LEE- Kimya Mühendisliği-Yüksek Lisans
Bu koleksiyon için kalıcı URI
Gözat
Son Başvurular
1 - 5 / 51
-
ÖgeLityum-iyon piller için yüksek performanslı anot malzemesi olarak kalay oksit-indirgenmiş grafen oksit kompozitlerinin geliştirilmesi ve karakterizasyonu(Lisansüstü Eğitim Enstitüsü, 2025-07-09)Gelişen teknolojiler ve bu paralelde ivmelenerek artan enerji ihtiyacı, enerji depolama sistemlerinin verimliliği ve uzun ömürlü kullanım gereksinimini her geçen gün daha kritik bir noktaya taşımaktadır. Taşınabilir elektronik cihazlardan elektrikli taşıtlara, yenilenebilir enerji sistemlerinden çeşitli endüstriyel uygulamalara kadar uzanan alanlarda, yüksek enerji potansiyeline sahip, hafif, düşük maliyetli ve dayanıklı yapıların kullanımı ön plana çıkmaktadır. Lityum-iyon piller, sahip oldukları yüksek enerji yoğunluğu, geniş çalışma gerilimi aralığı, uzun döngü ömrü benzeri önemli avantajları nedeniyle enerji depolama alanında en sık tercih edilen sistemlerden biri haline gelmiştir. Bu avantajlarına rağmen, günlük yaşamdaki kullanımlarında performanslarında geliştirmeye açık alanlar bulunan lityum-iyon pillerde elektrot malzemelerinde, özellikle de anot yapısında iyileştirici düzenlemelere gidilmesi bir zorunluluk haline gelmiştir. Günümüzde ticari pillerde kullanılan grafit bazlı anot malzemeleri artan enerji ihtiyacı karşısında 372 mAh/g teorik kapasitesi ile yetersiz kaldığından, daha yüksek kapasite sunabilen, döngü dayanımı yüksek ve yapısal olarak kararlı yeni nesil anot malzemelerinin geliştirilmesine yönelik çalışmalar yoğunluk kazanılmıştır. Bu bağlamda, grafit malzemesinin yaklaşık iki katı teorik kapasite sunan (782 mAh/g) kalay oksit (SnO₂) önemli bir alternatif olarak göze çarpmaktadır. Ne var ki, kalay oksit sahip olduğu yüksek kapasite potansiyeline rağmen, döngüler sırasında meydana gelen hacimsel genleşme sebebiyle yapısal bozulmaya uğrayabilmekte ve bu da zamanla elektrotun aktifliğini kaybetmesine yol açmaktadır. Bu bilgiler doğrultusunda, literatürde kalay oksit anotların yapısal kararsızlığı ve hacim değişimi gibi dezavantajlarının karbon bazlı malzemelerle kompozit yapılar oluşturularak minimize edilebileceği ifade edilmektedir. Bu yaklaşım, kalay oksitin yüksek teorik kapasitesi ile grafenin üstün iletkenlik ve mekanik dayanım özelliklerinin kombinasyonunun, pil performansını olumlu yönde etkileyebileceği yönünde bir varsayım ortaya koymaktadır. Bu tez çalışması kapsamında, farklı oranlarda kalay oksit ve indirgenmiş grafen oksit (rGO) kompozit yapıları hidrotermal yöntem ile sentezlenmiş ve sentez başarıları XRD, SEM ve TGA karakterizasyon teknikleriyle doğrulanmıştır. XRD analizleri ile kompozit yapının kristal ve amorf özellikleri tespit edilmiş; SEM görüntülemelerinde rGO levhalarının SnO₂ partikülleri ile homojen biçimde dağıldığı gözlemlenmiştir. Ek olarak yapılan TGA analizleri kompozitlerdeki karbon içeriğinin nicel olarak belirlenmesini sağlayarak sentez başarısının ölçülmesine katkıda bulunmuştur. xxiKalay ve grafen oranının pil performansı üzerindeki etkisinin gözlemlenmesi amacıyla numunelerin elektrokimyasal performans analizleri, galvanostatik şarj-deşarj testleri, elektrokimyasal empedans spektroskopisi (EIS) ve çevrimsel voltametri (CV) araçları ile gerçekleştirilmiştir. Gerçekleştirilen elektrokimyasal analizler sonucunda, indirgenmiş grafen oksit içeriğinin artırıldığı kompozitlerde aktif kalay oksit miktarının azalmasına bağlı olarak pil kapasitesinin azaldığı belirlenirken; düşük indirgenmiş grafen oksit içerikli kompozitlerde hacimsel genleşmenin etkin bir biçimde tamponlanamaması nedeniyle yapısal bozulmaların arttığı tespit edilmiştir. Farklı oranlarda hazırlanan numunelere ait çevrimsel voltametri (CV) ve elektrokimyasal empedans spektroskopisi (EIS) verileri, şarj-deşarj test sonuçlarıyla tutarlı bir şekilde değerlendirilmiş; bu kapsamda, SnO₂/rGO = 1/2 oranına sahip kompozitin ilk döngüde en yüksek kapasiteyi sergilediği, SnO₂/rGO = 1 oranlı kompozitin ise 40 döngü sonunda en düşük kapasite kaybına sahip kompozit olarak öne çıktığı belirlenmiştir. Ek olarak, bu tez çalışması kapsamında ilk döngüde kaybedilen lityum iyonlarını telafi edebilmek amacıyla kompozit numunelere ön lityumlama işlemi uygulanmış ve bu işlemin elektrokimyasal performans üzerindeki etkileri incelenmiştir. Elde edilen bulgular, ön lityumlama uygulanmış numunelerin, aynı bileşim oranına sahip özdeş numunelere kıyasla daha yüksek performans sergilediğini ortaya koymuştur. Bu çerçevede, SnO₂/rGO = 1 oranına sahip