See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/235000960

Quantum Yield of the Ferrioxalate Actinometer

Article  in  The Journal of Chemical Physics · January 1964

DOI: 10.1063/1.1725147

CITATIONS READS
105 980

2 authors, including:

John W Lee
University of Georgia
144 PUBLICATIONS   2,980 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

bioluminescence View project

All content following this page was uploaded by John W Lee on 08 July 2017.

The user has requested enhancement of the downloaded file.

Quantum Yield of the Ferrioxalate Actinometer
John Lee and H. H. Seliger

Citation: J. Chem. Phys. 40, 519 (1964); doi: 10.1063/1.1725147

View online: http://dx.doi.org/10.1063/1.1725147
View Table of Contents: http://jcp.aip.org/resource/1/JCPSA6/v40/i2
Published by the American Institute of Physics.

Additional information on J. Chem. Phys.

Journal Homepage: http://jcp.aip.org/
Journal Information: http://jcp.aip.org/about/about_the_journal
Top downloads: http://jcp.aip.org/features/most_downloaded
Information for Authors: http://jcp.aip.org/authors

Downloaded 11 Oct 2012 to 128.192.114.19. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions
THE JOURNAL OF CHEMICAL PHYSICS VOLUME 40, NUMBER 2 15 JANUARY 1964

Quantum Yield of the Ferrioxalate Actinometer

JOHN LEE* AND H. H. SELIGER
McCollum Pratt Institute and Department of Biology, The Johns Hopkins University,
Baltimore, Maryland 21218
(Received 16 September 1963)

The quantum yield of the potassium ferrioxalate actinometer at 365/6 m,u has been measured to be
1.26±0.03 relative to a cali~rated thermopile. Comparisons of thermopiles were carried out at the Eppley
La~or~tory and at the National Bureau of Standards to ensure the flatness of thermopile response. Ir-
radlatlOns were made over the range 1012 to 3 X 1014 total absorbed photons.

