Background Art The invention relates to a continuous H2 production system in which photosynthetic 02 evolution and H2 photoproduction are separated physically in two separate bioreactors, wherein the first bioreactor operates as a sulfur chemostat to generate photosynthetically competent and active cells, which are continuously fed to a second photobioreactor wherein the culture produces H2 under anaerobic conditions, does not accumulate toxic levels of fermentation products, and is continuously replenished by fresh cells from the first chemostat to prevent complete inactivation of H2-production activity and claims benefit of the 10/1/02 filing date of provisional application P60/414,993.

Technical Field Many classes of microorganisms are able to produce H2. Among them are some anaerobic and photosynthetic bacteria, cyanobacteria, and green algae.

However, sustained algal H2 photoproduction can only occur if Oz, a co-product of photosynthesis, is continuously removed from the medium by the addition of reductants or Oz scavengers, or purging with neutral gases, all costly propositions. This is necessary due to the sensitivity of the algal hydrogenase enzyme to 02. Two possible ways to circumvent the O2- inhibition problem in algae are currently being pursued. In the first, molecular biology approaches are being used to engineer an 02-tolerant hydrogenase that functions under aerobic conditions. The second approach consists of separating the 02-evolution and H2-production activities temporally in the cultures.

Sustained algal H2 photoproduction for up to 4 days is currently attained in batch cultures, where the algal cells transition from a photosynthetic, 02-evolution phase to an anaerobic, H2-production phase. The transition is achieved by depriving the cultures of sulfate.

This is done by rigorously washing the cells 1-5 times in minus-sulfur medium, and then re- suspending them in the same medium. Under sulfur-deprived conditions, photosynthetic O2- evolution activity is partially (but reversibly) inhibited, and the cultures become anaerobic due

to the continuous respiration of accumulated starch and/or other endogenous substrates.

Anaerobiosis is necessary and sufficient to induce the synthesis of a reversible hydrogenase enzyme, which catalyses H2 production, using reductants originated mostly, but not entirely, from residual photosynthetic-water splitting function. In the batch mode, H2 photoproduction is transient, lasting from 3-4 days, since sulfur deprivation eventually affects all other cellular activities (besides photosynthetic Oz evolution). However, algal cultures can be rejuvenated, at that point, by re-adding sulfate to the medium for a 2-day period.

At the present, the batch sulfur-deprived algal system is not commercially viable due to its high cost. Some of the reasons for this are: a) the low specific rates of H2 photoproduction; b) the short period of time during which cultures actually produce H2 ; and c) the expensive centrifugation step required to deprive the culture of sulfate.

In T. V. Laurinavichene et al. Different Methods To Deprive Clalamydomonas reifahardtii Cultures Of Sulfur For Subsequent Hydrogen Photoproduction. J. Int. Hydro. Energy, in press.

(2002), it is demonstrated that it is possible to deprive the cultures of sulfate without relying on centrifugation.

Kosourov et al. in Sustained Hydrogen Photoproduction By Chlamydomonas reinhardtii : Effects of culture parameters. Biotechnol Bioeng 78: 731-740. (2002), disclose that the re-addition of small amounts of sulfate to the cultures at the start of sulfur- deprivation resulted in higher specific rates and yields of H2 photoproduction.

Summary One object of the present invention is to provide a continuous H2-production system based on the effects of sulfur-deprivation on algal photosynthesis and physical separation of the net Oz and H2 production processes in algal cultures.

Another object of the present invention is to provide a continuous H2-production system based on the effects of sulfur deprivation on algal photosynthesis and physical separation of the net O2 and H2 production, wherein the yield of the continuous system is comparable or better than that of the batch system with substantially lower cost due to the elimination of many centrifugation steps.

A further object of the present invention is to provide a continuous H2-production system based on the effects of sulfur-deprivation on algal photosynthesis and physical separation of the

net °2 and H2 production, wherein the algal H2 production is stable for at least 14 days and is maintained at a stable rate of about 4.7 jumols H2 mg Chl-l h-l.

In general, the invention is accomplished by: a) starting the two-bioreactor system and initiating hydrogen photoproduction in the second vessel without centrifuging ; b) running both vessels under conditions where just enough healthy cells are continuously delivered to the second vessel to support maximal hydrogen production; c) circulating fluids to recycle as much liquid as possible; d) obtaining valuable by-products by removing and recovering cells from the second vessel and separating compounds from the spent medium; e) removing build-up materials in the return fluid that negatively affect cell growth in the first vessel and subsequent H2 production in the second vessel; recycling acetate produced in the second vessel back to the first vessel where acetate is required to produce cells capable of evolving hydrogen in the second vessel; g) transferring cells from the first to the second vessel in a manner that does not carry over oxygen; and h) delivering the cells to the second vessel in a state that allows them to immediately produce hydrogen without the delay normally present in current processes.