kompozit numune, ön lityumlama sonrasında hem ilk döngüde hem de 40 döngü sonunda ulaştığı kapasite değerleriyle bu etkinin en belirgin şekilde gözlemlendiği örnek olmuştur. Elde edilen sonuçlar, ön lityumlama işleminin, pil döngüsünün başlangıcında oluşan katı elektrolit ara fazının (SEI) daha dengeli ve homojen bir şekilde oluşumunu destekleyerek kapasite kaybını azalttığını ve genel pil performansını olumlu yönde etkilediğini ortaya koymuştur. Elde edilen tüm bu bulgular ışığında, SnO₂/rGO kompozitlerinin uygun kalay-grafen oranlarda sentezlenmesi ve ön-lityumlama işlemiyle elektrokimyasal olarak desteklenmesi sonucunda, bu yapıların gelecekteki çalışmalar için yüksek potansiyele sahip enerji depolama sistemleri olabilecekleri tespit edilmiştir.
-
ÖgeMachine learning-based prediction of FTIR spectral peaks for biomass characterization(ITU Graduate School, 2025)This thesis explores how machine learning (ML) can be integrated with Fourier Transform Infrared (FTIR) spectroscopy to rapidly characterize lignocellulosic biomass. Traditional wet-chemical methods (e.g., Soxhlet extraction, Klason lignin assay) are accurate but time-consuming, motivating the use of FTIR as a faster, non-destructive tool. By capturing specific transmittance dips (absorbance peaks) tied to functional groups (e.g., O–H, C=O, aromatic rings), FTIR provides a detailed "fingerprint" of biomass components such as cellulose, hemicellulose, lignin, and extractives. This research aims to develop machine learning models that transform FTIR spectra into meaningful compositional or structural information. Three phases of investigation are designed for increasingly focused prediction targets: •Phase 1 (Full Spectrum): A multi-output regression approach predicts every wavenumber's intensity (3551 points) from nine input features (including biomass category, moisture, ash, volatile matter, holocellulose, lignin, etc.). We compare algorithms such as Partial Least Squares (PLS), Ridge Regression, Random Forest, and a Multi-Layer Perceptron (MLP). •Phase 2 (Broad-Range Classification): Instead of predicting each intensity, we classify whether a strong absorbance peak appears within broad wavenumber intervals (e.g., 3700–3000 cm⁻¹, 1800–1500 cm⁻¹). We employ multi-label classification (Logistic Regression, Random Forest, Gradient Boosting, SVM) to determine "peak present/absent" in each interval. •Phase 3 (Targeted Narrow Ranges): We zoom in on crucial intervals (like 3000–2800 cm⁻¹, 1800–1500 cm⁻¹, 1150–900 cm⁻¹) that directly link to chemical properties (e.g., lignin's aromatic ring signals, carbohydrate fingerprints). Classification models identify specific dips within these smaller spectral windows. Key Findings •Full-spectrum regression (Phase 1) is challenging, yielding low R² values (~0.04–0.21). MLP performed best overall for this high-dimensional task. •Broad-range classification (Phase 2) achieved higher accuracies (Hamming accuracy up to ~0.75) since "peak vs. no peak" is less complex than predicting all intensities. •Targeted intervals (Phase 3) gave the most robust classification (Hamming accuracy up to ~0.81) and better interpretability, as each narrow band strongly correlates with a known chemical feature. Conclusion & Implications By tailoring the ML approach to the desired level of spectral detail—ranging from full-spectrum regression to broad or narrow-interval classification—this thesis demonstrates that FTIR combined with data-driven modeling can effectively screen and characterize biomass. The findings support the hypothesis that simplified or chemically targeted outputs (Phase 2 and 3) can outperform fully detailed predictions (Phase 1). Ultimately, ML-enhanced FTIR offers a rapid, cost-saving alternative to classical assays, facilitating large-scale, real-time biomass characterization for bioenergy and bioproduct applications.
-
