Featured Image: 

I'm Kevin Theisen, currently a graduate student in chemistry at the University of California: Berkeley. I care very much about students and about the environment, and I am currently studying solar energy and other renewable energy sources. I am also the President and Founder of iChemLabs. This article briefly records the events and software that define the history of this company and I hope this article is an interesting story about its conception and evolution.

In the fall of 2004, I was an undergraduate freshman at Rutgers University in New Brunswick. I had already decided to major in chemistry and minor in computer science. In chemistry, I was very interested in molecular visualization. I had asked my first professor (my later undergraduate advisor) how I could see molecules. He blatantly responded that we couldn't. This is due to the fact that the wavelengths of visible light are much larger in scale than the atoms they may or may not be reflecting off of. Therefore it is impossible (with our current technology) to fully resolve atoms with a visible light microscope, however many types of spectroscopy and microscopy do allow "visuals" on the Ångström scale. This lead me to actively study the very broad field of nuclear magnetic resonance (NMR).

In computer science, another year passed while I was debating with myself why I was minoring in it at all, when I finally began to learn about the graph data structure. Graph data structures are defined by any number of nodes containing data and links that connect two nodes in one or both directions. It occured to me that almost everything in chemistry was characterized by this simple structure. Atoms, molecules, nanostructures, synthetic routes, mechanisms, glassware apparati, and everything else I could think of was easily defined by this structure.

Out of curiousity, I began working on a simple method to optimize synthetic organic pathways. It quickly developed into a powerful algorithm for optimizing synthetic routes based on cost, yield, time, number of steps, hazard levels, and any other relevant criteria. It was a small Java application that was run in a command prompt and produced very interesting results (even though it was just text), as the image below shows.

A query for the synthesis of 2-bromo-2-chlorobutane from butane produced three results with varying costs. The best route calculated was the bromonation of butane, followed by the S_N2 replacement of bromide by chloride, followed by a second bromination. Additionally, when asked to calculate the most efficient route, the synthesis calculator suggested I start from 2-chlorobutane as it is readily available, allowing for an even easier and cheaper synthesis. I was surprised by the thoroughness and power of this application. Being a chemist, I realized there was an unlimited amount of possibilities to aid the scientific community and I began to dream up more ambitious and challenging applications.

It was now the end of my sophomore year and I had begun to look at commercially available applications to help me pursue chemistry and specifically to learn about NMR (I really wanted an NMR spectrometer, but then realized I'd need a bunker, molecule samples and a million dollars). The current choices were unacceptable. They were inaccessable to students in many ways: cost, availability and other hurdles in obtaining the software. I wanted to make sure that I and other students could use powerful software to help pursue our educations, and I was very dissappointed with our options. I wanted to use an NMR simulator to test my answers to excercises and develop ideas and other trends about NMR interpretation. I realized that I would have to write one myself, so I began.

In the summer of 2006, my application for predicting, explaining, and analyzing ¹H NMR, HNMRazor, was released. I wanted to make sure all students could use the application so I created iChemLabs to distribute the applications online for free. The predicted spectrum of Biotin produced in HNMRazor is displayed below.

HNMRazor was an incredibly accurate application and allowed me to really further my knowledge of NMR (it's very surprising how much information is kept from students in undergraduate chemistry, but mostly due to time constraints). The creation of this software was an incredibly fun excercise, and encouraged me to continue studying the topic. To date, I was able to develop a few algorithms for accurately predicting chemical shifts of various systems, a few of which I will be publishing. Looking into other nuclei, a few months later, HNMRazor's ¹³C NMR equivalent, CNMRazor, was released.

On an aside, an interesting web application I had been playing with, MolGrabber, became quite powerful. A screenshot is displayed below. Based off of an image of PubChem's public database, I was able to sort and search through 8 million molecules at my own pace. The ability to seamlessly search through and sort chemical data is very important (the chemical informatics industry makes a huge amount of money on it), and I wanted to be able to integrate that type of power with all of my applications. Since then, all iChemLabs applications have easy access to large databases of 2D and 3D compounds for all kinds of uses, a feature which is unparalleled by other software.

Over the next year I began to realize the limits of the Razors and of trying to use a two-dimensional palette to accurately describe our three-dimensional chemicals. There are two options to improve the description of 3D entities: I can continue to expand upon the 2D interface that describes the 2D structures or I can portray chemicals in a 3D system. I had long wanted to obtain a 2D chemical drawing application that worked the way I imagined, and I also wanted to finally be able to visualize and manipulate molecules in 3D. Since both were great ideas, it became my goal to create fully functional 2D and 3D "sandboxes" for chemical manipulations and investigations.

A few months later, in the fall of 2007, ChemDoodle, iChemLabs fully functional 2D chemical structure environment was released.

It was a completely native-acting Java application on all systems (most Java applications use very generic looking and acting GUIs because of the limitations involved in providing a cross-platform programming language) and surpassed all of the original expectations I had imagined (I actually didn't think I would have the time to create it). ChemDoodle was incredibly powerful as it was as functional as any other chemical drawing application, but was molded in a sensible and aesthetic way to allow for easy use. Widgets, small programs inside of ChemDoodle, provide additional and amazing functionality and new widgets are often added. Currently, ChemDoodle supports several widgets for SMILES interpretation, multiplet generation, database searching, template managing and many other functions. I also wanted ChemDoodle to be easily accessible and available to all students, so iChemLabs provides the complete version for only a very small fee, just to offset distribution costs and ensure continued development on the project.

The 3D "sandbox" is still under development, but trust me it is amazing. I won't release too many details, but it will be completely configurable and connected to a number of databases. All kinds of quantum simulations will be available and any functions will be easily tweaked.

iChemLabs is and always will be a company that creates high quality scientific software and distributes it at minimal cost to students and scientists around the world. We will always be pursuing new ideas and new applications and I welcome you to contact us with any ideas, suggestions or comments; you may do this by clicking on the CONTACT tab above. You are more than welcome to contact me personally at or at my Berkeley address, . It is with your continued support that we continue to feel enthusiastic about our goals.

Kevin Theisen
President and Founder, iChemLabs