Solar distillation of human urine to recover non-potable water and metal phosphate mineral

Global phosphorus (P) depletion through the mining of phosphate rock threatens the reliable supply of fertilizer nutrients to agriculture, which is essential to food security (Cordell & White 2011; Reijnders 2014; Anlauf 2022). Supply chain issues and rising fertilizer prices caused by the interdependence between countries for fertilizer production and the rising costs of natural gas have resulted in farmers turning to less conventional sources of crop nutrients (Timsina 2018; Barbieri et al. 2022). Urine contains the primary macronutrients required for crop production including nitrogen (N) as urea, P as phosphate, and potassium (Kirchmann & Pettersson 1995; Pandorf et al. 2019). Human urine may be a viable alternative to commercial fertilizer due to its availability as a renewable and readily accessible resource (Nagy & Zseni 2017), local and universal presence, energy efficiency as compared to new production from natural resources (Maurer et al. 2003), and elimination of wastes and environmental pollutants (Good & Beatty 2011). However, switching to urine-derived fertilizer poses new challenges such as the development of collection infrastructure (Berndtsson 2006; Ishii & Boyer 2015), social acceptance (Pahl-Wostl et al. 2003; Segrè Cohen et al. 2020), and treatment processes able to separate nutrients from water, salts, and trace organic contaminants such as pharmaceuticals (Patel et al. 2020; Martin et al. 2022).

Implementation of urine source separation may be most feasible in workplaces such as office buildings where restrooms are the main source of wastewater generation (Boyer & Saetta 2019). Collection and treatment of urine in buildings could produce useful amounts of non-potable water, which can be used for industrial purposes, such as cooling or irrigation, and thereby reduce requirements for potable water (Jagtap & Boyer 2020). One such technology for urine treatment is solar distillation, which provides a simple means of P and N recovery and may be particularly appealing in arid regions with abundant sunlight and limited water resources. Rather than using urine in its excreted form, distillation reduces the volume of the final product, reducing transportation costs, and the precipitation of phosphate as struvite produced through some methods of urine distillation is less susceptible to leaching than other fertilizer applications.

Previous studies investigating distillation or evaporation of human urine to recover nutrients and water are summarized in Supplementary Table S1. Urine distillation studies have focused on nutrient and water characterization by quantifying the amount of N and P in various forms after extraction and determining the quality of distilled water (Patel et al. 2020). Distillation for recovery of P and N in urine has mostly been conducted at the laboratory scale to evaluate the performance of nutrient extraction processes. For example, various forms of energy have been used to achieve urine distillation including pressure (Udert & Wächter 2012), heat (Udert & Wächter 2012; Bethune et al. 2014, 2015), air (Bethune et al. 2014, 2015, 2016), and artificial sunlight (Zhang et al. 2023), as well as natural sunlight (Antonini et al. 2012). Furthermore, studies have considered nutrient extraction by means of membrane separation (Patel et al. 2020), conversion films (Zhang et al. 2023), ion exchange resins (Patel et al. 2020), and electrolytic reactors in combination with other forms of energy (Jayakrishnan et al. 2021). A few studies have evaluated evaporation, relying on combinations of heat, air, and sunlight to derive nutrients most commonly in solid form through complete evaporation of water from the urine (Antonini et al. 2012; Bethune et al. 2014, 2015; Delhiraja et al. 2021). A few studies have considered various environmental conditions including temperature, wind, humidity, and airflow to determine how these factors affect evaporation rate (Bethune et al. 2015, 2016).

Urine distillation studies have evaluated P recovery as phosphate minerals such as struvite and hydroxyapatite (Bethune et al. 2016). Phosphorus was often recovered as a solid product (Udert & Wächter 2012; Bethune et al. 2014), though some studies ended with P present in the distillation bottoms (Delhiraja et al. 2021; Zhang et al. 2023). Phosphorus in the distillation bottoms was either as precipitate or remained in solution. Nitrogen was recovered in many forms including ammonia, ammonium, nitrate, nitrite, and urea (Bethune et al. 2016; Ren et al. 2021). Most studies recovered a combination of N forms. Nitrogen has also been recovered as a solid product (Udert & Wächter 2012; Ren et al. 2021).

