Executive summary

Overview 

Mesoporous engineered carbon supports (PCS) can be used in various applications including energy storage, adsorption, and as active metal supports for widely used catalysts in hydrogen fuel cells. Due to the increase in emissions from greenhouse gases, demand for PCS products is expected to grow from $178.9 million in 2022 to $6.5 billion by 2031. This demand can be attributed to the exceptional properties of PCS such as high surface area, stability, conductivity, tunable selectivity, and enhanced porosity enabling various emission reduction technologies. Pajarito Powders LLC, a New Mexico-based research and production start-up, manufactures PCS for hydrogen fuel cells and electrolyzers, at a lab scale of 1 kg per batch. To meet growing demand, they aim to increase their production to 20 kg/day. To aid in this effort, we have developed a process that targets this production rate and addresses the current system gaps such as reliance on batch processing, manual material handling, lack of process controls and quality checks, absence of unit operation scheduling, and environmental and safety hazards. 

Project Description 

Our project consists of scaling up Pajarito Powder LLC’s PCS manufacturing process in an existing plant located in northeast Albuquerque. To accomplish this goal, our team performed research and various calculations to determine the appropriate unit operations with optimized dimensions. Our project was constrained by the lack of experimental data and tests, plant location, material properties, and product specifications. In the end, we managed to achieve the 20kg/day production goal while also improving process safety and efficiency. 

Our process can be broken down into two main steps: embedding silica, a hard template, into a carbon structure to create pores, followed by the removal of silica to expose the resultant mesoporous properties of the carbon structure. The raw materials consist of colloidal silica, our carbon precursor chitosan, acetic acid, and water. The raw materials are mixed, dried, and then grinded using a ball mill to activate the materials through mechanochemical reactions. This grinded material then undergoes pyrolysis, which carbonizes the chitosan around the silica template. Silica is removed from the pores using hydrofluoric acid and nitric acid, resulting in highly porous carbon particles. This material is then filtered and further pyrolyzed, increasing surface area and solidifying pore structure, giving us our final product. 

Customer Profile and Market Analysis 

Because Pajarito Powder current production is limited to 1 kg/month their PCS is primarily sold for lab testing and validation. Their customers are particularly interested in developing hydrogen-powered commercial and passenger vehicles. These include Hyundai, General Motors, and various Department of Energy National (DOE) Labs across the US. This is in support of the DOE’s goal to commercialize hydrogen-powered technologies for heavy-duty vehicle applications. 

With our increase in production, we plan on breaking into a wide range of markets outside hydrogen fuel cell catalysts. With the PCS market expanding to $6.5 billion, we plan on selling our product for use in wastewater treatment and purification, catalysis supports in various chemical industries, energy storage, and carbon capture technologies to name a few. This wide range of potential markets ensures there is a sustained high demand for our products. 

Economic Results  

Pajarito Powder set an ambitious goal for this project: slash product price by 90% while remaining profitable. Achieving this goal means our PCS would be on the market at an impressive $4.7 per kg, projecting an annual revenue of $28.7 M. Our initial fixed capital investment is a modest $1.9 M, covering equipment and building enhancements thanks to our existing infrastructure. With manufacturing costs estimated at roughly $10 M, profitability is expected within the first four months of the plant operation. Our net present value, which normalizes all cashflows to the present day, is $76.2 M, resulting in an outstanding discounted cash flow rate of return is 246%, and an internal rate of return of 735%. Even when factoring in raw material price fluctuations and drops in PCS selling prices, our process remains resilient, offering a near guarantee of a sustained profit.  

Status of the Project 

Phase 1: Preliminary Process and Market Research – Completed. 

Phase 2: Process Design and Optimization – Completed. 

Phase 3: Process Safety – Completed.  

Phase 4: Environmental Safety and Compliance – Completed.  

Phase 5: Design Report and Company Presentation – Completed 

Meet the Team 

Our team consists of six Chemical Engineers: Luke Lucero, Raphael Reyes, Marelessis Palomino, Patrick Survis, Tin Nguyen, and Dhruv Patel. Together, we have experience ranging from material science R&D to large-scale manufacturing, which was all utilized for this project. Process research was led by Marelessis and Raphael. Process design and safety was led by Dhruv, Luke, and Tin, while Marelessis oversaw environmental safety and compliance. Market/Economic Analysis was led by Patrick.

Hydrogen fuel is a growing topic of interest due to its high energy content per unit weight and potential to support a sustainable carbon cycle by providing a clean source of energy. Hydrogen technology is critical in achieving net zero emissions which the global energy sector aims to meet by the year 2050. Various strategies are being implemented to reach this goal such as fuel shifts to nuclear, hydrothermal, or solar energy and 6% is attributed to hydrogen1. Despite the growing interest, there are major pushbacks due to the cost and need for consistent efficiency of the hydrogen fuel cell. Modifying current carbon supports for catalysis can lead to more efficient electrolysis of hydrogen. This can be achieved through mesoporous characteristics of 2nm-50nm, and large surface area which increases the rate of reaction. Additionally, carbon support is growing a market of its own known as Porous Carbon Support (PCS) which has various applications such as energy storage, carbon capture, and air purification. To meet the demand for hydrogen fuel cells and the growing new market of PCS, efforts towards scaling need to be implemented to generate components for hydrogen technology in large quantities. Pajarito Powder, an Albuquerque-based startup, is addressing this goal by developing PCS for hydrolysis and porous carbon supports that can be sold individually to markets. Their current goal is to scale production processes and relieve a portion of the manufacturing costs to make their processes more profitable and the product cheaper.   

Our product is classified as a premium product and will be sold directly and shipped to commercial vehicle manufacturers.  Our primary market will be semi-trucks, buses, and heavy goods vehicles due to their large contribution to carbon emissions.  The pricing of the PCS sold individually is currently on average $47/g at the kilogram scale which we aim to reduce by 90%. The price of the PCS varies with the amount (ordering larger quantities lowers cost) and is often sold in bulk due to economic constraints. Our optimization aims to reach a profit margin between 50% based on a range of manufacturing and material costs. Current PCS production sits at a Technology Readiness Level (TRL) of 4 in which the technology is only validated in a lab environment. By increasing production, we aim to reach a TRL of 8, signifying that the technology is complete and fully qualified for a commercial environment. We aim to increase Pajarito powders PCS production from 30g/day to 20kg/day. Through this scale up,  we anticipate challenges such as    

maintaining desirable properties (surface area, pore size, particle size, etc.) of the porous carbon support (PCS). An effective scale-up will require the implementation of i) automated machinery such as conveyor belts and gravity-feeding to reduce hands-on labor ii) optimization or substitution of unit operations iii) control systems and quality checks to maintain those favorable operating conditions, iv) substitution of hazardous chemicals such as hydrofluoric acid and parallel processing. These strategies will reduce energy consumption and negative environmental impacts. In addition to high surface areas, large pore volumes, and the low cost of porous carbon supports PCS also offer low mass transport resistance, high dry H+ conductivity, and reduced levels of ionomer poisoning2,3. The scale-up will take place in Pajarito Powder’s new facility located in Northeast Albuquerque. This facility will have 4000 ft2 of manufacturing space, which occupies two stories of vertical space. This more than doubles the existing space, not including the additional vertical space added. This is desired because it allows for an increase in unit operations sizing, availability to store and hold multiple batches, and containment of hazardous chemicals. A drawback to this location is it is located within a close radius of multiple communities, and educational buildings which can cause safety concerns.   

Design Constraints  

Technical:  

• Each unit operation in the process should be investigated for transition to a semi-batch/continuous process to increase production rate by 20 times the current rate, while maintaining catalyst quality. 

• Current furnace/pyrolysis stages are where many of the desirable surface characteristics form and are done so under reduced conditions, making automation of this specific operation difficult.  

• Bottleneck operations pose a challenge, and strategies for minimizing batch time from 2-3 weeks down to 1 day, including the consideration of parallel processing, should be explored while maintaining catalyst quality4. 

• Minimization of manual labor should be prioritized, with a goal of reducing manual input down from every unit operation to 1/3 of the operations.  

• The PCS specifications, including sizes in the range of 0.4 to 1 micron, BET surface area of 400 – 1000 m^2/g, pore size of 5 – 50 nm, and the mesoporous quality must be maintained. Simply scaling up the current process will lead to decreased control over size and shape, which leads to a less stable product2. Intricate design considerations are required for a successful scaled-up process. 

• The pursuit of a one-pot production method would simplify the overall process, making a scale-up more feasible. But it would demand a complete redesign of the synthetic process to obtain the desired surface characteristics and poses challenges in achieving the desired surface characteristics and maintaining control over properties2. The current Ex Situ method must be preserved. 

• Quality Checks are required to ensure compliance with specifications, involving time-consuming techniques like BET and XRD. Each Q.C takes up to a day. Strategic placement of quality control checks is needed to prevent excessive time delays. 

• Implementation of control systems will be required to maintain optimal operating conditions, specifically in the furnace where a consistent temperature is needed to maintain PCS quality parameters. As well as a constant reductive atmosphere in the furnace. Other time delay controls can ensure controlled mixing, drying, etching, and filtering.  

• The new PCS production area offers 2600 ft2 total of manufacturing space across three bays with vertical space. Efficient utilization of vertical space, along with incorporating gravimetric feeders where applicable, is needed to increase throughput. 

• The source of utilities is restricted to NM utilities. The Albuquerque water system has a certain capacity for company use of water and other resources5. 

Economic:  

• Porous Carbon Supports (PCS) are made of carbon with a nitrogen dopant, making the raw materials cheap and sustainable.  

• A price reduction of 80-90% is needed for our PCS to be economically viable on a large scale. 

• Fossil fuel-based technology currently is still cheaper than battery/electric-based technology, making it a challenge for the growth of fuel cell market. 

• 20 kg of PCS must be produced per day to be economically viable.  

Environmental and Safety:   

• Hydrogen fluoride (HF) reactant, which is an extremely corrosive and toxic chemical and is also difficult to dispose. HF has an EPA waste code of U134 and is disposed of by using absorbent materials6. In the current process, HF cannot be replaced, but automating the addition of HF to the etching stage will be considered to remove the safety risk associated with handling HF. 

• Recycling of the entire PCS is not economically viable once a defect or issue is found the support cannot be repurposed7. This is due to ionomer poisoning and contamination in the processing techniques.  

• The PCS is a fine powder when mishandled can create a dust hazard requiring proper containment and ventilation.  

• Workers must be properly trained to operate and briefed on the hazards. 

• Safe working conditions must be maintained. 

• The surrounding community must be aware of the potential hazards. 

• Emergency systems must be in place to mitigate risk and danger. 

Ethical and Societal:  

• Dangerous chemicals used in the process have been known to kill or injure the people handling them. The handling of HF has many standards and safety practices, ensuring the wellbeing of the workers and the environment8. Due to this we will adhere to state and federal regulations. 

• Ensure proper training and safety for all employees for hazardous chemical handling through OSHA regulations9. 

• Ensure safe and fair work environments for employees10. 

• Make surrounding communities aware of all chemical hazards through the Emergency Planning and Community Right-to-Know Act (EPCRA) of 198610. 

• Ensure proper disposal and treatment of any hazardous chemical waste stored on site. 

• Comply with OSHA hierarchy of controls to ensure an overall safety process11. 

• Waste management and pollution control systems must be in place. 

Schedule: 

• The DOE project spans a 4-year period. Stage one, subtask 1 is to be completed by May 2024. This includes benchmarking the existing PCS manufacturing process to establish a baseline. 

• Pajarito Powders has the ambitious goal of scaling up their process 30x by October 2024.  

