The polymorphism of glycerides

1 THE POLYMORPHISM OF GLYCERIDES T. Malkin Dt'aIyc. the past century, numerous workers have studied the interesting problem of the "double melting" of g]yccrides, and whilst it has long been known that this was due to po!~norphism, the exact nature of this, and the relationship between the various polymorphic forms, was little understood. Indeed, as we now know, a clear explanation was hardly possible before the advent of x-ray diffraction methods. Once, however, that this new weapon had revealed the structure and also the general habit of polymorphism of simple long chain compounds (see Vol. l, Ch. 1), the way was open for a new attack on this longstanding problem, and from 1934 on,yards there appeared a series of thermal and x-ray studies of glycerides, by 3[ALKIN a~(1 his eollabor,~_tors, which developed the present views on glyceride polymorphism. More recently, DAUBERT, Lt'TTON, and other U.S. workers have joined MALKINin this field.

The double melting of lriglycerides As far back as 1849, HELXTZ¢1) observed that stearin, after melting and quickly solidifying, melted at 51-52°C, but on being heated further, resoli(lified and meltcdagain at 62-62.5°C. A little later, Dt-FFY le~ gave, for his time, a remarkably clear account of the melting behaviour of tristearin, for which he reported three melting points, namely 52 °, 64.2 °, and 69.7 °. He showed that the highest melting point was identical with that of the solvent crystallized glyceride, and that the lowest practically coincided with the setting point. He also noticed that the more he purified llis tristearin, the more difficult it became to detect the intermediate melting point. This excellent work appears to have been overlooked by later investigators, and for many years only two melting points were rcp0rted for tristcarin (BERTIIELOT,13) SCHE1J: ~41 GUTH, (5) LAUTZ~)), for tripalmitin (3LxSKELIYE(7~),and for trimyristin (1),m.SIERand WILL(8~),thus giving rise to the well-knowr expression "double melting" of glycerides. In 1915, however, OTm~ll.:R(9) reported three melting points for tristearin and trimyristin, and a little later, LOSKIT(m} confirmed this, and also found three for tripahnitin. On the other hand, as recently as 1930, JOGLEICaRand ~VATSO-~,(11) in a careful study, reported only two melting points fi)r a number of triglycerides of the .highest degree of purity. Two years later, howe~;er, WEYGAND and GRUNTZlG~l°'~ claimed the existence of seven different fl,rms, for a number of triglycerides, as a result of a detailed microscopic examination. It should 1)e observed, however, that visual observation is a n(doriously unreliable method fiw the identification of different crystal forms, and hence these lattcr rcslflts must bc regarded with some reserve. 1 2

The Polymorphism of Glyeerides The position therefore at this time, just prior to the first x-ray studies, was still very confused, and it is of interest to note that the views of D v s r r , eighty years earlier, most nearly summarized the position, namely, simple triglycerides exist in three modifications of different melting points, the modification of intermediate melting point often being difficult to detect.

Thermal and x-ray investigation Briefly, the principle of this method is to establish the number and range of stability of the various polymort,hs, by means of cooling and heating curves, and, on the basis of this, to determine differences in structure between the v..rious forms, by x-ray examination. MAL~I~ c13) had earlier used this method in a study of the dimorphism of ethyl esters of fatty acids, and there was good reason to expect equal success when it was applied to glycerides, since these are merely slightly more complex esters. As expected, a close parallel was found. Ethyl esters had been shown to exist in two modifications, a stable form (ill in which the long chains are inclined with respect to the planes formed by the terminal methyl groups, and a less stable form (a), in which the chains are ve,'tical and rotating. Both these forms were found to exist in the triglycerides, (14) namely, a stable fl-form with inclined chains, and a monotropic :~-form, with vertical chains. In harmony with views on alternation advanced earlier by :M_~LKI.~,~13~the melting points of the inclined /3 forms were found to alternate, whilst those of the vertical a-forms were nonalternating. The third and lowest melting form of the triglycerides, which is obtained by rapid cooling of the molten glyceride, was found not to be truly crystalline; it possesses many characteristics of a glass, and was therefore termed a vitreous form (7). On extending the investigation to mixed saturated glycerides (containing two different acid radicals), ~IALKIN and M~:.~mA(I~L~ls~ and CARTER and MxI~ KL~Ca,, Cas~ found a fourth modification, with a melting point lying between those of the ~- and fl-forms. This they termed the fl~-form, since it possessed non-rotating, tilted chains (see Vol. l, p. l l , for distinction between a- and fl-forms). On re-examination of the simple glycerides in the light of this diseovery, the fl'-form was found for those containing odd membered acids (see GRUh'TZIO,(19} tO whom these results were communicated). Unfortunately the work had to be disc~mtinued at this stage, but some years later this form was also reported for simple even acid glycerides b y a number of U.S. workers.l~L (21), (22) Summarizing the position for saturated triglyeerides in 1948, CLm~zso.'," and MALKI.'~~2a~ stated that all the saturated glyeerides that they had examined (some fifty), whether simple or mixed, exist in four solid forms, viz. in order of increasing meh.ing point, vitreous, :¢- and fl'. (crystalline, monotropic), and {/- (stable crystalline) form. Except for those containing short clmins, i.e. Ca0 and C1~, rapid cooling of the melt brings about solidification mainly in the vitreous form, which on warnfing passes through the following transitions vitreous --~

Cooling and Heating Curves a -~ fl' -~ fi, and in some eases, vitreous ~ a -+ ft. These transitions are rapid for simple even acid, and s~nnmetrical mixed triglycerides, but slow for simple odd acid, and uns~nnmetrical mixed triglycerides, the speed of transitions diminishing in all cases with increasing length of acid chain. It was shown that a-forms of simple even acid triglycerides may "cha.nge directly into the stable fl-form without passing through the intermediate fl'-form, but the factors causing this somewhat exceptional transition could not be determined.

Cooling and heating curves These play an important part in the investigation of glyceride polymorphism, and a clear understanding of the subject cannot be arrived at without a detailed study of such curves, taken under a variety of temperature gradients. The technique should preferably not be too elaborate, for it is desirable to take a number of curves for each glyceride, and the following simple method can be recommended : 0.75-1 g of the specimen is placed in a thin-walled test-tube of about ~-z3 i in. diameter, and fixed by a cork in a tube of about three times this diameter to act as an air jacket. This is placed in a Dewar vessel, which may contain air, water, ice, or freezing mixtures, to give any desired cooling (or heating) gradient. A copper-constantan thermocouple, fixed centrally in the specimen, is attached to a Cambridge thread recording instrument, which marks a chart on a revolving drum every half-minute. The chart can be read to ¼°C, which is sufficient for most purposes, since all melting points can be checked by the standaxd capillary method. Cooling curves Although cooling curves differ in appearance considerably, according to the rate of cooling, there are, neve~hele~, two distinct types, which correspond t~) the separation of a- and vitreous forms, respectively (Fig. 1). Thesa are readily distinguished by the supercooling which almost invariably precedes a-crystallization. Supercooling cannot, of course, precede the separation of the vitreous form, for the temperature at which this occurs is essentially that below which the glyceride cannot remain liquid. Cooling curves normally give no indication of the separation of fl'- and fi-forms, and indeed, molten glycerides can be maintained for hours in the liquid state, at temi~ratures several degrees below the melting points of these forms. If, however, the melt is heavily seeded with fl-form, or ii" it has not greatly exceeded in temperature the fl-m.p. {i.e. there are still a number of minute fl-seeds present), an arrest will appear at the ~6-m.p., but crystallization is excessively slow: :It appears, therefore, that whether fl-seeds are present or not, the most facile route from the melt to the stable fl-form is ~qa the a-form, viz. :

# liquid -+ a < Crystallization in the a-form is, however, by no means rapid. For example,

~ne Polymorphism of Olycerides JOGLEKAR and WATSON(11} showed that 1 g specimens of tristearin took from two to three hours to solidify at the ~-solidifying point. It is clear, therefore, that the unwieldy glyeeride molecules do not readily assume the regular layers required by the ~-structurc. Consequently, if a molten glyceride is cooled rapidly, solidification takes place, because of loss of molecular kinetic energy, before the required regularity of the :c-form is achieved, and the form which has been termed vitreous separates. Heat is evolved during this separation, for in an irregular manner, hydrocarbon chains and parts of chains are held together much as they are in a true crystal, and the "heat of crystallization" of the contacting methylene groups in adjacent chains is evolved. The main difference, therefore, between the ers, stalT

I00 qO 80

70 o~ 60 50

A

~

30" 20" tO •

0 TIME

Fig. l.

Cooling curves for Tripalmitin.

line and the vitreous condition is that in the former, the regular arrangement permits the maximum number of methylene contacts, whereas in the latter, these are fewer, and may vary according to the rate of cooling. The heat evolved during vitreous solidification is consequently less than that evolved during :c-crystallization; moreover, because of the Variable methylene contacts, the heat content of the vitreous solid varies, and hence the setting point (and also tlze m.p.) is not constant, b u t varies over a small range. It will be clear from the above that the nature of the solid which separates, and the form of the cooling curve, will greatly depend on the rate of cooling; by varying this, solids may be obtained consisting of groups of moleculcs in all possiblearrangements from the disorder of the liquid to the regular layers of the a-form, and even fl'- and fl-fo-ms. A brief consideration of a few typical cooling curves will illustrate this. Fig. I.x Oripahnitin, separation of :c-form) is a typical curve: which can be obtained for most glycerides, if the cooling gradient is small. Considerable supercooling occurs, and the temperature may closely approach the setting point of the vitreous form before ~-scpara~ion commences. At this point, the t(~npcrature rises and remains steady at the ~.-sctting point for a short time, 4

Cooling G~arvea after which there m a y be a further rise owing to partial transitioa into the fl'or fl-forms, but the m.p. of the latter is not usually reached. Fig. 1B is also a curve for tripalmitin, but with a greater cooling gradient. There is no supercooling and separation of vitreous form occurs, with relatively small heat evolution. 80 o

7C~ bO:

5C~

/

Cj

40 °

30 o 1-

20 o ,

IO © -

/(

-

B'

O'

B"

TIMI:

Fig. 2. Cooling and heating curves for Trilaurin.

Fig. 2 is a curve fi)r trilaurin, showing :t-separation: in this case, crystallization sets in at a temperature well above the vitreous setting point (15°C), but, nevertheless, only after considerable supercooling. The cooling gradient is teo great to permit a steady temperature during ~-separation, and therefore neither the m a x i m u m nor the minimum on this curve represents a definite setting point. -70 ° bO ~ 50 ° ~O c

a~

. ..... • ~ o

":..:

"N

~

I--

tC7 2- MYRISTODIPALMITIN 2- LAUROOIMYRISTIN OO

2 - PALMITODIMyRISTIN I , I

I

IO

20

30

4O"

5O

0

I0

20

30

40

TIME MINUTES.

Fig. 3. Cooling and healing eurvca.

In a case of this kind, tile ~-m.p. can be determined by a heating curve (BC) or by adjusting the cooling gradient t.o obtain a steady arrest followed by a rise as in Fig. 1A. Fig. 3 a, b, are curve.~ for 2-myristodil)ahnitin; tbc former, cooling to room temperature, shows separation of vitrc.us fi)rm, whilst tile latter, cooled more slowly, shows first at-separation, f()lh)~ved by vitre,us seI)aration, and tinally a i,apid transition into fl'- and fl-forms. Fig. 3 c, c' (2-laurodimyristin) shows vitreous separation followed by transitions into ~- and fl'-fi)rms.

The Polymorphism of Glycerides The controlling factor in the separation of a crystalline solid from its melt is the rate of formation of crystal nuclei, and all the evidence from cooling curves of glycerides, shows that it is the rate of formation of nuclei of the a.form which determines the nature of the final solid. Formation of nuclei of t6'- and/if-forms,

°°°: J ~

~G

-;,

a0 B

TEHPERATURE

(a) (b) Fig. 4, (a) Free energies, (b) activation energies of nucleation of o:-, ~'- and ~forms of triglycerides (diagrammatic).

directly from the melt, is too slow to have any appreciable influence on the course of the solidification. Below the melting point of the stable fl-form, the liquid and the a- and fl'-forms are metastable, i.e. their free energies (G) are higher than that of the

¢¢'N,P.

VITREOUS H.P

RATE OF NUCLEATION

Fig. 5. ]Relation between temperature anti rate of nucleation of ~- forms o f triglyeerides {diagrammatic).

stable form (Fig. 4a). Nevertheless, crystallization does not commence spontaneously, for before this can begin, an energy barrier must be crossed, the activation energy of which, according to VOLMER,(24 ) i8 one-third of the surface energy of the newly formed nucleus. Fig. 4b illustrates this diagrammatically •~nd indicates tile greater difficulty of formation of fl'- and/q-nuclei.

Heating Curves I t is seen from Fig. 4a that the greater the supercooling, the greater the free energy difference between the liquid and the ~-form, and hence the greater the tendency to crystallize. On the other hand, increased supercooling results in increased viscosity, and hence the molecules cannot so readily move into positions favourable for nuclei formation. Consequently, somewhere between the ~- and the vitreous m.p. there will be a zone of optimum rate of nuclei formation, Fig. 5. The position of this zone varies a little according to the nature of the glyceride, but in general it lies nearer to the vitreous than to the co-setting point. This is illustrated by the sharp rise in temperature when specimens are maintained for any appreciable time in the neighbourhood of the vitreous setting point, cf. Fig. 3 a, b, c. Heating curves

The cype of heating curve obtained depends upon the previous cooling history of the specimen, and upon the heating gradient. I f the molten glyceride is cooled rapidly, it passes through the zone of a-nucleation too quickly to form a-nuclei of the necessary critical size for further growth, although numerous embryonic nuclei (i.e. less than the above critical size) may be present. I f a steep heating gradient is now applied to this solid, it melts at the vitreous m.p. and the embryonic nuclei break up and dissolve in ~he melt, as the temperature rises beyond the zone of rapid nuclei formation. The curve obtained shows a single slight arrest at the vitreous m.p. With a smaller heating gradient, however, the embryonic nuclei may remain in the nucleation zone for a sufficient length of time to grow and initiate ~crystallization. The heating curve in this case would show a slightly larger arrest at the vitreous m.p., followed by a steep rise to a second arrest at the ~-m.p. In a similar way, too rapid heating may cause the dissolution of embryonic nuclei of fl'- and fl-forms, and, in general, the smallest practicable heating gradient should be employed. As ~dready indicated, if t,he original melt is submitted to a slower rate of cooling, the resulting solid may consist of vitreous and a-forms, ~-form, or aand fl'-forms, and arrests will usually be found on the heating curves corresponding to the forms present. It should, however, be remembered that all the above forms are metastable, and they may change into higher melting forms when heated. These transitions take place the more rapidly the smaller the acyl chains, and the greater the s3"mmetry of the molecule: they may indeed take place in the solid state at room temperature, as in the case with glycerides of lauric and lower acids. This is illustrated by the curves given in Fig. 2, where BC, B'C', and B"C" are heating curves taken at ½, 1, and 3 hr respectively, from t h e commencement of the cooling curve. The ~-form, which first separates, stowly changes at room temperature into the stable fl-form, os shown by tbe disappearance of the a-arrest from the heating curve. Typical heating o,,rves are sho~m in Figs. 2~ and 6.

