Hydrogen Production by Means of an Artificial Bacterial Algal Symbiosis

Report on Experiments in the Sahara

Ingo Rechenberg

Technische Universität Berlin
Bionik und Evolutionstechnik
Ackerstraße 71-76
D-13355 Berlin, Germany




The idea of biophotolysis of water is based on the integrated metabolism between vegetative cells and heterocysts in blue algae, where hydrogen is required for nitrogen fixation. Nostoc muscorum is the biological model to design a two-stage photo-bioreactor.

Since 1987 details of the artificial bacterial algal symbiosis will be tested in Erg Chebbi, a dune region at the edge of the Sahara. The main results are: The rate of hydrogen production is considerably increased if one makes use of the reflected radiation from the dunes. Fluorescent laser dyes further amplify the hydrogen formation. Cooling of the bioreactor is done by passing the heat through a cooling tube into the depth of the dune. – The best result: one liter bacteria suspension will produce two liters of hydrogen per Sahara day.


Origin of the Erg-Chebbi-Project

Erg Chebbi is a small sand desert at the edge of the Sahara in South Morocco. I came for the first time 1982 to this area to collect purple bacteria. My friends and I look for hydrogen producing purple bacteria over the whole world to this day. Our test-strains come from Algeria, Antarctica, the Caribbean, Ghana, Israel, Japan, Mexico, Mauritius, Nigeria, Spitsbergen et cetera. Bacteria screening is made to get strains which are more efficient, robust, and easier to handle in the technical terms. The hydrogen is produced under the light of bulbs. Now, under the singing sun of the Sahara I surrendered to me the question: would the purple bacteria produce hydrogen also under these extreme conditions? The project was born. The desert named Erg Chebbi became the field-laboratory to test hydrogen producing photo-bioreactors.


Experiments at the Planning Stage

The task was posed. An apparatus must be developed for the Sahara. But how should the bacteria stand the heat? The use of an electrical driven thermostat was unrealistic. My idea was, to carry away the heat into the depth of the dunes. In autumn 1987 I checked this idea using a dummy reactor filled with red colored water. I stayed alone 10 days in the desert. When the sun was in the zenith I measured 60 degrees Celsius on the surface of the dunes but 30 degrees in the depth of 1 meter. The maximum temperature of the dummy bacteria suspension was 54 degrees Celsius. Due to the low heat conductivity of the sahara sand I left the idea of ground cooling. (In sommer 1994 I realized the principle of ground cooling with great success).

Next year I drived with my expedition car to the same experimental site in the desert. I had a small 140 ml reactor with me, which was cooled by a solar driven Peltier-element. Because it got broken I changed to simple evaporation cooling. This nomadic cooling principle worked excellent. Within 16 days I could win 0.6 liter bio-hydrogen in the Sahara.


Advanced experiments in the Sahara

1989 I was at my experimental site in spring and in autumn. The new bioreactor design had a professional evaporation cooler. The low relative humidity of the air in the desert makes this cooling principle very effective. The two reactor tanks with a total volume of 1.4 liter looked like a solar panel which could be aligned to the sun. But the result was disappointing. The amount of hydrogen production dropped down from hour to hour. It was supposed that the purple bacteria were damaged by the strong radiation. Therefore the reactor panel was directed to the dunes and the reflected light proved to be much more effective (Fig. 1).


    Fig. 1:

    Reflected light from the dunes gives the best solar efficiency.


Project ArBAS

The experiments ran quite successfully, but there is one very important point of critique. Purple bacteria require an organic substrate for hydrogen production. In the experiments presented lactate was used. The big goal is to split water into hydrogen and oxygen. All green plants do this. But the hydrogen becomes never existent. Plants do not produce molecular hydrogen with one exception: for nitrogen fixation. The blue alga Nostoc muscorum is my biological model to design a bioreactor which splits the water (Fig. 2). The vegetative cell of Nostoc corresponds with the flask containing green algae. Here the hydrogen of the water will be built in carbohydrates. The carbohydrates are transported through a special membrane into the heterocyst, which corresponds with the flask containing purple bacteria. Here the carbohydrates are broken into CO2 and H2 by means of nitrogenase. The process in the blue alga Nostoc as well as in the technical reactor is driven by the energy of light. It must be added: the process in the bioreactor works only, if the green algae excrete carbohydrates. In the literature algal strains are mentioned that excrete up to 50% of the total photosynthetic production into the surrounding. I have to point out the role of CO2 . It works as a carrier molecule which is loaded and unloaded with hydrogen.


