**3. O2 Removal**

The use of inert gas (such as N2 or Ar) is another type of strategy to remove the O2 in microalgal cultures [47,48]. Alternatively, O2 scavengers can be employed to remove the O2 in order to induce anaerobiosis in the culture. Paramesh and Chandrasekhar screened three O2 scavengers individually in order to improve H2 production in *Chlorococcum minutum* [49]. In the presence of all three O2 scavengers, efficient H2 generation was found. They found that sodium sulfite was the best one for enhancement of H2 production. Nagy et al. showed that the simultaneous addition of glucose, glucose oxidase, and ascorbate to the *C. reinhardtii* culture resulted in reduced O2 content in the headspace and tenfold-increased H2 production [30]. Su et al. created an O2-consuming sandwich-like layer by using tannic acid, polydopamine, and laccase, in order to generate anaerobiosis around the *Chlorella pyrenoidosa* cells [50]. This layer enabled the encapsulated cell to switch from O2 production to H2 production. Márquez-Reyes et al. found that the chemical reducing agent cysteine induced anaerobic H2 production in cultures of *Chlamydomonas gloeopara* and *Scenedesmus obliquus* cultures [51]. In the presence of cysteine, H2 production was 5 times higher compared to the sulfur-starvation protocol. Chen and coworkers found that *C. reinhardtii* produce H2 at a rate of 0.44 μmol H2/h/mgchl per month by using a chemoenzymatic cascade system (CEC). The CEC system contained four components: glucose oxidase, catalase, glucose, and Mg(OH)2. In this CEC, they combined O2 consumption, cell aggregation, and pH maintenance to activate hydrogenase [52]. Nagy et al. showed that the application of an iron-based O2 absorbent (O20TM) in *C. reinhardtii* cultures, in which the activation of the Calvin–Benson–Bassham cycle in the light was prevented, presented a H2 production yield of 2.58 mL/L/h, to which corresponded a mean LCE (light to H2) of 0.27% [30].

#### **4. Co-Cultures**

Another approach to create an anaerobic environment is the addition of living aerobic bacteria to the microalgae cultures (co-culture) [53]. Many works have proven the possibility of increasing H2 production by co-culturing microalgae and bacteria [28,54–57]. The main advantage of co-culturing microalgae with heterotrophic bacteria is the efficient removal of the O2 from the growth media. Simultaneously, the CO2 released during bacterial fermentation of an organic substrate can support microalgae growth. Moreover, many metabolites can be exchanged between these microorganisms, such as carbon, nitrogen, phosphorous, and sulfur sources, and vitamins [58,59]. The presence of bacteria inside the microalgal culture enhances starch accumulation [60]. Different *Chlamydomonas* WT co-cultures incubated in sulfur depleted TAP medium employing *Pseudomonas* sp. or *Bradirizhobium japonicum* have achieved high H2 production rates (165–170 mL H2/L) [54,60]. Fakhimi et al. evaluated H2 production by *C. reinhardtii* in cocultures with different bacteria strains [28]. They found that co-culturing *Pseudomonas* spp. with *Chlamydomonas* significantly improved microalgal H2 production. Interestingly, the integration of the photobiological and the fermentative H2 production in *Chlamydomonas* and *Escherichia coli* co-cultures resulted in H2 production 60% higher than the sum of the respective monocultures [28]. *Chlamydomonas* co-cultures with *Pseudomonas* sp. and *Bradyrhizobium japonicum* (not H2-producing bacteria) in sulfur-depleted TAP medium improved H2 production by 22.7 times and 32.3 times compared to the pure microalgal cultures, respectively [57]. Furthermore, the production of H2 by *C. reinhardtii* in nutrientreplete cultures is strongly limited by the O2 release, unless it is performed under very low

light irradiance (lower than 20 μmol photons/m2/s), but it may become feasible under higher light irradiance by using different consortia, which allow the maintenance of anaerobiosis conditions, thus creating an opportunity to use full medium and much higher light irradiance, enhancing the H2 output. Finally, it must be pointed out that large-scale production of H2 with *Chlamydomonas* and other microalgae will be necessarily carried out not with axenic cultures, but rather a microalgae–bacteria consortium, therefore understanding the complex interplay between microalgae H2 producers and bacteria is important for the economic exploitation of an industrial H2 production process.

