Supplementary Information for Publication

Electrochemically Induced Conversion of Urea to Ammonia

Fei Lu and Gerardine G. Botte*

Center for Electrochemical Engineering Research

Department of Chemical and Biomolecular Engineering, Ohio University, Athens, OH

* To whom correspondence: botte@ohio.edu

Contents

Section S1. eU2A/THU reactor fabrication, assembly, and operation 2

Section S2. Electrochemical Measurement 4

Section S3. Ammonia determination 5

Section S4. Urea determination 7

Section S5. Gas chromatography 9

Section S6. H2 production and energy consumption in eU2A 10

Section S7. Reaction rate of eU2A and THU in 24 hours experiments 12

Section S8. Notation 13

Section S9. References 14

1

Section S1. eU2A/THU reactor fabrication, assembly, and operation

The eU2A reactor vessel was assembled with a lid, a bottom and a hollow cylinder. All the three parts were machined at the Center for Electrochemical Engineering Research (CEER) from Teflon rods (Virgin Electrical Grade Teflon rod with 75 mm diameter from McMaster–Carr). The size of each part and the dimensions of the fittings on the lid are shown in Figure S1a. The O-ring seal was made of Fluorosilicone (Parker), which has good chemical resistance to ammonium hydroxide. Compression fittings were used to connect the thermocouple, gas outlet, sample port and electrodes’ current collectors as shown in Figure S1b. The reactor is a two electrodes system. The anode was built by filling a Ni mesh (DEXMET Corporation, 7 Ni 7-077) cylinder type basket with 2,500 Ni beads (New England Miniature Balls, with a diameter of 0.32 mm and purity 99 %) and a Ni rod (diameter 6.35 mm and purity 99.5 %) was used as the anode current collector as shown in Figure S1a. The anode is the working electrode in the reactor with a volume of 47 cm3, measured by water displacement. The cathode consisted of a round Ni mesh (bottom diameter: 6.0 cm; height: 9 cm) (DEXMET Corporation, 7 Ni 7-077) -the picture of the cathode is shown in Figure S1b. The cathode mesh was placed concentrically around the anode (the distance between the electrodes was 7.5 mm). A Ni wire (diameter 1.0 mm, purity 99.5 %, Alfa Aesar) was spot welded into the mesh of the cathode and used as cathode current collector (see “cathode” in Figure S1b). A silicone strip heater (ZORO tools, SHS80411) was pasted around the reactor vessel as heating element controlled by a temperature controller (Econo Temperature controller 12125-14) equipped with a type “K” thermocouple (OMEGA Engineering Inc.). The uncertainty of the temperature was ± 1 °C. The heater was covered by fiberglass (Frost King) and insulation tape (Polyken).

The reagent -250 ml DEF in the presence of 7 M KOH- was made by dissolving KOH (containing 10~15 wt% of water, Fisher Scientific) into DEF (32.5 wt% of urea, Old World Industries, LLC). The ammonia gas was trapped in a 1 L 1 M sulfuric acid solution made from dilution of concentrated sulfuric acid (98.0 wt%, Fisher Scientific). Between each two experiments, the cathode and anode were sonicated with 0.01 M HCl and followed by sonication in deionized water (DI) for 1 hour (Zenith Ultrasonics, G2-40) at 40kHz. Finally, the cathode and anode were rinsed with DI water and let dried naturally. In the eU2A experiments, voltage was applied once the reagent reached the desired operating temperature. The applied cell voltage was controlled by an Arbin BT2000 potentiostat. Ammonia generated and urea consumed during the heating cycle of the eU2A process were quantified and subtracted from the final results for the material balance and ammonia generation rate. The thermal hydrolysis process (THU) took place in the same reactor without applying a cell voltage between the electrodes.

2

Figure S1. Design of eU2A reactor. a: the features of the reactor vessel and working electrode. b: photograph of the cathode, anode, and reactor vessel from top view and side view.

