An interesting question, need a little clarification for a real good out of the box soln.

Are you expecting the so called "physical process" which prevents the current flow between the two metals to reversible or irreversible ?

Are we considering AC or DC volts.

Are you expecting the physical process develop on its own after some time upon the application of 0-2 volts, which set in and then blocks the current permanently,and stays on permanently as a block. It begins acts like like a fuse.

K. Sreenivas 

These are great questions, I am sorry I didn't specify them initially. 

There doesn't need to be any reversibility to whatever process stops the two metals from conducting charge between each other.

The situation is DC volts.

I don't want to limit the imagination here, but i suspect the current-blocking mechanism will not be the result of a chemical reaction at the interface that creates some dielectric layer.

A thin P type semiconductor and N type semiconductor embedded between two metals (at junction) creates three depletion layers (1) Metal 1 (say M1) to semiconductor (2) Between two semiconductor (say N to P (3) second semiconductor to Second metal ( say M2). There will be three potential barriers exists and with appropriate concentration of impurity, in turn, majority carriers, the flow of current may be prevented

What's wrong with an oxide - most metal oxides are insulators (thus oxidised Zr is commonly used as an insulator in high temperature electrochemistry)? All that is necessary is to ensure that the oxide is thick enough to prevent dielectric breakdown (this is exactly the reverse of the problem in switch contacts, where it is necessary to avoid problems due to oxide films). For metallic conductors you will get conduction in the absence of a dielectric layer or a suitable set of semiconductor layers.

I imagine that there must be some combination of intrinsic electron properties of various metals we can leverage simultaneously without resorting to a classic dielectric layer. For example, to first approximation, a metal/solution junction with a high ionic strength can act similarly to a metal/metal junction. As long as the metal/solution junction maintains a potential far enough away from any redox potential, it behaves similarly to this conditions I am trying to recreate here...except with two metals.

It is a crazy question, indeed!

If current doesn't flow across a metal/metal junction, how do you think electrical circuits function?

The fault in your crazy question arises due to your so called first approximation, where by you assume a metal solution junction with a high ionic strength to act similar to a metal/metal junction.

No current ever flow through a single metal/solution junction! Current can and does flow only through a proper combination of two metal/solution junctions. This is in contrast to current flow through a single metal/metal junction!

In electrochemistry parlance, electric current flows through a proper combination of two electronic/ionic conductor junctions, never through a single electronic/ionic conductor junction. However, electric current can and does always flow through a single electronic/electronic conductor junction. 

Therefore, no physical processes exist to prevent current from flowing between a metal/metal junction.

1) How about wires of special amorphous composistions like (NbxSi1-x) which exhibit superconductor -metal-insulator transition. there maybe some other compositions and definitely do need further extensive research.

    The question that arises:  

    1) Will these amorphous alloys will be ductile/malleable/ alloyable to other contracts to make fine metal-metal joints. Change in temperature will make this special alloy to transform from (superconducltor-metal-insulator transition)

    2) The current carrying capacity of such wires, and the temperature change required for the transition in the conduction mechanism.

    3) How good is the metallic conductivity, and how good is insulating property, to really prevent the blockage of current flow. 

2) How about thin wires of Vo2, (now this is not metal, but metal oxide). 

     Problem is with ducility/malleability properties of a typical metal/metallic alloy.

     But depending upon the geometry of the wire, pump sufficient current to heat up the wire, and bring about the changeover from metal-insulator transition.

     But the advantage is a reversible changover between the metallic and the insulator phase when heated  just beyond the low transition temperature of about

65 to 70 deg C. The phase transition can be induced/incluenced electronically, optically and mechanically.

    Enclosing a relevant reference of a paper by:

     Arash Joushaghani et al., in the Optical society of America, and available on the google.  They talk about a thin VO2 wire, which exhibits a current controlled optical modulation in thin VO2 wires. 

    There are some problems still left with VO2 to understand further, and improve upon them especially the switching speeds 

All these are known, and the ongoing development/improvements are still appearing, but mostly trying to solve technical problems, but not aimed at an interesting physical process.

I fully agree with Mathew Bruke Starr and convince myself that I am still away from that so called out of box solution.

To really achieve the prevention of current flow, we need to have a good transition towards a good insulator characteristic. (resistance must become extremely high). 

K. Sreenivas

Matthew

You have got a selection of answers, but we are hampered by not knowing why you want to do this - there are plently of simple ways of isolating two metallic conductors (air gap, oxide layer, polymer layer etc.) and of switching the connection (physical movement, semiconductor switch). What do you actually want to achieve?

The capacitance of this junction would be very high. I am thinking about high capacitive junctions.

More importantly, I think it is a great exercise to ask a simple question that on the face of it challenges a simple rule of thumb. questions like this make you dig into the fundamentals and challenge your beliefs.

What about spin valve

https://en.wikipedia.org/wiki/Spin_valve

or memristors

https://en.wikipedia.org/wiki/Memristor

Ivan, great! I looked into these and the phenomenon responsible for GMR and topological insulators. I think these are a great direction to go in for finding physical phenomena that will be useful.

From a practical standpoint, the change in resistance with spin can be impressive and significant (milliohm to ohm), but significantly more resistance is necessary to prevent appreciable leaking through this junction.