Films Built by Depositing Successive Monomolecular Layers on a Solid Surface by KATHARINEB

June, i935 ~EPOSITIONOF SUCCESSIVEMONOMOLECULAR LAYERS 1007

2. In addition to showing the above change, sulfuric acid or stannic chloride is not due to the xanthone and lluorenone possess new bands in ether oxygen. sulfuric acid sohtions which are believed to be 4. The color of ketone chlorides and sulfuric associated with the stabilization of the quinonoid acid or stannic chloride is of a different nature structure for these compounds. Solutions of from that of ketones and resembles that of tri- stannic chloride produce the same changes on the arylcarbinols and salts of these carbinols. It is absorption spectrum of xanthone that sulfuric postulated that in such solutions the ketone acid does. chlorides exist in a quinonoid modification. 3. The production of color with xanthone and ANN ARBOR,MICH. RECEIVEDMARCH 28, 1935

[CONTRIBUTION FROM THE RESEARCHLABORATORY, GENERALELECTRIC COMPANY ] Films Built by Depositing Successive Monomolecular Layers on a Solid Surface BY KATHARINEB. BLODGETT’

A previous paper2 has described briefly a 5.0, approximately, the Ca ions in the water com- method of depositing successive monomolecular bine with the carboxyl. group of the adsorbed layers of “stearic acid” on glass. Subsequent molecules, converting the film to calcium stear- experiment^,^ which will be discussed in this ate. The film may be neutral soap or an acid paper, have shown that the substance deposited soap, depending on the pH of the water. The on the glass was calcium stearate and not stearic term “acid soap” is used here in the sense in which acid. The layers were deposited one at a time; it is used by McBain6 and other writers to refer thus a film could be built having a thickness of 1, to compounds of the fatty acid and the neutral 2, 3 or more layers of molecules. Further de- soap. Both types of soap will be called “calcium velopment of the method has made it possible to stearate” in this paper. deposit more than 200 layers on glass and on The film is transferred from water to glass’ by various metals. raising a clean glass slide slowly out of the water This paper will describe experiments in which on which the film is spread. A constant surface the thickness of films containing many layers was pressure is maintained on the film as the glass is measured by means of the interference of mono- raised through the surface. If the water-bath chromatic light reflected by the films. These is alkaline (PH > 7.5), the stearate molecules measurements were made with films of calcium show a ready tendency to attach themselves to or barium soaps of long-chain fatty acids deposited glass, so that the surface pressure causes them to on glass having a high refractive index. Films can be shoved off the water and onto the glass. Lang- be built with fatty acids4 when deposited on a muir, Adam and others have demonstrated that metal, but fatty acids do not adhere readily to glass. the fatty acids tend to orient themselves on a A monomolecular calcium stearate film is water surface with the -COOH “head” of each formed on water by placing a small amount of molecule in contact with the water and the chain stearic acid, dissolved in benzene, on a clean sur- tilted at a more or less steep angle with the plane face of water containing calcium salts. The of the water surface. In the present experiments stearic acid spreads rapidly in a monomolecular the monomolecular layer of calcium stearate is film.6 If the PH of the water-bath is greater than similarly oriented and the layer is transferred to (1) The experiments described in this paper were commenced in glass with the (-C00)2Ca heads attached to the collaboration with Dr. Irving Langmuir and were continued while he was traveling in the Orient. The writer is indebted to Dr. Langmuir glass, and the paired chains perpendicular to the for urging her to develop further the method described in the previous plane of the glass. The exposed upper surface paper, and for contributing many important suggestions which have been included in this paper. of the film is composed of closely packed CHs (2) K. B. Blodgett, THISJOURNAL, 56,495 (1934). groups. (3) I. Langmuir, J. Franklin Insf.,l18, 153 (1834). (4) G. L. Clark, K. K. Sterrett and P. W. Leppla, THISJOURNAL, (6) J. W. McBain, M Taylor and M. E. Laing, J. Chcm. Soc., 121, 67, 330 (1935). 621 (1922), and other papers. (5) (a) I. Langmuir, ibid., 39, 1848 (1917); (b) N. K. Adam, (7) I. Langmuir, Trans. Faraday Soc., 15, 62 (1920): reprinted in Proc. Roy. Sor. (London), A99, 336 (lYZlJ,and succeeding papers. G. E. Rev., 24, 1025 (1921). 1008 KATHARINEB. BLODGETT 1‘01. 57

A film in which all the molecules are oriented glass is being raised. Dr. Langmuir pointed out with the CHB groups at the upper surface pos- that pressure could be most conveniently ap- sesses the exceptional property that neither water plied by placing a tiny drop of castor oil or oleic nor oil will wet the surface. The property of acid, or some other hydrophilic oil, at one end of shedding water is an important factor in the pres- the trough. A hydrophilic oil has the property ent experiments. As a glass slide is slowly raised that the molecules of the oil spread on water until out of the water on which a film is spread, the they have covered all the available area with a water recedes from the glass wherever the film monomolecular layer, crowding together into this attaches itself to the solid surface, and if the slide layer until they have established an equilibrium be withdrawn from the water fairly slowly (5 to surface pressure.8 The equilibrium pressure ex- 10 cm. per minute), it will emerge completely dry. erted by an oil is a characteristic property of each This enables the experimenter to regulate the individual pure oil; for example, oleic acid exerts speed at which he lifts the slide so that the stearate approximately twice as much pressure as castor film may be deposited uniformly over the entire oil. After the surface is covered with a mono- surface. molecular layer, the surplus oil remains gathered The speed with which a film attaches itself to in many small lenses. If, however, new space glass is governed by three factors, @H, calcium becomes available on the water surface, oil spreads concentration, and temperature of the water- instantaneously from these lenses and covers the bath. The speed increases with increasing values new area also with a monomolecular layer. Thus of one or more of these variables. Thus at 20’ a a drop of oil placed at one end of a trough on solution nearly saturated with calcium carbonate, which a calcium stearate film has been spread, 5 X molar, gives films which coat glass with acts as a surface piston to hold the stearate film fair speed at fiH 7.5, and with greater speed at under constant pressure at every instant. Oil pH S.0; whereas a solution 1 X molar cal- used for this purpose is called “piston-oil” and its cium carbonate needs to have a value of @Habout use is illustrated in Fig. 2. 0.5 higher in order to give the same results. The After glass has been coated with one layer, speed in every case is greater at 35 than at 20’. additional layers can be added one at a time by dipping the glass in and out of the water-bath on which a monomolecular layer of calcium stearate is spread. The successive operations will be described in the following section.

