The optical properties of evaporated gold in the vacuum ultraviolet from 300 Å to 2 000 Å

HAL Id: jpa-00205721

https://hal.archives-ouvertes.fr/jpa-00205721

Submitted on 1 Jan 1964

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

The optical properties of evaporated gold in the vacuum ultraviolet from 300 Å to 2 000 Å

L.R. Canfield, G. Hass, W.R. Hunter

To cite this version:

L.R. Canfield, G. Hass, W.R. Hunter. The optical properties of evaporated gold in the vac- uum ultraviolet from 300 Å to 2 000 Å. Journal de Physique, 1964, 25 (1-2), pp.124-129.

�10.1051/jphys:01964002501-2012401�. �jpa-00205721�

a la transition de la bande « d ) au niveau de Fermi.

Du reste des r6sultats r6cents de Ehrenreich et

Philipp [4] ^ont montre que des bandes d’absorption

par effet photoélectrique ^{interne sont les memes}

pour le nickel et le cuivre qui occupent ^des positions

similaires compar6es ^{a celles} occupees par I’argent- palladium ^dans le tableau p6riodique.

11 semble _que ce soit donc aussi le ^cas pour la

sequence palladium-argent.

Nous allons poursuivre ^{cette 6tude en examinant}

les propri6t6s ^du palladium ^{et de} 1’argent puis ^en

observant ce qui ^se produit lorsque ^{nous etudierons}

les propri6t6s ^de l’alliage argent-palladium.

Discussion

M. ELLIS.

1o Were these measurements of reflectance and transmittance taken in air or vacuum.

20 If in _vacuum, do the values change upon admission of air ?

M. LosTIs. - Values were taken in air.

M. BLANC. -- Distance approximative support

creuset ?

M. LOSTIS. - 35 cm.

BIBLIOGRAPHIE [1] ^LOSTIS (P.), ^Thèse Ing.-Docteur, juin ¹⁹⁵⁸ (Rev. Opt.,

1959, 38, 1).

[2] ^ABELÉS (F.), ^{J. O. S.} A., 1957, 47, ^473.

[3] ^MARÉCHAL (A.), ^DUPUY (O.) ^{et RENAULT} (M.), Opt.

Acta, 1962, 9, ^47.

[4] EHRENREICH et PHILIPP, ^à paraître ^dans Physical

Review.

THE OPTICAL PROPERTIES OF EVAPORATED GOLD IN THE VACUUM ULTRAVIOLET FROM 300 Å TO 2 000 Å (1)

By L. R. CANFIELD and G. HASS,

U. S. Army Engineer Research and Development Laboratories, ^Fort Belvoir, Virginia.

and

W. R. HUNTER,

E. O. Hulburt Center for Space Research, U. S. Naval Research Laboratory, Washington, ^{D. C.}

Résumé.

On a mesuré les facteurs de réflexion de couches d’or, obtenues par évaporation,

dans l’ultraviolet de 300 à 2 000 Å. On a déduit les constantes optiques ^{des mesures de facteur}

de réflexion effectuées _pour différents angles d’incidence. Contrairement à ce qui ^se passe pour

l’aluminium, le facteur de réflexion de couches d’or préparées ^sous vide varie peu ^{au cours de leur} exposition ^{à l’air.} Des couches semi transparentes, ^d’une épaisseur voisine de 150 Å, présentent

des maximums de réflexion, ^{dus à des} phénomènes d’interférence, pour ^de nombreuses longueurs

d’onde de l’ultraviolet lointain.

Abstract.

The reflectance ^{and optical} constants of evaporated gold ^{films were measured in}

the vacuum ultraviolet from 300 Å to 2 000 Å. The optical ^{constants were} determined from reflectance measurements made at various angles ^of incidence. In contrast to aluminium, gold

shows little change in reflectance after deposition ^{in vacuum and} during exposure to air. Due to interference, semitransparent ^{films of about 150} Å on glass ^{show the} highest reflectance at most

wavelengths ^{in the vacuum} ultraviolet.

