The adaptive and direct evolution cycles are the machinery that produce mutations in the proteins produced by NdmA and NdmB in order to produce variants of interest. NdmA catalyzes the N1-demethylation of caffeine to produce theobromine and formaldehyde, while NdmB catalyzes the N3-demethylation of theobromine to produce 7-methylxanthine and formaldehyde.

Directed evolution induces mutagenesis in each generation by first conducting EP-PCR on NdmB and PCR on the backbone, then performing Gibson Assembly on the pSB1C3 backbone with GFP, NdmA, and NdmB. Through EP-PCR, random mutations are induced, while the selectivity of primers specifies the region of DNA. After a successful Gibson Assembly, the NdmAB plasmid is transformed into the BL21 strain of E. coli and is cultured for 2-3 days. Next, Miniprep is conducted on the samples; then, the samples are put through EP-PCR and the process is repeated. Preparation for the ADE cycle is now complete.

Adaptive evolution induces mutagenesis one time, then allows natural mutagenesis to take over. First, caffeine cultures will be filled in 24-well plates, and the media will be changed every couple days. Then, 100 µL of the culture will be recultured in a new 24-well plate to allow natural mutagenesis to occur. This process allows each well to be more uniform, ensuring that a sample taken and sequenced will be representative of the majority of bacteria present. However, with adaptive evolution, there is no guarantee of a successful mutation. Thus, our combined approach of adaptive/directed evolution (ADE) ensures the generation of a large library of mutants while ensuring consistency and uniformity. Both aspects are very important to our project.

The directed evolution process can begin after 2 or 3 recultures. Using the top 24 cultures from the 5 plates, take 100 µL of each sample, and miniprep the samples. EP-PCR is done on the samples, using different concentrations of salt depending on the intended amount of variation. The samples are then cultured and recultured in the same method as was done for AE. With directed evolution, more variations are created in each cycle, increasing variation. This in turn increases the chances of a mutation occurring that would suit the purpose of this project. However, this excessive variation can become uncontrollable. To counter this, heavily controlled EP-PCR is needed.

After the mutants are isolated, enzymes will be produced by the bacteria, which will be immobilized within encapsulated calcium alginate beads to protect enzyme activity throughout the bioreactor process. The enzymes are dissolved in a sodium alginate solution to a concentration of 2 mg/mL concentration, then dropped into a calcium chloride solution via a syringe to form droplets, which are obtained after 30 minutes of incubation through filtration.

The design of the plasmid used a His-tagged NdmAB-GFP Plasmid. In the first trial, while building this plasmid, the His-tagged NdmAB inserts were unavailable, so they were substituted with NdmAB inserts. While waiting for the His-tagged parts to come in, we moved forward with creating a vector with NdmA and NdmB+GFP via Gibson Assembly to validate the Gibson Assembly procedure. The confirmation of the first trial was done using gel electrophoresis, which displayed that the assembly was not successful. A potential reason for this was due to the NdmA and NdmB parts we ordered not having the sticky ends needed at the time of assembly. The design remained the same, but we PCR amplified the parts to add the sticky ends and performed DNA cleanup on NdmA and NdmB+GFP prior to use. After testing the results of the DNA cleanup using the nanodrop, it was noted that the concentrations of the NdmA and NdmB+GFP were low: 19.9 ng/µl and 14.7 ng/µl respectively.

Through troubleshooting, it came to our attention that the DNA backbone used in our first trial was not an empty plasmid and therefore did not match the designed primers. We switched to a new backbone for trials using the traditional pSB1C3 from previous years. To prepare the vector, we performed a T4 DNA Ligase Protocol, without adding any inserts, to confirm a circular plasmid. The plasmid was then linearized through an EcoRI-HF digest protocol to get it ready for PCR amplification. Since the pSB1C3 had already been cut in a previous year, the backbone ligation and digest were needed to get a backbone DNA cut at the desired site. To test, DNA cleanup and gel electrophoresis were used and ultimately did not work. Upon reflection, we realized that the T4 DNA Ligase was not added last, as specified by the protocol. The ligation was repeated with the steps in the correct order. EcoRI cutting was also repeated, but with the materials being diluted to scale down by a factor of 20 to both conserve materials and better match the ratios of NEBioLabs protocol.