Project Deliverables  

D1 Preliminary PCS Process Benchmark: 

• Problem: Pajarito Powder’s production of the PCS is too low to be commercially viable. The goal is to scale up the process from ~1kg/month to 20 kg/day. Scaling up this process will be extremely difficult, and process optimization and scheduling is needed. 

Solution: Perform a preliminary process benchmark to determine current process inputs such as raw materials, water, energy, and fossil fuel usage using experimental plant data. This will establish a baseline which will allow us to identify areas where the process can be optimized. This benchmark will be used by future groups who will expand on these results.  

• Safety analysis:  

• The process involves working with nano-scale particles, which means employees and workers must be properly trained in how to safely work with nanoparticles. 

• The current manufacturing process involves HF, which is highly corrosive and dangerous. Both the workplace and the employees must be properly trained and uphold safety standards when working with HF. 

• Environmental assessments:  

• There are concerns regarding the environmental effects the plant can have on the local area. The community must be made aware of these concerns.   

• The current process is also very energy intensive. Our goal is to reduce the energy requirements during manufacturing by at least 25%.   

• Recycling PCS is currently not a viable option due to ionomer poisoning. 

• Determine commercial viability of process – have we reached a technology readiness level of 8 

D2 PCS Process Optimization: 

• Identify potential for energy and water usage optimization: 

• Ball milling unit operations can be made into semi-continuous operations with little to no downside. Considering the energy consumption and cost will be affected upon switching from batch and semi-continuous.  

• Batch furnaces can be made into semi-continuous tube furnaces, the advantages of this are increased time of reaction, and increased production rate; the disadvantages of using a semi-continuous tube furnace are higher cost, higher frequency of maintenance, and higher energy cost.  

• Etching will require a decantation stage which can be made semi-continuous based on process scheduling.   

• Vertical design: Gravity based feed system. 

• DI water lines to reduce manual addition.  

• Determine areas of difficulty for scale up: 

• How does scale up affect quality? – implement controls to mitigate loss in quality. 

• What unit operations can be modified or added to maintain product quality? 

• How will acid be handled and stored to reduce hazardous conditions.  

• How much raw product is needed for a production rate of 20 kg/day? 

D4 Product Market Analysis:  

• Identification of market outlook: 

• Main target market will be the commercial truck industry, in partnership with Million Mile Fuel Cell Truck. Kenworth’s trucking market is planning to expand current electric trucks models T640E–T680E12,13,14. 

• Expected buyers: 

• China will be our largest market, as they recently upgraded their fuel cell policy as well as unveiled plans for fuel cell technology deployment within the country. China also has the world’s largest automotive market that highly supports fuel cell vehicles. It is predicted that by 2025, there will be around 50,000 hydrogen fuel cell vehicles on road in China15. 

• Korea will be the second largest market as they recently agreed to reinforce the maintenance and operation of fuel cells. Korea is also the fastest-growing country within Asia15. 

• The Asia Pacific and North America regions will be our tertiary focus as both regions are anticipated to have the fastest-growing compound annual growth rate in fuel cell investment16,17,18. 

• Bekaert which is a steel company requires Pajarito’s current PCS for subcomponent for electrolyzes19. 

• Identification of Additional Markets: 

• Other potential markets besides automobile are oral drug delivery systems, contaminated water treatment, separation of large biomolecules, electrical double-layer capacitors, wastewater treatment, carbon capture, catalysis, and air purification20. 

• Sell carbon support directly to customers. Potential use outside of fuel cell catalysis: High surface area silicas, carbon blacks, activated carbons.  

D5 Product Outlook Analysis: 

• Determine economic viability PCS Production: 

• The estimated general fuel cell market value in 2023 is $7.2 billion, by 2027 it will be $9.1 billion and by 2033 it is projected to reach $35 billion15. 

• For hydrogen fuel cell vehicle market, we expect it to reach $42 billion in 2026, corresponding to an annual compound growth rate of 66.9%12,21. 

• The current selling price of PCS is $70 per gram or $47,000 per kilogram of PCS.  

• We are looking to further reduce manufacturing costs by 80 to 90%. 

D7 Synthesize PFD: 

• Fully developed PFD of scaled up process. 

• Benchmark data on current unit operations, including utility and energy usage. 

• Equipment summary and stream tables for both current and scaled up process. 

D8 Final Senior Design Materials: 

• Design project kickoff presentations by Dec 9th  

• Final report by May 7th  

• Final Presentation and Poster by May 2nd and 3rd 

Sources:  

(1) Bermudez, J. M.; Evangelopoulou, S.; Pavan, F. Hydrogen. IEA. https://www.iea.org/energy-system/low-emission-fuels/hydrogen (accessed 2023-11-26).  

(2) Xiao, Y.-X.; Ying, J.; Liu, H.-W.; Yang, X.-Y. Pt–C Interactions in Carbon-Supported Pt-Based Electrocatalysts. Front. Chem. Sci. Eng. 2023, 17 (11), 1677–1697. https://doi.org/10.1007/s11705-023-2300-5.  

(3) Zhang, W.; Ma, Z.; Zhao, X.; Zhou, L.; Yang, L.; Li, P. Textural Effect of Pt Catalyst Layers with Different Carbon Supports on Internal Oxygen Diffusion during Oxygen Reduction Reaction. Front. Chem. 2023, 11. https://doi.org/10.3389/fchem.2023.1217565. 

(4) Turton, R.; Bailie, R. C.; Whiting, W. B. Analysis, Synthesis and Design of Chemical Processes; Prentice Hall, 2012.  

(5) Olivas, E. C.; Councilor, T.; Fiebelkorn, C.; Pat, D.; Commissioner, B. B.; Councilor, T. E.; Jones Mayor, T. M.; Keller Commissioner, A. Water resources drought management plan. Abcwua.org. 

(6) BR search. US EPA. https://enviro.epa.gov/enviro/brs_codes_v2.waste_lookup (accessed 2023-11-27).  

(7) Miceli, M.; Frontera, P.; Macario, A.; Malara, A. Recovery/Reuse of Heterogeneous Supported Spent Catalysts. Catalysts. 2021, 11 (5), 591. https://doi.org/10.3390/catal11050591.  

(8) Swedwatch.org. https://swedwatch.org/wp-content/uploads/2017/12/Problematic-Platinum-web.pdf (accessed 2023-11-26).  

(9) Osha.gov. https://www.osha.gov/laws-regs/regulations/standardnumber/1910/1910.1200 (accessed 2024-04-21). 

(10) Deq.mt.gov. https://deq.mt.gov/files/DEQAdmin/DIR/Documents/EPCRA/epcra_fact_sheet.pdf (accessed 2024-04-21). 

(11) Osha.gov. https://www.osha.gov/sites/default/files/Hierarchy_of_Controls_02.01.23_form_508_2.pdf (accessed 2024-04-21). 

(12) Energy and Environmental Solutions. Fuel Cell Handbook, Fifth Edition; Office of Scientific and Technical Information (OSTI), 2000.  

(13) Powder, P. Hydrogen and fuel cell catalyst maker Pajarito Powder receives Series-B investment from Hyundai Motor. PR Newswire. https://www.prnewswire.com/news-releases/hydrogen-and-fuel-cell-catalyst-maker-pajarito-powder-receives-series-b-investment-from-hyundai-motor-301345606.html (accessed 2024-04-21). 

(14) T680E. Kenworth Trucks. https://www.kenworth.com/trucks/t680e/ (accessed 2024-04-21). 

(15) Hazlegreaves, S. Fuel cell electric vehicles: The genesis of a new era or myth-busting vehicle technology?. Open Access Government. https://www.openaccessgovernment.org/vehicle-technology/52116/ (accessed 2023-11-26).  

(16) Fuel cell market. MarketsandMarkets. https://www.marketsandmarkets.com/Market-Reports/fuel-cell-market-348.html (accessed 2023-11-26).  

(17) Fortune Business Insights. Global fuel cell market size worth USD 36.41 billion, globally, by 2029 at a CAGR of 29.7%. Fortune Business Insights. https://www.globenewswire.com/en/news-release/2023/04/03/2639625/0/en/Global-Fuel-Cell-Market-Size-Worth-USD-36-41-Billion-Globally-by-2029-at-a-CAGR-of-29-7.html (accessed 2023-11-26).  

(18) Luo, Y.; Wu, Y.; Li, B.; Mo, T.; Li, Y.; Feng, S.-P.; Qu, J.; Chu, P. K. Development and Application of Fuel Cells in the Automobile Industry. J. Energy Storage. 2021, 42 (103124), 103124. https://doi.org/10.1016/j.est.2021.103124.  

(19) Johnson, W. Bekaert makes Series-B investment in electrocatalyst maker. Pajarito Powder. https://pajaritopowder.com/2022/bekaert-makes-series-b-investment-in-electrocatalyst-maker-pajarito-powder/ (accessed 2024-04-21). 

(20) Nih.gov. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8123390/) (accessed 2024-04-21). 

(21) Giorgi, L.; Leccese, F. Fuel Cells: Technologies and Applications. The Open Fuel Cells Journal. 2013, 6 (1).  

1. Background and Market Analysis 

1. Introduction  

Engineered porous carbon supports (PCS) are becoming increasingly relevant for various applications owing to their relatively low cost and high tunability to allow for exceptional material properties1. Applications such as catalysis, adsorption, and energy storage are seeing an increase in demand for the development of sophisticated and efficient technologies. This is directly a result of the growing concern over rising man-made emissions that contribute to climate change and global warming. In 2022, the International Energy Agency reported emissions reaching an all-time high of 36.8 gigatons with large contributions primarily from fossil fuel energy, coal, agriculture, industrial and transportation sectors2,3. To mitigate this, emissions reduction strategies to create a sustainable carbon cycle can be categorized into the following: i) adopting low-carbon alternative sources of energy and energy storage ii) carbon and greenhouse gas capture and iii) utilization or sequestration.  

Pajarito Powders LLC, an Albuquerque, New Mexico-based startup, seeks to address the first strategy, replacing carbon-based energy sources, by developing hydrogen fuel cells. Diatomic hydrogen as an energy storage medium has a high energy content per unit weight which when split into protons via fuel cells, produces an electric current and enables the reduction of oxygen to generate water4. Pt and other precious metal-based catalysts are often used to split hydrogen due to their exceptional ability to overcome the binding energy of protons5. Consequently, such catalysts make up 59% of the cost of an average fuel cell stack6. PCS and particularly nitrogen-doped PCS serve as relatively inexpensive and effective supports for precious metal atoms or nanoparticle dopants which are susceptible to agglomeration or sintering and would otherwise be unstable in other support media7. In addition to maximizing dopant surface area, they provide enhanced electrocatalytic properties such as porosity, selectivity, and conductivity8. 

To address the second strategy, efficient carbon capture from point sources and the atmosphere requires porous materials with high surface areas. Again, nitrogen-doped PCS are excellent candidates as molecular “sponges” for capture owing to their high chemical stability, structural stability, distinct surface chemistry and selectivity8. Furthermore, CO2 utilization as a renewable resource can be an effective strategy to close the carbon cycle. CO2 conversion to added value products such as methanol are well studied in literature and are most effective through catalysis. Again, nanoparticles or active metal sites supported on various PCS have been shown to be highly effective in gas phase CO2 hydrogenation, electrochemical reduction, and oxidation8. 

Tackling the significant volume of rising emissions therefore requires large and regular outputs of these technologies and their materials of construction, including PCS. Unfortunately, current production of PCS are limited by their difficulty in scale-up resulting in small lab-scale processes that incur high operating and therefore selling costs. These difficulties arise due to many factors which this report aims to amend. 

