On the Formation of a Heat-Reversible Gel .
Summary .
1 .
By magnetic deflection of the a-particles from thorium active deposit the result of Marsden and Barratt , that the a-particles of 4'8 cm .
and 8'6 cm .
range do not occur in equal numbers , has been confirmed .
2 .
Chemical , physical and recoil experiments alike failed to separate the-atoms giving the a-particles of the two different ranges .
3 .
The experiments show that thorium C emits very penetrating / 3-rays ( / 1 = 13*5 cm.-1 Al ) practically unaccompanied by 7-rays , while thorium I\gt ; gives relatively soft / 3-rays ( / / .
= 2P5 ) with intense penetrating 7-rays .
4 .
The results suggest that of the atoms of thorium C , 35 per cent , emit a-particles of range 4'8 cm .
and become converted into atoms of thorium Dr while the remaining 65 per cent , emit / 3-particles and disintegrate into atoms , of a very short-lived a-ray product thorium C2 ( range 8-6 cm .
) .
We wish to acknowledge our indebtedness to Prof. Butherford for his ever ready advice and encouragement .
On the Formation of a Heat-Reversible Gel .
By W. B. Hardy , F.R.S. ( Received April 11 , \#151 ; Read May 16 , 1912 .
) In the course of his study of the cyclic ketonic compounds , I)r .
Ruhemann has discovered a substance which is , I think , destined to play an important part in the development of the theory of gelation .
The substance is 5-dimethylaminoanilo-3 :4 diphenylcyclopentene-1 : 2 dione , * the formula , being c6h5c\#151 ; c : nc6h4n(ch3)2 c6h5c CO \o It is a solid at ordinary temperature , and occurs either as orange-coloured needles or as small dark red plates ; the former separate from dilute solutions , the latter from concentrated solutions in absolute alcohol .
" If a little water is added to the hot alcoholic solution , no crystals * Ruhemann and Naunton , 'Chem .
Soc. Trans. , ' 1912 , vol. 101 , p. 42 .
[ Apr. 11 , Mr. W. B. Hardy .
separate on cooling , but the whole sets to a transparent jelly , which , according to concentration , is yellow or yellowish-red .
This gel , when kept at the ordinary temperature , gradually liquefies , and in the course of one or two days completely disappears , with separation from the resulting solution of the azomethine in orange needles .
" Dr. Ruhemann has therefore furnished us with a substance which offers special facilities for studying the relations between the crystalline and the colloidal state .
The colloidal characters do not appear in any of the substitution products of the azomethine which have been prepared .
They completely disappear when a single hydrogen atom is displaced by , e.g. , bromine .
Why so small a .change in the molecule should cause the complete disappearance of such striking physical characters is an important problem , which I understand Dr. Ruhemann is investigating .
I have to thank Dr. Euhemann for his kindness in providing me with a small quantity of the azomethine , and for allowing me the opportunity of examining the remarkable changes of state .
In the following pages is given an account of a purely qualitative study .
Influence of the Solvent.\#151 ; The action of water in promoting the change of solutions in alcohol to the colloidal state ( and I may also add solutions in aldehyde and glacial acetic acid ) led the authors to suggest that the phenomenon is accompanied by the formation of a hydrate , which spontaneously loses water , and so yields the original compound in the crystalline form .
My own experiments prove that , if water be necessary for the .change , the most minute traces are sufficient .
I found it easy to obtain a gel at 10 ' with a good sample of absolute alcohol , provided the solutions , made at the boiling point of the alcohol , were sufficiently concentrated .
Wrater lowers the concentration of solute needed for gelation .
It was immaterial whether red or orange crystals were used for making the solution .
The substance is completely insoluble in water , which , therefore , may be said to act by lowering the degree of super-saturation needed for gelation , which occurs only in solutions supersaturated with respect to crystals .
In unit volume the added water displaces so much alcohol , and combines with some of the alcohol actually present , and so in both ways increases the mass ratio of solute to alcohol .
Water also delays crystallisation , and thereby alters the time relation in such a way as to make it easier to reach the limit of overcooling , at which gelation occurs without deposition of crystals .
This is especially seen when ether is the solvent .