Apparatus and Technique The process of depositing the layers is carried out in a

/ // trough of the kind developed by Langmuir and by Adam Y for studying films on water. The trough T, shown in Fig. 1, is filled with water to the brim. It stands in a water- w bath D, the temperature of which is controlled. The Fig. 1.-Apparatus for transferring a monomolecular film edges and inner surface of the trough and the metal barrier from a water surface to a solid surface. B are coated with paraffin wax. Three detachable strips of glass K serve as a pen within which the benzene solution When the @Hof the water is less than pH 7.5, of stearic acid is placed on the water surface, to prevent approximately, the film shows less tendency to the benzene from reaching the waxed edges. detach itself from the water and attach itself to The glass slide G is raised and lowered by means of a the glass, and at PH < 7.0 it has very little ten- lever L operated by a hand windlass W. A simple me- chanical device of this sort is needed for the reason that it is dency to coat the glass in this manner. However, difficult to lift the slide slowly by forceps held in the hand excellent films can be deposited from water at pH without an occasional backward “hitching” motion. 6.0 to 7.0 by withdrawing the glass from the bath Hitching must be positively avoided since a 6lm is de- completely wet, the surface of the water on the posited each time the slide is raised and lowered. A glass being covered with calcium stearate, and pinch-clamp at L holds the rod H and a similar clamp J at the lower end of H holds the slide. The slide and the then removing the water from beneath the film clamp which hold it are both thoroughly cleaned in a hot by drying the slide slowly at a low heat. solution of concentrated sulfuric acid and potassium di- In each of these methods it is important that (8) A. Cary and E. K.Rideal, Proc. Ro)’. SOC.(London), 109A,301 the film be held under constant pressure while the (1925). June, 1935 DEPOSITIONOF SUCCRSSIVEMONOMOLECULAR LAYERS 10#9 chromate, washed in running water, rinsed in distilled erted by the piston-oil outside the fence, the spreading of water and lowered at once beneath the surface of the water piston-oil ceases. A thread & is then laid in the position in the trough. The clamp holds the slide faced in such a shown in the diagram as a double safeguard, so that if direction that the surface on which the iilms are to be de- piston-oil accidentally passes S1 it will remain in the area posited is perpendicular to the water surface and to the between SI and Sg. The process of depositing layers of length of the trough. calcium stearate on the slide G is then begun; the details of Figure 2 illustrates the method used to keep the area this process are described in the section entitled “Suc- occupied by piston-oil fenced off from the area occupied cessive Layers.” Each time one layer is deposited the film by calciutn stearate, by means of the waxed silk thread SI. is depleted by an amount equal to the total area of the This precaution is necessary for the reason that if piston- slide, front and back, and the thread is pushed forward a oil is permitted to reach the slide G, and if one layer of corresponding amount by the piston-oil until nearly a11 of this oil is accidenta:lly deposited on the slide, it will usually the calcium stearate is transferred to the glass (Fig. 2(e)). prevent the slide from taking up any more layers of calcium Since any dust or fuzz particles which fall from the air upon stearate. the water are transferred with the amto the slide, a glass The surface of water in the trough T is first cleaned by cover is laid over the trough after the threads are in posi- sweeping the metal barrier B over the surface. When tion, the cover being supported by two movable metal much oil is present on the surface, the oil that is swept bridges not shown in Fig. 1. against the waxed edges of the trough tends to adhere to the wax in spots ,and causes subsequent trouble. This a CALCIUM STEARATE FILM may be avoided by placing a strip of paper underneath B. ezr PisroN OIL FILM The strip is made wider than B so that it extends 6.3 mm. in front of B and takes up oil as it advances. The grade 8 of paper commonly used for scratch-pads works well. A waxed edge which has become oily may be cleaned toler- ably well by wiping the wax with clean wet ater paper. In order to test the water surface for the presence of oily films after it has been swept by B, lay a silk thread in r, a large circular loop on the surface at one end of the trough (not shown in Fig. 2) and blow on the thread. If the surface is clean the thread can be easily blown from one end of the trough to the opposite end. The amount of oily film remaining after a surface has been cleaned should Cii c Cover not more than 10% of the total area. The extreme precautions which are required to reduce this contamina- 0 tion to one per cent. are impractical in a trough in which Fig. 2.-Use of a waxed thread to mark the boundary of piston-oil is used constantly. Therefore the following a monomolecular surface am. procedure has been: devised whereby the residual film is prevented from contaminating the calcium stearate film. The oleic acid that was used as piston oil was purified A small metal clip is tied to each end of a waxed silk by fuller’s earth which was mixed with the oleic acid for thread S,. The thread is waxed by being soaked in a thirty minutes at the temperature of boiling water and was benzene solution of paraffin wax, and dried; it must not then separated in a centrifuge. The oleic acid exerted take up so much wax that it becomes stiff. After the a surface pressure of 29.5 dynes/cm. at 20”. Cary and water surface has been cleaned and the slide G has been Rideals found that the pressure of oleic acid (spread lowered into the water near one end of the trough, these on I/IOO N HCl) was nearly constant over moderate clips are attached to opposite edges of the trough at C ranges of temperature. Their measurements show a linear and the thread is laid on the surface in the position shown decrease in pressure with increasing temperature, of an in Fig. 2(a). Stearic acid dissolved in benzene is placed amount 0.07 dyne/cm. per degree C. The pressure of on the surface near G, and as the stearic acid spreads it castor oil was found to vary considerably among various pushes the fence before it, as shown in Fig. 2(b). After samples of the oil; the sample that was used exerted 16.5 spreading has ceased, the thread is fastened to the edge dynes/cm. at 20”. These pressures were measured by at points midway along the film by two small copper means of a surface pressure balance. hairpins F (Fig. 2(c)). The determinations of fiH values were made by means A tiny drop of piston-oil is now placed on the surface of indicators, the colors being matched to La Motte color at P which breaks up into many tiny lenses surrounded standards. Small samples of water taken from the trough by a monomolecular film that presses against the thread- were used for making these determinations. fence and causes the calcium stearate am to contract In the case of alkaline solutions, sodium bicarbonate was slightly in area (Fig. 2(d)). Calcium stearate forms the used as a buffer in a concn. ~/~ODOmolar, and sodium hy- type of film known as a “condensed” film, that is, it con- droxide was added when necessary to obtain the desired tracts only slightly as the external surface pressure is fiH. Tests made at @H9.0 showed that when sodium bi- increased from zero pressure to 30 dynes/cm. As soon as carbonate and sodium hydroxide were added to a solution the calcium stearate has contracted to an area at which the 1/1~,~~~molar in calcium carbonate in an amount which pressure within the thread-fence equals the pressure ex- made the concentration of Na ions l/200 normal, the pres- 1010 KATHARINEB. BLODGETT VOl. 57 ence of the Na ions in this concentration did not interfere placed on water that is slightly warmer than the oil One with the processes of depositing layers of calcium stearate should allow about one minute's time before judging the from the surface of the solution. When the hTa concen- spreading of a small lens in order to allow the lens to tration was increased to 1/1~normal, a first layer was become 4aturated at the water temperature. found to be somewhat reluctant to adhere to glass; at I/~Oiiormal the reluctance was very marked. The re- Successive Layers luctance was probably due to the fact that sodium stearate does not attach &?elf to glass, i. e., films cannot be de- Two distinct series, P and X, of successive posited from alkaline solutions containing Na but no Ca. layers of calcium stearate can be deposited, their Samples of stearic and palmitic acid from various sources occurrence depending mainly on the PH and tem- were tested for purity by a very sensitive method described perature of the water. in a previous paper.2 The stearic acid that was used for Y-Films Built of "Alternating Layers."