#### **5. Immobilization**

Microalgae immobilization can increase the H2 production yield [61]. The main reason for their higher performance is that such experiments are usually carried out by using a much higher chlorophyll concentration compared to liquid cultures. The process of switching between oxygenic photosynthesis (aerobiosis) and H2 production (anaerobiosis) can be facilitated by using cell immobilization systems [62]. One of the most used materials for microalgae encapsulation is calcium alginate [63]. Immobilization in calcium alginate matrix allows high cell density and protection from mechanical stress and contamination and is easy to scale-up [64]. Immobilization of microalgal cells could increase their LCE [62]. Using *Calothrix*, *Anabaena,* and *Chlamydomonas* cells immobilized on thin calcium alginate films gave an LCE of 2.5% of the photosynthetically active radiation [48,65]. *C. reinhardtii* immobilized on calcium alginate films in nutrient-depleted cultures (-P,-S) presented a H2 production rate of 12.5 μmol/mgChl/h [66]. Ruiz-Marin et al. proposed immobilization of *Chlorella vulgaris* and *Scenedesmus obliquus* cells in calcium alginate for the production of H2 [11]. These microalgae were grown in urban wastewater under sulfur starvation and blue or purple light conditions. The maximum H2 production obtained under red light was 204.8 mL H2/L/d for *Scenedesmus obliquus* and 39.1 mL H2/L/d for *Chlorella vulgaris* [11]. Maswanna et al. studied H2 production by *Tetraspora* sp. CU2551 cells immobilized in a 4% *w/v* calcium alginate matrix in their recent work. They obtained a maximum H2 production rate of 182 ± 20 nmol/mg of cell dry weight/h [67].

Our group recently tested the capability of immobilized *Chlorella vulgaris* (BEIJ G-120 strain) cells in a calcium alginate (3%*w/v*) gel matrix to produce H2 in a direct light-driven process under continuous illumination. Calcium alginate beads were stable, showing minimal cell leakage, and they measured 4.69 ± 0.02 mm in diameter and 54.01 ± 0.03 μL in volume, carrying 145.5 ± 8.9 μg of microalgal cells (biomass dry weight/bead) (Figure 2). Immobilized cells retained their viability for more than 30 days (Figure 3A). Immobilized *Chlorella* cells were capable of generating H2 without nutrient deprivation with a maximum rate of 162 mL/L (Figure 3B). Anaerobiosis was maintained by the presence of glucose and the high respiration rate of the strain.

**Figure 2.** *Chlorella vulgaris* (BEIJ G-120 strain) immobilized in calcium alginate beads. (**A**) Calciumalginate bead; (**B**) *Chlorella vulgaris* cells inside the calcium-alginate beads.

**Figure 3.** (**A**) Fv/Fm of immobilized *Chlorella vulgaris* cells as a function of time; (**B**) H2 production of the immobilized *Chlorella vulgaris* cells as a function of time.

#### **6. Hydrogen Production without Nutrient Starvation**

The H2 production protocol by Melis and coworkers based on sulfur starvation greatly improved light-driven, algal H2 production, and particularly the possibility for researchers to study the process [68]. However, in the recent years it has become clear that it is not adequate for an industrial development of the process since it requires one to eliminate sulfur residues. Moreover, the severe reduction of PSII activity caused by the sulfur deprivation greatly reduces the H2 production and thus the viability of the process. Awareness of these limits has prompted several workers to eliminate the sulfur deprivation phase by selecting strains with high respiration-to-photosynthesis ratios.

