The race to net zero is truly underway with over 127 countries now committed to net zero targets, accounting for 90% of global gross domestic product and 85% of the global population (1). By achieving net zero together, we can all limit global warming to 1.5ºC above pre-industrial levels; however, urgent action is required if we’re going to meet the targets of the Paris Agreement (2). At the 26th United Nations Climate Change Conference of the Parties (COP26), it was widely recognised that not enough progress is being made to achieve the goals agreed in Paris, and the Glasgow Climate Pact calls for each country to strengthen their climate plans this year to achieve 45% reductions in CO₂ emissions by the end of the decade (3).

Carbon pricing aims to reduce emissions of greenhouse gases (GHGs) by directly pushing the costs of emissions back onto emitters; where those operators may choose to continue to emit and pay for it through carbon pricing or invest in transforming their operations to mitigate emissions and cut their future cost of carbon (4). Approaches to carbon pricing include carbon tax credits, such as in the USA, and cap-and-trade schemes, such as the European Union (EU) Emissions Trading Scheme (ETS). These operate differently, but when used alongside policy and regulations they can help move funds from the biggest emitters towards innovation in clean technology, and their deployment at scale which is critical in driving down costs.

Direct emissions of CO₂ from industrials sits at approximately 9 billion tonnes per year, with steel and cement the top emitters, while chemicals and petrochemicals make the top three with 18% of the total (5). These sectors typically have hard-to‐abate CO₂; those that are technically challenging and therefore costly to reduce, locking in carbon from both the combustion (such as fuel) and use (such as feedstock) of fossil fuels within their processes.

All these sectors are looking for cost effective decarbonisation solutions to mitigate emissions, secure a license to operate and future proof against higher CO₂ costs. The adoption of low carbon solutions to both existing and future asset investments is critical to their success, and catalysis and process technology will sit at the heart of delivering those solutions.

As companies start to embark on their net zero strategy, a critical early step is to understand where they stand today in existing operations. Only by knowing the baseline can a company commit to their future GHG emissions reductions and define a pathway to achieving those commitments. There will be decisions taken on the extent of carbon reduction targets, at what pace and affordability, and the risk companies are prepared to take on the direction of future government policy.

To make reporting more transparent, the measurement of emissions is defined in ‘Scopes’: Scope 1 are direct GHG emissions from the company’s processes, Scope 2 are indirect GHG emissions from imported electricity and steam and Scope 3 are other indirect GHG emissions in a company’s upstream (supplier operations) or downstream (customer operations) value chain.

Following the definition of baseline and setting of strategy, the assessment of pathways to reduce Scope 1–3 emissions is started. These decisions on decarbonisation are complex as there are a very large number of alternative options now available at varying levels of technical and commercial readiness. Options can be segmented into three types:

In each of the three areas listed above, companies need to define their execution plans to replace or reduce carbon, which could be made on a whole site basis, or more typically from each of the process units.

With regard to ‘replace’, there are options to replace fuel and process heat with electrified heating (using renewable energy), or to replace fossil fuel feedstocks with renewables, such as recycled plastics, bio-based materials and municipal solid waste for conversion to sustainable fuels and chemicals. In the generation of green hydrogen, water and renewable energy are now new feedstocks, and when combined with CO₂ captured from the air there then exists a pathway to green methanol, such as in the Haru Oni project in Chile (6), and to sustainable synthetic fuels using HyCOgen^TM and FT CANS^TM technology (7).

Now when considering ‘reduce’, it is possible to change to a catalyst that reduces the overall carbon intensity of a product, such as CATACEL SSR^TM for the production of hydrogen (8). However, to maximise carbon reduction versus the baseline of today’s operations, capital investment may be required, and this is where CCS is being evaluated across the industrial space. However, issues remain with the levels of CO₂ reduction that can be achieved at a reasonable capital cost and plant footprint, though it is possible to retrofit plants, such as in hydrogen and methanol, so that up to 95% CO₂ reduction is achieved with significantly reduced capital cost and space requirement compared to post combustion carbon capture (9). Finally, there are options for blue hydrogen, which provides a long-term, scalable and cost-effective replacement for fossil fuels, to enable the decarbonisation of industry, transport and heat (10).

To make decisions on low CAPEX options for transformational technology, or whether a replace or reduce strategy is required, companies will evaluate the potential of process improvements, calculate the reduction in carbon intensity, and identify the technology readiness level and availability of proposed solutions. With the options now available, the industrials are well positioned to make big steps towards the decarbonisation targets of the next decade.

Fermented foods are defined as foods and beverages produced via controlled microbial growth, and by the conversion of food components through enzymatic actions (1). Fermentation enhances the preservation of foods as well as enables the transformation of raw materials into a new product with unique sensory properties (taste, aroma and texture), improved nutritional values and functional properties (2–4). Foods and beverages that are prepared via fermentation processes constitute an important part of human nutrition in virtually every food culture around the globe (3). Fermented or pickled fruits and vegetables, fruit juices, tea leaves, cereals, roots and tubers are very popular and widely consumed in many regions of Europe, Asia, the Americas, Africa and Middle East (5). Generally, several genera of lactic acid bacteria (LAB) including Lactobacillus, Streptococcus and Leuconostoc are predominant in fermented foods, but other bacteria as well as yeast and fungi also contribute to food fermentation (6). There are two main methods by which foods are fermented. First, foods can be fermented naturally, also known as ‘spontaneous ferments’ or ‘wild ferments’, a process whereby microorganisms are naturally present in the raw food ingredient or processing environment, for example, sauerkraut, kimchi and certain fermented soy products (7). On the other hand, foods can also be fermented by the addition of starter cultures, often referred to as ‘culture-dependent ferments’, for example kefir and kombucha (6). One method of preparing a culture-dependent ferment is ‘back-slopping’, a process in which a little amount of a previously fermented batch is inoculated into the raw food, for example sourdough bread (1).

With a surging interest in gastrointestinal health in recent years, the consumption of fermented foods containing live microorganisms has gained popularity as an important dietary strategy for improving human health (1, 7). This popularity may not be unconnected with the rising consumer awareness about the concept of probiotics. Probiotics (from Greek: meaning ‘for life’) are “Live microorganisms which, when administered in adequate amounts, confer a health benefit on the host” (8). Human probiotic microorganisms mostly belong to the Lactobacillus, Bifidobacterium, Streptococcus, Lactococcus and Enterococcus genera (9). Furthermore, some strains of Gram-positive bacteria in the genus Bacillus as well as some yeast strains in the genus Saccharomyces are also commonly used in probiotic products (10). The Food and Agriculture Organization (FAO) and World Health Organization (WHO) (11), the European Food Safety Authority (EFSA) (12) and the US Food and Drug Administration (FDA) have stipulated different criteria such as safety, functionality, technological usefulness and the absence of the risk of acquired resistance to antibiotics for probiotic strain selection that would exert beneficial effects on human or animal health and could be used in the probiotics industry (9, 10, 13, 14). Generally, one of the most important properties among the selection criteria is that a probiotic strain must produce antimicrobial substances and be antagonistic to pathogens. Antimicrobial activity of probiotics may be manifested by one or a combination of the following mechanisms including: (a) competition for limited nutrients; (b) competition for adhesion sites; (c) synthesis of various antimicrobial metabolites such as organic acids, hydrogen peroxide, bacteriocins; and (d) inhibition of toxin production in pathogenic microorganisms (15, 16). For instance, LAB produce lactic acid as well as other organic acids; thus, lower the environmental pH and consequently inhibit the growth of bacterial pathogens (17, 18).

Although commercially-produced fermented foods usually serve as carriers for probiotic bacteria (6), that a particular food or beverage is produced by fermentation does not necessarily imply that it contains live microorganisms. For example, bread, wine, beer and distilled alcoholic beverages are produced by yeast fermentation, but the producing organisms are either inactivated by heat (in the case of bread and some beers) or are physically removed by filtration or other means (in the case of wine and beer). Moreover, many fermented foods are heat-treated after fermentation to extend shelf-life or enhance food safety. Hence, soy sauce, sauerkraut and other fermented vegetables are made shelf-stable by thermal processing. Even non-thermally processed fermented foods may notwithstanding contain low levels of live and viable microorganisms simply due to unfavourable environmental conditions which decrease microbial populations over time (6). Furthermore, it is also pertinent to note that the absence of live microorganisms or probiotics in the final fermented product does not preclude a positive functional role. Several studies have reported some mechanisms through which fermented foods may exert beneficial effects on health and disease whether they contain live microorganisms or not. For instance, food-fermenting microbes may produce vitamins or other bioactive molecules in situ or inactivate anti-nutritional factors and yet be absent at the time of consumption (6). Prebiotics and other components found in fermented foods may also exert health benefits (1, 19). In addition, fermentation-derived metabolites may exert health benefits such as in the case of LAB which generate polyamines and bioactive peptides in both dairy and non-dairy fermented foods with potential effects on immune, metabolic and cardiovascular health (20).

There are sufficient historical and scientific records that have shown Turkey to be an origin country for a decent number of fermented probiotic beverages (21). Moreover, several fermented traditional foods are produced in many regions of Turkey from products that are indigenous to particular regions. Kefir, ayran, boza, hardaliye and shalgam juice seem to be the most widely known fermented traditional Turkish non-alcoholic beverages. The first two products are produced from milk while the last three are obtained from cereals, fruits and vegetables respectively, and their microbiota is composed mainly of LAB (22).

Kefir is a smooth, slightly foamy, whitish, viscous, slightly acidic, slightly carbonated mildly alcoholic fermented beverage that is widely consumed in Turkey (23, 24). The name kefir is derived from the Turkish word ‘keyif’ meaning ‘joy, pleasure or good feeling’ to express the feelings experienced after drinking it (21, 23–26). Kefir can be produced by fermentation of different kinds of milk including cow, ewe, goat or plant-derived milk (26). It is traditionally produced by adding a starter culture known as ‘kefir grains’ to pasteurised milk. Kefir grains is a symbiotic consortium of lactic and acetic acids-producing bacteria as well as lactose-fermenting yeasts (for example, Kluyveromyces marxianus) and non-lactose fermenting yeasts (for example, Saccharomyces cerevisiae, Saccharomyces unisporus) housed within a polysaccharide and protein matrix called kefiran (25, 27). Kefir has been reported to be tolerated well by people with lactose intolerance and maldigestion as it contains β-galactosidase enzyme expressing organism (for example K. marxianus) which hydrolyses lactose, and thus reduces lactose concentrations in the beverage (7).

Ayran is an indigenous Turkish fermented milk beverage which is more or less a ‘national drink’ widely consumed across the length and breadth of Turkey (21). It is produced in either of two ways, i.e. by the addition of water to yoghurt (homemade) or by the addition of Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus as starter cultures to standardised milk for fermentation (industrial production) (28). Basically, ayran is generated by blending yoghurt with 30–50% of water and 0.5–1% of salt (29).

Boza is a cereal-based fermented beverage produced by a combination of yeast and lactic acid fermentation of millet, maize, wheat, rye, rice semolina or flour and mixed with sugar or saccharine (30). It is a viscous liquid with a pale yellow colour and sweet slightly sharp to slightly sour taste which is widely consumed in Turkey due to its pleasant taste, flavour and nutritional properties (31). It is also popular and widely consumed in several Balkan countries on a daily basis (26, 32). In Turkey, boza is usually served alongside cinnamon powder and roasted chickpeas, and is generally considered to be beer’s ancestor (30). Boza is produced at both an artisanal and industrial scale (33). Its preparation typically involves six stages: milling of raw materials (cereals), boiling, straining, cooling, sugar addition and fermentation. A previously fermented boza is used as inoculum for the fermentation stage (21, 30, 31). Lactic acid fermentation by LAB produces antimicrobial metabolites such as lactic acid and increases acidity which supplies preservative effect whereas the metabolites from alcoholic fermentation by yeasts bring about the mouthfeel and odour of boza (26, 30, 31). Boza is a good source of fibre, carbohydrate, protein and vitamins including riboflavin, thiamine, niacin and pyridoxine (26, 30).

Hardaliye is a kind of fruit-based fermented non-alcoholic traditional beverage which originates from Thrace, the European part of Turkey, and is produced from red grape juice and 0.2% crushed black mustard seeds (34, 35). Hardaliye is mostly homemade by the traditional method. The ingredients are pressed and allowed to ferment at room temperature for 5–10 days in wooden or plastic barrels (22, 26, 33). Benzoic acid is sometimes added as a preservative especially at the industrial scale (26). Owing to its rich LAB flora, hardaliye has been described as a non-dairy probiotic beverage which aids digestion and also helps in the prevention of coronary heart disease (36). The nutritional value, functional properties and health benefits of hardaliye are derived from its ingredients and fermentation process. For example, grapes are rich in phenolic contents and provide strong antioxidant effects for the human body and thus, inhibit cancer cells formation, whereas the oils from mustard seeds exert medicinal effects on circulatory disorders, common cold and bronchitis as well as possess antimicrobial properties (34).

Shalgam juice (Şalgam) is a vegetable-based red coloured, cloudy and sour beverage produced by LAB and yeast fermentation of a mixture of black or purple carrots (Daucus carota), turnips (Brassica rapa), bulgur (broken wheat) flour, salt, sourdough and water (37). It is widely consumed in many cities across the Mediterranean region of Turkey but has also in recent years become popular in metropolises such as Istanbul, Ankara and Izmir (38). Production comprises two stages with the first one termed ‘first fermentation’ which involves mixing bulgur flour (3%), salt (0.2%) and sourdough (0.2%) together with water and allowing it to ferment at room temperature for 3–5 days. In the second stage, cleaned and chopped black carrots (10–20%), sliced turnips (1–2%), salt (1–2%) and water are then added to the extract obtained from stage one and left to undergo a ‘second fermentation’ for 3–10 days in a wooden barrel (37). The juice is then filtered and packaged in suitable containers, and chilli powder may also be added depending on consumer preference (26). Shalgam is typically made on a home-scale and consumed within three months from production time, although it is also produced commercially with an extended shelf life of 1–2 years using preservatives (26). Shalgam juice is a highly nutritional beverage due to its high mineral, amino acid, polyphenol and vitamin contents (37, 39).

There are quite a few studies and review papers that have reported the microbiological, chemical, nutritional and some probiotic properties of fermented traditional Turkish foods and beverages (22, 26). However, no study reporting the antimicrobial efficacies of the potential probiotic microbiota of these five major beverages (boza, kefir, hardaliye, ayran and shalgam) was encountered in the literature. The current study, therefore, aimed to consolidate past works on the probiotic properties of the selected beverages specifically by investigating the antimicrobial activities of their microbiota against selected standard pathogens, thereby helping to validate their widely acclaimed beneficial effects on human health. Ultimately, the present study aims to contribute to public knowledge of fermented functional foods and probiotics, increase awareness of the Turkish populace about the health benefits derivable from their own indigenous fermented dairy and non-dairy beverages. It is hoped that in the long run this knowledge and awareness will result in increased consumption of these functional food products and, hence, result in an improved general public health.

2.1 Isolation, Identification and Characterisation of Target Microorganisms from Beverage Samples

2.2 Evaluation of the Antimicrobial Activities of the Isolates

A total of 22 bacterial strains comprising lactic and acetic acid bacteria were isolated from four (boza, kefir, shalgam and ayran) of the five fermented traditional Turkish beverages analysed. No microbial strain was isolated from hardaliye. Boza and kefir contained the largest microbiota loads with eight isolates each followed by shalgam having five, which is due to their various raw materials and fermentation processes. The names of the 22 isolates as identified by molecular sequence analysis of their 16S rRNA and their distribution in the studied fermented traditional Turkish beverages are presented in Table I.

In the spot-on-the-lawn assay, the bacterial strains isolated from boza, kefir and shalgam exhibited antimicrobial activity against nearly all tested indicator pathogens at varying degrees. Moreover, the results presented in Table II clearly show that the obtained probiotic isolates displayed antagonistic activity against both Gram-positive (B. cereus, MRSA, S. aureus, S. epidermidis, VRE faecium) and Gram-negative bacteria (A. baumannii, E. coli, K. pneumoniae, P. aeruginosa, S. typhi). Lactobacillus nagelii (B24), Lactobacillus parabuchneri (B25) and Acetobacter peroxydans (B27) isolates obtained from boza displayed the strongest antagonistic effects with ≥20 mm diameter inhibition zones against all tested bacterial pathogens. However, they exhibited no antifungal activity against C. albicans. MRSA was highly sensitive to Lb. nagelii (B24) and Lb. parabuchneri (B25) isolates from boza as well as to Lactobacillus fermentum (K13) and Gluconobacter frateurii (K26) isolates from kefir with 26.00 mm, 22.10 mm, 26.40 mm and 25.90 mm inhibition zones, respectively, but minimally sensitive to Lactococcus lactis (S2) obtained from shalgam (12.00 mm zone). L. lactis strains B7 and B10 isolated from boza showed medium antimicrobial activity against the bacterial pathogens and Candida albicans with ≥ 10 mm ≤ 20 mm diameter inhibition zones. L. lactis (B7) strain exhibited the strongest inhibitory effect (28.60 mm zone) while L. lactis (B10) and Lb. fermentum (K13) strains displayed medium antagonism producing 16.80 mm and 11.90 mm inhibition zones respectively against C. albicans. G. frateurii (K26) and Lb. fermentum (K13) strains isolated from kefir also showed a very strong antibacterial activity creating 20.80–30.10 mm diameter inhibition zones against five bacterial pathogens comprising both Gram-positive (VRE, MRSA) and Gram-negative (P. aeruginosa, K. pneumoniae and E. coli). In contrast, another Lb. fermentum strain K12 also obtained from kefir displayed a weak antagonistic effect (<10 mm zones) against the same pathogens. The strain L. lactis (S2) isolated from shalgam exhibited a medium inhibitory effect (9.60–15.70 mm zones) on all indicator bacterial pathogens except A. baumannii.

