Note: this is rather technical.

This spring I had the pleasure of speaking briefly with a distinguished engineer, inventor, businessperson, and benefactor of science. He explained how he has recently become interested in the work of Prof. Gerald Pollack, who discovered what he calls the “4th phase of water”. The very term “4th phase of water” immediately raised an alarm bell in my head, since there are actually 19 or so known phases of water. I decided to check out what this “4th phase” was. It turns out this ‘phase’ has so far only been observed at the boundary with an odd material called Nafion, so really, it’s interfacial water with special properties, not a new phase of the liquid itself. My research focus the past three years has been understanding the microscopic details underlying the dielectric properties of water.  I am very interested in the structure and behavior of water around proteins and dissolved ions (and have read numerous papers on the subject) so naturally I am interested in Dr Pollack’s claims. Additionally, Pollack has shown that he can use the exclusion done phenomena to build a device that filters out nanospheres, and he claims his discovery can be used for desalination technology. He has not yet actually presented a functioning desalination apparatus, but he has filed a patent for the technology.

Background – pathological water science

There have been many spurious claims made about water over the years. I not referring just to the constant deluge of nonsense coming from hucksters and homeopaths who peddle ‘special’ water to cure various ailments. I am referring to research that has been pursued by highly respected scientists and published in top peer-reviewed journals like Science and Nature, but which ultimately turned out to be badly mistaken. History has shown that research having to do with water is very susceptible to what Irving Langmuir calls “pathological science“.  Two of the archetypal examples of pathological science are polywater and water memory. The case of polywater bears troubling similarities to exactly the claims Dr. Pollack has made.  Just like “EZ water”, polywater was purported to be a ‘new phase’ of water. Like EZ water, polywater only formed in special circumstances – ie. when water was condensed in tiny capillary tubes. Studying something that is confined to a tiny tube is also tricky, and in the same way, studying “EZ water” (water near a surface) is tricky. Years of painstaking research were dedicated to studying the properties of polywater. In the end it gradually became apparent that the tubes were contaminated with trace amounts of impurities, often from human sweat. In some cases, the ‘sample tubes’ contained very little water at all! The entire episode lasted about a decade before everyone admitted there was no real polywater.

I contend that the Mpbema effect, where hot water is observed to freeze faster than cold, is a modern day example of pathological water science. Invariably the experiments that found such an effect were later shown to potentially plagued by container variation, impurities, dissolved gases, and unwanted evaporation. The most carefully controlled experiment to date shows no such effect when an identical container is used (see my post on this). It was shown that “identical glass containers” are unavoidably non-identical – they have natural variation in their highest temperature nucleation site. (Brownridge, 2011)

A final example are the present day claims of the so-called autothixiotrpy of water – that containers of completely pure water will become more viscous after sitting still for a long time (months or years). The reported autothixiotropic effect is very small.

The main phenomena

<figure id=”” class=”thumbnail wp-caption aligncenter style=”width: 388px”>

<figcaption class="caption wp-caption-text">an example of an exclusion zone observed in Pollack’s lab. The microspheres (right) are repelled from the Nafion sheet on the left.</figcaption></figure>

Most of G. Pollack’s claims hinge on a single narrow class of experiments, where he measures how microspheres are repelled from different surfaces. In _Unexpected Presence of Solute-Free Zones at Metal-Water Interfaces_, for instance, he the various types of microspheres as follows:

“To quantify the size of the exclusion zone, various functionalized microspheres were used, all 1-µm diameter. These microspheres included carboxylate (2.65% solid-latex, Polysciences Inc.), polystyrene (2.65% solid-latex, Polysciences Inc.), amino (2.66% solid-latex, Polysciences Inc.), amidine (4.1% solid-latex, Invitrogen) and 488-nmexcitation fluorescent amine-modified microspheres (2% solid, yellow-green fluorescence, Invitrogen).”_

The main surface that he studies in most of his work is Nafion, a proton exchange membrane developed by DuPont which is used in fuel cell technology. More specifically, it is a “sulfonated tetrafluoroethylene based fluoropolymer-copolymer”.   Nafion, and its interaction with water, has been the subject of several studies__.Unfortunately, Pollack does not cite or discuss any of these studies, although I imagine he must be aware of them. Here is a list of some of relevant research that has been done on Nafion-water interaction:

  1. Quasielastic Neutron Scattering Study of Water Dynamics in Hydrated Nafion Membranes
  2. Properties of Nafion membranes under PEM water electrolysis conditions (review)]( [A Computer Simulation Study of the Mesoscopic Structure of the Polyelectrolyte Membrane Nafion
  3. Molecular Simulation Study of Nafion Membrane Solvation in Water and Methanol An infrared study of water in perfluorosulfonate (Nafion) membranes
  4. Atomistic Simulation of Nafion Membrane. 2. Dynamics of Water Molecules and hydronium ions
  5. Structural Organization of Water-Containing Nafion: The Integral Equation Theory
  6. Diffusion and Interfacial Transport of Water in Nafion

Nafion can absorb a lot of water, forming a matrix of nafion and water. It is postulated that water helps the exchange of protons via Grotthuss mechanism. The structure of fully hydrated Nafion is rather complex, with numerous water domains and channels inside. The exact structure is not known (see Wikipedia). Note that Nafion may contain SO4- groups which carry negative charge. It is clearly important how these charges are arrayed on the surface!

