Entanglement

Lately, I've been thinking a lot about entanglement. It started with my projects at work. The chemical industry has long recognized the importance of entanglements-often intuitively, without knowing the term or the polymer physics involved-within natural rubbers and silicone gums to toughen the final products used in automotive, construction, electronics, and even personal care markets. High molecular weight polymers essential for entanglement bring about many interesting properties that allow for the tuning of lubricity, adhesion, rupture, processing, and more. Recently, the concept of "entanglement engineering" has been revived in academia to better understand the detailed mechanisms starting from hydrogels and expanding to other industrial materials.

As an industrial scientist, I see tremendous value-not just intellectually, but also monetarily-in not only understanding the mechanisms, but also allowing formulators to easily leverage them in the labs and in manufacturing to get desirable properties within safety margins. To put this into perspective, in industry, many highly potent but toxic chemicals are being phased out due to regulations. And it's getting more difficult to develop new chemicals. This brings up a really important question: how can we use less toxic but often less potent chemicals to make products that perform just as well as those made with more hazardous substances?

I don't think we're quite there yet. One reason is that much of the soft matter physics knowledge remains in academia. And it is not widely recognized in the chemistry community. (Funny story: about 100 years ago, the father of rheology, E. Bingham, suggested to M. Reiner, civil engineer, that a new branch of physics should be established to deal with the flow of colloid chemistry. Reiner initially questioned the need, noting that such a branch already existed as "mechanics of continuous media". Bingham replied, "Such a designation will frighten away chemists!" He then consulted the professor of classical languages, which led to the term we now know: "rheology".)

​My interest now goes way beyond just the fundamentals of polymer science like I did in grad school. I'm really curious about where polymers actually show up in the real world-hot and tough (literally and figuratively) areas like data centers, electric vehicles, space exploration. In all these areas, getting the entanglement of polymer chains right, both before and after curing, is super important! I also keep an eye on the competitive landscape, like finding a durable "moat". It's getting rare in the fast-paced tech world, except for companies like NVIDIA and ASML (at least for now). In more traditional industries, though, companies like Graco and Rockwell Automation still have some pretty strong moats.

I don't know the nitty-gritty of neuroscience and deep learning, though I have a keen interest in their applications. Sometimes, my habit of thinking about all kinds of topics leads to what experts may find nonsense. I know the term "network topology" is used in both deep learning and polymer science. Interestingly, the concept is quite similar in both fields. In deep learning, it refers to the hypothetical space where nodes and layers (the building blocks) are interconnected via weights and biases to optimize model performance. In polymer science, it describes a physical arrangement where polymer chains (the building blocks) are interconnected through specific reaction pathways. In both cases, the topology of the network fundamentally determines the system's properties and behavior.

To me, entanglement is really just a way to get "systems" behavior-something you don't see if everything stays separated or poorly interconnected. In this sense, synapses in neuroscience which interconnect neurons to enable information transfer for survival, polymer chains trapped within a network which help dissipate applied stress and strain as heat energy in materials for durability, and parameters in deep learning which adjust variables within nodes and layers to make accurate predictions, are all examples of interconnected systems that give rise to new, collective behaviors. In neuroscience, that means human insights for survival; in deep learning; machine insights that predict better; and in polymer science, materials insights-even though materials aren't alive, they still figure out how to be durable through their structure under our supervision (unsupervised reaction pathways by conscious monomers can be developed in the future?). 

My post-graduation goals are still the same as they were 3 years ago: decipher global trends, resolve complex challenges, and generate financial opportunities by harnessing the synergies between academic and industrial research. I need more insights. I am very hungry. All these smart people around me have very strong insights. Could I ever reach that level? Maybe by entangling insights from lots of different areas-even though I'm not a top dog in each respective field-I could still add very unique value.