The Polymorphism of Glycerides Sufficient has now been said to indicate the value of the study of cooling and heating curves in the determination of the number and the stability of polymorphic forms. With the majority of triglycerides they readily reveal the existence of four solid forms, and it'is only in the case of simple triglycerides of even acids where they fail to do so. Here the fl' --> fl transition takes place rapidly in the solid state, during the course of the heating curve, and the fl'-form does not actually melt; hence, there is no indication of this form on the curve. Before leaving this section, it should again be strongly emphasized that without a detailed study of cooling and heating curves it is impossible to

,

"*,°,.

..../

~4d •

$

,F-: J /"

"v"

!

....I

x~ i O

IO

, -, ..............

@OO@eO@bO

.-"i....

;':

2O 3O TIME MINUTES 2- LAUROCdPALMfflN

40

2-MYRISTODISTEARIN 2 - P,,~.J41TC)DILAL~!N

Fig. 6. Cooling a n d h e a t i n g curves.

obtain a clear picture of glyceride polymcrphism, and the failure of certain workers(25), (26), (27) to observe the forms found by MALKIN is due mair.ly to the neglect of these studies.

Capillary meltimjpoints Early studies of glyeeride polymorFhism were carried out mainly by capillary m.p. determinations, and these affor,1 useful confirmatory information to that obtained from curves. In some eases, tile capillary method is the only means of determining the m.p., e.g. the fl'-forms of simple even acid triglycerides. There is no difficulty in the determination of the m.p. of the stable fl-form, which is usually obtained by slow crystallization from non-i)olar solvents. The only precaution necessary is to raise the temperature very slowly in the region of the m.p., otherwise, owing to the slow melting of glycerides, it is easily possible to exceed the true m.p. by as nmch as a degree. The determination of the m.p. of other forms requires a m()dified technique 1)eeausc of tlm possibility of transitions in the s()li(1 state, as the temperature (,f the specimen is being raised. Thus, the m.p. of the vitreous form is determined by immersing the capillary, containing the rapidly cooled glyceride, in a bath at a temporature at which melting and resolidification just takes place (vitreous -.~ ~-for,n). The capillary is now placed in a bath at a temperature at which the

X-ray Investigation

a-form just melts and resolidffies (a -->//'-form) and similarly for the m.p. of the fl'-form. This, of course, involves a large number of repetitions, as the various bath temperatures must be found by trial, but the work is greatly reduced if the m.p.s of the different forms are k n o ~ from heating curve studies. The m.p.s are not quite so clearly defined when any particular form contains an appreciable number of nuclei of the higher melting forms. If, for example, a molten glyceride is cooled a little too slowly, the resulting vitreous form may contain numerous a-nuclei; on being immersed in a bath at the vitreous m.p., these nuclei grow rapidly as the vitreous form melts, and the mass appears to soften rather than melt, as it changes into the a-form. The same apphes, of course, to a- and fl'-m.p.s, which have frequently been reported as softening points. This is a sure indication t h a t the particular specimens contained nuclei of higher melting forms. These difficulties are partly due to the fact that solid glycerides are extremely bad conductors of heat, so t h a t even when a capillary is immersed in a cold bath, the inner core of the specimen is protected by the solid outer shell, and may cool relatively slowly. Because of this, it is advisable to use thin walled capillaries of small diameler. As mentioned earlier, the capillary method affords the only means of determining/3'-m.p.s of simple triglycerides of even acids, and even so, it is not possible to determine those of trilaurin and lower homologues, because of the speed of the transitions. The fl'-m.p.s of higher members are determined as follows: The specimen is melted in a capillary, solidified in ice, and then placed in a bath one or two degrees above the m.p. of the vitreous form. Complete or nearly complete melting takes place, folluwed closely by rcsolidification in the :t-form, but after some three to six minutes longer at this temperature, a further elmnge takes place, denoted by a shrinking from the walls of the capillary. Before this change, the specimen melts ahnost c~m~pletely at the a-m.p., but after the shrinking, the ~.-m.p. is passed without any change in appearance until the fl',m.p, is reached. At this point., softening and very rapid resolidification in the stable ~-form takes place.

X-ray investigation The x-ray technique has been described in some detail in Vol. 1, p. 8, and need not be filrther elaborated. Thin layers, either pressed or melted, and rods are examined in order to determine long and short spacings respectively. In general, the determination of the spacings of the stable fl-forms offers no difficulty, for this form can usually be obtained suitably crystalline by slow crystallization from solvents (preferably non-polar). Layers and rods of other forms are prepared by coolingthin melted layers or capillaries of the glyccridcs, followed where necessary by beating, under conditions determined by previous cooling and heating curve and capillary m.p. studies. From what has been ~aid earlier, it will be apl)reciated that it is not always a simple matter to obtain the metastable forms ent irely free from higher melting forms, and in order to obt~lin homogeneous specimens, repetition is often necessary. It. must also be remembered t h a t metastable forms of glycerides containing the shorter acids eaprie 9¸

"

The Polymorphism of Olyeeride8

and laurlc may pass into stable forms at room teml~rature too quickly to permit the determination of x-ray data. In such cases, special arrangements for cooling the specimens during exposure must be made. In order to obtain the vitreous form for x-ray work, a small amount of molten glyccride is spread on a microscope cover slip, in as thin a layer as possible, by means of a warm metal spatula. Tile layer is then cooled rapidly by holding the underside of the cover slip in ice-salt-water or solid CO2-acetone, and the cover slip is then mounted on the spectrograph. Alternatively, the molten glyceride may be cooled in a fine capillary and ejected as a rod, although in this case, traces of a-form may be present owing to the relatively slower cooling in the centre of the rod. These specimens may be converted into the :~-form by holding ~hem at the ~mperature of the vitreous m.p., or the spceimeas may be melted again and cooled less rapidly. Similarly, the ~-forms may be converted near their m.p.s into the fl'-forms. In certain cases, it is possible to obtain fl'-forms by rapid crystallization from solvents, particularly ethanol or acetone. With an x-ray tube running at about 15 mA, exposures of 2ff-30 minutes each side for layers, and 10-15 minutes for rods, are usually adequatel but, sometimes melted layers, being not so well oriented, may require a little longer.

Long spaci~gs The long spacings of the various forms of simple triglycerides increase linearly with the carbon content of the constituent acids, and when plot~d, fall on straight lines which cut the ordinate at C=O, in the ncighbourhood of 4.5 A, see Fig. 7.

A~DS)

50--

~40

.... '

f

~" 3 0

o

~2o <

O

, I

2

3

4 S b 7 B 9 IO II 12 13 14 IS Ib 17 18 NU,"48[R OF CARBON ATOMS IN ACID RADICALS

Fig. 7. Long spacings of aimplo triglycerides.

The spacings are too large to be related to the length of a single acid chain, and this led CLARKSONand ~IALIKINt14} tO propose the tuning fork structure, where tile spacing is related to twice the length of the acid chain, viz. ] . Alternat;ve molecules are regarded as lying head to tail across ti~e reflecting 10

Long Spacings planes, eithcr verticaUy (all a-forms and a few ~'-forms) or tilted at a variety of angles ~ ' - and fl-forms).

/I/1i1'1'1/I ! /////////2 Since in Fig. 7 the spacings ale plotted against the number of in the acid chains, the intercept at C-- 0 (4.5 _~.)is the contribution molecule and the terminal hydrogen atoms to the long spacing. i tilt 0 is given by i-~ ---- sin 0, where i is the increment of spacing

carbon atoms of the glycerol The angle of per C atom in

(1-3 -~ is the maximum increment for a zig-zag chain of methylene groups, assuming the carbon atoms to be joined at an angle of 116 °, see Vol. 1, Ch. 1).* The above structUre holds for all saturated mixed glycerides ff the acids do not differ in length b y more than two carbon atoms, but where they differ by four or more, there is usually a l~rge increase in the long spacings. This led .~ISTTO~(za) to propose the follo~Sng triple chain length (T.C.L.) structures:

, ]

j

t

AND

I This latter group is of particular interest, for the s-forms and certain of the fl'- and fl-forms still appear to retain the previously described simpler structure. It may well be in these cases that thcre is some interpenetration of the chains, viz. : I,,

-

I

l:

Seine light might be thro~n on this problem by plotting the spacing against the et!eetive length of the gtyeeride unit, and eonsiderii,g the intercept at C = O. For the simple structure of CLA!a~SO~ and MAL~r~-: the inbereept is 4.5; any smaller intercept would indicate interpenetration of the chains, whilst the T.C.L. structure would give rise to an intercept of about 9 -~. ~rhen plotting£he long spacings of mixed triglycerides against carbon content, regularities can be expected only if structurally similar families are selected. Thus, the family of glycerides of acids Cr,C10CI~., C14C12C14, ClsC14C16, ('lsClsCls i i

Ci2 CIO CI~

Cl8 Ci6 Ct 8

have the same irregul',rity in structure at. the terminal planes, i.e. the central chain is two carbon atonis shorter than the two outer, and the incre,~me in total * All a n g l e s o f t d t g i v e n in t h i s m e m o i r a r e c a l c u l a t e d o n t h i s a s s u m p t i o n .

ll ¸

The Polymorphism of Glycerides length from member to member is four carbon atoms. The long spacings of such a series would be expected to fall on a straight line, and hence the molecules would have the same angle of tilt across the reflecting planes. Similarly for the series C10C12C10,CloCa4Ca2, C14(',16C1~: ClnCasCjn. On the other hand, a series such as C1sCl0Cls, C1sCaoCls,ClsC14Cls, ClsC16Cls, ClsClsCls is so markedly different from member to member that regularity in structure would not be. expected, e.g.

~L

(a)

1IV ....

(b)

CI8 Clo CIB

C18 ClB CI8

greater interpenetration of chains is possible in a than b, and this might alter both the tilt and the actual length of the diffracting unit; as already mentioned, the irregularity in this group is ~c great that the first three members adopt an entirely different structure (T.C.L.). The effect of tilt on thc long spacings appears to have been overlooked by some workers: thus, LUTTON(~8) has proposed three straight line equations for the calculation of the a-, fl'-, and fl-forms of saturated triglycerides in general, on the assumption that each form has the same tilt as in simple triglycerides. This is definitely not so, as will be seen from Table l, and it is therefore clearly imp~ssible to calculate the spacings on the obcvc simple basis. Again, BAILEY(29) finds it difficult to understand what configuration could gi~-e a greater long spacing for unsymmetrical dipalmitomyristin, than for tripalmitim This is also clear from the tilts, for although the two molecules have approximately the same length, the angles of tilt are respectively 70 ° and 61 °, and hence the spacing D, which is the distance between the reflecting planes, is greater for dipalmitomyristin.

,TF

l

7-f--/-- ~ -

, , "/j

LONG SPACING

C16 C16 (::16

CI6 CI6 CI4 Table 1 *

I

I

I

Simple t riglycerides Series C,.C1oCx2; CI,CI'2CI|; etc. Seri()s ClD(?12Clo;

ClyCllC]2 ; e t c .

Series Cl0C1zCl~; Cl.CI,CI4; etc. Series Cl:Cl0Clo; C1,CI.CI~; elm.

12

i .

i

Angle of tilt, lV

Angle of tilt, p

67 90 79 69 65

61 66 66 79 66

"

Shor~ Spacings

Short spacings Both the xitreous and the ~-form give a single strong line at -~ 4.19 .~, but the :c-form is distinguished by the presence of sharp lines due to the long spacing. Owing to the difficulty indicated earlier, of avoiding the presence o f small amounts of the :c-form in the vitreous form, faint long spacings may appear on the photogTaph of a specimen which is predominantly vitreous, and this has led some workers to question the existence of tile vitreous form. *~°), (21t It should, however, be remembered that there is no sharp dividing line between the crystalline and non-crystaUine state, and one visualizes a range of states from crystal -> crystallite -~ embryonic crystallite --> unit cell --> disorder. With diminishing size of unit, x-ray diffraction becomes more and more diffuse and as thc layer structure falls below a certain size, long spacings disappear. A diffused short spacing of -~- 4-2 still persists, owing to lateral diffraction by hydrocarbon chains, which does not nece~arily depend on a layer structure A vitreous form, containing embryonic nuclei, might well give rise to weak long spacings, yet nevertheless mclt at thc vitreous m.p. The interpretation of x-ray photographs, in these cases, is therefore a matter of experience and judgement, the extreme cases being ,~ single diffuse spacing of 4.2, accompanied l)y fogging due to general x-ray scattering, for a vitreous form, and a much clearer photograph showing sharp long and short spacings for an ~-form. The :c-form of glycerides has been so termed by analogy with co-forms of other long chain compounds which lie vertically across the reflecting planes (hydrocarbons, alcohols, ethyl esters) and which give rise to a single short spacing of 4-2 _~. MULLER suggested that the chains in such compounds were rotating about their long axis. ~-forms of glycerides are therefore vertical rotating forms, and since such forms are axially symmetrical, the existence ofthc so-called "tilted :c-form" of LUTTO.N,(a°~ is e.xircmely improbable. The fl'- and fl-forms give rise to groups of short spac;ngs which can usually, but not invariably, be sharply differentiated. ]n general, the/~-forms give lines somewhat widely separated, relented to as a "wide band," the strongest line of which is 4-6 .~. A ts~2ical wide band is that of the simple glyccridcs 3-7, 3-9, 4-6, 5"3 ~. The lines given by the/]'-forms are usually closer together, and they have been termed a "narrow band," e.g. the /?'-form of 2-decodilaurin gives 4-35, 4-16, 4-05, 3.82 .~, the strongest of which is 4.16/~. Examples are given in Fig. 8. The existence of the above three types of short spacings led LUTTO_N~z°} tO make them the basis for nomenclature, thus a-form - - a single strong iine ~- 4-15 tix. /]'-form-- usually two (but occasionally more) strong lincs corresponding to -"-4-2 and 3.8. fi-fi)rm - - a strong (usually strongest) line corresponding to -"- 4-6 ~. This has led to considerable confusion, for ~[ALKIN{15}]tad earlicr introduced these terms to dt.note the polym,~rphs in order of inercasing m.p.s viz. ~, fl', ft. Lva~roy-'s terminology is unsatisfactory fl'om many points of view, for as has already been shown both the vitrcous and the ~.forms give the single spacing of 13

The Polymorphh~a of Glyceride~ 4"19/~; moreover, both the fl'- and /?-forms of trihydnocarpin, triehaulmoogrin and all the diglyeerides give the strong 4.6 line which LvTTo~ ascribes to fl-forms only, and finally, all the ~-forms of the series of triglycerides CnCa+~Cn, Cn+~Cn+aCn+e, etc., give'the 4-2 and 3.8 lines. It would indeed be astonishing ff the great variety of structurally different glycerides were to fall into the three clear-cut crystallographic groups sugges~d above. LVTTO~ goes so far indeed, as to associate the various spacings with defimte planes in the crystal, and ascribes (by inference from his diagrams, Fig. 9) monoclinic and triclinic structures to the ~'- and ~-forms respectively. It will hardly be necessary to point out that it is quite impossible to deduce such structures on the basis of a few lines on a powder photograph. Whilst it is possible to identify planes in crystals of higher s3nnmetry--cubie, hexagonal, trigona], rhombohedral--on the basis of a large number of powder photograph

D ~ - 4-14A

aLPHA

D~ = 4-15A

D1 = 3-68 A

BETA PRIHE

D3~3 B b a BETA

Fig. 9. LL'TTO.~'Ssuggested .structures for glycerides.

lines, it is usually excessively difficult, or even impossible, to do so with orthorhombic, menoclinic and triclinic crystals. In the case of the latter, the necessary mathematical expression is so unwieldy that it is not worth while attempting to use it. Planes ofmonoclinic crystals of certain long chain compounds can be identified from powder photographs, with a high degree of probability, where single crystal data are available, as is the case with hydr,)carbons and fatty acids. :No single crystal data are available for glycerides, however, and until these are forthcoming, any association of fl'- and/~-forms with definite crystal structures is pure speculation. FtLER, Jr., SIDHU, CHE~-, and DAUBERT,(a~) who a!so find anomalies in LUTTON'S terminology, are more circumspect, and rightly state that "judgement as to the positive association of specific and characteristic sidespacings (i.e. short spacings) for distinguishing the fl- and fl'-phases of triacidtriglycerides, must be reserved until additional information can be o b t a i n e d . . . " In the present article the terms :c-, fl'-, and fl-forms are used, as has alway~ been quite clearly stated by 3[hLKr:% to denote the crystall!ne polymorphic forms in order of ascending melting points. Before passing on to the detailed consideration of t.l:e various groups of triglyceridcs, the general picture may briefly be summarized. On slowly cooling the molten glyccrides, considerable supercooling occurs and ~-nuclei slowly form and initiate a-crystallization. This form consists of layers of molecules, lying 14

:

j +

,

..L._.

