Micron = micrometer

• Molecular Weight of Water (H2O): 18.015 g/mol [2]
• Molecular Weight of Ethanol (C2H5OH): 46.07 g/mol [3]

---

Filtration Technique: Reverse osmosis

• Reverse Osmosis: Mechanism
• Reverse Osmosis: Process

<

>

Reverse osmosis (RO)

​​Mechanism
Liquid is pushed by a pump through a .001 μm membrane. In bever-age filtration, only water and ethanol can pass through, while everything else, including dissolved minerals are rejected. This can be done using a cross-flow filtration system or cartridge filtration system. In technical terms: “Osmosis is the spontaneous net move-ment or diffusion of solvent molecules through a selectively perme-able membrane from a region of high water potential to a region of low water potential, in the direction that tends to equalize the solute concentrations on the two sides” [4]. In other words, because the pore size is so small, the natural flow of water would go from the filtered side to the unfiltered side. By applying pressure, the wine, beer, or tap water is pushed through the membrane, thereby reversing the osmosis.

Filter Design [5]
Saha et al (2013) in a paper presented at the International Sympo-sium Oenoviti International Network, notes the following:

Filter Body

• Cross-flow filtration is a popular option, as it prevents membrane fouling due to colloids
• Dead-end filtration using a specialized plate and frame filter.
• Cartridge filters with hollow fiber or spiral wound filters.
• ​Heat exchangers for temperature control; the high pressure used leads to increases in temperature of the membrane surface and results in filtration.

Membrane [6]
Pore size: 0.1-1 nmMembrane

Material:

• Cellulose acetate blend: Not as selective or durable compared to synthetic polymersIt has a low flux rate.
• Polyethersulfone (PES) has extremely low protein binding properties, and therefore better color retention, as well as good physical robustness [7].
• Ceramics: Very strong and durable. An expensive material, de-signed originally for separation of uranium isotopes.
• Other filter materials: Polyamide or polyimide on polyester [8]. Membrane Construction
  • Thin-film composite membranes composed of:
  • Membrane layer of hydrophobic polyamide or polyethersulfone.
  • One or more supporting layers of polymeric support mate-rial.
  • These membranes have good flux characteristics, are very durable under the high pressure, and cleanable by back flushing.
  • Asymmetric, flat sheet membranes composed of cellulose acetate or cellulose triacetate.

These membranes are highly permeable by water and ethanol, and provide good rejection of high molecular weight compounds like proteins, polyphenol, and carbohydrates.

• Spiral wound membranes, which are essentially large, flat membranes that are rolled around a hollow retentate collection tube with alternating membrane layers separated by a feed spacer.
• Hollow fiber membranes of aromatic polyamides or cellulose triacetate.

Reverse Osmosis Process [9] [10]
1. The beverage is pre-filtered for particulates.

2. The beverage is then fed from the hold-ing tank into the filtration system.

3. Concentration phase: The beverage is sepa-rated/filtered by membranes into permeate, which goes into a separate tank, and retentate, which goes back to the original storage tank. This is repeated until the final desired alcohol concentration is achieved. As the permeate is low in alcohol (for example ~0.7-1.5 % v/v ethanol in wine), multiple passes through the filtration system are necessary for dealcoholization [5].
• The permeate: Water and alcohol.
• The retentate: Concentrated proteins, color, and flavor compounds
.
4. Diafiltration phase: Brewing water which is completely demineralized and deaerated, so that there is an oxygen content of less than 0.1 ppm, is added to the original storage tank. This both replaces water lost due to its removal with the alcohol, and helps to “wash out” the ethanol by adding more water to help carry it in its azeotropic mixture. Alternatively, discontinuous water addition at the beginning or end of the process can also be used [8].
​
5. Make-up phase/Blending: Additional deaerated water is added to obtain the final flavor profile. This is done because it is more precise than trying to achieve accuracy during the diafiltration phase.