INTRODUCTION cantly from ideal blackness which we reported pre-

viously.7 Second, we have found a reproducible, and as
HE most thorough examination to date of the use
T of the photoreduction of potassium ferrioxalate as a
chemical actinometer for the determination of fluxes of
yet unexplained, enhancement of the quantum yield
when the sides of the square fused silica cuvette are
incident radiation has been made by Parker and co- included in the irradiation beam.
workers,I·2 revising the initial work of Allmand and EXPERIMENTAL
Webb in 1929.3 These former experimenters recom-
mended a quantum yield of 1.21 at 365/6 Ill,LL. In the The preparations of materials and the procedures
course of our experiments on the measurement of the were followed according to the directions of Hatchard
spectral sensitivity of electron multiplier phototubes,4 and Parker. Solutions were prepared in total darkness
we decided to use Parker's technique as an independent or in the presence of weak red illumination. The strong
check of the photon fluxes previously measured using lines, 365/6 and 436 mM from a Hg source, were used for
thermopiles calibrated, in our laboratory and elsewhere, excitations, although only the former was used exten-
by means of a National Bureau of Standards radiant sively for the quantum yield determination. The yield
energy standard lamp. The ferrioxalate actinometer, as at other wavelengths can be determined by relative
reported by Parker,2 is approximately one thousand comparison with the yield at 365/6 mM.
times as sensitive as the older uranyl oxalate technique" The method is essentially a direct substitution tech-
and can be used easily in light fluxes of the same order nique, and therefore, should be independent of geometry
of magnitude as are used in the calibration of our or systematic error arising from nonhomogenous field of
thermopiles. However, recently the uranyl oxalate flux. Both a low-pressure "Osram" Hg discharge spec-
system has been made even more sensitive than the tral lampS and a collimated high-pressure "Osram"
ferrioxalate system by gas chromatographic determina- HBO 200 lamp were employed as sources with no differ-
tion of the photochemically produced carbon monoxide. 6 ence in the results. The quantum yield reported here
In our preliminary experiments, we found that the was determined with the low-pressure lamp.
photon fluxes determined with the ferrioxalate actinom- After passing through an aperture, the beam tra-
eter, using Hatchard and Parker's values of the versed a filter consisting of two centimeters of a copper
quantum yield, differed significantly from those esti- sulfate solution (10% weight/volume CuS04·5H20 in
mated with one of our calibrated thermopiles. 7 A more O.OlN H 2S04). This solution transmits mainly in the
complete investigation has revealed two sources of region 300-650 mM and, in combination with a glass
error in actinometry measurements: First. metal-black filter (Corning C.S. 1-69), absorbs completely all wave-
coatings, such as gold black and platinum black, evapo- lengths above 650 mM. The photon fluxes were measured
rated onto thermopile surfaces tend to deviate signifi- by means of a number of ac "Reeder" thermopiles9 with
t~e calibration light flux chopped at 10 cps, and also,
With an "Eppley" dc thermopile, calibrated in our
* Present address: The New England Institute for Medical
Research, Ridgefield, Connecticut. laboratory, with an NBS standard lamp, and also at
1 C. A. Parker, Proc. Roy. Soc. (London) A220, 104 (1953). the Eppley Laboratory.lo The actinometer solution at
2 C. G. Hatchard and C. A. Parker, Proc. Roy. Soc. (London)
A235, 518 (1956). distances from 50 to 150 cm from the light sources, ~as
3 A. J. Allmand and W. W. Webb, J. Chern. Soc. 1929, 1518. always irradiated in a steady light flux. A glass color
4 J. Lee and H. H. Seliger (to be published). filter (Corning C.S. 3-73 for 365/6 mM) was used to
6 W. G. Leighton and G. S. Forbes, J. Am. Chern. Soc. 52,
3139 (1930). isolate the appropriate Hg line. Both the thermopile
6 K. Porter and D. H. Volman, J. Am. Chern. Soc. 84, 2011
(1962). 6Ealing Corporation, Cambridge 40, Massachusetts.
7 Preliminary results were presented at the International Sym- 9Charles M. Reeder and Company, 173 Victor Avenue De-
posium on Photochemistry, University of Rochester New York troit, Michigan. '
March 1963 (unpublished). " 10 The Eppley Laboratory, Inc., Newport, Rhode Island.

519

Downloaded 11 Oct 2012 to 128.192.114.19. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions
520 J. LEE AND H. H. SELIGER