.:. .

.

_.&

i.,_..:

4 k&A&

__-.c

.’

:,.

i i

Simple Triglyceride~ Mternatively in head to tan position, vertically across the reflecting planes, with the long chains rotating freely about the long axis. The slow nucleation is due presumably to the requirement of a head to tail alignment, and to the increasing viscosity of the melt on cooling. If, at this stage, the heat of crystallization does not balance the external cooling, the chains become frozen in positions which on average, correspond to those of rotating chains, and the resulting a-form may be stable at room temperature for varying periods, according to the length of the chains. Such a solid gives long and short spacings corresponding to a hexagonal structure. If, however, the heat of crystallization balances or exceeds the external cooling, the molecules retain Sufficient mobility to move into the more stable positions of the fl'-structure. This in the majority of cases means that the chains lose their rotational energy and tilt across the reflecting planes (a few cases occur where the chains remain vertical). This a --> fl'-transition is probably a change from the hexagonal to a body-centred monoclinic form, a structure which is adopted by the majority of long chain compounds ; but, as mentioned earlier, there is no direct evidence for this in the case of glycerides. Finally, the //'-form changes into the stable fl-form at room temperature or near its melting point, and at different ra.tes, according to the varying lengths of the acyl chains and the general symmetry of the molecule. In general, the fl' -->/~-transitior. is most likely a change into a slightly more stable monoclinic structure, similar perhaps to the change of the B -+ C forms in fatty acids, see Vol. 1, Ch. 1. The picture of glyceride transitions is not so clear when the a --> fl' -+ fl transitions involve large increases in the length of the spacings, i.e. formation of T.C.L. structures, and this group is worthy of a more detailed study, but the most urgent need in this field is single crystal data. ]~[.P. AND X - R A Y I)ATA FOR SATURATED TRIGLYCERIDES

Simple triglycerides This group is the most complete homologous series of glycerides so far studied, and it exhibits the regularities in m.p. a n d x-ray data usually found in such series. Thus the m.p.s of the 0o tilted fl'- and fl-forms alternate in a regular manner, whilst those 6o of the vertical a-forms are nonf i ~ v~-ous alternating, Fig. t0. Again, the long spacings of tilte'! forms of 4o / / odd and even members, plotted / / against the carbon content of 2o • / the acyl chains, fall on different straight lines, Fig. 7, and the s h o ~ spacings of odd and even members fall into ~parate, closely similar gq'oups. Transitions of metastable

/1

-so

i ol--

I i

I 10

II

II

13

J4

J$

16

17

f|

Fig. 10. Melting points of slmplo triglyeeride~. 15

•

The Polymorphism of Glyceddes

forms, at their m.p.s, into higher melting forms, are parficuiarly rapid ~fith the even acid members of this group, and it is not easy to obtain homogeneous :t- and//'-forms. The even acid members also appear to be unique in passing direct b from the :t-form to the stable fl-form, and these factors make the determination of the fl'-m.p, difficult. Transitions in the odd acid group, particularly t h e fl'--~// transition, "are much slower, and it is rare to realize the//-m.p, arrest on a heating curve• No difficulty is experienced, therefore, in the determination of the m.p.s of the "various forms of this group. Tab~ 2. 3lellimy points of simple triglycerides (CLARKSON a n d 3IALKIN t23)

Tristearin Trimargarin Tripalmitin Tripent adecylin Trimyrist in Tritridecylin Trilaurin Triundeevlin Tridecylin .

Vitreous

=

fl"

"i"1

I

54.5 50.0

•~

45.0

.! I

40"0 33"0 25"0 15,0 1-0

65.0 61 "0 56-0 51 '5 46"5 41 "0 35.0 26-5

70-0 t;2-5 63-5 53"0 54"5 42"5 -29-0

15.0

18-0

. •I I --

i

--

72"0 64"0 65"5 54"G

57"0 44"0 46"4 30'5 31"5

if-melting points have not been determined for trilaurin and tridecylin.

Table 3. Lo~+9and short spaci~uls of simple triglvcerides (CLARK:ON aml 5IALKIN) (13|, (23) s Long

Short +__

Ct

Tristearin . Trhnargarin Tripahnitin Tripentadeeylin Trimvristin. Tritrideevlin Trilaurin . Tri,m,lecylin Trideeylin.

50-6 i 47-2 45.0 ".48"5i43"7 143"5 45-6 42.6 ' 40.6 42.9 ! 39.1 ,38.9 i 141 .o+il37-6 35.8 37.7 i 34-2 I 34.1 i 35-6 32.85'i 31.2 .! 33-0 i 29"5 i 29"6 26.8 I -- I -,,

Vg *

~

very

stD~ll2;

S

~

Mrozlff,

itl

,, ~

,,,,

4.2 4-2 4.2 4.2 4.2 4-2 4.2 4.2 .,,

1LiOn|crate;

3.g~s, 4.22s ; 3.7m, 3.9m, 4-6.% 5,3m 3.79s, 4.02w, 4-19s. 4-37w I 3.f;5m. 4.0m. 4-6.% 5.3m 3-8s, 4-22s i 3-7m, 3.9m, 4-6s, 5-3m 3-81s, 4-12~, 4.3s ! 3-05m, 4.0m, 4.6s, 5-3m 3.8s, 4.22s 3.7m, 3.9m. 4-6s, 5.3m 3.85m, 4.04w, 4.2fis, 4-43w 'i 3.65m. 4-0m, 4.6s, 5-3m i 3-78s, 4'18vs* 3-7m, 3.9,n, 4.6s, 5.3m 3"SSs. 4-25s, 4-52w i 3-65m. 4-0m, 4.6s, 5-3m ! 3.7m, 3.9m, 4.6~', 5,~n ,,,

,,

,

W

::

..

,

,

,

weak.

I,I-TTON.,=~)

Diacid tr(glyceride, with acid.s d i f f e r i ~ by two carbon atoms T h e s e fall i n t o t h e f o l l o w i n g fi)ur gr, m p s in w h i c h t h e s i n g l e a e y l g r o u p is s h o r t e r or longer than the remaining two aey] groups by two e,rbon atoms. 16

Diacid Triglyeeride~ with Acids differing by Two Carbon Atoms

ca,c.c., c.c.£...

Group (a) C~2C,oC~, C l ~ C l 2 C l l , Group (b) C~oCx~C~o, Cz2C~Cz2, Group (c) C~oC~2C)2, CzzC~C~, G r o u p (d) C12C1oC10 , C14C12C1.~ ,

C,,C,~C,~, C,~CzsC~,. C~eC~C~, C~sC~C,,.

In each group independently, the irregularity in structure due to the different lenKths of the chains, is the same for each member, and since thcre~is a uniform increase in the lengrth of the molecule from member to member, regularities in m.p. and x - r a y d a t a are probable. %

~! .

C18

--

__

C3b

c18 (c)

--

c~ b

c,~

CI~

C16

~,

~

~.~

j,----

3~8

cm

cls

"

C~

'

j,

------

'

qs

;'

Cla

(d)

t.______

CI 8

--

c~6 CI 8 _ _

%

¢18 clb ci 5

t._.__,

All the above glycerides exist in the vitreous, a-, fl'-, and fl-forms, the m.p.s of which fall on smooth curves for each of the above groups, Fig. 11, and the long spacings for each group show a linear relationship to carbon content. The M.P'S OF SYMMET~AI. TR;GLYCEIUDE$ .a'

/,/,4..L /2"//! /Y//

~d

sd

M.IE'S OF UNSYMH~RICAL TRIf~YCERIDES

I IM~,I

h ,~ I ,G ,~:/,,~ / "'7 °° y,d,,,

VIT~IECY~

~

%

f-

//4/

Jd

-~

i

~d

z

o ,

z_

z

_z

z

~

~

o

o

o

o

g

-

~ ~ -

~-. ~ :T,

-

=

/

/L m

z

z [ i ~L , .

///~( ..7/~,,.oo,.

z

L

/

z

z '-

z

";_

~

z

z

o_

5

~:

"

±

~ L

Z

'7

Fig. l 1. Melting points Of diacid triglyc~wides.

m a g n i t u d e of the long spacings suggests the same structure as t h a t of the simple triglyceride3, i.e. a head to tail tuning fork structure, and if the spacings are plotted against twice tile length of the longer acyl chain, tile intercepts at C ~ O for (a), (b), (c), and (d) are respectively 2, 1, 2, O, 2~-, which suggests some slight interpenetration of the chains. 17

T h e P o l y m o r p h ! s m o f Glycefides Short spacings are similar to those of simple triglycerides and consist of broad a n d n a r r o w b a n d s f o r / 6 - a n d fl' f o l m s r e s p e c t i v e l y , except, f o r t h e m e m b e r s o f g r o u p (b), w h e r e b o t h / 6 ' - a n d f l - f o r m s g i v e rise t o a n a r r o w b a n d o f s h o r t s p a c ings. B y w h a t a p p e a r s t o b e m o r ~ t h a n a c o i n c i d e n c e , t h e side s p a c i n g s o f t h e fl-forms of the members of th's group are practically identical with those of the fl'-fi)rms o f s i m p l e o d d a c i d g l y c c r i d e s o f t h e s a m e a c y l l e n g t h , e.g. t r i m a r g g r i n and 2-stearodipalmitin.

Table 4. Melting points of diacid triglycerides Unsymn~tricaln:J

SymmetricalaS* Vitreous

Vitreous

a

/ [ l~" !

DecodHaurin . Laurodimyristln Myristodipalmitin I'abnitodistearin Laurodidecoin Myristod ilaurin Pahnitodim)-ristin Stearodipalmitin

2 24 37 50 24 38 49

il °

23 35 46 56 25 37 49 59

33 45 55 64 34 44 55 ~5

38"5 50 60 68 37"5 43 58'5 68

5 22 36 50 0 19 34 46-5

26 37 47.5 57 17.5 33-5 45-5 55

i 31 [ 42 52 i 61 ! 26 39 50-5 59.5

35"5 46"5 57 65 30 43"5 54 62-5

Table 5. Long and stwrt spacings of diacid triglycerld~ "~1~, (17~ Short #'

2-Decodilaurin 2-Laurodimyristin 2-Myri.~todipalmitin 2-Palmitodistearin 2-Laurodideeoin . 2-Myristodilaurin. 2-Palmitodilnyristin 2-Stearodipalmitin l-Deeodilaurin

1-Laurodimyristin 1 -Myristodipa},nitin 1-Palmitodistearin ] -Laurodidocoin . 1-Myristodilaurin. l-Pahnitodimyrist in 1-St~arodipalmitin

#

--[--13o o --

3-79m, 4.06w, 4.35w, 4-62s 5.3m 39"636"734"714"19 3.82s. 4-05w, 4.16s, 3-84m, 3.89m, 4-61s, 5.3rn 4-35w 4 4 4 4 ° 4 3 9 0 4 1 9 ' 3-82s, 4.05w, 4.16s, 3-74m, 3.86m, 4.61s, 5.3m ' i --" • " , " 4-35w 50.547-544-24-19 3.82s, 4.05w, 4"] 6s, 3-68m, 3.86m, 4-61s, 5.3m 4.35w 3-87s, 4.17m, 4.39s 3.85s, 4"35* 3.86m, 4-06w, 4.26a, 4-45w 3-88s, 4-]3% 4-31s 3-81s, 4.13m, 4-31s 3-8hn, 4.03w, 4.2s, 4-48m .150.2 44-7 43-2 4.19 ! 3.81s, 4-35* I

3.67m, 3.86m, 4-6.% 5-3m 3-67m, 3.S6tn, 4-6s, 5-3m 3-G7m, 3.86m, 4-6s, 5.3m 3-~;7m. 3..~6m, 4-6s, 5-3m 3-75m, 3.86rn, 4-6s, 5.3m 3-75m, 3-$6m, 4-6s, 5.3m 3-67m, 3-86m, 4-6s, 5-3m 3.67m, 3.86m, 4.6s, 5-3m

3-83, 4-18, 4-35 3-83, 4'18, 4'35 3.83, 4'18, 4.35 .148.8!44-746-;5'4 •19' 3.83, 4-18, 4-35 3.83, 4-18, 4.35 • 134-5 33.0 - - ! 3.~3, 4-18, 4.35 . i42-8139-5 37.7~4.19: 3-~3, 4.22, 4.35 • 47-8'43.9 4.25 4.19 3.83, 4-18, 4-35 ,,,,

* = diffuse lines. 18

,u,,

,

Diacid Triglyeerides (Acids differing by more than Two Carbon Atoms) There is a marked difference in the speed of the transitions between the symmetrical (a) and (b) and the unsymmetrical (c) and (d) glycerides. With the former, the transition :~ -+/~' is particularly rapid, and cooling curves of group (a) in particular show an unusually large rise in temperature: due to this change (see Fig. 3). This marked rise is not observed for any of the unsymmetrical glycerides, which are characterized by very slow transitions. As a consequence, cayillary m.p.s of the various polymorphs of unsymmetricM glycerides are re,~dily determined, for tins group exhibits true "double mel~ing," and no trouble is experienced with "softening" points. The transition fl' --~ fl is so slow in this group that even from solvents, the fl'-form more commonly sep,~rates than the stable fl-form The latter is best obtained by extremely slow crystallization from non-polar solvents, preferably in a Dewar vessel. Unsymmetrical glyceridcs are easily distinguished, by their less crystalline appearance, from the symmetrical compounds, which form glistening masses of felted needles: under the microscope the former appear as short thickish prisms, and the latter as long slender, thin prisms, fl'-forms have the outward ~ppearance of precipitated chalk.

Dk¢cid triglycerides (acids differinq by more than two carbon atoms) The polymorphism of this group is similar to that of other diacid triglyceridcs, the transitions being rapid for symmetrical and slow for unsbmametrical compounds. This is illustrated in the cooling curves for the symmetrical compounds, Fig. 6, which show a sharp rise in ~mperature as crystallization sets in, not observed on curves for the unsyrnmetrical compounds. The fl' -+ fl change for the latter group is exceedingly slow, and in the case of 1-1aurodistearin, the fl-form has not been obtained, although a m.p. comparison with the symmetrical isomer, suggests that it should exis$. Table 6. Meltinfi points of diacid triglycerides m.

