Procedure for Testing Pharmaceutical Removal by Duckweed Plants via HPLC-MS Analysis

A modified version of the EPA 1694 Method: Pharmaceuticals and Personal Care Products in Water, Soil, Sediment, and Biosolids by HPLC/MS/MS

Pharmaceutical Standards

Primary Standards (raw chemical)

1. Nicotine obtained from: Cambridge Isotope Laboratories, Inc.; PN# N-008; $18 for 1mg/mL in MeOH; CAS#54-11-5


2. Cotinine obtained from: Cambridge Isotope Laboratories, Inc.; PN# C-016; $25 for 1mg/mL in MeOH; CAS#486-56-6


3. Acetaminophen obtained from: Sigma-Aldrich; 4-Acetamidophenol PN# A7302-5G-A; $16 for 5 grams; CAS# 103-90-2


4. Sulfamethoxazole obtained from: Bioworld; PN# 41910016-1 (762346-6); $41.90 for 25 grams; CAS#723-46-6


5. Carbamazepine obtained from: Sigma-Aldrich; PN# C4024-1G; $21 for 1 gram; CAS# 298-46-4


6. Fluoxetine hydrochloride USP obtained from: Spectrum Chemical Mfg. Corp.; PN# F1200; $152.90 for 5 grams; CAS# 59333-67-4


7. Progesterone obtained from: Sigma-Aldrich; PN# 850454-5G; $20 for 5 grams; CAS# 57-83-0


8. Atrazine-D5 (internal standard) obtained from: Fluka/Sigma-Aldrich; PN# 34053; $229 for 10 mg; CAS# 163165-65-1

Secondary Standards (Single chemical solution)

• Makeup individual solutions for each compound in HPLC grade methanol at 500,000 ng/mL, with the exception of the Atrazine-D5 internal standard make up at 50,000 ng/mL in MeOH

Intermediate Standard/Spiking solution (Combined chemical solution)

• Makeup combined solution with all seven compounds in HPLC grade methanol at 2500 ng/mL and 5000ng/mL(acetaminophen only)


• Makeup internal standard solution with Atrazine-D5 at 50 or 250 ng/mL

Working Standards (Quality control standards)

• Makeup HPLC/MS calibration standards in 0.1% Formic Acid at approximately: 10, 25, 50, 100, 250, 500, and 1000 ng/mL

Microcosms Setup

Figure 1: Test jars rinsed and labeled, filled with tap water and spiked with nutrient solution and pharmaceuticals, ready for plants to be added to specific jars

1. Rinse sample jars (wide-mouth pint-size Mason jars) with HCl, MeOH, and then tap water


2. Label sample jars(triplicate jars with treatments: A = dark no plant; B = dark with plant; C = light w/ plant; D = light no plant; E = light w/ macerated plant; X=control w/ lids)


3. Record tare weight of jars


4. Fill jars with approximately 400 mL tap water (fill jars for macerated duckweed with only 200 mL tap water)


5. Spike jars with Farrell.v2 nutrient soln. (1:400 "A" and 1:1000 "B,C,D" stock solutions)


6. Spike jars with 500ng-pharmaceuticals, except for 1000ng-acetaminophen


7. Wait 1 hr. to chemicals to equilibrate in solution


8. Weigh jars with solution


9. Cover jars with paper shields that are white on outside and black on the inside

Add fresh duckweed to jars

Figure 2: Fresh duckweed added to jar spike with nutrient solution and 7-pharmaceutical compounds in tap water

1. Acclimatize duckweed in container with 10+ liters of Farrell.v2 nutrient soln. (1:400 "A" and 1:1000 "B,C,D") for 24 hours


2. Spoon approximately 300g-fresh duckweed and strain excess water with butterfly-netting


3. Place squeezed duckweed into tared aluminum dish, adjust weight to approximately 300g-fresh, and record fresh weight.


4. Dump duckweed into appropriate jar with nutrient soln. and pharmaceutical compounds.


5. Set aside an aliqout of fresh material prepared the same way to obtain the dry weight (60 degC).

Add macerated duckweed to jars

Figure 3: Macerated duckweed being added to test jars after 2 minutes in blender and continuously mixed in 1000 mL beaker to maintain homogenous mixture

1. Prepare fresh duckweed according to instruction above for enough jars that will require macerated duckweed (i.e. triplicate jars x 3 time periods x 300 g-fresh/jar = 2700 g-fresh duckweed)


2. Dump duckweed into a blender with no more than one-third the required tap water (i.e. 9 jars required * 400mL/jar * 0.33 = 1200 mL)


3. Macerate fresh duckweed in the blender at low speed (to reduce foaming) for approximately 2 minutes


4. Decant liquid into a glass beaker, rinsde blender to recover all the solids.


5. Mix contents of beaker with glass stir-rod to obtain a uniform suspension of solids.


6. While mixing, pour the proper amount of solution into a graduated cylinder and then to the proper test jar for macerated duckweed (e.g. 2700g-fresh/1200mL per 9 jars = 300g-fresh/133mL per jar)


7. Rinse walls of the graduated cylinder with tap water in order to recover all solids and then bring the volume up to approximately 400 mL like the other test jars.

