*3.1.2 Highly concentrated NH4 solution*

The above operation was repeated by replacing the spent formulated sample a new one in the stripping column and dissolving the stripped NH3 gas by reaction with sulfuric acid in the recovery column until the NH4 <sup>+</sup> concentration in the recovery column no longer increased. The NH4 <sup>+</sup> concentration in the recovery column is determined by the solubility of the product, (NH4)2SO4, which is 744 g/L at ambient temperature. With this concentration, the concentration of NH4 <sup>+</sup> is 19%, a theoretical number. We reached 18% in our experiment. If nitric acid is used, the NH4 <sup>+</sup> concentration would be more than 33%, given the solubility of NH4NO3 in water, 1500 g/L.

To demonstrate the liquefaction of the highly concentrated NH4 <sup>+</sup> solution we prepared, we set up an experiment to produce liquefied NH3. We heated 100 mL of the 18% NH4 <sup>+</sup> solution at 90°C, vaporizing NH3, and sent the NH3 gas, through a glass condenser into a metal cylinder half-submerged in iso-propanol which was cooled to 60°C by dry ice. The moisture generated by heating the N solution at 90°C was captured by a desiccant inside the condenser. At 60°C, the NH3 gas is liquefied, while the other gases N2 and O2 stay as gas, being released to the atmosphere. About 50 mL of liquefied NH3 was collected. The volume of the recovered liquid NH3 was limited by the volume of the metal cylinder. Hence, no quantitative recovery rate was assessed. Yet, this demonstrates a possibility of NH3 liquefaction from an 18% NH4 <sup>+</sup> solution. The technology for NH3 liquefaction and the liquid NH3 transportation infrastructure already exist. The same experiments should be repeated by using a real DS sample for validation.

Very few studies have been published to report the nitrogen concentration as high as 18% in recovering nitrogen from manure liquid. The high nitrogen concentration was made possible by dissolving the NH3 gas into the water with a highly soluble acid. The other gases such as N2, O2, and CO2 gases inside the air bubble can interfere with the NH3 gas dissolving into the water. When the rising velocity of the bubbles is high, the NH3 gas can be carried away by the other gases which are not water soluble except for CO2 which dissolves somewhat. Using plastic packing materials inside the recovery column can help slow down the rising velocity.

As to the economic benefit, it is difficult to estimate since there is no market for renewable ammonia at this moment. Still, the price of N fertilizers has been going up significantly, due to the increase in the price of natural gas (NG), and can be unpredictable, given geopolitical reasons such as the economic sanction against Russia, a large exporter of N fertilizers. Using the recycled ammonia can help farmers save money. In addition, many large companies are investing in what is called Green Ammonia which uses water electrolysis followed by the Haber-Bosch process without using NG [23]. It is known that the production of Green Ammonia can cost four to five times as much as the conventional ammonia due to the high cost of water electrolysis [23]. Our process does not use electrolysis, nor does it produce NH3.

It simply recovers NH3 from the wastewater. Though a comprehensive cost-benefit comparison is not straightforward for Green Ammonia and our renewable NH3, it should be clear that recovering NH3 is much cheaper than producing it, given the extreme chemical stability of water and N2, both of which are the raw materials for Green Ammonia. Both our renewable NH3 and Green Ammonia should be qualified as non-fossil-based ammonia, and the demand for such ammonia is expected to grow massively high in the future [24].

### **3.2 Protein recovery**

#### *3.2.1 Composition of DS*

**Table 3** summarizes the compositions of the original DS sample and the leftover solid after the extraction of protein by THP on a dry matter basis. The condition for THP was the following: heating the DS sample at *T*<sup>1</sup> = 100°C for 1 hour followed by heating it further at *T*<sup>2</sup> = 160°C for 1 hour.

Almost 60% of the original protein was extracted by THP. Phosphorous mostly stayed in the leftover solid, while potassium dissolved in the solution after THP.