Water’s behavior at interfaces

What do we know about the behaviour of water near interfaces? An enormous amount of research has been done on this topic, but water is complex liquid and behaves differently depending on the type of interface and the microscopic details in many cases are not fully understood. This complexity is due to the hydrogen bond network and the many different ways that water molecules can orient themselves at the interface, which alter the average structure of the H-bond network near the interface. After surveying the literature, it seems that more research has been done on hydrophobic interfaces (review here). At a very hydrophobic surface, a ‘hydrophobic gap’ is formed between the water and the interface. This gap consists of both a tiny bit of vacuum between the water and the surface and a very thin layer of lower-density water.  A 2006 study found that “the size (of the layer) is the diameter of a water molecule” and “the integrated density deficit at the interface amounts to half a monolayer of water molecules”. This result has been independently verified (see here and here) although because it is so small, it is hard to actually measure. The largest layer where the density is perturbed has an extent of about 5 Ang, at the interface with highly hydrophobic self assembled monolayers. A similar story holds for the air-water interface, which is a type of ‘hydrophobic’ interface. Water can exhibit complex behaviour at metal interfaces as well, where water molecules are known to adsorb themselves to form a surface layer. Counter-intuitively, the adsorbed water monolayer can be hydrophobic.  In any case, much less work has been done on studying water at hydrophillic interfaces.

“secondary effects”

It needs to be emphasized that there are many possible ‘secondary effects’ that can contaminate a microsphere system. Microsphere systems have been heavily studied to better understand the hydrophobic effect. Referring to research that uses plastic microspheres one scientist in the field says the following:

these systems are notoriously plagued by secondary effects, such as bubble adsorption and cavitation effects or compositional rearrangements… Unfortunately, even if bubbles and other complications can be excluded, the very short-ranged portion of the hydrophobic between micrometer-sized surfaces can typically not be resolved experimentally because of mechanical instabilities of the measuring device. Likewise, the intricate scale dependence of the hydrophobic effect makes it nontrivial to relate the force between micrometer-sized particles to the one between molecules.”

Clearly these are temperamental systems!

Long aside: repulsive van der Waals forces?

It occurred to me that the repulsion of the micropheres from the nafion (and metals) may be due to the repulsive van der Waals force (which in this context is also called the Casmir-Polder force). The possibility that two objects of different composition may feel a repulsive force when submerged in a liquid was first realized by Hamaker in 1937. The full theory for such forces, for arbitrary dielectric media, was worked out by Lifshitz in 1954. Lifshitz’s equations allow for a repulsive force between two objects if the dielectric susceptibility of the medium between the two plates is intermediary between the two. Recent calculations using Lifshitz theory show that the finite thickness of the slabs does not effect the repulsion between them.(Zhao, 2011)(van Zwol, 2010)

In widely-reported work in 2009, a repulsive Casmir force was measured between a gold plate and a silica sphere submerged in bromobenzene.(Munday, 2009)  This sphere-plate geometry is easier to study since one doesn’t have to worry about precisely aligning two plates.  Similar repulsion has been found in follow up work with cyclohexane and other liquids.(Meurk, 1997)(Lee, 2002) More recently, so-called intermediate-range repulsion was observed between a ZnO nanorod and a SiO2 nanorod in bromobenzene.

I managed to find a very interesting study from 1996 that measured the Casmir force between gold and PTFE submerged in several liquids, including water. Milling et al. (1996) measured the force between a gold sphere and PTFE block submerged in several liquids, including water. They state:

For most of the polar solvents used ( water, ethanol, and dimethyl sulfoxide ) the forces between the surfaces was always attractive, albeit weakly in the case of ethanol. The only instance of repulsion between the gold and PTFE surfaces with a polar solvent was observed with dimethyl formamide…”_

Although the force they measured for water was attractive, Table 2 of their study shows that theoretically a gold sphere and PTFE should repel each other in water (indicated by a negative Hamaker constant). Interestingly, the sign of the Hamaker constant predicted by theory only matches the sign found by experiment in 3/10 cases, suggesting problems with either the theory or experiment. At the end of the paper, they explain that this is likely due to incomplete knowledge about the high frequency (UV) dielectric function of the materials (the theory requires the frequency dependent dielectric function as input).