2. Pajarito Powders Goal 

Pajarito Powders develops PCS primarily for hydrogen fuel cell and fuel cell electrolyzers. They specifically target mesoporous nitrogen-doped PCS (2-50nm pores) due its high dry H+ conductivity, low mass transport resistance, long interparticle distance, and reduced ionomer conductivity which are desirable properties in hydrogen fuel cells. They claim to have excellent control of critical mesopore size, pore volume, and surface area which is exhibited by their wide range of various PCS products sold on the market9. Their goal is to augment their current production of 1 kg over multiple days in a lab process to 20 kg/day in support of the US Department of Energy’s target to develop 20,000 fuel cell stacks a year. This would advance this technology from lab scale validation to a full commercial launch.  

3. Current System Gaps  

The current process entails a risk-heavy use of hydrofluoric acid (HF) as a key component in removing the silica template. Employees frequently come to handle HF, exposing them to severe health hazards including acute toxicity in cases of oral, dermal, and inhalation as well as skin corrosive and severe eye damage10. In addition to the health risk posed to employees, the HF etching step is lengthy with three days spent on the etching process, thus identifying the process as the bottleneck step of the process. Due to the usage of HF, the manufacturing space is shut down, halting the manufacturing process. There are no quality checks currently incorporated within the process. Batch quality is determined using variable criteria of intuition and product coloring, rather than quantitative assessment. Additionally, no process scheduling has been completed, resulting in increased time between unit operations.  This leaves unit operations offline when it otherwise could be used in parallel processing strategies to maximize output.    

4. Project Constraints  

Current lab scale production is a batch process that poses numerous constraints in scale-up: 

1. Although many have reported various methods for minimizing the steps in PCS synthesis (10), Pajarito Powder has developed a hard template PCS synthesis that require key unit operations essential to their ability to have precise control over material properties and high purity.  

2. Our design must be able to produce relatively high purity materials with minimal contamination or loss due to material handling.  

3. Material and process hazards including their implications on the environment and human safety are increased with larger quantities. 

5. SMART Goals 

SMART is a strategy of goal making that ensures goals are attainable and meaningful. It does this by breaking down the goals into 5 sections: Specific, Measurable, Achievable, Relevant, and Time Based. Each of these sections helps frame each goal and sets reasonable expectations for the team. This process was done for our project and is summarized in Table 1 below. 

Table 1: SMART goals

Specific - Automate etching step to decrease manual handling of Hydrofluoric Acid  

Measurable - The economic viability of the product 

Achievable - Well-researched, large market  

Relevant  - Meet the growing demand for PCS 

Time Based - Increase production from 30g per day to 20kg per day in the next three years  

6. Market analysis  

Pajarito Powder’s previous research showed that a 90% cost reduction in PCS would open more markets for their product. While Pajarito Powder currently focuses on selling their PCS to long-haul trucking companies for hydrogen fuel cell catalysts, their technology has the potential to revolutionize various industries. The hydrogen fuel cell market holds promise, with an expected reach of $1.45 billion by 203011. However, the true potential lies in the mesoporous carbon market, which is projected to reach a staggering $6.3 billion by 203112. We see significant potential for Pajarito Powder's PCS beyond long-haul trucking. By providing a consistent mesoporous carbon support, they can empower major automakers to refine their own catalyst technology. But the applications extend far beyond vehicles. Mesoporous carbon shows promise in a multitude of fields, including controlled oral drug delivery systems, adsorption of harmful metals and other molecules from water, separation of large biomolecules, electrical double-layer capacitors, wastewater treatment, carbon capture, catalysis, and air purification13 (13). By lowering the price of their PCS by 90%, Pajarito Powder can unlock a vast consumer market across these diverse applications. This price reduction would make their technology more accessible and allow it to have a wider impact on various industries. 

7. Plant Location 

The Pajarito Powder facility is located in the metropolitan Albuquerque area in the state of New Mexico (NM). NM has an executive order effective as of 2022 which directs various state departments to support the hydrogen economy and foster collaborations with national labs and academic institutions14. This executive order along with ongoing efforts of the state to reach net zero emissions makes Albuquerque, NM an ideal location for the Pajarito Powders facility. The facility features a 4,000 sq. ft manufacturing space with large vertical space and multiple rooms equipped to facilitate production growth. The facility is not limited to PCS production and will also house other areas of hydrogen fuel cell production and testing. An additive bonus of this location is the direct proximity of engineers and scientists at Sandia National Labs, Los Alamos National Labs, and the University of New Mexico. These government and educational institutions feature a large pool of competent future employees to support the growth of the company. The facility is however, located in a residential area which is in close proximity to various educational buildings, commercial companies, and apartment complexes posing significant safety concerns. This will be discussed further in section four of this report. 

Sources:  

(1) Xia, Y.; Yang, Z.; Mokaya, R. Templated Nanoscale Porous Carbons. Nanoscale 2010, 2 (5), 639. https://doi.org/10.1039/b9nr00207c. 

(2) CO2 Emissions in 2022. IEA. https://www.iea.org/reports/co2-emissions-in-2022 (accessed 2024-04-19). 

(3) Us Epa, O. Global Greenhouse Gas Overview. 2016. 

(4) Hydrogen benefits and considerations. Energy.gov. https://afdc.energy.gov/fuels/hydrogen_benefits.html (accessed 2024-04-19). 

(5) Kazemi, A.; Manteghi, F.; Tehrani, Z. Metal Electrocatalysts for Hydrogen Production in Water Splitting. ACS Omega 2024. https://doi.org/10.1021/acsomega.3c07911 

(6) Kleen, G.; Gibbons, W.; Fornaciari, J.; Papageorgopoulos, D.; Satyapal, S. Energy.gov. https://www.hydrogen.energy.gov/docs/hydrogenprogramlibraries/pdfs/23002-hd-fuel-cell-system-cost-2022.pdf (accessed 2024-04-19). 

(7) Hu, Y.; Lu, Y.; Zhao, X.; Shen, T.; Zhao, T.; Gong, M.; Chen, K.; Lai, C.; Zhang, J.; Xin, H. L.; Wang, D. Highly Active N-Doped Carbon Encapsulated Pd-Fe Intermetallic Nanoparticles for the Oxygen Reduction Reaction. Nano Res. 2020, 13 (9), 2365–2370. https://doi.org/10.1007/s12274-020-2856-z. 

(8) Matsagar, B. M.; Yang, R.-X.; Dutta, S.; Ok, Y. S.; Wu, K. C.-W. Recent Progress in the Development of Biomass-Derived Nitrogen-Doped Porous Carbon. J. Mater. Chem. A Mater. Energy Sustain. 2021, 9 (7), 3703–3728. https://doi.org/10.1039/d0ta09706c. 

(9) Products. Pajarito Powder. https://pajaritopowder.com/products/ (accessed 2024-04-19). 

(10) Liu, H.; Zhao, J.; Li, X. Controlled Synthesis of Carbon-Supported Pt-Based Electrocatalysts for Proton Exchange Membrane Fuel Cells. Electrochem. Energy Rev. 2022, 5 (4). https://doi.org/10.1007/s41918-022-00173-3. 

(11) VERIFIED MARKET RESEARCH. Global Hydrogen Fuel Cell Catalyst Market Size by Catalyst Type, by Application, by Description, by Geographic Scope and Forecast. 

(12) Transparency Market Research. Mesoporous Carbons Market is Predicted to Reach US$ 6.3 billion in 2031, Expanding at a 6.6% CAGR: TMR Report. Transparency Market Research. https://www.globenewswire.com/en/news-release/2023/10/10/2757622/32656/en/Mesoporous-Carbons-Market-is-Predicted-to-Reach-US-6-3-billion-in-2031-Expanding-at-a-6-6-CAGR-TMR-Report.html (accessed 2024-04-19). 

(13) Rahman, M. M.; Ara, M. G.; Alim, M. A.; Uddin, M. S.; Najda, A.; Albadrani, G. M.; Sayed, A. A.; Mousa, S. A.; Abdel-Daim, M. M. Mesoporous Carbon: A Versatile Material for Scientific Applications. Int. J. Mol. Sci. 2021, 22 (9), 4498. https://doi.org/10.3390/ijms22094498. 

(14) Hydrogen. Env.nm.gov. https://www.env.nm.gov/climate-change-bureau/hydrogen/ (accessed 2024-04-21).

3. Unit Op Designs 

2.3.1- Mixer 

Mixing is the first step used in making the pre-pyrolysis mixture. This is a simple step that will ensure the homogeneity of our reactants. A homogenous mixture will result in the formation of a mesoporous template that will have novel catalytic properties8. Due to our reactant selection ensuring the suspension of our ceramic in the solvated carbon precursor is a critical step to creating the desired porosity, surface area, and pore diameter. 

A mixer is a common unit operation that is used in the food and beverage, pharmaceutical, and cosmetics industries. Typically, viscous or dry powders blend into a homogenous mixture to be sold to the general public. Common industrial mixers include batch, continuous, planetary, and ribbon mixers that are designed for specific fluid parameters. Pajarito Powder currently has a batch mixer with a spiral blade impeller. This was a concern and area of improvement as the specific impeller did not have up and down motion and scraping blades at the bottom and sides of the mixing bowl to ensure unmixed material did not remain. The unmixed mixture could result in contamination between batches or reduce catalyst properties. The mixing of our reactants is an aggressive step that is second to the ball milling step. Fed into the mixer is dry 90% DA chitosan with 1% (w/v) acetic acid to ensure the biopolymer is able to dissolve readily in water. Next colloidal silica and water are added carefully to the mixing tank to reduce to possibility of dust formation and slip hazards. A key assumption made during the mixer design was no heat transfer occurs and no reaction takes place between the reactants. 

Material of Construction (MOC):  

Colloidal silica is a suspended mixture of silica particles in water. The silica particles have a Mohs hardness of 7 which will require a strong material of construction that can withstand the abrasiveness of silica9. Our initial decision was to have a carbon steel mixing vessel since this is a cheap metal to construct and is abrasion resistant. Consulting compatibility charts of carbon steel at various carbon weight percentages has a severe effect with acetic acid10. Due to its poor resistance to acids, it could cause acid corrosion in the mixing vessel resulting in hazardous conditions and product contamination.  

Stainless steel was the next choice for the mixer vessel since this material exhibits excellent corrosion resistance to acetic acid11 and can handle the abrasiveness of silica particles. The caveat of stainless steel is in cost and construction will become difficult since stainless steel has an increased hardness and strength compared to carbon steel. However stainless steel 316 will be a better choice due to its rigid properties and chemical resistance for both the vessel and impeller construction. Unloading the mixer will consist of a polypropylene pipe fitted to the bottom of the mixing vessel where the flow will be directed to a progressing cavity pump. This will allow for the system to be more streamlined and equate to less handling of the homogenized material.  

Mixing Vessel Sizing  

The mixer vessel is in the shape of a dished bottom vessel. This specific design will account for our impeller and will allow for both the up and down mixing and off-bottom suspension of silica particles. Our mixer design was sized using the volume of the solid and liquid materials for an approximate volume of 600 liters. Since we chose a dished vessel, we can solve for the radius using the equation for a hemisphere which comes out to 0.66 meters. Then from Perry’s Chemical Engineering Handbook an anchor agitator is approximately 95% of the vessel’s diameter12. Using this heuristic the diameter of the agitator and the vessel diameter can be solved which come out to 1.32 and 1.38 meters respectively. Using the known volume of 600 L the height of the vessel can also be solved using the radius which comes out to 0.44 meters and the approximate wall thickness of 316 stainless steel needed is 0.035 meters which is inches13. Although since we have a dished bottom vessel a cylindrical volume will be added to the hemispherical volume which will calculate to a total volume of 740 L. This is an obscure size and to find a commercial product a 1000 L vessel would be a better approximation. This way Pajarito Powder will have a capacity of 75% to reduce the possibility of overfilling and bubble/foam formation during mixing. The 1000 L vessel will stand on four legs with a clearance of 0.6 meters off the ground to ensure the outlet flow is making it to the progressing cavity pump.  