While performing DNA cleanup in this trial, the NdmA was diluted to 80 µL total so each µL had 10 mM instead of 100 mM. After using the nanodrop machine to test the DNA cleanup and PCR, we noted an over 10-fold higher concentration: 161.2ng/µL for NdmA, 104.7ng/µL for NdmB+GFP, and 140.6ng/µL for the backbone pSB1C3. We ran gel electrophoresis to confirm the DNA parts. Wells B and C roughly match up with the expected length of the respective sequences, and there are no other bands, suggesting that we obtained a pure sample of both target sequences NdmA and NdmB+GFP. Since we were able to obtain great samples of our DNA parts, we proceeded to move onto combining the parts into the pSB1C3 vector backbone.

Wells B and C roughly match up with the expected length of the respective sequences, and there are no other bands, suggesting that we obtained a pure sample of both target sequences NdmA and NdmB+GFP. Since we were able to obtain great samples of our DNA parts, we proceeded to move onto combining the parts into the pSB1C3 vector backbone.

To continue building, we did a Gibson Assembly, followed by a transformation. To test the success of this, we observed both a negative control of just BL21 E. coli cells and NdmAB Full Plasmid in the same strain plates. Neither of the plates showed any growth. We concluded the lack of growth was due to the Gibson Assembly failing as the negative control did not grow colonies as expected. A positive control of competent cells transforming a known plasmid would have confirmed that the BL21 cells were working and that the transformation protocol is adequate. However, we avoided the positive control as the gel electrophoresis revealed gel bands in the wrong place, suggesting that the Gibson Assembly failed. Through more troubleshooting, we repeated the Gibson assembly with adjusted times for annealing. In one assembly, we ran for 45 minutes, and in the other assembly, we ran for an hour. We then ran both products of the assembly in a gel electrophoresis. Both products ran the same in the gel, with separate NdmA and NdmB, but no backbone vector visible.

The first part of the project focused on getting NdmA and NdmB. However, the importance of NdmD in the pathway was overlooked as it is not directly involved in the demethylation pathway. By doing further literature review, it was revealed that NdmD plays an important role in proton transfer that would allow NdmA and NdmB to work more effectively.

Once we received the his-tagged NdmA, NdmB, and NdmD parts, we ran a Gibson Assembly on these constructs, this time using a newly ordered redesigned pSB1C3+RFP backbone in a E. coli agar stab. The reason behind this was to increase the cohesion between pSB1C3 and the DNA parts in order to raise the efficiency of the Gibson Assembly. The culture was plated on LB with a 1000:1 ratio of chloramphenicol, which helped confirm the presence of RFP.

Two cell cultures (pSB1C3 A and pSB1C3 B) were inoculated into a liquid culture before the cells were lysed and cleaned. PCR was performed on the lysate and the newly obtained 6x his NdmA/B/D parts to obtain a good purity of the backbone for the Gibson Assembly. The concentrations and purity of the products are listed below.

Figure 3: Concentrations of all DNA parts

The PCR amplification seems to have gone well for all the DNA parts. We then proceeded with a Gibson Assembly attempting to insert the 6x His-NdmA/B/D into the pSB1C3+RFP backbone. We performed three Gibson Assemblies, one using pSB1C3 A and the other using pSB1C3 B. The last one was a negative control with just the backbone and no additional parts. We performed gel electrophoresis to confirm success shown below.

Oddly enough, only NdmD showed a visible band revealing there may be an issue with the primer design or the PCR protocol. The issues of NdmA/B and the pSB1C3 backbone not showing up were reflected downstream in the Gibson trials in wells 7 and 8 not showing up either. This was noted for future troubleshooting of the Gibson Assembly protocol.