detector and the actinometer solution were alternated ferrioxalate system were measured for incident radiation
in position immediately behind the filters, suitably of 365/6 mp. from a low-pressure Hg lamp:
shielded from light originating from outside the solid
angle of the irradiation system. The stray radiation, as 0.OO6M
determined by the use of auxiliary, sharp cutoff,
Corning filters, was much less than 1% of the incident Ka [Fe (C 20 4) aJ 1.26±0.03 (17 determinations)
intensity.
At first, the actinometer solution (3.0 ml) was con- a.1SM
tained in a 10-X lO-mm fused silica fluorescence cuvette
Ka [Fe (C 20 4)SJ 1.20±0.03 (13 determinations).
("Pyrocell") with the entire front surface exposed to
the radiation field. This technique gave rise to the en- These values are exactly the same as those obtained
hancement of quantum yield. Later, this was modified by Hatchard and Parker using a thermopile furnished
by placing a small aperture of known area in front of by the National Physical Laboratory, although their
the cuvette thus eliminating the sides from the direct recommended values, including lower results based on a
light. Irradiations were made for periods ranging from comparison with the uranyl oxalate system, were 1.21
10-15 min in fluxes of 1-3XlOla photons/sec-cm2 (5-15 and 1.15, respectively.
p.W/cm2 at 365/6 mp.). With these irradiations, suffi-
cient ferrous ions were produced to give an optical
DISCUSSION
density (510 mp.) of the ferrous-phenanthroline com-
plex in the range 0.3-0.6. The optical density of the
A. Thermopile Blackness
actinometer solution at 365 mp. was such that the inci-
dent radiation was completely absorbed. As the photo- The thermopile provides a method of measuring
chemically induced change in ferrioxalate concentration fluxes of radiation greater than about 0.1 p.W/cm2. Ideal
was only about 0.1 %, nitrogen bubbling was found to operation of such a device would require that absorption
be unnecessary. Unirradiated, but otherwise identical, be independent of the wavelength of the incident pho-
blanks were run in matched cuvettes. The method of tons. Justification of this ideal "blackness," of course,
estimating ferrous ion produced was one of substitution. will depend on the black coating applied to the thermo-
It was based on a titration with accurately standardized pile surface, neglecting for the moment any effect of the
permanganate, and therefore, the requirement of the thermopile window. These black coatings are usually
spectrophotometer operation was only that its response thin deposits of lamp black or gold black. The optical
was linear in optical density. This eliminated systema- properties of these types of coatings have been studied
tic errors due to uncertainties in the absolute value of and show that the assumption of nonselectivity in the
the optical density measurement. Nevertheless, our visible and near infrared regions should not lead to an
measured value of the extinction coefficient at 510 mp. of error of more than a few tenths of one percent.H These
the ferrous-phenanthroline complex was within 0.5% results apply only to the bulk properties of the sub-
of that previously reported. 2 stances themselves, and not to an evaporated film on
In checking experiments, it was found that the aver- the thermopile junction surface. To make the thermo-
age beam intensity did not vary by more than ±1 % pile response rapid the coating must be thin. If it is too
over periods of several hours. The areas of the thermo- thin, the backing material can be "seen" through it
piles used (0.1 cm2) were much smaller than the 1-10 leading to spectral selectivity. It is not feasible, how-
cm2 areas presented by the actinometer solution to the ever, to determine the reflectivity of each thermopile,
beam. However, the entire area was scanned by the but, at least in the visible region, a good test is how
thermopile, and it was established that the radiation black the coating looks to the eye; any visible coloration
field was uniform to within one percent. Corrections indicates selectivity.
were made for light reflections at all surfaces. At the Unexpectedly, the problem with the thin metal-black
source-to-sample distances involved, refraction (lens) coating lies not in the visible, but in the infrared. Our
effects at the air-to-glass interface of the cuvette could thermopiles are calibrated by placing them in the field
be neglected. of an NBS radiant energy standard lamp.12 As this is
operated at a color temperature around 2300 oK, most
RESULTS of the energy lies in the infrared between 1-3 p.. There-
fore, if the calibrated thermopile is to be used to deter-
The thermopiles used in the final measurements were mine a flux of radiation in any other spectral region,
known to be nonselective. The calibration was corrected
for window absorption and, using a uniform, slightly
diverging or parallel radiation flux of intensity 1-3X lOla 11 L. Harris, R. T. McGinnes, and B. M. Siegel, J. Opt. Soc.
Am. 38, 582 (1948).
photons/sec·cm2 incident on shielded areas of 0.1-10 1J R. Stair and R. G. Johnston, J. Res. Nat!. Bur. Std. (U.S.)
cm2, the following quantum yields of the potassium 53, 211 (1954).

Downloaded 11 Oct 2012 to 128.192.114.19. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions
 QUANTUM YIELD OF FERRIOXALATE ACTINOMETER 521

there must be no difference in the reflectance of the

black coating between this region and the 1-3 p. region.
Our results show that some, but not all, gold-black
'" V
/

coatings have a "hole" in the infrared, i.e., the reflec-

tance is increased significantly, and when a thermopile
having such a coating is calibrated, it yields light flux
//~ v /

estimates for the visible region which are too high.

Slit~idth-
t-> ~ QUARTZ
CELL

For a thermopile to have good responsivity, it must // / /

be of the vacuum type necessitating a window. The
optical properties of this window must also be taken / // ~ /
into account when the wavelength region of measure-
ment is different from that used in the calibration.
Quartz, for instance, is often used for thermopiles em- lL.. 5 1.0 r--
I
-i,
0_ .9 I
ployed for ultraviolet measurement. It starts to absorb >- I-
I
I
iI
U U .8 I I
in the infrared at about 3 p. making a correction to the z::J I I