Many urine distillation studies have attempted to collect and analyze the quality of the distillate (i.e., water) as a byproduct of nutrient recovery. Important to assessing the quality of the water product is the quantification of various ions excreted in urine. For example, Zhang et al. (2023) considered the presence of sodium, potassium, ammonia, and total P in both the urine concentrate (i.e., distillation bottoms) and product water (i.e., distillate) to better understand the fate of urine constituents. Other studies have considered the presence of organics in the distillate (Udert & Wächter 2012), while others have sought to track the fate of pharmaceuticals that are present in urine such as diclofenac during precipitation (Ronteltap et al. 2007). Notable additions to urine distillation studies include evaluation of solar distillation performance for nutrient collection using both urine and feces (Delhiraja et al. 2021), and comparison of crop growth rates for recovered urine versus commercial fertilizers (Antonini et al. 2012; Ren et al. 2021).

The review of previous research on distillation and evaporation of human urine highlights several gaps in the literature. Foremost, the coupling of phosphate mineral precipitation with distillation has not been thoroughly investigated. This is an important consideration because phosphate minerals, such as struvite and calcium phosphate, have the potential to be used as alternatives to commercial fertilizer (De Boer et al. 2018). Across urine treatment studies, phosphate mineral precipitation in urine with the addition of calcium and/or magnesium has been extensively studied (Yan et al. 2021; Zhang et al. 2022). However, many other metal salts are known to precipitate with phosphate (Perry 2011). This research provided the opportunity to evaluate phosphate precipitation with different metal salts and thereby advance urine treatment literature. Distillation studies have been conducted on both synthetic urine and real urine, yet few distillation studies have directly compared the two, and similarly, differences between hydrolyzed urine and fresh urine are often overlooked. In addition, acidified fresh urine is another form of urine and is recommended to stabilize the N in urine as urea (Saetta et al. 2020), which is important to minimize loss of N during distillation or evaporation. No previous studies have been comprehensive enough to compare synthetic hydrolyzed and synthetic fresh urine with real hydrolyzed and real fresh urine. Furthermore, urine distillation research could benefit from exploring methods of distillation at smaller and more accessible scales that allow for a greater number of variables to be studied. In addition, distillation via solar heating is understudied with no previous solar distillation studies conducted in North America.

The goal of this research was to provide an improved understanding of solar distillation of human urine to recover water and nutrients. The specific objectives of the research were to (1) evaluate the influence of distillation conditions on distillate (water) recovery and composition for distillation of synthetic urine; (2) evaluate the influence of different metal salts on phosphate precipitation in distillation bottoms for distillation of synthetic urine; and (3) evaluate the influence of synthetic versus real urine composition on distillate (water) recovery and phosphate precipitation results during distillation. Solar distillation performance was based on volume and composition of the water recovered, extent of P removed from solution via phosphate precipitation, and fate of N. Both synthetic urine and real urine were used for both fresh urine and hydrolyzed urine types including acidified fresh urine.

The following metal chlorides were used in this research to investigate phosphate mineral precipitation: aluminum chloride hexahydrate, AlCl₃·6H₂O (Fisher Chemical); calcium chloride dihydrate, CaCl₂·2H₂O (Fisher Chemical); cerium(III) chloride heptahydrate, CeCl₃·7H₂O (Sigma-Aldrich); cobalt(II) chloride hexahydrate, CoCl₂·6H₂O (Fisher Chemical); ferric chloride hexahydrate, FeCl₃·6H₂O (Fisher Chemical); ferrous chloride tetrahydrate, FeCl₂·4H₂O (Spectrum Chemical); indium(III) chloride anhydrous, InCl₃ (Alfa Aesar); lanthanum(III) chloride heptahydrate, LaCl₃·7H₂O (Alfa Aesar); magnesium chloride hexahydrate, MgCl₂·6H₂O (Fisher Chemical); manganese(II) chloride tetrahydrate, MnCl₂·4H₂O (Fisher Chemical); lead(II) chloride, PbCl₂ (Acrōs Organics); zinc chloride, ZnCl₂ (Fisher Chemical); nickel chloride, NiCl₂·6H₂O (Acrōs Organics); and copper(II) chloride, CuCl₂·2H₂O (Acrōs Organics). The metals were selected because they are known to precipitate with phosphate (Perry 2011), and the chloride salt of the metal was used because urine has a high concentration of chloride so additional chloride would have a minor impact on urine composition.

