‘:a_
`
`1::
`IEE SPECIES FLOCKS OF
`ERICA’S GREAT LAKES
`ét-emlttmgdiodesfromorganic polymers
`.f‘ant transcrlptlon factor. Pit-1.dwarfsmice
`
`Z)
`’3
`' )1
`\J
`
`Opv
`
`)’
`
`,7 —‘—--‘;.A~¥— ,r
`
`R
`“7?
`.2!
`$5
`
`II
`
`n
`4‘.-
`5!
`5%
`it
`-§
`%‘
`‘J1
`
`lD.
`
`4
`:n
`{A
`.4
`\J"
`U.)
`\l
`
`O0
`
`‘
`
`Z9 OJ
`
`!\r
`
`u
`,_.
`
`\O"
`
`V ZJ
`
`>—
`
`.
`
`1
`
`TFS1075
`
`

`

`4 A sample ofthe variety offish in Lake
`
`Malawi. On page 550 Meyer et al.
`report on mitochondrial DNA variation
`
`in the cichlid fish of the great lakes of
`
`East Africa. These ‘species flocks’ are
`
`regarded as natural laboratories for the
`
`study of speciation, and the new data
`i .
`r KICKS OF
`
`
`a, was may resolve important questions relat-
`ing to their origins. See also News and
`
`Views. Photo: Andreas Spreinat
`
`
`“ EEK.
`.
`. THIS WEEK.
`reen polymer?
`
`.velopment of an electri-
`umped luminescent diode
`
`g a conjugated polymer
`'539 and 518) is a step
`
`*
`~ realizing the potential of
`semiconductors. The new
`
`{has structural and opto-
`
`ic properties suited to
`ea light—emitting displays.
`
`
`
`.
`
`. THIS WEEK. ..
`
`Gallo investigation I Ulysses’ lonely path I Public access
`to genetic release information I Germany and
`the Gulf I CNRS tracing visitors I Hubble’s
`progress I Industrial use of universities I Museum research
`I Forensics in science I Nobel Prize winners I New
`
`Sellafield inquiry I Japanese welcome
`for foreigners
`
`CORRESPONDENCE
`
`502 — 508
`
`Early AIDS I Science selling out
`
`N EWS AN D VIEWS
`
`Pulsar seen in SN1987A remnant?
`Paul Murdin
`Evolution: Flocks of African fishes
`John C Avise
`
`Astrophysics: X-ray burster is theory buster
`Michael Garcia
`
`Obituary: J S Bell 1928 — 1990
`C H Llewellyn Smith
`Transcriptional control: Scissors and helical forks
`Robert T Sauer
`
`General relativity: Space-based gravity tests
`Clifford M Will
`
`Plant pathology: The race for resistance genes
`Geoffrey North
`Electronics: Making light of polymers
`Lewis Rothberg
`Geology: A reprieve for ocean crust
`Norman H Sleep
`Cell biology: Phospholipid transfer market
`James E Rothman
`
`Daedalus: The magic bullet
`SCIENTIFIC CORRESPONDENCE
`
`Leukaemia risk and plutonium D Newton, A J Warner
`& R J Talbot I Nutrients in the early Cambrian
`M D Brasier
`
`HIV mutation rate M Nowak I Radiating bodies
`AE Rout
`
`
`NEW JOURNALS REVIEW
`
`Communication problems
`Michael Buckingham
`This issue contains Nature’s annual review of new
`
`journals. Reviews begin on page 582, with an index on
`page 583 and a list of other journals received on page 599
`
`ARTICLES
`
`Trace-element fractionation in plumes and the origin of
`HIMU mantle beneath the Cameroon line
`
`Guide to Authors
`Page 600.
`
`A N Halliday, J P Davidson, P Holden,
`C DeWolf, D-C Lee & J G Fitton
`
`,
`
`Island race
`Long-distance dispersal of seeds
`by birds may be responsible for
`the great distance between two
`subspecies of the flowering plant
`Peperomia
`ben‘eroana
`found
`respectively in the Juan Fernan-
`dez Islands in the Pacific and
`5,000 km away on the Tristan da
`Cunha archipeligo in the souh
`Atlantic. Page 549.