Figure 4: Jars ready for experiment in triplicate for treatments (L to R) A-dark no plants, B-dark w/ plants, C-light w/ plants, D-light no plants, and E-light w/ macerated plants
Begin Microcosm Experiment

1. Place all light-treatment jars 40" below 1000 W HPSLs (200 umol/m^2-sec)


2. Place all dark-treatment jars underneath a cardboard box close to the light jars in order to maintain at similar temperatures.


3. Record the testing start time.


4. Stop tests and separate liquids/solids at time = approximately 2 hours, 24 hours, and 4 days

Figure 6: Microcosm experiment started with treatment jars spread randomly 40 inches beneath a 1000 W High Pressure Sodium Lamp

Separate liquids/solids* for jars containing duckweed

Video 1: Separating duckweed solids from the liquid

1. Record stop time and jar weight


2. Cover open top of jar loosely with butterfly netting (leave a little pouch of netting above the jar), hold in place with rubber band around the mouth of the jar


3. Invert jar into a 1000 mL beaker


4. Squeeze excess liquid from duckweed solids without removing the butterfly netting from the jar


5. Rinse the inside of the jar and butterfly netting with tap water (waste this liquid) for approximately 10 seconds to rinse solids sticking to the jar and capture as much solids as possible in the bottom of the netting.


6. Remove the netting from the jar (don't squeeze yet), pour the liquid captured in the 1000 mL beaker back into the testing jar


7. Rinse the 1000 mL beaker and hand to dry for the next harvest


8. Twist the netting and squeeze the excess liquid from the duckweed back into the testing jar for a count of seven.


9. Tare an aluminum weighing dish


10. Put the clump of freshly harvested duckweed solids onto the tared dish, scrape excess solids from the netting


11. Record fresh weight of duckweed and set in 60 degC oven (dry for 10 hours until a stable dry weight is obtained, try not to exceed 24 hours). Once dry, solids are ready for extraction preparation steps**
    


12. Record the weight of the test jar plus liquid, place a lid on the jar which is now ready for the cleanup step (perform cleanup step within 6 hours of harvesting)

*Note: macerated plant sample are filtered to separate solids using several glass fiber pre-filter pads (e.g. AP40 or Whatman GF/A). Filter pads are tared, collected in 15 mL centrifuge tubes, dried at 60 degC, weighed after drying (total dry weight - (filter pad tare(s) + centrifuge tare) = dry macerated plant solids), and filled to 15 mL with methanol preparatory to extracting the solids.
**Note: thermal degradation at 60 degC should not be an issue. Thermal degradation (TD) charts for each compound show that it does not create significant removal below 100 degC. On the other hand, the table below shows that the percent of compound recovered after evaporating the compound to dryness at 60 degC and then re-constituting to 1 mL and analyzing by HPLC/MS gives lower than expected recoveries for several of the compounds. The reason for the loss is probably more a function of evaporating to dryness than it is thermal degradation since none of the compounds should lose more than 10% mass at 60 degC (see quantities below). During the concentration step, the EPA 1694 method only recommends evaporating to 1 mL and then bringing the volume up to 4 mL. If the sample is concentrated below 0.5 mL then it extraction efficiencies decrease.

Table: Extraction Efficiency post 60 degC Evaporation to Dryness

Expected temperature when thermal degradation becomes an issue:

Sulfamethoxazole (above100 degC)

Carbamazepine (above 170 degC before transformation and less than 10% loss under 70 degC)

Acetaminophen (above150 degC)

Fluoxetine (99% @ 185degC and 95% @ 211 degC)

Progesterone (above130-200)

An explanation about "Thermogravimetric analysis"

Table: Liquid Extraction Efficiency

Table: Compound Percent Recovery and Loss due to not rinsing/silinizing glassware during Liquid Solid Phase Extraction procedure


Figure 7: Jar with duckweed plants at conclusion of test. Roots lengthen as nutrients deplete and necrotic duckweed tissue settles to the bottom of the jar.

Figure 8: Fresh duckweed recently separated from the solids and ready to dry for 10-24 hours in a 60 degC oven until a stable dry weight is obtained.

Liquid cleanup (Modified from the EPA 1694 Test Method):

Figure 9: Liquid cleanup retaining organics from the liquid phase onto Waters Oasis HLB 6cc SPE cartridges

• Note: If solids are visible in the liquid then it is highly recommended to filter the samples before passing contents through the Waters Oasis HLB 6cc Solid Phase Extraction cartridge which will speed up extraction times and allow for more liquid to pass through the cartridge


• Note: When pre-filtering liquids, glass-fiber filters (AP40 and GF/A 1.6 pore size used for these tests) are recommended; however, perform tests to determine the pharmaceutical retention by the filters.