Dissolved oxygen management
During this process, dissolved oxygen can be mitigated by sparging with nitrogen gas. Steve Peck, Vice President of Winemaking at J. Lohr, noted via personal communication that sparg-ing can be used in the tank return line and at the end of the reverse osmosis process. Their ARIEL dealcoholized wines, in particular, use a process which reduces dissolved oxygen to below 1 ppm. Peck also notes “We really don’t see much oxidative degradation of the wines in the ARIEL process. Probably because the amount of time spent at high DO is relatively short and the temperature is always under 50ºF.”

​Final processes for beer: As the reverse osmosis process strips the beer of CO2, it must be re-carbonated. Additionally, color and flavor adjustments like dry hopping or the addition of hop oils can be performed at this step.

Reverse osmosis is not economically feasible for the production of beer with an alcohol percentage less than 0.45

Operating parameters that influence the efficacy of ethanol removal
• Feed pressure, where higher pressure results in higher solvent permeation and thus higher ethanol flux, but also increases the permeation of aroma compounds
.• Operating temperature range: 10-20ºC (50-68ºF).
• Temperature increases will increase ethanol and aroma compound permeation. However, higher temperatures can cause losses in volatile aroma compounds
.• The lower the temperature during alcohol removal, the more the Beta-glucan gel (a humectant that attracts water) is formed.

For a great video on reverse osmosis, as well as the corresponding documentation from GEA and the Brewer’s Association: Video (membrane filtration starts at 21:35) GEA. (2021, August 24). Diversifying Your Craft Portfolio with Seltzer and De-alcoholized Beverages. YouTube. Retrieved December 14, 2022, from www.youtube.com/watch?v=rBEY-5rrhgI

Corresponding slides: GEA. (2020). Craft Beer and Beyond. Brewers Association. Retrieved December 14, 2022, from www.brewersassociation.org/wp-content/up-loads/2020/04/CBC-Online-Sponsored-Seminar-Presentation-Craft-Beer-Beyond-Presented-by-GEA-North-America.pdf

​General Benefits

• Processing occurs at low temperatures of 1-5°C.
•  Requires less energy than thermo-distillation processes. For an assessment on the energy consumption and environmental impact of SCC, EP, and RO:
  
  Margallo, M.; Aldaco, R.; Barceló, A.; Diban, N.; Ortiz, I.; Irabien, A. Life Cycle Assess-ment of Technologies for Partial Dealco-holisation of Wines. Sustain. Prod. Consum. 2015, 2, 29–39. from: https://grupos.unican.es/tab/Publicaciones/2015/4.%20Margallo.pdf

Drawbacks

• Water addition is required to achieve effi-cient dealcoholization. Problematically, the addition of water to wine can be illegal or restricted in many wine-producing countries.
• Membrane fouling can pose technical challenges, with more colloidal liquids increasing the rate of membrane fouling. Alcantara et al. (2016) found that membrane fouling restricts flow rate and increases with time and permeate volume [11].

Filtration Technique: Evaporative Perstraction 

• Evaporative Perstraction: Mechanism
• Evaporative Perstraction: Process

<

>

Evaporative Perstraction/ Osmotic Distillation/ Isothermal membrane distillation
While evaporative perstraction (EP) membrane can be used as a stand-alone process for alcohol removal, its usage of heat is detrimental to volatile aromatic compounds. It is more commonly used to remove the ethanol from the permeate fraction following reverse osmosis treatment. Once the ethanol is removed, the remaining liquid, which contains trace amounts of aromatic volatiles, can be recaptured and returned to the original beverage. Additionally, the usage of EP, as noted by David Wollan in the Memstar patent for reduction in alcoholic beverages, notes this is particularly important in wine as [14]:

• The retentate from reverse osmosis requires the addition of diafiltration water, however, the addition of water to wine is not allowed in some jurisdictions. This can be circumvented by stripping out the ethanol from the re-verse osmosis permeate and using that liquid to rehydrate the retentate.
• As distillation requires expensive licensing, and some jurisdictions require the de-alcoholized permeate to only be recombined with the wine from which it was derived, evaporative perstraction is an alternative to shipping the permeate to a distiller and having the dealcoholized permeate shipped back.