After continuous difficulty utilizing the backbone, the focus shifted to assisting the product development subteam with enzyme encapsulation. The new backbone including mRFP1 was used to e test our sodium alginate beads. The red fluorescence proteins were transformed via liquid culture. 20 falcon tubes containing 15mL of liquid culture with 1:1000 ratio of chloramphenicol (15 µL) were centrifuged at 1200 rpm for five minutes using a Beckman Coulter Allegra X-22R. A couple of the tubes were re-centrifuged for an extra 3 minutes in order to form a visible pellet. Out of the 20 tubes used for the liquid culture, 19 were able to develop large E. coli pellets visibly showing a pink tinge to confirm expression of RFP. The one liquid culture that was unable to grow pellets was kept in the same tube to incubate longer. For the rest of the tubes, the old LB was thrown out and replaced with a new LB. After two days of incubation, there was sufficient growth of the pellets in all 20 tubes. A nanodrop was performed to ensure the concentration of the pellets was sufficient. After a high concentration was confirmed, glycerol stock was used to store the bacteria. This was done by first spinning down the cells and removing the excess LB. 1 mL of LB was then used to resuspend the liquid culture followed by adding 1 mL of 50% glycerol. This was stored in the -80℃ freezer for future uses. The liquid cultures of pSCB1C3-mRPF1 are shown below.

During this process, it was also discovered that a plasmid backbone construct had been delivered from GenScript a few weeks prior. As a result, attempts to create our own plasmid were momentarily paused. Instead, we shifted focus onto using the plasmid already containing the NdmA/B/D from GenScript for transformations.

With the GenScript raffle, Team Cornell was able to order our first iteration of pENGM42, the finalized plasmid design for our team. While the final design includes mRFP, the first iteration includes GFP but both fluorescence proteins serve the same qualitative purpose of viewing expression of NdmA/B/D. To confirm the plasmid received was pENGM42, Team Cornell did a transformation and gel electrophoresis in parallel. The transformation confirmed chloramphenicol resistance within the plasmid while the gel confirmed the plasmid was the proper size. The transformation was done in quadruplicate. A picture of the gel is shown below where well 1 is pENGM42 while well 8 is the ladder. Other parts were also tested within this gel as told below.

Figure 7: Wells of the gel electrophoresis of Figure 6

While there is a smear, the general size of pENGM42 correlates to what was expected. The same issue of NdmB and NdmA not appearing in the gel can be seen and well 4 shows multiple parts but were not combined into the desired full plasmid. A couple issues could have been the voltage (ran at 135 V) was too high causing smears and the same Gibson Assembly issues as before. To retest pENGM42, we performed liquid 12 cultures (3 from each plate) of the successful transformation of pENGM42 to qualitatively see if GFP conjugated with NdmB would properly express. Successfully, bright green from each plate was shown from the cell cultures when placed under UV light as shown below. If GFP was able to keep its function, it can be assumed that the conjugation would not affect the function of NdmB, the protein of interest, as well.

These cultures were then mini prepped to collect pENGM42 and had a DNA gel performed once again with a well per trial. After 50 mins at a lower voltage (120 V), the gel confirmed each plate successfully expressed the proper plasmid and solidified that the design received from GenScript was the proper design from Team Cornell.

With wiki freeze approaching, our web lab became even more determined to obtain the enzymes of interest (NdmA/B/D) to be encapsulated for the next phase of testing enzyme encapsulation and the reactor. As time closes in, each plate’s liquid cultures were combined and spun down to prepare them for lysing. To save materials, all the cells were only suspended in 5 mL of LB per plate and had the optical density of them taken in order to properly add the correct amount of lysis buffer. The OD600 values are shown below.