LE JOURNAL DE PHYSIQUE ^TOME 25, JANVIER-FÉVRIER 1964,

Introduction.

A study ^{of the} optical prop- erties of gold ^{in the vacuum} ultraviolet is of interest for practical ^and theoretical reasons. The reflectance of gold ^is higher ^than that of many (1) ^{This work was} supported ⁱⁿ part by ^{the Goddard} Space Flight Center, NASA, Greenbelt, Maryland.

other film materials for wavelengths shorter than 900 A and exhibits almost no change during

extended use and storage ^{in air. In} addition, gold

films are easy ^to deposit ⁱⁿ any desired thickness and have _{proven to} be an excellent medium in which tojrule diffraction gratings ^{for use in the}

extreme ultraviolet.

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphys:01964002501-2012401

125 From the theoretical point ^of view, ^measure-

ments of the reflectance and the optical ^constants

of gold ^{in the vacuum} ultraviolet are of interest to the solid state physicist ^since they furnish infor- mation on the absorption spectrum ^and energy band structure of the metal.

It is the purpose ^of this paper ^to present ^the

results of recent measurements of the reflectance and optical ^{constants of} gold ^{in the} wavelength region ^{from 300 A to 2 000} A, ^{and to discuss the}

effect of film thickness on the optical properties ⁱⁿ

this region.

Experimental techniques.

^All evaporations

were performed ^{in an} evaporator-reflectometer

combination in which the reflectance of the films could be measured within seconds of ^the comple-

tion of their deposition. ^The instrument, ^shown

in figure 1, is attached ^to a one-meter normal incidence monochromator. A detailed description

FIG. 1.

Section of evaporator-reflectometer combination

of the system ^{has been} published ^elsewhere [1],

therefore only ^the salient features of this instru- ment will be mentioned here. Basically ^the evapo- rator-reflectometer consists of two vacuum cham-

bers separated by ^{a barrier with a hole} in it. The vapor source, in this ^{case a} tungsten ^{boat con-} taining gold, ^{is located} in the lower chamber, ^while

the upper chamber contains the reflectometer.

The substrate holder may be disengaged ^{from the}

reflectometer mechanism and placed ^{over the hole}

in the barrier between the two chambers. Thus, during ^an evaporation, the substrate faces the boat which is approximately ^{40 cm} away from it. The

substrate and an additional shutter serve to shield the upper chamber ^{and its contents from the} metal vapor. Immediately after the completion

of an evaporation the substrate may be swung up into position ^for measurement.

The polished glass plates ^{used as} substrates were

. given ^a preliminary cleaning ^{with a} detergent and, just prior ^{to the} evaporation, ^{a final} cleaning ^with

a high-voltage ^dc glow discharge.

The evaporation ^{rates of} gold ^{were varied so}

that the deposition ^rates ranged ^{from 2 to}

1 600 Å/sec; ^{the thickness was monitored} by measuring ^the light transmittance at À

5 000 A.

During ^the evaporations the pressure ^was approxi- mately ^10-6 torr, except ^{when it was} deliberately

increased to study the effect of pressure ^on the

optical properties.

Two types ^of light sources were used ^to cover the

spectral range from 300 A to 2 000 A. One was a dc glow discharge (1) ^{and the other a} pulsed discharge ^source (2). Figure 2 shows the spectra

of the radiation available for measurement. with the two light ^{sources. The radiation emitted from}

the glow discharge ^{is shown at the} top ^of ( fig. ^2).

The many-line hydrogen ^molecular spectrum,

which within the limits of resolution of the mono-

chromator is a quasi-continuum, ^{and two atomic}

lines at 1216 A and 1 026 A, ^{cover the} spectral region ^from approximately ^{900 A to} beyond ^the

upper wavelength limit used in this investigation.