,i

Symmetrical~

Decodimyristin Laurodipalmitin Myristodisteariu Myristodidecoin Palmitodilnurln Stearodimyristin Decodipahnitin Laurodistearin PalmitodidecoL, a Stearodilaurin. S~arodideeoin Decodistearirx .

U t ~ y r n m e t ~ c d n*j

TM

Vitreous

a

fl"

fl

I'itreous

16 34 47 3 19" 33 20 36 6 21 5* 30*

37 47 56 21 35* 47 42 52 27 3S 34 47*

40 50 59 30 42-5* 53 48* 58" 36 43 40 53*

43.5 53-5 62.5 34 45-5 55.5 hi.5 60.5 40 47 44.5 57

15 32 44 3 20 36 23 36 2 20 13 33

i

p

* Difficult to dek.rmlneowtng to speed of transitions, probably accurate to I*C~

I9

~t

32 45 54 20 33 46 37 47 24 3l 32 42.5

fl"

fi

38 49-5 57-5 31 43 52 41 52 32 41"5 38 46

43"5 54-0 62-0 34-5 46"5 56 45-5 35 45 41 49

The Polymorphism of Glycerides When members of this group are arranged in appropriate ~omilies, e.g. C14C10C14, CleC12C16, C1sC14Cls, they exhibit regularities in m.p. and x-ray data, but some divergences occur among the x-ray data, owing to possibilities of both the simple and the triple-chain-length structui'e, and a more detailed x-ray investigation will be required before some of these points can be cleared up. The greatest difficulty in this group, is to obtain a clear picture of the change Prom the simple to the T.C.L. structure by a mechanism not involviug a large heat change since this would appear to involve the breaking and re-making of a large number of lateral methylene linkages. Moreover, the conditions under which glycerides of apparently similar structure adopt one or other of these structures, is not clearly understood, e.g. the fl'-forms of 2-decodipalmitin and 2-1aurodistearin.

Table 7. Long a~ui short spacings of diacid triglycerides <1~1,(18) i

.

I Longspacings

Side spacings

2-Decodimyristin . 33-7 52"5 ! 3.92 4-15 4.39w 2 - L a u r o d i p a l m i t i n .!44.6 77-0 59-0 3.81 4.08 4.36 2 - M y r i s t o d i s t e a r i n . 149.5 144-7'65"81 3.77 4-12 4-26 4-3w 4.6w 2 - M y r i s t o d i d e c o i n . - - 130.3 46-5! 3.9 2 - P a l m i t o d i l a u r i n .I - - !36"6'35"51 3.9w 4.09w 4.29 4.44 4.34 2 - S t e a r o d i m y r i s t i n . '144.0 41.0~40.0 3 . 8 8 w 4 . 1 7 2 - D e e o d i p a l m i t i n .i39"0 ' 7 4 " 0 5 6 ' 5 ~ 3-82 4-08 4.31 2 - L a u r o d i s t e a r i n . 47.1 42-463.71 3.82 4.08 4.31 2-I'almitodidecoin tI--149'51 4-41w 2-Stearodilaurm .!40 8 37-5156.8 i 3.83 4-02w 4-2 2.Stearodidecoin i -- !151.6~ 2-Decooistearin ~76-3'61-2! 3.83 4-]1 4.34 a.12 4.33 1 - D e c o d i m y r i s t i n .IF - - 33.8 3 5 . 2 3.8 4" 5"~" 1 - L a u r o d i p a l m i t i n . 43.4 138-5 39-8! 3.78 4 - 0 9 w 4 . 2 6 . • . I I ' q 4.5w l - M y r i s t . o d l s t e a r m . i48-5 j43-4 45-01 3.78 4 - 0 9 w 4 . 2 6 4.1w 4-26 1 - M y r i s t o d i d e c o i n .] - - 131.3 47-51 3-8 1-ealmitodilauri..[1~36.2;54.6] 3-78 5.09w 4-26 4.15 4-27 l - S t e a r o d i m y r i s t i n !46-4 141-7 61-4 3-8 4.35 1-Deco
3.84 3"84 3"84 3-86 3"85 3.82 ,3'84 3"84 3"85 3-$5 3"87 3.86 3.72 3"72 3.72 3"84 3.84 3.84 3"84

4-2w 4"2w 4"2w 4.09w 4"17

4-3vw 4'3vw 4-24w 3"9 3"9 3"9

4.27w

4-6 4-6 4-6 4.6 4"29 4-34 4.61 4.61 4-58 4.58 4-58 4'61 4'53

5.25w 5"25w 5-25w 4"96w 5'29w 4.44w

5-32w 5"32w 4-Sow 5-27w 4.Sow 5-27w 4-9w 5"29w 5-34 4"67 5-4 4"534-67 5"4 4"53 4"67 5-4 4-6 4 - 8 4 w 5 " 2 1 w 4.6 4 - 8 4 w S ' 2 1 w 4-6 4 - 8 4 w 5 . 2 1 w 4"6 4.85w 5"26w

3"84 4.27w 4-6 4-85w 5"26w 3.84 4.27w 4.6 4 . 8 5 w 5 " 2 6 w 3-84 4.6 5"3 3'74 4-02 4'33 4.6w

All ~.forms give a single spacing 4"19 A. w = weak lines. * Possibly ft.

The vitreous form of triglycerides With the exception of a few additional data of CLARKSON and MALKIN~231 in 1948, those so far discussed are due to 3IALKINd al. up to 1939~ and they lead to the satisfactory and harmnnious explanation of the polymorphism of triglycerides which was summarized on pp. 2, 3. In 1945, however, LUTTO.~~2°~and 20

The Vitreous Form of Triglycerides other American workers, ~21~,~22~none of whom had had any previous experience in this difficult field, advanced the view, on the flimsiest of evidence, "that the concept of the glassy state of triglycerides should be eliminated" and that therefore triglycerides exist in only three solid modifications. But since 5L~L~rn eta/., over the previous dozen years, had published m.p. data for some forty triglycerides, all existing in four forms, this new suggestion he.e naturally given rise to some confusion, which has not been lessened by LvTTo.n's misuse and misapplication of MALKIS'S terminology (see p. 13). Nor ~ere matvers impro~-ed when LVTTO.n,(2s~, (32) who for long insisted on the existence of only three forms, began at last to find four, and was consequently compelled to introduce a still more confusing terminology. This view has retarded progress so greatly that it has become an urgent matter to settle the question finally, and in order to do so LCTTO.~'S view will now be examined a little more closely. There is, of course, nothing inherently improbable in the view that glyceridcs may exist in a vitreous form. TA_',I_~A.n.~',one of the greatest authorities on the phenomena of melting and crystallization, expressed the view that most substances could probably be obtained in the vitreous form. To illustrate this he took, more or less haphazardly, as many organic substances as he could find about his laboratory (153), and melted and cooled them at different rates, and f.3und that 59 could exist as glasses. (~3~ Many of the early workers recognized that the lowest form of glyccrides was non-crystalline, thus HELXTZ(1) speaks of tristearin cooling to a transparent porcelain-like mass, and tlm same expression was used by REI~ER and WILL Cs~ for trimyristin. LOSKIT~1°~stated that the lower glyccrides solidify as glasses, whilst GUTtl (5}speaks of an amorphous labile condition. Quite recently, RAV~CH, ZVRI.~OV, VOL.'~OVA,and P~T~ov, (34) in a thoughtflfl paper, which sho~s a clear understanding of the principles of glyceride crystallization, adduce optical and 6thor evidence for the existence of the vitreous form of glycerides. Part of their summary is worthy of quotation. "A thin layer of molten trilaurin between covcrglass and slide, upon being cooled off at a given rate (by placing it on a cooling stage of temperature 4°C) yielded the transparent vitreous y-form, which does not reveal anisotropy on microscopic examination in polarized light. This form is exceedingly brittle, shatters into fragments at a blow and upon softening assumes a salve.like consistency. Upon extremely rapid cooling of trilaurin melts (by immersing the samples in liquid air) they acquire the property of double refraction, like compressed glass plates.": Fig. 12 taken from the above paper, shows the curiously cracked appearance of the solidified trilaurin referred to, and the shattering effect when this is struck. In a recent paper from L~TTTO~'S laboratory, QuI~IB~ (35) also refers to the rigidity and brittleness of the lowest melting form, which lie finds fractures under high pressure. Now it is filirly common knowledge tlmt the most characteristic property of long chain compounds in general is their softness and "soapy" feel. This is due to the regular layers of weakly attracting methyl groups in the crystal, which form the main cleavage planes: the fact that the above low melting forms fracture and shatter is the strongest evi(lencc that these planes 21

The Polymorphism of Glyceride~ are absent, and that the solid is not truly crystalline. Qv~BY, who also considers that the vitreous form does not exist, refers frequently to the well-known spherulitic formations which are so characteristic of glycerides, but he does no¢ seem to have considered the !mpli.cations of this. Spherulitic g r o ~ h is a direct consequence of crystal gro~%h in a highly viscous melt. Practically all the naturally occurring spherulites are kno~-n to have been formed in rapidly cooled viscous magmas, and they are common in rock glasses whose chief components are alkali feldspars and quartz, Artificial spherulitic structures often develop in badly annealed glass. In short, the formation of spherulites is diagnostic of a highly viscous melt, and it can safely be predicted that a substance which gives rise to spherulites will always form a glass when its melt is cooled quickly. As is well known, the ability of a substance to supercool and finally form a rigid glass can roughly be related to its viscosity, or to the activation energy constant (B) in the relation between viscosity and temperature ~ = Ae BraT. A high value of B means that t h e molecules cannot easily change their positions, and this is clearly a measure of resistance to crystallization: it also means a high rate ofincrease in viscosity with falling temperature, and hence a high probability of forming a rigid glass at some lower temperature. The striking resistance to crystallization of glycerides has been stressed earlier (pp. 3, 4). It will be of interest to now consider LUTTON'S views more fully. According to him, most glyccrides exist in three forms, and there is no vitreous form. It was, however, necessary for him to explain the data in Tables 4 and 6, wlfich give the m.p.s of four forms for some forty glycerides. I.VTTON therefore concluded that MXLKr~ and his collaborators had c.ompletely misin¢~erpreted their results , and that their vitreous, ~- and fl-forms were really ~¢-, fl'-, and fl-forms respectively: moreover, all the fl'-m.p.s given in the above tables were fictitious, since owing to a remarkable and extremely convenient effect k n o ~ as "degree of stabilization," fl-m.p.s of pure glycerides may vary over a range of anything up to 5 °. This iatter view is too naive to warrant discussion, but it will be worth while to consider the fl'- and fl-m.p.s a little further. Crystallographically, the fl'- and fl-forms are similar• in that they possess tilted, non-rotating chains, and they are also probably both monoelinie. It is not likely therefore that their free energy difference, nor their difference in m.p. will be great. From the previous tables, the average m.p. differences are: Simple even glycerides, 2.2°; mixed glycerides (Table 4), 4-1 ° (Table 6), 3-5 °, and for a further large group of mixed triglyccrides investigated by C~EN and D.aVBERT(36~ (Table 11), 3-3 °, the extreme limits of difference being 1.2 ° and 5.5 °, and it may be mentioned, in passing, that the differences between the 8'- and fl-m.p.s of mono- and diglycerides arc of the same order. The differences between/~' and fl-m.p.s found b y LUTTON(2°~, (a~L C3=~are as follows: PSS, 0-2; PSP, 1"0; SSS, 9.1; PPP, 10.0; 5LMM, 9.9; LLL, 10-0; PMP, 13.3; SMM, 9-9; PI~IM, 7-9; SLL, 13-6; where L=='.,aurie; M = m y r i s t i c ; 1~ ~ palmitic; S ~ stearic; PSS = palmitodistearin, etc. With the exception of the first two examples, these differences are of an 22

Fig. 12. The vitreous form of trilaurin.

The Vitreous Form of Trlglycerides

entirely different order from those of MALK~, and CKE.~ and DACBEaV, and they clearly refer to a difference between the m and fl m.p.s. This would be expected to be larger, beca use of the high energy content of the vertical rotating chains. In other wards, Luwro.x's fl'-form is in fact the at-form as MALKn~ originally stated. That tiffs is so, is also clear from a consideration of the simple triglycerides. Both LUTTO.~"and ~[ALKI.'~ agree t h a t the fl'- and fl-forms are tilted. Hence the melting points of these forms should alternate, for there is no known exception to the rule that tilted forms alternate (see Vol. l, p. ]5). According to LuTro.~, however, the fl'-m.p.s fall on the smooth curve marked in Fig. 10. Before concluding this section, a little ~hould be said concerning L y t t o N ' s interpretation of cooling and heating curves. In only one case, t h a t of 2-stearo-

70

g

I

'

~

4c

~v"-

- - - -

1

~, 3c

2(3

~

, TIME

Fig. 13. C,ooling and heating curves for 2-stearodipalmitin (L~'rTO.~).

dipahnitin, has he carried out any systematic study of curves, and these were taken too rapidly to be of any real value. LUTTO.nconsiders that only two forms exist for this glyceride, namely the ¢¢-form of m.p. 46.5 and the stable form of m.p. 68.6 ° which he wrongly germs the/F-form, contrary to M_~LKLn'Soriginal dcfinition. On the o~her hand, ~ L r : k n and MrAga c15~found vitreous, ~-, fl'-, and/3-forms, melti.ng respectively at 49, 59, 65, and 68°C. L v r T o s could give no explanation whatever for curve 1, Fig. 13, reproduced from his paper, for neither arrest corresponds with either of his stated m.p.s; it is, however, a typical curve showing separation of ~-form, similar to those discussed on pp. 4, 5. After supercooling, almost to the vitreous set point, which, as has been pointed out, is often the region of maximum ~-nucleation, crystallization of the ~-form sets in, and the curve rises to the ~-m.p. In this region, some ~ --~ fi' transition takes place, but the heat of crystallization and transit.;on is not sufficient to o~ercome the external cooling and the temperature slowly falls. With a slightly smaller cooling gradient, a second rise would almost certainly appear on this cooling curve. The beating curve 2, docs in fact show a slight inflexion at the 0c-m.p. (59°), but as the entire rise in temperature to the main m.p. occupied barely a minute 23

The Polymorphism of Glycerides and a half, it is not surprising that this arrest is so slight, and that it was overlooked by Lvl~ro~. In both of the heating curves the melting begins at about 65 ° (the fl'-m.p.s el MXLKIN and ~IEhRA) and at no point is there a fiat portion at 68.6 °, the m.p. claimed by LVT~'O~-. Clearly, this arrest is due mainly to the melting of the fl'-form which is aucompanied by a small amount of fl' -+ fl transition, but at no time is there any appreciable amount of the fl-form present. In a later paper, LUTTON(a°) gives a schematic cooling curve for tristearin (Fig. 14), showing the forms separating at the different points of the curve. This is the ideal curve for all glycerides discussed earlier, except that the arrests marked fl' and fl should be :¢ a~d fl'. I f LI:TTO.~'S interpretation of the curve is applied to the case of PSP, first eonsidcrcd, it will be seen that the :t-form is crystallizing out some degrees

TIHll

Fig. 14. Schematic cooling curve for tristearin (Lt"~o.~).

above its m.p. A eonsidcration of the heat s of crystallization of the various forms shows that LUTTO_~"S interpretation of the above curve is wholly untenable. KI~-O and GARNER(37) have shown that the heats of crystallization per methylene group for ~- and fl-forms is 0-708 calorics and 1-04 calories respectively, which agrees reasonably well with the ratio of the heats of fusion determined by CMA~BO~'.~ET and SI~,-C.LETO~"(3s) for c£ and /5 tristearin, namely 38"9 and 54-5 calories per gram, i.e. 38-9 for ~ £ormatiou and 15'6 for ~ ~ fl' -~ fl transitions. In other words, "10/o o/ of the heat evolved during crystallization is due to xformation. According to Fig. 14 the heat ev()lution during ~-formation is not more than 10% of the whole. The same point is demonstrated by the steep rise characteristic of the ]mating curve when the vitreous form melts (Fig. 13 ; cf. also Fig. 3). I f the vitreous form wcre the ~-form as LI3TTON suggests, tbe only heat available to produce this rise would bc the relatively small heat of transition a - > / 3 ' , for by far the larger part of the ]teat of crystallization wouht have already bcen cvolved during the crystallization of the ~-form. Quite cicarly, the explanation is that LUTTON'S so-called ~-fi~rm is vitreous, and the steep rise in temperature i~ due to the delayed crystallization of the a-form. 24:

:Further Diacid Triglycerides Sufficient has now been said to show that the view that there is no vitreous form leads to contradictions and errors from whatever angl e it is approached. Its introduction has caused the greatest confusion, and in the interests of the subject it should be discarded.