calibration necessary. This correction has been closely

w g .7
I
I
I
I
I
I
~ ~ .6 I
1
I
1
studied, and it was found, for a loS-mm-thick crystal ~ Z .5
1
1
I
I
1
quartz window, to amount to an absorption of about w 0 .4
1
1
1
1
1
I
W
8.5% for the light from an NBS radiant energy > (/) .3
_::J
I
,
10 ,
I
01
0 o
standard lamp operated at 0.35 or 0.40 A.12 We calcu- ~ ~ .2
I
I
1
1
1
1
lated the absorption in other thicknesses of fused ~ 5
.1 I
I
1
1
I
a: "- 0 I I I I I I I I
quartz by assuming that the extinction coefficient was -5 -~ -3 -2 -I a 1 2 3 4 5 mm
the same as for crystal quartz and that Lambert's law LATERAL DISTANCE FROM CENTER
was· obeyed. The reflection from the two sides of the OF CELL OF NARROW SLIT
fused quartz window, amounting to a further 6%, must
be added to this absorption to obtain the total radiant FIG. 1. Upper-scale drawing of the irradiation cell and slit
dimensions. Lower-variation in photoreduction efficiency with
energy attenuation. The 6% reflectance correction for slit position.
fused quartz in the infrared is essentially the same in
the visible region.
Potassium bromide windows are also used on some of responsivity of the thermopile. Perhaps because lamp
our thermopiles. There is no absorption correction to black and "Parson's black" surfaces are normally
make to the calibration of a thermopile having one of heavier than gold-black evaporations, they do not give
these windows as the transmission is flat out to at rise to the same disability.
least 15 p.. It should be observed, however, that the In a preliminary experiment, the facilities of the
transmission starts to decrease below 400 mp., and does Eppley Laboratory were made available to us in order
so markedly by 300 mp.. The thermopile does not have a to compare our thermopiles with one of theirs which
constant response across its sensitive area due to non- had been coated with "Parson's black." We used radia-
uniformity of coating, variable heat losses by conduc- tion greater than 700 mp. from an Hg source-filter
tion, etc. It is, therefore, essential to use and to cali- combination, for normalization of the response of the
brate the thermopile in identical radiation fields, a re- thermopile in the infrared, for comparison of the "Par-
quirement which means that these should be uniform son's black" thermopile with our own at 365/6 mp. and
fields. 436 mp.. Since it had been shown that this "Parson's
The divergence of our initial results from Hatchard black" thermopile was equally black for all wave-
and Parker's value, together with discussions with the lengths, a relative comparison in this way should give
Radiometry Section of NBS and the Eppley Labora- an approximation to the selectivity. It was found that a
tory, led us to suspect the ideality of the blackness of 1S% correction for lack of infrared blackness was re-
our gold-blacked thermopile surface. In independent quired for one of our Reeder thermopiles and about
measurements made at the National Physical Labora- 10% for our own Eppley thermopile. This latter thermo-
tory in England and the NBS and Eppley Laboratory pile was then recoated with lamp black to be used as a
in this country, both lamp black and a proprietary com- standard in our later blackness determinations. At this
pound called "Parson's Black" have been found to be, point, it should be emphasized that the reproducibility
essentially, ideally black from the ultraviolet through of thermopile calibrations between our laboratory and
about 10 p. in the infrared. In reflectivity measurements, the Eppley Laboratory was ±O.S%.
gold black has also been found to exhibit these flat We have found that the selectivity correction for a
characteristics. However, the difficulty with gold black gold-blacked thermopile can vary from zero to 20%.
apparently arises in the method of evaporation of a Our precision was such that deviation from blackness
thin film of this substance that is required for high greater than 1% could be measured. Only those thermo-

Downloaded 11 Oct 2012 to 128.192.114.19. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions
522 J. LEE AND H. H. SELIGER

vector over one or two wavelengths of light is nonzero

e
~ 4,0
and can, therefore, interact with the adsorbed mole-
X cules/a the additional wall effect would amount to

..
T3.0 approximately a 1 % increase in photoreduction. We
"- can suggest no other mechanism to explain this effect .
~