Filter liquid

1. Rinse 500mL Erlenmeyer flask/filter apparatus with methanol and tap water


2. Place AP40 or GF/A filter pad on the filter apparatus
   


3. Slowly pass liquid sample through the filter(s) and the rinse gently with deionized water


4. Put all used filter pads for each sample into a 15 mL centrifuge tube


5. Pour Erlenmeyer flask contents back into the test jar, rinse with deionized water, and record liquid weight (to be used in case not all contents pass through the Oasis HLB cartridge)


6. Clean filter apparatus and flask with methanol and then decant into the 15 mL centrifuge tube containing the used filter(s), and bring to the 15 mL mark with methanol.


7. Liquid now ready for cleanup. Vortex (5 min) and sonicate (30 min) in an ultrasonic bath before evaporating and HPLC/MS anaslysis.

Load Cartridge

1. Clean vacuum tubes by rinsing with methanol, blowing air through the tubing, and rinsing again with deionized water


2. Precondition Oasis HLB 6cc (Waters PN# 106202) solid phase extraction cartridges: **Note: due to the nature of the pharmaceuticals selected for this study, all liquid samples were extracted under neutral conditions (i.e. no pH change) in order to obtain the highest recoveries for all seven compounds


3. Rinse cartridges with at least 3 cartridge volumes of methanol (approximately 6 mL per cartridge), do not let it go dry


4. Rinse cartridges with at least 2x the methanol rinse volume (approximately 30 mL), do not let it go dry


5. Attach vacuum tubes to cartridges and sample bottles


6. Load cartridge with sample(s) at a rate of 5-10 mL/min (approximately 7 mmHg)--watching constantly to prevent cartridge from going dry


7. Dry cartridge under vacuum for approximately 5 minutes


8. Elute cartridge with 10 mL of methanol at a slow rate (approximately 3 mmHg), capture in borosilicate vial


9. Re-condition cartridges with 20-30 mL methanol and save for future use

Concentration

1. Clean receiver flask by soaking in soapy water for at least 3 hours, rinsing with deionized water, followed by rinsing with methanol. To improve recoveries by approximately 10% then silanize the receiver flask.
1a. Silanize receiver flask (Baker, 2011) by pouring dichloromethyltoluene (DCMT) solution into the receiver flask and then decanting back into the DCMT bottle
1b. Rinse receiver flask with methanol, fill with methanol, and let stand for at least 1 hour
2. Pour 10 mL eluted volume into the receiver flask and dry to 0.25 mL (minimal precipitation) or 1mL (visible precipitation/floc) under a gentle stream of nitrogen gas (10-15 psi) at 55 degC (TurboVap)
2a. While drying, thoroughly rinse the receiver flask sidewalls (increases percent recoveries by more than 50% for several compounds) as the solution evaporates. Rinse with the solution in the flask, adding more methanol is not typically necessary.
2b. As solids precipitate to the glassware, rinse and sonicate the receiver flask.
2c. When the solution reaches approximately 2-5 mL then rinse the receiver flask using a disposable pasteur pipette

Figure 10: Solids precipitating from solution during the evaporation step. Continuous rinsing and ultrasonication is recommended to prevent buildup of solids on the receiver flask.

3. Bring volume up to 1 mL (minimal precipitaiton) or 4 mL (visible precipitation/floc) with 0.1% formic acid
3a. If the dilluted solution still appears cloudy then it will be necessary to centrifuge the solution.
3b. Pipette dilluted solution into a polypropylene microcentrifuge tube
3c. Centrifuge for 3 min at 10,000-11,000 rpm
4. Pipette sample into 2 mL HPLC glass vials with teflon septum. Add spikes and/or internal standards, if necessary. Samples are now ready for HPLC/MS analysis.

Figure 11: Agilent HPLC/MS

Figure 12: 2 mL HPLC autosampler vial representing several hours of work including filtration/separation, cleanup, concentration, and centrifugation.

HPLC/MS Instrument Configuration

1. Devices: Agilent 1200 Series High Performance Autosampler SL (PN# G1367C); Agilent 1260 Series Binary Pump (PN# G1312B); Agilent 1200 Series Thermostatted Column compartment SL (PN# G1316B); Agilent G6220A Time of Flight Detector (PN# G6220A)


2. TOF/Q-TOF Mass Spectrometer: Ion Source: Dual ESI, Ion Polarity: Positive Mode; Gas Temp. 350 degC; Gas flow 12 L/min; Nebulizer 25 psi; Scan Vcap = 3500 V; Reference masses 121.050873 to 922.009798; Chromatagram TIC type; 9.2 min stop time; 10 min run time w/ 3 min runtime between runs