#### *3.2.2 Protein recovery yield*

We analyzed the protein recovery yield, *η protein recovery*, defined by the following equation:

$$\eta\_{recover}^{protein} = \mathbf{100} \times \frac{\left[\mathbf{W}\_{hydrolystate}\right]}{\left[\mathbf{W}\_{protein}\right]} \tag{7}$$

where [*Whydrolysate*] and [*Wprotein*] refer to the weights of the protein hydrolysates (PHs) in the reaction solution after THP and the weight of the protein in the original sample prior to THP, respectively. The THP condition was the following: heating the DS sample at *T*<sup>1</sup> = 100°C for 1 hour followed by heating it further at *T*<sup>2</sup> = 160°C for 1 hour. **Table 4** summarizes the recovery yield under this condition, showing a reasonably high yield.

The numbers listed in **Table 4** were determined by AAA. The experiments were performed in triplicate and an error of *η protein recovery* was within 3%. Vanotti et al. did not include the recovery yield in their patent [17].

The protein in manure digestate solid may be embedded in a complex solid matrix or trapped in a web of lignocellulosic components such as cellulose, hemicellulose, and


*a Dry matter basis.*

*b Obtained by multiplying the Kjeldahl nitrogen by 6.25. For the analytical method, refer to our earlier publication [20]. c Alkali metals such as Na, Ca, and Mg.*

*d The leftover solid after THP recovered by filtration by a screen with 90 μm mesh, dried in an oven overnight, and ground by a pestle for analysis.*

#### **Table 3.**

*Compositions of DS samples before and the leftover solid after THP (wt.%)a .* *Mitigation of Environmental Impact of Intensive Animal Farming through Conversion… DOI: http://dx.doi.org/10.5772/intechopen.105131*


#### **Table 4.**

*Protein recovery yields by THP.*

lignin. The interactions between the protein and the rest of the components in the solid may be hydrophobic in nature or electrostatic in nature associated with the functional groups of constituent amino acid residues of the protein. It is known that the dielectric constant of water decreases at high temperatures [25]. This creates two unusual characteristics for water: water favoring hydrophobic interactions and obstructing electrostatic interactions [25]. Hence, the first heating step of THP may interfere with the hydrophobic or the electrostatic interactions which may keep the protein trapped inside the solid matrix. It is also known that the pH of water goes down at high temperatures from around 7 [25]. Once the protein is released from the solid phase and dissolved into the solution phase, it should experience an acidic environment created by a low pH which may cause hydrolysis of the extracted protein. Accordingly, the dissolved protein may undergo hydrolysis, yielding short-chain peptides or individual amino acids. Although this is only speculation, the results show that protein can be extracted at a reasonably high yield by the two-step THP and what was extracted from DS was a mixture of oligopeptides and amino acids, as is shown below.

#### *3.2.3 Molecular weight (MW) distributions*

**Figure 4** displays the SDS-PAGE band images for the PH prepared under the THP condition described above. The PH exhibited very few lines, indicating very few fractions within the range analyzed.

**Figure 5** exhibits the MALDI-TOF mass spectra for the PH. The reference peptide, shown at 1046.79 m/z, was added to the sample prior to the measurements for the concentration of PH relative to the reference below the MW of 1000 Da. Contrary to the SDS-PAGE images, there are a number of peaks below 1000 Da. Peptides in this region were low-MW peptides such as oligopeptides or free amino acids. The concentration of the PH in **Figure 5** can be calculated from the peak positions and the intensities of each signal relative to the reference. It was about 3.9 g/L which is about 70% of *Whydrolysate* in **Table 4**.

Based on the results from SDS-PAGE and MALDI-TOF mass spectroscopy, we conclude that our PH had more low MW fractions than the higher MW fractions. Earlier studies reported low MW peptides exhibited antioxidant activities [26–28].

Next, we will subject our PH to antioxidant activities.

### *3.2.4 ORAC against the peroxyl and hydroxyl radicals*

**Figure 6a** and **b** show the inhibition of the peroxyl radical attack against fluorescein protein by Trolox and PH, respectively, as a function of the logarithm of the sample concentration, *C*. This assay is important since lipid molecules constituting cell membranes are prone to become peroxyl radicals that attack DNA, protein, and other molecules in a cell [29]. The curve profile shown in **Figure 6b** is very similar to that in **Figure 6a**. In fact, the values of IC50 for PH and Trolox were very close: 7.67 and

#### **Figure 4.**

*SDS-PAGE image of PH extracted from DS. The measurements were performed in triplicate. The numbers on the right side are the MW markers in kDa.*

8.08 mg/L, respectively. This observation demonstrates that the antioxidant activity of PH was as strong as Trolox against the peroxyl radicals. IC50 refers to the concentration of the sample at which the inhibition is 50%. The experimental error for IC50 was within 1 mg/L which was estimated over triplicate experiments.