Mechanism

• A temperature differential creates a vapor pressure differential between the sides of the hydrophobic membrane.
• The ethanol removal rate during EP depends on processing conditions, primarily membrane surface area, feed flow rate, the stripping solution flow rate, and temperature. Aroma loss can occur with prolonged treatment times and/or at elevated temperatures [13].
• The transport of volatile compounds through the membrane, besides the membrane selectivity, is highly correlated with the Henry’s law constant (KH) [17]. Also called the air–water partition coefficient, this is the ratio of a compound's partial pressure in air to the concentration of the com-pound in water at a given temperature.
• For more on the Henry’s law constant: Avishay DM, Tenny KM. Henry's Law. [Updated 2022 Oct 10]. In: StatPearls [Internet]. Treasure Is-land (FL): StatPearls Publishing; 2022 Jan-. Available from: www.ncbi.nlm.nih.gov/books/NBK544301/

Filter design
Filter body: Dead-end filter

Membrane

• Pore Size: 30-50 nm
• Material: Polypropylene or polyvinylidene fluoride hollow fiber membrane.
• Polarity: Hydrophobic–this property helps to retain the liquid by creating a vapor gap that permits ethanol vapor to permeate across the membrane into the water ‘stripping’ solution. 
  
• The membrane used inhibits the passage of liquid water, while allowing permeability for vapor. The temperature used in dealcoholization will predominantly vaporize ethanol while remaining below the vapor point of water.

Process [13]
The permeate from reverse osmosis is the feedstock for this process.

• The EP feedstock of water and ethanol is degassed, heated to between 45-55°C (113-131°F), then passed across the perstractive membrane.
• The stripping liquid of filtered, degassed water is passed across the other side of the membrane.

This:

1. Vaporizes from the permeate.
2. Moves across the perstractive membrane by convection or diffusion because of the differences in vapor pressure.
3. Condenses the ethanol in the strip water.

• The EP-treated permeate is cooled, recombined with the retentate and returned to the feed tank.
• The beverage is circulated through the RO-EP unit until the desired alcohol level is achieved.
• ​The advantage of this method is that most processing can be conducted at ambient pressures, thereby mitigating cost.

​Factors influencing the rate of ethanol re-moval and the final aroma composition:

• Feed velocity influences time and temperature. The challenge to finding the ideal feed velocity is that aroma losses occur with extended time (and lower feed velocity) and with higher temperature (caused by higher feed velocity).
• Stripping solution velocity.

Dealcoholization by Filtration's Influence on Wine

• Influence on Polyphenols and Acidity
• Influence on Aroma Compounds

<

>

Reverse Osmosis and RO/EP Influence on Wine
Reverse osmosis, often coupled with evaporative perstraction, influences the final wine in a multitude of ways. Generally, however, it results in losses of aroma com-pounds. It was found that in wine, espe-cially when compared to beer, it is difficult to adjust post-dealcoholization because nothing can be added, unlike beer, which can be adjusted using hops. The following summarized studies have come to similar conclusions:

Consequences of Partial Dealcoholization of Red Wine by Reverse Osmosis-Evaporative Perstraction by Pham D-T, Stockdale VJ, Wollan D, Jeffery DW, Wilkinson KL. (2019) studied the influence of 0.7-2.6% dealcoholization by reverse osmosis with evaporative perstraction on Shiraz, Cabernet Sauvignon, and a Shiraz-Cabernet Sauvignon blend.

Influence of partial dealcoholization by reverse osmosis on red wine composition and sensory characteristics by Gil, Mariona & Estévez, S. & Kontoudakis, Nikolaos & Fort Marsal, Maria Francesca & Canals, Joan-Miquel & Zamora, Fernando (2013) studied the effects of 1-2% dealcoholiza-tion of oak-aged Cabernet Sauvignon, Grenache, and Carignan by reverse osmosis.

Partial dealcoholization of red wines, sensory and composition quality by Lisanti, M. T., Gambuti, A., & Moio, L. (2013) which studied the red wine Aglianico, which was dealcoholized by almost 2%, 3%, and 5% v/v of ethanol.