Figure 10: OD600 values of pENGM42 liquid cultures combined

Following the lysis protocol for NEBio Labs, the lysates of each plate were created and the proteins were spun down for storage at 4C as shown below.

Our next steps after wiki freeze will be to follow our designed protocol for Ni-NTA Affinity Chromatography utilizing the 6x his tags of NdmA/B/and D to isolate them from the rest of the proteins expressed by our bacteria. These will then be ready for encapsulation by the product development subteam where they will interact with caffeine flowthrough to convert it into 7-MX. In parallel, epPCR will begin using our circular plasmid method to generate mutations in NdmB and begin the adaptive/ directed evolution (ADE cycle). Once those are prepared, they will follow the same process of enzyme isolation to be encapsulated and run through the reactor by product development for 7-MX generation. More information will be provided during the 2023 iGEM Grand Jamboree presentation of Team Cornell of how far our next steps take us!

At its very core, this idea consists of an enzyme and its reactants. We want to flow a solution of caffeine, the caffeine breakdown enzyme, or both of these molecules through our bioreactor such that the product (or a solution that consists primarily of product) flows out.

The goal of this design process was to determine the superior way to allow the enzymes and reactants to interact inside our bioreactor. The frequency and speed of reactions, as well as the cost and ease of use of materials were all considered while determining the structure of the bioreactor.

We know that enzyme immobilization while we flow reactants into and products out of the bioreactor is the most conceptually and mathematically simple approach to this bioreactor [4]. Immobilizing enzymes while reactants flow through at a regular rate allows us to calculate the residence time that reactants spend in the reactor interacting with an enzyme, the amount of reactants that each enzyme encounters on average, and more.

Several methods of enzyme immobilization were considered during the design process. Utilizing magnetic nanoparticles, hydrogel bead encapsulation, and entrapment in a porous hydrogen-bonded organic framework were all researched. We spoke to an expert in reaction mechanisms, Dr. Phillip Milner, an assistant professor in the department of Chemistry and Chemical Biology at Cornell University, to help us determine which encapsulation method to use. Dr. Milner emphasized that more complex methods of enzyme immobilization are only ever used as a last resort in most products. If a cheaper, simpler way of immobilizing enzymes is possible, that is the method we should pursue. With this in mind, we decided to move forward with hydrogel bead encapsulation.

More specifically, we know that sodium alginate, when dropped into calcium chloride, will form a porous bead that allows reactants to flow in, interact with the encapsulated enzyme, and flow back out [4]. Calcium chloride and sodium alginate are both relatively inexpensive materials, and relatively stock materials in most laboratories. As a result, we concluded that sodium alginate beads would be the best method to encapsulate enzymes.

Before the sodium alginate beads can be formed, we have to make the calcium chloride and sodium alginate solutions.

The calcium chloride solution is made by dissolving a set amount of calcium chloride powder in water depending on the desired concentration of the solution. For the purpose of testing, we created five different calcium chloride solutions ranging in even increments of 0.5 M from 0.5M to 2.5 M.

Sodium alginate solution is also created by dissolving a set amount of sodium alginate powder in water. Once again, the amount of powder used depends on the desired concentration of the solution. For testing purposes, we created five different sodium alginate solutions ranging in even increments of 0.5% from 1.5% weight by volume to 3.5% weight by volume.

When dissolved in water, sodium alginate has a tendency to form lumps that are very difficult to break apart and add time onto the already long dissolving process. In initial testing procedures, sodium alginate was added entirely to the distilled water solution, resulting in the formation of large, undissolved particles. Further research found that despite the hydrophilic nature of alginate salts, the high hydratability led to the formation of lumps and non-uniform dissolution of the powder [1]. Thus in later batches, the protocol was altered and sodium alginate power added very slowly, allowing the powder already in the solution to dissolve almost entirely before adding any more powder. This increased production efficiency by quartering the amount of time required to make the solution.