3

Section S2. Electrochemical measurement

Three electrodes experiments were conducted in a closed beaker using a potentiostat (Solartron 1470 E) to identify the maximum current and cell voltage associated with the formation of NiOOH in the presence and absence of urea. The working and counter electrodes were Ni foils (2 cm × 2 cm, Alfa Aesar, 0.025 mm thick, 99.98 %) and the reference electrode was Hg/HgO. The solution was heated to 70 °C using a hot plate. The cyclic voltammograms (CVs) of the Ni electrode from 0 V to 0.4 V vs. Hg/HgO were obtained in 7 M KOH and DEF in the presence of 7 M KOH at a scan rate of 10 mV s -1. Any gas generated was trapped into a 1 L 1 M sulfuric acid trap. As shown in Fig. S2, in 7 M KOH, a pair of redox peak appears in the anodic and cathodic region. The anodic peak at 0.315 V (1.545 V cell voltage measured by the auxiliary channel) corresponds to the oxidation of Ni(OH)2 (Ni+2) to NiOOH (Ni+3), whereas the cathodic peaks 0.170 V and 0.200 V correspond to NiOOH reduction to α-Ni(OH)2 and β-Ni(OH)2, respectively.1 When urea is present (DEF), the oxidation current increases significantly after the peak for the formation of NiOOH, which corresponds with the electrolysis of urea as reported in the literature.1 However, in the cathodic scan the reduction peaks corresponding to the reduction of NiOOH are not observed, which indicates that the catalyst is participating in the hydrolysis of urea. As discussed in a DFT study,2 with the adsorption of OH- on the NiOOH molecule, urea decomposes to ammonia and CNO- making NiOOH reduced to Ni(OH)2, which is in agreement with the mechanism of oxidation of organic molecules on Ni proposed by Fleischmann et al.3 Therefore, to promote the hydrolysis of urea to ammonia a cell voltage of at least 1.545 V (for Ni anode/Ni cathode) should be applied to maintain the formation of Ni 3+ on the working electrode while minimizing the electrolysis of urea. Because, the cell voltage is affected by the hydrodynamics of the reactor and the electrode architecture (which could cause up to 100 mV overpotential), the ammonia production rate was measured to determine the optimum cell voltage (as demonstrated in Figure 2b of the manuscript).

Figure S2. Cyclic voltammogram of Ni in 7 M KOH and DEF in the presence of 7 M KOH at 70 °C at a scan rate of 10 mV s-1

4

Section S3. Ammonia determination

The generated ammonia was determined using an ammonia ion selective electrode (ISE) (Thermo 710A+, measurement error is ± 2 %). The ammonia ISE was calibrated by 0.121, 1.21, 12.1, 121 ppm as ammonia standard solutions (1 ppm = 5.87×10-5 M), made by dilution from 1000 ppm as nitrogen standard (Thermo Scientific, Orion 951007). At each sample point, 0.4 ml electrolyte sampled from the reactor was diluted to 100 mL (volumetric flasks were used for accuracy). The diluted solution was divided into two 50 mL solutions. One for ammonia ISE testing and the other was used as the testing sample for UV-vis spectroscopy to determine the concentration of residual urea in the reactor. On the other hand, 0.4 mL of solution from the acid trap was diluted to 100 mL for ammonia ISE testing. By adding the moles of ammonia from the

reactor and the trap, the total produced ammonia, , could be obtained as shown in eqns. (1) to (3). For example, Table S1.1 and S1.2 exhibit the produced ammonia during the 24 hours experiments in the eU2A and THU, respectively.