Brief Description of the Drawings FIG. 1 is a drawing depicting the concept of the two-bioreactor, continuous flow system for algal H2 production of the invention.

FIG. 2 is an example of a schematic of the computerized photobioreactor system of the invention for continuous cultivation of algae under sulfur-limiting conditions and for H2 photoproduction.

FIG. 3 is a schematic showing the principles of operation of the double chemostat system of the invention, inclusive of D1 and D2 dilution rates for photobioreactor 1 (PhBRl) and photobioreactor 2 (PhBR2).

FIG. 4 is a graph showing biophysical and biochemical parameters associated with establishment of steady-state growth in PhBR1 under continuous dilution with TAP buffer containing 75RM of sulfate.

FIG. 5 is a graph showing biophysical and biochemical parameters measured in PhBR2.

The experiment was started in the batch mode. The chemostat mode began at the time indicated by the arrow, as cells from PhBRl were used to replace those in PhBR2.

FIG. 6 is a graph showing single turnover, flash-induced fluorescence yields of samples taken from the two photobioreactors after steady-state growth is established. The highest fluorescence yield, Flax, is a measure of the PSII capacity of the cultures.

FIG. 7 is a graph showing comparison between the rates of H2 production by cells grown under batch and continuous (chemostat) modes.

FIG. 8 is a graph showing the amount of H2 gas collected and changes in the rate of H2 gas collected and changes in the rate of H2 photoproduction during the sulfur-deprivation process in the batch mode. The maximal rate of H2 production is equivalent to 9.4 immoles H2/(mg/Chl~l lx h-l), but this is not sustainable.

Materials and Methods Cell Growth Conditions Wild type Chlamydomoaas reirahardtii, strain cc124 (mt), was pre-grown photoheterotrophically in batch cultures on a standard Tris-Acetate-Phosphate (TAP) medium in 1-L Erlenmeyer flasks at about 25°C ; however, other H2-producing, oxygenic photosynthetic organisms can also be used.

The culture volume was about 800 mL, and algal cultures were bubbled with air containing 2% CO2. The gas mixture was sterilized using membrane filters with a 0. 2 uM pore size (Acro 37 TF, Gelman Sciences, Inc., Ann Arbor, MI). During growth, algae were continuously illuminated with cool-white fluorescent light of about 200 µE # m-2 # s-1 PAR and mixed by magnetic stirring (PC-131, Coming Inc., NY). After reaching the late logarithmic phase (4-6 x 106 cells mL-1), cells were harvested by centrifugation at 3000 x g for 5 minutes.

Harvested cells were washed once in TAP-minus-sulfur medium and resuspended in a small volume of the same medium. The cultures could also have been grown under limiting sulfate conditions to eliminate the washing step. This culture material was used to inoculate two bioreactors: the first bioreactor for chemostatic culture growth and the second for H2 photoproduction.

TAP-minus-sulfur medium was modified from a standard TAP medium by

replacing all sulfates with chloride salts at the same concentrations. All procedures with algal cultures mentioned above (i. e. washing and transfer from one flask to another) were performed under sterile conditions.

Continuous Flow Photobioreactor System FIG. 1 is a drawing showing the concept for the two reactor, continuous flow system for algal H2 production.

A schematic of the computerized photobioreactor system designed for continuous cultivation of algae under sulfur-limiting conditions and for H2 photoproduction is shown in FIG. 2. The system employs two (1,2) specially fabricated, glass photobioreactors (4-cm optical path length, 1.1-L culture volume) with built-in ports (3) for dissolved °2 (pO2), redox (Eh) and pH plus temperature (pH-T) sensors. Each photobioreactor has three inlets for fresh medium (4) and titrants (5-base, 6-acid) and one outlet for gas and cell suspension (7).

The first photobioreactor has an additional inlet for bubbling the algal culture with a sterile COz/air gas mixture (8). Excess cell suspension in the first bioreactor, if any, together with any produced gas are disposed of into a collector flask (9). Alternatively, at the example dilution rates used in this example, the excess material could be used to supply a second H2- producing photobioreactor. The first photobioreactor (1) supplies the second photobioreactor (2) with the cell biomass and sulfate-depleted medium (12, top arrow).