The authors also make this revealing remark:

In some instances data collected using water as the intervening liquid showed a long-range exponential repulsion suggesting contamination of the surfaces by charge bearing groups. These may have been already present on the PTFE surfaces (21) as residual carboxylic groups from the polymerization process or other impurities, which would readily transfer to the gold surface. Clearly, van der Waals interactions between the surfaces are insufficient in describing the observed forces.”_

Also, interestingly, in table 1, the dielectric constant of gold is listed as 200. I am not sure where this number comes from (metals are often considered to have an infinite dielectric constant), but I will use this number to make a point. Let’s assume the dielectric constant of the metals Dr. Pollack used is very high (at least 100). The dielectric constant of water is 78. The dielectric constant of a polystyrene microsphere is about 2.5, and it is safe to assume the others have dielectric constants between 1.5 and 3. Thus, the metal-microsphere-water system obeys the conditions necessary for Casmir Pollard repulsion.

Unfortunately, there is a fly in the ointment of this theory – # retardation effects.#   van der Waals forces are due to quantum mechanical charge fluctuations in atoms and molecules. For an attractive (or repulsive) force to be set up, two bodies must be able to respond to each other’s random fluctuations. If the travel time due to the speed of light becomes similar the timescale (period) of the fluctuations, then the force is weakened.  For example, for the usual der Waals force between two atoms, the force changes from falling as 1/r^7 to fall as 1/r^8.  The speed of light is slower in a liquid, which makes the situation even worse. For example, Lee and Sigmund reported measuring retardation effects at 4-5 nanometers. This is much smaller than the exclusion zone reported by Pollack (which ranges up to 100 micrometers). However, full Lifshitz theory for macroscopic bodies (ie. microspheres) seems to give a different picture. vdW forces are highly non-additive in nature and this non additive can greatly enhance the force experienced between macroscopic objects (so, for example, the force between a collection of N atoms  (microsphere) with another collection of N atoms is not just multiplied by N). In his book Intermolecular and Surface Forces, Isrealachvili notes that here is also a non-retarded zero frequency component to the vdW force. According to Isrealachvili, in many cases the progression in the vdW energy is 1/r^6 -> 1/r^7 -> 1/r^6.

In one study, the non-retarded vdW force between a microsphere and an insulating wall in water persisted up to ~200nm, about exactly the same length scale as the exclusion zone!

The repulsive van der Waals forces probably are present in the microsphere system, and they may in fact be causing the exclusion zone. There is a remaining highly speculative reason though that I believe van der Walls physics may be relevant – Pollack claims that the EZ grows when infrared radiation is shined on it. This is reminiscent of a paper I found while researching the repulsive van der Walls force. Based on the abstract of the paper, the van der Walls forces between silver nanoparticles can be enhanced by radiation, through a process involving induced dipole moments. Additionally, the argument above about retardation effects probably needs to be is modified if charge is fluctuating along the entire nano-article, as may be possible if it has conduction electrons. Overall, I am left with the impression that there are likely weird aspects of van der Waals physics / Lifshitz theory still waiting to be illuminated. Surprisingly, it was recently discovered that retardation effects can change an attractive vdW force into a repulsive one.

Debunking the quackery from Dr. Pollack

It seems that Dr. Pollack has alienated himself from the rest of the scientific community by peddling grandiose nonsense. Instead of with focusing on his discovery and getting other scientists to replicate it, Dr. Pollack has built an entire edifice of nonsense on top of it, which he documents in his book, The Fourth Phase of Water.  He discusses his ideas in a TEDx talk, another example of why TEDx is not TED. Dr. Pollack sees his ‘fourth phase’ Many of his claims are easy to refute. The first idea he has is that the hydrogens somehow lie directly between the oxygens, even though are most sophisticated quantum chemistry simulations have never shown any such behavour. Such behaviour isn’t predicted to occur in ice until extremely high pressures are reached (so called ‘superionic phase‘) The ‘phenomena’ that he attributes to EZ water are actually just mundane surface tension. Most notably, the electrically-induced water bridge (described in an earlier post) which Dr. Pollack claims is made of EZ water has been shown by researchers to have the same internal structure as regular water – implying that the water bridge is supported by enhanced surface tension.  Both molecular dynamics simulation and X-ray crystallography support this conclusion.

Pollack belongs to a group of people who think that confined water may have a special structure and special properties. Recently, the moniker ‘biological water’ has emerged to describe how cellular water may be different than normal, bulk water. It is true that confined water does exhibit different thermodynamic properties, such as reduced freezing point, and understanding how the effects of confinement change with confinement volume and geometry is an active area of research. The thermodynamic changes can largely be explained by the macroscopic phenomena of Laplace pressure of the solid-liquid boundary (detailed ref). Most of the claims about biological water have scant experimental evidence to back them up and contradict findings from computer simulation. Growing research on the hydration water around proteins shows evidence of dipolar ordering being disturbed up to a few nanometers from the protein, but structural ordering (ie, as shown in the Oxygen density PDF/RDF) is only disturbed a few angstroms from the surface.

