PRINTING STOCHASTIC MASA
By Alan S. Brown

Inspiration struck Jeffrey Lipton on a flight home from a conference. The Ph.D. candidate in mechanical engineering at Cornell had just delivered a paper on what he described as "things that were foodish."
That was two years ago, shortly after Lipton joined fab@home, an open-source project to develop cheap 3-D desktop printers. The group had a following among engineers, designers, artists, and hobbyists.

Then fab@home started printing food. "We got orders of magnitude more response," Lipton said. “It fell into one of two categories. People either thought it was the coolest thing ever, or the most vile and disgusting concept they ever heard.”

Fab@home started with chocolate, frosting, and Cheez Whiz. Think of them as low-hanging fruit. While all are complex liquids, they solidify when deposited. Lipton, however, wanted to print foods from more basic starting ingredients.

This proved difficult because most foods are chunky. The only way to force then through the syringe of a 3-D printer is to turn them into runny emulsions that flow all over the table when deposited. Lipton decided to attack the problem head-on by learning to print gels made of viscosity modifiers, flavors, and food coloring.

Fried scallops, printed from a CAD drawing of the Space Shuttle.


The results were indeed "foodish": edible, though not mouth-watering, and totally free of nutritional value. Lipton's creations were, alas, just a science experiment. Then, on the flight home from his meeting, Lipton began reading about chef Dave Arnold.

A former fine arts graduate of Columbia University, Arnold had tapped his inner geek. As director of technology at the very cosmopolitan French Culinary Institute in New York, Arnold uses laboratory centrifuges to clarify juice, blogs about hacking his pressure cooker, and sounds poetic on the virtues of meat glue (also known as transglutaminase).

Arnold wanted a 3-D printer, and Lipton and his colleagues were happy to oblige. They immediately saw what they were missing. Within hours, Arnold had emulsified scallops and celery in agar and printed a layered treat. Complex shapes made from turkey and bacon (with meat glue to bind it together) followed. "They can be really delicious if you know what you’re doing," Lipton said with obvious respect.

Lipton went back to Cornell and was soon printing cookies with complex shapes and words that showed up when diners cut or bit into them.

Arnold meanwhile, kept looking for something he could only make with a 3-D printer. At a second meeting, Lipton told him about his work printing filters and catalyst supports with randomized thicknesses and patterns. Arnold jumped right on it, using this stochastic process to print filter-like blocks of masa, a corn flour used to make tortillas. He then fried them up. "I could eat a bajillion of those," Arnold wrote on his blog, cookingissues.com.

Granted, printing food sounds like the most trivial of pursuits. Yet according to fab@home's founder, Cornell professor Hod Lipson, the challenges are similar to those found in printing replacement bone, cartilage, skin, and even organs.

Printing replacement body parts makes surprising sense. Researchers know that cells will not form body parts unless they grow on a 3-D structure, or scaffold. But while growing liver cells on scaffolds is easy, creating a useable liver with its branching blood vessels and other substructures is not.

Three dimensional printing gives medical researchers a way to build substructures and scaffolding on the fly by depositing different seed cells in the proper locations. Other researchers are investigating depositing artificial skin on burn victims to promote healing and prevent infection.

Working with food gives researchers the experience they need to handle cells and other fragile biomaterials. This starts with developing ways to solidify emulsions runny enough to flow through a syringe. "You don't want an ear or a cookie to ooze out," Lipson said. His lab has experimented with crosslinking agents as well as heating and cooling to keep liquid foods in place.

Printing foods also requires equipment modifications. Commercial printers use only one material at a time. These materials are so well characterized, printers do not use feedback loops. Food, on the other hand, achieves flavor and texture from ingredient mixtures. And those ingredients are anything but characterized.

"If they begin to ooze, the printer should sense that it needs to slow down, reduce the temperature, or deposit material in a different pattern," Lipson said. “We want it to adapt, because it will never know what kind of foods it will have to deal with in advance.”

Lipson envisions a world where people buy 3-D printers and download recipes from famous chefs. "Food printing for 3-D printers could be what gaming was for computers,” he said. “It looked frivolous, but the market was huge.”