Further diacid trlglycerides JACKSON and LUTTON~39}have reported the m.p. and x-ray data for behenyl. palmitic and behenyl-stearie glycerides given in Tables 8 and 9. Tkeir lowest melting forms are probably vitreous, and to preserve a uniform terminology,

Table 8. Melting points of behenyl mixed triglyceride~s

Vitreous

fl

PBP

SBS

47.4 61.5 66-6

56"0 64"0 70"6

•.PBB

55"9 66-1

SBB

61-3, 71-5 73.5

Table 9. Long a~wlshort spacings of behenyl mixed triglycerides

Sub -alpha Vitreous .

PBP

SBS

50"0 46-0 70"4

532 49-4 75-1

B'

58-9 57"4 52"4

4.2s

Sub-alplm Vitreous

PBB

.: 4 - 1 4 v s 4.19vs 3.76s 4.58s 3.79vs

4-14vs 4.18vs 3.78s 4.58s 3.78vs

3.72m 4"15vs 4-20vs 3-78s

SBB

61"4" 59"1" 53-3 54-6 4-2s 3-77ra 4-15vs 4.20vs 3.78s 4-598 3.878 3-70s

* It is doubtful whether this differe,ce in the long spacings of the sub-alpha and vitreous forms is real: only I wo orders were found for the former d = (i2'58 and 60"3, and therefore the exlu~rimental error is of a high order.

they have been so designated in the table. The authors report the discovery of a new low temperature form, termed sub-alph,% but no m.p.s were found for these forms, and the only evidence for their existence is the appearance of a short spacing in the region of 3.7-3.9 ~, at low temperatures (-- 50°). At such tow temperatures, however, ice condenses on the specimen, and gives rise to a spacing in this region, and it wouhl be unwise to accept the existence of the aub-ali)ha form on this evidence alone. A later paper by JACKSON, WILLE, and LUTTO_~¢4°1gives m.p. and x-ray data for 2-acetyl, 2-butyryl, and 2-caproyl distearin and dipalmitin (Table 10). 25

•

TLe PolyrnGrphism

of Glycerides

Although six different polymorphic forms are claimed to exist in this group, none

exhibits

more

than

three

m.p.s,

and

a critical

examination

d a t a s u g g e s t s t h a t o n l y t h e :¢- a n d t h e s t a b l e , 8 - f o r m s a r e d e f i n i t e l y

of the

x-ray

characterized.

Table 10. T h e r m a l a n d x - r a y data of acetyl, "butyryl, a n d caproyl diglycerideg 4°) SC,S 1'C21)

8C,S

I ,l

p C , P I S C , S ;I SC,S C,P '.... I'. . .I'C,P ............

SC,S

r

l

SC,S II.C,P I ]'C~I' .....

Melllng point.a. °C } I

Sub -alpha Alpha . Super-alpha Sub-A-beta Sub.B-beta Beta

40.4

-

•! • } •

I

I o

ij

~,

6-'8 ~ __

13-4

3341

206

42.7 ! t

32-6

J-

0 , ' 8 i a4.8

b

_,

I I

28-4 ~

I

t

2':

46-0 46-0 r, o3.1

-

46.0

E

). . . . . .

277 40-5 40.5 44-5

. . . . . . . . . . . . . . . . . . . . . .

X.ray d,Tta : long spacings, ,4 4

S u b - a l p h a - 1 (-3) Super-alpha-I Sub-A-beta-3 Sub-B-beta-3 I~eta-3

28'2

25.7

• 52.1 . . . .

47.2

29.8 28-5 26-6

26.9 i 30.4 25.9 ] 30-0

24.2

48-4

155-6 • 54.5 45-0 56.8

! . . . . . . . . . . . . . . . . . . . . . . . . . . .

53-7

52.8 49.2 52.5

I

Short ~pacings, .4 4,20s 3,78m 2.50vw

Sub-alpha

'

Alpha .

4-14s .4 2-44~,vi ! 15.75s* 5-00w 4-14vs 2-43vw

Super-alpha .

Sub-A-beta

.

Sub-B-beta

.

Beta

4.15vs 3-71m

4.20vs 3-70m ', 4.15vs 2-50vw

5-15m 4.69s 4.55s 3.86a + 3-55m 5-38m 4-59vs 3-87s 3-66s 4.59s 3-87m 3-72vs

4-60va

j 4-04s 3-85s 3"69m

3.5a~ ° "/'his is obviously not a short spacing. 26

5.19m

4-6]vs 3.82vs

Triacid Triglyeexides The authors' description of the behaviour of SC~S and PCaP is of considerable interest and is worthy of quotation. "The super-alpha form, observed for SCtS, and PC4P, is unique thus far for triglycerides. The name was chosen because the new polymorphic form had the characteristic 4-14 ~ short spacing of alpha while its melting point was 10-12 ° higher than that of alphaThe super-alpha form was obtained by tempering the alpha form at 0-5 ° below its melting point for 2-4 minutes." Thus the authors unwittingly confirm MXLKrN'S view as to the nature of their alpha form. As was pointed out earlier, the lowest melting forms are vitreous, but they often contain sufficient crystallites of the ~-form to give an =~-ray pattern of this form. "When this form is held near its m.p. it changes into the higher melting ~-form (Lu.~To~'.~ super-alpha) which naturally gives the spacings of the a-form. Far from being unique, this is the general behaviour of all triglycerides.

Triacid triylyverides CHE~ and D~u~ERT (~6~ and SIDHU and DAUBERT(41) have determined the m.p. data for a ~eries of ttiacid triglycerides, given in Table l l . I t is not clear how

Table 11. Melling points of triacid triglycerides (3~. (tu II

I

CI.---CI.---CI. il Cx.---Cl.----Cl. CI.----CIo----Ci. CIa---C1.---CI. i CIs--C~.---Cz, CIB--q310--CI, C18---C1,--C12 • C3.---CI(--CIffi • C18---C1o~-C1, Cl~--Cl.---Cao Cx,----C1,--CIo Cas---CI.---CIo C16---C~-----C~ C~---C~f---C~o

56-1 56.0 53-8 56-0(54.0) 51-9 50-1 47.0 45-5 41.8 46-5 42-0 40-0 44-0 34-0

59-5 57-5 55-0 58.5(59.0) 55.0 52-5 52.0 49.5

o

50.0 45-0 4"~-0

49-0 37.0

llI 40"6 32"0 20"1 40"3(43"5-44) 28"8 14"0 33-4

27.5 22.3 26.1 21-5 14.5

37.0 22.0

far the glyeerides were investigated b y cooling and heating curves, for no details are given; but m.p.s are given for three forms termed I, II, III. CHES and DaU~ERT made ~he interesting observation that several members of the group could be obtained in form II, by crystallization from solvents, whereas, of course, tile stable form I would be expected, and in a later paper, FILER, Jr., SIDHU, C~IE~, and D.~UBERT(31~ showed, that forms I and I I correspond to r- and/3'forms respectively. Form I I I was not identified, but there is little doubt that this is the vitreous form, except for the case of C~6----C14--C12, where it is probably the ~-form. This can be deduced by plotting the m.p.s of the related series, C14--Clz---C:e , Cx6---Cl~Caz, C18---C1~---C14. which have the same terminal 27

"I

'I "i

Ct,--C,.--C,, C'~--Ct+--CI°

i[

Cls--Ct:--Ct°

C,~--Ct.--Ct~

C,s--C,o--C,~ C,e--C,+--C,o

C~--C,4--Ct2

C,+--C,+--C,+

C,s--C,o--Cl.

C,B--C,2--C~

C,~--CL.--C,+ C,.--C,2--C,+ Cts--C,o--CI+ C'1~ --C 1 .--Cl~.

63-4 {;t.2 60.0 40.7 59.8 57.3 62-4 59.6 54-9 38.5 56.1 35.6 35.7 33.4

Long

59.0 54.3 53.0

43.3 40.1

41.5 60.0

3.83vn, 3.8.Iv.% 3.82vs, 3.72vs, 3.83vs, 3.82vs, 3.82vs, 3.82vs, 3.82vs, 3.85v.% 3.83va, 3.81s, 3.74.% 3.82s, 4.62vs, 4.33m, 4.2hn, 3.82vs, 4.25s, 4.3m, 4.34m, ,l.32s, 4.6vs, ,t.14vs, 4.2vs, 4.19vs 3.83s, 4.13vs, 5,02m, 5.42m

5.37m 5.18s, 5.35m

5.19m 5.41m

4.82m, 5.35m, 5.55m 5.6s

4.83m, 4.78m, 5.34m 4.8m, 4.82m, 5.2m 4.82s,

4.57vs, 5.08m, 5.33m 4.33vs, 4.6m

5.24m 4.61vs, 4.61vs, 4.57vs, 4.61vs, 4.62s, 4.62vs, 4.61vs, 5.28m 4.33s, 4.45s,

Shore

3.76s, 3.73s, 3.83s,

3.86s, 4.22vs, 5.14m 3.84s, 4.05s, 4.19s, 4.16vs, 4-29s, 4.46s

3.82s, 4.1Is, 4.29vs 3.81vs, 4.26~-s, 4.45m

3.8Ova, 4.11vs, 4.34va

"l'nblc 12. Lo~lff aml short spaciugs of lriacid lriglyccrides (m), (41>

4.45s,

.

4.58s

t~

o

o

'2

1-3Ionogiycerides of Saturated Acids irregularity, yet increase regularly in length (Fig. 15). Forms I and II lie on smooth curves, whereas the centre point of the lower curve is clearly too high. From the absence of experimental details, this series does not appear to have

/

40

3o

2o i

IC v

v

v.~

v

%

%

Z)

¢~J

tJ

:Fig. 15. Melting points of triaeid triglyeeridcs.

been suhmitted to a detailed thermal study ; such a study would most probably reveal other a-m.p.s lying between those of forms II and III. x-ray data for the above glyeerides are given in Table 12. ~[ONOGLYCERIDES

1-monoglycerides of saturated acids FISCItER, BERGMANN, and ~B~-R%VL\'D(42} w e r e the first to report the double

melting of 1-mono-st~arin and -l)ahnitin in 1920. Ten years later, REWADIKAR and WATSO-~(4"~)made a systematic study of the melting phenomena of the even acid 1-mono-glycerides from monolaurin to monostearin, and showed that they exist in two distinct forms; they also suspected the existence of a third form, without, however, expressing a definite opinion. This work was confirmed and extended by ~L~LKIN and SHURBA(;Y,~4~) who, by means of cooling and heating curves and x-ray examination, found three solid modifications for all lomono glyeerides from monodeeoin to monostearin, viz. a tow melting 0t-form (vertical rotating chains) and two higher melting forms, //'- and /~- (inclined chains, alternating m,p.s). They showed that when the molt,on monoglyecride is cooled 29

The Polymorphism of Glycerides the a-form is the first to separate; this then changes into a more stable fl'-form, which finally changes much more slowly ~,to the stable fl-form. The latter can also be obtained by very slow crystallization from solvents; rapid crystallization from solvents may give rise to a mixture of all three forms, but as a rule gives the fl'-form. LVTTO.~ and.JAcKso.n(*s) later claimed to have found a fourth form which they termed sub-alpha. Cooling and h eatir~2 curves Tim curves for monoundecoin in Fig. 16 illustrate the general behaviour of monoglyeerides. AB, cooling to room temperatures, shows the separation of a-form, and BC shows the melting of a-, fl'-, and fl-forms. On increasing the 80* 70*

,A "'"

:-"'""

.C

,/"

\

Ef"'°"

,.~ S-O~

4c~ 30 ~

"--,/

/

:E v- 20'

B "'..."

",:

I(:7

0

q

30

6o TIME MINUTES

F i g . 16. C o o l i n g a n d h , , a t i n g c u r v e s o f 1 - r n o n o g l y o c r i d e a . l-monoundecoin, :-----; 1-monopalmitirt, .......... .

cooling gradient by the use of ice, a lower arrest is observed (BD) and the subsequent heating curve, DE, now rises to tim m.p. of the fl-form. No arrest now appears at the a-m.p, and only the slightest indication of an arrest at the fl'-m.p. Clearly, as the a-form cools, it changes into fl'- and fl-forms, and fl'om BDE it appears that these transitions are completed by further cooling. The curves for monolaurin are very similar to the above, and curves for other members differ in a gradated manner, according to the length of the acid chain, changes being more rapid with shorter than with long chains. Thus the a and fl" arrests on the heating curve of monodecoin are less pronounced, because of the rapid a --> fl', fl' -~ fl changes, whereas the corresponding arrests are more marked in the case of monotridecoin, and the final fl-arrest is slight. On ascending the series to monomyristin and monopentadecoin, the fl-arrcst disappears altogether, owing to the slow fl'--> fl change. If, however, the specimen is first cooled very slowly, then the latter transition takes place, and a fl-arrest is observed on the heating curve. With still higher members, both tlie fl' and fl arrests are usually absent; after very slow cooling, the/6'-arrest may be observed but the transition fl' --->fl can only be brought about by maintaining the specimen in the region of its fl'-m.r, for several days. 30

1-Monoglycerides of Saturated Acids The cooling curves for the series are alike in showing t w o arrests, the first of which represents the changes Equid -~ a-form. T h e changes a t the second a r r e s t are more complex, b u t for the lower m e m b e r s a p p e a r to be ~ - > fl' --> fl-transitions. F o r the higher members, however, these transitions do not a p p e a r to t a k e place, a n d the lower arrest is reversible (d;scontinuous curve, Fig. 16, 1-monopalmitin). Table 13

•

Docoin

Lol/~ a ~re~ff

#'

l . M onoglyeeride

27 ° 36.5 44 50 56 62 66.5 70

Undecoin Laurin Tridecoin Myristin Pentadecoin Palmitln Heptadecoin Stearin.

49 ° 52 59-5 61 67.5 69 74 74.5 79

74

53 ° 56-5 63 65

70-5 72 77 77 81-5

8o

3 15 9

24 17 34 28 47-5

and 42* * 1-Monostearin differs from the othqr monoglyeerides in t h a t there arc two distinct reversible cLanges in |.his region. ].UTTON~°j has now confirmed this, having previously failed to observe the two changes. "~D

-

,< 3?','" i

I 4C

i

• ! A"

J ,s~

# iO II 12 I]I 14 I$ 16 h~JHg~l OF C A I ~ N AiOt~ KADICAI.

17 III iq IN ACYL

Fig. 17. Mc]ting p e i n t s and" t r a n s i t i o n i e m f ) e r a t u r e s o f 1 - m o n o g ] y c e r i d e s .