-A the measurements given in Table I was prepared by Dr. layer of a Y-film is deposited each time glass is R. E. Burnett in this Laboratory, the palmitic acid was given by Dr. E. Emmet Reid. The arachidic acid that was raised through a surface layer, and also each time used was not pure; the melting point was 70". it is lowered through the surface layer. The jirst The test for purity of stearic or palmitic acid consisted layer must be deposited on the first up-journey, in determining the temperature at which mineral oil as previously described; the second layer is de- saturated with the fatty acid in question would spread on posited on the next down-journey; the third on '/IOC N hydrochloric acid. Cary and Rideals have shown that a pure fatty acid lowers the surface tension of acid the next up-journey, and so on. After a slide has water only if the temperature of the water is above a criti- been coated with an initial layer, which must cal value, the critical temperature increasing with the usually be deposited rather slowly, the succeeding length of chain of the fatty acid. In the present method layers may be deposited at a rate ten to fifteen one per cent. of fatty acid was dissolved in Squibb's per minute on a slide 2.5 cm. in height. Liquid Petrolatum wax by warming the Petrolatum slightly. The solution was then cooled to 25' in the case The layers of odd number, U', are oriented in of stearic acid and to 10" in the case of palmitic acid at the opposite direction from layers of even num- which temperatures the greater part of the fatty acid ber, Y", for the following reason. The first layer crystallized. The solubility of stearic acid in Petrolatum is oriented with the (-C00)2Ca "heads" next the is 0.1 6 per cent. at 25'. glass and the CH, groups away from the glass. A drop of the mixture of oil and stearic acid crystals was As stated in the earlier paper, when glass coated placed on a clean surface of N hydrochloric acid at 25", and if the drop remained in a lens having a diameter less with a first layer is next lowered into the water, than 1 cm. the temperature of the acid water was raised the film on the water surface attaches itself to the and new drops were tested at successive temperatures until slide, but, since water makes a contact angle of the critical temperature was reached at which drops approximately 90' with a CHs-surface, the film commenced to spread to disks of diameter much greater than 1 cm. If the temperature was raised 1" above this is turned upside down as the slide carries it down critical temperature, the increase in spreading was very into the water. The phenomenon is striking when great, a single drop spreading out to cover many hundred the motion of the film is observed by means of square centimeters. If stearic acid contained a trace of scattered talc particles. This film remains at- palmitic or other lower fatty acid, the critical spreading tached upside down to the underlying layer, that temperature was lowered, usually by several degrees. is, with the CH3-groups toward the glass and the The purest stearic acid that has been tested commenced to spread between 28 and 29", the purest palmitic acid be- heads away from the glass, and forms the second tween 12 and 13". The method has the advantage that layer of the series. As the glass is raised from the several samples of a given fatty acid may be tested simul- water the t,hird layer attaches itself with the heads taneously on the same water surface: the sample which adjoining the heads of the second layer and the spreads the least contains the least amount of impurity CH?-groups away from the glass. All layers of consisting of lower fatty acids. The test does not reveal the presence of higher acids. odd number are oriented similarly to the first, Caution.-The theory on which the preceding test is with the heads toward the glass, and all layers of based requires that the Petrolatum solution shall be satu- even number similarly to the second with the rated, and shall not be supersaturated. If the solution heads away from the glass. This type of alternate be supersaturated, it will spread at a temperature lower orientation of successive layers is usual in fatty than the critical temperature, and the observations will be deceiving. The crystallization of fatty acids from Petrola- acids. M~ller~*~~and others have shown by x-ray tum takes place slowly; therefore the oil should always be analysis that the molecules in crystals of fatty acids cooled to a temperature at least 3" lower than the critical are oriented end to end, in opposite directions. temperature. Since the oil mixture is full of tiny crystals, ('1) 4 hIiiller, J Lhein 5oc 123, 2043 (1023) it becwniw ?attirated rapidly when a drop of the mixture is (101 h Miillrr uiid C; bhedrrr, rbiri US, (156 (1923) June, 1935 DEPOSITIONOF SUCCESSIVEMONOMOLECULAR LAYERS 101 1

Since the molecules in the Y”-layers, deposited was prepared in the trough and air was bubbled as the glass travels downward into the water, are through it to remove the excess carbon dioxide oriented with the heads away from the glass, until the PH rose to the value of 6.4. The bath these layers adhere tightly to water. This prop- was maintained at PH 6.4 to 6.6, at 22 to 33”. erty can be demonstrated by the following test. When this solution is in equilibrium with the pres- An even number of layers, such as four layers, sure of carbon dioxide contained in ordinary air are deposited 0x1 a slide and after the down- the PH is 6.9, approximately; therefore, a small journey on which the fourth layer is deposited amount of carbon dioxide was passed through the slide is kept under water while the water the bath about every forty-five minutes to main- surface is swept clean. The slide is then raised tain the desired value. The reason for preferring through the clean surface. If the calcium con- the lower value was that at PH 6.9 the film on the tent and $H of the water are high, glass coated water surface crumpled wherever it was rudely with two or four layers emerges from the water broken, so that as the glass slide rose and fell completely wet. If the number of layers is an through the surface broken bits of the film gradu- even number higher than four, say 10 layers, the ally gathered on the glass, giving it a fogged, glass emerges dry, shedding water as it rises. soapy appearance. At pH 6.4 this did not occur Close examination shows, however, that the 10th when castor oil was used as piston oil. When layer does not remain attached to the glass in this oleic acid was used, which exerted a higher pres- case, but is ripped off the glass by the adhering sure and had therefore, a greater tendency to pro- water and returns to the surface of the water- duce a crumpled film, it was found advisable to bath, leaving the slide coated with 9 layers. This lower the PH to 6.2 or 6.0. is readily shown by sprinkling the clean water When this solution is used the first layer must surface with talc or with sulfur before the slide be “dried on” the glass in the manner previously is withdrawn. As the slide rises the particles described. Subsequent layers, however, are de- recede from the slide, whereas if perfectly clean posited very readily, so that on the second ascent, glass be withdrawn through a clean water surface, as the third layer is deposited, the glass emerges the particles do not recede. If the calcium con- from the surface shedding water nearly as rapidly tent and PH of the water are not high, but have as though the glass were coated with paraffin wax. the values commonly used for depositing Y-films There is plainly a marked difference between the which will be given in this paper, the layers of forces which attach a film to underlying layers of even number almost always return to the surface calcium stearate and those which attach it to of the water-bath when this surface is not covered clean glass, since a film will not detach itself from with a surface filrn under pressure. When, how- water at PH 6.4 in order to attach itself to glass ever, the water is covered with a surface film under (unless the water be removed by drying), but pressure, the even-numbered layer remains on the attaches itself readily to glass coated with one or glass and the next odd-numbered layer is de- more layers of calcium stearate. posited on top of i,t as the glass is raised. There- X-Films Built of “Non-Alternating” Layers.-- fore films built of Y-layers are considered to exist These layers have the property of being trans- out of water only in odd numbers of layers. ferred to the glass slide only as the slide travels Y-Layers are the type of layers which are com- downward into the water. As the glass is raised monly deposited and are most easily investigated. out of the water no layer is deposited. The They are the only type that have been observed question of whether a layer is deposited or is not in the case of barium and magnesium stearates. deposited is decided by watching the motion of With calcium stearate they occur on water at the thread which encloses the calcium stearate $H < 7.0, and often occur at pH = 7.0 to 8.5. film on the water surface (Fig. 2). The slide The solution wbich was used for depositing the sheds water perfectly in either case as it rises layers from which the data of Table I were through the water surface, but if the slide receives obtained was l/lo,ooo molar calcium carbonate. a layer of calcium stearate as it travels upward, It was made from a more concentrated solution of the thread moves toward the slide, if it receives Ca(HCO& obtained by bubbling carbon dioxide no layer the thread does not move. through a mixture of 0.5 g. of calcium carbonate X-Layers are obtained from films spread on per liter of water. The l/lo,ooa molar solution alkaline water containing calcium or strontium. 1012 KATHARINEB. BLODGETT VOl. 57

A solution of l/lo,ooo molar calcium carbonate, better at PH < 6.8. It may usually be avoided pH 9.0, gives layers of this type at room tempera- with X-films by working at pH > 8.5, and by ture. Other values of PH and temperature at using fresh solutions; but the conditions which which these layers have been investigated will be govern the X-series are less clearly defined than discussed in a later section of this paper. those which govern the Y-series. For example, Layers of this type are not deposited next to the water which has been in an open trough for two or glass, for the first 5 or 10 or more layers occur as three days has frequently ceased to give X-films alternating layers even on alkaline water. After at any PH, but when the trough was cleaned and these first layers are deposited the film com- filled with a fresh solution, X-films were obtained mences to “slip” on the ascent of the glass, that is, at once. On the other hand, three-day-old solu- it fails to attach itself to the glass as the glass tions have often given perfect X-films. Presum- rises through the surface. When this occurs the ably certain substances present in the air or in water receding from the slide commences to form the walls of the trough enter the water and affect a noticeably steeper contact angle with the glass its properties with regard to X-films. than when the water recedes depositing a film as it goes. Measurement of Film Thickness Under suitable conditions X-layers may be de- Films composed of several layers of molecules posited as regularly as Y-layers, a hundred or more exhibit interference colors when deposited on consecutive layers being deposited on descents and glass, provided that the refractive index of the none on ascents. This means that each layer de- glass is sufficiently different from the refractive posited on a descent acquires the property of index of the film so that light is plentifully re- shedding water before the glass rises from the