SOME APPLICATIONS OF X-RAYS

ORGANIC COMPOUNDS *

J. J. Trillat, Paris

In this article we propose to consider investigations of organic compounds carried out with the aid of X-rays, and to show in sufficient detail what theoretical and practical conclusions may be drawn on the basis of this study in a field which until now has remained closed to spectrographic investigations by X-rays.

Here we shall not touch upon anything that is not directly related to the determination of the crystalline structure of organic substances, except in those cases when such consideration will be directly connected with our immediate task. We shall consider only relationships of a physico-chemical order, such as: determinations of molecular lengths, orientation phenomena, the study of surface reactions, analytical applications, the study of lubrication phenomena, etc.

ORIENTATION OF MOLECULES

The arrangement of molecules and their orientation are closely connected with capillarity phenomena. The study of certain thin films—such as soap bubbles—has shown that these films are built up from a series of sheets of equal thickness and that the surface tension of the film depends on this layered structure. There also exists a large number of substances,

* Chapter from the book J. J. Trillat, Les applications des Rayons X.

possessing such a lamellar structure both in the solid and in the viscous state. It is sufficient, as an example, to mention that fatty acids, when melted on glass, orient themselves in such a way that their molecules are arranged between rectilinear, equally spaced planes, forming a true layered structure.

Since the study of a considerable number of organic compounds by means of X-rays is directly connected with the presence of such a layered state, it is not useless to give a brief survey of works devoted to this question; there is, moreover, a number of more important problems connected with the study of surface layers of thin films, such as: capillarity phenomena; so-called contact phenomena, through which a very large number of chemical reactions occur; phenomena of friction, lubrication, etc.

The state of surface layers, which accounts for a considerable number of facts, is in many cases more important for study than the internal structure of bodies.

We have already mentioned soap films; J. Perrin succeeded in showing that under certain conditions these films are composed of superposed sheets, analogous to sheets of paper of equal thickness sliding one over another. On the other hand, Langmuir,^2 studying such layers of fatty acids on the surface of water, suggested that the molecules situated on the surface of water at the moment of maximum tension have an elongated form and are always arranged on the surface in accordance with the affinity of water for the hydroxyl group. Such states of a peculiar polarization for solid bodies are characterized by the appearance of a surface capable of being wetted by water, and an external surface which does not retain liquid (Devaux).^3 In all these thin films the molecules are arranged side by side, forming a sheet solid in one direction, while the presence of the plane of \(\mathrm{CH}_3\) causes the formation of layers sliding one over another, thereby creating properties of a peculiar flow in a direction perpendicular to that indicated above.

The study of crystalline forms by means of X-rays shows that a peculiar arrangement of molecules in such thin surface films is found both in substances in a state intermediate between liquid and solid, and in solid crystals. Such an arrangement is characteristic of a very large number of substances.

Experiment has shown that these substances, especially when they exhibit a lamellar structure, possess lattice planes capable of reflecting X-rays in the same way as regularly formed crystals.

For substances that do not have a crystalline state, the origin of these lattice planes must be sought in the formation of chains from long molecules, as occurs in some liquid crystals. Arranging themselves parallel to one another, the molecules become such that their ends form a system of equidistant planes; since these ends have a more considerable capacity for scattering X-rays than the rest of the molecule, the planes along which these ends are situated form a lattice system capable of reflecting X-rays.

In other substances which, while not possessing a truly crystalline structure, nevertheless display certain signs of crystalline organization, the appearance of reflecting planes occurs as a result of a preferential arrangement of microcrystals caused by an external influence, such as, for example, melting on an orienting substrate, pressure, etc. In this case the molecules are not arranged freely, but form, as it were, a crystalline lattice. One may imagine that a large number of small prismatic crystals with parallel bases, situated next to one another and layer upon layer, gives rise to the appearance of equidistant planes in which the ends of the molecules are located. Substances oriented in this way also possess a lamellar or stratified structure, which can also be detected optically.

Thus, owing to this phenomenon, it proved possible to apply X-ray analysis methods to a broad class of substances that do not possess a true crystalline structure, and the study of these substances was carried out with the same ease as the study of properly formed crystals such as rock salt and mica.

STUDY OF FATTY ACIDS WITH A LONG CHAIN

The first series of investigations in this direction was carried out in 1928 by Müller and Shearer on fatty acids of homologous series from \(C_{20}\) to \(C_{22}\). The investigations have recently been extended down to acids with two C-atoms.

The technique of the investigation was as follows: a small amount of the molten substance was applied to a glass plate so as to obtain strips 2–3 mm wide, 1 cm long, and from 0.1 to 0.2 mm thick. Apparently, under these conditions the substance under investigation is arranged in layers parallel to the surface of the glass and capable of producing sufficiently intense reflection of X-rays. \(^1\)

To prove this, the method of the rotating crystal was used.

Monochromatic rays fall on a glass plate covered with the substance under investigation, which slowly rotates about a vertical axis; as soon as the angle formed by the incident ray with the surface of the substance under investigation corresponds to the wavelength used, reflection of the X-rays occurs; falling on a photographic plate placed behind, they produce a definite effect on it.

Reflection also occurs at angles corresponding to double, triple, etc. values of the wavelength, according to the well-known law

\[ n\lambda = 2d\sin\theta. \]

Crystalline films, being oriented parallel to the surface of the glass, act as a single crystal

and give a series of lines characteristic of the given substance. In photographs obtained in this way one can see (Fig. 1) a group of well-defined lines symmetrical with respect to the central spot. It is important to note that the mutual arrangement of these lines changes with the substance. It decreases as the number of \( \mathrm{CH}_2 \) increases, which corresponds to an increase in the distances between the lattices. In the photograph one can see lines belonging to the 10th order of reflection.

In addition to these intense lines, arising from reflection on the planes of greatest density, there are others, which are situated between the preceding ones and originate from planes separated by smaller distances; the position of these lines is constant for one and the same series of substances.