On the other hand, in the agar well diffusion assay, the CFS of 10 out of 22 isolates exhibited medium antimicrobial effects (≥ 10 mm ≤ 20 mm zones) against bacterial and C. albicans pathogens (Table III). However, P. aeruginosa was strongly inhibited by Lb. fermentum (B4), Lactobacillus pentosus (S5) and Lactobacillus plantarum (S8) strains with 25.10 mm, 23.40 mm and 19.10 mm diameter zones respectively. The CFS of Lb. fermentum (B6) with a 25.40 mm zone of inhibition displayed the strongest antagonistic activity against the yeast pathogen C. albicans. Furthermore, L. lactis strains (K18 and A22), Lb. fermentum strains (B4 and K13), Lb. plantarum strains (K14 and K15) had medium inhibitory effects with 10.20–15.80 mm zones on the pathogenic yeast C. albicans. The weakest antibacterial effects were displayed by the supernatants of L. lactis (K18) and Lb. plantarum (S8) both with an 8.30 mm inhibition zone against P. aeruginosa and K. pneumoniae indicator strains respectively. Representative images of the isolates are shown in Figures 1–3.

Fig. 1.

Fig. 2.

Fig. 3.

Among the known mechanisms of action by which probiotics exert their beneficial effects on human health is antagonism against microbial pathogens via the production of antimicrobial metabolites. In the present study, the antimicrobial efficacy of some potential probiotic bacteria isolated from fermented traditional Turkish beverages against 11 common human pathogens has been evaluated. Thus, with respect to antimicrobial effects, the potentials of boza, kefir and shalgam to bestow some probiotic health benefits to the consumers have been validated through the present study. Other probiotic properties of Turkish boza and kefir were reported in a previous study (43). Kefir was also reported in a previous study to exhibit antimicrobial activity which is attributable to the lactic acid, acetic acid, bacteriocins, hydrogen peroxide, acetaldehyde, volatile acids, diacetyl and CO₂ produced by the bifidobacteria and LAB strains present in the drink (44). Although ayran is the most widely consumed amongst the fermented Turkish beverages examined, its inherent LAB strains exhibited no significant antimicrobial effects. No lactic acid bacterial or other potential probiotic strain was isolated from commercially available hardaliye in this study. This reason is thought to be due to the presence of the preservative benzoic acid which is usually added at the start of the fermentation process. It has, therefore, been determined that commercial hardaliye lacks probiotic properties considering the fact that probiotics must be found alive and in sufficient quantities in final products. In order to retain the potentially probiotic natural microflora of commercial hardaliye, production without the use of benzoic acid or other chemical preservatives is recommended. In the same vein, no live LAB strain was found in some brands of commercial shalgam that were examined. Similar to hardaliye, the reason is also thought to be due to the presence of preservatives and additives such as sodium benzoate which is declared in the ingredients list on label.

Pathogenic bacterial multidrug resistance as well as biofilm formation have led to the ineffectiveness of the antibiotics available in the treatment of infections whereas the application of probiotics has been considered functional in preventing and counteracting biofilm-related infections (45). Antagonism by antimicrobial metabolites has been considered as an important property in the selection of potential probiotics for the maintenance of a healthy microbial balance in the gut. LAB, mostly the lactobacilli due to their capacity to alienate bacterial pathogens via the production of some antimicrobials such as organic acids (mainly lactic acid), bacteriocins and hydrogen peroxide, acquire probiotic potential (46). Lactic acid typically diffuses into pathogenic bacterial cell which disrupts the cell membrane integrity thereby causing damage to it as well as retarding their metabolic processes and preventing growth (47). Within the scope of this study, the antimicrobial activities of LAB and acetic acid bacteria strains isolated from fermented Turkish beverages were assessed by two methods namely spot-on-the-lawn and agar well diffusion methods.

Through this study, it has been possible to evaluate the potential beneficial effects of fermented traditional Turkish beverages on human health as regards antimicrobial efficacy. 18 strains namely: Lb. fermentum (B4), L. lactis (B7), Lb. fermentum (B6), L. lactis (B10), Lb. fermentum (K13), Lb. fermentum (K12), Lb. nagelii (B24), Lb. parabuchneri (B25), L. lactis (S2), L. lactis (K18), L. lactis (A22), Lb. plantarum (K14), Lb. plantarum (K15), Lb. fermentum (B3), Lb. pentosus (S5), Lb. plantarum (S8), G. frateurii (K26) and A. peroxydans (B27) exhibited varying degrees of antimicrobial activities against the tested pathogens. Furthermore, differences in the antibacterial effect of different strains of the same species of isolates were observed. This result has confirmed that the antibacterial activity and potential health benefits imparted by probiotic bacteria are strain-specific rather than being species- or genus-specific. It is therefore crucial to note that no single strain will provide all the proposed benefits, not even strains of the same species, and also not all strains of the same species will be effective against defined health conditions (48, 49). This specificity has also been confirmed by previous studies in terms of the microorganism strain, the metabolites it produces, and even the susceptibility pattern of the test organism to antimicrobials (50). It therefore goes without saying that in order to obtain maximum health benefits, the addition of a mixture of various strains of probiotic organisms to diets is imperative.

Many bacteriocins isolated from LAB are usually active against Gram-positive bacteria while Gram-negative bacteria generally exhibit little sensitivity to bacteriocins. The difference in resistance between Gram-positive and Gram-negative bacteria may be due to differences in the cell envelopes (51). However, production of lactic acid in high concentrations in combination with bile salts has an inhibitory effect on the growth of pathogenic Gram-negative bacteria in the intestinal tract (52). The present study has reported antagonistic activity of LAB strains against both Gram positive and negative bacteria alike. This finding is in accordance with a similar study (53) which reported that the CFS of a Lb. plantarum strain displayed broad spectrum antimicrobial activities against Gram-positive and negative bacteria as well as against yeast C. albicans. P. aeruginosa, K. pneumoniae and E. coli were reported to be sensitive at medium levels to Lb. plantarum and L. lactis strains isolated from Turkish boza and B. cereus (a Gram-positive bacterium) was also strongly inhibited by the same LAB strains (54). Also, the inhibitory effect of Lb. plantarum and Lb. fermentum strains used as starter organisms in the production of bread and bakery products against rope-forming B. cereus was previously reported (55, 56). Many Lb. plantarum strains have been reported to produce bacteriocins known as plantaricins (57) which are compounds that are highly diverse in their activities and structures and have been reported to be particularly active against gastrointestinal as well as food-borne pathogens (58). In the present study, L. lactis strains (B7 and B10), Lb. fermentum strains (K12 and K13), G. frateurii ( K26), Lb. nagelii (B24), Lb. parabuchneri (B25), A. peroxydans (B27) and L. lactis (S2) isolates showed better antibacterial activity by spot-on-the-lawn method. The excellent antibacterial activity reported of Lb. parabuchneri against all indicator pathogens in this study is in contrast to the results obtained by Meira et al. (59) who had reported that Lb. parabuchneri strain isolated from a Brazilian ovine cheese had little inhibitory effect against all tested pathogens. Lb. parabuchneri is an obligatory heterofermentative bacterium occasionally isolated from cheeses (60). Considering that heterofermentative LAB typically ferment glucose to yield ethanol, acetic acid and CO₂ in addition to lactic acid as byproducts as against homofermentative LAB which produce only lactic acid, they possess the ability to antagonise highly pathogenic bacteria. It is, therefore, noteworthy that this is the first study to have reported the presence of Lb. parabuchneri in a Turkish boza drink. LAB strains that produced larger inhibition zones against E. coli, S. aureus and Salmonella enteritidis in a previous study were reported to be heterofermentative (61).

Comparing the results of the two methods adopted in this study, L. lactis strains (B7 and B10) as well as Lb. fermentum (K13) strain showed antifungal activity against C. albicans by spot-on-the-lawn method, but the same effect was not observed with agar well method. The lack of positive results by agar well diffusion method may be due to the indiffusibility of the secreted molecules since this is the basis of agar well assay. In contrast, Lb. plantarum (K14, K15), Lb. fermentum (B3, B4, B6), L. lactis (K18), Lb. pentosus (S5) and Lb. plantarum (S8) strains showed antimicrobial activity by agar well diffusion but no antagonism was recorded with spot-on-the-lawn method. Since they showed negative results by spot-on-the-lawn method, it is thought that the isolates may not secrete any primary or secondary metabolites other than bacteriocins. Therefore, spot-on-the-lawn method has been found in this study to be more effective than agar well diffusion method in the determination of antimicrobial activity. This finding was consistent with the results reported by (62) who investigated the antagonistic effects of LAB strains against Gram-negative bacteria using these two methods and found spot-on-the-lawn to be more effective. However, utilising the probiotic strain Bifidobacterium bifidum against S. enterica serovar Enteritidis, agar well diffusion method was demonstrated to be better in determining antagonism than the other two methods (disk diffusion and spot-on-the-lawn) employed (63). The inhibitory activity on tested bacteria by the spot-on-the-lawn method is seen as better, but it could be as a result of the synergy of all metabolites lactic acid, acetic acid, diacetyl and bacteriocin, as they are being produced during the assay period (64). The variation in antibacterial activities as depicted by different authors might be due to the number of the CFU of the LAB used (in spot-on-the-lawn) and the quantity of culture supernatant used (in agar well diffusion) as well as the activity and diffusibility of the bacteriocins possessed in it (46, 65).

Probiotics are increasingly gaining attention from both the food industry and academia. Probiotic foods constitute a significant part of the functional foods market worldwide. The presence of these good microbes in fermented food products all over the world, and their ability to combat pathogenic and spoilage microbes using different mechanisms of action, has been validated with several scientific studies. From the perspective of antimicrobial health benefits, the present study consolidates past studies demonstrating the probiotic potentials of fermented traditional Turkish beverages. The study has substantiated the antimicrobial efficacy of the potential probiotic strains isolated from kefir, boza and shalgam against an array of human pathogens comprising both Gram-positive and negative bacteria as well as yeast (C. albicans). In total, 16 LAB and two acetic acid bacteria isolates were found to be antagonistic at varying degrees against the tested bacterial and yeast pathogens. In the current study, boza and kefir followed by shalgam were found to be more effective in terms of antimicrobial activity against human pathogens. The outstanding antimicrobial efficacy of boza and kefir in particular is believed to be connected with the greater diversity of their microflora as well as the absence of chemical preservatives. It has, therefore, been determined that boza, kefir and shalgam compared to other fermented traditional Turkish beverages analysed, are the most promising probiotic candidates. In addition, going by the results obtained in this study as well as previous studies on other probiotic properties, boza, kefir and shalgam may also be applicable as bio-therapeutics or nutraceuticals against bacterial and yeast infections in humans, although in vivo studies would be useful in validating their efficacy. Due to the presence of beneficial nutrients and substances such as probiotics and antioxidants, boza, kefir, ayran, shalgam and hardaliye can also be generally regarded as functional foods. Other probiotic properties of the microbiota of these fermented Turkish beverages have been reported individually in previous studies. However to the best of our knowledge, this is the first study that evaluated the antimicrobial properties of the isolates obtained from all five beverages against a broad spectrum of human microbial pathogens. The study has to a large extent achieved its objective.

Future research efforts should be directed toward animal testing and clinical trials to better evaluate the antimicrobial as well as overall probiotic effects of Turkish boza, kefir and shalgam on humans. Furthermore, suitable biopreservatives for commercial hardaliye and shalgam drinks should be investigated in order to retain their beneficial microflora whilst extending their shelf lives. Alternatively, probiotic strains and starter cultures resistant to benzoic acid may be investigated in further studies for fortification of commercial hardaliye since no LAB strain was isolated from any hardaliye sample which is thought to be due to the presence of chemical preservative. Ayran which is a ‘national drink’ is also recommended for fortification with carefully selected and promising probiotic strains from the Bifidobacterium and Lactobacillus genera in order to enhance its probiotic functional qualities.

The pgms have been the cornerstone of many catalytic applications and will continue to be so as processes become more sustainable. Platinum is a highly reactive catalytic metal with applications in highly contemporary processes, such as low temperature water-gas shift (1, 2) proton exchange membrane fuel cells (3) and green hydrogen production through the hydrogen evolution reaction (4). Ruthenium and iridium oxides are state-of-the-art catalysts for the associated oxygen evolution reaction (5, 6). The pgms are also found to be highly active for aqueous phase hydrogenation and reforming reactions, of relevance for biomass platform chemical utilisation (7–12). Yet, it is often considered that pgm-rich catalysts cannot satisfy the requirements needed for the employment of these new technologies at scale, because of their relative scarcity, high cost and poor catalytic stability in certain key reactions. Carbon monoxide poisoning of platinum in fuel cell applications is an excellent example of catalyst poisoning and instability. Regarding scarcity and environmental cost, Nuss and Eckelman in a life cycle assessment showed that the global warming potential of rhodium, platinum, iridium and ruthenium are 35,100 kg_CO2 kg_M^–1, 12,500 kg_CO2 kg_M^–1, 8860 kg_CO2 kg_M^–1 and 2110 kg_CO2 kg_M^–1 (M = metal) respectively (13). Although the scale of use of these materials in relation to iron, steel and aluminium must be placed in context.

Strategies to mitigate against these issues can be broadly summarised as: (a) improving atom efficiency of pgms through control of atom cluster or nanoparticle sizes, i.e. improving performance per gram of pgm; (b) pgm-non-noble metal alloying to reduce pgm content and improve stability and poisoning resistance; and finally (c) complete replacement of pgms with other catalytic materials. Advancements in synthetic procedures and characterisation has seen a proliferation of work around controlled pgm clusters and single atom catalysts that incorporate the concept of a ‘single site catalyst’ (14–19). The term, coined by the late Sir John M. Thomas, refers to a single energetically equivalent active site providing maximum efficiency of catalytic material (20). However, control over metal nuclearity in an applied context is far from trivial, catalyst stability is challenging and proof of the existence or retention of ‘single-sites’ requires robust in situ or operando characterisation (21). Utilisation of non-noble metals (alone or in alloys) clearly reduces the environmental burden of catalysts. Straight replacement of pgms with first row transition metals can result in acceptable catalytic performance, but notably poorer than a pgm regarding activity and stability (22–24). Specifically, metals such as nickel and copper are more prone to oxidation and leaching into liquid reaction media. Alloying of these metals with pgms often removes these weaknesses and even on occasion produces superior catalytic performance, but perhaps fail in their main remit of sufficiently reducing reliance on pgms (25–28). It is worth noting that these strategies are not mutually exclusive, and combinations can be quite successful for specific applications (vide infra).

An alternative strategy is to diversify into abundant metal compounds. Transition metal carbides are a class of Group IV–VI metal compounds with carbon occupying interstitial sites of the relevant metal. In addition to being known for their hardness (29), the early transition metal carbides have been shown by Levy and Boudart to have ‘platinum-like’ catalytic behaviour (30). Since the seminal work by Levy and Boudart on tungsten carbide catalysed hydrogen oxidation and hydrocarbon isomerisation, metal carbides have been widely used as catalysts for a host of hydrogenation (31–34), dehydrogenation (35–38), hydrogenolysis (39–41), carbon monoxide or nitric oxide reduction (42), isomerisation (43, 44) and water-gas shift reactions (45–48). Molybdenum carbide phases, which are the focus of this review, have received significant attention for contemporary applications in green and sustainable processes. Interest in molybdenum carbide catalysis is driven by its catalytic properties, low cost, earth abundancy and resistance to poisoning from sulfur, nitrogen and carbon monoxide impurities present in both hydrocarbon and biomass feedstocks (49–52). Such properties make these catalysts appealing as robust alternatives to pgm catalysts. However, this generalisation of properties is overly simplistic, given the numerous possible structures of carbides, a vast array of potential applications that have subtly different requirements, and the fact that while similar, the properties of metal carbides and platinum are clearly not identical. The reader is directed to several excellent reviews which discuss these complexities in detail, such as discussing metal carbides properties in the context of other alternative catalytic systems (such as nitrides, phosphides and boron alloys (53)) and those that focus on specific catalytic applications or resistance to poisoning (49, 54, 55). A further recent observation is the exceptionally strong metal support interaction between platinum and some first row transition metals and molybdenum carbides, in particular the metastable α-MoC_1–x phase. Several reports show that single-atom platinum can be supported on α-MoC_1–x to improve catalytic properties (56–58).