Before conducting tests about bead formation, batches of sodium alginate and calcium chloride solutions at several different concentrations needed to be made. After being fully dissolved, the calcium chloride solutions exhibited almost no distinct properties. In fact, there were no physical properties that made it possible to distinguish between the different concentrations.

Unlike calcium chloride, the sodium alginate solutions began exhibiting different properties based on concentration immediately. Solutions with a higher weight by volume concentration of sodium alginate demonstrated increased viscosity and more air bubbles, most likely due to the increased amount of time it takes for said bubbles to travel through the solution. Furthermore, the more viscous sodium alginate solutions - more specifically those at a 2.5% weight by volume concentration or higher - were much more difficult to handle. Frequently, the solution would adhere to the sides of pipettes and syringes, and not all of the solution would drain into the desired receptacles.

Originally, a serological pipette was used to transfer the sodium alginate solution between containers and to add the sodium alginate solution dropwise into the calcium chloride solution for the formation of encapsulated beads. However, when testing solutions above a 2.5% weight by volume concentration, the viscosity of the solutions made it almost impossible to allow the sodium alginate to drip out of the pipette one bead at a time. The solutions exhibited an extreme surface tension that proceeded to only allow large globs to be pushed out of the pipette. Even with the minimal extrusion of droplets, the pipette would remain lined with excess solution that would cling to the sides and cause waste, resulting in lower production of beads.

Upon discovering these difficulties, the decision was made to switch from a serological pipette to a twenty milliliter syringe with a four inch needle. The twenty milliliter syringe could hold just as much solution as the twenty milliliter serological pipette, and the smaller circumference of the output allows us to create a smaller, more uniform bead size. Dr. Phillip Milner also told us that smaller bead size will increase reaction frequency, since there will be more beads at a higher concentration in the reactor, and thus reactants will encounter the beads more often. Thus, the production of smaller, more uniform bead size is desirable.

The sodium alginate solution is added dropwise to the calcium chloride solution by pressing slowly down on the syringe until the bead forming at the tip of the needle gathers enough mass to break away and fall into the calcium chloride. Initially, this was all done over a stir plate. The stir bar in the calcium chloride solution was kept even after it had been fully dissolved, because initial testing demonstrated that the calcium chloride required additional time from the initial addition to fully dissolve, and we wanted to ensure that the calcium chloride stayed dissolved, and that the solution was entirely uniform.

However, upon making a batch of beads in this solution, we noticed that the stir bar was negatively affecting the shape of the beads. The sodium alginate beads would enter the solution almost perfectly spherical, but would later be knocked around by the bar, pinned between the bar and the container, and, generally, deformed. As a result of these observations, the protocol was adjusted with the stir bar removed to create the rest of the beads in a calcium chloride solution at rest.

When creating different concentrations of sodium alginate solutions, we added different amounts of coomassie blue dye. When the dye and alginate was fully dissolved, it became clear that the 3.0% weight by volume sodium alginate contained enough dye to make the beads visible in clear solution, but little enough dye to allow us to see through the solution, and thus into the beads. As a result, the varying calcium chloride concentrations were all tested with 3.0% weight by volume sodium alginate. In addition to taking qualitative observations of the beads formed in each calcium chloride solution, we measured the average bead diameter of each batch. These findings are listed below:

Figure 12: A table displaying the average alginate bead diameter at varying calcium chloride concentrations.

As seen above in Figure 12, the concentration of the calcium chloride solution had no noticeable effect on the average diameter of each batch of beads. Furthermore, there were no significant qualitative differences between concentrations of sodium alginate.

Beads formed using different sodium alginate concentrations yielded much more interesting results.

Figure 13: A table displaying the average alginate bead diameter at varying sodium alginate concentrations.

As can be seen above in Figure 13, sodium alginate concentration actually had an effect on the average diameter of the beads in different batches. The greater the sodium alginate concentration, the greater the diameter of the bead. One noticeable outlier in this trend is the 3.5% weight by volume sodium alginate concentration, which actually yielded a lower average bead diameter than the 3.0% weight by volume concentration. This can likely be attributed to experimental error, including the transition from use of a serological pipette to a syringe and needle.