More rigorous experiments to eliminate water are the following:\#151 ; Absolute alcohol , prepared by standing on calcium chloride for a week , and prolonged boiling with calcium chloride under a vertical condenser , and 1912 .
] On the Formation o f a Heat-Reversible Gel .
subsequent distillation , was again kept over a large quantity of freshly fused chloride for a week .
It was then distilled slowly , the first quantity to come over being rejected , and the middle portion received directly , with exclusion of air as far as possible , into a quartz test-tube , which had been strongly heated , and allowed to cool in a desiccator .
Crystals of azomethine which had been exposed to an air current for four hours at 120 ' were at once added , and a solution made by heat , but without boiling .
In three minutes the solution was poured off the remaining crystals into another dry quartz tube , and placed in a freezing mixture of ice and salt .
It at once set to a firm gel .
Ethylicether made anhydrous by distillation from sodium and exposure to a large surface of sodium for a week was then tried in a similar way , but as the substance is only slightly soluble in ether it was necessary to concentrate by rapid boiling .
When the solution saturated at the boiling point of ether had been concentrated to one-third its volume it was placed in the freezing mixture and gelation occurred .
Carbon tetrachloride is a very good solvent for the substance and deep orange-red solutions are readily obtained .
This solvent is immiscible with water .
A solution made in a dry quartz tube of red crystals with solvent which had stood for 48 hours over a large excess of freshly fused calcium chloride gave a firm gel in a freezing mixture .
The most probable conclusion from these and similar experiments with other solvents is that water is not necessary .
With the help of a freezing mixture a gel has been obtained from solutions in ethylic alcohol , ethylic ether , aldehyde , acetone , carbon tetrachloride , carbon bisulphide and glacial acetic acid , and , doubtfully , in chloroform and benzine .
The substance is insoluble in the paraffins ( petroleum ether ) and in water .
Some of the solvents which give gels are associating liquids , that is to say the liquid molecule is a complex of simple molecules , * others are not .
Gelation occurs at least as readily in carbon tetrachloride , which does not associate , as it does in alcohol , w'hich is an associating liquid .
An important negative conclusion follows , namely , that gelation is not founded upon an intrinsic tendency of the solvent molecules to form complexes amongst themselves .
Nor , if the view of the mechanism of solution put forward by Walden and Abeggf be accepted can gelation be traced to the association of solute molecules with molecules of solvent , for , according to this view , such associations occur only in solvents which themselves associate and do not occur in non-associating solvents .
From whatever point of view it be regarded the slight influence exerted by the molecular state of the solvent is remarkable * Ramsay and Shields , ' Phil. Trans. , ' 1893 , p. 647 .
t ' Zeits .
f. An.org .
Chem. , ' 1904 , vol. 29 , p. 330 .
Mr. W. B. Hardy .
[ Apr. 11 , and unexpected .
The process of gelation and the curious structural features of the gels are the same in solvents of both types .
All the gels examined liquefy on standing with deposition of crystals .
The change occurs at room temperature ( 15 ' ) in a few minutes from an ether gel , a few hours from gels of absolute alcohol or aldehyde , and in some days from a carbon tetrachloride gel .
The presence of water delays the liquefaction of the gel , that is it stabilises the gel state so that gels of aldehyde water and alcohol water may persist for days .
Sooner or later at about 15 ' the gel liquefies ; it is , therefore , at this temperature labile to the system saturated solution-crystals .
At temperatures below say +5 ' the gels change very slowly .
At higher temperatures the gel melts and will set again on cooling .
If it remains melted crystallisation rapidly occurs .
Above a certain temperature gelation will not occur at all , but crystallisation occurs direct .
The limiting temperature rises as concentration of solute increases .
For instance , , alcohol 97'5 per cent , ( by specific gravity ) as solvent , a strong solution supposed to be saturated at the boiling point set to a gel at 32 ' , but deposited crystals only at 36 ' .
On again testing after further boiling with crystals it now formed a gel at 35*17 ' which completely liquefied with deposition of crystals-in three minutes , the temperature rising to 35*55 ' in this process .
At this temperature the borderland between gel and crystals is as narrow as it is with ether at temperatures about the freezing point .
