Chapter 6

Structural Studies of Agar-Gelatin Complex

Coacervates by Small Angle Neutron Scattering,
Rh.eology and Differential Scanning Calorimetry

6.1 Introduction
Novel and smart biomateriali; can be made through inter and intra-molecular
complexation of bio-polyelectrolytes. Coacervates are a new class of polymeric
materials formed via electrostatic interactions between complementary_polyelectrolytes
or a polyelectrolyte and colloid pair [1-6]. Such materials have found applications in
protein purification [7], drug and enzyme immobilization [8] and in trapping organic
plumes [9]. A variety of coacervates have been made in the past by suitably tuning and
enhancing the electrostatic interactions between pairs of polyions. In these transitions,
the screened electrostatic interactions existing between oppositely charged
polyelectrolytes yield soluble intermolecular complexes that precede coacervation.
During such a process, the homogeneous solution undergoes a liquid-liquid phase
separation with the polymer-rich phase compartmentalizing itself into a dense state
withdrawn from the bulk of the solvent, called the supernatant that co-exists
thermodynamically with the coacervate. The solvent present in the concentrated phase
largely constitutes the solvation liquid. The supernatant is a very dilute polyelectrolyte
solution.

This phenomenon has been conceptualized in various models. In Veis model [10] the
aggregates are referred to as symmetrical aggregate polymer (SAP) whereas Tainaka's
revised model [11] refers to these as asymmetrically aggregated polymers (AAP). In
Nakajima-Sato model [12] the solute-solvent interactions was incorporated in to the
earlier proposed Overbeek-Voorn model [13] through Flory-Huggins prescriptions.
Regardless, it is accepted that the coacervate phase owes its origin to the formation of
intermolecular soluble aggregates. Tainaka model [11] assumes Gaussian distribution

112
of segments in AAP aggregates independent of their size and that all the counter-ions
are bound' to the AAP aggregates. It should be realized that when two oppositely
charged segments join together, some amount of counter ion is always released in to the
solvent, there by increasing the entropy of the solution. This will assist the process to
move towards coacervation. A more rigorous model describing coacervation [14] and
its conceptualization in the free-energy landscape formalism was proposed recently
[15].

In chapter 4, we have undertaken a qualitative, and yet a systematic study on

intermolecular complexation and phase separation in aqueous solutions of gelatin (a
low charge density polyampholyte molecule) interacting with agar, a polysaccharide,
under various thermodynamic conditions of the solvent with an objective to understand
the kinetics of complex coacervation [16]. Agar, a poly-anion was shown to undergo
liquid-liquid phase separation (complex coacervation) in presence of a polyampholyte,
gelatin, both for pH<pI and pH >pI of gelatin. Agar sols form thermo-reversible
physical gels with large hysteresis between melting and gelling temperatures with the
constituent unit being anti-symmetric double helices [17,18]. Gelatin, a polyampholyte
obtained from denatured collagen, is a polypeptide. The gelation kinetics of gelatin has
been well studied and characterized in the past [19,20]. The length scale hierarchy that
exists in gelatin sol, gel and coacervate phases has been extensively disc~ssed in an
earlier work [21]. The mapping of microscopic structure of coacervates requires further
probing in order to understand the spatial ordering of polyelectrolytes inside the
coacervate phase. In the past very little attention has been paid to this aspect. In this
chapter, we examine the physical characteristics of agar-gelatin coacervates by
physically probing these samples through an array of experimental techniques like:
Small Angle Neutron Scattering (SANS), rheology and Differential Scanning
Calorimetry (DSC).

Thermal experiments were carried out using a Labsys TG-DSCI6 differential scanning
calorimeter instrument manufactured by Setaram, France. Samples (typically =50 mg)
were heated in aluminum cups sealed with lids at a controlled rate of 1°C/min up to a
maximum temperature of 100°C starting from room temperature (20°C).