Impeller Design 

As mentioned previously the current impeller Pajarito Powder uses does not facilitate efficient up and down motion as well as ensuring the material deposits on the side of the mixing bowl are fully incorporated. The new impeller that we recommend to Pajarito Powder is a U-shaped anchor impeller with polypropylene scrapers on the bottom and sides as well as a centered 45° pitched paddle impeller with three blades. This specific impeller will ensure solid material does not cake onto the walls and bottom of the mixer. The diameter of the anchor impeller is 1.32 meters and a height of 0.4 meters. Scrapers were specifically designed to be polypropylene due to its compatibility with acetic acid, water, and high viscosity mixtures as well as high durability. Two scrapers will be attached to the bottom of the impeller as well as two scrapers on the sides of the impeller. Due to the design of the impeller adding baffles became difficult since traditionally baffles help form von Kármán vortex streets to increase mixing for low-viscosity solutions. However, these prove difficult to include since baffles aren’t optimal since vertexing of the mixture will occur reducing blend time and increasing blend time. Our recommendation is to include an unbaffled vessel to ensure reactants are fully mixed and homogenized.

Power Consumption  

The operational speed for the mixer relies on multiple fluid parameters. Since we have a mixture of chitosan, silica, and water the effective density or viscosity is calculated using equation 1. Which uses each respective weight fraction (wt%) per component multiplied by the literature values for the density or viscosity.

Using the effective density and viscosity the Reynolds number can be calculated that will help solve for the power per unit volume and the actual power used by the agitator. The actual power requires a motor that is 50 Hertz and 2 poles although this will provide more than enough rotations per minute (rpm) to the agitator. The speed we recommend Pajarito Powder uses is in the range of 60 rpm to 100 rpm however for the calculations we have done use 100 rpm. The Reynolds number for the impeller is shown below in equation 214: 

Where n refers to the rotational speed in rpm, d is the diameter of the impeller, and is the density and is the mixture viscosity. The Reynolds numbers for both the paddle and anchor impeller in laminar flow regimes. This is desired since the sensitivity of our material has not been researched further however, from calculating the shear stress, of 0.6 Pascals a low power input is required to achieve a homogeneous mixture. The time it will take to have 95% homogeneity is less than a couple of minutes. This is due to the viscosity of the mixture which must be studied further to obtain an accurate blend time.  

After calculating other parameters, the power number, can be calculated for an unbaffled vessel using equation 314.

Where 𝛽 is the dimensionless coefficient, D is the diameter of the vessel, d is the impeller ,H is the liquid depth, and f is the friction factor. The values from this equation are 0.24 and 0.29 for the paddle and anchor respectively. However, these are not in terms of watts which is easily converted using equation 4 shown below. 

Assumptions: 

1. Uniform heating occurs in the dryer.  

2. No shrinkage of material due to its porosity.  

3. The volume of acetic acid is dilute.  

4. The duct length is 25ft and the duct diameter is 5 inches.

Tray Dryer Sizing 

The tray dryer is designed around the volume of post-mixed material. This way the dryer will have enough capacity to hold and dry the material for both drying stages. The total volume of homogenized material from the mixer is approximately 550 L and a mass of 621 kg. Since we have assumed the acetic acid becomes dilute due to the bulk volume of water being much larger than acetic acid. The moisture content must be calculated to ensure the dried material can be fed into the ball mill. This is because dry ball milling for powders is more beneficial than wet ball milling15. The precise reaction is denoted as mechanochemistry, which describes the phenomenon of thermal reactions in fluids due to mechanical forces. With this knowledge, the desired dried material must have a moisture content of 1%. However, the initial moisture content is approximately 77% which was calculated by the ratio of the wet reactants divided by the total volume. Thus, 76% of the total mass which is primarily water must be evaporated. Which equates to 472 kg of water resulting in 149 kg of dried material with trace amounts of water.  

The temperature of the dryer will be held is 90℃. Which was chosen based on drying curves for the raw materials. Since silica has a high melting point of 1700℃ it would be trivial to find drying curves for silica. However, the drying curves for chitosan must be explored since this provides the template for silica to be suspended. The specific drying curve is shown in Figure 5.

The operating temperature and the capacity to dry being known we sized the trays based on heuristics and commercial tray sizing. Maximizing surface area rectangular trays were chosen which are also commercially available. From commercial sizing, trays with a width of 0.45 meters, a length of 0.64 meters, and a depth of 0.04 inches were used. This resulted in a tray surface area of approximately 0.4m2 and a volume of 10 liters. Using a heuristic from the Handbook of Industrial Drying the maximum loading depth is 40 mm to follow this heuristic we decided to have a loading depth of 20 mm which gives us a volume of material per tray of 6 liters17. Using the tray volume with material loaded the number of trays can be calculated which is exactly 93 trays. However, we have two drying stages one after the mixer and the other after the filter. Using a material balance the volume that was estimated from the filter was approximately 35 liters of filter cake. Using the same tray dimensions the second drying step will require approximately 7 trays bringing the total amount of trays to 100 which is common in industrial dryers. Even so, this is the minimum number of trays required, it would be more beneficial to have 120 trays to ensure we have enough space to hold material for both drying steps. Based on heuristics the approximate spacing per tray will be 0.1 m giving a total drying surface area of 45m2 without mixture. Since the trays will be used every day to ensure efficient drying of the material and durability, the trays will be 304 stainless steel. The residence time for a tray dryer is between 1-6 hours however the time to dry is approximated to be 4 hours17. Where it was determined that 1 hour will account for the loading and unloading of the material.  

Power Consumption 

Traditionally tray dryers are electric which provide power to heaters or fans. The tray dryer that we designed will have an internal fan that will recirculate the heated air between the trays. Ensuring containment of heat, the dryer will be double walled with proper insulation to minimize heat loss to the surroundings. The first parameter that must be calculated is the Reynolds number of the heated air which was approximately 70,000 which has a turbulent flow regime. Using the Reynolds number the friction factor, f can be calculated using equation 5: 

Where A and B are friction factor parameters from the Churchill 1977 equation18. The friction factor will be used to calculate f1 which is analogous to the head loss or pressure drop. The pressure drop depends on the length and diameter of the duct as well as the density of heated air. The pressure drop is then used to calculate the pressure loss of air in the dryer. This is done using equation 6: 

Where f1 is the pressure drop in the vent, Z0 is the loading depth and V is the drying air velocity. The pressure loss of air in the dryer is roughly 578 pascals which is negligible since the loss is less than a kilopascal. From the results of equation 6, the fan power can be calculated using equation 7 shown below.  

Where Ff is the amount of air recycled in the dryer it was assumed that the total dryer area without mixture multiplied by the velocity of airflow will give the amount of recycled air in kilograms per second. Where changes of heated air will happen around every hour based on the ASHRAE guidelines dryers need to be well-ventilated with approximately 30 changes per hour19.  

The power that was calculated from equation 1.3 was approximately 1.3 kW from the fan although to heat the air and trays the heat transfer rate must be calculated which is shown in equation 8.

Where T2 and T1 are the dry bulb temperature of 90℃ and room temperature respectively and d refers to the thickness of material on the tray. Using the heat transfer rate to calculate the heat flow in watts which was roughly 35 W. Resulting in a total of 1.3 kW of power supplied to the dryer however, from commercial products the actual power supplied to the dryer is 6 to 10 kW.  

However, more experimentation must be done to accurately size a tray dryer. Primarily the drying curves for the desired raw material as well as the more accurate drying times that achieve the final moisture percentage more readily rather than an estimate. This way the drying kinetics can be studied to pick a dryer that will be more suitable for the inlet composition.  

Our ball mill sizing centers around the mass of product being processed at this stage. In a typical setup, the ball mill is charged with 50% griding media. Accounting for roughly 26% void space between the spherical media, this translates to 13% of the total volume in the mill is void space (0.5*0.23=0.13). It is important to ensure sufficient product loading above the griding media to guarantee continuous contact between the milling media and the product. For proper loading, there must be an adequate amount of product above the grinding media. This is demonstrated in Figure 6 below. Inadequate solid loading leads to the media striking each other, causing excessive wear on the balls. 

Proper loading of the ball mill ensures that each piece of griding media is always in contact with solid material. To achieve proper loading, an additional 12% volume above the media charge is recommended22. Including the 13% volume taken up by void space, which the added product will occupy, this brings our total product charge to 25% of the ball mill volume. Considering the 65L of solid being added to the ball mill, we determined that the ball mill required a volume of 160L. This ensures optimal grinding conditions and a prolonged lifespan. 

Grinding Media 

Silica, with a Mohs hardness value of 7, requires a grinding media of comparable hardness23. It was determined that zirconia, with a hardness of 9, would be sufficient for this system24. In addition to hardness, zirconia beads exhibit exceptional resistance to wear and have high chemical resistance24. These characteristics significantly lower the risk of zirconia contamination, which is an issue with steel balls. 

The size of griding media will determine the size of the output feed. To determine the size of the media, we employed the Fred Bond model shown in equation 9, expressed as: 

Where Pf is the size of feed after the pass, Wi is the work index of the feed, C is an empirical constant depending on the shape of the grinding media (200 for balls), Vcr is the percentage of mill critical speed, Sgs is the specific gravity of feed, and Db is the diameter of the ball mill25.  

Using this equation, we determined that this ball mill would require half inch zirconia for the grinding media. 

Ball Mill Speed and Power 

The operational speed of a ball mill depends on its size, calculated based on the critical speed of the device. This critical speed is the point at which centrifugal forces within the mill are equal to gravity, resulting in the particles staying on the wall, and you effectively get no grinding26. At this speed, the ball mill behaves like a centrifuge. The critical speed can be expressed as a function of diameter using equation 10: 

Where Db is the diameter of the ball mill in feet. This gives us a critical speed of 57 rpm.  

Extensive experimentation has been conducted to determine the optimal range of ball mill speed and has found it should stay between 60 and 90 percent of this critical speed. At higher speeds, more energy is inputted into the system generating greater mechanochemical reactions, however, this puts great wear on the grinding media and mill liner. It has been concluded that a mill speed of 72 percent of the critical speed (41.2 rpm) will increase the life of our system27. 

The power requirement of our system was estimated using an empirical model designed to provide a rough estimate when exact system parameters are unknown. It is important to acknowledge that this estimation solely considers particle size and does not account for mechanochemical reactions. More accurate models require knowledge of complex parameters such as the system’s work index of the system, dynamic angle of change, and dynamic center of gravity, which all require extensive testing to determine. Assumptions made for material coefficients further highlight the need for laboratory validation.  

The empirical power model, shown below in equation 11, is expressed as:

Where K is the grinding medium coefficient, M is the mass of material loaded in the ball mill, and Db is the diameter of the mill27. K changes based on both the grinding media and product being grinded, making this approximation difficult. Using this equation along with our assumptions, our ball mill motor is sized at 3.05 kilowatts. 

It is necessary to conduct further milling tests to fully understand the power required for the expected mechanochemical reactions to occur, as well as how exactly zirconia and the carbon material interact with each other in the mill. These tests would provide data that could be used to further enhance the accuracy of our calculations. 

Where V is the terminal velocity of the particle, d is the diameter of the particle, g is the gravitational constant, ρs

is the density of the particle, ρl is the density of the liquid, and μ𝜇 is the viscosity of the liquid (34). 

From this calculation, and using the values in found in Appendix B, a terminal velocity of 0.44 mm/s is calculated. However, considering the substantial volume of material within the tank, the influence of hindered diffusion cannot be ignored. Hindered diffusion occurs when particles interact with each other as they settle within a packed tank35.  