The various curves can be modified b y varying the r a t e of cooling. Slower cooling increases t h e velocity of tile transitions, so t h a t if, for example, monopalmitin is cooicd v e r y slowly, it gives the t y p e of c u r v e n o r m a l l y given b y trimyristin, and shows evidence of the fl'-form on the heating curve. Co/~versely 31

The Polymorphism of Glycerides the curves for monomyristin cooled rapidly are similar to those shown in Fig. 16 for monopalmitin, and the lower arrest appears to bc reversible. This reversible effect is therefore due to the excessively slow rate of the tra o.sition ~ - + fl', and if the palm|tin curve is continued as sho~na in Fig. 16, F.G.H., the magnitude of the reversible effect diminishes, .and the presence of the fl'-form is shown by the final heating •curve. Melting points of the various forms and temperatures of the lower arrests on the cooling curves are given in Table 13 and plotted in Fig. 17. X-ray investigation

The spacings of the stable fl-f)rms are b2st obtained from specimens which have been slowly crystallized from hcxane or ether. Rapid Crystallization gives rise to mixtures of forms, mainly the fl', mixed v'ith the fl-form for the lower members of the series, and the a-form for the higher. Melted layers give rise to a-, fl'-, or/3- forms, according to the rate of cooling and the temperature during the x-ray exposure. Thus, a specimen allowed to soil(lily on a temperature controlled mount, just below its a-m.p., gives the a-spacing, together with that of the fl' or//-form, transition into which is favoured by the h;gh temperature. Within experimental error, the long spacings of fl'- and fl-forms are the same; they correspond to double molecules tilted at an angle of 54} °. The long spacings of a-forms correspond to double molecules lying vertically across the reflecting planes. Table 14. Lo~uy spacings of sab~ratcd 1-monoglycerides (44~ No. of C atoms in acid .

16 I

fl' and fl-forms . u-form

I

I

] 1 1 3 7 " 240-2

43.2

46-2

51-3

17 I 18 I 20*

Ir

45-8 48-.'2 50-0 54.4 58.3 I

f

St,~ort spacings of ,'aturated 1-monoglyeerides fl' = 3"86m, 4-24s: fl ---- 3"88s, 4-12, 4-37, 4-58s. * fl-form determined by SIDIIUand DAUBERT.~i*t

The 0t-form is stable only over a small temperature range near its m.p. and at this temperature it rapidly changes into the more st.able fl'-form. ]t is rare to obtain more than two orders on the x-ray photograph, and as a rule only tile first order appears, accompanied by several orders of the f - f o r m . Although the long spacings of the ['I'- and/3-forms are the same, the two forms are readily distinguished by their short spacings, tim strongest line_~ of which are 3"$6m 4-24s, ;rod 3"88s 4-58s, respectively. The increasing rate of transition fi' -+ fl with decreasing length of ~lvid chain is well shown 1,y the short spacings. From Ca5 upwards, fl'-forms appear to be indefinitely stable at room temperature, with monomyristin the chaTlge takes place overnight and with lower members the transition usually occurs during the time of an x-ray exposure. The or-form may occasionally give rise to a single strong line at 4.18 ~_, but ~s 32

1-Monoglycerides of Saturated Acids a rule this is accompanied by weaker lines on either side, which suggests that transitions are taking place during the exposure. Ltra~ro~n and JACKSO:~~45~ state that the ~-form is reasonably stable douaa to the lower transition temperature, but it is clear from their x-ray results that they were not dealing ~ith the somewhat fleeting a-form, e.g. the long spacing for their alleged a-form of monostearin is 8/~ too short, and it cannot therefore be the spacing of a rotating vertical form. The instability of the a-form is well demonstrated by microscopic examination. I f a monoglyceride is melted on a microscope slide under a coverslip, and placed on a heated stage just below the m.p. of the a-form, it appears through crossed nicols as a greyish liquid, without structure ; yet nevertheless, it gives a strong mfiaxial interference figure. This condition usually lasts olfly a few minutes, and

~- 4

o

~

oL1221_1_ < 20

O

I

2 3 4

5

6 7 8 9 IO II 12 13 14 15 16 17 ;S 19 2 0

NUMBFR OF

CARBON ATOMS IN AC-'YL RAOtCAL

Fig. l:q. L o n g spaeing~ of s a t u r a t e d l-monoglycerides.

considerable repetition is necessary in order to observe i~ This is the true a-form, with vertical rotating chains, such as has been observed for fatty acid esters ~a6Jand aleohols.(~TL (48~ Although the ~-forms of triglyeerides give rise to hexagonal spacings similar to those of 0c-f(~rms of monogly(;erides, they do not exist in a form which gives a single uniaxial interference figure, and hence their chains are to be regarded as non-rotating, but frozen into positions which sinmlate the structure of true 0c-forms. Under the microscope, the liquid-like ~-form of monoglyeerides rapidly passes into a variety of irregular shaws, typical of the mesomort)hie forms associated wi;h liquid erj'stals , and 3IALKIN and SIIURBAGY considered that this form changed on cooling into a senti-vitreous solid, slightly m~)re ordered than the vitreous form of triglycerides. LUT'CO_,," and J.~cKSO.X- consider that this latter solid is a new crystalline modification, termed sub-alpha, which changes reversibly on heating into the a-form. They find that it gives a distinctive group of short spacings (4-13v~, 3-92m., 3-75m, 3-58m), and there is no doubt that these spacings can be obtained for myristin upwards, when the molten glyeeride is solidified fairly quickly. The cooling and heating curves already discussed, particularly those for monopalmitin, show, however, that the a - > sub-all)ha 33

The :Polymorphlsm of Glycerides cha::g, is not genuinely reversible, but only appears to be so because of the slow rate of the 0t - > fl' transition; cf. for example, the interesting dielectric constant study by CR0WE and S.~I"¢TH,(49~ where it is shown that !-monopalmitin and stearin take several hours to pass from the 0¢- to the fl'-form in the neighbourhood of the ~-m.p. Whatever the true nature of this sub-alpha form, it is mainly non-crystalline in the ordinary sense of the term, for taking similar acid chains, the monoglyceridcs crystallize much mere slowly than triglycerides, and 4re thus more likely to solidify in a vitreous phase on being rapidly cooled. This slow crystallization is probably due to the possibility of a variety of modes of hydrogen bonding between the two free hydroxyl groups of each molecule, which make it more difficult for the molecules to assume the correct position for crystallization, e.g. the slow crystallization of polyhydroxy compounds such as glycerol and sugar is well known. Moreover, the high m.p. of 1-monoglycerides, compared with alcohols and acids, suggests that the structure is more complex, and may involve lateral hydrogcn bonding in addition to the normal head to head bonding. The still higher melting acid amides afford a parallel example. It is of interest to note that the 2-monoglycerides, which on grounds of symmetry might have been expected to melt higher than the 1-isomers, actually melt several degrees lower. This suggests that the hydrogen bonding of the former is of the head to head type.

2-~.lo~wglyccridcs Tbe only thermal investigation of 2-monoglycerides is due to DAVBElCT and C L A R K E , (51) w h o examined a series from monocaprin to monosbearin, and who came to the conclusion that they were not po!ymorphic. FrLE~ Jr., S I D I I U , DAUBERT, and LO~-GENECKER(~) had earlier determined the long and short spacings of the solvent crystallized forms of this series, and although their data are complicated by an omission to filter fl-radiation adequately, their results are in excellent agreement with the unpublished data of BnvxY arid 5L~LKiN given in Table 15.

Table 15

2-Monocaprin 2-Monolaurin 2-3lonomyristin 2-Monopahnitin 2-3lonostcarin

M.P.

S.P.

40.2

34-0

Long

spacing* 29.4

3~.S . .

" 69.0 75.2

58"0 65-5 70.0

3(;.2 40.4 43.9

Short spacings

4-07, 3-96, 3.92, 3.88, 386,

4-38, 4-40, 4.42, 4.42, 4-43,

4-58 4"58 4-60 4-60 4-60

* These correspond (o double moleeule~ tilted a t an angle of 44:30 ". When plot;ed rgalust tile t~rl)on content of the acid% tile intercept at C : 0 is ----- 1 l/t.

A close examination of cooling curx-cs shows, however, that 2-monoglycerides probably scparate from the melt in the 0~-fi)rm. For m(mo-caprin to monomyristin, this changes rapidly int.() the fl-fi)rm, and the only arrest on the heating curve is at tire fl-m.i). Cooling curves for mon,)palmitin anti monostcarin show a 34

1,3-Diglycerides second arrest at a lower temperature. This appears to represent a reversible change, for there is an arrest at the same temperature on the heating curve, which then rises to the fl-m.p. There is thus a certain parallel with the behaviour of the 1-monoglycerides, but the transitions are much more rapid, and there is no indication of the existence of the fl'-form. Curves for 2-monostearin and 2-monomyristin are given in Fig. 19. All the 2-monoglycerides exhibit spheruhte formation ff cooled at a suitable rate, and with a little care, 2-monocaorin may be sho~aa microscopically, to exist in a uniaxial form.

80

.u

~2

2O TIME

Fig. 19. Cooling and heating curves of 2.monoglyeeridos. 2-monos~arin (AA); 2-monomyristin (BB).

sc

!

/

4C

/ 3c

/

2o / /

/

J

L

J

I

Ic

J

2 4 b 8 IO 12 14 Ib 18 NUMBER OF CARBON A'T.ONF IN ACYL RADICAL

Fig. 20. Long spacing~ of 2-monoglyeerides.

1,3- Diglycerides The polymorphism of simple diglycerides was first investigated by ~L~LK[N, SHvRm~c¥, and 3IEARA, (52) who showed that they were similar to the mono- and triglyccrides in existing in x-,/3'- and fl-forms. They reported an apparent discontinuity at dipcntadccoin, since only one fl-form was obtained for tbe highcr members, but later, BACR, JACKSO_~',KOLr, and LVTTO-~(s3) observed the missing fl-form for dipalmitin and distearin: hence, the entire series from dicaprin to 35

The ro]ymorphism of Glyceridea distearin, behaves normally, • and exists in ~-, fl'-, and fl-forms. The latter authors showed that diglycerides often crystallize from solvents in the fl'-forms, and they also observed a new type of short spacing, very similar to those described by 5IALK~ et al., but as'there were now three types of short spacings to describe two forms (fl' and fl), there remained some slight confusion, which was removed by a later re-examination of the problem by H o w e and MXLKIN.(54~ 70'

-\

50 ~

~ 30© 20 ° I0 o 0

IO

20 "fllqE, MINUTES.

30

Fig. 21. Cooling and heating curves for 1,3.dilaurin.

80* =

9 IO II 12 13 14 15 lib NUNBER O1: CARBON ATON$ IN ACYL RADICALS

17

18

Fig. 22. Melting points of 1,3-diglyeerides.

SIDntr and DaUBERT(sS) have shown that a number of mixed dig!yeerides exhibit tile same type of po}ymorphism as tim simple diglycerides. Diglyceridcs give rise to spherulites which are distinguished from those given by triglycerides by a curious rippled effect.

Cooling and heating carves Cooling curves show only one arrest (~-separation), but two usually appear on the heating curves. For dm lower members, with not too rapid cooling, these 36

1,3-Diglycerides are at the fl'- and fl-m.p.s, whilst for the higher members, or for the lower members after rapid cooling, they occur at the :c- and fl'-m.p.s The curves sho~m in Fig. 21 for dilaurin are typical of those for the lower members. In general, :c -~ ~' transitions are rapid, but fl' --> fl transitions are unusually slow and for the higher members, occur only near the fl'-m.p. BAVRct al. have sho~-n that the latter change takes place at rates which are inversely related to the chain length. Melting points are given in Table 16, and those of the series dicaprin to distearin are plotted in Fig. 22.

X-ray investigation x-ray data for fl'- and fl-forms only have been recorded. Owing to the rapidity of tile transition :c -->/~', it has not been found possible to obtain x-ray data for

30

z 2o I0

0

-

I

2 3 4 5 6 7 8 9 IOII 12 I) 14 15 Ib 17 18 19 20 NUMBER OF CARBON ATOMS IN ACYL RADICAL

Fig. 23. Long spacings of 1,3-diglycerides.

the :c-form, and the identity of the low melting form as an :c-form, can only be inferred by analogy with the bchaviour of other glycerides, and by the/'act that the m.p.s are non-alternating (see Fig. 22). Tlie long spacings of the simple diglycerides correspond with double molecules tilted at angles of 75~ (fl) and 64~- (fl'). When plotted against earbon atoms in ttle acyl chain, tile intel'ccpts are ~- .~A, see Fig. 23. It is seen from Table 16 that there are three distinct groups of short spacings, which have been termed the a, b, and c types. They occur as follows: a type: fl-forms of all odd acid 1,3-diglyceri,les, and fl'-forms of dipalmitin and distearin. b type: fl'-forms of all 1,3-diglycerides except dipalmitin and distearin. c type: fl-forms of all even acid diglyceridcs. These results emphasize the difficulty of attenipting to classify polymorphic forms ofglycerides according to the type of short spacing (see 1). 13). Both fl'and/%forms of diglyccrides give rise to tile strong 4-6 line, which according to LYTTON'S classification, is characteristic of/~-forms only. 37

00

¢.o

* SIDtIU a n d ] ) h U B E R T . ~6b~

Didocoirt . Diundocoin D:laurin Ditridecoin Dimyristin D~pontadccoin Dipalmitin Diheptadccoin Distcarin 1-Mvristyl-3-caprin* 1-Pa!mityl-3-!aurin* 1-8tearyl-3-myristin*

#'

42 ° 47 54 57 63 66.5 71'5 72"5 77 44 56 63

a

37 ° 43.5 49-5 54.5 60 63.5 68 71"5 74 39 51 56

Melting points

30.5 33.2 35.7 38.1 40.5 42.5 44.7 47-7 49.5

44.5 ° 49 56.5 59.5 65.5 6b.5 72"5 74'5 78 48 59"5 67 --

--

B'

#

32.5 35.2 37.4 40.4 42.6 45.0 47.5 50.3 52.8 38.0 42.7 47.6

#

Long spacings A

b b b b b b a b a

4.65s, 4.63s, 4.66s, 4.58s, 4.64s, 4.58~, 4.6s, 4.6s, 4.6s,

3.93m, 3.88w, 3-94m, 3.9w, 3.88m, 3.86w, 3.87s, 3.85m, 3.88s,

#'

i

3.7~, 3.72s, 3.73s, 3.71s, 3.68s, 3.70s, 3.65s 3~7s, 3.68~

3.54w

3.46w 3.58w 3.61vw 3.53w 3.53vw 3-52w

Short spacings, A*

Table 16. Melting points and x.ray data for 1,3-diqlyceride, (52), ("~

c 4.55 a'4.53 c 4.6 a 4.55 c 4.57 a 4.58 c 4.6 a 4.57 e 4.6 o 4.55 o 4'63 a 4'6

3.75 3.81 3.66 3.75 3.83 3.68 3.73 3.84 3.08 3.75 3.82 3.66 3.73 3.84 3.73 3"8 3.76 3.76

#

3.