flected from the optical boundary between the water. Therefore, the molecules in these layers film and the glass. The glass used in the present must acquire end-groups which have a very experiments had a refractive index 1.64 and iilms different composition or packing from the mole- on this glass showed vivid color. A film contain- cules deposited in the case of the Y series, for it ing only 9 layers can be seen to show a faint was shown previously that each YK-layer ad- yellow-brown color, due to the diminished inten- heres tightly to water until the YN-layer takes sity of blue light. As the number of layers is in- up a Y’-layer which sheds water. Some experi- creased the color of the film changes, correspond- ments will be described in a later section which ing to the change of the absent color to the suc- afford evidence that the molecules in an X-layer cessive longer wave lengths green, yellow, red. undergo a change after the layer has become at- As still more layers are added the spectrum is re- tached to the slide. peated. A slide coated with Y-layers taken from a solu- A method has been developed for recording the tion l/lo,oo~ molar in calcium carbonate at PH < successive stages by depositing on a single slide 7.0 will immediately take up X-layers if the PH “steps” of 21, 41, 61, etc., layers, or any other be changed to 9.0, and will resume the Y-series chosen sequence. The steps are deposited by if the $H be restored to PH < 7.0. At interme- covering all of a clean wet slide (I), except 0.5 diate values of $H, particularly at PH 7.0 to 7.5, cm. length, with another clean wet slide (11) either type of iilm may be deposited. For ex- having the same dimensions. After 20 layers are ample, in this range one may deposit 40 or more deposited the covering slide (11) is moved 0.5 alternating layers, and then for no apparent rea- cm. while the slides are under water, so that it son the films commence to slip on one part of the then covers all but 1.0 cm. of (I), and the next slide and one obtains non-alternating films over 20 layers are deposited on both the first and the that area for the next 10 or 20 layers, and then second areas. The colors of the steps are made the alternating process may be resumed; in the to appear much more vivid by pressing a slide of meantime the remainder of the slide will have wet cobalt glass against the back of (I) while the continued taking up alternating films without films are being deposited on the front of (1). interruption, so that one obtains a slide coated The slide (1) then forms the center of a sandwich, with a greater number of layers in one place than between the cobalt glass and the cover glass (II), in another place. This difKculty may be avoided all three being held together by the clamp J (Fig. in the case of Y-films by working at PH < 7.0, or 1). The sandwich should be assembled under June, 1935 DEPOSITIONOF SUCCESSIVEMONOMOLECULAR LAYERS 1013 water, and the covering slide should be moved and step No. 2 are 28 layers apart, No. 2 and No. 3 are only when the slides are under water, to pre- 30 layers apart, No. 3 and No. 4 are 32 layers apart. vent contaminating oily films from reaching the When any two neighboring steps match exactly, a maximum or minimum lies at a thickness half-way between the glass. two. Thus since No. 1 and No. 2 match when they have A slide coated in steps shows bands of vivid thicknesses 217 and 189 layers, respectively, the minimum color when viewed by white light. When the lies at 203. When No. 1 and No. 2 have thicknesses films are built of calcium stearate the first eight slightly less than these, such as 215 and 187, the diagram shows that the intensity of No. 1 is less whereas that of bands have the following colors when viewed by No. 2 is greater than the matching intensity. Due to this light of perpendicular incidence : 21 layers “seesaw” of intensities about a maximum or minimum, yellow-brown, 41 dark blue, 61 light blue, 81 the contrast observed between two neighboring steps yellow, 101 red, 121 blue, 141 green, 161 red- changes so rapidly with changes in thickness that an yellow. When monochromatic light is used the addition of two molecular layers to both steps is plainly visible to the eye, and under good conditions the addition steps appear as a series of dark and bright bands, of one layer can be detected. the intensity of the reflected light varying from After No. 1 and No. 2 have been matched, more layers step to step. The series of intensities ranges be- are added to all four steps until No. 2 and No. 3 are brought tween a maximum and a minimum value, varying with the thickness of the step in the manner illustrated by the curve .“6 in Fig. 3. The abscissa scale Y8 c of the diagram in Fig. 3 has H been drawn for the case of calcium stearate steps illuminated by sodium light at per- 0 50 100 150 200 250 Film thickness (no. of layers). pendicular incidence. For this Fig. 3.-Two methods (a) and (b) of determining the thickness of a film which case the first three intensity reflects minimum intensity of monochromatic light. The curve is drawn for the minima were found to occur at case of calcium stearate illuminated by sodium light at perpendicular incidence. steps having the thicknesses 39, 121, 203 layers (Table I). The light source to matching positions on opposite sides of the minimum, was a 6000-lumen sodium vapor lamp. and then No. 3 and No. 4 are matched; thus three inde- pendent determinations of the location of the minimum The number of layers corresponding to an in- are obtained. These three determinations usually agree tensity minimurn may be determined by either exactly; when the values disagree they differ by not more of two methods (a) or (b). than one layer. The interval between successive steps should be chosen in such a way as to cause the intensity (a) Symmetrical Intensities.-Alternating films are de- of the steps to lie posited in a series of five steps having a uniform step-rise on a steep part of the intensity curve when the steps are being matched. Widely different for the series. A step-rise of ten layers is convenient for intervals may be used if desired, but it has been found this purpose. When the number of layers reaches 15, 25, most practical to choose three which are not very differ- 35, 45, 55 the steps reflect light with intensities corresponding to the open circles in Fig. 3(a). The 35 step is ent, such as the intervals 28, 30 and 32 illustrated in the diagram, for the observer soon learns to gage intensity seen to be the darklest of the five steps, but the intensity of differences in terms thickness differences if the intensi- 35 is not symmetrical with respect to 25 and 45, the 45 of ties that are used lie always on about the same part of the step appearing considerably darker than the 25 step. When four more layers are added to the entire slide, bring- curve. ing the steps to 19, 29, 39, 49, 59, the 39 step is found to Films of calcium palmitate, calcium stearate be symmetrical with respect to 29 and 49. The thickness and calcium arachidate have been measured by of 39 layers therefore corresponds to minimum intensity for sodium light. As more layers are added to the slide, method (b). Table I gives the results of these the steps are seen to pass through a maximum of in- measurements for alternating hs. The quality tensity at 80 layers, a second minimum at 121, and so on. of these substances as “building material” for (b) Matched Intensities.-The second method consists films is markedly affected by the length of the in bringing two neighboring steps to exactly equal in- chain. Calcium palmitate, having 16 carbon tensity. A series of steps Nos. 1, 2, 3, 4 are built (Fig. atoms in the chain, is troublesome to work with 3(b)) in which the step-rise between consecutive steps may be any known even number of layers and is prefer- and seldom gives satisfactory films after about ably not uniform for the series. In the diagram step No. 1 100 layers have been deposited, the small imper- 1014 KATHARINEB. BLODGETT VOl. 57 fections in many of the first 100 layers causing a Columns (7), (8), (9) give the values of At/AN cumulative roughening of the surface. Calcium calculated from these data for the three sub- stearate ((21,) works well, and calcium arachidate stances. Columns (lo), (ll), (12) give At/AN (C,,) works the best of the three. Subsequent divided by the number of carbon atoms in the experiments have shown that for a given length chain of the substance. The results in the last of chain one obtains the best results by using the three columns are seen to be closely similar for the soap of a heavy metal. Barium stearate films three substances, having an average value 1.36 A. are more easily built than films of calcium stear- Mullerl2 found by x-ray measurement of stearic ate, whereas magnesium stearate films are very acid crystals that the length of a pair of molecules troublesome. arranged end to end was c = 48.84 A. He found TABLEI that the axis of the chains was inclined at an angle = with the plane in which the MEASUREMENTSOF “ALTERNATING”(Y) FILMS OF CAL- /3 63’38‘ CIUM SOAPS carboxyl groups lay, so that the spacing between Water-bath 1/1O,OOO molar calcium carbonate, PH 6.4, consecutive carboxyl planes was c sin /3 = 43.76 temp. 22’. Piston oil: castor oil. Refractive index of The value At/ANs = 24.4 in Col. (8) gives films /1 = 1.471. k. (1) (2) (3) Np (4)ANp NE (51ANs NA (6)ANA the thickness of a single layer of calcium stearate. Calcium Calcium Calcium The agreement between this value and c/2 of palmitate stearate arachidate v PI, A. A. (cis) (Cia) (CW) Muller’s measurements implies that /3 = 90’ for 0 1473 1001 45 39 35 the calcium stearate films. That is, the surface 1 4420 3005 138 93 121 82 108 73 pressure exerted on the molecules as the films are 2 7367 5008 203 82 181 73 deposited causes the molecules to assume positions 3 10313 7011 255 74 Mean 73.3 perpendicular to the plane of the glass surface. PiperESmeasured by x-ray diffraction the long (k) B (2 (E) (li) - -- spacings of potassium soaps of the fatty acids. AX AN8 AX 16ANp 18ANs -A A. A. A. A. A. A. His results for acid soaps are given below, for 21.5 24.4 27.4 1.35 1.36 1.37 comparison with the corresponding measure- 24.4 27.4 1.36 1.37 ments made by the present method (Cols. (7), 27.0 1.35 - - (8), (9)). The data are given in terms of the Mean 27.3 Mean 1.36 spacing corresponding to a single layer.