3. Binary Pump: Flow 0.35 mL/min; Pmax = 500 bar; Solvent "A" 0.1% Formic Acid and 0.1% MeOH; Solvent "B" 90/10 Acetonitrile/H2O + 0.1% Formic Acid; Pump Time Table: 0 min. @ 3% "B", 1 min. @ 3% "B", 5 min. @ 100% "B", 8 min. @ 100% "B", and 10 min. @ 3% "B"


4. Column: Agilent Eclipse Plus C18 (2.1 x 55 mm, 1.8 um) PN# 959741-902; Column-SL temperature 30 degC


5. Autosampler: Injection Volume 5 uL w/ 3 sec. needle wash; Draw position 3 mm; Draw/Eject speed 200 uL/min; Equilibration time 0-2 sec. ***Custom Injection Program for Internal Standard Runs: 1st Eject, 2nd Draw 5 uL Atrazine-d5 (instd), 3rd Needle wash 3 sec., 4th Draw 5 uL sample, 5th Needle wash 3 sec, 6th Wait 2 sec., 7th Inject 10 uL
   

Solids Extraction
(still in the process of developing this method since recoveries are very poor)

Video 2: Demonstrating the process of preparing dried duckweed for extraction (demonstrating steps 1-4 below)
Figure 13: Crushed duckweed following 60 degC drying

1. Place 60 degC dried duckweed into mortar/pestle


2. Crush duckweed


3. Pour crushed duckweed into weighing dish, record mass, and then pour into 15 mL centrifuge tube


4. Rinse aluminum weighing dish (used for drying) and mortar/pestle with methanol. Save methanol and add to the 15 mL centrifuge tube. Repeat rinsing until centrifuge tube is full.


5. Allow solids to sit in methanol for >24 hours


6. Vortex sample 5 min


7. Ultrasonic bath for 30 min


8. Centrifuge 5000 rpm for 5 min


9. Decant solution into receiver flask


10. Add 10 mL methanol + 3 mL deionized water


11. Repeat vortex/ultrasonic bath/centrifuge/decant steps


12. Add 10 mL methanol and repeats vortex/ultrasonic steps


13. Pour all contents through a filter into the receiver flask


14. Rinse solids/filter with methanol


15. Evaporate solution to 5-10 mL under a gentle stream of nitrogen gas (10-15 psi) and 55 degC water bath (TurboVap) *15a. Constantly rinse and ultrasonicate the glassware as the volume decreases and more solids precipiate from solution


16. Bring the volume to 200-500 mL with deionized water, the solution is now ready for the cleanup/elution/concentration steps

Figure 14: Extracted solids following three repetitions of vortexing, sonicating, and centrifuging solids in a mixture of methanol and water. The chlorophyll comes out of the plants and turns the solution very green. Solids begin to precipitate out of solution below 10 mL.

Figure 15: Solids have been extracted in MeOH, concentrated, and now being re-dilluted to 200-500 mL

Figure 16: Solids have been extracted, concentrated, re-dilluted and are ready for SPE cleanup

Figure 17: My lab assistants and the future generation of duckweed-experts. Plus, the reason all my waking (and non-waking) hours are not spent in the lab figuring out pharmaceutical extractions from solids.

Figure 18: Four 2 mL HPLC autosampler vials ready for HPLC/MS analysis representing at least 8 hours of work to obtain 200 uL to 1 mL of sample.

Figure 19: Sonication by a probe immersed direction in the solid/solvent solution

Figure 20: Sonication by an ultrasonic bath

Table: Solids Extraction Efficiencies using EPA 1694 Method

*** Added ***

Extracting Pharmaceuticals from Solids:

1. Follow steps outlined previously to dry/crush solids. Store crushed solids (approx. 500 mg-dry) in a 15 mL centrifuge tube with 12mL-methanol plus 2 mL-deionized water.

2. After storage in MeOH for 24 hours, vortex for 5 min, centrifuge, and then:
a) if the expected ng/mL concentration is more than 100 ng/mL, run 1 mL sample on HPLC/MS (approximately 1:14 dilution).
Note: even though the instrument can detect concentrations around 5 ng/mL +/- 10%, the matrix effects produce readings that are 5-10x's less than equivalent concentrations samples without matrix interferences.
b) if the expected concentration is less than 100 ng/mL, decant supernatant from 15 mL centrifuge tube into 15 mL disposable glass culture tube. Place in 60 degC oven and evaporate to 4 mL. Add 1 mL-methanol, vortex in order to clean the sidewalls of the vials, and then micro-centrifuge sample. Analyze 1 mL sample on HPLC/MS (approximately 1:5 dilution).
Note: Formic acid causes solids to precipitate from solution therefore was not added to the solids extracted samples.

Figure 21: Disposable culture tubes after evaporating sample to 4 mL in 60 degC oven. Afterwards 1 mL MeOH added to each vial and samples were centrifuged to clean the sidewalls of the vials. Finally, samples were micro-centrifuged and 1 mL was analyzed by HPLC/MS.