The application of molybdenum carbides is more complicated than may be initially perceived, as demonstrated by the lack of its widespread adoption in industrial processes. One aspect to consider is the challenge in identifying and selecting a specific phase and morphology of molybdenum carbide, which is desirable for a particular process. In this context, theoretical studies of reaction mechanism with respect to termination and morphology of different carbide phases has provided great insight. Specifically, they provide understanding of the thermodynamic stability of specific surfaces with respect to synthesis, reaction conditions and Mo:C composition, then used this as a basis to elucidate a preferred structure to a specific catalytic application. A further challenge is a reliable, scalable and affordable preparation method that allows for the synthesis of such specific carbide phases and morphologies. This current review will provide a brief overview of molybdenum carbide phases and their general properties, then discuss the properties and synthetic routes of the metastable molybdenum carbide (α-MoC_1–x) phase that has recently received significant interest. An overview of the potential synthetic strategies to produce α-MoC_1–x is given in Scheme I.

Scheme I.

As alluded to above, all molybdenum carbide phases are generally characterised by outstanding mechanical properties and, like other transition metal carbides has (in parenthesis Mo₂C values): high hardness (17 Gpa), tensile strength (530 GPa), high melting point (2520°C) and good thermal (15 W m^–1 k^–1) and electrical conductivity (57 μΩ cm) (59). The latter point on electrical conductivity being defined by the delocalisation of the molybdenum d-band due to increased Mo–Mo distances, which result in an increased density of states near the Fermi level (60). The novel catalytic properties of molybdenum carbide, which as we already have said, have been found to be comparable to the noble metals, can be explained by the surprising similarity of its electronic properties to platinum (61, 62) (Figure 1). Specifically, these metallic properties are due to the hybridisation of carbon sp and molybdenum d-orbitals, which allows the delocalisation of the d-band of the molybdenum, similar to the d-band of metallic platinum (61). It is this broad partially occupied d-band in MoC_x that makes this material particularly successful as catalysts, in hydrogen evolution and purification, reforming reactions, hydrogenations and Fischer-Tropsch synthesis (53, 57, 61, 63).

Fig. 1.

Taking these general aspects and applications for which MoC_x has grown in popularity into account, a more thorough analysis of the different phases is necessary. Phases will have differences according to geometric factors (Hagg’s rule regarding the ratio of the hard ball radii of nonmetal to metal (64)) and electronic factors (Engel and Brewer theory on s–p electron count (65)), which define their structure, stability and activity as illustrated by Oyama (60, 66). In addition, it is worth remembering that, as with any material, bulk and specific surface structures can have different features (67).

Molybdenum carbide forms a range of phases with 1:2 or 1:1 Mo:C stoichiometry (68, 69), although significant nonstoichiometry has been observed. In all cases they are considered interstitial compounds and changes in the content of carbon and a variation of the oxidation states of molybdenum are observed accordingly. Across this Mo:C composition range seven main phases are reported, as shown in Figure 2 and Table I. Some of them are thermodynamically stable at room temperature, others are only stable in a very small range of temperatures and compositions, while others are metastable. Importantly, the stability of the phases varies according to the stoichiometry. Hugosson et al. affirmed that a substoichiometric phase is not favourable for the γ and γ’ phase, while η and δ are present only with a certain amount of vacancies (δ-MoC is experimentally found only between MoC_0.66–MoC_0.75 (70–73)). Nonstoichiometry influences multiple properties, such as the bulk moduli and the electronic properties, with vacancies changing the density of states (DOS) due to the unscreened Mo–Mo (74, 75).

Fig. 2.

It is worth noting that terminologies are inconsistent across the literature, with the orthorhombic Mo₂C being previously assigned as α (76, 77) or β (68, 69, 78). Experimentally and theoretically (79) it has been observed that orthorhombic Mo₂C is the most thermodynamically stable phase and is extensively reported in numerous publications. Confusion arises as this phase is sometimes referred to as α-Mo₂C to differentiate it from a high-temperature phase discussed in the relevant papers as β-Mo₂C (80, 81). Also an intermediate temperature phase between these two, called ɛ, has also been observed, which has a hexagonal or trigonal structure (71, 81, 82). Due to the low scattering cross-section of carbon, laboratory source X-ray diffraction (XRD) patterns of these phases are quite similar and only recently advanced neutron scattering experiments have classified these structures conclusively (80–83). Consequently, materials focused research refers to the stable orthorhombic phase as α-Mo₂C, while the catalytic community refers to this phase is as β-Mo₂C as a way to easily differentiate it from the metastable face-centred cubic MoC_1–x designated as α-MoC by Clougherty et al. (84). Further confusion is caused by this face-centred cubic MoC_1–x being interchangeably referred to as α or δ. We will follow the commonly used notation of β-Mo₂C (hcp) and α-MoC_1–x (fcc).

The most commonly catalytically studied phase is the orthorhombic β-Mo₂C. However, there has been increasing interest and research into the metastable MoC_1–x phases, in particular the cubic α-MoC_1–x, although the hexagonal η-MoC and γ-MoC phases have also been shown to be of interest in a limited number of studies (85–87). The stability and properties of these phases and the influence of Mo:C ratio (45, 88, 89) and nonstoichiometry in this context has been thoroughly investigated (71, 90, 91).

1.2 Surface Properties of the α and β Phases

1.3 An Overview of the Catalytic Application of α-MoC_1–x

1.4 Thermal Catalytic Applications of α-MoC_1–x

1.5 Hydrogen Evolution Reaction

Historically, interstitial metal carbide synthesis techniques have been grouped (66, 145) as: (a) those associated with high temperature techniques; (b) temperature programmed techniques; and (c) thin film synthesis by plasma or chemical vapour deposition. Oyama provides, in his 1992 review, an excellent general overview of the first two techniques at the time (66). In summary, high-temperature methods, such as traditional carbonisation of oxides or metals with solid carbon, reaction of vapourised metals with hydrocarbons or solid combustion methods produce thermodynamically stable carbides of generally low surface area and are of limited scalability. Prior to the introduction of temperature programmed synthesis, the only procedure known for the synthesis of α-MoC_1–x was quenching a mixture of Mo–C from around 2000°C (70).

Boudart and his group developed in 1973 the temperature programmed synthesis route to making tungsten and molybdenum carbides (30). In the direct temperature programmed reaction of molybdenum trioxide and various methane/hydrogen mixtures the β-Mo₂C phase can be produced with surface areas between 50–100 m² g^–1 (146). Depending on the CH₄:H₂ ratio used a clean Mo₂C was formed or a mixture of β-Mo₂C and surface polymeric carbon, which could be removed by controlled treatment in hydrogen. The reaction proceeds through an initial reduction of MoO₃ to MoO₂, first via the formation of sub-MoO₃ oxides, followed by carburisation to form Mo₂C. Boudart and coworkers showed that control of ramp rate and the reduction potential of the gases is key to controlling the kinetics of oxygen removal, carbon migration into the structure and molybdenum species sintering (146). Synthesis of the α-MoC_1–x phase was not observed by the group using this methodology, which can be rationalised by an aggressive carburisation process that facilitated the rearrangement of the molybdenum cations to form the thermodynamically favoured β-Mo₂C phase. The method developed by Boudart and coworkers to produce α-MoC_1–x via a modification of this temperature programmed synthesis remains today the most studied and adopted synthesis strategy, mainly for the very high purity of the carbides produced as well as for the high surface area.

2.1 Temperature-Programmed Synthesis of α-MoC_1–x

2.2 Temperature-Programmed Synthesis of α-MoC_1–x Using Carbon Supports and Amine Precursors

2.3 Temperature-Programmed Synthesis of α-MoC_1–x With Non-Carbon Supports and Additives

2.4 Alternative Preparation Routes to α-MoC_1–x

Passivation of as-synthesised α-MoC_1–x (or any other transition metal carbide) is often required due to its highly reactive and pyrophoric nature. Without passivation there is significant surface instability in the presence of air, exacerbated by the high surface areas of α-MoC_1–x (40). The consequence of uncontrolled exposure to air are, at a scientific level, a lack of understanding of the reactive surface of the catalyst and at an applied level a significant safety issue. Commonly, within the academic open literature, passivation procedures are given little discussion, or else outright ignored, in synthetic procedures. While this is not universally true, the result is often a lack of clear and precise protocols. Further, regarding this literature review on α-MoC_1–x synthesis, the few dedicated passivation studies are related to the β-Mo₂C phase, despite it still being an essential process for the application of α-MoC_1–x.

Conventionally, passivation of pyrophoric materials is conducted at room temperature in a controlled atmosphere with low concentrations of oxygen (0.5–2% oxygen in helium or nitrogen) over relatively long reaction times (2–20 h) (70, 149, 181–183) The process can also be done by gradually increasing the oxygen partial pressure of the mixture 1–20% before exposing the sample to air (117). Alternatively, passivation can be done by using milder oxidants such as water/CO₂ with and without oxygen at 400°C (181, 184), but while this can be less disruptive of the carbide’s subsurface structure, it has reduced effectiveness regarding long term stability on exposure to air (185).

Studies on passivation of β-Mo₂C and WC phases show that it can strongly change the surface properties and the reactivity of the material (40, 181). Jentoft and coworkers showed that 0.1% and 1% oxygen passivation was also insufficient for long term storage, with reactivation with hydrogen showing significantly more oxygen content in the old sample than fresh (Figure 15). Demczyk et al. showed in the analogous case of nitrides an actual distinct subsurface phase of Mo₂N_3–xO_x was formed during the passivation (144, 186). As noted by Hargreaves in his recent review, this observation raises the question as to how these subsurface phases influence activation and carburisation (144). Further, passivation is also a crucial consideration when wishing to deposit reactive metals on the MoC_x surface. Simply, this is because the metal is no longer interacting with a MoC surface, but with an oxidised surface. This was highlighted clearly by the studies of Thompson et al. (47, 187, 188) with platinum on Mo₂C, where platinum assumed different structural configurations on a passivated surface compared to an unpassivated one: namely forming agglomerates on the passivated Mo₂C, while remaining finely dispersed on the unpassivated Mo₂C. The water-gas shift activity in the two samples was notably different even though the surface areas were similar.

Fig. 15.

Passivation of catalysts frequently requires an accompanying reactivation step prior to further surface treatments, catalyst characterisation and employment in a catalytic reaction. However, the passivation-activation procedure is not innocuous and the reactivated surface is rarely the same as in the fresh catalyst (40). In the process of removing surface oxygen through the use of hydrogen, carbon atoms are lost from the near-surface (189, 190). The use of mixtures of methane/hydrogen appears to be milder, with reduced but still significant surface carbon loss (191). It is common to reactivate passivated catalysts by flowing hydrogen or alkene/hydrogen mixtures at 450–700°C for one or two hours prior to catalytic use (56, 182, 189, 192).

Potentially it is possible to avoid a direct passivation step, for example Gao et al. report a method to grow β-Mo₂C on carbon nanotubes in a way that no passivation was needed afterwards (193). Another example is work by Malmstadt and coworkers where the synthesised α-MoC_1–x was reportedly capped by organic compounds (34). Alternatively, some literature studies perform catalytic testing directly after synthesis and within the same reactor to avoid the need for passivation (167), although this is unlikely to be practically viable in an applied industrial reactor. In reality, such unpassivated catalysts will have significant surface reconstruction upon exposure to catalytic reactants (167). Although on a partially tangential topic to carbide synthesis, Román-Leshkov and coworkers have recently shown how reactive carbide surfaces are during hydrodeoxygenation reactions (194). Therefore, it is obvious and possibly not unexpected to think that a reactive carbide catalyst will have notably different surface and possibly subsurface structure as compared to an as-synthesised material.

Since Levy and Boudart’s seminal study of the ‘platinum-like’ catalytic behaviour of early transition metal carbides, there has been a significant scientific push to apply these compounds to contemporary catalysis. It has been noted that, while these compounds are generically platinum-like, they have their own unique properties and further that different transition metal carbides and carbide phases have their own specific catalytic behaviours and applications. Therefore, it has become clear that particular carbide structures and morphologies should be targeted using the correct synthetic procedures. This review focuses on the metastable fcc structured α-MoC_1–x which has excellent OH bond activation properties for application in reforming, water-gas shift and biomass conversion technologies. In addition, the structure provides exceptionally strong metal support interactions with transition metals and can be used to produce highly dispersed metal supported catalysts. Table VI provides the collated synthetic procedures, parameters and structural properties of the resultant α-MoC_1–x materials from this review.

Controllable synthesis of this phase is highly challenging on both a laboratory scale and on an applied industrial scale. Most carbide synthetic strategies revolve around a temperature programmed methodology, in which a molybdenum precursor (most often MoO₃) is heated at a slow ramp rate with a high space velocity of hydrogen and an alkane. In the absence of specific synthetic strategies the process will result in the formation of the thermodynamically stable β-Mo₂C. In order to successfully produce the metastable α-MoC_1–x, the process is believed to have to undergo a topotactic reaction of carbon insertion into an equivalent cubic precursor phase. fcc γ-Mo₂N, MoO_xC_y and MoO_xH_y have all been observed as such topotactic precursors. Certainly, in the case of γ-Mo₂N a clear, measurable scientific correlation between this precursor phase and the final α-MoC_1–x product has been reported. The process produces high surface area (225 m² g^–1) and high purity α-MoC_1–x. However, it is frequently asserted that Boudart’s original ammonolysis of MoO₃ prior to carburisation of the resultant γ-Mo₂N is impractical. Handling of high-temperature pure ammonia gas streams is clearly challenging, but as demonstrated by numerous scientific groups, possible to achieve on the laboratory scale. Application on an industrial scale would however require significant control of process heat flows and would undoubtably be challenging.

Avoiding the use of ammonolysis to form α-MoC_1–x, via hydrogen pretreatment, incorporation of hydrogen dissociating transition metals, use of higher alkanes (i.e. butane and higher) in carburisation and decomposition of amine containing precursors is well reported. A number of these synthetic protocols produce relatively high purity α-MoC_1–x, although traces of β-Mo₂C are frequently observed. Reported surface areas vary significantly, with to our knowledge only one higher (relative to Boudart’s method) reported value of >500 m² g^–1, several at comparable surface areas to the original 200 m² g^–1, some with dramatically lower surface areas and several papers not reporting surface area. The presence of additional graphitic carbon addressed in Boudart’s original publications is often not discussed in these papers. In some instances, evidence has been provided that certain cubic intermediates are produced during the temperature-programmed reaction of these ammonolysis free routes. Unfortunately, this is not always the case. This is particularly notable in more recently reported, non-conventional preparation routes, including using alternative molybdenum precursors and the formation of hybrid hierarchical materials, for example in the use of hard templating. Our perspective is that several control experiments require performing and that in situ characterisation studies of these synthesis routes would provide a solid scientific basis as to how they produce α-MoC_1–x. Further, it is not clear if the use of more costly molybdenum precursors or processes makes these alternative routes more applicable on an applied scale than Boudart’s ammonolysis strategy. Each of these alternative routes will have specific advantages and disadvantages, for example the addition of <0.2 wt% platinum or first row transition metals might be viable or even beneficial in some catalytic applications. Indeed, several publications report excellent metal-support interaction between α-MoC_1–x and platinum to make highly active catalysts. The incorporation of so called ‘single atom’ platinum onto α-MoC_1–x presents an exciting topic area which bridges other contemporary areas of science that aim to maximise pgm efficiency.

DFT based theoretical studies have shown that there are interesting correlations between morphology, particle size and Mo:C ratio in regard to the α-MoC_1–x phase. Validation of these findings through experiments requires further work. Specifically, the predicted morphologies are based entirely on the thermodynamic stability of specific surface terminations. Practically, access to the metastable phase requires topotactic rearrangement of a precursor phase (i.e. kinetic control) and so the morphology is dictated by that of the precursor. Most synthesis methods have used orthorhombic MoO₃ which has large platelets with a high fraction of the (010) surface termination. Given that the (100) plane of Mo₂N or α-MoC_1–x is parallel to this, these products exhibit pseudomorphism. The validity of these very interesting theoretical studies can be interrogated by suitable characterisation of synthesised materials using modern advanced microscopy. In addition, very little research has been published on exploiting the interdependence of α-MoC_1–x morphology on that of the precursor. It should be noted that amine-MoO₃ composite synthesis appears to be one of the few synthetic strategies to date that can produce different α-MoC_1–x morphologies.

Noticeable throughout a significant proportion of the literature is an absence of discussion around α-MoC_1–x passivation. Given that the α-MoC_1–x surface is highly reactive, passivation will create permanent surface and subsurface changes to catalysts, influencing performance and the ability to support additional metal nanoparticles. However, from a practical perspective the process is entirely necessary to ensure easy handling of this pyrophoric material. Fully exploring this parameter space, while less exciting than development of new preparation routes, is equally important. Systematic studies that consider correlations between passivation conditions (oxidant, its concentration and reaction temperature), specific carbide properties (surface area, surface terminations, Mo:C ratio and particle size), along with difference between different phases (α-MoC_1–x vs. β-Mo₂C), are required. Characterisation to elucidate these correlations could include the use of near-ambient XPS (NP-XPS) which would be highly beneficial for studying surface changes in situ.

