Here, we report a ∼60-fold increase in the photovoltage through use of a bipolar membrane structure consisting of a cation-exchange membrane affixed to an anion-exchange membrane. The junction between the layers was characterized in detail using electrochemistry, scanning electron microscopy, spectroscopy, and thermal gravimetric analysis. Our results represent considerable progress toward a device that directly converts sunlight into ionic electricity, which has implications for direct solar desalination of salt water. …


It has been over three-quarters of a century since the discovery of the semiconductor pn-junction diode, which later led to the advent of other functional circuit elements, such as the transistor and the integrated circuit.1 These circuit elements enable many of today’s advanced technological capabilities, including the digital camera, computer, cell phone, and solar cell. Less than 15 years after the discovery of the semiconductor pn junction, a publication by Fuller compared water to a semiconductor,2 indicating that the equations that approximate equilibrium carrier concentrations in non-degenerate semiconductors are the same as those that dictate the concentration of protons and hydroxides in water. Several years later, the first electronic pn-junction solar cells were developed and within just 2 years after their discovery, a sunlight-to-electricity power conversion efficiency of >6% was demonstrated.3,4 Around this time, work by Bockris and colleagues showed that water, with the proper permselective membrane contacts for protons and hydroxides, could behave like traditional diodes and exhibit ionic current rectification.5 Although the analogy of water to an electronic semiconductor had been clearly articulated,6 there was little evidence to suggest that researchers attempted to fabricate solar cells based on water as the “protonic semiconductor,” most likely because the properties of water result in rates of ion transport and ion lifetimes that are much smaller than rates of electron transport and electron lifetimes in high-quality electronic semiconductors. It was not until 20 years later, in 1978, that Prof. George Murphy briefly outlined several possible means to convert light into ionic power through photodriven transport of protons and hydroxides;7 however, there are no publications that report experiments related to Murphy’s purely ion-transporting concepts in the absence of Faradaic charge transfer. Today, 40 years later, no one has demonstrated experimentally or theoretically that an ionic analog to the electronic pn-junction solar cell can be made for sunlight-to-ionic-electricity power conversion. Here, we describe the fabrication and photoelectrochemical evaluation of such a device and also include a perspective on outlooks for its practical use.

To further support the notion that water is a “protonic semiconductor,” we compare the fundamental concepts of doped electronic semiconductors with water-containing ion-exchange membranes. Standard physical equations that describe properties of traditional electronic pn-junction semiconductors are introduced along with analogous concepts for water-containing ion-exchange membranes.