113
6.2 Results and Discussions

6.2.1 Small Angle Neutron Scattering data (SANS)

The differential scattering cross section arising from a collection of scattering
....
~-,
particles
consists of two terms [22-25]. The first term depends on the intra-particle scattering
which in tum depends on the shape, and size of the particle and in principle can be
calculated for any geometry. The second term depends on the inter-particle scattering
that is governed by the inter-particle interactions prevailing inside the system.
Incoherent background correction in SANS data normally poses a serious challenge.
Samples prepared in deuterated water helps in minimizing this effect.
The present objective was to study the microscopic structure of the agar-gelatin
coacervates. The expression for scattering cross-section can be simplified for the
experimental situation of the coacervate that permits the SANS data to be analyzed in
random phase approximation, in two distinct ranges of scattering vectors (q): low q-
range (Debye-Bueche behaviour [26]) and the high-q range (Ornstein-Zernike
behaviour [27]). Analysis of SANS data within the model of mean field theory reveals
that for the polymers in a good solvent, at equilibrium, the structure factor of
concentration fluctuations in thee Ornstein-Zernike (O-Z) region is given by

(6.1)

Where, ~ defines is the correlation length of the concentration fluctuations or mesh size
of the network and Ir (0) is related to the cross-link density and longitudinal osmotic
modulus. Experiments carried out in the semi dilute regime of polymer solutions have
shown deviations from the Ornstein-Zernike function.

Long wavelength concentration fluctuations in these systems often gives rise to excess
scattering in the low-q region of the SANS data. The origin of these fluctuations
remains debatable, and a clear cause that generates such excess scattering is yet to be
established. However, Koberstein et al. [28] have suggested long-range random
inhomogeneities with correlation length many times larger than the radius of gyration

114
of the dissolved polymer to cause this excess scattering. If the spatial scale of density
fluctuations due to the presence of inhomogeneities is large compared to the correlation
length ~, then the two contributions can be treated separately and added to give the total
structure factor as

Seq) = SL(q) + Sex(q) . (6.2)

Where SL(q) is the Ornstein-Zernike (O-Z) function, and the Debye-Bueche (D-B)

structure factor has the form Sex ( q) given by

S ( ) - lexCO)
ex q - Cl+q2f)2
(6.3)

where IexCO) is the extrapolated structure factor at zero wave vector. It is possible to
Study low-q domain of the structure factor provided a high instrumental resolution

SANS spectrometers is used to collect data. We have examined neutron scattering data

obtained from agar-gelatin coacervate (20°C) as function of salt concentration. The

experimental data are presented in Figures 6.1 and 6.2 reveal structure factor profile for
all the samples. Let us recall that coacervates are amorphous substances, and devoid of
any spatial ordering (coacervation transition is a first-order phase transition).This would
imply identical scattering profiles for all the samples, which was observed (Figure 6.2)
if and only if the inter-molecular interactions are so strong that mobile ion induced
Debye-Huckel screening is not too effective. Since, I(q) depends on the square of the
difference of neutron scattering-length densities of the scatterer and the solvent, it can
not distinguish between microscopic structures of two samples with identical
concentrations, particularly, when both are amorphous materials.

1 15
 10
0 0.01 M
0 0.02M
b. 0.05M
<> 0.1 M

,.....,
m0.1
0-

0.01

0.001
0.01 0.1

q I(A Or 1
Figure 6.1: Plot of static structure factor, S( q) versus scattering wave' vector, q for
agar-gelatin coacervates prepared in D20 detennined from SANS measurements
perfonned at 20°e. Notice the invariance of scattering profiles with ionic strength of
the coacervate samples.

A least-squares fit of the structure factor data in the low-q range, 0.018 kl nm- I :s q :s
0.072 kl (see Figure 6.2), to the D-B function yields size of the heterogeneities
(s=220± 20 A) present inside the coacervate phase. Secondly, the measured correlation
length (Fig. 2) detennined by fitting high-q region data to O-Z function gave, S =12±2
A much smaller than the persistence length of gelatin, which is ~ 25 A-I. For gelatin'
coacervates, gels and sols the correlation length values reported [21] are 12 A, 26 A and
26 A respectively (see Table-I)_ Size of heterogeneities reported for such systems were
~ 200 A which is comparable to the size of the heterogeneities, s=220 ± 30 A reported
now. Thus, it can be inferred that the simple and complex coacervates of gelatin bear
identical structural signatures. On the other hand, agar gels are associated with
contrastingly different length scales; the correlation length and heterogeneity size are
59 A and 700 A respectively [29]. These values remained unchanged as the

116
concentration of CaCh added to these gels was varied from 0.01 to 1M. The agar-
gelatin samples did not exhibit any such dependence either.