ÖgeEffect of operating pressure on design and control of extractive distillation process separating DMC-MeOH azeotropic mixture(Graduate School, 2022)Number of industrial facilities increase rapidly which leads to rise in their negative impact on their environmental destruction. In order to reduce this negative impact, there is a noteworthy increase in the usage of environmentally friendly raw materials and chemical processes. Dimethyl carbonate (DMC), due to its favorable properties such as having low ecotoxicity and being biodegradable, stands out as an environmentally friendly "green chemical". Since DMC has no harmful impact on environment, it is commonly used as substitute material of dimethyl sulphate and phosgene in methylation and carbonylation reactions. In addition to this, it is used as a co-solvent for non-aqueous electrolytes for lithium rechargeable batteries and adequate side material for internal combustion engine fuels. Transesterification of propylene carbonate and methanol is a preferable path for DMC synthesis. However, reaction with excess methanol (MeOH) being fed to the system results in azeotropic mixtures with DMC which eventually leads to difficulties in separation of DMC-MeOH mixture. Extractive distillation which is one of the most preferred methods for separation of azeotropic mixtures is known as an expensive process due to requirement of regeneration of extractive agent. Nevertheless, since DMC-MeOH azeotropic mixture is sensitive to pressure changes, a reduction in extractive agent requirement is considered achievable by operating the extractive column at higher pressures. For that reason, it is aimed to design an increased-pressure extractive distillation process which provides DMC with 99.8% purity and MeOH with 99.99% purity. The proposed process consists of two columns. The purpose of extractive distillation column is to separate methanol from azeotropic mixture using methyl isobutyl ketone (MIBK) as an extractive agent, while the recovery column is used to purify DMC and recycle the regenerated extractive agent MIBK back to the extractive column. The thesis consists of two stages. In the first stage, it is aimed to simulate alternate DMC-MeOH separation process configurations using Aspen Plus where the extractive distillation column operates at different pressures such as 1 bar, 5 bar, 7.5 bar and 10 bar. All increased-pressure extractive distillation process options and the base case where both columns operate at atmospheric pressure are optimized based on the total annual cost (TAC). As per the simulation results, significant amount of reduction in entrainer requirement is observed by increasing the operating pressure of extractive distillation column. By operating extractive distillation column at 10 bar pressure, a 34.1% decrease in total annual cost and 29.8% reduction in carbon dioxide emissions are observed compared to the base case where extractive distillation column operates at atmospheric pressure. Among the design studies of 1 bar, 5 bar, 7.5 bar and 10 bar, control structures are implemented for the case with 10 bar operating pressure which results in the lowest TAC and CO2 emissions. Prior to exporting steady-state design from Aspen Plus to Aspen Dynamics, equipment sizing for reflux drums and column sumps are completed. As steady-state simulation is exported to Aspen Dynamics, necessary controllers such as flow, level, pressure and temperature controllers are implemented to the system. ATV test is used to tune the temperature controllers. After that, seven different control structures which are distillate to reflux ratio control, feed to reflux ratio control, feed to reboiler duty ratio, feed to reflux & reboiler duty ratio controller, combined ratio control of distillate to reflux, feed to reflux and feed to reboiler duty, and feed to entrainer flow ratio are designed, and two different types of disturbances such as ± 20% change in the feed flowrate and ± 3% change in the feed composition are introduced to the system. Each of the simulations is conducted for 50 hours in which first two hours are operates in steady-state conditions. At the end of each simulation, generated data is exported to MATLAB to produce graphical results. According to outcomes of this study, it is seen that in case of change in the feed flowrate, solely feed to reflux ratio, combination of feed to reflux and feed to reboiler duty ratio and combination of distillate to reflux ratio control of extractive column, feed to reflux ratio control of recovery column and feed to reboiler duty ratio control in both of the column gives the best results for product purities by taking into account the lowered offset values and oscillations in addition to quicker response time. In more detailed observation, although there are very small differences between the control strategies yielding good results, combination of distillate to reflux ratio control of extractive column, feed to reflux ratio control of recovery column and feed to reboiler duty ratio control in both of the column gives better results with respect to other two control structures. Two