There is a great deal of literature and understanding of transition metal carbide catalysts and an evolving understanding of specific phases, such as α-MoC_1–x. However, controllable and practical synthesis methods still require further work. These include methodologies that make use of relatively low temperature and simple pyrolysis steps, which ideally will have greater tolerance to deviation in gas flow dynamics and heating ramp rate than current temperature-programmed techniques. In addition, it is important that any preceding wet chemical processing to make molybdenum precursors that facilitate ‘simple’ heat treatments are clearly understood and reproducible at scale. Ensuring a systematic understanding of the process and how it influences the final phase and surface properties of the catalysts is essential to scale up and application.

Driven by the desire to develop novel catalyst formulations which are applicable for localised, more sustainable routes, the area of heterogeneously catalysed ammonia synthesis has attracted much attention in the academic literature in recent times. One of the key incentives for this has been the idea that ammonia synthesis for the production of synthetic fertiliser can be conducted on, for example, a farm close to its point of application with the required hydrogen feedstream being derived from sustainable sources such as electrolysis of water accomplished using electricity produced using wind turbines or solar energy sources. Further drivers are the possible application of ammonia as a non-fossil based fuel and also as a means to indirectly store intermittent over-supply of sustainably derived electricity. In the literature, the energy intensive nature of the Haber-Bosch process, frequently quoted to be 1–2% of global energy demand, and its carbon dioxide footprint, stated to comprise 2.5% of fossil fuel based emissions, are statistics that are often quoted in justification for the search for new routes to ammonia production (1, 2). However, due recognition has to be given to the highly efficient nature of the Haber-Bosch process as currently operated. In relation to this, large scale synthesis of ammonia is highly optimised and it can be credited with the sustenance of ca. 40% of the global population. These considerations, coupled to the recently reported UK CO₂ supply chain shortage, related to a reduction in commercial fertiliser production (3), underline the importance of the highly integrated nature of the process.

In developing smaller localised sustainable ammonia synthesis units operating under milder reaction conditions, a step change in catalyst technology would be necessary requiring the development of catalysts being able to operate in the highly desirable lower temperature–lower pressure reaction regimes which have proved elusive for so long. Such catalysts should also be able to withstand a series of rapid start-up and shut-down procedures corresponding to the intermittent nature of sustainably derived electricity supply. A further consideration, not generally acknowledged nor widely explored, is the desirability of such new catalysts to withstand poisoning. In the following, we detail some of the approaches which have been described in the literature in recent years to bypassing the inherent limitations of the more conventional catalysts. While electrocatalysis and photocatalysis are increasingly explored in relation to nitrogen activation, we have concentrated on heterogeneous catalysis.

The paradoxical nature of scaling relationships in relation to conventional metal-based ammonia synthesis catalysts is widely acknowledged. This relates to the occurrence of a Sabatier volcano type relationship whereby catalysts of optimum activity possess intermediate nitrogen binding energies. In this way, the relative performance of metals known to be effective ammonia synthesis catalysts (such as iron, ruthenium and osmium) can be rationalised and the design of new active catalyst formulations can be undertaken by alloying metals of low nitrogen adsorption strength with those of high nitrogen adsorption strength (4). This rationale has been applied to explain the comparatively high performance of the Co₃Mo₃N ammonia synthesis catalyst (5, 6) on the basis of the combination of cobalt (low nitrogen adsorption enthalpy) with molybdenum (high nitrogen adsorption enthalpy) as expressed in the (111) surface plane (4). This explanation implies structure-sensitivity, which has not generally been explored with this material, and also considers the role of the lattice nitrogen to be induction of ordering such that the active surface crystallographic plane is exhibited (4).

An alternative viewpoint of the origin of the activity of this system is the potential occurrence of a Mars-van Krevelen mechanism. The Mars-van Krevelen mechanism, which is perhaps most frequently encountered in oxidation catalysis using metal oxide catalysts, involves the direct participation of reactive lattice species generating transient vacancies which are replenished to complete the catalytic cycle (7). For ammonia synthesised from metal nitrides, this would involve the synthesis of ammonia directly by hydrogenation of lattice nitrogen leading to vacancy sites where nitrogen is activated. To this end, direct hydrogenation of Co₃Mo₃N to produce ammonia occurs and can lead to the previously unprecedented Co₆Mo₆N phase (8). Co₃Mo₃N and Co₆Mo₆N are two line phases (with no intermediate stoichiometries being observed) with the transformation involving the loss of 50% of the lattice nitrogen upon reduction, with the residual lattice nitrogen relocating from the 16c to 8a crystallographic site (8). At ammonia synthesis temperature, using a hydrogen/nitrogen reaction feed, the rapid regeneration of Co₆Mo₆N to Co₃Mo₃N occurs. Such regeneration can also be accomplished by treatment with nitrogen alone (9) thereby opening the possibility of a chemical looping system for ammonia production (albeit with a very low gravimetric nitrogen content). Heterolytic isotopic exchange studies wherein ¹⁵N₂ is exchanged with the lattice nitrogen of Co₃Mo₃N (in which the nitrogen is predominantly ¹⁴N reflecting normal isotopic abundance) demonstrate the lattice nitrogen to be highly exchangeable with 40% exchange occurring at 600ºC within 40 min (10). Computational modelling lends further support to the possibility of the participation of reactive lattice nitrogen for this system in ammonia synthesis with high concentrations of surface nitrogen vacancies being predicted at the temperatures applied (11). Recently, the possibility of an associative mechanistic pathway occurring has been demonstrated through modelling (12) and this proposal is consistent with the observation that Co₃Mo₃N is not effective for homomolecular ¹⁴N₂/¹⁵N₂ exchange (which directly relates to the material’s ability to dissociate nitrogen) at ammonia synthesis temperature despite being an active catalyst (10). Generally, for heterogeneously catalysed ammonia synthesis, dissociative pathways (wherein nitrogen is dissociated on the catalytic surface often in the rate determining step) are invoked and the occurrence of an associative pathway (in which hydrogenation of surface bound nitrogen occurs prior to its dissociation) is an exciting development since it potentially provides a premise for catalyst design and a link to enzymatic nitrogen fixation which operates, albeit comparatively very slowly, under ambient conditions.

The role of nitrogen vacancies as being of interest for the development of novel nitride-based catalysts has been exemplified in the recently reported nickel/lanthanum nitride catalyst (13). In this catalyst, which is reported to exhibit high activity and extended stability, it is proposed that the scaling limitation is circumvented via the occurrence of a dual site mechanism in which nitrogen activation occurs at the nitrogen vacancies of the lanthanum nitride with hydrogen dissociation being accomplished by the nickel component. This concept has been expanded to cerium nitride and yttrium nitride also, with cerium nitride being an effective catalyst in its own right in which both nitrogen activation and hydrogen activation can be accomplished on nitrogen vacancy sites (14). The application of dual sites to circumvent limiting scaling relations has also been applied to the development of catalysts based on the combination of metals or metal nitrides with lithium hydride in which activated N is believed to transfer to the lithium hydride component which is progressively hydrogenated to LiNH₂ and then ammonia, regenerating the lithium hydride component (15). In terms of this strategy, combination of various alkali and alkaline earth hydrides with manganese nitride has been undertaken with LiH-Mn₄N being particularly effective (16). Elsewhere, in a chemical looping based study, lithium-ion doping has been shown to significantly enhance lattice nitrogen lability in Mn₆N_5+x where computational modelling indicates that its presence reduces the nitrogen vacancy formation energy (17). Overall, it is arguable that the role of alkali metal doping and hydrogen activation have been less extensively studied for novel catalysts. For Co₃Mo₃N, low levels of Cs⁺ ion doping have significant promotional effects but there are challenges in the preparation of the material related to phase stability issues (5).

High activities for ammonia synthesis have been reported for solid state hydrides and oxyhydrides in recent years. To this end, the performance of TiH₂ (18), BaTiO_2.5H_0.5 (18), BaCeO_3–x H_y N_z (19), VH_0.39 (20) and NbH_0.6 (20) is documented within the literature. In a number of cases, hydrogen-based Mars-van Krevelen mechanisms have been invoked as the source of activity. The surfaces of some of these materials may possibly nitride in operation, as demonstrated by small shifts in lattice parameters and elemental analyses. In the case of BaTiO_2.5H_0.5, the formation of BaTiO_2.5N_0.2H_0.3 is documented although BaTiO_2.5N_0.3 when tested exhibited no activity, thereby emphasising the initial requirement for the hydride containing phase to be present. Lattice mobility is an important consideration in the Mars-van Krevelen mechanism wherein diffusion of species within the bulk of materials to their surfaces occurs. This consideration was applied to the development of the vanadium and niobium hydride based systems on the basis that their ‘more open’ body-centred crystal structures facilitate diffusion to a greater extent than the face-centred cubic based crystal structures based upon TiH₂ and the perovskite structure based BaTiO_2.5H_0.3 and BaCe_3–x H_y N_z systems. This consideration, coupled with advantageous metal-nitrogen bond strengths, is invoked to explain the enhanced performance of these materials (20). This is an interesting design consideration from which novel catalyst compositions of apparent high activity have been developed from components which, in isolation, might not be expected to display good performance. Lanthanum oxyhydrides have been applied as supports for ruthenium where it is reported that catalytic performance relates to high surface hydride ion mobility leading to low work function electrons at hydride ion vacancies near the ruthenium interface enhancing nitrogen activation and reducing the effect of hydrogen poisoning on the ruthenium component (21). Surface hydride is proposed to be formed by the reaction of electrons at vacancies reacting with hydrogen adsorbed on the ruthenium component forming hydride ions which subsequently react with adsorbed nitrogen species releasing electrons back to the vacancy sites. It was also stated that the choice of oxyhydrides is important in reduction of the deleterious effect of nitridation. Ternary ruthenium hydrides have also very recently been reported as effective catalysts with Ba₂RuH₆/MgO outperforming the most active catalysts published to date (22). It is proposed that an associative pathway occurs for this system, which is stabilised by the [RuH₆] centres, lattice hydrogen and alkaline earth metal cation. Experiments performed for Li₄RuH₆/MgO demonstrated the requirement for feeding both nitrogen and hydrogen for the generation of ammonia.

In addition to interest in the role of nitrogen and hydrogen vacancies, activation of dinitrogen at dicarbide vacancies has been proposed to occur in relation to nickel-loaded rare earth dicarbides (23). In the case of Ni/CeC₂, the nickel component was reported to be important in promoting the formation of C₂ vacancies in cerium carbide by virtue of its enhanced hydrogen activation functionality. Furthermore, differences in nitrogen adsorption energy and configuration were noted in comparing Ni/CeN and Ni/CeC₂ although they were found to possess comparable ammonia synthesis activity under the testing conditions applied (23). In relation to the catalytic performance of carbides, earlier studies had shown β-Mo₂C to be an effective catalyst and more active than γ-Mo₂N, unlike the α-MoC_1–x phase which was not stable under reaction conditions (24). In a comparison between the performance of Co₃Mo₃N and Co₃Mo₃C, which had been carefully prepared so as to avoid any complicating effects of differing morphology, the carbide was shown to be less active requiring a higher reaction temperature and with activity developing after an initial lag period (25). Upon extended testing, gradual nitridation of the Co₃Mo₃C phase was observed to occur. Ammonia synthesis was related to the presence of nitrogen occupying the 16c crystallographic site although it was not possible to distinguish whether the presence of lattice nitrogen arose from the production of ammonia or was the cause of it (25). Co₆Mo₆C, which comprises carbon in the 8a lattice site, was found to be inactive under the conditions tested and the phase remained stable (25). This is contrary to the observations for Co₆Mo₆N referred to previously.

In concluding this brief summary of some of the recent advances in relation to the development of ammonia synthesis catalysts, it is fair to say that there are very exciting and tantalising opportunities for the design of novel catalysts. The ability to synthesise more complex nitrides such as the quaternary phases NiCoMo₃N (26) and (NiM)₂Mo₃N (M = copper or iron) (27) provides the possibility to tune catalytic performance. Taking ternary metal nitrides as an example, in the comparison of the behaviour of Co₃Mo₃N, Fe₃Mo₃N, Ni₂Mo₃N and Co₂Mo₃N it can be observed that the role of composition and structure is complex (28) and computational modelling has a role to play in rationalising the origin of performance leading to the design of new catalysts. It is also noteworthy that the possibility of associative pathways is being invoked more widely. Such pathways are closer to enzymatic systems and, as such, may be related to the possibility of lower temperature catalytic routes. While surveying the literature uncovers many interesting and exciting advances in the area of alternative materials which can be extended to the further discovery of novel catalysts, the development of technology suitable for small scale sustainable ammonia synthesis has yet to be accomplished and remains an exciting and potentially highly rewarding challenge. In terms of application, consideration would need to be given to longevity of performance and also to poison tolerance as well as catalyst handling, storage and preparation. In relation to preparation, it is interesting to draw attention to the considerations of heat transfer in relation to the synthesis of active materials as has been discussed, for example, for binary molybdenum nitrides where nitridation of MoO₃ via ammonolysis and treatment with nitrogen/hydrogen mixtures has been compared (29, 30).

“Sustainable Materials for Transitional and Alternative Energy” is a 294-page, five-chapter book which forms one part of the Modern Materials and Sensors for the Oil and Gas Industry Series. The book is authored and edited by Mufrettin Murat Sari (Texas A&M University, USA), Cenk Temizel (Saudi Aramco, Kingdom of Saudi Arabia), Celal Hakan Canbaz (Ege University, Turkey), Luigi A. Saputelli (ADNOC Frontender Corp, USA) and Ole Torsæter (Norwegian University of Science and Technology, Norway), in collaboration with a further 15 contributors.

Mufrettin Murat Sari is a Chemistry Professor at the Department of Chemistry, Texas A&M University, and Life and Health Science Department, University of North Texas at Dallas, USA. Beginning with his PhD degree from Hacettepe University, Turkey, in 2005, he has 20 years of experience in the fields of materials chemistry, applied biochemistry and nanotechnology. Throughout his illustrious career, he has published about 40 scientific articles and proceedings with hundreds of citations. Cenk Temizel is a Senior Reservoir Engineer with Saudi Aramco. His career spans around 15 years in industry working on reservoir simulation, smart fields, heavy oil, optimisation, geomechanics, integrated asset modelling, unconventionals and enhanced oil recovery (EOR) across the Middle East, USA and UK, prior to which he was a teaching and research assistant at the University of Southern California, USA, and Stanford University, USA. Like Temizel, Celal Hakan Canbaz is a Senior Reservoir Engineer with 16 years industrial experience. He is highly regarded in the fields of special core analysis, reservoir wettability characterisation, well testing analysis, perforation and testing design, multiphase flow meters, carbon dioxide/oil/water interactions, wellbore flow dynamics and pressure-volume-temperature data interpretation. He has made contributions to industrial projects, conference and journal papers, two books and a US patent. Luigi A. Saputelli is a Reservoir Engineering Expert Advisor with 30 years of experience as a reservoir, drilling and production engineer. Alongside his industrial contributions, he is a researcher, lecturer and active volunteer in the Society of Petroleum Engineers, and has published in excess of 100 industry papers on digital oilfield, reservoir management and real-time production optimisation. Ole Torsæter is a Professor of reservoir engineering at the Norwegian University of Science and Technology and Adjunct Professor at the University of Oslo, Norway, as well as a research associate at PoreLab, a Norwegian Centre of Excellence. He has supervised a staggering 220 Masters and 25 PhD candidates, alongside publication of 200 research papers himself, with the most recent focusing on nanofluids for EOR.

“Sustainable Materials for Transitional and Alternative Energy” addresses today’s energy needs in a fast-paced world in which nanotechnology is of vital importance for maintaining environmental sustainability and protecting human health. This book keeps pace with advancements in the energy industry by addressing the latest research involving advanced nanomaterials which engineers can apply to nanoparticle applications beyond the petroleum industry. Additional topics in Volume 2 include carbon capture-focused, green-based nanomaterials, the importance of coal gasification in terms of fossil fuels and advanced materials for fuel cells.

This book has been written with the intention of targeting a wide range of readers from academics to researchers, and undergraduate to graduate students, from various backgrounds, including petroleum engineers and researchers, nanotechnology researchers in the oil and gas industry, chemical engineers and material scientists.

This book will be reviewed in order of chapter due to its methodological breakdown starting with the oil and gas industry, before moving onto ‘greener’ technologies involving nanomaterials.

The objective of this chapter is to define the field of smart and state-of-the-art materials together with their current status and potential benefits. In the future, more focus will be placed on additives, nanoparticles, shape memory materials and piezoelectric materials: those that can generate electricity on the application of mechanical stress.

Chapter coverage is equally divided among smart materials and state-of-the-art materials. The former refers to those materials that can change their composition or structure, their electrical and mechanical properties or even their functions in response to environmental stimuli. The state-of-the-art materials included in this chapter are additives (bacterial control, corrosion inhibitor, fluid loss, lubricants, fluid viscosifiers, synthetic-based muds, clay stabilisers, antifreeze, odorisation and defoamers) and nanoparticles (for improving sweep efficiencies, stabilised foam, polymer flooding and reduced water production). Their properties and applications are discussed in detail throughout this book, and specifically, this chapter.