,'~- .

o-z fitting: ~ =16 A 0

2
\

o O.01M
O.02M
O.05M
O.1M
-1

0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14

q2/(AO r1
Figure 6.2: Plot of reciprocal of static structure factor, l/S(q) versus scattering wave
vector, q for agar-gelatin coacervates prepared in D20 determined from SANS
measurements; data is from Figure 1. Low-q data fitted to Debye-Bueche and high-q
fitted to Ornstein-Zernike function. This defines a clear q-cut off.

6.2.2 Rheology data

Isochronal temperature sweep (at the rate 0.3 0C/min.) measurements were undertaken
on various samples to measure their melting behaviour. The first derivative (dG'/dT) of

storage modulus (G') with temperature (T) gave unambiguous values for two melting
transition temperatures and the data are shown in Figure 6.3 for various salt
concentrations. Let us compare these temperatures with the thermal properties of agar
[30] and gelatin gels [31]. Agar has a gelation temperature ~ 40°C and melting
temperature ~ 85°C. On the other hand, gelatin gels at ~ 30°C. The data shown in
Figure 6.3, thus, represents a linear combination of the thermal behaviour of agar and

117
gelatin gels. One can also argue that such behaviour owes it origin to the presence of
gel-like (physical) network structures inside the coacervate phase. Agar and gelatin gels
are known to comprise of hydrogen bond stabilized double [32] and triple helix [33]
units that constitute the three dimensional interconnected network. The gelatin gels are
rubbery and since, these molecules are rich in glycine, proline and hydroxyproline
residues, inside the triple helix, the individual molecules feature in poly(L-proline II)
(trans) helix conformation in contrast, agar gels are composed of micro-domains of
polymer-rich spinodal phases that imparts it a non-rubbery structure [18]. The shear
induced flow behaviour of coacervate samples were quite revealing. The shear rate
dependent viscosity data of these samples is plotted in Figure 6.4 which reveals the
non-Newtonian yield.

0.25 r--------~====:__---_, 1.8

-0-- O.01M
~ 0.01 & 0.02M 1.6
-[}- O.05M
0.20 -b-- O.02M 1.4
-<>- O.1M
1.2 ...-..
("');;-0.15 ("')
0
1.0
--
~ ~

C X
I- 0.10 0.8 I-
l! "'0
:::--.
(!) 0.6 (!)
"'0 0.05 "'0
0.4
0.2
0.00
0.0
>
10 20 30 40 50 60 70 80 90
TO/C
Figure 6.3: Plot of the first derivative of isochronal storage modulus in a temperature
sweep experiment. Notice the near invariance of transition temperatures with ionic
strength of the samples. It appears that the strong associative interactions present inside
the dense medium are not effected by screening.

I 18
 3r-----------------------~~~~~,
o O.01M
k o O.02M
11=110 (dy/dtr
Ll O.05M
2 <) O.1M
k=1.2
- - fitting

-1

-2 -1 o 1 2

(dy/dt)/s-1

Figure 6.4: Plot of viscosity versus shear rate for samples prepared with different ionic
strengths. The Carreau model provided excellent fitting to the data within experimental
error. For clarity fitting is shown for one sample only having salt concentration
=0.05M. The exponent, k changed by less than 10% as the salt concentration increased
from 0.01 to O.1M.