**Figure 7a** and **b** display the inhibition of the hydroxyl radical attack against fluorescein protein by Trolox and our PH, respectively, as a function of the logarithm of *C*. Hydroxyl radicals are often generated inside a cell in the presence of metal ions such as Fe(II), Cu(I), and Co(II) and attack organic molecules involved in metabolic reaction pathways [30]. We observed a significant difference between the two samples: the inhibition by our PH reached 100% when log *C* was 1.5, while Trolox did not reach 100% inhibition even when log *C* was 2. IC50 of 107.6 mg/L for our PH was less than 1/7 of that of Trolox, 741 mg/L. The values of IC50 imply that the antioxidant activity of our PH was more than seven times as strong as Trolox. The strong antioxidant activities of our PH are consistent with the previous studies on peptides

*Mitigation of Environmental Impact of Intensive Animal Farming through Conversion… DOI: http://dx.doi.org/10.5772/intechopen.105131*

**Figure 5.** *MALDI-TOF-mass spectrum of PH. A signal for a peptide with a known MW is included as a reference at 1046.79 m/z.*

#### **Figure 6.**

*The inhibition of the peroxyl radical attack against fluorescein protein by (a) Trolox and (b) PH, respectively, as a function of the logarithm of the sample concentration, C.*

[26–28, 31–40]. A theoretical study on the antioxidant activity of peptides has been published [41].

The DS sample included some non-protein nitrogen compounds which were not removed from the PH sample prior to the ORAC assay; therefore, their contributions to the inhibition of the radicals cannot be ignored. Our data only demonstrates that the extracted compounds from the DS sample by our THP and recovered by UF with a 150 kDa membrane inhibited both peroxyl and hydroxyl radicals to the extent that the ability to inhibit the former radical was comparable to that of Trolox and the ability to inhibit the latter was seven times stronger than Trolox.

Our assay is an in-vitro test, and the results should be considered preliminary. Further study is warranted to confirm the antioxidant activity of our PH. If confirmed, the recovered protein hydrolysate could be sold as antioxidant feed additives. Currently, a 50% feed-grade vitamin E is sold at \$13.5/kg [42]. Using the numbers in **Table 2**, the volume of the recovered protein from manure digestate solid generated

**Figure 7.** *The inhibition of the hydroxyl radical attack against fluorescein protein by (a) Trolox and (b) PH, respectively, as a function of the logarithm of C.*

on a dairy farm would be about 5000 tons/year at the recovery yield of 60%. This would provide an annual revenue of \$137 million/year which could potentially overshadow the revenue from selling the milk. It should be noted that this is only a rough estimate under a hypothetical scenario.

The GHG emissions from agricultural activities in the U.S. were 641 Mt of CO2eq in 2019 [43]. Of that volume, 58% was due to the N2O emission caused by spraying nitrogen fertilizers including manure/digestate liquids through the mechanism mentioned earlier, and 13% was primarily due to the N2O emissions by manure management mostly from manure storage such as lagoons. N2O generated by manure management is mainly produced from organic nitrogen, mostly protein, in manure. By recovering protein from manure before manure management through the protein recovery process described in this chapter, we could potentially reduce about 83 Mt of CO2eq emissions, assuming the above 13% was generated by decomposition of protein in manure. 30 to 50% of the N2O emissions caused by spraying N fertilizers and manure/digestate liquids originate from applications of animal manure which includes organic nitrogen [3, 4]. Hence, practicing a combination of the protein and NH3 recovery processes described in this chapter could potentially reduce about 109– 186 Mt of CO2eq emissions. Altogether, up to 269 Mt of CO2eq emissions could be removed by the combination of the two processes. This number is about 42% of the total GHG emissions from agricultural activities in the U.S. To evaluate these estimates, we assumed 100% protein and NH3 recovery rates by both recovery processes. Though the actual number will be lower, a significant volume of GHG emissions can be still reduced by the two processes. The potential reduction of eutrophication caused by nitrogen runoff cannot be ignored through our processes.