0

1,2-Diglyceridea 1,2-Diglycerides

The only study of the polymorphism of 1,2-diglycerides is due to ]=[OWE and M_~Krs, cu~ who showed t h a t they exist in two solid modifications, a and fl, the transition ~ --> fl being much slower than is the case with the 1,3-isomers. Thermal examination

Solvent crystallized material melts at the fl-m.p, and resolidifies at the ~¢-m.p., remelting again at the ~¢-m.p. The transition ~ -+ fl is very slow, but ff a-forms are maintained overnight, just below their m.p.s they change into the fl-forms. Cooling and heating curves normally show arrests at the a-m p. only• 1,2-Diglycerides exhibit spherulite formation, but the rippled appearance characteristic of spheruhtes of 1,3-diglycerides is absent. X - r a y examination

Long spacings of a- and fl-forms are obtained from "melted" and pressed layers respectively and short spacings from "melted" and pressed rods. When ~,lotted against the carbon content of the acyl chain, they fall on two straight lilies which cut the axis C=O, at -~- 6.5 (fl) and 9.5/~ (a). The results agree with an arrangement of double molecules lying vertically (:¢) or inclined a~ an angle of -~- 64 ° (fl) to t h e reflecting planes. Short spacings are practically identical for the series, and they differ in type from those given by the 1,3-diglycerides. 3I.p. and x-ray data are given in Table 17. Table 17. Mel~i~] points and x-~'ay data for 1,2-diglycerides ( ~

I

a-Form

M.P.

Lon~ . Short ~pacing,A lepacing, A

"

~-Form

M.P.

Long spacing, .4

39 ° 54 63.5 71

34-1 38-8 43.5 48-3

Short spacing, A

I Dilaurin Din,yristin . Dipahnitin . Distearin

20.0° 37.5 50.0 59-5

59.2 44-4

49-3 54-5

4.13 4-12 4.10 4-12

m

vs

s

4.31 4.31 4.27 4.27

4-01 4.06 4-03 4-05

3.79 3.82 3-76 3.81

Uns&turalezl glycerides

Glycerides containing unsaturated acids exhibit the same type of polymorphism as the saturated glycerides, and there are no major differences. Indeed, tlmse containing trans unsatura*~d acids are very closely similar in behaviour and structure to glycerides of the corresponding saturated acids. As might be expected, however, from the slight irr¢,gularity which a cis bond introduces into the chain, there are structural diffcrcnces in the case of cls.acid compounds• This is particularly noticeable with oleo-disaturated triglycerides, which have a general tendency to crystallize in triple-chain-length structures, ilTcspective of the relative lengths of the chains. 39

el

ell.

.I .

.

.!

!

i

' 1 - - 32* "1 15.5 'i 6 . I 43

12-9

-- lO-5(fl) -- 24"2

--

-- 13"1

--12" 37 17 50

25

B'

Form .¢

4.9 42 30 59 59'3

45"2

-- 27-29°(~)

I

54"7

45"8

3'

Long epacinffs

43"3 44"1 51"1 53"6

43 -- 45"6 i -- ~ 3 - 4 7 ° ( v i t r e o u s ) -- 44.6 --

Form I I

4"53, 4'28, 3 ' 8 8

4'35, 3"87

a'

5.28, 5.3, 5.24, 5.3,

4.57s, 4.6s, 4.6s, 4.6~,

Short spacings

* FgRC;IYSO,naml Lva"ro,~ term these x and /J', but It is unlikely that there is ao great a dlfl'erenee in melting point between #'- and B-forms (seo p. 22),

RAV[Clt etal. Trilinolenitt

DAUBEItT

Trilim)h,ia : WHrzELIet~ et al.

Trioloin Trielaidin . Trierucirx Tribrassi(lin

Vitreous

Table 18

3.97, 3-9,. 4.03, 3.9,

#

3-84, 3.71 3.7 3.84, 3.7 3.7

O m~

0

Umqa'.urated Mixed Triglycerides

Simple trigly~rides The poiymorphism of siml=le unsaturated triglycerides has been investigated by CXaTER and MX~KL~,(56~ who found three forms for trielaidin and tribrassidin, and four forms for trierucin. WHEELER,I~IEMENSCltNEIDER,and S~'~DO(s~ reported three forms for triolein and two for trilinolein, and the latter result was confirmed lager by DAURERT and BALDWJ~/5s~ who also found two forms for trilinolenin. RAVICHand Zum~'ov, csg~however, have since reported three forms for trilinolein. The above investigations were carried out by means of m.p.s and cooling and heating ourves, and o ~ y that of CARTERel al. included x-ray resulgs. However, in 1947, FEROUSO~ and LVTTOXc6°~succeeded in obtaining characteristic x-ray data for the various forms of triolein. They confirmed the m.p.s reported by WHEELERet at., and brought out similarities in x-ray data with the results of CARTERa al. for trierucin, x-ray and m.p. data are given in Table 18. In general, transitions with the above gtycerides are more rapid than for the corresponding saturated compounds, hence the difficulty in completing the above table. It will be noticed that the short spacings of the fl-form of triolein and trierucin, are identical within experimental error.

Unsaturat~ mixed triglyc~rides DAU~ERT and CLARKE(nl) showed by means of cooling and heating curves that 2-oleo-disaturaged glycerides of capric, laurie, myristic, palmitic, and stcaric acids, exist in four solid modifications, and this was supported by the work of MEARA,(6z~ who found four m.p.s fi)r 2-oleodistcarin, isolated from various natural fats. Later, FILER Jr., SIDHU, ])AUBFRT, and LONGENECKER(21) recorded x-ray data for certain of the above glycerides, and shortly afterwards LUTTON(6~) examined 2-oleodistearin thermally and with x-rays, and concluded that only three forms exist. In additien to ihe difference concerning the number of polymorphs, LUTTO.N'Sx-ray results were not in agreement with those of"FILER et al., and this led ~L~LKINand Wmso.x" to re-examine 2-oleo-dimyristin, -dipahnitin, and -distearin. They were able to report excellent agreement with the thermal investigalions of DAUBERT and CLARKE,whose curves they were able closely to reproduce. T h e y also found a further modification, and showed that these compounds exist in the following five modifications: vitreous, a,/~",/~',/~, in order of ascending m.p. 2-Elaido-dipalmitin and -distearin, which were examined at the same time, were found to resemble the related 2-stvaro(tipalmitin and trisgearin respectively, and to exist in four forms, viz. vitreous, ~,//', and//.

Cooling and t~eating curves The curves of D~r BERT and CLARKEfor 2-oleodim3~istin reproduced in Figs. 24 and 25 are typical for the ser~cs. They are the normal curves for low melting glycerides, showing separation of either the ~ (AA') or the vitreous form (CC') according to the rate of cooling, followed by transitions into other forms, according to the time interval before the heating curve is taken. The series is very similar to the s)qnmetrical mixed triglyceridcs of the 2-caprodilaurin 41

The Polymorphism of Glycerides series, which is characterized b y rapid transitions vitreous - + a -+ fl', e.g. eL Figs. 3 and 25 (very slow cooling). The i n t e r p r e t a t i o n of these curves will be clear from the earher discussion on page 3ft. T h e curves given in Fig. 24

o

$ o

"HME Fig. 24. Cooling and heating curves for 2.o|eodimyristin (DAuBmRTand CLARKm).

M

X

29 w a_

"'2. TIME

Figs. 05 Cooling and heating curves for 2-ole(~timyristin (DAv~IgRTand CnXRKE). definitely establish the existence of four form% y e t the u p p e r m.p. (26.5) is still two degrees lower t h a n the m.p. of the solvent crystallized material. This led ~[ALKIN and Wmso_n to the discovery of the fifth i%rm, t h e stable

Table 19. Meltingpoints of 2-oleo. and 2-elaido-disaturated glycerides ( ~ Vitreous 2-Oleodimyristin 2-Oleodipahnitin 2.Oleodistearin

2-0 (2.1) 12-0(12"0)

23-0(22'3)

2.Elaidodipalmitin 2-E]aidodistearin 2-Oleodicaprin 2-Oleodilaurin Va]ues in parentheses

33"0 40"0 IV (-- 16"4) •

( - - 7'3)

are due to ])AUBEI',T and

11-0(]2.3) 265(2o-8) 20.5(29.8)

19-0(21.5) 29.0(30.4) 37-0(37.0) (37.6)

42"0 46-0

III (-- 10.2) (1.4) C L A R K K . ~tH

42

I= (0.6) (11-0)

26.5(°6.3) 35.o(35.2) 41-5(41.6)

28"5 37"5 43"5

52"5 58.0 I (6.2) (16.5)

55-0 61-0

Unsatura~ed Mixed Triglycerides

Table 20. X-ray dala for 2-oleo. and 2-ela do-drsat trated glycerides i

Short spacings, A

,Form rpacir,j, A

~

2-Oleodimyristin* 2-Oloodimyristin* 2-Oleodipal~Litint 2-Oleodipalmitint 2-Oleodipalmitin~ 2-Oleodistearin 2-Oleodistearin 2-Oleodistearin 2-Oloodistearin 2-Elaidodipalmitint 2-Elaidodipalmitin~ 2-Elaidodistearin 2-Elaidodistearin 2-Elaidodistearin

t

39"1 56"7 68'5 42"1 60"7 50"3 72"5 45"2 64"8 44"7 42-7 51 "2 47"2 44"9

! !

3.92s,

4-23s

3.68m, 3-58m, 3-88m, 3.7m,

3.84m, 3-87vs, 4-14m, 3.8w,

4-04m, 4-58vs, 5-17w, 5-43m 4-42m, 4-74s, 5-21w 4.35s 4-0m, 4-56vs, 5-45m

3.88vs, 4-11m, 3.84m, 4-18s 4-04w,

4-47m, 4.74s, 5-22w 4.32s 4.02m, 4.58vs, 5.44m

4.17s 3.58m, 3.85m, 3.66rr, 3-8m, 3.77s, 4.14s 3.8m, 3.66m,

4.17s, 4.34m

4-17s 3-85m, 4-00w, 4.57s,

5.31m

s = strong; vs ~ very strong; m = moderate; w = weak. * =- and fl"-forms unstable at room temperature. t ~-form unstable at room temperature.

fl-form, which is normally only obtained from solvents, since the change fl' -~ fl is exceedingly slow. This is true also for other members of this group, and for 2-oleodistearin, the change ~ " -> fl' is also slow, and only takes place when the j6"-form is held near its m.p. for a few hours. A heating curve for this compound, therefore, sho~vs only one arrest---at the/~" m . p . M . p , and x-ray data are given in Tables 19 and 20.

40

30

20

IO

o • x

MALKIN AND WILSON "UTTON

Fig. 26. M-elti;~gpoints of 2,oleo-disaturat,,d glycerldes.

In a later paper dealing with 2.oleo-dipahnitin and -distearin, L~TTO~ and JACKSO.~{6s} failed to observe the forms reported b y 3LtLnn; and %ViLSON"and recorded Widely divergent m.p. and x-ray dat~. Since, however, their m.p. 43

The P o l y m o r p h i s m o f Glycerides d a t a for 2 - o l e o d i s t e a r i n a r e g r e a t l y d i f f e r e n t f r o m t h o s e o f DAUBERT a n d CLARKE, )].ALKLN"a n d WILSON, a n d M E , ~ A , w h i c h a r e in a g T e e m e n t , a n d f u r t h e r , since t h e y s h o w no r e l a t i o n s h i p t o t h o s e o f t h e r e l a t e d o l e o d i p a l m i t i n (see F i g . 26), s o m e d o u b t s m a y b e e x p r e s s e d c o n c e r n i n g t h e p u r i t y o f t h e i r compounds. S i m i l a r d o u b t s arise c o n c e r n i n g t h e r e s u l t s r e p o r t e d in a n x - r a y s t u d y for f u r t h e r o l e o - d i s a t u r a t e d g l y c c r i d e s , b y LUTTO,~', te~) m . p . d a t a for w h i c h a r e g i v e n in T a b l e 21. Table 21. 2-Oleo.disaturated glycerides POP

POS

Alpha-2 . Sub-beta prime-2 Beta prime.2 . Beta-3

. 18.1 i 26-5 33-5 38.3

A,p',a-2

SOS

:18;2

.

Beta prime-2 Beta prime-3 Beta-3

Alpha-3 . Beta prime-3 Sub-beta-3 Beta-3

33 38

. 22.4 i 35 36.2 44-3

"

1-Olco.disaluraled glyceridz~s OPS

OSP

oss

Alpha-2 .! 25-3 Sub-betai~ prime-3 37 Beta prime-3 40

Sub-alpha-2 56.3 I~eta prime-2 . 40.2 Beta prinm-3 . 39.8

Alpha-3 i 30.4 Beta prime-3 43-5

oPP

I Alpha.3 . Beta prime-3

18-5 35-2

I t is n o t e a s y t o see w h y t h e m . p . s o f t h e ~ - f o r m s o f t h e i s o m e r i c S O S a n d 0 S S differ b y so m u c h , w h e r e a s t h o s e o f t h e r e l a t e d i s o m e r s P O P a n d O P P a r e t h e s a m e , n o r is i t clear w h y t h e :c-m.p.s o f P O S , O P S , a n d 0 S P also differ Table 22

SSO PPO MMO LLO CCO SOO POO MOO LOO COO CyOO C~O0

26"7

38-5 34-5 25-0 16.0 -I •I

4.o _03-5

•I "1 ii

19.0 13.5 6.5 0.5 --5.6 --10

IV

III

11

22-7 4.8 --2.5

-

-

-

-

- -

-- -

29-8 18-6 10-0

18"5 3"8

8-6

15"5 -- 27"0 -1"5

_0- 5

--

15-0

- - 4 . 2

- -

--

- -

--- 16.5

10.9

--

-- 34-2

--

Co = caproic; Cy = capry'dc; C ~ capric; L ~ laurie, etc. 44

13"2

21"8 -- 22"0 -- 40-5 -- 50.0 - - 56-5 --

1-Mono- and 1,3-Di-glycerides

greatly. As a rule, isomerism of this type results in a difference of only 2 or 3 °. I t is also extremely unlikely t h a t a n y member of this group exists in only two forms, and it is clear t h a t this group should be submitted to a thorough investigation of cooling and heating curves before x-ray d a t a are allocated to the various forms. M.p. d a t a for a, number of triglycerides containing the oleic radical, determined by DAVBERT and C'L~m.~v.teT)are given in Table 22. These authors did not carry out any x-ray investigation, and therefore denoted the different forms by I, II, III, and IV.

1-Mono- and 1,3-di-glycerides 1-Monoglycerides---The upper and lower curves in Fig. 16, p. 30, are practically identical in form with those of 1-monobrassidin and 1-monoerucin, respectively. In the case of the latter, the transitions vitreous -,- ~, ~ --> fl', fl' -+ fl take place at room temperature, and according to the duration of the cooling curve, arrests for e- and fl'-forms m a y or m a y not appear on the heating curves. The slowest transition is the fl'---> fl and an arrest at the fl-m.p, is not observed unless the curves are taken slowly. 1-Monoolein can occasionally be obtained in a form which gives a uniaxial interference figure. 1,3-Diglycerides--Cooling and heating curves are similar to those given in :Fig. 21 for ditaurin except t h a t the arrest at the//-m.p, is slight, or more usually, absent. The transition fl'--> fl is particularly slow and tq-forms are best obtained by very slow cryslallization from solvents. x - r a y and m.p. data are given in Table 23. Glycerides of cl~<~uImoogric and hydr~arpiv, acid. GUPTA and .~IALK1N~71) have recently examined the 1-mono, 1,3-di-, and triglycerides of the above acids and ~nd tlmt t h e y exhibit the same type of polymorphism as the gly(vridcs of straight chain f a t t y acids. Their results are summarized in Table 24.

