1
Introduction to Reverse Osmosis: History, Challenges, and Future Directions

1.1 Introduction

Reverse osmosis (RO) is a demineralization technique (also known as a "desalination" technique) used to separate solutes in solution from solvents. As a demineralization technique, the solutes are defined as dissolved ions and organics, while the solvent is usually water. RO relies on a semipermeable membrane that is responsible for the separation. The membrane allows water to pass through it while retaining most of the dissolved solids. The driving force for RO is an applied pressure that forces water through the membrane in the direction opposite of that via the natural process of osmosis (detailed in Chapter 2).

Figure 1.1 shows how the separation performance of RO compares to other membrane- and conventionally-based separation/filtration technologies. RO is the finest "filtration" technique currently available, capable of removing monovalent ions from solution to yield demineralized water (as discussed in Chapter 3, RO does not rely on size-exclusion filtration to separate solutes from solution but uses the most-cited Solution-Diffusion Model of separation to describe how solutes pass through a membrane.

Reverse osmosis is the leading worldwide technology for demineralization for both industrial and municipal applications today. Figure 1.2 shows that membrane techniques (including RO, electrodialysis, electrodialysis reversal, continuous electrodeionization, membrane distillation, etc.) have been on the rise since 2000 while thermal processes have been on the decline during the same time period (note that world-wide capacity for both types of techniques were about equal just prior to the year 2000) [1]. RO offers several advantages over other demineralization processes. Total energy requirements for RO are lower than that for thermal processes [2]. Further, RO systems have a smaller footprint and are modularized for each of installation, use, and expansion [2].

Figure 1.1 Filtration spectrum comparing various membrane-based technologies (italics) and conventional multimedia filtration (bold) for separation capabilities based on approximate size of removal and nature of the dissolved solute or suspended solid to be removed.

Figure 1.2 New contracted capacity of membrane- and thermal-based desalination techniques from 2002 through June 2022. Courtesy of IDA Desalination and Reuse Handbook, 2022-2023, Page 7 [1].

RO has also replaced ion exchange in many plants for brackish water demineralization to avoid handling acid, caustic and regeneration waste. Today, RO is commonly used to compliment ion exchange by removing the bulk of solutes from water prior to treatment with ion exchange, thereby greatly reducing the chemicals and neutralization needed for the ion exchange polishing.

Since commercialization in the mid-1960s, RO has seen developmental strides in selectivity and water permeability, thereby producing better quality water at lower applied pressure. For example, the 1965 pilot test at Coalinga, CA, USA (conducted prior to start-up of the first commercial RO facility at this location) demonstrated solute passage (measured as total dissolved solids (TDS)) through the cellulose acetate membranes of about 9% [3]. Today, TDS passage through commercial, brackish water, polyamide composite membranes is as low as 0.2% [4], and, for commercial, thin film nanocomposite seawater membranes, 0.11% [5]. RO operating pressures have decreased from about 41 bar in the 1970s to less than 16 bar today. Specific energy consumption (SEC) has also decreased from the range of 7.0-9.0 kWh/m3 to about 2.5-3.5 kWh/m3 in 2016 [6], owing to more water-permeable membranes and efficiencies in pumping equipment.

1.2 A Brief History of Reverse Osmosis

1.2.1 Early Development

The first recorded description of osmotic properties of semipermeable membranes occurred in 1748, when Jean-Antoine Nollet observed the phenomenon of osmosis [7]. Wilhelm Pfeffer, in his book, Osmotic Investigations: Studies on Cell Mechanics, published in 1877, describes the osmotic properties observed in plant cell membranes [8]. At about the same time, Moritz Traube developed artificial membranes made of cupric ferrocyanide (carbon, copper, nitrogen, and iron), and demonstrated that these membranes interacted differently with water than with dissolved solutes [8]. In 1948, Dr Gerald Hassler, at the University of California, Los Angeles (UCLA), proposed an "air film" barrier between two cellophane membranes where he surmised that osmosis involves evaporation of water at one membrane, followed by transport through the air film as a vapor, and then the vapor condensed at the other membrane [7].