Liu et al. presented a work on H2 production of *Chlorella pyrenoidosa* using NaHCO3 as a carbon source and N -(3,4-Dichlorophenyl)-N,N-dimethylurea (DCMU) [69]. In this work, *Chlorella pyrenoidosa* cells showed an overall H2 production of 93.86 mL/L. In a recent work, Li et al. constructed a transgenic *C. reinhardtii* strain (amiRNA-D1) with a heatinducible expression system targeting D1 gene (*psbA*). After a heat-shock, the transgenic *C. reinhardtii* strain presented a 73% decrease of *psbA* gene expression and a 60% increase of H2 content compared to the WT strain [70]. Ben-Zvi et al. explored the in vivo H2 production of HydA–SOD fusion phenotype in *C. reinhardtii* and found that expression of an active hydrogenase superoxide dismutase fusion protein resulted in sustained H2 production with a rate of 20 mL H2/L/d for 8 days [71]. Hwang et al. showed that the over-expression of the hydrogenase gene in *Chlorella vulgaris* resulted in H2 production under aerobic conditions with continuous illumination using CO2 as the sole source of carbon [72]. Under 5% O2 and 10% CO2, *Chlorella vulgaris* strains YSL01 and YSL16 produced 1.9 mL H2/h and 1.2 mL H2/h in 3 and 4 days, respectively. In another of their works, this group studied and compared the photosynthetic activities of *C. reinhardtii* and *Chlorella sorokiniana* with different acetate/Cl<sup>−</sup> ratios [73]. They found that maintaining acetate/Cl<sup>−</sup> ratios greater than 60–100 led to continuous O2 depletion. Using fermenter effluents, at an acetate/Cl<sup>−</sup> ratio of 150, *Chlorella sorokiniana* and *C. reinhardtii* presented an H2 production rate of 0.25–0.33 mmol/L/min and 0.20–0.38 mmol/L/min, respectively. Kosourov et al. demonstrated sustained H2 production by *C. reinhardtii* by shifting the culture light conditions from continuous illumination to a set of light pulses interrupted by longer dark phases [29]. In a recent work, Sirawattanamongkol et al. demonstrated that *Chlorella* sp. strain KLS Sc59 was able to produce up to 750 mL H2/L in the presence of reducing agents such as ethanol and dithionite [74]. H2 production rates in various microalgae strains are summarized in Table 1.


**Table 1.** Comparison of H2 production rates in various microalgae strains.

In our recent work, we reported H2 production by *Chlorella vulgaris* (strain BEIJ G-120) without the use of nutrient deprivation [8]. This *Chlorella* strain presents two main properties: high respiration rate and high light compensation point. By exploiting these two properties, it was possible to efficiently consume the photosynthetically produced O2, thus maintaining anaerobiosis, even under light conditions. In this work, *Chlorella* cells presented a maximum H2 production rate of 12 mL/L/h and an average rate of 4.98 mL/L/h. The strain was capable of producing H2 in the dark as well, by fermentation of glucose. The excessive accumulation of byproducts of the fermentation (e.g., acetate, formate, lactate, ethanol) may inhibit H2 production. However, the possibility of also producing H2 in the dark by microalgae is desirable for the development of the process under natural light/dark cycle. On the other hand, some of the byproducts of dark fermentation, such as acetate, can be used as substrate for mixotrophic grown during the following light phase.

#### **7. Theoretical Limit for Biological Hydrogen Production**

Only a small fraction of the total solar light radiation (>1,100,000 EJ per year) can potentially be transformed into H2 energy using the process of photosynthesis, according to the following general equation:

$$2\text{H}\_2\text{O} + 8\text{ photons} = \text{O}\_2^{\uparrow} + 4\text{e}^- + 4\text{H}^+ = 2\text{H}\_2\text{}^{\uparrow}$$

Step 1 indicates the total incident radiation received at the surface of the culture (100%) (Figure 4). It follows that:


by photosynthetic pigments. As a result, the total amount of available light drops to 41%.


**Figure 4.** The energy losses of the incident solar light during the different steps of the photobiological H2 production process.