`
`Escape clause
`Variants of foot-and-mouth dis-
`ease virus selected for their abil-
`ity to escape neutralization by a
`monoclonal antibody are shown
`to adopt a novel mechanism of
`antigenic variation via sequence
`changes at residues not directly
`involved in antibody binding. The
`antibody binds to a distinct site,
`which
`crystallography
`shows
`adopts a different conformation
`in the variants from the wild-type
`virus. Page 568.
`
`We have a problem
`The
`so-called solar neutrino
`problem, the fact that detection
`rates for solar neutrinos are
`about one-third of those pre-
`dicted by standard models,
`is
`tackled by Elsworth et al. using
`precise seismological measure-
`ments of the Sun's core. They con-
`clude that the ‘problem' must be
`resolved within neutrino physics,
`not by revising current models of
`the Sun’s structure. Page 536.
`
`Sun and sea
`Ocean transparency measure-
`ments from the Nimbus-7 satel-
`lite, together with climatological
`density and heat—flux data, show
`that solar radiation in visible fre-
`quencies, usually assumed to be
`absorbed at the sea surface,
`in
`fact penetrates tens of metres
`beneath the surface, This may
`account
`for
`the
`discrepancy
`between predicted and observed
`sea surface temperatures. Pages
`543 and 515.
`
`c growth
`arf mutant mice are shown
`e mutations in the pit-1
`
`'
`iption factor gene, which
`. a protein containing a
`
`, meodomain and is essen-
`
`
`_r pituitary differentiation.
`
`nemonstrates a direct link
`
`
`
`
`
`
`
`
`
`n a transcription factor
`
`commitment and progres—
`
`'
`ntsin mammalianorgano-
`Is. Page 528.
`
`
`rtiger
`* .9; battle between printed
`
`"
`s and electronic inform-
`Ill systems, the casualties so
`
`ave been on the electronic
`
`. Leading off this year’s New
`
`, als Review (on traditional
`
`'
`Michael Buckingham
`
`
`
`replacements
`
`.
`rplified laboratory method
`; etermining the ozone de-
`
`.
`potentials of halocarbon
`
`eunds is described on page
`
`.4, and used to estimate the
`
`‘
`es of two halocarbons suit-
`
`:for use in fire extinguishers
`, One intended as solvent. All
`
`‘ new compounds are broken
`
`efficiently in the tropo-
`
`6, and so show promise as
`
`' cements for today’s CFCs
`halons.
`
`‘
`' " (ISSN 0028-0836) is published weekly on Thursday. except the last week in December, by Macmillan Magazines Ltd (4 Little Essex Street, London WC2R3LF). Annual subscription for USA
`nada US$350 (institutional/corporate) , US$135 (individual making personal payment). USA and Canadian orders to: Nature, Subscription Dept, PO Box 1733, Riverton, NJ 08077-7333, USA.
`
`l’ders to Nature, Brunei Road, Basingstoke, Hants RGZl 2X8. UK. Second class postage paid at New York, NY 10012 and additional mailing offices. Authorization to photocopy material for
`”I Personal use, or internal or personal use of specific clients, is granted by Nature to libraries and others registered with the Cop right Clearance Center (CCC) Transactional Reporting
`
`OVided the base fee of $1.00 a copy plus $0.10 a page is paid direct to CCC, 27 Congress Street, Salem, MA 01970, USA. Identi ication code for Nature: 0028-0836/90 $l.00 + $0.10. US
`5‘“ Send address changes to: Nature. 65 Bleecker Street, New York, NY 10012. Published in Japan by Nature Japan K.K,, Shin-Mitsuke Bldg, 361chigaya Tamachi, Shinjuku—ku, Tokyo 162.
`, @1990 Macmillan Magazines Ltd.
`2
`
`2
`
`

`

`
`
`
`beyond the standard model of the electroweak
`l:
`
`H JimE; accepted 24 August 1990.
`.. )saak. G. R.. McLeod, C. P., van der Raay. H. B. & Roca Cortes, T. Nature 232, 591—594
`
`'
`
`
`r, F. L, & Gough. D, 0. A. Rev. Astr. Astrophys. 22. 5934619 (1984).