The temperature range of solutions in alcohol may , therefore , be divided into an upper region above about 35 ' in which gelation does not take place , a middle region in which the gel is .
labile to crystals , and a lower region below about +5 ' in which the gel persists for so long a time as to be sensibly stable .
The azomethine is , so far as I know , unique in the wide range of solvents which it will convert into gels .
It is true that Graham described gels of silica and alcohol or silica and sulphuric acid , etc. , but these were obtained by displacing the water of an already formed hydrogel by diffusion\#151 ; a process probably akin to the replacement of one fluid in the interstices of a sponge of solid by another .
Euhemann 's substance seems capable of forming gels directly with any solvent .
The Structure of the Gel.\#151 ; Whatever the molecular process underlying gelation may be , it in many most remarkable features resembles crystallisation .
To this I now turn .
When a strong solution is cooled , it becomes supersaturated with respect to crystals and to gel .
One characteristic of over-cooling is that , if the fall of temperature below the point of change be sufficiently great , the appearance of the new phase is prevented.* ' In this sense a solution may be over-cooled with respect to gelation .
When a * Cf H. A. Wilson , \#163 ; Phil. Mag. ' 1912 .
] On the Formation of a Heat-Reversible Gel .
solution in 97-5-per-cent , alcohol was rapidly cooled in ice , it remained fluid , but set to a gel when allowed to warm to room temperature , or at once on rapid shaking .
I have observed the same feature of over-cooling in solutions of gelatine in water rapidly brought to the freezing point ; and Garrett , working under Quincke , found that the changes in a cooled solution of gelatine in water were hastened by sowing with already formed gel\#151 ; an observation which clearly points to a similar state of affairs * An overcooled and therefore supersaturated solution contains a quantity of heat Hi which must be dissipated before gelation can occur , and a further quantity , H2 , which is dissipated when the gel liquefies with separation of crystals , and , from what has been already said , the ratio Hi/ H2 must be regarded as varying with temperature , and becoming equal to zero at some temperature about 359 for solutions in strong alcohol .
Since the gels of this type , that is heat-reversible gels , are liquefied by a rise of temperature , gelation may be .taken to be an exothermic change .
Attempts were made to measure the heat given off or absorbed , but the experimental difficulties proved to be considerable , and the results were not satisfactory .
They will be renewed with a new form of calorimeter .
Gelation of the opposite type , namely , that which is not reversed by warming but produces a gel which shrinks on warming , occurs with absorption of heat .
For instance , when allowance is made for the heat of mixing of the solvent with chloroform , it is easily possible to prove experimentally that the gelation of a solution of celloidin in absolute alcohol and ether , which is caused by chloroform , is accompanied by a small absorption of heat .
Nuclei , as is well known , favour crystallisation , and nuclei seem necessary for gelation .
The gel starts always at distinct points on the wall of the vessel which holds the solution .
I failed to discover the nature of the nuclei , but their influence is unmistakable and profound .
If a test-tube of glass or quartz containing a solution about to gel be tilted , the gel may be watched rapidly forming at a number of points on the wall .
Each mass grows rapidly until it meets neighbouring masses , on which it seems to press .
The result is that a fully formed gel , though it appear transparent and homogeneous , has in reality a remarkable structure determined by the number and nature of the nuclei .
It is composed of masses more or less imperfectly joined together and of very various sizes in alcohol gels , and very uniform in size in carbon tetrachloride gels .
The formation of these masses may be readily followed , and they are easily obtained isolated , either by pouring away the fluid from a partially formed gel or by disintegrating a gel with careful shaking .
* Dissertation , Heidelberg , 1903 .
VOL. LXXXVII.\#151 ; A. D Mr. W. B. Hardy .
[ Apr. 11 , For example:\#151 ; 130 c.c. of a 0'7-per-cent .
solution was made in 97'5-percent .
alcohol , and allowed to cool slowly by standing on the laboratory bench .
It set to a continuous , firm , transparent gel the colour of amber .
At the end of an hour about 0'5 c.c. of fluid separated and floated on top of the gel .
In the fluid were fine crystals .
Seen by transmitted light the gel was seen to be composed of facetted masses , the base of each where it abutted on the glass being a regular hexagon .
A sketch , the actual size , is given in fig. 1 , and the arrangement is shown diagrammatically in fig. 2 , \gt ; Fig. 1 .
which is an optical and horizontal section of part of the gel .