In addition shear thinning behaviour was exhibited by these samples. The data
presented in Figure 6.4 were fitted to Carreau model expression [34]

(6.4)
with k = 1.2±0.2 independent of ionic strength. In fact, k reveals the viscous response
of the samples to applied shear: k=O gives Newtonian, k<O indicates shear-thickening
and k>O implies shear-thinning behaviour. Thus, shear-thinning features are clearly
manifested in these samples. It is interesting to note that alike the SANS results, the
rheology data did not exhibit any ionic strength dependence. Alcohol induced simple
coacervates of gelatin formed due to self-charge neutralization has been shown to
exhibit non-Newtonian behaviour in the past [21). Thus, the agar-gelatin coacervate has
rheological features not very different from that of gelatin coacervate indicating that the

119
Carreau model can be universally applied to describe such material. In contrast, the gels
of agar and gelatin have been shown to exhibit viscoelastic response (see Table-I).

6.2.3 Differential Scanning Calorimetry data (DSC)

The DSC data pertaining to the coacervate samples prepared with different salt
concentrations is displayed in Figure 6.5. These indotherms depict near invariance of
data with sample ionic strength. The weak melting behavior observed at T~ 30°C could
be ascribed to melting of the network structures present inside the coacervate phase.
The rheology data supported the existence of such network structures which did show a
melting temperature around ~ 32°C. Thus, the indotherms seen in Figure 6.5 can be
identified as that corresponding to the gelation temperature of gelatin network. The area
under the endotherm quantifies the enthalpy of the transition. A closer look at Figure
6.5 reveals that this enthalpy is invariant of ionic strength of the medium. A
comparative data is presented in Table-I that distinguishes various phases of agar and
gelatin. It is quite interesting to note that the endotherm observed at T ~ 30°C has direct
correspondence to a similar melting temperature observed in gelatin gel and coacervate
systems [21,31]. This can be safely attributed to the gelation temperature of gelatin.
In another study an exotherm was observed at the same temperature for agar gel
samples that was attributed to local crystallization of helices in agar gels [35].
However, the agar gels, used by us, had two characteristic transition temperatures [30]:
gelation temperature ~ 40°C and melting temperature ~ 80°C. The agar-gelatin
coacervate samples did not yield signature of such an exotherm. Thus, it appears that
the thermal properties of this coacervate were completely governed by the same of
gelatin regardless of the presence of agar that was expected to be strongly interacting
with this polypeptide. Another anomaly pertains to the observation of invariance of the
transition temperature with ionic strength of the coacervate. Gelation temperature of
gelatin gels is strongly influenced by the nature and concentration of the salt present
[36].

Table-I: Summary of physical characteristics describing various phases of agar and

gelatin. Data reported pertain to 20°C for samples prepared in aqueous medium (D20

120
for SANS experiments). The values listed are representative. The transition
temperatures are associated with an uncertainty of ±30C and the same with
characteristic sizes is ±15% of the listed data, typically.

Sl. ParameterlProperty Agar Gelatin Gelatin Agar-

No. Gel Gel Coacervate3 gelatin
Coacervate
1. Mesh size 59Ac 26A3 12A 12A

2. Size of 700Ac 200A3 200A 220A

heterogeneity

3. First Transition 41°C d 28° Cb 33° C 30°C

Temp. (30°C)g
(DSC)
4. Second Transition 75 uCd - -
75 u C
Temp.
(DSC)
5. First Transition 33 u C 28° Ct 34u C 33° C
Temp.
(Rheology)
6. Second Transition -- -- -- 80u C
Temp.
(Rheology)
7. Viscoelastic feature Visco- Visco- Viscous Viscous
elastic elastic 11(Y·) - (y·r 11 (y.) -
k k
(y·r
k= 1.4 k= 1.2

3: ref[21], b: ref[31], C: ref [29], d: ref [30] and e: ref[41], f: ref [36], g: ref [35]

121
 2

~ -2
<-
3:
0
LL -4
.....
co
Q)
I -6

-8

-10
0 20 40 60 80 100 120
TPC
Figure 6.5: DSC thennograms for agar-gelatin coacervates prepared in water. The rate
of heating was 1DC Imin. The initial concentration of each of these biopolymers was
0.1 % w/v. Notice absence of any endothenn at the higher end of temperature. A single
endothenn was located close to 30DC invariant of ionic strength of the coacervate
samples.