~
en
...'"
2.0

1.0

...... /
./ Areas up to 10 cm2 were irradiated in a Lucite cell.
The path length was small in this case, 1.59 mm, and a
more concentrated actinometer solution, O.l5M, was
used in order to effect complete absorption at 365/6 mJL .
10 Fig. 2 demonstrates a completely linear relationship
between rate of Fe++ production and irradiation area for
FIG. 2. Relative photoreduction as a function of area. a constant intensity beam. In this case, a slightly lower
quantum yield was obtained in agreement with Hatch-
ard and Parker's data.
piles which were black to within ± 1% were used for
the quantum yield determinations reported in this
paper. C. Sources of Error
B. Irradiation Area The absolute accuracy of Hatchard and Parker's
thermopile measurement is not known, but their pre-
In the preliminary measurements of the quantum
cision appears to be of the order of 3%. No mention
yield the cuvette was used unmasked, although well
was made of the possibility of thermopile selectivity.
shielded from stray light from any direction, and the
However, their recommended quantum yield was 4%
irradiation area was assumed to be the area presented
lower than their thermopile result, in order to take
by the front surface of the cuvette containing the 3.0 cc
into account the results obtained by comparison with
of solution, i.e., 3.0 cm2 • High results, about 1.5, were
the uranyl oxalate actinometer. The recently improved
obtained for the quantum yield. In the more detailed
sensitivity of the uranyl oxalate system should enable
experiments, the cuvette was used in a holder with an its quantum yield to be measured by the same method
aperture such that only a central portion, of total area
we have used for potassium ferrioxalate.
about 1.0 cm2, of the cuvette was irradiated. Experi- The present quantum yield measurement is subjected
ments with a central portion of the cuvette exposed with
to the following sources of uncertainty:
varying width of mask, demonstrated a sudden increase
in quantum yield as the edge of the cuvette was exposed (1) Uncertainty in calibration of the NBS
to the flux. This effect is shown in Fig. 1 which is a plot radiant energy standard lamp including the
of the derivative of results obtained as the aperture was uncertainty in the international radiation stand-
narrowed from a total width larger than the cell to one ard I2 ,14,15 1%
including only a fraction of the solution portion of the (2) Thermopile calibration 1%
cell. The ordinate is the relative efficiency of ferrous (3) Thermopile selectivity 0.5%
ion production. If there were no wall effect, the relative (4) Thermopile window absorption 0,5%
efficiency should fall to zero at the inner edge of the (5) Radiant energy estimation with thermo-
cuvette as shown by the solid dark line. If we assume pile including stray radiation effects 1%
that the area of the wall exposed to the beam is as (6) Estimation of total radiation incident on
effective as an equivalent area of solution in producing the solution considering cell wall reflection and
the photoreduction, the dashed curve of Fig. 1 would absorption, and the area of irradiation 1%
result. Our experimental points, as indicated by the (7) Spectrophotometric estimation of Fe++ 1%
black dots, are intermediate between these two extremes (8) Standard deviation of quantum yield
and indicate a surprising enhancement of photoreduc- results 1%
tion by light striking the edge of the cuvette. This edge
effect cannot be explained by the possibility of fluores- rms uncertainty: ±2,5%,
cence of the quartz sides of the cuvette as this cannot
be expected to contribute more than a few percent. CONCLUSIONS
Diffraction effects and scattered light are also negligi- Our confirmation of Hatchard and Parker's result,
ble. using a calibrated thermopile for the quantum yield of
The concentration of ferrioxalate ions in the bulk
solution is 3.6X 1018 molecules/cc. Even if we assume
that a monomolecular layer of ions is adsorbed on the JaN. J. Harrick, Ann. N.Y, Acad. Sci. 101, 928 (1963).
14 C. L. Sanders and O. C. Jones, J. Opt, Soc. Am. 52, 731
glass surface, and that, upon total reflection at the (1962).
interface between the wall and the solution, the electric 15 E. J. Gillham, Proc, Roy. Soc. (London) A269, 249 (1962).