Sections cut with the free hand of a gel show , under the microscope , polygonal areas , and in the centre of some of these a confused mass appears , from which darker lines radiate regularly to the periphery ( fig. 3 ) .
This figure was drawn from a gel entirely free from true crystals .
These masses have all the appearance of crystals .
They are not crystals , however .
They are singly refractive , and when they form in free fluid the surface is rounded and not at all facetted .
In fact , each nucleus is the centre of a sphere of gelation which continues as a sphere until it meets neighbouring masses , when mutual pressure produces a polygon .
The true crystals of the azomethine are doubly refractive .
The appearance of radiating lines reproduced in fig. 3 puzzled me greatly until I came to study gels of carbon tetrachloride and carbon bisulphide .
In these gels the gel masses are easily studied , for they are always very distinct , giving to the gel a curious cloudy opaque appearance .
The following description applies to gels of tetrachloride .
The gel is seen by a hand lens of good power to be composed of spheres , very uniformly 0T4 mm. in diameter .
These spheres hang together to form a framework , and there is fluid between them .
I described gels 1912 .
] On the Formation of a Heat-Reversible Gel .
with similar structures many years ago.* Each sphere is opaque and cloudy , and the cause of this is found when a sphere is examined under a higher magnification ( Oc .
4 , Ob .
D ) .
Each sphere is then seen to be built of close packed smaller masses , which are polygonal owing to the close packing , and are very uniform in size , an axis being 10 The smaller masses are arranged in a pattern which radiates outward from the centre of the larger sphere .
They are singly refractive , and at the free surface the facetted type seems to give place to a spherical type\#151 ; to globuliten , in fact .
The gels of azomethine are thus built up of at least two orders of structure , and the air surface of a transparent and apparently homogeneous alcohol gel exhibits by reflected light a pattern of large hexagons .
The further history of a gel can be followed under the microscope .
The gel never changes directly into crystals .
The true gel is completely stable with respect to crystals , but the component large masses melt at their surfaces only , and the fluid so formed deposits crystals .
These facts concerning the gels of azomethine , together with what is known of gelation in aqueous solutions , lead to the following account of the process in heat-reversible gels .
The over-cooled fluid solution changes to a solid solution by a process which , starting about some nuclei , spreads equally through the fluid , so that spherical masses of solid are formed .
Each solid mass is isotropic , and it may be compared to a glass .
The foundation of gelation , therefore , is a formation of masses of glass , which start from independent centres .
The process has many points of resemblance to crystallisation , but the masses are not crystals unless the sphere be admitted as a crystalline form .
To the masses the old term globuliten fitly applies , and the change may be called vitrification .
The particular glass of azomethine gel is a solid solution Plut the globuliten which appear in a cooling colloidal solution may be at first fluid , and continue fluid over a determinate range of temperature , just as ordinary glass will melt to a viscous fluid .
This appears very conclusively from Garrett 's investigation of the viscosity of solutions of gelatine at various temperatures .
The temperature range divides itself into three parts : ( 1 ) from 100 ' to about 25 ' the viscosity of a 2-per-cent , solution of gelatine does not change with time ; ( 2 ) a middle region , between about 25 ' and 21 ' , in which viscosity rises with time to a maximum\#151 ; in this region , although the viscosity is high , gelation does not occur ; and ( 3 ) below 21 ' , in which viscosity continually rises with time , until it becomes infinite when the state of a solid gel is reached .
The existence of the middle region may , * ' Journ. Physiol. , ' 1899 , vol. 24 , p. 167 ; 'Roy .
Soc. Proc. , ' 1900 .
D 2 Mr. W. B. Hardy .
[ Apr. 11 , without doubt , be traced to the separation of a colloid rich phase , which is fluid , as Garrett himself points out .
It is the region in which vitrification is incomplete , a fluid glass separating .
Vitrification may involve all the solution , or a portion of fluid may persist , in which case two phases result .
Both conditions are readily obtained and studied in carbon tetrachloride gels of azomethine .
The phase which separates on cooling as spheres is not always the colloid rich phase .