#### **8. Chlorophyll Fluorescence Measurements as a Tool for Monitoring Changes of Photochemical Efficiency during the Hydrogen Production Process**

Chlorophyll fluorescence is a fast and non-invasive tool for monitoring residual photosynthetic activity during the H2 production process through changes in the maximum quantum yield of PSII (Fv/Fm) and the effective quantum yield (ΔF/F m) [38]. In particular, it has been observed that in *C. reinhardtii*, at the occurrence of anaerobiosis, the value of ΔF/F <sup>m</sup> rapidly declined, and this drop could be ascribed to the state 1 to state 2 transition, controlled by the redox state of plasto-quinone (PQ)-pool [36,84–86]. This mechanism regulates the migration of the light-harvesting complex (LHC) from PSII (state 1) to PSI (state 2), and it is induced under high level of PQ-pool reduction and excess of light energy. The start of the H2 production induces a partial oxidation of the photosynthetic electron chains, comprising PQ-pool, with a partial recovery of ΔF/F m. In *C. reinhardtii*, migration of LHC can involve up to 80% of the total LHC. The redox potential of the cells represents another important parameter related to the cell physiology under anaerobiosis, as it is the result of a balance between starch degradation, the capacity of PSII to perform photosynthesis, and the ability of cell to dissipate electrons from PQ-pool. Indeed, after establishing anaerobiosis, the value of the redox potential changes from a positive initial value to a very low value (about –550mV in *C. reinhardtii*). The changes of the values of the redox potential lag behind the changes in the yield and are less rapid that the chlorophyll fluorescence changes. In Figure 5, an example of the typical kinetics of chlorophyll fluorescence yield and redox potential in the different phases of the H2 production process is reported.

**Figure 5.** *C. reinhardtii* kinetics of chlorophyll fluorescence and redox potential during induction of H2 production under sulfur deprivation. (**A**) Time courses in dissolved oxygen (empty circle) and output of hydrogen (H2) (filled square). (**B**) The time courses in the effective quantum yield of PSII (ΔF/F m) (filled circle) and the redox potential (Eh) (empty diamond) in *C. reinhardtii* under sulfur deprivation with 70 μmol photons/m2/s, supplied on both sides of the reactor.

Other important information on the changes of the photosynthetic efficiency can be provided by the chlorophyll fluorescence rise kinetics (OJIP curve), strictly reflecting the progressive reduction of the photosynthetic electron transport chain [60,87], which can indicate and quantify the reduction of electrons transport for each step [88]. The most evident change occurs at the J-step level, indicating the reduced transfer of electrons further than QA, measured by VJ parameter, and thus, an accumulation of reduced QA<sup>−</sup> [88]. An example of the changes of the shape of the OJIP curve during the occurrence of anaerobiosis in *C. reinhardtii* is reported in Figure 6.

**Figure 6.** Effect of sulfur deprivation on the chlorophyll a fluorescence transient in *C. reinhardtii* cultures. Start (dark squares); 40 h of sulfur deprivation (blue triangles); 110 h of sulfur deprivation (red diamonds). Relative variable fluorescence (Vt = (Ft − F0)/(Fm − F0)) [89].

In *Chlorella sorokiniana,* nutrient starvation conditions reduce photosynthetic activity and induce anaerobiosis and H2 production, indicated by the decrease of both Fv/Fm and ΔF/F m. Similarly, to what was observed in *C. reinhardtii*, the maintenance of a residual PSII activity provides electron to hydrogenase enzyme [90]. The same behavior was observed in *Chlorella vulgaris* without nutrient starvation with a strain able to reach anaerobiosis in complete medium, showing a decline of Fv/Fm and ΔF/F m within 24 h [8].

In conclusion, the use of fluorescence measurements to monitor changes in photosynthetic activity can help us to better understand the physiological status of microalgae during the H2 production process, making it easier to interfere in the cell metabolism or enhance the production process. Moreover, the application of chlorophyll fluorescence helps in selecting strains more resistant to the stress imposed by anaerobic conditions, and with higher potential H2 output.