To determine the hindered settling rate, equation 13 is applied:

Where V0 is the hindered settling rate and C is the volume ratio of solid to liquid in the tank. For our system, this ratio is approximately 0.05.  

The constant Z is calculated using equation 14 where:

Here, D is the diameter of the tank, and Ar is known as Archimedes number and can be calculated using equation 15:

Using these equations, a hindered velocity of 0.31 mm/s was determined. Considering the height of the material, this results in a settling time of approximately 24.4 minutes. To ensure adequate settling of all solids, the system will wait 30 minutes. After this time, P-106 will be turned on and the etching liquid will be pumped out of the tank.  

Once removed, 100L of fresh water will be added to the tank. The impeller will be turned back on for 15 minutes, after which the settling and liquid removal process will be repeated. The cycle continues until the system reaches a pH of 3, monitored by a pH sensor in the tank.  

After reaching the desired pH level, the solution will be pumped to the filter stage while the impeller remains active to minimize solid loss. Following this, the tank undergoes flushing with water to remove any remaining solids. 

Etching Quality Check  

This system has the added benefit of allowing quality checks to be conducted concurrently with the etching process. A separate QA step is done by pumping a small quantity of solution from the tank during etching into V-103. This extracted material can undergo thorough washing and subsequent utilization in characterization tests.  

The main test that will be run is the ashing test, which serves as a primary means of assessing the completeness of the etching process. During this procedure, the material is burned in an oxygen rich environment which turns all the carbon into CO2. Any silica present within the material remains unaffected by this treatment. The post-ash material can then be weighed to determine how much silica remains.  

This analytical approach allows lab techs to determine the completeness of the etching process without requiring the washing of the entire batch, as is current practice. By implementing this quality check, significant time savings will be seen, eliminating the potential for days of delay typically associated with problematic etching batches.  

2.3.7- Scrubber

Where Gi and Go are the gas flow rates entering and exiting the column respectively, in cubic meter per second. Gmol, i and Gmol, o

are the molar gas flow rates entering and exiting the column respectively, in moles per second. Li and Lo are the solvent flow rates entering and exiting the column respectively, in cubic meter per second. Lmol, i and Lmol,  o are the molar solvent flow rates entering and exiting the column respectively, in moles per second. Xi and Xo are the inlet and outlet ratios of moles of pollutant per moles of pollutant-free solvent in the liquid stream respectively, Yi and Yo are the inlet and outlet ratios of moles of pollutant per moles of pollutant- free gas in the gas stream respectively, xi and xo are the inlet and outlet pollutant mole fractions in the liquid stream respectively, yi and yo are the inlet and outlet pollutant mole fractions in the gas stream respectively. Ls and Gs are the pollutant free solvent and gas flow rates respectively, in cubic meter per second.

To begin the design procedure, we must first assume some of the stream conditions of the inlet and outlet of both gas and liquid streams. The first assumption we made is that the mole fraction of acid vapor in air in the etching station is 0.004 molar of nitric acid, we ignored hydrofluoric acid in the calculations as its concentration is much less compared to nitric acid; this value is also the pollutant molar concentration of the inlet gas stream and needs to be confirmed and adjusted via a chromatography test. The value for the pollutant molar concentration of the outlet gas stream and the inlet liquid stream is assumed to be 0. We assumed the ventilation has a 0.427 cubic meter per second per square meter flow, based on the standard air flow for fume hoods38. The etching station is roughly 3.363 square meter which means ventilation the flow rate Gi is around 0.427 cubic meter per second. We also assumed that the solvent has the same physical properties as water due to it being a very diluted aqueous solution of sodium bicarbonate. 

After assuming the above basis, we begin the calculations. First, we calculate the quantity of nitric acid in terms of moles of HNO3 per moles of HNO3- free gas (Yi) using equation 16:

Where [HNO3] is the molar concentration of nitric acid in the air. We then calculated the quantity of pollutant in the exit gas stream (Yo) using equation 17, which is written as:  

Where 𝜂 is the removal efficiency which we assumed to be 99 percent. Finally, we know that the pollutant quantity in the liquid inlet stream Xi

is 0. After defining the Yi, Yo, and Xi values, we need to consult the McCabe-Thiele equilibrium curve of nitric acid and water below to determine the X∗o value, which is the liquid mole fraction of nitric acid at equilibrium with water at the designated nitric acid vapor mole fraction:  

From Figure 10, we can find the Xo* value when Yi is equal to 0.004 which is around 0.0023. From the plot, we can find its slope which is the following ratio:  

Where (Ls/Gs)min is the ratio of moles of pollutant-free solvent Ls per moles of pollutant-free gas Gs. From equation 18, we obtained a (Ls/Gs)min value of 1.72. This ratio is usually an unrealistic value and needs to be multiplied with an adjustment factor which commonly ranges from 1.2 to 1.5. However, because the equilibrium curve at this section is very close to a straight line, for this calculation, we used an adjustment factor of 1.0 to obtain the (Ls/Gs)actual value:  

From the above calculation, obtained the (Ls/Gs)actual value of 1.72. To find Ls, we calculated Gs using the below formula:

Where ρG is the density of the gas stream and MWG is the molecular weight of the gas stream. Due to the low pollutant concentration, we can assume these values equal to those of ambient air. From the equation, we obtained a Gs value of 0.022 mol/s. We then calculated the Ls value by using equation 21 below:  

We obtained an Ls value of 0.038 mol/s. With the Gs and Ls values, we can calculate the Gmol, i and the Lmol, i values using equations 22 and 23:  

We then obtained the Gmol, i and the Lmol, i values of 0.022 and 0.038 respectively. Next, we calculate the pollutant mole fraction in the liquid stream outlet Xo using equation 24:  

Where Yo∗ is the outlet mole fraction of the pollutant in the vapor phase in equilibrium. We then obtained the values of Xo of 0.11 and Yo∗ of

7.0∗10−5. 

The last stream condition that we needed to calculate is the inlet solvent flow rate Li, which can be calculated using equation 25: 

Where MWL and ρL are the molecular weight and density of solvent respectively which we assumed to be equal to those of water. We obtained an Li value of 7.0∗10−7 m3/s.

We then calculated the absorption factor AF which describes the relationship between the equilibrium line and the liquid-to-gas ratio using equation 26: 

Where m is the equilibrium line and can be calculated using equation 27:  

Where Y∗i is the inlet mole fraction of the pollutant in the vapor phase in equilibrium which we assumed to be 0. From equations 26 and 27, we obtained the absorption factor value of 45.71. 

After we determined the inlet and outlet stream conditions, we began modeling the packed bed column. First, we chose our packing type to be the 2-inch ceramic Raschig rings as they are one of the more common packing materials, also our calculations down below also indicate that the liquid flow rate is sufficient in wetting this packing type. However, we recommend further testing to find the most appropriate type of packing for this process. Then, we started calculating the Abscissa value and the Ordinate using the below equations:

We obtained the Abscissa and Ordinate values of 0.062 and 0.21 respectively. We then calculate the Gsfr,  i, which is a superficial mass gas flow rate per area using the below equation:  

Where gc is the gravitational constant, FP is the packing factor which we obtained from Barbour et al., and 𝜓 is the density ratio of the scrubbing liquid to water which we assumed equal to 1 due to the low sodium bicarbonate concentration. We obtained a Gsfr,  i value of 55.81 g/s/m2.

We then calculated the dimensions of the column using the following equations:

Where A is the cross-sectional area of the column and D is the column’s diameter. From the above equations, we found the values of A and D to be 0.015m2 and 0.14 m respectively. We then calculated the superficial inlet liquid mass flow rate per area of the Lsfr,  i using the formula below:  

We obtained a Lsfr,  i value of 44.40 g/s/m2. To make sure the column operates correctly, this value needs to be high enough to wet the packed bed. To check this, we calculated the (Lsfr,  i)min value using the following formula: 

Where MWR is the minimum wetting rate in square meter per second which for common packing materials is equal to 3.4∗10−5 m2/s. 𝛼 is a packing constant, for 3/8-in plastic Pall rings this value is 3.82. We found that the (Lsfr,  i)min value is 0.13 g/s/m2, which is lower than the Lsfr,  i value, which means that our current solvent flow rate is sufficient to wet the packed bed and our choice of packing material is appropriate.

For the final step of sizing the packed bed column, we calculated the packing height and the column height using the below equations:

The fundamental design of a plate and frame filter is a sequence of plates and frames with the same dimensions. The design of the plate depends on the fluid properties of the slurry as well as the desired degree of clarification. The common filter plates that are used are recessed, membrane, and recessed chamber filter plates. The specific plate that was researched was a flat plate with a filter cloth that will accumulate solid material onto the hollow filter frame. Figure 11 describes the components of the plate and frame filter however a plate and frame filter press works by pressurizing a chamber that is created by the series of plates and frames. Plates are covered in filter cloth or filter paper with a specific mesh size, the mesh size for our process is approximately 40040. The methodology behind choosing a mesh size of 400 that targets 40-micron-sized particles is based on the particle self-filtering the slurry from the post-etching slurry. As the particles leech from the feed port to the filter cloth the small particles will clump on the surface of the filter medium. Creating an artificial filter via adsorption filtering the slurry over time, a higher mesh filter could be used although this would be costly and redundant. However, if the pressure in the filter press is low, adsorption will not occur which could result in higher mesh filters in the 1000 to 1200 size range.  

Filter sizing was estimated using the theoretical volume from the etching stage. The sizing of both the plates and frames we determined from commercial-sized plates and frames and heuristics. Heuristics for plate sizes can range from 10 by 10 centimeters to 2 by 4 meters and from industry, 400 mm by 400 mm plates are the most common12. Consulting material compatibility charts the material of the plates and frames will be polypropylene as well as the pump, vessel, and piping29.  

The theoretical volume of liquid feed from etching is approximately 67 liters. Where it was assumed that the plates have a width of 0.03 meters which is roughly one inch and the depth of the filter cake that is accumulated onto the filter cloth being ½ an inch.  Which resulted in the total filter area being calculated from simple division. Due to the plate and frame giving an area of 0.16m2 the number of plates and frames totaling 36 with 17 individual plates and frames in series including a start and end plate. Fundamentally, each plate requires a filter cloth to catch the material therefore the plate and frame filter will have approximately 17 filter cloths that are approximately $20/cloth41. 

Filter Press Power Consumption 

Estimating the power requires the density of solids in the filter cake to be determined which is shown in equation 41:

Whereas ρs is the density of the solids in the filter cake, mC and mH2O being the mass of carbon and mass of water. Which assumes a basis of 1 gram of carbon for the post-etching carbon composition and VT being the total volume of carbon and water. Emanating a density of 1287 kg/m3 which is similar to water. The density is used in equation 42 shown below to calculate the specific cake resistance. 

