a r X i v :a s t r o -p h /9911042v 1 3 N o v 1999
Hard X-ray Emission From Low Mass X-ray Binaries
D.Barret,J.F.Olive &L.Boirin
Centre d’Etude Spatiale des Rayonnements,CNRS/UPS,9Avenue du Colonel Roche,31028Toulouse Cedex 04,F
Didier.Barret@cesr.fr
C.Done
Department of Physics,University of Durham,South Road,Durham DH13LE,England,UK
chris.done@https://www.doczj.com/doc/1c3464362.html,
G.K.Skinner
School of Physics and Astronomy,University of Birmingham,Edgbaston,Birmingham B152TT,UK
gks@https://www.doczj.com/doc/1c3464362.html,
J.E.Grindlay
Harvard Smithsonian Center for Astrophysics,60Garden Street,Cambridge,MA 02138,USA
josh@https://www.doczj.com/doc/1c3464362.html,
ABSTRACT
We report on Rossi X-ray Timing Explorer observations of four type I X-ray bursters;namely 1E1724–3045,GS1826–238,SLX1735–269and KS1731-260.The ?rst three were in a low state,with 1-200keV X-ray luminosities in the range ～0.05?0.1L Edd (L Edd :Eddington luminosity for a neutron star =2.5×1038ergs s ?1),whereas KS1731-260was in a high state with a luminosity ～0.35L Edd .The low state sources have very similar power spectra,displaying high frequency noise up to ～200Hz.For KS1731-260,its power spectrum is dominated by noise at frequencies 20Hz;in addition a quasi-periodic oscillation at 1200Hz is detected in a segment of the observation.The 1-200keV spectra of the low state sources are all consistent with resulting from thermal Comptonization with an electron temperature (kT E )around 25-30keV.For KS1731-260,the spectrum is also dominated by thermal Comptonization,but with a much lower kT E ～3keV and no signi?cant hard X-ray emission.With the exception of GS1826–238,there is an underlying soft component,carrying at most ～25%of the total 1-200keV luminosity.For all sources,we have detected an iron K αline at 6.4keV (although it is weak and marginal in 1E1724–3045).A re?ection component is present in the spectra of GS1826–238and SLX1735–269,and for both we ?nd that the re?ecting medium subtends only a small solid angle (?/2π～0.15,0.28).The origin of the line and the re?ection component is most likely to be irradiation of the accretion disk by the X-ray source.
We suggest a model in which the region of main energy release,where hard X-rays are produced would be an optically thin boundary layer merged with an Advection Dominated Accretion Flow (ADAF),and would be responsible for the rapid variability observed.The soft component observed probably represents the unscattered emission from an optically thick accretion disk of variable inner radius.When the accretion rate increases,the inner disk radius shrinks and the strength of the re?ected component and associated iron line increase.At the same time,the Comptonization region cools o?in response to an increased cooling ?ux from the accretion disk and from the reprocessed/re?ected component,thus leading progressively to a quenching of the hard X-ray emission.If low state NSs accrete via ADAFs,the observation of X-ray bursts,indicating that all the accreting matter actually accumulates onto the NS surface,argues against the existence of strong winds from such accretion ?ows.
Finally,we discuss two criteria recently proposed to distinguish between non-quiescent Black Holes (BHs)and Neutron Stars (NSs),and that are not contradicted by existing observations.The ?rst
1.Introduction
It is now well established that hard X-ray emis-
sion(E 30keV)from X-ray binaries is not exclu-
sively associated with Black Hole systems(BHs).
The major breakthrough came with the SIGMA
and BATSE observations,which provided the?rst
unambiguous detections of type I X-ray bursters
(hence Neutron Star systems,NSs)at～100keV
(Barret&Vedrenne1994,Barret&Tavani1997
and references therein).However,due to the mod-
erate sensitivity and low spectral resolution of
these instruments and the lack of simultaneous X-
ray observations,it was impossible to investigate
the conditions under which NSs emit hard X-rays.
Similarly,very little could be inferred about the
accretion geometry and the relative contribution
of the potential emitting regions to the total emis-
sion(NS surface,boundary layer,accretion disk,
corona).In addition,the hard X-ray data alone
were not good enough to discriminate between the
thermal and non-thermal models that have been
put forward to account for this emission.Finally,
if the emission of hard X-rays is indeed common
to BHs and NSs,there may remain some di?er-
ences either in the shape of the hard tails(posi-
tion of the energy cuto?;e.g.Tavani&Barret
1997,Churazov et al.1995),or in the luminosities
that can be radiated simultaneously in soft and
hard X-rays1(Barret et al.1996,Van Paradijs
&Van der Klis1994,Barret&Vedrenne1994,
Zhang et al.,1996).All these potential di?erences
have yet to be quanti?ed precisely.They can now
be addressed through observations performed by
the Beppo-SAX and the Rossi X-ray Timing Ex-
plorer(RXTE)satellites;the latter combining
the Proportional Counter Array(PCA)and the
High Energy X-ray Timing Experiment(HEXTE)
(Bradt et al.1993).These two instruments o?er
for the?rst time the possibility of observing NSs
with good sensitivity and excellent timing capa-
bilities simultaneously from2to～150keV.
In this paper,we report on the RXTE ob-
servations of four type I X-ray bursters;namely
1E1724–3045,GS1826–238,SLX1735–269and
KS1731-260.All these sources have in common
that they have already been detected up to～100
cosθ=2.7
√
km(at10kpc corresponding to R in
nent was?tted with a Comptt model(Titarchuk 1994)with an electron temperature of～27keV, an optical depth～3.3for a spherical scattering cloud,and a temperature of0.6keV(kT W)for the seed photons.The ratio of the bolometric(0.1-100 keV)luminosity of the soft component to the total luminosity was11%and36%for the BB and MCD models respectively.However,as pointed out by Guainazzi et al.(1998),whereas the radius de-rived from the BB?t is close to the NS radius,the very small value inferred for R in2appears unphys-ical(for any plausible values of the source inclina-tion);hence the hypothesis that the soft compo-nent comes from the boundary layer around the NS might be preferred.
An ASCA observation con?rmed the hardness of the source in X-rays,and allowed an accurate determination of N H towards the source(～1.2×1022H atoms cm?2).This value is consistent with that expected from the optical reddening of the cluster(Barret et al.,1999a).
cosθ(for an as-
sumed distance),which are the color temperature of the in-ner accretion disk and the projected inner disk radius(note
that in the case of a pseudo-newtonian disk with zero-stress
inner boundary conditions around a Schwarzschild BH,the
actual inner disk radius is2.73?1times less than the mea-
sured R in,see Gierlinski et al.1999).Correction for spec-
tral hardening must be made to kT in and R in to account
for the fact that the inner disk opacity is dominated by elec-
tron scattering.We use a spectral hardening factor f=1.7, a value consistent with the source luminosity(Shimura&
Takahara1995),and determine the e?ective temperature
as kT e?=kT in/f,and the e?ective inner disk radius as
R e?=f2R in=2.9R in.For1E1724–3045,this means that
kT e?=0.82keV,R e?=11.4km for an assumed inclination
angle of60?(the source is not a dipper,therefore its inclina-
tion must be less than75?).If we further assume that the
disk terminates at the last stable orbit,then as described in Ebisawa et al.(1994),R S(the Schwarzschild radius)is
related to R e?as3R S≈9M NS km∝3
cosθby relativistic e?ects,see Ebisawa
et al.1994,here we takeη=0.6as in Shimura&Takahara
(1995)).From this,the low inferred value of R e?implies
a very low mass of0.4M⊙for the NS(for a1.4M⊙NS, R e?should be37km,equivalent to R in～13km).This does not seem to favor the MCD model used to?t the soft
component in the Beppo-SAX data of1E1724–3045.How-
ever,one has to keep in mind that R in is derived from the
normalization of the MCD model,and therefore would be
underestimated if some fraction of the disk?ux is scattered
in a corona.The measured value of R in derived from the
unscattered fraction would then depend on the geometry, and the optical depth of the scattering corona.2.2.GS1826–238
When GS1826–238was discovered serendipi-tously by GINGA in1988,it was?rst classi?ed as a transient source,and further classi?ed as a Black Hole Candidate(BHC)based on its rapid and in-tense?ickering and hard Power Law(PL)spec-trum in X-rays(Tanaka1989).The source was later detected by TTM(In’t Zand1992),ROSAT-PSPC(Barret et al.1995a),and at hard X-ray energies by OSSE(Strickman et al.1996).The BH nature of GS1826–238was questioned(e.g. Strickman et al.1996,Barret et al.1996),but the?rm evidence that GS1826–238did not con-tain a BH came with the discovery of type I X-ray bursts(the unmistakable signature of a NS)with the Beppo-SAX Wild Field Cameras(WFC,Uber-tini et al.1997).In the WFC data,the70bursts observed over more than2.5years of source mon-itoring exhibit a quasi-periodicity of5.76hours in their occurrence times(Ubertini et al.1999).In addition,the so-calledαparameter which is the ratio between the average persistent X-ray?ux to the average?ux emitted during X-ray bursts(as-suming that both are isotropic)was estimated to be60±7for two of the70bursts analyzed by Ubertini et al.(1999).Similarly In’t Zand et al. (1999),using also a sample of two bursts,observed α=54±5consistent with the previous estimate. This value is consistent with a picture in which all the accreting material is accumulated onto the NS and helium is burned during the bursts(αshould be～20and～80for pure Hydrogen and Helium burning respectively).
As is the case for1E1724–3045,broad band observations of GS1826–238already exist.First, Strickman et al.(1996),combining non simul-taneous GINGA(X-ray)and OSSE(hard X-ray) data showed that the2-200keV spectrum of the source could be?tted by an exponentially cuto?power law(CPL:Γ=1.76±0.02,E Cuto?=58±5 keV,Γis the photon index of the power law, i.e.the energy index would beΓ-2)with a weak (and marginally signi?cant)re?ection component. More recently,Beppo-SAX observed the source, following a BATSE trigger which indicated that GS1826–238was emitting hard X-rays(Frontera et al.1998,Del Sordo et al.1999).The0.5-200keV spectrum was well?tted by a compos-ite model consisting of a blackbody of0.94±0.05 keV,and a PL of photon indexΓ=1.34±0.04,
exponentially cuto?at49±3keV.The column
density measured by Beppo-SAX was0.47×1022 H atoms cm?2,consistent with the ROSAT PSPC
value(0.5±0.04×1022H atoms cm?2,Barret et al.1995);both being slightly larger than the value expected from the optical reddening towards the
source(E(B-V)～0.4corresponding to0.22×1022 H atoms cm?2,Predehl&Smith1995,Barret et al.1995).Another Beppo-SAX observation
performed half a year before by In’t Zand et al. (1999)yielded best?t spectral parameters consis-
tent with those reported in Del Sordo et al.(1999).
GS1826–238has a V=19.1±0.1optical coun-terpart(Barret et al.1995,Homer et al.1998)
in a possible2.1hour orbital period system,and is therefore a Low Mass X-ray Binary(LMXB).If
the2.1hour modulation is the true orbital period of the system,one can try to estimate the source distance following the approach of Van Paradijs&
McClintock(1994).In LMXBs,the bulk of the op-tical light originates from the reprocessing of the X-rays in the accretion disk.For systems for which
reliable distance estimate exists,there is a good correlation between the absolute V magnitude and
the X-ray luminosity(L X)and the size of the ac-cretion disk(and hence the orbital period P orb, L V∝L1/2X P orb2/3).Most LMXBs have M V in the range0-2,whereas some short period systems
have M V in the range3-5.With a2.1hr orbital period,GS1826–238is a short period system,and
if one assumes M V in the above range(and an ab-solute dereddened V magnitude of17.9),one gets a formal distance range between～4and～10kpc. Constraints on the source distance can also be in-ferred from the observation of X-ray bursts.In’t Zand et al.(1999)derived an unabsorbed bolo-metric peak?ux of a burst of2.7±0.5×10?8ergs s?1cm?2.The fact that this burst did not show evidence for photospheric expansion implies that its luminosity is below the Eddington limit(L Edd) which we assume to be2.5×1038ergs s?1(this value is appropriate for helium-rich material,a1.4 M⊙NS,and a moderate gravitational redshift cor-rection,Van Paradijs&McClintock1994).This in turn implies an upper limit on the source dis-tance of9.6kpc.In the remainder of this paper, we assume a distance of7kpc for GS1826–238.2.3.SLX1735–269
SLX1735–269was reported for the?rst time in 1985,when it was detected by the Spacelab2X-ray Telescope(Skinner et al.1987).However,it was present in the Einstein Slew Survey in ob-servations performed～5years earlier(Elvis et al.1992).SLX1735–269was later detected by TTM(In’t Zand1992),by ART-P,ROSAT-PSPC (Grebenev et al.1996)and ASCA(David et al. 1997).SIGMA observations have shown that it is a persistent hard X-ray source of the Galactic center region(Goldwurm et al.1996).The weak-ness of the source(～15.4mCrab in the35-75 keV range)did not allow tight constraints to be put on the time averaged(1990-1994)hard X-ray spectrum,which could be?tted either by a simple PL of photon indexΓ～2.9±0.3,or alternatively by a Comptonization model(Compst in XSPEC, Sunyaev and Titarchuk,1980)with kT E of26+35
?9 keV(Goldwurm et al.1996).Although early sus-pected to contain a NS based on the softness of its hard X-ray spectrum(Goldwurm et al.1996), its nature remained uncertain until type I X-ray bursts were discovered with the Beppo-SAX WFC (Bazzano et al.1997).The ASCA observations of SLX1735–269revealed that its0.6-10keV spec-trum could be well?tted with a PL of index2.15 (David et al.1997),absorbed through an N H of ～1.4?1.5×1022H atoms cm?2,a value which is consistent with a source location near the Galactic center(i.e.at a distance of～8.5kpc,David et al. 1997).The N H value derived with ASCA is also consistent with the one derived from the ROSAT-PSPC and ART-P observations(1.2?1.4×1022 H atoms cm?2,Grebenev et al.1996).The rapid variability of the source was recently investigated by Wijnands&Van der Klis(1999),using a short 10kilosecond Rossi X-ray Timing Explorer ob-servation performed between February and May, 1997(see below).
