Some Methods of Luminescence Efficiency Measurements*

Methods of absolute and relative radiant and quantum efficiency measurements are described for ultraviolet, visible, cathode-ray, and x-ray excitations. Data on some standard luminescent materials are given.


. Introduction
Me th ods of absolut e radi a nt a nd qu a ntum e ffi cie ncy meas ure me nts are giv e n togeth e r with meth ods of relati ve e ffi c ie ncy meas ure me nts. Th e me th ods are es pec ia ll y s uit ab le fo r powd e r ma te ri als fo r whi c h th e a ngular di s tributi on of th e e mitted lumin esce nt radi ati on is La mbe rti an.
Th e re la ti ve meas ure me nts a re perfo rm ed with th e aid of s ta nd a rd ph os ph ors, whose effi cie ncies have previously bee n de te rmin ed by a bsolute meas ure me nts. Methods are give n fo r excita ti on of th e ph os phors by ultraviole t a nd vis ible radiati on , cathod e rays and x rays.
F or samples with non-La mbe rti a n e mi ss io n di stributio ns, a me thod is desc ribed in whi c h a n Ulbri c ht 's s phere or an e lliptical mirror is used.

. Ultraviolet Excitation
All powder phosphors are meas ured us in g a thi c k layer (thi c kn ess ab out 2 mm) at th e irradi ated side. Th e detection tak es place pe rpe ndi c ular to th e plan e of the phos phor, t he excitati on is at a n a ngle of 50° with th at plan e (see fi g. 1). Th e e xcita tio n wa vele ngth ( A exc) or regio ns a re iso la ted from a hi gh pressure merc ury la mp by inte rfe re nce filt ers, the arc being foc used on th e phos phor with a qu artz le ns. In thi s way a hi gh excita ti on de nsity is reached , but generally well below th e excitation region where s aturation e ffects s ta rt. This is es peciall y ad vantageo us wh en a rela ti vely in se ns iti ve th erm oele me nt is used as a d etector.
* P aper p re se nt ed at th e W ork s hup Sem in a r 'Slandard izat ion in S pec trophutometry a nd L u m in esce lwe Meas ure rn e nl s' held a t the Nat iona l Burea u uf Standards. Ga ithersbu rg , Md .. Nove mbe r 19-20. 1975. The radiant e fficiencies, from whic h the quantum e fficiencies are calculated, are de te rmined directly (whe n the s pectral power distribution is known).

. 1. Relative Measurements, Giving Absolute Efficiency Values
Phos phors can be meas ured with res pec t to th e following standard sa mples whose e ffi cie ncy is ge ne rally agreed upon.
Th is ph os ph or is also s uit able for exc itati on in the far ultraviolet (vacuum ultraviolet) because of its constant efficiency as a function of Ae xc up to 350 nm. (c) The standard "Ekta S10"1 proposed by Grum [6]. (d) "Lumogen T red GG," which can be used in the excitation region between 190 and 550 nm [7].

Absolute Measurements
The absolute radiant efficiency can in fact be determined with the aid of a relative measurement, being the ratio of the amount of emitted power and that of the absorbed exciting power [IV For one or two wavelengths the absolute efficiencies can be determined. For other Aexc the relative excitation spectrum can be determined from which the absolute efficiency at any Aexc can be derived.
. Forthis determination three quantities are measured: (a) The diffuse reflection of the exciting radiation against BaS04 for which the reflection is known.
(b) The luminescence + reflection of the exciting radiation (without using a filter). (c) The luminescence of the phosphor, using a filter between phosphor and detector that passes only the luminescence.
From these three measured quantities the reflection and radiant efficiency of the phosphor can be determined. The expressions found for the radiant efficiency YJ p and the reflection rp are given here for the case of using as a detector a thermopile or thermoelement with flat radiant response. Three emfs are measured, viz, VII due to the reflection standard (e.g., BaS04 [8], reflection R) , VI' due to the phosphor (luminescence intensity L + reflected exciting radiation of intensity l) and Vp , v due to the phosphor when a filter F absorbing the exciting radiation is placed in front of the detector. We assume that the filter has a transmission T in the emission region of the phosphor. This leads to the following equations: After solving for fp and L we find