𝛼 is the specific cake resistance that functions on the porosity, ε and S0 is the specific surface area. Although to solve for the specific surface area, Pajarito provided information on the specific area from BET. Which is the surface area per mass, but we need the surface area per volume which was calculated by dividing the surface area per mass by the density of the slurry. However, this does not give an accurate specific cake resistance which resulted in the assumption that the cake resistance is negligible due to its low resistance. Although, it is negligible the filter medium resistance must be used instead but has its limitation since it is a difficult parameter to find which was why it was assumed that the filter medium resistance (RM) is 1E05cm−1. Using this assumption the rate of volume change over time per area can be solved which is denoted as:

Where ΔPt is the pressure difference across the filter press, gc is the gravitational constant, and w is the weight of the solids in the slurry per volume of liquid in the slurry42. Where the derivative of volume over time per area is the velocity, it can be easily solved to obtain the flow rate. Once this simple algebraic step is made the power to pump the slurry into the filter can be solved using equation 44 below:

Where P is the power for pumping liquids, Q is the flow rate in m3/min, ΔP is the pressure drop in bar, and ε is the pump efficiency43. As a diaphragm pump is not typical, it is like a rotary pump which has an efficiency of 55%. Giving the power to pump the slurry through the plate and frame filter press to be approximately 0.002 kW. However, plate and frame filter presses operate using hydraulic power as well which squishes the plates and frames together to achieve a cake on the filter cloth. The estimate of power consumed by the hydraulic press is around 0.756 kW which is a total of 0.76 kW of power needed to operate the plate and frame filter press. The power to operate the plate and frame filter press is not far off from commercial values of approximately 1.5 kW however increasing the number of plates and frames will bring the value closer to the literature value. The approximate time to filter the post-etched material is 2 hours although the common issue with plate and frame filter presses is the cleaning process. Therefore, we have the system being cleaned twice with water that totals 76 liters despite washing the filter cake it must be blow-dried. The compressed air will be supplied from the machinery, blow drying will ensure the filter cake doesn’t have water or recycled filtrate. Including these steps in our filtration scheduling the total time is estimated to be 4 hours.   

While a plate and frame filter press could be a better solution to filter the post-etched material the specific dynamic need to be modeled and researched. Specifically, the volume of filtrate collection over various pressures to determine if the filter cake is compressible and the volume of filtrate over constant pressure for an incompressible cake. The operating pressure was chosen from Pajarito Powder’s current filtration data although the typical operation pressure for filter presses is at 100 psig. Even so, the amount of product being produced per day is low compared to typical chemical plants and increasing the pressure could deform or destroy the product. If the pressure tolerance of the post-etched material as well as the raw powder can be further researched this could aid in less filtering time and transportation of powder from unit operation to unit operation.

Sources: 

1) Pavlenko, V.; Khosravi H, S.; Żółtowska, S.; Haruna, A. B.; Zahid, M.; Mansurov, Z.; Supiyeva, Z.; Galal, A.; Ozoemena, K. I.; Abbas, Q.; Jesionowski, T. A Comprehensive Review of Template-Assisted Porous Carbons: Modern Preparation Methods and Advanced Applications. Mater. Sci. Eng. R Rep. 2022, 149 (100682), 100682. https://doi.org/10.1016/j.mser.2022.100682. 

(2) Lee, J.; Kim, J.; Hyeon, T. Recent Progress in the Synthesis of Porous Carbon Materials. Adv. Mater. 2006, 18 (16), 2073–2094. https://doi.org/10.1002/adma.200501576. 

(3) Feng, J.; Yan, Y.; Chen, D.; Ni, W.; Yang, J.; Ma, S.; Mo, W. Study of Thermal Stability of Fumed Silica Based Thermal Insulating Composites at High Temperatures. Compos. B Eng. 2011, 42 (7), 1821–1825. https://doi.org/10.1016/j.compositesb.2011.06.023. 

(4) Ourworldindata.org. https://ourworldindata.org/co2-emissions#:~:text=In%201950%20the%20world%20emitted,yet%20to%20reach%20their%20peak. (accessed 2024-04-19). 

(5) Fuel cells. Energy.gov. https://www.energy.gov/eere/fuelcells/fuel-cells (accessed 2024-04-19). 

(6) Taguchi, A.; Schüth, F. Ordered Mesoporous Materials in Catalysis. Microporous Mesoporous Mater. 2005, 77 (1), 1–45. https://doi.org/10.1016/j.micromeso.2004.06.030. 

(7) Steinert, M.; Acker, J.; Oswald, S.; Wetzig, K. Study on the Mechanism of Silicon Etching in HNO3-Rich HF/HNO3 Mixtures. J. Phys. Chem. C Nanomater. Interfaces 2007, 111 (5), 2133–2140. https://doi.org/10.1021/jp066348j. 

(8) Banham, D.; Feng, F.; Fürstenhaupt, T.; Pei, K.; Ye, S.; Birss, V. Novel Mesoporous Carbon Supports for PEMFC Catalysts. Catalysts 2015, 5 (3), 1046–1067. https://doi.org/10.3390/catal5031046. 

(9) Silica powder. Avani Group of Industries. https://avanitalc.com/minerals/silica-powder/ (accessed 2024-04-19). 

(10) Carbon steel chemical compatibility. Calpaclab.com. https://www.calpaclab.com/carbon-steel-chemical-compatibility-chart/ (accessed 2024-04-19).
(11) Stainless steel chemical compatibility. Calpaclab.com. https://www.calpaclab.com/stainless-steel-chemical-compatibility-chart/ (accessed 2024-04-19). 

(12) Green, D. W.; Southard, M. Z. Perry’s Chemical Engineers’ Handbook, 9th Edition, 9th ed.; McGraw-Hill Education, 2018. 

(13) Tank heads. Baker Tankhead, Incorporated. https://www.bakertankhead.com/products/tank-heads/ (accessed 2024-04-19). 

(14) Furukawa, H.; Kato, Y.; Inoue, Y.; Kato, T.; Tada, Y.; Hashimoto, S. Correlation of Power Consumption for Several Kinds of Mixing Impellers. Int. J. Chem. Eng. 2012, 2012, 1–6. https://doi.org/10.1155/2012/106496. 

(15) Kotake, N.; Kuboki, M.; Kiya, S.; Kanda, Y. Influence of Dry and Wet Grinding Conditions on Fineness and Shape of Particle Size Distribution of Product in a Ball Mill. Adv. Powder Technol. 2011, 22 (1), 86–92. https://doi.org/10.1016/j.apt.2010.03.015.
(16) Srinivasa, P. C.; Ramesh, M. N.; Kumar, K. R.; Tharanathan, R. N. Properties of Chitosan Films Prepared under Different Drying Conditions. J. Food Eng. 2004, 63 (1), 79–85. https://doi.org/10.1016/s0260-8774(03)00285-1. 

(17) Handbook of Industrial Drying; Mujumdar, A. S., Ed.; CRC Press, 2006.
(18) Offor, U. H.; Alabi, S. B. An Accurate and Computationally Efficient Explicit Friction Factor Model. Adv. Chem. Eng. Sci. 2016, 06 (03), 237–245. https://doi.org/10.4236/aces.2016.63024.
(19) Rampey, M. What is an Air Exchange Rate and Why is it Important? —. Air Assurance. https://www.airassurance.com/blog/2018/04/26/air-exchange-rate (accessed 2024-04-19). 

(20) Takacs, L. Hyperfine Interactions 1998, 111 (1/4), 245–250. https://doi.org/10.1023/a:1012670104517.
(21) Laurent Delafontaine, P. A. ACID-FREE PYROLYTIC SYNTHESIS OF M-N-C CATALYST. 2023/0416934 A1, December 28, 2023
(22) Ball mill loading - dry milling. Pauloabbe.com. https://www.pauloabbe.com/ball-mill-loading-dry-milling (accessed 2024-04-19)
(23) Silica powder. Avani Group of Industries. https://avanitalc.com/minerals/silica-powder/ (accessed 2024-04-19).
(24) Zirconium beads. Final-materials.com. https://www.final-materials.com/gb/179-beads-zirconium (accessed 2024-04-19).
(25) Bond, F.c. - 1958 - grinding ball size selection.Pdf. Pdfcoffee.com. https://pdfcoffee.com/bond-fc-1958-grinding-ball-size-selectionpdf-pdf-free.html (accessed 2024-04-19). 

(26) Michaud, D. Ball Mill Critical Speed. 911 Metallurgist. https://www.911metallurgist.com/blog/ball-mill-critical-speed (accessed 2024-04-19).
(27) Jxscmachine.com. https://www.jxscmachine.com/new/ball-mill-parameter-selection-calculation/ (accessed 2024-04-19). 

(28) Versatile split tube furnace FST / FZS - carbolite Gero. Carbolite-gero.com. https://www.carbolite-gero.com/products/tube-furnace-range/split-tube-furnaces/fst-fzs/ (accessed 2024-04-20). 

(29) Graco.com. https://www.graco.com/content/dam/graco/ipd/literature/misc/chemical-compatibility-guide/Graco_ChemCompGuideEN-B.pdf (accessed 2024-04-19). 

(30) Polypropylene chemical compatibility. Calpaclab.com. https://www.calpaclab.com/polypropylene-chemical-compatibility-chart/ (accessed 2024-04-19). 

(31) Labs, W. Why choose polyvinylidene fluoride (PVDF) for conveyor belting?. Food Engineering. https://www.foodengineeringmag.com/articles/97320-why-choose-polyvinylidene-fluoride-pvdf-for-conveyor-belting (accessed 2024-04-19). 

(32) Polyvinylidene Fluoride (PVDF) Market - industry analysis and forecast (2024-2030). MAXIMIZE MARKET RESEARCH. https://www.maximizemarketresearch.com/market-report/polyvinylidene-fluoride-pvdf-market/40002/ (accessed 2024-04-19).
(33) Polypropylene price 2022. Statista. https://www.statista.com/statistics/1171084/price-polypropylene-forecast-globally/ (accessed 2024-04-19). 

(34) Cadence CFD Solutions. Deriving stoke’s law for settling velocity. Cadence.com. https://resources.system-analysis.cadence.com/blog/msa2022-deriving-stokes-law-for-settling-velocity (accessed 2024-04-19). 

(35) EBSCOhost Login. Ebscohost.com. https://eds.p.ebscohost.com/eds/ebookviewer/ebook/bmxlYmtfXzIxMTUxMl9fQU41?sid=6ec89415-7088-4c7d-a7f0-6d761b86e5bd@redis&vid=0&format=EB&lpid=lp_206&rid=0 (accessed 2024-04-19). 

(36) The 3 most common types of wet scrubbers. Slyinc.com. https://blog.slyinc.com/the-3-most-common-types-of-wet-scrubbers (accessed 2024-04-19). 

(37) Sixth Edition. Epa air pollution control cost manual. Epa.gov. https://www.epa.gov/sites/default/files/2020-07/documents/c_allchs.pdf (accessed 2024-04-19). 

(38) Memarzadeh, F. Fume Hood Requirements and Testing. National Institutes of Health. 

(39) Sciencedirect.com. https://www.sciencedirect.com/topics/engineering/pressure-filter (accessed 2024-04-19).
(40) Filter cloth. Cleveland Wire Cloth. https://www.wirecloth.com/filter-cloth/ (accessed 2024-04-19).
(41) Nylon polyester screen mesh. miaqua. https://www.miami-aquaculture.com/screen (accessed 2024-04-20)
(42) Foust, A. S.; Wenzel, L. A.; Clump, C. W.; Maus, L.; Andersen, L. B. Principles of Unit Operations, 2nd ed.; John Wiley & Sons: Nashville, TN, 1980.
(43) Turton, R.; Shaeiwitz, J. A.; Bhattacharyya, D.; Whiting, W. B. Analysis, Synthesis and Design of Chemical Processes, 5th ed.; Pearson: Upper Saddle River, NJ, 2022.

The control system for our etching stage is shown in Figure 17 below. This system includes various control modes designed to manage the addition of fluids into the tanks, the dilution stage, and transport to the filter and hazardous waste storage tank. Each of these control modes operates independently to prevent interference between the controllers. 

Mode A: Tank Filling 

During tank filling, Control Mode A facilitates the addition of acid and water into the tanks. A flow controller manipulates CV-2 to regulate acid addition, followed by CV-1 to introduce the required amount of water. A level indicator in the tank is used as a redundancy to safeguard against over filling, ensuring the tank never exceeds 80 % of its capacity.  

