Is there any unilateral heat exchanger?

Given the details you gave in the comments of the question.

An active solution can be designed and realized. Heat exchange is driven dy a temperature variation. Say you have this system:

• chamber 1 at T1=50°C (this is the chamber we want to keep at fixed T)
• chamber 2 at variable T2 (this is the outside environment)
• wall between the two chambers on which we have a water cooled heat exchanger with water running at temperature TE

Case 1: T2=20°C (T2<T1)

In this case the water running in the exchanger will be at TE=50°C(an heater will be needed) so T1-TE=0 that means no heat flowing. At the same time TE-T2>0 so chamber 2 will cool the water running in the exchanger.

Case 2: T2=80°C (T2>T1)

Also in this case the water will be kept at TE=T1=50°C so again TE-T1=0 and no heat is flowing. Now T2-TE=30°C meaning chamber 2 will heat up the water thus a chiller will be needed.

Clearly both the chiller and the heater might be fit into the same controlling device. An heat exchanger in this case might be a rather trivial plate with channels for water to flow into. As @Solar Mike suggested a peltier device might be used but I think it will be more complex as it will need cooling, maybe liquid, in order to work properly.

---

Passive solutions may be possible but I can't say much without any detail on temperatures and time for which the system must work. If it work for an indefinite time then the active solution is the only way possibile as for long times any system will reach an equilibrium point at a common temperature.