Geothermal energy is one of the largest renewable energy sources in existence. Chapter 2 addresses investigation methods used in the exploration, discovery and monitoring phases of geothermal systems, the structural characterisation of which can be implemented through the use of geophysical tools. The latest technology is compared with conventional counterparts in terms of efficiency and cost-effectiveness. This includes the use of nanotechnological materials and advanced tools (for example, fibre optic sensors, distributed temperature sensing systems, thermal infrared remote sensing tools and advanced technology carriers such as drones, aircrafts and satellites) for measuring physical properties of geothermal fluids and rocks. Furthermore, advanced materials and nanomaterials starting to be used in geothermal downstream parts, such as heat transfer and energy conversion in thermoelectrical power plants, are introduced and discussed in appropriate detail.

CO₂ capture and sequestration (CCS) technologies have received increasing interest in recent years due to the pressing global demand to reduce CO₂ emissions and slow down global warming. This chapter explores the development of reusable, low-cost and green-based adsorbents which are important for efficient and environmentally-friendly removal of CO₂ from the atmosphere. The absorbents discussed in detail include halloysite nanotubes, nanofibrillated cellulose, enzyme immobilised on bioinspired nanosorbents, green metal-organic frameworks and bio-derived porous carbons. The advantages of each are given, alongside research study results and their many uses in CCS, which are well summarised in Sub-chapter 3.7, ‘Conclusion and Outlook’. Looking forward to the future, there is a need to investigate CO₂ selectivity of nanomaterials over other gases (for example nitrogen and methane) as research currently focuses on adsorption characteristics.

As global energy needs increase, so does the importance of addressing energy security, environmental problems and the cost of energy. Since renewable energy sources are yet to reach the required energy demand, and issues of radioactive waste from nuclear energy remain, the gasification process (conversion of hydrocarbon fuels into gases) stands out as an alternative technology that enables the production of clean gas products that can be used in many areas. This book’s penultimate chapter provides a comprehensive background for nanocatalysts and sensors in the coal gasification process. Emphasis is placed on the use of catalysts, specifically nanocatalysts with their superior physical properties (namely, large surface area to volume ratio), in the gasification process which have the potential to increase carbon conversion rates and economically reduce processing times.

Inconsistent with the first three chapters of this book and the subsequent final one, Chapter 4 does not have a ‘Conclusions’ section. This emphasises the point raised at the end of this review which highlights the book’s lack of editorial consistency and disruptive flow of content.

The fifth and final chapter of this book introduces the reader to fuel cell technology and its importance as one of the main solutions for the next generation of environmentally friendly energy. It provides a general perspective on fuel cell types, namely polymer electrolyte membrane, microbial, alkaline, phosphoric acid, solid oxide and protonic ceramic, their mechanisms and applications, alongside the nanostructures used to produce catalysts in this field.

For background context, fuel cells are electrochemical devices that directly convert chemical energy from oxygen and hydrogen into electrical energy in a single step. Several improvements and developments regarding the increase in hydrogen production, improvements in cell design, removal of membrane, utilisation of microbial catalysts or platinum-free catalysts have been successfully demonstrated in the field of electrocatalysts. These are addressed in this chapter and summarised in Figure 1 and Table I.

Fig. 1.

While proton exchange membrane fuel cell development is the most highly anticipated area, traditional catalysts, such as platinum nanoparticles on carbon (Pt/C) are still being used in current fuel cells as well as nonprecious metal catalysts. Nanoscale materials used in fuel cell technology include nanoparticles, nanoframes, nanorods and core-shell structures – the latter of which has shown enhanced results in the cathodes and anodes in fuel cells electrodes – which have contributed immensely to fuel cell commercialisation expansion. Current and potential fuel cell applications are discussed in detail in this chapter and summarised in Figure 2.

Fig. 2.

In summary, this book provides a detailed introduction and assessment of the use of sustainable materials in the energy industry. Scientific papers and research are heavily referenced throughout; these are highly detailed and often require the reader’s full attention and specialist background knowledge in order to be fully understood. This book is best suited to a subject matter expert interested in making technological advances in the energy industry using smart and state-of-the-art materials, namely nanomaterials.

In general, this is a poorly written book which made for an unnecessarily difficult read. It is riddled with grammatical errors, such as the omission of conjunction words which greatly impacts upon the flow of the material. Material flow is also disrupted by poor figure placement throughout; there are multiple examples where figures are placed in the middle of sentences rather than at the end of paragraphs making it difficult for the reader to link information.

Inconsistencies exist between chapters due to having multiple contributory authors. The most significant of these concerns the level of background or introductory information provided. For example, in Sub-chapter 4.2, fossil fuels, CO₂ formation and global warming are explained in detail, but assumptions are often made in other chapters of a high level of understanding on much more complex and less familiar topics. This is inconsistent with the fact that the book claims to be targeted to undergraduate students from various backgrounds.

Furthermore, the impact of multiple authors on the fluidity of this book is evident through the inclusion of several introductions to nanotechnology and nanomaterials across numerous chapters. This is tedious for the reader when reading the book as a whole; however, if chapters are viewed in isolation, a case can be made for the need for several introductions. Nevertheless, repeated content could be removed, thus saving on space, by cross-referencing information between chapters; this would also have the added benefit of improving the fluidity of the book’s content.

Resultantly, this book could have been better edited to ensure consistencies between chapters and eliminate repetitive information.

Refined, white beeswax pearls and retinyl palmitate (vitamin A) of 1.7 MIU g^–1 (MIU = milli-international units) were purchased from Bulk Apothecary (Aurora, OH, USA) and Fisher Scientific USA (Pittsburg, PA, USA), respectively. Ethanol was obtained from Decon Laboratories, Inc (King of Prussia, PA, USA). Compression fabric (warp knit: 77% nylon and 23% spandex) was obtained from the Marena Group (Lawrenceville, GA, USA). Pampers^® Aqua Pure^TM nonwoven wipes were also used as a carrier to transfer microparticles from the substrate to skin.

We microencapsulated retinyl palmitate by melt dispersion technique and investigated the effect of four process variables on the produced microcapsules, such as different theoretical loading capacity (10%, 15%, 25%), types of wax (beeswax, carnauba wax, paraffin wax), emulsifier concentrations (0%, 1%, 2%) and stirring speeds (180 rpm, 230 rpm, 280 rpm) in our previous study (12). The statistical analysis showed that theoretical loading capacity and surfactant (%) were the most significant factors and we were able to determine that the highest theoretical loading (25%) and highest surfactant (2%) selected in that study can provide us high actual loading with the small size of the particles. There was no significant difference found among the effects of type of wax on loading capacity, encapsulation efficiency, antioxidant activity or mean size of particles. Hence we decided to conduct further study selecting beeswax as the shell material because of its natural skincare benefits as well as operational convenience due to low melting point (65°C). We selected 280 rpm stirring speed to facilitate dispersion of the oil-in-water emulsion and formation of small size particles.

Thermal analysis of the beeswax, retinyl palmitate and retinyl palmitate-loaded beeswax microcapsules was carried out by using Mettler-Toledo GmbH DSC821e (Greifensee, Switzerland) instrument, where a standard empty aluminium pan was used as the reference. The weight of the samples was within 2–9 mg, and the samples were scanned from 25°C to 100°C under nitrogen atmosphere with a heating rate of 10°C min^–1.

After preparing the microcapsules with 25% theoretical loading, we looked into the shelf life of microcapsules by measuring the actual loading percentage, i.e. the content of retinyl palmitate in a fixed amount of capsules over a period of time, both in powder and dispersion forms. We evaluated the shelf life of the beeswax microcapsules (approximately 71% encapsulation efficiency) in powder form, where they were filtered and dried before storing in an enclosed petri dish under room temperature; and also in dispersion form (approximately 75% encapsulation efficiency), where the particles were kept dispersed within the emulsion during preparation, stored inside dark vials in refrigerator and a portion was filtered on each day of measurement (Day 1, Day 4, Day 8, Day 15 and Day 31).

An extraction from 0.1 g of microcapsules was performed, by heating the capsule in 20 ml of ethanol solution to release the vitamin content and then filtering the wax residue. The concentration of supernatant aliquots was measured at 327 nm by a Shimadzu Corporation UV-2401PC spectrophotometer (Kyoto, Japan). The amount of retinyl palmitate was determined from a standard curve of known concentrations.

We conducted an in vitro kinetic release study similar to prior literature (27, 43) with some modification based on particle content, solvent type and machine parameters. The retinyl palmitate release profile from 3 g of suspended particles (approximately 77% encapsulation efficiency) was examined in 600 ml of pure ethanol. The study was performed in a New Brunswick Scientific C24 (Eppendorf, Germany) incubator shaker with a speed of 100 rpm and temperature set at 37±2°C. Supernatant aliquots of 2 ml were withdrawn and replaced by the fresh medium at appropriate time intervals (1 min, 5 min, 15 min, 30 min, 1 h, 2 h, 4 h). The supernatants containing dissolved retinyl palmitate were diluted and analysed by ultraviolet-visible (UV–vis) spectroscopy at 327 nm. The results were compared with a standard to calculate the vitamin A concentration and to evaluate the release ratio.

We used a robotic transfer replicator (Figure 1) to simulate the transfer of microparticles from a nonwoven wipe to the skin and evaluate the transfer percentage, by means of a similar method as described by Yu et al. (44). 1 g of microparticles was spread as evenly as possible by a spatula over a commercial nonwoven wipe containing 99% water that acted as a donor surface with a diameter of 133 mm. The receptor material was a compression fabric, i.e. a warp knit with a composition of 77% nylon/23% spandex (fabric weight 276 g cm^–2). This fabric was chosen because the study by Yu et al. (44) regarding transfer of particulates from carpet surface to human skin-like receptors revealed that this fabric replicated the human skin, particularly finger pads best as a receptor material. The receptor fabric was attached to an aluminium nose piece with the help of O-ring made of rubber. After the activation of the replicator, the nose piece descended the receptor fabric onto the donor surface and rubbed the receptor against the donor by executing a certain number of motions (imitating an hourglass pattern) under a constant pressure maintained by the programmed hydraulic system. After performing this rub cycle, the nose piece was raised and delivered onto a glass jar containing 20 ml of ethanol. The fabric was released into ethanol and shaken vigorously, followed by sonication for 2 h so that all the particle content is released into ethanol. Then aliquots were removed for assay in an UV–vis spectrophotometer to measure the content of retinyl palmitate. Finally, the amount of transfer of retinyl palmitate was calculated in percentage.

Fig. 1.

All the measurements for shelf life study were performed in triplicates, whereas the measurements of kinetic release study and simulated transfer study were performed in duplicates. The results have been reported as the mean values and their corresponding standard deviations.

Figure 2 shows differential scanning calorimetry (DSC) scans of beeswax, retinyl palmitate and beeswax microcapsules with 25% theoretical loading capacity. In the thermogram of retinyl palmitate, a sharp endothermic peak is observed at 34.33°C, which corresponds to its melting point. However, it is observed that the microcapsules show no endotherms corresponding to the melting point of retinyl palmitate. This implies that retinyl palmitate dissolved within the matrix of beeswax when the temperature reached its melting point (45). This observation was consistent with the result found by Milanovic et al. (46), where encapsulated ethyl vanillin dissolved in the carnauba wax matrix. Untreated beeswax and retinyl palmitate-loaded beeswax microcapsules show their melting peaks at 65.67°C and 63°C, respectively. The slight decrease in the melting point of the particle should be due to the mixing of retinyl palmitate and wax because of plasticisation. A second peak is observed for microcapsules at higher temperature (slightly higher than melting temperature), which could be because of fraction of large crystallites formed after encapsulation process that showed higher melting.

Fig. 2.

Figure 3 shows the shelf life study of the beeswax microcapsules in: (a) powder form stored under room temperature; (b) dispersion form stored in a refrigerator. When the particles were evaluated in powder form under room temperature, the microcapsules lost their active content within 8 days (Figure 3(a)). This phenomenon can be attributed to the diffusion of retinyl palmitate through the wax shell. The high compatibility between lipophilic, low molecular weight active ingredients with wax is the major cause of diffusion (47). Diffusion can be accelerated in small-sized particles due to the availability of larger contact areas as well as due to pores existing in the shell matrix (48).

Fig. 3.

Djordjević et al. (31) described the internal structure of particles produced by melt dispersion with the wax shell to be nonhomogeneous with matrix or hollow-shell morphology. Therefore in the prepared microcapsules, retinyl palmitate is distributed within the wax shell matrix. With the course of time, the core content comes up to the surface and diffuse through the shell. From Figure 4(a), the gradual change in the colour of beeswax microcapsules supports the phenomenon of diffusion as a plausible explanation. The particles stored as powder form appear to be bright yellow after the retinyl palmitate diffuses to the surface and they turn white (beeswax) when almost all of the core content leaches out.

Fig. 4.

On the other hand, when the retinyl palmitate-beeswax particles were stored in the dispersed aqueous emulsion in a refrigerator, they retained the core material and showed no significant decrease in retinyl palmitate content until the 15th day (Figure 3(b)). The variability in size distribution of different batches of filtered particles may account for the slight increase observed in actual loading capacity (Figure 3(b)). After 30 days, a decrease in loading was observed, which can be explained by ester hydrolysis of the beeswax while stored in aqueous emulsion resulting in the release of the content (49). retinyl palmitate-beeswax particles stored in the dispersed aqueous emulsion in the refrigerator do not show a significant visual difference in colour when filtered (Figure 4(b)).

The release profile (Figure 5) of retinyl palmitate-beeswax microcapsule showed an initial burst followed by a slower release of the vitamin entrapped inside the beeswax matrix. Due to the initial burst effect, 7% of the retinyl palmitate released at the first minute, leading to around 55% release in the first half an hour. After the initial rapid release, the release profile showed a sustained release over time. Within 4 h, approximately 98% of retinyl palmitate was released. A similar pattern of release was found by Kheradmandnia et al. (49) from ketoprofen-loaded solid lipid nanoparticles incorporated in the matrix of beeswax-carnauba wax mixture. Zigoneanu et al. (50) described the phenomenon of such initial burst as the result of the cumulative effect of diffusion of the core through the matrix, penetration of dissolution medium into the particle, and degradation of the shell matrix. As retinyl palmitate is soluble in ethanol, this explanation is agreeable to our result. Permeation of ethanol through the pores of the shell matrix and simultaneous diffusion of retinyl palmitate through the matrix facilitated the fast dissolution of the vitamin into ethanol. Duclairoir et al. has reported similar release profile for α-tocopherol from wheat gliadin nanoparticles, where mathematical models were demonstrated for the bistep release, i.e. the burst effect and the slower diffusion process (51). While the initial burst could not be described by their model, the time-dependent slow release showed a good fit (R² = 0.90) for the model in Equation (i):

(i)

Fig. 5.

Here, M₀ is the amount of active content incorporated, M_t is the amount of release core at time t, D is the diffusion coefficient and R is the radius of the particle. Thus the sustained release was related to the diffusivity of the active core inside the matrix system, the surface area of the particle and the loaded content.

From this result, we can understand that alcohol-based cosmetic formulations will not be stable over time as the core content would be released in the carrier substrate during the storage period, making retinyl palmitate susceptible to oxidation and degradation. On the contrary, as we already observed in the shelf life study, an aqueous medium prevents the active content from releasing from the capsule because of having no affinity to the lipophilic content. As a result, a water-based formulation would be suitable to contain the particles for cosmetic applications.

From the transfer study, we found that 21.7±0.02% of retinyl palmitate was transferred to the receptor material from the donor surface of wet nonwoven wipe after the preprogrammed rubbing cycle. The percentage falls within the range reported by Yu et al. in their study of transfer of particulates from carpet surfaces to human skin. Although this amount may vary depending on encapsulation efficiency, method of particle incorporation, and the amount of particle incorporated, this study demonstrates the potential of using such microparticles into facial wipes to impart skincare properties. Knaggs, in his skin-ageing handbook, mentioned that 0.05–0.1% tretinoin (retinoic acid) was effective to reduce signs of ageing in Asians (52). Oliveira et al. demonstrated in their study that topical application 1% retinyl palmitate has promising results for the treatment of skin ageing (53). According to Gangurde et al., the recommended concentration for topical semisolid formulation of vitamin A palmitate is 0.05%–0.3% (27). Thus, considering the approved dosage of retinoids, absorption and conversion rate of retinyl palmitate to retinoic acid within the skin, a proper formulation has to be developed in further study.

Although benzene, which is the hydrogen-lean form of the benzene/cyclohexane LOHC system (Figure 1), is a known carcinogen (the most significant drawback of this specific system), several studies have been conducted into the properties of effective hydrogenation/dehydrogenation catalysts for this model feedstock.

Fig. 1.

Hydrogenation is generally accepted to occur via associative adsorption of benzene onto the catalytic surface. During the hydrogenation step, it is believed the carrier molecules are arranged as ï-complexes and are hydrogenated in a stepwise fashion, through a series of cycloolefins (Figure 2) (44). This is supported with the detection of cyclohexene (44).