Mode B: Dilution Stage 

Control Mode B is used when etching has been completed, and the solution is diluted prior to filtration. The level indicator controller adjusts CV-4 to drain the required amount of fluid, ensuring no solid is removed in this draining process. A flow controller then adjusts CV-1, which adds the required amount of water once the initial drain is complete. This process repeats until a pH indicator in the tank detects a pH of 3 in the tank. At this point, CV-5 can be opened, and material is pumped to the filter. It is important to note that a pressure switch is placed at the outlet of P-105. As discussed in the previous section, this was done to ensure that this positive displacement pump is shut off if pressure increases beyond a setpoint (This controller is running regardless of mode). 

Mode C: Quality Check 

For quality checks, Control Mode C manages the filling of the quality check tank. Once a technician is ready to perform a quality check, V-103 will be filled. A level controller will ensure that the tank is only filled to 30% of the tank’s volume.  

Given the extensive use of the onsite waste storage, a level indicator and controller has been added to this tank. If the level in the tank exceeds 80% capacity, CV-6 and CV-3 will be closed. This controller takes precedence over all other controllers, meaning the process cannot be run if there is not enough space in the waste tank. At this point, the waste must be treated or removed. 

Additional Controls 

To mitigate against human error, time delay switches have been added to process equipment including the ball mill, furnace, dryer, and mixer. These time delay switches ensure each piece of equipment can only run for a certain amount of time. This is to protect against a technician forgetting to shut off a certain operation, or when an operation is completed at night. 

Limit Switches are also present in all pieces of equipment, and set maximum operational parameters for devices, protecting against incorrect inputs, and prolonging the life of our equipment. Pumps will be limited to 20% above their normal operational output, furnaces will be limited to 1200°

C, and dryers to 120°C.

To protect employees from accidentally entering etching rooms under hazardous conditions, go-no go panels will be added outside each room. These indicate whether the room is safe to enter and will be based on two parameters: Air quality, and tank pH. These parameters will be monitored by acid gas analyzers and pH indicators respectively. If the gas analyzer reads above 3ppm and/or the pH in the tank is below 3, the panel will indicate it is not safe to enter the room, and the doors will be locked.

2. Federal regulation  

Clean Air Act:  

Through the mass balance for the 20kg per day production of PCS, we have determined that there are only trace amounts of flue gas which consists of NOx, CO, and Co2 being released into the atmosphere during the pyrolysis steps. The trace amounts currently don't pose a safety threat to employees or communities.  We will continue to monitor these emissions levels and comply with the Clean Air Act of 1963 (CAA) and New Mexico's NMED Air Quality Bureau17,18.  

Compressed gas:  

Our system uses two main sources of compressed gas including nitrogen and hydrogen gas due to this we will comply with OSHA regulations for compressed gas specifically 29 CRF 1910.101, and 29 CRF 1910.1022. Which states regulations for handling, storing, proper employment training, and how to secure gas cylinders to avoid physical hazards. Pajarito Powder will ensure that all gas cylinders will be properly secured while stored and in use. This includes being upright and held with chains to prevent the cylinder from following2. Gas cylinders will also be kept 20 feet away from any combustible material and in well-ventilated areas2. All cylinder tanks will be well and properly labeled allowing for easy identification of any potential hazard2. Lastly, both tanks will include pressure relief valves devices and proper CGA fitting should be used for all hydrogen cylinders3,4. 

Community Right to Know and Emergency Planning:  

The Emergency Planning and Community Right-to-Know Act (EPCRA) of 1986 was developed to make the public aware of any potential chemical hazards around them.  Since the facility exceeds the minimum Threshold Planning Quantity (TPQ) of 100 lbs. of Hydrofluoric acid and 1,000 lbs of Nitric acid storage we will comply with energy planning highlighted in sections 301-303, 304, 311-312, and 313 of EPA.  

Sections 301-304    

Sections 301 through 303 of the EPCRA highlight the need for evacuation procedures in a chemical emergency that ensures the safety of all employees and the community. In compliance with this Pajarito Powder will have an evacuation plan in place with all surrounding communities and employees aware of how to proceed in the event of an emergency. In addition, all employees will be trained in how to identify a chemical emergency and how to adhere to all appropriate reporting guidelines and procedures outlined in section 304. We will ensure that emergency reporting procedures are easily accessible to all employees. 

Sections 311-313   

Due to the large amount of hydrofluoric acid and nitric acid within the facility and hydrogen tanks, we will comply with the chemical reporting and inventory requirements highlighted in sections 311-313 of the EPCA. This includes providing floorplans that highlight all where chemicals are located within our facility along with Material Safety Data Sheets (MSDS) to Albuquerque, NM fire departments. MSDS will also be given to neighboring residential communities, educational buildings, and businesses, and easily accessible to the public. Lastly, Pajarito Powder will comply with all appropriate labeling of waste containers and chemicals. This includes using the globally harmonized system for all containers containing chemicals including all appropriate picograms allowing anyone to understand any potential hazards.  

3.  Employment safety 

OSHA regulations  

The system involves mostly manual steps besides the etching and filtration steps along with the usage and storage of multiple EHS, due to this we will comply with all OSHA standards. By complying with this administration and the Hazard Communication Standard (HCS), Pajarito Powder will ensure a safe, and healthy work environment for our employees. All employees will be aware of any hazards within the process, and hazardous chemical locations within the facility and be required to attend workshops and or safety courses to ensure adequate training.  All employees will also be required to follow appropriate personal protective equipment (PPE) beyond a certain portion of the facility and be familiar with NFPA ratings.    

Manual Handling  

As mentioned, within our report most material transfer from and to unit operations involves manual transfer. To ensure that employees do not come in direct contact with any of the material that can pose a potential safety hazard, contaminate the product or result in product loss we will use equipment to facilitate the transfer of material. This includes the usage of large trays that can be loaded into movable carts, reducing any physical hazard associated with having to carry an individual large tray around the facility. This is specifically useful in transferring the material from the mixer to the dryer and the filter to the dryer. Additionally, there will be large funnels that can facilitate the addition of any material and chemical needed within our process. Ultimately, we will use revolving belts to transport material within unit operations, which would reduce all physical and safety hazards associated with the transportation of material. 

Sources: 

(1) Hendershot, D. C. Inherently safer design: The fundamentals. Aiche.org. https://www.aiche.org/sites/default/files/cep/20120140-1.pdf (accessed 2024-04-19)
(2) Sciencedirect.com. https://www.sciencedirect.com/topics/chemistry/hydrofluoric-acid (accessed 2024-04-19) 

(3) Osha.gov. https://www.osha.gov/sites/default/files/Hierarchy_of_Controls_02.01.23_form_508_2.pdf (accessed 2024-04-19). 

(4) Hydrogen Fluoride. (n.d.). SAFETY DATA SHEET. Airgas.com. Retrieved April 23, 2024, from https://www.airgas.com/msds/001077.pdf 

(5) Rev., G. H. S. (n.d.). Safety Data Sheet. Fishersci.com. Retrieved April 23, 2024, from https://www.fishersci.com/content/dam/fishersci/en_US/documents/programs/education/regulatory-documents/sds/chemicals/chemicals-n/S25860.pdf 

(6) (N.d.-e). Fishersci.com. Retrieved April 23, 2024, from https://www.fishersci.com/msdsproxy%3FproductName%3DAC423220025%26productDescription%3DACETIC%2BACID%252C%2BGLACIA%2B2.5L%26catNo%3DAC42322-0025%2B%26vendorId%3DVN00033901%26storeId%3D10652 

(7)  Hazard Identification. (n.d.). Section 1: Product and company identification product name: Colloidal silica suspension, 0.06 m synonym: CO colloidal silica company name. Tedpella.com. Retrieved April 23, 2024, from https://www.tedpella.com/SDS_html/815-110_sds.pdf 

(8) (N.d.-f). Sigmaaldrich.com. Retrieved April 23, 2024, from https://www.sigmaaldrich.com/US/en/sds/aldrich/448869?userType=anonymous 

(9) (N.d.-d). Airgas.com. Retrieved April 23, 2024, from https://www.airgas.com/msds/001040.pdf 

(10) (N.d.-g). Airgas.com. Retrieved April 23, 2024, from https://www.airgas.com/msds/001026.pdf 

(11) (N.d.-h). Fishersci.com. Retrieved April 23, 2024, from https://www.fishersci.com/store/msds?partNumber=BP28191&productDescription=WATER+MOLECULAR+BIOLOGY+GRADE&vendorId=VN00033897&countryCode=US&language=en 

(12) Dioxide, C. (n.d.). SAFETY DATA SHEET. Airgas.com. Retrieved April 23, 2024, from https://www.airgas.com/msds/001013.pdf 

(13) Product, 1. (n.d.). Safety Data Sheet (SDS). Siliconesolutions.com. Retrieved April 23, 2024, from https://siliconesolutions.com/media/pdf/SS-5061SDS.pdf 

(14) N.d.-b). Fishersci.com. Retrieved April 23, 2024, from https://www.fishersci.com/store/msds?partNumber=AAL148290E&productDescription=HXFLROSILICIC+AC+35%25+W+.+2.5KG&vendorId=VN00024248&countryCode=US&language=en 

(15) Sodium Bicarbonate. (n.d.). Safety Data Sheet. Fishersci.com. Retrieved April 23, 2024, from https://www.fishersci.com/content/dam/fishersci/en_US/documents/programs/education/regulatory-documents/sds/chemicals/chemicals-s/S25533.pdf 

(16) Tetrafluoride, S. (n.d.). SAFETY DATA SHEET. Airgas.com. Retrieved April 23, 2024, from https://www.airgas.com/msds/001076.pdf 

(17) Us Epa, O. (2015). Overview of the Clean Air Act and air pollution. https://www.epa.gov/clean-air-act-overview 

(18) Air Quality Bureau. (n.d.). Env.nm.gov. Retrieved April 25, 2024, from https://www.env.nm.gov/air-quality/ 

5. Economic Analysis 

To determine the feasibility of our process, an extensive economic analysis was conducted. Our aim is to achieve a 90% reduction in the selling price of $47/g, with a production rate of 20kg/day over 323 operating days per year to account for holidays and maintenance. This section will fully detail our analysis and conclude with the plausibility and profitability of this price reduction.  

Capital Cost Summary:  

Our economic analysis will utilize Turton’s CAPCOST program, designed to estimate capital costs, provide cash flows, and risk analyses through various equations. The reported values and graphs in this section are derived from CAPCOST. Preliminary equations utilize Hurton et al., Analysis and Design of Chemical Processes (1) and material/equipment costs are derived from supplier quotes or online market sources. 

The following details how the program calculates the Fixed Capital Investment (FCI) of our process with user input. We first had to determine the tax rate for this process. Federal tax rate is 21% combined with New Mexico’s rate of 6%, therefore, we determined our process to have a taxation rate of 27%. We then calculated the annual interest rate using equation 45:

where ic is the annual interest rate and DR is the debt ratio, which was calculated using the monthly fees of the company along with the annual income and was found to be 18%. iD is the interest rate due on debt, which was found to be 0.2, and 𝑖e is the cost of equity which was found to be around 0.08. The annual interest rate is calculated as 4%.

Since CAPCOST was developed in 2017, its prices need to be adjusted to the present time using CEPCI values. The equation we used to adjust any prices is below in equation 46: 

Where C1 is the past cost, C2 is the present cost, CEPCI2024 is 796.8, and CEPCI2017 is 567.5.

Since our process includes unit operations that are not listed in CAPCOST, quotes were retrieved from manufacturers for the furnaces, the ball mill, filter, and the mixer. All the other unit operations were added using the built-in equipment already present in the program. Table 17 below summarizes the cost of equipment used for the following calculations. 

