HomeVideos

Inside The Startup Reinventing America’s Trillion Dollar Chemical Industry

Now Playing

Inside The Startup Reinventing America’s Trillion Dollar Chemical Industry

Transcript

391 segments

0:00

This reactor in Houston, Texas,

0:02

represents a whole new approach to

0:04

chemical manufacturing. We make

0:05

chemicals in a completely new way. We

0:08

started off by making hydrogen peroxide.

0:10

Now we sell products for water

0:11

treatment, national defense,

0:13

infrastructure, agriculture, you name

0:15

it, we make a chemical for it.

0:16

>> The story of Cyogen is as scrappy as it

0:19

gets. It all started with this prototype

0:21

built out of PVC pipes the founders

0:23

bought from Home Depot.

0:25

>> In its heyday, this did 12,000 bucks a

0:27

month in revenue. Fast forward to now,

0:29

Cyogen is a billiondoll company shipping

0:32

out tanker trucks of products that power

0:35

critical US industries. So, how did they

0:37

go from this single beaker of hydrogen

0:40

peroxide to this full-scale

0:42

manufacturing plant? I visited Solen HQ

0:45

in Houston, Texas to find out.

1:00

What do you guys do here? What is solen?

1:02

>> We use biology to create chemicals that

1:05

allow us to create smaller chemical

1:06

plants and have a cleaner, safer, more

1:09

environmentally friendly footprint.

1:10

We've invented a process called chematic

1:13

processing where we take the specificity

1:16

of biology by taking an enzyme and

1:18

pairing it with a metal catalyst uh

1:21

which is what's traditionally used in

1:22

industry. What we've done with these is

1:24

by marrying these two, you can actually

1:26

create more efficient reactions. So

1:28

instead of having a 60% yield, you can

1:30

have 96% yield, which is what we have at

1:33

scale precisely because of these two

1:35

catalysts.

1:35

>> Syogen is the first company to fuse

1:37

biology and chemistry in this way.

1:40

Pulling enzymes from living cells, in

1:42

this case corn syrup, and pairing them

1:44

with novel metal catalysts. The output

1:46

is chemicals that can be used in

1:48

everything from agriculture to skinare.

1:50

Traditionally, chemical plants relied on

1:52

fossil fuel feed stock, which has led to

1:54

all kinds of unintended consequences.

1:56

This new approach is cleaner, safer, and

1:59

more efficient. We receive rail cars of

2:02

corn syrup, and then we run our utility

2:04

system. We change the parameters of the

2:06

plant, the way that the enzymes are

2:08

reacting, the way that the metal

2:09

catalysts are reacting to oxidize the

2:11

corn syrup how we want. And then at the

2:12

end, we evaporate water. We take the

2:14

final products, we store them in

2:16

finished good tanks or we send them to a

2:18

blend farm to get blended with other

2:19

chemicals and that's the process.

2:22

>> Sogen started with a true Eureka moment,

2:24

the kind of freak invention that

2:26

normally happens only in science

2:27

fiction. Shawn was working in the

2:29

chemicals industry and Gorb was in

2:31

medical school studying pancreatic

2:33

cancer.

2:34

>> I was working on this like skunk works

2:35

project to try to do direct hydrogen

2:37

peroxide synthesis where you react

2:39

hydrogen and oxygen gas directly

2:41

together over a metal catalyst. And like

2:42

you know I was working on this project

2:43

and then Gorb was like oh yeah no like I

2:45

found this like really crazy mechanism

2:46

in pancreatic cancer where it's like

2:48

locally 50% hydrogen peroxide

2:50

concentration. I was like what that's

2:51

wild.

2:52

>> It turned out that an obscure enzyme

2:54

found in pancreatic cancer cells that

2:56

Gorb was studying was the key to a new

2:58

process for making industrial hydrogen

3:00

peroxide.

3:01

>> So these pancreatic cancer cells they

3:03

put out peroxide. Peroxide creates

3:06

almost like an invisibility cloak around

3:08

the pancreatic cancer that makes it

3:10

difficult for immune cells to come in.

3:11

It was like that's a very interesting

3:13

discovery, but why? And so once we asked

3:16

the question why, we found out the

3:18

reason was an enzyme. It was just a

3:20

mutated enzyme that's found only in

3:22

pancreatic cancer. And so that's when we

3:24

said, what if the two worlds could

3:26

collide, right? What if enzymes and

3:28

metal catalysts could coexist?