Fig. 2.

Early research concluded that monometallic iron-catalysts with body-centred structures were essentially inactive for the hydrogenation of benzene, while the activity of a mixed iron‐cobalt catalyst decreased with increasing iron content (45). Since cobalt and nickel, both with face-centred cubic structures, were found to be active for the reaction, it was assumed the body-centred structure of iron was responsible for its inactivity (45). However, a monometallic, face-centred copper catalyst was deemed to be completely inactive unless in the presence of a suitable promoter, such as nickel (45). The inactivity of the copper catalyst has since been attributed to difficulties in obtaining a suitably high dispersion on the catalyst support, while the mechanism of the promotional effect observed with the addition of nickel remains a topic for debate (35). Several hypotheses for this have arisen, including a changing in copper crystal structure, or an alteration in surface concentration as a consequence of nickel collection (45). Interestingly, even in the very early studies, it was clear that palladium has a high activity for the reaction, but only on the condition that the surface area was sufficiently high, while the conversion is also reported to be catalysed by rhodium(I) complexes (45, 46).

Nickel-based catalysts are frequently considered for ring-hydrogenation reactions, with several detailed studies revealing key information for future catalyst development. For example, the hydrogenation of benzene using nickel catalysts has been suggested to be a structure insensitive reaction, while an unexpectedly low activity at low nickel loadings (less than 5 wt%) is the result of a nickel aluminate spinel formation, when an alumina support is used (47). More commonly however, this spinel structure is formed at high reaction temperatures. Although expected, these findings are important, signifying not only the choice of metal, but also the properties of the support (such as acidity, interaction with the metal) and reaction conditions play a role in determining catalytic performance. The move from pgm-based catalysts to base metal catalysts could also lower the cost associated with the LOHC technology as well as reliance on highly fluctuating costs of pgms (48, 49). This is provided that comparable or even improved catalytic performance can be achieved.

Nickel is also widely referenced for its activity in catalysing the dehydrogenation reaction of cyclohexane. Given the low boiling point of cyclohexane (81°C), this reaction is often performed in the gaseous phase. The studies have shown that when using a cyclohexane feedstock, a bimetallic nickel/platinum combination (20 wt% nickel and 0.5 wt% platinum) can improve catalytic dehydrogenation activity 60-fold when compared to using a monometallic platinum catalyst (0.5 wt% platinum on activated carbon cloth), or 1.5-fold when compared to a monometallic nickel catalyst (20 wt% nickel on activated carbon cloth) (34). In addition, selectivity in the dehydrogenation reaction was improved using various monometallic nickel catalysts with the addition of platinum (50). For these cyclohexane dehydrogenation studies, a spray-pulsed reactor was used at a temperature of 300°C (34).

The concept of improved catalytic performance through the use of bimetallic combinations has also been reported for silver-based catalysts, employed for the benzene hydrogenation reaction (34). Here, the same spray-pulse reactor and reaction temperature (300°C) as discussed for the platinum catalyst above was used. For the most significant catalytic enhancement, it was concluded that the secondary promoter must be classified as a pgm (34). As an example, the hydrogen production rate was doubled when using a platinum-promoted silver catalyst on activated carbon cloth (1 wt% platinum + 10 wt% silver) compared to the non‐promoted 10 wt% silver equivalent (34). This was suggested to be the result of a synergistic effect between the two metals for breaking C–H bonds, high hydrogen reverse-spillover, or hydrogen recombination abilities of the catalyst (51). The improvement in catalytic activity of the silver catalysts promoted with 1 wt% of a pgm were found to follow the trend: platinum > rhodium > palladium, although catalyst stability improved in the order platinum > palladium > rhodium (51). The study did not comment on expected lifetime of the catalysts, highlighting the need for further research regarding key performance indicators (activity, selectivity, lifetime).

To catalyse the hydrogenation of toluene to methylcyclohexane, palladium-based catalysts have been reported to be effective. For example, in 1 h a 10 wt% palladium/carbon catalyst can achieve a 90% toluene conversion at 80°C and 15 bar hydrogen pressure, when the hydrogenation reaction is performed in a batch reactor (52). Interestingly, when this palladium on carbon catalyst was coated with a liquid coordination complex (LCC), the conversion of toluene increased to 99.9%, under otherwise identical reaction conditions (52). LCCs are described as ‘ionic liquid-like’ Lewis acid species, with their equilibrium composition containing cations, anions and neutral components (52). In this experiment, the LCC was synthesised from aluminium chloride and urea (52). It was concluded that only a thin film of the liquid (ionic liquid coating at 13 wt%) was required to improve the catalytic performance; when larger quantities were applied to the catalyst a decrease in activity was observed as a result of pore blocking and mass diffusion limitations (52).

As for benzene hydrogenation, platinum-based catalysts have also been found to be successful in catalysing the hydrogenation of toluene to methylcyclohexane. For example, a 0.3 wt% platinum catalyst supported on zeolite CBV‐780 (silicon:aluminium ratio of 40) is capable of achieving full hydrogenation at just 120°C, with a hydrogen pressure of 30 bar (53). The activity of the catalyst notably decreased when alloyed with palladium. A trimetallic combination of nickel, cobalt and molybdenum on a zeolite support was also found to effectively catalyse this hydrogenation reaction (54). In a batch reactor, with a reaction temperature of 200°C and a hydrogen pressure of 20 bar, it was found that a HY support (silicon:aluminium ratio of 5.1) allowed for superior catalytic activity after 30 min on-stream, when compared to using mordenite, HY (silicon:aluminium ratio of 80) or ZSM-5 alternatives. In explanation, the authors propose the larger pore volume and pore diameter of the HY support (silicon:aluminium ratio of 5.1) reduces pore diffusion limitations (54).

Moreover, monometallic nickel-based catalysts are effective for toluene hydrogenation (55). For example, one study highlighted the increase in catalytic activity observed for a 20 wt% nickel catalyst, compared to a 5 wt% nickel catalyst, both supported on gamma-alumina (55). Since it is generally accepted that an increase in the number of metallic active sites increases catalytic activity, this result is unsurprising (55). However, it was also found that employing supports of different alumina phases (γ-Al₂O₃ or κ-Al₂O₃) has different effects on catalytic activity when different nickel loadings are used. For instance, a significantly higher catalytic activity was reported for a 5 wt% Ni/κ-Al₂O₃ catalyst than for a 5 wt% Ni/γ-Al₂O₃ (toluene conversions of 98% and 28%, respectively), under the same reaction conditions (55). Yet at higher nickel loadings an opposite trend was observed: a 20 wt% Ni/κ-Al₂O₃ catalyst resulted in a lower toluene conversion than that of a 20 wt% Ni/γ-Al₂O₃ catalyst, although the difference in activities between the catalysts is smaller at this higher nickel loading (55). In explanation, it was concluded that changes in reducibility of the surface nickel sites has a larger effect on the observed activity than simply the number of active sites present (55). These reactions were performed in a plug-flow reactor, using 0.1 g of catalyst (55).

Nickel on kaolinite catalysts have also been considered for the hydrogenation of toluene (56). Using a flow-type system at ambient pressures, it was found that modification of the catalyst with a small amount (2 wt%) of either potassium or zinc (using KNO₃ or Zn(NO₃)₂) increased catalytic activity, which was attributed to a decrease in interaction strength between the nickel and kaolinite. Yet, a larger modification (3–7 wt%) resulted in a significant decrease in activity (56). This was proposed to be a result of potassium covering the nickel active sites, or an unfavourable zinc-nickel interaction (56).

The dehydrogenation reaction (methylcyclohexane to toluene, Figure 3) typically employs a platinum or nickel-based catalyst, supported on alumina (1). Reaction temperatures for such catalysts often vary between 350°C and 450°C (with a general increase in temperature increasing the conversion of methylcyclohexane to toluene before equilibrium is reached) (57, 58). Another study highlighted the influence of process conditions on increasing conversion of methylcyclohexane and lowering operating temperatures by using efficient palladium or palladium-alloy membranes in catalytic membrane reactors. Such reactors combine the catalytic dehydrogenation reaction with the extraction of hydrogen in a single unit (58).

Fig. 3.

Interestingly, a potassium-platinum on alumina catalyst was reported to achieve a hydrogen yield of 95%, when employed at 320°C in a fixed-bed reactor, which is slightly above the typical yield range given above (1). Moreover, an excellent selectivity (>99.9%) was reported for this reaction (1). A Raney-nickel catalyst has also been reported to achieve a 65% yield after 30 min on-stream, under multiphase reaction conditions (59). Although this reaction was performed at the lower reaction temperature of 250°C, the Raney-nickel catalyst is not a suitable candidate to catalyse the reactions of the LOHC technology (1). This is in view of the accompanying isomerisation and disproportionation reactions which would ultimately necessitate the need for more frequent LOHC replacement (1). It should be noted that from a thermodynamic viewpoint, the dehydrogenation of methylcyclohexane, like all LOHC candidates, should be performed at the lowest possible pressure as this facilitates the use of the lowest possible temperature for full conversion to toluene (34). However, in practice, it has been found that such conditions may not be compatible with the catalyst, potentially resulting in side reactions and catalyst deactivation (34).

As has already been stated, the catalyst support materials can be modified to increase catalytic performance (34). For the toluene/methylcyclohexane system, it was concluded that the hybrid composite support alumina-titania was responsible for a huge increase in catalytic dehydrogenation activity, when compared to alumina (99% and 16.5% methylcyclohexane conversion, respectively, where nickel is the active site) (34). In addition, properties of the perovskite La_0.7Y_0.3NiO₃, and metal oxides including La₂O₃, CeO₂ and MnO₂ have been evaluated in terms of their suitability as catalyst supports for the dehydrogenation of methylcyclohexane (34, 60). Most notably, it was found the hydrogen production rate more than doubled when comparing the Pt/La₂O₃ and Pt/La_0.7Y_0.3NiO₃ catalysts (21.1 mmol g_met^–1 min^–1 and 45 mmol g_met^–1 min^–1, respectively) (60). These reactions were conducted in a spray-pulsed reactor.

This particular LOHC system has also been studied alongside a palladium membrane reactor, which allows for purification of the hydrogen released from the LOHC (34). Contaminants of the hydrogen can include carbon monoxide, CO₂, methane and cyclic hydrocarbons and have been reported to be present in quantities between 100 ppm and 1000 ppm, depending upon the operating conditions of the LOHC system (61). The exact origins of the impurities remain unknown, but it is believed a combination of atmospheric oxygen and residual moisture in the LOHC is responsible for the formation of carbon monoxide and CO₂, while partial decomposition of the LOHC at high reaction temperatures, or contaminants from the production of the LOHC, can explain the hydrocarbon presence (61). The quality of the hydrogen released depends upon the intended application: for example, several fuel cells have specific regulations on compatible hydrogen purity (61). Despite this, only a few examples of hydrogen purification processes exist in the literature, most of which employ palladium membranes. This is not surprising, considering palladium has a high hydrogen solubility, permitting effective separation of hydrogen and contaminants at high temperatures (typically above 300°C) (61).

A palladium membrane reactor has been reported to have a dual functionality, acting both as a dehydrogenation reactor and a hydrogen purification system (34). The dehydrogenation of methylcyclohexane using such a reactor and a 1 wt% platinum on alumina catalyst, permitted the use of a 20°C decrease in temperature than would otherwise be required to achieve the same conversion (70%) without employing the membrane (225°C and 245°C, respectively) (34). In addition, the platinum catalyst demonstrated good stability and selectivity within the temperature range 150–325°C, with no significant deactivation for approximately 600 h time on-stream (62). A 5 μm palladium-silver membrane, coupled with a microstructured system, has also been reported to effectively purify the hydrogen released (61).

Boasting a high theoretical hydrogen storage capacity of up to 7.4 wt%, a naphthalene/decalin LOHC system has been considered (Figure 4). However, in practice, dilution with a solvent such as toluene is necessary to keep the LOHC cycle within the liquid phase (melting point of naphthalene: 80°C). This lowers the overall hydrogen capacity to 3.8 wt% (1). For this reason, the naphthalene/decalin system is less widely studied than other possible carriers, although important learnings have been gained, which can be transferred to other LOHC systems. It is worth noting that decalin exists as structural isomers, and thus an approximately equimolar mixture of the cis- and trans- isomers are produced upon hydrogenation of naphthalene, despite the thermodynamic favourability of the trans- isomer (1). The exact cis :trans ratio is naturally dependent on reaction conditions, in addition to the catalyst itself (1).

Fig. 4.

Much like other LOHC systems, platinum-based catalysts have also been the focus of most studies on the hydrogenation of naphthalene (1, 63). Yet, in contrast to other systems discussed, an aluminium Mobil Composition of Matter No. 41 (Al-MCM-41) support was investigated (63). Harsher reaction conditions (temperatures of 300°C and pressures of 69 bar) than employed for alternative systems were required to achieve full hydrogenation to decalin in 150 min, in a batch reactor (1). It was found that the reaction temperature could be lowered to 200°C under the same reaction pressure, but this imposed the requirement of a drastically increased reaction time (480 min) (1). Generally, conclusions from this study are similar to those of other systems: hydrogenation activity and selectivity are both strongly dependent on support properties (63).

Although the hydrogenation process presents its relative challenges, it is the dehydrogenation step which confirms the incompatibility of this system with the LOHC technology. For instance, this step is likely to produce intermediates including tetralin (Figure 5). Although tetralin itself can be dehydrogenated and become part of the LOHC process, it is unlikely that no tetralin molecules will remain following the dehydrogenation reaction. As tetralin is still partially hydrogenated, the hydrogen storage capacity of the LOHC is lowered in subsequent cycles. Thus, regular replacement of the carrier material would be required to ensure a constant hydrogen storage capacity (34). This would increase the operating expenditure of the technology (more LOHC required) in the case of naphthalene. In light of these challenges, a pilot scale demonstration is yet to be achieved (1).

Fig. 5.

Research into the dehydrogenation reaction of naphthalene has, however, produced some interesting results. Typically, when employing a platinum catalyst supported on carbon at conditions slightly milder than those used for the hydrogenation reaction (280°C for 150 min at atmospheric pressure), almost full conversion of decalin to naphthalene can be achieved in a batch reactor system. Furthermore, adding rhenium to the catalyst (platinum-rhenium on carbon) can decrease the reaction time from 150 min to 120 min under otherwise identical conditions (1, 64).

A 3 wt% platinum on carbon catalyst was also used to investigate the effect of varying the catalyst preparation method on catalytic performance for the dehydrogenation of decalin (34). It was found that advanced methods of preparation, including ion-exchange and polyol-assisted synthesis (in which ethylene glycol was used as a solvent and reducing agent), resulted in a greater dispersion of platinum (19.6% and 14%, respectively) than the more conventional precipitation and impregnation methods (10% and 5.4% platinum dispersion, respectively) (34, 65). Increasing the platinum dispersion increases the surface area of the metal over which the reaction can occur, consequently resulting in a greater rate of hydrogen release from decalin (34, 65). Again, optimisation of the support must also be considered. Several carbon-based possibilities including nanofibers, carbon black, carbon xerogel and ordered mesoporous carbon have been evaluated, with the high surface area of ordered mesoporous carbon believed to be responsible for demonstrating the highest activity for the dehydrogenation of decalin (34). Yet, over longer operation times catalyst deactivation was observed as a result of pore blockages with the bulky LOHC feedstock (34). Thus, it can be argued that as carbon black has wider pores than the ordered mesoporous carbon, this would be the most suitable support, despite a lower activity. These experiments were performed at 260°C, in a batch reactor (34).

Catalytic activity, in addition to selectivity, has also been reported to be improved with the addition of tin to platinum on activated carbon (66). The effect is two-fold: (a) the electronic modification of platinum by tin prevents the cleavage of C–C bonds on the catalyst surface, facilitating the adsorption and desorption of reactant and product molecules; (b) the addition of tin as a catalytic promoter improves catalyst stability by preventing the sintering and agglomeration of platinum at high temperatures (34).

The most significant advantage of the NEC/perhydro-NEC LOHC system (Figure 6) is the relatively low energy required for dehydrogenation (53.2 kJ mol^–1 hydrogen, Table I) (1). This lower reaction enthalpy facilitates a lower temperature for the dehydrogenation reaction and consequently improves the overall efficiency of the system. As such, interest into the potential applications of NEC has significantly grown in recent years, culminating in its use as a feedstock for the LOHC technology. However, like other LOHC candidates discussed thus far, NEC also possesses unfavourable properties for its employment within LOHC technology. Arguably the most significant of these is the solid nature of NEC at ambient temperatures. To ensure the key liquid property of the LOHC technology is not lost, employing NEC would thus require NEC dilution (1). This reduces the efficiency of the LOHC technology (1). Moreover, NEC is significantly more expensive than alternative LOHCs, such as toluene (€40.00 kg^–1 and €0.30 kg^–1, respectively) (1).

Fig. 6.