Because we are moving into an existing space, we determined that the use of the Lang Factor technique should be employed. This technique considers an expansion to an existing chemical plant. It considers aspects such as equipment installation and major building renovations needed for plant startup. This accounts for Pajarito Powder’s existing facility which can be retrofitted and modified with piping, electricity, and waste storage for our process.  

Our process is dominated by solid processing which requires a Lang factor of 3.1 as stated in literature1. The FCIL is then calculated using equation 47 below:

Where FLang is the Lang factor, and Cp represents the price of equipment, which is summed together. Using this equation, our fixed capital investment was calculated to be $1.9 million.   

Cost of Manufacturing: 

Next, the Cost of Manufacturing (COMD) is calculated using equation 48 below:

It considers FCIL, cost of operating labor (COL), the cost of utilities (CUT), cost of water treatment (CWT), and the cost of raw materials (CRM).   

The cost of operating labor was estimated using the amount of equipment, and assuming an average technician salary of $70,000 per year, coming out to $1.1 million a year. The cost of utilities was calculated using energy input for each unit operation and was calculated to be $122,000 per year. A breakdown of these utilities can be found under Appendix B. The price of raw materials was determined using current market prices, with flow rates being calculated using our batch stream table. CAPCOST assumes continuous processes, so an hourly rate was found using daily rates and the operating days of the year. The full material table is shown below in Table 18. This gives us a yearly cost of raw materials of $5.4 million.

Considering all these costs, our yearly cost of manufacturing was determined to be $10.1 million. 

Revenue 

Revenue is the measure of how much money is generated from the sale of our product. As was discussed earlier, we are assuming that the plant produces 20 kg/day, 323 days out of the year. With our reduced selling price of $4.7/g, our yearly revenue is a staggering $28.7 million. To put this in perspective, Figure 18 below shows our annual revenue compared to our annual costs including the cost of raw products, waste, operating labor, and utilities.  

This shows that even with a 90% cost reduction, our annual costs are minimal compared to the revenue. To further analyze the profitability of our process, a discounted cash flow diagram was constructed, shown below in Figure 19. This discounted cash flow diagram considers the time value of money using of 4% annual interest rate, along with taxation using a 10 year-MACRS depreciation. This specific analysis was done assuming a 2-year plant start up, with the capital investment being spread across this start up, and a plant life of 10 years.

Our profit is linear after the startup stage. The data shown in the cash flow diagram is shown in the Table 18. below.

The NPV Monte Carlo shows that our process is promising and has low risk to investors with a low NPV of $54.9 M. There is a 0% chance of failure (considering successful process startup), along with a 50% probability of exceeding $80M. Furthermore, analysis was done with the discounted payback period (DPBP). 

The DPBP Monte Carlo shows that within 0.1-0.3 years of the process being completed, we can expect a profit. Finally, rate of return on investment (ROROI) Monte Carlo simulation was performed as shown in Figure 5. The analysis also shows a very large rate of return on the investment, with the lowest being 394% and the highest being 875%.  

This risk analysis demonstrates the near guarantee of profit for our investors, which is incredible given the high rewards provided by our process. 

Conclusion and Recommendations 

Cash flow analysis and Monte Carlo simulation show our process with a 90% price reduction is highly profitable, with expected profit to be up to $100 million, and minimal risk to investors with a 0% chance of failure as well as an absence of debt. This process provides immense potential to investors and is one of the few opportunities where large gains come with low risk.  

Considering the broad and ever-growing market for PCS, our product is comparatively much cheaper, and the high profitability suggests further price reduction can be accomplished. This will enable high accessibility and potentially revolutionize the previously discussed technologies. However, we do not expect this level of profit margin to be sustainable. We projected the current selling price to be viable for only a few years after startup, after that, the saturation of the market as well as the rise in competition will make the current selling price unviable.  

While the hydrogen/catalyst market is still relatively young, there are already many companies that we consider to be potential competitors which include but are not limited to MAXIMATOR Hydrogen GmbH, one of the world’s leading suppliers and developers of hydrogen technologies based in Nordhausen8. SEGULA Technologies is a German-based company with global operations in over 30 countries and 140 branches, specializing in fuel cell development, integration, and infrastructure8. Finally, Fuel Cell Systems Ltd. is an expanding company that focuses on designing and manufacturing fuel cell systems, so far, they have successfully integrated more than 2500 fuel cell systems for their customers8.  

We expect this list of potential competitors to only grow even more in the next five years, further decreasing the value of hydrogen fuel cell technology-related products, including our PCS. Due to this, we are looking into further decreasing the selling price of our PCS while still retaining economic viability. By optimizing around the rate of return to be at a minimum of 12% (current reports show that the overall rate of return of the hydrogen fuel cell market to be 8%, making our 12% assumption to be still viable), we found that a price reduction up to 96% (1.75 $/g) can still be extremely profitable while also being sustainable in a still growing market.  

Sources: 

(1) Turton, R.; Shaeiwitz, J. A.; Bhattacharyya, D.; Whiting, W. B. Analysis, Synthesis and Design of Chemical Processes, 5th ed.; Pearson: Upper Saddle River, NJ, 2022 

(2) LUDOX TMA colloidal silica 34wt. suspension water 7631-86-9. Sigmaaldrich.com. https://www.sigmaaldrich.com/US/en/product/aldrich/420859?utm_source=google&utm_medium=organicshopping&srsltid=AfmBOoqGqd02StRSHNkeWQAxqA-gd9U9XhCdWyZYgx2LtPf6ltTqIOCsOX4 (accessed 2024-04-29) 

(3) Hydrofluoric Acid 50% solution, technical/Lab Grade. Lab Alley. https://www.laballey.com/products/hydrofluoric-acid-50-lab?variant=39804337619099 (accessed 2024-04-29) 

(4) Acetic Acid Glacial 99% Lab Grade. Lab Alley. https://www.laballey.com/products/acetic-acid-glacial-lab?variant=39822211842203&currency=USD&utm_medium=product_sync&utm_source=google&utm_content=sag_organic&utm_campaign=sag_organic&srsltid=AfmBOooMYxIeYah7lPVcuvfkZQEK_hkga_uA5uN8_nbQopMIxUj_nnqF1a4 (accessed 2024-04-29) 

(5) Nitric acid 40% – alliance chemical. Alliancechemical.com. https://alliancechemical.com/product/nitric-acid-40/?attribute_pa_size=55-gallon&attribute_pa_packaging-type=drum&gad_source=1&gclid=CjwKCAjwte-vBhBFEiwAQSv_xaUH-a2tfiHQEoBE0NjMCA2eBVhBodEUqpv4vq-8pgxBXKRjvUB44hoCSdIQAvD_BwE (accessed 2024-04-29) 

(6) Hydrogen Calibration Gas H2 95 PPM Balance Nitrogen in a 305 liter steel disposable cylinder. Cal Gas Direct Incorporated. https://www.calgasdirect.com/hydrogen-calibration-gas-h2-95-ppm-balance-nitrogen-in-a-304-liter-steel-disposable-cylinder/ (accessed 2024-04-29) 

(7) Chitosan Powder, 90+% (Deacetylated), 2.5kg. eBay. https://www.ebay.com/itm/254424918577?chn=ps&mkevt=1&mkcid=28&srsltid=AfmBOooQu5ElK80hdUJfCsYEIkHxvv18nJ2_6wysDum4jYNplqzARiCFF0Y (accessed 2024-04-29) 

(8) Hyfindr.com - the B2B marketplace for fuel cell components, hydrogen system and services. Hyfindr.com. https://hyfindr.com/en/ (accessed 2024-04-29)

6. Conclusion 

In this report, our objective was to develop a large-scale process to produce 20 kg/day of Pajarito Powder’s engineered mesoporous carbon support. We have detailed the specifications for scaled-up equipment and implemented the following principles in our design:  

i) Implement sustainable alternatives 

Chitosan was carefully chosen as a starting material that is nitrogen-rich and highly capable as a porous carbon precursor as reported in literature. Most importantly, chitosan is derived from biomass and biomass waste providing a renewable alternative to synthetic organics or fossil fuel derived hydrocarbons. Because our process is relatively small in comparison to large scale chemical plants, we have opted to use electrically powered unit operations which improves their respective efficiencies and eliminates emissions from burning fossil fuels.  

ii) Ensure the health and safety of employees 

An increase in material amounts results in increased risk that arise from chemical and physical hazards. In regards to chemical hazards, dust hazards associated with dry fumed silica are eliminated by substitution with colloidally suspended silica which is widely commercially available in various particle sizes enabling high tunability. A significant chemical hazard arises from the use of highly corrosive acids, HNO3 and HF, required for etching our material. Our design features a fully automated system that reduces exposure, eliminates manual acid handling, and alerts employees of potential hazards as oppose to Pajarito Powder’s current process. Additionally, the room is well ventilated to remove any exhaust gasses or loose powder. The majority of our process still requires manual handling of kg scale amounts which can pose physical strain and potential injury resulting from dropped materials. By implementing conveyers and rolling carts between unit operations the potential for this hazard can be significantly decreased. 

iii) Minimize negative environmental effects 

As mentioned previously, emissions are significantly reduced by utilizing electrically powered equipment. Most environmental hazards arise primarily from fumes and waste generated from the etching process. Our scrubber system has been designed to minimize the release of toxic gasses and waste is properly stored in contained areas for later removal. To protect the nearby residential and commercial areas, hazardous chemicals are stored safely within the facility. Additionally, continuous monitoring of pollutants via sensors will ensure that we do not violate regulation and release toxins into the environment.  

iv) Maintain Product Specifications 

Unit operations were carefully designed to maintain operating conditions that can be easily adjusted to resemble lab scale conditions. This was done to enable Pajarito Powder’s ability to specifically tune material properties. Automation of the etching stage and sieve, implementation of conveyers and rollers will reduce any impurities introduced from manual handling. Unit operation material of constructions were also carefully chosen to prevent any contamination as detailed previously. Additionally, quality checks at various stages of the process is feasible particularly at post-pyrolysis and through a built in system that allows operators to deposit a small amount of material during the etching process. This ensures adequate removal of silica from the pores and provides information of pore characteristics. Unfortunately, analytical and characterization tools used in quality checks are time restricted and require the use of specialized devices that are difficult and expensive to implement in situ. It is therefore recommended that prior to full plant start-up, testing and optimization of operating conditions that yield in spec material is necessary to reduce mid-process quality checks and waste material. 

v) Improve Efficiency 

All unit operations were designed to result in improved efficiencies however, the process is still constrained by the duration of each individual process. Our Gantt chart fully details an efficient schedule maintaining daily operations that results in our desired 20 kg/day output. A significant limitation in our efforts to improve efficiency is the lack of specific information available regarding material properties between unit operations. Due to the novelty and absence of useful reported data on porous carbons, properties such as density, porosity, purity, reaction kinetics (pyrolysis, etching), diffusivity, and heat transfer between unit operations were largely estimated in this report. To further optimize the efficiency, potentially model the process, and facilitate simple quality checks, extensive characterization and experimentation of the material is needed. This includes answering questions experimentally such as the necessity of a hydrogen reducing environment in etching. This has only been predicted to facilitate and enhance pore structure however, as a result, silica is reduced to silicon which therefore requires nitric acid to etch. Another question is the effect on various ball mill speeds on the product specifications and the extent of mechanochemical reactions. 

Although further experimentation needed, our design is highly equipped for changes in operating conditions and economic analysis has shown that, at the rate our product is being sold, this process is highly profitable. Further price reduction can potentially be done to facilitate high accessibility in the aforementioned industries and markets. Emissions reduction technologies that utilize PCS can be implemented at a global scale making PCS a key component in efforts to fight the negative effects climate change.