4x30
PAUL J. FLORY AND ROBERT R. GARRETT
Vol.
so
[CONTKIBLTION FROM THE DEPARTMENT O F ClIEMISTRY UF CORSELL USIVERSITY]
Phase Transitions in Collagen and Gelatin Systems1 BY PAUL J. FLORY . ~ N DROBERTK. GARRETT RECEIVED MAKCH 24, 1958 The transformation of native beef .ichilles tendon (BAIT)and r a t tail tendon (RTT) collagcn in admixture with ethylene glycol to the amorphous, shrunken state has been investigated over t h e range of volume fraction a2 of collagen from 0 t o 0.83. Precision dilatometric methods employed for z2 > 0.08 reveal small b u t unmistakable latent volume changes (ca. 5 X cc. per g. of collagen), T h e transformation is partially reversible, the latent volume change being slowly recovered in part upon cooling. Return t o the previous liquidus occurs within 1 2 ' of the original transformation temperature T,n. These observations show conclusively t h a t the transformation is properly represented as a phase transition, involving melting of crystalline regions, rather than as a rate process. Melting temperatures T,n also have been ascertained b y polarizing microscopy and, a t high dilution, from the decrease in viscosity. The dependence of T, on ' ~ ' 2conforms rather well with polymer melting theory, which yields 24 f 5 cal. g,-1 for the heat of fusion of collagen. -4 glass-type transition observed both in native collagen-diluent mixtures below T,,, and in shrunken d r y collagen a t 95' probably is associated with freezing-in of side chain disorder, Melting of gelatin gels resembles the re-melting o f collagen allowed to re-crystallize by cooling after the initial transformation. T h e identity of the processes involved is indicated by the observation t h a t T, for gelatin-glycol is within 2" of t h a t for collageti-glycol of the same composition.
creases the concentration of diluent ( 2 .e., water) iii Introduction The phenomenon of shrinkage of native colla- the fiber in the shrunken state. The hypothesis that thermal shrinkage of collagel1 upon heating while immersed in a swelling agent such as water is well known. Diversity of gen represents fusion of a crystalline state of oropinion has persisted, however, concerning the na- ganization of oriented polypeptide chains was adture of the process responsible for shrinkage. The vanced b y Wohlisch4 and Kiintzel.14 It was subentire shrinkage, amounting to as much as 807, of sequently set forth in a little-known paper by Van the initial length, occurs within a temperature in- Hook15in 1947. U'iederhorn and co-workers'" emterval of only 2 or 3" if the temperature is raised ployed this concept in their interpretation of the slowly. The X-ray diffraction pattern disappears influence of stress on the transformation temperasimultaneously with shrinkage, but may reappear ture of kangaroo tail tendon collagen. X7eirj6on the on lowering the temperature,2 or, preferably, with other hand, has advanced the view that the process drying or re-elongation.3 Absorption of heat dur- is a rate-controlled chemical reaction having an exing shrinkage is revealed by calorinietric ineasure- traordinarily high temperature coeficient. Thies ments.?,5 X small increase in volume during and Steinhardtl? preferred t o consider that rupture shrinkage has been reported by m'eir.6s7 The of inter-chain bonds is responsible for the shrinkage, shrinkage temperature has been observed to in- or "denaturation," a state of equilibrium being attained a t each temperature. crease with the concentration of the p ~ l y m e r , ~quickly ,~ The series of investigations reported in part in with the tensile force on the fiber4,9-12 and with the degree of cross-linking introduced by tanning.I3 this paper was undertaken with the primary object The last mentioned observation probably is con- of pursuing the hypothesis t h a t shrinkage phenomnected with the first inasmuch as cross-linking de- ena in fibrous proteins generally, and t h a t displayed so strikingly by collagen in particular, are manifesf l ) F r o m a thesis submitted t o t h e Graduate School of Cornril tations of the melting of crysta!line regions, and University, 1957, by Robert R. G a r r e t t in partial fulfillment of t h e requirements f o r t h e P h . D . degree. T h e paper was presented before that the process may be appropriately treated as a t h e Division of Polymer Chemistry a t t h e 132nd meeting of t h e Ameriphase transition. It was decided to apply the exSeptemher 10, 1957. can Chemical Society, New York, S . T., perimental and theoretical niethods successfuly T h e work described in this paper was supported in p a r t b y t h e adapted in recent years to the melting of convenAllegany Ballistics Laboratory, Cumberland, M a r y l a n d , a n establishment owned b y t h e United S t a t e s N a v y and operated b y t h e Hertional polymers.'O Specifically, precise dilatometcules Powder Company under Contract h* Ord 10431. ric measurements of voluine changes with teni(2) A. L. Zaides, Kolioid Z h i c ~, 12, 317 (1050). perdture iii the tr,insfor~nationregion were underBear, ".4dvances in Protein Chemistry," V u l . V I I , Acu taken, first on beef Achilles tendon ( R A T ) and suhdemic Press, Inc., h-ew E'ork, N . Y . , 1932, I>. (iil. ( 4 ) E. Wohlisch a n d IIiir i zin :: mi cr . 1)il.rtometry
Dilatometry Ililatonietry Polarizing micr I'ohrizing micr. 1)ilatomrlry \'isromel r y
Fig 6.-Relativc viscosity of collngcn-glycnl iolution, 5 X 10 - 4 n~ a fiinction of tciiiperotiit c
it'?