The gas produced by the second bioreactor is collected in an upside-down graduated cylinder filled with water (10). At the same time, excess cell suspension from the second bioreactor is disposed of in a second collector flask (11). In a commercial system, this material would be recycled back to the first photobioreactor. The flow of all liquids in the system is maintained by pumping through Tygon LFL tubing (0.89 mm i. d. , 96429-26, Cole-Parmer Instrument Co, Illinois) with six peristaltic pumps (12) (MasterFlex C/L 77120-70, Cole-Parmer Instrument Co, Illinois) operating under the control of a microprocessor system. Finally, the microprocessor system (13,14), consisting of two computers, is used for data storage and processing. The first computer (13), consisting of four PCA-6135/L microprocessor motherboards and an IPC-620 chassis (Advantech, Sunnyvale, CA) processes and stores the data obtained by the sensors, controls the dilution rates in both photobioreactors independently, and holds the pHs inside bioreactors at a given value. The second computer (14) (a P-II, 400 MHz, Toshiba notebook) is used for calibrating the sensors, changing the operating parameters, and

on-line monitoring of the growth conditions inside the photobioreactors. Both computers communicate via the RS-232 port.

PhBRl operates as a classical chemostat under sulfate-limiting conditions and supplies PhBR2 with the healthy cell biomass (FIG. 3). The cell density in the first bioreactor will depend on the sulfate level in the medium and the dilution rate, Dl. The purpose of PhBR2 is to produce H2 continuously. The algal cells in this bioreactor experience severe sulfur-deprivation conditions due to the constant uptake of sulfate in the first bioreactor by the growing culture, and due to the lower rate of culture flow between the two bioreactors, D2 (where D > D2). Under severe sulfur deprivation, algal cells partially inactivate their 02-evolving activity and, as a result, make their environment anaerobic enough to allow H2 to be produced. The establishment of such an anaerobic environment inside the second photobioreactor is necessary for induction of the hydrogenase and, hence, for H2 photoproduction. The algal culture does not grow under these conditions. The cell density in the second photobioreactor will depend on the cell density in the first bioreactor, the rate of culture flow between the two bioreactors and out of PhBR2 (D2), and the rate of cell death under sulfur deprivation. The system in this configuration allows for the continuous cultivation of healthy algal cells in PhBRl and constant H2 production in PhBR2.

Procedures The experiment begins with a period of pre-cultivation of the algal culture in the first photobioreactor under sulfur-limiting conditions. For this, PhBRI, containing 900-1000 mL of TAP-minus-sulfur medium with 75 tM sulfate, was inoculated with 100-200 ml of ChlamyeloTnonas reinhardtii culture to a final chlorophyll concentration of about 5-10, ug/mL.

The inoculum used was prepared as described above (see"Cell Growth Conditions"Section).

The photobioreactor was placed under two-sided illumination of # 200 - 220 µE # m-2 # s-1 PAR (eight 40-W cool-white fluorescence lamps), connected to input and output tubing and cultivation was started by adjusting the dilution rate to 0.04 h-1 (DI). The bioreactor was fed with TAP-minus-sulfur medium containing 75 uM of sulfate. The algal culture inside the photobioreactor was mixed by magnetic stirring and bubbled with 3% C02 in air. The gas mixture was sterilized by passing it through membrane filters. Once the culture in the first bioreactor was close to steady-state, cultivation of algal cells in PhBR2 was started. The photobioreactor was filled with 1.1 L of cell suspension in TAP-minus-sulfur medium and placed under the same light and mixing conditions as the first photobioreactor.

The algae in the second bioreactor were cultivated initially in the batch mode. When the algal culture in this bioreactor began producing H2, PhBR2 was connected to PhBRl and the liquid flow between the two bioreactors was adjusted to a dilution rate of 0.02 h-1 (D2). Under our example conditions, the transit time for the culture to travel from PhBRl to PhBR2 was 0.3 min. In this mode, the photobioreactor system was operated for a prolonged period (up to 14 days). The pHs in both bioreactors were kept constant by automated addition of 0.2 M NaOH or 0.2 M acetic acid at, respectively, 7.5 and 7.8.

Flash-Probe Chl a Fluorescence The fluorescence yield of Chl a following a saturating single-turnover light flash reflects changes in the level of QA-, the reduced form of the primary stable electron acceptor in PSII.

The amount of QA-, generated under our experimental conditions, is a measure of the photochemical capacity of PSII in the cultures and is related to Fmax (see below). Flash-induced Chl a fluorescence yields were measured with a home-built instrument. Samples used for fluorescence measurements were taken directly from each PhBR, concentrated by a centrifugation to get a total chlorophyll concentration of 40 1 g/ml and adapted in the dark for 5 minutes under atmospheric 02 partial pressure. The samples were placed in a 4 mm x 10 mm plastic cuvette, and DCMU was added to a final concentration of 30 juM to block an electron transfer between QA-and QB (the secondary electron acceptor) on the reducing side of photosystem II. Fluorescence was initiated by a saturating actinic flash and the corresponding fluorescence yield was measured as (F-Fo) lFo. The initial maximum fluorescence level, Fax, appears within 20-40 jn. s following the saturating flash and decays to a low final level, F,, o/ within 3 s.