~is,A. ~1s,A. Cm,A. The thickness t of a film which shows minimum Piper 22.64 25.23 27.82 intensity when illuminated by light of wave Cols. (7), (8), (9) 21.5 24.4 27.3 length X at perpendicular incidence is t = (2% + 1) X/(4 p) where n has the values 0, 1, 2, . . ., and Piper found that the spacings for neutral potas- p is the refractive index of the substance of which sium soaps were approximately 20% less than the film is composed.” Column (2) in Table I the spacings for acid soaps. In the case of calcium gives the values of pt which correspond to the stearate films no change was found in the spacing values of n in Col. (1) when sodium light was used, of Y-films when the PH of the water was changed taking as the value of X a mean of the sodium D within the range of pH in which Y-films could be lines, A = 5893 A. The refractive index of the deposited. Barium stearate Y-films can be de- films was determined by a method which will be posited over a wider range of pH than calcium described; it was found to be p = 1.471 (Table stearate Y-films. No difference in thickness was 11). Column (3) gives the values of t calculated found between barium stearate films formed at for this value of p. PH 6.9 and at pH 8.5. Columns (4)’ (5), (6) give the number of layers When the data for film thickness, given above, N at which minima occur for the three substances, are plotted against the number of carbon atoms and the values of the difference AN between in the chain, Piper’s data lie on a straight line minima. The values of DJV are estimated to having a slope tc = 1.275 A. per C atom. The have a probable error of k0.5 layer for the Go complete range of Piper’s data extends from the and CIS chains, and =+=1layer for the Cle chain. Cq chain to the C24 chain. The writer’s data give (11) This equation applies only to cases, such as the present case, a slope 1.45 A.; however, if the estimated prob- where II has a value less than the refractive index of the substance on (12) A. Muller, Proc. Roy. Soc. (London), All& 546 (1927). which the film is deposited. (18) S. H. Piper, 1.Chcm. Soc., 2.36 (1929). June, 1935 DEPOSITIONOF SUCCESSIVEMONOMOLECULAR LAYERS 1015 able error is taken into consideration, it is found films which they have measured by means of x- that the possible slopes range from 1.34 to 1.55. ray diffraction. They obtained the value 47.55 That is, the range of data from Cle to CPOis too A. for the long spacing of built-up layers of cal- short to afford an exact determination of tc. cium stearate, and 47.53 A. for the long spacing The lowest value tc = 1.34 is higher than most of pure synthesized calcium stearate studied by of the values previously determined by other the older methods. These long spacings cor- workers from x-ray measurements of fatty acids respond to the thickness of two monomolecular and of soaps. The highest value obtained from films turned in opposite directions. They found the measurements made by Slagle and Ott“ is for calcium palmitate a long spacing 43.50 A. and approximately 1.2 A. The value of the length for myristate 39.55 A. A plot of their values for which each CH2 group contributes to the chain, the three chains against the number of carbon calculated on the theoryl0 that the chain is a zig- atoms in the chain has a slope tc = 1.0 A. zag in which the lines joining the centers of suc- Films of barium stearate deposited from a solu- cessive carbon atoms are inclined at the tetra- tion l/lo,~molar in barium carbonate gave the hedral angle 1109”28‘, is 1.257 A.,the diameter of same value of AN, as films of calcium stearate. the carbon atom being 1.54 A. The higher value Barium stearate films were found to have a re- obtained in the present experiments seems to fractive index p = 1.46 (Table 11). Piper13 ob- indicate that the molecules, being held under served, when he compared the long spacing of much constraint from the pressure applied by the thallium stearate with that of potassium stearate, piston oil, assume orientations and lengths which that “a very small change in spacing is produced have a greater “long spacing” than when films are by the addition of a very heavy metallic atom.” formed by other methods. Piperl6 and others The change which he observed was 2.3% for the have shown that the long spacing is affected by normal palmitate, 0.5% for the normal stearate. the degree of constraint imposed on the molecules Films of magnesium stearate in the present ex- of a fatty acid while crystallization is taking place. periments were also the same as calcium stearate The surface pressure exerted by oleic acid, 29.5 within the limits of error of the measurements; dynes/cm., on a molecular wall 25 A. in thickness, it was found difficult to produce homogeneous is 118 atmospheres. In the case of castor oil, films of magnesium stearate when the number of 16.5 dynes/cm., the pressure is 66 atmospheres. layers exceeded about 50. Castor oil was used for the series of measurements Clark has pointed out that films of stearic acid given in Table I for the reason that calcium palmi- may be deposited from the surface of purest dis- tate tends to collapse slowly on the water surface tilled water. He obtained for these films a long when subjected to the pressure of oleic acid. spacing 46.05 A. The writer has deposited films Only a very slight difference was found between of stearic acid on polished chromium (i. e., chro- series of films obtained when castor oil was used mium-plated brass) from distilled water, pH 5.2 and those obtained with oleic acid. In the case of to 5.4, and from the same water to which sufficient calcium stearate, for example, the number of hydrochloric acid was added to lower the pH layers between successive minima for sodium to 4.2. These films were found to have a value light was ANIS = 82 in the case of castor oil of $1 approximately 8% less than the calcium or (Table I), and AN, = 80 to 81 in the case of barium stearate layers, where tl is the thickness oleic acid. This observed increase in thickness per layer. The films at pH 4.2 gave the same re- of approximately 2% for a change in pressure of sult as films at PH 5.2. The value of ph was es- 13 dynes/cm. is in accord with Adam’s6 measure- timated in the following manner. ments of the compressibility of monomolecular Two color scales were made on separate slides, films of palmitic acid on pure water. He found to serve as gages. On the first slide there was that the force required for 1% decrease in area built a series of’ten “20-layer” steps of barium (corresponding to 1% increase in thickness in stearate, i. e., the steps had 21, 41, . . ., 201 layers. the present experiments) was 5.9 dynes/cm. On the second slide ten “18-layer” steps of barium Clark, Stenett and Leppla4 have used the stearate were built, i. e., the steps had 19, 37, . . ., method of the present experiments to deposit 181 layers. When the slides were viewed side (14) F. B. Slagle and E. Ott, TrsJOURNAL, M,4399 (1933). by side, by white light at normal incidence, they (15) S. H. Piper, ‘I.. Malkin and H. E. Austin, J. Chcm. Sa., 2313 (1928). formed two different color series. A series of ten 1016 KATHARINEB. BLODGETT Vol. 57

20-layer steps of stearic acid was then built using angle, the light reflected from the surface of the oleic acid as piston-oil, and it was found that this film changed 180’ in phase; but the light re- series nearly matched the 18-layer series of barium flected from the film-glass boundary did not stearate which was also built at oleic acid pres- change phase because refraction of the light as it sure (AN, = 80 for oleic acid pressure). The entered the film caused the light to arrive at this difference in ptl between stearic acid and barium boundary at an angle of incidence far less than stearate was estimated as being 8% of ptl for the polarizing angle characteristic of a glass-film barium stearate. The refractive index of stearic boundary. Therefore when the difference of acid was found to be p = 1.48. From the meas- path length between the ray reflected from the urement AN = 80/0.92 and the refractive index surface of the film and the ray reflected from the one obtains the result At/AN = 22.9 8. for the film-glass boundary was such that the two rays thickness per layer of stearic acid. produced minimum intensity at an angle slightly Determination of Refractive Index.