`J. N. Prac. Int Cant. on Neutrino Physics and Astrophysics Vol. 2 (eds Cence, R. J..
`Roberts. A.) 253 (University of Hawaii, 1981).
`
`K. 81 Rhodes, E, J. Jr Astrophys. J. 265, 551663 (1983).
`o. 0. & Novotny, E. Sol. Phys. 128. 1437160 (1990).
`
`"sen-Dalsgaard, J, in Solar Interior and Atmosphere (eds Cox, A. N. Livingston. W. C. &
`M.) (University of Arizona Press. in the press),
`
`N, Guzlk, J. A. & Raby, S. Astrophys. J. 353, 698—711 (1990).
`,
`,E. & Maeder. A. Astr. Astrophys 96, 1-16 (1981).
`
`Y., Berthomieu, G. & Provost, J, Advances in Helio and Asteroseismolagy, IAU 123 (eds
`n-Dalsgaard, J, 81 Frandsen, 8.) 95—98 (Reidel, Dordrecht, 1988),
`
`D. N. & Press, w. H. Astrophys. J. 294. 663—673 (1985),
`a. Gilliland. R. L. Astrophys. J. 299, 994—1000 (1985).
`
`
`ande. K., Lee, C. K., Cleveland, B. T. & Ullman, J, Inside the Sun (eds G. Berthomieu,
`er, M.) 1717177 (Kluwer, Dordrecht, 1990).
`
`K. S. et al. Inside the Sun (eds Berthomieu, G. & Cribier, M.) 179-186 (Kluwer, Dordrecht,
`
`R, H. 8: Goldenberg. H. M. Astrophys. J. Supp. 21, 1317182 (1974).
`on, D. s. 3. Carter, w, E. Nature 310, 572—574 (1984).
`‘
`,J. H. & Weisberg, J. M. Astrophys. J. 345, 434—450 (1989).
`
`,R. H,, Kuhn, J, R, & LiberCht, K. G. Astrophys. J. 318. 451—458 (1988).
`a, A., Isaak, G. R,, McLeod, C. P., van der Raay, H, B, & Roca Cortes, T. Nature 293, 443—445
`
`)d, M. F. Nature 309, 530—532 (1984).
`
`ll, J. N. Neutrino Astrophysics Ch. 1 (Cambridge University Press, 1989).
`sen-Dalsgaard, J. Proc. Symp. Seismology of the Sun and Sun-like Stars (ed. Rolfe,
`431—450 (European Space Agency, 1988).
`
`H. A. Phys. Rev. Lett. 56. 1305—1308 (1986).
`'
`l-n, R. 9., Noyes, R. W, & Simon, G, W. Astrophys, J. 135, 474—499 (1962).
`5, J. R,, lsaak G. R. & van der Raay, H. B. Mon. Not R. astr. Soc. 185, 1—17 (1978).
`
`rth, Y., Howe, R., Isaak, G. R., McLeod, C. P. & New, R. Nature 345, 322—324 (1990).
`
`; rd. M, F. & Noyes. R. W. Nature 318, 449450 (1985).
`5., Fossat, E. & Grec, G. Astr. Astrophys. 200, L29—L31 (1988).
`
`P. L., Regulo, C. & Roca Corte, T. Astr. Astrophys. 224, 2537258, (1989).
`,J. N. & Ulrich, R, K. Rev. mod. Phys. 60, 297—372 (1988).
`
`, J., Gough. D. 0, & Vania, M. N. Nature 32.1. 226—229 (1986).
`W., Gilliland, R. 81 Christensen-Dalsgaard, J. Nature 321, 229—231 (1986),
`
`, A. et aI. Advances in Helio and Astroseismology, IAU 123 (eds Christensen-Dalsgaard,
`randsen. S.) 205—209 (Reidel, Dordrecht, 1988),
`T. K, & Pantaleone, J, Rev. mod. Phys, 61, 937-979 (1989).
`
`.: ENTS. We thank all members, past and present, of the Birmingham and Institute
`
`. ca de Canaries solar oscillations groups for their assistance, and J. Christensen—Dalsgaard,
`and D. O. Gough for their comments. This work was funded by SERC and by the CAICYT.