There were also significant quantitative differences between each batch of beads. A demonstration of some of these qualitative differences is pictured below.

The top row contains beads split with a sharp tool, the second contains untouched beads, and the lowest row shows beads upon which a uniform pressure was applied. One important thing to notice is that beads made with 1.5% and 2% weight by volume sodium alginate concentration split under this pressure, whereas the greater concentrations could only be split with the sharp tool. Below is another photo that demonstrates what these beads looked like before experimentation.

When the sodium alginate concentration exceeded 2.5% weight by volume, the beads demonstrated an incredibly heightened amount of “tailing”, the property we used to describe the issue where beads form a teardrop shape rather than a perfect sphere.

No physical or chemical differences were noticed between the beads formed at different calcium chloride concentrations in these experiments. As a result, future production of the beads will use a standard 1 M concentration of calcium chloride.

The benefits of each concentration of sodium alginate were much more difficult to weigh. Beads of a higher concentration are more durable and split less easily, but also are larger and the solution itself is more difficult to work with. We concluded that the best concentration to utilize in further bead making is 2.5% weight by volume. This is because, for present purposes, it is the lowest concentration that will not split under pressure. As a result, it is the smallest bead at the concentration that is easiest to work with out of the most durable beads.

With an established uniform procedure for making beads, optimization and speed of the process was the next goal. An estimated fifty to one hundred beads would be required for each bioreactor test session, and our current bead-making procedures were very slow. In order to optimize our use of crucial lab time, we needed to create an automated bead-making machine.

Several designs were considered during the first drafts of this machine. We initially considered using a syringe attached to the bottom of a container of the sodium alginate solution, and allowing gravity to drip the sodium alginate out of the solution. However, the viscosity of the sodium alginate solution proved to be much too high to allow gravity alone to push the solution out of the syringe in a timely manner. As a result, reducing the amount of solution to only that in the syringe and then adding pressure in the form of weight on top of the syringe plunger. However, there were still concerns that this would not produce beads reliably in a short amount of time.

Finally, the idea to use a syringe infusion pump was suggested. Utilizing one of the syringe infusion pumps in our lab, except instead of connecting the syringe to a larger system, the sodium alginate would be allowed to drip out of the needle, similar to the process of making the beads by hand.

Constructing this system was quite simple; the same syringe that was utilized when making the beads by hand would be put into the syringe infusion pump, and a container of 10 mL of calcium chloride would be put underneath the tip of the needle attached to the syringe so that the sodium alginate will easily drip directly into the solution. A video of the finalized setup is shown below.

Testing the success of this automated bead maker was also relatively simple. To determine whether or not it was effective in increasing our rate of bead making, the approximate rate at which beads are produced, both by hand and by the automated bead maker, needed to be compared.

By hand, approximately forty beads could be made in two minutes. That resulted in bead-making by hand exhibiting a rate of twenty beads per minute.

Alternatively, the automated bead maker was observed to produce about seven beads in ten seconds, putting it at a bead-making rate of approximately seventy beads per minute.

Additionally, by simple qualitative analysis, it was observed that the beads made by the automatic bead maker were more consistently sized, had less tailing, and had less air bubbles. Consistent sizing and lack of tailing allows the beads to be loaded into the reactor more easily, and lack of air bubbles ensures that reactions between the reactants and enzymes will not be disturbed by gaseous molecules in the way.

It is evident that the automated bead maker is more successful in efficiently making beads. The uniformity and speed with which these beads are made allow us to use time more efficiently in the lab, and will speed up processing by tenfold when scaleup occurs. As a result, we will move forward with making beads with the automated bead maker only.