The dehydrogenation reaction is again most commonly performed with pgm-based catalysts, particularly palladium on alumina or platinum on alumina at metal loadings of around 5 wt%. These catalysts have been reported to have the highest catalytic activities, when compared to ruthenium and rhodium equivalents (67). Specifically, catalyst dehydrogenation activity follows the trend palladium > platinum > ruthenium > rhodium under atmospheric pressure and at 180°C (67). Full dehydrogenation to NEC was observed for palladium and platinum catalysts, after reaction times of 240 min and 300 min, respectively, while employment of a ruthenium catalyst over the same time frame produced a mixture of both the fully dehydrogenated NEC (71.28%) and the partially hydrogenated species, 4H-NEC (28.54%) (67). Using the catalyst with the lowest dehydrogenation activity (rhodium on alumina), even less of the fully dehydrogenated NEC is obtained (10.64%) (67). Important to note, all reactions are carried out below 270°C. Beyond this temperature, NEC becomes susceptible to dealkylation reactions, producing the byproduct carbazole (1, 68).

Like the dehydrogenation reaction, the hydrogenation of NEC is frequently catalysed by a pgm. However, given the comparative ease of the hydrogenation reaction (in comparison to the dehydrogenation reaction), a palladium or ruthenium-based catalyst is more frequently employed than for the dehydrogenation reaction. For example, full hydrogen loading onto NEC has been achieved in 3 h at 150°C and 50 bar hydrogen pressure using a 5 wt% ruthenium on alumina catalyst in a batch reactor (68). The molar ratio of LOHC to ruthenium used in these experiments was 400:1 (68). The catalyst Pd₂Ru on silicon carbonitride has also been reported to successfully catalyse the hydrogenation of NEC (69). In comparison to the 5 wt% ruthenium on alumina catalyst discussed previously, a milder reaction temperature (110°C) and hydrogen pressure (20 bar) can be employed (69). However, a significantly longer reaction time (36 h) is required for full hydrogenation of NEC to perhydro-NEC, and a greater amount of active metal is also required, compared to the ruthenium on alumina catalyst described above (0.52 mol% and 0.25 mol%, respectively) (68, 69). The same Pd₂Ru@SiCN catalyst has also been reported to successfully catalyse the dehydrogenation of perhydro-NEC at 180°C in a reaction time of 7 h, in a batch reactor (69).

More recently, full hydrogenation has also been attained using ruthenium supported on a rare earth hydride catalyst, Ru/YH₃ (70). Reporting both the mildest conditions (100°C, 10 bar) and the highest catalytic activity for the hydrogenation of NEC to date, it is clear that such rare earth supported catalysts have potential for LOHC applications (70). This was highlighted by concluding the high stereoselectivity of the Ru/YH₃ catalyst for the all-cis product would be advantageous for any subsequent dehydrogenation reactions, since the cis product is more easily dehydrogenated than the trans equivalent. In explanation, the authors propose the cis product is less sterically hindered on the catalyst surface (70). However, despite an excellent catalytic performance and satisfactory stability, a high hydrogen pressure is required to achieve high selectivity (70). Manufacturing costs, handling and scale-up considerations were not discussed and may present challenges for such a material.

The Ru/YH₃ catalyst has also been reported to effectively catalyse the hydrogenation of another N-heterocycle: 2-methylindole, which has been suggested as a suitable LOHC candidate (70, 71). However, much like NEC, 2-methylindole is a solid at ambient temperatures and possesses a slightly lower hydrogen storage capacity than NEC (5.7 wt% and 5.8 wt%, respectively) (71).

To achieve an effective and efficient hydrogen storage and transportation system, the choice of LOHC is of critical importance. For commercial implementation of the LOHC technology, DBT can be seen as the most suitable candidate, considering its key advantages: relatively low-cost, low toxicity and high hydrogen storage capacity (6.2 wt%, Table I). In addition, DBT is already mass produced as it is used as a heat transfer agent. As a result, there is a vast exploration of DBT as a potential carrier molecule in the literature.

Several catalysts have been evaluated for the hydrogenation and dehydrogenation reactions of the DBT/perhydro-DBT system (Figure 7). Much like NEC, DBT hydrogenation is also typically catalysed with an alumina-supported pgm, namely platinum or ruthenium (1). As an example, in a batch reaction, a ruthenium on alumina catalyst can achieve full hydrogen loading of DBT in 4 h at 150°C with a 50 bar hydrogen pressure, which is longer than that of NEC (3 h) under the same conditions (68).

Fig. 7.

The more in-depth studies on this feedstock have also revealed that the catalytic hydrogen-loading does not have to be performed with pure hydrogen. As DBT hydrogenation is selective, and provided the catalyst is not negatively affected by the presence of other components, a mixed gas stream (for example, including methane, CO₂) can be used (34). From an industrial perspective this is very attractive, enabling otherwise low-value hydrogen present in waste gas streams, from processes including reforming and gasification reactions, to be stored and transported. The first example of such a process was reported in 2017 by Dürr et al., in which a mixture of hydrogen and methane from the decomposition of methane (obtained from offshore drilling) was fed directly to the hydrogenation unit containing the DBT feedstock (72). Separation of the loaded (hydrogenated) LOHC and gaseous methane can then be facilitated. It has also been reported that the presence of methane does not negatively affect the hydrogenation or dehydrogenation of DBT/perhydro-DBT (30). In contrast, methane was remarkably found to slightly improve the hydrogenation rate, postulated to be a result of a lower DBT viscosity and thus improved hydrogen mass transport (34).

In a similar vein, hydrogenation of DBT has been investigated using a hydrogen/CO₂ stream (up to 30% CO₂) (73). However, with such a hydrogenating mixture, both methanation (CO₂ to methane) and reverse water gas shift (CO₂ to carbon monoxide) processes were found to occur, with the extent of such side-reactions being strongly dependent on the catalyst employed (34). Even with the most promising palladium on alumina and rhodium on alumina catalysts, the degree of hydrogenation in a batch reactor reached only 0.8 and required elevated temperatures (210°C and 270°C for the rhodium and palladium catalysts, respectively) when compared to using a pure hydrogen stream (34). For the rhodium catalyst, moderate methane formation (methane:CO₂ ratio of less than 0.1) can be achieved at temperatures between 120°C and 150°C (rhodium on alumina), while the palladium catalyst showed a lower selectivity for methane formation, with a methane:CO₂ ratio of less than 0.1 between 120°C and 270°C (palladium on alumina) (34, 73).

Despite the frequent employment of platinum and ruthenium catalysts with pure hydrogen streams, these catalysts were found to be ineffective for the selective hydrogenation of DBT using a hydrogen and CO₂ stream (34). Their respective unsuitability was concluded to be a result of platinum catalyst poisoning from the reduction product carbon monoxide formed via reverse water gas shift of CO₂ and facilitation of CO₂ methanation by ruthenium.

Palladium-based catalysts have low activities for the dehydrogenation reaction of perhydro-DBT, and have thus been identified as catalysts to avoid (34). Under the same reaction conditions (270°C, 3.5 h, batch reactor), a stark difference in the degree of dehydrogenation between a palladium on carbon (5 wt%) catalyst and its platinum equivalent has been observed (16% and 55%, respectively) (68). In a different study, a lower metal loading has proven to be advantageous (lowering the metal content of the platinum on carbon catalyst to 1 wt% from 5 wt% increased the degree of dehydrogenation from 55% to 71%), while the support preference for this particular dehydrogenation reaction was found to follow carbon > alumina > silica (34, 68). In these experiments, the catalysts were used in batch reactions in quantities of 0.15 mol% with respect to perhydro-DBT (68).

Interestingly, the wider research around the use of DBT as a LOHC has included a study in which the same platinum on alumina (0.3 wt%) catalyst has been used for both hydrogenation and dehydrogenation reactions (34). Under the same reaction temperature (291°C), Jorschick et al. reported that the two reactions could be interchanged by varying the pressure between 1.05 bar (dehydrogenation) and 30 bar (hydrogenation) in a hot pressure swing reactor (30). However, the use of a non‐optimised catalyst for the dehydrogenation process resulted in a lengthy reaction time (20 h), when compared to the examples discussed previously. In addition, it must be noted that the productivity of the dehydrogenation (and thus catalyst activity) decreased significantly over the first cycle, before stabilising for the following three cycles. This was concluded to be a consequence of bulky, high boiling point byproduct formation via thermal cracking of DBT. Such molecules, including diphenylmethane and 2,6-dimethyldiphenylmethane, were found to increase per cycle and can be expected to block active sites of the catalyst (34, 74). It has also been suggested that some of the possible byproducts of the dehydrogenation reaction are unstable and thus undergo cracking during a subsequent hydrogenation reaction. Such an example would be the conversion of perhydromethylfluorene to perhydrobenzyltoluene (74).

DBT exists as structural isomers and consequently, more than 24 stable intermediates can be found in a partially hydrogenated solution. These can be categorised into four main groups: DBT (H0-DBT), hexahydro‐DBT (H6‐DBT), dodecahydro‐DBT (H12‐DBT) and octadecahydro‐DBT (H18-DBT) (75). Laboratory studies regarding selectivity would require advanced analytical detection methods. DBT conversion can be studied using ultraviolet-visible spectroscopy (i.e. the degree of ring hydrogenation, providing information about conversion levels only), but detailed studies probing catalyst selectivity would be challenging, requiring either a complex high-performance liquid chromatography (HPLC) method or ¹H nuclear magnetic resonance (NMR) analysis (29, 76, 77).

As such, alternative model carriers are often sought for fundamental academic studies aimed at deepening the chemical understanding of how to improve the performance of catalytic materials for the LOHC technology. In addition to being easily characterisable, the ideal carrier would have high boiling points and low melting points of both hydrogen-rich and hydrogen-lean molecules. This enables the whole cyclic process to be carried out in the liquid phase. A high melting point would result in the formation of a solid at ambient temperatures, with a dissolving process or incomplete hydrogen-unloading significantly lowering storage efficiency. Whereas a low boiling point would require extra economic expense in gas condensation equipment. Frequently, toluene is employed as such a model compound. Given its structural similarities to the central motif of DBT, this is a logical choice. Yet, the environmental health and safety hazards associated with its usage are quite significant, being noted for serious concern regarding both human and aquatic toxicity (23). Thus, to lower the hazards associated with the feedstock, other aromatic compounds similar in structure to toluene might be chosen.

Energy-intensive industries as well as the transportation sector contribute significantly to global greenhouse gas (GHG) emissions (1). To mitigate climate change and achieve the goals of the Paris Agreement (2), it is necessary that each sector develops pathways towards GHG emission reductions and accelerate the transition towards deep decarbonisation. The production and use of a low-carbon hydrogen is seen as a ground-breaking aspect of a low carbon future, especially for hard-to-decarbonise sectors (3–14). A significant amount of renewable electricity will be required to enable energy-intensive industries and the transportation sector to reduce their emissions and meet decarbonisation goals when deploying green hydrogen produced via water electrolysis using renewable energy sources (11). Given that capacities of renewables and electricity costs for the production of green hydrogen are extremely heterogeneous (15–18), it is expected that the production of green hydrogen in all required locations at adequate costs will be challenging in the future. Similar to fossil fuels, which are imported and exported across the world, geographical locations with high renewable potential and low costs of electricity are expected to be focal points for the production of green hydrogen. Robust hydrogen storage and transportation systems are among the key components in the successful transition from fossil fuel-based energy systems towards hydrogen-based alternatives (19–23). Storage and transportation of hydrogen at scale are yet to be addressed. An innovative method for long-distance transport and long-term high density hydrogen storage is to use LOHCs. This process is a two-step cycle, which is based on loading of hydrogen via catalytic hydrogenation into LOHCs, such as unsaturated organic compounds, followed by unloading of hydrogen via catalytic dehydrogenation after transport and storage (24–27).

In our previous work we provided an overview and perspectives on the LOHC technology among different hydrogen storage and transportation technologies (27). This study describes the potential deployment and integration of LOHCs within different industries. These include: the transportation sector (automobiles, ships, trains); steel and cement industries; the use of stored hydrogen to produce fuels and chemicals from flue gases; a system integration of fuel cells and LOHCs for energy storage.

Renewable sources, typically wind or solar, provide the energy required for the electrolysis of water to produce hydrogen (26). For the electrolysis stage, polymer electrolyte membrane (PEM) electrolysis is the preferred method, due to the ability of the system to respond to the characteristic fluctuations in renewable energy power supplies (25). The hydrogen produced can then be stored in LOHCs (through the hydrogenation step, Figure 1).

Fig. 1.

Interestingly, it has been reported that when dibenzyltoluene is employed as the LOHC, the hydrogen produced via electrolysis does not need to be dried before the catalytic hydrogenation reaction (28). When a ruthenium-based catalyst was used in a pellet form rather than as a powder, the activity of the catalyst was found to be virtually unaffected by water, while only a small decrease in hydrogenation activity was recorded when employing platinum catalysts with wet hydrogen (28). Costs associated with energy-intensive hydrogen drying processes can therefore be avoided. To better understand this observation, the tolerance of the catalyst should be investigated with other LOHC candidates.

When required, the stored hydrogen can be released from the LOHC (through the dehydrogenation step) and converted back into electricity, again using fuel cells. In this step, either PEM fuel cells (PEMFCs) or solid oxide fuel cells (SOFCs) can be utilised. PEMFCs are better designed to produce variable quantities of energy and thus meet changes in energy demand (25). Thus, the SOFC technology is only advantageous when there is a constant electricity demand (25). Nonetheless, using the SOFC technology to convert green hydrogen back into electricity is widely believed to become a future conventional method for stationary, green electricity production (29).

Furthermore, the waste heat from SOFCs (operated at 600–1000°C) is of the correct level to allow its use in the dehydrogenation step of the LOHC process (25). The coupling of LOHC technology with SOFCs can therefore improve the overall efficiency of the LOHC technology. To emphasise this, scale-up calculations for a LOHC–SOFC integrated system predict that 1 kg h^–1 of hydrogen is capable of producing around 18.04 kW of power, corresponding to a SOFC efficiency of 54.1% (29). Another study suggests that an overall electrical efficiency of 45% is achievable with a 10-year SOFC lifespan (30). Such a long lifespan can be maintained if LOHC vapour does not damage the SOFC (31).

In contrast, a combination of LOHC systems with PEMFCs would require heat from another source, such as burning a portion of the hydrogen produced, to facilitate the dehydrogenation reaction (25). This is because the waste heat of PEMFCs (below 180°C) is much lower than the heat required for the endothermic dehydrogenation process needed for hydrogen release. Hence an integrated LOHC–PEMFC system would not be as efficient as a LOHC–SOFC equivalent (31).

Interestingly, the SOFC technology has also shown promise when combined with a mixed LOHC feedstock (29). In a temperature cascade dehydrogenation process, a eutectic mixture of N-ethylcarbazole (NEC), N-phenylcarbazole, ammonia and biphenyl-diphenylmethane increases the energy generated per unit mass of the LOHC (kilowatt hour per kilogram LOHC) by 1.3–2 times, compared to an individual LOHC (29). The system integration of the LOHC technology, fuel cell technologies and green hydrogen produced via electrolysis using electricity from intermittent renewable sources, can thus enable the local storage of excess energy or renewable electricity. Such a process is entirely sustainable as no harmful emissions are produced.

The concept of system integration can be used as a safety device for the storage of electricity on a large-scale, such as the National Grid, UK. There are currently several reports detailing the problem of electricity fluctuation in the National Grid, due to the lowered electricity demand during the COVID-19 pandemic (32). Despite a rise in the percentage of people working from home, resulting in an increased demand for domestic electricity, an overall decrease in electricity demand was observed (32). This can be attributed to the closure of non-essential offices, schools, hospitality and leisure venues. As a consequence, measures such as temporarily shutting down flexible windfarms are expected to be taken, in order to lower the excess of energy in the National Grid (32). Too much electricity in the National Grid is equally as concerning as too little electricity, as the rise in frequency increases the potential to damage infrastructure (33). As the proportion of renewable energy in the National Grid will predictably increase in the future, the fluctuations in electricity are also expected to increase. An energy storage technology, such as LOHC combined with fuel cells, could therefore be extremely beneficial. Similar would apply to ammonia and methanol, where such hydrogen carriers can be deployed in combination with fuel cells.

The above-mentioned energy storage ability of the LOHC technology has been reported to have implementation potential in residential and commercial buildings (localised energy storage) (Figure 2) (34, 35).

Fig. 2.

For such an application, heterocyclic aromatic compounds, such as NEC, have been considered. Although the fully dehydrogenated form is a solid at room temperature, which limits dehydrogenation to 90% and reduces the efficiency of the overall system, NEC is a safer feedstock than other possibilities (for example a toxic toluene/methylcyclohexane system). Thus, the requirements for domestic implementation are better accommodated (34).