results can be reached from this observation. First, addition of feed to reboiler duty ratio control improves response time of the process against flow rate disturbances although single application of feed to reboiler duty ratio alone does not provide satisfactory results. Secondly, although distillate to reflux ratio alone gives bad results, addition of feed to reflux ratio improves the control structure response to a reasonable level. On the other hand, each of the simulated scenarios except for the scenario seven have converged to its new steady state value in around 20-25 hours and stabilized for any kind of disturbances. It is aimed to control the entrainer make-up flow via a certain ratio from feed flow in 7th control structure, however, proper control of the system was not achieved. As the disturbanced are intoruduced to the system, it is not possible to conduct column operations decently.
-
ÖgeInvestigation of PD-CU alloying application in pd-based dense metallic membranes with h2 and sulfur interaction(Graduate School, 2024-08-28)Energy consumption is increasing day by day all around the world as a result of rise in human population, civilization and quality of social life. Almost 77% of energy is produced from fossil fuels which negatively impacts our environment. To prevent the hazards of fossil fuels, the governments, industries and scientific world are trying to find better alternatives which are sustainable. Hydrogen is a strong candidate as an sustainable energy carrier due to its renewability, high energy content and high efficiency. However, hydrogen cannot be found in pure state naturally. It is generally found in mixture with different gases such as N2, CO, CH4, and CO2. Therefore, to utilize hydrogen in energy production such as in fuel cells, it is crucial to separate and purify it from these gas mixtures. There are different commercially available techniques such as pressure swing adsorption (PSA) and cryogenic distillation which require intense energy. Therefore, membrane separation technology has received more attention than other processes since it consumes less energy. Advanced membrane technology is also cost-effective for extracting pure hydrogen from a stream with low purity or low pressure, compared to PSA or cryogenic technologies. The properties of membranes affect the hydrogen yield at the end of the process, so they must possess excellent chemical, thermal and mechanical stability, high hydrogen permeability, and be cost-effective to prepare, energy efficient, and durable. Dense metallic Pd membranes have demonstrated great potential to achieve ultra-high pure hydrogen via separation from gas mixtures. However, susceptibility of Pd to sulfur poisoning and expensive production cost are common drawbacks for these membranes. Therefore, to decrease the cost and to create a surface which is sulfur-tolerant, Pd can be combined to form an alloy with other metals. In this study, palladium-copper alloy surfaces on support were prepared using sequential electroless plating (ELP) method due to low cost and high sulfur resistance of these alloys. α-Al2O3 was used as support material. Prior to plating, α-Al2O3 supports were cleaned, dried and then activated with Pd seeds. Following the seeding, activated supports were plated first with Pd and then, with Cu on Pd plated layer. The total plating time of Pd was 3 hours which was followed by 30 minutes, 1 hour and 3 hours of Cu plating. In some experiments, the plating procedures were repeated several times. Different drying temperatures were applied between these sequential plating procedures in oven. At the end of the deposition procedures, the supports contains both Pd and Cu were annealed under stagnant gas. Annealing is crucial to create a well-mixed Pd-Cu alloy. It was observed that as Cu layer thickness increased, the total annealing time and temperature had to be increased to provide enough driving force and time for Pd atoms to diffuse into Cu layer to create a well-mixed alloy. Our experiment results show the correlation between the film thickness and total annealing time which is directly proportional to achieve well-mixed Pd-Cu alloy. The total annealing time was found as 96 hours to create an FCC Pd-Cu alloy at 600 °C for 3 hours Pd and maximum 1 hour Cu deposition. However, these time and temperature conditions are not sufficient for the samples were undergone several 3 hours depositions such as 3 times 3 hours Pd deposition then 3 hour Cu