S
Upon cooling the transformed solution to 2 1 " and holding overnight, a small but definite increase in viscosity was noted. In analogy with the dilatometric experiments, rcvcrsion to the proto5brillar (crystalline) form is suggested. Upon warming, the sample was observed to re-mclt at a tcmpcrature within 5 2 " of the origiiial T,. The observation of slightly lower viscosities above T, after the second melting may br due t o solvolytic tlegradntion. Table I presents a summary oi the melting points for various collagen-glycol mixtures. IVcight fractions have been converted to volume fractions calculated for the amorphous mixture a t 25' using for the densities of glycol and collagen (amorphous) the respective valucs 1.11 and 1.33. Null volume changcs on mixing are implicitly assumed. .\lso, possible consequences of changes in the volume fraction with temperature are disregarded. Errors of interpretation arising from these sources are trivial. The melting temperatures arc plotted against the volume fraction in Fig. 7. Numbers identify the points with entries in Table I. The m i n t s fall smoothlv on the curvc, (calculated, cf. &q ), with the exception of points 5 , 6 and 10. Point 0 is believed to be
equilibration. The latent volume change, though small, is unmistakable. The melting point is only :t degree below that f o u r i d for the I i T T collngcm glycol mixture of the same concentration. Thus, while the degree oi crystallinity is less for the gelatin, the melting point appears to be nearly the saiiic~ as for collagen a t the sainc concentration. Alelting temperatures similarly determinecl for gelatin-water mixtures are given in Table 11. The tlctermination a t the highest temperature is soiiicwhat in doubt owing to evidence of tlecompositioti . Latent volume changes wcre small, and the forms of the volume-temperature curves were suggestivc of those obtained on second melting of the colla-
PHASE TRANSITIONS IN COLLAGEN AND GELATINSYSTEMS
Sept. 20, 1958
4843
T h e observation of a glass type transition in native collagen was unexpected in view of the high degree of crystallinity in this form. While it is possible that the transition is confined to the very small fraction of the material which may be in the amorphous condition, i t appears more likely to us that the side chains of the collagen molecules are responsible for the transition. According to this interpretation, the side chains lack sufficient mobility below T , to achieve a state of equilibrium. The results of an experiment conducted on dried, shrunken BAT collagen are shown in Fig. 10. A sample of tendon similar to that used in the experiment previously tried was dried by heating a t 175" for several minutes. Shrinkage was observed to occur during drying. The data were taken rapidly in order to complete the experiment before decomposition became excessive.
I
I
I
30
40
60
50
T["C).
Fig. 8.-Gelatin-ethylene glycol, z1-2 = 0.10. Sample weight ca. 190 g.; capillary diameter 1.00 mm.
The results shown in Table I1 are in substantial agreement with those of NeimanZ6covering approximately the same composition range. They accord also with melting temperatures obtained by Hirai,30who applied a variety of methods a t compositions ranging up to w2 = 0.75. A Glass-like Transition in Collagen.-Preliminary dilatometric studies on native BAT containing various quantities of water revealed discontinuities in the slope of the dilatometer reading plotted against the temperature. An example of such an experiment is presented in Fig. 9. A two-inch length of tendon was placed in a dilatometer after washing the sample with distilled water. Observed volumes were constant with time over the range investigated, and, as indicated in Fig. 9, the volume-temperature rerationship was accurately reproducible in succeeding cycles. 20 -
'I
Initial
e
2
A
Heating
' 5
4
Cooling Second Healing
/*'
D
15-
/"
d
.. -
/.
E
dy
a
& ,W ,
lo
,* /= I
30
35
40
,
45
I
50
J
55
I
90
95
I
100
T('C). Fig. 10.-Evidence for a glass-like transition in air dried, shrunken BAT in absence of diluent.
An attempt to conduct a similar experiment on the tendon prior to shrinkage failed owing to instability of the volume a t each temperature. The slow disappearance of voids is presumed to be responsible for this difficulty. The observed glass transition temperature in the neighborhood of 95' for dry, shrunken collagen agrees within experimental error with that reported recently by Hirai30for dry gelatin. Discussion The process of melting of collagen-diluent mixtures presents a complicating feature not ordinarily encountered in the melting of other polymers. This complication arises from the fact t h a t the crystalline phase contains diluent. We do not refer here to diluent present in the interfibrillar cavities, which may indeed serve as repositories of diluent (reluctantly, perhaps, as previously noted), but to the interprotofibrillar spaces within the crystalline phase itself. Absorption of solvent by the ordered phase in this manner is evidenced by the increase in equatorial X-ray spacings with solvation (in water).3@ I n prescribing the thermodynamic con(40)
A. L. Zaides, Rolloid Zhur., 16, 265 (1954).
4844
l7Ol. so
P L i uJ.~ FLORY ,IND ROBERTI