Fig. 1.

Fig. 2.

Calculation of the large distances between the lattices shows that their increase with the number of \( \mathrm{CH}_2 \) is constant (Fig. 2): for saturated fatty acids,

\[ \frac{\Delta d}{\Delta N} \]

proved to be equal to one angstrom. Thus the addition of each \( \mathrm{CH}_2 \) group manifests its influence in a constant and measurable lengthening of the molecule.

Another series of experiments was carried out with saturated fatty acids, with saturated hydrocarbo-

...amides, ethers, ketones, alcohols, and many other substances with a long chain. For all the substances listed it turned out that an increase in the number of \(CH_2\) groups is accompanied by a regular change in the distance between the nets.

Thus it is now possible to put forward one more fundamental law, valid for all substances of this type.

“The length of the carbon chains of hydrocarbons of one and the same homologous series and, consequently, the distance between the lattice planes formed by these compounds, increases regularly with the number of carbon atoms.”

Let us now examine these phenomena more deeply, in order to show what the spectrographic study of such bodies gives to physics and chemistry. We shall consider first the physical side of the question, then the chemical.

Physical study of the orientation of fatty acids with a long chain

The study of compounds with a long chain by means of X-rays can provide new information concerning the orientation of molecules and the various factors that determine it.

Let us take, for example, palmitic acid, apply it to a glass plate, and record a spectrogram by the rotating-metal method described above. In doing so the following can be observed:

1. The moist condition of the surface of the plate hinders the orientation of the molecules. M. Devo succeeded in showing by another method that the adhesion of fatty substances to glass occurs only under conditions of direct contact, and that even a monomolecular layer of water between the glass and the substance can prevent it from adhering.

Heating, which makes it possible to dry the plate completely, promotes adhesion and, consequently, orientation. The author was able to confirm this experimentally. Probably this phenomenon is connected with the appearance, in fatty substances, of two surfaces—one wettable and one non-wettable.

1. If the preparation is cooled rapidly, or if the acid is applied to a cold plate, the layers will not have time to form because of the excessively high rate of crystallization; the molecules will be fixed in the positions in which they reached the glass. But the same thin layer, heated and slowly cooled, gives a good orientation spectrum.

2. Acidic or basic properties of the surface. Glass is an alkaline silicate, and a fatty acid wets it sufficiently well. It is interesting to trace whether the phenomena will be repeated in the case of an acidic surface, as for example with quartz. Experience shows that in this case the fatty acid wets the surface poorly and collects into small spheres because of the increase in surface tension. The COOH groups, instead of being attracted, are to a certain extent repelled from the surface, and the spectrum obtained characterizes this disorganization. Thus the chemical action either favors or hinders the orientation of the molecules.

Fig. 3.

1. If this phenomenon really occurs, one may expect to detect it also in the case of fatty acids on the surface of metals (Trillat, loc. cit.).

In this case a spectrum of complex form is often observed,

in the study of which one can detect, besides the ordinary spectrum of the fatty acid, yet another spectrum, evidently from oriented molecules of the salt, formed when the acid comes into contact with the plate (Fig. 3). This latter is often very intense and characterizes the substance under investigation.

For different metals it is different. Thus, the spectrum of the soap formed is sufficiently intense for Pb and Sb, still more intense for Sn, and very bright for Bi, Fe, Cu. Its intensity is small for Mo and Ni, which are only weakly corroded by the acid, while Au and Pt do not form soap spectra at all. These spectra also do not form for Al. Al retains three molecules of acid in a compound, and this prevents the formation of a molecule elongated into a single line, which is necessary for the appearance of layers.

Below we shall see applications of this phenomenon; before leaving it, however, let us note that by this method one may try to study the affinity of a metal for an acid and certain contact reactions.

1. Finally, the orientation of acids on water also gives interesting results. If a small amount of palmitic acid is placed on heated water, allowed to cool, and a spectrogram is taken by the Laue method, a series of concentric rings is obtained, corresponding to the spacings of the principal reticular planes of the acid, i.e. 36 Å. On the contrary, if one operates with the same substance finely pulverized, only rings of large diameter are obtained, characterizing microcrystals.

To explain this result, one has to assume that in contact with water the first rows of molecules become oriented, and that all the molecules are arranged parallel to one another. These phenomena of orientation occur for a large number of substances; they, as Langmuir assumes, underlie the phenomena of catalysis.

1. In conclusion, let us note that the study of the distribution of intensities among the various orders of spectra (Scherrer,⁶ Trillat⁵) points to one important circumstance. Elementary lamellae of compounds, whose molecule ...

ends in a chemically active group (fatty acids, alcohols, glycerides, etc.), are bimolecular, the two molecules being turned toward one another by their “backs.” The leaflets are monomolecular if at the end of the molecule there is no such active group (hydrocarbons, some ketones); a single glance at the spectra obtained makes it possible to say to which class the product under study belongs (Fig. 4).

Fig. 4.

We shall also see further the application of the study of the distribution of intensities in the spectrum to determining the position of the ketone \(C=O\) in the carbon chain (Schrieb).

Fig. 5.

Study of chemical properties

Let us now turn to the chemical side of the question.

Fatty acids. By the spectrographic method a complete series of fatty acids with an even number of \(C\) atoms was studied. It was from this study that the law was derived which was expounded by us at the beginning. Supplementing this series by studying saturated acids with an odd number of \(C\) atoms, the author found that they are not situated on the same straight line as the even acids (Fig. 5). A certain difference in the properties of even and odd—

even acids had already been noted by chemists: in fact, even acids are often encountered in nature and are easily prepared, whereas odd acids occur more rarely; alternations are also observed in molecular volumes and melting points (Garner, Madden, Rushbrooke^7).

The following table gives some of the results obtained:

Fig. 6.