Thus , when gelatine is dissolved in 40-percent .
alcohol , if the percentage by weight is less than about 25 , spheres of a solid solution rich in gelatine separate , and adhere together in anastomosing lines , so as to form a brittle framework of solid .
But when the concentration of gelatine lies between about 25 and 36 , spheres of fluid poor in gelatine , and uniformly 10 yu .
in diameter , may be seen to separate at 20 ' , and the remainder vitrifies as a continuous solid mass .
I succeeded in pushing the concentration of gelatine to 40 per cent. ; the mixture now melted at 70 ' and set at 35 ' to a milky gel with the remarkable feature that increasing the concentration of the colloid increases the volume of the spheres of fluid\#151 ; some being 25 g in diameter.* Agar water containing 1 to 2 per cent , of agar unquestionably vitrifies incompletely .
The colloid poor phase can be expressed from the gel with the utmost readiness either by centrifuging , by squeezing the gel in one 's hands , or by gravity , as e.g. if a hole is made in a mass of such gel it rapidly fills with fluid .
Why do not the masses of the colloid rich phase fuse completely ?
I suggested in an earlier paperf that the " passive resistance to change\#151 ; introduced by the formation of a solid phase " would account for it .
The influence of this resistance is undeniable ; once the solid state is reached , progress to complete equilibrium between the phases will take place mainly by slow interchange due to diffusion\#151 ; mixing by convection becomes impossible .
Difficulties in the way of dissipation of surface energy will also play their part\#151 ; electrification of the surface of the spheres would , for instance , delay or prevent the diminution of surface which follows fusion .
All these features have been already insisted upon , | but the azomethine series of gels afford convincing evidence that the distinct vitreous masses which start from the nuclei , although they may and do adhere to one another , still do not fuse in the sense of losing their individuality .
Each continues to persist as a quasi-independent system of molecules grouped about and related to its own nucleus , and each , when the mass changes , changes independently of the others at its own rate .
Whether the mass would be annealed to a truly * Hardy , loc. t ' Roy .
Soc. Proc. , ' 1900 .
I Loc .
tit .
1912 .
] On the Formation of a Heat-Reversible Gel .
continuous glass by very slow cooling is a matter for speculation , but it may be some such process of slow annealing , that is the adjustment of the individual vitreous masses of different size and nature to a common level , which accounts for the slow change in vapour tension and elasticity which reversible gels are known to exhibit .
If gels of the azomethine are typical of all reversible gels , then these are the glass state of solutions supersaturated with respect to crystals .
Why therefore , do not gels of gelatine or agar slowly liquefy and deposit crystals ?
It is not because of the large size of the chemical molecules , for many proteins can be obtained in the crystalline form .
The answer lies , I think , in the fact that , in the case of solutions of azomethine , between the region of concentration and temperature in which the gel is permanent , and the region of simple saturated solutions , there is interposed a region in which the glass melts to a true fluid which is supersaturated with respect to crystals .
In the case of solutions of gelatine or agar , this intervening region does not exist ; instead , we have the glass softening to a viscous fluid , which passes insensibly to colloidal solution .
That solutions of azomethine are not singular in the fact that gelation starts from centres , and only slowly involves the whole mass , is proved by an observation made by me many years ago .
A solution of silicic acid was dialysed for some days against running tap-water until it formed a firm gel in the dialysing tubes .
The gel was then removed from the tubes , broken up , and centrifuged for three hours , with the result that a small quantity of fluid separated at the top of the gel .
This was pipetted off and , on adding a trace of magnesium sulphate , it at once set to a firm gel .
Vitrification proceeding from nuclei in the solution of silicic acid had therefore not yet had time to involve the whole mass of the solution .
A stage in which a melted viscous " glass " separated was directly observed by me in a solution of 55 parts by weight of gelatine in equal parts by volume of absolute alcohol and water .
When sufficiently heated , this forms a fluid which appears homogeneous under a magnification of 450 diameters .
On cooling , it may be seen to separate into two fluids , one of which , as the temperature continues to fall , becomes solid .
[ \#166 ; Postscript added May 20.\#151 ; The gels described in the preceding pages behave like ordinary watery gels in the fact that dyes such as eosine and methylene blue diffuse into them from alcoholic solution and " stain " the substance of the gel .
]