3:30

>> Here's how it works. Cyoggen feeds corn

3:33

syrup to the enzymes, the same ones from

3:35

the pancreatic cancer cells, which

3:37

transform it into new compounds. They

3:39

then use metal catalysts to further

3:41

process those into the final chemicals

3:43

that can be used to build all kinds of

3:45

products we use in our everyday lives.

3:46

In most chemical plants, the feed stock

3:48

comes from oil and gas, which inevitably

3:51

produces toxic byproducts. Solugen

3:54

literally starts with sugar. We have to

3:57

suspend belief for a second. People

3:59

believe that anything to do with biology

4:01

and chemicals, it's just a bad mix

4:03

because, oh, biology is too sensitive.

4:05

It's going to break down blah blah blah

4:07

blah blah. But we said let's just

4:08

suspend that criticism for a second and

4:11

just look at the numbers and we say if

4:13

we can make the enzyme last this long on

4:16

stream and the product is this

4:18

concentration we can make a lot of money

4:20

and that's where we started.

4:22

>> This insight that organic enzymes could

4:24

operate at industrial scale and

4:26

efficiency was the company's first key

4:28

breakthrough. Today Scen operates both a

4:31

biology and metals lab where it produces

4:33

its own enzymes and metal catalysts

4:36

inhouse. These are actually enzyatic

4:38

reactors. So what we do, we grow bugs,

4:40

we break the bugs open, we take the

4:42

enzymes and we put them in these

4:43

reactors and we can stress them out and

4:46

figure out what they're capable of doing

4:47

at scale. We have outfitted this with

4:50

probably some of the best uh analytics

4:52

that you can ever have for enz enzymes

4:54

which gives us a good insight into how

4:56

things will scale.

4:57

>> Across from the biology lab is Cogen's

4:59

metals lab. We just went to the the

5:01

enzyme lab. Yeah. But basically now we

5:03

pair that enzyme with the right metal

5:05

and so you can basically mix and match

5:07

which metal and which enzyme you want to

5:10

pair together.

5:10

>> Their next big insight was a commercial

5:12

one. Previous startups that had tried to

5:14

do something like this all started by

5:16

raising a huge amount of funding and

5:18

building a large scale plant. Sigen took

5:20

a different approach. They built their

5:21

first reactor for just $10,000 and then

5:24

started selling to customers almost

5:25

immediately. Gradually they scaled up to

5:28

larger and larger plants. So you did

5:30

this technoeconomic analysis and you're

5:32

like, "Wait a second. This could

5:34

actually work."

5:35

>> So May 2016, we got unofficial second

5:38

place in the MIT 100K competition.

5:40

>> We lost.

5:40

>> We got We lost uh gloriously, but we got

5:43

10,000 bucks.

5:44

>> $10,000. That doesn't seem like very

5:46

much money.

5:47

>> I think capital constraint forces very

5:49

creative thinking cuz like with 10

5:50

grand, like you have a very confined

5:52

space of what you can afford to buy to

5:54

try to make the product. And so ours

5:56

ended up being PVC like a Walmart shelf.

6:00

You know, we couldn't even afford the

6:01

metal catalyst parts. It was just the

6:02

enzyme portion.

6:05

>> This was the very first solen reactor.

6:08

>> Yep. This is the first one complete with

6:11

schedule 80 PVC from Home Depot. This is

6:13

like a bubble column with a membrane. So

6:16

you sparge air in to the bottom. Inside

6:20

there's the liquid with the corn syrup

6:22

and the enzyme and it's kind of spinning

6:24

in a loop. Then they're reacting which

6:26

makes the peroxide and then this is a

6:28

membrane and so the membrane keeps the

6:31

enzyme in the bubble column and then the

6:34

permeate on the membrane is the peroxide

6:36

product. Armed with just their $10,000

6:39

reactor but no customers, Shawn and Gorb

6:41

applied and were accepted into YC. But

6:43

they deferred a few months and set out

6:45

to first try and sell the tiny volumes

6:47

of peroxide they could make.

6:49

>> We made our first product in like

6:50

September 2016. We couldn't afford any

6:53

controls, right? This is a total manual

6:55

operation. We'd come in in the morning.

6:57

We would try to get this reactor to a

6:59

steady state. And then I'd go to work.

