June, 1939
ELECTRICAL PROPERTIES OF MULTIMOLECULAR FILMS
that @alanineis actually a cleavage product of pantothenic acid.
Discussion.-The fact that @-alanine, as a cleavage product of pantothenic acid, may show considerable biological activity in certain media and not in others, serves to explain some of the contradictions in the literature with regard to the stability of “bios” in alkaline and acid condition. It seems likely from our findings that @alanine is of widespread importance in biochemistry and that its importance for various organisms is connected with the fact that it makes up a part of the pantothenic acid molecule. This question is being investigated more fully in this Laboratory.
[CONTRIBUTION FROY THE RESEARCH
1425
summq 1. Pantothenic acid appears to be synthesized by yeast only when @-alanineis furnished in the culture medium. 2. Several lines of evidence indicate @-alanine to be a cleavage product of pantothenic acid. It finally was isolated from this source in the form of fl-naphthalenesulfo-@-alanine. 3. The yield of 8-alanine from the pantothenic acid preparation of which an analysis has been reported, indicates that it was not more than 90% pure. 4. The significance of these findings for “bios” studies has been discussed briefly. CORVALLIS. -ON
RECEIVED FEBRUARY 21,1939
LABORATORY OF TEE GENERAL ELECTRIC COMPANY ]
Electrical Properties of Multimolecular Films BY H. H.
RACE AND
A. Introduction In 1935 we made some preliminary attempts to measure the dielectric constant and dielectric loss of 101 layer barium stearate films deposited on a polished chromium surface. Measurements of such films should be extremely interesting because of the known arrangement of organic molecules which normally can be measured only in conditions of random orientation. We immediately found three experimental limitations. First our existing bridge equipment had insufficient sensitivity and accuracy. Second, we needed more accurate measurement of the area of contact of a small drop of mercury used as one electrode. Third, the metal and water surfaces must be kept dust free while the film is being prepared. Occasional dust particles have little effect on optical measurements but cause conducting holes or cracks in the film which prevent electrical measurements. Nevertheless, the preliminary results were encouraging enough so that last year we refined the necessary bridge equipment, developed a means for accurate measurement of a few square millimeters of mercury drop contact area and built a box in which films could be deposited with a much lower concentration of dust particles. The results of recent studies on films made from a number of metal soaps, using this improved equipment, are described in this paper. (1) S. I . Reynolds and II 1%.Race, G E. Rev , 41, 52D (1938).
s. I. REYNOLDS
B. Preparation of Specimens.-Multilayer films were built by depositing monolayers on a clean polished chromium slide as described by Blodgett2 except that we took special precautions to prevent the deposition of dust particles in or on the film. It was found necessary to control very closely the acidity of the water-bath from which the various types of films were made in order to obtain good reproducibility. This was done by measuring the PH of the water with a glass electrode.a C. Apparatus and Technique of Measurements Electrodes.-The metal slide on which the multimolecular films were laid down constituted one electrode. A number of types of contacts were tried on the surface of the film for the other electrode. Using very thin pieces of metal foil, we were unable to get intimate contact without mechanically damaging the film, and the accurate determination of area was very difficult. Using drops of aqueous electrolytes, we were unable to find any of d c i e n t l y low resistivity not t o affect the dielectric loss measurements particularly a t high frequencies. A mercury drop is the only satisfactory contact which we have found. The electrode assembly is shown in Fig. 1. The chromium plated slide on which the film is built is held in a microscope stage, which is insulated from the base plate by mica strips. The mercury drop upper electrode is suspended from an amalgamated copper rod which can be lowered by a micrometer screw to m a k e contact with the film surface. The upper electrode assembly is insulated from the base plate by long Pyrex rods. The whole assembly is mounted on a heavy cast iron base to prevent vibration and changes in area of the mercury drop contact during measurements 1.
(2) K . B . Blodgett, THISJOURNAL, 57, 1007 (1936). (3) D. A. MacInnes and M.Dole, i b i d . , 69, 20 (1030).
112(i
€1.