Impact of dealcoholization on quality properties in white wine at various alcohol content levels by Liguori, L., Albanese, D., Crescitelli, A., Di Matteo, M., & Russo, P. (2019) studied the dealcoholization of the 12.5 vol% white grape Falanghina by evaporative perstraction/osmotic distilla-tion (Liquicel membrane) to a final ABV of 9.8% ABV to .3% ABV .

Reverse osmosis’ general influence on wine's polyphenols
Polyphenols, and therefore anthocyanins, generally concentrate during reverse osmosis, since their effects are primarily influenced by ethanol removal [13]. And while polyphenols can collect on the membrane, the losses are minimal, relative to the concentration effect.

Reverse osmosis’ general influence on wine's acidity

• Total Acidity (TA): Slight increase in total acidity was observed above 2% removal [13] [15]. Gil et al. (2013) noted that this acidity is influenced by polysaccharide concentration, as wine polysaccharides can inhibit the growth of potassium hydrogen tar-trate crystals, and polysaccharide concentration is influenced by etha-nol concentration. Additionally, po-tassium bitartrate solubility decreas-es as ethanol content increases [15]. Because of this, Gil et al (2013) suggest that reverse osmosis dealcoholization should be applied as soon as possible in order to prevent the usual losses of titratable acidity which happens in wine during cold stabilization.
• pH and volatile acidity changes were not significant [13] [15] [17].

Reverse osmosis’ general influence on wine's volatile compounds
As all fermentation volatiles are generally capable of permeating the RO membrane, volatile concentrations in dealcoholized wines are generally lower than the levels observed in wine prior to RO-EP treat-ment [13]. The amount of reduction was dependent on the quantity of alcohol re-moved and the compound class [16] [17].

General Trends by Wine Color

• In red wine dealcoholized by RO, Lisanti et al. (2013) found that a 2% reduction minimally affected the sensory proper-ties of wine, whereas a 5% reduction of alcohol caused great modifications of the sensory profiles and losses of up to 100% of some volatile aroma com-pounds. Additionally, the reduction in concentration occurred at different rates [16].
• In white wine dealcoholized by EP, Liguori et al. (2019) found that almost 50% of higher alcohols with acids and lactones were preserved in dealcohol-ized wine at 9.8 vol% alcohol content. This percentage reduced to 30% in the sample at 6.8 vol% and was even lower in the dealcoholized wine with lower alcohol content.

Carboxylic Acids
Carboxylic acids generally decrease in con-centration, though not at equal proportions [13] [17].

• Lower molecular weight acids acetic, propionic, and butanoic acids, for example are more diminished than higher molecular weight acids like hexanoic, octanoic, and decanoic acids, as a result of dealcoholization [13].
• The experimental results found most carboxylic acid concentrations decreased due to their loss through EP, even though acid hydrolysis of ethyl esters would hypothetically increase as ethyl esters = carboxylic acid + ethanol [13]
• The quantity of fatty acid reduction increases with additional dealcoholization [17].
• EP, when used without RO to reduce alcohol by 5.5 vol%, resulted in a reduction of acids by 73.6% [17].

Higher Alcohols
Generally, higher alcohols decreased [16]. The total reduction in higher alcohols increased with additional reduction. Specific higher alcohols with more specific trends:

• β-phenylethyl alcohol (rose aroma) did not decrease, probably due to π-π stacking interactions [16].
• In white Falanghina wine dealcoholized by EP, Liguori et al. (2019) found that 2-phenylethanol remained at higher concentrations until later in de-alcoholization because of its low volatility, high solubility in water and π–π interactions (non-covalent bonds that act similarly to an electrostatic charge) with polyphenols [16].
• EP, when used without RO to reduce alcohol by 5.5 vol%, resulted in a reduction of higher alcohols by 72.8% [17].

​Esters
Ester loss generally increased as more ethanol was removed. This results in a loss of fruit notes. For example, Lisanti et al. (2013) found the berry fruit notes of red wines are related to the additive effect of several esters, rather than to a specific compound, the loss of esters after the dealcoholization process was quite definitely responsible for the decrease of “cherry” and “red fruit” olfactory notes, as found by sensory analysis.

