12 October 1954

Dear Peter:

We had a visit from Rosalind Franklin the other day, and she showed some beautiful x-ray photographs. I have been thinking about them since then, and you might be interested in some vague ideas that I have had.

First, she has made a wonderful photograph of oriented tactoids of tobacco mosaic virus. The photograph is like one obtained by Bernal and Fankuchen about 15 years ago, but is much better. The material is tobacco mosaic virus plus some water. A liquid crystalline phase is produced, which is birefringent, although fluid. This is due to the lining up of the long rod-like molecules parallel to one another. The molecules are about 150 A in diameter, and I think that there is about a 100 A layer of water in between them, so that they are about 250 A apart. They are in hexagonal packing, with the axes nicely parallel to one direction, but with arbitrary orientation of the molecules around there axes, and arbitrary location of the ends in the axial direction. As a result, the x-ray photograph shows excellent layer lines, with a spacing corresponding to an identity distance of 68 A. The layer lines are, however, essentially continuous, and their intensities represent the form factor of a single rod-like molecule. You remember that Watson interpreted Bernal and Fankuchen's photograph a couple of years ago as showing probably 31 units in 3 turns of a helix.

She made the photograph by sucking the liquid crystalline material into a fine capillary, and then allowing it to stand for a few weeks, in the sealed-off capillary. The glass walls produce orientation of the rods in the direction of the axis of the capillary. The structure is apparently quite perfect.

We should like to do something similar on the tactoids formed by sickle-cell-anemia hemoglobin, but I am not sure that anyone here has good enough technique and enough patience to do the job. Also, our microfocus tube is not yet in operation. Dr. Franklin took photographs on a microfocus tube, using exposure times of about 40 hours.

It has occurred to me that it might be worth while to investigate some ordinary liquid crystals with x-rays. I have a dim memory that there is some stuff in the literature about diffraction patterns of liquid crystals, but I have not looked it up. Probably some German did it.

I do not remember what the good examples of liquid crystals are -- a great many have been discovered. In particular, I do not remember what substances it is that form liquid crystals of the nematic type, but I think that a lot of these are known. If you wanted to check up, you could look in the book by O. Lehmann, Flüssige Kristalle, published about 1920. I thought that we had the book in our chemistry library, but it seems not to be here. You no doubt have it in Cambridge.

I think that it might well be that nematic liquid crystals formed by simple organic compounds, such as p-azoxyanisole (I am relying on my memory here, and am probably wrong about the compound) contain rods with a helical structure. These rods might be formed by an interaction of a molecule with the next molecule, probably through the formation of hydrogen bonds, that represents a rotation and translation, and then by the repetition of this operation a helix would be made. I doubt that anyone has tried to study nematic liquid crystals by the technique used by Dr. Franklin. The preparation of some photographs of liquid crystals of this sort, and their analysis to obtain the structure of the helical rod, and even, through the interpretation of the intensity distribution along the layer lines, the arrangement of the atoms in each molecule, would be something new. I should think that it would be worth while to make preparations of a large number of organic compounds that form nematic liquid crystals, and to get x-ray photographs of them, and attempt to determine the number of molecules per turn in the helix, the pitch of the helix, and the identity distance along the rod. Then one or two of them might be completely analyzed. I think that in general the job of determining a structure of a molecule in this way would be simpler than that of determining the structure of a true crystal of the same substance. The reason for this is that the determination of the structure of a molecule by analysis of the electron diffraction pattern given by the gas is simpler than that of determining the structure of the crystal, because the orientations of the molecules and their relative positions in the crystal do not need to be determined in the investigation of the gas.

I am sure that most of the liquid crystals that Lehmann describes have a range of stability of the liquid crystalline state that is above room temperature. That is, of course, an experimental difficulty.

You probably could get the compounds from the chemistry department -- I think that a good number of them are rather ordinary ones, but might well be available. Others might be purchased.

It may well be that you do not want to waste time checking up on this point. If so, let me know, and perhaps I shall ask somebody here to look into the matter.

I have just rewritten Chapter 7 of College Chemistry. I had revised it somewhat last month -- it is a chapter on descriptive chemistry, and I had put the alkali metals, alkaline-earth metals, and several other groups of elements into the chapter. It seemed to me to be rather dull, however, and I decided to scrap the whole thing, and to start over again. The result is a chapter on the chemistry of carbon, including a good bit of elementary organic chemistry. I think that it is a much more interesting part of chemistry to have at this place in the course than the survey of descriptive chemistry for a number of elements.

Mama and I are planning to be in Pasadena during most of the winter. At present we have only two trips scheduled, to Princeton next month, and to Harvard in late April. I expect also to make a short trip to Boston in December, to attend the meeting of the Scientific Advisory Board of Massachusetts General Hospital.

Dr. Corey and I are still feeling bad that there is no one here who knows how to work with crystalline proteins, and to get good x-ray photographs of them. Holmes made a nice x-ray photograph of carbonmonoxyhemoglobin many years ago -- I think, in fact, before Bernal and Crowfoot did their first work, but probably it was about simultaneously. He did not publish anything about it. He is, of course, too busy to tackle such a difficult job, requiring lots of time, now. Also, I think that Dr. Corey has had some experience, but he is busy with other things. Tang started off well, but after one year as a post-doctoral fellow he went back to China, with only one protein crystal investigated.

Love from

[Linus Pauling]