Synthetic fresh urine and synthetic hydrolyzed urine were prepared following the recipes in Supplementary Tables S2 and S3 with additional details provided. Chemicals used to prepare synthetic urine included: urea, CH₄N₂O (Fisher Chemical); sodium chloride, NaCl (Fisher Chemical); sodium sulfate anhydrous, Na₂SO₄ (Fisher Chemical); potassium chloride, KCl (Fisher Chemical); magnesium chloride hexahydrate, MgCl₂·6H₂O (Fisher Chemical); sodium phosphate monobasic anhydrous, NaH₂PO₄ (Fisher Chemical); calcium chloride dihydrate, CaCl₂·2H₂O (Fisher Chemical); ammonium hydroxide, NH₄OH (Fisher Chemical); and ammonium bicarbonate, NH₄HCO₃ (Acrōs Organics). Diclofenac sodium salt (Alfa Aesar) was added to synthetic urine in select experiments and was used to create the diclofenac calibration curve for UV photometry. Diclofenac is a well-studied pharmaceutical in urine, especially in phosphate precipitation studies (Ronteltap et al. 2007), so it was included in the experimental design to track its fate during distillation of synthetic urine. (Analyzing pharmaceuticals in real urine was not in the scope of this study.) The pH of synthetic urine was adjusted using sodium hydroxide solution 10 N, NaOH (Fisher Chemical) and acetic acid, CH₃COOH (Columbus Chemical Industries, Inc.). Acetic acid was added to fresh urine to mimic acidified urine used to stabilize urea in urine (Saetta et al. 2020).

The combined reagent used to measure phosphate was prepared with 5 N sulfuric acid, H₂SO₄ (Fisher Chemical); antimony potassium tartrate trihydrate, C₈H₄K₂O₁₂Sb₂·3H₂O (Fisher Chemical); ammonium molybdate tetrahydrate, (NH₄)₆Mo₇O₂₄·4H₂O (Fisher Chemical); and ascorbic acid, C₆H₈O₆ (Fisher Chemical). Potassium phosphate monobasic, KH₂PO₄ (Fisher Chemical) was used to create the calibration curve for measuring phosphate.

Real hydrolyzed urine was collected using a non-water urinal connected to a storage tank (Saetta et al. 2020). Real fresh urine was collected using specimen cups within 48 h of the experiment start time. Real fresh urine was collected in four batches with the first batch being used for control, pH 5, AlCl₃, and CaCl₂ conditions, the second batch being used for CeCl₃, CoCl₂, FeCl₃, FeCl₂, and InCl₃ conditions, the third batch being used for LaCl₃, MgCl₂, MnCl₂, PbCl₂, and ZnCl₂ conditions, and the fourth batch being used for NiCl₂ and CuCl₂ conditions. Real fresh urine at pH 5 was achieved by adding acetic acid to mimic acidified urine used to stabilize urea in urine (Saetta et al. 2020). Real hydrolyzed urine was collected and used as one batch. Collection of real human urine was approved by the Arizona State University Institutional Review Board. Tables 1 and 2 list the concentration of constituents in synthetic and real urine used in this research.