Pollack pronounces is that when sunlight is shined on EZ water, it causes positive and negative charges to separate, and the EZ water region to grow. This seems rather dubious since water is a good conductor. In another experiment, Pollack shines light on a tube and finds flow of water.  Instead of trying to understand this experimental result through modeling the induced convection currents or by asking others to replicate it, he draws an outrageous conclusion – that the human body may absorb light to cause the blood to flow! If this was the case, mammals that live in darkness would die, along with animals with fur coats.. Most troubling, Dr. Pollack has no qualms about associating with Dr. Mercola, an anti-vaxxer and alternative medicine huckster.

The journal of which he is editor, WATER is essentially an outlet for work on water that is too speculative or unscientific to appear in other journals. (WATER should not to be confused with the other bottom tier journal with the same name, water, which has been known to publish papers on water memory and other pseudoscience used to justify homeopathy.) Along the same lines, Pollack has helped organize a conference with an all-star lineup of the world’s top water crackpots. This is done to promote ‘speculative research’, something which I agree we need. Such conferences are dangerous though, in that they can legitimize each other’s research in an echochamber,and isolate researchers from the scientific scrutiny of the scientific community at large.

Kernel of truth 

The frustrating thing about Dr. Pollack’s research is that clearly he is observing some effect, but we can’t really say with confidence that it the type of effect he purports until it is reproduced by independent researchers!

My specific advice to Dr. Pollack (or his coworkers), is:

  1. Have someone independently try to reproduce your work.

  2. Cite previous work.  In one paper,  11/12 of the references are self citations! (with the remaining reference being to Einstein’s famous 1905 paper on Brownian motion). This is despite the fact that, as I mentioned, much work has already been done studying water-nafion interaction and water-microsphere systems.

  3. Associating with Dr. Mercola will give you a bad reputation. Avoid him.


A very interesting and very promising theory for the exclusion zone can be found in these papers:

(1) Phenomena Associated with Gel-Water Interfaces. Analyses and Alternatives to the Long-Range Ordered water Hypothesis ( J. M. Schurr, J. Phys. Chem. B 2013, 117, 7653-7674)

— for background, see A Theory of Macromolecular Chemotaxis (J. M. Schurr et al., J. Phys. Chem. B 2013, 117, 7626-7652)

(2) Long-range Repulsion of Colloids Driven by Ion Exchange and Diffusiophoresis (D. Florea et al., Proc. Natl. Acad. Sri. USA 2014, 111, 6554-6559)

The last paper, published in PNAS, includes an independent experimental verification of the exclusion zone phenomena and neat videos. Another independent experimental verification was published here:

Exclusion Zone Dynamics Explored with Microfluidics and Optical Tweezers (I. N. Huszar, et al., Entropy 2014, 16, 4322-4337)


  1. Pollack GH. Chai B, Mahtani AG. Unexpected presence of solute-free zones at metal-water interfaces. Contemporary Materials, # 31, 2012.
  2. H.C. Hamaker. The london van der waals attraction between spherical particles. Physica, 4(10):1058 – 1072, 1937.
  3. Irving Langmuir and Robert N. Hall. Pathological science. Physics Today, 42:36, 1989.
  4. Anders Meurk, Paul F. Luckham, and Lennart Bergs ̈om. Direct measurement of repulsive and attractive van der waals forces between inorganic materials. Langmuir, 13(14):3896–3899, 1997.
  5. Andrew Milling, Paul Mulvaney, and Ian Larson. Direct measurement of repulsive van der waals interactions using an atomic force microscope. Journal of Colloid and Interface Science, # 180# (2):460-465, 1996
  6. Capasso Federico Munday, J. N. and V. Adrian Parsegian. Measured long-range repulsive casimir-lifshitz forces. Nature, # 457:# 170, 2009.
  7. P. J. van Zwol and G. Palasantzas. Repulsive casimir forces between solid materials with high-refractive-index intervening liquids. Phys. Rev. A, # 81# :062502, Jun 2010.
  8. Seung woo Lee and Wolfgang M. Sigmund. Afm study of repulsive van der waals forces between teflon af thin film and silica or alumina. Colloids and Surfaces A: Physicochemical __and Engineering Aspects, # 204# (13):43 – 50, 2002.
  9. R. Zhao, Th. Koschny, E. N. Economou, and C. M. Soukoulis. Repulsive casimir forces with finite-thickness slabs. Phys. Rev. B, # 83# :075108, Feb 2011.
  10. E. M. Lifshitz. The theory of molecular attractive forces between solids. Sov. Phys. JETP, 2:73–83, 1956.