Thus, in each series, even and odd, the increase in the distances between lattices with increasing number of C atoms is constant, but this increase is not the same in the two series. This result can be explained either by taking into account the manner of arrangement of the C atoms in these two series, or rather the different inclination of the molecules between the layers (Fig. 6).

An analogous phenomenon occurs in saturated acids; here too the even and odd acids do not lie on a single straight line.

Soaps. We saw above that fatty acids are capable of combining with metals, giving good diffraction spectra. The very light corrosion that results—weak, because very small amounts of substance interact—constitutes a salt of the metal, for example lead palmitate; the COOH groups anchor themselves in the metal, and the molecules, having obtained a solid base, orient themselves in parallel, grouping their ends in equidistant planes.

Let us take some metal, for example Pb, and study its etching by saturated acids. First of all, one interesting fact is found: fatty acids, which are liquid at ordinary temperatures, can be studied on glass only with the aid of complicated devices (the low-temperature chamber of Gibbs, loc. cit.¹⁹); these same acids, when simply applied to a lead plate, give very intense spectra, which are a natural continuation of the spectra of the soaps of solid acids. (By this method it is at present possible to reach the very beginning of the series and even to study acetic acid.)

Fig. 7.

We studied by this method all the fatty acids, beginning with the first members of the series and ending with the acid \(C_{32}\). In the photo—

As (Fig. 7) the convergence of the characteristic lines is clearly visible. In the case of distant terms of the series, with an exposure of \(3/4\) hour, reflections of the 10th–12th order can be obtained (the \(K_\alpha\) and \(K_\beta\) rays of Cu).

Fig. 8.

The points obtained lie very well on a single straight line; for each added C atom there is an increase in the distance between planes of 1.3 Å (Fig. 8). Only one straight line is obtained, regardless of whether the acids were even or odd; this result is explained by the fact that lead combines with two molecules of the acid, forming a salt. In addition, the dis-

distances between the lattice planes prove to be somewhat larger than those obtained for the acids on glass. Evidently, the lead atom is situated between two molecules and draws them apart.

Incidentally, let us note that the distances between lattice planes obtained in these cases are considerably greater than those that have ever been measured spectrographically: for example, the lattice constant of mica, which has the largest lattice constant among mineral crystals, is 10.1 Å; the greatest value obtained by Müller and Shearer is 54 Å; here the distance between the lattice planes reaches 92 Å. Such diffraction gratings have been used to study X-rays of large wavelength by purely spectrographic methods and have quite recently made it possible to close the gap between the extreme ultraviolet and X-ray regions of the spectrum (Dauvillier).

Fig. 9.

Glycerides. Let us pause further on fats, or glycerides. Their spectra are also characteristic and reveal the same law of increase. The fat molecule may, in rough outline, be regarded as a kind of “coat hanger” with three long hooks (Fig. 9). Such “hangers,” placed one upon another and folded back-to-back in pairs, form bimolecular layers.

Structure of the molecule of organic compounds with a long chain

All these results make it possible to indicate a general principle for the construction of molecules of substances with a long chain. These mole-

are built by the successive addition of a \(\mathrm{CH_2}\) group. This addition, according to some authors, proceeds along a broken line, in zigzags; in the opinion of other authors—along a straight line. The orientation occurs owing to the presence of an active group at the end of the molecule: the group of an acid, alcohol, glycerol radical, etc. Molecules, at first arranged at random, become normal or inclined on their substrate, while remaining parallel to one another, and in this way form a series of leaflets. Melting the substance or slight pressure on its surface helps to improve such a structure; in this way one can artificially obtain true crystals from substances which in external appearance are quite amorphous, such as fats, tallow, oil, lecithin.

Recent investigations have made it possible to establish with complete accuracy the position and form of carbon chains by a thorough study of well-built crystals of fatty acids and hydrocarbons (Mark, Meyer, Brill, Uaykov, Müller\(^9\)).

The results obtained have entailed a whole series of consequences, some of which have proved quite unexpected. However, let us leave here the field of molecular physics in order to proceed to the consideration of various applications of X-ray spectrography to organic substances with a long molecule.

Analytical Applications

The first application of spectrography by means of X-rays, one which is of rather great interest, is analysis.

If one thinks of the difficulties and the time required by chemical analysis of organic substances, with all its inaccuracy, one cannot but arrive at the thought that precisely here the radiographic method could have numerous practical applications.

We have already seen how, in a definite homologous series, it proved possible to determine the number of carbon atoms

of one of the compounds of this series, calculating the distances between nets from the positions of the lines on the X-ray photograph. In doing this, however, a certain caution must be observed, since the spectra may be influenced by the method of preparing the oriented layers (see below, isomorphism). Nevertheless, with small precautions, by this method one can determine within a few minutes the number of carbon atoms in a given chain: we applied this method to daturic acid, whose formula had been discussed many times and which proved to be a saturated acid with \(C_{17}\); the investigation was carried out by means of spectrograms of this substance, first on a glass plate, then on a lead plate (Appendix,\(^{10}\) Figs. 10 and 11).

Fig. 10.

Fig. 11.

a) Determination of the position of the ketone or ether group in the chain

A remarkable application of the spectrographic method was given by Sherrer and Seville.\(^{5}\) Studying the distribution of intensities in the spectra of various orders of ketones of the form

\[ \mathrm{CH_3(CH_2)_m—CO—(CH_2)_n—CH_3}, \]

they observed that the elementary lamellae are monomolecular (there are no active groups at the ends). Further, depending on the position of the CO group in the chain, certain orders were weakened or disappeared altogether: thus, if the CO group is situated in the middle of the symmetrical ketone chain

\[ \mathrm{CH_3—(CH_2)_n—CO—(CH_2)_n—CH_3} \]

the even orders disappear; but if the CO group occupies a position determined by the ratio:

\[ \frac{\text{distance between the CO group and the end of the molecule}} {\text{length of the molecule}}, \]

some orders were strengthened, others weakened, so that the very first glance at the spectra made it possible at once to determine the position of the CO group (Figs. 12 and 13).