Downloaded 11 Oct 2012 to 128.192.114.19. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions
 QUANTUM YIELD OF FERRIOXALATE ACTINOMETER 523

the potassium ferrioxalate system under the conditions ACKNOWLEDGMENTS

specified, leads us to recommend the value of 1.26±O.03
as a standard for actinometric work having an absolute The authors are grateful to A. J. Drummond of the
accuracy indicated by the rms uncertainty. Owing to Eppley Laboratory, and R. Stair and W. Schneider of
the uncertainties in the radiometry, it is not possible N.B.S. for discussions and use of facilities.
to reduce the error in quantum yield to below ±2% This work was supported by the Division of Biology
at this time. and Medicine of the U.S. Atomic Energy Commission.

THE JOURNAL OF CHEMICAL PHYSICS VOLUME 40, NUMBER 2 15 JANUARY 1964

Effects of Anharmonicity on Vibrational Energy Transfer

FREDERICK H. MIES

National Bureau of Standards, Washington, D. C. 20234

(Received 17 October 1963)

The diagonal matrix elements of the interaction potential which enter into the quantum theory of vibra-
tional energy transfer generally have been assumed to be identical. Apparently the assumption is based
on mathematic convenience rather than on any physical model. If Morse potentials are used to describe
the intramolecular forces the elements are approximately but not identically equal. The transition prob-
ability is found to decrease markedly when the ratio of the diagonal elements of the initial and final oscillator
states is allowed to deviate even slightly from one. Morse potentials were chosen to reproduce the observed
anharmonicities of a variety of diatomic molecules and an anharmonic correction factor to the transition
probability was calculated. It is found that the generally used calculated transition probabilities should be
reduced by a factor of 10-1 to 10~.

I. INTRODUCTION used to describe the collision.- Our concern in this

HE quantum theory of vibrational energy transfer paper is with a practice first introduced by Jackson
T in molecular collisions is founded on two major
suppositions. The first is that the collision may be
and Mott, and used either explicitly or implicitly in
all subsequent calcUlations, which we choose to call
described by an adiabatic electronic interaction poten- the "symmetric approximation." It consists of equat-
tial (and further it is assumed that the principal con- ing the diagonal matrix elements W nn = (n I Win) of
tributions to energy transfer come from the strongly the interaction potential W which enter into the DWA,
repulsive portions of this potential where the largest i.e., W jj = W nn , where j and n refer to the initial and
changes in force during the course of a molecular vi- final oscillator states. Thus the" average" interaction
bration are expected). The second supposition is that is assumed to be independent of the quantum state
since the observed efficiency of vibrational energy of the oscillator, which implies that the potential is in
transfer is very small the dominant term in the wave- a sense" vibrationally adiabatic." Principally this is
function is the elastically scattered wave, and there- introduced as a mathematical convenience and leads
to a simple, closed-form expression for the transition
fore only the coupling between the initial and final
oscillator states need be considered. Together these
2 K. Takayanagi, Progr. Theoret. Phys. 8, 497 (1952); J. Phys.
assumptions dictate the use of the distorted wave ap- Soc. Japan 14, 74 (1959).
proximation (DWA) . 3 R. N. Schwartz and K. F. Herzfeld, J. Chern. Phys. 22,~767 j
The DWA was first applied to the vibrational exci~ (1954) .
4 J. C. McCoubrey, R. C. Milward, and A. R. Ubbelohde,
tation of an oscillator in a head-on collision with a Trans. Faraday Soc. 57, 1472 (1961).
structureless incident particle by Jackson and Mott. l 5 K. F. Herzfeld and T. A. Litovitz, Absorption and Dispersion

Many subsequent DWA calculations have been made, of Ultrasonic Waves (Academic Press Inc., New York, 1959).
6 T. L. Cottrell and J. C. McCoubrey, Molecular Energy Trans-
but primarily they are founded on this original work fer in Gases (Butterworths Scientific Publications, Ltd., London,
and represent further refinements of the physical model 1961).
7 R. N. Schwartz, Z. I. Slawsky, and K. F. Herzfeld, J. Chern.
Phys. 20, 1591 (1952).
1 J. M. Jackson and N. F. Mott, Proc. Roy. Soc. (London) 8 E. Bauer and F. W. Cummings, J. Chern. Phys. 36, 618
A137,703 (1932). (1962). ~

View publication stats Downloaded 11 Oct 2012 to 128.192.114.19. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions