39 HyaluronicALURONIC acid water binding
Studies of high concentration HA solutions using DSC reached firm and consistent
conclusions about the amount of water strongly bound by the polysaccharide (i.e., non-
freezing water). Joshi found that up to 21 molecules of water were strongly bound to each
hyaluronic acid disaccharide monomer (i.e., repeat unit), and that the amount bound was
0.6 g H
2 O/g hyaluronic acid.13 Molecular dynamics simulations agree with the number of
water molecules per disaccharide monomer suggested by Joshi.14 Ikada found 0.51 g non-
freezing H
2 O/g hyaluronic acid, which corresponds to 11.5 molecules of water per repeating
HA disaccharide.15 He also made one of the few measurements of freezing bound water,
which he found to be 0.59 g H
2 O/g hyaluronic acid. Yoshida determined a value of 0.5 g
H
2 O/g hyaluronic acid for strongly bound water, which corresponds to 11 molecules per
repeat unit, and also measured the freezing bound water, which he found to be 0.5-1.8 g
H
2 O/g hyaluronic acid.16 Kučerík determined there are 17.2–19.1 water molecules bound
per polymer repeat unit, and that the amount bound was 0.77–0.86 g H
2 O/g hyaluronic
acid.17,18 One interesting feature of this earlier report is that water binding was studied as
a function of the HA molecular weight, and it was found to be approximately the same at
most molecular weights.
Multiple DSC studies of HA hydration used high concentrations which were needed to see
thermal transitions associated with bound water. A single study by Cowman used solutions
of HA as low as 0.5% concentration,19 approaching closest to the situation studied herein.
Cowman found an anomalously large amount of water associated with HA, calculated from
conventional analysis of the thermal analytical data. She explained this finding by the idea
that freezing bound water and non-freezing bound water are coincident at low concentrations.
Assuming based on past work 0.6 g H
2 O/g HA for non-freezing bound water, the data led
her to conclude contributions for freezing bound water are 44 g H
2 O/g HA.
Another, early method used to interrogate the amount of water bound by HA used ultrasound
to measure the speed of sound in HA solutions to determine their compressibility. Davies
found that in aqueous solution at 25°C, HA is “closely associated with not less than 9
molecules of water of hydration.”20 The mass of that number of water molecules would
be 144 g/mol, which compares to the 403 g/mol molecular weight of the hyaluronic acid
disaccharide. That corresponds to 0.36 g H
2 O/g hyaluronic acid. Jouon found that the
“amount of non-freezable water associated with the polysaccharide was about 0.7 g H
2 O/g
dry solids which corresponded to 15 water molecules per disaccharide unit.”21
While the foregoing reports used different methods to measure the amount of water
strongly bound to hyaluronic acid, which produced somewhat different results, they are all
similarly in the range of 0.36–0.86 g H
2 O/g hyaluronic acid. These results are all obviously
quite far from the 1,000 g per g of HA in the commonly made claim.
OH
HO OH
O
O
HO
OH
NaO2C O
HO
O
NH
OH
O
molecular weight
=403 amu
molecular weight
=92 amu
Chart 1. Chemical structures of the hyaluronic acid repeating disaccharide salt and glycerol. Hydrogen
bonding to water is possible for each N and O atom. Solvation by water is possible for the Na ion.

40 JOURNAL OF COSMETIC SCIENCE
It is appropriate to consider the genesis of the erroneous claim of water binding by HA in the
Sutherland review earlier discussed. Key early studies of HA were conducted by Ogston.22–25
His work was not primarily aimed at determining the water-binding ability of HA, but rather
at its particle size and structure. It used several measurements (centrifugation, viscosity) to
measure particle size, finding that HA, despite being a linear polymer that adopts a coiled
form in solution, behaves hydrodynamically like a large solvated sphere that encompasses a
far larger volume than the molecular volume of the polymer. Ogston states “solvent entrained
within it will be largely carried with the particle in dynamic measurements” and notes the
dynamic entrainment of water in the interstices of a random-coil particle is what makes it
behave like a solid spheroid.26 Such entrained (definition: transport of a fluid by shear-induced
turbulent flux) water molecules would not be considered bound in any chemical sense to
the polymer. They move with the particle of HA and are neither bound to it by chemical
forces nor trapped within it. Ogston amplified this point in later work,27 stating “the coiled
particle carries a great deal of water with it when it moves, through the operation of purely
viscous forces.” He finds the total volume occupied by the solvated coil is 6,000 mL/g HA,
which agrees with the dimensions of the particle measured by light scattering, and notes the
coil is permeable to solvent. It seems quite likely that this particular report forms the basis
of the statement by Sutherland that HA binds 6,000 times its weight in water, given water’s
density of 1 g/mL. It also corresponds to the 6 L per gram of HA that was earlier mentioned.
Ogston also states that HA serves to immobilize water in connective tissues, and it “might
be said, in a dynamic sense, to ‘bind’ it.” It is important to realize he is not referring to a
chemical binding, under a thermodynamic definition, but merely to restrict its movement
within tissue. Hence, this is the basis for him to place the bind in his statement within
quotations, since he is not referring to binding in any classical chemical sense.
EXPERIMENTAL DESIGN
This work examines the binding of water by HA using two experimental methods that
have earlier been widely applied to the investigation of the physical chemistry of solutions of
carbohydrate polymers. To address the specific claim that HA has exceptional properties as a
humectant because of its ability to bind a 1,000 times its weight in water, we revisited past
thermal analysis studies of hyaluronic acid solutions conducted at the supposedly special
ratio of 1,000 g of water per gram of HA. HA water binding at higher concentrations (using
a variety of modalities, including thermal analysis and colligative properties, reviewed above)
has been extensively reported so it was unnecessary to study those higher concentrations
or the concentration dependence, though that has been done with other spectroscopic
methods.28 Fallacies might also persist among practitioners that HA would form a unique
material (such as a solid complex or a gel) at the supposedly special 1,000:1 ratio, which can
be dispelled by simple observation of mixtures at this concentration.
The current work experimentally investigated aqueous solutions of HA at the key
concentration of interest, as well as a 10% aqueous solution of a more mundane humectant,
glycerol. These solutions have comparable viscosity.
MATERIAL AND METHODS
A commercial sample of fermentation-derived, cosmetic-grade HA (specifically, sodium
hyaluronate) from Shandong AWA Biopharm (MW 0.92 mDa) (Binzhou City, China) was