• Ethyl esters = ethanol + carboxylic acid
  Therefore, ethanol removal causes hydrolysis. The research has additionally found higher molecular weight esters like ethyl octanoate and ethyl decanoate, decrease in concentration more rapidly as molecular weight increases (ethyl decanoate > ethyl octanoate > ethyl hexanoate [13]. Pham et al. (2019) suggested, based on Makhotkina and Kilmartin (2011), that the ethyl esters of branched acids are more readily hy-drolyzed than lower molecular weight esters. This result is contrary to the initial belief that smaller volatile esters would decrease faster because they can more easily transition across the RO membrane
• EP, when used without RO to reduce alcohol by 5.5 vol%, resulted in a reduction of higher alcohols by 86.3% [17].

Terpene concentration
In the study by Lisanti et al. (2013), alcohol reduction by up to 5% in red wine did not provide a definitive pattern, as some terpenes were found in one sample but not the other. No hypothesis was provided by these authors.

Ketones and Lactones
Losses occur in both ketones and lactones [17].

• White wine dealcoholized by more than 8.7% ABV using EP was found by Liguori et al. (2019) to only contain Pentadecane-2,4-dione with 92% loss
• Red wine was noted by Liguori et al. (2019) to also have losses, but to a lesser degree, due to π–π interactions with polyphenols and aromatic rings.
• EP, when used without RO to reduce alcohol by 5.5 vol%, resulted in a reduction of lactone by 53.2% [17].
• EP, when used without RO to reduce alcohol by 5.5 vol%, resulted in a reduction of ketones by 78.44% [17].Wine Attributes
• Viscosity decreases due to both the removal of ethanol and other nonspecified compounds [13].
• Free sulfur dioxide decreased significantly [13].

Dealcoholization by Filtration's Influence on Beer

• Influence on Filtration and Acidity
• Influence on Aroma Compounds

<

>

​Reverse Osmosis of Beer
In an insightful study by Ramsey et al. (2021) who found “a clear differentiation between a standard alcohol beer and its lower alco-hol counterpart” when studying the reverse osmosis (RO) dealcoholization using a polyamide membrane cross-flow filter operated at 20°C to a final concentration of 0.08 to 0.35% ABV of:

• 5.5.1% ABV pilsner made with Carapils malt, Mittelfrüh leaf hops and Saaz leaf hops, SafLager W-43/70 yeast.
• 4.3% ABV oatmeal stout with pale ale, chocolate, wheat, light crystal and Carafa Special III malts, flaked oats, roasted barley, Fuggles hops, and SafAle US-05 yeast.

Differences in the filtration process caused by the type of beer being filtered.

• Filtration times: Stouts were found to have longer processing times compared to lagers, which suggested membrane clogging. The stout contained five different malts, as well as flaked oats and roasted barley adjuncts, which were previously found to contain high molecular weight β-glucans which clog membranes.
• Final volatile aroma compound concentration after dealcoholization of the same beer reduced to the same ABV was highly variable.

​This suggests membrane fouling is inconsistent and influences aroma compound reduction.

• Operational suggestions to improve the final product and increase its comparability with the standard beer include:
  • Deep cleaning of the membrane between trials is required.
  • Separate membranes for different product matrices, to avoid product contamination associated with membrane fouling resulting in taints.
  • Altering the brewing process to account for volatile aroma losses.
  • Using special yeasts, which can produce higher levels of higher alcohols and esters during fermentation.
  • Changing the composition of brewing raw materials or selecting an RO membrane with a different composition.

Reverse osmosis’ general influence on beer’s acidity

• pH generally decreased, except for a small increase in stout trials. While the authors do not have a hypothesis as to why this increase occurred, it may be due to changes in acid solubility caused by the lack of ethanol, as was found by Gil et al. (2013).Reverse osmosis’ general influence on beer’s polyphenols
• Total polyphenols were reduced.
• Bitterness units (IBU) were reduced.

Reverse osmosis’ general influence on beer’s volatile aroma compounds
The volatile compounds which were permeable through the membrane are influenced, not by polarity, but by: Surface area + volume + shape.