Fig. 2: Cell linkage of Nostoc muscorum (a) and composed bioreactor (b)

I formulated this idea 1981 in a German article. Others may have done this before. I make propaganda for this trivial process because biological evolution has not found a more intelligent method to get hold of hydrogen. Preservation and transportation of hydrogen in carbohydrates might be not a bad idea. Since the birth of the idea we tried to realize the Artificial Bacterial Algal Symbiosis (ArBAS). In 1989 some of my students - supervised by Gerald Koch - could produce the first hydrogen from a bacterial algal linkage in batch operation. The alga Chlamydomonas oblonga (known to emit carbohydrates) growes under strong light and 3% CO2 in the inlet gas. After 200 hours the culture was inoculated with purple bacteria (Fig. 3). In this experiment algae and bacteria work in one tank at different time. Later we were able to do the same with the filtrated carbohydrates, excreted by C. oblonga. — It must be said: there is no hope that an ordinary algal culture, inoculated with purple bacteria, will produce hydrogen. We wasted so much time between 1981 and 1989, because we didn’t know the two tricks. First: all effort has to be done to get an exceptional high carbon/nitrogen ratio in the growing algal culture. Second: one must be very carful that not parasitic bacteria will consume the excreted carbohydrates.



Fig. 3:

Biophotolysis of water by a bacterial algal linkage.



Algae and bacteria in Erg Chebbi

The success with the bacterial algal composition encouraged me to repeat the experiment at my lonely desert place. The flat reactor vessels contained the algae while the purple bacteria grow in a cylindrical glas . This columnar reactor exploits the radiation from the sun and from the dunes. I had a small laboratory with me: algae, purple bacteria and nutrient solution in the solar driven refrigerator, small gas cylinders with CO2, pH-meter, photometer, solarimeter, distilled water et cetera. I worked hard and the experiment ran well. The algae became dark green with a small yellow tinge at the end. There seemed to be no difference to the laboratory culture. But the triumph to have achieved the first biological water-split in the sahara remained denied to me. After inoculation with purple bacteria no smallest hydrogen bubble could be detected.

I think that as a result of the day and night cycle parasitic bacteria had enough time to consume the produced carbohydrates so that the carbon/nitrogen ratio did not exceed the required threshold.


Laser dyes bioreactors in the Sahara

The frustration about the unsuccessful algae experiment pushed the test with the purple bacteria into the background. Just here a breakthrough happened. In the column-shaped reactor tank the lactat feeded Meski-purple-bacteria produced liters for liters of hydrogen. Meski is the name of our best purple bacteria strain. I isolated this strain in a small Moroccan oasis named Meski (preserved as Rhodobacter sphaeroides DSM 9483 at the German Collection of Microorganisms, D-38124 Braunschweig). It was the second experiment which became so very successful. For this experiment I covered the reactor on the sunlit side with an orange color foil. The first test with the unprotected reactor in the solar radiation was a failure.

The bioreactor design 1990 was made up of two glas flasks with transparent cooling jackets, which I didn´t use for cooling. To study the filter effect I filled the outer ring-volume with divers colored liquids. In Fig. 4 a laser dye filter is running against a grey filter. Both have the same energy absorption. The optimum laser dye has an absorption range from 420 to 520 nanometer, which corresponds with the absorption range of the carotenoids. These red and yellow pigments, that is known, protect cells from damaging sun rays. In addition to this the hydrogen production may be enhanced, because the laser dye transforms the absorbed wavelengths into longer ones, which are more effective in photosynthesis.


Fig. 4:

Step up of hydrogen evolution by a laser dye filter.

Color filter and grey filter have the same energy absorption.