The timeline to today's membranes begins in 1955 as shown in Figure 1.3. Professor Charles Reid at the University of Florida together with Ernest Breton, demonstrated a pressure-driven process of reversing osmotic flow through cellulose acetate membranes [10]. They had investigated several materials in a trial-and-error process, including cellophane, rubber hydrochloride, and polystyrene, in addition to cellulose acetate. They focused on available polymers at the time, with hydrophilic groups to facilitate water transport. Some polymers investigated exhibited no product flow or passed 75% of the feed water chloride concentration at up to 55 bar applied pressure, which clearly were not effective [10]. The cellulose acetate membrane prepared by Reid and Breton from DuPont (88 CA-43 (E.I. du Pont de Nemours, Wilmington, DE USA)) exhibited chloride passage of less than 4% at applied pressures of only 27.5 bar [10]. Water throughput ranged from 0.08 m3/m2-d (m/d) for a 22-µm thick membrane up to 0.56 m/d for a 3.7-µm thick membrane tested at 41 bar on a 0.1 M sodium chloride solution [10]. Reid and Breton concluded that their cellulose acetate membranes exhibited requisite semipermeable properties for practical applications, but improvements in durability and throughput were required [10].

Figure 1.3 Milestones in the history of RO development.

Figure 1.4 Sidney Loeb's "big dripper", cellulose acetate flat sheet membrane equipment. Courtesy of Julius "Bud" Glater. (a) disassembled module and (b) completely assembled module.

The breakthrough resulting in commercially-viable membranes for "reverse osmosis" (a term first used in a 1956 UCLA Engineering Report by Hassler [9, 11]) was achieved by Sidney Loeb and Srinivasa Sourirajan over the years of 1958-1960 while working in Professor Samuel Yuster's UCLA lab [9]. After months of work, Loeb and Sourirajan developed a suitable cellulose acetate membrane with higher throughput and lower solute passage than the Reid and Breton membranes [12]. The membranes were initially hand-cast and characterized as a homogeneous material with a physically-asymmetric structure [13]. Figure 1.4 shows Loeb and the flat-sheet membrane equipment dubbed the "big dripper". Later, tubular configurations of the membrane were achieved. Figure 1.5 shows a schematic of the tubular casting system [14], while Figure 1.6 shows permeating water from the tubular membrane. Figure 1.7 shows the capped, in-floor immersion preserved in Boelter Hall at UCLA.

Figure 1.5 Schematic of the tubular cellulose acetate membrane casting device used at UCLA. Courtesy of Julius "Bud" Glater.

Figure 1.6 Productivity of tubular cellulose acetate RO membrane. Courtesy of Julius "Bud" Glater.

Figure 1.7 Capped immersion tube used in the tubular membrane casting shown in Figure 1.4 at UCLA, 2008.

In 1961, the first company to apply RO was Havens Industrials in Southern California, USA [9]; details of this first application were not found in other historical references. However, on June 4, 1965, the City of Coalinga, CA USA, began a pilot test of the tubular, Loeb-Sourirajan cellulose acetate membranes, under the direction of Loeb and Professor Joseph McCutchan of UCLA (see Figure 1.8) [15]. The 3-year pilot dubbed "Raintree" generated 5,000 gallons per day of drinking water for the city [14].

Figure 1.8 Sidney Loeb (left), UCLA professor Joseph McCutchan (right) and other team members at the Raintree RO pilot test facility at Coalinga, CA USA, cir. 1965. Courtesy of Julius "Bud" Glater.

1.2.2 Advances 1970s-1980s

Cellulose acetate membranes exhibited good throughput with low solute passage, but they had some serious operational design limitations. Limitations included high operating pressure requirements (24-31 bar), narrow operating pH range (4-6), and a low maximum temperature limit of 35°C. And, although solute passage was less than 5%, even lower solute passage was desired. Hence, for RO to truly grow in application, membranes that could achieve performance beyond the limitations of cellulose acetate membranes were needed.

In 1967 E. I. du Pont de Nemours & Company...