• Straight chain/linear compounds were removed at a higher level than branched chain compounds or those with the presence of a benzene ring. For example, 3-methylbutyl acetate, 2-methylpropyl acetate, 3-methyl-1-butanol and 2-methylpropan-1-ol ( found to be the compounds with the highest retention, which was believed to be due to the additional methyl group within these molecules.
• This was deduced by Ramsey et al. (2021) from the finding that esters and higher alcohol removal appeared to increase with increasing size (e.g. ethyl acetate up to ethyl decanoate). A find-ing which contradicts the hypothesis and results from previously reported studies that showed compounds with a similar structure and molecular weight to ethanol would be removed during membrane dealcoholization, while anything more complex would be retained.
• While not experimentally confirmed, using RO techniques for a beer with increased levels of branched flavor com-pounds, like stouts, may result in less sensory and physicochemical changes in comparison to lagers and other beers with linear flavor compounds.

Higher alcohols
For both the lager and the stout, increased removal of the higher alcohols propan-1-ol, 2-methylpropan-1-ol and 3-methyl-1-buta-nol was found. Esters Ethyl acetate, ethyl butanoate, and 3-meth-ylbutyl acetate showed increased removal for the stout. Carboxylic acids and aldehydes. Both were significantly reduced Aroma attributes found by using a trained 12 member panel

• A decreased presence of the sensory attributes ‘fruity/estery’, ‘alcoholic/solvent’, and ‘sweetness’. Ethanol has been found in previous research to enhance the perception of fruity flavor, alcoholic/solvent, sweetness, and full-ness/body, with previous research also showing that RO removes volatiles like esters that contribute to these attributes.
• Significantly higher aroma attributes of ‘Malty aroma’ and ‘malty flavor’ were significantly higher in the standard beers
•  ‘Sour’ attributes, especially with lagers. Previous research suggests that the significant increase in perceived acidity is due to the removal of key esters and higher alcohols.
• ‘Sulfur flavor’, particularly in the lager. This may be attributed to the lack of other flavors which normally work synergistically to cover up such ‘off-flavors.’ No volatile compounds were identified by GCMS to correlate to the attribute of ‘sulfur flavor’. Additionally, the increased attribute may be caused by highly odor active sulfur compounds like dimethyl sulfide, dimethyl disulfide, and sulfur dioxide, which are present but not clearly defined by GC–MS analysis. This was less prevalent in the stout than the lager and may be attributed to increased amounts of other flavor compounds like pyrazines and furans.
• Sensory perceptions of ‘body’ were found to be significantly lower in the dealcoholized samples, suggesting that mouthfeel enhancers, such as sugars, were removed by the membrane due to their molecular size.

​For a thorough technical analysis, Dr. Ramsey’s Ph.D. dissertation: https://eprints.nottingham.ac.uk/64533/3/Final%20Thesis_After%20Corrections%20120221.pdf

​Sources and Suggested Reading

1. Hancock, N. (2017, January 23). Ultrafiltration, Nanofiltration and Reverse Osmosis — Safe Drinking Water Foundation. Safe Drinking Water Foundation. Retrieved December 19, 2022, from https://www.safewater.org/fact-sheets-1/2017/1/23/ultrafiltrationnanoandro

2. National Center for Biotechnology Information (2022). PubChem Compound Summary for CID 962, Water. Retrieved December 20, 2022 from https://pubchem.ncbi.nlm.nih.gov/compound/Water.

​3. National Center for Biotechnology Information (2022). PubChem Compound Summary for CID 702, Ethanol. Retrieved December 20, 2022 from https://pubchem.ncbi.nlm.nih.gov/compound/Ethanol.

4. Zoecklein, B., Fugelsang, K. C., Gump, B. H., & Nury, F. S. (2013). Wine analysis and production. Springer Science & Business Media. Retrieved from: https://www.apps.fst.vt.edu/extension/enology/downloads/wm_issues/Winery%20Filtration.pdf

5. Saha, B., Torley, P., Blackman, J., & Schmidtke, L. (2013). Review of technological strategies to reduce alcohol levels in wines. In International Symposium Oenoviti International Network (pp. 78-86). Vigne et Vin Publications Internationales. www.fondation.univ-bordeaux.fr/wpcontent/uploads/2016/01/2013-09-06-oenoviti-revue.pdf