The neutral resonance lines of He and Ne produced

isolated lines at 584 A and 736 Å, respectively,

when those gases ^were introduced into the light

source. More complete coverage of the region

below 900 A is afforded by using ^the pulsed ^dis- charge ^{source. The remainder of} figure ^{2 shows}

the radiation available using ^the pulsed discharge

with different gases. The spectral ^{range down to}

300 A is. covered rather fully using the four gases shown. With _oxygen, wavelengths ^{as short as}

246 A _may be obtained.

The reflectance of the films could be measured at any angle of incidence between 60 and 860 using

a 1P21 photomultiplier sensitized with sodium sali-

cylate ^{as a} detector. For the determination of

optical ^constants by the reflectance method des- cribed by Tousey [3] ^{and Simon} [4], reflectance measurements were made at several angles ^{of inci-}

dence at each wavelength. ^The resulting ^values

of reflectance were then used to determine the

possible values of the index of refraction n and of the extinction coefficient k (the complex ^index

of refraction, ^{N = n}

ik) ^{for each} angle ^{of inci-}

dence by ^{means of} previously computed ^curves.

The values of n and k common to each determi- nation for several angles of incidence then gave the optical ^{constants at the} wavelength ^under study. These initial values were then fed into ^a

digital computer ^{which was} programmed ^{to cal-}

FIG. 2.

Spectra produced by ^{the two} light ^sources.

culate the reflectance for each angle ^{at which}

measurements were made and ^to evaluate the de- viation between the measured and calculated values.. The constants ^were then varied by ^small

amounts and the calculation repeated ^{until the}

deviation ^was minimized. ^The program ^{also com-}

puted the reflectance ^at normal incidence which,

in all _cases, agreed ^within experimental ^{error with}

the measured values.

Results.

Evaporated gold shows little change

in vacuum ultraviolet reflectance after its depo-

sition in _vacuum, and during exposure ^{to and} pro-

longed storage ^{in air. The marked contrast in} behavior ^to that of a reactive metal such as alumi-

num may be ^seen in figure ^{3. In separate exper-} iments, ^the reflectance of _{freshly deposited} ^alumi-

num and gold ^{were studied as a function of time}

at 1216 A, before and after exposure ^to air. The reflectance of aluminum drops rapidly, particu- larly upon exposure ^to air, ^{due to} the formation

FIG. 3.

Comparison of the effect of aging ^{on the reflec-}

tance of evaporated ^{films of} gold and of aluminium at 1 216 A.

of a strongly absorbing ^oxide layer [5]. The reflec-

tance of gold, however, ^decreases only ^{about one}

percent ^{while in vacuum and} stays essentially

constant after exposure ^to air. At other wave-

lengths ^the degree ^of change ^{was about the. same,}

but at some wavelengths ^a slight increase could be observed.. Contamination resulting ^from storage

in air sometimes caused a slight change ^{in reflec-}

tance at all wavelengths, however, ^a simple ^rinse

with ethyl alcohol restored the reflectance prac-

tically ^{to its} original value. Thus the conclusion

was reached that measurements in situ were unnec-

essary, and that the ^true optical properties ^of gold

could be studied using films that had been exposed

to air.

The effect ^of deposition parameters ^on the reflec- tance of gold ^{films was} investigated, ^{and it was}

found that neither the pressure during deposition

nor the rate of deposition ^were critical until extremes were reached. Deposition ^{rates of from}

50 to 1 600 A _per second with pressures of from less than 10-6 to above 10-5 torr were employed

with practically ^{no effect on} the reflectance of the

resulting ^film. Only ^{when a} rather thick film ^was made under poor conditions did the reflectance

drop, undoubtedly ^{due to increased surface} rough-

ness. Ip ^{figure 4} the upper solid curve represents

the reflectance ^{of a} film made in a more or less

typical evaporation ; the lower is the reflectance of a film made under the relatively poor conditions

given. ^The reflectance values of Robin [6], ^obtain-

ed from visibly opaque films and shown as the dashed curve of figure 4, ^are considerably ^lower

than those measured here.