n. RACEA N D S. I. REYNOLDS
VOl. GI
The low potential test lead is connected t o the chromium plate, and the guard lead is conncctcd to the base plate Of the film holder (see Fig. 1) with the high potential lead connected t o the mercury electrode holder. By balancing the guard and test circuits of thc hridgea the chromium plate and base plate of film holder arc brought to the same potential. thus diminating from the measurenients any possible losses in the mica insulation between the movable stage and b a w plate. At the same time the lossrs caosrtl by the glass supports from thc hirh potential electrode arc shunted to the giiardcd base plate and thus do not affrct the loss mwsurcrnents uf the film Zero meast~remcnts are made with the iiiercury drop just above the film. 5. DielehlcStrengthTests.-D. c. . lirrakdown tests are also made using thcclrctrodcsshowun in Pig. 1. Voltage may he vmictl from 0.01 v. u p and is ohtained from a fine potentionicier and I I I C B S U ~ P on ~ a high resistance v d t n ~ ~ t c ~ . For a. c. breakdown measiireinenlc thc voltage is s!rpplicd by a bcat frequency occillalor. The output voltI'ig..l age is conwolled by a potentiometer vcry hiall. Also, in order t o calculatc thr didectric cor>- on the oscillator panel and is measured with a vacuum stant from such capacitance mea~iircnients the contact tube voltmeter.e area and the thickness of the film must be measured vcry D. Experimental Results accuratcly. It was found by experience that the contact 1. Reproducibility of Dielectric Constant arca might not be circular. Therefore a filar microscope with a horizontal micrometer feed was set up on B separate Measurements.-To obtain a cleaii drop of support so that l h e drop could be vieacd from two 911' mercury as an upper electrode the amalgamated positions at an angle slightly ahovc the plane of thc film copper rod was washed with clean mercury before surfare. In each position the drop is illuminated from the rear to give a sharp silhouette in the field of the micro- each measurement. Also the surface of tlie slide scope. These diameters were averaged t o calculate the was blown dust free by a small jet of filtered air. areaof contact of the drop. If dust particles, visible under the microscope, are 3. Measurement of Film Thickness.-The thickness left between the drop and the film, reproducible per layer of various metal soap films has hcen accurately electrical measurements are impossible. With determined by index of refraction measurements.' T a o acids of dilfrrent carbon chaiii length were used t o form thc extreme care typical variations between succesfilms reporled in this paper. The thickness per layer for sive dielectric constant (e') measurements at stearic acid is 24.2 X Ill* rm., and for arachidic acid is different points of the same film are shown in 26.9 X IflFcm. Figs. 2 and 3. The standard deviatioiis are 2% The total thickness of a givcn film is obtained by miiltior less, which represents the over-all accuracy of plying these values by thr number of layers applied. To check these thicknesses, inlrrfercncc color step ~ a g e chave the area, thickness arid electrical measurements. been prepared for direct optical comparison w i t h the films Variations of dielectric constant measurements t o bc measarrd. with area of the mercury drop 011 the same film 4. Electrical Bridge Measurements.-Capacitanre and are shown in Fig. 4 and indicate no systematic 'resistance measurcmenls were made over a frequency range error. cycles pcr second by nicans of special from 40 to 1,000,011~1 Each of the above comparisons was made by giiardcd bridges.' Froni these measurements the dielectaking measurements on different areas oil tlie tricconstant (e') anddielectric 1 0 s (tan6) of the films wcre calculated from the equivalent parallrl circuit." same film. Wider variations of electrical prop2. Measurement of Mercury Drop Contact Area.--In order to work within the capacitance limits of the bridge equipment, the mercury drop contact arcn had to be small since the Capacitance per unit area of these thin films is
(4) (5)
K.E.Blcdgctt. "Built-up Films 01 Barium Stesmtc:' H.H. Race end S. C.Leonard. Elrr. En#.. 6s. 1347 (IllBal.
erties are often found on different films of the same (61 General Radio Type 726A.
June, 1939
-i
ELECTRICAL PROPERTIES OF MULTIMOLECULAR FILMS 1
1
i
x
2.4
(Y) layers of barium stearate. We think that the difference between these two sets of data does not result from differences in the direction of orientation but rather b a t it is a function of the PH of the bath from which the films were made since we know that the proportion of soap to acid in the films varies considerably with changes in hydrogen ion concentration7 of the bath.
'
2.6--iT i I
.
,
-"
I
a
)I
I
x I
x t
x
/
-
2.7,
1427
2.7 I -* 2
.,1
I
I
5
I
I
I
I
av
144
23
60
100 140 180 Molectlar layers. Fig. 6.-Variation in E' measurements on different thicknesses of (X) and (Y) barium stearate films (buffered): 0 for Y-films--pH 6.9, bath lo-' M BaC12.2Hz0 and 2 X 10-4 M KHC03; X for X-films--pH 9.0, bath lo-' M BaCl2.2H10and 3 X 10-3 M KHCOs and 2 drops per liter of concd. NHIOH.