`
`
`
`:1 t-emitting diodes based on
`i iugated polymers
`
`Burroughes*:l:, D. D. C. Bradley*, A. R. Brown*,
`
`— "'Marks*, K. Mackay*, R. H. Friend*, P. L. Burnsi‘
`B. I-lolmesiL
`
`endish Laboratory. Madingley Road, Cambridge CBB OHE, UK
`
`, rsity Chemistry Laboratory, Lenstield Road,
`.
`1 ridge CB2 1EW, UK
`
`‘GATED polymers are organic semiconductors, the semicon-
`"l“ 1 behaviour being associated with the 11 molecular orbitals
`‘ 7W1ized along the polymer chain. Their main advantage over
`‘
`.7' olymeric organic semiconductors is the possibility of process-
`
`“f the polymer to form useful and robust structures. The response
`‘
`e System to electronic excitation is nonlinear—the injection
`
`l electron and a hole on the conjugated chain can lead to a
`‘
`
`~0calized excited state which can then decay radiatively, sug-
`7 llg the possibility of using these materials in electrolumines-
` ‘ Jdevices. We demonstrate here that poly(p-phenylene vinylene),
`. rated by way of a solution-processable precursor, can be used
`Jule active element in a large-area light-emitting diode. The
`
`displays.
`There has been long-standing interest in the development of
`solid-state light-emitting devices. Efficient light generation is
`achieved in inorganic semiconductors with direct band gaps,
`'such as GaAs, but these are not easily or economically used in
`large-area displays. For this, systems based on polycrystalline
`ZnS have been developed, although low efliciencies and poor
`reliability have prevented large-scale production. Because of the
`high photoluminescence quantum yields common in organic
`molecular semiconductors, there has long been interest in the
`possibility of light emission by these organic semiconductors
`through charge injection under a high applied field (electro-
`luminescence)”. Light-emitting devices are fabricated by
`vacuum sublimation of the organic layers, and although the
`efficiencies and selection of colour of the emission are very
`good, there are in general problems associated with the long-term
`stability of the sublimed organic film against recrystallization
`and other structural changes.
`One way to improve the structural stability of these organic
`layers is to move from molecular to macromolecular materials,
`and conjugated polymers are a good choice in that they can, in
`principle, provide both good charge transport and also high
`quantum efiiciency for the luminescence. Much of the interest
`in conjugated polymers has been in their properties as conduct-
`ing materials, usually achieved at high levels of chemical dop-
`ing}, and there has been comparatively little interest in their
`luminescence. One reason for this is that polyacetylene, the most
`widely studied of these materials, shows only very weak photo—
`luminescence. But conjugated polymers that have larger semi-
`conductor gaps, and that can be prepared in a sufliciently pure
`form to control non-radiative decay of excited states at defect
`sites, can show high quantum yields for photoluminescence.
`Among these, poly( p-phenylene vinylene) or PPV can be con-
`veniently made into high-quality films and shows strong photo-
`luminescence in a band centred near 2.2 eV, just below the
`threshold for 77' to 7r* interband transitionsg'lo.
`We synthesized PPV (I) using a solution-processable precur-
`sor polymer (II), as shown in Fig. 1. This precursor polymer is
`conveniently prepared from a,a’-dichloro-p-xylene
`through polymerization of the sulphonium salt intermediate
`(IV)“’13. We carried out the polymerization in a water/ methanol
`mixture in the presence of base and, after termination, dialysed
`the reaction mixture against distilled water. The solvent was
`removed and the precursor polymer redissolved in methanol.
`We find that this is a good solvent for spin-coating thin films
`of the precursor polymer on suitable substrates. After thermal
`conversion (typically 2250 °C,
`in vacuo, for 10 h), the films of
`PPV (typical thickness 100nm) are homogeneous, dense and
`
`(.3
`CIHZCQCHZCI ———>
`MeOH,50 0
`(III)
`
`08+ C,-
`
`(IV)
`
`1.