HPLC/MS Analysis using matrix standards:

Note: Previous attempts to extract the solids using the TurboVap received poor extraction efficiencies due to matrix effects and solids precipitating from solution as the evaporation volume decreased. Running samples in Methanol without severe evaporation steps provided as good or better results.

1. Add 500 mg (dry) duckweed--that has not been exposed to pharmaceuticals--to a 15 mL centrifuge tube. Add 12 mL-methanol plus 2 mL-deionized water. Store for 24 hours.
2. After storage, vortex 5 min. and centrifuge samples.
3. Evaporate to 5 mL if extracted solids are also being evaporated, vortex, and centrifuge.
4. Prepare at least five 1-mL aliquots of solution from which to make the matrix standards. Spike in the appropriate amount of secondary pharmaceutical standards to make the working standards.
Note: Also prepare one 250 ng/mL matrix-free methanol standard in order to obtain the correct retention times. This is recommended because the matrix standards have more interferences and it's harder to be certain of the retention times without running a matrix-free methanol standard). This matrix-free methanol standard will not be used to construct the actual calibration curve.
5. Run the matrix-standards with the extracted solids samples to create a matrix-calibration curve which takes into account interferences that reduce the signal sensitivity of the HPLC/MS. The matrix-sample signals are typically 2-10x's less than matrix-free signals.


Note: Multiply the results by the dilution factor (if necessary) and then by the Extraction Efficiency multiplier (i.e. the inverse of the extraction efficiency). If appropriate, determine the half-life of the compound and adjust the final concentration depending on how long the sample was stored in methanol before extraction/analysis. The half-life values in the table below were calculated by measuring the concentration at t = 1 day and then 34 days later

Half-life equation: [Ct=34] = [Ct=1]*exp(-0.693*t/t0.5)

Table: Solids Extraction Efficiencies using Matrix Standards and Degradation at 4 days due to Halflife decay

Final Results

Fraction Recovery in the:

Liquid

Solids

Liquid+Solids

Methods Conclusion:

1. Liquid analysis. The modified preparation and analysis procedures are acceptable.
2. Solids extraction. The efficiencies by EPA 1694 and modified method (matrix standards without TurboVap and SPE in order to reduce steps) are poor. Neither method is robust. Extraction efficiencies improved in the modified method (except Cotinine).
3. Mass balance. Pharmaceuticals not accounted for in the liquids were not completely accounted for in the solids. This may be due to poor extraction (i.e. inability to desorb compounds), physical/chemical reactions that degrade the parent compound over time (as seen in the half-life degradation study), or due to biological degradation by duckweed or attached microorganisms. The solid samples for this study remained in methanol for 46 days before analysis--allowing sufficient time for all but cotinine and acetaminophen to degrade by at least 50%.

Results/Discussion:

Treatments

"X" Control reactor. Dark. No plant. Covered.
"A" Dark control reactor. Dark. No plant. Uncovered.
"B" Dark plant reactor. Dark. With plant. Uncovered.
"C" Light plant reactor. Light. With plant. Uncovered.
"D" Light control reactor. Light. No plant. Uncovered.
"E" Macerated plant reactor. Light. Macerated plant. Uncovered.

Chemicals

Acetaminophen = "ACT"
Sulfamethoxazole = "SLF"
Fluoxetine = "FLX"
Carbamazepine = "CRB"
Progesterone = "PRG"
Nicotine = "NIC"
Cotinine = "COT"

Physical/chemical removal

In the case of the control reactors, there were three types. First, controls in the dark without plants and covered with a lid. Second, controls in the dark without plants and uncovered. Third, controls in the light without plants and uncovered. There was no significant difference between the "A" dark and "D" light control reactors (p-value > 0.05 in all treatments as seen in these significance results testing for significance between treatments. Note: significance occurs when "p-adj" is less than 0.05). The uncovered plants were significantly less than the covered reactors, suggesting that some removal occurred due to exposure to the air (e.g. volatilization loss).

Loss due to sorption to glassware = LGS = (1 - reactor "X") x 100
Note: Sorption to glassware can be a significant loss during treatment and analysis. When evaporating samples after solid phase extraction rinsing and siliconization was an important step to prevent up to 70% of compounds from remaining on the glassware. This study (Baker 2011) researches this phenomenom.



In the case of acetaminophen, the results suggest that active uptake by the plant may play a major role in compound loss--which is expected due to its low pKow value (0.46)--see Dettenmeir 2008 (USU Thesis and Dissertations @ Digital Commons). The liquid removal charts show that whole plants removed significantly more compound than the macerated plants. One study showed that plants are able to metabolize acetaminophen by glucosidation (Huber 2009). The metabolites listed in the study were searched using Agilent's Qualitative MassHunter program. Some of the metabolites were identified in the liquid samples and increased (with respect to the area of the chromatograms) as the parent compound decreased. However, these studies require more careful analysis to ascertain whether the compounds identified were actually the metabolites.