#### **9. Photobiological Hydrogen Production in Outdoor Photobioreactors**

Until now, H2 production experiments using *C. reinhardtii* have been carried out mostly under laboratory conditions. Mean LCE in sulfur-deprived laboratory cultures grown in well-mixed PBR has hardly surpassed 1% (light to H2). The necessity to downregulate the PSII activity to the level of the respiration is considered the main reason for such a low efficiency. As a matter of fact, the LCE strongly increased when it was possible to use microalgal strains with high respiration-to-photosynthesis ratio. This was the case of the *Chlorella* strain G-120, which averaged 3.2%, over the 8-day period [8].

The utilization of solar energy is mandatory for the economical scale-up of the H2 production process. However, under solar light, the light energy received by microalgae cells exceeds their ability of light conversion into valuable biomass. This leads to either energy dissipation as heat or to photodamage and cell death, which strongly reduce the LCE. To reduce the "saturation effect", a number of PBR designs have been proposed [91]. Torzillo and coworkers reported H2 production of about 21% of that attained under laboratory in an outdoor 50 L tubular PBR using *C. reinhardtii* under sulfur deprivation (Figure 7) [92]. The PBR consisted of ten glass tubes (2.0 m length and 4.85 cm internal diameter) placed horizontally and connected by polyvinylchloride U-bends (Figure 7). The PBR was placed in a stainless-steel container with temperature-controlled water. A polyvinyl chloride pump allowed the culture to circulate.

**Figure 7.** The tubular photobioreactor (working volume 50 L) used for the outdoor H2 production experiments. The photobioreactor was equipped with probes for measurement and control of pH and temperature. Culture speed can be adjusted to reach the desired turbulence.

The low performance of the culture was explained by the rapid PSII inactivation by the high light irradiation, which during the experiments reached as much as 1850 μmol photons/m2/s in the middle of day. In order to avoid the problem of light saturation, Giannelli and Torzillo, 2012 [79] proposed a 110 L PBR in which the culture tubes were immersed in water with light-scattering silica nanoparticles. The PBR contained 64 glass tubes (length 2.0 m, internal diameter 27.5 mm) arranged on an 8 × 8 square pitch cell and connected by polyvinylchloride U-bends. The PBR was immersed in a rectangular parallelepiped tank made of isotactic polypropylene, except for the opposite square faces, which were made of transparent Plexiglas. The culture was circulated with a peristaltic pump. The light scattering promoted by nanoparticles permitted a homogeneous distribution of light on the surface of the PBR (Figure 8). Solar light was collected by two sun-tracking mirrors, which delivered light to the opposite faces of the reactors through two light ducts (Figure 8c).

**Figure 8.** The 110 L photobioreactor utilized for the hydrogen production experiments. (**a**) General view of the photobioreactor; (**b**) frontal view of the photobioreactor; (**c**) sun-tracking mirrors collecting light; (**d**) light ducts delivering light to the photobioreactor from the opposite sides; (**e**) nocturnal view of the photobioreactor with two light bulbs (1000 W each) placed in the light duct to provide illumination to the culture during the night.

The total amount of H2 collected reached 3.5 L, which was almost 2-fold higher than that obtained with the 50 L tubular PBR per unit of volume of reactor. Furthermore, in the scaled-up PBR, the LCE of the process increased from 0.055% in the 50 L horizontal PBR to 0.213% in the 110 L PBR, which was much closer to that attained in the laboratory with sulfur-deprived *C. reinhardtii* cultures. This still-low efficiency was the result of a number of factors: (i) the necessity to carry out the process according to the two-step protocol (sulfur starvation), which strongly reduces the contribution of PSII; (ii) the need to attain sulfur starvation by culture dilution, which normally yields lower H2 output; and (iii) the longer mixing time of tubular PBRs, which delays reaching H2-saturation levels in the cultures.

#### **10. Optimal Photobioreactor Design for the Hydrogen Production**

The experience acquired from experiments with different indoor and outdoor PBRs has provided some guidelines useful for the optimal design of future PBRs for H2 production [93,94]. An in-depth discussion on the influence of the PBR design on the H2 output can be found in a recent book on the subject of H2 production [94].