Fig. 12.

A similar kind of investigation can be carried out on ethers with the formula

\[ \mathrm{CH_3—(CH_2)_m—CO—O—(CH_2)_n—CH_3}. \]

These two interesting examples characterize the importance of the method, which makes it possible, from purely physical considerations, quickly to determine the position of chemical groups in a chain in those cases where chemical methods are too long and uncertain. The complete theory of these phenomena has been given in part by Cevit and Pirenne, and in part by de Broglie and Trillat. \(^{12,6}\)

Fig. 13.

b) Analysis of mixtures

One may think a priori that the spectrographic method can render great service in such a delicate question as the study of mixtures of substances that are close in composition. Here, in a very large number of cases, the phenomena of isomorphism distort the results and do not allow one to obtain anything except doubtful numbers. Nevertheless it is possible, by adopting

some precautions, to draw correct conclusions, if one operates under strictly defined conditions. Thus, for the case of a mixture of triglycerides, we succeeded in obtaining spectra with very distinctly separated lines of trilaurin, tripalmitin, and tristearin (Trillat, Ann. de Phys., loc. cit. and C. R. Acad. Sc. 10). Likewise, in the case of a mixture of stearic and palmitic acids, it was possible to obtain lines different from the lines of daturic acid with \(C_{17}\) (Figs. 10 and 11). However, these methods are still only qualitative and do not make it possible to detect small quantities.

Phenomena of isomorphism

In obtaining all the above-mentioned results, it was assumed as self-evident that a definite long-chain organic compound, studied by the rotating-crystal method, always gives one and the same spectrum. However, this assertion is true only within certain limits. Piper, Malkin, Austin, de Boer, Clark, and Tibo 13 showed that one and the same substance may give two or even three series of characteristic lines corresponding to two or even three different spacings between the lattice planes, depending on the method of preparing the specimen (by melting, pressure, or evaporation of the solvent), depending on the temperature, and also depending on the purity of the product. Thus, for palmitic acid, in addition to the usual spacings of 35.6 Å, one may also find spacings of 41 Å and 39.08 Å; for stearic acid—39.9, 43.9, and 46.6 Å. Normal paraffin hydrocarbons likewise give, depending on the temperature, two different lattice constants.

These results show that certain compounds can crystallize in several forms; each of these forms is stable under quite definite external conditions, which are difficult to establish exactly. The different forms are probably obtained owing to different inclinations of the molecules with respect to the normal to the lamellae.

Studies of mixtures of fatty acids (stearic acid, palmitic acid) carried out by Piper and by the author (Trillat) showed that the higher acid compels the lower one to be arranged in the same lattice in which its own molecules are arranged; thus, with the exception of a very high content of palmitic acid, in its mixture with stearic acid there is observed a spectrum corresponding to spacings close to, or somewhat smaller than, the spacings of the lattice planes of stearic acid. Conversely, in a mixture of acids with a sufficiently different number of C atoms—for example, in the case of a mixture of lauric or myristic acid with stearic acid—the spectra of each of the acids are observed. Here there is present the same phenomenon as in the paraffins studied by Clark: the spacings between planes vary depending on the rate of crystallization—the more time given to the molecules for orientation, the greater these spacings become.

One remarkable fact should be noted: a product such as paraffin, containing up to 18 hydrocarbons with different chain lengths, forms equidistant layers and gives a spectrum as if it were a single substance. In this direction, further investigations are evidently necessary.

Thus we are faced with a large number of problems that require explanation. The study of long-chain compounds is apparently not as simple as it seems at first glance. Measurements of the spacings between lattice planes of these bodies, if they are used as a means of identification, must be carried out with great care; nevertheless it should be noted that the “normal” spacings between lattice planes, i.e. those discussed in the preceding paragraphs and those most often obtained in investigations, are always obtained when fatty-acid preparations are made by melting. Applying this method, the author observed, in many hundreds of experiments, always only one group of lattice spacings—namely those that follow the laws stated above. In addition, the method of separating met--

crystal plates with fatty acids makes it possible, as it were, to avoid errors connected with isomorphism. In the soaps formed in this way this phenomenon has not yet been observed.

APPLICATION TO THE PHENOMENON OF LUBRICATION

Oiliness

According to Bragg’s assumption,^16 the explanation of the phenomena of lubrication should be sought in the layered structure of thin oriented films. In such a formation the lateral sides of the molecules are firmly bound to one another, whereas the methyl ends have weak cohesion; therefore the layers readily slide over one another. Graphite has a similar structure: the study of its spectrum indeed leads to the idea of the existence of sheets of equal thickness, the cohesion between which is very weak. It is also known that graphite is often used as a lubricant.

It would be interesting to test this hypothesis on actual lubricating fats, and not on crystallizing substances such as stearic acid (Trillat.^14).

It was indicated above that if a fatty acid—whether solid or liquid—is applied to a metal plate, and the spectrum of this acid is then taken by the rotating-crystal method, then, in addition to the spectrum of the acid itself, even in the case where this acid is solid, one obtains a spectrum characterizing the corrosion of the metal by the acid. In Fig. 3 we have already seen some of these spectra. Their interpretation and calculations show that, when a metal is corroded by an acid, very thin films are formed, in most cases invisible, and that these films possess a structure resembling a pack of playing cards: COOH penetrates into the metal, and the molecules, having thus obtained a solid base, are oriented in parallel, grouping their ends in equidistant planes.

Exactly the same phenomena occur when one is dealing with a mineral oil possessing free acidity,

or with oil in which a fatty acid is dissolved. The acid molecules, very active and possessing a marked affinity for the metal, are attracted to the metallic surface, become anchored in it, become oriented, and form a series of lamellae of one and the same thickness, arranged one upon another. The metallic surface is thus replaced by a surface composed of groups \(CH_3\), and the intense force field of the metal proves to be partly neutralized and replaced by the considerably weaker field of the ends \(CH_3\); therefore the friction caused by molecular attraction is reduced.