1-Monoglycerides The first form to separate from the melt is the e-form which changes slowly, at room temperature, into the fl'-form. When the experiment starts with the molten glyceride, these are the only forms observed. The high-melting /3-form is obtained by slow crystallization from solvents, and it is not easy to obtain it entirely free from fl'-form. In contrast to the monoglycerides of straight-chain f a t t y acids, whose long spacings fi)r fl'- and /q-forms are identical, the long spacings of t h e two forms are quite distinct, and c~rrespond to tilts of 31 ° for the fl-fi~rm and 44 ° for the fl'-form. ~.-Forms, befi)re changing t.o Om/J-form, pass quickly into an intermediate phase, the short spacings of which are the same as those given by saturated 1-nmnoglycerides, i.e. a strong line at 4-2 A, associated with a few weaker lines close on either side. 45

•

........

----

12'5 29.5 15 37 -----

25 42 36 62 18 49 41 63.5

62

53 44.5 66.5

32 56 47 68.5

Melting points

25 55 46.5 68.5 52 65.9 28'5 21.5 --2.6 -- 12.3

71

35 58.5 5O

a

49.8 47.8 58.8 39.0 5O.0

49.5 50.3 58.3 58.8 39.7 52.6 46.4 63.1 48.2 52.6 48'4 39'3 45.2 40.3

Long spacings

3.6, 3.66m, 3.57, 3.88s, 3.75vs,

3.59, 3.68, 3.59, 3.75s, 3.85m, 3.75s, 4.67vs 3.9vs,

3.73, 3.87, 3.75,

B'

4.6vs

3.95, 4.04, 4.0,

3.88, 4.05, 3.9,

" BF..~r:J)WT, S[DHU a n d DAVJBERT,G68~see also DA~BERT an~i 8IDHU (6a} for d a t a for 1,3-d|claidin. |~AUI~ERT a l ~ [,UTTON,~te~. Ot her d a t a from CARTEE a n d I~ALKIN. (ie~

/

1-SIonoohqn . . l-Monoelaidin i 1-Monoerucin 1-,Monobrassidin 1,3.Dioloin . 1,3-Dielaidin •I 1,3-Dierucin i/ 1,3-Dibrassidin t l-Stcaryl 3-okdn* .~ l-Stearyl 3.elaidin* i 1-Olcyl 3.elaidin* 1,3-Dioleint . i 1,3-Dilinoloin'[" i 1,3.Dilinolenin'l"

[ I'itre. • [ O[~#t

I

Table 23

4.61s 4.24, 4.68s 3.59s

4.13s 4.3, 4.65s 4.14s

3'58, 3"93s, 3"61, 3'93s, 3'64, 3.7s, 3.64m, 3.72s, 3.78vs, 3.75vs, 3.70s, 3.82, 4.2, 4.4,

3"77, 4.14, 3'75, 4-15, 3.85ms, 3.9s, 3.95ms, 3.9l~, 4.62vs, 3.85vs, 3.81s, 4.73 4.69 4.7

8hor| spacings

4.05, 4"38, 3"9, 4'4, 4.06m, 4.6s 4.05m, 4.6s 4.74s 4.58vs 4.65s

4.28,

4.38,

4.7s

4.32ms, 4.58s 4.55s 4"08, i . ~ 3 m , 4.57s 4.56s 4.23, 4.44m, 4.7s

0

0

1-Monoehauhnoogrint l-Monohydnoearpin . 1 : 3-Di(.hauhnoogrin I : 3-Dihydnocarpin . Trichaulmoogria Trihydnocarpin

27 o 15

Vitreous

53.5 ° 39.5 52 42 35 24

57.5 ° 47 57 47 41.5 31

~lcltlng points

58.5-59 ° 49 59 49 44.5 34 39"2 36'1

43.5 40.0

38'1 35.4 36'7 32.7 39'2 36"0

Long spacings (/k)

3.98s, 4.0s,

4.58vs 4.6vs

4.0m, 4.3m, 4.6s 4.05m, 4.32m, 4.6s

3.8w, 3.8w,

3.9s, 3.9s, 3.1w, 3.4w, 4. ls, 4.1s,

4.Is 4.1s 3.9m, 4.2m, 4.6s 3.7m, 4.Ore, 4.6s 4.6s 4.58s

Short spacings (~ )

Table 24. Melting points and x.ray data for glycerides of chaulmoogric and hydnocarpic acids

o c:

o

C~ ~o ~°

The Polymorphism of Glyceridea 1,3. Diglyc~rides

These differ from most other diglyeerides in crystallizing in large flakes, very similar in appearance to straight-chain f a t t y acids. Under the microscope, between crossed nicols, t h e y exhibit a striking spherulite formation. Three distinct forms can be detected by the capillary melting-point methods, viz. ~, ~', and/5, b u t the changes :¢ --->f~' -~ ~ are so rapid t h a t only the :¢- and the ~-forms are observed on the cooling and heating curves. ~,~,'hcn the molten glyceride is cooled in a capillary, there is solidification at the a-m.p, and, if the temperature is raised when only a small portion has solidified, remclting occurs at the same temperature. I t is not easy, however, to avoid transition to the /5'- and even to the ~-form, and some repetition is usually necessary to observe the m.p.s of the :¢- and the fl'-forms, x-ray data could be determined only for the stable//-form, and ~ho allocation of the :c-structure to the lowest-melting form is based on analogy. From the long spacings, the tilt of thc chains is found to be ~- 50 °. Triglycerides

Starting with the molten glycerides, both capillary m.p.s and cooling and heating curves show the existence of vitreous, ~.- and fl'-forms. The fl-form can be obtained only by slow crystallization from non-polar solvents, or by holding the /~'-form near its m.p. for some hours. Crystallization from ethanol or acetone usually gives the/~'-form. Cooling curves (cooling jacket from 0 ° to room temperature) fall to ~he vitreous m.p. and then rise to the ~-m.p. and, if the heating curve is taken u h e n the curve again begins to fail, there is a single arrest at, the fl'-m.p. Thus, the change 0¢->/5' is moderately rapid. Within experimental error, the long spaeings of the ~'- and the fl-forms are identical, but the two forms are distinguished by their short spacings. Both triglyeerides exhibit typical spherulite formation. I/EFERENCES

m HEI~'TZ, ~V.; Jahrcsber. 2 (1849) 342 12~ DUFFY, P . ; J. Chem. Soc. 5 (1853) 197 [a) BERTItELOT, 5I.; A n n . $8 (1853) 304; 92 (1.q54) 301 m ,'qCHEIJ, L. T.; Rye. lrav. ,5ira. 17 (1898) 200 15~ GUTtI, F.; Z. Biol. 44 (1902) 78 (6i LAu'rz, It.; Ph. Ch. 84 (1913) 631 (71 ,~IA.~KEI,INE, N. S.; J a h r e s b e r . S (1855) 519 {sl I~EIMER, C. L. a n d WILL, W . ; Be,'. 18 (ISSS) 2(113 ~91 OTItMER, P . ; Z. anorg. Chem. 91 (1915) 237

(lo~ IX)SKIT, K.; Z. phys. Chem. 134 (1928) 135 R. B. ltlld ~VAT,CON,1]-. ]~.; J. Italian Inst. Sci. 13 A (1930) 119 (1-*~ ~VEs"GA.~'D,C. and GRUNTZI(~, \~*.; Z. al~org. Chc~L 206 (193~) 301,313 ~:~ 3IALKIN, T.; J. Chem. Soe. (1931) 2796

111) JOGLEKAR,

(1~ CLARKSON, C. E. a n d 3IALKIN, T . ; J . C h ( ' ~ . ,~oc. (1934) 666 [15~ MALKIN, T. a n d 31E~RA, 3I. L . ; J . Ch~'m. S o c . (1939) 103 I1,) ; J . C h e m . Noe. (1()39) l l l l ttT) CARTER, 3I. G. H. a n d ,][ALKIN, W.; J . C])~t)~. ,~oc. (1939) 577 (1~) , J . C h e m . S o c . (1939) 1518

48

References ~19~ GrtL-.x-]7.1o, W . ; Z. anorg. Chem. 240 (1939) 3 1 3 [20~ LUTTO.X', E. S.; J . A n w r . Chem. Sac. 67 (1945) 524 ¢21~ FILER, J r . L. J., SIDrrC, S. S., DAUBERT, B. F. a n d LONGENECKER~ J . E . ; J ' A n w r . Chem. Sac. 68 (1946) 167 ~:~} BAILEY, A, S . , JEFFERSON, .~I. E., KREEGER, F. B. an d BAVFm, S . T . ; Oil and Soap 22 (1945) 10 i23~ CLArtKSON, C. E. a n d 5IALKI~, T . ; J . Chem. Sac. (1948) 985 t ~ VoLxrER, M.; Kinetic d~r Pb~senbildung, S¢~einkopff, Dresden a n d Leipzig, 1939, p. 18~. ~5~ Lu'~'l'o~', E. S.; J. Am~r. Chem. Sac. 68 (1940) 676 (-~6~ _ _ , JACKSOn', F. L. a n d QuIMBY, O. T . , J. A , w r . Chem. Sac. 70 (1948) 2411 (27~ JACKSON, F. L. an d LL-TTON, E. S.; J. A n w r . Chem. Sac. 71 (1949) 1976 (es, LUTTON, E. S.; J. A m e r . Chem. Sac. 70 (19i~) 248 ~29~ BAII~Y, A. E . ; Mclling and Solidification of Fats, Interscience, N e w York, L o n d o n (1950) p. 132 (30) Lr]-roN, E. S.; J. A m c r . Oil Chem. Sac, 27 (1950) 276 ~31) ]tILER, J r. L. J., SIDHU, S. S., CKEN, C, a n d DAUBERT, ~B. ~F.; J . A n w r . Chem. Sac, 67 (1945) 2085 (321 JACKSON, F. L. an d LtTrro.~, E. S.; J . Amer. Chem. Sac. 71 (!949) 1976 caa} TA~XI.XlANN,G.; Z. Phys. Chcm. 25 (1898) 472 (34} I~AVICH, G., ZURINOV, C., ~,'OLNOVA, ~r. a n d ]DETItOV, V.; Acta T'hysicochim. 21 (1940) 101 (see also p. 321) {351 QuI3~BY, OSCAR T . ; J. A m v r . C]~cm. Sac. 72 (1950) 5064 (a6} Cm-;~', C. an d DAUBERT, B. F.; J. Amvr. Chem. Sac. 07 (1945) 1256 137) KIN(;, ~ . 5[. an d GARNER, ~V. E.; J. Chem. Sac. (1936) 1372 ~38~ CHARBON.XET,G. It. a n d SI~'GLETON, ~V. S. ; J. A ,~cr. Oil Chem. Sac. 24 (1917) 140 ~aa} JACKSON, F. L. a n d LU'rTO.~, E. S.; J. Amvr. Chcm. Sac. 72 (1950) 4519 ,0~ -, W'.~LE, R. L. a n d Lv~ro.x', E. ,S.; J. Amvr. Chem. Sac. 7 3 ( 1 9 5 1 ) 4280 {~]} S]DIt~', S. S. an d DAUBERT, B. 1".; J . Amcr. Chem. Sac. 69 (1947) 1451 (~u) FISCHEr, E., BERe;.~IAN'.~, 3I. a n d BXI~WL~D, H . ; Ber. 53 (1920) 1591 [43~ REWAI)IKAR, 1~. S- a n d "~VATSON,I~. E . ; J. I n d i a n Inst. Sci. 13 A (1930) !28 (~a} 5|ALKIN, T, a n d SttURBAGY, M. R. EL; J . Chen~. Sac. (1936) 1628 [4~ LUTTO.~', E. S. an d J-ACKSO:;, F. L.; J . A n w r . Chem. Sac. 70 (1948) 2 i 4 5 (a6) 5[ALKIN, T.; Trans. Faraday Sac. 29 (1933) 977 (~:~ Bt:~:.',.-_~L,J. D. ; Z. krist. 83 (1932) 153 (~81 5IALKIN, T. ; J. Chem. Sac. (1935) 726 (tg~ C~OWE, R. "~V. a n d SSIYTH, C. P . ; J. Amer. Chem. Sac. 72 (1950} 4427 (~"~ ,q~I)~tt;, S. S. and DAUI~EItT, B. F . ; J. Amcr. Chem. Sac. 68 (1946) 1975 (51~ I;AUI~ERT, B. F. a n d CLARKE, T. ]{.; Otl attd Soap 22 (1945) 113 I~-~ MALKLX',T., SIIUJ~BAGY, .M. 11. EL a n d MEARA, 5I. L.; J. Chcm. Sac. (1937) 1409 I~a) BA~:R, F. ,L, JACK~ON, F. L., KOLP, D. G. a n d Lu]-ro~', E. S.; J . Amer. Chem. Sac. 71 (1949)3363 i~) t t o w E , R. J. a n d 5IALKI.~, T . ; J. Chem. Sac. (1951) 2663 (~) S~D~t', S. S. an d DAUBERT, n . F . ; J . Amer. CSom. Sac. 68 (1946) 2603 (~s) CnwrrR ' M. G. R. a n d 5IALKIN, T,; J. Chcm. Sac. (19t7) 551. See also ]~ILDITCH'fl Chemical Co~stitution of Natural Fats, C h a p m a n and lIall, 1910, p. :~58 ~gr~ ~VHEFI~m, D. 11., I{IE.XIENSCHNEIDER, R. ~V. and SAN'DO, C. E.; J . Biol. Chem. 132 (19:10) 6S7 (ss~ DAI;IqERT, I~. F. ~lnd I}ALDXVIN, A. 11.; J. A n w r. Chcm. Sac. 66 (1944) 997 ~59~ I{AVICH, G. a n d Ztyi:L~Ov, C.; Acta Physicochi,u 21 (19t6) 321 ~0~ ]?].:~t(;t'so.~-, R. 1I. a n d LU'rTON, J~. S.; J . . A m c r . Chem. Sac. 69 (1917) 1~45 {61) DAUI~EI{T, B. F. a n d CLAltKE, T. I L ; J. Amer. Cl/em. Sac. 66 (1914)690 (62~ .MEAlt~, M. L.; J. Chem. Sac. (19|5) 22 ~ss~ LUTTON, E. S.; J. A m c r . Chcm. Sac. (iS (1946) 676 49

The P o l y m o r p h l s m of Glycer]des (") ~IALKI~, T. a n d WIL~O.~, B. R . ; J . Ct~cm. Soc. (1949) 369 (6~) Lu2"roN, E. S. a n d JAcxso~, F. L.; J . Amvr. Chem. Soc. 72 (1950) 3254 ("J . . . . ; J . A~nvr. Cl, cr,~. Soc. 73 {1951) 5595 (67) D.AUBERT, B. F. a n d C L t m K E , . T . H . ; Oil a n d S o a p 22 (1945) 113 (6s) BENEDICT, J. I:I., SIDHU, S. S. a n d DAUBERT, B. F . ; J . A ~ w r . Oil Chem. Sac 27 (1950) 91 (69) DAUBERT, B. F. a n d SIDHU, S. S.; J . A , T e r . CJtem. Soc. 70 (1948) 1848 (70) ~ a n d L u ~ o N , E. S.; J . A?nvr. Che~n. Soc. 69 (1947) 1449 (71) GUPTA, A. a n d MALKIN, T.; J . Chem. Soc. (1952) 2405

50