`2. H+
`3, Dialysis
`
`to“, «— we8 + Cl
`
`(‘I
`
`(I)
`
`250 °C, vacuum
`
`(ll)
`
`'nt address: IBM Thomas J. Watson Research Centre. Yorktown Heights, New York 10598, USA.
`
`FIG. 1 Synthetic route to PPV.
`
`RE - VOL 347 - 11 OCTOBER 1990
`
`3
`
`3
`
`

`

`LETTERS TO NATURE
`
`
`
`
`allow formation of a thin oxide coating, gold and indium 5.3
`can all be used as the positive electrode material, and"
`aluminium, magnesium silver alloy and amorphous gig;
`hydrogen alloys prepared by radiofrequency sputtering
`
`suitable as the negative electrode materials. The high Sta...
`of the PPV film allows easy deposition of the top contact
`
`and we were able to form this contact using thermal evapor
`
`for metals and ion-beam sputtering for indium oxide.
`Figures 2 and 3 show typical characteristics for devices h
`
`indium oxide as the bottom contact and aluminium as
`
`contact. The threshold for substantial charge injection 13
`below 14 V at a field of 2 X 106 V cm1,and the integrate -.'
`
`output is approximately linear with current. Figure 4 sh
`
`spectrally resolved output for a device at various temper
`The spectrum is very similar to that measured in photolu...
`
`cence, with a peak near 2.2 eV and well resolved phono
`ture9 '0.These devices therefore emit in the green—yellow 1
`
`of the spectrum, and can be easily seen under normal labo
`
`lighting The quantum efficiency (photons emitted per e
`
`injected) is moderate, but not as high as reported for s
`the structures made with molecular materialsz‘7. The qu '
`
`efficiences for our PPV devices were up to 0.05% We
`that the failure mode of these devices15 usually associateeir
`
`failure at the polymer/thin metal interface and1s probab,
`
`_
`to local Joule heating there.
`The observation and characterization of electrolumin,
`in this conjugated polymer is of interest in the study
`
`fundamental excitations of this class of semiconductor.
`the concept of self-localized charged or neutral excited
`
`in the nonlinear response of the electronic system has
`useful one. For polymers with the symmetry of PPV
`
`excitations are polarons, either uncharged (as the .
`
`exciton) or charged (singly charged as the polaron, am
`charged as the bipolaron)‘5"6. We have previously assign '
`
`photoluminescence in this polymer to radiative recomb'
`of the singlet polaron exciton formed by intrachain excita
`
`and, in view of the identical spectral emission here, we
`the electroluminescence to the radiative decay of the
`
`excited state. The electroluminescence is generated by r
`
`nation of the electrons and holes injected from oppos
`
`of the structure, however, and we must consider what th
`
`carriers are. We have previously noted that bipolarons, th
`
`stable of the charged excitations in photoexcitation and on
`doping studies, are very strongly self-localized, with m
`
`of the associated pair of energy levels deep into the semi
`
`tor gap, to within 1 eV of each otherg. In contrast, the mo 2
`of these levels into the gap for the neutral polaron
`
`which one- electron models predict to be the same as
`bipolaron15, is measured directly from the photoluminmi
`
`
`(mA)
`
`Voltage (V)
`FIG. 2 Current-voltage characteristic for an electroluminescent device hav~
`ing a PPV film 70 nm thick and active area of 2 mmz, a bottom contact of
`indium oxide, and a top contact of aluminium. The forward-bias regime is
`shown (indium oxide positive with respect to the aluminium electrode).