Table 1

Concentration of chemicals in synthetic fresh urine based on recipe

Chemical .	Formula .	Concentration (mmol/L) .
Urea	CH4N2O	500 as N
Sodium chloride	NaCl	44
Sodium sulfate anhydrous	Na2SO4	15
Potassium chloride	KCl	40
Magnesium chloride hexahydrate	MgCl2·6H2O	4
Sodium phosphate monobasic anhydrous	NaH2PO4	20
Calcium chloride dihydrate	CaCl2·2H2O	4
Sodium, total	Na+	94
Potassium, total	K+	40
Magnesium, total	Mg2+	4
Calcium, total	Ca2+	4
Chloride, total	Cl−	100
Sulfate, total		15
Phosphate, total		20

Chemical .	Formula .	Concentration (mmol/L) .
Urea	CH4N2O	500 as N
Sodium chloride	NaCl	44
Sodium sulfate anhydrous	Na2SO4	15
Potassium chloride	KCl	40
Magnesium chloride hexahydrate	MgCl2·6H2O	4
Sodium phosphate monobasic anhydrous	NaH2PO4	20
Calcium chloride dihydrate	CaCl2·2H2O	4
Sodium, total	Na+	94
Potassium, total	K+	40
Magnesium, total	Mg2+	4
Calcium, total	Ca2+	4
Chloride, total	Cl−	100
Sulfate, total		15
Phosphate, total		20

View Large

Table 2

Concentration of chemicals in synthetic hydrolyzed urine based on recipe

Chemical .	Formula .	Concentration (mmol/L) .
Sodium chloride	NaCl	60
Sodium sulfate anhydrous	Na2SO4	15
Potassium chloride	KCl	40
Ammonium hydroxide	NH4OH	250
Sodium phosphate monobasic anhydrous	NaH2PO4	13.6
Ammonium bicarbonate	NH4HCO3	250
Sodium, total	Na+	103.6
Potassium, total	K+	40
Chloride, total	Cl−	100
Inorganic carbon, total		250
Sulfate, total		15
Phosphate, total		13.6
Ammonia, total	NH3 +	500

Chemical .	Formula .	Concentration (mmol/L) .
Sodium chloride	NaCl	60
Sodium sulfate anhydrous	Na2SO4	15
Potassium chloride	KCl	40
Ammonium hydroxide	NH4OH	250
Sodium phosphate monobasic anhydrous	NaH2PO4	13.6
Ammonium bicarbonate	NH4HCO3	250
Sodium, total	Na+	103.6
Potassium, total	K+	40
Chloride, total	Cl−	100
Inorganic carbon, total		250
Sulfate, total		15
Phosphate, total		13.6
Ammonia, total	NH3 +	500

View Large

Batch distillation experiments were conducted as follows. Ten distillation chambers were created using 27 cm diameter clear plastic bowls and placing one 125 mL amber glass bottle inside each bowl. Plastic wrap was stretched smoothly over the top of the mixing bowls. A weighted object was placed in the center of the plastic wrap on each bowl, directly above the collection bottle, directing the evaporated liquid into the center catchment. The bowls were transported outdoors to the upper level of a parking structure for 6 h in direct sunlight. The approximate surface area of urine liquid exposed to the sun was 89 cm².

Along with a control condition at pH 6, a condition of pH 5 with acetic acid addition, and the condition of diclofenac addition at pH 6 in synthetic fresh urine, there were 14 metal conditions tested for synthetic fresh urine and real fresh urine at pH 6: AlCl₃, CaCl₂, CeCl₃, CoCl₂, FeCl₃, FeCl₂, InCl₃, LaCl₃, MgCl₂, MnCl₂, PbCl₂, ZnCl₂, NiCl₂, and CuCl₂. Control conditions were tested for synthetic hydrolyzed urine and real hydrolyzed urine at pH 9 in addition to a diclofenac condition for synthetic hydrolyzed urine, and 12 metal conditions at pH 9: AlCl₃, CaCl₂, CeCl₃, CoCl₂, FeCl₃, FeCl₂, InCl₃, LaCl₃, MgCl₂, MnCl₂, PbCl₂, and ZnCl₂ for each of the hydrolyzed urines. All experimental conditions are summarized in Supplementary Table S4. Metal salts were added based on the phosphate concentration in urine (i.e., 1:1 molar ratio of metal to phosphate), which was calculated for synthetic urine and measured for real urine. Each condition was tested in duplicate. For the distillates and distillation bottoms of each condition, four measurements were collected: temperature, volume, pH, and conductivity. When diclofenac was added to synthetic urine, the distillate and distillation bottoms were analyzed for diclofenac. Phosphate was measured in the distillation bottoms for all distillation conditions. UV absorbance was measured for the distillate samples from real urine.