A futuristic vision

Between 1988 and 1990 I produced 124.31 liters hydrogen in the Sahara. This is energetically equivalent to 0.04 l gasoline. Only a super economical car would reach with 0.04 l gasoline the oasis Hassi Labiad in 8 km distance. How can the bioreactor be enlarged? For an art exhibition 1986 I designed a purple bacteria bioreactor in the form of a cone. The exhibited object was named "heliomites in the Sahara". The three heliomites with 300 l volume ran under the floodlights for several weeks. The discovery that the reflected radiation from the dunes is so effective gives rise to the idea to build heliomites for the real Sahara. The design in Fig. 5 has a volume of 1500 liters. With a production rate of 190 ml H2 per liter bacteria suspension and hour (average production rate of the reactor ‘90) one heliomite produces a power of 1 kilowatt. The energy farm below with 100 heliomites would produce a power of 100 kilowatts.



Fig. 5:

Heliomites in the Sahara.

Futuristic view of a 100 kW biological hydrogen farm.



There is no doubt that hydrogen production by an algal bacterial copound is uneconomic today. But intensive research should increase the efficiency definitely. Algae and bacteria utilize different wavelengths of the spectrum. They can be arranged one upon the other without stealing each other the light. Further I see a big chance to improve the photobiological efficiency, which is so low (2%) because the light intensity is to high at the surface and to low in the depth of the reactor. Sun radiation has to be scattered uniformly into the reactor volume. Wavelength transformators should also increase the efficiency. And the world is changing. 50 years later we look with other eyes to the sketch of a biological hydrogen farm.

It is the thought of bionics to use artificial systems in place of the evolved originals in nature. For a future biophotolysis of the water I could imagine that hydrogen will be binded to a carbohydrate-analogue in a first stage. The hydrogen-complex, unable to re-react with oxygen, then moves to a second stage, where the hydrogen will be released. The blank carrier molecule moves back to the first stage, where it is reloaded with hydrogen. — But a solution à la bionics is not in sight. At present the substitution of the process by biological components should be permitted.



Fogg GE (1966). Extracellular products of algae. Oceanogr Mar Biol Ann Rev 4: 195-212.

Klemme JH (1991). Wasserstoffproduktion mit Mikroorganismen. Naturwissenschaftliche Rundschau 44: 52-58.

Koch-Schwessinger G, Rechenberg I (1990). Photoproduction of hydrogen by means of an algal bacterial symbiosis. 2nd Joint Schlesinger Seminar on Energy and Environment. Technische Universität Berlin: 97-110.

Koch-Schwessinger G (1993). Photobiologische Wasserstoffproduktion durch Purpurbakterien und Biophotolyse [Dissertation]. Technische Universität Berlin, Fachgebiet Bionik und Evolutionstechnik.

Koch-Schwessinger G (1993). Aspekte zur photobiologischen H2-Produktion im technischen Maßstab. In: Reiß T, Hüsing B, eds. Biologische Wasserstoffgewinnung; Schriftenreihe Zukunft der Technik. Köln: TÜV Rheinland: 309-336.

Miyake J, Kawamura S (1987). Efficiency of light energy conversion to hydrogen by the photosynthetic bacterium Rhodobacter sphaeroides. Int J Hydrogen Energy 12 (3): 147-149.

Oelze J (1994). Biologische Wasserstoff-Produktion. Erschließung einer alternativen Energiequelle? Biologie in unserer Zeit 24 (3): 131-138.

Rechenberg I (1981/1987). Wasserstofferzeugung mit Purpurbakterien. Wissenschaftsmagazin TU-Berlin (1): 36-43 / Schweizerische Laboratoriums-Zeitschrift 44 (11): 417-425.

Rechenberg I (1994). Photobiologische Wasserstoffproduktion in der Sahara. Werkstatt Bionik und Evolutionstechnik Band 2. Stuttgart: Frommann Holzboog.

Watanabe Y, Noüe J de la, Hall DO (1995). Photosynthetic performance of a helical tubular photobioreactor incorporating the cyanobacterium Spirulina platensis. Biotechnol Bioing 47: 261-269.

Watanabe Y, Hall DO (1995). Photosynthetic production of the filamentous cyanobacterium Spirulina platensis in a cone-shaped helical tubular photobioreactor. Appl Microbiol Biotechnol 43.