I(l-rp) T(I-fp) VII
I In order to desc ribe mat e rial s a nd ex perimen tal procedures adequately, it is occasionally necessary to ide ntify co mm ercial products by manufacturer's name or labeL In no instance d oes s uc h id e ntifi cation impl y e ndorse ment by the Nati ona l Bureau of S tandard s. nor does it imply that th e particular produc t or equipm ent is ne cessa ril y the best available for that purpose.
2 Fi gures in bracket s indicate th e literature referenc es at the e nd of thi s paper.
As a cross-check the reflection found in this way can be compared with that measured directly with a spectrophotometer. The method described can be used in the same way for the case of a varying spectral response of the detector and/or a varying spectral transmission of the filter, even when the filter transmits partly in the region of the exciting radiation. Of course the equations become somewhat more co mplicated in this case.
The quantum efficiency qp is found from the radiant efficiency by where P (A) is the emitted luminescent power and Aexc is the exciting wavelength.
The NBS standards mentioned in 2.1 are not excited in the visible (only No. 1030 would be suitable in the blue-violet) region. Therefore. other standards are necessary in the visible region. These can be found among the "lumogen" phosphors. A yellow luminescent lumogen was described by Kristianpoller and Dutton [9], yellow and red ones by Vavilov [10] . Morgenshtern, Neustruev and Epshtein [11] and Kuttner, Selzle and Schlag [12]. The latter used 5-(p-dimethylaminobenzyliden)-barbituric acid as a red lumogen; they found a quantum efficiency of 45 percent at Aexc = 405 nm.
We chose the red luminescent" Lumogen T red GG" which was already mentioned in section 2.1. It is commercially available from the Badische Anilin und Soda Fabrik (Ludwigshafen. Germany). The . properties of the phosphor are described in reference [7], It has a red luminescence and shows a quan-'tum efficiency which is not quite constant but varies in a limited range between 40 percent and 60 percent in the spectral region between 220 nm and 550 nm (see fig. 2). Th e absorption of rhodamine B is given in figure 6, s howin g th e e normous variation through the spectrum leadin g to a similar large variation in light output. Another drawb ac k of liquid samples is the different geo me try of the set-up needed for the measurement. Va ri ous authors ha ve reported meas ure me nts usin g 254 nm merc ury vapour di sc harge excitation. H ere we giv e additional meas ure men ts o n so me standards for lon ger wav elength excitation at A exc= 365 nm. The phos phors meas ured were sodium salic ylate, the " Ekta SlO" sa mple, introdu ced by Crum [6] and " Lumogen T r ed GC" [7] (see tables 1 and 2).
Th e res ults for diffuse refl ec ti on at the excitin g and e mi ss ion wavele ngth , th e practi cal and intrin sic rad iant e ffi cie ncies, and th e prac ti ca l and intrins ic quantum e ffi cie ncies are gi ven. Becau se of the thic k layer used , a correcti on has to be made for the loss of th e lj ght absorbed in th e layer. The intrinsic radiant effi cie ncy YI i can th e n be approximated by [2] 2 Yli = l + r", YIp where r x is . the reflection coefficient of the phosphor for an infinitely thick layer.
The diffus e reflection of "Ekta SlO" is given In figure 7, the spectral power distribution in figure 8. The efficiency data for N a-salicylate at " exc = 260 nm ca n be co mpared with the data given in Samson' s book (ref. [3]) which are disc us sed by us in reference [7], together with so me additional data.
Polarization effects in our measurements proved to be negligible, as may be expected for powder materials.
Measurements we re carried out with incident polarized UV radiation, in two directions perpendicular to each other.
The stability of the lumoge n was also tested as well as the dependence on excitation density. During one month the efficiency of the lumogen was measured e very two days. The stability in time proved to be very good; no changes were observed within the error of measurement, which was of the order of ± 10 percent. The effici ency values were not affected e ven when the intensity of the UV -radiation was attenuated a thousand times.

Excitation in Selected Narrow Absorption Peaks
A method to determine the efficiencies of phosphors that have a small absorption of a few percent in narrow, well defined excitation levels (for the normal case of "exc ~ " em) was described earlier by us [13,14] .
Examples of these powders are rare-earth activated phosphors, such as YV04 -Eu 3 + and NaYF4-Er3 + , where the (visible) excitation peaks are those of the rare-earth ion. The host lattice absorbs in the UV region.
A diagram of the se t-up is shown in figure 9. The phosphor is irradiated via a scanning monochromator. Two measure ments have to be carried out, diffe ring only in the filter use d in front of the photomultiplier. One filter transmits only the light refl ected from the sample, giving the absorption spectrum_ In the second measurement the other filter selects the emission wavelength region, thus obtaining the excitation s pectrum of that e mission_ The curves are of the type shown in figure 10 for YV04 -Eu3+. The efficiency is calculated as follows.
The radiant efficiency is the ratio of the emitted power E to the absorbed exciting powe r A. The latter is determined by the area under the absorption curve of a certain peak with correction for the transmission T A of the filter used and for the photomultiplier re-

Cathode-Ray Excitation
Th e radiant e ffici e ncy 7)" for cath ode -ray excitation [2-4 , 15 , 16] is ge nerally defi ne d as th e ratio of th e a mount of e mitte d lumin esce nt power in the spectral regio n unde r co nsideration to the power of th e in c id e nt cath ode-ray beam (a nd not to th e powe r absorbed by th e phos ph or laye r). Thus no correctio n is made for the loss due to re fl ec ti on of primary e lec trons [2, 15,16].
In thi s case two reaJly absolu te meas ure me nts are necessar y, viz., th a t of th e e milte d power a nd th at of th e pow e r of th e cath ode·ray beam (fi g. 11). The meas urements are carried out on thick layers at the irradiated side. Precautions should be taken to e nsure that charging up of the layer is negligible.
The radiant output of the phosphor was compared with the radiation of a standard lamp which was diffusely reflected by a MgO layer. A thermopile was used [2] las a detector.
, i

X-Ray Excitation
To measure radiant efficiencies with x·ray excitation [17 -19] thin phosphor layers are used (= lOOp-m). This is necessary to minimize the loss in light output due to scattering and absorption of the emitted luminescence. The total back·screen emission is collected by a 27T-geometry elliptical mirror and focused onto the photomultiplier detector (see fig. 12), which is calibrated in absolute units (A/W). The x-ray absorption coefficients . are measured ,with a scintillation crystal as well as calculated from the tables of Storm and Israel [20].

Crystals, etc
In cases where the angular distribution of the e mitted radiation does not obey Lambert's law it is not sufficient to measure the emitted radiation in one direction but the total radiation should be determined. This can be carried out with the aid of an Ulbricht's sphere or with an elliptical mirror.
(2) the luminescent output is compared with the output of a calibrated standard lamp, e.g., a 200 W or 1000 W tungsten· halogen lamp; calibrated by the National Bureau of Standards in Washington, D.C. (W /nm . cm 2 ). In this case the diode to be measured is replaced by a BaS04-coated screen S.
The use of a 27T-geometry elliptical mirror [21] instead of an Ulbricht's sphere gave nearly the same results.