentS,7 and the suggestion has been made
`that it is due to valence exchange. 7
`In our work the temperature dependence and
`anisotropy of the upper peak have been examined
`in some detail. Our results provide strong evidence
`that the high temperature resonance losses are,
`indeed, due to valence exchange, a conclusion that
`would seem to exclude this mechanism as the one
`also responsible for the low temperature peak.
`We have measured the linewidth of the uniform
`precession at 13.4 kMHz for several single crystals
`of silicon-doped YIG (YaFes-IlSiIl012) between room
`temperature and the Curie point. The crystals were
`grown from a PbO-PbF2 flux using high purity
`starting materials to which controlled amounts of
`Si02 were added. Silicon concentration was de(cid:173)
`termined from spectrographic analysis of several
`crystals (not the actual samples) selected from the
`growth run. Measurements were made on samples
`prepared as well polished spheres, 0.025" in diam(cid:173)
`eter. These were mounted in a TE-l 08 transmission
`cavity and linewidth measurements were made
`using the half-power method. 8
`Our results for the composition 6 = 0.06 are
`shown in Fig. 1. A sample containing a smaller
`silicon content (6 = 0.04) showed the same tempera(cid:173)
`ture dependence but with the magnitude of the
`losses reduced roughly in proportion to the doping.
`
`55
`TCL 1025, Page 4
`
` Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 198.65.204.101 On: Mon,
`14 Nov 2016 15:02:16
`
`