FIG. 4.

Effect of evaporation conditions on the reflec- tance of evaporated gold ^{as a} function of wavelength

from 1 000 A to 2 000 A.

If the surface of the evaporated ^{gold film is} rough, ^{it will scatter} the incident radiation and

cause erroneous reflectance measurements which will result in inaccurate determinations of n and k.

Since roughness increases with increasing ^film thickness, the films should be kept ^{thin to avoid} roughness but, ^{at the same} time, ^thick enough ^to

be _opaque. Interference effects and transmission losses will be negligible ^if less than 0.1 % ^{of the}

energy incident of the vacuum-film interface is transmitted through ^{the film to the} glass ^substrate.

Calculations of the film thickness required ^to satisfy

this condition were made using preliminary ^values

for n and k obtained from a film 1600 A thick.

A plot the of thickness necessary ^to reduce trans- mission to 0.1 % ^{as a} function of wavelength ^from

300 Å ^to 2 000 A is shown in figure ^{5. The} optical

constants given by ^Schulz [7] ^and Schulz ^and Tangherlini [8] ^{were used to} calculate the trans- mittance ^of gold ^{at 5 000} A as a function of thick- ness, and ^the right-hand ^{scale of} figure ⁵ gives ^the

transmittance at 5 000 A corresponding ^{to the}

FIG. 5.

Calculated thickness of gold ^films required ^to

decrease transmittance to 0.1 % ^{as a} function of wave-

length ^{from 300 A to 2 000} A. Transmittance data at 5 000 A are included for comparison.

thickness given ^{in the left-hand} scale..It is seen

that films that are opaque in the ^vacuum ultra- violet are semitransparent ^{at 5 000 A. Thus}

monochromatic light ^{of that} ’wavelength may be used to monitor the thickness of gold ^{films that}

will be opaque in the ^vacuum ultraviolet.

Once the desirable thickness for gold ^{films was}

determined for the various wavelengths, ^meas-

urements of n and k were made using ^{films of the} appropriate thicknesses. The results are shown

FIG. 6.

Optical constants and normal incidence reflec-

tance of gold ^{in the vacuum} ultraviolet.

in figure 6, ^where n, k, ^{and the} measured normal incidence reflectance are plotted ^{as a} function of

wavelength ^{from 300 A to 2 000 Å. Cole and} Oppenheimer [9] ^{have measured the} optical ^cons-

tants of gold ^{at the} wavelengths 1216, 1048, 920,

584 and 304 A. Their values of n and k _are, with the exception of the indices at 584 A and 304 A, ^{lower than the ones} reported ^here.

Reflectance measurements from 1770 A to 450 A

have been reported by Walker, Rustgi ^and

Weissler [9a] ^who obtained the same reflectance

versus wavelength characteristic ^as shown in

figure ^{6. Their} reflectances were equal ^{to those}

shown in the figure ^for wavelengths greater ^than

700 A, ^but somewhat lower for wavelengths ^shorter

than 700 A.

The extinction coefficient, k, ^{cannot be corre-}

lated directly ^with- absorption processes occurring

in the metal, however, ^{Frohlich and Pelzer} [10]

have shown that a phenomenological relation exists between absorption processes and the bulk optical

constants. According ^to them, ^the quantity,

when plotted against wavelength, ^{should have}

maxima corresponding ^to absorption processes.

This expression ^was calculated using ^{the n and k}

values of figure 6 and the results are shown in

figure ^{7. There are distinct} maxima ; ^{at 32.6} eV,

25.8 eV, ^16.3 eV, ^{and 6.7} eV, corresponding ^to wavelengths 380, 480, 760 and 1 880 A.

FiG. 7.

2nk/(n2 + k2)2 ^{as a} function of wavelength

for gold ^{in the vacuum} ultraviolet.