No accurate distance estimate exists so far. From the burst reported in Bazzano et al.(1997a) (1.5×10?8ergs s?1cm?2,2-10keV,Bazzano et al.1997b),an upper limit of～10kpc can be de-rived(after bolometric corrections and assuming a blackbody of2keV for the burst).In the remain-der of the paper,we will assume a distance of8.5 kpc.
2.4.KS1731-260
KS1731-260was discovered in October1988by TTM(Sunyaev et al.1990),and then classi-?ed as a transient.KS1731-260is located in the Galactic center region,about5?away from the Galactic nucleus and only1?from the X-ray pul-sar GX1+4.Above the mean persistent level of 80mCrab(2.7×10?10ergs s?1cm?2,1-20keV), TTM observed several X-ray bursts from KS1731-260,thus indicating that it contains a weakly mag-netized NS.The TTM spectrum was?tted by a Thermal Bremsstrahlung(hereafter TB3)of5.7 keV,absorbed through an N H of2.2×1022H atoms cm?2.Follow-up ROSAT and optical ob-servations indicated that the source is a likely LMXB(Barret et al.1998).The N H derived from the ROSAT-PSPC all sky survey observations of 1.3×1022H atoms cm?2for a PL?t;is about a factor of2less that the value derived with TTM, thus possibly indicating intrinsic N H variations within the source.More recently,we observed KS1731-260with ASCA(September27th,1997) and con?rmed the N H found by ROSAT(Narita et al.1999).Additional X-ray observations can be found in Yamauchi&Koyama(1990)and Alek-sandovich et al.(1995).KS1731-260was detected only once in hard X-rays between35and150keV by SIGMA,and has remained undetectable since then at these energies(Barret et al.1992).Unfor-tunately,no simultaneous X-ray observations exist for the SIGMA detection.
RXTE?rst observed KS1731-260in July-August1996.It was then in its standard high state.These observations led to the discovery of a highly coherent524Hz periodic X-ray signal at the end of the contraction phase of an X-ray burst that showed photospheric expansion(Smith et al. 1997).This signal was tentatively interpreted as an indication of the NS spin(Smith et al.1997). However,in the same data set,two simultane-ous High Frequency Quasi-Periodic Oscillations (HFQPOs)at898±3.3Hz and1158±9Hz were
Fig.1.—The RXTE/ASM long term light curves of 1E1724–3045,GS1826–238,SLX1735–269and KS1731-260.Time is given as days after January 1st,1994.Data have been selected using fselect to satisfy the criterion that the measured rate for each individual “dwells”is positive and that the reduced χ2of the ?t to recover the source inten-sity is less than 1.5.Data from all three Scanning Shadow Cameras are combined within lcurve .The binning time of the light curves is 2days.The gaps in the light curves are due to the sun getting too close to the source region (see Levine et al.1996for information about the ASM).The dates of our pointed observations are indicated with arrows.3.
Observations and Results
The PCA instrument consists of a set of 5iden-tical Xenon proportional counters covering the 2-60keV energy range with a total area of about 6500cm 2(Bradt et al.1993).The HEXTE instru-ment is made of two clusters of 4NaI(Tl)/CsI(Na)phoswich scintillation detectors providing a total e?ective area of 1600cm 2in the 15-200keV range 4(Rothschild et al.1998).The two clusters rock
97mar20c
so far(i.e by ASCA/ROSAT/Beppo-SAX).In or-der to account for the uncertainties in the relative calibration of the PCA and HEXTE experiments, we have also left the relative normalization of the two spectra as a free parameter of the model used.
To investigate the systematics in the PCA data and to validate our analysis scheme,we have ?rst analyzed a15kilo-second Crab observation (March22,1997).We have found that to get ac-ceptable?ts,it was necessary to add a system-atic error to the data to reduce the e?ects of the imperfect knowledge of the instrument response near the K and L edges of xenon.These errors derived from looking at the residuals of the power law?t are0.5%between2.5and15keV,1%be-tween15and25keV and2.5%above.Fitting the Crab spectrum between2.5and25keV thus yields aχ2d.o.f of1.0(50d.o.f.),a power law in-dex of2.18,and a normalization at1keV of13.5 Photons cm?2s?1keV?1(N H=3.3×1021H atoms cm?2being frozen during the?t).The ratio be-tween the data and the power law folded through the response matrice is shown in Fig.8.A sim-ilar level of systematics has been used by several groups(e.g.Rothschild et al.1999,Wilms et al. 1999,Bloser et al.1999).Based on this,before the?tting,we have combined quadratically to the Poisson errors of our data,this level of system-atics using the FTOOLS grrpha version5.7.0.No systematic errors were added to the HEXTE data.
3.1.The RXTE observations
The RXTE observations are summarized in Fig.2,3,4and5,for1E1724–3045,GS1826–238,SLX1735–269,and KS1731-260.These?g-ures represent the HEXTE background-subtracted light curve(top panel),the PCA2-40background-subtracted light curve(underneath),the PCA color-color diagram(soft color:7-15keV/2-7 keV,hard color15-40keV/7-15keV(bottom left panel),and a PCA hardness intensity diagram (bottom right panel).For clarity and homogene-ity,the data from orbits contaminated by bursts have been?ltered out(only data recorded with the5PCA units ON are shown).
The observation of1E1724–3045started on November4th,1996(T0=11h54mn39s),and was spread over more than4days.During the observa-tion,an X-ray burst occurred on Nov.8th at05h 30mn07s(UT).As said above for KS1731-260,a few NSs exhibit nearly coherent pulsations during X-ray bursts.Sampling the burst pro?le each sec-ond,we have searched for such a coherent signal in both the science event and burst catcher mode data.No signi?cant signals were detected between 200and1500Hz.Data from the orbit when the burst occurred and the next one have been re-moved from the present analysis.The source did not display any signi?cant variability over the ob-servation.In addition,looking at the color-color diagram,one can see that it was observed in a sin-gle spectral state(see also,Olive et al.1998).A single PCA spectrum averaged over the whole ob-servation has thus been considered for the spectral
analysis.
Fig.2.—Top HEXTE25-100keV hard X-ray light curve with underneath the2-40keV PCA light curve of1E1724–3045over the4.5day span of our observation.Bottom left The color-color diagram (the soft color HR1is the ratio7-15keV/2-7keV and the hard color HR2is15-40keV/7-15keV) and Bottom right the hardness-intensity diagram for the soft color.Only PCA data recorded with the5PCA units ON are shown(this explains the ?rst gap in the PCA light curve while HEXTE was recording data).
GS1826–238was observed between November 5th and6th,1997for about50ksec also(T0=08h
22mn39s,UT).There are two bursts in the middle of the observation.As for1E1724–3045,no coher-ent pulsations were detected in those bursts.The time separation between the two bursts is～5.6 hours,consistent with the5.76hours(1σspread of0.26h)periodicity reported by Ubertini et al. (1999).In addition,from this periodicity,one ex-pects a burst to have occurred～1600seconds be-fore the beginning of our observation.The tail of this burst is clearly seen in the?rst orbit of data. Data from this orbit,as well as from those contam-inated by the two bursts,have been removed from the present analysis.The spectral analysis of these bursts and their impact on the persistent emission will be reported elsewhere.As shown in Fig.3the source intensity varies by as much as～10%in the 2-40keV range,but the source did not display sig-ni?cant spectral variability.As for1E1724–3045,a single PCA spectrum was then used in the spectral analysis.In the hard X-ray band,it is the bright-est of the sources considered here(the count rate in HEXTE-Cluster A is8.0cts s?1in the25-100 keV range).
SLX1735–269was observed in October10th, 1997(T0=04h37mn19s);its observation which was scheduled for50ksec ended2.5days later. The mean PCA count rate is165.5cts s?1(2-40 keV),and this is the faintest of the four sources considered here.No X-ray bursts were observed. The source is clearly variable on minute time scales.However it is clear that these intensity variations are not accompanied by strong spectral changes,as the source remains in a limited region in the color-color diagram.
Finally,the observation of KS1731-260started on October28th,1997(T0=22h20mn15s,UT), lasted for about50ksec,and ended on October 30th.In the PCA,KS1731-260has the largest count rate of the sources considered here.The so-called banana shape is clearly visible on the color-color and hardness intensity diagrams.It moves from the lower to the upper parts of the banana as the count rate increases(by up to～25%,2-40 keV).Given the source variability,we have consid-ered three PCA spectra in the spectral analysis; one for each day of the observation.It is detected by HEXTE only up to～50keV.No X-ray bursts were
detected.Fig.3.—Top HEXTE25-100keV hard X-ray light curve with underneath the2-40keV PCA light curve of GS1826–238over the1.2day span of our observation.Bottom left The color-color diagram (the soft color HR1is the ratio7-15keV/2-7keV and the hard color HR2is15-40keV/7-15keV) and Bottom right the hardness-intensity diagram for the soft color.The data contaminated by X-ray bursts have been removed.Only PCA data recorded with the5PCA units ON are shown. 3.2.Timing properties
The normalized Power Density Spectra(PDS) averaged over the whole observation of all four sources are shown in Fig.6.The?rst three sources show noise up to～200Hz.The integrated power of these PDS is29.1%,26.1%,27.6%in the2-40keV band(0.005-300Hz).In addition,they also show a break in the PDS at low frequen-cies(typically around0.1-0.2Hz),and a QPO-like feature around1Hz.Although the complete analysis is still under way(for GS1826–238and SLX1735–269),no strong HFQPOs were detected from any of these three systems.For1E1724–3045, for which a sensitive search has already been con-ducted,an upper limit of2.5%on the RMS of a 1000Hz QPO has been derived(5-30keV)(Barret et al.1999b).
Fig.4.—Top HEXTE25-100keV hard X-ray light
curve with underneath the2-40keV PCA light
curve of SLX1735–269over the2.5day span of our
observation.Bottom left The color-color diagram
(the soft color HR1is the ratio7-15keV/2-7keV
and the hard color HR2is15-40keV/7-15keV)
and Bottom right the hardness-intensity diagram
for the soft color.Only PCA data recorded with
the5PCA units ON are shown.
The PDS shown in Fig.6for SLX1735–269is
very similar both in shape and normalization to
the one reported by Wijnands&Van der Klis
(1999a)from a short RXTE observation per-
formed in February-May1997.Although the
source count rate did not change a lot during our
observation(see Fig.4),we have computed two
PDS for segments of the observation associated
with the highest and lowest count rates:172cts
s?1(156cts s?1)versus162cts s?1(150cts s?1),
2-40keV(2-16keV),respectively.Fitting these
two PDS with broken power laws yields a break
frequency of0.15±0.1Hz,and0.08±0.02Hz
respectively.Thus,within our limited range of in-
tensity variations,the break frequency decreased
with the count rate,a behaviour which is generally
observed from atoll sources(Van der Klis,1994).
At lowest count rate,Wijnands&Van der Klis
(1999a)reported the opposite behaviour
between
Fig.5.—Top HEXTE25-50keV hard X-ray light
curve with underneath the2-40keV PCA light
curve of KS1731-260over the1.5day span of our
observation.Bottom left The color-color diagram
(the soft color HR1is the ratio7-15keV/2-7keV
and the hard color HR2is15-40keV/7-15keV)
and Bottom right the hardness-intensity diagram
for the soft color.Only PCA data recorded with
the5PCA units ON are shown.
two segments of observations of count rates141
and113cts s?1(2-16keV);νBreak increased from
0.11to2.3Hz.Therefore,this means that like
in the case of the millisecond pulsar SAXJ1808.4-
3658(Wijnands&Van der Klis1998),it appears
thatνBreak correlates with the source count rate
(the latter most likely tracking the mass accre-
tion rate),down to a certain value,below which
νBreak and the source count rate anti-correlates.
For both SLX1735–269and the millisecond pul-
sar,this transition occurred at a luminosity of
～2?4×1036ergs s?1(3-25keV).
KS1731-260is characterized by weaker variabil-
ity with an integrated power of～3.0%(2-40keV).