The chlorophyll a + b content was assayed spectrophotometrically in 95% ethanol extracts by the method of Spreitzer (Harris, 1989).'The cell density was measured indirectly by the absorbance of the cell suspension at 750 nm and related to the corresponding chlorophyll concentration.

Results As observed, it is possible to deprive algal cells of sulfate without using costly centrifugation. This can be done by diluting sulfur-replete cultures either in sulfur-minus medium or in sulfur-limiting medium. In the experiments described here, however, we utilized I E. H. Harris The Chlamydomonas sourcebook: a comprehensive guide to biology and laboratory use. Academic Press, San Diego (1989).

centrifugation to carefully wash out sulfate and re-suspended the cultures in medium of defined sulfate concentration (75 VLM in PhBRl), in order to better examine the experimental parameters.

The algal cultures in the first photobioreactor were grown chemostatically under constant dilution with medium containing 75 gM sulfate, at a rate of 0.04 h-1. These conditions resulted in a rapid transition of the culture to an anaerobic environment (where the rate of photosynthetic 02 evolution-rate of respiratory Oz consumption in the cells), as monitored by changes in redox potential and dissolved 02 content of the culture medium (FIG. 4, PO2 and Eh). The redox potential of the medium decreased to about-250 mV (measured against an Ag/AgCl reference electrode) and then equilibrated at about-150 mV. No dissolved 02 was detected in solution during the rest of the incubation period. The pH of the culture was kept constant at 7.5 by automatically titrating in 0.2 M acetic acid and 0.2M NaOH. That pH value has been shown to be optimal for photosynthetic growth of the cells. After the transition to anaerobiosis, the growth rate of the algal culture decreased and eventually reached a steady-state level (FIG. 4, OD75o and Chl), at a stable chlorophyll concentration of about 10 llg/mL. When the cultures reached steady state, at t = 113 h, they were pumped into the second photobioreactor at a constant dilution rate of 0.02 h-1.

The second photobioreactor was started in the batch mode. As seen in FIG. 5, under complete sulfur deprivation, the algal culture quickly becomes anaerobic. The redox potential (Eh) reached a significantly more negative value (-450 to-500 mV) as compared to PhBRl (where the steady-state Eh--100 mV and no significant H2 production was observed), and then climbed steadily to about-350 mV after H2 production started. After 24 h of H2 photoproduction in PhBR2, the flow of fresh cells from the first photobioreactor was initiated.

In our system, tubing with a total volume of 1 ml connects the two photobioreactors, delivering cells from PhBRl to PhBR2 in about 0.3 minutes. The chlorophyll concentration reached steady-state (about 8-9 ug/mL) at t = 100 h, and the pH of the medium was kept at 7.8 by titration with 0.2 M NaOH. After an initial decrease in the H2-production rate, the culture in the second bioreactor reached a steady-state rate of H2 production of 4. 7, umoles/(mg Chl h), compared to 12. 1 umoles/(mg Chl h) initial rates in the batch mode. The former rate was sustainable for the remainder of the experiment (another 220 h, and 600 ml of H2 gas was produced. The Eh in PhBR2 did not decrease below the steady-state level after 2 days of H2 production. A decrease in Eh is characteristic of decreased H2-production rates and cell death in batch cultures (data not shown). Upon optimization, both higher H2 production rates and longer

durations of H2 production can be expected. This will involve adjusting the dilution rates, sulfate concentration in PhBRl, and transfer times between PhBRl and PhBR2. The photosynthetic capacity of the cultures in both reactors was assessed by means of flash-induced fluorescence yield measurements. FIG. 6 shows that the photosynthetic photochemical capacity of the cultures in PhBRl (Fmax = 0.6), measured after steady-state conditions were established, is equal to that of algal cultures grown under fully aerobic, sulfur-replete, photosynthetic conditions. Moreover, the culture in PhBR2 maintains photosynthetic activity at a value (Fmax = 0.2) that corresponds to or is slightly lower than the level measured in the batch culture during the period of peak H2 production.