Robins [11] ^{used a} reflection technique ^to study monoenergetic ^electrons elastically and inelas-

tically ^scattered by ^an evaporated gold ^surface

and observed four characteristic losses of about the same energy ; 32.6 eV, ^25.8 eV, 16.0 eV and 6.3 eV.

Interference effects may be observed _{when very} thin gold ^{films are} deposited ^on glass. Figure 8

shows the calculated and measured normal inci-

FIG. 8.

Calculated and measured reflectance _{of opaque} and semi-transparent (t

¹⁵⁰ A) gold ^{films on} glass ^as

a function of wavelength ^{from 1} 000 A to 2 000 A.

dence reflectances of opaque and 150 Å thick films from 1000 A to 2 000 A. The agreement ^between calculated and measured values for the 150 A thick film indicates that n and k of the bulk material

are still correct for films this thin. The curve also shows that highest reflectance is obtained with films that are not opaque ^{in the extreme} ultra- violet. The same situation applies ^{to films of} platinum [12].

The calculated reflectance of evaporated gold ^on glass ^{as a function} of film thickness at several

wavelengths is shown in figure ^{9. At all wave-}

Fic. 9.

Calculated reflectance of gold ^on glass ^{as a func-}

tion of film thickness for various wavelengths ^{in the}

vacuum ultraviolet.

lengths ^shown except ⁷⁰⁰ Å there is an increase in reflectance at a gold thickness less than about 250 Å. It can be concluded that for use as a

reflector throughout ^{the vacuum} ultraviolet a gold

thickness of about 150 A on glass represents ^a good

compromise.

129

REFERENCES [1] ^MADDEN (R. P.) and CANFIELD (L. R.), ^J. Opt. ^Soc.

Am., 1961, 51, ^838.

[2] ^HUNTER (W. R.), Proc. Xth Colloquium Spectrosco- picum Internationale, Spartan ^Books, Washington,

D. C., 1963, p. 247.

[3] ^TOUSEY (R.), ^J. Opt. ^Soc. Am., 1939, 29, 235.

[4] ^SIMON (I.), ^J. Opt. ^Soc. Am., 1951, 41, 336.

[5] ^MADDEN (R. P.), ^CANFIELD (L. R.) ^{and HASS} (G.),

J. Opt. ^Soc. Am., 1953, 53, 620.

[6] ^ROBIN (S.), C. R. Acad. Sc.,1953, 236, 674.

[7] ^SCHULZ (L. G.), ^J. Opt. ^Soc. Am., 1954, 44, ^357.

[8] SCHULZ (L. G.) and TANGHERLINI (F. R.), ^J. Opt. ^Soc.

Am., 1954, 44, ^362.

[9] ^COLE (T. ^{T.) and} OPPENHEIMER (F.), Applied Optics, 1962,1, ^709.

[9a] ^WALKER (W. C.), ^RUSTGI (O. P.) ^{and WEISSLER} (G. L.), ^J. Opt. ^Soc. Am., 1959, 49, 471.

[10] ^FROHLICH (H.) ^{and PELZER} (H.), ^Proc. Phys. Soc., 1955, ^A 68, ^525.

[11] ^ROBINS (J. L.), ^Proc. Phys. Soc., ^1961, 78, ^1177.

[12] ^JACOBUS (G. F.), ^MADDEN (R. P.) ^{and CANFIELD} (L. R.), ^J. Opt. ^Soc. Am., 1963, 53, ^1084.

THE ANGULAR AND SPECTRAL DISTRIBUTION OF TRANSITION RADIATION FROM THIN SILVER FOILS

By ^E. T. ARAKAWA, ^{N. O.} DAVIS, ^{L. C.} EMERSON and R. D. BIRKHOFF,

Health Physics Division, ^Oak Ridge ^National Laboratory (1), ^Oak Ridge, ^Tennessee.

Résumé.