Temperature data collection began 15 min before the 6-h distillation was complete and lasted until 15 min after the 6-h mark. Temperatures were gathered at the location of distillation, still exposed to direct sunlight for the duration of data collection. Volume data was collected most often on the day after the experiment and was determined by pouring the distilled contents into graduated cylinders. Conductivity and pH readings were most often collected on the day after the experiment and were determined by using the Thermo Scientific Orion Versa Star Pro pH and conductivity probes. Phosphate was measured following Standard Method 4500-P. E. Ascorbic Acid Method (Eaton et al. 2005). The amount of phosphate removed from the urine by precipitation through the addition of metals during distillation was determined for all distillation bottoms by adding 8 mL of combined phosphate reagent to 50 mL of samples at a dilution of 1 mL of filtered (0.45 μm) sample per 500 mL deionized water. Ten minutes after the combined reagent was added to the diluted sample, UV absorbance at 880 nm was used to determine phosphate concentrations using the Thermo Scientific Orion AquaMate 8000 UV–VIS spectrophotometer. The phosphate measurement was most often taken the day after the experiment. The amount of phosphate precipitated was calculated as the difference in concentration between initial concentration in urine and final concentration after filtration. The diclofenac concentration in distillate and distillation bottoms samples was analyzed by UV absorbance at a wavelength of 275 nm using the Shimadzu UV-2600 UV–VIS spectrophotometer (Landry & Boyer 2013). A calibration curve for diclofenac was created using concentrations of 0.2, 0.1, 0.05, 0.02, 0.01, and 0.005 mmol/L. Diclofenac concentrations were measured the day or days after the experiment occurred. Real fresh urine and real hydrolyzed urine distillates received additional testing for the presence of organic compounds post-distillation based on UV absorbance at a wavelength of 254 nm using the Thermo Scientific Orion AquaMate 8000 UV–VIS Spectrophotometer. Measurement of organics was taken the day or days after the experiment occurred. All urine conditions were performed in duplicate with averages and standard deviations calculated.

The results in Figures 1 and 2 indicate that the strongest influence on the volume of distillate collected was the type of synthetic urine, with hydrolyzed urine achieving greater volume recovery than fresh urine. The distillate volume collected averaged 9.5 mL for synthetic fresh urine and 16.0 mL for synthetic hydrolyzed urine. Synthetic hydrolyzed urine contained the volatile species of ammonia and carbon dioxide, which may have facilitated water evaporation (Seaward et al. 1984; Pahore et al. 2012). The impact of metal salt addition on phosphate precipitation is discussed in the next section. Synthetic hydrolyzed urine had greater conductivity values in both the distillate and bottoms than synthetic fresh urine. The conductivity of bottoms samples was nearly twice as high for synthetic hydrolyzed urine than for synthetic fresh urine. The conductivity of distillate samples of synthetic hydrolyzed urine was 70× greater than synthetic fresh urine. The disparity in distillate conductivity between the two urine types suggests that there was greater volatilization of chemicals from hydrolyzed urine that became ions in the distillate from hydrolyzed urine (Pahore et al. 2012). This was most likely from ammonia and carbon dioxide, thereby resulting in less pure distillate (water) product. Synthetic hydrolyzed urine had higher pH values and less variability in pH values than synthetic fresh urine due to its buffering capacity from ammonia and bicarbonate. Synthetic fresh urine had high pH variability in the bottoms between the different metal conditions and between the bottoms and distillates due to the absence of buffering chemicals. Synthetic fresh urine had a difference of 2.3 pH units between the pH values of distillates and bottoms, whereas synthetic hydrolyzed urine had a difference of 0.2 pH units between the pH values of distillates and bottoms.