-The re- less than cpp, the relative phase change of 180’ at fractive index p of the calcium stearate films was (pp caused these rays to produce maximum in- determined by measuring the “polarizing angle” tensity at an angle slightly greater than pap. of the films. The method was based on certain Vue versa, a maximum for cp < cpp changed to a laws of optics which will be stated briefly. minimum for cp > (pp. Thus two neighboring When light is reflected from a surface, the com- steps which were seen as a minimum and a maxi- ponent of vibration parallel to the plane of inci- mum for cp < cpp exchanged brightness as cp dence is reflected with an amplitude different from passed through the polarizing angle; and at the that of the component perpendicular to the plane polarizing angle the steps matched exactly. The of incidence. The component parallel to the angle at which the steps matched was found to plane of incidence has the following properties. be very critical, and could therefore be measured (i) The intensity is zero when (cp + x) = 90°, with a probable error less than 0.1 degree. where cp is the angle of incidence of the light In order to obtain the greatest sensitiveness it illuminating the reflecting surface, and x is the was desirable that the steps be built so that the angle of refraction. The angle of incidence at thicknesses would correspond exactly to maxi- which (cp + x) = 90’ is the “polarizing angle” mum intensity. The theory of the method is the cpp. At this angle sin xp = cos (pp, hence p = same for steps having any difference in intensity, tan cpp, since by definition p = sin (p/sin x. but obviously the method is most sensitive when (ii) When cp < (pp, the amplitude of the re- the difference in intensity is as great as possible. flected light is in phase with that of the incident The number of layers which were needed to build light. the steps was calculated from the data of Table (iii) When cp > cpp, the amplitude of the re- I by means of the formula for the path difference flected light has a difference of phase of 180’ 1 between the rays reflected from the upper and from that of the incident light. from the lower surface of a film of thickness T, In the present measurements the reflected which is 1 = 2 p T cos x. For the data of Table light was polarized by means of a Nicol prism I, cos x = 1. When cp = 55O, x = 33’50’, cos which transmitted the component parallel to the x = 0.83. The steps were therefore built as plane of incidence. The polarizing angle pp was follows: minimum at cp = 55’, 47 layers; maxi- determined by measuring the angle at which the mum, 97 layers; minimum 147 layers. The change of phase of 180” between Case (ii) and steps were deposited on glass having a refractive Case (iii) occurred. This was done by viewing index 1.64, from a solution ‘/IO,OOO molar in cal- two neighboring steps, one of which was built cium carbonate, fiH 6.4, temp. 22’. The slide with a thickness which reflected minimum inten- was mounted at the center of a spectrometer sity when illuminated by sodium’light at an angle table in such a position that the boundary line of incidence cp = 55O, and the other with a thick- between two steps lay in the axis of the table. A ness which reflected maximum intensity at the horizontal beam of unpolarized light illuminated same angle. The angle cp = 55’ is less than the the slide. The observer viewed the slide by look- polarizing angle of calcium stearate and corre- ing through the slit in the end of a collimator sponds to Case (ii). tube, a Nicol prism being mounted in the axis of When cp was made greater than the polarizing the tube. The plane of the light transmitted by June, 1935 DEPOSITIONOF SUCCESSNEMONOMOLECULAR LAYERS 1017 the prism WiaS parallel to the plane of incidence, Stearic arid does not attach itself readily to that is, the plane was horizontal. The angle (pp glass, and no method was found of depositing a was determined by measuring the difference 2pp first layer which would adhere to the glass satis- between two positions of the slide in which a maxi- factorily while 50-100 layers were built on top of mum and a minimum were seen to exchange bright- it. Imperfections in the first layer always al- ness, one position occurring when the observer lowed water to work its way underneath this layer viewed the slide at the angle of reflection (pop and and gradually to rip sections of the film off the the other position at -pp. glass. The difficulty was obviated by depositing The results of these measurements are given in a first layer of barium stearate on the glass from a Table 11. The measurements were not made for barium carbonate solution, $H = 8.6, and then the Cle and CZOsoaps. The refractive indices of building on top of this first layer successive stearic molten palmitic and stearic acids are nearly equal, acid layers taken from distilled water. the valuesls being p = 1.4273 for molten palmitic Decrease in Thickness Produced by Heat.- acid at 79.8' and p = 1.4299 for molten stearic Significant changes occur in the thickness of Y- acid at 80.2'. It seems probable, therefore, that films when the films are heated. The record of variations in p with the chain length of the soaps four tests will indicate the nature of the results are negligible for the purposes of the present recently obtained; a systematic investigation of measurements. the effects of heating has not yet been made. The films were baked in an oven the temperature TABLEI1 of which was held constant with an accuracy of REFRACTIVEINDEX p OF FILMS No. of layers of two 1' by thermostat control. Substance neighboring steps Bp tan 9p = p (1) Calcium stearate, Y-films. Water PH 6.6, Calcium stearni e 47 and 97 55'45' 1.469 Y-Films 97 and 147 55'50' 1.473 T 22O, l/lo,ooo molar calcium carbonate. Pres- - sure: castor oil. Films deposited on polished (PH6.4) Mean 1.471 chromium in seven 20-layer steps. The slide Calcium stearate was baked for fifteen minutes at 100' and during approx. 42 and 87 55'47' 1.471 baking the thickness of each step decreased 20%. X-Films approx. 87 and 132 55'45' 1.469 The decrease was measured by using as a gage the (PH 8.8) Mean 1.470 colors of an unbaked slide on which seven 16- Barium stearate 47 and 97 55'35' 1.460 layer steps had been built. Further prolonged Y -Films 97 and 147 55 '40' 1.464 baking at 100' produced no further change in (PH 6.9) Mean 1.462 thickness. A change took place, however, in the Stearic acid 51 and 103 55'55' 1.478 structure of the film, for when viewed by trans- Y-Films 103 and 155 56"5' 1.487 verse light each step scattered light of a color complementary to the color of the same step when (PH 5.2) Mean 1.482 viewed by reflected light. Baking at 150' pro- In order to measure the refractive index of duced a gradual transition from a transparent stearic acid by the method used for calcium stear- film to a white, opaque film, which caused the ate it was necessary to deposit the stearic acid on colors to appear to be "fading out," the transition glass rather than on chromium. The component taking place in an uneven, spotty manner over of the light reflected from the surface of a film the surface. which is used for making these measurements has (2) Barium stearate, Y-films. Water PH 8.5, such low intensity at angles of incidence near the T 22', '/IO,OOO molar barium carbonate, l/lwo polarizing angle that if the intensity of the light molar sodium bicarbonate, plus a very small reflected from the film-solid boundary be great, amount of sodium hydroxide. Pressure: oleic the very weak ray has little effect in modifying the acid. Surface : polished chromium. The slide intensity of the strong ray. Therefore in order to was baked for ten minutes at 100' and showed a obtain as much contrast as possible between steps 10% decrease in thickness. Further baking at of minimum and maximum intensity, the ray from 100' for two hours produced no further change the film-solid boundary must be weak, and glass in color. Complementary colors seen by scat- serves this purpose far better than chromium. tered light, Baking at 150' caused gradual (16) "International Critical Tables." change to opaque film. 1018 KATHARINEB. BLODGETT VOl. 57

(3) Barium stearate, Y-films. Water PH 6.9, Dr. Langmuir has pointed out that a decrease T 22', '/IO,OOO molar barium carbonate. Pres- in thickness of 20% in a film which remains trans- sure : oleic acid. Surface : polished chromium. parent must be accompanied by an accumulation Baking at 65O for ten minutes caused no change of about 20% of the material in small mounds or in thickness. Steps viewed by scattered light ridges on the surface, these excrescences occupy- showed not complementary colors but colors ing a comparatively small fraction of the area of which would be seen by reflected light from steps the surface. The strong forces exerted by the having approximately 20% less thickness. After molecules in their constrained upright positions twenty minutes at 65" the difference in color be- become sufficient to expand the film in a lateral tween scattered and reflected light was less direction when the forces are increased and the marked, after thirty minutes at 65' the differ- rigidity of the film is diminished by heat. The ence had disappeared. After ninety minutes at expansion produces breaks and ridges in the films 65" the steps continued to retain the same thick- analogous to the pressure ridges which are com- ness which they had had originally. The slide monly seen in the ice of a large frozen lake. The was then baked at 70" and after ten minutes following brief study was made of the structure much of the film was opaque, after 20 minutes as seen with the aid of a microscope. every step except the 21 and 41 layer steps was A barium stearate film of the type described completely white. It was observed in three sepa- in Case (2) was baked at 100' for ten minutes, rate tests that the two thinnest steps were the after which it showed a decrease in thickness of last to lose color and become opaque. lo%, the color of the light scattered by the sur- The transition temperature 70' in this test is face being then complementary to the color of the the melting point of stearic acid. It is higher, reflected light. The film was viewed through a however, than the transition temperature ob- microscope by dark-field illumination, at a mag- served for films of stearic acid (see Case 4). The nification 270:l. The steps having a thickness films in Case (3) had the same thickness as the of 81 or more layers were seen to be crossed by a films in Case (2), and did not have the thickness series of parallel ridges. The direction of the of stearic acid films. It seems probable, there- ridges was parallel to the line of intersection which fore, that in Case (3) the films were an acid barium the plane of the water surface made with the stearate composed of neutral barium stearate and plane of the slide when the film was being built. stearic acid combined in a proportion which is not Mr. L. L. Wyman has measured the spacing of known, and that in Case (2) the films were neu- the ridges on a film of 141 layers, thickness T = tral stearate. Barium stearate which was made 3.4 X cm. He found an average spacing from sodium stearate by precipitation with barium of 6.56 X 10-4 