(1)

(2)

(3)

Table S1.1 Ammonia generation during eU2A (1.65 V, 70 °C) in 24 hours experiment*

Time,h

, ppm

, M

, ppm

, M

, mol

,mol

0 5.69 0.069 0 0 0.022 00.25 14.3 0.210 0.103 0.002 0.054 0.0320.5 21.0 0.308 0.130 0.002 0.079 0.057

0.75 29.5 0.433 0.321 0.005 0.113 0.0911.0 38.3 0.562 0.887 0.013 0.153 0.131

1.25 43.1 0.632 0.956 0.014 0.172 0.1501.5 49.5 0.726 1.62 0.024 0.204 0.1833.0 68.4 1.004 7.27 0.107 0.356 0.3344.0 75.7 1.111 13.8 0.203 0.478 0.4565.0 91.1 1.337 21.1 0.310 0.640 0.6186.0 111 1.629 34.1 0.500 0.902 0.880

18.0 68.4 1.004 77.7 1.140 1.385 1.36424.0 67.3 0.988 94.8 1.391 1.631 1.609

* Measurement error is: ± 2 %

5

Table S1.2 Ammonia generation during THU (70 °C) in 24 hours experiment*

Time,h

, ppm

, M

, ppm

, M

, mol

,mol

0 7.96 0.117 0 0 0.030 00.25 12.4 0.182 0.035 0.001 0.046 0.0160.5 17.8 0.261 0.042 0.001 0.066 0.036

0.75 21.9 0.321 0.078 0.001 0.081 0.0521.0 24.7 0.362 0.106 0.002 0.092 0.062

1.25 28.8 0.423 0.257 0.004 0.109 0.0791.5 29.3 0.430 0.263 0.004 0.111 0.0813.0 36.5 0.536 1.19 0.017 0.151 0.1214.0 42.3 0.621 1.53 0.022 0.177 0.1475.0 43.4 0.637 1.54 0.023 0.181 0.1516.0 46.8 0.687 1.59 0.023 0.194 0.164

18.0 67.8 0.995 1.67 0.025 0.271 0.24124.0 66.6 0.977 2.99 0.044 0.286 0.256

* Measurement error is: ± 2 %

6

Section S4. Urea determination

UV-vis spectroscopy was performed by using a Hewlett Packard Spectrophotometer HP 8452A to determine the amount of urea converted during the experiments. 50 ml testing samples were prepared as described in Section S3. A coloring agent4 was used to color the urea solutions. The agent was made by mixing 2 g p-dimethylaminobenzaldehyde (Fisher Scientific,), 100 mL ethyl alcohol (95%, Denatured w/IPA and Methanol, Fisher Scientific), and 10 mL hydrochloric acid (12.1 N, Fisher Scientific). A series of standard urea solutions are shown in Fig. S3. The standard solutions, 0.008, 0.016, 0.024, 0.032, 0.040 M, were prepared by dilution from 1 M urea solution made by urea powder (≥ 99 %, Fisher Scientific) to calibrate the UV-vis spectroscopy. The calibration spectra are demonstrated in Fig. S3. It is evident that the broad maximum absorbance region (420 nm to 435 nm) presented by urea containing solutions and p-dimethylaminobenzaldehyde can be used to quantify the concentration of urea. At the point with maximum absorbance, 420 nm, the urea concentration is proportional to the absorbance as shown in Fig. S3 (inset).

Testing samples were prepared as follow: 0.4 mL from the 50 mL diluted testing samples were transferred to UV-vis cuvettes and mixed with 1.6 mL coloring agent and 2.0 mL DI water to make 4.0 mL colored samples. The UV spectra of the samples were then obtained with the UV-vis spectrophotometer. Based on the linear equation in Fig. S3 (inset) and the absorbance of

each sample at 420 nm, the concentrations of the diluted samples, , were obtained.

The concentration of residual urea in the reactor, , was then calculated according to eqn. (4) and the results are shown in Table S2.

(4)

Figure S3. UV-vis spectra of standard urea solutions from 0.008 M to 0.040 M. The linear relationship between absorbance and urea concentration was obtained at 420 nm as shown in the inset figure.