The hydrocarbon molecules slide easily over this new surface. The lamellae formed in this way from oriented molecules readily slide over one another with zero, or almost zero, friction, while resisting deformations in all other directions. Comparison with soap-bubble films gives a sufficiently accurate conception of this structure.

The affinity of a fatty acid for the metallic surface makes it possible to ensure the stability of the oil film. It has been experimentally established that mineral oil containing no active molecules is not stabilized, and in fact there is no, or very little, reason for the molecules to become fixed in the metal. In contrast to this, the very same oil containing fatty acids will be stabilized and will adhere firmly, owing to the formation of a layered and oriented structure inside the film (Bowden and Tabor).

However, fatty acids are not the only substances capable of being vigorously oriented on the surface of metals. We have been able to show, for example, that the same phenomenon occurs for triglycerides. Here the matter is not one of chemical affinity, but of a purely physical interaction caused by the presence of active ends. This fact was checked for all solid glycerides: tripalmitin, tristearin, trimyristin, and so on. As indicated above, the molecule of these substances may be likened, as it were, to a coat rack with three hooks, the active part being the glycerol radical;

namely, it is fixed on the support, then three chains rise vertically (Fig. 9); the second molecule faces the first and is arranged in the opposite direction. The elementary leaflet here contains two molecules.

The planes of CH₃ are planes of easy slip; indeed, the mutual activity of the methyl ends is very weak, and moreover, from the distribution of intensities in spectra of different orders one can discover that between two CH₃ planes arranged in parallel there exist, as it were, empty spaces.

With a minimal force tangent to these planes, slip or easy folding occurs; it is precisely for this reason that these substances produce the sensation of fatness and slipperiness. It is mainly in this peculiar structure that one must seek the explanation of the phenomena of oiliness.

The property of oiliness, according to the results obtained with the aid of X-rays, is determined chiefly by the ease with which a substance can orient itself and assume a layered or lamellar structure. Whether oiliness is a permanent property of substances or something accidental, artificially induced—X-rays must give us the answer to this question.

Let us turn to such substances as lard, lubricants for taps, thick greases, etc., which have an amorphous structure and possess a viscous consistency.

Let us place a small quantity of one of these substances on a glass or metal plate, cover it with another plate, and give it a to-and-fro motion relative to the first, while at the same time pressing the plates against one another. In doing so one may notice that the resistance to motion, large at first, decreases progressively as the lubricant is distributed and the layer becomes thinner. The two plates can then be separated by sliding; each of them proves to be covered with a thin layer of lubricant. The specimen prepared in this way can be investigated by the rotating-crystal method (Trillat¹⁴).

For the majority of the substances studied by this method, the appearance of orientation spectra has been established, sometimes

very intense. These spectra show that the lubricating film has become stratified owing to the motion of one plate relative to the other. In some cases a simple rubbing with the finger is sufficient to produce the appearance of these spectra.

Finally, if the lubricant is applied to a heated plate and then allowed to cool, stratification occurs only very weakly, or does not occur at all, and the spectra disappear.

When machines are in motion, a stratified film of lubricant is formed between the rubbing parts in the same way. If lubricant is applied to the axle of a wheel, then, when the wheel rotates, there occurs first of all a kind of “rolling out” of this lubricant into a thin layer, as a result of which the substance acquires a layered structure. It is precisely to this stage that the resistance corresponds which freshly lubricated parts of a machine oppose to motion. After stratification has occurred, the sheets that have formed slide one over another, like playing cards.

The stratification observed for lubricating materials involves a fairly thick layer, since, in order for distinct diffraction phenomena to take place, a stack of at least 100–200 sheets is required. Moreover, it is very probable that the principal substances entering into the composition of a lubricant stratify each in its own way, as may be seen in some cases where two or even three spectra superposed upon one another are observed. Apparently, the molecular orientation which underlies these phenomena is established at the surface of glass or metal in exactly the same way as in the case of fatty acids deposited on metals.

Finally, it is appropriate to note that the stratification phenomena observed in lubricants can be used for analysis. Indeed, by calculating the distances between the lattices and studying the distribution of intensity in the spectra, one can judge the composition of the lubricating substance. Likewise, by using successively glass and metal plates, one can detect the substances included in the lubricant.

acids, which manifest themselves by the appearance of new spectra when deposited on metal.

Finally, this extremely sensitive method also makes it possible to investigate changes in lubricated surfaces...

From these investigations the following important circumstance follows: oiliness is a hidden property of a substance; it manifests itself under the action of external physical agents, such as pressure or friction, and in essence the old method of assessing the oiliness of a lubricant by rubbing it with a finger on the palm of the hand here receives its full foundation.

Despite the rather considerable difficulties of the investigation, these results have been extended to the case of liquid lubricating substances. In liquid lubricants, such as vaseline oil containing oleic acid, compressed between metallic or glass plates, we also found a layered structure. Evidently all molecules, provided only that they are sufficiently asymmetric, are capable of orienting themselves under pressure, and in some cases, moreover, of arranging themselves as parallel leaflets. This is precisely what must occur each time a thin layer of oil is compressed and “rolled out” between moving parts. Such stratification is obtained not only in the case of lubricating oils—we detect it by means of X-rays in rolled metals, in films, and when investigating the structure of rubber or drawn glass. Asymmetric molecules of a liquid mass subjected to the action of considerable tensions or pressures respond to this action by a more or less perfect orientation, which manifests itself in almost all cases in significant optical birefringence (Trillat[^17]). This phenomenon of orientation, accompanied by stratification, is a general property of lubricating oils, regardless of whether the molecules of these substances are active or inactive. However, the presence of active molecules significantly intensifies this phenomenon; in this case, in addition, adhesion occurs—or, more precisely, adsorption on the surface of metals—which imparts to the layers those special qualities which, properly speaking, characterize ...