Diclofenac was included as a model pharmaceutical in the distillation experiments using synthetic urine because diclofenac is often detected in urine (Winker et al. 2008) and monitored during nutrient recovery processes (Wei et al. 2018). For example, with magnesium addition to hydrolyzed urine and subsequent struvite precipitation, diclofenac remained in the liquid phase and was not incorporated into the phosphate solid (Ronteltap et al. 2007). Therefore, it was of interest to understand the fate of diclofenac during distillation with the caveat that it is just one pharmaceutical type. Compared to synthetic hydrolyzed urine, synthetic fresh urine had a higher concentration of diclofenac in the distillate product and a lower concentration of diclofenac in the distillation bottoms. Synthetic fresh urine had an average diclofenac concentration of 0.017 mmol/L in the distillate and 0.233 mmol/L in the bottoms, while synthetic hydrolyzed urine had an average diclofenac concentration of 0.004 mmol/L in the distillate and 0.337 mmol/L in the bottoms. The results for diclofenac show a different trend than the results for conductivity where the distillate from hydrolyzed urine contained more ions than the distillate from fresh urine.

Distillate volume recovery did not correlate to increasing daily high air temperature. As daily high air temperature rose from 99 to 115°F, no change in distillate volume recovered was observed among all urine conditions, with an R² value of 0.0019. The trend for average humidity during the distillation period showed a small increase in distillate volume recovered as humidity rose from 7 to 34%, with an R² value of 0.1518. There was an increase in distillate and bottom liquid temperatures as the daily high air temperature increased.

Neither real fresh urine nor real hydrolyzed urine appeared to have a clear advantage in precipitating phosphate from the solution. Real fresh urine achieved an average phosphate removal of 45% and real hydrolyzed urine averaged a phosphate removal of 39%. The phosphate precipitation results in this study may have been hindered due to the lack of mixing, although high phosphate removal by precipitation was archived for synthetic hydrolyzed urine, and especially for Mg addition. Similar to the precipitation results for synthetic urine, real fresh urine had a few outstanding metal phosphates precipitate whereas real hydrolyzed urine did not have any outstanding metal phosphates. The metals that precipitated phosphate well between both synthetic fresh and real fresh urines were Fe(III) and In. Only a few studies have compared phosphate precipitation in real fresh urine and real hydrolyzed urine (Jagtap & Boyer 2018; Patel et al. 2020). In general, phosphate precipitation is more effective in hydrolyzed urine than fresh urine because the higher pH results in a higher concentration of orthophosphate, which increases the ion activity product of the metal phosphate.

The methods of solar distillation in combination with phosphate precipitation by metal salt addition used in this study can create three products from human urine: non-potable water, metal phosphate solid, and N-containing concentrate. Although not all metal phosphate solids are safe or effective as fertilizer, iron phosphate was recovered in greater amounts from fresh urine while struvite (magnesium ammonium phosphate) was recovered in greater amounts from hydrolyzed urine. Fe(III) showed high phosphate removal in synthetic and real fresh urine while Ce showed high phosphate removal in synthetic and real hydrolyzed urine. The metal that performed best in all urine types was Ce. Metal addition may also be useful in the distillation process for removal of organics from hydrolyzed urine as the metals reduced the concentration of UV-absorbing organics in the distillate from real hydrolyzed urine when compared to the control. Distillation of fresh urine results in a higher purity product water than distillation of hydrolyzed urine; however, distillation of hydrolyzed urine resulted in a higher volume recovery of product water than fresh urine. Synthetic and real fresh urine showed similar distillation trends, as did synthetic and real hydrolyzed urine. Similar results for pH and conductivity between real and synthetic urine did not translate to similar results between phosphate precipitation results for most metals used. This suggests that synthetic urine cannot be used to simulate real urine for phosphate precipitation portions of distillation. The volume of distillate (i.e., product water) recovered was unique to each type of urine for each condition.

This work was supported by the Science and Technologies for Phosphorus Sustainability (STEPS) Center, a National Science Foundation Science and Technology Center (NSF STC, CBET-2019435).

P.B.: Investigation, Visualization, Writing – original draft. M.S.: Investigation, Writing – review and editing. V.T.: Investigation, Writing – review and editing. C.G.: Investigation. L.C.: Investigation, Writing – review and editing, Supervision. T.H.B.: Conceptualization, Writing – review and editing, Supervision, Funding acquisition.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.