Proc. Nat. Acad. Sci.

USA
Vol. 68, No. 6, pp. 1093-1094, June 1971

Experiments on Osmosis with Magnetic Fluid

P. F. SCHOLANDER AND M. PEREZ

ABSTRACT Experiments on a ferromagnetic colloidal cover, the surface deforms and papillae rise toward the magnet
fluid at equilibrium showed equality between magnetic (3). The film maintains a flat surface without imposing a
and osmotic force; this result identifies solute pressure
against the free surface as the cause of the negative solvent perpendicular pressure vector.
pressure. Except for water of hydration, there is no other PROCEDURE
osmotic interaction between solute and solvent.
The magnet, mounted above the solution, is elevated until
It has been argued repeatedly and independently that the the pointer shows no lift; this is taken as the null point on the
impact of solute molecules on the free or unrestrained surface scale. One of the weights is hung under the cup, and the re-
of a solution gives negative hydrostatic pressure to the water sulting depression of the spring is compensated for by lowering
and causes osmotic flux (1). There are several lines of evidence the magnet until the pointer is back to the null mark (Fig. 1).
for this view, such as the lowering of the vapor pressure of a The vertical vector of the magnetic force will now exactly
liquid by solutes, the equality between hydrodynamic and balance a 5-mm decrease in the manometer pressure, i.e.,
osmotic flux, and the fact that this mechanism accounts for provided the entire osmotic effect were due to surface cou-
the negative pressure of the water in matrices and suspensions pling. The procedure is repeated with one weight added at a
(2). According to this theory, which invokes forces acting on time.
a free surface, it can be predicted that any force that changes With the magnet underneath the cup, three of the weights
the dispersal pressure of the solute would affect the osmotic are hung on the sides of the cup and the null line is determined
pressure. Besides the well-known effect of temperature, one with the magnet removed. One at a time, the weights are
would expect osmotic effects from acceleration, electrostatic, removed and the magnet is elevated until the pointer is back
or magnetic forces. We report here on experiments with a to the null mark; the balancing pressure is read on the manom-
self-suspending colloidal ferromagnetic solution obtained from eter.
the Ferrofluidics Corp. (3).
RESULTS
METHODS (FIG. 1) In experiments with the magnet above the cup (Fig. 2), the
A dialyzing membrane is mounted over a lens paper sup- change in weight by the magnetic force is exactly matched
ported by a rigid rubber disk that is bulged so as to stretch the by the manometer pressure, both read as mm of water. At
membrane (4). The rigidity of this mounting permits mea- zero magnetic force, we have the osmotic pressure of the solu-
surements of balancing pressures within +1 mm of H20. This tion, which was 11 mm of water. With the magnet below the
cup-shaped unit is fastened horizontally to one end of an cup, a pull of 6 mm or more entirely abolishes the magnetic
aluminum spring, the other end being clamped in a vise.
A fine plastic tubing connects the osmometer with a reading
capillary that is kept exactly level with the fluid surface.
The other end of the capillary connects to an adjustable 8
water manometer (5). In the center, below the cup, is a hook
on which weights may be hanged. These weights are made from 3
a roll of soldering tin and cut to weigh exactly as much as a
5-mm depth of water does in the cup. To the spring is fastened
a fine wire pointer, which extends to a movable null scale.
The entire assembly is scrupulously nonmagnetic.
A strong permanent magnet (Westinghouse 6440048) is
mounted on a lab jack so that it can be raised and lowered
with the poles facing up or down. The magnet is positioned so
as to maintain, as nearly as possible, a flat fluid surface in FIG. 1. Experiments with magnet above ferromagnetic fluid
in osmometer cup. (1) Dialyzing cup with surface covered by
the cup along the total path of travel of the magnet. With thin film, 2; (3) meniscus; (4) manometer; (6) aluminum spring;
the magnet above the cup, it is necessary to cover the mag- (6) vise; (7) adjustable null scale; (8) magnet on lab jack. When
netic fluid with a thin plastic film that is clamped tightly corrected for a capillary blank (3), the manometer reading (4) is
around the edges but rests slack on the surface. Without this negative.
1093
 1094 Physics: Scholander and Perez Proc. Nat. Acad. Sci. USA 68 (1971)

I I I ~~~~~~I
I I CONCLUSIONS
We see how the osmotic pressure produced by the magnetic
0 colloidal substance relates to the free surface and not to the
-l0
membrane. As the magnet approaches the surface from above,
I-C.,
the osmotic pressure increases by the same amount as the
weight decreases, whereas the solute pressure against the
I membrane diminishes; conversely, as the magnet pulls the
Ig\ solute down from the surface it abolishes the osmotic effect,
w
LL.- -201
IL while the solute molecules increasingly press on the mem-

C,) -30f
I
I KI \ brane. The idea of osmotic force being driven by a difference
in water "concentration" across the membrane must be
abandoned. Evidently the water fabric is only interrupted by
0
solute molecules but is not "diluted" by them (2).
.s I I . I
a I a The perfect match between the magnetic and osmotic force
-4A in these experiments gives no room in osmotic terms for a
+10 0 -10 -20 -30 solute-solvent interaction, except by coupling with the free
MAGNETIC EFFECT MM H20 surface.
This research was supported by National Institutes of Health
FIG. 2. Magnetic effect on water-pressure in magnetic solu- Grant no. GM-10521.
tion. Left: magnet pulling down, right: magnet pulling up. OP
is osmotic pressure with zero magnetic force. The diagonal is 1. Noyes, A. A., Z. Phys. Chem., 35, 707 (1900); Hulett, G. A.,
drawn at 450 and is the theoretical line for surface coupling be- Z. Phys. Chem., 142, 353 (1902); Herzfeld, K. F., Phys. Z.
tween solute and solvent. Spread of observations is indicated by 38, 58 (1937); Mysels, K. J., Introduction to Colloid Chemistry
the length of the vertical bars. (Interscience Publishers, Inc., New York, 1959); Scholander,
P. F., Proc. Nat. Acad. Sci. USA, 55, 1407 (1966).
2. Scholander, P. F., Microvascular Research, 2, 1 (1971).
effect, leaving 5 mm due to nonmagnetic solutes. Two runs 3. Rosensweig, R. E., Ind. Res., October, 36 (1970).
gave identical results and the blank value without magnetic
4. Hargens, A. R., and P. F. Scholander, Microvascular Re-
search, 1, 417 (1969).
force remained the same. No magnetic particles were found 5. Scholander, P. F., Topics in the Study of Life: The Bio Source
to have penetrated the membrane. Book (New York, 1971), pp. 138-147.