deposition. Besides sulfur resistance of Cu, metal organic frameworks are also promising for gas separation processes with their good properties to prevent the sulfur poisoning of surface. They have variety of applications and this is one of them. Therefore, as a final step, UiO-66 (Zr) was plated on the Pd-Cu surfaces by growing the frameworks on the membrane surface. UiO-66 was preferred for the studies because of its high working capacity, good selectivity, cost-effective regenerability, and ease of modification. Through experimentation, a correlation was observed between the amount of Cu and the MOF deposits on the surface: when Cu amount is higher, the surface has a better characteristic to coat MOF. The Pd-Cu samples and MOF coated Pd-Cu samples were subjected to sulfur environment at 250-300 °C for 3 hours. SEM-EDS analysis was employed to characterize the surface morphology of the samples and determine the composition. According to SEM-EDS results, the chosen samples were also characterized by XRD to validate alloy formation and monitor changes in crystal structure upon sulfur exposure.
-
ÖgeEnzymatically synthesized poly(δ-valerolactone) and poly(δ-valerolactone) nanohybrid usage in controlled drug delivery systems(Graduate School, 2025-01-23)There is an increasing interest for the biopolymers these days, because oppose to synthetically/artificially formed ones, they are generated by renewable resources. It is clear that with the increasing population and the times when we need to pay more attention to the world's depleting resources, why the people are tending to use biopolymers instead of synthetic polymers, which are petroleum-based. In other words, with the developing technology, oppose to fossil resources, which are very traditional, people started to care to eliminate their carbon footprint. In order to make this valuable approach happen, they beginned to use natural resources. Due to their better biodegradability, reproducable, biocompatibility and chemical degradability, they indicate non-toxicity and they have low-weight. These qualities lead them to be used in not only packaging, textile, medical, agricultural industries, but also wearable devices, sensors and mostly in drug delivery systems. Due to their many benefits, including biocompatibility, nontoxicity, thermoplasticity, semi-crystalline and hydrophobic form, flexibility, along with controlled decomposition, PVL is an intriguing option and frequently utilized in biomedical implementations. PVL is a fundamental component in the production of anthracycline type of antibiotics and the controlled medications (like doxorubicin (DOX)), which are prescribed to avoid different types of cancer. Nanohybrid systems are composite substances constructed up of two or more different nanoscale constituents that are closely joined or blended at the molecular or atomic scale. When these elements are combined, a fresh substance with specific qualities that may be superior to those of the constituent parts individually can be produced. The resultant nanohybrids frequently show mutually beneficial impacts, which enhance efficiency across a range of applications. Having particle sizes varying from one to thousand µm in diameter, microspheres are a form of particle dispersion mechanism, where molecules of drug are incorporated in a polymer matrix by adsorption or dispersion. Numerous medical compounds, including peptides, proteins, nucleic acids and molecules of small size medicine could be encapsulated in microspheres and have the relevant therapeutic properties. The microspheres have garnered a lot of interest lately upon account of their tiny size of particles, unique surface features, and substantial surface-to-volume ratio. Chalcones are polyphenolic chemicals of various structures found in plants. There are several possible uses for chalcone-based compounds, including antibacterial, antiviral, antioxidant, anti-inflammatory, antidiabetic, antiulcer, anticancer properties. The remarkable chemical structure and intriguing pharmalogical properties of trans-chalcone (TC), also known as 1,3-diphenyl-2-propen-1-one, make it a prominent focus of scientific investigation alongside other chalcones. In this study, physical adsorption was employed in the first part to immobilize lipase on surface-modified rice husk ash (RHA) to be used in enzymatic ring opening polymerization. The silica based material underwent surface modification when amine groups were added to its surface by utilizing 