In contrast to an application in which the loaded-LOHC is not stored in the vicinity of intended use, a ‘decentralised energy storage’ system is proposed. This technology also uses fuel cells to meet the local demands for electrical energy but has the added economic advantage that any unused energy can be sold back to the National Grid, for example. Furthermore, provided the building has such a resource as solar panels fitted onto the roof, the potential for a completely self-sufficient system exists, while any waste heat generated from the fuel cells, electrolyser and exothermic hydrogenation process may be used to heat the building (34). Yet, with the requirement of fuel cells, electrolysers, hydrogenation and dehydrogenation units, in addition to LOHC storage tanks, the physical space requirement and initial economic investment are high (34). It has, however, been suggested that houses already possessing a crude-oil storage tank may avoid the requirement for a subsequent tank (34).

The transformation of green hydrogen back into electricity is just one example which demonstrates how the LOHC technology can facilitate energy storage. Green hydrogen, stored and transported in LOHCs, can also be used as a ‘green’ feedstock for the synthesis of fuels and chemicals. Moreover, green hydrogen production and storage is a vital part of most carbon capture and utilisation (CCU) technologies which are focused on capturing CO₂ from either air or industrial flue gases and converting it into chemicals or fuels. For the moment, the hydrogen required for CCU technologies is considered to be generated from sustainable energy resources, however detailed integrated processes have often not yet been developed. The availability of green hydrogen for CCU processes is limited by competing demands such as hydrogen used in fuel cells for the transportation sector, and hydrogen used as domestic and industrial fuel supply.

Green hydrogen production is a key accelerator of CCU for production of chemicals and fuels at a commercial scale. Therefore, the production of hydrogen via electrolysis and its storage and transportation using LOHCs (or other hydrogen carriers such as ammonia or methanol) can be viewed as an integral part of sustainable chemicals and fuels manufacturing. One of such examples is methanol production, with a global production capacity of around 85 million metric tonnes in 2016, which is expected to rise in the coming years (36). Conventionally, methanol is synthesised using synthesis gas (syngas) produced from fossil fuels. However, a move towards the use of renewable hydrogen for sustainable methanol synthesis, using CO₂ captured either from the air or from industrial flue gases, would enable a reduction of global GHG emissions.

Thus, a scenario under which low-cost green hydrogen (i.e. produced at locations with an abundant energy supply and so cheaper electricity) is transported using LOHCs to the industrial sites (for example, cement, steel, refinery industries) with high CO₂ emissions to produce sustainable chemicals and fuels, might become viable in the future. Nevertheless, technoeconomic assessments and market penetration studies are required in order to understand under which circumstances this scenario can be realised.

In the coming years, the proportion of electricity to be obtained from renewable sources is expected to increase. By 2024, it is predicted as much as 30% of the global electricity demand will be met with renewable energy sources: an increase of 4% in four years (37). The expected 15–30% decrease in the cost of solar power within the same time frame is also expected to accelerate the growth of renewable energy generation sites (37). However, the fluctuations in renewable electricity supply make it unreliable for direct industrial use. Integration of the LOHC technology within a cement plant has thus been studied as method of energy storage, which can be utilised to equalise the plant’s energy output (Figure 3) (38).

Fig. 3.

The electricity for a cement plant can be supplied from renewable sources. During favourable conditions, the excess electricity is converted into hydrogen via electrolysis, followed by loading the hydrogen onto the LOHC and storing it. The dehydrogenation reaction is then performed at times of insufficient power supply from the renewable sources. The hydrogen released can be converted into electricity using fuel cell technology, a combustion engine or turbine (38). If an adequate power supply still cannot be achieved with the added electricity from the storage, the plant can be connected to the National Grid.

The higher temperatures required for the endothermic dehydrogenation reaction (i.e. hydrogen unloading) is a bottleneck of the LOHC technology, with the requirement of an external heating source resulting in a lowered system efficiency. In order to improve the efficiency of the LOHC system, the heat required for the dehydrogenation reaction can be coupled to the waste heat of a cement plant (Figure 3) (25). Arguably, this is a more suitable solution than the coupling of the dehydrogenation reaction to the waste heat from the exothermic hydrogenation reaction for heat recovery, as this is typically not of the same temperature level as is required for the dehydrogenation reaction. The waste heat from a cement plant, however, is more suitable (temperature level of up to 600°C) for the dehydrogenation process. Attractively, this avoids the need to reduce the efficiency of the process by burning a portion of the released hydrogen and there are no extra costs associated with employing an external heating source (25, 38).

Integrating a LOHC system with a cement plant allows an overall reduction in the working electricity cost of a cement plant. For a plant with an average power demand of 12.5 MW, it has been found that converting the hydrogen released in the dehydrogenation reaction to electricity and using this to power the plant would reduce electricity costs by around €1 million per annum (38). Although this saving has not been expressed as a percentage of the total electricity cost, it has been suggested that successful coupling of wind energy and the LOHC technology to a cement plant would achieve amortisation in fewer than 10 years. This is on the condition that investment costs are kept to a limit of €3.5 million. In addition, it should also be noted that the savings in electricity costs are calculated on the assumption that a thermal energy storage system is also installed, which ensures hydrogen can be released from the LOHC even in times of lower exhaust heat from the cement plant (38). In the absence of such a thermal storage system, fluctuations in the temperature of the waste heat must be provided for with the addition of a hydrogen burner (38).

The iron and steel industry is responsible for an annual output of approximately 2.5–3.0 Gt_CO2 year^–1, with up to 10% originating from within the European Union (EU). This represents 6% of total global CO₂ emissions, and 16% of total industrial CO₂ emissions. To reach the EU climate targets, the iron and steel industries must decrease their CO₂ emissions by up to 90% by 2050. Several processes are being explored to reduce CO₂ emissions from the steel industry. They can be broadly divided into two categories: carbon-based (coal- or natural gas-based) and hydrogen-based steel production (39). In carbon-based steel production, the residual gas emissions from the iron and steel industry can be transformed into valuable products, such as fuels or chemicals, or captured and stored or both. Hydrogen-based technologies, which use hydrogen as the reducing agent instead of carbon, avoid carbon emissions altogether, provided that hydrogen used in these processes is carbon-free hydrogen, produced by electrolysis of water using renewable electricity.

Numerous steel manufacturers have started to explore hydrogen-based technologies. As an example, voestalpine AG, Austria, has set up a goal of direct avoidance of CO₂ emissions in their steel manufacturing over the coming years by moving towards the use of green hydrogen for steel production (i.e. direct reduction of iron (DRI)). To this end, voestalpine together with their project partners have commenced the production of green hydrogen at the voestalpine premises in Linz, Austria within the framework of the EU-funded project called H2FUTURE (40, 41). In this project, the proton exchange membrane electrolysis technology is demonstrated on an industrial scale (6 MW), simulating rapid load changes in electricity generated from renewable energy sources and from electric arc furnace steelmaking (grid balancing). Thyssenkrupp Steel Europe AG, Germany’s biggest steelmaker, is also looking into using hydrogen for steel manufacturing. RWE AG, a German multinational energy company, and Thyssenkrupp Steel Europe AG have agreed to collaborate towards a longer-term hydrogen partnership to supply green hydrogen for steel manufacturing (42, 43). RWE plans to build a 100 MW electrolyser which can produce 1.7 tonnes of hydrogen per hour for Thyssenkrupp Steel Europe AG. This could potentially cover 70% of the quantity required by the Thyssenkrupp steelmaker’s blast furnace in Duisburg, Germany.

Another example is a joint venture between SSAB, LKAB and Vattenfall, all in Sweden, within the framework of the Hydrogen Breakthrough Ironmaking Technology (HYBRIT) project. The aim is again to reduce CO₂ emissions and decarbonise the steel industry by replacing coal with hydrogen in the steelmaking process to produce fossil-free steel at Sweden’s pioneering fossil-free steel production plant (Figure 4) (44–46).

Fig. 4.

Interestingly, the HYBRIT project faces two major challenges: (a) to develop an effective process to use 100% hydrogen on an industrial scale; (b) to produce hydrogen in an energy efficient way that is economically justifiable and commercially viable (46). To this end, the HYBRIT project has recently invested SEK 200 million (£17.5 million) in a pilot plant for storage of green hydrogen in Lulea, Sweden near the SSAB steel manufacturing site (47). The implementation study for the HYBRIT initiative has recognised the need for large-scale storage of hydrogen. To ensure an even flow of green hydrogen produced from renewable energy for steel manufacturing at the SSAB site, a large-scale hydrogen gas storage facility is required to balance the electricity system with an increasing proportion of weather-dependent power generation. It is expected that the integration of large-scale storage of green hydrogen to a fossil-free value chain for steel manufacturing will allow the fossil-free steel to be price competitive.

In the HYBRIT project, the 100 m³ subterranean (25–35 m underground) pilot hydrogen gas storage facility is built in a bedrock cavern with a steel lining as a sealing layer to store pressurised hydrogen (47). In contrast to SSAB, other steel manufacturing companies might not have the facilities for the underground storage of green hydrogen. Furthermore, geographically some of the steel manufacturing sites might not have access to low-cost surplus renewable electricity. In addition, if they are located in energy demanding industrial districts, numerous sectors will compete for green hydrogen or renewable electricity to reduce their CO₂ emissions in the near future. To ensure a competitive production cost for the fossil-free steel (i.e. DRI technology), an effective hydrogen storage and transportation technology, such as the LOHC technology, might thus be required to allow a reliable and on-demand supply of green hydrogen.

As large-scale storage of hydrogen is an important part for a fossil-free value chain for steel manufacturing, it should become an integral part of the DRI technology. For the deployment of the LOHC technology in the steel industry, two scenarios can be foreseen. One is an onsite hydrogen production and storage in LOHCs. Such a scenario can be realised when a steel manufacturing site has access to surplus renewable electricity from intermittent resources to produce hydrogen that should be stored to balance the electricity. Another one is using LOHCs to transport hydrogen, which is produced at geographical locations with high availability of renewable electricity, to a steel manufacturing site with low availability of renewable electricity but high demand for green hydrogen.

BMW AG, Germany, has been developing automobiles that employ hydrogen technology for around 40 years (48). The first model built using such a technology was named the BMW Hydrogen 7, and comprised the storage of liquid hydrogen in a cryogenic tank. However, during its early development serious technological challenges were realised (48). For instance, sufficient hydrogen storage space must be provided to enable longer-distance travel, but size and weight limitations for a practical motor vehicle must also be considered. In addition, the safety risks associated with burning hydrogen in the internal combustion engine must also be adequately minimised. As only 100 vehicles of the Hydrogen 7 model were ever released, it can be deduced that the disadvantages of the technology outweighed the advantages (48). More recently, BMW has announced a partnership with Toyota Motor Corporation, Japan, to develop a fuel cell-based system suitable for integration within its motor vehicles (49). It is predicted that the new model will be commercially available by 2022 (49).

As an alternative for onboard liquid hydrogen storage in mobility applications, the LOHC technology has been suggested (Figure 5) (50). Advantageously, onboard storage of hydrogen in the LOHC would resemble that of gasoline and diesel (liquid state of the LOHC at ambient pressures), which is widely understood, and the safety hazards of storing high-pressured hydrogen are removed (48). Moreover, a range of 500 km is reportedly achievable using 100 l of NEC loaded with hydrogen (equivalent to 5 kg of hydrogen) as the LOHC (48).

Fig. 5.

For motor vehicles, the loaded LOHC would be transported to the refuelling station, before a subsequent release of hydrogen from the carrier in a catalytic dehydrogenation reaction occurring onboard the vehicle (50). In contrast to diesel and gasoline fuels, the dehydrogenated LOHC would not be consumed, but stored within the vehicle until replaced with new hydrogen-loaded material at a designated station (48). The unloaded LOHC can then be transported back to a hydrogenation site and reloaded. Therefore, either two tanks, or a tank capable of separating the loaded and unloaded forms of the LOHC, is required. The hydrogen produced could then be used in an internal combustion engine or combined with fuel cell technologies. Despite the burning of a portion of hydrogen to meet the working temperature of a typical fuel cell, it has been predicted that the overall efficiency would still be higher than that of a combustion engine (48).

Highlighting the attractiveness of implementing the LOHC technology within the mobility sector, Hyundai Motor Company, South Korea, has recently announced plans to develop an onboard LOHC storage technology (51). This encompasses a partnership with Hydrogenious LOHC Technologies GmbH, Germany, who will supply dibenzyltoluene to be used as the LOHC. Initially, the technology will be introduced in South Korea, before being extended to the European market (51). It is expected that the development of LOHC compatible automobiles will raise the profile of the technology as an important tool in the transition to the hydrogen economy.

The use of hydrogen in maritime applications is also an active research area aimed at reducing the pollution from the maritime industry. One of the major challenges is the storage of hydrogen on board of ships (52). The LOHC technology is seen as one of the potential hydrogen storage solutions (52–54). For instance, the use of a double chamber tank system has been proposed, which is capable of separating the loaded and unloaded carriers during the fuelling process into different sections of the tank (53). Nevertheless, providing a suitable level of heat to facilitate the dehydrogenation reaction remains problematic.

Similarly, the use of LOHCs in trains has also been considered (55). Since a significant portion of trains are currently operated using diesel, or a combination of electricity and diesel, future environmental-focused objectives are likely to concentrate heavily on finding alternatives to these fuels (55). Although trains can be powered electrically with renewable electricity to meet zero-emission transportation goals, building the infrastructure of overhead wiring is relatively expensive. Currently, 42% of the UK railway routes are electrified and can become zero-carbon when using renewable electricity (56). The remaining 58% still rely on diesel (56). An alternative approach to electrification is the use of hydrogen fuel cells to generate electricity onboard to power trains (57). Hydrogen powered trains have a potential to revolutionise railway operations in Europe (56).

Using LOHCs for onboard hydrogen storage, coupled with hydrogen fuel cells for electricity production to power trains would avoid the production of hazardous emissions (i.e. CO₂, soot, nitrogen oxides), while still permitting long-distance travel, an essential criterion for trains. In a recent study, the LOHC technology (with dibenzyltoluene as the LOHC) was found to be a very promising option for hydrogen storage, transport and release and can be combined with electricity generation by hydrogen fuel cells to power trains (55). The choice of dibenzyltoluene as a LOHC was influenced by its favourable properties, such as low flammability, low toxicity and liquid-state within the range of hydrogenation and dehydrogenation temperatures, in addition to its commercial availability as a heat-transfer oil (55). Notably, this technology has been supposed to be favourable over alternatives, such as batteries, which are typically characterised by low energy densities (55). Furthermore, the hydrogen fuel cell technologies required for the integration with the LOHC technology for onboard hydrogen storage are expected to become a lower cost alternative to battery and diesel options in the second half of the decade.

Even though the studies into hydrogen-powered aviation are somewhat immature in comparison to trains and cars, it is anticipated that sustainable aviation will quickly become a central research focus in the coming years. Like other mobility sectors, hydrogen is again expected to be named as a primary energy source for propulsion in the aviation industry. This can be realised either through powering aircrafts with electricity generated through hydrogen fuel cells, direct burning of hydrogen in gas turbines or using hydrogen as a building block for production of sustainable synthetic aviation fuels (58). Notably, Airbus has already revealed three hydrogen-powered planes to lower aircraft emissions, which comprise the use of hydrogen gas-turbine engines (59, 60). In addition, Airbus SE, The Netherlands, has postulated the combined use of hydrogen fuel cells with a gas-turbine engine to create a ‘highly efficient hybrid-electric propulsion system’ (61). The urgency of this sustainable transition has also been emphasised by the Swedish government. A mandatory reduction in GHG emissions, originating from aviation fuels, will be introduced for fuel to be sold in the country from 2021 (62). This will start at a 0.8% reduction in 2021 and reach 27% by 2030, in preparation for reaching their national fossil-free target in 2045 (62).

In summary, the LOHC technology is an attractive solution for the storage and transportation of hydrogen to allow a reliable and on-demand hydrogen supply, enabling industrial decarbonisation. The potential deployment and integration of the LOHC technology within different industries, such as the transportation sector, steel and cement industries, the use of stored hydrogen to produce sustainable fuels and chemicals from flue gases, and a system integration of fuel cells and LOHCs for energy storage, is depicted in Figure 6.

Fig. 6.

The possibility of deployment and integration of LOHC systems within different industries is reviewed in this study. These include: the transportation sector, steel and cement industries, the use of stored hydrogen to produce fuels and chemicals from flue gases and system integration of fuel cells and LOHCs for storing renewable electricity. An effective system integration of the LOHC technology with different industries might help with the cost reduction of the LOHC technology, when for example, waste heat is used for dehydrogenation of LOHCs. Importantly, the deployment of the LOHCs for storage and transportation of hydrogen to allow a reliable and on-demand hydrogen supply might enable energy-intensive industries to reduce their emissions and meet decarbonisation goals.

Numerous possibilities for the deployment and integration of LOHCs within different industries might necessitate the use of different LOHC carriers in each instance. While a carrier choice offers a large amount of flexibility in the LOHC technology, the myriad of possible carriers and catalysts combined with reactor technologies might be considered as one of the factors impeding the integration and commercial deployment of the LOHC technology across different industries. Customer-tailored solutions and offerings might need to be developed to accommodate specific requirements. A review of the most prominent LOHC systems, focusing on properties of LOHCs and catalytic materials used for hydrogenation and dehydrogenation of LOHCs, is presented in our following work dedicated to the analysis of LOHC systems (63).