The development of an economically feasible PBR system is the most important factor for successful H2 production. Closed PBRs (such as flat-panel and tubular) are mandatory for H2 production. Tubular PBRs contain one or more glass tubes arranged in various configurations and orientations. Flat-panel PBRs consist of one or more transparent panels containing the culture, which is circulated between the panels by a pump. Flat-panel PBRs probably meet most of the above-mentioned requirements for H2 production. They present high area-to-volume ratio and good biomass productivity, and shorter mixing time compared to tubular PBRs, which reduces the risk of H2 oversaturation in the reactor. A drawback is the difficulty of controlling the temperature, and the high power consumption for mixing. Although the setup costs of closed PBRs are high, they provide several advantages, such as optimal growth, minimizing the risk of culture contamination, and reduced water and CO2 consumption.

The International Energy Agency established the commercial cost target for H2 production at 0.3 USD/kg. According to James et al., with 10% LCE, the cost could be 2.99 USD per gallon of gasoline equivalent [95]. The authors estimated that by using robust microalgae strains presenting 1.5% LCE, the cost of H2 would be 8.44 USD per gallon of gasoline equivalent. Greater costs for PBR construction are materials, manufacturing, and personnel costs. H2 production prices should be more promising than market prices. Show et al. showed that the costs of fabrication materials and chemical nutrients are the main expenses (84% of the total cost) for PBR development [96]. Recycling metabolic products of PBRs (such as organic acids) and/or considering potentially cheaper nutrient sources are possible ways to reduce the chemical cost of nutrients necessary for microalgal growth [97]. Finally, production cost based on direct bio-photolysis, were estimated to be, 18.45 \$/kgH2, for the Netherlands, which is expected to drop significantly in the future (potential cost of 3.10 \$/kgH2) [98].

#### **11. Concluding Remarks**

This review provides evidence that photobiological H2 production by microalgae and cyanobacteria might be a viable option. The discovery of the sulfur starvation method has allowed maintenance of H2 production for several days, and thus, it has represented an opportunity to study the process beyond scientific curiosity.

Nowadays, however, a substantial change of paradigm in photobiological H2 production is necessary. The opportunity to improve the economic feasibility of the process could come from employing strains which do not need sulfur starvation. These strains feature a high respiration-to-photosynthesis ratio and a higher level of saturation irradiance compared to the *C. reinhardtii* strains currently available. These characteristics are usually found in microalgal strains with reduced antenna size, which is a very important biotechnological condition to allow penetration of the light deep through the culture layers [99–102]. Important achievements could be expected from microalgal cultures growing both mixotrophically and heterotrophically in PBRs and fermenter, respectively, in the presence of glucose. A *Chlorella* strain with such characteristics was recently studied by us under laboratory conditions, but its better H2 performance needs to be proved under solar light in PBRs. Of course, the use of an expensive source such as glucose as a respiratory substrate to maintain anaerobiosis, and thus the functioning of the hydrogenase, halves

the efficiency of the process and strongly reduces its sustainability. Therefore, it will be important to consider the potential of much cheaper sources of organic substrates such as wastewater from sugar factories, baker's yeast, and breweries. The O2 consumption, through respiration of organic substrates, produces H2 with high purity (close to 98%), which strongly reduces the investment cost for H2 purification. The use of molasses, which is very rich in glucose and sucrose, could represent an option.

In conclusion, until a hydrogenase resistant to oxygen is discovered, the selection of strains with higher resistance to oxygen and/or with high respiration-to-photosynthesis ratio represent nowadays the only realistic possibility for the success of photobiological H2 production. This research should proceed in parallel with efforts to engineer organisms with O2-resistant hydrogenases. Success in either direction will lead to expected improvements in technologies to: (1) increase effective conversion efficiency of photosynthesis; (2) reduce or possibly eliminate competing pathways, such as CO2 fixation; (3) increase starch biosynthesis.

**Author Contributions:** Writing—original draft preparation, E.T., G.T., C.F. and A.M.S.B.; writing review and editing, E.T. and G.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work received no funds by any research agency.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data that support the findings of this study are available from the corresponding author upon reasonable request.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