7:02

He was He'd go to the hospital. He was

7:03

in his last year of med school.

7:04

>> I was on surgery rotation.

7:06

>> On surgery rotation.

7:06

>> Wow.

7:07

>> Yeah. 36-hour shifts.

7:08

>> Good times. And then in the evenings, go

7:10

in and try to retune it to a steady

7:12

state. And then our first three

7:14

customers were float spa hot tub owners

7:17

in Dallas. We discovered the supply

7:19

chain dislocations because for these hot

7:21

tub owners, they were buying 3% peroxide

7:23

in the brown bottle on the store that

7:25

went through multiple distributors,

7:27

multiple downpackers, putting in little

7:29

brown bottles, shipping it to the store,

7:30

retail markup. It's like, oh, we only

7:32

have 10 grand, but like we're actually

7:34

manufacturing the chemical and we're

7:35

like bypassing like huge distribution

7:37

value chains. So on weekends, we would

7:39

put pour the chemicals in people's hot

7:40

tubs. We looked at a bunch of markets to

7:43

be like what can we just like wedge

7:45

ourselves into? When we first accepted

7:47

Soligen into YC, they had zero revenue,

7:49

but they deferred by a batch and spent 6

7:52

months getting those first customers. By

7:54

the time they started at YC, they were

7:56

gaining traction. Typically, when people

7:58

think about hard tech companies like you

8:00

guys in YC, they're like, "That makes no

8:02

sense. What could you possibly

8:03

accomplish for like a few hundred,000?"

8:05

To most people, it intuitively seems

8:07

hilariously mismatched to the like costs

8:09

and timelines of something like a new

8:11

chemical plant. This fundamentally goes

8:13

back to the customer experience, which

8:15

is what YC taught us to do, right? It

8:17

was really like for me it was like grad

8:18

school for customers is how I I look at

8:20

YC where it's like the second you have a

8:22

PhD in your customer and you're an

8:24

expert in their world, then you know

8:26

exactly what you can and can't build.

8:27

It's that simple. If you know this

8:29

customer is not the right fit, that's

8:30

okay. Go to the next customer.

8:32

>> I remember the picture of the beaker

8:34

because in your like demo video, that's

8:36

what it was.

8:37

>> Do you remember this? The blue beaker

8:38

and then we had the color change back

8:40

and forth with the metal catalyst. But

8:41

but I remember another thing about your

8:43

application, which is even though the

8:46

only thing you'd actually made was like

8:48

one beaker full of this stuff, you had

8:51

the idea completely worked out. Like you

8:53

totally understood the reason that this

8:55

would work. You'd worked out all the

8:57

technoeconomics. Like all the math was

8:59

done. The idea was fully flushed out. I

9:01

don't think the idea like the core idea

9:03

has changed one bit since the blue

9:05

beaker phase.

9:06

>> Literally the same.

9:06

>> Yeah. It's just like a million times

9:08

larger.

9:08

>> Yeah. Is this larger?

9:13

During YC, you guys are still driving

9:15

around with like buckets of peroxide and

9:18

selling them to spas. You finish YC, you

9:21

raised your $4 million seed round. What

9:23

What happens then?

9:24

>> So, we moved to where our customers

9:25

were, which was Houston, Texas, and we

9:27

used the seed capital to build our first

9:29

big pilot reactor, not just a PVC pipe

9:31

reactor, but a proper 1500gallon type

9:34

reactor.

9:35

>> And then our first oil and gas field

9:36

trial was January 2018. And so we had

9:39

billboards targeting this guy so we

9:41

could go get our first oil and gas field

9:42

truck.

9:42

>> Wait, you had billboards targeting one

9:44

guy?

9:44

>> So what we did, there was one guy um at

9:47

at the saltwater disposal company who

9:50

controlled all of the chemical spent.

9:51

What we figured out was where he

9:53

frequents like he actually goes to the

9:55

field where he goes and where he lives

9:57

or at least the neighborhood he lives

9:59

in. and we bought up the billboard and

10:01

billboards are cheap

10:03

>> along the highway that he commutes to

10:04

work in so that every day as he's

10:06

commuting to work he's just seeing your

10:07

billboards

10:08

>> and then eventually he gets a call from

10:09

us. It's a it's like a priming aspect of

10:11

it that look we knew this was the guy. I

10:14

was willing to spend 10 to 15K just to

10:16

make sure that we can get in front of

10:18

the guy. And the second that he saw that

10:20

and he got a call from us, he was like,

10:21

"Oh, I see your billboards everywhere."