Its PDS can be approximated by a PL of index
α=1.3on which a QPO centered at7.6±0.7Hz
is superposed(FWHM=3.1±2.0Hz for a gaus-
sian?t).This QPO is very similar to the so-called
“Horizontal branch”QPO observed in Z sources.
Fig. 6.—The power density spectra of all four sources averaged over the whole observations. 1E1724–3045,GS1826–238and SLX1735–269dis-play high frequency noise extending up to200-300 Hz with RMS amplitudes of29.1%,26.1%,27.6% in the2-40keV band(0.005-300Hz).KS1731-260displays Very Low Frequency Noise with a RMS amplitude of～3.0%(2-40keV).These PDS are consistent with the?rst three sources being in the so-called island state(low state),and KS1731-260being in the banana state(high state).
In addition,at higher frequencies,for the segment of the observation performed on October28th(i.e. at the lowest count rate～1050cts s?1,5-30keV), we have detected at the4.7σlevel a HFQPO cen-tered at～1200±10Hz(FWHM=45.0±20Hz, RMS=2.5±0.4%)5.This is the highest HFQPO ever detected from KS1731-260.No HFQPOs were detected in the subsequent observations performed on October29th(1285cts s?1)and30th(1355cts s?1).We have derived an upper limit of～2%on the RMS of any HFQPOs around1000Hz for the
1E1724–3045
kT E28.1+3.0
?3.025.6+3.1
?2.0
kT W 1.1+0.1
?0.11.6+0.2
?0.2
τ 2.9+0.2
?0.23.3+0.3
?0.3
kT BB,kT in0.6+0.1
?0.11.2+0.1
?0.2
R BB,R in
√
F1?20keV 1.52 1.58
F20?200keV0.900.91
f BB(%)14.427.2
cosθis the projected inner disk radius(Mitsuda et al.1984);they are both given in kilometers and scaled at the source distance (6.6kpc).EqW is the equivalent width of the iron line in eV.The1-20keV?ux(F1?20keV)and hard X-ray20-200keV?ux(F20?200keV)are given in units of×10?9ergs s?1cm?2.f BB is the contribution of soft component(BB or MCD)to the total luminosity computed in the1-200keV range.Errors in all the tables are quoted at the90%con?dence level(χ2=χ2+2.7).
and a hard Comptonized component well?tted by the Comptt model in XSPEC(Titarchuk1994, Guainazzi et al.1998,see also Barret et al.1999). The soft component is well described by a BB (χ2=88.1,81d.o.f).A MCD could?t it as well, although theχ2is slightly larger(χ2=92.3,81 d.o.f).For the Comptonizing cloud,we have de-rived an electron temperature(kT E)of～27?30 keV and an optical depth of～2.8?2.9and～1.1 assuming that it has a spherical or disk-like ge-ometry respectively.In both cases,a temperature of～1.0keV for the seed photons was derived. These parameters,which are listed in Table1,are strikingly close to those found by Guainazzi et al.(1998),suggesting that in addition to its low variability in intensity(see Fig.1),the source is also stable from the point of view of its energy spectrum.For the soft component,the best?t parameters are strikingly similar to those derived from the Beppo-SAX observations.
Independently of the model?tting the contin-uum,an excess around6keV seems to remain in the residuals.Assuming that it results from iron K?uorescence(6.4keV),and that it is nar-row(σFe=0.1keV),the Comptt+BB+Line model
yields aχ2of79.9(80d.o.f).The line equivalent width is21eV.In the remainder of this paper, we will assess the signi?cance of additional com-
ponents in the?tted model by means of the stan-dard F-test.What we call the F-test probability
will always refer to the probability of rejecting the correct hypothesis,which states that,with the ad-ditional component the?t is better.We de?ne
F m as(χ21?χ22)/(ν1?ν2)/(χ22/ν2)withχ21>χ22 andν1>ν2(whereχ21,2are the Chi-squared val-ues of the two?ts,withν1,2number of degrees of
freedom,d.o.f.);F m follows a Snedecor’s F distri-bution.In the present case,we have F m=8.2,and P(F>F m)=0.005,which means that the addition of the line to the Comptt+BB model is signi?cant at the level of～99.5%.It is worth pointing out that the signi?cance of any features in the spec-trum depends strongly on the level of systematics added to the data,and that there are some uncer-tainties in the estimate of these systematics.To illustrate this,if instead of using0.5%around6 keV,one uses1%,as for example in Rothschild et al.(1999),the signi?cance of the line would drop
Fig.7.—The data and folded model(top panel), the residuals(center panel)and the unfolded spec-trum(bottom panel,Comptonized(Comptt)com-ponent is the dashed line,multi-color disk black-body model is the dot-dashed line,the Gaussian line is the dotted line)of1E1724–3045.
down to85%,and hence the line detection would not be signi?cant.Leaving the line energy and its width as free parameters does not improve the?t signi?cantly.The presence of the line is also signif-icant at the level of98.5%for the Comptt+MCD model.However,since no such a line was found in ASCA and Beppo-SAX observations,we consider our detection very marginal.Fig.7shows the data and folded model(Comptt+BB+L),the residuals and the unfolded spectrum of1E1724–3045.
For comparison with SIGMA data(35-200 keV),we have?tted the HEXTE data alone in the25-150keV range.Although the?t is not sat-isfactory due to the presence of a clear high energy cuto?(～70keV),a PL?t yields a photon index of 2.7±0.1,to be compared with the time-averaged value observed by SIGMA(3.0±0.3,Goldwurm et al.1993,1994).Fitting the2.5-25keV PCA spec-trum with a simple PL(note an acceptable?t) yields a photon index of～2.0.We therefore con-clude that the steepness of the hard tail observed by SIGMA was arti?cial,and due to the pres-ence of a high energy cuto?around60-70keV.We have also?tted the continuum with the relativis-tic Comptonization model developed by Poutanen &Svensson(1996)(CompPS in XSPEC).For the Comptonized tail,this yields best?t parameters of 35keV for kT E and2.1forτ(spherical geometry assumed).
From this observation,we derive L1?20keV= 8.1×1036ergs s?1,and L20?200keV=4.8×1036 ergs s?1(d=6.6kpc).L1?20keV is consistent with the time-averaged value derived from the ASM light curve.
3.3.2.GS1826–238
For GS1826–238,we have again set N H to the value observed by ROSAT and Beppo-SAX(0.5×1022H atoms cm?2).To illustrate the complexity of the spectral shape,in Fig.8,we show the ra-tio between the PCA data and a PL model folded through the PCA response matrix.There is a clear excess around6.0keV.First,of the single compo-nent models,the CPL is the one that provides the best?t.This yields a cuto?energy at～95 keV and a photon index of1.7(χ2=195.0for87 d.o.f).A?t using a comptonization model(e.g. Compst in XSPEC)yields kT E of20keV and an optical depth of4.6.Adding a narrow6.4keV line (σFe=0.1keV)to account for the feature of Fig.8 improves the?t(χ2=149.4for86d.o.f)but still, the residuals show systematic deviations that in-dicate that this model is not the right description of our data.
Since a re?ected component had been found in a combined analysis of the OSSE and GINGA data (Strickman et al.1996),we have tested for the presence of such a component by adding an ab-sorption edge at7.1keV.This leads to a signi?cant decrease ofχ2(χ2=100.5for85d.o.f,F m=41.3, P(F>F m) 7.9×10?9).Thus we have substi-tuted the CPL model by the XSPEC Pexrav model (Magdziarz and Zdziarski1995),which is the sum of a CPL plus a Compton re?ected component. The inclination angle to the source is unknown, but since GS1826–238is not a dipper,we have as-sumed a standard value ofθ=60?,and we leave the re?ection scaling factor as a free parameter(we assumed solar abundance for the re?ecting ma-terial).This yields a re?ection scaling factor of ～0.20,and a narrow6.4keV line(σFe=0.1keV)
GS1826–238
Γ 1.70+0.01
?0.021.72+0.01
?0.01
E Cuto?98.7+8.0
?7.090.0+8.0
?7.0
f re?...0.15+0.03
?0.04
Line Energy 6.4(?xed) 6.1+0.3
?0.2
σFe0.1(?xed)0.48+0.25
?0.23
EqW36+8?1050+17
?13
χ2(d.o.f)149.4(86)85.2(83)
Table2:Best?t spectral results for GS1826–238.N H has been set to0.5×1022H atoms cm?2.The CPL+L model is the sum of a Cuto?Power Law(Γis the photon index,E Cuto?is the cuto?energy in keV),plus a narrow6.4keV line(σFe=0.1keV).The CPL+L+R is the same but with the re?ected component added (the CPL is substituted by the Pexrav model in XSPEC,Magdziarz and Zdziarski1995),and the line energy and width(σFe)left as free parameters of the?t.f re?is the re?ection scaling factor normalized to the CPL. It should be equal to1for an isotropic source above an in?nite?at disk.Same units for F1?20keV and F20?200keV as in Table1.
of EqW of37eV(χ2=97.1,85d.o.f.).The sig-
ni?cance of the re?ection component is very high
(F m=45.4,P(F>F m) 1.8×10?9).Leaving
the line energy and its width as free parameters
of the model,we getχ2=85.2(83d.o.f.)with the
line parameters:σFe=0.48keV,centroid energy
at6.1+0.3
?0.1keV(90%con?dence level,still con-
sistent with a?uorescent iron Kαline),and an EqW of50eV.This decrease ofχ2is signi?cant at the level of99.7%(F m=6.0,P(F>F m)=0.003), we therefore conclude that this is the best?t to our data.As a further step,we have tried a re-?ection model including ionisation:Pexriv model in XSPEC(Magdziarz and Zdziarski1995).This model?ts the data as well as the Pexrav model (χ2=95.5,84d.o.f.),and yields a disk ionisation parameter consistent with0.Therefore,with both models,our data are consistent with re?ection from a cool neutral medium.
We have also reanalyzed the GINGA spectrum of GS1826–238used in Strickman et al.(1996)and Zdziarski et al.(1999).The re?ection component is highly signi?cant,and consistent with coming also from a neutral medium also.A?t with a sim-ple power law and a6.4keV narrow line yields χ2=39.6for31d.o.f.,while adding the re?ection component yieldsχ2=16.1for29d.o.f.(F m=21.2, P(F>F m)=2.1×10?6).The best?t parame-ters areΓ=1.88±0.05,and f re?=0.8±0.3,and an equivalent width for the iron line of15eV6.
There is no soft component in the source spec-trum.With the Pexrav model,we have set a90% con?dence limit of about1%on the fraction of1-200keV luminosity in the soft component modeled by a1.5keV BB component.The results of the best?t using the CPL+Line+Re?ection model are given in Table2.The data and folded model,the residuals and the unfolded spectrum of GS1826–238are shown in Fig.9.
For comparison with hard X-ray observations of similar systems,we have?tted the HEXTE spec-trum with a PL.Although the?t is poor due to the presence of a clear cuto?in the spectrum,one gets a photon index of～2.3±0.2,similar to the value derived for1E1724–3045.Fitting the hard tail with the CompPS model yields kT E=41keV andτ=1.9.
For this observation,one derives L1?20keV=
SLX1735–269
Γ 2.06+0.01
?0.012.09+0.03
?0.04
kT in0.53+0.06
?0.070.44+0.10
?0.08
R in
√
F1?20keV0.720.75
F20?200keV0.390.30
f BB(%)9.012.6
7For SLX1735–269,theχ2associated with the best?ts are
lower than1(see Table3).This makes naturally question-
able the results of our F-test.Similarly,the errors com-
puted on the best?t parameters should not be considered
as true90%uncertainties(these errors are computed un-
der the assumption that the errors on the data are Poisso-
nian).The lowχ2indicates certainly that for SLX1735–
269,which is the faintest of the4sources,the systematics
assumed are too large.However,for consistency in the
analysis,we have chosen to set them to the values used for
the other sources.
Fig.8.—The ratio between the data and a folded power law model for the Crab,SLX1735–269and GS1826–238.The response matrice used in the Crab analysis has been obtained using the same procedure as for the other sources.For GS1826–238and SLX1735–269,this plot indicates the pres-ence of a strong feature around 6-7keV.
ponent is statistically signi?cant based on our F-test.The EqW found (67eV if the line is narrow,σFe =0.1keV)is well below the upper limits of 150eV (σFe =0.1keV)derived by David et al.(1997)using ASCA SIS data.