...call forth the phenomena of oiliness. In the case of mixtures of active and inactive molecules, the former are selectively adsorbed owing to the intensity of the molecular field that they form, or to the affinity that they possess with respect to the metal. The fact that, in a mixture of molecules, spectra of several kinds can be observed shows that the stratification proceeds according to rather complex laws, and that the layers contain both substances intermixed. As for inactive molecules, having a weak affinity, they are situated on top of these layers.

Thus one may put forward a hypothesis which is partly confirmed by the results of X-ray investigations, as well as by numerous other experiments: molecules react to pressure by opposing it and, in doing so, orient themselves so that their total energy is minimal; they arrange themselves perpendicular to the metal surface and side by side, forming rows of molecular packing. The orientation should not be limited only to one layer, since the ends of the molecules turned toward the liquid also have a force field, although a weaker one in comparison with that formed by the molecule itself, and of course considerably weaker than that produced by the solid body. But, owing to its regularity, the field of the ends of the molecules of the first layer is sufficient to orient a new layer of molecules or to permit the formation of a new layer from molecules capable of orienting themselves; this leaflet is in turn covered by a third layer of molecules, etc., with each layer being oriented under the influence of the preceding one, until some external disturbances prevent the formation of new tiers. Molecules with identical ends, which are little active, as for example in the case of saturated hydrocarbons, form monomolecular, poorly adhering layers, since the field of such molecules is weak. In the case of acids, triglycerides, whose molecules are asymmetric, the leaflets are bimolecular and their adhesion is stronger owing to the greater magnitude of the field they form.

Application to the Oxidation of Unsaturated Acids

Yet another application of this method of investigation can be made to the drying of paints, or essentially to the oxidation of unsaturated fatty acids. This question is extremely complex and far from clear. The results obtained, however, provide some new information (Trillat, Ann. d. Phys.).

Unsaturated fatty acids thicken in air; this phenomenon is analogous to the hardening of drying paints. The hardening is accompanied by polymerization and oxidation reactions, which make investigations almost impossible.

Various fatty acids extracted from linseed oil were poured onto a lead plate; two or three drops are quite sufficient for interaction to occur and for the surface of the metal to be cleaned. Lead was not chosen by chance—it oxidizes very rapidly and consequently should promote the hardening of drying paints. It is precisely for this reason that all paints contain small quantities of lead, or manganese and cobalt, which act in a similar way.

If the lead plate is placed on the table of the spectrograph and radiographs are taken one after another over the course of 2 or 3 hours, it can be observed that, as the liquid layer thickens, the spectra obtained change (Fig. 14).

In Fig. 14 one can see that the number of bands decreases and that they become less and less distinct as drying proceeds. Finally a moment arrives when the layer becomes elastic and solid; the substance has set; it has dried. From this moment it is no longer possible to obtain a spectrum; on the prints there is a continuous veil, not intersected by any even barely noticeable lines.

The study of this phenomenon led to the following hypotheses: the molecules of fatty acids lengthen upon the addition of an oxygen molecule. Each molecule can retain a definite number of oxygen molecules, determined by the degree of its unsaturation. With each addition, it

lengthens by a certain amount. This growth creates an unstable structure, like a building to which ever more new stories are being added. Physically, the lengthening of the molecule is manifested in a decrease in the number of lines, and its instability—in their blurring.

Fig. 14.

Thus there comes a moment when the structure breaks and collapses; the fragments join together, forming complex and varied compounds; this stage corresponds to the setting of the paints. Naturally, in the photographs, from this moment on, the characteristic lines are no longer obtained, since it was precisely the presence of long molecules, parallel to one another, that caused the appearance of planes reflecting X-rays. Now before us is a more or less homogeneous mass, without a well-pronounced stratification—in a word, “a true amorphous substance.”

Fig. 14a.

Fig. 14b.

We have already seen how interesting the application of X-ray spectrography to organic chemistry has proved, both from the physical and from the chemical side. In particular, as regards the so frequently encountered state of stratification, we have shown what applications the method of X-ray analysis may find here in the study of molecules, and also of certain phenomena, such as the phenomenon of lubrication. The study of these layers has proved very fruitful for the new physics, and most of the works carried out in this direction have found their justification precisely thanks to the application of the method of spectrography by means of X-rays, which is one of the best for investigating these phenomena.

Before concluding this article, let us try to point out also the general properties of organic compounds with long

molecule and substances having a plate-like structure (Bragg^16).

General properties of organic compounds with a long molecule and of substances having a plate-like structure

By the method of spectrography it has been possible to show that substances with a long chain, oriented by glass or metal, possess a structure analogous to the structure of soap bubbles or films of soap on the surface of water. The conditions that lead to the formation of films on the surface of water, to the layered structure of fatty acids and of other similar bodies, also occur in the case of the formation of a large number of solid crystals. They give rise to a special layered state; the crystals split very easily into thin layers, which slide over one another, producing a sensation of oiliness. In this, the lateral bonds of the molecules must be stronger in comparison with the bonds at the ends; thus the layers can be very easily separated from one another. The layered structure of these substances is in all cases caused by one and the same reasons—the general orientation of the molecules across the layer and the presence of active ends (COOH, OH, etc.).

If we review the principal properties of this large class of plate-like crystals, we shall see that Langmuir monomolecular layers and Perrin’s multiple spreading, as well as the whole class of “thin films,” have much in common with the solid bodies of this class. There is, however, also a certain difference between them.