6. Schmidtke, L. M., Blackman, J. W., & Agboola, S. O. (2012). Production technologies for reduced alcoholic wines. Journal of Food Science, 77(1), R25-R41. https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1750-3841.2011.02448.x)

7. Corning. Corning Filtration Guide. (n.d.). Retrieved May 17, 2022, from https://www.corning.com/catalog/cls/documents/selection-guides/t_filterselectionguide.pdf

8. Ramsey, I., Yang, Q., Fisk, I., Ayed, C., & Ford, R. (2021). Assessing the sensory and physicochemical impact of reverse osmosis membrane technology to dealcoholize two different beer styles. Food Chemistry: X, 10, 100121. https://www.sciencedirect.com/science/article/pii/S2590157521000092

9. Lowal de-alcoholizer. Alfa Laval. (n.d.). Retrieved December 12, 2022, from www.alfalaval.us/products/separation/membranes/membrane-filtration-systems/lowal-de-alcoholizer/

10. Montanari, L.; Marconi, O.; Mayer, H.; Fantozzi, P. 6—Production of Alcohol-Free Beer. In Beer in Health and Disease Prevention; Preedy, V.R., Ed.; Academic Press: San Diego, CA, USA, 2009; pp. 61–75. https://www.researchgate.net/profile/OmbrettaMarconi/publication/286327980_Production_of_alcohol-free_beer

11. Alcantara, B. M., Marques, D. R., Chinellato, M. M., Marchi, L. B., da Costa, S. C., & Monteiro, A. R. G. (2016). Assessment of quality and production process of a non-alcoholic stout beer using reverse osmosis. Journal of the Institute of Brewing, 122(4), 714-718. https://onlinelibrary.wiley.com/doi/10.1002/jib.368

12. Russo, P.; Liguori, L.; Corona, O.; Albanese, D.; DiMatteo, M.; Cinquanta, L. Combined Membrane Process for Dealcoholization of Wines: Osmotic Distillation and Reverse Osmosis. Chem. Eng. Trans. 2019, 75, 7–12. Retrieved from: https://core.ac.uk/download/pdf/224993613.pdf

13. Pham D-T, Stockdale VJ, Wollan D, Jeffery DW, Wilkinson KL. Compositional Consequences of Partial Dealcoholization of Red Wine by Reverse Osmosis-Evaporative Perstraction. Molecules. 2019; 24(7):1404. https://doi.org/10.3390/molecules24071404

14. Wollan, D. Alcohol Reduction in Beverages. U.S. Patent 20080272041, 6 November 2008. Retrieved from: https://patents.google.com/patent/US20160097024/en

15. Gil, Mariona & Estévez, S. & Kontoudakis, Nikolaos & Fort Marsal, Maria Francesca & Canals, Joan-Miquel & Zamora, Fernando. (2013). Influence of partial dealcoholization by reverse osmosis on red wine composition and sensory characteristics. European Food Research and Technology. 237. 10.1007/s00217-013-2018-6. Retrieved from https://www.researchgate.net/publication/257373302_Influence_of_partial_dealcoholization_by_reverse_osmosis_on_red_wine_composition_and_sensory_characteristics

16. Lisanti, M. T., Gambuti, A., & Moio, L. (2013). Partial dealcoholization of red wines, sensory and composition quality. Alcohol level reduction in wine, . In International Symposium Oenoviti International Network (pp. 89-97). Vigne et Vin Publications Internationales. www.fondation.univ-bordeaux.fr/wp-content/uploads/2016/01/2013-09-06-oenoviti-revue.pdf

17. Liguori, L., Albanese, D., Crescitelli, A., Di Matteo, M., & Russo, P. (2019). Impact of dealcoholization on quality properties in white wine at various alcohol content levels. Journal of food science and technology, 56(8), 3707-3720. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6675842/

18. Wikipedia contributors. (2022, May 20). Pi-interaction. In Wikipedia, The Free Encyclopedia. Retrieved 06:33, December 13, 2022, from https://en.wikipedia.org/w/index.php?title=Pi-interaction&oldid=1088893493