On a déterminé expérimentalement les distributions angulaire ^et spectrale, ^{dans le}

domaine de longueurs d’onde de 2 500 Å à 5 600 Å, ^de _photons émis par ^des feuilles d’argent ^de

660 Å et 1 980 A d’épaisseur lorsqu’on les bombarde avec des électrons de 40 keV. La distribution

spectrale de la lumière polarisée ^{dans le} plan d’incidence a montré un maximum aigu ^à

3 300 Å ± ¹² Å. Cette valeur est une moyenne évaluée ^sur 1’ensemble des directions faisant avec

la normale à la feuille des angles ^allant de 10° à 50°. En plus ^{de ce maximum, le} spectre présentait

un fond continu dont l’intensité décroissait lentement lorsque ^la longueur ^d’onde augmentait,et

un faible minimum à 3 200 A, suivi, pour les longueurs ^d’onde plus ^{courtes, d’une} augmentation

d’intensité. La distribution angulaire ^des photons émis, pour la longueur ^d’onde correspondant

au maximum, présente ^{un maximum à 30°} de la normale à la feuille. Aucun photon ^{n’est émis} perpendiculairement ^ou tangentiellement ^à la feuille. En revanche, la lumière émise pour d’autres

longueurs d’onde du fond continu, par exemple ^{2 700 Å et 4 500} A, était surtout intense à 50°

de la normale. On a déterminé le rendement lumineux absolu après avoir étalonné le spectro- mètre, l’analyseur ^{et Ie} photomultiplicateur ^{avec une} lampe ^à filament de tungstène provenant

du U. S. National Bureau of Standards. En ce qui ^concerne les directions faisant un angle ^{de 10°.}

à 40° avec la normale, le rendement lumineux était en tout point ^{en accord avec les} prévisions théoriques pour toutes les longueurs ^d’onde sauf pour celle correspondant ^{au maximum} aigu

pour laquelle la valeur expérimentale ^est inférieure de 30 % ^environ.

Abstract.

The angular ^and spectral distributions of photons ^emitted by Ag ^{foils 660 Å and}

1 980 Å in thickness when bombarded by ⁴⁰ keV electrons have been determined experimentally

in the wavelength region ^{from 2 500 A to 5 600 Å. The} spectral distribution of light polarized

in the plane ^of incidence showed a sharp peak ^{at 3 300 Å ± 12} Å, ^this value being ^an average

over photon directions from 10° to 50° from the foil normal. In addition to the peak, ^the spectrum

showed a continuum which decreased slowly ^with increasing wavelength ^{and a} deep ^minimum

at 3 200 Å with a rise in the shorter wavelengths. ^The angular distribution of photons ^emitted

at the peak wavelength ^{showed a maximum at 30° from the} foil normal with zero intensity ^{at 0°}

and near 90°, whereas the photons emitted at other wavelengths ^{in the} continuum, e.g., 2 700 A

and 4 500 Å, ^{were most} intense at 50° from the foil normal. The absolute photon yield ^{was deter-}

mined by calibrating ^the spectrometer, analyzer, ^and photomultiplier ^{with a} tungsten ^filament lamp obtained from the U. S. National Bureau of Standards. For photon directions from 10°

to 40° the photon yield ^{was found} to agree with the theoretical predictions ^{in all} respects ^{at all} wavelengths except at that of the sharp peak ^{where the} experimental ^{values were about 30} %

lower.

JOURNAL DE PHYSIQUE TOME 25, JANVIER-FÉVRIER 1964,

Introduction.

In 1945 Frank and Ginsburg [1]

predicted that radiation would be emitted when ^a

uniformly moving ^electron passed ^{from one}

medium into another. To explain ^the pheno-

(1j Operated by Union Carbide Corporation ^{for the}

U. S. Atomic Energy Commission.

menon of this " transition radiation ", they ^consi-

dered an electron passing ^{from a vaccum into a} perfect conductor. While approaching ^{the con-}

ductor the electromagnetic ^{field in the vacuum is}

equal ^{to the field of the} electron and of its image

moving towards it. From the point ^{of view of}