To improve our current batch algal H2-production system, we devised a process based on two sequential chemostats where (a) centrifugation of the cultures to deprive them of sulfate after start-up was eliminated, (b) fresh, photosynthetically-competent cells (PhBRl) continuously replace dying, H2-producing organisms (PhBR2), (c) circulating fluids can be recycled, (d) valuable co-products can be generated, (e) the accumulation of inhibitors of growth and/or H2 production is prevented, (f) acetate, required for rapid decrease of the 02-production function and subsequent anaerobiosis in PhBR2, can be recycled, (g) oxygen is eliminated by continuous respiratory activity of the cells, and (h) the cells delivered to the H2-producing reactor maintain the culture producing H2 continuously.

The first reduction to practice led to two important observations. The first was that the continuous addition of 75 x sulfate to the first chemostat under our dilution rates resulted in a steady-state Chl concentration and no detectable 02 inPhBRl, and a continuous H2 photoproduction in PhBR2. This was surprising since a one-time addition of even 1, uM sulfate directly to a H2-producing bioreactor during the H2-production phase in batch cultures is known to significantly inhibit H2 production. In our concept of a continuous system, we thought that aerobiosis in the culture would be required to maintain the cells photosynthetically active in PhBRl. This prediction was based on experiments in our lab, showing that 02-evolution capacity of algal cells is inactivated by incubation under dark, anaerobic conditions even in the presence of sulfate. However, as seen in FIG. 6, the photosynthetic capacity of PSII is very high inPhBRl, comparable to that of photoheterotrophically-grown cells at normal sulfate concentrations. We also tried other limiting sulfate concentrations in PhBRl at the previous dilution rates, with negative results. Below 50 uM sulfate, the culture in PhBRl reached a much lower steady-state Chl concentration, and was not able to maintain high rates of H2 production

by PhBR2. At 90 pu sulfate or above, the steady-state redox potential in PhBRl was too high and the transport of the culture to PhBR2 actually inhibited H2 production. Transit times between PhBRl and PhBR2 of >> 0.3 min. may be necessary at these higher sulfate concentrations in PhBRl to maintain H2 production in PhBR2.

It is important to note that the steady-state Chl concentration in the first chemostat depends both on the amount of sulfate with which it is continuously diluted and on the dilution rate utilized. Initial experiments at dilution rates of 0.02, 0.03 and 0.04 h-l led to the choice of the latter, but this was done at 75 uM sulfate, and this example may not be optimal yet from the process perspective. Additionally, the optimal sulfate concentration in PhBRl to support H2 production by the cultures in PhBR2 will vary depending on the time that it takes to deliver the cells to PhBR2, a factor that will determine the overall activity of the delivered cells.

The second important observation resulting from our experiments is that the total yield of H2 using the continuous system is higher than the corresponding average amount using the batch system, as shown in FIG. 7. We point out that the results for the batch system in FIG. 7 includes a 2-day"down period"used to rejuvenate the culture. This is done by adding sulfate back to the cultures, allowing photosynthetic 02 evolution activity to recover, and submitting the cultures to a second round of sulfur deprivation by centrifugation. The yield of the continuous system compares quite favorably with the corresponding yield of the batch system, without the accompanying disadvantages of the latter's high cost of operation. Moreover, the total Ha yield from the continuous system is actually double that shown in FIG. 7, because, under the described operating procedures (Dl = 2 x D2, FIG. 2), each PhBRl can be used to run two PhBR2s. As can be seen, specific rates of H2 production over the time of the experiments are: 4.7 gmoles H2/ (mg Chl x h) -continuous mode; 3 jumoles H2/ (mg Chl x h) -batch mode.

Even though the current rate of continuous H2 photoproduction is high, our aim is to achieve the highest possible rate. This target is shown in FIG. 8, where the rate of H2 photoproduction is plotted as a function of time under sulfur deprivation in a medium of pH = 7.7 using A TYPICAL"batch culture"experiment. The rates initially increase, as the medium becomes anaerobic, reach a maximum value, and promptly decrease under the batch growth conditions. At its peak, this culture produced H2 at 4 ml/h (or 9. 4 umoles. mg Chl-l. h-l), which represents close to the upper limit for the reaction under these particular experimental

conditions. These maximum specific rates are about twice as high as that of the continuous system and represent our goal for H2-production rates in the latter.

In summarization, we have developed a continuous H2-production system based on the effects of sulfur-deprivation on algal photosynthesis and physical separation of net 02 and H2 production in the cultures. The yield of the continuous system is comparable to or better than that of the batch system with substantially lower cost due to the elimination of many centrifugation steps. Algal H2 production was shown to be stable for at least 14 days and to be maintained at a stable rate of about 4.7 µmoles H2 # mg Chl-1 # h-1.