However,by looking at the residuals above ～7keV the data show systematic deviations with an edge-like shape.An edge around these energies is suggestive of the presence of a re?ection compo-nent.We have therefore ?rst substituted the PL model with the Pexrav model in XSPEC (assum-ing no energy cuto?in the model,and a 6.4keV line with a σFe =0.1keV,an inclination angle of 60?).This leads to χ2=31.4(58d.o.f.),indicating that the presence of a re?ected component is sig-ni?cant at a level greater than 99.99%(F m =23.1,P (F >F m )=1.110?5).The solid angle inferred is ～0.26and the line EqW is 47eV.Whether or not the re?ector is ionised has been tested by
us-
Fig.9.—The data and folded model (top panel),the residuals (center panel)and the unfolded spec-trum (bottom panel)of GS1826–238.The iron line at 6.4keV is represented by the dashed line,the re?ection component by the long dashed line.ing the Pexriv model.The statistical quality of the data does not allow us to determine the line energy simultaneously with the Compton re?ected com-ponent.We have assumed the line at 6.5keV (ex-pected from a moderately ionised medium).This yields a χ2=28.4(57d.o.f.),an ionisation param-eter,albeit poorly constrained,of ξi ～70+740?70,a solid angle of ～0.28and a line EqW of 39eV (see Table 3).The improvement of χ2from the Pexrav to the Pexriv model is signi?cant at the level of 98.3%(F m =8.8,P (F >F m )=4×10?3).It is worth pointing out that leaving out the line at 6.4keV increases the signi?cance of the ionisa-tion of the re?ector,but then the model becomes unrealistic,as the line is not expected to be at 6.4keV for the ionisation parameters derived from the ?t (ξi 70).To conclude,one can say that our data for SLX1735–269are consistent with re?ec-tion from a neutral or moderately ionised medium.In the absence of any cuto?s in the useful en-ergy range of the present measurements,comp-tonization models cannot really be used.We note
Fig.10.—The data and folded model (top panel),the residuals (center panel)and the unfolded spec-trum (bottom panel)of SLX1735–269.The power law component is represented by the dashed line,the multi-color disk blackbody model by the dot-dashed line,the iron 6.4keV line by the dotted line and the re?ection component by the long-dashed line.The Raymond-Smith thermal spectrum cor-responding to the galactic di?use emission around SLX1735–269is also indicated as a dot-dot-dot dashed line.
however that a Comptonization model for which the electron temperature would be ～30keV (as seen in 1E1724–3045)could ?t the continuum as well as the PL.This would yield an optical depth of ～3for the scattering cloud.Similar results were obtained by combining simultaneous ASCA and SIGMA data (Goldwurm et al.1995).The best ?t parameters are listed in Table 3while the data and folded model,the residuals and the un-folded spectrum are shown in Fig.10.
For this observation,we derive L 1?20keV =6.0×1036ergs s ?1,and L 20?200keV =3.5×1036ergs s ?1(d=8.5kpc).L 1?20keV is about a factor of two lower than the time-averaged value derived from the ASM light curve.
3.3.
4.KS1731-260
KS1731-260was observed in a high state,and its spectrum is clearly much softer than the other three sources examined before.We have assumed the N H measured by ROSAT and ASCA (i.e.1.2×1022H atoms cm ?2).As the source inten-sity increased smoothly during our 3-day observa-tion,and its intensity variations are accompanied with spectral variations (see Fig.5),as said above we have made a PCA spectrum for each of the three days.These PCA spectra were ?tted simul-taneously with a single HEXTE spectrum aver-aged over the whole observation to increase the statistics at high energies.
We ?rst analyzed the October 28th spectrum.As there is a clear cuto?in the spectrum around 10keV,we have ?rst ?tted the continuum with a simple comptonization model (Compst ).This model alone is clearly rejected (χ2 977for 55d.o.f).This is mainly due to the presence of a broad feature again centered around 6.4keV.We have therefore added an iron K ?uorescence line (6.4keV)and ?tted its width and intensity.This improves the ?t signi?cantly (χ2=156.2for 53d.o.f).However,looking at the residuals indicates that the low energy part of the spectrum is not well accounted for.As in similar systems,a soft com-ponent (either a blackbody or a disk blackbody)is usually observed (e.g.White et al.1988),we have included such a component in the ?t.This again improves the ?t (χ2=48.6for 51d.o.f.,F m =56.5,P (F >F m )=1.2×10?13)for addition of the BB model.We have tried to ?t the “hard”compo-nent with a MCD model (Mitsuda et al.1984),instead of using Compst ,but the ?t is rejected,as the reduced χ2exceeds 4.On the other hand,the soft component could equally well be ?tted with a MCD.For both models,the equivalent radius of the BB and the projected inner disk radius of the MCD are small;R BB 5km and R in
√
KS1731-260
C+L+MCD C+L+BB C+L+MCD C+L+BB C+L+MCD C+L+BB
cosθor R BB 2.2+0.5
?0.33.8+0.5
?0.4
2.5+0.8
?0.5
3.8+1.2
?0.2
3.9+1.9
?1.5
4.7+4.4
?1.4
EqW154+29
?26136+32
?29
118+30
?25
106+32
?27
94+35
?26
86+37
?28
χ2(d.o.f)50.6(51)48.6(51)45.7(51)43.9(51)51.4(51)51.2(51) Table4:Best?t spectral results for KS1731-260for three segments of the observation(Oct28th,29th,30th). N H has been set to1.3×1022H atoms cm?2.The best?t model is the sum of a Compst model(C),a MCD or a BB,and a6.4keV gaussian line of?tted width(σFe).R in
√
proved the?t either.The non detection of such a
component might be simply related to the curvey shape of the spectrum(in particular the contin-
uum is not a CPL as assumed in Pexrav);re?ec-tion is easier to observe when the continuum is PL-like(Note that if there is indeed a re?ection
component;this will a?ect the?tted parameters for the continuum and hence the strength of the iron line).
The same model is also the best?t of the Oc-tober29th and30th spectra.In Table4,we list
the best?t parameters for the Compst+BB+Line and Compst+MCD+Line models.The data and folded model,the residuals and unfolded spectrum
KS1731-260for October28th are shown in Fig.11.
For KS1731-260,L1?20keV was～8.0×1037ergs
s?1at the beginning of the observation and reached～9.1×1037ergs s?1at the end(d=8.8 kpc).Such X-ray luminosity is slightly larger
than the time averaged ASM value.Throughout the observation,L20?200keV contributed just at
the level of～1%to the1-200keV luminosity of the source.
4.Discussion
We have here reported RXTE PCA+HEXTE timing and spectral observations of four type I
X-ray bursters(hence NSs).In the remainder,?rst we discuss their timing properties,then their
broad band spectral properties,and?nally address issues concerning possible di?erences between BHs and NSs.
4.1.Timing properties
1E1724–3045,SLX1735–269and GS1826-238 were in a so-called low state(LS)whereas KS1731-
260was in a high state(HS).The LS sources all have very similar PDS,characterized by High Fre-
quency Noise(HFN,or?at top noise,Van der Klis1994)with large RMS amplitudes(30%,2-40 keV).1E1724–3045and SLX1735–269are charac-
terized by the presence of a～1Hz QPO,and a break(atνBreak)at0.1-0.2Hz.For GS1826–238, depending on the modeling of the PDS,νBreak～0.02Hz or～0.2Hz,andνQPO～0.2or～1.0 Hz.KS1731-260has a very di?erent PDS,dis-
playing both Very Low Frequency Noise(VLFN) and a QPO at～7.6Hz.These properties are all consistent with those of“atoll”sources at vary-ing luminosities;the3LS sources would be in the “island”state,whereas KS1731-260would be in the“banana”state,according to the terminology de?ned by Hasinger&Van der Klis(1989).
4.1.1.Origin of the high frequency noise?
The high frequency noise is seen only in sources displaying signi?cant hard X-ray emission(all sources except KS1731-260),and hence must be related to the existence of a hot scattering cloud. This noise in the PDS is very similar in shape and normalization to that seen from LS BHs,strongly suggesting that the same physical mechanism is responsible for the spectrum and variability in these systems.Since the spectral shape strongly suggests thermal Compton up scattering,then the variability must have something to do with vary-ing either the seed photons,or optical depth,or the dissipation in the hot region.
The overall PDS shape can be described by a superposition of randomly occuring?ares(or “shots”),generally with some distribution of event durations.Such“shot noise”models can give a good description of the PDS(Lochner et al.1991 and references therein,Miyamoto et al.1992, Nowak et al.1999,Olive et al.1998),but are only a phenomenological,rather than a physical, description of the variability.To get a physi-cal description requires associating the“shots”to changes in the comptonizing cloud.
The simplest variability to envisage is a change of the soft photon input.Thermal comptonization involves multiple scattering of these seed photons on the hot electrons,so the spectrum at lower en-ergies responds?rst,followed after several scat-tering timescales by the higher energy spectrum. Such models predict that there should be time lags between the hard and soft energy bands,but that these lags(which are simply a measure of the size of the scattering cloud)should be constant irre-spective of whether the input variability was a short?are or a longer event.This directly con?icts with the observed lags in the BHCs(Miyamoto et al.1988)and the NSs(Ford et al1999),which clearly show longer lags for longer timescale vari-ability(Miyamoto et al.1991).
A satisfactory explanation of the long timescale lags is di?cult.If the variability arises from vary-ing the seed photons then the length of the lag
directly implies that the region is large.This led Kazanas,Hua and Titarchuk(1997)to develop the “Extended Atmosphere Comptonization Model”(see also Hua,Kazanas&Cui1999).This assumes a source of white noise at the center of the scatter-ing cloud which has a density pro?le n(r)∝1/r out to radii of a few light seconds.Their inhomo-geneous density distribution appears to match the observed time lags,and source spectra,but the physical situation is very hard to envisage.The majority of the gravitational potential energy is close to the compact object,so how can this power a hot corona whose size is many orders of magni-tude larger?Another weakness of that model,is that it produces di?erent PDS at di?erent ener-gies,contrary to what is observed(e.g.Nowak et al.,1999,Olive et al.1998).
An alternative model which can?t the PDS and time lags but with a small source size has been de-veloped by Poutanen&Fabian(1999).Here they associate the“shots”with magnetic?ares above a cold accretion disk,and use the spectral evolution of the?ares to produce the long lags.The?are begins with electrons being heated but the back-ground disk seed photon density is high,so the early time?are spectrum is soft.As the heating progresses,the energy dissipated in the?are dom-inates that of the disk,so the spectrum becomes hard.The drawback with such models is that they rely on details of the magnetic dissipation,which are not well known.Nonetheless,it is encouraging that at least under some circumstances the lags and PDS can be matched by the envisaged small source.
4.1.2.Origin of low frequency QPOs?
The origin of the～1Hz QPO(atνQPO)in the LS sources,once suspected to be a BH sig-nature,is also unknown,and is not accounted for by the above models.Chen&Taam(1994)have proposed that they might arise from disk lumi-nosity oscillations resulting from thermal viscous instabilities developing within the inner disk re-gion.Vikhlinin et al.(1994)have also suggested that they could result from a weak interaction be-tween the“shots”when the instability is triggered in a region of stable energy supply.To keep the energy released constant on large timescales,the appearance of a strong shot should a?ect the am-plitude and the probability of occurence of sub-sequent“shots”.Recently,developing a model speci?c for NSs(i.e.not accounting for the sim-ilarities between BHs and NSs),Titarchuk&Os-herovich(1999)have associated the QPOs with radial oscillations in a boundary layer.
It has been shown that in many sources there exists a strong correlation betweenνBreak and νQPO(Wijnands&Van der Klis1999b).Our three LS sources follow this correlation(Olive& Barret1999).This correlation suggests that the longest?uctuations(scaling asνBreak?1)and the QPOs are related to the same unknown physical mechanism or are produced in regions interacting with each other.Since the same correlation is ob-served among BHs and Z-sources,this mechanism cannot be related to the presence of a hard sur-face or even a small magnetosphere.In addition, in several systems(e.g.4U1608-52,Yoshida et al.1993,see however above for SLX1735–269),a strict correlation betweenνBreak and the inferred mass accretion rate has been observed.For BHs, Esin et al.(1998)have speculated that the rapid variability that is associated with the hard X-ray emission could originate from an ADAF,with the variability timescales determined by the viscous or dynamical timescales in the ADAF.They have further proposed that the1-10Hz QPOs could re-sult from the interactions at the boundary(the so-called transition radius in ADAF terminology) between the ADAF and the outer cool accretion disk,in which case theνQPO would be some mul-tiples of the Keplerian frequency at the transition radius.Hence,a shrinking of the inner disk ra-dius(possibly due to an increase in the mass ac-cretion rate through the disk)should result in an increase of the QPO frequency.The correlation between the position of the inner disk radius in-ferred from the kilo-Hz QPO frequency andνQPO, observed for instance in4U1728-34(Ford et al. 1998),is consistent with the above picture.Fur-thermore,ifνBreak is somehow related to the size of the ADAF,the observed correlation between νBreak andνQPO would imply that the ADAF con-tracts when the accretion rate increases(possibly due to an increase of cooling?ux from the disk).
