A new recipe to shape magnetisation using network structure
[Posted on 20 November 2026]
Article by Unnati Ashar
Illustrations by Arunima V
The word magnet might bring a souvenir fridge magnet to mind. But there are other forms of magnetism, and scientists at the Raman Research Institute (RRI) have cracked a way to tune an exotic form of quantum magnetism in a network by changing the connectivity of the network. This may have applications in quantum devices for example in storing quantum information.
“Many systems around us are composed of interacting objects, and that can range from gas in this room to social networks; it can also be biological, technological networks,” says Revathy B. S., the lead author of the new study and a postdoctoral fellow at RRI, an autonomous institute of the Department of Science and Technology. And depending on how the components are connected, the collective behaviour can change.
The hub-and-spoke diagrams showing flight routes in glossy aeroplane magazines illustrate how connectivity allows people and goods to move across cities. Another relatable example, but one that could evoke unhappy memories, is that of superspreaders during COVID, who acted as hubs in spreading the disease. These are familiar instances that show how different patterns of connections can determine collective behaviour.
Why study quantum networks?
But these are all classical systems. “There’s nothing holding you from asking a similar network-level question for quantum systems. Can you get new behaviour by connecting [quantum units] in an unorthodox way?” says Shovan Dutta, co-author and assistant professor in the Theoretical Physics group at RRI.
The work of Steven Strogatz, an American mathematician and author who studied classical networks, seems to have greatly inspired Dutta. “I’ve been reading Strogatz since childhood. … So I was always looking up to his work. … Those influenced me a lot,” Dutta says with childlike enthusiasm. While Strogatz worked on classical systems, Dutta’s path took him to quantum systems.
“From a fundamental point of view, it’s a question of what happens when many quantum objects interact with each other? What sets the collective ordering? How do they organise themselves?” says Revathy. “And how does quantum mechanics change things compared to classical systems—that’s the broader question that motivated us to pursue this research.”
Forming new bonds
To understand the system studied by Revathy and Dutta, picture a square lattice with one vacant site (hole) and all other sites occupied by one electron. The electrons have an intrinsic property called spin, which can point up or down. The hole can hop from one site to another, and this motion of the hole can influence how the electron spins arrange themselves, giving rise to what scientists call “kinetic magnetism”. Unlike conventional magnetism which arises from direct interactions between the spins, kinetic magnetism arises from the quantum motion (kinetic energy) of particles. Kinetic magnetism is a true quantum-mechanical effect with no classical counterpart.
If all the spins point along the same direction, it is called ferromagnetic order, and if adjacent spins have opposite directions, it is called antiferromagnetic ordering. For a square lattice with one hole, it has been shown before that the lowest-energy state is ferromagnetic.
Revathy and Dutta showed that the magnetic ordering can be varied by the connectivity of the network. For example, introducing a diagonal bond in the square lattice makes the spins of the two electrons involved antiparallel, creating a quantum mechanically anti-aligned pair of spins, called a “singlet”. The remaining sites retain the ferromagnetic ordering, and the hole keeps hopping across the network.

Fig. 1: The top row shows the possible magnetic ordering of a many-body quantum system within a single square lattice. The bottom shows that the introduction of a diagonal in the square lattice yield
Towards deterministic predictions
“What’s amazing is that I can almost deterministically predict that wherever I [introduce a diagonal bond], I will get [singlets],” says Revathy. By introducing a diagonal bond and changing the connectivity locally, they could engineer the global magnetic order. These findings demonstrate that collective quantum order can be shaped by geometry alone.
“This is far from obvious, because if you had a classical system, similar changes in geometry do not really affect the magnetic order,” says Dutta.
Adding a diagonal bond creates triangles in a square network. A triangle is an example of an odd (three-sided) loop, while a rectangle is an even (four-sided) loop.
“The fact that an even-sided loop has ferromagnetic ordering was discovered in the 60s, believe it or not,” says Dutta. “But it is still not known to this day that if you have a random network which has odd-sided loops, not just even-sided loops, what is the ordering that you get.” This work partly answers that historical question and demonstrates that even random networks could be engineered to control their magnetic behaviour.

Fig. 2: Global magnetic order can not only be determined in a quantum many-body system, but also engineered
Taking up the challenge
The computational complexity involved prevented physicists from tackling the problem for decades. A lattice with 20 sites yields a mind-boggling two million configurations. Revathy and Dutta used a combination of advanced numerical and analytical approaches to uncover these microscopic patterns and their physical origin.
The research published in Physical Review B is also particularly timely because experimental advances are reaching a stage where researchers can possibly observe the complex theory in action.
Towards applications
Although this work is theoretical, it opens new avenues for shaping quantum states of interacting particles, which may have applications in quantum devices. Our familiar hard drives, for example, use magnetic patterns to store information, but only as classical bits. The singlets in these networks are entangled quantum states and may be used to store quantum information.