Metabolites

Click here to view the analysis for metabolites, particularly the Acetaminophen metabolites acetaminophen-glucoside (molecular formula C14H19NO7, m/z = 314.1234) and acetaminophen-cysteine (molecular formula C11H14N2O4S, m/z = 271.0747). Other metabolites searched for included: Progesterone metabolite 5-alpha-pregnane3,20-diol (molecular formula C21H32O2, m/z = 317.2475); Fluoxetine metabolite Norfluoxetine (molecular formula C16H16F3NO, m/z = 296.1257); and Sulfamethoxazole metabolite N4-acetyl-sulfamethoxazole (molecular formula: C12H13N3O4S, m/z = 296.0700), and Carbamazepine metabolite Carbamazepine-10,11-epoxide (C15H12N2O2).

In the case of progesterone, the liquid recovery results show that progesterone levels decrease immediately in the macerated samples and then whole plant sample reach the same removal levels within 24 hrs. This trend suggests that removal is achieved by sorption to duckweed cells and is supported by the fact that progesterone removal by duckweed fits a Freundlich isotherm at 26 degC well--this is expected due to its large pKow value. The rapid decrease in the macerated samples is attributed to more sorption sites due to the intracellular tissue in solution. Recovery from the liquid was poor most likely due to: inability to desorb progesterone from duckweed, matrix effects, or possibly metabolism by microrganisms attached to duckweed tissue.

In the case fluoxetine, the removal also fits a Freundlich isotherm (removal was repeated for three different concentrations) suggesting removal by sorption. However, unlike progesterone, fluoxetine recovery in the solids was better. Sorption is expected because this compound is positively charged with a large pKow.

In the case of carbamazepine, removal was not significant in any treatment. Typically, carbamazepine has poor removal from wastewater treatment plants. However, another reason for poor removal may be due to concomitant use with fluoxetine. Fluoxetine inhibits the reaction of carbamazepine (drugs.com):

"Fluoxetine^{{116}{117}{119}{121}} or Fluvoxamine^{127}{128}
Concurrent use with carbamazepine may inhibit the metabolism of carbamazepine, resulting in increased plasma concentrations and toxicity; carbamazepine plasma concentrations should be monitored."

In the case of sulfamethoxazole, it appears that sorption and desorption both occurred. This may be attributed to the multiple pka's of sulfamathoxazole. Poor sorption is expected when the compound is negatively charge. The fluctuating sorption/desorption could take place if there is an equilibrium process going on and the compound is going back and forth between neutral and negative charges. The compound should always be negative at pH 7.5-8.0 is the pKa is approximately 5.8; however, www.chemaxon.com predicts a pKa value near 7.7 which would explain the back and forth charge.

Preliminary Literature Review

Title: Literature Review for Duckweed Systems: P-removal, Growth, and Harvest
Created: 29 April 2010
Author: Jon Farrell

Background:

Nutrient removal from wastewater prevents eutrophication from occurring downstream where the wastewater is discharged into water bodies such as rivers and reservoirs. One nutrient removal system that has been researched extensively over the past 40 years (Culley) utilizes duckweed plants (Lemnaceae) which uptake nutrients like N, P, K, Ca, and Mg into its biomass as it grows. Duckweed systems rely on three basic principles: nutrient uptake, harvesting, and solids management.

Duckweed plants typically contain more phosphorus in its tissue than other floating plants, which makes them suitable for phosphorus removal (Alaerts, Reddy). Duckweed systems usually treat sewage lagoons that receive weak municipal wastewater containing 1-4mg-P/L; however, duckweed is also used to treat swine lagoon waste containing 62.5-135mg-P/L (Chaiprapat).

Harvesting is an essential component of duckweed nutrient removal systems because it physically removes the phosphorus from the system via the biomass. Without harvesting, the plant tissue would die, settle to the bottom of the lagoon, decompose and then release the phosphorus and other nutrients back into the water column. This harvested biomass can be used as compost (Donahue), fodder rich in protein (Culley), or to generate fuel like methane (Clark).

Duckweed:

Duckweed grows naturally in almost every region with a growing season of at least five months. Most studies involving duckweed take place in climates with 9-10 month growing seasons; however, several also take place in regions with only 5-7 month growing seasons (Culley). Duckweed is a monocot, it floats on water, and has one of the fastest growth rates of any of the macrophytes. Duckweed is the common name for the Lemnaceae family of plants, with species like Lemna minor, Lemna Gibba, Spirodela Polyrhizza, and Wolffia (genus name). Duckweed studies range from full-scale operations with ponds covering 200m2 (Edwards) to 11 acres (Donahue); to pilot scale operations with only a few m2 (Reddy, Zimmo), to lab scale tests in jars with only 0.004 m2 surface area (Chaiprapat).