`
`uniform. Furthermore, they are robust and intractable, stable
`in air at1 room temperature, and at temperatures >300 °CIn a
`vacuum11
`Structures for electroluminescence studies were fabricated
`with the PPV film formed on a bottom electrode deposited on
`a suitable substrate (such as glass), and with the top electrode
`formed onto the fully converted PPV film. For the negative,
`electron-injecting contact we use materials with a low work
`function, and for the positive, hole-injecting contact, we use
`materials with a high work function. At least one of these layers
`must be semi-transparent for light emission normal to the plane
`of the device, and for this we have used both indium oxide,
`deposited by ion-beam sputtering14 and thin aluminium (typi-
`cally 7—15 nm). We found that aluminium exposed to air to
`
`units)
`
`E”
`S
`,
`f
`S
`3
`5
`a
`2
`E
`_
`:l
`g
`a
`
`
`
`
`I'E'\
`
`E
`D
`C
`8
`A
`
`120K
`147K
`200K
`235K
`285K
`
`\
`./
`\
`/
`/
`. ,\
`./ /D \\ /
`/
`/
`\,l
`/ /
`\\
`/
`/
`./-
`//J ,’€\\
`\\
`.//
`I,
`z
`I
`‘/
`llé‘\\\\
`///
`//
`X
`// /,»"’ I,“
`Xxx/m“ /\~—
`
`/'
`
`\
`
`v
`
`4
`
`

`

`.
`,1, rge carriers, their mobilities are low and the strong
`
`ization of the bipolaron evident in the positions of the
`
`< probably does not leave sufficient energy for radiative
`. m the photon energies measured here. Therefore, the
`
`" arriers involved are probably polarons. The evidence
`y can combine to form polaron excitons requires that
`
`7
`on gap states move no further into the gap than those
`‘laron exciton and may account for the failure to observe
`
`w, 3] transitions associated with the polaron.
`V Lphotolumincscence quantum yield of PPV has been esti-
`
`i. to be ~8%. It has been shown“"17 that the non-radiative
`(es that limit the efficiency of radiative decay as measured
`
`filuminescence are due to migration of the excited states
`sites which act as non-radiative recombination centres,
`, at high intensities, to collisions between pairs of excited
`
`ese are processes that can, in principle, be controlled
`
`design of the polymer, and therefore there are excellent
`ties for the development of this class of materials in a
`E
`electroluminescence applications.
`
`ey,
`by,
`~
`
`
`
`
`
`August: accepted 18 September 1990,
`,P. 5., Barlow, W, A., Hann, R. A, & Roberts. G. G. Thin Solid Films 94, 4767488 (1982).
`W. & VanSlyke. S, A. Appl. Phys. Lett 51, 913—915 (1987).
`
`W., VanSlyke, S. A. & Chen. C. H. J. appl. Phys, 65, 3610A3616 (1989).
`i, C Tokito. S., Tsutsui, T. Gt Saito. 5. lap. J. app]. Phys, 21, 59—61 (1988).
`
`,C., Tsutsui, T. & Saito, 8, Appl. Phys. Lett. 55, 1489—1491 (1989).
`C., Tsutsui. T. & Saito, S. Appl. Phys. Lett 56, 799—801 (1989).
`
`Hasegawa. M.. Hosohawa, C., Tokailin, H, E. Kusomoto, T. Chem. Lett. 189—190(1990).
`et al. Nature 327, 403—405 (1987).
`
`, Bradley, D. D. C, & Townsend. P. D, J Phys. 020. 1367-1384 (1987).
`. D, C. & Friend, R, H, J. Phys: Condensed Matter 1, 36713678 (1989).
`. D. c. J. Phys. 020. 13894410 (1987).
`
`Ohnishi, T., Noguchi. T. 8. Hirooka, M. Synthetic Metals 11, 6394544 (1987).
`mlth, J. D.. Lenz, R. W. & Wegner, G. Polymer 30, 10484053 (1989).
`m. J. R.. Phillips, W. A. 8: Ad<ins. C. J. J Phys: Condensed Matter 2, 620743221 (1990).
`. K.. Bishop. A. R. & Campbell, D, K. Phys, Rev. B21. 4804—4825 (1983).
`Bkii, S. A. & Kirova, N. N. JEPT Lett. 33. 4—8 (1981).
`ey, D. D, C. et al, Springer Ser. Solid 3L Sci. 76, 107—112 (1987).