6.3 Conclusions
Agar-gelatin coacervate samples were probed by three techniques in order to detennine
the micro-structure of this phase. The experimental results reveal that the coacervate
phase is a heterogeneous viscous material. The polymer-rich phase comprises
physically crosslinked networks of agar and gelatin molecules. There is finite
possibility of existence of asymmetric double helices of agar coexisting with triple
helices of gelatin. The presence of inter-penetrating networks of these biopolymers can
not be claimed with certainty at this stage. However, it will be appropriate to make a
physical comparison between agar-gelatin complex coacervate with agar, and gelatin
gels, and gelatin coacervates.Such a comparison has been provided in Table-I. The
agar-gelatin coacervate has been found to retain the thennal properties of its

122
constituents to a remarkable extent. This indicates the presence of physica11y entangled
networks of agar and gelatin in the coacervate phase. The viscoelastic response of this
phase to external stress is dependent on the specific nature and density of these
cross links. Thus, detennination of the degree of helicity present in the coacervate phase
wi11 be of significant importance. Coacervates, like gels, are biphasic in nature
comprising the solvated polymer and solvent in various structural fonns. Thus, the
exact detennination of the amount of interstitial and free water present in this material
is required to be evaluated which wi11 provide a better understanding of the thennal and
temporal stability of this material.

It is interesting to observe that gelatin coacervates (simple) and agar-gelatin

coacervates (complex) share a generality as far as their microscopic structures are
concerned, which is clearly evident from the -data shown in Table-I. Though the
characteristic temperatures differ by a few degrees~ the rest of the physical signatures
are identical. Thus it will be appropriate to argue that the microscopic structure of the
coacervate material comprised of crosslinked polymer-rich zones separated by
polymer-poor regions having characteristic viscoelastic length. Such systems are
associated with two characteristic relaxation processes [37, 38]: one due to
concentration fluctuation and another arising from viscoelastic relaxation. This has
been adequately described by models and supported by experiments in the past. The
model involves dynamic coupling between stress and diffusion in a complex and
rigorous way [39,40]. Thus, the coacervate phase is in a dynamica11y evolving state that
makes this system extremely interesting [41]. The results presented provide a
significant insight into the distinctive microscopic features of this coacervate vis-a-vis
gelatin and agar gels, and gelatin coacervates. It does not answer a11 the questions
related to the structure of coacervates, yet it makes an attempt to give some foundation
to its understanding.

123
6.4 References
[1] H.G. Bungenberg de Jong; In Colloid Science; Vol-IT; Ed.; Kruyt, H. R.,Elsevier,
New York, 1949.
[2] K.Kaibara, T.Okazaki, H.B. Bohidar and P.L.Dubin; Biomacromolecules, 1,100,
2000.
[3] H.B. Bohidar, P.L. Dubin, P. Majhi, C.Tribet and W. Jaeger; Biomacromolecules, 6,
1573,2005.
[4] B. Mohanty and H.B. Bohidar; Biomacromolecules, 4, 1080,2003.
[5] E. Kokufuta; Prog. Polym. Sci, 17, 647, 1992.
[6] A Eisenberg and F.E. Bailey; Coulombic Interactions in Macromolecular Systems,
Ed.; ACS Vol. 302, 1986.
[7] H.L Hinze and D.W.Armstrong; Ordered Media in Chemical Separation, ACS
Vol. 342, 1987.
[8] J. Xia, K.W. Mattison, V. Romano, P.,L. Dubin and B. Muhoberac; Biopolymers,
41,359,1997.
[9] Y.Wang, J. Banziger,G. Filipelli and P.L. Dubin; Environ. Sci. Tech., 35,
2608,2001.
[10] (a) A Veis and J. Cohen; J. Phys. Chem.,78, 6238,1956 (b) AVeis.; J. Phys.
Chem.,65, 1960,1963 (c) A Veis; J. Phys. Chem., 61, 1798,1961 and (d) A. Veis
and C. Aranyi; J. Phys. Chem., 64, 1203,1960.
[11] K.Tainaka; Biopolymers, 19, 1289,1980; J. Physics. Soc. Japan, 46,1899,1979.
[12] ANakajima and H. Sato; Biopolymers, 10, 1345,1972.
[13] J. Overbeek and MJ. Voom; J. Cell. Compo Physiol., 49, supp-l, 7,1957.
[14] AGupta and H.B. Bohidar; Phys. Rev. E.,72, 011507,2005.
[15J AGupta, Reena and H.B. Bohidar; J. Chem. Phys.,125, 054904,2006.
[16] S.S. Singh and H.B. Bohidar; Int. J. BioI. Macromolecules, 41,185,2007.
[17J J.S. Craigie, C. Leigh; in Hellebust and J.S. Craigie, editors Hand Book of
Phycological Methods, Cambridge Univ. Press; Cambridge; 109-131,1978.
[18J E.Pines and W. Prins; Macromolecules, 6,888,1972.