10:24

And so that's when he felt special,

10:26

right? And this goes back to the

10:27

customer experience. Do you think the

10:28

CEO of DAO would do that? No. No way.

10:31

They're not going to do that. Soon, Syen

10:33

had signed enough customers that they

10:35

were able to raise money again and build

10:37

their first state-of-the-art plant, the

10:39

Bioforge 1. So, this is Bioforge.

10:42

Everything you're looking at here was

10:44

built in five locations simultaneously

10:46

and then shipped here on trucks. We

10:49

rented a crane for 4 months and just

10:51

stacked it up like Legos. So, all of the

10:53

tan tanks are filled with corn syrup.

10:56

That's our raw material. We were able to

10:58

hold four rail cars of corn syrup at any

11:00

one time, which is about 800,000 pounds.

11:03

>> So it's like 800,000 pounds of corn

11:05

syrup in these four large tan tanks.

11:08

That's the starting material for the

11:09

whole plant.

11:09

>> That's the starting material for the

11:11

whole plant. And then the plant runs

11:12

continuously 24/7. So this is the full

11:15

scale version of the PVC reactor. Is

11:18

this the bubble column?

11:19

>> This is the bubble column.

11:20

>> It's so tall.

11:21

>> 60 ft tall. It's identical to the PVC

11:25

reactor from Y Combinator. It's just

11:27

10,000 gallons instead of seven gallons.

11:29

We sparge air in the bottom, corn syrup

11:31

and enzyme go in the top and the two

11:33

react together.

11:34

>> All the other stuff is necessary, but

11:36

like this piece is the magical piece,

11:38

right?

11:38

>> This is the reactor that makes it

11:40

happen. Yep. We feed in one Coke bottle

11:43

of enzyme and you get two to four tanker

11:44

trucks of product. That's how efficient

11:46

the enzyme is.

11:47

>> Wow. After filling up the tanks, Syogen

11:50

loads up trucks to distribute the

11:52

chemicals to nearby customers.

11:54

>> So, this is the end of the line. This is

11:55

where customers pull up uh either with

11:58

their own trucks or Solent supplies

12:00

trucks and does all the logistics for

12:01

them and we're able to fill up trucks at

12:03

about 300 gallons per minute.

12:05

>> Building out lots of factories near

12:07

customers to keep shipping costs down is

12:10

a key part of how Syen has managed to

12:12

undercut its larger competitors. What

12:14

has it been like to build new like

12:17

physical stuff in America? It is

12:19

definitely challenging, but if you're in

12:22

a part of the country that wants that

12:23

manufacturing back that's favorable to

12:25

manufacturing, like it is absolutely

12:27

possible to to build in America. What

12:29

does Syogen look like in 10 more years?

12:31

>> It's actually going to be multiple

12:32

different assets. So, we're actually

12:34

like we've taken enzymes and metal

12:36

catalyst and applied them to not just

12:37

bioforges, but other types of

12:40

manufacturing assets that combine them.

12:41

And we'll be solving the most

12:43

fascinating customer problems that like

12:45

right now like we're not even aware of

12:46

them. some of the problems that we're

12:47

going to solve, they don't exist yet.

12:48

And so it's like just creating a culture

12:50

that's willing to like be wrong and

12:52

solve those problems is actually what's

12:53

most important right now.

Interactive Summary

Solugen is a chemical manufacturing company that uses a novel approach called chematic processing, which combines the specificity of biology (enzymes) with the efficiency of metal catalysts. This process allows for smaller, cleaner, safer, and more environmentally friendly chemical plants with higher yields compared to traditional methods that rely on fossil fuels. The company's journey began with a scrappy prototype made from PVC pipes and a $10,000 grant. Their breakthrough came from realizing that organic enzymes could operate at an industrial scale. They developed a process using corn syrup as a feedstock, enzymes derived from pancreatic cancer cells, and metal catalysts to produce various chemicals. Solugen's innovative customer acquisition strategy, including targeted billboards, and their focus on building manufacturing assets near customers have contributed to their success. The company has grown from a small startup to a billion-dollar company with a state-of-the-art plant, Bioforge 1, and plans to expand into new areas of manufacturing.

Suggested questions

5 ready-made prompts