3-aminopropyltriethoxysilane (3-APTES). Then, pysical adsorption methodology was used to immobilize the free form of Candida antarctica lipase B. By using the monomer ratios from prior study, poly(δ-valerolactone) and poly(δ-valerolactone) nanohybrid were generated by enzymatic ring opening polymerization. The reactions are conducted at 80oC for 24 hours to get the highest molecular mass for both polymer and nanohybrid. These resultant materials were ultimately found to be applied in the fabrication of microspheres. As following, the methodology of O/W emulsion was employed to fabricate TC-loaded PVL microspheres, TC-loaded PVL nanohybrid microspheres, drug-free PVL microspheres and drug free PVL nanohybrid microspheres. Afterwards, drug release mechanism of these fabricated microspheres were investigated. For this investigation, with three different PVA concentrations (0.1, 0.5, 1 (w/v)% ), there different TC:PVL ratio (10, 20, 40%) and TC:PVL nanohybrid ratio (10, 20, 40%) have been evaluated in order to ascertain greatest encapsulation efficiency, and thereby drug release profile. To define the thermal, chemical and mechanical properties, TGA, DSC, SEM and XRD were applied to these microspheres. The greatest encapsulation efficiencies were observed with specimen A-2 (1% PVA and 20% ratio of TC:PVL) as 98.6 ± 8.4 (%) and specimen C-2 (1% PVA and 20% ratio of TC:PVL nanohybrid) as 82.7 ± 6.4 (%). This research, also demonstrates that drug release mechanisms are not affected majorly by the pH-changes. The DSC analysis shows that addition of TC into microspheres has lowered the melting points but there were no any extra TC-peak on the curves, so it is concluded that TC is successfully encapsulated and molecularly dispersed into PVL and PVL nanohybrid microspheres. TGA analyses were used to compare the thermal degradation pattern of drug-free and TC-loaded PVL microspheres with PVL. Additionally, TGA analyses were used to compare the thermal degradation pattern of drug-free and TC-loaded PVL nanohybrid microspheres with PVL nanohybrid. Decompositions and weight losses have been evaluated and interpreted based on the scientific literature and the analyses that were conducted. The chemical groups that indicate the existence of TC, PVL, TC-loaded PVL microspheres, drug-free PVL microspheres, PVL nanohybrid, TC-loaded PVL nanohybrid microspheres, and drug-free PVL nanohybrid microspheres were all observed by using FT-IR as characterization technique. Since FT-IR spectras are alike, it shows that TC was determined to be encapsulated within the microspheres. XRD analyses were used besides to several other characterizations to investigate the impact of TC loading on the crystalline structures and the crystallinity of microspheres. The SEM images reveal that every single one of the microsphere formulations had spherical shape. To observe the drug release behaviour of the microspheres fabricated in various surroundings, pH-dependent drug release investigations were conducted in the present research by employing two pH levels that were 5.6 and 7.4. The total cumulative release of TC was enhanced by the PVL microsphere formulations, reaching 42.1% in pH 7.4 ambient and 43.5% in pH 5.6 ambient. Moreover, the total cumulative release of TC was enhanced by the PVL nanohybrid microsphere formulations, reaching 39.4% in pH 7.4 ambient and 34.2% in pH 5.6 ambient. The TC release was conducted for an overall duration of 744 hours in each case. Finally, the release kinetics were examined. The release was found to be compatible with the kinetic model of Korsmeyer-Peppas. All microsphere formulations had TC release that was controlled by Fickian diffusion, according to the calculated n value. In conclusion, the goal of this study is to develop new controlled drug delivery systems by integrating a transchalcone into biological in origin polymeric carriers. A biocompatible, environmentally friendly, high-molecular-weight polymer and nanohybrid containing home-made immobilized enzymes that are compatible for human body will be produced in order to fabricate microspheres. Transchalcone will be incorporated with the polymer and nanohybrid generated by biocatalyst for use in therapy. Following the completion of all characterization tests and investigation of drug release profiles, it can be clearly said that this research's findings suggest that PVL microspheres or PVL nanohybrid microspheres may find implementation in prolonged therapeutic applications.