124
[19] J.P.Busnel, E.R. Morris and S.B. Ross-Murphy; Int. J. BioI. Macromol.,
11,119,1989 and J.P.Busnel and S.B. Ross-Murphy; Int. J. BioI. Macromol.,
10,121,1988.
[20] A. Veis; The Macromolecular Chemistry of Gelatin, Academic Press, New York,
1964.
[21] B. Mohanty, V.K. Aswal, 1. Kohlbrecher and H.B. Bohidar; J Plym. Sci: Part B,
44, 1653,2006.
[22] P.S. Goyal, V.K. Aswal and J.V. Joshi; Current Sci., 79, 947,2000.
[23] P.Thiyagaranjan, J.E. Epperson, R.K. Crawford, J.M. Carpenter, T.E. Klippert and
D.G. Wozniak; J Appl. Crstyllogr.,30, 280,1997.
[24] G.L.Squires; Thermal Neutron Scattering, Cambridge University Press:
Cambridge, 1987.
[25] L.A.Feigin and D.I.Svergun; Structure Analysis by Small-Angle Neutron X-Ray
Scattering and Neutron Scattering; Plenum Press: New York, 1987.
[26] P. Debye and A.M. BuchE; J Appl. Phys., 20, 518,1949.
[27] P.G. de Gennes; Scaling Concepts in Polymer Physics, 2nd Ed.; Cornell University
Press: Ithaca, New York, 1985.
[28J1.T. Koberstein, C. Picot and H. Benoit; Polymer,26,673,1985.
[29] Nicolas, Fatin-Rouge, KJ.Wilkinson and J. Buffle; J Phys. Chem. B,110,20133,
2006.
[30] K. Prasad, A.K. Siddhanta, A. K. Rakshit, A. Bhattacharya and P.K. Ghosh; In(l. J
Bioi. Macromolecules, 35,135,2005.
[31] H.B.Bohidar and S.S. Jena; J Chem. Phys., 98, 8970,1993.
[32] M. Djabourov, A.H. Clark, D.W. Rowlands and S.B. Ross-Murphy;
Macromolecules, 22, 180,1989 ..
[33] M. Djabourov, J. Leblond and P. Papon; J Phys. France,49,319,1988 and ibid
49,433.
[34] H.A. Barnes; A handbook of Elementary Rheology, University of Wales Press,
Wales, England, 2000.
[35] M. Watase, A.H. Nishinari, A.K. Clark and S.B. Ross-Murphy; Macromolecules,
22,1196,1989.

125
[36] S. Chattetjee and H.B. Bohidar; Int. 1. Bioi. Macromolecules, 35,81,2005.
[37] N.Toyoda, M. Takenaka, S. Saito and T. Hashimoto; Polymer, 42, 9193,2001.
[38] T.J. Hashimoto; Polymer Sci: Part B: Polymer Phys., 42,3027,2004.
[39] A. Onuki and T. Taniguchi; J Chem Phys., 106,5761,1997. ,<:,

[40] M.Doi and A.Onuki; J. Phys. II (France), 2,1631,1992.

[41] K.C.Labropou1os, D.E. Niesz, S.c. Danforth and P.G. Keverkidis; Carbohydrate
Polymers, 50,393,2002 and ibid 50 ,407.

126