4.1.3.Spectral states and high frequency QPOs
Finally,in the three sources displaying high frequency noise and a hard X-ray tail,no HFQ-POs(above300Hz)are detected(e.g1E1724–
3045,Barret et al.1999b).More generally,it seems that no HFQPOs are seen whenνBreak is low( 6?7Hz; e.g.4U1705-44,Ford et al. 1998).The lack of HFQPOs might therefore be related to the presence of a hot scattering corona. Smearing of the HFQPO signal in such a corona is a possible mechanism,which is however invoked at higher accretion rates(i.e.for larger optical depthτ 5,cooler corona,Brainerd&Lamb, 1987).This might indeed explain the disappear-ance of the HFQPOs after it has reached satura-tion(i.e.when the inner disk is at the last stable orbit,e.g.4U1820-30,Bloser et al.1999),or more generally when the source enters the upper branch of the“banana”state(Miller et al.1998).The anticorrelation between the QPO amplitude and the QPO frequency observed in KS1731-260(see above)is consistent with this picture(note that there is a weak indication that the optical depth derived from the spectral?tting increases when the source luminosity increases,see Table4).For the LS sources,from the spectral analysis,we have derived an optical depth of a few for the Comp-tonizing cloud.Could that cloud,through smear-ing,account for the lack of HFQPOs?Using our upper limit,one can set a lower limit on the size of the scattering region.If the radiation is scattered, the RMS amplitude at in?nity(A∞)of a luminos-ity oscillation with frequencyνand amplitude A0 at the center of a spherical region of radius R C and optical depthτis A∞≈(23
3πνR Cτ/c(νis the frequency of the os-cillation,Kyla?s&Phinney,1989;Miller et al., 1998).For a beaming oscillation,the attenuation is stronger(due to the fact that the scattering pro-cess tends to isotropize the photon distribution), and a factor2/(1+τ)has to be multiplied to the?rst term of the previous equation(Kyla?s& Phinney,1989).Let us?rst considerν=300Hz, from the previous expression,taking A∞=2.5% (our upper limit),assuming R C=200km,one can set an upper limit～4and～6%on A0for a luminosity and beaming oscillations,respectively. For a signal at1000Hz,these upper limits become 12%and20%respectively.Alternatively,if we as-sume A0=10%,for the signal to be attenuated down to our upper limit,the size of the corona has to be larger than100km and130km for a beaming and luminosity1000Hz oscillation.Clearly,this indicates that with the low optical depth derived for the comptonizing cloud,and for any plausible sizes of such a cloud,the attenuation is not very strong,and if signal there is,it has to be intrinsi-cally weak.
Another possibility which may explain the lack of HFQPOs in LS sources might be related to the position of the inner disk radius.Obviously,if the HFQPOs are generated in the disk,and if the inner edge of the disk lies at large radius(see be-low),no signals at such high frequencies should be seen.In that respect,it is worth pointing out that the bump(around10Hz)in the PDS of1E1724–3045(Fig.6)was interpreted by Psaltis et al.(1999)as a low frequency kilo-Hz QPO(the same bump is seen in GS1826–238and SLX1735–269).Their conclusion was derived from the fact that the0.8Hz QPO and the bump at10Hz?tted in the global correlation observed over several fre-quency decades betweenνQPO and the frequency of the lower kilo-Hz QPOs(Psaltis et al.1999).
4.2.Spectral properties
To?rst order,the broad band LS spectra of NSs can be approximated with a PL followed by an exponential cuto?at50-80keV.There is also a weak soft component in2of the3LS sources, with temperature of～0.5?1keV contributing less than～30%of the total luminosity.The lat-ter quantities,when derived from the?t of PCA data are subject to uncertainties because the?t starts at2.5keV,because they depend on the N H values assumed,and because there are some calibrations uncertainties of the PCA at low en-ergies(E 5keV).However,as for example in the case of1E1724–3045,the values derived from our spectral?tting are consistent with the ones obtained from observations that are more sensi-tive to such a soft component(e.g.by Beppo-SAX or ASCA,Guainazzi et al.1998,Barret et al.,1999a).Thermal Comptonization of these soft photons by hot electrons then appears to be the most plausible emission mechanism for these sys-tems,and for a spherical scattering cloud,the de-rived optical depths and electron temperatures are in the ranges of2 τ 4and15 kT E 30keV. Thermal Comptonization also dominates the spec-tral formation in KS1731-260,but with a signi?-cantly lower kT E(～3keV)and largerτ(～10). The main di?erence between a source in the HS and a source in the LS is best illustrated in Fig.
第18章X射线光电子能谱分析 18.1 引言 固体表面分析业已发展为一种常用的仪器分析方法，特别是对于固体材料的分析和元素化学价态分析。目前常用的表面成分分析方法有：X射线光电子能谱（XPS）, 俄歇电子能谱(AES)，静态二次离子质谱(SIMS)和离子散射谱(ISS)。AES 分析主要应用于物理方面的固体材料科学的研究，而XPS的应用面则广泛得多，更适合于化学领域的研究。SIMS和ISS由于定量效果较差，在常规表面分析中的应用相对较少。但近年随着飞行时间质谱（TOF-SIMS）的发展，使得质谱在表面分析上的应用也逐渐增加。本章主要介绍X射线光电子能谱的实验方法。 X射线光电子能谱（XPS）也被称作化学分析用电子能谱（ESCA）。该方法是在六十年代由瑞典科学家Kai Siegbahn教授发展起来的。由于在光电子能谱的理论和技术上的重大贡献，1981年，Kai Siegbahn获得了诺贝尔物理奖。三十多年的来，X射线光电子能谱无论在理论上和实验技术上都已获得了长足的发展。XPS已从刚开始主要用来对化学元素的定性分析，业已发展为表面元素定性、半定量分析及元素化学价态分析的重要手段。XPS的研究领域也不再局限于传统的化学分析，而扩展到现代迅猛发展的材料学科。目前该分析方法在日常表面分析工作中的份额约50%，是一种最主要的表面分析工具。 在XPS谱仪技术发展方面也取得了巨大的进展。在X射线源上，已从原来的激发能固定的射线源发展到利用同步辐射获得X射线能量单色化并连续可调的激发源；传统的固定式X射线源也发展到电子束扫描金属靶所产生的可扫描式X射线源；X射线的束斑直径也实现了微型化，最小的束斑直径已能达到6μm大小, 使得XPS在微区分析上的应用得到了大幅度的加强。图像XPS技术的发展，大大促进了XPS在新材料研究上的应用。在谱仪的能量分析检测器方面，也从传统的单通道电子倍增器检测器发展到位置灵敏检测器和多通道检测器，使得检测灵敏度获得了大幅度的提高。计算机系统的广泛采用，使得采样速度和谱图的解析能力也有了很大的提高。 由于XPS具有很高的表面灵敏度，适合于有关涉及到表面元素定性和定量分析方面的应用，同样也可以应用于元素化学价态的研究。此外，配合离子束剥离技术和变角XPS技术，还可以进行薄膜材料的深度分析和界面分析。因此，XPS 方法可广泛应用于化学化工，材料，机械，电子材料等领域。 18.2 方法原理 X射线光电子能谱基于光电离作用，当一束光子辐照到样品表面时，光子可以被样品中某一元素的原子轨道上的电子所吸收，使得该电子脱离原子核的束缚，以一定的动能从原子内部发射出来，变成自由的光电子，而原子本身则变成一个激发态的离子。在光电离过程中，固体物质的结合能可以用下面的方程表示： E k = hν- E b - φs （18.1）