Phosphorus Removal:

Many studies have pointed out a direct correlation between %P in the plant tissue and the available P in the water column (Alaerts, Culley). As the PO4-P (bioavailable P) concentration in the water column decreases so does the %P in the tissue. While 1%-P is very common in oven dried duckweed, values have been reported from 0.3 up to 2.6%-P. Percent dry matter ranges from 5.4-8% with 69-86% being the organic (volatile) fraction. The N:P ratio is typically 5:1 (Alaerts, Edwards). Knowing the %P and %N in the duckweed tissue helps to construct a mass balance identifying the fate of phosphorus in the system.

Up to 100% phosphorus removal has been reported in bench scale tests (Chaiprapat); however, 60-75% phosphorus removal (Alaerts, Zimmo, and Kadlec) is more common. These same reports have identified duckweed biomass as contributing 13-47% of the total phosphorus removal, and one account attributes all of it to duckweed. The phosphorus concentration in the effluent coming from duckweed systems almost always falls below 1mg-TP/L and frequently less than 0.53mg-P/L down to 0.05mg-P/L (Willet, Edwards, Alaerts). Edwards observed that duckweed growth decreases when phosphorus levels fall below 0.3mg/L.

Duckweed Growth:

Most studies recommended starting and maintaining duckweed systems with enough duckweed to fully cover the surface area. Full coverage provides some of the highest growth rates (Reddy), but perhaps more importantly, it prevents algae proliferation that out competes the duckweed (Edwards, Al-Nozaily, Lemna Corp.) and leads to decreased productivity. Starting densities should be kept in the linear range between 10-120 g(dry)/m2 for Lemna minor (Reddy). Reddy and Edwards recommended starting with 10-11.9 g(dry)/m2; Culley, Chaiprapat, and Zimmo recommended 30-40 g(dry)/m2; while Willet, Lemna Corp., and Alaerts recommended 80-132 g(dry)/m2. Starting densities with fresh duckweed ranged from 500 to approx. 1500 g/m2. Typical seasonal yields ranged from 3-9.5 tons(dry)/ac·yr. Maximum yields between 17-25 tons(dry)/ac·yr have also been reported (Alaerts, Edwards). The relative growth rate (RGR, gnew/gold·day) of duckweed ranges from 0.06-0.121 for many systems (Chaiprapat, Culley, Willet) up to 0.24-0.31 for lab experiments. Al-Nozaily observed that light intensity was the single most important variable controlling RGR, and recommended providing 200-300 umol/m2·sec (ppf) for highest growth rates indoors.

Several factors limiting growth rates have been observed. Growth rate decreases as biomass accumulates to the point that fronds start overlapping each other (Al-Nozaily, Chaiprapat, Culley, Reddy). Growth rate decreases with nutrient depletion (Chaiprapat, Culley, Edwards). Duckweed prefers ammonium (NH4) to ammonia (NH3), and growth decreases when NH3>NH4 or when pH exceeds 9.25 (Al-Nozaily, Culley). Phosphorus precipitation also occurs at pH near 9.3, which also leads to nutrient deficiencies and lower growth rates. Several studies indicated that wind or movement decreased growth (Edwards, Willet). Biomass started depleting at temperatures below 17°C, and completely disappeared below 5°C (Donahue, Zimmo). Growth rate also decreased due to competition between species. Edwards noticed that Wolffia out-competed the Lemna species and yielded less biomass due to its smaller plant size. Aphids living atop duckweed mats in some instances were associated with decreased growth as well (Zimmo, Edwards).

Harvesting:

The frequency of harvesting and the amount of biomass removed per harvest varies from study to study. However, consistent observations include: 1) Maintain 100% coverage to reduce algae growth; 2) Harvest at least once every 20 days—the more frequent the better for nutrient removal; and 3) Harvest frequency and amount often depends on the available manpower and equipment available to harvest.
Continuous harvesting prevents overcrowding, biomass death, and release of nutrients back into the water column. Culley reported that up 50% of the N & P in the biomass gets released if more than 20 days go by between harvests. Alaerts harvested approx. 4.5mg(dry)/m2·day, Willet harvested 50% after the biomass had doubled the starting density, while Edwards harvested every 2-15 days depending on whether it was the dry(warm) or wet(cool) season, respectively.

Harvest rates depend not only on duckweed growth, but also on the ability to physically harvest the system. Donahue, superintendent of a Lemna Corp. duckweed covered lagoon in Boulder City, NV, reported harvesting the entire lagoon every week. This required harvesting 11acres/wk. at a rate of 37g/m2·week. This yielded approx. 71 tons (dry)-duckweed per year. Two people worked 10 hr. shifts M-Th and used mechanical harvesters with 4 ft. wide conveyors to remove the fresh duckweed that was then loaded into trucks and composted at the local landfill. Donahue reported that the duckweed system was used for approx. 10 years before being shut down because they could not keep up with the quantity of duckweed produced. Hence, careful solids management programs are necessary to guarantee sustainable and long-lasting duckweed systems.