In fact, the layers which we have studied and which represent perfect crystals of this class of bodies in the plate-like state are bimolecular; thus it is clear why, for example, a regular crystal of oleic acid has twice as many oriented particles as a monomolecular film of the acid on water; a monomolecular layer is composed of molecules raising their chains upward, whereas for a complete crystalline structure the presence of molecules with an opposi-

in the opposite direction of the chains. If it is assumed that two films of oleic acid—one with molecules in one direction, the other with molecules in the opposite direction—come into contact, we shall have all the conditions necessary for the formation of a crystalline structure. In this case the carboxyl groups meet and combine; each molecule of the upper group is bound to two or four molecules of the lower one and conversely. The form obtained in this way is much more stable than a monomolecular film; owing to this circumstance the two films tend to increase the area of their contact, which is the cause of the phenomenon of the spreading of fatty acids on water.

The layer obtained in this way represents a true crystal, despite the fact that oleic acid melts at \(10^\circ\); its crystalline structure is reproduced also in the liquid state.

Other films may form on the first film, both owing to the weak attraction of the methyl groups among themselves and owing to the pressure of the air. These layers slide easily over one another and possess the same structure as genuinely crystalline oleic acid. In solid crystals, however, the layers do not slide so easily; even if they do slide over one another as a whole, like a deck of playing cards, there is friction between them which restrains them. The cause of this friction is not entirely precisely known; possibly it is a molecular effect, and then sliding does not occur until the bonds, already weak, have been weakened still further (for example, by proximity to the melting point). This effect may also be caused by an imperfection of the crystalline structure.

The phenomena of molecular orientation are, finally, of great interest in the study of catalysis and adsorption.

Indeed, reactions depend essentially on the orientation of the molecules and on their distances, as well as on the nature of the planes that are in contact, as Langmuir showed.

Therefore, as the basis for these contact actions it is necessary to discuss the arrangement of molecules on the surface: both those which existed before the reaction and those which appear there after it. Precisely these ideas were laid at the foundation of the works of Harkins, Adam, Perrin, and many others. Investigations carried out with the aid of X-rays have provided new data for this theory and have made it possible to hope that this question, still obscure until now, may be somewhat illuminated.

BIBLIOGRAPHY

Application of X-rays to the study of organic compounds

1. Perrin, Annales de Physique, 9 série, t. X, p. 160, 1918.

2. Langmuir, Journ. Amer. Chem. Soc., C. S. 39, 1865, 1917; Journ. Amer. Chem. Soc. p. 1865, July-Dec. 1922.

3. Devaux, J. de Phys., t. IV, 9, sept. 1923; Rev. gén. Sc., No. 4, p. 143, 1913.

4. A. Müller, Trans. Chem. Soc. 123, 2043, 1923. G. Shearer, Trans. Chem. Soc. 123, 3152, 1923. A. Müller et G. Shearer, Trans. Chem. Sos. 123, 3156, 1923; Chem. Soc. Report on Crystallography, 1923.

5. Müller. Structure des acides stéarique et stéarolique. Nature, p. 45, 11 juillet, 1925. W. Bristol Saville et G. Shearer. An x-Ray investigation on saturated aliphatic ketones. Journ. of the Chem. Soc., vol. 127, 1925. R. E. Gibbs. An x-Ray investigation of the lower members of the fatty acid series. Journ. of the Chem. Sos., vol. 127, 1925, Müller and Saville. Further x-Ray measurements on long chain Compounds (N-hydrocarbons). Journ. of the Chem. Sos. 127, 1925. Jönes, Chemistry and Industry, I août 1924, La molécule des acides gras. Piper et Grindley. Structure fine de quelques sels de sodium des acides gras dans l’eau de savon. Proc. Phys. Soc. Londres, 15 août 1923. J. J. Trillat, Ann. de Physique X-série, t. VI, p. 1 à 101, juillet—août 1926.

6. G. Shearer, Proc. of the Roy. Soc. A., vol. 108, p. 655, 1925.

7. Garner, Madden et Rushbrooke, Journ. of the Chem. Soc. p. 2491, 1926.

8. Dauvillier, journal de Physique, Janvier 1927.

9. Müller. A further x-Ray investigation of long chain Compounds (N-Hydrocarbon), Proceed. of the Royal Soc. A., vol. 120, p. 437, 1928, 114, 542, 1927. Meyer et Brill. Röntgen Untersuchungen an Laurinsäure. Z. Krystallographie, B. 67, Heft 5/6, p. 570. Wyckoff, Hunt et Merwin, Z. Krist. 62, 553, 1925. K. Meyer. Neue Wege in der organischen Strukturlehre und in der Erforschung hochpolymerer Verbindungen. Z. f. angew. Chem., No. 34, 1928.

1. J. J. Trillat, C. R. Acad. Sciences, 184, 812, 1927. Sur l’interprétation analytique des spectres d’acides gras et de leurs mélanges. Ann. de Phys., X-e série, t. VI, p. 81, juillet-août 1926.

2. Shearer, Proceedings of the Royal Soc., A., vol. 108, p. 655, 1925.

3. L. de Broglie et J. J. Trillat, C. R. Acad. Sciences, p. 1485, 4 mai 1925.

4. Clark. The variabiability of long diffraction spacings in paraffin Waxes Public. from the Massachusetts Institute of Technology, sept. 1927. J. Thibaut, C. R., Acad. Sciences, 184, 24. 1927; 184 96, 1927. Piper, Malkin et Austin. An x-Ray Srudy of some structural modification of long chain compounds. Journ. of te Chem. Sos., p. 2310, 1926. De Boer. Différentes modifications des acides gras, Nature, janvier 1927; x-Ray investigation of the polymorphism of fatty acids, Nature. 30 aprill 1927.

5. J. J. Trillat, Recherches, au moyen des rayons X, des phénomènes de lubrification C. R. Acad. Sc., p. 843, 29 mars 1926; et la technique du graissage. Science Industrie, No. 158, 1927 et No. 169, 1928.