Our goal was to design and create a fixed-bed bioreactor in which our selected enzyme would be immobilized in alginate beads and those beads then immobilized within the bioreactor [2][3]. Going into this project, there were a few properties we wanted our design to have: modularity, ease of scaling, and cost effectiveness. These were all necessary because we were seeking to replace an existing process with our project: if our project was more costly or difficult to scale or maintain, what improvement would we truly be making?

The first step in our design was determining the structure we would use to immobilize the alginate beads. With an assumed bead diameter of a millimeter, each team member created a prototype CAD model to present to the rest of the product development team. We determined several desirable properties in these alginate bead immobilization systems; for maximum efficiency, our reactor required maximum exposed bead surface area, individual bead separation, and relative ease of bead replacement.

Several rounds of this brainstorming process were completed to fully flesh out the various strengths and weaknesses of each design including covalent adhesion, adsorption, entrapment, as well as dynamic bead swelling. Ultimately, an adsorption method was selected where encapsulated enzyme beads would be pushed against a mesh-like spiral matrix to produce separation of beads and allow fluid flow, while keeping beads immobilized in a relative space of the reactor. The mesh-like properties of this immobilization system allowed for maximum exposed bead surface area, and the attachment of immobilization matrix to central supports theoretically provided easy removal of the entire mesh system should beads need to be replaced or examined due to its simple design of simply sliding into the reactor column. The CAD model is pictured below in Figure 5.

In parallel, we also came up with a rough idea of what we wanted the bioreactor assembly to look like. To model our tube, we researched various materials and landed on PVC piping for its affordability, lack of reactivity, and ease of use. The relatively large size of our piping (1½ inches in diameter) allowed us to test more parameters as well.

From this, we came up with a mock-up of the bioreactor using easily-acquired materials to determine the necessary parameters for optimal synthesis. With the preliminary immobilization CAD models completed, the prototype immobilization methods were 3D printed using PLA on a Prusa i3 MKS3+ printer. PVC pipe was used to model the reactor chamber, and two mason jars were placed at each end of the pipe to store water. A peristaltic fluid motor pump was fixed to the jar that contained water, and was programmed to flow water through the mock reactor at a constant rate. Plastic beads were used to simulate free-flowing alginate beads in the reactor for initial testing for ease-of-testing. Later, encapsulated beads were exchanged for the plastic beads, proving modularity of the system. Burlap cloth was fixed to either end of the reactor to hold the beads and immobilization matrix inside.

The mockup reactor was tested by measuring the amount of time it took for this simple motor setup to move a known volume of water through the tube. By altering various elements of the setup, such as the incline of the tube (or lack thereof), the power being delivered to the motor, and the amount of “beads” in the tube, we gained valuable insight into the way that our final product would have to operate. An example of these trials is shown below (Figure 18).

Figure 18: Measurements taken to qualify residence time of various reactor tube sizing

After testing, we moved forward with creating a more finalized version of the mock reactor. Our new design consists of a similar immobilization design scaled to our updated testing column dimensions. This design was 3D printed and put inside of a 10mL syringe with a stopper made from the syringe plunger connected to a tube.

[1] How to dissolve Alginate. KIMICA Corporation. (n.d.). https://www.kimica-algin.com/alginate/usage/

[2] Making gel beads - science & math investigative learning experiences. (n.d.). https://smile.oregonstate.edu/sites/smile.oregonstate.edu/files/gel_beads_1.pdf

[3] The ultimate guide to sodium alginate beads - gino gums. Gino Gums & Stabilizers. (2023, April 20). https://gumstabilizer.com/guide-to-sodium-alginate-beads/

[4] Urrea, D. A. M., Gimenez, A. V. F., Rodriguez, Y. E., & Contreras, E. M. (2021, October 20). Immobilization of horseradish peroxidase in CA-alginate beads: Evaluation of the enzyme leakage on the overall removal of an azo-dye and mathematical modeling. ScienceDirect. https://www.sciencedirect.com/science/article/pii/S0957582021005279?via%3Dihub