从劳厄发现晶体X射线衍射谈起 摘要：文章从劳厄发现晶体X射线衍射的前因后果谈起。劳厄的这个发现产生了两个新学科，即X射线谱学和X射线晶体学。文中还回顾了布拉格父子对这两个新学科所作的重大贡献，并阐述了X射线晶体学的深远影响。 今年是劳厄（von Lane M）发现晶体X射线衍射九秩之年。 从1895年伦琴（R0ntgen W C）发现X射线到1926年薛定愕（Schrodinger）奠定量子力学基础的30多年是现代物理学诞生和成长的重要时期。在此期间的众多重大发现中，1912年劳厄的发现发挥了极为及时而又十分深远的影响，是很值得我们通过回顾和展望来纪念它的。 我们先来了解一下劳厄发现的前因后果。1912年劳厄发现晶体X射线衍射时是在德国慕尼黑大学理论物理学教授索未菲（Sommerfeld）手下执教。除理论物理教授索未菲外，在这个大学中还有发现X射线的物理学教授伦琴和著名的晶体学家格罗特（Groth）。当时，劳厄对光的干涉作用特别感兴趣，索末菲则在考虑X射线的本质和产生的机制问题，而格罗特是晶体学权威之一，并著书Chemische KristallograPhic （化学晶体学）数卷。身在这样的学府中，劳厄当时通过耳闻目睹也就对 晶体中原子是按三维点阵排布以及X射线可能是波长很短的电磁波这样的想法不会感到陌生或难于接受了。而且看来正当而立之年的他是很想在光的干涉作用上做点文章的。真可谓机遇不负有心人了。这时，索末菲的博士生埃瓦尔德（Ewald P P）来请教劳厄，谈到他正在研究关于光波通过晶体中按三维点阵排布的原子会产生什么效应。这对劳厄有所触发并想到：如果波长短得比晶体中原子间距离更短时又当怎样？而X射线可能正是这样的射线。他意识到，说不定晶体正是能衍射X射线的三维光栅呢。现在劳厄需要考虑的大事是做实验来证实这个想法。当时索末菲正好有个助教弗里德里希（Friedrich W） ，他曾从伦琴教授那里取得博士学位。 他主动要去进行这样的实验。经过几次失败后，他终于取得了晶体的第一个衍射图「（见图1）」。晶体是五水合硫酸铜（CuSO4·5H2O）。 劳厄的发现经过进一步的工作很快取得了一箭双雕的效果：既明确了X射线的本质，测定了波长，开创了X射线谱学，又使测定晶体结构的前景在望，从而将观察晶体外形所得结论经过三维点阵理论发展到230个空间群理论的晶体学，提升为X射线晶体学。这个发现产生的两个新学科，几乎立即给出了一系列在科学中有重大影响的结果。英国的布拉格父子（Bragg W H和Bragg W L）在奠定这两个新学科的基础中起了非常卓越的作用。他们使工作的重心从德国转到英国。将三个劳厄方程（衍射条件）压缩成一个布拉格方程（定律）的小布拉格曾把重心转移的原因归之于老布拉格设计的用起来得心应手的电离分光计”。既然晶体是X射线的衍射光栅，那么，为了测定X射线的波长，光栅的间距当如何得出？1897年巴洛（Barlow W）预测过最简单的晶体结构型式，其中有氯化钠所属的型式。根据当时已知的NaCI的化学式量（58.46）和阿伏伽德罗常数（6.064×1023）以及晶体密度（2.163g/cm2），可以推算出氯化钠晶体（10）原子面的间距d＝2.814×10－8cm。 布拉格父子的工作是有些分工的：老布拉格用他的电离分光计侧重搞谱学，很快发现X射线谱中含有连续谱和波长取决于对阴极材料的特征谱线。此后，测定晶体结构主要依靠特征射线。同时还观察到同一跃迁系特征射线的频率是随对阴极材料在元素周期系中的排序递增的，这种频率的排序给出了原子序数。这是对化学中总结出来的元素周期律作出的呼应。小布拉格的工作是沿着X射线晶体学的方向发展的。他一生中从氯化钠和金刚石一直测到蛋白质的晶体结构。从1913年起，他在两年中一连测定了氯化钠、金刚石、硫化锌、黄铁矿、荧石和方解石等的晶体结构。这一批最早测定的晶体结构虽然极为简单，但很有代表性，而且都足以让化学和矿物学界观感一新。同时为测定参数较多和结构比较复杂的晶体结构也进行了理论和技术方面的准备。X射线晶体学能不断采用新技术和解决周相问题的新方法，使结构测定的对象
锂硫电池的研究现状 近年来，随着不可再生资源的逐渐减少，清洁能源的利用逐渐得到重视，而电池作为储能装置也受到越来越多的考验。锂硫电池与传统的锂离子电池相比，优势主要在于硫的高比容量，单质硫的理论比容量为1600mAh/g ，理论比能量2600Wh/kg。并且硫是一种廉价且无毒的原材料。而与此同时，硫作为锂电池的正极材料也存在着诸多问题[1]： 1、单质硫以及最终放电产物都是绝缘的，如果与正极中掺入的导电物质结合不好，就会导致活性物质不能参与反应而失效； 2、单质硫在反应过程中会生成长链的聚硫化物离子S n2-，这种离子容易溶解在电解液中，并与锂负极反应，产生“穿梭效应”，引起自放电并使库伦效率降低； 3、在每次放电过程结束之后，都会有一些Li2S2/Li2S沉淀在正极上，并且这些不溶物随着循环次数的增加，在正极表面发生团聚，并且正极结构也会发生变化，导致这部分活性物质不能参与电化学反应而失效，并且使电池的内阻增加； 4、硫正极随充放电的进行会产生约22%的体积变化，从而导致电池物理结构破坏而失效。 针对硫作为正极材料的种种弊端，研究者们分别采用了多种方法予以解决，其中将硫与碳材料复合的研究较多。针对几种典型方法，分别举例介绍如下：一、石墨烯-硫复合材料 Wang等人采用石墨烯包覆硫颗粒的方法制作复合材料电极[2]。如图1所示，他们首先采用化学方法制备了硫单质，并利用一种特殊的表面活性剂Triton X-100在硫颗粒的表面修饰了一些PEG高分子，然后再用导电炭黑和石墨烯的分散液对硫颗粒进行包覆。这种方法的优点在于：首先，石墨烯和导电炭黑具有优异的导电性能，可以克服硫以及硫反应产物绝缘的问题；第二，导电炭黑、石墨烯和PEG高分子对硫颗粒进行了包覆，可以解决硫在电解液中溶出的问题；第三，PEG高分子具有一定的弹性，可以在一定程度上缓解体积变化带来的影响。 二、碳纳米管-硫复合材料 Zheng等人用AAO做模板制备了碳纳米管阵列[3]，随后将硫加热使其浸入到碳纳米管中间，然后将AAO模板去掉，得到碳纳米管-硫复合材料，如图2所示。这种方法的优点在于碳纳米管的比表面积大，有利于硫化锂的沉积。并且长径比较大，可以较好地将硫限制在管内，防止其溶解在电解液中。碳纳米管的导电性好管壁又很薄，有利于离子导通和电子传输。同时，因为制备过程中先沉积硫，后去除模板，这样有利于使硫沉积到碳管内，减少硫在管外的残留，从而防止这部分硫的溶解。
X射线光电子能谱分析 1 引言 固体表面分析业已发展为一种常用的仪器分析方法，特别是对于固体材料的分析和元素化学价态分析。目前常用的表面成分分析方法有：X射线光电子能谱（XPS）, 俄歇电子能谱(AES)，静态二次离子质谱(SIMS)和离子散射谱(ISS)。AES分析主要应用于物理方面的固体材料科学的研究，而XPS的应用面则广泛得多，更适合于化学领域的研究。SIMS和ISS由于定量效果较差，在常规表面分析中的应用相对较少。但近年随着飞行时间质谱（TOF-SIMS）的发展，使得质谱在表面分析上的应用也逐渐增加。本章主要介绍X射线光电子能谱的实验方法。 X射线光电子能谱（XPS）也被称作化学分析用电子能谱（ESCA）。该方法是在六十年代由瑞典科学家Kai Siegbahn教授发展起来的。由于在光电子能谱的理论和技术上的重大贡献，1981年，Kai Siegbahn获得了诺贝尔物理奖。三十多年的来，X射线光电子能谱无论在理论上和实验技术上都已获得了长足的发展。XPS已从刚开始主要用来对化学元素的定性分析，业已发展为表面元素定性、半定量分析及元素化学价态分析的重要手段。XPS的研究领域也不再局限于传统的化学分析，而扩展到现代迅猛发展的材料学科。目前该分析方法在日常表面分析工作中的份额约50%，是一种最主要的表面分析工具。 在XPS谱仪技术发展方面也取得了巨大的进展。在X射线源上，已从原来的激发能固定的射线源发展到利用同步辐射获得X射线能量单色化并连续可调的激发源；传统的固定式X射线源也发展到电子束扫描金属靶所产生的可扫描式X射线源；X射线的束斑直径也实现了微型化，最小的束斑直径已能达到6 m 大小, 使得XPS在微区分析上的应用得到了大幅度的加强。图像XPS技术的发展，大大促进了XPS在新材料研究上的应用。在谱仪的能量分析检测器方面，也从传统的单通道电子倍增器检测器发展到位置灵敏检测器和多通道检测器，使得检测灵敏度获得了大幅度的提高。计算机系统的广泛采用，使得采样速度和谱图的解析能力也有了很大的提高。 由于XPS具有很高的表面灵敏度，适合于有关涉及到表面元素定性和定量分析方面的应用，同样也可以应用于元素化学价态的研究。此外，配合离子束剥离技术和变角XPS技术，还可以进行薄膜材料的深度分析和界面分析。因此，XPS方法可广泛应用于化学化工，材料，机械，电子材料等领域。 2 方法原理 X射线光电子能谱基于光电离作用，当一束光子辐照到样品表面时，光子可以被样品中某一元素的原子轨道上的电子所吸收，使得该电子脱离原子核的束缚，以一定的动能从原子内部发射出来，变成自由的光电子，而原子本身则变成
第十四章 X-射线光电子能谱法 14．1 引言 X-射线光电子谱仪（X-ray Photoelectron Spectroscopy，简称为XPS），经常又被称为化学分析用电子谱（Electron Spectroscopy for Chemical Analysis，简称为ESCA），是一种最主要的表面分析工具。自19世纪60年代第一台商品化的仪器开始，已经成为许多材料实验室的必不可少的成熟的表征工具。XPS发展到今天，除了常规XPS外，还出现了包含有Mono XPS (Monochromated XPS, 单色化XPS，X射线源已从原来的激发能固定的射线源发展到利用同步辐射获得X射线能量单色化并连续可调的激发源), SAXPS ( Small Area XPS or Selected Area XPS, 小面积或选区XPS，X射线的束斑直径微型化到6μm) 和iXPS（imaging XPS, 成像XPS）的现代XPS。目前，世界首台能量分辨率优于1毫电子伏特的超高分辨光电子能谱仪（通常能量分辨率低于1毫电子伏特）在中日科学家的共同努力下已经研制成功，可以观察到化合物的超导电子态。现代XPS拓展了XPS的内容和应用。 XPS是当代谱学领域中最活跃的分支之一，它除了可以根据测得的电子结合能确定样品的化学成份外，XPS最重要的应用在于确定元素的化合状态。XPS可以分析导体、半导体甚至绝缘体表面的价态，这也是XPS的一大特色，是区别于其它表面分析方法的主要特点。此外，配合离子束剥离技术和变角XPS技术，还可以进行薄膜材料的深度分析和界面分析。XPS表面分析的优点和特点可以总结如下： ⑴固体样品用量小，不需要进行样品前处理，从而避免引入或丢失元素所造成的错误分析 ⑵表面灵敏度高，一般信息采样深度小于10nm ⑶分析速度快，可多元素同时测定 ⑷可以给出原子序数3-92的元素信息，以获得元素成分分析 ⑸可以给出元素化学态信息，进而可以分析出元素的化学态或官能团 ⑹样品不受导体、半导体、绝缘体的限制等 ⑺是非破坏性分析方法。结合离子溅射，可作深度剖析 目前，XPS主要用于金属、无机材料、催化剂、聚合物、涂层材料、纳米材料、矿石等各种材料的研究，以及腐蚀、摩擦、润滑、粘接、催化、包覆、氧化等过程的研究，也可以用于机械零件及电子元器件的失效分析，材料表面污染物分析等。 14.2 基本原理 XPS方法的理论基础是爱因斯坦光电定律。用一束具有一定能量的X射线照射固体样品，入射光子与样品相互作用，光子被吸收而将其能量转移给原子的某一壳层上被束缚的电子，此时电子把所得能量的一部分用来克服结合能和功函数，余下的能量作为它的动能而发射出来，成为光电子，这个过程就是光电效应。 该过程可用下式表示: hγ=E k+E b+E r（14.1） 式中: hγ：X光子的能量（h为普朗克常数，γ为光的频率）；
全固态锂电池技术的研究进展与展望 周俊飞 （衢州学院化学与材料工程学院浙江衢州324000） 摘要：现有电化学储能锂离子电池系统采用液体电解质，易泄露、易腐蚀、服役寿命短，具有安全隐患。薄膜型 全固态锂电池、大容量聚合物全固态锂电池和大容量无机全固态锂电池是一类以非可燃性固体电解质取代传统锂离 子电池中液态电解质，锂离子通过在正负极间嵌入-脱出并与电子发生电荷交换后实现电能与化学能转换的新型高 安全性锂二次电池。作者综述了各种全固态锂电池的研究和开发现状，包括固态锂电池的构造、工作原理和性能特 征，锂离子固体电解质材料与电极/电解质界面调控，固态整电池技术等方面，提出并详细分析了该技术面临的主要 科学与技术问题，最后指出了全固态锂电池技术未来的发展趋势。 关键词：储能；全固态锂离子电池；固体电解质；界面调控 1 全固态锂电池概述 全固态锂二次电池，简称为全固态锂电池，即电池各单元，包括正负极、电解质全部采用固态材料的锂二次电池，是从20 世纪50 年代开始发展起来的[10-12]。全固态锂电池在构造上比传统锂离子电池要简单，固体电解质除了传导锂离子，也充当了隔膜的角色，如图 2 所示，所以，在全固态锂电池中，电解液、电解质盐、隔膜与黏接剂聚偏氟乙烯等都不需要使用，大大简化了电池的构建步骤。全固态锂电池的工作原理与液态电解质锂离子电池的原理是相通的，充电时正极中的锂离子从活性物质的晶格中脱嵌，通过固体电解质向负极迁移，电子通过外电路向负极迁移，两者在负极处复合成锂原子、合金化或嵌入到负极材料中。放电过程与充电过程恰好相反，此时电子通过外电路驱动电子器件。目前，对于全固态锂二次电池的研究，按电解区分主要包括两大类[13]：一类是以有机聚合物电解质组成的锂离子电池，也称为聚合物全固态锂电池；另一类是以无机固体电解质组成的锂离子电池，又称为无机全固态锂电池，其比较见表1。通过表1 的比较可以清楚地看到，聚合物全固态锂电池的优点是安全性高、能够制备成各种形状、通过卷对卷的方式制备相对容易，但是，该类电池作为大容量化学电源进入储能领域仍有一段距离，主要存在的问题包括电解质和电极的界面不稳定、高分子固体电解质容易结晶、适用温度范围窄以及力学性能有提升空间；以上问题将导致大容量电池在使用过程中因为局部温度升高、界面处化学反应面使聚合物电解质开貌发生变化，进而增大界面电阻甚至导致断路。