6. P. Woog, Contribution à l’étude du graissage, onctuosité et influences moléculaires. Delagrave, éd. (Paris) 1926.

7. Bragg, Nature, 21 février 1925.

8. J. J. Trillat, C. R. Acad. Sciences. Sur les phénomènes d’orientation dans les gels colloïdaux, t. 188, p. 1246, 1929; Journal de Phys., janvier 1929.

9. J. J. Trillat. Orientations de composés organiques par des surfaces cylindriques de verre. Applications, C. R. Acad. Sc., 18 février 1929 p. 555.

Thin films. Monomolecular layers. Capillarity

Langmuir, Journ. Amer. Chem Sos., 39, 1865, 1917.

Wells, Ann de Phys., t. XVI, p. 108 (soap films), 1921.

Devaux. Détermination du poids moléculaire de l’acide abiétique au moyen de lames minces. Étude des propriétés de la surface du mercure. Procès verbaux, Société Sciences physiques de Bordeaux, 1926—1927.

R. Dubrisay. Applications de la mesure des tensions superficielles à l’analyse chimique. Mémorial des Sciences physiques, fasc. VIII, 1929.

Devaux, Structure moléculaire de la cellule et des tissus vivants. Journ. de Phys. p. 34, 2 mars 1928.

Mouillabilité des surfaces solides. Soc. de Phys., 16 février 1923. Membrane de coagulation par simple contact de l’albumine avec l’eau. Soc. Sc. phys. et nat. de Bordeaux, janvier 1904.

Perméabilité des lames minces, Soc. de phys., mai 1921.

A. Marcelin. Solutions superficielles, fluides à deux dimensions. Soc. Ch. Phys. 10 décembre 1924.

Solutions superficielles et loi de Gay-Lussac (Marcelin et Delaplace) Soc. Ch. Phys. 23 décembre 1925.

Épaisseur des couches très minces à la surface de l’eau. Diplôme d’études supérieures, Paris, 1913.

Solutions superficielles à deux dimensions. Ann. de Phys., t. IV, X-e série, nov.-déc. 1925.

R. Marcelin. Structure des cristaux en lames très minces. Ann. Phys., 9-e série, t. X, déc. 1918.

J. J. Trillat. Les lames minces, leurs structures et leurs dimensions. Rev. scientifique, № 17, septembre 1926.

Frazer, Phys. Rev., janvier 1929, p. 97.

Bouhet. Polarisation elliptique produite par réflexion à la surface des solutions d’acides gras dans l’eau. C. R. Acad. Sc., 2 janvier 1929, p. 59; 1 juillet 1929, p. 43; 26 décembre 1928, p. 200.

Fahir Emir. Nouvelle détermination de l’épaisseur d’un film d’acide oléique, C. R. Acad. Sc., 13 mai 1929, p. 1284; 24 juin 1929, p. 1667.

Lecomte du Noüy. Equilibres superficiels des Solutions colloïdales. Masson, éd. Paris 1929.

M. Kladam, Proc. Roy. Soc., vol. 101, p. 452, 1922; vol. 110, 1925; vol. 120, 1928; vol 110, 1928.

Mécanisme moléculaire des phénomènes capillaires, 4 avril 1925. Cohésion in surfaces films. Trans of the Faraday Soc., 81, XXIV, février 1928.

Insoluble films on water surfaces (Afam et Jossop), ibid, 71, XXII, décembre 1926. Angles of contact and Folarity of Solid Surfaces. Journ. o the Chem. Soc., v. 127, 1926.

Kennard. La cause de la tension superficielle. Nature, 31 octobre 1925.

Schofield et Rideal, Proc. Roy. Soc, A, 109, 1925; 110, 167, 1926.

Cary et Rideal, Nature, 28 mars 1926.

Schultze. Capillarité et mouillage. Koll. Z. 37, 10, 1925.

Labrouste. Transformations moléculaires dans les couches minces à la surface de l’eau. Thèse, Paris, 1920.

Lord Rayleigh, Proc. of the Roy Soc., t. XLVII, mars 1890.

Guyot. Effet Volta métal-électrolyte et couches monomoléculaires, Thèse, Paris, 1924.

Delaplace. Extension de la loi de Gay-Lussac aux solutions superficielles. C. R., v. CLXXX, p. 2020.

Nietz. Molecular orientation at Surfaces of Solids. Journ. of Phys, Chem. février-april 1928.

Keenan. Formation de films mines sur les surfaces de mercure. Journ. of Cham., XXXIII, № 3, p. 371, 1929.

Organic substances with a long chain

Becker et Rose, Z. f. Physik, 14, 369, 1923; et 5, p. 61, 1921.

Prins. Higher Order X-Ray reflections from fatty acids. Nature, 17 juillet 1926; Nederl. Tydschrift voor Natuurkunde, p. 305, 1926.

Prins. Ueber die Beugung von Röntgenstrahlen in Flüssigkeiten und Lösungen. Z. Physik, B. 56, H. 9/10, 1929 (see p. 642 et seq.).

Lubrication

Hardy and Bircumshaw, Proc. Roy. Soc., 108, May 1925.

Hardy and Doubleday. Boundary Lubrication. Proc. Roy. Soc., A, 100, p. 550; A 101, p. 487; A 104, p. 25.

J. J. Trillat. Researches, by means of X-rays, on the phenomena of lubrication. C. R. Acad. Sci., 29, p. 843, March 1926. The technique of greasing. Science et Industrie, 158, 1927, and No. 159, 1928.

P. Woog. Contribution to the study of greasing, oiliness and molecular influences. Delagrave, ed., 1926.

Petrow, Reynolds, Sommerfels, Michell. Hydrodynamic Theory of Lubricant Friction. Ostwald’s Classics of the Exact Sciences, No. 218, Leipzig, 1927.

Von Dallwitz-Wegener. On New Ways to Investigate Lubricants, R. Oldenbourg, Munich—Berlin, 1910.