7

Table S2 Concentration of residual urea in the reactor during 24 hours experiments

Section S5. Gas Chromatography (GC)

The gaseous products from both electrodes were collected together in the gas collector and qualified by GC (SRI 8610 multi-gas). The presence of ammonia in the gas phase was not expected since the system was designed to trap this gas in the acid solution. Similarly, any CO2

produced was absorbed by the KOH solution to form K2CO3 or KHCO3 in the reactor (7 M KOH was sufficient to offer OH- ions in these experiments). A 10 ml gas sample was extracted from sample port III in Fig. 1.c and injected into a molecular sieve 5A column with a thermal conductivity detector for analysis. The gas chromatograph of the gas sample is presented in Fig. S4. The sample gas consists of H2, O2 and N2. Table S3.1 gives the integrated area for each component and the corresponding volume and percentage volume of each component. In

8

Time, heU2A THU

, MSt. dev.,

M , M St. dev., M

0 4.648 0.082 4.653 0.1660.25 4.571 0.126 4.596 0.0340.5 4.454 0.083 4.530 0.124

0.75 4.305 0.120 4.480 0.0731.0 4.154 0.032 4.449 0.178

1.25 4.058 0.026 4.444 0.0011.5 4.027 0.015 4.419 0.2283.0 3.575 0.046 4.257 0.1414.0 3.534 0.079 4.242 0.1795.0 3.243 0.015 4.224 0.0226.0 2.897 0.085 4.103 0.039

18.0 1.843 0.063 3.831 0.08224.0 1.311 0.016 3.786 0.054

addition, by applying the ratio of O2 and N2 (79:21) in air, the percentage of the gases was recalculated in Table S3.2 indicating the volumetric percentage of air present.

Figure S4. Gas chromatograph of gas products in eU2A, there are no ammonia and CO2 signals since those gases are trapped in the acid trap and the electrolyte reactor, respectively

Table S3.1 Composition of gas sample through GC analysisArea Volume, mL Percentage, % St. Dev., %

Hydrogen 1319.015 6.19 61.9 1.2Oxygen 44.011 0.72 7.2 0.1Nitrogen 189.16 3.09 30.9 0.6

Table S3.2 Composition of gas sample through GC analysis normalized with airVolume, mL Percentage, % St. Dev., %

Hydrogen 6.19 61.9 1.2Air 3.39 33.9 0.7

Nitrogen 0.42 4.2 0.1

Section S6. H2 production and energy consumption in eU2A

During the 24 hours eU2A experiments at 70 oC, a cell voltage of 1.65 V was applied by the potentiostat (Arbin BT2000). The current response is shown in Fig. S5. An average current, 0.364 ± 0.002 A, was calculated using the First Mean Value Theorem for Integrals over the total time of the process (24 hours). Ammonia and H2 produced on the electrodes generate bubbles that affect the surface of the electrodes which results in unstable current. When applying the voltage, H2 evolution occurs on the cathode while active catalyst NiOOH is produced on the anode (working electrode). The amount of H2 can be calculated according to Faraday’s Law based on 100 % current efficiency as reported by others.5

(5)

The H2 production rate in 24 hours was calculated according to

9

(6)

0 5 10 15 20 25

0.25

0.30

0.35

0.40

0.45

0.50

0.55

Cur

rent

, A

Time, h

Figure S5. Current response at 1.65 V applied potential for 24 hours eU2A experiments at 70 oC

The analysis of the gases indicates that nitrogen is produced at the applied cell voltage, which indicates that a fraction of the current is used to promote the urea electrolysis reaction (reaction 2 in the paper). Eqn. (7) was used to calculate the urea consumption and N2 production due to reaction 2 in the paper. The results obtained from Eqn. (7) were used to determine the moles of urea converted into ammonia during the eU2A process.

(7)

The material balance for urea and ammonia after the 24 hours experiments in the eU2A and THU processes is summarized in Table S4. The stoichiometric ratios of generated ammonia and converted urea are 1.96 and 1.19 for the eU2A and THU, respectively as shown in eqn. (9) and (10).