同时，具有隔膜作用的电解质层的力学性能的下降将引起电池内部发生短路，从面使电池失效[14-15]。无机固体电解质材料具有机械强度高，不含易燃、易挥发成分，不存在漏夜，抗温度性能好等特点；同时，无机材料处理容易实现大规模制备以满足大尺寸电池的需要，还可以制备成薄膜，易于将锂电池小型化，而且由无机材料组装的薄膜无机固体电解质锂电池具有超长的储存寿命和循环性能，是各类微型电子产品电源的最佳选择[10]。采用有机电解液的传统锂离子电池，因过度充电、内部短路等异常时电解液发热，有自燃甚至爆炸的危险（图3）。从图 3 可以清楚地看到，当电池因为受热或短路情况下导致温度升高后，传统的锰酸锂或钴酸锂液体电解质锂离子电池存在膨胀起火的危险，而基于纯无机材料的全固态锂电池未发生此类事故。这体现了无机全固态锂电池在安全性方面的独特优势。以固体电解质替代有机液体电解液的全固态锂电池，在解决传统锂离子电池能量密度偏低和使用寿命偏短这两个关键问题的同时，有望彻底解决电池的安全性问题，符合未来大容量新型化学储能技术发展的方向。正是被全固态锂电池作为电源所表现出来的优点所吸引，近年来国际上对全固态锂电池的开发和研究逐渐开始活跃[10-12] 2 全固态锂电池储能应用研究进展 在社会发展需求和潜在市场需求的推动下，基于新概念、新材料和新技术的化学储能新体系不断涌现，化学储能技术正向安全可靠、长寿命、大规模、低成本、无污染的方向发展。目前已开发的化学储能装置，包括各种二次电池（如镍氢电池、锂离子电池等）、超级电容器、可再生燃料电池（RFC：电解水制氢-储氢-燃料电池发电）、钠硫电池、液流储能电池等。综合各种因素，考虑用于大规模化学储能的主要是锂二次电池、钠硫电池及液流电池，而其中大容量储能用锂二次电池更具推广前景。。 全固态锂电池、锂硫电池、锂空气电池或锂金属电池等后锂离子充电电池的先导性研究在世界各地积极地进行着，计划在2020 年前后开始商业推广。在众多后锂离子充电电池中，包括日本丰田汽车、韩国三星电子和德国KOLIBRI 电池公司对全固态锂电池都表现出特别的兴趣。图 4 为未来二十年大容量锂电池的发展路径，从图 4 可以看出，全固态电
X射线荧光光谱分析 X射线是一种电磁辐射，其波长介于紫外线和γ射线之间。它的波长没有一个严格的界限，一般来说是指波长为0.001-50nm的电磁辐射。对分析化学家来说，最感兴趣的波段是0.01-24nm，0.01nm左右是超铀元素的K系谱线，24nm则是最轻元素Li的K系谱线。1923年赫维西(Hevesy, G. Von)提出了应用X射线荧光光谱进行定量分析，但由于受到当时探测技术水平的限制，该法并未得到实际应用，直到20世纪40年代后期，随着X射线管、分光技术和半导体探测器技术的改进，X荧光分析才开始进入蓬勃发展的时期，成为一种极为重要的分析手段。 1.1 X射线荧光光谱分析的基本原理 当能量高于原子内层电子结合能的高能X射线与原子发生碰撞时，驱逐一个内层电子而出现一个空穴，使整个原子体系处于不稳定的激发态，激发态原子寿命约为10-12-10-14s，然后自发地由能量高的状态跃迁到能量低的状态。这个过程称为驰豫过程。驰豫过程既可以是非辐射跃迁，也可以是辐射跃迁。当较外层的电子跃迁到空穴时，所释放的能量随即在原子内部被吸收而逐出较外层的另一个次级光电子，此称为俄歇效应，亦称次级光电效应或无辐射效应，所逐出的次级光电子称为俄歇电子。它的能量是特征的，与入射辐射的能量无关。当较外层的电子跃入内层空穴所释放的能量不在原子内被吸收，而是以辐射形式放出，便产生X射线荧光，其能量等于两能级之间的能量差。因此，X射线荧光的能量或波长是特征性的，与元素有一一对应的关系。图1-1给出了X射线荧光和俄歇电子产生过程示意图。
K层电子被逐出后,其空穴可以被外层中任一电子所填充，从而可产生一系列的谱线，称为K系谱线：由L层跃迁到K层辐射的X射线叫Kα射线,由M层跃迁到K层辐射的X射线叫Kβ射线……。同样,L层电子被逐出可以产生L系辐射(见图1-2)。
全固态3D薄膜锂离子电池的研究进展 作者：邓亚锋钱怡崔艳华刘效疆来源：本站浏览数：289 发布时间：2013-8-8 16:28:16 0 引言 全固态薄膜锂离子电池主要由正/负极薄膜、电解质和集流器薄膜组成．整个电池厚约10 μm，可设计成任意形状和大小集成在IC电路中，是便携式电子设备、微电子机械系统(MEMS)以及微型国防技术装备(如微型智能武器)的理想能源。全固态平面薄膜电池(图1)受限于几何结构，能量和功率密度难以满足快速发展的MEMS、微型医疗器械、无线通信、传感器等领域对微电源的要求。全固态三维薄膜锂离子电池(简称3D锂电池)通过独特的构架设计(图2)，增大单位立足面积内电极活性物质负载量，并缩短锂离子扩散半径，提高了电池的容量和充放电速率。是解决未来微电子器件能量需求的一种有效方式，引起了人们的极大关注。 1 不同构架的全固态3D薄膜锂电池 1．1 叉指碳柱3D电池 叉指碳柱3D电池由加利福尼亚大学Wang小组于2004年首次提出(图3)，在Si/SiO2衬底上涂覆感光胶，光刻得到图形，再经过高温热解及后处理，即制得正/负极叉指状碳柱3D电池。叉指碳柱既可以直接作为电极，又可以作为集流器，在其表面沉积各种电化学活性物质。2008年，Min等研究了在叉指碳柱上电镀十二烷基苯磺酸盐掺杂聚吡咯(PPYDBS)导电聚合物薄膜的方法。结果表明，覆盖约10 μm厚PPYDBS的叉指阴极(C-PPYDBS)，电极电位从碳电极的3．2 V提高到了3．7 V(相对于Li/Li+)，但自放电较为严重，电池的放电容量远小于充电电容。 为改善叉指碳柱电极性能，Teixidor等制备出包覆中间相碳微球的叉指碳柱(C-MCMB)，有效提高了电极不可逆容量，但可逆容量仍较低。Chen等在叉指碳柱上包覆碳纳米管(CNT／C-MEMS)使单位立足面积电容达到8．3 F/cm2，充放电循环性能得到显著提高。 叉指碳柱电极成本低、热力学和化学稳定性好、易制成各种形貌、能包覆不同的活性材料(图4)，光刻-热解工艺较为成熟，适合工业化生产。但是，叉指结构放电不均匀、漏电流较大、碳柱在锂离子嵌入和脱出过程中易变形破损，这些问题需进一步研究解决。 1．2 微通道衬底3D电池 1998年，以色列特拉维夫大学的Peled小组首次报道了微通道衬底3D 电池(3D-MCP)；在Si片或玻璃上蚀刻出均匀分布、直径为15～50 μm的微通
论 著8 全固态薄膜锂离子二次电池的研究进展 耿利群任岳*朱仁江陈涛 （重庆师范大学物理与电子工程学院，重庆 400047） 摘 要：本文综述了全固态薄膜锂离子二次电池的研究进展，主要阐述了薄膜锂电池的结构设计以及正极、负极和固体电解质材料研究现状，并对其今后的发展趋势及研发热点进行了展望。 关键词：全固态薄膜锂离子二次电池；固体电解质；电池结构 DOI：10.3969/j.issn.1671-6396.2013.01.004 1 引言 随着电子信息工业和微型加工技术快速发展，对其所需的微型能源则提出了特殊微型化的要求。其中全固态薄膜锂离子二次电池因其高的能量密度、强的安全性、长的循环寿命、宽的工作电压和重量轻等优点，成为微电池系统需求的最佳选择[1]。本文主要介绍了全固态薄膜锂离子二次电池的关键性薄膜材料及电池结构的研究现状，并对其的开发应用及研究前景作了分析。 2 全固态薄膜锂离子二次电池结构的研究 薄膜电池结构的设计，对整个电池性能将产生直接的影响；同样对提高电池的能量密度、循环寿命和锂离子的传输速率也起到至关重要的作用。所以优化薄膜电池结构的设计，则是对构造高性能薄膜锂离子电池做到了强有力的支撑。 1993年美国橡树岭国家实验室（ORNL）Bates等[2]研制出了一种经典的薄膜锂离子电池叠层结构（见图1）。在衬底上先沉积两层阴阳极电流收集极薄膜，而后依次沉积阴极、固体电解质和阳极薄膜，最后在薄膜电池外表面上涂一层保护层，以此来防止阳极上金属锂和空气中的一些物质发生化学反应。 图1 薄膜锂离子电池结构剖面示意图 Baba等[3]研发出另一种典型的薄膜锂离子电池结构（见图2）。其较图1薄膜锂电池结构设计更为简单，制作更为容易。在不锈钢衬底上依次沉积各层薄膜电池材料，而在图示中有两个引线端子则是为了便于薄膜电池的连接使用。这种结构设计很好地提高了整个电池的有效面积，进而也极大地改善了薄膜电池的性能。 Nakazawa等[4]利用直流溅射和射频溅射的方法，研制出一种“直立型”全固态薄膜锂离子电池结构（见图3）。该研究小组利用该薄膜电池结构设计，成功制备出有效面积更大的全固态薄膜锂离子电池，这样也使得薄膜电池的能量密度和循环寿命等电化学性能得到大幅度提升。 图2 全固态薄膜锂离子电池结构剖面示意图 图3 “直立型”全固态薄膜锂离子电池剖面示意图 Hart等[5]设计了柱状电极交替排列的微型锂电池结构（见图4）。并对几种不同的正极、负极排列方式进行了相关的研究计算，得出了此薄膜电池的结构能够大大提升薄膜电池本身的能量密度。然而Eftekhari[6]则研制出了一种3-D微型锂电池结构的LiMn2O4电极，与以往微型锂电池结构的LiMn2O4电极在电池容量方面得到了提升。 图4 3-D微电池柱状结构示意图 [正极（灰色） 、负极（白色）交替排列分布]
晶体X射线衍射实验报告全解
中南大学 X射线衍射实验报告 材料科学与工程学院材料学专业1305班班级 姓名学号0603130500 同组者无 黄继武实验日期2015 年12 月05 日指导教 师 评分分评阅人评阅日 期 一、实验目的 1)掌握X射线衍射仪的工作原理、操作方法； 2)掌握X射线衍射实验的样品制备方法； 3)学会X射线衍射实验方法、实验参数设置，独立完成一个衍射实验测试； 4)学会MDI Jade 6的基本操作方法； 5)学会物相定性分析的原理和利用Jade进行物相鉴定的方法； 6)学会物相定量分析的原理和利用Jade进行物相定量的方法。 本实验由衍射仪操作、物相定性分析、物相定量分析三个独立的实验组成，实验报告包含以上三个实验内容。 二、实验原理
1 衍射仪的工作原理 特征X射线是一种波长很短(约为20～0.06nm)的电磁波，能穿透一定厚度的物质，并能使荧光物质发光、照相乳胶感光、气体电离。在用电子束轰击金属“靶”产生的X射线中，包含与靶中各种元素对应的具有特定波长的X射线，称为特征（或标识）X射线。考虑到X射线的波长和晶体内部原子间的距离相近，1912年德国物理学家劳厄(M.von Laue)提出一个重要的科学预见：晶体可以作为X射线的空间衍射光，即当一束X射线通过晶体时将发生衍射，衍射波叠加的结果使射线的强度在某些方向上加强，在其他方向上减弱。分析在照相底片上得到的衍射花样，便可确定晶体结构。这一预见随即为实验所验证。1913年英国物理学家布拉格父子(W. H. Bragg, W. L Bragg)在劳厄发现的基础上，不仅成功地测定了NaCl、KCl等的晶体结构，并提出了作为晶体衍射基础的著名公式──布拉格定律： 2dsinθ=nλ 式中λ为X射线的波长，n为任何正整数。当X射线以掠角θ（入射角的余角，又称为布拉格角）入射到某一点阵晶格间距为d的晶面面上时，在符合上式的条件下，将在反射方向上得到因叠加而加强的衍射线。 2 物相定性分析原理 1) 每一物相具有其特有的特征衍射谱，没有任何两种物相的衍射谱是完全相同 的 2) 记录已知物相的衍射谱，并保存为PDF文件 3) 从PDF文件中检索出与样品衍射谱完全相同的物相 4) 多相样品的衍射谱是其中各相的衍射谱的简单叠加，互不干扰，检索程序能 从PDF文件中检索出全部物相 3 物相定量分析原理 X射线定量相分析的理论基础是物质参与衍射的体积活重量与其所产生的衍射强度成正比。 当不存在消光及微吸收时，均匀、无织构、无限厚、晶粒足够小的单相时，多晶物质所产生的均匀衍射环上单位长度的积分强度为： 式中R为衍射仪圆半径，V o为单胞体积，F为结构因子，P为多重性因子，M为温度因子，μ为线吸收系数。 三、仪器与材料 1)仪器：18KW转靶X射线衍射仪 2)数据处理软件：数据采集与处理终端与数据分析软件MDI Jade 6 3)实验材料：CaCO3+CaSO4、Fe2O3+Fe3O4
能源材料课程业 ——薄膜锂电池的研究进展 院系：材料科学与工程学院 专业：金属材料与成型加工 班级：2012级金属材成1班 学号：20120800828 姓名：吴贵军
薄膜锂电池的研究进展 摘要：微电子机械系统(MEMS)和超大规模集成电路(VLSI)技术的发展对能源的微型化、集成化提出了越来越高的要求.全固态薄膜锂电池因其良好的集成兼容性和电化学性能成为MEMS和VLSI能源微型化、集成化的最佳选择.简单介绍了薄膜锂电池的构造,举例说明了薄膜锂电池的工作原理.从阴极膜、固体电解质膜、阳极膜三个方面概述了近年来薄膜锂电池关键材料的研究进展.阴极膜方面LiCoO2依旧是研究的热点,此外对LiNiO2、LiMn2O4、LiNixCo1-xO2、V2O5也有较多的研究;固体电解质膜方面以对LiPON膜的研究为主;阳极膜方面以对锂金属替代物的研究为主,比如锡的氮化物、氧化物以及非晶硅膜,研究多集中在循环效能的提高.在薄膜锂电池结构方面,三维结构将是今后研究的一个重要方向.。 关键词：薄膜锂电池；微系统；薄膜：微电子机械系统随着电子集成技术的飞速发展，SO C (System on chi p) 成为 现实，电子产品在不断地小型化、微型化。以整合集成电路及机械系统，如各种传感器于同一块晶片上的技术，即微机电技术，受到了普遍重视。微小型飞行器、微小型机器人和微小型航天器等都在源源不断地出现和进一步地改进。这些微型系统的功能强大，必然对其能源系统提出了微型化的