`
`I
`= ENTS. We thank J. R. Gellingham, C. J. Adkins and W. A. Phillips for their help in
`
`the indium oxide films. We thank SERC and Cambridge Research and Innovation Ltd for
`
`
`
`x.pospheric lifetimes of three
`
`: pounds for possible
`
`Iacement of CFC and halons
`
`1. Brown, C. E. Canosa-Mas, A. D. Parr, K. Rothwell
`
`k" P. Wayne
`‘.| Chemistry Laboratory, South Parks Road, Oxford OX1 3QZ, UK
`
`
`RINE and bromine have been implicated in the massive
`
`, me depletion of stratospheric ozone over Antarctica]. As
`‘i'lu'ce of these halogens is anthropogenic emissions of halo-
`
`-’es,
`the problem has acquired political and economic
`‘t-‘mcancez. The chemical industry has been forced to consider
`
`my 1y possible replacements for conventional halocarbons, and
`
`7 potential effects on the environment of proposed alternative
`
`_
`. Iunds have been evaluated recently in the Alternative
`l'ocarbon Environmental Acceptability Study (AFEAS)3.
`
`such compounds that are not
`included in the AFEAS
`are CFzBrH (halon 1201). CF3CFBrH (halon 2401) and
`CF2CCIZH (HCFC 225ca). Removal of these compounds from
`
`tmosphere will occur primarily by reaction with the hydroxyl
`" 1, OH (ref. 4). Here we determine, from laboratory studies,
`41 solute rate of reaction between these three species and OH.
`
`netic data are vital for assessing their viability as replace-
`
`
`‘RE ' VOL 347 - 11 OCTOBER 1990
`
`‘
`
`Our approach shows how laboratory measurements can provide a
`useful first estimate of the environmental acceptability of com-
`pounds of this sort.
`CFZBrH and CF3CFBrH are being considered as substitutes
`for CFgBr in fire extinguishers, and CF3CF2CC12H has useful
`properties as a solvent. The efficiency with which a unit mass.
`of each halocarbon will destroy stratospheric ozone relative to
`the most important ozone—depleting molecules CFC-11 (CFC13)
`and CFC-12 (CFZCIZ) will depend strongly on the atmospheric
`lifetime of the compound. The daytime degradation of these
`hydrogen-containing molecules, RH, will occur primarily by
`reaction with the OH radical4
`
`k
`OH+RH —l> HZO+R
`
`to know the rates of the
`therefore important
`is
`and it
`homogeneous gas-phase reactions of OH with RH.
`We have determined values of the second-order rate constants,
`k1, for the hydrogen-atom abstraction process (1) and their
`variation with temperature for
`the three hydrohalocarbon
`molecules. Our experimental procedure and data analysis are
`described fully elsewhere5'7. We used an absolute discharge
`flow technique with a movable injection point for the reactant
`to provide time resolution; OH concentrations were measured
`by resonance fluorescence. The variation of OH concentration
`with injector position was monitored with a known excess of
`hydrohalocarbon added. The logarithm of the OH signal was a
`linear function of calculated contact time, and the gradient was
`plotted against the concentration of hydrohalocarbon to give a
`straight line of slope k1. Our values for [C1 are given in Table 1
`along with the experimental conditions. We used a conventional
`Arrhenius equation, In k1=lnA—E / RT, to analyse the tem—
`perature-dependence. Thus a plot of In k1 as a function of 1/ T
`should give a straight line of slope E / R and intercept in A.
`Values of the activation energies E in the form E / R and
`pre-exponential factors A are given in Table 1. Errors on the
`slope, calculated in a linear least-squares analysis, are the 95%
`confidence limits. The 95% confidence limits on the intercept
`were unrealistically large, as is often the case in calculating
`Arrhenius parameters: we do not quote the errors here. The
`predictive capacity of the experimental Arrhenius expression is
`not properly represented by the individual errors on E and A,
`but rather by the combined expression incorporating both
`parameters.
`We paid particular attention to the possible presence of fast-
`reacting impurities in the hydrohalocarbon samples, which
`could seriously afiect
`the measurement of k1. All of the
`
`.\0
`
`
`
`
`
`O
`
`\\\ \
`
`
`
`
`
`
`\ \K 2
`
`CF CF CO H
`
`.\
`
`e—CF3CFBrH
`
`
`
`
`
`O
`
`
`
`2-5
`
`3-0
`1.000 / T (K)
`
`35
`
`FIG. 1 Arrhenius plots for reactions of RH with 0H.
`
`5
`
`5
`
`