(9)

(10)

Table S4 Material balance of urea and ammonia after 24 hours experimentseU2A THU

Moles of urea converted, mol 0.834 ± 0.021 0.216 ± 0.044Moles of urea converted to ammonia*, mol 0.823 ± 0.021 0.216 ± 0.044

10

Moles of ammonia produced, mol 1.609 ± 0.032 0.256 ± 0.005Ammonia generation rate per effective volume**

* Moles of urea converted to nitrogen due to electrolysis (0.011 mol) were subtracted** 47 cm3 for eU2A and 250 cm3 for THU

The ammonia generation rate per effective volume for both processes is summarized in Table S4. The effective volume is defined as the active volume for the reactions, 47 cm3

(electrode volume) for the eU2A process and 250 cm3 for the THU process. In the calculation of the ammonia generation rate per effective volume of the eU2A process the contribution of the thermal hydrolysis (0.256 moles) was subtracted.

The electrical energy consumed in the eU2A process to generate the active catalyst was calculated by:

(11)

The electrical energy consumed per gram of ammonia produced in the eU2A process is:

(12)

Section S7. Reaction rate for eU2A and THU in 24 hours experiments

First order reaction as a function of the concentration of urea6 was assumed for the calculation of the overall reaction rate constants for eU2A and THU. Because pH is higher than 12, no backward reaction occurs. Let - rA be the rate of forward reaction, CA,0 is the initial concentration, CA,i is the concentration at any sample time (see Table S2), Δti is the time interval between two samples. The average rate of forward reaction, k, can be obtained through eqns. (13) and (14) and the results are shown in Table S5.

11

(13)

(14)

Table S5 Forward reaction rate constants at different reaction conditions

Operation conditionReaction

ratek, min-1

St. dev.,min-1

eU2A 70 °C, 1.65 V 0.00140 0.00068THU (in 7 M KOH) 70 °C 0.00039 0.00033

Urea hydrolysis7 140°C, 100 kg cm-2 0.00457 -

Section S8. Notation

Concentration of ammonia, M

Concentration of ammonia in the trap, M

Concentration of ammonia in the diluted sample from the trap, ppm and M

12

Concentration of ammonia in the reactor, M

Concentration of ammonia in the diluted sample from the reactor, ppm and M

Concentration of urea, M

Concentration of urea in the reactor, M

Concentration of urea in the diluted sample removed from the reactor, M

Initial concentration of solution A, M

Concentration of solution A at sample point i, MEnergy, WhFaraday constant, 26.8 Ah mol-1

Average current, A

Reaction rate constant for urea dissociation at sample point i, min-1

Average reaction rate constant for urea dissociation, min-1

Mass of ammonia, g

Moles of hydrogen produced, mol

Hydrogen generation rate, g h-1

Moles of urea consumed by urea electrolysis, mol

Moles of nitrogen generated by urea electrolysis, mol

Moles of ammonia produced, mol

Moles of urea consumed, mol

Ammonia generation rate per effective volume, g h-1 L-1

Initial moles of ammonia, mol

Forward reaction rate for urea dissociation, M min-1

Time, min and h

Time interval between every two sample points, minApplied voltage in eU2A, V

Volume of solution in the trap vessel, mL

Volume of reagent in the reactor, mLElectrochemical equivalent, z = 2 for hydrogen evolution

Section S9. References

13

1. V. Vedharathinam and G. G. Botte, Electrochimica Acta, 81, 292 (2012).2. D. A. Daramola, D. Singh and G. G. Botte, The Journal of Physical Chemistry A, 114, 11513 (2010).3. M. Fleischmann, K. Korinek and D. Pletcher, Journal of Electroanalytical Chemistry and Interfacial Electrochemistry, 34, 499 (1972).4. G. W. Watt and J. D. Chrisp, Analytical Chemistry, 26, 452 (1954).5. B. K. Boggs, R. L. King and G. G. Botte, Chem. Commun., 4859 (2009).6. R. C. Warner, J. Biol. Chem., 705 (1942).7. J. N. Sahu, P. Gangadharan, A. V. Patwardhan and B. C. Meikap, Industrial & Engineering Chemistry Research, 48, 727 (2008).

14