要求。当电池系统被微型化，电池底面积小于10 m m2、功率在微瓦级以下时，被称为微电池。微电池的制备通常是将传统的电池微型化、薄膜化。目前，用于微电池的体系有：锌镍电池、锂电池、太阳能电池、燃料电池、温差电池和核电池。锂电池是目前具有较高比能量的实用电池体系，因此人们对薄膜化的锂电池投入了大量的研究。 优点： （1）成本低，根据Photon 的预测，预计到2012 年下降到2.08 美元/w；预计薄膜电池的平均价格能够从2.65 美元/w 降至1.11 美元/w，与晶体硅相比优势明显；而相关薄膜电池制造商的预测更加乐观，EPV 估计到2011 年，薄膜组件的成本将大大低于1 美元/w；Oerlikon 更估计2011 年GW 级别的电站其组件成本将降低于0.7 美元/w，这主要是由转化率提高和规模化带来的。 （2）弱光性好 （3）适合与建筑结合的光伏发电组件(BIPV)，不锈钢和聚合物衬底的柔性薄膜太阳能电池适用于建筑屋顶等，根据需要制作成不同的透光率，代替玻璃幕墙。 缺点： （1）效率低，单晶硅太阳能电池，单体效率为14%-17%(AMO)，而柔性基体非晶硅太阳电池组件（约1000平方厘米）的效率为 10-12%，还存在一定差距。
一、X光电子能谱分析的基本原理 X光电子能谱分析的基本原理：一定能量的X光照射到样品表面，和待测物质 发生作用，可以使待测物质原子中的电子脱离原子成为自由电子。该过程可用 下式表示： hn=Ek+Eb+Er （1） 其中:hn：X光子的能量；Ek：光电子的能量；Eb：电子的结合能；Er：原子的 反冲能量。其中Er很小，可以忽略。 对于固体样品，计算结合能的参考点不是选真空中的静止电子，而是选用费米 能级，由内层电子跃迁到费米能级消耗的能量为结合能Eb，由费米能级进入真 空成为自由电子所需的能量为功函数Φ，剩余的能量成为自由电子的动能Ek，式(1)又可表示为： hn=Ek+Eb+Φ（2） Eb=hn-Ek-Φ（3）仪器材料的功函数Φ是一个定值，约为 4 eV，入射X光子能量已知，这样， 如果测出电子的动能Ek，便可得到固体样品电子的结合能。各种原子，分子的 轨道电子结合能是一定的。因此，通过对样品产生的光子能量的测定，就可以 了解样品中元素的组成。元素所处的化学环境不同，其结合能会有微小的差别，这种由化学环境不同引起的结合能的微小差别叫化学位移，由化学位移的大小 可以确定元素所处的状态。例如某元素失去电子成为离子后，其结合能会增加，如果得到电子成为负离子，则结合能会降低。因此，利用化学位移值可以分析 元素的化合价和存在形式。 二、电子能谱法的特点 (1)可以分析除H和He以外的所有元素；可以直接测定来自样品单个能级光电 发射电子的能量分布，且直接得到电子能级结构的信息。(2)从能量范围看，如果把红外光谱提供的信息称之为“分子指纹”，那么电子能谱提供的信息可称 作“原子指纹”。它提供有关化学键方面的信息，即直接测量价层电子及内层 电子轨道能级。而相邻元素的同种能级的谱线相隔较远，相互干扰少，元素定 性的标识性强。 (3)是一种无损分析。 (4)是一种高灵敏超微量表面分析技术，分析所需试样约10-8g即可，绝对灵敏
【干货】玩转XPS丨案例解析X射线光电子能谱（XPS）八大应用！ 表面分析技术 (Surface Analysis)是对材料外层(the Outer-Most Layers of Materials (<100nm))的研究的技术。 X射线光电子能谱简单介绍 XPS是由瑞典Uppsala大学的K. Siegbahn及其同事历经近20年的潜心研究于60年代中期研制开发出的一种新型表面分析仪器和方法。鉴于K. Siegbahn教授对发展XPS领域做出的重大贡献，他被授予1981年诺贝尔物理学奖。 X射线激发光电子的原理 XPS现象基于爱因斯坦于1905年揭示的光电效应，爱因斯坦由于这方面的工作被授予1921年诺贝尔物理学奖； X射线是由德国物理学家伦琴（Wilhelm Conrad R?ntgen，l845-1923）于1895年发现的，他由此获得了1901年首届诺贝尔物理学奖。
X射线光电子能谱（XPS ，全称为X-ray Photoelectron Spectroscopy）是一种基于光电效应的电子能谱，它是利用X射线光子激发出物质表面原子的内层电子，通过对这些电子进行能量分析而获得的一种能谱。 这种能谱最初是被用来进行化学分析，因此它还有一个名称，即化学分析电子能谱（ESCA，全称为Electron Spectroscopy for Chemical Analysis）。XPS谱图分析中原子能级表示方法 XPS谱图分析中原子能级的表示用两个数字和一个小字母表示。例如:3d5/2（1）第一个数字3代表主量子数(n)； （2）小写字母代表角量子数; （3）右下角的分数代表内量子数j
第二十三章 X射线光电子能谱 1954年以瑞典Siegbahn教授为首的研究小组观测光峰现象，不久又发现了原子内层电子能级的化学位移效应，于是提出了ESCA(化学分析电子光谱学)这一概念。由于这种方法使用了铝、镁靶材发射的软X射线，故也称为X-光电子能谱(X-ray Photoelectron Spectroscopy)。X光电子能谱分析技术已成为表面分析中的常规分析技术，目前在催化化学、新材料研制、微电子、陶瓷材料等方面得到了广泛的应用。 23.1 基本原理 固体表面分析，特别是对固体材料的分析和元素化学价态分析，已发展为一种常用的仪器分析方法。目前常用的表面成分分析方法有：X射线光电子能谱（XPS）, 俄歇电子能谱(AES)，静态二次离子质谱(SIMS)和离子散射谱(ISS)。AES分析主要应用于物理方面的固体材料（导电材料）的研究，而XPS的应用面则广泛得多，更适合于化学领域的研究。SIMS 和ISS由于定量效果较差，在常规表面分析中的应用相对较少。但近年随着飞行时间二次离子质谱（TOF-SIMS）的发展，使得质谱在表面分析上的应用也逐渐增加。 X射线光电子能谱最初是由瑞典科学家K.Siegbahn等经过约20年的努力而建立起来的，因在化学领域的广泛应用，被称为化学分析用电子能谱（ESCA）。由于最初的光源采用了铝、镁等的特性软X射线，该技术又称为X射线光电子能谱（XPS）。1962年，英国科学家D.W.Turner等建造出以真空紫外光作为光源的光电子能谱仪，在分析分子内价电子的状态方面获得了巨大成功，同时又用于固体价带的研究，与X射线光电子能谱相对照，该方法称为紫外光电子能谱（UPS） XPS的原理是基于光的电离作用。当一束光子辐射到样品表面时，样品中某一元素的原子轨道上的电子吸收了光子的能量，使得该电子脱离原子的束缚，以一定的动能从原子内部发射出来，成为自由电子，而原子本身则变成处于激发态的离子，如图23-1所示。在光电离过程中，固体物质的结合能可用下面的方程式表示： E b=hγ- E k -φs（23-1） 式中: E k为射出的光子的动能；hγ为X射线源的能量；E b为特定原子轨道上电子的电离能或结合能（电子的结合能是指原子中某个轨道上的电子跃迁到表面Fermi能级（费米能级）所需要的能量）；φs为谱仪的功函数。 由于φs是由谱仪的材料和状态决定，对同一台谱仪来说是一个常数，与样品无关，其平均值为3 eV ~4eV。因此，（1）式可简化为： E b =hγ- E k’ （23-2） 由于E k’可以用能谱仪的能量分析器检出，根据式（23-2）就可以知道E b。在XPS分析中，由于X射线源的能量较高，不仅能激发出原子轨道中的价电子，还可以激发出内层轨道电子，所射出光子的能量仅与入射光子的能量及原子轨道有关。因此，对于特定的单色激发光源及特定的原子轨道，其光电子的能量是特征性的。当固定激发光源能量时，其光子的能量仅与元素的种类和所电离激发的原子轨道有关，对于同一种元素的原子，不同轨道上的电子的结合能不同。所以可用光电子的结合能来确定元素种类。图23-1表示固体材料表面受X射线激发后的光电离过程[1]。
《X射线荧光光谱分析的基础知识》讲义 廖义兵 X射线是一种电磁辐射，其波长介于紫外线和γ射线之间。它的波长没有一个严格的界限，一般来说是指波长为0.001-50nm的电磁辐射。对分析化学家来说，最感兴趣的波段是0.01-24nm，0.01nm左右是超铀元素的K系谱线，24nm则是最轻元素Li 的K系谱线。1923年赫维西(Hevesy, G. Von)提出了应用X射线荧光光谱进行定量分析，但由于受到当时探测技术水平的限制，该法并未得到实际应用，直到20世纪40年代后期，随着X射线管、分光技术和半导体探测器技术的改进，X荧光分析才开始进入蓬勃发展的时期，成为一种极为重要分析手段。 一、X射线荧光光谱分析的基本原理 元素的原子受到高能辐射激发而引起内层电子的跃迁，同时发射出具有一定特殊性波长的X射线，根据莫斯莱定律，荧光X射线的波长λ与元素的原子序数Z有关，其数学关系如下： λ=K(Z? s) ?2 式中K和S是常数。 而根据量子理论，X射线可以看成由一种量子或光子组成的粒子流，每个光具有的能量为： E=hν=h C/λ 式中，E为X射线光子的能量，单位为keV；h为普朗克常数；ν为光波的频率；C为光速。 因此,只要测出荧光X射线的波长或者能量,就可以知道元素的种类，这就是荧光X射线定性分析的基础。此外,荧光X射线的强度与相应元素的含量有一定的关系，据此，可以进行元素定量分析。 图1为以准直器与平面单晶相组合的波长色散型X射线荧光光谱仪光路示意图。 图1 平面晶体分光计光路示意图 A—X射线管；B—试料；C—准直器；D—分光晶体；E—探测器 由X射线管（A）发射出的X射线（称为激发X射线或一次X射线）照射到试料（B），试料（B）中的元素被激发而产生特征辐射（称为荧光X射线或二次X
X射线荧光光谱仪结构和原理 第一章 X荧光光谱仪可分为同步辐射X射线荧光光谱、质子X射线荧光光谱、全反射X射线荧光光谱、波长色散X射线荧光光谱和能量色散X射线荧光光谱等。 波长色散X射线荧光光谱可分为顺序（扫描型）、多元素同时分析型（多道）谱仪和固定道与顺序型相结合的谱仪三大类。顺序型适用于科研及多用途的工作，多道谱仪则适用于相对固定组成和批量试样分析，固定道与顺序式相结合则结合了两者的优点。 X射线荧光光谱在结构上基本由激发样品的光源、色散、探测、谱仪控制和数据处理等几部分组成。 §1.1 激发源 激发样品的光源主要包括具有各种功率的X射线管、放射性核素源、质子和同步辐射光源。波长色散X射线荧光光谱仪所用的激发源是不同功率的X射线管，功率可达4~4.5kW，类型有侧窗、端窗、透射靶和复合靶。能量色散X射线荧光光谱仪用的激发源有小功率的X射线管，功率从4~1600W，靶型有侧窗和端窗。靶材主要有Rh、Cr、W、Au、Mo、Cu、Ag等，并广泛使用二次靶。现场和便携式谱仪则主要用放射性核素源。 激发元素产生特征X射线的机理是必须使原子内层电子轨道产生电子空位。可使内层轨道电子形式空穴的激发方式主要有以下几种：带电粒子激发、电磁辐射激发、内转换现象和核衰变等。商用的X射线荧光光谱仪中，目前最常用的激发源是电磁辐射激发。电磁辐射激发源主要用X射线管产生的原级X射线谱、诱发性核素衰变时产生的γ射线、电子俘获和内转换所产生X射线和同步辐射光源。 §1.1.1 X射线管 1、X射线管的基本结构 目前在波长色散谱仪中，高功率X射线管一般用端窗靶，功率3~4KW，其结构示意图如下：
X 射线衍射分析 1实验目的 1、 了解X 衍射的基本原理以及粉末X 衍射测试的基本目的； 2、 掌握晶体和非晶体、单晶和多晶的区别； 3、 了解使用相关软件处理XRD 测试结果的基本方法。 2实验原理 1、 晶体化学基本概念 晶体的基本特点与概念：①质点（结构单元）沿三维空间周期性排列（晶 体 定义），并有对称性。②空间点阵：实际晶体中的几何点，其所处几何环境和 物质环境均同，这些“点集”称空间点阵。③晶体结构 =空间点阵+结构单元。非 晶部分主要为无定形态区域，其内部原子不形成排列整齐有规律的晶格。 对于大多数晶体化合物来说，其晶体在冷却结晶过程中受环境应力或晶核数目、 成核方式等条件的影响，晶格易发生畸变。分子链段的排列与缠绕受结晶条件的 影响易发生改变。晶体的形成过程可分为以下几步：初级成核、分子链段的 图1 14 种Bravais 点阵 表面延伸、链松弛、链的重吸收结晶、表面成核、分子间成核、晶体生长、晶体 生长完善。Bravais 提出了点阵空间这一概念，将其解释为点阵中选取能反映空 间点阵周期性与对称性的单胞，并要求单胞相等棱与角数最多。满足上述条件棱 间直角最多，同时体积最小。1848年Bravais 证明只有14种点阵。 Bravais lattice Cryslal DerBCfliptlcn CSlnwle) cubic Cubic a - b=? E , a = = 90B BaO/-rentered F*韓?“nl ?喇 (Simple) T etr-Aa^n^ i = b*C i .a=g = 7=SCi , Boa^-tentered ler^go>nal CGlnwle) DfthDmunbic Orlharhanbir 目日即亡时恒创 Orlhorhcmbic Oase-ceMered Offliorhombic Fac$-LBnter$d OnhorhChibie (Simple) Rhombol*edrai Rhombohedr^i (Trigonal) R = b = (;&= p= 7^ 90' 闭卿闾扫D 城帕1 H 曲g 肿对 a = b?iC fc tt=ft=0I]-a 7=12D - ⑻側 1.) MOnlQClHiC Monoclinic a c s a = y= go ■，&4 ger Monodlnlc ： (Slrrwle)AnDithl￡： Anotlhic (Triclinic) a * ： UH p 9D" PDFhAAiNT uaes Ehe 他 dfescnbing lhe pallerrt I Ellice cubic tatnigond orthortiombic rhonibdhedral hdxa.goinal (trigcnal) anortluc (tricl i me) iriufiOchrr ic