During the latter part of this work the films were taken off from buffered solutions so that no 2.5. Q ---_Po4 Q changes should occur in the constitution of sucw 0 0 2.4 cessive layers. Such changes may account for some of the variations observed when films were 0.02 0.04 0.06 built up from unbuffered solutions. Contact area, sq. cm. 2. Dielectric Constant Independent of ThickFig. 4.-Variation in E' measurements ness.-Figures 5 and 6 show that within the with area of mercury drop on a cadmium reproducibility and accuracy of our measurements arachidate film: av. e' = 2.46, u = 0.08%. the dielectric constant is independent of the film that, in general, there is better agreement betwcen thickness between 51 and 181 layers. We have measurements taken a t different spots of the same additional data corroborating this statement for film than is found between different films. It was films varying from 21 to 201 layers. This indifound that control of the acidity of the bath was cates also that, when properly dust free, there is very important not only in the building of multi- no air between the film and the mercury drop. layer films but also in the electrical properties o f 3. Dielectric Constant Independent of Fresurh films. For the niorc recent tests the #H o€ quency.--The dielectric. constant of -a number the bath is recorded on the data sheets. In Fig. of films was measured within the frequency range (i are shown dielectric constant data for (X) arid from 40 to IO" cycles per second. The data are plotted in Figs. 7 , 8 and 9. Again, within the reproducibility and accuracy of our measurements, the dielectric constant is independent of frequency up to IO6cycles per second. 4. Dielectric Loss Independent of Thickness and Frequency.-Each measurement of dielectric constant reported above was accompanied by a diFilm no. 1 2 3 4 5 6 electric loss measurement made simultaneously. Layers 101 61 61 101 101 101 6.12 ? 6.2 7.7 These represent independent properties of the PH I 6.8 Fig. 5.-Variation in E' measurements on differfilms and are plotted against frequency in Figs. ent films of cadmium stearate: bath lo-' M 10, 11 and 12. The losses show larger deviations -d
CaCL.2H20 (not buffered), av. 4.3%.
e'
= 2.59, u =
(7) I . Langmuir and V (19363
J
Schaefer, THIS J O U R N A L . 68, 284
H. H. RACEAND
1428
s. 1. REYNOLDS
Vol, 61
the two sets of data are caused by the electrolyte in which their films were immersed.
Fig. 7.-Variation in E' with frequency on copper stearM CuC12.2H20, 0 ate films (not buffered): bath for 61 Y-layers with pH 5.38, X for 101 Y-layers with pH 4.68. 4 5 6 Log,, (f) cycles per sec. Fig. lO.--Variation of tan 6 with frequency on copper M CuC12.2Hz0, 0 101 Ystearate films: bath layers a t PH 4.68, 1361 Y-layers a t PH 5.38.
3
----CI-.
3 4 5 6 Loglo (f) cycles per sec. Fig. 8.-Variation in e' with frequetlcy ofl cadmium M CdC12.2H20, stearate hlms (not buffered): bath p&I 6.0-7.7, l3 101 layers on 3-338, 0 61 layers on 4-25-38, X 61 layers on 5-2-38, @ 101 layers on 6-27-38. 2
w' 0.11010 -f--
Q
3 4 5 6 Log10 (f) cycles per sec. Fig. O.-Varistion in d with frequency on cadmium arachidate films (not buffered).
2
between measurements on the same film or on different films than was found for the dielectric constant. However, these deviations appear to be random and show no dependence upon either thickness or frequency. The ordcr of magnitude of tan 6 is low, being usually less than 0.001. The accuracy and sensitivity of the bridges uscd are considerably better thaii this value,' so that we believe the films are electrically non-uiiiforin to this extent. The dielectric loss measurements on dry multilayer films are of a much smaller order of magnitude than those obtained by H a s k i d and his coworkers. Assuming that they reported equivalent parallel resistance and capacitance of the films, their data indicate a two-fold variation in dielectric constant and a 15-fold variation in tan 6 as functions of the number of layers, and the order of magnitude of tan 6 is from 1000 to 10,000 times higher than the values reported in this paper. We believe that the wide differences in ( 8 ) Buc'kwald, Gallagher, Haskins, Thatcher and Zahl, Pror S a l Acnd Sci , 24, 204 (1938).
o
3 4 5 Loglo (f) cycles per sec. Fig. 11.-Variation lh tatl 6 with frequency on cadmium stearate films: av. t a n 6 = 0.00093 and u = 17%, bath lo-' Y CaCl~.lHzO. 2
Symbol
Layers
0 A
61 101 101 59
El
X
6.12 ? ?