cm., which is 19 T. chloride showed no sharp melting point, but be- The spacing of the ridges was nearly the same came sticky at about 140'. for all the steps which were thick enough to show (4) Stearic acid, Y-fis. Water PH 5.2 to any ridges, but on the thickest step (141 layers) 5.4, T 22'. These films gradually deteriorated additional ridges were seen which were more pro- at 65'. The transition temperature was some- nounced than the first set. The second set ran what less sharp than in Case (3). parallel to the first but with much wider spacing, The marked decrease in thickness which OC- and had occasional cross ridges at right angles. curred in Cases (1) and (2) seems to indicate that These preliminary observations have shown that in these cases the molecules were able to re-orient the growth of the ridges may be studied as a themselves as a result of molecular agitation at the function of film thickness and conditions of bak- higher temperature, and that they then departed ing, and further experiments will continue this from the constrained positions in which they were study. deposited and assumed orientations more nearly The phenomenon of films which scatter light characteristic of the substance in its normal state. having a color complementary to the color of re- Piper1*found from his studies of potassium palmi- flected light has been observed by Wood'' in the tate, stearate and eicosoate that "in contrast with case of "frilled" transparent films on metallic the acids, which possess a number of crystalline surfaces. The phenomenon is attributed by him forms, only one form appears to exist for these (17) R. W. Wood, "Physical Optics." 2d ed., p. 172; 3d ed., p. pure anhydrous salts at the ordinary temperature." 206; R. W. Wood, Phil. Men.. 7,386 (1904). June, 1935 DEPOSITIONOF SUCCESSIVEMONOMOLECULAR LAYERS 1019 to the effect of a mesh structure which is “so not appeared in the range pH 7.0 to 11.0, at tem- small that it can be detected only with the highest peratures 20 to 30’. powers of the microscope.” In the present case Layers in an X-series are deposited only as the it is due to the ridges on the surface, and a fur- slide descends into the water, as previously de- ther study of these ridges should lead to an ex- scribed. Let n be the number of journeys which planation of ithe phenomenon. the slide must make in order that a thickness At It can be argued that water is held by the = 29471~8. may be added to the film on the films as they are deposited, and that the presence slide, counting each ascent as one journey and of layers of water accounts for the high value of each descent as one. Then in the case of Y- tc and loss of water accounts for the shrinkage at films n = N where N is the number of layers cor- high temperature. Films of barium stearate responding to this value of At. In the case of placed in a idesiccator for three days at room X-films n = 2N provided that no films whatso- temperature showed no change whatsoever in ever are collected on ascents. color. Further tests are needed to decide this If N’ films are collected on ascents, n = question. 2(N - N’). Optical Properties of the Film-Glass Boun- The question of whether a slide collects a small dary.-If the values of N in Cols. (3), (4), (5) amount of calcium stearate on some of its ascents of Table I are plotted against the values of pLt in is not easy to decide, for the following reason. Col. (2),the result is a linear plot which does not Experiments have shown that under certain con- pass precisely through the origin. At t = 0, the ditions n may range nearly all the way from n = plot shows Np = -1.5, Ns = -2, NA = -1.7. 2N to n = N. That is, a slide may take up an Intercepts N = -1 to -2 layers were obtained amount of calcium stearate on successive ascents from every set of data that was taken in the which corresponds to the thickness of N’ layers course of making these measurements. That is, N’ having any value from 0 to N/2. This occurs the mechanism by which reflection of light takes for the reason that the amount taken up by an place at the glass-film boundary is of such a na- ascending slide depends on the length of time ture that the reflection appears to take place at which the slide has spent under water since the an average distance approximately 20 to 40 A. preceding descent. If the time under water is beneath the surface of the glass. short, the glass collects a film on the ascent; if Films deposited on glass may be loosened from the time is long it collects little or no film. the glass by acid and the entire film floated off This property of X-films was first observed in on a water surface. This is accomplished by cases of “graded” films. In these films the upper placing a few drops of acid (20 to 50% hydro- edge of the slide (the edge which emerges first chloric acid) on one end of the slide from which from the water) takes up more calcium stearate the film has been wiped off so that the acid comes than the lower edge, for the reason that the upper in direct contact with the glass. The slide is part of the slide always spends less time under tipped slightly so as to bring the acid up to the water than the lower part. The result is a film edge of the film, and the acid then creeps slowly having a thickness which slopes continuously from under the film, disrupting the bond which at- the upper to the lower edge of the slide. The taches the film to the glass but not dissolving the continuous grading of thickness is made apparent film. A film which exhibits a given interference by the continuous grading of color. The follow- color on glass exhibits the complementary color ing experiment tested this property of X-films when floating on acid or water, since the index more exactly. of refraction of glass is greater and the index of A fresh solution was used, 6/~~,~molar refraction of water is less than that of the stearate. calcium carbonate, pH = 7.4. Pressure: oleic Measurement of X-Films.-These films occur acid. Values of n/2 were measured at various on alkaline water containing calcium or strontium. temperatures. That is, journeys of the slide They have not been observed in the case of mag- were counted while thicknesses of two “matched nesium or barium. The latter substances have layers” (method (b) Fig. 3) were changed from not been so thoroughly investigated as calcium, values on each side of the first minimum for so- so it cannot be said with certainty that X-films dium light to values on each side of the next never occur with these stibstanctts, hut they have maximum. The results are given in Table 111. 1020 KATHARINEB. BLODGETT VOl. 57

The seven sets of measurements were made con- T 25’, pressure castor oil. The slide was dipped secutively with the same solution. The comment at the usual rate. The alkalinity was secured by “Dipped at usual rate” means that the slide was adding sodium hydroxide, and a concentration of allowed to descend rapidly and was raised as l/1000 molar sodium bicarbonate to serve as rapidly as it would shed water, the two operations buffer. The values of n/2 obtained for this solu- requiring about ten seconds for completion. tion were 74 to 76. If these values represent perfect X-films, the difference between Nx = TABLE111 n/2 = 74 to 76 and NY = n = 80 represents the MEASUREMENTSOF “NoN-ALTERNATING”(X) FILMSOF CALCIUMSTEARATE difference in packing between the layers of mole- Solution 6/1~,mmolar calcium carbonate, pH 7.4. cules in the two cases, the thickness of the layers Piston-oil: oleic acid. being 5 to 8% greater for X-films than for Y-films. TEmp., C. nl2 A slide coated with steps of X-films was baked Piston-oil: oleic acid in the oven at 100’ at the same time as the slide 33 68 Dipped at usual rate of Y-films described in Case (1) of the section 26 66 Dipped at usual rate dealing with baked films. The X-films showed 20 Graded thickness : Dipped at usual rate no decrease in thickness due to baking. They 42 top, 62 bottom 15 44 Dipped at usual rate deteriorated at 150’ in the same manner as the 17.5 72 Held under water 30 sec. after Y-films. each descent 17.5 42 Dipped at usual rate Miscellaneous Films 17.5 66 Held under water 30 sec. after each descent Mr. V. J. Schaefer has assisted the writer in studying several different varieties of films. Those values of n/2 in Table I11 which are Films can be built on a metal surface as readily high correspond to cases in which comparatively as on glass, provided the metal has a mirror polish. little stearate was collected on ascents. High If the metal is not highly polished, a first layer is values were obtained consistently with this solu- reluctant to adhere to the surface. Films have tion at T 2 24’. At T 7 17.5’ high values were been built on chromium, nickel, brass, steel, cast obtained only when the slide was left under water iron, silver, platinum. Films on polished chro- for an exceptionally long period of time. When mium exhibit remarkably fine colors and this the slide was withdrawn at the usual rate at 17.5’, metal was used for building the step-gages pre- the count n/2 = 42 was obtained. This is al- viously described. most exactly the value for Y-films, n = 80 for The colors of films on chromium are especially oleic acid pressure. These results are interpreted brilliant when the slide is illuminated at angles as meaning that a change takes place in the mole- of incidence greater than about 65’ by polarized cules which form the layer collected on a descent, light, the colors of the steps when the light is polar- the rate of change depending on the temperature ized in the plane of incidence being in most cases of the water. As a result of this change the layer complementary to the colors when the light is which would ordinarily adhere on an ascent (as it polarized perpendicular to the plane of incidence. does adhere in the case of Y-films) does not ad- A greater variety of color effects is obtained when here, or adheres only at various spots on the sur- the slide is illuminated at these angles of incidence face. by light polarized in a plane at azimuth 45’ and At the higher temperatures the change occurs is viewed through a Nicol prism. The colors of in a time which is short compared with the time the light transmitted by the analyzing Nicol of dipping; at 20’ the time is approximately the change as the Nicol is rotated. same as the time of dipping so the change occurs When steps of barium stearate having 1,3,5,. . . , more completely on the lower portion of the slide layers are built on polished chromium, and the which spends the longer time under water; at slide is illuminated at angles near grazing inci- 17.5’ the time is more than ten seconds less than dence by light polarized in the plane of incidence, thirty seconds. the difference in intensity of light reflected from The maximum values of n/2 which have been the successive steps is easily seen, and the differ- obtained for X-films were obtained with solutions ence between a single layer and clean chromium l/lo,oM) molar in calcium carbonate, PH 8.8 to 9.0, is plainly visible. June, 1935 DEPOSITIONOF SUCCESSIVEMONOMOLECULAR LAYERS 1021