Experimental Setup:

Background:

This study looked into the practicality of using a duckweed system to remove phosphorus from the Wellsville (UT) Municipal Sewage Lagoons. These lagoons were constructed in the 1960’s and cover 56 acres. Currently, this is a 0.5 MGD system that is expected to increase flow during the next 10 years to the point that the Utah Dept. of Environmental Quality (UDEQ) is concerned that it will not be able to meet its allowable 432kg-P/yr. discharge permit. A duckweed system for phosphorus removal seems promising in Wellsville for two principal reasons: 1) native duckweed plants (a mixed culture of Lemna minor and Wolffia) already cover the entire surface of the lagoons for at least 6 months (May through October); and 2) Wellsville has weak wastewater with approx. 4mg-P/L which results in a loading of approx. 12.2g-P/m2·yr. which is in the recommended <20g-P/m2·yr. range (Kadlec).

Material and Methods:

Phosphorus removal and duckweed growth:

A mixed culture of L. minor and Wolffia was seeded into approx. a 113 L acrylic reactor (3 ft. L x 2 ft. W x 8 in D) and divided into 3 sections simulating 3 lagoons; an identical reactor was placed next to it without duckweed as a control. The experiment took place for one year in a 25°C constant temperature room. High-pressure sodium lamps (HPSLs) were suspended 48 in. above the plants and provided 300 umol/m2/sec (ppf) 16 hrs/day. Raw wastewater from Wellsville influent was continuously fed with peristaltic pumps at a rate of approx. 1.77 Lpd. An average 66% of the influent flow evaporated per day, and so dilution tap water was continuously fed at 0.64 Lpd to provide enough effluent. The effluent was captured in 15 L buckets. Duckweed was re-seeded a few times at starting densities ranging from 15-90 g(dry)/m2 . Plant harvesting occurred every 7-14 days and removed 25-75% of the coverage. Plants were oven dried at 105°C to get dry mass of duckweed.

Measurements:

Total Phosphorus measurements were made with HACH test kit method 10127. Reactive Phosphorus (PO4-P) measurements were made with ascorbic acid APHA Standard Method 4500P-E. Total Nitrogen, Ammonia, and Nitrate measurements were made with HACH test kit methods 10071, 10031, and 10020, respectively. Alkalinity, TSS, and VSS measurements followed APHA Standard Methods. pH and DO measurements were made with Corning and Hanna probes, respectively. Duckweed tissue samples were measured by the Utah State University Analytical Lab (USUAL). Phosphorus concentrations in plant tissue and sediments were also measured for PO4-P following dry ashing at 550°C with subsequent wet aqua regia digestion; these results were verified with standard grape petiole leaves with a known 0.38%P dry weight.
References:

Alaerts, G. J., M. R. Mahbubar, and P. Kelderman. 1996. Performance analysis of a full-scale duckweed-covered sewage lagoon. Wat. Res. 30(4):843-852.

Al-Nozaily, F. G. 2001. Performance and Process Analysis of Duckweed-Covered Sewage Lagoons for High Strength Sewage. Rotterdam, NL: A. A. Balkema.

Chaiprapat S., J. J. Cheng, J. J. Classen, and S. K. Liehr. 2005. Role of internal nutrient storage in duckweed growth for swine wastewater treatment. Transaction of the ASAE. 48(6):2247-2258.

Clark, P. B., and P. F. Hillman. 1996. Enhancement of anaerobic digestion using duckweed (Lemna minor) enriched with iron. J. of the Chartered Institution of Water and Environmental Management. 10(2):92-95.

Culley Jr., D. D., E. Rejmankova, J. Kvet, and J. B. Frye. 1981. Production, chemical quality, and use of duckweed (Lemnaceae) in aquaculture, waste management, and animal feeds. J. World Maric. Soc. 12(2):27-49.

Donahue, Don. Superintendent of the Boulder City (NV) Wastewater Treatment Plant. Personal correspondence. 3/18/2009.

Edwards, P., M. S. Hassan, C. H. Chao, and C. Pacharaprakiti. 1992. Cultivation of duckweeds in septage-loaded earthen ponds. Bioresource Technology. 40:109-117.

Kadlec, R. H. 2009. Treatment Wetlands. 2nd ed. Boca Raton, FL: CRC Press.

Lemna Corporation. 1996. Operation and Maintenance Manual for Boulder City, Nevada . St. Paul, MN: Lemna Corporation.

Reddy, K. R., and W. F. De Busk. 1985. Growth characteristics of aquatic macrophytes cultured in nutrient-enriched water: II. Azolla, duckweed, and salvinia. Economic Botany 39(2): 200-208.

Willet, D. 2005. Duckweed-based Wastewater Treatment Systems: Design Aspects and Integrated Reuse Options for Queensland Conditions. Brisbane, AU: DPI&F Publications.

Zimmo, O. R. 2002. Process performance assessment of algae-based and duckweed-based wastewater treatment systems. Water Sci. and Tech. 45(1):91-101.