5.20
-
-
3 4 5 ti Log,, (f) cycles per sec. Fig. 12.-Variation in tan 6 with frequency on cadmium arachidate films: bath lo-* M CdC12.2Hz0 (not buffered). 2
Symbol
Layers
9H
0
61 61 61 93 101 143
4.92
X Q
l3 X
+
5.92 5.92 5.62 4.92 5.62
June, 1939
5. Attempt to Use Glass Slide with Evaporated Chromium Surface.-In an attempt to get a highly polished metal surface on which to form the films we tried microscope slides on which a coating of chromium had been obtained by vaporization. It was found that metal films thin enough to adhere to the glass without peeling or cracking had an electrical resistivity which was too high to permit high frequency dielectric loss measurements, as indicated in Fig-. 13. 6 . Effects of Ske1etonization.-Langmuir and Schaefer7 have shown the dependence on the hydrogen-ion concentration of the bath of the conversion of fatty acids to metallic soaps. For example, if a film of cadmium arachidate which has been built a t PH 5.62 is soaked for one to ten seconds in a solvent such as benzene and is then taken from the benzene, the color of the soaked film is much different from that of the original film. The change is due, according to Blodgett aiid L a n g m ~ i r to , ~ a large decrease in the refractive index of the film, the actual thickness being 11naltered. The benzene dissolves the arachidic acid, leaving cadmium arachidate as a skeleton with air filling the spaces previously occupied by the arachidic acid. For this concept of the structure of a skeletonized film the remaining soap can be considered to be paralleled by the air spaces, and the dielectric constant of such a film is given by the relation 6'1
1429
ELECTRICAL PROPERTIES OF MULTIMOLECULAR FILMS
=
€'O(dl/dO)
+ e',(l
- dl/dO)
(1)
where 6'0 = dielectric constant of unskeletonized film, d1 = dielectric constant of skeletonized film, e l g = 1 = dielecti-ic constant of air, and (dJd0) = ratio of densities of skeletonized to unskeletonized films. Therefore calculations of E' can be made on the above assumptions for comparisoii with actual measuremeiits on films after skeletonization. In order to do so it is necessary to convert the observations of changes in apparent thickness caused by skeletonization to changes in density in order to determine the proportion of the volume of the skeleton occupied by air. Such calculations have been made by BlodgettlO and the results are given in Table I. The proof that a skeletonized film did not collapse lay in the fact that the original optical thickness was restored by allowing a drop of oil to run over the film and thus refill the spaces originally (9) K B Blodgett a n d I I-angrnuir, P J ~ YRsc u , 51,980 (1937) (IO) K B Blodgett, "Use of Interference t o Extmgulsh Reflection of I i g h t f r o m Glass" (not yet puhlished )
3 4 Loglo (f) cycles per sec. Fig. l3.-Effect of using vaporized chromium surface on glass slide for film base: bath M CdCI2.2H20. Ordinate
Symbol
Layers
pH
tan d ian 6
A 0
E'
+
45 81 45 81
5.75 5.63 5.75 5.63
6'
X
occupied by acid molecules. After this refilling the dielectric constant can again be measured and a check calculation made by substituting for E', in Equation 1 the dielectric constant of the oil. TABLE I FILMS FROM APPARRELATIVE DEXSITYOF SKELETONIZED ENT THICKNESS (AXGLE OF INCIDENCE = 80'; T O = 1 50) ? ! ! ?O
-1
1 0 0 9 8
7
. (i .5 4
Ratio of apparent Ratio of densities thickness after and after and before before skeletonization skeletonization (11 0 0 ) (di /do)
1 0 0 9404 8793 8159 7500 6804 6058
1.0 0 9136 .8242
7317 63*58 5.169 4340
Table I1 shows the data obtained on several films of cadmium arachidate made a t PH 5.7, with check calculations made as indicated above. Columns (1) and (2) give the apparent thickness in number of layers before and after skeletonization obtained by visual comparison with step gages. Column (4) gives the ratio of densities as determined from a plot of the data in Table I. Columns (5) and (6) give the measured dielectric constants before and after skeletonization, calculatd on the basis that the actual thickness
H. H. RACEAND S. I. REYNOLDS
1430
Vol. 61
TABLEI1 ELECTRICAL DATAON UNSKELETONIZED (0) AND SKELETONIZED (1) FILMS OF CADMIUM ARACHIDATE (1)
(3)
(2)
(4)
Optical thickness
(5)
t'o
€11
(7) .'I
(9)
(%led with non-polar oil C'Z
e'(
lo
tl
(tl/b)
(&/do)
Measd.
Measd.
Calcd.
Measd.
Calcd.
A
75
53
0.707
0.574
2.48 2.48 2.47 2.49
1.69 1.74 1.71 1.87 1.73
1.85
2.29 2.20 2.19
2.34
B
61
43
.705
.571
2.44 2.43 2.47
1.62 1.68 1.69
1.82
2.25
2.31
C
61
51
.837
.764
2.50 2.50 2.50
1.96 1.94 2.02 1.97
2.15
2.32 2.38
2.42
D
59
61
.86
.797
2.49
2.10
2.18
e
Y h
0
L4
5C
M
l
o
- --
:;
3()()()-
c.'
;
u u-
.*
D /