Before the films are deposited the metal is to form expanded films. At $H 3.6 the film on a cleaned electrolytically by being made cathode water surface collapsed so rapidly under oleic in a solution 2.5% in sodium hydroxide and in acid pressure that films could not be built at water and alcohol, with a current 1.5 amp. for room temperature. Probably by lowering the about ten seconds. Only a small concentration water temperature films could be obtained at of water is used to dissolve the sodium hydroxide. lower values of PH. Both Langmuir and Adam In the case of glass the first layer is deposited have found fatty acid films less stable under pres- most readily from alkaline water, in the case of sure on acid water than on neutral water. metal most readily from water that is neutral or Mr. Schaefer has deposited films of stearic slightly acid. The solution used for making the acid on glass from distilled water ($H 5.2) by step-gages on chromium was ~/IO,OM)barium drying the slide in an oven at 50'. The slide carbonate in equilibrium with the pressure of was held in a horizontal position in the oven. carbon dioxide present in ordinary air (pH 6.9 After a first film is dried by this method, a few approximately, at T 22'). If the metal has a successive films of stearic acid can be deposited high polish and is thoroughly clean, the first layer on top of it in the usual manner. If the slide is attaches itself readily and water recedes rapidly dried at temperatures higher than about 60°, suc- as the slide is raised on the first ascent. The first cessive films cannot be deposited satisfactorily. layer is reluctant to attach itself to glass under the He found it necessary to use oleic acid for he could same conditions and must usually be dried on. not form a first film at the pressure of castor oil A number of different kinds of glass were used which would take up successive films. At PH in these experiments. Every kind of glass that 5.2, T 22O, the film on the water surface contracted was used took up a first layer readily from alka- at a rate which was initially very slow but be- line water. Small differences were observed in came quite rapid after about five minutes at oleic the case of films deposited from acid water, the acid pressure. This difficulty was avoided by differences being very small, however, compared working at 15', at which temperature there was with the great difference between glass and metal almost no contraction. in this respect. The glass used for all the meas- The experiments on the effect of raising films urements of film thickness, p = 1.64, took up a to a high temperature, described in a previous first layer from acid water more satisfactorily section, demonstrated that films taken up from than any other glass, the first layer being de- neutral or alkaline water were not composed of posited by drying the slide in an oven. This free fatty acid, but of acid soap or neutral soap. glass contained about 50% of lead. Bausch and The substance of which the film on the water sur- Lomb microscope slides, p = 1.52, took up a first face is composed may be studied very simply by layer under these conditions which was slightly skimming it off the water. The entire length of less perfect than a first layer deposited on the a trough is covered with a monomolecular film glass containing a high percentage of lead. The and then a barrier is used to shove all the film to layer was apparently attached less firmly to the one end of the trough. As the film is jammed by microscope slide than to the lead glass. A sample the barrier it breaks and wrinkles, and much of plate glass of thickness 0.7 cm. was very trouble- water is enclosed in the wrinkles. Thus the very some, the first layer being usually not at all uni- small amount of stearic acid required to cover a form. The ccmposition of this glass was not large surface, 0.3 mg. for 1300 sq. cm., assumes known to the writer. an appreciable volume when the jammed mass is Stearic acid is taken up by chromium very collected at one end of the trough. It may then readily from acid water. Step-films of this sub- be readily skimmed off the water by a strip of stance have been built at PH 4.2 using the pres- metal, and heated over a flame to observe the sure of oleic acid. High pressure must be used melting point. When the films are calcium stear- with stearic acid, for at low pressure fatty acids ate, the soap remains white and solid while the form expanded films on acid water and the ex- water held by the soap boils away. When the pansion apparently interferes seriously with the films are stearic acid, the acid is seen to melt be- building of successive layers. Adams gives the fore the water commences to boil. value PH 5.5 approximately, as being the hydrion It has not been found possible to build films of concentration at which palmitic acid commences sodium stearate by any of the ordinary methotls 1022 KATHARINE€3. BLODGETT VOl. 57 using solutions containing Na but no Ca. Since then 2n layers of barium stearate, and so on in sodium stearate is soluble in water, the strong alternating succession. The stearic acid would affinity for water presumably prevents it from be more transparent to radiations of short wave leaving the water to attach itself to a solid. A length than the barium stearate and would solution '/1000 molar in sodium hydroxide gave therefore serve to space the series of layers of films which showed no tendency to adhere to a barium stearate at known intervals apart. solid, but when a concentration of 3 X molar calcium carbonate was added to this solution, summary moderately good films were obtained at once. Films were built by transferring monomolecu- Calcium may be added to a solution in the form lar layers from a water surface to a solid at a rate of calcium chloride instead of calcium bicarbonate ten to fifteen per minute. The films were made and gives the same results for corresponding of calcium soaps of palmitic, stearic and arachidic values of PH. The carbonate is more satisfactory acids. The Mg, Sr, Ba, Pb and Zn soaps and the in most cases since it serves as an alkali. fatty acids were also used. Two different types The soaps which have given satisfactory films of layers of calcium stearate were observed, their have all been the soaps of bivalent metals. In occurrence depending on the PH and temperature addition to the soaps of Mg, Ca, Sr, Ba, good of the water from which the films were transferred. films have been built using Zn and Pb. The Films were deposited on glass, chromium, films of Sr, Zn and Pb have not been measured. nickel, brass, steel, cast iron, silver and platinum Films from solutions containing aluminum chlo- surfaces. They were built in steps having 21, ride have the property that these films will not 41, 61, . . ., 201 layers, and in other chosen se- adhere to a solid either on an ascent or on a de- quences of odd-numbered layers. These steps scent of the slide. If a slide coated with stearic showed vivid interference colors when viewed acid or with the soap of another metal be lowered by white light. A single layer could be seen on into a solution containing aluminum chloride in polished chromium by means of polarized light. a concentration 1/~~,~molar (PH 5.4) on which The thickness of step-films was measured by a surface film is spread, no layer adheres on means of the interference of reflected monochro- the descent. This property of an aluminum matic light. The refractive index was determined soap is markedly different from the properties of by measuring the polarizing angle by a new and soaps of the alkaline earth metals, which always sensitive method. The results gave values of film adhere to a descending slide. thickness a little greater than the long spacings of Composite films can be built by depositing one the corresponding fatty acids determined by other substance on top of another. A palmitate can workers by means of x-ray diffraction. They were be deposited on top of a stearate, or ~ceversa. 2 to 4% lower than Piper's measurements of the Stearic acid can be deposited on top of barium potassium acid soaps of fatty acids. The melting stearate, etc. Dr. Langmuir has suggested that point of barium stearate films indicated that films films could be built for use as diffraction gratings taken from water, @H8.5, were probably neutral for soft x-rays by depositing (2n + 1) layers of stearate; from water, PH 6.9, acid stearate. barium stearate, then 2n layers of stearic acid, Scmrncrmy, NEW YORK RECEIVEDAPRIL 1, 1935